This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!
Aln, ( p - 1-5 pm) As is clear from Figure 15 in this region too, both the normal modes of elliptically polarized light get absorbed by the dyes, although q,d, in this case, is smaller compared than that in the region p IAln,. However, due to moderate operating voltages, this is the region in which phase change displays are operated. The cells are usually prepared with either ho-
286
3.4 Guest-Host Effect
mogeneous or homeotropic alignment, with a cell thickness of usually more than 4-5 times the pitch of the dichroic mixture (-2-5 pm). Sometimes hybrid alignment (homeotropic on one plate and homogeneous on the other) is given. Irrespective of the alignment, the central layers adopt the planar Grandjean arrangement, with the helix axis perpendicular to the glass surfaces.
3.4.11.1 Threshold Characteristic and Operating Voltage The threshold characteristic of a phase change dichroic display is shown in Fig. 16. With slowly increasing voltage, the cell seems to go from Grandjean to a scattering focal conic and from there to a homeotropic nematic texture. On reducing the voltage the same effect is observed in reverse order. The first major jump in the threshold characteristic corresponds to the transition from Grandjean or planar to fingerprint or focal conic texture, and the second is for focal conic to nematic texture. The critical unwinding voltage V, for the cholesteric -ne-
matic phase transition can be calculated as [103]:
This equation was derived for an infinitely thick cholesteric layer where there is no torque exerted by the boundaries and cholesteric material has its natural pitch. However, it holds reasonably well for practical devices with large d/p ratios, and in particular for White-Taylor displays with homeotropic boundaries having no forced pitch. The threshold characteristic shows a noticeable hysteresis. The winding voltage V, is given by (see Fig. 16)
where k22and k,, are the twist and bend elastic constants, respectively. V, is found to With homeotropic be smaller than V,.
Figure 16. Threshold characteristic of a phase change dichroic display: (--) homogeneous alignment; (- - - - ) homeotropic alignment. Voltage (V)
287
3.4.1 1 Phase Change Effect Dichroic LCDs
alignment the threshold characteristic shifts towards lower voltage. Homeotropic alignment generates lower threshold and operating voltages (see Fig. 16). Eq. (17) also shows that threshold and operating voltages are directly proportional to the cell gap and inversely proportional to the pitch of the mixture and the square root of A&.Thus, to reduce the operating voltage, the cell thickness must be reduced and the pitch and A& increased. However, increasing the pitch reduces the contrast. To obtain the appropriate contrast and operating voltage we recommended that dlp be kept at about 4-6 for directly driven applications, especially for avionic use. With smaller d/p ratios (d/pO >O O
>O O O
Theoretical contrast ratioa) ~~
1 2 3 4 5 6
Vth) and the use of polarized light. In planar cells the E vector is parallel to n. In homeotropic cells the E vector agrees with n after the dielectric reorientation. b, The director n is tilted by a small angle with respect to the layer normal.
3.4.13 Positive Mode Dichroic LCDs
3.4.13.2 Positive Mode Dichroic LCDs Using a i1/4 Plate
W4 displays do not need polarizers. One can get a positive mode A/4 display using the same two-cell geometries and materials as discussed in an earlier section.
Positive Mode Double Cell Dichroic LCD
3.4.13.3
The same geometry and materials can be used as discussed for single cell positive mode LCDs (see Sec. 3.4.13).
Positive Mode Dichroic LCDs Using Special Electrode Patterns
3.4.13.4
A positive mode dichroic LCD [16, 1271 can be demonstrated by using special electrode patterns even in the case of pleochroic dyes embedded in a liquid crystal of positive dielectric anisotropy. In this case the voltage is applied to the whole visible area and is removed only from the desired segmented areas. 3.4.13.5 Positive Mode Phase Change Dichroic LCDs
The molecular structure of a long-pitch cholesteric mixture is determined both by the cell gap and the type of alignment [ 13- 161. If the cell gap is sufficiently large, the structure is always helicoidal in the bulk of the layer, irrespective of the homeotropic or homogeneous alignment on the surface [ 13- 161. If the cell gap is below a critical thickness d,, where
the helix is unwound in the presence of a homeotropic alignment and the whole structure is homeotropic nematic [ 1201. Two versions of cells using this property have been proposed using a pleochroic dye
295
in a liquid crystal host of positive dielectric anisotropy [ 1281. The operational principle ofboth types of cell is that the area surrounding the picture element has homeotropic alignment and a cell gap less than d,, resulting in a nonabsorbing homeotropic nematic structure. In the first geometry the areas surrounding the pixels undergo homeotropic treatment, while the pixel areas receive homogeneous alignment [ 1291.The cell gap is less than d,. In the second geometry the alignment on the whole cell is homeotropic, but the cell gap is adjusted in such a way that the pixel areas have a cell gap more than d,, while the remaining area is less than d,. These proportions result in an absorbing helicoidal structure in the pixel areas and a nonabsorbing homeotropic one in the remaining area [ 1281. On application of the electric field, the helicoidal structure is destroyed and the liquid crystal molecules adopt a homeotropic structure, with the result that only the nonactivated pixels remain dark. The control logic for this display is inverted with respect to that required for the White-Taylor type with negative contrast. Another way of obtaining apositive mode phase change type display is to use a pleochroic cholesteric mixture of negative dielectric anisotropy [ 16, 1291. The surface treatment is homeotropic and cell thickness d is chosen to be less than d,. In the quiescent mode the cell is in the homeotropically aligned nematic state, and hence pleochroic dyes do not absorb the light. On aplication of the field, the molecules become parallel to the glass plates and also adopt the heliocoidal structure. In this situation the light is absorbed by the dyes. A display of this type has the advantages of positive contrast, brightness, wider viewing angle, lower operating voltage, and multiplexing capability [ 16, 1291, but has low contrast. A very high order parameter dichroic mixture with a very low optical anisotropy is need-
296
3.4 Guest-Host Effect
ed to obtain sufficient contrast for this display.
3.4.13.6 Dichroic LCDs Using an Admixture of Pleochroic and Negative Dichroic Dyes These dichroic LCDs have a dichroic mixture containing both pleochroic (L) and negative dichroic (T) dyes, these having absorption wavelengths in different parts of the spectrum [53, 1181. The display can be made in Heilmeier, TN, and phase change mode. If a liquid crystal mixture of positive dielectric anisotropy is used to make a dichroic mixture using these L and T dyes, and is filled in a cell having homogeneous alignment, both L and T dyes will absorb in the quiescent state, and the cell will have the complementary color of the dichroic mixture. This color can be changed by changing the ratio of L and T dyes. On application of the field, the dye molecules, together with the liquid crystal molecules, become aligned in the direction of the field. Only the T dye absorbs in this geometry, and its complementary color becomes the observed color. These displays do not require polarizers and can be operated at low voltage like TN displays. Their excellent brightness and contrast are much less dependent on the order parameter compared to other guest-host displays [ 1181. Schadt [ 1181 also derived the detailed analytical expressions describing the dependence of transmission and color effects on the dye order parameter, the direction of the dye’s transition moment, and the elastic and dielectric properties of the liquid crystal mixture. The viewing angle and threshold voltage have been calculated and compared with experimental values. Using tetrazine dye (T) and anthraquinone dye D27 (L), Schadt developed a display exhibiting bright-red characters on a bright-blue background [ 1181. This display
can also be operated with a single polarizer in the Heilmeier mode [ 1181.Pelzl et al. [54] have reported a similar display.
3.4.14 Supertwist Dichroic Effect Displays Supertwist dichroic effect (SDE) displays have a twist of approximately 3 x/2, instead of n/2 as in the case of the popular TN mode LCDs, and hence are called supertwisted LCDs 129-311, The 3n/2 twist imparts a very steep slope to the threshold characteristic of the cell. This means it has a lower value of AV/Vth, and hence can handle a higher level of multiplexing. However, the threshold characteristics of 3 n/2 displays, unaided by dye or the interference of ordinary and extraordinary rays, has low contrast. The dyes absorb the leakage and enhance the viewing angle and contrast. Two types of SDE display can be made, one by using a Heilmeier geometry (by using a single polarizer), and the second by using the phase change or White-Taylor mode. However, SDE displays did not become popular due to their lower contrast, lower brightness, and slower switching speed compared to other supertwist displays.
3.4.15 Ferroelectric Dichroic LCDs Surface stabilized mode ferroelectric LCDs [31-351 show bistability, i. e. both the 8and the -8 position are equally stable. A pulse of one polarity is used to switch from 8 to -8,while a pulse of opposite polarity is used to switch from -8 to 8. The molecules can remain in either of these states (8 or -8)for a long time without any applied voltage. In the case of GH ferroelectric LCDs, the ferroelectric mixture contains an appropriate
3.4.16 Polymer Dispersed Dichroic LCDs
amount of dichroic dyes. Ferroelectric dichroic LCDs have a very fast switching speed (-100 ps or less), wide viewing angle, and good contrast [35].
3.4.15.1 Devices Using A Single Polarizer For a dye dissolved in a ferroelectric host in a single polarizer device, the orientation is rotated through twice the tilt angle (28) of the guest - host material on changing from one switched state to the other. If the polarizer is aligned with the dye direction to generate maximum absorption (dark state), then the switched state ( @ = 2 8 )would be lighter. One can also make a display exhibiting dark characters on a light background by initially aligning the polarizers perpendicular to the long axes of the pleochroic dye molecules. For maximum contrast a tilt angle of 45" is required. Only a few ferroelectric liquid crystals have such a high tilt angle. Fortunately, a 10-90% change in intensity may be achieved with a tilt of only 27" [35], making commercially available ferroelectric host and dichroic mixtures, with a typical tilt angle of 22-25', useful for this application. Another single polarizer DGHFE mode LCD uses an anisotropic fluorescent dye in the ferroelectric mixture [35].
3.4.15.2 Devices Using No Polarizers
-
A ferroelectric liquid crystal with 8 22.5" and a high pleochroic content can be used entirely without a polarizer by using a 2 4 plate [32] with a reflector. In the off-state the incoming light is selectively absorbed along the homogeneous alignment n in the cell. A quarter wave plate is placed along or perpendicular to the director. The nonabsorbed radiation (- 50%)in the ideal case vibrates perpendicular to the dichroic director and is reflected back unaffected. In the onstate n is turned 45" to the axis of the A14 plate; the nonabsorbed orthogonal vibration
297
is thus split up and components retarded corresponding to a double pass 2 i1/4=A/2. The polarization plane is therefore rotated by 90" and the radiation is absorbed on its way back into the liquid crystal. Coles et al. [35] made another ferroelectric dichroic device without polarizer, using a double layer surface stabilized mode DGHFE. The optimum tilt angle required for this geometry is 22.5".
3.4.16 Polymer Dispersed Dichroic LCDs A polymer dispersed liquid crystal (PDLC) film [36-39, 130- 1321, with a pleochroic dye dissolved in it, possesses a controllable absorbance as well as a controllable scattering. This combination can be exploited to generate high-contrast displays. For polymer dispersed dichroic LCDs, Fergason's NCAP systems are found to be better than Doane's PDLC systems [36 - 391. Unlike to other dichroic LCDs where we strictly require dichroic dyes only, one can use isotropic as well as dichroic dyes in these kinds of devices. Vaz observed a larger than expected color change between the on and off states of dichroic PDLC shutters, even with dichroic dyes with small order parameters (0.2) and dichroic ratios (1.7) [37]. This is explained on the basis of increased path length (i.e. more than the film thickness) due to multiple scattering in the off mode, which consequently increases the off-state absorbance. The contrast ratio of the dyepolymer system can be improved significantly by incorporating dichroic dyes with high order parameters and extremely low solubilities in the polymer [36-39, 1311. These displays provide much higher brightness and wider viewing angles than TN displays. Two optical processes are at work in these displays: switchable absorbance due to dye alignment, and electrically controlled
298
3.4 Guest-Host Effect
light scattering. When the display is excited, the light entering the film experiences a low absorbance and minimal scattering. The color reflector in the back is seen clearly with high brightness and color purity. In the quiescent state, incoming light is both strongly absorbed and scattered. The strong absorption by the pleochroic dyes gives the film a dark appearance. The scattering effect enhances the absorption and causes a fraction of the light to be reflected before reaching the colored reflector. This ensures that the light reflected from the display in the unexcited state possesses a very low color purity and hence looks more neutral or black. The color difference properties of the film can be optimized by varying various parameters of the film, such as film thickness, dye concentration, dye order parameter, and the level of scattering. The scattering is controlled by the liquid crystal birefringence and the droplet size distribution.
3.4.17
Dichroic Polymer LCDs
It is now well established that polymer nematic liquid crystals can be aligned by an electric field and surface forces, in more or less the same way as low molecular weight nematic liquid crystals [ 133- 1351. To generate dichroic polymeric LCDs, either dichroic dyes are dissolved in nematic polymers, or side-chain dye moieties are attached in nematic copolymer structures [ 133- 1351.As the pleochroic dye has its absorption transition dipole defined relative to its chemical structure, rotation of the dye produces a change in absorption of the guest-host cell when observed in polarized light. Several types of effect have ben reported and many more are possible, analogous to the low molecular weight nematic counterpart (as discussed in Sec. 3.4.83.4.16). Some of these are: Heilmeier type,
using pleochroic dyes in liquid crystal polymer of positive dielectric anisotropy; guest-host cells, using a dichroic liquid crystal polymer in a low molecular weight liquid crystal host; and polymeric ferroelectric dichroic LCDs. However, their use in display devices is strongly restricted by their slow response time. The reason for this is that the viscosity of side-chain liquid crystals is some orders of magnitude greater (depending on the temperature) in relation to the glass transition temperature and the degree of the polymerization of the polymer. One of the most promising potential uses of liquid crystal polymers is in optical storage devices [133, 1351. The information is written using a He-Ne laser. Images are stable over a long period of time (at least 4 years under ambient conditions), and are erasable by applying an electric field. The absorption of the laser beam is facilitated by incorporating a dye into the polymer [ 1361. Ferroelectric polymeric materials are fascinating because, in principle, they can have the fast switching speed of ferroelectric materials together with the processibility of polymers. Dichroic properties can be achieved either by dissolving dichroic dyes in ferroelectric polymers or by inserting dichroic groups in the polymer. Using a little imagination one can form all types of dichroic ferroelectric LCDs using polymeric ferroelectric dichroic mixtures in place of ferroelectric dichroic liquid crystal mixtures.
3.4.18 LCDs
Smectic A Dichroic
In these displays usually a thin layer of SmA material (- 10- 15 pm) is sandwiched between I T 0 coated glass plates with homeotropic alignment. When the layer is heated above the isotropic temperature and
3.4.20
is cooled back to the SmA phase via a nematic phase, it generates: (1) a focal conic scattering texture in the absence of the field, and (2) a clear homeotropic texture if cooled in the presence of an electric field [137]. These properties of SmA phases have been used in making various kinds of laseraddressed [ 137, 1381 and thermal-addressed [40, 41, 1381 SmA displays. Incorporation of pleochroic dyes converts the scattering state into an absorbing state [40, 41, 137, 1381. The homeotropic state will remain clear in this case too. In spite of a lot of research and demonstrations, thermally and electrically addressed displays have not become a commercial product [40, 41, 1371. The main reasons for this are their high power consumption and limited operating temperature range. Moreover, the advent of supertwist displays closed the market sectors where they could have been utilized.
3.4.19 Fluorescence Dichroic LCDs Some attempts have been made to replace normal dichroic dyes with fluorescent type dichroic dyes [16, 59-61, 139- 1411. Fluorescence is a two-step process involving absorption at a given wavelength and reemission at a longer wavelength, and is associated with electronic transitions which, relative to dye-molecule geometry, are directional in space. Thus the change in the orientation of the fluorescent dye molecule changes the absorption and subsequent emission intensity. Depending on the polarization of the incident and transmitted light, sample thickness, and concentration, it is possible to produce an increase or decrease in the light level. These fluorescent dyes can be excited either by UV light or light of shorter wavelengths in the visible spectrum
References
299
(such as blue light for exciting green, and yellow for exciting red fluors). These fluors can be excited either directly or indirectly through a liquid crystal. For example, a green fluorescent dye can be excited directly by blue light or indirectly with ultraviolet light. In the latter case the UV energy is absorbed by an appropriately formulated liquid crystal host, and subsequently transferred to the fluorescent dye. The fluorescent LCDs combine the visual impact and good viewing angle of emissive displays with the desirable features of LCDs, such as long voltage operation and low power consumption. Both Heilmeier [ 1391 and phase change [61] type fluorescent dichroic displays have been reported. When stimulated from the rear, indirect excitation is preferred because the exciting light is invisible and the display then has a dark appearance on a bright colored background. The display can also be excited from the front. In the frontlit mode, the phase change version is preferred over the Heilmeier mode, as the polarizer blocks UV light and attenuates any blue light passing through it. Acknowledgments Thanks are due to Urmila, Shivendra, and Shachindra Bahadur for their untiring help in preparing this manuscript. This article is dedicated in loving memory to my aunt, Mrs Sushila Srivastava, who passed away recently after a brave fight against cancer.
3.4.20 References [ l ] G. H. Heilmeier, L. A. Zanoni, Appl. Phys. Lett. 1968, 13, 91. [2] G. H. Heilmeier, J. A. Castellano, L. A . Zanoni, Mol. Cryst. Liq. Cryst. 1969, 8, 293. [3] B. Bahadur, Mol. Cryst. Liq. Cryst. 1983, 99.
34s. [4] E. Sackmann, in Applications ($Liquid Crystu/s (Eds.: G. Meier, E. Sackmann, J. G. Grabmaier), Springer-Verlag, New York, 1975. [S] J. W. Emsley, J. C. Lindon, N M R Spectroscopy Using Liquid crystal Solvents, Pergamon Press, New York, 1975.
300
3.4 Guest-Host Effect
161 C. Schumann, in Handbook of Liquid Crystals (Eds.: H. Kelker and R. Hatz), Verlag Chemie, Weinheim, 1980. [7] C. L. Khetrapal, R. G. Weiss, A. C. Kunwar, Liquid Crystals- Applications and Uses, Vol. 2 (Ed.: B. Bahadur), World Scientific, Singapore, 1991. [8] A. Saupe, G. Engler, Phys. Rev. Lett. 1963, 11, 462. [9] A. Carrington, G. R. Luckhurst, Mol. Phys. 1964,8,401. [lo] D. L. Uhrich, Y. Y. Hsu, D. L. Fishel, J. M. Wilson, Mol. Cryst. Liq. Cryst. 1973, 20, 349. [ I13 W. J. Leigh, in Liquid Crystals - Applications and Uses, Vol. 2 (Ed.: B. Bahadur), World Scientific, Singapore, 1991. [12] T. J. Scheffer, J. Nehring, in Physics and Chemistry of Liquid Crystal Devices (Ed. J. G. Sprokel), Plenum Press, New York, 1979, p. 173. [13] T, J. Scheffer, Phil. Trans. R. SOC., London Ser. A, 1983,309, 189. [14] T. Uchida, M. Wada, Mol. Cryst. Liq. Cryst. 1981, 63, 19. [15] B. Bahadur, Mol. Cryst. Liq. Cryst. 1991, 209, 39. [16] B. Bahadur, in Liquid Crystals - Applications and Uses, Vol. 3 (Ed.: B. Bahadur), World Scientific, Singapore, 1992. 1171 D. L. White, G. N. Taylor, J. Appl. Phys. 1974, 45,4718. [I81 H. S. Cole, R. A. Kashnow, Appl. Phys. Lett. 1977,30,619;SID Digest, 1977,8,96; GE-CRD Report No. 77CRD059,1977. [19] B. Bahadur, SPIE, Opt. Test. Meteorol. II, 1988, 954, 574. 1201 B. Bahadur, K. Wan, C. Prince, L. Krzymien, J. Tunnoch, Mol. Cryst. Liq. Cryst. 1992,220,155. [21] T. Uchida, H. Seki, C. Shishido, M. Wada, SlD Digest, 1980, 11, 192. [22] H. Seki, T. Uchida, Y. Shibata, Proc. SID 1984, 25, 275. 1231 K. Sawada, Y. Masuda, Proc. SID 1982,23,241. [24] K. Sawada, S. Kudoh, F. Kadoo, Y. Endoh, Y. Nakagawa, Y. Masuda, SID Digest 1985,78. 1251 G. A. Dir, V.J. Hull, Proc. Int. Disp. Res. Con$, 1985, 83. [26] B. Bahadur, Development of a High Resolution (Phase II) Liquid Crystal Display System - Llquid Crystal Mixture and Process Development, LSL Report, April 1989. [27] T. J. Scheffer, J. Nehring, J. Appl. Phys. 1984, 908,56. [28] K. Wan, D. Lewis, B. Bahadur, SID Digest 1994, 25, 601. B. Bahadur, K. Wan, US Patent 5589965,1996. [29] C. M. Waters, E. P. Raynes, V. Brimmell, Mol. Cryst. Liq. Cryst. 1985, 123, 303. [30] C. M. Waters, V. Brimmell, E. P. Raynes, Proc. SID 1984, 25, 261, Jpn. Displ. 1983,398.
1311 B. Bahadur, Supertwisted Dye Effect, Supertwisted Birefringence and Ferroelectric Liquid Crystal Displays, Litton Data Images Report, March 1986. [32] S. T. Lagerwall, J. Wahl, N. A. Clark, Proc. Int. Displ. Res. Con$ 1985,213. [33] K. Kondo, S. Era, M. Isogai, A. Mukoh, Jpn. J. Appl. Phys. 1985,24, 1389. [34] H. J. Coles, H. F. Gleeson, Mol. Cryst. Liq. Cryst. Lett. 1988, 6, 53. 1351 H. J. Coles, H. F. Gleeson, J. S. Kang,Liq. Cryst. 1989, 5, 1243. 1361 P. S. Drziac, R. Wiley, J. McCoy, SPZE, Liq. Cryst. Chem. Phys. Appl. 1980,41. [37] J. W. Doane, in Liquid Crystals - Applications and Uses, Vol. 1 (Ed.: B. Bahadur), World Scientific, Singapore, 1990. [38] J. L. Fergason, SZD Digest 1985, 68; US Patent 4435047, March 6,1984. 1391 P. S. Drzaic, Liquid Crystal Dispersions, World Scientific, Singapore, 1995. 1401 S. Lu, D. H. Davies, Mol. Cryst. Liq. Cryst. 1983, 94, 167. 1411 W. Harter, S. Lu, S. Ho, B. Basler, W. Newman, C. Otaguro, SID Digest 1984, 15, 196. [42] B. Bahadur, R. K. Sarna, V. G. Bhide, Mol. Cryst. Liq. Cryst. 1981, 75, 121. 1431 H. Seki, T. Uchida, Y. Shibata, Mol. Cryst. Liq. Cryst. 1986, 138, 349. 1441 R. J. Cox, Mol. Cryst. Liq. Cryst. 1979,55, 1. [45] G. W. Gray, Chimica, 1980, 34, 47; Dyes Pigments 1982,3,203. 1461 H . V. Ivashchenko, V. G. Rumyantsev, Mol. Cryst. Liq. Cryst., 1987, 150A, 1. [47] M. G. Pellatt, I. H.C . Roe, J. Constant, Mol. Cryst. Liq. Cryst. 1980, 59, 299. [48] F. C. Saunders, K. J. Harrison, E. P. Raynes, D, J. Thompson, Proc. SID 1983,24, 159. [49] J. Cognard, T. H. Phan, Mol. Cryst. Liq. Cryst. 1981, 68, 207. [50] X. Q. Yang, J. R. Wang, G. D. Ren, Mol. Cryst. Liq. Cryst. 1991, 199, 1. 1511 S. Imazeki, M. Kaneko, T. Ozawa, A. Mukoh, Mol, Cryst. Liq. Cryst. 1988, 159, 219. [52] G. Pelzl, H. Schubert, H. Zaschke, D. Demus, Krist. Techn. 1979, 14, 817. [53] G. Pelzl, H. Zaschke, D, Demus, Displays 1985, 141. [54] Literature on Dichroic dyes, Mixtures and Hosts from BDH (England), E. Merck (Germany), Hoffman La Roche (Switzerland) Mitsui (JaPan). [55] T. J. Scheffer, J. Appl. Phys. 1982, 53, 257. 1561 B. S. Scheuble, G. Weber, L. Pohl, R. E. Jubb, SIR Digest 1983, 176; Proc. SID 1984, 25, 25. [57] G. Heppke, B. Knippenberg, A. Moller, G. Scherowsky, Mol. Cryst. Liq. Cryst. 1983, 94, 191.
3.4.20 References [58] U. Claussen, A. Brockes, E. F. Kops, F. W. Krock, R. Neeff, SlD Digest 1984, 209; Proc. SlD 1985,26, 17. [591 E. Wolarz, D. Bauman, Mol. Cryst. Liq. Cryst. 1991, 197, 1. [601 S. Sato, M. M. Labes, J. Appl. Phys. 1981, 52, 3941. R. W. Filas, M. M. Labes, 1981,52,3949. [611 M. R. Lewis, M. C. K. Wiltshire, A. Cuddy, J. Stanford, I. Sage, J. Griffiths, S. Tailor, Proc. SlD 1988, 29, 233; Eurodisplay 1987, 149. [62] S . Aftergut, H. S. Cole, Jr., Mol. Cryst. Liq. Cryst. 1982, 89, 37. [63] H. Seki, C. Shishido, T. Uchido, M. Wada, Mol. Cryst. Liq. Cryst. 1981, 66, 209. [64] S. Ohmi, A. Watanabe, K. Hara, Jpn. Displ. 1986, 282; Proc. SlD 1987, 28, 123. (651 H. Seki, T. Uchida, C. Shishido, Jpn. J . Appl. Phys. 1980,19, L501. [66] S . Aftergut, H. S. Cole, Mol. Cryst. Liq. C y s t . 1990, 188, 147; Proc. SID 1990, 31, 317; SlD Digest 1990, 21, 95. [67] F. Jones, T. J . Reeve, Mol. Cryst. Liq. Cryst. 1981, 78, 201. [68] J. K. Foitzik, W. Haase, Mol. Cryst. Liq. Cryst. 1987,142,55; 1987, 149,401. [69] F. Jones, F. A. Kirby, Mol. Cryst. Liq. Cryst. 1984,108, 165. [70] S. Imazeki, A. Mukoh, T. Yoneyama, M. Kaneko, Mol. Cryst. Liq. Cryst. 1987, 145, 79. [71] A. Bloom, P. L. K. Hung, Mol. Cryst. Liq. Cryst. 1978,44, 323. [72] S. Yano, S. Kato, M. Kano, Y. Kamijo, Mol. Cryst. Liq. Cryst. 1985, 123, 179. [73] T. K. Attwood, J. E. Lydon, F. Jones, Liq. Cryst. 1986, I , 499. [74] M. A. Osman, L. Pietronero, T. J. Scheffer, H. R. Zeller, J . Chem. Phys. 1981, 74, 5377. [75] B. Bahadur, in Liquid Crystal -Applications and Uses, Vol. 2 (Ed.: B. Bahadur), World Scientific, Singapore, 1991. [76] B. Bahadur, Mol. Cryst. Liq. Cryst. 1984,109, 1. [77] S. Morozumi, in Liquid Crystals -Applications and Uses, Vol. 1 (Ed.: B. Bahadurj, World Scientific, Singapore, 1990. [78] J. Cognard, Mol. Cryst. Liq. Cryst. 1981, 7 8 (Suppl 11, 1. [79] T. Uchida, H. Seki, in Liquid Crystals - Applications and Uses, Vol. 3 (Ed.: B. Bahadurj, World Scientific, Singapore, 1992. [80] J. M. Geary, J. W. Goodby, A. R. Kmetz, J. S . Patel, J. Appl. Phys. 1987, 62, 4100. [81] S. Matsumoto, M. Kawamoto, N. Kaneko,Appl. Phys. Lett. 1975,27, 268. [82] Homeotropic aligning agents from E. Merck, Germany (such as ZLI-2570), and Nissan Chemical Ind., Japan (such as RN 722), etc. [83] H. Seki, Y. Itoh, T. Uchida, Y. Masuda, Jpn. J. Appl. Phys. 1990, 29, L2236; Mol. Cryst. Liq. Cryst. 1991, 199, 151.
30 1
[84] Low tilt (for TN) and high tilt (for STN) producing polyimides from Hitachi (Japan), Japan Synthetic Rubber (Japan), Nissan Chemicals Inc. (Japan). [851 H. Fukuro, S . Kobayashi, Mol. Cryst. Liq. Cryst. 1988, 163, 157. [861 M. R. Lewis, M. C. K. Wiltshire, Appl. Phys. Lett. 1987, 51, 1197; R. W. Filas, SID Digest 1984,206. [87] B. 0. Myrvold, Mol. Cryst. Liq. Cryst. 1991, 202, 123. [881 C. H. Gooch, H. A. Tarry, Electron. Lett 1974, 10, 2; J. Phys. D 1975, 8 , 1585. [89] P. Kassubek, G. Meier, Mol. Cryst. Liq. Cryst. 1969, 8, 305. [90] A. J. Hughes, D. G. McDonnell, S. Hedges, Displays, April 1987, 69. [91] M. Kawachi, K. Kato, 0. Kogure, Jpn. J. Appl. Phys. 1978, 17, 1245. [92] Information sheets for polarizers from Polaroid Corp., USA, Nitto, Japan, Sanritsu, Japan. [93] R. W. G. Hunt, Measuring Colouv, Ellis Horwood, Chichester, 1987. [94] A. Bloom, E. B. Priestley, ZEEE Trans. Electron. Dev., 1977, ED24, 823. [95] Euro-Fighter Project Definition Phase, Final Report-Report No. EFJ-R-EFA-000-107, October 1987. [96] E. Jakeman, E. P. Raynes, Phys. Lett. 1972, 39A, 69. [97] K. M. Bofingar, H. A. Kurmier, M. R. Romer, J. D. E. Beubert, German Patent 4699636 (E. Merck, Darmstadt, Germany). [98] US Military Standard MIL-L-85762. [99] F. Gharadjedaghi, J . Appl. Phys. 1983, 54, 4989. [loo] H. L. Ong,J. Appl. Phys. 1988,63, 1247; H. L. Ong, Proc. SlD 1988, 29, 161. [ l o l l G. Schauer, W. Wiemer, Eurodisplay 1990, 122. [lo21 D. Potzsch, H. Baeger, F. W. Nickol, E. Purtzel, H. Voss, Proceedings of 12th Freiburger Arbeitstagung Fliissigkristulle, Freiburg, Germany, 1982. [lo31 P. G. de Gennes, J. Prost, The Physics o f L i q uid Crystals, 2nd edn. Oxford University Press, London, 1993. [lo41 M. J. Press, A. S. Arrott,J. Phys. 1976,37,387. [lo51 B. Kerllenvich, A. Coche, Mol. Cryst. Liq. Cryst. 1985, 124, 149; 1983, 97, 103. [ 1061 P. R. Gerber, Z. Naturforsch. a 1981, 36, 7 18. [lo71 A. Saupe, J . Chem. Phys. 1980, 72, 5026. [ 1081 H. S. Cole, Jr., S . Aftergut, J . Chem. Phys. 1978, 68, 896. [lo91 S . Aftergut, H. S. Cole, Jr., Appl. Phys. Lett. 1981, 38, 599. [I101 T. Ishibashi, K. Toriyama, K. Suzuki, M. Satoh, Proc. lnt. Display Res. Conf. 1990, 186; Proc. SID 1983, 24, 152.
302
3.4 Guest-Host Effect
11111 D. Jones, B. Desai, SOC.Autorn. Engrs 1980, 91; J. Mech. Engng 1981, 331. [112] M. Schadt, P. Gerber, Mol. Cryst. Liq. Cryst. 1981, 65, 241. [113] P. W. Ross, B. Needham, D. Coates, Eurodisplay 1981, 55. 11141 Data sheets from dichroic LCD manufactures such as Litton Data Images, Ottawa, Canada, Goodrich Aerospace Displays, Hatfield, USA, and Stanley Electric Co., Tokyo, Japan. [ 1151 C. Mauguin, Bull. SOC.Fr. Mineral. 1911, 34, 71. 11161 H. de Vries, Acta Crystallogr. 1951, 4, 219. [117] D. W. Berreman, J. Opt. SOC.Am. 1973, 63, 1374. 11181 M. Schadt, J. Chern. Phys. 1979, 71,2336. [119] J. L. Fergason, Appl. Opt. 1968, 7, 1729. [120] F. Fischer, Z. Nuturforsch. a 1976,31,41. 11211 J. F. Clerc, Displays, July 1985, 148. [122] G. Gharadjedaghi, A. E. Lagos, Proc. SID 1982,23,237. 11231 W. A. Crossland, B. Needham, P. W. Ross, Proc. SID 1985,26,237. 11241 S. Mitsui, Y. Shimada, K. Yamamoto, T. Takamatsu, N. Kimura, S. Kozaki, s. Ogawa, H. Morimoto, M. Matsuurd, M. Ishii, K. Awane, SID Digest 1992,437. [125] C. Casini, US Patent 4516834, May 14, 1985. [126] H. L. Ong, Jpn. J. Appl. Phys. 1988,27, 2017. [127] C. S. Oh, G. Kramer, Displays, January 1982, 30. [128] F. Gharadjeadaghi, E. Saurer, IEEE Trans., Electron. Dev. 1980, ED27, 2063.
[129] F. Gharadjedaghi, Mol. Cryst. Liq. Cryst. 1981, 68, 127. [130] J. L. West, R. Ondris, M. Erdmann, SPZE, Liq. Cryst. Disp. Appl. 1990, 57, 76. [ 1311 P. S. Drzaic, A. M. Gonzales, P. Jones, W. Montoya, SID Digest 1992, 571. [132] P. Jones, W. Montoya, G. Garza, S. Engler, SID Digest 1992,762. [133] H. Finkelmann, W. Meier, H. Scheuermann, in Liquid Crystals -Applications and Uses, Vol. 3 (Ed.: B. Bahadur), World Scientific, Singapore, 1992. [134] R. Simon, H. J. Coles, Liq. Cryst. 1986, I , 281. 113.51 H. J. Coles, Faraday Discuss. Chern. SOC. 1985, 79, 201. 11361 C. Bowry, P. Bonnett, M. G. Clark, presented at the Euro Display Conference, Amsterdam, September 1990. [ 1371 D. Coates in Liquid Crystals-Applications and Uses, Vol. 1 (Ed.: B. Bahadur), World Scientific, Singapore, 1990. [138] T. Urabe, K. Arai, A. Ohkoshi, J. Appl. Phys. 1983,54, 1552; Proc. SID 1984,2.5,299. [139] R. L. Van Ewyk, I. O’Connor, A. Moseley, A. Cuddy, C. Hilsum, C. Blackburn, J. Griffiths, F. Jones, Displays 1986, 155; Electron. Lett. 1986,22,962. 11401 L. J. Yu, M. M. Labes, Appl. Phys. Lett. 1977, 31,719. [141] D. Bauman, K. Fiksinski, Eurodispluy 1990, 278.
Handbook ofLiquid Crystals D. Demus, J. Goodby, G. W. Gray, H.-W. Spiess, V. Vill Copyright 0WILEY-VCH Verlag GmbH. 1998
Chapter IV Chiral Nematic Liquid Crystals
1 The Synthesis of Chiral Nematic Liquid Crystals Christopher J. Booth
1.1 Introduction to the Chiral Nematic Phase and its Properties Chiral nematic liquid crystals, as the name suggests, are optically active variants of nematic liquid-crystalline compounds; the incorporation of a chiral centre imparts properties which are unique to the chiral nematic phase and are responsible for their utilisation in a variety of differing display technologies and other related applications. The term cholesteric liquid crystal was originally used to describe this phase, and originates from the structural nature of the earliest chiral nematic liquid crystals which were derivatives of cholesterol [ 1,2]. Nowadays, the term chiral nematic is used primarily because the materials are clearly derived from nematic type liquid crystals [3, 41. Despite these differences in definition, the terms cholesteric and chiral nematic phase are interchangeable and it is common to find references to either term in the literature. The incorporation of a chiral centre into a formerly nematic structure (or alternatively of a chiral non-mesogenic dopant into a nematic host) results in the induction of a
macroscopic helical twist distortion in the bulk of the sample [ 5 , 6 ] .As a consequence the chiral nematic phase can be described as having helical orientational order or being a single twist structure [4]. The director ( n ) may precess through 360"; the distance over which this occurs is called the pitch length ( p ) , and this is shown simplistically in Fig. 1
1
pitch length, p
Figure 1. Diagram of the structure of the chiral nematic phase.
304
1 The Synthesis of Chiral Nematic Liquid Crystals
All interest in the unique optical properties of the chiral nematic phase stems from the two-fold optical activity of the phase [5, 6-91. That is, the mesophase displays (1) molecular optical activity - the phase being composed of optically active molecules and (2) macromolecular optical activity - arising from the macroscopic helical twist induced by the chiral molecules in the phase (sometimes termed form chirality). These features are of immense importance technologically, as many of the applications of such materials depend on one if not both of these phenomena. As would be expected for any optically active molecule, a chiral nematogen should rotate the plane of plane polarized light. This is indeed the case, but in contrast to other optically active materials (e.g. sugars), very large values of rotation are observed (ca. lo3 deg. mm-’ [5]); this may be attributed to the relative phase retardation of the right and left handed components of the plane polarized light as they encounter different refractive indices associated with the macroscopic helical twist of the phase. Rotation of the plane of polarized light usually occurs when the pitch length of the chiral nematic phase is very much longer than the wavelength of the incident light (i.e. p %-A).A second optical property associated with this phase is circular dichroism, whereby the chiral nematic phase will transmit one circularly polarized component of light whilst reflecting the other circularly polarized component. Which circularly polarized component becomes reflected or transmitted depends on the handedness of the macroscopic twist associated with the chiral nematic phase; a right handed helix will transmit the right-handed component whilst reflecting the left-handed circularly polarized component [4]. However, the most noted feature of the chiral nematic phase is its ability to reflect incident white
light selectively, provided that the incident light has a wavelength of approximately the same order as the pitch length of the phase (i.e. il= p ) . Coherently reflected light components experience constructive interference from other reflected wavefronts, giving rise to iridescent colour play (various analogies with Bragg reflection are possible, assuming normal incidence of the light beam). The wavelength or waveband of the reflected light is related to the pitch length by Eq. (1) [8]:
AA=An.p
(1)
where An = optical anisotropy of the phase (or birefringence). The optical purity (or the width) of the selectively reflected component ( A n ) may be modulated to some degree by manipulation of the optical anisotropy ( A n ) of the phase (this is usually achieved by controlling the sp2 or sp3 nature of the mesogen’s chemical structure). Special mention must also be made for the case of obliquely incident light on a chiral nematic phase; here the reflection band is shifted to shorter wavelengths and results in an angle-dependent effect to the viewer. However, this is outside the scope of this chapter and readers are referred to references [lo-121. These optical properties are observed only when the mesophase is in the planar or Grundjeun texture, that is where the helical/optical axis is perpendicular to the glass substrates when a sample is constrained between a microscope slide and cover slip [ 131, and the molecules lie in the substrate plane. There is a second common texture associated with the chiral nematic phase, where the helical axes of domains in the bulk sample are randomly aligned, and this is sometimes referred to as the pseudo focalconic texture; light undergoes normal scattering when the sample is in this texture [5, 131. This texture is easily mechanically
1.2 Formulation and Applications of Thermochromic Mixtures
disturbed to give the selectively reflecting planar texture. Certain chiral nematic liquid crystals (or mixtures) are found to exhibit thermochromism, that is, temperature dependent selective reflection; here the pitch length of such materials or mixtures is often found to have an inverse relationship with temperature and can depend on the nature of the material’s mesomorphism. Thermochromism is found to be at its most spectacular in materials which display chiral nematic phases and underlying smectic phases; this is exemplified for a material with an I-N*SmA* sequence (on cooling) and is shown schematically in Fig. 2. Here the pitch length gradually lengthens from the blue end of the electromagnetic spectrum (i.e. p = 0.4 pm), but an altogether more rapid increase in pitch length is experienced as the chiral nematogen is cooled towards the smectic A phase. This is best explained by the gradual build-up of smectic order on approach to the N*-SmA* transition, and results in startling colour play through to the red end of the electromagnetic spectrum (p=O.7 pm) as the helical structure of the phase becomes unwound (indeed the SmA* phase may be considered as having an infinite pitch length). It should be noted that not all chiral nematogens are either as responsive or behave in precisely
305
the same manner, and indeed some materials are notable for being exceptions to this rule of thumb [ 141. As an aside, it may be of interest to the reader that although selective reflection is usually a phenonemon associated with calamitic liquid crystalline systems, various multiyne materials have recently been demonstrated to show selective reflection as well as a helix inversion in chiral discotic nematic phases [15, 161.
1.2 Formulation and Applications of Thermochromic Mixtures Rarely, if ever, does one particular chiral nematic compound display “ideal” behaviour, which would make it suitable for use in a given application, and normally carefully formulated mixtures of a variety of mesogenic and non-mesogenic components, which may have very different thermodynamic properties, are used. Formulation of mixtures is frequently a complex trade-off of one property over another and is usually achieved by application of the Schrodervan Laar equation [ 17 -201, which allows the thermodynamic characteristics of a par-
1
I
I I I I
I I
Pitch
I I
Length
I
I
P)
I I I I
Smecuc A*
1 I I
I
C h i d Nematic
I
Temperature (“C)
I I
Isouoplc Liquid
Figure 2. The pitch length- temperature dependence of a chiral nematic liquid crystal.
306
1 The Synthesis of Chiral Nematic Liquid Crystals
ticular mixture to be predicted theoretically (with some accuracy) before recourse to actual experimental formulation of the mixture. Practical commercial thermochromic mixtures must have the following general characteristics: low melting points, wide operating ranges (i.e. -50- 150“C), short pitch lengths and underlying smectic A* phases to induce the spectacular pretransitional unwinding of the helix of the chiral nematic phase, which results in a visible thermochromic effect. It is usual to employ racemic or partially racemized compounds, as these have the effect of compensating the twist induced by other chiral components, enabling the fine tuning of the colour playtemperature characteristics, although this sometimes has the undesired side effect of extending the blue selective reflection range [3,5,21]. Similarly, the mixing of polar and non-polar liquid-crystalline components may result in the formation of injected smectic phases; these may also be exploited to give wider operating temperature ranges (they are also reported as preventing the extension of the “blue-tail’’ phenomenon) 1221. The colour purity of a thermochromic mixture (An),is related to the birefringence (An) of the mixture, and as mentioned earlier, this may be controlled by judicious selection of components which are either aromatic (sp2 hybridised), or alicyclic (sp3 hybridised) or a combination of the two (sp2- sp3) [ 171. The commercial synthesis of a wide variety of modern chiral nematic materials has made many of these points a little easier to address, in contrast to times when the choice of materials was limited to unstable cholesterol derivatives, azo-compounds, and Schiff’s bases; nevertheless, the final formulation of a mixture is still a complicated process. As may now be more clearly appreciated, the potential uses of chiral nematic liquid crystals depend on either the twisting pow-
er or the thermochromic response of the chiral nematic phase (N”). The former macroscopic property was initially the reason for most of the commercial interest in the use of chiral nematic liquid crystals; here chiral components were employed in electro-optic displays to counter reverse twist domain defects (in low concentrations ca. 1-10% wt.wt) and to sustain the helical twist induced by surface alignment in both the twisted nematic display and supertwist display modes [23, 241. Chiral nematic dopants may also be employed to induce short pitch lengths in White - Taylor Guest - Host displays (this prevents waveguiding of light in the off-state 117, 251) or as circularly dichroic filters 1261. However, the major applications of chiral nematic liquid crystalline materials are in thermometry, medical thermography, non-destructive testing, pollutant sensing, radiation sensing, surface coatings, cosmetics, and printing inks for decorative and novelty applications [3, 51; all of these applications employ either the selective reflective properties or the thermochromic response of the chiral nematic phase, or both. In many of these applications, the thermochromic liquid crystal mixture is rarely utilised as a thin, neat film, but rather as a microencapsulated slurry which is incorporated into a coating; microencapsulation of the thermochromic mixture is usually achieved by complex coacervation using gelatin and gum-arabic to form the microcapsule walls [27]. The reasons for microencapsulation are firstly economic and secondly to protect against and prevent attenuation of the optical properties of the chiral nematic liquid crystals, which are often prone to chemical and photochemical degradation during use. However, this may be reduced to some extent by use of UV filters or free-radical scavengers [3]. The availability of microencapsulated ‘inks’ have led to the increased use of the materials simply
1.3 Classification of Chiral Nematic Liquid Crystalline Compounds
because they may be applied to a variety of surfaces (paper, plastic, fabrics, and metal) by a variety of widely used techniques, e.g., airbrush, screenprinting, gravure, or flexographic means) [3, 281.
1.3 Classification of Chiral Nematic Liquid Crystalline Compounds Despite the wide variety of chiral nematic liquid crystals (and of course low molar mass chiral liquid crystals in general), it is possible to sub-divide them into three class types, according to the relationship of the chiral moiety and the liquid-crystalline core [29]. Firstly, we consider type I; here the chiral centre (or multiplicity of chiral centres) is situated in a terminal alkyl chain attached to the effective liquid-crystalline core. As will be seen later, it is entirely possible that the compound can have two or more chiral terminal groups associated with the molecule. Type I materials are probably the most frequently encountered chiral nematic liquid-crystalline compounds, often
3 07
because of their relative ease of synthesis and because of the availability of suitable chemical precursors. Secondly, type I1 materials: here the chiral centre may be trapped between two liquid-crystalline core units, and the structure resembles a dimer or a twinned molecule. Here, in principle, it is possible to modulate the properties of the molecule by variation of the length of the flexible spacer which carried the chiral group; this is believed to occur via processes which restrict the freedom of rotation about this central axis. However, this type of material has the disadvantage that the core units, which are usually polarisable in nature (i.e. sp2-hybridised carbon skeleton) are separated by a non-polar region (i.e. sp3hybridised carbon skeleton), preventing effective conjugation of the cores which usually leads to destabilisation of any liquidcrystalline characteristics. Thirdly and finally, type 111: here the chiral centre is in some way incorporated into the effective liquid-crystalline core: for example see Fig. 3 . In type 111 materials asymmetry may be achieved either by the presence of a chiral atom in the core or by a particular structural feature which results in gross molecular asymmetry. This stuctural type may be
Type I; terminal chiral chains appended to a liqmd-crystallme core
Type II: flexible chiral spacer chain between two liquid-clystalline cores
Type III: chiral point or structural molecular asymmetry within the liquid-crystalline core
Figure 3. The classification of chiral liquid crystals according to the position of the core unit and chiral moiety.
308
1
The Synthesis of Chiral Nematic Liquid Crystals
of use when attempting to gain the greatest coupling effect between the chiral centre and core to produce strongly twisting dopants. All three types of material are schematically illustrated in Fig. 3. Many of the compounds used as examples in this chapter to illustrate the three classes of chiral nematic structure are not necessarily compounds in which the presence of a chiral nematic phase or its properties were of primary importance to the researchers concerned in the work; for instance some of the materials were developed as potential ferroelectric dopants, but happened to show a chiral nematic mesophase. Also in many cases, the synthetic techniques used and the hazardous nature of some of the reagents employed, along with the poor chemical or photochemical stabilities of the mesogens, would almost automatically preclude their use in any commercial applications (i.e., thermography). However, such materials are of scientific interest in that they help to illustrate the many potential variations of novel molecular structures which may result in a chiral nematic mesophase.
1.3.1 Aspects of Molecular Symmetry for Chiral Nematic Phases Having defined the different forms of chiral nematogen structure with regard to the positioning of the chiral moiety relative to the liquid-crystalline core unit, it is now necessary to describe further and more precisely the exact nature of this structural relationship. It has been understood for many years that optical properties, such as helical twist sense and the direction of rotation of plane polarised light, depend intimately on the absolute spatial configuration of the chiral centre, the distance the chiral centre is sep-
arated from the core, the electronic nature of the substituents attached to the chiral centre and the overall enantiomeric purity of the system in question. These molecular features may be more clearly appreciated with reference to the schematic structure of a type I chiral liquid crystal shown in Fig. 4. Firstly, the absolute configuration of an organic compound is determined using the Cahn, Ingold and Prelog sequence rules which relate the atomic number of the groups attached to the chiral centre to their relative priority over one another (the higher the atomic number, the higher the priority; i.e. C1 has higher priority than H). The resulting order of increasing priority may then be designated as being either (R)- or (S)-; that is rectus or sinister (right or left forms) [4, 301. Additionally, it is necessary to define the distance of the chiral centre from the core (rn + l), as it has been found that the helical twist sense of the resultant chiral nematic phase alternates from odd to even numbers of atoms [ 311. This may be illustrated by considered the following two (S)-4-alkyl-4’-cyanobiphenyls(A and B), in Fig. 5 . Subsequent experimentation by contact studies with other chiral systems (R)- and (S)-, of differing parity revealed the following helical twist sense relationships, entries 1 to 4 in Table 1. These results form the basis of a set of empirical rules termed the Gray and McDonnell rules [ 3 2 ] and are particularly useful in predicting the properties of a chiral nematic phase of a given com-
X and Y may represent polar or non-polar groups; m and p = integers.
Figure 4. A representation of a type I chiral liquid crystal.
1.3 Classification of Chiral Nematic Liquid Crystalline Compounds
309
mathematically as
P = (P .C J 1 A
Abs. config ...(3)-;parity (m+1) ...2 (even, e); rotation of plane polarized light ...d; twist sense...LH.
B Abs. config ...(Sj-; parity (m+1) ...3 (odd, 0):
rotation of plane polarized light ...I; twist sense...RH
Figure 5. The relationship between twist sense, parity, and absolute configuration demonstrated for two chiral nematogens.
pound. The first four entries in Table 1 all correspond to materials which have an electron donating substituent at the chiral point (i.e. CH,-, +I). Electron withdrawing groups, such as C1 or CN (-1) were later demonstrated effectively to reverse these empirical twist sense rules; entries 5 to 8 in Table 1 all reflect this electronic structural change [ 3 3 ] . Finally, the other property which is also strongly dependent on molecular symmetry associated with the chiral centre is the twisting power, P, which is usually expressed Table 1. Rules relating twist sense to absolute configuration, parity and the electronic nature of the substituents at the chiral centre. Absolute configuration
Parity
Electronic nature of substituent
1. ( S ) (S)-
e
+I +I +I +I -I -I -I -I
2. 3. 4. 5. 6. 7. 8.
(R)(R)(S)(5’)(R)(R)-
0
e 0
e 0
e
o
Rotation of plane polarized light
Helical twist sense LH RH RH LH RH LH LH RH
(2)
where p =pitch length, c, = concentration of the chiral solute species, i; a right or left handed system is usually denoted by P being either positive (+) or negative (-), respectively [34, 3.51. This mathematical relationship is usually only applicable to systems which contain low concentrations of the chiral dopant (i), usually bicyclo[2.2.2]octyl> cyclohexyl [5]. This entire class of stable, colourless chiral nematogens turned out to be commer-
During the late 1980s the highly optically pure intermediate (R)-2-(4-hydroxyphenoxy)propanoic acid (RHPP; ee >97% [70]) was employed as the basis of a series of novel thermochromic liquid crystals. RHPP is a bifunctional structure and theoretically allows for extension of the molecule along the molecular long axis at either terminal functional group, to give structures which may be classified as type I or type I1 materials. Additionally, the chemical nature of these groups could be modified to give other varieties of polar linkage (e.g. C 0 2 H converted to either CN or CH,OH), to give variation in the thermodynamic properties of the different structures. Initial work revealed that chiral nematic phases were only obtained when the carboxylic acid moiety of RHPP was converted into either a simple alkyl ester or a cyano group. This behaviour is highlighted by the ( R ) and (R,S)-alkyl 2-( 4-[4-(truns-4-pentylcy-
320
1 The Synthesis of Chiral Nematic Liquid Crystals
Table 5. Transition temperatures and phase assignments for a series of (R)-and (R,S)-2-{4-[4-(truns-4-pentylcyclohexyl)benzoyloxy]phenoxy) propanoates and a propanonitrile.
Compound no.
61 62 63 64
65
Z C02C3H7
Configuration
Transition temperature ("C)
(R)-
Cr 67.2 (SmA* 60.5) N* 77.9 I Cr 77.0 (Cryst B* 35.8 SmA* 67.9) N 82.0 I Cr 58.1 (SmA* 46.2) N* 67.3 I Cr 70.6 Cr 77.5 (Cryst B* 19.2 SmA* 51.4 N 70.9) I Cr 122.3 (SmA* 92.9) N* 151-6 I
C02C3H7 C02C4H9
(R)-
C02C4H9
(RS)-
CN
(R)-
clohexyl) benzoyloxy] phenoxy } propanoates (61 - 64) and (R)-2-{4-[4-(truns-4pentylcyclohexyl)benzoyloxy]phenoxy } propanonitrile (65) [71, 721, whose chemical structures and transitions are given in the Table 5. However, analytical determinations performed using 'H nmr and the chiral shift reagent, (+)-europium tris-(D-3-heptafluorocamphorate) revealed that these materials had inherently low enantiomeric excesses (ee). The ee values of the propyl (61) and butyl esters (63) were found to be between 0.69-0.75 on integration of the areas for the aromatic protons ortho to the carboxylate moiety of the trans-4-pentylcyclohexylbenzoate core. The low values of ee have been attributed to the presence of the electronwithdrawing carbonyloxy or cyano groups a to the chiral centre, which readily facilitates racemization, even during the relatively mild synthetic processes performed previously during the synthesis of the compounds. It is also interesting to note that the independently synthesized racemic modifications of these propyl and butyl esters (62 and 64) have noticeably higher melting points and clearing points when compared with their (R)-enantiomers (61 and 63). This behaviour has been noted before in other chiral systems and is an indication of the general influence and importance of enan-
tiomeric purity on a compound's physical properties [72]. This problem of low enantiomeric purity (ee) was circumvented by removing the electron withdrawing group. This was achieved by the reductive cleavage of a benzyl ester function to give the alcohol, using lithium aluminium hydride, and resulting in a series of materials derived from the compound (R)-2-(4-benzyloxyphenoxy)propan1-01(68) [73]. The synthetic route employed is outlined in Scheme 6. The first step involves heating (R)-benzyl 2-(4-hydroxyphenoxy)propanoate (66) with benzyl bromide and potassium carbonate in butanone to give the dibenzyl compound 67. Compound 67 was the reduced to (R)-2-(4-benzyloxyphenoxy)propan-1-01 (68) using lithium aluminium hydride in dry tetrahydrofuran at room temperature. The alkylation of compound 68 was then achieved using sodium hydride and the appropriate alkyl halide in dry N,N-dimethylformamide at room temperature to give the (R)-1-alkoxy-2-(4-benzyloxyphenoxy)propanes (69-75) as colourless oils. These compounds were then debenzylated using hydrogen and palladium-on-charcoal at room temperature and pressure to give the chiral phenols (76-82) which were then subsequently esterified with the appropriate two ring core acid using dicyclohexylcarbo-
1.5 Type I Chiral Nematic Liquid Crystals
32 I
PhCH?Br,
LiAM4, THF, Nr NaH, AlkylBr, H3C
H3C 6 8 0 H
OCnHznil
69-15
H2, Pd-C, EtOH Core Acid, DCC, DMAP, CH2C12,RT H3C
OCnHznil
76-82
83-101
H3C
OCnHzn+1
a = trans- 1,4-cyclohexyl, 1.4-phenyl or 2-fluorophenyl.
b = 1,4-phenyl or 2,3-difluoro-1,4-phenyl. n=lt06
diimide (DCC) and 4-N,N-dimethylaminopyridine (DMAP) to give the target compounds 83-101. Proton NMR and chiral shift reagent studies on various (R)-1-alkoxy-2-{4-[4-(truns4-pentylcyclohexyl)benzoyloxy]phenoxy ] propanes (83 and 84) failed to resolve the presence of any of the other enantiomer, indicating that the material was optically pure (ee 20.98, within experimental error) and that this system was now inherently stable to racemization.
Scheme 6. Synthesis of ( R ) -1alkoxy-2-[4-(4-substituted)phenoxylpropanes.
The methyl to hexyl terminal chain homologous all showed chiral nematic phases, although their phase ranges decreased from 15.2 to 4.4"C as the homologous series was ascended. The most interesting feature of the phase behaviour of these materials was the occurrence of Blue phases near the clearing points of the higher homologues of the series (butyl to hexyl, compounds 83 to 88). This is clearly the result of damped or restricted rotation of the bonds attached to the chiral centre caused by both the prox-
Table 6. The transition temperatures and phase assignments for the ( R ) -1-alkoxy-2- (4-[4-(tran.s-4-pentylcyclohexyl)benzoyloxy]phenoxy Jpropanes.
Compound no.
n
Transition temperatures ("C)
83
I 2 3 4 5 6
Cr, 69.3 Cr, 83.5 Cryst B* 85.5 SmA* 107.5 N* 122.7 I Cr, 43.4 Cr, 59.7 Cryst B* 80.1 SmA* 105.1 N* 112.4 I Cr 54.8 Cryst B 73.9 SmA* 95.8N* 100.8 I Cr 48.8 Cryst B* 73.6 SmA* 89.5 N* 93.7 BP 94.0 I Cr 46.3 Cryst B* 68.1 SmA* 83.2 N* 86.7 BP 86.7 I Cr 44.5 Cryst B* 62.8 SmA* 74.0 N* 78.4 BP 78.4 I
84
85 86 87 88
322
1 The Synthesis of Chiral Nematic Liquid Crystals
imity of the liquid-crystalline core and the presence of longer peripheral alkyl chains [29]. The transition temperatures and phase assignments are given in Table 6. The esters with 4’-pentylbiphenyl cores were somewhat disappointing in their phase behaviour; all materials showed smectic A* and crystal B* phases; and only the methoxy compound 89 displayed a short chiral nematic phase (3.8 “ C ) .
H3C‘
bCH3
89; Cr 73.3 CrystB* 93.3 SmA* 131.6 N* 135.4 I (“C) [73]
The use of lateral fluoro-substituents in liquid crystals chemistry is well known for influencing physical properties, particularly in supressing smectic phases, as well as depressing both melting points and clearing points [74-791. By use of 2’-fluoro- and 2,3-difluoro-substituents in the biphenyl core, it proved possible not only to reduce both the melting points and clearing points, but also drastically to lower the thermal stability of the smectic A* phases, often to such a degree that room temperature iridescent chiral nematic phases were observed for a number of compounds in two different, but related, series. The acids providing the two different fluoro-substituted cores (4’pentyl-2’-fluorobiphenyl-4-carboxylicacid and 4’-pentyl-2,3-difluorobiphenyl-4-carboxylic acid) were synthesized by well-documented techniques for low temperature lithiation, carboxylation and formation of the boronic acids from aryl bromides as well as palladium (0) catalysed cross-coupling procedures [80-831. It is important to stress the versatility of these methods, the increased use of which has led to the accessibility of numerous structures throughout liquid crystal chemistry, which were previously thought to be too difficult to synthesize easily.
In the first of these fluoro-substituted series, the (R)-1-alkoxy-2-[4-(2’-fluoro-4’pentylbiphenyl-4-carbonyloxy)phenoxy]propanes (90-95), it was found that the 2’-fluorosubstituent destabilizes the ability of the molecules to pack efficiently to such a degree that they prefer to form chiral nematic phases over the more ordered smectic phases observed in their non-fluoro-substituted parents (see compound 89) [841. The first five materials show monotropic smectic A phases at correspondingly low temperatures; furthermore the butyl, pentyl, and hexyl homologues did not crystallize above a temperature of -40 “C (potentially an advantage for microencapsulation). The transition temperatures of this series are shown in Table 7; the clearing points are seen to decrease very rapidly from 8 1.2 to -6.1 “C. A similar, but less rapid decrease in thermal stability of the SmA phase is also observed. The propyloxy and butyloxy compounds (92 and 93) are both of interest in that they show highly iridescent chiral nematic phases at or around room temperature. Spectroscopic measurements, indicate that the homologues of this series have pitch lengths which vary in magnitude between 0.30 and 0.32 ym. Table 7. The transition temperatures and phase assigments of (R)-1-alkoxy-2-(4-(2’-fluoro-4’-pentylbiphenyl-4-carbonyloxy)phenoxy]propanes. F
Compound no.
n
Transition temperatures (“C)
1
Cr 54.3 (SmA* 24.5) N* 81.2 I Cr 40.6 (SmA* 4.9) N* 50.1 I Cr 36.2 (SmA* 6.0) N* 47.4 I Cr -(SmA* -4.1) N* 18.6 I Cr - (SmA* -10.0) N* 7.7 I Cr N* -6.1 I
90 91 92 93 94
4 5
95
6
2 3
1 .5
Turning now to the second fluoro-substituted series, (R)-l-alkoxy 2-[4-(2,3-di-
Type I Chiral Nematic Liquid Crystals
323
propanonitrile [72] and (R)-1-methoxy-2( 4- [4-(trans-4-pentylcyclohexyl)-benzoylfluoro-4’-pentylbiphenyl-4-ylcarbonyloxy)- oxyl-phenoxy }propane (83) [73], revealed phenoxylpropanes (compounds 96 to 101), that they show opposite helical twist senses the transition temperatures listed in Table 8, (the compounds are left handed and right show that as would be expected, the inhanded, respectively) as would be predictsertion of a second fluoro-substituent aped for materials with electron acceptor and pears to have little extra effect on the cleardonor groups at the chiral centre [32, 33, ing point over that of the 2’-fluoro-substi851. Interestingly, certain composition mixtuted series [84]. tures displayed either TGBA* phases mediThe melting points, clearing points and ating the chiral nematic to smectic A* tranthe N*-SmA* temperatures are indeed all sitions or twist inversion phenomena in the slightly higher than for the corresponding chiral nematic phase. 2’-fluoro homologues; the SmA* phases all The use of the stable, optically pure ( R ) -1-alkoxy-2-(4-substituted-phenoxy)occur enantiotropically, indicating greater propane system in conjunction with the apthermodynamic stability, unlike the precedpropriately fluoro-substituted liquid-crysing SmA* phases of the 2’-fluoro series (90 talline cores offers great potential for mateto 95). This overall result may be partly due rials which are suitable for use in thermoto the shielding of the 3-fluor0 substituent chromic mixture formulations [SS]. by the carboxylate linkage and the fact that the second fluoro-substituent effectively makes the liquid crystal core no broader 1.5.5 Miscellaneous Type I (i.e. the second fluoro-substituent has little Chiral Nematic Liquid Crystals effect on the overall geometrical anisotropy of the molecule). The former examples of type I materials are Binary mixture studies involving the not the only examples of the class; many chiral nematogens (R)-2- [(2’-fluor0-4’other forms of novel side chains have also pentylbiphenyl-4-carbonyloxy)phenoxy lbeen investigated. These miscellaneous chiral nematic systems include the following types of material: (R)-and (5’)-l-methylalkTable 8. The transition temperatures and phase assignments of (R)-1-alkoxy-2-(4-(2,3-difluoro-4’-pen- oxy derivatives (102) [87, 881; derivatives of (S)-P-citronellol (103) [89]; (R)-2-chloty l biphenyl-4-carbonyloxy)phenoxy]propanes. ropropanol and (S)-2-ethoxypropanol(lO4) F F 190, 9 11; (S)-2-chloropropyl derivatives (105) [29, 921; (S)-2-halogeno-4-methylpentyl derivatives (106) [29, 931; materials with (2S,3S)-3-propyloxirane groups (107) Compound n Transition temperatures (“C) [29,94]; and (R)-4-(1-propoxyethyl)phenyl no. derivatives (108) [95]. These materials will 96 1 Cr 34.3 SmA* 41.4 N* 85.6 I not be covered in great detail, as many of 97 2 Cr 17.5 SmA* 45.6 N* 74.5 I them were synthesized as potential ferro98 3 Cr 18.2 SmA* 25.1 N* 54.1 I electric materials, nonetheless, they high99 4 Cr 1.4 SmA* 9.5 N* 32.8 I 100 5 Cr 2.6 SmA* 13.6 N* 35.8 I light the use of other novel chiral side 101 6 Cr 0.6 SmA* 15.8 N* 42.1 I chains.
324
1 The Synthesis of Chiral Nematic Liquid Crystals
102; (S)-Abs config.; Cr 71.8 SmC* 89.8 N* 137.4 I ("C) [87, 881 A
103; Cr 79.5 SmC* 116.6 N* 150.0 I ("C) [89]
OCzH5
104; Cr 27.9 (Sm2 16.9) SmA* 45.2 N* 50.3 I ("C) [90,91]
ci 105; Cr 102.6 SmA* 137.0 N* 166.0 I ("C) [29,92]
106;Cr 67.8 SmC* 79.9 SmA* 96.8 TGB A* 100.4N* 128.3 BP 129.4 I ("C) [29, 931
107; R=C,H,; Cr 59 (SmCg 46.2 SmCZ 46.8 SmCt) Ng 106.3 NE 112.1 NE 158.1 BPI 162.9 BPI1 164.6 I ("C) [29, 941
Some of these materials were readily available from commercial sources, others were not and have had to be synthesized from convenient precursors. This has often led to two distinct types of approach: firstly, by employing mild synthetic methods which somehow preserve the optical activ-
ity of a particular system; secondly, where the former method was not possible, asymmetric induction methods have been employed. Here the chiral moiety may be synthesized stereospecifically, for example by employing a-amino acids such as (S)-alanine, (S)-leucine, (S)-valine, or (2S,3S)-isoleucine [29, 92, 931, or by using chiral catalysts or auxiliaries [94,95]. These catalysts and auxiliaries allow a reaction to take place at one specific face of the transition state complex during the reaction, thereby leading to a preponderance of one isomer over others (see also the Sharpless asymmetric epoxidation in Sec. 1.7.3 of this Chapter). One of the more unusual and elegant examples of the uses of a chiral catalyst is in the synthesis of the novel mesogen (R)-1(4'-nonyloxybiphenyl-4-y1) 4-( 1 -propoxyethy1)benzoate (108) [95]; this is outlined in Scheme 7. The achiral starting material, methyl 4-acetylbenzoate (109) was enantioselectively reduced using borane in the presence of the chiral oxazaborolidine (110) (itself synthesized from (S)-(-)-2-diphenylhydroxymethy1)pyrolidine and BH3 : THF [96]) to give the optically active benzyl alcohol (lll),in 89% ee. The (R)-4-( l-propoxyethy1)benzoic acid (112) was then obtained in a 25% yield after the sequence of base hydrolysis, propylation and base hydrolysis procedures. Compound 112 was then esterified in the standard manner to give the target compound (108). (R)-1-(4'nonyloxybiphenyl-4-y1) 4-( 1-propoxyethyl)benzoate (108) showed only a very short chiral nematic (N*) phase of range 0.5 "C, after undergoing a direct SmC*-N* phase transition at 104.5 "C.
1.6 Type 11 Chiral Nematic Liquid Crystals
325
110, BH,, THF.
* H3CO&
H3C02C H:* 3
111
109
I NaOH, MeOH; (ii) NaH, CSH,I; (iii) NaOH, MeOH (I)
(i) SOCI,; (ii) ArOH, pyridine.
C9H,90 ~
O
Z
C
~
-
c OC3H7
H
3
H 0 2 c ~ c H 3 OC3H7
112 108; Cr98 SmC' 104.5 N* 105 I ("C) [95]
110 [961
Scheme 7. Asymmetric synthesis of (R)-l-(4'-Nonyloxybipheny1-4yl) 441-propoxyethyl)benzoate.
1.6 Type I1 Chiral Nematic Liquid Crystals
1.6.1 Azomethine Ester Derivatives of (R)-3-Methyladipic Acid
Achiral twin compounds, which possess two liquid-crystalline cores separated by a flexible spacer, are now well known and documented. As well as being pre-polymer model systems, they also give large variations in transition temperatures which are dependent on the length of the flexible spacer and frequently give incommensurate smectic phases [97- 1001. However, examples of chiral materials which may be categorized in this manner are not numerous, and the majority of these examples concern the use of flexible, chiral linking groups based upon (R)-3-methyladipic acid [ 1011051, derivatives of lactic acid [106, 1071or optically active diols 11081 between two liquid-crystalline moieties. This class of chiral materials, however, remain an interesting subject in that, in principle, it is possible to modulate and sterically restrict motion about the chiral centres by varying the length of the flexible linking spacer unit.
The first examples of this kind of material belong to a homologous series of bis-azomethines (113-114) which have the general structure represented in Table 9 [loll. The materials were prepared by esterification of (R)-3-methyladipic acid with 4-hydroxybenzaldehyde to the bis-[bforrnylphenyl] (R)-3-methyladipate. This was then condensed with the appropriate 4-alkoxyaniline to give the desired bis-azomethine dimer. In this homologous series, only the early members (ethoxy, butoxy and hexyloxy, compounds 113 to 115) showed chiral nematic phases; the transition temperatures of these materials are given in Table 9.
326
1 The Synthesis of Chiral Nematic Liquid Crystals
Table 9. The transition temperatures and phase assignments of diesters derived from the chiral3-methyl adipic acid and 4-hydroxybenzylidene-4'-alkoxyanilines.
Compound no.
n
Transition temperatures ("C)
113
2 4 6
Cr 156.3 N* 236.6 I Cr 154.6 Sml 158.5 Sm2 169.3 N* 210.9 I Cr 122.8 Sml 140.0 Sm2 174.7 SmA* 190.0 N* 195.5 I
114 115 ~~~
~
Where Sml and Sm2 represent unidentified smectic phases.
1.6.2 Novel Highly Twisting Phenyl and 2-Pyrimidinylphenyl Esters of (R)-3-Methyladipic Acid A more recent continuation of this theme using (R)-3-methyladipic acid, was the systematic study of dimeric or twin structures which are capable of forming highly twisted helical structures [ 102- 1051.These examples, differ from the previously mentioned bis-azomethines in that they employ 2-pyrimidinylphenyl (116 and 117) and phenyl core units (118); they were made simply by esterification of (R)-3-methyladipic acid with the appropriately substituted phenol using dicyclohexylcarbodiimide (DCC) and the catalyst 44,N-dimethylaminopyridine (DMAP). The structures and phase transitions of members of these two classes are given in structures 116 to 118.
The transition temperatures for the 2-pyrimidinylphenyl based compounds (116 and 117) clearly show that they display only rather short range, monotropic chiral nematic phases. Bearing this in mind, it is not surprising that the smaller twin with a phenyl core (118) does not form a chiral nematic phase. Studies which employed these three materials as dopants have shown conclusively that the chiral nematic phases of the test mixtures have particularly short pitch lengths ( p = 16- 22 pm) and right handed twist senses. This contrasts noticeably with the monomeric materials which have similar structures, for example (S)-4-(4-[2-(4hexyloxyphenyl) - 5 -pyrimidinyl] -phenyl } 3-methylpentanoate (119), which has a considerably longer pitch length ( p = 79 pm).
119; Cr 118.2 (SmC* 185.4) N* 195.8 I ("C); p = Fma, LH-helix [102, 1031
It has been suggested that the marked difference in pitch lengths between the twins and monomeric materials is due in part to the further restriction of rotational motion about the chiral centre caused by the presence of a second bulky core moiety, result-
116; n = 7; Cr 139.6 (N* 136.9) I ("C); p = 19 pm', RH-helix [102-1041 117; n = 8; Cr 130.9 (N* 127.7) I ("C); p = 22 pm",RH-helix [102-1041
C8H170 - @ 0 2 c & C 0 2 ~
OC,H
17
118; mp = 56.7"C, p = 16 pna, RH-helix [102-1041
a Corresponds to a measurement made with a 2 wt% mixture in 6CB.
1.7 Type 111 Chiral Nematogens
ing in a more tightly twisting helical structure. However, if this was entirely true, then a similar increase in spontaneous polarization of the induced SmC* phases, which are also present in the test mixtures doped with twins, should also be observed. This is not the case, and this may possibly be due to the transverse dipole of the twin aligning perpendicularly to the C2 axis of the chiral smectic C phase [ 1021.
1.6.3 Chiral Dimeric Mesogens Derived from Lactic Acid or 1,2-Diols
327
(R,R)-2,3-butandiol, has shown the diesters to have high helical twisting powers (p) when used as dopants in suitable nematic mixtures. However, only one of these materials is known to show a liquid-crystalline phase; this is a derivative of (R,R)-2,3-butanol (122) [IOS]. The high thermal stability of the chiral nematic phase is not surprising, considering the size of the liquid-crystalline core (4'-(trans-4-pentylcyclohexyl)biphenyl-4-carboxylic acid); the chiral nematic phase occurs between 212 and 255 "C. C5H11
C5Hll
Other attempts have been made to employ the (R)-2-oxypropanoyloxy moiety as a chiral spacer between identical [ 1061 and dissimilar [7 1, 1071 liquid-crystalline core systems, typical structures are given by compounds 120 and 121.
122; Cr 212 (SmX* 197) N* 255 I ("C) [lo81
This compound has been demonstrated to give highly twisted chiral nematic phases when used as a dopant and shows a helix in-
120; Cr 104.9 (SmA* 80.4 N* 83.7) I ("C) [lo61
Compound 120 shows a short monotropic chiral nematic and smectic A* phase, whilst 121 shows a short enantiotropic chiral nematic and a monotropic smectic A* phase. From this it may be concluded that the lactic acid derived (R)-2-oxypropanoyloxy moiety is not particularly suited to sustaining liquid crystal phases; this may either be partly due to poor conjugation between the liquid crystalline cores or because of the non-linear nature of molecules. Research on diesters derived from the optically active diols, (S)- 1,2-propandiol and
version at approximately 60°C in a commercially available wide range nematic mixture.
1.7 Type I11 Chiral Nematic Liquid Crystals This class of chiral mesogen contains many structurally unusual and interesting liquid
328
1 The Synthesis of Chiral Nematic Liquid Crystals
crystal materials. Here the chirality is imparted by the presence of either a chiral atom situated within the core, or by the generation of gross molecular asymmetry as a result of the core’s overall spatial asymmetry. It may also be recalled from an earlier section on molecular symmetry (Sec. 1.3.1), that the chirality of a liquid-crystalline system may be increased if the chiral centre is brought closer to the mesogenic core, due to restricted motion of groups around the chiral centre [29]. This increase in chirality may manifest itself in many ways, such as the shortening of helical pitch length or the increase in spontaneous polarisation of a chiral dopant. Therefore, it may be appreciated that incorporation of the asymmetric centre in the core, to create a chiral core, is an attractive method of trying to develop a mesogen capable of forming a very highly twisted helical structure (i.e. a short pitch chiral nematic phase). Examples of these less common systems include the following compound types: twistanol (or tricycl0[4.4.0.0~~~]decane) derivatives [ 1091; cyclohexylidene ethanones [ 110, 1111; chiral oxiranes [112,1131; chiral 1,3-dioxolan4-ones [ 1141; substituted stilbene oxides [ 1151; chiral dioxanyl derivatives [ 1161; nitrohydrobenzofuran derivatives [ l 171 and cyclohexanes [ 118,1191. Although many of these systems show chiral nematic phases, their usefulness is severely limited for a number of reasons; the awkward and expensive synthetic routes, the necessity for inefficient and frequently impractical resolution techniques, and in many cases their ultimate chemical or photochemical instability.
1.7.1 Tricyc10[4.4.0.0~~~]decane or Twistane Derived Mesogens
has been incorporated into a number of liquid-crystalline (R)-(+)-8-alkyltwistanyl 4-(trans-4-pentylcyclohexyl)benzoates or 4’-pentylbiphenyl-4-carboxylates(130-132), some of the members of these series are reported as showing chiral nematic phases [ 1091.The twistane ring system has D, symmetry and may be viewed as consisting of four fused boat form cyclohexyl rings which all twist in the same sense; using the Cahn, Ingold and Prelog selection rules, the two enantiomeric forms may be classified as having P- or M-helicity [30]. The (R)-(+)8-alkyltwistanol moiety was prepared by use of a six step synthetic route, starting from the disubstituted cyclohexan- 1,5dione (123) shown below in Scheme 8. The first two steps, that is the stereoselective cyclization in the presence of (S)-(-)proline to give the bicyclic diketone (124) and the hydrogenation of the double bond formation during the condensation step to give compound 125, are crucial in setting up the appropriate stereochemistry necessary for the acid-catalysed cyclization step which results in the formation of the basic twistance core (126). Of the five materials synthesized in this way, only three show chiral nematic liquid crystal phases, as listed in Table 10. It is also of interest to note that compounds 131 and 132 are reported to show anomalous pitch dependency in their chiral nematic phase; the pitches are both reported to increase from 1.1 to 1.6 pm with increasing temperature.
1.7.2 Axially Chiral Cyclohexylidene-ethanones
Solladie and Zimmerman have reported a series of novel chiral liquid crystals which The optically active twistane ring, more are believed to be the first to incorporate a correctly named tricyc10[4.4.0.0~,~]decane, molecular unit which possesses axial chiral-
1.7
128
329
126
127
I LiAlH4
Type 111 Chiral Nematogens
ArCOCI,
OH pyridine
C5Hll
129
130-132 where R = C3H7 or CSH11.
Scheme 8. Synthesis of (R)-(+)-8-twistanol containing mesogens.
Table 10. The mesomorphic behaviour of (R)-(+)-8twistanol derivatives.
where R1 = alkyl, alkoxy, nitrile and R2 = alkyl or mcthyleneoxy.
Compound no.
X
R
Transition temperatures ("C) H
130 131 132
Ph") Ph Ch")
C3H, C,H,, CSH,,
Cr 94.2 N* 119.2 I Cr71.8N" 126.41 Cr 65.0 N* 124.7 I
Figure 7. (a) The general structure of a cyclohexylidene ethanone. (b) Axial chirality of the cyclohexylidene moiety.
Where Ph denotes 1,4-phenyl and Ch denotes trans1,4-cyclohexyl.
')
ity; these are the substituted biphenylyl cyclohexylidene ethanones of the general structure shown in Fig. 7(a) [110, 11 11. The axial chirality is the consequence of two perpendicular planes associated with the R, and hydrogen substituents on the C-4 carbon atom of the cyclohexyl ring and the alkene substituents, R , and H; this is more clearly seen in Fig. 7(b). The optically active compounds were obtained by a series of elegant stereoselective reactions, outlined in Scheme 9. The chiral sulfoxide (134) was obtained from the reaction of the Grignard reagent of
a bromo-(4-substituted cyclohexy1)methane (133) and (S)-(-)-methyl sulfinate. Compound 134 was then stereoselectively acylated using lithium di-isopropylamide (LDA) at -78 "C with an appropriately substituted aroyl chloride, to give the diastereoisomeric sulfoxides (135- 136). The (R,R)-and (S,R)-diastereoisomers may both be obtained as either the thermodynamic or the kinetic product of the reaction process. The final step involves the stereoselective pyrolytic elimination of the sulfoxide group to give the ( S ) - or (R)-cyclohexylidene ethanones (137 and 138) from the diastereoisomeric (R,R)- or (S,R)-sulfoxides (135 and 136).
330
1 The Synthesis of Chiral Nematic Liquid Crystals
(i) LDA, THF, -78 OC; (ii) ArCOC1, -78 'C.
Table 11. Mesomorphic behaviour of a variety of arylcyclohexylidene ethanones.
Compound no.
Configuration
(0 (0
139 140 141
(9-
Rl
R2
Transition temperatures ("C)
C5H1 1
CHZOCZH,
CH3 CN
C5H1 1 C5H11
Cr 43 SmX* 63 N* 67 I Cr 65 N* 124 I Cr 102 SmX* 113 N* 135 I
SmX* = represents an unidentified smectic phase.
Of the seven optically active cyclohexylidene ethanones synthesized, only three showed chiral nematic phases; these materials are listed in Table 11 . However, the conjugated nature of these materials makes them particularly photochemically unstable; irradiation of a racemic cyclohexylidene ethanone with UV light has been shown to result in the rapid isomerization of its double bond in a few hours to give compounds of the general structure (142), as well as the oxidation product, 4'-pentylbiphenyl-4-carboxylic acid.
142
Nonetheless, despite this poor photochemical instability, these materials un-
equivocally prove that materials possessing axial chirality are capable of forming chiral nematic phases.
1.7.3 Chiral Heterocyclic Mesogens A number of heterocyclic type I11 systems have been investigated, primarily as ferroelectric dopants, but a number of the materials do display short (sometimes unstable) chiral nematic phases. Notable examples of systems based on three-membered ring heterocycles are probably best demonstrated by the materials containing the (2R,3S)-2oxirane carboxylic acid unit [ 1121 and trans-stilbene oxides [ 1 151. The synthesis of the oxirane units is of interest as it is achieved by use of the Sharp-
1.7 Type 111 Chiral Nematogens
less asymmetric epoxidation procedure by treatment of either cis- or trans-allylic alcohols with a stoichiometric mixture of L-(+)diethyl tartrate, titanium isopropoxide and t-butyl hydroperoxide at low temperature to give the appropriate epoxide [ 112,113,120, 1211. This process is demonstrated in the synthesis of (2R,3S)-(-)-4-heptoxyphenyl 3- [truns-4-(truns-4-pentylcyclohexyl)cyclohexyl]oxirane-2-carboxylate (146) shown in Scheme 10 [113]. The use of a further mild oxidation and neutral esterification methods ensure that ring opening of the epoxide ring is minimized. As can be seen from the transition temperatures of compound 146, the chiral nematic phase has quite a high thermal stability, presumably a direct result of the two cyclohexane rings which are known to promote high clearing points [49]; the material also has an underlying smectic A phase. Interestingly, this form of epoxide is somewhat more stable than other examples (i.e. compound 107 [94]) because the electron-withdrawing carboxyl moiety reduces the basicity of the oxirane ring, reducing the tendency towards ring opening processes.
-
CSHll
OH
143
Ti('OPr)4,L-(+)-DET, 'BuOOH, CH2Cl2,-23 "C.
144
C5hl
' I
RUC13, NaI04, CC14, CH,CN, H20, RT. 0 CO2H
* *
145
4-hcptoxyphenol, DCC, DMAP, C H Z C I ~RT. ,
33 1
Larger chiral heterocyclic core systems have been synthesized, for example chiral 1,3-dioxany1-4-ones (147) [ 1141 methyldioxanyl (148) [I 161 and 2-alkyl-2,3-dihydrobenzofuran (149) [ 1171 derived mesogens, and the phase behaviour and transition temperatures of these materials are shown below.
147: Cr 71 SmC* 75 N* 85 I ("C) [ I 141
148; Cr 52 SmA* 57.4 N* 100.5 I ("C) [ I 161
CeHi70
w z C6H13
149; Cr 115 SmC* 140 SmA* 183 N * 185 I ("C) 11171
1.7.4 Chiral Mesogens Derived from Cyclohexane A number of elegant attempts have also been made to employ chiral cyclohexane rings as part of the liquid-crystalline core; the approaches to these materials have usually involved the use of either enantiospecific aluminium trichloride catalysed Diels -Alder reactions (150) [ 1181 or the chiral reducing agent alpine boramine on unsaturated cyclohexyl derivatives (151) followed by palladium (0) catalysed coupling procedures [119]. However, the full development and potential of such mesogens still remains a relatively unexplored field.
146; Cr52.0 SmB* 121.0 SmA* 159.7 N* 167.5 I ("C) [I131
Scheme 10. Synthesis of (2R,35')-(-)-4-Heptoxyphenyl 3-[trans-4-(truns-4-pentylcyclohexyl)cyclohexyl Ioxirane-2-carboxylate.
150: Cr 79 S2 94 S , 113 SmC* I32 SmA* I33 N* 150 I ("C) [I181
332
1 The Synthesis of Chiral Nematic Liquid Crystals
H,CO‘
151; Cr 120.1 SmA* 127.4 N* 139.9 I (“C) [119]
1.8 Concluding Remarks As has been stressed throughout this chapter, the development of chiral nematic liquid-crystalline materials has often taken its lead from the inadequacies and unsuitability of many of the existing materials which have been employed in particular technological roles. This materials development process is perhaps most clearly evident in the development of the stable, colourless ester derivatives which utilise the (S)-2-methylbutyl side chain, particularly as this evolution is easy to trace from the early sterolbased mesogens, to the chiral azobenzenes through to the chiral azomethine mesogens. As well as being commercially successful, these materials also led to the development of ground rules which relate the nature of the chemical structure to the physico-chemical properties. This has most importantly led to a certain degree of selectivity in the manipulation of properties during the process of choosing a material for any given application. The fact that these materials have relatively simple structures, especially when compared with those of the ‘exotic’ chiral nematogens, is of benefit and importance in that is makes them more readily accessible and therefore more economically viable at a commerical level. Nevertheless, the advent of certain synthetic methods (i.e. the palladium (0) catalysed cross-coupling reactions) has led to a number of notable issues such as improvements in the synthesis of known materials, improvements of physical characteristics as a result of minor
structural modifications (i.e., use of lateral fluoro-substituents) and the generation of entirely novel and frequently exotic chiral mesogens (i.e., twistanol derivatives or cyclohexylidene ethanones). These ‘exotic’ chiral systems demonstrate the myriad of structural variations possible, despite some being highly impractical. Nevertheless they are of great interest from a purely scientific stand point; these materials have usually been used to probe experimentally some facet of the relationship between chirality and molecular structure.
1.9 References [1] F. Reinitzer, Monatsh. Chem. 1888, 9,421. [2] L. Lehmann, Z. Phys. Chem. 1889,4,462. [3] I. Sage in Liquid Crystals: Applications and Uses, Vol 3 (Ed.: B. Bahadur), World Scientific, Singapore 1992, Chapter 20. [4] J. W. Goodby, J. Muter. Chem. 1991, 1, 307318. [5] D. G. McDonnell in Thermotropic Liquid Crystals (Ed.: G. W. Gray), John Wiley, Chichester 1987, Chapter 5. [6] D. M. Makow, Colour Research andApplication 1979,4,25-32. [7] J. L. Fergason, Sci. Am. 1964,211, 77-85. [8] J. L. Fergason, Mol. Cryst. 1966, 1 , 293-307. [8] J. L. Fergason, Mol. Cryst. 1966, 1, 309-323. [9] J. L. Fergason, Appl. Optics 1968, 7, 17291737. [lo] D. W. Berreman, T. J. Scheffer, Phys. Rev. Lett. 1970, 25, 557. [ l l ] R. Dreher, G. Meier, Phys. Rev. A 1973, 8, 1616- 1623. [12] A. Saupe, G. Meier, Phys. Rev. A 1983, 27, 2196 -2200. [13] Y. Bouligand, J. Phys. (Paris) 1973, 34, 603614. [14] T. Harada, P. Crooker, Mol. Cryst. Liq. Cryst. 1975,30,79-86. [15] D. Kriierke, H.-S. Kitzerow, G. Heppke, V. Vi11, Ber. Bunsenges. Phys. Chem. 1993, 97, 13711375. [16] M. Langner, K. Praefcke, D. Kriierke, G. Heppke, J. Muter: Chem. 1995,5,693-699. [17] Ian Sage in Thermotropic Liquid Crystals (Ed.: G. W. Gray), John Wiley, Chichester, 1987, Chapter 3.
1.9 References [ 181 D. Coates in Liquid Crystals: Applications an,d
Uses, Vol 1 (Ed.: B. Bahadur), World Scientific, Singapore, 1990, Chap. 3. [I91 D. S. Hulme, E. P. Raynes, K. J. Harrison, J. Chem. Soc., Chem. Comm. 1974,98-99. [20] I. Schroder, Z. Phys. Chem. 1893, / I , 449-465. [21] Liquid Crystal Information Booklet, Merck UK Ltd, Poole, Dorset, UK. [22] J. Constant, D. G. McDonnell, E. P. Raynes, Mol. Cryst. Liq. Cryst. 1987, 144, 161 -168. [ 2 3 ] E. P. Raynes, Electron. Lett. 1973, 9, 118. [24] P. A. Breddels, H. A. van Sprang, J. Bruinink, 1. Appl. Phys. 1987, 62, 1964-1967. [25] D. L. White, G. N. Taylor, J. Appl. Phys. 1974, 45,4718-4723. [26] M. Schadt, Ben Sturgeon Memorial Lecture, The British Liquid Crystal Society Conference. 28-30 March, 1993, Manchester, UK. [27] J. E. Vandegaer in Microencupsulation ProceSse.7 and Applications (Ed.: J. E. Vandegaer). Plenum Press, New York, 1974, Chapter 2. [28] Uncredited article, Polym. Paint. Col. J., 1990, 180 (4255), 118. [29] J. W. Goodby, A. J. Slaney, C. J. Booth, I. Nishiyama, J. D. Vuijk, P. Styring, K. J. Toyne, Mol. Cryst. Liq. Cryst. 1994, 243, 231 -298. [30] R. S . Cahn, C. K. Ingold, V. Prelog, Angew. Chem. Int. Ed. 1966, 5, 385-415. [31] G. W. Gray, D. G. McDonnell, Electron Lett. 1975, 11, 556. 1321 G. W. Gray, D. G. McDonnell, Mol. Cryst. Liq. Cryst. 1977,34. 21 1-217. [33] J. W. Goodby. Science 1986, 231, 350. [34] G. SolladiC, R. G. Zimmermann, Angew. Chem. Int. Ed. Engl. 1984, 348-362. [35] G. Gottarelli, G. P. Spada, Mol. Cryst. Liq. Cryst. 1985,123,377-388. [36] J. March, Advanced Organic Chemistry: Reactions, Mechanisms and Structure, 3rd ed. WileyInterscience, New York, 1985, Chap. 4, pp 107 109. [37] C. J . Booth, D. A. Dunmur, J. W. Goodby, J. S . Kang, K. J. Toyne, J. Muter. Chem. 1994, 4 , 747-759. [3 8 ] G. Friedel, Ann. Phys. (Paris) 1922, 18, 273. [39] F. M. Jaeger, Rec. Trav. Chim. 1906,25, 334. [40] E. M. Barrall 11. R. S . Porter, J. F. Johnson, J. Phys. Chem. 1967, 71, 1224- 1228. [41] G. W. Gray, J . Chem. Soc. 1956, 3733. [42] H. Baessler, M. M. Labes, J. Chem. Phys. 1970, 52,631-637. [43] H. Baessler, P. A. G. Malya, W. R. Nes, M. M. Labes, Mol. Cryst. Liq. Cryst. 1970,6,329-338. [44] F. F. Knapp, H. J. Nicholas, J . Org. Chem. 1968, 33,3995 -3996. [45] F. F. Knapp, H. J. Nicholas, J. P. Schroeder, J . Org. Chem. 1969,34,3328-3331. [46] F. F. Knapp, H. J. Nicholas, Mol. Cryst. Liq. Cryst. 1970, 6, 319-328.
333
[471 D. Demus, H. Zaschke, Fliissige Kristalle in Tabellen /I, VEB Deutscher Verlag fur Grundstoffindustrie, Leipzig, 1984. [481 C. J. Booth, J. P. Hardy, 0. C. Lettington, K. J. Toyne, J. W. Goodby, G. W. Gray, Unpublished results, 1991. [49] K. J. Toyne in Thermotropic Liquid Crystals (Ed.: G. W. Gray), John Wiley, Chichester, 1987, Chapter 2. [50] J. March, Advanced Organic Chemistry: Reactions, Mechanisms and Structure, 3rd ed., Wiley Interscience, New York, 1985, Chapter 1 I , p471. [S11 M. Tsukamoto, T. Ohtsuka, K. Morimoto, Y. Murakawi, Jup. J. Appl. P h y . 1975, 14, 13071312. [S2] Y. Y. Hsu, D. Dolphin, Mol. Cryst. Liq. Cryst. 1977,42,319-325. [53] J . A. Castellano, Ch. S . Oh, McCaffrey, Mol. Cryst. Liq. Cryst. 1974, 27, 419-429. [S4] G. Heppke, H. Marschall, P. Nurnburg, F. Oestreicher, G. Scherowsky, Chem. Ber. 1981, 114,2501 -2518. [ S 5 ] P. Keller, Ann. Phys. (Paris) 1978,3, 139- 143. [56] D. Dolphin, Z. Muljiani, J. Cheng, R. B. Meyer, J . Chem. Phys. 1973,58, 413. [57] W. Helfrich, Ch. S. Oh, Mol. Cryst. Liq. Cryst. 1971,14,289-292. [58] D. Coates, K. J. Harrison, G. W. Gray, Mol. Cryst. Liq. Cryst. 1973, 22, 99- 122. [59] D. Coates, G. W. Gray, Mol. Cryst. Liq. Cryst. 1976, 34, 1 - 5 . [60] E. M. Elashvili, G. S . Chilaya, S . M. Aronishidse, M. I. Brodzeli, K. G. Dshaparidse, Svohshch. Akad. Nuuk. Gruz. SSR 1976,81, 105. [61] Y. Y. Hsu, D. Dolphin, Mol. Cryst. Liq. Cryst. 1977,42,327-340. [62] J. Doucet, P. Keller, A. M. Levelut, P. Porguet, J . Phvs. (Paris) 1978, 39, 548-549. [63] G. W. Gray, K. J. Harrison, J. A. Nash, Electron. Lett. 1973, 9, 130. [64] G. W. Gray, D. G. McDonnell, Mnl. Cryst. Liq. Cryst. 1976,37, 189-21 1. [6S] B. H. Klanderman, T. R. Criswell, J . Am. Chem. SOC.1975,97, 1585-1586. [66] G . W. Gray, D. G. McDonnell, Mol. Cryst. Liq. Cryst. 1978, 48, 37 -52. [67] Y. Y. Hsu, Mol. Cryst. Liq. Crvst. 1977, 42, 263-271. (681 M. A. Tius, X. Gu, J. W. Truesdell, S . Savariar, D. K. Yang, P. P. Crooker, Mol. Crjst. Liq. Cryst. 1987, 150h, 247-255. (691 M. A. Tius, X . Gu, J. W. Truesdell, S. Savariar, P. P. Crooker, Synthesis 1988, 36-40. 1701 P. Le Gras, Personal communication, Zeneca FCMO, Huddersfield, UK, 1988. 1711 C. J. Booth, G. W. Gray, K. J. Toyne, J. P. Hardy, Mol. Cryst. Liq. C y s t . 1992, 210, 3 1 51.
334
1 The Synthesis of Chiral Nematic Liquid Crystals
[72] C. J. Booth, J. W. Goodby, J. P. Hardy, 0. C. Lettington, K. J. Toyne, Liq. Cryst. 1994,16,925 -940. [73] C. J. Booth, J. W. Goodby, J. P. Hardy, 0. C. Lettington, K. J. Toyne, J. Muter. Chem. 1993, 3,821 -832. [74] P. Balkwill, D. Bishop, A. Pearson, I. Sage, Mol. Cryst. Liq. Cryst. 1985, 123, 1- 13. [75] V. Reiffenrath, J. Krause, H. J. Plach, G. Weber, Liq. Cryst. 1989,5, 159- 170. [76] C. Viney, R. J. Twieg, T. P. Russell, L. E. Depero, Liq. Cryst. 1989,5, 1783- 1788. [77] C. Viney, T. P. Russell, L. E. Depero, R. J. Twieg, Mul. Cryst. Liq. Cryst. 1989,168,63 - 82. [78] C. Vauchier, F. Vinet, N. Maiser, Liq. Cryst. 1989,5, 141-151. [79] M. A. Osman, Mol. Cryst. Liq. Cryst. 1985,128, 45-63. [SO] J.-M. Fu, M. J. Sharp, V. Snieckus, Tetrahedron Lett. 1988, 29, 5459-5462. [81] S. Gronowitz, A. B. Hornfeldt, Y. H. Yang, Chemicu Scriptu 1986,26, 3 11- 3 14. [82] G. W. Gray, M. Hird, D. Lacey, K. J. Toyne,Mul. Cryst. Liq. Cryst. 1989,172, 165- 189. [83] G. W. Gray, M. Hird, D. Lacey, K. J. Toyne, J. Chem. Suc., Perkin Trans. 2 1989,2041 -2053. [84] C. J. Booth, J. W. Goodby, J. P. Hardy, 0. C. Lettington, K. J. Toyne, J. Muter. Chem. 1993, 3,935-941. [85] C. J. Booth, I. W. Goodby, J. P. Hardy, K. J. Toyne, Liq. Cryst. 1994, 16, 43-51. [86] C. J. Booth, J. W. Goodby, G. W. Gray, J. P. Hardy, K. J. Toyne, Eur. Put. Appl. 1992,230, 392. [87] K. Terashima, M. Ichihashi, M. Kikuchi, K. Furukawa, T. Inukai, Mol. Cryst. Liq. Cryst. 1986, 141,237-244. [88] T. Inukai, S. Saitoh, H. Inoue, K. Miyazawa, K. Terashima, K. Furukawa, Mol. Cryst. Liq. Cryst. 1986,141,251-266. [89] E. Chin, J. W. Goodby, J. S. Patel, J. M. Geary, T. M. Leslie, Mul. Cryst. Liq. Cryst. 1987, 146, 325 -339. [90] B. Otterholm, C. Alstermark, K. Flatischler, A. Dahlgren, S . T. Lagerwall, K. Sharp, Mol. Cryst. Liq. Cryst. 1987, 146, 189-216. [91] C. Alstermark, M. Nilsson, B. Otterholm, Mul. Cryst. Liq. Cryst. 1987, I50b, 277-287. [92] A. J. Slaney, J. W. Goodby, Liq. Cryst. 1991, 9, 849- 861. [93] A. J. Slaney, I. Nishiyama, P. Styring, J. W. Goodby, J. Muter. Chem. 1992,2, 805-810. [94] P. Styring, J. D. Vuijk, A. J. Slaney, J. W. Goodby, J. Muter Chem. 1993,3, 399-405. [95] T. Kusumoto, T. Ueda, T. Hiyama, S. Takehara, T. Shoji, M. Osawa, T. Kuriyama, K. Nakamura, T. Fujisawa, Chem. Lett. 1990, 523-526. [96] E. J. Corey, R. K. Bakshi, S. Shibata, J. Am. Chem. SOC. 1987,109,5551-5553.
[97] A. C. Griffin, T. R. Britt, J. Am. Chem. SUC. 1981,103,4957-4959. [98] J. W. Emsley, G. R. Luckhurst, G. N. Shilstone, I. Sage, Mul. Cryst. Liq. Cryst. 1984,102,223233. [99] J. L. Hogan, C. T. Imrie, G. R. Luckhurst, Liq. Cryst. 1988, 3, 645 - 650. [loo] G. S. Attard, S. Garnett, C. G. Hickman, C. T. Imrie, L. Taylor, Liq. Cryst. 1990, 7,795 -508. [loll J. Barberi, A. Omenat, J. L. Serrano, Mol. Cryst. Liq. Cryst. 1989, 166, 167-171. [lo21 A. Yoshizawa, I. Nishiyama, J. Muter. Chem. 1994,4,449-456. [lo31 I. Nishiyama, H. Ishizuka, A. Yoshizawa, Ferruelectrics 1993, 147, 193-204. [lo41 I. Nishiyama, A. Yoshizawa, Liq. Cryst. 1994, 17,555-569. [105] I. Nishiyama, M. Saito, A. Yoshizawa, Poster at the 15th ILCC, Budapest, Hungary, 1994. [lo61 M. Marcos, A. Omenat, J. L. Serrano, Liq. Cryst. 1993,13,843-850. [lo71 C. J. Booth, J. P. Hardy, J. W. Goodby, K. J. Toyne, Unpublished results. [lo81 G. Heppke, D. Lotzsch, F. Oestreicher, Z. Nuturforsch. 1987,42a, 279-283. [lo91 R. Ch. Geivandov, I. V. Goncharova, V. V. Titov, Mol. Cryst. Liq. Cryst. 1989, 166, 101103. [110] G. SolladiB, R. Zimmermann, J. Org. Chem. 1985,50,4062- 1068. [I 111 G. SolladiC, R. G. Zimmermann, Angew. Chem., Int. Ed. Engl. 1985,24, 64-65. [112] G. Scherowsky, J. Gay, Liq. Cryst. 1989, 5, 1253- 1258. [I131 G. Scherowsky, J. Gay, N. K. Sharma, Mol. Cryst. Liq. Cryst. 1990, 178, 179-192. [I141 G. Scherowsky, M. Sefkow, Mul. Cryst. Liq. Cryst. 1991,202,207-216. [115] B. F. Bonini, G. Gottarelli, S. Masiero, G. P. Spada, Liq. Cryst. 1993, 13, 13-22. [116] R. Buchecker, Oral Presentation at the 13th ILCC, Vancouver, British Columbia, Canada, 1990. [117] H. Matsutani, K. Sato, T. Kusumoto, T. Hiyama, S . Takehara, H. Takazoe, T. Furukawa, Poster at the 15th ILCC, Budapest, Hungary, 1994. [ 1181 Z. Li, B. M. Fung, R. J. Twieg, K. Betterton, D. M. Walba, R. F. Shao, N. A. Clark, Mol. Cryst. Liq. Cryst. 1991,199, 379-386. [ I 191 S. J. Lock, M. Hird, J. W. Goodby, K. J. Toyne, J. Muter. Chem. 1995,5,2175-2182. [ 1201 Y. Gao, R. M. Hanson, J. M. Klunder, S. Y. KO, H. Masamune, K. B. Sharpless, J. Am. Chem. SOC. 1987,109,5765-5780. [121] K. B. Sharpless, T. Katsuki, J. Am. Chem. SUC. 1980,102,5974-5976.
Handbook ofLiquid Crystals D. Demus, J. Goodby, G. W. Gray, H.-W. Spiess, V. Vill Copyright 0WILEY-VCH Verlag GmbH. 1998
2 Chiral Nematics: Physical Properties and Applications Harry Coles
2.1 Introduction to Chiral Nematics: General Properties The discovery of thermotropic liquid crystals is generally attributed to Reinitzer [ l ] who studied compounds such as cholesteryl benzoate and cholesteryl acetate and noted that the former compound exhibited two melting points and bright iridescent reflection colors between these points. Lehmann [2] coined the phrase “liquid crystal” for such materials as a result of his extensive work on liquid-like substances and there is a lot of interesting debate in the early literature as to who actually discovered liquid crystals [3]. Lehmann certainly constructed specialized polarizing microscopy apparatus which was of fundamental importance in recognizing the selective reflection properties of such materials. It is interesting to note that at much the same time as this discovery research was being carried out on the reflection colors of birds and insects, and that Michelson also noted [4] the selective reflection in ordinary daylight of circularly polarized light from the beetle Plusiotis resplendens. It is now generally thought that
this reflection layer is formed at the late chrysalis stage by a helicoidal liquid crystalline glandular secretion that hardens into a pseudomorph on the surface [ 5 ] . This reflecting layer has a thickness of 5-20 pm, which is typical of the reflecting cholesteric films discussed later. Further, Caveney showed that this reflection was enhanced by a nontwisted half-wave plate layer naturally formed between the two such helicoidal layers [6]. This biological example of enhanced selective reflection serves to show that we have many lessons to learn from Nature which is probably the true inventor of liquid crystals and, as noted in [7], “the farther backward you can look the farther forward you are likely to see”. Studies of the early thermotropic and biological systems have led to much of our present understanding of the general properties of cholesteric liquid crystals. These will be considered herein and we will show how such properties lead to optically linear and nonlinear devices, thermal imagers, radiation detectors, optical filters, light modulators, and other technologically interesting applications in Sec. 2.5 of this Chapter.
336
2 Chiral Nematics: Physical Properties and Applications
Historically, the helicoidal liquid crystals recognized by Reinitzer, Lehmann, Michelson, Caveney et al. were called “cholesterics” after the original cholesterol esters studied by Reinitzer [l]. However, many other steroidal and nonsteroidal compounds exhibit these so-called cholesterical phases (see, for example, [8-121). Thermodynamically, the cholesteric phase is the same as the nematic phase described in the previous sections of this volume and, although both notations are often used, the cholesteric phase can be considered as just the chiral form of the nematic phase. This can be understood by considering a racemic mixture of exactly equal amounts of enantiomeric forms (S and R) of the same nematic molecule. Such a mixture would be achiral, however, a slight excess of one form over the other leads to a chiral nematic phase. It therefore seems more rational to use the notation “chiral nematic” or N*. Thus a chiral nematic phase (N*) is exhibited by neat nematic compounds or mixtures where at least one of the compounds is chiral (i.e., each molecule not superimposable on its mirror image) or where there is an excess of one enantiomer over the other. In this chapter we will not consider in depth chiral nematic systems where only a small percentage of chiral material is added to give a weak helical structure to prevent ‘reverse twist’, as in twisted nematic or supertwisted nemat-
ic devices [ 131. We will primarily be concerned with systems with a high twisting power, such that the N* helicoidal structure formed macroscopically has a pitch length ( p ) of the order of the wavelength of light (A) (see Fig. l), or more generally exhibits a high value of optical rotatory dispersion. For comparison, a typical N* material would give an optical rotation of >lo3 degrees/mm compared with -10 degreedmm for a typical sugar solution. Chiral nematic (or cholesteric) liquid crystals have the orientational order of achiral nematics, but the director (n)is constrained in a layer-like structure to precess around the helical axis perpendicular to n. A ‘layer’ one molecule thick would give a director rotation of -10-20 min of arc per layer. Thus successive ‘layers’, with molecules ‘on average’ preferring to sit at such an angle to each other, twist and describe a helix of pitch p over distances that are large in comparison to the molecular dimensions. Here it should be remembered that such ‘layers’ are composed of molecules in constant thermal agitation and the director only describes a mean macroscopic or continuum picture. Molecules can exchange between near neighbors in the N* phase just as in the N phase. The helix pattern described by the director then dominates the optical properties of the N* phase. Depending on the enantiomeric excess, the materi-
AXIS
Figure 1. Representation of the helicity in chiral nematic structures. The helix defines the z axis and the periodicity in the structure is pi2 due to the condition n =-n.
2.1
Introduction to Chiral Nematics: General Properties
als may have a right-handed (clockwise director rotation) helical structure, and there is a clear relationship between helical sense and molecular structure [ 141. If the director n (a unit vector) describes a right-handed screw along the z axis, in a right-handed coordinate system then the director in the N*< phase is given by
where @ is a constant defining an arbitrary angle with x in the x, y plane and p is the helix pitch, which is positive or negative for right-handed or left-handed materials, respectively. Note that the helix only extends in one direction and that in any one ‘layer’ the molecules appear nematic-like. Further, since n =-n the ‘true’ period of the N*phase corresponds to p12 or a 180” rotation of n . We may also define the helical wave vector k = 2 d p , which then appears in the twist term of the free energy density equation for an N* phase, i.e.,
+ X k2, ( n . V xn + k), +% k3, ( n x V xn),
,,
(2)
Here k, k,,, and k,, are the usual splay, twist, and bend elastic constants of Leslie [lS, 161 and Ericksen [17, 181 continuum theory. This theory will be considered further in Sec. 2.2.2. It is the dependency of k on p , dpld T, and d p l d P that leads to interesting reflection or transmission filters and thermometry devices or temperature and pressure ( P ) sensors, respectively. These will be considered further in Sec. 2.5. If an electric or magnetic field is applied to the system, then an additional term -W is also added to the Eq. ( 2 ) , where W=XArl ~ X ( E. n)2 or %AX ( H n12, respectively. This again has important ramifications
337
which will be considered in Secs. 2.2 and 2.4. The modified free energy density equation refers to static or time-independent properties. If pulsed fields are used, then transient changes lead to switchable bistable states, dielectric and hydrodynamic instabilities and flexoelectric phenomena [ 191, which again may be used in modulation devices. These will be considered in Secs. 2.3 and 2.5, respectively. The optical properties of chiral nematics are remarkable and are determined by the pitch, p , the birefringence, An, and the arrangement of the helicoidal axis relative to the direction and polarization of the incident light. Here we use the term light to indicate that spectral region between the material’s ultraviolet and infrared absorption bands, since this leads, for example, to the possibility of near-infrared modulators. The helical pitch may be much greater or much less than the wavelength &, provided we remain in the limit of high rotatory dispersion discussed briefly earlier. In an external field changes may occur in both the direction of the helix axis, which leads to a textural transition, or in the helical pitch p , which then leads to an untwisting of the helix itself. These effects have been reviewed recently [20, 211 and led to numerous device applications (see Sec. 2.5). The observed optical properties of chiral nematic films depend critically on the direction of the director at the surface interface and on how this propagates to the bulk material. If the director is oriented along the surface of the cell using suitable alignment agents, such as rubbed polyimide, PVA, or PTFE, then the helix axis direction (see Fig. 1) is perpendicular to the substrates, as shown in Fig. 2a. In this case, an optically active transparent planar texture is obtained. It is this texture that is normally used to observe the bright iridescent reflection colors initially observed by Reinitzer and Leh-
338
2 Chiral Nematics: Physical Properties and Applications
mann. If, however, the molecules are oriented normal to the substrate using lecithin, quaternary ammonium surfactants (e.g., HTAB), or silane derivatives, the competition between the helix twist and the surface forces constrains the helix axis parallel to the substrate. This is turn leads to two possible textures, i.e., fingerprint and focal conic (Fig. 2 b, c). In the fingerprint texture the helix axis, z , is aligned uniformly and a side view of the helix is observed. As a result of the director rotation, the refractive index varies in an oscillatory manner and through crossed polarizers this appears as banding reminiscent of a fingerprint. On the other hand, if the helical axis is random in the horizontal plane, we would have a 'polycrystalline' sample which becomes focal conic if the axis of the helix tilts and forms ellipses or hyperbolae. We will discuss these textures further in Sec. 2.2.1. The focal con-
ic texture is highly light-scattering and for initially describing the optical properties of chiral nematics we consider optical phenomena observed with helically twisted planar textures. Under the conditions that the pitch @) is of the order of the wavelength (A) in the medium and that the incident light (Ao)propagates along the helix axis (z), we have two primary characteristic features: a) There is a strong selective reflection of circularly polarized light having the same handedness and wavelength as the pitch inside the chiral nematic medium (see Fig. 3). This corresponds to a snapshot of the E vector of the circularly polarized light matching the helicoidal structure. The reflected light is circularly polarized with the same handedness as the incident light, which is the exact opposite of a normal mirror reflection, which has a 7c phase change on reflec-
+ LINEAR
Figure 3. Schematic diagram of selective reflection from aright-handed chiral nematic planar texture for ;1=&ln inside the liquid crystal. The linearly polarized input light may be considered as counter-propagating right-handed (RH) and left-handed (LH) circular components. The RH component in which the E field matches the sense of the helicoidal structure is back reflected due to director fluctuations, whilst the LH component is almost totally transmitted. By convention, the handedness is defined in terms of the progression of the E field vector in time relative to the observer. The RH rotating wave therefore has the same spatial structure as the chiral nematic at any time.
2.1
Introduction to Chiral Nematics: General Properties
tion. Thus right circularly polarized light would be reflected by a right-handed helix. On the other hand, left circularly polarized light would be almost wholly transmitted through the medium without reflection. Thus if white light is normally incident on such a chiral nematic film, it appears vividly colored on reflection. The effect is best observed, as in thermometry devices, against a black background. The selective circularly polarized reflection occurs around the incident wavelength, &, (in air) as given by ;h=iip
(3)
with a bandwidth A L centered on A, given by
AA =Anp
(4)
where A n = n l ,-nl and E, the mean refractive index, is given by (n,,+ nJ2. Note that here rill is defined as parallel to the director and n1 as perpendicular to it in the same plane, and that the ratio An/& is defined only in terms of the birefringence and refractive indices of the chiral nematic material, i.e., An/;. This is of considerable importance in the design of narrow band optical filters. If p becomes very temperature-sensitive, such as, for example, when the N* phase approaches an SmA* phase on cooling, then since in the latter phase p +00 the wavelength & also diverges towards infinity. This is the basis of many thermometry devices [ 1 I]. At oblique incidence, the reflection band is shifted to shorter wavelengths and side-band harmonics appear [22]. This situation is complex to analyze because of the effect of the twisted refractive index ellipsoid associated with each ‘layer’ on the different polarization components of the incident light. We have in fact to deal with complex elliptically polarized light rather than circularly polarized reflec-
339
tions and their interactions with the refractive index. However, to a good approximation Fergason [22] showed that &, varies for polydomain samples as
where Oi and 8, are the angles of incidence and reflection at the chiral nematic polydomain sample and m is an integer. The equation is only approximate, since the derivation assumes small O,,, and A n . If Oi= Or, i.e., a planar N* sample then
4)= Pmk cos ~
[ (n)l sin
-I
sinO,
(6)
which is of the form of the often quoted ‘Bragg-like’ condition where
m A = p cos $r
(7)
where sin Oi=E sin $,., $r is the internal angle of refraction at the aidliquid crystal interface, and ilis the wavelength in the medium [23]. Experimentally, all orders (m= 1, 2,3, etc.) are observed and again the reflected polarizations are elliptical rather than circular [23-251. We will return to more exact derivations of the selective reflection properties of chiral nematic materials in Sec. 2.2.1 by solving Maxwell’s equations inside the structure. b) There is a strong optical rotatory power for incident wavelengths A,, away from the central reflection maximum at ;lo,subject to the two limiting conditions that Ai >>p or Ai AA/2
2.1
Introduction to Chiral Nematics: General Properties
applies. This equation also shows that the polarization plane rotates in the same sense as the helix for A
p. In the case of A c p (Mauguin's regime), the plane of the polarization actually follows the rotation of the director, and the angle of its rotation on emergence from the layer corresponds exactly to the number of turns of the helix. This is the waveguide regime used in twisted nematic (TN) devices [13], where the helical pitch is much greater than the wavelength of light. In this device the full twist in the director between the upper and lower plates is 90". Thus in the field off state linearly polarized light is rotated through 90" between crossed polarizers to give a transmission state. If the chiral nematic has a positive dielectric anisotropy (i.e., A E= - and E~~and are the dielectric constants parallel and perpendicular to the director at the frequency of an applied field E ) , then the planar texture on the application of E transforms to the homeotropic state and the twist ordering is lost. Thus the incident polarization is no longer rotated in the cell and the transmitted light intensity falls to zero, i.e., extinction between crossed polarizers. Although we will not consider such a 'dilute' regime ( A c p ) further, in great detail, it represents a major applications area of display devices and illustrates another reason why we prefer to call the present materials chiral nematic or N* rather than cholesteric. There is a further geometry of practical interest for light incident on chiral nematic films, related to the pitch of the helix in which we consider light propagating in a direction normal to the helix axis, i.e., as in the fingerprint texture, but with a pitchp less than A. In this short pitch chiral nematic case, the chiral optical tensor is averaged in space and the macroscopic optic axis is collinear with the helix axis [27]. The macro-
34 1
scopic refractive index, nia, for light waves with their polarization orthogonal to the helix axis is an average of the microscopic indices nI1and n I . Since in any one 'layer' nematic liquid crystals are optically positive (nll>n,), the short pitch chiral nematic is macroscopically optically negative, i.e., R ;;"< RY. Electric fields may then be applied to such short pitch systems to deviate the effective optic axis. At low fields this gives a linear electrooptic effect [28] due to flexoelectricity [29], whilst at higher fields dielectric coupling leads to a quadratic effect, i.e., the helix deforms as the pitch increases in the field to unwind completely above a critical field. We will consider these different dielectric and flexoelectric effects in Sec. 2.2.3, and their implications for light modulation devices in Sec. 2.5. In this introduction we have outlined the important physical properties, i.e., optical, elastic, etc., of chiral nematic liquid crystals and indicated the vital role that the pitch plays in determining these macroscopic properties. In the following sections we will present more rigorous derivations of the origins of these properties. Firstly we will consider the static properties, i.e., those exhibited at equilibrium either with or without an applied field (Sec. 2.2). We will then consider time-dependent or dynamic properties, where the systems are in continuous flux or responding to a transient field (Sec. 2.3). As outlined above, both the static and the dynamic properties lead to thermo-optic, electrooptic, magneto-optic (Sec. 2.4), and even opto-optic phenomena in chiral nematics, which in turn lead to devices where the chiral nematic pitch varies from much less than the wavelength of light to much greater than A. These applications will be considered in the Sec. 2.5, where we will also speculate on possible future developments.
342
2 Chiral Nematics: Physical Properties and Applications
2.2 Static Properties of Chiral Nematics In the previous section we discussed the preferential use of the term chiral nematic over cholesteric. At zero pitch the chiral nematic (or cholesteric) becomes nematic. This is true on the molecular scale where the local twist is very small. In the simple 'layer' model, ignoring molecular fluctuations, the twist between layers is of the order of 0.1" for pitch lengths at optical frequencies. Therefore within this limit, and including fluctuations, the cholesteric behaves as a nematic and locally the order parameter definitions are the same for each system. For this reason, the definitions of local birefringence ( A n =q- nl) and dielectric anisotropy (A&= 6, - E ~ remain ) the same. Parallel (11) and perpendicular (I) refer to the local director inside the material. Macroscopically we consider the director, now averaged on a long length scale, to vary smoothly throughout the material, as depicted in Fig. 1. The local order parameter is described as for a nematic, i.e., S = -(3cos2 1 8 - 1) 2
(9)
where 8 is the angle between the long axis of each molecule and the local director. Here we ignore local molecular asymmetry to define the direction of the long axis. In chiral nematics we make the same assumptions, as in nematics, that (1) the centers of mass have no long range order, (2) there is local order as defined in Eq. (9) and therefore the director n has a direction, and (3) that the states n and-n are indistinguishable. For the chiral nematic, the director is no longer arbitrary but has a preferred helical conformation (Fig. l , as defined by Eq. (1)). The helical axis z and @ are arbitrary unless the director is aligned preferentially at
a surface. We will consider how this is done experimentally in the next section. As a result of the n =-n condition, the structure is periodic along z with a spatial variation of p/2, as previously discussed. The helical wave vector k (=2n/p) has magnitude and sign. The magnitude determines the macroscopic physical properties we are interested in herein, whilst the sign defines the handedness of the helix. As pointed out in reference [30] (p. 15) for materials in which k changes sign, the macroscopic properties, such as the specific heats, remain steady across the transition from -k through zero to + k . At k=O the material behaves like a conventional nematic. Further there is no evidence from X-ray scattering of any significant differences between cholesteric and nematic phases save a slight broadening on the cylindrical distribution curve due to reduced parallelism caused by the helical twist [31]. It was Friedel [32] who first noted the similarity between the local molecular arrangements of the so-called cholesteric (or N*)and nematic phases. This is further support for the use of the more general term chiral nematic in preference to cholesteric. For very tight pitch materials (i.e., when p approaches to within 10-100 times the molecular dimensions), the structure is more akin to a layered smectic phase; locally, however, it is still nematic although the scale of the repeat unit of the helix will give X-ray scattering peaks similar to SmA materials, but for a wider layer spacing [33]. These structures are then of considerable interest for flexoelectric and dielectric devices [29]. By definition then a racemic or achiral material leads to a nematic N phase, whilst a nonracemic or chiral material leads to a chiral nematic N*phase. In the preceding discussion chirality was used only to differentiate between righthanded and left-handed structures, that is, structures in which their mirror images can-
2.2
not be superimposed. This is the definition introduced by Lord Kelvin [34]. There is, however, no absolute measure of chirality and there is no obvious way of absolutely predicting how much twist comes from the molecular chirality. Chirality is important herein in that if the individual liquid crystals are chiral (assuming a nonracemic mixture) then there will be a tendency to form a helicoidal structure of the director superimposed on the local nematic order in the nematic (and tilted smectic) phase. As discussed above, the induced helicoidal structure has a defined axis orthogonal to the local director and a defined pitch p . It is the existence of this helix and the variation of p under different conditions that leads to the use of chiral nematics in many different applications and devices (see Sec. 2.5). Before such devices can be understood, it is necessary to consider how the bulk macroscopic properties are determined by this helicoidal structure. In this section we will consider the static properties, that is, those properties as observed at equilibrium in the presence of alignment forces, applied electric or magnetic fields, and at different temperatures and pressures. We have used the heading static to indicate, under any given conditions, that we are not considering transient changes between a ‘ground’ and an ‘excited’ state, but the conditions in either end state. Time-dependent phenomena will be considered in Sec. 2.3 under the general heading of dynamic properties. The static properties important in the following section will relate to how the helicoidal pitch is a) measured and b) influences the optical properties and leads to different textures and defects. We will then examine in detail the theoretical approach to optical propagation, Bragg reflection, and transmission for normal and nonnormal incidence in chiral nematic materials. We will outline how the
Static Properties of Chiral Nematics
343
bulk elastic properties are modified by the helicoidal structure by using the continuum theory to deduce a free energy density expression both for the quiescent state and in the presence of external fields. This will then allow us to determine the helicoidal pitch behavior in the presence of external influences such as temperature, pressure, and electric or magnetic fields. Finally, we will describe how these external fields lead to different dielectric, diamagnetic, and flexoelectric phenomena in chiral nematics.
2.2.1 Optical Properties As described earlier, the spectacular optical properties of chiral nematics are determined by the helicoidal pitch, the birefringence (and refractive index), the direction and polarization of the incident light, and the arrangement of the helix axis. For normally incident light the direction of the helix axis gives rise to the three classical textures depicted in Fig. 2, and typical photomicrographs taken with crossed polarizers are shown in Fig. 6; these are: a) The planar or Grandjean texture, in which the helix axis is uniformly orthogonal to the confining glass plates. The surface alignment is often induced by rubbed polymer films. b) The focal conic texture, in which chiral nematic domains are formed with the helix axes arranged in different directions. These may be achieved without surface treatment on rapid cooling from the isotropic phase. c) The fingerprint texture, in which the helix axes are parallel to the glass plates. Surfactants are often used to produce this texture, which may be uniform or polycrystalline depending on the degree of alignment control at the surface. The
344
2 Chiral Nematics: Physical Properties and Applications
Figure 6. Photomicrographs of (a) planar, (b) focal conic, and (c) fingerprint textures in chiral nematics, and the schlieren texture demonstrating brushes in an apolar nematic (d); observations through crossed polarizers.
2.2 Static Properties of Chiral Nematics
name fingerprint arises, in the case of long pitch materials (i.e., p A) from the stripes observed as the helix pitch rotates through optically equivalent states with a period of p/2. IN.B. In short pitch mate-. rials ( p a i l ) , the pattern is not resolved and often takes on a fairly uniform colored appearance depending on the cell thickness, A n macroscopic (i.e., ni"-n;I"), and the direction of the input polarization.]
-
All of these textures may be switched to produce electro-, magneto, or flexo-electric effects, which will be discussed at the end of this section. The key to the use of these various effects and indeed observation of the optical properties lies in control of the director orientation at the liquid crystal-substrate interface. For the techniques to be described, we will be in the strong anchoring regime where we can ignore the k , , and k,, terms in the full free energy density expansion (see Sec. 2.2.2.1). As a result of the local similarity between nematic and chiral nematic phases, most of the alignment techniques that work for nematics [3S] work for chiral nematics. We are principally interested in two types of alignment of molecules at the surface, i.e., planar and homeotropic, in which the preferred orientation of the molecules pinned to the surface is parallel or perpendicular to the surface. Planar alignment in chiral nematics can normally be achieved by three well-known techniques: (1) SiO evaporation, (2) polymer rubbing, or (3) microgrooved surfaces. In (1), silicon monoxide ( S O ) is evaporated onto a surface to give a layer, typically 20-100 nm thick, at an oblique angle 8. For 8=60" the molecules align parallel to the evaporation plane, whilst for 0=80" they are perpendicular [36, 371. If random planar alignment is required, the SiO is evaporated at normal incidence. SiO gives a surface of
34s
very high thermal stability. In (2), i.e., polymer rubbing techniques [35, 381, a polymer such as PVA (polyvinyl alcohol), which is water soluble, or polyimide is spin-coated onto the substrate and then rubbed with a spinning velvet cloth in a unique direction. In these cases the molecules are aligned in the rubbing direction and, if necessary, speed and pressure may be used to control a small 1-5" surface pretilt. The commercially available polyimides have good thermal and chemical stability, whilst PVA is convenient for rapid testing in the laboratory. If a barrier layer is required for electrooptic applications to prevent ionic flow from the electrode layers, it is possible to evaporate the SiO normally and then spin-coat on top the polymer prior to rubbing, in order to obtain good planar alignment. A new polymer alignment technique [39] using frictiondeposited polytetrafluoroethylene (PTFE) gives excellent planar alignment of nematic and chiral nematic materials with almost zero surface tilt [40]. In ( 3 ) , as will be discussed in the thermochromic devices section (Sec. 2.5.2) fine microgrooved surfaces may be used in embossed laminated sheets [41] to give excellent planar alignment. Homeotropic alignment is usually obtained by coating the substrate surfaces with surfactants such as lecithin or HTAB [42, 431. A practical laboratory method is to deposit egg lecithin in ethanol or chloroform on the surface. The solvent evaporates to leave the surfactant with its polar head group stuck to the surface and the hydrocarbon chains pointed, on average, into the cell perpendicular to the substrates. The liquid crystal-hydrocarbon chain interaction then gives the desired homeotropic alignment. Via the elastic forces, the surface alignment then propagates into the bulk volume. Silane derivatives [44] or chromium complexes [4S] that chemically bind to glass surfaces have equally been used for homeotropic alignment.
346
2 Chiral Nematics: Physical Properties and Applications
The different optical textures induced by the different surface alignment conditions lead us to three common methods of measuring the helical pitch:
1. Fingerprint texture: Using homeotropic alignment and a polarizing microscope we can measure the distance between adjacent dark lines (see Fig. 6) provided the sample is illuminated with monochromatic light. As discussed earlier, the line separation is thenp/2. Since this is a visual technique, it is only easy to observe pitch lengths greater than -1.5 pm. If a polarized laser beam is used and the fingerprint texture treated as a diffracting (grating) element then this technique may be extended down to the diffraction limit to measure pitch lengths -0.6 pm, depending on the wavelength used. 2. Planar texture: Using planar alignment and by measuring the reflection or transmission spectrum of a chiral nematic
film it is possible to measure the Lax (re(transmitted) using a flected) or kin monochromator either to control the input light wavelength [ 101or using a white input light to monitor the reflected or transmitted spectrum. In the latter case we [46] have found it particularly useful to couple a simple single pass or spinning prism monochromator to a reflection microscope to measure such spectra for a large variety of materials. This is shown schematically in Fig. 7. This technique allows the spectrum to be monitored on an oscillocope instantaneously and allows pitch measurements in the range 0.2 p m < p < 1.O pm. The pitch p is related to &, by p=&,-n (see Eq. 3), where Z= (rill + n,)/2. This technique necessitates an independent measurement of nll and nI. Since these are normally measured for order parameter measurements using an Abb6 refractometer, this is no great problem. Simple diffracting ele-
Figure 7. Schematic diagram of a microscope spectrometer for measuring transmitted or reflected spectra from chiral nematic textures.
2.2
ments coupled to CCD arrays now make the detection simple and fast so that dynamic phenomena may be monitored [47]. A related technique for spectral measurement using phase-sensitive detection has also been presented [48]. 3. Focal-conic texture: Based on the diffraction equation derived by Fergason [22] for such a polydomain texture (Eq. 5 ) , e.g., HeNe laser at A=632.8 nm, the helix pitch can readily be determined. Again Z has to be determined independently, but the technique may be used for dynamic electrooptic studies. This method is limited to materials of small An,but the range of p readily measured is 1 p m < p < 2 0 pm. Supplemental to these techniques, it is also possible to use a wedge-shaped cell to measure the helical pitch in a chiral nematic [49, 501. This is a further example of the use of planar boundary conditions. In a wedge cell, with the angle a=h/L previously determined, the planar alignment produces the classical Grandjean-Cano texture with disclinations separated by a distance 1 (see Fig. 8). These disclinations arise since as the cell thickness increases the number of ‘half pitches’ through the cell also inreases to minimize the free energy. As a result of the alignment conditions, it can only do this in a quantized way. The increase thus takes
Static Properties of Chiral Nematics
347
place in steps where disclination lines are created every time the number of half pitches is increased by one (see Fig. 9). The pitch p is then given by p = 21a =
~
21h L
where I is the distance between two disclination lines and a is the wedge angle determined by interference fringes in an empty cell. Using this technique, pitch measurements can be made readily in the range 0.8 p m
afew microm-. eters). For shorter pitch systems, in which( case the fingerprint may be too fine to observe directly (other than by diffraction), an electric field [below the critical helix unwinding (see Sec. 2.2.3)] may be used in conjunction with the shearing process to unidirectionally align the helix for materials of positive dielectric anisotropy. This is particularly useful for chiral nematics used in flexoelectric switching [29]. Pate1 and Meyer [28] showed that rubbed polyimide layers could be similarly used (N. B. without shearing) and that, on cooling from the isotropic phase in the presence of an AC field, the helix could again be uniformly aligned in the plane of the substrates. Although, as discussed in Sec. 2.2.1, the fingerprint texture may be used to measure the helical pitch, care has to be taken lest the helix axis is inclined to the substrate plane or local distortions occur at the substratel liquid crystal interface [73]. An interesting use of the fingerprint texture is to examine pitch changes with concentration in a contact preparation between a chiral and an achiral nematic. As the pitch varies from infinity in the nematic to the value of p in the neat chiral compound, the fingerprints become closer together. This effect is shown in Fig. 11b for a very short pitch cholesta-
Static Properties of Chiral Nematics
353
nol-based organo-siloxane [74] in contact with pentylcyanobiphenyl (5CB). We will now consider in more detail some of the alignment or director field patterns around different defect structures in chiral nematics. Using the simple one elastic constant approximation (i.e., k as for the nematic case above) and the definition of the chiral director (i.e., n =(cos 6, sin 6, 0), 6 = k z , and @=O; see Eq. (1)) in the free energy density expression, (Eq. 2) gives
and V26=0. For X-screw disclinations (see the nematic wedge disclinations discussed earlier) there is a singular line along the z axis (i.e., parallel to the helix axis) and the director pattern is now given by
where m is integral as before. The simplified director patterns for s=X and 1 are shown in Fig. 12. If the elastic anisotropy is now included in the free energy density (see, for example, [59], p. 139 et seq.), then one
Optic Axis
/@//a/
/ s/I/
41t
/
/ e l f [a / /’ S=I/2
s=1
Figure 12. Representation of director patterns for s= X and 1 X-disclinations in a chiral nematic. The unfilled or filled dumbbells or pins denote the spiraling director.
354
2 Chiral Nematics: Physical Properties and Applications
particular value of c is favored for a disclination pair. This is important for chiral nematics where each layer may be regarded as nematic-like, but rotated about the twist axis. Since each layer is allowed only one value of c (0 or d 2 ) (see Fig. lo), the disclination pairs are expected, theoretically, to adopt a helicoidal configuration and this has been observed experimentally for s=% [75] and s= 1 [76] pair disclinations. For edge disclinations, the singular line disclination is perpendicular to the twist axis. On rotation around this line, an integral number of half pitches ( p / 2 ) are gained or lost and the pattern for s=% is as shown schematically in Fig. 13. The director pattern around such a disclination was first proposed by de Gennes [77], again in the one elastic constant approximation, with the following twist disclination type of solution
x
n = (cos 8,sin 8,o)
1
and m is again an integer. This approach has been extended to take into account elastic anisotropy by Caroli and Dubois-Violette 1781.
----c----OQ~cO----OCg.----OQCg
.............................. -*o--..o-.c+.o0
cs
-0-
4--cg
~rg-J-o-o-o.-o-o
cc
..........................
+-0.----0cOcO--o.0.c-c-c-c--mo-
...........................
h
L-c--c+-c-c-+
-----c------
@ - - - - a ~ ~ ~ o - - - - o o Figure 13. The director pattern for s = % X-edge disclinations in a chiral nematic. The filled in dots signify that the director is orthogonal to the plane of the figure, the dumbbells that it is in the plane, and the pins that it is tilted in a spiraling structure.
As for achiral disclinations, the Volterra process may be used to create screw or edge disclinations by cutting parallel or perpendicular, respectively, to the chiral nematic twist axis. As discussed when considering the focal conic texture, the chiral nematic may be considered as having a layered structure of period p / 2 for a given direction of n. This layered lattice can then have disclinations similar to SmA materials. In the Volterra process, if the cut is made such that the disclination line (L) is along the local molecular axis then the longitudinal (A) disclinations created are designated A+ (indicating material has to be removed) or d- (for material added) to arrive at the final configuration. If, however, L is perpendicular to the local molecular axis then these transverse (7) disclinations are similarly denoted z+ and z-. These configurations are shown schematically in Fig. 14, and it should be noted [59] that the coreless d disclinations have lower energies than the cored z structures in a chiral nematic. Also the and z(') disclinations occur in pairs to give five possible forms of dislocations and pincements or pinches (i.e., a transformation from a wall to a line) (see Fig. 15). These disclinations are most easily observed in the fingerprint texture (see Fig. 6 c), where many of these defects may be readily identified by the observant reader. As a result of the layered nature of the chiral nematic structure, like the smectic A, it can also exhibit focal-conic textures [79] and both phases exhibit screw and edge dislocations. A dislocation corresponds to a displacement of the layered structure in a - plane - o orthogonal to the layer and may be formed by the pairing of two disclinations of opposite sign. A screw dislocation has a singular line along the screw axis and is equivalent to ax-screw disclination in a chiral nematic. An edge dislocation corre-
2.2
2j.i .. 8 ::1
0-20-40-0~
................... =:'B ..................... b =..\
G-Qu-oo--o D-0C-c
L p ooo-*
q... %:;%
c+-oc-0
d........... -0 +---7F
..... o-. -6"
'
. . . .- . .0. . % .
t--S .......... -0
Y ...........
sponds to a line defect in the y direction for a planar (x, y ) chiral nematic with its helix axis in the z direction. This line defect then forces the layer (or planes of equivalent n ) to tilt with respect to the I or unperturbed helix axis. Schematic diagrams of such singular defects are shown in Fig. 16 a,b. These edge dislocations under favorable energetic conditions can combine (Fig. 16c,d) to form a 'grain boundary' with zig-zag and quadrilateral defects [79] in chiral nematics. These are clearly evident in Fig. 17 from
Static Properties of Chiral Nematics
355
Figure 14. The director configurations around defects leading to (a) X', (b) A+, (c) z-,and (d) r+ disclinations in a chiral nematic. The dumbbells, pins, and dots have the same significance as in Fig. 13 and the black star circles L represent the disclination lines.
Figure 15. Schematic representations of the pairing of iland z disclinations of opposite signs in a chiral nematic. The schematic diagrams show edge disclinations comprised of (a) il-and A+, (b) z-and T+, (c) r- and A+, (d) X and r+ and (e) pinches or pincements of z+ and 5-. The pins, dots, and layer lines represent the spiraling director field.
[80], where the dark banding is due to the layers of equivalent n in the chiral structure. The texture also shows in-plane linear and hyperbolic defects generated by coalescing edge dislocations in the z direction. These are similar to the focal conics formed in smectic A materials, but in the chiral nematics the length scale of the layers of equivalent n are separated by pI2, which is the order of micrometers rather than nanometers, i.e., it is a macroscopic rather than a microscopic feature.
356
2 Chiral Nematics: Physical Properties and Applications
I
(c)
(4
Figure 17. Optical micrograph of a chiral nematic texture exhibiting zig-zag and quadrilateral defects.
2.2.1.2 Optical Propagation (Wave Equation Approach) The large and anomalous optical rotations observed in the chiral nematic phase are not due to intrinsic spectroscopic properties, i.e., absorption or emission, of the constituent molecules, since they do not persist in the true isotropic phase (i.e., away from pre-
Figure 16. Simplified schematic diagram of (a) and (b) singular defects which combine to form grain boundaries with (c) zig-zag and (d) quadrilateral defects.
transitional effects [8 11). They must therefore be due to the transmission and reflection properties of light propagating in the helicoidal or twisted optically anisotropic liquid crystalline matrix. Based on the spiraling structure depicted in Fig. 1, this propagation has been examined theoretically in the seminal contributions of Mauguin [60], Oseen [61], and de Vries [26]. More recently Kats [62] and Nityananda [63] presented a highly accessible theoretical treatment of this work, and this will be considered below. This approach, based on the spiraling ellipsoid model for the dielectric tensor, gives ready access to the two experimentally interesting conditions of a) waveguiding in the Mauguin limit where p > >iland b) selective optical reflection when p=A. Solution of Maxwell’s equations in this formalism leads, not only to the conditions for strong reflection of circularly polarized light and anomalous optical rotatory dispersion, which are of interest in this chapter, but also to the treatment of absorbing chiral
2.2
nematic systems, e.g., dye guest-host [82]. In the present treatment we assume (1) am ideal helix (i.e., the director of n rotates regularly and sinusoidally, (2) a semi-infinite planar structure bounded on its upper surface by an isotropic dielectric medium of the same mean refractive index as that of the chiral nematic [i.e., E=X(nll+n,)l, (3) that the birefringence is small (An=nll-n, A , the material behaves macroscopically as a variable birefringence film with A n (z) varying as a function of displacement of the orthogonal light beam relative to z, the helix axis. For p e a , the light wave interacts with many turns of the helix and the polarization experiences an average refractive index in, say, the x, y plane. In this case, if we define nfA as the macroscopic refractive index parallel to the helix axis (z) and n;OA as orthogonal, then the macroscopic birefringence is nfA-nroA. For this situation, nfA =nI in the local director n notation. The macroscopic orthogonal component is then given by (37) averaged over one pitch E(Z)= n2 (z>
where rill is parallel to the microscopic director n and e ( ~is) the in-plane (x,y ) angle between nI1(z=O) or y and n (z). Thus a microscopic positive birefringence material (A n =nI1-nl> 0) will become macroscopically negative, since nfA=nl il. In the next section we will discuss the practical Bragg diffraction properties of chiral nematic films in more detail.
2.2.1.3 Optical Propagation (‘Bragg’ Reflection Approach) In the previous section we considered the propagation of light in a chiral nematic material by solving Maxwell’s equations under the limiting conditions of (a) a planar sample, (b) a perfect helicoidal structure with its axis orthogonal to the liquid crystal layer, and (c) for A e p - d a n d A-p. For A-p we showed that circularly polarized light is reflected in a wavelength range defined by & + p A n / 2 . This formalism is the only exact way of describing the reflective, refractive, and transmissive properties of light interacting with a chiral nematic. In the present section we will use a ‘Bragg’ diffraction approach to describe these properties in the practical situation where the chiral nematic material is polydomain in nature, and where these domains may be inclined to the direction of the incident light. The term Bragg reflection is used throughout the literature, but it must be stressed that this is only a convenience. True Bragg diffraction occurs when the radiation incident on a crystal lattice has the same order of wavelength as the repeat unit of the lattice, and where the density waves of the scatter-
Static Properties of Chiral Nematics
2.2
ing medium are discrete. In chiral nematics,, the material is in a continuum and it is the ‘local’ director that gives a layer-like structure as it precesses around the helix axis (2). Fluctuations ensure that the density wave is constant with respect to z along the axis and on the length scale of visible light. In the Bragg reflection approach, as in the figures that follow, the layers used to define the scattering or reflection geometries refer to planes of equivalent optical properties separated by d , where the local directors are orientationally aligned in the same direction with n=-n. There is a continuum of identical material, orientationally rotated ’in plane’, with respect to and in between these layers, separated byp/2 (see Eq. 1) and Anp defines the bandwidth of specularly reflected light at a wavelength A(=&/$ in the continuum. The simplest case of Bragg diffraction is shown in Fig. 20 a for an ideal planar chiral nematic. If k , is the scattering vector, arising from the structure in, and local fluctuation of, ~ ( r )the , dielectric tensor, then
k , = k1- k ,
363
and scattered light, respectively. A more exact treatment is given in Chap. 3 of [30]. If 19~=$~=0, then for ki=-k,=k, the only allowed nontrivial solution, we have
k,=2ik or
2d=A
(41)
which is the simple ‘Bragg’ condition. As discussed in the previous section, if k>O (i.e., a right-handed helix), then right circularly polarized light is back scattered or reflected whilst the left-handed circularly polarized light is transmitted. Secondly, there is only one order of back-reflected light. For Oi > $i= &, Eq. (40) gives solutions corresponding to
2d cos $i,s = rn A or piicos$=rn&
(42)
where rn is an integer and A=&/Z. In this case [84] higher orders are allowed and the reflected light is elliptically polarized. As discussed, when solving the wave equation, the propagation of light at oblique angles is complex, since the light polarization couples neither with nIl nor nI but with values
(40)
where ki and k , are the wave vectors, which are assumed to be coplanar with the incident
I
(a)
(b)
Figure 20. Geometry for selective reflection for plainar chiral nematic films for director layers (a) perfectly aligned for internal scattering and (b) inclined at an angle a to the surface interface. All rays are assumed to be in the plane of the figures.
364
2
Chiral Nematics: Physical Properties and Applications
of n given by an angled slice across the refractive index ellipsoid. For the following discussion we will ignore these complex polarization properties of the reflected light and consider only the wavelength behavior. Unless we are considering some of the specialized reflecting films to be discussed in Sec. 2.5 of this Chapter, from a practical point of view we are only interested in intensities and wavelength. One interesting consequence of Eq. (42) is that if the monodomain is illuminated with white light and reflects back normally, say, red wavelengths, then observation at angles of 8 (or Cp internal) leads to a blue shift and this is observed in practice. Indeed, this is used in security devices, since it is an effect that is impossible to photocopy. The angular dependence of the selective reflection is one of the key features of chiral nematic thin films. In the practical world we rarely have perfect semi-infinite planar samples. Surface imperfections, preparative techniques, etc., normally lead to polydomain samples in which the planar texture is locally inclined and Cpi#Cps. This is depicted in Fig. 20b. ' O oriented . There will be many polydomains with their helix axis at different angles (a) to the substrate or liquid crystal/air interface, and each domain will selectively reflect light at some specular angle [22]. In the original model, not so far from reality in microencapsulated or polymer-dispersed chiral nematic films, Fergason assumed fictitious chiral nematic domains embedded in a medium of mean refractive index H,,, the helix has infinite pitch.
2.4
387
Field-Induced Distortions in Chiral Nematics
= L j [ [ s - k )2k 2 2 - A ~ H 2
+ const. where, as before, k=27clp, the undistorted helix wave vector. The equilibrium condition is then 10 0
I 0
H/Hm
Figure 34. Magnetic field dependence of the relative pitch p (H)/p(0) in a chiral nematic. These data come from [ 1371 and correspond to PAA doped with cholesteryl acetate.
H < H C N , with H in the x direction [i.e., H = ( H , 0, O)], the directors will deform to distort the helix, as shown in Fig. 33b. These directors originally pointing (on average) with a greater anisotropy component in the x direction will deform to be parallel to H , i.e., in the x direction (at B), having a tightly twisted region at the ‘points’ or positions (A) with the director along y , i.e., orthogonal to H . These correspond to a succession of 180” twist walls. With further increasing field, the regions or ‘walls’ such as B become unwound so that the directors realign along the field and the areas A on average expand, thereby increasing the effective pitch. The helix becomes fully unwound for H>H,. Mathematically this is described as follows. The field H is along z (i.e., H = h , 0, 0) and the director (see Eq. 1) is n = (cos v, sin v,0), where y = k z + $. For thick cells in which d>>p, $ is a constant defined by the surface alignment and can be ignored. We will consider pitch deformations ‘far away’ from the substrate. The free energy density of the system is then given by
k22(
d2W - A X H 2 s i n t y c o s v = 0 g)
(127)
which gives
z = 2a5,F, (a) for
and R12
dv
Fi(a)= j (I-a 2 sin . 2 y)Il2 In this notation, a is a constant determined from the condition that 3 is a minimum, z = p (H)/2, i.e., the half pitch of the field distorted structure or the distance between twist walls, F , ( a ) is a complete integral of the first kind, and 5, relates to the wall ‘thickness’, which is of the order of 2 t 2 ( H ) [30, 1391. The free energy is then given by
1
(q+p)
kz
kz
where
a2 = (I + p)-’ and h = (t2k)-’
(128)
388
2 Chiral Nematics: Physical Properties and Applications
and A12
fi (a)=
5 (1- a2sin2yl)1’2dyl 0
Here F2 (a)is an elliptic integral of the second kind. Minimizing the free energy as in the previous sections for A p= 0 leads to
For z(H=O)=n/p(O), i.e., the undisturbed helix, and z (H)=7t/p (H) we get
If z(H) +m, then a+ 1, F 2 ( a ) and for H=HCN,i.e., the helix unwinding field, we have h =7t/2 or -03,
where p=p(O), as defined in Eq. (126). Experimentally it has been verified that HCNis inversely proportional to p [ 1401 for relatively long pitch systems and that Eq. (131) is experimentally valid [137, 1401. It should also be noted that the pitch divergence only occurs close to HCNas the number of pitch walls rapidly decreases to zero. A further interesting observation is that for low field H dichroism (where 5 is the absorption coefficient parallel or perpendicular to n) may be used to give negative [259] or positive [260-2631 (i.e., dark symbols on a light background) optical contrast. In these devices the input polarization is the same as the alignment direction at the 'input' planar alignment layer.
2.5.3.2 Intermediate Pitch Length Systems ( p -A)
-
Intermediate pitch systems ( p A) give rise to focal-conic and planar or Grandjean textures (see Sec. 2.2.1.1), and these may be deformed in electric of magnetic fields (see Sec. 2.4). The earliest reports of electric field induced reorientation leading to electrooptic effcts are reported for scattering or focal-conic textures in [136, 1381 and for planar textures in [ 141,2641. Similar effects in magnetic fields were reported in [129, 137, 1401. The different optical effects observed at the time are accounted for by the direction of the applied field relative to the helix axis (see Sec. 2.4). As was discussed in Sec. 2.2.1.2,, for a chiral nematic phase exhibiting selective reflection, the ORD becomes anomalously large and therefore, if the helix pitch can be unwound or deformed in an electric field, the circular polarization and ORD should equally change. The latter was clearly observed and studied in some depth in [265]. These early observations set the scene for considering electrooptic devices in intermediate pitch ( p 2) chiral nematics. The focal-conic texture is inherently light-scattering in the forward direction [266] and this in itself does not provide sufficient optical contrast between 'off' and 'on' states when switched to the homeotropic texture, without the use of external polarizers. The planar to homeotropic field in-
-
402
2 Chiral Nematics: Physical Properties and Applications
Sec. 2.4.4) unwinds the helix to give a homeotropic state. The axis of the dye is then in the field direction and the device becomes nonabsorbing, assuming positively dichroic dyes (see Fig. 38a). In a second variant of this device, planar orientation is used to spiral the axis of the dye, following n , in the plane of the device and for zero fields this leads to a colored absorbing state. For such systems, where the pitch is too short to give light polarization guiding (see the Mauguin limit, Eq. (166)), the polarization becomes elliptical and this is readily absorbed by the dye (Fig. 38 b). For a high enough field the transition to homeotropic is again induced, and this leads to a nonabsorbing (or weakly absorbing, due to limiting 6)or clear state. The threshold voltage for helix unwinding is given by
duced transition does provide contrast with a suitable black background, but on field removal tends to relax back to a focal-conic rather than a planar texture in a fairly slow dynamic process. There have, however, been recent developments in which these effects have been utilized or improved upon to produce interesting displays. The first was the White-Taylor device [267] using anisotropic dyes and either homeotropic or planar surface alignment. The second is in using polymer-stabilized chiral nematic films to exploit the bistable nature of the planar and focal-conic textures [268]. In the White-Taylor device the chiral nematic ( p A) is doped with an anisotropic dichroic dye. With homeotropic boundary conditions and low voltages, the focal-conic texture becomes axially aligned in the plane of the device. The dye spirals with the director and the random directions of the helix axis in the plane of the device ensure that unpolarized light is absorbed uniformly in this state. Application of a high field (see
-
Input White Light c_ ____ j---____ mPolariser Glass Substrate .Transparent Electrode Homeotropic Alignment Layer Chiral Nematic (Focal Conic)
-
and for d - 10 pm and p 3 pm gives typical threshold voltages (to complete the un-
Input White Light
Input White Light
-_-I
output Coloured Light
Glass Substrate Transparent Electrode Planar Alignment Layr Chiral Nematic (Grandjean) outout Coloured Light
Output White Light
0 : LiquidCrystal With Little Attenuation Molecule
I (a) Unenergized State Focal Conic
(b) Unenergized State Grandjean
: DyeMolecule
(c) Energized State (Homeotropic)
Figure 38. Schematic operation of the White-Taylor dye guest-host chiral nematic electrooptic cell. In (a) for zero applied field the axis of each focal-conic domain is random in the x, y plane, as therefore is the dye, using homeotropic surface alignment. In (b) the texture is planar for the zero field state and therefore the dye spirals around the z direction. In (c) the focal conic (a) or planar (b) transition to homeotropic nematic has taken place above the threshold voltage V,, (WT). The black ellipses represent the dyes in the chiral nematic matrix.
2.5
winding) of the order of 1OV. These devices are polarizer-free and have better viewing angles than TN devices, although V,, (TN) < V,, (WT) (see Eq. 170). They can also be matrix-addressed to a limited extent, and the performance of the two modes has been compared in references [269]. Negamay also be used to protive dyes duce reverse contrast. It is not, however, possible to decrease the pitch ad infinitum in these devices because of the dependence of V,,(WT) on dlp. A recent innovation in the use of chiral nematic phases has been to stabilize the focal-conic or planar textures in situ with a polymer network [268]. The helix pitch may be chosen to give bright specular reflection, and that can then be switched into a focalconic (forward-scattering) texture. Using a black background, this has led to bright, high resolution colored images with high contrast and wide viewing angles, and without the need for backlights. Devices have been fabricated using flexible polymer substrates to make lightweight displays with a 320 x 320 pixel writable-erasable screen. Applications are foreseen in electronic publishing and, using a stylus, it is possible to hand write erasable information on these displays. As with the White-Taylor device, dyes can be incorporated into the polymerstabilized chiral nematic device [270] to improve contrast between the on and off states, and it is anticipated that such dyed PDN*LC devices will find use in projection displays. The use of flexible substrates without backlighting, combined with the specular reflection properties (in unpolarized light) of chiral nematics that may be electrically switched from one bistable state to another, seems to hold great promise in the search for the Holy Grail of an electronic newspaper (2711. In the latter reference, reflective chiral nematic liquid crystals were used to produce a 200 dpi full page, passive matrix,
(tl> t,,)
Applications of Chiral Nematics
403
bistable glass display with a 2240 row by 1728 column resolution. Whilst these are slow update displays, taking several seconds, they would seem ideally suited to this application. A further interesting use of the focal-conic to homeotropic texture transition is in infrared modulation [272]. Here it was found possible to modulate infrared light at A= 8 - 12 pm with a maximum transmission of 8796, a contrast of 93%, and turn on and off times of 1 ms and 125 ms, respectively. A further window examined was 3-5 pm, and this work suggests that other chiral nematic electrooptic effects could be exploited in the near infrared. In communications technology a 2x2 optical switch for fiberoptics has been developed [273] using a chiral nematic film and two switchable nematic waveplates. It has been demonstrated that this is suitable for LED or laser sources. The device worked at 1.3 18 pm and had switching times of 40 ms with -26 dB crosstalk between unselected fibers. There will clearly be further advances in this use of the unique optical properites of chiral nematics.
2.5.3.3 Short Pitch Systems ( p 2)/21
(1 1)
Here qo = 27c/Z0, the wave vector related to the layer thickness. Unlike (u2(r)),this quantity converges. Rather than a Bragg peak like a delta junction for a crystal, the calculated X-ray peaks should exhibit the following profile:
~ ( q=) l/[q,
qx = q y = 0 (12a)
- q0l2-" 2 2-11
I ( q ) cc 1/[q,2+qy1
9
q2=0
(12b)
where
17
= q i k B T/[8Z(KB)'/2]
(13)
This theoretical prediction has been confirmed by detailed X-ray diffraction studies which also enable accurate measurements of [ 111. Figure 2 displays the X-ray diffraction intensity Z(q,) as a function of the scattering wave vector qz obtained at two temperatures in the SmA phase of 8 0 C B [4'-n-octyloxy-4-cyanobiphenyl].The quasi-long-range smectic layer order gives the
2.2
I I
deviation of the experimental data from the dashed line which represents the scattering profile for a system with true long-range order. The best fits of the data to Eq. (12) convoluted with the resolution function are shown as two solid lines. The values of 77 determined from the best fits agree with the experimentally determined values of qo, K , and B. This absence of true long-range smectic order is a result of the simultaneous existence in the free energy of a solid-like elastic term for wave vectors perpendicular to the layers and a nematic term for wave vectors parallel to them. Such a system exhibiting quasi long-range order in three dimensions is at the lower marginal dimensionality below which long-range order no longer exists.
445
Smectic A Phase
Previous discussions demonstrate the importance of the compressional elastic constant B in characterizing the SmA phase. Moreover, the critical behavior of B in the immedite vicinity of the nematic (N)-SmA transition remains an important and unresolved issue. To obtain high resolution measurements of B , several experiments have been designed, (e.g. second-sound resonance [ 121 and the line width of Rayleigh scattering [13]). Well into the SmA phase, the typical value of B is about lo8 dyn/cm2. The most recent sets of experimental data resemble a power-law approach to a finite value at a continuous N-SmA transition. Figure 3 displays the compressional elastic contant B versus temperature just below the N - SmA transition temperature of SOCB. The data were obtained from the line width of Rayleigh scattering measurements [ 13). Whether the value of B approaches a finite value or zero at the SmA - N transition remains an experimentally and theoretically important question. The enhancement of surface ordering at the liquid crystalhapor interface has been well documented experi-
-m
7,8........................................................
7.4..............................
0
7
- 1 which indicates stronger surface damping of the layer fluctuations. Holyst et al. [15] first carried out the numerical analysis on a discrete version of the free-energy (see Eq. 14). The model offers a satisfactory explanation for the X-ray scattering data obtained from the SmI/C phase of free-standing 70.7 films at 72.5 “C [16]. The 70.7 compound is one member of the homologous series of N-(4-n-aZkyZoxybenzylidene)-4-n-aZkylanilines(n0.m).The phase consists of monolayer two dimensional SmI surface layers on SmC interior layers [ 171. High resolution data from four film thicknesses, namely 3-, 5-, 15-, and 35-layers, were obtained. The data can be fitted by a model including molecular tilt angle profiles to account for the finite tilt angle of both the SmI and SmC phases. However, the effect of surface hexatic order was not considered. Consequently, the critical test of the model should be conducted on a simpler system, namely, free-standing SmA films. Shalaginov and Romanov [ 181 have recently developed a continuum model and obtained an explicit formula for the displacement-displacement correlation function. Furthermore, both the uniaxial properties of the SmA phase and multiple reflection from the interfaces have been included in this theoretical model. It offers excellent agreement with the X-ray scattering data [9] acquired at a temperature well into the SmA phase of FPP (or H7F6EPP) [5-n-heptyl-2(4-n-perfluorohexylethanophenyl)pyrimidine]. Both the specular reflectivity and offspecular diffuse diffraction from 4-, 20-, and
2.3
Hexatic Smectic B Phase
447
[9]. Nevertheless an independent measurement of the compressional elastic constant would be extremely important. The low value of y is due to the perfluorinated tail. Subsequently, a direct measurement of surface tension confirmed this value of y [ 191. In comparison with terminal aliphatic compounds, the reduction of yand enhancement in B leads to v(=y/(BK)"2)< 1. Thus surface damping of layer fluctuations is weaker.
2.3 Hexatic Smectic B Phase Figure 4. Log-log plot of the X-ray scattering intensity versus transverse scans at fixed qz for 4- and 34layer H7F6EPP films. The films are in the SmA phase. Circles and crosses indicate positive and negative q x , respectively. The values (in ,k ') of qL are 0.235 (a), 0.292 (b), 0.348 (c), 0.448 (d), 0.216 (e), 0.287 (0, 0.355 (g), and 0.429 (h), respectively. Data have been shifted for clarity. The solid lines are the best fits to the model. (Adapted from [9]).
34-layer films were measured. Figure 4 shows transverse scans obtained from 4- and 34-layer films at various fixed qz's. A broad peak with a long tail is characteristic of the Peierls -Landau instability. The theoretically fitted results are shown as solid lines. The best fitting results for both film thicknesses and various qz values yield K = l . O x IO-'dyn, B = l . O x 10'' dyn/cm2 and y= 13.0 dyn/cm. The value of K is approximately the same as that for the ordinary nematic elastic constant. The coefficient B is about two orders of magnitude larger than for ordinary liquid crystal compounds with saturated alkyl groups [ 12, 131. The more rigid part of the perfluorinated tail in the H7F6EPP compound may be the origin of this enhancement. This is consistent with the fact that the layer structure of this perfluorinated compound is better defined. The evidence is that the specular reflectivity exhibits pronounced second order peaks
2.3.1 Macroscopic Properties In 1971 Demus et al. [20] first reported the existence of the SmF phase in 2-(4-npentylphenyl)-5-(4-n-pentyloxyphenyl)pyrimidine. About eight years later, utilizing X-ray diffraction, Leadbetter et al. [21] and Benattar et al. [22] studied the molecular arrangement in this mesophase. A pseudohexagonal molecular arrangement was found within the smectic layers in both the SmF and SmI phases. The basic understanding of this observed pseudo-hexagonal molecular arrangement required the concept of hexatic order, developed in the context of the two-dimensional melting theory which was proposed at about the same time by Halperin and Nelson [23]. While the theorists put forth these exciting hypotheses, experimentalists [24], performing X-ray diffraction studies on various liquid crystal compounds, also demonstrated the existence of two different types of B phases with different ranges of interlayer positional correlation lengths. This observation was subsequently followed by several high resolution X-ray diffraction investigations on free-standing liquid crystal films [24-271. First, the conventional B phase in 4 0 . 8 (N-(4-n-butyloxybenzyli-
448
2 Physical Properties of Non-Chiral Smectic Liquid Crystals
dene)-4-n-octylaniline) was demonstrated to be a three-dimensional crystalline phase [25, 261. The more interesting liquid crystal mesophase exhibiting long-range bondorientational order, but only short-range positional order, the SmB,,, phase, was subsequently identified in 650BC [27]. In Fig. 5, the X-ray scattering scan along one of the “crystalline” axes (Q,, scan) displays short-range positional order, while the QI scan exhibits very weak interlayer coupling, similar to that for the SmA phase. The development of a six-fold modulation in an angular scan (with the rotation axis parallel to the layer normal) becomes apparent below the transition temperature TAB=67.9 “C (determined by high resolution heat-capacity measurements by Huang et al. [28]). This six-fold modulation indicates that the phase below TABexhibits 3 D long-range bond-orientational order. However, the size of the hexatic order domain is about 1 mm2, which was much smaller than the X-ray beam size. Extensive averaging is required to demonstrate the existence of six-fold modulation
4
650
using X-ray diffraction. This makes quantitative studies and analyses of the bondorientational order practically impossible. Employing electron diffraction with a beam size of about 1 pm2, Ho and coworkers [29, 301 obtained beautiful six-fold arcs. Figure 6 displays a reproduction of one of these photographs [30] obtained from the (a) SmA, (b) SmB,,,, and (c) B phases of
BC
Figure 5. %-averaged X-ray scattering intensity for a Qll scan (solid dots) and a QL scan (open circles) in the HexB,, phase of 650BC. Note that different scales are used for the QII and QI axes. The inset illustrates the directions of three different scans, namely Qll and QI scans. (Adapted from [27]).
x,
Figure 6 . Electron diffraction patterns from a 9-layer 54COOBC film. The entire film is in the SmA phase (a), in the HexB phase (b), and in the crystal B phase (c). (Adapted from Ref. 30).
2.3
a nine-layer 54COOBC [n-pentyl 4’-n-
pentanoyloxybiphenyl-4-carboxylate]film. The constant intensity ring found in the high temperature SmA phase indicates the liquid-like in-polar molecular arrangement. Such a diffraction pattern is seen ubiquitously in fluids. The low temperature crystal B phase gives one set of six sharp spots. Sharp diffraction peaks are the signature of a crystal. However, much weaker second order diffraction peaks, which cannot be seen in this photograph, demonstrate that the Debye-Waller factor is very small; that is, the crystal is very “soft”. The intermediate SmB,,, phase yields the characteristic six-fold arc for the hexatic order. Although the inplane translational order remains short-ranged, the positional correlation length shows a significant increase (from approximately 2 nm to 7 nm for 650BC) through the SmA-SmB,,, transition [27]. As a result of this experimental evidence, the traditional smectic B phase has been reclassified into two different phases: the SmB,,, and B phases. By employing a symmetry argument, the existence of long range positional order in the B makes the SmA-B [28,3 11 and SmB,,,-B [30] transitions first order. The order parameter associated with the SmA- SmB,,, transition is bond-orientational order. It can be represented by YH= YHoexp (i 6 06).This order parameter places the SmA - SmBh,,-transition in the XY universality class. Thus the transition can be continuous in three dimensions and should, furthermore, exhibit Kosterlitz Thouless-like behavior in two dimensions [32]. Mechanical responses obtained from a low frequency torsional oscillator yield a finite in-plane shear modulus for 4 0 . 8 films in the crystal B phase [33, 341 indicative of a solid-like response. The hexatic B phase of 650BC yields no in-plane shear modulus, just like a liquid [35].
Hexatic Smectic B Phase
449
The molecular bond direction is defined by an imaginary line between one pair of the nearest neighbour molecules [23]. This means that the measurement of bond-orientation correlation requires a four-point correlation function. To date no appropriate experimental tool allows us to conduct this kind of measurement. Thus the degree of bond-orientation correlation can only be inferred indirectly. Moreover, no available physical field exists which can couple directly to the hexatic order in the hexatic B phase. Consequently, the heat-capacity measurement is one of the most powerful experimental probes used to investigate the nature of the SmA-SmB,,, transition and to complement structural identification by X-ray or electron diffraction. The pioneering X-ray work on 650BC also revealed the existence of herringbone order 1271,but a detailed investigation to determine the range of the herringbone order has not yet been performed. Despite the indication of the herringbone order, this phase is simply denoted as the SmB,,, phase. Subsequent electron diffraction studies on 8-layer 3(10)OBC films also revealed the characteristic diffraction pattern due to herringbone order [36]. As a consequence, the hexaticB phase found in nmOBC compounds may be more complicated. To test the theoretically predicted thermal properties [37,38]related to the two-dimensional liquid-hexatic transition critically, we have carefully characterized the hexatic B phase in compound 54COOBC 1301 and found that it does not display herringbone order. To make a clear distinction between these two cases, we propose to use hexatic B to denote the phase found in 54COOBC and use SmB for the phase found in the nmOBC compounds which has a clear indication of some degree of herringbone order.
450
2 Physical Properties of Non-Chiral Smectic Liquid Crystals
ian proposed by Bruinsma and Aeppli [43] to describe the three-dimensional SmASmB,,, transition, we considered the two dimensional version of the Hamiltonian. In the appropriate parameter space of this model, a single transition from the disordered phase to the phase with hexatic and herringbone order can be identified. Furthermore, the heat capacity anomaly can be characterized by a=0.35. Although this coupled XY model provides a plausible explanation for the liquid-hexatic transition found in the nmOBC's, no existing theory is suitable to explain the results from 54COOBC two-layer films, which lack any indication of herringbone order. Thus, our experimental results offer a unique contrast. The structural data clearly demonstrate the existence of the liquid-hexatic transition predicted by the two-dimensional melting theory [23], while the calorimetric results disagree with the related theoretical prediction. Heat capacity data from thin 3(10)OBC free-standing films of various thicknesses are shown in Fig. 8. Using information from specular X-ray diffraction [ 161,electron diffraction [44], and the evolution of this set of
2.3.2 Thin Hexatic B Films Employing our state-of-the-art ac differential free-standing film calorimeter [39 -411, we have conducted high resolution heat capacity and optical reflectivity measurements on liquid crystal films as thin as two molecular layers (thickness approximately 2.5 ndlayer). Figure 7 displays the results from two-layer films of 3( 10)OBC [41] near the liquid-hexatic transition. This heat capacity anomaly shows divergent behavior, while the optical reflectivity displays a sign change in curvature in the immediate vicinity of the heat capacity peaks. These heat capacity data are very different from the theoretical prediction [37,38] which shows only an essential singularity at the two-dimensional liquid-hexatic transition temperature and a broad hump on the high temperature side. Moreover, these heat capacity data can be well-characterized by a simple power-law expression with the heat capacity critical exponent a = 0.30 f 0.05. Using Monte-Carlo simulations, we have investigated the role of herringbone order on the liquid-hexatic transition of nmOBC [42]. Based on the coupled XY Hamilton-
~"
"
73
'
"
"
'
"
'
"
'
75 76 TEMPERATURE ("C)
74
"
"I
Figure 7.Temperature variation of heat capacity and optical reflectivity obtained from 2-layer films of 3(10)OBC. The inplane molecular density was obtained directly from the optical reflectivity data. (Adapted from [41]).
2.3 7.5
5.2
(0
1
(d)
64
1 f
N=7
N=5
68 72 TEMPERATURE (“C)
76
t 9
3.2
(b)
64
N=3
68 72 TEMPERATURE (“C)
76
Figure 8. Temperature dependence of heat capacity near the SmA-SmB,,, phase transition of 3( l0)OBC. The data were obtained from 2- (a), 3- (b), 4- (c), S- (d), 6- (e) and 7-layer (f) films. Both layer-by-layer transition and sharpness of the heat capacity peaks are clearly shown.
heat capacity data from free-standing films, we can draw the following conclusion. Upon cooling from the high temperature SmA phase, hexatic order is established at the outermost layers and proceeds toward
Hexatic Smectic B Phase
45 1
the interior layers in a layer-by-layer fashion, for at least the first four sets of the outer layers [45]. The fact that three-layer films show two well separated transitions indicates that only two-layer films possess twodimensional thermal behavior. Careful thermal hysteresis measurements on various samples indicate that the SmA - SmB, transitions in bulk nmOBC samples and thin free-standing films are continuous [40, 41, 461. The following is a list of several counterintuitive, yet salient features: (1) Distinct layer-by-layer transitions are found associated with this continous transition. This suggests no interlayer coupling. (2) The layer-by-layer transition can be characterized by an exponent v =: 113, suggesting van der Waals-type interlayer interaction (see the discussion below). (3) Under the influence of the surface hexatic order, the transitions associated with the interior layers remain as sharp as the surface transaction. This also suggests no interlayer coupling. (4) The lower temperature heat capacity peak of 3-layer films is more than four times smaller than that of 4-layer films. This indicates a strong interlayer interaction for 3-layer films, but not for the 4-layer ones. Further experimental and theoretical advances are required to address these counterintuitive questions. Although surface enhanced order is commonly found for various liquid crystal transitions, the layer-by-layer transitions have only been characterized in the following five cases: the SmA-SmI transition of 90.4 1471, the SmA-SmB,,, transition of nmOBC [45], the SmA-SmB,,, transition of 54COOBC [48], the SmA-B transition of 40.8 [49] and the SmA-isotropic transition [50, 511. Thus far the sequence of the transition temperatures can be described by the following simple power law [ 5 2 ] :
452
2
Physical Properties of Non-Chiral Smectic Liquid Crystals
2.4
Smectic C Phase
2.4.1 Physical Properties near the Smectic A-Smectic C Transition
\
I
* '4
lo-'
IO-~
lo-*
lo-'
(T(L) -To)/To Figure 9. The number ( L ) of layers of a 90.4 thick film in the SmI phase which form at the SmA/vapour interfaces on cooling is plotted as function of the layer transition temperature (T(L)).The temperature Tois a freezing temperature which is one of the fitting parameters in Eq. (15). (Adapted from [47]).
Here L gives the separation (measured in units of layers) between the nearest film/ vapor interface and the layer in question. The fittings yield L0=0.24k0.01, u=O.37 k0.02 for 90.4 [47], L0=0.31 k0.02, u=0.32k0.02 for 3(10)OBC [45], L0=0.34 k0.05, u=0.30k0.05 for 54COOBC [48] andLo=0.32k 0 . 0 1 , =0.36 ~ k0.02 for 40.8 [49]. The experimental data and fitting results near the SmA-SmI transition of 90.4 are shown in Fig. 9. The critical exponent u = 113 indicates that simple van der Waals forces are responsible for the interlayer interaction. Although the layer-by-layer transition near a first order transition has been theoretically predicted [52], that near a second order transition is totally counterintuitive. Further experimental and theoretical work is necessary to address this unresolved puzzle.
2.4.1.1 Bulk Properties
The major difference between the SmA and SmC phases is that the latter exhibits a finite molecular tilt angle from the layer normal. While the SmA phase possesses uniaxial properties, the SmC phase shows biaxial properties. Upon heating, one usually finds the following three possible transitions from the SmC phase: SmC-isotropic, SmC-N, and SmC- SmA. In the first two cases, more than one degree of order vanishes in the given transition. For example, both the layer structure and molecular tilt disappear at the SmC-N transition. These cases are, in general, first-order transitions. Even though pretransitional behavior is not uncommon, the tilt angle shows a finite jump at the transition temperature. The order parameter associated with the SmC - SmA transition can be written as Yc= Oo exp(i4,) [53]. The amplitude Oo is the molecular tilt angle relative to the layer normal and the phase factor GC denotes the azimuthal angle of the molecular director. In 1972, de Gennes [53] argued that the SmA - SmC transition should belong to the three-dimensional XY universality class (helium-like) and might be continuous. This important observation stimulated numerous experimental efforts to investigate the nature of this transition. However, the novelty of this phase transition was revealed ten years later by careful X-ray diffraction studies [54]. Subsequent high resolution calorimetric studies by Huang and Viner [55] enabled them not only to confirm the meanfield nature of this transition, but also to pro-
2.4
pose an extended mean-field model to characterize this transition fully. The majority of the SmA - SmC transitions are found to be mean-field [56-611 and also to be in the close vicinity of the mean-field tricritical point. Thus, to describe the physical properties associated with this transition, we need the following extended mean-field free energy expansion [55, 621.
G = Go + a t I YCl2 + b I YCI4+ c JYclh (16) Here t (=(T-T,)IT,) is the reduced temperature with T, being the transition temperature. The coefficients a and c are usually positive constants. For a continuous SmA-SmC transition b > O and for a first order one b < 0. The special case, b = 0, is the mean-field tricritical point. Huang and Viner [55] proposed a dimensionless parameter to=b2/(ac) to describe the crossover temperature between the mean-field tricritical region (I t l 9 to) and the ordinary meanfield regime ( I t 1 4 to). The value of to is usually very large for most other mean-field transitions [55], (i.e. paraelectric-ferroelectric, conventional normalsuperconducting, Jahn-Teller-type transitions). Thus the I YCI6 term can be ignored. To our great surprise, most of the SmAor even as SmC transitions have to < low as to= lo-'. Thus they are extremely close to a mean-field tricritical point. Various research groups have not only characterized continuous and first order SmASmC transitions, [63] but also identified systems in which the transition is tricritical [64-661. The width of the critical region may be estimated from the following Ginzburg criterion [67]:
Due to the unusually large value of the meanfield heat capacity jump (AC > lo6 erg/ cm3 K), the critical region of the SmA-SmC
Smectic C Phase
45 3
transition becomes experimentally inaccessible (tc < lo-') provided that > 1.3 nm which is slightly larger than the effective size of the molecules. Since the SmC ordering is presumably not driven by long-range interactions and the SmA-SmC transition is not at or above the upper critical dimensionality (d,=4 for the XY universality class), it is very difficult to explain the observed mean-field behavior. If the coupling to the layer undulation could be neglected, the SmA- SmC transition would be helium-like. Although the effect of layer undulation on the nature of the SmA - SmB,,, transition was considered by Selinger [68], it is extremely important to conduct extensive studies of the effect of the quasi-long-range SmA order on the molecular tilt order (SmA- SmC transition) and the bond-orientation order (SmA -SmB,,, transition). Meanwhile, it is very important to identify experimentally a system showing criticalfluctuations associated with the SmA-SmC transition [69]. In principle, with sufficient resolution, the SmA- SmC transition should exhibit an extremely rich series of cross-over behavior in an appropriate liquid crystal compound: mean-field tricritical-like ( 1 t (> to) - ordinary meanfield (to> I t I > tc) - Gaussian fluctuations (I tl= tc)-critical fluctuations (tc> I t l ) upon approaching the transition temperature. In light of the abnormal behavior of ultrasound velocity and attenuation near the SmA-SmC transition [70, 711, Benguigui and Martinoty [72] advanced a theory to explain the experimental data. They concluded that the Ginzburg crossover parameter ( t G ) determined by the static properties, (e.g. heat capacity) could be much smaller than that obtained from the measurement of the elastic constant. However, a quantitative comparison between the theoretical prediction and the experimental data is still lacking.
tetf
454
2
Physical Properties of Non-Chiral Smectic Liquid Crystals 8
From Eq. (16), one can calculate the temperature variation of tilt angle, heat capacity (C) and susceptibility They are given as follows:
(x).
For T > T,,
eo= o c=c,
L
0 -1.0
(184 (194
-0.6
-0.2 0
0.2
0.6
1.0
T-Tc (lo
a)
and
x-'= 2 a t
(20 4
For T < T,, 00 = { [-b+b
(1 + 3 I tI/to)'/2]/(3~)} 1'2 (18 b)
c = co+ TA((T, - T)/T,)-'/~ and .. .
x-' = 8 a ( t l + (8ato/3)
305
. [ l -(1 +31tl/to)]"2
(20b)
b)
310
31 5
320
325
TEMPERATURE ( K )
Figure 10. (a) Temperature variation of tilt angle
Here Co is the non-singular part of the (circles) for T < T, and reciprocal of the susceptibility heat capacity. A = L Z ~ / ~ / [ ~ ( ~ C )and ' ~ ~ T (triangles) ~ / * ] for T>T, near the SmA-SmC transition of 40.7. The transition temperature Tc=49.69"C. T, =T, (1 + t0/3). These equations demonThe tilt angles are measured by X-ray (open circles) strate that the effect of mean-field triand light (solid circles) scattering. The solid line is the criticality only shows up in the region T < T, best fit to Eq. (18) with to= 1 . 3 ~ (Adapted from in which I Ycl is non-zero. Except for the [56]).(b) Heat capacity in J/cm3K versus temperature temperature variation of for T < T, many near the SmA-SmC transition of racemic 2M450BC. The solid curve is the best fit to Eq. (19) with high resolution experiments [55 -611 have to= 3 . 9 ~ 1 0 - ~ (Adapted . from [58]). been conducted to test these predictions critically. Temperature variation of both Oo(Tlo0 V and the frequency of the field increased, a focal conic texture eventually appeared. This is the point at which the material changed from positive to negative dielectric anisotropy. This was repeated for a range of temperatures (Fig. 12). In nematics, this temperature dependence is a major problem, because addressing schemes have to be complex to cope with changing voltages and frequencies as the ambient temperature changes. In smectics, the additional problem is that in the negative mode (switching from homeotropic to planar or focal conic) the focal conics are too large to cause light scattering and yet not
Polymer dispersed liquid crystal (PDLC) devices usually contain N phases as the liquid crystal material [71] and are used in vision products [72], e. g., privacy windows, projection displays [73], and direct view displays [74, 751. Cholesteric liquid crystals have also been used [76]. All these devices relax back to the original ground state when the field is removed. Ideally such films consist of diroplets of liquid crystal in a polymer matrix; the reverse situation (reverse phase) consists of a liquid crystal continuum with polymer balls dispersed within it. The latter films are not desirable, because they do not provide reversible electrooptic effects. Ferroelectric SmC* liquid crystals have also been used [77], and these give bistable
488
3 Nonchiral Smectic Liquid Crystals - Applications
devices not too dissimilar to conventional FLC devices. The light scattering appearance of a thick (> 20 pm) film of an SmA liquid crystal provides an opaque film. If doped with a colored dichroic dye, an intense color can be produced. On heating this film to the isotropic liquid, the light scattering disappears and the light path through the film is reduced such that a fairly clear film is produced. An appropriately colored background, previously hidden by the scattering film, would now be visible. This is the basis for a temperature indicator. One problem with this concept is that the liquid crystal is fluid; this problem can be overcome by containing the smectic phase in a PDLC film. By heating a PDLC film made from a liquid crystal having a smectic and an N phase, it is possible to make an electrically addressed storage device [78]. The PDLC film is heated to the temperature at which the liquid crystal is nematic and a small electric field applied to the appropriate areas using I T 0 conductors; this homeotropically aligns the N phase into a clear state and on cooling with the field applied, a transparent image on a scattering background is produced. A PDLC film containing an SmA liquid crystal can also be addressed with light from a suitable laser to produce a clear line under the influence of an applied electric field when the film cools [79]. Very recently a ‘memory type liquid crystal’, which has the characteristics of an
SmA phase, has also been used to produce a memory type device [80]. Interestingly, this device uses reverse phase PDLC so that a finer line resolution can be obtained, reversal of the process being of no importance. This device consists of a thin (6 pm) film of memory type liquid crystal and has submicrometer sized polymer balls dispersed within it and then a polymer skin on top to give.a memory type liquid crystal polymer composite (M-LCPC) (Fig. 13). This film is prepared on top of an I T 0 coated glass substrate. The film switches from scattering to clear between 200 and 300 V to provide many grey levels. The film is used in the same way as a photographic film, but in a special camera. Held 10 pm above the M-LCPC layer is a structure consisting of an organic photoconductor and an I T 0 electrode mounted on a glass substrate. A large DC field (700 V) is applied between the two I T 0 layers. When an area of the photoconductor is exposed, its resistivity drops in accordance with the energy of light and allows a field to be applied to the M-LCPC layer, which responds to the strength and duration of the field applied to it. Typical exposure time is 1/250 s for the camera shutter and the field is applied for about 1/25 s. The camera contains a lens to separate the incident image into red, green, and blue images, each of which is captured on three M-LCPC films. The films are removed from the camera, scanned with a 5000-CCD line sensor with a pixel size of 6 pm and the analog signals converted to
Figure 13. Submicrometer polymer balls in a continuum of memory type liquid crystal and a polymer cover film.
3.10 References
digital data, enhanced, and the three images combined; the final image can then be displayed on monitors or printed on conventional color printers. A resolution of 6 ym was demonstrated and the sensitivity was about I S 0 10-50; future films are expected to be I S 0 100. Compared to conventional silver halide photographs (from negative film), the images produced were much less granular. Thus this system for capturing images provides high resolution color prints very quickly, and the master images can be stored indefinitely.
3.9
Conclusions
Basic research to discover how smectic liquid crystals are affected by heat and electric fields has resulted in several effects being found, and a few of them have been extensively studied and developed into working display prototypes and early commercialization. The property of bistability is both useful and, at the same time, a problem. As yet a really long term application involving smectic liquid crystals has not been found. However, they continue to hold a fascination because they can offer many unique properties which are continually being reexplored for new applications.
3.10 References A . J. Leadbetter in Therniotropic Liquid Crystals (Ed.: G . W. Gray), Society of Chemical Industry, Great Britain 1987, Chap. 1 . G. W. Gray, J. W. Goodby, Smectic Liquid Crystals, Leonard Hill, Glasgow 1984. D. Coates, W. A. Crossland, J. H. Morrissey, B. Needham, Mol. Cryst. Liq. Cryst. 1978,41, 151. D. A. Dunmur, M. R. Manterfield, W. H. Miller, J . K. Dunleavy, Mol. Cryst. Liq. Cryst. 1978,45, 127.
489
151 D. Demus, H. D. Demus, H. Zaschke, Flin’ssige Kristalle in Tabellen, VEB Deutscher Verlag fur Grundstoffindustrie, Leipzig 1974. 161 F. J. Kahn, Appl. Phys. 1973, 22, 1 1 1. 171 G. N. Taylor, F. J. Kahn, J. Appl. Phys. 1974,45, 4330. [8] G. W. Gray, K. J. Harrison, J. A. Nash, Electron. Lett. 1973, 9, 130. [ 91 G. W. Gray, Advances in Liquid Crystal Materiuls ,for Applications, BDH Chemicals, Dorset 1978. [ 101 G. W. Gray, K. J. Harrison, J. A. Nash, J. Chem. Soc. Chem. Commun. 1974,431. [ I 11 Merck Ltd., Licrilite Catalogue 1994. [ I21 S. LeBerre, M. Hareng, R. Hehlen, J. N. Perbert, Displuys 1981, 349. [ 131 M. Hareng, S. LeBerre, R. Hehlen, J. N. Perbert, SlD Digest of Technical Papers 1981, 106. [ 141 H. Scherrer, A. Boller, US Patent 3 952046,1976. [IS] I. Shimizu, K. Furukawa, M. Tanaka, EP Patent 037 468 B 1, 1988. [16] J. Newton, H. Coles, P. Hodge, J. Hannington, J . Muter: Chern. 1994,4(6), 869. [ 171 M. Ibn-Elhaj, H. Coles, D. Guillon, A. Skoulios, J . Phys. I / (France) 1993,3, 1807. [18] J. C. Dubois, A. Zann, A. Beguin, M o l . Cryst. Liq. Cryst. 1977, 42, 139. 1191 M. Chambers, R. Clemitson, D. Coates, S. Greenfield, J. A. Jenner, I. C. Sage, Liq. Cryst. 1989, 5(1), 153. 1201 K. Kida, M. Uchida, N. Kato, A. Mori, H. Takeshita, Chem. Express 1991, 6(7),503. [21] H. Melchior. F. Kahn, D. Maydan, D. B. Fraser, Appl. Phys. Lett. 1972, 21, 392. [22] D. F. Aliev, A. K. Zeinally, Zh. Tekh. Fiz. 1982, 52, 1669. [23] T. Urabe, K. Arai, A. Ohkoshi, J . Appl. Phys. 1983,54, 1552. [24] A. G. Dewey. J. T. Jacobs, B. G. Huth, SID Digest Technicul Papers 1978, 19. 1251 W. H. Chu, D. Y. Yoon, Mol. Cryst. Liq. Cryst. 1979, 54, 245. [26] R. Daley, A. J. Hughes, D. G. McDonnell, Liq. Cryst. 1989, 4(6), 585. [27] C. W. Walker, W. A. Crossland, Displuys 1985, 207. (281 M. Hareng, S . LeBerre, Electron. Lett. 1975, I / , 73. 1291 F. Kahn, P. N. Kendrick, J. Leff, J. Livioni, B. E. Lovcks, D. Stepner, SID Digest Technical Pupers 1987, 18, 254. [30] J. Harold, C. Steele, Proc. SID 1985, 2 6 , 141. 1311 D. Mash, GB Patent 2093206A, 1982. [32] A. Sasaki, N. Hayashi, T. Ishibashi, Jpn. Display 1983,497. [ 33] J. Harold, C. Steele, Eurodispluy Proc. (Paris) 1984, 29. [34] M. R. Smith,R. H. Burns, R. C. Tsai, Proc. SPIE 1980,200, 17 1.
490 [35] [36] [37] [38] [39] [40]
3 Nonchiral Smectic Liquid Crystals - Applications
A. G. Dewey, Opt. Engl. 1984,23, 230. G. J. Sprokel, US Patent 3 999 838, 1976. D. Armitage, Appl. Phys. 1981,52,4843. D. Coates, GB Patent 2091 1753 A, 1982. Eastman Kodak, US Patent 3 774 122, 1973. T. Nishizawa, R. Tukahara, R. Morinaka, T. Hidaka, GB Patent 2069518,1981. [41] T. Urabe, GB Patent 2 140023A, 1984. [42] V. T. Muller-Westerhof, A. Nazzal, R. J. Cox, A. M. Giroud, Mol. Cryst. Liq. Cryst. 1980,56,249. [43] K. Nakamura, 0. Yashuhiro, M. Sugimoto, J. Appl. Phys. 1989,59(2), 593. [44] A. G. Dewey, S. F. Anderson, G. Cherkoff, J. S. Feng, C. Handen, H. W. Johnson, J. Leff, R. T. Lynch, C. Marinelli, R. W. Schmiedeskamp, SID Digest Technical Papers 1983, 36. [45] A. G. Dewey, IEEE Electron Devices 1977,24, 918. [46] D. Mayden, N. Melchior, F. J. Kahn in Proc. IEEE Conference in Display Devices, 1972, p. 166. [47] M. Hareng, S. LeBerre, B. Mourey, P. C. Moutou, J. N. Perbert, L. Thirant in Proc. of the IEEE International Displays Research Conference 1982, p. 126. [48] M. Hareng, R. Hehlen, S. LeBerre, J. P. LePesant, L. Thirant in Proc. of the Electronics Displays Conference, Vol. 1 1984, p. 28. [49] S. Lu, D. H. Davies, R. Albert, D. B. Chung, A. Hochbaum, C. H. Chung, SID Digest Technical Papers, 1982, p. 238. [50] S. Lu, D. H. Davies, C. H. Chung, D. Evanicky, R. Albert, R. Traber in IEEE International Display Research Conference 1982, p. 132. [51] W. Harter, S. Lu, S. Ho, B. Basler, W. Newman, C. Otaguro in SID Symp. Digest. 1984, p. 196. [52] S. Lu, D. Buuvinhchu, US Patent 4461715, 1984. [53] M. Hareng, S. LeBerre, L. Thirant, Appl. Phys. Lett. 1974,25, 12. [54] M. Hareng, S. LeBerre, L. Thirant, Appl. Phys. Lett. 1975,27, 575. [55] W. A. Crossland, J. H. Morrissey, D. Coates, GB Patent 1569688, 1977. [56] D. Coates, A. B. Davey, C. J. Walker, Eurodisplay 1987, 96. [57] V. N. Chirkov, D. F. Aliev, G. M. Radzhabov, A. K. Zeinally, Mol. Cryst. Liq. Cryst.Lett. 1979, 49, 293.
(581 D. Coates, W. A. Crossland, J. H. Momssey, B. Needham, J. Phys. D: Appl. Phys. 1978, 11, 2025. [59] W. A. Grossland, J. H. Morrissey, D. Coates, GB Patent 1569686, 1977. [60] G. H. Heilmeier, L. A. Zanoni, L. A. Barton, Appl. Phys. Lett. 1968, 13,46. [61] B. Bahadur in Liquid Crystals - Applications and Uses (Ed. B. Bahadur) World Scientific, Singapore 1990, Vol. 1, Chap. 9. [62] C. Tani, Appl. Phys. Lett. 1971, 19, 241. [63] M. Steers, A. Mircea-Roussel, J. Physique Coll. 1976,37(3), 145. [64] W. Helfrich, J. Chern. Phys. 1969, 51,4092. [65] W. A. Crossland, D. Coates, GB Patent 1 601 601, 1978. [66] S. Matsumoto, M. Kawamoto, T. Tsukada, Jpn. J. Appl. Phys. 1975,14,7. [67] J. A. Geurst, W. J. A. Goosens, Phys. Letts. 1972,41(4), 369. [68] W. A. Crossland, P. J. Ayliffe, Proc. SID 1979, 23,9. [69] W. A. Crossland, S . Cantor, SID Digest Technical Papers 1985,16, 124. [70] D. Coates, Mol. Cryst. Liq. Cryst. 1978,49, 83. [71] J. L. Fergason, SID Digest Technical Papers 1985, 16,68. [72] Taliq Varilite and 3M Privacy Film. [73] Y. Nagoe, K. Ando, A. Asano, I. Takemoto, J. Havens, P. Jones, D. Reddy, A. Tomita, SID Digest Technical Papers 1995,26,223. [74] P. Nolan, M. Tillin, D. Coates, E. Ginter, E. Leuder, T. Kallfass, Eurodisplay 1993, 397. [75] P. Drzaic, R. C. Wiley, J. McCoy, Proc. SPIE 1989,1080,41. [76] P. P. Crooker, D. K. Yang, Appl. Phys. Lett. 1990,57,2529. [77] V. Y. Zyryanov, S. L. Smorgan, V. F. Shabanov, SID Digest Technical Papers 1992, 776. [78] M. Koshimizi, T. Kakinuma, Japanese Patent 07 36007 (95 36007), 1995; Chem. Abs. 1995, 123,70512. [79] T. Kakinuma, M. Koshimizu, T. Teshigawara, Jpn. Display (Hiroshima) 1992, 85 1. [80] M. Utsumi, M. Akada, E. Inoue, IS&T’s48th Annual Conference Proceedings 1995, p. 499.
Handbook of Liquid Crystals D. Demus, J. Goodby, G. W. Gray, H.-W. Spiess, V. Vill
VOl. 1: Fundamentals Vol. 2 A: Low Molecular Weight Liquid Crystals I Vol. 2 B: Low Molecular Weight Liquid Crystals I1 VOl. 3: High Molecular Weight Liquid Crystals
Further title of interest:
J. L. Serrano: Metallomesogens ISBN 3-527-29296-9
D. Demus, J. Goodby, G. W. Gray, H.-W. Spiess, V. Vill
Handbook of Liquid Crystals
8WILEY-VCH
D. Demus, J. Goodby, G. W. Gray, H.-W. Spiess, V. Vill
Handbook of Liquid Crystals Vol. 2 B: Low Molecular Weight Liquid Crystals I1
8WILEY-VCH Weinheim New York Chichester Brisbane Singapore Toronto
Prof. Dietrich Demus Veilchenweg 23 06 1 18 Halle Germany Prof. John W. Goodby School of Chemistry University of Hull Hull, HU6 7RX U. K. Prof. George W. Gray Merck Ltd. Liquid Crystals Merck House Poole BH15 1TD U.K.
Prof. Hans-Wolfgang Spiess Max-Planck-Institut fur Polymerforschung Ackermannweg 10 55 128 Mainz Germany Dr. Volkmar Vill Institut fur Organische Chemie Universitat Hamburg Martin-Luther-King-Plat2 6 201 46 Hamburg Germany
This book was carefully produced. Nevertheless, authors, editors and publisher do not warrant the information contained therein to be free of errors. Readers are advised to keep in mind that statements, data, illustrations, procedural details or other items may inadvertently be inaccurate.
Library of Congress Card No. applied for. A catalogue record for this book is available from the British Library. Deutsche Bibliothek Cataloguing-in-Publication Data: Handbook of liquid crystals I D. Demus ... - Weinheim ; New York ; Chichester ; Brisbane ; Singapore ; Toronto : Wiley-VCH ISBN 3-527-29502-X Vol. 2B. Low molecular weight liquid crystals. - 2. - (1998) ISBN 3-527-29491-0
0 WILEY-VCH Verlag GmbH. D-60469 Weinheim (Federal Republic of Germany), 1998 Printed on acid-free and chlorine-free paper.
All rights reserved (including those of translation in other languages). No part of this book may be reproduced in any form - by photoprinting, microfilm, or any other means - nor transmitted or translated into machine language without written permission from the publishers. Registered names, trademarks, etc. used in this book, even when not specifically marked as such, are not to be considered unprotected by law. Composition and Printing: Fa. Konrad Triltsch Druck- und Verlagsanstalt GmbH, D-97070 Wiirzburg. Bookbinding: Wilhelm Osswald & Co., D-67433 Neustadt Printed in the Federal Republic of Germany.
The Editors D. Demus studied chemistry at the Martin-Luther-University, Halle, Germany, where he was also awarded his Ph. D. In 1981 he became Professor, and in 1991 Deputy Vice-Chancellor of Halle University. From 1992-1994 he worked as a Special Technical Advisor for the Chisso Petrochemical Corporation in Japan. Throughout the period 1984-1991 he was a member of the International Planning and Steering Commitee of the International Liquid Crystal Conferences, and was a non-executive director of the International Liquid Crystal Society. Since 1994 he is active as an Scientific Consultant in Halle. He has published over 310 scientific papers and 7 books and he holds 170 patients.
J. W. Goodby studied for his Ph. D. in chemistry under the guidance of G. W. Gray at the University of Hull, UK. After his post-doctoral research he became supervisor of the Liquid Crystal Device Materials Research Group at AT&T Bell Laboratories. In 1988 he returned to the UK to become the Thorn-EMI/STC Reader in Industrial Chemistry and in 1990 he was appointed Professor of Organic Chemistry and Head of the Liquid Crystal Group at the University of Hull. In 1996 he was the first winner of the G. W. Gray Medal of the British Liquid Crystal Society.
G . W. Gray studied chemistry at the University of Glasgow, UK, and received his Ph. D. from the University of London before moving to the University of Hull. His contributions have been recognised by many awards and distinctions, including the Leverhulme Gold Medal of the Royal Society (1987), Commander of the Most Excellent Order of the British Empire (1991), and Gold Medallist and Kyoto Prize Laureate in Advanced Technology (1 995). His work on structure/property relationships has had far reaching influences on the understanding of liquid crystals and on their commercial applications in the field of electro-optical displays. In 1990 he became Research Coordinator for Merck (UK) Ltd, the company which, as BDH Ltd, did so much to commercialise and market the electro-optic materials which he invented at Hull University. He is now active as a Consultant, as Editor of the journal “Liquid Crystals” and as authorleditor for a number of texts on Liquid Crystals.
VI
The Editors
H. W. Spiess studied chemistry at the University of FrankfurVMain, Germany, and obtained his Ph. D. in physical chemistry for work on transition metal complexes in the group of H. Hartmann. After professorships at the University of Mainz, Miinster and Bayreuth he was appointed Director of the newly founded Max-Planck-Institute for Polymer Research in Mainz in 1984. His main research focuses on the structure and dynamics of synthetic polymers and liquid crystalline polymers by advanced NMR and other spectroscopic techniques.
V. Vill studied chemistry and physics at the University of Miinster, Germany, and acquired his Ph. D. in carbohydrate chemistry in the gorup of J. Thiem in 1990. He is currently appointed at the University of Hamburg, where he focuses his research on the synthesis of chiral liquid crystals from carbohydrates and the phase behavior of glycolipids. He is the founder of the LiqCryst database and author of the LandoltBornstein series Liquid Crystals.
List of Contributors Volume 2B, Low Molecular Weight Crystals I1
Boden, N.; Movaghar, B. (IX) Centre for Self-Organising Molecular Systems (SOMS) Dept. of Chemistry University of Leeds Leeds LS9 9JT U.K.
Fukuda, A.; Miyachi, K. (VI:3) Shinshu University Faculty of Textile Science and Technology Dept. of Kansei Engineering Tokida 3-15-1, Ueda-shi Nagano-ken 386 Japan
Bushby, R. J. (VII) University of Leeds Leeds LS2 9JT U.K.
Giroud, A.-M. (XIV) Chimie de Coordination Unit6 de Recherche AssociCe au CNRS N. 1194 CEA Grenoble-DRFMC/SCIB 17, rue des Martyrs 38054 Grenoble Cedex France
Cammidge, A. N. (VII) University of East Anglia Norwich NR4 7TJ U.K. Chandrasekhar, S. (VIII) Centre for Liquid Crystal Research P. 0. Box 1329 Jalahalli Bangalore - 560 013 India Diele, S.; Goring, P. (XIII) Martin-Luther-Universitat Haile-Wittenberg FB Chemie Inst. f. Physikal. Chemie 06099 Halle Germany
Imrie, C. T. (X) University of Aberdeen Dept. of Chemistry Meston Walk, Old Aberdeen AB24 3UE U.K. Kato, T. (XVII) University of Tokyo Institute of Industrial Science 7-22- 1 Roppongi, Minato-ku Tokyo 106 Japan
VTII
List of Contributors
Kelly, S. M. (V1:l) The University of Hull Liquid Crystals & Advanced Organic Materials Research Group Dept. of Chemistry Hull HU6 7RX U.K . Lagerwall, S. T. (VI:2) Chalmers University of Technology Physics Department Liquid Crystal Group 41296 Goteborg Sweden Luckhurst, G. R. (X) University of Southampton Dept. of Chemistry Highfield Southampton SO17 1BJ U.K. Lydon, J. E. (XVIII) University of Leeds Dept. of Biochemistry & Molecular Biology Leeds LS2 9JT U.K. MalthCte, J. (XII) Institut Curie Section de Recherche UMR - CNRS 168 1 1 , Rue Pierre et Marie Curie 75231 Paris Cedex 05 France
Nguyen, H. T.; Destrade, C. (XII) Centre de Recherche Paul Pascal Avenue A. Schweitzer 33600 Pessac France Praefcke, K.; Singer, D. XVI TU Berlin IOC StraBe des 17. Juni 124 10623 Berlin Germany Sadashiva, B. K. (XV) Raman Research Institute CV Raman Avenue Bangalore - 560 080 India Weissflog, W. (XI) Max-Planck-Gesellschaft Arbeitsgruppe Flussigkristalline Systeme Martin-Luther-Universitat Mahlpforte 1 06108 Halle Germany
Outline Volume 1
..................
Chapter I: Introduction and Historical Development . George W Gray
1
Chapter 11: Guide to the Nomenclature and Classification of Liquid Crystals . . . . . 17 John W Goodby and George W Gray Chapter 111: Theory of the Liquid Crystalline State . . . . . . . . . . . . . . . 1 Continuum Theory for Liquid Crystals . . . . . . . . . . . . . . . . Frank M . Leslie Molecular Theories of Liquid Crystals . . . . . . . . . . . . . . . . 2 M. A. Osipov Molecular Modelling . . . . . . . . . . . . . . . . . . . . . . . . . 3 Mark R. Wilson Chapter I V General Synthetic Strategies . Thies Thiemann and Volkmar Ell
. . . . 25 . . . . 25 . . .
.
40
. . . . 72
....................... ................
115
..............
133
................ .........,.......
189 189
...............
204
...........
215
............
23 1
.........
253
Chapter V Symmetry and Chirality in Liquid Crystals. John W Goodby
Chapter VI: Chemical Structure and Mesogenic Properties Dietrich Demus Chapter VII: Physical Properties . . . . . . . . . . . Tensor Properties of Anisotropic Materials 1 David Dunmur and Kazuhisa Toriyama 2 Magnetic Properties of Liquid Crystals . . David Dunmur and Kazuhisa Toriyama 3 Optical Properties . . . . . . . . . . . . . David Dunmur and Kazuhisn Toriyama 4 Dielectric Properties . . . . . . . . . . . . David Dunmur and Kazuhisa Toriyama 5 Elastic Properties. . . . . . . . . . . . . . David Dunmur and Kazuhisa Toriyama 6 Phase Transitions. . . . . . . . . . . . . .
87
.
..
... ... .....
. ... .. ..
. . . . . . . . . . . . . . . . . 28 1
X
Outline
6.1 6.2 6.2.1 6.2.2 6.2.3 6.2.4 6.3 6.4 7
8 9 10 11
12 13
Phase Transitions Theories . . . . . . . . . . . . . . . . . . . . . . . . . . Philippe Barois Experimental Methods and Typical Results . . . . . . . . . . . . . . . . Thermal Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Jan Thoen Density . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Wolfgang Wedler Metabolemeter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Wolfgang Wedler High Pressure Investigations . . . . . . . . . . . . . . . . . . . . . . . . . P. Pollmann Fluctuations and Liquid Crystal Phase Transitions . . . . . . . . . . . . P. E . Cladis Re-entrant Phase Transitions in Liquid Crystals . . . . . . . . . . . . . . P. E . Cladis Defects and Textures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Y: Bouligand Flow Phenomena and Viscosity . . . . . . . . . . . . . . . . . . . . . . . Frank Schneider and Herbert Kneppe Behavior of Liquid Crystals in Electric and Magnetic Fields . . . . . . . Lev M . Blinov Surface Alignment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Blandine Je'rSme Ultrasonic Properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Olga A . Kapustina Nonlinear Optical Properties of Liquid Crystals . . . . . . . . . . . . . . R Pal&-Muhoray Diffusion in Liquid Crystals . . . . . . . . . . . . . . . . . . . . . . . . . l? Noack
'
281
. 310 310 334 350 355
. 379 . 391 406 454
. 477 535 549
. 569 582
Chapter VIII: Characterization Methods . . . . . . . . . . . . . . . . . . . . . . . . 595 1 Magnetic Resonance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 595 Claudia Schmidt and Hans Wolfgang Spiess X-Ray Characterization of Liquid Crystals: Instrumentation . . . . . . . . 619 2 Richard H . Templer Structural Studies of Liquid Crystals by X-ray Diffraction . . . . . . . . . 635 3 JohnM.Seddon 4 Neutron Scattering . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 680 Robert M . Richardson Light Scattering from Liquid Crystals . . . . . . . . . . . . . . . . . . . . 699 5 Helen E Gleeson Brillouin Scattering from Liquid Crystals . . . . . . . . . . . . . . . . . . 719 6 Helen E Gleeson Mossbauer Studies of Liquid Crystals . . . . . . . . . . . . . . . . . . . . 727 7 Helen F. Gleeson
Outline
XI
Chapter IX: Applications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 731 1 Displays . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 731 Ian C. Sage Nondisplay Applications of Liquid Crystals . . . . . . . . . . . . . . . . . 763 2 William A . Crossland and TimothyD . Wilkinson 823 Thermography Using Liquid Crystals . . . . . . . . . . . . . . . . . . . . 3 Helen F: Gleeson Liquid Crystals as Solvents for Spectroscopic. Chemical 4 Reaction. and Gas Chromatographic Applications . . . . . . . . . . . . . 839 William J. Leigh and Mark S. Workentin Index Volume 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
897
Volume 2 A Part I: Calamitic Liquid Crystals . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
1
Chapter I: Phase Structures of Calamitic Liquid Crystals . . . . . . . . . . . . . . . . . 3 John W Goodby Chapter 11: Phase Transitions in Rod-Like Liquid Crystals . . . . . . . . . . . . . . . 23 Daniel Guillon Chapter 111: Nematic Liquid Crystals . . . . . . . . . . . . . . . . . . . . . . . . . . 47 Synthesis of Nematic Liquid Crystals . . . . . . . . . . . . . . . . . . . . . 47 1 Kenneth J . Toyne 2 Physical Properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60 Elastic Properties of Nematic Liquid Crystals . . . . . . . . . . . . . . . . 60 2.1 Ralf Stannarius Dielectric Properties of Nematic Liquid Crystals . . . . . . . . . . . . . . . 91 2.2 Horst Kresse Diamagnetic Properties of Nematic Liquid Crystals . . . . . . . . . . . . . 113 2.3 Ralf Stannarius 2.4 Optical Properties of Nematic Liquid Crystals . . . . . . . . . . . . . . . 128 Gerhard Pelzl 2.5 Viscosity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 142 Herbert Kneppe and Frank Schneider 2.6 Dynamic Properties of Nematic Liquid Crystals . . . . . . . . . . . . . . 170 R. Blinc and Z. Mus'evic' 3 Applications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 199 199 TN, STN Displays . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.1 Harald Hirschmann and Volker Reiffenrath Active Matrix Addressed Displays . . . . . . . . . . . . . . . . . . . . . . 230 3.2 Eiji Kaneko
XI1 3.3 3.4
Outline
Dynamic Scattering . . . . . . Birendra Bahadur Guest-Host Effect . . . . . . . Birendra Bahadur
.......................
243
.......................
257
Chapter IV: Chiral Nematic Liquid Crystals . . . . . . . . . . . . . . . . . . . . . . 303 The Synthesis of Chiral Nematic Liquid Crystals . . . . . . . . . . . . . . 303 1 Christopher J. Booth Chiral Nematics: Physical Properties and Applications . . . . . . . . . . . 335 2 Harry Coles Chapter V: Non-Chiral Smectic Liquid Crystals . . . . . . . . . . . . . . . . . . . . Synthesis of Non-Chiral Smectic Liquid Crystals . . . . . . . . . . . . . . 1 John W Goodby Physical Properties of Non-Chiral Smectic Liquid Crystals. . . . . . . . . 2 C. C. Huang Nonchiral Smectic Liquid Crystals - Applications . . . . . . . . . . . . . 3 David Coates
41 1 41 1 441 470
Volume 2B Part 2: Discotic Liquid Crystals. . . .
.........................
Chapter VI: Chiral Smectic Liquid Crystals . . . . . . . . . . . . . Synthesis of Chiral Smectic Liquid Crystals. . . . . . . . 1 Stephen M. Kelly Ferroelectric Liquid Crystals. . . . . . . . . . . . . . . . 2 Sven i? Lagerwall Antiferroelectric Liquid Crystals . . . . . . . . . . . . . 3 Kouichi Miyachi and Atsuo Fukuda Chapter VII: Synthesis and Structural Features. . . Andrew N. Cammidge and Richard J. Bushby
49 1
. . . . . . . . . 493 . . . . . . . . . 493 .
........
. . .
5 15
. . . . . . 665
..................
693
Chapter VIII: Discotic Liquid Crystals: Their Structures and Physical Properties . . . 749 S. Chandrasekhar Chapter IX: Applicable Properties of Columnar Discotic Liquid Crystals Neville Boden and Bijou Movaghar
,
. . . . . . 781
Outline
XI11
Part 3 : Non-Conventional Liquid-Crystalline Materials . . . . . . . . . . . . . . . . 799 Chapter X: Liquid Crystal Dimers and Oligomers . . . . . . . . . . . . . . . . . . . 801 Corrie 7: Imrie and Geoffrey R. Luckhurst Chapter XI: Laterally Substituted and Swallow-Tailed Liquid Crystals . . . . . . . . 835 Wolfgang Weissflog Chapter XII: Phasmids and Polycatenar Mesogens . . . . . . . . . . . . . . . . . . . 865 Huu-Tinh Nguyen. Christian Destrade. and Jacques Malth2te Chapter XIII: Thermotropic Cubic Phases . . . . . . . . . . . . . . . . . . . . . . . Siegmar Diele and Petra Goring
887
Chapter XIV Metal-containing Liquid Crystals . . . . . . . . . . . . . . . . . . . . Anne Marie Giroud-Godquin
901
Chapter XV: Biaxial Nematic Liquid Crystals . . . . . . . . . . . . . . . . . . . . . B. K. Sadashiva
933
Chapter XVI: Charge-Transfer Systems . . . . . . . . . . . . . . . . . . . . . . . . Klaus Praefcke and D . Singer
945
Chapter XVII: Hydrogen-Bonded Systems . . . . . . . . . . . . . . . . . . . . . . . Takashi Kato
969
Chapter XVIII: Chromonics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . John Lydon
981
Index Volumes 2 A and 2 B . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
1009
Volume 3 Part 1: Main-Chain Thermotropic Liquid-Crystalline Polymers . . . . . . . . . . . . . 1 Chapter I: Synthesis. Structure and Properties . . . . . . . . . . . . . . . . . . . . . . 3 Aromatic Main Chain Liquid Crystalline Polymers . . . . . . . . . . . . . . 3 1 Andreas Greiner and Hans- Werner Schmidt Main Chain Liquid Crystalline Semiflexible Polymers . . . . . . . . . . . . 26 2 Em0 Chiellini and Michele Laus Combined Liquid Crystalline Main-Chain/Side-Chain Polymers . . . . . . . 52 3 Rudolf Zentel BlockCopolymers Containing Liquidcrystalline Segments . . . . . . . . . 66 4 Guoping Ma0 and Christopher K. Ober
XIV
Outline
Chapter 11: Defects and Textures in Nematic Main-Chain Liquid Crystalline Polymers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Claudine Noel
93
Part 2: Side-Group Thermotropic Liquid-Crystalline Polymers . . . . . . . . . . . . 121 Chapter 111: Molecular Engineering of Side Chain Liquid Crystalline Polymers by Living Polymerizations . . . . . . . . . . . . . . . . . . . . . . . . . . Coleen Pugh and Alan L. Kiste
123
Chapter I V Behavior and Properties of Side Group Thermotropic Liquid Crystal Polymers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Jean-Claude Dubois, Pierre Le Barny, Monique Mauzac, and Claudine Noel
207
Chapter V: Physical Properties of Liquid Crystalline Elastomers . . . . . . . . . . . 277 Helmut R. Brand and Heino Finkelmann
Part 3: Amphiphilic Liquid Crystals . . . . . . . . . . . . . . . . . . . . . . . . . .
303
Chapter VI: Amphotropic Liquid Crystals . . . . . . . . . . . . . . . . . . . . . . . Dieter Blunk, Klaus Praefcke and Volkmar Vill
305
Chapter VII: Lyotropic Surfactant Liquid Crystals . . . . . . . . . . . . . . . . . . . 34 1 C. Fairhurst, S. Fullec J. Gray, M. C. Holmes, G. J. 7: Eddy Chapter VIII: Living Systems. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Siegfried Hoffmann
393
Chapter IX: Cellulosic Liquid Crystals . . . . . . . . . . . . . . . . . . . . . . . . . Peter Zugenmaier
45 1
Index Volumes 1 - 3. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
483
Contents Volume 2 A
Part I: Calamitic Liquid Crystals . . . . . . . . . . . . . . . . . . . . . . .
1
Chapter I: Phase Structures of Calamitic Liquid Crystals . . . . . . . . . . . . . . 3 John W Goodby 1
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
2
Melting Processes of Calamitic Thermotropic Liquid Crystals . . . . . . . . . 4
3 3.1 3.2 3.3 3.3.1 3.3.2 3.3.3 3.3.4 3.4 3.4.1 3.4.2 3.4.3 3.4.4 3.4.5
Structures of Calamitic Liquid Crystals . . . . . . . . . . . . . . . . . . . . . 6 The Nematic Phase . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 Structures of Smectic Liquid Crystals . . . . . . . . . . . . . . . . . . . . . . 7 The Structures of the Orthogonal Smectic Phases . . . . . . . . . . . . . . . 7 Structure of the Smectic A Phase . . . . . . . . . . . . . . . . . . . . . . . . 7 Structure in the Hexatic B Phase . . . . . . . . . . . . . . . . . . . . . . . . 10 Structure of the Crystal B Phase . . . . . . . . . . . . . . . . . . . . . . . . 10 Structure of Crystal E . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12 Structures of the Tilted Smectic Phases . . . . . . . . . . . . . . . . . . . . . 13 Structure of the Smectic C Phase . . . . . . . . . . . . . . . . . . . . . . . . 13 Structure of the Smectic I Phase . . . . . . . . . . . . . . . . . . . . . . . . 16 Structure of the Smectic F Phase . . . . . . . . . . . . . . . . . . . . . . . . 16 Structures of the Crystal J and G Phases . . . . . . . . . . . . . . . . . . . . 17 Structures of the Crystal H and K Phases . . . . . . . . . . . . . . . . . . . . 18
4
Long- and Short-Range Order . . . . . . . . . . . . . . . . . . . . . . . . . .
18
5
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
20
3
Chapter 11: Phase Transitions in Rod-Like Liquid Crystals . . . . . . . . . . . . . 23 Daniel Guillon 1
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
2 2.1 2.2 2.3 2.4
Isotropic-Nematic (Iso-N) Transition . . . . . . . . . . . . . . . . . . . . . 23 Brief Summary of the Landau-de Gennes Model . . . . . . . . . . . . . . . . 23 Magnetic Birefringence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24 Light Scattering . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24 Deviations from the Landau-de Gennes Model . . . . . . . . . . . . . . . . . 25
23
XVI
Contents
3 3.1 3.2
Nematic-Smectic A (N-SmA) Transition . . . . . . . . . . . . . . . . . . . . 26 26 The McMillan-de Gennes Approach . . . . . . . . . . . . . . . . . . . . . . Critical Phenomena: Experimental Situation . . . . . . . . . . . . . . . . . . 26
4 4.1 4.2 4.3 4.4 4.5 4.6
Smectic A-Smectic C (SmA-SmC) Transition . . . . . . . . . . . . . . . . . 29 29 General Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30 Critical Behavior . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30 Experimental Situation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Smectic A-Smectic C* (SmA-SmC*) Transition . . . . . . . . . . . . . . . . 32 The Nematic-Smectic A-Smectic C (NAC) Multicritical Point . . . . . . . . 33 35 SmA-SmC Transition in Thin Films . . . . . . . . . . . . . . . . . . . . . .
5 5.1 5.2
Hexatic B to Smectic A (SmBhex-SmA) transition . . . . . . . . . . . . . . 36 General Presentation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36 SmBhex-SmA Transition in Thin Films . . . . . . . . . . . . . . . . . . . . 37
6 6.1 6.2 6.3
Induced Phase Transitions . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38 Mechanically Induced SmA-SmC Transition . . . . . . . . . . . . . . . . . . 38 Electrically Induced Transitions . . . . . . . . . . . . . . . . . . . . . . . . 39 39 Photochemically Induced Transitions . . . . . . . . . . . . . . . . . . . . . .
7 7.1 7.2 7.3 7.4
Other Transitions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Smectic C to Smectic I (SmC-SmI) Transition . . . . . . . . . . . . . . . . Smectic C to Smectic F (SmC-SmF) Transition . . . . . . . . . . . . . . . Smectic F to Smectic I (SmF-SmI) Transition . . . . . . . . . . . . . . . . Smectic F to Smectic Crystalline G (SmF-SmG) Transition . . . . . . . . .
8
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
43
Chapter 111: Nematic Liquid Crystals . . . . . . . . . . . . . . . . . . . . . . . . .
47
1 1.1
1.2 f.3 1.4 1.5 1.6 1.7 1.8 1.9 1.10 1.11 1.12
. . . .
41 41 41 42 42
47 Synthesis of Nematic Liquid Crystals . . . . . . . . . . . . . . . . . . . . . . Kenneth J. Toyne Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47 Benzene. Biphenyl and Terphenyl Systems . . . . . . . . . . . . . . . . . . . 48 49 Cyclohexane Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1,4-Disubstituted-bicyclo[2.2.2]octanes. . . . . . . . . . . . . . . . . . . . 50 51 2,5.Disubstituted.l, 3.dioxanes . . . . . . . . . . . . . . . . . . . . . . . . . 2,5.Disubstitute d.pyridines . . . . . . . . . . . . . . . . . . . . . . . . . . . 51 2,5.Disubstituted.pyrimidines . . . . . . . . . . . . . . . . . . . . . . . . . . 52 3,6.Disubstituted.pyridazines . . . . . . . . . . . . . . . . . . . . . . . . . . 52 52 Naphthalene systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Unusual Core Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53 Ester Linkages . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53 54 Lateral Substitution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Contents
1.13 1.14 1.15
XVII
4.c.(trans.4.A1kyicyc1ohexyl).1.alkyl.r.1.cyanocyclohexanes . . . . . . . . . 55 Terminal Groups . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56 56 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
60 Physical Properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Elastic Properties of Nematic Liquid Crystals . . . . . . . . . . . . . . . . . 60 Ralf Stannarius 2.1.1 Introduction to Elastic Theory . . . . . . . . . . . . . . . . . . . . . . . . . 60 2.1.2 Measurement of Elastic Constants . . . . . . . . . . . . . . . . . . . . . . . 63 64 2.1.2.1 Frkedericksz Transition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68 2.1.2.2 Light Scattering Measurements . . . . . . . . . . . . . . . . . . . . . . . . . Other Experiments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68 2.1.2.3 . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69 Experimental Elastic Data 2.1.3 71 2.1.4 MBBA and n-CB . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73 2.1.5 ‘Surface-like’ Elastic Constants . . . . . . . . . . . . . . . . . . . . . . . . . 79 2.1.6 Theory of Elastic Constants . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.1.7 Biaxial Nematics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83 2.1.8 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84 2 2.1
2.2 2.2.1 2.2.2 2.2.3 2.2.4 2.2.5 2.2.6 2.3 2.3.1 2.3.2 2.3.2.1 2.3.2.2 2.3.2.3 2.3.2.4 2.3.2.5 2.3.3 2.3.4 2.3.5 2.3.6 2.4 2.4.1 2.4.2
Dielectric Properties of Nematic Liquid Crystals . . . . . . . . . . . . . . . . 91 Horst Kresse 91 Rod-like Molecules in the Isotopic State . . . . . . . . . . . . . . . . . . . . Static Dielectric Constants of Nematic Samples . . . . . . . . . . . . . . . . 92 The Nre Phenomenon and the Dipolar Correlation . . . . . . . . . . . . . . . 98 99 Dielectric Relaxation in Nematic Phases . . . . . . . . . . . . . . . . . . . . Dielectric Behavior of Nematic Mixtures . . . . . . . . . . . . . . . . . . . 102 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 109 Diamagnetic Properties of Nematic Liquid Crystals . . . . . . . . . . . . . Ralf Stannarius Magnetic Quantities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Measurement of Diamagnetic Properties . . . . . . . . . . . . . . . . . . Faraday-Curie Balance Method . . . . . . . . . . . . . . . . . . . . . . . . Supraconducting Quantum Interference Devices Measurements . . . . . . NMR Measurements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Magneto-electric Method . . . . . . . . . . . . . . . . . . . . . . . . . . . Mechanical Torque Measurements . . . . . . . . . . . . . . . . . . . . . . . Experimental Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Increment System for Diamagnetic Anisotropies . . . . . . . . . . . . . . Application of Diamagnetic Properties . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
. 113 113 . 116 116 . 116 117 117 118 118 . 124 125 126
Optical Properties of Nematic Liquid Crystals . . . . . . . . . . . . . . . . 128 Gerhard Pelzl Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 128 Experimental Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 128
XVIII 2.4.3 2.4.4 2.4.5 2.4.6 2.4.7 2.4.8 2.4.9 2.5 2.5.1 2.5.2 2.5.2.1 2.5.2.2 2.5.2.3 2.5.2.4 2.5.2.5 2.5.3 2.5.3.1 2.5.3.2 2.5.3.3 2.5.3.4 2.5.4 2.5.4.1 2.5.4.2 2.5.4.3 2.5.5 2.6 2.6.1 2.6.2 2.6.3 2.6.4 2.6.5 2.6.6 2.6.7 2.6.8 2.6.9 2.6.10 2.6.1 1 2.6.12 2.6.13 2.6.14
Contents
Temperature Dependence of Birefringence and Refractive Indices . . . . . . 132 133 Dispersion of ne. no and An . . . . . . . . . . . . . . . . . . . . . . . . . . Refractive Indices of Mixtures . . . . . . . . . . . . . . . . . . . . . . . . . 135 136 Birefringence in Homologous Series . . . . . . . . . . . . . . . . . . . . . Determination of Molecular Polarizability Anisotropy and Orientational 136 Order from Birefringence Data . . . . . . . . . . . . . . . . . . . . . . . . Relationships between Birefringence and Molecular Structure . . . . . . . . 137 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 140 Viscosity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Herbert Kneppe and Frank Schneider Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Determination of Shear Viscosity Coefficients . . . . . . . . . . . . . . . General Aspects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Mechanical methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Light Scattering . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Other Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Experimental Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Determination of Rotational Viscosity . . . . . . . . . . . . . . . . . . . . . General Aspects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Experimental Methods with Permanent Director Rotation . . . . . . . . . Relaxation Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Experimental Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Leslie Coefficients . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Determination from Shear and Rotational Viscosity Coefficients . . . . . . Determination by Means of Light Scattering . . . . . . . . . . . . . . . . Other Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
142 142
. 142 142 143 147 1SO 150 155 155 . 156 157 160 165 . 165 . 166 167 167
Dynamic Properties of Nematic Liquid Crystals . . . . . . . . . . . . . . . R. Blinc and I . MuSevic' Quasielastic Light Scattering in Nematics . . . . . . . . . . . . . . . . . . . Nuclear Magnetic Resonance in Nematics . . . . . . . . . . . . . . . . . . . Quasielectric Light Scattering and Order Fluctuations in the Isotropic Phase Nuclear Magnetic Resonance and Order Fluctuations in the Isotropic Phase . Quasielastic Light Scattering and Orientational Fluctuations below Tc . . . Nuclear Magnetic Resonance and Orientational Fluctuations below Tc . . . . Optical Kerr Effect and Transient Laser-Induced Molecular Reorientation . . Dielectric Relaxation in Nematics . . . . . . . . . . . . . . . . . . . . . . . Pretransitional Dynamics Near the Nematic-Smectic A Transition . . . . . . Dynamics of Nematics in Micro-Droplets and Micro-Cylinders . . . . . . . Pretransitional Effects in Confined Geometry . . . . . . . . . . . . . . . . . Dynamics of Randomly Constrained Nematics . . . . . . . . . . . . . . . . Other Observations. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
170 170 173 174 175 177 177 181 182 183 184 188 189 190 191
Contents
3 3.1 3.1.1 3.1.2 3.1.2.1 3.1.2.2 3.1.2.3 3.1.3 3.1.3.1 3.1.3.2 3.1.3.3 3.1.3.4 3.1.4 3.1.4.1 3.1.4.2 3.1.4.3 3.1.4.4 3.1.4.5 3.1.4.6 3.1.4.7 3.1.4.8 3.1.5 3.1.5.1 3.1.5.2 3.1.5.3 3.1.5.4 3.1.6 3.2 3.2.1 3.2.1.1 3.2.1.2 3.2.1.3 3.2.1.4 3.2.1.5 3.2.2 3.2.3 3.2.4 3.2.4.1 3.2.4.2 3.2.5 3.2.6
XIX
Applications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 199 199 TN. STNDisplays . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Harald Hirschrnann and Volker Reiffenrath Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 199 200 Twisted Nematic Displays . . . . . . . . . . . . . . . . . . . . . . . . . . . Configuration and Operation Principles of Twisted Nematic Displays . . . . 200 Optical Properties of the Unactivated State . . . . . . . . . . . . . . . . . . 200 Optical Properties of the Activated State . . . . . . . . . . . . . . . . . . . 202 204 Addressing of Liquid Crystal Displays . . . . . . . . . . . . . . . . . . . . 205 Direct Addressing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 205 Passive Matrix Addressing . . . . . . . . . . . . . . . . . . . . . . . . . . . The Improved Alt-Pleshko Addressing Technique . . . . . . . . . . . . . . 207 Generation of Gray Levels . . . . . . . . . . . . . . . . . . . . . . . . . . . 207 208 Supertwisted Nematic Displays . . . . . . . . . . . . . . . . . . . . . . . . Influence of Device and Material Parameters . . . . . . . . . . . . . . . . . 208 Configuration and Transmission of a Supertwisted Nematic Display . . . . . 211 Electro-optical Performance of Supertwisted Nematic Displays . . . . . . . 213 Dynamical Behavior of Twisted Nematic and Supertwisted Nematic Displays 2 13 Color Compensation of STN Displays . . . . . . . . . . . . . . . . . . . . . 215 Viewing Angle and Brightness Enhancement . . . . . . . . . . . . . . . . . 218 218 Color Supertwisted Nematic Displays . . . . . . . . . . . . . . . . . . . . . Fast Responding Supertwisted Nematic Liquid Crystal Displays . . . . . . . 219 Liquid Crystal Materials for Twisted Nematic and Supertwisted Nematic 220 Display Devices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Materials with High Optimal Anisotropy . . . . . . . . . . . . . . . . . . . 221 Materials with Positive Dielectric Anisotropy . . . . . . . . . . . . . . . . . 221 Materials for the Adjustment of the Elastic Constant Ratio K33/Kll . . . . 225 226 Dielectric Neutral Basic Materials . . . . . . . . . . . . . . . . . . . . . . . 227 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Active Matrix Addressed Displays . . . . . . . . . . . . . . . . . . . . . . Eiji Kaneko Thin Film Diode and Metal-Insulator-Metal Matrix Address . . . . . . . . . Diode Ring Matrix Address . . . . . . . . . . . . . . . . . . . . . . . . . . Back-to-back Diode Matrix Address . . . . . . . . . . . . . . . . . . . . . . Two Branch Diode Matrix Address . . . . . . . . . . . . . . . . . . . . . . SiNx Thin Film Diode Matrix Address . . . . . . . . . . . . . . . . . . . . Metal-Insulator-Metal Matrix Address . . . . . . . . . . . . . . . . . . . . CdSe Thin Film Transistor Switch Matrix Address . . . . . . . . . . . . . . a-Si Thin Film Transistor Switch Matrix Address . . . . . . . . . . . . . . . p-Si Thin Film Transistor Switch Matrix Address . . . . . . . . . . . . . . . Solid Phase Crystallization Method . . . . . . . . . . . . . . . . . . . . . . Laser Recrystallization Method . . . . . . . . . . . . . . . . . . . . . . . . Metal-oxide Semiconductor Transistor Switch Matrix Address . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
230 230 230 230 231 231 232 233 234 237 238 238 239 240
xx 3.3 3.3.1 3.3.2 3.3.2.1 3.3.2.2 3.3.2.3 3.3.3 3.3.4 3.3.4.1 3.3.4.2 3.3.5 3.3.6 3.3.6.1 3.3.6.2 3.3.6.3 3.3.6.4 3.3.6.5 3.3.6.6 3.3.7 3.4 3.4.1 3.4.2 3.4.2.1 3.4.3 3.4.4 3.4.4.1 3.4.4.2 3.4.5 3.4.6 3.4.6.1 3.4.6.2 3.4.6.3 3.4.6.4 3.4.7 3.4.7.1 3.4.7.2 3.4.8 3.4.8.1
Contents
Dynamic Scattering . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 243 Birendra Bahadur Introduction and Cell Designing . . . . . . . . . . . . . . . . . . . . . . . . 243 Experimental Observations at DC (Direct Current) and Low Frequency AC (Alternating Current) Fields . . . . . . . . . . . . . . . . . . . . . . . . . . 244 Homogeneously Aligned Nematic Regime . . . . . . . . . . . . . . . . . . 244 Williams Domains . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 244 Dynamic Scattering . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 246 Observations at High Frequency AC Field . . . . . . . . . . . . . . . . . . 247 247 Theoretical Explanations . . . . . . . . . . . . . . . . . . . . . . . . . . . . 248 Carr-Helfrich Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Dubois-Violette, de Gennes, and Parodi Model . . . . . . . . . . . . . . . . 249 Dynamic Scattering in Smectic A and Cholesteric Phases . . . . . . . . . . 250 Electrooptical Characteristics and Limitations . . . . . . . . . . . . . . . . 251 Contrast Ratio Versus Voltage, Viewing Angle, Cell Gap, Wavelength, and Temperature . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 251 Display Current Versus Voltage, Cell Gap, and Temperature . . . . . . . . . 252 Switching Time . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 253 Effect of Conductivity, Temperature and Frequency . . . . . . . . . . . . . 253 Addressing of DSM (Dynamic Scattering Mode) LCDs 254 (Liquid Crystal Displays) . . . . . . . . . . . . . . . . . . . . . . . . . . . Limitations of DSM LCDs . . . . . . . . . . . . . . . . . . . . . . . . . . . 254 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 254 Guest-Host Effect . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Birendra Bahadur Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Dichroic Dyes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Chemical Structure, Photostability. and Molecular Engineering . . . . . . . Cell Preparation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Dichroic Parameters and Their Measurement . . . . . . . . . . . . . . . . . Order Parameter and Dichroic Ratio of Dyes . . . . . . . . . . . . . . . . . Absorbance, Order Parameter, and Dichroic Ratio Measurement . . . . . . . Impact of Dye Structure and Liquid Crystal Host on the Physical Properties of a Dichroic Mixture . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Optical, Electro-Optical, and Life Parameters . . . . . . . . . . . . . . . . . Luminance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Contrast and Contrast Ratio . . . . . . . . . . . . . . . . . . . . . . . . . . Switching Speed . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Life Parameters and Failure Modes . . . . . . . . . . . . . . . . . . . . . . Dichroic Mixture Formulation . . . . . . . . . . . . . . . . . . . . . . . . . Monochrome Mixture . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Black Mixture . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Heilmeier Displays . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Threshold Characteristic . . . . . . . . . . . . . . . . . . . . . . . . . . . .
257 257 259 260 266 268 268 269 271 271 272 272 273 273 274 274 274 275 276
XXI
Contents
3.4.8.2 Effects of Dye Concentration on Electro-optical Parameters . . . . . . . . 3.4.8.3 Effect of Cholesteric Doping . . . . . . . . . . . . . . . . . . . . . . . . . . 3.4.8.4 Effect of Alignment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.4.8.5 Effect of Thickness . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.4.8.6 Impact of the Order Parameter . . . . . . . . . . . . . . . . . . . . . . . . . 3.4.8.7 Impact of the Host . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.4.8.8 Impact of the Polarizer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.4.8.9 Color Applications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.4.8.10 Multiplexing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.4.9 Quarter Wave Plate Dichroic Displays . . . . . . . . . . . . . . . . . . . . . 3.4.10 Dye-doped TN Displays . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.4.11 Phase Change Effect Dichroic LCDs . . . . . . . . . . . . . . . . . . . . . 3.4.11.1 Threshold Characteristic and Operating Voltage. . . . . . . . . . . . . . . 3.4.1 1.2 Contrast Ratio, Transmission Brightness, and Switching Speed . . . . . . 3.4.11.3 Memory or Reminiscent Contrast . . . . . . . . . . . . . . . . . . . . . . . 3.4.11.4 Electro-optical Performance vs Temperature . . . . . . . . . . . . . . . . 3.4.1 1.5 Multiplexing Phase Change Dichroic LCDs . . . . . . . . . . . . . . . . . 3.4.12 Double Cell Dichroic LCDs . . . . . . . . . . . . . . . . . . . . . . . . . . 3.4.12.1 Double Cell Nematic Dichroic LCD . . . . . . . . . . . . . . . . . . . . . . 3.4.12.2 Double Cell One Pitch Cholesteric LCD . . . . . . . . . . . . . . . . . . 3.4.12.3 Double Cell Phase Change Dichroic LCD . . . . . . . . . . . . . . . . . . 3.4.13 Positive Mode Dichroic LCDs . . . . . . . . . . . . . . . . . . . . . . . . . 3.4.13.1 Positive Mode Heilmeier Cells . . . . . . . . . . . . . . . . . . . . . . . . 3.4.13.2 Positive Mode Dichroic LCDs Using a AJ4 Plate . . . . . . . . . . . . . . 3.4.13.3 Positive Mode Double Cell Dichroic LCD . . . . . . . . . . . . . . . . . 3.4.13.4 Positive Mode Dichroic LCDs Using Special Electrode Patterns . . . . . . 3.4.13.5 Positive Mode Phase Change Dichroic LCDs . . . . . . . . . . . . . . . . 3.4.13.6 Dichroic LCDs Using an Admixture of Pleochroic and Negative Dichroic Dyes . . . . . . . . . . . . . . . . . . . . . . . . . . 3.4.14 Supertwist Dichroic Effect Displays . . . . . . . . . . . . . . . . . . . . . . 3.4.15 Ferroelectric Dichroic LCDs . . . . . . . . . . . . . . . . . . . . . . . . . . 3.4.15.1 Devices Using A Single Polarizer . . . . . . . . . . . . . . . . . . . . . . . 3.4.15.2 Devices Using No Polarizers . . . . . . . . . . . . . . . . . . . . . . . . . 3.4.16 Polymer Dispersed Dichroic LCDs . . . . . . . . . . . . . . . . . . . . . . 3.4.17 Dichroic Polymer LCDs . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.4.18 Smectic A Dichroic LCDs . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.4.19 Fluorescence Dichroic LCDs . . . . . . . . . . . . . . . . . . . . . . . . . 3.4.20 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Chapter 1% Chiral Nematic Liquid Crystals . . . . . . . . . . . . . . . . . . . . 1 1.1
. 276
. . . . . . . . . .
278 278 279 279 279 281 281 281 282 283 284 286 287 289 291 291 291 291 292 292 292 292 295 295 295 295 296 296 296 297 297 297 298 298 299 299
303
The Synthesis of Chiral Nematic Liquid Crystals . . . . . . . . . . . . . . . 303 Christopher J . Booth Introduction to the Chiral Nematic Phase and its Properties . . . . . . . . . 303
XXII 1.2 1.3 1.3.1 1.4 1.5 1.5.1 1.5.2 1.5.3 1.5.4 1S . 5 1.6 1.6.1 1.6.2 1.6.3 1.7 1.7.1 1.7.2 1.7.3 1.74 1.8
1.9 2 2.1 2.2 2.2.1 2.2.1.1 2.2.1.2 2.2.1.3 2.2.1.4 2.2.2 2.2.2.1 2.3 2.3.1 2.3.2 2.3.3 2.4 2.4.1 2.4.2 2.4.3 2.4.4 2.5 2.5.1
Contents
Formulation and Applications of ThermochromicMixtures . . . . . . . . . . 305 Classification of Chiral Nematic Liquid Crystalline Compounds . . . . . . . 307 Aspects of Molecular Symmetry for Chiral Nematic Phases . . . . . . . . . 308 310 Cholesteryl and Related Esters . . . . . . . . . . . . . . . . . . . . . . . . . Type I Chiral Nematic Liquid Crystals . . . . . . . . . . . . . . . . . . . . 311 Azobenzenes and Related Mesogens . . . . . . . . . . . . . . . . . . . . . 312 313 Azomethine (Schiff’s Base) Mesogens . . . . . . . . . . . . . . . . . . . . Stable Phenyl. Biphenyl. Terphenyl and Phenylethylbiphenyl Mesogens . . . 3 14 (R)-2-(4-Hydroxyphenoxy)-propanoicAcid Derivatives . . . . . . . . . . . 319 Miscellaneous Type I Chiral Nematic Liquid Crystals . . . . . . . . . . . . 323 Type I1 Chiral Nematic Liquid Crystals . . . . . . . . . . . . . . . . . . . . 325 Azomethine Ester Derivatives of (R)-3-Methyladipic Acid . . . . . . . . . . 325 Novel Highly Twisting Phenyl and 2-Pyrimidinylphenyl Esters 326 of (R)-3-Methyladipic Acid . . . . . . . . . . . . . . . . . . . . . . . . . . Chiral Dimeric Mesogens Derived from Lactic Acid or 1.2.Diols . . . . . . 327 Type 111 Chiral Nematic Liquid Crystals . . . . . . . . . . . . . . . . . . . . 327 Tricyclo[4.4.0.03,8]decane or Twistane Derived Mesogens . . . . . . . . . . 328 Axially Chiral Cyclohexylidene-ethanones . . . . . . . . . . . . . . . . . . 328 330 Chiral Heterocyclic Mesogens . . . . . . . . . . . . . . . . . . . . . . . . . Chiral Mesogens Derived from Cyclohexane . . . . . . . . . . . . . . . . . 331 332 Concluding Remarks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 332 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Chiral Nematics: Physical Properties and Applications . . . . . . . . . . . . Harry Coles Introduction to Chiral Nematics: General Properties . . . . . . . . . . . . . Static Properties of Chiral Nematics . . . . . . . . . . . . . . . . . . . . . . Optical Properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Textures and Defects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Optical Propagation (Wave Equation Approach) . . . . . . . . . . . . . . . Optical Propagation (‘Bragg’ Reflection Approach) . . . . . . . . . . . . . Pitch Behavior as a Function of Temperature, Pressure, and Composition . . Elastic Properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Continuum Theory and Free Energy . . . . . . . . . . . . . . . . . . . . . . Dynamic Properties of Chiral Nematics . . . . . . . . . . . . . . . . . . . . Viscosity Coefficients . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Lehmann Rotation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Macroscopic Flow . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Field-Induced Distortions in Chiral Nematics . . . . . . . . . . . . . . . . . Magnetic Fields Parallel to the Helix Axis . . . . . . . . . . . . . . . . . . Magnetic Fields Normal to the Helix Axis . . . . . . . . . . . . . . . . . . Electric Fields Parallel to the Helix Axis . . . . . . . . . . . . . . . . . . . Electric Fields Normal to the Helix Axis . . . . . . . . . . . . . . . . . . . Applications of Chiral Nematics . . . . . . . . . . . . . . . . . . . . . . . . Optical: Linear and Nonlinear . . . . . . . . . . . . . . . . . . . . . . . . .
335 335 342 342 350 356 362 365 368 369 374 374 377 379 382 382 386 388 391 394 394
Contents
2.5.2 2.5.3 2.5.3.1 2.5.3.2 2.5.3.3 2.6 2.7
XXIII
397 Thermal Effects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 399 External Electric Field Effects . . . . . . . . . . . . . . . . . . . . . . . . . 400 Long Pitch Systems (p>A) . . . . . . . . . . . . . . . . . . . . . . . . . . Intermediate Pitch Length Systems (p =A). . . . . . . . . . . . . . . . . . . 401 403 Short Pitch Systems ( p d ) . . . . . . . . . . . . . . . . . . . . . . . . . . 404 Conclusions and the Future . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 405
Chapter V Non-Chiral Smectic Liquid Crystals . . . . . . . . . . . . . . . . . . 411 Synthesis of Non-Chiral Smectic Liquid Crystals . . . . . . . . . . . . . . . 411 John K Goodby Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 411 1.1 Template Structures for the Synthesis of Smectogens . . . . . . . . . . . . . 411 1.2 414 1.2.1 Terminal Aliphatic Chains . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.2.2 Polar Groups Situated at the End of the Core . . . . . . . . . . . . . . . . . 415 1.2.3 Functional Groups that Terminate the Core Structure . . . . . . . . . . . . . 417 418 1.2.4 Core Ring Structures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 420 1.2.5 Liking Groups . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 422 1.2.6 Lateral Substitutions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Syntheses of Standard Smectic Liquid Crystals . . . . . . . . . . . . . . . . 426 1.3 1.3.1 Synthesis of 4-Alkyl- and 4-alkoxy-4’-cyanophenyls:Interdigitated Smectic A Materials (e.g., 8CB and (80CB) . . . . . . . . . . . . . . . . . 426 Hexatic 1.3.2 Synthesis of 4-Alkyl-4-alkoxybiphenyl-4’-carboxylates: Smectic B Materials (e.g., 650BC) . . . . . . . . . . . . . . . . . . . . . . 427 1.3.3 Synthesis of 4-Alkyloxy-benzylidene-4-alkylanilines:Crystal B and 428 G Materials (e.g., nOms) . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.3.4 Synthesis of Terephthalylidene-bis-4-alkylanilines:Smectic I. Smectic F. Crystal G and Crystal H Materials (e.g., TBnAs) . . . . . . . . . . . . . . . 428 Smectic C Materials . . . 429 1.3.5 Synthesis of 4-Alkoxy-phenyl-4-alkoxybenzoates: Smectic C. 1.3.6 Synthesis of 4-Alkylphenyl-4-alkylbiphenyl-4’-carboxylates: 430 Smectic I. Hexatic Smectic B Materials . . . . . . . . . . . . . . . . . . . . 4-(2-Methylutyl)phenyl-4-alkoxybiphenyl-4’-carboxylates: Synthesis of I .3.7 Smectic C. Smectic I. Smectic F. Crystal J. Crystal K and Crystal G 430 Materials (nOmIs. e.g., 8OSI) . . . . . . . . . . . . . . . . . . . . . . . . . 1.3.8 Synthesis of 2-(4-n-Alkylphenyl)-5-(4-n-alkoxyoxyphenyl)pyrimidines: Smectic F and Crystal G Materials . . . . . . . . . . . . . . . . . . . . . . 431 1.3.9 Synthesis of 3-Nitro- and 3-Cyano-4-n-alkoxybiphenyl-4’-carboxylic Acids: Cubic and Smectic C Materials . . . . . . . . . . . . . . . . . . . . 433 1.3.10 Synthesis of bis-[ 1-(4’-Alkylbipheny1-4-y1) -3- (4-alkylpheny1)propane- 1 3-dionato]copper(11): Smectic 433 Metallomesogens . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Synthesis of Smectic Materials for Applications . . . . . . . . . . . . . . . 435 1.4 1.4.1 Synthesis of Ferroelectric Host Materials . . . . . . . . . . . . . . . . . . . 435 1
.
XXIV
1.4.2 1.5 1.6
Contents
Synthesis of Antiferroelectric Host Materials . . . . . . . . . . . . . . . . . 437 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 438 439 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Physical Properties of Non-Chiral Smectic Liquid Crystals . . . . . . . . . . 441 C. C. Huang 2.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 441 2.2 Smectic A Phase . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 443 443 2.2.1 Macroscopic Properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.2.2 X-ray Characterization of Free-Standing Smectic Films . . . . . . . . . . . 446 2.3 Hexatic Smectic B Phase . . . . . . . . . . . . . . . . . . . . . . . . . . . 447 2.3.1 Macroscopic Properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . 448 450 2.3.2 Thin Hexatic B Films . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 452 2.4 Smectic C Phase . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.4.1 Physical Properties near the Smectic A-Smectic C Transition . . . . . . . . 452 2.4.1.1 Bulk Properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 452 2.4.2 Macroscopic Behavior of the Smectic C Phase . . . . . . . . . . . . . . . . 457 457 2.4.2.1 Bulk Properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 461 2.5 Tilted Hexatic Phases . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.5.1 Identification of Hexatic Order in Thin Films . . . . . . . . . . . . . . . . . 461 2.5.2 Characterization of Hexatic Order in Thick Films . . . . . . . . . . . . . . . 462 464 2.5.3 Elastic Constants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 464 2.6 Surface Tension . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 467 2.7 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
3 3.1 3.2 3.3 3.3.1 3.3.1.1 3.3.1.2 3.3.2 3.3.2.1 3.3.2.2 3.3.2.3 3.3.3 3.3.3.1 3.3.3.2 3.3.3.3 3.3.3.4 3.3.3.5 3.4 3.5 3.5.1
Nonchiral Smectic Liquid Crystals - Applications . . . . . . . . . . . . . . 470 David Coates Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 470 471 Smectic Mesogens . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Laser-Addressed Devices . . . . . . . . . . . . . . . . . . . . . . . . . . . 473 Basic Effects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 473 473 Normal Mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 475 Reverse Mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 476 Materials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Liquid Crystals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 476 476 Lasers and Dyes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 477 Additives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Physical Characteristics and Applications . . . . . . . . . . . . . . . . . . . 478 478 Line Width and Write Speed . . . . . . . . . . . . . . . . . . . . . . . . . . 478 Contrast Ratios . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 478 Projection Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 478 Color Devices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 478 Commercial Devices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Thermally and Electrically Addressed Displays . . . . . . . . . . . . . . . . 481 482 Dielectric Reorientation of SmA Phases . . . . . . . . . . . . . . . . . . . . Materials of Negative Dielectric Anisotropy . . . . . . . . . . . . . . . . . 482
Contents
3.5.2 3.5.3 3.6 3.6.1 3.6.2 3.6.3 3.7 3.8 3.9 3.10
xxv
Materials of Positive Dielectric Anisotropy . . . . . . . . . . . . . . . . . . 482 A Variable Tilt SmA Device . . . . . . . . . . . . . . . . . . . . . . . . . . 482 Dynamic Scattering in SmA Liquid Crystal Phases . . . . . . . . . . . . . . 483 Theoretical Predictions . . . . . . . . . . . . . . . . . . . . . . . . . . . . 485 Response Times . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 485 Displays Based on Dynamic Scattering . . . . . . . . . . . . . . . . . . . . 486 Two Frequency Addressed SmA Devices . . . . . . . . . . . . . . . . . . . 486 Polymer-Dispersed Smectic Devices . . . . . . . . . . . . . . . . . . . . . 487 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 489 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 489
XXVI
Contents
Volume 2 B
Part 11: Discotic Liquid Crystals . . . . . . . . . . . . . . . . . . . . . . .
491
Chapter VI: Chiral Smectic Liquid Crystals . . . . . . . . . . . . . . . . . . . . . 1 Synthesis of Chiral Smectic Liquid Crystals . . . . . . . . . . . . . . . . . Stephen M . Kelly 1.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.2 Long Pitch Chiral Smectic Liquid Crystals or Dopants . . . . . . . . . . . . 1.2.1 Schiff’s bases . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.2.2 Aromatic Esters with Alkyl Branched Alkyl Chains . . . . . . . . . . . . . 1.2.3 Aromatic Heterocycles with Alkyl-Branched Alkyl Chains . . . . . . . . . . 1.2.4 Esters and Ethers in the Terminal Chain . . . . . . . . . . . . . . . . . . . . 1.2.5 Halogens at the Chiral Center . . . . . . . . . . . . . . . . . . . . . . . . . 1.2.6 Cyclohexyl a-Fluorohexanoates . . . . . . . . . . . . . . . . . . . . . . . . 1.2.7 Gyano Groups at the Chiral Center . . . . . . . . . . . . . . . . . . . . . . 1.2.8 Optically Active Oxiranes and Thiiranes . . . . . . . . . . . . . . . . . . . 1.2.9 Optically Active y-Lactones . . . . . . . . . . . . . . . . . . . . . . . . . . 1.2.10 Optically Active &Lactones . . . . . . . . . . . . . . . . . . . . . . . . . . 1.2.11 Miscellaneous Optically Active Heterocycles . . . . . . . . . . . . . . . . . 1.3 Short Pitch Chiral Smectic Liquid Crystals or Dopants . . . . . . . . . . . . 1.3.1 Optically Active Terphenyl Diesters . . . . . . . . . . . . . . . . . . . . . . 1.3.2 Optically Active Methyl-Substituted Dioxanes . . . . . . . . . . . . . . . . 1.4 Antiferroelectric Liquid Crystals . . . . . . . . . . . . . . . . . . . . . . . 1.5 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
493 493
2 2.1 2.2 2.2.1 2.2.2 2.2.3 2.2.4 2.2.5 2.2.6 2.2.7 2.3 2.3.1 2.3.2 2.3.3
Ferroelectric Liquid Crystals . . . . . . . . . . . . . . . . . . . . . . . . . Sven T. Lagerwall Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Polar Materials and Effects . . . . . . . . . . . . . . . . . . . . . . . . . . Polar and Nonpolar Dielectrics . . . . . . . . . . . . . . . . . . . . . . . . The Nonpolarity of Liquid Crystals in General . . . . . . . . . . . . . . . Behavior of Dielectrics in Electric Fields: Classification of Polar Materials Developments in the Understanding of Polar Effects . . . . . . . . . . . . The Simplest Description of a Ferroelectric . . . . . . . . . . . . . . . . . Improper Ferroelectrics . . . . . . . . . . . . . . . . . . . . . . . . . . . . The Piezoelectric Phase . . . . . . . . . . . . . . . . . . . . . . . . . . . . The Necessary Conditions for Macroscopic Polarization in a Material . . . The Neumann and Curie Principles . . . . . . . . . . . . . . . . . . . . . . Neumann’s Principle Applied to Liquid Crystals . . . . . . . . . . . . . . The Surface-Stabilized State . . . . . . . . . . . . . . . . . . . . . . . . . .
493 495 496 497 500 501 503 503 506 506 508 508 508 509 509 510 510 512 515
.
.
.
. .
.
515 520 520 522 523 527 531 536 539 541 541 542 544
Contents
2.3.4 2.3.5 2.3.6 2.3.7 2.4 2.4.1 2.4.2 2.4.3 2.4.4 2.4.5 2.4.6 2.4.7 2.4.8 2.4.9 2.5 2.5.1 2.5.2 2.5.3 2.5.4 2.5.5 2.5.6 2.5.7 2.5.8 2.5.9 2.5.10 2.5.11 2.6 2.6.1 2.6.2 2.6.3 2.6.4 2.6.5 2.6.6 2.6.7 2.6.8 2.6.9 2.7 2.7.1 2.7.2 2.7.3 2.7.4 2.7.5 2.7.6 2.7.7 2.7.8
XXVII
Chirality and its Consequences . . . . . . . . . . . . . . . . . . . . . . . . 548 The Curie Principle and Piezoelectricity . . . . . . . . . . . . . . . . . . . . 550 Hermann’s Theorem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 552 The Importance of Additional Symmetries . . . . . . . . . . . . . . . . . . 553 The Flexoelectric Polarization . . . . . . . . . . . . . . . . . . . . . . . . . 555 Deformations from the Ground State of a Nematic . . . . . . . . . . . . . . 555 The Flexoelectric Coefficients . . . . . . . . . . . . . . . . . . . . . . . . . 556 The Molecular Picture . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 557 Analogies and Contrasts to the Piezoelectric Effect . . . . . . . . . . . . . . 558 The Importance of Rational Sign Conventions . . . . . . . . . . . . . . . . 559 The Flexoelectrooptic Effect . . . . . . . . . . . . . . . . . . . . . . . . . . 560 Why Can a Cholesteric Phase not be Biaxial? . . . . . . . . . . . . . . . . . 563 Flexoelectric Effects in Smectic A Phases . . . . . . . . . . . . . . . . . . . 564 Flexoelectric Effects in Smectic C Phases . . . . . . . . . . . . . . . . . . . 564 The SmA*-SmC* Transition and the Helical C* State . . . . . . . . . . . . 568 The Smectic C Order Parameter . . . . . . . . . . . . . . . . . . . . . . . . 568 The SmA*-SmC* Transition . . . . . . . . . . . . . . . . . . . . . . . . . 571 The Smectic C* Order Parameters . . . . . . . . . . . . . . . . . . . . . . . 573 The Helical Smectic C* State . . . . . . . . . . . . . . . . . . . . . . . . . 574 The Flexoelectric Contribution in the Helical State . . . . . . . . . . . . . . 576 Nonchiral Helielectrics and Antiferroelectrics . . . . . . . . . . . . . . . . . 577 Simple Landau Expansions . . . . . . . . . . . . . . . . . . . . . . . . . . 578 The Electroclinic Effect . . . . . . . . . . . . . . . . . . . . . . . . . . . . 583 The Deformed Helix Mode in Short Pitch Materials . . . . . . . . . . . . . 587 The Landau Expansion for the Helical C* State . . . . . . . . . . . . . . . . 588 The Pikin-Indenbom Order Parameter . . . . . . . . . . . . . . . . . . . . . 592 Electrooptics in the Surface-Stabilized State . . . . . . . . . . . . . . . . . 596 The Linear Electrooptic Effect . . . . . . . . . . . . . . . . . . . . . . . . . 596 The Quadratic Torque . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 599 Switching Dynamics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 600 The Scaling Law for the Cone Mode Viscosity . . . . . . . . . . . . . . . . 602 Simple Solutions of the Director Equation of Motion . . . . . . . . . . . . . 603 Electrooptic Measurements . . . . . . . . . . . . . . . . . . . . . . . . . . 604 Optical Anisotropy and Biaxiality . . . . . . . . . . . . . . . . . . . . . . . 608 The Effects of Dielectric Biaxiality . . . . . . . . . . . . . . . . . . . . . . 610 The Viscosity of the Rotational Modes in the Smectic C Phase . . . . . . . . 613 Dielectric Spectroscopy: To Find the yand E Tensor Components . . . . . . 617 Viscosities of Rotational Modes . . . . . . . . . . . . . . . . . . . . . . . . 617 The Viscosity of the Collective Modes . . . . . . . . . . . . . . . . . . . . 618 The Viscosity of the Noncollective Modes . . . . . . . . . . . . . . . . . . 620 The Viscosity yq from Electrooptic Measurements . . . . . . . . . . . . . . 622 The Dielectric Permittivity Tensor . . . . . . . . . . . . . . . . . . . . . . . 622 The Case of Chiral Smectic C* Compounds . . . . . . . . . . . . . . . . . . 623 Three Sample Geometries . . . . . . . . . . . . . . . . . . . . . . . . . . . 625 Tilted Smectic Layers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 626
XXVIII
Contents
2.7.9 2.7.10 2.8 2.8.1 2.8.2 2.8.3 2.8.4 2.8.5 2.8.6 2.8.7 2.8.8 2.8.9 2.8.10 2.8.1 1 2.9 2.10
Nonchiral Smectics C . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Limitations in the Measurement Methods . . . . . . . . . . . . . . . . . . . FLC Device Structures and Local-Layer Geometry . . . . . . . . . . . . . . The Application Potential of FLC . . . . . . . . . . . . . . . . . . . . . . . Surface-Stabilized States . . . . . . . . . . . . . . . . . . . . . . . . . . . . FLC with Chevron Structures . . . . . . . . . . . . . . . . . . . . . . . . . Analog Gray Levels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Thin Walls and Thick Walls . . . . . . . . . . . . . . . . . . . . . . . . . . C1 and C2 Chevrons . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The FLC Technology Developed by Canon . . . . . . . . . . . . . . . . . . The Microdisplays of Displaytech . . . . . . . . . . . . . . . . . . . . . . . Idemitsu's Polymer FLC . . . . . . . . . . . . . . . . . . . . . . . . . . . . Material Problems in FLC Technology . . . . . . . . . . . . . . . . . . . . Nonchevron Structures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Is There a Future for Smectic Materials? . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
3
Antiferroelectric Liquid Crystals . . . . . . . . . . . . . . . . . . . . . . . 665 Kouichi Miyachi and Atsuo Fukuda Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 665 Origin of Antiferroelectricity in Liquid Crystals . . . . . . . . . . . . . . . 665 Biased or Hindered Rotational Motion in SmC* Phases . . . . . . . . . . . 665 Biased or Hindered Rotational Motion in SmCA* Phases . . . . . . . . . . . 669 Spontaneous Polarization Parallel to the Tilt Plane . . . . . . . . . . . . . . 671 Obliquely Projecting Chiral AlkylChains in SmA Phases . . . . . . . . . . 673 Thresholdless Antiferroelectricity andV-Shaped Switching . . . . . . . . . 675 Tristable Switching and the Pretransitional Effect . . . . . . . . . . . . . . . 675 Pretransitional Effect in Antiferroelectric Liquid Crystal Displays . . . . . . 679 Langevin-type Alignment in SmCR* Phases . . . . . . . . . . . . . . . . . 682 Antiferroelectric Liquid Crystal Materials . . . . . . . . . . . . . . . . . . . 684 Ordinary Antiferroelectric Liquid Crystal Compounds . . . . . . . . . . . . 684 Antiferroelectric Liquid Crystal Compounds with Unusual Chemical Structures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 688 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 689
3.1 3.2 3.2.1 3.2.2 3.2.3 3.2.4 3.3 3.3.1 3.3.2 3.3.3 3.4 3.4.1 3.4.2 3.5
627 628 629 629 630 634 637 639 644 648 650 651 653 655 658 660
Chapter VII: Synthesis and Structural Features . . . . . . . . . . . . . . . . . . 693 Andrew N . Carnrnidge and Richard J . Bushby 1
General Structural Features . . . . . . . . . . . . . . . . . . . . . . . . . .
693
2 2.1 2.1. I
Aromatic Hydrocarbon Cores . . . . . . . . . . . . . . . . . . . . . . . . . Benzene . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Esters and Amides . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
694 694 694
Contents
XXIX
2.1.2 2.1.3 2.1.4 2.2 2.3 2.4 2.5 2.5.1 2.5.2 2.5.3 2.5.4 2.5.5 2.6 2.7 2.8 2.9 2.10 2.11 2.12 2.13
Multiynes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 697 Hexakis(alkoxyphenoxymethy1) Derivatives . . . . . . . . . . . . . . . . . 698 Hexakis(alkylsu1fone) Derivatives . . . . . . . . . . . . . . . . . . . . . . . 698 Naphthalene . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 699 Anthracene (Rufigallol) . . . . . . . . . . . . . . . . . . . . . . . . . . . . 700 Phenanthrene . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 701 Triphenylene . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 702 Ethers. Thioethers and Selenoethers . . . . . . . . . . . . . . . . . . . . . . 702 Esters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 703 Multiynes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 704 Unsymmetrically Substituted Derivatives . . . . . . . . . . . . . . . . . . . 705 Modifications of the Number and Nature of Ring Substituents . . . . . . . . 708 Dibenzopyrene . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 712 Perylene . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 712 Truxene . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 713 Decacyclene . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 714 Tribenzocyclononatriene . . . . . . . . . . . . . . . . . . . . . . . . . . . . 715 Tetrabenzocyclododecatetraene . . . . . . . . . . . . . . . . . . . . . . . . 717 Metacyclophane . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 719 Phenylacetylene Macrocycles . . . . . . . . . . . . . . . . . . . . . . . . . 719
3 3.1 3.2 3.3 3.4 3.5 3.6 3.7 3.8 3.8.1 3.8.2 3.8.3 3.9 3.9.1 3.9.2 3.9.3 3.9.4 3.9.5 3.9.6 3.9.7 3.9.8 3.9.9 3.9.10 3.9.1 1
Heterocyclic Cores . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 720 Pyrillium . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 720 Bispyran . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 722 Condensed Benzpyrones (Flavellagic and Coruleoellagic Acid) . . . . . . . 723 Benzotrisfuran . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 723 Oxatruxene and Thiatruxene . . . . . . . . . . . . . . . . . . . . . . . . . . 724 Dithiolium . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 725 Tricycloquinazoline . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 726 Porphyrin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 727 Octa-Substituted Porphyrin . . . . . . . . . . . . . . . . . . . . . . . . . . 727 meso.Tetra(p.alkyl.phenyl).porphyrin . . . . . . . . . . . . . . . . . . . . . 729 Tetraazaporphyrin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 729 Phthalocyanine . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 730 Peripherally Substituted Octa(a1koxymethyl)phthalocyanine . . . . . . . . . 730 Peripherally Substituted 0cta.alkoxyphthalocyanines . . . . . . . . . . . . . 731 Peripherally Substituted Octa-alkylphthalocyanine . . . . . . . . . . . . . . 733 Tetrapyrazinoporphyrazine . . . . . . . . . . . . . . . . . . . . . . . . . . . 734 Peripherally Substituted Octa(alkoxycarbony1)phthalocyanines . . . . . . . 734 Peripherally Substituted Octa-(p-alkoxylpheny1)phthalocyanine . . . . . . . 735 Peripherally Substituted Tetrabenzotriazaporphyrin . . . . . . . . . . . . . . 735 Tetrakis[oligo(ethyleneoxy]phthalocyanine . . . . . . . . . . . . . . . . . . 736 Non-Peripherally Substituted Octa(alkoxymethy1)-phthalocyanines . . . . . 736 Non-Peripherally Substituted Octa-alkylphthalocyanine . . . . . . . . . . . 737 Unsymmetrically Substituted Phthalocyanines . . . . . . . . . . . . . . . . 739
xxx 4 4.1 4.2 4.3 5
Contents
Saturated Cores . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Cyclohexane . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Tetrahydropyran (Pyranose Sugars) . . . . . . . . . . . . . . . . . . . . . . Hexacyclens and Azamacrocyles . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
739 739 741 742 743
Chapter VIII: Discotic Liquid Crystals: Their Structures and Physical Properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . S . Chandrasekhar
749
1
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
749
2 2.1 2.1.1 2.1.2 2.2 2.3 2.4 2.5 2.6
Description of the Liquid . Crystalline Structures . . . . . . . . . . . . . . 750 The Columnar Liquid Crystal . . . . . . . . . . . . . . . . . . . . . . . . . 750 NMRStudies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 752 High Resolution X-Ray Studies . . . . . . . . . . . . . . . . . . . . . . . . 753 Columnar Phases of ‘Non-discotic’ Molecules . . . . . . . . . . . . . . . . 755 The Nematic Phase . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 757 757 The Columnar Nematic Phase . . . . . . . . . . . . . . . . . . . . . . . . . 758 The Chiral Nematic Phase . . . . . . . . . . . . . . . . . . . . . . . . . . . The Lamellar Phase . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 759
3
Extension of McMillan’s Model of Smectic A Phases to Discotic Liquid Crystals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
760
4
Pressure-Temperature Phase Diagrams . . . . . . . . . . . . . . . . . . . .
762
5
Techniques of Preparing Aligned Samples . . . . . . . . . . . . . . . . . . 764
6
Ferroelectricity in the Columnar Phase . . . . . . . . . . . . . . . . . . . .
7
The Columnar Structure as a One-Dimensional Antiferromagnet . . . . . . . 766
8
Electrical Conductivity in Columnar Phases . . . . . . . . . . . . . . . . . . 766
9
Photoconduction in Columnar Phases . . . . . . . . . . . . . . . . . . . . .
10 10.1 10.2 10.3 10.4 11 11.1 11.2 12 13 14
Continuum Theory of Columnar Liquid Crystals . : . . . . . . . . . . . . . 769 769 The Basic Equations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Acoustic Wave Propagation . . . . . . . . . . . . . . . . . . . . . . . . . . 770 Fluctuations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 771 772 Mechanical Instabilities . . . . . . . . . . . . . . . . . . . . . . . . . . . . 773 Defects in the Columnar Phase . . . . . . . . . . . . . . . . . . . . . . . . Dislocations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 774 Disclinations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 774 The Properties of Discotic Nematic Phases . . . . . . . . . . . . . . . . . . 775 776 Discotic Polymer Liquid Crystals . . . . . . . . . . . . . . . . . . . . . . . 777 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
765
768
Contents
XXXI
Chapter IX: Applicable Properties of Columnar Discotic Liquid Crystals . . . . 781 Neville Boden and Bijou Movaghar I
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
2 2.1 2.2
Molecular Structure-Property Relationships . . . . . . . . . . . . . . . . . . 782 HATn Materials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 782 783 The Phthalocyanines . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
3 3.1 3.2
Electrical Properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . HATn Materials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Chemically Doped Materials . . . . . . . . . . . . . . . . . . . . . . . . . .
784 784 787
4
Dielectric Properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
788
5
Fluorescence Properties . . . . . . . . . . . . . . . . . . . . . . . . . . . .
790
6
Gaussian Transit Characteristics . . . . . . . . . . . . . . . . . . . . . . . .
791
7 7.1 7.2
Anisotropy of Charge Mobility and Inertness to Oxygen . . . . . . . . . . . 792 Xerography . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 792 Phthalocyanine Based Gas Sensors . . . . . . . . . . . . . . . . . . . . . . 792
8
Selforganizing Periodicity, High Resistivity, and Dielectric Tunability . . . . 792
9 9.1 9.2
Ferroelectrics and Dielectric Switches . . . . . . . . . . . . . . . . . . . . . Nematic Phases . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Ferro- and Antiferroelectric Phases . . . . . . . . . . . . . . . . . . . . . .
10
Novel Absorption Properties, Fluorescence. and Fast Exciton Migration . . . 795
11
Applications of Doped Columnar Conductors . . . . . . . . . . . . . . . . . 795
12
Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
795
13
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
796
781
794 794 794
Part 3: Non-Conventional Liquid-Crystalline Materials . . . . . . . . 799 Chapter X: Liquid Crystal Dimers and Oligomers . . . . . . . . . . . . . . . . . 801 Corrie 7: Imrie and Geoflrey R . Luckhurst 1
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
2 3 3.1 3.2 3.3
Structure-Property Relationships inLiquidCrysta1 Dimers . . . . . . . . . . 801
801
Smectic Polymorphism . . . . . . . . . . . . . . . . . . . . . . . . . . . . Conventional Smectics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Intercalated Smectics . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
804 804 807
Modulated Smectics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
812
XXXII
Contents
4
Chiral Dimers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
5
Oligomeric Systems and Relation to Dimers . . . . . . . . . . . . . . . . . 813
6 6.1 6.2
Molecular Theories for Liquid Crystal Dimers . . . . . . . . . . . . . . . . 814 The Generic Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 815 822 A More Complete Model . . . . . . . . . . . . . . . . . . . . . . . . . . .
7
Molecular Shapes of Liquid Crystal Dimers . . . . . . . . . . . . . . . . . . 829
8
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
813
832
Chapter XI: Laterally Substituted and Swallow-Tailed Liquid Crystals . . . . . 835 Wolfgang Weissflog 1
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
2 2.1 2.2 2.3
Laterally Alkyl Substituted Rod-Like Mesogens . . . . . . . . Long-Chain 2-Substituted 1.4.Phenylene bis(Benzoates) . . . . Further mesogens bearing one Long-Chain Group in the Lateral Two Long-Chain Substituents in Lateral Positions . . . . . . .
3 3.1 3.2
Mesogens incorporating Phenyl Rings within the Lateral Segments . . . . . 843 Mesogens with One Lateral Segment containing a Phenyl Group . . . . . . . 843 Mesogens with Two Lateral Segments each containing a Phenyl Ring . . . . 850
4
Swallow-Tailed Mesogens . . . . . . . . . . . . . . . . . . . . . . . . . . .
850
5
Double-Swallow-Tailed Mesogens . . . . . . . . . . . . . . . . . . . . . .
855
6
Further Aspects and Concluding Remarks . . . . . . . . . . . . . . . . . . . 857
7
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
835
. . . . . . . 835 . . . . . . . 835 Position . . 837
. . . . . . . 841
860
Chapter XII: Plasmids and Polycatenar Mesogens . . . . . . . . . . . . . . . . . 865 Huu-Tinh Nguyen. Christian Destrade. and Jacques Malth2te 1
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
865
2
Nomenclature . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
865
3
Synthesis Principles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
866
4 4.1 4.1.1 4.1.2 4.1.3
Mesomorphic Properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . Polycatenars with Only Aliphatic Chains . . . . . . . . . . . . . . . . . . . Phasmids or Hexacatenar Mesogens 3mpm-3mpm . . . . . . . . . . . . . . Pentacatenars . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Tetracatenars . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
866 867 867 869 869
Contents
XXXIII
4.2 4.2.1 4.2.2
875 Polycatenars with Polar Substituents . . . . . . . . . . . . . . . . . . . . . Polycatenars with Hydrogenated and Fluorinated Chains . . . . . . . . . . . 875 Polycatenars with Other Polar Substituents . . . . . . . . . . . . . . . . . . 877
5 5.1 5.2
The Core. Paraffinic Chains. and Mesomorphic Properties . . . . . . . . . . 879 879 Core Length . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 880 Number and Position of Chains . . . . . . . . . . . . . . . . . . . . . . . .
6 6.1 6.1.1 6.1.2 6.1.3 6.1.4 6.2
Structures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Structures of Mesophases . . . . . . . . . . . . . . . . . . . . . . . . . . . Nematic . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Lamellar Mesophases . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Columnar Phases . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Cubic Mesophases . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Crystalline Structures . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
882 882 882 882 882 883 883
7
Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
884
8
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
884
Chapter XIII: Thermotropic Cubic Phases . . . . . . . . . . . . . . . . . . . . . Siegmar Diele and Petra Goring 1 Historical Remarks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
887 887
2
Chemical Structures and Phase Sequences . . . . . . . . . . . . . . . . . . 888
3 4
On the Structure of the Cubic Phases . . . . . . . . . . . . . . . . . . . . . Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
895 899
5
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
899
Chapter XIV Metal-containing Liquid Crystals . . . . . . . . . . . . . . . . . . 901 Anne Marie Giroud-Godquin 1 901 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
Early Work . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
3 3.1 3.2 3.2.1 3.2.1.1 3.2.1.2 3.2.1.3 3.2.2 3.2.2.1
Metallomesogens with Monodentate Ligands . . . . . . . . . . . . . . . . . 902 Organonitrile Ligands . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 902 902 n-Alkoxystilbazole Ligands . . . . . . . . . . . . . . . . . . . . . . . . . . Distilbazole Ligands . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 903 Palladium and Platinum . . . . . . . . . . . . . . . . . . . . . . . . . . . . 903 Silver . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 903 Iridium . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 903 903 Monostilbazole Ligands . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Platinum 904
901
XXXIV
Contents
3.2.2.2 3.2.2.3 3.3 3.3.1 3.3.2 3.4 3.5
Rhodium and Iridium . . . . . . . . . . . . . . . . Tungsten . . . . . . . . . . . . . . . . . . . . . . . Other Pyridine Ligands . . . . . . . . . . . . . . . Rhodium and Iridium . . . . . . . . . . . . . . . . Silver . . . . . . . . . . . . . . . . . . . . . . . . . Acetylide Ligands . . . . . . . . . . . . . . . . . . Isonitrile Ligands . . . . . . . . . . . . . . . . . .
............. ............. .............
4 4.1 4.1.1 4.1.2 4.1.3 4.1.4 4.2 4.2.1 4.2.2 4.2.3 4.2.4 4.2.5 4.2.6 4.3 4.4 4.4.1 4.4.2 4.4.2.1 4.2.2.2 4.4.2.3 4.4.3 4.5 4.5.1 4.5.2 4.5.3 4.5.4 4.6 4.6.1 4.6.2 4.6.3 4.6.4 4.6.5 4.7 4.7.1 4.7.1.1 4.7.1.2 4.7.1.3
Metallomesogens with Bidentate Ligands . . . . . . . . . . . . . . . . . . Carboxylate Ligands . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Alkali and Alkaline Earth Carboxylates . . . . . . . . . . . . . . . . . . . . Lead. Thallium. and Mercury Carboxylates . . . . . . . . . . . . . . . . . Dinuclear Copper Carboxylates . . . . . . . . . . . . . . . . . . . . . . . . Dinuclear Rhodium. Ruthenium. and Molybdenum Carboxylates . . . . . PDiketonate Ligands . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Copper Complexes with Two or Four Peripheral Chains . . . . . . . . . . Copper Complexes with Eight Peripheral Chains . . . . . . . . . . . . . . Malondialdehyde Complexes . . . . . . . . . . . . . . . . . . . . . . . . . Dicopper Complexes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Polymers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Other Metal Complexes . . . . . . . . . . . . . . . . . . . . . . . . . . . . Glycoximate Ligands . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Sulfur Containing Ligands . . . . . . . . . . . . . . . . . . . . . . . . . . . Dithiolene . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Dithiobenzoate . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Nickel and Palladium . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Zinc . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Silver . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Other Dithio Ligands . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . N-0 Donor Sets: Salicylaldimine Ligands . . . . . . . . . . . . . . . . . Copper, Nickel, and Palladium . . . . . . . . . . . . . . . . . . . . . . . . . Platinum, Vanadyl, and Iron . . . . . . . . . . . . . . . . . . . . . . . . . . Rhodium and Iridium . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Polymeric Liquid Crystals Based on Salcylaldimine Ligands . . . . . . . . Cyclometalated Complexes . . . . . . . . . . . . . . . . . . . . . . . . . . Azobenzene . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Arylimines . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Orthopalladated Diarylazines . . . . . . . . . . . . . . . . . . . . . . . . . Aroylhydrazine . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Orthopalladated Pyrimidine Complexes . . . . . . . . . . . . . . . . . . . . Metallocen Ligands . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Ferrocene . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Monosubstituted Ferrocene . . . . . . . . . . . . . . . . . . . . . . . . . . 1,1’Disubstituted Ferrocene . . . . . . . . . . . . . . . . . . . . . . . . . . 1,3 Disubstituted Ferrocene . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
....... ....... ....... .......
904 904 904 904 904 905 906
. 906 906 906 . 906 907 . 908 909 . 909 . 910 911 911 911 911 913 913 913 913 914 914 914 914 . 915 915 917 917 . 917 918 918 919 920 921 921 921 921 921 922 922
Contents
xxxv
Ruthenocene . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Iron Tricarbonyl Derivatives . . . . . . . . . . . . . . . . . . . . . . . . . .
923 923
5.1 5.1.1 5.1.2 5.1.3 5.1.4 5.2 5.3
Metallomesogens with Polydentate Ligands . . . . . . . . . . . . . . . . . . Phthalocyanine Ligands . . . . . . . . . . . . . . . . . . . . . . . . . . . . Copper Complexes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Manganese. Copper. Nickel. and Zinc Complexes . . . . . . . . . . . . . . Lutetium Complexes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Silicon. Tin. and Lead Complexes . . . . . . . . . . . . . . . . . . . . . . . Porphyrin Ligands . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Other Amine Ligands . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
923 923 923 924 924 924 925 926
5
Lyotropic Metal-Containing Liquid Crystals . . . . . . . . . . . . . . . . . 926
7
Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
927
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
928
Chapter X V Biaxial Nematic Liquid Crystals . . . . . . . . . . . . . . . . . . . . B. K . Sadashiva
933
1
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
933
2
Theoretical Prediction of the Biaxial Nematic Phase . . . . . . . . . . . . . 933
3
Structural Features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
934
4
Synthesis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
935
> 5.1 5.2
Characterization Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . Optical Studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . X-Ray Diffraction Studies . . . . . . . . . . . . . . . . . . . . . . . . . . .
937 937 938
5
Concluding Remarks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
941
7
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
942
Chapter XVI: Charge-Transfer Systems . . . . . . . . . . . . . . . . . . . . . . . K . Praefcke and D . Singer 1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
945
2
Calamitic Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
946
3
Noncalamitic Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
952
4
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
964
4.7.2 4.7.3 )
945
XXXVI
Contents
Chapter XVII: Hydrogen-Bonded Systems . . . . . . . . . . . . . . . . . . . . . Takashi Kato Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
969 969
2 2.1 2.1.1 2.1.2 2.1.3 2.1.4 2.2 2.2.1 2.2.2 2.2.3 3 3.1 3.2
Pyridine/Carboxylic Acid System . . . . . . . . . . . . . . . . . . . . . . . Self-Assembly of Low Molecular Weight Complexes . . . . . . . . . . . . . Structures and Thermal Properties . . . . . . . . . . . . . . . . . . . . . . . Phase Diagrams . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Stability of Hydrogen Bonds . . . . . . . . . . . . . . . . . . . . . . . . . . Electrooptic Effects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Self-Assembly of Polymeric Complexes . . . . . . . . . . . . . . . . . . . Side-Chain Polymers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Main-Chain Polymers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Networks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
969 969 969 972 973 973 973 973 974 974
Uracil/Diamino-pyridine System . . . . . . . . . . . . . . . . . . . . . . . Low Molecular Weight Complexes . . . . . . . . . . . . . . . . . . . . . . Polymeric Complexes . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
975 975 975
4
Miscellaneous Thermotropic H-Bonded Compounds by Intermolecular Interaction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
976
5
Lyotropic Hydrogen-Bonded Complexes . . . . . . . . . . . . . . . . . . .
977
6
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
978
981 Chapter XVIII: Chromonics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . John Lydon 981 1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.1 A Well-Defined Family Distinct from Conventional Amphiphiles . . . . . . 981 982 1.2 The Chromonic N and M Phases . . . . . . . . . . . . . . . . . . . . . . . 982 1.3 Drug and Dye Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.4 Molecular Structure of Chromonic Species . . . . . . . . . . . . . . . . . 983 2 2.1 2.2 2.3
The History of Chromonic Systems . . . . . . . . . . . . . The Early History . . . . . . . . . . . . . . . . . . . . . . . Disodium Cromoglycate and Later Studies . . . . . . . . . The 3-Way Link Between Drugs, Dyes, and Nucleic Acids
........ ........ ......... . . . . . . . . .
3 3.1 3.2
The Forces that Stabilize Chromonic Systems . . . . . . . . . . . . . . . . 986 Hydrophobic Interactions or Specific Stacking Forces? . . . . . . . . . . . 986 The Aggregation of Chromonic Molecules in Dilute Solution 987 and on Substrat Surfaces . . . . . . . . . . . . . . . . . . . . . . . . . . .
4
Phase Diagrams . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
988
5
Optical Textures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
989
983 983 984 985
Contents
6
X-Ray Diffraction Studies . . . . . . . . . . . . . . . . . . . . . . . . . .
7 7.1 7.2 7.3 7.4 7.5 7.6
The Extended Range of Chromonic Phase Structures The P Phase . . . . . . . . . . . . . . . . . . . . . . . Chromonic M Gels . . . . . . . . . . . . . . . . . . . Chiral Chromonic Phases . . . . . . . . . . . . . . . . More Ordered Chromonic Phases . . . . . . . . . . . Chromonic Layered Structures . . . . . . . . . . . . . Corkscrew and Hollow Column Structures . . . . . .
8
The Effect of Additives on Chromonic Systems: Miscibility and Intercalation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . .
XXXVII
. . . . . . . . . . . .......... .......... .......... .......... .......... ...........
991 993 993 998 998 998 999 999 1000
9
The Biological Roles of Chromonic Phases . . . . . . . . . . . . . . . . . 1001
10
Technological and Commercial Potential of Chromonic Systems . . . . . . 1005
11
Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
1005
12
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
1006
Index Volumes 2 A and 2 B . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
1009
General Introduction
Liquid crystals are now well established in basic research as well as in development for applications and commercial use. Because they represent a state intermediate between ordinary liquids and three-dimensional solids, the investigation of their physical properties is very complex and makes use of many different tools and techniques. Liquid crystals play an important role in materials science, they are model materials for the organic chemist in order to investigate the connection between chemical structure and physical properties, and they provide insight into certain phenomena of biological systems. Since their main application is in displays, some knowledge of the particulars of display technology is necessary for a complete understanding of the matter. In 1980 VCH published the Handbook of Liquid Crystals, written by H. Kelker and R. Hatz, with a contribution by C. Schumann, which had a total of about 900 pages. Even in 1980 it was no easy task for this small number of authors to put together the Handbook, which comprised so many specialities; the Handbook took about 12 years to complete. In the meantime the amount of information about liquid crystals has grown nearly exponentially. This is reflected in the number of known liquid-crystalline compounds: in 1974 about 5000 (D. Demus, H. Demus, H. Zaschke, Fliissige Kristalle in Tabellen) and in 1997 about 70000 (V. Vill, electronic data base LIQCRYST). According to a recent estimate by V. Vill, the cur-
rent number of publications is about 65000 papers and patents. This development shows that, for a single author or a small group of authors, it may be impossible to produce a representative review of all the topics that are relevant to liquid crystals on the one hand because of the necessarily high degree of specialization, and on the other because of the factor of time. Owing to the regrettable early decease of H. Kelker and the poor health of R. Hatz, neither of the former main authors was able to continue their work and to participate in a new edition of the Handbook. Therefore, it was decided to appoint five new editors to be responsible for the structure of the book and for the selection of specialized authors for the individual chapters. We are now happy to be able to present the result of the work of more than 80 experienced authors from the international scientific community. The idea behind the structure of the Handbook is to provide in Volume 1 a basic overview of the fundamentals of the science and applications of the entire field of liquid crystals. This volume should be suitable as an introduction to liquid crystals for the nonspecialist, as well as a source of current knowledge about the state-of-the-art for the specialist. It contains chapters about the historical development, theory, synthesis and chemical structure, physical properties, characterization methods, and applications of all kinds of liquid crystals. Two subse-
XL
General Introduction
quent volumes provide more specialized information. The two volumes on Low Molecular Weight Liquid Crystals are divided into parts dealing with calamitic liquid crystals (containing chapters about phase structures, nematics, cholesterics, and smectics), discotic liquid crystals, and non-conventional liquid crystals. The last volume is devoted to polymeric liquid crystals (with chapters about main-chain and side-group thermotropic liquid crystal polymers), amphiphilic liquid crystals, and natural polymers with liquid-crystalline properties. The various chapters of the Handbook have been written by single authors, sometimes with one or more coauthors. This provides the advantage that most of the chapters can be read alone, without necessarily having read the preceding chapters. On the other hand, despite great efforts on the part of the editors, the chapters are different in style, and some overlap of several chapters could not be avoided. This sometimes results in the discussion of the same topic from
quite different viewpoints by authors who use quite different methods in their research. The editors express their gratitude to the authors for their efforts to produce, in a relatively short time, overviews of the topics, limited in the number of pages, but representative in the selection of the material and up to date in the cited references. The editors are indebted to the editorial and production staff of WILEY-VCH for their constantly good and fruitful cooperation, beginning with the idea of producing a completely new edition of the Handbook of Liquid Crystals continuing with support for the editors in collecting the manuscripts of so many authors, and finally in transforming a large number of individual chapters into well-presented volumes. In particular we thank Dr. P. Gregory, Dr. U. Anton, and Dr. J. Ritterbusch of the Materials Science Editorial Department of WILEY-VCH for their advice and support in overcoming all difficulties arising in the partnership between the authors, the editors, and the publishers. The Editors
Part 2: Discotic Liquid Crystals
Handbook of Liquid Crystals Edited by D. Demus, J. Goodby, G. W. Gray, H.-W. Spiess, V. Vill Copyright 0 WILEY-VCH Verlag GmbH. 1998
Chapter VI Chiral Smectic Liquid Crystals
1 Synthesis of Chiral Smectic Liquid Crystals Stephen M. Kelly
1.1 Introduction Although it was known from seminal investigations early in this century [ 11 that smectic phases can be characterized by a one dimensional density wave, the exact layer structure was not known. Most smectic phases were characterized as smectic A, B or C (SmA, SmB and SmC) until the 1960s, depending on their optical characteristics under polarized light [2-4]. At this time it was known that in the SmA phase the timeaveraged director orientation is perpendicular to the layer plane. Therefore, uniformly aligned domains appear optically uniaxial. The structure of the SmB phase was unknown and this term was used to describe a range of orthogonal and tilted smectic phases [2-4], whose number and nature were determined much later [5, 61 The structure within the layers of the SmC phase, where the molecules are tilted at an azimuthal angle, 0, to the layer normal [7-91 giving rise to the observed optical biaxiality [8], had
just been determined [7-91. Liquid Crystals (LCs) incorporating an optically active center and exhibiting chiral smectic phases including the chiral smectic C phase (SmC*) had been synthesized along with nonoptically active LCs from the very beginning of LC research [lo-131. They were the result of investigations of the relationship between molecular structure and mesomorphic behavior and transition temperatures [ 10-1 31. The first SmC* compounds synthesized intentionally and recognized as such [ 141were prepared in order to study its structure. These were optically active versions [ 141 of laterally substituted Schiff's bases (see Structure 1) [7]. The lateral substituent had been introduced in order to reduce the high melting point (T,) of the original Schiff's bases [15,16], which probably also possess an SmC phase. Theoretical studies had indicated that the molecular orientation (i.e. the director) in an SmC* phase should spiral around an axis normal to the layer plane giving rise to a helical structure [ 171. If the pitch, p , the distance for a rotation of the di-
Cr-SmC* 29°C; SmC*-N* 94°C; N*-I 146.5"C [14] Structure 1
494
1 Synthesis of Chiral Smectic Liquid Crystals
Table 1. Comparison of the transition temperatures and spontaneous polarization measured at TSmC* - 10°C of (S)-4-decyloxybenzylidene-,(S)-a-methyl4-decyloxybenzylidene-, (S)-2-methylbutyl a-chloro 4-decyloxybenzylidene-, (S)-2-methylbutyl a-cyano 4-decyloxybenzylidene- and (S)-2-methylbutyl4-decyloxy-2-hydroxybenzylidene-4-amino-cinnamate[21, 351 and (S)-1-methylpropyl 4-decyloxybenzylidene-4-amino-cinnamate [37, 381.
c1
n 1 1 1
1 1 O
X H H H H OH H
Y H CH, C1 CN H H
""
I=\
v N Y& O (
CH 2)" ' C H ( C b
)C,H5
Cr-SmC* ("C)
H-SmC* ("C)
SmC*-SmA* ("C)
SmA*-I
P,
("C)
(nC crn-*)
76 45 51 92 88 82
(63)
92 68 (45) (70) 114 91
117 99 71 104 145 106
-
-
3
DOBAMBC
-
-
=3 10
HDOBAMBC DOBA- 1-MPC
Values in parentheses represent monotropic transition temperatures.
rector through 360°, is of the same magnitude as the wavelength of visible light, this can lead to the selective reflection of circularly polarized light in the visible part of the electromagnetic spectrum. 6 was found to be temperature dependent and normally increases with decreasing temperature [9]. The other tilted smectic (SmI* and SmF*) or crystal (J*, G*, H* and K*) phases or orthogonal smectic (hexatic SmB) or crystal (B and E) phases exhibit varying degrees of higher order within and between the layers [5, 61. The ordered chiral smectic phases can exhibit the same form chirality as the SmC* phase (see Sec. 2 and 3 of this chapter) [18]. The local C2 symmetry in tilted chiral smectic layers leads to a polar inequivalence within the layers due to the chiral and polar nature of the individual molecules despite a considerable degree of molecular reorientation about the long axis (see Sec. 2 and 3 of this chapter) [19-221. The consequence of this time-dependent alignment of the dipoles along the C2 axis is a spontaneous polarization (P,) parallel to the layer planes, which is averaged out macroscopically to
zero by the helical superstructure. However, a permanent polarization arises in an homogeneous monodomain, if the helix is unwound by an electric field or suppressed mechanically. Based on these symmetry arguments the ferroelectric properties of chiral smectic phases with a helical structure such as the SmC* phase were predicted and then found [21, 221 for the Schiff base 4-decyloxybenzylidene-4-amino-2-methylbutylcinnamate (DOBAMBC), see Table 1 . The first eight members of this series had already been prepared [12]. This was before the structure and optical characteristics of the SmC* phase had been established. Soon after the initial discovery of ferroelectricity in chiral smectic LCs it was predicted that, if the helix of an SmC* phase were suppressed by surface forces in very thin layers between two glass electrodes, then this would pin the molecules in their positions and allow switching between two energetically equivalent polarization directions, thereby giving rise to an electro-optic memory effect [22]. This is the basis of the electro-optic display device called the surface stabilized ferroelectric liquid crystal
1.2 Long Pitch Chiral Smectic Liquid Crystals or Dopants
device (SSFLCD) [23] utilizing ferroelectric LCs (see Sec. 2 and 3 of this chapter). In this device the helix is suppressed by surface forces, if the cell gap is considerably lower than the pitch of the SmC* helix. Therefore, long-pitch SmC* materials are used. Compounds exhibiting the SmC* phase are chosen due to the low rotational viscosity (li, of this phase compared to that of the more ordered tilted smectic phases. Chiral LCs with an SmA* phase are also receiving increased attention for potential applications utilizing the electroclinic effect, where the molecular tilt is coupled directly to the applied field [24].
1.2 Long Pitch Chiral Smectic Liquid Crystals or Dopants Although several attempts were made soon after the initial discovery of ferroelectricity in DOBAMBC to reduce T, and increase P, in order to facilitate the study of this new phenomenon at temperatures closer to room temperature, the vast majority of syntheses of chiral smectic LCs were undertaken later in order to provide stable SmC* materials with a defined spectrum of physical properties for SmC* mixtures designed for electro-optic display devices. Although the first eutectic mixtures created for prototype SSFLCDs consisted of purely optically active SmC* components [25], it was soon found that yand response time (z) for mixtures consisting of purely nonracemic SmC* components are generally too large for practical applications. It was also necessary to mix SmC* compounds with the same sign of P,, but opposing p , so that the P , values are additive and the p of the resultant mixture would be large. Both of these parame-
495
ters depend on the polar character of the substituents at the chiral center and the position of the chiral center in the terminal chain (see Sec. 2 and 3 of this chapter) [ 5 , 6, 181. Current commercially available SmC* mixtures designed for SSFLCDs consist of a nonoptically active base SmC mixture with a low value for a high TSmC* and low T , doped with at least one optically active (chiral) dopant [26-291. The chiral dopant should induce the desired value of P, and a long p in the base SmC mixture, in order to avoid the necessity of pitch compensation, without depressing TSmC* or increasing or the birefringence (An) of the mixture excessively. The dopant can also affect 6, which should be approximately 22" in order to produce a good optical contrast. All the mixture components must be chemically, thermally and photo- and electrochemically stable. The resultant induced P, of the mixture must also not be too large in order to avoid charge effects, but a large intrinsic value of P, for the chiral dopant allows smaller amounts to be used. Since many chiral dopants designed for LCD applications do not themselves possess a SmC* phase, parameters such as the P , or zoften have to be determined in a standard SmC mixture. This is often of more relevance to display device applications than the value for the pure dopant, if it does possess an SmC* phase, as the P, is matrix, concentration and temperature dependent. The P, in a mixture often reaches a maximum at a certain concentration and then remains constant or even decreases as more dopant is added [26-291. In contrast to the expectation that a higher P, necessarily means a shorter z an increase of P , which is due to an increase in 8,leads to longer z. If both z and P, increase (or decrease) upon changing the side chain within an homologous series of dopants, then this is probably a change of 8 of the mixture induced by the
496
1 Synthesis of Chiral Smectic Liquid Crystals
dopant. If zdecreases and P, increases, this is probably due to a change of Po (where P, = Po sine), which is a constant characteristic of the dopant, especially for low dopant concentrations, where changes of viscosity are small [30]. Both microscopic and steric models of the SmC phase postulate a zigzag shape for the terminal aliphatic chains of essentially symmetric compounds for SmC formation [20, 3 11. The microscopic model assumes additionally that antiparallel permanent dipoles at an angle to the molecular axis promote SmC behavior due to an induced torque 1201. It has been shown that while such dipole moments are not always absolutely essential for SmC formation, they are to be found in most SmC components and still mostly lead to higher SmC transition temperatures than those of the corresponding compounds without these dipoles [32, 331. This is not valid for compounds incorporating isolated, nonconjugated dipoles next to aliphatic rings [34]. There are also indications that alternately a cisltruns, linearly-extended conformation of the terminal chains is preferred in the SmC phase [34].
1.2.1 Schiff’s bases To facilitate the study of ferroelectricity in chiral smectics lateral substituents were introduced into the phenyl ring next to the cinnamate group of DOBAMBC and other members of the series. This was in order to reduce the degree of rotation at the chiral center and to increase the lateral dipole moment in the core of the molecule in an attempt to increase P, and at the same time lower the high T, of the original Schiff’s bases, see Table 1, [35]. Although the lateral substituents resulted in large decreases in T,, TSmC* and the clearing point ( T J , except for the nitriles, where T, increases, and the suppression of ordered smectic phases,
no significant effect on P, could be determined. This is most probably due to the relatively large d,istance between the substituents Y and the chiral center in the (9-2methylbutyl moiety. The introduction of a hydroxy group in a lateral position next to the central bond results not only in an improved stability , but the transition temperatures and the SmC* temperature range are also increased significantly, see Tables 1 and 2 [36-40]. This is caused by the formation of a pseudo-sixmembered-ring at the core of the molecule due to hydrogen bonding between the hydroxy group and the nitrogen atom of the central linkage. This gives rise to a higher degree of planarity of the aromatic rings, without leading to a broadening of the molecular rotation volume. However, the large dipole (2.58 D) associated with the hydroxy group in a lateral position does not lead to a significant increase in P,. This implies that a nonchiral dipole moment in the middle of the core of the molecule is of significantly less importance for the P , value than the dipole moment at the chiral center. Related Schiff’s bases incorporating commercially available optically active secondary alcohols such as (R)-2-butanol and (R)-2-pentanol instead of the primary alcohol (S)-2-methylbutanol exhibit a broad SmC* phase at high temperatures [37, 381, see Table 1 . A consequence of the closer proximity ( n = 0) of the methyl group to the 1core in 4-decyloxybenzylidene-4-aminomethylbutylcinnamate (DOBA- 1 -MBC) is a higher degree of coupling between the dipole moment at the chiral center and that of the core as well as a lower degree of free rotation. This results in a relatively high value for the P, and slightly lower transition temperatures than those of DOBAMBC ( n = 1) [21]. To further increase the chemical and photochemical stability and reduce T,, more to-
1.2 Long Pitch Chiral Smectic Liquid Crystals or Dopants
497
Table 2. Comparison of the transition temperatures, tilt angle and spontaneous polarization measured at 25 "C for (S)-4-(2-methylbutyl)-2-hydroxyresorcylidene4-octylaniline, (S)-4-(4-methylhexyl)-2-hydroxyresorcylidene 4-octylaniline and (S)-4-(6-methyloctyl)-2-hydroxyresorcylidene4-octylaniline 139,401.
~~
n 1 3 5
~
Cr-SmC* ("C)
Sm2-SmC*
36 -
=20
20
-
("C)
SmC*-SmA* ("C)
SmA*-I ("C)
8
50 17 86
57 -
8 =40
-
=5
wards room temperature to yield SmC* substances more suitable for use in SSFLCDs, new Schiff's bases without the cinnamate group were synthesized [39,40]. The effect of chain length and proximity of the point of branching to the molecular core, which are general for SmC* LCs, is shown for these Schiff's bases in Table 2. The TSmC* is highest for the compounds with the longest chain and the chiral center furthest from the molecular core ( n = 3). The steric effect of the methyl group at the point of branching decreases for positions further from the core due to the greater number of chain conformations possible. The P, also decreases due to the smaller energy difference between the various positions of the dipole moment and the lower degree of coupling with the dipoles of the core. This is shown more clearly when differences in 8 are taken into consideration. The low value for T , and wide SmC* temperature range of MHxRA 8 allowed studies of the ferroelectric properties of these LCs to be carried out at room temperature.
1.2.2 Aromatic Esters with Alkyl Branched Alkyl Chains To prepare stable SmC* LCs with comparable P , values to those determined for
PS
(degrees) (nC cm-') 3 2 =O. 1
P, I e (nC cm-' rad-I) 22.0 2.9 0.2
MBRA8 MHxRA8 MORA8
Schiff's bases, but with greater resistance towards hydrolysis, higher resistivity and better alignment properties a wide variety of aromatic esters with alkyl-branched and, above all, methyl-branched terminal chains were synthesized, see Table 3 [41-511. Several ester series with a nonracemic chiral center had already been prepared for thermochromic applications [44]. Several homologues also exhibited SmC* as well as N* phases at high temperatures. The optically active part of the molecules was usually derived form the same alcohols such as (S)-2-methylbutanol, (R)-2-butanol and (R)-2-pentanol as for the Schiff's bases, see Scheme 1 . Similar dependencies of the transition temperatures, P, and yon the nature of the methyl-branched chain were found. Additionally it was found that the value of P, increased with the number of carbon atoms in the chain after the point of branching, probably due to the dampening effect of the longer chain [6, 41-44]. Large substituents like an ethyl group at the point of branching reduce the transition temperatures substantially more than a methyl group. The absolute transition temperatures and the presence of other smectic phases or the nematic phase depends on the direction of the ester linkage in the molecular core, the presence and nature of polar groups between the core and the terminal chain and
498
1
Synthesis of Chiral Smectic Liquid Crystals
Table 3. General structures for typical smectic C* liquid crystals with nonracemic, methyl-branched terminal chains derived from primary and secondary alcohols. Structure
the number of rings in the core [6,41]. Aliphatic rings in the core such as l, 4-disubstituted cyclohexane [45], bicyclo(2, 2, 2)octane [46], dioxane rings [47] or dioxaborinane rings [48] lead to the elimination of the SmC* phase in two-ring compounds and to lower TSmC* in three-ring compounds compared to the corresponding fully aromatic analogs. However, yand/or An are lower [45]. This results in lower Tor allows thicker cells to be used. The introduction of a cis carbon-carbon double bond in the middle or in a terminal position of the alkoxy chains of two-ring and three-ring esters leads to lower transition temperatures in general and sometimes to broader SmC* phases close to room temperature [49].
Ref.
The additive effects of dipole moments near the chiral center and at the chiral center itself are demonstrated by a reference to Tables 4 and 5. The introduction of a large dipole moment in the form of a carbonyl function between the aromatic core and the chiral center to yield a number of alkanoylsubstituted (keto-) esters led to higher P, values than those of the corresponding ethers and esters, see Table 4 [50]. The presence of a second optically active group in the second terminal chain can also lead to additive effects and a still higher P,, although y can be expected to be considerable [50].A strong dependence of the observed P , values on the number of carbon atoms in both terminal chains and the presence and
1.2 Long Pitch Chiral Smectic Liquid Crystals or Dopants
499
H
Br
a) Mg/Et20 b) BrC4H$r
I
* H
HO
2
I
Br
NaOEt/EtOH/CH2(COZEt)z
o* OyO
f
HO
Ix
LiAIH4/Et20 H+/H20
H
O
A
HO
HO
Ai
HBr/H2S04
b)H+/H20 1 H
H
Br
O
A
1
HBr/H2SO4
position of various lateral substituents was also observed. Dipole moments incorporated in the form of lateral substituents in the aromatic core next to the terminal chain, but
Scheme 1. Typical reaction pathways to important reaction intermediates derived from naturally occurring (S)-2-methylbutanol and used to synthesize chiral smectic liquid crystals or dopants with nonracemic, methylranched terminal chains.
further away from the chiral center can also give rise to a higher p,, see Table 5 [28, 5 11. However, the effect is smaller for similar dipole moments and the increase is al-
500
1 Synthesis of Chiral Smectic Liquid Crystals
Table 4. Comparison of the spontaneous polarization measured at TSmC*-I 0 “C of (S)-4-([2-octyl]oxy)phenyl 4-octyloxy- l,l’-biphenyl-4’-yl-carboxylate(X = 0),(S)-2-octyl4-(4-octyloxy- l,l’-biphenyl-4’-yl-carboxy)benzoate (X = COO) and (S)-4-(2-methyloctanoyl)phenyl 4-octyloxy-l,l’-biphenyl-4’-yl-carboxylate (X = CO) [sol.
X -0-
0.68 1.23 1.99
-coo-co-
38 70 112
130 195 313
Table 5. Comparison of the transition temperatures, tilt angle and spontaneous polarization measured at TSmC*-10 “C of 4-octyloxy- l,l’-biphenyl-4’-yl 3-substituted-4-([(S)-2-octyl]oxy)benzoates [5 11. X
X
(D)
Cr-SmC* (“C)
SmC*-SmA*/N* SmA*-I (“C) (“C)
N*-I (“C)
0 (degrees)
PS
(nC cm-2)
Po (nC cm-2)
0 -1.47 -1.59 -1.57 -4.05
78 65 13 43 68
99 97 95 87 90
122 100 97 89
45 36 40 41 25
45 78 123 131 160
64 134 191 202 386
P
~
H F
c1 Br CN
-
105
most certainly also due to steric effects as the energy difference between the most stable, time-dependent conformations of the chiral center is increased. This is seen as p decreases and yincreases with the size of the lateral substituent [28,5 11. However, the small p requires pitch compensation for practical application in SSFLCDs. This contributes to the higher y and zvalues caused by the presence of the lateral substituents.
1.2.3 Aromatic Heterocycles with Alkyl-Branched Alkyl Chains A further development was the introduction of an optically active center into the terminal chains of two- and three-ring SmC com-
-
pounds with a phenyl-pyrimidine, pyrazine or pyridazine group in the molecular core to produce the corresponding methylbranched SmC* materials [52-551. Some of these new compounds exhibited a broad SmC* phase, sometimes close to room temperature (Structure 2). They were characterized by low yvalues. The corresponding alkenyloxy-substituted phenylpyrimidines with a cis-carbon-carbon double bond or a carbon-carbon double bond in a terminal position often possessed an SmC* phase closer to room temperature [56].
Cr-SmC*: 23°C; SmC*-N*: 34°C; N*-I: 47°C [ S 2 ]
Structure 2
1.2 Long Pitch Chiral Smectic Liquid Crystals or Dopants
NCfOH
NaCN
a) LiA1H4/Eb0
b) Ht/H20
\
TosCl/N(C,H,),
/
a)KOH/EtOH b)H+
/
A
l
2
\
\ H+/H20
a) LiA1H4/Et20
C02H CnHZn+lo-r(
1.2.4 Esters and Ethers in the Terminal Chain Large P , values can be achieved by increasing the polarization of the branching methyl group at the chiral center by the introduction of the electron withdrawing oxygen atom. Simple esters and ethers were usually prepared from naturally occurring alkyl L-(+)-lactates (alkyl (S)-2-hydroxypropa-
Cr-SmC*: 88°C; SmC*-I: 148°C [58]
Structure 3
50 1
Scheme 2. Typical reaction pathways to important reaction intermediates derived from naturally occurring alkyl L-(+)-lactates (alkyl (S)-2-hydroxypropanoates) and used to synthesize chiral smectic liquid crystals or dopants with nonracemic terminal chains with polar substituents.
noates), see Scheme 2 [57-601 or readily accessible, synthetic alkyl (R)-3-hydroxybutanoates [61]. More sophisticated syntheses of two- and three-ring phenylpyrimidines with an alkoxy group and a methyl group at the chiral center in the terminal chain (Structure 3) involved resolution of a prochiral ester into the corresponding optically active alcohol and ester by an enzyme catalyst [62]. Lactate ethers [60] exhibit higher P ,
Cr-SmC*: 51°C; Sx-SmC*: 46°C; SmC*-N: 69°C; N*-I: 91.5"C [60]
502
1
Synthesis of Chiral Smectic Liquid Crystals
Table 6. Comparison of the transition temperatures and spontaneous polarization of 4'-heptyloxy- 1, 1'-biphenyl-4-yl (S)-2-chloro-3-methylbutanoate, (2S,3R)-2-chloro-3-methylpentanoateand (S)-2-chloro-4-methylpentanoate r64, 651.
R
Cr-SmC*/SmA* ("C)
S3*-SmC*/SmA* ("C)
SmC*-SmA* ("C)
SmA*-I* ("C)
PS ( n c cm-2)
16
-
(71)
80
-98
48
(36)
56
66
-210
71
-
(65)
74
+S8
Values in parentheses represent monotropic transition temperatures.
a) NaN02/H2S04 a) LiAlH,
a) LiAlH,
b)
b) H+/H20
J H
NaN02/H2S04
O
T
c2H50H!Y
(CF3SO2)20
cF3s020~2"5
/U*N-F
a) LiAlH,
CHzOH
Scheme 3. Typical reaction pathways to important reaction intermediates derived from naturally occurring amino acids such as (2R,3s) or (2S, 3R)-2-amino-3-methylpentanoic acid (D- or L-alloleucine) and used to synthesize chiral smectic liquid crystals or dopants with nonracemic terminal chains with polar substitu-
503
1.2 Long Pitch Chiral Smectic Liquid Crystals or Dopants
Table 7. Comparison of the spontaneous polarization (extrapolated to loo%),response times (10 V,,/pm square wave, time to maximum current), pitch and tilt angle measured at 25°C for mixtures of 7 wt% 4-octyloxy-1,l’biphenyl-4’-yl (R)-2-fluorohexanoate and 7 wt% trans- 4-(4-octyloxyphenyl)cyclohexyl (R)-2-fluorohexanoate and a host mixture [69].
e
X
(degrees)
-Q-
56
a..,, 35
950
0.75
24.5
700
1.52
22
and higher y values than comparable lactate ester mixtures with similar 8 and TSmC* values. This results in similar z for similar mixtures containing one or the other chiral dopant. The high TSmC* temperatures for the pure lactate esters [58] are not reflected in high TSmC* values for mixtures containing them. This is typical of the nonlinear and nonideal behavior of chiral dopants in SmC* mixtures.
1.2.5 Halogens at the Chiral Center If the dipole at the chiral center is increased by replacing the methyl branching group by a more polar moiety such as a halogen atom, then Po increases due to the larger dipole moment perpendicular to the long axis of the molecule, see Table 6. Most ferroelectric LCs incorporating halogen atoms at the optically active center are esters or ethers derived either from naturally occurring lactic acid or amino acids, see Schemes 2 and 3 [63-671. There is still a considerable degree of freedom of rotation at the chiral center and the observed value for P, is still considerably lower than that
expected if the dipole moment at the chiral center is taken into account. For a large value of the P, it is important that the dipoles in the vicinity of the optically active center point in the same direction, see Table 6. The dipoles due to the two chiral centers in the (2S,3R)-2-chloro-3-methyl-pentanoyloxysubstituted esters are additive and lead to the highest P, values [63-671. The pure esters exhibit high negative P , values and enantiotropic SmC* phases. The values extrapolated to 100% from SmC* mixtures are often considerably lower. The small van der Waals radius (1.47 A) of fluorine and relatively large dipole moment (1.41 D) of the C-F bond lead to SmC* dopants with a relatively high P,, but with a lower y and higher TSmC* than the corresponding chloro- or bromo-substituted compounds [68, 691.
1.2.6 Cyclohexyl a-Fluorohexanoates Substitution of the aliphatic trans- 1, 4-disubstituted cyclohexane ring for the aromatic benzene ring leads to a lower electron density at the chiral center and to a higher
504
1 Synthesis of Chiral Smectic Liquid Crystals
degree of conformational mobility of the dipole moment, but P, is only reduced by about 40%, see Table 7 [30, 691. However, the cyclohexane ring gives rise to lower values for An, y and a substantially longer pitch, as expected. The data collated in Table 8 allow a valid comparison of the relative effects of either a fluorine or a chlorine atom attached directly to the optically active center of the chiral dopant [69]. The transition temperatures and P, of the mixture containing the Table 8. Comparison of the transition temperatures, spontaneous polarization and response times (10 V,,/pm square wave, time to maximum current) measured at 25°C for mixtures of 7 wt% of the truns4-(2',3'-difluoro-4'-decyloxy- 1,l '-biphenyl-4-yl)cyclohexyl (R)-2-fluorohexanoate and truns -4-(2', 3'Difluoro-4'-decyloxy- 1, l'-biphenyI-4-yl)cyclohexyl (R)-2-chlorohexanoate and a host mixture (SmC-SmA: 76°C; SmA-N: 81 "C; N-I: 103 "C) ~301. F
F CI
6.2 4.5
F
415 75.5 630 74.5
88.3 86.3
104.3 102.3
a-fluoroester are all higher than those observed for the corresponding mixture containing an equal amount of the otherwise identical a-chloroester. This could be due to the stronger electronegativity of the fluorine atom. Additionally, y of the a-fluoroester must also be lower than that of the a-chloroester as shown by the significantly lower 2, which is only partially due to the higher value of Po. This is a result of the smaller size of the fluorine atom and the shorter carbon-fluorine bond compared to that of chlorine. Despite the presence of the cyclohexane ring, which promotes SmB behavior, several optically active a-fluoroesters incorporating the aliphatic cyclohexane ring and a number of different cores possess an enantiotropic SmC* phase at elevated temperatures [69]. However, the tendency to form the SmC* phase is much reduced and a number of dipole moments in various parts of the molecular core are required, see Table 9. This is consistent with the microscopic, dipole-dipole theory of the SmC phase [20]. Only the combination of the phenylpyrimidine core and an alkoxy chain in the ester gives rise to the SmC* phase. Tworing compounds incorporating a cyclohexane ring do not generally exhibit an SmC phase.
Table 9. Comparison of the transition temperatures for the truns-4-(4'-decyI-l, l'-biphenyl-4-yl)cyclohexyl ( R ) 2-fluorohexanoate, truns-4-(4-[5-decylpyrimidin-2-yl]phenyl)cyclohexyl(R)-2-fluorohexanoate and rrans-4[4-(5-nonyloxypyrimidin-2-yl)phenyl]cyclohexyl(R)-2-fluorohexanoate (SmC*-SmA: 76 "C; SmA-N: 8 1 "C; N-I: 103 "C) [30].
R
X
C-SmB*/SmC*/I ("C)
SmB*-SmC*/SmA* ("C)
SmC*-SmA* ("C)
I16 90
108
SmA*-I ("C) I45
I65
1.2 Long Pitch Chiral Smectic Liquid Crystals or Dopants
505
Table 10. Comparison of the spontaneous polarization, response times (10 V,,/pm square wave, time to maximum current) and pitch measured at 25 "C for mixtures of 7 wt% of the ether rrans-4-[4-decyloxyphenyl]cyclohexyl (R)-2-fluorohexanoate and the ester truns-4-[4-decanoyloxyphenyl]cyclohexyl(R)-2-fluorohexanoate and a host mixture (SmC-SmA: 76°C; SmA-N: 81°C; N-I: 103°C) [30].
CHZ
co
3.8 5.7
460 400
64.5 68.0
82.6 81.5
97.8 97.1
Table 11. Comparison of the transition temperatures, spontaneous polarization and response times ( 1 0 V,,/pni square wave, time to maximum current) measured at 25 "C for mixtures of 7 wt% of the diester tuuns-4-[4-(2,3difluoro-4-decyloxybenzoyloxy)phenyl]cyclohexyl (R)-2-fluorohexanoate and 7 wt% of the ester truns-4-(2',3'difluoro-4'-decyloxy- l,l'-biphenyl-4-yl)cyclohexyl (R)-2-fluorohexanoate and a host mixture (SmC-SmA: 76°C; SmA-N: 81 "C; N-I: 103°C) [30]. F
coo -
5.4 6.2
F
650 415
The data collated in Table 10 allow the effect of an ester group in a terminal position in the core of a chiral dopant to be determined. The ether [30] and the diester [69] differ only in the presence of a carbonyl group (CO) instead of a methylene unit (CH,), that is, the chain lengths are the same. TSmC* for the mixture containing the ester (X = CO) is higher than that incorporating an equal amount of the ether (X = CH,). The P , is higher for the ester than for the ether, indicating a greater value for Po as shown by the shorter z. The effect of an ester group in a central position of a chiral dopant can be elucidat-
76.7 75.5
83.5 88.3
103.9 104.3
ed from the data in Table 11 [69]. TSmC* for the mixture containing the a-fluoro diester with a second ester group in a central position is higher than that incorporating an equal amount of the a-fluoro (mono-) ester. This is not surprising as the a-fluoro di-ester exhibits an enantiotropic SmC* phase at elevated temperatures in the pure state. The P, is higher for the ester than for the diester. This is probably due to a change of Po as well as of because of presence of the second carboxy (ester) group. Even if Po were constant and Owere fully responsible for the increase of P,, yof the mono ester would still be substantially lower.
506
1 Synthesis of Chiral Smectic Liquid Crystals
The P, values for related fluoro-alkyl- or alkoxy-substituted compounds are lower due to the lower dipole moment at the chiral center [70-751. The presence of the trifluoromethyl group at the chiral center in the side-chain of three-ring compounds leads to high P , values and an SmC* phase for the pure substances [76]. Although a trifluoromethyl group at the chiral center in the central linkage between two aromatic rings in the core of a similar three-ring phenylpyrimidine leads to very low transition temperatures and the absence of a SmC* phase for the pure substance and to a low TSmC* for mixtures containing these chiral dopants the values of P,, z and 8 are otherwise attractive [77].
1.2.7 Cyano Groups at the Chiral Center The large dipole moment and steric bulk of the cyano group at the chiral center gives rise to a larger energy difference between the most probable time-averaged conformations of the chiral center and therefore to a high P , value, although low transition temperatures are also observed [78-811. This is a consequence of the large steric effect of
CN
Cr-SmC*: 102°C; SmC*-SmA: 145°C;SmA-I: 155°C [79]
Structure 4
the cyano group close to the molecular core (Structure 4) leading to low transition temperatures for the pure compounds. Therefore, enantiotropic LC phases are only observed for three-ring compounds.
1.2.8 Optically Active Oxiranes and Thiiranes Another way to maximize P , is to limit the degree of free rotation at the chiral center and increase the energy difference between the two major conformations of the dipole at the chiral center. This can be achieved by incorporating the chiral center in a cyclic system such as in an oxirane 127, 82, 831. The molecular dipole associated with an epoxide moiety is oriented normal to the tilt plane. The rigidity of the epoxide ring minimizes the conformational averaging out of the dipole moment and should lead to a high value for the P,. Oxiranes are prepared by Sharpless epoxidation of ally1 alcohols,
Table 12. Transition temperatures and spontaneous polarization of mixtures of 7.37 wt% of 4-[(R)-l-fluor0-(2s, 3S)-epoxynonyl]pheny14-decyloxybenzoate(X = H;Y = F) or 7.58 wt% of 4-[(S)- I-fluoro-(2S, 3s)-epoxynonyllphenyl 4-decyloxybenzoate (X = F; Y = H) and a host compound (Cr-SmC: 35 "C; SmC-SmA: 70.5 "C; SmA-I: 72 "C) [83].
X
H F
Y
F H
Cr-SmC* ("C)
S mC *-SmA * ("C)
SmA*-I ("C)
ps (nC cm-2)
53 25
66 61
70 66
+2 -13
507
1.2 Long Pitch Chiral Smectic Liquid Crystals or Dopants
Table 13. Helical twisting power (HTP) (measured just above SmA*-N*), rotational viscosity, response time and spontaneous polarization measured at 25 "C for mixtures of 10 mol% of 5-([cis-(2S, 3S)-epoxy]octanoyloxy)phenyl-2-octyloxyprimidineor 5-([truns(2R, 3S)-epoxy]octanoyloxy)phenyl-2-octyloxyprimidine and a host mixture (C-SmC: 10 "C;SmC-SmA: 84.5 "C; SmA-N: 93.5 "C; N-I: 105 "C) [27].
Configuration
HTP pm-'
ps
z
y (mPa)
ps
(2S, 3s) cis (2R, 3s) trans
6.2 0.8
38 422
240 36
77 2.3
(nC cm-2)
which can be easily attached to a phenol, or oxidized to the corresponding acid and attached to the phenol in a normal esterification reaction [27, 821. If the dipoles associated with the fluorine atoms and the epoxy moiety in the two diastereomeric oxiranes shown in Table 12 [83] point in the same direction, then a high P, value is observed. If they are opposed, then they cancel each other out to some extent, and a low P, value is the result. Converting the ether linkage to an ester group
does not lead to a significant change in the observed P, value, see Table 13 [27]. However, changing the configuration at the second carbon atom of the epoxy group (i.e. changing from trans to cis) results in an enhanced stereo coupling between the chiral moiety and the lateral molecular dipole. Thus, significantly higher P, values are observed for the cis-substituted oxiranes [27]. The cis configuration at the epoxy moiety leads automatically to a nonlinear (i.e. fully extended) conformation of the chain. This leads to a substantially higher value for y However, y i s increased to a lesser extent than the P, is increased and this results in significantly shorter z for the cis-substituted esters. Replacement of the oxygen atom in the epoxy moiety by a sulfur atom to produce the analogous thiiranes [84] results not only in lower transition temperatures, but also significantly lower values for P,, if the optically active unit is situated on the periphery of the molecule. The opposite is observed with respect to P,, if the chiral entity is situated between two mesogenic units such as six-membered rings. A Walden inversion during the reaction produces an inversion of configuration at both asymmet-
Table 14. Transition temperatures, response time (10 Vpp/ymsquare wave, time to maximum current), tilt angle and spontaneous polarization measured at 25°C in a 2 pm cell with polyimide alignment containing mixtures of 2 mol% of the (2 S , 4R)-truns- (a-[4'-pentyl-l,l'-biphenyl-4-yl]-y-pentyl)lactone or (2R,4R)-cis- (a-[4'-pentyl- l,l'-biphenyl-4-yl]-y-pentyl)lactoneand a host mixture (Cr-SmC: O and those with PO
eb>O
Figure 32. Geometrical relation between the local polarization density and the director deformation in the case of positive values for the flexoelectric coefficients. For negative values the direction of the induced dipole should be reversed. The size and sign of e , and eb (alternatively called e , and e,) are a molecular property.
x
Figure 33. The symmetry of twist. If we locally twist the adjacent directors in a nematic on either side of a reference point, there is always a twofold symmetry axis along the director of the reference point. Therefore a twist deformation cannot lead to the separation of charges. Thus a nematic has only two nonzero flexoelectric coefficients.
vector (see our earlier discussion of additional symmetries in Secs. 2.3.6 and 2.3.7). Hence a twist deformation cannot lead to a separation of charges. This property singles out the twist from the other deformations, not only in a topological sense, and explains why twisted states are so common. The well-known fact that a nematic which is noninvariant under inversion is unstable against twist and normally adopts a helical (cholesteric) structure, would not have been the same if the twist had been connected
558
2
Ferroelectric Liquid Crystals
with a nonzero polarization density in the medium. This also means that a “twisted nematic” as used in displays is not polarized by the twist.
2.4.3 The Molecular Picture Recognition of the flexoelectric effect is due to Meyer [18], who in 1969 derived an expression equivalent to Eq. (97). He also had a helpful molecular picture of the effect, as illustrated in Fig. 34a. To the left in this picture, the unstrained nematic structure is shown with a horizontal director in the case of wedge-shaped and crescent-shaped molecules. To the right the same molecules are shown adjusted in their distribution corresponding to a splay and a bend distortion, respectively. In either case the distortion is
Bend
Splay
Bend
Splay
coupled to the appearance of a nonzero polarization density. As we can see, this is a steric (packing) effect due to the asymmetry of the molecular shape. The inverse effect is shown in Fig. 34b. When an electric field is applied it induces a distortion. In this figure we have illustrated a hypothetical compound splay-bend deformation and also used a slightly more general shape for the model molecules. It is clear that the observable effects will depend on the molecular shape as well as the size and distribution of the dipoles in the molecules. The inverse flexoelectric effect simply offers another mechanism for polarizing the medium, since the distortion will imply a polarization parallel to E . It is therefore a special kind of dielectric mode involving orientational polarization. In this respect, the inverse effect bears a certain resemblance to the fieldinduced lining up of dipoles in a paraelectric material, as the dipoles are already present in the molecules in both cases (see Fig. 35). Flexoelectric polarization leads to a separate contribution to the general term G = - E . P in the free energy. With Eq. (97) this contribution takes the form (98) G f = -e,E . n ( V . n )- e,E . [n x ( V x n)]
Bend
Figure 34.(a) The flexoelectric effect. A polarization is coupled to a distortion (from [18]). (b) The inverse flexoelectric effect. An applied field induces a distortion (from [I 121). With our sign conventions, (a) corresponds to e,>O, ebo
c3
Optic axis
E=a
E=o
Figure 38. Field-induced tilt (left) and corresponding splay- bend distortion when looking at the Bouligand plane along the field direction (right).The same director pattern will be found in any cut made perpendicular to the tilted optic axis, whereas the director is hornogeneous in any cut perpendicular to the nontilted axis.
Irnl
It can be shown 1117, 1181 that the induced tilt is linear in E according to
4 ) -eE Kk where 1 e = -(es +q,) 2
and 1
K = - (4 1 + K33) 2 and k is the cholesteric wave vector. The response time is given by
r=- Y
Kk2
(''I)
where y is the characteristic viscosity. We note, in advance of discussing the same feature in the electroclinic effect, cf. Sec. 2.5.8, that the response time does not depend on the value of the applied electric field. The temperature independence is seen immediately from Eq. (log), because both e and K should be proportional to S2, the square of the scalar nematic order parameter. Therefore a mixture with temperature-independent pitch (not too difficult to blend) will have a @ ( E )that is independent of temperature. Since the original work of Patel and Meyer [ 1171, the effect has been further investigated [119- 1221, in the last period [ 112, 118, 123- 1261 with a special empha-
2.4 The Flexoelectric Polarization TI 827
563
ferroelectrics. They all belong to the category of "in-plane switching" with a continuous grayscale and still have a great potential to be exploited in large scale applications.
@(E) T=2S°C @(E) T=44"C
2.4.7 Why Can a Cholesteric Phase not be Biaxial?
0 0 0 0
E = 5OV/pm 10
25
30
35 40 Temperature ("C)
1 45
Figure 39. The flexoelectrooptic effect measured on the Merck mixture TI 827. (a) Induced tilt as a function of E. (b) Induced tilt as a function of temperature.
sis on ruling out the normal dielectric coupling and increasing the linear range, which is now larger than for any other electrooptic effect in liquid crystals. With the available cholesteric materials, the applied field is quite high but can be estimated to get lowered by at least a factor of ten if dedicated synthetic efforts are made for molecules with convenient shapes and dipoles. The flexoelectrooptic effect belongs to a category of effects that are linear in the electric field, and which all have rather similar characteristics. These include the electroclinic effect, the deformed helical mode in the C* phase, and the linear effect in anti-
The nonchiral nematic is optically a positive uniaxial medium. A cholesteric is a nematic with twist. The local structure of a cholesteric is believed to be the same as that of the nematic except that it lacks reflection symmetry. This means that the director and therefore the local extraordinary optic axis is rotating around the helix axis making the cholesteric a negative uniaxial medium with the optic axis coinciding with the twist axis. The question has been asked as to why the nematic with twist could not be biaxial, and attempts have been made to measure a slight biaxiality of the cholesteric phase. In other words, why could the twist not be realized in such a way that the long molecular axis is inclined to the twist axis? Why does it have to be perpendicular? Section 2.4.6 sheds some light on this question. As we have seen, as soon as we make the phase biaxial by tilting the director the same angle out of the plane everywhere, we polarize the medium. Every fluctuation 6n out of the plane perpendicular to the helix axis is thus coupled to a fluctuation 6P raising the free energy, -(6P)2. The energy of the state is thus minimized for 4=0, which is the ground state. In other words, the N* state cannot be biaxial for energetic reasons, just as the N state cannot be polar for the same reason (see the discussion in Sec. 2.2.2). In other words, the cholesteric phase is uniaxial because this is the only nonpolar state.
564
2
Ferroelectric Liquid Crystals
2.4.8 Flexoelectric Effects in Smectic A Phases In smectics such deformations which do not violate the condition of constant layer thickness readily occur. In a smectic A (or A*) this means that V x n = 0 and the only deformation is thus pure splay, in the simplest approximation. We will restrict ourselves to this case, as illustrated in Fig. 40. Special forms are spherical or cylindrical domains, for instance, the cylindrical structure to the right in the figure, for which a biologically important example is the myelin sheath which is a kind of “coaxial cable” around a nerve fiber. Let us consider such a cylindria1 domain. The director field in the xy plane is conveniently described in polar coordinates p and 0 a s
n = (np, no) = (120)
(112)
from which we get the divergence
(1 13)
This gives us the flexoelectric polarization density
P = e, n ( V .n) = e, n
(1 14)
b) \
Figure 40. The easy director deformation in a smectic A is pure splay which preserves the layer spacing everywhere.
The polarization density thus falls off as Ilp. In the center of the domain we have a disclination of strength one. Because the disclination has a finite core we do not have to worry about P growing infinite. It is interesting to take the divergence of this polarization field. We get
v. P = l ~
a (pP,) = 0
-
P aP
Thus the polarization field is divergence free in this geometry, which makes it particularly simple, because a nonzero V - P would produce space charges and long range coulomb interactions. Hence, a myelin sheath only has charges on its surface from the flexoelectric effect. It is further a well-known experience when working with both smectic A and smectic C samples, under the condition that the director prefers to be parallel to the boundaries, and that cylindrical domains are quite abundant. In contrast, spherical domains do generate space charge because, as is easily checked, v . P is nonzero, falling off as - ~ r * .
2.4.9 Flexoelectric Effects in Smectic C Phases Flexoelectric phenomena in smectic C phases are considerably more complicated than in smectic A phases, even in the case of preserved layer thickness. Thus under the assumption of incompressible layers, there are not less than nine independent flexoelectric contributions. This was first shown in [ 1271. If we add such deformations which do not preserve the layer spacing, there are a total of 14 flexocoefficients [128]. If we add chirality, as in smectic c*,we thus have, in principle, 15 different sources of polarization to take into account. In a smectic C , contrary to a smectic A, certain bends and twists are permitted in the
2.4 The Flexoelectric Polarization
director field without violating the condition of constant smectic layer spacing. Such deformations we will call “soft” in contrast to “hard” distortions which provoke a local change in layer thickness and therefore require much higher energy. As in the case of the N* phase, soft distortions may be spontaneous in the C* phase. Thus the chiral smectic helix contains both a twist and a bend (see Fig. 41) and must therefore be equal to a flexoelectric polarization, in contrast to the cholesteric helix which only contains a twist. This may even raise the question of whether the spontaneous polarization in a helical smectic C* is of flexoelectric origin. It is not. The difference is fundamental and could be illustrated by the following example: if starting with a homogeneously aligned smectic C you twist the layers mechanically (this could be done as if preparing a twisted nematic) so as to produce a helical director structure identical to that in a smectic C*, you can never produce a spontaneous polarization. What you do achieve is a flexoelectric polarization of a
Figure 41. Model of the director configuration in a helical smectic C*. To the left is shown a single layer. When such layers are successively added to each other, with the tilt direction shifted by the same amount every time, we obtain a space-filling twist - bend structure with a bend direction rotating continuously from layer to layer. This bend is coupled, by the flexoelectric effect, to an equally rotating dipole.
565
certain strength and sign, which is always perpendicular to the local tilt plane, just as the spontaneous C* polarization would be. However, the flexoelectric polarization is strictly fixed to the deformation itself and cannot, unlike the spontaneous one, be switched around by an external electric field. On the other hand, it is clear from the example that the two effects interfere: the flexoelectric contribution, which is independent of whether the medium is chiral or nonchiral, will partly cancel or reinforce the spontaneous polarization in a smectic C* in all situations where the director configuration is nonhomogeneous in space. This is then true in particular for all chiral smectic samples where the helix is not unwound, and this might influence the electrooptic switching behavior. The simplest way to get further insight into the flexoelectric polarization just described is to use the “nematic description” [46] of the C* phase. We then look, for a moment, at the smectic as if it were essentially a nematic. This means that we forget about the layers, after having noticed that the layers permit us to define a second vector (in addition to n), which we cannot do in a nematic. The second vector represents the layers and is the layer normal, z (or k ) , which differs from then direction by the tilt angle. The implicit understanding that z is constant in space represents the condition of undeformed smectic layers. We use the Oseen-Frank elastic energy expression [Eq. (96)] for a nematic medium as a starting point. Now, according to our assumption, the medium is chiral, and an ever so slight chiral addition to a nematic by symmetry transforms the twist term according to [ l l l ]
.)I2 + [n . ( V x n )+ qI2
[n . ( V X
(1 16)
The ground state now corresponds to a twisted structure with a nonzero value of
566
2
Ferroelectric Liquid Crystals
n . ( V x n ) given by a wave vector q, the sign of which indicates the handedness. Equation (1 16) is written for a right-handed twist for which n . ( V Xn ) =-q. Note that the reflection symmetry is lost but the invariance condition [Eq. (6)] is still obeyed. Chirality thus here introduces a new scalar quantity, a length characteristic of the medium. If the medium is also conjectured to be polar, it might be asked if it is possible, in a similar way, to introduce a true vector (n and z are not true vectors, since there is one symmetry operation that changes the sign of both). A look at the first term of Eq. (96) clearly shows that this would not allow such a thing. In fact, it is not possible to add or subtract any true scalar or vector in the splay term without violating the invariance of ( V . n ) 2 under the operation n + - a . Thus no ground state can exist with a spontaneous splay. Is there a way to introduce a vector in the bend term? There is. The bend transforms, obeying Eq. (6), as follows
[n x ( V x n ) I 2-+[ n x ( V x n ) - BIZ
(117)
where B is a vector. Only a vector B parallel to n x ( V x n ) can be introduced in the bend term. For undeformed smectic layers, n x ( V x n ) lies in he lzxnl direction, and B can be written
B=pzxn (1 18) where p is a scalar that must be zero in the nonchiral case. If p is nonzero, the medium is characterized by the local vector B and the reflection symmetry is lost. The form of the bend expression in the presence of a local polarization then corresponds to a constant spontaneous bend in the local frame of the director. The converse of this is the flexoelectric effect. Note that Eq. (117) conforms to Eq. (100). From the above reasoning we see two things: First, that this description permits the smectic C* to be polar and requires the
polarization vector to be perpendicular to the tilt plane, a result that we achieved before. Second, that the chiral and polar medium will be characterized by both a spontaneous twist and a spontaneous bend. The smectic C* is, in fact, such a medium where we have a space-filling director structure with uniform twist and bend. This nematic description has been very helpful in the past and permitted rapid solutions to a number of important problems [129, 1301. We will return to it in Sec. 7.1. If we insert the new twist term according to Eq. (1 16) into Eq. (96), the free energy attains its minimum value for n . V x n = -4 if splay and bend are absent (the cholesteric ground state). It means that we have a spontaneous twist in the ground state. This is connected to the fact that we now have a linear term in the free energy. For, if we expand the square in the twist term we could write the free energy as 1 q2 + K22 qn . ( V x n )(119) Gpp = (& + -K22 2
The last term is not reflection invariant and secures the lowest energy for right-handedness, which is therefore the lowest energy state. In the same way, if we expand the bend term (Eq. 117) we get the linear term -K33B [nx ( V Xn)] in the free energy. With both these linear terms present, the liquid crystal thus has both spontaneous twist and spontaneous bend. The general problem of tracing the deformations coupled to polarization in smectic C phases is complex, as has already been stated. We refer to [127] for the derivation in the incompressible case and to the discussion in [ 1281 for the general case. In the following we only want to illustrate the results in the simplest terms possible. This is done in Fig. 42, where we describe the deformations with regard to the reference system, k ,
2.4 The Flexoelectric Polarization
(3)
567
(4)
.
& (c.Vk.c)c (c.Vk.c)k
Figure 42. The six soft and the three hard deformations in a smectic C; these are coupled to the appearance of local polarization. Below each deformation is stated the covariant form of the independent vector field corresponding to the deformation. Five of the distortions (1,2,4,7,8) create a dipole density along the c (tilt) direction as well as along the k direction. The four other distortions (3,5,6,9) create a dipole density along thep direction, which corresponds to the direction of spontaneous polarization in the case of a chiral medium (C*). By the inverse flexoelectric effect, the distortions will be provoked by electric fields along certain directions. In more detail they can be described as follows: (1) is a layer bend (splay ink) with the bending axis alongp, ( 2 ) is the corresponding layer bend with bending axis along c. Both deformations will be provoked by any electric field in the tilt plane (c, k). (3) is a saddle-splay deformation of the layers, with the two bending axes making 45 ' angles to c and p . This deformation will be caused by a field component along p . (4) is a splay in the c director (corresponding to a bend in the P field if the smectic is chiral). It is generated by any field component in the tilt plane. ( 5 ) is a bend in the c director (in the chiral case coupled to a splay in P ) generated by ap-component in the field. (6) is a twist in the c director. It generates a dipole in the p direction (and is itself thus generated by a field in this direction). As this distortion is the spontaneous distortion in the helicoidal smectic C*, it means that we have a flexoelectric polarization along the same direction as the spontaneous polarization in such a material. Whereas these six deformations do not involve compression, the last three are connected with changes in the interlayer distance, described by the variable y=dulaz, where u is the layer displacement along the layer normal (z). Thus (7) is a layer compression or dilatation which varies along k and thereby induces a bend in the n director, connected with a dipole density, in the tilt plane. (8) is a layer splay inducing a splay-bend deformation in the n field, and thereby a dipole density, in the tilt plane. (7) and (8) are thus generated by field components in that plane. Finally, (9) is a layer splay perpendicular to the tilt plane, which induces a twist-bend in the n field alongp. The bend component then means a dipole alongp. Distortions (l), (2), and (4) are coupled among themselves as are (3), (3,and (6). The latter three, and in principle also (9), occur spontaneously in chiral materials.
c , p , where k is the local layer normal (along the direction of the wave vector for a helical smectic C * ) ,c is the local tilt direction, i.e., n tilts out in this direction (hence k, c is the tilt plane), and p is the direction of k xc (corresponding to the direction along which the spontaneous polarization has to
be in the chiral case). It should be stressed that k , c , andp are all considered unit vectors in this scheme, thus c does not give any indication of the magnitude of the tilt, only its direction. All in all there are 14 independent flexoelectric coefficients, 10 of which describe
568
2 Ferroelectric Liquid Crystals
five distinct deformations generating dipolar densities in the tilt plane. By the converse effect, these deformations are generated by electric field components lying in the tilt plane. A field component perpendicular to the tilt plane will generate the other four deformations (3, 5, 6, 9). which, consequently, themselves generate dipole densities along thep direction. In the figure, the deformations have been divided into three categories: those with gradients in k (1, 2, 3), those with gradients in c (4, 5, 6), and
au
those with gradients in y- -, i.e., gradients
aZ
in layer spacing (7,8,9) hard deformations. However, it is illuminating to make other distinctions. Thus deformations (l),(2), (4), (7), and (8) give quadratic terms in the free energy and are not influenced by whether the medium is chiral or not. As we have repeatedly stated, the flexoelectric effect is not related to chirality. However, a certain deformation which gives rise to a flexoelectric polarization, whether the material is chiral or not, may turn out to be spontaneous in the chiral case, like the C* helix. Thus the four deformations (3), (9,(6) and (9) turn out to give linear terms (like Eq. 119) in the free energy in the chiral case. Therefore we may expect them, at least in principle, to occur spontaneously. They all give contributions to the flexoelectric polarization in the direction alongp. At the incompressible limit, deformation (9) may be omitted. The remaining three deformations (3), ( S ) , and (6) are coupled to each other and have to be considered as intrinsic in the smectic C* state. Of these, the only space-filling structure is (6) (the same as illustrated in Fig. 41) and will therefore give rise to the dominating flexoelectric effect. Of the other two, deformation (3) in the chiral case means an inherent spontaneous tendency to twist the flat layers. This deformation actually only preserves a constant layer
thickness very locally, but not in a macroscopic sample, and will therefore be suppressed. In cyclindrical domains it will tend to align the c director 45 O off the cylinder axis [ 1271.Finally, deformation ( 5 )amounts to a spontaneous bend in the c director (corresponding to a splay in P).This means that P has the tendency to point inwards or outwards at the edges of the smectic layers. It means, on the other hand, that the effects are transformed to a surface integral so that the term does not contribute to the volume energy. Thus the general contributions to the flexoelectric effect are exceedingly complex and, at the present state of knowledge, hard to evaluate in their relative importance, except that the dominating contribution by far is the twisted (helicoidal) deformation. The consequences of the inverse flexoelectric effect are even harder to predict, especially in the dynamic case, without any quantitative knowledge of the flexocoefficients. When strong electric fields are applied to a smectic C (or C*) any of the nine deformations of Fig. 12 will in principle be generated, if not prohibited by strong boundary conditions. This consideration applies equally well to antiferroelectric liquid crystals as to ferroelectric liquid crystals. The fact that sophisticated displays work well in both cases seems to indicate that the threshold fields for these detrimental deformations are sufficiently high.
2.5 The SmA* - SmC* Transition and the Helical C* State 2.5.1 The Smectic C Order Parameter By changing the temperature in a system we may provoke a phase transition. The ther-
569
2.5 The SmA*-SmC* Transition and the Helical C* State
modynamic phase that is stable below the transition generally has a different symmetry (normally lower) than the phase that is stable above. The exceptions to this are rare, for instance, the liquid-gas transition where both phases have the same symmetry. In a number of cases of transitions between different modifications in solids there is no rational relation between the two symmetries. In such cases it may be difficult to construct an order parameter. In the majority of cases, however, the transition implies the loss of certain symmetry elements, which are thus not present any longer in the low temperature (condensed or ordered) phase. The symmetry group of the ordered phase is then a subgroup of the symmetry group of the disordered phase. In such cases we can always construct an order parameter, and in this sense we may say, in a somewhat simplifying manner, that for every phase transition there is an order parameter. Thus, in liquid crystals, the transitions isotropic ++ nematic ++ smectic A ++ smectic C f~smectic F, etc. are all described by their different specific order parameters. (There may be secondary order parameters, in addition.) The order parameter thus characterizes the transition, and the Landau free energy expansion in this order parameter, and in eventual secondary order parameters coupled to the first, has to be invariant under the symmetry operations of the disordered phase, at the same time as the order parameter itself should describe the order in the condensed phase as closely as possible. In addition to having a magnitude (zero for T > T,, nonzero for T< T J , it should have the same symmetry as that phase. Further requirements of a good order parameter are that it should correctly predict the order of the transition, and that it should be as simple as possible. As an example, the tensoria1 property of the nematic order parameter
correctly predicts that the isotropic - nematic transition is first order. Very often, however, only its scalar part S is used as a reduced order parameter, in order to facilitate discussions. In the case of the smectic A-smectic C transition, the tilt 8 is such a reduced order parameter. The symmetry is such that positive or negative tilt describes identical states in the C phase (see Fig. 43), hence the free energy can only depend on even powers in 8 and we may write, in analogy with the discussion in Sec. 2.2.5
For the time being, this simple expansion will be sufficient to assist in our discussion. A first-order SmA- SmC transition occurs if b 0 and T - T, = 0, 8 will be small and the aG = 0 then e6term can be dropped. Putting gives
a ( T - T,)B + be3 = 0
ae
(122)
with two solutions
e=o
(123)
which corresponds to the SmA phase (a check shows that here G attains a maximum value for T T,), and
\ // corresponding to the SmC phase (the extremal value of G is a minimum for T < T,). The transition is second order and the simple parabolic function of AT= T, - T at least qualitatively describes the temperature behavior of the tilt for materials having an SmA - SmC transition. However, the director has one more degree of freedom and we have to specify the tilt direction. In Fig. 43 we have chosen the director to tilt in the plane of the paper. By symmetry, infinitely many such planes can be chosen in the same way, and evidently we have a case of continuous infinite degeneracy in the sense that if all the molecules tilt in the same direction, given by the azimuthal angle cp, the chosen value will not affect the free energy. The complete order parameter thus has to have two components, reflecting both the magnitude of the tilt Band its direction cpin space, and can conveniently be written in complex form
y = Oe'q
(125)
With a complex scalar order parameter, the SmA-SmC transition is expected [131] to belong to the 3D XY universality class, which does not have the critical exponent p equal to 1/2. Nevertheless, experiments show [132] that it is surprisingly well described by mean field theory, although with an unusually large sixth order term. As all tilt directions are equivalent, the free energy can only depend on the absolute value squared of the order parameter. Actually, the first intuitive argument would say that there could be a linear term in the absolute value. However, physical descriptions avoid linear terms in absolute values, since these mean that the derivatives (with
tl
Figure 44. Functions of the kind G - 1 ~ 1 ,i.e., linear dependence in an absolute value, are with few exceptions not allowed in physics, because they imply discontinuities in the first derivative, which are mostly incompatible with the physical requirements.
physical significance) would have to be discountinous (see Fig. 44). Therefore, we assume that the free energy can only depend on
1yl2=Y * Y = Be-@x Be@= e2
(126)
The Landau expansion then has to be in powers of IY'I2= Y*Y and we see that the free energy is independent of cp, as it has to be. This is just equivalent to our first expansion. Generally speaking, G is invariant under any transformation Y - + Ye'W
(127)
This is an example of gauge invariance and the azimuthal angle is a gauge variable. The complex order parameter Y is shown schematically in Fig. 45 illustrating the conical degeneracy characteristic of the SmC
fluctuationsin 3!
Figure 45. Illustrations of the two-component order parameter Y = Qetqdescribingthe SmA-SniC transition. The thermodynamic variable Qandthe gauge variable qare completely different in their fluctuation behavior. The fluctuations in qmay attain very large values and, being controlled by an elastic constant scaling as Q 2 , may actually become larger than 27t at a small but finite value of 8.
2.5 The SmA*-SmC* Transition and the Helical C* State
phase. The gauge variable cp is fundamentally different in character from 8.The latter is a “hard” variable with relatively small fluctuations around its thermodynamically determined value (its changes are connected to compression or dilation of the smectic layer, thus requiring a considerable elastic energy), whereas the phase cp has no thermodynamically predetermined value at all. Only gradients Vcp in this variable have relevance in the energy. The result is that we find large thermal fluctuations in q around the cone, of long wavelengths relative to the molecular scale, involving large volumes and giving rise to strong light scattering, just as in a nematic. The thermally excited cone motion, sometimes called the spin mode (this is very similar to the spin wave motion in ferromagnets), or the Goldstone mode, is characteristic of the nonchiral SmC phase as well as the chiral SmC* phase, but is of special interest in the latter because in the chiral case it couples to an external electric field and can therefore be excited in a controlled way. This Goldstone mode is of course the one that is used for the switching mechanism in surface-stabilized ferroelectric liquid crystal devices. The tilt mode, often, especially in the SmA phase, called the soft mode (although “hard” to excite in comparison with the cone mode, it may soften at a transition), is very different in character, and it is convenient to separate the two motions as essentially independent of each other. Again, this mode is present in the nonchiral SmA phase but cannot be detected there by dielectric methods, because a coupling to an electric field requires the phase to be chiral. In the SmA* phase this mode appears as the electroclinic effect.
57 1
2.5.2 The SmA* - SmC* Transition If the medium is chiral the tilt can actually be induced in the orthogonal phase by the application of an electric field in a direction perpendicular to the optic axis, as shown in Fig. 46. The tilt is around an axis that is in the direction of the electric field. This is the electroclinic effect, presented in 1977 by Garoff and Meyer [21]. Soft mode fluctuations occur in the SmA phase just as in the SmA* phase, but only in the SmA* phase can we excite the tilt by an electric field. This is because a tilt fluctuation in the medium with chiral symmetry is coupled to a local polarization fluctuation, resulting from the transverse dipoles being ever so slightly lined up when the tilt disturbs the cylindrical symmetry of the molecular rotation. Superficially, it might be tempting to think that chiral and nonchiral smectic A (SmA* and SmA) would be very similar or even have the same symmetry, because they are both orthogonal smectics with the same organization of the molecules in the layers and with the same cylindrical symmetry around the layer normal. Nothing could be more wrong. As we have seen (in one striking example - several more could be given),
Figure 46. If an electric field E is applied perpendicular to the optic axis in a smectic A* (or any chiral orthogonal smectic), the optic axis will swing out in a plane perpendicular to the plane E , n defined by the optic axis and the field (electroclinic or soft-mode effect).
572
2
Ferroelectric Liquid Crystals
their physics is very different, due to the fact that they have different symmetry. Reflection is not a symmetry operation in the SmA* phase, and the electroclinic effect does not exist in the SmA phase just because the SmA phase has this symmetry operation. We may also apply the Curie principle to this phenomenon. The SmA phase has D,, (or -lm) symmetry, with one C, axis along the director (optic axis), infinitely many C2 axes perpendicular to this axis, and in addition one horizontal and infinitely many vertical mirror planes. The mirror planes are absent in the SmA* phase (D, or 0322). If we apply an electric field E (of symmetry C,) perpendicular to the C, axis, the only common symmetry element left is one C2 axis along E , which permits a tilt around C2. In Fig. 46 we also see that, in particular, the plane ( E , n) is a mirror plane in the nonchiral SmA phase, and consequently neither of the two tilting directors shown in the figure are allowed. In the SmA* phase ( E ,n ) is not a mirror plane, hence one of these tilt directions will be preferred. (Which one cannot be predicted as this is a material property.) An interesting aspect of this phenomenon is that it means that an electric field acting on the chiral medium actually exerts a torque on the medium. This is further discussed in Sec. 2.5.8. The torque is of course inherent in the chirality. The axial symmetry character is provided by the medium. The electroclinic effect is a new form of dielectric response in a liquid crystal. If we increase the electric field E, the induced polarization P will increase according to the first curve of Fig. 5. With a small field, P is proportional to E, and then P saturates. As the coupling between tilt and polarization is also linear at small values of tilt, we get a linear relationship between 8 and E. With 8 = e*E, the proportionality factor e* is called the electroclinic coefficient. It has an intrinsic chiral quality. (We have used an
asterisk here to emphasize this and also to clearly distinguish it from the flexoelectric e-coefficients.) In nonchiral systems, e * is identically zero. With present materials, 8is quite small (115 ”), but this may change with new dedicated materials. The electroclinic effect is also appropriately called the soft-mode effect, because the tilt deformation is a soft mode in the SmA phase, the restoring torque of which softens when we approach the SmA*SmC* transition, at which the deformation starts to “freeze in” to a spontaneous tilt. When we have such a spontaneous tilt as we have in the SmC* phase, we will also have a spontaneous polarization because of the rotational bias. It is important to note that the molecular rotation is biased in the nonchiral SmC phase as well as in the SmC* phase, but in a different manner. We may start with the “unit cell” of cylindrical symmetry of the SmA (or SmA*) phase and the corresponding cell of monoclinic symmetry of the SmC (or SmC*) phase, as in Fig. 47. We put a molecule in each one and ask whether the molecular motion (which, like the shape of the molecule, is much more complicated than here indicated) is directionally biased in its rotation around the optic axis (only roughly represented by the core part of the molecule). One simple way of representing this bias is to draw the surface segments corresponding to some interval in the rotation angle. Let us imagine that the molecule has a dipole perpendicular to its average rotation axis. We then find that the probability of finding the dipole in any particular angular segment is same in all directions in the SmA (SmA*) phase, as it has to be, because the optic axis is a C, axis. In the nonchiral C phase, the rotation cannot have this circular symmetry, but it must be symmetrical with respect to the tilt plane, as this is a mirror plane. This means that the rotational bias has a quadrupolar symmetry,
2.5 The SmA*-SmC* Transition and the Helical C* State
573
=t
9
A phase
Y
J
chiral molecule in A or A* matrix
phase is uniaxial
non-chiral molecule in C matrix
phase is biaxial and has quadrupolar order
chiral molecule in C or C* matrix
phase is biaxial and has polar order
Figure 48. Origin of the spontaneous polarization in a smectic C* phase. The illustration shows the directional bias for a lateral dipole in the rigid rod model of a molecule, with the directionality diagram for the short axis when the molecule is rotating about its long axis (from Blinov and Beresnev [45]).
I
I
tilt plane
Figure 47. The origin of the spontaneous polarization. The probability for a lateral dipole (averaged in time and space) to be in a unit angular segment is compared for the orthogonal, nonchiral tilted and chiral tilted phases. The common statement that the origin of polarization is a “hindered rotation” is quite misleading because, hindered or not, a rotation will only result in nonzero polarization for a specific (polar) directional bias.
which we can describe by a quadrupolar order parameter. If we now drop the mirror plane by making the phase chiral we will, in addition to this quadrupolar dissymmetry, have a polar bias giving a different probability for the dipole to be directed at one side out of the tilt plane rather than in the opposite direction. This will give a nonzero polarization density in a direction perpendicular to the tilt plane (see also the illustration in Fig. 48).Which direction cannot be predicted a priori but it is a molecular property. It is fairly obvious that the bias will increase as we increase the tilt angle, and
hence the polarization will grow when we lower the temperature below the SmA* SmC* transition. This transition occurs when the tilt 8 becomes spontaneous and is only weakly influenced by whether the phase is chiral or not. However, chirality couples 8 to P (which is a dramatic effect in a different sense) and brings P in as a secondary order parameter. There might also be a coupling between P and the quadrupolar order parameter, which will, however, be ignored at this moment.
2.5.3 The Smectic C* Order Parameters If we keep the tilt angle 8 in Fig. 48 fixed (just by keeping the temperature constant), we can move the molecule around the tilt cone, and we see that the rotational bias stays sterically fixed to the molecule and the tilt plane. The direction of the resulting P lies in the direction z x n . If we change the azimuthal angle by 180”, this corresponds to a tilt -8 (along the -x direction), and results in a change of P direction from y to -y. This means that we could tentatively write the relation between the secondary and pri-
574
2
Ferroelectric Liquid Crystals
mary order parameters as
P-0
(128)
or more generally that P is an odd function of 0 P = aO+ b e
+ ...
(129)
I
I$
' t e
8
Figure 49. Under the action of the twofold symmetry axis in the smectic C* state the tilt changes from 8 to 8+n.
As sin0 is an odd function, an expression that has this form is
P = Poz x rz = (PosinO)8
(130)
where we have introduced Po(with its sign, see Fig. 23 in Sec. 2.3.4), the magnitude of spontaneous polarization per radian of tilt. With 0 r z x n we may write &hetilt as a vector 0 = 8 8 to compare its symmetry with that of P. 0 is then an axial vector, while P is a polar. As we have no inversion symmetry in the chiral phase, this difference does not cause any problem, and if P and 0 transform in the same way we may make a Landau expansion according to the improper ferroelectric scheme of Sec. 2.2.6. If we take the scalar part of Eq. (130), it takes the form
P = Po sin0
(131)
This expression is often used but does not have the correct symmetry because it is not invariant under the C2 symmetry operation. For, if we change 8 to 8 + x, sin 0 changes to -sin 0 although P cannot change, because this is the same state (see Fig. 49). Therefore Eq. (131) should be replaced by 1 P=-&sin20 2
(132)
which is a C 2 invariant expression. In practice the expressions do not differ very much; P just increases somewhat slower at large tilts according to Eq. (132). For small values of tilt the expressions do of course give the same results, i.e., P=Po8. Experimentally, this simple linear relation between P and 0 often holds quite well, as shown in
Figure 50. (a) Schematic temperature dependence of the tilt angle 8 and polarization P at a second-order SmA* - SmC* transition. (b) Tilt angle versus temperature for DOBAMBC below the SmA*-SmC* transition at 95 "C (from Dumrongrattana and Huang [ 1331).
Fig. 50. However, there are exceptions, for instance, cases where P is a linear function of temperature rather than parabolic. Although the tilt angle has certain limitations, we will continue to use 0 as the order parameter for the time being. Finally, however, we will construct a different order parameter (Sec. 2.5. l l ) to describe the tilting transition.
2.5.4 The Helical Smectic C* State We did not change anything in the nonchiral order parameter (Eq. 125) when we pro-
2.5 The SmA*-SmC* Transition and the Helical C* State
ceeded to the chiral case, except that in addition we introduced a secondary order parameter P which does not exist in the nonchiral case. Now the normal state of the smectic C* is the helical C* state, i.e., one in which the tilt has a twist from layer to layer. At constant 6, q~is therefore a function of z (see Fig. Sl), with @z)=qz, q being the wave vector of the helix
575
z
(133) where the helix is right-handed for q > 0 and Z is the helical periodicity. For this case there is an obvious generalization of the nonchiral order parameter (Eq. 125) to y = eeiln(r) q
with
4 a= qz
(134)
(C)
Figure 51. (a) In the helical smectic C* state the constant tilt Bchanges its azimuthal direction cp such that cpis a linear function in the coordinate z along the layer normal. (b) The c director is a two component vector (cx, c,,) of magnitude sin 0, which is the projection of n on the smectic layer plane. (c) P makes a right angle with the tilt direction. The P direction here corresponds to a material with Po > 0.
where we have introduced a new length q-’ as an effect of chirality. Yqis now chiral by construction but does not make any change in the Landau expansion since Yq*Yqis still equal to 82, and we will still have the SmA* -SmC* transition as second order. We can also write Yqas
As the polarization vector is phase-shifted by 90” relative to the tilt vector Yq,we see from Eq. (136) that it can be written (see Fig. 5 1c) as
Yq= B(cosqz + i sinqz)
P = (-P sinqz, Pcosqz, 0)
(135)
It is interesting to form the divergence of this vector. We find
and as a vector Yq= (Qcosqz, esinqz, 0)
(1 37)
(136)
or
Yq= ec
(136a)
where we have introduced the “c director”, the projection of the director n on the smectic layer plane (see Fig. 51b). Note that the c director in this description (which is the common one) is not a unit vector, as it is in the movable reference frame k , c , p used, in contrast to the space-fixed reference frame x, y, z, in Sec. 2.4.9 to describe the flexoelectric deformations.
This is an important result as it means that the helicoidal C* state is a divergence-free vector structure P(r) in space. Hence we have no appearance of space charges anywhere. Thus not only does the helix cancel the macroscopic polarization and thereby any external coulomb fields, although we have a local polarization P everywhere, but the fact that V -P = 0 also secures that there are no long range coulomb interactions in
576
2
Ferroelectric Liquid Crystals
the material due to space charge. The helical state is therefore “low cost” and “natural” from several points of view. A homogeneous state of the director, as can ideally be realized by surface stabilization, of course also has V . P = O , although it has an external coulomb field and therefore is not stable by itself. However, as we will see later, the condition V . P = O cannot be maintained everywhere if we have a local layer structure of chevron shape (see Sec. 2.8.3). For the sake of completeness, but also for distinction, let us finally make a comment on an entirely different kind of chiral order parameter. Let us imagine that the substance we are dealing with is a mixture of two enantiomers (R) and (S). We can then define a scalar quantity R-S K=-----R+S
(139)
where R and S stand for the relative concentration of (R) and (S) enantiomers. This quantity can in a sense be regarded as an order parameter. K = 1 for “all R’, -1 for “all S”, and zero for a racemic mixture. Thus in general, 0 I I K I 5 1. This quantity is nothing less than what organic chemists introduced long ago, but in a completely different context. It is sometimes called “optical purity”, but is today rather called enantiomeric excess. It is an extremely useful quantity in its right place, but it is not relevant as an order parameter because it does not depend on the temperature T, only on the concentration, which is trivial in our context. It does not help us in introducing chirality and we will not have use for it.
2.5.5 The Flexoelectric Contribution in the Helical State We note from Fig. 5 1 that the projection of the director on the smectic layer plane is
n sin 6= sin 6. If we write the components of n in the x,y,z system, we therefore have n, = sin6 cosq n,, = sin0 sinq n, = cos6 If we insert v>(z) = qz for the helical state this gives,
n, = sin6 cosqz n,, = sin6 sinqz n, = cos6 This is a twist-bend structure, as we have repeatedly stated, and is therefore connected to a flexoelectric polarization of size (see Eq. 97) P, = ebn x ( vx n )
(142)
We will now calculate this contribution. Forming the curl as (143)
Y = - q sin 6 sin 6 cos qz sin &in qz cos 6
‘ i
X .
cos qz
sin qz
o i
I
2.5
The SmA*-SmC* Transition and the Helical C* State
Therefore we can write the flexoelectric polarization due to the helical deformation as
1 Pf =-eb.-qsin28 2
(147)
Note that this polarization grows with the tilt according to the same function as the spontaneous polarization P, in Eq. (132). We may then write the total polarization density P = P, + Pfas
P=-(Po 1 -ebq)sin28 2
(148)
or, for small angles P = (Po - e,q) 8
(148 a)
If Po and eb have different signs, the two contributions will cooperate, otherwise they tend to cancel. (This is not changed if we go from a certain chiral molecule to its enantiomer, as q and Po will change sign simultaneously.) In the absence of a smectic helix, q = 0, and the flexoelectric contribution will vanish. The sign of Po has been determined for a large number of compounds, but sign (and size) determinations of flexoelectric coefficients are almost entirely lacking. Such measurements are highly important and should be encouraged as much as possible. However, recent measurements by Hoffmann and Kuczyn'ski [134] show that the flexoelectric polarization is of the same magnitude as the spontaneous one. Equation (148) may also be a starting point for a new question. In the absence of chirality, thus with Po=O, could the flexoelectric effect lead to a helical smectic state in a nonchiral medium? This will be the topic of our next section.
577
2.5.6 Nonchiral Helielectrics and Antiferroelectrics As we have repeatedly stressed, flexoelectricity is a phenomenon that is a priori independent of chirality. But we have also seen that some flexoelectric deformations do have a tendency to occur spontaneously in a chiral medium. All except the helical C* state are, however, suppressed, because they are not space-filling. A flexoelectric deformation may of course also occur spontaneously in the nonchiral case, namely, under exactly the same conditions where the deformation is space-filling and does not give rise to defect structures. In other words, in creating the twist-bend structure which is characteristic of a helielectric. Imagine, for instance, that we have mesogens which have a pronounced bow shape and, in addition, some lateral dipole. Sterically they would prefer a helicoidal structure, as depicted in Fig. 52, which would minimize the elastic
Figure 52. Space-filling twist -bend structure of strongly bow-shaped achiral molecules equally split into right-handed and left-handed helical domains. (Here a right-handed domain is indicated.) A bowshaped molecule of this kind is illustrated in Fig. 53. The kind of domain depicted in this figure would correspond to the normal helical antiferroelectric organization in a smectic C*. As a result of the two-dimensional fluidity of these smectics, the cancellation can, however, be expected to occur on the smallest possible space scale.
578
2 Ferroelectric Liquid Crystals
energy, because the spontaneous bend B would cancel the bend term (Eq. 117) in the Oseen-Frank expression. The resulting local polarization from Eq. (148) with Po= 0 may still not be too costly because the external field is cancelled and V. P = 0. This would have an interesting result: because the starting material is nonchiral, we would observe a spontaneous breaking of the nonchiral symmetry leading to equal regions with left-handed and right-handed chirality. It would correspond to the well-known examples of SiO, and NaC10, discussed in Sec. 2.3.4. It seems that finally this kind of phenomenon may have been observed in liquid crystals [135]. However, in contrast to the cases of SiO, and NaClO,, the helical structures are probably possible on a molecular level as well as on a supermolecular level. Thus we may expect domains of nonchiral molecules in different conformations, right and left-handed, which behave as if they belong to different enantiomeric forms. The possibility that a space-filling flexoelectric deformation will be spontaneous for certain molecular shapes and thereby create a chiral structure out of nonchiral molecules will be much enhanced if the deformation can take place in the layer rather than in the interlayer twist -bend structure of Fig. 52, and may then lead to antiferroelectric (rather than helielectric) order similar to that in antiferroelectric liquid crystals made of chiral compounds. The polarization may very
QrN ‘gH17
Figure 53. Bow-shaped nonchiral molecule which may create chiral C* domains (from [135]).
well be switchable, because it is not connected to a supramolecular director deformation. In other words, the deformation represented to the lower right of Fig. 32 now applies to the single molecule and this can be flipped around by the electric field. Further investigations of these new materials are important and will shed light on a number of problems related to polarity and chirality. Whether we will find mechanisms for complete spatial separation reminiscent of the classic discovery by Pasteur, who could separate crystallites of tartaric acid into right-handed and left-handed pieces, is too early to speculate about, but the possibilities, as always in liquid crystalline systems, are quite rich.
2.5.7 Simple Landau Expansions In order to sharpen the observance for inconsistencies in a theory, it is sometimes instructive to present one which does not work. For instance, let us see if we can describe the transition to a polar nematic phase by a Landau expansion. It might also be thought of as describing the transition isotropic-polar smectic A or nematic-polar smectic A. (In either interpretation it will suffer from serious deficiencies.) We write the free energy as
G=-aS 1 2+-bP 1 2 -cSP2 2 2 where S is the (reduced) nematic order parameter and P the polarization. The same abbreviation a =a (T- To) and sign conventions are used as before. The form of the coupling term is motivated by the fact that S and P have different symmetries, quadrupolar and polar, respectively. (It is also easy to persuade oneself that a term -SP would lead to absurdities.) G now has to be minimized with respect to both order parame-
2.5
The SmA* -SmC* Transition and the Helical C* State
ters: putting aG/aP and aG/aS equal to zero gives b P - 2~ S P = 0
(150)
a S- cp2= 0
(151)
from which we deduce
(153) Whether it is already not clear physically how a non-zero S-value would couple to a non-zero polarization, we see that Eq. (153) describes a P(7‘) increasing below the transition point only if b < 0, at the same time as S is essentially independent of temperature. In order to give S > 0, we have then to require that c T,. As the tilt induced in the A* phase by an external electric field is quite small we first discard the O3 term and get
a(T-TC)8=~o~c*E
(182)
This gives the induced tilt angle
It is thus linear in the electric field in this approximation, but strongly temperature
dependent, diverging as we approach T, from above. In the A* phase the relation between 8 and E is almost as simple as that between P and 8 in the C* phase, while the relation between P and 8 is not as simple. Equation ( 183) describes the electroclinic effect. We may write it 8 = e*(T)E
(1 84)
where e* is the electroclinic coefficient
showing the close relationship between the chiral material parameters e*, c* and s*. Specifically we see that the structure coefficient s* is the non-diverging part of e*.
2.5.8
The Electroclinic Effect
If we assume (which is not quite true, but empirically fairly well justified) that the same simple relation P = s* 8 is valid also for T > T,, the combination
p=s*e, 8=e*E
(186)
is equivalent to P=s*e*E With e* taken from Eq. (185) we see that
i.e., that part of the polarization which corresponds to the induced tilt, diverges in the same way as the primary order parameter. (There is of course also a non-diverging part of P). While 8 - s * in Eq. (184), P-s**, which has to be the case as P , unlike 8, cannot depend on the sign of s*. We note, furthermore, that as e* and s* only differ in the temperature factor they always have the same sign. This means that the electroclin-
584
2
Ferroelectric Liquid Crystals
ically induced tilt or polarization will never counteract a spontaneous tilt or polarization, except in the sense that in a switching process - because the electroclinic effect is so fast - the electroclinic contribution may temporarily get out of phase from the ferroelectric contribution. In order to get the correct relation between P and 8 for T > T, we have to start from the generally valid Eq. (174) which we write as
sponding to Eq. (1 88). The first is the “background” part corresponding to s* = 0, ie.e to a racemic or non-chiral substance. The electroclinic effect is linear only if we do not approach too closely to T,. At the transition, T = T,, the first term in Eq. (1 8 1) vanishes and the relation between 8 and E becomes
If we assume the linear relationship 8 = e* E to hold - which is not true near T, - then
This nonlinear behavior with saturation is well known experimentally. It means that the electroclinic coefficient becomes fielddependent and falls off rather strongly (-E-2’3) at high fields. It also means the same for the electroclinically induced polarization which acquires relatively high values at low applied fields but then shows saturation. It explains the shape of the measured P ( T ) curves near T, as illustrated in Fig. 18 (dashed line). For T# T, but near T, we have to keep all the terms in Eq. (18 1). We may write this expression in the form
P=s*8+s*8 c*e* and inserting e* from Eq. (185)
The first term slowly decreases with in- 1/T) whereas the creasing temperature second slowly increases. This explains the empirically found result that P / 8 is essentially the same (=s*) in the C* phase as well as in the A* phase. The ratio is thus practically the same whether P and 8 are spontaneous or induced. Within the approximation of small tilt, Eq. (182), we now insert the value for Bfrom Eq. (183) in the expression Eq. (189) for P. This gives
(xo
be3 = S* E
( 1 94)
or
A 8 + Be3 = E
( 196)
with
a ( T - T,) A =< S
(197)
and B=- b S”
or (193) The second part is the electroclinic contribution to the dielectric susceptibility corre-
Generally it can be said that Eq. (196) very well describes the experimental results. A particularly careful study has been made by Kimura, Sako and Hayakawa, confirming both the linear relationship Eq. (197) and the fact that B, and therefore b as well as s*, are independent of temperature [ 1891.
2.5
The SmA*-SmC*Transition and the Helical C* State
The electroclinic effect is intrinsically very fast. The torque acting on the director, and giving rise to a change in tilt, is obtained by taking the functional derivative of the free energy with respect to the tilt, (199)
Equation (157) combined with Eq. (164) then gives
585
axis relaxes back to its state along the layer normal from a beginning state of nonzero tilt. Equation (202) in this case simply reads
which is directly integrated to
with the characteristic time constant
-re=a0 + b02 - c*P = a ( T - T,)O+ b03 - s*E
(200)
= a0- x O q c * 2 0+ be3 - x O q c *E
As evident from the middle one of these equations, we confirm, by Eq. (194) that this torque vanishes when we approach T = T,, where 0 shows a diverging tendency and we can expect a critical slowing down in the response. The viscous torque r" counteracting any change in 0 is, by definition,
If instead a constant field Eo is applied at t= 0, the growth of 8 will be given by
which is integrated to
with the saturation value of 0 equal to
ao ru= -yQ at where '/e is the electroclinic or soft mode viscosity. In dynamic equilibrium the total torque r=re+T v is zero. This gives
This equation describes the dynamics of the electroclinic effect in a material where we have a lower-lying phase (T< T,) characterized by a spontaneous tilt of the optic axis. If we divide all terms by s*, it can be written
A 8 + B03 +($)b = E This is the dynamic equation corresponding to Eq. (196). For small induced tilts, however, we skip the O3 term in Eq. (202). If we further put E=O, we will see how the optic
which of course also corresponds to Eq. (183). If we finally apply a square wave, where we reverse the polarity from -EO to +Eo at t=O, the optic axis will change direction by an amount 20, according to
corresponding to the initial and final values
e(o) = -eO,o ( ~= )+oo.
We note that the response time given by Eq. (206) is independent of the applied electric field. In comparison, therefore, the response time due to dielectric, ferroelectric and electroclinic torque is characterized by different powers of E, as
586
2
Ferroelectric Liquid Crystals
The reason for this apparent peculiarity in the electroclinic effect is of course that if the field E is increased by a certain factor, the saturation value of the tilt is increased by the same factor. The angular velocity in the switching is doubled if we double the field, but the angle for the full swing is also doubled. Thus, electrooptically speaking, the speed of a certain transmission change in a modulator certainly does increase with increased field. But one mystery remains to be solved. How come, that the characteristic time zin Eq. (205) describing the relaxation back to equilibrium, is the same as the “response time” z in Eq. (208) describing the change of 8 under the influence of the electric field? Should not the relaxation back be much slower? This is what we are used to in liquid crystals: if a field is applied to a nematic, the director responds very quickly, but it relaxes back slowly when we take the field off. The answer is: the electric field exerts no torque at all on the director in the electroclinic action. Its effect is to shift the equilibrium direction in space of the optic axis. If the field is high the offset is large and the angular velocity in the director motion towards the new equilibrium state will be high. Equation (207) describes this motion and tells that the rate of change of t9is proportional to the angular difference between the initial and final director state. In fact Eqs. (204) and (207) are the same equation and can be written
The difference is only that the final state 0, in Eq. (204) is zero, whereas in Eq. (207) 80 is the value given by the electric field according to Eq. (209). Note that in Eq. (214) the electric field does not appear at all. The electric field E does, however, exert a torque on the whole medium during the electroclinic action. This torque, by conservation of an-
gular momentum, is opposite in sense to the torque exerted by the “thermodynamic force” on the director. As already mentioned in Sec. 1 of this Chapter, the electroclinic effect was first announced as a ‘‘piezoelectric’’effect [20], but shortly thereafter renamed. The similarities are interesting but the differences sufficient to characterize the electroclinic effect as a new dielectric effect. The variable 8 in which the “distortion” takes place is here an angular variable, cf. Fig. 55. If E is reversed, P and 8 are reversed. But there is no converse effect. (One cannot choose 26 in the distortion variable to induce a P of a given sign.) The electroclinic tilt is connected with a mechanical distortion, a shrinking of the smectic layer thickness d. In the simplest model, assuming tilting rigid rods instead of molecules, thus exaggerating the effect, the thickness change is
6d = d ( i -case) - e2
(2 15)
-
hence 6d E2, not -E and we have an electrostrictive, not a piezoelectric distortion. This is also consistent with the fact that no
t
1
t
I
Figure 55. Analogy between the piezoelectric and the electroclinic effect. A translational distortion in the former corresponds to an angular distortion in the latter. If the sign of the applied field E is reversed, the sign of P and 8 is also reversed. But there is no converse effect in the electroclinic case. The mechanical deformation is electrostrictive, i.e. proportional to E’. In the smectic C* case there is also a component proportional to E.
2.5
587
The SmA*-SmC* Transition and the Helical C* State
dilatation is possible in the A* phase, thus the effect is inherently unsymmetric. (Note that this is not so in the C* phase; with a finite 0 dilatation as well as compression is allowed.) The electrostriction may be a problem in the practical use of the electroclinic effect, because the field-induced layer shrinking, at least sometimes, leads to the appearance of layer kink defects, which may be observed as striations in the texture [ 1901. However, very little is known today of how the electrostrictive properties vary between different materials. The electroclinic effect is the fastest of the useful electro-optic effects so far in liquid crystals, with a speed allowing MHz switching rates in repetitive operation. Its electro-optical performance is discussed and compared with other modes in the review of reference [191]. A number of device applications are discussed in reference [140]. The effect appears in all chiral orthogonal smectics but has been studied in a very restricted class of materials. It suffers so far, except from the fairly strong temperature sensitivity, in particular from a limitation in the value of induced tilt angle (12 to 15 degrees). This means that only a very small part of the typical V-shaped transmission curve can be utilized for a modulator using this kind of material, cf Fig. 56. In this respect short pitch C* or N* materials are much better (exploiting the deformed helical and flexoelectro-optic mode, respectively) with a sweep in either direction of more than 22 degrees in the first and more than 30 degrees in the second case. However, new A* materials may very soon change the scene. The ideal material for the electroclinic effect would be a so-called de Vries compound [191a] in which the molecules have a large but unbiased tilt in the A phase (adding up just as in the short pitch C* or antiferroelectric case, to an optic axis along the layer normal) but a biased tilt in the C* phase.
-450
-22.5"
00
22.5"
45"
Figure 56. Modulation of the transmitted light intensity I for all electro-optic effects describing a tilt of the optic axis in the plane parallel to the cell plates (in-plane switching), for the case that the initial axis direction is along one of the two crossed polarizer directions. The curve is drawn for a cell thickness corresponding to A/2 plate condition, for which a tilt sweep of 45 degrees gives full modulation (from zero to 100 percent transmitted light). If this condition is not met, the degree of modulation will be reduced.
WewillreturntosuchmaterialsinSec. 2.8.10 because they are also of eminent interest for high resolution displays. Up to now they have been somewhat hypothetical but seem recently to have been confirmed [ 1921. Such de Vries materials would be ideal even if they have to be somewhat slower (the torque counteracting tilt fluctuations would not be so strong for such compounds) because first of all the electroclinically induced tilt would be very large, namely up to the full value of unbiased individual tilt, and this for an applied field that can be expected to be very low. Furthermore the tilt biasing does not involve any electrostrictive effect or, at least this would be minimal. Finally the temperature sensitivity can be expected to be very low as it is in the absence of collective behavior, essentially T-' instead of (T - T J ' .
2.5.9 The Deformed Helix Mode in Short Pitch Materials An interesting electro-optic effect utilizing the helical SmC* state of a short pitch ma-
588
2 Ferroelectric Liquid Crystals
terial was presented in 1980 by Ostrovskij, Rabinovich and Chigrinov [193]. In this case the twisting power was so high that the pitch came below the wavelength of light. In such a densely twisted medium the light wave averages out the twist to feel an optic axis directed along the helical axis, i.e. in the direction of the layer normal z (Fig. 57). When an electric field is applied perpendicular to this axis it starts to partly untwist the helix and in this way perturb the spatial direction of the average. The result is a linear
E-0
E>O
E pitch) (from Beresnev et al. [194]).
effect very similar to that of the electroclinic effect and the flexoelectro-optic effect: the angular deviation (8)of the effective optic axis is proportional to the applied field. Like the other two mentioned effects this deformed-helix mode (DHM) has no memory but a continuous grey scale. Though not as rapid as the electroclinic effect it has at least two distinctive advantages. First, (0) can attain much higher values than the soft-mode tilt. Second, the apparent birefringence (An) is quite low, allowing a larger cell thickness to be used and still matching the ;1/2 condition. The short-pitch materials have shown another highly interesting property. When they are used in the SSFLC mode, i.e. when a field sufficient for complete unwinding is applied and then reversed, in order to actuate the switching between the two extreme cone positions, a new mode of surface-stabilization is found, in the limit Z+d, where 2 is the pitch or helical periodicity and d is the cell gap thickness. It may at first seem surprising that surface stabilization can be achieved not only for d IZ but also for d Z, in fact the smaller the value of 2 the better. In the limit Z 4 d it is a pinning phenomenon which makes it distinctly different from the first case of surface-stabilization. In other words, whereas for Z> d the non-helical state is the equilibrium state, for 2 6 d it is a metastable state. The helix thus tends to rewind but, for topological reasons (the back-transition is defect-mediated) it may stay for a considerable time. This means that short-pitch materials are very interesting alternatives to pitch-compensated materials for display applications. Their demonstrated inconvenience so far has been the difficulty to align then well, generally giving somewhat scattering textures. One further advantage of this effect is that the response time is only weakly temperature dependent. A general discussion of this and other fea-
+
2.5
The SmA*-SmC*Transition and the Helical C* State
tures is found in chapter VI of reference [41]. The DHM has been developed further by the Roche and Philips groups and applied in active matrix displays with grey scale, cf references [195] and [196].
2.5.10 The Landau Expansion for the Helical C* State As we repeatedly pointed out, the Landau expansion Eq. (157) used so far
1 2 +-be 1 4 +- l P 2 G=-aO 2 4 2x0 EO -c* P O - PE
(157)
refers to the non-helical C* state. Even if the non-helical state is somewhat special - the general case is one with a helix - and if it is true that for many purposes we may even regard the presence of the helix as a relatively small perturbation, it is now time to take the helix into account. The first thing we have to do then is to add a so-called Lifshitz invariant. This invariant is a scalar
589
Inserting this in Eq. (216) we find that the Lifshitz invariant (which has a composition rule slightly reminiscent of angular momentum, cf. L,=xp,-yp,) in the cholesteric case has the value equal to q, the wave vector. In fact we can gain some familiarity with this invariant by starting from an expression we know quite well, the Oseen-Frank expression for the elastic free energy Eq. (96). Because of its symmetry, this expression cannot describe the cholesteric state of a nematic which lacks reflection symmetry and where the twisted state represents the lowest energy. Now, if there is a constant twist with wave vector q, the value of n . V x n in the K2, term equals -4. The expression Eq. (96) therefore has to be "renormalized" to
G=1 Kl I (V.n)2 + 1 K22 (n . v x n + q)2 2 2 + -21 K33 ( n x v x n)2 (2 18) which, if there is no splay and bend present, becomes zero for n V x n = -4. If we expand the square in the K22term we can write the energy in the form GN* = GN + K22 q n . V X12 + 1 K22 q2 (219) 2
permitted by the local C2 symmetry which describes the fact that the structure is modulated in space along the z-direction. If in Fig. 49 we let the x direction be along the C, axis, we see that the C2 symmetry operation involves y +-y, z +-z, ny +-ny and nz+-n,, of which the first and last do not interfere and the middle two leave Eq. (216) invariant. We remember from Fig. 29 that the cholesteric structure has infinitely many C, axes perpendicular to the helix axis. The expression Eq. (2 16) is therefore also an invariant for the N* phase. If q is the cholesteric wave vector, we write the director as
n = (cos$, sin$) = (cosqz, sinqz)
(217)
where G, corresponds to the energy expression Eq. (96) for a non-chiral nematic. Two new terms have appeared in the energy as a result of dropping the reflection symmetry. The first is a linear term which is the Lifshitz term, the second is an energy term quadratic in the wave vector. In fact, a simple check shows that
590
2
Ferroelectric Liquid Crystals
We could therefore alternatively write
GN*=GN -K22q
n x z - n ya nx x ) + 1K22Y(221) 2
(
any
or use L=-n.Vxn
(222)
instead of Eq. (216). We may finally note that if we go from a right-handed to a lefthanded reference frame then the Lifshitz invariant (like angular momentum) changes sign because Vxn does. We can also see this in Eq. (216) because in addition to the C, operation the inversion involves that x +-x and n, +-nx. In the helical smectic C* case, again with q(z) = qz, the director components were stated in Eq. (141) n, = sin0 cos@ ny = sin0 sin$ n, = cos0
Inserting this in Eq. (216) we find that the Lifshitz invariant equals
L = q sin20
(223)
which is also often written
aq . 0 L=-sin
aZ
for the more general case that the twist is not homogeneous. We may also note, that in the smectic C* case we can equally well write Eq. (216) in the form
seen in Sec. 2.5.5, with a helical deformation we also have a flexoelectric contribution to P. This is opposite in sign to P, for e,>O, cf. Eq. (148), and proportional to the wave vector q. These two contributions must now be separated in the Landau expansion because we have to regard q as a new independent variable. In that case we not only have a coupling term -c*PBbut also one corresponding to the flexoelectric contribution which we could write eqP0. With this convention, Eq. (157) would be extended, for the helical C* case, to G = - 1a 0 2 + - 1b e 4 +- 1 p 2 2 4 2x0 Eo
where the -PE term has been skipped since we are here only interested in the field-free case. For the Lifshitz invariant we have used Eq. (223) in the limit of small 0, but chosen the minus sign, corresponding to A* > O in Eq. (219). In accordance with Eq. (219) we have also added a term proportional to q2. The origin of the last term, proportional to q202 is the fact that the smectic helix is a combined twist-bend deformation and that its energy therefore has to depend on the tilt angle as well as on q. By symmetry this dependence has to be -0’ (or, actually proportional to the square of sin 20). But it is illustrative to examine these last terms in a less ad hoc way. Thus, if we go back to our “nematic” description, introduced in Sec. 2.4.9, we can write Gc.
i.e. using the two-dimensional c director, as the n, component does not appear in L. Before we add the Lifshitz invariant to our Landau expansion Eq. (157) another consequence of the helix may be pointed out. It is the fact that, as we have already
=z1 K 2 2 ( n . V x n + q ) 2
(227)
+ 21 K33(n x V x n - B ) 2 ~
= Gc
+ 21 K22q2 + 21 K33 B2 -
~
+ K 2 2 q n .V x n - K33B . n x V x n
2.5
The SmA*-SmC* Transition and the Helical C* State
where Gc corresponds to the expression (96) without splay term. As the spontaneous 1 bend B is equal to - q sin 2 8 = q e for small 2 tilts, cf. Eq. (147), the third term on the right side is equal to the last term in Eq. (226). From Eq. (145) we further know that n x V x n =-qsinecose z x n for the helical C* case, hence the last term in Eq. (227) can be written K3,Bqe for small tilts. With P= ebB for the flexoelectric contribution we write this term eqPB, with the abbreviation K33/eb=e. This term was already included in Eq. (226). Before continuing with the Landau expansion we note that with n V x n =-q and n x V x n = B inserted in Eq. (227), the free energy in the helical C* state can be written in the form 1
G p = G,
- (K22
+ K33e2)q2
(228)
(K33q - Xo%e2q)e2+ K22q = (a* - xo%c*) e2
59 1
(234)
or K f q 2 0 2+ K22q= A*,@
(235)
We have here introduced the renormalized coefficients Kf = K33 - xo%e2 ,
a*-- a* - Xo%ec* (236) f
which contain the flexoelectric contributions. The wave vector is then obtained as
g e2 '=
KZZ+Kfe2
(237)
For a non-chiral material, Af* = 0 (both A* and c* are chiral coefficients in Eq. (236), and hence q = O . The pitch of the smectic helix Z equals 27c/q, thus
With the new terms in Eq. (226), minimizing with respect to P, 8 and q gives
z=K22 + Kf O2 n;-* e2
a'ae = a 0 + b e 3 - c* P + eq P - 2 a * q e + K33e q 2 = 0
indicating that the pitch should diverge at the transition C* +A*, cf Fig. 58. Experimentally, the situation is still not very clear regarding Z(T). While some measurements do indicate that a pure divergence exists if the layers are parallel to the cell plates, the majority of measurements give a result that the pitch goes through a maximum value
(230)
The first of these yields
ZPP) J
I I
Comparing this with Eq. (148) which, for small tilts we can write P=(Po-qeb)8
(233)
we find that Po corresponds to ~ o & o c * , equal to s* in Eq. (167), and that t?b corresponds to Xo%e, where e is the coefficient Tc in the Landau expansion' we now inFigure 58. The measured helical pitch Z at the transert p from Eq. (232) into Eq. (231) which sition C*-tA*. The pitch should diverge according gives to Eq. (238).
592
2
Ferroelectric Liquid Crystals
slightly below the transition and attains a finite value at T= T,. If we write the pitch dependence
z=z,+CC2
(239)
there should be a saturation value Z, = ~ ~ K ~ at I Alow; temperatures. Z, as well as C= 2 7 ~ K ~ ~ shows l A ; the balance between the non-chiral elastic coefficients ( K ) wanting to increase the pitch and the chiral coefficients (A*)wanting to twist the medium harder. When we finally insert the value of P from Eq. (232) into Eq. (230) we get
xosoc * 8~ + 2 x o E~ ec* q e - xosoe2 q28 a e + be3 -
- 2 a * q e + ~ ~ ~ e q ~ = o (240) After skipping the solution 8= 0, we are left with -
xosoc * -~2 (A* - xosoec*) q
(241)
+ ( K - xo s g e 2 )q2 + b e 2 a -xo so c * -~2 g q + Kf q2 + b e 2 = 0 Together with Eq. (237) this leads to a complicated expression for 8( T ) ,which we will not pursue further here. However, in the approximation that we take the low temperature limit for the wave vector a*
4 40 = Kf (corresponding to 2,) we see that Eq. (241) simplifies to a - x , ~ c *- Kfqi ~ + b82
(243)
which is reshaped in the usual way to
a ( T - T,)
+ b82
(244)
with a phase transition shift T, - Toequal to (245)
While AT is still expected to be generally small, it is certainly true that the contribution from unwinding the helix may raise the transition temperature more than the contribution from the chiral coupling, at least in the case of hardtwisted media like shortpitch smectics C* where the pitch is smaller than the wavelength of light.
2.5.11 The Pikin-Indenbom Order Parameter A group theoretical symmetry analysis by Indenbom and his collaborators in Moscow [197,198] led, around 1977, to the introduction of a very attractive order parameter for the A*-C* transition which usually is referred to as the Pikin-Indenbom order parameter. It is attractive because it presents, in the simplest form possible, the correct symmetry and a very lucid connection to the secondary order parameter P.It was adopted by the Ljubljana group around Blinc for the description of both static and dynamic properties of the C* phase [199-2041. The basic formalism has been described in particular detail by Pikin [40] and by Pikin and Osipov [41]. In the smectic C* phase, the layer normal z and the director n represent the only two natural directions which we can define. That is, z and n are the only natural vectors in the medium. Using these we now want to construct an order parameter which is a vector, is chiral, has C2 symmetry, leads to a second order A* - C* transition and is proportional to 8for small values of 8.We can do this by combining their scalar and vector products, multiplying one with the other,
N = (n . z ) (n x z)
(246)
We then obtain a vector in which the first factor is invariant, while the second changes sign on inversion. N thus has the two first
593
2.5 The SmA*-SmC*Transition and the Helical C* State Y
properties. Performing the multiplication we find
N = nz(ny,-n,, 0)
(247)
which only has two components, hence is written
N = (nzny,-nznx)
(248)
The n,ny and nznx combinations do not change on inversion but the x and y directions do, thus we see that N changes sign on inversion. The angle between n and z is 8. Therefore, taking the absolute value in Eq. (246) 1 N =cos8sin8=-sin28 2
(249)
We then recognize that this order parameter automatically reflects the C2 symmetry, cf. the discussions around Eqs. (131), (132) and Figure 49, where we had to introduce this symmetry in an ad hoc way. Clearly N = 8 is valid for small tilts according to Eq. (249). Finally, we can check that N and -N describe the same state from Eq. (248), by virtue of the fact that n and -n describe the same state. Thus, if we make a Landau expansion in N , only even powers can appear and the transition will be one of second order. Before going on to this expansion we should, however, make the connection to the secondary order parameter P . Looking onto the smectic layer plane in Figure 59 we have (for a positive material, Po > 0) a relation between the P direction and the tilt direction such that P advances in phase by 90 degrees in the positive cp direction. If we express P as a complex quantity P, + iP,, we can relate P directly to the C director expressed in complex form, c = c, +icy, by P,
+ iPy - -i(cx + icy)
Figure 59. The tilt vector n,c lags P in phase by 90 degrees (for Po> 0), corresponding to a multiplication by -i in the complex plane.
(250)
Let us now introduce the tilt vector 6 as the complex quantity nzc= (n,c,+inzcy) =
(n,n,
+ in,ny) and rewrite Eq. (250) as
P, + iPy - -i(n,n,
+ in,ny) = (n,ny - in,n,)
(251)
,
In vector notation the tilt vector is 6 = ( 5 , { 2 ) = ( n 2 n x , n p Y ) and Eq. (251) in vector form reads (PX?py>- ( 5 2 , -51)
(252)
This is the relation between the secondary order parameter P = (P,, Py)and the primary order parameter N = (nzny,-nznx),when the latter is expressed in the tilt vector 6 = ( 52).
We may now proceed to ask what invariants we may form in N and P , and their combinations, to include in the Landau expan1 1 sion. N 2 and N4 give -a({:+ 5;) and -b 2 4 corresponding to our earlier 8' and O4 terms, P 2 gives in the same way the
(ct+{;)2
(P,'+ Py")term. But because P and N 2x0 EO transform in the same way (cf. Eq. (252) under the symmetry operations, their scalar product P . N will also be an invariant,
P . N = PXt2- Py5,
(253)
594
2
Ferroelectric Liquid Crystals
t2,
This corresponds to our previous PO and gives a term -c* (P,52-Py51) in the Landau expansion. In contrast to PO the form Py5,directly shows the chiral character. Thus, this invariant changes sign when we change from a right- to a left-handed reference frame, which means that the optical antipode of a certain C* compound will have a polarization of opposite sign. Further, we have the Lifshitz invariant
P, and P,. convenient expressions for C,, This can be done in several ways if we only correctly express the spatial relationship of 6 and P . For a positive material (P,>O), P is always 90 degrees ahead. If the director tilts out in the x direction, this will cause a polarization along y, symbolically expressed by
a5 51 a 5 2 - 52 A aZ
and similarly, with tilt along y
P,t2-
(254)
corresponding to Eq. (225) and, because both 5 and P behave like the c director, also one in P ,
5 = (0, o), P = (0, PI
(257)
5 = (0, el, P = (-P, 01
(258)
Generally, with cp = qz, and small tilts, = n,n, = cose sine coscp = 8 cosqz
t2= n,ny = cose sine sincp
(255)
corresponds to
Both have the same chiral character as the form in Eq. (253). Only one of these can be chosen, because they are linearly dependent, according to Eq. (252). This results in a Landau expansion
P, = -P sinqz P, = P cosqz
=:
8 sinqz (259)
Inserting these in Eq. (256) immediately gives G = -1a e 2 + -1b e 2 4
4
+ - L P 2 2x0 Eo
+ 1(P,’ + Py?)+ c* (P, 5 2 - Py 51) 2X
This is the same as Eq. (226) except that 1 K2,q2 is lacking. While minhere the term 2 imizing with respect to P gives the same equilibrium value as before, aG/aq = 0 gives
a*e2+ K33qe2= o (262) Since P - 6, all terms will be proportional to eP8-
In this expansion the last two terms describe the flexoelectric (previously eq PO) energy and the elastic energy due to the helix. The wave vector q in these terms is hidden in the deriatives ag,/az and a52/az. The expression Eq. (256) is the “classical” Landau expansion in the formulation of Pikin and Indenbom. We now have to insert
O2 and the @dependencesimply cancels out. This means that the wave vector, according to this model, will simply be q = A;/K,, equal to the low temperature value of Eq. (242). Thus the pitch Zis temperature independent, equal to Z,. This is of course in serious disagreement with experiment and the reason
2.5 The SmA*-SmC* Transition and the Helical C* State
for this deficiency is the failure to recognize the twist component of the helix. We can learn from this that a good order parameter will not by itself take care of all the physics. But our previous Landau expansion is likewise far from perfect, even if it gives a more reasonable behavior for the pitch. We should therefore stop for a moment to contemplate how capable the theory is, so far, to reproduce the experimental facts. As we already discussed, the pitch has a diverging behavior as we approach T, (this is a priori natural, as no helix is allowed in the A* phase) but seems to change its behavior (within about 1 K from the transition) to decrease to a finite value at the transition. Another important fact is that the ratio PI0 = s* experimentally is constant, as predicted by theory, up to about 1 K from the transition after which it decreases to a lower value [ 1331. A third fact, which we had no space to develop, is that the dielectric response, according to the Landau expansion used so far, should exhibit a cusp at the transition, which does not correspond to experimental facts. What is then the deficiency in the theory? Did we make an unjustified simplification when taking sin 8 = 8 in Eq. (259)? Probably not since, as the problem seems to lie in the very vicinity of T,, where 8 is small, it would not do any good to go to the more accurate expressions valid for large values of 19.Then how do we go further? Should we extend our expansion to include O6 beyond e4,or P4beyond P2? No, rather the solution ought to come from physical reasoning instead of formal procedures. This step was taken by iekS in 1984 [201]. The argument is as follows. Our Landau expansion so far has failed to take into account a quite important non-chiral contribution. At the tilting transition rotational bias appears, whether the material is chiral or not, which is quadrupolar in character, cf.
595
Fig. 47. But if there is a strong quadrupolar order it will certainly contribute to increasing the polar order, once we remove the reflection symmetry. This leads to a coupling term in P and 8 which is biquadratic, -P202. If we now only consider those terms in the free energy which contain P and q, we can write
1 17p4- a * q e 2 1 e2 +2 4 1 +-2 K33q282-v*qe4
--a p 2
(263)
1 The biquadratic coupling term is - 5L.2 (P,
t2
- Py=(n,,n,). Althoughljandcare parallel they are not identical: g=nzc. The 8 dependence of l j and c is different but for small tilts the difference between the vectors is not big. After having defined 5, the first thing we did in using it in Eq. (256) was in fact to go to the small tilt limit. Thus in practice we have not used anything else than
the c director as order parameter. Which means that we have in essence, all the time, in different shapes, used the two component complex order parameter from the beginning of Sec. 2.5, cf. Eq. (125) Y = 8eiq = 8eiqz= Bcosqz + isinqz (269)
As primary and secondary order parameters we have thus used two two-component vectors lying in the smectic plane, perpendicular to the layer normal. The number of degrees of freedom, i.e. the number of components, is evidently more important than subtle differences in symmetry, if the order parameters are used with sound physical arguments.
2.6 Electrooptics in the Surface-Stabilized State 2.6.1 The Linear Electrooptic Effect The basic geometry of a surface-stabilized ferroelectric liquid crystal is illustrated in Fig. 60. The director n (local average of the
-cl)
Figure 60. Local symmetry of the srnectic C* phase. The direction of P is shown for a positive material (P>O); by definition, one in which z, n, and P make a right-handed system. If electrodes are applied as shown, P will follow the direction of the electric field generated between them.
2.6 Electrooptics in the Surface-Stabilized State
molecular axes) may be identified with the optic axis (we disregard the small biaxiality) and is tilted at the angle 8 away from the layer normal (z). In the figure the director is momentarily tilted in the xz plane with a tilt angle 8,which is fixed for a fixed temperature. Its only freedom of motion is therefore described by the azimuthal variable cp indicated in the xy plane. If the phase is chiral, no reflections are symmetry operations and therefore a local polarization P is permitted in the (positive or negative) y direction. If electroded glass plates are mounted as in the figure, such that the layer normal z runs between them in the middle, we see that this most peculiar geometry with P perpendicular to the optic axis is ideal for a display: by applying the electric field across the thin slice, we can move the optic axis back and forth between two positions in the plane of the slice. This so-called in-plane switching cannot be achieved in a solid ferroelectric (nor in a nematic) without the use of a much more complicated electrode configuration. A typical cell structure is shown in Fig. 61. It illustrates the fact that we have two equivalent stable states; this is the reason for the symmetric bistability. The coupling is linear in the electric field and described by a torque
T=PxE
(270)
driving n (which is sterically coupled to P) both ways on reversal of the field direction and, as we will see below, leading to a microsecond response speed according to
dominating over any quadratic (dielectric) terms at low fields. In Fig. 62 this linear electrooptic effect is compared with the quadratic effect controlling the state of a twisted nematic device.
597
Figure 61. Originally assumed FLC cell structure. The field E is applied between the electroded glass plates e. The smectic layers are perpendicular to these glass plates and have a layer thickness d(T) of the order of 5 nm, which is temperature-dependent. The thickness of the smectic slab is 1-2 pm and equal to L, the distance between the plates. The shaded area illustrates the achievable switching angle of the optic axis. The symbols below illustrate that the polarization vector can be actively switched in both directions (1 -2). Strangely enough, this simple geometry, which corresponds to the original concepts of [93 -951, has only been realized in recent times, thanks to a devoted synthetic effort by chemists over the last 15 years. Before these materials are coming into use (not yet commercial) the layer structure is much more complicated. These problems are discussed in Sec. 2.8.
With a torque -E2 the latter effect is insensitive to the direction of E and is driven electrically only in the direction from twisted to untwisted state, whereas it has to relax back elastically in the reverse direction. The difference is also striking optically. The important feature of the in-plane switching is that there is very little azimuthal variation in brightness and color, because the optic axis is always reasonably parallel to the cell plane. If the optic axis pointed out of this plane (see Fig. 63), which it often does in the field-on state of a twisted nematic, a considerable variation with viewing angle will occur, including reversal of contrast when the display is viewed along angular direc-
598
2 Ferroelectric Liquid Crystals
TRANSMITTING STATE
NoN-TRANSMITIWGSTATE
ANALYZER
GLASS
TWISTED NEMATIC
1
GLASS
POLARIZER
ANALyzwl
GLASS
FLC
GLASS
tions nearly coinciding with the optic axis. The basic FLC optical effect is also illustrated in Fig. 64. In-plane switching has another advantage than the superior viewing angle, which may be equally important but has so far passed fairly unnoticed. This lies in its color neutrality. In the top part of Fig. 62, it is seen that when intermediate values of the electric field are applied to a twisted nematic in order to continuously vary the transmission, the tilt of the optic axis and thus the birefringence are changed at the same time. Thus the hue is influenced at the same time as the grey level in a twisted nematic, which makes the color rendition less perfect in a display where the colors are generated either by red - green -blue filter triads or by sequential color illumination. In-plane switching, on the other hand, in principle allows perfect separation of color and grey shades.
Figure 62. Comparison between twisted nematic and FLC switching geometry.
p"
'tl .-
Figure 63. The presentation of the first Canon monochrome A4 size panel in 1988 not only demonstrated the high resolution capability based on speed and bistability, it was also a definite breakthrough in LCD optics, demonstrating the essentially hemispherical viewing angle connected with in-plane switching.
Figure 64. The basic FLC optic effect is illustrated by the 1988 Canon monochrome prototype in which the front polarizer sheet has been removed; this also makes the drivers along the bottom visible. The optical contrast between the two polarization states is given by the small hand-held polarizer to the lower left.
2.6 Electrooptics in the Surface-Stabilized State
2.6.2 The Quadratic Torque When we apply an electric field across an FLC cell there will always be a dielectric torque acting on the director, in addition to the ferroelectric torque. This is of course the same torque as is present in all liquid crystals and in particular in nematics, but its effect will be slightly different here than it is in a nematic. This is because in the smectic C phase the tilt angle 8 is a “hard” variable, as earlier pointed out. It is rigid in the sense that it is hardly affected at all by an electric field. This is certainly true for a nonchiral smectic C and also for the chiral C*, except in the immediate vicinity of TC***,where we may not neglect the electroclinic effect. Hence we will consider the tilt angle 8 uninfluenced by the electric field, and this then only controls the phase variable cp of Fig. 60. In order to appreciate the difference, it is illustrative to start with the effect in a nematic. We will thus study the torque on the director due to an electric field E that is not collinear with n, as illustrated in Fig. 65. If we introduce the dielectric anisotropy
AE =
- EL
(272)
we may write the dielectric displacement D = ~0 EEcaused by the field in the form of the two contributions
(273) Subtracting E ~ Efrom ~ E the~first ~ and adding it to the second term gives
D = % A E E+ ~ ~~ E L E = % A E ( .~E ) n + EOELE
(274)
Figure 65.The dielectric torque will diminish the angle a between the director n and the electric field for A&>O.
599
This gives a contribution to the free energy (which is to be minimized for the case of a fixed applied voltage)
f,
- - E~ A & ( n .E )2 - 1 cO.cLE2 L
L
We discard the second term as it does not depend on the orientation of n and write the relevant free energy contribution (276) 1 1 G = - - EO A E ( .~E)2= - -- EO AE E2cos2a 2 2 where a i s the angle between n and E. With positive dielectric anisotropy, A&>0, we see that this energy is minimized when n is parallel toE. If AE 0. The dieelectric torque T Eis obtained by taking the functional derivative of G with respect to a
(277) which yields
r E= - - 1E ~ A E E ~sin2a 2
(278)
The viscous torque is always counteracting the motion, and can be written
(279) where yl is the so-called twist viscosity counteracting changes in a. In dynamic equilibrium the total torque is zero. From this we finally obtain the dynamic equation for the local director reorientation in an ex-
600
2
Ferroelectric Liquid Crystals
ternally applied field E -1- ~ ~ A ~ E ~ s i y1 n 2-d= aO + 2 dt
In the nematic case we would have, with
n . E = E cos a and n x E = - E sina (280)
Equation (280) describes the dielectric response to an electric field. It is easily integrated, especially if we look at the case of very small deviations of the director from the field direction (a),for which we have sin2a= 2 a and
In this case, a will approach zero simply as
rE= -A&%E~ sina cosa
(286)
where a i s the angle between n and E , which is equivalent to Eq. (278). In the smectic case the rigidity of the tilt angle 19prevents any other change than in the phase variable cp. Consequently, we have to take the z component of the torque working on cpand causing the conical motion of the director. From Fig. 51 and Eq. (140) we find, with the external field E applied in the y direction, i.e., E = (0, E, 0)
n . E = E sine sincp
(287)
and where a, is the initial angular deviation in Fig. 65 and zis given by
n x E = E(n,, 0, -n,)
z=-
from which Eq. (285) takes the shape
Yl
A E E ~E2
= E(cose,O, -sine coscp)
(288)
I"=A&q,E2(sinecos8,0, The characteristic response time is thus inversely proportional to the square of the applied field. What we have derived so far is valid for a nematic, where there are no particular restrictions to the molecular motion. In smectic media there will be such restrictions, and therefore Eqs. (278) and (280) will be modified in a decisive way. In any liquid crystal the induced polarization Pi caused by an applied electric field P, is obtained from D = E ~ E E = ~ , E +as P , = D - q , E or, from Eq. (274)
Pi = q,A&(n.E ) n + ~ , E -~q ,EE = ( E ~ -I ) q , E + %AE(IZ. E ) n
.E ) ( n x E )
(289)
The z component of this torque is thus
r,"= -A&%E~ sin2e sincp cosq
(290)
Compared to the nematic case we get a formally similar expression, except for the appearance of the sin28 factor, but the involved angle q has a completely different meaning than a. The azimuthal angle q is not the angle between the director and the field, but in the SmC* case the angle between the spontaneous polarization and the field, as seen when we imagine the director to move out in direction cp in Fig. 60.
(284)
The first term represents a component parallel to the electric field and does not therefore give any contribution to the torque P x E . From the second we get
r"= Pi x E = AEq,(n
-sin20 sincp coscp)
(285)
2.6.3 Switching Dynamics We are now able to write down the equation of motion for the director moving on the smectic cone and being subjected to
60 1
2.6 Electrooptics in the Surface-Stabilized State
ferroelectric and dielectric torques, as well as the elastically transmitted torques KV2q, which tend to make n spatially uniform. Disregarding the inertial term J@ (of negligible importance) and the elastic torques from the surfaces (a far more serious omission), we can write the dynamic director equation
’ud,
x
Figure 66. If the xz plane is parallel to the electrodes and the electric field is applied in the positive y direction, the ferroelectric torque will turn the tilt plane (given by cp) in the negative cp direction.
y -=-PEsinq acp at -AEEOE 2 sin2 8 sinq cosq+ K V2cp (291) 9
write Eq. (293) as
acp = -sinq z-
For E = 0 this reduces to
at
1 --sin20sinqcosq+{2V2q
(294)
X which has the form of the equation for heat conduction or diffusion and means that spatial gradients in q will damp out smoothly with a rate determining quantity Kly, corresponding to the heat conductivity or diffusion coefficient, respectively. With regard to the signs, we note that (see Fig. 60) q is the angle between the tilt plane (or c director) and x, and therefore also between P and E . The torque P x E = - P E sinq is consequently in the negative z direction for q > O (see Fig. 66). Equation (291) describes the overdamped approach towards equilibrium given by the field E, together with the material parameters P, 0, y, K , and AE. Let us divide all terms by PE
.sin2Qsinqcosq+-V K PE
z q (293)
yq Introducing a characteristic time ZE -, PE
a characteristic length 5 = mensionless quantity
x = EOAEE’ we can -~
Evidently we can write Eq. (294) in an invariant form using the variables t’ = tlz and l a i a r‘=r/{. For, with - = 7 and - = - _
a
at
at
a ax 6 ax”
etc., Eq. 294 takes the dimensionless shape
. coscp(r’, t’)
+ V2q(r’, t’)
(295)
The new variables are scaled with
&,r‘-.-PE
(296
and 7xY 2-
PE
(297
which set the basic length and time scales for the space-time behavior of cp(r, t). In addition, the behavior is governed by the dimensionless parameter
X=-
P EO AEE
The characteristic time z determines the dynamic response towards the equilibrium state when we apply an electric field. If we put in the reasonable values
602
2
Ferroelectric Liquid Crystals
y,=50 CP (=0.05 Ns/m2) (see Fig. 67), P = 10 nC/cm2, and E = 10 V/pm, we get a z value of about 5 ys. The characteristic length Sexpresses the balance between elastic and electric torques. An external field will align P along the field direction except in a thin layer of the order of 5, where a de-
J
viating boundary condition will be able to resist the torque of the field. As this layer thickness falls off according to 5 - E-"2, it will be much smaller than the sample thickness d for sufficiently high fields. In fact, it also has to be much smaller than the wavelength of light in order for the sample to be optically homogeneous. Only then will Eq. (297) also apply for the optical response time. It is easy to check that this is the case. With the P and E values already used and with K = 5 x N, we find from Eq. (296) that 5 is less than a nanometer. Finally, the parameter ( expresses the balance between the ferroelectric and the dielectric torque. As it appears in the dielectric term of Eq. (295) in the combination s i n 2 0 / ~ which , is = 0 . 1 5 / ~for 0=22", would have to be less than about 0.5 in order for us to have to keep this term. With the same P and E values as before, this requires a (A&(value as high as 22.5. We will therefore skip this term for the moment, but return to this question later.
x
2
2.6.4 The Scaling Law for the Cone Mode Viscosity
1o ~ / T ( K )
Figure 67. (a, b) An external electric field E perpendicular to the smectic layer normal is rotating the local polarization P as indicated in (a), forcing P to end up parallel to E . The corresponding reorientation of the optic axis takes place through the variable qaround a conical surface, i.e., the tilt cone, with no change in tilt 8. In contrast to a pure emotion (corresponding to viscosity y& or the equivalent twist viscosity y, in the nematic case), the q reorientation has a significant component around the long molecular axis, which leads to low effective viscosity. (c) Arrhenius plot of the viscosity y, in the nematic phase, as well as soft mode and Goldstone mode viscosities yo and y, in the smectic C* phase of a two-component mixture (from Pozhidayev et al. [148]).
Figure 67 gives an indication that, for the motion on the cone, the effective viscosity ought to depend on the tilt angle. Evidently, for small values of tilt it approaches the limiting case, where all the motion is around the long inertial axis, which must have the lowest possible viscosity. We also note from the experimental data of that figure that the cone mode viscosity, y, is even lower than the nematic viscosity at a higher temperature, which at first may seem puzzling. We will look at the viscosity in a more general way in Sec. 2.6.9, but here only try to illuminate this question. We will treat the viscosity in the simplest way possible, regarding it for a moment as a scalar.
603
2.6 Electrooptics in the Surface-Stabilized State
To a first approximation, any electric torque will excite a motion around the 8-preserving cone, i.e., around the z axis of Fig. 60. The essential component of the viscous torque will therefore be r,, which is related to dpldt (-@).We can thus write for the cone motion
2.6.5 Simple Solutions of the Director Equation of Motion
-rz = Y,@
y -+PEsinq=O a9
(299)
For a general rotational motion of the director n , the relation between the viscous torque and the viscosity has the form -r=ynxn
(300)
If we discregard both the dielectric and elastic terms in Eq. (291), the dynamics is described by the simple equation (305)
q at
For small deviations from the equilibrium state, s i n p z p and Eq. (305) is integrated directly to
p = pee-'"
(306)
Taking the z component we get
-r, = yln x lil, = yIc x C l
(30 1)
expressed by the two-dimensional c-director, which is the projection of n onto the plane of the smectic layer. Their relation means that c=n sin8 = sin8 (see Fig. 51). However Ic x i.1 = c2@
(302)
Hence
-rz= ysin28@
(303)
Comparing Eqs. (299) and (303) we obtain the important scaling law with respect to the tilt angle for the viscosity y, describing the motion around the cone (Goldstone mode)
y, = ysin 2 8
(304)
The index prefers to the p variable, and the index-free y corresponds to the nematic twist viscosity. Equation (304) means that the cone mode viscosity is lower than the standard nematic viscosity ( y,< It also seems to indicate that the viscosity y, tends (8-0). to zero at the transition T-T,, This, of course, cannot be strictly true, but we will wait till Sec. 2.6.9 to see what actually happens.
where z= YV and cpo is the angle between PE E and P at time t= 0. The response time z is inversely proportional to the field, instead of being inversely proportional to the square of the field, as in the dielectric case. Generally speaking, this means that the FLC electrooptic effect will be faster than normal dielectric effects at low voltages, and also that these dielectric effects, at least in principle, will ultimately become faster when we go to very high applied fields. The unique advantage of FLC switching is, however, that it is not only very fast at moderate fields, but that it switches equally fast in both directions; a feature that cannot be achieved if the coupling to the field is quadratic. If 9 is not small in Eq. (305), we change the variable to half the angle (sin9 = 2 sin 9 cos -) 9 and easily find the analytical 2 2 solution ~
n).
(307) with the same z=
yq ~
PE ly be expressed as
. This can alternative-
604
2
Ferroelectric Liquid Crystals
2.6.6 Electrooptic Measurements
is directly obtained as
Equations (307) and (308) are a solution to Eq. (305) if the electric field is constant with time. The expression is therefore also useful when the field is a square wave. Such a field is commonly used for polarization reversal measurements when switching between +P, and -P, states. In order to switch from -Ps to + P , we have to supply the charge 2 P, to every unit area, i.e., 2 P,A over the sample, if its area is A. The reversal of the vector P, thus gives rise to an electric current pulse i(t)= dQ through the curcuit in dt which the sample cell is connected. This current is given by the time derivative of the charge due to P,on the electrodes of area A. When P, makes an angle cp with E , the polarization charge on the electrodes is Q = P,A cos cp. The electric current contribution due to the reversal of P, is thus given by (309)
When cp passes 90 O during the polarization reversal, i(t) in Eq. (310) goes through a maximum. The shape of the function i(t) is that of a peak. This is shown in Fig. 68 where Eqs. (308) and (3 10) have been plotted for the case cpo= 179 '. By measuring the area under the polarization reversal peak, we obtain the total charge transferred and can thus determine the value of the spontaneous polarization P,. The time for the appearance of the current peak then allows a determination of the viscosity y,. The curves in Fig. 68 have been traced for the case where P makes an angle of 1 O relative to the surface layer normal, corresponding to the same pretilt of the director out of the surface. This pretilt also has to be included as a parameter determining the electrooptic properties. On the other hand, the integration of Eq. (305) giving the current peak of Fig. 68 can only give a crude estimation of the electrooptic parameters, because the equation does not contain any term describing the surface elastic torques. The value of the method lies therefore in the
~
i(t)= P,A-[coscp(t)] d = -P,A-sincp(t) dcp dt dt Substituting the time derivative of cp from Eq. (305), the polarization reversal current
-1.0 - 0.8
> -
1°.6 - 0.4 50-
i0.2 I
0
2
4
6
tl
8
10
,
,
,
0.0
12
Figure 68. The polarization reversal q(r) according to Eq. (304) for a half-period of an applied square wave and the resulting current pulse i(t). The current peak appears roughly at t=5zand the reversal is complete after twice that time. The latter part of the cp curve corresponds to the simple exponential decay of Eq. (32). The optical transmission is directly related to Ht). If the state at t=O is one of extinction, the transmission will increase to its maximum value, but in a steeper way than ( d q l d t l (from Hermann 11361).
2.6
Electrooptics in the Surface-Stabilized State
fact that it is simple and rapid. The same can be said about the alternative formalism of [ 1371, which has proven useful for routine estimations of electrooptic parameters, at least for comparative purposes. In [137] a hypothetical ad hoc surface torque of the form Kcoscp has been added to the ferroelectric torque in Eq. (305). With only these two torques present, the equation can still be easily integrated in a closed form. However, the chosen form of the elastic torque means that it cannot describe the properties of a bistable cell. Numerous other approaches have been proposed in order to describe the switching dynamics, for instance by neglecting the surface elastic torques, but including the dielectric torque using the correct representation of the dielectric tensor [138]. Common to most of them is the fact that they describe the shape of the polarization reversal current and the electrooptic transmission curve reasonably well. They are therefore easy to fit to experimental data, but may give values to the electrooptic parameters that may deviate within an order of magnitude. (An important exception to this is the polarization value P,, which is essentially insensitive to the model used because it is achieved by integration of the current peak.) Hence no model put forward so far accounts generally for the switching dynamics, and there is a high degree of arbitrariness in the assumptions behind all of them. Experiments in which the parameters in the equation of motion are measured may be carried out with an electrooptic experimental setup, usually based on a polarizing microscope. In the setup, the sample is placed between crossed polarizers and an electric field E(t) is applied across the cell. The optical and electric responses of the sample to the field are then measured. The light intensities Z ( t ) through the crossed polarizers and the sample are mea-
605
sured by a photodiode after the analyzer. The electric current i(t)through the sample is obtained by measuring the voltage u ( t ) across a resistance R connected in series with the liquid crystal cell (see Fig. 69). All three signals E ( t ) , Z ( t ) , and i ( t ) are fed into an oscilloscope so that the switching dynamics of the liquid crystal cell can be monitored. The oscilloscope is connected to a computer so that the waveforms can be stored and analyzed in detail. From the optical transmission, the tilt angle and the optical response time can be measured. From the electric response the spontaneous (and induced) polarization can be measured, as well as the response time, as discussed earlier. If we also include the SmA* phase (the electroclinic effect), the coefficients of the Landau expansion of the free energy density may be derived. Estimations of the viscosities for the cp- and &motions may also be obtained from the electrooptic measurements. In Fig. 69 we see how the quasi-bookshelf (QBS) smectic constitutes a retarder with a switchable optic axis. In the smectic A* phase this axis can be switched continously, but with small amplitude. The same setup is also used to study the DHM effect in short pitch chiral smectics and the flexoelectrooptic effect in the cholesteric phase, in both of which we achieve much higher tilts. Inserted between crossed polarizers, the retarder transmits the intensity
z = z0 sin2 2~
sin
TcdAn
2 ~
A
where Y is the angle between the polarizer direction and the projection of the optic axis in the sample plane, An is the birefringence, ilis the wavelength of light in a vacuum, and d is the thickness of the sample. The function Z ( Y ) is shown in Fig. 70. A very important case is the lambda-half condition given by d A n = 112. A cell with this
606
2
Ferroelectric Liquid Crystals OPTICAL
DETECTOR I
i c ! -- - - - Optimum optical setting: p22.5'
i(t)=u(t)/li
Figure 69. (a) The experimental setup for electrooptic measurements. The sample is placed between crossed polarizers. (b) The RC-circuit used in the switching experiments. The current response from the sample is measured across the resistor R (from Hermann [ 1361).
1.0
0.4 CI W
0.2
0.0 0
22,5
45 (deg)
67,5
90
Figure 70. Transmitted light intensity I ( Y )for a linear optical retarder with switchable optic axis placed between crossed polarizers. Y is the angle between the polarizer direction and the switchable optic axis, and maximum transmitted light intensity is achieved for Y=45" combined with a sample thickness fulfilling the A12 condition. A smectic C* in QBS geometry and tilt angle 8=22.5 could ideally be switched from Y=O to Y=45 O in a discontinuous way. DHM and flexoelectrooptic materials could be switched in a continuous way. Z(Y)is proportional to Y in a region around Y =22.5 ', which is the linear regime (from Rudquist [ 1391).
value of d A n will permit maximum transmission and flip the plane of polarization by an angle 2 Y. Since the position of the optic axis, i.e., the angle Y', may be altered by means of the electric field, we have electrooptic modulation. As seen from Eq. (311) and from Fig. 70, the minimum and maxi-
mum transmission is achieved for Y=O and Y'=45 O , respectively. If we adjust the polarizer to correspond to the first state for one sign of E across an SSFLC sample, reversal of sign would switch the polarization to give the second state, if 28=45". For natural grey-scale materials (electroclinic, DHM,
2.6
Electrooptics in the Surface-Stabilized State
flexoelectrooptic, antiferroelectric), the electrooptic response to a field-controlled optic axis cell between crossed polarizers is modelled in Fig. 71. Here Y= Yo+@ ( E ) , where Yois the position of the optic axis for E=O and @ ( E ) is the field-induced tilt of the optic axis. The tilt angle in the SmC* phase can be measured by the application of a square wave of very low frequency (0.2- 1 Hz), so that the switching can be observed directly in the microscope. By turning the angular turntable of the microscope, the sample is set in such an angular position that one of the extreme tilt angle positions, say -$, coincides with the polarizer (or analyzer), giving Z=O. When the liquid crystal is then
607
switched to +Oo, I becomes #O. The sample is now rotated so as to make I= 0 again, corresponding to +Oo coinciding with the polarizer (or analyzer). It follows that the sample has been rotated through 2 Oo, whence the tilt angle is immediately obt ained . The induced tilt angle in the SmA* phase is usually small, and the above method becomes difficult to implement. Instead, the tilt can be obtained [ 1401 by measuring the peak-to-peak value AZpp in the optical response, from the relation 1 arc sin MPP Oind = ~
4
10
which is valid for small tilt angles.
Electrooptic signal / Transmitted light intensity
k
III
I
/ I
'
I
1
LA-! '
I I
I
I
vt
Applied voltage / Input signal, giving
II
Y = Yo+ @ ( E )
m
i I
Figure 71. Electrooptic modulation due to the field-controlled position of the optic axis between crossed polarizers. I) The optic axis is swinging around Yo=Ogiving an optical signal with double frequency compared to the input signal. 11) The linear regime. When the optic axis is swinging around Y0=22.5 the variation of the transmitted intensity is proportional to @. For a liquid crystal A / 2 cell with quasi-bookshelf (QBS) structure, it follows from Eq. (311) that the intensity I varies linearly with @ if b, is small (1/1,=1/2+2@(E)), which, for instance, is the case if the electroclinic effect in the SmA* phase is exploited. As electroclinic tilt angles are still quite small, certain other materials, though slower, have to be used for high modulation depths. 111) Characteristic response for @ ( E )exceeding 222.5 ". (from Rudquist [139]). O,
608
2
Ferroelectric Liquid Crystals
On applying a square wave electric field in the SmC* phase the electric response has the shape of an exponential decay, on which the current peak i(t) (see Fig. 68) is superposed. The exponential decay comes from the capacitive discharge of the liquid crystall cell itself, since it constitutes a parallelplate capacitor with a dielectric medium in between. The electrooptic response i(t) on application of a triangular wave electric field to a bookshelf-oriented liquid crystal sample in the SmC* and SmA* phases is shown as it appears on the oscilloscope screen in Fig. 72, for the case of crossed polarizers and the angular setting 'Yo= 22.5". Applying such a triangular wave in the SmC* phase gives aresponse with the shape of a square wave on which the polarization reversal current peak is superposed. The square wave background is due to the liquid crystal cell being a capacitor; an RC-circuit as shown in Fig. 69b delivers the time derivative of the input. The time derivative of a triangular wave is a square wave, since a triangular wave consists of a constant slope which periodically changes sign. The optical response Z ( t ) also almost exhibits a square wave shape when applying a triangular wave, due to the threshold of the two stable states of up and down polarization corresponding to the two extreme angular positions -8 and +8 of the optic axis (the
director) in the plane of the cell. The change in the optical transmission coincides in time with the polarization reversal current peak, which occurs just after the electric field changes sign. In the SmC* phase, the optical transmission change in the course of the switching is influenced by the fact that the molecules move out of the plane of the glass plates, since they move on a cone. They thus make some angle C(t) with the plane of the glass plates at time t. The refractive index seen by the extraordinary ray n , ( n depends on this angle. The birefringence thus changes during switching according to A n [ C(t)]= n, [ [ ( t ) ]-no. After completion of the switching, the molecules are again in the plane of the glass plates and {=O, so that the birefringence is the same for the two extreme positions -8 and +8 of the director. When raising the temperature to that of the SmA* phase, the polarization reversal current peak gradually diminishes. In the SmA* phase near the transition to SmC*, a current peak can still be observed, originating from the induced polarization due to the electroclinic effect. On application of a triangular wave, the change of the optical response can most clearly be observed: it changes from a squarewave-like curve to a triangular curve, which is in phase with the applied electric field. This very clearly illustrates the change from the nonlinear cone switching of the SmC* phase to the linear switching of,the SmA* phase due to the electroclinic effect.
2.6.7 Optical Anisotropy and Biaxiality (a)
SmC* phase
(b) SmA* phase
Figure 72. The electrooptic response when applying a triangular wave over the cell in the SmC* and the SmA* phases. The ohmic contribution due to ionic conduction has been disregarded.
The smectic A phase is optically positive uniaxial. The director and the optic axis are in the direction of highest refractive index, n11-n21=An>0. At the tilting transition
2.6 Electrooptics in the Surface-Stabilized State
SmA +SmC, the phase becomes positive biaxial, with rill -+n3 and nl splitting in two, i.e., nl+nl, n,. The E ellipsoid is prolate and n3 > n2 > a1
(313)
or A n > 0 , dn>O
(3 14)
where we use, with some hesitation, the abbreviations An = n3 - n1
(315)
an = n2 - n1
(316)
The hesitation is due to the fact that the refractive index is not a tensor property and is only used for the (very important) representation of the index ellipsoid called the optical indicatrix, which is related to the dielectric tensor by the connections E,
= n 2l ,
2 2 = n2, E~ = n3
(3 17)
between the principal axis values of the refractive index and the principal components of the dielectric tensor. For orthorhombic, monoclinic, and triclinic symmetry the indicatrix is a triaxial ellipsoid. The orthorhombic system has three orthogonal twofold rotation axes. This means that the indicatrix, representing the optical properties, must have the same symmetry axes (Neumann principle). Therefore the three principal axes of the indicatrix coincide with the three crystallographic axes and are fixed in space, whatever the wavelength. This is not so for the monoclinic symmetry represented by the smectic C . Because the symmetry element of the structure must always be present in the property (again the Neumann principle), the crystallographic C, axis perpendicular to the tilt plane is now the twofold axis of the indicatrix and the E tensor, but no other axes are fixed. This means that there is ambiguity in
609
the direction of along the director, (which means ambiguity in the director as a concept) as well as in the direction E, perpen) the tilt dicular to both the “director” ( E ~ and axis (E,). Only the latter is fixed in space, and therefore only &2 can be regarded as fundamental in our choice of the principal ax. es of the dielectric tensor ( E ~ &,, E ~ ) Now, it is well known that the triaxial character, i.e., the difference between and (or n2 and n l ) ,&=E,-E,, is very small at optical frequencies. The quantity &is now universally called the biaxiality, although this name would have been more appropriate for the degree of splitting of the two optic axes as a result of the tilt. Thus optically speaking we may still roughly consider the smectic C phase as uniaxial, with an optic axis (director) tilting out a certain angle 8from the layer normal. However, as has recently been pointed out by Giesselmann et al. [141], due to the ambiguity of the .cl and E~ directions, the optical tilt angle must be expected to be a function of the wavelength of the probing light. As they were able to measure, the blue follows the tilt of the core, whereas the red has a mixture of core and tail contributions. The optical tilt is thus higher for blue than for red light, and the extinction position depends on the color. While this phenomenon of dispersion in the optic axis is not large enough to create problems in display applications, it is highly interesting in itself and may be complemented by the following general observation regarding dispersion. At low frequencies, for instance, lo5 Hz, the dielectric anisotropy A&= E~ - is often negative, corresponding to a negatively biaxial material. As it is positively biaxial at optical frequencies (- 1015Hz), the optical indicatrix must change shape from prolate to oblate in the frequency region in between, which means, in particular, that at some frequency the E tensor must become isotropic.
610
2
Ferroelectric Liquid Crystals
2.6.8 The Effects of Dielectric Biaxiality At lower frequencies we can no longer treat the smectic C phase as uniaxial. Therefore our treatment of dielectric effects in Sec. 2.6.2 was an oversimplification, valid only at low values of tilt. In general, we now have to distinguish between the anisotropy, A&= E~ - E, , as well as the biaxiality, = E* - E , , both of which can attain important nonzero values typically down to -2 for A& and up to +3 or even higher for The latter parameter has acquired special importance in recent years, due to a special addressing method for FLC displays, which combines the effects of ferroelectric and dielectric torques. As for the background, it was successively discovered in a series of investigations by the Boulder group between 1984 and 1987 that the smectic layers are generally tilted and, moreover, form a so-called chevron structure, according to Fig. 73, rather than a bookshelf structure. The reason for the chevron formation is the effort made by the smectic to fill up the space given by the cell, in spite of the layer thickness shrinking as the temperature is lowered through the SmA - SmC transition and further down into the SmC phase. The chevron geometry (to which we will return at length in Sec. 2.8) reduces the effective switching angle, which no longer corresponds to the optimum of 28= 45 O , thereby reducing the brightness-contrast ratio (we have in the figure assumed that the memorized states have zero pretilt, i.e., that they are lying in the surface). This can, to some extent, be remedied by a different surface coating requiring a high pretilt. Another way of ameliorating the optical properties is to take advantage not only of the ferroelectric torque (-E) exerted on the molecules, but also of the dielectric torque (-E2). This torque always tries to turn the
a&.
highest value of the permittivity along the direction of the electric field. The tilted smectic has the three principal E values in the directions shown in the figure, E , along the chosen tilt direction, .s2along the local polarization, and &3 along the director. If is large and positive, it is seen that the dielectric torque exerted by the field will actually lift up the director away from the surface-stabilized state along the cone surface to an optically more favorable state, thus increasing the effective switching angle. The necessary electric field for this action can be provided by the data pulses acting continuously as AC signals on the columns. This addressing method thus uses the ferroelectric torque in the switching pulse to force the director from one side to the other between the surface-stabilized states (this could not be done by the dielectric torque, because it is insensitive to the sign of E ) , after which high frequency AC pulses will keep the director dynamically in the corresponding extreme cone state. This AC enhanced contrast mode can ameliorate the achievable contrast in cases where the memorized director positions give an insufficient switching angle, which is the case in the socalled C2 chevron structures (see Sec. 8.8). It requires specially engineered materials with a high value of Its main drawback is the requirement of a relatively high voltage (to increase the dielectric torque relative to the ferroelectric torque), which also increases the power consumption of the device. The AC contrast enhancement is often called “AC stabilization” or “HF (high frequency) stabilization”. A possible inconvenience of this usage is that it may lead to a misunderstanding and confusion with surface-stabilization, which it does not replace. The dielectric contrast enhancement works on a surface-stabilized structure. There is no conflict in the concepts; the mechanisms rather work together.
a&.
2.6 Electrooptics in the Surface-Stabilized State
61 1
Figure 73.The simple bookshelf structure with essentially zero pretilt would lead to ideal optical conditions for materials with 2 0equal to 45". In reality the smectic layers adopt a much less favorable chevron structure. This decreases the effective switching angle and leads to memorized P states that are not in the direction of the field (a, b). A convenient multiplexing waveform scheme together with a properly chosen value of the material's biaxiality d~ may enhance the field-on contrast relative to the memorized (surface-stabilized) value (c), by utilizing the dielectric torque from the pulses continuously applied on the columns (after [142, 1431).
In Eq. (291) we derived an expression for the director equation of motion with dielectric and ferroelectric torques included. If the origin of the dielectric torque is in the dielectric biaxiality, the equation (with the elastic term skipped)will be the closely analogousone a9 2 y -=-PEsincp-&oa&E sinqcosq at (3 18)
In the present case, where the layers are tilted by the angle Grelative to E, as in Fig. 73a, the effective applied field along the layer will be Ecos 6, thus slightly modifying the equation to
y -=-PEcosGsinq acp 9 at
- E~
a&E2cos2 6 sin cp cos q (3 19)
612
2
Ferroelectric Liquid Crystals
In addition to the characteristic time rand length 5, we had earlier introduced the dimensionless parameter describing the balance between ferroelectric and dielectric torques. Its character appears even more clearly than in Eq. (298) if we write it as
x
nc rc
= &g a&E2
If we again replace E by E cos 6, our new will be
x
P = &o a&Ecos6
FromX and z=
yq we can form a charPE cos 6
acteristic time
x
and from and Ed (cos 6 factors canceling in Ed), a characteristic voltage
The values of the parameters z, and V, are decisive in situations where the ferroelectric and dielectric torques are of the same order of magnitude, and they play an important role in the electrooptics of FLC displays addressed in a family of modes having the characteristics that there is a minimum in the switching time at a certain voltage (see Fig. 74a), because the ferroelectric and dielectric torques are nearly balancing. This means that the switching stops both on reducing the voltage (due to insufficient ferroelectric torque) and on increasing it (due to the rapidly increasing dielectric torque, which blocks the motion, i.e., it increases the delay time for the switching to take off). The existence of a minimum in the switching time when dielectric torques become important was discovered by Xue et al. [138] in 1987, but seems to have appeared in novel addressing schemes some years later in the British national FLC collaboration within the JOERS/Alvey program. It turns out [ 1451 that the minimum switching time
log v Vmin
.
.
lot
10
1
20
30
40
i .
. :
50
60
70 80 90 100
Pulse voltage, V, / V
Figure 74. (a) Typical pulse switching characteristic for a material with parameters giving ferroelectric and dielectric torques of comparable size; (b) Possible routes for engineering materials to minimize both zmi, and Vmin(from [144]).
2.6 Electrooptics in the Surface-Stabilized State
613
a&,
depends on y,, and P according to Eq. (322), and that this minimum occurs for a voltage that depends on and P according to Eq. (323). At present, considerable effort is given to reduce both zo and V,. Figure 74b indicates how this could be done: either (1) by lowering the viscosity y, whilst or (2) by inincreasing the biaxiality creasing both & and P by a relatively large amount.
a&
a&,
2.6.9 The Viscosity of the Rotational Modes in the Smectic C Phase In Sec. 2.6.4 we derived the scaling law for the cone mode viscosity with respect to the tilt angle 8. In this section we want to penetrate a little deeper into the understanding of the viscosities relevant to the electrooptic switching dynamics. Let us therefore first review the previous result from a new perspective. With 8 = const, an electric torque will induce a cone motion around the z axis (see Fig. 75a). We can describe this motion in different ways. If we choose to use the angular velocity ci, of the c director, that is, with respect to the z axis, then we have to relate it to the torque component
rl
-c= Y, ci,
(324)
In Fig. 75b (y, is denoted as q ) we have instead illustrated the general relation, according to the definition of viscosity, between the angular velocity & of the director n and the counteracting viscous torque rv
-rv = y&
(325)
As a is the coordinate describing the motion of the head of the n arrow (on a unit sphere), a = n x i , and the relation can be written
-r"=ynxli
(326)
Figure 75. (a) The variable cpdescribes the motion of the c director around the layer normal. (b) The relation between the rate of change of the director i ,the angular velocity CU, and the (always counteracting) viscous torque r.
We have used this before in Eq. (300). Several things should now be pointed out. The difference between y, and y is, in principle, one of pure geometry. Nevertheless, it is not an artifact and the scaling law of y, with respect to 8 has a real significance. In a nematic we cannot physically distinguish between the 8 motion and the cp motion on the unit sphere. Therefore yin Eq. (325) is the nematic viscosity. In a smectic C, on the other hand, we have to distinguish between these motions, because the director is moving under the constraint of constant 8. As is evident fromFig. 75b, however, this motion is not only counteracted by a torque Tz;instead r must have nonzero x and y components as well. We might also add that in general we have to distinguish between the 8 and cp motions as soon as we go to a smectic phase, either SmA or SmC. For instance, we have a soft mode viscosity y,in these phases, which has no counterpart in a nematic.
614
2
Ferroelectric Liquid Crystals
Let us now continue from Eq. (326). With sin 0 cos cp (327) we have
One of the drastic oversimplifications made so far has been in treating the viscosity as a scalar, whereas in reality it is a tensor of rank 4. In the hydrodynamics, the stresses oGare components of a second rank tensor, related to the velocity strains vk,/ 0.. = q-avk
- sin 0 sin cp@
z~
hence
(329) -sin0 cos2 cp -rv=y @ -sinOsincpcoscp sin2 0 cos2 cp + sin2 0 sin2 a,
I:
Therefore the z component is
-r; = y @sin20
1
(330)
A comparison of Eq. (330) with Eq. (324) then gives 2
y, = y sin 0
(331)
which is the same as Eq. (304) ( y is ' a nematic viscosity). We may note at this point that Eq. (331) is invariant under O + O+n, as it has to be (corresponding to n +-n). We have still made no progress beyond Eq. (304) in the sense that we have stuck with the unphysical result y, +0 for 0 +0. The nonzero torque components r, and ry in Eq. (329) mean that the cone motion cannot take place without torques exerting a tilting action on the director. With 0 constant, these have to be taken up by the layers and in turn counteracted by external torques to the sample. We also note that since O=const in Eq. (327), we could just as well have worked with the two-dimensional c director from the beginning c = (c,, cy)=
sin 0 cos cp sin 0 sin cp
to obtain the result of Eq. (329).
(332)
(333)
ax,
in which q is thus a tensor of rank 4 just like the elastic constants Cijkl for solid materials, for which Hooke's law is written (334)
o i.j . = c rjkl . . &kl
As is well known, however, the number 34= 8 1 of the possibly independent components is here reduced by symmetry, according to (335) giving only 36 independent components. They therefore permit a reduced representation using only two indices Cijkl
=cp,
0 p
= cp" &"
(336)
which has the advantage that we can write them down in a two-dimensional array on paper, but has the disadvantage that cpv no longer transforms like a tensor. The same is valid for the Oseen-Frank elastic constants and the flexoelectric coefficients in liquid crystals, which are also tensors of rank 4, but are always written in a reduced representation, Kq and e,, respectively, where K , and eijcannot be treated as tensors. The viscosity tensor in liquid crystals is a particular example of a fourth rank tensor. The viscous torque r i s supposed to be linear in the time derivative of the director and in the velocity gradients, with the viscosities as proportionality constants. This gives five independent viscosities in uniaxial ( 1 2 in biaxial) nematics, the same number in smectics A, but 20 independent components in the smectic C phase [146]. With nine inde-
2.6 Electrooptics in the Surface-Stabilized State
pendent elastic constants and as many flexoelectric coefficients, this phase is certainly extraordinary in its complexity. Nevertheless, the viscous torque can be divided in the rotational torque and the shearing torque due to macroscopic flow. When we study the electrooptic switching of SSFLC structures, we deal with pure rotations for which macroscopic flow is thus assumed to be absent. Even if this is only an approximation, because the cone motion leads to velocity gradients and backflow, it will give us an important and most valuable description because of the fact that the rotational viscosity has a uniquely simple tensor representation. If we go back to the Eq. (325) defining the viscosity, both the torque and the angular velocity are represented by axial vectors or pseudovectors. This is because they actually have no directional property at all (as polar vectors have), but are instead connected to a surface in space within which a rotation can only be related to a (perpendicular) direction by a mere convention like the direction for the advance of a screw, “righthand rule”, etc. In fact they are tensors of second rank which are antisymmetric, which means that
r.. V = -r.. J‘ and
&.IJ = -&..J‘
(337)
with the consequence that they have (in the case of three-dimensional space) only three independent components and therefore can be written as vectors. As yin Eq. (325) connects two second rank tensors, it is itself a fourth rank tensor like other viscosities, but due to the vector representation of T and a, the rotational viscosity can be given the very simple representation of a second rank tensor. Thus in cases where we exclude translational motions, we can write yas
(338)
615
Unfortunately, no similar simple representation exists for the tensor components related to macroscopic flow. We will now finally treat ynot as a scalar, but as a tensor in this most simple way. In Eq. (338) y is already written in the “molecular” frame of reference in which it is diagonalized, as illustrated in Fig. 76, with the principal axes 1,2, and 3. Because the C, axis perpendicular to the tilt plane has to appear in the property (Neumann principle - we use it here even for a dynamic parameter), this has to be the direction for y*. As for the other directions, there are no compelling arguments, but a natural choice for a second principal axis is along the director. We take this as the 3 direction. The remaining axis 1 is then in the tilt plane, perpendicular to n. The 3 axis is special because, along n,the rotation is supposed to be characterized by a very low viscosity. In other words, we assume that the eigenvalue y3 is very small, i.e., fi+y,, E. Starting from the lab frame x, y, z, y is diagonalized by a similarity transformation
f =T - ’ ~ T
(339)
where T is the rotation matrix
uj.,
cos8 0 -sin8 1 0 sin8 0 case
j
(340)
G Figure 76. The principal axes 1, 2 , and 3 of the viscosity tensor. In the right part of the figure is illustrated how the distribution function around the director by necessity becomes biaxial, as soon as we have a nonzero tilt 8.
616
2
Ferroelectric Liquid Crystals
whereas for 8= z/2 we would find (formally) that
corresponding to the clockwise tilt 8 (in the coordinate transformation the rotation angle 8 is therefore counted as negative) around the y (2) axis, and T-' is the inverse or reciprocal matrix to T; in this case (because rotation matrices are orthogonal) T is simply equal to its transpose matrix T (i.e., = TV).Hence we obtain the viscosity components in the lab frame as
Yq (">=y, 2
corresponding to the rotational motion in a two-dimensional nematic. There is another reason than Eq. (347) not be neglect E : as illustrated to the right of Fig. 76, a full swing of the director around the z axis in fact involves a simultaneous full rotation around
zj
y =TFT-~
(348)
(34 1)
which gives cos8 0 -sin8 y=[o 1 0 sin8 0 cos8
0 sin8
0
0
y3
-sin8
o case
1
y, cos2 8 + y3 sin2 8
o
(yl - y3)sinocos8
Y2
(yl -y3)sin8cos6
0
0 ylsin28+y3cos28
Assuming now the motion to be with fixed 8, the angular velocity a is
a= (O,O, ql)
(343)
and Eq. (325) takes the form
(344)
y1 sin$cosB@
- P = y[;]=[o
(yl sin2 8 + y3 cos28 ) @
the z component of which is
-r; = (nsin28 + Y,COS%)@
1
(345)
Comparison with Eq. (324) now gives y,(~) = yl sin28+ f i cOs%
(346)
This relation illustrates first of all that the cp motion is a rotation that occurs simultaneously around the 1 axis and the 3 axis, and it reflects how the rotational velocities add vectorially, such that when 8 gets smaller and smaller the effective rotation takes place increasingly around the long axis of the molecule. For 8=0 we have Y,(O) = 73
(347)
(342)
the 3 axis. Hence these motions are actually intrinsically coupled, and conceptually belong together. If, on the other hand, we had neglected the small f i in Eq. (342), the viscosity tensor would have taken the form
y1cos28
y=[o y1sinecose
o
y1sin8cosB
Y2
0 y1sin28
o
(349)
which would have given the same incomplete equation (304) for y, as before. The first experimental determination of the cone mode viscosity was made by Kuczyhski [147], based on a simple and elegant analysis of the director response to an AC field. Pyroelectric methods have been used by the Russian school [136, 1481, for instance, in the measurements illustrated in Fig. 67. Later measurements were normally performed [ 1491 by standard electrooptic methods, as outlined earlier in this chap-
2.7
Dielectric Spectroscopy: To Find the ?and ETensor Components
ter. The methodology is well described by Escher et al. [l50], who also, like Carlsson and kekS [ 1511, give a valuable contribution to the discussion about the nature of the viscosity. The measured viscosities are not the principal components yl, y2, y3 in the molecular frame, but y, and y2, which we can refer to the lab frame, as well as y3, for which we discuss the measurement methodin Sec. 2.7. As y2 refers to a motion described by the tilt angle 8, we will from now on denote it y , and often call it the soft mode viscosity (sometimes therefore denoted It corresponds more or less, like yl, to a nematic viscosity (see Fig. 67 for experimental support of this fact). The cone mode viscosity y,, which is the smallest viscosity for any electrooptically active mode, will correspondingly often be called the Goldstone mode viscosity (sometimes denoted yG).As yl and y, must be of the same order of magnitude, Eq. (304) makes us expect y," ye. The difference is often an order of magnitude. The value y3, finally, corresponding to the rotation about the long molecular axis, is normally orders of magnitude smaller than the other ones. The value should be expected to be essentially the same as in a nematic, where this viscosity is mostly denoted xl or yl. We will here prefer the latter, thus in the following we will write yl instead of y3.The value for twist motion in a nematic is mostly called yl but also % or, as opposed to xl, yL. As y, and y2 cannot be distinguished in the nematic, the twist viscosity corresponds both to yl and y2 (yo) in the smectic C. The meanings of yo and yl are illustrated in Fig. 77. If we symbolize the director by a solid object, as we have to the right of Fig. 76, we see that the coordinates 8, 9,and y,the latter describing the rotation around the long are three inaxis, thus corresponding to dependent angular coordinates which are re-
x).
x,
617
Figure 77.Rotations that correspond to viscosities yo and yl, as drawn for the uniaxial SmA (SmA*) phase.
quired for specifying the orientation of a rigid body. Except for the sign conventions (which are unimportant here), they thus correspond to the three Eulerian angles. More specifically, 9 is the precession angle, 8 the nutation angle, and y is the angle of eigenrotation. The three viscosities y,, yo, and y, determine the rotational dynamics of the liquid crystal. Only y, can easily be determined by a standard electrooptic measurement. All of them can, however, be determined by dielectric spectroscopy. We will turn to their measurement in Sec. 2.7.
2.7 Dielectric Spectroscopy: To Find the and 6 Tensor Components 2.7.1 Viscosities of Rotational Modes Using dielectric relaxation spectroscopy, it is possible to determine the values of the rotational viscosity tensor corresponding to the three Euler angles in the chiral smectic C and A phases. These viscosity coefficients (ye, y, yl) are active in the tilt fluctuations (the soft mode), the phase fluctuations (the Goldstone mode), and the molecular re-
618
2
Ferroelectric Liquid Crystals
orientation around the long axis of the molecules, respectively. It is found, as expected, that these coefficients obey the inequality ye> y,+ yl. While the temperature dependence of and y, in the SmC* phase can be described by a normal Arrhenius law as long as the tilt variation is small, the temperature dependence of yo in the SmA* phase can only be modeled by an empirical relation of the form ye=AeEJkT+ $(T-T,)-". At the smectic A* to SmC* transition, the soft mode viscosity in the smectic A* phase is found to be larger than the Goldstone mode viscosity in the SmC* phase by one order of magnitude. The corresponding activation energies can also be determined from the temperature dependence of the three viscosity coefficients in the smectic A* and SmC* phases. The viscosity coefficient connected with the molecular reorientation around the long axis of the molecules is not involved in electrooptic effects, but is of significant interest for our understanding of the delicate relation between the molecular structure and molecular dynamics.
2.7.2 The Viscosity of the Collective Modes Different modes in the dipolar fluctuations characterize the dielectric of chiral smectics. As the fluctuations in the primary order parameter, for instance represented by the tilt vector 1 2 1 sin 28 sin q
= n, n, = - sin28 cosq
t2= n, ny =
(350)
are coupled to P,some excitations will appear as collective modes. Dipolar fluctuations which do not couple to lj give rise to non-collective modes. From the tempera-
ture and frequency dependence of these contributions, we can calculate the viscosities related to different motions. In the standard description of the dielectric properties of the chiral tilted smectics worked out by Carlsson et al. [ 1521, four independent modes are predicted. In the smectic C* the collective excitations are the soft mode and the Goldstone mode. In the SmA* phase the only collective relaxation is the soft mode. Two high frequency modes are connected to noncollective fluctuations of the polarization predicted by the theory. These two modes become a single noncollective mode in the smectic A* phase. There is no consensus [153] as yet as to whether these polarization modes really exist. Investigations of the temperature dependence of the relaxation frequency for the rotation around the long axis show that it is a single Cole-Cole relaxation on both sides of the phase transition between smectic A* and smectic C* [ 1541. The distribution parameter a of the Cole-Cole function is temperature-dependent and increases linearly (a = aTT + bT) with temperature. The proportionality constant aT increases abruptly at the smectic A* to SmC* transition. This fact points to the complexity of the relaxations in the smectic C* phase. The dielectric contribution of each of these modes has been worked out using the extended Landau free energy expansion of Sec. 2.5.11. The static electric response of each mode is obtained by minimizing the free energy in the presence of an electric field. The relaxation frequency of the fluctuations in the order parameter is obtained by means of the Landau-Khalatnikov equations, which control the order parameter dynamics. The kinetic coefficients r, and r2of these equations have the dimensions of inverse viscosity. The soft mode and Goldstone mode are assumed to have the same kinetic coef-
2.7
Dielectric Spectroscopy: To Find the 7 and ETensor Components
ficient, r,,which, however, does not mean that they have the same measured value of viscosity. The viscosity of the soft mode in the smectic A* phase can be written, according to [152]
(351) where C* is the coupling coefficient between tilt and polarization, es is the dielectric contribution from the soft mode, andf, is the relaxation frequency of the soft mode in the smectic A* phase. In order to calculate the viscosity, a measure of the coefficient C* is needed. It can be extracted from the tilt angle dependence on the applied field, from Sec. 2.5.7
(352) The corresponding expression for the soft mode viscosity in the SmC* phase is not equally simple, and reads
(353) where b,lb, is the following combination of parameters from the expansion (263) C* - e q + 252P8 I ne3+37~2
($)-=xo -
(354)
Eo
This ratio can be obtained by fitting the experimentally measured values of E, versus temperature. In the frame of this Landau theory, the soft mode dielectric constant is
where the term wois a cut-off parameter.
619
Assuming the ratio b,lb, is independent of temperature, ye can be calculated in the smectic C* phase. This approximation is valid deeper in the smectic C* phase, where P , 8,and q have their equilibrium values. The Goldstone mode viscosity can be calculated [ 1521 according to the expression
(356) where P is the polarization, 8 is the tilt angle, and % andf, are the dielectric constant and relaxation frequency of the Goldstone mode. The polarization and the tilt angle need to be measured in order to calculate the viscosity. Figure 78 shows a plot of the soft mode and Goldstone mode rotational viscosities measured on either side of the phase transition between the smectic A* and SmC*. It can be seen that, except in the vicinity of the phase transition, the viscosity ye seems to connect fairly well between the two phases. The activation energies of these two processes are, however, different. This result may be compared to results obtained by Pozhidayev et al. [148], referred to in Fig. 67. They performed measurements of ye beginning in the chiral nematic phase of a liquid crystal mixture with corresponding measurements in the SmC* phase, and have shown the viscosity values on an Arrhenius plot for the N* and SmC* phases. Despite missing data of in the smectic A* phase they extrapolate the N* values of down to the smectic C* phase and get a reasonably smooth fit. Their measurements also show that yeis larger than y,, and this is universally the case.
620
2 Ferroelectric Liquid Crystals 10
Y (Nsec/m2)
1
Figure 78. The soft mode and Goldstone mode rotational viscosities as a function of temperature. The material is mixture KU-100 synthesized at Seoul University, Korea(courtessyofProf. Kim Yong Bai) (from Buivydas [lSS]).
0s 20
30
40 50 60 Temperature (“C)
2.7.3 The Viscosity of the Noncollective Modes
(357) where ft is the relaxation frequency of the molecular rotation around the short axis and 1 is the length of the molecule, as shown in Fig. 79. The viscosity related to the molecular rotation around the long axis has the corresponding expression
kT
80
by Nippon Mining Inc., which we may call LC1. It has the structural formula
The viscosity of the noncollective modes can be estimated, as shown by Gouda [ 1541 and Buivydas [155], using the original Debye theory of dielectric response. Assuming that the Stokes law is valid, the viscosity for the long axis rotation around the short axis scales as l r 3 , where 1 is the length of the long axis. For this transverse rotation we have
Yl =-87c2jid3
70
and the phase sequence I 105.0 SmA* 68.1 SmC* 8.0 Cr. The relaxation frequencies f i (of the order of 3 MHz) and f , (of the order of 110 MHz) are now measured as a function of temperature, the former in quasi-homeotropic alignment (layers parallel to the cell plates), the latter in QBS geometry. With a cross section d taken from a simple mothe correlecular model to be about 4.5 sponding function y l ( T ) is obtained from (358). Now, the length and width of a molecule used in Eqs. (357) and (358) are not self-ev-
w,
(358)
where fi is the relaxation frequency and d is the cross diameter of the molecule, cf. Fig. 79. The substance used in these measurements is a three-ring compound synthesized
L d J
~
Figure 79. The molecular dimensions d and 1 The depicted length/ width ratio corresponds to the substance LC1 (see later).
2.7
62 I
Dielectric Spectroscopy: To Find the ?and &TensorComponents
ident values. Therefore the Landau description may be used to calculate the viscosity related to the molecular rotation around the long axis. In the smectic C* phase, the viscosity is then expressed by
Landau method. In order to calculate yl ( T ) we also need a value for the molecular length 1. This can be obtained by the requirement that we can have only one single viscosity in the isotropic phase. Combining (357) and (358) gives
(359) In order to determine the same viscosity in the smectic A* phase, the normalization factor (He)*should be preserved, otherwise the viscosity will show a jump at T, [ 1541. Thus the corresponding equation will read
where (P/8), is the limiting value of the P/8 ratio measured at the transition. The viscosity '/I calculated by the Landau theory and the '/I evaluated by using the Debye formulas are shown for comparison in Fig. 80. We see that they agree fairly well in the limit of small tilt, thus particularly in the A* phase. The non-Arrhenius increase at higher tilts is here an artifact inherent in the
On putting %= we find
From this 1 is obtained as 15 A.When we check the molecular length by searching the lowest energy conformation of LCI in a simple computer program, we see that this value actually corresponds to the length of the core. We could of course also have slightly adjusted the input value of d such that the Debye and Landau viscosity values coincide exactly in the A* phase. This is found to occur for d = 4.7 A,corresponding to 1=16 The results from all viscosity measurements on the substrate LC1 are summarized in Fig. 8 1.
A.
lo2
Smectic C*
Smectic A*
10' ."
I
I
I
Y loo (Ns/mz) 10' 10-2
t 2,87
I 2,91
2,95
' 2,99
3,02
3,06
3,lO
1000/T (K-l)
Figure 80. Viscosities of noncollective motions evaluated by two different methods. The substance is LC1 (from Buivydas [155]).
2,62
2,70
2,79
2,87
2,95
3,04
1000/T (K-')
Figure 81. The three components of the viscosity tensor measured by dielectric relaxation spectroscopy. The substance is LC I (from Buivydas [ 1551).
622
2
Ferroelectric Liquid Crystals
The noncollective rotational viscosities are denoted by yl (longitudinal) and '/t (transverse). The collective motions have the corresponding designations (for soft mode) and yG (for Goldstone mode). Equivalently, yo and y, could also have been used. It is remarkable that the Goldstone mode viscosity is almost as low as the viscosity around the long axis of the molecule. As pointed out earlier, this is explained by the fact that a large component of the angular velocity is around this axis. The soft mode viscosity lies much higher and, as the results show, is essentially the same as the viscosity for the corresponding individual molecular rotation.
2.7.4 The Viscosity yq from Electrooptic Measurements In the smectic C* phase, the rotational viscosity y, can be estimated by observing the polarization reversal or the electrooptic properties of the cell, as described in Sec. 2.7.6. The estimation may, for instance, be based on the approximation mentioned there, using the elastic torque [ 1371
pS Esincp - K cosq + y, @ = 0
(363)
Solving this equation and fitting the observed polarization reversal process to the solution, the viscosity can be estimated and compared to the one measured by means of dielectric relaxation spectroscopy. This method is fast, but less accurate. The polarization reversal measurement is a large signal method requiring full switching of the liquid crystal, and therefore cannot be expected to coincide with the much more precise results from the low signal dielectric relaxation spectroscopy measurements. Basically, a &, value from the polarization reversal technique involves spurious contributions of elastic effects due to the
2,o .' ' ' ' '
1,s r In Y$
' ' '
0
1
I
s
- - . - .,
*
:
I
I
I
*
a
a
I
I _
O
I
1
-
a
: -
non'
-
a
I
-
/
4 5 2 measurements from dielectric a
I
-
4".
' (Nsec/mz) 0,s 1
-l,o
I
from polarization reversal
0
LO :
0,o
I
Smedic C*
"
1
a
W
*
L
T
*
* I*
I -
surfaces of the sample. However, the values from the polarization reversal method do not differ by more than typically a factor of two, and the difference tends to be smaller at higher temperatures (see Fig. 82).
2.7.5 The Dielectric Permittivity Tensor Phases having biaxial symmetry (tilted smectic phases) exhibit dielectric biaxiality in particular. At frequencies of 1 MHz and below, the biaxiality becomes important and critically influences the electrooptic switching behavior of the SmC* phase. It is therefore important to be able to measure the biaxiality at these frequencies. Being a symmetrical second rank tensor, the dielectric permittivity can always be diagonalized in a proper frame and described by three components along the principal directions. The three principal values can then be expressed by a single subscript and can be determined by three independent measurements performed at three different orientations of the director relative to the measuring electric field. In practice, this may not be that
2.7
Dielectric Spectroscopy: To Find the ?and &TensorComponents
straightforward, for instance, in the case of a thick sample (measured on a scale of the pitch) of a smectic C* when the helix is not quenched by the surfaces. In such a case the local frame of the E tensor is rotating helically through the sample. Y
2.7.6 The Case of Helical Smectic C* Compounds
(364)
0
E3
As for y in Sec. 2.6.9, we may consider 2 being brought to diagonal form (starting from the laboratory frame) by a similarity transformation
~(1,2,3)=T-'&(~,y,z)T
E ( 8 ) = TO 2 Tg'
(366)
with the rotation matrix TO cos8 0 0 1 sin8 0
(365)
where T is a rotation matrix and T-' is the inverse or reciprocal matrix to T, in this case identical to its transpose.
E(e) =
Figure 83. The definitions of the molecular axes and components of the dielectric tensor.
In Fig. 83 we may imagine a laboratory frame with the z axis along the layer normal (thus along ell in the figure), such that we have turned the director n a certain angle 8 around the y direction (along E~ in the figure). The x direction is along c. In this laboratory frame xyz, the permittivity tensor is then given by
In the molecular frame (Fig. 83) we write the tensor E as
0
EI
623
- sin8
(367) cos8
Performing the multiplication - the same as in Eq. (342) - we find
- E ~ sin0 ) coso c1cos2 o + E~ sin2 8 o 0 E2 0 ( ~ ~ - ~ ~ ) s i n o~ cE ]osin28+E3cos2e s ~
which can be written in the form
0 -A&sin8cosO E2 0 &(el = -A& sin8 cos8 0 Ell
1
(369)
by introducing the dielectric anisotropy A&= E~ -E' and the abbreviations cos28+ E ~ C O 2S8
(370)
E,, = I ,sin28+ c3cos2e
(371)
=
and If we now further turn the director by an angle cp around the z direction, the permittivity tensor will change to
624
2
Ferroelectric Liquid Crystals
~ ( 89) , = q,E(e)Tml
(372)
As this rotation is performed counterclock-wise (positive direction) in the xy plane, the rotation matrix is now instead
(373)
This gives
&(e,q) = T,T,~T,-~T,-~
(374)
or ) cosq -A&sin8 cos8 cosq cos2 q + E~ sin2 q -(el - E ~ sinq - ( ~ 1 - ~ ~ ) s i n q c o s q sin2 q + E~ cos2 q AEsin8 cos8 sinq AE sin 8 cos 8 sin q - A&sin q cos 8 cos q ElI
In the helical smectic C * , the tilt 8 is a constant while the phase angle q is a function of the position along the layer normal, q= q ( z ) . Averaging over one pitch, with (cos2 q) = (sin2 q) =
+
and (sinq) = (cosq) = (sinqcosq) = 0 E
simplifies to
0
(Ell)
with the new abbreviations 1 2
( E ~=) - (cl CO?
8 + E~ + E~ sin28)
and (E,,) =
sin28+ E ~ C O S ~ O
Equation (378) reveals that the helical state of the smectic C* is effectively uniaxial, if the pitch is sufficiently short to al-
1
(375)
low the averaging procedure. The shorter and more perfect the helix, the more completely the local biaxiality is averaged out (both dielectric and optical). This is consistent with the empirical evidence, showing almost perfect uniaxiality for helical samples of short pitch materials. Before we can use our derived relations to determine E ~ E~ , and E ~ we , have to perform one more refinement. The simple measurement geometry where smectic layers are parallel to the glass plates is easy to obtain in practice. On the other hand, it is very hard to turn them into being perfectly parallel to the plates; normally, the result is that they meet the surface with a slight inclination angle 6. To account for this case we have to see how the tensor ( ~ ( 0 )transforms )~ when we turn the layer normal by a certain angle 6 around the x axis. We perform the rotation clockwise (in this case the direction does not matter), corresponding to
2.7
625
Dielectric Spectroscopy: To Find the ?and CTensor Components
and giving 0
( ~ ( 86))q , = Ts ( ~ ( 8 T; )' ) ~ =
0 ( E ~cos2 ) 6 + ( ~ 1 1 ) sin26 (EL) - ( ~ ~ 1sin ) 6 cos 6 (EL) - (q) sin6 cos6 (eL)sin26 + ( ~ ~ , ) ( c6o) s ~
o
In the case where we assume a helix-free sample and fully addressed or memorized director states lying on extreme ends of the smectic cone, so that we might put cp equal to 0 or n in Eq. (375), would instead have given the tensor components cos2 8 + E~ sin2 8 A~sin8cos8sin6 - A&sin 8 cos 8 cos 6 E'
A&sin 8cos 8sin 6 ~2 cos2 6 + ( ~ 1 1 ) sin2 6 ( ~ -(elI))sin6cos6 2
-A& sin 8 cos 8sin 6 (12- (ell))sin 6 cos 6
(383) ~~sin~6+(~~~)cos~6
2.7.7 Three Sample Geometries
Eq. (378) or, by Eq. (379)
With the expressions derived in the last section, in the first version worked out by Hoffmann et al. [ 1561 and further developed by Gouda et al. [157], we can calculate E ~ E, ~ and E~ from three independent measurements, using three different sample geometries. Consider first that we have the smectic layers oriented perpendicular to the glass plates so that the applied field is perpendicular to the layer normal. In the Smectic A* phase we would call this planar alignment, and a dielectric measurement would immediately give the value in the SmA* phase. If we cool to the SmC* phase, the helix will appear in a direction parallel to the glass plates. By applying a bias field we unwind the helix and the signal field will now measure &nw, which is identical to ~2
&helix = -(el 1
Eunw
= EZ
(384)
the permittivity value along the direction of the polarization P . If we take away the bias field the helix will eventually relax. The value of the permittivity measured in the presence of a helix is denoted by &helix. It is equal to E~~ in the lab frame, i.e., to (el) in
2
,
cos28 + e2 + ~3 sin28 )
(385)
The third geometry normally requires a second sample aligned with the layers along the glass plates. We may call this geometry homeotropic and in the SmA* phase we measure the value &h = E,, along the layer normal. After cooling to the SmC* phase we may call the geometry quasi-homeotropic and denote the measured value &horn. The helix axis ( z ) is now parallel to the measuring field, which means that &horn is ,&, in the lab frame, which is (4,) in Eq. (378) or, by Eq. (380) =
&horn
sin28+ ~3 c0s28
(386)
From Eqs. (384) to (386) the three principal values of the dielectric tensor can thus be E ~ and , r,, calculated. After solving for the components of the diagonalized tensor can be expressed in the measured values as
1 1- 2sin2 8
&I = '
[(2&helix - &,nw ) Cos2 8 - &horn Sin281 (387)
&2 = ELinw
(388)
626
2
Ferroelectric Liquid Crystals
1 I - 2sin2 e
E3 =
be
. [(2&he1ix - E , , ~)sin 2 e - &horn cos2 e] (389) The three measurements require preparation of two samples. By using a field-induced layer reorientation, first investigated by Jakli and Saupe [158], Markscheffel [158a] was able to perform all three measurements on a single sample, starting in the quasi-homeotropic state, and then, by a sufficiently strong field, turning the layers over to the quasi-bookshelf structure. Generally, one would start with the quasibookshelf geometry and then warm up the sample to the smectic A* phase and strongly shear. The advantage of this technique is a very homogeneous, homeotropic orientation. The quality of homeotropic orientation may even improve (without degrading the quality of planar orientation) if the glass plates are coated by a weak solution oftenside.
= ( E , ) C O S+ ~~ (q,)sin26
(391)
while &horn stays unaffected, i.e., ( E ~ ~given ), by Eq. (386), because of the absence of the layer tilt in this geometry. The value of 6 is often known, growing roughly proportional to 6, and can in many cases be approximated as 6 ~ 0 . e. 8 The principal values of qj can then be extracted from Eqs. (386), (390), and (391). It turns out that the values are not very sensitive to small errors in 6. A special case of tilted layers is the chevron structure, for which the tilt makes a kink either symmetrically in the middle or unsymmetrically closer to one of the surfaces. The chevron structure means that two values +6 and -6 appear in the sample. As can be seen from Eqs. (389) and (390), the angle 6 only appears as a quadratic dependence, hence the relations are also valid in the chevron case. Solving for the principal values in the chevron or tilted layers case, we obtain 1
(392)
&'
=
~
1 (&nw - &horn sin' 6 ) cos2 6
(393) (394)
2.7.8 Tilted Smectic Layers When the layers make the inclination 6with the normal to the glass plates, we see from Eq. (383) that q,nw (= &yy) will be expressed by (E,,,
= &,cos26
whereas
+ (ql)sin26
390)
= E~~ in Eq. (382) turns out to
Typical measured variations of the permittivities &,, e2,and E ~as, we pass the phase boundaries from isotropic to tilted smectic phase, are shown in Fig. 84. The values of and ~2 have a similar temperature variation. Therefore their difference, the biaxiality 6&, has a weak but noticeable temperature dependence. The difference between the values and E~ is the dielectric anisotropy A&.This depends on temperature much
2.7
40
80 100 Temperature (“C)
60
627
Dielectric Spectroscopy: To Find the ?and 2.Tensor Components
120
Figure 84. The principal values of the dielectric permittivity tensor at 100-kHz measured by the short pitch method. The material is the mixture SCE12 by BDHlMerck (from Buivydas [155]).
scribed by Sambles et al. [159]. They found the profile to be uniform or nearly so within each surface-stabilized domain. However, other reports show that the surface-stabilized state is more accurately modeled by some triangular director profile [ 1601. When a uniform director profile is assumed, then the smectic director field n can be described by the three angles q, 6,and 6. These angles are uniform throughout the sample and the dielectric tensor component E~~ corresponds to the dielectric permittivity q,measured in the planar orientation. Under certain conditions this may be written Ep = &2 -8&
~
more strongly than the dielectric biaxiality. As a result of the fact that the relaxation frequency of E~ in the nematic phase lies below 100 kHz, the dielectric anisotropy is negative at that frequency. A negative value of A& is a common feature, which may greatly affect the electrooptic switching properties at room temperature. In contrast, the relaxation frequency of E~ lies much above 100 kHz, whence the positive biaxiality 8~(which may rather be considered as a low frequency value). This means that for information about A& we have to go to relatively low frequencies and inversely to high frequencies for corresponding information about 8 ~ .
sin2 6 sin28
(395)
where
8&= &2 - &I
(396)
is the dielectric biaxiality. When measuring in the quasi-homeotropic orientation, we will observe the component &h = E,, and, assuming the layer tilt 6 to be zero, &h takes the form gh = ~2
+A E C O S ~ ~
(397)
Again we have to look for an independent measurement of a third value, in order to cal-
2.7.9 Nonchiral Smectics C When the cell has another director profile than the uniform helix, averaging of the angle q over the pitch length cannot be performed in a straightforward way. It is necessary to know about the director profile through the cell before making simplifications. This profile is hard to get at but may be estimated by optical methods, as de-
SmC
30
40
SmA N
50 60 Temperature (“C)
70
Iso-
-
80
Figure 85. The dielectric biaxiality of the nonchiral smectic C mixture M4851 (Hoechst), measured by the long pitch method (from Buivydas [155]).
628
2
Ferroelectric Liquid Crystals
culate all three principal components of the dielectric permittivity tensor. At the second order transition SmA -+SmC, the average value of E is conserved, thus (398) With knowledge of the tilt angle 8 and the layer inclination 6, we then can calculate E ~ and , E~ from Eqs. (395), (397) and (398).
2.7.10 Limitations in the Measurement Methods
The reason for this failure is that for a 45" tilt angle 8 there is no difference in average dielectric anisotropy between the planar orientation and the quasi-homeotropic orientation, and thus the two equations in the system (Eqs. 385 and 386) are linearly dependent as far as that they give the same geometric relation between the direction of the measuring field and the direction of the principal axes and E~ (& is not involved in Equations (385) and (386) in this case give &helix
=
51 (El -tE 2 )
&2
(399)
and Each of the models serving as basis for the dielectric biaxiality measurements has been derived using certain assumptions. The method described by Gouda et al. [ 1571 assumes an undisturbed ferroelectric helix, and this assumption is best fulfilled in short pitch substances. Actually, the pitchhell thickness ratio is the important parameter. For measurement convenience, the cells cannot be made very thick, hence the method works well for materials with a helical pitch of up to about 5 pm. This includes practically all cases of single smectic C* substances. The method cannot be applied to nonchiral materials. The method described by Jones et al. [161] (long pitch method) was developed for materials with essentially infinite pitch. It may be used for long pitch materials in cells where the helical structure is unwound by the surface interactions and a uniform director profile is established. The weakness of the method is that it requires the director profile to be known. On the other hand, it works for the nonchiral smectic C. It also works for materials with 8=45" (typically materials with N 4 SmC transitions, lacking the SmA phase), which is a singular case where the short pitch method fails.
&horn =
z1
(El + E3)
(400)
respectively, whence it is not possible to , only measure their sum. split and E ~ but Using tilted layers (6#0) or knowing the average value E is not remedy for this, because only the combination &, + E~ appears in all equations. As a result of this the errors in the calculated values of and E~ increase and diverge when 8 approaches the value 45'. This is also evident from the denominator (1-2 sin28) in Eqs. (392) and (394). In Fig. 86 we show how the errors in and E~ grow with the tilt angle under the assumption that the value used in the calculation formula was measured with a 2% error and that the value of was assumed to be exact. The presence of this singularity limits the usability of the short pitch method to substances with tilt angle 8 up to about 35-37'. In practice, however, this is not too serious, as materials with a tilt approaching 40" are rare.
2.8 FLC Device Structures and Local-Layer Geometry 30
25 20 Error 15 (%)
10
5 0 0
10
20 30 Tilt angle 9 (deg)
40
50
Figure 86. The dependence of the errors in the tensor components E, and E~ calculated by the short pitch method when the smectic tilt angle @approaches45 ’. The input value is assumed to be measured with a 2% error (from Buivydas [ 1551).
2.8 FLC Device Structures and Local-Layer Geometry 2.8.1 The Application Potential of FLC Ten years of industrial development offerroelectric liquid crystal (FLC) devices have now passed. This may be compared with 25 years for active matrix displays. The potential of FLC was certainly recognized early enough following the demonstration of ferroelectric properties in 1980. In principle the FLC has the potential to do what no other liquid crystal technology could do: in addition to high-speed electrooptic shutters, spatial light modulators working in real time, and other similar hardware for optical processing and computing (all applications requiring very high intrinsic speed in repetitive operation), the FLC (and the closely related and similar AFLC) technology offers the possibility to make large-area, highresolution screens without the need for tran-
629
sistors or other active elements, i.e., in passive matrix structures using only the liquid crystal as the switching element. Several kinds of such structures (Canon, Matsushita, Displaytech) have now been demonstrated which give very high quality rendition of color images, and one polymer FLC technology has been developed (Idemitsu) capable of monochrome as well as multi-color panels. When the SSFLC structure was presented in 1980, some obvious difficulties were immediately pointed out: the liquid crystal had to be confined between two glass plates with only about 1.5 pm spacing or even less and, furthermore, in a rather awkward geometry (“bookshelf” structure), which meant that both the director alignment and the direction of smectic layers had to be carefully controlled. Luckily, a number of more serious problems were not recognized until much later. In addition, only a few molecular species had been synthesized at that time; in order to develop a viable FLC technology, a considerable effort in chemical synthesis would have had to be started and pursued for many years. (We may recall that the submicrosecond switching reported in the first paper had been observed on a single substance, HOBACPC, at an elevated temperature of 68 “C.) This chemical problem alone would have discouraged many laboratories from taking up the FLC track. Nevertheless, in 1983 excellent contrast and homogeneity in the optical properties were demonstrated on laboratory samples of DOBAMBC, one of the other single substances available (thus not on a room temperature mixture) and sufficiently well aligned for that purpose by a shearing technique [96]. Shortly afterwards, two Japanese companies, Canon Inc. and Seiko Instruments and Electronics, were engaged in R&D. By 1986 about twenty Japanese and five European companies were engaged in
630
2
Ferroelectric Liquid Crystals
FLC research, supported by about ten chemical companies worldwide. Common to almost all these ventures was the fact that the FLC teams were very small. When really troublesome obstacles began to show up, one company after the other left the scene, favoring STN or TFT technology. It was clear that the obstacles were not only in manufacturing technology, but there were tremendous scientific difficulties of a very basic nature as well. After having presented the first mature FLC prototypes in 1988 (black and white) and 1992 (color), Canon in Tokyo is now manufacturing the first 15 in. (37.5 cm) FLC panel with 1280x 1024 pixels, where each pixel, of size 230 ymx230 ym, can display 16 different colors because it is composed of four dots or subpixels. This screen (Fig. 87) has a remarkable performance and does not resemble anything else in its absolute freedom of flicker. At the other end of the scale, Displaytech Inc., U.S.A.,
is marketing FLC microdisplays, slightly larger than 5 mm by 5 mm in size with VGA resolution (640x480), where every dot is capable of 5 12 colors (see Fig. 88). Another remarkable display is Idemitsu Kosan’s fully plastic monochrome screen in reflection, of A4 size (15 in) with 640x400 dots (see Fig. 89). This is a passively multiplexed (1 :400) sheet of FLC polymer, thus entirely without glass plates and all in all 1 mm thick. Regarding these achievements, it may be time to look at the real SSFLC structures that are used in these devices.
2.8.2 Surface-Stabilized States It is important to point out that for a given material there is a variety of different realizations of surface-stabilized states (states with nonzero macroscopic polarization), on the one hand, and also a variety of other states. Among the former, those with a non-
Figure 87.Canon 15 in (37.5 cm) color screen which went into production 1995. High quality rendition of artwork is extremely sensitive to flicker. The FLC freedom of flicker therefore gives an unusual means of presenting art, here illustrated by the turn of the century artist Wyspianski (self-portrait) and Bonnard (Dining Room with Garden).
2.8 FLC Device Structures and Local-Layer Geometry
63 I
Figure 88. The first version of Displaytech’s miniature display (ChronoColor), which went into production in early 1997, in comparison with a conventional active matrix display. This microdisplay utilizes sequential color to produce full color on each pixel, resulting in a brighter, crisper image than that of an AMLCD, which uses a triad of red, green, and blue pixels to form a color. In the insets, the pixels of the ChronoColor display and AMLCD are magnified to show the difference in fill factors. The color is what makes the image that you see, and it occupies 75% of the display on the left, but only 35% of the corresponding active matrix display. (Courtesy of Displaytech, Inc.)
Figure 89. All-polymer FLC prototype presented by Idemitsu Kosan Co., Ltd., in 1995. This reflective monochrome display is 0.5 mm thick, has a 1:400 multiplex ratio, and a size of 20 cmx32 cm (400x640 pixels each 0.5 mmx0.5 mm), corresponding to a diagonal of 15 in (37.5 cm).
zero P along the applied field direction will have a strong nonlinear dielectric response in contrast to other states, which behave linearly until saturation effects dominate. Only the former may be bistable, i.e., show
a stable memory (P#O in the absence of an applied field E). We will give two examples to illustrate this. The first is a case with supposedly strong anchoring of the director along a specified azimuthal direction cp, cor-
632
2
Ferroelectric Liquid Crystals
responding to parallel rubbing at top and bottom plate of the cell 11271. This case is illustrated in Fig. 90, where successive n -P states are shown across the cell. Clearly we have two completely symmetrical arrangements, which have a certain effective polarization down and up, respectively, and which are, each of them, a stable state. They can be switched back and forth between each other by the application of convenient pulses, and have a symmetrical relationship similar to the two cases of Euler buckling of a bar in two dimensions under fixed boundary conditions. The cell is thus surface-stabilized and ferroelectric with a symmetrically bistable reponse. However, it is immediately clear that this configuration may not be the most desirable one, because we have a splayed P state with only a small part of P that can actually be switched by a short pulse, requiring a high voltage for a large optical effect. This is because the optical
contrast between the two memorized up and down states is very low. In order to enhance this contrast, we would have to apply a considerable AC holding voltage across the cell after application of the switching pulse, which seems to make this use of ferroelectric structure rather pointless. Clearly the structure is intrinsically inferior to the structure in Fig. 61 where the director profile is supposed to be homogeneous across the cell. However, it not only serves the purpose of illustrating that surface-stabilized structures can be of many kinds and that they might be characterized by their electrical (speed) as well as their optical efficiency, but also this structure actually turns out to be of high practical interest, as we will see in Sec. 2.8.6. The second example is what we call a twisted smectic [162], and is illustrated in Fig. 91. Ideally a smectic C* with 8=45" should be used as a twisted smectic. In Fig. 9 1 the anchoring is again assumed to be strong (although this is not a necessity for
@=-n
FIELD OFF
FULL FIELD ON
(TRANSMISSIVE)
(MINIMUMTRANShfIlTANCE) ,
polarizer glass
~
w
,
Nbbing direction
smectic layer rubbing direction glass
Figure 90. Surface-stabilized configuration with less than optimum efficiency, switchable between two symmetric states with low optical contrast. The surface pretilt angle has been chosen equal to the smectic tilt angle 8 i n this example. For a strong boundary condition with zero pretilt, a different extreme limiting condition with @approachingzero at the boundary is also conceivable, without any essential difference in the performance of the cell.
polarizer
Figure 91. The twisted smectic preferably uses a 45 O tilt material. It i s a fast-switching device giving an electrically controlled, continuous grayscale. This is a smectic C* working in a dielectric mode. It should not be mistaken for a ferroelectric device (after Pertuis and Pate1 [ 1621).
2.8
FLC Device Structures and Local-Layer Geometry
the working principle), and the alignment directions are at right angles (with zero pretilt at the surface) at top and bottom, introducing a 90” twist across the cell. This also results in zero macroscopic polarization along the direction in which the electric field is applied. With no field applied, the light will be guided in a way similar to that in a twisted nematic, and the cell will transmit between crossed polarizers. On application of an electric field, the twisted structure will unwind, until the whole structure except for one boundary layer has a homogeneous arrangement of the director, giving zero transmittance. The response is linear at low electric field and we have a fast switching device with a continuous grey scale. It has to be pointed out that this, of course, is not a ferroelectric mode. Thus a smectic C* liquid crystal does not have a ferroelectric response per se, but it can be used in many different ways. As already pointed out, bulk structures of smectic C* tend to have an antiferroelectric-like ordering in one way or another. The point about surface stabilization is that it transfers macroscopic polarization to the bulk, giving it ferroelectric properties in certain geometries. These properties are best recognized by the appearance of spontaneous domains of up and down polarization, and by the fact that the response to an electric field is now strongly nonlinear. Let us finally look back to Fig. 61 and consider the material parameters 6, P, and d, i.e., tilt angle, polarization, and smectic layer thickness. They all depend on temperature: B(T),P ( T ) , and d ( T ) . The temperature variations of 0 and P , if not desirable, at least turned out to be harmless. On the other hand, the much smaller temperature dependence of d ( T )turned out to be significantly harmful. If the liquid crystal molecules behaved like rigid rods, the layer thickness would diminish with decreasing temperature accord-
633
ing to d,cos 6, as the director begins to tilt at the SmA +SmC transition, i.e., we would have a layer shrinkage
A d ( T ) - 62
(401)
in the smectic C phase. In reality, the shrinkage is less (i.e., the molecules do not behave like rigid rods), but still a phenomenon that is almost universal, i.e., present in almost all materials. If, in Fig. 61, we consider the translational periodicity d to be imprinted in the surface at the nematic - smectic A transition, and if we further assume that there is no slip of the layers along the surface, the only way for the material to adjust to shrinking layers without generating dislocations is the creation of a folded structure in one direction or the other on entering the SmC phase, as illustrated in Fig. 92. As the layer thickness dc decreases with decreasing temperature, the chevron angle 6 increases according to
dc cos6 = dA
Figure 92. The shrinkage in smectic layer thickness due to the molecular tilt O(Q in the SmC* phase results in a folding instablity of the layer structure (“chevrons”). Even if the fold can be made to go everywhere in the same direction (in the figure to the right) to avoid invasive zigzag defect structures, the switching angle is now less than 2 0, which lowers brightness and contrast.
634
2
Ferroelectric Liquid Crystals
where 6 i s always smaller than 8. Often 6 and 8 have a similar dependence, such that 6= A@ with A= 0.85 -0.90 in typical cases. The resulting “chevron” structure constitutes one of the most severe obstacles towards a viable FLC technology, but its recognition in 1987 by Clark and his collaborators by careful X-ray scattering experiments [163, 1641 was one of the most important steps both in revealing the many possibilities of smectic local layer structures and in controlling the technical difficulties resulting from these structures.
2.8.3 FLC with Chevron Structures The extremely characteristic zigzag defects had been observed for many years before the
subtleties of the local layer structure were unveiled. Very often they appear as long “streets” or “lightning flashes” of thin zigzag lines, with the street going essentially along the smectic layers, whereas the lines themselves are at a small angle or almost perpendicular to the layers. In Fig. 93 the angle between the lines and the layer normal is roughly 5 ” . Often these thin lines are “short-circuited” by a broad wall, with a clear tendency to run parallel to the layers, as in Fig. 93a where it is almost straight and vertical, or slightly curved, as in Fig. 93b. It is evident that if the layers are not straight in the transverse section of the sample but folded, as in Fig. 92, then where we pass from an area with one fold direction to an area with the opposite direction there must be a defect region in between, and hence it is likely that the zigzags mediate such
(b)
Figure 93. Appearance of zigzag walls in a smectic C structure with smectic layers initially (in the SmA phase) standing perpendicular to the confining glass plates, running in a vertical direction on the micrographs (thin lines to the left, broad walls to the right). (Courtesy of Monique Brunet, University of Montpellier, France).
2.8 FLC Device Structures and Local-Layer Geometry
changes. The question is how. This may be illustrated starting from Fig. 94. In (a) we show the typical aspect of a zigzag tip with ybeing the small angle between the line and the layer normal (which is often the same as the rubbing direction). In (b) is shown the folded structure making chevrons of both directions, suggesting that at least two kinds of defect structure must be present. In the picture the fold is, as is most often the case, found in the middle of the sample - we will call this fold plane the chevron interface although its depth may vary, as in (c). We will mainly restrict ourselves to the symmetric case (b) in the following discussion of the general aspects. As we will see, the chevron interface actually acts as a third surface, in fact the one with the most demanding and restrictive conditions on the local director and polarizations fields. It is clear that the transition from one chevron to the other cannot take place abruptly in a layer as sketched in (d), as this would lead to a surface (2D) defect with no continuity of layer structure nor director field. Instead it is mediated by an unfolded part of width w (this region will be observed as a wall of thickness w ) in (e), looked at from the top as when observing the sample
through a microscope. The wall thickness \.t’ is of the order of the sample thickness L. A more adequate representation of the threedimensional structure is given in Fig. 95. The vertical part of the layer has the shape of a lozenge (in the general case, with unsymmetric chevrons it becomes a parallelogram). The sterically coupled fields n (r) and P ( r ) can now be mapped on this structure. Along the folds, n ( r )must simultaneously satisfy two cone conditions (i.e., to be on the cone belonging to a particular layer normal) indicated in the figure. This is always possible for 61 8 and gives a limiting value for the chevron angle 6. After the discovery of the chevron structures, a number of complexities in observed optical and electrooptical phenomena could be interpreted in these new terms. For a more thorough discussion in this matter, we refer to references [ 166- 1681. Our emphasis will be on the important consequences for the physics due to the presence of chevrons, but even this requires dealing with at least some structural details. First of all, the chevron fold forces the director n to be in the interface, i.e., to be parallel to the plane of the sample in this region, regardless of how n ( r ) may vary through the rest of the sample and independent of the surface conditions. Although the director n is continuous at the chevron interface, the local polarization field P cannot be, as shown in Fig. 96. It makes a jump in direction at the interface, nevertheless, in such a way that
V.P=O (c)
Irt)
(e)
Figure 94. Thin zigzag lines run almost perpendicular to the smectic layers making a small angle ywith the layer normal as shown in (a). In (b) we look at the chevron folds in the plane of the sample, in (e) perpendicular to the plane of the sample. Asymmetric chevrons (c) cannot preserve anchoring conditions at both surfaces and are therefore less frequent.
635
(403)
which means that the discontinuity is not connected with a build-up of local charge density. If we now apply an electric field between the electrodes, i.e., vertically as in (a), we see that there is an immediate torque because E and P are not collinear. Hence no nucleation is needed for switching from state 1 to state 2. Neither are fluctuations
636
2
Ferroelectric Liquid Crystals
Figure 95. Section of a thin wall mediating the change in chevron direction. The layers in the chevron structure make the angle 6 (the chevron angle) with the normal to the glass plates. Along a chevron fold where two surfaces meet, two cone conditions have to be satisfied simultaneously for n, which can be switched between two states. At the two points of the lozenge where three surfaces meet, three conditions have to be satisfied and n is then pinned, thus cannot be switched.
Ej
f r - r r l
(3
kK--Nk
@
4 ______________________________
Figure 96. (a) At the chevron interface the local polarization P is discontinuous, making a jump in direction. (b) When switching from the down to the up state, P rotates everywhere anticlockwise above and clockwise below the chevron interface (time axis to the right). The director is locked in the chevron plane and can move between n , and n2.
needed for this. In other words, the switching process is not fluctuation-controlled. On the other hand, the chevron region requires an elastic deformation of the local cone in order to switch: the local tilt 8 must dimin-
ish to a smaller value (midway between 1 and 2), which gives a contribution to the threshold for the process. Generally speaking, the bulk switching on either side of the chevron interface preceeds the switching in the interface. The latter contributes to the latching and thus to the bistability. As seen from Fig. 96b, the switching process is unambiguous as regards the motion of IZ and P (sterically bound to n): on the upper side of the chevron, P rotates counterclockwise, on the lower side it rotates clockwise when we switch from 1 to 2; everything turns around in the reverse switching direction. This explains why there are no twist and antitwist domains like the ones observed in twisted nematics prior to the time when chiral dopants were added in order to promote a certain twist sense. So far we have described the switching concentrating on the chevron interface, completely disregarding what could happen at the two bounding (electrode) surfaces. In fact, if the anchoring condition on the surfaces is very strong, switching between up and down states of polarization will only
2.8
FLC Device Structures and Local-Layer Geometry
take place at the chevron interface. At high voltage this will more or less simultaneously take place in the whole sample. At low voltage it will be possible to observe the appearance of down domains as “holes” created in an up background, or vice versa, in the shape of so-called boat domains (see Fig. 105) in the chevron interface (easily localized to this plane by optical microscopy). The walls between up and down domains have the configuration of strength one (or 27c) disclinations in the P field. It should be pointed out that the uniqueness of director rotation during the switching process is not a feature related to the chevron per se, but only to the fact that the chevron creates a certain P-tilt at the chevron interface. If the boundary conditions of the glass surfaces involved a similar P-tilt, this will have the same effect. A glance at Fig. 97 reveals another important consequence of the chevron structure. As P is not along E (applied field) there will always be a torque PxE tending to straighten up the chevron to an almost upright direction. Especially in antiferroelectric liquid crystals, which are used with very high P values, this torque is sufficiently strong for almost any applied field, for instance, normal addressing pulses, to raise and keep the structure in a so-called quasibookshelf structure (QBS) under driving conditions. In ferroelectric liquid crystals,
Figure 97. The fact that even after switching the polarization P is not entirely in the direction of the applied field will tend to raise the chevron structure into a more upright position, so decreasing the effective 6 but breaking up the layers in a perpendicular direction. This gives a characteristic striped texture from the newly created, locked-in defect network.
637
presently with considerably lower P values, the same effect was previously employed to ameliorate contrast and threshold properties, by conditioning the chevron FLC to QBS FLC by the application of AC fields 11691. The effect on the switching threshold can be extracted from Fig. 96. When the chevron structure is straightened up, 6 decreases and the two cones overlap more and more, leading to an increasing distance between 1 and 2, as well as further compression of the tilt angle 6 in order to go between 1 and 2. The threshold thus increases, in agreement with the findings of the Philips (Eindhoven) group. On the other hand, this straightening up to QBS violates the conservation of smectic layer thickness d,, which will lead to a breaking up of the layers in a direction perpendicular to the initial chevrons, thus causing a buckling out of the direction running perpendicular to the paper plane of Fig. 97.
2.8.4
Analog Grey Levels
As we just pointed out, in the chevron structure the polarization is no longer collinear with the external field. This can be used (for materials with a high value of P ) to straighten up the chevron into a so-called quasibookshelf structure, combining some of the advantages from both types of structure. For instance, it can combine a high contrast with a continuous gray scale. How to produce analog gray levels in an SSFLC display is perhaps not so evident, because the electrooptic effect offers two optical states, hence it is digital. Nevertheless, the shape of the hysteresis curve reveals that there must be small domains with a slightly varying threshold, in some analogy with the common ferromagnetic case. Normally, however, the flank of the curve is not sufficiently smeared out to be controlled and to
638
2
Ferroelectric Liquid Crystals
accommodate more than a few levels. Curve I of Fig. 98 shows the transmission -voltage characteristics for a typical SSFLC cell with the layers in the chevron configuration [ 1651. The threshold voltage is fairly low, as well as the achievable transmission in the bright state, leading to a low brightnesscontrast ratio. The position and sharpness of the threshold curve reflect the relatively large and constant chevron angle S, in the sample. If a low frequency AC voltage of low amplitude (6- 10 V) is applied, the smectic layers will be straightened up towards the vertical due to the P - E coupling, so that the local polarization vector increases its component along the direction of the field. This field action, which requires a sufficiently high value of P , breaks the layer ordering in the plane of the sample and introduces new defect structures, which are seen invading the sample. The result is that the chevron angle 6 is reduced, on average, and the threshold smeared out, as shown by curve 11. Lower 6 means a larger switching angle (and higher threshold), and thus higher transmission. Still higher transmission can be achieved by an additional treatment at a somewhat higher voltage (+25 V), giving threshold curve 111, corresponding to a new distribution around a lower &value and a microdomain texture on an even finer scale. The actual switching threshold is a complicated quantity, not fully understood (no successful calculation has been presented so far), and usually expressed as a voltagetime area threshold for the switching pulse. For a given pulse length it is, however, reasonable that the amplitude threshold increases according to Fig. 98 when the average value of 6 decreases. There are at least two reasons for this, as illustrated by Fig. 96. First, it is seen that the distance between the two positions n l and n2 in the chevron kink level (which acts as a third,
T
t
Figure 98. Amplitude-controlled gray scale in SSFLC. The chevron structure is transformed to a quasi-bookshelf (QBS) structure by external field treatment. In addition to giving gray shades, the QBS structure increases the brightness and the viewing angle. This method of producing gray levels was developed by the Philips (Eindhoven) group, who called it the “texture method”.
internal surface), as well as the corresponding positions at the outer surfaces and in between, increase when 6 decreases. It would therefore take a longer time to reach and pass the middle transitory state, after which the molecules would latch in their new position. In addition, it is seen that the local deformation of the cone i.e., a decrease of the tilt angle 6,which is necessary to actuate the transition from n , to n2, increases when Gdecreases. (A paradox feature of this deformation model is that it works as long as 6+0, whereas 6=0 gives no deformation at all -but also no chevron - at the chevron kink level.) . The smectic layer organization corresponding to curves I1 and I11 of Fig. 98 is generally characterized as a quasi-bookshelf (QBS) structure, denoting that the layers are essentially upright with only a small chevron angle. The QBS structure has a very large gray scale capacity. This might, however, possibly not be utilized to advantage in a passively driven display (as it can in the AFLC version). Its drawback in this
2.8 FLC Device Structures and Local-Layer Geometry
respect is that the shape of the threshold curve is temperature-dependent, which leads to the requirement of a very well-controlled and constant temperature over the whole area of a large display. Furthermore, the QBS structure is a metastable state. Finally, the microdomain control of gray shades requires an additional sophistication in the electronic addressing: in order to achieve the same transmission level for a given applied amplitude, the inherent memory in the microdomains has to be deleted, which is done by a special blanking pulse. Using this pulse, the display is reset to the same starting condition before the writing pulse arrives. As a result of these features, it is not clear whether the microdomain method will be successfully applied in future FLC displays. A combination FLCTFT might be required; if so it will be quite powerful.
2.8.5 Thin Walls and Thick Walls The micrographs of Fig. 93 indicate that the thin wall normally runs with a certain angle y to the smectic layer normal. What determines this angle 3: what determines the thickness of the wall, and what is the actual structure of it? To answer these questions w e have to develop the reasoning started for Figs. 94 and 95. This is done in Fig. 99 where we look at the layer kink (the lozengeshaped part in Fig. 95) from above. By only applying the condition of layer continuity, the following geometric relations are easily obtained, which describe the essential features of thin walls. Starting with Fig. 99a and remembering that the layer periodicity along the glass surface is dA, we obtain a relation between yand the layer kink angle p with the chevron angle 6 as a parameter. With dcld, = cos 6, this is writ-
639
ten (404) For small angles we can solve either for j3
or for y
This relation shows that y increases monotonously with p. Thus the larger the smectic layer kink, the more the wall will run obliquely to the layer normal. The fact that there is a physical upper limit for p (see below) explains why we never observe zigzag walls with a large inclination, y For ease in computation, when we make estimations Eq. (406) can more conveniently be written (407) Next we get the obvious relations between the width w of the wall and the width b of the lozenge or the length c of the cross section of the wall cut along the layer normal w = bcos(p- y) = csiny
(408)
as well as the width b of the lozenge from
c tan y = b cos6
(409)
But this width b is also related to the kink angle and the sample thickness L (see Fig. 99c, d and Fig. 94e). b sinp = Ltan6
(410)
We may now use Eqs. (408), (410), and (404) to obtain the relation between the wall thickness and the sample thickness
w = L cosy sin 6 sin P
640
2
Ferroelectric Liquid Crystals
y7:
L Y
Y (d)
(C)
Finally, we can express the angle o between the adjacent surfaces M and N in Fig. 40c by taking the scalar product of the corresponding layer normals u and v [marked in (a) and (d)]. This gives COSO
= cosp C OS S
(412)
Let us now extract the physical consequences contained in all these relations and start with the last one. As numerical examples we will use the data from Rieker and Clark [168] on the mixture W7-W82, for which the tilt angle 8 saturates at about 2 1O at low temperature, and the corresponding saturation value for the chevron angle Shas been measured as 18". To begin with we see that Eq. (412) sets an upper limit to the layer kink p, because of the requirement of uniqueness of the director n at each chevron interface. Figure 96a illustrates the fact
=
Figure 99. (a) Chevron structure like the one in Fig. 36, as seen from above. A thin wall of width w joins the region with a chevron bend in the negative z direction (left part of the figure) with another having the chevron bend in the positive z direction (right part of the figure). (b) Section through the wall cut along the z direction, (c) projection along the z direction, and (d) a cut parallel to the xz plane.
that where the two inclined layers meet they can share the director only to a maximum layer inclination of up to twice the tilt angle 8. At this point the two cones touch, and beyond it no uniqueness in the director field can be maintained. Thus o i n Eq. (412) has a maximum value equal to 2 6, which gives a maximum value, in practice only somewhat lower than 2 8, for the allowable layer kink p. This, .in turn, gives a limiting value for y. With the data for W7-W82 we get o=42", which gives (Eq. 412), p=39" and (Eq. 407) y= 15". The width of the wall in this case [Eq. (41 1)] is found to be roughly 0.5 L, i.e., half the sample thickness. This is thus the limiting case for an inclined zigzag in this material. Let us now assume that we observe another wall running at a 5" inclination to the layer normal. In this case Eq. (405) gives p=24" and Eq. (41 1) gives
2.8
FLC Device Structures and Local-Layer Geometry
15'1,'
J
r
i
I w = Wmax = L
w = 0.8 L. Finally, let us check a wall running exactly perpendicular to the layers (y= 0). In this case Eq. (404) shows /?to be equal to 6, hence p= 18", and w = L according to Eq. (411). This is the other limiting case with a wall of maximum width. Summarizing (see Fig. 100) we may say that a thin wall has its maximum width (equal to the sample thickness) when running perpendicular to the layers. The layer kink then has its minimum value, equal to 6, i.e., -- 8.Inclined walls are thinner, represent a higher energy density, and cease to exist when the layer kink approaches = 2 8,which sets the maximum value of obliqueness to about 15-20'.
Figure 101. Thin wall curving until it finally runs parallel to the smectic layers, then being a broad wall. The character changes from to >>> Oeq, and thus with a tilt that is larger than the equilibrium tilt angle. Because the layers are upright in the middle of a broad wall, the optical contrast between the two states is particularly high, and may be further enhanced if the width is smaller than the equilibrium value W corresponding to Eq. (413). As is evident from Fig. 102b, a broad wall is thicker the smaller the chevron angle 6. This is also most clearly expressed by Eq. (414). In contrast, the width of a thin wall is not very much affected by 6. However, it is useful to take a closer look at its behavior with varying '/. For this, let us go back to Fig. 101. We have already seen that the character of the wall changes from to >>>
J
-
2 2
c
c
2
-
c
c
\
\
UP
DOWN
Figure 107. Polarization switching taking place in the chevron plane at fixed polar boundary conditions. In this case the condition is one of high pretilt a for n with the P vector pointing into the liquid crystal from the boundary.
up or down state, respectively. These halfsplayed states, which are most often just called twist states, occur in both C1 and C2 structures and are then abbreviated C 1T and C2T, in contrast to the uniform states C l U and C2U. The twist states are normally bluish and cannot be brought to extinction, hence they give very low contrast in the two switchable states. Moreover, due to the different sign of V . P , there will be an unsymmetric charge distribution between adjacent areas switched to the up and the down state, tending to burn in any already written static picture. For high P , materials, even the electrostatic energy has to be taken into account which, together with the elastic energy, may tend to shift the chevron plane to an unsymmetric state as illustrated to the right of Fig. 107, in order to relieve the high local energy density. Such a state is then completely unswitchable (monostable). It is evident that, whether we choose C1 or C2, a twisted state has to be avoided. In their first analysis of C1 and C2 structures, Kanbe et al. [ 1701 put forward the essential criteria for their stability. Both are allowed at low pretilt a. Whereas C1 can exist at high pretilts, the cone condition ( n must be on the cone) cannot be fulfilled for C2 if a i s larger than (6-S), as illustrated in
2.8
FLC Device Structures and Local-Layer Geometry
647
Figure 108. Stability conditions for chevrons of C 1 and C2 type (after [170])
Fig. 108. This means that for a#O it cannot be fulfilled at the phase transition point S m A j S m C when 8 and 6 are both zero. Hence the chevron that is first created when the sample is cooled down to the SmC phase is always C1. However, C1 is not stable when the tilt angle 8 increases. Therefore C2 appears together with C1 at lower temperatures and the sample is marred by zigzag defects. One of the reasons for this is that C 1 has a higher splay - twist elastic deformation energy, which increases with increasing 8. Another reason is that the director is more parallel to the rubbing direction in the C2 case; it has to split more to fulfill the cone condiditon for C1. This effect will favor C2 under strong anchoring conditions. Finally, as pointed out in the cited Canon paper [170], while a gradual transition C 1 - + C2 to an almost chevron-free state might be observed on cooling, the once such-created C2 state tends to be quite stable if the temperature is subsequently raised; it is almost necessary to go back to the SmA* phase in order to recreate C 1. Thus C2 is stable over a much broader range of temperature than C1. All this speaks for the C2 state, in spite of the considerably lower effective switching angle between its memorized states. Several FLC projects have also been based on C2 (JOERUAlvey, Thorn/CRL, Sharp). The fact that bistable switching only takes place through being latched by the chevron surface is compensated for by a number of advantages. First of all, the surface alignment is relatively simple, as the outer surfaces do not switch but only work togeth-
er with the chevron. It is also much simpler to avoid twist states for C2 than for C1. This is easily seen from Fig. 106. Here the structure has been sketched for a=(6-8) such that the director is exactly along the rubbing direction and strongly anchored. As already pointed out, the nonchevron version of this is shown in Fig. 90. In practice it is found [ 1711 that a medium pretilt a of about 5" is convenient for achieving this cone surface condition, whereas a values near zero give both C2U and C2T. For several materials, (8-6) is of the order of about 5"; however, observations have been made [ 1721of chevron-free C2U structures for a= 5", although (8-6) seems to be only about 1.5", thus violating the stability condition for C2. Whether this is due to a change of 6 under driving conditions (6 decreasing towards QBS case), is unclear. A further advantage with the C2 structure is that zigzags induced by (light) mechanical shock, which causes a local transition to C1, easily heal out by themselves in the C2 structure [170], whereas the opposite is not true. Fortunately, C2 also has a substantially higher threshold to so-called boat- wake defects, which appear on the application of a very high voltage. The main disadvantages with the C2 structure are obviously the low optical contrast in the true memorized state and the quite high voltages applied so far in the electronic addressing of the cell, when the socalled z - V ~addressing ~ ~ modes are used. This is also combined with a high power consumption. Thus the switching pulse is
648
2
010
~
Ferroelectric Liquid Crystals
-
1
5 10 15 AC voltage vd [v]
Figure 109. Effective switching angle in a C2 structure as a function of the amplitude V, of the applied data pulses for the so-called Malvern-3 scheme. The switching pulse amplitude is 30 V. V,= 0 corresponds to the memorized states as latched by the chevron Interface. The P, value of the material is 4.4 nC/cm2 (after 171).
used to accomplish latching in the chevron plane, whereas the bipolar AC data pulses are used to force the director as much as possible into the in-plane condition between the chevron surface and the outer surfaces (see the lower part of the right of Fig. 106). In this way the effective switching angle can be increased from a typical value of 14", corresponding to the memorized state, to the more significant value of 24" (Fig. 109), when data pulses of 10 V amplitude are applied [171]. This higher switching angle thus corresponds to what we might call "forced memorized states".
2.8.7 The FLC Technology Developed by Canon As indicated above, the first Canon prototype from 1988 had a=O giving a hemispherical viewing angle, which was a complete novelty at that time. However, the contrast (- 7:l) was only good enough for a monochrome display. Moreover, the C 1 structure could not be made sufficiently stable at low temperatures. To solve this problem the team decided not to abandon C1
in favor of C2, but instead to pursue the much more difficult road to go to a very high pretilt. The idea was to squeeze out C2 completely from the material because its cone condition a c ( 8 - 6 )can no longer be satisfied. Thus, by going to the rather extreme pretilt of 18" (a technological achievement in itself), Canon was able to produce zigzag-free structures, as the layers can only be folded to C1 chevrons. A considerable development of FLC materials also had to take place, among others, mixtures with (6-6) values between 3 and 5". In this way C2 could be successfully suppressed under static conditions, for instance, in a material with 8 = 15" and 6 = lo", which is not surprising, because a i s not only larger than (0-6) but is even larger than 8 alone. However, it turned out [173] that C2 would appear to some extent under dynamic, i.e., driving conditions. This obviously cannot be explained by assuming that 6 gets smaller under driving conditions. Thus, it seems that the stability criterion can hold at most under static conditions. This indicates that the underlying physics of the CUC2 structure problem is extremely complex. In order to avoid C2 at any temperature, several parameters had to be optimized, such that both 0 and Sshowed only a small variation as functions of temperature. The obvious advantage of working with C 1 is that the switching angle is high, at least potentially, between the memorized states. This is particularly true if the cell switches not only at the chevron surface, but also at both outer boundaries. However, surface switching does not normally occur at polymer surfaces with a low pretilt (although it does at SiO-coated surfaces). This is another advantage of going to high pretilt. The situation may be illustrated as in Fig. 110. At high pretilt the surface state not only switches, but the switching angle may be just as large as permitted by the cone angle of the
2.8
(a)
(b)
FLC Device Structures and Local-Layer Geometry
(C)
Figure 110. Switching angle under different conditions. (a) To the left QBS geometry is supposed. If the surface cannot be switched (for instance on polyimide), this will lead to a strongly twisted state. In the middle (b) a chevron geometry is assumed. If the surface is nonswitchable, it leads to a smaller twist than i n (a), if it is switchable, the switching angle is less than in a switchable version of (a). With high pretilt (c) a polymer surface normally switches and the switching angle may be optimized, similar to the situation in (a).
material used. This cone angle is limited by the possible appearance of C2 at low temperatures and also by the fact that the twisted state C1T would appear at higher 8. Thus 8 is limited, quite below the ideal performance value, in the C 1 state as well as in the C2 state. Whether the twisted state appears in the C1 case or not is, however, not so much determined by the tilt 6, but preeminently by the polar strength of the surface condition. It is therefore important to realize that this influence may be at least partly neutralized by a suitable cross-rubbing, i.e., obliquely to what is going to be the direction of the smectic layer normal. The idea is quite simple, but can give rise to a number of different interpretations, e.g., suppose that the lower surface demands the P vector to point into the liquid crystal. However, if n at this surface is tilted in a direction clockwise (seen from above) relative to the layer normal, then this n direction corresponds to the same direction of P (for P - Am=9Oo(YO;a>] (YO;a ) + Am=90° (v0;a>
(7)
3.2 Origin of Antiferroelectricity in Liquid Crystals (a)
0.39
0.38 0 0)
m
$ 0.37 In 2 0.36
0.35' 0
'
"
90
'
.
'
180
.
'
'
270
'
'
360
67 1
from that in SmC*, provided that a does not change during the field-induced phase transition from SmCi to SmC*. Using the experimentally obtained degree of polarization, D ( y o )= -0.03, and assuming that the degree of biasing or hindrance is a =: 0.3, we determine that yo= 135" (45") for the partially racemized ( S ) ( ( R ) )enantiomer in the SmCi phase. If a = 0.2, then yo=100" (80"). In the SmC* phase of (S)-MHPOBC at 120°C however, the experimentally obtained value is D (yo)= 0.05, and hence we obtained yo= 175" (5") for a =: 0.3 and yo= 165" (15") for a = 0.1.
0
3.2.3 Spontaneous Polarization Parallel to the Tilt Plane
-0.05
-0.10 0
30
60
90
vo(deg.)
Figure 8. (a) Normalized absorbance versus the polarizer rotation angle calculated for SmC* and SmC; for radiation incident along the smectic layer normal by using the distribution function given in equations (3) and ( 5 ) . The molecular tilt angle and the degree of hindrance are assumed to be 8=25" and n = 0.2, respectively, and w = 0 is the tilt-plane normal. (b) Degree of polarization versus the most probable orientation for several values of a .
for various values of a. Note that positive and negative signs of D ( y o ;a) correspond, respectively, to the out-of-phase and inphase cases with respect to the phenyl peak. As the molecules are tilting from the smectic layer normal along which the IR radiation is incident, the C=O peak shows the out-of-phase angular dependence even when the rotational motion is free. Nevertheless, the conspicuous out-of-phase and in-phase angular dependencies in SmC* and SmCi, respectively, given in Figure 7, clearly indicate that yo in SmCi differs
We can conclude, therefore, that the chiral C=O has a tendency to lie on the tilt plane in SmCi, while it assumes a fairly upright position in SmC*. For several years antiferroelectricity has been modelled as arising from the pairing of transverse dipole moments at the smectic layer boundaries [17, 181. Although the chiral C=O in question is not at the end of the molecule, the unexpected bent shape of MHPOBC in its crystal phase just below SmCi, as revealed by X-ray crystallographic studies, has been considered to support the pairing model [ 18, 191. The model based on the pairing of transverse dipole moments is physically (or, rather, chemically) intuitive in relation to an understanding of the stabilizaton of antiferroelectric SmC;. However, because yois as large as 45" or more, as concluded above, it is not reasonable to ascribe an essential role to the pairing alone, and we must look for other causes for the stabilization of SmCi ~151. The symmetry arguments that Meyer used to develop his speculation about ferroelectric SmC* may help us to clarify anti-
672
3 Antiferroelectric Liquid Crystals
ferroelectric SmCi. There are two types of two-fold axis: one at the layer boundary is parallel to the X axes in the tilt plane (C2X,b), and the other is located in the centre of the layer, perpendicularly ( C , , ,) (Figure 9). In the pairing model, the C2y,caxis is the only one taken into consideration explicitly; the spontaneous polarization along C,,, , P,, alternates in sign from layer to layer [ 16- 18,20,21].However, Cladis andBrand [22, 231 insisted on the importance of the C,,,, axis in the tilt plane, and the above
Z
c2y, b
(b)
(a) Figure9. The two types of symmetry axes in unwound SmC; and SmC*. (a) Unwound SmC; has two-fold axes, C 2 , b and C2y,c,at the smectic layer boundary and at the smectic layer centre, respectively. (b) Unwound SmC* has two-fold axes, c , , and C2Y,c,at the smectic layer boundary and at the smectic layer centre, respectively.
(a)
Smc'
SmC;
conclusion about biased or hindered rotational motion of the chiral C=O supports their indication. The chiral C=O in question, although not in the core, is not at the end of the molecule either. Nevertheless, it is natural to consider that the in-plane spontaneous polarizations exist at smectic layer boundaries; they are parallel to the tilt plane along &X,b in SmCi, but are perpendicular to the tilt plane along C,,,b in SmC*. Their magnitude in SmCi may not be equal to that in SmC*. Note that C 2 y , b exists, as well as C2y,c,even in SmC*, as illustrated in Figure 10. An unanswered question is whether or not the Coulomb interaction among the permanent dipoles plays an important role. If the interaction is considered to result in in-plane spontaneous polarizations parallel to the tilt plane Px,we can designate this viewpoint as the Px model. For the stabilization offerroelectric SmC*, the Coulomb interaction among the permanent dipoles has not been considered essential, as insisted upon initially by Meyer [lo]. It is also possible that this interaction does not play any essential role in the stabilization of SmCi [24]. A SmC'
Figure 10. In-plane spontaneous polarizations in SmC; and SmC*. (a) Previous dipole pairing model. Only C2y,c,and hence the in-plane spontaneous polarization along the tilt-plane normal P,, are taken into consideration. (b) The present P, model. Both C2,,, and C2y.b are considered; hence in-plane spontaneous polarizations are considered to exist at the smectic layer boundaries, P, in SmCi and P , in SmC*.
3.2
Origin of Antiferroelectricity in Liquid Crystals
tempting feature of the P, model is, however, the possibility of explaining the herringbone structure, where the tilt planes in adjacent smectic layers are parallel to each other; we could not devise any other simple explanation for the structure. When permanent dipoles of equal magnitude are distributed on a plane, they may align in the same direction in order to minimize the Coulomb interaction energy, resulting in the in-plane spontaneous polarization parallel to the tilt plane Px. The sign of Px alternates from boundary to boundary; this fact also contributes, through fluctuation forces, to minimizing the Coulomb interaction energy between P , planes 125,261. Let us consider three dipoles which are located at the apices of a regular triangle with one side d, and oriented in the same direction and sense as illustrated in Figure 11. When another dipole approaches along a line perpendicular to the triangle and passing through its centre, it will align parallel to the three dipoles if zdl&, in order to minimize the Coulomb interaction energy. Although, in the pairing model, we have to invoke rather ad hoc assumptions (i.e. the local spontaneous optical resolution and the conformational chirality), the present P, model can naturally explain the existence of SmC,; the in-plane spontaneous polar-
Z
k
Figure 11. Illustration of three dipoles at the apices of a regular triangle with one side d and another dipole approaching on the Z axis (see the text for details).
673
ization parallel to the tilt plane P, is independent of chirality [22, 23, 271 and emerges even in racemates and non-chiral swallow-tailed compounds with two terminal chains of equal length [24].
3.2.4 Obliquely Projecting Chiral Alkyl Chains in SmA Phases The above discussion shows that it is most probable that in-plane spontaneous polarizations emerge at smectic layer boundaries and stabilize antiferroelectricity in the SmCi phase. In fact, Jin et al. [28] have quite recently shown unambiguously that the chiral chain projects obliquely from the core, even in SmA* and hence is considered as precessing around the core long axis; the angle between the chiral chain and core axes is more than the magic angle 54.7'. Such a bent molecular structure may allow the permanent dipoles in adjacent layers to interact adequately through Coulombic forces. By deuterating the chiral and achiral alkyl chains separately, Jin et al. observed the IR absorbance as a function of polarizer rotation angle for the CH, and phenyl ring stretching peaks in the SmA* phase of MHPOBC in a homogeneously aligned cell. Their results are summarized in Figure 12. Because of the selective deuteration, Jin et al. were able to obtain information about the achiral and chiral chains independently. The angular dependence of the CH, stretching peaks, both asymmetric (2931 cm-') and symmetric (2865 cm-'), shows a marked contrast between the achiral and chiral chains. In the achiral chain it is out of phase with that of the phenyl ring stretching peak, while in the chiral chain it is in phase. Moreover, the degree of polarization is very small. Jin et al. calculated the normalized absorbance by assuming an average transition di-
674
3
Antiferroelectric Liquid Crystals
x 0.4
pole moment of the CH, stretching vibrations, which is perpendicular to the average chain axis and is rotating freely about it, and by neglecting the sample anisotropy. Several necessary angles are defined as depicted in Figure 13: a is the angle between the chain and core axes; is the azimuthal angle of the chain axis around the core axis; I+Y is the angle of rotation of the transition dipole moment about the chain axis; and 13 is the tilt angle in SmC* which is zero in SmA*. The relationship between the absorbance and the polarizer rotation angle is given by
x
0.2
b
0'01'80
' -90' ' ' ' $0 ' ' lA0 Polarizer rotation angle,w (deg)
0
1.5
0
2x 2x
A ( w ' ; a ) =1G J
J f(x>
0 0
. (sin o' cos x sin I,Y - sin o'cos a sin I+Y (8) - cosw' sin a cos I+Y),d x d t,u
-180
-90 0 90 180 Polarizer rotation angle,w (deg)
Figure 12. Absorbance versus the polarizer rotation angle in SmA* measured at 130°C using a homogeneously aligned cell. (a) 4.0 pm thick cell, (R)-MHPOBC-d,, (enantiomeric excess 30%); (b) 9.6 pm thick cell, racemic MHPOBC-d,, . (0)Phenyl ring stretching; (A) CH, asymmetric stretching; ( 0 )CH, symmetric stretching.
Here the distribution function in SmA* is f ( x )= 1 4 2 ~ )the ; chain axis rotates freely about the core axis because of the uniaxial SmA* symmetry. Figure 14 shows the calculated normalized absorbance in SmA*. At the magic angle a = c0s-l ( U 6 ) = 54.7", the absorbance does not depend on the polarizer rotation angle A (o; a) = 1 / 6 . The aforementioned inphase and out-of-phase relationships corre-
0.6 a, t
2 0.4 P
n u N ._ 1 0.2 E
b
2
0.0
Figure 13. A model of the transition dipole moment for the CH, stretching vibrations, situated on the alkyl chain and freely rotating about the chain axis. The chain itself makes an angle a with the core axis, and precesses around it.
-180
-90 0 90 180 Polarizer rotation angle,w'(deg)
Figure 14. Normalized absorbance versus the polarizer rotation angle in SmA*, calculated for various values of a.
3.3 Thresholdless Antiferroelectricity and V-Shaped Switching
spond to a> 54.7" and a < 54.7", respectively. In this way, as shown unambiguously in Figure 12(b), the chiral chain in the SmA* phase projects obliquely, precessing around the core axis; the angle between the chain and core axes is more than 54.7", but is probably close to the magic angle because of the fairly small observed degree of polarization. Figure 12 (a) also clearly shows that, in this case, the other terminal chain in SmA* makes an angle of a < 54.7" with the core axis. By taking into account of the fact that not only the average transition moment, but also the average chain axis itself, is actually fluctuating, the appreciable degree of polarization observed may indicate that the angle is greatly different from the magic angle and rather close to zero. Jin et al. also analysed how the chiral chain precession becomes biased or hindered in SmC*, and suggested that the biasing or hindering may occur to the tilt plane normal = 0'). Although no experimental study of the biased or hindered precession in SmCi has yet been made, it is tempting to speculate that the biasing or hindrance may occur to the tilt plane (xo= 90"). As described in Sec. 3.2.1 and 3.2.2, the biased or hindered direction of the C = O group near to the chiral carbon atom in SmC* is quite different from that in SmCi. Anyway, we can safely conclude that in a prototype AFLC, MHPOBC, the molecules have a bent structure so that the C=O groups near the chiral carbon atoms, i.e. the permanent dipoles, in adjacent smectic layers can adequately interact through Coulombic forces to stabilize the antifen-oelectric SmCi (SmCA), and that in-plane spontaneous polarizations are produced at layer boundaries along the two-fold C2X,baxes which are parallel to the molecular tilt plane. Figure 15 presents an illustrative Px model based on bent shaped molecules.
(xo
675
n
Figure 15. An illustrative Px model based on bent shaped molecules.
3.3 Thresholdless Antiferroelectricity and V-Shaped Switching 3.3.1 Tristable Switching and the Pretransitional Effect Ordinary antiferroelectricity in liquid crystals results in tristable switching, which is the electric-field-induced transition between antiferroelectric SmCi and ferroelectric SmC* and has the characteristic d.c. threshold and hysteresis shown in Figure 16 [29-351. When a weak electric field is applied perpendicularly to the page, the dielectric anisotropy unwinds the helicoidal structure of SmCi and the tilt plane of the unwound SmCi becomes perpendicular to the page; all in-plane spontaneous polarizations, p b , have the same magnitude, but point upwards and downwards alternately. After the phase transition into SmC*, the molecules are parallel to the page, tilting uniformly either to the right or to the left; all p b have the Same magnitude and point uniformly either downwards or upwards. During the pretransitional effect, the Pbstay
676
3
Antiferroelectric Liquid Crystals
Electric field
m
-
I
i Electric field H -Bh
F
L -€th
L Bh
H
Bh
Pretransitional AF Pretransitional F
Figure 16. Ordinary tristable switching; (a) apparent tilt angle versus applied electric field; (b) optical transmittance versus applied electric field; (c) molecular tilt directions during the pretransitional effect and in the antiferroelectric (AF) and ferroelectric (F) states.
perpendicular to the page, but their magnitude and sense may change alternately from boundary to boundary, so that a slight net spontaneous polarization, either downward or upward, may emerge. The molecular tilting changes slightly from the unwound SmCi directions towards either of the unwound SmC* directions, and the tilt planes in neighbouring layers are no longer parallel to each other; Pbat a boundary is neither par-
allel nor perpendicular to the tilt plane in either of the neighbouring smectic layers. The unwinding process and the pretransitional effect may be virtually indistinguishable if the helicoidal pitch is short and the threshold field is low, or if the applied field changes very quickly. Because of the antiferroelectric unwound herringbone structure, however, the unwinding process in SmCi alone does not induce any apparent tilt angle; the small induced apparent tilt angle in SmCi clearly results from the pretransitional effect and increases almost linearly with the applied field from zero up to the threshold. The pretransitional effect causes a parabolic response in the optical transmittance versus applied electric field in a homogeneously aligned (smectic layer perpendicular to substrate) cell between crossed polarizers, one of which is set parallel to the smectic layer normal. In order to see more clearly the temporal behaviour of electric-field-induced phase transitions between SmCi and SmC*, it is useful to observe the optical transmittance when the applied field is changed stepwise across the threshold, &,h or &$the results for which are summarized in Figure 17 [36]. The antiferroelectric to ferroelectric (AF+F) switching (Figures 17(a) and 17(c)) critically depends on E h , but not on E l ; it contains two components, fast and slow. The fast component increases at the expense of the slow component as E h becomes higher. The slow component is seriously affected by E h and the response time representing the transmittance change from 10% to 90% is approximately proportional to ( E h - &,h)4. The reverse switching (F+AF) (Figure 17b and d) critically depends on E,, but not on E,; it consists of the slow component only. The response time is approximately proportional to (Eth,l-E1)7.A hump is observed when E, becomes sufficiently negative, i.e. when E1S-1.6 V . pm-l.
3.3 Thresholdless Antiferroelectricity and V-Shaped Switching
0
200
0
100
LOO 600 Time ( ~ s )
800
P m C c
'E 2
e
I-
200
300
LOO
Time (us1
0
200
LOO 600 Time ( ~ s l
800
: g
2 . L ,
The observed ratio of the fast component to the total change in transmittance was confirmed to be almost consistent with the estimated ratio from the maximum induced apparent tilt angle during the pretransitiona1 effect and the tilt angle after the transition to SmC*. Therefore, the fast component represents the pretransitional effect, which involves no macroscopic domain formation; when observing the effect through a microscope, the viewing field changes uniformly. The slow component represents a change due to the SmCi-SmC* (AF-+F) phase transition. Stroboscopic microscope observation has revealed that the transition always accompanies the movement of the domain (phase) boundaries of characteristic shape, as shown in Figure 18 [36, 371. The hump observed in Figure 17(d) originates from the F + F switching (process) which occurs temporarily. Since the domain boundary propagates rather slowly, so that the middle part of the cell stays ferroelectric for a while (Figure 18), the director in this part may not switch directly from F to AF, but tends to rotate on the cone from F (+) to F(-), or vice versa, and then to return to AF, if E, is negative. Therefore, the apparent tilt angle overshoots into the opposite side, crossing x=O".The overshoot can actually be seen in the stroboscopic photomicrographs shown in Figure 18 (b and c), which were obtained by choosing the crossed polarizer directions in such a way that F(+) becomes dark. The bright (green) area in the middle is F(-) and the intermediate (blue) area appearing at the electrode edges (the right and left sides of the photomicrographs) is AF. The boundary of characteristic shape between the AF and F(-) regions moves towards the middle. The dark (black) area observed in Figure 18 (b and c) clearly indicates that the F(-) +F(+) process really does occur; although not shown in Figure 18, the F(+)+AF process follows
x
P
I-
0
LO
80 120 Time (PSI
160
Figure 17. Optical transmittance change with crossed polarizers as a function of time after changing the electric field stepwise. Measurements were made using a 2.5 pm thick, homogeneously aligned cell of (R)-TFMNPOBC at 80°C by setting the polarizer in such directions that the antiferroelectric state becomes dark. The stepwise changes are (a) EFX= 0 V pm-' -+ Eh = 11.2- 15.2 V pm-'; (b) Eh=11.2-15.2Vpm-' -+ Ef'"=OVpm-I; (c) E l = -3.2-0.8 V pm-' + E F = 12 V pm-'; (d) E z X= 12 V pm-' + El=-3.2-0.8 V pm-I. U refers to uniform ferroelectric and 3rd to helicoidal antiferroelectric.
677
678
3
Antiferroelectric Liquid Crystals
Figure 18. Stroboscopic photomicrographs recorded at (a) 0.145 V (253 ms), (b) 0.290 V (256 ms) and (c) 0.339 V (257 ms) by linearly changing the voltage from -48.4 V (0 ms) to 48.4 V (500 ms) in a 4.6 pm thick, homogeneous cell of (S)-MHPOBC at 80°C. One of the ferroelectric states is chosen to be dark, overshooting to which is actually observed as clearly seen in (b) and (c). UI and Ur refer to ferroelectric states tilted uniformly to the left F(-) and to the right F(+), respectively, and 3rd refers to helicoidal antiferroelectric state (A).
afterwards, and the whole area becomes blue. These switching processes are also identified in the polarization reversal current measured by applying the triangular electric field (Figure 19). If the temporal rate of change of an applied electric field dEldt is not large, two bell-shaped peaks may typically appear, which correspond to the two processes F(-)+AF and AF+F(+). As dEldt increases, the F(-) +AF bell-shaped
J
0
200
LOO Time (ms)
600
Figure 19. Polarization reversal current in a 4.6 pm thick cell of (S)-MHPOBC at 80°C by applying a 548.4 V triangular wave voltage of 1 Hz. Overshooting to the other ferroelectric state is observed as a conspicuous sharp peak near 0 V followed by a backward current peak.
peak shows a cusp; with further increase in dEldt, the cusp becomes a conspicuous sharp peak, indicating the F(-) + F(+) process, and even the negative polarization reversal current, which results from the backward process F(+) + AF, is observed (Figure 19). Note that the small cusp was observed in the initial measurement made by Chandani et al. [29], although its significance was not realized at that time. In this way, it becomes clear that, phenomenologically, the pretransitional effect, as well as the phase transition between SmCz (AF) and SmC* (F), can be described by the rotation of the azimuthal angle 4 ; tilted molecules in alternate smectic layers rotate in the same direction, but those in the neighbouring smectic layers rotate in the opposite direction [38,39]. Let us consider what happens during biased or hindered rotational motion about the molecular long axis. In Sec. 3.2.3 we concluded that the biased or hindered direction yoin S m C i (I,$") is substantially different from that in SmC* (vz); more generally speaking, tytFand yc may be regarded as representing the rotational states about the molecular long axes in S m C i and SmC*, respectively. From this
3.3 Thresholdless Antiferroelectricity and V-Shaped Switching
point of view, the pretransitional effect is considered as a slight change in yofrom ytF towards .:yt Note the essential point that, not only the magnitude and direction of Pb at a boundary, but also the molecular tilt directions (azimuthal angle) in the neighbouring smectic layers of the boundary, are determined by the molecular rotational states. We have tried to observe this change in yo directly by means of polarized fourier transform IR (FTIR) spectroscopy, but have so far been unsuccessful. In the rotation of the azimuthal angle @ during the F(-) -+ F(+) process, or vice versa, the molecular rotational states have been observed to some extent by means of polarized dispersive IR spectroscopy in the submicrosecond regime [40].
3.3.2 Pretransitional Effect in Antiferroelectric Liquid Crystal Displays Antiferroelectric liquid crystal displays (AFLCDs) have been relevant to flat panel displays since the discovery of AFLCs. Their characteristic features can be summarized as follows: In the SmCi phase stable at zero electric field, the optical axis is almost parallel to the rubbing direction. This favours the ease and stability of alignment of the molecules. Nakagawa et al. [41] have shown that the smectic layer normal in SmA* does not coincide exactly with the rubbing direction, deviating clockwise by several degrees for (S)-MHPOBC relative to the substrate. There is no permanent spontaneous polarization, but polarization emerges when needed. Hence the switching may not be hampered by the so-called ‘ghost effect’ or the unnecessarily strong, difficult to erase, memory effect.
679
The switching is characterized by the distinctive d.c. threshold and hysteresis, which allow matrix addressing of large capacity displays, if use is made of a positive or negative bias field. The use of a symmetrical bias field, positive in one frame and negative in the next, also ensures an ideal viewing angle, much wider than that in SSFLDs, although the pretransitional effect that induces a small increase in the apparent tilt angle in SmCi may slightly decrease the contrast ratio. The quasi-bookshelf smectic layer structure is realized and the twisted state of the director alignment is inherently avoided. Hence the relatively high contrast ratio of 20-30 is easily obtained. The switching response time may become as short as 50 ps or better under acceptable driving conditions by optimizing the materials used. Stripe domains parallel to the smectic layer appear from the electrode edges quite regularly during switching, and the speed of movement of their boundaries depends critically on the field applied. This fact is utilized in controlling the grey scale, in combination with the distinctive d.c. threshold and hysteresis. Self-recovery from alignment damage caused by mechanical and thermal shocks is observed during operation. This results from the reversible layer switching between the bookshelf and the quasibookshelf structures, which occurs in the network of characteristic stripe defects parallel to the layer normal. Because of these attractive characteristics, much attention has been paid to AFLCDs, and two prototype AFLCDs with passive matrix (PM) addressing have been developed and exhibited. One is a 6-inch, video rate, full-colour display by Nippondenso
680
3 Antiferroelectric Liquid Crystals
Co., Ltd. [42], and the other is a 5.5-inch, VGA monochrome display by Citizen Watch Co., Ltd. [43], both of which have an extremely wide viewing angle with a relatively high contrast ratio of 30. What prevents AFLCDs from achieving much higher contrast ratios is the pretransitional effect, which manifests itself as a slight increase in the transmittance below the threshold (see Sec. 3.3.1). Ironically, during attempts to develop materials that can suppress the pretransitional effect, we have encountered materials which show a large pretransitional effect and, at the same time, a remarkable decrease in the threshold field strength. Hence we came to consider that enhancement of the pretransitional effect, in combination with the use of active matrix
(AM) addressing, is another way of endowing liquid crystals with attractive display characteristics [34,35]. Taking some compounds and their mixtures as examples, let us show in a concrete way how the threshold becomes lower as the pretransitional effect is enhanced. Table 2 [44,45] shows that, in a series of compounds with the general molecular structure, an ether linkage and its position in the chiral end chain influence the threshold field Eth,h in a subtle way. When the ether linkage is located at the chain terminus, &h,h becomes higher than that of the reference substance of equal chain length but having no ether linkage. As the linkage moves towards the core, E t h , h becomes lower and, finally, SmC* (F) appears instead of SmCi (AF).
Table 2. The effect of an ether linkage position on the threshold field strength.
3.4
15.7 6.5
Table 3. The effect on the contrast ratio of substituting the phenyl rings in 1 (1 = 1, m = 4, n = 1, la) with fluorine.
c,1H23o&CO1&
la
CO2-C*H(CF3)C,H@CH, y2
Contrast ratio a 36
27 12 31 28 10 a Measured at 20°C below the SmC* (SmA*)-SmCi phase transition temperature by using a 1.7 pm thick cell and applying a 0.1 Hz, k 4 0 V triangular wave voltage.
3.3 Thresholdless Antiferroelectricity and V-Shaped Switching
Blending further reduces E t h , h as illustrated in Figure 20 [35].Furthermore, the pretransitional effect can be enhanced by suitably substituting phenyl rings with fluorine [35]. In order to evaluate the pretransitional effect we measured the contrast ratio TF/TAF just below E t h , h ; the smaller the contrast ratio, the larger the pretransitional effect. Table 3 summarizes the results, indicating that substitution at Yl decreases the contrast ratio and hence enhances the pretransitional effect most remarkably. 100
I SmA
!
I-
'/020 4 0
L
0
20
40 60 80 Mixing ratio(%)
100
In this way we can enhance the pretransitional effect and reduce the threshold Eth,h.An unanswered question is that of whether or not we can prepare a system where the threshold becomes zero, so that the pretransitional effect prevails and the AF-F phase transition occurs continuously. Surprisingly, it was not difficult to realize this property in a three-component mixture; the field-induced F-AF-F phase transition looks like a V-shaped switching (Figure 21) [34, 351. Hereafter we will call this property 'thresholdless antiferroelectricity'. Although no detailed investigations have yet been made, the response of the system to an applied stepwise field is rather fast, as illustrated in Figure 22. When we observed the switching by means of polarizing optical microscopy, the visual field varied uniformly and continuously without showing any irregularities, indicating the boundary movement characteristic of tristable switching or the disclination lines caused by the helicoidal unwinding of a ferroelectric liquid crystal. The V-shaped switching is totally different from tristable switching and helicoidal unwinding.' We still need to optimize the V-shaped switching characteristics by further developing suitable materials, but the results obtained with the materials available so far appear to indicate the possibility of realizing:
F C , H I ~ OW
C
0
2
G CO2- & W C & O C , H s
3
Figure 20. The phase diagram of a binary mixture of 2 and 3 and the threshold field versus mixing ratio. The phase boundary between the AF and F regions is parallel to the temperature axis and hence the stability of the AF phase becomes equal to that of the F phase in the binary mixture at this particular boundary mixing ratio. As the mixing ratio approaches the boundary, decreases, although it could not be confirmed that it diminished to zero.
68 1
0
0
a tilt angle of more than 35"; a field of less than 2 V ym-' to complete the AF-F switching; light transmission that is almost linear with respect to the applied field, and free from hysteresis; a AF-F switching time of less than 50 ps; a contrast ratio as high as 90; and an extremely wide viewing angle of more than 60".
682
3 Antiferroelectric Liquid Crystals
1.o
0.8
-
0.6
c .-
0.4
! +
0.2 0.0 -6
-4
0
-2
2
4
6
Electric field (V/pm)
Figure 21. The V-shaped switching observed at 25°C by applying a 0.179 Hz, +7 Vpp triangular wave in a 1.4 pm thick, homogeneous cell of a mixture of 4, 5 and 6 in the mixing ratio 40 : 40 : 20 (wt%), which has a tilt angle 0 of 35.7'.
3.3.3 Langevin-type Alignment in SmCg Phases
1
Transmittance-.
\
Time Figure 22. Temporal behaviour observed at 25 "C by applying a +4 V rectangular wave to the same cell as that in Figure 21: (a) 2 msldivision; (b) 50 psldivision.
A small cell, although very primitive, has actually been made, and demonstrates these attractive display characteristics.
We will begin with speculation about thresholdless antiferroelectricity in terms of the molecular rotational motion which plays an essential role in the emergence of inplane spontaneous polarizations in liquid crystals. We concluded in Sec. 3.2.3 that the biased or hindered direction yo in SmCi (ykF)is substantially different from that in SmC* (y:); more generally, ytF and y: may be regarded as representing the rotational states around the molecular long axes in SmCz and SmC*. We also concluded in Sec. 3.3.1 that the pretransitional effect is considered as a slight change in yo from ykF towards y:. Based on these conclusions, we speculate as follows. In the phase showing thresholdless antiferroelectricity, the biased or hindered rotational motions (more generally, the molecular rotational states in SmC; and SmC*) characterized by ytF and y: have nearly the same energies and, furthermore, the barrier between them diminishes, so that the states characterized by any yobetween yeFand are equally thermally excited at zero electric field.
683
3.3 Thresholdless Antiferroelectricity and V-Shaped Switching
As noted in the previous section, this arbitrariness in tyo makes the molecular tilting direction uncorrelated between adjacent layers. Figure 23 illustrates a simplified model of the phase SmCg that shows thresholdless antiferroelectricity. The director tilting is uniform and has constant polar and azimuthal angles in a smectic layer, but its azimuthal angle varies randomly from layer to layer. In-plane spontaneous polarizations exist at the smectic layer boundaries, their orientation being random and their magnitude varying from boundary to boundary. Random tilting of the smectic layers ensures the disappearance of the net spontaneous polarization, but an applied electric field induces it according to a Langevintype equation [46]. The difference from the ordinary Langevin equation is that we are dealing with very large effective dipole moments of variable magnitude because many molecules interact cooperatively and produce in-plane spontaneous polarizations, and their rotation is restricted in the twodimensional space.
The light transmittance was obtained by assuming that the effective dipole moments have the same magnitude peffand follow the same distribution as the c director. The two-dimensional and three-dimensional Langevin alignments are given by 7l
exp (x cos 4) cos @ d@
(9)
(cos@)=O exp(x cos@>d# 0
and
respectively, where x = peffE / ( k T ) ( k is the Boltzmann constant and T is temperature) [46]. Figure 24 shows the numerically calculated results; the two-dimensional case aligns more steeply than the three-dimensional one. In a rough approximation, it is convenient to use an apparent tilt angle (cos #) = cos oaap
(1 1)
We will further assume that the index of the ellipsoid is uniaxial with its eigen values, nl, and n,, independent of the applied field.
1.o
_______*----
0.8 c
$ 0.6
2-dimensional 3-dimensional
E
01
0.4 -E
E=O
+E 0.2
0 Langevin-type alignment
Random from boundary to boundary
Langevin-type alignment
Figure23. A simplified model of the phase that shows thresholdless antiferroelectricity.
'
0.0
0
I
5
10
15
20
X
Figure 24. The two- and three-dimensional Langevin alignments.
684
3 Antiferroelectric Liquid Crystals
$
;::m
+
0.2
X
X
; (m 0.6 ._
Figure 25. Calculated switching behaviour for cell thickness d = 1.55 pm,n,= 1.5, wavelength of incident light L=500 nm: (a) rill= 1.65, And/L=0.45; (b) n i l =1.71, And/L=0.63; (c) rill= 1.75, And/L=0.75.
0.4
L
0.0 -10
-5
0
5
10
X
Some calculated results are shown in Figure 25. The switching behaviour in the final stage near the bright ferroelectric state naturally depends on a parameter A n . d/A, where d is the cell thickness and ilis the wavelength of the incident light. In this approximation, when An . d / A= 0.63, the behaviour becomes ideal.
3.4 Antiferroelectric Liquid Crystal Materials 3.4.1 Ordinary Antiferroelectric Liquid Crystal Compounds Because of their potential application in display devices, a large number of, probably almost 1000, AFLC and related compounds have been synthesized in Japan. Most of these have been reported in Japanese Patent Gazettes; some have been reported in scientific reports and papers, but many of these are written in Japanese, in particular, those
in the Abstracts of Japanese Liquid Crystal Conferences. As access to these publications may not be easy for non-Japanese scientists and engineers in the liquid crystal community, and as these sources contain valuable information, we will try to summarize those AFLC compounds that have been reported mainly in Japan. Note that the materials characterized by the V-shaped switching due to thresholdless antiferroelectricity, which form one of the main topics of this discussion, are reported in these publications. Almost all the related Gazettes and Abstracts so far publicized are listed in [47] and [48]. Fortunately, the Abstracts contain titles and authors together with institutions written in English and chemical structural formulae used in the Gazettes and the Abstracts are international. The phase sequence of a prototype AFLC, MHPOBC, has been known to be complex since the early stages of the investigation of antiferroelectricity in liquid crystals [ l , 6, 491. The origin and the structures of the subphases have been well documented [6, 501. However, it is not easy to identify the sub-
3.4 Antiferroelectric Liquid Crystal Materials with Unusual Chemical Structures
phases properly in newly synthesized compounds. Moreover, the existence of subphases as well as the structures may be critically influenced by the substrate interfaces. Both the subphases, SmC;, SmCT, etc.. and SmCz itself are quite similar to SmC*, and the differences in free energy between them must, therefore, be very small. As MHPOBC was first reported as a ferroelectric liquid crystal, the phase identifications given in the Gazettes and Abstracts are usually primitive and not always correct; they frequently disregard the details of the subphases, simply specifying them as ferroelectric or antiferroelectric. Therefore, when we refer to these publications, we must be aware of this ambiguity; nonetheless, these reference sources are very useful. Note that the most general subphase sequence currently known is [50-521:
-AF
- Ferri/anti
77G
-c=C-C-O-
685
(-CHZCHC02-)
-[SmC*I- SmC; - SmA*
The chemical structural formulae of MHPOBC and another prototype AFLC, TFMHPOBC, are given in Figures 1 and 26, respectively. Most AFLC compounds so far synthesized have similar molecular structures to those of MHPOBC and TFMHPOBC. The structure can be considered to consist of three sub-structures: the chiral chain, the central core, and the nonchiral chain (Figure 26). In the chiral chain, it may be convenient to differentiate between the connector to the central core and the remaining part of the chiral chain; here-
(d) Non-chiral chains ~
non-chiralchain
core
connecter chiral chain
Figure 26. Structural formula of TFMHPOBC, illustrating four substructures: non-chiral chain, central core, connector, and chiral chain.
R-
RO-
RC02-
ROCO-
R:C,H2"+1
Figure 27. Some typical sub-structures of AFLC compounds: (a) chiral chains; (b) connectors; (c) cores; (d) non-chiral chains.
686
3 Antiferroelectric Liquid Crystals
after we will designate these the 'connector' and the 'chiral chain', unless otherwise stated. The types of substructure so far studied (i.e. chiral chains, connectors, cores and non-chiral chains) are summarized in Figure 27. In this section, the chemical structural formulae are drawn in such a way that the chiral chains are located on the righthand side. Chiral chains play a critical role in determining the stability of the antiferroelectricity. The variety of chains is rather limited. The most common chiral chain is -*CH(CF,)C,H2,+1, which stabilizes the antiferroelectricity much more than does -*CH(CH3)C,H2,+1, as illustrated in Figure 28; the odd-even effect is clearly observed in -*CH(CH,)C,H2,+I, while only the SmCi phase occurs in - *CH(CF,)C,H,,+, . Trifluoromethylalkoxy alkyl chains, -*CH(CF3)C,H&OC,H2,+ give a low threshold field and fast response
(see Sec. 3.3.2). The pretransitional effect is very small in some compounds containing trifluoromethylcycloalkane chains, such as -*CH(CF,)-C,H, and -*CH(CF,)-CsH9 (see e.g. 7 and 8).
Monofluoro- and difluoromethylalkyl chains do not stabilize antiferroelectricity. Racemates do not exhibit antiferroelectricity, but they do frequently have the herringbone structure. Moreover, Nishiyama and coworkers [53, 541 have shown that chirality is not essential for the appearance of the herringbone structure; the swallow-tailed compounds (e.g. 9), C ~ H , , O ~ C ~ C - C O ~ ~ C O ~ 3-~ 7C) 2H ( C 9
which do not have an asymmetric carbon atom, have the herringbone structure, and
C l o H z l ~ C O z ~ ~ C 'H(CF3)C,H2,+l 0 2 C
10BIMFn 2
n-6 n=7 n=8 n-9 n.10
20
40
60
80
100
120
140
160
180
n-2 n=3 n-4 n-5
n-6 n-7
2
n-6 n=7 n-8
chiral end chains (Please note: In the figure the old nomenclature - S instead of Sm and Is0 instead of I - is used).
3.4 Antiferroelectric Liquid Crystal Materials with Unusual Chemical Structures
the addition of a small amount of chiral dopant induces tristable switching. Some connectors are listed in Figure 27(b). Almost all AFLC compounds have a carbonyloxy connector between the asymmetric carbon atom and the ring in the core, but some compounds [55] have a carbonyl or oxycarbonyl connector (Figure 29). Inui [56] studied the alkylcarbonyloxy connectors and found that -C2H,COO- exhibits antiferroelectricity but -CH,COO- does not; although -CH=CH- is better regarded as included in the core, cinnamates also show antiferroelectricity. A characteristic feature of core structures is that there are three or more rings, at least two of which are not directly connected but are separated by a flexible group. Some examples are given in Figure 27(c). AFLC
687
compounds containing only two rings are quite exceptional. Four-ring compounds usually show antiferroelectricity at temperatures above 100°C. The flexible connecting groups known so far are -C(O)O-, -OC(O)-, -OCH,-, -CH,O-, C(0)S-and -C(O)NH-. Of these, -COO- is the most commonly used. An increase in the number of flexible connecting groups may stabilize the antiferroelectricity, but makes the compound viscous. Rings used in the core are benzene, cyclohexane, tetralin, naphthalene, 1,3-dioxane, piperidine, pyrimidine, etc. Of these the benzene ring is the most commonly used. Six-membered rings may be connected directly, forming biphenyl, phenylpyrimidine, pyrimidine phenyl, dioxane phenyl, phenylcyclohexane, piperidine phenyl, etc. Halogen-substituted rings are also used, as illustrated in Figure 29. In this way, a variety of core structures can be used to obtain AFLC compounds.
8
SCA SCA
sc
10 SCA
Figure 29. Some AFLC compounds containing a carbonyl, oxycarbonyl or alkylenecarbonyloxy connector.
Figure 30. Odd-even effect in compounds containing non-chiral and chiral end chains (Please note that the old nomenclature - S instead of Sm - is used).
688
3 Antiferroelectric Liquid Crystals
Some non-chiral chains are listed in Figure 27(d). Alkoxy chains, RO-, are the most common. The alkanoyloxy chain, RCOO-, increases both the stability and the viscosity, while alkoxycarbonyl chains, ROCO-, decrease them. Simple alkyl chains, R-, may slightly decrease the stability and viscosity at the same time. The odd-even effect is less clear in non-chiral chains than in chiral chains, as illustrated in Figure 30.
3.4.2 Antiferroelectric Liquid Crystal Compounds with Unusual Chemical Structures A homologous series of AFLC compounds, which have entirely different chemical
n
-_
Ps lnCicrn'1.b, 0 ideal b, -
Phase transition temperatures on cooling stepa)
104 129 129 147 159 194
50
40
60
70
80
90
100
'z
Cr SrncA* SmA I a) cooling rate = SKlmin. b) Tc-T = 5K.
structures from the ordinary ones characterized by having a carbonyl group in the connector, have been synthesized by Aoki and Nohira [57-591. The chemical structures and the corresponding phase sequences are summarized in Figure 3 1. They identified the phases by checking miscibility with a TFMHPOBC-like compound as shown in Figure 32. The ordinary tristable switching, as illustrated in Figure I6(b), was observed optically, and is characterized by a d.c. threshold and hysteresis. Tristable switching was also observed electrically by detecting two current peaks that correspond to the appearance and disappearance of the spontaneous polarization, Aoki and Nohira further confirmed that the racemates have the SmC, phase. According to the Px model de-
110
30 34 35 37 38 40
Figure 31. Structural formula of AFLC compounds synthesized by Aoki and Nohira [57-591 and their phase sequences and physical properties (Ps, spontaneous polarization; 0, tilt angle).
150
100
. ?
b
50
Figure 32. Miscibility test for 10 with n = 1 (see Figure 31) with a reference AFLC.
0 0
20 Weight 40 percent 60 of801 100
C,oH210 ~ C O , ~ C O & ( C F ~ ) C ~ H E 10 a
3.5
scribed in Secs. 3.2.3 and 3.2.4, the molecules are expected to be bent at the chiral carbon atom. It is an interesting problem for the future to study whether or not this is the case. A second class of materials are pyrimidine host mixtures containing pyranose guests, some examples of which are given in Figure 3 3 . Although the SmCi phase is not exhibited by the guest compounds themselves, Itoh et al. [60,61]* have clearly confirmed the appearance of the SmCi in the mixture, the composition of which is specified in Figure 33, by observing the typical tristable switching. This finding is very important, suggesting as it does the possibility of developing practical guest-host type mixtures of AFLC materials. Although detailed investigations have not yet been performed, some of the subphases associated with the SmCi phase seem to have been observed in the series of AFLC
Cry
71'C
Is0
Figure 33. Pyrimidine host mixtures containing pyranose guests.
References
689
compounds 11 synthesized by Wu et al.
[a.
* Note added in proof: Quite recently, Itoh et al. reported that the phase in these pyrimidine host mixtures is not antiferroelectric but ferroelectric and that the typical tristable switching results from an ultrashort pitch of about 50nm. (K. Itoh, Y. Takanishi, J. Yokoyama, K. Ishikawa, H. Takezoe and A. Fukuda, Jpn. J. Appl. Phys. 1997,36, L784).
3.5 References [ l ] A. D. L. Chandani, E. Gorecka, Y. Ouchi, H.Takezoe, A. Fukuda, Jpn. J. Appl. Phys. 1989,29, L1265. [2] J. W. Goodby, J. Muter. Chem. 1991, 1 , 307. [3] R. Blinc, CondensedMatferNews 1991, 1, 17. [4] P. E. Cladis, Liq. Cryst. Today 1992, 2, 7. [5] A. Fukuda (Ed.), The Future Liquid Crystal Display and its Materials, CMC, Tokyo, 1992 (in Japanese): A. Fukuda, J. Li, Chap. 1 , p. 1; Y. Yamada, Chap. 8, p. 102; K. Itoh, Chap. 13, p. 287; M. Johno, T. Yui, Chap. 14, p. 297; S. Inui, H. Iwane, Chap. 15, p. 308; and S . Nishiyama, Chap. 16, p. 318. [6] A. Fukuda, Y. Takanishi, T. Isozaki, K. Ishikawa, H. Takezoe,J. Muter. Chem. 1994,4,997. [7] I. Nishiyama, Adv. Muter. 1994, 6, 996. [8] K. Miyachi, Ph. D. Thesis, Tokyo Institute of Technology, 1996. [9] R. B. Meyer, L. Liebert, L. Strzelecki, P. Keller, J. Phys. (France) 1975, 36, L69. [ l o ] R. B . Meyer, Mol. Cryst. Liq. Cryst. 1977, 40, 33. [I 11 M. Luzar, V. Rutar, J. Seliger, R. Blinc, Ferroelectrics 1984,58, 115. [I21 A. Yoshizawa, H. Kikuzaki, M. Fukumasa, Liq. Cryst. 1995, 18, 351. [13] K. H. Kim, K. Ishikawa, H. Takezoe, A. Fukuda, Phys. Rev. E 1995,51,2166. (Erratum: Phys. Rev. E 1995, 52, 2120.) [14] K. H. Kim, K. Miyachi, K. Ishikawa, H. Takezoe, A. Fukuda, Jpn. J. Appl. Phys. 1994, 33, 5850. [I51 K. Miyachi, J. Matsushima, Y. Takanishi, K. Ishikawa, H. Takezoe, A. Fukuda, Phys. Rev. E 1995,52, R2153.
690
3
Antiferroelectric Liquid Crystals
[16] A. D. L. Chandani, E. Gorecka, Y. Ouchi, H. Takezoe, A. Fukuda, Jpn. J. Appl. Phys. 1989, 28, L1265. [17] A. Fukuda, Y. Takanishi, T. Isozaki, K. Ishikawa, H. Takezoe, J. Muter. Chem. 1994,4,997. [l8] Y. Takanishi, H. Hiraoka, V. K. Agrawal, H. Takezoe, A. Fukuda, M. Matsushita, Jpn. J. Appl. Phys. 1991,30, 2023. [19] K. Hori, K. Endo, Bull. Chem. Soc. Jpn. 1993, 66,46. [20] L. A. Beresnev, L. M. Blinov, M. A. Osipov, S. A. Pikin, Mol. Cryst. Liq. Cryst. l988,158A, 1. [21] Y. Galerne, L. L. Liebert, Phys. Rev. Lett. 1991, 66, 2891. [22] P. E. Cladis, H. R. Brand, Liq. Cryst. 1993, 14, 1327. 1231 H. R. Brand, P. E. Cladis, H. Pleiner, Macromolecules 1992,25, 7223. [24] I. Nishiyama, J. W. Goodby, J. Muter. Chem. 1992, 2, 1015. [25] J. Prost, R. Bruinsma, Ferroelectrics 1993,148, 25. [26] R. Bruinsma, J. Prost, J. Phys. I1 (France) 1994, 4, 1209. [27] Y. Galerne, L. Liebert, Phys. Rev. Lett. 1990,64, 906. [28] B. Jin, Z. H. Ling, Y. Takanishi, K. Ishikawa, H. Takezoe, A. Fukuda, M. Kakimoto, T. Kitazume, Phys. Rev. E 1996,53, R4295. [29] A. D. L. Chandani, T. Hagiwara, Y. Suzuki, Y. Ouchi, H. Takezoe, A. Fukuda, Jpn. J. Appl. Phys. 1988,27, L729. [30] A. D. L. Chandani, E. Gorecka, Y. Ouchi, H. Takezoe, A. Fukuda, Jpn. J. Appl. Phys. 1989, 28, L 1265. [31] M. Johno, A. D. L. Chandani, J. Lee, Y. Ouchi, H. Takezoe, A. Fukuda, K. Itoh, T. Kitazume, Proc. SlD 1990,31, 129. [32] M. Yamawaki, Y. Yamada, N. Yamamoto, K. Mori, H. Hayashi, Y. Suzuki, Y. S. Negi, T. Hagiwara, I. Kawamura, H. Orihara, Y. Ishibashi in Proceedings of the 9th International Display Research Conference (Japan Display '89, Kyoto), 1989, p. 26. [33] K. Miyachi, J. Matsushima, Y. Takanishi, K. Ishikawa, H. Takezoe, A. Fukuda, Phys. Rev. E 1995,52, R2153. [34] A. Fukuda in Proceedings of the 15th International Display Research Conference (Asia Display '95, Hamamatsu), 1995, p. S6-1. [35] S. Inui, N. Iimura, T. Suzuki, H. Iwane, K. Miyachi, Y. Takanishi, A. Fukuda, J. Muter. Chem. 1996,6, 67 1 . [36] M. Johno, K. Itoh, J. Lee, Y. Ouchi, H. Takezoe, A. Fukuda, T. Kitazume, Jpn. J. Appl. Phys. 1990,29, L107. [37] X. Y. Wang, P. L. Taylor, Phys. Rev. Lett. 1996, 76, 640.
[38] T. Akahane, A. Obinata, Liq. Cryst. 1993, 15, 883. [39] H. Pauwels, A. de Meyere, J. Fornier, Mol. Cryst. Liq. Cryst. 1995,263, 469. [40] M. A. Czarnecki, N. Katayama, M. Satoh, T. Watanabe. Y. Ozaki, J. Phys. Chem. 1995.99, 14101. [41] K. Nakagawa, T. Shinomiya, M. Koden, K. Tsubota, T. Kuratate, Y. Ishii, F. Funada, M. Matsuura, K. Awane, Ferroelectrics 1988, 85, 39. [42] N. Yamamoto, N. Koshoubu, K. Mori, K. Nakamura, Y. Yamada, Ferroelectrics 1993, 149,295. (431 E. Tajima, S. Kondon, Y. Suzuki, FerroelecrricJ 1994,149,255. [44] S. Inui, T. Suzuki, N. Iimura, H. Iwane, H. Nohira, Ferroelectrics 1993, 148, 79. [45] M. Johno, T. Matsumoto, H. Mineta, T. Yui, Abstracts 11: Chemical Society of Japan 67th Spring Meeting, Chemical Society of Japan, Tokyo, 1994, Abstract 1 1B31 12. [46] G. M. Barrow, Physical Chemistry, 5th edn, McGraw-Hill, New York, 1988, p. 67 I . [47] Japanese Patent Gazettes: H 1-213390; H 13 16339; H1-3 16367; H1-316372; H2-28128: H2-40625; H2-78648; H2-131450; H2-153322; H2-160748; H2-173724; H2-176723; H2230117; H2-270862; H3-11041; H3-68686; H38395 1; H3- 121416; H3- 123759; H3- 141323; H3- 1675 18; H3-18 1445; H3-197444; H3197445; H3-2 18337; H3-223263; H3-291270; H3-292388; H4-29978; H4-46158; H4-46159; H4-55495; H4-96363; H4-82862; H4-89454; H4-103560; H4-108764; H4-117345; H4159254; H4-173779; H4-178353; H4-198155; H4-217942; H4-235950; H4-244047; H4261 162; H4-261163; H4-266853; H4-266855; H4-266876; H4-282344; H4-305554; H4305555; H4-312552; H4-3 16562; H4-327576; H4-338361; H4-338362; H4-342546; H4342547; H4-360850; H4-360854; H4-372929; H5-1041; H5-4945; H5-17407; H5-17408; H532675; H5-32973; H5-39246; H5-58959; H565484; H5-70392; H5-85988; H5-85989; H597821; H5-98258; H5-105644; H5-105877; H5132449; H5-140046; H5-140119; H5-155880; H5-163208; H5-163248; H5-186401; H5186402; H5-201929; H5-201983; H5-202028; H5-213881; H5-221928; H5-230032; H5230033; H5-230035; H5-247026; H5-255 193; H5-247026; H5-255193; H5-255196; H5311171; H5-320125; H5-331107; H6-25098; H6-25099; H6-25100; H6-32764; H6-32765; H6-32770; H6-65 154; H6- 108051; H6- 1 16208; H6-128236; H6-145111; H6-211749; H6228056; H6-247898; H6-247901; H6-247902; H6-271852. [48] Abstracts of Japanese Liquid Crystal Conferences: 1990,16,2K101,2K102,2K103,2K104, 2K105,2K106; 1991,17,2F302,2F315,3F313,
3.5 3F314, 3F316, 3F318, 3F321, 3F322, 3F324; 1992,18,2B408,2B409,2B410,2B411,2B412, 2B417, 3B416, 3B417, 3B419, 3B420, 3B421, 3B422,3B423,3B424,3B425; 1993,19, l A l l , 1A12, 1A14, 1B13, 2A03, 2A08, 2A10, 3D08; 1994,20, 1F603,1G303,2F610,2F614,2F617, 3G201, 36203,3G205,3G206,36208,3G213, 3G215, 3G216, 36218; 1995,21, 2C01, 2C04, 2C05, 2C09, 2C10, 2C12, 2C13, 2C14, 2C18, 3C07,3C08. [49] M. Fukui, H. Orihara, Y. Yamada, N. Yamamoto, Y. Ishibashi, Jpn. J. Appl. Phys. 1989, 28, L849. [50] K. Itoh, M. Kabe, K. Miyachi, Y. Takanishi, K. Ishikawa, H. Takezoe, A. Fukuda, J. Muter. Chem. 1996, 7,407. [51] H. Hatano, Y. Hanakai, H. Furue, H. Uehara, S. Saito, K. Muraoka, Jpn. J. Appl. Phys. 1994, 33, 5498. [52] T. Sako, Y. Kimura, R. Hayakawa, N. Okabe, Y. Suzuki, Jpn. J. Appl. Phys. 1996, 35, L114. 1531 I. Nishiyama, J. W. Goodby, J. Matel: Chem. 1992,2, 1015.
References
69 1
[54] Y. Ouchi, Y. Yoshioka, H. Ishii, K. Seki, M. Kitamura, R. Noyori, Y. Takanishi, I. Nishiyama, J. Muter. Chem. 1995,5,2297. 1551 K. Kondo in Extended Abstracts of 52nd Research Meeting, JSPS 142 Committee of Organic Materials for Information Science (Sub-Committee of Liquid Crystals) 1993, 1 . [56] S. Inui, Ph. D. Thesis, SaitamaUniversity, 1995, p. 1 (in Japanese). [57] Y. Aoki, H. Nohira, Liq. Cryst. 1995, 18, 197. [58] Y. Aoki, H. Nohira, Liq. Cryst. 1995, 19, 15. [59] Y. Aoki, H. Nohira, Ferroelectrics 1996, 178. 213. [60] K. Itoh, M. Takeda, M. Namekawa, S. Nayuki, Y. Murayama, T. Yamazaki, T. Kitazume, Chem. Lett. 1995, 641. [61] K. Itoh, M. Takeda, M. Namekawa, Y. Murayama, S. Watanabe, J. Muter. Sci. Lett. 1995, 14. 1002. [62] S.-L. Wu, D.-G. Chen, S.-J. Chen, C.-Y. Wang, J.-T. Shy, G. Hsiue, Mol. Cryst. Liq. Cryst. 1995, 264, 39.
Handbook of Liquid Crystals Edited by D. Demus, J. Goodby, G. W. Gray, H.-W. Spiess, V. Vill Copyright 0 WILEY-VCH Verlag GmbH. 1998
Chapter VII Synthesis and Structural Features Andrew N. Cammidge and Richard J. Bushby
1 General Structural Features The discovery of discotic liquid crystals is generally dated to the work of Chandrasekhar on the hexaesters of benzene [l], although the concept that liquid crystal phases should be formed by disk-like as well as rodlike molecules had already been recognized by DeGennes and others [2]. Structural similarities also exist with the carbonaceous phases, which have been known for some time [3]. The least ordered (usually highest temperature) mesophase formed by disk-like molecules is the nematic ( N D ) phase which is directly comparable to the nematic phase formed by calamitics. The molecules in such a phase have orientational order but no positional order. Examples of N D phases remain rather few, and more common are those phases in which the molecules stack into columns, the columns then being organized on a two-dimensional lattice. In some respects these columnar phases are analogous to the smectic mesophases of calamitic liquid crystals. However, whereas the smectic phases can be thought of as two-dimensional liquids (they have positional order in one dimension and are disordered in the other two), the columnar phases can be thought of as one-dimensional liquids (they have positional order in two dimensions and are disordered in the third). Columnar phases show rich polymesomorphism and they are normally classified at three levels; according to the symmetry of the two-dimensional array, the or-
ientation of the core with respect to the column axis, and (sometimes) the degree of order within the column. The latter criterion has been the subject of a variety of different interpretations. The main types [4] are illustrated in Figure 1 and they are discussed in more detail in Chapter 8 in this volume. Typically, if a compound forms both a D,' phase (rectangular columnar) and a D, phase (hexagonal columnar) the D, phases are found at lower temperatures. It should be noted, however, that exceptions exist. Determination of columnar mesophase structure can often be difficult, since the differences in the X-ray diffraction results can be quite subtle. The archetypal discogenic molecule has a rigid, planar aromatic core with three-, four- or six-fold rotational symmetry, and generally six or more flexible side chains, each more than five atoms. However, exceptions to all these 'rules' are now quite common. Systems are known with low symmetry, a non-planar, non-aromatic core, and with side chains as short as three atoms. We survey here the different classes of discotic mesogens, the methods by which they have been synthesized, and the phase behaviour claimed for each system. Interest in the development of new routes to discotic mesogens has provided a welcome stimulus to a somewhat unfashionable area of organic
'
In this contribution the nomenclature D for a columnar mesophase of disk-like molecules has been used, but the reader should appreciate that D,, D,, D, etc. are the same as Col,, Col,, Colt etc.
694
VII
Synthesis and Structural Features
2 Aromatic Hydrocarbon Cores
Nematic Phase ND
Disordered Columnar Hexagonal Phase Dhd
2.1 Benzene 2.1.1
Common variations in stacking within the columns
Ordered (0)
Disordered (d)
Tilted (t)
Common variations in stacking of the columns
Hexagonal (h)
Rectangular (r)
Oblique (ob)
Figure 1. Schematic diagram illustrating the basis for the classification and nomenclature of the discotic mesophases. For the columnar phases the subscripts are often used in combination with each other. Hence D,, refers to a (two dimensional) rectangular lattice of columns within which the discs are stacked in a disordered manner
synthesis, i.e. of polynuclear aromatic and heterocyclic systems. Quite a lot of what has been developed so far has exploited old, and in some cases very old, chemistry in which the polynuclear core is assembled through some ‘one-off’ condensation process. There is clearly a need to develop rational general synthetic strategies based on modern methods. Undoubtedly more of these will emerge in the next few years.
Esters and Amides
The hexasubstituted benzenes were the first discotic liquid crystals to be reported. Chandrasekhar and coworkers [ 1, 51 prepared a series of hexaesters of hexahydroxybenzene and found that the heptanoyl and octanoyl homologues were liquid crystalline. The synthetic route used, which has been adapted by others to prepare related derivatives, is shown in Scheme 1 . Hexahydroxybenzene [6] is prepared from glyoxal and treated with an acid chloride to give the esters. Yields of the hexaester are improved if an argon atmosphere and/or a partial vacuum is used. Typically, no solvent is used for these reactions. The mesophase behaviour of a range of hexaesters is summarized in Table 1. All derivatives that have been characterized so far form an hexagonal mesophase, but the stability (temperature range) of the mesophase is extremely sensitive to
HC-CHNaO
OH
HO
0
OH OH
ic ROCO$L:
ROC0 OCOR
Scheme 1. Synthesis of the hexaesters of benzene. Reagents: (a) Na,SO,/O,; (b) SnC1,IHCI; (c) RCOCl 14-87% [ 7 ] .
2
Table 1. Transition temperatures for hexaesters
695
Aromatic Hydrocarbon Cores
0
0
OH
(RCOO-) of benzene. R
Phase transitionsa
Refs.
Cr 81 Dh 86 I [I, 5, 81 Cr 82 Dh 84 I [ l , 5, 81 Liquid [71 0C6H I3 CH2OCSHl1 Cr 68 Dh 94 I [71 C2H40C4H9 Cr 30 D, 32 I [71 C3H60C3H7 Cr 43 Dh 44 I [71 [71 CSHIOOCH3 CI' 16 Dh 25 I CH2OC3H7 CrllODh1121 171 CH20C7HlS Cr 74 Dh 94 I [71 Cr 66 Dh 67 I (Cr 69 I) [7,9] CH2SCSHI I C2H4SC4H9 Cr 101 I PI C3H6SC3H7 ? 53 D 57 I [91 CsH,,SCH3 Cr 32 D 37 I [91 CH(Me)C6H13 Liquid 1101 CH,CH(Me)C,H, Cr 96 I [lo1 C,H4CH(Me)C4H, M 96 I I101 M 1021 I101 C3H6CH(Me)C,H, C4H8CH(Me)C,H, Cr 61 M 87 I [I01 [lo1 C,H,,CH(Me)CH, Cr 85 M 87 I C,H,CH(Et)C,H, Cr 60 MI 86 M, 87 I [lo] [lo1 CsH,,CH(Me)C,H, Cr 52 M 75 I C6H I 3 C7HlS
a
Scheme 2. Synthesis of the tetraesters of 1,4-benzoand I ,4-hydro-quinone. Reagents: (a) RCOCl/pyridine reflux 30-60% [lo]; (b) H+ 90-95% [ l 11.
Table 2. Transition temperatures for the tetraesters (RCOO-) of 1,4-benzo- and 1,4-hydroquinone.
Quinones Hydroquinones
H
O
G
o
M, mesophase. HO
/ NHz
heteroatoms in the side chain. The greatest stabilization is found for derivatives bearing a P-oxygen atom in the side chain. If the position of this atom is moved towards or away from the core, either the mesophase stability is drastically reduced or a mesophase is not observed at all [7]. Branched side chains stabilize the mesophase if the branch point is close to the middle, but destabilization is observed as it approaches the core [lo]. Some partially (2,3,5,6-) acylated derivatives of hexahydroxybenzene have also been found to exhibit discotic liquid crystal phases, as have the related tetraesters of 2,3,5,6-tetrahydroxy- 1,4-benzoquinone [ 1 I]. Their synthesis, shown in Scheme 2, is relatively straightforward. 2,3,5,6-Tetrahydroxy- 1,4-benzoquinone [ 121 is acylated with an excess of the acid chloride in refluxing pyridine. Neutral work-up yields the quinone, whereas the hydroquinone is iso-
:
a
R
Phase transitions
Refs
C,H,, C,H,, C6HI3 C7Hl,
Cr55D57I Cr68D70I
[Ill [Ill
Cr 80 D 85 I Cr79D82I
[13, 141 1141
RCOp@
NHCOR
RCOp
/
OCOR
NHCOR
1
Scheme 3. Synthesis of N,N'-dialkanoyl-2,3,5,6-tetrakis(alkanoy1oxy)-1,4-benzene diamines. Reagents: (a) RCOCUpyridine [ 131.
lated from an acidic work-up [ 131. Both sets of compounds, however, have very narrow mesophase ranges (Table 2). Another variation, employed by Matsunaga and coworkers, has been to replace some of the ester linkages of the benzene hexa-alkanoates with amide groups. The most heavily substituted members of this class of compounds, the N,N'-dialkanoyl2,3,5,6-tetrakis(alkanoyloxy)-1,4-benzene diamines (1) [15], are prepared by total acylation of 3,6-diamino-1,2,4,5-tetrahydroxybenzene [ 161 (Scheme 3). The mesophase behaviour of this series is summarized in Table 3. It is immediately apparent that the introduction of the two amide groups stabiliz-
696
VII
Synthesis and Structural Features
A
Table 3. Transition temperatures of N,N'-dialkanoyl2,3,5,6-tetrakis(alkanoyloxy)(1) (RCONH-/RCOO-). 1,4-benzene diamines
C4H9 CSHI I C6H 13
C7H IS C8H17 C9H19
CI,HZl CIIH23 1 3H27
ClSH3,
Cr 134 D, 208 I Cr 71 D, 209 I Cr 64 D, 208 I Cr 75 D, 205 1 Cr 77 D, 200 1 Cr 77 D, 199 1 Cr 81 D, 198 I Cr 89 D, 197 1 Cr 92 D, 189 I Cr 94 D, 190 I
a
"J2&NO2
-
NH2*
R
CHI
CH3
m
b
_c
N
H
CH3
i
CHI NHCOR
NHz
Scheme 5. Synthesis of 1,3,5-trimethylbenzene triamides [19]. Reagents: (a) HN03/H2S0,; (b) Sn/HCI; (c) RCOCUpyridine.
Table 5. Transition temperatures for 1,3,5-trimethylbenzene triamides (RCONH-) (see Scheme 5 for structure). R
Phase transitions [ 191
C3H7 C4H9 CSHI 1 C6H13
I
OH
I OCOR
Scheme 4. Synthesis of benzene 1,3-diamide-2,4diesters [18]. Reagents: (a) NaN02, -2°C [17]; (b) SnCI2/HCI. (c) RCOCl/pyridine.
C7H 15 C9H19
CIOHZI c l lH23
I 3H27
C15H31
Cr 380 D, 410 1 Cr 315 D, 380 1 Cr 300 D, 357 I Cr 257 D, 357 I Cr 239 D, 346 I Cr 200 D? 222 D, 340 I Cr 189 D? 206 Dh 342 1 Cr 185 D? 191 D, 338 I Cr 183 D? 189 D, 342 I Cr 118 D? 182 D, 338 I Cr 120 D? 175 D, 328 I
Table 4. Transition temperatures for benzene 1,3diamide-2,4-diesters (RCONH-/RCOO-) (see Scheme 4 for structure). R
Phase transitions 1181
C6H I3
C7HlS C9H19
CIOHZI C,lH23 13H27 C1SH31
Cr 89 D, 124 I Cr 87 D, 122 1 Cr 90 D, 121 I Cr 93 D, 122 I Cr 95 D, 120 I Cr97Dh1171 Cr 101 D, 114 I Cr 103 D, 114 I
NHCOR
& y y J J L ~ z f f o R
es the discotic mesophase substantially with respect to the parent hexaester. The melting points Of the two sets Of are very Similar for derivatives O f the Same chain length (e.g. for C,H,,CO-, hexaester m.p.
NO2
NO2 RCONH
NHCOR
C
Scheme 6. Synthesis of mixed ester and amide derivatives of benzene [21], Reagents: (a) Sn/HCI; (b) RCOCl/pyridine; (c) HN03 [20].
2
Aromatic Hydrocarbon Cores
697
Table 6. Transition temperatures for mixed ester and amide derivatives of benzene (RCONH-/RCOO-) (see Scheme 6 for structures). Phase transitions [21]
R
A
B
C
144 Dhd 222 I 133 D,, 225 I 129 Dh, 229 I 123 D,, 227 I 109 D,, 226 I 1 1 1 D,, 227 I Cr 107 D,, 225 I Cr 112 D,, 225 I Cr 113 Dh, 218 I
Cr 100 Dh, 241 1 Cr 55 Dh, 238 1 Cr 63 D,, 241 I Cr 49 D,, 242 I Cr 73 D,, 227 I Cr 65 D,, 225 I
Cr 113 D,, 196 I Cr 67 D,, 208 I Cr 88 Dh, 208 I Cr 77 Dh, 208 I Cr 93 Dh, 206 I Cr 92 Dh, 203 I Cr 96 D,, 202 I Cr 101 D,, 203 I Cr 103 Dhd 198 I Cr 103 Dhd 193 I
Cr Cr Cr Cr Cr Cr
C4H9 CSH, 1 C6H13 C7H I S C9H 19
C10H21 CI,H,, 1 3H2?
C15H31
COCl
ClCO
Cr 78 Dh, 224 I Cr 84 Dh, 224 1 Cr 66 D,, 225 I
Table 7. Transition temperatures for N,N',N"-1,3,5benzene tricarboxamides (RCONH -).
CONHR
COCl
RNHCO
CONHR
Scheme 7. Synthesis of N,N',Nf'-l,3,5-benzene tricarboxamides [22]. Reagents: (a) RNHJpyridine.
R
Phase transitionsa [22]
C5Hl I C6H 13
C7H1s C8Hl?
C9H 19
82 "C, bisamide m.p. 75 "C) and the mesophase extension results primarily from an elevation of the clearing points. The mesophase-stabilizing effect of the amide group allows discotic benzene derivatives to be prepared that carry fewer long chains. The syntheses of such materials are shown in Schemes 4-6, and their mesophase properties are given in Tables 4-6. The N,N',N"- 1,3,5-benzene tricarboxamides [22] also form mesophases, but it should be noted that their columnar structure has not been confirmed unambiguously. These compounds are simply synthesized from commercially available 1,3,5benzene tricarbonyl chloride and alkylamines by refluxing in pyridine (Scheme 7). Their thermal behaviour is shown in Table 7. It should be pointed out that other 1,3,5-trisubstituted benzenes have been synthesized and shown to form monotropic mesophases [23].
CIOH2, CI 1H23 c 12H2.5 C13H2? 14H29
CISH31 I 6H33 I 7H35 1 sH37
a
Cr119M206I Cr 99 M 205 I Cr116M208I Cr 102 M 204 I Cr 65 M 215 I Cr 49 M 208 I Cr 72 M 216 I Cr 88 M 212 I Cr 81 M 216 I Cr 61 M 209 1 Cr 88 M 214 I Cr 73 M 205 I Cr87M211I Cr 68 M 206 I
M, mesophase.
2.1.2 Multiynes A widely studied class of discotic benzene derivatives are the multiynes prepared by Praefcke and coworkers [24].Like the structurally similar naphthalenes, these materials are prepared by palladium-catalysed coupling of polybromobenzenes with acetylene derivatives (Scheme 8). Their mesophase behaviour is given in Table 8. The hexakisalkynylbenzenes themselves exhib-
698
VII
Synthesis and Structural Features
Table 8. Transition temperatures for benzene multiynes (see Scheme 8 for structures). X
R
Phase transitions
Refs.
Cr 170 N, 185 Idecomp Cr 124 N, 142 I Cr 98 N, 131 I Cr 80 ND 96 I Cr 109 N, 193 I Cr 170 D >240 Idecomp Cr 246 D >260 Idecomp Cr 171 D 230 N, 250 Idecomp Cr 106 D 285 Idecomp [I-16 D] [I-8 D]
-
%: R
R
I
+ RONa-
Br%Br
Br
Er
/
RO
OR
A Scheme 8. Synthesis of benzene multiynes. Reagents: (a) RC=CH/Pd[O].
Scheme 9. Synthesis of hexakis[(alkoxyphenoxy)methyllbenzenes [28].
it only monotropic mesophases [ 3 ] . Formation of ‘superdisk’ structures such as benzene hexakisphenylacetylenes [24, 251 gives rise to materials that exhibit relatively stable discotic nematic phases. Enlarging the core even further (biphenyl analogues) leads to columnar discotic mesophases [27].
Table 9. Transition temperatures for hexakis[(alkoxyphenoxy )methyl]benzenes (ROCH, -).
2.1.3 Hexakis(a1koxyphenoxymethyl) Derivatives One further class of discotic benzenes comprises the hexakis[(alkoxyphenoxy)methyl] derivatives [28]. These compounds are simply prepared from hexakis(bromomethy1)benzene and an excess of p-alkoxyphenoxide (Scheme 9). The thermal behaviour of these materials is given in Table 9 [28], but it should be noted that the authors were un-
R
Phase transitions [28]
Cr 68 D,, 97 I p-C5H110C6H4Cr 76 D,, 83 I p-C6H I 3°C6H4p - ~ 7 ~ 1 5 ~ ~ 6 ~ Cr 4 67 - D,, 71 I
able to confirm unambiguously the presence of a discotic mesophase and point out that the ‘mesophase’ might be a special type of crystal [28].
2.1.4 Hexakis(alkylsu1fone) Derivatives Finally, a class of benzene derivatives that also form discotic mesophases is the hexa-
2
-
clQ
a
CI
61
Cl
.'"@
Rs+
SR
RS
699
A
S02R
RS4
SR
Aromatic Hydrocarbon Cores
W2R
Scheme 10. Synthesis of hexakis(alkylsu1fone) derivatives of benzene [29]. Reagents: (a) RSNdHMPT 80-95%; (b) m-CPBAICHCI, 60-70%. Scheme 11. Synthesis of naphthalene multiynes. Reagents: (a) RC6H4C=CH/Pd[O]. Table 10. Transition temperatures for hexakis(alky1sulfone) derivatives of benzene (RS02-). R
Phase transitions
Table 11. Transition temperatures for naphthalene multiynes (see Scheme 11 for structures).
Refs.
R C6H 13
C7HE CXH17 C9H19
ClOH2, C11H23
CI2H2S C13H27
C14H29 C15H31
I hH33
I 164 Cr I 138 D, 120 Cr I 134 D, 108 Cr 1 130 D, 87 Cr I 125 D, 77 Cr I 115 D, 65 Cr I 103 D, 45 Cr I 9 0 D, 44 Cr 1 77 D, 42 Cr I 63 D, 49 Cr I 6 9 Cr
kis(alkylsu1fone) derivatives prepared by Praefcke and coworkers [29]. The synthesis (Scheme 10) is straightforward. Hexachlorobenzene is treated with an excess of sodium alkylthiolate, typically in HMPT, to give the hexakis(alky1thio)benzenes. Oxidation with rn-CPBA affords the sulfones. The compounds appear to form relatively disordered hexagonal mesophases [30-321 (Table lo), which show significant supercooling.
2.2 Naphthalene Two general series of naphthalene derivatives are known to give discotic liquid crystal phases, the multiynes [27, 331 and the p-alkoxyphenoxymethyl derivatives [28]. The multiynes are synthesized through
Phase transitions
Cr 121 D, 156 D, -260 Idecomp Cr 69 N, 98 D, 134 D, -245 Idecomp Cr 49 N, 95 D, 159 D, -230 Idecomp Cr 60 N, 112 D, 137 D,, -200 I C9H19 Cr68N,llOI CSHIIO Cr 144 N, -250 Idecomp HC6H4 Cr 240 D >260 dI CSH11C6H4 Cr 122 D >240 dI C8H17C6H4 Cr 75 D >240 dI C,HI,C,H, Cr 58 D -180 N, -230 Idecomp CSHll C~HI, CTHl, C,H,,
Refs. [33] [33] [33] [33] [33l ~331 ~271 ~271 1271 [27]
a palladium-catalysed coupling of hexabromonaphthalene with aryl acetylenes [33] (Scheme 11) and their mesophase behaviour is summarized in Table 11. The multiynes exhibit a rich polymesomorphism and tend to form the relatively rare nematic discotic phase, which may be due to the 'super-disk' structure of the molecules [27,33]. The most striking feature, however, is the reverse phase sequence for the lower homologues in the alkylphenylacetylene series. It should be noted that the related alkylacetylene derivatives do not give any liquid crystal phases. The p-(a1koxyphenoxy)methylnaphthalenes are easily synthesized by a nucleophilic substitution on bromomethylnaphthalenes (prepared by bromination of polymethylnaphthalenes [34]) with p-alkoxyphenoxides [28] (Scheme 12). Their mesophase behaviour is summarized in Table 12 and it can
700
VII
Synthesis and Structural Features
lb
OR
R
O
z
C
W
N
O
R
A
c
O
W
OR
CCOR
R02C
R' = H or CH20PhOR
6R
0
AcO
OAc
OCOR
0
OR
4
Scheme 12. Synthesis of (p-alkoxyphenoxymethyl) derivatives of naphthalene [28]. Reagents: (a) Br2/hv ;
4
(b) NaOC,H,OR/DMF 80°C, 2 h 20-63%.
Table 12. Transition temperatures for [(p-alkoxyphenoxy)methyl]naphthalenes (see Scheme 12 for structures). R C5H,, C,H,:, C7HI5 C,H,, C,H,, C6H13
C7H15
R'
Phase transitions [28]
C5H,,0PhOCH2 Cr 162 D, 164 I C6H130PhOCH, Cr 135 D, 167 I C,H,,OPhOCH, Cr 105 D, 167 I C,H,,OPhOCH, Cr 78 D, 167 I C,H,,0PhOCH2 Cr 62 Dh 165 I H Cr 65 D,, 147 D, 184 I H Cr 79 D,, 92 D:, 138 D, 182 I
be seen that the symmetrical (octa-substituted) derivatives give a single hexagonal mesophase. The unsymmetrical (hepta-substituted) materials, however, are polymesomorphic, with the higher homologue giving two distinct rectangular phases in addition to the high-temperature hexagonal phase.
2.3 Anthracene (Rufigallol) The rufigallol esters occupy an important place in the history of discotic liquid crystals, as they were one of the earliest systems
OAC
RO
0
RO
OR
0
OR
Scheme 13. Synthesis of Anthracene (Rufigallol) based discotic liquid crystals. Reagents: (a) 98% H,S0,/100 "C/2 h [36]; (b) Ac20/trace H,SO,/reflux12 h [36]; (c) pyridinehoil [36]; (d) Hf 1361; (e) RBr/Na,C03/DMF/160"C/16 h [38]; (f) RCOCI/ pyridine - acetonel40 "CI1 h [37].
reported to form a columnar mesophase but to have a nucleus with only C, symmetry [35]. Rufigallol (3) is made in low yield through the acid-catalysed self-condensation of gallic acid (2) [36] and it is normally purified through its hexa-acetate (4). The esters are prepared from the hexaphenol(3) [35, 371, but the ethers are better made directly from the acetate (4) [38]. The syntheses are shown in Scheme 13 and the mesophase behaviour of the products is summarized in Table 13. All the ethers C, to C,, [38] and esters C7 to CI41371 show enantiotropic behaviour, and the esters C, to C, [35, 371 also show monotropic behaviour, a more ordered mesophase being formed when the normal columnar phase is cooled to 108 "C in the case of the heptanoate [37], 99 "C in the case of the octanoate 1351 and 83 "C in the case of the nonanoate [37].
2
Table 13. Transition temperatures of anthracene (rufigallol) based discotic mesogens. Substituent
Phase transitions a
Cr 105 Dh 131 I 1381 Cr 83 D, 117 I [381 Cr 53 Dh 105 1 [381 C7HlS0 Cr 49 D, 102 I [381 C8H170 Cr 37 D, 96 I [381 C9H190 Cr 32 Dh 92 I [381 c IOH2lO Cr 48 D, 87 I [381 Cr 26 D, 84 I [381 CI I H2,O C12H250 Cr 37 Dh 79 1 [381 C13H270 Cr 62 D, 72 I 1381 C3H7CO2 Cr 216 I I371 C4H9C02 Cr 170 I [371 CSHl lc02 Cr 152 I [371 C6H I 3c02 Cr 112 M b 134 I [371 C,HlsC02 Cr 110 (108) M C133 (128) I [37 (35)] C8H17C02 Cr 81 M d 128 I [371 C9H 19c02 Cr 109 M 128 I [371 c lOH21CO2 Cr 92 126 I [371 c l 1H23C02 Cr 99 M 124 I 1371 I 2H25C02 Cr 96 M 120 I 1371 13H27C02 Cr 103 M 117 I [371 a
Me0 OMe
-
OMe
Refs.
C4H9O CSHIIO C6H I3O
M, mesophase. Gives a different mesophase on cooling to 108 "C. Gives a different mesophase on cooling to 99 "C. Gives a different mesophase on cooling to 83 "C.
M
e
0
8
H
OMe
MeO
d
HO
OH
(OH)
(OW
RCO,
O OMe Z
-
\
/=\
PCOR
RCO*p+&
)"
RC02
"(OCOR
Scheme 14. Synthesis of phenanthrene based discogens [39]. Reagents: (a) LiOMe/MeOH; (b) 12/hv; (c) BBr,; (d) DCCI/DMAP/RCOOH.
Table 14. Transition temperatures for phenanthrene based discotic mesogens (see Scheme 14 for structures). R
R'
Phase transitions (391
H H RC02RC02-
D51 I D 75 I Cr 76 I Cr -13 D 145 I
2.4 Phenanthrene Discotic phenanthrene derivatives have been prepared by Scherowsky and Chen [39]. Their synthesis is illustrated in Scheme 14. The phosphonium salt prepared by bromination of trimethoxybenzyl alcohol and reaction of the benzyl bromide with triphenyl phosphine was deprotonated and a Wittig reaction with 2,4-dimethoxyor 2,3,4-trimethoxybenzaldehydeproduced the stilbene. Photocyclization gives the phenanthrene, which can be demethylated with boron tribromide and acylated using the Steglich method. The mesophase behaviour of discotic phenanthrenes is given in Table 14. It was
70 I
Aromatic Hydrocarbon Cores
~~
C,H,,OCH*(CH3) C6H,3CH*(CH3)0CH,C7HI5OCH*(CH,)C6H13CH*(CH,)OCH,-
~
found that straight-chain esters of hexahydroxyphenanthrene gave only narrow, monotropic mesophases. Enantiotropic mesophases were formed for branchedchain penta- and hexaesters. A helical 'Dr' mesophase structure is proposed for these chiral derivatives, which show electrooptic switching.
702
Synthesis and Structural Features
VII
2.5 Triphenylene The hexasubstituted derivatives of triphenylene were among the first discotic mesogens to be discovered. The most commonly studied derivatives are the hexaethers and hexaesters of 2,3,6,7,10,1 l-hexahydroxytriphenylene, and their basic synthesis is illustrated in Scheme 15.1,2-Dimethoxybenzene (veratrole) is oxidatively trimerized using chloranil in concentrated sulfuric acid to give hexamethoxytriphenylene [40]. The reaction takes about a week. Demethylation with boron tribromide [41] or hydrobromic acid [42] yields the hexaphenol,
a
-
CH3O
Ib RC02 OCOR OCOR
OH
b(R=CH3)
o JR (
' OR
e,f
which is then alkylated or acylated to give the required mesogen. Alternatively, treatment of the hexamethoxy compound with trimethylsilyl iodide gives the hexatrimethylsilyloxy derivative, which can be alkylated or acylated directly [43]. Under the chloranil/concentrated sulfuric acid conditions, longer chain 1,2-dialkoxybenzene derivatives also undergo oxidative trimerization to give the required hexaethers directly, but the products are heavily contaminated with impurities [44] (mainly dealkylation products) which are difficult to remove so that, in general, the three-step procedure has been preferred. An improved synthesis of hexaalkoxytriphenylenes has been developed more recently in our laboratories [45], allowing dialkoxybenzenes to be converted directly to the corresponding triphenylene in a single high-yielding step using ferric chloride as the oxidizing agent. The reaction is typically carried out in dichloromethane solution, is complete in about an hour, and is worked up reductively with methanol. Alternative oxidizing agents used for the synthesis of related triphenylenes include ferric chloride/sulfuric acid [46] and electrosynthesis [47]. Indeed, it was in these electrochemical syntheses that the advantages of and the need for reductive work-up procedures was first recognized 1481.
2.5.1 Ethers, Thioethers and Selenoethers
''
Ro@R RO
OR
OR
Scheme 15. Synthesis of ethers and esters of triphenylene. Reagents: (a) chloranil/70% H, S04/73% [46]; (b) 48% HBr/HOAc/reflux/lS h [42]; (c) RBrlbase [49, 501; (d) RCOCl/base [49, 501; (e) FeCI,/CH,CI, [45]; (f) MeOH/86% (R=CH,) 73% (R=C6H1,)two steps [45].
The mesophase behaviour of simple symmetrical hexaether derivatives of triphenylene is summarized in Table 15. A11 derivatives with side-chains longer than propyloxy form a single hexagonal mesophase, with both the transition temperatures and the mesophase range decreasing as the length of the side chain is increased. The corresponding thioethers and selenoethers are made
2
703
Aromatic Hydrocarbon Cores
Table 15. Transition temperatures for ethers of triphenylene. Phase transitions
Refs.
/
Cr 177 I [44,45,50] Cr 89 Dh, 146 I [44,45, 501 Cr 69 Dh, 122 1 [44, 50, 5 11 Cr 68 Dh, 97 (100) 1 [43, 44, 50 (45)l [43, 44, 501 Cr 69 D,, 93 I [44, 45, 501 Cr 67 Dh, 86 I Cr 57 D,, 78 I [44, 45, 501 Cr 58 Dh, 69 I [44, 501 Cr 54 Dh, 66 1 [43,44, 501 Dh, 49 I ~44,501
Scheme 16. Synthesis of thioethers and selenoethers of triphenylene. Reagents: (a) FeCl, 1531; (b) hv/I,
from the reaction of hexabromotriphenylene [52] and the requisite nucleophile. As shown in Scheme 16, the hexabromide is readily obtained from triphenylene itself. Unfortunately, this is an expensive starting material and troublesome to make on any scale. Of the several routes available [57] perhaps the most convenient for a smallscale laboratory preparation is the ferric chloride oxidative cyclization of o-terphenyl [53], a relatively cheap starting material. Transition temperatures for the thioethers and selenoethers are summarized in Table 16. Interest in the ether derivatives arises from the fact that they are photoconductive and that, when doped with a few mol per cent of an oxidant such as aluminium trichloride, they become semiconductors [59]. Oriented samples of the D, phase can be produced, and it is found that the preferred direction of conduction is parallel to the director; the system behaves rather like a molecular wire, with the conducting stack of aromatic cores being surrounded by an insulating sheath of alkyl chains. The hexahexylthioether derivative gives a 'helical' ordered columnar phase at low temperatures 158, 601, which shows remarkably high charge-carrier mobility in experiments in
[54]; (c) Br2/Fe/C,H,N02/reflux/2 h, [ 5 2 ] ; (d) RSNa/DMEU/10O0C/1-2 h, 40-55% [55]; (e), (i) RSeNa/HMPA/2 h/80"C, (ii) RBr/15 h/60 "C, 38-55% [56].
Table 16. Transition temperatures for thioethers and selenoethers of triphenylene. Substituent
Phase transitions
Refs.
which the charge carriers are generated photochemically [61].
2.5.2 Esters The mesophase behaviour of the discotic triphenylenehexakisalkanoatesis summarized in Table 17. Simple, straight-chain derivatives tend to form a rectangular mesophase with the higher homologues also forming a hexagonal phase at higher temperatures. Melting and clearing temperatures do not alter significantly through the series. The effect of heteroatom substitution into the side
704
VII
Synthesis and Structural Features
chain depends on its nature [9] (oxygen or sulfur) and position [9, 641, and it is worth noting that mesophase behaviour is destroyed if a bromine atom is present in the side chain [65]. The hexabenzoate esters of triphenylene have received particular attention because they tend to form nematic phases. Their mesophase behaviour is summarized in Table 18. Increasing the chain length of the p-alkyl or p-alkoxy substituent in the benzene ring tends to reduce the transition temperatures slightly, but more pronounced effects are observed when lateral (ortho or meta) methyl substituents are introduced into the benzene rings. The number and position of these additional methyl groups influences both the transition temperatures and the phase sequence [68, 691. The corresponding cyclohexane carboxylates (in which the benzene ring is formally reduced
to a trans-l,4-disubstituted cyclohexane) show wider mesophase ranges and they exhibit mesophase behaviour with shorter alkyl chain substituents. This is illustrated in Table 19 [51, 701.
2.5.3 Multiynes As shown in Scheme 17, palladium catalysed cross-coupling of hexabromotriphenylene [52] and p-alkyl derivatives of phenylacetylene gives liquid crystalline hexaalkynyl derivatives of triphenylene [24]. These discogens, which have a core diameter of about 2.4 nm, give enantiotropic nematic phases only (7:R = C,H, Cr 157 N D 237 I; R=C7HI5,Cr 122 N D 176 I). Straightchain alkylacetylene derivatives (like 7,but without the benzene rings) are not liquid crystalline [24].
Table 17. Transition temperatures of esters of triphenylene Substituent
Phase transitions
Cr 146 I Cr I08 D, 120 I Cr 64 D, 130 I (Cr 66 D 126 I) Cr 62 (66) D, 125 (129) I Cr 75 (65) D, 126 I Cr 67 Dra 108 D, 122 I Cr 80 D? 93 D, 1 1 1 D, 122 I Cr 83 Drb 99 D, 118 I Cr 87 D, 96 D, 11 1 I Cr 59 D 65 D,97 I Cr 75 D 138 I Cr 86 D 198 I Cr 57 D 202 I D, 203 I Cr 98 I Cr 94 I Cr -14 Dh 227 1 Cr ? D 93 I Cr ? Dh 156 I a
Gives another columnar phase on cooling to 56 "C. Gives another columnar phase on cooling to 81 "C.
Refs.
2
705
Aromatic Hydrocarbon Cores
Table 18. Transition temperatures for benzoate esters of triphenylene of the general type 5.
C
X
5 X
A-D
Phase transitions
A-D=H A-D=H
Cr 210 I (Cr 130 D 210 I) Cr 208 N, 210 I (Cr 179 N, 192 I) [Cr 125 D, 179 D, 222 I] Cr 175 D, 183 N, 192 I Cr 185 D, 189 I Cr 257 N, 300 I Cr 224 N, 298 I Cr 186 D, 193 N, 274 I Cr 168 N, 2.53 I (Cr 142 N, 248 I) Cr 152 D, 168 N, 244 I Cr 154 D, 183 (180) N, 227 (222) I [Cr 147 D, 182 D2 230 I] Cr 142 D, 191 (155) N, 212 (177) I Cr 145 D, 179 N, 18.5 I Cr 146 D, 174 I 345 I) [70 (51)1
ucts [45, 50, 711. If the two reactants are used in equal proportions the ratio of these products is almost exactly statistical, and often the mixture is very difficult to separate, although in the case of electro-oxidations it is possible to manipulate the product ratio to some extent by varying the reaction conditions [72]. In other cases it may be advantagous to use one reactant in excess, or favourable solubility properties may mean
that separation can be achieved fairly easily [45]. An alternative 'statistical' strategy that has been successfully exploited is that of partial alkylation of the hexa-acetate of triphenylene, as shown in Scheme 19 [41]. The reaction gives a mixture of hexa-alkoxy, penta-alkoxy, tetra-alkoxy, etc., derivatives, but the penta-alkoxy derivative can be isolated in low yield. Hydrolysis and acylation or alkylation of the product has led to a series of penta-alkoxymonoacyloxy and unsymmetrical hexa-alkoxy derivatives. The introduction of a single ester (particularly a branched ester) side chain increases the clearing temperatures, but more notably it surpresses crystallization such that these systems readily form glasses [41, 731. The simplest and most straightforward of the 'rational' syntheses is the oxidative coupling of a tetra-alkoxybiphenyl and a 1,2-dialkoxybenzene derivative (74 - 761 using the ferric chloride/methanol work up protocol shown in Scheme 20. The reaction is easy to perform, offers a good degree of regiocontrol, and has been adapted to yield products in which both the number and nature of the substituents have been varied. Through this reaction a variety of different ring substitution patterns has also been produced [75]. The biphenyl derivatives required for these phenylhiphenyl coupling
6 R
a
Br Br
R 7
Scheme 17. Synthesis of multiyne derivatives of triphenylene [24]. Reagents: (a) RC6H,C = CH/Pd(PPh,),Cl,/PPh,/Cu(l)/ 1OO0C/38h, 52-5770.
2
-
RO R
OAc
\
O\
‘
RO \ RO@
g
\ \
OR
707
Scheme 18. ‘Statistical’ route to unsymmetrically substituted derivatives of triphenylene. Reagents: (a) electrochemical, FeCl,, or chloranil oxidation.
OR2
OR2
Aromatic Hydrocarbon Cores
OR
OR
OR
Scheme 19. ‘Statistical’ synthesis
+ isomeric diacetates
OR OR
gR‘ -
OR‘
OR’
a,b
/ \
OR’
OR2
of unsymmetrically substituted derivatives of triphenylene [41]. Reagents: (a) RBr/K,CO,/ PhCOMe/l16”C/17 h, 26% monoacetate.
R
.
~ 3 \ 0
0
8
R2
\
\
OR2
OR‘
reactions can be prepared either through Ullman or Heck coupling reactions, and the only significant limitation on this strategy arises from a lack of regiospecificity in the coupling step itself, such that, if neither the phenyl nor the biphenyl component has C2 symmetry, a mixture of regioisomers will result. Longer and more elaborate strategies [77,78] have been developed that allow full regiocontrol. In these, ‘Heck’ chemistry is
Scheme 20. ‘Rational’ synthesis of unsymmetrically substituted derivatives of triphenylene [74-761. Reagents: (a) FeCI,/CH,Cl,; (b) MeOH
used to assemble a terphenyl, which is then subjected to oxidative cyclization to yield the triphenylene. One of these routes is illustrated in Scheme 21 [78]. A third, and even longer, route to unsymmetrically substituted triphenylenes was originally developed by Wenz in his synthesis of discotic liquid crystalline polymers [79- 821, but has since been adapted for the synthesis of low molar mass discogens [83]. Scheme 22 il-
708
VII
Synthesis and Structural Features
F O R 2 OR'
OR'
R
5
0
4 R3
b
R60
~50%
R60
'
OR'
\
Scheme 21. 'Rational' route to unsymmetrically substituted derivatives of triphenylene [78]. Reagents: (a) Pd,(dba),/Ph,P/THF/ \
0
OR'
OR'
~
2
reflux/4.5 h, 45%; (b) FeC1,/CH,Cl2/1 h, then MeOH (dba = dibenzylidine acetone, THF = tetrahydrofuran).
Scheme 22. 'Rational' synthesis OC5H11
C
~
C4Hg02C
H
~
Ol 1
~
___)
C
~
' 'I \
OC5H11 OC5H11
lustrates its use in the synthesis of the diester 8, which gives a columnar phase (Dhd) between room temperature and 167 "C and shows very interesting self-organizing behaviour at the phase-interfaces [83 1861. Transition data for unsymmetrically substituted systems are given in Table 21.
8
OCSH1 OC5H11
of unsymmetrically substituted derivatives of triphenylene [79, 831. Reagents: (a) CH,COCH,/'BuOK/EtOHI reflux/4 h; (b) TsOH/C6H,C1/C,H,O2CC = CC0&4H,; (c) 12lhv.
2.5.5 Modifications of the Number and Nature of Ring Substituents Reactions have now been developed that allow the properties of triphenylene hexaeth-
709
2 Aromatic Hydrocarbon Cores
Table 21. Transition temperatures for unsymetrically substituted derivatives of triphenylene of the type 9.
R20
'' '
OR^
OR^
9 R'
R2
R3
R4
Phase transitions
Refs.
Cr 72 Dh 101 1 Cr 55 D, 106 I Cr? D,, 161 1 Cr 44 D,, 177 I D,, 182 I D,, 182 I D,, 191 1 Cr 56 Dh127 I Cr 60 Dh 104 I Cr 51 D, 100 I Cr 32 D,, 55 1 Cr 67 Ia Cr 53 D 72 I Cr 50 D,, 117 I Cr 80 I Cr 54 D,, 74 1 Cr 61 D,, 63 I Cr 61 Cr 128 Dh 139 1 Cr 68 I b Cr 47 Dho 84 1 Cr 47 D, 84 I Cr 51 D 61 I Cr 53 D,, 71 I Cr 56 I Cr 87 D, 99 D, 118 I Cr 65 I' a
Gives D,, phase on cooling to 51 "C. Gives D, phase on cooling to 64 "C. Gives D phase on cooling to 55 "C.
er discogens to be modified either by introducing extra substituents into the aromatic nucleus through electrophilic aromatic substitution or by removing some of the alkoxy groups through selective reduction. The outcome of electrophilic substitution reactions on the unsubstituted hydrocarbon triphenylene seems to be determined by a conflict between electronic effects that favour a attack and steric factors (involving mainly the
interaction with the peri hydrogen atoms) that favour P attack, but normally P attack dominates. In those mesogens in which all the P-positions bear substituents steric crowding must be severe, but the systems can be readily nitrated, chlorinated and brominated [92 -941. The introduction of the nitro substituent is a particularly valuable step since, as shown in Scheme 23, it can be readily transformed into a wide range
710
VII
Synthesis and Structural Features
Table 26. Transition temperatures for hepta-substituted derivatives of triphenylene of the type 11. RO
C6H130@
L
OC6H 1 3 l3
OR
OR
%HI30 \
OC6H13 OC6H13 11
X
if
\ OR
OR
OR
Scheme 23. Modification of the substituents through electrophilic substitution. Reagents: (a) HNO,/Et,O/ AcOH/15 min, 73% 1921; (b) Sn/AcOH/reflux/4 h, 87% 1921; (c) NaNO,/AcOH [94]; (d) KN,, 58% [94]; (e) NaNO,/AcOH, 45%, [92]; (f) R'COCl/K,CO,/ CH,C1,/18 h, 95% (R=Me) 1921.
of other substituents and many of the products show enhanced mesophase properties as detailed in Table 26. Note also those compounds, shown in Scheme 24, that are not themselves mesomorphic but which give mesomorphic nitro derivatives ( e g 12-14, the second and third of which are readily available through variations on the 'biphenyl' route shown in Scheme 20) [93]. Hence 12 shows Cr 96 I, but its nitration product gives a columnar phase on cooling to 88 "C. Compound 13 shows Cr 83 I and its nitration product Cr 81 I, but gives a columnar phase on cooling to 64°C. Compound 14 shows Cr 70 I and gives a columnar phase on cooling to 60 "C, but its nitra-
NO, NH, N, NHCOCH, NHCOC,H, NHCOC3H7 NHCOC,H, NHCOCSH, 1 NHCOC,HI, NHCOC7HIs NHCOCSHI, NHCOC,H,, NHCOC 1 1 H,,
c1
CHZOH
Phase transitions
Refs.
350 I [I911 Cr 119 (102) Dh, >345 1 1188 (192)l Cr 120 Dh, >345 I [I881 Cr 104 Dh, >345 1 11881 Cr 110 Dh, >345 1 [188l Cr 94 Dho ? Idecomp [186, 1881 Cr l2 Dho ? Idecomp [186, 1881 Cr 59 D,, 84 I [ 1891 Cr 101 (107) Dh, >345 1 1188 (189)I Cr 106 Dh, >345 I [I881 [I881 Cr 94 Dho 345 Idecomp Cr 104 D,, >345 I 11881 Cr 58 Dhd 87 I [ 1891 Cr 83 Dho 334 Idecomp [I881 Cr 92 Dh, >345 1 [1881 Cr 83 Dh, 309 I [186, 1881 Cr 95 - 110 D,, 333 I [187, 196, 1971 Cr 75 D,, 269 I [ 1841 Cr 85-91 D 345 I 11871 Cr 96-98 D 254 I [ 1871 Cr 105 D,, 310 I [ 1871 Cr 95 Dho ? Idecomp 11861 Cr 95 Dh, >345 I 11881 Cr 78 Dh, 3 10 1 11841 Cr 99 D 375 I 11871 [I891 Cr 59 Dhd94 I Cr 55 Dhd 107 I ~391 Cr 56 Dh, 89 1 [I891 Cr 56 Dh, 87 I [I891 [190 (194)] Cr 170 (158) Dtetd 223 N, 270 (255) I Cr 204 D, 242 N, 290 I [194, 1951 Cr 300 I
Scheme 54. Synthesis of peripherally substituted [209]. Reagents: tetrabenzotriazaporphyrins (a) CH,(CH,),,MgCVether; (b) H+; 19% overall.
n= 3: Cr 78 &,>300 I n= 8 : Cr 11 Dho 170-180 I
stable mesophase (possibly with lamellar structure [2lo]). The synthesis presumably yields 47 as a mixture of regioisomers, which could account for the broad melting range.
3.9.8 Tetrakis[oligo(ethyleneoxy )]phthalocy anine McKeown and Painter 12111 have synthesized mesomorphic phthalocyanines bearing four oligoethyleneoxy substituents in the peripheral positions. Their synthesis is shown in Scheme 55 along with the mesophase behaviour of the final products. The ethyleneoxy substituted phthalonitrile is prepared via a base-catalysed nitro displacement of 4-nitrophthalonitrile [212] with polyethylene glycol monomethyl ether. Cyclization under normal conditions yields the phthalocyanines, presumably as a mixture of regioisomers. The compounds give a single, thermotropic, hexagonal mesophase, as well as a range of lyotropic mesophases in water.
3.9.9 Non-Peripherally Substituted Octa(alkoxymethy1)phthalocyanines In general, the synthesis of the non-peripherally substituted phthalocyanines is less
Scheme 55. Synthesis of tetrakis[oligo(ethyleneoxy)]phthalocyanines [21 I]. Reagents: (a) K,CO,/DMF; (b) C,H,,OLi/C,H, IOH, reflux; (c) H+.
straightforward than their peripherally substituted counterparts, requiring routes to 1,4-disubstituted phthalonitriles. For the liquid crystalline materials, this has been achieved using variations on a synthesis which uses the Diels- Alder reaction between a furan or thiophene- 1,l-dioxide and fumaronitrile (Scheme 56). The synthesis of non-peripherally substituted octa(alkoxymethy1)phthalocyanines is outlined in Scheme 57 [213, 2141. 2,5bis(Alkoxymethy1)furan can be prepared by lithiation of furan followed by quenching the anion with (bromomethy1)alkyl ether to give the mono-substituted furan, and repetition of the procedure to give the required product [213]. A more simple synthesis involves chlorination of 2,5-furan dimethanol with thionyl chloride, followed by nucleophilic displacement of chloride with sodium alkoxide [214]. The furan is equilibrated with fumaronitrile for about a week and the Diels- Alder adduct aromatized by treatment with lithium bis(trimethylsily1)amide (a non-nucleophilic base), followed by an acidic work-up. The 3,6-bis(alkoxymethy1)phthalonitrile is cyclized under standard conditions (lithium/pentanol).
3 Heterocyclic Cores
737
Table 51. Transition temperatures for non-peripherally substituted octa(alkoxymethy1)phthalocyanines (ROCH,-) (see Scheme 57 for structures). R
M
C4H9
H, Cr 185 D,, >300 I HZ Cr 123 D,, >300 I HZ Cr 85 D,, >300 I H, Cr 79 Dr, ? (90- 120) Dh 4300 I cu Cr 84 D >300 I Zn Cr 70 D > 300 I Hz Cr 67 D,, ? (80 - 220) Dh > 300 I HZ Cr 75 Dh, >300 I H, Cr 64 D,,j > 300 I Hz D,, 236 I H, D,, 106 I
Phase transitions [214, 2151
R
R
C5H11 C6H 13
C7H15 C7H15
Scheme 56. The synthesis of 1,il-disubstituted phthalonitriles via a Diels- Alder reaction between a furan or thiophene- I , I -dioxide (diene) and fumaronitrile (dienophile).
11 1
h,i
C7H15
C8H17 C9H I9
ClOHZI CH,CH(Et)C,H, CH(Me)C,H,,
oxymethyl) form both, with the hexagonal phase being formed at higher temperatures. Chain branching leads to derivatives (produced as mixtures of diastereomers) that are liquid crystalline (rectangular) at room temperature. These low melting points contrast with the peripherally substituted octa(2-ethylhexyloxy)phthalocyanine, which has a significantly higher melting point than its straight-chain isomer.
R O C H ~ ~ C H ~ O R
3.9.10 Non-Peripherally Substituted Octa-alkylphthalocyanine ROCH~-(&CH~OR
Scheme 57. Synthesis of non-peripherally substituted octa(alkoxymethy1)phthalocyanines [214]. Reagents: (a) SOCl,, (b) RONdROH, 36-78% over two steps; (c) BuLi; (d) ROCH,Br, 20% overall (C& (e) fumaronitrile; (f) LiN(SiMe&/THF; (g) H', 4-23%; (h) ROLi/ROH, A; (i) H+, 7-34%; (j) M' acetate, 72-87%.
It can be seen from Table 51 that shortchain derivatives form rectangular mesophases and longer homologues form hexagonal mesophases. The intermediate compounds (n-heptyloxymethyl and n-octyl-
3,6-Dialkylphthalonitriles have been prepared fromboth 2,5-dialkylfuran [213,216] and 2,5-dialkylthiophene [216]. Both heterocycles are synthesized in a similar manner by lithiation and alkylation of either furan or thiophene, respectively. Generally, the thiophene route is preferred because dialkylation can be performed in a single step, and the final Diels - Alder reaction (between fumaronitrile and the corresponding sulfone) goes to completion and yields the required phthalnotriles without the need for a strong base or inert conditions. The dialkylthiophenes can be oxidized to the correspond-
738
VII
Synthesis and Structural Features
?
Scheme 58. Synthesis of non-peripherally substituted octaalkylphthalocyanines. Reagents: (a) BuLi; (b) RBr; (c) fumaronitrile; (d) LiN(SiMe,),/THF, [213]; (e) H', 25-30%; (f) [O], 35-51% [216]; (8) fumaronitrile, A, 40-48% [216]; (h) C,H,,OLi/ C5H,,0H, reflux; (i) H+, 18-25%; (j) M' acetate, 80-94%.
ing sulfones using either m-CPBA, sodium perborate or dimethyldioxirane [2171. The phthalonitriles are cyclized using the lithium/pentanol method. The synthesis is shown in Scheme 58. An homologous series of compounds, each substituted with a variety of central metal ions, has now been synthesized. Their mesophase behaviour is summarized in Table 52. A number of trends are apparent. Melting points decrease as the side-chain length increases, the effect being most marked for the lower homologous. At longer chain lengths (octyl and above) a pronounced odd-even fluctuation is observed. Clearing temperatures decrease linearly with increasing chain length. The central ion(s) also has a significant influence on mesophase behaviour. The largest mesophase stabilization is observed when zinc is incorporated in the discogen, giving derivatives with higher transition temperatures
4
of mesophase stabilization, the order is Zn>Co>Cu>Ni, H, [217]. This class of phthalocyanine discogens shows a rich polymesomorphism. All samples give an hexagonal phase when cooled from the isotropic liquid. In some cases, further cooling leads to a second hexagonal or a rectangular phase, and, more rarely, an additional, monotropic re-entrant hexagonal phase is observed. It is interesting to note that the closely related non-peripherally substituted octa-alkoxyphthalocyanines and non-peripherally substituted octa-alkoxyphthalocyanines are non-mesomorphic [217].
3.9.11 Unsymmetrically Substituted Phthalocyanines A number of unsymmetrically substituted phthalocyanines have been synthesized and found to be mesogenic. Their synthesis is generally difficult, involving cyclization of a mixture of phthalonitrile derivatives, followed by tedious separation of the products (Scheme 59). ROCH,
CHpOR
d M
or
NH
Saturated Cores
739
NH
+ many other products Scheme 59. Routes to unsymmetrically substituted phthalocyanines. Reagents: (a) RO-; (b) N,N-dimethylaminoethanol, reflux.
nine 48 and found that the compound has a remarkably stable hexagonal mesophase. Cook and coworkers [216] have synthesized a range of unsymmetrical, non-peripherally substituted phthalocyanines 49. Their mesophase behaviour (Table 53) is extremely sensitive to the degree of asymmetry (substituents) in these non-peripherally substituted systems. Mesophase behaviour is destroyed if the diester 50 is saponified to give the diacid (R2= R, = CH2CH2CH2C02H).Similarly, substitution of an additional alkyl group in the penta-alkyl mono-3-hydroxypropylphthalocyanines leads to materials (R2= R, = (CH,),OH) that give only monotropic mesophases [2 17, 22 I].
4 Saturated Cores H
ROCHp CHpOR 48 R = C,,H,,.Cr40 Dh>300 I
Simon and coworkers [ 180, 2201 have synthesized peripherally substituted dicyanohexakis(dodecyloxymethy1)phthalocya-
49
4.1 Cyclohexane The synthesis of discotic liquid crystalline cyclohexanes is relatively straightforward. Both classes of derivative (the scylloinosito1 ethers and esters) are synthesized from commercially available scylloinositol (Scheme 60). The esters are prepared by treatment with an excess of acid chloride in trifluoroacetic acid [222], whereas the ether is synthesized using potassium hydroxide and hexyl bromide [223]. The mesophase
740
VII
Synthesis and Structural Features
Table 53. Transition temperatures for unsymmetrically substituted phthalocyanines 49. R,
R2
R3
CIOH21 (50) C6H13
C7H I5 C8H17
Cd2I C6H 13
C7H15 C8H I7 C9H19
CinH21
R4, R5
Phase transitions
Refs.
H H H H H -OC(Me),O-OC(Me),O- OC(Me),O- OC(Me),O -OC(Me),O-
Cr 71 D 120 I Cr 94 D 148 I Cr75D7117Dh 1291 C r 5 5 D 1 7 4 D h 1121 Cr 39 D 82 I Cr 244 D 259 I Cr 186 D7 235 Dh 237 I Cr 183 D, 208 Dh 226 I Cr 155 D, 192 Dh 216 I Cr 138 D, 180 D, 207 I
[2 161 [217, 2211 [2211 [2211 [2171 1217, 2211 [221l [22 11 [22 11 [217,221]
behaviour of these materials (Table 54) has a number of interesting features. The fact that mesogenic materials are produced with side chains as short as acetyl is itself suprising. Comparing the ether and ester derivatives with similar side-chain lengths reveals a much enhanced mesophase stability for the ester. It is worth noting that neither the myoinositol ester (51) [224] or the mytilitol derivative (52) [224], each bearing an axial substituent, gives any discotic mesophases. Many related hexa (equatorial) substituted cyclohexanes have been synthesized (e.g. 53) in which the oxygen linkage to the ring has been replaced by carbon, but to date all are non-mesogenic [225]. R
RO
R2&XI
OR 52
OR 51 R = COC5H11:Cr 48 I
R = COC9H19:Cr 75 I
53 R = C6Hl,: Cr 66 I
6H
Scheme 60. Synthesis of cyclohexane (inositol) based discogens. Reagents: (a) KOH/RBr; (b) RCOCUTFA.
Table 54. Transition temperatures for cyclohexane (inositol) based discotic mesogens (see Scheme 60 for structures). Substituent
Phase transitions
Cr 18 Dh, 91 I Cr 288 D 292 I Cr 212 D 276 I Cr 213 D 259 I Cr 185 D 208 I Cr 69 D 200 I Cr 70 D 202 I Cr 76 D 199 I Cr 81 D 196 I Cr 84 Dh 189 I Cr 88 D 183 I Cr 92 D 176 1 0C0C11H23 OCOC,H4C9Hl, Cr 184 H 191 I Cr 42 D 116 I OCOCH(Me)C,H, OCOCH2CH(Me)C5Hll Cr 48 D 195 I D 18 1 I OCOC2H,CH(Me)C4H9 OCOC3H,CH(Me)C3H7 D 186 I OCOC4H,CH(Me)C2H, Cr 54 D 197 I OCOC5Hl,CH(Me)CH3 Cr 87 D 201 I OCOC4HsCH(Et)C2H5 Cr 38 D 186 I OCOC4HgCp Cr 139 D 199 I OCOC3H6CH(Me)C2H5 Cr 68 D 196 I OCOCSHllCH(Me)C2HS Cr 44 D 192 I OCOC,H,OC,H, Cr 47 D 87 I OCOC,H,oOCH3 Cr 44 D 144 I OCOCH2SC5HIl Cr 57 D I06 I OCOC,H,SC,H, Cr 86 D 183 I OCOC,H$C,H, Cr 64 D 189 I OCOCSH,oSCH, Cr 56 D 172 I OCO(CH,),Cl Cr 205 D 235 I OCO(CH2),Br Cr 190 D 250 I OCO(CH,),Br Cr 93 D 230 I OCO(CH,),Br Cr 94 D 199 I OCO(CH,),Br Cr 92 D,, 175 I OCO(CH,),,Br Cr 90 D,, 117 I 0C6H I3
OCOCH, OCOC2H5 OCOC3H7 OCOC4H9 OCOC5H, I OCOC,H,, OCOC7H15 OCOCsH17 OCOC9Hl9 ococ,OH21
Refs.
74 1
4 Saturated Cores
Collard and Lillya have made systematic changes to the side chains of cyclohexanebased discotic mesogens by introducing branching [ 101 and heteroatoms [9]. Introduction of a methyl group into the side chain tends to reduce the melting point, the largest effect being observed when the branch point is in the middle of the side chain. Moving the branch point closer to the core tends to lower the clearing temperature, presumably because steric effects begin to disfavour a columnar arrangement. Introduction of a heteroatom into the chains also tends to reduce both transition temperatures, with the effects depending on its nature (oxygen or sulfur) and position. Halogen substitution at the chain terminus stabilizes the mesophase for lower homologues, but a destabilization is observed for longer chain derivatives [227].
‘z0-
RCO2CHp
RC02CH2 OCOR
RCOO H
OCOR 54 a-glycopyranosidederivative
55 pglycopyranoside derivative
RC02CH2 RCO2 H 56 cellobiose derivative 57 chitobiose deivative
RC02CH2
RRCOZc
X = OCOR X = NHCOR
RC02CH2
oX 2
RC02CH2 ~ o
~
o
&
$
X
$H, OCOR
H
58 cellotriose derivative 59 chitotriose deivative
X = OCOR X = NHCOR
R R@C 6O ~* COHa~ O C 1 ~ H , ,
4.2 Tetrahydropyran (Pyranose Sugars) The pyranose sugars provide a ready-made source of laterally substituted cyclic nuclei and hence of potential discotic systems and the esters 54-61 are all mesogenic. All are made simply by treating the appropriate natural sugar with the required acid chloride in pyridine/chloroform for 1-3 days [228]. The C,o to CISesters of the a- and p-glucopyranosides 54 and 55 show complex phase behaviour which is found to depend on the thermal history of the sample as well as the structure of the mesogen. There is a degree of disagreement in the literature over the phase behaviour [228 - 2301. In view of the fact that these systems represent fairly minor deviations from the six-fold symmetric inositol structures, the formation of D, phases does not seem too surprising. Much
RCOO
RC02CH2
n 60 cellobiose derivative, X = RC02 61 cellobiose derivative, X = HO
less expected and less easy to understand is that the cellobiose 56 [23 1,2321, chitobiose 57 [233], cellotriose 58 [231, 2321 and chitotriose 59 [231] derivatives behave in a similar way. It is proposed that in the columnar phases formed by these di- and trisaccharides the pyran rings lie across the column, as in a conventional discotic phase, but that in the hexagonal columnar phases formed by higher oligomers, such as the pentamer in the cellulose series, the polysaccharide backbone lies parallel to the axis of the column 12341. The mesophase behaviour of the esters of cellubiose is summarized in Table 55 [23 11. The chitin-based oligomers generally give wider mesophase ranges than do the cellulose-based oligomers [233].
742
VII
Synthesis and Structural Features
Table 55. Transition temperatures of liquid crystalline sugar derivatives. Sugar 54
55 56
Acyl substituent C9H19C0 C9Hl 9c0
C6H13C0 C7H15C0 CXH17C0 C9H19C0
C9H19CO
c I IH&O 58 57
13H27C0 C9H19C0 C9H19C0
C13H27C0
59
CI7H33CO C9H19CO C13H27C0 C17H33C0
60
CIlH23CO C12H250C6H4CO
61
C12H250C6H4C0
Phase transitions
Refs.
Cr 27/8 D 34 I Cr 32/3 D 38/40 I (Cr 40 I) Cr4D91 I Cr 25 D 62 D,, 87 I Cr 37 D,, 86 I Cr 43 Dh, 83 I Cr 58 Dh, 93 I Cr 49 Dho81 I Cr 56 D,, 73 I Cr 45 D, 93 I Cr 65 M 207 I Cr 70 M 196 I Cr 73 M 185 I Cr 60 M 195 I Cr 65 M 175 I Cr 70 M 170 I Cr 74 D 93 I C r ? D 1201 Cr ? D 145 I
~301 [230 (235)] ~2311 12311 12311 12311 [232, 2341 12311 12311 [232, 2341 12331 W31 V331 “2331 P331 W31 ~291 ~2291 r2291
4.3 Hexacyclens and Azamacrocycles Mesophase behaviour of hexa-n-acylated hexacyclens (62) was first observed by Lehn et al. [236] for the hexa-(p-n-dodecyloxybenzoyl) derivative. A columnar arrangement was assumed, but the lack of high-order X-ray reflections prohibited absolute assignment. Nevertheless, the term ‘tubular mesophase’ (T) was introduced to describe the proposed structure, which is wholly analogous to that of the D, mesophase. In the case of hexacyclens (and similar compounds), however, the ‘cores’ possess a central void region, giving rise to supramolecular ‘tubes’. Many other derivatives have been prepared, but in only two cases (63 and 64, see Table 56) has the hexagonal columnar structure been confirmed [237]. More evidence to support the columnar structure 65 (Table 56) has come from X-ray measurements of mixtures of 65 with brominated analogues [238].
The hexacyclen mesogens are easily synthesized from commercially available hexacyclen by condensation with a benzoyl chloride in dimethylacetamidelp-dimethylaminopyridine [236] (Scheme 61). The mesophase behaviour of the discotic hexacyclens is shown in Table 56. It should be noted that an additional feature of these materials is that their thermal behaviour is dependent on the history/recrystallization conditions of the sample, and different transition temperatures are reported from different laboratories [239].
62 18aneN~
63: R = (3-Br-4-C,,H,,0)C6H,CH:CHC0 64: R = (3,5-CI~-4-Cl,-H,50)C6H4CH:CHC0 65: R = C12H,,0C,H4C0
Scheme 61. Acylation of hexacyclen. Reagents: (a) R’COCUbase.
5
References
743
Table 56. Transition temperatures for hexacyclen based discotic mesogens 62. R C,H,7OC6H.+CO c l lH230C6H4C0
C I ~ H ~ % H , C O(65) C,4H290C6H4C0 (3-Br-4-C I ,H,,O)C,H,CH:CHCO
(63) (3,5-CI+-Cl 2H,,O)C,HzCH:CHCO (64) C,4H,,C6H,CH:CHCO CaH,,OC,H,N:NC,H,CO 3,4-(C,&2,O)2C&,CO a
Phase transitions a
Refs.
Cr 120 T 139 I Cr ? T 144 I Cr 122 (97) T 142 I Cr 106 D,, 136 I Cr ? D, ? I Cr ? D, ? I Cr 217 D, 233 I b Cr 237 D 245 I Cr 104 D 140 I
124I 1 12401 [236 (240)] ~421 12371 12371 W31
12421 W41
T, tubular mesophase. Mesophase behaviour dependent on E/Z ratio.
R
66 12aneNa Cr 38 D
w 112441 R=
~
R 67 9aneN3
68 14aneN4
D 44 D 66 1[244]
D95 0 132 If2451
~
~
~
H
2
,
The presence of the benzamide moiety appears essential for mesophase production. If the benzamides are reduced to the benzylamines [240], or straight-chain amides ( C , ,H,,CO) [241] are used, all mesophase behaviour disappears. Some smaller ring azamacrocycles 66-68 have been prepared and found to give discotic phases. Their synthesis is wholly similar to that of the hexacyclens. Benzamides of larger macrocycles (e.g. [30] N,, [241]) have also been synthesized, but these show only monotropic mesophases.
5 References [ I ] S . Chandrasekhar, B. K. Sadashiva, K. A. Suresh, Pramana, 1977, 9, 471. [2] P. G . DeGennes (1974) The Physics of Liquid Crystals, Oxford University Press, Oxford. 131 J. D. Brooks, G. H. Taylor, Carbon, 1965,3, 185.
[4] C. Destrade, N. H. Tinh, H. Gasparoux, J. Malthete, A. M. Levelut, Mol. Cryst. Liq. Cryst., 1981, 71, 111-135. [ 5 ] S. Chandrasekhar, B. K. Sadashiva, K. A. Suresh, N. V. Madhusudana, S. Kumar, R. Shashidhar, G. Venkatesh, J. Phys. Colloq., 1979, C120. [6] A. J. Fatiadi, W. F. Sager, Org. Synth. 1973, 595. [7] I. Tabushi, K. Yamamura, Y. Okada, J. Org. Chem. 1987,52,2502-2505. [8] A. Maliniak, S. Greenbaum, R. Poupko, H. Zimmermann, Z. Luz, J. Phys. Chem. 1993, 97, 4832-4840. [9] D. M. Collard, C. P. Lillya, J. Org. Chem. 1991, 56,6064 - 6066. [lo] D. M. Collard, C. P. Lillya, J. Am. Chem. Snc. 1989, I l l , 1829-1830. [ 1I] C. P. Lillya, R. Thakur, Mol. Cryst. Liq. Cryst. 1989,170, 179-183. [12] A. J. Fatiadi, W. F. Sager, Org. Synth. 1973, V , 1011. [13] 0. B. Akopova, G . G. Maidachenko, G . A. Tyuneva,L. S. Shabyshev, J. Gen. Chem. U.S. S. R., 1984,54, 1657- 1659. [14] 0. B. Akopova, G. A. Tyuneva, L. S. Shabyshev, Y. G . Erykalov, J. Gen. Chem. U. S. S . R., 1987,57,570-574.
744
VII
Synthesis and Structural Features
[I51 Y. Kobayashi, Y. Matsunaga, Bull. Chem. SOC. Jpn, 1987,60,3515-3518. [16] R. Nietzki, T. Benckiser, Berichte, 1885, 18, 499- 5 15. [17] A. Fitz, Berichte, 1875, 8, 631-634. [ 181 H. Kawada, Y. Matsunaga, Bull. Chem. Soc. Jpn, 1988,61,3083-3085. [I91 Y. Harada, Y. Matsunaga, Bull. Chem. Soc. Jpn. 1988,61,2139-2741. [20] F. Kehrmann, N. Poehl, Helv. Chim. Acta 1926, 9,485-491. [21] J. Kawarnata, Y. Matsunaga, Mol. Cryst. Liq. Cryst., 1993, 231, 79-85. [22] Y. Matsunaga, N. Miyajima, Y.Nakayasu, S . Sakai, M. Yonenaga, Bull. chem. Soc. Jpn. 1988, 61, 207. [23] S. Takenaka, K. Nishimura, S. Kusabayashi, Mol. Cryst. Liq. Cryst. 1984, 111, 227-236. [24] K. Praefcke, B. Kohne, D. Singer,Angew. Chem. Int. Ed. Engl. 1990, 29, 177-179. [25] B. Kohne, K. Praefcke, Chimia, 1987, 41, 196-198. [26] M. Ebert, D. A. Jungbauer, R. Kleppinger, J. H. Wendorff, B. Kohne, K. Praefcke, Liq. Cryst. 1989,4, 53. [27] K. Praefcke, D. Singer, B. Gundogan, K. Gutbier, M. Langner, Ber: Bunsenges. Phys. Chem. 1993,97, 1358-1361. [28] D. M. Kok, H. Wynberg, W. H. De Jeu, Mol. Cryst. Liq. Cryst. 1985, 129, 53-60. [29] V. W. Poules, K. Praefcke, Chem. Z., 1983,107, 310. [30] K. Praefcke, W. Poules, B. Scheuble, R. Poupko, Z. Luz, Naturjorsch., Teil b, 1984,39,950-956. [31] M. Sarkar, N. Spielberg, K. Praefcke, H. Zirnmermann,Mol. Cryst. Liq. Cryst., 203, 159-169. [32] N. Spielberg, Z. Luz, R. Poupko, K. Praefcke, B. Kohne, J. Pickard, K. Horn, Z. Naturforsch., Teil a, 1986, 41, 855-860. [33] K. Praefcke, B. Kohne, K. Gutbier, N. Johnen, D. Singer, Liq. Cryst. 1989, 5, 233-249. [34] A. P. Krysin, N. V. Bodoev, V. A. Koptyug, J. Org. Chem. U. S. S. R., 1977,13, 1183-1 186. [35] J. Billard, J. C. Dubois, C. Vaucher, A. M. Levelut, Mol. Cryst. Liq. Cryst., 1981, 66, 115 - 122. [36] J. Grimshaw, R. D. Haworth, J. Chem. Soc., 1956,4225-4232. [37] G. Carfugua, A. Roviello, A. Sirigu, Mol. Cryst. Liq. Cryst. 1985, 122, 151-160. [38] C. Carfagna, P. Iannelli, A. Roviello, A. Sirigu, Liq. Cryst. 1987, 2,611 -616. [39] G. Scherowsky, X. H. Chen, Liq. Cryst. 1994, 17,803-810. [40] I. M. Matheson, 0. C. Musgrave, C. J. Webster, J. Chem. SOC.,Chem. Cornmun. 1965,278-219. [41] W. Kreuder, H. Ringsdorf, Mukromol. Chem. Rapid Commun. 1983,4,807 - 8 15. [42] M. Piattelli, E. Fattorusso, R. A. Nicolaus, S. Magno, Tetruhedron 1965, 21, 3229- 3236.
[43] L. Y. Chaing, C. R. Safinya, N. A. Clark, K. S. Liang, A. N. Bloch, J . Chem. Soc., Chern. Commun., 1985,695-696. [44] C. Destrade, M. C. Mondon, J. Malthete, J . Phw. Colloq., 1979, C3, 17-21. [45] N. Boden, R. C. Borner, R. J. Bushby, A. N. Cammidge, M. V. Jesudason, Liq. Cryst. 1993, 15,851-858. [46] 0. Karthaus, H. Ringsdorf, V. V. Tsukruk, J. H. Wendorff, Lungmuir, 1992, 8, 2279-2283. [47] A. Ronlan, B. Aalstad, V. D. Parker, Acta Chem. Scand., 1982,36B, 317-325. [48] V. LeBerre, L. Angely, N. Simonet-Gueguen, J. Simonet, J . Chem. SOC.,Chem. Commun., 1987, 984-986. [49] C. Destrade, M. C. Mondon-Bernaud, N. H. Tinh, Mol. Cryst. Liq. Cryst. Lett. 1979,49, 169. [50] C. Destrade, N. H. Tinh, H. Gasparoux. J. Malthete, A. M. Levelut, Mol. Cryst. Liq. Cryst., 1981, 71, 1 I 1 - 135. [51] C. Vauchier, A. Zann, P. IeBarny, J. C. Dubois. J. Billard, Mol. Cryst. Liq. Cryst., 1981,66,103114. [52] R. Breslow, B. Jaun, R. Q. Kluttz, C.-Z. Xia, Tetrahedron, 1982, 38, 863 -867. [53] R. C. Borner, Ph. D. thesis, Leeds, 1992. [54] W. H. Laarhoven, Rec. Trav. Chim. Pays Bas, 1983,102, 185-204 and 241-254. [55] B. Kohne, W. Poules, K. Praefcke, Chem. Z. 1987,108, 113. [56] B. Kohne, K. Praefcke, T. Derz, W. Frischmuth, C. Gansau, Chem. Z., 1984,108,408. [57] C. M. Buess, D. D. Lawson, Chem. Rev., 1960, 60,313-330. [58] E. F. Gramsbergen, H. J. Hoving, W. H. de Jeu, K. Praefcke, B. Kohne, Liq. Cryst. 1986, 1. 397 -400. [59] N. Boden, R. J. Bushby, J. Clements, M. V. Jesudason, P. F. Knowles, G. Williams, Chem. Phys. Lett., 1988, 152, 94-99. [60] E. Fontes, P. A. Heiney, W. H. de Jeu, Phys. Rev. Lett., 1988, 61, 1202-1205. [61] D. Adam, P. Schuhmacher, J. Simrnerer, L. Haussling, K. Siemensrneyer, K. H. Etzbach, H. Ringsdorf, D. Haarer,Nature 1994,371,14I -143. [62] A. M. Levelut, P. Oswald, A. Ghanen, J. Malthete, J. Phys. 1984,45, 745 - 754. [63] H. Bock, W. Helfrich, Liq. Cryst. 1992, 12, 697-703. [64] I. Tabushi, K. Yamamara, Y. Okada, Tetrahedron Lett., 1987,28, 2269-2212. [65] C. P. Lillya, D. M. Collard, Mol. Crysf. Liq. Cryst., 1990,1828,201-207. [66] N. H. Tinh, H. Gasparoux, C. Destrade. Mol. Cryst. Liq. Cryst. 1981,68, 101- 11 1. [67] N. H. Tinh, C. Destrade, H. Gasparoux, Phys. Lett. 1979, 72A, 251 -254. [68] T. J. Phillips, J. C. Jones, D. G. McDonnell, Liq. Cryst. 1993, 15, 203-215.
5 References 1691 P. Hindmarsh, M. Hird, P. Styring, J. W. Goodby, J . Muter. Chem., 1993,3, 1117-1128. [70] D. R. Beattie, P. Hindmarsh, J. W. Goodby, S. D. Haslam, R. M. Richardson, J. Muter. Chem., 1992,2, 1261- 1266. 1711 N. H. Tinh, M. C. Bernaud, G. Sigand, C. Destrade, Mol. Cryst. Liq. Cryst., 1981, 65, 3073 16. [721 J. M. Chapuzet, N. Simonet-Gueguen, I. Taillepied, J. Simonet, Tetrahedron Lett., 1991, 32, 7405 - 7408. 1731 M. Werth, S. U. Vallerien, H. W. Spiess, Liq. Cryst. 1991, 10, 759-770. [74] N. Boden, R. J. Bushby, A. N. Cammidge, J. Chem. Soc., Chem. Commun. 1994,465-466. I751 N. Boden, R. J. Bushby, A. N. Cammidge, G. Headdock, Synthesis, 1995,31-32. [76] N. Boden, R. J. Bushby, A. N. Cammidge, J. Am. Chem. Soc., 1995,117,924-927. 1771 R. J. Bushby, C. Hardy, J. Chem. Soc., Perkin Trans. I 1986, 721 -723. 1781 R. C. Borner, R. F. W. Jackson, J. Chem. Soc., Chem. Commun., 1994,845-846. 1791 G. Wenz, Mukromol. Chem. Rapid Commun. 1985,6,577- 584. [SO] H. Ringsdorf, B. Schlarb, J. Venzmer, Angew. Chem. Int. Ed. Engl., 1988,27, 113-158. 1811 W. Kreuder, H. Ringsdorf, P. Tschirner, Makromol. Chem. Rapid Commun., 1985,6,367-373. 1821 H. Ringsdorf, R. Wusterfeld, Phil. Trans. R. Soc., Ser A 1990,330,95- 108. (831 J. Y. Josefowicz, N. C. Maliszewskyj, S. H. J. Idziak, P. A. Heiney, J. P. McCauley, A. B. Smith, Science 1993,260, 323-326. [84] N. C. Maliszewskyj, P. A. Heiney, J. Y. Josefowicz, J. P. McCauley, A. B. Smith, Science 1994,264,77- 79. [SS] N. C. Maliszewskyj, P. A. Heiney, J. K. Blasie, J. P. McCauley, A. B. Smith, J. Physique (Sex 2) 1992,2,75- 85. 1861 E. Orthmann, G. Wegner, Angew. Chem. Int. Ed. Engl., 1986, 12, 1105. 1871 P. Henderson, H. Ringsdorf, P. Schuhmacher, Liq. Cryst. 1995, 18, 191-195. [88] N. Boden, R. J. Bushby, A. N. Cammidge, G. Headdock, UK Patent 9411828.8,1994. [89] W. Kranig, B. Huser, H. W. Spiess, W. Kreuder, H. Ringsdorf, H. Zimmermann, Adv. Muter. 1990,2, 36-40. [90] F. Closs, L. Haussling, P. Henderson, H. Ringsdorf, P. Schuhmacher, J. Chem. SOC., Perkin Trans. I, 1995, 829-837. I911 5 . W. Goodby, M. Hird, K. J. Toyne, T. Watson, J. Chem. Soc., Chem. Commun., 1994, 17011702. 1921 N. Boden, R. J. Bushby, A. N. Cammidge, Liq. Cryst. 1995,18,673-676. 1931 N. Boden, R. J. Bushby, A. N. Cammidge, G. Headdock, J. Muter Chem. 1995,5,2275- 228 1 .
745
1941 N. Boden, R. J. Bushby, A. N. Cammidge, G . Headdock, UK Patent 9405793.2, 1994. 1951 P. Henderson, S. Kumar, J. A. Rego, H. Ringsdorf, P. Schuhmacher, J. Chem. SOC., Chern. Commun. 1995, 1059. 1961 F. G. Mann, M. J. Pragnell, J. Chem. Soc., 1965,4120-4 127. [97] W. J. Musliner, J. W. Gates, J. Am. Chem. Soc., 1966,88,4271-4273. [98] 0. C. Musgrave, C. J. Webster, J. Chem. Soc.. Chem. Commun., 1969,712-713. 1991 C. Goltner, D. Pressner, K. Mullen, H. W. Spiess, Angew. Chem. Int. Ed. Engl. 1993, 32, 1660-1 662. [loo] U. Anton, C. Goltner, K. Mullen, Chem. Ber, 1992,125,2325-2330. [I011 C. Destrade, H. Gasparoux, A. Babeau, N. H. Tinh, Mol. Cryst. Liq. Cryst., 1981,67,37-48. [I021 C. Destrade, J. Malthete, N. H. Tinh, H. Gasparoux, Phys. Lett., 1980, 78A, 82-87. [I031 J. Chenault, C. Sainson, Compt. Rend. C, 1978, 287, 545-547. [I041 N. H. Tinh, P. Foucher, C. Destrade, A.M. Levelut, J. Malthete, Mol. Cryst. Liq. Cryst. 1984, I l l , 277-292. [lo51 W. K. Lee, B. A. Wintner, E. Fontes, P. A. Heiney, M. Ohba, J. N. Haseltine, A. B. Smith, Liq. Cryst. 1989,4, 87 - 102. [I061 C. Destrade, P. Foucher, J. Malthete, N. H. Tinh, Phys. Lett., 1982,88A, 187- 190. [lo71 E. Fontes, P. A. Heiney, M. Ohba, J. N. Haseltine,A. B. Smith, Phys. Rev.A, 1988,37,13291334. [IOS] N. N. Tinh, J. Malthete, C. Destrade, Mol. Cryst. Liq. Cryst. Lett., 1981, 64, 291-298. 11091 N. H. Tinh, J. Malthete, C. Destrade, J. Phys. Lett., 1981,42, L417-L419. [110] E. Keinan, S. Kumar, R. Moshenberg, R. Ghirlando, E. J. Wachtel, Adv. Mat., 1991, 3, 25 1 254. [I 111 H. Zimmermann, R. Poupko, Z. Luz, J. Billard, Z. Nuturforsch., Teil a, 1995,40, 149- 160. [112] H. Zimmermann, R. Poupko, Z. Luz, J. Billard, Z. Nuturforsch., Teil a, 1986, 41, 11371140. [ I 131 J. Malthete, A. Collet, J. Am. Chem. Soc. 1987, 109,7544-7545. [ I 141 R. Poupko, Z. Luz, N. Spielberg, H. Zimmermann, J. Am. Chem. Soc., 1989, I l l , 60946105. [115] G. M. Robinson, J. Chem. Soc., 1915, 267-276. 11161 A. Oliverio, C. Casinovi, Ann. Chim. Roma, 1952,42, 168- 184; 1956,46,929-933. [117] F. Bertinotti, V. Carelli, A. Liquori, A. M. Nardi, Struct. Chimi. Suppl. Ricerca Sci. 1952, 22, 65. [I181 A. S. Lindsey, Chem. Znd., 1963,823-824. [119] A. S. Lindsey, J. Chem. Soc., 1965, 1685.
746
VII
Synthesis and Structural Features
[120] J. Malthete, A. Collet, Nouv. J. Chim., 1985, 9, 151- 153. [I211 A . M . Levelut, J. Malthete, A. Collet, J. Phys. Paris, 1986,47, 351 -357. [I221 A. Collet, Tetrahedron, 1987, 43, 57255759. 1231 C. Garcia, J. Malthete, A. Collet, Bull. Soc. Chim. France, 1993,93-95. 1241 J. D. White, B. D. Gesner, Tetrahedron 1974, 30, 2273-2277. 1251 V. Percec, C. G. Cho, C. Pugh, J. Mater. Chem. 1991, I , 217-222. 1261 V. Percec, C. G. Cho, C. Pugh, Macromolecules, 1991, 24, 3227-3234. [ 1271 H. Zimmermann, R. Poupko, Z. Luz, J. Billard, Liq. Cryst. 1988,3,759-770. [I281 V. Percec, C. G. Cho, C. Pugh, D. Tomazos, Macromolecules, 1992, 25, 1164- 1176. [I291 N. Spielberg, M. Sarkar, Z. Luz, R. Poupko, J. Billard, H. Zimmermann, Liq. Cryst. 1993,15, 31 1-330. [ 1301 H. Zimmermann, R. Poupko, Z. Luz, J. Billard, Liq. Cryst. 1989, 6, 151- 166. [131] S. Bonsignore, A. Du Vosel, G. Guglielmetti, E. Dalcanale, F. Ugozzoli, Liq. Cryst. 1993,13, 471 -482. 11321 G. Cometti, E. Dalcanale, A. Du Vosel, A. M. Levelut, J. Chem. SOC.,Chem. Commun. 1990, 163- 165. [ 1331 S. Bonsignore, G. Cometti, E. Dalcanale, A. Du Vosel, Liq. Cryst. 1990, 8, 639-649. [I341 E. Dalcanale, A. Du Vosel, A. M. Levelut, J. Malthete, Liq. Cryst. 1991, 10, 185-198. [135] L. Abis, V. Arrighi, G. Cometti, E. Dalcanale, A. Du Vosel, Liq. Cryst. 1991, 9,277-284. [136] J. Zhang, J. S. Moore, J. Am. Chem. Soc., 1994, 116,2655-2656. [137] J. Zhang, J. S. Moore, Z. Xu, R. A. Aguirre, J. Am. Chem. Soc., 1992,114, 2273-2274. [I381 J. S. Moore, J. Zhang, Angew. Chem. Int. Ed. Engl. 1992,31,922-924. [139] P. Davidson, C. Jallabert, A. M. Levelut, H. Strzelecka, M. Verber, Liq. Cryst. 1988, 3, 133- 137. [ 1401 H. Strzelecka, C. Jallabert, M. Verber, P. Davidson, A. M. Levelut, Mol. Cryst. Liq. Cryst., 1988,161,395-401. [141] H. Strzelecka, C. Jallabert, M. Verber, Mol. Cryst. Liq. Cryst. 1988, 156, 355-359. 11421 M. Verber, P. Sotta, P. Davidson, A. M. Levelut, C. Jallabert, H. Strzelecka, J. Phys. France, 1990,51, 1283-1301. [I431 H. Strzelecka, C. Jallabert, M. Verber, P. Davidson, A. M. Levelut, J. Malthete, G. Sigaud, A. Skoulios, P. Weber, Mol. Cryst. Liq. Cryst. 1988,161,403-411. 11441 J. Barry, G. Bram, G. Decodts, A. Loupy, P. Pigeon, J. Sansoulet, Tetrahedron Lett., 1982,23, 5407-5408.
11451 R. Fugnitto, H. Strzelecka, A. Zann, J.-C. Dubois, J. Billard, J. Chem. Soc., Chem. Commun. 1980, 27 1-272. 11461 F. D. Saeva, G. A. Reynolds, L. Kaszczuk, J. Am. Chem. SOC., 1982,104, 3524-3525. 11471 G. A. Reynolds, C. H. Chen, J. Heterocyclic Chem., 1981,18, 1235- 1237. [148] F. D. Saeva, G. A. Reynolds, Mol. Crysi. Liq. Cryst., 1986, 132, 29-34. [149] H. Zimmermann, J. Billard, H. Gutman, E. J. Wachtel, R . Poupko, Z. Luz, Liq. Cryst. 1992. 12,245-262. [150] C. Destrade, N. H. Tinh, H. Gasparoux, L. Mamlok, Liq. Cryst., 2, 229-233. 11511 L. Mamlock, J. Malthete, N. H. Tinh, C. Destrade, A. M. Levelut, J. Phys. Lett., 1982, 43. L641 -L647. [152] C. Destrade, N. H. Tinh, L. Mamlock, J. Malthete, Mol. Cryst. Liq. Cryst., 1984, 114, 139150. 11531 R. Cayuela, H. T. Nguyen, C. Destrade, A. M. Levelut, Mol. Cryst. Liq. Cryst. 1989, 177, 81-91. [154] N. H. Tinh, R. Cayuela, C. Destrade, Mol. Cryst. Liq. Cryst., 1985, 122, 141-149. [155] N. Boden, R. C. Borner, R. J. Bushby, J. Clements, J. Am. Chem. SOC., 1994, 116, 1080710808. [I561 G. G. Hall, W. R. Rodwell, J. Theor. Biol.. 1975,50, 107-120. 11571 R. W. Baldwin, G. J. Cunningham, M. W. Partridge, H. J. Vipond, Br. J. Cancer. 1962, 16, 275 -282. [I581 S. Kumar, E. J. Wachtel, E. Keinan, J. O r , . Chem., 1993,58,3821-3827. [I591 E. Kienan, S. Kumar, S. P. Singh, R. Ghirlando, E. J. Wachtel, Liq. Cryst. 1992, 11, 157173. [160] B. Franck, Angew. Chem. Int. Ed. EngI., 1982, 21,343-353. [I611 B. A. Gregg, M. A. Fox, A. J. Bard, J. Chem. Soc. Chem. Commun. 1987, 1134- 1135. [162] B. A. Gregg, M. A. Fox, A. J. Bard, J. Am. Chem. SOC.,1989,111,3024-3029. [163] J. C. Fanning, F. S . Mandel, T. L. Gray, N. Datta-Gupta, Tetrahedron, 1979, 35, 1251 - 1255. [164] K. M. Smith, G. H. Barnett, B. Evans, Z. Martynenko, J . Am. Chem. Soc., 1979,101,59535961. [165] Y. Shimitzu, M. Miya, A. Nagata, K. Ohta, A. Matsumura, I. Yamamoto, S. Kusabayashi. Chem. Lett., 1991, 25-28. [166] A. D. Adler, F. R. Longo, J. D. Finarelli, J. Goldmacher, J. Assour, L. Korsakoff. J. Or,. Chem., 1967,32,476. [I671 G. Bahr, G. Schleitzer, Chem. B e t , 1957, YO, 438 -443. I1681 G. Morelli, G. Ricciardi, A Rovjello, Chem. Phys. Lett., 1991, 185, 468-472.
5 [169] F. Lelj, G. Morelli, G. Ricciardi, A. Roviello, A. Sirigu, Liq. Cryst. 1992, 12, 941 -960. [170] P. Doppelt, S. Huille, New J. Chem., 1990, 14, 607-609. [171] C. Piechocki, J. Simon, A. Skoulios, D. Guillon, P. Weber, J. Am. Chem. SOC., 1982, 104, 5245 - 5 247. [ 1721 G. Pawlowski, M. Hanack, Synthesis, 1980, 287-289. [173] E. Klingsberg, Synthesis, 1972, 29-30. [ 1741 W. Wenner, J. Org. Chem., 1952,17,523-528. [ 1751 W. Offermann, F. Vogtle,Angew. Chem. Int. Ed. Engl., 1980,19,464-465. 11761 M. Hanack, A. Beck, H. Lehmann, Synthesis, 1987,703-705. [I771 P. Weber, D. Guillon, A. Skoulios, J. Phys. Chem., 1987,91,2242-2243. [178] C. Piechocki, J.-C. Boulou, J. Simon, Mol. Cryst. Liq. Cryst., 1987, 149, 115- 120. [179] D. Guillon, A. Skoulios, C. Piechocki, J. Simon, P. Weber, Mol. Cryst. Liq. Cryst., 1983, 100,275-284. [180] D. Guillon, P. Weber, A. Skoulios, C. Piechocki, J. Simon, Mol. Cryst. Liq. Cryst. 1985, 130,223-229. [181] C. Sirlin, L. Bosio, J. Simon, J. Chem. SOC. Ckem. Commun., 1987,379-380. I1821 C. Piechocki, J. Simon, Nouv. J. Chim., 1985, 9, 159-166. [183] I. Cho, Y.Lim, Chem. Lett., 1987,2107-2108. I1841 I. Cho, Y. Lim, Mol. Cryst. Liq. Cryst., 1988, 154,9-26. (1851 K. Ohta, T. Watanabe, S. Tanaka, T. Fujimoto, I. Yamamoto, P. Bassoul, N. Kucharczyk, J. Simon, Liq. Cryst. 1991,10, 357-368. [186} J. F. Van der Pol, E. Neeleman, J. W. Zwikker, R. J. M. Nolte, W. Drenth, Rec. Truv. Ckim. Pu~s-Bus,1988,107,615-620. [I871 L. M. Severs, A. E. Underhill, D. Edwards, P. Wight, D. Thetford, Mol. Crysr. Liq. Cryst., 1993,234,235- 240. [ 1881 J. F. Van der Pol, E. Neeleman, J. W. Zwikker, R. J. M. Nolte, W. Drenth, J. Aerts, R. Visser, S. J. Picken, Liq. Cryst. 1989, 6,577-592. 11891 T. Sauer, G. Wegner, Mol. Cryst. Liq. Cryst., 1988,162B, 97- 118. [ 1901 T. Komatsu, K. Ohta, T. Watanabe, H. Ikemoto, T. Fujimoto, I. Yamamoto, J. Muter. Ckem. 1994,4,537-540. [I911 P. G. Schouten, J. M. Warman, M. P. de Haas, C. F. van Nostrum, G. H. Gelinck, R. J. M. Nolte, M. C. Copyn, J. W. Zwikker, M. E. Engel, M. Hanack, Y. H. Chang, W. T. Ford, J.Am. Chem. Soc., 1994,116, 6880-6894. [I921 D. Lelievre, M. A. Petit, J. Simon, Liq. Cryst., 1989,4,707-710. 11931 P. G. Schouten, J. F. Van der Pol, J. W. Zwikker, W. Drenth, S. J. Picken, Mol. Cryst. Liq. Cryst. 1991,195,291-305.
References
747
[194] W. T. Ford, L. Sumner, W. Zhu, Y. H. Chang, P.-J. Um, K. H. Choi, P. A. Heiney, N. C. Maliszewskyj, New J. Chem., 1994,18,495-505. [195] K. Ohta, T. Watanabe, H. Hasebe, Y. Morizumi, T. Fujimoto, I. Yamamoto, D. Lelievre, J. Simon, Mol. Cryst. Liq. Cryst. 1991, 196, 13. [196] D. Masurel, C. Sirlin, J. Simon, New J. Ckem. 1987,1I, 455-456. [197] K. Ohta, L. Jacquemin, C. Sirlin, L. Bosio, J. Simon, New J. Chem. 1988,12,751-754. [198] E. A. Cuellar, T. J . Marks, lnorg. Ckem., 1981, 20,3766-3770. [199] K. Tamao, K. Sumitani, Y.Kiso, M. Zembayashi, A. Fujioka, S.-I. Kodama, 1. Nakajima, A. Minato, M. Kumada, Bull. Chem. SOC. Jpn. 1976, 49, 1958- 1969. [200] M. K. Engel, P.Bassoul, L. Bosio, H. Lehmann, M. Hanack, J. Simon, Liq. Cryst. 1993, 15, 709-722. [201] K. Ohta, T. Watanabe, T. Fujimoto, I. Yamamoto, J. Ckem. SOC., Chem. Commun., 1980, 1611-1613. [202] S. M. McElvain, Org. React., 1948, 4, 256268. [203] D. E. Ames, G. Hall, B. T. Warren, J. Ckem. SOC.(C), 1968,2617-2621. [204] K. Kanakarajan, A. W. Czarnik, J. Org. Chem., 1986,51,5241. [205] L. Dulog, A. Gittinger, MoZ. Cryst. Liq. Crysr., 1992,213,31-42. [206] D. Wohrle, U. Hundorf, Mukromol. Ckem., 1985,186,2177-2187. [207] A. J. Saggiomo, J. Org. Chem., 1957, 22, 1171 - 1175. [208] P. A. Barrett, R. P. Linstead, G. A. P. Tuey, J. M. Robertson, J. Ckem. SOC. 1939, 18091820. [209] C. C. Leznoff, N. B. McKeown, J. Org. Ckem. 1990,55,2186-2190. [210] N.B. McKeown, C. C. Leznoff, R. M. Richardson, A. S . Cherodian, Mol. Cryst. Liq. Cryst., 1992, 213, 91-98. [211] N. B. McKeown, J. Painter, J. Muter: Chem., 1994,4, 1153-1156. [212] T. W. Hall, S . Greenberg, C. R. McArthur, B. Khouw, C. C. Leznoff, New J. Ckem., 1982,6, 653. [213] M. J. Cook, M. F. Daniel, K. J. Harrison, N. B. McKeown, A. J. Thomson, J. Ckem. SOC., Chem. Commun., 1987, 1086- 1088. [214] A. N. Cammidge, M. J. Cook, K. J. Harrison, N. B. McKeown, J. Ckem. SOC., Perkin Trans. I , 1991,3053-3058. [215] A. N. Cammidge,M. J. Cook, S. D. Haslam,R. M. Richardson, K. J. Harrison, Liq. Cryst. 1993,14, 1847-1862. [216] N. B. McKeown, I. Chambrier, M. J. Cook, J. Chem. SOC., Perkin Trans. I , 1990, 11691177.
748
VII
Synthesis and Structural Features
[217] M. J. Cook, J. Muter. Sci., 1994, 5, 117-128. [218] A. S. Cherodian, A. N. Davies, R. M. Richardson, M. J. Cook, N. B. McKeown, A. J. Thomson, J. Feijoo, G. Ungar, K. J. Harrison, Mol. Cryst. Liq. Cryst., 1991, 196, 103-114. [219] M. J. Cook, S. J. Cracknell, K. J. Harrison, J . Mater. Chem. 1991,1,703-704. [220] C. Piechocki, J. Simon, J. Chem. Soc., Chem. Commun. 1985,259-260. [221] I. Chambrier, M. J. Cook, S. J. Cracknell, J. McMurdo, J. Mater. Chem., 1993,3,841-849. [222] B. Kohne, K. Praefcke, Angew. Chem. Int. Ed. Engl., 1984, 23, 82-83. [223] K. Praefcke, B. Kohne, W. Stephan, P. Marquardt, Chimia, 1989,43, 380-382. [224] B. Kohne, K. Praefcke, W. Stephan, P. Nurnberg,Z. Naturforsch., Teilb, 1985,40,981-986. 12251 K. Praefcke, P. Psaras, B. Kohne, Chem. Ber., 1991,124,2523-2529. [226] B. Kohne, K. Praefcke, Z. Naturforsch., Teil b, 1986,41, 1036- 1044. 12271 C. P. Lillya, D. M. Collard, Mol. Cryst. Liq. Cryst., 1990, 182B, 201-207. [228] N. L. Morris, R. G. Zimmerman, G. B. Jameson, A. W. Dalziel, P. M. Reuss, R. G. Weiss, J.Am. Chem. SOC., 1988,110,2177-2185. (2291 V. Vill, J. Thiem, Liq. Cryst. 1991,9,451-455. [230] R. G. Zimmerman, G. B. Jameson, R. G. Weiss, G. Dermailly, Mol. Cryst. Liq. Cryst. Lett., 1985, 1, 183-189. [231] A. Takada, T. Fukuda, T. Miyamoto, Y. Yakoh, J. Watanabe, Liq. Cryst. 1992, 12, 337-345. [232] T. Itoh, A. Takada, T. Fukuda, T. Miyamoto, Y. Yakoh, J. Watanabe, Liq. Cryst. 1991, 9, 221228.
[233] M. Sagiura, M. Minoda, J. Watanabe, T. Fukuda, T. Miyamoto, Bull. Chem. SUC.Jpn., 1992, 65, 1939- 1943. [234] A. Takada, K. Fujii, J. Watanabe, T. Fukuda, T. Miyamoto, Macromolecules, 1994, 27, 1651 1653. [235] B. Kohne, K. Praefcke, Chem. Z., 1985, 109, 121 - 127. [236] J.-M. Lehn, J. Malthete, A. M. Levelut, J. Chem. Soc., Chem. Cummun., 1985, 17941796. [237] S. H. J. Idziak, N. C. Maliszewskyj, G. B. M. Vaughan, P. A. Heiney, C. Mertesdorf, H. Ringsdorf, J. P. McCauley, A. B. Smith, J. Chem. Soc., Chem. Commun., 1992,98-99. [238] J. Malthete, A.M. Levelut, J.-M. Lehn. J. Chem. Soc., Chem. Commun. 1992, 14341436. 12391 M. Zhao. W. T. Ford. S. H. J. Idziak, N. C. Maliszewskyj, P. A. Heiney, Liq.Cryst., 1994, 16, 583-599. [240] S. H. J. Idziak, N. C. Maliszewskyj, P. A. Heiney, J. P. McCauley, P. A. Sprengeler, A. B. Smith, J. Am. Chem. Soc., 1991, 113, 76667672. [241] J. Malthete, D. Poupinet, R. Vilanove, J.-M. Lehn, J. Ckem. Soc., Chem. Commun. 1989, 1016-1019. [242] C. Mertesdorf, H. Ringsdorf, Liq.Cryst,, 1989, 5, 1757-1772. [243] C. Mertesdorf, H. Ringsdorf, J. Stumpe, Liq. Cryst. 1991, 9, 337-357. [244] G. Lattermann, Mol. Cryst. Liq. Cryst. 1990, 182B, 299-311. [245] G. Lattermann, Liq. Cryst., 1989,6, 619-625.
Handbook of Liquid Crystals Edited by D. Demus, J. Goodby, G. W. Gray, H.-W. Spiess, V. Vill Copyright 0 WILEY-VCH Verlag GmbH. 1998
Chapter VIII Columnar, Discotic Nematic and Lamellar Liquid Crystals: Their Structures and Physical Properties S. Chandrasekhar
1 Introduction We begin this chapter with a few remarks about terminology. Soon after the discovery of liquid crystals of disk-like molecules [l-31, the word 'discotic' was introduced to distinguish these mesogens from the familiar rod-like or calamitic-type mesogens. However, before long, terms such as 'discogen', 'discotic phase', etc., began to be used extensively in the literature, and the word 'discotic' and related terms came to be applied loosely to describe the molecules as well as the mesophases formed by them. Strictly speaking, it is the molecules that are discotic and not the mesophases, which may be columnar, nematic or lamellar. Moreover, as we shall see later, certain molecules, which themselves are not discotic, have been found to exhibit columnar phases [4-61. Thus, care is necessary in the use of these terms in order to avoid any ambiguity. The first examples of thermotropic mesomorphism in discotic compounds were observed in the hexa-alkanoyloxy benzenes (1, R=CnH2,+,) [l] and the hexa-alkoxyand hexa-alkanoyloxytriphenylenes (2, R=C,H,,+,O and C,H2,+,C0 0) [2, 31, and it was established by X-ray studies [ 1, 71 that the mesophases of these compounds have a columnar structure. About 1500 discotic mesogens are known to date [8] and a variety of new mesophase structures has been identified. The simplest of these compounds have flat or nearly flat cores surrounded by six or eight, or sometimes four,
long-chain substituents. Much more complex discotic molecules have since been synthesized; for example, the star-like triphenylene derivatives (3) reported recently [9]. Available experimental evidence indicates that the presence of long side chains is crucial to the formation of discotic liquid crystals. The synthetic procedure and other chemical aspects of the subject are dealt with in Chapters VII and IX in this volume, and are not therefore considered here. The discussion here is limited to the structural classification of the mesophases and a description of their physical properties, and just a few illustrative examples of discotic molecules that are relevant to the topics discussed, are presented.
I
R 1
+
A
"UR
OR R
RO
2
750
VIII
Columnar, Discotic Nematic and Lamellar Liquid Crystals
n = 6,9,11
3
2 Description of the Liquid Crystalline Structures Discotic liquid crystals may be classified broadly into three types: columnal; nematic (including its chiral modification) and lamellar.
2.1 The Columnar Liquid Crystal The basic columnar structure is illustrated in Figure 1 (a): the disks are stacked one on
top of another to form columns, the different columns constituting a two-dimensional lattice. The structure is somewhat similar to the hexagonal phase of soap-water and lyotropic systems [lo], but several variants of the structure have been identified - upright columns (Figure 1 a), tilted columns (Figure 1b), hexagonal, rectangular, etc. In some cases, the columns are liquid-like, i.e. the molecular centres are arranged aperiodically within each column, as depicted in Figure 1 (a and b), while in others they are arranged in a regular, ordered fashion. The columnar phase is denoted by the symbol D (according to new guidelines, see Chap. I1 of Vol. I of this Handbook, it should
2
(a)
Description of the Liquid - Crystalline Structures
75 1
(b)
Figure 1. Schematic representation of columnar structures of discotic mesogens: (a) upright columns, (b) tilted columns.
be denoted as “Col” instead of “D” in future) with appropriate subscripts to distinguish the different modifications. For example, the subscript ‘h’ stands for a ‘hexagonal’ lattice of columns, ‘r’ for a ‘rectangular’ lattice, ‘d’ for a ‘disordered’ stacking of the molecules in each column, ‘0’ for an ‘ordered’ stacking and ‘t’ for tilted columns. Thus: D,, signifies columnar, hexagonal, disordered - D,, signifies columnar, rectangular, disordered - Dho signifies columnar, hexagonal, ordered - D, signifies columnar, tilted -
A tilt of the molecular core with respect to the column axis is a common feature in these structures. In the very first report on the hexa-alkanoyloxy benzenes [I] it was noted that the X-ray photographs contained a few weak diffraction spots not quite conforming to true hexagonal symmetry, but the quality of the X-ray patterns in these early studies did not warrant a more refined analysis. Subsequently, Frank and Chandrasekhar [ 1I ] noted that, when seen under the polarizing microscope, the extinction brushes (or crosses) in the optical textures of themesophases do not coincide with the polarizer - analyser directions, but are usually tilted
Figure 2. Circular domains with curved column axes showing oblique extinction brushes in pure benzene hexa-n-hexanoate (a b) and in its mixture with a few per cent by weight of benzene (c). The mesophase of the mixture is slightly more mobile than that of the pure material, but has the same structure. Polarizer - analyser directions are vertical and horizontal ((a) Courtesy of K. A. Suresh; (b and c), from Frank and Chandrasekhar [l I], reproduced by permission of the Commission des Publications Frayaises de Physique).
with respect to these directions (Figure 2). From this and other optical observations, they concluded that the plane of the rigid aromatic core (in which resides the anisotropy of molecular polarizability, and the diameter of which is only about a third of the whole molecule) is not normal to the column axis, but inclined to it at an angle of about 55-60’, indicating that the lattice should be pseudo-hexagonal. This was confirmed by Levelut [12] who found that the
752
VlII
Columnar, Discotic Nematic and Lamellar Liquid Crystals
columnar phase of the hexa-n-octanoate of benzene (1) has a rectangular lattice with an axial ratio of 1.677, which departs from the true hexagonal value of 6 by about 3%. Figure 3 shows a schematic representation of the structure of a column in which the cores are tilted with respect to the column axis and the chains are in a highly disordered state. Figure 4 presents the different types of two-dimensional lattice structures that have been identified by Levelut [12], and Table 1 gives the space groups of the columnar structures formed by some derivatives of triphenylene. (These are pEanar space groups that constitute the subset of the 230 space groups when symmetry elements relating to translations along one of the axes, in this case the column axis, are absent.)
(el
(d)
Figure 4. Plan views of two-dimensional lattices of columnar phases; ellipses denote disks that are tilted with respect to the column axis: (a) hexagonal (P6 2Im 2lm);(b) rectangular (P 2,Ia); (c) oblique ( P , ) ;(d) rectangular (P,la); (e) rectangular, face-centered, tilted columns (C2lm). (From Levelut [12],reproduced by permission of the division des Publications Soci6t6 de Chimie Physique).
2.1.1 NMR Studies
Figure3. The structure of a column in which the molecular cores are tilted with respect to the column axis and the chains are in a highly disordered state.
Deuterium NMR spectroscopy of the columnar phase of hexa-n-hexyloxytriphenylene has revealed the marked difference between the order parameters of the cores and the tails [13]. Spectra of two selectively deuteriated isotopic species, one in which all aromatic positions are substituted and the other in which only the a-carbon side chains
Table 1. Two-dimensional lattices of columnar structures formed by some hexa-substituted triphenylenes [ 121. R
Space group
Lattice parameters (nm)
Temperature (“C)
C,H,lO
P6 21m 2Im P6 2fm 2lm P6 2Im 2Im P2,fa P6 2lm 2lm P2,la P2 / a C2lm
a=1.895 a = 2.22 a=2.33 a=4.49, b=2.64 a=2.63 a=3.78, b=2.22 a=5.18, b=3.26 a = 3.07, b = 2.84
= 80 = 80 = 80
C7H150 C8H170
I IH23C00
C7HI ,COO cl O-C6H4-COOa C,&, 3 - O-C&- coo” a
,
The C6H4 groups are para substituted.
117 105 100 165 185
2
Description of the Liquid - Crystalline Structures
are substituted, were investigated. Figure 5 gives the quadrupole splittings of the aromatic and the a-aliphatic deuterons versus temperature in the mesophase region. It is seen that the rigid core is highly ordered, the orientational order parameter S ranging from 0.95 to 0.90, where
s=
::.
These results indicate that in any realistic theory of the statistical mechanics of columnar systems the molecules cannot be treated as rigid disks, but the conformationa1 degrees of freedom of the hydrocarbon chains must also be taken into account.
2.1.2 High Resolution X-Ray Studies
-(3c0s2e-i)
and 8 is the angle which the molecular-symmetry axis makes with the director or column axis, whereas the a-aliphatic chains are in a disordered state. Relaxation studies have also been reported [14].
- 0.8
-
High resolution X-ray studies have yielded important additional information about the columnar structures. The measurements were made on very well oriented monodomain samples obtained by preparing freely
(C,,H,,CO . 0)6-truxene (4, R=C,,H&O * 0)
50 -
40
753
-
$9
R
0'50i"'-----0.45 0.40
I
I
'
'
'
'
R
'
vc
Figure 5. Quadrupole splittings for aromatic and a-aliphatic vg deuterons of deuteriated hexa-n-hexyloxytriphenylene (THE6) as functions of temperature (T,-T) in the mesophase region, where T, is the columnar- isotropic transition point. (0) Neat THE6-ard6 amd THE6-a-dI, separately; ( 0 ) a 2 : 1 mixture of the two isotopic species. The scale on the upper righthand side gives the orientational order parameter of the aromatic part. The curve at the bottom gives the ratio of the quadrupole splittings for the a-aliphatic and aromatic deuterons (From Goldfarb et al. [ 131,reproduced by permission of the Commission des Publications Franpise de Physique).
K 4
The correlation length of the two-dimensional lattice of the hexagonal columnar phase of this compound is greater than 400 nm (or approximately 200 columns), the lower limit being set by the resolution of the instrument [ 191.Within each column, on the other hand, the flat molecular cores from an oriented, one-dimensional liquid,
754
VIII
Columnar. Discotic Nematic and Lamellar Liquid Crystals
i.e. they are orientationally ordered but translationally disordered. In contrast, the flexible hydrocarbon chains are highly disordered and produce a nearly isotropic scattering pattern. This striking difference between the ordering of the cores and the tails is also seen in the X-ray scattering from the columnar mesophases of triphenylene compounds, in conformity with the results of deuterium NMR spectroscopy, but it is not so apparent in the hexa-n-octanoate of benzene, which has a much smaller core [ 16, 171. The structure of the columnar phase of the truxene compound is essentially the same as that proposed originally by Chandrasekhar et al. [ 11. The possibility of such a structure i.e. a three-dimensional body in which the density is a periodic function of only two dimensions, was in fact envisaged many years ago by Landau and Lifshitz [20] (see Sec. 10.3). (C,,H,,CO . O),+riphenylene (2, R=C,,H&O. 0)
This compound exhibits a transition from a hexagonal (Dh)to a rectangular (D,) columnar phase. The transition, which is weakly first order, is associated with a small distortion of the lattice, consistent with a herringbone arrangement in the rectangular structure, with only the core of the molecule tilted with respect to the column axis (see Figure 3 ) . However, high resolution synchrotron X-ray studies on monodomain strands [16, 171 have established that the tilt of the molecular core persists in the D, phase as well, except that the tilts in neighbouring columns are rotationally uncorrelated, i.e. they are free to assume different azimuthal angles. Thus the D,-D, transition may be looked upon as an orientational order- disorder transition. Theoretical models have been developed to describe this transition [21-231.
It may be mentioned that deuterium NMR and dielectric spectroscopy studies [24,25] on columnar liquid crystals indicate the existence of rotational motions of the molecules or groups of molecules about the column axes. Hexuhexylthiotriphenylene (2, R = SC,H,
3)
This compound (HHTT) shows a transition from a hexagonal ordered phase (Dho)to a hexagonal disordered one (Dhd). As mentioned earlier, the former is a phase in which there is regularity in the stacking of the triphenylene cores in each column, and the latter one in which the column is liquidlike. X-ray studies using freely suspended strands reveal that in the ordered phase there is a helicoidal stacking of the cores within each column, the helical spacing being incommensurate with the intermolecular spacing [ 181. In addition, a three-column superlattice develops as a result of the frustration caused by molecular interdigitation in a triangular symmetry (Figure 6). This helical phase of HHTT is also referred to as the H phase. Ideally, if there is no intercolumn interaction, true long-range order cannot exist within a column because of the Peierls - Landau instability. The existence of a regular periodicity in the stacking in each column therefore implies that neighbouring columns must be in register. Thus ordered columnar phases can probably be compared with the highly ordered smecticlike phases of rod-like molecules, e.g. the crystal B, E, G, H, etc. phases, which possess three-dimensional positional order. However, further high resolution studies are necessary before general conclusions can be drawn. An unusual feature observed in the D,, phase of this compound is worth mentioning. The hexagonal lattice exhibits a negative coefficient of thermal expansion (Figure 7). This has been attributed to the
2
Description of the Liquid - Crystalline Structures
stiffening of the hydrocarbon tails, which therefore become longer and may provide the driving mechanism for the Dhd-Dho transition by enhanced intercolumn coupling.
I--4
2.2 Columnar Phases of “on-discotic’ Molecules Columnar phases are also formed when the flat core is replaced by a conical or pyramidal shaped one (5 and 6) [26,27], and, rarely, even when the central core is absent altogether, as in certain macrocyclic molecules (7)[28]. In the latter case, the columns are in the form of tubes (Figure 8), and the mesophase has been called ‘tubular’.
R = Cg HlgCOO-
R
=
Cll H2$O0 -
R =C , 2 H 2 5 0 e C 0 0 -
5
Figure 6. Structure of the D,, phase of HHTT the diagram represents a three-column superlattice (not to scale), with column 1 displaced b y p / 2 relative to columns 2 and 3 , neighbouring columns having correlated helical phases. (From Fontes et al. [ 181, reproduced by permission of the America1 Physical Society).
R
R
R
21.8
21.6
-
“5
21.4 21.2 21.0 20.8
I
R = C,,H2,,+1-0 R z C,H2,,+,-C0.O
I
-
LWdm&---v
HEATNG
0 COOLlNC
1
-. *
....
60
65
70
75
80
85
90
TPC)
Figure 7.The intercolumn separation d i n the columnar phases of HHTT plotted versus temperature, showing the negative coefficient of thermal expansion. ( 0 ) Heating mode; (0) cooling mode. (From Fontes et al. [18], reproduced by permission of the American Physical Society).
755
6
756
VIII
Columnar. Discotic Nematic and Lamellar Liquid Crystals
Figure 9. Structure of the hexagonal columnar mesophase of 8. The mesophase has been described as ‘phasmidic’. (From Malthete et al. [ 6 ] ,reproduced by permission of Commission des Publication Franpises de Physique).
Figure 8. (a) Stacking of macrocyclic molecules: for clarity of presentation the molecules are drawn with the same orientation about the tube axis. (b) The hexagonal ‘tubular’ mesophase so formed. (From Lehn et al. [ 2 8 ] ,reproduced by permission of the Royal Society of Chemistry).
Malthete and coworkers [6,29] have prepared a series of mesogens shaped like stick insects called ‘phasmids’ (8). Some of them form columnar mesophases; the structure proposed for the hexagonal phase is shown in Figure 9.
Historically, an interesting case is that of diisobutylsilane diol (9). In the 1950s, CH3 H3C
H
I
I
-C -C
I
I\
H
H3C -C-C
I
H
H
t
H
9
/OH
Eaborn and Hartshorne [30] observed that the compound showed a mesophase which they were unable to classify as belonging to any of the then known liquid crystal types. Twenty-five years later, Bunning et al. [4] carried out miscibility studies of 9 with the hexa-n-heptanoate of benzene (Figure 10) and proved conclusively that the mesophase is a columnar liquid crystal. They have suggested that the molecule forms a dimer [4] or a polymer [ 5 ] that favours the occurrence of a columnar phase.
>
100 */. DllBSD
>
, , , . , . Cornposit ion
100 % BHH
Figure 10. The miscibility diagram of diisobutylsilane diol (DIIBSD) with benzene hexa-n-heptanoate (BHH). (From Bunning et al. [4], reproduced by permission of Springer-Verlag).
2 Description of the Liquid - Crystalline Structures
2.3 The Nematic Phase The nematic phase formed by disk-like molecules, designated by the symbol N,, is a fluid phase showing schlieren textures similar to those of the classical nematic phase of rod-like molecules (Figure 11). Its structure is represented in Figure 12: it is an orientationally ordered arrangement of disks with no long-range translational order. However, unlike the usual nematic phase, N, is optically negative, the director n now representing the preferred axis of orientation of the molecular short axis or the disk normal. The ND phase is exhibited by relatively few compounds: examples are hexa-n-alkyl- and hexa-n-alkoxybenzoates of triphenylene (2, R=C,H2,+,-C6H4-C0 * 0, and C,H,,+,O-C,H,-CO~ 0) [31, 321 and hexakis(4-octylphenylethyny1)benzene (10) [ W . R
/
9Q R
R
10
i
757
Figure 12. Schematic representation of the nematic phase (N,) of disk-like molecules.
In some compounds (usually the lower members of a homologous series), the transition to the ND phase takes place directly from the crystal, while in others (the higher members of the series) it takes place via a columnar phase (Table 2) [31, 321. The trend is somewhat analogous to the behaviour of the smectic A-nematic transition for a homologous series, and as will be discussed in Sec. 3 this trend can be explained by an extension of the McMillan model of the A-N transition to systems of discotic molecules. There is also X-ray evidence of skew cybotactic groups in the ND phase prior to the transition to the D, phase, very much like the situation seen near a nematic-smectic C transition [34]. The hexa-n-alkanoates of truxene show an unusual sequence of transitions [35,36]. For the higher homologues the phase sequence on cooling is I - D, - D, - N, - Reentrant D, - Cr, where I and Cr are the isotropic and crystalline phases, respectively. It has been suggested that the truxene molecules are probably associated in pairs, and that these pairs break up at higher temperatures, which might explain this extraordinary behaviour.
2.4 The Columnar Nematic Phase
Figure 11. Schlieren texture of the nematic phase of hexa-n-hexyloxybenzoate of triphenylene. (From Destrade et al. [ 3 2 ] , reproduced by permission of Heyden and Son).
The addition of electron acceptor molecules such as 2,4,7-trinitrofluorenone (TNF) to electron donating discotic molecules results in the formation of charge transfer complexes, which stabilize (or, in non-mesomorphic materials, induce) mesophases [37 -391. A
758
VIII
Columnar, Discotic Nematic and Lamellar Liquid Crystals
Table 2. Hexa-n-alkoxybenzoates of triphenylene: temperatures and heats of transition” [3 1, 321. Phase
4 5 6
145 2.15
183 0.31
7 8 9 10
11 12
257 224 4.55 I86 2.24 168 2.25 152 14.58 154 3.24 142 8.30 145 6.0 146 1.16
193 1.12
>300
0
298
0
274
0
253
168
244
183
227
181
212
179 1.21 114
185 0.054
~
~~
0
~~
~
The phases exhibited by a compound are indicated by points in the appropriate columns. The transition temperatures are given in degrees Celsius, and for those cases for which data are available the corresponding heats of transition are given immediately below in kilocalories per mole.
new type of induced mesophase identified in such systems is the columnar nematic (N,) phase, the structure of which consists of columns (Figure 13). Transitions from
N, to D,, have been observed in mixtures of TNF and a pentakis(phenylethyny1)phenylalkyl ether [37, 381.
2.5 The Chiral Nematic Phase
NC
Figure 13. Schematic diagram of the structure of the columnar nematic phase induced by charge transfer interactions between an electron donor, pentakis(pheny1ethynyl)phenylalkyl ether (in black), and an electron acceptor, 2,4,7-trinitrofluorenone (hatched) [37 391.
Optically active esters of benzene, triphenylene and truxene have been studied [40]. The triphenylene and truxene derivatives (structures 2 and 4, respectively, with R=(+)- or (-)-CH,-(CH,),-CH(CH,)CH,-CO . 0) do show mesophases, but of the non-chiral columnar type. In other words, this chiral substituent does not change the nature of the phases. The benzene derivative is non-mesomorphic. However, when these compounds are mixed with
2
Description of the Liquid - Crystalline Structures
the hexa-n-heptyloxybenzoateof triphenylene, which in the pure state shows the N, phase, typical cholesteric textures are obtained. This was the first example of a twisted nematic or cholesteric phase of discotic mesogens. It is designated by the symbol NG and its structure is illustrated schematically in Figure 14. However, unlike the classical nematic, a relatively high concentration (- 50%) of the chiral additive is required to produce an appreciable twist, and, furthermore, the cholesteric texture can be seen only at higher temperatures after the initial appearance of a nematic texture, implying a very strong temperature dependence of the pitch. A pure discotic cholesterogen, a triphenylene derivative (2) with R = (S)-(+)-CH, CH2 - CH(CH,)(CH,),O - C6H4 - CO * 0, has been reported [41]. The cholesteric phase occurs between 192.5 and 246.5 "C, with a pitch of about 30 pm at 200°C. On the other hand, the optically active derivative of triphenylene with R = (S)-(+)-CH, (CH2),-CH(CH,)(CH2)2O-C,H4-CO. 0 exhibits only a non-chiral D,, phase, whereas that with R = (S)-(+)-CH, - (CH,), CH(CH,)CH20-C,H4-C0. 0 is nonmesomorphic. Thus the occurrence of a cholesteric phase depends rather critically on the structure of the chiral chain.
759
2.6 The Lamellar Phase Giroud-Godquin and Billard [42], who prepared and investigated the first discotic metallomesogen, bis( p-n-decylbenzoy1)methanato copper(I1) (11, R=C,,H,,), suggested that its mesophase has a lamellar structure. Similar conclusions have been drawn by Ohta and coworkers [43, 441 from X-ray studies and by Ribeiro et al. [45] from NMR investigations. Sakashita et al. [44] have proposed a tilted smectic C type structure, as depicted in Figure 15. R
R
A variety of discotic metal complexes has been synthesized which exhibit columnar or lamellar phases [46]. X-ray determina-
5
Figure 14. Schematic representation of the twisted nematic phase (N:) of disk-like molecules.
Figure 15. Schematic representation of the structure of the lamellar phase of the copper complex 11, with R = C 12H25,proposed by Sakashita et al. [44]. (Reproduced by permission of Gordon and Breach).
760
VIII
Columnar, Discotic Nematic and Lamellar Liquid Crystals
Figure 16. (a) Crystal structure of bis[(1-p-n-heptylphenyl, 3-p-n-heptyloxypheny1)propane-1,3-dianato]copper(I1): arrangement of the molecules perpendicular to the crystallographic a axis (From Usha and Vijayan 15 I], reproduced by permission of Taylor and Francis Ltd.). (b) Crystal structure of bis[(1-p-n-heptylphenyl, 3-p-nheptyloxypheny1)propane-1,3-dianato]copper(II): arrangement of the molecules perpendicular to the crystallographic b axis. (From Usha et al. [52], reproduced by permission of Taylor and Francis Ltd.).
tions of the crystal structures have been useful in elucidating the structures of the complexes and the mesophases formed by them [47 - 531. Some of them have a layered arrangement in the crystalline phase (Figure 16), but it would be fair to say that the disposition of the molecules in the lamellar mesophase is not yet completely understood.
3 Extension of McMillan’s Model of the Smectic A Phase to Columnar Liquid Crystals As mentioned in Section 7.2.3, the observed trend of the columnar - nematic - isotropic sequence of transitions in some homologous
3 Extension of McMillan’s Model of the Smectic A Phase to Columnar Liquid Crystals
series of discotic compounds is rather reminiscent of the behaviour of the smectic A - nematic - isotropic sequence in calamitic systems. This can be understood qualitatively by a simple extension of McMillan’s mean field model of the smectic A phase [54], which takes account of the fact that the density wave is now periodic in two dimensions [55-581. The hexagonal disordered columnar phase (Dhd)can be described by a superposition of three density waves with the wavevectors
C=A+B where d is the lattice constant. The appropriate single particle potential, which depends on both the orientation of the short axis of the molecule and the position r of its centre of mass, may be written in the mean field approximation as
v , ( ~case) , ~ ,= - V ~ P ~ ( C O S ~ ~ . (77 + C X ~ [ C O S*( rA) + cos ( B . r ) + cos (C . r)] } ( 1 ) retaining only the leading terms in the Fourier expansion. Here V0 is the interaction energy which determines the nematic -isotropic transition, a is the McMillan parameter given by
r, being the range of interaction which is of the order of the size of the aromatic core,
76 1
is the orientational order parameter, P2 being the Legendre polynomial of order 2, and
is an order parameter coupling the orientational and the translational ordering. The angle brackets represent the statistical average over the distribution function derived from the potential given by Eq. (l), the spatial integrations being carried out over a primitive cell of the hexagonal lattice. This form of the potential ensures that the energy of the molecule is a minimum when the disk is centered in the column with its plane normal to the z axis. The free energy can then be calculated using standard arguments: ~~
AF - vo NkBT-2kBT
(q2+ao2) 1
dx dy d(cos8) 0
Numerical calculations yield three possible solutions to the equation:
(1) 77 = o = 0 (isotropic phase) ( 2 ) q f 0, o = 0 (nematic phase) (3) 77 z 0, o f 0 (hexagonal columnar phase) For any given values of a and T,the stable phase is the one that minimizes the free energy given by Eq. ( 5 ) . Interpreting a as a measure of the chain length, as in McMillan’s model, the phase diagram as a function of a (Figure 17) reflects the observed trends for a homologous series of compounds. The temperature range of the nematic phase decreases with increasing a, and for a > 0.64 the columnar phase transforms directly to
762
VIII
Columnar, Discotic Nematic and Lamellar Liquid Crystals
4 Pressure-Temperature Phase Diagrams
-
0
02
0.4
0.6
0.8
Model parameter (a)
Figure 17. Theoretical plot of the reduced transition temperature against the model parameter a showing the hexagonal, nematic and isotropic phase boundaries. All the transitions are of first order [57, 581. (Reproduced by permission of the Royal Society).
the isotropic phase. It should be noted that, since the wavevectors of the hexagonal lattice satisfy the relation A +B-C=O, the columnar-nematic transition is always of first order; a Landau expansion of the free energy contains a non-vanishing cubic term in the order parameter 0. Ghose et al. [59] have extended this theory using a variational principle to solve the problem with the full potential rather than the one truncated to the first Fourier component. This method, which is closely analogous to the extension of McMillan's model of the smectic A phase by Lee et al. [60], leads to some qualitative improvements in the phase diagram for a homologous series. More recently, Ghose et al. [61] have presented a simpler mean field version of their theory, which yields essentially similar results.
Pressure studies have been carried out on the hexa-alkanoyloxybenzenes(1, R = C,IH2rl+l). Only three homologues ( n = 6, 7 and 8) are mesomorphic at atmospheric pressure, the last ( n = 8) being monotropic. On application of even small pressures, the n = 8 homologue becomes enantiotropic [62]. For all three compounds, the temperature range of the mesophase decreases with increasing pressure, and finally disappears (Figure 18). The crystal - mesophase - isotropic triple points occur at 2.96 kbar and 116.5"C, 1.43 kbar and 97.4"C, and 1.3 kbar and 101 "C for n =6, 7 and 8, respectively. On the other hand, the n = 5 homologue, which is non-mesomorphic at atmospheric pressure, shows pressure-induced mesomorphism [62]. The mesophase -isotropic tran-
T("C)
Figure 18. Pressure- temperature phase diagram of benzene hexa-n-octanoate. The circles and triangles represent two independent sets of measurements. (From Chandrasekhar et al. [62], reproduced by permission of the Commission des Publications Franpises de Physique).
4
I
0
I
50 Benzene hexa-n- heptanoate (96)
Pressure-Temperature Phase Diagrams
763
sition line extrapolated to atmospheric pressure yields a virtual transition temperature of 89°C (Figure 19a) [63]. Remarkably, this value agrees very well with that obtained from miscibility studies carried out by Billard and Sadashiva [64] (Figure 19b).
100
Figure 19. (a) Pressure- temperature diagram of benzene hexa-n-hexanoate. The mesophase- isotropic transition line extrapolated to atmosphere pressure yields a virtual transition temperature of 89 "C. (From Chandrasekhar et al. [63], reproduced by permission of Academic Press). (b) Miscibility diagram of benzene hexa-n-hexanoate and the heptanoate. The virtual mesophase-isotropic transition temperature for the hexanoate is 89 "C, in agreement with the value obtained from the pressure- temperature diagram (a). From Billard and Sadashiva [64], reproduced by permission of the Indian Academy of Sciences).
Figure 20. Pressure- temperature diagrams: (a) hexan-octyloxytriphenylene; (b) hexa-n-decanoyl-triphenylene. (- - - -) Monotropic transitions. (From Raja et al. [65], reproduced by permission of the Indian Academy of Sciences).
764
VIII
Columnar. Discotic Nematic and Lamellar Liquid Crystals
Pressure - temperature diagrams have been studied for two triphenylene compounds, hexa-n-octyltriphenylene (HOT; 2, R = C8H,,0) and hexa-n-decanoyloxytriphenylene (HDOT; 2, R = C9H19COO),both of which exhibit the columnar phase Dhoat atmospheric pressure. Two interesting results, common to both compounds, have been obtained [65]. Firstly, in contrast to what is usually observed in other liquid crystals, the columnar - isotropic (D- I) transition, which is enantiotropic at atmospheric pressure, becomes monotropic at high pressures. The triple points occur at 0.64 kbar and 84.5"C for HOT and 1.25 kbar and 123°C for HDOT. Secondly, dTldP is exremely small for the D - I transition, implying that, despite the drastic change in the molecular order at this transition (the heats of transition being 1.04 kcal mol-' for HOT and 0.69 kcal mol-' for HDOT [3, 66]), the associated volume changes are extremely small for both compounds (Figure 20).
5 Techniques of Preparing Aligned Samples Progress in precise quantitative experimental studies on discotic materials has been hindered to some extent because of the difficulty in preparing aligned samples. Vauchier et al. [67] investigated the effect of surface treatment on the anchoring of the director of the mesophases of certain triphenylene compounds. Glass surfaces coated with flat molecules (e.g., hexahydroxybenzene, mellitic acid, hexahydroxytriphenylene and rufigallol) orient the ND phase, and in some cases the D phases as well, with the director perpendicular to the wall (i.e. with the disks parallel to the
surface). Similar results have been obtained by slow cooling from the isotropic phase to the columnar phase of samples contained between aluminium-coated glass or quartz plates [68, 691. Freshly cleaved surfaces of crystals such as apophyllite and muscovite also have the same effect. Orientation of the N D phase with the director parallel to the surface has been achieved by the standard technique of oblique deposition of SiO. Twisted ND cells could be prepared in this manner [67]. As most D and ND phases are diamagnetically negative, the application of a magnetic field does not result in an oriented Sample as it does in the case of a calamitic system. The sample consists of domains the directors of which are distributed in a plane perpendicular to the field direction. Goldfarb et al. [13] demonstrated that a single domain can be obtained by allowing the mesogen to cool slowly from the isotropic phase in a magnetic field while spinning the sample about an axis perpendicular to the field direction. The single domain is formed with its director parallel to the spinning axis. Another technique, which has proved to be extremely useful for high resolution X-ray studies of columnar liquid crystals (see Sec. 2.1.2), is the preparation of highly oriented freely suspended strands [ 15- 19, 701. Figure 21 (a) shows a drawing of the temperature-controlled oven designed by Fontes et al. [ 191 for growing and annealing strands of the liquid crystal material in an inert atmosphere. The material is placed in a cup and a pin inserted into it. When the pin is slowly pulled out, a fibre or strand of about 150-200 pm diameter is drawn (Figure 21 b), with the column axes parallel to the axis of the strand (Figure 21 c). The oven has cylindrical beryllium walls, which permit 360 O X-ray access. Shearing of the glass plates has been found to produce aligned films of the co-
6 Ferroelectricity in the Columnar Phase
VACUUM SHROUD
UTER HEATER NNER HEATER
STRAND 1 CUP
(b)
. (c)
Figure 21. (a) The temperature-controlled oven designed to grow and anneal strands of the liquid crystal in an inert atmosphere. When the pin is slowly pulled out, a thin strand of the material is drawn (b), with the column axes parallel to the axis of the strand (c). (From Fontes et al. [ 191, reproduced by permission of the American Physical Society.).
765
metry arguments (see Sec. 2 of Chap. VI) it follows that each column is spontaneously polarized perpendicular to its axis. Macroscopic polarization is observed when the polarizations of the different columns do not annul each other. It is only recently that a ferroelectric columnar liquid crystal was realized experimentally. Bock and Helfrich [72, 731 demonstrated ferroelectric switching in the columnar phases of two compounds, namely 1, 2, 5,6, 8 , 9 , 12, 13-octakis ((S)-2-heptyloxypropanoyloxy)dibenzo[e,llpyrene (12) and a lower homologue with one less methylene unit in the hydrocarbon chain. Sheared samples, with the columns parallel to the glass surfaces and to the shear direction, exhibit the full optical tilt angle (the optical tilt angle being defined as the angle which the extinction cross makes with the polarizer- analyser directions (see Figure 2). This implies that the disk normals are inclined to the column axis, as discussed in Sec. 2.1 [ l 11. The polarization is then normal to the cell walls. On applying an alternating voltage, the optical tilt angle reverses with the field, going through an optically untilted state.
lumnar phase [71-731. The columns are oriented parallel to the glass surfaces and to the direction of shear. H3CAOn5Hl5
6 Ferroelectricity in the Columnar Phase Columnar phases can be ferroelectric if the discotic molecules are chiral and tilted with respect to the column axis [74]. From sym-
12
Compound 12 exhibits a field-induced transition from one ferroelectric phase to another at about 10 Vpm-'. The high field phase relaxes slowly to the low field phase when the field is switched off. The polariza-
766
VIII
Columnar, Discotic Nematic and Lamellar Liquid Crystals
tions of the low and high field phases are - 60 and - 180 nC cm-2, and the corresponding optical tilt angles are k24.5" and &37", respectively, and practically independent of temperature. The mechanism of reversal of the polarization is still not quite clear. It could be caused by (a) a rotation of the column as a whole, (b) the molecular tilt direction reversing through an untilted state (without rotation about the column axis), (c) the molecular tilt direction rotating around the column axis, or (d) a combination of these effects. The switching times ranged from 0.1 ms to 100 s in these experiments, depending strongly on temperature and field strength.
7 The Columnar Structure as a One-Dimensional Antiferromagnet The copper complex 11 with R = C,H,, contains a paramagnetic metal ion at its centre. Single crystal X-ray analysis has established that in this homologue the disks are stacked in columns in the crystalline state [50], indicating the possibility of formation of a one-dimensional exchange coupled system. Eastman et al. [75]have carried out detailed electron spin resonance studies on this copper complex in the crystalline and liquid-crystalline phases. Using the method discussed by Bartkowski and Morosin [76] for one-dimensional systems, Eastman et al. have analysed the angular dependence and line shapes, and drawn the important conclusion that the spectra of single crystals have the features of a spin-ione-dimensiona1 Heisenberg antiferromagnet, and, furthermore, that the exchange interactions are quite significant (i.e. an appreciable degree
of antiferromagnetic long-range order persists) even in the columnar mesophase.
Electrical Conductivity in Columnar Phases The columnar phases of metallomesogens have the electrical properties of molecular semiconductors [77 - 841. Figure 22 presents the Arrhenius plots of the a.c. conducof unaligned samples of copper tivity (0) phthalocyanines (13) determined by Van der Pol et al. [83]. The conductivity (T is of the order of 5 x lo-* Sm-' at 175 "C, and increases with increasing temperature. The activation energy is approximately 0.5 0.6 eV at 175 "C for the compounds 13 with R=OC,H,, and OC12H25. R
R
d
R 13
Like the triphenylene compounds, the columnar phases of pure polynuclear aromatic mesogens are insulators, but they can be made to conduct by doping [84-861. The d.c. conductivities of unaligned samples of HHTT in the pure form and after saturation doping with iodine (by exposure to iodine vapour) have been studied by Vaughan et al. [85]. The conductivities of pure HHTT and HHTT/I, are plotted in Figure 23, and as can be seen (T increases by 4-5 orders of magnitude as a result of doping. Figure 24 gives a schematic representation of the structure
8 Electrical Conductivity in Columnar Phases
0
t o
767
I
J
20
2.0
2.2
2.6
2.4 (
2.0
3.0
K-')
Figure 22. Arrhenius plots of the electrical conductivity (0) of unaligned samples of copper phthalocyanines(13).Datapoints: (A) R=H; ( 0 )R=OCH8H17; (0) R=OC12H2,. A slight increase in the conductivity is observed for the last compound at the transition from the mesophase to the crystalline phase (indicated by the arrow). (From Van der Pol et al. [83], reproduced by permission of Taylor and Francis Ltd.).
Figure 24. Proposed structure of the iodine-doped columnar (Dhd)phase of HHTT. The molecular cores are assembled in columns with short-range intracolumnar order and long-range hexagonal intercolumnar order. The black spheres in sets of three represent I; ions, which are assumed to occupy the space between the columns in a disordered fashion. For the sake of clarity, the liquid-like tails filling the space between the columns are not shown in the diagram. (From Vaughan etal. [SS], reproduced by permission of the American Physical Society).
1 0 ' o f G 7
,
0
1
5
I
60
40
Temperature
\'
HHTT:I~.I
7;
\ .,/ ?07F
ld'_ 1dST
167
.'
.*"%
K
t ( ! z
. I
I
I
of the iodine-doped columnar D,, phase, as envisaged by Vaughan et al. [85]. Figure 25 presents the a.c. conductivity measurements carried out by Boden et al. [86] on aligned samples of hexa-n-hexyloxytriphenylene (HAT6) (2, with R = C6H130) doped with a small quantity (mole fraction x = 0.005) of AlCl,. The conductivity oIl measured parallel to the column axis is about lo3times greater than o,,the conductivity measured in the perpendicular direction. This is a striking demonstration of the quasi-one-dimensional nature of the electrical properties of these materials. Calculations indicate that the hopping probability of the electron or hole is several orders of magnitude greater along the column axis than transverse to it, as is of course to be expected [82]. Boden et al. [86] have also investigated the dependence of the conductivity as a function of frequency and interpreted the results in terms of the Scher-Lax model [87] of hopping transport. The conduction along the columns is identified with
768
I
VIII
Columnar, Discotic Nematic and Lamellar Liquid Crystals
I
I
T (I0
Figure 25. Low frequency (100 Hz) conductivities of aligned samples of hexa-n-hexyloxytriphenylene doped with a small quantity (mol fraction x=0.005) of AlCI, in the columnar and isotropic phases plotted versus temperature. The conductivity measured parallel to the column axis (o,,) is about lo3 times greater than that measured in the perpendicular direction (oL). (From Boden et al. [86], reproduced by permission of Chapman & Hall).
a single transport process in which the carries hop between localized states (radical cations) associated with counterions (AlCI,) distributed alongside the columns. The subject is, needless to say, of considerable fundamental and practical importance, so much so that rapid advances may be expected to take place in the near future.
9 Photoconduction in Columnar Phases Photoconductivity of the discotic columnar phase is of much interest because of its potential applications, and several studies have been reported in recent years [68, 69, 88 -901. Probably the most significant of these is the discovery by Adam et al. [69] of fast photoconduction in the highly ordered columnar phase of a discotic mesogen. The experiments were performed on
HHTT, which exhibits the following sequence of transitions: Cr 62 Dho70 Dhd 93 I, where Cr is the crystalline phase, D,, is the highly ordered discotic phase, the structure of which is depicted in Figure 6, and I is the isotropic phase. The photocurrent transients were studied using a time-of-flight method. HHTT was sandwiched between two glass plates coated with semi-transparent aluminium electrodes. The cell thickness was about 30 pm. The interaction of the molecules with the flat electrodes resulted in homeotropic orientation with the column axes normal (or the disks parallel) to the glass surfaces. A pulsed (10 ns) N, laser of wavelength 337 nm, which coincides with the main absorption band of HHTT, generates electron hole pairs in the electrically biased cell. Depending on the polarity of the external field, the electrons or holes drift across the cell, causing displacement currents that can be recorded in an external circuit. Figure 26 gives the measured photocurrents for holes as a function of time t for the different phases of HHTT. The electric field applied was 2.0 x lo4 V cm-’ . Here tT represents the transit time. The form of the photocurrent in the Dhd and Dho phases can be explained by assuming Gaussian transport. A fairly compact charge carrier ‘packet’ drifts across the sample at a constant velocity, broadened only by the usual thermal diffusion, leading to a broadening of the photocurrent decay when the packet reaches the counter electrode. On the other hand, in the polycrystalline Cr phase, there is dispersive behaviour; the charge carrier packet is rapidly smeared out in time and space because of trapping events, and this results in a featureless decay. The hole mobilities are plotted versus temperature for the different phases in Figure 27. We will discuss the cooling mode first. In the isotropic phase the carrier mo-
10 Continuum Theory of Columnar Liquid Crystals
0
I
I
c
I
c n -
I
28 "c
Crystn1
0.0
I
J
1.0
2.0
t (ms)
Figure 26. Photocurrent as a function of time in the different phases of HHTT. A change of slope in the profile marks the transit time (tT) of the charge-cmier packet. The value of tT could be determined in the I, D,, and D,, phases, but not in the Cr phase which gives a featureless decay profile. (Adapted from Adam et al. [69], reproduced by kind permission of MacMillan Magazines Ltd.).
T(T)
60
!
2.60
2.80
..~00000
3.00
40
0
0
3.20
1/T(1O3K-'1
769
gesting a slight increase in the intracolumnar correlation. At the transition to the highly ordered D,, phase, p increases by almost two orders of magnitude and reaches values of the order of 0.1 cm2V-' s-l. With the exception of organic single crystals, these are the highest values known to date of electronic mobility for photoinduced charge carriers in organic systems. The mobilities are independent of temperature and the externally applied field. The high mobility and the rapid transit time in the D,, phase indicate an efficient charge transport mechanism with no major perturbation by disorder or defects, and good n-orbital overlap along the columns. In the crystalline phase the formation of grain boundaries and other defects inhibits efficient charge transport, and the behaviour is dispersive. In the heating mode, the defects in the crystalline phase persist in the D,, phase, so that the photocurrents remain dispersive. However, in the D,, phase, which consists of liquid-like columns, the mobilities measured in the heating and cooling modes remain the same. As Adam et al. [69] point out, these studies will no doubt have an impact on applications such as electrophotography, and on the development of new devices such as organic transistors and active layers in light emitting diodes.
Figure 27. Arrhenius plot of the hole mobility (p) versus the reciprocal of the temperature in the different phases of HHTT. Electric field = 2 x lo4 V cm-'. (From Adam et al. [69], reproduced by permission of MacMillan Magazines Ltd.).
bility is of the order of cm2V-'s-1, a value which is indicative of short-range orientational order. At the transition to the liquid crystalline D,, phase, p jumps by an order of magnitude, and increases with decreasing temperature in the mesophase, sug-
10 Continuum Theory of Columnar Liquid Crystals 10.1 The Basic Equations The hydrodynamic equations of the hexagonal columnar phase composed of liquid-
770
VIII
Columnar, Discotic Nematic and Lamellar Liquid Crystals
10.2 Acoustic Wave Propagation
like columns are [91]:
a& ao + P ) at at
In the absence of a temperature gradient, and assuming that permeation is negligible, i.e. U i = V;,and that damping is weak, it turns out that under adiabatic conditions there are three propagating acoustic modes for any arbitrary direction not coinciding with an axis of symmetry. Of these, one is the familiar longitudinal wave arising from density fluctuations, the velocity (V,) of which is practically independent of the direction of propagation. Another is a transverse wave, which is exactly analogous to the ‘second sound’ of the smectic A phase [92,93]. The velocities V1 and V2 of these two waves are given by the roots of the equation
Q. = - + -( E
dV 2= dt
ayP + ai @: + Vyjkla, aj v k
where:
a
+s i n 2 y co s 2 y [(B +D)A-C 2 ]=O
a a,aj=- a2 ax ax, x j a = -a22 a2 A a2~ a2= ax ay a2 32; = axax, (x; = x, y , z ) =-
~
+
ZZ
Q is the heat density; 0 is the volume strain; p is the density; V; are components of the velocity; KIIand Kl are the thermal conductivities parallel and perpendicular to the column axis, respectively; u; are components of the column displacement; P is the pressure; E is the energy density expressible as F + A O2 + C O ( u , , + u ~ , ~where ), F is given by equation (6) (see Sec. 7.10.3); A is the bulk modulus; C is a coupling constant; cis the permeation coefficient; 5 is the thermomechanical coefficient; and qtlijkl are viscosity coefficients.
4
where y is the angle between the direction of propagation and the column axis, and B and~ D are defined in Eq. (6) (see Sec. 10.3). We have ignored the coupling between the lattice and the curvature deformation of the columns. The third mode is a transverse wave the polarization of which is orthogonal to that of the second sound. This wave propagates because the two-dimensional lattice can sustain a shear. Its velocity for any arbitrary direction is given by 112
&=():
siny
The velocities of the three waves and their angular dependence can be studied by Brillouin scattering, as has been demonstrated for the smectic A phase [94]. In the present case, there should be one, two or three pairs of Brillouin components, depending on the orientation, but no experiments have yet been reported.
10 Continuum Theory of Columnar Liquid Crystals
10.3 Fluctuations In his treatment of the general problem of the stability of 'low-dimensional' systems, Landau also considered the case of a threedimensional body in which the density is a periodic function of two dimensions only. In the light of his analysis, he concluded: 'Thus, such a structure could in theory exist, but it is not known whether they do in fact exist in Nature' [20]. The structure envisaged by Landau is, of course, an accurate description of the columnar phase of discotic molecules that was proposed originally [l], i.e., liquid-like columns forming a twodimensional lattice. We shall now discuss the theory of thermal fluctuations in such a columnar structure. Let us suppose that the liquid-like columns are oriented along the z axis and that the two-dimensional lattice (assumed to be hexagonal) is parallel to the xy plane. The two basic deformations in such a structure are: (1) the curvature deformation (or bending) of the columns without distortion of the lattice, and ( 2 )the lattice dilatation (or compression) without bending of the columns. There can also be coupling between the two types of distortion, but the coupling term merely rescales the bend elastic constant of the columns [95]. Considering only the vibrations of the lattice in its own plane, the free energy may be written as [9 1, 961
77 1
lattice in its own plane, u, and uy are the displacements along x and y at any lattice point, and k,, is the Frank constant for the bending of the columns. In the notation of the standard crystal elasticity theory, B = 71 (cl +c12)and D = (cI I -c12). We neglect here the splay and twist deformations because they give rise to a distortion of the lattice and involve considerable energy. We also neglect any contributions from the surface of the sample. Writing the displacement u in terms of its Fourier components
0 )=
c 4 4 )exp(i!Z. r ) q
and substituting in Eq. (6) we get, in the harmonic approximation
F=
1 ~
c (44: + ko Si")
(u;)
2 ,
and from the equipartition theorem
where
B o = B + 2 D , k0=2k33 and q l =( q x2+q v2)
1/2
The mean square displacement at any lattice point is given by
where L' is the length of the columns, L is the linear dimension of the lattice in the xy plane and d is its periodicity. Assuming that L ' s L ,
where B and D are the elastic constants for the deformation of the two-dimensional
where A = (k,/Bo)"2 is a characteristic length [57, 96, 971. The structure is there-
772
VIII
Columnar, Discotic Nematic and Lamellar Liquid Crystals
fore stable as L -+w. As is well known from the classical work of Peierls [98] and Landau [99], the two-dimensional lattice itself is an unstable system with ( u 2 ) diverging as In L [loo, 1011. Thus the curvature elasticity of the liquid-like columns stabilizes the two-dimensional order in the columnar liquid crystal. In the two-dimensional lattice and the smectic A phase, the displacement - displacement correlation is of logarithmic form, and therefore the Bragg reflections (or the &function singularities in the X-ray structure factor) are ‘washed out’ for large samples. Instead, there results a strong thermal diffuse scattering with a power-law singularity [ 100- 1031. This has been verified quantitatively for the smectic A phase by the fine X-ray work by Als-Nielsen et al. [104]. On the other hand, the columnar liquid crystal gives the usual Bragg reflection, as borne out by the excellent high resolution X-ray studies by Fontes et al. [19] using oriented liquid crystal strands. The mean square amplitude of the orientational fluctuations of the director and the corresponding intensity of light scattering can be worked out in a similar way [55]. For example, when the incident and scattered beams are both polarized perpendicular to the column axis, the scattered intensity is r
n
qi
plates, a mechanical dilatation normal to the film gives rise to an undulation instability, which has been studied in detail [105]. An equivalent type of instability may be expected to occur in the columnar liquid crystal [95]. The columns are parallel to the glass plates, and the separation d between the plates is increased (Figure 28). If u,=ax (a being positive), then at a critical value a, given by 112
a periodic distortion should set in with a wavevector qz given by
Another type of instability may be described as a buckling instability. A compressive stress is applied parallel to the columns. At a critical value of the stress given by 7c2 k33 0,= _ _
d2
the columns buckle. This has actually been observed [ 1061,but there is a significant discrepancy between theory and experiment which has not been explained satisfactorily.
1
.1
- T W 4 [ ( B + D)q: + k33 qz
where qL= (q2+q;)1’2,and LC) is the angular frequency of the light. In certain geometries, the scattering is similar to that from the smectic A phase.
t“
10.4 Mechanical Instabilities When a smecticA film is taken between glass plates with the layers parallel to the
Figure 28. Undulation instability in a columnar liquid crystal subjected to a mechanical dilatation normal to the columns.
11 Defects in the Columnar Phase
773
11 Defects in the Columnar Phase The symmetry elements in the hexagonal structure composed of liquid-like columns are shown in Figure 29. The director n is parallel to the column axis; a (or, equivalently, a’ and a”) is the lattice vector of the two-dimensional hexagonal lattice; L,, T2 and 13, are two-fold axes of symmetry; L , is a three-fold axis and L6 a six-fold axis. Any point on L , or L6 is a centre of symmetryAny plane normal to n and the planes 156, T2 and L6, 0, are planes of symmetry.
b=a
2‘
Figure 29. The hexagonal columnar phase with the symmetry elements in the structure.
b
Figure 30. Dislocations in the columnar phase: (a, b) longitudinal edge dislocations; (c, d) transverse edge dislocation; (e, f) screw dislocations; (g) a hybrid of screw and edge dislocations. It should be noted that the Burgers vector b = a for (a), (c), (e) and (g), a ndb=a -a ’ for (b), (d) and (f). (From Bouligand [ 1071, reproduced by permission of the Commission des Publications Franpises de Physique).
774
VIII
Columnar, Discotic Nematic and Lamellar Liquid Crystals
The nature of the defects in such a structure has been investigated in detail by Bouligand [lo71 and by KlCman and Oswald [95, 108, 1091.
30 e and f). A hybrid composed of screw and edge dislocations is shown in Figure 30 g.
11.2 Disclinations 11.1 Dislocations The Burgers vector b of a dislocation in the columnar structure is normal to the columnar axis n. In general,
b = la +ma' where 1 and m are positive or negative integers. For edge dislocations, b is perpendicular to the dislocation line L. Two types of edge dislocation are possible: longitudinal edge dislocations (L parallel ton, Figure 30a and b), and transverse edge dislocations (L perpendicular to n, Figure 30c and d). For screw dislocations b is parallel to L (Figure
-
Longitudinal wedge disclinations are the standard crystal disclinations in a hexagonal lattice. The roQtion vector is L , or L , or L,, parallel to the columns. Two examples of such defects are shown in Figure 31 (a and b). The lattice becomes compressed near a positive disclination and stretched near a negative one. In general, most of these disclinations have prohibitively large energies, and hence they occur as unlike pairs at the core of a dislocation. In transverse wedge disclinations the rotation vector is 6, or T2,normal to n. Examples of k7t disclinations are shown in Figure 31 (c, d and e).
Figure 31. Disclinations in the columnar phase: (a, b) -7r.13 and a/3 longitudinal wedge disclinations about the six-fold axis L,; (c, d) a transverse wedge disclinations about the binary axes T2 and 02,respectively; (e) -n transverse wedge disclination leading to the formation of walls; (f) two n disclinations at right angles to each other, one about T2 and the other about 02.(From Bouligard [107]. reproduced by permission of the Commission des Publications Frdnpise de Physique).
12 The Properties of Discotic Nematic Phases
Two transverse wedge disclinations may occur in association, as shown in Figure 3 1 (f). The angle between the rotation axes may be 90", 60" or 30". Such defects have been observed experimentally [ 1091. The symmetry of the columnar phase also permits the occurrence of twist disclinations in the hexagonal lattice and of hybrids consisting of a twist disclination in the hexagonal lattice and a wedge disclination in the director field. According to Bouligand, these defects are not likely to exist. The energetics of defects in the columnar structure have been discussed in detail by Chandrasekhar and Ranganath [ 1101.
12 The Properties of Discotic Nematic Phases The symmetry of the discotic nematic phase ND (Figure 12) is the same as that of the classical nematic phase of rod-like molecules. Thus identical types of defects and textures are seen in both cases. The ND phase is optically, and, as a rule, diamagnetically, negative. However, its dielectric anisotropy A&(= ell - el, where and are the dielectric constants parallel and perpendicular to the director, respectively) may be positive or negative, depending on the detailed molecular structure. For example, the hexa-n-heptyloxybenzoate of triphenylene and hexa-n-dodecanoyloxytruxene are both dielectrically positive in the nematic phase [ I l l , 1121. This is because, while the contribution of the electronic polarizability of these disk-shaped molecules to A& is negative, the permanent dipoles associated with the six ester linkage groups make a stronger positive contribution, so that A&> 0 for the two compounds. On the other hand, hexakis(4-octylphenyl-
775
ethyny1)-benzene (lo),which does not have any permanent dipoles, does in fact show a negative A&, as expected [ 131. Very few quantitative measurements of the physical properties have been reported. The Frank constants for splay and bend have been determined using the Freedericksz method [ 111 - 1141. Interestingly, the values are of the same order as for nematic phases of rod-like molecules. The twist constant k,, has not yet been measured. The fact that the diamagnetic anisotropy is negative makes it somewhat more difficult to measure these constants by the FrCedericksz technique. However, by a suitable combination of electric and magnetic fields, it is possible, in principle, to determine all three constants [112]. The hydrodynamic equations of the classical nematic phase are applicable to the ND phase as well. There are six viscosity coefficients (or Leslie coefficients), which reduce to five if one assumes Onsager's reciprocal relations. A direct estimate of an effective average value of the viscosity of the ND phase from a director relaxation measurement [ 1 1 1, 1 141 indicates that its magnitude is much higher than the corresponding value for the usual nematic phase. In ordinary nematic phases of rod-like molecules, the Leslie coefficients p 2 and p3 are both negative and ( p J >lp31.Under planar shear flow, the director assumes an equilibrium orientation Oo given by tan2eo= p 3 / p , where 6, lies between 0" and 45" with respect to the flow direction (Figure 32 a). In practice, 0, is usually a small angle. In certain nematics, it is found that p3> O at temperatures close to the nematic - smectic A transient point. Under these circumstances there is no equilibrium value of e0,and in the absence of an orienting effect due to the walls or a strong external field the director
776
VIII
/ -v
-
(4
__ -- - ___
Columnar, Discotic Nematic and Lamellar Liquid Crystals
x----A!,,
(b)
Figure 32. Flow alignment of the director in nematic liquid crystals. For rod-shaped molecules the alignment angle 19with respect to the flow direction lies between 0" and 45" (a), while for disk-shaped molecules it lies between -90" and 45" (or, equivalently, between 90" and 135") (b).
13 Discotic Polymer Liquid Crystals
We end this chapter with a very brief account of a new class of liquid crystal polymers, i.e. discotic polymers [ I 18- 1261.The basic monomer units are discotic mesogenic moieties, which are components of the polymer main chain itself or are attached to the polymer backbone as side groups. A few examples are shown as structures 14-16. Besides the columnar phase, some new types of mesophase have been identified. A
starts to tumble and the flow becomes unstable. Consequently, such materials have two distinct flow regimes, which have been investigated in detail both theoretically and experimentally [ 1151. The suggestion has been made that in the N, phase, the disk-like shape of the molecule may have a significant effect on p2 and p3 [ 116, 1171. The stable orientation of the director under planar shear will now be as shown in Figure 32(b). Thus it can be argued that both p2 and p, should be positive, and the flow alignment angle Oo should
I
CH
14 Annealed:
glass 19 D 39 I [118]
kHJ$ ?+I 15 Glass 60 D 150 I [120]
lie between -45" and -90". It then follows that when p2c 0 the director tumbles and the flow becomes unstable. As yet, however, no experimental studies have been carried out to verify any of these ideas.
N+C
R,
R1
P
R,:-C-O-C,,Y,
R2 ,R,:-O-C,H,
16 D, 67 DS 130 I I1211
n
14 References
polyester with triphenylene as the repeating unit in the main chain separated by flexible spacers [ 1201, or with triphenylene units attached as side group to the polymer chain via flexible spacers [ 1261, forms a hexagonal columnar mesophase (Figure 33 a). On the other hand, rigid aromatic polyamides and polyesters with disk-shaped units in the main chain appear to form a nematic-like, or sanidic, structure [122], with boards stacked parallel to one another (Figure 33b). The addition of electron acceptor molecules such as 2,4,7-trinitrofluorenone, to discotic polymers results in the formation of charge transfer complexes that induce the columnar nematic phase (Figure 33 c) [124, 1251 similar to what is observed for low molar mass system (see Sec. 2.4). X-ray diffraction studies of the macroscopic alignment of the columnar structures produced by stretching a film or by drawing fibres (about 1 m long and several micro-
777
Fiber axis or Stretch direction
fb)
Figure 34. Orientations of the molecules in stretched samples of discotic polymers: (a) main chain discotic polymer; (b) side group discotic polymer. (From Huser et al., [126], reproduced by permission of the American Chemical Society).
metres thick) from the polymer melt have been reported [ 120, 1261. Huser et al. [ 1261 have demonstrated that almost perfect alignment can be achieved by proper mechanical and thermal treatment. Interestingly, in the main chain polymer, the columns are oriented perpendicular to the stretch direction (or the chain direction), whereas in the side group polymer they are oriented parallel to it (Figure 34). A study of the mechanical and other physical properties of these oriented polymers is of much interest. Acknowledgements
I am greatly indebted to Dr Geetha Nair and Dr S. Krishna Prasad for their valuable cooperation throughout the preparation of this chapter.
References
(C)
Figure 33. Mesophases of discotic polymers: (a) hexagonal columnar phase; (b) the board-like, or sanidic, nematic phase; (c) the columnar nematic phase.
[I] S. Chandrasekhar, B. K. Sadashiva, K. A. Suresh, Pramana 1977, 7,471 -480. [2] J. Billard, J. C. Dubois, N. H. Tinh, A. Zann, Now. J. Chim., 1978, 2, 535-540. [3] C. Destrade, M. C. Mondon, J. Malthete, J. Phys (France) 1979,40, C3-Cl7. [4] J. D. Bunning, J. W. Goodby, G. W. Gray, J. E. Lydon in Proceedings Conference on Liquid Crystals of One- and Two-Dimensional Order,
778
VlII
Columnar. Discotic Nematic and Lamellar Liquid Crystals
Garmisch-Partenkirchen (Eds.: W. Helfrich, G. Heppke), Springer-Verlag, Berlin, 1980, p. 397. [5] J. D. Bunning, J. E. Lydon, C. Eaborn, P. M. Jackson, J. W. Goodby, G . W. Gray, J. Chem. Soc., Faraday Trans. 1, 1982, 78, 713-724. [6] J. Malthete, A. M. Levelut, N. H. Tinh, J. Phys. Lett. (France) 1985,46, L875 -L880. [7] A. M. Levelut, J. Phys. Lett. (France) 1979,40, L8 1-L84. [8] V. Vill, LiqCryst - Liquid Crystal Database, LCI, Hamburg, 1995. [9] T. Plesnivy, H. Ringsdorf, P. Schumacher, U. Nutz, S. Diele, Liq. Cryst. 1995, 18, 185- 190. [lo] P. Spegt, A. Skoulios, Acta Crystallogr. 1966, 21,892-897. [ 1 I] F. C. Frank, S. Chandrasekhar, J. Phys. (France) 1980,41, 1285- 1288. [12] A . M . Levelut, J. Chim. Phys. 1983, 80, 149161. [ 131 D. Goldfarb, Z. Luz, H. Z. Zimmermann, J. Phys. (France) 1981,42, 1303- 1311. [14] S. Zumer, M. Vilfan, Mol. Cryst. Liq. Cryst. 1981, 70,39-56. 1151 D. H. Van Winkle, N. A. Clark, Phys. Rev. Lett. 1982,48, 1407- 1410. [16] C . R. Safinya, N. A. Clark, K. S. Liang, W. A. Varady, L. C. Chiang, Mol. Cryst. Liq. Cryst. 1985,123,205-216. [17] C. R. Safinya, K. S. Liang, W. A. Varady, N. A. Clark, G. Anderson, Phys. Rev. Lett. 1984, 53, 1172- 1175. [ 181 E. Fontes, P. A. Heiney, W. H. de Jeu, Phys. Rev. Lett. 1988, 61, 1202- 1205. 1191 E. Fontes, P. A. Heiney, M. Ohba, J. N. Haseltine, A. B. Smith, Phys. Rev. Ser. A , 1988, 37, 1329- 1334. [20] L. D. Landau, E. M. Lifshitz, Statistical Physics, Part Il, 3rd edn., Pergamon, Oxford, 1980, p. 435. [21] E. I. Kats, M. I. Monastyrsky, J. Phys. (France) 1984,45,709-714. 1221 Y. F. Sun, J. Swift, Phys. Rev., Ser. A, 1986,33, 2735-2739,2740-2744. [23] D. Ghose,T. R. Bose,M. K. Roy,M. Saha, C. D. Mukerjee,Mol. Cryst. Liq. Cryst. 1989,173,1729. [24] W. Kranig, C. Boeffel, H. W. Spiess, Macromolecules 1990,23,4061-4067. [25] M. Moller, J. H. Wendorff, M. Werth, H. W. Spiess, H. Bengs, 0. Karthaus, H. Ringsdorf, Liq. Cryst. 1994, 17, 381-395. [26] A.M. Levelut, J. Malthete, A. Collect, J. Phys. (France) 1986,47,351-357. [27] H. Zimmerman, R. Poupko, Z. Luz, J. Billard, Z. Naturforsch., Teil a 1985, 40, 149- 160. [28] J. M. Lehn, J. Malthete, A. M. Levelut, J. Chem. Soc., Chem. Commun. 1985, 1794- 1796. [29] N. H. Tinh, C. Destrade, A. M. Levelut, J. Malthete, J. Phys. (France) 1986,47, 553-557.
[30] C . Eaborn, N. H. Hartshorne, J. Chern. Snc. 1955,549-555. 1311 N. H. Tinh, H. Gasparoux, C. Destrade. Mol. Cryst. Liq. Cryst. 1981, 68, 101- 11 1. [32] C. Destrade, M. C . Bernaud, H. Gasparoux, A. M. Levelut, N. H. Tinh, in Proceedings ofthe International Conference on Liquid Crystals. (Ed.: S. Chandrasekhar), Heyden, London, 1979, p. 29-32. [33] B. Kohne, K. Praefcke, Chimia 1987,41, 196. [34] A. M. Levelut, F. Hardouin, H. Gasparoux, C. Destrade, N. H. Tinh in Proceedings ofrhe Conference on Liquid Crystal One and Two-Dimensional Order, Garmisch-Partenkirchen (Eds.: W. Helfrich, G. Heppke), Springer-Verlag, Berlin, 1980, p. 396. 1351 C. Destrade, H. Gasparoux, A. Babeau, N. H. Tinh, J. Malthete, Mol. Cryst. Liq. Cryst. 1981, 67,37-48. 1361 N. H. Tinh, J. Malthete, C. Destrade, J. Phys. Lett. (France) 1981,42, L417-419. 1371 K. Praefcke, D. Singer, B. Kohne, M. Ebert. A. Liebmann, J. H. Wendorff, Liq. Crysr. 1991, 10, 147- 159. [38] K. Praefcke, D. Singer,M. Langer, B. Kohne, M. Ebert, A. Liebmann, J. H. Wendorff, Mol. Cryst. Liq. Cryst. 1992, 215, 121-126. [39] H. Bengs, 0. Karthaus, H. Ringsdorf, C. Baehr, M. Ebert, J. H. Wendorff, Liq. Cryst. 1991, 10, 161- 168. 1401 C . Destrade, N. H. Tinh, J. Malthete, J. Jacques, Phys. Lett., Ser. A 1980, 79, 189-192. [41] J. Malthete, C. Destrade, N. H. Tinh, J. Jacques. Mol. Cryst. Liq. Cryst. Lett. 1981, 64. 233. [42] A. M. Giroud-Godquin, J. Billard, Mol. Cryst. Liq. Cryst. 1981, 66, 147- 150. [43] K. Ohta, H. Muroki, A. Takagi, K. I. Hatada, H. Ema, I. Yamamoto, K. Matsuzaki, Mol. Cryst. Liq. Cryst. 1986, 140, 131-152. 1441 H. Sakashita, A. Nishitani, Y. Sumiya. H. Terauchi, K. Ohta, I. Yamamoto, Mnl. Crysr. Liq. Cryst. 1988, 163, 211-219. [45] A. C. Ribeiro, A. F. Martin, A. M. Giroud-Godquin, Mol. Cryst. Liq. Cryst. Lett. 1988, 5 , 133-139. [46] A. M. Giroud-Godquin, P. M. Maitlis, Angen: Chem., lnt. Ed. Engl. 1991, 30, 375-402. [47] K. Usha, K. Vijayan, B. K. Sadashiva, Mol. Cryst. Liq. Cryst. Lett. 1987, 5, 67 -7 1. [48] K. Usha, K. Vijayan, Mol. Cryst. Liq. Cryst. 1989, 174, 39-48. 1491 K. Usha, K. Vijayan, B. K. Sadashiva, P. R. Rao, Mol. Cryst. Liq. Cryst. 1990, 185, 1 - 11. [SO] K. Usha, K. Vijayan, B. K. Sadashiva, Mol. Cryst. Liq. Cryst. 1991, 201, 13-21. [51] K. Usha, K. Vijayan, Liq. Cryst. 1992, 12. 137-145. [52] K. Usha, K. Vijayan, S. Chandrasekbar, Liq. Cryst. 1993, 15,575-589.
14 References [53] B. Muhlberger, W. Haase, Liq. Cryst. 1989, 5, 25 1- 263. [54] W. L. McMillan, Phys. Rev., Ser. A, 1971, 4, 1238- 1246. 1551 E. I. Kats, Sov. Phys.-JEZ'P 1978,48,916-920. [56] G. E. Feldkamp, M. A. Handschy, N. A. Clark, Phys. Lett., Ser. A 1981, 85, 359-362. [57] S. Chandrasekhar, Phil. Trans. R. Soc., London, Ser. A 1983,309,93. [58] S. Chandrasekhar, K. L. Savithramma, N. V. Madhusudana, ACS Symp. Sel: 1984, 4, 299309. [59] D. Ghose, T. R. Bose, C. D. Mukherjee, M. K. , Roy, M. Saha, Mol. Cryst. Liq. Cryst. 1986,138, 379 - 392. [60] F. T. Lee, H. T. Tan, Y. M. Shih, C. W. Woo, Phys. Rev. Lett. 1973,31, 1117-1120. [613 D. Ghose, T. R. Bose, M. K. Roy, C. D. Mukherjee, M. Saha, Mol. Cryst. Liq. Cryst. 1988, 154, 119- 125. [62] S. Chandrasekhar, B. K. Sadashiva, K. A. Suresh, N. V. Madhusudana, S. Kumar, R. Shashidhar, G. Venkatesh, J. Phys. (France) 1979, 40, C3-120-124. 1631 S. Chandrasekhar, Adv. Liq. Cryst. 1982, 5, 47-78. [64] J. Billard, B. K. Sadashiva, Pramana 1979, 13, 309-3 18. [65] V. N. Raja, R. Shashidhar, S. Chandrasekhar, R. E. Boehm, D. E. Martire, Pramana J. Phys. 1985,25, L119-L122. 1661 C. Destrade, N. H. Tinh, H. Gasparoux, J. Malthete, A. M. Levelut, Mol. Cryst. Liq. Cryst. 1981, 71, 111-135. 1671 C. Vauchier, A. Zann, P. Le Barny, J. C. Dubois, J. Billard, Mol. Cryst. Liq. Cryst. 1981, 66, 103- 1 14. [68] D. Adam, F. Closs, T. Frey, D. Funhoff, D. Haarer, H. Ringsdorf, P. Schuhmacher, K. Siemensmeyer, Phys. Rev. Lett. 1993, 70,457. 1691 D. Adam, P. Schuhmacher, J. Simmerer, L. Haussling, K. Siemensmeyer, K. H. Etzbach, H. Ringsdorf, D. Haarer, Nature 1994, 371, 141143. [70] P. Davidson, M. Clerc, S. S. Ghosh, N. C. Maliszewskyi, P. A. Heiney, J. Hynes, A. B. Smith, J. Phys. 11(France) 1995, 5, 249-262. [7 I ] 1. Itoh, A. Tanaka, F. Fukuda, T. Miyamoto, Liq. Cryst. 1991, 9, 221 -228. [72] H. Bock, W. Helfrich, Liq. Cryst. 1995, 18, 387 -399. [73] H. Bock, W. Helfrich, Liq. Cryst. 1992, 12, 697 - 703. [74] J. Prost in Symmetries and Broken Symmetries (Ed.: N. Boccara) IDSET, Paris 1981, p. 159. 1751 M. P. Eastman, M. L. Horng, B. Freiha, K. W. Shen, Liq. Cryst. 1987,2, 223-228. 1761 R. R. Bartkowski, B. Morosin, Phys. Rev., Sel: B 1972,6,4209-4212.
779
[77] J. Simon, C. Sirlin, Pure Appl. Chem. 1989, 61, 1625- 1629. 1781 C. Piechocki, J. Simon, A. Skoulios, D. Guillon, P. Weber, J. Am. Chem. Soc. 1982, 104, 52455247. 1791 B. A. Gregg, M. A. Fox, A. J. Bard, J. Am. Chem. SOC.1989,111,3024-3029. 1801 P. G. Schouten, J. M. Warman, M. P. deHaas, M. A. Fox, H . L . Pan, Nature 1991, 353, 736737. [8 11 S. Gaspard, P. Millard, J. Billard, Mol. Crysr. Liq. Cryst. 1985,123,369-375. 1821 Z. Belarbi, M. Maitrot, K. Ohta, J. Simon, J. J. Andre, P. Petit, Chem. Phys. Lett. 1988, 143, 400-403. [83] J. F. Van der Pol, E. Neelemen, J. W. Zwikker, R. J. M. Nolte, W. Drenth, J. Aerts, R. Visser, S. J. Picken, Liq. Cryst. 1989, 6 , 577-592. 1841 N. Boden, R. J. Bushby, J. Clements, M. V. Jesudason, P. F. Knowles, G . Williams, Chem. Phys. Lett. 1988, 52, 94-99. 1851 G. B. M. Vaughan, P. A. Heiney, J. P. McCauley Jr, A. B. Smith 111, Phys. Rev., Ser. B, 1992, 46, 2787-2791. 1861 N. Boden, R. J. Bushby, J. Clements, J. Matel: Sci. Matel: Electron. 1994, 5, 83 - 88. 1871 H. Scher, M. Lax, Phys. Rev., Sel: B 1973,7,4491 -4502. [88] N. Boden, R. J. Bushby, A. N. Cammidge, J. Clements, R. Luo, K. J. Donovan, Mo/. Crysr. Liq. Cryst. 1995, 261, 25 1- 257. [89] H. Bengs, F. Closs, T. Frey, D. Funhoff, H. Ringsdorf, K. Siemensmeyer, Liq. Cryst. 1993, 15,565-574. 1901 F. Closs, K. Siemensmeyer, T. H. Frey, D. Funhoff, Liq. Cryst. 1993, 14, 629-634. 1911 J. Prost, N. A. Clark in Proceedings of the International Conference on Liquid Crystals (Ed.: S. Chandrasekhar), Heyden, London, 1979, 53-58. 1921 P. G. deGennes, J. Prost, Physics of Liquid Crystals, 2nd edn., Clarendon Press, Oxford, 1993, 421 -423. [93] S. Chandrasekhar, Liquid. Crystals, 2nd edn., Cambridge University Press, Cambridge, 1992, 324 - 326. [94] Y.Liao, N. A. Clark, P. S . Pershan, Phys. Rev. Lett. 1973, 30, 639-641. 1951 M. KICman, P. Oswald, J. Phys. (France) 1982. 43,655 -662. 1961 G. S. Ranganath, S. Chandrasekhar, Curl: Sci. 1982,51,605-606. 1971 V. G. Kammensky, E. I. Kats, Zh. Eksp. Teor. Fi: 1982,56,591-594. [98] R. E. Peierls, Helv. Phys. Actu. Suppl. 1934, 7, 81. [99] L. D. Landau in Collected Papers of L. D. L a w dau (Ed.: D. Ter Haar), Gordon & Breach, New York, 1967, p. 210.
780
VIII
Columnar, Discotic Nematic and Lamellar Liquid Crystals
[loo] B. Jancovici, Phys. Rev. Lett. 1967, 19,20-22. [ 1011 L. Gunther, Y. Imry, I. Lajzerowicz, Phys. Rev., Ser. A 1980,22, 1733- 1740. 11021 J. D. Litster in Proceedings of the Conference on Liquid Crystals of One- and Two-dimensional Order, Garmisch-Partenkirchen, FRG (Eds.: W. Helfrich, G . Heppke), Springer-Verlag, Berlin, 1980, 65 -70. [ 1031 J. Als-Nielsen in Symmetries and Broken Symmetries (Ed.: N. Boccara) IDSET 1981, 107122. [I041 J. Als-Nielsen, J. D. Litster, R. J. Birgeneau, M. Kaplan, C. R. Safinya, A. Lindegaard-Andersen, S. Mathiesen, Phys. Rev., Ser. B 1980, 22,313-320. [lo51 G. Durand, Pramana, (Suppl. l), 1973, 23. [I061 M. Cagnon, M. Gharbia, G. Durand, Phys. Rev. Lett. 1984,53, 938-940. [I071 Y. Bouligand, J. Phys. (France) 1980, 41, 1297-1306, 1307-1315. [ 1081 M. KlCman, J . Phys. (France) 1980, 41, 737 745. [I091 P. Oswald, J. Phys. (France) 1981, 42, L-171. [ I 101 S. Chandrasekhar, G. S. Ranganath,Adv. Phys. 1986,35,507-596. [ I 1 11 B. Mourey, J. N. Perbet, M. Hareng, S. Le Berre, Mol. Cryst. Liq. Cryst. 1982, 84, 193199. [ 1 121 V. A. Raghunathan, N. V. Madhusudana, S. Chandrasekhar, C. Destrade, Mol. Cryst. Liq. Cryst. 1987, 148, 77-83. [ I 131 G. Heppke, H. Kitzerow, F. Oestreicher, S. Quentel, A. Ranft, Mol. Cryst. Liq. Cryst. Lett. 1988,6,71 -79.
[114] T. Warmerdam, D. Frenkel, J. J. R. Zijlstra, J. Phys. (France) 1987,48,319-324. [ I 151 S. Chandrasekhar, U. D. Kini in Polymer Liquid Crystals (Eds.: A. Ciferri, W. R. Krigbaum, R. B. Meyer) Academic Press, New York, 1982,201-246. [116] T. Carlsson, J. Phys. (France) 1983, 44, 90991 1. 11171 G. E. Volovik, Zh. Eksp. Teor. Fiz Lett. 1980, 31,273-275. [ 1181 W. Kreuder, H. Ringsdorf, Makrornol. Chem. Rapid Commun. 1983,4,807- 8 15. [119] G. Wenz, Makromol. Chem. Rapid Cammun. 1985,6,577-584. [120] 0. Herrmann-Schonherr, J. H. Wendorff, W. Kreuder, H. Ringsdorf, Makromol. Chem. Rapid. Cornmun. 1986, 7, 97- 101. [121] 0. Herrmann-Schonherr, J. H. Wendorff, H. Ringsdorf, P. Tschirner, Makromol. Cheni. Rapid. Commun. 1986, 7,791 -796. [122] M. Ebert, 0. Herrmann-Schonherr, J. H. Wendorff, H. Ringsdorf, P. Tschirner, Makrornol. Chem. Rapid Commun. 1988,9,445-45 1. 11231 B. Kohne, K. Praefcke, H. Ringsdorf, P. Tschirner, Liq. Cryst. 1989,4, 165- 173. [ I 241 H. Ringsdorf, R. Wustefeld, E. Zerta, M. Ebert, J. H. Wendorff, Angew. Chem., Int. Ed. E q I . 1989, 28, 914. [125] H. Ringsdorf, R. Wiistefeld, Molecular Chemistry f o r Electronics, The Royal Society, London, 1990, p. 23. [126] B. Huser, T. Pakula, H. W. Spiess, Macromolecules 1989, 22, 1960- 1963.
Handbook of Liquid Crystals Edited by D. Demus, J. Goodby, G. W. Gray, H.-W. Spiess, V. Vill Copyright 0 WILEY-VCH Verlag GmbH. 1998
Chapter IX Applicable Properties of Columnar Discotic Liquid Crystals Neville Boden and Bijan Movaghar
1 Introduction We start by reminding ourselves that columnar discotic liquid crystals are comprised of disordered stacks (1-dimensional fluids) of disc-shaped molecules arranged on a twodimensional lattice (Fig. 1) [ 11. This structure imparts novel properties to these materials from which applications are likely to stem. One such property is the transport of charge along the individual molecular stacks [2-71. The separation between the aromatic cores in, for example, the hexaalkoxytriphenylenes (HATn), the archetypal columnar discotic mesogen, is of the order of 0.35 nm, so that considerable overlap of 7c* orbitals of adjacent aromatic rings is
insulating liquid hydrocarbon matrix
conducting aromatic cores
-LUMo -HOMO Molecular
expected to lead to quasi-one-dimensional conductivity [8]. It turns out, however, that the HATn have a very low intrinsic carrier concentration, owing to the large band-gap in these materials. They are therefore good insulators (Table 1) with high breakdown fields of =4 MV cm-' [9]. The mobility of an injected charge along the molecular columns (wires) is, however, fairly high [lo-121. The combination of these properties, together with other unique properties of columnar liquid crystals such as the selforganizing propensity of the molecular columns [ 131 which eliminates long lived grain boundaries which trap charge, and the ability readily to wet the surface of metals and
8
Condensed State
Figure 1. Schematic view of a discotic columnar liquid crystal Col, phase, a simplified electronic structure of an isolated molecule, and the electronic structure in the gap in the condensed state showing the band gap Eg and the bandwidth AE.
Table 1. Comparison of some electrical properties of HAT6 with those of the organic metal TTF/TCNQ [8] and pure silicon [ 9 ] . OK-'
HAT6 TTF/TCNQ Si
cm-'
p/cm2 V-' s-' 4 x lo4
500
1
10-3
102
n/~m-~ 10'0 102' 1013
E,IeV
AEIeV
4 2
0.2 0.2-0.5 3-4
1.1
782
IX
Applicable Properties of Columnar Discotic Liquid Crystals
semiconductors which enables charges to be injected from electrodes into the molecular wires [ 141, could be harnessed in a variety of new applications. In this short review of applicable properties, we will examine the electrical properties of columnar liquid crystals in some detail and discuss their potential application in light scanners, xerography, chemical and light sensors, molecular scale FETs, p-n junctions, ferroelectric memories and electrical contacts [8, 91. We will also consider fluorescent and ferroelectric properties.
2 Molecular StructureProperty Relationships We will consider some physical examples of columnar liquid crystals so as to estab-
lish the relationship between molecular structure and their electrical and optical properties.
2.1 HATn Materials Much work has been focused on the HATn materials (Scheme 1). Their electrical properties are typical of wide band-gap transparent columnar liquid crystals in general and can be considered as a model system. The HAT-n are chemically stable, important for electro-optical applications; their chemistry is often more accessible that that of other discogens such as the phthalocyanines. These materials can now be produced in large quantities, in high purity and at an economic price. Routes for synthesizing low molar mass discotics, polymeric
RO
RO
OR Cr (67)Col(98)1
Cr(GT)Col(136) I
Scheme 1. HAT-n materials
2
Molecular Structure-Property Relationships
Cr (78) C01,,(305) I
Cr (83) Colh,(309)I
R = C12H25 Pc Zn
R = C12Hz
783
Scheme 2. The Phthalocyanines.
discotics, and for introducing substituents into the triphenylene ring are now available [ 15-19]. The latter is important not only for fine tuning their electrical and optical properties, but also for producing room temperature liquid crystals essential for applications.
2.2 The Phthalocyanines The PhthalocYanine discogens (Scheme 2) have been extensively studied and their synthesis and Properties are well d o c ~ ~ n t e d in books and articles [20, 2 1J and references therein. These, together with the porphyrim and metal-porphyrins form phases which typify narrow band-gap columnar liquid crystals [21-241.
784
IX
Applicable Properties of Columnar Discotic Liquid Crystals
Table 2. Comparison of the main properties of wide band-gap and narrow band materials. ~
Key properties
HATn and related compounds Phthalocyanines and related compounds
Band gap Electronic mobility Mobility anisotropy Undoped conductivity Doped conductivity Fluorescence yield 0, binding Charge carrier gen. efficiency Clearing temperatures Birefringence Ionization energy (condensed phase)
4 to 5 eV lo-' to cmz V-' s-' > lo3 in undoped state lo3 in doped state Sfcm T=300 K S/cm T = 300 K Qf-O.1, T=300 K =
C C, COS, @
(26)
where # is the torsional angle which is defined to be zero in the trans conformation and b takes integral values from 0 to 5. The molecular field equations for the continuous potential can be obtained more or less from those which we have given for the discrete model (see Eqs. 15, 17, 18, 20-23) by replacing the summations over the discrete
828
X *. 30
070
060
Liquid Crystal Dimers and Oligomers
,
I
1
I
,
3
5
7
9
71
13
15
17
19
n
Figure 26. The dependence of the nematic-isotropic transition temperature, T$l1, scaled with that for the sixth homologue on the number of atoms, n, in the spacer of the cyanobiphenyl dimers predicted with the continuous model for the torsional potential (-) and by the discrete model (. ...). Figure 27. The variation of the transitional entropy,
conformations with integrations over the torsional angles [82]. For long chains the number of such integrations is clearly large which complicates the calculations considerably and makes it impossible to include all conformational states, unlike the discrete model. This difficulty may be overcome by using a biased Monte Carlo scheme to select the torsional angles based on the internal energy, Uint(@),alone and with the additional weighting resulting from the orientational partition function (see Eq. 17) being introduced specifically, as in umbrella sampling [84]. The results of these calculations for the N-I transition temperature based on the continuous torsional model are shown in Fig. 26 for the methylene linked cyanobiphenyl dimers with the spacer containing from 3 to 18 atoms. For ease of comparison, the transition temperatures have been scaled with the value for the sixth member of the homologous series. The results reveal that the alternation in TNIis rapidly attenuated with increasing spacer length and that for spacers containing more than about 11 atoms, the alternation in TN-I is essentially
removed. This is, of course, the behavior which we had anticipated, at least, qualitatively, for a continuous model. However, to ensure that this behavior does not result from some other feature of the molecular field calculations they were repeated for the discrete model in which the spacer adopts just the trans and gauche conformations of the Flory model. In this molecular field calculation, the energy difference between the trans and gauche conformations E,, was given the value predicted by the RyckaertBellemans torsional potential, all other parameters were kept the same as for the continuous model. The dependence of TNPIon the number of atoms in the spacer is also shown in Fig. 26 and we can see that the alternation is larger and that the attenuation is weaker. This confirms, therefore, the major influence which the continuous nature of the torsional potential has on the N-I transition temperature for dimers containing relatively long spacers. The entropy of transition was also calculated for the continuous and discrete torsional potential models with the results shown in Fig. 27. The results for the continuous model reveal that as the spacer length increases the larger alternation in ASIR is attenuated although this does not occur to any significant extent until the spacer contains 15 or more atoms. The results predicted by the discrete model are also given in Fig. 27 and these exhibit an enhanced alternation in the transitional entropy with increasing spacer length. It is of interest that for short spacers the values for ASIR predicted by the two models are very similar. However, as the spacer length grows so the transitional entropy predicted by the discrete model become significantly greater than those given by the continuous model for even dimers. In contrast, the results predicted by both models for the odd dimers are essentially the same. This is analogous to the behavior
7 Molecular Shapes of Liquid Crystal Dimers
AS/R
829
7--. I not change ~
0' 1
' 3
~
' 5
"
" 7
"
'
" 9
"
' 11
' 13
17
15
19
,I
ASIR, for the cyanobiphenyl dimers with the number of atoms, n, in the flexible spacer predicted with the continuous model for the torsional potential (-) and by the discrete model (. ...). For comparison the corresponding predictions for the alkyl cyanobiphenyl monomers based on the continuous potential are also shown (-).
significantly at the N-I transition. Of greater interest is the observation that the alternation in ASIR for the monomers is essentially identical to that predicted for the dimers, again for the continuous torsional potential for spacers containing 19 or more atoms. It would seem, therefore, that the ability to undergo fluctuations around the torsional minima decouples the mesogenic groups and makes the dimers behave like monomers for relatively long spacers. Such dimers have yet to be made which would allow this fascinating prediction and, by implication, the validity of the Flory rotational isomeric state model to be tested.
Figure 28. The images created for the average mo-
found for the N-I transition temperature where the values predicted by the discrete model for even spacers are significantly larger than for the continuous model whereas the predictions for the odd dimers are essentially the same for the two models (see Fig. 26). Another way to consider these predictions for the continuous potential model is that in the limit of long spacers the relative orientations of the two mesogenic groups will be decoupled. In this limit it would then be expected that the dimers should behave like monomers. To see if this expectation is likely to be met the transitional entropy for an alkyl cyanobiphenyl was calculated using the continuous model with the same parameters as for the dimers. The results are shown in Fig. 27 and reveal that as the length of the terminal chain is increased so ASIR exhibits an overall increase with a small alternation superimposed on this. In addition, the transitional entropy for the monomers is comparable to that for the odd dimers, in keeping with the fact that the conformational distribution for the dimers does
7 Molecular Shapes of Liquid Crystal Dimers The anisometric shape of a mesogenic molecule is thought to play an important role in determining its liquid crystal behavior. Indeed, as we have seen it is the differing shapes of odd and even dimers in their alltrans conformations which are often invoked to explain the strong alternation in the transitional behavior along a homologous series. However, as we have also seen, this difference in the shape of a single conformer is not the entire story and it is also necessary to allow for the many conformational states adopted by the spacer to account for the characteristic odd-even effect exhibited by liquid crystal dimers. The question then arises as to how the average shape of a nonrigid molecule which can assume different shapes can be defined and whether such average structures could be of value in understanding the behavior of liquid crystal dimers. One definition for the average shape of a nonrigid molecule which has been proposed is the following [ 8 5 ] .The
830
X
Liquid Crystal Dimers and Oligomers
coordinates of the atoms, r:, in the nth conformer are averaged over all conformers
Fi = Cpnr," n
(27)
Then, in order to obtain a chemical-like structure, those atoms which are covalently bonded are linked together. Clearly in such an average structure the bond lengths and bond angles will differ from their values in rigid molecules. So as to emphasize still further that the image represents an average molecular structure and that the molecule should not be thought of as existing in this single structure each atom is decorated with an ellipsoid, rather like the thermal ellipsoids used in crystallographic structures. The dimensions of the ellipsoids are proportional to the weighted root mean square displacement of the atom away from its average position
a? = Cp,300 0 0 0
Table 14. Transition temperatures ["C] of compounds 38. Compound
Number and positions of chains
a b
'P a) 2mp b) 2mm 3mpm
C
d a)
On heating;
b,
on cooling.
Cr
SmC 111 122
261 142 118
0
89 71
Cub
-
-
-
N
@h
170 161
(0
163 80) 112
-
-
I 297 0 0 0 0
879
5 The Core, Paraffinic Chains, and Mesomorphic Properties
Table 15. Transition temperatures ["C] of compounds 39. Compound
Number and positions of chains
a b
1P 2mP 2mm 3mpm
C
d
Cr
Cub 107 120 72 81
0 0
0
114
-
-
I
@h
@?
-
177
0
125 50) 84
0 (0
0
0
Table 16. Transition temperatures ["C] of compounds 40. Compound
Number and positions of chains
Cr
1P 2mp 2mm 3mpm
0
a b C
d
Cub 74 89 80 47
0
-
Table 17. Transition temperatures [TI of compounds 41. n 1 2 3 4 5 6 7 8 9 10 I1 12
Cr
0
210 203 160 154 141 128 120 116 118 118 119 119
SmC
SmA
-
-
(0
-
I
-
121
-
107 60
-
147 0 0
0
owing to the 2mp cyanohydrogenated chains. In the same way an SmA phase was also observed with the pentacatenar 42b in which there is a second meta chain.
-
-
(0
N
99
I
Cub
@h
@?
113)
136 124
144) 154.5 156 154 153 153 151 147 147
5 TheCore, Paraffinic Chains, and Mesomorphic Properties 5.1 Core Length
o-Cyano Chains. The w-cyano chains associate the flexibility of the paraffins with the polarity of the cyano group. They are used in order to obtain orthogonal lamellar mesophases such as an SmA phase instead of tilted SmC phases. In fact, the biforked mesogens 42a displayed an SmA phase
For the reviewed compounds, at least three benzene rings are needed in order to obtain mesomorphic properties (Sec. 4.1.4 of this chapter, compounds 27-29 and Sec. 4.2.1 of this chapter, compounds 31 -33) of clasRl
/
42 a RI = H, R2 = RS = n-C12Hzj0 ; K 127OC Sc 1 5 8 T SA 189°C I b R1 = R2 = R3 = n-ClzH250 ; K 10l°C SA 155°C I
R3
880
XI1 Phasmids and Polycatenar Mesogens
sical elongated molecules. It is clear that a balance must be kept between the number of chains directly grafted on the core and the length of the latter in order to obtain mesomorphic properties. For example, with four chains, a minimum of a four-ring core is necessary to obtain mesomorphic properties. With five and six chains, at least five rings are necessary for mesomorphism. Let us also recall that, owing to its stiffness and its incompatibility with hydrocarbons, a fluorinated chain acts simultaneously as a core and as a chain and thus allows the ring number to be reduced without losing the mesomorphic properties (Sec. 4.2.1.).Both core length and chains are determinant for the existence of lamellar and columnar phases.
5.2 Number and Position of Chains This discussion only concerns hydrogenated chain compounds with five- and six-benzene cores, which present the richest polymorphism. Table 18 shows that when Nm>Np, columnar phases are obtained, when Nm 2OO0C) a SmC phase. Replacing the two chloride ligands in the Pd(I1) complexes by aliphatic carboxylic acids lowers melting and clearing temperatures [ 14al. Trisubstituted ligands lead to complexes showing columnar organisation [ 14b].
903
show SmA and SmC mesophases, while for shorter chain length the nitrate salts show only a SmA phase and the triflate salts only a nematic phase [ 151. The dodecyl sulfate (Y = C,,H2,SOp) salts formed mesophases and showed unusual behavior. A rich polymorphism is found: N, SmA, SmC and cubic phases [20]. The single crystal X-ray structure of the methoxy homologue has been determined [16]. This rich and varied polymorphism in these ionic systems may have some electrochemical applications. This mesomorphism is disturbed by lateral fluorination of the ligand: 3-fluorinated isomers promote smectic phases and suppress the cubic phase while the 2-fluorinated isomers promote both the nematic and the cubic mesophase [ 171.
3.2.1.3 Iridium Iridium complexes of dimeric stilbazole, 9, have been synthesized recently. These materials show avery large odd-even effect in their melting behavior, and are called metal-containing ‘Siamese Twin’ liquid crystals [ 181.
9
‘Siamese Twin’ complexes
7 Palladium carboxylate complexes
3.2.2
3.2.1.2 Silver Reaction of the stilbazoles with silver salts, AgY, led to a quite unexpected series of linear mesogenic materials: the ionic bis-ligand silver salts, 8, [A~(IZ-OST)~]Y with Y = BF?, c , , H ~ ,sop,CF,OSO,O. The long chain homologues of each series
NOP,
8 Silver complexes
Monostilbazole Ligands
A way to lower transition temperatures is to incorporate some asymmetry by replacement of one of the stilbazoles with an alkene ~91.
904
XIV
Metal-containing Liquid Crystals
3.2.2.1 Platinum In the Pt complexes, 10, the chain lengths n and m were systematically varied ( n = 3 - 12; m = 0 - 8). SmA phases were seen for most derivatives. It was also noted that complexes with n + m < 8 were not mesomorphic; those with 8 < n + rn < 13 showed monotropic mesomorphism, while those with n + rn > 13 showed enantiotropic SmA phase [14].
ored and the disubstituted one shows an unusual behavior on cooling giving an unidentified liquid crystal mesophase [22].
oc--'w -co
co I c'o
L-'w
-co
I \co co
10 Platinum alkene complexes
3.2.2.2 Rhodium and Iridium A series of unsymmetric rhodium and iridium stilbazoles, 11, was prepared.
12 Tungsten carbonyl complexes
3.3 Other Pyridine Ligands I c1 11 Stilbazole complexes of rhodium and iridium
The rhodium complexes were yellow/ orange in the solid state characteristic of mononuclear Rh(I), while the burgundy iridium complexes became yellow both in solution and on melting. Both series of materials showed very similar mesomorphism: nematic phases were found at short chain lengths giving way to the SmA phase for higher homologues for the two metals 1201. Similar 2- and 3-fluorinated derivatives have been synthesized to lower the transition temperatures. The 2-fluorinated iridium complexes show monotropic N and SmA phases while the 3-fluorinated compounds are not mesomorphic [21].
3.2.2.3 Tungsten Monosubstituted and disubstituted tungsten carbonyl complexes, 12, are intensely col-
3.3.1 Rhodium and Iridium Serrano et al. [23] have used other substituted pyridines as ligands for metallomesogen formation. Thus the cis-iridium complexes, 13, with short chains formed monotropic nematic phases, while those with longer chains led to SmA phases. The related rhodium complexes are also mesogenic. This is an example of a mesogenic complex formed with a nonmesogenic ligand.
13 Alkoxypyridylbenzylideneaniline complexes
3.3.2 Silver The silver complexes, 14, with the same iminopyridine ligand (n-OIP) and with
3 Metallomesogens with Monodentate Ligands
905
14 Silver iminopyridine and carboxylatepyridine complexes
the pyridine carboxylate ligand (n-OCP) showed mesogenic behavior: Y =BF4, CF,, NO,, PF,. Only the salts (Y = BF,, n = 2) are nematic: all the others show SmA and SmC mesophases. The ester salts show wider mesophase range and greater smectic polymorphism than the imino salts [24].
sessed a negative diamagnetic anisotropy and can be aligned in a magnetic field [26]. This work with polymeric acetylides was extended to the study of related low molar mass systems and the platinum(I1) complexes, 16, were found to exhibit thermotropic N and SmA phases [27].
iMe3
16 Acetylide platinum complexes
3.4 Acetylide Ligands The very unusual organometallic liquid crystals, 15, [M, M'=Pd, Pt or Ni; M=M' or M #M'] were made by Takahashi et al. in 1978 [25], by a copper chloride-triethylamine coupling of the appropriate metal halide with an alkyne. These materials do not show thermotropic properties, however they are lyotropic nematic in trichloroethylene solution. PBu3
PBu3
I
I PBu,
I
PBu~
17
PBu,
I I
I [(-c+~-M-c~c-c~c-M'-c~c-c~c-M -c r c-)
Bruce et al. [28] looked at the effect of changing the acetylide ligand to make it longer and more structurally anisotropic, 17. They found that platinum complexes show enantiotropic nematic phase, and that increasing the size of the phosphine decreases the temperatures of transition.
Polymeric acetylide complexes
"1
PBu3
15 Takahashi's acetylide complexes
The physical properties of these 1,4-butadiyne metal polymers were studied using both homometallic polymers (M = M') and heterometallic polymers (M = Pt, M'= Pd or Ni), which showed that all of them pos-
Changing the direction of the ester groups strongly influences the thermal stability of the mesophases of the complexes [29]. The mono substituted complexes were not mesomorphic.
906
XIV
Metal-containing Liquid Crystals
3.5 Isonitrile Ligands Gold-isonitrile complexes, 18, as a new family of metallomesogens, are prepared and even nonmesomorphic isonitrile ligands have been found to form N and SmA liquid crystals when coordinated to an Au(1)Cl moiety [30a].
18 Gold-isonitrile complexes
The same group [30 b] prepared platinum and palladium complexes. The coordination of the isonitrile to a metal increases the thermal stability of the ligand. Both Pt and Pd complexes show mesomorphic properties. The liquid crystalline complex [ P ~ I * ( C N C ~ H ~ C ~ H ~ O ~has C H , been ~)*] crystallographically characterized. Compounds where the gold atom links an alkynyl and an isonitrile, 19, give rise consistently to SmA mesophases [31].
oxylates, and in 1938 Lawrence [32] noted that copper stearate melted to a plastic fluid before clearing.
4.1.1 Alkali and Alkaline Earth Carboxylates Structural data, indicating the existence of columnar and even disc-like mesophases for the anhydrous alkali, alkaline earth, and cadmium salts of long-chain fatty acids were first reported by Skoulios et al. [9]. For example the potassium salts can be lamellar, ribbon-like or disc-like in a three dimensional lattice [33], see Fig. 1.
4.1.2 Lead, Thallium, and Mercury Carboxylates SmC phases have been reported for the lead(I1) carboxylates, [Pb(C,H2,z+ for n =7,9,11,15, 17 [34]. The compounds were investigated further by various methods, particularly X-ray diffraction, and were found to transform from a lamellar crystal-
X =H,OCnHzn+l 19 Gold-isonitrile- alkynyl complexes
4 Metallomesogens with Bidentate Ligands 4.1
Carboxylate Ligands
Metal soaps have long been known to have mesogenic properties. Thus, Vorlander [8 a] in 1910 reported the existence of lamellar phases in the anhydrous salts of alkali carb-
Figure 1. Structure of potassium alkanoate.
4
Metallomesogens with Bidentate Ligands
907
line to another lamellar highly viscous phase and then to melt to a SmC phase [35]. Later, Attard et al. [36] describes the synthesis, thermal properties, and response to UV radiation of the lead(I1) salts of alkadiyonic acids
phase bearing only four alkyl chains. All the compounds [Cu2(p-02CC,H2,+ for n=4-24, show the same structure, see Fig. 2. The X-ray shows that the complexes crystallize in a lamellar structure in the [(CH~(CH2),-C-C-C~C-(CH2)~COo)~Pb].solid state and above 120°C exist as a coThese compounds were found to exhibit lumnar two dimensional hexagonal strucmesomorphism over a wide temperature ture in which polar groups are surrounded range. They also appear to form glassy meby disordered alkyl chains [40]. sophases at lower temperature. All the maThe phase transition in these materials terials could be polymerized by low intenhas been investigated by a variety of techsity UV radiation. niques: X-ray single crystal diffraction The mercury carboxylates shows them to be dinuclear with each cop[Hg(C,H2,+,COO),] for n = 7 - 17, are not per surrounded by five oxygen atoms, one mesogenic, while thallium carboxylates belonging to the neighboring dimer [41]. [T1(C,H2,+1C00)2] gave stable lamellar Cu-K EXAFS (extended X-ray absorpphases with branched alkyl chains [37]. tion fine structure) confirmed this and also showed that the dinuclear arrangement of the core was preserved on passing from the crystal into the mesophase [42]. 4.1.3 Dinuclear Copper Magnetic susceptibility measurements Carboxylates show a small, sharp decrease in the magnetic moment at the transition to the mesophase, implying that there was some change The liquid crystalline phase of dinuclear which was probably attributable to a modcopper stearate has been known since 1938, ification of the relative disposition of the and was reinvestigated in 1964 [38]; but it dimers within the columns [43]. was not until 1984 [39] that this mesophase Isotopic labeling has facilitated band aswas definitely characterized by Giroudsignment in the infrared spectra [44a] and Godquin et al. as columnar discotic by Xdynamic behavior of the alkyl chains studray diffraction. This was the first example of a thermotropic hexagonal discotic mesoied by quasielastic neutron scattering [44 b].
Figure 2. Structure of copper alkanoate in the discotic hexagonal mesophase.
908
XIV
Metal-containing Liquid Crystals
Copper laurate can be processed into oriented fibers by melt spinning of its thermotropic columnar mesophase, X-ray diffraction and electron microscopy revealed a high degree of orientation of the spun fibers both in the crystalline and in the columnar states [45 a]. Neutron experiments have been performed on these fibers to study the motion of the hydrogen atoms of the chain [45 b] . The first example of a fully characterized metallomesogen giving an authentic columnar mesophase near room temperature in a pure state, was described by Maldivi and coworkers. These mesogens are obtained by simple insertion of carbon-carbon double bonds in the carboxylate ligands [46]. In contrast to the case of straight chain salts, complexation of the copper(I1) branched chain carboxylate with a difunctional ligand as pyrazine or 4,4'-dipyridyl result in a liquid crystalline material whose mesophase structure has not been established [47].
4.1.4 Dinuclear Rhodium, Ruthenium, and Molybdenum Carboxylates The rhodium carboxylates [Rh,(p-02CC,H2, for n = 4 - 24 behave very similarly to the copper complexes and form, on heating, columnar mesophases. Rh-K-edge EXAFS measurements showed no change in the core structure passing from the solid state to the mesophase [48]. The complexes Rh, [(p-O2CC6H,-4 OC,H,,+ J4] for n = 8 - 14, were investigated by optical microscopy, DSC and X-ray diffraction. The complexes display a columnar mesophase of rectangular symmetry. The eight chain complex shows two columnar phases, one rectangular at room temperature and a hexagonal one above 138 "C [49]. +
Mixtures of rhodium eicosanoate or copper dodecanoate with solvents such as toluene, decahydronaphthalene, and (+) camphene have been studied. The mesophases found at high temperature turns from columnar to nematic when the weight fraction of the solvent is increased beyond a value of about 50% [50]. A wide series of diruthenium (11, 11) tetracarboxylates of general formula [Ru~(~-O,CC,H,~+ has been made by the Grenoble group; the carboxylate substituents include linear alkyl chains (n=4- 19), unsaturated chains, and fluorinated chains [5 13. The mesomorphic behavior of these compounds was investigated by DSC, optical microscopy and X-ray diffraction. A thermotropic columnar mesophase was observed for all complexes except the perfluoro-compounds. Unsaturation in the peripheral chains depresses transition near room temperature. The magnetic susceptibility was measured. A sharp change in the magnetic moment at the solid-mesophase transition is ascribed to a core distortion similar to Cu(I1) and Rh(I1) complexes. Important work was carried out by Cukiernik et al. of the Grenoble group on mixed valence ruthenium carboxylates of general formula [Ru2(0,CC,H2,+ Xo, with X=anion [52]. The anion has a strong influence on the mesomorphic behavior: the chloro-complexes (X = ClO) are not mesomorphic. With a carboxylate (X = R'COOO) or an alkylsulfate (X=C,,H,,OSO~) as the anion, the compounds show a hexagonal columnar mesophase on heating. To favor the interdimer interactions an adduct pyrazine has been intercalated between two dimers. The compound, 20, is supposed to have a polymeric structure with the pyrazine axially coordinated to the two dimers [53].
4 Metallomesogens with Bidentate Ligands
909
4.2.1 Copper Complexes with Two or Four Peripheral Chains
R
R
d
R
20 Diruthenium tetracarboxylate complexes with an adduct pyrazine
The quadruply bonded molybdenum carboxylates [Mo2(O2C,H2,, present a similar behavior to that observed for M=Cu, Rh, Ru. When a side chain is branched on the complex the transition temperature is lowered to about 37 “C. Substitution of fluorine increases the temperatures but an enantiotropic hexagonal discotic mesophase is still found [54].
The mesogenic copper (11) pdiketonate, 22, first described by Giroud-Godquin et al. [56] in 1981 has a square planar geometry about the metal and a d9 configuration (one unpaired electron). Both the symmetrically and the unsymmetrically substituted complexes (m=n or m f n ) exhibit very organized discotic mesophase. Later work and particularly NMR investigation and X-ray diffraction suggested that the discotic phase was, in fact, a disordered lamellar crystal
WI. Ohta et al. [58] reported later the related complexes (R=R’=C,H,,, M=Cu; R=R’=OC,H,,, M=Cu) comiscible with Giroud-Godquin’s compound and therefore of the same type.
4.2 /?-Diketonate Ligands P-Diketonates form flat bis-ligand complexes with a variety of metals (including Ni, Pd and Cu) that can take up square planar geometries. When appropriate substituents are present, these complexes can give rise to nematic, discotic and occasionally to lamellar phases. The first P-diketonate, 21, examined for mesophase behavior was the palladium complex of Bulkin et al. [55]. They suggested that it may be mesomorphic but have been unable to confirm this. ,GH17
CBHI~~
21 Bulkin’s pdiketonate complexes
The structures of the single crystal of two complexes of this series (M = Cu, Pd) have been established. According to X-ray diffraction data, the copper atom has a square planar configuration and the packing of the molecules in the crystal is of the layered-columnar type [59]. An interesting development in this field was made by Chandrasekhar et al. [60] in
9 10
XIV
Metal-containing Liquid Crystals
1986 with the synthesis of copper complexes, 23, which were monotropic nematic and paramagnetic. These complexes incorporate features of both rod-like and disc-like molecules and are claimed to be the first example of biaxial nematic metallomesogen [61]. Unfortunately no other workers have been able to reproduce this observation of biaxiality. The effect of different terminal substituents (R=C,H2,+1 or OC,H,,+ 1; R'= CN, OCH,, OC2H5,C2H5,C3H7, C4H9; M = Cu or Pd) on the mesomorphic properties were studied later: mesophases of the SmC, SmA and N types were observed [62].
this complex curiously shows a dimer discotic rectangular ordered columnar mesophase. A more systematic investigation of such structures was performed by Thomson et al. [65] who found that complexes, 25, with the rod-like structure A showed monotropic nematic mesophase of discotic type while those with the more disc-like molecule B show enantiotropic phases of calamitic tYPe.
0,
R
,o
F'
25 Thomson et al.'s Pdiketonate complexes A Ring X = phenyl; R = C,,H,,, R' = CH,, H Ring X = cyclohexyl; R' = CH,, C,H, 23 Chandrasekhar's Pdiketonate complexes CIOH21
Muhlberger and Haase [63] synthesized the saturated derivatives, 24, which also show a monotropic nematic phase. X-ray studies showed that the diketonate ring was planar and that identical substituents were trans to each other.
0,
o'c".o
.o
CnH2n+1
3 FcmH2m CmHzm+l
0.
*o
B Ring Y = phenyl; R" = C,,H,,CH,, Ring Y = cyclohexyl; R" = OCH,
C,H,, F
o'c".o
4.2.2 Copper Complexes with Eight Peripheral Chains
CnH2n+1
24 Miihlberger and Haase's Pdiketonate complexes
Ohta [64] synthesized the biphenyl equivalent ( n= 16, m = 1, M = Cu) and found that
Accounting for the empirical fact that at least six substituents are needed to get columnar mesophases, Giroud-Godquin et al. [66] reported synthesis and properties of
91 1
4 Metallomesogens with Bidentate Ligands
new copper diketonate complexes, 26, substituted on each phenyl, in 3,4 positions, by two alkoxy chains giving a resultant complex with eight chains. These complexes were shown to have a hexagonal columnar mesophase, D,, by miscibility studies and X-ray diffraction. The corresponding nickel complex is not mesogenic neither is the complex in which the phenyl is substituted in the 3,5 positions. Similar complexes were studied by Ohta [67] who found discotic mesophases for alkoxy (R=OC8H,7) chains and no mesophase for alkyl tails (R = C,H,,).
arrangement in the crystalline and liquid crystalline phases are discussed based on magnetic data [70].
27 Malondialdehyde complexes
4.2.4 Dicopper Complexes Dicopper complexes, 28, have been synthesized and characterized by Lai et al. [71]. These complexes melt to give birefringent fluid phases with columnar superstructures. ?R
OR
I
OR
26
Giroud-Godquin's eight chained Pdiketonate complexes
OR
RO RO
4.2.3 Malondialdehyde Complexes Other interesting compounds are copper and nickel complexes of (p-alkoxypheny1)-malondialdehydes, 27. The synthesis, characterization, thermal behavior, X-ray diffraction studies and magnetic properties of a number of these complexes were performed by Haase et al. [68]. It was found that nickel complexes show liquid crystalline SmA phases over a broad temperature range with low melting points (85"C), while copper complexes show nonidentified mesomorphic behavior. This behavior is different from that observed for n = 5 copper complex reported by Blake et al. [69] who observed a nematic phase. Models for the molecular
OR
28 Dicopper Pdiketonate complexes
4.2.5 Polymers Polymers containing Pdiketonate metal complexes of copper(II), nickel(II), and cobalt(II1) were synthesized and examined by DSC, polarizing microscopy and X-ray diffraction. The copper complex exhibits a SmA phase while both the nickel and the cobalt complexes do not work as mesogens ~721.
4.2.6 Other Metal Complexes Synthesis, mesomorphic properties and crystal structures of copper(II), palladi-
912
XIV
Metal-containing Liquid Crystals
um(II), and nickel(I1) complexes have been performed. Despite the unambiguous structural isomorphism only the palladium and the copper complexes are mesogenic. The chains are fully extended in an all trans conformation [73]. The first octahedral metallomesogen was an iron(II1) tris-diketonate, 29, which showed a birefringent structure and endothermic transitions but the mesophase has not been defined [74].
Iron(III), manganese(II1) and chromium(III), 31, have been also investigated and the bargraph shown in Fig. 3 illustrates the phase behavior of these octahedral metallomesogens [77]. C12H250
RI
4
P
M = Fe, Mn,Cr
31 Iron, manganese and chromium
R
tris-pdiketonate complexes
29 Iron tris-pdiketonate complexes
..as been shown that a Pdiketonate liquid crystal containing palladium(I1) and oxovanadium(IV), analogous to known copper complexes (which display discotic lamellar and columnar mesophases), has been prepared and characterized; palladium complexes are mesomorphic but their identification is difficult due to their decomposition in the isotropic phase; the complex n = 10 seems to present a N, phase on the basis of optical microscopy. The vanadium complexes are enantiotropic discotic [75], while the vanadium complex, 30, presented by Styring et al. [76] (n= lo), exhibited a short monotropic columnar discotic phase. 0
I1
The effect of the number of side chains on the mesophase stability has been investigated by the same authors. The results indicate that complexes with four side chains are not mesomorphic (in disagreement with-
160T
140/ 120
-sz 2
100 80
L
Lu
F
60 40
20 0 Fe N=6
30 Vanadium Pdiketonate complexes
Fe Mn Cr N=12
Fe Mn Cr hk15
Fe Mn Cr k l 8
Figure 3. Bar graph showing the phase behavior of octahedral metallomesogens.
4 Metallomesogens with Bidentate Ligands
in Styring's work), while those with eight chains can display limited mesomorphism and complexes with ten side chains exhibit a discotic hexagonal disordered mesophase over a wide range of temperature [78]. Several diketonate thallium(1) complexes with a different number of alkoxy groups on the aromatic ring have been synthesized [79]. The mesogenic behavior of the thallium(1) and of the copper(I1) derivatives are qualitatively similar. The authors propose a dimer structure.
913
4.4 Sulfur Containing Ligands 4.4.1 Dithiolene
The first systematic study of d-block metallomesogens were carried out by Giroud and Mueller-Westerhoff [ 1 11 and marked, in 1977, the beginning of interest in metal-containing systems. They were dithiolene complexes of nickel, palladium and platinum with two p-alkyl phenyl substituents (see structure 3). Nickel and platinum complexes showed nematic and smectic phases, 4.3 Glycoximate Ligands depending on the chain length. Palladium analogs do not possess any mesomorphic Ohta et al. [80] reported the discotic mesoproperties [81]. morphism, thermochromism and solvatoAnalogous di- and tetrasubstituted nickchromism of bis[ 1,2-bis(3,4-dialkoxyphenyl) el dithiolenes prepared by Strzelecka et al. ethanedione dioximato]metal complexes, [82] were originally described as discotic 32 (abbreviated to [M { (C,O),dpg} 2]), and mesogens. Later, X-ray data showed that of bis[ 1,2-bis(3,4-dialkylphenyl)ethanedithese compounds were not mesomorphic one dioximato] metal complexes (abbrevi[83]. However, the related nickel dithioated to [M { (C,),d~g}~] with M = Pd, Ni and lenes, 33, with four 3,4-bis(dodecyloxy)Pt. phenyl substituents was shown by X-ray diffraction to give rise to a hexagonal disorRO OR dered columnar mesophase D,, between 84 and 112 "C for R = CI2H2,by Ohta et al. [841. O
RO
OR
32 Glyoxamate platinum complexes
All the complexes showed a discotic hexagonal disordered columnar mesophase (Dhd).The mesophases were identified by X-ray diffraction. Platinum complexes exhibit a very clear thermochromism: the color of the liquid crystalline discotic mesophase turns from green to red with increasing temperature from room temperature.
33 Dithiolene complexes with four chains
4.4.2 Dithiobenzoate Most of the work on dithiocarboxylates have been centered on metal complexes of 4-alkoxydithiobenzoic acids (abbreviated to n-OdtbH) 34.
914
XIV
Metal-containing Liquid Crystals
Lateral fluorination on these compounds promotes the SmC phase [89]. 34
4-alkoxydithiobenzoic acid
4.2.2.2
Many complexes [M(n-odtb)2] have been synthesized for M = Ni, Pd, Zn, Hg, Ag, and some planar gold(II1) complexes showing SmA phases [AuX2(n-odtb)2] with X = C1, Br, Me [85]. 4.4.2.1
Nickel and Palladium
Studies by the Sheffield group on the bis(4-alkyloxy-dithiobenzoato)complexes,35, showed that the Ni(I1) and the Pd(I1) complexes were mesogenic. Palladium complexes were shown by X-ray studies to be composed of monomeric species and to have a square planar geometry about the metal. These complexes show a SmC phase for long chains and a N phase for shorter chains [85]. Complexes with branched chains have a melting point lower than those with straight chains (861.
Zinc
X-ray diffraction studies on alkoxy-dithiobenzoato zinc complexes [Zn(n-odtb),] showed that in the solid state the material is a dimer containing an eight-membered Zn- S -C-S -Zn-S -C-S ring; the dimer held together in the nematic phase and disappeared in solution where is found a monomeric structure [85]. This work is confirmed by EXAFS studies [90]. Later, 4-substituted-piperazine- 1-dithiocarboxylate complexes, 37, have been synthesized by Hoshino-Miyajima [91] with different metals (M=Ni, Pd, Cu, Zn) and different chain lengths. They all show N and SmC mesophases at high temperature.
37 Piperazine dithiobenzoate complexes M = Pd, Ni, Cu, Zn
Y = Phenyl, C6H40C,H2,,+,,CnH2,,+, 35 Alkoxydithiobenzoate complexes
The palladium complexes were further studied both in the solid state and in the mesophase using EXAFS technique [87]. The data confirmed the known crystal structure. The nickel complexes have the same mesogenic properties than the related palladium. The major difference is that at high temperature, the blue bis-(dithiobenzoates) rearranged to form the red trithiobenzoate complexes, 36, which are also mesogenic [88].
36 Trithiobenzoate complexes
4.4.2.3
Silver
Primary silver thiolate compounds [AgSC,H,,,,] (n=4,6, 8, 10, 12, 16, 18), behave as thermotropic liquid crystals. On heating they display successively lamellar (SmA), cubic, and micellar mesophases. X-ray studies show that the micellar phase is an hexagonal columnar mesophase [92].
4.4.3
Other Dithio Ligands
Other dithio systems are the alkyldithioacetate and alkylxanthates of nickel described by Ohta [93]. Dimeric tetrakis (nalkyldithiocarboxylato) dinickel complexes (R=C,H,, to C,,H,,), 38, show mono-
4 Metallomesogens with Bidentate Ligands
91.5
tropic lamellar mesophases, while the related monomeric xanthate [Ni(ROCS,),] (R =CH, to C,2H25),39, exhibit complex double- and triple-melting behavior.
41 Ovchinnikov’s salicylaldimate complexes
38 Dimeric dithiocarboxylate complexes
39 Alkylxanthate complexes
4.5 N - 0 Donor Sets: Salicylaldimine Ligands Schiff-bases derived from substituted salicylaldehydes, 40, are very versatile ligands which form ( N - 0)chelates with many metals: copper, nickel, palladium, vanadium and iron.
40
Salicylaldehyde
4.5.1 Copper, Nickel, and Palladium Most of the paramagnetic salicylaldimine copper(I1) complexes, first reported by Ovchinnikov et al. in 1984, showed smectic mesophases [94]. However, nematic behavior was observed by Galyametdinov et al. for 41 [9.5]. They were the first examples of materials with a paramagnetic nematic phase.
All the complexes show both smectic and nematic phases. This was confirmed by Xray diffraction studies. Following this work, the study of mesomorphic salicylaldimate complexes became very popular [96]. Marcos et al. [97] and Hoshino et al. [98] examined at the same time several series of copper and nickel complexes, 42. Both Cu(I1) and Ni(I1) complexes derived from the n-alkylamine (R = C,H,, show mainly nematic mesophase, while those derived from arylamine (R = 4-C6H,0C,H,, + ,) show largely smectic mesophases. +
42 Marcos and Hoshino’s salicylaldimate complexes
Then, Sirigu [99] and later Courtieu [ 1001 synthesized a series of copper, nickel and palladium complexes, 43, with high-temperature smectic and nematic behavior.
43 Sirigu and Courtieu’s salicylaldimate complexes
916
XIV
Metal-containing Liquid Crystals
Courtieu [ l o l l lowered the melting and the clearing temperatures by halogenation of the complexes (X = C1, Br), 44.
44
Halogenated salicylaldimate complexes
Sadashiva and coworkers [ 1021 investigated a series of salicylaldimate complexes of nickel, copper and palladium with R = ROPH-CH20, 45, showing SmA and SmC phases.
also been made by Marcos et al. [ 105, 1061 and Ghedini et al. [107].
47 u
45
Marcos' chiral salicylaldimate complexes M = Cu, Pd, VO Z = -CH=CH-COO
Sadashiva's salicylaldimate complexes
The copper and the vanadium complexes have been studied by ESR [ 1031. Polar Schiff-base complexes, 46, have been synthesized to study the relation between mesogenic behavior and the molecular structure. This family showed SmA and N phases [104]. The lateral polar groups X, Y, Z can be F, CF,, CN, CH,CN.
46
Polar salicylaldimate complexes
Chiral ferroelectric complexes, 47 and 48, of palladium, copper and vanadium have
48
Ghedini's chiral salicylaldimate complexes (RCH,)* = (-)-myrtalyl = (-)-menthy1 = S(-)-p-citronellyl = R(-)-2-0~tyl
Some of these materials display the potentially chiral ferroelectric SmC phase in addition to a SmA or cholesteric mesophase. Pyzuk [lOS] found for copper and nickel complexes some blue phase or novel type of amorphous phase between a tightly twisted chiral nematic phase and the isotropic liquid. These ferroelectric metallomesogens are interesting as they can be aligned par-
4
allel or perpendicular to a magnetic field, depending on the anisotropy of the magnetic susceptibility of the complex. Other N-substituted salicylaldimines have been used as complexing ligands. All copper, nickel and vanadium complexes studied, 49, show N andlor SmC phases depending on the chain length [ 1091, and characterized by X-ray diffraction studies [ 1101.
Metallomesogens with Bidentate Ligands
917
formed SmA phases. A p-0x0-bridged complex of Fe(III), 50, has also been synthesized by the same authors and the magnetic behavior investigated. It is the first complex of Fe(II1) exhibiting a nematic mesophase [114].
50 Galyametdinov's iron complexes
49
Marcos and Hoshino's N-substituted salicylaldimate complexes
4.5.2 Platinum, Vanadyl, and Iron Of the complexes, 43 (M=Zn, Co, Ni, Cu, VO, Pd, Pt), only the copper, the vanadyl, the platinum and the palladium complexes showed smectic mesophases. Both platinum(I1) and oxovanadium(1V) complexes containing only a chain on one biphenyl moiety (R'=H) exhibit nematic phase [lll-1131.
Note that many magnetic studies have been performed as well as ESR and X-ray diffraction studies [115, 1161 on all the paramagnetic complexes.
4.5.3 Rhodium and Iridium Mesomorphic Schiff-bases can also complex two square planar cis-[M(CO,)]moities with M = rhodium or iridium [ 1171. The complexes, 51, show SmA behavior.
n
1.
,co
*m
51 Rhodium and iridium salicylaldimate-carbonyl complexes
43 Sirigu and Courtieu's salicylaldimate complexes
R = CnHZn+l R' = Ph-CmH2,+1
R = OCnH2n+l R' = CmH*m+,
The mesogenic iron(II1) complex reported by Galyametdinov is paramagnetic and
4.5.4 Polymeric Liquid Crystals Based on Salycylaldimine Ligands Carfagna [ 1181 reported the first polymeric metallomesogen with a Schiff-base ligand. The polymer showed a monotropic phase presumably smectic. Later, Caruso et al. [ 1191 showed that a copper complex could give a polymer with nematic behavior. Dif-
918
XIV
Metal-containing Liquid Crystals
ferent series of homopolymers containing paramagnetic units ofcopper(I1) salicylaidiminates have been synthesized and studied by ESR [ 1201. Later, magnetic susceptibilities of copper and iron containing metallopolymers have been investigated [121]. There were two reports of the same bis(a1koxylsalicylidene) ethylenediamine complexes of Ni(I1) and Cu(II), 52.The material was reported as giving a SmA phase at elevated temperature [ 1221 and a nonidentified mesophase [123]. Ohta later [124] found that the nickel complexes showed two smectic liquid crystalline phases, SmE and SmA. The SmE-SmA phase transition in the nickel complex is accompanied by a reversible stretching of the interlayer distances. This phenomenon can be ascribed to a dimer SmE - monomer SmA phase transition.
52 Bis(alkoxysa1icylidene)ethylenediamine complexes
4.6 Cyclometalated Complexes
?
53 Ghedini's ortho-palladated complexes
R = EtO or Et; R' = CnHZn+,
of the bridging halogens was also investigated: the chloride showed a nematic phase while the bromide and the iodide showed enantiotropic SmA and N phases. X-ray diffraction [ 1261, quasielastic neutron scattering [ 1271,and electron energy loss spectroscopy [128], have been performed on these compounds. Chiral dinuclear cyclopalladated complexes have also been synthesized [ 1291. SmC* mesophases have been found. NMR and circular dichroism have been performed on these chiral molecules. The investigation of the viscoelastic properties has been studied and compared with the uncomplexed ligand [ 1301. 54 The monomeric complexes, (L =PPh,), obtained by cleaving 53 with PPh, are not mesogenic while the amine adducts (L = pyridine or quinoline) form smectic and nematic mesophases.
4.6.1 Azobenzene Up to now, most of cyclometalated mesogens are palladium-containing. Only one mercury complex is known. Ghedini et al. [ 1251 in Calabria prepared the first cyclopalladated liquid crystal, 53. These dinuclear chlorobridged complexes formed nematic phases at about 200 "C at much higher temperature and in a narrower range than the free ligand. The role
I
R
54 Mononuclear palladated complexes
4 Metallomesogens with Bidentate Ligands
Mononuclear cyclopalladated liquid crystalline materials whose molecular structure consists of two different thermotropic ligands, connected by a palladium atom, 55, have been also prepared. The mesomorphic properties depend on the length of the chain. Monotropic or enantiotropic nematic and/or smectic phases can be found [131]. Other orthometallated azobenzenes were reported by Hoshino [132]. The complexes presented the same nematic phase as the ligand, at higher temperature.
919
melting and especially the clearing temperatures of the complexes are very significantly higher than those of the free ligand. The substitution of the alkoxy substituents by polar groups (Cl, CN, NO,) modified the behavior only slightly [135]. By substitution of the halogeno bridges by chiral chains they prepared compounds giving cholesteric mesophases at about 145- 150 "C. Recently, Espinet and coworkers reported a comparative study of metal-containing liquid crystals for which the ferroelectric behavior has been fully characterized [ 1361.
R'
X = C1, Br, SCN 55
Azpac compounds
Azoxymercury complexes, 56, have also been synthesized [ 1331. The ligand bears in 4 and 4' position, an alkoxy chain and the chiral S(+)-2-octyloxy, respectively. Both the ligand and their HgCl derivatives are liquid crystals. The mercury complexes show an enantiotropic SmC* mesophase.
'c1 56
4.6.2
Azoxymercury complexes
Arylimines
Related complexes based on imines were reported by Espinet [134]. The complex 57 showed smectic phases (usually SmA) when X=C1, Br and SCN. The acetato complexes were generally nonmesogenic. It was suggested that this was due to the nonplanarity of the molecules. Again for this series, the
57 Espinet's orthopalladated imine complexes
Baena et al. [137] introduced the idea that perturbing the high symmetry of the molecular shape of this family of compounds might lead to lower melting points and less ordered mesophases. Azomethines were orthopalladated to give the binuclear complex and these were treated with Tl(acac) (acac = acetylacetonate) to give a mononuclear species. These mononuclear complexes gave rise to mesophases below 90°C. N, SmA and SmC phases were found. New compounds containing two or even four palladium atoms and eight or twelve flexible side chains have been synthesized by Praefcke et al. [ 1381. The compound, 58, shows an enantiotropic discotic hexagonal columnar mesophase over a wide range of temperature and a lyotropic nematic phase in various organic solvents. X-ray diffraction studies reveal an oblique arrangement of disordered columns.
920
XIV
Metal-containing Liquid Crystals
RO
OR
RO
OR
OR
RO X = OAc, C1, Br, I, SCN, N,
0
1,4-Phenylene 4,4'-Stylbenylene
structure due to its six alkoxy groups [ 1401. Compared to the analogous palladomesogens, the mesophases observed are of the same type, nematic-discotic and hexagonal columnar, but are more stable in the case of the platinum liquid crystals. X-ray diffraction has been performed on the dinuclear complex [ 1411. Manganese(1) and rhenium(1) complexes, 60, have been synthesized by Bruce et al. [142]. They are the first examples of calamitic thermotropic liquid crystals containing a metal with a simple octahedral coordination environment. Manganese and rhenium complexes show nematic phases between about 120 and 190°C.
M=Pd,Pt 58
Tetranuclear metal complexes
The dinuclearpallado compounds, 59, are the first (monotropic) cases of metallomesogens exhibiting the nematic discotic phase. On doping with a strong electron acceptor, the stabilization and/or induction of mesophases was observed [ 1391.
R = alkyl or CnHzn+10 M = Mn, Re
60 Bruce's manganese and rhenium complexes
?R
4.6.3 Orthopalladated Diarylazines
o
%:@OR
RO RO
59 Dinuclear pallado complexes
Dinuclear platinum and tetraplatinum compounds have also been studied, 58 (M = Pt). Synthesis and lyotropic-crystalline behavior (nematic), are discussed for this platinum compound, disc-shape in
These studies were then extended to the synthesis of palladium complexes of symmetric azines. In contrast to the above-mentioned compounds the di-p-acetato complexes, 61, showed SmC, and in two cases, also nematic, mesophases. The p-acetato ligands constrain the complexes to be nonplanar, and a novel type of structure based upon open-book-shaped molecules [ 1431 has been proposed. With R' optically active, a mixture of cis- and trans-isomers was observed.
4
92 1
Metallomesogens with Bidentate Ligands
dine ligands, 63. When X - Y = acac (n= 0) a material with a monotropic SmA phase resulted. When X - Y = 2 - 2’ bipyridine ( n= 1 and the counterion = BF,) a material with an enantiotropic nematic phase is produced.
L
J
63 Mononuclear 2-phenylpyrimidine complexes
61 Open-book shape structure of the chiral trans isomer of the p-carboxylato palladium azines
4.6.4 Aroylhydrazine A series of aroylhydrazinato-nickel and -copper complexes, 62,was synthesized in high yield and shown to form metallomesogens with SmC and N phases around 150 “C. The nickel complexes were found to be highly stable while the copper complexes decompose soon after entering the isotropic phase [ 1441.
Further studies [ 1461 of the same system looked at the dimeric precursors to such monomeric species. The molecular structure consists of two ligands in a transoid geometry with respect to the Pd-Pd axis, connected by the Pd,X, bridges, 64. On the basis of X-ray diffraction, the dinuclear derivatives show on heating, first a solid phase, then a SmA phase. R
YR’
I
I
I
OR’ R 64 Dinuclear 2-phenylpyrimidine complexes
4.7 Metallocen Ligands 62 Aroylhydrazine complexes
4.6.5 Orthopalladated Pyrimidine Complexes Other work in this area has been done by Ghedini and coworkers [ 1451 with pyrimi-
The largest group of materials in this family are derived from the ferrocene unit [ 1471.
4.7.1 Ferrocene 4.7.1.1 Monosubstituted Ferrocene The first examples of mesogenic ferrocenes were reported by MalthCte and Billard in
922
XIV
Metal-containing Liquid Crystals
1976 [lo] and showed SmC mesophases. Another ferrocenyl Schiff-base derivative, 65, described by Galyametdinov gave rise to enantiotropic nematic phases [ 1821.
65 Monosubstituted ferrocene
Interestingly these compounds can complex a copper anion to lead to trimetallic compounds, 66, with a high temperature nematic phase. In these series the ferrocene moiety is at the end of the molecule rather than being incorporated into the core.
F FP
66
Copper complexes of 1,l’ monosubstituted ferrocene
4.7.1.2
The crystal structure reveals that these molecules adopt the extended S geometry, with the two alkyl chains extending in the opposite direction. Those with X = Y = p C,H40R were shown to give monotropic SmA or SmC phases. With an additional benzene ring link by an ester, the mesomorphic properties depend of the direction of the ester: with X = Y =p-OCOC,H,OR, none of the ferrocenes showed liquid crystal behavior on heating; with X = Y = p COOC,H,OR (R=C,H2,+1) ( n = 2 - IS), the derivatives with n = 2-4 showed monotropic nematic phases, for n = 6 N and SmA phases. Increasing the alkyl chain length caused the disappearance of the nematic mesophase [ 1491.The unsymmetrical derivatives X Z Y led to compounds giving SmA and SmC mesophases [ 1501. Replacing the ester group of Y by a Schiff-base X = Y = p-CH = N C,H,OC,H,,+I led to nematic mesophase for short chains and smectic for longer chain lengths [ 1511.
4.7.1.3 1,3 Disubstituted Ferrocene
1,l’ Disubstituted Ferrocene
In 1988 Bhatt et al. 11481 reported a series of diesters in which the ferrocene moiety is
Compounds of 1,3 disubstituted ferrocene have been reported by Deschenaux et al., 68. The symmetrical compound [152] led to SmA and N phases, the unsymmetrical
Fe
& 68
at the center of the molecule, 67.
0
67
I, 1’-ferrocene diesters
1,3 disubstituted ferrocene
[153] to SmC and SmA as well as N phases. They showed the strong influence of structural isomerism on the liquid crystal behavior by comparison between 1,l’ and 1,3 disubstituted ferrocenes. X-ray diffraction studies on the SmA phase suggested a monolayer molecular organization [ 1541.
5 Metallomesogens with Polydentate Ligands
923
Synthesis and mesomorphic properties of the first ferrocene-containing side-chain liquid crystalline polymers, 69, obtained by grafting either a 1,l'- or a 1,3-disubstituted ferrocene derivative functionalized by a vinyl group onto a polysiloxane have been reported [ 1551. These polymers showed enantiotropic SmC and/or SmA mesophases.
69 Ferrocene containing liquid crystalline polymers
4.7.2 Ruthenocene Synthesis and mesomorphic behavior of ruthenocene substituted in the 1,l' positions have been reported by Deschenaux et al. [ 1561. With long substituted alcoxy chains ( n = 11 - 14), compounds give rise to enantiotropic SmA phases. Comparison of the mesogenic properties with those of the analogous ferrocene shows that both series, on heating, give SmA mesophases. The ruthenocene derivatives melt at higher temperature and in a shorter anisotropic domain than the corresponding ferrocene.
4.7.3 Iron Tricarbonyl Derivatives Another group were the butadiene iron tricarbonyl complexes, 70, reported by Zimin-
Fe
70
Iron tricarbonyl complexes
sky and MalthCte [157]. All the terminallysubstituted complexes showed nematic phases over a wide range. The disubstituted complexes showed a nematic phase for short chains and a SmA phase for longer chains.
5 Metallomesogens with Polydentate Ligands 5.1 Phthalocyanine Ligands 5.1.1
Copper Complexes
Most work on metallophthalocyanines has been carried by Simon and coworkers [ 1581 who synthesized in 1982 the first octasubstituted phthalocyanines and their copper complexes, 71. These compounds were reported to show a stable mesophase over a very wide range of temperature and which was characterized by X-ray diffraction as discotic columnar mesophases.
924
XIV
Metal-containing Liquid Crystals
@l")
R + N $ - R
R
,
R
71
\
complexes have also been reported [161]. The fact that these compounds are brightly colored has been considered for laser addressed devices.
/N.
R
5.1.3 Lutetium Complexes
1
R
Octasubstituted metallophthalocyanines
Copper phthalocyanines bearing four henzo(l5)crown-5 substituents, 72, have been made: they stack to give channels through which ions may be transported [159].
Simon et al. 11621 found that bis(phtha1ocyanines) lutetium, 73, formed on heating discotic mesophases which were molecular semiconductors. X-ray data for the mesophase indicated an intercolumnar distance twice the thickness of the phthalocyanine ring.
J $ 3 \ 73 Luthetium phthalocyanines
5.1.4 Silicon, Tin, and Lead Complexes 72 Crown-ether metallophthalocyanines
5.1.2 Manganese, Copper, Nickel, and Zinc Complexes Complexes of Co, Ni and Pb with the phthalocyanine (R= CH20C8H,,) exhibited mesophases below 100°C. Cobalt compound reacts with NaCN to give a polymeric material [ 1601. Discotic zinc and manganese
Columnar phthalocyanines linked by 0 - S i - 0 or 0 - S n - 0 spacers have also been described. The tin complex shows a columnar mesophase between 54 and 114°C with a rectangular packing of the columns. Silicon forms a dihydroxy silylphthalocyanine, 74, which shows a hexagonal columnar mesophase. On heating, loss of water occurred and led to polymeric materials with a polysiloxane spine [ 1631.
925
5 Metallomesogens with Polydentate Ligands
CHzOR
CHZOR
ROHzC CHzOR
76 Octasubstituted metalloporphyrins
I 74 Silylphthalocyanines
Lead forms a columnar discotic liquid crystal with the metal sitting out of the plane of the molecule, 75. X-ray data led to the suggestion that the molecules were stacked antiferroelectrically in a tilted stack [ 1641.
75 Lead phthalocyanines
Phthalocyanines with a different substitution pattern [ 1651were also described and again stable discotic mesophases were obtained.
5.2 Porphyrin Ligands The first most systematic work in this area were the porphyrin complexes, 76, with Cu(II), Cd(II), Zn(II), and Pd(I1) reported and carefully characterized by Gregg and coworkers [ 1661. The results indicate a clear influence of the metal and of the length of the alkyl chains on the mesomorphic behavior.
The synthesis and mesomorphic properties of Co, Cu, Ni, and Zn complexes of octakis(a1kylthio)tetraazaporphyrins has been reported by Doppelt and Huille [ 1671 and by Morelli and coworkers [ 1681 showing the hexagonal columnar nature of the mesophases. Some 5, 10, 15, 20-tetrasubstituted tetraphenyl porphyrins and their metal complexes (Co, Ni, Zn, and Pd) with square planar geometry were reported to exhibit discotic lamellar phases in contrast to the nonmesomorphic alkoxy derivatives [ 1691. In the D, phase, molecular planes align in a parallel manner to form a layered structure. However, Kujimiya et al. [ 1701reported that some alkoxy compounds are mesomorphic [171]. This is to compare with the zinc and copper complexes of tetrakis(carb0xyphenyl) porphyrins which exhibit thermotropic phase transition with no evidence of liquid cristallinity [ 1721.More recently, Bruce et al. [ 1731 reported zinc(I1) complexes of 5,15disubstituted porphyrins, 77. By extending the porphyrin along one axis via substitution at the 5 and 15 positions, the disc-like porphyrin can behave as a rod, showing calamitic N and SmA phases at elevated temperatures. Recently the same authors, to reduce these temperatures, introduced lateral chains that folded over the faces of the porphyrin. These new compounds melt, for example with n=7, at 140°C to a nematic phase, which
926
XIV
Metal-containing Liquid Crystals
clears at 177"C, whereas the previous one melts into a nematic phase at 305°C and clears with decomposition at 433°C [173b].
Copper(1) complexes show a lamellar structure not conclusively assigned [ 1751. Liquid crystalline complexes with chromium, molybdenum and tungsten as metallic centers have also been reported. The 1,4,7trisubstituted 1,4,7-triazacyclononane and three carbonyl groups, 80, are coordinated in an octahedral geometry.
77 Disubstituted metalloporphyrins
5.3 Other Amine Ligands
RO'
80 Triazacarbonyl complexes
The annelides form mesogenic copper complexes, 78. The lamellar phases, intermediate between crystalline solids and liquid crystals, have been termed tegma crystals [ 1831.
Aza and aza-thia macrocyclic complexes have been developed. The induction of new mesomorphic phases in saturated macrocyclic ligands was introduced by Lehn et al. [ 1 841. Macrocyclic [ 181ane N204 and [ 18lane N,S4 complexed with palladium give cationic complexes, 79, showing smectic mesophases [ 1741.
f
S
' S J
W
6 Lyotropic Metal-Containing Liquid Crystals While thermotropic metallomesogens are widely studied, there are very few examples of lyotropic' systems. Some amphiphilic Fe(I1) complexes, 81, have been synthe24
n
L
The observed mesophases are characterized as disordered rectangular columnar of a pyramidic type of X-ray measurements [176].
Y = CorO
79 Macrocyclic complexes
7
sized and are shown to give a hexagonal (H,) mesophase in water [177].
2 M+
r---
NC---M
Conclusions
927
phases [181]. A review on transition metal based ionic mesogens has been published in 1996 by F. Neve [185].
CN
CN 81 L
Surfactant complexes of Cr(II1) and Co(III), 82, with long chain and bidentate ligands have been reported and lyotropic liquid crystal properties have been determined for a few of them [ 1781.They have to be compared with Simon et al. annelide complexes.
82 Surfactant Co(I1) complexes
It has also been seen that disc-shape tetrapalladium compounds [ 1791 form charge transfer complexes exhibiting lyotropic cholesteric properties. Interesting surfactant tris(bipyridine) derivatives of ruthenium(II), 83, containing one alkyl chain show lyotropic mesomorphism in water [180]. r
1 2+
83 Lyotropic ruthenium complexes
New surfactants incorporating paramagnetic 0x0-vanadium centers, 84, were studied and some of the complexes were found to display both hexagonal and lamellar
84
7
J
Surfactant 0x0-vanadium complexes
Conclusions
Metallomesogens now form an interdisciplinary topic which is properly established. The introduction of a metal into a liquid crystal may reasonabily be expected to introduce a substantial number of novel features, such as color, paramagnetism, or conductivity. Other interesting and useful properties, arising from the large polarizable electron density associated with the metal center, may also be expected. A very interesting new development is the strong evidence that a continuum may exist between the rod- and the disc-like shapes and that rodlike shapes can give rise to discotic phases. A topic of increasing interest, so far not really explored, is the field of lyotropic metal-containing liquid crystals. Although many metals have already been incorporated into metallomesogens many others have not yet been used. Designs of new ligands have to be imagined. New applications will require new types of complexes. The field is young and there is still so much to do before finding applications for these materials.
Acknowledgements I should like to thank all our colleagues who have been working in the field of metallomesogens and with
928
XIV
Metal-containing Liquid Crystals
whom I always have valuable discussions. I am particularly grateful to P. Maitlis (Sheffield) for the first review we did together, D. Bruce (Sheffield) and J. C. Marchon for reading the manuscript, W. Moneta for his kind assistance with the drawings and all my coworkers for fruitful collaborations. I also thank the CNRS and the CENG for their financial support.
References [I] M. Marcos, J. L. Serrano, Adv. Matel: 1991, 30(5), 256-257. [2] A. M. Giroud-Godquin, P. M. Maitlis, Angew. Chem. Int. Ed. Engl. 1991,30,375-402. [3] P. Espinet, M. A. Esteruelas, L. A. Oro, J. L. Serrano, E. Sola, Coord. Chem. Reviews 1992,117, 215 - 274. [4] S. A. Hudson, P. Maitlis, Chem. Rev. 1993, 93, 861 -885. [5] D. W. Bruce inInorganicMuterials (Eds.: D. W. Bruce, D. O’Hare), 1992, Wiley, Chichester, pp. 405 -490. [6] A. P. Polishchuk, T. V. Timofeeva, Russian Chern. Rev. 1993,62(4), 291 -321. [7] D. W. Bruce, J. Chem. SOC.Dalton Trans. 1993, 2983-2989; D. W. Bruce, Adv. Matel: 1994, 6(9), 699-701. [8] (a) D. Vorlander, Bel: Dtsch. Chem. Ges. 1910, 43,3 120; (b) D. Vorlander, Z. Phys. Chem. Stoechiom. Verwandtschaftsl. 1923, 105, 21 1. [9] A. Skoulios,Ann. Phys. Paris 1978,3,421-450. [lo] J. MalthCte, J. Billard, Mol. Cryst. Liq. Cryst. 1976,34, 117. [I I ] A. M. Giroud, U.T. Mueller-Westerhoff, Mol. Cryst. Liq. Cryst. (Lett.) 1977, 41, 1 1 . [12] H. Adams, N. A. Bailey, D. W. Bruce, D. A. Dunmur, E. Lalinde, M. Marcos, C. Ridgway, A. J. Smith, P. Styring, P. M. Maitlis, Liq. Cryst. 1987,2, 381. [13] D. W. Bruce, S . M. Fan, G. R. Luckhurst, Liq. Cryst. 1994, 16(6), 1093- 1099. [14a] J. P. Rourke, F. P. Fanizzi, N. J. Salt, D. W. Bruce, D. A. Dunmur, P. M. Maitlis, J. Chem. Soc. Chem. Commun. 1990, 229. J. Chem. Soc. Dalton Trans. 1992, 3009-3014. [14h] B . Donnio, D. W. Bruce, J. Chem. Soc. Dalton Trans 1997, 2745-2755. [ 151 D. W. Bruce,D. A. Dunmur, S. A. Hudson, P. M. Maitlis, P. Styring, Adv. Muter for Optics and Electronics 1992, 1, 37-42. [16] H. Adams, N. A. Bailey, D. W. Bruce, S. C. Davis, D. A. Dunmur, P. D. Hempstead, S. A. Hudson, S. Thorpe, J . Muter Chern. 1992, 2(4), 395-400; D. W. Bruce, B. Donnio, S. A. Hudson, A. M. Levellut, S . Megtert, D. Petermann, M. Veber, J. Phys. (France) 1995,5(2), 289.
[I71 D. W. Bruce, S. Hudson, J. MateK Chern. 1994, 4(3), 479-486. [18] D. W. Bruce, M. D. Hall, Mol. Cryst. Liq. Cryst. 1994,250,373- 375. [19] J. P. Rourke, F. P. Fanizzi, D. W. Bruce, D. A. Dunmur, P. M. Maitlis, Dalton J. Chem. Soc. Trans 1992,3009-3014. [20] D. W. Bruce, S. C. Davis, D. A. Dunmur, S . A. Hudson, P. M. Maitlis, P. Styring, Mol. Crysr. Liq. Crysf. 1992,215, 1 - 11. [21] H. A. Adams, N. A. Bailey, D. W. Bruce, S. A. Hudson, J. R. Marsden, Liq. Cryst. 1994, 16(4), 643 -653. [22] D. W. Bruce, J. P. Rourke, Polyhedron 1995,14, 1915-1922. [23] M. A. Esteruelas, E. Sola, L. A. Ora, M. B. Ros, J.L. Serrano, J. Organomet. Chem. 1990, 387, 103. [24] M. Marcos, M. B. Ros, J. L. Serrano, M. A. Esteruelas, E. Sola, L. A. Oro, J. Barbera, Chern. Mater 1990,2, 748. [25] S. Takahashi, M. Kariya, T. Yakate, K. Sonogashira, N. Hagihara, Macromolecules 1978, 11, 1063. [26] S. Takahashi, Y.Takai, H. Morimoto, K. Sonogashira, N. Hagihara, Mol. Cryst. Liq. Cryst. 1982, 82, 139-143. [27] T. Kaharu, H. Matsuhara, S. Takahashi, J. Muter: Chem. 1991, 1 , 145-146; ibid. 1992, 2(1), 43-47. [28] D. W. Bruce, M. S. Lea, J. R. Marsden, Mol. Cryst. Liq. Cryst. 1996, 275, 183. [29] T. Kaharu, H. Matsuhara, S . Takahashi, Mol. Cryst. Liq. Cryst. 1992,220, 191 - 199. (a) T. Kaharu, R. Ishii, S. Takahashi, J. Chern. Soc. Chem. Commun. 1994, 1349- 1350; (b) T. Kaharu, T. Tanaka, M. Sawada, S. Takahashi, J . Muter. Chern. 1994,4(6), 859-865. P. Alejos, S. Coco, P. Espinet, New J. of Chem. 1995, 19(7), 799. A. S. C. Lawrence, Trans. Faraday Soc. 1938, 34,660. B. Gallot, A. Skoulios, Kolloid Z. Z. Polyrn. 1966,209(2), 164- 169. H. A. Ellis, Mol. Cryst. Liq. Cryst. 1986, 139, 281 -290. . C. G. Bazuin, D. Guillon, A. Skoulios, A. M. Amorim da Costa, H. D. Burrows, C. F. G. C. Geraldes, J. J. Teixeira-Dias, E. Blackmore, G. J. T. Tiddy, Liq. Cryst. 1988, 3(12), 16551670. G. S. Attard, Y. D. West, Liq. Cryst. 1990, 7(4), 487-494. J. Lindau, W. Hillman, H. D. Doerfler, H. Sachmann, Mol. Cryst. Liq. Cryst. 1986, 133, 259. R. F. Grant, Can. J. Chem. 1964,42,951-953. A. M. Giroud-Godquin, J. C. Marchon, D. GuilIon, A. Skoulios, J . Phys. Lett. (Orsay, F K )1984, 45, L681 -L684.
8 References [40] H. Abied, D. Guillon, A. Skoulios, P. Weber, A. M. Giroud-Godquin, J. C. Marchon, Liq. Cryst. 1987, 2(3), 269-279; M. Ibn-Elhaj, D. Guillon, A. Skoulios, A. M. Giroud-Godquin, P. Maldivi, Liq. Cryst. 1992, 11(5), 731-744. [41] J. R. Lomer, K. Perrera, Acta Crystullogruphicu 1974, B30,2912-2915. [42] P. Maldivi, D. Guillon, A. M. Giroud-Godquin, H. Abied, H. Dexpert, A. Skoulios, J. Chim. Phys. 1989,86, 1651 - 1664. [43] A. M. Giroud-Godquin, J. M. Latour, J. C. Marchon, lnorg. Chem. 1985,24,4452. [44] (a) 0. Poisat, D. P. Strommen, P. Maldivi, A. M. Giroud-Godquin, J. C. Marchon, Inorg. Chem. 1990, 29, 4853; (b) A. M. Giroud-Godquin, P. Maldivi, J. C. Marchon, M. BCe, L. Carpentier, Mol. Phys. 1989, 68, 1353. [45] (a) A. M. Giroud-Godquin, P. Maldivi, J. C. Marchon, P. Aldebert, A. Peguy, D. Guillon, A. Skoulios, J. Phys. (Orsay, Fr.) 1989, 50, 5 13 - 5 19; (b) M. BCe, A. M. Giroud-Godquin, P. Maldivi, J. Williams, Mol. Phys. 1994,81(1), 57-68. [46] P. Maldivi, L. Bonnet, A. M. Giroud-Godquin, M. Ibn-Elhaj, D. Guillon, A. Skoulios, Adv. Muter. 1993,5(12), 909-912. [47] G. S. Attard, P. R. Cullum, Liq. Cyst. 1990,8(3), 299-309. [48] J. C. Marchon, P. Maldivi, A. M. Giroud-Godquin, D. Guillon, A. Skoulios, D. P. Strommen, Phil. Trans. Roy SOC. Lond. A 1990, 330, 109- 116. [49] J. Barbera, M. A. Esteruelas, A.M. Levelut, L. A. Oro, J. L. Serrano, E. Sola, Inorg. Chem. 1992,31(5), 732-737. [50] M. Ibn-Elhaj, D. Guillon, A. Skoulios, A. M. Giroud-Godquin, J. C. Marchon, J. Phys. France 1992,2,2197-2206. [51] F. D. Cukiernik, P. Maldivi, A. M. GiroudGodquin, J. C. Marchon, M. Ibn-Elhaj, D. Guillon, A. Skoulios, Liq. Cryst. 1991, 9(6), 903 - 906. [52] L. Bonnet, F. D. Cukiernik, P. Maldivi, A. M. Giroud-Godquin, J. C. Marchon, M. Ibn-Elhaj, D. Guillon, A. Skoulios, Chem. Muter. 1994, 6, 31-38. [53] F. D. Cukiernik, A. M. Giroud-Godquin, P. Maldivi, J. C. Marchon, lnorg. Chim. Acta 1994, 215,203-207. [54] D. V. Baxter, R. H. Cayton, M. H. Chisholm, J. C. Huffman, E. F. Putilina, S. L. Tagg, J. L. Wesemann, J. W. Zwanziger, F. D. Darrington, J. Am. Chem. SOC. 1994,116,4551-4566. [55] B. J. Bulkin, R. K. Rose, A. Santoro, Mol. Cryst. Liq. Cryst. 1977,43, 53. [56] A. M. Giroud-Godquin, J. Billard, Mol. Cryst. Liq. Cryst. 1981,66, 147. [57] A. C. Ribeiro, A. F. Martins, A. M. Giroud-Godquin, Mol. Cryst. Liq. Cryst. (Lett) 1988,5, 133.
929
[58] K. Ohta, A. Ishii, H. Muroki, I. Yamamoto, K. Matsuzaki, Mol. Cryst. Liq. Cryst. 1985, 116, 299; K. Ohta, A. Ishii, I. Yamamoto, K. Matsuzaki, J. Chem. SOC. Chem. 1984, 1099. [59] K. Usha, B. K. Sadashiva, K. Vijayan, Mol. Cryst. Liq. Cryst. 1994,241, 91 - 102. [60] S. Chandrasekhar, B. K. Sadashiva, B. S . Srikanta, Mol. Cryst. Liq. Cryst. 1987, 151, 93. [61] S. Chandrasekhar, B. R. Ratna, B. K. Sadashiva, V. N. Raja, Mol. Cryst. Liq. Cryst. 1988, 165, 123. [62] A. Ghode, U. Shivkumar, B. K. Sadashiva, Bull. Muter. Sci. 1994, 17(3), 283-297. [63] B. Miilberger, W. Haase, Liq. Cryst. 1989,5,25 1. [64] K. Ohta, H. Akimoto, T. Fujimoto, I. Yamamoto, J. Mat. Chem. 1994, 4(1), 61-69; K. Ohta, Y. Morizumi, H. Akimoto, 0. Takenaka, T. Fujimoto, I. Yamamoto, Mol. Cryst. Liq. Cryst. 1992,214, 143- 149. [65] N. J. Thomson, G. W. Gray, J. W. Goodby, K. J. Toyne, Mol. Cryst. Liq. Cryst. 1991,200, 109. [66] A. M. Giroud-Godquin, G. Sigaud, M. F. Achard, F. Hardouin, J. Phys. Lett. (Orsay, France) 1984, 45, L 387; A.M. Giroud-Godquin, M. M. Gauthier, G. Sigaud, F. Hardouin, M. F. Achard, Mol. Cryst. Liq. Cryst. 1986, 132, 35. [67] K. Ohta, H. Ema, H. Muroki, I. Yamamoto, K. Matsuzaki, Mol. Cryst. Liq. Cryst. 1987,147,61. [68] W. Haase, V. S. Reyes, M. Athanassopoulou, G. S. Garrido, K. Griesar, E. Schuhmacher,MoE. Eng. 1995, in press. [69] A. B. Blake, J. R. Chipperfield, S . Clark, P. G. Nelson, J. Chem. SOC. Dalton Trans 1991, 1159. [70] V. S. Reyes, G. S. Garrido, C. Aguilera, K. Griesar, M. Athanassopoulou, H. Finkmann, W. Haase, Mol. Muter. 1995, in press. [71] C. K. Lai, A. G. Serette, T. M. Swager, J . Am. Chem. SOC. 1992,114,7948-7949. [72a] K. Hanabusa, T. Suzuki, T. Koyama, H. Shirai. Makromol. Chem. 1992,193,2149-2161. [72b] L. Oriel, J. L. Serrano, Adv. Muter. 1995, 7, 348 - 369. [73] K. Usha, B. K. Sadashiva,Mol. Cryst. Liq. C n x t . 1994,241,91- 102. [74] A. M. Giroud-Godquin, A. Rassat, C. R. Acnd. Sci. Ser. 2 1982,294,241. [75] S. N. Poelsma, A. H. Servante, F. P. Fanizzi, P. M. Maitlis, Liq. Cryst. 1994, 16(4), 675-685. [76] P. Styring, S. Tantrawong, D. R. Beattie, J. W. Goodby, Liq. Cryst. 1991, Z0(4), 581-584. [77] H. Zheng, T. M. Swager, J. Am. Chem. SOC. 1994,116,761-762 [78] H. Zheng, P. J. Carroll, T. M. Swager, Liq. Cryst. 1993, I4(5), 1421 - 1429. [79] J. Barbera, C. Cativiela, J. L. Serrano, M. M. Zurbano, Adv. Muter. 1991,3(12), 602-605. [80] K. Ohta, M. Moriya, M. Ikejima, H. Hasebe, T. Fujimoto, I. Yamamoto, Bull. Chem. Soc. Jpn. 1993,66,3553-3564.
930
XIV
Metal-containing Liquid Crystals
[81] U. T. Mueller-Westerhoff, A. Nazzal, R. J. Cox, A. M. Giroud, Mol. Cryst. Liq. Cryst. (Lett.) 1980, 249-255. 1821 M. Veber, R. Fugnitto, H. Strzelecka, Mol. Cryst. Liq. Cryst. 1983, 22 1 . 1831 M. Veber, P. Davidson, C. Jallabert, A.M. Levelut, H. Strzelecka, Mol. Cryst. Liq. Cryst. 1987, I . [84] K. Ohta, H. Hasebe, H. Ema, T. Fujimoto, I. Yamamoto, J. Chem. Soc. Chem. Commun. 1989, 1610. 1851 H. Adams, A. C. Albeniz, N. A. Bailey, D. W. Bruce, A. S. Cherodion, R. Dhillon, D. A. Dunmur, P. Espinet, J. L. Feijoo, E. Lalinde, P. M. Maitlis, R. M. Richardson, G. Ungar, J. Muter. Chem. 1991, I , 843. [86] K. Ohta, Y. Morizumi, H. Ema, T. Fujimoto, I. Yamamoto, Mol. Cryst. Liq. Cryst. 1992, 214, 151 - 159. [87] D. Guillon, D. W. Bruce, P. Maldivi, M. Ibn-Elhaj, R. Dhillon, Chem. Muter. 1994,6,182- 189. 1881 K. Ohta, H. Ema, Y. Morizumi, T. Watanabe, T. Fujimoto, I. Yamamoto, Liq. Cryst. 1990, 8, 3 1 1 - 330. [89] D. W. Bruce, R. Dhillon, D. A. Dunmur, P. M. Maitlis, Muter. Chem. 1992,2(1), 65-69. 1901 J. P. Fackler Jr, J. A. Fetchin, D. C. Fries, J. Am. Chem. Soc. 1972, 94,7323. [91] N. Hoshino-Miyajima, J. Chem. Soc. Chem. Commun. 1993, 1442. 1921 M. J. Baena, P. Espinet, M. C. Lequerica,A. M. Levelut, J. Arner. Chem. Soc. 1992, 114, 4182-41 85. [93] K. Ohta, H. Ema, I. Yamamoto, K. Matsuzaki, Liq. Cryst. 1988, 3, 1671. [94] I. V. Ovchinnikov, Yu. G. Galyametdinov, G. I. Ivanova, L. M. Yagfarova, Dokl. Akad. Nauk SSSR,Sex Khim 1984,276, 126. [95] Yu G. Galyametdinov, I. V. Ovchinnikov, B. M. Bolotin, N. B. Etingen, G. I. Ivanova, L. M. Yagfarova, Izv. Akad. Nauk SSSR, Ser. Khim. 1984,2379. 1961 U. Caruso, A. Roviello, A. Sirigu, Liq. Cryst. 1991, 10(1), 85-93. [97] M. Marcos, P. Romero, J. L. Serrano, C. Bueno, J. A. Cabeza, L. A. Oro, Mol. Cryst. Liq. Cryst. 1989, 167, 123; M. Marcos, P. Romero, J. L. Serrano, Chem. Muter. 1990,2,495. 1981 N. Hoshino, H. Murakami, Y. Matsunaga, T. Inabe, Y. Maruyama, Inorg. Chem. 1990, 29, 1177. 1991 U. Caruso, A. Roviello, A. Sirigu, Liq. Cryst. 1988, 3, 1515. [loo] J. P. Bayle, E. Bui, F. Perez, J. Courtieu, Bull. Soc. Chim. Fr. 1989,4,513. [I011 E. Bui, J. P. Bayle, F. Perez, L. Liebert, J. Conrtieu, Liq. Cryst. 1990,4, 513. [I021 V. Prasad, B. K. Sadashiva, Mol. Cryst. Liq. Cryst. 1993, 225, 303-312; V. Prasad, B. K.
Sadashiva, Mol. Cryst. Liq. Cvst. 1994, 241, 167- 174. [I031 E. Campillos, M. Marcos, J. L. Serrano, J. Barbera, P. J. Alonso, J. I. Martinez, Chem. Mnrer. 1993,5, 1518-1525. [lo41 E. Camppillos, M. Marcos, L. T. Oriol, I. L. Serrano, Mol. Cryst. Liq. Cryst. 1992, 215. 127-135. [lo51 M. Marcos, J. L. Serrano, T. Sierra, M. J. GimCnez, Chem. Matel: 1993,5, 1332- 1337: M. Marcos, J. L. Serrano, T. Sierra, M. J. Gimenez,Angew. Chem. Int. Ed. Engl. 1992,3/( 1 I ) , 1471- 1472. [lo61 M. J. Baena, P. Espinet, M. B. Ros, J. L. Serrano, A. Ezcurra, Angew. Chem. Int. Ed. Engl. 1993,32(8), 1203- 1205. [lo71 M. Ghedini, D. Pucci, E. Cesarotti, 0. Francescangeli, R. Bartolino, Liq. Crysr. 1993, / 5 ( 3 ) , 33 1 - 344. [lo81 W. Pyzuk, Yu Galyametdinov, Liq. Cryst. 1993, I5(2), 265-268. [lo91 N. Hoshino, R. Hayakawa, T. Shibuya, Y. Matsunaga, Inorg. Chem. 1990,29,5 129 - 5 I3 I ;N. Hoshino, A. Kodama, T. Shibuya, Y. Matsunaga, S . Miyajima, Inorg. Chem. 1991, 30, 3091-3096; E. Campillos, M. Marcos, A. Omenat, J. L. Serrano, J. Mat. Chem. 1996, 6(3),349. [110] E. Campillos, M. Marcos, J. L. Serrano, J. Mute,: Chem. 1993,3(10), 1049- 1052; S. Armentano, G. De Munno, M. Ghedini, S. Morrone, Inorg. Chim. Acta 1993, 210, 125- 127. [ I l l ] J. Barbera, A. M. Levelut, M. Marcos, P. Romero, J. L. Serrano, Liq. Cryst. 1991, /0(1), 119- 126. [112] M. Ghedini, S. Morrone, R. Bartolino, V. Formoso, 0. Francescangeli, B. Yang, D. Gatteschi, C. Zanchini, Chem. Muter. 1993, 5, 876- 882; M. Ghedini, S. Morrone, 0.Francescangeli, R. Bartolino, Mol. Cryst. Liq. Cryst. 1995,250,323. [113] B. K. Sadashiva, A. Ghode, Liq. Cryst. 1994, 16(1), 33-42. [ 1 141 Y. Galyametdinov, G. Ivanova, K. Griesar, A. Prosvirin, I. Ovchinnikov, W. Haase, A h . Muter. 1992,4(11), 739-741. [I IS] K. Griesar, Y. Galyametdinov, M. Athanassopoulou, I. Ovchinnikov, W. Haase, Adv. Muter. 1994, 6(5),381 -384. [ 1161 P. J. Alonso, M. Marcos, J. L. Martinez, V. M. Orera, M. L. Sanjuan, J. L. Serrano, Liq. Cryst. 1993, 13(4), 585-596; P. J. Alonso, M. Marcos, J. I. Martinez, J. L. Serrano, T. Sierra, Adv. Muter. 1994,6(9), 667-670; P. J. Alonso, J. A. Puertolas, P. Davidson, B. Martinez, J. I. Martinez, L. Oriol, J. L. Serrano, Macromolecules 1993,26,4304-4309. [ 1171 P. BerdaguC, J. Courtieu, P. M. Maitlis, J. Chem. Soc.: Chem. Commun. 1994, 1313- 1314.
8 References [ I 181 C. Carfagna, U. Caruso, A. Roviello, A. Sirigu, Makromol. Chem. Rapid. Commun. 1987, 8, 345. [ 1 191 U. Caruso, A. Roviello, A. Sirigu, Mczcromolecules 1991, 24, 2606-2609. [I201 M. Marcos, L. Oriol, J. L. Serrano, Macromolecules 1992, 25, 5362-5368; L. Oriol, P. J. Alonso, J. I. Martinez, M. Pinol, J. L. Serrano, Macromolecules 1994, 27(7), 18691874. [I211 W. Haase, K. Griesar, M. F. Iskander, Y. Galyametdinov, Mol. Mat. 1993,3, 115- 130. [I221 R. Paschke, H. Zaschke, A. Madicke, J. R. Chipperfield, A. B. Blake, P. G. Nelson, G. Gray, Mol. Cryst. Liq. Cryst. (Lett.) 1988,6,81. See also: A. B. Blake, J. R. Chipperfield, W. Hussain, R. Paschke, E. Sinn, Inorg. Chem. 1995,34, 1125- 1129. [I231 T. D. Shaffer, K. A. Sheth, Mol. Cryst. Liq. Cryst. 1989, 172, 27. [ 1241 K. Ohta, Y. Morizumi, T. Fujimoto, Mol. Liq. Cryst. 1992,214, 161- 169. [125] M. Ghedini, F. Neve, D. Pucci, E. Cesarotti, M. Grassi, J. Organomet. Chem. 1992, 438, 343-351. [I261 V. Formoso, M. C. Pagnotta, P. Mariani, M. Ghedini, F. Neve, R. Bartolino, M. More, G. Pepy, Liq. Cryst. 1992, 11(5), 639-654. [127] R. Bartolino, G. Coddens, F. Rustichelli, M. C. Pagnotta, C. Versace, M. Ghedini, F. Neve, Mol. Cryst. Liq. Cryst. 1992, 221,
101 - 108. [ 1281 A. Amoddeo, R. Bartolino, L. S . Caputu, E. Co-
lavita, V. Formoso, M. Ghedini, A. Oliva, D. Pucci, C. Versace, Mol. Cryst. Liq. Cryst. 1992, 221,93-99. 11291 M. Ghedini, D. Pucci, E. Cesarotti, 0 . Francescangeli, R. Bartolino, Liq. Cryst. 1994, 16(3), 373-380. [130] C. Versace, V. Formoso, D. Lucchetta, D. Pucci, C. Ferrero, M. Ghedini, R. Bartolino, J. Chem. Phys. 1993,98(1 l), 8507- 85 13. 11311 M. Ghedini, S. Morrone, 0. Francescangeli, R. Bartolino, Chem. Muter. 1992, 4(5), 11191123. 11321 N. Hoshino, H. Hasegawa, Y. Matsunaga, Liq. Cryst. 1991, 9(2), 267-276. [ 1331 A. Omenat, M. Ghedini, J. Chem. Soc.: Chem. Commun. 1994, 1309- 1310. [ 1341 J. Barbera, P. Espinet, E. Lalinde, M. Marcos, J. L. Serrano, Liq. Cryst. 1987, 2, 833; M. B. Ros, N. Ruiz, J. L. Serrano, P. Espinet, Liq. Cryst. 1991, 9(1), 77-86. [135] M. J. Baena, J. Buey, P. Espinet, H. S. Kitzerow, G. Heppke, Angew. Chem. lnt. Ed. Engl. 1993,32(8), 1201- 1203. [ 136) M. J. Baena, J. Barbera, P. Espinet, A. Ezcurra, M. B. Ros, J. L. Serrano, J. Am. Chem. SOC. 1994, 116(5), 1899- 1906.
93 I
[137] M. J. Baena, P. Espinet, M. B. Ros, J. L. Serrano, Angew. Chem. Int. Ed. Engl. 1991, 6, 71 1-712. [138] K. Praefcke, D. Singer, B. Gundogan, Mol. Cryst. Liq. Cryst. 1992, 223, 191- 195: B. Gundogan, K. Praefcke, Chem. Ber. 1993,126. 1253- 1255; K. Praefcke, S. Diele, J. Pickardt, B. Gundogan, U. Nutz, D. Singer, Liq. Crysf. 1995, 18, 857. [139] D. Singer, A. Liebrnann, K. Praefcke, J. H. Wendorff, Liq. Cryst. 1993, 14(3), 785-794; K. Praefcke, D. Singer, B. Gundogan, K. Gutbier, M. Langner, Be,: Bunsenges. Phw. Chem. 1994,98, 118- 122. [140] K. Praefcke, B. Bilgin, J. Pickardt, M. Borowski, Chem. Be,: 1994, 127, 1543-1545; K. Praefcke, B. Bilgin, N. Usol’tseva, B. Heinrich, D. Guillon, J. Muter. Chem. 1995, 5 , 2257. [141] K. Praefcke, B. Bilgin, J. Pickardt, M. Borowski, Chem. Ber. 1994,127, 1543- 1545. [142] D. Bruce, X. H. Liu,J. Chem. SOC.Chenz. Commun. 1994,729-730; D. Bruce, X. H. Liu, Liq. Cryst. 1995, 18, 165. [143] P. Espinet, E. Lalinde, M. Marcos, J. Perez. A. Rernon, J. L. Serrano, Organometallics 1990, 9, 555. [144] M. N. Abser, M. Bellwood, M. C. Holrnes, R. W. McCabe, J. Chem. SOC.:Chem. Cornmuti. 1993, 1062. [145] M. Ghedini, D. Pucci, G. De Munno. D. Viterbo, F. Neve, S. Armantano, Chem. Marex 1991, 3, 65. [146] M. Ghedini, D. Pucci, R. Bartolino, 0. Francescangeli, Liq. Cryst. 1993, 132). 255 -263. [147] A. Togni, T. Hayashi (eds.), Ferrocenes, VCH. Weinheim, 1994. [148] J. Bhatt; B. M. Fung, K. M. Nicholas. C. D. Poon, J. Chem. Soc.: Chem. Commun. 1988, 1439; J. Bhatt, B. M. Fung, K. M. Nicholas, Liq. Cryst. 1992, 12, 263. [149] R. Deschenaux, J. L. Marendaz, J. Santiago, Helv. Chim. Acta 1993, 76, 865 - 876. [150] R. Deschenaux, M. Rama, J. Santiago, Tetrcihedron Lett. 1993, 34(20), 3293-3296. [151] K. P. Reddy, T. L. Brown, Liq. Crysr. 1992. 12(3), 369-376. [152] R. Deschenaux, J. Santiago, D. Guillon, B. Heinrich, J. Matel: Chem. 1994,4(5),679-682. [I531 R. Deschenaux, J. Santiago, Tetrahedron Lett. 1994,35(14),2169-2172. [ 1541 R. Deschenaux, J. L. Marendaz, J. Chem. Soc.: Chem. Commun. 1991, 909-910; R. Deschenaux, 1. Kosztics, J. L. Marendaz, H. StoeckliEvans, Chimia 1993,47,206-210. [I551 R. Deschenaux, I. Kosztics, U. Scholten, D. Guillon, M. Ibn-Elhaj, J. Mute,: Chem. 1994, 4(8), 1351-1352. [156] R. Deschenaux, J. Santiago, J. Muter. Cheni. 1993, 3(2), 219-220.
932
XIV
Metal-containing Liquid Crystals
[I571 L. Ziminski, J. MalthCte, J. Chem. Soc.: Chem. Cornmun. 1990, 1495. [ 1581 C. Piechoki, J. Simon, A. Skoulios, D. Guillon, P. Weber, J. Am. Chem. Soc. 1982,104,5245. [159] C. Sirlin, L. Bosio, J. Simon, V. Ahsen, E. Yilmazer, 0. Bekgoglu, Chem. Phys. (Lett.) 1987, 139, 362. 11601 M. Hannack, A. Beck, H. Lehmann, Synthesis 1987,703. [161] D. Guillon, P. Weber, A. Skoulios, C . Piechoki, J. Simon, Mol. Cryst. Liq. Cryst. 1985,130, 223. [ I621 C. Piechocki, J. Simon, J. J. AndrC, D. Guillon, P. Petit, A. Skoulios, P. Weber, Chem. Phys. (Lett.) 1985, 122, 124; Z. Belarbi, C. Sirlin, J. Simon, J. J. AndrC, J. Phys. (Lett) 1989, 93, 8105. [163] C. Sirlin, L. Bosio, J. Simon, J. Chem. Soc. Chem. Communn. 1988,236. [164] P. Weber, D. Guillon, A. Skoulios, J. Phys. Chem. 1987, 91,2242. [165a] M. J. Cook, M. F. Daniel, K. J. Harrison, N. B. McKeown, A. J. Thompson, J. Chem. Soc.: Chem. Communn. 1987,1086. [165b] A. Sirigu, Liq. Cryst. 1993, 14(1), 15-36. [166] B. A. Gregg, M. A. Fox, A. J. Bard, J. Am. Chem. Soc. 1989, I l l , 3024-3029. 11671 P. Doppelt, S. Huille, New J. Chem. 1991, 14, 607. [168] F. Lelj, G. Morelli, G . Ricciardi, A. Roviello, Chem. Phys. (Lett.) 1991, 185,468; F. Lelj, G. Morelli, G. Ricciardi, A. Roviello, A. Sirigu, Liq. Cryst. 1992,12(6), 941-960. [I691 Y. Shimizu, M. Miya, A. Nagata, K. Ohta, A. Matsumara, I. Yamamoto, S. Kusabayashi, Chem. (Lett.) 1991, 25; Y. Shimizu, T. Matsuno, M. Miyia, A. Nagata, J. Chem. Soc. Chem. Commun. 1994,2411. 11701 S. I. Kugimiya, M. Takemura, Tetrahedron (Lett.) 1990, 31(22), 3157-3160. [I711 Y. Shimizu, M. Miya, A. Nagata, K. Ohta, I. Yamamoto, S. Kusabayashi, Liq. Cryst. 1993, 14(3), 795-805.
11721 R. Ramasseul, P. Maldivi, J. C . Marchon, M. Taylor, D. Guillon, Liq. Cryst. 1993, 13(5), 729-733. [173a] D. Bruce, D. A. Dunmur, L. S. Santa, M. A. Wali, J. Muter. Chem. 1992, 2(3), 363-364; D. Bruce, M. A. Wali, Q. Min Wang, J. Chem. SOC.:Chem. Cornmun. 1994, 20892090. [173b] Q. M. Wang, D. W. Bruce, Angew. Chem. Int. Ed. Engl. 1997, 36, 150. 11741 F. Neve, M. Ghedini, J. Inclusion Phenomena and Molecular Recognition in Chemistry 1993, 15,259-272. [175] F. Neve, M. Ghedini, A. M. Levelut, 0. Francescangeli, Chem. Muter. 1994, 6(1), 70-76. [176] S. Schmidt, G. Lattermann, R. Kleppinger, J. H. Wendorff, Liq. Cryst. 1994, 16(4), 693- 702. [I771 D. W. Bruce, D. A. Dunmur, P. M. Maitlis, J. M. Watkins, G . J. T. Tiddy, Liq. C y s t . 1992, 11(1), 127-133. 11781 D. W. Bruce, I. R. Denby, G. J. T. Tiddy, J. M. Watkins, . I . Mate,: Chem. 1993, 3(9), 91 1-916. 11791 N. Usol’tseva, K. Praefcke, D. Singer, B. Giindogan, Liq. Cryst. 1994, 16(4), 617-623. [I801 D. W. Bruce, J. D. Holbrey, A. R. Tajbakhsh. G. J. T. Tiddy, J. Muter. Chem. 1993, 3(8), 905 - 906. [I811 S. S. Zhu, T. M. Swager, Adv. Mater. 1995, 7, 280. [I821 Yu. G. Galyametdinov, 0. N. Kadkin, 1. V. Ovchinnikov, Izv. Akad. Nuuk., Set Khim. (Russia) 1990, 10, 2462; Yu. G . Galyametdinov, 0. N. Kadkin, I. V. Ovchinnikov, lit: Akad. Nauk., Ser. Khim. (Russia) 1992, 2, 402. [ 1831 D. Markovitsi, J. J. AndrC, A. Mathis, J. Simon, P. Spegt, G . Weill, M. Ziliox, Chem. Phys. Leu. 1984, 104,48. [I841 J. M. Lehn, J. MalthCte, A. M. Levelut, J. Chem. Soc. Chem. Commun. 1985, 1794. [ 1851 F. Neve, Adv. Muter. 1996,8(4), 277 - 289.
Handbook of Liquid Crystals Edited by D. Demus, J. Goodby, G. W. Gray, H.-W. Spiess, V. Vill Copyright 0 WILEY-VCH Verlag GmbH. 1998
Chapter XV Biaxial Nematic Liquid Crystals B. K. Sadashiva
1 Introduction The nematic mesophase is exhibited by a large number of compounds composed of rigid rod-like molecules. This mesophase possesses a high degree of long range orientational order, but no long range positional order. In other words, it is somewhat similar to the ordinary liquid except that the molecules are arranged approximately parallel to one another. Because of the lack of positional order of the molecules, the nematic phase is quite fluid and the molecules can freely slide past one another. However, a well aligned sample of a nematic phase is optically uniaxial [ l , 21. It is well known that a smectic C liquid crystal which has a monoclinic symmetry is optically biaxial [3]. A biaxial nematic (N,) phase in which the molecules are oriented along the three directions in space, i.e., they have three directors which are perpendicular to one another, is possible. Thus the shape of molecules exhibiting the N, phase should be different from those exhibiting the uniaxial nematic phase.
2 Theoretical Prediction of the Biaxial Nematic Phase Based on the interaction employed in the Maier-Saupe theory of the nematic state, Freiser [4] was the first to predict the possible existence of an N, phase as an intermediate between two uniaxial nematic phases. Later, a number of other theoretical investigations were carried out 15-71 using various models to predict the possibility of obtaining an N, liquid crystal. In all these models, a system consisting of hard rectangular plates was considered. These approaches gave the same result, an N, phase should be obtained between two uniaxial nematics of opposite sign, i.e., those made up of rod-like and plate-like molecules. They also predicted that a transition from a uniaxial nematic to a biaxial nematic would be second order. There was not much progress in the design and synthesis of rectangular plate-like molecules. However, the theoreticians [8 - 121 continued their efforts by considering systems in which mixtures of rod-like (N,) and rectangular plate-like (Nd) molecules were used in varying proportions. These investigations led to the conclusion that mixtures containing rod-like and platelike molecules should produce an N, phase, and a typical phase diagram is shown in Fig. 1.
934
X V Biaxial Nematic Liquid Crystals
Figure 1. Phase diagram of the uniaxial and biaxial nematic phases as predicted by the microscopical theories (N, calamitic nematic; N, biaxial nematic; and N, discotic nematic). In the case of systems of hard rectangular plates, the parameter x is the shape anisotropy of the elementary units (i.e., the width to length ratio of the rectangles). In the case of mixtures of rodlike and disk-like particles, x is the relative concentration of the disk-like particles. The first order transition to the isotropic phase is marked as a dashed line. The second order N,-N, phase transitions are represented with solid lines (from [8, 131).
3 Structural Features The N, phase was first observed experimentally by Yu and Saupe [14] in an amphiphilic system consisting of potassium laurate + 1-decanol + D,O. It was suggested [ 151 that a convenient way of obtaining the N, phase in low molecular weight thermotropic systems is by bridging the gap between rod-like and disk-like molecules i.e., by preparing a mesogen that combines the features of the rod and the disk. It may also be possible to obtain an N, phase from discotic systems in which the constituent molecules are elliptical in shape rather than circular. By some chemical ingenuity it may be possible to synthesize compounds comprised of elliptically shaped molecules that exhibit an N, phase. The compounds designed in this manner are likely to exhibit the N, phase as a result of hindrance in the rotation of the constituent molecules along their long axis.
It is perhaps appropriate to mention here that both nematic liquid crystalline side chain polymers with laterally attached mesogenic groups [I61 and sanidic aromatic polyamides [ 171 have been shown to exhibit N, phases. Praefcke et al. [ 181 have discussed a few structural features or molecular shapes of compounds that have been synthesized with a view to obtaining an N, phase, and these are shown in Fig. 2. They are (a) elliptically shaped stretched disk, (b) two-dimensionally flat bones or disk-rod-disk, (c) spoons or disk-rod, and (d) a pipet or rod-disk-rod. It should be emphasized here that these are only a few examples and various other shapes are also possible. Malthete et al. [ 191 were the first to make use of theoretical ideas and they synthesized 4-[3',4',5'-tri(p-n-dodecyloxybenzyloxy)] be n z oy 1ox y - 4"- dodec y 1ox y b en z o y 1ox y biphenyl (compound l), which combined
""'I2+
I H25C120
the features of the rod and the disk. Though they reported an N, phase for this compound, it has now been established that the characterization of the mesophase is uncon-
a
b
C
d
Figure 2. Shapes of low molecular weight compounds that have been synthesized with a view to obtaining an N, phase: (a) ellipses, (b) hones, (c) spoons, and (d) pipets (from [IS]).
4
vincing and that there is a smectic Cnematic phase transition [20]. However, the possibility exists that the nematic phase could have an undetectable weak biaxiality.
4 Synthesis Several low molecular weight compounds have been synthesized incorporating the features described above in an attempt to observe biaxiality in the nematic phase. Many such copper(I1) complexes were synthesized [21] exhibiting the nematic phase in a metastable state. The overall shape of the molecules of these complexes resemble that of a pipet. One of these complexes, bis[ 1-(p-n-decylbiphenyl)-3-(p-ethoxypheny1)propane-1,3-dionato] copper(I1) (compound 2) showed evidence of biaxiality in HZlClO
935
Synthesis
boxylic acids exist as dimers. These dirners are in trans conformation and are assumed to be more or less planar as a result of the conjugation between their phenyl and carboxylic groups. They also represent a new structural type of bone-shaped mesogens. The transition temperatures of these cinnamic acids are given in Table 1. Praefcke et al. [18] have examined a number of other bone-shaped hexaalkoxy compounds (4-7), and their transition temperatures are given in Table 2. The nematic phases exhibited by these compounds have not yet been investigated.
RO
OR
% &+% c”c‘cH
~
0
c
*
H
5
II
I
0, ”0 oI‘u\o
H5C20
2
ClOH2l
its nematic phase [22] and the transition temperatures of this complex are as shown Cr (168.5 N) 186.6 I
(1 a)
The experimental evidence put forward for the existence of the N, phase in this complex will be described later. Praefcke et al. [ 18,231synthesized a novel series of 2,3,4-trialkoxycinnamic acids (compound 3) and, as expected, these car-
They also synthesized and examined 1241 a number of multiethyne discotic nematogens with the hope of obtaining an N, phase. The pentakis [(4-pentylphenyl)ethynyllphenyl ethers (compound 8), which are spoon-shaped, do exhibit nematic phases and the transition temperatures together
936
XV
Biaxial Nematic Liquid Crystals
Table 1. Transition temperaturesa and enthalpiesb of a series of truns-2,3,4-trialkoxycinnamicacids (compound 3)" (from [IS]). R
C
I
Compound number
R
C
N
4a
203.7 52.2
(103.4)d
(67.9) 0.8
4b
95.8 43.8
(49.1)
51.2 14.I
59.6 0.9
4c
62.7 39.2
(33.6)
54.2 48.7
(50.9) 0.5
5a
232.0 50.7
292.1 1.2
5b
145.4 39.3
154.4 1.3
6a
225.2 57.3
268.1 1.4
6b
150.0 41.1
( 144.2)d
6c
123.8 41.9
(116.3)d
7a
204.1 55.4
7b
74.3 63.9
CH3
173.2 32.2
C4H9
75.6 18.6
ca,, C8H17
Nb
Table 2. Transition temperatures a and enthalpies of hexaethers (compounds 4-7)" (from [ 181).
Transition temperature in "C (top figure). Enthalpy in kJ mo1-I (bottom figure). " Heating rate 1 K/min; C:crystalline phase, N,: biaxial nematic phase, I: isotropic phase, the temperature in parentheses are for a monotropic transition. a
with their enthalpies are summarized in Table 3.
I
Transition temperature in "C (top figure). Enthalpy in kJ mol-' (bottom figure). C: crystalline phase, N: nematic phase, and I: isotropic phase. Monotropic transitions determined by hot stage polarizing microscopy; enthalpies could not be measured. a
The a,w-bis[penta(4-pentylphenylethynyl)phenoxy]alkanes (compound 9), which are bone-shaped were obtained in reasonably good yields from palladium-catalyzed
coupling reactions [ 181. The transition temperatures of these twin ethers are summarized in Table 4. A plot of the transition tem-
5
Characterization Methods
937
Table 3. Transition temperatures a and enthalpies' of pentakis [(4-pentylphenyl)ethynyl]phenyl ethers (compound 8)' (from 1181). R
C
H
86.4 42.9
109.9 0.3
CH=CH,
76.7 36.2
101.4 0.2
ND,b
I
' Transition temperature in "C (top figure). I,Enthalpy in kJ mol-' (bottom figure). Heating rate 5 K/min, C: crystalline phase,
ND,h:
discotic nematic biaxial phase, I: isotropic phase.
loo
u 8
10
9
11
12
n
'
Table 4. Transition temperatures a and enthalpies of a series of a,o-bis[penta(4-pentylphenylethynyl) phenoxylalkanes (compound 9)' (from [ 181).
8
127.0
127.8
61.8
(...)*
9
131.0 67.5
(112.5)
10
129.1 57.0
153.5 0.3
11
118.2 62.2
140.8 0.3
12
121.4 67.9
155.0 0.4
(...)"
Transition temperatures in "C (top figure). Enthalpy in kJ mol-' (bottom figure). C: crystalline phase, N D , b : discotic nematic biaxial fhase, I: isotropic phase. The enthalpy value for this transition could not be determined from the melting peak. This monotropic transition could only be determined by hot stage polarizing microscopy (heating rate 10 Klmin). a
peratures of these twin ethers is shown in Fig. 3, which depicts the interesting oddeven effect of the clearing temperatures versus the number of methylene units between the disk-shaped pentayne head units.
Figure 3. Plot of the transition temperatures (y-axis) versus the number of methylene units (x-axis) in the alkyl bridge of the twin ethers (compound 9). C: crystal; ND,b: discotic nematic biaxial; I: isotropic; x: melting points; 0 : clearing points (from [18]).
5 Characterization Methods 5.1 Optical Studies The uniaxial nematic (Nu)phase usually exhibits a schlieren texture with two and four point singularities (two and four brushes). However, the texture exhibited by complex 2, shown in Fig. 4, often consists of [S] = 1/2 (two-brush) disclinations similar to that exhibited by a smectic C phase. Also, this complex showed zig-zag disclinations [24] occasionally, and since this type of disclination has been observed in Nu phases [25] it cannot be regarded as conclusive evidence for biaxiality. It is also possible that an entanglement of disclination lines may lead to topological rigidity, as discussed by Toulouse [26] and others [27-291. These possibilities have not been examined experimentally. Hence optical textures that
938
XV
Biaxial Nematic Liquid Crystals
Figure 4. Schlieren texture of the nematic phase of copper(I1) complex 2 (from [22b]).
have been reported for nematic phases so far do not seem to give direct evidence of biaxiality. One of the most important and useful techniques for confirming the existence of an N, phase is conoscopy. From a good interference pattern, it is possible to establish whether or not a system is biaxial. The evidence put forward using conoscopic observations for the compounds discussed in the previous section is given below. A thick sample (-100 pm) of complex 2 aligned homeotropically using the combined effect of a silane coating and a 3 kHz AC (alternating current) electric field was used for conoscopic observations. Figure 5 a shows the conoscopic pattern for the pure material about 1.5 "C below the clearing point. An addition of a very small quantity of the uniaxial nematogen 4"-n-pentyl-4cyano-p-terphenyl (5CT) to this complex seems to produce an optically positive Nu phase between the isotropic and N,phases. Figures 5 b and c show the interference patterns for a mixture at two different temperatures. Figure 6 shows the phase diagram for the binary mixture for a concentration
range of 0-1% of 5CT. As can be seen, the temperature range of the Nu phase increases with increasing concentration of 5CT. These delicate experiments suggest a weak biaxiality of the nematic phase of complex 2. In the case of 2,3,4-trialkoxycinnamic acids (compound 3) [18,23], the conoscopic observations were made on much thinner samples (-20 pm). These acids, sandwiched between two glass plates, the inside surfaces of which were coated with transparent tin oxide, were aligned homeotropically using an electric field. The interference pattern obtained under these conditions is shown in Fig. 7. The position of the dark brushes is independent of the temperature. Rotation of the sample around 90" makes the brushes coalesce and reappear in the perpendicular direction, as shown. Perhaps, among all the observations made so far, those conoscopic investigations carried out on (1) nonyl pentakis[(4pentylphenyl)ethynyl]phenyl ether (compound 8) R = H and (2) 1,12-bis(pentakis [(4-pentylphenyl)ethynyI]phenyloxy)dodecane (compound U), n = 12 are the most
5
939
Characterization Methods
172
ISOTROPIC
BlAXlAL NEMATIC 160 0
'\
0.5 X ( w t . % of 5CT in complex
1.0
I)
Figure 6 . Partial phase diagram of the binary mixture of complex2 and 5CT in the concentration range 0-1% 5CT, showing the I-Nu and Nu-N, phase boundaries. The dashed portion of the curve represents an extrapolation (from [22 b]).
Figure 5. Conoscopic figures: (a) biaxial nematic phase of pure complex 2 at T ~ - ~ - T1.5"c; = (b) uniaxial nematic phase of a mixture of 99.8% complex 2 +0.2% 5CT at TN_~-T=0.7"C;(c) biaxial nematic phase of the same mixture at TN-I-T=5.5"C. Sample thickness -125 pm (from [22a]).
interesting cases [24]. In these two cases, a homogeneously aligned sample (23 pm thick) between two glass plates was used. The interference pattern obtained in the nematic phase for the bis ether (compound 9), n = 12, is shown in Fig. 8. The isogyres form a cross when the trace of the optic axial plane lies parallel to the extinction position, as shown in (a). Rotating the sample by 245" breaks the cross into hyperbolic brushes which are illustrated in (b) and (c). The clear splitting into two arcs could be observed only when the glass plates were rubbed in a uniform direction. The sign of biaxiality has been shown to be negative for both the compounds using sensitive red plates. It has been pointed out that the separation of the two arcs can be enhanced by a flow of the nematic fluid in the direction of rubbing, and the pattern becomes diffuse when the schlieren texture is formed from the homogeneous texture. These two compounds represent the first examples of thermotropic optically negative N, phases [24].
940
XV
Biaxial Nematic Liquid Crystals
Figure 7. Conoscopic figures of an aligned sample of 2,3,4-trihexyloxycinnamic acid dimer (compound 3, R=n-C,H,,) in its nematic phase: on the left rotated 0"; on the right 90" (from [23]).
5.2 X-Ray Diffraction Studies
Figure 8. Interference patterns of the biaxial nematic phase of compound 9 (n= 12) at 137 "C. (a) The trace of the optic axial plane lies parallel to the polarizer or analyzer (extinction position); (b) and (c) represent the trace of the optic axial plane enclosing an angle of +450 or -450 with the polarizer position (from [24]).
The wide angle X-ray diffraction method was employed in the case of a sanidic aromatic polyamide [ 171 which exhibited a nematic phase. The characteristic feature of the X-ray pattern was the observation of two amorphous halos, which indicated the absence of rotational symmetry about the long chain axis. Based on this observation as well as the conoscopic interference pattern, it was concluded that the polyamides exhibit the Nb phase. Assuming the N, phase to be an orthorhombic fluid, X-ray diffraction methods have been used to distinguish it from the Nu phase. The X-ray diffraction patterns of the uniaxial nematic phase of 4'-n-octyloxy-4cyanobiphenyl (80CB) and that of the nematic phase exhibited by complex 2 are shown in Figs. 9 and 10, respectively. The additional pair of peaks obtained in the equatorial scan have been attributed to an orthorhombic structure and hence to the biaxiality of the nematic phase [22c]. Figure 11 shows the space filling molecular mode] of 2,3,4-trihexyloxycinnamic acid in its dimeric form and Fig. 12 shows the X-ray diffraction pattern Of a magnetically aligned sample of this acid. In Fig. 12,
6 Concluding Remarks
Figure 9. Raw microdensitometer scans of the X-ray intensity versus the diffracting angle (28) for the uniaxial nematic of 80CB at 77 oc: (a) meridional scan (parallel to H); (b) equational scan (perpendicular to H) (from [22c]).
94 1
the maxima (a) and (c) perpendicular to each other reveal the nematic character of the mesophase (also seen from the optical texture), whereas the third maximum (b) at small angles and perpendicular to the magnetic field points to peculiarities of the structure. The X-ray data, however, clearly indicates that the 2,3,4-trihexyloxycinnamic acid is dimeric in nature and has a flat boardlike (sanidic) structure, which in principle should exhibit the N, phase [18, 231.
6 Concluding Remarks
-111111111111(1111( -12 0 12 -2L -12 0 12
2L
2B(degrees)
Figure 10. Raw microdensitometer scan of the X-ray intensity versus the diffracting angle (20) for the nematic phase of complex 2 at 166.5"C: (a) meridional scan (parallel to H); (b) equatorial scan (perpendicular to H). M represents the diffraction peak from the mylar film (from [22c]).
The controversy concerning the existence of the N, phase in low molecular weight thermotropic liquid crystalline compounds is not yet resolved. However, the occurrence of the N, phase in amphiphilic and polymeric systems has been clearly established. Many compounds possessing both rod-like and disk-like structural characteristics have been synthesized exhibiting the nematic phase. All these compounds have a direct transition from the nematic to the isotropic phase, in contrast to the theoretical predic-
Figure 11. Photograph of a space filled model of the 2,3,4trihexyloxycinnamic acid dimer (compound 3, R = n-C,H,,) including three estimated molecular parameters (from 1231).
942
XV
Biaxial Nematic Liquid Crystals T. R. Taylor, J. L. Fergason, S. L. Arora, Phys. Rev. Lett. 1970,24,359. M. J. Freiser, Phys. Rev. Lett. 1970,24, 1041. C. S. Shih, R. Alben, J. Chem. Phys. 1972,57.
3055. R. Alben, Phys. Rev. Lett. 1973,30,778. J. P. Straley, Phys. Rev. A 1974,10, 1881. R. Alben, J. Chem. Phys. 1973,59,4299. R. G. Caflitsch, Z. Y. Chen, A. N. Berker, J. M. Deutch, Phys. Rev. A 1984,30,2562. Z.Y. Chen, J. M. Deutch, J. Chem. Phys. 1984,
80,2151.
Y.Rabin, W.E. McMullen, W. M. Gelbart, Mol. Cryst. Liq. Cryst. 1982,89,67. A. Stroobants, H. N. W. Lekkerkerker, J. P h y . Chem. 1984,88,3669. Y.Galerne, Mol. Cryst. Liq. Cryst. 1988, 165, 131. L. J. Yu, A. Saupe, Phys. Rev. Lett. 1980,45, Figure 12. X-ray diffraction pattern of an oriented sample of the 2,3,4-trihexyloxycinnamicacid dimer (compound 3,R=n-C,H,,) in its nematic phase. According to the scattering maxima, the periods of the density waves in angstroms are as follows: (a) 4.4, (b) 19.5,and (c) 14.0 (from [23]).
1000. S. Chandrasekhar, Mol. Cryst. Liq. Cryst. 1985, 124,1. F. Hessel, H. Finkelmann, Polym. Bull. 1986, 15, 349; F.Hessel, R. P. Herr, H. Finkelmann, Makromol. Chem. 1987,188,1597. M.Ebert, 0.H. Schonherr, J. H. Wendorff. H. Ringsdorf, P. Tschirner, Makromol. Chem., Rapid Commun. 1988,9,445. K. Praefcke. B. Kohne. B. Gundopan. D. Sine" er, D. Demus, S. Diele, G. Pelzl, U. Bakowsky, Mol. Cryst. Liq. Cryst. 1991,198,393. [19] J. Malthete, L.Liebert, A. M. Levelut, C. R. Acad. Sci. Paris 1986,303,1073. [20] J. C. Bhatt, S . S. Keast, M. E. Neubert, R. G. Petschek, Liq. Cryst. 1995,18,367. [21] S. Chandrasekhar, B.K. Sadashiva, S. Ramesha, B. S. Srikanta, Pramana J. Phys. 1986,27, L713;S.Chandrasekhar, B. K. Sadashiva, B. S. Srikanta, Mol. Cryst. Liq. Cryst. 1987,151, 93. [22] (a) S. Chandrasekhar, B. K. Sadashiva, B. R. Ratna, V. N. Raja, Pramana J. Phys. 1988,30, L491;(b) S. Chandrasekhar, B. R. Ratna, B. K. Sadashiva, V. N. Raja, Mol. Cryst. Liq. Cryst. 1988,165,123;(c) S.Chandrasekhar, V. N. Raja, B. K. Sadashiva, Mol. Cryst. Liq. Cryst. Lett. Y
tion that the N, phase should occur between two Nu phases. Moreover, the thermodynamic data reported for these transitions is also first order, though weak in some cases. Though some evidence has been shown that these compounds exhibit the Nb phase, the same is yet to proved unequivocally. As mentioned in Sec. 2, there have been a number of theoretical approaches to the N,phase and some aspects of the physical properties have also been studied. For those who are interested in these aspects of the Nb phase, a list of references [31-531 is included.
7 References [ 11 P. G.de Gennes, The Physics of Liquid Crystals,
Clarendon, Oxford 1974. [2] S.Chandrasekhar, Liquid Crystals, 2nd ed., Cambridge University Press, Cambridge 1992.
1990,7,65. [23] K. Praefcke, B.Kohne, B. Gundogan, D. Demus, S. Diele, G. Pelzl, Mol. Cryst. Liq. Cryst. Lett. 1990, 7, 27. [24] K. Praefcke, B.Kohne, D. Singer, D. Demus, G. Pelzl, S. Diele, Liq. Cryst. 1990,7,589. [25] Y. Galerne, L. Liebert, Phys. Rev. Lett. 1985,55,
2449. 1261 Y.Galerne, J. Itoua, L. Liebert, J. Phys. 1988, 49,681. [27]G.Toulouse, J. Phys. Lett. (Paris) 1977, 38, L-67;V.Poenaru, G. Toulouse, J. Phys. (Paris) 1977,8,887. [28] N. D.Mermin, Rev. Mod. Phys. 1977,51.591.
7 References 1291 H. R. Trebin, Adv. Phys. 1982, 31, 195. 1301 M. Kleman, Points, Lines and Walls,Wiley, New York 1983. [31] P. B. Vigrnan, A. I. Larkin, V. M. Filev, Sov, Phys. JETP 1976,41,944. [32] A. Saupe, P. Boonbrahm, L. J. Yu, J. Chim. Phys. 1983, 80, 7. [33] D. L. Johnson, D. Allender, R. DeHoff, C. Maze, E. Oppenheim, R. Reynolds, Phys. Rev. 1977, B16,470. [34] P. H. Keyes in Proc. Eighth Symp. on Thermo Physical Properties - Thermophysical Properties of Fluids, Vol. 1 (Ed.: V. Sengers), The American Soc. of Mech. Engineers, New York 1981, p. 419. [35] E. A. Jacobsen, J. Swift, Mol. Cryst. Liq. Cryst. 1982, 87, 29. [36] J. W. Doane in NMR of Liquid Crystals (Ed.: J. W. Emsley), Riedel, Dordrecht 1985, p. 441. [37] S. R. Sharrna, P. Palffy-Muhoray, B. Bergersen, D. A. Dunrnur, Phys. Rev. 1985, A23, 3752. [38] D. W. Allender, M. A. Lee, Mol. Cryst. Liq. Cryst. 1984, 110, 331; D. W. Allender, M. A. Lee, N. Hafiz, Mol. Cryst. Liq. Cryst. 1985,124, 45. [39] T. C. Lubensky, Mol. Cryst. Liq. Cryst. 1987, 146, 55.
943
1401 A. Saupe, J. Chem. Phys. 1981, 75,5118. [41] M. Liu, Phys. Rev. 1981, A24, 2720. [42] H. Brand, H. Pleiner, Phys. Rev. 1981, A24. 2777. [43] E. A. Jacobsen, J. Swift, Mol. Cryst. Liq. Crysf. 1981, 78, 311. [44] G. E. Volovik, E. I. Kats, Zh. Eksp. Teor. Fi:. 1981, 81, 240. [45] U. D. Kini, Mol. Cryst. Liq. Cryst. 1984, 108, 71; U. D. Kini, Mol. Cryst.Liq. C r y ~ t1984,112, . 265. [46] E. Covers, G. Vertogen, Phys. Rev. 1984, A30, 1998. [47] A. Chaure, J. Eng. Sci. 1985, 23, 797. [48] U. D. Kini, S. Chandrasekhar, Physicu 1989, 156A, 364; U. D. Kini, S. Chandrasekhar, Mol. Cryst. Liq. Cryst. 1990, 179, 27. [49] A. V. Kaznacheev, A. S. Sonin, Sov. Phys. Crjstallogr. 1988, 33, 149. [50] L. G. Fel, Sov.Phys. Crystallogr. 1989,34,737; 1990, 35, 148. [5 I] D. Monselesan, H. R. Trebin, Phys. Sfatus Solidi 1989, 155, 349. [52] S. Chandrasekhar, G. S. Ranganath, Adv. Phys. 1986,35, 507. [53] G. S . Ranganath, Curr. Sci. 1988, 57, 1 .
Handbook of Liquid Crystals Edited by D. Demus, J. Goodby, G. W. Gray, H.-W. Spiess, V. Vill Copyright 0 WILEY-VCH Verlag GmbH. 1998
Chapter XVI Charge-Transfer Systems Klaus Praefcke and Dirk Singer
1 Introduction In principle, a combination of short range repulsive forces and anisotropic dispersive interactions between mesomorphic molecules is sufficient to understand their selforganization in mesophases [l]. This basic approach is used for theoretical descriptions of mesomorphic states, especially for their simulations by molecular dynamics or Monte Carlo calculations of systems composed of idealized particles. Even on this simplified level, realistic pair potentials constitute one of the crucial problems. For real systems, the occurrence of liquid-crystalline properties is determined by the shape of the constituent mesogenic units and, as realized to an decreasing degree, also by specific interactions of various types between them. Thus, for example, (1) electrostatic ionic forces, (2) dipole/dipole interactions, ( 3 ) hydrogen bonding, and (4) charge-transfer situations are often important in supporting the formation and stability of supramolecular structures in organized systems. In this context, in recent times, charge-transfer interactions have attracted increasing interest in the field of liquid crystal research; on the one hand, due to the fascinating possibilities concerning the variation or induction of mesdmorphic selforganization, and on the other hand, connected to strategies successfully applied in materials science related to synthetic metals, organic semiconductors, molecular electronics, and nonlinear optics. Attractive
interactions between n-systems are well known and have significant impact on such diverse phenomena as the vertical basebase interaction in DNA (deoxyribonucleic acid), drug-intercalation into DNA, the packing of aromatic molecules in crystals, the tertiary structures of proteins, conformational and binding properties of polyaromatic macrocycles, complexation in some types of guest-host systems, and porphyrin aggregation [2, 31. In addition, strong n--71:interactions are crucial for the formation of sufficiently large bands of electronic states required in charge-transfer complex based organic conductors [4], and might also control the molecular self-organization of such systems. However, the basic contributions to such interactions, usually denoted as n-n, charge-transfer (CT), or electron donor-acceptor (EDA) interactions, are not very well understood, despite the fact that they are often used to describe and explain the aggregation of molecules with z-electron systems in solution or even in higher condensed states [2, 31. The self-organization of molecules in such ordered states is in general controlled by several types of interaction; some of the basic contributions are (1) van-der-Waals forces, (2) electrostatic interactions due to static distributions in the molecules, and ( 3 ) charge-transfer interactions [ 5 ] ; the latter ones might be described as a stabilization due to mixing the ground state of two interacting molecules A and B with a charge-separated excited state A+B-.
946
XVI
Charge-Transfer Systems
In systems, composed of good electron donors and good electron acceptors, the formation of charge-transfer complexes is characterized either by charge-transfer transitions in the UV/VIS absorption spectra or at least by a broadening of the absorption. However, relatively little is known about the relative significance of electrostatic versus charge-transfer interactions for the molecular self-organization in condensed phases [2, 3e]. Especially for systems without spectroscopic evidence for the presence of strong CT-interactions, an approach based on electrostatic forces involving the out-ofplane 7c-electron density seems to be sufficient to explain geometrical requirements for interactions between aromatic molecules [2]. In the following, we will present and discuss examples where the liquid-crystalline self-organization is significantly governed by 7c-7c interactions. However, the relative strength of the charge-transfer versus electrostatic dipole-dipole or quadrupolequadrupole interactions is generally not known. Thus we use the synonyms chargetransfer or electron donor acceptor interactions or complexes here to denote and classify those systems that show (1) spectroscopic evidence, such as, for example, a CTband or a broadening in the UV/VIS absorption spectrum or (2) at least a significantly deeper color than the neat components as an indication of strong 7c-7c interactions in a kind of “chemists approach”. In addition, CT interactions are especially relevant for binary or multi-component systems in the context of mesophase induction or variation [6]. These terms denominate a significant change in the mesomorphic self-aggregation in the mixture compared to those of its constituting components, for example, the occurrence of a new mesophase; we also include here the stabilization of a mesophase of one of the com-
pounds. However, these expressions do not cover the observation of a “latent” mesophase due to eutetic melting point depression. It should also be mentioned here that the terms “liquid crystalline mixed phases” [6, 7 a] of “flussigkristalline Mischphasen” [7b], sometimes used to designate mesomorphism in multi-component systems, are misleading as they imply a mixture of phases, but in fact we are dealing with phases of mixtures, often with specific molecular aggregation.
2 Calamitic Systems Charge-transfer interactions in liquid-crystalline systems first gained interest more than twenty years ago in connection with the report of enhanced thermal stability of the nematic phase in mixtures of N-(4-methoxybenzylidene)-4-butylaniline (MBBA) with 4-cyano-4’-pentylbiphenyl [CPB (more often referred to as 5CB)] by Park et al. [8 a]. The phase diagram of this binary system shows two eutectics and a stabilized solid state, later found to be a smectic B phase (see [Sc]). The maxima of the “melting” as well as of the clearing temperature are observed near to the equimolar composition. This particular phase behavior has been explained in terms of the formation of a molecular 1:1 electron donor (MBBA)/acceptor (CPB) complex due to charge-transfer interactions, evident by a broad absorption band around 550 nm. The molecular complexing is also thought to be responsible for the increased dielectric anisotropy observed for the equimolar MBBA/CPB mixture [8b]. A similar stabilization of the nematic phase of CPB has been found by doping this nematogen with 4-aminobiphenyl, a nonmesogenic electron donor compound [9] (see Table 1).
2
Calamitic Systems
947
Table 1. Examples of electron donor pairs with rod-like mesogenic units exhibiting stabilized or induced mesophases. Stabilized andlor induced mesophases
Ref.
N
[91
N, SmA
[I01
02N*iO*OR
N, SmA
[14a, e]
RO--@COO-@NO2
N, SmA
[ 14al
N, SmA
[14a, bl
N, SmA
u4bi
N, SmB, SmE
[14c]
Electron acceptor
Electron donor
X = CH or N, Y = CH, N, or NO R, R’ or R’, R = R, R’ or R’, R = NO, and alkyl or alkyloxy N(CH,), and alkyl or alkyloxy or R = R’= NO,
R, R’ or R’, R = NO2 and alkyloxy
N(CH,), and alkyloxy
5) R
H
N
~
N
H
R
R = alkyl
0 RO-@-!-O, N=C-@CN
I C”3 CN
R 0 *COO& H @ 17c8x-@
CN CN
X = 0 or NH
948
XVI Charge-Transfer Systems
Matsunaga et al. have studied the phase diagrams of a wide variety of electron donor acceptor systems in which both components are more or less rod-shaped and represent at least “potential” mesogens (see Table 1 and [lo]). Most of these compounds under evaluation were composed of two phenyl rings connected by a rigid, twomembered double bond spacer; the components designed as electron acceptors usually carried a nitro or cyano group, while the electron donors had dimethylamino substituents (see Table 1). The phase diagrams of these binary EDA systems are characterized by congruent melting points of a molecular complex, usually near to the 1:1 composition. Furthermore, most systems show stabilized/induced nematic and/or smectic A phases. The molecular complexing in these binary mixtures is accompanied by chargetransfer bands in the visible region, or at least a broadening of the spectrum, supporting the model of strong x-x interactions of the charge-transfer type. In another study of the phase behavior of Schiff base donor compounds, with 4,4’dinitro- or 4,4’-dicyanobiphenyl as electron acceptors, or of N,N,N’,N’-tetramethylbenzidine as an electron donor with suitable Schiff base acceptors, similar mesophase inductions/stabilizations are reported [ 111 (see Table 1). Even p-substituted derivatives of nitrobenzene and benzonitrile, respectively, are able to induce or stabilize smectic mesophases (SmA and/or SmB) in mixtures with, for example, the mesogen N(4-propyloxybenzylidene)-4-hexylaniline [12]. However, in these systems a dipolar mechanism for the enhanced molecular aggregation seems more likely. Iida examined and explained the induced/stabilized nematic phase in the MBBAICPB system and those of donor/ acceptor mixtures based on N-(4-X-benzyIidene)-4-Y-aniline derivatives (see Table 1)
by a modified solid solution model [ 131; the charge-transfer interaction energies between donor and acceptor in the nematic state were estimated as 0.16 kJ/mol for the MBBA/CPB complex [13a] and as 0.380.52 kJ/mol for the second type of EDA mixtures [ 13bl. Clear evidence of charge-transfer interactions in binary mesomorphic mixtures was provided by the homologous series of 4,4’-bis(alkylamino)-biphenyls studied by the former Halle group (Table 1). These mesogens are excellent electron donors and form colored EDA complexes with a number of liquid-crystalline compounds as electron acceptors, usually accompanied by the stabilization or induction of SmA phases [14a-c, el, but for certain combinations even induced SmE and B phases are ob. et al. have shown that served [ 1 4 ~ 1Sharma the charge-transfer bands in the visible region of the absorption spectrum show negative dichroism, thus the electron transition moments responsible for the absorption are oriented perpendicular to the main axis of the molecules [14a]. Demus et al. estimated the charge-transfer complex stability constant and the formation enthalpy of 4,4’bis(butylamino)-biphenyl/4-(2,2-dicyanoethenyl)phenyl-4’-nonyloxy-benzoate via the temperature dependence of the CT band, and of the strongly negative excess volumina in this system [ 151. The complex formation enthalpies were found to range from -58 to -219 kJ/mol depending on the phase type (SmA, N, or Iso). These values are of the same order of magnitude as those of nonmesogenic systems in the isotropic or solid states, respectively 141. In addition, complex stability constants of liquid crystalline electron donor acceptor pairs of dialkylazoxybenzene/PCB, determined by a dielectric method in cyclohexane solution [16a], are also comparable to those reported by Demus et al. [15]. Density measure-
2
ments of these binary systems confirm the formation of molecular 1: 1 complexes [16b]. It is noteworthy that the EDA interaction energies here are significantly higher than those reported by Iida [ 131, but he assumed a complete dissociation of the complex at the phase transition N-Iso. A model for induced smectic phases in binary mixtures of liquid crystals due to molecular complexing, especially due to charge-transfer interactions, has been proposed some time ago [ 171.Within the mean field approximation, the influence of complex formation on the N-SmA transition temperature and the types of phase diagrams are discussed. In this context it seems noteworthy that in mixtures of a nematogen carrying a lateral dicyanoethenyl substituent with electron donor nematogens, no phase induction or stablization occurs, despite the color change on mixing the components, indicating the formation of a CT complex [14d]. Thus the relative orientations of the donor and the acceptor moieties have to fulfill geometric requirements supporting a parallel arrangement of the molecular main axes for the formation of induced mesophases. For a mixture of 4,4’-dinonanoylbiphenyl with N,N’-didecyl-4,4’-diaminobiphenyl, showing an induced smectic A phase and a slightly positive deviation of the clearing temperature from ideal linear behavior, an interesting synergistic combination of EDA interactions and hydrogen bridging has been proposed [ 181. It should be emphasized here that electron donor-acceptor combinations are not the only ones giving rise to mesophase induction or variation. Similar phase behavior is also known for binary systems composed of suitable proton donor-acceptor pairs due to hydrogen bridging [19, 201, or for combining components of complementary molecular shapes as, for example, in the so-called “filled smectics” [21]. In
Calamitic Systems
949
mixtures of usually unpolar smectic mesogens, nematic phases are often induced if the aggregation to higher ordered layered structures is hindered by, for example, a significant difference in the molecular length [221. Another common type of phase inductionhariation is observed in mixtures of a terminal polar compound, usually a nematogen carrying a nitro or a cyano group, and a nonpolar one [23]. In such systems induced SmA-phases are often formed, but they might also exhibit higher ordered smectic phases, for example SmB, SmC, SmE, and SmI or form nematic ones, sometimes even reentrant. The phase behavior of these binary systems is often explained in terms of dipole-dipole or dipole-induced dipole interactions, partially in analogy to the specific mesophase structures observed for highly polar neat mesogens [6, 241. According to Prost, the mesomorphic properties of such dipolar mesogens are a result of the competition between the incommensurable length of the molecules and their dimers [24e]. The electron acceptors discussed so far resemble these terminal polar compounds in their structure, or are even identical. Cladis, for example, has studied the phase behavior of mixtures of butyloxybenzylidene octylaniline with cyanooctyloxybiphenyl showing stabilized smectic phases of the A- and B-type, as well as an induced SmE phase; the results are summarized in a Landau description [23g]. However, the phase behavior of this particular system, strongly related to that studied by Park et al. [8], is discussed in terms of dipolar pair formation. Furthermore, the phase behavior and macroscopic properties, e.g., densities and rotational viscosities, in mixtures of polar 4-cyano derivatives of biphenyl with apolar azoxy compounds, were found to differ significantly from those comprising the relat-
950
XVI
Charge-Transfer Systems
ed cyanophenylcyclohexane derivatives, but in both type of systems the formation of mixed associates occur [23i]. These examples raise the question of whether the z-z interactions of the chargetransfer type are the main cause for the enhanced aggregation in mesomorphic EDA systems or if dipole-dipole or dipole-induced dipole interactions are more important. If the latter is true, charge-transfer contributions, evident by a related absorption band, would just constitute a “side effect” of a suitable molecular arrangement as a result of dipolar forces. In this context it should be noted that up to now neither the mesomorphic molecular organization nor the dynamic behavior of the mesogens in calamitic EDA-systems have been studied systematically. Thus it is not clear if these systems show any peculiarities compared to those of binary mixtures of compounds polar-polar or polar-nonpolar in nature. Finally, an anisotropic, two-dimensional solvation model developed by Szabon for the classification of induced smectic phases, based only on the requirement of a quasihexagonal close-packing and a sufficient difference of the rigid and the flexible aliphatic segments of the components, should be mentioned because no specific intermolecular interactions are required within its framework [25]. Based on the scarce experimental data available, a final decision concerning the relative significance of charge-transfer versus dipolar interactions for the self-organization in these EDA-systems is not possible yet. However, recently it could be shown that 2,4,7-trinitrofluorene (TNF), an electron acceptor with only a small dipole moment and well known for inducing/stabilizing columnar aggregates of disc-shaped molecules due to charge-transfer interactions (see next section), can also induce or stabilize smectic mesophases. In contrast
to the electron acceptors discussed so far, this one is much stronger and its molecular shape is far from that of even a “potential” calamitic mesogen. However, with bromobridged dinuclear palladium organyls, the formation of deeply colored CT complexes and a significant stabilization of the SmA phase of these centrally fused twin-like mesogens has been observed [26]. Furthermore, the bis(2,3,4-trisdodecyloxyphenyl)substituted Schiff base of 4,4’-diaminostilbene, a nematogen, exhibits an induced SmA phase with TNF, showing the maximum clearing temperature for the 1: 2 donor-acceptor ratio (see Fig. 1) [20]. With 2-(2,4,5,7-tetranitro-9-fluorenylideneaminooxy)-propionic acid (TAPA) as electron acceptor, the induction of a cholesteric nematic phase is observed instead, with a maximum clearing temperature around 45 mol% (-)-TAPA [20]. Further examples of a TNF-induced mesophase is provided by the SmA phase of various polyether macrocycles, a dimesogenic twin, and of numerous calamitic heterocycles [27]. MBBA, on the other hand, forms a red CT complex with TNF, but shows only a strong depression of the N-Iso transition temperature [26]. A similar behavior is observed for chloranil EDA complexes of 1,4-bis(phenylethiny1)benzene mesogens [28 a] and for tetracyanoethylene (TCNE) in 4-cyano-4’pentylbiphenyl (PCB) [28b]. In the latter system, TCNE was found to lower the isotropization temperature much more strikingly than, for instance, 2,3-dimethylbutane which shows no evidence of complex formation with PCB, suggesting its peculiar effect [28b]. Furthermore, despite the electron acceptor character of PCB itself, the charge-transfer interaction is clearly evident from related absorption bands; the complex formation enthalpy was determined as 14.98 kJ/mol. The unusually high order parameters found for some anthraqui-
2
95 1
Calamitic Systems NO2
02N
X
9 N
RO
RO
e
O
R
N
02N
d& 0
r
NO2 N
O
II
I
NO2
o -*CHC-OH I
TNF
C12H2!i
OR
02N
(-I-TAPA
CH3
T
0
10 20 30 40
50
60
70 80 90 100
mol% TNF
0
10 20 30 40
50 60
70 80
90
100
mol X (-I-TAPA
Figure 1. Phase behavior of a nematic bisimine donor with TNF and (-)-TAPA, respectively, as electron acceptors [20]; Cr: crystalline, SmA: smectic A, N and N*: nematic and cholesteric nematic phase, respectively, and Iso: isotropic liquid.
none dyes in the nematic phase of cyanophenylcyclohexane derivatives have also been attributed to CT interactions [29]. The examples discussed above indicate that the formation of induced or stabilized mesophases in binary systems composed of calamitic compounds is usually the result of a complex combination of intermolecular interactions and packing effects. Therefore separation and analysis of the different contributions is difficult as long as none of them is clearly dominating the mesomorphic selforganization of the molecules. Especially generalizations concerning the relative significance of electron donor-acceptor properties versus permanent dipole moments of components in such binary systems, for the formation of inducedhtabilized mesophases, should be viewed carefully. However, taking into account the (so far) small number of calamitic compounds investigated for charge-transfer complexation and mesophase inductionhariation, with strong non-
calamitic electron acceptors like TNF, TAPA, or 7,7,8,8-tetracyano- 1,4-quinodimethane (TCNQ), a great multitude of similar binary systems can be expected to exhibit charge-transfer induced mesophases. The induction or stabilization of liquidcrystalline phases in binary CT systems, where the donor and the acceptor compounds are rod-shaped, has been utilized in the design of polymers to enhance their mesomorphic properties; the combination of mesomorphic properties and those of polymeric systems, e.g., glassy behavior, has attracted interest due to possible applications, for example, in the field of nonlinear optics. Side-chain copolymers with a polyester or polyacrylate backbone containing 4-methoxyazobenzene and 4-cyanoazobenzene units, for example, show the formation of an SmA phase with an enhanced mesomorphic temperature range [30]. Similar behavior is observed for copolystyrene derivatives carrying electron-rich donor units (4-methoxy-
952
XVI
Charge-Transfer Systems
azobenzene), as well as for electron acceptor moieties (4-nitro- or 4-cyanoazobenzene) via flexible spacers [31]. A series of similar copolymethacrylates with ethylene spacers and carbonate linkage only shows a negative deviation from the clearing temperature of the copolymers compared to the two homopolymers [32a]. Co-methacrylates containing both an electron-donating carbazol group and an electron-accepting nitrophenyl or cyanophenyl unit show stabilized or induced SmA phases [32b], most probably due to interactions between these functions, to their longer spacer and the ether linkage to the mesogenic side groups; this tendency on thermal behavior was not seen in a related copolymer having carbazol and (4’-methoxybenzy1idene)anilin groups both of which are of electron-donating nature. The mesomorphic self-organization in such polymeric systems is not only governed by the interaction profiles of the mesogenic groups, but also by backbone-related structural features leading, for example, to a variety of variations of the SmA type of mesophase [33]. Asymmetric dimeric liquid crystals with charge-transfer groups constitute good models for the mesomorphic self-organization of the related polymers discussed above. Their mesophase structures have been studied in some detail [34]. In this context, the formation of intercalated smectic phases of the A, C, and I type is especially noteworthy [34c]. However, the nature of the specific interactions between the unlike mesogenic groups, as well as the conformation of the spacer, are still to be explained [34d]. A nematic, main-chain polycarbonate polymer comprising azobenzene mesogenic units shows an induced SmA phase in mixture with a nematic nitroazobenzene dimeric liquid crystal [35]; a behavior resembling that of related monomeric mesogenic mixtures. A more specific example of
charge-transfer contributions to the mesomorphic self-aggregation of a main chain polymer has been observed for rigid polyesters based on 1,4-dialkylesters of pyromellitic acid (1,2,4,5benzenetetracarboxylic acid) copolymerized with 4,4’-biphen01 [36a]. These rigid-rod polyesters are dimesomorph; the layered, low temperature mesophase shows a high degree of ordering attributed to CT interactions between the biphenyl donor units and the pyromellitic acid ester acceptor moiety [36b]. Finally, doping surface-stabilized ferroelectric liquid crystals with charge-transfer complexes, for example, tetramethyltetrathiafulvalene/octadecyltetracyano1,4-quinodimethane, has been utilized to improve their bistability [37]: Ions from the CT complex form an internal electric field in reverse to the applied pulse. By applying this phenomenon, a ferroelectric liquid crystal cell with perfect bistability and inverted memory characteristics was designed [37]. Ono and Nakanowatari developed a method for the determination of the internal electric field and studied a TCNQ doped ferroelectric liquid crystal [38]. In another approach towards enhanced switching properties of ferroelectric liquid crystals, a conductive overlayer of the tetrathiafulvalene/TCNQ CT complex on the SiO layers was used to shorten the response time [39].
3 Noncalamitic Systems In the area of noncalamitic liquid crystals, especially those with a planar, sheet-like or discoid structure, the impact of chargetransfer interactions on the mesomorphic self-organization is much higher due to the favoritism of a face-to-face type, intercalat-
953
3 Noncalamitic Systems
ed stacking in columnar supramolecular structures. However, surprisingly, the importance of CT interactions for the design of such mesomorphic systems has only been recognized in very recent years. In this context, it is interesting that the first examples of mesomorphic CT complexes of discoid or sheet-like donor compounds and a strong electron acceptor [7,7,8,8-tetracyano- 1,4-quinodimethane (TCNQ)] constitute an exception from the usually observed behavior. These complexes of bi-4-H pyran or bi-4-H thiopyran derivatives carrying four alkyl- or alkyloxyphenyl substituents with TCNQ (see Fig. 2), have been prepared as organic conductors [40]. The neat alkylphenyl derivatives in the bi-4-H pyran series and alkyl- or alkyloxyphenyl-substituted bi-4-H thiopyran donors are mesomorphic [40], but the structures of their mesophases are still not clear. The equimolar charge-transfer complex of the dodecyloxyphenyl substituted bi-4-H thiopyran with tetracyano-1,4-quinodime-
thane (TCNQ) received special interest due to combining of conductivity with mesomorphic properties (Fig. 2). In the crystalline state the conductivity was found to be 0=0.7 Q-' cm-' for a single crystal and Q-' cm-' for a powder sample [40d]. The latter order of magnitude is also found for the conductivity in the mesophase range [40e, f]. Based on X-ray diffraction studies, the mesophase of this CT complex has been identified as lamello-columnar, combining features of a smectic mesophase with those of a columnar one [~OC]. Up to now only a few examples [41] of lamellar mesophases, first postulated by Billard 1421, have been observed. Two variations are known so far: The L, phase, without any intralamellar order of the sheet-like molecules, and the lamello-columnar Lcol phase with columnar stacking within each layer. A sketch of the donor self-organization of the dodecyloxyphenyl substituted bi-4-H thiopyran with TCNQ is depicted in Fig. 2. The TCNQ acceptor molecules probably fill the space
xR P +
R
R
R = 0 ~ 1 2 ~ 2 5'R
TCNQ Figure 2. Schematic representation [4Oc] of the tetra(4-dodecyloxyphenyl) dithiapyranylidene donor self-organizatjon in the lamello-columnar (L& mesophase of an equimolar CT complex with TCNQ; the TCNQ molecules are not shown in this drawing.
954
XVI
Charge-Transfer Systems
More significant with respect to the impact of charge-transfer interactions on the mesomorphic self-organization of discoid moieties, however, is a look at electron-donor acceptor systems composed of low-molecular-weight compounds. 2,3,6,7,10,11Hexaethers of triphenylene, for example, forming columnar hexagonal mesophases with intracolumnar stacking order (Col,,) [48]show a stabilization of the existing columnar structures and an extension of the mesomorphic range on addition of TNF due to CT interactions [49]. The clearing temperature steadily increases with the TNF 2,2’,6,6’-Tetra(4-d0decylphenyl)-A~~~’-concentration and maximum stabilization is observed for the 6 :4 electron donor accepbipyran also forms a charge-transfer comtor ratio. The TNF-CT complexes of the triplex with TCNQ as well as cation-radical phenylene hexaethers with this composition charge-transfer salts, the first ones of disclike shape showing columnar mesomorphic behave like a single compound, showing a sharp transition from the Colh,, to the isobehavior over a broad temperature range [44]. This shows liquid-crystalline propertropic phase; higher or lower acceptor concentrations in such mixtures lead to biphasties, but no evaluation of the mesophase structure has been reported up to now. The ic ranges. The same type of mesophase stabilization is also observed in mixtures of TCNQ charge-transfer salt shows only a 2,3,6,7,10,11-hexaester derivatives of trivery small electric conductivity due to the complete charge transfer between donor and phenylene with TNF, while analogous nonliquid crystalline hexaethers with polyethyacceptor. More recent is the idea that induced columnar stacking interactions could be ’ The nomenclature of columnar mesophases formed beneficial for the design of mesomorphic by disc-like molecules, up to now (falsly) called dissystems composed of disc- or sheet-shaped cotic, and their short notation with D and indices demolecules. This approach was first realized scribing the intercolumnar arrangement and the intracolumnar stacking clearly give more and more rise to by Ringsdorf et al. in (1) charge-transfer inmisconceptions. Since the utilization of this “D-noduced nematic columnar (NCJ phases exmenclature”, a wide variety of molecular shapes, ofhibited by mixtures of amorphous sideten supported by specific types of intermolecular interactions, e.g., hydrogen bonding, have been found chain polymers carrying 2,3,6,7,10,11to result in the formation of columnar mesophases. hexasubstituted triphenylene moieties with Therefore, a more systematic nomenclature based on TNF and ( 2 )TNF-induced columnar hexagthe type of mesomorphic self-organization, e.g., lamellar or columnar, was strongly desirable and has onal (colh,) phases of related main-chain also been introduced recently [45] which quickly seems polymers [46]. This type of doping also to gain growing acceptance. Although questions in this makes possible the formation of one-phase context are still under discussion we use here the novel type of notation for columnar (Col) situations blends of otherwise incompatible polymers in mesophases: N,,, (nematic columnar), ND (nema[46] and opens new ways in molecular entic discotic), Col, (columnar with, e.g., x=ob for -obgineering of liquid-crystalline polymers lique, -tilted etc.), L,,, (lamello-columnar), L, (lamellar discotic). [471. between two stacks within a layer and may be in segregated stacks of their own. However, also an intercalated columnar structure common in mesophases of charge-transfer complexes of disc-like electron donors with strong electron acceptors cannot be ruled out, taking into account the observation of such an alternate donor-acceptor stacking in the crystalline state of the semiconductor 2,2’,6,6’-tetramethyl-A434‘-bisthiapyran/ TCNQ [43] and the X-ray diffraction pattern of charge-transfer based hexagonal columnar phases of TNF complexes of different discoid donor compounds discussed below.
’
3 Noncalamitic Systems
leneoxy chains show the induction of a Col,, phase instead, even with TNF contents of only about 15 mol% [49]. X-ray studies prove the formation of mixed stacks of donor and acceptor molecules with a regular, alternating packing in the induced or stabilized columnar mesophases [49]. Very strong support for the nonstatistical, alternating donor-acceptor stacking results from the observation of X-ray diffraction maxima corresponding to a d-spacing of 6.92 A (0.692 nm) in the mesophase of a 1: 1 2,3,6,7,10,1 l-hexakispentyloxytriphenylene/TNF complex, besides the usual intracolumnar reflex at 3.46 A (0.346 nm) [49a]. The fact that this diffraction is observed emphasizes the dissimilarity in the electron density of the donor and acceptor molecules. Not only 2,4,7-trinitro-9-fluorenone (TNF) but also other electron acceptors, for example, TCNQ, 172,4,5-tetracyanobenzene, the enantiomers of 2-(2,4,5,7-tetranitro-9-fluorenylideneaminooxy)-propionic acid (TAPA) or 2,4,5,7-tetranitro-9-fluorenone, and derivatives of the latter one, are effective in this type of mesophase induction [50]. In particular, a TNF derivative carrying two hexadecanoyl side chains (2,4,7-trinitrofluorenylidene-9-malonic acid bis(hexadecy1) ester) is remarkable because of the induction of a nematic columnar type (Nco,) mesophase with hexakispen-
NCol
955
tyloxytriphenylene [5 I]. The formation of an electron donor-acceptor complex here also stabilizes the columnar aggregates, but at the same time the self-organization of these columns in a two-dimensional lattice is prevented by the long chains of the acceptor molecules. With 2,3,6,7,10,11-hexakishexadecyloxytriphenylene, however, a Col,, phase is induced by this electron acceptor because its two flexible chains are well within the aliphatic periphery of the CT-based columnar aggregates. A schematic representation of the columnar electron donor-acceptor aggregates and their selforganization in the two-dimensional lattice of a Col,, phase or their nematic arrangement (Ncol phase) is sketched in Fig. 3. While the hexagonal arrangement of such charge-transfer based columns in the absence of intercolumnar restraints is the most common one, there is also the TNF complex of a triphenylene main chain polymer, for example, forming a rectangular two-dimensional lattice (Col,, phase) [49b]. The structural features of charge-transfer based mesophases, especially those of polymeric as well as low molecular weight triphenylene derivatives, as a function of the donor-acceptor ratio, have been studied by X-ray diffraction in some detail [49]. Moreover, using the model proposed in the eighties [ 171to describe the induction of smectic phases as a result of complex for-
Colho
Figure 3. Schematic representation of thermotropic mesophases based on charge-transfer induced formation of intercalated donor-acceptor stacks; Nc,,: nematic columnar and Col,,,: columnar hexagonal ordered mesophase (see footnote 1 in Sec. 3 commenting on the nomenclature of such columnar mesophases). a : electron donor, G 2 : electron acceptor.
956
XVI
Charge-Transfer Systems
mation, Wendorff et al. could account qualitatively for the experimental findings concerning charge-transfer induced mesophases of discoid compounds [49b]. An interesting type of mesomorphic selforganization based on CT interactions is observed for charge-transfer twins consisting of triphenylene 2,3,6,7,10,1l-hexaether [52a] or radial multiyne [52b] units (as discoid electron donors) each covalently linked via flexible spacers to trinitrofluorenone acceptor moieties. The structural model proposed for their mesophase assumes a columnar arrangement of the molecules in such a way that mixed stacks of donor and acceptor units are formed. The columns containing triphenylene parts are arranged in an orthorhombic two-dimensional lattice; neighboring columns are connected along specific directions via the flexible spacers, giving rise to highly anisotropic properties [52a]. In addition, these columnar aggregates and their self-organization in ordered monolayer films have been studied by atomic force microscopy [52c]. A possible borderline case with respect to these two latter examples of twin CT complexes [52a,b] could be a series of sevenfold substituted triphenylene derivatives, very recently synthesized and studied for the first time [52d], carrying one acceptor function (e.g., NO,) in a so-called a-position and six ethergroups at the six outer (P-)positions of that aromatic ring system. These new molecules are sterically heavily congested and are, therefore, helically deformed, but show strong stabilization of their columnar phase behavior in comparison to their parent hexaether without an acceptor attached to it by only one covalent bond [52d]. Main-chain polyesters based on 2,4,7trinitrofluorenylidene malonate are also effective electron acceptors able to stabilize the columnar hexagonal mesophase of 2,3,6,7,10,1l-hexakispentyloxytripheny-
lene [53a]. As an interesting note, a polysiloxane with 2,4,7-trinitrofluorenylidenemalonate side groups forms a nematic mesophase on its own, most likely due to microphase separation between the polysiloxane backbone and the side groups [53b]. The molecular motion in the chargetransfer induced/stabilized columnar hexagonal mesophases of 2,3,6,7,10,1l-hexakisheptyloxytriphenylene, as well as of a related dimer and a polymer, each doped with TNF, have been studied by deuterium NMR [54a]. For the monomer, fast rotation around the column axis is observed, this motion is quenched for the dimer and the polymer. The TNF molecules, however, show free rotation in all three CT complexes. Further studies by ,H-NMR combined with dielectric spectroscopy have been performed (1) on mixtures of 2,3,6,7,10,1 l-hexakispentyloxytriphenylene with a TNF derivative with two long aliphatic chains, (2) on mixtures of this hexakispentyloxytriphenylene with a TNF-derived acceptor mainchain polymer, and (3) on two donor-acceptor twin compounds with a different spacer length between the triphenylene donor and the acceptor moiety, giving some insight into the molecular dynamics of such chargetransfer systems [54b]. The optical properties of charge-transfer complexes of the above hexakispentyloxytriphenylene and related discotic electron donor compounds with TNF-type acceptors have been studied in solution and in thin films [55]. CT complexes of that hexaalkoxytriphenylene derivative with TNF, for example, show in both cases a charge-transfer absorption corresponding to 2.51 eV, the polarized absorption spectra obtained from oriented thin films prove that the chargetransfer interaction occurs along the alternate donor-acceptor stacking [55 a]. In addition, picosecond time-resolved absorption spectroscopy based on Kerr ellipsome-
3 Noncalamitic Systems
try revealed information about spectral and temporal characteristics of the oxidized donodreduced acceptor radical ion pairs resulting from laser excitation, and confirmed that the electron transfer occurs in a direction perpendicular to the molecular planes of the donor and acceptor molecules [55b]. The photophysical properties of the TNF complexes of 2,3,6,7,10,11 -hexakisalkyloxy derivatives of triphenylene differ significantly from those of the neat triphenylene hexaalkoxy mesogen; in particular the photoconductivity is lost [55 c]. The chirooptical properties of CT complexes of hexaalkoxytriphenylene main-chain polyesters carrying four chiral or achiral alkyloxy side chains, with TNF, as well as with the enantiomers of 2-(2,4,5,7-tetranitro9-fluorenylideneaminooxy)-propionic acid propyl ester [(+)- and (-)-TAPAE], have been studied by UV and CD (circular dichroic) spectroscopy in thin films [56]. All chiral CT mixtures show high optical activity arising from helical superstructures of the columns induced by the chiral acceptor in the case of achiral polymers or as a result of the chirality of the polymer itself. TNF-CT complexes of 2,3,6,7,10,11hexakispentyloxytriphenylene or related amphiphilic derivatives are also able to form Langmuir-Blodgett (LB) films [57]. Polarized UV/VIS and infrared spectra of those of TNF complexes of hexakisalkyloxytriphenylenes discussed here, for example, indicate a preferred orientation of the columnar axes parallel to the surface and to the dipping direction [57a]. An amphiphilic triphenylene derivative with a pair of o-OH chains and four pentyloxy groups, forming a lamellar mesophase in the neat state, a nematic discotic main chain polymer derived from the aforementioned monomer both exhibit induced columnar hexagonal (colh,) mesophases with TNF in the bulk state [57b]. The X-ray diffraction pattern of the
957
LB films of these CT complexes indicate a dense packing of the columnar aggregates lying parallel to the surface in a multilayer superstructure virtually identical to that in the bulk Colh, phase, quite different to the double layer structure observed for LB films of the neat monomeric and polymeric triphenylene centered mesogens [4 1f]. Radial multiynes constitute another class of large discoid electron donors intensively studied under the aspect of charge-transfer induced mesophases [58-601. The CT complex formation in all the examples described here is indicated by a deep red to brown color on mixing the donor and acceptor compounds. Hexaynes with benzene (1) [58a], naphthalene (2) [58b], or triphenylene (3) [58c] (see Fig. 4) as the “anchors” of “super-disc-cores” with six 4-alkylphenylethinyl “spokes” each form nematic discotic (N,) phases; in the case of the naphthalene hexaynes with peripherical alkyl chains up to n = 8, these are accompanied by columnar rectangular and hexagonal phases (Col,, and Colh,, respectively) at higher temperatures. The addition of an electron acceptor, for example, TNF, leads to the formation of charge transfer complexes and the induction of columnar hexagonal phases with intracolumnar stacking order (Colho)[26,49a]. The phase behavior of some members of these three radial hexayne series (1, 2, and 3) is shown in Fig. 4. The maximum clearing temperature of the CT-induced Col,, phases is found for the equimolar composition. X-ray diffraction studies of the charge-transfer induced mesophases of the TNF complexes of 1 with IZ = 6 proves the hexagonal arrangement of the columnar aggregates and the high intracolumnar stacking order [49 a]. Interestingly, the naphthalene and triphenylene-centered radial hexaynes (2) and (3) also show induced columnar mesophases in binary mixtures with a variety of dipolar carbocy-
958
XVI
Charge-Transfer Systems R
R
R
R
R
R
2 1, 2, 3: R = CnHzn+q
--.
150
--_ 0
Cr
50
Cr
Cr 50
m o l % TNF
100
0
50
mol% TNF
100
0
100
50
m o l % TNF
0
50 50
molX TNF
100
0
50
100
mot% TNF
Figure 4. Schematic phase diagrams of members of the radial hexayne series 1,2, and 3 as electron donors with TNF as the electron acceptor [ 2 6 ] ;Cr: crystalline, N,: nematic discotic, Col,: columnar rectangular, and Col,: columnar hexagonal mesophase (0:intracolumnar ordered), Iso: isotropic liquid (see footnote 1 in Sec. 3 commenting on the nomenclature of columnar mesophases).
clic compounds [61]. However, no indications for charge-transfer interactions are observed in these systems; the inducedhtabilized columnar phases here are more likely the result of improved space filling in combination with dipolar interactions. Even more interesting is the phase behavior of radial pentayne derivatives as, for example, that of the pentayne ethers of type 4 shown in Fig. 5. These pentaynes with unsubstituted peripheral phenyl rings are nonmesomorphic in their pure states, but form liquid crystalline charge-transfer complexes with TNF [26,59 a, d]. The type of charge-transfer induced mesophase depends strongly on the length of the one alkyloxy chain present in the electron donor; an element of steric disturbance in the aggregation of the CTbased columns leads to a two-dimensional lattice (see Fig. 5 ) .
In principle, similar behavior is observed for pentayne ethers carrying methyl groups in the para-positions of their peripheral phenyl rings with TNF, but here the neat pentayne donors are mesomorphic, exhibiting the nematic discotic (N,) type of mesophase [26, 59dl. The members of this series with a pentyloxy or a nonyloxy chain show a slightly decreased clearing temperature of their N, phases and induced Col,, phases below the nematic state. The tridecyloxy and hexadecyloxy derivatives only exhibit the nematic state in mixtures with TNF of acceptor concentrations of more than 60mol%. No phase boundaries are observed, but X-ray studies clearly show the presence of columnar aggregates in the nematic phases of mixtures of, for example, hexadecyl pentayne ether with more than 27 mol% TNF [26, 59dl. Several other radial
150pq 15~ 959
3 Noncalamitic Systems
200
Colho
~~
Cr
100
150
0
20 30 40
I0
50
80
70 80
90 100
0
20
10
m o t % TNF
'
'
S
150
C
Cr
100
0
30 40 50 SO 70 00 90 100
30 40 50 60 70 00 90 100
10 20
m o l % TNF
F
m o l % TNF
T
i
i
NCol
NCol
F
!
NCol
Cr
100
0
10 20 30 40 50 60 70 80 90 100
0
10
0
20 30 40 50 00 70 00 90 100
mol% TNF
mol% TNF
10 20 30 40 50 60 70 00 90 100
m o l % TNF
T/'C
200
150
4 5
10
15
n
Figure 5. Schematic phase diagrams of members of the radial pentayne ether series 4 as electron donors with TNF as the electron acceptor and a plot of the transition temperatures of the equimolar CT complexes with TNF as a function of the number of carbon atoms IZ in the alkyloxy chain of 4;Cr: crystalline, Col,,: columnar hexagonal ( 0 : intracolumnar ordered), N,,,: nematic columnar mesophase, and Iso: isotropic liquid (see footnote I in Sec. 3 commenting on the nomenclature of columnar mesophases).
pentakis(phenylethiny1)phenyl derivatives without side chains in the periphery also show induced nematic and hexagonal columnar mesophases with TNF as the electron acceptor [59b]. Amphiphilic pentayne ethers carrying a hydroxyl group, an ester, or a carboxylic acid function at the end of
the ether chain behave similarly [59 c] . These neat pentaynes form nematic discotic (N,) phases when carrying peripheral alkyl side chains, and are nonmesomorphic without such substituents. The TNF-CT complexes of the first group show induced Colh, phases, while those of the latter one
960
XVI
Charge-Transfer Systems
The dielectric and elastic properties of the nematic columnar (NcOl)phase of chargetransfer mixtures of pentayne ether (4) ( n = 13) with TNF have been studied and compared with those of the N, type of mesophase [60]. The elastic constants K , and K3 are both one order of magnitude higher in the Ncol phase; this dependence on the acceptor concentration indicates that the interaction is strongest for the equimolar composition. Dinuclear palladium organyls with a discoid molecular shape constitute a class of electron donor compounds displaying an unusual and complex phase behavior with TNF as a strong electron acceptor [64] (see Fig. 7). The mesomorphism of these mesogens, exhibiting monotropic nematic discotic phases in their pure forms, in TNF-CT
display induced nematic phases, most probably of the Ncol type. A nematic discotic twin-pentayne ether [62] also exhibits a charge-transfer induced Col,, phase with TNF (see Fig. 6). For other examples of multiyne electron donors see [631. By using the chiral electron acceptor (-)-TAPA, even the cholesteric versions of the nematic discotic phase N;";or of the nematic columnar one NGol can be obtained, for example, by doping a nematic discotic pentayne ether with up to 30 mol% (-)-TAPA or by inducing the N& phase in a ternary mixture composed of pentayne ether (4) ( n = 16) with TNF and (-)-TAPA; the latter electron acceptor does not itself lead to the induction of a columnar mesophase with such pentayne donors [26, 59 d].
T/" C 150
100
-~
0
10
20
30
40
50
60
70
80
90
100
mo1% TNF Figure 6. Schematic phase diagram of a radial twin-pentayne ether as electron donor with TNF as the electron acceptor [26]; Cr: crystalline, N,: nematic discotic, Col,,: columnar hexagonal mesophase ( 0 : intracolumnar ordered), Iso: isotropic liquid (see footnote 1 in Sec. 3 commenting on the nomenclature of columnar mesophases).
3
Noncalamitic Systems
96 1
R
5a: X = CI
b: X = Br c X = l d: X = SCN R = hexyl
k T
T
I T 1 I
t
-i mol% TNF
T
mol% TNF
“TI
@
I 100
-
50
-
I
I
I
I
I
I
I
I
I
I
complexes strongly depends on the bridging group X (Fig. 7). For the members a and b of series 5 with X = C1 or Br, for example, the induction of columnar hexagonal (Col,,) mesophases is observed in mixtures with TNF, while the phase diagram of the binary system iodo-bridged compound 5c/TNF contains an additional nematic discotic area. The thiocyanato-bridged member 5d (X = SCN) shows a strong stabiliza-
Figure 7. Schematic phase diagrams of the dinuclear palladium organyls 5a, c, and d with TNF as the electron acceptor [64]; Cr: crystalline, N,: nematic discotic, Col,,: columnar hexagonal mesophase (0: intramolecular ordered), Iso: isotropic liquid (see footnote 1 in Sec. 3 commenting on the nomenclature of columnar mesophases). Shaded areas mark biphasic regions.
tion of the nematic discotic (ND) phase with TNF. Based on X-ray diffraction experiments, here the acceptor and donor molecules most probably aggregate only to charge-transfer pairs, not to bigger columnar aggregates. However, the ND-phaSe of such an equimolar CT complex of Sd/TNF is completely miscible with the induced Ncol phase of a 1 :1 complex of a pentayne ether (4) with TNF. Thus taking into account
962
XVI
Charge-Transfer Systems
the absence of observable phase transitions or phase boundaries within the nematic regions in the phase diagrams of type 4 pentayne ethers/TNF systems (Fig. 5), a continuous development between these two nematic types of mesomorphic associations seems likely. Phase diagrams similar to those of type 5 pallado-nematogens with TNF are also observed for a related series of dinuclear palladium organyls with hexyloxy instead of the hexyl chains [20] and for similar dinuclear platinum organyls [65]. Members of the 1,2,5,6,8,9,12,13-octa(alky1oxy)dibenzopyrene series, each exhibiting a columnar hexagonal (colhd) mesophase in the pure state, show strong increases of the stability range of their mesophases on the addition of TNF, with a maximum stability for the equimolar donor/ acceptor mixtures [66]. Besides being thermally more stable, the mesophases of the CT complexes are also more highly ordered and exhibit well defined intramolecular stacking order, classifying them as Col,,. Deuterium NMR studies revealed planar diffusion around the columnar axis as the main molecular reorientation process in the mesophase. An X-ray diffraction peak at a position corresponding to twice the normal intracolumnar distance, as observed for the colh, phase of TNF-CT complexes of triphenylene hexaethers, indicates that the donor and acceptor molecules are stacked alternately in the columns. Moreover, when alkanes are added to these 1 :1 charge transfer complexes, a lyotropic nematic phase of the Ncol type with columnar structures as the mesogenic units is induced [66]. Such nematic columnar lyomesophases with apolar organic solvents are only formed by compounds with especially strong intracolumnar interactions, such as, for example, long chain hexaamides of hexaaminobenzene [67a], dinuclear copper(I1)- and rhodi-
um(1I)carboxylates [67b, c], tetranuclear palladium [67d, el, and platinum organyls [45] or amides of azacrown ethers [67f]. However, charge-transfer interactions between disc- or sheet-shaped electron donor compounds and strong electron acceptors might lead to sufficient stability of columnar aggregates. Thus equimolar TNF-CT complexes of the hexaynes 1 or 3 with heptyl chains and benzene or triphenylene as the “anchor” display nematic lyomesophases with alkane solvents [26]. Even charge-transfer complexes of triphenylene hexaethers with TNF form nematic columnar lyomesophases with alkanes [68]. Tetranuclear palladium organyls exhibiting thermotropic columnar mesophases with an oblique two-dimensional lattice and without intracolumnar stacking order (Col,,, d) [69] also form lyotropic mesophases with alkane solvents; in particular, some members exhibit two different lyotropic nematic phases side by side [67d, el. Adding TNF results in a significantly increased stacking order in, for example, a 1 :2 donor-acceptor complex. This CT complex forms only one nematic lyomesophase with, for example, pentadecane over an extended temperature range; by using the chiral electron acceptors (+)- or (-)-TAPA, a cholesteric nematic N* phase is observed [70a]. A dinuclear bisimine palladiumorganyl with two acetylacetonato ligands represents another example of charge-transfer induced thermotropic and lyotropic mesomorphism [70b]. This nonmesomorphic compound exhibits a charge-transfer induced columnar mesophase with TNF; the two-dimensional arrangement of the columns in this mesophase is still not clear. However, in ternary mixtures with alkanes such as, for example, pentadecane, a lyotropic nematic phase is observed [70b]. A tricyclyne, i.e., three phenyl rings connected via three ethinyl links forming an annulene type macrocycle and carry-
963
3 Noncalamitic Systems
ing six hexyloxy chains in its outer phenyl positions, nonmesomorphic in its pure state, also forms a highly viscous, most probably columnar mesophase with TNF [71]. Based on the observations described in this section, it seems that for a qualitative discussion of the interactions leading to mesomorphic self-organization of disc- or sheet-like compounds, separation into an intracolumnar potential and an intercolumnar one is usually sufficient. Neat discoid mesogens show only one type of potential along their column axis; the ordered or disordered nature of a columnar phase depends on the width of the potential minimum in this description, while the thermal stability reflects the depth of the minimum. For charge transfer induced or stabilized columnar phases, three types of intermolecular interactions - donor/donor, acceptor/acceptor, and donor/acceptor - must be considered. In the case of the intercalation of electron acceptors, e.g., TNF, the mixed interaction is the dominant one, with a deeper and narrower potential minimum than for the donor/donor combination of the neat mesogen, leading to alternating stacking and increased stability and intracolumnar positional order. A dominance of the donoddonor and/or acceptor/acceptor interactions might lead to the formation of separate columnar aggregates of the donor and acceptor compounds, as discussed for the mesomorphic tetracyanoquinodimethane
(TCNQ) complexes of dithiapyranylidene derivatives (see Fig. 2 and [40]). The intercolumnar interactions in a mesomorphic system built of columns of sufficient stability then can be reduced by the addition of an alkane solvent or by attaching sterical restraints. The potential minimum normal to the column axis becomes flattened and broadened, resulting in the formation of a nematic columnar mesophase and finally an isotropic solution of columnar aggregates. This behavior is well known for the so-called chromonic lyomesogens in aqueous systems [72]. Finally, dinitrophenylhydrazones of trialkyloxy derivatives of benzaldehyde should be mentioned here as mesogens with charge-transfer contributions to their mesomorphic self-organization in the pure state [20, 731. With a 3,4,5 substitution pattern of the alkyloxy chains, a columnar hexagonal mesophase with very high intracolumnar stacking order (Dho phase) is formed as a result of a regular head-to-tail packing of subsequent molecules [73]. This packing mode has been attributed to a planar conformation of the molecules supported by hydrogen bridge formation and charge-transfer interactions between the donor and acceptor moieties of subsequent molecules. The same type of intracolumnar organization is observed for a 2,3,4-trialkyloxy derivative, but here the columns form a rectangular two-dimensional lattice (Col,, phase) (see Fig. 8 and [20]).
Don
-
Acc
Figure 8. Model of the mesomorphic self-organization of 2,3,4-trisdodecyloxy-benzaldehyde-2',4'-dinitrophenylhydrazone in its columnar rectangular mesophase 1201; Don: donor part, Acc: acceptor part.
964
XVI
Charge-Transfer Systems
It is especially noteworthy that this dinitrophenylhydrazone can act as an electron acceptor in mixtures with the hexayne nematodiscogens 1 or 2 ( n = 9, see Fig. 4), but also constitutes an electron donor in mixtures with TNF. Both types of binary systems show an induced mesophase with textures typical for columnar hexagonal mesophases [20]. Partially related to the topics discussed here is the doping of, for example, triphenylene 2,3,6,7,10,1l-hexaetherswith small amounts of bromine or the Lewis acid aluminum trichloride to induce electrical conductivity [74]; however, no change of the mesomorphic self-organization is caused by these dopants. In another very recent work [75] trifluoroacetic acid, a representative of a new family of dopants, i.e. that of Brprnsted acids, was found to stabilize or induce columnar mesophases in disc-like alkoxy compounds, such as aforementioned hexaalkyloxytriphenylene and also hexa-alkyloxytribenzocyclononene. This acid forms 1 : 1 molar complexes with these two types of arene hexaethers. The mesophases exhibited by these complexes show enhanced properties for the mesogenic homologues and induced mesomorphism for the lower ones, which in the neat form are not mesogenic. It is suggested that the enhancement of the mesomorphic properties is due to the formation of oxonium complexes. On the other hand, it is well known that perfluoro arenes, for example, hexafluorobenzene or octafluoronaphthalene, form some type of “complex” with their hydrocarbon parent compounds, having melting points higher than those of the pure components [76]. Very recently, the deuterium NMR spectra of the octafluoronaphthalene! naphthalene system have been explained in terms of a highly ordered columnar mesophase [76d]. However, the face-to-face stacking in such solids and also in liquids is
better described as a result of interactions based on different molecular electric quadrupole moments [2, 7 6 ~ 1 .
References See, for example: a) G. R. Luckhurst, Ber: Bunsenges. Phys. Chem. 1993,97, 1169; b) A. P. J. Emerson, G. R. Luckhurst, S. G. Whatling, Mol. Phys. 1994, 82, 1 1 3 . C. A. Hunter, J. K. M. Sanders, J. Am. Chem. Soc. 1990,112,5525 and references cited therein. For related recent reports, as an introduction see, for example, a) M. Harmata, C. L. Barnes, J. Am. Chem. Soc. 1990, 112, 5655; b) S . B. Ferguson, E. M. Sanford, E. M. Seward, F. Diederich, J. Am. Chem. Soc. 1991, 113, 5410; c) F. Cozzi, M. Cinquini, R. Annuziata, J. S. Siegel, J. Am. Chem. Soc. 1993,115,5330;d) A. W. Schwabacher, S. Zhang, W. Davy, J . Am. Chem. SOC. 1993, 115, 6995; e) C. A. Hunter. Angew. Chem. 1993,105, 1653;Angew. Chem.. lnt. Ed. Engl. 1993, 32, 1584; f) F. Cozzi, F. Ponzini, R. Annuziata, M. Cinquini, J. S. Siegel, Angew Chem. 1995,107, 1092;Angew. Chem., Int. Ed. Engl. 1995, 34, 1019. See for example a) Z. G. Soos, Ann. Rev. Phys. Chem. 1974,25, 121; b) P. Delhaes,Mol. Cryst. Liq. Cryst. 1983,96, 229. M. Rigby, E. B. Smith, W. A. Wakeham, G. C . Maitland, The Forces between Molecules, Clarendon, Oxford 1986. For a short introduction, see: S. Diele, Ber: Burtsenges. Phys. Chem. 1993, 97, 1326. a) G. Pelzl, D. Demus, H. Sackmann, Z. Phys. Chem. 1968,238,22;b) H. Stegemeyer, Nachr: Chem. Tech. Lab. 1993,41, 966, see p. 970. a) J. W. Park, C. S. Bak, M. M. Labes, J . Am. Chem. SOC.1975,97,4398;b) J. W. Park, M. M. Labes, J. Appl. Phys. 1977, 48, 22; c) see related comments in [23d] and [23f]. J. W. Park, M. M. Labes, Mol. Cr?/st.Liq. Cryst. 1977, 34, 147. a) K. Araya, Y. Matsunaga, Bull. Chem. Soc. Jpn. 1980, 53, 3079; b) K. Araya, Y. Matsunaga, Mol. Cryst. Liq. Cryst. 1981, 67, 153; c) K. Araya, Y. Matsunaga, Bull. Chem. Soc. Jprt. 1981,54,2430;d) K. Araya, Y. Matsunaga, Bull. Chem. Soc. Jpn. 1982, 55, 1710; e) K. Araya. N. Homura, Y. Matsunaga, Bull. Chem. Soc. Jpn. 1982,55,1953;0 Y. Matsunaga, I. Suzuki, Bull. Chem. Soc. Jpn. 1984,57, 141 I ; g) Y. Matsunaga, Mol. Cryst. Liq. Cryst. 1985, 125. 269; h) N. Homura, Y. Matsunaga, M. Suzuki, Mol.
4 References Cryst. Liq. Cryst. 1985,131,273; i) Y.Matsunaga, N. Kamiyama, Y. Nakayasu, Mol. Cryst. Liq. Cryst. 1987, 147, 85. [ 111 a) M. Fukui, Y. Matsunaga, Bull. Chem. Soc. Jpn. 1982, 55, 3707. 1121 a) E. Chino, Y. Matsunaga, Bull. Chem. SOC. Jpn. 1983, 56, 3230; b) E. Chino, Y. Matsunaga, M. Suzuki, Bull. Chem. Soc. Jpn. 1984, 57, 237 I . [I31 a) Y. Iida, Bull. Chem. Soc. Jpn. 1982, 55, 313; b) Y. Iida, Bull. Chem. SOC.Jpn. 1982,5.5,2661. [ 141 a) N. K. Sharma, G. Pelzl, D. Demus, W. Weissflog, Z. Phys. Chem. 1980, 261, 579; b) D. Demus, G. Pelzl, N. K. Sharma, w . Weissflog, Mol. Cryst. Liq. Cryst. 1981, 76, 241; c) D. Demus, A. Hauser, G. Pelzl, U. Bottger, S. Schonburg, Cryst. Res. Technol. 1985,20, 381; d) W. Weissflog, G. Pelzl, D. Demus, Cryst. Res. Technol. 1986, 21, 117; e) G. Pelzl, U. Bottger, S . Diele, D. Demus, Cryst. Res. Technol. 1987,22, 1321. [I51 D. Demus, A. Hauser, G. Pelzl, U. Bottger, S. Schonburg, Cryst. Res. Technol. 1985, 20, 381. [ 161 a) W. Waclawek, R. Dabrowski, A. Domagala, Mol. Crysr. Liq. Cryst. 1982, 84, 255; b) K. W. Sadowska, A. Zywocinski, J. Stecki, R. Dabrowski, J. Physique 1982,43, 1673. [I71 W. H. de Jeu, L. Longa, D. Demus, J. Chem. Phys. 1986, 84, 6410. [ 181 A. Kolbe, G. Pelzl, W. Weissflog, Adv. Mol. Relax. Interact. Proc. 1982,24, 25 1. [ 191 a) T. Kato, J. M. J. Frechet, J. Am. Chem. SOC. 1989, 111, 8533; b) T. Kato, J. M. J. Frechet, Macromolecules 1989, 22, 3818; c) M.-J. Brienne, J. Gabard, J.-M. Lehn, I. Stibor, J. Chem. Soc., Chem. Commun. 1989, 1868; d) T. Kato, A. Fujishima, J. M. J. Frechet, Chem. Lett. 1990,919; e) T. Kato, P. G. Wilson, A. Fujishima, J. M. J. Frechet, Chem. Lett. 1990, 2003; f) L. Y. Yu, J. M. Wu, S. L. Wu, Mol. Cryst. Liq. Cryst. 1991, 198,401; g) C. Fouquey, J.-M. Lehn, A.-M. Levelut, Adv. Muter. 1990, 2, 254; h) C. M. Paleos, J. Michas, Liq. Cryst. 1992, I I , 773; i) H. Kresse. C. Dorscheid, I. Szulzewski, R. Frank, R. Paschke, Ber. Bunsenges. Phys. Chem. 1993, 97, 1345; j) C. M. Paleos, D. Tsiourvas, Angew. Chem. 1995,107, 1839; Angew. Chem., Int. Ed. Engl. 1995,34, 1696. [20] K. Praefcke, B. Gundogan, D. Singer, unpublished results; Bilgi Gundogan, Synthesen und Eigenschaften neuer flachiger Flussigkristalle, Koster Verlag, Berlin 1996, ISBN 3-89574-1566, Ph. D. Thesis, Technische Universitat Berlin, Germany. 1211 a) S. Diele, G. Pelzl, W. Weissflog, D. Demus, Liq. Cryst. 1988,3, 1047; b) G. Pelzl, A. Humke, s. Diele, D. Demus, W. Weissflog, Liq. Cryst. 1990, 7, 115; c) G. Pelzl, A. Humke, S. Diele, S . Ziebarth, W. Weissflog, D. Demus, Cryst. Res. Technol. 1990, 25, 587.
965
[22] See for example: a) M. E. Neubert, K. Leung, W. A. Saupe, Mol. Crysr. Liq. Cryst. 1986,135,283; b) B. Ziemnicka, A. de Vries, J. W. Doane, S. L. Arora, Mol. Cryst. Liq. Cryst. 1986,132,289; c) S. Diele, G . Pelzl, A. Humke, S. Wunsch, W. Schaefer, H. Zaschke, D. Demus, Mol. Crysr. Liq. Cryst. 1989, 173, 113. [23] a) J. P. Schroeder, D. C. Schroeder, J. Org. Chem. 1968, 33, 591; b) J. S. Dave, P. R. Patel, K. L. Vasanth, Mol. Cryst. Liq. Cryst. 1969, 8,93; c) C. S. Oh, Mol. Cryst. Liq. Ctqst. 1977, 42, 1; d) G. Heppke, E.-J. Richter, Z. Naturforsch. 1978,33a, 185; e) A. C. Griffin, R. F. Fisher, S. J. Havens, J. Am. Chem. SOC. 1978,100,6329;f') M. Domon, J. Billard, J. Physique Colloque C3 1979, 40, C3-413; g) P. E. Cladis, Mol. Cryst. Liq. Cryst. 1981, 67, 117; h) L. Songsheng, G . Jezhen, W. Liangyu, Mol. Cryst. Liq. Cryst. 1983, 100, 285; i) V. V. Belyaev, T. P. Antonyan, L. N. Lisetski, M. F. Grebyonkin, G . G . Slashchova, V. V. Petrov, Mol. Cryst. Liq. Cryst. 1985, 129, 221; j ) L. K. M. Chan, G . W. Gray, D. Lacey, T. Srithanratana, K. J. Toyne, Mol. Cryst. Liq. Cryst. 1987, 150b, 335; k) K. Czuprynski. R. Dabrowski, B. Sosnowska, J. Baran, Liq. Cryst. 1989,5,505. [24] a) F. Hardouin, A.-M. Levelut, M. F. Archad, G. Sigaud, J. Chim. Phys. 1983, 80, 53; b) J. Prost, P. Barois, J. Chim. Phys. 1983, 80, 65; c) H. T. Nguyen, J. Chim. Phys. 1983, 80, 83; d) R. Shashidhar, B. R. Ratna, Liq. Cryst. 1989, 5,421; e) J. Prost, Adv. Phys. 1984,33, 1. [25] J. Szabon, Mol. Cryst. Liq. Cryst. 1985,124,343 and references cited therein. [26] K. Praefcke, D. Singer, unpublished results; D. Singer, Neue makrodiscotische Flussigkristalle - mesomorphe Aggregationsformen und ihre Beeinflussung, Koster Verlag, Berlin 1993, ISBN 3-929937-79-4, Ph. D. Thesis, Technische Universitat Berlin. [27] B. Neumann, D. Joachimi, C. Tschierske, Adv. Matel: 1997, 9, 241, and Liq. Cryst. 1997, 22, 509. [28] a) C. Pugh, V. Percec, Polym. Bull. 1990, 23, 177; b) K. Kato, S . Kobinata, S . Maeda, Liq. Cryst. 1989, 5, 595. [29] S. Imazeki, A. Mukoh, N. Tanaka, M. Kinoshita, Mol. Cryst. Liq. Cryst. 1993, 225, 197. [30] a) M. Portugall, H. Ringsdorf, R. Zentel, Makromol. Chem. 1982, 183, 2311; b) A. c. Griffin, A. M. Bhatti, R. S. L. Hung in Nonlinearcznd Electroactive Polymers (Ed.: P. N. Prasad, D. R. Ulrich, Plenum, New York 1988; c) B. Beck, H. Ringsdorf, Liq. Cryst. 1990, 8, 247. [31] a) C. T. Imrie, F. E. Karasz, G. S . Attard, Liq. Cryst. 1991, 9, 47; b) T. Schleeh, C. T. Imric, D. M. Rice, F. E. Karasz, G. S. Attard, J. Polym. Sci. 1993,32A, 1859.
966
XVI
Charge-Transfer Systems
[32] a) M. Sato, T. Nakano, K.-I. Mukaida, Liq. Cryst. 1995, 18, 645; b) Y. Kosaka, T. Kato, T. Uryu, Macromolecules 1994,27, 2658. [33] See for example C. T. Imrie, T. Schleeh, F. E. Karasz, G. S. Attard, Macromolecules 1993,26, 539. 1341 a) J. L. Hogan, C. T. Imrie, G. R. Luckhurst, Liq. Cryst. 1988, 3, 645; b) G. S . Attard, S. Garnett, C. G. Hickmann, C. T. Imrie, L. Taylor, Liq. Cryst. 1990, 7,495; c) G. S . Attard, R. W. Date, C. T. Imrie, G. R. Luckhurst, S . J. Roskilly, J. M. Seddon, L. Taylor, Liq. Cryst. 1994, 16, 529; d) A. E. Blatch, I. D. Fletcher, G. R. Luckhurst, Liq. Cryst. 1995, 18, 801. [35] S. Ujiie, H. Uchino, K. Iimura, Chem. Lett. 1995, 195. [36] a) M. Sone, B. R. Harkness, J. Watanabe, T. Yamashita, T. Torii, K. Horie, Polym. J. 1993, 25, 997; b) B. R. Harkness, J. Watanabe, Macromolecules 1991, 24, 6759. 1371 a) M. Nitta, N. Ozaki, H. Suenaga, K. Nakaya, S. Kobayashi, Jpn. J. Appl. Phys. 1988, 27, L477; b) K. Nakaya, B. Y. Zhang, M. Yoshida, I. Isa, S . Shindoh, S. Kobayashi, Jpn. J. Appl. Phys. 1989, 28, L116; c) B. Y. Zhang, M. Yoshida, H. Maeda, M. Kimura, H. Sekine, S . Kobayashi, Mol. Cryst. Liq. Cryst. 1991, 202, 149. [38] K. Ono, J. Nakanowatari, Jpn. J. Appl. Phys. 1991,30,2832. [39] E. Matsui, K. Nito, A. Yasuda, Ferroelectrics 19Y3, 149, 97. [40] a) V. Gionis, R. Fugnitto, G. Meyer, H. Strzelecka, J. C. Dubois, Mol. Cryst. Liq. Cryst. 1982, 90,153; b) V. Gionis, R. Fugnitto, H. Strzelecka, Mol. Cryst. Liq. Cryst. 1983, 96, 215; c) P. Davidson, A.-M. Levelut, H. Strzelecka, V. Gionis, J. Physique 1983,44, L-823; d) H. Strzelecka, V. Gionis, J. Rivory, S. Flandrois, J . Physique, Colloque C3 1983, 44, C3-1201; e) R. Kormann, L. Zuppiroli, V. Gionis, H. Strzelecka, Mol. Cryst. Liq. Cryst. 1986, 133, 283; f) V. Gionis, H. Strzelecka, M. Veber, R. Kormann, L. Zuppiroli,Mol. Cryst. Liq. Cryst. 1986, 137, 365. 1411 a) K. Ohta, H. Muroki, A. Takagi, K.-I. Hatadd, H. Ema, I. Yamamoto, K. Matsuzaki, Mol. Cryst. Liq. Cryst. 1986, 140, 131; b) H. Sakashita, A. Nishitani, Y. Sumiya, H. Terauchi, K. Ohta, I. Yamamoto, Mol. Cryst. Liq. Cryst. 1988,163, 21 I ; c) A. C. Ribeiro, A. F. Martins, A,-M. Giroud-Godquin, Mol. Cryst. Liq. Cryst. Lett. 1988, 5, 133 and references therein; d) G. Latterman, S. Schmidt, R. Kleppinger, J. H. Wendorff, Adv. Muter. 1992, 4, 30; e) J. Malthete, A. Collet, A.-M. Levelut, L. Liebert, Adv. Muter. 1992,4,37;f) 0. Karthaus, H. Ringsdorf, V. V. Tsukruk, J. H. Wendorff, Langmuir 1992, 8, 2279.
[42] J. Billard, C. R. Acad. Sci. Paris 1984, 11-299, 905. [43] B. F. Darocha, D. D. Titus, D. J. Sandman. Actu Cryst. 1979, B35, 2445. [44] F. D. Saeva, G. A. Reynolds, L. Kaszczuk, J. Am. Chem. Soc. 1982,104,3524. [45] K. Praefcke, B. Bilgin, N. Usol’tseva, B. Heinrich, D. Guillon, J. Mater. Chem. 1995,5, 2257; K. Praefcke, J . D. Holbrey, N. Usol’tseva, Mol. Cryst. Liq. Cryst. 1996, 288, 189, see p. 194. K. Praefcke, J. D. Holbrey, J. Incl. Phen. Mol. Recogn. Chem. 1996, 24, 19, see p. 25. - K. Praefcke, J. D. Holbrey, N. Usol’tseva, D. Blunk, Mol. Cryst. Liq. Cryst. 1997, 292, 123. see p. 129-130. [46] a) H. Ringsdorf, R. Wustefeld, E. Zerta, M. Ebert, J. H. Wendorff,Angew. Chem. 1989,101. 934; Angew. Chem., Int. Ed. Engl. 1989,28,914: b) H. Ringsdorf, R. Wiistefeld, M. Ebert, J. H. Wendorff, ACS Polym. Prep. 1989, 30,479. [47] H. Ringsdorf, R. Wustefeld, Phil. 7ran.s. R. Soc. Lond. 1990, A330,95. [48] See C. Destrade, P. Foucher, H. Gasparoux, H. T. Nguyen, A.-M. Levelut, J. Malthete, Mol. Cryst. Liq. Ctyst. 1984,106, 12 1 and references cited therein. [49] a) H. Bengs, M. Ebert, 0. Karthaus, B. Kohne, K. Praefcke, H. Ringsdorf, J. H. Wendorff. R. Wiistefeld,Adv. Mater. 1990,2,141;b) M. Ebert, G. Frick, C. Baehr, J. H. Wendorff, R. Wiistefeld, H. Ringsdorf, Liq. Cryst. 1992, 11. 293. [50] M. Ebert, Ph. D. Thesis, TH Darmstadt 1990. [51] H. Bengs, 0. Karthaus, H. Ringsdorf, C. Baehr, M. Ebert, J. H. Wendorff, Liq. Cryst. 1991, 10, 161. 1521 a) M. Moller, V. Tsukruk, J. H. Wendorff, H. Bengs, H. Ringsdorf, Liq. Cryst. 1992, 12, 17; b) D. Janietz, Chem. Commun. 1996, 713: c) V. V. Tsukruk, D. H. Reneker, H. Bengs, H. Ringsdorf, Langmuir 1993, 9, 2141; d) K. Praefcke, A. Eckert, D. Blunk, Liq. Cryst. 1997. 22, 113, and SPIE-Proceedings, in press. - N . Boden,R. J.Bushby,A. N.Cammidge,S. Duckworth, G. Headdock, J . Mater. Chem. 1997, 7, 601. [53] a) H. Bengs, R. Renkel, H. Ringsdorf, C. Baehr, M. Ebert, J. H. Wendorff, Makromol. Chem.. Rapid Commun. 1991, 12, 439; b) M. Miiller, V. V. Tsukruk, J. Wendling, J. H. Wendorff, H. Bengs, H. Ringsdorf, Makromol. Cheni. 1992, 193,2659. [54] a) W. Kranig, C. Boeffel, H. W. Spiess, 0. Karthaus, H. Ringsdorf, R. Wiistefeld, Liq. C r w . 1990, 8, 375; b) M. Moller, J. H. Wendorff, M. Werth, H. W. Spiess, H. Bengs, 0. Karthaus, H. Ringsdorf, Liq. Cryst. 1994, 17, 381. [55] a) D. Markovitsi, H. Bengs, H. Ringsdorf, J . Chem. Soc., Faraday Trans. 1992, 88, 1275: b) D. Markovitsi, N. Pfeffer, F. Charra. J.-M.
4 References Nunzi, H. Bengs, H. Ringsdorf, J. Chem. Soc., Faraday Trans. 1993, 89, 37; c) See for example D. Markovitsi, A. Germain, P. Millie, P. Lecuyer, L. K. Gallos, P. Argyrakis, H. Bengs, H. Ringsdorf, J. Phys. Chem. 1995,99, 1005. 1561 M. M. Green, H. Ringsdorf, J. Wagner, R. Wustefeld, Angew. Chem. 1990,102, 1525; Angew. Chem., Int. Ed. Engl. 1990,29, 1478. [57] a)K. Ogawa,H. Yonehara,C. Pac,E. Maekawa, Bull. Chem. Soc. Jpn. 1993, 66, 1378; b) V. V. Tsukruk, J. H. Wendorff, 0. Karthaus, H. Ringsdorf, Langmuir 1993, 9, 614. [58] a) B. Kohne, K. Praefcke, Chimia 1987,41,196; b) K. Praefcke, B. Kohne, K. Gutbier, N. Johnen, D. Singer, Liq. Cryst. 1989, 5, 233; c) K. Praefcke, B. Kohne, D. Singer, Angew. Chem. 1990, 102, 200; Angew. Chem., Int. Ed. Engl. 1990,29, 177. [59] a) K. Praefcke, D. Singer, B. Kohne, M. Ebert, A. Liebmann, J. H. Wendorff, Liq. Cryst. 1991, 10, 147; b) K. Praefcke, D. Singer, M. Langner, B. Kohne, M. Ebert, A. Liebmann, J. H. Wendorff, Mol. Cryst. Liq. Cryst. 1992, 215, 121; c) D. Janietz, K. Praefcke, D. Singer, Liq. Cryst. 1993, 13, 247; d) K. Praefcke, D. Singer, A. Eckert, Liq. Cryst. 1994, 16, 53. [60] B. Sabaschus, D. Singer, G . Heppke, K. Praefcke, Liq. Cryst. 1992, 12, 863. [61] K. Praefcke, D. Singer, B. Kohne, Liq. Cryst. 1993, 13,445. [62] K. Praefcke, B. Kohne, B. Gundogan, D. Singer, D. Demus, s. Diele, G. Pelzl, U. Bakowsky, Mol. Cryst. Liq. Cryst. 1991, 198, 393. [63] K. Praefcke, D. Singer, B. Gundogan, K. Gutbier, M. Langner, Ber. Bunsenges. Phys. Chem. 1993, 97, 1358; Corrigendum 1994, 98, 118. [64] D. Singer, A. Liebmann, K. Praefcke, J. H. Wendorff, Liq. Cryst. 1993,13,785. [65] a) K. Praefcke, B. Bilgin, unpublished results; Belkiz Bilgin, Neueflachige Mesogene des Palladiums and Platins, Koster Verlag, Berlin 1996, ISBN 3-89574-155-8, Ph. D. Thesis, Technische Universtitat Berlin, Germany; b) K. Praefcke, B. Bilgin, J. Pickardt, M. Borowski, Chem. Ber. 1994,127, 1543. [66] S. Zamir, D. Singer, N. Spielberg, E. J. Wachtel, H. Zimmermann, R. Poupko, Z. Luz, Liq. Cryst. 1996, 21, 39. [67] a) B. Kohne, K. Praefcke, T. Derz, H. Hoffmann, B. Schwandner, Chimia 1986, 40, 171;
967
b) M. Ibn-Elhaj, D. Guillon, A. Skoulios, A.-M. Giroud-Godquin, J.-C. Marchon, J. Phys. 11 France 1992, 2 , 2197; c) R. Seghrouchni, A. Skoulios, J. Phys. II France 1995, 5, 1385; d) N. V. Usol’tseva, K. Praefcke, D. Singer, B. Gundogan, Liq. Cryst. 1994, 16, 601; e) N. Usol’tseva, G. Hauck, H. D. Koswig, K. Praefcke, B. Heinrich, Liq. Cryst. 1996, 20, 731; f) H. Ringsdorf, K. Schimossek, M. Wittenberg, J. H. Wendorff, H. Fischer, 24. Freiburger Arbeitstagung Flussigkristalle, 05.-07. April 1995, Freiburg (Germany), Poster P42. 1681 K. Praefcke, D. Singer, Institute of Organic Chemistry, Technische Universitat Berlin, unpublished results. [69] K. Praefcke, S. Diele, J. Pickardt, B. Gundogan, U. Nutz, D. Singer, Liq. Cryst. 1995, 18, 857. [70] a) N. V. Usol’tseva, K. Praefcke, D. Singer, B. Gundogan, Liq. Cryst. 1994, 16, 617; b) N. V. Usol’tseva, K. Praefcke, D. Singer, B. Gundogan, Mol. Muter. 1994,4, 253. [71] S. Pollack, Howard University, Washington DC, U. S. A., private communication to D. Singer. [72] See for example T. K. Attwood, J. E. Lydon, C. Hall, G. J. T. Tiddy, Liq. Cryst. 1990, 7, 657. [73] a) W. Paulus, H. Ringsdorf, S. Diele, G. Pelzl, Liq. Cryst. 1991, 9, 807; b) W. G. Scaife, G. McMullin, H. Groothues, Mol. Cryst. Liq. Cryst. 1995, 261, 5 1. [74] See for example: a) L. Y. Chiang, J. P. Stokes, C. R. Safinya, A. N. Bloch, Mol. Cryst. Liq. Cryst. 1985,125,279;b) N. Boden, R. J. Bushby, J. Clements, M. V. Jesudason, P. F. Knowles, G. Williams, Chem. Phys. Lett. 1988,152,94; c) N. Boden, R. Borner, D. R. Brown, R. J. Bushby, J. Clements, Liq. Cryst. 1992, 11, 325; d) N. Boden, R. J. Bushby, J. Clements, J. Chem. Phys. 1993, 98,5920; e) N. Boden, R. J. Bushby, J. Clements, J. Movaghar, K. J. Donovan, T. Kreouzis, Phys. Rev. 1995, B52, 13274. [75] L. Calucci, H. Zimmermann, E. J. Wachtel, R . Poupko, Z. Luz, Liq. Cryst. 1997,22, 621. [76] a) C. R. Patrick, G. S . Prosser, Nature 1960,187, 1021; b) T. Dahl, Acta Chem. Scand. 1973,27, 995; c) J. H. Williams, Acc. Chem. Res. 1993, 26, 593; d) G. R. Luckhurst, W. M. Tearle, Abstracts: Poster NM-20, European Conference on Liquid Crystals Science and Technology, ECLC 95, 5.-10. March, 1995, Bovec, Slovenia.
Handbook of Liquid Crystals Edited by D. Demus, J. Goodby, G. W. Gray, H.-W. Spiess, V. Vill Copyright 0 WILEY-VCH Verlag GmbH. 1998
Chapter XVII Hydrogen-Bonded Systems Takashi Kato
1 Introduction Hydrogen bonding is one of the key interactions for chemical and biological processes in nature due to its stability, directionality, and dynamics. The history of liquid crystals and hydrogen bonding has started in the early time of 20th century when mesomorphic properties were observed for the derivatives of cinnamic acids [ l ] and benzoic acids [2], and argaric acid [3]. For monosaccharides with long alkyl chains, Fischer observed double melting behavior in 1911 [4]. This behavior was recognized later as thermotropic liquid crystallinity [S]. This behavior was recently established as an effect of hydrogen bonding [6]. Some compounds containing hydrogen bonding moieties, such as polyols [7], silanediols [8], amides [9], and hydrazines [lo] have been shown to exhibit thermotropic mesophases. These compounds show thermotropic liquid-crystalline phases due to the intermolecular hydrogen bonding between identical molecules. A new type of thermotropic “supramolecular” liquid crystal obtained by molecular recognition processes through hetero-intermolecular hydrogen bonds, which function between different and independent molecules species, were reported by Kato and FrCchet [ 11-13) and Lehn and co-workers [ 14, 151 in 1989 and 1993. After these findings, a wide variety of structures of molecular complexes has been prepared by supramolecular self-assembly.
This chapter focuses on supramolecular hydrogen-bonded liquid crystals built by the formation of hetero-intermolecular hydrogen bonds. The H-bonded liquid crystals obtained by the association of identical molecules are described in other chapters.
2 PyridineKarboxylic Acid System 2.1 Self-Assembly of Low Molecular Weight Complexes 2.1.1 Structures and Thermal Properties Liquid-crystalline supramolecular complexes with well-defined structures can be obtained by self-assembly through heterointermolecular hydrogen bonding between carboxylic acid and pyridine moieties [ 111. For example, the complexation of 4-methoxybenzoic acid (1) and trans- 1,2-bis(4-pyridy1)ethylene (2) results in the formation of mesogenic complex 3, which exhibits a nematic phase from 166 to 185 “C (Scheme 1) [ 161. In this H-bonded mesogenic structure, nonmesogenic compound 2 functions as the core unit of the mesogen through the hydrogen bond, which is directional and stable. Figure 1 shows the plot of phase behavior against the length of the alkyl chain for the complex of 4-alkoxybenzoic acid
970
XVII
Hydrogen-Bonded Systems
nonmesogenic
nonmesogenic
180
-
. 6
U
160-
nematic 166-185 'C
E
'8
Scheme 1
140
and truns-l,2-bis(4-pyridyl)ethylene. These complexes show stable smectic and nematic behavior over 100°C [16]. The relationship between the core structure and the transition temperature for the complexes of 4-hexyloxybenzoic acid (4) and bipyridines is shown in Table 1 [16]. The highest clearing temperature is achieved for the complex based on 2, which has the most coplanar structure. Table 2 compares the mesomorphic temperature ranges of hydrogen-bonded complexes (8, 9 and 10) with different core lengths [ 16-19]. The thermal stability of the mesophase increases when the aromatic ring number increases, as is observed for normal liquid crystals. Significant mesophase stabilization is seen for complex 8 consisting of four aromatic rings, which was first reported as the dissimilar hydrogen-bonded
-
-
c
"'*
sc
\
' loo 0
2 *
4
6O
Cr
carbon number (n)in the alkyl chain for the series of hydrogen-bonded complexes obtained from 4-alkoxybenzoic acid and trans-l,2-bis(4-pyridyl)ethylene. (SA: SmA; S,: SmC; K: Cr.) w: Ii; 0 : I,.
mesogenic complex [ 121. In contrast, the mesophase near room temperature was achieved by the complexation of 4-hexyloxybenzoic acid (4) and 4-octylpyridine [ 191. The transition temperatures of the hydrogen-bonded mesogens 10 and 11, consisting of two rings, are compared with those of the structurally related phenyl benzoates
SmC
5 (X=none)
0 0
140
7 (X=-CH,-CH,-)
0
123
a
Transition temperatures in "C.
I
SmA
6 (X=-CH=CH-)
105
10L
Figure 1. Plot of transition temperature against the
Table 1. Liquid crystalline behavior of supramolecular 2 :1 hydrogen-bonded complexes derived from 4-hexyloxybenzoic acid and bipyridines (4). a
Complex
8
130
159 177 152
0 0 0
2 PyridineICarboxylic Acid System
97 1
Table 2. Liquid-crystalline behavior of supramolecular 1 :1 hydrogen-bonded complexes formed through intermolecular hydrogen bonding. a
Complex
8 9 10 a
Cr
N
SmA
136 102
I
238 155 48
160 130 38
0
Transition temperatures in “C.
Table 3. Transition temperatures of liquid-crystalline complexes 10 and 11based on 4-hexyloxybenzoic acid and 4-octylpyridine, and of 4-octylphenyl4-hexyloxybenzoates(OPHOB and OPHOFB). X
O
/ \ c*
C H 3 * H 2 k\ 0 o-H 0. . . . ~-~ C H 2 + C H 3
11 :X=H 10 X=F
non-covalent bond
OPHOB : X=H OPHOFB : X-F
covalent bond
Complex
Transition temperature a (“C) SmC
SmA
38 43
-
-
28 57
-
Cr
10 OPHOB 11 OPHOFB a
0 0 0
0
31
0
I
N 0
33
-
0 (0
48 61
39 45)
0 0 0 0
Transition temperatures of monotropic phases are in parentheses.
OPHOB and OPHOFB, which are normal liquid crystals built only from covalent bonds, in Table 3 [ 191. The melting temperatures for the H-bonded liquid crystals are lower than those of the normal liquid crystals, and the wider temperature range of the
mesophase can be achieved for the H-bonded complex. This difference may be caused by “softness” of the complex structure. Ferroelectricity is induced by the self-assembly of a chiral benzoic acid which is nonmesogenic and 4-alkoxystilbazoles [20].
972
XVII
Hydrogen-Bonded Systems
Complex 12 exhibits a chiral smectic C phase from 109 to 123°C. The value of spontaneous polarization for 12 at 115 "C is 33.0 nC/cm2. A similar structure was reported to show an SmC* phase [21].
Molecular structures such as twins 13 [22] and angular structures 14 [23] have been prepared by the self-assembly of appropriate H-bonding components.
tion through the noncovalent interaction is observed in the diagram. The smectic A phase, which does not appear for either of the individual components, appears in the range of about 20-70 mol% of truns-1,2bis(4-pyridy1)ethylene. The highest clearing temperature is observed at the point where equimolar amounts of donor and acceptor moieties is present. This hydrogenbonded complex system can be obtained using a nonstoichiometric donor and acceptor mixure [ 11,13,16,24]. This concept can be used efficiently to control specific properties of H-bonded mesogenic materials.
13
180
14
.
160
4J
2.1.2 Phase Diagrams The formation of supramolecular 1:1 complexes with well-defined structures was described in the preceding chapter. However, mixtures containing any ratios of donor and acceptor moieties can be prepared because each component is independent. Figure 2 presents a binary phase diagram for 4-hexyloxybenzoic acid (4) and trans- 1,2-bis(4pyridy1)ethylene (2) [ 161. The significant effect of the mixing on mesophase stabiliza-
3
g 140
k
120
I
loo: 20
1
,
60 MoI% Of 4
40
80
100
Figure 2. Binary phase diagram for 2 and 4. (SA: SmA; K: Cr.)
2
2.1.3 Stability of Hydrogen Bonds It is noteworthy that the hydrogen-bonded complexes of carboxylic acids (pK, - 4) and pyridines (pK, = 9) exhibit thermally stable mesophases though the structure consists of noncovalent bonding [ll-13, 16-23]. In the infrared specta for the complex, the 0 -H band at 2500 cm-' is indicative of the strong hydrogen bond [16-18, 25, 261, while for benzoic acid dimers the 0 - H band appears at ca. 3000 cm-' [18, 271. Variable temperature infrared spectra for hydrogen-bonded liquid-crystalline complexes have shown that the structure of the hydrogen-bonded mesogen is stabilized by molecular ordering and packing in the liquid-crystalline state [16,18,27]. The stabil. ity of the hydrogen bond suddenly decreases once the complex becomes isotropic disordered [16, 18, 271.
2.1.4 Electrooptic Effects Room-temperature hydrogen-bonded liquid crystals consisting of 4-alkoxybenzoic acids and simple alkylpyridines align in twisted nematic cells and show electrooptic effects [ 191. This observation suggests that the hydrogen bonding is stable in electric fields. It is expected that the dynamic behavior of the complex is different from that of normal liquid crystals, because the core structure consists of noncovalent bonding, which is "soft" bonding between interacting molecular species [28].
Pyridine/Carboxylic Acid System
2.2 Self-Assembly of Polymeric Complexes 2.2.1 Side-Chain Polymers Biomacromolecules, such as nucleic acid, polypeptide, and cellulose, all have hydrogen bonding functional groups that play key roles for molecular recognition, molecular self-organization, and supramolecular selfassembly. Therefore hydrogen bonding would be useful for the molecular design of functional synthetic polymeric materials. The first approach to "supramolecular" liquid-crystalline polymers involved polyacrylates with a benzoic acid moiety through flexible spacers in the side chain [ l l , 131. Polysiloxanes have also been used to build side-chain polymeric structures [29, 301. Table 4 summarizes the thermal properties of hydrogen-bonded side-chain polyacrylates [ l l , 13, 17, 311, which exhibit stable mesomorphic behavior. These complexes show thermally stable mesophases. In particular, for complex 15, the mesophase is stable up to 252 "C [ 131. The ferroelectric rlnisrrnolscular Hydrogen Bond
CHIFH+m
d'-
15: R= - " * C ~ O C ~ ~ 1 6 : k -OCH3 17:R=
-CN
Table 4. Liquid-crystalline behavior of supramolecular side-chain polymers formed through intermolecular hydrogen bonding. a Polymeric complex 15 16 17 a
G
SmX
Transition temperatures in "C.
14 38
I
N
SmA 140
38
973
252 194 200
974
XVII
Hydrogen-Bonded Systems
side-chain polymer 18 has been formed by self-assembly of the H-bonding polysiloxanes and chiral stilbazoles [30]. Another example of a side-chain supramolecular polymer is prepared from poly(4vinylpyridine) and H-bonding side chains [32-351. In this case, mesogenic side-chain groups are directly attached to the polymer backbone and liquid-crystalline polymers such as 19 are formed by the noncovalent interaction [33]. This molecular design has been used to incorporate functional molecules into liquid-crystalline, host-guest polymeric systems [32]. fCH2-CHh
4 $0
'60
1
20
183°C. The behavior of such main-chain complexes was theoretically studied 1371.
2.2.3 Networks A liquid-crystalline network (21) has been constructed by the self-assembly of polyacrylate and 4,4'-bipyridine [38]. Although complex 21 has a fully cross-linked structure through the noncovalent interaction, it shows a glass transition at 95 "C and a subsequent smectic A phase up to 205 "C. The formation of a hydrogen-bonded intermo-
19
2.2.2 Main-Chain Polymers The concept of the hydrogen-bonded complexation has been extended to the preparation of main-chain type polymeric complexes by Griffin and co-workers [36, 371. Bifunctional H-bonding donor and acceptor moieties were used for the formation of a polymeric chain structure. Complex 20 exhibits a nematic phase from 174°C to
8 'N
21
3 Uracil/Diaminopyridine System
lecular mesogen should contribute to the induction of the stable mesophase. The smectic phase is viscous but shows shear flow. It should be noted that reversible phase transitions between the mesophase and the isotropic phase are observed for the network complexes due to the reversibility of the noncovalent interaction. The dynamic nature of the hydrogen bonding contributes to such behavior. For covalently bonded polymeric networks with a high density of crosslinks, the molecules are fixed in the solid state and no phase transitions are observed. Multifunctional low molecular weight compounds also form a liquid-crystalline network that is dynamic and thermally reversible [39, 401. Self-hydrogen bonding and dimerization of the benzoic acid side chain of polysiloxanes also induce network mesogenic structures [29].
3 Uracil /Diaminopyridine System 3.1 Low Molecular Weight Complexes Triply hydrogen-bonded mesogenic complexes can be obtained through the molecular recognition of complementary compo-
975
R
O d
22
nents of uracil and diacylaminopyridine moieties [14, 151. Complex 22 with R=Me, k = Z= 10, rn = 11, and n = 16 exhibits a monotropic mesophase below 72°C. For this complex structure, long alkyl chains are needed to induce mesomorphism. An X-ray diffraction study has shown that complex 22 forms a columnar hexagonal phase with a diameter of 40.4 A (4.04 nm). Each column contains two complexes side by side.
3.2 Polymeric Complexes Liquid-crystalline polymeric supramolecular species 23 has been prepared by the association of bifunctional H-bonding components [41, 421. These bifunctional individual components were derived from L(+), D(-), and meso (M) tartaric acids, respectively. It is interesting that the liquid-crystalline phases of the polymeric complexes consisting of chiral species are more stable than those of polymers consisting of meso components. The complex based on L-(+)tartaric acids shows a mesophase from room temperature to 254 “C, while the mesophase
R= &12,25 0
A,
Y \
976
XVII
Hydrogen-Bonded Systems
is observed from room temperature to 222 "C for the complex consisting of mesotartaric acids. The X-ray patterns show that a triple helix structure is organized by the complex ased on L-(+)-tartaric acid, while the complex consisting of meso-tartaric acid forms a twofold structure.
4 Miscellaneous Thermotropic H-Bonded Compounds by Intermolecular Interaction Doubly hydrogen-bonded supramolecular mesogenic complexes have been derived by the formation of the hydrogen bonds between 2-aminopyridyl and carboxylic acid moieties [43,44]. Complex 24 sharply melts
The interaction between phenol and pyridine, which is weaker than the carboxylic acid/pyridine interaction, has been used to effect mesomorphism [45-491. The 1 : 1 mixture (26)of 4-alkoxy-4-stilbazoles and 4-cyanophenol shows a monotropic nematic phase, while the stilbazoles exhibit a very narrow temperature range of ordered smectic phases and 4-cyanophenol is nonmesogenic [45]. This is due to the interaction between the phenol and pyridine moieties. The complex based on 3-cyanophenol (27)
A, #O c
24
Y
26: k H , Y=CN 27: X=CN, Y=H
CH3
at 90 "C and exhibits a monotropic smectic B phase from 83 to 67°C [43]. The structure of the complex, which does not have an overall rod shape, is unique for calamitic liquid crystals. This supramolecular mesogen is attached to the polyacrylate backbone [44]. For polymeric complex 25, a monotropic smectic phase was observed between 98 and 80 "C. It should be noted that in this polymeric structure the mesogenic group is expected to align perpendicular to the polymer chain.
i
5
shows more thermally stable mesophase than that based on 4-cyanophenol because molecular linearity is well kept for the metasubstituted system [46]. A phenohtilbazole complex (28) based on 2,4-dinitrophenol exhibits a smectic A phase between 97 and 117 "C [47]. Variable-temperature electronic spectroscopy shows that ionic hydrogen-bonded state by proton transfer is stabilized in the mesophase. The induction of smectic phases has also been observed for the mixture of 4,4'-bipyridine and 4-alkoxyphenol [44]. Liquid-crystalline polymer blends (29)involving phenol/pyridine interactions have been prepared from a polyester containing a pyridyl lateral group and poly(4-vinylphenol) [45]. The formation of mesogenic side-chain polymer complexes has been reported for pyridine-N-oxidelcarboxylic acid [46] and trialkylamine/phenol [47]. Complex 30 exhibits smectic phases between 29 and 123 "C ~461. CH3
30
Stabilization of a columnar phase for a calix[4]arene by a hydrogen bonding inter,But
BU',
Lyotropic Hydrogen-Bonded Complexes
977
action with a low molecular weight component has been described for bowlic liquid crystals [48]. Self-assembly of these host and guest molecules leads to mesomorphism.
5 Lyotropic HydrogenBonded Complexes Self-assembly and molecular recognition of some biomolecules result in the formation of lyotropic liquid-crystalline phases with water [49]. For example, nucleic acids such as DNA and RNA, which form supramolecular complexes, show lyotropic mesophases [49]. An interesting example of a biomolecule that shows lyotropic liquid-crystallinity by association of identical molecules is a guanosine residue that exhibits columnar mesophases due to the formation of hydrogenbonded tetramers [50]. Foric acid salts exhibit columnar liquid-crystalline phases in the presence of water [5 11. Synthetic supramolecular lyotropic polymer 31 has been built through triple H bonds. The polymer shows liquid-crystalline behavior with an organic solvent, whereas it melts with decomposition at high temperatures due to its rigidity [52]. BU'
978
XVII Hydrogen-Bonded Systems
References A. C. de Kock, Z. Phys. Chem. 1904, 48, 129. D. Vorlaender, Bea Dtsch. Chem. Ges. 1908,4I, 2033-2052. [V. Vill, LiqCryst - Database of Liquid Crystal Compounds for Personal Computers, 19941. A. E. Bradfield, B. Jones, J. Chem. SOC. 1929, 2660-2661. G. M. Bennett, B. Jones, J. Chem. SOC.1939,420-425. M. Gaubert, C. R. Acad. Sci. 1919,168,277-279. E. Fischer, B. Helferich, Justus Liebigs Ann. Chem. 1911,383,68-91. C. R. Noller, W. C. Rockwell, J. Am. Chem. Soc. 1938,60,2076-2077. G. A. Jeffrey, Acc. Chem. Res. 1986, 19, 168173. J. W. Goodby, Mol. Cryst. Liq. Cryst. 1984, 110,205-219. S. Diele, A. Madicke, E. Geissler, D. Meinel, D. Demus, H. Sackmann, Mol. Cryst. Liq. Cryst. 1989, 166, 131-142. K. Praefcke, P. Marquardt, B. Kohne, Mol. Cryst. Liq. Cryst. 1991, 203, 149- 158. 3. D. Bunning, J. E. Lydon, C. Eaborn, P. H. Jackson, J. W. Goodby, G. W. Gray, J. Chem. SOC.,Faraday Trans. 1 1982,713-724. Y. Matsunaga, M. Terada, Mol. Cryst. Liq. Cryst. 1986, 141, 321-326. J. MalthCte, A.-M. Levelut, L. Litbert, Adv. Matea 1992,4,37-41. H. Schubert, S. Hoffmann, J. Hauschild, I. Marx, Z. Chem. 1977,17,414-415. T. Kato, J. M. J. Frtchet, Macromol. Symp. 1995,98,311-326. T. Kato, J. M. J. Frtchet, J. Am. Chem. SOC. 1989,111,8533-8534. T. Kato, J. M. J. Frtchet, Macromolecules 1989, 22,3818-3819; 1990,23,360. J.-M. Lehn, Makromol. Chem., Macromol. Symp. 1993,69, 1- 17. M.-J. Brienne, J. Gabard, J.-M. Lehn, I. Stibor, J. Chem. Soc., Chem. Commun. 1989,1868- 1970. T. Kato, J. M. J. Frtchet, P. G. Wilson, T. Saito, T. Uryu, A. Fujishima, C. Jin, F. Kaneuchi, Chem. Matea 1993, 5, 1094-1100; T. Kato, P. G. Wilson, A. Fujishima, J. M. J. FrCchet, Chem. Lett. 1990,2003 -2006. T. Kato, H. Kihara, T. Uryu, A. Fujishima, J. M. J. FrCchet, Macromolecules 1992, 25, 6836 -6841. T. Kato, T. Uryu, F. Kaneuchi, C. Jin, J. M. J. FrCchet, Liq. Cryst. 1993,14, 1311-1317. T.Kato, M. Fukumasa, J. M. J. FrCchet, Chem. Matel: 1995, 7 , 368-372; T. Kato, M. Fukumasa, T. Uryu, J. M. J. FrCchet, Chem. Lett. 1993,65-68. T. Kato, H. Kihara, T. Uryu, S. Ujiie, K. Iimura, J. M. J. FrCchet, U. Kumar, Ferroelectrics 1993, 148,161 - 167.
[21] L. J. Yu, Liq. Cryst. 1993,14, 1303-1309. [22] T. Kato, A. Fujishima, J. M. J. FrCchet, Chem. Lett. 1990, 919-923. [23] T. Kato, H. Adachi, A. Fujishima, J. M. J. FrCchet, Chem. Lett. 1992, 265 -268. [24] H. Kresse, I. Szulzewsky, S. Diele, R. Paschke, Mol. Cryst. Liq. Cryst. 1994,238, 13- 19. [25] S. L. Johnson, K. A. Rumon, J. Phys. Chem. 1965,69,74-89. [26] S. E. Odinokov, A. A. Mashkovsky, V. P. Glazunov, A. V. Iogansen, B. V. Rassadin, Spectrochim. Acta 1976,32A, 1355-1363. [27] T. Kato, C. Jin, F. Kaneuchi, T. Uryu, Bull. Chem. SOC.Jpn. 1993,66,3581-3584. [28] S . Machida, T.-I. Urano, K. Sano, T. Kato, J. M. J. FrCchet, Langmuir 1997, 13, 576-580. [29] U. Kumar, T.Kato, J. M. J. FrCchet, J. Am. Chem. Soc. 1992,114,6630-6639. [30] U. Kumar, J. M. J. FrCchet, T. Kato, S. Ujiie, K.Iimura, Angew. Chem. 1992, 104, 15451547; Angew. Chem., Int. Ed. Engl. 1992, 31, 1531- 1533. [31] T. Kato, H. Kihara, S. Ujiie, T. Uryu, J. M. J. Frtchet, Macromolecules 1996,29,8734-8739. [32] T. Kato, N. Hirota, A. Fujishima, J. M. J. FrCchet, J. Polym. Sci., PartA: Polym. Chem. 1996, 34, 57-62. [33] C. G. Bazuin, F. A.Brandys, Chem. Muter. 1992, 4,970-972. [34] C. G. Bazuin, F. A. Brandys, T. M. Eve, M. Plante, Macromol. Symp. 1994, 84, 183- 196. [35] D. Stewart, C. T. Imrie, J. Muter. Chem. 1995, 5,223-228. [36] C. Alexander, C. P. Jariwala, C. M. Lee, A. C. Griffin, Macromol. Symp. 1994, 77, 283-294. [37] P. Bladon, A. C. Griffin, Macromolecules 1993, 26,6604-6610. [38] T. Kato, H. Kihara, U. Kumar, T. Uryu, J. M. J. Frtchet, Angew. Chem. 1994,106, 1728- 1730; Angew. Chem., Int. Ed. Engl. 1994, 33, 16441645. [39] H. Kihara, T. Kato, T. Uryu, J. M. J. FrCchet, Chem. Muter. 1996,8,961-968. [40] L. M. Wilson, Macromolecules 1994,27,6683 6686. [41] C. Fouquey, J.-M. Lehn, A.-M. Levelut, Adv. Muter. 1990,2,254-257. [42] T. Gulik-Krzywicki, C. Fouquey, J.-M. Lehn, Proc. Natl. Acad. Sci. USA 1993, 90, 163-167. [42a] T. Kato, Y. Kubota, M. Nakano, T. Uryu, Chem. Lett. 1995, 1127-1128. [42b] T. Kato, M. Nakano, T. Moteki, T. Uryu, S. Ujiie, MacromoZecuZes 1995,28, 8875 - 8876. [43] D. W. Bruce, D. J. Price, Adv. Muter. Opt. Electron. 1994,4,273-276. [43a] K. Willis, D. J. Price, H. Adams, G. Ungar, D. W. Bruce, J . Matea Chem. 1995, 5 , 21952199.
6 References [43b] D. J. Price, T. Richardson, D. W. Bruce, J. Chem. Soc., Chem. Commun. 1994,1911 - 1912. [44] T. Koga, H. Ohba, A. Takase, S. Sakagarni, Chem. Lett. 1994,207 1 - 2074. [45] A. Sato, T. Kato, T. Uryu, J. Polym. Sci., Part A, Polym. Chem. 1996,34,503- 505. [46] U. Kumar, J. M. J. FrCchet, Adv. Muter: 1992,4, 665-667. [47] S . Malik, P. K. Dhal, R. A. Mashelkar, Mucromolecules 1995,28, 2159-2164. [48] B. Xu, T. M. Swager, J. Am. Chem. Soc. 1995, 117,5011-5012.
979
[49] S. Hoffrnann in Polymeric Liquid Crystals (Ed.: A. Blumstein), Plenum, New York, U. S.A. 1985, p. 423. [50] P. Mariani, C. Mazabard, A. Garbesi, G. P. Spada, J.Am. Chem. Soc. 1989,111,6369-6373. [51] F. Ciuchi, G. D. Nicola, H. Franz, G. Gottarelli, P. Mariani, M. G. P. Bossi, G. P. Spada, J. Am. Chem. Soc. 1994,116,7064-707 1. [52] M. Kotera, J.-M. Lehn, J.-P. Vigneron, J. Chem. Soc., Chem. Commun. 1994, 197-199.
Handbook of Liquid Crystals Edited by D. Demus, J. Goodby, G. W. Gray, H.-W. Spiess, V. Vill Copyright 0 WILEY-VCH Verlag GmbH. 1998
Chapter XVIII Chromonics John Lydon
1 Introduction 1.1 A Well-Defined Family Distinct from Conventional Amphiphiles Over the last twenty years it has become increasingly clear that there is a well-defined family of lyotropic mesogens embracing a range of drugs, dyes, nucleic acids, antibiotics, carcinogens, and anti-cancer agents. In contrast to conventional amphiphile mesogens, such as soaps, detergents, and biological lipids, these chromonic materials do not generally have significant surfactant properties. Their molecules are disc-like or plank-like as opposed to rod-like, they are aromatic rather than aliphatic, and the hydrophilic ionic or hydrogen-bonding solubilizing groups are arranged around the peripheries of the molecules, not at the ends. The molecules can be regarded as being insoluble in one dimension, and the basic structural unit of all chromonic phases is a molecular stack of some kind, rather than a micelle. Chromonic systems have characteristic phase structures, phase diagrams, optical textures, X-ray diffraction patterns, and miscibility properties. The tendency of chromonic molecules (Fig. 1) to aggregate into columns is present even in dilute solution (just as for amphiphile systems, where micelle formation occurs before a mesophase is formed). How-
disodium cromogylcate
\
RU31156
Sinus Supra Brown RLL
methyl orange COG Na'
Cob Na*
Copper-teeacarboxyphthalocyanine Figure 1. A selection of chromonic molecules: The antiasthmatic drug, disodium cromoglycate (DSCG), the antiallergic drug RU 31 156, the dye, Sirius Supra Brown RLL, the dye, methyl orange, and copper tetracarboxyphthocyanine.
982
XVIII
Chromonics
ever, although there may be a threshold concentration before aggregation begins to occur, there is no optimum column length and hence no critical concentration directly analogous to a cmc. A further distinction is the absence of a Krafft temperature. Since the process of mesophase formation does not depend on the presence of flexible alkyl chains, there is no threshold temperature below which mesophases cannot form because the vital flexibility of the molecules has been frozen out.
M
N
a i
ii
b C
1.2 The Chromonic N and M Phases There are two principal chromonic mesophases, which have become known as the N and M phases (Fig. 2) (although there are some rarer additional phases). The designations N and M date back to the early polarizing microscopy studies of the mesophases of the drug disodium cromoglycate (DSCG) [l, 21. The N phase was so-named because it forms a schlieren texture, similar to that shown by the thermotropic nematics. X-ray diffraction studies indicate that it is a nematic array of columns and that the only regular repeat distance is the stacking repeat of 0.34 nm along each column [3]. The M phase was so called because it can form an optical texture similar to that of the hexagonal H I (middle) soap phase. X-ray diffraction investigations show that in this phase the columns are arranged on a hexagonal lattice.
'
Note that the widely-used name "cromoglycate" was invented for patent reasons and does not conform to IUPAC nomenclature.
Figure 2. The structure of the chromonic N and M phases: The basic structural unit of both phases is the untilted stack of molecules. The N phase is a nematic array in which these stacks lie in a more or less parallel pattern, but where there is no positional ordering. The M phase is a hexagonal array of these columns. The six-fold symmetry is a result of orientational (but not positional) disorder. A schematic diagram of a localized region, as shown in (a) has only orthorhombic symmetry, but, averaged over the whole structure, each column actually lies in a site with sixfold symmetry (b). The restrictions to the possible orientations of the columns are shown in (c). Because of packing considerations, for any particular orientation of a column, as shown on the left, an adjacent column (right) can take up only two of the three possible orientations (i) and (ii). A representation of the orientationally disordered state of the M phase is given in Fig. 9. Note that the molecular columns are shown here in a highly stylized way. They are not necessarily such simple one-molecule-wide stacks.
1*3 Drug and Dye Systems A general correspondence between the optical textures, the X-ray diffraction patterns, and the forms of the phase diagrams has been observed for chromonic drugs and many dye/water systems [4-71. The only significant difference between these two families of mesogens appears to be in some optical textures and rheological properties
2 The History of Chromonic Systems
of the M phase (and I suggest that these distinctions are essentially quantitative rather than qualitative). Whilst the drug M phases usually form a well-defined herringbone texture and flow in a homogeneous manner when mechanically disturbed, some dye M phases tend to form a poorly developed grainy texture and flow as distinct, fairly rigid domains. Such behavior suggests a phase tending towards a more gel-like state. However, this need not be regarded as a fundamental distinction and may be due to the difference in dimensions of the two sets of mesogens - the dye molecules being, in general, two or three times as long as those of the chromonic drugs.
1.4 Molecular Structure of Chromonic Species At one time it was thought that DSCG was a unique mesogen and that the V-shaped structure of this molecule was vital for its mesogenic properties [3]. The subsequent discovery of other mesogenic drugs with simpler structures made it clear that this view was wrong [ 5 ] . It appears that the requirements for chromonic behavior are straightforward, i.e., the molecules must have sufficiently large aromatic regions to have a tendency to stack in columns and they must carry solubilizing groups to enable the columns to disperse in water. Just as for amphiphile systems, the solubilizing groups can utilize the ability of water to dissolve either polar or hydrogen-bonding groups (i.e., they can be ionic or nonionic). The majority of chromonic drugs and dyes studied to date are ionic, hence the major emphasis in this account will lie with such systems, but this should not be taken to imply that nonionic chromonic systems (such as aromatic nuclei with peripheral polyethylene-
983
oxide chains [S]) are likely to prove less important. A survey of chromic species, especially dyes, makes it clear that there is no need for the molecules to even approximate to circular discs, as might be expected if chromonic systems are regarded merely as lyotropic discotic phases; extended blade-like molecules are capable of forming chromonic mesophases with apparently perfect hexagonal order.
2 The History of Chromonic Systems Up until the last twenty years, the family of chromonic mesophases lay hidden like an uncharted continent. Occasionally explorers returned to give some partial account, but their reports usually fell on deaf ears, and they were largely unaware of each other’s discoveries. A coherent overall picture was remarkably late coming. It is significant that in the 500 pages of the two-volume earlier attempt at a comprehensive survey of liquid crystals by Gray and Winsor, published in 1974, there is no mention whatsoever of the systems we now call chromonic [9].
2.1 The Early History The story of the three-way connection between drugs, dyes, and nucleic acids begins in the early decades of this century when the optical microscope was still the foremost research tool in mineralogy, biology, and medicine. The use of specific stains for highlighting and identifying biological tissue had become a highly developed art form. It was apparent that stains could be highly selective for particular tissues and
984
XVIII Chromonics
for distinguishing between different microorganisms [ lo]. Chromosomes were sonamed, not because they were naturally colored, but because they could be readily stained. Gram-positive and gram-negative bacteria were distinguished by the way in which they accepted or rejected a particular stain. Slowly, the idea grew that, since some reagents could specifically stain harmful bacteria, perhaps they could specifically kill them also. The concept of “Dr Ehrlich’s magic bullet” was born. Ehrlich and co-workers began a search through the range of artificial dyes then available, hoping to find some with medically useful antibacterial properties [ 111. They achieved some success with the dye, trypan red which cured trypanosomiasis in mice and rats. (Tyrpanosomes are parasitic protozoa causing such diseases as sleeping sickness.) This started a large-scale search for other medically effective dyes, leading directly to the first generation of wonder drugs, the sulphanilamides. Note that the British synthetic dye industry was accidentally born in an almost mirror-image situation in 1856 when the young Perkin was attempting to synthesize the antimalarial drug quinine and instead produced the dye aniline purple, better known as mauve The first report of the mesogenic properties of dyes appears to have come from Sandquist in 1915. He was investigating the properties of 10 chloro- and 10 bromophenanthrene-3 (or 6)-sulphonic acid and he noticed ‘nematic threads’ in a solution of this material when viewed under polarized light [ 131. The Bayer company then produced the antitrypanosomiasis drug Bayer 205, which was examined by Fourneau et al. [14]. Following this, Balaban and King investigated the properties of Bayer 205 and its derivatives in an attempt to discover the mode of action. In their subsequent paper, a short dis-
cussion was included describing the mesogenic properties of the drug and associated derivatives [15]. All the mesogenic compounds were salts of disulphonic acids. Commenting on their optical textures they said: “The intermediate, anisotropic fluid appears to posses a certain amount of rigidity, but always flows when under pressure of a coverslip”, and noted “alternations of dark and light patches”. Between 1933 and 1935, a succession of liquid crystalline dye solutions were reported by Gaubert [16]. This was immediately followed by studies carried out independently by Jelley [ 171 and Schiebe et al. [ 181 on the aggregation of pseudo-isocyanine dyes in solution. Jelley describes the appearance of solutions of 1: l’diethyl-p-cyanine chloride in terms of “- a fine thread-like structure of a brilliant greenish yellow colour showing streaming birefringence”. He concluded that the mesophase structure had one-dimensional order and that the molecules had either “rotational freedom or random orientation” and that the mesophase was “analogous to the (thermotropic) nematic phase described by other workers”. Although his paper does not include a micrograph of this texture, from the description given it is clear that the threads are actual structures (probably M ribbons) rather than lines of disclination within the single phase nematic region.
2.2 Disodium Cromoglycate and Later Studies After the work of Jelley, no significant work appears to have been done in the field of chromonics for thirty years. In 1967, following the epic (and courageous) investigations of Althounyan [19], Fisons began to market the new antiasthmatic drug disodium cro-
2
moglycate (DSCG) under the trade name INTAL. At the instigation of Cox, the McCrone Research institute in London began studies aimed at maximizing the fluidity of the inhalant and improving the dosage formulations. This involved an optical and X-ray diffraction study of the mesophases and the hydrated crystals. Cox et al. noted the great affinity that DSCG had for water and they described two mesophases, the N and M phases observed using optical microscopy, eventually producing the peritectic phase diagram now regarded as characteristic of the simplest type of chromonic system [l]. As a postscript to their paper they remarked upon the 3.4 8, (0.34 nm) spacing found in the nematic phase (but did not identify this with a stacking repeat distance) and noted that the inner reflections in the diffraction pattern of the M phase correspond to spacings in ratios characteristic of a hexagonal phase. In the following year a more detailed optical and X-ray diffraction study by Hartshorne and Woodard was published, proposing models for the N and M phases [2], but there was no immediate follow-up. The DSCG mesophases were quietly relegated to the oddity pile. The DSCG/water system was reconsidered in 1980 by Lydon [3] and reinvestigated experimentally in 1984 by Attwood and Lydon. New models for the N and M phases were proposed and, when the study was extended, it was found that a similar pattern of mesophase formation occurred for a number of other antiasthmatic and antiallergic drugs [5]. The 7,7’ analog of DSCG was extensively studied by Perahia et al. and found to have closely similar mesogenic properties [20]. There have been more recent reports of other mesogenic drugs, such as nafoxidine chloride (by Mlodozeniec [21]), nafcillin sodium (Bogardus [22]), diethylammonium flufenamate (Eckert and Fischer [23]), and
The History of Chromonic Systems
985
fenprofen sodium (Rades and Muller-Goymann [24]), where the mesophases are apparently lamellar. For all of these compounds, the molecules look distinctly chromonic and some form of lamellar brick-wall structure would appear likely. The distinction between these systems and those of the DSCG type is reinforced by the observation of Kustanovitch et al. [25] who found that diethylammonium flufenate forms a mesophase with positive diamagnetic susceptibility, implying that the molecular planes lie parallel to the director. (This is in contrast to the negative diamagnetic susceptibility found by Goldfarb et al. [26] for DSCG mesophases where the director is parallel to the column axes, i.e., perpendicular to the molecular planes.) More recently, following a preliminary survey using polarizing microscopy, each of the two azo dyes, Sirius Supra brown and “acid red”, was shown to form a single lyotropic mesophase at room temperature [4]. The latter system in particular was studied in some detail and, on the basis of its optical texture, the mesophase formed was characterized as nematic. Moreover, on heating and cooling, arichvariety of optical textures was exhibited and the overall form of the phase diagram appeared to be complex. It was the similarities between the mesogenic properties of these dyes and DSCG that gave rise to the concept of the family of chromonic phases [6].
2.3 The 3-Way Link Between Drugs, Dyes, and Nucleic Acids In view of the number of studies of dye aggregation and mesophase formation, and the presence in the literature of reports of liquid-crystalline drug solutions, the idea
986
XVIII
Chromonics
that there was a common pattern of behavior came surprisingly late. Once the point had been made, however, it was natural that studies of mixed drug/dye systems would follow 127-291. Similarly, the “formal” inclusion of nucleic acids into the chromonic family occurred later than might have been expected considering that there was a growing amount of literature describing nucleic acid mesophase formation in liquid crystal journals [30-331. The first general review of chromonic mesophases (by Vasilevskaya et al.) was published in 1989 [34].
3 The Forces that Stabilize Chromonic Systems 3.1 Hydrophobic Interactions or Specific Stacking Forces? A simple view of the state of a chromonic system is that the molecules are effectively insoluble in one dimension. The energy involved in their aggregation is more or less comparable to kT for each aromatic ring and only at extremely low dilution are there individual, unaggregated molecules. For ionic systems, the columns attract counterions, forming electrical double layers. This results in column-column repulsion, which causes the columns to behave as if they were separated by compressed springs. This effect presumably stabilizes the hexagonal structure of the M phase and causes the columns to move apart equally when the mesophase is diluted. It was natural that early workers on chromonic systems would attempt to explain their properties in terms of the then famil-
iar picture of molecular association and amphiphile mesophases, and there were repeated references to “hydrophobic interactions” and “micelle formation”. However, there has been a growing realization that the forces that cause chromonic molecules to stack, although in a sense “hydrophobic” (because water is excluded from the space between stacked molecules), must be radically different. The concept of hydrophobic “bonding” is a convenient myth (especially useful for physical chemists when talking to biologists). It pictures attractive forces between “oily” groups such as alkyl chains, which cause them to cluster together in the interiors of micelles and membranes. In reality, however, these are not the dominant interactions. The major interactions are those between water molecules. It is so thermodynamically unfavorable for a water molecule not to be in contact with other water molecules that hydrophobic species are excluded from the aqueous part of the phase. They are squeezed out and the process of micelle formation is said to be solvent-driven. Hydrophobicity is the absence of hydrophilicity [35]. Taken to (or perhaps beyond) its logical conclusion, this picture would suggest that there can only be one kind of hydrophobicity and that in mixed systems all hydrophobic molecules should be forced to aggregate together. This does not appear to be the case. Hydrophobic aromatic and aliphatic species do not in general mix or aggregate together in the presence of water. It appears that there are effectively two distinct types of hydrophobicity. Although one case study is hardly definitive, an investigation of a mixed amphiphile and chromonic system gave no evidence that there is any tendency for mixed micelles to form [36]. In addition, it appears that the association of chromonic molecules in solution is exothermic (i.e., en-
3 The Forces that Stabilize Chromonic Systems
thalpically driven). It must be concluded therefore, that the interactions between parallel n-systems set aromatic systems apart from aliphatic ones [37].
3.2 The Aggregation of Chromonic Molecules in Dilute Solution and on Substrate Surfaces The formation of conventional amphiphile mesophases is preceded by the aggregation of molecules in solution. Groups of two or three amphiphile molecules associate with reluctance. There is still an unfavorably large fraction of the hydrophobic parts of the molecules exposed to water. This situation continues until the formation of aggregates approaching the size of a micelle (Fig. 3).
--
aggregation of amphiphiles
aggregation of chromonic molecules
Figure 3. The contrast between the pattern of aggregation of amphiphiles and chromonic molecules: For amphiphiles, the energetically unfavorable contact between the hydrophobic alkyl chains and water molecules is reduced to more or less zero when a closed micelle is formed. In contrast, for chromonic systems, the fraction of the total molecular surface exposed to the solvent decreases as the columns lengthen, but there is no optimum aggregate size directly comparable to a micelle.
987
There is then a sudden drop in the free energy as virtually all of the amphiphile molecules are incorporated into micelles, enabling their hydrophobic alkyl chains to be more or less completely shielded from the aqueous part of the phase. This leads to the familiar abrupt change in physical properties at the cmc. In chromonic systems there is also the aggregation of molecules in dilute solution before mesophase formation, but the pattern of association is different. The hydrophobic surfaces of the molecules cause them to aggregate in stacks like packs of cards. As these stacks grow, the fraction of the total hydrophobic surface area exposed to the aqueous part of the phase steadily falls, but there is no minimum free energy state, no cmc, and there is no structure directly comparable to the micelle [38]. There is widespread evidence of the association of dye molecules in solution (and of the tendency of many dyes to be adsorbed on substrates as aligned, structured aggregates rather than as individual molecules). Although the bulk phase may be isotropic, the increasing asymmetry of the aggregates causes flow alignment and there is often an easily detectable flow birefringence. Spectroscopic studies such as those of Stegemeyer and Stoke1 (on pseudoisocyanine dyes systems) confirm the presence of aggregates in solution above the N-I transition [39]. The changes in the absorption spectra of dye systems with concentration are of major interest to the dye industry (and to the photographic industry, because of the use of adsorbed dyes as photographic sensitizers). The situation is complex, varying from system to system, with changes occurring in both the intensity and the wavelength of maximum absorption. A range of different aggregate structures have been postulated to explain these effects. Three distinct spectroscopic effects have been identified, and re-
988
XVIII
Chromonics
peated attempts have been made to explain these in terms of different patterns of molecular aggregation [40,41]: 1) A change to shorter wavelength accompanying the formation of dimers (the D band), 2) a bathochromic shift (towards the red), thought to accompany the formation of tilted (“staircases”) termed “J (Jelley) [ 171 aggregates”, and 3) a hypsochromic shift (towards the blue) thought to accompany the formation of untilted or less tilted stacks (“ladders” or “H aggregates”).
In completely separate investigations, early osmometry studies of the association of the nucleic acid bases and nucleosides in solution indicated that the molecules were also aggregating in columns [42]. Sedimentation equilibrium studies showed that this process is reversible and that there is a near constant free energy increment when each additional molecule is added, irrespective of the existing length of the column. Such behavior was termed “isodesmic”. For purine and pyrimidine nucleosides, the association constants indicate weak associations where both A H and A S are negative (i.e., the aggregation process is enthalpically rather than entropically driven). The standard free
Free Energy
I
energy change is of the order of the thermal energy, k T (- 2.5 kJ per mole). Compared with the complicated mass of literature concerning the aggregation of dyes, that for chromonic drugs is sparse, but there is a study by Champion and Meeten [43] on the aggregation of DSCG in solution, dating back to the time of the Cox, Woodard, McCrone and Hartshorne investigations and an infrared study of the same system by Hui and Labes [44]. Both of these confirm the growth of chromonic aggregates in solution. Simple isodesmic behavior (Fig. 4) is expected to be characteristic of chromonic systems, and this appears to be generally true. However, the situation can be more complex. Extensive studies of dye/water systems have shown that there is sometimes a distinct threshold concentration with some resemblance to the amphiphile cmc, where dimers first appear [41].
4
Phase Diagrams
The classic chromonic phase diagram is that of the DSCG/water system. This has a double peritectic form in contrast to the standard multi-eutectic form of conventional
Free Energy Micelle Formation IncreasingCnlumn Length ---c
Aggregation Number Ic
Non-Isodesmic Hehaviour
Aggregation Number
+
Isodesmic Rehaviour
Figure 4. Isodesmic behavior: For isodesmic behavior, the addition of each molecular unit to a stack is accompanied by a constant free energy increment. There is therefore a gradual fall in free energy with column length. This is in contrast with the (nonisodesmic) behavior of amphiphiles, where micelle formation occurs and a particular aggregation state represents a pronounced free energy minimum.
5
amphiphile system phase diagrams. The distinction between these two patterns of behavior can be explained in terms of the different patterns of molecular aggregation. For conventional amphiphiles (at a particular pressure and temperature), each mesophase has an optimum composition corresponding more or less to the peak in the phase diagram. In the hexagonal HI phase, for example, the cylinders have an optimum diameter and there will be an optimum inter-cylinder separation. Higher or lower mesogen content strains the system and lowers the free energy and hence the HIisotropic clearing temperature. The picture for chromonic systems is different: as the phase diagram is traversed from left to right, it is not a matter of micelle-building, destruction, and rebuilding in a different manner, there is a progressive increase of order: first the aggregation into columns and then increasing levels of ordering of these columns (as indicated by the progressive rise in the clearing temperatures). As is usually found for mesophase systems, the boundaries on the phase diagram are not all equally easy to detect. All of the known chromonic mesophases are birefringent and the clearing points can therefore be readily seen by optical microscopy. The crystalline solid-M phase boundary is often harder to detect optically, but it does show up readily in X-ray diffraction studies. Note also that some early chromonic phase diagrams determined by 2H NMR did not show the N + I +M + I boundary [45]. At first sight there appears to be a paradoxical feature in the standard chromonic phase diagram (Fig. 5). This concerns the N + I/M + I boundary, where on heating the N phase regions become M, i.e., a less ordered phase changes to a more ordered phase. But note that the two mesophases do not have the same composition, i.e., the M phase is more concentrated than the N. This
70
-
60
.
989
Optical Textures
50. 40
-
30
-
20
.
M+N , 10
.
.
.
.
, 20
.
,
,
.
i o '
"
Concentration
40
"
( % w/w )
Figures. The phase diagram of the DSCG/water system (as determined by hot stage polarizing microscopy). The double peritectic form of this diagram has come to be regarded as the classic chromonic pattern. Note, in particular, the characteristic sequence of phases formed by samples within the composition range indicated by the shaded band.
process is an inevitable consequence of the way in which, as the temperature rises, the phase in equilibrium with the isotropic liquid becomes progressively less hydrated. At low temperatures it is the N phase, at higher temperatures the M, and at still higher temperatures, the crystalline solid. The double peritectic form of the chromonic phase diagram often produces an interesting feature not found for conventional amphiphile systems. Samples of DSCG for example, with a particularly narrow composition range, pass on heating through the phase sequence M + M + N + N - + N + I + M + I + I . Thus one single sample can display all of the one-phase and two-phase liquid crystalline regions of the system.
5 Optical Textures In optical examinations of chromonic systems, the textures of both the one-phase and the two-phase regions tend to be highly characteristic [46]. The N phase usually
990
XVIII
Chromonics
assumes a nematic schlieren texture, but homeotropic or homogeneous samples can be prepared by surface alignment. Occasionally an N phase sample will form the striated ‘tiger skin’ texture (Fig. 6). We interpret this texture in terms of a twisted rope-like arrangement of the columns. It appears that such an arrangement represents the way in which any initial misalignment in the system can be accommodated, whilst at the same time keeping the splay distortion to a minimum [47]. The characteristic M texture is much more grainy in appearance than the N schlieren. One of the sources of confusion to early workers in the field was the way in which chromonic textures depend on the past history of the sample. A one-phase or twophase region may have dramatically different textures depending on whether it has been reached by heating or cooling, by con-
Figure 6. The explanation for the tiger skin texture sometimes adopted by chromonic N (and P) phases: The banded pattern seen with crossed polars is thought to result from a twisted rope-like arrangement of the columns. Such a pattern could be a way of accommodating initial misalignment in the sample with a minimum of splay distortion. (a) The postulated arrangement of molecular stacks in two adjacent twisted ropelike assemblies. (b) The gross structure of the sample showing the large-scale organization of the assemblies represented in (a). (c) The banded appearance of a sample when viewed vertically between crossed polars.
centration or by dilution. When the N phase is heated and passes into the two-phase N + I region, more or less circular islands of isotropic phase appear, giving the reticulated texture. For compositions where the N + I becomes M + I on heating, the continuous network develops a grainy texture. Cooling produces different textures. The I-N transition produces isolated birefringent droplets which grow and coalesce to produce the bulk N phase. In more concentrated samples, which undergo the transition from I to M + I, the M phase appears as birefringent ribbons. M phases grown by allowing an N phase sample to become more concentrated by evaporation can develop the detailed herringbone texture2 (Fig. 7). This is very similar to the texture adopted by the middle (HI and H2) phases of conventional amphiphiles, and a similar explanation of its origin is proposed [48]. As the M phase develops, the columns are packing closer and growing in length. It is energetically more favorable to add molecules to an existing column than to create another column. This causes mechanical strain to build up and eventually this is relieved by the structure buckling and fracturing, creating a series of “geological” fault lines. Both N and M phases are uniaxial and optically negative. The N phase is uniaxial because of the nematic ordering (i.e., the random orientations of the columns about their long axes and the random column-tocolumn separations). The M phase is uniaxial as a consequence of the transverse hexagonal ordering. The negative sign indicates that in both mesophases the planes of the Note that the term “herringbone” is used in two different contexts in chromonic systems: at the molecular level, to describe the transverse arrangement of columns in the M phase, and on a much larger scale, to describe the diagonally striped optical texture of the M phase grown by concentration of the N.
6 X-Ray Diffraction Studies
99 1
there is a striking correspondence between the range of optical textures exhibited by this system and those of chromonic N and M phases (N-schlieren, reticulated N + I, Ndroplets in I, herringbone-M, etc.). In retrospect, this appears very reasonable, since the mesogenic units in both cases are long semi-rigid columns, but note that 40 years later, some workers were still talking in terms of smectic patterns of aggregation to explain M-phase optical textures. Figure 7. The origin of the herringbone texture of the M phase: This texture develops when the M phase is formed by the concentration of the N phase (by allowing water to evaporate from the edge of the sample). The explanation is the same as that proposed to explain the similar texture of the amphiphile H , phase. It is easier to extend a column than to create a new one. This leads to strain in the specimen which is relieved by the creation of multiple fault lines, giving the herringbone appearance.
molecules lie more or less perpendicular to the director (i.e., the column axis). The birefringence due to the molecular orientation would be expected to be reduced to some extent by the form birefringence of the columns. The way in which an M + I sample formed by heating an M phase has a continuum of N, whereas the N + I phase formed by cooling the N has a continuum of I, appears to indicate that the surface tension at the N/I interface is relatively low. An early paper, which could have been most influential, was apparently overlooked by most of the liquid crystal fraternity (possibly because of the journal in which it was published) [49]. This described the work of Bernal and Fankuchen on the aqueous mesophases formed by suspensions of tobacco mosaic virus. These virus particles are long cylindrical assemblies [ 180 A (18 nm) in diameter and -3000 A (300 nm) long] and
6 X-Ray Diffraction Studies Chromonic phases can in general be oriented by surface alignment or magnetic fields, but when samples are introduced into untreated glass tubes for an X-ray study, there is usually sufficient spontaneous alignment for axial and equatorial reflections to be easily distinguished. (In rare cases a sample in the M phase will form a large scale zig-zag texture like that observed by Bernal and Fankuchen for TMV solutions [49].) The diffraction patterns of both the N and M phases show fairly sharp 3.4 A (0.34 nm) axial arcs. This spacing corresponds to the thickness of an aromatic ring and indicates that the molecules lie in untilted columns in both mesophases. (Accordingly, the position of this reflection does not vary with dilution of the sample and tends to be constant from one mesophase to another.) The 3.4 A reflection is not as sharp as a typical reflection from a crystalline solid (but it is appreciably sharper than the broad stacking repeat reflections of the D,, phase). From the width of this peak, Goldfarb et al. inferred that that the column lengths in the N phase of DSCG range from 80-200nm [ S O ] . However, it must be admitted that since the
992
XVIII
Chromonics
peak profile is a function of both the column length and any local irregularities of spacing, these values must reflect both of these parameters. In addition, there are indications from NMR studies that, for the DSCG/ water system at least, the assembly of molecules into columns is “complete” in the N phase (i.e., there is no sudden increase in column length at the N +M transition) 1441. The N phase gives a diffuse inner axial reflection corresponding to the continuous range of column-to-column distances. The M phase, in contrast, gives a series of sharp axial reflections corresponding to spacings in the ratio 1:I / d 3 : l&: 1 / f i: l/&, which is typical of a hexagonal lattice (Fig. 8). The hexagonal spacing varies from mesogen to mesogen and increases linearly with the square root of the dilution (indicating that the structure is swelling in two dimensions only and that the column length is effectively infinite). Note that in the current model for the M phase, although the centers of the columns lie on a hexagonal lattice, the herringbone
N
M
Figure 8. Drawings of the X-ray diffraction patterns of N and M phases: In both diffraction patterns, the axial (vertical) arcs correspond to the 3.4 (0.34 nm) stacking repeat distance along the columns. In the N phase patterns, the diffuse low angle equatorial reflections correspond to the range of column-column separations. The sharp low angle equatorial reflections of the M phase correspond to spacings in the ratio 1:l / f i : l/h: l / f i (characteristic of a hexagonal lattice). The inner reflections change with dilution. Detailed studies confirm that the M phase structure swells in two dimensions as more water is added.
array itself has only orthorhombic symmetry. To explain this discrepancy, an analogy had been drawn with the arrangement of molecules within a layer of the smectic B structure. In this case comparison of the molecular dimensions with the observed hexagonal spacing appears to leave little doubt that there is a local herringbone arrangement, and that the higher hexagonal symmetry must therefore result from a pattern of rotational disorder. It is the space-averaged (and to a lesser extent, time-averaged) picture rather than the short range snapshot view, which has hexagonal symmetry. (As a corollary to this picture, there appears to be evidence of an orthorhombic chromonic 0 phase of certain dyes, well-formed, homeotropic, hexagonally shaped monodomain plates of the M phase grow into butterfly shapes with two-fold symmetry at the M-0 transition [46].) In addition to the outer axial 3.4 A reflection and the inner axial equatorial hexagonal reflections, M phase diffraction patterns often show diffuse axial and off-axial reflections in the 0.4-2 nm range 1511. These are difficult to interpret. They must relate to some structural feature of the stacks, but they do not generally lie on the layer lines corresponding to multiples of 3.4 8, and hence can not be explained in terms of simple regular rational repeats along the stack. The history of the structural studies of the DSCG/water mesophase system is unusually complex. In retrospect it can be seen that the interpretation of the early X-ray diffraction results was complicated by three factors: 1) The misleading picture given by the crystal structure 1521. In the crystalline solid the molecules are not flat. They are butterfly-shaped with an angle of 52” between the two chromone rings. It was uncertain whether the molecules had the same conformation in the mesophases or
7 The Extended Range of Chromonic Phase Structures
whether the repeat distance along the stack of molecules in the M phase would be the same as that in the crystalline solid. 2) The apparatus used was a specially constructed low angle powder camera. This gave one-dimensional information only and did not indicate the geometrical relationship between the high angle and low angle reflections. 3) The 3.4 A reflection was not at first identified with the column repeat distance and the original picture suggested for the N phase was of a nematic discotic array of individual free molecules.
7 The Extended Range of Chromonic Phase Structures 7.1
The P Phase
In some dye/water systems the N-M transition is difficult to detect by optical micros-
N / Distinct Phase Roundary
993
copy. The textures are very similar and in the two-phase regions, fine details within the textures often extend across the phase boundaries from one phase into the other. When the coverslip is mechanically disturbed, there is no obvious discontinuity in viscosity to distinguish the two phases. This gives the impression that the N/M transition has become so weakly first order that it is virtually second order, and the N and M phases have effectively amalgamated into one. This combined phase has been termed the P phase [47,53, 541. A plausible (but as yet unsubstantiated) explanation for the distinction between N/M and P type behavior lies in terms of the effective cross sections of the columns in the two cases (Fig. 9). For conventional N/M behavior, the columns in the M phase are pictured as being blade-like and lying in a herringbone pattern. The transition from N to M therefore involves the loss of both lateral positional order and orientational order. In contrast, the P phase columns are pictured as being more or less circular and the N-M transition need only involve the loss of positional order.
2
1
0
_____.
Indistinct Phase Boundary
3
A
0
M
N/M
P
Figure 9. The postulated distinction between NIM and P-type chromonic behavior: This figure shows a schematic view of NIM and P-type phases viewed down the column axes. The distinction is explained in terms of the effective cross-sectional shapes of the columns. In the P system it is thought that the columns are effectively circular (and are able to rotate more or less independently), whereas in NIM systems they are more blade-like (and can only rotate co-operatively).
994
XVIII
Chromonics
I
The schlieren texture of the N phase. (DSCG/water, x300, crossed polars + 1 il red plate.)
11
The tigerskin texture of the N phase. (DSCGlwater, x300, crossed polars+ 1 1 red plate.)
111 The typical grainy texture of the M phase. (DSCG/water, x300, crossed polars + 1 il
red plate.)
IV
The herringbone texture of the M phase obtained by allowing a sample in the N phase to become more concentrated by peripheral evaporation. , polars+ 1 1 (DSCG/water, ~ 3 0 0crossed red plate.)
7 The Extended Range of Chromonic Phase Structures
995
V
The reticulated texture of a two-phase N + I sample obtained by heating the N phase. (DSCGlwater, x300, crossed polars + 1A red plate.)
VI
The grainy reticulated texture of a twophase M + I sample obtained by heating a sample with the reticulated N + I texture. (DSCG/water, x300, crossed polars+ 1A red plate.)
VII The two-phase N + I droplet texture formed on cooling an I phase, with droplets of N in an isotropic continuum. (DSCG/water, x300, crossed polars + 1A red plate.)
VIII The grainy M droplet texture formed on cooling the N+I droplet texture. (DSCGlwater, x300, crossed polars + 1 A red plate.)
996
XVlII
Chromonics
M ribbons formed on cool-
IX
ing, when the isotropic solution passes directly into the M + I two-phase region. (DSCG/water, x300, crossed polars.)
a
X The N/M boundary
b .
and P-phase behavior: (a) An example of a very distinct N/M boundary, (b) an example of P-type behavior, where the N/M boundary has become indistinct and where features within the optical texture of one phase are continued across the phase boundary. (x300, crossed polars (Ared plate))
7
The Extended Range of Chromonic Phase Structures
XI
997
The alignment of intercalated dye molecules in a host N phase. These three plates show a dilute solution of the dye methylene blue in a host N phase of the drug RU 31 156. Because of the dichroism of the dye, the alignment of guest and host species can be observed independently. (a) Viewed between crossed polars - the pattern of light and dark areas results mainly from the orientation of the host phase. (b) Viewed with the polarizer alone - the colors seen result from the dichroism of the guest dye molecules. They appear blue when the electric vector lies along the major axis of the molecules and pale yellow when it is perpendicular. (c) The sample as for (b) but rotated through 90", showing the alignment of the dichroism of the guest dye molecules. (x300)
XI1 The fingerprint texture of a chirally doped chromonic drug N phase. This sample shows the N + I phase of a 5% solution of arginine in a 22% sample of DSCG at room temperature. The separation of the striations corresponds to 1/2 the pitch of the helicoidal structure. ( ~ 3 0 0crossed , polars.)
998
XVIII
Chromonics
7.2 Chromonic M Gels The conditions for the formation of gels may not be very different to those required for the formation of chromonic M phases, and there may well be a smooth gradation of phase characteristics between these two states. Two quite separate observations support this notion: firstly, an appreciable number of drugs and dyes with molecular structures very closely related to those of chromonogenic species form viscous gels rather than mesophases, and secondly, it is found that the addition of electrolytes gradually increases the viscosity of the M phase and causes it to become more gel-like. The addition of 1% sodium chloride to the DSCG/ water system, for example, whilst not appreciably affecting the schlieren texture of the N phase, inhibits the formation of the M-phase herringbone texture, giving a grainy texture virtually indistinguishable from that found with dye M phases [6] (see also Sec. 8).
7.3 Chiral Chromonic Phases There appear to be no reports of small molecule chiral chromonic mesogens, but it is easy to convert the nonchiral N phase into a cholesteric phase by adding a small watersoluble chiral solute such as an amino acid. Lee and Labes carried out an extensive study of the relationship between the pitch of the induced chiral phase and the concentration of various dopants [%I. They found that the twist varies linearly with concentration over large concentration ranges, and the “helical twisting power” is characteristic of the particular dopant. This approach offers a novel form of assay for small quantities of water-soluble chiral compounds. Low pitch chiral phases have the same characteristic optical properties as thermo-
tropic cholesteric phases, i.e., high optical rotatory powers and selective wavelength reflection. High pitch chiral N phases (where the pitch >1 pm) have the characteristic fingerprint texture, where the spacing of the bands corresponds to half of the pitch. Preliminary reports of a blue phase [56] have been retracted [57],but there appears to be no reason why the chromonic analogs of all three thermotropic blue phases should not exist, especially since lyotropic blue phases have been reported for DNA solutions (Leforestier and Livolant [58])and for a chiral micellar system (Radley [59]). The effect of adding chiral dopants to dye/water systems has been examined for both small chiral molecules (e.g., tartrates) and helical polymers (poly L-glutamic acid) [37].The effect is to create a pronounced circular dichroism in the J absorption band. This is interpreted in terms of the assembly of corkscrew columns of molecules (a recurrent theme in the disjointed history of chromonic phases). Both simple helices and more complicated intertwined assemblies have been proposed. The biologically occurring nucleic acids, DNA and RNA, can be regarded as polymerized chromic columns and, because of the chirality of the ribose sugar units, the entire assemblies are chiral and in a particular sample (under defined conditions) there are helices of only one handedness present. At low concentrations, these systems formchiral N (i.e., cholesteric) phases, but at higher concentrations, an apparently untwisted hexagonal M phase is formed [31].
7.4 More Ordered Chromonic Phases If the analogy between the herringbone structure of the M phase and the molecular
7
The Extended Range of Chromonic Phase Structures
999
arrangement within the layer of a smectic B phase is valid, then more ordered chromonic modifications would be expected analogous to the E and higher smectic phases. There is a growing mass of evidence that this is the case and that other chromonic structures exist at lower temperatures and higher concentrations than the M phase. Goldfarb et al. have reported an additional low temperature smectic-like phase of DSCG (which they provisionally labeled phase 111). This phase is difficult to study. It is metastable and it requires the addition of glycerol as an antifreeze to prevent the formation of ice [60]. A subsequent study by Lee and Labes suggested that there is a further high concentration phase to the lower right of the M phase in the phase diagram [611.
Figure 10. The brickwork structure proposed for the layered mesophase of a cyanine dye.
7.5 Chromonic Layered Structures
7.6 Corkscrew and Hollow Column Structures
Throughout the history of chromonic mesophase studies there have been sporadic reports of smectic phases. Some of these have usually been modified later and the phase concerned later re-identified as M. However, others, including nafoxidine and the salts of fluhenamic acid mentioned earlier (Sec. 2.2), appear to be genuinely lamellar. A recent study of a cyanine dye/water system by Tiddy et al. produced unequivocal evidence of a layered chromonic phase [53]. Samples with concentrations > 1.5% give X-ray diffraction patterns containing a series of inner reflections corresponding to spacings in the ratio 1 : 1/2 : 1/3 : 114 characteristic of a simple one-dimensional lamellar spacing. A layer structure with an overlapping pattern of molecules resembling open lattice brickwork has been proposed (Fig. 10).
A further recurrent theme in the history of chromonic structures is the proposal of complex helical and hollow column structures. An early model for the DSCG columns with a hollow square arrangement was proposed as a way of enabling V-shaped molecules to form stacks with a more or less circular cross section (a feature that at that time was thought to be necessary for the formation of a hexagonal array) [3]. This model was subsequently retracted [5] for a combination of reasons: It was realized that the hexagonal symmetry could result from a disordered herringbone packing of noncircular columns; the X-ray evidence did not support such large columns and the salt bridges proposed did not seem to be a viable way of holding the molecules together in a heavily hydrated environment.
@b,, so;
iChh 60;
(C,H,hN*H
Cyanine Dye.
1000
XVIII
Chromonics
There have been a number of proposals of complex dye columns, including a recent study of the azo dye C. I. acid red 266, where the X-ray diffraction evidence strongly indicates columns with a cross sectional area six times larger than an individual molecule [53,54]. Furthermore, there is a sharp 0.37 nm axial reflection. This is sufficiently far removed from the aromatic ring thickness of 0.34 nm to rule out a simple untilted columnar structure. A hollow chimney made up from a number of intertwined helical stacks appears to be the most likely structural unit (Fig. 11). Note that since the molecules in this case are non-chiral, a racemic mixture of right- and left-handed columns is probable.
OH NaO$' CI Acid Red 266
Figure 11. The hollow brickwork chimney model proposed for the mesophase of the azo dye, acid red 266.
8 The Effect of Additives on Chromonic Sy stems: Miscibility and Intercalation From the point of view of their interactions with chromonic host phases, guest dopant molecules can be grouped into three distinct categories:
1) Electrolytes: The state of aggregation and hence the mesogenic properties of chromonic systems are sensitive to the addition of electrolytes. The effects clearly involve more than simply a modification of the ionic strength of the aqueous part of the phase. Both the column structure and the intercolumn forces seem to be affected. In most (but not all) cases, it appears that chromonic stacks are lengthened and stiffened. Systems with strongly associating chromonic molecules, which already form mesophases, tend to be converted to gels or pseudo-crystalline solids [62] and weakly associating species, which would otherwise not be mesogenic, can be induced to aggregate sufficiently to form liquidcrystalline phases. The addition of NaCl to the DSCG/ water phase diagram has been described by Yu and Saupe [45]. The general effect is to move the phase boundaries upwards and to the left, i.e., a sample that would otherwise have been in the N phase will tend towards the M or M+I. In the dye industry, the final products are usually salted out and therefore tend to contain large quantities of salt. This reduces the solubility of the dye and can turn a respectable mesophase into a curdled gel. As might be expected, if the chromonic species is anionic like DSCG, the cation
8 The Effect of Additives on Chromonic Systems: Miscibility and Interaction
of the added electrolyte appears to be more critical than the anion. (For example, changing the added anion from C1to Br- causes only a small movement of the phase boundaries, but changing the cation from Na+ to K+ makes an appreciable difference.) Note that some aqueous nucleic acid systems appear to require added electrolyte to form mesophases. 2) Nonionic solutes: Nonionic water-soluble species, such as small molecules like urea and glycerol or hydrophilic polymers such as PEG have an effect different to that of electrolytes. The addition of such compounds tends to leave the qualitative features of the phase diagram unaltered, but causes the boundaries to move in the opposite direction to electrolytes. In the dye industry, the removal of salt is generally not a commercially viable option and urea has traditionally been added as a cheap way of countering its effect. Nonionic solutes also tend to suppress the freezing point and have been used to make metastable low temperature chromonic phases accessible. For example, Goldfarb et al. used glycerol as an antifreeze in order to obtain the low temperature chromonic phase, 111,of the DSCG/ water system [601. 3 ) Intercalating species: By analogy with the general mutual solubility of lipids and conventional alkyl chain amphiphiles or with the co-miscibility of smectic phases of the same type, it might be expected that the co-miscibility of chromonic species would be widespread. Complete miscibility has been found in only a few cases, but the intercalation of solThe derivation of the word “intercalation” is interesting. It has the same root as the word calendar. Originally there was no geometrical connotation, it referred to the process of realigning an errant calendar by inserting an additional day (Oxford English Dictionary).
1001
uble aromatic species in chromonic systems over limited solubility ranges appears to be widespread. An intercalating soluble aromatic molecule could be regarded as showing potential chromonic character. This view appears to be justified by the way in which intercalation occurs. In general, the guest species do not aggregate together in groups. They appear to be distributed as individual molecules within the columns of the host system. As is usual in liquid crystal systems, there may be stringent constraints for miscibility, but there do not appear to be stringent geometrical constraints. And for soluble aromatic molecules to be intercalated is not required an exact match with the dimensions of the host column. The intercalation of colored dyes into colorless drug mesophases has proved to be a rewarding study [27, 281. There is generally a metachromic color change (similar to that which occurs when large aggregates of dye molecules are formed). The absorption spectrum of the dye depends on the liquid crystalline state of the host phase, with the dye molecules acting in a loose sense like spin probes and giving information about their surroundings. The mixed system of methylene blue and DSCG is a good example: If solutions containing the same concentration of dye, but increasing concentrations of DSCG are prepared, there is a marked color change from green to blue as the absorption spectrum changes. In addition, the appearance of the sample depends on the optical conditions used and this system provides an interesting example of the use of polarizing optics. Three different situations can be explored. With no polarizer or analyzer, the sample appears homogeneously colored, showing that the dye is distributed evenly throughout the sample. With a polarizer but no analyzer, the blue/yellow
1002
XVIII
Chromonics
Absorbance
-
2.0
1.81.6-
1.41.2-
Erhidium Bromide
1.0-
0.8
-
0.6-
I
400
420
440
460
480
500
520
540
560
580
Figure 12. The metachromic color change occurring for an intercalated dye in a colorless host N phase: Absorption spectra for samples containing the same concentration (0.002%) of the DNA marker, ethidium bromide, in different concentrations of a DSCG/water host phase.
600
Wavelength I nm
dichroism of the dye indicates the local alignment of the guest molecules, whereas with both analyzer and polarizer (in the crossed position) the orientation of the host species can be seen. One system recently investigated is that of the biological stain, ethidium bromide, dissolved in the DSCG mesophases [63]. This reagent is widely used in biochemistry as a marker for nucleic acids. Stained material (where the stain is intercalated between the nucleic acid bases) shows a strong fluorescence, whereas the reagent on its own does not. In the mixed dye/drug system there is a metachromic color shift and the fluorescence appears (Fig. 12). A particularly dramatic example of intercalation occurs when DNA is mixed with the square-planar platinum complex, bipyridylethylenediamine platinum(I1) nitrate. Molecules of the metal complex intercalate regularly at every third position, giving the regular L (ladder) form of DNA [64].
9 The Biological Roles of Chromonic Phases Although the mesogenic behavior of some of the most widely used antiasthmatic and antiallergic drugs has been widely studied, the mode of action of these drugs is still unclear and there is little evidence that the mesogenic properties are the direct cause of the remedial properties. (Although it may well be that the molecular properties that give rise to the formation of liquid-crystalline phases are also those that give these drugs their effectiveness.) I suggest that in the future the major biological role of the chromonic systems will be seen with relevance to the properties of nucleic acids. The liquid crystalline state of the biological membrane (as exemplified in the Singer and Nicolson fluid mosaic model [65]) appears to be vital for its functioning. The membrane is more than just a passive hydrophobic barrier enclosing cell organelles, it must act as a substrate for membrane proteins, be flexible, self-ordering, and self-
9 The Biological Roles of Chromonic Phases
healing. These are all features of a spontaneously self-ordering system, i.e., a liquidcrystalline phase. An analogous list of the physical properties of DNA and RNA could be drawn up. It appears that the biological functioning of the nucleic acids requires them to be liquid crystalline also. They must be flexible enough to wind and unwind, to separate during DNA replication, and to pass though the ribosome reading head when coding for protein synthesis. (A crystalline solid structure would not have the flexibility and a liquid phase would lack the essential tendency towards ordering; only the liquid crystalline state combines both of these essential properties.) In this context, DNA can be regarded as a mesogenic side chain polymer, i.e., essentially a chromonic stack held together by the two sugar-phosphate chains (Fig. 13), rather than as simply a helical polymer rod. There is at least some measure of justification for this view. As the early workers attempting to prepare aligned-fiber samples for X-ray diffraction study found, the molecule is structurally labile; the alignment of the paired bases changes with concentration, temperature, and ionic strength. The
a
C
Figure 13. The chromonic nature of DNA: (a) A stylized sketch of the B form of DNA showing the central stack of base pairs and the two helical sugar-phosphate chains. (b) A chromonic column of unpolymerized molecules, to be compared with the central stack of bases in DNA. (c) The intercalation of a guest molecule in DNA. Note the untwisting of the sugarphosphate chain to accommodate the host molecule. (d) The intercalation of a guest molecule in an unpolymerized chromonic column.
1003
familiar, sturdy molecular models of DNA in biology teaching laboratories (invariably of the B form) give a false impression of the rigidity of the actual structure. A more realistic picture of the physical state of the molecule model would be given by a stack of greasy plates tied together with two loose, helically wound ropes. Just as for membrane lipids, there are two distinct experimental approaches. Firstly, the study of the physical properties of the separated components and secondly, the study of the properties of the biological assembly in the living organism. The crucial studies justifying the view that the nucleic acids are chromonic can be listed as:
1) Investigations of the aggregation of (unpolymerized) purine and pyrimide bases in solution which show that there is a strong tendency for these molecules to stack into columns. (Note that the degree of aggregation of separate (unpolymerized) purine and pyrimidine bases appears to be, in general, just too weak to give rise to mesophases or gels. The exception is for guanosine (and many of its derivatives, but not guanine itself). The difference arises because of the formation of square, planar rings of hydrogenbonded tetramers. The structural unit can therefore be regarded as containing four bases rather than one. When viewed in this light, the hydrogen-bonded dimer base pair units produced by WatsonCrick hydrogen bonding at the center of the DNA double helix are just on the edge of mesogenic behavior.) The discovery of the liquid-crystalline phases of RNA and DNA solutions. Whilst preparing materials for X-ray diffraction studies (which eventually led to the Watson-Crick double helix model), Spencer et al. noted that liquid crystal
1004
XVIII
Chromonics
3) The ease of intercalation of other chromonic and potentially chromonic materials. A number of biochemical reagents, such as acridines and ethidium bromide, intercalate readily between the stacked bases in DNA and RNA. Anticancer drugs, such as the square planar platinum complexes, intercalate avidly (Fig. 14) and naturally occurring antibiotics, such as actinomycin, similarly act by intercalation into the stack of bases [67]. They act as tailor-made “spanners in the works” and prevent the reading and replication of DNA. 4) The liquid-crystalline nature of nucleic acids in vivo. In lower organisms, the nucleic acids are naked and exist for much of the time as concentrated droplets of cholesteric phase. In electron micrographs of the nuclei of some dinoflagellates, for example, the characteristic bands of nested arcs, “Bouligand Pat-
phases of t-RNA show banded optical textures [30], implying a low pitch cholesteric phase. Subsequent investigations by Robinson showed that the optical properties were comparable to those of cholesteric solutions of synthetic a-helical polypeptides [31]. Later studies notably by Livolant showed that there are at least two types of nucleic acid mesophase, i.e., a relatively low concentration cholesteric phase and a high concentration hexagonal phase [32]. In these systems, the hexagonal lattice spacing is large enough for the structure to be seen directly by electron microscopy. It is significant that the so-called peptide nucleic acids (where the central stack of bases is the same, but where the usual sugar-phosphate chain has been replaced by a modified protein chain) form a similar range of columnar structures to the conventional nucleic acids [66].
\
H,N’
\
Pt
/
2+
z H )
H2C-CHz [(bipy)Pt(en)y+
Figure 14. The structure of L-DNA: The planar platinum complex [(bipy)Pt(en)] [21] readily intercalates in DNA. The saturated structure has the L (ladder) form in which the DNA helix is completely unwound and the intercalation appears to occur regularly in every third position (redrawn from [64]).
11 Conclusions
terns” of a helicoidal structure are visible [68]. For higher organisms, the situation is more complex. Coiled coils of DNA are wound round histone bobbins and there is no bulk mesophase formed.
10 Technological and Commercial Potential of Chromonic Systems Many compounds of commercial interest, e.g., dyes, drugs, antibiotics, and anticancer agents, are either chromonic or have sufficiently chromonic character to be able to intercalate into chromonic systems. However, there is a distinction between commercially useful compounds, which incidentally happen to form liquid-crystalline phases, and compounds that have technological value specifically because of their mesogenic properties. We have hardly scratched the surface of the technological potential of chromonic systems. An occasional dye or food coloring is marketed in the form of a convenient V phase, and the drug companies know that :oncentrated solutions of many drugs form iighly viscous M phases, but these cases are rivial compared with the potential applicaions. To recognize that the dye/water sysem can be liquid-crystalline opens the door o a wide range of possible forms of exploiation. All of the necessary ingredients ippear to be present for the production of :hromonic information display and storage jevices, i.e., surface alignment, guesthost ilignment, orientation by electric or magietic fields, birefringence, and dichroism. Jse could be made of the conflict between he surface alignment of aggregates and the ilignment caused by externally applied ilectric or magnetic fields. There is, for ex-
1005
ample, high security printing, where it may be possible to control not only the position of the printed image but also the orientation and state of aggregation of the dyes. Although it does not seem likely that lyotropic display devices will replace those of mainstream thermotropic technology, it would be surprising if there were not situations where chromonics had something unique to offer. As printing techniques become more sophisticated, novel processes perhaps incorporating features of both jet printers and photocopiers could utilize the mesogenic nature of dye solutions. Finally, there are indications that viable light-harvesting devices can be constructed from mixed chromonic systems involving dyes and transition metal complexes [69].
11 Conclusions When the century of diverse literature dealing with the solution chemistry of drugs, dyes, and nucleic acids is surveyed, the way in which the same concepts recur independently, time after time is amazing -the same problems reconciling molecular structure to mesogenic properties, the same pictures of molecular aggregation, i.e., stacks of cards, piles or coins, helical aggregates, the same puzzling optical textures, “limpid fluids possessed of a sheen”, “worm-like growths”, and the same observations of flow birefringence, i.e., the same observations about the effects of added electrolyte. The gathering together of drug, dye, and nucleic acid systems under the heading of chromonics has been more than just a taxonomic exercise. It has enabled observations from one area of study to throw light on another. The term “isodesmic”, for example, is not generally used in the dye industry, but it neatly describes a pattern of behavior that
1006
XVIII
Chromonics
has no other name and will, I am sure, become more widely used in future. (There was, of course, no need for such a term in the days when conventional aliphatic amphiphiles were the only recognised type of lyotropic mesogen and where only one pattern of aggregation was known.) The identification of chromonic systems has led to a reappraisal of the meaning of the term hydrophobic, and it appears that there are two distinct forms of hydrophobic behavior, i.e., that shown by conventional amphiphiles with largely saturated alkyl chains and that shown by aromatic species where n-n interactions occur. Concerning the term chromonic, this word was coined to replace phrases such as “lyotropic mesophases of the types formed by soluble aromatic mesogens”. It was intended to echo the bis-chromone structure of DSCG and to carry connotations of dyes and chromosomes (and hence nucleic acids). It was also intended to mark Hartshorne’s studies of DSCG (which although by no means the first such mesophase system to be explored, was one of the most extensively examined), and it is to Norman Hartshorne, who taught me and many others a love of optical microscopy, that I respectfully dedicate this chapter. Acknowledgements I am grateful to Zeneca (formerly ICI Dyestuffs, Blackley, U. K.) and to Fisons Ltd., Holmes Chapel, U. K.) for their continued interest and support, and to my friend Gordon Tiddy for his encouragement. I gratefully acknowledge advice from Dr. Richard Bushby and Dr. John Griffiths (Department of Chemistry and Colour Chemistry respectively, University of Leeds) with the preparation of this chapter. I have been privileged to have had a succession of very able and enthusiastic co-workers. I am indebted to Teresa Attwood (who was there at the beginning) to Jane Turner and Paul Alder, to Julie Cox, and finally to Kevin Mundy.
12 References [ 11 J. S . G. Cox, G. D. Woodard, W. C. McCrone. 1.
Pharm. Sci. 1971,60, 1458- 1465. [2] N. H. Hartshorne, G. D. Woodard, Mol. Cryst. Liq. Cryst. 1973,23, 343 - 368. [3] J. E. Lydon, Mol. Cryst. Liy. Cryst. Lett. 1980, 64, 19-24. [4] F. Jones, D. R. Kent, Dyes Pigm. 1980. 1, 39-48. [5] T. K. Attwood, J. E. Lydon, Mol. Cryst. Liq. Cryst. 1984, 108, 349-357. [6] T. K. Attwood, J. E. Lydon, F. Jones, Liq. Crysr. 1986,1,499-507. [7] J. E. Turner, J. E. Lydon, Mol. Cryst. Liq. Cryst. Lett. 1988, 5,93-99. [8] N. Boden, R. J. Bushby, C. Hardy, J. Phys. Lett. 1985, 46, L325-328; N. Boden, R. J. Bushby, C. Hardy, F. Sixl, Chem. Phys. Lett. 1986, 123 (5), 359-364; N. Boden, R. J. Bushby, L. Ferris, C. Hardy, F. Sixl, Liq. Cryst. 1986, I (2), 109- 125. [9] G. W. Gray, P. A. Winsor, Liquid Crystals and Plastic Crystals, Ellis Horwood, Chichester 1974. [lo] See, for example, H. J. Conn, Biologicul Stains, Biotech. Publications, Geneva, N. Y., U. S. A. 1953, or E. Gurr, Encyclopaedia of Microscopic Stains, Leonard H ill (Books), London 1960. [ l 11 For a short review of the work of Ehrlich with detailed references see: A. S. Travis, The Biochemist 1991, 13 (5), 9. [12] A. S. Travis, The Rainbow Makers - The Origin of the Synthetic Dyestuffs Industry in Western Europe, Lehigh University Press, Bethlehem 1993; A. S. Travis, Text. Chem. Color. 1988, 20 (8), 13-18; A. S. Travis, Text. Chern. Color. 1990,22 (12), 18-20; 0. Meth-Cohn, A. S. Travis, Chem. Britain 1995,31, 541-549. [13] H. Sandquist, Berichte 1915,48,2054-2055. [ 141 F. Fourneau, J. TrCfouel, (Mme) J. TrCfouel, I. VallCc, Ann. Znst. Pasteur 1924, 38, 8 1. [15] I. F. Balaban, H. King, J. Chem. Soc. 1927,127, 3068. [16] P. Gaubert, C. R. 1933, 197, 1436-1438: 1934, 198,951-953; 1935,200,679-680. [17] E.E. Jelley, Nature 1936, 138, 1009-1010; E. E. Jelley, Nature 1937, 139, 63 1-632. [ 181 G. Scheibe, L. Kandler, H. Ecker, Nururwissenschaften 1937,25,75;G. Scheibe, Koll. Z. 1938, 8 2 , l ;G. Scheibe,Angew. Chem. 1938,52,631637. [I91 R. E. C. Altounyan, Clin. Allergy 1980,10, 481; J. Pepys, A. W. Frankland, Disodiuin Crornglycute in Allergic Airways Disease, Butterworths, London 1970. [20] D. Perahia, Z. Luz, E. J. Wachtel, H. Zimmermann, Liq. Cryst. 1987, 2,473 -489. [21] A. R. Mlodozeniec, J . Soc. Cosm. Chem. 1978, 29,659-683.
12 References [22] J. B. Bogardus,J. Pharm. Sci. 1982,71,105-109. [23] T. Eckert, W. Fischer, Colloid Polym. Sci. 1991, 259, 553 and 260, 880-887. [24] T. Rades, C. C. Muller-Goymann, Pharm. Pharmucol. Lett. 1992, 2, 131- 134. [25] I. Kustanovitch, R. Poupko, H. Zimmermann, Z. Luz, M. M. Labes, J. Am. Chem. Soc. 1984, 107,3494-3501. [26] D. Goldfarb, M. E. Mosley, M. M. Labes, Z. Luz, Mol. Cryst. Liq. Cryst. 1982, 89, 1 19- 135. [27] J. A. Cox, Ph. D. Thesis, University of Leeds 1992. [28] M. Kobayashi, A. Sasagawa, Mol. Cryst. Liq. Cryst. 1993, 225, 293-301; M. Kobayashi, H. Matsumoto, T. Hoshi, I. Ono, J. Okubo, Mol. Cryst. Liq. Cryst. 1990, 180B, 253-263. [29] K. Mundy, J. C. Sleep, J. E. Lydon, Liq. Cryst. 1995,19, 107- 112. [30] M. Spencer, W. Fuller, M. H. F. Wilkins, G. L. Brown, Nature 1962,194, 1014-1020. [31] C. Robinson, Tetrahedron 1961, 13, 219-234. [32] F. Livolant, J. Mol. Biol. 1991, 218, 165-181; F. Livolant, A. M. Levelut, J. Doucet, J. 0. P. Benoit, Nature 1989,339, 724-726. [33] S. Bonazzi, M. De Morais, A. Garbesi, G. Gottarelli, P. Mariani, G. P. Spada, Liq. Crysf. 1991, 10,495-506; A. Garbesi, G. Gottarelli, P. Mariani, G. P. Spada, Pure Appl. Chem. 1993, 65, 641 -646; G. Gottarelli, G. P. Spada, P. Mariani, M. M. Demorais, Chirality 1991,3,227-232; P. Mariani, M. M. Demorais, G. Gottarelli, G. P. Spada, H. Delacroix, L. Tondelli, Liq. Cryst. 1993, 15, 757-778. See also: Y. M. Yevdokimov, S. G. Skuridin, V. I. Salanov, Liq. Cryst. 1988, 11, 1443- 1459, and references therein. [34] A. S. Vasilevskaya, E. V. Generalova, A. S. Sonin, Russian Chemical Reviews 1989, 58 (Y), 904-916; translated from Usp. Khim. 1989,57, 1575-1596. [35] See, for example, C. Tanford, The Hydrophobic Effect, Wiley, New York 1973. [36] T. K. Attwood, J. E. Lydon, C. Hall, G. J. Tiddy, Liq. Cryst. 1990, 7, 657-668. [37] C. A. Hunter, J. K. M. Sanders, J. Am. Chem. Soc. 1990,112,5525-5534. [38] D. F. Evans, H. Wennerstrom, The Colloidal Domain, VCH, New York 1994, Chap. 4. [39] H. Stegemeyer, F. Stokel, Ber. Bunsenges. Phys. Chem. 1966,100,9- 14. (401 E. G. McRae, M. Kasha, J. Chem. Phys. 1958, 28,721-722. [41] See, for example, N. Tyutyulkov, J. Fabian, A. Melhorn, F. Dietz, A. Tadjer, Polymethine Dyes: Structure and Properties, St. Kliment Ohridski University Press, Sofia 1996. [42] W. Saenger, Principles of Nucleic Acid Structure, Springer, Berlin 1984, Sec. 6.6 and references therein, including: P. 0. P. Ts’o in Basic Principles in Nucleic Acid Chemistry (Ed.:
1007
P. 0. P. Ts’o), 1974, I, 453; P. 0. P. Ts’o, I. S. Melvin, A. C . Olson, J. Am. Chem. Soc. 1963, 85, 1289- 1296. [43] J. V. Champion, G. H. Meeten, J. Pharm. Sci. 1973,62 ( l o ) , 1589-1595. [44] Y. W. Hui, M. M. Labes, J. Phys. Chem. 1986, 90,4064 - 4067. [45] L. J. Yu, A. Saupe, Mol. Crysf.Liq. Cryst. 1982. 91,53-55. [46] N. H. Hartshorne, The Microscopy of Liquid Crystals, Microscope Publications, London 1974. [47] J. E. Turner, Ph. D. Thesis, University of Leeds 1988. [48] J. Rogers, P. A. Winsor, J. Cull. Inter- Sci. 1969, 30, 500-510. [49] J. D. Bernal, L. Fankuchen, J. Gen Physiol. 1941,25, I 1 1- 146. [50] D. Goldfarb, Z. Luz, N. Spielberg, H. Zimmermann, Mol. Cryst. Liq. Cryst. 1984,126,225-246. [51] T. K. Attwood, Ph. D. Thesis, Universit of Leeds 1984. [52] S. Hamodrakas, A. J. Geddes, B. Sheldrick, J. Pharm. Pharmacol. 1974,26,54-56. [53] G. J. T. Tiddy, D. L. Mateer, A. P. Ormerod, W. J. Harrison, D. J. Edwards, Langmuir 1995 11 (2), 390-393. [54] W. J. Harrison, D. L. Mateer, G. J. T. Tiddy, J. Phys. Chem. 1996,100,2310-2321. [55] H. Lee, M. M. Labes, Mol. Cryst. Liq. Cryst. 1982,84, 137-157. [56] H. Lee, M. M. Labes, Mol. Cryst. Liq. Cryst. Lett. 1983, 82, 355-359. [57] M. Kuszma, H. Lee, M. M. Labes, Mol. Cryst. Liq. Cryst. Lett. 1983, 92, 81 -85. [58] A. Leforestier, F. Livolant, Liq. Cryst. 1994,17, 651-658. [59] K. Radley, Liq. Cryst. 1995, 18, 151-155. [60] D. Goldfarb, M. M. Labes, Z. Luz, R. Poupko, Mol. Cryst. Liq. Cryst. 1982, 87, 259-279. [61] H. Lee, M. M. Labes, Mol. Cryst. Liq. Cryst. 1983,91,53-58. [62] D. E. Sadler, M. D. Sannon, P. Tollin, D. W. Young, M. Edmonson, P. Rainsford, Liq. Cryst. 1986, I , 509- 520. [63] K. Mundy, J. C. Sleep, J. E. Lydon, Liq. Cryst. 1995,19, 107-112. [64] S. Arnott, P. J. Bond, R. Chandrasekaran, Nature 1980,287,561-563. [65] S. J. Singer, G. L. Nicolson, Science 1972, 175, 720-731. [66] P. Wittung, P. E. Nielsen, 0. Buchardt, M. Egholm, B. Norden, Nature 1994,368,561-563. [67] See, for example: W. Saenger, Principles of . Nucleic Acid Structure, Springer, Berlin 1984, Chaps. 8 and 16 and the references therein. [68] Y. Bouligand, M. 0. Soyer, S. Puiseux-Dao, Chromosoma 1968,24,251-287. [69] D. Gustand, T. A. Moore, Science 1989, 244, 35-41.
Handbook of Liquid Crystals Edited by D. Demus, J. Goodby, G. W. Gray, H.-W. Spiess, V. Vill Copyright 0 WILEY-VCH Verlag GmbH. 1998
Index
positions, dichroic dyes 266 Abbe refractometer 129, 348 absorbance - antiferroelectrics 667 - guest-host effect 269 absorption band, chromonics 998 absorption properties, columnar discotics 795 absorption spectra - charge transfer systems 946 - guest-host effect 259 f, 270 acetylene derivatives, synthesis 697 acetylenes, hydrocarbon cores 719 acetylide ligands, metallomesogens 905 f achiral host materials, ferroelectric mixtures 435 achiral materials, nematics 342 acid derivatives, chiral nematics 3 19 acid esters, swallow-tailed mesogens 851 aciapyridine complexation, hydrogen bonding 970 acids - dielectric behavior 105 - heterocyclic cores 722 f acoustic waves propagation, columnar phases 770 activated states, TN displays 202 active matrix addressing - AFLCD 679 - STN 219 - LCD 199 active matrix devices, STN 224 active matrix displays 230-242 acylation - hexacycles 742 hydrocarbon cores 714 - nematics 50 acylphenoxyethyl esters, substituted mesogens 848 additional symmetries, ferroelectrics 554 additives - chromonics 1000 f laser addressed devices 477 addressing - AFLCD 679 DSM 254 dual frequency 294 FLC displays 610 - nonchiral smectics 480 STN 219 - TNdispIays 204 CI
~
~
~
~
~
~
admixtures, dichroic dyes 296 aggregation - charge transfer systems 945 chromonics 981 ff, 987 ff, 1000 alcohols, chiral smectics 496 alicyclic rings, smectogens 4 19 alignment - chiral nematics 345 - dye molecules 257 ff - homeotropic 471 - sample preparation 764 f - shear viscosity 143 alignment damage, AFLCD 679 alignment effect, Heilmeier displays 278 aliphatic chains - chiral smectics 496 - nematics 122 - phasmids 865 - smectogens 412 - terminal 414 aliphatic deuterons, splitting 753 aliquat, reagents 72 1 alkali carboxylate Iigands 906 alkanes, biaxial nematics 936 alkene complexes, ligands 904 alkoxy chains, hydrocarbon cores 710 alkoxystilbazole ligands, metallomesogens 902 alkyl biphenyl carboxylates, synthesis 427 alkyl branched alkyl chains, chiral smectics 497 alkyl bridges, biaxial nematics 937 alkyl chains - chiral 673 - cubic phases 888 - dimers 801 hydrogen bonding 970 f - lateral substituents 835 ff - relaxation 101 alkylation, nematics 50 alkylbenzylidenemalonate 85 1 alkylcyclohexanoates 837 alkylenecarbonyloxy connectors 687 alkylphenyl derivatives 953 all-trans conformation 814, 825, 836 f all-trans molecular length, dimers 808 f Alt-Pleshko addressing 2 19 Alt-Pleshko law 206 f -
~
1010
Index Vol. 2A and 2B
alternating current fields, dynamic scattering 244 alternating smectic phases 14 amides 694 - cubic phases 890 hydrogen bonding 969 smectogens 422 amine ligands, metallomesogens 926 amino acids, chiral smectics 502 aminoester 866 ammonium surfactants, chiral nematics 338 amphiphiles chromonics 981-1007 - substituted mesogens 859 amphiphilic iron complexes 927 analog gray levels, ferroelectric devices 637 anchoring - chevron interfaces 636 - discotics 764 - ferroelectric devices 631 - nematics 74, 188 - shearviscosity 144 angle dependence - STN 208 TNdisplays 202 aniline derivatives, charge transfer systems 948 aniline purple, chromonics 984 anisotropic elastic ratio, nematics 7 1 anisotropic guest-host interaction 257 anisotropy - diamagnetic 124 - ferroelectrics 608 nonchiral smectics 481 optical 221, 304 annelide, amine ligands 926 anthracene, synthesis 700 anthracenyl, substituted mesogens 837 anthraquinone, dichroic dyes 260 ff antibiotics, chromonics 981 anticancer agents, chromonics 98 1 antiferroelectric host materials 437 antiferroelectric mixtures, nonchiral smectics 41 1 antiferroelectric smectics 10 antiferroelectrics 510 f, 525, 577 ff, 665-691 - columnar discotics 794 antiferromagnetism, columnar phases 766 apophyllite, sample preparation 764 applications - chiral nematics 305, 335409 - columnar discotics 781-798 - diamagnetic properties 125 - ferroelectrics 629 laser addressed devices 478 nematics 199-229 - nonchiral smectics 470-490 - smectic materials 435 aromatic esters, chiral smectics 497 aromatic heterocycles, chiral smectics 500 aromatic hydrogen cores, discotics 694 ff aromatic rings, smectogens 412 -
-
-
-
-
-
-
-
aroylhydrazine, cyclometalated complexes 92 1 Arrhenius plots - copper phthalocyanines 766 - ferroelectrics 619 - nematics 99 aryl iodides, hydrocarbon cores 7 19 aryl segments, substituted mesogens 849 arylboronic acids, nematics 55 arylimines, cyclometalated complexes 9 19 aspect ratio, shear viscosity 143 associations, thermotropic cubic phases 888 ff asymmetry - chiral nematics 307 - ferroelectrics 559 atmospheric pressure chemical vapor deposition 238 awochromic groups, dichroic dyes 266 Avogadro number 114 azacrowns, tubular 720 azamacrocycles, synthesis 742 azobenzene - cyclometalated complexes 9 18 - isomerization 40 azobenzened-mesogens, chiral nematics 3 12 azodyes 260ff - chromonics 985 azomercury, cyclometalated complexes 9 19 azomethine 260 - substituted mesogens 844 azomethine ester derivatives 325 azomethine mesogens 3 13 azoxy, dynamic scattering 243 azoxybenzenes, shear viscosity 152 azpac compounds, cyclometalated complexes 9 19 azulene, dichroic dyes 260 azulene dicarboxylic esters 841
p positions, dichroic dyes
266 10 back to back diode, active matrix displays 230 backflow, shear viscosity 150 bacteria, chromonics 984 bandgap, columnar discotics 78 1 f bandwidth, columnar discotics 781 bathochromic groups, dichroic dyes 266 bathochromic shift, chromonics 988 BCBOn series, dimers 802 ff bend deformations, nematics 184 bend director distortions, nematics 62 bend distortions, nonchiral smectics 464 bend elastic constants, nonchiral smectics 47 1 bend elastic deformations, chiral nematics 370 f bend modes, nematics 172 bending, columnar phases 77 1 benzate esters, synthesis 705 benzene derivatives, discotics 749 ff benzenes 693 ff - dielectric behavior 107 - nematics 48
B phases, calamitics
Index Vol. 2A and 2B rotational viscosity 161 benzoic acid derivatives, hydrogen bonding 969 benzoic acids, swallow-tailed substituted 859 benzoinylanilines, substituted mesogens 843 benzonitrile derivatives 948 benzonitriles 221 benzoquinone, dichroic dyes 260 benzotrisfuranes, synthesis 723 benzpyrones, condensed 723 biased rotational motion, antiferroelectrics 665 ff biaxial nematics 83, 933-943 biaxiality, ferroelectrics 608 f, 654 biaxiylity, optical 493 bicyclo(2.2.2) octanes, disubstituted 50 bidentate ligands 906 biforked polycatenars 865 ff bifurcation point, dimers 820 bilayer structures, smectics 8 biological roles, chromonics 1002 biomedical thermography, chiral nematics 398 biphenyl derivatives, synthesis 706 biphenyl mesogens, chiral nematics 3 14 biphenyls 48, 94 bipolar configuration, nematics 185 bipyridenes, hydrogen bonding 970 birefringence - chiral nematics 337, 357 chromonics 987 columnar discotics 784 - displays 288 f dynamic scattering 243 ferroelectrics 588, 598, 654 - 1-N transitions 24 - nematics 7, 128 nonchiral smectics 471 bispyran, synthesis 721 f bistability, ferroelectrics 525 f, 547 f, 636 black brushes, chiral nematics 350 black mixtures - dichroic dyes 260 guest-host effect 274 black operations, STN 215 blue mode, STN 2 13 blue phases - chiral nematics 365 -- chromonics 998 blue shift, chromonics 988 blue tail phenomennon, chiral nematics 306 boat domains, chevron interfaces 637 boat-shaped domains, ferroelectric devices 644 boat-wake defects, ferroelectric devices 647 Boltzmann factor - dimers 817 ferroelectrics 520 bonds, STN 224 bone-shaped biaxial nematics 934 bookshelf structures - AFLCD 679 ~- ferroelectrics 610, 626, 629, 653 ~
~
~
~
101 I
Born approximations, nematics 171 Bose-Einstein condensation, ferroelectrics 5 17 Bouligand pattern 1004 Bouligand plane 560 boundary conditions, director fields 73 Bragg angle, cubic phases 895 Bragg peaks, nonchiral smectics 444 Bragg reflection - chiral nematics 339, 358 - columnar phases 772 - optical propagation 362 f branching - saturated cores 741 - smectogens 425 - terminal aliphatic chains 417 break, columnar phases 786 breakdown, active matrix displays 230 brightness, displays 287 brightness contrast ratio, ferroelectric devices 638 brightness enhancement, STN 2 18 Brillouin scattering 770 broad band dyes, guest-host effect 274 bromination - heterocyclic cores 733 - hydrocarbon cores 709 bromo derivatives, polycatenars 877 bromobiphenylcarboxylic acids, synthesis 433 brushes, chiral nematics 350 bulk properties, smectic AIC 452,457 Burgers vector 773 f butanol, chiral smectics 496
~
~
~
~
ClIC2 chevrons, ferroelectric devices 644 cadmium selenium TFT address 233 Can0 wedge, chiral nematics 347 f, 35 I Canon technology, ferroelectric devices 648 capillary flow, shear viscosity I43 carbocyclic compounds, charge transfer systems 958 carbocyclic rings, smectogens 412 carbon atoms, intercalated smectics 808 carbon-carbon bonds, dimers 823 carbon-carbon double bonds, chiral smectics 498 carbonaceous phases 693 carbonyl connectors, antiferroelectrics 687 carbonyl groups - antiferroelectrics 666 - chiral smectics 505 carboxylate ligands, metallomesogens 906 carboxylatepyridine complexes 905 carboxylic groups, biaxial nematics 935 carcinogenicity, heterocyclic cores 726 carcinogens, chromonics 98 1 Carr-Helfrich theory - chiral nematics 389 - dynamic scattering 247 - smectic A 483 cell gap 251,281 cell preparation, guest-host effect 266 f
1012
Index Vol. 2A and 2B
cell spacing 267 cell thickness, displays 287 cellobiose derivatives, synthesis 741 cellooligosaccharide derivatives 866 cells - dynamic scattering 243 STN 216 chain branching, heterocyclic cores 733 f chain number, polycatenars 880 chain position, polycatenars 880 chain-spacer interactions, dimers 806 chains - chiral smectics 497 smectogens 411 characterization methods, biaxial nematics 937 charge carriers, columnar discotics 784 charge mobility anisotropy, columnar discotics 792 charge transfer systems 945-967 charge transport, columnar discotics 78 1 Chatelain method 129 chemical structures - cubic phases 888 - dichroic dyes 260 chemical vapor deposition, silicon nitride layer 23 1 chemically doped materials 787 chevrons dynamic scattering 246 - ferroelectrics 576, 610, 626, 633 ff chiral alkyl chains, antiferroelectrics 673 chiral center incorporation, nematics 303 chiral chains, antiferroelectrics 686 chiral chromonics 998 chiral compounds, swallow-tailed mesogens 853 chiral coupling, ferroelectrics 582 chiral derivaties, synthesis 701 chiral dimers 813 chiral dopants, chromonics 998 chiral electron acceptors 960 chiral nematic phases, columnar 758 f chiral nematics, synthesis 303-324 chiral salicylaldimine ligands 916 chiral smectics, synthesis 493-5 14 chirality, ferroelectrics 516 ff, 543 ff, 548 ff, 653 Chisso- 1013 mixture, ferroelectric devices 656 chitobiose derivatives, synthesis 741 chloranil, reagents 702,712 chlorination, hydrocarbon cores 707 f cholesteric configurations, guest-host effect 273 cholesteric doping, Heilmeier displays 278 cholesteric helix, displays 288 cholesteric nematics, chiral 303 cholesteric phases, dynamic scattering 250 cholesterics - Saupe theory 61 - symmetry 554 see also: chiral nematics 336 ff cholesterol acetate 335 cholesterol benzoate 335 cholesterol derivatives 303 -
-
-
cholesteryl/-esters 3 10 chrominance, guest-host effect 272 chromium - diketonate ligands 9 12 - lyotropic metallomesogens 927 chromium compounds, cell preparation 267 chromonics 981-1007 cinnamic acid esters, swallow-tailed mesogens circular components, chiral nematics 338 circular dichroism, chiral nematics 304 cis-trans back reactions, chiral nematics 368 clamped crystal condition, ferroelectrics 540 classification, chiral nematics 307 clearing process, nematics 108 clearing temperatures - alkyl substituted mesogens 835 ff - biaxial nematics 937 - charge transfer systems 946,957 - chiral smectics 496 - columnar discotics 784 - dimers 804 - heterocyclic cores 727, 738 - metallation 734 - phasmids 869 - shear viscosity 150 - STN 226 cleaving, sample preparation 764 clusters, isotropic phase 92 coatings - chiral nematics 345 - tin-indium oxide 484 cobalt, lyotropic metallomesogens 927 coherence length, MBBA 25 Cole-Cole equation 183 Cole-Cole plots, nematics 100 Cole-Cole relaxation, ferroelectrics 6 18 collective modes, ferroelectrics 618 colodial solutions, nematics 170 color - charge transfer systems 946 - chiral nematics 306 - chromonics 988 - dichroic dyes 259 - ferroelectric devices 630,645, 653 - Heilmeier displays 281 - laser addressed devices 478 - STN displays 215,218 color filters, cell preparation 267 color neutrality, ferroelectrics 598 colored dyes, chromonics 1001 colored twisted nematic cells 395 column-column repulsion, chromonics 986 column length, chromonics 982 columnar nematics 757 ff columnar phases - discotics 693, 750 ff - polycatenars 880 ff - swallow-tailed mesogens 857 columns aggregations, chromonics 98 1
85 1
Index Vol. 2A and 2B complexation, hydrogen bonding 969 ff composition dependence, chiral nematic pitches 365 compounds - antiferroelectric 684 f biaxial nematics 934 chiral nematics 306 f, 51 1 - dielectric properties 120 polycatenars 876 conductivities - columnar discotics 784 - columnar phases 766 f - dynamic scattering 243, 249, 253 - nonchiral smectics 484 conductors, doped columnar 795 cone mode viscosity, ferroelectncs 602 f, 613 configurations - active matrix displays 230 - nematics 185 - STN 209 - TN displays 199 conformers - dimers 816 - substituted mesogens 836 f - swallow-tailed mesogens 855 f connectors, antiferroelectrics 686 conoscopy, biaxial nematics 938 constituent units, charge transfer systems 945 constrained nematics, randomly 189 continuum theory - chiral nematics 337, 369 f - columnar phases 769 ff contrast enhancement, ferroelectrics 6 10 contrast ratio - AFLCD 680 displays 287 - dynamic scattering 251 - guest-host effect 272 - Heilmeier displays 277 - laser addressed devices 478 - positive mode dichroic LCDs 294 - TN displays 204 converse piezoeffect, ferroelectrics 524 copper, ligands 907 f, 9 15,923 copper carboxylates, charge transfer systems 962 copper complexes - biaxial nematics 935 - lamellar phases 759 copper tetracarhoxyphthalocyanine 98 1 f core length, hydrogen bonding 970 core systems - nematics 53 - smectogens 418 core units - chiral nematics 307 - smectogens 411 cores - dicotics 693 f polycatenars 866, 879 ff - semirigid anisometric 801 -
-
-
-
-
1013
corkscrew structures, chromonics 999 correlation length, cyanobiphenyls 27 coruleoellagic acid, heterocyclic cores 722 Coulomb interactions, antiferroelectrics 665 f, 672 coupling, chiral nematics 377 critical behavior - N-SmA transitions 26 f - nematics 71 - SmA-SmC transitions 30 cross relaxation, nematic-polymer interface 187 crystal BIC materials, synthesis 428 crystal phases (B,G,H,...) nonchiral smectics 441 ff crystal structures chromonics 992 polycatenars 883 cubic phases - polycatenars 880 ff - thermotropic 887 14 Curie law, ferroelectrics 521, 541 ff, 550 f Curie temperature, ferroelectrics 517 f, 532 ff Curie-Weiss law 25 - ferroelectrics 528, 532 f cut-off frequencies, nematics 180 cymo derivatives, polycatenars 878 cyano end groups, smectogens 4 18 cyano groups, chiral smectics 506 cyanobiphenyls - biaxial nematics 940 f - charge transfer systems 946 f - dimers 803 f, 825 - elastic properties 60, 7 1 - ferroelectrics 522 f - metallomesogens 902 - nematics 48 - shear viscosity 152 - synthesis 426f - viscoelastic properties 177 cyanocyclohexanes, nematics 55 cyanophenyl derivatives 84 1 cybotactic groups, substitued mesogens 835 cyclohexanate esters 706 cyclohexane mesogens, chiral nematics 33 1 cyclohexanes 49,739 cyclohexanones 95 cyclohexyl a-fluorohexanoates, chiral smectics 503 cyclometalated complexes, metallomesogens 9 18 ff -
-
dark brushes, chiral nematics 350 de Gennes theory - ferroelectrics 556 - nematics 171 Debye description, ferroelectrics 531 ff, 621 decacyclenes, synthesis 714 defects - chiral nematics 350 f - columnarphases 773 ff - ferroelectric devices 634, 647 SmA-SmC transitions 35 -
1014
Index Vol. 2A and 2B
deformations columnar phases 771 - ferroelectrics 555 f 567 f - flexoelectric smectic phases - nematics 67, 184 deformed helix mode (DHM), ferroelectrics 587 f dendrimers, smectogens 412 deposition, silicon nitride layer 23 1 deprotection, hydrocarbon cores 7 19 deuterated derivatives, hydrocarbon cores 719 deuterons, aliphatic 753 device structures, ferroelectric 629 ff dialkylazoxybenzene/PCB, charge transfer systems 948 dialkylbenzenes, synthesis 733 dialkylketonoxime benzoate 853 diamagnetic properties, nematics 113 15 diamines, synthesis 696 diarylazines, cyclometalated complexes 920 diarylmercury complexes, metallomesogens 901 dibenzopyrene, synthesis 7 12 dichlorobenzene, synthesis 733 dichroic color sensitive polarizer 216 dichroic dyes, guest-host effect 259 ff dichroic LC displays, guest-host effect 258 f dichroic mixture formulations 274 f dichroic parameters, guest-host effect 268 ff dichroic phase change effect display 204 dichroic polymer LCDs 298 dichroic ratio, dyes 259 ff, 268 ff dicopper complexes, diketonate ligands 9 1 1 dicyanoethenyl substituents 949 dielectric anisotropy - negative 243 - nonchiral smectics 481 - positive 221 dielectric biaxiality, ferroelectrics 610 dielectric neutral basic materials, STN 226 dielectric properties charge transfer systems 960 788 ff, 794 - columnar discotics - columnarphases 775 - ferroelectrics 520 f, 599,622 - nematics 91-112 dielectric regime, dynamic scattering 249 dielectric relaxation ferroelectrics 622 nematics 182 dielectric reorientation, smectic A 48 1 dielectric spectroscopy, ferroelectrics 6 17 ff dielectric states, ferroelectrics 524 Diels-Alder reaction - heterocyclic cores 736 f - hydrocarbon cores 713 f diethylammonium flufenamate, chromonics 985 diffraction approach, optical propagation 362 diffraction patterns, chromonics 991 difluorophenyl pyrimidines, synthesis 436 digital full color displays 657 ~
~
~
~
diisobutylsilane diol (DIIBSD) 756 f diketonates 434, 909 ff diketone complexes, smectogens 413 dilute solutions, chromonics 981,987 ff dimeric mesogens, chiral nematics 327 dimers 802 - biaxial nematics 935 - chromonics 988 - cubic phases 888 - discotic 801-833 - heterocyclic cores 725 - hydrocarbon cores 716 f dimethyl aminopyridine, hydrocarbon cores 7 14 dinonanoyl biphenyl mixtures 949 diodes, active matrix displays 230 diol derivatives, cubic phases 889 diols, chiral nematics 327 dioxanes - disubstituted 51 - optically active 510 dipolar correlation, nematics 98 dipolar groups, smectogens 4 16 dipole dipole interactions - charge transfer 945 ferroelectrics 519 dipole moments - chiral smectics 498 - nematics 98 dipole reorientation, nematics 91 direct adressing, TN displays 205 direct current fields, dynamic scattering 244 director alignment, shear viscosity 143 director distortions, nematics 61 director fluctuations, nematics 173 f, 186 director orientation, shear viscosity 142 director rotation - chiral nematics 336 f - viscosity measurement 156 director states, ferroelectric devices 645 directors - biaxial nematics 933 f - calamitics 14 - chiral nematics 303 - chiral smectics 493 - chromonics 991 - dyes 268 - ferroelectrics 548, 556, 603 - nematics 171 disc-like molecules, discotics 749 disc shape, cubic phases 894 disclinations - biaxial nematics 937 - chiral nematics 350 ff columnarphases 774 discotics (discogens) 693-748 - physical properties 749 ff dislocations - chiral nematics 354 - columnar phases 774 ~
~
Index Vol. 2A and 2B disodium cromoglycate (DSCG) 981 ff disordered crystals, calamitics 19 dispersion, nematics 133 dispersive interactions, charge transfer systems 945 displacements, columnar phases 772 displays - antiferroelectrics 679 dynamic scattering 252,486 nonchiral smectics 480 - twisted nematics 199 see nlso: Heilmeier displays Displaytech microdisplays 650 distilbazole ligands, metallomesogens 903 distortions - chiral nematics 303, 382 ff - columnarphases 771 - dynamic scattering 244 - ferroelectric devices 650 nonchiral smectics 464 dithio ligands, metallomesogens 914 dithiobenzoate, sulfur containing ligands 913 dithiolene ligands, metallomesogens 9 13 dithiolium, synthesis 725 divergence - rotational viscosity 160 - shear viscosity 152 divergence temperature, dimers 814 f DNA, chromonic nature 1003 DOBAMBC, ferroelectric properties 494 dodecycloxy derivatives 857 donor/acceptor (EDA) interactions 945 ff donor sets, metallomesogens 915 ff dopants - chiral nematics 303, 495, 509 chromonics 998 nonchiral smectics 484 doped columnar conductors, applications 795 doping - charge transfer systems 952 - chemically 787 dynamic scattering 243, 250 double bond effects, STN 224 double cell dichroic LCDs 29 1 ff double layer compensated STN displays 2 I6 double pertitectic form, chromonics 988 double swallow-tailed mesogens 855 ff Drehung, nematics 63 drugs, chromonics 981 ff dual frequency adressing, displays 294 Dubois-Voilette-de Gennes-Parodi model 246 f Dupin cyclides, chiral nematics 365 dye concentration, Heilmeier displays 275 dye doped twisted nematic displays 277,283 dye guest-host (DHG) displays 399 dyes - chromonics 981 ff dichroic 259 ff - guest-host effect 257 ff, 271 f -- laser addressed devices 476
1015
dynamic properties - chiral nematics 374 ff - nematics 170-198 - STN/TN 213 dynamic scattering 243-256 - SmAphases 483 dynamic scattering mode (DSM) 243
~
~
~
~
~
~
~
E phases, calamitics 12 edge disclinations, chiral nematics 354 Ehrlich magic bullet, chromonics 984 elastic constants 63, 79 ff - adjustment 225 - ferroelectrics 592, 654 - nematics 69, 171 - nonchiral smectics 464 elastic properties - charge transfer systems 960 - chiral nematics 368 ff - nematics 60-90 electric fields, axis orientated 388 f electrical conductivity, columnar phases 766 f electrical properties - columnar discotics 784 ff - HAT6 781 electrically addressed displays 480 electrically controlled birefringence (ECB) 199 electrically induced transitions 39 electroclinic effect - ferroelectrics 583 ff - SmA-SmC* transitions 33 electrode patterns, positive mode dichroic LCDs 295 electrolytes, chromonics 1000 electron acceptor molecules, columnar nematics 757 electron donor pairs, charge transfer systems 947 electron donorslacceptors 945 ff electron spin resonance (ESR) - guest-host effect 257 - nematics 137 - viscosity 159 electronic mobility, columnar discotics 784 electrooptic applications, HATn 782 f electrooptic switching, hydrocarbon cores 701 electrooptical effect - chiral smectics 494 - displays 290 - hydrogen bonding 973 electrooptical performance, STN 213 electrooptical properties - dynamic scattering 251 - ferroelectrics 604 - guest-host effect 271 f - Heilmeier displays 275 - TNdisplays 202 - viscosity 622 f electrooptics, surface-stabilized states 596 ff electrostriction, ferroelectrics 524 ellagic acids 866, 894
1016
Index Vol. 2A and 2B
elliptically shaped biaxial nematics 934 elongated groups, dichroic dyes 266 enantiomeric excess, ferroelectrics 576 enantiomers, chiral nematics 336 enantiomorphic properties, chiral nematics 371 enantiotropic behavior, anthracenes 700 enantiotropic phases, diketonate ligands 9 10 encapsulation technique, chiral nematics 398 energy conservation, chiral nematics 375 enthalpies - biaxial nematics 936 phasmids 868 erase mechanism, high frequency 390 Ericksen-Leslie theory, nematics 170 escaped radial director fields, nematics 75, 188 ester groups, smectogens 421 ester linkages, nematics 53 esters 694 aromatic 497 - dynamic scattering 243 - phasmids 866 - shear viscosity 152 - substituted mesogens 841 - terminal chains 501 ethanones, chiral nematics 328 ether linkage, AFLCD 680 ethers 702 f - biaxial nematics 935 f terminal chains 501 ethylene complexes, hydrogen bonding 969 Euler angles, ferroelectrics 617 Euler buckling, ferroelectric devices 632 Euler-Lagrange equations, nematics 60 excitons, columnar discotics 790 ff experimental methods - optical properties 128 f - shear viscosity 145 extended range, chromonics 993 extended viewing angle (EVA) 216 external electric field effect 399 extinction coefficient, dichroic dyes 259 -
-
-
F phases, smectics 16 FA-006, ferroelectric devices 656 Faetti theory, nematics 77 failure modes, guest-host effect 273 Faraday-Curie balance, nematics 116 fast exciton migrations, columnar discotics 795 fast responding STN displays 2 19 Felici model, dynamic scattering 248 Fergason approximation 339 ferrielectric states, ferroelectrics 525 ferrocene, metallocen ligands 92 1 ferrocene derivatives, metallomesogens 902 ferroelectric dichroic LCDs 296 ferroelectric host materials, synthesis 435 ferroelectric mixtures, nonchiral smectics 41 1 ferroelectric switching, columnar mesogens 712
ferroelectricity columnar phases 765 ff - hydrogen bonding 97 1 ferroelectrics 5 15-664 - chiral smectic 494 - columnar discotic 794 - doping 952 - smectic 487 fibers, sample preparation 764 field induced distortions, chiral nematics 382 ff filled smectics - charge transfer systems 949 - swallow-tailed mesogens 855 film compensated STN displays 2 17 fingerprint textures - chiral nematics 338 ff, 343 f - chromonics 997 Fizeau interference fringes 130 flavellagic acid, heterocyclic cores 722 flexoelectricity - chiral nematics 382, 392,404 - ferroelectrics 516,555 f, 653 flexoelectrooptic effect 560 f Flory rotameric state model 822 ff flow alignment - columnarphases 776 - shear viscosity 144, 154 flow properties, chiral nematics 379 fluctuations - chiral nematics 342 - columnar phases 771 - shear viscosity 149 - smectic A-C transitions 453 fluhenamic acid salts 999 fluorene derivatives, swallow-tailed mesogens 85 1 fluorescence, columnar discotics 784, 790, 795 fluorescence dichroic LCDs 299 fluorescent dichroic dyes 259 f fluorescent materials, hydrocarbon cores 712 fluorinated chains, polycatenars 875 fluorination, ferroelectric devices 656 fluoro aromatics, STN 221 fluoro substitution, smectogens 424 flygare method, nematics 115 focal conic textures - chiral nematics 343 ff - nonchiral smectics 470 f Foerster transfer rate, columnar discotics 790 formaldehyde, reagents 716 formulations, chiral nematics 305 fourth rank tensor, ferroelectrics 614 frame rate control, Gray levels 207 frame time, TN displays 205 Frank constants - chiral nematics 350 - columnar phases 771 - nematics 79, 175 - nonchiral smectics 464 -
Index Vol. 2A and 2B Frank theory, nematics 60 Frank-Oseen energy 27 Frederiks threshold - nematics 117 - STN 209 TN displays 202 Frederiks transitions chiral nematics 369 ff, 382 ff - nematics 64 ff, 80, 191 free energy, chiral nematics 369 f free standing film, smectic A/C 446,456,460 freezing point, chromonics 1001 Frenkel conduction, active matrix displays 232 frequencies, dynamic scattering 253 frequency splitting, nematics 91 friction coefficients, chiral nematics 374 Friedel-Crafts acylation - hydrocarbon cores 714 - smectogens 428 Friedel-Crafts synthesis 50 functional groups, smectogens 41 1 furanes 723 - heterocyclic cores 736 -
-
G phases, smectics 17 gallic acid, heterocyclic cores 723 gauche conformers - substituted mesogens 836 - swallow-tailed mesogens 855 f gauche form 8 15 gauche link 823 Gaussian transit characteristics, columnar phases 79 1 Gay-Berne theory, nematics 80 gels, chromonics 998 generic model, dimers 8 15 ff ghost effect, AFLCD 679 Ginzburg criterion, smectic A-C transitions 453 glass, sample preparation 764 glass transition, hydrogen bonding 974 globular shape, cubic phases 894 glycoparanoside derivatives, synthesis 74 1 glycoside, cubic phases 889 glycoximate ligands, metallomesogens 9 13 gold, isonitrile ligands 906 Goldstone mode - ferroelectrics 571 f, 603,617 ff - nematics 170, 182, 191 Gooch-Tany curve - cell preparation 267 - TN displays 201 grain boundaries - chiral nematics 355 - columnar discotics 784 grainy textures, chromonics 994 Grandjean-Can0 disclinations - chiral nematics 348 - guest-host effect 267
Grandjean textures chiral nematics 304, 338 ff, 343 f - displays 290 see also: textures grating, transient 182 gray levels - addressing 207 - ferroelectric devices 636 grayscale, ferroelectric devices 632 grid pattern, dynamic scattering 25 I Grignard reagents, heterocyclic cores 733 f group dipole moments, nematics 94 guest-host effect 257-302 guest-host systems - charge transfer systems 945 - substituted mesogens 849 - swallow-tailed mesogens 855 -
Hagenbach-Couette effect, shear viscosity 145 Haller fit - chiral nematics 348 - nematics 115, 122, 129 - rotational viscosity 160 halogen substitution, saturated cores 741 halogenated salicylaldimine ligands 9 16 halogeno end group, smectogens 41 8 halogens, chiral centers 503 handedness - chiral nematics 337 - ferroelectrics 544 HATn materials, electrical properties 782 ff H e N e laser addressed devices 478 head-tail equivalence, antiferroelectrics 665 f heat capacity, OCB 450 Heck coupling, hydrocarbon cores 707 Heilmeier displays - cell preparation 267 - dichroic 204,292 - guest-host effect 258 ff, 275 ff Heisenberg ferromagnet 170 Helfrich model, chiral nematics 380 ff helical C* states, ferroelectrics 568 ff helical smectic C* compounds 623 helical superstructures, chiral smectics 494 helical twist distortions, chiral nematics 303 helical twisted power (HTP) - chiral smectics 507 - chromonics 998 - ferroelectrics 549 helicity, chiral nematics 309, 336, 382 helielectrics - ferroelectrics 525 - nonchiral 577 Helmholtz energy - chiral nematics 376 - dimers 817, 824 heptanoate, synthesis 700 Hermann theorem, ferroelectrics 553 f
1017
1018
Index Vol. 2A and 2B
herringbone textures, chromonics 983 ff, 990, 994 heteroatoms, saturated cores 741 heterocyclic cores, discotics synthesis 720 ff heterocyclic mesogens, chiral nematics 330 heterocyclic rings aliphatic 854 aromatic 500 optically active 508 smectogens 412,420 hetero-intermolecular hydrogen bonding 969 hexa-n-octanoate (BHH), benzene 75 1 ff hexaalkanoyloxy benzenes, discotics 749 ff hexaalkoxy compounds, biaxial nematics 935 hexaalkoxytriphenylenes (HATn) - columnar discotics 781 ff - electrical properties 782 ff hexaamides, charge transfer systems 962 hexaaminobenzene, charge transfer systems 962 hexacatenar mesogens (hexacatenars) 866 hexacyclens 742 hcxaesters 694 hexaethers - biaxial nematics 936 charge transfer systems 954 hexagonal columnar phases, discotics 693, 750 hexagonal phases, polycatenars 883 hexahexylthiotriphylene (HHTT) 754 ff hexakis(alkoxyphenoxymethy1) derivatives 698 hexakis(alkylsu1fone) derivatives 698 hexakispentyIoxytriphenyleneiTNF complexes 955 hexalkanoxytriphenylenes, discotics 749 ff hexatic B-smectic A transition 36 hexatic phases, calamitics 5 hexatic smectic B phase 440 see also: smectic phases hexatic structures 10 hexaynes, charge transfer systems 957 hexylcarbonyl group, substituted mesogens 837 HHTT, electical properties 766, 784 high-frequency stabilization, ferroelectrics 6 10 high frequencies, dynamic scattering 247 f high-resolution X-ray studies 753 hjndered rotation - antiferroelectrics 665 ff - ferroelectrics 573 holes - chevrons 637, 644 - columnar discotics 784 - positive mode dichroic LCDs 294 hollow column structures 999 hollow prism method 129 homeotropic alignment - chiral nematics 345 - displays 284 homeotropic orientations, nonchiral smectics 470 homodyne set-up, shear viscosity 148 homogeneous alignment - cell preparation 266 - displays 284 -
-
-
nematic regime 244 homologous series - birefringence 136 - elastic data 72 Hooke law, ferroelectrics 539, 614 hopping columnar discotics 790 - excitons 795 - Scher-Lax model 767 host birefringence, displays 288 host impact, Heilmeier displays 279 host materials, synthesis 435 host mesogens, double swallow-tailed 855 host mixtures - antiferroelectrics 689 - chiral smectics 507 host molecules, guest-host effect 257 hue, dichroic dyes 259 Hurst exponent, nematics 191 hybrid aligned nematics (HAN) 78 ff hybrid alignment, displays 284 hydrazine derivatives, cubic phases 893 hydrocarbon chains - columnar phases 754 - substituted mesogens 840 hydrocarbon groups, chiral nematics 345 hydrodynamic equations, nematics 170 hydrogen bonded systems 969-978 hydrogen bonding - charge transfer systems 945 - cubicphases 888 hydrogen bridging, charge transfer systems 949 hydrogenated chains, polycatenars 875 hydrophobic interactions, chromonics 986 f hydroquinones, synthesis 695 hydroxy groups, chiral smectics 496 hydroxy substitution, smectogens 424 hypsochromic elanogated groups 266 hypsochromic shift, chromonics 988 hysteresis curves, ferroelectrics 525 -
-
I phases, smectics 16 Idemitsu polymer FLC 65 1 iminomethyl, spacers 846 iminopyridine complexes, metallomesogens 905 improper ferroelectrics 537 f improved Alt-Pleshko addressing technique (IAPT) 207 in plane switching (IPS) 199, 597 inclinations, ferroelectric devices 639 increment systems, diamagnetic anisotropies 124 indanone, synthesis 7 13 index ellipsoid, ferroelectrics 609 indium-tin oxide (ITO) - coatings 484 - laser addressed devices 474 - TN displays 200 induced phase transitions 38 ff
Index Vol. 2A and 2B infinite periodic minimal surface (IPMS) 897 inositiol, synthesis 740 instabilities, columnar phases 772 intensity reflection coefficient, chiral nematics 360 interaction parameters, mesogenic groups 8 16 interaction types, charge transfer systems 945 f intercalated dye molecules, chromonics 997 intercalated smectics, dimers 807 f intercalation, chromonics 1000 intercolumn coupling/separation 755 interdigitated smectics, dimers 808 interference method 129 f interference pattern, biaxial nematics 938 interlayer stacking structures, calamitics 12 interlayers, nonchiral smectics 441, 451 intermediate pitch systems, chiral nematics 40 1 intermolecular interactions - cubic phases 888 hydrogen bonding 976 nematics 186 inversion centers, ferroelectrics 552 inversion walls, rotational viscosity 156 ionic forces, charge transfer systems 945 ionization energy, columnar discotics 784 iridium, ligands 903, 91 7 iron ligands 912, 917 - lyotropic metallomesogens 927 iron tricarbonyl derivatives, ligands 923 isocontrast curves, Heilmeier displays 280 isodesmic behavior, chromonics 988 isonitrile ligands, metallomesogens 906 isopentyl nitrite, reagents 727 isotropic liquids 19 isotropic phases, order fluctuations 174 isotropic states, nematics 91 isotropic-nematic transition (I-N) 23 f -
-
J phases, smectics 17 Jahn-Teller type smectic A-C transitions Jelley aggregates, chromonics 988
453
Keating theory, chiral ncmatics 366 Kelvin chirality, ferroelectrics 543 Kerr effect, nematics 175, 18 1 Ken ellipsometry, charge transfer systems 956 keto groups, hydrocarbon cores 714 ketoximesters, substituted mesogens 841 kink angles, ferroelectric devices 639 Knorr condensation, heterocyclic cores 727 Krafft temperature, chromonics 982 lactic acid, chiral nematics 327 lactones optically active 508 swallow-tailed mesogens 854
-
-
1019
Lambert reflectors, displays 289 lamellar phases - discotics 759 f - polycatenars 880 ff Landau expansion - ferroelectrics 578 ff, 618 f - helical C* state 589 - nematics 174 - smectic A discotics 762 Landau theory - ferroelectncs 516 ff, 530 ff - helical C* state 569 Landau-de Gennes theory - chiral nematics 366 I-N transition 23 f nematics 60, 175 Landau-Khalatnikov equations 6 18 Landold-Bornstein tables - nematics 93 - refractive indices 128 Langevin alignments, antiferroelectrics 683 Langevin functions, ferroelectrics 520, 527, 530 Langmuir crystals 78 Langmuir-Blodgett films 957 Langsverbiegung 63 Larmor frequency 173 f laser-addressed devices, nonchiral smectics 473 f laser recrystallization method 238 lateral segments, substituted mesogens 843 ff lateral substituents nematics 54 smectogens 422 laterally substituted liquid crystals 835-863 lattice parameters - columnar phases 752 - cubicphases 896 lattice vibrations, columnar phases 771 layer shrinkage, ferroelectric devices 633 layer spacing - dimers 806 - ferroelectrics 564 - smectogens 418 layer thickness - ferroelectric devices 654 Gooch-Tarry curves 201 - substituted mesogens 850 layered structures, chromonics 999 layers chiral nematics 336 - cubic phases 893 - nonchiral smectics 471 - rotational viscosity 157 STN 216 - tilted smectic 626 lead, ligands 906, 924 leakage, active matrix displays 23 1 lecithin, chiral nematics 338, 345 Lehmann nomenclature, nematics 128 Lehmann rotation, chiral nematics 377 ff -
-
-
-
-
-
1020
Index Vol. 2A and 2B
Leitz-Jelley microrefractomer 129 Lenard-Jones theory, nematics 80 Leslie coefficients - chiral nematics 374 - columnar phases 775 - rotational viscosity 165 Leslie viscosity, nematics 175 Leslie-Ericksen theory - chiral nematics 337 f, 377 - rotational viscosity 165 life parameters, guest-host effect 271 f Lifshitz terms, ferroelectrics 589 ligands, metallomesogens 902 ff light intensities, ferroelectrics 605 light scattering - Leslie coefficients 166 N-I transitions 24 f - nematics 67 - quasielastic 170 - shear viscosity 147 lightning defects, ferroelectric devices 634 limitations, dynamic scattering 25 1 line singularities, chiral nematics 350 line width, laser addressed devices 478 linear electroptic effects, ferroelectrics 596 f linear optics, chiral nematics 394 f linkages, benzenes 695 linking - chiral smectics 496 - drugs/dyes/nucleic acids 985 f - substituted mesogens 847 f linking groups nematics 47, 137 - smectogens 413,420 liquid crystal displays (LCDs) 199 f local layer geometries, ferroelectrics 629 ff long-chain phenylene bis(benzoates) 835 f long pitch - chiral nematics 400 f chiral smectics 495 ferroelectrics 628 long range order, calamitics 18 Lorentzian band, shear viscosity 149 low frequencies, dynamic scattering 244, 249 low pressure chemical vapor deposition (LPCVD) 238 Luckhurst potential, dimers 816 luminance, guest-host effect 272 lutetium, phthalocyanine ligands 924 lyotropic blue phases, chromonics 998 lyotropic systems - chromonics 981 - cubic phases 887, 893 hydrogen-bonded 977 metallomesogens 926 ~
~
~
~
~
~
M gels, chromonics 998 M phases, chromonic 982
M ribbons, chromonics 996 Mach-Zehnder interferometer 132 macrocycles, hydrocarbon cores 7 19 f macrocyclic complexes, metallomesogens 926 macroscopic flow, chiral nematics 379 macroscopic polarization, ferroelectrics 54 1 ff macroscopic properties - hexatic smectic B 447 - smectic A 443 - smectic C 457 magnetic birefringence see: birefringence magnetic fields, axis orientated 382 f magnetic properties, nematics 113 magnetoelectric methods, nematics 117 magnetooptical methods, nematics 65 Maier-Saupe theory - biaxial nematics 933 - dimers 816 ff - nematics 80, 92 f main chain polyesters, charge transfer systems 956 main chain polymers, hydrogen bonding 974 malondialdehyde, diketonate ligands 91 1 Malvern-3 scheme, ferroelectric devices 648 manganese - cyclometalated complexes 920 - ligands 912,924 Marcelja theory, dimers 823 mass conservation, chiral nematics 375 material parameters - device influences 208 - TN displays 203 materials - ferroelectric devices 653, 657 - laser addressed devices 476 - STN displays 220 matrix addressing, AFLCD 679 Mauguin limit - chiral nematics 341, 356 - nematics 65 - pitch systems 400 - TN displays 200 mauve, chromonics 984 Maxwell equations 356 f MBBA - charge transfer systems 946 f - coherence length 25 - dielectric properties 119 - dynamic scattering 246 - elastic properties 60, 71 - isotropic phase 174 - rotational viscosity 164 McMillan theory - smectic A discotics 760 f - smectogens 416 McMillan-de Gennes theory 26 f mean field behavior, SmA-SmC transitions 3 1 mean field model, smectic A discotics 760 f mechanical instabilities, columnar phases 772 mechanical torque measurements, nematics 1 18
Index Vol. 2A and 2B mechanically induced SmA-SmC transitions 38 Meissner effect 184 melting process, calarnitics 4 melting temperatures - biaxial nematics 937 - chiral nematics 306 - heterocyclic cores 727 - shear viscosity 150 memory effect - chiral nematics 390 displays 289 mercury, carboxylate ligands 906 mesogenic properties, chromonics 1000 mesomorphic properties, polycatenars 866 f, 879 ff mesophase behavior, discotics 694 ff meso-tetra@-alkyl-pheny1)porphyrins 729 metu position phasmids/polycatenars 866 - substituents 848 metachromic color change, intercalated dyes 1002 metacyclophanes, synthesis 719 metal containing liquid crystals 901-932 metal containing materials, cubic phases 893 metal insulator metal matrix address 230 ff metal oxide semiconductor transistor (MOSFET) 239 metal porphynns, applications 783 metallocen ligands 92 1 ff metallogens, synthesis 433 metallomesogens 837, 901-932 - lamellar phases 759 - nonchiral smectics 41 1 methoxy links, smectogens 422 methoxybenzoic acid, hydrogen bonding 969 methyl branched primary alcohols, chiral nematics 313 methyl groups - chiral smectics 496 - smectogens 423,425 methyl orange, chromonics 981 methyl substituents, synthesis 704 methyl sustituted dioxanes, optically active 5 10 methylene groups, intercalated smectics 807 methylene units, biaxial nematics 937 micelle formation, chromonics 981, 986 microdensitometer scanning, biaxial nematics 94 1 microdisplays, ferroelectric devices 650 microdropletsicylinders, nematic dynamics 184 microgrooved surfaces, chiral nematics 345 Miesowicz viscosity 147 chiral nematics 380 migration, columnar discotics 790, 795 Miller indices, cubic phases 896 miscellaneous type I chiral nematics 323 Mischphasen, fliissigkristallin 946 miscibility, chromonics 1000 mixed phases, charge transfer systems 946 mixtures - antiferroelectrics 680, 689 - benzene derivatives 696 -
-
-
biaxial nematics 938 chiral nematics 305 - chiral smectics 495 - Cole-Cole plots 100 - dichroic 271 f - dynamic scattering 243 - hydrogen bonding 976 - MBBMCPB 946 - NAC multicritical point 34 - nematics 102 - nonchiral smectics 41 1,472 - refractive indices 135 - rotational viscosity 163 - W7-W82 640 mobility anisotropy, columnar discotics 784 modes - displays 284 - laser addressed devices 473 - normally whiteblack 204 see also: Goldstone see also: White-Taylor modulated smectics, dimers 812 moieties - charge transfer systems 949 - chiral nematics 307 - dimers 801 molecular aggregation, chromonics 988 molecular engineering, dichroic dyes 260 molecular motion, charge transfer systems 956 molecular packing, substituted mesogens 849 molecular reorientations, local 187 molecular shapes - cubicphases 894 - dimers 829 f molecular structures - birefringence 137 - chromonics 983 - columnar discotics 782 ff rotational viscosity 160 molecular theories - dimers 814 ff - ferroelectrics 558 ff - nematics 60 molybdenum, carboxylate ligands 908 momentum conservation, chiral nematics 375 monoacrylhydrazide, cubic phases 889 monochromic mixtures, guest-host effect 274 monoclinic symmetries, ferroelectrics 609 monodentate ligands, metallomesogens 902 monolayer structures, smectics 9 monomers, dimers 8 18 monosaccharides, hydrogen bonding 969 monostilbazole ligands, metallomesogens 903 monotropic nematic phases, diketonate ligands 910 multiplexing, Heilmeier displays 28 1 multiplexing phase change dichroic LCDs 291 multiynes 697, 704 - charge transfer systems 957 -
-
1021
1022
Index Vol. 2A and 2B
muscovite, sample preparation 764 myoinositol ester, synthesis 740 mytilitol derivatives, synthesis 740 N phases, chromonic 982 N-0 donor sets, metallomesogens 9 15 ff nafcillin sodium, chromonics 985 nafoxidine 999 nafoxidine chloride 985 naphthalene compounds, ferroelectric devices 657 naphthalene hexaynes 957 naphthalenes 699 f - nematics 52 naphtoquinone 260 naphtyl, substituted mesogens 837 narrow band dyes, guest-host effect 274 negative dichroic dyes, guest-host effect 259 ff negative dielectric anisotropy materials 48 1 Nehring-Saupe theory - ferroelectrics 559 - nematics 61 nematic double cell dichroic LCDs 291 f nematic ground state deformations, ferroelectrics 555 f nematic-isotropic transitions - chromonics 987 - dimers 803 f nematic phases - calamitics 6f - discotic 694 f, 757 f - ferroelectric 794 polycatenars 882 nematic-smectic A transition (N-SmA) 26 ff nematic-smectic A-C multicritical point (NAC) 33 nematics 47-59 - chiral 303-334 - material requirements 654 - melting processes 5 networks, hydrogen bonding 974 Neumann principles, ferroelectrics 541 ff, 609, 615 nickel, ligands 914 f, 924 nitration, hydrocarbon cores 709 nitrile ligands, metallomesogens 902 nitro derivatives, polycatenars 878 nitrobenzene derivatives, charge transfer 948 nitrobiphenylcarboxylic acids, synthesis 433 nitroester, phasmids 866 nomenclature - charge transfer systems 954 - discotics 750 - phasmidsipolycatenars 865 ff nonaoate, synthesis 700 noncalamitic charge transfer systems 952 nonchevron structures, ferroelectric devices 655 f nonchiral helielectrics, ferroelectrics 577 nonchiral smectics, synthesis 41 1 4 4 0 nonchiral smectics C, ferroelectrics 628 noncollective modes, ferroelectrics 620 -
nondiscotic molecules, columnar phases 755 f nonenantiomorphic properties, chiral nematics 37 1 nonionic solutes, chromonics 1001 nonlinear optics (NLO) - chiral nematics 394 f - ferroelectrics 527 f nonmesognic dopants, chiral nematics 303 nonmetallomesogens, nonchiral smectics 4 1 1 nonpolarity, ferroelectrics 522 nonracemic terminal chains, chiral smectics 498 nonselect states, transmission spectra 2 13 nonyl groups, substituted mesogens 841 normal mode, laser addressed devices 473 normally whiteblack, TN displays 200 nuclear magnetic resonance (NMR) columnar LC 752 f - guest-host effec 257 - nematics 116 f, 137 - viscosity 159 nucleation sites, guest-host effect 273 nucleic acids, chromonics 981 ff -
oblique mesophases, discotics 694 oblique phases, polycatenars 883 OCB compounds, nonchiral smectics 442 OCBs - nonchiral smectics 442 - synthesis 426 f octa substituted porphyrins 727 octa(a1koxy)phthalocyanine 73 1 octa(a1koxycarbonyl)phthalocyanine 734 octa(a1koxymethyl)phthalocyanine 730, 736 octa@-alkoxylpheny1)phthalocyanine 735 octa(alky1)phthalocyanine 733 ff octanoate 700 odd-even effect - antiferroelectrics 686 - dimers 803 oily streaks, chiral nematics 35 1 Oldane-Barber0 paradox, nematics 76 oligo(ethy1ene oxide), spacers 80 1 oligomers - chitin/cellulose-based 74 1 - discotic 801-833 oligo(siloane), spacers 80 1 OmIs, synthesis 430 Oms, synthesis 428 one pitch double cell dichroic LCDs 292 Onsager theory, nematics 80, 103 operating voltage, displays 286 operation modes, TN displays 199 OPHOBIOPHOFB complexation, hydrogen bonding 971 optical activity, ferroelectrics 549 optical anisotropy - ferroelectrics 608 - STN 221 optical applications, chiral nematics 394 f
Index Vol. 2A and 2B optical indicatrix, ferroelectrics 609 optical propagation - Bragg approach 362 f - chiral nematics 356 f optical properties biaxial nematics 937 charge transfer systems 956 - chiral nematics 303 f, 343 f, 382 guest-host effect 271 f - nematics 128 14 - unactivated states 200 optical purity, chiral nematics 309 optical reflectivity, OCB 450 optical rotary dispersion (Om) - chiral nematics 336 - ferroelectrics 549 optical rotation, chiral nematics 340, 399 optical textures chiral nematics 343 f chromonics 982, 989 ff optical transmission, ferroelectrics 604 f optically active centers, chiral smectics 493 order breakdown, calamitics 3 order fluctuations, isotropic phase 174 order parameters calamitics 18 - chiral nematics 342 - dichroic dyes 259 f, 266 - ferroelectric helical C* state 568 f ferroelectrics 537 guest-host effect 268 ff Heilmeier cells 279 - I-N transitions 23 - Pikin-Indenvom 592 - smectic A-C transitions 452 thin films 461 ordered phases, chromonics 998 organonitrile ligands, metallomesogens 902 organso siloxanes, chiral nematics 396 orientation - chiral smectics 493 - nonchiral smectics 470 f - shear viscosity 142 orientational correlation, calamitics 19 orientational fluctuations, below T, 177 orientational order - calamitics 3 discotics 693 generic model 819 - long-range 170 - nematics 136 oriented gas, guest-host effect 257 ortho position - phasmids/polycatenars 866 substituents 841 orthogonal phases, calamitics 5 orthogonal smectics 7,470,493 orthopalladated metallomesogens 837 orthorhombic symmetries, ferroelectrics 609 -
-
-
Oseen theory chiral nematics 375 - ferroelectrics 556 f - nematics 60 Oseen-Frank expression - antiferroelectrics 578 - ferroelechics 559 f, 614 - helical C' state 589 OS1, synthesis 430 Osipov model, ferroelectrics 592 osmometry, chromonics 988 out-board terminal dipoles, smectogens 4 16 oxarole derivatives, static dielectric constants 97 oxatruxenes, synthesis 724 oximesters, substituted mesogens 837, 841 oxiranes, optically active 506 oxycarbonyl connectors, antiferroelectrics 687 oxygen inertness, columnar discotics 792 -
-
-
-
-
-
-
-
-
-
-
1023
K systems, charge transfer 945 P phase, chromonics 993 f PAA - domain pattern 245 - elastic data 65 - NMR 176 - shear viscosity 155 packing - calamitics 11 - chromonics 982 - ferroelectrics 558 - smectogens 418 - substituted mesogens 849 page blanking, displays 486 palladium - cyclometalated complexes 91 8 - ligands 903,914 f palladium organyls 960 f pallado nematogens 962 para position - phasmids/polycatenars 866 - substituents 846 paraelectric phases, ferroelectrics 525, 539 paraffinic chains, polycatenars 865 f, 879 ff paramagnetism, ferroelectrics 5 17 parity, chiral nematics 309 Parodi model chiral nematics 377 - rotational viscosity 165 partial fluorination, ferroelectric devices 656 passive layers, STN 2 17 passive matrix adressing, TN displays 205 Peierls-Landau instability columnarphases 754 - nonchiral smectics 444, 465 pentacatenars 869 pentanol, chiral smectics 496 pentayne esters, charge transfer systems 958 peptide nucleic acids, chromonics 1004 -
-
1024
Index Vol. 2A and 2B
perceived contrast ratio (PCR), guest-host effect 272 perfluoro arenes, charge transfer systems 964 perfluoro chains, polycatenars 875 periodicity - columnar discotics 792 - cubic phases 896 permeation, chiral nematics 380 permittivity - columnar discotics 789 ferroelectrics 622 tilted smectic layers 626 perturbations, chiral ferroelectrics 583 - dichroic dyes 260 perylene 712 - dichroic dyes 260 phase change effect dichroic displays 284,292,295 phase compensation, STN 216 phase diagrams - chromonics 988 ff pyridineicarboxylic acid systems 972 f phase ranges, ferroelectric devices 654 phase sequences, cubic phases 888 phase structures - calamitics 3 19 chromonics 981 phase transitions - induced 38ff - rod-like crystals 23 24 phasmids 865-885 - columnar phases 755 phenanthrenes, synthesis 701 f phenolipyridine interactions, hydrogen bonding 976 phenyl groups, biaxial nematics 935 phenyl mesogens, chiral nematics 314 phenyl rings - AFLCD 680 charge transfer systems 948 substituted mesogens 840 ff phenylacetylene macrocycles, synthesis 7 19 phenylazosalicylic acid esters, substituents 837 phenylbenzoates 429 - cyan0 terminated 414 phenylethylbiphenyl mesogens, chiral nematics 3 14 phenylsilane compounds, cell preparation 267 phoney dyes, guest-host effect 269 photochemically induced transitions 39 photoconduction, columnar phases 768 photocyclization, phenantrenes 701 photodecomposition, dichroic dyes 265 f photostability, dichroic dyes 260 ff phthalocyanine based gas sensors 792 phthalocyanine ligands, metallomesogens 923 f phthalocyanines 730 f, 739 applications 783 physical properties - chiral nematics 335 - columnar phases 775 f dichroic mixtures 271 f - discotics 749 32 -
-
-
-
-
-
-
guest-host effect 271 f laser addressed devices 478 - nonchiral smectics 441 29 - SmAmixtures 472 - smectic A-C transition 452 piezoelectric effect 524, 539 f, 550, 586 piezoelectric/flexoelectric analogies 559 Pikin-Indenbom order parameter 592 f pincements, chiral nematics 354 pinches, chiral nematics 354 pipet shaped biaxial nematics 934 pinmidine, cyclometalated complexes 92 1 pitch lengths - biphenyls 317 - chiral nematics 305, 336 pitch lines, cell preparation 267 pitchlcell thickness ratio, ferroelectrics 628 pitches - chiral nematics 365 f, 400 ff, 495 - ferroelectric devices 654 guest-host effect 273 pixels - active matrix displays 230 f - displays 486 - ferroelectric devices 630 - positive mode dichroic LCDs 295 - TN displays 205 planar alignment, chiral nematics 345 planar configurations, nematics 75 planar radiavpolar director fields 188 plate movement, shear viscosity 146 plates, positive mode dichroic LCDs 295 platinum, ligands 903,913, 917 platinum complexes, chromonics 1004 pleochroic dyes - guest-host effect 259 ff, 265 f - positive mode dichroic LCDs 292 pleochroichegative dichroic dyes admixtures 296 Pockel modulators, ferroelectrics 527 point defects, chiral nematics 350 point symmetries, ferroelectrics 546 Poiseuille flow 380 ff polar compounds, charge transfer systems 949 polar functional groups, smectogens 41 1 polar groups, core end 415 polar materials, ferroelectrics 520 ff polar salicylaldimine ligands 9 I6 polar substituents, polycatenars 875 f polarizability 114, I36 - columnarphases 775 polarization ferroelectrics 516 ff - macroscopic 541 ff - spontaneous 494 f, 665 polarized IR spectroscopy, antiferroelectrics 666 polarized modes, displays 284 polarizers - cell preparation 267 f electrooptic measurements 606 -
-
-
-
-
Index Vol. 2A and 2B ferroelectric LCDs 297 Heilmeier displays 281 polyamides, biaxial nematics 940 polycatenar compounds, thermotropic cubic phases 891 polycatenars 865-885 polydentate ligands, metallomesogens 923 ff polyesters, charge transfer systems 956 polyimides, chiral nematics 337 polymer dispersed dichroic LCDs 297 polymer dispersed liquid crystal (PDLC) 400 polymer dispersed smectic devices 487 polymer rubbing, chiral nematics 345 polymeric complexes - self-assembly 973 - uracilidiamino pyridine systm 975 polymeric salicylaldimine ligands 917 polymers - diketonate ligands 9 11 - discotic 776 polymesomorphisms, discotics 693 polymorphism, smectic 805 f polysiloxaneistilbazole, self-assembly 974 polytetrafluoroethylene (PTFE) 345 polyvinyl alcohol (PVA) cell preparation 266 - chiral nematics 345 Poniewierski-Stecki relations, nematics 79 porphyrins 727 aggregation 945 applications 783 ligands 925 ff positional order, calamitics 3, 19 positive dielectric anisotropy materials 48 1 positive mode dichroic LCDs 292 f potassium, alkanoate ligands 906 pressure dependence chiral nematic pitches 365 rotational viscosity 164 pressure-temperature phase diagrams, discotics 762 ff pretransitional dynamics, N-SmA phase I83 pretransitional effects - antiferroelectrics 675 f confined geometry 188 projection systems, laser addressed devices 478 propanes, chiral nematics 32 1 propanonitriles, chiral nematics 320 proton donor-acceptor pairs 949 pulse modulation, Gray levels 207 f purine bases, chromonics 1003 purine nucleosides, chromonics 988 purity dichroic dyes 259 - optical 304, 309 Jyramidics, hydrocarbon cores 715 iyranose sugars, synthesis 741 jyrenes, columnar discotics 794 iyridazines, disubstituted 52 -
-
-
-
-
-
-
-
~
~
1025
pyridineicarboxylic acid systems, hydrogen bonding 969ff pyridines - disubstituted 51 - ligands 904 f - reagents 695, 724 pyridinium salts, dynamic scattering 243 pyrillium, synthesis 720 f pyrimide bases, chromonics 1003 pyrimidine host mixtures, antiferroelectrics 689 pyrimidine nucleosides, chromonics 988 pyrimidines 436 - disubstituted 52 pyroelectricity, ferroelectrics 516 f, 528, 552 pyrogallol, synthesis 7 19 pyrrole trimethyl ester, synthesis 728 quadratic torques, ferroelectrics 599 f quadrupole-quadrupole interactions 946 quadrupole splitting, columnar phases 753 quasibookshelf structures (QBS), chevrons 637 quasielastic light scattering, nematics 170 quasihomeotropic geometries, ferroelectrics 625 quater wave plate dichroic displays 282 f Querverbiegung, nematics 63 quinones, synthesis 695 racemic compounds chiral nematics 306 - swallow-tailed mesogens 853 racemic materials, chiral nematics 342 racemic structures, ferroelectrics 549 radial configuration, nematics 185 radiation, guest-host effect 259 Raman experiments, antiferroelectrics 667 randomly constrained nematics 189 Rapini-Popular terms 77 rational routes, hydrocarbon cores 705 Rayleigh interferometer 131 Rayleigh scattering 172 reagents, hydrocarbon cores 694 ff rectangular phases - biaxial nematics 933 - columnar discotics 693 f, 750 - polycatenars 883 red-green-blue (RGB) filters 2 18 red-green-blue (RGB) polarization 395 red shift, chromonics 988 reductive sustituent removal, hydrocarbon cores 7 10 re-entrant phases, chiral nematics 365 f re-entrant phenomenon, nematics 98 reflection band, chiral nematics 340 reflections - chromonics 992 - cubic phases 896 reflective mode, diyplays 480 reflective monochrome displays 63 I -
1026
Index Vol. 2A and 2B
reflectivity, OCB 450 reflectors - cell preparation 267 - dynamic scattering 244 reflux, hydrocarbon cores 708 refractive indices - chiral nematics 304, 341 f, 348 - nematics 67, 128, 182 refractometers - nematics 129, 348 see also: Abbt refractometers Reinitzer cholesterol, chiral nematics 335 relaxation - columnar discotics 789 - ferroelectrics 618 - nematics 91, 99 relaxation method, rotational viscosity 157 relaxation time, dynamic scattering 249 reminescent contrast, displays 289 reorientation - ferroelectrics 520 - molecuar 187 - smectic A 481 - transient laser-induced 181 resistivity - columnar discotics 792 - dichroic dyes 259 - dynamic scattering 243 resolution, ferroelectric devices 630, 657 response times - AFLCD 679 - nonchiral smectics 485 retardation, chiral nematics 304 retardation factor, nematics 92 retardation plates, chiral nematics 395 reticulated textures, chromonics 995 reverse mode, laser addressed devices 475 reverse twist, chiral nematics 336 rhenium, cyclometalated complexes 920 rhodium, ligands 904, 908, 917 rhodium carboxylates, charge transfer systems 962 ring bromination, heterocyclic cores 733 ring configuration, active matrix displays 230 ring structures - nematics 47 - smectogens 412,418 ring substituents, hydrocarbon cores 708 RNA, chromonic nature 1003 rochelle salt, ferroelectrics 5 15 rod-like crystals, phase transitions 23 24 rod-like mesogens, laterally substituted 835 ff rod-like molecules, isotropic states 91 rod-like phases, biaxial nematics 933 rotational modes, ferroelectrics 613 ff rotational viscosity 155 f RU3 1156, chromonics 98 1 rubbing - chevrons 644 - ferroelectric devices 649
mffigallol, synthesis 700 ruffigallol hexaalkanoates, ellagic acid 866 ruthenium - carboxylate ligands 908 - lyotropic metallomesogens 927 ruthenocene, metallocen ligands 923 Ryckaert-Bellemans potential, dimers 827
saccharides, pyran rings 741 salicylaldimine ligands, metallomesogens 9 15 ff sample geometries, ferroelectrics 625 f sample preparation, discotics 764 sanidic aromatic polyamides, biaxial nematics 940 saturated cores, discotics 739 ff Saupe theory, nematics 79 Saupe-Nehring theory - ferroelectrics 559 - nematics 61 scaling law, cone mode viscosity 602 f scattering centers, nonchiral smectics 485 scattering geometries, shear viscosity 147 Scher-Lax model, columnar phases 767 Schiff base compounds, charge transfer systems 948 Schiff bases - chiral nematics 306, 313 - chiral smectics 493, 496 - dichroic dyes 260 - dynamic scattering 243 f - laser addressed devices 476 - salicylaldimine Iigands 9 I5 - shear viscosity 152 - smectogens 418 schlieren textures - biaxial nematics 937 - chiral nematics 344 ff - chromonics 982, 994 - nonchiral smectics 470 - smectics 461 - triphenylene 757 Schroder-van Laar equation 305 Schwarz P surface, cubic phases 898 screw disclinations, chiral nematics 353 scylloinositol, saturated cores 739 second harmonic generation (SHG) - chiral nematics 396 - ferroelectrics 554 f seignette electricity, ferroelectrics 530 selective reflection, chiral nematics 335 ff, 349 selenoethers, synthesis 702 self-assembly, hydrogen bonded complexes 969 ff self-beating scattering, nematics 172 self-diffusion 173, 186 self-organizing periodicity, columnar discotics 792 shapes - biaxial nematics 934 -- charge transfer systems 945 shear effect, cholesteric symmetry 5 54 shear viscosity coefficients, determination 142 f
Index Vol. 2A and 2B shifts, chromonics 988 short-pitch chiral nematics 403 short-pitch chiral smectics 509 short-pitch materials, DHM 587 short-range order, calamitics 18 side chain polymers, hydrogen bonding 973 sign convention, ferroelectrics 559 f silane derivatives, chiral nematics 338 silanediols, hydrogen bonding 969 silicon, phthalocyanine ligands 924 silicon monoxide evaporation, chiral nematics 345 silicon nitride layer, active matrix displays 23 1 silicon TFT matrix address 234 siloxane compounds, cubic phases 889 siloxanes, chiral nematics 396 silver, ligands 903 f silyl, phthalocyanine ligands 925 Simon-Glatzel equation 165 single bond, nematics 47 Sirius Supra Brown RLL 98 1 SmA*-SmC* transition, ferroelectrics 571 f SmA-SmC transitions 29 ff SmA-SmC* transitions 32 smectic A dichroic LCDs 298 smectic C* states, ferroelectrics 546 smectic materials 427 ff requirements 654 smectic mesogens, nonchiral 471 smectic phases - dynamic scattering 250 - ferroelectrics 613 - flexoelectrics 613 - nonchiral 441 ff substituted mesogens 848 - swallow-tailed mesogens 855 smectic polymorphism, dimers 804 smectic structures, calamitics 7 smectic transitions (C,I,F,G) 41 f smectics - melting processes 5 nonchiral 411440 orthogonal 7,470,493 smectogens synthesis 41 1 ff soap phases, chromonics 982 soft modes, ferroelectrics 617 Soleil-Babinet compensator 216 solid phase crystallization method 238 solubility, dichroic dyes 259 f solutes, chromonics I001 space-time behavior, ferroelectrics 601 spacer chains, chiral nematics 307 spacers - charge transfer systems 952 dimers/oligomers 801 ff ferroelectric devices 650 substituted mesogens 843 ff spacing ferroelectrics 564 -- interlayers 441 ~
~
~
~
~
~
~
~
1027
spectral bandwith - chiral nematics 349 - dichroic dyes 259 spectral transmission, dynamic scattering 25 1 spectroscopic techniques, guest-host effect 257 spin lattice relaxation - MBBA 177 - nematics 173, 188 spin-spin coupling, guest-host effect 258 splay bend deformation, ferroelectrics 558 splay director distortions, nematics 62 splay distortions, nonchiral smectics 464 splay elastic constants - nematics 171 - smectic A 444 splay elastic deformations, chiral nematics 370 splay-twist director states 646 splitting, columnar phases 753 spontaneous polarization - antiferroelectrics 665 - chiral smectics 494 ff - ferroelectrics 526 ff, 533 ff, 573, 654 spoon shaped hiaxial nematics 934 stability - dichroic dyes 259 - hydrogen bonding 973 stabilization - chromonics 986 f - ferroelectrics 610 stacking - calamitics 12 - charge transfer systems 962 - chromonics 982,986 - columnar discotics 78 1 - discotics 694, 750 static dielectric constants, nematics 92 static properties, chiral nematics 342 f statistical synthesis, hydrocarbon cores 705 Steglich method, hydrocarbon cores 701 steric packing, ferroelectrics 558 sterically coupled fields, ferroelectric devices 635 sterol based mesogens, chiral nematics 3 11 stilbazole complexes, ligands 904 Stokes law, ferroelectrics 620 storage mode, chiral nematics 390, 396 strands, sample preparation 764 streaks 351 street defects, ferroelectric devices 634 striations, dynamic scattering 247 structures biaxial nematics 934 f chromonic 981 ff, 999 - dichroic dyes 260 - dimers 801 ff - discotic 749-780 hydrogen bonding 969 - nematics 6 - nonchiral smectics 41 1 - polycatenars 882 ff ~
~
~
1028
Index Vol. 2A and 2B
subphase sequences, antiferroelectrics 685 substituents - lateral 422, 836 - smectogens 411 substituted derivatives, unsymmetrically 705 substrate layers, columnar discotics 793 substrate surfaces, chromonics 987 suface stabilized states, ferroelectrics 544 ff sugar derivatives, synthesis 742 sugar phosphate chain, DNA chromonics 1003 sulfanilamides, chromonics 984 sulfolane, reagents 726 sulfur containing ligands, metallomesogens 9 13 ff superconductivity - ferroelectrics 528 - nematics 183 superdisk benzene, hydrocarbon cores 720 superdisk structures - heterocyclic cores 735 - hydrocarbon cores 698 supertwist dichroic effect displays (SDE) 296 supertwisted nematic displays 208 supertwisted nematics (STN) 199 f, 399 supraconducting quantum interference device (SQUID) 116 supramolecular hydrogen bonding 969 surface alignment chromonics 991 - ferroelectric devices 647 - shear viscosity 143 surface defects, ferroelectric devices 635 f surface-like elastic constants, nematics 73 surface relaxation, columnar discotics 792 surface stabilized ferroelectric LC (SSFLC) 545 ff - device (SSFLCD) 494,630 surface tension, nonchiral smectics 464 surface tensor model, dimers 823 surface treatment, chevrons 645 surfactant complexes, lyotropic metallomesogens 927 susceptibility 118 - ferroelectrics 520 f, 535 f, 584 f - N-SmA transitions 27 nematics 113 swallow-tailed liquid crystals 835-863 swallow-tailed mesogens 850 ff switch matrix address, MOSFET 239 switching chevron interfaces 636 - chiral nematics 345 - columnar discotics 794 - columnarphases 765 ferroelectric devices 645 ferroelectrics 597, 600 f - hydrocarbon cores 7 12 - QBSgeometry 649 - V-shaped 675 f switching speed - displays 284, 287 guest-host effect 273 -
-
-
-
-
switching time, dynamic scattering 253 symmetries - chiral nematics 308 - ferroelectrics 546 ff, 609 synthesis - biaxial nematics 935 ff - chiral nematics 303-334 - discotics 693-748 - nonchiral smectics 411440 - phasmidsipolycatenars 866 Talbot-Rayleigh interferometer 13 1 tantalum pentaoxide layer, active matrix displays 232 TAPA, charge transfer systems 950 ff TBnAS, synthesis 428 temperature dependence - chiral nematic pitches 365 - columnar discotics 784 - director fluctuations 179 - displays 291 - optical properties 132 - rotational viscosity 160 - shear viscosity 150 - SmB,,,-SmA transitions 36 temperature ranges, dynamic scattering 243,25 I template structures, nonchiral smectics 4 11 tensor components, ferroelectrics 6 17 f tensor increment, susceptibility 124 terminal aliphatic chains, chiral smectics 496 terminal aliphatic rings, smectogens 4 14 terminal branches, swallow-tailed mesogens 853 terminal chains, chiral nematics 307 terminal groups - nematics 47, 56 - smectogens 413,417 terminal polar compounds, charge transfer systems 949 terphenyl diesters, optically active 509 terphenyl mesogens, chiral nematics 3 14 terphenyls, nematics 48 tetraalkanoates, ellagic acids 866 tetraazaporphyrin, synthesis 729 tetrabenzocyclododecatetraene, synthesis 7 17 tetrahenzotriazaporphyrin, synthesis 735 tetracatenars 869, 876 tetracyano-p-quinodimethane (TCNQ) 722, 950 f tetracyanoethylene (TCNE) 950 f tetraesters, synthesis 695 tetrahydropyran, synthesis 74 1 tetrakis(oligo(ethy1ene-0xy)phthalocyanine) 736 tetramethylbenzidine, charge transfer systems 948 tetrapyrazinoporphyrazine 734 tetrazine, dichroic dyes 260 textures - chiral nematics 304, 338 ff, 350 f - chromonics 982 ff, 989 ff, 994 ff - columnar phases 775 - displays 290
Index Vol. 2A and 2B ferroelectric devices 638 nonchiral smectics 470 smectics 461 see also: schlieren textures thallium, carboxylate ligands 906 thermal conductivity, chiral nematics 377 thermal effects, chiral nematics 397 f thermally addressed displays 480 thermochromic mixtures, chiral nematics 305 thermomechanical coupling, chiral nematics 377 thermotropic calarnitics, melting processes 4 f thermotropic hydrogen bonded compounds 976 thermotropic phases - cubic 887-900 optically negative 939 thiatruxenes, synthesis 724 thick films, orderparameters 462 thick layers, rotational viscosity 158 thick walls, ferroelectric devices 639 thickness, Heilmeier cell 279 thin film diode, active matrix displays 230 thin film silicon PIN diode (TFD) 230 thin film transitor (TFT), active matrix displays 233 thin films, hexatic B 450 thin films, SmA-SmC transition 35 thin films - SmB,,,-SmA transitions 37 - smectic C 460 thin layers, rotational viscosity 157 thin walls, ferroelectric devices 639 thioester groups, smectogens 42 1 thioethers, synthesis 702 f thiopyran derivatives, noncalmitic systems 953 thiranes, optically active 506 three ring materials, STN 223 threshold characteristics, displays 275, 286 threshold voltage, nonchiral smectics 485 thresholdless antifenoelectricity (TLAF) 665 f, 675 f, 682 tigerskin textures 990, 994 tilt alignment, nonchiral smectics 47 1 tilt angles - AFLCD 681 chiral smectics 496 - ferroelectrics 599 f, 607, 654 SmA-SmC transitions 30,452 -- STN 208 tilt plane, antiferroelectrics 671 tilt SmA device 482 tilted columns, discotics 750 tilted hexatic phases, smectics 461 tilted phases - calamitics 5 - smectic 13 tilted smectic layers, ferroelectrics 626 f tilted smectics 470,493 tilted stacks, chromonics 988 tin, phthalocyanine ligands 924 toluene, reagents 712 -
-
-
-
-
1029
torque, rotational viscosity 155 torque coupling, ferroelectrics 597 torsional shear flow, nematics 150 total reflection method, nematics 129 trans-cis back reactions, chiral nematics 368 trans conformation - biaxial nernatics 935 - polycatenars 875 trans form, dimers 8 15 trans link, dimers 823 transient laser-induced molecular reorientation 181 transition sequences, isotropic 760 transition temperatures - alkyl chain homologs 472 - biaxial nematics 936 f - chiral nematics 3 11 ff - chiral smectics 493 - dimers 803 f - generic model 819 - heterocyclic cores 720 ff - hydrocarbon cores 695 ff - I-N 24 - phasmids 868ff - smectogens 418 - substituted mesogens 836 f - triphenylene derivatives 758 f translational diflksion induced rotation (TDIR) 186 translational order, calamitics 3 translational properties, dimers 80 1 transmission - AFLCD 681 - displays 287 - dynamic scattering 251 ferroelectrics 604, 645 - guest-host effect 270 f - Heilmeier displays 277 - STN 210 - TN displays 202 transmittance, optical 677 transport properties, columnar discotics 79 1 trapping times, columnar discotics 792 hialkoxycinnamic acids, biaxial nematics 935 ff triazacarbonyl, amine ligands 926 tribenzocyclononatrienes, synthesis 7 15 tricarboxamides, synthesis 697 tricatenars 874 f triclinic symmetries, ferroelectrics 609 tricritical behavior, SmA-SmC transitions 30 f, 453 tricyclodecane, chiral nematics 328 tricycloquinazolines, synthesis 726 triesters, phasmids 868 trimers - hydrocarbon cores 7 16 oligomeric systems 813 trinitrofluorene (TNF) - charge transfer systems 950 - columnar nematics 757 triphenylene compounds, columnar phases 766 triphenylene derivatives, discotics 749 fF -
-
1030
Index Vol. 2A and 2B
triphenylenes 702 f, 954 triplet excitons, columnar discotics 790 f, 795 tristable switching, antiferroelectrics 675 truxenes 713,753 f trypan red, chromonics 984 tubular mesophases 742, 755 tumbling, anisotropic 257 tunability, columnar discotics 792 tungsten, stilbazole ligands 904 turn on delay, STN 214 twist angle, STN 208 twist director distortions, nematics 62 twist distortions, chiral nematics 303 twist elastic constants - nematics 171 - nonchiral smectics 471 twist elastic deformations, chiral nematics 370 f twist sense, chiral nematics 309 twist-bend structure, ferroelectrics 576 twistane derived mesogens, chiral nematics 328 twistanol derivatives, chiral nematics 329 twisted Heilmeier cell 278 twisted nematics 199-229 - chiral 341 f - displays 277 - ferroelectrics 598, 632 - guest-host effect 273 twisting power, ferroelectrics 549 two branch configuration, active matrix displays 23 1 two frequency addressed SmA devices 486 two ring materials, STN 222 type I chiral nematics 3 I 1 type IIiIII chiral nematics 325 f Ullman coupling, hydrocarbon cores 707 ultrasoud methods, shear viscosity 150 unactivated states, TN displays 200 uracilidiamino pyridine system 975 ff
V shaped switching, antiferroelectrics 675 f, 682 van der Waals forces, charge transfer systems 945 van der Waals interlayer interactions 45 1 van der Waals models, nematics 80 van der Waals spheres, dimers 824 vanadium - diketonate ligands 9 12 - lyotropic metallomesogens 927 vanadyl, salicylaldimine ligands 917 veratole, reagents 716 viewing angle - AFLCD 679 - dynamic scattering 251 - Heilmeier displays 277 - STN 216f - TN displays 202 viscosity - chiral nematics 374 f
chiral smectics 495 dynamic scattering 243 - ferroelectrics 547, 654 - nematics 142-169 - rotational modes 613, 617 scaling laws 602 see also: Leslie viscosity Vogel-Fulcher plot 190 Volterra process 350 Vuks equations, nematics 133 Vuks field correction factor 348 -
-
-
W7-W82 mixture, ferroelectric devices 640 walls, ferroelectric devices 639, 634 f water molecule interaction, chromonics 986 Watson-Crick double helix 1003 wave equation approach, optical propagation 356 f wavelength - chiral nematics 339 - dichroic dyes 259 - dynamic scattering 251 wedge disclinations - chiral nematics 35 1 - columnar phases 774 wedge parameters, nematics 130 Weiss theory, ferroelectrics 532 f Wenz synthesis, hydrocarbon cores 707 white operation, STN 2 15 WhiteTaylor displays 204,286 ff White-Taylor dye guest host cell 402 White-Taylor modes - cell preparation 268 - guest-host effect 258 f Wilhelmy plate technique 465 Williams domains, dynamic scattering 243 ff write mechanism, low frequency 390 write speed, laser addressed devices 478 Wulf model, smectogens 417
X ray diffraction biaxial nematics 940 f chromonics 991 ff X ray scattering, critical exponents 28 X ray studies, antiferroelectrics 671 xerography, columnar discotics 792 xylene, reagents 7 12 -
-
YAG lasers 478 yellow mode, STN 2 13 ylidenes, charge transfer systems 954 Zener breakdown, active matrix displays 230 zigzag defects, ferroelectric devices 634 zinc, ligands 914,924
D. Demus, J. Goodby, G. W. Gray, H.-W. Spiess, V. Vill
Handbook of Liquid Crystals
8WILEY-VCH
Handbook of Liquid Crystals D. Demus, J. Goodby, G. W. Gray, H.-W. Spiess, V. Vill
VOl. 1: Fundamentals Vol. 2 A: Low Molecular Weight Liquid Crystals I Vol. 2 B: Low Molecular Weight Liquid Crystals I1 Vol. 3: High Molecular Weight Liquid Crystals
Further title of interest:
J. L. Serrano: Metallomesogens ISBN 3-527-29296-9
D. Demus, J. Goodby, G. W. Gray, H.-W. Spiess, V. Vill
Handbook of Liquid Crystals VOl. 3: High Molecular Weight Liquid Crystals
6BWILEY-VCH Weinheim New York Chichester Brisbane Singapore Toronto
Prof. Dietrich Demus Veilchenweg 23 061 18 Halle Germany Prof. John W. Goodby School of Chemistry University of Hull Hull, HU6 7RX U. K. Prof. George W. Gray Merck Ltd. Liquid Crystals Merck House Poole BH15 1TD U.K.
Prof. Hans-Wolfgang Spiess Max-Planck-Institut fur Polymerforschung Ackermannweg 10 55 128 Mainz Germany Dr. Volkmar Vill Institut fur Organische Chemie Universitat Hamburg Martin-Luther-King-Platz 6 20146 Hamburg Germany
carefully produced. Nevertheless, authors, editors and publisher do not warrant the information contained therein to be free of errors. Readers are advised to keep in mind that statements, data, illustrations, procedural details or other items may inadvertently be inaccurate.
Library of Congress Card No. applied for A catalogue record for this book is available from the British Library. Deutsche Bibliothek Cataloguing-in-Publication Data:
Handbook of liquid crystals / D. Demus ... - Weinheim ; New York ; Chichester ; Brisbane ; Singapore ; Toronto : Wiley-VCH ISBN 3-527-29502-X Vol. 3. High molecular weight liquid crystals. - 1998 ISBN 3-527-29272-1
0WILEY-VCH Verlag GmbH. D-60469 Weinheim (Federal Republic of Germany), 1998 Printed on acid-free and chlorine-free paper. All rights reserved (including those of translation in other languages). No part of this book may be reproduced in any form - by photoprinting, microfilm, or any other means - nor transmitted or translated into machine language without written permission from the publishers. Registered names, trademarks, etc. used in this book, even when not specifically marked as such, are not to be considered unprotected by law. Composition and Printing: Fa. Konrad Triltsch Druck- und Verlagsanstalt GmbH, D-97070 Wurzburg. Bookbinding: Wilhelm Osswald & Co., D-67433 Neustadt Printed in the Federal Republic of Germany.
The Editors D. Demus studied chemistry at the Martin-Luther-University, Halle, Germany, where he was also awarded his Ph. D. In 1981 he became Professor, and in 1991 Deputy Vice-chancellor of Halle University. From 1992-1994 he worked as a Special Technical Advisor for the Chisso Petrochemical Corporation in Japan. Throughout the period 1984-199 1 he was a member of the International Planning and Steering Commitee of the International Liquid Crystal Conferences, and was a non-executive director of the International Liquid Crystal Society. Since 1994 he is active as an Scientific Consultant in Halle. He has published over 310 scientific papers and 7 books and he holds 170 patients.
J. W. Goodby studied for his Ph. D. in chemistry under the guidance of G. W. Gray at the University of Hull, UK. After his post-doctoral research he became supervisor of the Liquid Crystal Device Materials Research Group at AT&T Bell Laboratories. In 1988 he returned to the UK to become the Thorn-EMI/STC Reader in Industrial Chemistry and in 1990 he was appointed Professor of Organic Chemistry and Head of the Liquid Crystal Group at the University of Hull. In 1996 he was the first winner of the G. W. Gray Medal of the British Liquid Crystal Society.
G. W. Gray studied chemistry at the University of Glasgow, UK, and received his Ph. D. from the University of London before moving to the University of Hull. His contributions have been recognised by many awards and distinctions, including the Leverhulme Gold Medal of the Royal Society (1987), Commander of the Most Excellent Order of the British Empire (l991), and Gold Medallist and Kyoto Prize Laureate in Advanced Technology (1995). His work on structure/property relationships has had far reaching influences on the understanding of liquid crystals and on their commercial applications in the field of electro-optical displays. In 1990 he became Research Coordinator for Merck (UK) Ltd, the company which, as BDH Ltd, did so much to commercialise and market the electro-optic materials which he invented at Hull University. He is now active as a Consultant, as Editor of the journal “Liquid Crystals” and as authodeditor for a number of texts on Liquid Crystals.
VI
The Editors
H. W. Spiess studied chemistry at the University of FrankfurVMain, Germany, and obtained his Ph. D. in physical chemistry for work on transition metal complexes in the group of H. Hartmann. After professorships at the University of Mainz, Munster and Bayreuth he was appointed Director of the newly founded Max-Planck-Institute for Polymer Research in Mainz in 1984. His main research focuses on the structure and dynamics of synthetic polymers and liquid crystalline polymers by advanced NMR and other spectroscopic techniques.
V. Vill studied chemistry and physics at the University of Munster, Germany, and acquired his Ph. D. in carbohydrate chemistry in the gorup of J. Thiem in 1990. He is currently appointed at the University of Hamburg, where he focuses his research on the synthesis of chiral liquid crystals from carbohydrates and the phase behavior of glycolipids. He is the founder of the LiqCryst database and author of the LandoltBornstein series Liquid Crystals.
List of Contributors Volume 3, Polymeric Liquid Crystals and Lyotropic Crystals
Blunk, D.; Praefcke, K. (VI) Technische Universitat Berlin Institut fur Organische Chemie StraBe des 17. Juni 135 10623 Berlin Germany
Finkelmann, H. (V) Albert-Ludwig-Universitat Institut fur Makromolekulare Chemie Stefan-Meier-StraBe 3 1 79104 Freiburg Germany
Brand, H. R. (V) Theoretische Physik I11 Universitat Bayreuth 95440 Bayreuth Germany
Gray, J. (VII) Division of Chemical Sciences Science Research Institute University of Salford Salford, M5 4WT U.K.
Chiellini, E. (I:2) Diparitmento di Chimica e Chimica Industiale Universith di Pisa via Risorgimento 35 56126 Pisa Italy
Greiner, A. (I:1) Philipps-Universitat Marburg Fachbereich Physikalische Chemie 35032 Marburg Germany
Dubois, J. C.; Le Barny, P. (IV) Thomson CSF Laboratoire Central de Recherche Domaine de Corbeville 9 1404 Orsay Cedex France Fairhurst, C.; Holmes, M. C. (VII) Department of Physics and Astronomy University of Central Lancashire Preston, PRl 2HE U.K.
Hoffmann, S. (VIII) Martin-Luther-Universitat Halle-Wittenberg FB Biochemie/Biotechnologie Kurt-Mothes-StraBe 3 06120 Halle (Saale) Germany Laus, M. (I:2) Diparitmento di Chimica Industriale e dei Materiali UniversitB di Bologna via Risorgimento 4 40137 Bologna Italy
VIII
List of Contributors
Mauzac, M. (IV) CRPP-CNRS Avenue du Dr. Schweitzer 33600 Pessac France
Tiddy, G. J. T.; Fuller, S. (VII) Department of Chemical Engineering Institute of Science and Technology University of Manchester Manchester, M60 1QD U.K.
Noel, C. (I1 and IV) Laboratoire de Physico-Chimie Structurale de Macromoleculaire Laboratoire Associk au CNRS 10, Rue Vauquelin 75231 Paris France
Vill, V. (VI) Universitat Hamburg Institut fur Organische Chemie Martin-Luther-King-Platz 6 20 146 Hamburg Germany
Ober, C. K.; Mao, G. (I:4) Cornell University Dept. of Materials Science and Engineering 322 Bard Hall Ithaca, NY 14853-1501 USA Pugh, C.; Kiste, A. (111) The University of Michigan Dept. of Chemistry Ann Arbor, MI 4809- 1055 USA Schmidt, H.-W. (1:l) Universitat Bayreuth Makromolekulare Chemie I UniversitatsstraSe 30 NW I1 95440 Bayreuth Germany
Zentel, R. (I:3) BUGH Wuppertal Fachbereich Chemie GauSstraBe 20 42097 Wuppertal Germany Zugenmaier, P. (IX) Institut f. Physikalische Chemie Technische Universitat Clausthal Arnold-Sommerfeld-S traBe 4 38678 Clausthal-Zellerfeld Germany
Outline Volume 1
Chapter 1: Introduction and Historical Development . . . . . . . . . . . . . . . . . . . 1 George W Gray Chapter 11: Guide to the Nomenclature and Classification of Liquid Crystals . . . . . 17 John W. Goodby and George W. Gray Chapter 111: Theory of the Liquid Crystalline State . . . . . . . . . . . . . . . . . . . 2.5 1 25 Continuum Theory for Liquid Crystals . . . . . . . . . . . . . . . . . . . . Frank M. Leslie Molecular Theories of Liquid Crystals . . . . . . . . . . . . . . . . . . . . 40 2 M. A. Osipov 3 Molecular Modelling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72 Mark R. Wilson Chapter IV: General Synthetic Strategies . . . . . . . . . . . . . . . . . . . . . . . . Thies Thiemann and Volkmar Vill
87
Chapter V: Symmetry and Chirality in Liquid Crystals. . . . . . . . . . . . . . . . . 1 1.5 John W Goodby Chapter VI: Chemical Structure and Mesogenic Properties . . . . . . . . . . . . . . 133 Dietrich Demus Chapter VII: Physical Properties . . . . . . . . . . . . . . Tensor Properties of Anisotropic Materials . . 1 David Dunmur and Kazuhisa Toriyama Magnetic Properties of Liquid Crystals . . . . 2 David Dunmur and Kazuhisa Torijiama Optical Properties . . . . . . . . . . . . . . . . 3 David Dunmur and Kazuhisa Toriyama 4 Dielectric Properties . . . . . . . . . . . . . . . David Dunmur and Kazuhisa Toriyamn 5 Elastic Properties. . . . . . . . . . . . . . . . . David Dunmur and Kazuhisa Torigamn 6 Phase Transitions. . . . . . . . . . . . . . . . .
.............. 189 . . . . . . . . . . . . . . . 189 . . . . . . . . . . . . . . . 204
. . . . . . . . . . . . . .
2 1.5
..............
23 1
..............
253
. . . . . . . . . . . . . .
28 1
X
Outline
6.1 6.2 6.2.1 6.2.2 6.2.3 6.2.4 6.3 6.4 7 8 9 10 11
12 13
Phase Transitions Theories . . . . . . . . . . . . . . . . . . . . . . . . . . 281 Philippe Barois Experimental Methods and Typical Results . . . . . . . . . . . . . . . . . 310 Thermal Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 310 Jan Thoen Density . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 334 Wo2fgang Wedler 350 Metabolemeter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Wovgang Wedler 355 High Pressure Investigations . . . . . . . . . . . . . . . . . . . . . . . . . P. Pollmann Fluctuations and Liquid Crystal Phase Transitions . . . . . . . . . . . . . 379 P. E . Cladis Re-entrant Phase Transitions in Liquid Crystals . . . . . . . . . . . . . . . 391 P. E . Cladis 406 Defects and Textures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . I.: Bouligand Flow Phenomena and Viscosity . . . . . . . . . . . . . . . . . . . . . . . 454 Frank Schneider and Herbert Kneppe Behavior of Liquid Crystals in Electric and Magnetic Fields . . . . . . . . 477 Lev M . Blinov Surface Alignment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 535 Blandine Je'rBme 549 Ultrasonic Properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Olga A . Kapustina Nonlinear Optical Properties of Liquid Crystals . . . . . . . . . . . . . . . 569 P. Palffy-Muhoray 582 Diffusion in Liquid Crystals . . . . . . . . . . . . . . . . . . . . . . . . . F: Noack
595 Chapter VIII: Characterization Methods . . . . . . . . . . . . . . . . . . . . . . . . 1 Magnetic Resonance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 595 Claudia Schmidt and Hans Wolfgang Spiess 2 X-Ray Characterization of Liquid Crystals: Instrumentation . . . . . . . . 619 Richard H . Templer Structural Studies of Liquid Crystals by X-ray Diffraction . . . . . . . . . 635 3 John M . Seddon 680 4 Neutron Scattering . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Robert M . Richardson Light Scattering from Liquid Crystals . . . . . . . . . . . . . . . . . . . . 699 5 Helen E Gleeson Brillouin Scattering from Liquid Crystals . . . . . . . . . . . . . . . . . . 719 6 Helen E Gleeson Mossbauer Studies of Liquid Crystals . . . . . . . . . . . . . . . . . . . . 727 7 Helen E Gleeson
Outline
XI
Chapter IX: Applications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 731 1 Displays . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 731 Ian C. Sage 2 Nondisplay Applications of Liquid Crystals . . . . . . . . . . . . . . . . . 763 William A. Crossland and Timothy D . Wilkinson Thermography Using Liquid Crystals . . . . . . . . . . . . . . . . . . . . 3 823 Helen F: Gleeson 4 Liquid Crystals as Solvents for Spectroscopic, Chemical Reaction. and Gas Chromatographic Applications . . . . . . . . . . . . . 839 William J . Leigh and Mark S. Workentin Index Volume 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
897
Volume 2A Part I: Calamitic Liquid Crystals . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
1
Chapter I: Phase Structures of Calamitic Liquid Crystals . . . . . . . . . . . . . . . . . 3 John W. Goodby Chapter 11: Phase Transitions in Rod-Like Liquid Crystals . . . . . . . . . . . . . . . 23 Daniel Guillon Chapter 111: Nematic Liquid Crystals . . . . . . . . . . . . . . . . . . . . . . . . . . 47 47 1 Synthesis of Nematic Liquid Crystals . . . . . . . . . . . . . . . . . . . . . Kenneth J. Toyne 2 Physical Properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60 Elastic Properties of Nematic Liquid Crystals . . . . . . . . . . . . . . . . 60 2.1 Ralf Stannarius Dielectric Properties of Nematic Liquid Crystals . . . . . . . . . . . . . . . 91 2.2 Horst Kresse Diamagnetic Properties of Nematic Liquid Crystals . . . . . . . . . . . . . 113 2.3 Ralf Stannarius 2.4 Optical Properties of Nematic Liquid Crystals . . . . . . . . . . . . . . . 128 Gerhard Pelzl 142 2.5 Viscosity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Herbert Kneppe and Frank Schneider 2.6 Dynamic Properties of Nematic Liquid Crystals . . . . . . . . . . . . . . 170 R. Blinc and I . MuSevic' 3 Applications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 199 3.1 199 TN, STN Displays . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Harald Hirschmann and VolkerReiffenrath Active Matrix Addressed Displays . . . . . . . . . . . . . . . . . . . . . . 230 3.2 Eiji Kaneko
XI1
3.3 3.4
Outline
Dynamic Scattering . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 243 Birendra Bahadur Guest-Host Effect . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 257 Birendra Bahadur
Chapter IV: Chiral Nematic Liquid Crystals . . . . . . . . . . . . . . . . . . . . . . 303 1 The Synthesis of Chiral Nematic Liquid Crystals . . . . . . . . . . . . . . 303 Christopher J. Booth 2 Chiral Nematics: Physical Properties and Applications . . . . . . . . . . . 335 Harry Coles Chapter V: Non-Chiral Smectic Liquid Crystals . . . . . . . . . . . . . . . . 1 Synthesis of Non-Chiral Smectic Liquid Crystals . . . . . . . . . . John W Goodby Physical Properties of Non-Chiral Smectic Liquid Crystals. . . . . 2 C. C. Huang 3 Nonchiral Smectic Liquid Crystals - Applications . . . . . . . . . David Coates
. . . . 41 1 . . . . 41 1 . . . . 441 . . .
. 470
Volume 2B Part 2: Discotic Liquid Crystals. . . . . . . . . . . . . . . . Chapter VI: Chiral Smectic Liquid Crystals . . . . . . 1 Synthesis of Chiral Smectic Liquid Crystals. Stephen M. Kelly 2 Ferroelectric Liquid Crystals. . . . . . . . . Sven is Lagerwall 3 Antiferroelectric Liquid Crystals . . . . . . Kouichi Miyachi and Atsuo Fukuda
.....,.......
491
. . . . . . . . . . . . . . . . 493 . . . . . . . . . . . . . . . . 493 . . . . . . . . . . . . . . . . 515 . . .
.............
665
Chapter VII: Synthesis and Structural Features. . . . . . . . . . . . . . . . . . . . . 693 Andrew N. Carnmidge and Richard J. Bushby Chapter VIII: Discotic Liquid Crystals: Their Structures and Physical Properties . . . 749 S. Chandrasekhar Chapter IX: Applicable Properties of Columnar Discotic Liquid Crystals . Neville Boden and Bijou Movaghar
......
781
Outline
XI11
Part 3: Non-Conventional Liquid-Crystalline Materials . . . . . . . . . . . . . . . . 799 Chapter X: Liquid Crystal Dimers and Oligomers . . . . . . . . . . . . . . . . . . . 801 Corrie T Imrie and GeofSrey R. Luckhurst Chapter XI: Laterally Substituted and Swallow-Tailed Liquid Crystals . . . . . . . . 835 Wolfgang Weissflog Chapter XII: Phasmids and Polycatenar Mesogens . . . . . . . . . . . . . . . . . . . 865 Huu-Tinh Nguyen, Christian Destrade, and Jucques Malthzte Chapter Xlll: Thermotropic Cubic Phases . . . . . . . . . . . . . . . . . . . . . . . 887 Siegmar Diele and Petra Goring Chapter XIV: Metal-containing Liquid Crystals . . . . . . . . . . . . . . . . . . . . 90 1 Anne Marie Giroud-Godquin Chapter XV: Biaxial Nematic Liquid Crystals . . . . . . . . . . . . . . . . . . . . . 933 B. K. Sudashiva Chapter XVI: Charge-Transfer Systems . . . . . . . . . . Klaus Praefcke and D. Singer
,
. . . .
. .
. .
. . . . . 945
Chapter XVII: Hydrogen-Bonded Systems . . . . . . . . . . . . . . . . . . . . . . . 969 Takashi Kato Chapter XVIII: Chromonics. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98 1 John Lydon Index Volumes 2A and 2B. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1009
Volume 3 Part 1: Main-Chain Thermotropic Liquid-Crystalline Polymers . . . . . . . . . . . . . 1 Chapter I: Synthesis, Structure and Properties . . . . . . . . . . . . . . . . . . . . . . 3 Aromatic Main Chain Liquid Crystalline Polymers . . . . . . . . . . . . . . 3 1 Andreas Greiner and Hans- Werner Schmidt Main Chain Liquid Crystalline Semiflexible Polymers . . . . . . . . . . . . 26 2 Emo Chiellini and Michele Laus Combined Liquid Crystalline Main-Chain/Side-Chain Polymers. . . , . . . 52 3 Rudolf Zentel Block Copolymers Containing Liquid Crystalline Segments. . . . . . . . . 66 4 Guoping Ma0 and Christopher K. Ober
XIV
Outline
Chapter 11: Defects and Textures in Nematic Main-Chain Liquid Crystalline Polymers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Claudine Noel
93
Part 2: Side-Group Thermotropic Liquid-Crystalline Polymers . . . . . . . . . . . . 121 Chapter 111: Molecular Engineering of Side Chain Liquid Crystalline Polymers by Living Polymerizations . . . . . . . . . . . . . . . . . . . . . . . . . . Coleen Pugh and Alan L. Kiste
123
Chapter IV: Behavior and Properties of Side Group Thermotropic Liquid Crystal Polymers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Jean-Claude Dubois, Pierre Le Barny, Monique Mauzac, and Claudine Noel
207
Chapter V: PhysicalProperties of LiquidCrystallineElastomers . . . . . . . . . . . 277 Helmut R. Brand and Heino Finkelmann
Part 3: Amphiphilic Liquid Crystals . . . . . . . . . . . . . . . . . . . . . . . . . .
303
Chapter VI: Amphotropic Liquid Crystals . . . . . . . . . . . . . . . . . . . . . . . Dieter Blunk, Klaus Praefcke and Volkmar Vill
305
Chapter VII: Lyotropic Surfactant Liquid Crystals . . . . . . . . . . . . . . . . . . . 34 1 C. Fairhurst, S. Fuller, J. Gray, M. C. Holmes, G. J. 7: Tiddy Chapter VIII: Living Systems. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Siegfried Hoffmann
393
Chapter IX: Cellulosic Liquid Crystals . . . . . . . . . . . . . . . . . . . . . . . . . Peter Zugenmaier
45 1
Index Volumes 1 - 3. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
483
Contents
Part I: Main-Chain Thermotropic Liquid-Crystalline Polymers . . .
1
Chapter I: Synthesis. Structure and Properties . . . . . . . . . . . . . . . . . . . .
3
Aromatic Main Chain Liquid Crystalline Polymers . . . . . . . . . . . . . . . 3 Andreas Greiner and Hans- Werner Schmidt Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 1.1 1.2 Structural Modification Concepts of Liquid Crystal Polymers (LCPs) . . . . . 4 Aromatic LC (Liquid Crystal) Polyesters . . . . . . . . . . . . . . . . . . . . 1.3 7 1.3.1 Polyesters with Moieties of Different Lengths . . . . . . . . . . . . . . . . . 10 1.3.2 Polyesters with Kinked or Double Kinked Moieties . . . . . . . . . . . . . . 10 13 1.3.3 Polyesters with Crankshaft Moieties . . . . . . . . . . . . . . . . . . . . . . 15 1.3.4 Polyesters with Lateral Substituents . . . . . . . . . . . . . . . . . . . . . . 16 1.3.4.1 Polymers with Flexible Substitutents . . . . . . . . . . . . . . . . . . . . . . 19 1.3.4.2 Polymers with Stiff. Bulky Substituents . . . . . . . . . . . . . . . . . . . . Polyesters with Noncoplanar Aromatic Moieties . . . . . . . . . . . . . . . . 20 1.3.5 Para-Linked Aromatic Polyamides . . . . . . . . . . . . . . . . . . . . . . . 22 1.4 1.5 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
I
Main Chain Liquid Crystalline Semiflexible Polymers . . . . . . . . . . . . . 26 Emo Chiellini and Michele Laus Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26 2.1 2.2 Preparation of Main Chain Liquid Crystalline Polymers (MCLCPs) . . . . . . 27 2.2.1 Generalities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27 28 2.2.2 Polyesters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Polyesters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32 2.2.3 2.2.4 Polyurethanes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33 2.2.5 Polyamides . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36 2.2.6 Polysiloxanes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36 2.2.7 Poly(p-aminoester)s and Poly(b-thioester)s . . . . . . . . . . . . . . . . . . . 38 2.3 Structure-Liquid Crystalline Property Correlations . . . . . . . . . . . . . . 39 40 2.3.1 Mesogenic Group Effects . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.3.2 Flexible Spacer Effects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42 42 2.3.2.1 Polymethylene Spacers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.3.2.3 Polysiloxane Spacers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45 2.3.3 Spacer Substituent Effects . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46 2.3.4 Copolymerization Effects . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46
2
XVI 2.3.5 2.4
Contents
Molar Mass and Molar Mass Distribution Effects . . . . . . . . . . . . . . . 47 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49
Combined Liquid Crystalline Main-ChainlSide-Chain Polymers . . . . . . . Rudolf Zentel Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.1 Molecular Structure of Combined LC Polymers . . . . . . . . . . . . . . . 3.2 3.2.1 Polymers with Side-Chain Mesogens Linked at the Main Chain Spacer . . . 3.2.1.1 Achiral Combined LC Polymers . . . . . . . . . . . . . . . . . . . . . . . . 3.2.1.2 Chiral Combined LC Polymers . . . . . . . . . . . . . . . . . . . . . . . . . 3.2.1.3 Cross-Linked LC Elastomers . . . . . . . . . . . . . . . . . . . . . . . . . . Polymers with Side-Chain Mesogens Linked at the Main Chain Mesogen . . 3.2.2 3.3 Properties of Combined LC Polymers . . . . . . . . . . . . . . . . . . . . . 3.3.1 Structure-Property Relationships and Types of LC Phase . . . . . . . . . . 3.3.2 Interaction of Main-Chain and Side-Chain Mesogens . . . . . . . . . . . . 3.3.3 Rheology and LC Elastomers . . . . . . . . . . . . . . . . . . . . . . . . . . 3.4 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
4 4.1 4.1.1 4.1.2 4.1.3 4.1.4 4.2 4.2.1 4.2.2 4.2.3 4.2.4 4.3 4.3.1 4.3.2 4.3.2.1 4.3.2.2 4.3.2.3 4.3.2.4 4.4 4.4.1 4.4.2 4.5 4.6
. 52 . .
. . .
52 53 53 53 55 56 57 58 58 60 62 64
Block Copolymers Containing Liquid Crystalline Segments . . . . . . . . . . 66 Guoping Muo and Christopher K. Ober Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66 Block Copolymers: a Brief Review . . . . . . . . . . . . . . . . . . . . . . . 66 Liquid Crystalline Polymers: Architecture . . . . . . . . . . . . . . . . . . . 67 Architecture of Liquid Crystalline Block Copolymers . . . . . . . . . . . . . 67 General Features of Liquid Crystalline Block Copolymers . . . . . . . . . . . 68 70 Rod-Coil Diblock Copolymer Systems . . . . . . . . . . . . . . . . . . . . . Polypeptides as the Rod Block . . . . . . . . . . . . . . . . . . . . . . . . . 70 Rod-Coil Block Copolymers with Short Rod Blocks . . . . . . . . . . . . . . 72 Rod-Coil Block Copolymers Based on Polyisocyanates . . . . . . . . . . . . 74 76 Discussion of Rod-Coil Systems . . . . . . . . . . . . . . . . . . . . . . . . Side Group Liquid Crystal - Coil Diblock Copolymer Systems . . . . . . . . 78 Synthesis and Characterization of the Side Group Liquid Crystal 78 Coil Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Properties of Side Group Liquid Crystal - Coil Systems . . . . . . . . . . . . 82 Liquid Crystal Properties in Side Group Liquid Crystal - Coil Systems . . . . 82 Phase Diagram for Side Group Liquid Crystal - Coil Systems . . . . . . . . . 83 86 Interface Thickness . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Interplay of Liquid Crystallinity and Macrophase Separation . . . . . . . . . 86 Applications of Liquid Crystal - Block Copolymers . . . . . . . . . . . . . . 87 Microphase Stabilized Ferroelectric Liquid Crystal Displays . . . . . . . . . 88 Self-healing, Stable, Low Surface Energy Materials Based on Liquid Crystal - Block opolymers . . . . . . . . . . . . . . . . . . 89 Future Work . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90
Contents
XVII
Chapter 11: Defects and Textures in Nematic Main-Chain Liquid Crystalline Polymers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Claudine Noel
93
1
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
93
2 2.1 2.2 2.3 2.3.1 2.3.2
94 Textural Assignments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94 The Basic Equations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Some Consequences of Elastic Anisotropy . . . . . . . . . . . . . . . . . . 97 Topology of Director Fields: Homotopy Groups and Classification of Defects 99 99 Uniaxial Nematics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101 Biaxial Nematics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
3
Defects in Nematic Main-Chain Liquid Crystalline Polymers . . . . . . . . . 102
4
Biaxial Nematic Main-Chain Liquid Crystalline Polymers . . . . . . . . . . 104
5
Defect Associations and Textures at Rest . . . . . . . . . . . . . . . . . . . 105
6
Flow-induced Textures and their Relaxation Behavior . . . . . . . . . . . . 114
7
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
118
Part 11: Side-Group Thermotropic Liquid-Crystalline Polymers . . I 2 1 Chapter 111: Molecular Engineering of Side Chain Liquid Crystalline Polymers by Living Polymerizations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Coleen Pugh and Alan L. Kiste
123
1 1.1 1.2
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Chain Polymerizations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Living Polymerizations . . . . . . . . . . . . . . . . . . . . . . . . . . . .
2 2.1 2.2 2.3 2.4 2.5
‘Living’ Polymerizations used to Synthesize Side Chain Liquid Crystalline Polymers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 127 Anionic and Group Transfer Polymerizations of Olefins . . . . . . . . . . . 127 Polymerizations with Metalloporphyrins . . . . . . . . . . . . . . . . . . . 133 136 Cationic Polymerizations of Olefins . . . . . . . . . . . . . . . . . . . . . . Ring-Opening Metathesis Polymerizations . . . . . . . . . . . . . . . . . . 142 Polymer Analogous Reactions on Well-Defined Precursor Polymers . . . . . 149
3 3.1 3.2 3.3 3.4 3.5 3.6
Structure/Property Correlations Determined using ‘Living’ Polymerizations The Effect of Molecular Weight . . . . . . . . . . . . . . . . . . . . . . . . The Effect of the Mesogen . . . . . . . . . . . . . . . . . . . . . . . . . . . The Effect of the Spacer . . . . . . . . . . . . . . . . . . . . . . . . . . . . The Effect of the Nature of the Polymer Backbone . . . . . . . . . . . . . . The Effect of Tacticity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The Effect of Polydispersity . . . . . . . . . . . . . . . . . . . . . . . . . .
123 123 125
1 52 152 1.56 159 163 167 169
XVIII
4 4.1 4.1.1 4.1.2 4.1.3 4.1.4 4.1.5 4.2 4.3
Contents
Chain Copolymerizations . . . . . . . . . . . . . . . . . . . . . . . . . . . 171 Block Copolymers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 172 Anionic and Group Transfer Copolymerizations . . . . . . . . . . . . . . . 174 Cationic Copolymerizations . . . . . . . . . . . . . . . . . . . . . . . . . . 177 Copolymerizations with Metalloporphyrins . . . . . . . . . . . . . . . . . . 181 Ring-Opening Metathesis Copolymerizations . . . . . . . . . . . . . . . . . 181 Morphology and Thermotropic Behavior of Side-Chain Liquid Crystalline Block Copolymers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 183 Graft Copolymers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 190 Statistical Binary Copolymers . . . . . . . . . . . . . . . . . . . . . . . . . 192 Other Factors Controlling the Thermotropic Behavior of SCLCPs as Studied using Living Polymerizations: Induction of Smectic Layering using Immiscible Components . . . . . . . . . . . . . . . . . . . . . . . . . . . . 197 The Future . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
197
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
199
Chapter I V Behavior and Properties of Side Group Thermotropic Liquid Crystal Polymers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Jean-Claude Dubois. Pierre Le Barny. Monique Mauzac. and Claudine Noel
207
1
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
207
2 2.1 2.1.1 2.1.2 2.1.3 2.1.4 2.2 2.2.1 2.2.2 2.2.3 2.2.4 2.2.5 2.3 2.3.1 2.3.2 2.3.3
Ferroelectric Liquid Crystal Polymers’ Behavior . . . . . . . . . . . . . . . Chemical Structures of SmC* Liquid Crystal Polymers . . . . . . . . . . . . Homopolymers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Copolymers and Terpolymers . . . . . . . . . . . . . . . . . . . . . . . . . Oligomers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Combined Polymers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Phase Behavior . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Means of Investigation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Chemical Structure-Phase Behavior Relationships . . . . . . . . . . . . . . Influence of the Molecular Weight . . . . . . . . . . . . . . . . . . . . . . . Influence of the Dilution of the Mesogenic Groups . . . . . . . . . . . . . . Occurrence of Unusual Mesophases in Chiral Side Chain Polymers . . . . . Ferroelectric Properties of SmC* Polymers . . . . . . . . . . . . . . . . . . Uniform Alignment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Experimental Methods for Measuring the Spontaneous Polarization . . . . . Spontaneous Polarization Behavior of FLCPs (Ferroelectric Liquid Crystal Polymers) . . . . . . . . . . . . . . . . . . . . Tilt Angle and Spontaneous Polarization . . . . . . . . . . . . . . . . . . . Electrooptic Behavior . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Ferroelectric Switching . . . . . . . . . . . . . . . . . . . . . . . . . . . . Electroclinic Switching . . . . . . . . . . . . . . . . . . . . . . . . . . . .
208 208 208 208 212 212 212 212 213 213 214 215 217 218 218
7
2.3.4 2.4 2.4.1 2.4.2
218 222 223 223 225
Contents
XIX
2.4.3 2.4.4 2.5 2.5.1 2.5.2 2.5.3
Antiferroelectric Switching . . . . . . . . . . . . . . . . . . . . . . . . . . Broadband Dielectric Spectroscopy . . . . . . . . . . . . . . . . . . . . . . Potential Applications of FLCPs . . . . . . . . . . . . . . . . . . . . . . . . Nonlinear Optics (NLO) . . . . . . . . . . . . . . . . . . . . . . . . . . . . Pyroelectric Detectors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Display Devices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
3 3.1 3.1.1 3.1.2 3.2 3.2.1 3.2.2 3.3 3.3.1 3.3.2 3.3.3 3.3.4
Side Chain Liquid Crystalline Networks and Mechanical Properties . . . . . 229 Theoretical Approaches . . . . . . . . . . . . . . . . . . . . . . . . . . . . 229 Landau-de Gennes Description of Nematic Elastomers . . . . . . . . . . . . 229 Models for Nematic Rubber Elasticity . . . . . . . . . . . . . . . . . . . . . 230 231 Characterization of Networks . . . . . . . . . . . . . . . . . . . . . . . . . General Features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 231 231 Effective Crosslinking Density . . . . . . . . . . . . . . . . . . . . . . . . Mechanical Field Effects on Liquid-Crystalline Networks . . . . . . . . . . 234 Response of the Network as a Function of the Stress . . . . . . . . . . . . . 235 Thermal Evaluation of the Elastic Modulus . . . . . . . . . . . . . . . . . . 236 Mechanical Orientability of the Mesogens . . . . . . . . . . . . . . . . . . 237 240 Electromechanical Properties . . . . . . . . . . . . . . . . . . . . . . . . .
226 227 227 227 227 228
Nonlinear Optical Properties . . . . . . . . . . . . . . . . . . . . . . . . . . 242 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 242 Theoretical Approach . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 242 Second Harmonic Generation (SHG) . . . . . . . . . . . . . . . . . . . . . 242 Theoretical Models for the Electric Field Poling . . . . . . . . . . . . . . . 244 The Electrooptic Effect . . . . . . . . . . . . . . . . . . . . . . . . . . . . 245 Liquid Crystalline Systems . . . . . . . . . . . . . . . . . . . . . . . . . . 246 Molecular Design for High NLO Activity . . . . . . . . . . . . . . . . . . . 246 247 NLO Liquid Crystalline Systems . . . . . . . . . . . . . . . . . . . . . . . Guest-Host Materials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 247 LC Copolymers Containing Both Nematogenic (or Smectogenic) and Active Side Groups . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 249 4.3.2.3 SCLCPs Wherein the NLO Active Possess Mesogenic Properties Themselves . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 254 265 4.3.2.4 Ferroelectric LC Polymeric Systems. . . . . . . . . . . . . . . . . . . . . . 4.3.2.5 Rigid Rod-Like Polymers . . . . . . . . . . . . . . . . . . . . . . . . . . . 267 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 References 270
4 4.1 4.2 4.2.1 4.2.2 4.2.3 4.3 4.3.1 4.3.2 4.3.2.1 4.3.2.2
Chapter V. Physical Properties of Liquid Crystalline Elastomers . . . . . . . . . 277 Helmut R. Brand and Heino Finkelmann 1
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
2 2.1
Phase Transitions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 277 The Nematic-Isotropic Phase Transition . . . . . . . . . . . . . . . . . . . 277
277
xx
Contents
2.1 .1 2.1 .2 2.1 .3 2.1 .4 2.1 .5 2 .I .6 2.1.7 2.1.8 2.2 2.3 2.4 2.5
Mechanical Properties of Polydomains . . . . . . . . . . . . . . . . . . . . Mechanical Properties of Monodomains: Liquid Single Crystal Elastomers . Optical Properties of Polydomains . . . . . . . . . . . . . . . . . . . . . . . Optical Properties of Monodomains: Liquid Single Crystal Elastomers . . . Electric and Dielectric Properties . . . . . . . . . . . . . . . . . . . . . . . NMR Investigations of Monodomains . . . . . . . . . . . . . . . . . . . . . Models for the Nematic-Isotropic Transition for Poly- and Monodomains . X-ray Investigations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The Cholesteric-Isotropic Phase Transition . . . . . . . . . . . . . . . . . . The Smectic A-Isotropic Phase Transition . . . . . . . . . . . . . . . . . . The Discotic-Isotropic Phase Transition . . . . . . . . . . . . . . . . . . . Other PhaseTransitions in LiquidCrystallineElastomers . . . . . . . . . .
277 279 281 284 285 285 287 288 288 289 289 290
3 3.1 3.1.1 3.1.2 3.1.3 3.1.4 3.1.5 3.1.6 3.1.7 3.1.8 3.1.9 3.2 3.2.1 3.2.2 3.3 3.4 3.4.1 3.4.2 3.4.3 3.5
Macroscopic Properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . Nematic Phase . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Mechanical Properties of Polydomains . . . . . . . . . . . . . . . . . . . . Mechanical Properties of Monodomains: Liquid Single Crystal Elastomers . Linear and Nonlinear Optical Properties of Polydomains . . . . . . . . . . . Linear andNonlinear OpticalProperties of Monodomains . . . . . . . . . . Effects of Electric and Magnetic Fields . . . . . . . . . . . . . . . . . . . . Diffusion Effects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Macroscopic Equations . . . . . . . . . . . . . . . . . . . . . . . . . . . . Rubber Elasticity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . X-ray Investigations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Cholesteric Phase . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Mechanical Properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Electromechanical Properties . . . . . . . . . . . . . . . . . . . . . . . . . Smectic A and Smectic C Phases . . . . . . . . . . . . . . . . . . . . . . . Smectic C* Phase . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Mechanical Properties of Polydomains . . . . . . . . . . . . . . . . . . . . Nonlinear Optical Properties . . . . . . . . . . . . . . . . . . . . . . . . . . Electromechanical Properties of Polydomains . . . . . . . . . . . . . . . . . Discotic Phases and Hexagonal Lyotropic Phases . . . . . . . . . . . . . . .
291 291 291 292 293 293 294 294 295 295 295 295 295 296 297 298 298 299 300 300
4
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
301
Part 111: Amphiphilic Liquid-Crystals . . . . . . . . . . . . . . . . . . 303 Chapter VI: Amphotropic Liquid Crystals . . . . . . . . . . . . . . . . . . . . . Dieter Blunk. Klaus Praefcke and Volkrnar Vill
305
1
Introduction. Remarks on History . . . . . . . . . . . . . . . . . . . . . . .
305
2 2.1
Principles of Molecular Structures . . . . . . . . . . . . . . . . . . . . . . . Simple Amphiphiles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
309 311
Contents
XXI
2.2 2.3 2.4 2.5 2.6 2.7
Bolaamphiphiles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ‘Y’-Shaped Materials . . . . . . . . . . . . . . . . . . . . . . . . . . Disc-shaped Amphiphiles . . . . . . . . . . . . . . . . . . . . . . . . Metallomesogens . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Polyphilic Liquid Crystals . . . . . . . . . . . . . . . . . . . . . . . . Further Molecular Architectures . . . . . . . . . . . . . . . . . . . . .
... ... ... . . . . . . ...
324 327 330 332 334 335
3
General Phase Behavior . . . . . . . . . . . . . . . . . . . . . . . . . . . .
335
4
Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
336
5
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
336
Chapter VII: Lyotropic Surfactant Liquid Crystals. . . . . . . . . . . . . . . . . 341 C. Fairhurst. S. Fuller; J . Gray. M . C. Holmes. G . J . 7: Tiddy 1
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
341
2
Surfactant Solutions: Micelles . . . . . . . . . . . . . . . . . . . . . . . . .
342
3 3.1 3.2 3.3 3.4 3.5 3.6
Liquid Crystal Structures . . . . . . . . . . . . . . . . . . . . . . . . . . . Lamellar Phase (La) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Hexagonal Phases (Hl, H2) . . . . . . . . . . . . . . . . . . . . . . . . . . Cubic Phases . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Nematic Phases . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Gel Phases (Lp) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Intermediate Phases . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
348 349 350 350 353 354 356
4
Phase Behavior of Nonionic Surfactants . . . . . . . . . . . . . . . . . . . .
359
5
Block Copolymer Nonionic Surfactants . . . . . . . . . . . . . . . . . . . .
373
6
Zwitterionic Surfactants . . . . . . . . . . . . . . . . . . . . . . . . . . . .
376
7
Ionic Surfactants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
376
8 8.1 8.2 8.3 8.4 8.5 8.6
The Influence of Third Components . . . . . . . . . . . . . . . . . . . . . . Cosurfactants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Mixed Surfactants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Oils . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Hydrotropes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Electrolytes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Alternative Solvents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
381 381 383 384 384 385 386
9
Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
388
10
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
389
XXII
Contents
Chapter VIII: Living Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Siegfried Hoffmann 1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
393 393
2 2.1 2.2 2.3 2.4 2.4. I 2.4.2 2.4.3 2.5 2.6 3
Biomesogens and the Grand Process . . . . . . . . . . . . . . . . . . . . . 394 Historical Dualities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 395 “Genesis” and Phase Transitions . . . . . . . . . . . . . . . . . . . . . . . . 399 Evolution of Amphiphilic Patterns . . . . . . . . . . . . . . . . . . . . . . . 402 Transition to Life and (Bio)mesogenic Reflections . . . . . . . . . . . . . . 404 Molecular developments . . . . . . . . . . . . . . . . . . . . . . . . . . . . 406 Molecular Matrices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 411 (Bio)mesogenic Order-Disorder Patterns . . . . . . . . . . . . . . . . . . . 418 The Developed Biomesogenic Systems . . . . . . . . . . . . . . . . . . . . 429 The Human Component . . . . . . . . . . . . . . . . . . . . . . . . . . . . 435 Outlook . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
438
4
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
439
Chapter IX: Cellulosic Liquid Crystals . . . . . . . . . . . . . . . . . . . . . . . Peter Zugenmaier
453
1
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
453
2
The Structure of Cellulosics . . . . . . . . . . . . . . . . . . . . . . . . . .
453
3 3.1 3.2
The Chiral Nematic State . . . . . . . . . . . . . . . . . . . . . . . . . . . 455 Methods for the Detection of Cellulosic Mesophases . . . . . . . . . . . . . 455 Models of Chiral Nematic Cellulosics . . . . . . . . . . . . . . . . . . . . . 461
4
Mesophase Formation of Cellulosics . . . . . . . . . . . . . . . . . . . . .
5
Mesophases of Cellulose . . . . . . . . . . . . . . . . . . . . . . . . . . . .
6 6.1 6.2 6.3 6.4 6.5 6.6 6.7 6.8
Lyotropic Liquid-Crystalline Cellulose Derivatives . . . . . . . . . . . . . . 464 Supermolecular Structure and Optical Properties . . . . . . . . . . . . . . . 464 Handedness . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 465 Temperature Dependence of the Pitch . . . . . . . . . . . . . . . . . . . . . 470 Concentration Dependence of the Pitch . . . . . . . . . . . . . . . . . . . . 472 Molecular Mass Dependence of the Pitch . . . . . . . . . . . . . . . . . . . 473 Solvent Effects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 474 Order Parameter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 475 Phase Behavior . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 475
7
Thermotropic Liquid Crystals of Cellulosics . . . . . . . . . . . . . . . . . 477
8
Concluding Remarks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
480
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 Index Volumes 1-3 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
481
462 463
483
General Introduction
Liquid crystals are now well established i n basic research as well as in development for applications and commercial use. Because they represent a state intermediate between ordinary liquids and three-dimensional solids. the investigation of their physical properties is very complex and makes use of many different tools and techniques. Liquid crystals play an important role in materials science, they are model materials for the organic chemist i n order to investigate the connection between chemical structure and physical properties. and they provide insight into certain phenomena of biological systems. Since their main application is in displays, some knowledge of the particulars of display tcchnology is necessary for a complete understanding of the matter. In 1980 VCH published the Halidhook o f Liquid Crystnls, written by H. Kelker and R. HatL. with a contribution by C. Schuniann, which had a total of about 900 pages. Even in 1980 i t was no easy task for this small number of authors to put together the Hcriidhook, which comprised so many specialities; the Haridhook took about 12 years to complete. In the meantime the amount of information about liquid crystals has grown nearly exponentially. This is reflected in the number of known liquid-crystalline compounds: in 1974 about 5000 (D. Demus, H. Deinus, H. Zaschke. Fliissigr~KI-istrrlle i r z 7irheIIer7) and in 1997 about 70000 (V. Vill. electronic data base LIQCRYST). According to ;I recent estimate by V. Vill. thc C I I F
rent number of publications is about 65000 papers and patents. This development shows that, for a single author or a small group of authors, it may be impossible to produce a representative review of all the topics that are relevant to liquid crystals on the one hand because of the necessarily high degree of specialization, and on the other because of the factor of time. Owing to the regrettable early decease of H. Kelker and the poor health of R. Hatz, neither of the former main authors was able to continue their work and to participate in a new edition of the H m d b o o k . Therefore, it was decided to appoint five new editors to be responsible for the structure of thc book and for the selection of specialized authors for the individual chapters. We are now happy to be able to present the result ol'the work of more than 80 experienced authors from the international scientific conimunity. The idea behind the structure of the Hurzdhook is to provide in Volume I a basic overview of the fundamentals of the science and applications of the entire field o f liquid crystals. This volume should be suitable as an introduction to liquid crystals for the nonspecialist, as well as a source of current knowledge about the state-of-thc-art for the specialist. It contains chapters about the historical development, theory, synthesis and c h e in i c a1 s tr it c t ure , physic a 1 properties, cliaracterir.ation methods, and applications of all kinds of liquid crystals. Two subse~
XXIV
General Introduction
quent volumes provide more specialized information. The two volumes on Low Molecular Weight Liquid Crystals are divided into parts dealing with calamitic liquid crystals (containing chapters about phase structures, nematics, cholesterics, and smectics), discotic liquid crystals, and non-conventional liquid crystals. The last volume is devoted to polymeric liquid crystals (with chapters about main-chain and side-group thermotropic liquid crystal polymers), amphiphilic liquid crystals, and natural polymers with liquid-crystalline properties. The various chapters of the Handbook have been written by single authors, sometimes with one or more coauthors. This provides the advantage that most of the chapters can be read alone, without necessarily having read the preceding chapters. On the other hand, despite great efforts on the part of the editors, the chapters are different in style, and some overlap of several chapters could not be avoided. This sometimes results in the discussion of the same topic from
quite different viewpoints by authors who use quite different methods in their research . The editors express their gratitude to the authors for their efforts to produce, in a relatively short time, overviews of the topics, limited in the number of pages, but representative in the selection of the material and up to date in the cited references. The editors are indebted to the editorial and production staff of WILEY-VCH for their constantly good and fruitful cooperation, beginning with the idea of producing a completely new edition of the Handbook of Liquid Crystals continuing with support for the editors in collecting the manuscripts of so many authors, and finally in transforming a large number of individual chapters into well-presented volumes. In particular we thank Dr. P. Gregory, Dr. U. Anton, and Dr. J. Ritterbusch of the Materials Science Editorial Department of WILEY-VCH for their advice and support in overcoming all difficulties arising in the partnership between the authors, the editors, and the publishers.
The Editors
Part 1: Main-Chain Thermotropic Liquid-Crystalline Polymers
Handbook ofLiquid Crystals D. Demus, J. Goodby, G. W. Gray, H.-W. Spiess, V. Vil Copyright 0WILEY-VCH Verlag GmbH. 1998
Chapter I Synthesis, Structure and Properties
1 Aromatic Main Chain Liquid Crystalline Polymers Andreas Greiner and Hans- Werner Schmidt
1.1 Introduction Aromatic main chain liquid crystalline polymers consist of a sequence of directly connected aromatic moieties, namely, polyarylenes, or a sequence of aromatic moieties linked by an even number of atoms or heterocyclic units. Typical linkage groups in combination with aromatic moieties are, for example, ester and amid groups (Fig. 1). These polymers are semirigid to rigid materials with respect to the conformational freedom along their polymer backbone, and are thermotropic or lyotropic according to the Flory theory, depending mainly on chain extension, melting temperature, and solution behavior [I]. The properties, in particular the solubility and fusibility and as consequence the processability, are controlled by the molecular architecture and chemical structure of the aromatic moieties. Aromatic paralinked polymers with benzene rings as the aromatic moieties, without linking atoms, for example, poly-p-phenylene, or with an even number of linking atoms between the benzene rings, for example, poly-p-hydroxybenzoic acid, are characterized at relatively low molecular weights by high melting temperatures (often in the range of decomposi-
tion) and by poor solubilities. Consequently, in these systems thermotropic or lyotropic phases can only be observed (if at all) under extreme conditions, often in the range of polymer degradation. Improvement of the solubility and the fusibility are of fundamental interest in order to fully exploit the potential of aromatic liquid crystal polymers (LCPs). Therefore, in the past two decades considerable effort has been made to obtain tractable aromatic LCPs by different structural modification concepts. It is far beyond the scope of this chapter to present a comprehensive overview on the different structures of aromatic main chain LCPs and take into account the various property and application aspects. The objective of this chapter is to discuss the impact of structural concepts in modifying the properties of aromatic LCPs, focusing here on aromatic thermotropic LC polyesters. This will be discussed for selected examples. Conclusions from the work on polyesters are transferrable to other classes of thermotropic and lyotropic aromatic polymers.
4
1 Aromatic Main Chain Liquid Crystalline Polymers
tiAi Aromatic units
Aromatic units
Linkage group A
O
H
H
O
I
H
Figure 1. Typical aromatic units and linkage groups in aromatic liquid crystalline polymers.
1.2 Structural Modification Concepts of Liquid Crystal Polymers (LCPs) Under thermodynamic considerations, high melting temperatures are the result of high melting enthalpies and low melting entropies. Consequently, the structural modification concepts are based either on a de-
crease of the melting enthalpy and/or an increase of the melting entropy. These can be achieved mainly by a controlled decrease of the symmetry along the polymer backbone, so reducing interchain interactions, or by an increase in the chain flexibility, but without destroying the formation of LC phases. The various structural modifications, which have been extensively investigated, are schematically illustrated in Fig. 2.
I .2 Structural Modification Concepts of Liquid Crystal Polymers (LCPs) (a)
R
R
R = H or alkyl
5
6
1 Aromatic Main Chain Liquid Crystalline Polymers
Figure 2. Modification concepts for para-linked aromatic polymers including typical monomer structures: (a) monomer units of different length, (b) kinked comonomers, (c) double kinked comonomers, (d) ‘crankshaft’ comonomers, (e) flexible lateral substituents, (f)bulky and stiff lateral substituents, and (g) monomers with noncoplanar conformation.
The incorporation of monomer units of different length, such as p-phenyl, p-biphenyl, or p-terphenyl moities (Fig. 2a), causes the linkage groups to be positioned at random distances along the polymer chain. As a consequence, the intermolecular interactions, particularly of the polar linkage groups, are partially excluded. The overall chain stiffness increases with the amount of the longer aromatic unit. A different structural modification, which is mainly based on lowering the chain stiffness and as a result also on reducing the intermolecular interactions, is the incorporation of kinked and double kinked comonomers (Fig. 2 b and c). Typical monomers with kinks are meta-substituted phenylene derivatives or 4,4’-functionalized biphenylethers and 4,4’-functionalized biphenylsulfides. Monomers with double kinks are,
for example, 4,3’-functionalized biphenylethers or 4,3’-substituted functionalized benzophenones. Substitution in the 4,3‘ position allows for a linear conformation in which the kink is compensated to some degree. However, in both cases these monomers can only be incorporated as a comonomer in para-linked aromatic polymers up to a critical amount, until the polymer chain becomes too flexible and the LC formation is destroyed. An industrially important concept for the structural modification of LCPs involves using the so-called ‘crankshaft’ monomers (Fig. 2d), particularly 2,6-substituted naphthalene derivatives. These monomers result in a step-wise shift along the polymer backbone. The intermolecular interactions can be lowered substantially by lateral substituents, which can be either flexible (Fig. 2e)
1.3 Aromatic LC (Liquid Crystal) Polyesters
or bulky and stiff (Fig. 2f). The number of substituents and symmetry of substitution can result in positional isomerism along the polymer backbone and, as a consequence, in a different direction of the substituents with respect to each other. In some cases, the substituent has an influence on the conformation of the linkage group and lowers the chain stiffness. If should be pointed out that these modifications also result in an increase in the chain diameter. The main differences between flexible and bulky substituents are their influence on the thermal stability and the glass transition temperature ( T J . Flexible chains act as an internal plasticizer, whereas bulky substituents increase the T g . Monomers with noncoplanar conformation are, for example, 2,2’-substituted biphenylenes or binaphthyl derivatives. Substitution in the 2,2’-position causes the phenyl units to be in a noncoplanar conformation. This reduces the intermolecular interactions between the chains very effectively. These monomers reduce the chain stiffness far less than the bulky substituents. It is obvious that combinations of these different structural modifications can be used and have been utilized in numerous examples to modify LCPs’ properties. An other important concept for the modification of LCPs is the incorporation of flexible spacers between mesogenic units. These semiflexible LCPs have been extensively investigated and have to be distinguished from the above systems. As a consequence, a separate contribution in this handbook is devoted to this class of LCPs (Sect. 2 of this Chapter). In the following sections, the different concepts of structural modifications of LCPs will be discussed in more detail for thermotropic LC polyesters and briefly for lyotropic polyamides. The same structural modifications have been applied to other
7
classes of para-linked aromatic polymers, e. g., polyimides, poly-(benzobisoxazole)s, poly-(benzobisthiazole)s, poly-p-phenylenes, poly(p-phenylenevinylene)s, and poly( p-phenyleneethiny1ene)s.
1.3 Aromatic LC (Liquid Crystal) Polyesters Aromatic para-linked LC polyesters have been investigated extensively and represent the most important class of aromatic LCPs. The historical development, the chemistry, and the physics of aromatic LC polyesters have been summarized in several excellent reviews and book chapters. Some examples are given in the literature [2]. Several synthetic routes are known for the synthesis of aromatic LC polyesters. Melt, solution, and slurry polycondensations are mainly used. Most significant are the polycondensation of terephthalic acid diesters and aromatic diols, the pol ycondensation of terephthalic acids and acetates of aromatic diols with the addition of transesterification catalysts, and the polycondensation of aromatic diols and aromatic diacid dichlorides [2]. A method successfully utilized for laboratory synhesis is the polycondensation of silated aromatic diols and aromatic diacid dichlorides [ 3 ] . Molecular weights depend significantly on the reaction conditions and on the solubility as well as the fusibility of the polyesters, which is relatively poor for para-linked unsubstituted aromatic polyesters. Polyesters prepared by the polycondensation of monomers based on terephthalic acid (TA) and hydroquinone (HQ) or p-hydroxybenzoic acid (HBA) are insoluble and exhibit very high melting temperatures, which are in the decomposition range (Fig. 3 ) [4]. Melting temperatures of 600 “C
8
1
Aromatic Main Chain Liquid Crystalline Polymers
HB
t0a-q Poly(HBA)
for poly(TA/HQ) and 610 "C for poly(HBA) were observed by applying DSC heating rates of 80 K/min in order to minimize decomposition [2 b]. This melting temperature is by far too high to obtain stable liquid crystalline phases on processing from the melt. However, these polymers and also poly(4'hydroxybiphenyl-4-carboxylic acid) received attention due to the formation of whiskers and the potential applications of such whiskers in composites [ 5 ] . General interest in thermally processable LC polymers is driven by a combination of properties that distinguish LC-polymers from random coil polymers. In particular, the formation of thermotropic LC phases reduces the melt viscosities during processing, and the increased chain stiffness reduces the thermal expansion coefficients. Both properties are of technical significance in combination with the high temperature stability. This is the basis for the application of LC polyesters as high precision injection molding materials and high modulus fibers. The different structural modification concepts, which will be discussed below, also have to be looked at in context with this property profile. Considerable synthetic efforts have been undertaken in order to decrease the melting temperatures of aromatic LC polyesters while retaining LC properties. Melting temperatures below 350 "C are necessary in order to obtain stable, melt-processable poly-
Figure 3. Monomers and polymer structures of the unsubstituted para-linked aromatic polyesters, poly(p-phenylene terephthalate) [poly(TA/HQ)] and poly(p-hydroxybenzoic acid) [poly (HBA)I.
esters. The different structural modifications shown in Fig. 2 were realized for aromatic LC polyesters. Also, combinations of different concepts are used for copolymers in order to modify their thermal properties and solution behavior. In the following sections, selected examples for each structural concept will be discussed focusing mainly on the thermal properties and the formation of thermotropic mesophases. It is important to note that the thermal properties are a function of the molecular weight, and that for a comparison of the structure - property relations the molecular weight has a significant impact. The inherent viscosity dependence of the melting temperature (T,) and the clearing temperature (Tc),and the molecular weight dependence of the glass transition temperature (T,) are shown in Fig. 4 for two examples. Nematic poly(2-n-decyl- 1,4phenylene-terephthalate) levels to a plateau for T, and T, above an inherent viscosity of 1.0- 1.5 dl/g [6]. The polyester poly(2,2'-
dimethyl-4,4'-biphenylene-phenylterephthalate) forms a nematic melt. Crystallization from the melt is completely suppressed and the polyester only has a T,. The Tg values reach a plateau at a molecular weight about 8000 g/mol (GPC versus PST standards), which corresponds to an inherent viscosity of 1.2 dl/g [7].
1.3 Aromatic LC (Liquid Crystal) Polyesters
Jn
L
350
9
isotropic
"
0 0
Tc
0
nematic
-9 300 -
A
Tm
n
-
crystalline
$
?
250-
200 I
I
I
1
180
160nematic 140-
F
f
YTg nematic glass
d I-
0
100
- l _ 80 I 60
I
'
I
'
I
'
I
.
I
'
Figure 4. (a) Inherent viscosity dependence of the melting temperature (T,) and the clearing temperature (T,) of nematic poly(2-n-decyl- 1,4-phenyleneterephthalate) [6] and (b) the molecular weight dependence of the glass transition temperature (T,) of nematic poly(2,2'-dimethyl-
10
1 Aromatic Main Chain Liquid Crystalline Polymers
1.3.1 Polyesters with Moieties of Different Lengths The combination of monomers such as p-HBA, TA, HQ, and 4,4'-dihydroxybiphenyl (DHB) can be used for the synthesis of copolyesters (Fig. 5). The polycondensation of such monomers causes a structural irregularity in the direction and distance of the ester groups, thus reducing the interchain interactions. The melting temperatures of polyesters synthesized by the copolymerization of p-HBA, TA, and HQ are still above 500 "C and the polyesters are far from forming stable, processable melts [8]. Depending on the composition, the melting temperatures were decreased by 100- 150 "C if HQ was replaced by DHB. These polyesters are injection-moldable and were commercialized under the trade name Ekkcel 1-2000 (Carborundum) and currently as Xydar (Amoco). These materials will gain further commercial interest as the cost of DHB decreases. They belong to the class of high temperature LC polyesters. For example, a special Xydar grade (SRT-500) has a melting temperature of 358 "C and is an injection-moldable resin with a tensile modulus of 8.3 GPa and a tensile strength of 126 MPa.
The molecular structure of these copolyesters is complex and is affected significantly by the synthesis conditions, thermal history, and processing conditions. The sequence distribution of the different repeating units is often found to be more or less in blocks, as indicated by cross polarization solid state I 3C-NMR and X-ray diffraction [9]. The as-polymerized polyester is highly crystalline, indicating block-like ordered sequences, but becomes less crystalline after processing from the molten state, which is due to further transesterification reactions and the formation of a more random sequence distribution.
1.3.2 Polyesters with Kinked or Double Kinked Moieties A reduction of the melting temperatures of aromatic LC polyesters can also be accomplished efficiently by the partial incorporation of kinked comonomers. Suitable monomers are meta- or ortho-functionalized benzene rings, meta- or ortho-functionalized six-member heterocycles, and 2,5-functionalized five-member heterocycles. A frequently used alternative is two benzene rings connected by linking groups such as
Figure 5. Examples of unsubstituted monomers of different lengths and the corresponding LC copolyesters.
1.3 Aromatic LC (Liquid Crystal) Polyesters
-CR,-, -0-, -S-, and -CO-, which are functionalized in the 4,4’-position or the 3,4’-position. Monomers with functional groups in the 3,4‘-position, such as 3,4’-dihydroxy benzophenone (3,4’-DHBP) or 3,4’-dicarboxy-diphenylether (3,4’-CDPE) compensate the kink to a certain extent. These comonomers are also referred to as double kinked comonomers. These types of comonomers can only be incorporated in
RES
oHQ
4,4-ODP
para-linked aromatic polyesters up to a critical amount until the polymer backbone becomes too flexible and self-organization into liquid-crystalline order is no longer possible. Examples of structures of kinked and double kinked monomers are given in Fig. 6. Poly-(p-hydroxybenzoicacid) [poly(HBA)] and poly-(p-phenylene-terephthalate) [poly(TA/HQ)] were modified with monomers
IA
mHBA
ThA
4,4’-TD P
Ho2cGo&C02H
3,4‘- DHBP
11
3,4‘-CDPE
Figure and kinked 6 . Examples double kinked of comonomers for aromatic LC polyesters.
12
1
Aromatic Main Chain Liquid Crystalline Polymers
such as isophthalic acid (IA), resorcinol (RES), or rn-hydroxybenzoic acid (rnHBA) by several research groups [ 101. A liquidcrystalline melt was not found for all compositions. Jackson reported that the content of rnHBA necessary in order to reduce the melting temperature of poly(TA/HQ) below 400 "C destroys the liquid crystallinity [ 10d]. Copolyesters of HQ, TA, and rnHBA with more than 50% rnHBA form isotropic melts. At these concentrations, the chain conformation became too flexible due to the angle within the isophthalate moiety. The ability to pack the molecules in an ordered fashion is hindered and, consequently, no liquid-crystalline phases can be formed. Similar results were obtained with copolyesters of HQ, TA, and RES [lOd], HQ/IA/ HBA (liquid crystallinity is lost above an HQ-IA content >67%) [lOf], TA/RES/ HBA (liquid crystallinity is lost above a TA-RES content of >33%) [log], and HBAIrnHBA (liquid crystallinity is lost above an rnHBA content of >SO%) [lOh]. Lenz et a]. reported that the content of rnphenylene-terephthalate in poly(ch1oro-pphenylene-co-rn-phenylene-terephthalate) cannot exceed 60% in order to maintain liquid crystallinity [ 10 el. However, poly( p phenylene-terephthalate-co-p-phenylene isophthalate) containing about 85% of kinked isophthalate units melts at 355 "C, but exhibits a narrow mesophase range of 30°C [IOi]. Also, the crystallization of para-linked aromatic copolyesters is efficiently suppressed by the incorporation of large amounts of ortho-substituted aromatic units [ 1I]. Compared to meta-substituted monomers, ortho-substituted monomers are more favorable to form an extended chain conformation. For example, a glass transition temperature of 90 "C was reported for a copolyester of TA/methyl substituted HQIoHQ [ 11 b]. This copolyester, described as amorphous, formed a nematic melt about
19 "C above Tgwith a clearing temperature of 380°C. The broad nematic phase was found to have a content of 50% of oHQ and 50% of methyl-substituted HQ. Kinks in phenylene-based polyesters were also obtained by 2S-thiophene dicarboxylate (ThA) moieties, which exhibit a core angle of 148" [12]. These copolymers exhibit stable nematic phases, but their transition temperatures are higher than those of the corresponding isophthalates. For example, the polyester composed of chlorohydroquinone and 25thiophene dicarboxylate [poly(ClHQ)/ThA] exhibits a melting transition around 410 "C with a clearing temperature above SO0 "C (Fig. 7). In contrast, the corresponding polyester poly(ClHQ/TA) shows only a solid state transition at 375 "C and does not melt below SO0 "C. The poly-
100
200
300 400 Temperature ["Cl
500
Figure 7. DSC investigations of three different aromatic homopolyesters: (a) LC polyester composed of chlorohydroquinone and terephthalic acid [poly(CIHQ/TA)], (b) LC polyester composed of chlorohydroquinone and 2,5-thiophene dicarhoxylic acid [poly(ClHQ)/ThA], (c) polyester composed of chlorohydroquinone and isophthalic acid [poly(CIHQ/IA)] (adapted from [ 121).
1.3 Aromatic LC (Liquid Crystal) Polyesters
ester composed of ClHQ/IA has a glass transition endotherm at 146 "C, a recrystallization exotherm around 240 "C, and a melting temperature of 308 "C, but does not exhibit liquid crystallinity. The incorporation of comonomers based on two benzene rings with single atom linking bridges and functional groups in the 4,4'-position were also investigated for modifying the thermal properties of LCPs [ 1 31. The deviation from the chain linearity of aromatic LC polyesters by such comonomers depends on the type of linking bridge and the amount of incorporated comonomer. Compared to the bent, rigid comonomers discussed above, these comonomers, also referred to as 'swivel'-type monomers, are more flexible. A comparative study based on poly(ch1oro-p-phenylene-terephthalate), modified by the incorporation of various amounts of 4,4'-functionalized bisphenols of different structures was performed by Lenz and Jin (Fig. 8) [10e]. For example, copolyesters composed of TA/ClHQ/BPA showed melting temperatures of the order of 340 "C with
13
a BPA content of 10- 30 mol% with respect to the amount of ClHQ. These copolyesters form anisotropic melts, whereas copolyesters with a BPA content of more than 40% were amorphous, forming isotropic melts. Copolyesters with bis(4-hydroxyphenyl) methane (BPM) or 4,4'-sulfonyldiphenol (SDP) have a stronger tendency to form anisotropic melts. It is possible to replace more of the ClHQ with these comonomers without destroying the liquid crystallinity. Based on these results, different 4,4'-functionalized bisphenols destabilize the liquid-crystalline phase quantitatively in the following order:
0 < CH2 - S < SO2 < C(CH,),
1.3.3 Polyesters with Crankshaft Moieties An important structural modification which results in a parallel offset within the polymer backbone is the incorporation of 2,6difunctionalized naphthalene monomers.
L
A n
7% X=
-C-
0
-CH2-
-0-
-S-
CH,
BPA
BPM
ODP
TDP
dSDP
Figure 8. Examples of LC copolyesters derived from terephthalic acid, chlorohydroquinone, and various kinked 4,4'-functionalized bisphenols.
14
1
Aromatic Main Chain Liquid Crystalline Polymers
mco2H
HO
/
HNA
/
HO
DHN
Figure 9. Examples of different crankshaft monomers.
This class of monomers is also referred to as 'crankshaft' monomers. Examples of such monomers are: 6-hydroxy-2-naphthoic acid (HNA), 2,6-dihydroxynaphthalene (DHN), and 2,6-naphthalene dicarboxylic acid (NDA) (Fig. 9) [14]. Figure 10 illustrates the influence of 2,6-naphthalene derivatives in lowering the melting temperature. A copolymer series based on p-HBAITAIHQ has melting temperatures in the range of 420-500 "C. The melting temperatures can be lowered by about 100"C by replacing the terephthalic acid with 2,6-naphthalene dicarboxylic acid (NDA), but the melting point minimum is still about 323°C. A further reduction was achieved in the copolymer series p-HBAITAIDHN. The lowest melting points were observed in a two-component polyester based on 6-hydroxy-2-naphthoic acid (HNA) and p-hydroxybenzoic acid (p-HBA). In this series the melting point could be lowered to a minimum of almost 230 "C. In particular, the results of the poly (HNAIp-HBA) series are the basis for the commercial development of a family of thermotropic LCPs by the Hoechst Celanese Cor-
450 500
-
E
2 400 3
c
c!
E
350 -
ICI
.-c
300 -
a
250-
zoo&
,
,
,
,
20
40
60
80
p-Hydroxybenzoicacid(HBA) [rnol%] poly(p-HBA/TAIHQ)
-.- . - . -
poly(p-HBAINDAIHQ)
..
poly(p-HBA/TA/DHN)
------
........ '
poly(p-HBAIHNA)
Figure 10. Influence of 2,6-difunctionalized naphthalene comonomers on the melting temperatures of LC polyester based on p-hydroxybenzoic acid (adapted from [14a]).
1.3 Aromatic LC (Liquid Crystal) Polyesters
poration under the trade name of Vectra [ 14bl. The Vectra A and Vectra C materials are based on a poly(naphthoate - benzoate) structure. In contrast, the Vectra B series is based on a poly(naphthoate - aminophenolterephthalate), an aromatic copolyester amide. These LCPs can be processed in a similar manner to conventional thermoplastics using almost all of the common polymer processing techniques. The thermotropic melts have a low viscosity, providing good molding characteristics. Molded parts exhibit low warpage and shrinkage, along with high temperature and dimensional stability, as well as chemical resistance. This property profile opened up commercial applications in areas such as electronics/electrical (connectors, capacitor housing, coil formers), fiber-optics (strength members, compilers, connectors), automotive (full system components, electrical systems), and others (compact disc player
15
components, watch components, microwave parts, etc.) [14c]. In addition to the 2,6-naphthalene based LCPs, an interesting aromatic LC polyester with 2,6-substituted DBD moieties has been disclosed in several patents 115, 161.
1.3.4 Polyesters with Lateral Substituents Different types of lateral substituents can be incorporated by monomers with substituents of different size, flexibility, and polarity. In addition, the number of substituents and the symmetry of the substitution can result in a different positional isomerism along the polymer backbone (Fig. 11). For example, in polyesters derived from monosubstituted monomers, the substituents can point towards each other (syn) and away from each other (anti),resulting in a random
positional isomorphism with mono-substituted moieties
R
positional isomorphism with 2,6-di-substituted moieties
polymer backbone with 2,5-di-substituted moieties
~
~
\1 /4
-
~
-
,
A
~
-
R
\~ /-
,
A
,
-
~
~
.
.
\; r /
Figure 11. Structural diversities of erally aromatic substituted LC polymers monomers. with lat-
16
1 Aromatic Main Chain Liquid Crystalline Polymers
sequence along the polymer backbone. A similar, but even more pronounced effect, is found in monomers with 2,6-disubstitution, whereas in a polymer backbone composed of symmetrical 2,5-disubstituted monomers, the substituents are regularly positioned along the backbone and obtain their irregularity by rotational isomerism. A selection of substituted poly( 1,4-phenylene terephthalates) with different types of lateral substituents is summarized in Tables 1 and 2.
1.3.4.1 Polymers with Flexible Substituents As flexible substituents, linear and branched alkyl, alkoxy, or thioalkyl side chains of different lengths have been utilized to modify para-linked aromatic polyesters. The main difference to the systems discussed above is the limited thermal stability caused by the substitution with alkyl chains. Also, the mechanical poperties are substantially lowered with increasing
length and number of the alkyl chains [ 171. The dramatic loss of the mechanical properties is obvious due to the reduced number of load-supporting polymer main chains in a given cross-sectional area. In addition, the intermolecular interactions of the alkyl side chains are relatively weak. Lenz et al. demonstrated for a series of monosubstituted poly(2-n-alkyl- 1,4-phenylene-terephtha1ate)s (Table 1) with the number of C-atoms in the side chain ranging from 6 to 12, that the melting temperatures and the clearing temperatures are lowered, as expected, with increasing length of the alkyl chain [6]. For different samples with inherent viscosities of about 0.5 dl/g (not in the plateau region; see Fig. 4a), the melting temperatures decrease from 277 "C to 228 "C and the clearing temperatures from 323 "C to 291 "C. All the polyesters were found to be nematic. Corresponding polyesters with cyclic and branched alkyl groups (cyclohexyl, cycloheptyl, cyclooctyl, and 3-methylpentyl) [ 181 showed melting temperatures in the range of 308-
Table 1. Glass transition temperatures, melting temperatures, and clearing temperatures of selected mono- and disubstituted poly-(p-phenylene terephthalate) esters with flexible substituents.
R
R'
Tg ("C)
T, ("C)
T, ("C)
Reference
n-hexyl n-heptyl n-ocytl n-nonyl n-undecyl
H H H H H
-
H H H H H H
210 212 237
- (CH,),-phenyl - (CH,),-phenyl
60 110
323 302 307 29 1 292 39 1 390 385 377 > 340 b 340 b 340 b 340
[61 [61 [61
cyclohexyl cycloheptyl cyclooctyl 3-methylpentyl - (CH,),-phenyl - (CHd,-phenyl
277 257 257 237 228 363 352 321 308 340 325 195 225
- (CH2)2-phenyl
t-butyl
-
-
161 [61 [I81 [I81 [I81 1181 [I91 1191 r191 1191
1.3 Aromatic LC (Liquid Crystal) Polyesters
17
Table 2. Glass transition temperatures, melting temperatures, and clearing temperatures of selected mono- and disubstitued poly-(pphenylene terephthalate) esters with bulky and stiff substituents.
C1 H BI H phenyl H phenyl phenyl
c1
phenyl
c1
t-butyl phenox y phenox y phenoxy phenox y t-butyl a)
H c1 H Br H phenyl phenyl C1 phenyl Br Br CF,
c1
Br CF3 phenyl Br
220 220 -
170 130 122 108 113 120 120 141 85 90 88 112 131
312 402 353 405 346 265 206 233 222 313 228 24 1 220
510 475 475 490 475 343 23 1 368 360 376 362 >350 355 > 350
-
-a
-
-a
-a
No LC phase observed.
363 "C. The clearing temperatures from the nematic melt are between 377 and 391 "C. In some cases glass transitions were found. As already mentioned, positional isomerism is important for the solubility and fusibility of aromatic LC polyesters. Consequently, polyesters made from symmetrical 2,5-disubstituted or 2,3,4,5-tetrasubstituted monomers should result in polymers that are less soluble and less fusible. This is in general the case with short lateral substituents. Ballauff and others reported that the series of poly( 1,4-phenylene-2,5-dialkoxy terephtha1ate)s with long flexible alkoxy side chains at the terephthalic moiety result in tractable LC polyesters [20] (Fig. 12). These polyesters exhibit a high degree of crystallinity with melting temperatures below 300 "C. Polyesters with short side chains (2<m 350 "C for m = 2
and T, = 250 "C for m = 6. Polyesters with side chains m 38 result in the formation of layered mesophases, which can also be found in the solid state. The layer spacing in the mesophase, as revealed by X-ray analysis, increases linearly with increasing length of the alkoxy side chain. Longer alkyl side chains (m212) result in sidechain crystallization with melting transitions around 50 "C. Tetrasubstituted polyesters with long alkyl side chains on the hydroquinone moiety connected via ether or ester linkages have been reported [ ~ O C(Fig. ] 12). In these polyesters, the tetrasubstituted moiety can be considered as a discotic mesogen. The polyesters with Z = H are highly soluble and have low melting temperatures ( T , < 100 "C) and low clearing temperatures in the range of 123-260°C. The type of mesophase formed is referred to as "sanid-
18
1 Aromatic Main Chain Liquid Crystalline Polymers
R
R
d
k
R = - CnHPn+l n=7,9,11
Z=Hor
ic". If the terephthalic acid moiety is disubstituted with two long alkoxy side chains (Z = -O-C,H1,), the mesophase is destroyed. This observation can be explained by a reduction of the chain stiffness in combination with the effect of the additional two alkoxy side chains. Similar observations have been made with poly( p-phenylene terephtha1ate)s with one, two, and four thioalkyl side chains of various length [21]. Depending on the degree of substitution and the substitution pattern, thermotropic as well as isotropic polyesters have been obtained. Chiral sanidic polyesters derived from copolymerizations of (S,S)-2,5-bis(2-methylbuty1oxy)terephthalic acid, 2,5-bis(dodecy1)terephthalic acids, and 4,4'-dihydroxybiphenyl have been reported [22]. These
-0-
-C8H17
Figure 12. Structure of 2,5-disubstituted polyesters with flexible alkoxy substituents and of polyesters with tetrasubstituted moieties [20 a-c].
copolyesters formed a solid sanidic layer structure with melting temperatures above 200 "C. A broad enantiotropic cholesteric phase is formed above T,,, with isotropization temperatures in the range of 275325 "C. A different type of flexible substituent are phenylalkyl substituents [ 191. Monosubstituted polyesters with phenylalkyl substituents with different lengths of the alkyl linkage are characterized by high melting temperatures (Table 1). The thermal transitions can be lowered significantly if the terephthalic moieties and the phenylene moieties are substituted by phenylalkyl moieties. The replacement of the phenylalkyl substituent at the hydroquinone moiety by t-butylhydroquinone results in an increase of the glass transition temperature. The polycondensa-
1.3 Aromatic LC (Liquid Crystal) Polyesters
tion of monosubstituted 1,4-hydroquinone derivatives with monosubstituted terephthalic acid derivatives (AABB monomer system) results in polyesters with substituents arranged randomly along the polymer backbone (Fig. 1 1 ) . In contrast, the polycondensation of monosubstituted 4-hydroxybenzoic acid derivatives (AB monomer system) leads to polyesters with defined positions of the substituents. A random arrangement of the substituents is also obtained for copolyesters made from 2- and 3-substituted 4-hydroxybenzoic acid derivatives. The role of positional isomerism was clearly demonstrated by a comparison of homo- and copolyesters based on phenethyl substituted poly(4-hydroxybenzoats)~ (Fig. 13) [23]. Solubility, melting temperature, and the LC phase were almost identical for the homopolyesters of the 2- and 3-substituted homopolyesters exhibiting melting temperatures of 280-297 "C; these give highly viscous birefringent melts with T, > 340 "C. In contrast, the solubility of the corresponding copolyester is enhanced and the crystallinity is depressed, forming an amorphous solid with Tg at 75 "C and a nematic schlieren texture with T, at 280 "C. It should be mentioned that the inherent viscosity of this copolyester was 2.80 dl/g and therefore the observed properties cannot be attributed to a low molecular weight.
1.3.4.2 Polyesters with Stiff, Bulky Substituents The influence of a stiff and bulky substituent on the thermal and the solution properties is governed by the size, polarity, position, and number of the substituents [24]. The main differences between flexible and bulky substituents are their influence on the thermal stability and the glass transition temperature. Whereas flexible chains act as an internal plasticizer and lower the thermal stability, most bulky substituents increase Tg and, depending on the chemical nature, do not lower the thermal stability. A selection of monosubstituted and disubstituted poly-(p-phenylene-terephtha1ate)s is compared in Table 2. Poly( p-phenyleneterephtha1ate)s with methyl, methoxy, chloro, or bromo substituents on either the hydroquinone or the terephthalic acid moiety exhibit melting temperatures of 350 "C or higher. Thermotropic liquid crystalline behavior is observed in these samples, although it is in the range of thermal decomposition. A comparison of the mono- and diphenyl substituted polyesters reveals an important trend. The monosubstituted poly-(p-phenylene-terephthalate) with the phenyl substituent in the hydroquinone moiety melts at 3 4 6 ° C also forming a nematic melt up to a clearing temperature of
L
homopolyesters
19
n
copolyester
Figure 13. Phenethyl-substituted homopolyesters and copolyesters based on poly(p-hydroxybenzoate) [23].
20
1
Aromatic Main Chain Liquid Crystalline Polymers
T, = 475 "C, which is in the range of decomposition. In contrast, the melting temperature of the polyester with the phenyl substituent on the terephthalic moiety is 265 "C and T, = 343 "C, which indicates that the chain stiffness is lowered if the bulky substituent is ortho to the carbonyl group of the ester linkage. The substitution of both the hydroquinone and the terephthalic acid moiety with a phenyl substituent suppressed the crystallization and resulted in a polyester glass transition temperature of 122 "C. Below the glass transition temperature, a glass with nematic order is obtained. The clearing temperature of this polyester was detected at 231 "C, which is considerably lower than those of the monosubstituted polyesters. In other disubstituted poly-(p-phenylene terephthalate)s, particularly with polar substituents such as halogen or trifluoromethyl, relatively weak melting transitions in the temperature range of 210-230 "C are usually observed. In all cases, glass transition temperatures ranging from 108 to 141 "C are observable. The glass transition increases as expected with the volume of the bulky substituents. The clearing temperature of the nematic melts is in all cases above 350 " C . However, polyesters with trifluormethyl-, phenyl-, or bromo-substituents on the terephthalic moieties in combination with a bulky substituent on the hydroquinone moiety were not found to be thermotropic as a result of overly reduced chain stiffness. Monomers with phenyl substituents based on 4,4'-dihydroxybiphenyl, 4-hydroxy-4'-carboxybiphenyl, and 2,6 functionalized naphthalene have been synthesized and incorporated in various LC copolyester systems. Bhowmik et al. used 3-phenyl-4,4'-dihydroxybiphenyl and 3,3'dihydroxybiphenyl as comonomers to modify, for example, poly-(p-hydroxybenzoic acid) and poly-(6-hydroxy-2-naphthoicac-
id) [25]. Jin and co-workers reported on wholly aromatic polyesters derived from 6hydroxy-5-phenyl-2-naphthoic acid and 4'-
hydroxy-3'-phenylbiphenyl-4-carboxylic acid [26]. With these monomers the melting temperatures of copolymers could be lowered substantially without destroying the formation of thermotropic melts, while maintaining thermal stability.
1.3.5 Polyesters with Noncoplanar Aromatic Moieties Noncoplanar aromatic monomers, such as 2,2'-disubstituted biphenylene derivatives and 4,4'-functionalized 1,1'-binaphthyl derivatives (Fig. 14) have been used as comonomers in para-linked aromatic polyesters with remarkable effects on the phase transition temperatures, the crystallinity, and the solubility. The incorporation of these noncoplanar monomers will not initially reduce the chain stiffness. The phenyl rings are forced by the 2,2'-substitution into a noncoplanar conformation which strongly decreases the intermolecular inter-
' O m BDP o H
Figure 14. Examples of noncoplanar diol and diacid monomers.
1.3 Aromatic LC (Liquid Crystal) Polyesters
actions. The noncoplanarity correlates with the size of the substituents. As a consequence, the crystallization tendency and transition temperatures are lowered remarkably and the solubilities are significantly enhanced. Sinta and co-workers 1271 demonstrated that in a variety of para-linked aromatic polyesters and copolyesters with 2,2’-disubstituted-4,4’-biphenylene units, mainly based on 2,2’-ditrifluoromethy1-4,4’-biphenyl diacid (BCFDA), the crystallization can be completely suppressed. The glass transition temperatures are generally between 100 and 150°C with wide nematic phases, and clearing temperatures usually in the range of decomposition. For example, the copolyester based on BCFDA (0.5 mol%), trifluormethyl-substituted terephthalic acid (0.5 mol%), and chlorohydroquinone is amorphous with Tgat 118 “C, and forms an easily shearable nematic melt at 180 “C with clearing temperatures in the decomposition range. Similar results are even obtainable in homopolyesters if the 2,2’-disubstituted biphenylene unit is combined with a monosubstituted phenylene unit [28]. For example, the homopolyester derived from 2,2’-dimethylbipheny1-4,4’biphenyl diacid and t-butylhydroquinone has a glass transition temperature of 150 “C and does not crystallize from the nematic melt. LC polyesters with this substitution pattern are highly soluble in common organic solvents such as THF, chloroform, methylene chloride, and tetrachloroethane up to maximum solubilities in the range of 50% (w/v). Another example of such a type of polyester is shown in Fig. 4. The polyester po-
ly(2,2’-dimethyl-4,4’-biphenylene-phenylterephthalate) derived from 2,2’-dimethylbiphenyl-4,4’-diol (BDMDH) and phenylterephthalic acid forms a nematic melt. Crystallization from the melt is completely
21
suppressed and the polyester displays only one glass transition at around 140°C. This polyester has been obtained with intrinsic viscosities of up to 9.60 dl/g (CHCl,, 20 “C), corresponding to a weight-average molecular weight (M,) of about 80 000 g/ mol [7 a]. The Mark-Houwink-Kuhn parameters of this polyester (CHCl,, 20°C) were determined as [n] =4.93 x 10-4M,0.877. The formation of anisotropic thermoreversible gels was observed with this polyester, for example, in 3-phenoxytoluene as the solvent. These mixtures also retained their anisotropy above the gel melting point at a concentration of 35% or more [29]. In highly aligned fibers spun from the nematic melt and post-drawn, a tensile modulus of 40-45 GPa and a tensile strength of up to 550-650 MPa was reached [28c]. Even higher glass transition temperatures are achievable if 4,4’-functionalized l,l’-binaphthyl monomers are used [28 b, 301. For example, the homopolyester derived from 4,4’-dihydrox- 1, 1’-binaphthyl (BDP) and phenylterephthalic acid with an inherent viscosity of 1.90 dl/g has a Tgof 183 “C. The polyester is nematic up to its decomposition temperature and forms a nematic glass below Tg.In a similar homopolyester based on 4,4’-dicarboxy- 1,1’-binaphthyl and t-butylhydroquinone the T, was found to be 189 “C. In both cases, good solubility in common chlorinated solvents is maintained. This type of monomers was, for example, also used to modify wholly aromatic copolyesters based on 6-hydroxy-2-naphthoic acid (HNA) and 2,6-naphthalene dicarboxylic acid (NDA) [25 a]. In conclusion, the selected examples of this chapter clearly demonstrate that it is possible with the discussed structural modifiations to modify in a systematic manner the melting behavior over a wide temperature range, the degree of crystallization from highly crystalline to amorphous, and the so-
22
1
Aromatic Main Chain Liquid Crystalline Polymers
lution behavior from solvent-resistant to highly soluble, while maintaining thermotropic liquid crystalline behavior. In particular, the combination of different structural modifications in copolymers allows the systematic tailoring of the thermal properties, the solution behavior, and the chain stiffness of thermotropic aromatic LC polyesters.
1.4 Para-Linked Aromatic Polyamides The same structural modification concepts, which were utilized to modify the properties of para-linked aromatic LC polyesters, have also been applied to aromatic polyamides. Para-linked aromatic polyamides are an important class of LC polymers. In contrast to thermotropic LC polyesters, para-linked aromatic polyamides form lyotropic solutions. Due to the formation of intermolecular hydrogen bridges, these polymers are in most cases unable to melt below their thermal decomposition temperature. Infusibility and limited solubility of unsubstituted para-linked aromatic polyamides are characteristic properties which limit synthesis, characterization, processing, and applications. Already in 1965, the formation of lyotropic liquid crystalline solutions of poly(p-aminobenzoic acid) in concentrated sulfuric acid was observed. This work was the basis for the commercialization of high strength, high modulus, and heat-resistant poly-(p-phenylene terephthalate) (PPDT) fibers under the trade name Kevlar (Dupont) and Twaron (Akzo) (Fig. 15) [31]. PPDT is typically prepared by the polycondensation of terephthaloyl chloride and 1,4-phenylenediamine in tertiary amidic solvents such as N-methylpyrrolidinone or dimethylacet-
Figure 15. Structure of poly( 1,4-~henyleneterephthalamide).
amide containing lithium or calcium chloride. PPDT forms a lyotropic nematic solution in concentrated sulfuric acid solution when a minimum concentration is exceeded. The PPDT fiber is spun from a lyotropic solution in concentrated sulfuric acid. The solution behavior has been significantly enhanced by the same structural modifications as reported previously for aromatic LC polyesters. For example poly-(pphenylene terephthalamide) has been modified by bulky, stiff substituents [32], flexible alkyl side chains [33], the incorporation of kinked and double kinked comonomers, and comonomers of different lengths [34], as well as the use of noncoplanar biphenylene monomers [35]. To develop high performance materials, modifications that increase the solubility while maintaining the rod-like character, high glass transition temperatures, and the temperature stability are of particular interest. The solubility and the chain stiffness are critical factors in achieving lyotropic solutions. Depending on the modifications of poly(p-phenylene terephthalamide) and analogous aromatic polyamides, it is possible to vary the solubility from concentrated sulfuric acid, to nonpolar aprotic solvents with inorganic salts, or to nonpolar aprotic solvents and to common organic colvents. In particular the incorporation of noncoplanar 2,2’disubstitution has proven to remarkable enhance the solubility and lower the crystallinity. Gaudiana et al. [35] demonstrated that para-linked aromatic polyamides containing noncoplanar 2,2’-bis(trifluoromethy1)biphenylene units are highly soluble in
1.5 References
aprotic polar amide solvents such as DMAc or NMP, and even in common solvents such as THF and acetone. Although high polymer concentrations were achieved, a general lack of lyotropic behavior was observed. In contrast, it was found that para-linked aromatic pol yamides containing 2,2’-dimethylbiphenyl units are in some cases capable of forming lyotropic solutions in amide solvents [32 h-j]. The critical concentrations to form a lyotropic phase are strongly influenced by the size and position of the substituents, which govern the overall chain stiffness and capability of chain extension. For some systems, concentrations of up to 40-50 wt.% are required to obtain a lyotropic solution. These results demonstrate the complex interplay of chemical substitution, chain stiffness, choice of solvent, and solvent/polymer chain interaction. The substituted polyamides with long alkyl- and alkoxychains [33] are highly soluble and form anisotropic melts above their melting temperature. These polyamides are typical examples of sanidic liquid crystalline polymers. Generally, no lyotropic behavior is observed. The temperature stability is obviously substantially lowered due to the substitution with alkyl chains. It should also be mentioned that a combination of ester and amide linkages was frequently used to prepare thermotropic LC polyesteramides. Frequently, the inexpensive p-aminophenol is incorporated. For example, the Vectra B series is a poly(naphthoate-aminophenolterephthalate) derived from 6-hydroxy-2-naphthoic acid (HNA), p-aminophenol, and terephthalic acid [ 141. It is obvious that only a limited amount of p-aminophenol can be incorporated in order to maintain thermotropic liquid crystalline behavior. The amide linkage enhances the intermolecular interactions via hydrogen bridges and improves the solid state properties of the material.
23
1.5 References P. J. Flory, Proc. R. Soc. London 1956, 73, A234. See, for example: a) M. G. Dobb, J. E. McIntyre, Adv. Polym. Sci. 1984,60/61,6 1;b) T.-S. Chung, Polym. Eng. Sci. 1986,26,901;c) S. L. Kwolek, P. W. Morgan, J. R. Schaefgen in Encyclopedia of Polymer Science, Vol 9 (Eds.: H. F. Mark, N. M. Bikules, C. G. Overberger, G. Menges, J. 1. Kroschwitz) Wiley, New York 1987,l;d) M. Ballauff,Angew. Chem. (Int. Ed.) 1989,28, 253; e) C. M. McCullagh, J. Blackwell in Comprehensive Polymer Science, 2nd Suppl. (Eds.: S. Aggarwal, S. Russo), Elsevier, Oxford, U. K. 1996; f) H. H. Yang, Aromatic High-Strength Fibers, Wiley-Interscience, New York 1989. H. R. Kricheldorf, Silicon in Polymer Synthesis, Springer, Berlin 1996. a) W. M. Eareckson, J. Polym. Sci. 1959, 40, 399; b) W. J. Jackson, Jr., B KPolym. J. 1980,12, 154; c) J. Economy, R. S. Storm, V. I. Matkovich, S . G. Cottis, J. Polym. Sci., Polym. Chem. Ed. 1976, 14, 2207; d) G . Lieser, G. Schwarz, H. R. Kricheldorf, J. Polym. Sci., Polym. Phys. Ed. 1983, 21, 1599; e) H. R. Kricheldorf, G. Schwarz, Makromol. Chem. 1983, 184, 475; f) G. Schwarz, H. R. Kricheldorf, Makromol. Chem., Rapid Commun. 1988,9,717; g) J. Economy, W. Volksen, C. Viney, R. Geiss, R. Siemens, T. Karris, Macromolecules 1988, 21, 2171. a) G. Schwarz, H. R. Kricheldorf, Macrornolecules 1995, 28, 3911; b) H. R. Kricheldorf, F. Ruhser, G. Schwarz, T. Adebahr, Makromol. Chem. 1991, 192, 2371; c) C. Taesler, J. Petermann, Makromol. Chem. 1991, 192,2255. J. Majnusz, J. M. Catala, R. W. Lenz, E U K Polym. J. 1983, 19, 1043. a) R. Schmitt, A. Greiner, W. Heitz, Makromol. Chem. Symp. 1992,61, 297; b) W. Heitz, H.-W. Schmidt,Makromol. Chem. Symp. 1990,38,149. J. Economy, S . G. Cottis, B. E. Nowak, German Patent 2,025,971, 1972. W. Volksen, J. R. Lyerla, Jr., J. Economy, B. Dawson, J. Polyrn. Sci. Pulym. Chenz. Ed. 1983, 21, 2249. a) S. G. Cottis, J. Economy, B. E. Novak, U. S. Patent 3,637,595, 1972; b) S . W. Kantor, F. F. Holub, U. S. Patent 3036990, 1962; c) I. J. Goldfarb, D. R. Bain, R. McGuchan, A. C. Meeks, Org. Coat. Appl. Polym. Sci. Proc. 1971, 31, 130; d) W. J. Jackson, Jr., BK Polym. J . 1980, 12, 154;e) R. W. Lenz, J.-I. Jin,Macromolecules 1981, 14, 1405; f) A. B. Eredemir, D. J. Johnson, J. G. Tomka, Polymer 1986, 27, 441; g) J.-I. Jin, S . H. Lee, H. J. Park, Polym. P r e p (Am. Chem. Soc. Polym. Div.) 1987,28(1), 122; h) R. Rosenau-Eichin, M. Ballauff, J. Grebo-
24
1
Aromatic Main Chain Liquid Crystalline Polymers
wicz, E. W. Fischer, Polymer 1988,29,518; i) R. Cai, E. T. Samulski, Macromolecules 1994, 27, 135. [11] a) B. P. Griffin, R. G. Fessey, EU Patent 0008855, 1980; b) C. Noel, C. Friedrich, F. Lauprete, J. Billard, L. Bosio, C. Strazielle, Polymer 1984, 25, 263; c) J.-I. Jung, J.-H. Chang, Macromolecules 1989, 22, 93; d) J.-I. Jung, E.-J. Choi, C.-S. Kang, J.-H. Chang, J. Polym. Sci. Chem. Ed. 1989, 27, 2291; e) F. Navarro, J. L. Serrano, J . Polym. Sci.: Part A: Polym. Chem. 1992,30, 1789. [I21 R. Cai, J. Preston, E. T. Samulski, Macromolecules 1992,25, 563. [ 131 a) B. P. Griffin, M. K. Cox, Br. Polym. J. 1980, 12, 147; b) H. R. Kricheldorf, V. Doring, Polymer 1992,33,5321;c) R. 0.Garray, P. K. Bhowmik, R. W.Lenz, J. Polym. Sci.: PartA: Polym. Chem. 1993,31, 1001; d) see also [4b] and [loel. [ 141 a) G. W. Calundann in High Performance Polymers: Their Origin and Development (Eds.: R. B. Seymour, G. S . Kirshenbaum) Elsevier, Amsterdam 1986, p. 235; b) selected patents: G. W. Calundann (Celanese), U. S. Patents 4,067,852, 1978; 4,185,996, 1980; 4,161,470, 1979; and4,256,624,1981; G. W. Calundann, H. L. Davis, F. J. Gorman, R. M. Mininni (Celanese), U. S . Patent 4,083,829, 1978; A. J. East, L. F. Charbonneau, G. W. Calundann (Celanese), U. S . Patent 4,330,457, 1982; c) information package on Vectra products, Hoechst Celanese Corporation. [15] R. S. Irwin, EU Patent 26,991, 1981. [16] J. F. Harris, Jr., US Patent 4,447,592, 1984. [17] A. R. Postema, K. Liou, F. Wudl, P. Smith, Macromolecules 1990, 23, 1842. [18] H.-R. Dicke, R. W. Lenz, J. Polym. Sci.: Polym. Chem. Ed. 1983,21, 258 1 . [ 191 W. Briigging, U. Kampschulte, H.-W. Schmidt, W. Heitz, Makromol. Chem. 1988, 189, 2755. [20] a) M. Ballauff, Makromol. Chem., Rapid Commun. 1986,7,407;b) M. Ballauff, G. F. Schmidt, Mol. Cryst. Liq. Cryst. 1987, 147, 163; c) H. Ringsdorf, P. Tschirner, 0. Hermann-Schonherr, J. H. Wendorff, Makromol. Chem. 1987, 188, 1431. [21] H. R. Kricheldorf, A. Domschke, Macromolecules 1996, 29, 1337. [22] H. R. Kricheldorf, D. F. Wulff, J. Polym. Sci. Part A - Polyrn. Chem. 1997,35,947. [23] W. Vogel, W. Heitz, Makrornol. Chem. 1990, 191, 829. [24] a) I. Goodman, J. E. McIntyre, D. H. Aldred, BP 993 272, 1975; b) W. R. Krigbaum, H. Hakemi, R. Kotek, Macromolecules 1985,18,965 (1985); c) W. Heitz, N. NieBner, Makromol. Chem. 1990, 191, 225. [25] a) P. K. Bhowmik, E. D. T. Atkins, R. W. Lenz, H. Han, Macromolecules 1996, 20, 1919;
b) P. K. Bhowmik, E. D. T. Atkins, R. W. Lenz, Macromolecules 1993, 26, 447; c) P. K. Bhowmik, E. D. T. Atkins, R. W. Lenz, Macromolecules 1993,26,447;c) P. K. Bhowmik, E. D. T. Atkins, R. W. Lenz, H. Han, Macromolecules 1996,29, 3778. [26] a) J. I. Jin, S.-M. Huh, Macromol. Symp. 1995, 96, 125; b) S.-M. Huh, J. I. Jin, Macromolecules 1997,30,3005. [27] a) R. Sinta, R. A. Gaudiana, R. A. Minns, H. G. Rogers, Macromolecules 1987, 20, 2374; b)R. A. Gaudiana, R. A. Minns, R. Sinta, N. Weeks, H. Rogers, Progr. Polym. Sci. 1989, 14, 47. [28] a) H. W. Schmidt, D. Guo, Makromol. Chem. 1988, 189, 2029; b) W. Heitz, H.-W. Schmidt, Macromol. Chem. Macromol. Symp. 1990, 38, 149; c) F. Montamedi, U. Jonas, A. Greiner, H.-W. Schmidt, Liq. Cryst. 1993, 14, 959. [29] a) A. Greiner, W. E. Rochefort, K. Greiner, G. W. Heffner, D. S. Pearson, H.-W. Schmidt i n Integration of Fundamental Polymer Science 5 (Ed.: P. J. Lemstra, L. A. Kleintjens), Elsevier, New York 1991; b) A. Greiner, W. E. Rochefort, K. Greiner, H.-W. Schmidt, D. S. Pearson, Makromol. Chem. Rapid Commun. 1992, 13, 25; c) A. Greiner, W. E. Rochefort in Polymer Liquid Crystals - Mechanical and Thermophysical Properties (Ed.: W. Brostow), Chapman & Hall, London 1997,43 1. [30] M. Hohlweg, H.-W. Schmidt, Macromol. Chem. 1989,190, 1587. [31] a) D. Tanner, J. A. Fitzgerald, B. R. Phillips,Angew. Chem. Adv. Muter. 1989,101,665; b) S. L. Kwolek, P. W. Morgan, J. R. Schaefgen, L. W. Gulrich, Macromolecules 1977, 10, 1390; c) T. I. Blair, P. W. Morgan, F. L. Killian, Macromolecules 1977,10, 1396; d) see also [2c] and ~
Vfl. [32] See, for example: a) J. Y. Jadhav, W. R. Krigbaum, J. Preston, J. Polym. Sci., Part A: Polym. Chem. 1989,27, 1175; b) N. Nagata, N. Tsutsumi, T. Kyotsukuri, J. Polyrn. Sci., Part A: Polym. Chem. 1988,26,235; c) W. R. Meyer, F. T. Gentile, U. W. Suter, Macromolecules 1991,24,633; d) W. R. Meyer, F. T. Gentile, U. W. Suter, Macromolecules 1991, 24, 642; e) B. R. Glomm, G. C. Rutledge, F. Kiichenmeister, P. Neuenschwander, U. W. Suter, Macromol. Chem. Phy5. 1994, 195, 475; f) H. R. Kricheldorf, B. Schmidt, R. Burger, Macromolecules 1992, 25, 5465; g ) H . R. Kricheldorf, B. Schmidt, R. Burger, Macromolecules 1992, 25, 5471; h) W. Hatke, H.-W. Schmidt, Makromol. Chem., Macromol. Symp. 1991, 50, 41; i) W. Hatke, H.-W. Schmidt, W. Heitz, J . Polym. Sci., PartA: Polym. Chem. 1991, 29, 1387; j) W. Hatke, H.-W. Schmidt, Makromol. Chem. Phys. 1994, 195, 3579.
1.5 References
[33] See, for example: a) M. Ballauff, G. F. Schmidt, Makrornol. Chem., Rapid Commun. 1987,8,93; b) M. Ballauff, G. F. Schmidt, Mol. Cryst. Liq. Cryst. 1989, 147, 163; c) M. Ebert, 0.HermannSchdnherr, J. H. Wendorff, H. Ringsdorf, P. Tschirner, Makromol. Chem., Rapid Commun. 1988, 9, 445; d) H. Ringsdorf, P. Tschirner, 0. Hermann-Schonherr, J. H. Wendorff, Mukromol. Chem. 1987, 188, 1431; e) J . M . RodriguezParada, R. Duran, G. Wegner, Macromolecules 1989,22,2507; f) A. Adam, H. W. Spiess, Mukromol. Chem. Rapid Commun. 1990, 11, 249; g) L. Yu, Z. Bao, R. Cai, Angew. Chern. 1993, 105, 1392; h) P. Galda, D. Kistner, A. Martin, M. Ballauff, Macromolecules 1993,26, 1595; i) M. Sone, I. B. R. Harkness, J. Watanabe, T. Yamashita, T. Torii, K. Horie, Polym. J . 1993,25,997; j ) H. R. Kricheldorf, A. Domschke, Macromolecules 1994, 27, 1509; k) I. G. Voigt-Martin, P.
25
Simon, S. Bauer, H. Ringsdorf, Macromolecules 1995, 28, 236; 1) M. Steuer, M. Horth, M. Ballauff, J . Polym. Sci., PartA: Polym. Chem. 1993, 31, 1609. [34] See, for example: a) J. Preston, Polyrn. Eng. Sci. 1975, 15, 199; b) J. Preston, R. W. Smith, S. M. Sun, J . Appl. Polym. Sci. 1972, 16, 3237; c) J. Blackwell, R. A. Cageao, A. Biswas, Mucromolecules 1982, 20, 667; d) see also [2 c] and
Vfl.
[35] See, for example: a) R. A. Gaudiana, R. A. Minns, H. G. Rogers, P. S. Kalyanaraman, C. McGowan, W. C. Hollinsed, J. S. Manello, R. Sahatjian, Macromolecules 1985, 18, 1085; b) R. A. Gaudiana, R. A. Minns, H. G. Rogers, R. Sinta, L. D. Taylor, P. S. Kalyanaraman, C. McGowan, J. Polym. Sci., PurtA: Polym. Chem. 1987, 25, 1249; c) see also [27b], [32h], [32i], and [32j].
Handbook ofLiquid Crystals D. Demus, J. Goodby, G. W. Gray, H.-W. Spiess, V. Vil Copyright 0WILEY-VCH Verlag GmbH. 1998
2 Main Chain Liquid Crystalline Semiflexible Polymers Emo Chiellini and Michele Laus
2.1 Introduction Since the discovery [l] that high performance polymeric fibers could be obtained from rigid-rod polymers forming liquid crystalline mesophases in solution, an enormous research effort has been addressed to the preparation and study of main chain polymers containing rigid anisometric segments. This has given rise to such commercial polymers as poly(p-phenylene terephthalamide) from which Kevlar (Du Pont) and Twaron (Akzo) fibers are spun, most of which are constituted of aromatic sequences. As all the poly(aromatic amide)s (aramid) are unmeltable and can only be spun from very aggressive solvents such as sulfuric acid, the macromolecular design was successively directed to the development of melt-processable liquid crystalline polymers [2].Melt processability can be very effectively obtained by disrupting the regular structure of the main chain, while not substantially reducing the chain stiffness, by random copolymerization, and/or the introduction of rigid nonlinear comonomeric units or of lateral substituents in the chain. This “frustrated chain packing” concept was successfully applied to the design
of the copolyesters consisting of 2-OXY-6naphthoyl and p-benzoyl units, from which Vectra fibers [ 3 ](Hoechst-Celanese) are obtained, and in the preparation of poly(ethylene terephthalate) (PET) modified with ripid p-hydroxybenzoic acid units, from which X7G fibers [4] (Eastman-Kodak) are prepared. The depressed efficiency of chain packing in the solid state results in relatively low melting temperatures and unusual ease in processing. The mechanical properties of these polyesters, including the ultrahigh modulus and strength, high thermal stability, and exceptionally low thermal expansion coefficient, were the subject of many investigations and several reviews [5 - 81. However, the synthetic procedures leading to the above copolymers do not guarantee the achievement of the same microstructure. The as-obtained sequence distributions are probably unstable because the presence of transesterification reactions leads to highly complicated solid state behavior, which typically depends on the thermal history and processing conditions of the sample. Definite relations between the phase behavior, the molar mass, and the molar mass distribution are not easily available because
2.2
27
Preparation of Main Chain Liquid Crystalline Polymers (MCLCPs)
of the poor solubility of these polymers. Finally, on account of the relative incompatibility of the different copolymer sequences, driven by their distinct inherent rigidities, multiphase domain structures in the submicrometer scale are obtained whose size and shape affect the physicochemical properties substantially. Considering the above-mentioned problems, the rigid rod polymers are very interestig for numerous technological opportunities, but they are not suitable candidates for the assessment of reliable correlations between the molecular and structural characteristics and the mechanical and physiochemical properties, including the tendencies to self-assemble and form mesophases. As an alternative way to reduce the phase transition temperatures, the main chain structure can also be disrupted by the insertion of flexible segments, i. e., molecular segments capable of assuming a variety of different configurations at accessible energy levels. The regularly alternating arrangement of flexible spacers and rigid anisometric groups gives rise to thermotropic main chain semiflexible liquid crystalline polymers (MCLCP). In the great majority of these polymers, rodlike units were employed with their main direction lying along the polymer chain, according to the basic structure illustrated in'Fig. 1 a. However, a few examples of polmers in which the rodlike units are placed with their long axes lying perpendicular to the polymer chain [9], according to the structure illustrated in Fig. l b, and in which the anisometric units present were of a disk-like shape (Fig. 1 c), were also reported [lo]. A collection of papers and rather comprehensive reviews were published on these polymers [11-221. In general, for these semiflexible polymers, the molecular packing features responsible for their unique thermodynamic
0 Flexible spacer: Rigid disklike group: 0 Rigid rodlike group:
JVV\P
Figure 1. Schematic representation of the main chain polymers obtained from the combination of rigid anisometric and flexible elements.
and mechanical properties arise from a crucial interplay between the ordering propensity, imparted by the interactions among the rigid groups that are orientation-dependent, and the intra and/or interchain constraints dictated by the flexible spacers connecting the mesogenic groups, including the specific effects exerted by the linking groups. Accordingly, the overall polymer molecule must adopt a spatial configuration whose features define the structure, symmetry, and stability of the mesophase. In addition to these structural modeling features, the polymeric nature influences the liquid-crystalline behavior of main-chain polymers through the molar mass and its distribution characteristics.
2.2 Preparation of Main Chain Liquid Crystalline Polymers (MCLCPs) 2.2.1 Generalities Semiflexible MCLCPs were prepared following two basic conceptual procedures. The first one (melt transesterification) involves the chemical modification of pre-
28
1t
0
2 Main Chain Liquid Crystalline Semiflexible Polymers
o0hOCH2CH20
+C
P
?
H ~ C co OH ~
BC- OCH2CH2O X
Figure 2. Chemical modification of PET via a melt transesterification reaction with p-acetoxybenzoic acid, to obtain a liquid-crystalline copolyester.
formed nonliquid crystalline polymers. By this approach, the poly(ethy1ene terephthalate) (PET) was reacted [23] with p-acetoxybenzoic acid, via a melt transesterification reaction performed at temperatures up to 270°C and catalyzed by zinc or antimony acetates, to provide a high molar mass liquid crystalline copolyester, according to Fig. 2. The structure of this copolyester consists of a sequence of ethylene, terephthalic, and oxybenzoic units. Depending on the overall composition and on the reaction conditions, which strongly affect the polymer microstructure, a variety of mesogenic aromatic dyads, triads, and even longer units can be obtained, interconnected by ethylene spacers. Accordingly, on one hand, the overall chain stiffness increases with respect to the starting PET, thus increasing the tendency of the polymer chains to pack into ordered structures. On the other hand, the chain periodicity is destroyed, thus somewhat inhibiting crystallization. The net result of these two conflicting tendencies is the reduction of the melting point and the increase of the isotropization temperature, thus leading to the formation of a liquid-crystalline melt which extends over a wide temperature range. The second synthetic approach (monomeric components step-growth) is based on more conventional step-growth poly-
merization reactions between appropriate monomers. In this case, the mesogenic units can be prebuilt in one precursor species, or they are formed during the polymerization itself. These reactions can be realized through high temperature transesterifications in the melt, for example, of diacetylated diphenols and diacids, or by a variety of conventional polymerization reactions carried out in solution at low or medium temperature, with or without catalysts, in homogeneous or heterogeneous conditions. These latter reactions allow for better control of the polymer microstructure, as long as they are performed under relatively mild conditions, thus preventing molecular scrambling and structure rerrangements. In addition, it is possible to predetermine the end group nature and, to some extent, the molar mass values, by performing the reactions under controlled nonstoichiometric conditions. Following this second approach, a wide variety of MCLCPs with highly differentiated structures were prepared, including polyesters, polyethers, polyurethanes, polyamides, polysiloxanes, poly(o-aminoester)s, and poly(Pthioester)s. It should be observed that most of the above polymers contain a variety of different groups in the repeating unit. For example, most of the described polyurethanes contain an ester function, in addition to the urethane functions. Accordingly, this classification is not unequivocal but is made on the basis of the chemical group or functionality, which is particularly important in determining the thermal and mesophasic behavior of the corresponding polymer.
2.2.2
Polyesters
Polyesters represent the majority of the prepared liquid-crystalline polymers. The
2.2
Preparation of Main Chain Liquid Crystalline Polymers (MCLCPs)
29
Table 1. Representation of some typical structures of polyesters. Structure” -
Ref.
R = polymethylene segments, segments containing etheroatoms, unsaturation, aliphatic, or aromatic rings.
a
Figure 3. Synthesis of an LC polyester through a melt transesterification reaction between the mesogenic 4,4’-diacetoxybiphenyl and a diol.
structures of representative examples of these polymers are reported in Tables 1-7. LC polyesters are generally prepared by following the monomeric component stepgrowth procedure involving condensation in the melt or in solution of stoichiometrically defined mixtures of the monomeric
components. A typical example of condensation in the melt can be represented by the reaction of transesterification [24, 251 of the mesogenic 4,4’-diacetoxybiphenyl and a polymethylene diol, as illustrated in Fig. 3 . These reactions are usually carried out at temperatures in the 200-280°C range under an argon or nitrogen atmosphere and at a reduced pressure. Catalysts such as zinc, titanium, or antimonium salts are often employed to speed the reaction. This synthetic approach presents great advantages in that it is possible to obtain high molar mass polymers in quantitative yields without the necessity of solvent removal, and can be recommended when the polymer does not contain chemically different ester functions. However, if distinct ester groups are present within the polymer chain, the above transes-
30
2 Main Chain Liquid Crystalline Semiflexible Polymers
Table 2. Representation of some typical structures of polyesters. S tructurea
Ref.
0
OC-R-C
P
co 0
PCO-R-0 -f
[38, 391
[32,40, 41 -461
a? P
OC-R-C$
a R = polymethylene segments, segments containing etheroatoms, unsaturation, aliphatic, or aromatic rings.
terification reactions lead to polymers with not well defined structures. In this case, to obtain polymers with controlled structures it is necessary to run the reaction under milder conditions in the presence of solvent, employing either solution or interfacial polycondensation methods. The former reaction was successfully employed in the preparation of polyesters by reacting the mesogenic dihydroxy-4,4’-phenyl benzoate with several a,o-alkanediacid chlorides in achloronaphthalene as the solvent (Fig. 4). Alternatively, the same polyester can be prepared by reacting polymethylene bis(phydroxybenzoates) and polymethylene bis(pchloroformy1benzoates) in tetrachloroethane using pyridine as the hydrochloric acid acceptor, according to the reaction scheme illustrated in Fig. 5.
Figure 4. Synthesis of LC polyesters by reacting the mesogenic dihydroxy-4,4’-phenyl benzoate with a,walkanediacid chlorides.
However, the majority of polyesters was prepared using the latter reaction, that is, by interfacial polycondensation. Usually, this reaction is carried out at room temperature in a solvent mixture consisting of water and an immiscible organic solvent such as, for example, tetrachloroethane, chloroform, or
2.2
31
Preparation of Main Chain Liquid Crystalline Polymers (MCLCPs)
Table 3. Representation of some typical structures of polyesters.
R = (CH,),, (CH2CH,O),CH,CH,
[52-541
R=(CH,), X=H, Y =H; X=CH,, Y =H; X=CH,, Y =CH,; X = CHZCH,, Y = CHZCH,
[55-591
Figure 5. Synthesis of LC polyesters by reacting polymethylene bis(p-hydroxybenzoates) and polymethylene bis(pchloroformy1benzoates).
dichloroethane. A monomer, usually a diacid chloride, is dissolved in the organic solvent whereas the other monomer, usually a diphenol, is dissolved in water after reaction
with sodium or potassium hydroxide. The two solutions are then mixed together in the presence of a tertiary amine, such as pyridine or triethylamine, or in the presence of
32
2
Main Chain Liquid Crystalline Semiflexible Polymers
Table 4.Representation of some typical structures of polyesters. Structure
Ref.
0
0
+
Br(CH2),Br
a phase transfer catalyst such as the benzyltriethylammonium chloride.
2.2.3 Polyethers O(CH,),,+ Figure 6. Synthesis of LC polyethers by aphase transfer catalyzed polyetherification reaction.
Polyethers are usually prepared by a phase transfer catalyzed polyetherification reaction between the sodium salt of the mesogenic diphenol dissolved in water and an a,lvdibromoalkane dissolved in o-dichlorobenzene or nitrobenzene in the presence of
2.2
Preparation of Main Chain Liquid Crystalline Polymers (MCLCPs)
33
Table 5. Representation of some typical structures of polyesters. Structure
0
Ref.
0
0
0
0
0
Figure 7. Synthesis of LC polyurethanes by a stepwise polyaddition reaction of a mesogenic diol with an aliphatic or aromatic diisocyanate.
tetrabuthylammonium hydrogen sulfate or tetrabutyl ammonium bromide as the phase transfer catalyst, according to Fig. 6. The structures of some prominent examples of LC polyethers are collected in Table 8.
2.2.4 Polyurethanes The preparation of thermotropic LC polyurethanes is performed essentially according to two synthetic pathways involving the stepwise polyaddition reaction of a meso-
34
2
Main Chain Liquid Crystalline Semiflexible Polymers
Table 6. Representation of some typical structures of polyesters. Structure
Ref.
[70 - 721 X=CH,, C1, Br, O(CH,CH,O),CH,, O(CH2CH,O),CH2CH,, n = 0, 1 , 2
X &=
Y
x,Y= c1,
f&O & Ar & +co-R-0 0
0
0
-
-
"
Table 7. Representation of some typical structures of polyesters and copolyesters. Structure
Ref.
0
0
n
1
2.2
Preparation of Main Chain Liquid Crystalline Polymers (MCLCPs)
35
Table 8. Representation of some typical structures of polyesters and copolyesters. Structure
Ref.
HO-~I-OH
=
: Mesogenic group
+
-NCO
T I:
Non mesogenic group
Figure 8. Synthesis of LC polyurethanes by a stepwise polyaddition reaction of a mesogenic diisocyanate with a diol.
genic diol with an aliphatic or aromatic diisocyanate (Fig. 7), or alternatively of a mesogenic diisocyanate with an aliphatic diol (Fig. 8). Some examples of LC polyurethanes are collected in Table 9. The polymerization reactions were carried out in DMSO or DMF at 120 "C for 24 h without any added catalyst. However, substantial irreproducibility of the molecular characteristics of the poly-
mers was observed in successive preparations, and sometimes an induction period was detected. These effects are probably due to the presence of adventitious water in these solvents, which is able to react with the diisocyanate monomer leading to reactive and unreactive by-products. The addition of catalytic quantities of 1,4-diazabicyclo[ 2.2.21octane considerably increases the reaction rate, but only oligomeric products
36
2 Main Chain Liquid Crystalline Semiflexible Polymers
Table 9. Representation of some typical structures of polyurethanes. Structure
Ref.
-(CH&-
0'"' CH2-
[97- 1001
with
:
were obtained. Polymeric products with reproducible and high molar masses were obtained by running the reactions at 90 "C in dry chloroform in a nitrogen atmosphere. Alternatively, LC polyurethanes were prepared [ 101, 1021 by reacting dichloroformates and diamines.
2.2.5 Polyamides Liquid crystalline main chain polyamides with a strictly alternating structure were prepared [ 1031by reacting aliphatic diacids and
bis(aminobenzoates) in the presence of pyridine and triphenylphosphyne according to the Yamazaki procedure at moderate temperatures of 100- 115 "C (Fig. 9).
2.2.6 Polysiloxanes Polysiloxanes are generally prepared, as illustrated in Fig. 10, by reacting equimolar amounts of a mesogenic monomer containing allylic functions with a-dimethylsilanyl-a-hydrogen-oligodimethylsiloxanesin the presence of a chloroplatinic catalyst or
37
2.2 Preparation of Main Chain Liquid Crystalline Polymers (MCLCPs)
Figure 9. Synthesis of LC polyamides by the polycondensation reaction between aliphatic diacids and bis(aminobenzoate)s.
Figure 10. Synthesis of LC polysiloxanes by the polycondensation reaction between a mesogenic monomer containing allylic functions and oligodimethylsiloxanes.
Table 10. Representation of some typical structures of polysiloxanes. Structure"
a
x=2-5, 13, 38, respectively.
Ref.
38
2 Main Chain Liquid Crystalline Semiflexible Polymers
Table 11. Representation of some typical structures of poly(paminoester)s and poly(p-thioester)s. Structure
4
Ref.
51
CH2CH2CO-[ Mesogenic Group
51
m
1OCCH2CH2-NUN] [105, 1061
4
n v
0
1
Y
II
CH2CH2CO-[ Mesogenic Group
1&CH~CH~-N-(CH&-p+ CH,
Mesogenic Group
:
C
H
of the platinum divinyltetramethyldisiloxane. The structures of some representative examples of LC polysiloxanes are collected in Table 10.
2.2.7 Poly(P-aminoester)s and Poly(P4hioester)s These functional polymers were prepared, according to Fig. 11, by reacting a,w-diacrylates containing different mesogenic groups with secondary bisamines, such as piperidine, 2-methylpiperidine, 1,6-dime-
CH, =
e
thyl-hexamethylenediamine or a,w-bisthi01s at room temperature in 1,4-dioxane according to a Michael-type reaction. When bisamines are employed no catalyst is necessary, whereas in the case of bisthiols the use of a catalytic amount of triethylamine is necessary to bring the reaction to completion. The structures of some representative examples of LC poly(p-aminoester)s and poly(Pthioester)s are collected in Table 11.
2.3
Structure - Liquid Crystalline Property Correlations
39
0
0 H H&=HCC
II
a"'
OCCH=CH2
+
HS(CH2),SH
lI
I
Figure 11. Synthesis of LC poly(paminoester)s and poly(Pthioester)s by the polycondensation reaction between a mesogenic diacrylic monomer and bissecondary amines and bisthiols, respectively.
A
Spherical molecule
V
n-n-0 U -
V
Rodlie molecule
Platelike molecule
Figure 12. Schematic representation of different molecular shapes.
2.3 Structure - Liquid Crystalline Property Correlations Although liquid crystallinity is a phase characteristic and not a molecular property, some structural features are well assessed as being conductive to liquid crystal phase formation. Common to all these substances is the asymmetry of molecular shape that produces orientation-dependent interac-
tions both of a hard (self-exclusion) or soft (anisotropy of polarizability) nature. This deviance from spherical symmetry results in two main classes of structurally different liquid crystals, that is, rodlike and platelike molecules (Fig. 12). General concepts on structure and liquidcrystalline properties in mesomorphic materials based on platelike mesogens are out of the scope of the present chapter and will be treated in other sections of this handbook. The most common liquid crystal substances are those with an elongated, cylinder-like shape. These liquid crystals are referred to as calamitic. Their anisometric features can be measured by the aspect ratio x = L/d where L and d are the length and the diameter of the molecule, respectively. The higher the aspect ratio, the higher the tendency of the molecules to arrange in an ordered assembly at temperatures higher than the melting point, because of the establish-
40
2 Main Chain Liquid Crystalline Semiflexible Polymers
ment of orientation-dependent interactions of a hard (self-exclusion) nature. In the case of low molar mass liquid crystals, the overall phase transition behavior depends on a critical balance between the soft and hard interactions. In contrast, for liquid-crystalline polymers, the orientationdependent interactions of a hard-nature are the prevailing ones [ 112, 1131 and the soft interactions can only affect fine details of the molecular and structural organization within the liquid-crystalline mesophase. Depending upon the specific molecular structure, several different types of liquidcrystalline phases or mesophases may occur. The mesophases generated by rodlike molecules usually feature long range uniaxial ordering, which includes both nematic and twisted nematic or cholesteric, or a two-dimensional layered ordering, which includes the low order smectics (smectic A and smectic C) or a three-dimensional crystal-like ordering for the high order smectics (smectics B, E, F, etc.). From a purely structural viewpoint, a few basic elements can be identified as the fundamental constituents of main chain liquid crystalline polymers, as illustrated in Fig. I . The essential building blocks which constitute the molecular architecture are the rigid mesogenic groups and the spacers, including also the relevant linking groups. The effects of each individual constituent of the polymer unit, including the molar mass and the molar mass distribution, will be described briefly in the following sections.
tions, and their aspect ratio. Indeed, the rodlike or cylinderlike part of the molecule must be substantially rigid, even at high temperature, and accordingly it is usually constituted of a sequence of para-linked aromatic (or cycloaliphatic) rings joined together by a rigid link which maintains the linear alignment of the aromatic groups. Linking groups such as imino, azo, azoxy, ester, and trans-vinylene groups, etc., as well as a direct link, such as a biphenyl and terphenyl units, are frequently employed since they restrict the freedom of rotation. In addition, linking groups containing multiple bonds can conjugate with phenylene rings, enhancing the anisotropic polarizability and introducing dipole moments which impart orientation-dependent interactions of a soft nature. The ester and amide groups are also effective, since resonance interactions confer a double bond character at the link, thereby restricting rotation from the trans to cis conformer. In all the above examples, the extended nature of the mesogenic group is maintained through the linearity and rigidity of its constituents. Flexible rodlike mesogens are also introduced when the linking unit consists of ethane or methyleneoxy units. Although these groups are quite flexible, they are able to preserve the colinearity of the aromatic groups through the adoption of the anti conformation (l),which is in dynamic equilibrium with the less preferred gauche conformation.
2
2.3.1 Mesogenic Group Effects
1
Anti
The structural features of the mesogenic groups affect the onset, nature, and stability of the liquid-crystalline mesophases through their rigidity, their polarizability along either the main or minor axis direc-
X= CH,, 0
Gauche
It is practically impossible to draw definite correlations between these structural
2.3
Structure-Liquid Crystalline Property Correlations
Table 12. Representation of some typical structures of
,"
polyesters containing the flexible spacer - (CH,) and their melting and isotropization temperatures.
?
-
?
co-
-0c
P
154
160
256
311
210
-
225
245
200
-
216
265
118
163
196
218
0 II
-0c
a-
0 n -0c
0 1 I
N=N
N=N+OC-
El-
9
N = p G & O 0 0 N=PGCX!0
H, H C N=?&-OC-
9
0
0c-
0
features and their specific effects. However, some interesting correlations can be observed from the comparison of the mesophase behavior of various polymers in which different mesogenic units are connected by the same flexible spacer, that is, a polymethylene chain consisting of ten methylene groups. The structures of the mesogenic groups and the relevant phase transition data are collected in Table 12. The aspect ratio of the mesogenic group has a strong influence on the both the melting and isotropization temperatures. In particular,
41
both transition temperatures increase as the axial ratio increases, as clearly observed for the biphenyl and terphenyl units, in which the introduction of a phenylene unit increases the mesophase stability by 150 "C, whereas the melting temperature increases by about 100 "C. This effect produces an increase in the mesophase existence range on increasing the aspect ratio of the mesogenic unit. The effect of reversing the ester group can be evidenced by comparing the polymers containing the azo and the azoxy groups as the bridging group, respectively. In the case of carboxy-terminated flexible spacers, polymers with stable and persistent mesophases are obtained. In contrast, no mesophase can be seen in polymer samples with the reverse ester orientation. This effect can derive from both a polarizability decrease of the mesogenic group and the introduction of strong dipoles perpendicular to the major axis of the mesogenic groups. The introduction of methyl substituents on the mesogenic group decreases both the melting and the isotropization temperatures, because they prevent the mesogenic groups from close packing by steric hindrance. In addition, the position of the methyl groups on the mesogen plays an important role in determining the mesophase stability. The depression in LC mesophase forming tendency is maximum when the methyl groups are placed in adjacent positions on the mesogen, because they can cause a steric effect that results in a twisting of the molecular structure and in a reduction of the collinear disposition of the phenyl groups, thus substantially decreasing the inherent propensity of the mesogenic groups to liquid crystal phase formation. The effect of substituents other than the methyl group on the mesophase behavior was investigated in very few studies. In general, it appears that small substituents such as Br or C1 do not substantially affect the
42
2 Main Chain Liquid Crystalline Semiflexible Polymers
isotropization temperature, but reduce the melting temperature, thus increasing the mesophase's existence range. This effect was believed to derive from two conflicting tendencies, that is, from the decrease in the mesophase stability due to the steric interactions that reduce the packing efficiency, and from the increased polar interactions which should provide an additional stabilization to the mesophase. Indeed, the introduction of very polar substituents such as NO2- or -CN decreases both the melting and isotropization temperatures, but the decrease in the latter is less pronounced, as in the case of the methyl group whose size is similar to that of the nitro or cyano groups. Also, in this case, the occurrence of polar interactions was claimed to be responsible for the partial mesophase stabilization. The effect of the introduction of lateral alkyl chains of variable length on the mesogenic group was systematically studied for a polyester [70] with structure 2.
With respect to the unsubstituted polymer, the introduction of a methyl group reduces both the melting and isotropization temperatures by about 80 "C.An increase in the alkyl chain further reduces the transition temperatures, which gradually level off and reach their saturation values corresponding to the polymer homolog with n = 5.
2.3.2 Flexible Spacer Effects The regular insertion of segments of flexible molecules to separate the mesogenic units along the polymer chain preserves the chemical periodicity of the molecule, al-
though the repeat distance is increased. The flexible spacers provide extra flexibility to the polymer backbone; however, they also impart an effective decrease of the overall aspect ratio of the chain. There is nowadays a huge amount of evidence that the spacer segments do not merely dilute the mesogens, but rather they play a key role in affecting the global thermotropic behavior as far as the stability and order of the mesophases are concerned. Due to the commercial availability of a great variety of aliphatic diols, dichlorides, and diacids with different chain lengths, most typical spacer segments consist of flexible polymethylene -(CH2)n- sequences of varying length (n). Spacer segments other than polymethylene have also been introduced into liquid-cryst a l h e poIymers. Polyethylene oxide segments extend the spacer length greatly, up to as much as -(CH,CH,0)8- without completely destroying the mesophase [40]. Polysiloxane spacers appear to be more effective per unit length in reducing the transition temperatures, on account of their greater flexibility and larger diameter. Less frequently, functional spacers such as diamino-alkylene [ 1061 or disulfidealkylene segments [ 1091 have been used. Specific reactions performed on the reactive sites of both spacers allow for the chemical modification of liquid-crystalline polymers [1141. The possibility of establishing correlations between the liquid-crystalline behavior, as indicated by the transition temperatures, and the nature and relative properties of the mesogenic and flexible sequences is, therefore, a central issue in designing and developing liquid-crystalline polymers of this type.
2.3.2.1 Polymethylene Spacers The effect of the polymethylene flexible spacer on the liquid-crystalline behavior of
2.3
Structure-Liquid Crystalline Property Correlations
43
260 Ti’
”‘
240
220 200
180 160 140
Figure 14. Limit conformations of odd-numbered and even-numbered flexible spacers. Figure 13. Variation of the isotropization temperature Ti with the flexible spacer length n ( 0 with thermal decomposition).
main chain polymers is well assessed in a number of publications. Both T, and Ti decrease significantly with increasing length of the flexible spacer, and Ti normally decreases in a zig-zag fashion in homologous series in which the spacer length is regularly increased [63]. By this so-called oddeven effect, Ti tends to be higher when there is an even number n of methylene groups in the spacer, but this oscillation is attenuated on ascending the series. This effect is exemplified in Fig. 13 for nematic polymers of structures 3.
9
9
0C(CH2),C+
3
Evidently, the even members allow for a more favorable packing of the polymer chains in the mesophase; this results in higher transition temperatures. This effect is best understood by assuming the conformation of the methylene spacer to be all trans. This conformation has the lowest energy, and there is considerable evidence that it is in fact a very likely conformation of these spacers in both liquid-crystalline and crystalline phases. The conformational distribution of a linear polymethylene chain, when
connected to a rigid group at both ends, is strongly dependent on its number ( n ) of methylenes. In a nematic phase, the more extended conformers are selectively preferred for steric packing reasons. This preference results in a change of their statistical weight in the anisotropic phase with respect to the isotropic one [ I 1.51. An even-numbered polymethylene spacer possesses a set of low energy trans conformers that force the rigid units to adopt a collinear disposition. In contrast, an odd-numbered spacer displays a relatively small fraction of trans conformers, which places the two mesogenic groups in angled orientations disfavoring the ordering of a nematic phase (Fig. 14). This basic difference between the configurational [ 1161 characters of the extended conformers of the flexible spacer determines the largely oscillating transition temperatures and can also explain the higher order assumed by the even members. Indeed, the clearing enthalpy and entropy (AHi and ASi) follow a zig-zag rising trend with increasing n, the even members exhibiting consistently greater values (Fig. 1.5). The increase in ASi is a manifestation of the greater flexibility and conformational freedom of the polymer chain in the isotropic phase than in the anisotropic one, and confirms the participation of the flexible spacer in the ordering process in such a way that even members are endowed with great-
44
2 Main Chain Liquid Crystalline Semiflexible Polymers 30
20
10
15
5
Figure 15. Dependence of the clearing enthalpy A H l and the entropy ASl on the flexible spacer length n.
er stability (Ti) and order (AS,). Order parameter measurements have detailed these general conclusions for particular classes of main-chain polymers [ 171. Whereas the above trend is well documented for the nematic - isotropic transition, unequivocal experimental evidence is not available for the smectic - nematic transition of semi-rigid polymers [ 1171. Correlations of the thermodynamic parameters (TSm-N,AHSmpN) with the structure of the polymer repeat unit must take into account very fine structural details (4).
portant to recognize that the oxyethylene unit comprises three atoms and, accordingly, comparisons are only possible within polymers with the same spacer length. Comparing polymers containing poly(ethy1ene oxide) spacers with analogous polymers containing polymethylene spacers with the same number of atoms in the spacers, lower isotropization temperatures and enthalpy and entropy changes are observed, unless the mesogenic group is so effective at dictating the phase transition behavior that the flexible spacers play a very minor role in affecting the mesophase onset and stability. This result clearly indicates that the poly(ethy1ene oxide) spacers have an inherently low mesogenic propensity relative to the polymethylene segments, which is in agreement with the high energy and low probability of existence associated with the extended conformers of poly(ethy1ene oxide) due to the preference of the gauche over the trans state relative to the rotational state around the C-C bond [118]. The effect of the “parity” of the poly(ethy1ene oxide) spacer was studied for a polymer system with structure 5 , in which the two mesogenic p-oxybenzoate dyads are interconnected by both polymethylene and poly(ethy1ene oxide) spacers in a regularly alternating fashion. This polymer system is particularly interesting because independent varia-
0
0
0
0
co
C-(OCH,CHdn-OC
oc
n
U
n
2.3.2.2 Poly(ethy1ene oxide) Spacers Various mesogenic groups were connected through oligomers of poly(ethy1ene oxide), even with a polydisperse distribution of chain lengths. However, to understand the individual effect exerted by the poly(ethylene oxide) spacers it is necessary to employ monodisperse molecular species. It is im-
U
tions are allowed in both the chemically different spacers, with n between 2 and 4 and m between 5 and 10, thus elicidating the distinct effects of each spacer. Within each series, keeping n constant, the isotropization temperatures varied with a typical even odd effect in which the even member possessed the higher temperature. Interestingly, the odd members are characterized by an
2.3
Structure-Liquid Crystalline Property Correlations
20
A Sil Jmol-’K
I
15 10 5
0
4
5
6
7
8
g r n 10
11
Figure 16. Dependence of the nematic isotropic transition entropy AS, on the flexible spacer length rn: 0,W n=2; 0 , n=3; A, A n = 4 .
increase in the isotropization temperature as the series is ascended, whereas the isotropization temperature decreases for the even members. This dual behavior derives from two conflicting tendencies, that is, the usual decrease due to the increased conformational flexibility of the macromolecular chain on increasing the length of the flexible spacers and the increase in the transition temperature due to a better packing possibility of the mesogenic groups through the establishment of favorable conformational states in the flexible spacer, which enhances the anisotropy in the molecular interactions. Figure 16 shows collectively the trends of the phase transition entropies of all the polymer samples as a function of the number of methylene units in the polymethylene chain. Within each series, the isotropization entropies for homologs with the same parity lie on smooth curves which increase with increasing spacer length, even members giving the upper curve. The isotropization entropies of the members of the series with n = 2 and 4 containing the same number rn of methylene units are very similar to each other, but significantly lower than those of the corresponding members of the series with n = 3. In addition, the slope of the iso-
45
tropization entropy of the even members of the series with n= 3 is substantially higher than that of the ones of the other series, irrespective of the parity of the polymethylene spacer. This indicates that in the case of the even members of the series with n= 3 the most favourable combination of the conformational properties of both flexible spacers occurs. In fact both spacers are of “even” parity and accordingly both can propagate the orientation correlation from a rigid group to the successive one. As a consequence, better collinearity and packing of the mesogens in the nematic state is realized. In contrast, the presence of an “odd” element in the macromolecular chain seems to attenuate the cooperativity between adjacent units.
2.3.2.3
Polysiloxane Spacers
Polysiloxane spacers are much more flexible than polymethylene and poly(ethy1ene oxide) spacers. In addition, the bulky structure and irregular conformation of the siloxane unit reduce the effective chain packing efficiency and interchain forces. Accordingly, lower transition temperatures and enthalpies are usually observed in polymers with polysiloxane spacers with respect to analogous polymers based on the more conventional polymethylene and poly(ethy1ene oxide) spacers. However, the introduction of polysiloxane spacers decreases the melting temperature and strongly inhibits the crystallization propensity of the polymeric materials, thus producing polymers with liquid-crystalline phases extending over a wide temperature region from the glass transition temperature, sometimes located below room temperature, to the isotropization temperature.
46
2
Main Chain Liquid Crystalline Semiflexible Polymers
The introduction of substituents such as methyl or ethyl groups in the flexible spacers decreases the steric packing efficiency of the chain molecules in the liquidcrystalline state and produces a perturbation of the conformational characteristics of the flexible spacer which partly loses its ability of propagating the orientation correlation from a rigid group to the successive one. As a result, a decrease of the isotropization temperature with respect to the unsubstituted polymer sample is usually observed, accompanied by a decrease in the propensity of giving rise to smectic phases. However, the steric packing efficiency in the solid state is even more reduced and accordingly, polymers with a very low, if any, melting temperature are observed, thus leading to mesophases extending over wide temperature regions. Another important effect associated with the introduction of lateral substituents in the spacer is the creation of chiral groups along 0
0
OC(CH2),k)f II
0
0
?
P
2.3.3 Spacer Substituent Effects
co
oc
-C
8 CO-R-0 6
CH;
CH,
nize that within these two polymer systems, the basic mesogenic behavior is independent of the enantiomeric excess which, however, affects fine details of the mesophase structure and physical properties.
2.3.4 Copolymerization Effects Copolymerization is a very effective way of tailoring phase transitions and tuning the properties of desired levels. In principle two or more flexible spacers can be inserted along the polymer chain, keeping the mesogenic group constant or, alternatively, two or more mesogenic groups can be inserted keeping the flexible spacer constant. Concerning the former case, a complete study was performed on the statistical copolymer samples given by 7. 0
0
II 0C(cH2),b0
8
7
X
the polymer chain. When a prevalent chirality is present, the nematic mesophase becomes cholesteric and the smectic mesophases are turned into their chiral counterparts. The influence on the mesomorphic properties of polymer structural parameters, for example, the length and nature of the polymer, the substituent’s structure, the structural isomerism enantiomeric excess, e. g., of the chiral spacer segment, has been elucidated for the polymer systems (6) [42-461 based on the 4,4’-terephthaloyldioxydibenzoyl mesogen. It is important to recog-
It was found that copolymerization lowers the crystal -mesophase transition temperature and increases the mesophase temperature range. This effect becomes progressively more pronounced as the difference in length of the flexible spacers increases. In addition, spacers of comparable length but different parity are not compliant with the steric-packing requirements of the crystalline solid state, thus decreasing the crystallization propensity of the copolymer with respect to the corresponding homopolymers. However, the nematic and smectic mesophases can accommodate different
2.3
Structure-Liquid Crystalline Property Correlations
47
units with minimal disruption of the mesophase structure provided that the two flexible spacers are not totally different. Accordingly, for both mesophases the transition temperatures and corresponding enthalpy changes are the same as the weightaveraged values of the similar parameters of the parent homopolymers. In this context, it is worth mentioning that by appropriate adjustment of the gross composition of random copolymers comprising a chiral spacer with prevalent chirality with an achiral one, such as, for example, in copolymer system 8, it is possible to produce chi-
these polymers [119- 1251. One reason for this is certainly related to the intrinsic experimental difficulty of obtaining thermotropic main chain polymers with predetermined molar mass and narrow molar mass distribution. This task has mainly been pursued through a time-consuming fractionation technique using conventional chromatographic techniques, e. g., for polyesters with structure 9. Such polymer samples ( M , = 2800-37 200; M,/M,= 1.27 - 1.84) form a nematic phase, and both T , and Ti increase, steeply at first
ral liquid crystalline materials with precisely predetermined mesophase sequence and tunable properties. Also, for copolymers consisting of either two different mesogenic units, or a mesogenic unit and a nonmesogenic rigid one, interconnected by a flexible spacer, a depression of the crystallization tendency is observed, whereas the liquid-crystalline transitions are affected very little, in agreement with substantial thermodynamic control of the transitions involving liquid-crystalline mesophases.
and then more gradually, with increasing molar mass up to a saturation value for M , = 10 000. Beyond this value, the transition temperatures remain essentially unaffected (Fig. 17). The increase in Ti with molar mass is more pronounced than for T,, and the mesophasic range is consistently wide for samples with greater M , . A quite similar dependence on M , is observed for the ASi, which increases from 1 1.7 J mol-' K-' to a saturation value of 17.5 J mol-' K-I. Such an increase in the thermodynamic phase transition parameters with molar mass is ascribed to an increased cooperativity between distant units belonging to the same macromolecular chain, as mediated by the orientational field of the nematic phase [ 124, 1251. This cooperative ordering of the rigid groups enhances the overall rigidity of the polymer chain and consequently imparts an increasing first-order character to the
2.3.5 Molar Mass and Molar Mass Distribution Effects Comparatively little attention has been devoted to the elucidation of the dependence of the phase transition parameters on the molar mass and molar mass distribution of
48
Main Chain Liquid Crystalline Semiflexible Polymers
2 160I
I
I
.I 150
a c
130;
a N
-I
m
I
10
I
I
20
I
l
30
l
40
Mn* 1 0
Figure 17. The melting ( 0 ) and nematic-isotropic (m) temperatures as functions of the molar mass.
isotropization transition. While an M , of 10 000 seems to be a very general value for main chain nematic polymers at which both Ti and ASi saturate, the specific responses of these transition parameters in the region of lower M, values cannot be anticipated. Therefore, samples of sufficiently high M , (and low M,/M,), beyond the saturation limit, must be considered when making reliable comparisons of the transitional behaviors of different polymers. The above-described behavior is well accepted and is considered a general one, although there are some interesting exceptions. For example, within the homologous series of polyesters with the general structure [ 1261 given by 10, and marked CrnCn,
and the nematic -isotropic transitions of these polymers is in agreement with the above-described general trend. In complete contrast, the stability of the smectic mesophase of polymers C,C, and C5C7 increases, whereas that of their nematic mesophase decreases with increasing number-average molar mass. This is particularly evident for the former polymer, for which the nematic mesophase changes from enantiotropic to monotropic with increasing number-average molar mass, and then disappears on reaching molar mass values of about M , = 7000 g mol-'. In addition, a wide molar mass distribution tends to increase the stability of the nematic mesophase. These effects clearly indicate that the increase in the molar mass tends to stabilize specific conformations of the flexible spacers which create an overall polymer chain configuration that is no longer compatible with the onset of a nematic mesophase, but is highly compatible with a smectic mesophase. This unconventional behavior of polymers C,C, and C5C7 was supposed to be mainly related to the short length and odd parity of both the flexible spacers, which should place the mesogenic groups in angled orientations unfavorable to the onset of nematic ordering. The conformational selection ap-
L
where rn and n identify the number of methylene units in both the polymethylene spacers; the phase transition parameters relevant to both the smectic C-nematic and the nematic - isotropic transitions of polymer samples C7C7 and C7C,, increase with the number average molar mass up to the limiting value of M,= 6000 - 8000 g mol-' and then remain constant. Accordingly, the molar mass dependence of the phase transition parameters of both the smectic C -nematic
pears to be more and more effective as the molar mass increases and the molar mass distribution narrows. Biphasic and segregation phenomena [ 127 - 1311 at the nematic-to-isotropic transition reveal the multicomponent nature of the main-chain polymers as due to the wide molar mass distribution. The thermodynamic width of the biphasic gap was delineated by annealing the samples inside the apparent biphasic gap, giving a result
2.4
of about 7-15°C. The use of polymer “blends” with a specifically designed molar mass and degree of polydispersity was found to be very effective in elucidating this behavior. Samples with high molar mass ( M , > 20 000 g mol-I) and narrow molar mass distribution ( M J M , < 1.5) and samples with low molar mass ( M , < 6000 g mol-l) and narrow or relatively wide molar mass distribution (M,/M,< 1.6) do not show nematic to isotropic biphase separation. In contrast, samples with intermediate molar mass ( M , = 8000 - 15 000 g mol-’) and relatively narrow or wide molar mass distributions ( M J M , = 1.7-2.4) show nematic isotropic biphase separation. Therefore, the phase segregation behavior appears to be due to the effects of the molar mass and the molar mass distribution of the polymer. The location of the biphasic region depends on the number-average molar mass, whereas the width of the biphasic region and the separation between the isotropization peak temperatures of the biphase demixed components depend on the width of the molar mass distribution.
2.4 References S . L. Kwolek (DuPont), U. S. Patent 3,600,350, 1971. S. G. Cottis, J. Economy, B. E. Nowak (Carborandum), U. S . Patent 3,637,595, 1972. G. W. Calundann (Celanese), U. S. Patent 4,161,470, 1979. W. J. Jackson, Jr., F. Kuhfuss, J. Polym. Sci., Polym. Chem. Ed. 1976,14,2043. G. W. Calundann, M. Jaffe in Proc. of the Robert A. Welch Conference on Chemical Research XXVI. Synthetic Polymers (Houston, Texas, November 1982) 1982, p. 247. M. Ballauff, Angew. Chem. 1988, 28, 253. M. G. Northolt, D. J. Sikkema, Adv. Polym. Sci. 1990, 98, 118. C. L. Jackson, M. T. Shaw, Inf. Muter: Rev. 1991, 36, 165. V. Krone, B. Reck, H. Ringsdorf, presented at 16. Freiburger Arbeitstagung Flussigkristalle, Freiburg 1986.
References
49
[lo] W. Kreuder, H. Ringsdorf, P. Tschirner, Makromol. Chem., Rapid. Commun. 1985, 6, 367. A. Blumstein (Ed.), Liquid Crystalline Order in Polymers, Academic, New York, 1978. A. Ciferri, W. R. Krigbaum, R. B. Mayer, Polymer Liquid Crystals, Academic, New York 1982. C. K. Ober, J. I. Jin, R. W. Lenz, Adv. Polym. Sci. 1984, 59, 104. M. Gordon, N. A. Plate (Eds.) Advances in Polymer Science, Vols. 59 -61, Springer, Berlin 1984. A. Blumstein (Ed.), Polymeric Liquid Cryslals, Plenum, New York 1985. L. L. Chapoy (Ed.), Recent Advances in Liquid Crystalline Polymers, Elsevier, London 1986. M. Jaffe in Encyclopedia of Polymer Science and Engineering, Vol. 7 (Eds.: J. I. Kroschwitz, A. Salvatore, A. Klingsberg, R. Piccininni), Wiley, New York, 1987, p. 699. C. Noel, Makromol. Chem. Macromol. Symp. 1988,22, 95. R. A. Weiss, C. K. Ober (Eds.) Liquid Crystalline Polymers, American Chemical Society, Washington 1990. W. Brostow, Polymer 1990,31, 979. J. Franek, Z. J. Jedlinski, J. Majnusz in Hundbook ef Polymer Synthesis (Ed.: H. R. Kricheldorf), Marcel Dekker, New York 1992. P. Magagnini in Thermotropic Liquid Crystal Polymer Blends (Ed.: F. P. la Mantia), Technomic, Basel 1993, p. 1. W. J. Jackson, H. F. Kuhfuss, J. Polym. Sci., Polym. Chem. Ed. 1976, 14,2043. D. Van Luten, L. Strzelecki, Eur. Polym. J. 1980, 16, 303. J. Asrar, H. Toriumi, J. Watanabe, W. R. Krigbaum, A. Ciferri, J. Polym. Sci., Polym. Phys. Ed. 1983,21, 1119. W. R. Krigbaum, J. Asrar, H. Toriumi, A. Ciferri, J. Preston, J. Polym. Sci., Polym. Lett. Ed. 1982, 20, 109. W. R. Krigbaum, J. Watanabe, Polymer 1983, 24, 1299. P. Meurisse, C. Noel, L. Monneri, B. Fayolle, BK Polym. J . 1981, 13, 55. L. Bosio, B. Fayolle, C. Friedrich, F. LauprCtre, P. Meurisse, C. Noel, J. Virlet in Liquid Crystals and Ordered Fluids (Ed.: A. C. Griffin, J. F. Johnson), Plenum, New York 1984. J. S. Moore, S . I. Stupp, Macromolecules 1988, 21, 1217. W. R. Krigbaum, R. Kotek, T. Ishikawa, H. Hakemi, Eur: Polyrn. J. 1984, 20, 225. C. K. Ober, J. I. Jin, R. W. Lenz, Polym. J . (Jpn.) 1982, 14, 9. A. C. Griffin, S . J. Havens, Mol. Cryst. Liq. Cryst. Lett. 1979, 49, 239. A . C. Griffin, S . J. Havens, J . Polym. Sci., Polym. Phys. Ed. 1981, 19, 95 1.
2 Main Chain Liquid Crystalline Semiflexible Polymers J. I. Jin, S. Antoun, C. K. Ober, R. W. Lenz, Br. Polym. J. 1980, 12, 132. A. Y. Bilibin, A. A. Shepelevskii, S. Y. Frenkel, S . S . Shorokhodov, Vysokomol.Soed. 1980, B22, 739. Q. F. Zhou, R. W. Lenz, J. Polym. Sci., Polyrn. Chem. Ed. 1983,21, 3313. Z. Jedlinksi, J. Franek, P. Kuziw, Makromol. Chem. 1986,187,2317. P. Kuziw, J. Franek, Chem. Stos. 1986, 2, 273. G. Galli, E. Chiellini, C. K. Ober, R. W. Lenz, Makromol. Chem. 1982,183,2693. A. Y. Bilibin, A. V. Tenkovtsen, 0. N. Piraner, S. S. Skorokhodov, Vysokomol.Soed. 1984,A26, 2570. E. Chiellini, G. Galli, C. Malanga, N. Spassky, Polym. Bull. (Berlin) 1983, 9, 336. E. Chiellini, P. Nieri, G. Galli, Mol. Cryst. Liq. Cryst. 1984, 113, 213. E. Chiellini, G. Galli, S. Carrozzino, S. Melone, G. Torquati, Mol. Cryst. Liq. Cryst. 1987, 146, 385. E. Chiellini, R. Po, S. Carrozzino, G. Galli, B. Gallot, Mol. Cryst. Liq. Cryst. 1990, 179, 405. E. Chiellini, G. Galli, S. Carrozzino, B. Gallot, Macromolecules 1990, 23, 2106. A. H. Al-Dujaili, A. D. Jenkins, D. R. M. Walton, Makromol. Chem., Rapid Commun. 1984,5, 33. A. Roviello, A. Sirigu, J. Polym. Sci., Polym. Lett. Ed. 1975, 13,455. A. Roviello, A. Sirigu, Eur. Polym. J . 1979, 15, 423. A. Roviello, A. Sirigu, Gazz. Chim. It. 1980,110, 403. P. Iannelli, A. Roviello, A. Sirigu, Eur. Polym. J. 1982, 18, 753. C. Carfagna, A. Roviello, S. Santagata, A. Sirigu, Makromol. Chem. 1986, 187,2123. A. Roviello, S . Santagata, A. Sirigu, Makromol. Chem., Rapid Commun. 1984,5, 141. W. J. Jackson, J. C. Morris, J. Appl. Polym. Sci., Appl. Polym. Symp. 1985,41, 307. A. Roviello, A. Sirigu, Makromol. Chem. 1984, 180, 2543. K. N. Sivaramakrishnan, A. Blumstein, S. B. Clough, R. B. Blumstein, Polym. Prep. 1979, 19(2), 190. A. Blumstein, K. N. Sivaramakrishnan, R. B. Blumstein, S. B. Clough, Polymer 1982, 23, 47. A. Roviello, S. Santagata, A. Sirigu, Makromol. Chem., Rapid Commun. 1983,4,281. M. Sato, J. Polym. Sci., Polym. Chem. Ed. 1988, 26, 2613. K. Iimura, N. Koide, R. Ohta, M. Takeda, Makromol. Chem. 1981, 182,2563. S. Vilasagar, A. Blumstein, Mol. Cryst. Liq. Cryst. Lett. 1980,56, 263.
[62] A. Blumstein, S. Vilasagar, S. Ponrathnam, S. B. Clough, R. B. Blumstein, J. Polym. Sci., Polym. Phys. Ed. 1982, 20, 877. [63] A. Blumstein, 0. Thomas, Macromolecules 1982,15, 1264. [64] D. Sek, Polym. J . (Jpn.) 1985, 17, 427. [65] H. R. Kricheldorf, R. Pakull, Macromolecules 1988, 21, 55 1. [66] H. R. Kricheldorf, R. Pakull, Polymer 1987,28, 1772. [67] H. R. Kricheldorf, R. Pakull, Mol. Cryst. Liq. Cryst. 1988, 13, 157. [68] H. R. Kricheldorf, R. Pakull, Polymer 1989,30, 659. [69] H. R. Kricheldorf, R. Pakull, S. Bruckner, Macromolecules 1988,21, 1929. [70] S. Antoun, R. W. Lenz, J. I. Jin, J. Polym. Sci., Polym. Chem. Ed. 1981,19, 1901. [71] R. W. Lenz, A. Furukawa, C. N. Wu, E. D. Atkins, Polym. Prep. 1988, 29(1), 480. [72] B. W. Jo, R. W. Lenz, J. I. Jin, Makromol. Chem. Rapid. Commun. 1982, 3, 23. [73] Q. F. Zhou, R. W. Lenz, J. Polyrn. Sci., Polyrn. Chem. Ed. 1983,21,3313. [74] J. I. Jin, B. W. Jo, R. W. Lenz, Polymer (Korea) 1982, 6, 136. 1751 E. Chiellini, G. Galli, A. S . Angeloni, D. Caretti, M. Laus, Liq. Cryst. 1989,5, 1593. [76] D. Caretti, A. S. Angeloni, M. Laus, E. Chiellini, G. Galli, Makromol. Chem. 1989,190, 1655. [77] A. Roviello, A. Sirigu, J. Polym. Sci., Polym. Lett. Ed. 1979, 15, 61. [78] J. Watanabe, W. R. Krigbaum, Macromolecules 1984, 17,2288. [79] S. Vilasagar, A . Blumstein, Mol. Cryst. Liq. Cryst. L e i . 1980, 56, 263. V. Percec, T. D. Shaffer, H. Nava, J. Polym. Sci., Polym. Lett. Ed. 1984, 22, 637. V. Percec, T. D. Shaffer in Advances in Polymer Synthesis (Eds.: B. M. Culbertson, J. M. McGrath), Plenum, New York 1984, p. 133. P. Keller, Makromol. Chem., Rapid Commun. 1985, 6, 255. V. Percec, R. Yourd, Macromolecules 1989, 22, 3229. V. Percec, R. Yourd, Makromol. Chem. 1990, 191,49. V. Percec, H. Navara, H. Jonsson, J. Polym. Sci., Polym. Chem. Ed. 1987,25, 1943. M. Tanaka, T. Nakaya, Makromol. Chem. 1986, 187,2345. M. Tanaka, T. Nakaya, J. Macromol. Sci., Chem. 1987, 24,777. M. Tanaka, T. Nakaya, Makromol. Chem. 1989, 190, 3067. C. M. Brunette, S. L. Hsu, W. J. McKnight, Macromolecules 1982, 15, 7 1. P. J. Stenhouse, E. M. Valles, S. W. Kantor, W. J. McKnight, Macromolecules 1989, 22, 1467.
2.4 [911 S. K. Pollack, D. Y. Shen, S. L. Hsu, Q. Wang, H. D. Stadham, Macromolecules 1989, 22, 551. [92] G. Smyth, E. M. Valles, S. K. Pollack, J. Grebowicz, P. J. Stenhouse, S . L. Hsu, w . J. McKnight, Macromolecules 1990,23, 3389. [93] R. Lorenz, M. Els, F. Haulena, A. Schmitz, 0. Lorenz, Angew. Makromol. Chem. 1990, 180, 51. [94] E. Chiellini, G. Galli, S. Trusendi, A. S. Angeloni, M. Laus, 0. Francescangeli, Mol. Cryst. Liq. Cryst. 1994, 243, 135. [95] A. S. Angeloni, M. Laus, E. Chiellini, G. Galli, 0. Francescangeli, B. Yang, Eur: Polym. J. 1995, 31, 253. [96] 0.Francescangeli, B. Yang, M. Laus, A. S . Angeloni, G. Galli, E. Chiellini, J. Polym. Sci., Polym. Phys. 1995, 33, 699. [97] W. Mormann, Makromol. Chem., Rapid Commun. 1988, 9, 175. [98] W. Mormann, M. Brahm, Makromol. Chem. 1989,190,631. [99] W. Mormann, S. Benadda, Makrornol. Chem., Macromol. Symp. 1991,50, 229. [loo] W. Mormann, M. Brahm, Macromolecules 1991,24, 1096. [I011 H. R. Kricheldorf, J. Awe, Makromol. Chem. 1989,190, 2579 and 2597. [lo21 H. R. Kricheldorf, J. Awe, Makromol. Chem., Rapid Commun 1988, 9,68 1. [lo31 S. M. Aharoni, Macromolecule,s 1988, 21, 1941. [lo41 C. Aquilera, J. Bartulin, B. Hisgen, H. Ringsdorf, Makromol. Chem. 1983, 184, 253. [lo51 G. Galli, M. Laus, A. S. Angeloni, P. Ferruti, E. Chiellini, Makromol. Chem., Rapid Commun. 1983, 4, 681. [lo61 A. S. Angeloni, M. Laus, C. Castellari, G. Galli, P. Ferruti, E. Chiellini, Makrornol. Chem. 1985,186,977. [I071 G. Galli, E. Chiellini, M. Laus, A. S . Angeloni, Polym. Bull. 1989, 21, 563. [lo81 A. S. Angeloni, M. Laus, E. Burgin, G. Galli, E. Chiellini, Polym. Bull. 1985, 13, 131. [I091 G. Galli, E. Chiellini, A. S. Angeloni, M. Laus, Macromolecules 1989, 22, 1120. [ I 101 A. S . Angeloni, E. Chiellini, G. Galli, M. Laus, G. Torquati, Makromol. Chem., Macromol. Symp. 1989,24, 3 1 1.
References
51
[ 11 11 M. Laus, A. S . Angeloni, V. Milanesi, G. Gal-
li, E. Chiellini, Polym. Rep. 1987, 28, 82. [ I 121 L. Onsager, Ann. N . !I Acad. Sci. 1949,51,627. [ I 131 P. J. Flory, Proc. R. Soc. 1956, A234, 73. [ I 141 E. Chiellini, G. Galli, A. S . Angeloni, M. Laus in Liquid Crystalline Polymers, ACS Symposium Series (Eds.: R. A. Weiss, C. K. Ober), 1990, p. 435. [115] D. Y. Yoon, S . Bruckner, W. Volksen, J. C. Scott, A. C. Griffin, Faraday Discuss. Chem. Soc. 1985, 79,41. [ I 161 D. Y. Yoon, S . Bruckner, Macromolecules 1982, 18, 651. [ I 171 P. Esnault, M. M. Gauthier, F. Volino, J. F. D’Allest, R. B. Blumstein, Mol. Cryst. Liq. Cryst. 1988, 157, 273. [ 1 181 P. J. Flory, Statistical Mechanics of Chain Molecules, Interscience, New York 1969. [ I 191 W. R. Krigbaum, T. Ishikawa, J. Watanabe, H. Toriumi, K. Kubota, J. Preston, J. Polym. Sci., Polym. Phys. Ed. 1983, 21, 185 1. [120] R. B. Blumstein, E. M. Stickles, M. M. Gauthier, A. Blumstein, F. Volino, Macromolecules 1984, 17, 177. [ 1211 G. Galli, E. Chiellini, A. S. Angeloni, M. Laus, Macromolecules 1989,22, 1120. [I221 M. Laus, A. S. Angeloni, G. Galli. E. Chiellini, Makromol. Chem. 1990, 191, 147. [ 1231 V. Percec, A. Keller, Macromolecules 1990,23, 4347. [I241 G. Sigaud, D. Y. Yoon, A. C. Griffin, Macromolecules 1983, 16, 875. [ 1251 P. G. de Gennes, Mol. Cryst. Liq. Cryst. Lett. 1984, 102,95. [I261 M. Laus, A. S. Angeloni, A. Spagna, G. Galli, E. Chiellini, J. Muter: Chem. 1994, 4 , 437. [127] J. F. D’Allest, P. P. Wu, A. Blumstein, R. B. Blumstein, Mol. Cryst. Liq. Cryst. Lett. 1986, 3, 103. [I281 J. F. D’Allest, P. Sixou, A. Blumstein, R. B. Blumstein, Mol. Cryst. Liq. Cryst. 1988, 157, 229. [I291 D. Y. Kim, J. F. D’Allest, A. Blumstein, R. B. Blumstein, Mol. Cryst. Liq. Cryst. 1988, 157, 253. [I301 M. Laus, D. Caretti, A. S. Angeloni, G. Galli, E. Chiellini, Macromolecules 1991, 24, 1459. [I311 M. Laus, A. S. Angeloni, G. Galli, E. Chiellini, Macronzolecules 1992, 25, 5901.
Handbook ofLiquid Crystals D. Demus, J. Goodby, G. W. Gray, H.-W. Spiess, V. Vil Copyright 0WILEY-VCH Verlag GmbH. 1998
3 Combined Liquid Crystalline Main-Chain/Side-ChainPolymers Rudolf Zentel
3.1 Introduction The starting point for the preparation of combined liquid crystalline (LC) mainchain/side-chain polymers (‘combined’ LC polymers) was merely a curiosity about what would happen if the structural properties of main-chain and side-chain polymers were combined (see Figure 1) [ 11. As a result of these syntheses a variety of new LC polymers were obtained which showed a surprisingly broad range of LC phases. Further interest in combining main-chain and side-chain structures arose from the
search for LC polymers with properties, such as the orientation in drawn fibers [ 1, 21, intermediate between those of these two classes of LC polymers. As the mesogens are oriented parallel to the fibre axis in main-chain polymers, but perpendicular to it in most side-chain polymers [1, 21, it seemed possible that the combined LC polymers would show a biaxial orientation and thus good mechanical properties in all directions. The initial results [ 11 seemed to indicate that such an orientation did indeed occur, with the side-group mesogens oriented perpendicular to the main-chain mesogens.
B
A
\
/ Figure 1. Schematic represen-
C
tation of the formation of combined LC polymers (C) from side-chain LC polymers (A) and main-chain LC polymers (B) [I].
53
3.2 Molecular Structure of Combined LC Polymers
However, this conclusion later proved to be false. Due to the interesting LC properties of combined LC polymers (i.e. broad LC phases, and the occurrence of different smectic phases and a nematic phases at different temperatures) and their intermediate nature between that of side-chain and that of mainchain polymers, a lot of research has been undertaken on these materials. Most of the research has been directed towards the preparation of cross-linkable polymers and LC elastomers [3-111 and of chiral combined LC polymers [4, 6, 7, 9, 12-16]. Quite early on, structural investigations showed [ 171 that the mesogens in the main chain and the side chain orient parallel to one another. This was confirmed by electron microscopy [18], X-ray diffraction [6,7, 13, 14, 17, 19, 201 and *H NMR spectroscopy [2 11. Following the general classification of LC side-chain polymers (Figure 2) 1221, these polymers should be classified as type 111. In these combined LC polymers, the polymer chains and mesogens orient parallel to one another to define the LC director.
3.2 Molecular Structure of Combined LC Polymers 3.2.1 Polymers with Side-Chain Mesogens Linked at the Main Chain Spacer 3.2.1.1 Achiral Combined LC Polymers The first combined LC polymers prepared by Reck and Ringsdorf [ 11 were obtained by means of a melt polycondensation, as shown in Scheme 1. The phase transitions of some of the polymers prepared in this way are collected in Figure 3. It can be seen that the LC phases observed exist mostly over a broad temperature range. Different smectic phases and a nematic phase can often exist in one material at different temperatures. The preference for smectic or nematic phases depends on whether the main chain and side chain spacers are of comparable (smectic) or strongly different lengths. The synergistic stabilization of the LC phase due to the presence of mesogens in the main chain and as side groups is best demonstrated by comparing the corresponding polymers shown in Figure 4. The clearing temperature of the
0
Figure 2. Schematic representation of different types of polymer with mesogenic side groups that form LC phases. In case I (NI), only the mesogenic side groups give rise to LC ordering. In case I1 (NII),only the polymer chain orients, whereas in case I11 (NIIl)both the main chain and the side groups orient parallel to one another [22].
0
II
II
0 (CH2jm OH + HSC20-C-CH-C-OC2H~
HO (CH& 0
I (CHzjn I 0
+ TI(O-I-Prj4
melt
I
U
0
II
"+
0
0 (CH2), 0-C-CH-C
I
(CHI),
1
-0
Scheme 1. Melt polycondensation for the preparation of combined LC polymers.
54
3 Combined Liquid Crystalline Main-ChainlSide-Chain Polymers Phase assigment of combined main chain /side chmn polyesters and wsualisation of the orientation of both types of mesogens
11
!o.
4 -
llh,;;
:;$ions
a) i'C
k 124/129 $c 153 SA 162 n 181 i 10 k 162 (n 159)i 2 klC91120n153i 6 k108/112s~131s~136n1SSi 10 g 7 2 S x 1 2 0 s ~ 139S~i541
6 6 6
2
1-6
a) k. Ciysrdline I highci oidrird m e & phLssc; 5. sincclic A 01 C. sx unidrnokied oldcxd meclic phase. sr sineclic C. SA: S ~ C E A, ~ ~n:Cnzmauc. I aolmpic
Mesophase Type strongly dependent on the ratio of main-chain and side-group Spacer Length
Smectic Phases
Figure 3. Phase assignment of combined main-chaidside-chain polyesters, and visualization of the orientation of the two types of mesogen.
Nematic Phase
G67 K158
L
ATLC = 0
I o=c
? I
HC-(CH&O
o=c
I
G9SmA71N77i ATLC= 68°C
-t;
0
G53 S 202 i
AT,c = 149°C N
N N
CN
Figure 4. Comparison of a combined main-chainhide-chain polyester with the corresponding main-chain and side-group polymalonates. ATLc, the temperature range of the LC phase.
3.2
Molecular Structure of Combined LC Polymers
B
A
Figure 5. Schematic representation of the arrangement of the main-chain (=) of combined LC polymers of type A (3.2.1) and B (3.2.2).
combined LC polymers is much higher than that of the corresponding main-chain and side-chain polymers. Obviously the temperature range of the LC phase is strongly broadened in the combined LC polymers due to the synergistic effect. For a detailed discussion of the structure-property relationships see Endres et al. [19] and Diele et al. [20]. The structural model derived from these observations is presented in Figure 5 (type A).
and side-chain (0) mesogens
a preformed polyphenolic polymer 19, 15, 16,231. The phase transitions of some of the polymers prepared in this way are summarized in Table 1. From Table I [16] the following structure-property relations can be extracted. Polymers without lateral substituents tend to be crystalline at room temperature. However, this crystallization can be suppressed
OH
3.2.1.2 Chiral Combined LC Polymers Because of the interest in chiral LC phases in general, which give rise to selective reflection of light (cholesteric phase) or ferroelectricity (chiral smectic C* phase), chiral combined LC polymers were prepared quite early on [4]. Polymers with cholesteric and chiral smectic C* phases could be prepared easily. As these polymers were synthesized using to the polycondensation process shown in Scheme 1, the chiral groups had to be selected carefully in order to prevent racemization during polycondensation [4, 7, 12, 131. In order to make more labile (referring to racemization) chiral groups accessible, a new synthetic route to combined LC polymers was developed (Scheme 2), which involves the esterification of chiral acids with
55
1
I
0
0
It
I/
+-C-CH-C-O
\ -0-,-,*
II
Scheme 2. Synthesis of chiral combined LC polymcrs by esterification of chiral acids with a preformed polyphenolic polymer. DCC, dicyclohexylcarbodiimide.
56
3 Combined Liquid Crystalline Main-Chain/Side-Chain Polymers
Table 1. Phase assignment of chiral combined main-chainlside-chain polymers.
R*
1. -C*H.C*H.C2Hs I C1
I CH3
X
Y
Phase transition temperature ("C)
H
-
K 115 SmX 124 SmC* 135 I
K 134N* 145 I
3,
4.
6
0-C*H
C 111
N
7H3
dm
+H~-CH$+CH~-C+
K
+H,
0
\CH2)6
E
c) E2
$ R
R=-OMe, -CN
SMCNG-x/y (R=-CN)
6” A
E3
Scheme 7. Structures of well-defined SGLCcoil systems synthesized via: ( A ) polymer analogous reaction: A1 [8]; A2 [421; (B) GTP [77]; (C) living cationic polymerization [78, 791; (D) living ROMP [80]; (E) anionic polymerization [83, 85-87].
80
4 Block Copolymers Containing Liquid Crystalline Segments
ples [83]. It is worth noting that polymers synthesized by GTP of the same mesogenic methacrylate [76] displayed LC properties similar to anionically produced LC polymers [83]. This major difference in LC properties obtained from different polymerization methods may be attributed to tacticity. Anionic polymerization of mesogenic monomer can produce LCPs with high syndiotactic (rr) diad content which favors a smectic mesophase compared with the atactic from radical polymerization. It is worth noting here that Watanabe et al. [85] first employed additives to promote the living anionic polymerization of mesogenic methacrylates (i.e. they added LiCl to their system to obtain narrow molecular weight distribution for the anionic polymerization of mesogenic methacrylate). We believe
1
Q
that recent advances in anionic polymerization [88, 891 will have a strong impact on the synthesis of well-defined LC -BCPs via direct anionic polymerization. Since Xerox researchers [90] reported the stable free radical polymerization (SFRP), controlled 'living' radical polymerization has drawn great attention [91]. Recently atom transfer radical polymerization (ATRP) was developed in several groups [92-961 as another type of 'controlled'/'living' radical polymerization method. Bignozzi and Ober [97] have employed this SFRP method to successfully synthesize liquid crystalline block copolymers. Their synthetic approach is shown in Scheme 8. The GPC traces shown in Fig. 7 clearly indicate the block copolymer formation occurs in a controlled/'living' fashion. We believe that controlledl'living' radical polymerization [both SFRP and recent work by Matyjaszewski's group [94] on atom transfer radical polymerization (ATRP)] will have great
-,C \=O Y
I"
I
0 (CH2)6
8
I I
a: ~oly(4-acetoxystyrene) Mn=7000, MJ Mn=l. I 8 , b: M,,=12600 (1.19)
T=135"C 1 2-dichlorobenzene FMPTS
I o\ 10-
0, ,c=o
oeC'o
10-
(CH2)6 0
I 4.5 10- 4'
ooc'o
35
40
45
5
Retention Time (min.)
0 0,
Scheme 8. Living radical polymerization [97]. (Courtesy of M. C. Bignozzi).
Figure 7.GPC traces of LC-coil BCPs synthesized by controlled/'living' radical polymerization (for structure, see Scheme 8): a: Coil precursor; b-d; LC-coil diblock copolymers with the same M, of coil block. (Courtesy of M. C. Bignozzi).
4.3
81
Side Group Liquid Crystal-Coil Diblock Copolymer Systems
Table 3. Comparison of different synthetic methods for liquid crystalline block copolymers Synthetic method for SGLC-coil BCP
Advantages
Disadvantages
Wide range of MW, narrow MWD Wide range of functional group
Small defect in LC block
Living Anionic
Good control of MW, narrow MWD
Moderate MW, limited functional groups Difficult monomer purification
GTP, ROMP Living cationic
Good control of MW, narrow MWD
Moderate MW, limited functional groups Special monomer
Living radical
Good control of MW, rather narrow MWD, tolerates functional groups
Needs more study
Direct coupling (rod-coil)
Good control of MW, narrow MWD
Limited MW, to purify
Polymer analogous reaction
impact on the synthesis of liquid crystalline block copolymers since radical polymerization is relatively simple and can tolerate water and many functional groups. In fact, we note that ATRP has been used in the synthesis of well controlled LC homopolymers by Pugh and coworkers [98]. Controlled/ ‘living’ radical polymerization will play an important role in meeting the challenge of synthesizing well-defined liquid crystalline block copolymers. A comparison of the methods of producing SGLC-coil BCPs is listed in Table 3. Besides the work by Gronski and coworkers [S, 99- 1021, polymer-analogous reactions have also been employed by Fischer et al. [42, 103- 1071(see Scheme 7 A2) and Ober et al. [43, 44, 62, 108- I l l ] (Scheme 9). Systematic morphology studies were only carried out in these two SGLC-coil systems. Table 4 lists some of the LC-BCPs synthesized and characterized by Ober et al. Well-controlled SGLC -coil structures with a wide range of molecular weight and a narrow molecular weight distribution (MWD) have been reported. Typical GPC traces are shown in Fig. 8. In our system, purification of the final block copolymers was successfully conducted with Soxhlet extraction us-
1
i) 9-BBN; ii) NaOH, HO ,,
-20°C
$343 t C H z - C H ~ b l t C H z - - F ~ C H , - C HI t t
Q
r
FHz
CHCH, I
FHZ
YHZ
OH
OH
RCOCl PyridineTTHF, 0°C
-(CHz)p(CFdqF
H,F,
7’0
c=o
R
R
p=3, q=X: p=Y. q=6:
SIH,F,-x/y
p=5. q=8: p=Y, q-10: etc.
Scheme 9. Synthetic route adopted for producing well-defined LC-coil BCPs over a wide range of molecular weight with narrow molecular weight distribution [43, 44, 108- 1 I I].
82
4
Block Copolymers Containing Liquid Crystalline Segments
Table 4. Summary of typical SGLC-coil diblock copolymers synthesized by Ober et al. Sample")
SICN5-176/55 SICN5 - 59/62 SICN5-176/78 SIC,$-13/56 SIC 105-41/53 SIC*10-41/63 SIH,F,-41/64 SIH,"Fl"-41/85
Theoretical M , (PS/LC)b'
MJM,
176K/55K 59K/62K 176K/78K 13K156K 41K153K 41K163K 4 1K/ 64K 41K/85K
1.13 1.10 1.15 1.09 1.08 1.13 1.05 -e ,
fLcc)
0.22 0.49 0.28 0.79 0.53 0.61 0.53 0.61
Thermal transitions (T)d)
Bulk Morphology
g45 g102 S, 179 I g45 glOl S A 158 I g45 glOl S, 185 I g-50 g97 S,110 S, 170 I g-50 g102 S,93 S, 169 I g55 glOl Sg 118 S i 130 I S,48 S,67 I glOl S,97 glOl S, 114 I
LC cylinder Lamella Bicontinuous Coil cylinder Coil cylinder Coil cylinder Lamella Coil cylinder
Ref.
") SI**-x/y: S stands for styrene block, I** for LC block (see scheme 9 for structures), x for the real molecular weight of styrene block, and y for M , of the LC block in kglmol. b, Absolute molecular weight of SGLC-coil block copolymers. The M , for PS block was measured by GPC calibrated with polystyrene standards. M , for LC block was calculated from 'H NMR. ') The volume fraction of LC block. d l Data from DSC second heating run. ') Not determined due to insolubility in tetrahydrofuran (THF).
ing 95% ethanol to remove small molecule mesogens [43] as indicated by GPC and TLC (thin layer chromatography). However, in Fischer's system, HPLC was used to remove the small molecular mesogens [106]. This is due to a much higher Tg (>120"C) for the LC block in their system while the Tg for our system is -50°C. The low Tgof the LC block in our system offered chain mobility for the successful removal of any small molecular mesogenic groups via simple extraction.
4.3.2 Properties of Side Group Liquid Crystal - Coil Systems 25
30
35
40
45
5 0
Retention Time (min.)
Figure 8. Typical GPC traces of SGLC-coil BCPs via polymer analogous reaction [62] (see Scheme 9 for structures): (a) polystyrene precursor; (b) styrene isoprene diblock copolymer; (c) hydroxylated styrene -isoprene diblock copolymer (unimolecular micelle in tetrahydrofuran); (d) final FLC - coil diblock copolymer.
4.3.2.1 Liquid Crystal Properties in Side Group Liquid Crystal-Coil Systems In general, SGLC-coil diblock copolymers have mesophases similar to those of the parent LC homopolymer [80, 83, 851 in contrast to the fact that a 50 wt% random co-
4.3
83
Side Group Liquid CrystalLCoil Diblock Copolymer Systems
polymer of mesogenic monomer with anonmesogenic monomer does not show any LC mesophase [76]. This is due to the unique microphase separation of SGLC -coil systems. Another interesting feature is the clearing entropy loss of SGLC - coil block copolymers [8] compared to its LC homopolymer. The entropy loss was found to be highly dependent on the composition [43, 80, 831. This observation was originally [8, 80, 831 believed to be due to the much smaller LC domain size and a rather disordered interfacial region between the LC and the coil components. Later work by Gronski et al. [ 1011 using *HNMR showed no evidence of an isotropic region at the LC-coil interface. This unusual phenomenon needs further investigation.
4.3.2.2 Phase Diagram for Side Group Liquid Crystal - Coil Systems It was Fischer and coworkers who first reported a phase diagram [42] (strictly speaking, a morphology diagram) for LC - coil diblock copolymers (Fig. 9). Similar phase diagrams were also found in another LC - coil system with a lower T, coil block [ 1061 and even LC triblock copolymer systems [ 1071. In their phase diagram, PS spheres, PS cylinder, and lamellar morphology were observed. The LC cylinder morphology, however, was not observed. Although LC
I PSI coil -coil system 0.17
0.28 0.34
'
0.62 0.66
0.77
fPS
Fischer el al.. 1994
LC-Sphere
CYL
0.2
0.4
0.6
0.8
fPS
Figure 9. Phase diagram (strictly speaking, morphology diagram) of SGLC-coil diblock copolymer systems [42,43,62] and conventional coil-coil diblock copolymers [23].
spheres were observed, the LC mesophase in the sphere microdomain only showed a nematic phase rather than the smectic mesophase which appeared in other morphologies and the LC homopolymer. This interesting phenomenon was explained by arguing that a thermodynamically stable SmA layered structure cannot be formed within a cylindrical or a spherical microdomain. However, from simple packing considerations of layered mesogens in a cylindrical microdomain, we can form two kinds of structures as indicated in Fig. 10. One is ho-
w M PS matrix
Homogeneous orientation
Homeotropic orientation
Figure 10. Packing consideration for a LC cylinder embedded on a hexagonal lattice. Mesogen 11 cylinder IMDS, homogenous packing; mesogen Icylinder IMDS, homeotropic packing. (Reprinted from [43], with permission).
84
4
Block Copolymers Containing Liquid Crystalline Segments
mogeneous packing in which all the mesogens are aligned parallel to the intermateria1 dividing surface (TMDS) and in the other they are aligned perpendicular to the IMDS and have homeotropic packing. It can be understood easily that homogeneous packing is a more energetically favorable state than the homeotropic form due to the +1 disclination which would exist in every cylinder. Similar arguments were also proposed by Walther and Finkelmann [ 131 in their review paper. Based on Fischer's results, they pointed out that it would be interesting to see whether a nematic mesophase can be observed in an LC cylinder morphology if not a smectic mesophase. Thomas and coworkers employed symmetry arguments [44, 611 for the efficient packing in SGLC-coil systems to provide a rationale for predicting morphologies. Their work suggested that not only a smectic mesophase but also a columnar mesophase could exist in a cylinder microdomain because of a match of symmetry. Ober and coworkers [43] have first observed unambiguously a LC cylinder morphology with SICN5-175/55 (see Scheme 9 for structure) in their block copolymers by both TEM (Fig. 11 D) and SAXS (small angle X-ray scattering) [43]. WAXD (wide angle X-ray diffraction) clearly showed a smectic mesophase [43]. This is the first observation of a LC cylinder morphology with a SmA mesophase packed homogeneously in the cylinder microdomain. Other morphologies such as coil cylinders, lamellae, and a bicontinuous structure were also observed with TEM (Fig. 11) [43]. The packing models derived from SAXS were shown in Fig. 12. We have observed that the structure of the famous leaning Tower of Pisa (Fig. 13) is similar to that found in the LC cylinder morphology with a smectic mesophase in the LC domain. Interestingly, the clearing transition temperature for the LC cylinder (SICN5-
176/55, Ti = 178 "C) was found to be 22 "C higher than that of the LC lamellar structure (SICN5-66/60, Ti= 156 "C; SICN5-59/62, T,= 155 "C) [43]. This is believed to be due to the confining effect of the cylinder morphology which stabilizes the mesophase within it, since the ODT of this system is much higher than Ti. Similar confining effects were reported by Dadmum and Muthukumar in their small molecular nematic liquid crystal embedded in porous glasses [ 1121 in which an increase of the N - I transition temperature was observed and attributed to the surface induced ordering. Clearly, a smectic mesophase can exist within the cylinder microdomain, therefore, it would be interesting to explore the structure of a nematic SGLC block packed within the cylinder microdomain to see whether it can be stabilized or even turned into a smectic mesophase. A bicontinuous morphology was also observed in Ober's LC -coil systems with SICNS-176/78 [43]. This is interesting because it was previously suggested by Thomas et al. [44] that this morphology might not be able to form due to the periodic LC defect presented in the microphase separated structure. Ober et al. also reported a preliminary phase diagram [43] based on their LC -coil system as shown in Fig. 9. Besides the two new morphologies observed in their system, another interesting feature is that coil cylinder (PS) morphology was observed at much smaller volume fraction of coil (PS) (see Fig. 9). It is believed that a cylinder morphology can offer a larger area per chain for effective packing of mesogenic groups. Interestingly, the LC cylinder morphology was also observed by Watanabe et al. [86] with SMCN6-17/20 (for structure, see Scheme 7 E2) since the LC weight fraction of this sample is very close to 50% for which a lamellar morphology was expect-
4.3
Side Group Liquid CrystalLCoil Diblock Copolymer Systems
85
Figure 11. TEM of SGLC-coil system by Ober et al. (Reprinted from [43], with permission). A. SICNS-41/53 (Coil Cylinder); B. SICN5-.59/26 (Lamellar); C. SICNS- 176/78 (Bicontinuous); D. SICNS-I76/SS (LC Cylinder).
ed. Other samples in the same series, such as SMCN6-5.6h.5, SMCN6-11.7/9.6, and SMCN6-31/28 all showed a lamellar morphology with a nematic phase while the mesophase in the LC cylinder was smectic. The mesogen packing was also observed to be homogeneous as indicated by SAXS which means the mesogens are packed parallel to the IMDS (Fig. 10). Tacticity would not be responsible for the phenomenon because all the polymers were prepared under the same experimental conditions.
Another interesting observation by Fischer [ 1 131 is the existence of a tetragonal structure in the coil cylinder morphology in their SGLC -coil system (PS - ChEMA) rather than the hexagonal close packing in coil-coil cylinders. This is in contrast to our observation that hexagonal packing was observed for both coil cylinders and LC cylinders (Fig. 11 A, D) [43, 611. So far, no reported novel morphology has been observed in any SGLC-coil diblock copolymersystems [8,42-44,83,105,106]
86
4
Block Copolymers Containing Liquid Crystalline Segments
domain
ra
TR
31 A
L -4
02
crjLnM = 640 A
4.4A
:
to the rather flexible nature of the main chain and the spacer in side group LC polymers. This flexibility offers the SGLC polymer chain mobility in many ways similar to conventional coil - coil BCPs.We feel that novel morphologies would most probably be observed in block copolymers with increased chain rigidity.
4.3.2.3
4L
o
4.4A
Figure 12. Packing models for lamellar and LC cylinder morphology. (Reprinted in part from [43], with permission).
Interface Thickness
Fischer directly measured the SGLC -coil interface thickness (t=2.1? 1 nm) using TEM with a selective staining technique [ 1141. This interface thickness value was also found to be independent of morphology and molecular weight which was similar to a coil-coil system. Quantitative estimation of interface thickness was given by Ober et al. using the SAXS technique. For SICN5 system, values of 1.7 nm were reported [43] which is quite similar to that of coil-coil BCPs (1.8-2.4 nm) reported by Hashimoto et al. [ 141.A much sharper interface thickness of 1.1 nm was estimated for the SIH,F, -41/64 polymer (semifluorinated system, see Scheme 9 and Table 4) [ 111 1. This is due to a much larger parameter for the semifluorinated block copolymer system since semifluorinated compounds are very immiscible with hydrocarbons.
x
Figure 13. Structure of the famous leaning Tower of Pisa. Note that the packing is also identical to that found in LC cylinder morphology with a smectic mesophase in the LC domain. The tower represents the cylinder microdomain and the columns represent mesogens which are packed in a layer structure like the SmA phase.
in contrast to the rod-coil systems [40,41, 44,53,61] in which several novel morphologies were discovered. This is mainly due
4.3.2.4 Interplay of Liquid Crystallinity and Microphase Separation In SGLC - coil systems, the LC mesophase must form within the block microdomain and adapt to the domain boundary conditions [44, 1 1 1 1. One needs to realize that the ODT in this system is much higher than the isotropic transition temperature. This domain boundary condition can act to stabilize the orientation of the mesogen in one of the domains [111] as indicated in Fig. 14. In Fig. 14, WAXD patterns are shown for sample SIH,F,-41/64 (phase transition: SmB
4.4
Applications of Liquid Crystal-Block Copolymers
Figure 14. WAXD of a SGLC-coil with semifluorinated groups: SIH,F,-41/64. The sample was simply cast from dilute solution with trifluoromethyl-benzene over a period of one week coupled with drying for 2 days at room temperature and annealing at 140°C for 4 days in a vacuum oven: (a) at 30 "C SmB phase; (b) at 60 "C, SmA phase; (c) at 80 "C, isotropic phase; (d) cooled to 30°C from isotopic phase. (Courtesy of J. Wang).
48 "C SmA 67 "C I glO1 "C). As can be seen from Fig. 14, the smectic layer is oriented as a result of its well oriented lamellar structure. The mesophase was identified as SmB at room temperature [ 110, 1111 due to the sharp arcs in the wide angle of the WAXD pattern (Fig. 14a). At 60°C in the SmA phase, the LC block underwent a SmBSmA transition so that the arcs at wide angle become very diffuse (Fig. 14b). At 80 "C the isotropic transition happened in which the sharp smectic layer diffraction disappeared (Fig. 14c). A small degree of mesogen orientation can still be seen even in the isotropic state. After cooling from the isotropic state, the orientation of the mesogen was fully recovered (Fig. 14d). This is believed to be due to the block domain boundary which stabilizes LC orientation [ 1111.
87
If the block microstructure is disordered, it is impossible to generate a LC monodomain in the SGLC-coil system at equilibrium since the LC domain must adapt to the block microdomain boundary condition [43]. However, once the block microdomain has monodomain character (e. g. via single liquid crystal processing methods such as roll-casting [24, 115, 1161, shearing [ 1171201,E-field [ 1 2 I- 1241, parallel plate [ 125, 1261, etc.), the LC mesophase will also form a monodomain in the SGLC-coil system. This LC monodomain structure will be stabilized by the block microdomain structure since the ODT is higher than its Ti. This could be very useful for processing monodomain smectic structures for reducing light scattering. Called a 'stabilized monodomain structure', it has been taken advantage of for constructing FLC displays [109, 1301 (see next section for details). It is worth pointing out that the defect structure of both the block microdomain and the LC subphase and the interaction between these two kinds of structures will be very interesting. We believe this area will draw great attention in the near future.
4.4 Applications of Liquid Crystal -Block Copolymers It is our belief that block copolymers containing LC segments are materials with novel and unencountered properties which will offer great opportunities for developing high performance materials. Here we would like to give two examples. One example is a microphase stabilized ferroelectric liquid crystal (MSFLC) [lo91 for potential flat panel display applications, while the other is a material for stable, low surface energy [ 1 101 application.
88
4
Block Copolymers Containing Liquid Crystalline Segments
When chiral mesogenic groups [ 1271 are introduced into polymer systems, chiral mesophases such as the SmC* [ 1281 or cholesteric mesophase (i. e. chiral nematic) [ 1291 will form. For LC-BCPs containing chiral mesogenic groups, the question is what the effect microdomain structure will have on these chiral mesophases. One can imagine that in the LC -BCP system, microphase separated structures will form with one LC domain and a second amorphous domain. Since the pitch of the SmC* LCPs (in the range of 1-2 pm) is much larger than the domain size which is generally on the submicron scale (10-60 nm), the pitch will then be unwound by the microdomain structure. This feature would make the SmC* mesophase in the block copolymer different from that of the homopolymer in solid bulk state. This feature was realized by Ober and coworkers and used for the production of a microphase stabilized FLC (MSFLC) which exhibited bistable switching [109]. The cholesteric mesophase also has a characteristic pitch for selective light reflection if the wavelength (and polarization) of light matches the pitch of the twisted chiral nematic mesophase [129]. The pitch value is usually in the 400- 10000 nm range which is much larger than the typical block microdomain size. Therefore, the cholesteric pitch could also be unwound by the block microdomain. It is then expected that cholesteric mesophases may not exist in SGLC coil block copolymers because selective light reflection may not be observed unless the domain size is bigger than the pitch. However, it can still be called a chiral nematic mesophase since it is nematic and also optically chiral. It is worth noting that Hammond et al. [87] have reported that cholesteric mesophases were observed with POM (polarized optical microscopy) in their chiral SGLC -coil diblock copolymers. Based on the above argument, this iden-
tification has to be confirmed by light reflection studies. One also needs to realize that the unwinding power in the LC-BCP block system may depend on the composition.
4.4.1 Microphase Stabilized Ferroelectric Liquid Crystal Displays Ferroelectric liquid crystals (FLC) are of great interest due to their fast electro-optical response which is about 1,000 times faster than conventional twisted nematic cells [131]. The geometry used is called a ‘surface stabilized FLC’ cell which utilizes a very thin gap (=2 pm) to unwind the FLC supramolecular pitch (= 1- 2 pm) since the bulk FLC materials do not show macroscopic polarization. This very thin gap, however, leads to difficulties in manufacturing large panels and very poor shock resistance. Researchers have proposed the concept of ‘microphase stabilized FLC’ [79, 109, 1301 using FLC-coil diblock copolymers for electro-optical applications as shown in Fig. 15. This concept takes advantage of ferroelectric liquid crystallinity and block copolymer microphase separation since the block
Figure 15. Concept of microphase stabilized ferroelectric liquid crystal (MSFLC). The black domain represents the coil block. Only lamellar microdomain morphology is shown in the figure. The FLC supramolecular pitch is unwound by the block microdomain [109, 1301.
4.4
Applications of Liquid Crystal-Block Copolymers
microdomain is in the submicron length scale which is much smaller than the FLC pitch. If the FLC can be oriented with a bookshelf structure, bistable switching should be possible. However, the bistable switching was not reported in Omenat et al.'s FLC-coil block copolymers [79]. We speculate that it may be due to a relatively small LC segments and small domain sizes in their system. Ober et al. have prepared a FLC-coil diblock copolymer and demonstrated its bistable switching behavior using the concept of MSFLC [109, 1301 although their system is far from practical application due to the slow response time (-0.1 s) and high voltage (V,,,=700 V/10 pm cell) needed for the bistable switching. It can be seen clearly from Fig. 15 that MSFLC cells offer many advantages. For example, a large gap of 10 pm can be used [109, 1301 instead of 2 pm used in SSFLC cells [ 1311. The polymeric nature of the LC materials offers excellent shock resistance and mechanical stability. No rubbing of the substrate is needed. We believe that practically useful electro-optical cells can be made utilizing this concept if the FLC -coil diblock copolymers are properly designed, synthesized, and processed. Since the ODT of LC block copolymers is in general much higher than the LC isotropic transition temperature T, (unless the total molecular weight is too small, i.e. a small xiV) [43, 621, all the LC mesophase transitions lie below the block microdomain boundary conditions. These microdomain boundary conditions were shown [ 1111 to stabilize the LC mesophase. The orientation of the mesogen within the block microdomain can be recovered when cooling from the isotropic state as already shown in Fig. 14. On the other hand, the texture of the LC block copolymer is very homogeneous and almost defect-free compared to that of
89
the LC homopolymer which has many defects due to a polydomain structure for the LC subdomain. The orientation of block microdomains makes it possible for the mesogens to be oriented in the LC domain parallel to the IMDS. The stabilized smectic monodomain structure, we believe, will have great potential for flat panel display application.
4.4.2 Self-healing, Stable, Low Surface Energy Materials Based on Liquid Crystal - Block Copolymers The idea of using block copolymers for low surface energy materials is to take advantage of surface segregation [132, 1331. A semifluorinated block segment offers a low surface energy material. By introducing a LC forming semifluorinated group (Scheme 9, SIH,F,), the liquid crystallinity enables the fluorinated groups to self-assemble into surface mesophases with almoct all CF, groups at the air-polymer interface and it offers extra stability for the surface due to the formation of the SmB mesophase [ 1101. Extra energy is needed to destroy the LC phase in order to reconstruct the surface. Contact angle measurements showed an advancing contact angle (with water) as high as 123 which is much higher than that of Teflon@ (= 110 "). The calculated surface tension is - 8 mNm-' for SmB mesophase using Zisman analysis [ 1101. In this process, the introduction of a flexible spacer on the LC side group is extremely important. It provides enough mobility for the polymer to self-organize into well-ordered systems with a hierarchy of ordered structures to produce stable, low energy surfaces. Block copolymers also offer processing advantages due to micelle formation since fluorinated materials are famous for poor solubility O,
90
4 Block Copolymers Containing Liquid Crystalline Segments
in organic solvents. These micelle forming abilities in semifluorinated block copolymers enable the fluorinated materials to be synthesized and processed easily in organic solvents. The polymer analogous reaction approach also offers opportunity for the rationale design of semifluorinated block copolymers to generate materials with various properties [110].
4.5 Future Work The field of liquid crystalline block copolymers has undergone great development in the last 10 years showing increasing attraction for both polymer chemists and polymer physicists. More work needs to be carried out in order to fully explore this class of materials. Novel architectures such as LCBCP with a dendritic LC block, rod-rod block with different rod diameters, and other novel structures will prove to be very interesting. Acknowledgments Funding by NSF grants DMR94-0 1845 and DMR9201845 is gratefully acknowledged. We would like to thank Professor E. L. Thomas, Dr John T. Chen, Mary Jane O’Rourke, and Dr Jianguo Wang for their collaboration and enlightening discussions. The authors also thank Dr M. C. Bignozzi for providing unpublished data.
4.6
References
[l] A. Ciferri, Prog. Polym. Sci. 1995, 20, 10811120. [2] C. K. Ober, G. Wegner, Adv. Muter: 1997, 9, 17-31. [3] H. Ringsdorf, B. Schlarb, J. Venzmer, Angew. Chem. Int. Ed. Engl. 1988,27, 113-158. [4] R. A. Weiss, C. K. Ober (Eds), Liquid-Crystalline Polymers, ACS SymposiumSeries 435. ACS, Washington DC, 1996. [5] V. V. Tsukruk, J. H. Wendorff, Trends Polym. Sci. 1995,3, 81 -89.
F. S. Bates, Science 1991, 251, 898-905. E. L. Thomas, R. L. Lescanec, Phil. Trans. R. Soc. London, Sex A 1994,348, 149- 166. J. Adams, W. Gronski, Makromol. Chem., Rapid Commun. 1989,10,553-557. B. Perly, A. Douy, B. Gallot, Makromol. Chem. 1976,177,2569-2589. B. Gallot, Prog. Polym. Sci. 1996, 21, 10351088. E. Chiellini, G. Galli, A. S. Angeloni, M. Laus, Trends Polym. Sci. 1994,2, 244-250. H. Fischer, S. Poster, Acta Polymer: 1996, 47, 413 -428. M. Walther, H. Finkelmann, Prog. Polym. Sci. 1996,21,951-979. T. Hashimoto, M. Shibayama, H. Kawai, Macromolecules 1980,13, 1237- 1247. T. Hashimoto, N. Nakamura, M. Shibayama, A. Izumi, H. Kawai, J. Macromol. Sci., Polym. Phys. 1980, B17, 389-406. E. L. Thomas, D. B. Alward, D. J. Kinning, D. C. Martin, D. L. Handlin, Jr., L. J. Fetters, Macromolecules 1986, 19, 2197 -2202. H. Hasegawa, K. Tanaka, K. Yamasaki, T. Hashimoto, Macromolecules 1987, 20, 1651 - 1662. K. Almdal, K. A. Koppi, F. S. Bates, K. Mortensen, Macromolecules 1992.25, 1743- 1751. 1191 D. A. Hajduk, P. E. Harper, S. M. Gruner, C. C. Honeker, G. Kim, E. L. Thomas, L. J. Fetters, Macromolecules 1994, 27,4063 -4075. [20] D. J. Meier, J. Polym. Sci., Part C 1969, 26, 81-98. [21] E. Helfand, Z. R. Wasserman, Macromolecules 1978,11,961-966. [22] L. Leibler, Macromolecules 1980, 13, 16021617. [23] F. S. Bates, G. H. Fredrickson, Ann. Rev. Phys. Chem. 1990,41,525-557. [24] R. J. Albalak, Polymer 1994, 4115-4119. [25] S. L. Kwolek, P. W. Morgan, J. R. Schaefgen, L. W. Gulrich, Macromolecules 1977,10, 13901396. [26] C. K. Ober, J. I. Jin, Q.-F. Zhou, R. W. Lenz,Adv. Polym. Sci. 1984,59, 102- 146. [27] C. B. McArdle (Ed.), Side Chain Liquid Crystal Polymers, Chapman & Hall, New York 1989. [28] B. Reck, H. Ringsdorf, Macromol. Chem. Rapid Commun. 1985,6,291-299. [29] Q.-F. Zhou, H.-M. Li, X.-D. Feng, Macromolecules 1987, 20, 233 -234. [30] S. R. Clingman, C. K. Ober, PMSE (Polym. Muter: Sci. Eng.) 1995, 72, 238-239. [31] Y. H. Kim, J. Am. Chem. SOC.1992,114,49474948. 1321 V. Percec, P. Chu, M. Kawasumi, Macromolecules 1994,27,4441-4453. [33] S . A. Ponomarenko, E. A. Rebrov, A. Y. Bobrovsky, N. I. Boiko, A. M. Muzafarov, V. P. Shibaev, Liq. Cryst. 1996, 21, 1- 12.
4.6 [34] D. Adam, D. Haarer, F. Closs, T. Frey, D. Funhoff, K. Siemensmeyer, P. Schuhmacher, H. Ringsdorf, Ber: Bunsenges. Phys. Chem. 1993, 97, 1366- 1370. [351 H. Ringsdorf, R. Wustefeld, E. Zerta, M. Ebert, J. H. Wendorff, Angew. Chem. Int. Ed. Engl. 1989,28,914-918. 1361 T. Kato, N. Hirota, A. Fujishima, J. M. J. Frechet, J. Polym. Sci., Part A. Polym. Chem. 1996, 34,57-62. [37] W. B. Gratzer, P. Doty, J. Am. Chem. Soc. 1963, 85, 1193-1197. [38] A. Gottschalk, H.-W. Schmidt, Liq. Cvst. 1989, 5, 1619-1627. [39] D. Pospiech, L. Haubler, H. Komber, D. Voigt, D. Jehnichen, A. Janke, A. Baier, K. Eckstein, F. Bohme, J. Appl. Polym. Sci. 1996, 62, 18191833. [40] J. T. Chen, E. L. Thomas, C. K. Ober, S . S. Hwang, Macromolecules 1995,28, 1688- 1697. [41] J. T. Chen, E. L. Thomas, C. K. Ober, G. Mao, Science 1996,273, 343-346. [42] H. Fischer, S. Poser, M. Arnold, W. Frank, Macromolecules 1994, 27, 7133-7138. 1431 G. Mao, J. Wang, S . R. Clingman, C. K. Ober, J. T. Chen, E. L. Thomas, Macromolecules 1997, 30,2556-2567. [44] E. L. Thomas, J. T. Chen, M. J. E. O’Rourke, C. K. Ober, G. Mao, Mucromol. Symp. 1997, 117, 241. [45] K. I. Winey, S . S . Patel, R. G. Larson, H. Watanabe, Macromolecules 1993,26, 2542 - 2549. [46] J.-P. Billot, A. Douy, B. Gallot, Makromol. Chem. 1977,178, 1641- 1650. [47] A. Douy, B. Gallot, Polymer 1987,28,147- 154. [48] V. S. Ananthanayanan, G. Davenport, E. R. Stimson, H. A. Scheraga, Macromolecules 1973, 6, 559-563. 1491 T. J. Deming, Adv. Mater: 1997, 9, 299-31 1. [50] T. J. Deming, J. Am. Chem. Soc. 1997, 119, 2759 - 2760. [51] M. Gervais, A. Douy, R. Erre, B. Gallot, Polymer 1987,27, 1513- 1526. [52] L. H. Radzilowski, J. L. Wu, S . I. Stupp, Macromolecules 1993, 26, 879-882. [53] L. H. Radzilowski, S . I. Stupp, Mucromolecules 1994,27,7747-7753. [54] M. Lee, N.-K. Oh, W.-C. Zin, Chem. Commun. 1996, 1787- 1788. [55] M. Lee, N.-K. Oh, J. Mater. Chem. 1996, 6, 1079- 1086. [56] M. Lee, N.-K. Oh, H.-K. Lee, W.-C. Zin, Macromolecules 1996, 29, 5567-5573. [57] X. F. Zhong, B. Francois, Makromol. Chem. Rapid Commun. 1988,9,411-416. [58] G. Widawski, M. Rawiso, B. Francois, Nature 1994,369,387-389. [59] B. Gallot, A. Douy, H. H. Hassan, Mol. Crysf. Liq. Cryst. 1987, 153, 347-356.
References
91
[601 A. Halperin, Macromolecules 1990, 23, 2724273 1. [61] J. T. Chen, PhD Thesis 1997, MIT. [621 G. Mao, PhD Thesis 1997, Cornell University. [631 M. N. Berger, J. Macromol. Sci.-Rev. Macromol. Chem. 1973, c9, 269-303. [641 A. J. Bur, L. J. Fetters, Chem. Rev. 1976, 76, 727 - 746. [65] S. M. Aharoni, Macromolecules 1979, 12, 94103.
[661 S . M. Aharoni, J. Polym. Sci., Polym. Phys. Ed. 1980, 18, 1303-1310. [67] S . M. Aharoni, Macromolecules 1981, 14,222224. [68] T. E. Pattern, B. M. Novak, J. Amer. Chem. Soc. 1991,113,5065-5066. [69] A. Halperin, Europhys. Lett. 1989,10,549-553. [70] G. Zhang, M. J. Fournier, T. L. Mason, D. A. Tirrell, Macromolecules 1992, 25, 3601 -3063. [71] A. N. Semenov, S . V. Vasilenko, Sov. Phys. JETP 1986,63,70-79. 1721 D. R. M. Williams, G. H. Fredrickson, Macromolecules 1992, 25, 3561 -3568. [73] E. Raphael, P. G. de Gennes, Makromol. Chem., Macromol. Symp. 1992, 62, 1 - 17. 1741 M. Muller, M. Schick, Macromolecules 1996, 29, 8900-8903. [75] E. Gurovich, Macromolecules, in press. 1761 W. Kreuder, 0. W. Webster, H. Ringsdorf, Makromol. Chem., Rapid Commun. 1986, 7, 5 - 13. 1771 M. Hefft, J. Springer, Makromol. Chem., Rapid Commun. 1990, 11, 397-401. [78] V. Percec, M. Lee, J. Macromol. Sci.-Pure Appl. Chem. 1992, A29,723-740. [79] A. Omenat, R. A. M. Hikmet, J. Lub, P. van der Sluis, Macromolecules 1996, 29, 6730-6736. [80] Z. Komiya, R. R. Schrock, Macromolecules 1993,26, 1387-1392. [81] R. H. Grubbs, J . Mucromol. Sci., Pure and Ayplied Chemistry 1994, 1829- 1833. [82] R. H. Grubbs, Lectures at Cornell. [83] R. Bohnert, H. Finkelmann, Macromol. Chem. Phys. 1994, 195, 689 - 700. [84] H. Schlaad, A. H. E. Muller, Macromol. Rapid Commun. 1995, 16, 399-406. [8S] M. Yamada, T. Iguchi, A. Hirao, S . Nakahama, J. Watanabe, Macromolecules 1995,28, SO-58. [86] M. Yamada, T. Iguchi, A. Hirao, S . Nakahama, J. Watanabe, PMSE (Polym. Mater. Sci. Eng.) 1997, 76, 306-307. 1871 W. Y. Zheng, P. T. Hammond, Macromol. Rapid Cornmun. 1996, 17, 813-824. [88] J.-S. Wang, H. Zhang, R. Jerome, P. Teyssie, Macromolecules 1995, 28, 1758- 1764. 1891 S. Antoun, J.-S. Wang, R. Jerome, P. Teyssie, Polymer 1996,37,5755-5759. [90] M. K. Georges, R. P. N. Veregin, P. M. Kazmaier, G. K. Hamer, Macromolecules 1993, 26, 2987-2988.
92
4 Block Copolymers Containing Liquid Crystalline Segments
[ I 121 M. D. Dadmum, M. Muthukumar, J. Chem. [91] C. J. Hawker, J. Am. Chem. Soc. 1994, 116, Phy,~.1993, 98,4850-4852. 1185-1186. [ 1131 H. Fischer, Polymer 1994,35,3786-3788. [92] M. Kato, M. Kamigaito, M. Sawamoto, T. Hi[ 1141 H. Fischer, Macromol. Rapid Commun. 1994, gashimura, Macromolecules 1995, 28, 172115,949-953. 1723. [I151 R. J. Albalak, E. L. Thomas, J. Polym. Sci., [93] J.-S. Wang, K. Matyjaszewski, J. Am. Chem. Part B, Polym. Phys. 1993, 31, 37-46. SOC.1995,117,5614-5615. [116] R. J. Albalak, E. L. Thomas, J. Polym. Sci., [94] K. Matyjaszewski, Polym. Prep. 1997, 38, Part B, Polym. Phys. 1994,32,341-350. 683-711. [ 1171 G. Hadziioannou, A. Mathis, A. Skoulios, Col[95] V. Percec, B. Barboiu, A. Neumann, J. C. Ronloid Polym. Sci. 1979,257, 136- 139. da, M. Zhao, Macromolecules 1996,29,3665[ 1 181 F. A. Morrison, H. H. Winter, Macromolecules 3668. 1989,22,3533-3540. [96] V. Percec, B. Barboiu, Macromolecules 1995, [I191 K. A. Koppi, M. Tirrell, F. S . Bates, K. Alm28,7970-7972. dal, R. H. Colby, Journal De Physique 11 1992, [97] M. C. Bignozzi, C. K. Ober, 1997. 2, 1941-1959. [98] A. Kasko, A. M. Heintz, Y. Wang, C. Pugh, [120] R. M. Kannan, J. A. Kornfield, MucromolePolym. Prep. 1997, 38, 675 -676. cules 1994,27, 1177-1186. [9Y] J. Adams, W. Gronski in Liquid-Crystalline [121] K. Amundson, E. Helfand, D. D. Davis, X. Polymers ACS Symposium Series 435 (Eds.: Quan, S. S. Patel, S . D. Smith, MacromoleR. A. Weiss, C. K. Ober) American Chemical cules 1991,24, 6546-6548. Society, Washington DC 1990, p. 174- 184. [122] K. Amundson, E. Helfand, X. Quan, S . D. [ 1001 J. Adams, J. Sanger, C. Tefehne, W. Gronski, Smith, Macromolecules 1993,26,2698-2703. Macromol. Rapid. Commun.1994,15,879-886. [123] K. Amundson, E. Helfand, X. N. Quan, S. D. [loll A. Martin, C. Tefehne, W. Gronski, Macromol. Hudson, S. D. Smith, Macromolecules 1994, Rapid Commun. 1996,17,305- 3 11. 27,6559-6570. [lo21 J. Sanger, W. Gronski, Macromol. Rapid Com[124] E. Gurovich, Phys. Rev. Lett. 1995, 74, 482mun. 1997,18,59-64. 485. [lo31 B. Zaschke, W. Frank, H. Fischer, K. Schmutz[125] Y. M. Zhang, U. Wiesner, J. Chem. Phys. 1995, ler, M. Arnold, Polym. Bull. (Berlin), 1991,27, 103,4784-4793. 1-8. [ 1261 Y. M. Zhang, U. Wiesner, H. W. Spiess, Mac[lo41 M. Arnold, S . Poser, H. Fischer, W. Frank, H. romolecules 1995, 28, 778-781. Utschick, Macrornol. Rapid Commun. 1994, [ 1271 J. W. Goodby, J.Mar. Chem. 1991,1,307 - 3 18. 15,487-496. [128] M. V. Kozlovsky, L. A. Beresnev, Phase Tran[I051 H. Fischer, S . Poser, M. Arnold, Liq. Cryst. sition 1992,40, 129- 169. 1995,503-509. [I291 Ya. S. Freidzon, E. G. Tropsha, V. P. Shibaev, [lo61 H. Fischer, S. Poser, M. Arnold, MacrornoleN. A. Plate, Macromol. Chem., Rapid Comcules 1995,28, 6957-6962. mun. 1985,6,625-629. [lo71 S. Poser, H. Fischer, M. Arnold, J. Polym. Sci., [I301 M. Brehmer, R. Zentel, G. Mao, C. K. Ober, Purr A: Polym. Chem. 1996,34, 1733- 1740. Macromol. Symp. 1997,117, 175. [108] G. Mao, S. R. Clingman, C. K. Ober, T. E. [131] N. A. Clark, S. T. Lagerwall, Appl. Phys. Lett. Long, Polym. Prepr: (Am. Chem. Soc., Div. 1980,36,899-901. Polym. Chem.), 1993, 34, 710-711. [132] S. S . Hwang, C. K. Ober, S. Perutz, D. R. Iy[ 1091 G . Mao, J. Wang, C. K. Ober, M. J. O’Rourke, engar, L. A. Schneggenburger. E. J. Kramer, E. L. Thomas, M. Brehmer, R. Zentel, Polym. Polymer 1995,36, 1321- 1325. Prep. 1997, 38, 374- 375. [133] D. R. Iyengar, S . M. Perutz, C.-A. Dai, C. K. [I101 J. Wang, G. Mao, C. K. Ober, E. J. Kramer, Ober, E. J. Kramer, Macromolecules 1996,29, Macromolecules 1997,30, 1906- 1914. 1229- 1234. [111] C.K.Ober,J.Wang,G.Mao,E.J.Kramer,J.T. Chen, E. L. Thomas, Macromol. Symp. 1997, 117, 141.
Handbook ofLiquid Crystals D. Demus, J. Goodby, G. W. Gray, H.-W. Spiess, V. Vil Copyright 0WILEY-VCH Verlag GmbH. 1998
Chapter I1 Defects and Textures in Nematic Main-Chain Liquid Crystalline Polymers Claudine Noel
1 Introduction
71. Defects and textures in LMWLCs are al-
Liquid crystalline phases show a variety of beautiful optical patterns when observed with a polarizing microscope, each characteristic of defects (local imperfections that break the order of molecular arrangement) that are peculiar to the state of molecular order prevailing in that mesophase. The first analysis of defects in low molecular weight liquid crystals (LMWLCs) was carried out very early by Lehmann [I], Friedel [2], and Grandjean [ 3 ] . The accepted classification and terminology of liquid crystals (LCs) is based almost entirely on Friedel's observations. His pioneering work introduced and defined the terms used today to describe the LC state: nematic (N), cholesteric (N") and smectic (Sm). His conclusions were supported by citations of work in other fields (e.g., X-ray diffraction, behavior in electric and magnetic fields). Although Friedel noted the possibility of smectic polymorphism, he was sceptical of its actual occurrence. Much work on the characterization and classification of smectic phases has been done in Halle by Demus, Sackmann and associates [4, 51. The work consisted essentially of the attempt to determine experimentally the number of different smectic phases by miscibility studies and to characterize them by their textures. Two relevant books with photographic illustrations have been published dealing deeply with this subject [6,
so presented in certain general articles and reviews [2, 3 , 8- 131. Although physicists have learned much about defects and textures from optical studies, from solutions to elastic equations, and from simply drawing pictures, only since 1976 have researchers 114, 151 used the topology to introduce a general scheme for classifying defects into condensed-matter physics. Much progress was made during the 1980s in elucidating the relationships between defects and structure due to the work of KlCman and others [ 14, 16- 181 who elaborated a complete theory relating defects of any dimensionality (point, lines, wall-like defects ...) in ordered systems to the generic singularities of the order parameter, via the topology of an adequate order parameter space. Of late, the properties of defects and textures in LMWLCs have been discussed on the basis of their material-dependent properties including their energetical stability and the nature of the order in the central region or core of the defect and also from the point of view of topology [19-231. As pointed out by KlCman in early 1985 [24], liquid crystalline polymers (LCPs) and LMWLCs are equivalent from the point of view of topology since they exhibit the same symmetries and the same type of order parameters. Defects are indeed characteristic breaks in local symmetry of the order parameter and can be classified according to their dimensionality and the specific sym-
94
11 Defects and Textures in Nematic Main-Chain Liquid Crystalline Polymers
metry they break. This implies that defects of a given type have equal topological stability in LCPs and LMWLCs. However, the relative occurrence of defects of different types and the arrangement of defects into various textures depend on energetic considerations. Since these are directly related to the molecular structure, LCPs are not energetically equivalent to LMWLCs. Despite the fact that main-chain liquid crystalline polymers (MCLCPs) have been studied for two decades, relatively little information is available on the influence of the structure on the properties of defects and very few conclusions of a general nature have been reached, except that the defects observed pertain to the topological classes expected [25]. In this regard, it is important to note that microscopic observations are sometimes misleading owing to the difficulty with which MCLCPs give specific textures in the liquid crystalline state. This might be due to their multiphase nature (existence of polycrystalline and amorphous material), polydispersity and the high viscosities of the liquid crystalline melts. In most cases, samples must be annealed for hours or days at a suitable temperature if equilibrium textures reminiscent of those of LMWLCs are to be seen [26-281. Also, the high melting points of MCLCPs have limited progress on their study. These temperatures are often above the operating range of the sample heaters, and in any case, the extended heating of polymer samples at such temperatures can cause degradation. The effect of structural disclinations upon texture and the origins of these in local deformations of or discontinuities in the arrangement of the molecules have been studied only in the case of nematic MCLCPs. The primacy aim of this chapter is therefore to present to the reader defects and textures in nematic MCLCPs. Some of consequences of constructing a nematic phase from long
polymeric chains will be presented. We will concentrate on understanding the role of properties such as scarcity of chain ends, chain rigidity, high degree of anisotropy in the elasticity and viscosity that clearly distinguish MCLCPs from LMWLC materials. We will consider neutral, rigid or semiflexible polymers. In electing to discuss the geometry and topology of defects in nematic MCLCPs, this chapter does differ somewhat from others [29-321 dealing with textures in both side chain (SC) and MCLCPs, which were intended as practical and useful experimental guides to the textures and classification of nematic cholesteric and smectic LCs of different polymorphic types.
2 Textural Assignments 2.1 The Basic Equations In ideal nematics, the molecules are aligned along one common direction which is referred to as the director and represented by a dimensionless unit vector n. The system is uniaxial and the states (n)and (-n) are undistinguishable since there is no polarity. In most practical circumstances, however, this ideal configuration will not be compatible with the constraints which are imposed by the limiting surfaces of the sample and/or by external fields acting on the molecules. There will be some deformation of the alignment and the orientation of the director will change continuously and in a systematic manner from point to point in the sample. The distorted state may then be described by a variable director n (Y). Three types of distortions are present simultaneously in nematics. As illustrated in Fig. 1, one can distinguish three situations which are pure splay, pure twist and pure bend. The corre-
2 Textural Assignments
a
C
Figure 1. Distortions in nematics. (a) Pure splay, (b) pure twist and (c) pure bend [19].
95
the xy plane parallel to the glass surfaces, and is not a function of z , the normal to the film. Taking the components of the director to be nx=cos @, ny=sin @, nz=O, and making the simplifying assumption that the medium is elastically isotropic (i.e. K , =K2,=K,,), Eqs. (1) and ( 3 ) reduce to:
,
F = 1 K(V(I)~ 2 ~
V2@=0 sponding elastic constants (often referred to as Frank constants) are K , K22 and K , , , respectively. The free-energy per unit volume of a deformed specimen relative to the undeformed one is given by [ 3 3 , 341.
,
F = 1/2 [ K , (V . n)2 + K,,(n . V x n)2 + K,,(n x V x n ) 2 ] (1) This is the fundamental formula of the continuum theory for nematics. The total energy of the system is: E=jFdr
(2)
Minimization of the total energy yields the conditions for equilibrium in the bulk:
(3) where we use the Cartesian tensor notation, repeated tensor indices being subject to the usual summation convention, and the comma denotes partial differentiation with respect to spatial coordinates. Distortions and defects can be interpreted in terms of the continuum theory through equations derived from the expressions of the elastic energy and the imposed boundary conditions. Solutions are known in certain simple situations. Oseen [35] has found configurations, named ‘disinclinations’ by Frank [ 3 3 ] ,or ‘disclinations’ today, which are solutions of this problem for planar samples in which the director n is confined to
The solutions of Eq. ( 5 ) are @ = S 8 + 8 , where 8=tan-’ (ylx) and 13, is a constant (0 < 8,< x). This equation describes the director configuration around the disclination, the singular line is along the z-axis and the director orientation changes by 2 n S on going round the line. In the nematic phase, the orientational order is apolar and hence S must be an integer or a half integer. Fig. 2, taken from Frank’s article, shows the molecular orientation in the neighborhood of a disclination for a few values of S. The curves represent the projection of the director field in the xy plane. S is called the strength of the disclination. The elastic energy (per unit length) associated with an isolated disclination is: E = x K S 2 1 n ( R / r , ) + E,
(6)
where R is a typical dimension of the specimen or the distance to other disclinations and r, the radius of the inner region or ‘core’ (of the order of molecular dimensions). E, is the energy of the core which is hard to calculate with any accuracy. Lines of strength *1/2 are expected to be common since the energy scale as S 2 . However, in LMWLCs the most frequent disclinations have S= 21. The reason for this is that defects with integral values of S can ‘escape into the third dimension’ [36, 371, that is the director can simply rotate to lie along the line in the central region, obviating the need for a core.
96
I1 Defects and Textures in Nematic Main-Chain Liquid Crystalline Polymers
, , , / , , \ \ * . - - - - - , , / , I l l / \ \ . - - - - - - -
,
/-r-ksa I , , ,
, / - - - - - - -
, , I , , / I-------
,,,,
, / , , - - - - - - - -
, , , , , , I ,
---------
, , , , , I , , , - - - - , , , I ! I / , ~ - - - - - , / , , I l l , , , - - - - - -
s = +1/2
s = -112
s = -1
I , ,
,,,,,,,_----..-
2‘0
b)
a)
Figure 4. Twist disclinations: the director patterns for (a) S= 1/2, O,=O and (b) S= 1, O,=O. s = +1, 0 0 = 0
s = +1, 80 = n /4
s = +1, 0 0 = X I 2
twist. Other escaped configurations exist [22]. The energy (per unit length) is:
E={
s = +3/2
s = +2
Figure 2. Director field in the neighborhood of the core of a disclination line of strength S: Wedge disclinations (after Frank [33]).
Figure 3. (a) Wedge disclination of strength +1 in a capillary: the arrangement near the center has a discontinuity and must involve a core. (b) ‘Escape’ of the disclination in the third dimension: the arrangement is continuous with no core [34].
Fig. 3 b shows the director ‘escape’ at the center of a disclination of strength S = & l in a thin capillary of radius R. The arrangement is continuous with no singular line. The deformation involves splay and bend, but no
3 n K for S = + l T C K for S = - 1
(7)
On the other hand, for the simplest arrangement where the director is everywhere radial and the deformation is pure splay, the energy is easily calculated to be:
E = nKln(R/r,)
+ E,
(8)
Then, the escaped configuration is more favorable energetically than the line as soon as R is large enough. In addition to the above-discussed disclinations, which are referred to as wedge disclinations, there are twist disclinations. The director is always parallel to the xy plane, but the axis of rotation (z-axis) is normal to the singular line (y-axis). Figure 4 shows the director patterns for (a) S=1/2, $=O and (b) S = l , $=O. qj is a linear function of the angle 8=tan-’ (z/x>.
qj = s tan-’ ( z / x ) + e,
(9)
where S is the strength of the disclination and 0, a constant. The deformation involves simultaneously splay, twist and bend. If the medium is elastically isotropic ( K , =K2 = K33=K), then the expressions derived for the wedge disclinations are applicable to the present situation. It should be noted, however, that the energy (per unit length) for the escaped configuration is now 2 n K 1 SI.
97
2 Textural Assignments
b
a
C
Figure 5. Singular points in droplets: (a) spherically symmetric radial configuration with the director normal to the surface and (b) bipolar structure with the director tangential to the surface; (c) singular points in a capillary [22].
Points defects are another class of defects in nematic LCs. They occur in droplets and can be also seen in thin capillaries (Fig. 5 ) . The total energy stored in the elastic field around a point defect grows linearly with the radius, R , of the volume enclosed:
E=KR
(10)
where K is a suitable Frank constant coefficient. Point of singularities of equal and opposite strengths attract one another and are annihilated. For LMWLCs, it has long been recognized that degenerate anchoring at the constraining surfaces may allow formation of domains of different molecular orientation, these being separated by diffuse walls or by disclinations. Energy considerations suggest that walls, rather than disclinations, may be observed if the sample thickness, h, obeys the inequality: h < 2K/W,
(11)
where K is the Frank elastic constant and W , the anchoring energy. For LMWLCs this is hard to achieve because W ,-- lop3Jmp2 leading to the need for film thickness of less than 10nm. However, for MCLCPs, although measurements of W , have not been made, it seems that walls may be somewhat
easier to achieve. If rubbing of the constraining surfaces has been used to induce a homogeneous or planar alignment (i.e., one in which the molecules lie parallel with the interface) two types of wall may be identified. In both cases a reorientation of the molecules by 180" is observed on crossing the wall. These two types of walls are termed inversion walls of the first and second kind [4] depending whether the reorientation is accomplished by rotation about an axis normal to the sample plane so that the molecules remain in that plane at all points, or about an axis within the sample plane, so that the molecules themselves rotate out of the plane and pass through the homeotropic orientation (i.e., normal to the plane) at the center of the wall.
2.2 Some Consequences of Elastic Anisotropy
,
,
We have so far assumed K , = K2 = K, =K. Real nematics are of course, elastically anisotropic. From the values of the energies of the disclinations calculated for the case where K , =K, # K2 it follows that the wedge disclinations are more stable than the twist disclinations if K2 > K =K , =K, 3 , and vice versa. Anisimov and Dzyaloshinskii [38] have shown that lines of half-integral strength may be stable against three-dimensional perturbations if the twist elastic constant K22# 1 / 2 ( K ,,+K3,). More precisely,
,
,
(1) S = +1/2 wedge disclinations are energetically favored when K2 > 1/2 ( K , + K, ,) and are stable against any out-of-plane distortion. ( 2 ) S=+1/2 twist disclinations are favored when K2*< 1 / 2 ( K ,,+K3,), but are unstable against an out-of-plane distortion.
98
I1 Defects and Textures in Nematic Main-Chain Liquid Crystalline Polymers
The effects of elastic anisotropy ( K l l # K22# K3 3 ) have been investigated by Dzyaloshinskii [39]. He found that, qualitatively, the molecular orientation close to the line is the same as in the elastically isotropic case, except for S = l . The only S = + l wedge lines that are stable versus threedimensional perturbations have a purely circular (@= 0 + d 2 ) or radial (@= 0) configuration. Cladis and KlCman 1361 have calculated the energy of the nonsingular line for the bend-twist and splay-bend cases. Whether the stable geometry is circular or radial with or without a singular core will depend on the nature and the size of the core and on the degree of anisotropy of the elastic constants. It is worth noting, however, that if singular integral lines are favored, the energy E = S2 predicts these should dissociate into two lines of half-order strength. In LMWLCs, where the differences between K , K22 and K33 are usually small, nonsingular integral lines are the most commonly observed defects. By contrast, as was first pointed out by KlCman 1251, one can infer from the arguments above that in MCLCPs, where the elastic anisotropy is large, the half-integral lines will occur more frequently than the integral ones and that these integral lines will be singular if there exist core arrangements of moderate energy. De Gennes [40] has proposed a model for the splay modulus for semiflexible polymers, based on intermolecular strain free energy, which predicts K , I = L2 (where L is the contour length of the chains). K3 is expected to be normal and K22 independent of L. Meyer [41] has estimated the magnitude of K , I by an argument based on chain end entropy and has predicted that K , should increase essentially with L. K3 strongly reflects the chain rigidity and is limited by the persistence length. K Z 2should still remain small and independent of L. Although some
,,
,
disagreement appears between the qualitative description of de Gennes’s and Meyer’s models, it is clear that splay deformation becomes difficult in a nematic composed of long molecules. Only few experimental studies have been carried out on thermotropic MCLCPs [42-451. Measurements of the elastic constants with conventional methods, particularly with the Freedericksz distortion technique, have been faced with experimental difficulties. This has been due mainly to the lack of strong surface anchoring, large viscosity and thermal degradation of polymers. Zheng-Min and KlCman 1421 have evaluated the elastic constants of the polyester:
1 from the Freedericksz’ transition. The splay elastic constant ( K , = 3 x dyn) was found to be one order of magnitude larger than in LMW nematics while the twist elastic constant ( K Z 2 =3 xlOP7dyn) and the bend elastic constant (which was more difficult to measure because of chemical degradation) exhibited more conventional Values. Volino et al. [43, 441 have used a simple NMR technique to evaluate the ratio K33/Kl for the polyesters:
,
,
‘CH,
CH/
2 where m=7, 10. They found K,,lK, I = 0.3-0.5, avalue close to that of -0.1 reported by Zheng-Min and KlCman. It should be noted, however, that the three elastic constants. of the copolyesteramide Vectra B 950@is about 100 times higher than corresponding values obtained for the above thermotropic MCLCPs [45]. To what extent these large values are related to the chemi-
2 Textural Assignments
cal structure of this polymer or to its biaxiality are unclear. Windle et al. [46, 471 have developed a model that simulates the evolution of microstructure for different ratios of the elastic constants and thus for both LMWLCs and MCLCPs. The results obtained on two dimensional lattices are in good agreement with those predicted by the continuum theory both when the elastic constants are equal (LMWLCs) and when they are different (MCLCPs). For example, the trajectories of the directors around the S=-1/2 disclinations are observed to be insensitive to the magnitudes of the elastic constants while those around the S=+1/2 disclinations undergo a marked change. In 3-dimensional models simulated using equal elastic constants, most disclinations are of the type S=+l and have point character. The reason for this is proposed to lie in the process of escape into the third dimension. By contrast, a predominance of S=+1/2 line disclinations is found when 3-dimensional simulations are performed with free boundary conditions and with the splay energy set much higher than that of bend or twist. This is due to the fact that escape in the third dimension is no longer favored. However, a number of problems have been encountered with this approach. In particular, the model needs to account correctly for both splay-splay compensation and the differentiation between splay and bend distortions in three dimensions. Quite recently, new developments of the central algorithm of this model has enabled the splay and bend components to be successfully partitioned in three dimensions, and splay-splay compensated structures to be handled properly [48]. Modeling using the new algorithm for the case where the splay constant is dominant leads to the generation of structures in which twist escaped +1 line singularities are a predominant feature. 3-Dimensional simulations
99
were also performed with planar boundary conditions at the top and bottom surfaces and with splay energy one hundred times that of bend and twist [47]. The structures were found to form layered morphology with little matching from layer to layer and a sinuous trajectory of the directors within the layers. This undulating structure is very reminiscent of the banded [49] and tight [50, 5 11 textures observed in thermotropic copolyesters.
2.3 Topology of Director Fields: Homotopy Groups and Classification of Defects 2.3.1 Uniaxial Nematics The term ‘nematic’ is derived from the Greek word ‘nema’, a thread, and refers to the presence of a large number of apparent threads in relatively thick samples. Stirring a nematic leads to a pronounced increase in their number. When the motion stops, many of them disappear and other progressively stabilize. In uniaxial nematics, the threads are of two types: either thin or thick, clearly distinguishable, both directly observable in light microscopy (Fig. 6). The thin threads are line defects of half integral strength S=?1/2, ?3/2 ... about which the director rotates by an angle of 2 n S and which are topologically stable. They have the same topology as a Mobius strip. The central region, or core, of the disclination appears as a fine thread running through the material as it scatters light strongly. The threads never show any free extremity in the interior of the liquid crystal. Topological considerations show that they have to form closed loops or end at the boundary surfaces or any interface (i.e.,
100
I1 Defects and Textures in Nematic Main-Chain Liquid Crystalline Polymers
Figure 6. Thin and thick threads in a typical low molecular weight liquid crystal: threaded texture of p-methoxybenzylidene-p-n-butylaniline(MBBA).
Figure 7. Schlieren texture of terephthalidene-bis(4-n-butylaniline) (TBHA). Integral nuclei (four brushes) and half-integral nucleus (two brushes).
between isotropic liquid and mesophase). In some cases, they are nearly perpendicular to the boundary surfaces, their ends being attached to the glass slide and coverslip, respectively. Often, however, they lie more or less horizontally, their ends being attached to the glass and the other parts moving and floating freely around in the nematic. The threads have elastic properties and may be extended and curved due to flow of the liquid. Some threads may also be stuck to the glass surfaces. As already discussed, the lines of integral strength S = + l , +2, ..., about which the director rotates by an angle 27cS are not topologically stable. The director can escape into the third dimension obviating the need for a core. In this case, the integral line shows up a blurred contrast under the polarizing microscope and thus appears thick. A half-integral line of any strength can be continuously deformed into the half-integral line of opposite strength as illustrated in [19]. So it is with integral lines. The threaded texture changes to the Schlieren texture when the lines are all perpendicular to the sample boundaries. This texture observed between crossed polarizers displays dark brushes (also called black
stripes or Schlieren) which have irregular curved shapes and correspond to extinction positions of the nematic (Fig. 7). Accordingly, the director lies either parallel or normal to the polarizer or analyzer planes, respectively. These brushes meet at point singularities on the surface of the preparation or rather, point singularities are the origins of the brushes. For the configurations S = + l , the director field forms + 27c or -2 7c disclinations and one observes point singularities with four dark brushes (‘Maltese crosses’), due to the two orientations of extinction. The configurations S=k1/2 lead to singularities with only two brushes, corresponding to +7c or -IT disclinations. Neighboring singularities which are connected by brushes have opposite signs and may attract each other. They can vanish, if their S values add up to zero. It should be noted that twist lines are visible in threaded textures while wedge lines are observed in Schlieren textures. The continuous transformation of a wedge S line into a wedge -S line involves the passage through an intermediate twist IS1 line [19]. Point defects are another class of defects in nematics. When a line, such as the wedge disclination of Fig. 3a, with S = l , escapes
2 Textural Assignments
+ +
+ +-
+
+
- +
+
Figure 8. Singular points in a capillary with normal boundary conditions.
in the third dimension, it may do it in two ways: either upwards or downwards. As shown in Fig. 8, an upwards portion and a downwards portion will be linked by a singular point (1). The classification of defects in nematics represents a straightforward example of the applications of homotopic group theory [ 14, 151. The reader is referred to reviews of the subject [19-21, 23, 521. This topological approach confirms the absence of walls, the existence of Mobius lines, the mutual annihilation of thin threads and the existence of singular points. More importantly, it shows that defects combine and merge according to the rules of multiplication of the two-element Abelian group 2,.
2.3.2 Biaxial Nematics The applicability of homotopic theory becomes much less obvious for liquid crystal phases with more complicated order parameters such as biaxial nematics and cholesterics, which are both locally defined by three directors forming a tripod. This gives rise to a description of the line singularities in terms of the quaternion group, Q . This is particularly interesting because the quaternion group Q is non-Abelian, a property that
101
has important consequences when one considers combining defects or entanglements of defects [52]. In the biaxial nematic, there are three classes of line defects involving Lrotations about each of the three inequivalent symmetry axes and a class of line defects of equivalent 2 n rotations. Because there is an orthogonal set of axes, one cannot rule out a strength 1 or 2 n singular line as one could in the uniaxial nematic. However, the continuous deformations of the director that are equivalent to ‘escape’ into the third dimension make all 2 n lines equivalent. Lines of integral strength 2 or 4 n are not topologically stable and must appear as nonsingular in biaxial nematics. It should be noted that two defects of opposite signs can be continuously transformed into one another if they belong to the same class. Similarly, a n defect and the -n defect that annihilates it must be of the same class. Since the mobility of line defects can play an important role in determining the macroscopic properties of a medium, it is important to determine whether two given line defects can or cannot cross each other. In media with Abelian fundamental groups all line defects can cross each other without the formation of a connecting line. In the non-Abelian case of the biaxial nematic, an examination of the multiplication table for the quaternion group reveals that two mobile lines can always cross without the production of a connecting line except for the case of two 71: disclinations of distinct types, which are necessarily joined after crossing by a 2 n disclination. Finally, it is worth noting that singular points are not topologically stable in biaxial nematics. All the peculiar features of the defects in biaxial nematics are of importance since the characteristics of a few thermotropic MCLCPs have been related to biaxiality [45, 51, 53-56].
102
I1 Defects and Textures in Nematic Main-Chain Liquid Crystalline Polymers
3 Defects in Nematic Main-Chain Liquid Crystalline Polymers As a rule, thin lines of strength +1/2 or singular lines of strength +1 are seen in the threaded textures of nematic thermotropic MCPs [25,57,58].Rare cases of thick lines have been observed. As we have already pointed out, the reason for this is that the elastic anisotropy is large in these systems. The disclinations with S=+1/2 were also reported to be the most abundant in the Schlieren textures of nematic copolyesters 159- 651. The first experimental studies of the defects in the nematic phase of thermotropic MCPs were reported by Klkman and coworkers [57, 581. They examined in detail the threaded texture of polyester 1 for two specific values of the degree of polymerization. The molecular weight was found to affect the threaded texture of the polyester with n=5. The thin lines (S=+1/2) and thick lines (S=+l)observed in samples with lower molecular weight = 1000) are the familiar defects seen in ordinary LMW nematics. It should be noted that in these samples the threads have a strong tendency to disappear with time and give place to a welldefined texture of Friedel’s nuclei with only integral lines and singular points. Integral lines or nuclei are occasionally found in samples of higher molecular weights (Fw > 10 000), but half integral lines are the most abundant. This is in agreement with the fact that both the splay elastic constant, K , I , and the bend elastic constant, K33,are larger than the twist elastic constant, K Z 2 [42]. Under these conditions, any splitting of the core of an I S J =1 line tending to remove the core singularity is not favored and since the line energy is proportional to IS12
(rw
when a core singularity is present, IS]=112 lines are favored. In planar samples, the half integral lines appear as thin loops floating in the bulk. These loops, which have a twist character, become smaller and smaller with time and to a great extent disappear after a few minutes. In samples with strong planar anchoring, surface lines with their ends attached to the same glass plate are observed. Very thin free droplets show disclination lines, the length of which increases with time, thus suggesting that they terminate on the free surface rather than on the glass surface. In LMWLCs, as one approaches the center of disclination, there is an isotropic core region. Its radius is estimated to be a few molecular lengths far from the isotropization temperature, TN+ but increases rapidly at TN.I. In MCLCPs, the cores appear quite large. Mazelet and KlCman 1571 reported important differences between the cores of the wedge parts of the half integral lines observed in free droplets of polyester 1. The +1/2 cores are much larger than the -1/2 cores. Positive and negative defects also display different mobilities. While the +1/2 defects are practically insensitive to any gentle flow given in the sample, the -1/2 ones move immediately. Mazelet and KlCman interpreted their observations for S=+1/2 by a high concentration of chain ends inside the core, which scatters light. Taking into account that for polyester 1, K , S=K3 > K22,they assumed that there is no splay involved outside the core. For S=-1/2, they developed a model which minimizes elastic energy by concentrating it in some angular sectors outside the core. The energy balance between splay and bend, which are the only contributions, leads to a subtle geometry where most of the bend energy is concentrated in a manner that avoids too much splay energy. In central part of the core, the molecular configuration is verti-
3 Defects in Nematic Main-Chain Liquid Crystalline Polymers
b)
Figure 9. Schematic models for the cores of half integral lines in polyester 1. Chain ends segregate (a) in the core (S=+1/2) or (b) along three quasi walls (S=-1/2) [57].
cal. Schematic representations of S=+1/2 and S=-112 wedge disclinations are shown in Fig. 9. Integral lines have been observed in the threaded texture [25] of a random copolyester prepared by the Eastman Kodak Company from 40 mol% poly(ethy1ene terephthalate) and 60 mol% p-acetoxybenzoic acid and commercially known as X-7G:
103
esters [59, 60, 64, 661. The core of the nuclei is much larger and the contrast is much fuzzier than in the Schlieren textures of ordinary LMW nematics. However, no detailed experimental observations of the disclination structure have been made and it is therefore difficult to ascertain whether these lines are singular or nonsingular. As arule, onlyS=+1/2 andS=+l are seen in nematics. Rare cases of higher strength defects (IS1 up to 4) have been observed in lyotropic systems [67] as also in two-component or impure thermotropic systems [68]. In the former cases, the defects were unstable as the microemulsion broke up after several minutes and the high strength singularities divided into conventional disclinations of S=+l. In the latter cases, however, the texture was found to be stable. More recently, transient disclinations of strength + 3/2 were observed in the nematic phase of a single component, thermotropic, rigid rod polytolane [69].They were characterized by six extinction brushes. Their formation was promoted by a combination of mechanical agitation and smaller undercooling below the clearing point. They had a lifetime of =30 s. Their small number and short lifetime were discussed in terms of their energy per unit line length, and in terms of the force that causes them to be attracted to and annihilated at other disclinations. For thermotropic MCPs, stable Schlieren texture with high strength defects (ISl=3/2 and ISl=2) was first reported by Galli et al. [70] for the poly(P-thioester):
3 They show a blurred contrast in linearly polarized light, but remain weakly contrasted in unpolarized light which may be taken as an indication of their singular character. Integral lines have also been observed in the Schlieren textures of a number of copoly-
4
maintained at a temperature close to the I - N transition. These defects were characterized by a single thin core. However, sometimes they possessed a blurred inner region from which the dark brushes were seen to emerge.
104
I1 Defects and Textures in Nematic Main-Chain Liquid Crystalline Polymers
Under these circumstances, it was difficult to distinguish between true high strength singularities and 'apparent' singularities consisting of several points bunched together around an immiscible isotropic droplet. Quite recently, high strength defects were observed in the nematic polymers of the following structure:
(polymer 5A m=6, polymer 5B m=8)
5 For the polymer 5A, disclinations of strength S=+3/2, + 2 and +5/2 were observed [7 I]. Most of these high strength disclinations were connected with neighboring defects of lower strengths of S=+1/2 andS= +1. Rare cases of pairs of +3/2 and -312 and of +3/2 and -2 disclinations were noted. These disclinations were found to be quite stable. They persisted even after annealing for several hours. For the polymer 5B,pairs of + 2 and -312 and of +1 and -312 disclinations were observed [72]. An examination of the textures showed that the core of these high strength disclinations was wider than that of the ordinary ones with S=+1/2 and S=+l. However, further studies are required for a better understanding of the core size and topology.
Biaxial Nematic Main-Chain Liquid Crystalline Polymers In 1970 the existence of a biaxial nematic phase for thermotropic LCs was theoretically predicted by Freiser [73]: the first-order transformation from the isotropic state to
the uniaxial nematic state may be followed at lower temperature by a second-order transformation to a biaxial nematic state. At this phase transformation, the rotation of the molecules around their long axes becomes hindered and the mesogens achieve planar alignment. As far as thermotropic polymers are concerned, the presence of a biaxial nematic phase was first suggested by Viney and Windle [59] for copolyester 3. Between 340 and 350"C, the beginning of the growth of the isotropic phase, this polymer shows a continuous Schlieren texture with both 2112 and f 1 disclinations, the former being relatively more numerous than the latter. Such a behavior is consistent with a uniaxial nematic phase. Below 340 "C, however, the melt exhibits interrupted Schlieren textures with disclinations of strength + l , but absence of any of strength +1/2, and occasionally large areas of homogeneous alignment separated by well defined walls [59, 661. As pointed out by Viney and Windle [59] these features, which are similar to those seen in LMWLCs with SmC structure, may in fact indicate the presence of a biaxial nematic phase. Windle and coworkers [51, 54, 561 have then interpreted some discrepancies between optical microstructures and X-ray diffraction patterns of nematic domains in copolyesters 3 and 6 in terms of biaxiality. Thin microtomed sections of these random copolyesters were
6 examined by X-ray scattering and shown to have a high level of preferred molecular orientation. However, when the same thin slices were viewed in the polarizing microscope, no suggestion of macroscopic orientation was seen. Long range rotational correlations of the molecules about their long
5
axes, giving othorhombic or lower symmetry and thus biaxial optical properties, would account for these observations. Another indication of biaxiality was provided by examination of thin samples of copolyester 7 prepared directly from the melt. The microstructure of such samples showed domains clearly delineated by walls [5 1, 551.
Defect Associations and Textures at Rest
105
b
a
\E”
C
,Cl
‘I
Even if there was some indication that copolyesters 3, 6 and 7 could present biaxiality, no clear evidence concerning their defects signature was given. Recently, however, De’Nkve, KlCman and Navard [45, 531 reported what we believe to be the first unambiguous observations of biaxial defects in a conventional thermotropic polymer commercially known as Vectra B 950@(8). These authors showed the existence of three types of half integer disclination lines (called E x , Ey and Ez),sometimes associated with integer disclination lines. Figure 10 shows a structural model of the chain (assuming that the aromatic planes are all parallel) and Fig. 11 illustrates the three possible half integer twist disclinations in the biaxial nematic. Interestingly, if Vectra B 950@is biaxial at rest, an external mag-
Figure 10. Schematic representation of the chain (Vectra B 9SO@).N x , N , and N , are the three principal refractive indices [53].
Figure 11. The three possible half integer disclination lines in the biaxial nematic phase of Vectra B 950@.(a) Rotation around the X axis, (b) rotation about the Y axis and (c) rotation around the Z axis [ S 3 ] . It is worth noting that the Ey line is perpendicular to the main director Y.
netic field or a weak shear flow disrupts the biaxiality and transforms the liquid crystal into a uniaxial nematic, thus indicating that the biaxiality is weak.
8
5 Defect Associations and Textures at Rest Textures correspond to various arrangements of defects. When the isotropic liquid is cooled, the nematic phase may appear at the deisotropization point in the form of separate small, round objects called droplets (Fig. 12). These can show extinction crosses, spiral structures, bipolar arrangements, or some other topology depending on boundary conditions. Theoretical studies based on a simple model confirm the stability of radial or bipolar orientation (Fig. 5 ) [22]. Considerations based on improved theoretical models yield stable twisted
106
I1 Defects and Textures in Nematic Main-Chain Liquid Crystalline Polymers
Figure 13. Marbled texture of copolyester:
Figure 12. The separation of the nematic phase in the form of droplets from the isotropic liquid of (a) polyester:
and (b) polyester:
structures. Nematic droplets characterize a type-texture of the nematic phase since they occur nowhere else. Upon cooling, the droplets grow and join together to form larger structures from which stable texture finally forms. The marbled, threaded and Schlieren textures are the ones most often cited in the literature for thermotropic MCLCPs. Examples of typical polymeric textures are shown in Figs. 13-15. Nematic marbled texture consists of several areas with different molecular orientation. On observing the preparation between
Figure 14. Threaded ~ t u r OfcoPolYester: e
t.c+olo.5
+\d
+*% C(CW3
obtained at 255 "C. (a) t= 10 min and (b) t=30 min.
5
Figure 17. Closed loops viewed between crossed POlarizers at 250°C: (a) t = O , (b) t = 2 min, (c) t=5 min,
Defect Associations and Textures at Rest
107
(d) t=6 min, (e) t = 8 min, (f) t= 15 min, (8) t=22 min and (h) t=30 min (from 1751).
108
I1 Defects and Textures in Nematic Main-Chain Liquid Crystalline Polymers
Figure 15. Schlieren texture of copolyester:
crossed polarizers, one can note that the interference color is nearly constant within the individual areas, indicating quasihomogeneous regions. Marbled textures very often appear with glass surfaces which have not been specially treated. The threaded (or line) texture is observed in preparations where threads lie more or less horizontally, their ends being attached to the glass surfaces and the other parts floating freely in the liquid crystal. The threads may be extended and curved due to flow of the liquid. They can be closed in the form of loops (Fig. 16) and, in this case, generally shrink and disappear more or less rapidly (Fig. 17, s. page 107) [74, 751. The Schlieren texture also called nuclei texture ('plage 2 noyaux') appears when the threads lie vertically (Figs. 18 and 19). When viewed between crossed polarizers this texture shows an irregular network of black brushes branching out from a number of scattered points or 'nuclei' and passing continuously from one nucleus to another. Four brushes meet at integral nuclei and two brushes at half integral nuclei - see Sec. 2.3.1 of this chapter. From the observation of S=+1/2 lines, the mesophase can be identified as a nematic phase since these singu-
Figure 16. Closed loop viewed between crossed polarizers at 250°C. Copolyester l l (from [75]).
larities occur nowhere else. Smectic C phases, which can exhibit the Schlieren texture, only form singularities with four brushes corresponding to S = k l . Transmission electron microscopy (TEM) studies conducted by Thomas et al. [62,63] have permitted direct visualization of the molecular director distribution in floworiented thin films prepared from semiflexible thermotropic LCPs of the following structure:
polymer 9A X = H, polymer 9B X = CH,)
9 These authors have used the technique of 'lamellar decoration' [76] which enables detailed assessment of characteristic mesophase defects and texture on a much finer scale than previously possible with conventional electron-microscopy preparations. The defects and texture existing in the polymer melt state are first retained by thermal quenching of the polymer fluid to room temperature. The glassy LCP film is then annealed above its glass transition, but below the melting point. Crystalline lamellae grow perpendicular to the local chain axis and effectively 'decorate' the molecular director
5
field throughout the sample. The most common types of disclinations seen in these polyesters are those of strength S=k1/2. Dipoles consisting of a positive and a negative z disclination at a typical separation of micrometers are frequently observed. Apparently, annihilation occurs if pairs of opposite sign disclinations approach each other more closely. These observations are consistent with those reported by Ford et al. [77] and Dong and Deng [78] who have used the same technique to reveal disclinations of strength S=+1/2 in a MCLCP containing flexible spacers between the mesogenic groups, and in an aromatic copolyester, respectively. Interestingly, in addition to isolated and variously paired disclination dipoles, arrays involving equal numbers of positive and negative disclinations are observed [76]. The molecular director pattern in regions defined by two pairs of oppositely signed half integer disclinations bears a likeness to that of the ‘effective domain’ structure hypothesized by Marrucci [79] to account for certain rheological aspects of MCLCPs. More recently, Chen et al. [64] have used the combined techniques of ‘lamellar decoration’ and ruthenium tetraoxide staining to reveal the director trajectories around disclinations of strength S=+1/2 and +1 in polymer 9B. All six topologies represented in Fig. 2 as well as (1/2, -1/2) and (1, -1) defect pairs have been observed, consistent with the fact that disclinations of opposite signs attract. More surprisingly, pairs of two positive ( + I , $ = d 4 ; +1/2) or two negative (-1/2, -1/2) disclinations have been seen repeatedly. These defect pairs correspond presumably to a nonequilibrium state and to an unstable texture. Quite recently, Chen et al. [71b, 801 have developed the technique of banded texture decoration. These authors showed that in thick samples, when the disclination density is small enough due to prior annealing,
109
Defect Associations and Textures at Rest
Figure 18. Schlieren texture of copolyester formed by transesterification of poly(ethylene-1,2-diphenox yethane-p,p'-dicarboxylate) with p-acetoxybenzoic acid. Recognizable singularities S=+1/2 (crossed polarizers, 250°C x200) (from [60]).
Figure 19. Schlieren texture of the polyester:
at the I - N transition. Recognizable singularities S= k1 (crossed polarizers, 278”C, ~ 2 0 0 ) .
quenching from the nematic state can lead to the appearance of a banded texture similar to that seen in flowing or deformed systems after removal of the external orienting influence. The banded texture formation is not crystallization-induced as suggested by Hoff et al. [81], since the nematic polymer of the following structure C1
10
110
I1 Defects and Textures in Nematic Main-Chain Liquid Crystalline Polymers
Figure 20. Grainy texture of polyester 6.
Figure 21. Schematic representation of the polydomain structure produced by spatial distribution of disclination lines of S=+1/2 (open circles) and S=-1/2 (filled circles) [ 6 5 ] .
showed a banded texture in the quenched amorphous glassy state. The bands lie perpendicular to the chain direction, thus offering the possibility of mapping the director orientations [82]. As compared to the ‘lamellar decoration’, the banded texture technique provides an extremely useful mechanism for direct visualization of defects and textures in both crystalline and non-crystalline materials by polarizing optical microscopy. The detailed observations by Chen et al. [71b, 821 point to a large predominance of disclinations of strength S=k1/2, some disclinations of strength S = + l and a few disclinations of strength S=-1. They also reveal (+1/2, -1/2) and (+l, -1/2) defect pairs and the formation of inversion walls.
In addition to these well characterized textures, an ill defined grainy texture appearing as a fine speckle pattern between crossed polarizers is commonly obtained at the Z-N transition (Fig. 20). This texture has been called a tight texture, a dense texture or a polydomain texture. Only a fine scale mottle is apparent (of say 0.1 - 1 pm). In general, this texture is observed for high molecular weight specimens, in thick samples and/or at low temperatures [83]. Determination of the molecular organization associated with this microstructure presents a challenge to the microscopist because the domain size approaches the resolution limit of the optical microscope. Transmission electron microscopy studies [63, 651 have shown that this microstructure can be considered as an assembly of ‘effective domains’ due to a particular arrangement of disclinations. The fine grains seen are believed to correspond to regions of uniform orientation correlation within nematic structures bounded by disclinations as illustrated in Fig. 21. Disclination densities up to 50-100 pm-2 have been observed [62]. However, the number of disclination lines decreases with annealing time [27, 81, 841 leading to a nematic texture that approaches the texture exhibited by LMWLCs although the resulting increased size range still remains with only several microns [85]. These observations seem independent on the specific MCLCP, the only difference being the time scale of this kinetic behavior. The coarsening of this nematic texture seems to be related to the dynamics of disclinations [65, 861. When a nematic phase is formed from the isotropic phase, its structure is dependent on the kinetics and mechanism of the phase transition. The initial texture is usually not stable under the given boundary conditions. The annealing process at T > Tgcan lead to the relaxation of the excess free energy associated with the discli-
5
nation density. The coarsening of the initially grainy texture, which is an ordering process, seems to involve a decrease of the number density of disclinations, p ( t ) , and an increase in the domain size & ( t ) ,by annihilation of the disclination lines, which occurs as a consequence of the diffusion and coalescence of the lines of opposite signs. The two quantities are interrelated:
where d is a spatial dimensionality. For copolyester 3 (X-7G), Shiwaku et al. [6S, 861 by plotting the average spacing between neighboring disclinations measured by optical microscopy as a function of annealing time found that:
Rojstaczer et al. [28] established through small angle light scattering (SALS) experiments that the size of the domains, characterized by the reciprocal scattering vector, Qmax,scales with time as:
1/Qmax= for a semiflexible aromatic polyester based on triad mesogenic unit containing a methyl arylsulfonyl substituted hydroquinone and a decamethylene spacer. It should be noted that the rate of decrease of the number of disclinations per unit volume is strongly influenced by the molecular weight of the polymer [26] and the rigidity of the polymer backbone [ 841. Rieger [87] has proposed a simple mean field theory for the description of the decay of the density of disclinations in nematic films of thermotropic MCLCPs, which is based on results for diffusion reaction systems of the type A + B +0, where A and B correspond to two different types of end points of a disclination line, respectively. This approach was found to approximate the functional behavior of the time dependence of the number of
Defect Associations and Textures at Rest
111
disclination points slightly better than the . algebraic law. The problem of describing SALS patterns for MCLCPs was first approached by Hashimoto et al. [61]. These authors developed a scattering theory for a model having isolated disclination lines contained in discs. The results poorly described the observed scattering. Later on, Greco [88] proposed a model consisting in a collection of isolated discs, randomly oriented, each containing a disclination dipole together with the corresponding director field. This simple two-body interaction of the S=21/2 disclinations could explain all the essential features of SALS experiments. The only parameter of the model (i.e. the distance between two disclinations) was sufficient to explain the evolution of the patterns as a function of annealing time. A suitable treatment of the glass surfaces between which the polymer is observed allows one to obtain either planar layers in which the director lies parallel to the surface or homeotropic films in which the molecules are aligned normal to the surface. Whereas, for LMWLCs, several techniques are now available for developing uniform orientation at surfaces, their application to orientation in MCLCPs is proving more difficult to achieve. However, a number of techniques have been used to create reproducible planar alignment. Rubbing of a raw glass plate with a diamond paste leads to uniformly aligned samples. SiOx layers evaporated obliquely (at 60" incidence) [7S] and freshly cleaved mica surfaces [60] produce similar effects. Polymer coatings on glass substrates [ 5 8 ] and surface-deposited hexadecyltrimethylammonium bromide [75] can also be used to align MCLCPs homogeneously. However, since a nondegenerate planar alignment is generally required, rubbing of the substrate after deposition of the polymer is used. Viewing planar samples from the top between crossed polariz-
112
I1 Defects and Textures in Nematic Main-Chain Liquid Crystalline Polymers
Figure 22. Development of homeotropic texture from copolyester 11 at 330°C (crossed polarizers). Annealing time (a) t=4 min, (b) t=8 min and (c) t = 2 8 min (from [75]).
ers results in the observation of four positions of extinction. Homeotropic alignment can be achieved by simple treatments of glass slides with boiling chromic-sulphuric acid, acetone and methanol (sequentially interspersed with water rinses) and rinsing with hot distilled water in an ultrasonic cleaner [75]. Films thus prepared appear completely dark
when viewed vertically between crossed polarizers because the director is oriented perpendicular to the glass surfaces and parallel to the light beam. If the cover glass is touched, the originally dark field of view brightens instantly, thus distinguishing between homeotropic and isotropic texture. Homeotropic alignment was also achieved by using a rocksalt substrate [89, 901. It should be noted that ageing MCLCPs at high temperature seems to encourage the homeotropic orientation [58, 75, 89, 901. This effect has been observed for polyesters 1, 6 and 11 (Fig. 22). A factor which seems to control homeotropic orientation at surfaces is the density of chain ends. For polyester 6, the tendency towards overall homeotropy was found to increase as the average molecular weight of the specimens decreased [90]. For polyesters 1 and 11, clearly some degradation of the samples occurred at the high temperatures required for homeotropic alignment leading to the formation of shorter chains with their relatively high density of chain ends [42, 751. This was borne out by the experiments on polyester 1 which showed that samples aged in an atmosphere of argon did not become homeotropic [42]. As has been discussed by Meyer [41], a perfect, ideal polymer nematic composed of infinite chains (i.e. with no chain ends) must lie with the chains parallel to the boundary surfaces for energetic reasons. For finite, but long, macromolecular chains, entropy leads to keep chain ends ‘dissolved’ in the bulk rather than condensed on the surfaces. On the other hand, short chains (i.e. abundant chain ends) will favor a homeotropic texture: while entropy will tend to distribute chain ends randomly, energy will tend to place them so as to relieve strains.
11
5
Defect Associations and Textures at Rest
113
Figure 23. Schematic representation of the geometry at a wall with a 90" rotation of the molecules across the wall [58].
KlCman et al. [58] reported that walls formed during the slow transformation from the planar to the homeotropic texture induced by annealing samples of polyester 1 for long times. The walls separated untransformed and transformed regions, with a consequence of 90" rotation of the molecules across them. The proposed geometry, akin to a 90" Bloch wall in ferromagnets, is shown in Fig. 23. Clearly, the molecules suffer a practically pure twist inside the transition region. Walls have also been observed in thin films of polyester 6 [89,90]. In these films, the walls also developed as the molecules attempted to transform to a homeotropic alignment, but the geometry was more akin to the 180" Bloch walls of ferromagnets. In this system, the walls do not separate planar and homeotropic regions, but rather regions where the molecules in both domains have some finite misorientation with respect to the plane of the sample, but in opposite senses. The domain structure adopted seems to be an example of splay compensation in operation, with the splay distortions in one plane being compensated by a negative splay component in an orthogonal plane. Classification of LC alignment into planar or homeotropic is an oversimplification. Under suitable conditions, the nematic director may make an angle with the substrate surface (Fig. 24). The observations between crossed polarizers of such samples are the
Figure 24. Nematic sample with the director making a given angle with the glass surfaces.
(a)
( C )
Figure 25. Ellipsoid of indices fora positive (n,>n,) uniaxial LC. The typical conoscopic images with average beam direction perpendicular to the plates are shown for (a) homeotropic, (b) planar and (c) oblique orientation [31].
same as in the planar case. Such tests are insufficient to check the uniformity of alignment and observations of the conoscopic images formed by illuminating the specimen with highly convergent light and inserting a so-called Bertrand lens below the eyepiece are needed (Fig. 25) [91]. The conoscopic interference patterns are also quite useful to identify uniaxial and biaxial nematics. For a uniaxial specimen, the conoscopic image consists of a dark cross
114
11 Defects and Textures in Nematic Main-Chain Liquid Crystalline Polymers
Figure 26. Basal interference figures of biaxial media. Sections perpendicular to the bisectrix of the acute angle. The cases depicted are those for thin sections of material and axial angle of about 10”(a) and 45” (b), respectively [3 I].
coupled with concentric circles, its position depending on the ‘crystal’ orientation. The cross will have four-fold symmetry. Figure 25a corresponds to the special case where the unique optic axis of the specimen is parallel to the microscope axis. For a biaxial material, the symmetry of the conoscopic image will be two-fold as illustrated in Fig. 26. Whereas this technique was used successfully for side chain liquid crystalline polymers (SCLCPs) [92-941, its application to MCLCPs appears not so easy to achieve owing to the difficulty with which MCLCPs give ‘monodomains’ in the liquid crystalline state. This is due mainly to the lack of strong surface anchoring, large viscosity and thermal degradation of polymers. However, with the aid of conoscopic observations, Noel et al. [75] have proved the positive uniaxial character of polyester 11. A final remark will concern the ‘glassy liquid crystal’. All the textures observed in the nematic state can easily be quenched and supercooled to room temperature. In spite of the thermal shock, both the planar alignment and the homeotropic one can be retained in the glassy state [59, 60, 66, 951.
6 Flow-induced Textures and their Relaxation Behavior The evolution from the static state to different textures observed as a consequence of controlled shearing experiments and the subsequent relaxation behavior at the cessation of flow were first analyzed by Graziano and Mackley [96] for a number of thermotropic MCLCPs. This pioneering work introduced and defined the terms used today to classify shear-induced textures: ‘worm’ texture, ‘ordered’ texture and ‘banded’ texture. It should be noted that these textures have also been observed in lyotropic nematics. The shearing behavior of MCLCPs depends on the initial state of the material. When molten MCLCPs exhibiting a threaded texture in the quiescent state are subjected to shear, the threads first extend and curve and disclination loops distort within the shear. However, at a rather well defined shear strain multiplication of the defects occurs [83, 84, 96, 971. These defects appear as short, dark curled entities which flow along the streamlines whilst continually changing their shapes, and the overall view of them during shear resembles a ‘vat of
6 Flow-induced Textures and their Relaxation Behavior
teeming worms’. The resulting disordered polydomain texture at the scale of a few microns (i.e. below the size of the specimen and independent of it) has been named a ‘worm’ texture by Graziano and Mackley [96]. The density of worms increases with increasing shear rate. At still higher shear rates, a sharp transformation can be induced into a birefringent texture whose optical appearance is dependent on the direction of crossed polarized light, thus indicating that flow alignment of some species is occurring [83, 84, 96, 971. This texture has been labelled the ‘ordered’ texture [96]. In general, the ordered texture is favored at low temperatures, high shears and small thicknesses of the samples, whereas the worm texture exists under less severe shear conditions and higher temperatures. On cessation of flow, the ‘worms’ relax by vanishing or coalescing to form threads with a relaxation time that depends on the polymer, the molecular weight of the specimen, the value of the applied shear and the temperature. Typically this might range from a few seconds to tens of minute. Upon cessation of shear, in the ordered texture, the worm texture develops within less than 1 s and subsequently relaxes as described above. Starting with a molten polymer exhibiting an ill-defined grainy texture, shearing causes this texture to be modified by the distortion and multiplication of defects [83,84, 961. Further shearing at higher rates leads to the apparent disappearance of any defect textures and the emergence of pure birefringence with the optic axis in the plane of shear (ordered texture). Upon cessation of shear, the material will relax back to the original dense defect texture. It should be noted, however, that another relaxation behavior has been observed for high molecular weight specimens [96,98-1001: a banded texture characterized by bands aligned perpendicular to the prior flow direction ap-
115
peared after cessation of shear in the ordered texture. In general, it is a low temperature, high shear relaxation phenomenon. Very little is known about the defects which are present in the worm texture because a direct optical study is hampered by the difficulty of isolating such moving objects 1961. However, the analysis of defects in Vectra B 950@(polymer 8) carried out by De’Nkveet al. [101, 1021 suggested that the defects are mostly half-integer twist disclination loops lying in the shearing plane. The chains are highly oriented along the flow direction near the core of the defects, thus indicating that the core is composed of a ‘perfect’ nematic, an assumption previously made by Mazelet and KlCman [57] for polyester 1. A peculiar feature of the cores of the defects is their large macroscopic size (of the order of 1 pm), which is probably related to the fact that the lines are in motion. They may move by dragging a large amount of nematic fluid in order to minimize distortion energy. Far from the defects, the directors are more or less aligned in the flow direction. However, in the vicinity of the defects two perpendicular orientations are observed: the director is oriented along the flow direction inside the loops, but along the vorticity axis outside the loops. The nucleation process of the ioops is therefore related to a rheological regime in which these two perpendicular orientations are favored. As pointed out by De’Nkve et al. [97, 1021 the simplest mechanism for this is the occurrence of tumbling. Concerning the mathematical modeling of LCPs, most attempts have concentrated on lyotropic systems. Doi [ 1031 has developed a defect free model for lyotropic rodlike polymers in the isotropic and anisotropic states and Larson and Mead [ 1041 have proposed a polydomain phenomenological description of lyotropic systems. Marrucci and Maffettone [lo51 solved a diffusion
116
I1 Defects and Textures in Nematic Main-Chain Liquid Crystalline Polymers
equation for rodlike lyotropic molecules Their two-dimensional numerical simulation in the shearing plane predicts a periodic rotation of the director at low shear rates, the so-called tumbling regime, and a higher shear alignment regime. This two-dimensional approach was extended in three dimensions by Ottinger and Larson [106]. These authors found that, at low shear rates, the director reaches two attractors. The first one is the shearing-plane and the second one is the vorticity axis. Tumbling is also shown to occur in three dimensions, but in a more complex manner, carrying both in-plane and out-of-plane components. However, these molecular theories do not at this stage predict or include defect effects. To account for experimentally observed microscopic observations of the worm texture behavior at low shear rates, De’Nhve et al. [97, 1021 assumed that tumbling generates large distortions that decrease their free energy by producing a large number of defects. An equilibrium density of defects should be obtained when the distance between neighboring defects prevents the director from tumbling freely. At this point defects start to annihilate. After having reached a first minimum, the defect density increases again. Such an explanation is consistent with the model developed by Marrucci and Greco [ 1071 which is based on the hypothesis that defects prevent the free rotation of the director. It also accounts for the damped oscillation phenomenon observed when the intensity of the light transmitted through the sample is plotted as a function of strain. These oscillations, which are related to the light scattered by the core of the defects, would correspond to the alternate action of nucleation and annihilation of defects. In addition, this tumbling mechanism agrees with the fact that the shear strain period of the defect density does not depend on shear rate since the tumbling time peri-
od is inversely proportional to shear rate [108]. The ordered texture does not seem to appear suddenly at a given shear rate, since there is a progressive increase in flow orientation [97]. Experimental data suggest that this texture may contain defects of the type seen in the worm texture, but with dimensions smaller than the resolution of optical microscope [ 1021. Quite recently, Gervat et a]. [84] developed a director simulation that incorporates optical, X-ray and rheological data. Essential elements of the model are concerned with a local molecular anisotropy, a molecular correlation coefficient, and a local defect texture. The main prediction of the simulation is that the presence of defects influences orientation relaxation after shear and that the defects themselves do not appear to have a profound effect on rheology. Following thin film shear [49, 109- 1111, uniaxial fiber drawing [ 1 121, injection molding [ 1131, or elongational flow [ 1141, nematic polymers typically exhibit a banded microstructure. This is especially evident when thin specimens are observed between crossed polarizers (Fig. 27). Dark bands lie perpendicular to the shear or flow direction Z, and are regularly spaced a few microns apart if Z is aligned parallel to the vibration direction of either polarizer. Detailed analysis of this microstructure [49, 50, 1 15- 1181 indicates that the global molecular alignment introduced by processing has partially relaxed, and that the molecular orientation instead varies periodically relative to the macroscopic shear direction. Banded texture is one of the characteristics of oriented specimens of both lyotropic and thermotropic LCPs. This subject has been reviewed in a recent article by Viney and Putnam [ 1101. The bands can appear during continuous shear [ 109, 1191 or after cessation of shear [83, 1201. In either case, a critical deformation rate must be exceeded [83,
6
Flow-induced Textures and their Relaxation Behavior
Figure 27. Banded texture of polyester 6. The shear direction was horizontal.
109, 119-1231, which has been predicted theoretically [124, 1251. A finite time must elapse from the onset of shear [lo91 or the cessation of shear [120, 1221 before the bands can form. The molecular organization in the banded texture is understood in some detail as a result of optical microscopy [49, 50, 83, 110-112, 115-122, 1261, transmission electron microscopy [49, 76, 89, 90, 115, 1271 and scanning electron microscopy [50]. It has been shown that the orientation variation of the molecules in this texture is near sinusoidal about the shear direction with most of the misorientation within the shear plane. There is a high degree of register of the molecules between each of the layers [50]. The influence of a number of parameters on the formation of bands has been studied: the molecular weight [96], the sample thickness [ 1 10, 120- 1221, the shear rate [109], the time during which the shear stress was applied [ 1221 and the total shear deformation [ 12I]. Various theoretical models have been proposed to account for this texture. Orientation of the directors within alternate planes by creation of alignment-inversion walls has been invoked [109]. Subsequent experiments seemed to indicate rather a smooth transition of the orientation between domains of a given direction [126]. The tumbling instability has been suggested as an explanation [128], but would not fulfill this condition. The development of novel instabilities of the director pattern was also observed as the
117
response of uniformly aligned nematic LCPs to a suddenly applied reorienting magnetic field [ 129, 1301. With conventional LCs composed of moderate length-towidth ratio molecules, the spatially periodic response to reorientating fields is not easily observed because it decays quickly to the homogeneous ground state. However, for nematics composed of very long molecules, the large anisotropy in their elasticity and viscosity has two important consequences. First, the periodic instabilities appear under a very broad range of circumstances, and second the stripe patterns produced are very long lived [ 130- 1321. An extension of the theoretical model by Guyon et al. [ 1331 for LMWLCs was used to interpret the data and it was suggested that a similar mechanism could explain the banded textures of LCPs in a shear flow. Unfortunately, because there are so many variables it is rarely possible to compare the observations made in different studies and the literature contains many conflicting results. For example, there is some contention as to whether the bands form during shear [ 109,1191or on stress relaxation afterwards [83, 1201. Ernst and Navard [120] considered a rather pure relaxation process via a periodic distortion after shear flow has ceased. According to this model, the macromolecules are stretched in the flow and then relax back. It is not clear, however, which mechanism in the relaxation process is responsible for the observed periodic structure. There is also disagreement as to whether the shear threshold does [ 110, 1211 or does not [120, 1221 depend on sample thickness. A shear threshold inversely proportional to the sample thickness, as reported in [121], would support the instability concept. Zielinska and Ten Bosch [124, 1251 argued that, for simple shear flow, the band structures are due to an instability mechanism present in the Leslie-Ericksen
118
I1 Defects and Textures in Nematic Main -Chain Liquid Crystalline Polymers
equations. They considered the existence of an instability producing a periodic pattern with a distortion perpendicular to the shear plane as observed in the optical experiments. They determined the critical shear rate and the critical wavevector for the instability to occur and showed that these depend strongly on the values of the elastic constants and anisotropic viscosities. They also obtained solutions for the director and velocity field. As discussed in the previous section, the literature on banded textures contains many attempts to correlate change in a parameter describing a banded texture with change in a molecular characteristic or processing conditions. However, experiments have sometimes given conflicting results. Also, theory is not sufficiently developed to suggest which microstructural parameter will be most sensitive to changes in any given molecular or processing parameter, and anticipated correlations frequently cannot be demonstrated conclusively.
References 0. Lehmann, Fliissige Kristalle, sowie Plastizitat von Kristallen im Allgemeinen, Molekulare Umlagerungen und Aggregatzustandsiinderungen, W. Engelmann, Leipzig, 1904. G. Friedel, Ann. Phys. (Puris) 1922, 18, 273. G. Friedel, F. Grandjean, Bull. Soc. Fr. Miner. 1910,33, 192 and 409. H. Sackmann, D. Demus, Fortschr. Chem. Forsch. 1969, 12, 349; Mol. Cryst. 1966, 2, 81; Mol. Cryst. Liq. Cryst. 1973,21, 239. H. Arnold, H. Sackmann, Z. Physik. Chem. (Leipzig)1960,213, 137, 145 and 262. D. Demus, L. Richter, Textures ofLiquid Crystals, Verlag Chemie, Weinheim, 1978. G. W. Gray, J. W. Goodby, Smectic Liquid Crystals: Textures and Structures, Leonard Hill, Glasgow, 1984. M. KICman, Points, Lignes, Parois dans les Fluides Anisotropes et les Solides Cristallins, Vol. 1, Les Editions de Physique, Orsay, 1977.
[9] M. KICman, J. Friedel, J. Physique Colloq. 1969,30, C4-43. [lo] J. Friedel, M. KlCman in Fundamental Aspects ofDislocation Theory, Vol. I , (Eds.: J . A. Simmons, R. de Wit, R. Bullough), National Bureau of Standards USA, Special Publication No. 317, 1970, p. 607. [ 1 I] J. Friedel in Dislocations in Solids, Vol. 1 (Ed.: F. R. N. Nabarro), North-Holland Publishing Company, Amsterdam, 1979, p. 3. [12] M. KICman, ibid, Vol. 5, 1980, p. 243. [I31 Y. Bouligand, ibid, Vol. 5, 1980, p. 299. 114) G. Toulouse, M. KlCman, J. Phys. (Lett.) 1976, 37, L 149. [ 151 G. E. Volovik, V. P. Mineev, Sov. Phys. J. E. 7: P. (Lett.) 1976,24,561; Sov. Phys. J. E. 7: P. 1977, 45, 1186. [ 161 V. Poenaru, G. Toulouse, J. Phys. 1977,38,887. [17] M. KICman, L. Michel, Phys. Rev. (Lett.) 1978, 40, 1387. [IS] L. Michel, Rev. Mod. Phys. 1980, 52, 617. [ 191 Y. Bouligand in Physique des Dc!fauts/Physics of Defects, Les Houches Session XXXV (Eds.: R. Balian et al.), North-Holland Publishing Company, Amsterdam, 1981, p. 668. (201 W. F. Brinkman, P. E. Cladis, Physics Today May 1982,48. [21] M. KICman, Points, Linesand Walls,Wiley, ChiChester, New York, 1983. [22] S. Chandrasekhar, G. S. Ranganath, Adv. Phys. 1986,35, 507. [23] M. KICman, Rep. Progr. Phys. 1989, 52, 555. [24] M. KlCman, Faraday Dis. Chem. Soc. 1985, 79, 215. [25] M. KlCman in Liquid Crystallinity in Polymers: Principles and Fundamental Properties (Ed.: A. Ciferri), VCH, Weinheim, 1991, p. 365. [26] V. Percec, D. Tomazos, C. Pugh, Macromolecules 1989, 22, 3259. [27] V. Percec, R. Yourd, Macromolecules 1989,22, 3229. [28] S. Rojstaczer, B. S. Hsiao, R. S. Stein, Div. of Polym. Chem., Am. Chem. Soc. Polymer Preprints 1988, 29,486. [29] C. Noel in Polymer Liquid Crystals (Ed.: A. Blumstein), Plenum Press, New York, 1985, p. 21. [30] C. Noel in Recent Advances in Liquid Crystalline Polymers (Ed.: L. Chapoy), Elsevier Applied Science Publishers, London 1985, p. 135. [31] C. Noel in Side Chain Liquid Crystal Polymers (Ed.: C. B. McArdle), Blackie, Glasgow, 1989, p. 159. [32] C. Noel in Liquid Crystal Polymers: From Structures to Applications (Ed.: A. A. Collyer), Elsevier Applied Science, London, 1992, p. 31. [33] F. C. Frank, Discuss. Furaday Soc. 1958,25, 19. [34] P. G. de Gennes, The Physics ofLiquid Crystals, Clarendon Press, Oxford, 1974.
7 [35] C. W. Oseen, Truns. Faraday Soc. 1933,29,883. [36l P. E. Cladis, M. KlCman, J . Phvs. (Paris) 1972, 33, 591. [37] R. B. Meyer, Phil. Mag. 1973, 27, 405. [38] S . I. Anisimov, 1. E. Dzyaloshinskii, Sov. Phys. J . E. i? P. 1972, 36, 774. [39] 1. E. Dzyaloshinskii, Sov. Phys. J. E. 7: P 1970, 31, 773. [40] P. G. de Gennes in Polymer Liquid Crystals (Eds.: A. Ciferri, W. R. Krigbaum, R. B. Meyer), Academic Press, New York, 1982, pp. 1 15-13 1. [41] R. B. Meyer in Polymer Liquid Crystals (Eds.: A. Ciferri, W. R. Krigbaum, K. B. Meyer), Academic Press, New York, 1982, pp. 133-163. [42] Sun Zheng-Min, M. KICman, Mol. Crysi. Liq. Cryst. 1984, 111, 321. [43] A. F. Martins, P. Esnault, F. Volino, Phys. Rev. (Lett.) 1986, 57, 1745. [44] P. Esnault, J . P. Casquilho, F. Volino, A. F. Martins, A. Blumstein, Liq. Cryst. 1990, 7, 607. [4S] T. De’Nkve, M. KICman, P. Navard, Liq. Cryst. 1994, 18, 67. 1461 S. E. Bedford, T. M. Nicholson, A. H. Windle, Liq. Cryst. 1991, 10, 63. [47] S . E. Bedford, A. H. Windle, Liq. Cryst. 1993, 15, 31. [48] J. R. Hobdell, A. H. Windle, J. Chem. Soc., Furaday Trans. 1995, 91, 2497. [49] A. M. Donald, C. Viney, A . H. Windle, Polymer 1983, 24, 155. [SO] S . E. Bedford, A. H. Windle, Polymer 1990,31, 616. [Sl] A . H. Windle, C. Viney, R. Golombok, A . M . Donald, G. R. Mitchell, Faraday Discuss. Chem. Soc. 1985, 79, 55. [52] N. D. Mermin, Rev. Mod. PIzys. 1979, 51, 591. (531 T. De’Nkve, M. KICman, P. Navard, J. Phys. I1 Frunce 1992, 2, 187. [54] C. Viney, G. R. Mitchell, A. H. Windle, Polym. Comnzun. 1983,24, 145. [SS] A.M. Donald, C. Viney, A. H. Windle, Phil. Mag. B 1985,52,925. [56] C. Viney, G. R. Mitchell, A. H. Windle, Mol. Cryst. Liq. C y s t 1985, 129, 75. [57] G. Mazelet, M. Kltman, Polymer 1986,27,714. [SS] M. KICman, L. Liebert, L. StrzCltcki, Polymer 1983, 24, 29.5. (591 C . Viney, A. H. Windle, J . Mater. Sci. 1982, 17, 266 1. [60] C. Noel, F. Laupretre, C. Friedrich, B. Fayolle, L. Bosio, Polymer 1984, 25, 808. [61] T. Hashimoto, A. Nakai, T. Shiwaku, H. Hasegawa, S. Rojstaczer, R. S. Stein, Macromolecules 1989, 22, 422. [62] S. D. Hudson, E. L. Thomas, R. W. Lenz. Mol. Cryst. Liq. Cryst. 1987, 153, 63. [63] B. A. Wood, E. L. Thomas, Nature 1986, 324, 655.
References
119
[64] S . Chen, L. Cai, Y. Wu, Y. Jin, S . Zhang, Z. Qin, W. Song, R. Qian. Liq. Cryst. 1993, 13, 365. [65J T. Shiwaku, A. Nakai, H. Hasegawa, T. Hashimoto, Macromolecules 1990, 23, 1590. [66] M. R. Mackley, F. Pinaud, G. Siekmann, Pol))mer 1981,22,437. [67] H. Lee, M. M. Labes, Mol. Crvst. Liq. Crysr. (Leit.) 1982, 82, 199. [68] N. V. Madhusudana, R. Pratibha. Mol. Cryst. Liq. Crysi. 1983, 103, 3 1. 1691 C. Viney, D. J. Brown, C. M. Dannels, R. J. Twieg, Liq. Cryst. 1993, 13, 95. [70] G. Galli, M. Laus, A. S . Angeloni, P. Ferruti. E. Chiellini, ELMPolym. J. 1985, 21, 727. (711 W. Song, S . Chen, R. Qian, (a) Makromol. Chem., Rapid Commun. 1993,14,605;(b) Macronzol. Symp. 1995, 96, 27. [72] Qi-Feng Zhou, X.-H. Wan, F. Zhang, D. Zhang, Z. Wu, X. Feng, Liq. Cyst. 1993, 13, 851. 1731 M. J. Freiser, Phys. Rev. (Lett.) 1970, 24, 1041. 174) B. Millaud, A. Thierry, C. Strazielle, A. Skoulios, Mol. Cryst. Liq. C v s t . (Lett.)1979,49,299. [7S] C. Noel, C. Friedrich, F. Laupretre, J . Billard, L. Bosio, C. Strazielle, Polymer 1984, 25, 263. [76] E. L. Thomas, B. A. Wood, Furaday Discu.ss. Chem. Soc. 1985, 79, 229. [77] J. R. Ford, D. C. Bassett, G. R. Mitchell, Mol. Cnst. Liq. Cryst. B 1990, 180, 233. [78] Y. Dong, H. Deng, Acta Polym. Sin. 1991, 584. [79] G. Marrucci. Proc. IX Int. Congr. Rheol. (University of Mexico), 1984. [80] S . Chen, C. Du, Y. Jin, R. Qian, Q. Zhou, Mol. Cgist. Liq. Cryst. 1990, 188, 197. 1811 M. Hoff, A. Keller, J. A. Odell, V. Percec, Polymer 1993,34. 1800. [82) S . Chen, W. Song, Y. Jin, R. Qian, Liq. Cryst. 1993, 15, 247. [83] N. J. Alderman, M. R. Mackley, Faraduy Discuss. Chem. Soc. 1985, 79, 149. [ 841 L. Gervat, M. R. Mackley, T. M. Nicholson, A. H. Windle, Phil. Trans. R. Soc. Lond. A 1995, 350, 1. [ S S ] J. L. Feijoo, G. Ungar, A. J. Owen, A. Keller, V. Percec, Mol. Cryst. Liq. Crysf.1988,155,187. [86] T. Shiwaku, A. Nakai, H. Hasegawa, T. Hashimoto, Polym. Commun. 1987, 28, 174. [87] J. Kieger, Macromolecules 1990, 23, 1545. [88] F. Greco, Macromolecules 1989, 22, 4622. 1891 A. M. Donald, J . Mat. Sci. (Len.) 1984, 3, 44. [90] A. M. Donald, A. H. Windle, J. Ma/. Sci. 1984, 19, 2085. 1911 N. H. Hartshorne, A. Stuart in Crystals and the Polarising Microscope, 4th edn., Edward Arnold, London, 1970. [92] H. Finkelmann in Liquid Crystals of One- and Two-Dimensional Order (Eds.: W. Helfrich, G. Heppke), Springer-Verlag, Berlin, 1980. 1931 A. I. Hopwood, H. J. Coles, Polymer 1985, 26, 1312.
120
I1 Defects and Textures in Nematic Main-Chain Liquid Crystalline Polymers
[94] F. Hessel, H. Finkelmann, Polym. Bull. 1986, 15, 349. [95] R. W. Lenz, J. I. Jin, Macromolecules 1981,14, 1405. [96] D. J. Graziano, M. R. Mackley,Mol. Cryst. Liq. Cryst. 1984, 106, 73. [97] T. De’Nkve, P. Navard, M. KlCman, J. Rheol. 1993, 37, 515. [98] S. Chen, Y. Jin, S. Hu, M. Xu, Polym. Commun. 1987, 28, 208. [99] X. Liu, D. Shen, L. Shi, M. Xu, Q. Zhou, X. Duan, Polymer 1990,31, 1894. [loo] J. Hou, W. Wu, D. Shen, M. Xu, Z. Li, Polymer 1994,35,699. [ l o l l T. De’Nbve, M. KICman, P. Navard, C. R. Acad. Sci. (Paris) 1993, 316, SCrie 11, 1037. [lo21 T. De’Nkve, P. Navard, M. KlCman, Macromolecules 1995, 28, 1541. [lo31 M. Doi, J. Polym. Sci.; Polym. Phys. Ed. 1981, 19, 229. [ 1041 R. G. Larson, D. W. Mead, J. Rheol. 1989,33, 1251. [lo51 G. Marrucci, P. L. Maffettone, Macromolecules 1989,22,4076. [lo61 R. G. Larson, H. C. Ottinger, Macromolecules 1991,24,6270. [ 1071 G. Marrucci, F. Greco, J. Non-Newtonian Fluid Mech. 1992,44, 1. [ 1081 R. G. Larson, Macromolecules 1990,23,3983. [I091 G. Kiss, R. S. Porter, Mol. Cryst. Liq. Cryst. 1980, 60, 267. [110] C. Viney, W. S. Putnam, Polymer 1995, 36, 1731. [111] J. Hou, W. Wu, M. Xu, Z. Li, Polymer 1996, 37, 5205. [112] S. C. Simmens, J. W. S. Hearle, J. Polym. Sci.; Polym. Phys. Ed. 1980, 18, 871. [113] H. Thapar, M. Bevis, J. Muter. Sci. (Lett.) 1983, 2, 733.
[ 1141 E. Peuvrel, P. Navard, Macromolecules 1991,
24, 5683. [ 1151 C. Viney, A. M. Donald, A. H. Windle, Poly-
mer 1985,26,870. [116] C. Viney, A. H. Windle, Polymer 1986, 27, 1325. [117] M. Horio, S. Ishikawa, K. Oda,J. Appl. Polym. Sci.; Appl. Polym. Symp. 1985,41, 269. [118] P. Navard, A. E. Zachariades, J. Polym. Sci. Part B; Polym. Phys. 1987,25, 1089. [119] P. Navard, J. Polym. Sci.; Polym. Phys. Ed. 1986, 24,435. [120] B. Ernst, P. Navard, Macromolecules 1989,22, 1419. [121] G. Marrucci, N. Grizzuti, A. Buonaurio, Mol. Cryst. Liq. Cryst. 1987, 153, 263. [122] E. Marsano, L. Carpaneto, A. Ciferri, Mol. Cryst. Liq. Cryst. 1988, 158B, 267. [123] T. Takebe, T. Hashimoto, B. Ernst, P. Navard, R. S. Stein, J. Chem. Phys. 1990, 92, 1386. [124] B. J. A. Zielinska, A. Ten Bosch, Phys. Rev. A 1988,38,5465. [125] B. J. A. Zielinska, A. Ten Bosch, Liq. Cryst. 1989, 6, 553. [126] C. Viney, A. M. Donald, A. H. Windle, J. Mater. Sci. 1983,18, 1136. [127] A. M. Donald, A. H. Windle, J. Muter: Sci. 1983,18, 1143. [128] G. Marrucci, PureAppl. Chem. 1985,57,1545. [129] C. R. Fincher, Macromolecules 1986,19,2431. [130] A. J. Hurd, S. Fraden, F. Lonberg, R. B. Meyer, J. Phys. 1985,46, 905. [131] F. Lonberg, S. Fraden, A. J. Hurd, R. B. Meyer, Phys. Rev. (Lett.) 1984, 52, 1903. [132] F. Lonberg, R. B. Meyer, Phys. Rev. (Lett.) 1985, 55, 7 18. 11331 E. Guyon, R. B. Meyer, J. Salan, Mol. Cryst. Liq. Cryst. 1979, 54, 26 1.
Part 2: Side-GroupThermotropic Liquid-CrystallinePolymers
Handbook ofLiquid Crystals D. Demus, J. Goodby, G. W. Gray, H.-W. Spiess, V. Vil Copyright 0WILEY-VCH Verlag GmbH. 1998
Chapter I11 Molecular Engineering of Side Chain Liquid Crystalline Polymers by Living Polymerizations * Coleen Pugh and Alan L. Kiste
1 Introduction 1.1 Chain Polymerizations Unil recently, most side chain liquid crystalline polymers (SCLCPs) were prepared by either hydrosilations of mesogenic olefins with poly(methylsi1oxane)s or by free radical polymerizations of acrylates, methacrylates and chloroacrylates [ l , 21. Both routes involve chain polymerizations, either directly or prior to a polymer analogous reaction. Chain polymerizations involve the four elementary reactions shown in Scheme 1 [ 3 ] . In contrast to step polymerizations, 1. Initiation 2. Propgation
I + M
m* + n M
IM*
-LI(W,M*
+A
3. Transfer
I(M),M*
4. Termination
I(M),,M* + A L
P+A* P
Scheme 1. Four elementary reactions of a chain polymerization.
chain polymerizations require an initiator (I) to produce reactive centers. In order for a vinyl monomer to be polymerizable, it must contain a substituent capable of stabilizing the resulting active site. Therefore,
* This article is reprinted with permission from C. Pugh, A. Kiste, Progress in Polymer Science, Vol. 22: Molecular Engineering of Side-Chain Liquid Crystalline Polymers by Living Polymerization, (1997), Elsevier Science Ltd., Oxford, England.
most 1-substituted alkenes undergo facile radical polymerizations (although a-olefins do not), whereas an electron-donating substituent is required to activate olefins to cationic initiation and propagation, and an electron-withdrawing substituent is required to activate olefins to anionic polymerizations. In addition to initiation and propagation, chain polymerizations may also undergo transfer and termination reactions (with reagent A for example), in which inactive chains (P) are formed. Chain transfer may occur to monomer, solvent, polymer, or a chain transfer agent. In order for this to be a transfer reaction rather than termination, the molecule (A*) generated must be reactive enough to reinitiate polymerization. Solvent is not required in radical polymerizations, although it retards autoacceleration and helps to diffuse heat; solvent is required in ionic polymerizations to control the heat generated. Since solvent is the component present in the highest concentration in most ionic polymerizations, solvents which either terminate the active sites or which act as chain transfer agents must be avoided. In contrast to chain transfer to solvent, which would be prevalent from the initial stages of a polymerization due to the solvent’s high concentration, chain transfer to polymer often does not compete noticeably with propagation until the end of the polymerization when monomer is depleted. Termination can be avoided or suppressed in ionic and metathesis polymeriza-
124
111 Molecular Engineering of Side Chain Liquid Crystalline Polymers
tions under appropriate conditions. In contrast to ionic polymerizations, in which the growing species can not react with each other due to their like charge, termination does occur in radical polymerizations by combination or disproportionation of two growing polymeric radicals. Therefore, termination is unavoidable in radical polymerizations. In contrast to the synthesis of small molecules, it is impossible to obtain from a polymerization a completely pure product that has a unique chain length and therefore a unique molecular weight. Chains of vastly different lengths will form if there are multiple propagating species with varying activities which do not exchange rapidly. This distribution of chain lengths and molecular weights is described by the moments of a statistical distribution of molecular weights (Eq. 1). The first moment of the distribution is the number average molecular weight (M,), which is the total weight of the sample (Cwi) divided by the number of chains in that sample (Cni). The weight (M,) and z-average (M,) molecular weights are the second and third moments of the distribution, respectively. That is, M,>M,>M,.
ni=number of chains of length i, M,=molecular weight of chains of length i. However, the thermotropic behavior of SCLCPs is generally most accurately described as a function of the number average degree of polymerization (DP,), rather than the average molecular weight. As shown in Eq. (2), the number (DP,), weight (DP,) and z-average (DP,) degrees of polymerization correspond to the number, weight and
z-average molecular weights, respectively.
Fin= D p n M O M , = 2,M ,
M,= mzM ,
M , = molecular weight of repeat unit and/or monomer, (2)
Since the number of polymer chains resulting from a chain polymerization is proportional to the amount of initiator used, the degree of polymerization is inversely proportional to initiator concentration. The ratio of the concentration of reacted monomer and polymer chains is determined by the ratio of the rates of propagation (R,) and all chain forming reactions, including transfer (&) (Eq. 3 ) . __
DPn =
AIM1 [Polymer]
-
RP Ri + R,, + ...
(3)
As shown in Fig. 1, high molecular weight polymer forms rapidly in a chain polymerization if initiation is slow and propagation is relatively fast, as in classic radical polymerizations. The molecular weight is limited by one of the chain breaking reactions (transfer or termination). In this case, the polymerization system consists of monomer and high molecular weight polymer at all stages of the polymerization, with only the yield of polymer increasing with increasing monomer conversion. In a radical polymerization which terminates by coupling and which has a constant rate of initiation, high molecular weight polymer forms at low ( kd). In this case, the molecular weight distribution depends on only the degree of polymerization as defined by the Poisson distribution (Eq. 7). ~
~
Living polymerizations in which initiation is fast and quantitative and which have irreversible growth offer several advantages over conventional polymerizations. In addition to the ability to obtain polymers with controlled molecular weights and narrow molecular weight distributions, it is also possible to control the polymer architecture and chain end functionality. For example, diblock and triblock copolymers containing liquid crystalline blocks have been prepared by living polymerizations.
2
‘Living’ Polymerizations used to Synthesize Side Chain Liquid Crystalline Polymers
2 ‘Living’ Polymerizations used to Synthesize Side Chain Liquid Crystalline Polymers
127
2.1 Anionic and Group Transfer Polymerizations of Olefins
The elementary synthetic and structural principles of SCLCPs are now being elucidated using living polymerizations in which the effect of a single structural feature can be isolated, while other structural variables remain constant. In particular, living polymerizations are used to determine the molecular weight at which the thermotropic be-
Alkenes polymerize anionically by nucleophilic attack of a growing carbanion at the least hindered position of the double bond of the monomer. Therefore, the monomer must be electrophilic and capable of stabilizing the resulting negative charge. In addition, the double bond must be the most electrophilic functionality in the monomer. Some vinyl monomers which can polymerize anionically are listed in Eq. (8) in their
havior of SCLCPs no longer depends on the addition or subtraction of another repeat unit. Additional comparisons should then be made only between samples whose thermotropic behavior is independent of molecular weight. In addition, living polymerizations with fast and quantitative initiation can be used to minimize the effect of polydispersity, although it is questionable as to whether or not polydispersity has any effect on the thermotropic behavior of SCLCPs [22, 231. Living chain ends are also required for preparing block copolymers by sequential monomer addition, and are convenient for preparing statistical copolymers whose composition depends on conversion. If the copolymerization is living, copolymers with compositions equal to that of the initial monomer feed are guaranteed by complete comonomer conversion. Thus far, anionic, cationic, metalloporphyrin, and ring-opening metathesis polymerizations have been used to prepare SCLCPs in a controlled manner.
order of reactivity, which corresponds to the electron-withdrawing ability of their substituents. However, the reactivity of the growing carbanions (and enolate anions) follows the opposite order shown above, which the most stable carbanions being the least reactive [ 5 , 6, 241. The choice of initiator is very important for a controlled anionic polymerization. The initiator will be nucleophilic enough to initiate polymerization if the pK, of the conjugate acid is similar to or greater than that of the propagating anion [25].However, the initiator should not be so reactive that it causes side reactions. Common solvents for anionic polymerizations include hydrocarbons and ethers. Protonic solvents, dissolved oxygen, and carbon dioxide result in either transfer or termination in most anionic polymerizations, depending on the nucleophilicity of the anion generated, and are therefore not used. Halogenated solvents also terminate anionic polymerizations and should be avoided.
128
I11 Molecular Engineering of Side Chain Liquid Crystalline Polymers
l/n (-R-.
Mt')
n
Scheme 2
In contrast to radical polymerizations in which there is only one type of propagating species, ionic polymerizations may involve several active species, each with different reactivities and/or lifetimes. As outlined in Scheme 2, ionic polymerizations may potentially involve equilibria between covalent dormant species, contact ion pairs, aggregates, solvent separated ion pairs, and free ions. Although ion pairs involving alkali metal countercations can not collapse to form covalent species, group transfer polymerization apparently operates by this mechanism. In anionic polymerizations, free ions are much more reactive than ion pairs (k;= 105k;), although the dissociation constants are quite small (K,= lo-') [ 5 ] . The equilibria between these different species in anionic polymerizations of nonpolar alkenes are dynamic enough relative to the rate of propagation to generate monomodal and narrow molecular weight distributions [26, 271. Although anionic polymerizations of nonpolar monomers such as styrene and budadiene are the most inherently living polymerizations available [ 5 , 61, only one mesogenic styrene has been polymerized by an anionic mechanism [28]. As shown "BuLi, THF
X CH2=CH
0
FH3 CH3-~i-O-~i-(CH2hO a
3
CH3
-
--o-o\
\ O(CH;),H
8
(9)
"ButCH2-CHb yH3
CH3-~-O-S;i-(CH2h0 CH3
a
W O ( C H 3 d H
3
in Eq. (9), 4-{ 3-[3-((4'-n-butoxy-(4"-phenyl)phenoxy)propyl)tetramethyldisiloxanyl]styrene was polymerized using n-BuLi as the initiator in THF (unreported temperature) in order to determine the effect of a siloxane spacer. In this case, the polymerizations were not well-controlled, with degrees of polymerization much higher than that calculated from [M],/[I], despite the low conversions (62-80% yield); pdi =1.2- 1.6. However, the degree of polymerization increased steadily as [M],/[I], increased. With the previous exception [28], only mesogenic methacrylates have been used in an attempt to prepare well-defined SCLCPs directly by an anionic addition mechanism. In this case, the most troublesome side reactions to avoid are intermolecular (Eq. lo), and intra-
4
!IYH3
R-C-C=CH,
+ R'O-,Li+ (?I3
R = initiator fragment or -C$-C-
I
C02R
2
‘Living’ Polymerizations used to Synthesize Side Chain Liquid Crystalline Polymers
molecular (Eq. 1 l), termination by nucleophilic attack at the carbonyl group of the monomer or polymeric repeat unit to produce unreactive alkoxide anions [29]; other electrophilic functionalities within the mesogen are also susceptible to nucleophilic attack. This is minimized in the initiation step by using a bulky initiator such as 1,ldiphenylhexyl lithium at low temperature (-78 “C). Low temperature and more polar solvents such as THF also minimize nucleophilic attack by the relatively hindered enolate anion during propagation [30]. In addition, the propagating ion pairs and free ions of these polar monomers may exchange slowly relative to propagation to produce polymers with multimodal molecular weight distributions [31 - 341. For example, anionic polymerization of t-butylmethacrylate results in broad (pdi =2.002.72) and multimodal molecular weight distributions when performed in a mixture of toluene and THF at -78 “C using 1,ldiphenylhexyl lithium as the initiator [3S]. However, narrow and unimodal molecular weight distributions are produced in either pure THF or pure toluene, in which propagation apparently occurs exclusively by free ions or ion pairs, respectively. Although 1 , l diphenyl-3-methylpentyl lithium and 1 , l diphenylhexyl lithium initiated polymerizations of methyl methacrylate in THF at -78 “C effectively yield ‘living’ polymers
129
with narrow molecular weight distributions (pdi = 1.1S ) , addition of lithium chloride decreases the rate of polymerization and narrows the polydispersity to 1.09 [36]. These conditions are also effective at controlling the anionic polymerization of t-butyl acrylate [37]. As outlined in Scheme 3 , lithium chloride narrows the polydispersity by shifting the equilibrium to LiCl adducts, thereby breaking up less reactive aggregates of the propagating polymer and shifting the equilibrium from free ions to ion pairs [38]. In addition to determining the effect of molecular weight on the thermotropic behavior of SCLCPs, ‘living’ anionic polymerizations have been used extensively to study the effect of tacticity. Although toluene is generally used as the nonpolar solvent in anionic polymerizations of methacrylates to promote association and isotactic placement [39, 401 it is a poor solvent for mesogenic methacrylates and/or the corresponding polymers. This results in only low yields of low molecular weight polymers with insufficient levels of isotacticity. Highly isotactic poly { 6-[ { 4’-[4”-[(S)-2”’methylbutoxy]phenyl]phenoxy } hexyl methacrylate} and poly { [4-(4’-methoxypheny1)phenoxyl-n-alkyl methacrylate } s ( n = 2 - 6) were instead obtained in chloroform at -20 to -30°C using t-BuMgBr as the initiator (Table l), although termination evidently broadened the polydispersity to 1.54- 3.06
130
I11 Molecular Engineering of Side Chain Liquid Crystalline Polymers
Table 1. Anionic polymerizations of n-[4’-(4”-methoxyphenyl)phenoxy]alkyl methacrylates ( n= 2-6).
Initiating system
Solvent
Temp. (“C)
[MI, -
GPC a
[I], DP”
Tacticity
Ref.
pdi
t-BuMgBr/Et,O
CHC1,
-30 to -20
10 20
13-39 24-68
1.54-3.06 (mm)=0.90-0.93 2.35- 14.0 (mm)=0.90-0.96
[42] [42]
t-BuMgBrlEt,O
THF
-78 to -60
10 20
14-21 19-33
1.23-2.27 1.23-1.57
(rr)=O.79-0.81 (rr)=0.79-0.85
[42] [42]
CH, Ph I I CH,CH,CHCH,C-, Li+
THF
-40
11
13b 23
1.13 1.26
( Y T ) = 0.80
[44] [44]
22
(rr)=O.80
I
3 LiCl a
Ph
Relative to polystyrene standards. Relative to PMMA standards.
at [M],/[I],= 10, and to pdi=2.35- 14.0 at [M],/[I],=20 [41, 421. Termination of the propagating chains is also demonstrated by the lower conversions at higher attempted degrees of polymerization, whereas partial consumption of the initiator is demonstrated by the much higher degrees of polymerization obtained relative to the theoretical values, even at these low polymer yields. Table 1 demonstrates that t-BuMgBr initiated polymerizations in THF at -60 to -78 “C are more controlled, yielding polymers with narrower polydispersities [41, 421. These conditions involve freely propagating chains and therefore result in syndiotactic polymers at low temperature [43]. Although ‘living’ polymethacrylates with the highest syndiotacticities [ ( T T ) = 0.92 - 0.961 have been achieved using t-BuMgBrl(nC,H,,),Al (1 : 3) in toluene at -78 “C [40], these conditions have not been applied to mesogenic methacrylates, probably because of the corresponding polymers limited solubility in toluene. Instead, the initiating system developed by TeyssiC [36] of
l,l-dipheny1-3-methylpentyl lithium in THF in the presence of at least three equivalents of LiCl has been used most successfully for preparing well-defined syndiotactic poly { n-[4’-(4”-methoxyphenyl)phenoxyllalkyl methacrylate} s with narrow molecular weight distributions (pdi = 1.05- 1.46) (Table 1) [44]. Although the polymers precipitated during the polymerization at -78 “C, Fig. 3 demonstrates that the polymerization is controlled and transfer is not detectable at -40 “C. However, the polymers precipitate during the polymerization when [M],/[I],>40, even at -40 “C, resulting in bimodal molecular weight distributions. In the absence of LiC1, 1,l-diphenyl-3methylpentyl lithium initiated polymerizations of 6-[4’-(4”-methoxyphenoxycarbonyl)phenoxy]hexyl methacrylate [45] and 4- [4’-(4’’-methoxyphenylazo)phenoxy] butyl methacrylate [46] in THF at -70°C yielded polymers (Scheme 4) with pdi = 1.09- 1.31 and molecular weights close to the theoretical values up to [M],/[I],=50.
2
‘Living’ Polymerizations used to Synthesize Side Chain Liquid Crystalline Polymers
80 G
-
131
0
60-
-3.0 0
Figure 3. Degree of polymerization
Scheme 4
In the latter case, the (co)polymerization was improved by adding a trace amount of triisobutylaluminium to the monomer solution in order to remove any residual water ~461. SCLCPs have also been prepared by group transfer polymerization (GTP) of mesogenic methacrylates are room temperature [47,48]. The livingness of the nucleophilic catalyzed GTP was originally attributed to a new mechanism (Eq. 12), in-
volving a (symmetry-forbidden [49, 501) eight-member cyclic transition state in which the trialkylsilyl group is transferred from the growing chain end to the incoming monomer [5 11. Alternatively, GTP may be an anionic polymerization in which a small concentration of enolate anions are in dynamic equilibrium with dormant silyl ketene acetal chain ends (Scheme 5 ) . This dynamic equi-
132
I11 Molecular Engineering of Side Chain Liquid Crystalline Polymers
M ,F .: , . Scheme 5
librium between a small concentration of enolate anions and a high concentration of dormant chain ends would reduce the lifetime of the active enolate anions, thereby accounting for the reduced rate [52, 531 of polymerization compared to classic anionic polymerizations of methacrylates. This shift in the equilibrium to only one type of active species would also account for the narrower polydispersities. Since the countercations are bulky tris(dimethy1amino)sulfonium [51] or tetrabutylammonium [54]ions, the active species would be free ions stabilized by solvent, which are less reactive towards intramolecular backbiting (Eq. 11). Most experimental results indicate that GTP is a classic anionic polymerization operating by reversible termination (dissociative mechanism) [55, 561, and/or degenerative transfer 157, 581. For example, termination occurs once monomer is consumed by backbiting at the pen-penultimate carbonyl to generate cyclic Pketoester endgroups as in classic anionic polymerizations (Eq. 11) [59]. Acids with pK, < 18 also terminate ‘group transfer’ polymerizations, whereas
so similar to those in anionic copolymerizations [55, 61, 621 rather than r l = r 2 = 1 as expected for an associative mechanism. However, the most compelling evidence supporting an anionic mechanism is that the stereochemistry resulting from group transfer polymerizations is nearly identical to that of an anionic polymerization involving freely propagating enolate anions [55, 63, 641. That is, the polymers are not highly isotactic as expected for the coordinated chain ends of an associative mechanism. Instead, freely propagating chains favor syndiotactic placement, especially at low temperature. In addition, ‘H-NMR experiments recently demonstrated that the silyl end groups of two different living polymethacrylates exchange in the presence of tris(dimethy1amino)sulfonium bifluoride 1651.
acids with pK, = 18- 25 act as chain transfer agents 1601. The reactivity ratios in GTP copolymerizations of methacrylates are al-
Although the polymer yield was high, the polymerization was not well-controlled, apparently due to termination at the phenyl
6-[4’-(4”-Methoxyphenoxycarbony1)phenoxylhexyl methacrylate was first polymerized by GTP using 1-methoxy-l-(trimethylsiloxy)-2-methyl- 1-propene as the initiator and tris(dimethy1amino)sulfonium bifluoride as the catalyst in THF at room temperature (Eq. 13) 1471.
2
‘Living’ Polymerizations used to Synthesize Side Chain Liquid Crystalline Polymers
133
benzoate carbonyl group of the mesogen, resulting in broader molecular weight distributions (pdi = 1.64- 1.77) and molecular weights substantially different than that calculated by [M],/[I],. As stated previously, molecular weights closer to the theoretical values were obtained by anionic polymerization of this monomer up to [M],/[I],=50 using the bulky anionic initiator, l,l-diphenyl-3-methylpentyllithium, at low temperature (-70 “ C )in THF [45]. However, the molecular weight distribution was still higher than desired (pdi = 1.19- 1.31). Relatively controlled ‘group transfer’ polymerizations are possible when the monomer lacks a phenylbenzoate group (Eq. 14). Poly[6-(4’-methoxy-4”-a-methylstilbeneoxy)hexyl
methacrylate] was prepared using 1-methoxy- 1-(trimethylsiloxy)-2-methyl- 1-propene as the initiator and tris(dimethy1amino)sulfonium bifluoride as the catalyst in THF at room temperature (DPn=2.8-28, pdi=1.09- 1.36) [48].
2.2 Polymerizations with Metalloporphyrins Aluminum porphyrins initiate controlled ring-opening polymerizations of oxiranes [67 - 691 p-lactones [70-721, 6-valerolactone [74], &-caprolactone [74] and D-lactide [75], as well as controlled addition polymerizations of methacrylates [76] and methacrylonitrile [77] (Eq. 15). As shown in Eq. (16), propagation occurs by a coordinative anionic mechanism
R = -H, -CH3, -CH2CH3, CH2C1, R = -H. CH,
Ph
C % m 3
134
I11 Molecular Engineering of Side Chain Liquid Crystalline Polymers
Ph
Ph
Ph
with insertion of the monomer into the aluminum-axial ligand bond of (5,10,15,20-tetrapheny1porphinato)aluminum derivatives [(TPP)AlCI, (TPP)AlMe, (TPP)AlOMe] to generate an aluminum alkoxide (oxiranes [69], 6-valerolactone [73], &-caprolactone [78], D-lactide [75]), aluminum carboxylate (p-lactones) [7 11 or aluminum enolate (methacrylates) [76] as the growing species. Inoue coined the term ‘immortal’ [79] to describe these polymerizations because the active chain ends are not ‘killed’ by proton sources. For example, the oxirane polymerizations undergo rapid and reversible exchange with alcohols such as methanol, hydroxyethyl methacrylate and polyethylene glycol (Eq. 17) [78 - 8 11; lactone
chain ends exchange rapidly with carboxylic acids [81]. In this case, the degree of polymerization corresponds to the molar ratio of reacted monomer to the sum of the initial concentrations of initiator and proton source (A[M]/[TPP-AlX] + [ROH],). Since the reversible termination is rapid compared to propagation (k,lk,= 10) [81], the polydispersity of the resulting polymers remains narrow at pdi = 1.1. Although the aluminum porphyrin initiated polymerizations take days or weeks to go to complete monomer conversion in the presence of a proton source, their rate is increased substantially by adding a sterically hindered Lewis acid such as methylalu-
minum bis(2,6-di-t-butyl-4-methylphenolate) [82]. In contrast to the oxirane and p-lactone polymerizations, E-caprolactone and 2,2-dimethyltrimethylene carbonate are polymerized by methylaluminum bis(2,6diphenylphenolate) in the absence of an aluminum porphyrin; these polymerizations are relatively rapid and controlled in the presence of i-propanol or methanol [83]. In contrast to the heterocyclic polymerizations, (TPP)AlCI does not initiate polymerization of methacrylates [76] and methacrylonitrile [77], whereas (TPP)AlMe initiation requires irradiation by visible light. These extremely slow polymerizations are also accelerated by addition of
2
bulky Lewis acids such as methylaluminum bis(2,6-di-t-butylphenolate) [77, 84 - 861 triphenylphosphine [86], or organoboron [87] compounds such as triphenylboron or tris(pentafluoropheny1)boron [88]. These Lewis acids accelerate the polymerization by coordinating to the monomer’s carbonyl group, thereby increasing the olefin’s electrophilicity. In addition, light is not needed to polymerize methacrylates if more nucleophilic thiolatealuminum porphyrins such as propylthio- and phenylthio(5,10,15,20tetrapheny1porphinato)aluminum [(TPP)AlSPr, (TPP)AlSPh] are used, preferably in the presence of methylaluminum bis(2,6di-t-butyl-4-methylphenolate) [89]. In all cases, the molecular weight is determined by the ratio of the concentrations of reacted CH3 X
CHz’F
135
‘Living’ Polymerizations used to Synthesize Side Chain Liquid Crystalline Polymers
I
F=O O(cHd6
m)me’
hv+
cH2c12.25 ‘C
0-y.
2) MeOH
-
N-@N
monomer to initiator up to at least [M],/[I],= 300, and the polydispersity is narrow at pdi = 1.05 - 1.2. Photoirradiated (TPP)AlMe was used to polymerize 6-[4’-(4”-methoxyphenoxycarbony1)phenoxyl hexyl methacry late [90], 6- [4’-(4”-n-butoxyphenoxycarbonyl)phenoxylhexyl methacrylate [91] and 6-[4’-(4”c yanophenox ycarbon y1)phenoxy ] hexyl methacrylate [92] in order to determine the effect of molecular weight and tacticity [(YY) =: 0.751. Although ‘high speed’ conditions [84 - 891 were not used, the polymerizations reached 76- 92% conversion in 10 h to produce polymers with the expected molecular weight (DP, < 35) and pdi = 1.09 - 1.35. In contrast, that of 6-[4’-(4”-cyanophenylazo)phenoxy]hexyl methacrylate (Eq. 18)
FH3
MetCH2-CtH
(18)
I ”
f.‘=O
o(CHd60-(=+!+N -
O
C
N
required 70 h to reach 90% conversion [90]. Nevertheless, methylaluminum bis(2,4-di-tbutylphenolate) accelerated polymerization of only the first three monomers to achieve SO-95% conversion in 15 min following its addition (Eq. 19), but terminated polymeriza-
tion of 6-[4’-(4”-cyanophenylazo)phenoxy]hexylmethacry late [90].
136
111 Molecular Engineering of Side Chain Liquid Crystalline Polymers
2.3 Cationic Polymerizations of Olefins Alkenes polymerize cationically by electrophilic addition of the monomer to a growing carbenium ion [8]. Therefore, the monomer must be nucleophilic and capable of stabilizing the resulting positive charge. In addition, the double bond must be the most nucleophilic functionality in the monomer. Some vinyl monomers which polymerize cationically are listed in Eq. (20) in their order of reactivity, which corresponds to the electron-do-
a \
6 s s 7H3
> C H ~ = ~ H> CH2=
C%=FH
\
OR
> CH2=C
> CH?=CH
/
,CH3 C%=C,
>
=/= (20)
CH3
ocH3
nating ability of their substituents. Sufficiently nucleophilic alkenes such as vinyl ethers, styrenes and isobutylene polymerize cationically to generate stabilized carbenium ions as the propagating species. However, the reactivity of the growing carbenium ions follows the opposite order shown above, with the most stable carbenium ions being the least reactive. Initiators include strong protonic acids and the electrophilic species generated by reaction of a Lewis acid with water, an alcohol, ester or alkyl halide (Scheme 6). In this case, initiation occurs by electrophilic addition of the vinyl monomer to the proton or carbenium ion, which may be accompanied by reversible or irreversible collapse of the ion pair (Eq. 21).
As demonstrated by the styrene polymerization in Scheme 7, propagation occurs in carbocationic polymerizations by electrophilic addition of the vinyl monomer to a growing carbenium ion. The resulting carbenium ions are very reactive and therefore difficult to control, with rate constants of propagation kp = lo4- lo6 1 (mo1s)-'. In addition, a significant amount of the positive charge is distributed over the P-H atoms, making them prone to abstraction by either monomer (ktr,M)or counteranion (ktr). Chain transfer by P-proton elmination from the propagating carbenium ions is the most common and detrimental side reaction in cationic polymerizations of alkenes. The first requirement for a controlled cationic polymerization is therefore to use only components which are nonbasic in order to reduce the possiblity of transfer by P-proton elimination. In the case of styrene polymerizations,
Scheme 6
Scheme 7
2
‘Living’ Polymerizations used to Synthesize Side Chain Liquid Crystalline Polymers
transfer also occurs by intramolecular Friedel -Crafts cyclization (k,) to form polymers with indanyl end groups, especially at high conversion. However, due to their higher activation energies compared to propagation, these chain transfer reactions can be suppressed by polymerizing at lower temperatures in hydrocarbon and chlorinated solvents [93]; nitro solvents are also used. Protonic solvents and amines which result in either transfer or termination should be avoided. In the absence of impurities and/or terminating agents, termination generally occurs in cationic polymerizations by unimolecular collapse of the ion pair (Q. However, termination may be reversible depending on the nature of the ligand. For example, anions containing chloride and bromide ligands such as SnCl;, SbCl;, SnBr;, BCI, and BBr, usually decompose reversibly. Reversible systems which exchange fast in comparison to propagation provide ‘living’ systems with narrow molecular weight distributions. Cationic polymerizations often involve both contact ion pairs and free ions as the propagating species (Scheme 8); solventseparated ion pairs have not been identified
Scheme 8
Scheme 9
137
spectroscopically. Propagation via covalent species has also been proposed for many of the new living carbocationic systems [94]. However, ions have been detected indirectly using optically active compounds and/or by various salt, substituent and solvent effects [8], and more directly by ‘H-NMR of exchange reactions in polymerizations and model systems [95-991.Although the reactions of these ion pairs and free ions are similar ( k ; = k i ) [ 100- 1021,theirdifferent lifetimes lead to high polydispersities [ 10, 1031. In order to achieve more controlled cationic polymerization conditions, the overall rate of polymerization must be decreased, thereby increasing the time available for functionalizing the chain ends. This has been accomplished by decreasing the stationary concentrations of active carbenium ions by shifting the equilibrium to dormant species. Since most cationic polymerizations involve an equilibrium between covalent adducts and ionic species, the equilibrium can be shifted from free ions to dormant covalent species by adding a salt with a common counteranion [104].Alternatively, the carbenium ions can be deactivated with nucleophiles, such as phosphines [ 1051, hindered amines [ 1061 dialkyl sulfides [ 107, 1081, ethers [ 1091, sulfoxides [110],esters[111],andamides[112] togenerate dormant onium ions (Scheme 9).The rapid dynamic exchange between active and dormant species, and the trapping of free carbenium ions with salts or nucleophiles also decreases the polydispersities (pdi < l .2).
138
I11 Molecular Engineering of Side Chain Liquid Crystalline Polymers
Table 2. Relative rates of initiation in cationic polymerizations initiated by covalent initiators in the presence of a Lewis acid catalyst [I 131. Initiator
Fast initiation
Similar initiation and propagation rates
Slow initiation
CH3TH-X
Table 3. Examples of 'living' cationic polymerizations of alkenes. SPe
Polymerization system
HA
HI, C&=qH
HA or RXILA
HI, I,, CH,=CH I 0-i-Bu
HA or RX LAINCX-
HA/Nu
CH3-CH-Cl, I
SnC14, BbN+,Ct,
Solvent
Ref.
n-hexane
C H ~ = ~
I
CF,SO,H, SMe,, CH2=CH I 0-i-Bu
The choice of initiator is also very important for a controlled cationic polymerization since initiation is often not quantitative in classic systems. Covalent initiators will be reactive enough to initiate polymerization if
Temperature ("C)
CH2CIZ
- 15
~191
CHIC12
- 15
[I071
the initiator ionizes faster than the (dormant) propagating chain ends. Many of the new 'living' systems therefore use initiators with better leaving groups and tertiary versus secondary active centers. The emerging
2
‘Living’ Polymerizations used to Synthesize Side Chain Liquid Crystalline Polymers
relative rates of initiation that are summarized in Table 2 correlate with the order of monomer reactivity listed in Eq. (20); completion of the table should therefore be obvious. Nevertheless, the new ‘living’ systems also use a much higher concentration of these well-defined initiators. This results in lower molecular weight polymers than in classic systems, which makes transfer less easily detected [114], and decreases the probability of spontaneous coinitiation with adventitious moisture. Examples of the different types of ‘living’ cationic polymerization systems are listed in Table 3. All involve relatively fast initiation and optimal equilibria between a low concentration of active carbenium ions and a high concentration of dormant species (Scheme 9). Only hydroiodic acid initiated polymerizations of N-vinyl carbazol are controlled in the absence of a Lewis acid activator or a nulceophilic deactivator [ 1151.
P
fiC-i-OH
0
+
polymerization of isobutylene [ 1 171 to produce polymers with - C1 terminated chains since boron has a greater affinity for acetate ligands than for chlorine [118]. The BCI, system therefore has the added advantage that the initiator with an acetate leaving group is more easily ionized than the propagating chain with a chlorine leaving group. [ 1 181 Table 3 also provides two examples in which it is necessary to add a common ion salt [ 1191 or a nucleophile [ 1071 to decrease the concentration of carbenium ions. Although many controlled cationic polymerizations that have been developed [ 1201, only mesogenic vinyl ethers have been used in an attempt to prepare well-defined SCLCPs by a cationic addition mechanism [ 1211. Nevertheless, these polymerizations (and copolymerizations) provide the most complete series of SCLCPs with the widest range of structural variables. As shown in Eq. (22), most of these monomers, includ-
1) lOMe$i
x CHFCH
I
OR
-
CHzCh. 0 ‘C 2) MeOH
H+CH~-CH+OCH~
I
OR
p 7% R
= -F R =
In contrast, all other adducts generated by addition of HI to the monomer’s double bond require activation with a Lewis acid suchas I, [ I 161. Similarly, cumyl acetaterequires activation by a Lewis acid to initiate
139
x
7 7%
CqCHZCH-CHEt , -O&XHZCH-CHEt
*
*
*
*
ing n-[(4’-(4”-cyanophenyl)phenoxy)alkyllvinyl ethers (Dp,= 2.1 -32, pdi = 1.021.54) [122- 1271, 2-[(4’-biphenylox~)ethYllvinYl ether (Dp,= 3.8-22, Pdi= 1.07 - 1.111 [1281, n-[(4’-(s(-)-2-methyl-
140
Ill
Molecular Engineering of Side Chain Liquid Crystalline Polymers
2oi
1:m 15
00
5
10 15 20 25 30 35 1.0
Figure 4. Degree of polymerization ( 0 ) and polydispersity (0)resulting from cationic polymerizations of (a) 5-[(4’-(4”-cyanophenyl)phenoxy)pentyl]vinyl ether [ 125, 1261 and (b) 8-[(4’-(2R,3S)-2-fluoro-3-methylpentyloxycarbonyl)-3’-fluorophenyl-4”-phenoxy)octyl]vinylether [ 1391 initiated by triflic acid in CH,Cl, at 0 “C in the presence of dimethyl sulfide.
[ 1371, n-[(4’-(2R,3S)-2-fluoro-3-methylpentanoate)-3’-fluorophenyl-4”-phenoxy)alkyllvinyl ethers (DP, = 4.0- 16, pdi = 1.06 - 1.13) [138], n-[(4’-(2R,3S)-2-fluoro-3methylpentyloxycarbonyl)-3’-fluorophenyl-4”-phenoxy)alkyl]vinyl ethers (DP, = 4.5-27, pdi= 1.08-1.22) [139], 8-[(4’Scheme 10 (2R(2S))-2-chloro-4-methylpentyloxycarbony 1) p h e n y 1- 4”- p h e n o x y ) o c t y 1] v i n y 1 1-butoxy)-4”-a-methylstilbeneoxy)alkyl]ethers (DP, = 3.9 - 15, pdi = 1.12- 1.22) vinyl ethers (DPn=4-24, pdi= 1.07- 1.12) [ 1401, 8-[(4’-(2R(2S))-2-fluoro-4-methyl[ 1291, 1 1-[(4’-cyano-4”-a-cyanostilbene)- pentyloxycarbonyl)phenyl-4”-phenoxy)ocoxy)alkyl]vinyl ethers (DP, = 4- 30, pdi = tyllvinyl ethers (DP, = 4.1 - 16, pdi = 1.111.05- 1.09) [ 1301, n-[(4’-(S-(-)-2-methyl1.33) [ 1411, and 4-{ 2-[4’-( 1 1 -vinyloxyun1-butoxy)phenyl-4”-phenoxy)alkyl]vinyl decyloxy)biphenyl-4-yl]ethyl]benzo-15ethers (DP,=4-26, pdi= 1.04-1.10) [131, crown-5 (DPn=6- 19) [142], werepolymer1321, n-[(4’-(2S,3S)-(+)-2-chloro-3-meth- ized using triflic acid as the initiator and ylpentyloxycarbonyl)phenyl-4”-phenoxy)- dimethyl sulfide as a deactivator in CH,Cl, alkyllvinyl ethers (DP,=4-21, pdi= 1.07at 0 “C. 1.16) [133, 1341,n-[(4’-(2R(2S),35’)-2-fl~oAlthough Webster et al. found that triflic ro-3-methylpentyloxycarbonyl)phenyl-4”- acid/dimethyl sulfide initiated polymerizaphenoxy)alkyl]vinyl ethers (DP, = 3.4 tions of i-butylvinyl ether are not living at 16, pdi= 1.05- 1.20) [135, 1361, n-[(4’0 “C in CH,Cl, [ 1071, plots of the degree of (2R(2S),3S),3S)-2-fluoro-3-methylpenta- polymerization against [M],/[I], demonno ate )p h e n y 1- 4”- p h e n ox y ) a 1k y 1] v in y 1 strate that transfer is usually not detectable ethers (DPn=3.5-15, pdi=1.04-1.20) at the molecular weights attempted in these
2
‘Living’ Polymerizations used to Synthesize Side Chain Liquid Crystalline Polymers
polymerizations (e. g. Fig. 4 a); the polydispersities are generally narrow at pdi= 1.1. Exceptions ( e . g . Fig. 4b) occur when the monomers are not sufficiently pure [ 1351411. The CF,SO,H/SMe, initiating system has also been used to cyclopolymerize 11-[(4’cyanophenyl-4”-phenoxy)alkyl]undecanyl3,4-bis(ethenyloxyethoxy)benzoateto form SCLCPs with crown ethers in the polymer backbone, although the polymerization is accompanied by termination [ 1431. Polymerization of the chiral vinyl ethers shown in Scheme 10 were reported to give only oligomers with DP, = 5 under the same condi-
mCH2-CH-CH2- CH
9I
R
,,,,+Aa
___f
tions, except with an unreported [M],/[I], [ 1441.However, this is apparently due to the use of impure monomers, reagents and/or reaction conditions, since similar functional groups were tolerated in previously polymerized monomers (Eq. 22). That is, cationic polymerizations tolerate cyano groups, phenyl benzoates, olefins, crown ethers and chiral centers with alkyl, chloro or fluoro substituents. However, they do not tolerate azobenzene groups [ 1441. Percec et al. have demonstrated by ID and 2D ‘H-NMR (COSY) that chain ends due to transfer by P-H elimination (Scheme 7) or termination by alkyl abstraction (Eq. 23) are absent or barely detectable,
-2-CH-CHrC-H
I
J
141
9
R
8
+
8
R-O-i-CF3
(23)
0
respectively, for poly { 11-[(4’-cyanophenyl-4”-phenoxy)undecyl]vinyl ether}s with DP, I 30 [ 1451. The ability to endcap growing polymers of 3-[(4’-(4’’-cyanophenyl)phenoxy)propyl]vinyl ether with 2-hydroxyethyl methacrylate (Eq. 24), 10-undecene-1-01 and
2- [2’-(2”-allyloxyethoxy)ethoxy]ethanol also demonstrates that transfer and termination are negligible at low degrees of polymerization (DPn=5 -7) [ 1461. As shown in Eq. (25), well-defined poly{ 2-[(4’-biphenyloxy]ethyl]vinyl ether}s (DP, = 22 - 50, pdi = 1.2) [ 1471poly { 2-[(4’-n-alkoxyphenyI-4’’-phenoxy)ethyl]vinylether}s
142
I11 Molecular Engineering of Side Chain Liquid Crystalline Polymers
(DP,=7.0-24, pdi= 1.02- 1.4) [147, 1481 p o 1y { 2 - [ (4’- c y an op h e n y 1- 4”- p h e n o x y ) ethyllvinyl ether}s (DP, = 4.9 - 28, pdi=1.04-1.2) [149], and poly(4(2’-vinyloxyethyl)-4”-phenylbenzoate} s (DP,= 18- 32, pdi = 1.2- 1.3) [ 1471 were prepared using HI as the initiator and either I, or ZnI, as the catalyst in CH,Cl, at -5 “C. { 6- [(4’-n-Butoxyphenyl-4”-benzoate)hexyllvinyl ether (DP,=7.6-38, pdi= 1.101.15) was polymerized similarly using HUI,, except that toluene was used as the solvent at -40°C [150]. Low molecular weight oligo(viny1 ether)s (DP,= 10- 19, pdi = 1.07 - 1.22) containing olefinic, nitro and cyano groups were also prepared using residual H,O in CH2C1, as the initiator and AlC1,Et as catalyst in presence of dimethyl sulfide at 0°C [151, 1521.
2.4 Ring-Opening Metathesis Polymerizations As outlined in Scheme 11, cycloolefins polymerize by ring-opening metathesis polymerization (ROMP) via a [2 + 21 cycloaddition with a propagating metallaolefin to form a metallocyclobutane intermediate, which then ring opens to regenerate a metallaolefin at the terminus of the growing chain [7, 153, 1541. The overall ring-opening of the cyclic monomer therefore occurs by cleavage of the double bond. Since polymers with narrow molecular weight distributions are only obtained if propagation is irreversible (k,,> kd), only highly strained cyclobutene derivatives and derivatives of the bicyclo[2.2.l]hept-2-enering system, including norbornenes, norbornadienes, ”-oxanorbornenes and 7-oxanorbornadienes, are candidates for controlled ROMP (Eq. 26).
Classic initiating systems for ROMP involve a transition metal such as WCl,, WOCl,, or MoCl,, combined with an alkyl metal such as n-BuLi or a Lewis acid such as AlEtCl,, SnMe,, or SnPh,. These systems generate low concentrations of unstable metallaolefin complexes which decompose during the polymerization [7]. In contrast, polymers with stable propagating chain ends are generated using well-defined initiators containing either a metallacyclobutane or a metallaolefin stabilizing by bulky ligands such as tricyclohexylphosphine, tertiary alkoxide groups and arylimido ligands with bulky isopropyl substituents at the ortho positions [7, 55-1631. Most of the initiators which produce well-defined polymers are either ruthenium(I1) complexes [161-1631, or high oxidation state complexes of titanium, tungsten or molybdenum (Table 4) [7, 155- 1601. However, the active chain ends and initiators containing titana- [ 1551and tantallacyclobutanes [ 1561 rapidly decompose once monomer is consumed at the temperatures (50-75°C) required to open the metallocyclobutane in chain growth. In contrast, the coordinatively unsaturated molybdenum and tungsten alkylidene initiators are stable for weeks in an inert atmosphere [7, 164- 1671, whereas the RuCl,(CHCH = CPh,)(PR,), complexes are moderately stable in air and are even tolerant of water in an inert atmosphere [ 16 1, 1621. However, RuCl,(CHAr)(PPh,), complexes decompose in solution via bimolecular reactions [ 1681. In addition to preventing spontaneous termination, increased stability of the metal center also minimizes chain transfer and other termination reactions. The elementary reactions outlined in Scheme 11 demonstrate that each repeat unit of the polymer contains a double bond which may participate in secondary metathesis reactions. Therefore, the most troublesome chain
2
‘Living’ Polymerizations used to Synthesize Side Chain Liquid Crystalline Polymers
Propagation
LR ‘b n
LxM-
n
Transfer
L
x
M
143
W
W
I
Scheme 11
Table 4. Metathesis of cis-pentene in toluene or benzene at 25 “C.
Catalyst
Turnovers per second
Ref.
1.7~10’ 4.1~10’
7.2~ 6.1x10-, 1.4~10-~
5.5xlo-4 = 2 x 10-5 0
transfer reaction to eliminate from ringopening metathesis polymerizations is chain transfer to polymer. Intramolecular chain transfer to polymer results in macrocycle formation with elimination of a shorter linear chain endcapped with the metalla-
olefin (or metallacyclobutane). Intermolecular chain transfer to polymer results in randomization of the chain lengths and broadening of the molecular weight distribution to the most probable distribution (pdi = 2).
144
I11 Molecular Engineering of Side Chain Liquid Crystalline Polymers
[ 174, 1751 cyclobutenes generate polymers with sterically hindered double bonds, and are therefore polymerized in a controlled manner using Mo(CH-t-Bu)(N-2,6-C6H,-iPr,) (0-t-Bu), without an additive . Molecular oxygen and sometimes the carbony1 groups of aldehydes, ketones, esters, and amides terminate the active metallaolefin chain ends of all but the new ruthenium initiators. For example, reaction of the growing chain with molecular oxygen generators a metal oxide and an aldehyde terminated polymer (Scheme 11). The resulting aldehyde chain end can then react with the active end of another polymer chain to generate a second molecule of the metal oxide and a polymer of higher molecular weight [176, 177).Since the rate of reaction with molecular oxygen is slower than the rate of propagation in polymerizations using the molybdenum and tungsten alkylidene initiators, it occurs at the end of the polymerization, resulting in a double molecular weight fraction [176, 1771. Of the molybdenum and tungsten alkylidene initiators shown in Table 4, Mo(CH-tBu)(N-2,6-C6H3-i-Prz)(O-t-Bu), is the least reactive and therefore the most selective initiator for ROMP. Its polymerizations are endcapped with benzaldehyde, pivaldehyde and acetone (benzophenone is less efficient) [7], although it tolerates a variety of nonprotonic functional groups, including acetate, cyano, ethers, esters, phenyl benzoates, N-substituted imides and trifluoromethyl groups in both the monomer [175, 1771821 and endcapping [ 1831 agent. With the exception of cyano groups [ 1841, ruthenium(I1) complexes tolerate the same groups, W(CH-t-Bu)(N-2,6-C6H,-i-Pr2)(O-t-Bu,> as well as alcohols, aldehydes, and ketones initiated polymerizations, thereby decreas[ 1621. These polymerizations are terminating the reactivity of the initiator and growed with ethyl vinyl ether. ing chain ends and increasing the rate of The final requirement for a controlled initiation relative to that of propagation. In ring-opening metathesis polymerization is addition, 3,3- [ 1731 and 3,4-disubstituted that the rate of initiation be greater than or
Secondary metathesis is prevented by using initiators which do not metathesize internal olefins. That is, if the metallaolefin or metallacyclobutane complex is reactive enough to metathesize an internal olefin, it is generally too reactive to obtain a living polymerization. As shown in Table 4, the reactions and selectivities of the commonly used initiators correlate with their rates of metathesis of cis-2-pentene. In general, the reactivity increases and selectivity decreases with increasing electrophilicity of the metal center. This is determined by both the metal itself and the electronegativity of the li gands . For example, W (CH-t-Bu)( N-2,6C6H3-i-Pr2)[OCMe(CF,),], metathesizes cis-2-pentene at over 17 turnovers per second in toluene at 25 "C, and therefore does not produce a well-defined polynorbornene [ 1641. In contrast, Mo(CH-t-Bu)(N-2,6C&,-i-Pr,)(Ot-BU), metathesizes cis-2pentene at only = 2 x 1 0 - ~ turnovers per second at 25 "C and is therefore much more selective at producing well-defined polynorbornene [ 1661. These transfer reactions can also be minimized by working with bicyclic monomers such as norbornene derivatives, which generate polymers whose double bonds are sterically hindered and therefore resistant to secondary metathesis [ 1531. Polymerizations of cyclobutene monomers are less controlled than norbornene polymerizations because the double bonds in the resulting polymers are highly susceptible to secondary metathesis reactions. However, well-defined polymers of cyclobutene [169, 1701 and methylcyclobutene [171, 1721 are obtained by adding a Lewis base (PR,) to
2
'Living' Polymerizations used to Synthesize Side Chain Liquid Crystalline Polymers
145
Table 5. Relative rates of initiation and propagation in ring-opening metathesis polymerizations of norbornene in C6D6 at 17-22 "C. Initiator
k,lk,
RuCl,(CHCH = CPh2)(PPh,)?
0.006 0.083
Mo(CH-t-Bu)(N-2,6-C6H,-i-Pr,)( 0-t-Bu), RUCl,(CH-p-C6H,X)(PPh,), x=-Cl X=-NO, X =-NMe, X=-OCH, X=-CH, X=-F X=-H
Ref.
1.2 2.3 2.6 2.6 2.9 4.8 9.0
equal to that of propagation. As shown in Table 5 for the most selective initiators, only the RuCl,(CHAr)(PPh,), complexes initiate norbornene polymerizations faster than propagation occurs, thereby producing polymers with very narrow molecular weight distributions (pdi = 1.04- 1.10) [ 1851. Although one would expect the rate of initiation to be highest with benzylidenes substituted with electron-withdrawing groups (most electrophilic metal center), the electronic effect of the alkylidene substituents is obviously minor. The aryl alkylidene is therefore evidently less sterically demanding than the insertion product (Eq. 27). In contrast, the low initiation rates of the ruthenium diphenylvinyl alkylidenes produce poly-
norbornenes with polydispersities of approximately 1.25 1851, although the diphenylvinyl alkylidene appears to be less sterically demanding than the cyclopentylidene insertion product (Eq. 28). Its lower reactivity was attributed to conjugation [185]. The rate constant of
initiation using Mo(CH-~-BU)(N-~,~-C,H~-~-P~,)(O-~-BU)~ is also slightly less than that of propagation, producing polynorbornenes with pdi < 1.10 [ 1861. This is apparently because the neopentylidene ligand with a tertiary carbon j3 to the metal is more sterically demanding than the insertion products, which contain secondary carbons j3 to the metal (Eq. 29).
146
111 Molecular Engineering of Side Chain Liquid Crystalline Polymers
With one exception [ 1871, only mesogenic norbornenes have been used to prepare well-defined SCLCPs by a ring-opening metathesis mechanism. Norbornene monomers containing terminally attached p-methoxybiphenyl[22, 188- 1901andp-cyanobiphenyl [ 189- 1911 mesogens were polymerized using Mo(CH-t-Bu)(N-2,6-C6H3i-Pr,)(O-t-Bu), as the initiator in order to determine the effect of molecular weight, polydispersity, flexible spacer length and mesogen density on the thermotropic behavior or SCLCPs with polynorbornene backbones (Eqs. 30 and 31). As shown in Fig.
R = -C@(CHz)IIR-0
-
-
5 (a) using a 5-{[n[4’-(4”-methoxyphenyl)phenoxy]alkyl]carbonyl} bicycle[ 2.2.11 hept-2-ene as an example, the GPC-determined number average degree of polymerization (relative to polystyrene) agrees well with the initial ratio of monomer to initiator. In this case, the degree of polymeriza-
tion varied from 6 to 151, with polydispersities of 1.05- 1.28 [22, 1881. Polymerizations of 5- { [n-[4’-(4”-methoxyphenyl)phenoxy]alkyl]methyleneoxy } bicyclo[2.2.1] hept-2-enes (DP, = 9 - 342, pdi = 1.15 1.19) [ 1891 and 5- { [n-[4’-(4”-cyanophenyl)phenoxy]alkyllcarbonyl } bicyclo[ 2.2.1I hept-2-ene (DP, = 7 - 290, pdi = 1.08- 1.27) [189] yielded similar results. The higher polydispersities were usually the result of a minor amount of a double molecular weight fraction in addition to the expected molecular weight due to insufficient degassing of molecular oxygen.
R = -OCH, -CN
Higher polydispersities were obtained
in Mo(CH-t-Bu)(N-2,6-C6H3-i-Pr,)(O-tBu), and Mo(CH-t-Bu)(N-2,6-C6H,-iPr,)(OCH,(CF,),), initiated polymerizations of (+)-endo,exo-5,6-di { [n-[4’-(4”cyanophenyl)phenoxy]alkyl]carbonyl } bicyclo[2.2.1]hept-2-ene (pdi = 1.2 1- 1.60)
2
‘Living’ Polymerizations used to Synthesize Side Chain Liquid Crystalline Polymers
147
wx
‘4
t
1404 a 120
80 100
1204 100
13.0
-
0
-2.0
40 “&&O 4.l
20 0 0 0
g
-1.0
25 25
~
50 50
d
75 75
o
100 100
.
o
I
0
25
50
75
0.0 100
Figure 5. Degree of polymerization ( 0 )and polydispersity (0) resulting from ring-opening metathesis polymerizations of (a) 5-( [6’-[4”-(4”’-methoxyphenyl)phenoxy]hexyl]carbonyl)bicyclo[2.2.1 Ihept-2-ene [22] and (b) 5-[[[2’,5’-bis[(4”-methoxybenzoyl)oxy]benzyi]oxy~carbonyl~bicyclo[2.2.l~hept-2-ene [I821 initiated by Mo(CH-t-Bu)(N-2,6-C,H,-i-Pr2)(O-t-Bu), or Mo(CHCMe,Ph)(N-2,6-C6H,-i-Pr2)(O-t-Bu),, respectively, in THF at 25 “C.
[ 1911 and (+)-endo,exo-5,6-di{ [n-[4’-(4”methoxyphenyl)phenoxy]alkyl]carbonyl } bicyclo[2.2.1]hept-2-ene(pdi= 1.20-2.75) [190], respectively, due to oxygen and monomer impurities (Eq. 31). The degrees of polymerization (DP, = 41 - 266) were substantially different from the attempted ratio of monomer to iniator ([M],/[I],= IOO), which were attributed to difficulties in actually controlling this ratio. Surprisingly, the ‘absolute’ molecular weights determined using a GPC with a viscometry detector and a universal calibration curve were even higher, with degrees of polymerization (DP, = 87 - 372) approximately twice those calculated relative to polystyrene [ 1911. Several polynorborenes with laterally attached mesogens have also been prepared using Mo(CHCMe2Ph)(N-2,6-C,H,-iPr,)(O-t-Bu), as the initiator in order to determine the effect of molecular weight
[ 1821, the effect of the length of the n-alkoxy substitutent [182, 192- 1971 as well as to design polymers with SmC mesophases by proper choice of the mesogen [192, 193, 196, 1971 and to test various concepts for inducing smectic layering in nematic liquid crystals (Scheme 12) [192- 1951. In particular, polymerizations of both 5 4 [2’,5’b i s [ ( 4”- n - a 1k o x y b e n z o y 1)ox y ] be n z y 1] oxy]carbonyl]bicyclo[2.2. I jhept-2-enes ( D P , = 5.1 - 168, pdi = 1.12- 1.27) and 5,sbis[ (4’-n-alkoxybenzoyl)oxy]- 1,2,3,4-tetrahydronaphthalene (DP, = 40 - 101, pdi = 1.06- 1.21) are fairly well controlled as demonstrated by the similar degrees of polymerization and [M],/[I], (Fig. 5 b), and by the polymers’ narrow polydispersities [182]. Polymerization of 5-{[[2’,5’,-bis[2( 3 ”-f 1u or o - 4”-n - a 1ko x y p h e n y 1) et h y n y 11benzyl]oxy]carbonyl } bicyclo[2.2.1 Ihept2-enes also results in the expected molecu-
148
111 Molecular Engineering of Side Chain Liquid Crystalline Polymers
F
Scheme 12
lar weights (DP,=39-76, pdi= 1.09-1.29) at [M],/[I],-50) [196, 1971. In this case, higher polydispersities result when the monomers are viscous oils which can not be purified by recrystallization. 5- { [ [2',5'b i s [ (4"- n - ( (p e r f 1u or o a 1k y 1)a I k o x y )b en zoyl)oxy]benzyl]oxy]carbonyl} bicyclo[2.2.l]hept-2-enes were also polymerized by ROMP, although it was necessary to use higher temperatures (40 "C) in order to prevent the polymer from precipitating out of solution during the polymerization; this
Scheme 13
-
broadens the polydispersity to pdi 1.5 [195]. Both norbornene and butene monomers containing terminally attached p-nitrostilbene mesogens were recently polymerized using RuCl,(CHPh)(PPh,), as the initiator in order to determine the effect of molecular weight and flexibility of the polymer backbone on their thermotropic behaviour (Scheme 13) [187]. Although the polydispersities are quite narrow (pdi = 1.07- 1.11), Fig. 6 a shows that the molecular weight resulting from polymerizations of 5-[n-[4'(4"-nitrostilbeneoxy)alkyl]carbonyl }bicycl0[2.2.1]hept-2-enes (DP, = 15-47) using RuCl,(CHPh)(PPh,), as the initiator may be slightly less controlled than those using M o (CHCMe,R) (N-2,6-C 6H i- Pr,) ( 0-tBu), to polymerize norbornene monomers, although nitro groups do react slowly with the latter initiator [ 1831.That is, Fig. 5 demonstrates that the number average degree of polymerization of mesogenic polynorbornenes determined relative to polystyrene standards generally match [M],/[I], quite closely. In contrast to the 5-{ [n-[4'-(4"-nitrostilbeneoxy)octyl] alkyl]carbonyl]bicyclo[2.2.l]hept-2-enes, Fig. 6 b shows that polymerization of the corresponding cyclobutene monomers yield polymers with de-
,-
2
200 180 180160 140 12010080
‘Living’ Polymerizations used to Synthesize Side Chain Liquid Crystalline Polymers
4.0 a
-
i
-
200 180-
-3.0 3.0
160140
-2.0 2.0
120100
149
b
. I
-
80 60-
-1.0
400
I
I
I
I
20
40
60
80
I
0.0 100
20
0
I
I
I
I
20
40
60
80
0.0 loo
[MIdDlo Figure 6 . Degree of polymerization ( 0 )and polydispersity (0) resulting from ring-opening metathesis polymerizations of (a) 5- [ [8’-[4”-(4’”-nitrostilbeneoxy)octyl]carbony~)bicyclo[2.2. Ilhept-2-ene and (b) poly[3-((8’-(4”(4’”-nitrostilbeneoxy)octyl)carbonyl)methyleneoxy)methyl]cyclobutene] initiated by RuCI,(CHPh)(PPh,), in CH,Cl, at 25 and 45 “C, respectively [187].
grees of polymerizations which are two to six times higher than [M],/[I], (DP,= 3072, pdi = 1.1 1 - 1.38). Mo(CHCMe2Ph)(N-2,6-C,H,-iPr,)(O-tBu2) has also been used to prepare low molar mass liquid crystalline polyenes [ 1981.
2.5 Polymer Analogous Reactions on Well-Defined Precursor Polymers In order to synthesize homopolymers by polymer analogous reactions, the reaction must go to 100% conversion. Hydrosilations of mesogenic olefins are the most common polymer analogous reactions used to prepare SCLCPs [199]. However, the precursor poly(methy1 si1oxane)s are generally not prepared by living polymerizations, nor do the hydrosilations readily go to comple-
tion. Only recently has a communication appeared describing a well-defined side-chain liquid crystalline polysiloxane [200]. As shown in Scheme 14, the precursor polysiloxane (DP, [M],/[I],, pdi < 1.2) was prepared by anionic ring-opening polymerization of pentamethylvinylcyclotrisiloxane using lithium trimethylsilanolate as the initiator in THF at 0 “C, followed by termination with t-butyldimethylsilyl chloride. The resulting poly(dimethylsi1oxane-mmethylvinylsiloxane) was functionalized by hydrosilation of the vinyl groups with the monoadduct of a mesogenic olefin and 1,1,3,3-tetramethylsiloxane. The thermotropic behavior of these (co)polymers has not been reported yet. Polybutadiene was the first well-defined polymer used in a polymer analogous reaction to synthesize SCLCPs [201-2031, primarily for comparison to the corresponding block copolymers discussed in Sec. 4.1.1 of
-
150
111 Molecular Engineering of Side Chain Liquid Crystalline Polymers
this chapter. As shown in Scheme 15, butadiene was polymerized by a living anionic mechanism using s-butyl lithium as the initiator in THF at -78 “C to produce poly( 1,2\/=
FH3
M~Si--ttO-S~O-SifOSiMer’Bu FH3 3‘
Scheme 14
Scheme 15
)
CH3 CH3-qi-O-(i-(CH>,O-CN CH3 CH3
butadiene) (DP,,= 1.1 x lo3,pdi= 1.10). Poly[((2-cholesteryloxycarbonyloxy)ethyl)ethylene] [201] (pdi= 1.13), poly{ [4-(4’methoxypheny1azo)(2”,3”,4”,6”-D4)pheny1 glutaratelethylene} [202] (pdi = 1.15) and poly { [4-(4’-methoxyphenylcarbony1)phenyl glutaratelethylene} [203] were then prepared by hydroboration of the remaining double bonds with 9-borabicyclo[3.3.1]nonane (9-BBN), followed by oxidation and formylation or esterification with the corresponding mesogenic chloroformate or acid chloride. Another cholesterol-containing polymer, poly[(2-cholesteryloxycarbonyloxy)ethyl methacrylate], was prepared by anionic polymerization of a protected hydroxyethyl methacrylate for comparison to the corresponding diblock and triblock copolymers (Sec. 4.1.1) [204-2101. As shown in Scheme 16, 2-(trimethylsi1oxy)ethyl methacrylate was polymerized in THF
2
‘Living’ Polymerizations used to Synthesize Side Chain Liquid Crystalline Polymers
1
1.51
pyr, 20 ‘C. 20 h
Scheme 16
DMSO. C s O H . B u ~ z + B i100-120 , ‘C
Scheme 17
at -78 “C using l,l-dipheny1-3-methylpentyl lithium as the initiator, which was generated in situ by reacting s-butyl lithium with 1,l-diphenylethylene. The trimethylsilyl protecting group was removed when the polymer was quenched and precipitated in acidic methanol. The resulting poly(2-hydroxyethyl methacrylate) block was then quantitatively functionalized with cholesterylchloroformate. Although the molecular weight distribution of poly[(2-trimethylsiloxy)ethyl methacrylate] is reported to be broad (pdi = 1.30) when polymerized under
these conditions and narrow (pdi = 1.OS) when polymerized in the presence of 10 equivalents of LiCl [211], the molecular weight distribution of the final poly[(2-cholesteryloxycarbonyloxy)ethyl methacrylate]~ prepared in these studies varied from pdi = 1.O - 1.2 when DP, < 1.59 [204, 2081, and pdi=1.90 when DPn=461 [208]. In contrast to the direct cationic polymerizations of mesogenic vinyl ethers discussed in Sec. 2.3 of this chapter, poly { n-[(4’-cyanophenyl-4”-phenoxy)alkyl]vinyl ether}s
152
I11 Molecular Engineering of Side Chain Liquid Crystalline Polymers
t
“I
604
1-8 1.6
B a 40
1.4
20
1.2 1.o
0
0
20
40
60
80
100
[mdmo Figure 7. Degree of polymerization ( 0 ,). and polydispersity (0, 0)of the precursor polymer resulting from the cationic polymerization of 6-chlorohexylvinyl ether ( 0 ,0)initialed by HIII, in toluene at -40 “C, and of the final poly [ 6-[(4’-n-ethoxy-4”-azobenzene)hexyl]vinyl ether) s prepared by polymer analogous reactions (W, 0) [214].
can also be prepared by (quantitative) etherification of poly [n-(chloroalky1)vinyletherls as shown in Scheme 17 [212, 2131, with reportedly no change in the polydispersity (pdi = 1.1 - 1.3) and only the expected increase in molecular weight. However, the more complete set of data plotted in Fig. 7 shows that the degrees of polymerization of poly { 6-[4’-(4’’-ethoxyphenylazo)phenoxyhexyllvinyl ether}s and their precursor polymers are lower than expected from [M],/[I], [214]. In this case, the extent of grafting was less than quantitative of 80-97%. SCLCPs have also been prepared by alkylation and/or quaternization of linear poly(ethy1ene imine)s [215, 2161 and poly(4-vinylpyridine)~[217 - 2231 with mesogenic alkyl halides or carboxylic acids. Although these precursor polymers can be prepared by controlled cationic [224] and anionic [225, 2261 polymerizations, respectively, ‘living’ polymerizations were either not used to prepare the SCLCPs, or the mo-
lecular weight and polydispersity data were not reported for the precursor polymers and the resulting SCLCPs. These systems are therefore not discussed here.
3 Structure/Property Correlations Determined using ‘Living’ Polymerizations 3.1 The Effect of Molecular Weight The most basic and fundamental structure/property question that must be answered with any new system, especially one prepared by a living polymerization, is the molecular weight or degree of polymerization at which its thermotropic behavior no
3
153
S tructure/Property Correlations Determined using ‘Living’ Polymerizations
longer depends on the addition or subtraction of another repeat unit. (However, the isotropization enthalpy of SCLCPs is essentially molecular weight independent.) Comparisons of additional structural variables can then be made between samples whose thermotropic behavior is independent of molecular weight. The thermal transitions of liquid crystalline polymethacrylates reach their limiting values at less than 50 repeat units. For example, Fig. 8 shows that the glass and nematic - isotropic transition temperatures of syndiotactic [(rr)=0.70-0.771 poly{ 6[4’- (4”- met h o x y p h e n ox y c arb on y 1)p h e n oxylhexyl methacrylate} prepared by controlled anionic polymerization levels off at approximately 35 repeat units [45]. This is in contrast to the corresponding polymers prepared by radical polymerizations, which have transition temperatures that are substantially higher and level off at approxmately 25 repeat units [45], in addition to having different end groups and perhaps branched architectures, the radically prepared polymers are also less syndiotactic [ ( r r )=0.40-0.661. However, Fig. 8 also plots the transition temperature(s) as a function of the inverse degree of polymerization in order to calculate the transition temperatures of an infinite molecular weight polymer from the intercepts. Although the transition temperatures plateau as a function of the degree of polymerization, they do not correspond to the transitions of an infinite molecular weight polymer, which is similar to that of the radically prepared polymers. Nevertheless, the rapid increase in transition temperatures upon going from DP, = 20 to DPn=30 indicate that this discrepancy may be due simply to errors in the calculated molecular weights of the low molecular weight polymers, and therefore to an inaccurately high (absolute) slope used for the extrapolation.
-
DP,-
0
0.01
0.02 0.03 0.04
0.05
100
G
80
v
i260
a 2
p 20 0 0
10
20
30 -
40
50
60
DPn Figure 8. Dependence of the glass (0.0)and nematic - isotropic (D,0)phase transition temperatures of syndiotactic [(rr)=O.70-0.77] poly( 6-[4’-(4”-methoxyphenoxycarbonyl)phenoxy]hexyl rnethacry late ) as a function of the number average degree of polymerization ( 0 ,D) and the inverse number average degree of polymerization (0, 0)[45]. Infinite molecular weight transitions: G 44 N 105 I.
The G- SmA-N-I transition temperatures of syndiotactic poly { 6-[4’-(4”-n-butoxyphenoxycarbonyl)phenoxyl)phenoxy]hexyl methacrylate} prepared by aluminum porphyrin initiated polymerizations also level off at approximately 25 repeat units [9 I]. Similarly, the glass and nematic - isotropic transition temperatures of poly[6-(4’met h o x y - 4”- a - met h y 1st i 1be n e o x y )he x y 1 methacrylate] prepared by group transfer polymerization become independent of molecular weight at approximately 20 repeat units [48]. Both polymethacrylates reach the same transition temperatures as the corresponding polymers prepared by radical polymerizations, which have nearly identical tacticities.
154
Ill
Molecular Engineering of Side Chain Liquid Crystalline Polymers
Y
-
-
DPn-l
DPn0 1201
I“
0.1
0
0.2
i I
200
0.1 I
. b
0.2 .
.
.
.
I
c,
G 20
Figure 9. Dependence of the glass ( 0 , 0) and nematic-isotropic (w, 0)phase transition temperatures of (a) poly{ 5-( [6’-[4”-(4’”-methoxyphenyl)phenoxy]hexyl~carbonyl}bicyclo[2.2.l]hept-2-ene)[22] and (b) poly{ 5-[[[2’,5’-bis[(4”-methoxybenzoyl)oxy]benzyl]oxy~carbonyl~bicyclo[2.2.l]hept-2-ene} [ 1821 as a function of the number average degree of polymerization ( 0 ,w) and the inverse number average degree of polymerization (0, 0). Infinite molecular weight transitions: (a) G 41 N 95 I; (b) G 99 N 167 I.
The thermotropic behavior of liquid crystalline polynorbornenes also reach their limiting values at 50 repeat units or less [22, 182, 188-1901. For example, Fig. 9 demonstrates that the glass and nematic-isotropic transitions of both terminally and laterally attached systems level off at 25-50 repeat units, and correspond to the transition temperatures of the infinite molecular weight polymers. The same is true of the crystalline melting and smectic-isotropic transition temperatures of poly { (+)-endo, exo - 5,6- di { [n - [4’- (4”- met hox yp h e n y 1)phenoxy] hexyllcarbonyl } bic yclo[ 2.2.1 ] hept-2-ene) [190]. Although most well-defined SCLCPs have been prepared by cationic polymeriza-
tions of mesogenic vinyl ethers, most of these studies did not prepare polymers of high enough molecular weight to prove that the transition temperatures had leveled off. For example, the highest molecular weight poly { 6-[(4’-cyanophenyl-4”-phenoxy)hexyllvinyl ether] prepared by direct cationic polymerization contains approximately 30 repeat units [124]. Figure 10 compares the thermotropic behavior of these poly(viny1 ether)s to the same polymers prepared by polymer analogous [212, 2131 reactions. The higher molecular weight polymers prepared by polymer analogous reactions confirm that the glass and isotropization transitions saturate at approximately 30 repeat units, although the SmC-SmA transition
3
120
Structure/Property Correlations Determined using ‘Living’ Polymerizations
155
-
p loov
p!
2
80-
fi 60-
A 0
E
g
40-
20
-
v
3 130
k 120
I Figure 10. Dependence of the transition temperatures from the glass ( 0 , O),SmC (+, O ) , SmA (m, 0)and nematic (A,A) phases of poly { 6-[(4’-cyanophenyl-4”phenoxy)hexyl]vinyl ether) prepared by direct cationic polymerization ( 0 ,+, W) [ 1241and by polymer analogous reactions (0, 0,0)[212,213] as a function of the GPC-determined degree of polymerization. Infinite molecular weight transitions: G 3 1 SmC 8 1 SmA 122 1.
has not leveled off yet. Similarly, the crystalline - nematic and nematic - isotropic transitions of poly { 6-[4’-(4’’-ethoxyphenylazo)phenoxyhexyl]vinyl ether} prepared by polymer analogous reactions reach their limiting values at 27 repeat units [211]. The systems shown in Fig. 8 and 9 display very simple thermotropic behavior, with the same phase sequence observed at all molecular weights. However, Fig. 10 exemplifies systems in which the number and type of mesophases vary as a function of molecular weight. This is simply due to the phases having different dependencies on molecular weight. That is, the slope of the glass transition versus degree of polymerization of poly { 6-[(4’-cyanophenyl-4’’-
K
1
Figure 11. Dependence of the transition temperatures from the crystalline (+, 0)and SmA (m, 0)phases of syndiotactic [(rr)= 0.801 poly( n-[4-(4’-methoxypheny1)phenoxylhexyl methacrylate] as a function of the number average degree of polymerization (+, W) and the inverse number average degree of polymerization (0, 0)[441. Infinite molecular weight transitions: K 122 SmA 138 I.
phenoxy)hexyl]vinyl ether} is less than that of the SmC - SmA transition; the SmC mesophase is therefore observed only at higher molecular weights, whereas a nematic mesophase is observed only at very low molecular weights. Figure 1 1 shows a second example in which the SmA-I transition temperature of syndiotactic poly { n-[4-(4’-methoxyphenyl)phenoxy]hexyl methacrylate } increases and levels off an approximately 30 repeat units, whereas the crystalline melting temperature is relatively constant [44]. In this case, the lowest molecular weight oligomer melts directly to the isotropic state. The crystalline melting of isotactic poly { n[4’-(4”-methoxypheny1)phenoxy} hexyloxy methacrylate} fractionated into samples
156
111 Molecular Engineering of Side Chain Liquid Crystalline Polymers
with DP,= 16- 133 (pdi = 1.16- 1.24) is also essentially constant (3.4 “C variation) [421. As explained by Percec and Keller [227], the temperature range of at least the highest temperature mesophase therefore increases with increasing polymer molecular weight due to the greater dependence of isotropization on the degree of polymerization compared to crystalline melting and the glass transition. This is due to a greater decrease in entropy and increase in free energy of the isotropic liquid with increasing molecular weight compared to that of the more ordered phases.
3.2 The Effect of the Mesogen The extensive literature on low molar mass liquid crystals demonstrates that specific mesogens (specific chemical structures) tend to form specific mesophases, which vary somewhat with the length of the flexible substituent. We therefore expect that the type of mesogen, including the terminal substituent(s) and the length of the spacer should be the primary factors determining the specific mesophase(s) exhibited by a given SCLCP. The nature of the polymer
Table 6. Thermotropic behavior of 4-n-alkoxy-4’-cyanobiphenyls [228], poly( n-[(4’-(4”-cyanophenyl)phenoxy)alkyl]vinyl etherls (DP,= 17-32, pdi= 1.09- 1.21) [122- 127, 212, 2131, n-[(4’-(4”-cyanophenyl)phenoxy)alkyl]vinyl ethers [ 122- 1271 and a-ethoxy-w-(4-n-alkoxy-4’-cyanobiphenyl)s [ 122- 1271. Thermotropic behavior ( “ C )
n
1 2 3 4 5 6 7 8 9 10 11 12
K 104 K 102 K 71 K 78 K 53 K 58 K 54 K 54 K 65 K 61 K 71 K 69
(N 86) I (N 91) I (N 64) I (N 76) I N68I N 76 I N75 I SmA 67 N 80 I SmA 76 N 80 I N 84 I N 87 I SmA 89 I
G 81x86 G 64x68 G 47x58 G 38 G 29 G 21 G 23 G 14 G 17 G 21
I I N 77 1 (N 39) I (N 60) I (N 54) I N71 I (N 59) I N 70 I (SmA 61 N 71) I
K 69 K 65 K 65 K 54 K 65 K 56 K 63 K 75 K 69 K 69
N,, 69 SmA 104 SmC 67 SmA 125 SmA 139 SmC 67 SmA 155 SmA 152 K 56 (SmC 49) SmA 156 K 66 (SmC 52) SmA 165
I N 1041 N 88 I N 115 I I I I I I I
(SmA 58 (SmA 41 (SmA 65) SmA 60
I I (N 33) I (N 35) I (N 60) I (N 5 5 ) I N 61) I N SO) I I I
CH2=CH--O-(CH&3
2 3 4 5 6 7 8 9 10 11
K 118 K 79 K 73 K 54 K 65 K 59 K 54 K 63 K 65 K 71
3
Structure/Property Correlations Determined using ‘Living’ Polymerizations
7
K
20
0
0
157
2
4
6
8
0
1 0 1 2
I 1
I
I
I
I
I
0
2
4
6
8
10
n
I
n
1
I
-
K 0
2
4
6
8
1 0 1 2
n
0
K 1
I
I
1
2
4
6
8
I
1 0 1
n
Figure 12. Transition temperatures observed on heating 4-n-alkoxy-4’-cyanobiphenyls[228], poly (n-[(4’-(4”cyanophenyl)phenoxy)alkyl]vinyl etherls (DPn=17-32, pdi= 1.09- 1.21) [122- 127, 212, 2131, n-[(4’-(4”-cyanophenyl)phenoxy)alkyl]vinyl ethers [ 122- 1271 and cx-ethoxy-w-(4-n-alkoxy-4’-cyanobiphenyl)s[ 122- 1271 from the glassy (o),crystalline (O), SmC (+), SmA (I)and nematic (A) states.
backbone, tacticity, polydispersity, etc. should then be secondary factors in enhancing, altering, or disrupting the natural ordering of the mesogenic side chains. If the mesogen and the length of the spacer are the primary factors determining the specific mesophase(s) exhibited by a given SCLCP, then its thermotropic behavior should be predictable by comparison to an appropriate model compound. Three possibilities for an appropriate model compound are shown in Table 6 and Fig. 12 using poly { n-[(4’-(4’’-cyanophenyl)phenoxy)-
alkyllvinyl ether)s [122- 127, 212, 2131 as an example. Not surprisingly, the monomers themselves, which contain olefin endgroups that are very different from the chemical structure of the polymer backbone, are the least appropriate model compounds for poly { n-[(4’-(4’’-cyanophenyl)phenoxy)alkyllvinyl ether}s. That is, the polymer exhibits enantiotropic nematic mesophases at spacer lengths of n = 3 - 5 and enantiotropic SmA mesophases at spacer lengths of n = 5 - 12. In addition, enantiotropic and monotropic SmC mesophases are some-
158
111 Molecular Engineering of Side Chain Liquid Crystalline Polymers
times observed, as well as a higher ordered phase. In contrast, the monomer forms a SmA mesophase only at n = 11, which is monotropic; the nematic mesophases observed at n = 4 - 11 are also primarily monotropic. The more obvious choice for a model compound corresponds to exactly one repeat unit of the polymer. The a-ethoxy-w(4-n-alkoxy-4’-cyanobiphenyl)sexhibit nematic mesophases at n=4-9 and SmA mesophases at n = 8 - 11, and therefore match the thermotropic behavior of the polymer better than the vinyl ether monomer. However the SmA mesophase is enantiotropic only at n = 11, and the nematic mesophase is monotropic at all of these spacer lengths. Compounds which take into account only the mesogen and spacer are also good, if not better, models of the polymers. In contrast to the ethyl ethers, all of the SmA and most of the nematic mesophases are enantiotropic, which means that the melting temperature mimics the relative temperature of the glass transition of the polymer backbone better. However, the nematic mesophase still appears at n = 6 - 11, and the SmA mesophase doen’t appear at n=5-7, 10, 11.
Nevertheless, the poly(viny1 ether)s and all of these model compounds demonstrate that liquid crystals based on cyanobiphenyl mesogens tend to form nematic and SmA mesophases. Model compounds which take into account only the mesogen and spacer also mimic the thermotropic behavior of laterally attached SCLCPs well. For example, Table 7 demonstrates that both 2,5-bis[(4’-nalkoxybenzoyl)oxy]toluenes [228] and all corresponding SCLCPs, such as polynorbornenes [ 1821,laterally attached through a one carbon spacer with 2,5-bis[(4’-n-alkoxybenzyoyl)oxy]benzene mesogens exhibit only nematic mesophases. (However, Sec. 5 of this chapter will demonstrate that smectic layering can be induced in these SCLCPs and their low molar mass analogs by terminating the mesogen’s n-alkoxy substituents with immiscible fluorocarbon segments [193-1951). In order to determine whether or not the alkyl substituent is necessary for low molar mass liquid crystals to mimic the thermotropic behavior of laterally attached SCLCPs, both 1,4-bis[(3’-fluoro-4’-n-alkoxyphenyl)ethynyl]benzenes [ 193, 1961
Table 7. Thermotropic behavior of 2,5-bis[(4’-n-alkoxybenzoyl)oxy]toluenes [228]and poly[ 5-[[[2’,5’-bis[(4”methoxybenzoyl)oxy]benzyl]oxy]carbonyl]bicyclo[2.2.1]hept-2-ene)s (DP, =23- 100,pdi= 1.12-1.24)[ 1821. Thermotropic behavior (“C)
n
K 166 N 252 I K 187 N 248 I K 138 N 209 I K 115 N 206 I K 90 N 178 I K 88 N 173 I
G 98 N 164 I G 92 N 172 I G 83 N 140 I G 73 N 138 I G 60 N 123 I G 56 N 126 I
3
Structure/Property Correlations Determined using ‘Living’ Polymerizations
I59
R=H
I
50
1“1
tnlll-l 0 2 4 6 8 1012 n
0
n
2
4
6
8 1 0 1 2
n
Figure 13. Transition temperatures of 1,4-bis[(3’-fluoro-4’-n-alkoxyphenyl)ethynyl]benzenes[ 193, 1961, 1,4bis[(3’-fluoro-4’-n-aIkoxyphenyl)ethynyl]toluenes[ 1961 and poly [ 5-( [[2’,5’-bis[2-(3”-fluoro-4”-n-alkoxyphenyl)ethynyl]benzyl]oxy]carbonyl]bicyclo[2.2.l]hept-2-ene]s (DP, = 31 -59, pdi = 1.09- 1.60) [ 1971; from the glassy ( O ) , crystalline (0). SmE (A), SmC (+) and nematic (A) states as a function of the length of the n-alkoxy substituents.
and 1,4-bis[(3’-fluoro-4’-n-alkoxyphenyl)ethylnyl]toluenes [ 1961 were synthesized and their thermotropic behavior compared to that of poly { 5-( [[2’,5’,-bis[2-(3”-fluoro4”- n - a 1k o x y p h e n y 1) e t h y n y 1] b en z y 1] oxylcarbonyl } -bicyclo[2.2. IIhept-2-ene) s [197].As shown in Fig. 13, the 1,4-bis[(3’fluoro-4’-n-alkoxyphenyl)ethynyl]benzenes exhibit a SmE-SmC-N-I phase sequence when n = 6- 12. In contrast, the polymers exhibit a nematic mesophase, and those with n = 3 - 12 slowly organize into a crystalline phase which occurs over an extremely narrow temperature range. The 1,4bis[ (3’-fluoro-4’-n-alkoxyphenyl)ethynyl]benzenes are therefore not appropriate model compounds for these polynorbornenes, evidently because they do not take the benzylic spacer into account. In contrast, the corresponding 1,4-bis[(3’-fluoro-4’-n-alkoxyphenyl)ethynyl]toluenes mimic the polynorborne’s thermotropic behavior quite
well. For example, Fig. 13 demonstrates that only a nematic mesophase is observed at most substituent lengths, in addition to a monotropic or an enantiotropic SmC mesophase when n=9- 12. Further comparisons to appropriate model compounds and the effect of increased mesogen density are discussed in the next sections.
3.3 The Effect of the Spacer Increasing the spacer length has much the same effect on the thermotropic behavior of SCLCPs as increasing the length of the flexible substituent has on that of low molar mass liquid crystals. That is, it destabilizes some phases and stabilizes others. For example, just as increasing the length of the flexible substituent depresses the melting
160
I11 Molecular Engineering of Side Chain Liquid Crystalline Polymers
YR :
140
+CH2-iH+
OR
1-I
80 60 40 20
-
0
0 0
2
4
6 n
8
1012
2
4
6 n
8
1012 0
2
4
6 n
81012
Figure 14. Transition temperatures of poly [ (+)-endo,exo-5,6-di( [n-[4’-(4”-cyanophenyl)phenoxy]alkyl]carbonyl]bicyclo[2.2.l]hept-2-ene)s(DP,=41-266, pdi= 1.21 - 1.60) [191], poly( 5-( [n-[4’-(4”-cyanophenyl)phenoxy~alkyl]carbonyl)bicyclo[2.2.l]hept-2-ene]s (DPn=43-290,pdi= 1.08- 1.27)[189] andpoly(n-[(4’-(4”-cyanophenyl)phenoxy)alkyl]vinyl ethers (DP,= 17-32, pdi= 1.09- 1.21) 1122- 127, 212, 2131 as a function of the number of methylenic units in their n-alkyl spacers; from the glassy ( O ) , crystalline (O), SmC (+), SmA (m) and nematic (A) states.
point of low molar mass liquid crystals, increasing the spacer length depresses the glass transition of SCLCPs, and consequently often uncovers mesophase(s) that are not observed without a spacer. Therefore, many of the first SCLCPs synthesized with the mesogen directly attached to the polymer backbone did not exhibit liquid crystallinity [2]. It was only after Finkelmann, Ringsdorf and Wendorff introduced the spacer concept in 1978 [229], that SCLCPs were synthesized routinely. Rather than simply suppressing the glass transition and/or crystalline melting, they attributed the spacer concept to a decoupling of the motions of the mesogens, which want to order anisotropically, from that of the polymer backbone, which tends to adopt a random coil conformation. However, neutron scattering experiments have demonstrated that the polymer backbone is deformed in both nematic and smectic mesophases, and that it is therefore impossible to completely decouple its motions from those of the
mesogens to which it is chemically linked [230]. In spite of the number of well-defined SCLCPs described in Sec. 2 of this chapter, only a few poly(viny1 ether)s and polynorbornenes have been synthesized with a complete and homologous set of spacer lengths. Figure 14 plots the phase diagrams of three polymers with terminally attached 4-(4’cyanopheny1)phenoxy mesogens as a function of the spacer length: Fig. 15 plots the corresponding phase diagrams of three polymers terminally attached with 4-(4’methoxypheny1)phenoxy mesogens. Both figures demonstrate that the glass transition decreases as the length of the spacer increases. This is due to decreased packing density, and is often referred to as internal plasticization. The temperature of crystalline melting of tactic SCLCPs also decreases as the spacer length increases (Fig. 15), albeit in an oddeven alternation as in low molar mass liquid crystals as a function of the length of the
3
Structure/Property Correlations Determined using ‘Living’ Polymerizations
161
+Q+COzR
-
200
80 20
0
6 8 1012 0 2 4 6 8 1 0 1 2 0 2 4 6 8 10 ! n n n Figure 15. Transition temperatures of (presumably) isotactic poly ((&)-endo,exo-5,6-di([n-[4’-(4”-methoxyphenyl)phenoxy]alkyl]carbonyl)bicyclo[2.2.Ilhept-2-ene)s (DP,=30- 173, pdi= 1.20-2.75) [190], poIy(5-[ [n[4’-(4”-methoxyphenyI)phenoxy]alkyl]carbonyl}bicyclo[2.2.l]hept-2-ene)s (DP,=72- 151, pdi = 1.05 - 1.28) [22, 1881 and syndiotactic [(rr)=0.79-0.86] poly( n-[4’-(4”-methoxyphenyl)phenoxy]alkyl methacry1ate)s (DP,= 19-33, pdi= 1.14-2.08) [42, 441 as a function of the number of methylenic units in their n-alkyl spacer; from the glassy ( O ) , crystalline (O), SmA (B) and nematic (A) states. 2
4
flexible substituent. However, without additional order within the polymer backbone itself due to high tacticity (Sec. 3.4 and 3.5 of this chapter), the mesogenic side chains are generally not able to crystallize until the spacer is sufficiently long. Therefore, both poly [n-[(4’-(4”-cyanopheny1)phenoxy)alkyllvinyl ethers [122-1271 (Fig. 14) and poly { 5- { [n-[4’-(4”-methoxyphenyl)phenoxy]alkyl]carbonyl ] bicyclo[2.2.l]-hept-2ene)s [I881 (Fig. 15) undergo side chain crystallization only when n 2 10, or when polymers with singly shorter spacers are of low molecular weight. Depending on the specific mesogen, the nematic mesophase of low molar mass liquid crystals is generally destabilized by increasing substituent length, whereas smectic mesophases tend to be stabilized [228]. That is, long alkyl or alkoxy substituents favor smectic mesophases, whereas short substituents favor nematic mesophases. Figures 14 and 15 confirm that long spacers also favor smectic mesophases, and that short
spacers favor nematic mesophases. These plots also seem to confirm that the transition temperature of nematic disordering decreases slightly with increasing spacer length in SCLCPs, whereas that of the SmA mesophase tends to increase, with at least a slight odd-even alternation. Nevertheless, since the glass transition temperature decreases more rapidly than isotropization with increasing spacer length, the temperature window over which these mesophase(s) are observed increases. Figure 16 shows the change in enthalpy and entropy as a function of the length of the spacer for the three polymer series plotted in Fig. 14. It demonstrates that both the change in enthalpy and entropy of isotropization from the nematic and sA mesophases increase linearly with increasing spacer length. Although such increases in AHi and ASi have been attributed to more efficient decoupling of the mesogen from the polymer backbone and therefore to increased order [231], it simply corresponds to a con-
162
111 Molecular Engineering of Side Chain Liquid Crystalline Polymers
18
16
0
2
4
n
6
8 1 0 1 2
n
Figure 16. Change in enthalpy and entropy of isotropization from the nematic (A)and SmA (0) mesophases of poly { (+)-endo,exo-5,6-di{ [n-[4’-(4”-cyanophenyl)phenoxy]alkyl]carbonyl] bicyclo-[2.2.1]hept-2-ene] s (NBE2-CN, DP,=41-266, pdi= 1.21 - 1.60) 11911; from the nematic (Y)mesophase of poly{4-{ [n-[4’-(4”-cyanophenyl)phenoxy]alkyl]carbonyl]bicyclo[2.2.1]hept-2-ene]s (NBEl.CN, DP,=43-290, pdi= 1.08- 1.27) [189]; and from the nematic (A) and SmA (m) mesophases of poly{ n-[(4’-(4”-cyanophenyl)henoxy)alkyl]vinyl ethers (VE-CN, DP,= 17-32, pdi=1.09-1.21) [122-1271 as a function of the number of methylenic units in their n-alkyl spacers.
3.0
o
T-----l
i
.4
d
2.0
0.5 0.0
0.0
i,
n
0
n
Figure 17. Normalized [232] change in enthalpy and entropy of isotropization from the nematic (A)and SmA (n)mesophases of poly { (+)-endo,exo-5,6,di{ [n-[4’-(4”-cyanophenyl)phenoxy]alkyl]carbonyl]-bicyclo[2.2.1]hept-2-ene)s (NBE2-CN, DPn=41-266, pdi= 1.21 - 1.60) [191]; from the nematic (V) mesophase of poly(5{ [n-[4’-(4”-cyanophenyl)phenoxy]alkyl]carbonyl]bicyclo[2.2.l]hept-2-ene]s (NBE1-CN, DP,,=43 -290, pdi= 1.08- 1.27) [ 1891; from the nematic (A) and SmA (M) mesophases of poly{n-[(4’-(4”-cyanophenyl)phenoxy)alkyl]vinyl ethers (VE-CN, DP,= 17-32, pdi= 1.09-1.21) [122- 1271; from the SmA (0) mesophase of poly { (+)-endo,exo-5,6-di { [n-[4’-(4”-methoxyphenyl)phenoxy]alkyl]carbonyl]bicyclo[2.2.1]hept-2-ene] s (NBE2-OMe, DPn=30- 173, pdi = 1.20-2.75) [190]; and from the nematic ( 0 )and SmA (+) mesophases of poly { 5- { [n-[4’-(4”-methoxyphenyI)phenoxy]alkyl]carbonyl]bicyclo[2.2.l]hept-2-ene] s (NBE1-OMe, DP,, = 72 - 15l , pdi= 1.05- 1.28) [22, 1881 as a function of the number of methylenic units in their n-alkyl spacers.
stant increase in AHi and ASi per methylenic unit, and therefore to the disordering of an additional -CH2- unit of the spacer. The
slope of these lines (AAHi/-CH2- and -AASi/-CH2-) are nearly equivalent for both nematic and smectic disordering, re-
3 Structure/Property Correlations Determined using ‘Living’ Polymerizations
163
Table 8. Normalized [232] changes in enthalpy and entropy of isotropization per methylenic unit in the spacer of poly { (+)-endo,exo-5,6-di { [n-[4’-(4”-cyanophenyl)phenoxy]alkyl]carbonyl] bicyclo[2.2.l]hept-2-ene]s (NBE2-CN) [ 1911, poIy(5- { [n-[4’-(4”-cyanophenyl)phenoxy]alkyl]carbonyl} bicyclo[2.2. IIhept-2-ene)s (NBE1-CN) [189], poly{ n-[(4’-(4”-cyanophenyl)phenoxy)alkyl]vinyl ethers (VE-CN) [ 122- 1271, poly( (k)endo,exo-5,6-di { [n-[4’-(4”-methoxyphenyl)phenoxy]alkyl]carbonyl}bicyclo[2.2.I lhept-2-ene)s (NBE2-MeO) 11901, and poly( 5 - { [n-[4’-(4”-methoxyphenyl)phenoxylalkyl]carbonyl]bicyclo[2.2.l]hept-2-ene]s (NBEI MeO) [22, 1881. SCLCP
NBE2 - CN NBEI -CN VE - CN NBE2 - M e 0 NBE I - M e 0 Mean
AAHi/-CHz- (kJlmru) Nematic
Smectic A
0.044
AAS,/-CHz- (J/K mru) Nematic
Smectic A
0.28
0.12
0.70
-
0.14
-
0.14
-
0.28 0.32
-
0.62 0.76
0.038
-
0.1 1
-
0.048 ?0.008
0.29+0.02
0.13 k0.02
0.69 k0.06
0.05 1 0.058
spectively, of poly { (+)-endo,exo-5,6-di ( [n-
[4’-(4’’-cyanophenyl)phenoxy]alkyl]carbonyl}bicyclo[2.2.l]hept-2-ene} and poly{ n[(4’-(4’’-cyanophenyl)phenoxy)alkyl]vinyl ether, and are half those values of poly { 5{ [n-[4’-(4’’-cyanophenyl)phenoxy]alkyllcarbonyl } bicyclo[2.2.l]hept-2-ene} since the latter has half as many mesogenic side chains. Therefore, if the mesogen density is taken into account for the four amorphous polymer series plotted in Figs. 14 and 15 12321, AAHi/-CH,- and AASi/-CH2are equivalent for the same mesophase (Fig. 17 and Table S), regardless of the type of polymer [poly(vinyl ether), mono- and di-substituted polynorbornene] and regardless of the mesogen [4-(4’-cyanopheny1)phenoxy and 4-(4’-methoxypheny1)phenoxy mesogens]. As expected, Figs. 16 and 17 and Table 8 also demonstrate that the enthalpy and entropy of isotropization from the more ordered smectic mesophase is higher than those from the nematic phase. This discontinuity and/or change in the slope with a change in the type of mesophase can therefore be used as additional confirmation that a phase change has occurred with the addi-
tion or subtraction of one methylenic unit in the spacer of a homologous series. The effect of varying the nature of the spacer from hydrocarbon to siloxane is starting to be investigated using living polymerizations [28]. However, these polymers have not been characterized yet.
3.4 The Effect of the Nature of the Polymer Backbone Since the effect of the nature of the polymer backbone has not been studied extensively using well-defined polymers, it is worth summarizing the observations from less controlled systems based on comparisons of poly(phosphazene), poly(methylsiloxane), poly(viny1 ether), poly(acrylate), poly(methacrylate), pol y(chloroacrylate), and polystyrene backbones [233-2411. In general, the mesogen is more ‘decoupled’ from the polymer backbone as the latter’s flexibility increases. This is because the dynamics of ordering increase with increasing flexibility, which increases the ability of the mesogenic side-chains to form more ordered
164
111 Molecular Engineering of Side Chain Liquid Crystalline Polymers
mesophases and/or crystallize. (The relative dynamics of these SCLCPs correlate directly with the speed with which they form identifiable microscopic textures between crossed polarizers [235]). In addition, the side chains are able to crystallize at shorter spacer lengths as the polymer flexibility increases [236-2381. Therefore, SCLCPs with extremely flexible backbones may form only a monotropic or virtual mesophase, whereas an analogous polymer with a more rigid backbone is more likely to form an enantiotropic mesophase. However, as the flexibility of the backbone increases, the glass transition temperature usually decreases, whereas that of (melting and) isotropization typically increase. Additional mesophases may therefore be revealed if the glass transition is low enough, and if the mesophase is thermodynamically stable relative to the crystalline phase. Secondary structural variables, including the nature of the polymer backbone, will only be elucidated by synthesizing complete and homologous series of well-defined polymers. As discussed previously, only a few of the poly(viny1 ether)s, polynorbornenes and polymethacrylates contain both identical mesogens and spacers. These systems confirm the above trends. For example, when (4’-n-butoxyphenoxycarbony1)phenoxy mesogens are attached to polymethacrylate and poly(viny1 ether) back-
bones through an n-hexyl spacer, the more flexible poly(viny1 ether) forms a highly ordered SmF or SmI mesophase with no evidence of a glass transition (Table 9). In contrast, the less flexible polymethacrylate is glassy. Otherwise, both have nearly identical mesophases and transition temperatures, except poly { 6-[(4’-n-butoxyphenyl4”-benzoate)hexyl]vinyl ether} forms a SmC mesophase (X-ray evidence) in addition to the nematic mesophase, whereas poly { 6-[4’-(4”-n-butoxyphenoxycarbonyl)phenoxy]hexyl methacry late } reportedly forms a SmA mesophase in the same temperature range. Comparison of the phase diagrams plotted in Fig. 14 of poly{5-{[n-[4’-4”-cyanophenyl)phenoxy]alkyl]carbonyl } bicyclo[2.2.1]hept-2-ene}s [ 1891 and poly{n[(4’-(4”-cyanophenyl)phenoxy)alkyl]vinyl
ethers [122- 127, 212, 2131 which contain a single mesogen per repeat unit demonstrates that the glass transition temperature decreases as the flexibility of the polymer backbone increases from polynorbornene to poly(viny1 ether), whereas the isotropization temperature increases. In addition to revealing additional mesophases at lower temperatures, this increase in polymer flexibility enables the poly(viny1 ether)s to form more ordered mesophases. That is, poly{ 5{ [ n - [ 4 ’- ( 4 ”-c y an o p h e n y 1) p he n o x y ] a 1kyllcarbonyl ] bicyclo[2.2.1] -hept-2-ene}
Table 9. Comparison of the thermotropic behavior of poly { 6-[4’-(4”-n-butoxyphenoxycarbonyl)phenoxy]hexy1 methacrylate] and poly [ 6-[(4’-n-butoxyphenyI-4”-benzoate)hexyl]vinylether].
Polymer backbone Methacry late Vinyl ether
DP,,
pdi
Thermotropic behavior (“C)
Ref.
19 38
1.13 1.15
G 38 SmA 104 N 110 I S 47 SmC 106 N 1 1 1 I
[911 [1501
3
S tructure/Property Correlations Determined using ‘Living’ Polymerizations
exhibits only a nematic mesophase at all spacer lengths, whereas the corresponding poly(viny1 ether) exhibits SmA and SmC mesophases with increasing spacer lengths, and also undergoes side-chain crystallization at spacer lengths of 10- 11 carbons, (at which point the SmC mesophases becomes monotropic). When the polynorbornene is substituted with two mesogens rather than one, the internal mobility of the chain should not change, although its overall flexibility and free volume should decrease. As shown in Fig. 14, doubling the mesogen density has very little effect on the glass transition, but causes an increase in the temperature of isotropization, especially at longer spacer lengths. Increasing the mesogen density also enables the mesogens to form smectic layers when the spacer length reaches eight carbon atoms, which is typical of SCLCPs terminally attached with cyano-substituted mesogens [2]. Since the glass transition of a polymer is determined not only by the polymer flexibility, but also by the packing density (free volume), the interactions between chains and the chain length, it is difficult to assess the relative flexibility of two polymers from the glass transition alone. Therefore, it is not immediately obvious from the glass transitions of syndiotactic poly(methy1 methacrylate) (105 “C) and polynorbornene (43 “C) which of the two backbones are more flexible. That is, although polynorbornene has the lower T g , the double bonds and cyclopentane ring along the backbone should be more constrained than the single bonds along a polymethacrylate backbone. The phase diagrams of syndiotactic [(YY) =0.79-0.861 poly{ n-[4’-(4”-methoxyphenyl)phenoxy]phenoxy]alkyl methacrylate [42, 441, poly{ 5-{[4’-(4”-methoxypheny1)phenox y] -alkyl]carbonyl } bicycl0[2.2.l]hept-2-ene [22, 1881 and poly
I65
{ (+)-endo,exo-5,6-di { [n-[4’-(4”-methoxypheny1)phenoxyl alkyl]carbonyl ] bicycl0[2.2.l]hept-2-ene} [ I901 were plotted in Fig. 15. Assuming that the polynorbornene backbone is more rigid than the polymethacrylate backbone, increasing the backbone flexibility increases the ability of the side chains to form more ordered mesophases and crystallize at shorter spacer lengths as expected. However, the more ordered sidechains of the poly { n-[4’-(4”-methoxypheny1)phenoxylalkyl methacrylates may also be due to their high tacticity as discussed in Sec. 3.5 of this chapter. Increasing the mesogen density from one to two mesogenic groups per repeat unit of the polynorbornene backbone again allows the side chains to organize into smectic layers. However, in this case, Mo(CH-tBu)(N-2,6-C6H,-i-Pr2)(OCH,(CF3),), was used as the initiator to polymerize the (k)-endo,exo-5,6-di{[n-[4’-(4”-methoxyphenyl)phenoxy]alkyl]carbonyl } b i cyclo[2.2.1]hept-2-enes instead of Mo(CHt-B~)(N-2,6-C,H~-i-Pr,)(o-t-Bu)~ [ 1901;
MO(CH-~-B~)(N-~,~-C,H,-~-P~~)(OCH, polymerizes (+)-endo,exo-5,6-[disubstituted]bicyclo[2.2.1]hept-2-enes to produce isotactic polymers with approximately 85% cis double bonds, whereas the corresponding Mo(CH-t-Bu)(N-2,6-C6H,-i-Pr,)(O-tBu), polymerizations produce atactic polymers with approximately 95% trans double bonds [242]. Therefore, the poly { (*)-endo,exo-5,6-di { [n-[4’-(4”-methoxyphenyl)phenoxy]alkyl]carbonyl } bicyclo[2.2.1]hept-2-ene}s shown in Fig. 15 are presumably highly isotactic and able to undergo side-chain crystallization, in contrast to the atactic poly { (+)-endo,exo-5,6-di ( [n[4'-(4"-cyanophenyl)phenoxy ]alkyl]carbonyl}bicyclo[2.2.11hept-2-ene)s presented in Fig. 14. Although a more ordered mesophase is formed by the more tactic polymers, the increased ability of the side chains to
166
111 Molecular Engineering of Side Chain Liquid Crystalline Polymers
E 0
n
2
4
6
n
8 1012
n
Figure 18. Transition temperatures from the glass ( O ) , crystalline (0), nematic (A)and smectic (w) phases of poly( 5-1[n-[4”-(4’”-nitrostilbeneoxy)alkyl]carbonyl] bicyclo[2.2.l]hept-2-ene]s (DP,,= 27-42, pdi = l .08- l . l l ) [187], 4-n-alkoxy-4’-nitro(stilbene) (2281 and poly[3-[((n-(4’-(4”-nitrostilbeneoxy)alkyl)carbonyl)methyleneoxy)methyl]cyclobutene)s (DP,,=61-72, pdi= 1.14-1.16) [187] as a function of the number of methylenic carbons in their n-alkyl spacers.
crystallize also results in much narrower temperature ranges over which the liquid crystalline phase is observed. A final example demonstrating that more ordered phases are possible when more flexible polymer backbones are used is shown in Fig. 18. Although this is not a complete series with consecutive spacer lengths, it does demonstrate that the more flexible polybutadiene backbone allows n-[4’-(4”nitrostilbeneoxy)alkyl] side chains with 6- 12 carbons in the spacer to organize into smectic layers, whereas the corresponding polynorbornenes exhibit only nematic mesophases [ 1871. This is especially interesting since the corresponding low molar mass liquid crystals transform from a monotropic nematic phase when the alkoxy substituent contains six carbons, to an enantiotropic smectic and nematic sequence at 7 - 8 carbons, and finally to only enantiotropic smectic at ten carbons [228]. The corre-
sponding polymers with less than six carbons in the spacer (as well as the rest of the series) should therefore by synthesized and characterized to determine to what extent the polymer backbone overrides the natural ordering tendency of the side chains. As discussed in Sec. 3.3 of this chapter, the change in enthalpy and entropy of isotropization increases linearly with increasing spacer length. The slopes of these lines (AAHi/-CH2-- and AASi/-CH2-) therefore correspond to the disordering of one methylenic unit of the spacer, and are evidently equivalent for a given mesophase if they are normalized [232] over the mesogen density. This is true regardless of the type of polymer [polynorbornene or poly(viny1 ether)] and regardless of the mesogen (p-cyanobiphenyl or p-methoxybiphenyl). Since the slope of the normalized changes in enthalpy and entropy of isotropization are equivalent for a specific mesophase, the degree
3
Structure/Property Correlations Determined using ‘Living’ Polymerizations
167
Table 10. Changes in enthalpy and entropy of isotropization of poly ((+)-endo-exo-5,6-di ( [n-[4’-(4”-cyanophenyl)phenoxy]alkyllcarbonyl} bicyclo[2.2.1 lhept-2-ene } s (NBE2 - CN) [ 19 11, poly( 5-1 [n-[4’-(4”-cyanophenyl)phenoxy]alkyl]carbonyl }bicyclo[2.2.llhetp-2-ene)s(NBEl -CN) [ 1891, poly( n-[(4’-(4”-cyanophenyl)phenoxy)alkyl]vinyl ethers (VE-CN) [122- 1271. SCLCP
n
AH, (kJ/mru)
AS, (J/K mru)
Ti (“C)
Mesophase
NBE2 - CN NBEl -CN VE - CN NBE2 - CN NBEl -CN VE-CN
4 4 4 5 5 5
0.492 0.312 0.03 14 0.596 0.322 0.262
1.27 0.830 0.0870 1.56 0.884 0.675
114 103 104 110 91 115
Nematic
NBE2-CN VE - CN NBE2 - CN VE - CN NBE2 - CN VE - CN
9 9 10 10 11 11
1.58 1.26 2.00 1.63 2.53 1.61
4.05 2.96 5.17 1.70 6.46 1.68
116 152 1 15 156 118 165
Smectic A
of order of at least the spacer, and presumably the mesogen, within a specific mesophase must be equivalent for the five polymer series summarized in Fig. 17 and Table 8. However, the different absolute values (e. g. the intercepts) of these lines indicate that the disordering of the spacer is not independent of the rest of the polymer, and instead varies with the polymer backbone to which it is chemically linked. The data in Table 10demonstrates that for a constant spacer length and mesophase, both the change in enthalpy and entropy of isotropization decrease as the flexibility of the polymer backbone increases from polynorbornene to poly(viny1 ether). However, the change in entropy decreases more rapidly than the change in enthalpy, and the isotropization temperature (Ti = AHi/ASi) therefore increases with increasing flexibility. Since lower entropies of fusion are associated with more rigid structures, the lower entropy of isotropization of poly(viny1 ether)s is obviously not due to a lack of inherent flexibility of its polymer backbone, but rather to the more flexible backbone being more ordered and therefore more
constrained (anisotropic) within the mesophase, as was recently demonstrated by neutron scattering experiments [230].
3.5
The Effect Of Tacticity
The effects of tacticity were first observed and discussed in 1981 [243-2451. However, relatively little effort has been made in examining the effect of tacticity on well-defined SCLCPs prepared through living polymerizations. Although these limited results are contradictory, we believe that trends are emerging which indicate that high tacticity in SCLCPs will cause either of two effects. In most systems, the high tacticity evidently places the mesogenic side groups in the proper configuration to crystallize and/or form more orderes mesophases. However, if the mesogenic groups are not in the proper configuration to form ordered phases, the primary effect of increasing the tacticity is to decrease the flexibility of the polymer backbone relative to that of the atactic polymer.
168
111 Molecular Engineering of Side Chain Liquid Crystalline Polymers
Table 11. Comparison of the thermotropic behavior of isotactic and syndiotdctic poly (n-[4’-(4”-methoxyphenyl)phenoxy]alkyl methacry late) s.
n
2 2 3 3 4 4 5 5 6 6 6
GPC
Phase transitions (“C)
%
DP,
pdi
(mm)
(mr)
(rr)
37 23 68 33 24 26 36 19 20 28 33
2.35 1.43 2.80 2.08 3.22 1.23 14.0 1.57 2.21 1.41 1.14
94 2 97 5 95 2 96 2 90 1
6 19 3 12 4 12 3 18 9 14
0 79 0 83
-
-
Discussion of the effect of the polymer backbone in Sec. 3.4 of this chapter already provided examples of highly isotactic poly { (k)-endo,exo-5,6-di{[n-[4’-(4”-methoxyphenyl)phenoxy]alkyl]carbonyl } bicyclo[2.2.l]hept-2-ene}s [190] and syndiotactic poly { n-[4’-(4”-methoxyphenyl)phenoxylalkoxy methacry1ate)s [42,44] which crystallize and form more ordered mesophases than those of the corresponding atactic polymers (Fig. 15). Although more flexible backbones are more able to achieve the conformation necessary to order, the side chains are evidently already attached to the polymer backbone of these tactic polymers with the proper configuration to order, which obviates the need to distort their conformation for such purposes. As discussed in Sec. 2.1 of this chapter, isotactic poly { n-[4’-(4”-methoxyphenyl)phenoxylalkyl methacrylate} s have also been prepared, although the isotactic polymerizations are much less controlled than the syndiotactic polymerizations [42, 441.
1
86 1 80 1 85 = 80
K K K K
Ref.
K 193 I K 142 I K 147 I K 1831 K 141 I 133 SmA 145 I K I45 I 160 SmA 167 I K 1361 116 SmA 131 I 122 SmA 135 I
The thermotropic behavior of both the isotactic and syndiotactic poly { n-[4’-(4”methoxyphenyl)phenoxy]alkyl methacrylate}s is summarized in Table 11. All of the tactic polymers crystallize. With the exception of poly { 2-[4’-(4”-methoxypheny1)phenoxylethyl methacrylate} ( n = 2), the melting temperature of the isotactic polymers is almost independent of the spacer length. In contrast, the syndiotactic polymers melt with a large odd-even alternation. However, only the syndiotactic polymers with at least four carbons in the spacer exhibit an enantiotropic smectic mesophase, which occurs over only a very narrow temperature range. The greater order of the isotactic polymers is evidently due to the greater segmental mobility of isotactic versus syndiotactic polymethacrylate backbones [246, 2471. Surprisingly, the mesogenic groups of syndiotactic [ ( r r )=0.70-0.781 poly{ 6-[4’(4”-methoxyphenoxycarbonyl)phenoxy]hexyl met ha cry late}^ are not in the proper
3
169
Structure/Property Correlations Determined using ‘Living’ Polymerizations
Table 12. Comparison of the thermotropic behavior of poly { 6-[4’-(4”-methoxyphenoxycarbonyl)phenoxy]hexyl methacrylate} with different tacticities.
~
Polymerization mechanism Radical
GPC
UPn
pdi
(mm)
(mr)
(rr)
48
1.51 3.65 3.10 1.23
-
1
41 34
59 65
-
34
66
G38N112I
-
-
-
G 41 SmA 72 N 11 1 I
-
44
so
28 28 22
12 72
G 42 N 107 I G 39 N 82 I G 38 SmA 64 N 105 1 G 4 0 N 1071
84 290 953 GTP Anionic (TPP)AIMe ’‘ (TPP)AIMe-ALC’
‘’
~~
Phase transitions (“C)
C/c
42 31 21 30
1.71 1.24
~
1.31
-
1.43
-
78
Ref.
G 39 SmA 72 N 107 I
G 4 0 N 1071
Methyl(5, 10,15,20-tetraphenylporphinato)aluminum.
’ Methyl(5,10, 15,2O-tetraphenylporphinato)aluminumand methylaluminum bis(2,4-di-r-butylphenolate). configuration to crystallize [45, 901, and the primary effect is therefore that of decreased flexibility compared to the atactic polymer. As discussed in subsection 3.4, more rigid backbones suppress side-chain crystallization and the formation of more ordered mesophases. In addition, the glass transition temperature increases whereas that of isotropization tends to decrease. However, as shown in Table 12, the glass transition temperature of poly ( 6-[4’-(4”methoxyphenoxycarbon y1)phenox y] hex yl methacrylate} is essentially independent of tacticity. With the exception of the polymer prepared by anionic polymerization [45], the nematic -isotropic transition also appears to be independent of tacticity. However, as demonstrated by the data in Fig. 8, the extrapolated transition temperatures (G 44 N 105 I) of an infinite molecular weight polymer prepared by anionic polymerization are nearly identical to those of the rest of the polymers presented in Table 12.
Although the more ordered SmA mesophase is not detected in two of the three more syndiotactic poly { 6-14’-(4”-methoxyphenoxycarbonyl)phenoxy]hexyl methacrylate)s, it is also not detected in some of the less tactic polymers prepared by free radical and group transfer polymerizations. These results are therefore inconclusive as to whether or not increasing syndiotacticity and decreasing flexibility suppresses the formation of the more ordered SmA mesophase in poly { 6-[4’-(4”-methoxyphenoxycarbonyl)phenoxy]hexyl methacry late } s.
3.6 The Effect of Polydispersity One of the most fundamental questions still open regarding the structure/property relationships of SCLCPs is the effect of polydispersity. In some cases, the temperature and nature of the mesophase(s) depend not
170
I11 Molecular Engineering of Side Chain Liquid Crystalline Polymers
only on the degree of polymerization, but also on the molecular weight distribution [147- 149, 2121. For example, the sc mesophase of the broad polydispersity (pdi = 1.9) poly { 6-[ (4’-cyanopheny1-4’’-phenoxy)hexyllvinyl ether] prepared by direct polymerization was not detected by DSC, although it was readily apparent in samples prepared by polymer analogous reactions (pdi = 1.2) [212]. However, broad polydispersity is generally believed to manifest itself in broad phase transitions. In contrast to low molar mass liquid crystals (LMMLCs) which undergo phase transitions over a few degrees, SCLCPs often exhibit extremely broad transitions. This wide (and even split [249, 2501) biphasic region, in which a mesophase coexists with either the isotropic melt or another mesophase [25 11, is evidently due to the fact that polymers are polydisperse. For example, fractionated samples of poly { 6-[(4’-cyanophenyl-4”-phenoxy)undecyllacrylate) undergo (SmC- SmA, SmA -I) transition over a narrower temperature range (15 - 25 “C) than the original polymer (40 “C) prepared by free radical polymerization [252]. (Unfortunately, the actual molecular weight and polydispersity were not determined for any of the samples.) In contrast, the transitions of most welldefined SCLCPs prepared by controlled polymerizations are relatively narrow. The effect of polydispersity was therefore investigated by blending well-defined (pdi < 1.28) poly{ 5- { [6’-[4”-(4’”-methoxypheny1)phenoxyl alkyl]carbonyl}bicycl0[2.2.1]hept-2-ene}s of varying molecular weights (DPn=5, 10, 15, 20, 50, 100) to create polydisperse samples (pdi = 2.504.78) [22]. In this case, both monodisperse samples and multimodal blends underwent the nematic -isotropic transition over a narrow temperature range. Polydispersity also had no effect on the temperatures of transi-
tions, which were determined simply by the number average degree of polymerization. Since polydispersity created by blending linear polymers has no effect on the thermotropic behavior of polynorbornenes [22], whereas fractionation of polyacrylates prepared by radical polymerization results in narrower biphasic regions [252], we must consider what the various sources of polydispersity are. Most SCLCPs are still prepared by free radical polymerizations. Since mesogenic monomers are highly functionalized with a number of sites which can cause chain transfer, their free radical polymerizations should result in chain branching and broad polydispersity at high monomer conversion. In addition, entanglements are trapped in high molecular weight polymers. Therefore, the broad phase transitions of SCLCPs may be caused by the immiscibility of a mixture of molecular architectures resulting from chain branching at the end of the polymerization [23], and/or by entanglements; only chain branching increases the molecular weight distribution. This argument is supported by recent theoretical and experimental studies of polyethylene. That is, while polydisperse blends of linear polyethylene are homogeneous [253], blends of linear and branched polyethylene phase segregate [254 -2581 with the immiscibility increasing as the branch content of the two components becomes more dissimilar [259, 2601. In addition, entropic corrections to the Flory - Huggins theory predict that mixtures of linear polymer and low concentrations of branched polymer will phase separate [26 11.
4
Chain Copolymerizations
171
4 Chain Copolymerizations Polymers which contain more than one type of monomeric repeat unit are called copolymers. By copolymerizing two or more monomers in varying ratios and arrangements, polymeric products with an almost limitless variety of properties can be obtained [3]. As shown in Scheme 18, there are four basic types of copolymers as defined by the distribution of, for example, comonomers A and B. While the properties of a random copolymer are intermediate between those of the two homopolymers, block and graft copolymers exhibit the properties of both homopolymers. The properties of an alternating copolymer are usually unique. The first three types of copolymers can be prepared by polymerizing a mixture of the two monomers simultaneously. In this case, the distribution of comonomers is determined by their relative concentrations and reactivity ratios. The reactivity ratios ( r l , r2, etc.) are the ratios of the rate constants of homopropagation and cross-propagation (Eq. 32). q = - ,kl 1
r 2 = k22 -
kl2
k2 1
(32)
Using the terminal model of copolymerization, there are four possibilities for propagation (Scheme 19). RANDOM
poly(A-ran-B)
BLOCK
poly(A-Mock-B)
ALTERNATING
poly(A-alt-B)
G
poly(A-groff-ft-B)
W
Scheme 18
Scheme 19
If there are only two active sites (MT and M;) whose concentration are at steady state and if monomer is consumed entirely by propagation, the molar ratio of the two monomer units in the polymer (d[M,]/d[M,]) is defined by Eq. (33), in which [M,] and [M2]are the concentrations of the two monomers in the polymer feed.
Random copolymers in which the comonomer distribution follows Bernoullian or zero-order statistics are formed by ideal copolymerizations in which r1r2 = 1. However, a truly random distribution of the two units results only if the Bernoullian distribution is symmetric (i. e. when r1= r2 = 1). In this case, the two monomers have equal probability of reacting with a given active center, regardless of the monomer it is derived from, and the copolymer composition equals the comonomer feed composition at all conversions. Random copolymers are generally formed by radical copolymerizations, whereas ionic copolymerizations tend to favor propagation of one of the comono-
AABAABBBABBAAABABAABB AAAAAAAAAAABBBBBBBBBB ABABABABABABABABABABA AAAAAAAAAAAAAAAAAAAA
m
g
m m
m m
m m
m
m g m m
172
111 Molecular Engineering of Side Chain Liquid Crystalline Polymers
mers much more than the other, yielding blocky sequences of that comonomer. That is, it is usually difficult to copolymerize monomers by an ionic mechanism due to large differences in their reactivity ratios: r l % 1, r,41. Alternating copolymers result when both reactivity ratios are close to zero due to little tendency of either monomer to homopolymerize ( r , = rz= 0). Block copolymers containing long blocks of each homopolymer in a diblock, triblock, or multiblock sequence are formed by simultaneous polymerization of the two comonomers only when r1S- 1 and r2% 1. However, block copolymers are more effectively prepared by either sequential monomer addition in living polymerizations, or by coupling two or more telechelic homopolymers subsequent to their homopolymerization. Alternatively, if the two monomers do not polymerize by the same mechanism, a block copolymer can still be formed by sequential monomer addition if the active site of the first block is transformed to a reactive center capable of initiating polymerization of the second monomer. Living polymerizations have been used to prepare copolymers based on either two mesogenic monomers, or one mesogenic monomer and one nonmesogenic monomer, with the comonomers distributed either statistically or in blocks or grafts.
4.1 Block Copolymers Block and graft copolymers based on two or more incompatible polymer segments phase separate and self-assemble into spatially periodic structure when the product (xN)of the Flory - Huggins interaction parameter and the total number of statistical segments in the copolymer exceeds a critical value [25, 262, 2631. Since the incompatible
I
0
I
I
I
I
0.2
0.4
0.6
0.8
Weight fraction of PS Figure 19. Morphology of poly(styrene-block-butadiene) as a function of molecular weight and composition [ 2 5 ] .
blocks are chemically linked to each other, macroscopic phase separation is prevented, and phase separation is instead limited to the microscopic level. The resulting morphology minimizes the free energy of the system by minimizing unfavorable contacts and the interfacial curvature energy, while maximizing the conformational entropy and maintaining a uniform segment density throughout the phase [262 - 2661. The geometry and size of the microdomains are therefore determined not only by the degree of incompatibility x N , but also by the volume fraction (@A, #B) of the different blocks, the block length(s), the total molecular weight, the conformational asymmetry and fluctuations [262-2711. An isothermal morphology diagram of poly(styrene-block-butadiene) is shown in Fig. 19 as a function of molecular weight and copolymer composition; the classic morphologies include spherical microdomains (O 0 A
P,< 0
--
P, = PI (k A n )
(b)
Figure 11. (a) C2 symmetry operation applied to a single SmC* layer, and (b) sign convention for P,.
218
IV
Behavior and Properties of Side Group Thermotropic Liquid Crystal Polymers
However, due to the helical structure of the SmC* mesophase, no net spontaneous polarization results. Macroscopic nonvanishing polarization is only obtained when the helix is unwound. This requirement has been a severe limitation for the determination of P,, and we had to wait until 1988 for the first observation of ferroelectric switching in a side chain SmC* LCP [65]. Even now, the measurement of P, remains delicate and its reliability is directly connected to the quality of the uniform alignment achieved.
sured as a voltage drop across a series resistance of several kiloohms. This technique requires the applied voltage be sufficiently high and its frequency sufficiently low (lower than the inverse of the response time of the FLCP) in order to obtain saturation of the measured value of P,. On the other hand, the presence of ionic impurities in the FLCP may disturb and even prevent the P, measurement. Finally, because of the anchoring effect of the alignment layers, the P, value is underestimated when the cell thickness is insufficient [ 141.
2.3.1 Uniform Alignment
2.3.3 Spontaneous Polarization Behavior of FLCPs (Ferroelectric Liquid Crystal Polymers)
The general procedure for generating uniform alignment is as follows: the sample under study is sandwiched between two I T 0 glass plates which are coated with polyimide or nylon and rubbed to favor a planar alignment. Very often, the test cells are filled with the polymer in an isotropic state, by capillary forces. Good planar alignment is achieved by repeated heating and cooling of the sample from the SmA (or isotropic) to the SmC* phase while applying an AC electric field of low frequency (several hertz). Sometimes the polymeric film is sheared in the SmC* phase to promote the expected alignment. The cell gaps are generally in the range of a few micrometers [2].
2.3.2 Experimental Methods for Measuring the Spontaneous Polarization Several methods exist for measuring the magnitude of the spontaneous polarization [70], but the most widely used is certainly the electric reversal of the polarization obtained when a triangular voltage is applied to the cell. The polarization current is mea-
Ferroelectric LCPs obey the general rules known for low molecular weight SmC* liquid crystals, namely, that the larger the transverse dipole and the smaller its distance from the chiral center, the higher the spontaneous polarization P,. Due to a slight increase in the rotational hindrance of the mesogens around their long axes as a result of their link to the polymer backbone, the values of the spontaneous polarization for the FLCPs tend to be higher when compared to the corresponding low molecular weight SmC* liquid crystals [45]. The reported value of Ps=420 nC/cm2 (at T = 151 "C) for polymer VI (Fig. 12) is very likely the highest value so far achieved with FLCPs [42]. The influence of the polymer molecular weight on the spontaneous polarization has been studied for a polyacrylate [21] as well as for a polysiloxane series [ 181. The conclusions are somewhat different. Polyacrylates V- 11 (Table 3) show a decrease of their spontaneous polarization as their molecular weight increases (Fig. 13a), whereas polysiloxanes VII (Table 5 ) exhibit a little de-
2
Ferroelectric Liquid Crystal Polymers' Behavior
219
+
'+ 0
CH3-
Si-(CH,),,O
@@CQO~C&,~ CH3 NO2
G 21 K 57 Sc'
n
0'1)
Figure 12. Chemical structure of polysiloxane VI having a very high P, [42]; K: Cr.
161 S , 183 L
50
1
N
40
u
30
20
10
0
130
150
140
160
(4
T
170 C"C1
fi
Y
0 1 '
300
' .
310
I
I
320
*
'
330
m
340
1'
I
350
T /K pendence of their P, on molecular weight (Fig. 13b). A possible explanation of these results could lie in the difference of the viscosity of the samples, leading to some differences in the quality of the obtained uniform alignment required to measure the P,.
360
370
Figure 13. (a) Temperature dependence of the spontaneous polarization for samples of polyacrylates V- 1 1 of different molecular weight [211. (b) Temperature dependence of P, of polysiloxanes VII [18]. OM,=49300; 0 M, = 12 700; A M, = 8000; A 2, = 5600.
Further investigations seem necessary to clarify this point. Study of the variation of the spacer length in polyacrylates V-n has shown that the spontaneous polarization increases when the spacer is shortened [21]. This could be
220
IV
Behavior and Properties of Side Group Thermotropic Liquid Crystal Polymers
Table 5. Molecular weights and phase transitions for polysiloxanes VII. a
3
Polymer VII- 1 VII-2 VII-3 VII-4 a
M, (g/mol)
M, llGn
Transition temperatures ("c)
5 600 8 000 12700 49 300
1.08 1.16 1.19 1.82
G 42 a SmC* 90 SmA 110 L G 30 SmC* 89 SmA 120 L G 30 SmC* 86 SmA 122 L G 50 SmC* 88 SmA 132 L
SmX-SmC* transition temperature, SmX: high order smectic phase.
consistent with a decrease of the side chains mobility. The introduction of a second chiral center of the same absolute configuration R into the spacer of polymer V-1 1 causes an important shift of the SmC* phase of about 75 "C to lower temperatures and a decrease by half of the spontaneous polarization (Fig. 14). To prove whether there is an effect of partial compensation between the two asymmetric centers, with respect to the polarization, the configuration of the chiral center in the spacer was changed from R to S. It turned out that the ferroelectric properties were completely lost [21]. Cooling down from the SmC* to a more ordered phase should lead to an increase of
/r
the spontaneous polarization. This behavior has not been observed in several homopolymers [64, 651 (Fig. 13) nor in mixtures of a racemic SmC polyacrylate with a chiral low molecular weight dopant 1541. The recorded decrease in P, could be due to an increase of the stiffness of the polymer. The phenomenon of sign reversal of P s , which has already been demonstrated by Goodby et al. in a low molecular weight SmC* liquid crystal 1711, can also occur in an SmC* polymer 1211. Polymer IX is the first example of an FLCP having such a behavior (Fig. 15). This phenomenon is attributed to a temperature-dependent equilibrium between different conformers of the chiral side chains, which have opposite sign of P,.
(VIII)
CH,
Figure 14. FLC polymer VIII with two remote asymmetric carbons in its side chain; S,*: SmI*.
4CH2
I
CH -COO(CH,),$H
-$HCH,O
Y R.T.Sx 57 S,' or S', 122'S,
175 L
Figure 15. FLC polyacrylate exhibiting a sign reversal of P, phenomenon at around 132°C; S*,:SmF*.
2
Ferroelectric Liquid Crystal Polymers' Behavior
diluted polysiloxane XI [40]. For small electric fields (3-5 V,, pm-'), P, goes to zero at the ferro-to-para electric transition temperature, as expected. But, at higher applied fields (15-20 V,, pm-l), a spontaneous polarization can also be measured well above the phase transition in the SmA phase (Fig. 17). Two interpretations of this effect are suggested by the authors: either a shift in the phase transition temperatures to higher
Concerning the 'diluted' polysiloxanes, it appears that the spontaneous polarization is not directly proportional to the chiral mesogen content in one case [31] or to the volume percentage of mesogen in another case [40]. The same tendency is also observed in mixtures of the racemic polyacrylate X with a chiral dopant (Fig. 16) [54]. Finally, close to the SmC*-SmA transition, a voltage dependence of the spontaneous polarization has been observed for the
6
22 1
--
5--
Figure 16. Dependence of the maximum of the P, of binary mixtures of X and D- 1. I
0
3
2
5
4
0
8
7
0
1011IZ
1 3 1 4
I5
molX
180
135
,
-
,
,
,
,
,
.
.
,
,
.
3
.
.
I
.
.
, , I " .
I
=c* '
&
A
A
"
I
.
SA*
A
A A A
b2
-
90
-
0
35vwm
A
15-20Vwlpm
AA>.. . ' 0 0
0 4
?\
nu
3,
0
45
0 1
-
O
A
o
o
0
~
~
~
Q
b
A
~ AA
"
'
~
Figure 17. Temperature depen~ dence of the~spontaneous polariza-
~
222
Behavior and Properties of Side Groiup Thermotropic Liquid Crystal Polymers
IV
temperatures (phase induction), or a combination of the electroclinic and the cone switching mechanisms allowing the observation of P,.
2.3.4 Tilt Angle and Spontaneous Polarization The absolute values of the spontaneous polarization P, and the tilt angle 8 as a function of temperature are well fitted to the following power law equations
8 = 8,(T, - T)"
(1)
P, = Po(Tc- T)""
(2)
Up to now, a' and a" have only been determined for the two polysiloxanes XI1 and XIII, whose structures are depicted in Fig. 18 [12, 141. As shown in Table 6, a' and a" are respectively of the same order of magnitude for polymers XI1 and XIII. It has to be pointed out that in the case of low molecular weight SmC* liquid crystals, both d and d'
are found to be significantly lower than 0.5 [72, 731. With a' being different from a" for both polymers XI1 and XIII, the relationship between P, and 8 is nonlinear. Such behavior is typical of ferroelectric liquid crystal materials with high P,, and can be explained on the basis of the generalized Landau model for the free energy density. A complete treatment is available for polymers XI1 and XI11 and the different calculated coefficients. Depending on the T,-T value, two different approximate linear relationships between P, and 8 can be derived. When T is close to T, D
S-&C 8
(3)
when T 4 T C
(4) where L?and C are respectively the coefficients of biquadratic and piezoelectric coupling between P, and 8, E is the high frequency permittivity (here, E = 2.7 x 10" C2 N rnp2), and 17 is a coefficient introduced to stabilize the system. L2 and C have been found to be significantly lower for copolymer XI11 than for homopolymer XI1 (Table 6).
Sx 68 S G 88 S , 156 L
A'.' 0 I -SI-CH,
Table 6. Exponents of the power law equations fitting the tilt angle and the spontaneous polarization of polysiloxanes XI1 and XIII, as well as their piezoelectric ( C ) and biquadratic (Q) coefficients.
O(m
4 A'
Polymer
0
I
-SI-(CH~)~~O
W
O
O
C
Y
4
P
.
,
* . *. CI CH,
Sx29 S,'
61 S, 89 L
Figure 18. Structures of polymers XI1 and XIII.
XI1 XI11
a'
a"
C(V/m)
0.35 0.35
0.5 0.48
6 . 5 ~ 1 0 ~5 . 3 ~ 1 0 " 3 . 6 ~ 1 0 ~2 . 0 ~ 1 0 "
52(m/F)
2 Ferroelectric Liquid Crystal Polymers' Behavior
2.4 Electrooptic Behavior
223
tion [74] I,
Knowledge of the electrooptic behavior of the FLCPs is of the utmost importance for display device applications. One relevant parameter in this respect is the response time. As for the spontaneous polarization, the determination of the response time requires a uniformly aligned sample. The test cell is placed between crossed polarizers so that one tilt direction is parallel to the direction o f one polarizer. The electrooptic effect is achieved by applying an external electric field across the cell, which switches the side chains from one tilt direction to the other as the field is reversed. A photodiode measures the attenuation of a laser beam when the cell is switched between the two states. Generally, the electrooptical response time is defined as the time corresponding to a change in the light intensity from 10 to 90% when the polarity of the applied field is reversed ( t10-90). Several switching processes have been identified for FLCPs: ferroelectric switching, corresponding to the rotation of the side chains around the cone angle (the Goldstone mode), - electroclinic switching, where a change of tilt angle 8 is observed (the soft mode), and - antiferroelectric switching. -
where E is the external electric field. As there is no appropriate method to measure directly the rotational viscosity of ferroelectric liquid crystals, y is generally deduced from the electrooptic response time measurements [ 12, 18,441. The relationship between t,Op,, and z is not straightforward and requires the use of a theoretical model for the optical transmission based on the 'bookshelf geometry' briefly summarized in the following. The transmission T o f the SmC* cell placed between crossed polarizers is given as
T = I = sin2[ 2 ( 0 - 8 ) ]sin2 lo ~~
where 0 is the angle between the polarizer and the layer normal and 8 is the apparent tilt angle expressed by (see Fig. 19) 8 = tan-' (tan o0 cos cp)
(7)
An is the effective optical anisotropy, given by ~ne_no_ _ _ _ _ no An = (8) (nosin2 ~p + n, cos2 ~ p > " . ~ where Cp is the angle between the light ray direction and the director cos Cp = sin 8, sin cp
z through
(9)
2.4.1 Ferroelectric Switching
T is related to angle cp
the azimuthal
Ferroelectric switching is the most likely process encountered in FLCPs. A solution of the equation of motion of the SmC* director, when the dielectric anisotropy is supposed to be negligible, leads to the introduction o f a switching time z which is related to the rotational viscosity y and the spontaneous polarization P, by the following rela-
where cpo=cp(t=O)and cp, 8 and Cp are timedependent. Fitting the experimental optical response and Eq. (6) allows the determination of cpo and the rotational viscosity yprovided P, is known.
224
IV
Behavior and Properties of Side Group Thermotropic Liquid Crystal Polymers
I
If
1
Am'yser
I
I
tz
'sidec~ham
I
d
Figure 19. Definition of the different angular parameters involved in ferroelectric switching.
Using this approach, Endo et al. [18] established that the rotational viscosity of the SmC* polysiloxanes VII (Table 5 ) was approximately proportional to the second power of M, at 60 "C
Y = mi,'
Smectic layer
(1 1)
Typical response times of SmC* FLCPs are of the order of a few milliseconds to a few hundred milliseconds, depending (through the viscosity) on the temperature, the chemical nature of the polymer backbone, and the molecular weight of the polymer. The switching time z increases with the molecular weight and decreases as the temperature increases. On the other hand, the response time measured at the same reduced
temperature decreases with the flexibility of the polymer backbone in the order Ypolysiloxane < Ypolyoxyethylene < Ypolyacrylate
r7i
Fast switching times are also observed when the spacer length increases, thus decreasing the mesogen-backbone coupling [ 3 1,751. Practical applications of FLCPs require the retention of fast response times at room temperature. Attempts to reduce the response time by increasing the spontaneous polarization failed, since introduction of strong lateral dipole moments in the side chains increases not only P, but also the viscosity, the overall result being an increase in z.Hence the following materials have been developed to reduce the viscosity of
2
Ferroelectric Liquid Crystal Polymers' Behavior
the FLCPs: 'diluted' polysiloxanes [40,75], binary mixtures of FLCPs and low molecular weight liquid crystals [3 I], binary mixtures of SmC LCPs and chiral dopants [24], and SmC* oligosiloxanes [39]. Response times as short as 400 ps at 130 "C (7.5 Vpp/pm) have been obtained with an oligosiloxane [39] and, on the other hand, a mixture of diluted polysiloxane with a low molecular weight liquid crystal has led to a response time of 0.5 ms at 40°C [31].
where p is a structure coefficient, a is the first constant in the Landau free energy expansion, and T, is the SmX-SmA* transition temperature. In the linear regime, the electroclinic effect is characterized by a fast field-independent response time z given by
with y0 being the soft mode viscosity. Electroclinic behavior has been recognized in few SmA* LCPs [77] and response times in the sub-millisecond range have been observed. But, surprisingly, the SmCT,xphase of polyacrylate V-1 I exhibits an electroclinic switching process with a short response time (200 ps) that is changed into a ferroelectric one as the temperature is raised. Moreover, the electroclinic-ferroelectric transition is shifted towards lower temperatures when the voltage is increased (Fig. 20). SmC* polyacrylate XIV, having a chiral center in its spacer (Fig. 21), also shows, at
2.4.2 Electroclinic Switching The electroclinic effect is an induced molecular tilt observed in the chiral orthogonal smectic phases, such as the smectic A* phase, when an electric field is applied along the smectic layers [76]. The induced molecular tilt 8 is a linear function of the applied field E and gives rise to an induced polarization Pi
.
12 t
A
10-
E
Y
0 -
o
"electroclinic"
$
8F -
E
i=
5.0 V / p m 7.5 V / p m 10.0 V / p m
9
a, Q
2.5 V / p r n
6 -
v)
s
225
ferroelectric
4-
-
v)
a,
n 2Ok
152
144 * 156
158
160
162
164
'
Figure 20. Switching behavior of polyacrylate V- I I .
Temperature ["C]
-f ICOO -(CHz),CHCOO I
-fCH2 -CH
t
CH3
~
c
m
~
m
l
o
H
z
l Figure 21. Structure of polyacrylate XIV.
226
IV
Behavior and Properties of Side Group Thermotropic Liquid Crystal Polymers
first, electroclinic switching between 70 and 95 "C, but as the temperature increases above 95 "C, the switching depends on the applied voltage. At low voltage (E=7 V/pm), an electroclinic process takes place, whilst at higher voltages (E=lS V/pm), mixed switching is observed. Finally, at high voltages ( E =25 V/pm), ferroelectric switching results. Up to now, this behavior has not been explained.
2.4.3 Antiferroelectric Switching Antiferroelectricity [49, 77-79] was observed in FLCPs only two years after its dis-
covery in low molecular weight SmC* liquid crystals [SO]. Figure 22 shows the chemical structure of the tristable FLCPs that have been studied up to now. The apparent tilt angle of polymer XVI as a function of the applied voltage is depicted in Fig. 23. The main features of the antiferroelectric switching in FLCPs are: a third state, which shows an apparent tilt angle of zero, a less marked threshold between the three states when compared to the low molecular weight antiferroelectric liquid crystals, a hardly observed hysteresis, and an anomalous behavior of the spontaneous polarization with temperature (Fig. 24), which is not encoun-
Figure 22. Structures of LCPs exhibiting three states switching (XV-XVII).
-
50
a
25
v
-
-;'02 .-
-
5 -
L
c
F
o
g -5 a a -10 - 25 -50
-
2
40
SO
60
io
80
Ferroelectric Liquid Crystal Polymers' Behavior
90
Temperature [ O C I
tered in low molecular weight antiferroelectric liquid crystals. The existence of a threshold for switching between the antiferroelectric and the ferroelectric states could be of interest in display device applications.
2.4.4 Broadband Dielectric Spectroscopy Additional information on ferroelectric and electroclinic switching can be obtained with broadband dielectric spectroscopy. It appears that the molecular dynamics of FLCPs are comparable to those of low molecular weight compounds [67]. However, the experimental observations are made more difficult for FLCPs than for low molecular weight SmC* liquid crystals due to the conductivity contribution which takes place at frequencies below lo4 Hz and to the difficulty to get a macroscopically well-aligned sample. Nevertheless, several authors, in studying SmC* polyacrylates [22] or SmC* polysiloxanes [ 14,41,67],have observed the two expected collective relaxations in ferroelectric liquid crystals, namely the Goldstone mode and the soft mode. These two relaxations occur at frequencies lower than 1O6 Hz.
'0'
227
Figure 24. Spontaneous polarization versus temperature for polymer XVI.
2.5 Potential Applications of FLCPs Chapter V of this Handbook is entirely devoted to the potential applications of the side chain liquid crystal polymers, but it is interesting to mention here the main areas where FLCPs could play a role. As far as we know, the following applications can be considered for SmC* LCPs: nonlinear optics, pyroelectric detectors, and display devices.
2.5.1 Nonlinear Optics (NLO) Several FLCPs have been designed for NLO [16, 23, 431. Very low efficiency has been obtained. This point is discussed in Sec. 4.
2.5.2 Pyroelectric Detectors There is a real need for efficient pyroelectric materials for uncooled infrared detectors. At the moment, low-cost infrared cameras are made with poly(viny1idene fluoride-cotrifluoroethylene) [ P(VF,-TrFE)] as the pyroelectric material. To take the place of P(VF,-TrFE), FLCPs have to exhibit a bet-
228
Behavior and Properties of Side Group Thermotropic Liquid Crystal Polymers
IV
ter figure of merit than P(VF,-TrFE). The relevant figure of merit for this application being P E tg6
~-~
(14)
,
withp the pyroelectric coefficient, E the permittivity, and tg6 the loss angle. Comparison of P(VF,-TrFE) with polymer XI1 [12] (Fig. 25), which is an FLCP having one of the best pyroelectric coefficients, is made in Table 7. It turns out that the figure of merit of polymer XI1 is an order of magnitude lower than that of P(VF2TrFE). Three improvements are necessary to FLCPs to compete with P(VF2-TrFe), namely: -
to have an SmC* phase at room temperature, to have a lower loss angle,
-
and to exhibit a higher pyroelectric coefficient.
At the moment, the two first requirements seem possible to meet, but the last one is still questionable.
2.5.3 Display devices This is certainly the most advanced potential application of FLCPs. Taking advantage of the processability of FLC polymers, Idemitsu Kosan Co. has fabricated a largearea, static-driven display (12 x 100 cm) and a dynamically driven simple matrix display (13 x37 cm) [7, 191. The performance of these two demonstrators is summarized in Table 8. The electrooptic material used was plasticized polyoxyethylene XVIII (Fig. 26).
20
3
2
c
10
2
sn 5
0
ao
75
70
Figure 25. Evolution of the pyroelectric coefficient of polymer XI1 with temperature. a5
90
temperature (OC)
Table 7. Comparison of the pyroelectric coefficient and the figure of merit for copo(VF2-TrF) and polymer XII. Polymer
p (nC/cm2 K)
E
tg 6
PI(& tg6)".s
0.02 0.2
13.36 1.44
~~
Copo(VF,-TrFe) XI1
'
a
5 5
7 60' ~
~~~~~
"At room temperature; 'at T=70"C; 'value deduced from [12].
3
Side Chain Liquid Crystalline Networks and Mechanical Properties
Table 8. Characteristics of Idemitsu’s demonstrators. Driving type Size (cm) Pixels Pulse width (ms) a Applied voltage (V)
Static
Dynamic
12x100
12x37 96x288 3 +30
5
+ 20
Viewing angle (deg.) (LR) (UD) a
I0 60
70 60
Room temperature.
A -
(XVIII)
I
CH3
Figure 26. Structure of polyoxyethylene XVIII used by Idemitsu to make their demonstrators.
The displays were fabricated in three steps: first the FLCP was deposited onto an IT0 coated polymeric substrate, then the counter electrode (another ITO-coated substrate) was put onto the FLCP film, and the formed sandwich was laminated. Finally, the laminated three-layer film was treated by a bending process to obtain the required alignment of the side chains (Fig. 27). These
M FLCP
1
Figure 27. Fabrication of the FLCP display device. (a) An FLCP-coated substrate is laminated with another IT0 oolvmeric substrate. (b) The laminated sandwich is bent on a temperature controlled roll. 1
,
\
I
results are very promising and we can expect in the very near future the advent of large flexible displays with subsecond rate images.
3 Side Chain Liquid Crystalline Networks and Mechanical Properties Crosslinking liquid crystalline polymers yield materials with exceptional properties due to the coupling between elastic properties and mesomorphous behavior. These compounds, especially the mesogenic elastomers, have attracted considerable attention in recent years. Like conventional elastomers, they can sustain very large deformations causing molecular extension and orientation, but they can also exhibit spontaneous distortions, some memory effects, an unusual mechanical response, and coupling between mechanical, optical and electric fields.
3.1 Theoretical Approaches 3.1.1 Landau-de Gennes Description of Nematic Elastomers
(b)
plastic sunstrate
229
De Gennes [ 8 1, 821 first suggested that remarkable effects, such as mechanical critical points, shifts in the phase transition temperature, jumps in the stress-strain relation, and a spontaneous shape change, would take place in nematic elastomers. His proposal was based on symmetry arguments independent of any microscopic consideration. According to this model, the free energy expansion, in the case of uniaxial nematic order and incompressible strain of the sam-
230
IV
Behavior and Properties of Side Group Thermotropic Liquid Crystal Polymers
ple, takes the form [81-831
F ( S , e ) = Fo(S) - U S e + 1/2Ee2 - oe (15) where S is the order parameter, e the deformation, E the elastic modulus, and o the stress (force per unit area). F,(S) represents the nematic energy in the absence of any network effect and - U s e the coupling of strain to nematic order. Without mechanical stress (o=O),spontaneous deformation must occur in the nematic phase because of the linear coupling between S and e. Minimizing F with respect to e gives the equilibrium value of em= U S / E .Inserting this value and the form for F,, the above equation takes the following form
F ( S , em)=
AS2 BS3 CS2 u2_ s2 + -~ + -~ + ... - _ 2 3 4 2E
~~
With an applied stress, an additional deformation arises and the minimum condition yields the strain e,=(o+ U S ) / E . Neglecting terms in 02,the shift in F away from F, becomes
The formalism of the influence of the mechanical field on S is similar to the one found with an electric or a magnetic field [84]. Principal results include a shift in the transition temperature [83] and the occurrence of a mechanical critical point. Experimental verification [85] is found by plotting the order parameter as a function of temperature for different nominal stresses.
3.1.2 Models for Nematic Rubber Elasticity Modifications of the conventional Gaussian elastomer theory were proposed that include short range nematic interactions. These an-
isotropic interactions cause a change in the chain conformation, and hence increase the chain radius of gyration [86-881. The elastic free energy is written as a function of the probability, P ( R ) , of finding two crosslinks separated by a vector R . If the chains are crosslinked in the isotropic state, the probability Po, corresponding to the undistorted R,, is an isotropic Gaussian. In the nematic phase, the system is not able to adopt its natural dimensions and undergoes a bulk deformation A assumed to be affine ( R =AR,). The additional elastic free energy of a strand with length R is given by F,,=-k,T(lnP(R))P,
(18)
If lI1and 1, are the effective step lengths (monomer length) for the random walks parallel and perpendicular to the ordering direction (in the isotropic phase lI1 = I,= lo), respectively, and under the assumption that the deformation occurs without a change of volume, the following relation is obtained for the elastic component of the free energy in the nematic phase
Warner et al. [88, 891 give a full description of the free energy and recover, by minimization, the spontaneous strain and the mechanical critical point. They also show that, if the network is crosslinked in the nematic phase, a memory of.the nematic state is chemically locked. This causes a rise in the nematic-isotropic phase transition temperature compared with the uncrosslinked equivalent. After crosslinking in the isotropic state, the transition temperature (on the contrary) is lowered. The same idea of nematic interactions favoring the distortion of the chains is also the basis of phenomenological descriptions. Jarry and Monnerie [86] and Deloche and
3
Side Chain Liquid Crystalline Networks and Mechanical Properties
Samulski [90, 911 described these interactions in the mean field approximation by an additional intermolecular potential from the classical theory of rubber elasticity. A similar expression is proposed for the elastic free energy.
23 1
Table 9. Phase transition as a function of the crosslinking density xa.
9
3.2 Characterization of Networks 3.2.1 General Features Liquid-crystalline networks can be obtained in several ways, either from an isotropic middle or from materials in which the mesogenic groups were first macroscopically aligned prior to the final crosslinking (see Chap. V of this Volume). The latter networks should keep a complete memory of the original orientation and recover it even after been held well above the clearing temperature for extended periods [92, 931. As long as the crosslinking density is low, the liquid-crystalline phases of the corresponding uncrosslinked polymer are retained for the networks [94-971, with the same structure and without large shifts of the transition temperatures (see Chap. V of this Volume). As the crosslinking density increases, the smectic phases should disappear (Table 9) for the benefit of a less-ordered nematic state [98]; then all the liquidcrystalline phases should be destroyed for the most crosslinked networks, at least for materials crosslinked in an isotropic state [99-1011. Degert et al. [9S] pointed out that this evolution in the mesophase stability also depends on the nature (mesogenic or aliphatic) of the crosslinks.
Phase transition temperature ("C) x
I%]
SmA
0
149
5 10 15
9.5
N
L 0
128 127 116
0
0
Molar %, determined from the conditions of synthesis.
a
3.2.2 Effective Crosslinking Density In order to compare theories with experiments, the number of effective crosslinks should be determined. Actually, this parameter cannot easily be analyzed, because in any one elastomer, the ratio of linkages that do not contribute to an infinite network is difficult to evaluate. In addition, mesogenic groups induce sterical hindrance, which should modify the chemical reactivity of the netpoints as well as the theories referring to ideal rubbers that are used for the determination of the effective crosslinking density.
232
IV
Behavior and Properties of Side Group Thermotropic Liquid Crystal Polymers
Spectroscopic techniques (for example, infrared or solid-state I3C NMR) [102, 1031 allow control, with more or less accuracy, of the chemical crosslinking reaction. Greater efficiency of the crosslinking process for materials crosslinked in the isotropic phase over those crosslinked in a mesophase was reported [ 101, 1031, particularly in the case of a smectic phase. This fact can be related to the lack of spacially available sites for crosslinking in the mesophase. Furthermore, the efficiency increases with the length of crosslinking agent [ 1031, which confirms that spacial considerations are important in the crosslinking process. The effective crosslinking density, i.e., the ratio of linkages taking part in the realization of a network of ‘infinite’ molecular weight, is strongly dependent on the conditions of the crosslinking reaction. When the reaction is performed in bulk (isotropic or anisotropic state), sterical hindrances play an important role. When the crosslinking reaction takes place in solution, Degert et al. showed that the polymer concentration has a major effect [98]: due to the short length of the backbone (degree of polymerization generally around 100 or 200), complete gel formation needs high concentrations of the polymer in the solvent (see Table lo). At low concentrations, intrachain linkages are favored to the detriment of intermolecular reactions, and the proportion of the former compared with the latter can become very large. Under the concentration corresponding to the gel point, the formation of an ‘infinite’ network is not achieved, even if the potential branching points are much more than two per chain. Consequently, in this case, the samples either remain soluble or contain extractibles or exhibit abnormal high degree of swelling [98, 1001. These networks should display nonreversible behavior when they are stretched [122].
Table 10. Characteristics of a crosslinked polymer synthesized at various concentrations in the solvent 1981.
?
I
+Q
? CHs
Polymer concentration in the reaction bath mmoles of units/cm3 of toluene 0.24 0.87 1.2 1.6 1.8
Weight
Consistency of the final product
%
8 23.5 29 35 37.5
Weight swelling ratio in toluenea k0.3
soluble gel gel gel gel
4.4 4.0 3.6 3.6
a Weight ratio of the swollen network to the dry network.
Swelling experiments and stress-strain measurements are widely used to estimate the crosslinking density in liquid-crystalline networks [IOl, 106-1081. For ideal rubbers with tetrafunctional crosslinks, these techniques allow the evaluation of the number average molecular weight between crosslinks M , [104, 1051. From swelling experiments, Flory theory gives, for high degrees of swelling and polymer chains that are long compared to M,,the following relation (neglecting the weight of the linkages)
M , = d V(0.5 - X ) - 1 9 513
(20)
3
233
Side Chain Liquid Crystalline Networks and Mechanical Properties
where d is the density of the network in the unswollen state, V the molar volume of the swelling solvent, and q the ratio of the volumes of the equilibrium swollen network and of the unswollen network. The FloryHuggins X parameter can be obtained from the second virial coefficient (from lightscattering experiments, for example). This theory refers to ideal rubbers. The additional entropic component, introduced by the presence of the mesogenic cores [91], is not taken into account, and so the absolute values of M , are not relevant. From mechanical measurements, classical rubber elasticity theory gives, for low uniaxial deformations and polymer chains that are long compared to M,, the following relation [ 1041
where R is the universal gas constant, o the stress, T the absolute temperature, and A the deformation (d=I/lo, I and I, being the lengths of the sample after and before deformation, respectively). As above, this relation, predicted for perfectly formed networks, is not absolutely convenient for mesomorphous elastomers,
Table 11. Average molecular weight M," of the network subchains in two networks '.
a : x : y : z = 0.96 : 0 : 0.04
b : x : y : z = 0.94:0.02 : 0.04
I/
e
t
Cross - link unit b/(a+b) mol %
b
Network
Mc theoretical value
Mc value from swelling experiment
Mc value from mechanical experiment
a b
18 300 18300
118000 86 000
70 000 48 000
CH2= CH - GO,. CH2. CH2. &C. CH = CHP
Figure 28. Elastic modulus of liquid crystalline networks obtained from the polymerization of monomers a and b as the function of the composition. The solid line corresponds to the value predicted by Eq. (21) [991.
a
In g/mol;
[ 1071.
234
IV
Behavior and Properties of Side Group Thermotropic Liquid Crystal Polymers
Lastly, the question of homogeneity in the liquid-crystalline networks remains open. The distribution of the linkages in these materials, which are obtained in bulk or in concentrated solution, would be worth further investigation, especially when the nature of the crosslinking agent is essentially incompatible with the liquid-crystalline organization. Some neutron investigations of networks swollen with deuterated solvent are in progress on different types of elastomers, differing from one another in the nature of crosslinks and in the conditions of synthesis [ 1091. These studies would help, as was done for another network [ 1lo], to analyze the microstructure of liquid-crystalline networks.
even in the isotropic state. Moreover, networks should not be too highly crosslinked if a connection with a Gaussian description is to be sought [83]: If the subchains between links are too short, then they are not flexible enough to be treated by a classical theory. Actually, significant deviations from linearity for aversus the crosslink density (estimated from the conditions of synthesis) (Fig. 28) were found [99]. Due to the imperfections of the theoretical approaches, the value of M , obtained from swelling and mechanical experiments widely differ from one another (Tables 11 and 12) and from the values determined either by the conditions of synthesis or by spectroscopy [ 106-1081. On the other hand, the proportion of permanent entanglements cannot be estimated. The degree of polymerization above which entanglements should appear is unknown for side chain liquid crystal polymers and, because of the coupling of mesomorphous order, the entanglement slippage has to be regarded as dependent on the direction 1951.
3.3 Mechanical Field Effects on Liquid-Crystalline Networks In the isotropic state, as in classical networks, the elastic behavior is governed by
Table 12. Crosslinking density (cd) of various elastomers (in % repeat units) a , b CHZ= C*COz-(CH2)2
-0-@OZ+CN
CHz=C*COz-(CH2)3
-0 G
C
CHz = Clt-COz -(CHz)z
-OH
111
cd theoretical value I11 (%)
I
O
z
G
C
N
11
cd from IR analysis
cd from modulus [Eq. (2111
cd from swelling [Eq. (2011
Tg ("C)"
4.1 3.9 3.8 3.9 4.2 5 .O
1.27 k0.13 1.40k0.13 1.47k0.30 1.46k0.13 1.30k0.13 1.40k0.13
3.4k0.8 3.8k1.0 3.3k0.5 4.8k1.2 3.3 k0.8 3.9k0.8
67 67 71 72 77 82
TNI("C)
100 99 100 99 102 102
bnetworks synthesized with variable proportions of I and 11; 111= 6%; Tg:glass transition temperature; T ~nematic-isotropic ~ : transition.
a [ 1081;
3 Side Chain Liquid Crystalline Networks and Mechanical Properties
the conformational properties of the network chains. The presence of the mesogenic groups should only modify the entropy of the system, and hence the extensibility of the polymer subchains. In liquid-crystalline phases, the introduction of liquid-crystalline order is the decisive factor in the elastic response. The coupling between mechanical properties and mesomorphous order is the main theme of this chapter (for more details, see Chap. V of this Volume).
235
A
CH3
I
4--
I I 0
3.3.1 Response of the Network as a Function of the Stress
c.0
I
(7WZ
The potential of the mesogenic units for alignment has a marked effect on the stressstrain behavior [ 8 5 , 1 1 1, 1 121. Consider a uniaxial stress applied to a liquid-crystalline network synthesized in an isotropic state; this means without any macroscopic orientation. At low strains, the induced strain increases linearly with the stress (Fig. 29). The polydomain structure remains stable and the X-ray pattern resembles that for powder samples. Above a strain threshold, macroscopic orientation of the mesogens parallel or perpendicular to the direction of the stress results, as evidenced by X-ray analysis or IR dichroism measurements (see Sec. 3.3.3). Depending on the topology of the network and on the mesogenic units, the polydomain-monodomain transition is more or less sharp, and therefore the threshold deformation is more or less observable [108]. During this orientation, a remarkable change in the dimensions of the sample spontaneously occurs, which depends on the orientation of the mesogens with respect to the stress axis (for the example shown in Fig. 29, the mesogens are in line with the elongation stress axis).
0
I R5
I I (W 0
I I c.0 0
I
--c-
I w
I
-
WI
As--
OC-NH
7.
G..
-W
-
Figure 29. Nominal stress 0,versus strain ( L = / / l o ) for the nematic elastomer shown in [ 8 5 ] .
At higher stresses, a linear relationship between the stress and the strain is regained. The sample is uniformly aligned and the orientation process no longer takes place.
236
IV
Behavior and Properties of Side Group Thermotropic Liquid Crystal Polymers
3.3.2 Thermal Evolution of the Elastic Modulus The rheological properties of liquid-crystalline elastomers strongly depend on the mesomorphous order [95, 98, 102, 113-1181, even for strains that are too small to induce macroscopic orientation of the mesogens (strain amplitude symmetry was induced along the poling direction, which has been verified in many subsequent studies, and that homogeneous transverse alignment of the nematic host could be achieved, free from domain-induced light scattering ((P2)=0.3). SHG was used to measure the x(*) tensor elements. A comparison with 7 doped at 2% in poly(methylmethacry1ate) showed a large enhancement (- 100-fold) in the liquid-crystalline system. This result was interpreted as being due to the correlation between the dopant molecules and the principal axis of the nematic LC. Excellent agreement was obtained between the experimentally measured value of ( 3 ~ 1 0 esu') -~ for E,=1,3 V/pm, t,=6 h, and T p = T g - 15 "C, and that predicted by a simple model based on statistical dipolar alignment of the guest in the oriented host. The polar alignment that was induced in this system was shown to be stable for long periods of time. However, several complexities in the poling process were uncovered in this study. It was found that the induced SHG dropped precipitously as Tg was approached from below. This was speculated as being due to the establishment of a mono-
xyi3
'Note that lo-.' esu=4xl0-" m/V.
mer-dimer or higher order aggregate equilibrium. In addition to this SCLCP host, MCLCPs have also been used. Li et al. [ 1681dissolved molecules of an oligomeric fragment of poly(pphenylenebenzobisthiazo1e) 10, which was terminated by opposing nitro and dimethylamino groups, in the polymer itself. The resultant material has a f 2 ) value of 6x1OP9 esu. Stupp et al. [169, 1701 carried out a careful analysis of the system of Disperse Red 1 (11) dissolved in a chemically aperiodic nematic polymer containing the following three structural units (12).
-
Electrical poling ( E P 67 kV/cm) of the dye-polymer alloys below the melting point of the nematic host resulted in strong SH intensities from an infrared laser pulse passing through the medium. Interestingly, a significant increase in the SHG signal was observed in alloys that had been aligned by an external magnetic field prior to poling by the electric field.xy?, was tripled in 5 % dye samples when the high molar mass solvent was macroscopically ordered ((P2) 0.6-0.7) in the magnetic field 4 . 5 3 ~ 1 0 esu -~ instead of 1.5 1 x lop9esu). This susceptibility enhancement points to coupling between polar ordering of dye molecules in the electric field and the uniaxial organization of the nematic solvent in the magnetic field. In the extreme of a macroscopic order pa-
(xyi3=
-
4
rameter approaching 1.O, the magnetic field would produce an ‘Ising’ solution of nematic polymer and dye in which external poling would simply bias each chromophore’s dipole moment parallel rather than antiparallel to the electric field (Fig. 34). Additional evidence of this coupling is revealed in magnetically aligned samples by a more rapid rise of the SH intensity with increasing temperature under the electric field. Indeed, this difference supports the premise that dye molecules are themselves preordered in the uniaxial environment of the solvent, thus facilitating the removal of an inversion center by the electric field. It is of interest to note that the thermal stability of the polar dye network was also greater i n the aligned environment. The trade-off between concentration and p limits the applicability of guest-host materials for useful devices. The higher p molecules tend to be larger and less soluble in the hosts. Exceptionally high concentrations of up to 20 mol% of 4-hydroxy-4’-nitrostilbene (13) have been achieved in the polymer 14 without destroying the LC for-
14
DYE
-+ Figure 34. Schematic representation of the Ising solution of nematic polymer and dye formed in a magnetic field [170].
Nonlinear Optical Properties
249
mation [ 17 11. However, limiting values of the order of 5-10 wt% are more common. In addition, in guest-host materials a plasticizing effect was often observed, which lowered the glass transition temperature. More recently, the NLO molecules have been incorporated as part of the polymer structure. Attaching the NLO groups to the polymer reduces their mobility and tendency to crystallize or phase-separate at high concentrations, and leaves the glass transition temperature relatively unaffected.
4.3.2.2 LC Copolymers Containing Both Nematogenic (or Smectogenic) and Active Side Groups The design requirements for synthesizing copolymers containing both nematogenic (or smectogenic) and active side groups are a stable nematic or smectic phase for alignment and poling, the ability to quench to a glass, and long term stability of the glass. Suitable structures may be obtained by appropriately adjusting the flexibility of the polymer backbone and the length of the flexible spacer. If the mesogenic groups are attached directly and rather rigidly to the backbone, the dynamics of the backbone dominate and no LC phases are formed. If the mesogenic groups are appended to the backbone by flexible spacers such as -(CH&, the mesogenic groups may now adopt the anisotropic ordering of an LC phase. However, long spacers increase the degree of anisotropic LC ordering from nematic to smectic, favor the tendency of the polymer to crystallize, and lower the T g . This does not mean that the properties of the polymer are uninfluenced by the backbone. The broadest range of mesophase thermal stability is exhibited by polymers containing the most flexible backbone, namely, polysiloxanes. However, compared to more rigid polymers, polymers containing ex-
250
IV
Behavior and Properties of Side Group Thermotropic Liquid Crystal Polymers
mixture. With bulky SCs, this may not be so, and side chain ratios can be affected by reaction conditions.
tremely flexible backbones usually exhibit additional mesophases or even side chain crystalline phases and have lower Tgvalues. The review in [ 1721by Percec and Pugh may be of value to readers who are interested in this subject. Other difficulties which can be overcome, but may have been less than fully appreciated in the past are:
Factors (l), (2), and (3) clearly affect product reproducibility, which is vital if the properties of these polymers are to be correlated with structure and if they are to play a role in applications. A number of copolymers having nematogenic (or smectogenic) and NLO active groups have been synthesized [155, 1731811. Table 15 lists a number of chemical groups that have been used for their mesogenic properties. As can be seen from their formulas, highly polar terminal substituents
Retention of SC precursor in the polymer. The danger of cross-linking involving SC substituents such as CN. The need to demonstrate that the ratio of the mesogenic and active side groups is in fact the same as that in the reaction
Table 15. Typical mesogenic groups. Molecule no.
Reference
Mesogenic group -
[I741
16
[155, 173, 177, 178, 180, 1811
17
[lSS, 1781 ~751 X = CF3 , CN , OCH3
19
0
0 Y=
-Lo-,
-0-c-
I1
[I751
[173, 1771
21
-
n
24
4 Nonlinear Optical Properties
have been incorporated to allow the pendant groups to be aligned by applying an electric field. The most common systems so far considered are polyacrylates and polymethacrylates with medium-length spacers of two to six methylenic units. The ability to gain molecular control of the LC character through synthesis is clearly outlined in Table 16 for copolyacrylates having the structure 25 [ 1731.
These copolymers were selected according to the above outlined principles. Phenyl ring-CN terminal substituents are known to enhance nematic thermal stability, while terminal CF, groups favor smectic mesophase formation. The R2 side groups possess the basic structure associated with large hyperpolarizability (/?). They contain a strong electron accepting group (-NO,) and an electron-donating group (X=-0-,>N-) linked in a trans-stilbene (Y=-CH=CH-) or an azobenzene (Y=-N=N-) fashion. As shown in Table 16, up to 50 mol% of active R, groups can be incorporated in these copolymers without destroying LC formation. It should be noted, however, that X =-N(CH,)- lowers the LC stability. This
4.",? I"
,?
25
Table 16. Molecular structural control of the LC character" Polymer
X
Y
Z
x (mol%)
Transition temperatures ("C)
0 7.1 21.3 50.3
N 1061 N 119.8 I N 111.41 SmAd 90 N 107.5 I
Y
-
CN CN CN CN
6 6 6
-0-
CN CN
6 6
-0-
-CH
= CH-
-0-
-CH
= CH-
CN
5
-CH
=
CH-
5.5
N 113 I
CN
5
-CH
= CH-
19.7
N 103.5 I
CN
5
-CH
= CH-
32.1
N 109.5 I
CN
6
-CH
=
CH-
2.6
I
0 3.8 8.1
SmA, 99 1 SmA, 87 1 SmA, 110 I
CFS CF, CF, a
[173].
-0-
-0-
-
-COO5 years with only a 20% decay in the electrooptic coefficient if used at temperatures 90 "C below Tg. In a subsequent study, Koide et al. [176] reported on experiments on copolymers containing nonmesogenic azo side groups having Disperse Red 1 as the NLO component and the mesogenic side groups 22. In this example, the films were poled above the clearing temperature with fields of up to 0.125 MV/cm. The ratiox(f:/x?/ was found to be significantly greater than the theoretical value of three predicted for isotropic systems, thus confirming that it is possible to gain in the net polar ordering by starting from a liquid-crystalline system. Using azo aromatic groups to provide the long conjugation between the donor and acceptor groups, other side chain LC copolymers have been synthesized and their NLO properties have been tested. Again it has been shown that liquid crystallinity enhanced the NLO properties upon poling under similar conditions. The reader is referred to a recent review by Xie et al. [184]. Smith and Coles [177] looked at a number of liquid-crystalline polymers. The contact electric field poling method was used to achieve noncentrosymmetric ordering of the NLO side groups. The materials were orientationally aligned in an alternating field at a temperature just below the transition to isotropic liquid. This alignment was stored by cooling into the glass phase before applying the poling field of 10 kV/cm, at a temperature a few degrees above Tgfor a period of 24 h. The second order susceptibility was typically improved by a factor of three by the orientational alignment process when compared with the application of the poling field to a material with no alignment
-
history. Again the results indicated that the liquid crystallinity has some effect on the polar ordering, since xy,)< xF;l3 for many of the liquid-crystalline materials. Recently, workers at the Weizmann Institute of Science reported photochromic side chain liquid crystal copolymers containing spiropyran units 27 [178, 180, 185, 1861. -(
ence of the field with simultaneous UV irradiation [ 1801. The investigators demonstrated that this procedure provides a means of inducing SHG by photochromic spiropyran-merocyanine conversion, and that this effect can be enhanced by blending the copolymer with thermochromic quasi-liquid crystals (QLCs) containing the merocyanine chromophore as the high p moiety (28) [ 1781 or by using the copolymer as host for DANS 7 and another related substituted stilbene guest molecule (29) [178]. They also showed that in these samples component the SHG via the dominant parallel to the direction of electric field alignment decreases over a time scale of several weeks which can be attributed to both orientational relaxation and the spontaneous merocyanine fading to the spiropyran form, which has a much smallerp. Interestingly, it was found that the polar ordering could be enhanced by reapplying a DC electric field to the glassy films parallel to the aligning direction at room temperature. However, these special poled films exhibited asymmetry not only in the poling direction, but also perpendicularly to it. This resulted in more nonLero components of the second-order susceptibility tensor than are obtained through the usual poling techniques. The x ' ~component ) perpendicular
- CH)= -(CH, - C H ) y
CH,
I
I c=o
C=0
I
d
NH
I
I
(CH2)5
(CH2)6
x'*'
I
I
II
c=o II
27
I R, = H, NO2
R, = CN, OCH,
In this study, after alignment in an electric field of ~ 1 kVIcm 0 at a temperature above Tg but below the clearing point, thin films were cooled to room temperature in the pres-
C
/ \
H
cn,
q% CH,
-
spiropyrm-form
11 ,
~
253
Nonlinear Optical Properties
4
C
CH,
NO, Q1.C
28
R=CN,0CH3,0C6H13
A
O
O
~
C
H
=29
N
~
~
~
~
254
IV
Behavior and Properties of Side Group Thermotropic Liquid Crystal Polymers
to the poling direction was the strongest. This nonlinearity was attributed to merocyanine aggregates stacked normal to the film substrate. The effect could be large, for example, a x(*) of 6 x lo-’ esu/cm3 (25 pm/Vj was reported for a QLC/LCP blend containing 2% by weight of DANS, poled at 5x106 V/m, but it was short-lived with decay over about one hour. Longer term stability was, however, reported in the case of the larger substituted stilbene when used in a larger concentration of 10% by weight, but here the coefficient was 1.1 pm/V for a poling field of lo6 V/m [178]. Ou et al. [179] reported on the successful fabrication of X-deposited multilayer films using the Langmuir-Schaefer deposition technique [ 1871. They used a polysiloxane copolymer containing both a mesogenic and an NLO side chain (30).
sentially those factors that encourage liquid crystallinity in low molar mass compounds. Therefore considerable effort has been devoted to the synthesis of NLO polymers in which the NLO moieties themselves possess mesogenic properties. Examples of NLO active mesogenic groups are given in Table 17. They are depicted below in a schematic form along with examples of chemical structural subunits where X and W -x*Y-z-w
represent a range of electron-donating and electron-withdrawing substituents and Y-Z is a conjugated n: electron system with Y=-CH=CH-, -N=N-, -CH=N-, and Z a substituent selected from
30 0
Both X-ray and NLO studies indicated a highly polar noncentrosymmetric structure with nearly perpendicular orientation of the chromophore with respect to the film plane. The investigators found a d,, coefficient of 5 . 3 ~ 1 0 esu. - ~ All the measurements were found to be reproducible over a span of several weeks, indicating a stable structure.
4.3.2.3 SCLCPs Wherein the NLO Active Moieties Possess Mesogenic Properties Themselves The factors that enhance the second order NLO activity of the molecules (high axial ratio, planarity, charged end groups) are es-
where a has small integral value. Examples of electron-donating substituents are divalent radicals such as -0-, - S - , and -NR’-, where R’ is hydrogen or methyl. Typical acceptor substituents are -NO,, -CN, and -CF,. The majority of SC polymers with NLO mesogenic groups have the donor substituent in the spacer chain and the
4
Nonlinear Optical Properties
255
Table 17. Examples of NLO active mesogenic groups. Structure no.
NLO active group
References
31
[171, 177, 182, 189, 191-193, 19.5, 196, 198-200,2021
32
[189, 191-193, 195, 1991
33 34
[177, 189, 1951
35 36
[191, 1921
37 38
[193, 196, 1991
[1.52, 200, 201,2031 ~2041
acceptor at the end of the side chain, as in compounds 31-37 (Table 17). However, Zhao et al. [194] and Bautista et al. [I971 have prepared a number of SC polyacrylates and polysiloxanes with 4-mdimethy lamino)4'-stilbene carboxylic ester mesogens (39 and 40 in Table 17), which have the donor substituent at the end of the side chain and the acceptor in the spacer chain. There have been two main approaches toward the synthesis of SCLCPs containing NLO mesogenic groups:
1) Side groups with suitable polymerizable functional ends can be synthesized first, with subsequent polymerization. Free radical polymerization of a vinyl monomer, e.g., acrylate, methacrylate, or styrene derivative, has been widely used to produce NLO single-component SCLCPs. This type of monomer is suitable as it is relatively easy to prepare and the conditions for free radical polymerization of acrylates, methacrylates, and styrene derivatives have been well documented. There are, however,
256
IV
Behavior and Properties of Side Group Thermotropic Liquid Crystal Polymers
special difficulties associated with the free radical polymerization of an interesting class of NLO materials, namely, nitroaromatics: the nitroaromatic group is a retarder of free radical polymerization, intercepting the propagating radicals by reaction on the aromatic ring or on the nitro group itself. Nitrostilbenes are desirable for NLO, but the stilbene unit itself is potentially reactive with a free radical center. The presence of the Ar-CH=CH-Ar double bond may lead to an uncontrolled crosslinking reaction leading to a gel [193, 1941. This could prevent a good polarization of the material. As reported by Griffin and Bhatti [196], polymerization of nitroaromatic stilbene methacrylates in solution under nitrogen using a single charge of initiator (AIBN) can provide an acceptable route to NLO SCLCPs. The analogous polymerization of acrylates produces lower yields with crosslinking. This difference in the extent of crosslinking is attributed to the greater stability of the propagating radical from methacrylate monomer (tertiary) and its concomitant greater selectivity. Alternatively, polycondensation appears to be a facile synthetic entry into NLO polymers having nitroaromatic-containing SCLC groups. Griffin et al. [191, 1921 were successful in preparing three series of malonate SCLCPs by using a polycondensation route first described by Reck and Ringsdorf [206]. This synthetic approach obviates the difficulties described above for free radical polymerization, as the nitro group is not an inhibitor of polycondensations.
~ O - C H - C H ~
I
CH, CI
DMF.TBAH
*~O-FH-CH
cn, CI
2) The preformed polymer can be reacted with the side groups. This has the advantage of starting with well-defined polymer parameters of the molecular weight and distribution. However, incomplete reaction of all possible sites on the polymer can lead to unwanted crosslinking at a later stage of processing, unless they are mopped up by a more reactive component after the NLO group has been attached. Bautista et al. [ 1971 used the hydrosilylation reaction to synthesize NLO SCLC polysiloxanes. They prepared a homopolymer from poly(hydrogen methylsiloxane) and a copolymer from poly(hydrogen-methyldimethylsiloxane). However, the higher Tg values that are sought for use in NLO need the use of a backbone that is inherently less flexible than the Si-0 linkage. Saturated carbon-carbon backbones fulfill this requirement, and McCulloch and Bailey [ 189, 1951 grafted alkylbromochromophore units onto a preformed poly-4-hydroxystyrene polymer. It should be noted that SC copolymers based on nonmesogenic SCs and NLO mesogenic SCs offer the opportunity to finetune the polymer properties by varying the ratio of the NLO mesogenic SCs, although care must be taken to characterize the product, since the final ratios of the SCs may not correspond to the feed concentrations. An interesting example is the series of copolyethers prepared by chemical modification of atactic polyepichlorohydrin with the sodium salt of 4-cyano-4’ hydroxybiphenyl [ 188, 2071 (45). The data in Table 18 show that the thermal behavior of these copolyethers strong-
,+o--cH
I
-cn2k
45
4
Nonlinear Optical Properties
257
Table 18. Thermal properties of copolyethers derived from polyepichlorohydrin [207]. X%
0
Ts ("C) -26.5 ("C)
TNI
25.5
40
65
74
83
91
10
26
66.6 105
79.5 133.5
83 143
86 153.5
ly depends on the degree of substitution. There is a minimum substitution limit of approximately 65% necessary for LC formation. Below this limit, the copolymers all resemble typical amorphous copolymers. There is a sharp linear increase in the glass transition temperature with increasing amount of substitution, indicating that the bulky side groups cause severe hindrance to main chain motions. Once 65% substitution is reached, the increase in Tgdrastically decelerates and the copolymers exhibit a nematic phase. The copolyethers with a degree of substitution of about 50-60% display no liquid crystallinity at rest. However, miscibility studies demonstrated that they exhibit a virtual nematic phase, the virtual nematic-isotropic liquid transition point being a few degrees below Tg [208]. The backbones that have been most commonly employed are those of the acrylate [177, 190, 194, 1961, methacrylate [152, 171,196,198-200],andsiloxane[152,177, 1971 types. Polyethers [207-2091. polyesters [182, 191, 1921, and polystyrenes [177, 189, 1951 have also been reported. Typical spacer groups consist of between 3 and 12 methylene units. The phase transitions of a number of SCLCPs containing NLO mesogenic groups are collected in Table 19. Unfortunately, the molecular masses of many of these polymers have not been determined, and the influence of the polymer structure on the phase transitions can not therefore be quantitatively discussed. However, the general points to emerge from these data are as follows:
As the spacer length increases so Tgtends down wards. Any N phases normally arise at shorter spacer lengths; longer spacers tend to encourage smectic phases (see 41 PA 2-6; 42 PMA 3- 12). Tg decreases as the backbone flexibility increases (see 41 PA 2-41 Psi 3,42 PMA 5 -42 Psi 5), while AT = T, - Tg increases with increasing flexibility. The highest clearing temperatures are obtained with the most flexible backbones (42 Psi 5 > 42 PMA 5 , 4 1 Psi 3 > 4 1 PA 2, 41 PA6>41 PMA6, 39 PSi11>39 PA 10). Similar trends have been noted in many other SCLCP series involving different polymer components [ 1721. The first question to be decided is whether it is possible to polymerize a molecular sequence, effectively adding the NLO mesogenic groups onto a backbone, without degrading the optical and NLO properties of the active moiety. The UV/visible properties of SCLCPs containing NLO mesogenic groups are similar to those of their low molar mass analogs. In all cases, the chromophore appears unchanged by the polymerization. The absorption maximum wavelength (A,,,,,) of these polymers lies in the range 285420 nm. Polystyrenes [ 189, 1951, polyacrylates [193], and polyesters [191, 1931 based on the NLO mesogenic groups 31,32, and 34 (see Table 17) have similar absorption maxima (375-381 nm in THF or CHCI,).
25 8
IV
Behavior and Properties of Side Group Thermotropic Liquid Crystal Polymers
Table 19. Thermal behavior of SCLCPs containing NLO mesogenic groups. Chromo- Back- +CH& DP M , (XIO') Mw (x 10') Tg ("C) Mesophase phore bone a SmA SmA SmA
31 31 31
PS PS PS
6 8 10
85 65 50
31 31 31 31
PMA PMA PMA PMA
6 6 6 6
61 52 71 unspecified
31 31 31 31
PEsa PEsb PEsc PEsd
1 6 6 6
60
32 32 32
PS PS PS
6 8 10
32
PMA
32 32 32
250 250 250
T, ("C)
Ref.
165 159 155
[189, 1951 [189, 1951 [188, 1951
Sm? 156 SmA 106 unspecified 153 N unspecified
30 -
N N SmA Ch
85-90 106.5 88.5 71.5
75 65 40
SmA SmA SmA
118 119 101
6
55
SmA
116
PEsb PEsc PEsd
6 6 6
-
5 10
N Ch Ch
62 52.9 33.5
34 34
PS PS
6 8
85 70
SmA SmA
151 151
36 36 36
PEsb PEsc PEsd
6 6 6
SmA SmA SmA
131.2 71.4 68.0
38
PMA
6
39 39 39 39 39 39
PA PA PA PA PA PA
10 2 4 6 8 10
39 40
Psi Psi
11 11
41
PMA
41 41 41 41
-
250 250 250
250 250
80
?
79 129 119 109 91 81
unspecified unspecified unspecified unspecified unspecified unspecified
128 129 130 125 121 1 I7
T,= 167 T,= I10
N SmB 192N
215 25 8
6
64.4
N
115
PA PA PA PA
2 4 5 6
84 42 35 32
41
Psi
3
30
SmA
145
42 42
PMA PMA
3 5
85 45
N N
I00 72
13.3 5.7 2.8 3.9 3.4 5.2 60 60
186 14.2 I 9.7 9.7 10.11
N 113.5 N 110 SmA,j 120 NI 124.4 N,,80 S&, 124.5N 132
[2001 [i991 [i981
[171]
4
Nonlinear Optical Properties
259
Table 19. (continued). Chromo- Backphore bone
+CH&
42 42 42 42
PMA PMA PMA PMA
6 8 I1 12
42 42 42 42
PMA PMA PMA PMA
6 6 6 6
42
PSI
5
DP
Mn(x10') M,
Mesophase
T, ("C)
Ref.
10
N N SmA SmA
64 75 9.5 80
[152,201] [152,201] [152,201] [152,201]
51 47 47 54
Sm Sm Sm Sm
18
SmA
165
[152,201]
( X I @ ) Tg ("C)
40
35 20
147 771.5 849.8 1241 55 CH?
PS: $-CH-CH2
I
j,
l
PA:
-(-CH-CH?+.
I
PMA:
j
Replacement of shifts A,,,
I
-'"'0
O~c'O-
Psi:
(-C-CH&
+-
by - 4 3 ( 3 4 )
to 393-394 nm. Similarly, chang-
ing X =-0- by X =-COO
CN(38) re-
sults in a shift of Amax to longer wavelengths ( ~ 4 1 nm) 2 [193]. The effect of increasing the length of the conjugated linkage between the donor and the acceptor is clearly seen from compounds
42, 31, 32, and 34. A shift of ca. 40 nm towards longer wavelengths can be realized on going from nitrobiphenyl (A,,, z 338 nm) [201] to nitrostilbene or nitroazobenzene [189, 1911. One particular comparison of the monomer and polymer properties has been done by De Martino et al. [201], who measuredp solvatochromically for 4-( 12-dodecy1oxy)4'-nitrobiphenyl, its methacrylate, and the
260
IV
Behavior and Properties of Side Group Thermotropic Liquid Crystal Polymers
polymer. Their results are shown in Table 20. All the compounds showed an absorption band at about 338-340 nm. However, a noticeable decrease of /3 was observed after polymerization. McCulloch and Bailey [ 189, 1951 reported /3 values for a series of SCLC polymers based on apolystyrene backbone (Table 21). They showed that both the azo and stilbene link units exhibit a greater solvatochromic shift than their imine counterpart, which manifests itself as a larger hyperpolarizabilTable 20. Comparison of /l values measured for a monomer and its polymer ’. CH3 -&l-CHZ
~~
Compound
La,(nm)
p ( 1 o - ~ O esu)
Monomer Polymer
340 338
11 8
a
12011.
Table 21. Values of following structure ’.
fy-cn,
0 A N CH N N CH N N CH N N
am,,
and p for polymers of the
+
ity. This difference is mainly attributable to the two aromatic rings being twisted out of plane, thus reducing the 7c-orbital overlap. Also, an increase in the length of the flexible spacer is accompanied by a slight decrease of /3. In a subsequent study, Wijekoon et al. [ 1991 carried out a careful analysis of a series of LC polymethacrylates and corresponding monomers based on the NLO mesogenic groups 31,32, and 38. A noticeable trend in their results is that, even though the hyperpolarizabilities of the monomers and the corresponding polymers have similar magnitudes, the /3 values obtained for the monomers are somewhat higher than those for the polymers. This effect may be attributed partially to the different local molecular environment resulting from the solvent-solute interaction. Several additional interesting effects were noted in this study. First, it was observed that for the same donor-acceptor pair, replacement of an ethylene linkage -CH=CH- between two aromatic rings by an imine -CH=N- linkage is damaging, so reducing the /3 value. This is in agreement with the data reported by McCulloch and Bailey [189, 1951 and confirms that the extended and planar structure of the stilbene moiety is very effective for optical nonlinearity. Second, it was found that replacement of an ether linkage -0- by an amino linkage -N
B
rn
Lax (nm)
/3(10-30esu)b
N CH CH N CH CH N CH CH N
3 6 6 6 8 8 8 10 10 10
378 378 377 378 378 377 378 378 377 378
31.2 29.1 21.6 28.9 29.3 16.0 28.5 24.2 17.2 23.5
a [189, 1951. Estimated by a solvatochromic technique using the solvent shift between THF and hexane.
>J--
results
in an increase i n p , the donor strength of the former being inferior to that of the latter. It is difficult to compare the second order nonlinear response of different materials because of its dependence on processing and the electrical poling process. Nevertheless, we have collected the SHG values of d,, (=xy:/2) tensor components and electrooptic r33 coefficients for a number of SCPs containing NLO mesogenic groups in Table 22.
a d33 -xYi12, r33= 2 xYj/n4, where n is the refractive index.
35 Psi 6
383
18 (at 1.910 ym)
33 PS 11
100
100
100 I00
100
292 292 292
17 17 17
Corona poling
41 PA2 41 PA6 41 Psi 3
383 (377 in THF)
100
800
18
32 PS6
370
Corona poling
39 PA 10
14
31 PEsa
370 (378 in THF)
150
-
EP (kV/cm)
Corona poling
74
31 PS6
370
370
Characteristics I+,,,, (nm)
38 PMA 6
74
31 PMA6
Chromo hore B(10-3L at 1.06 ym)
74
Ref.
31 PMA6
Polymer TP ("C)
24 h
24 h
24 h 24 h 24 h
24 h
20 min
a few min
2 min
tP
Poling conditions
Table 22. Linear, poling, and nonlinear optical characteristics of some SCLCPs listed in Table 19.
0.26
0.6 4 3 , or
~0
No difference bemeasured tween after poling in the isotropic or the LC state
Decrease of the electrooptic effect on cooling through the isotropidnematic transition
xyi Enhancement of by a factor of 1.5
Evaluation of NLO properties enhancement by LC order
NLO measurements
262
IV
Behavior and Properties of Side Group Thermotropic Liquid Crystal Polymers
Griffin and co-workers developed SCLCPs [ 1961 that demonstrate some interesting SHG phenomena [211]. They are easily poled at room temperature, having a signifin the mesophase. With the poling icant x(2) field continuously applied, the x(2) diminishes as the temperature increases, going to zero at the mesophase-isotropic transition (125-145 "C). Upon cooling back into the mesophase, the x(2)reappears and grows larger as the temperature decreases. That is, the x(2)tracks with the orientational order parameter of the liquid-crystalline phase. A maximum value of d33 of 35 x esu (G 14.7 pm/V) is reached at room temperature in polymer 38 PA 5 , which also shows excellent temporal stability of SHG (Fig. 35). The preliminary interpretation of this observation is that dipoles in the nitroaromatic pendant groups that are aligned by the electric field can relax back easily in the isotropic phase where the pendant rods are 'loosely' organized. However, they cannot relax back in the mesophase with its 'closely packed' parallel orientation of neighbor-
ing rods. That is, the motions of the pendant is enNLO dipoles are retarded and hanced significantly. Conformation of the better temporal stability of SCLCPs compared to that of isotropic SCPs that have similar chromophores has been made by Carr et al. [198]. These workers found that the polarization of the SCLCP 31 PMA 6 is much more stable than that of the isotropic copolymer (31 PMA 3)0,9-(31PMA 6)0,1in spite of the fact that the Tg of the latter polymer is about 117 "C compared with 7 1 "C for the LC homopolymer. However, in this work, the better temporal stability of the LC polymer was related to the absence of p relaxation, as shown by dielectric measurements. Quite recently, Gonin et al. [188, 208, 2121 and Guichard [209] conducted a systematic study of factors influencing SH efficiency. In this work, SHG was measured for five polymers containing the NLO mesogenic groups 41: a polyacrylate 41 PA 3, a polymethacrylate 41 PMA 3, two copolyethers derived from polyepichlorohydrin 45 ( x = 0 . 8 ) and 45 (x=0.55)and a copolyether 46 derived from poly(3,3-bis(chloromethy1)oxetane). Optical texture observations and X-ray investigations showed that the polyacrylate 41 PA 3 and the copolymer 45 (x=0.8) form nematic phases. The polymethacrylate 41 PMA 3 and the copolyether 45 (x=0.55)display no liquid crystallinity at rest, but exhibit a virtual.isotropic liquid-nematic transition a few degrees below T g ,as evidenced by miscibility studies. Copolyether 46 does not show any threaded or schlieren texture, which would be characteristic of a nematic.
1
0.0 100
200
300
TIME (hours)
Figure 35. Temporal stability of polymer 38 PA5
x(2)
4
263
Nonlinear Optical Properties
CHzCI
4
CH,
-7-I
CH,O
46
CH,CI
(mole ratio x : y : z 0.14 : 0.43 : 0.43)
The glass transition is associated with weak birefringence only. A number of important findings were reported in this study. First, it was shown that homeotropic alignment of the NLO mesogenic groups, free from domain-induced light scattering, could be achieved by using the corona poling technique. As shown in Fig. 36, the order parameter (P2)increases with poling time before leveling off. A saturation limit for (P2) of about 0.65 was observed for the nematic polymers 41 PA3 and 45 ( x = 0 . 8 ) . It is of interest to note that the response times of the polymers increased as the flexibility of the polymer backbone was increased. Clearly, the openness of the polymer structure is an important factor governing the axial order. The microscopic order parameters (P2) derived for 4 1 PMA3 and 45 (x=0.55) were smaller (0.4-OS), but larger than those usually found for covalently functionalized isotropic polymers (0.1-0.3) [213], indicating that the poling field induced an isotropic/ nematic transformation in these polymers, which are isotropic at rest but exhibit a vir-
e O0 43 I
/
O01 I/
0 0
rb
40
30
40
50
60
l (,I”)
Figure36. Microscopic order parameter (P2) as a function of poling time for 41 PA 3.
~~
tual isotropic liquidhematic transition. This was predicted by the MSVP model [1571591. SHG was used to monitor the polar alignment that was induced in these systems. Values of the second-order NLO coefficients d,, and d,, are given in Table 23. The salient conclusions of the results are as follows: The experimental values of d,, are much larger than those reported for other materials suitable for blue conversion applications, like potassium titanyl phosphate, and are approximately equal to those determined for lithium niobate [153]. The thermodynamic model predicts that d,, =d,,/3 for a poled isotropic material and d,, =0, if the liquid crystallinity of the material is accounted for using the Ising model (Table 13). In agreement with these theoretical predictions, the results listed in Table 23 show that liquid crystallinity results in higher anisotropy for the x(2)tensor, since d,, is much lower than d3,/3 for all the polymers investigated. The observed difference of behavior between the polyacrylate 4 1 PA3 and the polymethacrylate 4 1 PMA 3 and the two copolyethers 45 ( x = O . 8 ) and 45 (x=0.55) is probably due to the smaller axial ordering induced by the poling field in the films prepared from the isotropic polymethacrylate 4 1 PMA 3 and copolyether 45 (x=0.55), even though these polymers undergo an isotropidnematic phase transformation during poling.
264
IV
Behavior and Properties of Side Group Thermotropic Liquid Crystal Polymers
Table 23. NLO coefficients d,, and d,, for 41 PA3,41 PMA3,45 (x=0.8), 45 (x=0.55), and 46".
41 PA3
51.5
41 PMA3
97
45 (x=0.8)
79.5
45 (x=0.55) 46 a
56 106
15 60 15
57 57 97
0.17 0.57 0.45
9.5 35.6 33.5
1.2 2.1 3.3
8k1.5 17+3 10+2
15 60 15 30
81 81 57 107
0.57 0.62 0.45 0.19
17.8 23.4 7.9 20
0.97 1.65 0.7 2.4
15+3 14+3
11+2 8+1.5
Corona poling, SHG measurements at 1.06 pm
The values of d33 and r=d33/d31predicted by the different theoretical models and the resulting evaluation of SHG enhancement with liquid crystalline axial order are given in Table 24 for polyacrylate 41 PA 3 and copolyether 45 (x=O.8). The data in Table 24 confirm that it is possible to gain in net polar ordering by starting from a liquidcrystalline system. The enhancement of the NLO performance with liquid-crystalline order is by a factor of 2.3-3.3 in qualitative agreement with the SKS [156] and MSVP [157-1591 model results. It should be noted, however, that the measured order parameters (PJ are lower than the calculated
ones, which may be due to the antiparallel dipole-dipole correlation evidenced in the nematic state by X-ray diffraction. Several additional interesting effects were noted in this study. First, it was shown that the NLO SCLCPs investigated exhibited enhanced stability over isotropic polymers. Second, it was found that the temperature at which the polymeric films were stored below Tgwas important in determining the rate of degradation of d,, . Excellent thermal stability was observed for 46 at room temperature, while a decrease in d,, was evidenced for 45 ( x = O . 8 ) , and, to a greater extent, 45 (x=0.55). Sub-T, mobil-
Table 24. Calculated and experimental values of (P2),(P4), d33,r = d,,/d3I , and A =calculated (or experimental) d,,/calculated isotropic d3,. Model 45 (x=0.8)
41 PA3
A
Isotropic SKS MSVP Ising Experimental "
0 0.68 0.77 1 0.62
0 0.33 0.44 1
Isotropic SKS MSVP Ising Experimental a
0 0.61 0.75' 1 0.57
0 0.26 0.40' 1
10.0 33.3 37.0 50.1 23.4 10.7 32.3 38.7 53.2 35.6
a At 1.064 pm; calculated with the MSVP model at zero field; Ep=2.9 MV/cm (f(o)=2).
3 10.8 13.0 m
14+3 3 9.2 11.9 M
17+3
1 3.3 3.7 5 2.34 1 3.0 3.6 5 3.3
calculated with the MSVP model at
4
ity of the polymer chain segments might account for these relaxation phenomena. However, it was found that the absorption spectra of the poled films did not change with time, implying that the axial order is frozen in the glassy state. It was suggested that the 6 relaxation process (reorientation of the side chains around the polymer backbone) could play a role in the stability of induced polar ordering. It can therefore be concluded that the temporal and thermal characteristics of the poling process must be well understood and controlled to maximize poling-induced nonlinearity and stability.
4.3.2.4 Ferroelectric LC Polymeric Systems Up to now we have considered the uniaxial nematic and smectic A phases. As already mentioned, these mesophases are centrosymmetric and in order to achieve second order NLO effects it is necessary to remove this centrosymmetry by inducing a net polar ordering through electric field poling. A challenge therefore is to modify an otherwise centrosymmetric system having a large first hyperpolarizability /3 to render it capable of exhibiting second-order NLO response. Molecular asymmetry imparts form chirality to LC phases, which results in the formation of helical ordering of the constituent molecules of the phase and imposes a reduction in the space symmetry [214], which leads to some phases having unusual NLO properties. Essentially, this section deals with chiral smectic C phases that have polar symmetry C2 and possess an easily measured macroscopic polarization once the helical structure is unwound. For low molar mass liquid crystals and SCLCPs,the common practice is to apply an external electric field or to use surface forces.
Nonlinear Optical Properties
265
Since the early days of ferroelectric low molar mass LC research, the exploration of the NLO properties of these substances has been a topic of interest. For a recent review on SHG in ferroelectric LCs,see [215]. Unfortunately, the SHG efficiency of most ferroelectric LCs investigated so far [2162191 is orders of magnitude below that of state-of-the-art inorganic crystals, such as lithium niobate. The main reason for the low electronic NLO activity is that the ferroelectric LC compounds and mixtures used in these investigations were optimized for display applications rather than for NLO performance. Quite recently, some research groups started the development of ferroelectric LC compounds specifically devoted to NLO applications, and indeed improved the SHG efficiency by orders of magnitude [220-2241. In order to create ferroelectric LC materials with improved second-order NLO activity, implanting well-known NLO chromophores into the core of standard ferroelectric LC molecules might be considered. However, since the direction of dipolar orientation of the ferroelectric LC molecules is parallel to the C2 axis (perpendicular to the director), these NLO active molecular groups have to be incorporated in such a fashion that the direction of maximum second-order polarizability is polarly oriented along this axis. This is expected to occur if the chromophore is coupled to the chiral center of the molecule, because the specific nature of the chiral group defines the tilt direction and hence the direction of the polar axis. This requirement is automatically fulfilled if the chirality is part of the NLO group. If this is not the case, it seems to be advantageous to locate the two groups as close to each other as possible. As already mentioned, NLO chromophores with large /3 values normally consist of a donor and an acceptor group which are
266
IV
Behavior and Properties of Side Group Thermotropic Liquid Crystal Polymers
linked by a conjugated spacer, often in the form of aromatic rings. The direction of large second-order polarizability is then roughly along the charge-transfer axis. In view of the symmetry requirements discussed above, the NLO chromophores have to be integrated in the ferroelectric LC structure with their long axis perpendicular to the long axis of the LC molecule. It follows that liquid crystallinity can only then be compatible with such a functionalization if the NLO group is sufficiently compact and the rigid core of the mesogen sufficiently long. On the basis of this design approach, Walba et al. [222] succeeded in synthesizing the (o-nitroa1koxy)phenyl biphenylcarboxylate 47 exhibiting a d22=d23value of 0.6 pm/V (de,=0.23 pm/V).
In order to achieve amorphous polar solids, some research groups have explored ferroelectric SCLCPs, and have shown that the basic rules governing the relationships between molecular structure and macroscopic ferroelectric LC properties are the same [225-2281. The main difference between low molar mass and polymer ferroelectric LCs, however, is the existence of stable glassy phases in most of the latter. Recently, Wischerhoff et al. [229] reported the synthesis, the thermal behavior, and the ferroelectric properties of liquid-crystalline copolysiloxanes 49 containing chiral
47
Quite recently, Schmitt et al. [223] presented highly efficient NLO ferroelectric LC materials (48) whose SHG properties were investigated in 13 pm thin, homeotropically aligned layers.
Positioning the chiral center in the alkoxy terminal chain adjacent to the nitroaniline dipole moment strongly enhances both P, and the SHG efficiency, leading to the largest SHG coefficients (e.g., d22=5 pm V-' when R = 4 3 - ( 3 - 0 - C 9
HI9) re-
ported so far for ferroelectric low molar mass LCs.
0
c=o I
49
mesogens and NLO chromophores. The investigators showed that in all the polymer series a noticeable amount of chromophore (up to about 30 mol%) could be incorporated without losing the chiral smectic C phase. However, approaching high chromo-
4
phore contents, a tendency for the copolymers to form the nonferroelectric smectic A phase could be detected. SHG measurements were performed on thin homeotropically aligned polymer films, with the polar axis perpendicular to the incoming beam. However, in the geometry used, the applied electric field was necessarily limited and did not allow the helix to untwist completely, leading to low SHG efficiency. The techniques used to obtain the untwisted SmC* phase structure in low molar mass LCs and SCLCPs are limited to thin layers. In contrast to this, LC elastomers can be macroscopically uniformly oriented by mechanical deformations [230], and this orientation process is not limited to thin samples or suitable dielectric anisotropy of the material. Furthermore, for LC elastomers the oriented structure can be chemically locked in by crosslinking, resulting in the so-called liquid single crystal elastomers 12311. Quite recently, Finkelmann and co-workers [232, 2331 showed that an appropriate mechanical deformation of an SmC* elastomer 50 yields a permanent macroscopically uniform orientation. This process also unwinds the helicoidal superstructure, and accordingly, frequency doubling is observed where the intensity of the SHG is directly related to the perfection of the uniform smectic layer orientation. The d22 and dZ3 coefficients for a highly oriented sample were reported to be 0.1 pm/V and 0.15 pm/V, respectively. Taking into account that only 50% of the mesogenic units in the LC elastomer are active groups, these values are of the same order as those reported for low molar mass LCs containing similar chromophores [222].
267
Nonlinear Optical Properties
0
0
I
Q c=o
I
6
$ 0I
F="
I
c=o I I
0 FH2
I
CH3-CH*
I
cn3-
C2H.5
P
CH *
I
C6H13
50
4.3.2.5 Rigid Rod-Like Polymers Rigid rod-like polymers represent a special class whose phase behavior differs distinctly from that observed in flexible polymers. They are able to form LC phases, and highly ordered structures are generated due to their anisotropic molecular shape. In addition, they exhibit superior intrinsic mechanical and electrical properties and have specific advantages over polymers with a flexible backbone, such as high temperature and dimensional stability, toughness, high modulus, and tensile strength. This unique combination of profitable properties makes rigid-rod polymers particularly interesting for NLO applications. However, investigations have been mainly limited to third-order NLO properties and applications [234,235]. Another concept, in its early stages of exploration, for obtaining large values of x(2) through the poling process as well as longterm stability, is rigid rod-like aromatic
268
IV
Behavior and Properties of Side Group Thermotropic Liquid Crystal Polymers
polyesters and poly(esteramide)s having long alkyl or alkoxy side chains and NLO chromophores that are either directly incorporated into the main chain [236, 2371, or covalently linked to the backbone 1236, 238-2411 by flexible spacers. An overview including discussions on synthesis and properties is given in [242]. Three rigid-rod polymers with chromophores in the main chain (51, 52, and 53)
to an Arrhenius-like rather than a WLF-like behavior. This indicated that the relaxation process is based on a local mechanism, which is not coupled to a collective motion within the main-chain layers. A number of rigid rod copolymers, 54 and 55, with chromophores in the side chains were synthesized. According to the X-ray diffraction results, these polymers exhibit-
51
52
53
CN -CN
were used to cast transparent films with good mechanical and optical properties. Well-ordered structures were observed with polymer main-chain layers parallel to the substrate and side chains, extending normal to the layer plane (Fig. 37). One such polymer (52) exhibited a x$!: value of 15 pm/V at 0.632 y m for a poling field of z 4 0 V/pm. These new polymers showed promise for further development in that the materials once poled retained the NLO activity over long periods of time (up to 920 h at 100" C ) . It was also found that the temperature dependence of the relaxation times conformed
Substrate
W A l k y l side chain
Rigid-rod main chain
bc Donor and acceptor chromophores,respectively Figure 37. Suggested structure of polymers 51-53.
4
Nonlinear Optical Properties
269
54
55
Q
Q N
N
II
II
r:yO/l
n=-
110 0.5 10.5 0.3 10.7
14 16 16
(I ed a layered structure, the space between the layers formed by the main chains being filled by the interdigitated side chains (Fig. 38). Polymer solutions were spin cast into optical quality thin films and the NLO properties were measured by the attenuation total reflection (ATR) method. For series 54, it was found that strongly depends on the length of the alkyl chains. A maximum value ofXyi=11.8 p d V (at 0.632 pm) was obtained for polymer with rn = 5 and n = 7 at
~$2
the poling field of 40 V/pm. In series 55, the and x?,) second-order susceptibilities as well as the piezoelectric coefficient, were linearly proportional to the poling voltage at a given temperature. The maximum values of xy; were 140 (at 0.632 pm) and 91 pm/V (at 0.685 pm) at a poling field of 79 V/pm. The thermal relaxation behavior investigated at various temperatures demonstrated the enhanced stability of polymers
~0
270
IV
Behavior and Properties of Side Group Thcrmotropic Liquid Crystal Polymers
Figure 38. Suggested structures for (a) polymer 54 ( m = 2, n = 15) with long side chains and (b) polymers 5 4 ( m = 5 , n = 7 a n d m = 2 , n=5) withshortsidcchains.
55 compared to side chain polymer systems with flexible main chains. This enhanced stability was ascribed to hindering of the local mobility of the chromophores by the stiff main chains and the matrix of long alkoxy side groups in which the chromophores are tightly embedded.
5 References [ l ] C. B. McArdle, Side Chain Liquid Crystal Polymers, Blackies, 1989. [2] J. C. Dubois, Phys. Sci. 1988, 23, 299-305.
[3] C. Noel, P. Navard, Prog. Polym. Sci. 1991, 16, 55-1 10. [4] M. V. Kozlovsky, L. A. Beresnev, Phase Trans. 1992,40, 129-169. [5] E. Chicllini, G. Galli, Mol. Cryst. Lig. Cryst. 1994,254, 17-36. [6] V. P. Shibaev, L. A. Beresnev, L. M. Blinov, N. A. PlatC, Polym. Bull. 1984, 12, 299-301. [7] S . Hachiya, K. Tomoike, K. Yuasa, S. Togawa, T. Sckiya, K. Takahashi, K. Kawasaki, J. SlD 1993, IN, 295-299. [8] V. Pcrcec, Q. Zheng, Polym. Bull. 1992, 29, 501-508. [9] B. Gallot, G. Galli, E. Dossi, E. Chiellini, Liq. Cryst. 1995, 18,463. [ 101 E. Chicllini, G. Galli, E. I. Serhatli, Y. Yagci, M. Laus, A. Sante Angeloni, Ferroelectrics 1993, 148,3 11-322. [ 111 E. Chicllini, G . Galli, F. Cioni, E. Dossi, B. Gallot, J. Muter. Chem. 1993, 3 (lo), 1065-1073. [12] A. Kocot, R. Wrzalik, J. K. Vij, R. Zentel, J. Appl. Phys. 1994, 75 (2), 728-733. [13] G. H. Hsiue, P. J. Hsich, S . L. Wu, C. S . Hsu, Polym. Bull. 1994, 33, 159-166. [14] A. Kocot, R. Wrzalik, J. K. Vij, M. Brehmcr, R. Zentel, Phys. Rev. B 1994, 50 (22), 1634616356. 11.51 R. Zcntel, M. Brehmcr, Adv. Matel: 1994, 6 (7/8), 598-599. [I61 M. Ozaki, M. Utsumi, K. Yoshino, K. Skarp, Jpn. J. Appl. Phys. 1993,32, L852-L855. [17] T. Ikeda, 0. Zushi, T. Sasaki, K. Ichimura, H. Takczoe, A. Fukuda, K. A. W. Skarp, Mol. Cryst. Liq. Cryst. 1993, 225, 67-79. [18] H. Endo, S. Hachiya, T. Sekiya, K. Kawasaki, Liq. Cryst. 1992, 12 ( I ) , 147-155. [19] K. Yuasa, S. Uchida, T. Sekiya, K. Hashimoto, K. Kawasaki, Pruc. SPlE1992,1665, 154-165. [20] L. C. Chicn, L. G. Gada, Macromolecules 1994, 27,3721-3726. [21] G. Scherowsky, Polym. Adv. Technol. 1992, 3, 219-229. [22] M. Pfeiffer, L. A. Bercsnev, W. Haase, G. Scherowsky, K. Kiihnpast, D. Jungbaucr, Mol. Cryst. Liq. Cryst. 1992, 214, 125-141. [23] E. Wischerhoff, R. Zcntel, M. Redmond, 0. Mondain-Monval, H. Colcs, Macromol. Chem. Phys. 1994, 195, 1593-1602. [24] M. Ido, K. Tanaka, S. Hachiya, K. Kawazaki, Ferroelectrics 1993, 148, 223-232. [25] G. Dccobert, F. Soycr, J. C. Dubois, Polym. Bull. 1985, 14, 179. [26] G. Decobert, J. C. Dubois, S . Esselin, C. Noel, Liq. Cryst. 1986, 1, 307. [27] S . Esselin, L. Bosio, C. Noel, G. Dccobert, J. C. Dubois, Liq. Cryst. 1987, 2, 505. [28] S . Esselin, C. Noel, G. Decobert, J. C. Dubois, Mol. Cryst. Liq. Cryst. 1988, 155, 371. [29] J. M. Guglielminetti, G. Dccobcrt, J. C. Dubois, Polym. Bull. 1986, 16, 411.
5
[30] V. Shibaev, M. Kozlovski, N. Plate, L. Beresnev, L. Blinov, Vissokolol Soedin 1987, 29 (7), 1470. [31] M. Durnon, H. T. Nguyen, Polym. Adv. Technol. 1992,3, 197-203. [32] H. Endo, S . Hachiya, K. Kawasaki, Liq. Cryst. 1993, 13 (51, 721-728. [33] B. Hahn, V. Percec, Macromolecules 1987, 20 (12), 2961. [34] S. Ishibashi, K. Takahashi, S . Ishizawa, F. Yarnarnoto, Ferroelectrics 1993, 148, 277-283. [35] P. Keller, Ferroelectrics 1988, 85,425. [36] C.-S. Hsu, L.-J. Shih, G.-H. Hsiue, Macromolecules 1993, 26, 3 161-3 167. [37] Z. Z. Zhong, D. E. Schuele, W. L. Gordon, L.-C. Chien. A. J. Walz, Mol. Cryst. Liq. Cryst. 1994. [38] C.-S. Hsu, J.-H. Lin, L.-R. Chou, G.-H. Hsiue, Macromolecules 1992,25, 7126-7134. [39] H. Poths, E. Wischerhoff, R. Zentel, A. Schonfeld, C . Henn, F. Krerner, Liq. Cryst.: in press. [40] H. Poths, R. Zentel, Liq. Cryst. 1994, 16 ( 5 ) , 749-767. [41] A. Schonfeld, F. Krerner, H. Poths, R. Zentel, Proc. SPIE 1994,2175,60-70. [42] H. Poths, A. Schonfeld, R. Zentel, F. Krerner, K. Siemensrneyer, Adv. Mater: 1992, 4 (5), 35 1-354. [43] D. M. Walba, D. A. Zurnmach, M. D. Wand, W. N. Thurmes, K. M. Moray, K. E. Arnett, Proc. SPIE 1993,1911,21-28. [44] H. Poths, R. Zentel, D. Hermann, G. Andersson, K. Skarp, Ferroelectrics 1993, 148, 285295. [45] H. Kuhnpast, J. Springer, G. Scherowsky, F. GieRelrnann, P. Zugenrnaier, Liq. Cryst. 1993, 14 (3), 861-869. [46] P. Davidson, K. Kuhnpast, J. Springer, G. Scherowsky, Liq. Cryst. 1993, 14 (31, 901-910. [47] T. Kitazurne, T. Ohnogi, K. Ito, J. Am. Chem. SOC. 1990, 112,6608-6615. [48] T. Kitazume, T. Ohnogi, J . Flux Chem. 1990,47, 459-466. [49] J. Bornelburg, G. Heppke, J. Hollidt, Makromol. Chem., Rapid Commun. 1991,12,483-488. [50] H. Poths, R. Zentel, Macromol. Rapid. Commun. 1994, 15,433-440. (511 G. Scherowsky, B. Brauer, K. Grunegerg, U. Miiller, L. Komitov, S. T. Lagerwall, K. Skarp, B. Stebler, Mol. Cryst. Liq. Cryst. 1992, 215, 257-270. 1521 H. Coles, R. Simon, H. Gleeson, J. Bone, G. Scherowsky, A. Schliwa, U. Miiller, Polym. Prep. Jpn. (Engl. Ed.) 1990, 39, 1. [53] S. U. Vallerien, F. Kremer, G. Scherowsky, A. Schliwa, K. Kuhnpast, E. W. Fischer, Liq. Cryst. 1990, 8 (5). 719-725. [54] G. Scherowsky, K. Griineberg, K. Kuhnpast, Ferroelectrics 1991, 122, 159-166.
References
27 1
1551 S. Uiie, K. Iimura, Polym. J. 1991, 23, 14831488. [56] S. Uiie, K. Iirnura, Ferroelectrics 1993, 148, 263-269. [57] H. J. Coles, H. F. Gleeson, G. Scherowsky, A. Schliwa, Mol. Cryst. Liq. Cryst. 1990, 7, 117. [58] R. Zentel, G. Reckert, B. Reck,Liq. Cryst. 1987, 2 ( I ) , 83. [59] S. Bualek, R. Zentel, Makromol. Chem. 1988, 189, 797. [60] H. Kapitza, R. Zentel, Makromol. Chem. 1988, 189, 1793. [61] R. Zentel, Liq. Cryst. 1988, 3, 531. [62] R. Zentel, H. Poths, F. Krerner, A. Schonfeld, D. Jungbauer, R. Twieg, C . G. Willson, D. Yoon, Polym. Adv. Technol. 1992,3, 2 1 1-2 17. [63] P. E. Cladis, H. R. Brand, Liq. Cryst. 1993, 14 ( 5 ) , 1327-1349. [64] K. Kiihnpast, J. Springer. P. Davidson, G. Scherowsky, Makromol. Chem. 1992,193,3097-3 1 15. [65] S . Uchida, K. Morita, K. Miyoshi, K. Hashirnoto, K. Kawasaki, Mol. Cryst. Liq. Cryst. 1988, 155,93-102. [66] F. Hardouin, A. M. Levelut, M. F. Achard, G. Sigaud, J. Chim.Phys. 1983, 80, 53. [67] A. Schonfeld, F. Krerner, R. Zentel, Liq. Cryst. 1993,13 (3), 403-412. [68] R. B. Meyer, L. Liebert, L. Strzelecki, P. Keller, J. Phys. 1975, 36, L-69. [69] N. Clark, S. Lagerwall, Ferroelectrics 1984,59, 25. [70] P. Martinot-Lagarde, Ferroelectrics 1988,84,53. [71] J. W. Goodby, E. Chin, J. M. Geary, J. S. Patel, P. L. Finn, J. Chenz. Soc., Faruduy Trans. 1987, 83, 3429. [72] J. S. Patel, J. W. Goodby, Chem. Phys. Lett. 1987, 137, 91. [73] S . K. Prasad, G. G. Nair, Mol. Cryst. Liq. Cryst. 1991, 202, 91. [74] S . Kirnura, S. Nishiyarno, Y. Ouchi, H. Takezoe, A. Fukuda, Jpn. J. Appl. Phys. 1987, 26, L255. [75] S. Pfeiffer, R. Shashidhar, J. Naciri, S . Mery, SPlE 1992,1665, 166-174. [76] S. Garoff, R. Meyer, Phys. Rev. Lett. 1977, 38, 848. [77] E. Chiellini, G. Galli, Mol. Cryst. Liq. Cryst. 1994, 254, 17-36. [78] G. Scherowsky, K. Kuhnpast, J. Springer, Makromol. Chem., Rapid Commun. 1991, 12, 381385. [79] K. Skarp, G. Anderson, F. Gouda, S. T. Lagerwall, H. Poths, R. Zentel, Polym. Adv. Technol. 1992, 3, 241. [80] A. D. C. Chandani, T. Hagiwara, Y. Suzuki, Y. Ouchi, H. Takezoe, A. Fukuda, Jpn. J. Appl. P ~ J J S1988, . 27, L729. [ S l ] P. G. de Gennes, C. R. Acud. Sci. Paris 1975, 281, 101.
272
IV
Behavior and Properties of Side Group Thermotropic Liquid Crystal Polymers
[82] P. G. de Gennes in Polymer Liquid Crystals (Eds.: A. Ciferri, W. R. Krigbaum, R. B. Meyer), Academic Press, New York, U.S.A. 1982, Chap. 5. [83] M. Warner in Side Chain Liquid Crystal Polymers (Ed.: C. B. McArdle), Blackie, Glasgow, U. K. 1989, Chap. 2. [84] E. Gramsgergen, L. Longa, W. de Jeu, Phys. Rep. 1986,135, 195. [ 8 5 ] J. Schatzle, W. Kaufhold, H. Finkelmann, Makromol. Chem. 1989, 190,3269. [86] J. P. Jarry, L. Monnerie, Macromolecules 1979, 12 (2), 316. [87] B. Erman, L. Monnerie, Macromolecules 1989,22 ( 8 ) ,3348. [88] M. Warner, K. P. Gelling, T. A. Vilgis, J. Chem. Phys. 1988,88,4008. [89] M. Warner, Phys. Scr: 1991, T35, 53. [90] B. Deloche, E. T. Samulski, Macromolecules 1981, 14, 575. [91] B. Deloche, E. T. Samulski, Macromolecules 1988,21, 3107. 1921 J. Kupfer, H. Finkelmann, Makromol. Chem., Rapid Commun. 1991, 12,717. [93] C. H. Legge, F. J. Davis, G. R. Mitchel, J. Phys. I1 France 1991, 1, 1253. [94] H. Finkelmann, H. J. Koch, G. Rehage, Makromol. Chem., Rapid Commun. 1981, 2, 317. 1951 W. Gleim, H. Finkelmann in Side Chain Liquid Crystal Polymers (Ed.: C. B. McArdle), Blackie, Glasgow, U. K. 1989, Chap. 10. [96] R. Zente1,Angew. Chem., Adv. Muter. 1989,10, 1437. [97] R. Zentel, M. Benalia, Makromol. Chem. 1987, 188, 655. [98] C. Degert, H. Richard, M. Mauzac, Mol. Crysr. Liq. Cryst. 1992,214, 179. [99] G. R. Mitchell, F. J. Davis, A. Ashman, Polymer 1987,28,639. [loo] F. J. Davis, A. Gilbert, J. Mann, G. R. Mitchell, J . Polym. Sci., Polym. Chem. 1990, 28, 1455. [I011 M. Brehmer, R. Zentel, Mol. Cryst. Liq. Cryst. 1994,243,353. [I021 M. H. Litt, W. T. Whang, K. T. Yen, X. J. Qian, J. Polym. Sci., Polym. Chem. 1993,31, 183. [I031 A. J. Symons, F. J. Davis, G. R. Mitchell, Liq. Cryst. 1993, 14 (3), 853. [ 1041 P. J. Flory in Principles of Polymer Chemistry, Cornell University Press, Ithaca, New York 1953. [I051 L. Treloar in The Physics of Rubber Elasticity, Clarendon, Oxford 1975. [I061 R. Loffler, H. Finkelmann, Makromol. Chem., Rapid Commun. 1990,11,321. [lo71 J. Schatzle, W. Kaufhold, H. Finkelmann, Makromol. Chem. 1989,190, 3269. [lo81 W. Guo, F. J. Davis, G. R. Mitchell, Polymer 1994,35 (14), 2952.
[ 1091 C. Degert, Ph. D. Thesis, Universitt de Bor-
deaux I, France 1991. [ 1101 J. Bastide, F. Boue, M. Buzier in Proc. in Phys-
ics-N" 42: Molecular bases of polymer networks, Springer, Berlin 1989, p. 48. [ 1111 F. J. Davis, J. Muter. Chem. 1993, 3 (6), 55 1. [ 1121 R. V. Talroze, T. I. Gubina, V. P. Shibaev, N. A. Plate, V. I. Dakin, N. A. Shmakova, F. F. Sukhov, Makromol. Chem., Rapid Commun. 1990, 11, 67. [ 1131 W. Oppermann, K. Braatz, H. Finkelmann, W. Gleim, H. J. Kock, G. Rehage, Rheol. Acta 1982, 21,423. [ 1141 H. J. Kock, H. Finkelmann, W. Gleim, G. Rehage, Polym. Sci. Technol. 1985, 28, 275. [ 1151 W. Gleim, H. Finkelmann, Makromol. Chem. 1987,188, 1489. [ 1161 T. Pakula, R. Zentel, Makromol. Chem. 1991, 192,2410. [117] R. A. M. Hikmet, D. J. Broer, Polymer 1991, 32 (91, 1627. [ 1181 J. L. Gallani, L. Hillion, P. Martinoty, F. Doublet, M. Mauzac, J. Phys. IIFrance 1996,6,443. [119] D. J. Broer, R. G. Gossink, R. A. M. Hikmet, Angew. Makromol. Chem. 1990,45,3235. [120] H. Finkelmann, H. J. Kock, W. Gleim, G. Rehage, Makromol. Chem., Rapid Commun. 1984, 5, 287. [121] G. R. Mitchell, F. J. Davis, W. Guo, R. Cywinski, Polymer 1991, 32 (a), 1347. [122] C. Degert, P. Davidson, S. Megtert, D. Petermann, M. Mauzac, Liq. Cryst. 1992,12 (51,779. [123] W. Kaufhold, H. Finkelmann, H. R. Brand, Makromol. Chem. 1991,192, 2555. [I241 J. Schatzle, H. Finkelmann, Mol. Cryst. Liq. Cryst. 1987, 142, 85. [125] R. Zentel, M. Benalia,Makromol. Chem. 1987, 188, 665. [126] J. Kupfer, H. Finkelmannn, Macromol. Chem. Phys. 1994, 195 (4), 1353. [127] R. G. Kirste, H. G. Ohm, Makromol. Chem., Rapid Commun. 1985,6, 179. [128] L. Noirez, G. Pepy, P. Keller, L. Benguigui, J . Phys. France 1991, I , 821. [129] F. Moussa, J. P. Cotton, F. Hardouin, P. Keller, M. Lambert, G. Pepy, M. Mauzac, H. Richard, J. Phys. France 1987,48, 1079. [130] S. Lecommandoux, L. Noirez, M. Mauzac, F. Hardouin, J. Phys. II France 1994,4, 2249. [131] P. Davidson, A. M. Levelut, Liq. Cryst. 1992, I 1 (4), 469. [ 1321 X. J. Wang, M. Warner, . I Phys. . A: Math. Gen. ( U . K.) 1987, 713. [I331 G. G. Barclay, C. K. Ober, Prog. Polym. Sci. 1993,18, 899. [I341 R. A. M. Hikmet, J. A. Higgins, Liq. Cryst. 1992, 12 (5), 83 1. [135] H. R. Brand, Makromol. Chem., Rapid Commun. 1989, 10,441.
5 [ 1361 H. R. Brand, H. Pleiner, Makromol. Chem.,
Rapid Cornmun. 1990, 11, 607. [I371 E. M. Terentjev, M. Warner, J. Phys. 11 France 1994,4, 111. [ 1381 R. Zentel, Liq. Cryst. 1988, 3 (4), 53 1. 11391 R. Zentel, G. Reckert, S. Bualek, H. Kapitza, Makromol. Chem. 1989, 190,2869. [140] W. Meier, H. Finkelmann, Makromol. Chem., Rapid Commun. 1990,11, 599. [141] S. U. Vallerien, F. Kremer, E. W. Fischer, H. Kapitza, R. Zentel, H. Poths, Makromol. Chem., Rapid Commun. 1990,II, 593. 11421 A. Schonfeld, F. Kremer, S . U. Vallerien, H. ' Poths, R. Zentel, Ferroelectrics 1991, 121, 69. [ 1431 H. Hirschmann, W. Meier, H. Finkelmann, Makromol. Chem., Rapid Commun. 1992, 13, 385. [I441 R. A. M. Hikmet, Macronzolecules 1992, 25, 5759. [ 1451 W. Meier, H. Finkelmann, Macromolecules 1993,26, 1811. 11461 M. Brehmer, R. Zentel, Macromol. Chem. Phys. 1994,195, 1891. [147] M. Mauzac, H. T. Nguyen, F. G. Tournilhac, S. V. Yablonsky, Chem. Phys. Lett. 1995, 240, 461. [I481 A. F. Garito, K. D. Singer, C. C. Teng in Nonlinear Optical Properties of Organic and Polymeric Materials (Ed.: D. J. Williams), ACS Symp. Ser. 233, Washington, DC 1983, pp. 1 26. 11491 D. J. Williams, Angew. Chem., Int. Ed. Engl. 1984, 23, 690. [ 1501 K. D. Singer in Polymersfor Lighfwaveand Integrated Optics. Technology and Applications (Ed.: L. A. Hornak), Marcel Dekker, New York 1992, pp. 321-342. [I511 M. Goodvin, D. Bloor, S. Mann in Special Polymers for Electronics and Optoelectronics (Eds.: J. A . Chilton, M. T. Goosey), Chapman and Hall, London 1995, pp. 131-185. [I521 T. M. Leslie, R. N. Demartino, E. Won Choe, G. Khanarian, D. Haas, G. Nelson, J. B. Stamatoff, D. E. Stuetz, C. C. Teng, H. N. Yoon, Mol. Cryst. Liq. Clyst. 1987, 153, 451. [I531 P. Pantelis, J. R. Hill, S . N. Oliver, G. J. Davies, Br. Telecom Technol. J. 1988, 6 , no. 3. [I541 J. C. Dubois, P. Le Barny, P. Robin, V. Lemoine, H. Rajbenbach, Liq. Cryst. 1993, 14, 197. 11551 G. R. Meredith, J. G. Van Dusen, D. J. Williams, Macromolecules 1982,.15, 1385. 11561 K. D. Singer, M. G. Kuzyk, J. E. Sohn, J. Opt. Sac. Am. B: Opt. Phys. 1987,4, 968. [157] C. P. J. M. Van der Vorst, S . J. Picken, Proc. SPIE 1987,866,99. [I581 C. P. J. M. Van der Vorst, S . J . Picken, J. Opt. Soc. Am. 1990, B7, 320.
References
273
11591 G. R. Mohlmann, C. P. J. M. Van der Vorst in Side Chain LiquidCrystal Polymers (Ed.: C . B. McArdle), Blackie, Glasgow 1989, pp. 330356. [I601 P. N. Prasad, D. J. Williams, Introduction to Nonlinear Optical Effects in Molecules and Polymers, Wiley, New York 1991. [I611 G. R. Meredith, J. G. Van Dusen, D. J. Williams in Nonlinear Optical Properties of Organic and Polymeric Materials (Ed.: D. J. Williams), ACS Symp. Ser. 233, Washington, DC 1983, pp. 109-133. [I621 W. Maier, A. Saupe, Z. Naturforsch. 1958, 13A, 564; 1959,14A, 882; 1960, 15A, 287. [I631 J. B. Stamatoff, A. Buckley, G. Calundann, E. W. Choe, R. De Martino, G. Khanarian, T. Leslie, G. Nelson, D. Stuetz, C. C. Teng, H. N. Yoon, Proc. SPIE Int. Soc. Opt. Eng. 1987,682, 8 5 . [I641 J. C. Dubois in Conjugated Polymeric Materials: Opportunities in Electronics, Optoelectronics and Molecular Electronics (Eds.: J. L. BrCdas, R. R. Chance), Kluwer Academic, Dordrecht 1990, pp. 321-340. 11651 L. T. Cheng, W. Tam, G. R. Meredith, G. L. J. A. Rikken, E. W. Meijer, SPlE Nonlinear Optical Properties of Organic Materials I1 1989,1147,61. [I661 R. A. Huijts, G. L. J. Hesselink in Nonlinear Optical Effects in Organic Polymers (Eds.: J. Messier et al.), Kluwer Academic, Dordrecht 1989, pp. 101-104. [I671 I. Ledoux, J . Zyss, A. Jutand, C. Amatore, Chem. Phys. 1991, 150, 117. 11681 D. Li, T. J. Marks, M. A. Ratner, Chem. Phys. Lett. 1986, 131, 370. [ 1691 J. S. Moore, S . I. Stupp, Macromolecules 1987, 20, 273. [ 1701 S. I. Stupp, H. C. Lin, D. R. Wake, Chemistry of Materials 1992,4,947;Bull. Am. Phys. Soc. 1990, 35, 560. [171] M. Amano, T. Kaino, F. Yamamoto, Y. Takeuchi, Mol. Cryst. Liq. Cryst. 1990, 182A. 81. [ 1721 V. Percec, C. Pugh in Side Chain Liquid Crysfa1 Polynzers (Ed.: C. B. McArdle), Blackie, Glasgow 1989, pp. 30-105. [ 1731 C. Noel, C. Friedrich, V. LConard, P. Le Barny, G. Ravaux, J. C. Dubois, Makromol. Chern., Macromol. Syrnp. 1989, 24, 283. [ 1741 P. Le Barny. G. Ravaux, J. C. Dubois, J. P. Parneix, R. Njeumo, C. Legrand, A. M. Levelut, Proc. SPIE Int. Soc. Opt. Eng. 1986, 682, 56. [ 1751 C. Legrand, C. Bunel, A. Le Borgne, N. Lacoudre, N. Spassky, J. P. Vairon, Makrumol. Chem. 1990, 191, 2971. [ 1761 N. Koide, S. Ogura, Y. Aoyama, M. Amano, T. Kaino, Mol. Cryst. Liq. Crysf. 1991, 198, 323. 11771 D. A. McL. Smith, H. J. Coles, Liq. Crlvst.1993, 14,937.
274
IV
Behavior and Properties of Side Group Thermotropic Liquid Crystal Polymers
[178] S. Yitzchaik, G. Berkovic, V. Krongauz, Opt. Lett. 1990, 15, 1120. [I791 S . H. Ou, J. A. Mann, J. B. Lando, L. Zhou, K. D. Singer, Appl. Phys. Lett. 1992,61,2284. [ 1801 S. Yitzchaik, G. Berkovic, V. Krongauz, Chem. Mater. 1990, 2, 162. [181] M. R. Worboys, N. A. Davies, M. S. Griffith, D. G. McDonnell, Proc. of2nd Int. Con$ on Electrical, Optical and Acoustic Properties of Polymers, Canterbury, UK 1990, 18/1-18/6. [182] C. P. J. M. Van der Vorst, S. J. Picken, Proc. SPIE-Int. Soc. Opt. Eng. 1993,2025,243. [I831 L. G. P. Dalmolen, S. J. Picken, A. F. de Jong, W. H. de Jeu, J. Physique 1985,46, 1443. [I841 S . Xie, A. Natansohn, P. Rochon, Chem. Mater. 1993,5,403. [185] I. Cabrera, V. Krongauz, Nature 1987, 326, 582; Macromolecules 1987,20, 2713. [I861 S. Yitzchaik, G. Berkovic, V. Krongauz, Adv. Mater. 1990, 2, 33. 1871 I. Langmuir, V. J. Schaefer, J. Am. Chem. Soc. 1938,57, 1007. 1881 D. Gonin, C. Noel, A. Le Borgne, G. Gadret, F. Kajzar, Makromol. Chem., Rapid Commun. 1992, 13, 537. 1891 I. A. McCulloch, R. T. Bailey, Mol. Cryst. Liq. Cryst. 1991, 200, 157. 1901 W. T. Ford, M. Bautista, M. Zhao, R. J. Reeves, R. C. Powell, Mol. Cryst. Liq. Cryst. 1991,198,351. [191] A. C. Griffin, A . M . Bhatti, R. S. L. Hung, Proc. SPIE 1986,682,65. [192] A. C. Griffin, A. M. Bhatti, R. S . L. Hung,Mol. Cryst. Liq. Cryst. 1988, 155, 129. [I931 A. C. Griffin, A.M. Bhatti, R. S . L. Hung in Nonlinear Optical and Electrnactive Polymers (Eds.: P. N. Prasad, D. R. Ulrich), Plenum, New York 1988, p. 375. [194] M. Zhao, M. Bautista, W. T. Ford, Macromolecules 1991, 24, 844. [195] I. A. McCulloch, R. T. Bailey, Proc. SPIE 1989,1147, 134. [I961 A. C. Griffin, A. M. Bhatti, Org. Matel: Nonlinear Opt. 1989, 69, 295. [I971 M. 0. Bautista, W. T. Ford, R. S. Duran, M. Naumann, Div. Polyrn. Chem., Am. Chem. Soc., Polyrn. Prep. 1992,33, 1172. [I981 P. L. Carr, G. R. Davies, I. M. Ward, Polymer 1993, 34, 5. 11991 W. M. K. P. Wijekoon, Y. Zhang, S. P. Karna, P. N. Prasad, A. C. Griffin, A. M. Bhatti, J. Opt, Soc. Am. 1992, B9, 1832. [200] R. B. Findlay, T. J. Lemmon, A. H. Windle, Matel: Res. SOC.Symp. Proc. 1990, 175, 305; 1. Mater. Res. 1991, 6, 604. [201] R. N. De Martino, E. W. Choe, G. Khanarian, D. Haas, T. Leslie, G. Nelson, J . Stamatoff, D. Stuetz, C. C. Teng, H. Yoon in Nonlinear Optical and Electroactive Polymers (Eds.:
[202] [203] [204] [205] [206] [207] [208] [209] [210] [211] [212]
[213] [214] [215] [216] [217] [218] [219] [220]
P. N. Prasad, D. R. Ulrich), Plenum, New York 1988, pp. 169-187. M. Schulze, Trends Polym. Sci. 1994, 2, 120. T. M. Leslie, US Patent 4 807 968, 1989. R. N. De Martino, European Patent ApplicationO294706,1988(Priority 1987 US 58414). R. N. De Martino, H. N. Yoon, J. B. Stamatoff, European Patent Application 0271 730, 1987 (Priority: 1986 US 933425). B. Reek, H. Ringsdorf, Makromol. Chem., Rapid Commun. 1985,6,291. S. Piercourt, N. Lacoudre, A. Le Borgne, N. Spassky, C. Friedrich, C. Noel, Makromol. Chem. 1992,193,705. D. Gonin, Thkse de Doctorat de 1’UniversitC Paris, Paris 1994. B. Guichard, Thkse de Doctorat de 1’UniversitC Paris, Paris 1995. J. C. Dubois, G. Decobert, P. Le Barny, S. Esselin, C. Friedrich, C. Noel, Mol. Cryst. Liq. Cryst. 1986, 137, 349. D. R. Ulrich, Mol. Cryst. Liq. Cryst. 1990,189, 3. D. Gonin, G. Gadret, C. Noel, F. Kajzar in Nonlinear Optical Properties of Organic Materials VI (Ed.: G. Mohlmann), Proc. SPIE 1993,2025, 129. R. H.Page,M. C. Jurich,B. Reek, A. Sen,R. J. Twieg, S . D. Swalen, G. C. Bjorklund, C. G. Willson, J. Opt. Soc. Am. 1990, B7, 1239. J. W. Goodby, J. Muter. Chem. 1991, I , 307. I. Drevensek, R. Blinc, Condensed Matter News 1992,1, 14. J. Y. Liu, M. G. Robinson, K. M. Johnson, D. Doroski, Opt. Lett. 1990, 15, 267. A. Taguchi, Y. Oucji, H. Takezoe, A. Fukuda, Jpn. J. Appl. Phys. 1989, 28, L-997. M. Ozaki, M. Utsumi, T. Gotou, Y. Morita, K. Daido, Y. Sadohara, K. Yoshino, Ferroelectrics 1991, 121, 259. K. Yoshino, M. Utsumi, Y. Morita, Y. Sadohara, M. Osaki, Liq. Cryst. 1993, 14, 1021. D. M. Walba,M. B. Ros, N. A. Clark, R. Shao, K. M. Johnson, M. G. Robinson, J . Y. Liu, D. Doroski, Mol. Cryst. Liq. Cryst. 1991, 198, 51.
[221] J. Y. Liu, M. G. Robinson, K. M. Johnson, D. M. Walba, M. B. Ros, N. A. Clark, R. Shao, D. Doroski, J. Appl. Phys. 1991, 70, 3426. [222] D. M. Walba, M. B. Ros, N. A. Clark, R. Shao, M. G. Robinson, J. Y. Liu, K. M. Johnson, D. Doroski, J. Am. Chem. Soc. 1991, 113,5471. [223] K. Schmitt, R. P. Herr, M. Schadt, J. Fiinfschilling, R. Buchecker, X. H. Chen, C. Benecke, Liq. Cryst. 1993, 14, 1735. [224] M. Schadt, Liq. Cryst. 1993, 14, 73. [225] H. Kapitza, R. Zentel, R. J . Twieg, C. Nguyen, S. U. Vallerien, F. Kremer, C. G. Wilson, Adv. Mater. 1990, 2, 539.
5 12261 H. Kapitza, H. Poths, R. Zentel, Mukromol. Chem., Mucromol. Symp. 1991, 44. 1 17. [227] R. Zentel, H. Poths, F. Kremer, A. Schonfeld, D. Jungbauer, R. Twieg, C. G. Wilson, D. Yoon, Polym. Adv. Technol. 1992,3, 21 1. I2281 D. M. Walba, P. Keller, D. S . Parmar, N. A. Clark, M. D. Wand, J. Am. Chem. SOC.1989, 111,8273. [229] E. Wischerhoff, R. Zentel, M. Redmond, 0. Mondain-Monval, H. Coles, Mukromol. Chem. Phys. 1994,195, 1593. [230] W. Gleim, H. Finkelmann in Side Chain Liquid Crystalline Polymers (Ed.: C . B. McArdle), Blackie, Glasgow 1989, pp. 287-308. [231] J. Kupfer, H. Finkelmann, Mukromol. Chem., Rapid Commun. 1991, 12,717. [232] K. Semmler, H. Finkelmann, Polym. Adv. Technol. 1994, 5, 23 1. [233] I. BennC, K. Semmler, H. Finkelmann, Mucromol. Rapid. Commun. 1994, IS,295.
References
275
(2341 B. A. Reinhardt, Trends Polynz. Sci. 1993, I , 4. [235] H. S. Nalwa, Adv. Muter: 1993, 5, 341. [236] G. Wegner, D. Neher, C. Heldmann, H. J. Winkelhahn, T. Servey, M. Schulze, C. S. Kang, Muter: Rex Soc. Symp. Proc. 1994, 328, 15. [237] C. Heldmann, Ph. D. Thesis, University of
Mainz, Germany 1994. 12381 T. K. Servay, H. J. Winkelhahn, L. Kalvoda, M. Schulze, C. Boeffel, D. Neher, G. Wegner, Ber. Bunsenges. Phys. Chem. 1993, 97, 1272. [239] H. J. Winkelhahn, T. K. Servay, L. Kalvoda, M. Schulze, D. Neher, G. Wegner, Ber. Bunsenges. Phys. Chem. 1993, 97, 1287. [240] C.-S. Kang, H. J. Winkelhahn, M. Schulze, D. Neher, G. Wegner, Chem. Muter: 1994, 6 , 2159. [241] Z. Sekkat, C.-S. Kang, E. F. Aust, G. Wegner, W. Knoll, Chem. Muter. 1995, 7, 142. [242] M. Schulze, Trends Polym. Sci. 1994, 2, 120.
Handbook ofLiquid Crystals D. Demus, J. Goodby, G. W. Gray, H.-W. Spiess, V. Vil Copyright 0WILEY-VCH Verlag GmbH. 1998
Chapter V Physical Properties of Liquid Crystalline Elastomers Helmut R. Brand and Heino Finkelmann
1 Introduction In the preceding chapters the synthesis properties of linear liquid crystalline polymers are described, where different approaches exist to obtain the liquid crystalline state: rod-like or disc-like mesogenic units are either incorporated in the polymer backbone or are attached as side groups to the monomer units of the main chain. Following conventional synthetic techniques these linear polymers can be converted to polymer networks. Compared to low molar mass liquid crystals and linear liquid crystalline polymers, these liquid crystalline elastomers exhibit exceptional new physical and material properties due to the combination and interaction of polymer network elasticity with the anisotropic liquid crystalline state. In this chapter we review the physical properties of liquid crystalline elastomers using publications until 1996. For a review of the physical properties of strongly crosslinked liquid crystalline polymers, which give rise to anisotropic nonliquid crystalline materials (duromers or anisotropic solids) we refer to the recent review by Hikmet and Lub [ I ] . To organize this chapter efficiently we have split it into two sections: the behavior near phase transitions and the physical properties of the bulk phases. Accordingly, subsections contain material on one type of phase transition or on the bulk behavior of (typically) one specific liquid crystalline phase.
2 Phase Transitions 2.1 The Nematic-Isotropic Phase Transition 2.1.1 Mechanical Properties of Polydomains Much of the early physical work on liquid crystalline elastomers (LCE) focused on the mechanical properties of polydomain samples. Gleim and Finkelmann [2] studied the length of the sample as a function of temperature for a number of values of the applied load over a large temperature range including the vicinity of the nematic-isotropic transition. In addition the temperature dependence of the nominal stress was evaluated for constant sample deformation. As a result it was found that the length of the sample increased when cooling through the phase transition indicating an increasing degree of order in the orientation of the mesogens. In the pretransitional region above the transition it has been shown that a stressstrain relation similar to that of common rubbers occurs [ 3 ] . From the behavior of stress-strain curves just below the phase transition, but above the transition to a monodomain state it has been demonstrated that one can linearly extrapolate to stress zero and extract the length I,, of a corresponding monodomain sample [ 3 ] . When
27 8
V
Physical Properties of Liquid Crystalline Elastomers
this length is plotted as a function of reduced temperature (Fig. l), one finds a behavior that is reminiscent of the temperature behavior of the order parameter, a feature that will be discussed in more detail below. The change in length of an elastomer under a constant external load has been studied in detail in the immediate vicinity of the phase transition [4]. In Fig. 2 the strain
-.
1 u
cn 5 2 L
_J
" . " " " '
0.9L
0.96 0.98 1.00 1.02 Reduced temperature, Tred
Figure 1. Length 1, of a monodomain elastomer vs. the reduced temperature Tred(reproduced with permission from [3]).
il=L/Lk is plotted as a function of the reduced temperature Tred at constant nominal stress 0 , = 2 . 1 1 ~ 1 0 -N ~ mm-2. Here L ; is the loaded sample length at Tred = 1.03. These results will also be used below to establish a close connection between the strain tensor and the nematic order parameter. It has also been shown that a quadratic stressstrain relation yields in the isotropic phase above the nematic-isotropic phase transition a good description of the data for eleongations up to at least 60% [4]. It has been demonstrated that a uniaxial compression of cylindrical samples along their axis leads to a reorientation of the mesogenic groups orthogonal to their orientation under strain [5], thus complementing experiments done earlier [2]. Furthermore, it has been shown that this reorientation of the mesogenic side groups takes place when compression changes into strain. Measurements of the length and the diameter of cylindrical samples as a function of temperature were used to determine the linear expansion coefficients parallel (PI,) and perpendicular to the cylinder axis. It is found that both, (PI,)and peak sharply in the vicinity of the nematic-isotropic transition (Fig. 3). Measurements of the temperature dependent elongation ilover a large temperature range including the nematic-isotropic transition for fixed load have been reported [6]. Depending on the compound studied it was found that the elongation either peaks or dips at this phase transition [6]. Only comparatively few dynamic mechanical measurements in the vicinity of the nematic-isotropic transition have been reported; these include some for side-chain liquid crystalline elastomers [7, 81, and some for liquid crystalline elastomers with mesogenic groups in the main chains and in the side-chains (combined LCE) [9]. The latter reference describes the evaluation of
(Pl)
Reduced temperature, Tred
Figure 2. Strain il= L / L , vs. reduced temperature Tredat constant nominal stress (reproduced with permission from [4]).
(PL)
2 8.51
I
Temperature, T ( " C )
(4
"*/ 260
270
280
290 300 Temperature, J ( K )
t
0
-\.
0.2'io (b)
the shear storage modulus G' and the shear loss modulus G" as a function of temperature across various liquid crystalline phase transitions including the nematic-isotropic transition. It is found that the storage modulus always stays higher than the loss modulus in the vicinity of the nematic-isotropic transition. A torsional oscillator was used to measure G' and G" in a side-chain LCE and it was found that both, G' and G" show a jump at the phase transition and that G" is smaller than G' on both sides of the nematic-isotropic phase transition [7]. Dynamic mechanical and dynamic stressoptical measurements have been performed on side-chain LCEs with different crosslinking densities [8]. It was found that the relaxation strengths depend strongly on the crosslinking density demonstrating that the vicinity of the crosslinking points perturbs the liquid crystalline order. In this experiment the samples were exposed to a timedependent stress ( ~ ( t )and , the time-dependent responses, that is the time-dependent strain ~ ( tand ) the time-dependent birefringence A n ( t ) , were measured. The real part
D o o o ~ e
1.0
310
Figure 3. Linear thermal expansion coefficient as a function of the temperature T of a cylindrical network determined parallel (B,,)and perpendicular (&) to the longitudinal axis of the unloaded sample (reproduced with permission from [ 5 ] ) .
279
Phase Transitions
io
io
60
yo
8i
Temper at ur e , T ( C)
Figure 4. Temperature dependence of log (E') and tan (6) for nematic elastomers measured at 30 Hz (reproduced with permission from [S]).
of the dynamic elastic modulus, E', and the loss factor, 6, obtained at a frequency f=30 Hz are shown in Fig. 4. From this figure it emerges that the dynamic mechanical properties reveal no special feature around the nematic-isotropic transition (TNpI=: 60 "C) for the frequency chosen.
2.1.2 Mechanical Properties of Monodomains: Liquid Single Crystal Elastomers Static mechanical properties in the vicinity of the nematic-isotropic transition in liquid single crystal elastomers (LSCEs) have been investigated [lo, 111. In Fig. 5 the deformation LIL, (mon) is plotted as a function of the reduced temperature Tred.Here L,(mon) denotes the length of the LSCE at TN-,, the phase transition temperature of the nematic-isotropic phase transition and
280
V
Physical Properties of Liquid Crystalline Elastomers
0.85 0.90 0.95 Reduced temperature,
1.00
Tied
Figure 5. Deformation (LIL, (mon)) of a LSCE as a function of the reduced temperature Tred.L,(mon), length of the elastomer at TN-,(reproduced with permission from [lo]).
Tred= T/TN-,. The continuous behavior of the data in this graph in the vicinity of TN-I, as well as the fact that there is a finite deformation above TNpr,reflects the influence of the internal field, which is due to the synthetic process of LSCE described in section 3 of this chapter. In this process the mechanical stress applied during the second crosslinking step is frozen in giving rise to the internal strain field observed. The influence of the crosslinking density and annealing temperatures on the physical
properties of LSCEs has been studied [ 111. In Fig. 6 we have plotted the relative strength L,/L,,,, where L298is the length of the sample at the reference temperature Tref=298K, as afunction of temperature for various annealing temperatures. It can be seen from the curve taken at the highest annealing temperature ( T = 150 "C) that there is a much smaller length change in the vicinity of TNPIfor the higher crosslinking density reflecting the fact that actually a duromer is investigated. The nematic properties of the latter are confined to yielding an anisotropic material, which does not show, however, the director variability and dynamics of the more weakly crosslinked LSCE samples. We also note that one can infer from the other curves in the same figure that the crosslinking process is far from completion at lower annealing temperatures. Using X-ray investigations and optical measurements the reorientation behavior of nematic monodomain samples has been studied in detail by Kundler and Finkelmann [ 121. This reorientation behavior has been modeled using a bifurcation analysis of the macroscopic dynamic description by Wei-
r------
'1u m
(D
I
35 (a)
50
65
80
95
Temper at ur e , T(" C )
Figure 6. The relative length (LT/LZg8) of elastomers as a function of temperature: (a) Annealing temperature A = 150"C, * = 60 "C, (b) annealing temperature A = 150°C, + =75 "C, * = 4 0 T (reproduced with permission from [ I 11).
(b)
40 60 80 100 120 140 Tem per at ur e, T(" C )
2
Phase Transitions
28 1
lepp and Brand [13] and, subsequently, by Verwey et al. [ 141 using rubber elasticity for an anisotropic rubber. *
2.1.3 Optical Properties of Polydomains The first stress-optical measurements above the nematic-isotropic transition have been described shortly after the synthesis of side-chain LCE [15]. In this publication it was demonstrated that the stress-optical coefficient shows strong pretransitional effects, and that it changes sign when the spacer length between the mesogen and the polymeric backbone is changed from 3 to 4 indicating that the mesogenic groups are oriented predominantly parallel to the axis of deformation in one case and perpendicular to this axis in the other. Similar results for this stress-optical coefficient were obtained for a different family of LCEs 121. In addition it was shown that the birefringence, An, is linear in the applied stress for the range of temperatures investigated, a result one would expect on the basis of linear stress optics; these results are summarized in Fig. 7. Early measurements of the IR-dichroism in stretched samples of LCEs have been described [ 161. The dependence of the IR dichroism of the CN-stretching mode was used to obtain the absorption coefficients and to extract the order parameter as a function of temperature. However, while these measurements yield reliable results inside the nematic phase, the sensitivity of this method turns out to be not high enough to analyze quantitatively the influence of the applied stress on the order parameter and on TN-I.
Stress-optical experiments on cylindrical samples under uniaxial compression have also been performed IS]. It was found
287K
0 7 X
c -30
4 0; (r
289K
L
291K
al m 5 -20
E ._
m
293K -10 298K 303K 313K
0
7
0
0.5 1.0 1.5 2.0 2.5 3.0 S t r e s s , m l O * ( N rnm-Z)
Figure 7. Birefringence An versus true stress 0 of a nematic elastomer for different temperatures above clearing temperature T,-, as indicated (reproduced with permission from [2]).
that the stress-optical coefficient behaves qualitatively similarly to that in strained samples when approaching the nematicisotropic transition. There is, however, one important qualitative difference for the cylindrical samples under uniaxial compression: the reorientation of the mesogenic side groups is parallel (perpendicular) with respect to the direction of the mechanical force when it is perpendicular (parallel) to this direction for strain. These observations complement earlier results on the same family of compounds [2]. In addition it has been shown, using measurements of the birefringence An as a function of temperature, that unloaded cylindrical samples behave optically like biaxial crystals provided the director n of the nematic phase is oriented in the plane perpendicular to the cylinder axis
PI. The results and the analysis of detailed stress-optical measurements in the immediate vicinity of the phase transition have been reported [3,4]. In Fig. 8 the reciprocal
282
V
Physical Properties of Liquid Crystalline Elastomers
Reduced temperature, Tred
Figure 8. Reciprocal birefringence An-' vs. reduced temperature Tredof a nematic elastomer for different loads (* = 1 . 0 lo-', ~ + = 4 . 5 ~lO-'N mm-') (reproduced with permission from [3]).
birefringence is plotted as a function of the reduced temperature Tred. As one can see from this figure the extrapolated intersection of the curve through the experimental data with the abscissa depends on the applied stress. This implies in turn that the hypothetical second order phase transition temperature T* also depends on the applied stress as can be seen in Fig. 9. Using IR dichroism measurements in addition below TN-I it was shown that an increase in external stress leads to a shift of TN-I to higher temperatures, and to an increase in the nematic order parameter [3]. Both results are
in agreement with the Landau model for this phase transition by de Gennes [17]. Detailed stress-optical measurements have been analyzed to yield further information [4]. In Fig. 10 the birefringence (order parameter) was plotted as a function of reduced temperature for several nominal stresses on.These results were combined with the predictions of the Landau model and static stress-strain curves and led to a number of interesting consequences. In Fig. 11 the shift in the phase transition temperature is plotted as a function of nominal stress and shifts of up to 7.5 K were found; compared to maximum displacements by electric and magnetic fields of about 5 mK in low molecular weight materials. In Fig. 12 the birefringence An is shown as a function of strain k = L / L &at constant nominal N mm-'. A strictly stress on=2.11x I
1.0.
3.60
oo
1.18 v,
C
6
U
Y
a,
50
1.36
.-P
6 P
E .-m
L ..,, , w
,-.
,
,.,
J, ZL
0.2
1.12
C
?.
1
0
0.1
,,f
,,
3270
2
3
5
Q
YL
330
E
L
Stress,owxlO2(Nrnm-')
Figure 9. Phase transformation temperature T* vs. true stress owfor a nematic elastomer (reproduced with permission from [3]).
0.97 0.99 1.01 1.03 Reduced temperature, Tred
Figure 10. On the left side birefringence An and on the right side order parameter S vs. Tred= T/TN-,(o=o) for different nominal stresses on(reproduced with permission from [4]).
2
/.
/*
283
Phase Transitions
/
/
/
N
0
/
/ *
/
I
,
l A
1
3
7
L
Nominal stress,o,xl 02(Nmm-z)
0.5
1.0
1.5
2.0
& / ( T - T ; ) i n 10’/K
Figure 11. Shift of the phase transformation temperature AT=TN-I(u)-TN-I(u=O)vs. nominal stress on (reproduced with permission from [4]).
0‘
11
12
13
14
strain, h
Figure 12. Birefringence An vs. the strain 1 = L/Loat constant nominal stress on= 2.1 1x lo-* N mm-* (reproduced with permission from [4]).
linear relation is obtained for temperatures in the nematic phase as expected for quantities that can be used equivalently to characterize the state of order in a system [4]. The observed behavior in the isotropic phase can be analyzed using a Landau energy expression in the isotropic phase. A linear relation holds if the order parameter or the birefringence are plotted as a function of [(T-T;)]-’ (Fig. 13). This plot has also been used to determine the bilinear crosscoupling coefficient U between the strain and the order parameter in the Landau energy.
Figure 13. Birefringence An vs. & / ( T - T , )at two different constant stresses (reproduced with permission from [4]).
By making use of the form of the Landau energy proposed by de Gennes [ 171, all coefficients in this expression have been evaluated from experimental data [4]. Using these data to calculate U , one obtains a value that is considerably smaller than the one determined from the linearized analysis outlined above. As has been noted, however, a consistent analysis should not only include terms quadratic in the strains and bilinear coupling terms of the order parameter and the strain, but rather also nonlinear effects as well as nonlinear coupling terms between strain and the nematic order parameter [4]. The only existing publication describing dynamic optical properties in LCEs in the vicinity of the nematic-isotropic transition appear to be the measurements of the stressoptical coefficient [XI. It is found that there are two relaxation frequencies and strengths characterizing the dynamic birefringence of the LCE samples analyzed in the pretransitional regime. The authors ascribe the slow and fast processes observed to the lateral coupling of the spacers to the mesogens resulting in a difference in the rotational mobility for the two short axes. In addition they find that a higher crosslinking density leads to lower relaxation strengths reflecting the
284
V
Physical Properties of Liquid Crystalline Elastomers
t k
x x x
5 x
I I t n x
T ["CI
Figure 14. Temperature dependence of the relaxation frequency, ff,,,(reproduced with permission from [Sl).
perturbations of the nematic order in the vicinity of the crosslinking points. Both relaxation frequencies show critical behavior and it has been pointed out that this is in qualitative agreement with de Gennes' mean field analysis [8]. In Fig. 14 we have plotted the frequency of the fast relaxation process as a function of temperature in the isotropic phase. As will be discussed in the subsection on modeling the curvature in this figure can be interpreted in the framework of a mode-coupling theory.
2.1.4 Optical Properties of Monodomains: Liquid Single Crystal Elastomers Optical techniques to investigate the isotropic-nematic transition in liquid single crystal elastomers have been used [lo, 11, 181. The nematic order parameter S, (the modulus of the order parameter Q , ) has been determined by measuring the absorbances Alland A, using the integrated intensities of the CN absorption band with the incident light polarized parallel and perpendicularly to the director of the monodomain [lo, 181. The result is plotted in Fig. 15 as a function of reduced temperature. It is similar to these obtained for polydomains or-
dered macroscopically by an external mechanical field (compare the last subsection). The behavior in the immediate vicinity of the nematic-isotropic transition and, in particular, the presence of paranematic order above TNPI reflects the influence of the internal field of LSCE acquired in their synthetic process. In Fig. 16 we show the nematic order parameter S, obtained from the type of IR dichroism measurements just described as a function of the deformation (LIL,(mon)) of the same elastomer, where L,(mon) is the length of the LSCE at TNpI.As one can see
I/: L
aJ
Y 0.6. E
E
Ei
& 0.L.
B V
'Z
0.2'
i
8
.**.--.
0-
Reduced temperature, Tred
Figure 15. Nematic order parameter S, vs. reduced temperature Tredfor a LSCE (reproduced with permission from [IS]).
1.05
1.10
1.15
1.20
Deformation, L/L, (mon)
Figure 16. Nematic order parameter S, as function of the deformation (LIL, (rnon)). L,(mon), length of the LSCE at TN-, (reproduced with permission from [lo]).
2
A
A A A A
30
A
50 70 Temperature, (“C)
A
A
d
go
Figure 17. The birefringence, An, as function of temperature for differently crosslinked networks, * =weak, A = strong crosslinking density (reproduced with permission from [ I I]).
from this graph very clearly there is also for LSCE a strictly linear relationship between the deformation and the degree of nematic order underscoring once more the fact that the deformation can be used as an order parameter for liquid crystalline elastomers. The influence of the crosslinking density on the birefringence has been studied [ 111. For weakly crosslinked materials the results obtained coincide with the previous ones while for strongly crosslinked liquid crystalline elastomers the birefringence is only weakly affected in the vicinity of the phase transition as can be seen in Fig. 17. We stress, however, that the strongly crosslinked LSCE are duromers for which there are no independent macroscopic director degrees of freedom anymore, which are characteristic for the nematic liquid crystalline phase in low molecular weight materials, in sidechain polymers and in weakly crosslinked liquid crystalline elastomers.
2.1.5 Electric and Dielectric Properties While the effect of static electric and magnetic fields on the bulk of the nematic phase
Phase Transitions
285
has been investigated previously, such investigations, corresponding to the measurement of electric and magnetic birefringence, do not seem to exist in the pretransitional region of the nematic-isotropic transition. Their effect on equilibrium properties such as the transition temperature TN-Ican be expected to be even smaller than those for low molecular weight materials, for which T, can be shifted at best by about lop3K [19], since they also involve elastic deformations of the network via coupling effects between the fluctuations of the nematic order parameter and the network. Concerning the detection of fluctuations we note that the measurement of strain birefringence discussed earlier in this section yields much bigger effects. We close this short subsection by noting that there also appear to be neither dielectric nor electric conductivity measurements in the vicinity of the nematicisotropic transition in liquid crystalline elastomers.
2.1.6 NMR Investigations of Monodomains While there are not many NMR investigations of the isotropic-nematic transition in sidechain liquid crystal elastomers [20, 211, they turned out to be crucial in answering the question whether strained LCEs show a first order phase transition with a classical two phase region or whether there is a continuous change of the local degree of nematic order, since one is in a state beyond the critical point. As already mentioned earlier in this section this question could not be settled conclusively with the mechanical and optical investigations described above. In Fig. 18 we show the 2H-NMR spectra of both an uncrosslinked linear liquid crystalline polymer and a monodomain elasto-
286
V
372 K
Physical Properties of Liquid Crystalline Elastomers
1
320
330
(4
340
350
360
370
380
390
370
380
390
Temperature, J ( K )
z 0.6 vl
I
I
320 I
,
,
20
10
0
(a)
hulkHz
.
10
(b)
,
20
I
,
/
,
20
10
0
10
(b)
330
340
350
360
Temperature, T(K)
,
20
rlv/kHz
Figure 18. 'H-NMR spectra at different temperatures: (a) linear polymer; (b) LSCE (reproduced with permission from [2O]).
mer as a function of temperature. Comparing the two sets of spectra as the temperature is varied from a value above to one below the transition in several steps, one sees immediately, that the spectra of the polymer show a coexistence of the isotropic peak and of the two symmetric nematic peaks over a finite temperature range. This behavior is qualitatively different from that observed in the LSCE, where the peak in the isotropic phase is found to split symmetrically when passing into the nematic phase. In Fig. 19 we see that the degree of local nematic order (which is proportional to the quadrupolar splitting) shows a jump as a function of temperature as well as a two phase coexistence for the liquid crystalline polymer,
Figure 19. Order parameter as a function of temperature: (a) linear polymer; (b) LSCE (reproduced with permission from [ZO]).
while the LSCE with its built-in mechanical field shows a continuous decrease of the local degree of order as the temperature is increased and no two phase coexistence region. It is thus clear, that one is in a situation beyond a critical point for the LSCE due to the mechanical stress that was applied to the LSCE during its synthetic process. It should be pointed out that, in addition to the built-in mechanical field discussed so far, there is also a contribution from finite size effects on the nematic-isotropic transition in liquid crystalline elastomers due to the finite average distance between crosslinking points. This question has been discussed, and it has been pointed out that varying the crosslinking density aids in disentangling these two different types of effects [20-221.
2
1.0
-
z
,,
0.8 -
X /'
m
/
& + Q) 0.6 -
; 5
0.4
m
B
287
_ - - - - - - - - - - -x
x/--
W&
Phase Transitions
/
I I
I
____-__-----
&!+--g
-
---0
/ /
/ I
0.2-
I
i 0.0
'
I
'
I
'
I
.
I
'
I
'
I
.
In addition it was also stressed recently that one can deduce the degree of alignment of the director orientation from NMR measurements using a quantitative line shape analysis [21]. This is brought out in Fig. 20 where both, the degree of local nematic order (the order parameter modulus) and the degree of alignment of the director orientation are plotted as a function of an externally applied mechanical stress. While the degree of local nematic order is essentially constant as a function of the applied mechanical stress, the degree of alignment of the director orientation increases for low values of the external stress rather rapidly and saturates for higher values.
2.1.7 Models for the Nematic- Isotropic Transition for Poly- and Monodomains As the experimental investigations have focused predominantly on static properties of the nematic-isotropic transition, most of the theoretical papers have used a GinzburgLandau description involving an expansion in the nematic order parameter to describe static properties [4, 17, 22-25].
I
,
I
Figure 20. Nematic order parameter SN (0) and director order parameter S, (x) as function of internal mechanical stress o,,obtained from 2H NMR lineshape analysis (after [21]).
The most important coupling to deformations of the network is the one that is linear in both the strain of the network and the nematic order parameter. As has been discussed earlier in this section this leads to the consequence that the strain tensor can be used as an order parameter for the nematic-isotropic transition in nematic sidechain elastomers, just as the dielectric or the diamagnetic tensor are used as macroscopic order parameters to characterize this phase transition in low molecular weight materials. But it has also been stressed that nonlinear elastic effects as well as nonlinear coupling terms between the nematic order parameter and the strain tensor must be taken into account as soon as effects that are nonlinear in the nematic order parameter are studied [4, 251. So far, no deviation from classical mean field behavior concerning the critical exponents has been detected in the static properties of this transition and correspondingly there are no reports as yet discussing static critical fluctuations. To account for the mechanical reorientation properties of monodomains (LSCE), which differ depending on whether the crosslinking has been done in the nematic or the isotropic phase, the concept of frozen order has been suggested [22]. The essen-
288
V
Physical Properties of Liquid Crystalline Elastomers
tial idea is to introduce a splitting of the macroscopic nematic order parameter into two contributions, one associated with the vicinity of the crosslinking points and one with the conventional nematic order of the mesogenic units in the sidechains in the regions of the LSCE sufficiently far away from the crosslinking points. This decomposition turns out to be useful to account for a number of static and dynamic experimental results of these spatially inhomogeneous systems. As the vicinity of the crosslinking cannot reorient as easily, the elastic and the optic responses to external mechanical stress depend on the direction of the applied mechanical force as well as on the crosslinking density used. The polydomain-monodomain transition in nematic LCE was investigated [26] by incorporating a local anchoring interaction into the Ginzburg-Landau description of the nematic-isotropic transition in LCE [4, 17, 251. Since there are very few dynamic experimental investigations of pretransitional effects [8], not much modeling has been reported to date either. Based on work for the macroscopic dynamics of the nematic-isotropic transition in sidechain polymers [27 291, it has been suggested [28] that the nonmeanfield exponent observed in dynamic stress-optical experiments [8] can be accounted for at least qualitatively by the mode-coupling model [28, 291. Intuitively this qualitatively new dynamic behavior can be traced back to static nonlinear coupling terms between the nematic order parameter and the strain tensor.
2.1.8 X-ray Investigations While there have been numerous X-ray investigations of the bulk nematic phase, a systematic investigation of the nematic-iso-
tropic transition (or of any phase transition in liquid crystalline elastomers) using Xrays does not exist. We note, however, that already the second publication describing the properties of sidechain liquid crystalline elastomers compared X-ray data obtained in the isotropic phase above the phase transition with those in the nematic phase of a stretched elastomer [15]. It was pointed out that, in addition to the usual wide angle halo associated with short range positional order in both phases, there is a second halo which is associated with the elastomer, since it vanished upon swelling in a low molecular weight material. A similar cooling procedure was used to show that the mean value of the angle between the molecular longitudinal axis of the mesogenic sidechains and the director of the nematic phase is strongly reduced as one reduces the temperature indicating an increasing degree of order with decreasing temperature in the nematic phase [2].
2.2 The CholestericIsotropic Phase Transition While the first paper on liquid crystalline elastomers [30] already reports the detection of a cholesteric-isotropic transition using differential calorimetry and polarizing microscopy, comparatively little work has been done to characterize thy physical properties in the vicinity of this phase transition (compare, however, also the discussion of electromechanical effects in the next section) [9, 30, 311. Combined liquid crystalline elastomers have been synthesized and various of these materials show a cholesteric-isotropic transition using X-ray scattering, polarizing microscopy and differential scanning calorimetry [3 11. Dynamic mechanical investigations have been carried
2
out [9] for a number of main-chain liquid crystalline elastomers. The measurements of the shear storage and the shear modulus both reveal an increase in the vicinity of the isotropic-cholesteric phase transition.
2.3 The Smectic AIsotropic Phase Transition A SmA-I phase transition has been identified in a large number of liquid crystalline elastomers using differential scanning calorimetry, polarizing microscopy, and X-ray diffraction, both for side-chain liquid crystalline elastomers [30, 321 and for mainchainkombined liquid crystalline elastomers [31, 331. However, until very recently no detailed investigations of physical properties in the vicinity of the SmA-I transition have been reported 1341. Birefringence measurements under a static external mechanical stress were carried out above the SmA-I phase transition for a side-chain liquid crystalline elastomer for a range of values of the external stress [34]. The plot of the inverse birefringence versus temperature (Fig. 21) shows two different regimes
Phase Transitions
289
connected by a smooth cross-over. Using a simple model, it is suggested that these two regimes correspond to nematic fluctuations (at higher temperatures further above the transition) and SmA fluctuations (close to the SmA-I transition), respectively. It appears that such a cross-over behavior has never been seen before for low molecular weight and polymeric materials.
2.4 The Discotic-Isotropic Phase Transition While there is already a substantial body of work on phase transitions in liquid crystalline elastomers composed of rod-like molecules, much less has been done on columnar phases typically composed of disk-like molecules. In all cases published to date [35,36] discotic-isotropic phase transitions were found at higher temperatures followed by a glass transition at lower temperatures. Applying the technique developed by Kiipfer and Finkelmann [18], Disch et al. [36] were able to produce discotic monodomains as elucidated by X-ray diffraction and, in particular, by measuring the thermal expansion across the transition for a monodomain and by comparing these results with those obtained for a polydomain. As expected from the change in the degree of macroscopic order (compare the detailed discussion for the isotropic-nematic transition for poly- versus monodomains), one finds a drastic change in length in the vicinity of the isotropic-discotic transition for the monodomain while there is a small and very smooth change for the polydomain (Fig. 22).
2 151
.*
..
h ._ V
z x
Figure 21. The inverse of the birefringence plotted as function of the temperature for different values of the true stress (reproduced with permission from
W1).
290
V
1.3
Physical Properties of Liquid Crystalline Elastomers 1
1.25
..
1.2 1.15 Lli/Lim 1.1
1.3
..
1.05
1.2 RIIi 1.1
c
b
1
0 5
0.8
0.85
0.95
0.9
1
1 1.05
Reduced temperature, Jred
(a)
1.15 1.2
i
1-
(b)
1.4
....*.......
.**
.n...-*
Reduced temperature,
Tred
2.5 Other Phase Transitions in Liquid Crystalline Elastomers In addition to the phase transitions described so far, a number of other phase transitions involving liquid crystalline phases has been reported. These include for side-chain liquid crystalline elastomers: SmC-I [32], N-SmA [6,30], N-SmF [32], SmA-SmC [32], SmA-SmB [32] and SmC-SmF [32]. Furthermore, a phase transition between a SmA phase and a re-entrant
-
Figure 22. Thermal expansion of discotic networks: (a) LSCE; (b) polydomain permission from [36]).
nematic phase has been described [6]. The identification of these phase transitions was performed using differential scanning calorimetry, observations in the polarizing microscope, and X-ray scattering. The elongation was measured as a function of temperature for different loads and it was found that there appears to be a change in slope near the SmA-reentrant nematic phase transition [6]. In the field of main-chain and combined liquid crystalline elastomers the phase transitions observed include: cholesteric-SmC* [9,37], N-SmA [9,38], SmA-
SmC[9,38],N-SmC[39],SmA-SmB[38],
3 Macroscopic Properties
cholesteric-SmA [ 3 1] and cholestericSmC* [31]. The detection of all these phase transitions was based on polarizing microscopy, X-ray diffraction, and differential scanning calorimetry. In addition dynamic mechanical measurements were used [9] to show that the loss factor peaks below the cholesteric-SmC* and the SmA-SmC phase transition, respectively. Quite recently electromechanical and electro-optic effects have been studied in some detail for the SmA-SmC* transition in sidechain LCE [40]. The authors account for their observation using a Landau model, which contains an additional elastic energy associated with the tilt, when compared to the description of low molecular weight materi a1s. Generalizing the concept of side-chain liquid crystalline elastomers from thermotropic to lyotropic materials, various phase transitions have been observed in lyotropic liquid crystalline elastomers. A hexagonal to isotropic phase transition was described using polarizing microscopy and X-ray diffraction; in addition stress-strain measurements in the hexagonal phase demonstrated the rubber-like behavior of these lyotropic materials [41]. A material showing an isotropic-lamellar (La) phase transition in an elastomer has been examined [42]. The phase transition was identified with polarizing microscopy and NMR was used to analyze the change in the lineshape when one is passing from the isotropic into the lamellar phase in this elastomer. In addition it was shown by X-ray diffraction that the liquid crystalline phase could be oriented by a uniaxial mechanical compression.
29 I
3 Macroscopic Properties 3.1
Nematic Phase
3.1.1 Mechanical Properties of Polydomains It has been shown, for a number of fixed values of the compression, that the length of a cylindrial sample under compression shrinks as one cools from the nematic-isotropic transition temperature towards the glass transition [ 5 ] . Conversely the diameter of these samples grows throughout the nematic temperature range. Stress-strain measurements in the nematic phase showed that above a certain threshold stress associated with the polydomain-monodomain transition, a linear stress-strain relation results [3]. Extrapolating these linear curves to zero stress one was able to extract the fictitious length 1, of a monodomain sample. When this length I, was plotted as a function of reduced temperature, it was found that lo increased in the nematic phase similarly to the order parameter. The elongation, A, was measured as a function of temperature for fixed load for various nematic phases as well as for a reentrant nematic phase in sidechain LCEs [6]. In the reentrant nematic phase the observed elongation A increased monotonically (by up to 50%) as the temperature was decreased from the smectic A-reentrant nematic phase transition [6]. X-ray diffraction measurements have been used to extract the orientation parameter ( P 2 )as a function of the extension ratio [43]. ( P 2 )was found to increase rapidly first and then to saturate for higher deformations. These studies were generalized by varying
292
V
Physical Properties of Liquid Crystalline Elastomers
the crosslinking density and the length of the flexible spacer [44]. It was found that the orientation parameter (P2)is almost the same for different spacerlengths. In addition it ws shown that (P2)decreases when the nematic-isotropic transition is approached from well inside the nematic phase. Small angle neutron scattering measurements of the anisotropy of the main chain conformation of a side-on sidechain LCE under a mechanical stress are described in the isotropic and in the nematic phase [45]. It is found that a weak anisotropy of the main chain conformation exists only in the nematic phase and for sufficiently high values of the elongation. Very recently the first synthesis of a main chain LCE has been described [46]. It turns out, as was to be expected due to the completely different location of the mesogenic units, that the stress strain properties of these novel materials are rather different from what is known from side chain LCEs. For example, the main chain LCEs require a stretching by a factor of about 5 to get macroscopic alignment of the director. Such an extension is not even an option for side chain LCEs, which typically rupture when elongated by a factor between about 2 and 3. Thus in many ways the main chain LCEs might resemble more closely a classical rubber than the sidechain LCEs. Laser induced phonon spectroscopy has been used to study the ultrasonic properties of nematic sidechain LCEs [47, 481. It has been shown that both the ultrasonic velocity and the attenuation show a strong anisotropy and it was concluded that the coupling between density changes and reorientationa1 motion of the sidechains takes place on a time scale of lop7- lop9s [47]. Dynamic measurements of the storage (G’) and the loss modulus (G”) in the nematic phase of a sidechain LCE have been described [7]. It was found that throughout
the nematic phase G’>G” and that both were increasing monotonically as the temperature was lowered towards the glass transition.
3.1.2 Mechanical Properties of Monodomains: Liquid Single Crystal Elastomers Memory effects in LCE have been investigated [49]. Small monodomain LCE films were prepared by aligning the nematic polymer phase in a magnetic field prior to crosslinking. These LCE films preserved their original director alignment as well as their degree of preferred orientation, even when they were held for over two weeks in the isotropic phase and then cooled back down into the nematic phase. The degree of alignment was measured by wide angle X-ray scattering and was found to show a similar temperature dependence as in earlier investigations by the same group [44]. The influence of the spacer length on the nature of the coupling between sidechains and the elastomeric network has been investigated [50]. It turns out that a lengthening of the spacer is accompanied by a sign change of the coupling between the polymeric backbone and the mesogenic units. In addition it emerges from these investigations that the magnitude of this coupling decreases with increasing spacerlength, a result which is .in agreement with one’s intuition. Strain-induced transitions of the director orientation in nematic LCEs were studied [51]. These abrupt transitions with a director reorientation of 90” [51] were found to have a characteristic threshold strain and show qualitative similarities to predictions from a model based on rubber elasticity [52]. However, it should be noticed that the extension ratio was varied by 0.05 at a time
3
and that the actually observed variation of the director angle for this step size was 65" and not 90", thus rather suggesting a steep, but continuous change. Clearly further more detailed experiments are necessary to clarify this point. Nematic monodomains have been prepared using a variant of the method introduced by Kiipfer and Finkelmann [lo, IS]: instead of keeping the stress constant in the second crosslinking step the strain was held constant in the second crosslinking step 1.531. As for the case studied by Kiipfer and Finkelmann, it is found that the monodomain character is retained after heating the sample into the isotropic phase upon return into the nematic phase. Stress strain relation has been measured for two classes of polysiloxane elastomers. Ultrasonic experiments using laser induced phonon spectroscopy have been performed in a nematic liquid single crystal elastomer [48]. The experiments reveal a dispersion step for the speed of sound and a strong anisotropy for the acoustic attenuation constant in the investigated frequency range (100 MHz - 1 GHz). These results are consistent with a description of LCEs using macroscopic dynamics [54-561 and reflect a coupling between elastic effects and the nematic order parameter as analyzed in detail previously 1481.
3.1.3 Linear and Nonlinear Optical Properties of Polydomains It has been shown that changing the length of the flexible spacer, m, by one unit can change the optical properties of a deformed nematic LCE qualitatively 1151. While for m = 3 the deformed LCE is optically biaxial and the mesogenic units are oriented on average perpendicularly to the axis of defor-
Macroscopic Properties
293
mation, for m = 4 an optically uniaxial configuration arises in which the director is oriented parallel to the axis of deformation. These optical investigations have been complemented by X-ray investigations leading to the same conclusions. Further X-ray investigations on a similar family of compounds also showed the change from uniaxial to biaxial behavior as the spacer length was changed from m = 4 to m = 3 [2]. IR-dichroism measurements of the CNstretching mode revealed that the order parameter S extracted from the absorption coefficients along and perpendicular to the long axis of the molecules has a similar temperature dependence as in low molecular weight materials [16].
3.1.4 Linear and Nonlinear Optical Properties of Monodomains The reorientation behavior has been studied, which results when a nematic monodomain LCE is exposed to a mechanical field perpendicular to the preferred direction of the monodomain 1181. For this configuration the dichroic ratio R =AII/A, of the CNabsorption band as well as the external mechanical stress, o,,are plotted versus the elongation, LIL,, perpendicular to the director axis of the unstrained sample in Fig. 23. While the dichroic ratio saturates for small and large values of the external stress with a nearly linear change for intermediate values of the mechanical force, the nematic order parameter, S, changes from a value of S = 0.53 in the unstrained configuration to a value of S=0.36 for high values of the external stress thus indicating that the reorientation process does not involve all parts of the monodomain LCE. This last point has been investigated further and monodomain LCEs crosslinked in
294
V
Physical Properties of Liquid Crystalline Elastomers
is drastically reduced from S =0.5 in the unstrained sample to S=0.3, this reduction is not observed for the sample crosslinked in the isotropic phase. From these observations one concludes that the state of order in the vicinity of the crosslinking points also affects the macroscopic behavior of the nematic monodomain LCE [lo], a feature that has been modeled using a Landau approach and the concept of a frozen-in field [22].
****** 0
A
A "AA
I I
h6O0C
I I I I I
1.0
1.2
3.1.5 Effects of Electric and Magnetic Fields
,
1.1 1.6 Elongation, L/Lo
10 1.8
Figure 23. Dicroic ratio R (*) and nominal external stress o,(A) as function of the elongation LIL, of a nematic LSCE (reproduced with permission from [181).
the nematic and in the isotropic phase, respectively, were examined [ lo]. X-ray investigations performed for both types of samples show that the monodomains go for intermediate values of the elongation through a state with domains for which the director is either ordered by 0 = 4 5 " or 0=-45" to the stress axis. While for the nematic LCE crosslinked in the isotropic phase a complete reorientation is achieved, this is not the case for the sample crosslinked in the nematic phase for which the X-ray pattern shows, even for an elongation A = 1.6, clearly four distinct maxima. This qualitative difference in behavior between samples crosslinked in the isotropic and in the nematic phase is also clearly brought out by measurements of the dichroic ratio. While the order parameter S in the reoriented sample crosslinked in the nematic phase
Deformations of nematic elastomers in electric fields have been described [44, 57, 581. Reversible shape variations of a nematic sidechain LCE swollen with low molecular weight (LMW) materials in an electric field were reported, provided the elastomer is freely suspended [57]. If it is compressed between glass slides, no field effect is observed. Shape changes in mainchain LCEs swollen with a LMW material have been investigated in an electric field [44]. Similarly, reversible shape changes were observed for freely suspended LCEs on a time scale of one second and no shape changes were obtained for the unswollen elastomers [57]. More recently further experiments of the same type were done on swollen sidechain LCEs [58]. In addition it was checked that there was no response to an electric field if both, LMW material and LCE, were in their isotropic phase [58]. No reports on deformations and shape changes of nematic elastomers in external magnetic fields appear to exist yet.
3.1.6 Diffusion Effects It has been shown that the diffusion coefficient for salicyclic acid in a nematic side-
3 Macroscopic Properties
chain LCE can be controlled by varying the mass fraction of the mesogens thus demonstrating the potential of LCE for pharmaceutical applications [59].
3.1.7 Macroscopic Equations Macroscopic properties of nematic elastomers have been discussed [56, 601. De Gennes focused on the static properties, emphasizing especially the importance of coupling terms associated with relative rotations between the network and the director field [60]. The electrohydrodynamics of nematic elastomers has been considered generalizing earlier work by the same authors [54,55] on the macroscopic properties of nematic sidechain polymers [56]. The static considerations of earlier work [60] were extended to incorporate electric effects; in addition a systematic overview of all terms necessary for linear irreversible thermodynamics was given [56].
3.1.8
Rubber Elasticity
An extension of rubber elasticity (i.e. of the description of large, static and incompressible deformations) to nematic elastomers has been given in a large number of papers [52, 61-66]. Abrupt transitions between different orientations of the director under external mechanical stress have been predicted in a model without spatial nonuniformities in the strain field [52,63]. The effect of electric fields on rubber elasticity of nematics has been incorporated [65]. Finally the approach of rubber elasticity was also applied recently to smectic A [67] and to smectic C* [68] elastomers. Comparisons with experiments on smectic elastomers do not appear to exist at this time. Recently a rather detailed review of the model of an-
295
isotropic rubber elasticity has been given by Warner and Terentjev [69].
3.1.9 X-ray Investigations It has been shown using X-ray fiber patterns, that a weakly crosslinked sidechain LCE can be oriented by an external mechanical stress with the polymer chains oriented parallel to the stress axis and the director associated with the mesogenic units perpendicular to the stress axis [38]. X-ray investigations also have been done for nematic phases of combined elastomers [33]. X-ray investigations on strongly stretched and rather well-oriented samples using a two-dimensional detector have been reported for a polysiloxane series of LCEs [70].
3.2
Cholesteric Phase
3.2.1 Mechanical Properties Static mechanical measurements to evaluate the stress-strain relationship in cholesteric sidechain LCEs have been described [71, 721. In [72] it has been found, for example, that for 0.94 < A < 1 (where A = 1 corresponds to the unloaded sample), the nominal stress onis nearly zero as the polydomain structure must be converted first into a monodomain structure. For deformations A < 0.94, the nominal stress increases steeply. Similar results have also been reported elsewhere [7 11. The nominal mechanical stress as a function of temperature for fixed compression has also been studied for cholesteric sidechain elastomers [71]. It turns out that the thermoelastic behavior is rather similar as that of the corresponding nematic LCE [2, 51. The temperature dependence of the reciprocal length 1/Lmonof a macroscopically
296
V
Physical Properties of Liquid Crystalline Elastomers
uniform, oriented cholesteric LCE has been measured. Using independent optical measurements to determine the orientational order parameter S , it has been shown [73], that S is directly proportional to l/L,,,, similar to what has been found for nematic LCE [4], thus demonstrating that the deformation can also be used as an order parameter in ordered samples of cholesteric LCEs. The mechanical properties of polydomains of a main-chain cholesteric LCE, namely of crosslinked hydroxypropyl cellulose, have been studied 1741. The orientation parameter (P2), as extracted from X-ray measurements at fixed temperature, first increases as a function of extension ratio and then saturates at higher strains. As a function of temperature for fixed extension ratio, (P2)decreases monotonically. It has been shown for cholesteric combined LCE that polydomain samples can be untwisted by a sufficiently large external mechanical force leading to a strain of 150% [75].This has been checked by using X-ray fiber patterns. The same type of investigations has also been applied to other combined cholesteric LCE [31]. Dynamic mechanical measurements of the storage modulus G’ and the loss modulus G” in polydomain samples of cholesteric LCE have been described 191. It turns out that G‘> G” throughout the whole cholesteric range and that both G’ and G” increase when approaching the cholestericisotropic and cholesteric-chiral SmC* transition, respectively, thus indicating the presence of substantial pretransitional effects at both phase transitions.
3.2.2 Electromechanical Properties The electromechanical response in a cholesteric sidechain LCE under an external me-
chanical force has been studied 171,731. For sufficiently high values of the mechanical compression, needed to achieve a cholesteric monodomain with the helix axis parallel to the axis of the cylindrical sample, a linear electromechanical effect is found. When the temperature-dependent electric response is plotted for sufficiently high, but fixed compression, the resulting electric signal closely resembles the plot of the orientational order parameter as a function of temperature (Fig. 24). In addition to the electromechanical observations described so far, which have been confirmed [73], the dependence on the cholesteric pitch, p , has been studied [73]. It is found that the electromechanical response coefficient is directly proportional to l l p and vanishes for a racemic sample. Static measurements of the change in sample thickness, ALlL, (with Lo the thickness in the mechanically loaded state, but without external electric field), as a function of an applied electric field (parallel to the helical axis of a cholesteric monodomain sample) have been carried out [72]. To eliminate effects that are quadratic in the electric
I
01, 290
,
.
,
,
295
,
,
,
,
,
,
300
, ’ 305
,
, ? ,n
310
Temperature, T/K
Figure 24. Electromechanical response, U , vs. temperature for a compression of 25% of a cholesteric elastomer.
3
1.2
Electric field strength, E(Vpm-’)
field such as the attractive force between the plates bounding the sample and electrostrictive effects, the same experiments have also been done for the corresponding racemic compound. When the change in thickness of the cholesteric elastomer minus the change in thickness of the racemic elastomer ALIL, is plotted as a function of the electric field strength, a linear relationship results above a threshold field, Eth(Fig. 25). Alternating current measurements were used to investigate the electromechanical properties in combined photo-crosslinked cholesteric LCEs. In a compound showing both, a cholesteric and a chiral SmC* phase, the electromechanical response in the cholesteric phase was considerably higher. Various electromechanical effects in cholesteric and chiral SmC* phases were presented including a discussion of how to distinguish piezoelectricity from flexoelectricity and electrostriction [77, 781. It was examined for cholesteric structures in general for which structures one can get longitudinal piezoelectricity [79].
Macroscopic Properties
297
Figure 25. Relative fitted change of sample thickness of a cholesteric elastomer minus the relative fitted change of sample thickness of the racemic elastomer ALIL, as function of the electric field strength E (reproduced with permission from [72]).
3.3 Smectic A and Smectic C Phases In contrast to the extensive experimental investigations of nematic, cholesteric and chiral SmC* phases, comparatively little work has been done on the characterization of the physical properties of SmA and SmC elastomers. The elongation, A, has been measured as a function of temperature for constant external load for a number of different loads in SmA sidechain elastomers [6]. It was found that ilincreases monotonically as a function of temperature for a material, which has a SmA-I transition. In addition it was shown that the elasticity modulus, E, decreases monotonically with temperature. X-ray investigations on SmA phases in sidechain LCE have been performed [70]. It was found that for the family of compounds studied, the orientation of the mesogenic groups was always perpendicular to the direction of stretching. Second harmonic generation of a poled SmA elastomer has been demonstrated [80]. A uniform director orientation was achieved by a simple mechanical deformation of the elastomer, poling was done in an electric field and the NLO active chromophores (ni-
298
V
Physical Properties of Liquid Crystalline Elastomers
trogroups) were bounded covalently to the network. The layer spacing in the SmA and SmC phases of combined LCE has been investigated using X-ray diffraction [ 5 3 ] .The storage modulus G’ and the loss modulus G” have been studied in the SmA and SmC phases of combined elastomers and it was found that G’> G” throughout the whole temperature range of these two smectic phases [9]. The diffusion of gases through films of SmA elastomers has been analyzed [8 1,821. It was shown experimentally that the logarithm of the diffusion coefficient depends linearly on inverse temperature throughout all SmA phases and that the transport behavior is stress-dependent [8 ll. The diffusion of hydrocarbon gases through SmA elastomers was studied in detail [82]. It turns out that these LCE materials are suitable for the separation of gases due to their high selectivity and wide range of permeabilities. Physical properties of the higher ordered smectic phases SmF, SmI, SmB, SmH etc. have not been investigated as yet. Various physical properties of the layered La-phase in a lyotropic LCE have been studied [42]. Using NMR and X-ray investigations it was demonstrated that a lamellar phase can be oriented macroscopically by applying a uniaxial compression to the disordered swollen network.
uniaxial mechanical stress is applied to a swollen C* sample after the first crosslinking step, an X-ray pattern, which shows four distinct maxima in the small angle regime results (Fig. 26), indicating a layer orientation of the layer normal under an angle fa to the stress axis. In addition, the azimuthal wide angle intensities indicate an average alignment of the long molecular axis of the mesogens parallel to the stress axis. Thus, while the helical structure is untwisted, a monodomain is not obtained when applying only a uniaxial external stress. To come closer to achieving this goal a second deformation was carried out under an angle compared to the previous one in order to lift the remaining azimuthal degeneracy. As it turns out (Fig. 27) from analyzing the resulting X-ray pattern, a sample that is oriented to more than 95% (and thus close to a monodomain) results. This technique has been discussed and studied in more detail [84], where it is also emphasized that the samples resulting after the second crosslinking have macroscopically C,-symmetry and thus possess dipolar order giving rise to pyroelectricity and piezoelectricity. The mechanical behavior of combined SmC* polydomain samples was investigat-
3.4 Smectic C* Phase 3.4.1 Mechanical Properties of Polydomains X-ray investigations were used to study the orientation behavior of a sidechain SmC* elastomer under mechanical stress [83]. If a
Figure 26. X-ray pattern of a SmC*-elastomer after uniaxial deformation in the swollen state (reproduced with permission from [83]).
3 Macroscopic Properties
299
axis. This line of investigations was extended shortly thereafter to other compounds with similar results [37]. Dynamic mechanical measurements of the storage, G’, and the loss modulus, G”, in combined LCE [9] show, that both, G’ and G”, fall monotonically as a function of temperature throughout the entire range of existence of the C*-phase and that G’> G” is always valid. Figure 27. X-ray pattern of a SmC*-elastomer after second uniaxial deformation (reproduced with permission from [83]).
3.4.2 Nonlinear Optical Properties
ed for large mechanical deformations [75]. To detect the effect of the external stress on the helix, X-ray fiber patterns were used. It was found that for moderate values of the strain which was necessary to orient the fiber, 250%, the smectic layers were perpendicular to the fiber axis and the helix was still present. At very large values of the strain, >400%, however, X-ray data showed that the mesogenic groups were oriented on average parallel to the fiber axis and that thus the helix was unwound and that the layers were tilted with respect to the fiber
Using the technique described earlier [83] to get samples that are nearly monodomain, second harmonic generation has been investigated using a Q-switched Nd-YAG laser [ 8 5 ] . Since the helix is unwound in these samples and since they are close to a monodomain, they are macroscopically of C2symmetry and thus suitable for second harmonic generation. In Fig. 28 the intensity of the second harmonic signal is plotted as a function of the angle around the direction in the plane of the film, which is perpendicular to the first deformation direction. The results of the polarization dependent measure-
,?JT\ 0
A
*.
& \%
~
I
,
k
Figure 28. Marker fringe diagram for an annealed (A) and a non-annealed SmC*-
300
V
Physical Properties of Liquid Crystalline Elastomers
ments are consistent with overall C,-symmetry. The quality of the orientation was checked independently by X-ray measurements revealing the high degree of orientation. In contrast to other polymer networks, including centrosymmetric LCE such as nematic LCE, which require external poling fields, the unwound, uniformly oriented SmC* LCE has built-in noncentrosymmetry and thus possesses an intrinsic macroscopic electric polarization.
Making use of the technique by Semmler and Finkelmann [83, 841 to obtain a highly oriented SmC*-LCE film with untwisted helix and macroscopic C,-symmetry, Eckert et al. [87] investigated piezoelectricity and inverse piezoelectricity. They find that in these highly oriented and untwisted SmC* films a linear behavior results for the charge as a function of the applied force as well as for the displacement as a function of applied AC voltage.
3.4.3 Electromechanical Properties of Polydomains
3.5 Discotic Phases and Hexagonal Lyotropic Phases
Alternating current measurements were used to investigate the piezoelectric effect in combined photocrosslinked SmC* LCEs [76]. A linear relationship was obtained for the measured piezo voltage as a function of the mechanical deformation applied. In addition it was found that the measured piezo voltage is a monotonically decreasing function of temperature for a compound showing an I-SmC* phase transition. Piezoelectric and pyroelectric properties of polydomain sidechain LCEs in the chiral SmC* phase have been investigated [86]. The pyroelectric response was generated by heat pulses from a YAG laser. While the extracted saturation polarization falls monotonically as a function of temperature throughout the SmC* phase, the pyroelectric coefficient, y=aP/aT, goes through a maximum several degrees below the SmC*-I phase transition temperature. An AC piezoelectric effect has been measured throughout the entire temperature interval of the SmC* phase as well for the unoriented sample. An additional DC poling field led to an increase in the observed piezoresponse, probably due to a better degree of orientation of the macroscopic polarization throughout the sample.
There are only two papers describing the physical properties of discotic LCEs [35, 361. X-ray diffraction showed that a narrower azimuthal intensity distribution can be obtained when a second crosslinking step is done under an external load reflecting a macroscopic alignment parallel to the direction of the external stress [35]. X-ray diffraction experiments on monodomains (LSCE) show the occurrence of sharp maxima in the azimuthal distribution. Thermal expansion has also been studied for polydomain LCEs and for LSCEs in [361. It is found that well inside the discotic phase, that is sufficiently far below the discotic-isotropic transiton, the length parallel to the stress axis of both, poly- and monodomain samples, increases approximately linearly with increasing temperature as expected from the isobaric expansion of the network. In addition to the stress-strain measurements already mentioned, the hexagonal lattice constant was measured in a lyotropic LCE as a function of the water concentration [41]. Furthermore it was shown, using X-ray scattering, that a hexagonal phase can be oriented parallel to the preferred direction by applying an external mechanical field.
4
4 References [I] R. A. M. Hikmet, J. Lub, Progr. Polym. Sci. 1996,21, 1165- 1209. [2] W. Gleim, H. Finkelmann, Makromol. Chem. 1987,188, 1489- 1500. [3] J. Schatzle, W. Kaufhold, H. Finkelmann, Makromol. Chem. 1989,190,3269-3284; 1991,192, 1235- 1236. [4] W. Kaufhold, H. Finkelmann, H. R. Brand, Makromol. Chem. 1991,192, 2555-2579. [ S ] K. Hammerschmidt, H. Finkelmann, Makromol. Chem. 1980,190, 1089- 1101. [6] R. V. Talroze, T. I. Gubina, V. P. Shibaev, N. A. Plate, V. I. Dakin, N. A. Shmakova, F. F. Sukhov, Makromol. Chem., Rapid Comrriun. 1990, 11,67-71. [7] W. Oppermann, K. Braatz, H. Finkelmann, W. Gleim, H. J. Kock, G. Rehage, Rheol. Acfa1982, 21,423-426. [8] R. Sigel, W. Stille, G. Strobl, R. Lehnert, Macromolecules 1993, 26, 4226-4233. [9] T. Pakula, R. Zentel, Makromol. Chem. 1991, 192,2401-2410. [lo] J. Kupfer, H. Finkelmann, Makromol. Chem. 1994,195, 1353-1367. [ 111 J. Kupfer, E. Nishikawa, H. Finkelmann, Polym. Adv. Technol. 1994, 5, 110- 1 15. [I21 1. Kundler, H. Finkelmann, Macromol. Chem., Rapid Commun. 1995, 16,679-686. [13] J. Weilepp, H. R. Brand, Europhys. Lett. 1996, 34,495-500. [14] G. C. Verwey, M. Warner, E. M. Terentjev, J . Phys. 11 (France) 1996, 6, 1273- 1290. [IS] H. Finkelmann, H. J. Kock, W. Gleirn, G. Rehage, Macromol. Chem., Rapid Commun. 1984, 5, 287-293. [I61 J. Schatzle, H. Finkelmann, Mol. Cryst. Liq. Cryst. 1987, 142, 8 5 - 100. [17] P. G. de Gennes, C. R. Acad. Sci. (Paris) 1975, B281, 101- 103. [18] J. Kupfer, H. Finkelmann, Macromol. Chem., Rapid Commun. 1991,12,717-726. [19] E. Gramsbergen, L. Longa, W. de Jeu, Phys. Rep. 1986,135, 195-257. [20] S. Disch, C. Schmidt, H. Finkelmann, Macromol. Chem., Rapid Commun. 1994,15,303-310. (211 S. Disch, C. Schmidt, H. Finkelmann, Polymer Enzyclopedia, CRC Press, Boca Raton, FL, 1995. [22] H. R. Brand, K. Kawasaki, Macromol. Chem., Rapid Commun. 1994,15,251-257. [23] P. G. de Gennes in Polymer Liquid Crystals (Eds.: A. Ciferri, W. R. Krigbaum, R. B. Meyer), Academic Press, NY 1982, pp. 115- 131. [24] M. Warner, P. Ceiling, T. A. Vilgis, .I. Chem. Phys. 1988,88,4008-4013. [25] H. R. Brand, Macromol. Chem., Rapid Commun. 1989, 10,57-61,317.
References
30 1
[26] A. ten Bosch, L. Varichon, Makromol. Theory Simul. 1994,3, 533 - 542. [27] H. R. Brand, K. Kawasaki, J. Phys. I1 (France) 1992,2, 1789- 1795. [28] H. R. Brand, K. Kawasaki, J. Phys. I1 (Paris) 1994,4,543-548. 1291 K. Kawasaki, H. R. Brand, Physica A 1994,208, 407-422. [30] H. Finkelmann, H.-J. Kock, G. Rehage, Macrw mol. Chem., Rapid Commun. 1981,2,317-322. [31] R. Zentel, G. Reckert, S . Bualek, H. Kapitza, Makromol. Chem. 1989, 190,2869-2884. [32] Y. S. Freidzon, R. V. Talroze, N. I. Boiko, S . G. Kostromin, V. P. Shibaev, N. A. Plate, Liq. Cryst. 1988,3, 127-132. [33] S. Bualek, J. Meyer, G. F. Schmidt, R. Zentel, Mol. Cryst. Liq. Cryst. 1988, 155, 47-56. [341 M. Olbrich, H. R. Brand, H. Finkelmann, K. Kawasaki, Europhys. Lett. 1995, 31, 281286. [35J H. Bengs, H. Finkelmann, J. Kupfer, H. Ringsdorf, P. Schuhmacher, Macromol. Chem., Rapid Commun. 1993,14,445-450. [36] S. Disch, H. Finkelmann, H. Ringsdorf, P. Schuhmacher, Macromolecules 1995,28,24242428. [37] R. Zentel, Angew. Chem. Adv. Muter. 1989, 101, 1437- 1445. [38] R. Zentel, G. Reckert, Makromol. Chem. 1986, 187, 1915-1926. 1391 S. Bualek, R. Zentel, Makromol. Chem. 1988, 189, 791 -796. [40] M. Brehmer, R. Zentel, F. GieRelmann, R. Germer, P. Zugenmaier, Liq. Cryst. 1996, 21, 589596. [41] R. Loffler, H. Finkelmann, Macromol. Chem., Rapid Commun. 1990,11,321-328. [42] P. Fischer, C. Schmidt, H. Finkelmann, Macromol. Chem., Rapid Commun. 1995, 16, 435447. (431 F. J. Davis, G. R. Mitchell, Polymer Commun. 1987,28,8-11. [44] N. R. Barnes, F. J. Davis, G. R. Mitchell, Mol. Cryst. Liq. Cryst. 1989, 168, 13-25. [45] H. Finkelmann, H. R. Brand, Trends Polym. Sci. 1994,2,222-226. [46] G . H. F. Bergmann, H. Finkelmann, V. Percec, M. Zhao, Macromol. Chem., Rapid Commun. 1997,18,353-360. [47] F. W. Deeg, K. Diercksen, G. Schwalb, C. Brauchle, H. Reinecke, Phys. Rev. B 1991, 44, 2830-2833. [48] F. W. Deeg, K. Diercksen, C. Brauchle, Ber. Bunsenges. Phys. Chem. 1993, 97, 1312- 1315. [49] C. H. Legge, F. J. Davis, G. R. Mitchel1,J. Phys. 11 (France) 1991, I , 1253-1261. [SO] G. R. Mitchell, M. Coulter, F. J. Davis, W. Guo, J. Phys. 11 (France) 1992, 2, 1 121 - 1 132. 1511 G. R. Mitchell, F. J. Davis, W. Guo, Phys. Rev. Lett. 1993, 71, 2947-2950.
302
V
Physical Properties of Liquid Crystalline Elastomers
1521 R. Bladon, M. Warner, E. M. Terentjev, Phys. Rev. E 1993,47, R3838-R3840. [53] C. J. Twomey, T. N. Blanton, K. L. Marshall, S. H. Chen. S. D. Jacobs, Liq. Cryst. 1995, 19, 339 - 344. 1541 H. Pleiner, H. R. Brand, Mol. Cryst. Liq. Cryst. 1991,199,407-418. [55] H. Pleiner, H. R. Brand, Macromolecules 1992, 25,895-901. [56] H. R. Brand, H. Pleiner, Physica A 1994, 208, 359-372. 1571 R. Zentel, Liq. Cryst. 1986, I, 589-592. [58] R. Kishi, Y. Suzuki, H. Ichijo, 0. Hirasa, Chem. Lett. 1994, 12, 2257 -2260. 1591 H. Loth, A. Euschen, Macromol. Chem., Rapid Commun. 1988, 9, 35-38. [60] P. G. de Gennes in Liquid crystals of one- and two-dimensional order (Eds.: W. Helfrich, G. Heppke), Springer, NY 1980, pp. 231 -237. [61] P. Bladon, M. Warner, Macromolecules 1993, 26, 1078- 1085. [62] P. Bladon, M. Warner, E. M. Terentjev, Macromolecules 1994,27, 7067 -7075. 1631 R. Bladon, E. M. Terentjev, M. Warner, J. Phys. //(France) 1994,4, 75-91. [64] E. M. Terentjev, Europhys. Lett. 1993, 23, 27-32. 1651 E. M. Terentjev, M. Warner, P. Bladon, J. Phys. I1 (France) 1994,4, 667-676. 1661 M. Warner, R. Bladon, E. M. Terentjev, J. Phys. I1 (France) 1994,4, 93- 102. 1671 E. M. Terentjev, M. Warner, J. Phys. I1 (France) 1994,4, 849-864. 1681 E. M. Terentjev, M. Warner, J. Phys. ZI(France) 1994,4, 111- 126. I691 M. Warner, E. M. Terentjev, Progl: Polym. Sci. 1996,21,853-891. [70] C. Degert, P. Davidson, S. Megtert, D. Petermann, M. Mauzac, Liq. Cryst. 1992, 12, 779798.
[7 I] W. Meier, H. Finkelmann, Macromol. Chem., Rapid Commun. 1990,II, 599-605. [72] H. Hirschmann, W. Meier, H. Finkelmann, Macromol. Chem., Rapid Commun. 1992, 13, 385 394. 1731 W. Meier, H. Finkelmann, Macromolecules 1993,26, 1811-1817. 1741 G. R. Mitchell, W. Guo, F. J. Davis, Polymer 1992,33,68-74. 17.51 R. Zentel, Liq. Cryst. 1988, 3, 53 1-536. 1761 S. U. Vallerien, F. Kremer, E. W. Fischer, H. Kapitza, R. Zentel, H. Poths, Macromol. Chem., Rapid Commun. 1990, 11,593-598. 1771 H. R. Brand, Macromol. Chem., Rapid Commun. 1989,10,441-445. 1781 H. R. Brand, H. Pleiner, Macromol. Chern., Rapid Commun. 1990,II, 607-612. [79] H. Pleiner, H. R. Brand, J. Phys. //(Paris) 1993, 3, 1397-1409. [80] H. Hirschmann, D. Velasco, H. Reinecke, H. Finkelmann, J. Phys. II(France) 1991, I, 559-570. [81] H. Modler, H. Finkelmann, Be% Bunsenges. Phys. Chem. 1990, 94, 836-856. 1821 H. Reinecke, H. Finkelmann, Makromol. Chem. 1992,193,2945-2960. 1831 K. Semmler, H. Finkelmann, Polym. Adv. Tech. 1994,5,232-235. 1841 K. Semmler, H. Finkelmann, Makromol. Chem. 1995,196,3197-3208. 1851 I. BennC, K. Semmler, H. Finkelmann, Macromol. Chem., Rapid Commun. 1994,15,295-302. 1861 M. Mauzac, H.-T. Nguyen, F. G. Tournilhac, S. V. Yablonsky, Chem. Phys. Lett. 1995, 240, 461 -466. 1871 T. Eckert, H. Finkelmann, M. Keck, W. Lehmann, F. Kremer, Macromol. Chem., Rapid Comrnun. 1996,17, 767-773.
Part 3: Amphiphilic Liquid Crystals
Handbook ofLiquid Crystals D. Demus, J. Goodby, G. W. Gray, H.-W. Spiess, V. Vil Copyright 0WILEY-VCH Verlag GmbH. 1998
Chapter VI Amphotropic Liquid Crystals Dirk Blunk, Klaus Praefcke and Volkmar Vill
1 Introduction, Remarks on History Numerous organic compounds exhibit liquid crystalline properties in the pure state. In a defined temperature range between the melting and the clearing temperature the order parameter of such a system is between that of true crystals and that of isotropic liquids. This is what one usually describes with the term thermotropic liquid crystal. But temperature is not the only term which may be variable. The intrinsic property of liquid crystals having an order parameter between those of real crystals and liquids includes the possibility of other ways of creating that interesting ‘fourth state of matter’ than just by changing the temperature. Another possible way of generating this intermediate state is by the addition of solvents in appropriate amounts to convenient compounds leading to lyotropic liquid crystals. Under suitable conditions, especially in well defined concentration ranges, the molecules tend to organize themselves forming aggregates of various well defined geometries. These supramolecular assemblies behave as mesogenic units forming lyotropic liquid crystalline phases. In this type of liquid crystals the concentration range in which the system is liquid crystalline can be compared with the temperature range between the melting- and the clearing point of thermotropic liquid crystals; however, it must be mentioned here that
in case of lyotropic liquid crystals the temperature is an additional degree of freedom. In contrast to the technologically extremely important thermotropic mesogens known for about 110 years and quickly growing in number, lyotropic analogues are as yet poorly studied and are much more complicated as their mesophase units do not consist of single molecules, but are made up of more or less flexible and variable aggregates of molecules in (so far mostly aqueous) solution. Hitherto, most solvents used to form lyotropic liquid crystals have been polar in character. By far the most conventional (and also from the biological point of view certainly the most important) solvent is water. However, other solvents can be used ranging from methanol at the polar end of the scale up to long chained alkanes on the other, apolar side of this classification of solvents. Chemical compounds are said to act amphotropically if they give rise to both kinds of liquid crystal formation. Such particular compounds show thermotropic liquid crystalline behavior in their pure state on heating or the formation of lyotropic mesophases on the addition of a further component, mostly of an inorganic or organic solvent in certain amounts. Nearly all the compounds of interest for this chapter, beside their conventional form anisotropy possess general amphiphilic behavior. The hydrophilic (polar) part of a molecule, for example a carboxyl group or
306
VI
Amphotropic Liquid Crystals
a multiol substructure, is well segregated from the lipophilic (apolar) moiety. As will be discussed, three main factors of the molecular architecture have a significant influence on the thermomesomorphic properties of such compounds: the ratio of the space demands of the hydrophilic and the hydrophobic part of a molecule, - the sterical arrangement of the substituents at the hydrophilic headgroup, and - geometrical factors such as molecular shape or symmetry.
-
Beside this, in the recent past some investigations on amphotropic materials of nonamphiphilic structure were made as well (see ‘Further Molecular Architectures’ in this Chapter). Additionally, it is somewhat difficult to draw a sharp separation line between amphiphilic and nonamphiphilic compounds: the ability to act as an amphiphile depends mainly on the size and the polarity of the hydrophilic and hydrophobic parts of a molecule. This ‘internal equilibrium’ can be varied in a broad range. Thus, in this chapter, although mainly directed toward mesogens of the amphiphilic type, some apolar mesogens will be discussed as well. Due to their importance for life sciences, for example the formation of biological membranes, and their technical importance as surfactants and soaps, potential lyotropic properties of amphiphilic compounds are much better investigated than other ones of different molecular types; the broad field of surfactant liquid crystals is discussed in Chapter VII of this volume. Some reviews focusing on amphotropy have been published; they take the development of liquid crystal research in this field into account [ 1-12]. With regards to the history of amphotropic liquid crystals it seems that in 1854 (about 35 years before 0. Lehmann and F. Reinitzer established liquid
crystallinity as the fourth state of matter) the first observation of a liquid crystalline state had been described [13, 141 though not understood. The magnesium salt of myristic acid (tetradecanoic acid) is described here as ‘... ein lockeres, aus mikroscopischen Nadeln bestehendes Pulver, welches uber 100°C zu einer durchsichtigen, aber nicht flussigen Masse wird; noch vor dem eigentlichen ZusammenflieBen wird es zersetzt’ [ 131.Although this magnesium salt is described here as ‘a loose powder’ this could be taken as a strong hint towards possible liquid crystalline properties. In spite of this strong hint, its technological significance (about 20 patents related to this salt have been issued since) and also the knowledge about the liquid crystallinity of other magnesium soaps, e.g., magnesium stearate, that is magnesium octadecanoate, nobody has yet investigated the thermotropic properties of magnesium myristate more thoroughly. Although Muller [ 15, 161 looked somewhat closer on this historically very interesting amphiphile by carrying out temperature dependent X-ray measurements, differential thermal analysis, and thermogravimetry, clear statements about its thermotropic properties are still missing. Looking back to 1854, no planned and detailed research on self organizing molecules was done; at that time, this phenomenon was not yet known. Nevertheless, a few scientists started working on this topic in those early times. For example, R. Virchow studied biological systems by means of polarizing microscopy; he observed and described myelin figures [ 171, a type of texture of lyotropic liquid crystals. The phenomena of birefringence of such myelin structures was investigated by Mettenheimer in 1857 [ 181. The book by Valentin [ 191 published about 140 years ago can be considered as an early milestone about investigations of materials of various kind by this optical method.
1
The connection between observed myelin figures and mesomorphism was investigated much later. For instance, Quinke described the formation of myelin figures of salts of fatty acids [20]. Further milestones in the field of lyotropic liquid crystals are investigations of binary systems of soaps and detergents with water [21] and work on the lyomesomorphism of polypeptides and nucleic acids [22]. In 1933 the first review on lyotropic liquid crystals appeared [23]. The author of this review investigated the amphotropic properties of various compounds. He suggested that in amphotropic materials raising the temperature or adding a solvent (water) lead to the same partial break up of the crystal lattice, however, a certain order is maintained. Furthermore, colloidal solutions showing partial supramolecular structures are also described here. But, due to different parameters of order these latter systems differ from lyotropic liquid crystals. How the lateral alkyl chains are organized in the liquid crystal state of matter was already a relevant subject of studies starting in the early thirties initiating discussions in which way alkyl chains are conformed in it (e.g., static- or liquid-like all-trans) [24,25]. Somewhat later, even biological systems were investigated from a liquid crystal research point of view (e.g., the tobacco-mosaic-virus) [26], see also Chap. VIII of this volume. The main progress of lyotropic liquid crystal research in these times is connected to works on soap/water mixtures [27, 281 and the investigation of thermotropic mesophases of soaps [29]. The amphotropic character of such compounds was also studied. In such systems, no continuous transitions were observed between thermotropic and lyotropic mesophases, but always biphasic regions could be seen [30, 311. Thereafter, the interest in understanding biological cell membranes inspired research-
Introduction, Remarks on History
307
ers to study watery lipid systems as models for such membranes [32]. Here, for the first time, the 45 nm reflex from X-ray investigations was identified to stem from disordered paraffinic chains. Up to this time, research work was characterized by compiling facts originating from investigations of different amphotropic liquid crystals, but a systematization of research in this field began only in the midforties [33-351. Helpful tools for this structurization of liquid crystal research were temperature dependent X-ray investigations [36] of natural and synthetic lipids, and the discovery that mesophases may be identified by their different textures appearing in the microscope using crossed polarizers [37]. In the decade starting in about 1957 systematic screening of the concentration and temperature dependency of the major lyotropic mesophases was done and models of the molecular arrangement in the different phases were developed [38-451 (e.g., the so-called ‘middle’ or ‘neat’ phases [38], the cholesteric phase of polypeptides and nucleopeptides [44]). During this work Luzzati found hitherto unnoticed phases (e.g., the rectangular columnar mesophase in the concentration range between the so-called ‘neat’ and ‘middle’ phases [39]). In the year 1958 he observed an optically isotropic appearing mesophase. Although, he could not clear up the structure of this interesting mesophase by X-ray experiments in detail, he predicted a cubic geometry of the lattice of this liquid crystal phase. The first investigation of the inverted hexagonal phasetype in the quaternary system lecithin, cephalin, phosphatidylinositol and water in 1962 [41] was the next milestone in the field of ‘new phases’. During the same period Ekwall studied the mesophases previously discovered by McBain, in order to clear up their molecu-
308
VI
Amphotropic Liquid Crystals
lar architecture. He investigated mainly ternary systems of soaps, alcohols and water [46], describing for instance the coexistence of two phases in one area of a mixture of sodium caprylate, decanol and water [471. The early sixties were golden times for the progress of investigation techniques in this field of science: the first NMR studies in the anisotropic media of lyotropic liquid crystals [48] were done, the introduction of the freeze fraction technique for the preparation of electron microscopy probes (e.g., of cell membranes [49]) and the separation of different lyomesophases by centrifugation [47] were introduced all between 1960 and 1962. Since the beginning of the 20th century the research on liquid crystals has been closely connected with molecules of amphiphilic character. As early as 1910 Vorlander described the sodium salt of diphenyl acetic acid as a liquid crystal [50]. Much later, in 1970, Demus characterized its mesophase as being of the columnar type, which was the first description of a columnar arrangement in a thermotropic liquid crystal [5 I]. By 1919 agaric acid was explicitly described to be liquid crystalline [52] and later has been characterized as SmA [53]. An important example of a thermotropic liquid crystal possessing a columnar mesophase is diisobutylsilandiol, which originally was reported to possess an unknown mesophase [54]; its columnar type was determined later [55]. The interest in mesogens based on carbohydrates started only recently, although since 1911 several reports on ‘double melting points’ of sugar based surfactants are documented in the literature. For the first review on this particular field see [4]. Fischer reported in 1911 the double melting behavior of hexadecylglucopyranoside [56], two years later Salway verified
Fischers observations and extended the studies to steroidal glycosides [57]. In 1938 hexadecylglucopyranoside was described explicitly as being liquid crystalline [58] by Noller and Hori extended this term to various other families of glycolipids [59, 601. At this time Chemical Abstracts refused to use the expression ‘liquid crystal’ in their abstract about this reference, which may be the reason why the work of Hori has not been recognized. For amphotropic liquid crystals it is obvious that the so-called surfactants (including for example soaps) are the best investigated, as documented in two reviews [61, 621. Therefore, an own chapter of this book is dedicated to that field of supramolecular chemistry (see Chapter VII of this volume). Since the lyotropic properties of this type of amphotropic liquid crystals are discussed there in detail, we focus now mainly on the thermotropic behavior of such compounds. Amphotropic behavior can be found for a large number of different chemical structures. Additional information is given in other chapters of this Handbook. Typical classes of amphotropic materials are for instance classical soaps (see lyotropics), transition metal soaps (see metallomesogens), viologens, quarternary amines and other ionic surfactants (see lyotropics), block copolymers (see polymer liquid crystals), cellulose derivatives (see cellulose liquid crystals) and partially fluorinated paraffines, diols, peptide surfactants, lecithins, lipids, alkylated sugars and inositols, naturally occurring glycosides and silanols, which are discussed in this chapter. Some other compounds may also have the potential to possess amphotropic behavior, but because of too high melting points these properties can not be observed, for example polyamides such as Kevlar show lyotropic behavior only, and are not discussed here. Furthermore, one has to take into account
of Molecular Structures
HO
OH
1
that many existing chemicals have never been tested for amphotropic properties. An example for such a compound is solanin (1, the structure in Fig. 1) which for many years has been reported to possess two different melting points [63]. Only recently was it shown that these two temperatures are the melting and clearing points of a smectic liquid crystal [64]. Other compounds which might possess amphotropic behavior, but which have not been studied hitherto, can be found within tannines, saponines, gangliosides, naturally occurring membrane components or biopolymers. Also many synthetic surfactants have never been studied for their lyotropic and thermotropic liquid crystal properties. As a result, not all the compounds discussed in the following are already tested for both, lyo- and thermomesomorphism. Thus, in some cases their amphotropic behavior is not yet proved but nevertheless likely. On the other hand, it becomes more and more clear that the strong separation of both kinds of liquid crystallinity is fading [65]. The amphiphilic/amphotropic compounds seem to open a view on an uniform field of liquid crystals, regarding both lyoand thermotropic behavior of compounds together as one intrinsic principle of nature.
309
Figure 1. Molecular structure of solanin, 1, the poisonous compound in green potatoes recently proven to exhibit a SmA phase [63, 641.
2 Principles of Molecular Structures Most liquid crystals can be classified as being monophilic or amphiphilic in character, but the borderline between these extremes is not sharp, see Table 1. A comparison of the examples selected in Fig. 2, see also Table 2, may emphasize this difference. While 2 is a classical, monophilic liquid crystal, 3 with the same rigid core has an intramolecular contrast due to the perfluoration of one of its chains. This weak amphiphilic character leads to a higher stability, i.e. to a higher clearing temperature of the mesophase (84.5 "C + 134 "C) exhibited by this mesogen and to a change in phase (nematic to smectic). The same effect could be observed for the cholesterol derivative 4 which exhibits a SmA phase with a clearing temperature above 110 "C caused by its amphiphilic character. All the other cholesterol derivatives, monophilic in character, have only a cholesteric (and thus less ordered) mesophase with clearing points below 100 "C [69]. The thermotropic properties of amphotropic materials are determined by the equilibrium of the hydro- and lipophilic sections
310
VI
Amphotropic Liquid Crystals
Table 1. Comparison of the general properties of monophilic and amphiphilic mesogens. Properties
Monophilic mesogens
Amphiphilic mesogens
Origin of mesomorphic behavior
Molecular shape, space filling
Intramolecular contrast, phase separation
Molecular requirements
Rod-like or disc-like shape, rigid central part
Geometrically separated, hydrophilic and lipophilic sections
Nematic phases
Very common
Very seldom, because phase separation is not possible here
Polymorphism
Very common, stepwise degradation of the order from the crystalline to the isotropic phase
Seldom, normally only the least ordered type of a packing principle occurs
Normal
Normal
Rarely observed
Normal
Common (BP, N*, TGB,)
Seldom
Thermotropic behavior Lyotropic behavior Chiral macroscopic order Disordered tilted lamellar phases
Common (SmC)
Seldom
Influence of configuration
Determinating, normally only one isomer is mesogenic
Small, normally all isomers show comparable mesogenic properties
Influence of conformation
High
Very small
Technical applications (e.g., displays)
Biological systems, membranes, Langmuir-Blodgett-films
Relevance
Table 2. Phase transition temperatures ("C) of three mesogens (2-4) different in character (see Fig. 2). ~
Mesogen
2 3 4
Cr
M
57.5 N 64 Sm 96 SmA
Is0
84.5 134 114.2
Ref.
[66] 1671 [68]
Cr = crystalline, M = mesophase, Is0 = isotropic.
of their constitute molecules. A simple sketch depicting the influence of this equilibrium on the types of mesophases developed by such substances is shown in Fig. 3; this behavior is similar to the concentration dependent lyomesomorphic properties of surfactants. In order to exhibit amphotropic properties, liquid crystals require low melting
points and a suitable polar/apolar ratio of their molecular parts. Thus, this group of mesogens is a subset of amphiphiles showing also lyotropic properties. In the case of a balanced relation between the different molecular parts a SmA phase is observed. Excessive space demand of one of these parts results in the formation of a columnar mesophase or, in extreme situations, even in discontinuous cubic mesophases. Bicontinuous cubic mesophases may occur in the area between the SmA and the columnar regions. For pure unbranched paraffins which, of course, are monophilic in character, a dense packing of the SmB type of phase (rotator phase) is observed. A pure amphiphilic compound resembles a vertical line in the sketch of Fig. 3, thus,
2
normally it only exhibits one mesophase. Unusual geometries of the hydrophilic headgroup may lead to a deviation from that vertical orientation, thus, in exceptions even
I
I
Principles of Molecular Structures
31 1
polymorphism with two or three different mesophases can be observed. In case of comparable sizes of the headgroup and the lipophilic part, more dense packings of the molecules in the mesophases may occur leading to higher ordered liquid crystals (e.g., to a SmB phase). Addition of paraffinic or protic solvents shifts the system to the left or right side, respectively, in the scheme of Fig. 3. Thus, regarding the concentration gradient (a horizontal line in the sketch) one compound in different compositions, in principle, can show all the mesophases mentioned there.
2.1 Simple Amphiphiles Figure 2. Examples for the influence of differences in polarity of the molecular mesogen structures 2-4: whereas the monophilic ester 2 forms a nematic mesophase [66], the related amphiphilic compound 3 is smectic [67]; the amphiphilic cholesteryl derivative 4 exhibits a smectic mesophase on heating [68] in contrast to simple, nonamphiphilic cholesteryl compounds which are cholesteric [69].
'1 Alkanes
The term 'simple amphiphile' describes the principle of molecular constitution of compounds which will be discussed in this section. The simplest molecular architecture of an amphiphile consists of a hydrophilic group connected to a lipophilic substructure as shown in Fig. 4.
Is0
Amphiphile
H20
Figure 3. Sketch of the principal phase behavior of amphiphilic compounds. Usual amphiphiles are represented by a vertical line in this scheme; they exhibit only one type of mesophase. Extreme geometries of one of their molecular parts or the addition of solvents (linear alkanes or water) may lead to a deviation from the vertical orientation of that line, thus, amphiphilic compounds in such situations may form various types of liquid crystal phases: T = temperature, SmB phase (rotator phase), Cub,,, and Cub,, =cubic discontinuous or cubic bicontinuous phases, respectively, Col,, = columnar hexagonal phase, Is0 = isotropic phase.
312
VI
Amphotropic Liquid Crystals
B hydrophilic headgroup
isotropic
lipophilic tail
Figure4. Schematic drawing of the structure of a monotailed amphiphile. Very often, its polar head contains several hydroxyl groups whereas the lipophilic part, for example, is an alkyl chain.
hydrophilic area
Figure 6. The principal phase behavior of monotailed amphiphiles given by the clearing temperature T vs. the chain length n of the lipophilic molecular part. With increasing chain length liquid crystallinity appears (region A), reaches a maximum stability (plateau B ) , and vanishes (region C) in the case of very long substituents.
hydrophobic
area hydrophilic area
Figure 5. Schematic representation of the architecture of a bilayer SmA phase of intercalated amphiphilic molecules.
Amphiphiles of this type, thermotropically, form most often layered mesophase structures with different arrangements of the molecules within their layers. In these layers the molecules are arranged in such a way that the hydrophilic heads point towards each other as, similarly, the hydrophobic tails do. This special type of smectic arrangement is called bizayered because, on average, each layer in the direction of the layer-normal consists of two molecules. The notation of this phase type is often done by adding a subscript 2 to the appropriate index, that is SmA, for the smectic A bilayered phase type. A schematic representation of such an arrangement is shown in Fig. 5 . Often, the layer distance in such mesophases is not twice the molecular length but
less. This might be due to a tilt of the molecules with respect to the layer normal resulting in a SmC type of phase, but the same effect may occur because of an interdigitation of molecular parts. Such an interdigitation is indicated by the subscript d in the index of the phase notation, SmA,. With regard to amphotropic liquid crystals, this SmA, phase is by far the most observed type of phase, whereas SmC, SmA, and SmA, phases are found only rarely. For the arrangement of molecules within the layers the usual notation is used: in the SmA phase the molecules have no two-dimensional order within their layers, in the SmB phase the molecules within their layers are set in a hexagonal arrangement, characteristically in the SmC the molecules are tilted with respect to the layer normal but possess no other degree of order. An interesting fact is that the smectic phases of such amphiphiles are not miscible with the mesophases of the same type of nonamphiphilic smectic mesogens. The general phase behavior of mono-tailed amphiphiles is shown in Fig. 6. Starting at a certain chain length of the lipophilic part of the molecule, a thermotropic smectic phase
2
Principles of Molecular Structures
313
T I ["C]
100 Is0 -e
80
L
30'
I
2
7'
- -?!- -E--/f IV
o . . .
am
\
601
94
89
Cr
I
I
I
1
I
I
I
I
4
6
8
10
12
14
16
18
I
n
20
Figure 7. Phase transition temperatures of the fluor organyl series F(CF,),,(CH,),,H their alkyl part [74, 76, 771.
occurs. With increasing chain length the stability of this mesophase enhances rapidly (see Fig. 6, region A ) . The maximum clearing temperature is observed for a chain length appropriate to the size of the hydrophilic headgroup (plateau in region B ) ; for simple monotailed monosaccharides the optimum is reached with n= 16 (i.e., hexadecyl). Further increase of the chain length of the lipophilic part leads to a decrease of the clearing ternperature. Thus, the equilibrium between the hydrophilic and the lipophilic parts of the molecules determines the liquid crystalline behavior. Principally, the occurrence of liquid crystallinity in such amphiphiles can be compared to the miscibility gap of, for example, glycerol and alkanes. In the region where separation occurs in such two component systems, mesomorphic properties are observed in case of chemically linked multiolalkyl amphiphiles; since they can not separate, macroscopically, 'microphase separation' takes place leading to the formation of liquid crystalline mesophases. Lyotropically, several arrangements of their supramolecular aggregates (formations of layered, columnar and cubic phases) are observed for the simple amphiphilic type of molecules.
(type 5) over the length of
One way of synthesizing a simple aliphatic carbon compound amphiphilic in character results from the combination of a polyfluorinated alkyl group with a normal alkyl chain yielding F(CF,),(CH,),H. Several studies on such amphiphiles with perfluorocarbon segments have been published [70 -761. An overview of the transition temperatures of series 5 consisting of one F(CF,),, end and a hydrocarbon residue of variable length is given in Fig. 7 and in Table 3 . In the low temperature phase the fluorinated parts of these molecules are helically arranged whereas the hydrocarbon parts display a trans-planar conformation [74]. Raman spectra show that the hydrocarbon segments are loosely packed in a hexagonal structure as observed for alkanes in the rotator phase. This implies that the hydrocarbon segments of these compounds are not packed closely as in the polyethylene lattice, they are separated enough from each other allowing motionh-otation around their axis [70]. Above the transition temperature of compounds with n = 4 - 12 the fluorocarbon segments remain ordered whereas the hydrocarbon chains in this high temperature phase become disordered [74] leading to a
314
VI
Amphotropic Liquid Crystals
Table 3. Phase transition temperatures ("C)of multifluoroalkane amphiphiles (5a-j and 6). Amphiphile
5a 5b 5c
5d 5e 5f 5g 5h 5i 5.i 6 a
Perfluorodecyl-
Is0
Sma
Cr
ethane butane hexane octane decane dodecane tetradecane hexadecane octadecane eicosane
71.0 40 44 51 66 80 90 94 97 loob
-
Perfluorodecyldecane
38
SmB
0
[741 [741 [741 1741 [741 [741 [761 [761 ~761 174,761
0
[711
0
76.0 79.0 82.0 84.0 89.0 93 -
0 0 0 0 0 0 0
-
0
61
Ref.
The mesophase of 5b-g was characterized in reference [77] as smectic. This melting temperature depends on the thermal history of the sample.
smectic phasetype. It has to be quoted from literature that this phase transition is often denoted a solid-solid transition [70,72-761 although, for example, 13CNMR spectra show that the 'hydrocarbon segments have already gained a nearly liquid-like gauche/ trans ratio' [76]. Some years after liquid crystalline properties were described for the (perfluorodecy1)decane 6 [7 11, the mesomorphic nature of the phase exhibited by 5 e and 5f at temperatures above their 'solid-solid transitions' were also recognized [77]. In case of 6, polarizing microscopy and X-ray experiments prove the existence of a SmB mesophase in the temperature range 38-61 "C, see Table 3. A different thermal behavior is observed for those examples in which the hydrocarbon chain length exceeds the length of the fluorocarbon part. Here, no phase transitions other than that to the isotropic liquid occur at elevated temperatures. Indeed, in case of 6, F(CF2)12(CH2)16H, the thermally induced molecular motion of the long hydrocarbon chain is comparable to that of short chained examples at room temperature, shown by Raman spectra, for
example, F(CF2),2(CH2)8H compared to F(CF2)12(CH2)1gH [741. Generally, the miscibility of fluorocarbon and hydrocarbon molecules is very poor [76]. This fact is seen as one reason for the 'microphase separation' discussed above. However, mixtures of semifluorinated compounds with hydrocarbon solvents [72] or fluorocarbon solvents [76] form opaque, birefringent gels [72] which are highly viscous and have been suggested to consist of interdigitated crystallites enclosing large amounts of solvent [76] between the layers of the perfluoralkyl segments. It has also been shown [76] that micelle formation occurs in nonwatery systems offering the way to many interesting supramolecular assemblies. Thus, semifluorinated compounds act like tensides in water: they are surface active compounds with various types of micellar associates. More complex molecules are those carrying functional (e.g., hydroxyl) groups or other headgroups; such compounds have been studied in a great number, see some reviews [4, 7, 9-11]. Examples of the simplest class of them are the 1,2-dihydroxyalkanes 7a-h [78,
2
3 15
Principles of Molecular Structures
Table 4. Phase transition temperatures ("C) of selected alkane-1,2-diols, 7a-h [78, 791. T,, =clearing temperatures of these compounds saturated with water (50 wt%), ( ) =monotropic mesophases. Compound
Alkane- 1,2-diol
7a 7b 7c 7d 7e 7f 7g 7h
octanenonanedecaneundecanedodecanetridecanetetradecanepentadecane-
Cr
SmB 27 26 46 49 58 61 63 68
( 0
{* ( 0 ( 0
(* ( 0 ( 0
(ab
SmA -4 12 31 42) 50) 57) 60) 65)
Tr,
Is0
16) 25) 35)
Ref. a
39 63 72 77 80 80 76 86'
-
-
In reference [78] the temperatures are given in K. In reference [79] it is classified only as smectic. Since the longer chained members 7d-g of this series exhibit only the SmB phase [78], therefore, we feel it reasonable also to expect this type for 7h. This temperature value is the maximum clearing temperature in a contact preparation of 7 h with water. a
OH
7
Figure 8. Structure of 1,2-dihydroxyalkanes (7, R = alkyl). Thermotropically, these diols exhibit layered phases of the SmA and SmB type [78].
791, shown in Fig. 8. These carbinols are amphotropic, that is they exhibit both types of mesomorphism (thermotropic as well as lyotropic) [78-8 I]. The ethylene glycol derivatives exhibit their thermotropic mesophases only monotropically, see Table 4; interestingly, their mesomorphic phase range can be stabilized remarkably by addition of water. In the latter case, the clearing temperature increases whereas the transition temperature of the SmB-SmA transition decreases. The authors conclude, here, that the formation of hydrogen bonds between adjacent molecules is responsible for their mesophase formation [78]. By 1972 the thermal properties of monoalkylated glycerols (e.g., l-alkoxy-2,3propandiol) had been investigated [82] of which the mesophase was later identified as smectic [79]. The extension of the a-diol type, shown in Fig. 8, by insertion of a rigid, rod-like
Figure 9. Molecular structure of the hybrid amphiphile 8 and the nonamphiphilic, ordinary calamitic phenylpyrimidine 9 1831.
Table 5. Phase transition temperatures ("C) of the two phenylpyrimidine derivatives 8 and 9 (see Fig. 9). Compound
8 9
Cr
SmA 94 49
Is0 Ref."
N
-
52
66
1831 [83,841
In reference 1831 the temperatures are given in K.
phenylpyrimidine unit is illustrated in Fig. 9 [83], yielding the new 1,2-diol 8 of which the hydrophilic head is linked to it by a propyloxy spacer. Opposite this phenyl substituent an alkyl chain is attached as the flexible lipophilic part of the molecule. The nonamphiphilic analogue 9 without the hydroxyl groups, a typical calamitic liquid crystal, is also known. Thus, a direct comparison of these two compounds is possible
316
VI
Amphotropic Liquid Crystals
showing that the incorporation of, for instance, hydroxyl functions into a liquid crystalline molecule leads to remarkable changes of its mesomorphic behavior. In contrast to 9 which exhibits a SmA and a nematic mesophase on heating, the lyotropic 8 is thermotropically not liquid crystalline. After two solid-solid transitions the diol 8 melts at 94°C directly into the isotropic liquid. The thermotropic data of both compounds are given in Table 5. Addition of water to the amphiphile 8 induces several lyomesophases. With growing water content N, SmA, and SmC mesophases are observed [83]. Fig. 10 shows the phase diagram of 8 with water [83]. The uptake of water into the liquid crystalline phase is limited to 10 to 20 water molecules per diol molecule. This is slightly more than for other diol amphiphiles where the thermotropic mesophases could be stabilized by addition of water and where the amount of water in the mesophases is also limied [85,86]. The binary system of 8 and water has several eutectica and its melting temperature as well as that of the upper solid-solid transition (Cr, -+ Cr,) decreases with growing amounts of water, however, the lower solid-solid transition (Cr, +Cr2) remains constant. With growing water concentration the Cr, phase is replaced by the N mesophase which occurs in a very narrow concentration range from 1 to 2 wt% of water in the mixture (see Fig. 10). This was the first observation of an inverse nematic N,, phase in a binary system of an amphiphile with water [83]. At higher water concentrations lamellar phases occur: at first, starting at 2 wt% of water, the SmA and then, above 3 wt% of water, the SmC type of lyomesophase. The clearing temperature of the SmA phase increases with growing water concentration until about 16 wt% of water above which it remains constant. For the first time, the inverse phase sequence Cr -+N -+
SmA -+ SmC was observed here in lyotropic systems of this type of binary mixture [83]. The authors discuss three different models for the possible architecture of its nematic phase [83]: The nematic phase could represent the first example of an inverse nematic N,, phase, because it appears in the concentration range between the crystalline phase and the lamellar mesophases. In this case, the mesogenic structures would consist of small and flat, disc-like supramolecular units in which the molecules cling together by cooperative hydrogen bonds between their headgroups whereas their lipophilic tails point outwards of these associates. The geometry of this phase is similar to the N, phase in thermotropic liquid crystals. - The nematic phase could consist of linear associations which are formed by hydrogen bonds between the diol groups. The formation of large hydrogen bond networks in the phase is hindered because of the small amount of only about 1.5 wt% of water. - The nematic phase could occur because the hydrogen bond interactions between the diol groups are disturbed by the nitrogen atoms of the heteroaromatic ring, also hydrogen bond acceptors; a molecular situation which disfavors the formation of bilayers. In the water-free state the crystalline phase is so stable that the (monotropic) nematic phase is, so to speak, hidden in the crystalline phase and could not develop due to rapid recrystallization during cooling. The addition of small amounts of water decreases the melting temperature and the nematic phase gets more stable than the crystalline state.
-
Owing to the very tiny concentration range, it is difficult to investigate this nematic phase in detail [83].
2 Principles of Molecular Structures
317
, I
I
cr2 +W
240
t
crl+ I
,--1 - - 7 - --I----
I
2
6
L
8
1---------
a)
5
10
15
20
30
25
wt.% H20
49
is
340
0.2
OL
0.6
0.8
1.0
I
mol H20
"w molD bl
1
2
3
4
wt%
H20
Figure 10. Phase diagrams of the binary mixture of 8 and water 1x31. The lower drawing shows a detailed, enlarged section of the upper diagram; cr,. crz, cr,=different crystalline phases, s:, Sd=smectic phases, N=nematic phase, W=water, is=isotropic phase, for further details see [83]. Figure 11. Hexadecyl-P-D-glucopyranose [56], see Table 6
318
VI
Amphotropic Liquid Crystals
A more complex and larger headgroup is found in the family of monotailed glucopyranosides 10a-1 of which the phase transition data are compiled in Table 6. In this series, the hexadecylglucopyranoside 10k, shown in Fig. 11, is of historical interest. From our viewpoint today, it is one of the longest known amphiphilic mesogens: in 191 1, E. Fischer already described the behavior of this amphiphile [56]. Although the phenomenon of liquid crystallinity has been known since about 1890 [87, 881 it was not until 1938 that the thermotropic mesomorphic character of this multiol(10k) was discovered [58]. This compound is, so to speak, the prototype of amphiphilic (and amphotropic) liquid crystals of the great family of
CH,OH
hydrophilic headgroup
lipophilic tai 1
Figure 12. Schematic representation of three different lyomesophases, lamellar, cubic, or columnar hexagonal in type [97].
monotailed carbohydrate liquid crystals. Their hydrophilic headgroup consist of the carbohydrate moiety and their alkyl chain represents the lipophilic part of this amphiphile. In the meantime, a long homologous series of this family of carbohydrate derivatives has been studied. Some of them have gained great interest and are commercially available due to their ability to act as biodetergents, for example, for the solubilization of membrane proteins [89]. The general interest in this particular field of amphotropic liquid crystals is represented in several review articles [4, 7, 9-11]. Whereas the stability of the crystalline phase of members of series 10 remains roughly constant (their melting points deviate only in a temperature range of about 30 K), the behavior of the liquid crystal into isotropic phase transition is more complex. Their clearing temperatures rise with increasing length of the alkyl chain up to a maximum clearing temperature at the tetradecyl derivative. Afterwards they drop again and the longest (tridecyl) member of this family is no longer thermotropically liquid crystalline.
Table 6. Phase transition temperatures of various P-glucopyranosides (10a-1) and of one selected a-glucopyranoside (lorn). Structure
10a-1
the a-anomer a
Compound
10a 10b 1oc 10d 10 e 10 f 1% 10h 1Oi 10j 10k 10 1 10m
R
SmA
C6H13
73 66-67 88-91
C7H15
X
C9H I 9
CIOHZI
67.07 68 70.26
C11H23
X
C2H5
C4H9
Is0
C18H37
81.5 81.5
85 a 106.38 113 133.53 139a 144.9 155 147.5 120
C12HZ5
81.4
149.5
1 ZH25
80.4
14H29
X
C,6H33
In reference [92] this temperature is given in K. this temperature is not given in the literature.
X=
Cr
Ref.
2
319
Principles of Molecular Structures
Figure 13. Molecular structure of the dodecyl glucofuranoside 11, Both, the a- and p-anomer have been studied in their pure states [94].
Table 7. Phase transition temperatures of the a- and P-anomers of 1-0-dodecyl glucose with a pyranosidic and furanosidic headgroup, respectively (see Figs. 11 and 13). ~~
Sugar a-D-Glucopyranoside fi-D-Ghcopyranoside a-D-Ghcofuranoside 0-D-Glucofuranoside
Compound
10m 10i lla llb
Cr
SmA 81.4 80.4 69.5 14.4
The lyotropic behavior of these carbohydrate derivatives has been studied very carefully [96]. It follows the classical sequence of mesophases established for the ionic detergents such as soaps. In contact preparations of 1-0-octyl-P-D-glucopyranoside (10e) with water at room temperature three types of lyomesophases are observed (from high to low amphiphile concentration): lamellar, cubic and columnar hexagonal, their principal structures are depicted in Fig. 12. To explore the relationship between molecular structure and properties of this type of mesogens in detail several modified structures, different in their hydrophilic headgroups, have been synthesized and their mesogenic properties investigated 1981. For example, the glucopyranoside has formally been replaced by the furanoside as
Is0
149.5 144.9 114.2 147.8
Ho
~~
Ref.
L941
WI P41 P41
1 OH
11
Figure 14. Molecular structures of the two anomeric amphipiles 12 and 13 with each the bicyclic glucofuranosidurono-3,6-Iactoneheadgroup 1991.
depicted for the dodecyl glucofuranosides l l a and 11 b i n Fig. 13 [94]. The properties of both families 10 and 11 (see Table 7) are quite similar causing the authors to conclude that the increased flexibility of the fivemembered furanose ring compared to the more rigid one of the pyranose is not lowering the stability of the mesophase 1941. This phenomenon is generally valid for am-
320
VI
Amphotropic Liquid Crystals
a)
Figure 15. Basic molecular structure C) R=CioHzi some~O-alkyl-N-(2-hydroxd) R = c ~ ~ of H e) R = CI2HZS yethy1)-PD-glucofuranosiduronamides b,
12
13
OH
R=C8H17
= c9H19
P91.
Table 8. Phase transition temperatures ("C) of the 0alkyl-~-glucofuranosidurono-3,6-lactone series 12 and 13 [99].
The phase transition temperatures of the two 0-alkyl series 12 and 13 are given in Table 8 [99]. Whereas the 2,5-diols of the Compound Alkyl,R Cr SmA Is0 fi-anomers 12a to 12e are thermotropically liquid crystalline each exhibiting a SmA 71.8 { * 41.6) octane12 a phase either monotropically (12 a and 12b) 77.4 [ * 70.9) nonane12b 90.9 decane82.4 12c or enantiotropically (12c to 12e), the whole 104.9 0 undecane86.8 12d a-series 13 is thermotropically not liquid 89.5 116.0 12e dodecanecrystalline. The clearing temperatures of se91-92" 0 octane13 a ries 12 are related to the length of the alkyl 91.1 0 nonane-b 13b 94.8 13c decane-' chains in the usual way: their clearing tem95.8 dodecane-b 13 e peratures rise with increasing chain length. A possible explanation for the absence of { } = monotropic mesophases. a This temperature range was measured by polarizing thermomesogenic properties of series 13 microscopy. ' Solid-solid transition temperature given in the lit- (compared to series 12) could be their more stable crystalline phases covering the poserature is skipped here. sible mesophase [99]. The isotropic melts of members of series 13 can not be supercooled very much: crysphiphilic liquid crystals: the rigidity of tallization occurs just below their melting molecules is not as important as it is for claspoints. Moreover, the anomeric configurasical calamitic liquid crystals. The a-anotion probably prevents the formation of mer of the furanosidic glucose possess a strong intermolecular hydrogen bond netlower clearing temperature then the p a n o works and an efficient intercalation of the mer. Whereas this effect is rather small in alkyl chains [99]. These examples clearly case of the glucopyranosides, the clearing demonstrate the importance of phase stabiltemperatures of furanosides display a difity and stereochemistry for the formation of ference of up to 30 K. However, an explamesophases among these amphiphiles and nation for this sterical influence could not prove, for instance, the effect that a simple yet be given by simple models. change at one chiral center can have on the An interesting variation of the furanosidthermotropic behavior of members of both ic headgroup is realized in the amphiphiles anomeric series. 12a-e [99] shown in Fig. 14. These 1-0The solubility of the humpbacked lacalkyl-D-glucofuranosidurono-3,6-lactones tones 12a-e in water is very low (less than consist of a lipophilic tail joined to a bicy1 mM at 60 "C in the case of 12e) [99]; in clic headgroup which, in comparison to this respect, no comparison has been made those already discussed, is more rigid and with regard to the structurally stretched sterspace demanding.
14
I
antibiotic derivatives of 1-deoxynojirimycin. Here, the pyranosidic oxygen atom of glucose is substituted by nitrogen allowing
OH
eoisomers 13. The essential stereochemical differences between l2and l3have not been considered with regard to their thermotropic behavior. It is surprising that the more bowl-shaped examples of 12 are thermomesomorphic whereas those of the more linear ones of 13 are not. The cleavage of the lactone ring in 12 with 2-aminoethanol leads to members of the corresponding series 14 which possess a more hydrophilic headgroup (see Fig. 15). Their liquid crystalline properties are summarized in Table 9. Further modifications of the hydrophilic headgroup of such cyclic multiols have been realized. For example, the introduction of a pyranose ring in which the ring-oxygen atom is replaced by nitrogen offering for the first time the possibility of substitution at this particular position of the cyclic headgroup of an amphiphile (see Fig. 16) [ 1001. The thermal data of these interesting Nalkylated I-deoxynojirimycin derivatives 15 are summarized in Table 10. It is noteworthy to stress that these compounds are biologically active [ 1001: they possess antibiotic properties and inhibit various glucosidases. Several patents on antiviral effects of 1-deoxynojirimycin derivatives on HIV and other retroviruses have been registered. As expected for single-tailed rod-shaped amphiphiles, four of the investigated eight 1-deoxynojirimycin derivatives 15a-h exhibit an interdigitated SmA phase on heat-
32 1
Principles of Molecular Structures
2
Table 9. Phase transition temperatures ("C) of the 0alkyl-N-(2-hydroxyethyl)-~-~-glucofuranosiduronamides 14a-c [99].The typeofmesophase(M)exhibited by them was not determined explicitly. Compound
Alkyl,R
14a 14b 14c
decylundecyldodecyl-
M
Cr 39-41 38-41 58-60
Is0
83-85 108 133
ing. Furthermore in contact preparations with water, the long chained tetrols, 15e-h, develop lyotropic mesophases of layered, cubic, and columnar hexagonal architectures. In order to investigate the influence of the position of the alkoxy chain at the hydrophilic headgroup on the mesomorphic properties of such monotailed carbohydrate liquid crystals, two research groups systematically varied the structure of two series of model compounds (16 and 17) [9, 1011. Comparing the transition temperatures of the ten pentols 16a-e and 17a-e, shown in Fig. 17, there is a strong dependence of their melting temperatures on -
-
the position of this (lipophilic) alkyl ether function, and on the fixed stereochemistry of the hydrophilic headgroups.
On the other hand, the clearing temperatures of at least three mono-ethers of each series 16 and 17 (16b, 16c, and 16e as well as 17a-c) are very similar, at around 141 or
322
VI
Amphotropic Liquid Crystals
Table 10. Phase transition temperatures ("C) obtained by DSC of eight N-alkyl derivatives of l-deoxynojirimycin [IOO]. Structure
CHzOH H 0 e . R OH 15a-h
Compound
R
Cr
15 a 15b
methyl ethyl phen ylethyl cinnamyl nonyl decyl dodecyl tetradecyl
0
15 c 15 d
15e 15f 15 g
15h
CH20H
- + - O HHO OH
15
derivatizations of the hydrophilic head group at this particular position of the heterocycle [ 1001.
Figure 17. Plot of the phase transition temperatures, obtained by DSC or polarizing microscopy, against the structure of each five mono-dodecyl ethers of the
220 "C, respectively. As, in particular, can be seen in the bottom part of Fig. 17, relatively small structural/sterical changes at the inositol ring cause dramatic changes in the melting temperatures of these amphiphiles. For instance, the formal migration of the dodecyloxy group around the cyclohexane ring, away from the axial hydroxy function changes drastically the liquid crystalline properties of these natural product derivatives. Their melting points rise by about 90 to 123 K. Interestingly, the arrangements of the substituents in 17b and 17c of this carbocyclic series possessing the weakest molecular symmetry give rise to the formation of very broad, stable SmA phases with almost the same clearing temperatures. The three mono-ethers of highest molecular symmetry (17a, 17d, and 17e), however, do not or nearly not exhibit a mesophase due to their very high melting points.
0 0 0
0 0 0 0
SmA
151.0 153.4 179.7 161.4 105.6 88.3 98.1 101.9
Is0
0 0
0 0
0
149.5 155.3 164.6 166.7
This relationship can also be detected in the sugar series 16, e.g., considering 16c. Of much lower molecular symmetry but very similar constitution are in this heterocyclic series both 16a and 16d showing almost the same melting and clearing temperatures and also the broadest and most stable SmA phases. Thus, symmetry of these molecules is an important factor for their thermotropic properties: the higher their symmetry the more stable is the crystalline phase or the less probable is the appearance of a thermotropic mesophase. Another way of varying the molecular structure of such amphiphiles is the modification of their lipophilic part by formal substitution of an alkyl chain by an 4-alkylphenyl group [102a], see 18 and 19 in Fig. 18, or by an 4-alkoxyphenyl group [ 102b]. The phenyl ring makes the sidechain more rigid. It was found here, that this structural feature increases the transition temperatures, in particular, the melting points of the a-series 18. Since a phenyl group is equal to about four to five methylene groups in length a correspondence of the temperature increments with regard to simple alkyl glucosides (e.g., series 10) is discussed [102a]. For all of the six 4-alkylphenyl glucopyranosides of the series 18 and 19 investigated, lyotropic mesomorphism has been observed at room temperature. In contrast, this is not
2 Principles of Molecular Structures
323
16
........... (P.M.)-
- 162 0
OH
........... (P.M.)
dod - 0 HO
.....
6OH
OH
dod-0 HO
6
...........
Is0
0
OH
141,s
OH
......... . _......... _
"o*HO
OH 0 -dod
.I
. - - - - ...._ . - .
'+O-dod
'0
SmA
167.2
14j.4
80.4
I
.
.
["CI
.
100
150
, dod
17
216.7 _ _ _ _ . ...................................... /;,:23.5 HO
OH
.....
HO HO
dod
__.__
0
127.6
OH
. . _ . .___ . .
2,221.7
SmA
\
Is0 216.3
d0d-O'
.......................................... 14-2\. ...........( P.M.)Cr
- 250
OH
.........................................................
I
240.0 OH
. I
150
.
,
. I
. .
.
I
.
1
200
.
1
.
1
. 1
. .
. ["C] I
250
two series of D-glucopyranose [I011 (top) or myo-I scyllo-inositol [9] (bottom). Obviously, the appearance of the mesophase on heating is determined by their relative molecular symmetry originating from the different localization of the ether group in both series, hetero- or carbocyclic, respectively, in their molecular structures; Cr = crystalline, SmA = smectic A phase, Is0 = isotropic liquid.
324
VI
Amphotropic Liquid Crystals
HOCH2
H%&q
H;S&o*R
18
OH
19
HG+R
Figure 18. Molecular structures of the 4-alkylphenyl a-and P-glucopyranosides 18 and 19 [102a]; R=C,H,, C&, or C7H15.
HOCH2
OH*" OH
O
y
20
0
Figure 19. Molecular structure of the 4-alkyl umbelliferyl /?-~-glucosidederivatives 20 [103]; R=C,H,,, C,H,,, or C,,H,,.
Table 11. Phase transition temperatures ("C) obtained by DSC of the three derivatives 20a-c of umbelliferyl 0-D-glucoside (see Fig. 19). Tetrol 20a 20b 20c
Alkyl, R C,H,, C,H,, C,,H,,
Cr
SmA 107.2 163.2 163.4
{ ] = monotropic mesophase.
Is0
{a
0
158.8) 198.0
concentration) was found [ 102b]. Furthermore the nonyloxy derivative forms long ribbons in dilute aqueous solutions [ 102b]. Another series of carbohydrate liquid crystals incorporating a rigid group in their lipophilic part of the mesogen is shown in Fig. 19; the phase transition data of these umbelliferyl derivatives 20 [ 1031 are given in Table 11. The middle part of these glucosides consists of a flat, rigid coumarin unit which is linked to a hydrophilic tetrol (Dglucopyranose) and a lipophilic, aliphatic tail. Due to the location of the alkyl chain (4-position of the lactone ring) the overall molecular shape is somewhat bent, humpbacked in the region of the carboxylic group. In spite of this steric situation these coumarin derivatives can obviously exist in a conformation which is stretched enough and allow mesogenity [ 1031.
the case for short chained alkyl glucosides 2.2 Bolaamphiphiles (chains shorter than nonyl) [ 102a]. In conclusion it is emphasized that the inThe term bolaamphiphile [104, 1051 detroduction of a rigid phenyl group does not scribing the shape of this class of mesogens significantly affect the liquid crystalline very pictorial: it is related to the South properties, only the transition temperatures American hurl weapon consisting of two rise [102a]. stones or leather balls linked by a rope. Studies of the mesomorphic properties Bolaamphiphiles consist of two hydrophilof 4-alkoxyphenyl-~-~-glucopyranosides ic headgroups at both ends of the lipophilic [ 102bl reveal that these amphiphiles form a section of the molecules. SmA phase on heating. Lyotropically, the Formally, one could compare the strucusual phase sequence for single tailed carture with two simple amphiphiles (see bohydrates (isotropic, columnar hexagonal, Fig. 4) which are linked at the ends of their lipophilic parts. The principal structure of cubic, and lamellar with decreasing water
2
hydrophilic headgroup
lipophilic Pan
Principles of Molecular Structures
325
hydrophilic headgroup
Figure 20. Schematic drawing of a bolaamphiphilic molecular structure. The polar heads consist, for example, of cyclic or acyclic multiols or carbohydrate substructures which are separated by an alkyl chain.
such bolaamphiphiles is sketched in Fig. 20. For instance, the polar headgroups may each be formed by a diol substructure [80, 811 or by carbohydrate moieties [ 1061. As expected, these compounds form layered SmA or SmC phases on heating: see the remark upon the nomenclature used here [107] and the note with regard to the Lp,-phase in section 3, General Phase Behavior, at the end of this chapter. In contrast to single tailed simple amphiphiles forming bilayers, for bolaamphiphiles monolayers are characteristic. Since bolaamphiphilic molecules form hydrogen bond networks at both of their molecular ends the stability of their mesophase is remarkably increased compared to simple amphiphiles of half the length. The a,w-bisdiols 21 depicted in Fig. 22 are characteristic examples of this kind of amphiphile. Both ends of the molecules carry a hydrophilic, vicinal bisdiol unit. Except for the one with the shortest alkyl middle section (21a, n=4), all these multiols are thermotropic liquid crystalline showing at least a SmC phase [SO, 81, 1081. The thermotropic phase transition temperatures of members of this amphiphilic family are compiled in Table 12. Multimesomorphism is observed here for examples of which n 2 1 1; their SmC phase is followed by a narrow (1-5 K broad) SmA phase on heating. The mesophase stability increases with growing length of the alkyl spacer. X-ray investigations show that in the low tempera-
Figure 21. Schematic drawing of a SmA phase built up by monolayers of bolaamphiphiles.
21 Figure 22. Molecular structure of members of the bolaamphiphilic family 21 [IOS]; n 24, see Table 12.
ture smectic phase the alkyl chains are tilted by 28" with respect to the layers of the hydrogen bond network, hence, this mesophase is of the SmC type. The authors suggest a model of the molecular interactions on the basis of their experimental data and assume that the primary OH-groups interact with the secondary OH-groups of adjacent molecules by hydrogen bond formation leading to a tilt angle of about 30" [ 1081. In the high temperature phase, SmA, the alkyl part between the hydrophilic ends is liquidlike, unordered, and untilted. Owing to the big difference in the molecular order the transition between both smectic phases is indicated by relatively large transition enthalpies between =10 and =40 kJ mol-'. Since the melting of the polymethylene chains is accompanied by a reorganization of the hydrogen bond network, the energy for this process is part of the enthalpy measured here.
326
VI
Amphotropic Liquid Crystals
Table 12. Phase transition temperatures ("C) of the bolaamphiphilic tetrol series 21 [ 1081. Solid-solid transition temperatures on some of these tetrols, given in the literature [108], are skipped here. Tetrol
n
21 a 21b 21 c 21 d 21 e 21 f 21 g 21h 21 i 21j 21k 21 1 21 m 21 n 21 0 21 P
4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 20
Cr 0
0 0 0 0 0
0 0 0
0 0 0 0
0 0 0
SmC a 79 77 101 118 114 110 85 101 87 90 79 89 84 61 87 111
SmA a
i50
-
36 I 75 1 90 I 110) 115 124 127 132 131 134 135 138 139 140 140
10
{* {. 0 0 0
0 0 0 0
0 0 0 0
128 134 136 137 139 142 143 143 141
{ ] = monotropic mesophase. a
See the note on the nomenclature used here [ 1071
H
OOH~ 0 - oH
HO
21f
~
OH
24
OH
HO
22
25
0 /J
HO
23
OH
'OH
7 OH
26
OH
Figure 23. Molecular structures of various bolaamphiphilic multiols of the types 21f, 22-26 [SO].
The hydrophilic parts of such tetrols have also been modified leading to structures shown in Fig. 23 [8 11. The bolaamphiphilic tetrol22 with its two different headgroups has the most extended polymorphism of these rnultiols. One of its mesophases is hitherto unidentified due to rapid recrystallization.
All the multiols shown in Fig. 23 are liquid crystalline except the tetrol 23 with its bis- 1,3-diol headgroups [8 11. Apparently, this situation causes a high melting point (103 "C [Sl]) suppressing the occurrence of a mesophase. The same effect was discussed in subsection 2.1 'Simple Amphiphiles' of this chapter, for example, for members of a
2
monotailed pentol series derived from inositols [9]. The other multiols depicted in Fig. 23 form layered mesophases of various types in the water-free state which are partly monotropic, however, in some cases even multimesomorphism is observed [8 11. Addition of water to bolaamphiphiles of this type with long hydrophobic middle parts leads to a stabilization of mesophases, whereas mesophases of such multiols with short middle sections get destabilized. In all these cases, the addition of water leads to the formation of the less ordered SmA type of phase instead of lamellar mesophases of higher order described above [81, 1081.
2.3
The terms ‘Y’-shaped [ 1091, ‘tripodal’, ‘1,l-double-tailed’ or ‘peg-shaped’ [110] describe the molecular geometry of this type of amphiphile. At the hydrophilic head two hydrophobic tails are joined close together or even by a common link to the headgroup. Very famous examples for this molecular architecture are the biologically active phospholipids and sphingolipids [ 1111. Lipids occur in all biological cells and have common solubility properties: generally they are water-insoluble, amphiphilic molecules
ethanolamine
H3g4
28 choline
30
-0’
I
I
~ycero~ fatty acid
Po P ‘ 0
(CH,),N@ 4
HO
327
which may form colloids, micelles or liquid crystalline phases in water. They fulfill important functions in cell membranes and signal transduction and even seem to play a tremendous role in the origin of certain kinds of cancer; see an overview on the biological aspects of glycosphingolipids [ 1121. One of the best investigated phospholipids carries a cholin residue in its hydrophilic part and is, therefore, named phosphatidylcholin. Related phospholipids are phosphatidylethanolamine, phosphatidylinosito1 and phosphatidylserine. All these phospholipids have the same molecular architecture except for their hydrophilic substituent at the phosphate link, see Fig. 24. The degree of unsaturation of their fatty acyl chains and the chemically nature of their headgroups, the latter being responsible for the ability and strength of the hydrogen bond formation both between the lipid as well as the lipid and water molecules, do play a major role for the amphotropic behavior of such lipids. In water, these compounds spontaneously aggregate to bilayers in which the hydrophilic headgroups point outwards at each side whereas the lipophilic tails form the hydrophobic middle part of the layer. These bilayers may also form more complex aggregates, such as micelles, vesicles or lyome-
‘Y ’-Shaped Materials
phosphate
Principles of Molecular Structures
Figure 24. Sketch of the principal molecular structures of phospholipids. These molecules are the main components of cell membranes and may form complex supramolecular aggregates in water.
328
VI Amphotropic Liquid Crystals
HO HO OH
0
31 Figure 25. Molecular structure of the diacyl type 31 of galactosylglycerols [ 113- 1 181.
sophases. Without such supramolecular structures life would not be possible, a fact best describing the basic importance of such supramolecular structures. Up to 75% of the total lipid content of plant chloroplast membranes consists of galactoglycerolipids 31, see Fig. 25. They have been isolated and studied in their natural form and as hydrogenated derivatives. Synthetic analogues and homologues have also been prepared and studied in detail [ 113- 1181. These materials exhibit a very complex lyomesogenic behavior including metastable and time dependent phase equilibria. Their chemical structures, types of sugar, and chain length have significant influence on their mesogenic properties. In summary, one could make the following remarks about this very important class of
multiols: lamellar (Lp, La), cubic (Q,,)and inverse columnar hexagonal (HI,)lyomesophases were observed [ 1131, in annealed samples even an L, phase. At temperatures above their LpILa phase transitions, energetically weak bilayer-nonbilayer transitions may occur but, compared to the corresponding diacyl phosphatidylethanolamines, these phase transitions (Lp-L, and bilayer-nonbilayer) occur at lower temperatures [ 1 131. X-ray investigations of especially the L, phase of galactopyranosyl- and glucopyranosyl-glycerols reveal that not simply the size or orientation of the hydrophilic headgroups regulate the supramolecular packing and, thus, the structure of the mesophases, but rather a combination of these supported interactions of their alkyl chains are responsible in this matter [113]. A discussion of the thermotropic properties of this type of compounds began in 1983 [ 1 191with naturally occurring lipids. Again, the heterogeneity of the materials made syntheses of uniform derivatives desirable [ 114aI. Pure 1,2-di-O-acyl-3-O-(a-~-glucopyranosy1)-sn-glycerols (32) were investigated in view of their thermotropic properties. The thermotropic mesophase of the members of series 32 has been denoted as
Table 13. Phase transition temperatures ("C) of some selected a-D-glucopyranosyl-diacylglycerolsof type 32 [114a]. The temperature ranges in the first column (data for the transitions Cr+ 11) are indicated in reference [ 114al as so-called 'softening point'. Structure
0
32a-i
Lipid
n
32a 32b 3 2 ~ 32d 32e 32f 32g 32h 32i
12 13 14 15 16 17 18 19 20
Cr
Is0
11" 40-44 58-60 67-69 67-71 76-77 79-81 80-82 81-83 89-91
120-122 128-130 129-131 139- 140 142- 143 143-144 144- 145 145- 146 141-142
a The type of this mesophase has originally been denoted as probably smectic [114a] which later (without evidence) has been emphasized as columnar [ 10, 121.
2 R
I
X
R
I
R
R
I
1
HO
R
R
I
I
Principles of Molecular Structures
R
329
R
I
I
::{HofoH S
OH
HO
OH
33 x=o,s
OH
34
35
CH,
36
Figure 26. Acetals and thioacetals possessing a tripodal, Y-shaped structure; R=alkyl [123].
photropic materials, are the thio- and oxy‘probably smectic’ [ 114al; later it has been acetals 33-36 of carbohydrates shown in referred to be of a columnar type [lo, 121 Fig. 26 [94, 123- 1271. Interestingly, these which would be in better accordance with amphiphiles exhibit a columnar thermothe molecular structure of these amphitropic mesophase [ 123, 1261. The two space philes. It should be pointed out here, that demanding lipophilic chains enforce a curthis di-0-acyl series 32 was discussed in ref. vature of the interfacial region between the [ 10 and 121 as di-0-alkyl compounds synhydrophilic and the hydrophobic parts of the thesized by other workers [114b]. Selected mesophase resulting in a columnar type of phase transition data of this family of multheir mesophase. The architecture of this tiols are compiled in Table 13. phase is of an inverse micellar kind: the hyAn interesting example of a synthetic biodrophilic headgroups point inwards of the active representative of this family is 1,2-di0-9’-0ctadecynyl-3-O-~-~-galactopyrano- columns whereas the lipophilic chains do the opposite. syl-sn-glycerol [ 1201described by the authors A substitution of the CH,OH part by the of having a ‘sintering point’ at 55°C and a nonpolar methyl group, illustrated in 36 of melting point at 76 “C; here, a liquid crystalFig. 26, destroys the mesomorphic properline phase is much likely. Its lyotropic phase ties totally which is obvious since the hybehavior and also that one of mono-galactodrogen bond network (inside the columns) syl- or -glucopyranosyl-glycerol derivatives is severely perturbed, even disturbed in this [121, 1221 is very complex: a gel-phase and case [ 123, 1261. lyomesophases of the La, Qi, and HIIphasAll the Y-shaped mesogens described es have been observed. Here again, differenchitherto have two alkylchalcogeno chains es in the stereochemistry of the headgroups, as the upper part and a sugar moiety as the especially in the interfacial region between lower part of that ‘Y’. A different architecthe polar and the apolar part of the molecules, ture is found, for example, for N-dodecylare discussed to cause differences in the phase D-galacturonamide-didodecyl-dithioacetal behavior, see the discussion for the mono[ 1281 which also has a ‘Y’-shaped molecutailed amphiphiles [ 1211. lar structure, however, carrying now the Completely different in structure, but alsugar part in the center of that ‘Y’ surroundso belonging to this group of Y-shaped am-
Qi
330
VI
Amphotropic Liquid Crystals
2.4 Disc-shaped Amphiphiles
$H20H
HO
37
14H29
HO O* "
"2,
HO CHzOH
Figure 27. Formula of the unusual, Y-shaped amphiphile 37 with two hydrophilic headgroups [ 1291.
ed by three alkylhetero chains. This compound behaves similarly to disc-shaped amphiphiles, see below. An 'exotic' 'Y'-shaped mesogen is 37 [129], depicted in Fig. 27, possessing two hydrophilic (P-glucopyranosidic) units as the upper parts of the 'Y', both terminally linked to a branched (lipophilic) alkyl chain. The mesophase of this interesting glycoside is of a cubic type (Cr 111"C Cub 205°C I) [129].
Research on liquid crystals with a disc-like shape started in 1977 [130] when hexa(heptanoy1)benzene (38, first synthesized forty years earlier [ 1311) was proved to form a columnar mesophase (m. p. =I, 8 1.2 "C, cl.p.=Z,, 87.0"C) [132]. Disc-shaped amphiphilic molecules with saturated cores, such as the hexaesters 39 or even the hexaethers 40 of the naturally occurring scylloinositol [9, 109, 133-1371 form columnar liquid crystals much more easily with mesophases more stable and very much wider in range than known for the mentioned benzene derivatives. Table 14 compiles the phase transition data of some of these cyclic model compounds, 38-42 [ 133- 139) discussed here and shown in Fig. 28. The stability of the mesophase decreases drastically when the carbonyl functions of 39 are replaced by CH,-groups; this formal
R
'h~&----~~ I
RRO
0 38
40 CH,OR]
R10*OR'O R2 OR'
41
42
Figure 28. Some disc-shaped model compounds (38-42) amphiphilic and mesomorphic in character (38-41) or monophilic and not liquid crystalline (42) [9, 109, 131-1371; R=alkyl, R'=-CO-alkyl, R2=-0-alkyl or only alkyl.
2
Principles of Molecular Structures
33 1
Table 14. Phase transition temperatures (“C) of selected disc-shaped molecules (38-42). different in their amphiphilic character (see Fig. 28). Compound
38 39 a 39b 39 c 40 41 42
R %HI,
0
0
C6H 13
0
C9H19
0
C6H, 3
0
C9H I9
C6H13
Col,
Cr
0
81.2 68.5 68.0 84.0 18.4 37.5-39.5 37.5-40.0 66.2
transformation from esters into ethers weakens the amphiphilic character desisively. Moreover the star-shaped alkane 42 without any heteroatom is not liquid crystalline at all, apparently, due to a missing possibility of ‘microphase separation’. Again, an intramolecular contrast of hydrophilic and lipophilic parts is necessary for the occurrence of liquid crystallinity of such compounds, for which also a nearly perfect space filling of the substituents plays an important role [ 134, 1371. Peracylated sugar derivatives of type 41, studied independently by two groups [ 134, 1381, possess the required amphiphilicity, but owing to the gap around the pyranosidic oxygen the space-filling of the molecular periphery by only five substituents is not optimal and complicates the molecular situation. On the one hand [138, 1391, these compounds have been described as monotropically mesomorphic whereas on the other thermomesomorphism was not observed [134]. The monotropic mesomorphism is described to be very durable and allow detailed studies, for example by polarizing microscopy, DSC, X-ray diffraction, and circular dichroism spectra [ 1391. Both the a-and the p-anomer of type 41 form columnar mesophases of which the type depends on the molecular structure and on the thermal history as well as the thickness of the sample [ 1391.
Is0
87.0 199.5 200.0 188.7 90.8
Ref.
0 0 0
0 0
0
31.5-32.5)
0 0
With regard to the architecture of these phases the authors suggest a helical arrangement of the molecules in the columns. This chiral arrangement is more pronounced in case of the a-anomers, because the axially oriented chain at the anomeric position mesh into the gap at the pyranosidic oxygen of the neighboring molecule in the column [139]. In case of the p-anomer this ‘ratcheting’ is not expected, and indeed not observed, since all five substituents are in an equatorial position [139]. A comparative interpretation of the results of both research groups and a discussion of them with regard to tetraacylated- 10-alkylated glucosides is given in [140]. The latter materials form a highly viscous anisotropic paste which shows all the properties of a mesophase. Owing to this investigation [140] it was suggested that the strong anomeric effect of the ester groups in 41 should be considered. It reduces the population of the all-equatorial conformation and weakens the stabilizing gauche interactions between all the ester groups compared to the hexaesters of scyZZo inositol. It has to be mentioned, here, that short chained scyllo-inositol hexaesters 39 exhibit highly ordered columnar mesophases and a cubic phase as well [135]. In view of the chirality of compounds like 41 and other mostly carbohydrate based materials their ability to induce chirality in a
332
V1
Amphotropic Liquid Crystals
P I
R,
R’
R’s*’S
SI
s\R R
43
R
nematic mesophases was discussed elsewhere [ 1411. The phenomenon of increasing ability to form a mesophase as a function of amphiphilicity was also observed when the benzene hexakis(thi0ethers) 43 were oxidized into their hexakis(su1fones) 44 [ 142- 1441 (Fig. 29). In contrast to the nonliquid crystalline thioether series 43, the radial-polar hexasulfones 44 with chains containing between 7 and 15 carbon atoms exhibit a columnar hexagonal type of mesophase established by various methods, including X-ray diffractometry [ 142- 1441. Their lyotropic mesomorphic behavior is also of interest and currently under investigation [ 1451. Single crystal X-ray studies of derivatives of series 44 [ 1441 proved that the very dense packing of the sulfono groups form a broad, wavy quasi-macroheterocycle surrounding the benzene ring in a tight, spacefilling manner. The diameter of this superimposed covalently nonbonded ‘belt-structure’ of this new ‘core’ is about 107 nm [ 1441.
2.5
Metallomesogens
A completely different type of amphotropic materials is made up by various metal organyls, such as the linear ones illustrated by the di-palladium organyl 45, the twinned rod-like organyl 47, or the disc-like organyls 48 and 49 (Figs. 30 to 32). These and related metallomesogens exhibit thermo-
44
Figure 29. Molecular structures of the nonmesomorphic hexakis(thi0ether) series 43 and their hexasulfones 44 exhibiting a columnar type of mesophase [ 142, 1431; R = for example, nonyl, undecyl, or tridecyl.
tropic properties and are lyomesogenic in apolar organic solvents, such as pentadecane [146-1571. A survey about the mesomorphic behavior of these compounds including their amphotropic nature and their ability to form charge transfer complexes is given in [151, 152, 1581. The linear palladium organyl 45 is nonmesomorphic in its pure state, but a highly viscous thermotropic mesophase is induced by formation of charge transfer complexes with electron acceptors, such as trinitrofluorenone (46, TNF) [ 1481.In ternary systems of 45 with TNF and apolar solvents two types of mesophases, a nematic and a culumnar hexagonal one, are observed; the latter one exhibits a ‘herring-bone’ texture, typical for an M-chromonic type of phase. Owing to strong attracting forces in donor acceptor complexes a columnar nature of the nematic lyomesophase is reasonable in these cases [ 1481. Twin-like, bis-linear di-palladium mesogens of type 47 which thermotropically exhibit smectic phases [ 1591 form lyotropically different variants of supramolecular packings depending on the length and type of their four substituents as well as on the chain length of the organic solvent [154, 1551. In contrast to 45 such H-shaped metallomesogens are lyotropic liquid crystalline only without TNF. The formal addition of lateral substituents to the chemical structure of 47 (ringed in the formula of 48 in Fig. 3 1 ) leads to disc-
2
Yzc,2h
45
Principles of Molecular Structures
333
Figure 30. Molecular structures of the dinuclear palladium complexes 45 [148] and the electron acceptor trinitrofluorenone (46).
46
I
R
47
48
Figure 31. Molecular structures of the twinned rod-like palladium organyl47 [159] and the disc-like analogue 48 [149, 1561. The ringed lateral substituents of 48 are responsible for the change of the mesomorphic behavior; M = Pd or Pt, X=CI, Br, I (47) or X = C1, Br, I, SCN, N, (48).
shaped bis-metal complexes of type 48 with either palladium [ 1461 or platinum [149, 1561 in their molecular centers. In contrast to the bis-linear metalorganyl family 47 which behaves as calamitic mesogens (exhibiting mostly smectic phases), those ones of type 48 form a monotropic nematic-dis-
cotic (N,) phase in their pure states, the palladium organyls are the first metallomesogens with this property. However, mixtures of 48 with 2,4,7-trinitrofluorenone (46, TNF) shows also charge transfer complexes and the induction of columnar mesophases (Col,, or NcoJ [157].
334
VI
Amphotropic Liquid Crystals
PR
R?
OR
RO N
N
0 0 RO$
/
compatibility of their organometallic aromatic core with the surrounding alkyl chains. Thus, the latter has been denoted as 'internal solvent' [ 1581.
'
IM $O 'R
2.6 Polyphilic Liquid Crystals
OR
RO OR
RO
49 Figure 32. Molecular structure of the tetrametallomesogen 49 [146, 149, 1561; M=Pd or Pt, X=CI, Br, I. The spacer unit (n may ) he phenylene, stilbenylene or terphenylene.
The lath-like tetrapalladium or -platinum liquid crystals of structure 49 form columnar (Col,,,) mesophases on heating. Usually, the range of the mesophase this class of compounds exhibits is quite broad, typically about 200-250 K wide [155]. At high concentrations of 49 in solvents, such as in long-chain alkanes, a viscous two dimensionally ordered columnar mesophase of a similar type as under only thermal conditions is observed. At lower concentrations the formation of either one or two different nematic columnar mesophases appears [147, 148, 1501. In studies on such metallomesogens it became clear that the metal located deeply inside, quasi 'hidden', in the molecules has only a relatively weak influence on the mesogenic behavior of these organyls [149, 1601. Nevertheless, the amphiphilic character of these metallomesogens arises from the in-
Most of the amphiphiles discussed hitherto have only two distinct parts in their molecules which are different in their polarity. Polyphilic compounds possess more than two of them (see Fig. 3 3 ) depicting the calamitic biphenyl derivative 50 with three different parts: a polyfluorinated one, an alkoxy chain, and the aromatic (biphenyl) section. The mesomorphic properties of this triphilic examples are compiled in Table 15.
so Figure 33. A polyphilic mesogen consisting of three different parts: a fluorinated alkyl chain, a common alkyl, and an aromatic part. The terminal CH,-CF, group enables a smooth junction to the next layer. The phase sequence observed for this material is Cr 95°C (SmX92"C)SmA 113"CIso;Cr=crystalline,SmX= a ferroelectric smectic phase [162].
Figure 34. Sketch of the structure of a directed SmA phase of polyphilic mesogens.
335
3 General Phase Behavior
Table 15. Phase transition temperatures ("C) of the triphilic biphenyl derivative 50 [162] (see Fig. 33). Compound
50
Cr
SmX
95
{*
Is0
SmA
92)
a
113
51 The polyphilic quality leads to a further ordering of the molecules in their mesophase: Fig. 34 gives a simple schematic drawing of a directed lamellar phase. The molecules have a polar orientation within the layers and long range correlations between them. This special mesophase structure causes macroscopically polar properties. With regard to this kind of molecular arrangement some polyphilic compounds have been successfully studied in respect of ferroelectric properties [ 161- 1651.
2.7 Further Molecular Architectures Other types of interesting, sometimes even curious, amphotropic materials with unusual molecular architectures have also been found. Whereas Y-shaped [ 1091 molecular structures have flexible wings, T-shaped compounds are made of rigid lipophilic units [166]. In the T-shaped case, smectic instead of columnar phases are observed. While mono-alkylated amphiphiles usually show only SmA phases, an interesting exception are so-called 'banana shaped' amphiphiles of which 51 in Fig. 35 may be an example. This gentiobioside exhibits a bicontinuous cubic mesophase [ 1291. Hitherto, only monomeric, unimolecular compounds have been discussed in this overview article of which the amphotropic behavior is based on a balanced amphiphilic character. This is also true for polymeric materials [5, 11, 167, 1681 which, however, are not included here.
Figure 35. Molecular structure of a gentiobioside (51) possessing a bent geometry [95, 1291.
3 General Phase Behavior According to Fig. 3, classical thermotropic smectic phases of amphotropic liquid crystals are (SmA,), columnar hexagonal (Col,), bicontinuous cubic (Cub,,), or discontinuous cubic (Cub,,,) [ 1691. All these mesophases include a disclination surface between the hydrophilic and the lipophilic parts of the unordered molecules. This surface can be uncurved (SmA), curved in one direction (columnar), curved in two directions with the same sign (discontinuous cubic), or curved in two directions with opposite sign (bicontinuous cubic). Whereas monophilic liquid crystals can show a high diversity of smectic phases (SmA-SmQ), the amphotropic liquid crystals normally exhibit only the SmA, phase. Tilted smectic phases are only observed in a few cases. The first indication of possibly tilted phases was given in 1933 for thallium stearate [ 1701. A disordered SmC phase was also clearly described for mesogens containing a classical calamitic core aside to their amphiphilic structure [ 17 I]. Monophilic liquid crystals can show various ordered tilted smectic phases, for example, smectic I, F, G, J, H, and K. In the case of lipids only one mesophase, the p' phase,
336
VI
Amphotropic Liquid Crystals
with an ordered tilted lamellar structure was reported. A detailed analysis of this P’ phase in relation to the tilted smectic phases was given by Clark et al. [ 1721. Ordered orthogonal lamellar phases were reported for lecithins (Pphase) and for diols [78, 821. In these mesomorphic amphiphiles the diameter of each of the polar headgroups is comparable with the diameter of the paraffinic part. This allows a very dense packing of the molecules in their mesophase, thus, these phases are related to the rotator phase of simple paraffins. The columnar phase of amphiphilic liquid crystals is most often of a hexagonal disordered type. Only a very few examples are known for rectangular [173-1751 or monoclinic [ 1731 columnar phases or for ribbon phases [ 1761.
4 Summary In this chapter we described the functional principles of mainly amphiphilic compounds possessing the ability to form both lyo- and thermotropic liquid crystals. Although amphiphilicity and amphotropy are not synonymic, these properties are often coupled. Owing to the tremendous number of potentially amphotropic materials it is impossible to give a complete listing of all of them in such a chapter; but we have tried to show the functional principles which allow compounds to behave the way they do. Both, the lyo- and thermomesogenity are of high importance. While the technological breakthrough for thermotropic liquid crystals is obvious (for example, liquid crystal displays or thermography), some promising recent findings in the field of lyotropic liquid crystals concern the probable observation of a biaxial nematic phase (for the existence or doubts related to biaxial nemat-
ics see [ 177- 182]),and the spontaneous formation of a chiral order in lyomesomorphic compositions of nonchiral materials [ 1831.
References I. G. Chistyakov, Kristallografiya (russ.) 1960, 5,962. V. A. Usol’tseva, I. G. Chistyakov, Uspekhi Khim. (russ.) 1963, 32, 1124. H. Kelker, Mol. Cryst. Liq. Cryst. 1973, 21, 1. G. A. Jeffrey, Acc. Chem. Res. 1986, 19, 168. H. Ringsdorf, B. Schlarb, J. Venzmer, Angew. Chem. 1988, 100, 117; Angew. Chem., Int. Ed. Engl. 1988, 27, 113. A. S. Sonin, 100 Years - The History of Discovery and Research on Liquid Crystals (russ.) (Ed.: B. K. Vainshtein), Nauka Publishers, Moscow, Russia, 1988. G. A. Jeffrey, L. M. Wingert, Liq. Cryst. 1992, 12, 179. N. V. Usol’tseva, Lyotropic Liquid Crystals Chemical and Supramolecular Structure (russ.), Ivanovo University, Ivanovo, Russia, 1994. K. Praefcke, D. Blunk, J. Hempel, Mol. Cryst. Liq. Cryst. 1994, 243, 323. H. Prade, R. Miethchen, V. Vill, J. Prakt. Chem. 1995,337,427. C. M. Paleos, D. Tsiourvas,Angew. Chern. 1995, 107, 1839; Angew. Chem., Int. Ed. Engl. 1995, 34, 1696. C. Tschierske, Proc. Polym. Sci. 1996, 21, 775. W. Heintz, Ann. Chem. 1854, 92, 291. W. Heintz, J. Prakt. Chem. 1855, 66, 1. B. W. Miiller, Seven-Ole-Fette-Wachse 1976, 102,493. B. W. Muller, Arch. Pharm. (Weinheim, Ger.) 1977,310,693. R. Virchow, VirchowS Arch. Pathol. Anatom. Physiol. Klin. Med. 1854, 6, 57 1. C. Mettenheimer, Corr. Blatt des Vereins fiir gem. Arbeitz. Ford. der Wissensch. 1857, 24, 331. G. Valentin, Die Untersuchung der Pjlanzen und Thiergewebe im polarisierten Lichte, Engelmann, Leipzig, Germany, 1861. G. Quinke, Ann. Phys. Chem. 1894,53,593. J. W. McBain, Kolloid Z. 1926,40, 1. F. C. Bouden, N. W. Pirie, Proc. Roy. Soc. Lond. 1931,123,274. A. S. C. Lawrence, Trans. Faraday Soc. 1933, 29, 1008. G. S. Hartley, Aqueous Solutions of Para$n Chain Salts, Hermann et Cie, Paris, France, 1936.
5 L2.51 R. S . Bear, K. J. Palmer, F. 0. Schmitt, J. Cell. Comp. Physiol. 1941, 18, 3.55. [26] J. D. Bernal, I. Fankuchen, J . Gen. Physiol. 1941, 25, 1I I . [27] J. Stauff, Kolloid Z. 1939, 89, 224. 1281 H. Kiessig, W. Philippoff, Natur Wiss. 1939,27, 593. [29] R. D. Vold, M. J . Vold, J. Am. Chem. Soc. 1939, 61, 808. [30] R. D. Vold, J. Phys. Chem. 1939,43, 1213. [31] J. W. McBain, R. D. Vold, M. Frick, J. Phys. Chem. 1940,44, 1013. 1321 F. 0. Schmitt, R. S . Bear, K. J. Palmer, J. Cell. Comp. Physiol. 1941, 18, 31. (331 S. Ross, J. W. McBain, J. Am. Chem. Soc. 1946, 68, 296. [34] J. W. McBain, S. S . Marsden, Acta Cryst, 1948, I , 270. [35] S. S. Marsden, J. W. McBain, J . Am. Chem. Soc. 1948, 70, 1973. [36] J. B. Finean, Experimentia 1953, 9. 17. [37] F. B. Rosevear, J. Am. Oil Chem. Soc. 1954.31, 628. [38] V. Luzzati, H. Mustacchi, A. Skoulios, Nature 1957, 180,600. [39] V. Luzzati, H. Mustacchi, A. Skoulios, F. Husson, Actu Cryst. 1960, 13, 660. [40] F. Husson, H. Mustacchi, V. Luzzati, Actu Cryst. 1960, 13, 668. [41] A. E. Skoulios, V. Luzzati,Actu Cryst. 1961.14, 278. [42] P. A. Spegt, A. E. Skoulios, V. Luzzati, Acta Cryst. 1961. 14, 866. [43] V. Luzzati, F. Husson, J. Cell B i d . 1962, 12, 207. [44] V. Luzzati, A. Nicolaieff, J . Mol. Biol. 1963, 7, 142. [45] P. A. Spegt, A . E. Skoulios, Acta Cryst. 1964, 17, 198. [46] P. Ekwall, 1. Danielsson, Acra Chem. Scand. 1951, 5, 973. [47] K. Fontell, P. Ekwall, L. Mandell, I. Danielsson, Acta Chem. Scund. 1962,16,2294. [48] R. F. Grant, B. A. Dunell, Can. J. Chem. 1960, 38, 2395. [49] H. Moor, K. Muhlenthaler, H. Waldner, A. FreyWyssling, J. Biophys. Biochem. Cytol. 1961. 10,
I. [50] D. Vorlander, Ber. Dtsch. Chem. Ges. 1910.43, 3 120. [51] D. Demus, H. Sackmann, K. Seibert, Wiss. 2. Univ. Halle, Math.-Nut. R. 1970, 19 ( 5 ) ,47. [S2] P. Gaubert, Compt. Rend. 1919, 168, 277. [S3] V. Vill, Mol. Cryst. Liq. Cryst. 1992, 213, 67. [54] C. Eaborn,J. Chem. Soc. 1952,2840;C. Eaborn, N. H. Hartshorne, J . Chem. Soc. 1955, 549. [55] J. D. Bunning, J . W. Goodby, G. W. Gray, J. E. Lydon, Liquid Crystals of One- and Two-Diniensional Order, Springer Ser. Chem. Phys., Vol. 11
References
337
(Eds.: W. Helfrich, G. Heppke), Springer, Berlin, 1980. [561 E. Fischer, B. Helferich, Liehigs Ann. Chem. 1911, 383, 68. [571 A. H. Salway, J. Chem. Soc. 1913,103, 1022. [581 C. R. Noller, W. C. Rockwell, J . Am. Chem. Soc. 1938,60,2076. [591 R. Hori, Yakugaku Zusshi 1958, 78, 523; ihid., 1958, 78, 999; ibid., 1958, 78, 1171; R. Hori, T. Koizumi, ibid., 1958, 78, 1003. [60] R. Hori, Y. Ikegami, Yukugaku Zasshi 1959, 79, 80. [61] G. J. T. Tiddy, Phys. Rep. 1980, 57, 1. [62] H. Hoffmann, Ber. Bunsenges. Phyx Chem. 1994, 98, 1433. [63] W. L. Porter. Am. Potato J. 1972,49,403. [64] V. Grassert, V. Vill, Liq. Cryst. Today 1994,4,4. 1651 M. A. Marcus, P. L. Finn, Mol. Cryst. Liq. Cryst. (Lett.) 1985, 2, 1.59. [66] D. Demus, H.-J. Deutscher, D. Marzotko, H. Kresse, A. Wiegeleben, Liq. Cryst.; Proc. Int. Conf , Bangalore (Ed.: S. Chandrasekhar), 1980, p. 97. [67] T. I. Zverkova, E. I. Kovshev, L. I. Moklyachuk, Y. A. Fialkov, L. M. Yagupolski, Adv.Liq. Cryst. Rex Appl. (Ed.: L. Bata) 1980, 991. 1681 J. A. Szabo, A. I. Zoltai, 4th Liq. Cryst. Conf Soc. Countries, Tbilisi, 1981, abstract C 40. 691 V. Vill, LiqCryst - Database of Liquid Crystalline Compounds, LCI Publisher, Hamburg, Germany, Fujitsu Kyushu Systems (FQS) Ltd., Fukuoka, Japan, 1995. 701 J. F. Rabolt, T. P. Russell, R. J. Twieg, Macromolecules 1984, 17, 2786. 711 W. Mahler, D. Guillon, A. Skoulios, Mol. Cryst. Liq. Cryst. (Lett.) 1985, 2, 11 1. [72] R. J. Twieg, T. P. Russell, R. Siemens, J. F. Rabolt, Macromolecules 1985, 18, 1361. [73] R. J. Twieg, J. F. Rabolt, T. P. Russell, Polym. Prepr. (Am. Chem. Soc., Div. Polym. Chem.) 1985, 26, 234. [74] T. P. Russell, J. F. Rabolt, R. J. Twieg, R. L. Siemens, B. L. Farmer, Macromolecules 1986, 19, 1135. [75] J. F. Rabolt, T. P. Russell, R. Siemens, R. J. Twieg, B. Farmer, Polym. P r e p (Am. Chem. Soc., Div. Polym. Chem.) 1986, 27, 223. 1761 J. Hopken, C. Pugh, W. Richtering, M. MZiller, Makromol. Chem. 1988, 189, 91 I . [77] J. Hopken, M. Moller, Macromolecules 1992, 25, 2482. 1781 C. Tschierske, G. Brezesinski, F. Kuschel, H. Zaschke, Mol. Cryst. Liq. Cryst. (Lett.) 1989,6, 139. [79] C. Tschierske, G. Brezesinski, S . Wolgast, F. Kuschel, H. Zaschke, Mol. Cryst. Liq. Crysf. (Lett.) 1990, 7, 131. [SO] C. Tschierske, H. Zaschke,J. Chem. Soc., Chem. Commun. 1990, 1013.
338
VI
Amphotropic Liquid Crystals
[81] F. Hentrich, C. Tschierske, H. Zaschke, Angew. Chem. 1991,103,429;Angew. Chem., Int. Ed. Engl. 1991,30, 440. [82] A. Seher, Chem. Phys. Lipids 1972,8, 134. For a discussion of l-alkoxy-2,3-propandiol and related thioethers and amines see also H. A. van Doren, R. van der Geest, R. M. Kellogg, H. Wynberg, Reel. Trav. Chim. Pays-Bas 1990, 109, 197. [83] N. Pietschmann, A. Lunow, G. Brezesinski, C. Tschierske, F. Kuschel, H. Zaschke, Colloid Polym. Sci. 1991,269, 636. [84] D. Demus, H. Demus, H. Zaschke, Fliissige Kristalle in Tabellen, Deutscher Verlag fur Grundstoffindustrie, Leipzig, 1974, compound 4142, p. 260. [85] G. Brezesinski, A. Madicke, C. Tschierske, H. Zaschke, F. Kuschel, Mol. Cryst. Liq. Cryst. (Lett.) 1988, 5, 155. 1861 C. Tschierske, A. Lunow, D. Joachimi, F. Hentrich, D. Girdziunaite, H. Zaschke, A. Madicke, G. Brezesinski, F. Kuschel, Liq. Crysr. 1991, 9, 821. [87] F. Reinitzer, Monatsh. Chem. 1888, 18,421. [88] 0. Lehmann, Z. Phys. Chem. 1889,4,462. 1891 W. J. de Grip, P. H. M. Bovee-Geurts, Chem. Phys. Lipids 1979, 23, 321. [90] W. Schneider, J. Sepp, 0. Stiehler, Ber. Dtsch. Chem. Ges. 1918,51,220. 1911 L. C. Kreider, E. Friesen, J . Am. Chem. Soc. 1942,64, 1482. [92] A. Loewenstein, D. Igner, U. Zehavi, H. Zimmermann, A. Emerson, G. R. Luckhurst, Liq. Cryst. 1990, 7, 457. [93] J. W. Goodby, Mol. Cryst. Liq. Cryst. 1984, 110, 205. [94] L. F. Tietze, K. Boge, V. Vill, Chem. Ber: 1994, 127, 1065. 1951 V. Vill, Ph. D. Thesis, Universitat Miinster, 1990. [96] Y. J. Chung, G. A. Jeffrey, Biochim. Biophys. Acta 1989, 985, 300. [97] G. H. Brown, P. P. Crooker, Chem. Eng. News 1983, Jan. 3 I , 24. [98] V. Vill, T. Bocker, J. Thiem, F. Fischer, Liq. Cryst. 1989, 6, 349. [99] H. W. C. Raaijmakers, B. Zwanenburg, G. J. F. Chittenden, Reel. Trav. Chim. Pays-Bas 1994, 113, 79. [I001 D. Blunk, K. Praefcke, G. Legler, Liq. Cryst. 1994, 17, 84 1. 1011 R. Miethchen, J. Holz, H. Prade, A. Liptak, Tetrahedron 1992, 48, 3061. 1021 a) L. M. Wingert, G. A. Jeffrey, Jahangir, D. C. Baker, Liq. Cryst. 1993, 13, 467; b) E. Smits, J. B. F. N. Engberts, R. M. Kellogg, H. A. van Doren, Liq. Cryst. 1997, 23, 481. 1031 D. Blunk, K. Praefcke, G. Legler, Liq. Cryst. 1995, 18, 149.
[ 1041 J.-H. Fuhrhop, J. Mathieu,Angew. Chem. 1984,
96, 124;Angew. Chem., Int. Ed. Engl. 1984,23, 100. [I051 J.-H. Fuhrhop, H.-H. David, J. Mathieu, U. Liman, H.-J. Winter, E. Boekema, J . Am. Chem. Soc. 1986,108, 1785. [lo61 W. V. Dahlhoff, Z. Naturforsch. 1988, 43b, 1367. [ 1071 In their paper the authors follow the nomenclature by Luzzati and indicate the observed mesophases L, or Lpt, respectively. Here, however, to separate between lyotropic and thermotropic mesophases the indication was changed into SmA and SmC, respectively. [ 1081 F. Hentrich, S. Diele, C. Tschierske, Liq. Cryst. 1994, 17, 827. [I091 K. Praefcke, P. Marquardt, B. Kohne, W. Stephan, J . Carbohydr. Chem. 1991,10,539. [ 1 101 N. H. Hartshorne, The Microscopy cf Liquid Crystals, The Microscope Series, Vol. 48, Microscope Publications Ltd., London 1974, pp. 105-109; N. H. Hartshorne in Liquid Crystals and Plastic Crystals, Vol. 2, Chap. 2 (Eds.: G. W. Gray, P. A. Winsor), Ellis Horwood Ltd., Chichester, Halsted Press a div. of John Wiley & Sons, Inc., New York, London, Sydney, Toronto, 1974, pp. 5 1-54; K. Praefcke, B. Kohne, A. Eckert, J. Hempel, Z. Naturforsch. 1990, 45b, 1084. [ 11 I] In sphingolipids the glycerin component of the lipid is replaced by sphingosin. Main families of sphingolipids are glycosphingolipids (cerebrosides, gangliosides and sulfatides) and sphingophospholipids (sphingomyelins and inositolsphingophospholipids). [ I 121 S. Hakomori, Sci. Am. 1986, 254 (5),32. [ 1 131 D. A. Mannock, R. N. McElhaney, Biochem. Cell Biol. 1991, 69, 863. [ I 141 a) D. A. Mannock, R. N. A. H. Lewis, R. N. McElhaney, Chem. Phys. Lipids 1990,55,309; b) L. Six, K.-P. RueB, M. Lieflander, Tetrahedron (Lett.) 1983, 24, 1229 and references cited therein. [115] H. Kuttenreich, H.-J. Hinz, M. Inczedy-Marcsek, R. Koynova, B. Tenchov, P. Laggner, Chem. Phys. Lipids 1988,47, 245. [116] D. A. Mannock, R. N. A. H. Lewis, R. N. McElhaney, Chem. Phys. Lipids 1987, 43, 113. [117] E. Heinz, Biochim. Biophys. Acta 1971, 231, 537. [118] H. P. Wehrli, Y. Pomeranz, Chem. Phys. Lipids 1969, 3, 357. [ 1 191 P. J. Quinn, W. P. Williams, Biochim. Biophys. Acta 1983, 737, 223. [120] E. Heinz, H. P. Siebertz, M. Linscheid, Chem. Phys. Lipids 1979, 24, 265. [121] D. A. Mannock, A. P. R. Brain, W. P. Williams, Biochim. Biophys. Acta 1985,817, 289.
5
[I221 D. A. Mannock, R. N. A. H. Lewis, R. N. McElhaney, M. Akiyama, H. Yamada, D. C. Turner, S. M. Gruner, Biophys. J . 1992, 63, 1355. 11231 A. Eckert, B. Kohne, K. Praefcke, Z. Naturforsch. 1988,43b, 878. 11241 W. V. Dahlhoff, Z. Naturforsch. 1987,42b. 66 1. [I251 H. van Doren, T. J. Buma, R. M. Kellogg, H. Wynberg, J. Chem. Sac., Chem. Commun. 1988,460. [I261 K. Praefcke, A.-M. Levelut, B. Kohne, A. Eckert, Liq. Cryst. 1989, 6, 263. [ 1271 H. A. van Doren, R. van der Geest, C. A. Keuning, R. M. Kellogg, H. Wynberg, Liq. CqJst. 1989, 5, 265. 11281 C. Vogel, U. Jeschke, V. Vill, H. Fischer, Liebigs Ann. Chem. 1992, 1 171. 11291 S. Fischer, H. Fischer, S. Diele, G. Pelzl, K. Jankowski, R. R. Schmidt, V. Vill, Liq. Cryst. 1994, 17. 8.55. [130] S. Chandrasekhar, B. K. Sadashiva, K. A. Suresh, Pramaiia 1977, 9,47 1. [I311 H. J.Backer,S. vanderBaan,Recl. Trav. Chim. Pays-Bas 1937,56, 1 161. 11321 S. Chandrasekhar, B. K. Sadashiva, K. A. Suresh, N. V. Madhusudana, S. Kumar, R. Shashidhar, G. Venkatesh, J. Phys. (Paris) 1979, 40, C3-120. [133] B. Kohne, K. Praefcke, Angew. Chem. 1984, 96,70; Angew. Chem., Int. Ed. Engl. 1984,23, 82. [134] B. Kohne, K. Praefcke, Chem.-Ztg. 1985, 109, 121. 113.51 B. Kohne, K. Praefcke, J. Billard, Z. Nliturforsch. 1986, 41b, 1036. 11361 K. Praefcke, B. Kohne, W. Stephan, P. Marquardt, Chimia 1989, 43, 380. [137] K. Praefcke, B. Kohne, P. Psaras, J. Hempel, J . Carbohydr. Chem. 1991,10,523. [I381 R. G. Zimmermann, G. B. Jameson, R. G. Weiss, G. Demailly, Mol. Cryst. Liq. Cryst. (Lett.) 1985, I , 183. [139] N. L. Morris, R. G. Zimmermann, G. B. Jameson, A. W. Dalziel, P. M. Reuss, R. G. Weiss, J. Am. Chem. Soc. 1988,110,2177. [I401 V. Vill, J. Thiem, Liq. Cryst. 1991, 9,451. [I411 V. Vill, F. Fischer, J. Thiem, Z. Naturforsch. 1988,43a, 1119. [ 1421 W. Poules, K. Praefcke, Chem.-Ztg. 1983,107, 310. [ 1431 K. Praefcke, W. Poules, B. Scheuble, R. Poupko, Z. Luz, Z. Naturforsch. 1984, 39b, 9.50. 11441 N. Spielberg, Z. Luz, R. Poupko, K. Praefcke, B. Kohne, J. Pickardt, K. Horn, Z. Natiirfursch. 1986, 41a, 85.5; M. Sarkar, N. Spielberg, K. Praefcke, H. Zimmermann, Mol. Crysf.Liq. Cryst. 1991, 203, 159. 11451 K. Praefcke, D. Blunk, N. Usol’tseva, unpublished studies.
References
339
11461 K. Praefcke, D. Singer, B. Gundogan, Mol. Cryst. Liq. C q s t . 1992, 223, 18 1. 11471 N. Usol’tseva, K. Praefcke, D. Singer, B. Gundogan, Liq. Cryst. 1994, 16, 601, [I481 N. Usol’tseva, K. Praefcke, D. Singer, B. Gundogan, Mol. Muter. 1994,4, 253. [I491 K. Praefcke, B. Bilgin, N. Usol’tseva, B. Heinrich, D. Guillon, J. Muter: Chem. 1995,5,2257. [I501 N. Usol’tseva, G. Hauck, H. D. Koswig, K. Praefcke, B. Heinrich, Liq. Cryst. 1996,20,731. [I511 K. Praefcke, J. Holbrey, N. Usol’tseva, D. Blunk, Mol. Cryst. Liq. Cryst. 1997, 292, 123; see also the Festschrift in honor of Prof. A. Saupe, editors P. E. Cladis and P. Palffy-Muhoray, in press. 11521 K. Praefcke, J. D. Holbrey, J. Incl. Phen. Mol. Recogn. Chem. 1996, 24, 19. (1531 N. Usol’tseva, K. Praefcke, D. Singer, B. Giindogan, Liq. Cryst. 1994, 16, 617. [I541 N. Usol’tseva, K. Praefcke, P. Espinet, D. Blunk, J. Baena, poster B2P.21 at the 16th International Liquid Crystal Conference, Kent, USA, June 24-28, 1996, see the abstract volume, p. P-69. 115.51 K. Praefcke, J. D. Holbrey, N. Usol’tseva, D. Blunk, Mol. Cryst. Liq. Cryst. 1997, 292, 123. [156] K. Praefcke, B. Bilgin, J. Pickardt, M. Borowski, Chem. Ber. 1994,127, 1543. [IS71 D. Singer, A. Liebmann, K. Praefcke, J. H. Wendorff, Liq. Cryst. 1993, 14, 785. [ 1581 K. Praefcke, J. D. Holbrey, N. Usol’tseva, Mol. C v s t . Liq. Cryst. 1996, 288, 189. [I591 J. Barbera, P. Espinet, E. Lalinde, M. Marcos, J. L. Serrano, Liq. Cryst. 1987, 2, 833. [I601 S . A. Hudson, P. M. Maitlis, Chem. Rev. 1993, 93, 861 (p. 882). 11611 F. Tournilhac, L. M. Blinov, J. Simon, S . V. Yablonskii, Nature 1992, 359, 621. [I621 F. G. Tournilhac, L. Bosio, J. Simon, L. M. Blinov, S. V. Yablonsky, Liq. Cryst. 1993,14,405. [163] F. Tournilhac, L. M. Blinov, J. Simon, D. B. Subachius, S. V. Yablonsky, Synth. Metals 1993, 54, 253. [ 1641 P. Bassoul, J. Simon, New J. Chem. 1996, 20, 1131. [I651 P. Kromm, M. Cotrait, J. C. Rouillon, P. Barois, H. T. Nguyen, Liq. Cryst. 1996, 21, 121. [I661 F. Hildebrandt, J. A. Schrotcr, C. Tschierske, R. Festag, R. Kleppinger, J. H. Wendorff, Arigew. Chem. 1995, 107, 1780; Angew. Chem., Int. Ed. Engl. 1995, 34, 1631. [167] H. Finkelmann, M. A. Schafheutle, Colloid Polym. Sci. 1986, 264, 786. [168] Z. Ali-Adib. A. Bomben, F. Davis, P. Hodge, P. Tundo, L. Valli, J. Mater. Chern. 1996,6, 15. [I691 V. Vill, H. Kelkenberg, J. Thiem, Liq. Cryst. 1992, 11, 459. 11701 K. Herrmann, Trans. Faraduy SOC. 1933, 29, 972.
340
VI
Amphotropic Liquid Crystals
[171] D. Joachimi, C. Tschierske, H. Muller, J. H. Wendorff, L. Schneider, R. Kleppinger, Angew. Chem. 1993, 105, 1205; Angew. Chem., Int. Ed. Engl. 1993,32, 1165. [172] E. B. Sirota, G. S . Smith, C. R. Safinya, R. J. Plano, N. A. Clark, Science 1988, 242, 1406. [173] P. Marquardt, K. Praefcke, B. Kohne, W. Stephan, Chem. Ber 1991,124,2265. [ 1741 K. Praefcke, P. Marquardt, B. Kohne, W. Stephan, A.-M. Levelut, E. Wachtel, Mol. Cryst. Liq. Cryst. 1991,203, 149. [175] H. Fischer, V. Vill, C. Vogel, U. Jeschke, Liq. Cryst. 1993, 15, 733. [176] C. Tschierske, J . A . Schroter, N. Lindner, C. Sauer, S. Diele, R. Festag, M. Wittenberg, J.-H. Wendorff, Lecture 0-36 at the European Conference on Liquid Crystuls, Science and Technology, Zakopane, Poland, March 3-8, 1997; see the abstract volume, p. 61. [I771 K. Praefcke, B. Kohne, B. Gundogan, D. Singer, D. Demus, S. Diele, G. Pelzl, U. Bakowsky, Mol. Cryst. Liq. Cryst. 1991, 198, 393.
[178] S. M. Fan, I. D. Fletcher, B. Gundogan, N. J. Heaton, G. Kothe, G. R. Luckhurst, K. Praefcke, Chern. Phys. (Lett.) 1993, 204, 517. [ 1791 S . Chandrasekhar, Mol. Cryst. Liq. Cryst. 1994, 243, 1. [180] J. M. Goetz, G. L. Hoatson, Liq. Cryst. 1994, 17, 31. [ 1811 J. S . Patel, K. Praefcke, D. Singer, M. Langner, Appl. Phys. 1995, B60,469. [I821 With regard to discussions of nematic Schlieren pattern and possible two- or four-brush disclinations considered a diagnostic sign for the axialiality of nematic phases see S. Chandrasekhar, Geetha G. Nair, K. Praefcke, D. Singer, Mol. Cryst. Liq. Cryst. 1996, 288, 7 ; S. Chandrasekhar, Geetha G. Nair, D. S. Shankar Rao, S. Krishna Prasad, K. Praefcke, D. Blunk, Liq. Cryst. 1997, and Mol. Cryst. Liq. Cryst. 1997, both in press. 11831 N. Usol’tseva, K. Praefcke, D. Blunk, Lecture 0 - 6 0 at the European Conference on Liquid Crystals, Science and Technology, Zakopane, Poland, March 3-8,1997; see the abstract volume, p. 85.
Handbook ofLiquid Crystals D. Demus, J. Goodby, G. W. Gray, H.-W. Spiess, V. Vil Copyright 0WILEY-VCH Verlag GmbH. 1998
Chapter VII Lyotropic Surfactant Liquid Crystals Claire E. Fairhurst, Stuart Fuller, Jason Gray, Michael C. Holmes, Gordon J. 7: Tiddy
1 Introduction
the chain. The head group conveys watersolubility, while the hydrophobic chain drives the formation of self-assembled aggregates that are the subject of this article. A wide variety of structures can act as head groups. They can be ionic (anionic or cationic; single or multiple charges), zwitterionic, or nonionic or any mixture of these. The chain can be one or more alkyl groups (by far the most common), a perfluorocarbon group, or a polydimethyl siloxane. Examples of types of surfactant together with their commercial use are shown in Table 1.
Surfactants, or surface active agents, occur widely in nature where they are usually classified as lipids. They have been used for over 1000 years in everyday applications as emulsifiers in cleaning and in foods. Their functionality derives from a unique feature in the molecular structure: they are molecules that consist of two distinctly different regions. One is a polar (hydrophilic) moiety termed the head group, and the other a nonpolar (hydrophobic) part, referred to as
Table 1. Common surfactants and their applications. Name Sodium dodecyl sulfate Sodium dodecanoate (laurate)
Structure
Application
CH,(CH2),,OSO3Na
Detergents, emulsifier
CH,(CH,),&OJ'Ja
Soap Bars
Hexa-ethylene glycol monododecyl ether
CH,(CH,),,(OCH,CH,),OH
Detergents, emulsifier
1-octadecanoyl-sn-glycerol(monostearin)
CH,(CH,) ,,CO,CH,CH(OH)CH,OH
Food emulsifier
Hexadecyltrimethyl ammonium chloride
CH,(CH,),sN'(CH,),Cl-
Dioctadecyldimethyl ammonium chloride
CH3(CH2)17\ +/CH3 C1-
Hair conditioner Fabric conditioner
CH3(CH2),7/N\CH, Dodecyl dimethyl amine oxide
CH~CHZ)IIS]+(CH,)~
Speciality surfactant
0-
1,2-dioctadecyl-sn-glycero-3 phosphatidylcholine (distearoyl lecithin)
CH~(CHZ)~&O~CH~
I
Food emulsifier, mcmbrane lipid
CH~(CH,)I,CO~CH 0 I1
CH2 -P-OpCH,CH,N(CH,), 0
Perfluorooctanoic acid Octaethylene glycol mono tri siloxyl propyl ether
CF,(CF,),CO,H ((CH,),SiO),Si(Me)(CH2),(0CH,CH,),OH
Speciality surfactant Cosmetics, wetting agent
342
VII
Lyotropic Surfactant Liquid Crystals
When water-miscible surfactants are dissolved in water they form aggregates termed micelles above a well defined concentration. These aggregates are the building blocks of the liquid crystalline phases that occur at still higher concentrations. An understanding of micelle formation and micelle properties provides a good basis for a description of the liquid crystals, hence we begin with a discussion of this topic. This is followed by sections describing the various liquid crystal phases, the way that the mesophase type varies with surfactant chemical structure, and what happens with mixed surfactants. We also include the effects of additives such as electrolytes and other solutes. Because of the enormous number of publications in this area, space considerations, and the limited time of the authors, it simply is not possible to document all the available information here. Instead, it is our intention to give an introduction to the concepts and phenomena that underlie the formation of surfactant mesophases and to summarize the present knowledge of the area. As well as providing an up-to-date description our objective is to provide sufficient information so that anyone with a practical problem concerning mesophase formation or structure can see the general context. This should allow them to find some indication of experiments that may provide a solution. As a further comment we draw attention to several books [ 1-81 which provide a good general introduction to the area of surfactants and colloid science. Most of the books have been published in the last few years and summarize recent research.
2 Surfactant Solutions: Micelles When surfactants dissolve in water at low concentrations they exist as monomers (ionic surfactants are dissociated). As the concentration is increased aggregates termed micelles are formed. These appear at a welldefined concentration known as the ‘critical micelle concentration’ (CMC). This is not a critical point in the sense of modern physics since micelle formation occurs over a very narrow range of concentrations. This range is so small that for almost all practical purposes it can be represented by a specific value, the CMC. For pure single surfactants below the CMC, all of the dissolved surfactant exists as monomers, while above the CMC all added surfactant forms micelles. With surfactant mixtures the phenomenon is more complex because the components usually have different individual CMCs, but the same considerations apply PI. Micelles are aggregates of at least 15-20 monomers. Typical surfactants such as SDS (sodium dodecyl sulfate) or C,,EO, (Table 1) form micelles with aggregation numbers ( 4 ) in the range 50-100. The aggregation numbers can become very large up to lo4 or more, according to surfactant type, aggregate shape, temperature, and surfactant concentration (see below). Micelle formation arises from the hydrophobic effect [4,9-131. This is the term used to describe the interaction between nonpolar solutes and water. It is well known that nonpolar solutes are almost insoluble in water, with the limited degree of solubility decreasing rapidly with increasing solute size. A thermodynamic analysis of the process shows that the introduction of a hydrocarbon into water at ambient temperature is always associated with a decrease of entropy,
2
and an enthalpy of about zero, resulting in a large and positive free energy [ 10- 141. ‘Structuring’ of water molecules in the neighborhood of the alkane, akin to the formation of clathrate hydrates [15], is frequently cited as the origin of the entropy change. Unfortunately, attempts to locate the ‘clathrate structure’, for example by self-diffusion measurements of water and organic solutes such as tetraalkyl ammonium ions [16], have consistently failed to show that more than a single layer of water around the solute differs from bulk water: hardly evidence to support the ‘water structure’ concept. In a series of recent papers Kronberg et al. [ 10-121 have discussed this problem in detail. They view the hydrophobic effect as the resultant of two contributions, one arising from the ‘ordering’ of water molecules around the solute, the second from the energy required to make a cavity in the water large enough to accommodate the nonpolar solute. The first contribution is associated with a negative entropy because water molecules next to a nonpolar solute have fewer conformations available than ‘free’ water: they can not H-bond to the solute. It also gives a negative enthalpy, presumably because vicinal water molecules make stronger H-bonds. The second and opposite contribution arises from the large energy required to form a cavity to accommodate the solute. This is large, owing to both the high cohesion in water arising from H-bonding connectivity, and the small size of water molecules compared to, for example, alkanes. An important consequence of this mechanism is that the magnitude of the hydrophobic effect is proportional to the area of hydrophobic contact between water and the solute. As will be seen below, using this concept it is possible to estimate an approximate CMC for almost any novel surfactant, given that a few simple guidelines are followed.
Surfactant Solutions: Micelles
343
Micelles can only form when the surfactant solubility is equal to or greater than the CMC. In general this occurs only above a particular temperature known as the Krafft point (temperature). Below this temperature surfactant solubility increases slowly with increasing temperature because the surfactant dissolves as monomers. The limit to monomer solubility occurs when the chemical potential of the monomers is equal to that of the pure (usually crystalline) surfactant. Above this temperature the solubility increases very rapidly because the surfactant dissolves as micelles: the contribution of each micelle to the surfactant chemical potential being the same as that of a monomer. Micelle formation involves a dynamic equilibrium between monomers and aggregates represented by Eq. (1) [17-211: Sl
as,
(1)
(where S, represents an aggregate of n monomers). In fact micelles form by a series of step-wise reactions from monomers, Eq. (2).
Clearly, there is a range of aggregate sizes present in solution, with an average value of q. However, the size distribution is usually narrow (say 2 10%) particularly for globular micelles. Although the micelle formation process involves dimers, trimers, etc., the actual concentration of these species is very, very small. Hence, once formed, micelles have a reasonably large, but finite, lifetime. For dilute solutions, using fast relaxation techniques it is possible to measure both the exchange rate between monomers and micelles, and the rate of micelle forma-
344
VII
Lyotropic Surfactant Liquid Crystals
tionhreakdown. Both of these processes are related to the CMC, becoming faster as the CMC increases. Typically, the monomermicelle exchange rate is in the range 103-106 s-’, while micelle breakdownlformation rates are about lo-’ - 1O2 s-l (i.e. micelle lifetimes of lop2- 10 s). Note that from the kinetic analysis of fast relaxation studies an upper limit of mol dm-3 can be placed on the concentration of dimers, trimers etc. (i.e. aggregates with association numbers much smaller than q ) [21, 221. Figure 1 shows a schematic diagram of a small globular micelle. The rapid, continuous exchange of monomers with bulk solution means that the micelle is a very mobile, disordered, aggregate. There is a continuous movement of molecules jumping part way in and out of the interface [21] (this has recently been termed ‘protrusion’ [23,24]). Also, diffusion around the micelle surface is rapid [self-diffusion coefficient (D)=lO-ln m2 s-’1, as is the exchange between the various possible conformations of the alkyl chain (correlation time (7,) lo-’’ s). With an ionic surfactant, typically about 70-80% of counter-ions reside close (within =1 nm) to the micelle surface in a loosely ‘bound’ state due to the very high surface charge density. The exchange
-
Figure 1. Schematic representation of a spherical micelle.
between ‘bound’ and free counter-ions occurs on a time scale of =lop9s. There also exists a single layer of ‘bound’ water molecules associated with polar groups and ions. Again, the exchange between bound and free water occurs on a time scale of about lop9 s [25]. We have discussed the formation of micelles at a ‘critical micelle concentration’, and the absence of significant concentrations of small aggregates. In fact, micelle formation occurs over a very narrow range of concentrations, the range being too narrow to detect for all but the shortest chain (highest CMC) surfactants. The almost complete absence of small aggregates at concentrations above the CMC makes their occurrence below this highly unlikely. There are numerous claims in the literature of ‘pre-CMC’ surfactant aggregates formed due to a contribution from the hydrophobic effect. To date, none of these have withstood a thorough examination of the evidence. Usually, the deviations from expected behavior are due to the presence of impurities. A good illustration of how minor constituents can influence surfactant properties is the recent study of Thomas and Penfold who demonstrate [26] that doubts concerning the application of the Gibbs adsorption equation apply to the systems (i.e. the presence of impurities, which should be included in the equation) rather than the thermodynamics. Arguably the most important parameters for any surfactant is the CMC value. This is because below this concentration the monomer level increases as more is dissolved hence the surfactant chemical potential (activity) also increases. Above the CMC the monomer concentration and surfactant chemical potential are approximately constant, so surfactant absorption at interfaces and interfacial tensions show only small changes with composition under most
2
conditions. For liquid crystal researchers the CMC is the concentration at which the building blocks (micelles) of soluble surfactant mesophases appear. Moreover, with partially soluble surfactants it is the concentration above which a liquid crystal dispersion in water appears. Fortunately there are well established simple rules which describe how CMC values vary with chain length for linear, monoalkyl surfactants. From these, and a library of measured CMC values [27-291, it is possible to estimate the approximate CMC for branched alkyl chain and di- (or multi-) alkyl surfactants. Thus most materials are covered. This includes the ‘gemini’ surfactants, a ‘new’ fushionable group where two conventional surfactant molecules are linked by hydrophobic spacer of variable length [30]. The major determinant of the CMC is the hydrophobic effect, which is proportional to the area of the nonpolar chain exposed to water (see above) [lo-121. This leads to a logarithmic relationship between the CMC and alkyl chain length for linear surfactants: log(CMC) = A n + B
(3)
(where n=alkyl chain length; A , B are constants). The second factor is the valency ( s )of the head group charge (s = 0, 1, 2 ... for nonionic/zwitterionic, monoionic, di-ionic .. . etc.). According to the value of s, the constant A takes the values: A=0.5 (s=O), A=0.3 ( s = l ) , A=1.5-1.8 (s=2). Finally, the valency of the counterions is also important as this influences the value of B. Essentially, the CMC is reduced for multivalent counterions because fewer ions are required close to the micelle surface to (partially) balance the high surface charge density. Table 2 lists examples of CMC values for various surfactants. While there is some temperature dependence of the CMC [9- 121 with many materials showing a shallow
345
Surfactant Solutions: Micelles
Table 2. Typical CMC values from [2, 27-29] Surfactant
Temp. “C
CMC mol d m 8
C14E08
25 20 25 25 25
9 . 9 lo-’ ~ 9.5~ 6.8~10.~ 7.1 x 9.0 x 10-,
C,SO,Na CI W 4 N a C,,SO,Na
25 25 25
1.3x10-2 8.3 x lo-’ 2 1x ~ 0 - 4
CxNMe,Br C I ,NMe,Br C I ,NMe,Br C I ,NMe3C1
25 25 25 25
1 . 4 10-1 ~ 1 . 4 lo-’ ~ 3.3 x 1 0-4 1.7x10-2
C , ,NMe,O C I ,NMe,CH,SO,
27 25
2.1 x 103 . 6 lo-’ ~
unknown unknown unknown
1 . 6 lo-* ~ 1 . 6 lo-’ ~
70
3.3 x lo-’
CXEO,
c I OEO,
c I,EO, c I m,
di C, lecithin di C, lecithin
(C 1 ,),NMe,Cl C1,SWa
I .OX
minimum, the effect is small below 100 “C. It is clear that for nonionic and zwitterionic surfactants the CMC values reduce by about a factor of 10 for the addition of two CH, groups to the alkyl chain, while for monovalent ionic surfactants the factor is 4. Branched and multichain surfactants can be treated by estimating the equivalent linear chain having the same area of hydrophobic contact with water. There is a free energy penalty for constraining the conformations of multichain surfactants at the micelle surface [31, 321 which leads to a higher than expected CMC. This effect is small: equivalent to the loss of about one CH, group 1321. With mixed surfactants the CMC of the mixed micelle varies according to the CMCs of the individual surfactants, and their concentrations. Clearly, micelle composition varies with concentration since the micelles that form at the lowest concentration are rich in the lowest CMC surfactant, while the
346
VII
Lyotropic Surfactant Liquid Crystals
higher CMC materials become more abundant in micelles as the overall concentration is increased. The detailed dependence of CMC values on mixed surfactant composition varies according to whether there are specific interactions between head groups which lead to nonideal mixing in the micelle. This applies particularly with mixtures of nonionic and ionic surfactants, and ionic surfactants of opposite charge. Various treatments are available to describe the behavior (which are outside the scope of this article), for example as outlined in Clint [2], Ch 5 and 6. Given that micelles are present in solution above the CMC, the most important consideration for those concerned with liquid crystal phases is the micelle shape. There are three major types: spheres, rods and ‘discs’. They can be described using the packing constraint concepts [4, 331. These give a simple description of the relationship between micelle shape and molecular shape. The micelles are assumed to be smooth, with only the hydrophobic volume in the micelle interior. The main molecular parameters are the hydrophobic group volume, usually taken as being equal to the alkyl chain volume (v), the area that the molecule occupies at the micelle surface ( a ) and the maximum length of the alkyl chain (taken as the alltrans length, 1,). For a spherical micelle having a hydrophobic volume ( V )with radius r, a total surface area A and aggregation number q: A = q a = 4 r c r2
v = q v = -4n r 3 3
Hence
a = 3 -V r
(4)
Ignoring end or edge effects for circular cylinders (radius = r ) and bilayer/disc (thick-
ness = 2 r ) shapes, the equivalent relationship are:
a=2’
a=-
r
’r
(rod)
(5)
(disc)
(6)
The value of r can not be larger than I , , hence there are limitations on the lowest value of a for a given shape:
a 2 3 2 (sphere); r
a 2 2 2 (rod); r V a 2(disc)
(7)
r
Clearly, a surfactant with a given chain length can pack into spheres, rods or discs according to the size of the head group, with all three shapes being possible for the largest a values and only disc micelles for small a values. Entropy favors the formation of the smallest possible aggregate at the CMC (i.e. spheres over rods and rods over discs). The present authors are not aware of any exceptions to this. Thus large headgroup surfactants form spherical micelles, smaller headgroups give rods, and smaller headgroups still give discs. (Because of the flexibility of alkyl chains there does not appear to be a lower limit on the values of r for conventional surfactants, hence maximum values for a are not known.) It is a simple matter to estimate the volumes of hydrophobic groups from published density data for alkanes (normal and fluorinated) and polydimethyl siloxanes [4, 5 , 341. One simply sums the group volumes (see Table 3). Similarly, the maximum length of the hydrophobic group can be calculated from known bond lengths. Thus, the maximum micelle radius and hence the limiting a values for the various aggregates can be calculated. Note that there are small dif-
2
Surfactant Solutions: Micelles
347
Table 3. Molecular sizes of hydrophobic groups (25 “C). Hydrophobic group
Fragment volumes
(A)
Bond length (A)
Pol ydimeth yl siloxane
123
2.4 Limiting CL values [nm2 x 10-*]
Alkane Fluorocarbon Polydimethylsiloxane
Sphere
Rod
68 102 - 160
46 68 - 106
ferences between the parameters of Table 3 and those of other authors [4-61. These are unimportant since they result only in small differences between the limits to a values. Moreover, we recall the known roughness of the micelle surface (typically 0.20.3 nm), hence we emphasize that this model gives only an approximate description. An important consequence of the above model is that a simple increase in alkyl chain length should not alter micelle shape because both chain length and volume increase by a constant increment. However, in practice it is often observed that short chain (e.g., C ,*) surfactants form globular (‘spherical’) micelles while higher chain length materials with the same head group form long rod micelles. This probably arises from the influence of surface roughness on micelle shape and aggregation numbers. Micelle size (aggregation number) varies according to micelle shape and alkyl chain length. Spherical micelles always have low aggregation numbers, due to the micelle ra-
Disc
-
23 34 53
dius being limited by the all-trans alkyl chain length, hence the surface area is limited. For rod and disc micelles, where the fraction of ‘end’ or ‘edge’ molecules plays a significant part in micelle size, it is useful to consider a thermodynamic description employed by Israelachvili and collaborators [6, 33, 351. In a solution of aggregates with a range of aggregation numbers at equilibrium, the chemical potential ( p ) of all identical molecules is the same, whatever the aggregation number = constant, p,, = p ~+ , “KT n log n n = 1,2,3, ...
where pn is the mean chemical potential of a molecule in aggregates of aggregation number n, p,s) is the standard part of the chemical potential (i.e. the mean interaction free energy per molecule) and X , the concentration of molecules for the n-aggregates. Taking simple models for the interaction free energy within aggregates of vari-
348
VII
Lyotropic Surfactant Liquid Crystals
ous shapes, it can be shown that:
(9) where p takes the value 1/3, 1/2 or 1 for spheres, discs or rods respectively. The monomer-monomer bond energy within an aggregate is described by a k T . Above the CMC, with a > 1 (a reasonable assumption) a n d p < 1 (i.e. for spheres and discs), this approach leads to expressions which predict vanishingly low concentrations of aggregates having n values which are not small (say n > 10-20). Thus spherical and disc micelles remain small (or increase to ‘infinite’ aggregation numbers) i.e. they phase-separate. Only rod micelles can have large aggregation numbers ( n > -100). In practice this does appear to be generally true. In any case we have already seen that spherical micelles are prevented from becoming large by the alkyl chain packing constraints. Hence the major conclusion of this approach is that disc micelles either remain small, or grow to infinite size to form a lamellar liquid crystalline phase (see below). In real surfactant systems the interactions are much more complex that the simple picture used above. However, the general formalism still holds, but observed aggregation numbers for small micelles are often larger than expected for spheres, while the micelles do not grow very large as expected for rods. This is almost certainly due to micelle surface roughness. A surface roughness of 1-2 C-C bonds allows the micelle radius to be slightly larger than the all-trans chain length: with a significant increase in n. The roughness also allows repulsive interactions between adjacent head groups to be relaxed by the formation of a thick interfacial layer rather than a smooth surface. In addition, shape fluctuations can occur, which can also lead to larger aggregation numbers. However, if the alkyl chain lengths is long
enough, the micelles that have a rod shape do become very long, up to 100+ nm. This is not usually seen for common ionic surfactants because the longer chain homologues have high Krafft temperatures, hence they are insoluble at normal temperatures.
3 Liquid Crystal Structures There are six classes of liquid crystal phases, most of which now have well-established structures. These are the lamellar, hexagonal, cubic, nematic, gel, and intermediate phases. All except the intermediate phases have been recognized for many years, hence a comprehensive literature review is outside the scope of this article. In all of the states except the gel phases, both surfactant and water have a liquid-like molecular mobility, that is short-range rotational and translational diffusion on a time scale of s. They differ in the long-range symmetry of the surfactant aggregates and in the curvature of the micelle surface. Except for the phases with flat aggregate surfaces, each class can occur with either the polar regions or the nonpolar regions as the continuous medium, the former being referred to as normal, while the latter are reversed. As will be seen later, the phases occur in a particular composition sequence, occupying a specific region of the phase diagram. Each class of mesophases is usually labelled by a particular letter (see below) with the symbols having subscripts 1 or 2 to distinguish between the normal or reversed forms. Unfortunately there is no universally accepted nomenclature as is the case with thermotropic mesophases. In this review we employ the system that we have used in previous papers [36-381. As with thermotropic mesophas-
3
Liquid Crystal Structures
349
es, the most important technique to identify the mesophase type is polarizing microscopy. The birefringent phases usually have typical textures, while cubic phases have none, but they are very viscous.
3.1 Lamellar Phase (La) By far the most common surfactant mesophase is the lamellar phase ( L a ) ,also known as the neat phase from its occurrence during soap manufacture. In this phase, the surfactant molecules are arranged in bilayers frequently extending over large distances (a micron or more), which are separated by water layers (Fig. 2). This phase is similar to the thermotropic SmA phases. Its major repeating unit, the bilayer, forms the basic structural matrix of biological membranes [5,7, 81. While the lamellar phase does not usually flow under gravity, it has a fairly low viscosity, with the material being easily shaken into a container. It is readily identified from its characteristic optical textures (Fig. 3). The surfactant bilayer thickness can vary from about 1 .O- 1.9 times the all-trans alkyl chain length (1,) of the surfactant. Within this layer the ‘fluid-like’ characteristic of
Figure 2. Schematic representation of a lamellar phase.
Figure 3. Typical optical textures of a lamellar phase: (a) mosaic oily streaks, (b) Maltese crosses.
the alkyl chains is shown by a diffuse wide angle X-ray diffraction peak corresponding to a Bragg reflection of 0.45 nm 1391. The difference in layer thickness arises from differences in head group areas and gives rise to differing degrees of disorder within the alkyl chain region. For the bilayers of thickness 1, = 1.O the disorder is large. Alternative suggestions that these lamellar phases consist of ‘interdigitated monolayers’ are incorrect. By contrast, the water layer thickness varies over a much larger range, usually within the limits 0.8 and 20+ nm. The water thickness is the same throughout the sample, except at very high water contents [36] where low energy fluctuations can occur. The minimum water content is often that required to hydrate the polar groups, but
350
VII
Lyotropic Surfactant Liquid Crystals
very low or zero water content can occur with surfactants that form thermotropic lamellar phases. Sharp reflections in the ratio d : d/2: d/3 ... are observed in the low angle X-ray region due to the regular alternating layer structure with repeat spacings being the sum of the water and the alkyl chain layer dimensions.
3.2 Hexagonal Phases The next most common mesophase type is the hexagonal phase. There are two distinct classes of hexagonal phase, these being a ‘normal hexagonal’ (HI) also known as the middle phase (again from the soap industry) and a ‘reversed hexagonal’ (H2). The normal phase (H,) is water-continuous, while the reversed (H2)is alkyl chain-continuous. They consist of indefinitely long circular aggregates packed on a hexagonal lattice (Fig. 4). The normal micelles have a diameter of 1.3-2.0 times the all-trans alkyl chain length, with a typical intermicellar separation being in the region of 0.8-5 nm. The reversed micelles have a polar region diameter in the same range, but values above 3 nm are rare.
Figure 5. Typical optical textures of a hexagonal phase: (a) fan-like, (b) nongeometric.
X-ray diffraction of both phases shows Bra g reflections in the ratio 1 : l/&: 1hh: 1/ 7 : 1 / f i ..., again with a diffuse reflection at 0.45 nm. Both phases are rather viscous, much more so than the L,phase. It is usually not possible to shake a sample into a container by hand. The optical textures are similar for both types, again being distinctly different from those of L,phase (Fig. 5).
P
3.3 Norma HexagonalH i
Reversed HexagonalH2
Figure 4. Schematic representation of a normal (HI) and reversed hexagonal (H2) phase.
Cubic Phases
A third category of mesophases formed by surfactants comprises the cubic phases. They are also known as ‘ v i ~ c o ~ ~ As the name implies, these phases are based
3 Liquid Crystal Structures
Figure 6. Schematic representations of a primitive, face-centered and body-centered cubic lattices.
around one of three cubic lattices, namely the primitive, face centred and body centred (see Fig. 6). Therefore, they have no optical texture under polarized light. It is still not certain which symmetries form for the different cubic phase, but the overall picture has become much clearer during the past few years [37, 39-47] with more and more cubic structures being identified. There are at least four classes of cubic phase, these being the normal and reversed forms of two very different structures. One set of structures (I) is comprised of small globular micelles while the second (V) consists of a 3-D micellar network. Within each class several different structures occur. The simplest cubic phase is that of the I type where the surfactant aggregates are small globular micelles. For the water-continuous I, phases, primitive, body centred and face centred lattices [41] have all been proposed. Pm3n, Im3m and Fm3m lattice structures have been proposed [46,47] (see Fig. 8). In the nonionic C1,EO,,/water system [46] all three symmetries are reported. Here the micelles have diameters similar to normal micelles with a separation similar to that found in the H, phase. The struc-
(a)
(C)
35 1
ture of the Pm3n symmetry has been the attention of much debate [45-491. However, it is now thought that there are two different sizes of micelle present, one being slightly larger than the other [45]. Whether these micelles are short rods or flattened spheres is still a matter for debate. The short rod model fits better with their position in the phase diagram (between L, and H,). However, a model composed of two spherical micelles and six disc shaped micelles has been proposed [49]. It is this structure which is now thought to be the more accurate. As for the Im3m and Fm3m cubic lattices only one micelle type, a quasispherical structure, is proposed for the lattice [49] (see Fig. 7). A reversed I, structure of globular aggregates packed in a cubic array also has been reported recently for some surfactants [46-491, despite previous doubts about their existence. Here it appears that the micelles are spherical, but of two different sizes (Fig. 8). For reversed micelles the alkyl chain packing constraints no longer limit the micelle diameter, hence the coexistence of two spherical micelles of different sizes is more plausible than with the I, phases. The Fd3m phase is well established now [48, 491, but again, there appear to be several other distinct symmetries possible in the I, region [48]. It is likely that the confusion in this area will be resolved in the next few years as the surfactant structures necessary for both I, and I, phases are now clear, hence we can obtain the phases with many more surfactant types.
Figure 7. Polyhedral representations showing the micellar arrangements for (a) Pm3n, (b) Im3m and (c) Fm3m of I, cubic phases (reproduced from [46]). Micelles are located at the centres of the polyhedra
352
VII ,-,
Lyotropic Surfactant Liquid Crystals
Fd%
-
Figure 8. Schematic representation of the Fd3m I, phase (reproduced from [42]).
The second set of cubic phases has a ‘bicontinuous’ aggregate structure. The basic structures for these bicontinuous cubics have received more attention in previous years, with the structures being well established. With the three main spacegroups Pn3m, Im3m and Ia3d being well reported. The aggregates form a 3-D network extending throughout the sample, having curvature towards water (V,) or towards oil (V,), with many different structures being reported [41-441. These phases are structurally similar to the intermediate phases (see below) and although organized around the same cubic lattices as the I phases, have a completely different aggregate structure. The first one to be detailed was based on a body centred lattice, namely Ia3d [50] (Fig. 9). This
la3d
structure, originally proposed by Luzzati [50]was thought to be composed of rod-like aggregates joined three by three to form two independent networks. However, it is now believed that these structures are composed of infinite periodic minimal surfaces [5 I ] with the surfactant and water cylinders being interwoven and connected three by three. Another body centred cubic structure proposed for lipid/water systems [52], that of the Im3m (Fig. 9), is composed of a network of water channels connected six by six. There are two reported cubic phases composed around a primitive lattice. The first is the Pn3m, which first proposed to be composed of rod-like aggregates connected four by four at tetrahedral angles forming two independent diamond lattices. However, it is now believed to be composed around a minimal surface of two interwoven tetrahedral networks arranged on a double diamond structure. The second structure composed around a primitive lattice is that of the Pm3n. Its structure is still under debate, Luzzati and Tardieu [52] proposed a rodlike network of micelles forming a cage in which a spherical micelle is enclosed. However, this is now thought to be unlikely [43, 531. The cubic phases are all optically isotropic (as are micellar solutions), but they have
Pn3m
Figure 9. Schematic representations of cubic phases (reproduced from 1421).
lm3m
3
very high viscosities hence are easily distinguished from micellar solutions. The two classes of cubic phase, I and V, are distinguished from each other by their location in the phase diagram. I phases occur at compositions between micellar solutions and hexagonal phases, whilst V phases occur between hexagonal and lamellar phases. The factors that determine the particular structure within any set of I or V phases are not understood.
3.4
Nematic Phases
Lyotropic nematic phases were first reported by Lawson and Flautt [54] for mixtures of C, and C,, alkyl sulfates together with their corresponding alcohol in water. They are some what less common than the mesophases discussed so far. When they do form they occur at the boundary between an isotropic micellar phase (L,) and the hexagonal phase ( H I )or between L, and the lamellar phase (L,). As their name implies they have a similar long range micellar order to that of the molecules in a thermotropic nematic phase. They are of low viscosity, possessing long range micellar orientational order but reduced translational order com-
Liquid Crystal Structures
353
pared to the other lyotropic phases described above, and like the thermotropic phase can be aligned in a magnetic field. It is possible to identify nematic phases optically because of their characteristic schlieren texture. Lyotropic nematic phases are generally found for short chain surfactants, for both hydrocarbon or fluorocarbon derivatives [55, 561. Two different micelle shapes can occur (Fig. 10) [57]. One type (N,) is thought to be composed of small cylindrical micelles and is related to the hexagonal phase, while the other type of nematic (Nd) is composed of planar disc micelles and is related to the lamellar phase. Note that the ‘disc’ micelles are likely to be ‘matchbox’ or ‘ruler shaped’, rather than the circular discs. Hence the ‘disc’ nematic phase can have the director along the long axis of ‘ruler’ micelles or along the shortest micelle dimension, as with ‘match box’ micelles, while with N, phases the director always lies along the rod axis. As with thermotropic nematics, the addition of optically active species to lyotropic nematic phases gives lyotropic cholesteric phases. Whilst details of their structures are not fully established they appear to follow the general pattern outlined above. The cholesteric ‘twist’ would appear to derive from the packing of optically active mole-
Nd phase Figure 10. Schematic representation of a disc (Nd) and rod (N,) nematic phase.
N, phase
354
VII
Lyotropic Surfactant Liquid Crystals
cules within micelles, leading to a twisted micellar structure, rather than to the transmission of the anisotropic forces via solvent mediated forces. A recent report has shown evidence for the occurrence of cholesteric ‘blue’ phases, a remarkable observation ~581.
3.5 Gel Phases (Lp) The gel phase (Lp) closely resembles the lamellar phase in that it is comprised of surfactant layers, but it differs in its very high viscosity. The term ‘gel’ again originates from industry where these systems were observed to have a gel-like rheology. However, these states should not be confused with polymer gels or gels formed by hydrocolloid systems, since they are single phases in terms of the phase rule rather than being multiphase systems like polymer and colloid gels. Within the gel phase, the bilayers have rigid, mostly all-trans alkyl chains, as
shown by a sharp, wide angle X-ray spacing of about 0.42 nm and a large transition heat on melting, typically 25-75% of the crystalline surfactant melting transition. This indicates restricted chain motions, mostly limited to rotation about the long axis only. By contrast the water (polar medium) is in a ‘liquid-like’ state, with fast rotational and translational mobility. There are commonly reported to be three different structures of the gel phase as shown in Fig. 11. The first structure, with the bilayer normal to the liquid crystal axis, is the structure most commonly found in dialkyl lipid systems [59]. Here, the alkyl layer thickness is found to be approximately twice the alltrans alkyl chain length of the surfactant. The second structure shown, the tilted bilayer, is found in systems where the polar head group is larger than the width of the alkyl chain. This structure has been reported for monoglyceride systems [60]. The third structure, the interdigitated form, is found with long chain mono-alkyl systems such as potassium stearate [61].
Water
Water
Water Water
Normal
Tilted
Figure 11. Schematic representation of the possible gel phases.
Interdigitated
355
3 Liquid Crystal Structures
Whilst the occurrence of gel phases is commonly recognized for long chain dialkyl surfactants it is also a common occurrence for monoalkyl surfactants. Because of the lower packing order within the alkyl chain region in than the normal crystals, different chain length derivatives (usually up to four carbons) can mix within the L p phase. Additionally, different head groups can also mix within the Lp phases. Indeed, Lp phases can occur for mixed systems where none is observed for the individual constituents. For example sodium dodecyl sulfate and dodecanol form a mixed gel phase where as none is observed for SDS alone [62]. (Note that dodecanol does form a stable Lp phase termed the ‘a-crystalline phase’ with about 0.2 mol fraction of water [ 5 , 63, 641.) Whilst Lp phases have been accepted for years recently there has been debate about whether the state really does exist as a true thermodynamic equilibrium phase, based on very reasonable criticism of deficiencies in their location on properly determined phase diagrams [65]. However, in at least one case (the nonionic surfactant trioxyethylene hexadecyl ether [66]) the Lp phase in water melts at a higher temperature than the crystalline surfactant. It forms on mixing water and the liquid surfactant just above the crystalline surfactant melting point, clear proof that it is the equilibrium state. In fact, as Small has discussed in detail [ 5 ] ,the stability of the Lp state is determined by the packing of the alkyl chains. Polyethylene does not melt until ca. 140°C [67]. Long chain alkanes do form stable ‘a-crystalline’ (rotator) phases where the alkyl chain packing and mobility is similar to that of the Lp phase. The melting of alkanes can be regarded as being driven by the mismatch between CH, and CH, sizes within the crystal (2.3-2.6 nm and ca. 120 A respectively). Thus the transition tempera-
tures for the sequence: crystal
TA +
‘rotator’ phase
T8 -+
melt
increase with hydrocarbon chain length, the minimum chain size required to form a rotator phase being about C22.The value of Ts for alkanes of chain length 2 n represents an upper temperature limit for the Lp phase of surfactants with chain length n. Usually, the Lp phases melt at a lower temperature than the limit TB because the head groups are more hydrated in the molten phases than the Lp phase, and this free energy contribution is larger than the chain packing energy. To date the hydration of headgroups in the Lp phase has always been found to be lower than that of the higher temperature molten phases (usually lamellar). Hence Lp phases do not swell in water to the same extent as the La phases. Moreover, the size (area) of the hydrated head group determines which of the three structures occur (see Fig. 11). The perpendicular bilayer requires a head group area ( a ) of about 22 A2, whilst the tilted bilayer occurs with a =22-40 A2, and the monolayer interdigitated structure with a 2 4 4 A2. Hence increasing m in the series of polyoxyethylene surfactants C,EO, ( n > 16, rn=0-3), either singly or in mixtures, one expects to encounter all three phases. Whilst the monolayer and perpendicular bilayer structures are known, the tilted phase has not been reported. In fact, a careful examination of properties such as heat of melting, high angle Xray data and phase behavior for a series of closely related surfactants shows that considerable anomalies exist in both assumed alkyl chain packing structure and phase stability with the conventional picture. The area deserves a systematic broad study to give a proper molecular based understanding.
356
3.6
VII
Lyotropic Surfactant Liquid Crystals
Intermediate Phases
We have already described the occurrence of bicontinuous cubic phases having an aggregate curvature between those of hexagonal and lamellar phase. Over the past 40 years there have been sporadic reports of other structures with similar ‘intermediate’ curvature. A number of these so-called ‘intermediate’ phases have now been identified. They replace V, bicontinuous cubic phases for surfactants with longer or more rigid hydrophobic chains. It is likely that they replace V, phases under some conditions, but this area has yet to receive the same systematic attention as the V,/intermediate phases. Unlike the V,/V2 phases, intermediate phases are anisotropic in structure, and consequently birefringent; also they are often much more fluid than cubic phases. There is still discussion in the literature as to which structures are possible. The observed or proposed structures divide topologically into three broad types according to symmetry; rectangular ribbon structures, layered mesh structures, and bicontinuous structures which do not have cubic symmetry. Ribbon phases may be regarded as a distorted hexagonal phase (Fig. 12a). Mesh phases are distorted lamellar phases in
(4
which the continuous bilayers are broken by water filled defects which may or may not be correlated from one layer to the next (Fig. 12b). The bicontinuous phases are distorted cubic structures. Phases with noncubic structures were first identified by X-ray scattering from aqueous soap mixtures [68, 691 and anhydrous soap melts in a series of papers by Luzzati and Skoulios [69-731. In the first of these papers [68] the term intermediate was applied to a rectangular structure found in aqueous mixtures of potassium and sodium oleates and potassium laurate and palmitate. The structures found in the anhydrous soaps were reinterpreted as intermediate, tetragonal, rhombohedral, and ribbon structures in the paper by Luzzati and coworkers [74] in 1968. It was not until the early eighties that interest in these unusual phase structures was rekindled. Ribbon phases have been the most comprehensively studied of the intermediate phases. They occur when the surfactant molecules aggregate to form long flat ribbons with an aspect ratio of about 0.5 located on two dimensional lattices of oblique, rectangular (primitive or centred), or hexagonal symmetry. Ribbon phases were first proposed by Luzzati [68,69] in aqueous sur-
(b)
Figure 12. Schematic representations of (a) centered rectangular phase, (b) representation of six connected rhombohedral mesh phase.
3 Liquid Crystal Structures
factant systems and by Skoulios [71] in sodium stearate and tetradecane or cyclohexane. Hagslatt et al. [7S] have investigated ribbon phases in a number of ternary systems. In [7S] they review the results from a wide variety of studies on ribbon phases and show that these studies are consistent with the conclusion that all ribbon phases index to a centred rectangular cell, cmm. A 'hexagonal-rod' model of the ribbon cross section was suggested in which the ribbon structure is controlled by the competition between the requirement for a constant water layer thickness around each ribbon, the surface area per molecule and the minimization of total surface area. Note that given the anisotropic nature of the interaction forces between the ribbons, the assumption of a constant water layer must be an approximation. The first identification of intermediate mesh phase structures was by Luzzati [74] from the measurements by Spegt and Skoulios [70-731 in anhydrous soap melts. It was not until the work of KCkicheff and others [62, 76-81] on sodium dodecyl sulfate (SDS)/water and on lithium perfluoroocta-
357
noate (LiPFO)/water [82] that intermediate mesh phases were recognized in aqueous surfactant water systems. SDS/water has been extensively studied by optical microscopy [78], X-ray scattering [68, 77, 781, SANS [79] and NMR [77, 781 techniques. More recently, mesh intermediate phase structures have been identified in a number of nonionic surfactants with long alkyl chains [83-851. These reveal a very rich intermediate behavior [77-801. There are a number of possible mesh phase structures with both tetragonal and rhombohedral symmetry. These generate X-ray scattering patterns which, although they may show up to ten lines (Fig. 13) must be indexed with care because a variety of structures may be possible. Many intermediate phase regions are bounded by lamellar phases which contain water filled defects; the nonuniform curvature is retained although there is no longer any ordering of the defects within the bilayer and there are no correlations between the bilayers. These defected lamellar phases may also be regarded as random mesh phases. They have been seen in the SDS/water
Figure 13. The X-ray scattering from a 52 wt% C,,EO,/ water sample. (a) Lamellar phase; (b) random mesh phase; ( c ) rhombohedra1 mesh intermediate phase; (d) Ia3d cubic phase; and (e) hexagonal +gel two phase region. 0.05
0.15
0.10
Q
@-I)
0.20
0.25
358
VII
Lyotropic Surfactant Liquid Crystals
and LiPFO/water systems and as independently occurring phases in decylammonium chloride/NH,Cl/water [86, 871 and in caesium perfluorooctanoate/water [88-911. These phases are characterized by lamellar like Bragg reflections in the ratio 1: 1/2 : 113 ..., but with a broad liquid-like reflection from the intralamellar defects (Fig. 13). Whilst ribbon and mesh phases are now fairly well established, the bicontinuous noncubic structures are still elusive. The identification of tetragonal or rhombohedral phases of a mesh or bicontinuous type is ambiguous because there usually is insufficient information to make a definitive identification. There are only a few examples 'where authors have identified a bicontinuous phase usually because of its association with adjacent bicontinuous cubic phases [79,92]. Theoretically, Hyde has cast doubt on noncubic bicontinuous phases because periodic minimal surfaces with tetragonal or rhombohedral symmetries are expected to have a higher associated bending energy cost than their cubic phase counterparts. He suggests that the phase transition from ribbons to V, can be achieved by extra tunnels connecting the mesh layers in a rhombohedral mesh phase. Anderson and coworkers [93-951 have proposed that both rhombohedral and tetragonal bicontinuous structures of the type first proposed by Schoen [96] are not only possible but are the most likely intermediate phase structures with these symmetries. The existence of bicontinuous noncubic phases still remains an open question. Hyde et al. [97, 981 have considered the origin of intermediate phases in detail. All intermediate phase structures are characterized by nonuniform interfacial curvature. The forces between the polar head groups and those between the lipidic alkyl chains tend to impose a certain value for the interfacial curvature. This curvature may not be compatible with the molecular length. The
problem is resolved by either the system phase separating or by the formation of structures with nonuniform surface curvature. However, it is still not clear why increasing chain length or rigidity should favor the formation of these intriguing phases over bicontinuous cubic structures. Whilst the early reports of intermediate phases concerned systems with reversed curvature [73-761 these were for surfactants where some residual short range order in the polar groups was probably present. There are few definitive reports of fully molten intermediate phases with reversed curvatures. In fact the pattern of how intermediate phases replace the normal bicontinuous cubic phase as alkyl chain size increase only became recognized as systematic studies on homologous series were carried out [37,66]. Here it has required a combination of microscopy, multinuclear NMR and Xray diffraction to elucidate the structures. Such studies on reversed phases have yet to be carried out, particularly where small variations in alkyl chain structure are made. The key factor in studies of the normal systems was a gradual increase of chain length. Conformational restrictions on the chains in normal aggregates appear to be responsible for the reduced stability of the v, phase for long chain derivatives. With reversed structures the presence of water in the aggregate cores allows conformational freedom. Hence it is likely that reversed intermediate phases will be found for systems with low water contents and bulky head groups: inevitably with multiple alkyl chain compounds. Where the chains have identical lengths such surfactants have high melting temperatures. Hence multiple unequal chains are probably required to observe reversed intermediate phases.
4
4 Phase Behavior of Nonionic Surfactants In principle the aggregation properties and mesophases of nonionic surfactants should be easier to understand than the behavior of ionic surfactants because there are no long range electrostatic forces. As the intra- and intermicellar head group interactions operate over a much shorter range than for ionic compounds, the head groups can pack close together as is required for the transformation from highly curved aggregates such as those in I, and H I phases to geometries with a smaller or negative curvature as in the reversed phases [37,99-1011. The effective head group volume is given by its actual size, together with about one hydration layer. The most widely studied group of nonionic surfactants is that of the poly(ethy1ene oxide)alkyl ethers [n-CnH2,+,(OCH,CH2),0H; C,EO,] [37, 1011. One reason for this is that it is possible to study the phase behavior
0
25
50
75
composition (wt % C , 2 E 0 6 )
Phase Behavior of Nonionic Surfactants
359
whilst systematically varying the length of either the hydrophobic alkyl chain or the hydrophilic ethylene oxide groups. Surfactants with large head groups form cubic ( I I ) and hexagonal (HI) phases. With increasing temperature the head group dehydrates and reduces in effective size, often until the interfacial curvature becomes zero or even negative. This results in the formation of the lamellar (La) phase and a subsequent temperature increase (or further decrease in hydration and head group size) can lead to the formation of reversed phases. The variation of the phase behavior with head group size is illustrated with reference to the C,,EO, homologous series [37, 46, 102, 1031, where x=2-6, 8, 12. The simplest phase behavior is shown by Cl,EO, (Fig. 14) which has H,, V, and La phases and a region of partial miscibility or clouding (W+L,). The cloud temperature (critical point) is defined as the lowest temperature at which the region of partial miscibility is seen. The maximum temperatures of the H, and V, phases are 37" and 38 "C respectively while the La phase exists up to
100
Figure 14. Phase diagram of C,,EO,/water. Note that most two phase regions are small and not shown (L,, aqueous surfactant solution; W, very dilute surfactant solution; HI, normal hexagonal phase; V,, normal bicontinuous cubic phase; La, lamellar phase; S indicates the presence of solid surfactant) (reproduced from [37]).
360
VII
Lyotropic Surfactant Liquid Crystals
73 "C. Note that the pure surfactant forms a liquid, which is miscible with water over a certain composition/temperature range. The clouding region arises from the partial miscibility of the Cl2EO, micellar solution with water above about 49 "C. This is caused by a net intermicellar attraction arising from EO-EO interactions between adjacent micelles. It is discussed further in [lo31 and the references therein. Whilst the clouding phenomenon is important for many micellar properties of these surfactants, its only influence on the mesophases is to determine the limit to which they can swell in water, that is it fixes the boundary between mesophases and the dilute aqueous solution (Fig. 14). On increasing the EO size, an I, cubic phase is observed for C12E0, between -30 and 43 wt% surfactant on the dilute side of the large H , region (Fig. 15). A V, cubic is seen over a narrow concentration range and the L, phase is reduced to an even smaller concentration and temperature range. With a further increase in EO size to C12E0,2,
I
/
' 0
'1
;
1
/
1
/
I
75 composition (wt % C ~ , ~ E O , )
25
50
only I, and H I phases are observed with the cubic phase existence range increasing to =35-53 wt% surfactant (Fig. 16). Recently it has been shown that there are in fact three distinct I, phases present 1461. With increasing surfactant concentration these have spacegroups Fm3m, Im3m and Pm3m. As one might expect, the partial miscibility region shifts to higher temperatures as EO size increases. With very long EO groups the I , phase becomes even more dominant. However, eventually increasing the EO size is expected to lead to 'micelles' with small aggregation numbers and a tiny micelle core, which will not form ordered phases. There is a distinct change in the behavior as the number of ethylene oxide groups decreases below 6. For CI2EO5[102], H I , V, and La phases are all clearly observed (Fig. 17), but with the H I and V, phases existing over much narrower concentration ranges and to lower temperatures than for CI2EO6 (to 18.5 and 20°C respectively). The L, phase is formed over a much wider concentration range. Note that above the
I 100
Figure 15. Phase diagram of C,,EO,/water (reproduced from [37]) (I,, close-packed spherical micelle cubic phase; others as for Fig. 14).
Figure 16. Phase diagram of C,,EO,,/water (reproduced from [46]) (Fm3m, Im3m. Pm3n represent different I, cubic phases; others see Fig. 14).
4 100
(o)
~~
L,+L,80
60
u
k LO
Figure 17. Phase diagram of C,,EO,/water (L', and L;, surfactant solutions; L,, liquid surfactant containing dissolved water, not fully miscible with water: L,, sponge phase. isotropic solution not fully miscible with water or surfactant, otherwise as for Fig. 14) (reproduced from [ 1021).
cloud point the lamellar phase coexists with a dilute aqueous phase and just above this the L, so called 'sponge phase' occurs. This phase has received much attention from academic researchers, both because of its unique position in the phase diagram and because it is closely related to the occurrence of bicontinuous microemulsions [104-1071. It is comprised of large aggregates having a net negative curvature (the opposite of normal micelies). The aggregates extend throughout the phase, hence it is continuous in both the aqueous and surfactant regions, so it is bicontinuous. Also the 'aggregates' are extremely labile, hence L, phases have low viscosities but occasionally show shear birefringence at higher concentrations. Their formation can be slow, so the establishment of exact phase boundaries is sometimes tedious and time consuming. In some of the earlier mesophase studies it was not recognized just how slowly the L, forms under some conditions, hence
36 1
Phase Behavior of Nonionic Surfactants
L, regions may be underestimated (e.g. Fig. 18, below). In fact, the surfactant aggregate structure is closely related to that of the reversed bicontinuous cubic phases which are frequently found on the phase diagram adjacent to L, regions (e.g. CI2EO,, see below). Whilst there is a dramatic difference between the mesophase regions of C I ,EO, and C I ,E05, the difference arises just from the small difference in head group size, and the existence of partial miscibility above the cloud temperature. Note the marked decrease in the miscibility of CI2EO5and water above 80°C. This is a common feature with these surfactants. The phase behavior of CI2EO, is very similar to that of C ,2EO5 with the La phase existing over a similar concentration range (25-80 wt%), but to a slightly lower temperature (68°C). Clouding occurs at 4°C and the H, phase exists only at temperatures below -2 "C. The occurrence of a V, phase, however, can not be definitely proved. An L, phase is observed at similar concentrations but about 10 "C below that of C 2E0s. For C12E03 no micellar phase occurs (Fig. 18a), at least above 0 "C, while the only mesophase shown on the phase diagram is La which exists over a concentration range of =47-85% surfactant and to a maximum temperature of 52 "C. The L, phase is shown on the phase diagram as being continuous with L,. In many of these systems it is difficult to observe coexistence of L, and L, phases at high concentrations because of the similar refractive index of the two phases. Laughlin has observed that L, and L2 are not continuous in this system [ 1081. In the above three systems there are also a number of two phase regions in which various phases coexist with very dilute surfactant solution (W+L l, W + L 2 , W+L,, w + La). For CI2EO2[ 1031, in addition to La and L, phases there are two V, reverse bicon-
,
362
VII
Lyotropic Surfactant Liquid Crystals
composition (wt 70C,,EO,)
I
0
I
20
40
60
80
I
1oc
wt% C,,EO,
Figure 18. (a) Phase diagram of C,,EO,/water (reproduced from [37]. (b) Phase diagram of C,,EO,/ water (reproduced from [112]). Symbols as for Figs. 14 and 17.
tinuous cubic phases which exist between the other two phases over a temperature range of 24-36 "C (Fig. 18b). X-ray diffraction [lo91 has shown the cubic phases to have space groups Ia3d and Pn3m. Because of the reduced hydrophilicity of the head group, no mesophases exist above 36°C. Note that these occur at higher water concentrations than the L, phase over a small temperature region: so that addition of water promotes negative aggregate curvature. This is consistent with the view of L, as having aggregates with negative curvature, and is a common occurrence with many surfac-
tants that form reversed phases. It underpins the important general observation that there is not a similar regular phase sequence (I& V,, H, .,.) with concentration for the reversed structures as occurs f o r normal mesophases. For the remaining members of the series C12E01and C,,EO, (dodecanol) no mesophases occur, only separate water and amphiphile phases which show slight miscibility. With all the C12 surfactants no mesophase exists above 8O"C, while the maximum mesophase 'melting' temperature varies much less than the cloud temperatures. Like behavior is observed for homologous series of surfactants of different alkyl chain lengths, but the mesophase regions extend to higher temperatures with increasing n. The effect of curvature increases with EO number as expected. So for all alkyl chain lengths ( n > lo), I, cubic phases only exist for EO, when x > 8 and extensive L, regions occur for small EO groups (x 5 5 ) . The influence of alkyl chain length can be illustrated using C,E06 surfactants [37]. C,E06 gives just an H, phase which melts at ~ 1 2 ° C [ l l O , 1111. C,,E06 shows a substantial area of H1 phase but there is contention as to whether it additionally shows a low temperature L, phase [ 111,1121. No V, phase occurs with these two materials. The phase behavior of C 2E06is reported above. C14E06is very similar to C,,E06 only with the L, phase having a significantly higher melting temperature (95°C as opposed to 73°C). On increasing the chain length to C 16 there is a dramatic alteration in the phase behavior similar to that occurring between C12E06and C12E05.The behavior is made even more complex because a monolayer interdigitated gel (Lp) phase occurs below about 25°C. The H, and V, melting temperatures are slightly decreased (both 34 "C compared to 37 and 40°C for C14E06) and the L, melting temperature slightly in-
4
creased (102 "C). This is the first member of the C,E06 series to show an La+ W region and an L, phase (90->lo2 "C). Moreover, with an increase in chain size to c16, the bicontinuous cubic phase begins to be replaced by intermediate phases. In fact C16E06exhibits several phases not in the original report, including a rod micelle nematic phase, a random-mesh-lamellar phase and another intermediate phase, the latter being metastable [ 1131. Whilst the existence range of the nematic phase is small for CI6EO6,with CI,EOx a low viscosity long rod micelle nematic phase has been observed (Fig. 19) over a narrow concentration range at =34 wt% surfactant over the temperature range 28-54°C between the H, and L, phases [ 1141. Here the increase in EO size removes the intermediate phases and the La+ W and L, regions, but a gel (Lp) phase is present at temperatures below the molten phases. For much longer chain surfactants such as C,,,E06 (hexaethylene glycol cis- 13-docosenyl ether) [ 1151 or 'C,,EO,' (nonaethylene glycol mono( 1 1-oxa- 14,18,22,26tetramethylheptacosy1)ether [851 the corre-
8
2,. I
;l1:
0
25
, j
, 50
S
75
100
C,,EO, (wt.%)
Figure 19. Phase diagram of C,6E08/water (reproduced from [ 1141). (N:, rod nematic phase; otherwise as for Figs. 14, 17 and 18).
Phase Behavior of Nonionic Surfactants
363
lated defect mesh phases replace the V, region. No gel phases are expected here because of the disrupted alkyl chain packing. Note that several general conclusions are illustrated by the above behavior. First, increasing surfactant chain length has little effect on the composition ranges of phases with positive curvature, the major influence is to raise the upper temperature limit of the mesophases. Hence very short chain (C,, n < 8) surfactants do not form mesophases at all. Moreover, long rod nematic phases occur at the L,/Hl boundary for long chain surfactants. This pattern of behavior is true for all surfactant types. The reason why few N, phases are reported in the literature is because such long chain surfactants are usually insoluble. The C2,= and 'C30' derivatives referred to above have their molecular packing in the crystal disrupted by the presence of methyl side groups or unsaturation, both of which increase miscibility with water. Conroy et al. [lo31 investigated the effect of replacing the terminal OH in three polyoxyethylene surfactants by OMe (CI2EO,OMe, m = 4 , 6 , 8). For m = 4 the La phase melts at 27 "C for OMe (68°C for OH) and the L, phase is seen at lower temperatures (24-27" as opposed to 51.5-70°C). For m = 6 the H,, V , and La phases all melt at lower temperatures for OMe (24, 28, 43 "C as opposed to 37, 38, 73 "C for OH). For m = 8 the V, and La phases are not seen for OMe and the I, and H, melting temperatures reduced to 15 and 41 "C respectively from 16 and 59°C for OH. This shows the importance of the terminal OH of the EO groups, a factor often neglected. Commercial polyoxyethylene surfactants contain a wide range of EO sizes, often with a reasonably defined alkyl chain. Bouwstra et al. [ 1161 studied a commercial sample of C,=C9EO14 (i.e. a C I xchain with 9-10 cis double bond and a polydisperse EO group). It exhibits phase behavior similar to C r2EOx
364
VII
Lyotropic Surfactant Liquid Crystals
with I,, H,, V, and La phases. The H, phase exists between 35.6 and 74.8 wt% surfactant and up to 84 "C. The other phases exist over narrower concentration ranges and to lower temperatures. The commercial Synperonic A surfactants are a mixture of C I 3 (66%) and C15 (34%) alkyl chains while the hydrophilic chains are composed of polydisperse ethylene oxide groups. Investigations into the phase behavior of A7, A1 1 and A20 (containing an average of 7 , l l and 20 EO groups respectively) have been achieved using a number of experimental techniques [ 117, 1181. A7 shows H, and La phases, with the H, existing between 30 and 62 wt% surfactant up to a maximum of -30"C, while the La exists between 5 1 and 86 wt% surfactant to a temperature greater than 60 "C. With an increase of EO size to the A1 1 compound, a small area of V , cubic phase appears between the H, and La phases, while the latter is greatly reduced in stability (7383 wt% with a maximum temperature of -35°C). The H, phase exists at slightly higher concentrations than for A7 and up to =64 "C. The phase behavior of the A20 compound is radically different to the shorter chain homologues, with only I, and H, phases present. The I, phase exists between 24 and 62 wt% surfactant up to -70 "C and the H, between 60 and 8 1 wt% to over 80 "C. Unlike the other two surfactants the A20 compound shows no cloud point (below 100 "C) at low surfactant concentrations. Triton X-lOO@and X-l14@are industrial p-tert-octylphenol polyoxyethylene surfactants with about 9.2 and 7.5 ethylene oxide groups per molecule respectively. They are used in biochemical studies because of their ability to disrupt biological membranes without denaturing integral membrane proteins [ 1191. TX- 100 [ 1201forms an H, phase between 37 and 63% surfactant from 0 to
28 "C and a La phase between 65 and 78% up to 6°C. TX-114 [121] exhibits a larger area of La phase from 35 to 83% surfactant and from -20 to 65 "C, but does not form HI. Essentially these studies allow two general conclusions to be made when comparing pure and commercial nonionic surfactants. First, the stability of the V, phase is much reduced. Secondly, a commercial surfactant of average formula C,EO, resembles a pure surfactant with a slightly smaller EO groups such as C,EO,-,. (An example is given in the section on mixed surfactants below.) Thus the surfactant molecules with different EO sizes mix together in the same aggregates, rather than the long and short EO components separating into phases with large and small aggregate curvatures respectively. The small difference in hydrophilicity caused by changing the head group size by one EO unit is responsible for this. With other commercial surfactants where the head group is a small polymer of strongly hydrophilic residues, such as with alkyl polyglycerols, the difference in hydration and size between head groups differing by 1-2 polymer units is very often sufficient to cause the coexistence of phases with different curvatures (e.g. L,+H,) over a wide range of water contents. Obviously, this is only a rough guide for materials where the EO distribution follows some regular pattern. Mixed commercial nonionic surfactants can show marked deviations from that expected for the 'average' EO size if the distribution of EO sizes is bimodal (i.e. for mixed commercial surfactants). Whilst studies on linear alkyl surfactants are common, in recent years branched-chain nonionic surfactants have been studied as well as surfactants with novel head groups and surfactant mixtures. The phase behavior of a series of midchain substituted surfactants with the gen-
4
era1 structure CH3(CHz),CHEOx(CH2)SCH, where x = 3 , 4, 6, 8, 10 (denoted S-C,,EO,) has been studied by optical microscopy using the penetration technique [ 1221. These have much more bulky hydrophobic groups and show clear differences from the behavior of the linear materials described above, as expected from the packing constraints. No liquid crystal phases are seen for sC,,E03, just the coexistence of W and L,. For s-CI2EO4,a V, reversed bicontinuous cubic and a La phase are seen, both of which melt at low temperature (8.0 and 0.5 "C respectively) in addition to L2 and W. On increasing the EO size to that of s-C,,EO,, La and L, phases are observed. The L, phase exists between 35.3 and 53 "C on the lower surfactant concentration side of the La phase which exists up to 48.8"C. L, micellar solution is seen only below 5.9"C with a cloud temperature below 0 "C. The next compound s-C12E08(Fig. 20), shows similar phase behavior to s-C ,,E06
'eH13\
Phase Behavior of Nonionic Surfactants
with the melting temperatures of the La and L, phases increased (to 63.5 and 66.7"C respectively). A micellar solution with a cloud point at 19.2"C is present. Finally, s-Cl2EOI0(Fig. 21) gives H,, V , and La phases on penetration at 0 "C. The H, and V, phases melt to micellar solution at 100"C). C , , G exhibits an HI phase between 33 and 86 "C and V, and La phases from 39.5 "C and 49 "C respectively to >lOO"C. Similarly for C,,G, H I (45.575 "C), V, and La (50 "C and 56 "C respectively to >lOO"C). The phase behavior for C18=Gis much simpler exhibiting only an La phase from 22 "C to >lo0 "C. Raaijmakers et al. studied the mesogenic properties of some 3-0-alkyl derivatives of D-glucitol and D-mannitol [ 1371. Phase penetration scans have shown that the C10-16 glucitol derivatives and the C,, mannitol derivative all exhibit lyomesophases. 3-0-decyl-D-glucitol (C,,G) gives an H I phase on penetration at room temperature, and on heating a V, phase at 34 "C and a La phase at 46 "C. The solid bulk melts to La at 55 "C and all phases exist to at least 98 "C. CI2G gives an I, phase at 36 "C, H I (37 "C), V, (40 "C) and La (46 "C), with the solid melting to La at 65°C. All phases exist to >lo0 "C with the exception of the H I which melts to an I, phase at 62°C. C,,G gives only V, and La phases from 48" to >lo0 "C. Below T,,, the La phase cools to give a
372
VII
Lyotropic Surfactant Liquid Crystals
100
80
T/"C 60
40
20
L1
+ Cryst
0
J 20
40 wt. % ClO
60
80
Lp gel phase, a transition that is reversible on reheating. c16G gives a La phase from 52 to >lOO"C. A reversible La-Lp transition is observed on cooling at 44°C. The pattern of phase formation in 3-0-dodecylD-mannitol (C,,M) is similar to C,,G with all transition temperatures approximately 20 "C higher, apart from the melting temperature of the solid (97 "C). Finkelmann and Schafheutle studied monomeric and polymeric amphiphiles containing a monosaccharide headgroup [ 1381. The monomer, N-D(-)-gluco-N-methyl( 12-acryloyloxy)-dodecane1-amide, exhibits L, (>52 "C) and La phases (65-88%, 57-87 "C), with a biphasic region separating the two. The polymer, poly-(N-D(-)-gluco-N-methyl-( 12-acryloyloxy)-dodecane1-amide) again exhibits L, and La phases but these are formed at -0.1 "C. The lamellar phase exists between 37 and 100% polymer, and up to 184.8 "C. It is separated from
100
Figure 26. Phase diagram of C l 0 glucitol/water (reproduced from [136]). Symbols as for Fig. 14.
the L1 phase by a narrow biphasic region. There is a region of clouding (L, +L2) reaching from -0 to 31% polymer with a lower critical temperature of 17.9"C at 4.6% and an upper critical temperature of 105.5 "C at 9.6% polymer. Most of the above studies involve changes in alkyl chain size, with no alteration of the head group. In a very important early study, Sagitani et al. reported the behavior of pure alkyl polyglycerol [C,,0(CH2CHOHCH20),0H] surfactants with an ether-linked C12 alkyl chain, with n= 1-4 [138a]. The compounds with n = 1, 2 give just a lamellar phase to over 120 "C, while with n =4, a micellar solution and hexagonal phase occur. For n = 3 there is a cloud point at ca. 53 "C, with micellar solution and lamellar phase being present. The shape of the L1/Laphase boundary indicates the presence of H, below the cloud point, remarkably similar behavior to C12E06as the au-
5
thors point out. In fact, poly-OH compounds do resemble EO nonionics, but with a very much reduced influence of temperature. Much attention has been given recently to commercial polyglucoside surfactants (alkyl polyglucosides, APG's) [138b, c, d], where the average number of 'po1y'-glucoside units (termed d.p.) falls within 1-3. From the above studies, one expects short chain derivatives with d.p.=ca. 1 to form L,, H, and L, phases, while the low stability of the H, phase suggests that long chain compounds will form L, dispersions. Compounds with d.p.=2 will give H, and L, phases, while with d.p. = 3 or more the I , cubics will appear. Multi-phase coexistence is likely, along with the occurrence of metastable, sticky, viscous phases at high concentrations.
5 Block Copolymer Nonionic Surfactants Block copolymer surfactants show qualitatively similar phase behavior with temperature to conventional materials. However, solubility requires the presence of branched chains and/or ether links to reduce the occurrence of crystalline polymer with a high melting point. It must be emphasized that commercial polymeric surfactants are polydisperse in both alkyl chain and EO blocks, hence a range of molecular weights and head group sizes will be present. This will certainly modify the behavior compared to pure low molar mass surfactants, but should not qualitatively change it. It could lead to the occurrence of multiphase regions such as H I and L, or cubic regions, rather than single phases and 'intermediate' structures that might occur with very monodisperse polymers.
Block Copolymer Nonionic Surfactants
373
Alexandridis et al. have studied a number of block copolymers of ethylene oxide (EO), propylene oxide (-OCH,CHCH,-, PO) and butylene oxide (-OCH,CH(CH,CH3)-, BO) in water and in ternary systems with p-xylene. The polypropylene oxide and butylene oxide blocks are the hydrophilic portions. Obviously, the short methyl and ethyl branches will need to be included in any packing constraints considerations. The phase behavior of three ethylene oxide/propylene oxide ABA copolymers, (Eo)6(Po)34(Eo)6 (L62), (EO)13(PO),()(EO) 13 (L64) and (EO)37(PO),8(EO),, (P105) has been studied [139]. The number of mesophases formed increases with the poly(ethylene oxide) content (Fig. 27). L62 exhibits only a L, phase, between 5 1 and 76% polymer and from 85 "C. Note that simple AB block copolymers are expected to resemble even more closely the conventional surfactants. (The ABA blocks are likely to assume a U type conformation for the hydrophobic block.) The ternary system of L64/P105/water was also studied [I401 (Fig. 28). At 25 "C a lamellar phase is formed along the L64-water axis between 61 and 79% polymer, and between 73 and 87% polymer on the P105-water axis, and extends all the way from one axis to the other. Upon increasing the water concentration the hexagonal phase is formed. This extends between
374
VII
I
mlti phase
Lyotropic Surfactant Liquid Crystals
\
Figure 27. Phase diagrams of (a) (EO),(PO),,(EO), (L62), (b) (EO)I~(PO)~O(EO),, (L64), and (c) (EO),,(PO),,(EO),, (P105) block copolymers and water (reproduced from [ 1391). (E, normal hexagonal phase; I, cubic phase; P indicates presence of solid polymer; otherwise as for Figs. 14, 17 and 23.)
52-55% on the L64-water axis and 4767% on the P105-water axis. Additionally there is an I, cubic phase between 28 and 43% P105, and a narrow melted V, phase between 58 and 59% L64 between the H, and L, regions. The L64/water (D,O)/p-xylene ternary system exhibits L,, L, (containing a high p-xylene to water ratio), La, H,, H, and V, phases at 25 "C [ 1411. Most of the phase diagram consists of two and three phase regions (10 and 6 respectively). The L, solution phase forms up to ~ 4 6 % polymer, and can solubilize up to 4% xylene. The polymer is miscible with xylene in all proportions along the whole water-lean side of the phase diagram (the L, region): 12% water can be incorporated into the L, phase. The L, phase forms on the polymer-water axis between 61 and 81% polymer and extends well into the phase diagram, accommodating up to 20% xylene. The H, phase is formed on the polymer-water axis between the L, and L, phases (47-57% polymer). Like the L, phase, it can solubilize only a small amount of xylene. There is a fairly extensive H, phase between the La and L, phases at polymer concentrations of 4378%, and with water concentrations of 13- 22%. The V, phase forms in a small region between the La and H, phases at ~ 7 0 % polymer. It has the Ia3d space group. One other phase seen is an isotropic liquid phase (at 58-59% polymer) denoted here as L'. It has low viscosity and is an L, sponge phase. Pluronic 25R4 has the structure (PO),,(EO),,(PO),,, and has similar ternary phase behavior in waterlp-xylene as Pluronic L64 which has the opposite block sequence and a comparable molecular weight (see above). At 25 "C the 25R4/waterlp-xylene ternary phase diagram exhibits the same phases as L64 [ 1421. Again, the phase diagram is dominated by two and
5
Block Copolymer Nonionic Surfactants
375
Figure 28. Phase diagram of L64/P105/ water (reproduced from [140]). (L', L,, sponge phase; otherwise as for Figs. 14, 17, 23 and 27.)
three phase regions. The L, phase is formed along the polymer-water axis at polymer concentrations between 61 and 79%, and can accommodate up to 25% p-xylene. The H, phase is formed between the L1 and La phases at polymer concentrations of between 43 and 56%: 3-4% xylene is actually required for the phase to form. The H, phase is formed between the L, and L, phases, at polymer concentrations of between 48 and 73% polymer, with the water concentration varying in the range 13- 18%. The V, phase forms in a narrow concentration range of 65-68% polymer. It is thought to have a space group of Ia3d. In the (EO) 17(BO)lo/D,O/p-xylene Alexandridis et al. claimed the first, to their knowledge, reverse micellar cubic phase in a 'typical' ternary amphiphile/water/oil system [ 1431 before studying and reporting the system in more detail [ 1441. They report six mesophases and two solution phases at 25 "C. On increasing the polymer concentration along the polymer-water axis the phase sequence L, (140 "C. The C I 4 compound exhibits similar phase behavior [147] but also has a nematic phase (Nd) between 31 and 35% surfactant up to 60°C. A more typical surfactant is (dodecyldimethy1ammonio)propanesulfonate [ 1481. This forms an I, cubic phase between 49.5 and 63% surfactant from ~ 2 0 ° Cup to =85 "C. A hexagonal phase (65-87% surfactant) forms at a slightly higher temperature but exists up to >160"C. A lamellar phase is seen at very high concentrations and above 170°C. Below this is a hydrated solid surfactant phase which when viewed under a polarizing microscope has the optical appearance of a gel phase.
7
Ionic Surfactants
Ionic surfactants have been studied over a much longer period than nonionic materials, but the long-range electrostatic interactions and the insolubility of long chain materials have resulted in the general picture of mesophase behavior emerging only rather slowly. Typical surfactants with a single charge ionic group usually give small globular micelles and H, as the low concentration mesophase. As chain length increases (and if solubility allows), then rod micelles form and a rod nematic phase can occur at the L,/H, boundary. Occasionally, if the counter-ion is highly dissociated, an I, phase can occur as the first mesophase. Otherwise two charged groups are required for I, formation. At higher concentrations the sequence H,/V,/L, changes to H,/Intermediate/La (or a combination of the two) as
7
the alkyl chain size increases. In this region there can be long lived metastable phases that are difficult to work with and prevent easy determination of equilibrium boundaries. Sodium dodecyl sulfate is a very commonly used surfactant, yet for many years its mesophase behavior was not studied in detail. One reason for this is the number of phases to be dealt with. Besides the lamellar and hexagonal phases there are a number of intermediate phases existing over narrow concentration ranges [78]. The complete phase diagram was published by KCkicheff et al. [62] (Fig. 29). (Note that in this paper the hexagonal phase is represented by the symbol Ha.) Between the H a and L, phases they identify four intermediate phases, with biphasic regions separating most of the single phase regions. The H, phase is first seen at 37% (with L,) and at 40% in a single phase region at 25.7 "C. A two-dimensional monoclinic phase (Ma) exists with the H a phase from 57-59% up to 58 "C. Note that this is the only mesophase region that does not exist up to -100 "C. On increasing surfactant concentration between 59 and 69% the phase sequence is M, (40. I "C), Ma+%, (rhombohedral), %, (43.0 "C), Q, (cubic, 47.8 "C), Q,+T, (tetragonal), T,(48.5 "C), T a + L , L,(49.8 "C). The La phase exists up to 87% surfactant above which it coexists with a crystal phase. Despite the extensive work involved in this diagram, there remain questions. For example, the disappearance of the Ha+% two phase region is difficult to understand. Note that this system typifies those where equilibration between different states is expected to be slow, as mentioned above. Another system showing an intermediate phase between the H, and L, phase is the caesium tetradecanoate-water [ 1491. The phase behavior between 24 and 80°C was studied (Fig. 30). The H , phase is first seen in a two phase region with L, micellar phase
Ionic Surfactants
377
at 33% surfactant and in a single phase region between 37 and 67%. Above 65 "C the Ha phase is replaced by a two phase Ha+ V, region. However, below this temperature, a ribbon phase (R) is formed. The most likely structure of this phase is rodlike aggregates with an elliptic cross section. At 24 "C the R phase exists from 62-72% surfactant, with the concentration range decreasing with increasing temperature. Below 39 "C a biphasic R+L, region exists between 72 and 75% surfactant, and between 39 and 65°C the two phase region is R + V,. The V, region is very narrow (-2% surfactant) and is bordered by a two phase V, +La region up to 75% surfactant. The lamellar phase exists to over 90% surfactant. That the sequence of phases can be very complex and varies with chain size is illustrated by the 'penetration scan' table reported by Rendall et al. [150]. Here the 'intermediate' phase types have not been identified, but the general pattern and complexity of behavior is clear. One should also caution that minor levels of surface active impurities are likely to have a marked influence on the cubic/ intermediate phases, particularly if the impurities are uncharged cosurfactants (often likely to be present from the chemical synthesis). One of the first major surveys into the phase behavior of cationic surfactants was done by Blackmore and Tiddy who did penetration scans on a variety of surfactants including the -NH;, NMe,f, NEt;, NPr; and NBu; series [ 15I]. Previous studies had mostly been limited to alkylammonium and alkylmethylammonium salts [ 152, 1531. These show a qualitatively similar behavior to that of the anionic systems, with the sequence H,/V,/L, for short chain derivatives being replaced by H,/Int/La for long chain compounds. For the intermediate chain lengths, V, phases replace the intermediate structures as temperature increases.
378
VII
Lyotropic Surfactant Liquid Crystals
7
Ionic Surfactants
379
Figure 30. Phase diagram of CsTD/ water (reproduced from [ 1491).
Surfactants containing divalent headgroups have received little attention. Hagslatt et al. have investigated a number of these. Dodecyl- 1,3-propylene-bis(ammoniurn chloride) (DoPDAC) has a headgroup consisting of two protonated amine groups separated by a propylene group [ 1541. The surfactant shows conventional phase behavior in water (D20) (Fig. 31) forming L,, I, 160
140
y
120
23
loo
+d
n: w
B
g
x0
60 40
20
. . ... . .... ... . . . .._...
0
0
10 20
30 40 50 60 70 WT% DoPDAC
80
~
90 100
Figure 31. Phase diagram of DoPDAC/water (reproduced from [154]). (C+S indicates the presence of solid surfactant; otherwise as for Figs. 14. 15. 23 and 27.)
cubic (denoted I:), H I hexagonal (denoted E), V, cubic (denoted I;) and La lamellar (denoted D) phases. The I, cubic phase forms at 30% surfactant at ~ 3 0 ° C and exists up to =150"C. The hexagonal phase is seen at 42% surfactant from 38 to >160"C. Between 65 and 72% surfactant the H, phase melts to the V , cubic phase (which forms above =75 "C). Similarly above 72% surfactant the V, phase melts to a La lamellar phase. At 79% the La forms at 90°C. Note that the Krafft boundary has a rather large slope. Usually for monovalent surfactants the Krafft boundary is flat or has only a small slope [40]. They also studied the ternary phase diagram of the divalent surfactant dipotassium dodecylphosphate (K2DoP), monovalent surfactant potassium tetradecanoate (KTD) and D 2 0 [155]. At 25°C K2DoP exhibits L, (0-37% surfactant), I, (37-67%), H , (67-75%) and two phase liquid crystal/ hydrated surfactant crystal (75 - 100%). In the ternary system I, cubic is seen between 20 and 60% K2DoP. Up to 30% KTD can be incorporated into the phase at lower con-
380
VII
Lyotropic Surfactant Liquid Crystals
centrations of K2DoP. On decreasing the percentage of D,O along the D20/KTD axis, hexagonal (E), ribbon (R), intermediate (possibly orthorhombic) and lamellar (D) phases are seen. The hexagonal phase exists right down to the K2DoP/D20 axis (70-75% K2DoP), whereas the maximum K,DoP concentrations for the ribbon, intermediate and lamellar phases are 45, 24 and 22% respectively. Amphitropic (also referred to as amphotropic) surfactants are compounds containing both lyotropic (e.g. quaternary ammonium headgroup) and thermotropic (e.g. oxycyanobiphenyl group) moieties [ 1561. Fuller et al. report on two such compounds. 1O-(4'-cyano-4-biphenyloxy)decyltriethylammonium bromide exhibits only a La phase [157, 1581 from 22 to >90% surfactant with a temperature range of 33-100 "C. The second compound N,N'-bis(5-(4'-cya-
T I "c 105
85
Cct
65
45
L
25
0
10
20
30
40
50
60
70
Wt% 5-6-5 OCB
no-4-biphenyloxy)pentyl)-(N,N,N',N')-tet-
Figure 32. Phase diagram of 5-6-5 OCB/water (reproduced from [152]). Symbols as for Fig. 14.
ramethyl hexane diammonium dibromide (5-6-5 OCB) exhibits two La phases with a re-entrant L, micellar phase between them (Fig. 32). The low concentration lamellar phase (La) exists between 18 and 64% surfactant, up to 65°C. Between 35 and 47% surfactant the melting temperature of the La phase decreases from 65 to 50°C with increasing surfactant concentration. At this concentration the high concentration phase (L;) appears and exists to > 80% surfactant and >90"C. Note that the absence of HI and V, phases is due to the difficulty of packing the bulky oxycyanobiphenyl group into spherical or rod-like micelle interiors. The second of the above compounds is an example of a so-called 'gemini' surfactant. Recently there has been considerable interest in these surfactants which are doubleheaded cationic compounds in which two alkyl dimethyl quaternary ammonium groups per molecule are linked by a hydrocarbon spacer chain. (These are denoted as n-m-n
surfactants where n is the length of the terminal alkyl chains and m the length of the spacer chain.) In a second paper [158] Fuller et al. report the phase behavior of further m-n-m OCB surfactants and some straight chain 15-n-15 surfactants (n= 1,2,3,6). Note that these compounds have terminal hydrophobic chains of the same length as the oxycyanobiphenyl compounds. The m-n-m OCB surfactants all give just a lamellar phase from lOO"C. Additionally, a nematic phase is seen for m = 1, 2 and intermediate phases for m = 1, 2, 3. Zana et al. have investigated various properties of a series of gemini surfactants including their thermotropic and lyotropic phase behavior. They report that the 12-412 and 12-8-12 compounds form hexagonal
8 The Influence of Third Components
and lamellar phases in water [ 159, 1601. For the latter compound the H, phase exists between =55 and 78% surfactant and the La phase between 82 and 97% surfactant. The phase behavior of a wide range of gemini surfactants ( n = 12-18, m=2-20) has now been studied and a wide range of phases observed [ 1613. Generally the phases are qualitatively in accord with packing constraint expectations, where the spacer group is included in the hydrophobic volume. Howevere, unusual partial miscibility regions involving hexagonal, lamellar or concentrated surfactant solutions in coexistence with dilute solutions are also observed. Perez et al. [ 1621 have studied a series of novel gemini surfactants with guanidyl headgroups (from the amino acid arginine) and are referred to as bis(Args). They are made up of two symmetrical N“-acyl-Larginine residues of 10 [CJCA),] or 12 [CJLA),] carbon atoms linked by covalent bonds to an a,w-alkenediamine spacer chain of varying length (n,n=3,6,9). In the C,(LA), and C6(LA),/water systems an HI phase is seen at very low surfactant concentrations (lower than 5 % ) between 4 and 20°C. Penetration scans have shown that in C,(LA), the H, phase is replaced by an La phase.
8 The Influence of Third Components There is a vast body of data concerning the influence of third components on surfactant liquid crystals. Because of the possible great complexity of the inherent mesophase behavior this array of data can appear to be enormously difficult to rationalize. However, if we can consider the simple concepts described above (micelle formation, micelle
38 1
shape/packing constraints and the nature of intermicellar interactions) then a reasonably simplified picture emerges, at least for the water continuous phases. This section does not attempt to be comprehensive, it simply reports selected examples of behavior to illustrate general concepts. The simplest way to show the changes in mesophase behavior is to employ ternary phase diagrams. The reader should recall that the important factors are the behavior as a function of surfactant/additive ratio, and the volume fraction of surfactant + amphiphile (where present) for mesophase formation.
8.1
Cosurfactants
Cosurfactants are ‘surfactants’ that are insufficiently hydrophilic to form micelles or mesophases with water alone, but can have dramatic effects when mixed with normal surfactants. Examples of such materials are alcohols, fatty acids, and long chain aldehydes. Depending on the strength of the polar group hydration there is greater or smaller incorporation of the cosurfactant into mesophases. The extent to which the polar group resides at the micelle surface also has a profound effect on the mesophases. Weakly polar groups such as methyl esters can both occupy the micelle interior and reside with the ester group at the surface. They could be classified as ‘polar oils’. Thus cosurfactants span the range of properties from oils to surfactants. To illustrate the effects we show the behavior of a simple surfactant, sodium octanoate, with various additives taken from the extensive body of research produced by Ekwall, Fontell and coworkers [40, 1631. Sodium octanoate forms only hexagonal phase at room temperature, but with V , cubic and lamellar phases occurring at higher temperatures.
382
VII
Lyotropic Surfactant Liquid Crystals
Figures 33 to 35 show the ternary phase behaviors of sodium octanoate with decanol, octyl aldehyde (octanal), and methyl octanoate. All three additives have roughly the same alkyl chain length and volume as sodium octanoate (decanol is just a bit bigger). The phase behavior of octanol/sodium octanoate/water is available but the decanol system has received the most attention by far. We see that decanol mixes in the system to form a large lamellar region, and even an inverse hexagonal phase. This is because the alcohol group always resides at the water/
alkyl chain surface. Its contribution to the micelle area is ca. 12 A2, and since sodium octanoate has a head group size of ca. 58 A* in the mesophases, packing constraint calculations suggest a hexagonal/lamellar transition at a sodium octanoate/decanol weight ratio of 7 : 3 in excellent agreement with the phase diagram. (A theoretical calculation of this type of phase behavior has been described by Wennerstrom et al. [164]). The other two additives are dissolved within the hexagonal region, presumably because they occupy the micelle interior to some extent. Octanol can reside at the micelle surface to
Decanol
Figure 33. Phase diagram of sodium caprylate (octanoate)/decanol/water (reproduced from [40]). (B, ‘mucous woven’ lamellar phase; C, tetragonal phase; F, reversed hexagonal phase; G, isotropic phase; otherwise as for Figs. 14, 17, 23 and 27.)
8 The Influence of Third Components
383
appears rather than a V, cubic phase is a general observation. Both intermediate and bicontinuous cubic phases are much less frequently encountered in multicomponent systems than in binary surfactant/water mixtures.
8.2 Mixed Surfactants
Figure 34. Phase diagram of sodium caprylate (octanoate)/caprylaldehyde (octanal)/water (reproduced from [40]). Symbols as for Figs. 14, 23, and 27.
Methylcaprylate
40v
Wolcr
,
-,SO
sodium coprytolc
Figure 35. Phase diagram of sodium caprylate (octanoate)/methylcaprylate (octanoate)/water (reproduced from [40]). Symbols as for Figs. 14,23 and 27.
some extent because it forms a lamellar phase which is in equilibrium with L,, hence disc micelles can occur. Methyl octanoate mainly resides in the micelle interior, thus the lamellar phase does not have a boundary with L,. The fact that the lamellar phase
The behavior of mixed surfactants (at least in water-rich regions) can be understood from considering the nature of interactions between head groups and packing constraints. A simple example is that of the commercial nonionic surfactants reported by Bouwstraet al. [ 1651.They compared the phase behavior of technical grade C,,EO(,) and pure C12E0,. The ternary phase behavior with water and decane were studied and compared. Both exhibited I,, H I and L, phases but only the C,,EO, gives a very small region of V, cubic phase. Note that the I, phase is not seen in the binary C,,EO,/ water system [37]. Also the H, phase in C,,EO, is composed of infinite long rods, whereas short interrupted rods were found in the hexagonal phase of C 12E0(7). Where ionic surfactants are involved, like charges show only small changes in phase structures but overall solubility can be greatly increased because generally surfactants do not form mixed crystals, hence each increases the solubility of the other. For surfactants with opposite charges, usually the mixed salt is insoluble. Where solubility does occur, the average head group size is much reduced because of the neutralized electrostatic repulsions. Typically cationic and anionic surfactants, which alone form hexagonal or I , cubic phases, form lamellar phases in 1 : 1 mixtures [166a]. Obviously, with dialkyl surfactants more complex behavior can occur.
3 84
VII
Lyotropic Surfactant Liquid Crystals hexodecone
8.3 Oils Oils are solubilized into the interior of micelles where they allow the micelle to swell to a larger radius, hence giving rise to cubic (I,) and hexagonal (H,) phases at smaller a values than for the surfactant alone. Polar oils can also reside at the micelle surface to some extent, reducing micelle curvature and inducing the occurrence of lamellar and inverse phases. This behavior is typified by the behavior of the commercial nonionic surfactant nonylphenol-(probably branched)decaethylene oxide with hexadecane and p-xylene [40]. Figures 36 and 37 show that the hexadecane induces the formation of an I, cubic phase, whereas the much more polarizable p-xylene, which can reside at the chainlwater interface, induces a lamellar phase. As expected with the increase in micelle diameter, all the phases have their boundaries shifted to higher volume fractions with hexadecane addition. For surfactants having small polar groups and bulky chains there can be extensive ef-
p-Xylene
Figure 37. Phase diagram of EMU-O9/hexadecane/ water (reproduced from [40]). Symbols as for Figs. 14, 15,23 and 21.
fects with the addition of oils. Alone with water these surfactants form reversed micelles and/or reversed mesophases. Large volumes of oil can be incorporated into these systems because of the possibility of swelling the alkyl chain regions in these oilcontinuous phases (L,, H,, V,). While extensive research has been carried out on this area, it appears to be much more complex than for the water-continuous phases. Each different surfactant type can show individual behavior according to the curvature properties of the surfactant layer.
8.4 Hydrotropes
Water
aa
EMU-09
Figure 36. Phase diagram of EMU-09/p-xylene/water (reproduced from [40]). Symbols as for Figs. 14, 17,23 and 28.
Hydrotropes are small, highly water soluble additives that increase markedly the solubility of other components, including surfactants, in water. They are employed very widely in industry. In fact, they only work when the ‘insoluble’ phase is a mesophase with high molecular mobility (e.g. polymer coacervate or surfactant mesophase). They include weakly self-associating elec-
8 The Influence of Third Components
SO/
\ \ \ \ .
_.Water
--.
Sodium caprylate
Figure 38. Phase diagram of sodium caprylate (octanoate)/ethanol/water (reproduced from [40]). Symbols as for Figs. 15, 22, 23 and 27.
trolytes such as sodium xylene sulfonate and short chain alchols. Their influence can be illustrated by the effect of ethanol on the hexagonal mesophase of sodium octanoate (Fig. 38). Ethanol ‘solubilizes’ the hexagonal phase, shifting the boundary from ~ 4 0 % to =45% although it initially stabilizes it. There is almost no increase in maximum solubility of crystalline surfactant. Where mesophases occur over a wider concentration range (say 10-40%) then hydrotropes can sharply increase the isotropic solution range by removing the mesophase.
8.5
Electrolytes
The influence of electrolytes divides into two areas, the influence on nonionic (uncharged) surfactants and that on ionic materials. For nonionic surfactants the effects are either ‘salting out’ or ‘salting in’, in line with observations from the Hofmeister series of electrolyte effects on protein precipitation. Whilst there has been much discus-
385
sion over the past S O years on the molecular mechanism involved, including many words on the ‘structure’ of water. Ninham and Yaminsky have recently shown in a landmark paper that the phenomena can be explained by dispersion interactions [ 166bl. Essentially electrolytes that are adsorbed to surfactants increase their solubility, hence mesophases dissolve. Those that are desorbed from the aggregates raise the chemical potential and decrease solubility. Like hydrotropes, large effects are seen only with mesophase or ‘coacervate’ precipitates. For ionic surfactants there are two additional effects. Where a counter-ion that produces an insoluble surfactant salt is present then any mesophases are removed by an increase of the Krafft temperature. Also, the well known common-ion effect raises Krafft temperatures. Additionally, electrolytes have a marked influence on CMC values, reducing the relatively high CMC of ionic surfactants to the low values of zwitterionic derivatives. This happens typically when the added electrolyte level is of the same order as the CMC range (=1-300% of the CMC without electrolyte). Otherwise, where solubility is maintained, at high electrolyte levels, one observes salting in or salting out effects. These can be illustrated by the phase behavior of sodium soaps with water (Fig. 39) [167]. Added electrolyte induces coexistence between a dilute aqueous phase and micellar solutions (similar to a nonionic surfactant cloud temperature) and then changes hexagonal phase to the lamellar phase. Clearly, the ionic surfactant mesophase swells in water because of electrostatic repulsions. High salt levels negate the electrostatic repulsions, changing the behavior to that of a moderately polar surfactant (e.g. C,,EO,). Salting-in electrolytes, such as sodium thiocyanate are likely to cause the lamellar phase to swell possibly forming mi-
386
VII
Lyotropic Surfactant Liquid Crystals
Figure 39. Phase diagram of sodium laurate (dodecanoate)/sodium chloride/water (reproduced from [ 1671). (A) Salt + various curds, (B) Salt + curd + saturated lye electrolyte solution, (C) Grained soap (curd) on lye, (D) Neat soap (lamellar, La) +lye + curd, (E) Neat soap on lye, (F) Neat soap on nigre (L,), (G) Nigre on laye, (H) Salt+ saturated solution containing traces of soap.
celles again at very high levels. This merits further study.
8.6 Alternative Solvents Over recent years the question of whether or not surfactants form aggregates similar to micelles and mesophases in other polar solvents has received considerable attention [ 168- 1911. The solvents concerned are polar liquids such as glycerol, ethylene glycol, formamide or ethyl ammonium nitrate (a molten electrolyte at ambient temperature). The answer is yes, but the thermodynamics of the aggregation process is somewhat different. Figures40 and 41 show measurements of EMF in hexadecylpyridinium bromide solutions where water and ethylene glycol are solvents. These results clearly show that aggregates form over a very narrow concentration range in ethylene glycol, similar to micelle formation. However, the aggregation concentration is higher by two orders of mag-
nitude: simply because the ‘solvophobic’ effect is much smaller than the hydrophobic effect. In fact a useful ‘rule of thumb’ is that a surfactant with a C,, chain in the alternative solvents resembles a C12 surfactant in water. It seems that at least a C,, or C12chain is required for micelle-like aggregation. The mesophase behavior is remarkably similar
E m f (water)
E mf
(ethylene glycol)
I
loo
10’
lo2
I lo3
loL
lo5
lo6
Surfactant concentration / lo6 rnol dm” mV
Figure 40. EMF as a function of concentration for the cetylpyridinium bromide electrode in water and ethylene glycol (reproduced from [ 1681).
387
8 The Influence of Third Components 0-05 - I 0 0051-~
t
.
.
001
__ 0
005
01
015
0.25
02
Total surfactant concentration
Figure 41. The monomer concentration of cetylpyridinium bromide plotted against total concentration (reproduced from [ 1681).
to that in water, again with a C,, chain derivative in the solvent resembling that of a C,, surfactant in water. There are, of course, detailed differences. Thus cubic phases are stabilized over the intermediate phases (as expected for shorter chain surfactants). Polyoxyethylenes are not miscible with glycerol, so form no mesophases. In the main though, given the large differences between the nature of the polar solvents and water (molecular size, symmetry, polarity H-bonding ability, conformational freedom, influence on electrostatics) this similarity is remarkable. Table 4 shows the mesophase
Table 4. Phase penetration data of anhydrous C,,P,Br, showing the temperature range of the mesophases.
Water EG GLY FA EAN
-
-
-
-
-
-
48-61 -
57 -
43-100+ 44- 108 31- 150+ 50-100+ 49-120+
-
99 -
-
54-100+ 54- 106 48-150+ 61-100+ 74-120+
59-100+ 56-150+ 56-150+ 61-100+ 73-120+
The sequene of the columns indicates the order of the phases with increasing surfactant concentration. The first temperature given is Tpen;the second indicates the upper temperature limit of the mesophase. A '+' indicates that the mesophase is stable above the temperature given (the limit of measurement). L; indicates L , intrusion between the I, and H I phases. Similarly, L; indicates an L, intrusion between the H I and V, phases. The anhydrous surfactant forms L, at 74 to 75 "C.
120
100
80
60
Solid (S)
*"
(a)
0
25
50
Concentratwn
I 75
Sohd (S)
25
100
(b)
50
75
0
Concentration
Figure 42. Phase diagram of (a) C,,PyBr/water, (b) C,,PyBr/EAN (reproduced from [189]). Symbols as for Fig. 14.
388
VII
Lyotropic Surfactant Liquid Crystals
behavior of octadecylpyridinium bromide in a range of solvents. The phase sequence and types are all remarkably similar. Only formamide shows a real difference with I , cubic phases being observed. Even so, these are formed by shorter chain derivatives (the chloride salts) in water. The phase diagrams (Fig. 42) show clearly that in ethylammonium nitrate as solvent the boundaries are shifted to much higher concentrations, again resembling a shorter chain surfactant.
9 Conclusions It is clear that a framework based on packing constraints, surfactant chemical type, and water (solvent) - surfactant interactions now exists, so that the pattern of solventcontinuous surfactant mesophases can be rationalized for a wide range of materials, both for water and other polar solvents. Moreover, many data exist on the influence of additives, which can be seen to fit into
the general picture. Whilst no mention of theoretical computer modelling has been made here, this is an area that is evolving rapidly at present. Both molecular and mesoscopic models are under development that should enable semi-quantitative computer predictions of phase behavior to be made in the next few years. These should be able to shed light on remaining problems such as the structures of intermediate and gel phases. However, further experimental work is required on the structures of intermediate phases and gel phases. In addition, the knowledge of mesophase glasses and particularly other semi-solid mesophases is at a rather primitive level. Moreover, the general framework can now be employed to rationalize surfactant mesophase behavior with more complex additives such as welldefined biomolecules (e.g. proteins) as well as synthetic polymers of all types. Membrane protein crystals (where the protein hydrophobic region is covered by surfactant) certainly fall into this class. It is now timely to consider problems where mesophase kinetics are important.
.5
Time (5)
Figure 43. Series of X-ray patterns of a commercial nonionic lamellar phase, formed by fast mixing (reproduced from [192])
10 References
Mesophase formation is involved in the applications of many specialist formulated products, including detergents. Recent advances in instrumental techniques allow the initial events that occur on mixing surfactant and water to be monitored, for example using synchrotron X-ray diffraction and a fast mixing cell (Fig. 43) [ 1921. To develop a proper understanding of the behavior will require a consideration of diffusion processes in labile anisotropic media, an exciting opportunity for computational modelling. Rheology is perhaps the most important user property of surfactant mesophases, since almost every product is pumped, poured and stirred. Recent developments mean that it is now possible to make X-ray, neutron, optical microscope and other measurements on flowing systems to relate macroscopic orientation/morphology to rheological properties at the same time as molecular/aggregate reorientation is also studied. This offers the exciting prospect that in future not only will it be possible to formulate products for control of the mesophase structure through an understanding of ideas such as packing constraints etc., but rheological properties and dissolution rates will be controlled to deliver optimum benefit to customers. The future developments offer exciting intellectual challenges which will have an important scientific impact as well as being of widespread practical use.
10 References D. F. Evans, H. Wennerstrom, The Colloid Domain Where Physics, Chemistry, Biology and Technology Meet, VCH, New York, 1994. J . H. Clint, Surfactant Aggregation, Blackie, Glasgow, 1990. R. G. Laughlin, The Aqueous Phase Behavior uf Surfactants, Academic Press, London, 1994.
389
[4] C. Tanford, The Hydrophobic Effect: Formation of Micelles and Biological Membranes, Wiley Interscience, New York, 1980. [S] D. M. Small, The Handbook of Lipid Research - The Physical Chemistry of Lipids, Plenum Press, New York, 1988. [6] J. Israelachvili, Intermolecular und Surface Forces, Academic Press, London, 1992. [7] R. Lipowsky, E. Sackman, Eds., Handbook of Biological Physics, Vol. l a , Structure and Dynamics of Membrane, From Cells to Vesicles, Elsevier Science, Netherlands, 1995. [8] R. Lipowsky, E. Sackman, Eds., Handbook of Biological Physics, Vol. I b, Structure and Dynamics of Membrane, Generic and Specific Interactions, Elsevier Science, Netherlands, 1995. 191 D. F. Evans, Langmuir 1988,4, 3. [lo] M. Costas, B. Kronberg, R. Silveston, J. Chem. Soc. Faraday Trans. 1 1994, YO ( I I ) , 1513. [ I I ] B. Kronberg, M. Costas, R. Silveston, Pure Appl. Chem. 1995,67 ( 6 ) ,897. [I21 B. Kronberg, M. Costas, R. Silveston, J. Disp. Sci. Technol. 1994, 15 (3),333. [I31 K. Shinoda, M. Fujihara, Bull. Chem. Soc. Jpn. 1968,41,2162. (141 K. Shinoda, J . Phys. Chem. 1977,81, 1300. [151 D. N. Glew, J . Phys. Chem. 1962,66,605. [I61 P. Eriksson, G. Lindbom, E. Burnell, G. J. T. Tiddy, J . Chem. Soc. Faraday Trans. I 1988,84 (Y), 3129. [I71 E. A. G. Aniansson, S . N. Wall, M. Almgren, H. Hoffmann, I. Kielmann, W. Ulbicht, R. Zana, J. Lang, C. Tondre, J . Phys. Chem. 1976, 80, 905. [ 181 C. Elvingson, S . Wall, J. Colloid. Int. Sci. 1981, 121,414. [19] R . Palepu, D. G. H. Hall, E. Wyn-Jones, J . Chem. Soc. Faradaj Trans. I 1990,86, 1535. [20] D. M. Bloor, E. Wyn-Jones, unpublished data. [21] H. Gharabi, N. Takisawa, P. Brown, M. A. Thomason, D. M. Painter, D. M. Bloor, D. G. Hall, E. Wyn-Jones, J. Chem. Soc. Famday Trans. 1 1991, 87, 707. [ 2 2 ] E. A. G. Aniansson, J. Phys. Chem. 1978, 82, 2805. [23] J. Israelachvili, H. Wennerstrom, Langmuir 1990, 6, 873. [24] J. Israelachvili, H. Wennerstrom, J. Phys. Chem. 1992, 96, 520. [2S] G. J . T. Tiddy, M. F. Walsh, E. Wyn-Jones, J. Chem. Soc. Faraday Trans. I 1982, 78, 389. [26] Z. X. Li, E. M. Lee, R. K. Thomas, J. Penfold, J . Colloid Int. Sci. 1997, 187, 492. [27] P. Mukerjee, K. J . Mysels, CriticalMicelle Concentrations of Aqueous Surfactant Systems, NSRDS-NBS 36,1971. 1281 V. Degiorgio, Physics of Amphiphiles: Micelles, Vesicles and Microemulsions (Eds.: V. Degiorgio, M. Corti), NPC, 1985.
390
VII
Lyotropic Surfactant Liquid Crystals
[29] R. J. M. Tausk, Ph. D. Thesis, Physical Chemical Studies of Short-Chain Lecithin Homologues, State University, Utrecht, Netherlands, 1974. [30] M. J. Rosen, Chemtech. 1993, 23 (3), 16-27. [31] V. V. Kumer et al., J. Phys. Chem., in press. [32] T. J. Madden, Unilever Research Port Sunlight Laboratory, Personal Communication, 1990. [33] D. J. Mitchell, B. W. Ninham, J. Israelachvili, J. Chem. Soc. Faraday Trans. I 1976, 72, 1525. [34] CRC Handbook of Chemistry and Physics, CRC Press, Boca Raton, FL. [35] D. J. Mitchell, B. W. Ninham, J. Israelachvili, Biochem. Biophys. Acta 1977,470, 605. [36] G. J. T. Tiddy, Physics Report 1980, 57, 1. [37] D. J. Mitchell, G . J. T. Tiddy, L. Waring, T. Bostock, M. P. McDonald, J. Chem. Soc. Faraday Trans. 1 1983, 79, 975. [38] P. A. Winsor, Chem. Rev. 1968, 68, 1. [39] V. Luzzati, Biological Membranes. Physical Fact and Function, Academic Press, New York, 1968. [40] P. Ekwall in Advances in Liquid Crystals, Vol. 1 (Ed.: G . H. Brown), Academic Press, New York, 1975. [41] J. M. Seddon, Biochim. Biophys. Acta 1990, 1031, 1-69. [42] J. M. Seddon, R. H. Templer, Ch. 3 in Structure and Dynamics of Membranes: From Cells to Vesicles (Ed.: A. J. Hoff), Elsevier Science, Netherlands, 1995. [43] G. Lindblom, L. Rilfors, Biochim. Biophys. Acta 1989, 998, 221. [44] K. Fontell, K. K. Fox, E. Hanser, Mol. Cryst. Liq. Cryst. 1989, 1, 9. [45] J. Charvolin, J. F. Sadoc, J . Phys. (Paris) 1988, 49, 521. [46] P. Sakya, J. M. Seddon, R. H. Templer, R. J. Mirkin, G. J. T. Tiddy, submitted to Langmuir, 1997. [47] A. Gulik, H. Delacroix, G . Kischner, V. Luzzati, J. Phys. I1 France 1995,5, 445. [48] G. J. T. Tiddy, J. M. Seddon, J. Walsh, unpublished data. [49] J. M. Seddon, E. A. Bartle, J. Mingins, J. Phys. Condens. Matter 2, suppl. A, 1990. [50] V. Luzzati, P. A. Spegt, Nature 1967,215,707. [51] L. E. Scriven, Nature, 1976, 263, 123. [52] A. Tardieu, V. Luzzati, Biochim. Biophys. Acta 1970,219, 11. 1531 P. Eriksson, G . Linblom, G . Arvidson, J. Phys. Chem. 1985,89, 1050. [54] K. D. Lawson, T. J. Flautt, J. Am. Chem. Soc. 1967,89,5489. [55] B. J. Forrest, L. W. Reeves, Chem. Rev. 1981, 81, 1. [56] N. Boden, P. H. Jackson, K. McMullen, M. C. Holmes, Chem. Phys. Lett. 1979, 65,476. [57] B. Luhmann, H. Finkelmann, G . Rehage, Makromol. Chem. 1985,186, 1059. [58] K. Radley, Liq. Cryst. 1995, 18, 1, 151.
[59] D. Chapman, R. M. Williams, B. D. Labroke, Chem. Phys. Lipids 1967,1,445. [60] K. Larsson, Z. Phys. Chem. 1967, 56, 173. [61] J. M. Vincent, A. E. Skoulios,Acta Cryst. 1966, 20,432. [62] P. KCkicheff, C. Grabielle-Madelmont, M. Ollivan, J. Colloid Interface Sci. 1989, 131, 112. [63] A. S. C. Lawrence, M. P. McDonald, Liq. Cryst. 1966, 1,205. [64] A. S. C. Lawrence, Mol. Cryst. Liq. Cryst. 1969, 7, 1. [65] R. C. Laughlin, Paper presented to the SIS, Jerusalem, 1996. [66] C. D. Adam, J. A. Durrant,M. R. Lowry, G. J. T. Tiddy, J. Chem. Soc. Faraday Trans. 1 1984,80, 789. [67] L. Mandelkern in Comprehensive Polymer Science, Vol. 2 (Ed.: C. Booth, C. Price), Pergamon Press, Oxford 1989, p. 415. [68] V. Luzzati, H. Mustacchi, A. Skoulios, F. Husson, Acta Crystallogr. 1960, 13, 660. [69] F. Husson, H. Mustacchi, V. Luzzati, Acta Crystallogr. 1960, 13, 668. [70] A. Skoulios, V. Luzzati, Acta Crystallogr. 1961, 14, 278. [71] A. Skoulios, Acta Crystallogr: 1961, 14, 419. [72] P. A. Spegt, A. Skoulios, Acta Crystallogr. 1963, 16, 301. [73] P. A. Spegt, A. Skoulios,Actu Crystallogr. 1964, 17, 198. [74] V. Luzzati, A. Tardieu, T. Gulik-Krzywicki, Nature 1968, 217, 1028. [75] H. Hagslatt, 0. Soderman, B. Jonsson, Liq. Cryst. 1992, 12, 667. [76] 0. Soderman, G. Lindblom, L. B.-A. Johansson, K. Fontell, Mol. Cryst. Liq. Cryst. 1980,59, 121. [77] I. D. Leigh, M. P. McDonald, R. M. Wood, G . J. T. Tiddy, M. A. Trevethan, J. Chem. Soc. Faraday Trans. 1 1981, 77,2867. [78] P. KChcheff,B. CabaneJ. dePhys. 1987,48,1571. [79] P. KCkicheff, B. Cabane,Acta Crystallogr. 1988, B44, 395. [80] P. KCkicheff, J. Colloidlnterface Sci. 1989,131, 133. 1811 P. Quist, K. Fontell, B. Halle, Liq. Cryst. 1994, 16, 235. [82] P. KCkicheff, G. J. T. Tiddy, J. Phys. Chem. 1989,93,2520. [83] C. E. Fairhurst, M. C. Holmes, M. S. Leaver, Langmuir, 1997, 13,4964. [84] C. E. Fairhurst, M. C. Holmes, M. S. Leaver, Langmuir 1996,12,6336. [85] J. Burgoyne, M. C. Holmes, G. J. T. Tiddy, J. Phys. Chem. 1995,99,6054. [86] M. C. Holmes, J. Charvolin, J. Phys. Chem. 1984, 88, 810. [87] M. C. Holmes, J. Charvolin, D. J. Reynolds, Liq. Cryst. 1988, 3, 1147. [88] M. C. Holmes, P. Sotta, Y. Hendrikx, B. Deloche, J. de Phys. II 1993,3, 1735.
10 References
[89] M. C. Holmes, A. M. Smith, M. S. Leaver, J. de Phys. I1 1993,3, 1357. 1901 M. S. Leaver, M. C . Holmes, J . de Phvs. 11 1993,3, 105. [91] M. C. Holmes, M. S. Leaver, A.M. Smith, Langmuir 1995, I I , 356. [92] X. Auvray, T. Perche, R. Anthore, C. Petipas, I. Rico, A. Lattes, Langmuir 1991, 7, 2385. 1931 D. M. Anderson, P. Strom in Polymer Association Structures: Microemulsions and Liquid Crystals, ACS Symposium Series 384 (Ed.: M. A. El-Nokaly), American Chemical Society, Washington 1989, p. 204. [94] D. M. Anderson, Colloq. de Phys. 1990, 51, C7-1. [95] D. M. Anderson, H. T. Davis, L. E. Scriven, J. C. C. Nitsche. Advances in Chemical Physics, Vol. LXXVII (Eds.: I. Prigogine, S. A. Rice), John Wiley, New York 1990, p. 337. 1961 A. H. Schoen, Infinite Periodic Minimal Surfaces Without Selj-intersections, NASA Technical Note D-5541, Washington 1970. [97] S. T. Hyde, PureAppl. Chem. 1992,64, 1617. [98] S . T. Hyde, Colloq. de Phys. 1990,51, C7-209. [99] B. Lindman in Surfactants (Ed.: Th. F. Tadros), Academic Press, London, 1984. [ 1001 V. Degiorgio, M. Conti in Proceedings of the International School of Physics, ‘Enrico Fermi’, Italian Physical Society, 1985. [ l o l l J. Sjoblom, P. Stenius, I. Danielsson, Ch. 7 in Nonionic Surfactants (Ed.: M. J. Schick), Marcel Dekker, New York, 1987. [ 1021 R. Strey, R. Schomaker, D. Roux, F. Nallet, U. Olsson, J. Chem. Soc. Faraday Trans. 1990, 86, 2253. [I031 J. P. Conroy, C. Hall, C. A. Lang, K. Rendall, G. J. T. Tiddy, J. Walsh, G. Lindblom, Prog. Colloid. Polym. Sci. 1990,82, 253. [lo41 B. P. Binks, Annual Reports Section C, The Royal Society of Chemistry 1996, 92, 97. [I051 H. Kellay, B. P. Brinks, Y. Hendrikx, L. T. Lee, J. Meunier, Adv. Colloid lnt. Sci. 1994,49, 85. 11061 M. S . Leaver, U. Olsson, H. Wennerstrom, R. Strey, U. Wurz, J. Chem. Soc. Faraday Trans. 1995,91,4369. [I071 U. Olsson, H. Wennerstrom, Adv. Coloid lnt. Sci. 1994,49, 85. [ 1081 R. G. Laughlin, Personal Cornrnunicafions (1996). [lo91 J. S . Clunie, J. M. Corkill, J. F. Goodman, P. C. Symons, J. R. Tate, Trans. Faraday Soc. 1967, 63, 2839. [110] J. S. Marland, B. A. Mulley, J. Pharm. Pharmacol. 1971, 23, 561. [ 1 1 I] B. A. Mulley, A. D. Metcalf, J. Colloid Sci. 1964, 19, 501. [ 1121 S . S. Funari, G. Rapp, J. Phys. Chem. B 1997, 101, 732. [ 1131 S. S . Funari, M. C. Holmes, G. J. T. Tiddy, J. Phys. Chem. 1994,98,3015.
391
11 141 J. Corcoran, S. Fuller, A. Rahman, N. N. Shinde, G. J. T. Tiddy, G. S. Attard, J. Muter. Chem. 1992,2,695. [ 1151 S. S. Funari, M. C. Holmes, G. J. T. Tiddy, J . Phys. Chem. 1992,96, 11029. [ I 161 H. Jousma, J. A. Bouwstra, F. Spies, H. E. Junginger, Colloid Polym. Sci. 1987, 265, 830. [ 1171 G. T. Dimitrova, Th. F. Tadros, P. F. Luckham, Langmuir 1995,1I, 1101. [ l I81 G. T. Dimitrova, Th. F. Tadros, P. F. Luckham, M. R. Kipps, Langmuir 1996, 12, 315. [ 1191 C. Tanford, J. A. Reynolds, Biochim. Biophys. Acta 1976, 457, 133. [I201 K. Beyer, J. Colloid lnt. Sci. 1982, 86, 73. [I211 R. Heusch, F. Kopp, Ber. Bunsenges. Phys. Chem. 1987,91,806. 11221 L. Thompson, J. M. Walsh, G. J. T. Tiddy, Colloids and Surfaces A: Physicochemical and Engineering Aspects 1996, 106, 223. [ 1231 K. Kratzat, H. Finkelmann, J. Colloid lnt. Sci. 1996,181,542. 11241 K. Kratzat, H. Finkelmann, Colloid Polym. Sci. 1994, 272,400. [ 1251 K. Kratzat, C. Schmidt, H. Finkelmann, J. Colloid Int. Sci. 1994, 163, 190. [126] H. Kratzat, C. Stubenrauch, H. Finkelmann, Colloid Polym. Sci. 1995, 273, 257. 11271 H. Kratzat, H. Finkelmann, Liq. Cryst. 1993, 13, 691. [128] H. Soderlund, J. Sjoblom. T. Warnheim, J. Disp. Sci. Technol. 1989, 10, 13 1. 11291 H. Finkelmann, B. Luhmann, G. Rehage, Colloid Polym. Sci. 1982, 260, 56. 11301 H. Finkelmann, B. Luhmann, G. Rehage, Angew. Makromol. Chem. 1984,123,217. [I311 B. Luhmann, H. Finkelmann, Colloid Polym. Sci. 1987, 265, 506. [ 1321 B. Luhmann, H. Finkelmann, Colloid Polym. Sci. 1986, 264, 189. [133] J. Eastoe, P. Rogueda, A. M. Howe, A. R. Pitt, R. K. Heenan, Langmuir 1996,12,2701. 11341 P. Sakya, J. M. Seddon, R. H. Templer, J. Phys. 11 1994,4, 1311. [ 1351 F. Nilsson, 0. Sodermann, I. Johansson, Langmuir 1996, 12, 902. [136] C. Hall, G. J. T. Tiddy, B. Pfannemuller, Liq. Cryst. 1991, 9, 527. [I371 H. W. C. Raaijmakers, E. G. Arnouts, B. Zwanenburg, G. J. F. Chittenden, H. A. van Doren, Recueil des Travaux Chimiques des Pays-Bas 1995,114,301. [I381 H. Finkelmann, M. A. Schafheutle, Colloid Polym. Sci. 1986, 264, 786. [ 138aI H. Sagihari, Y. Hayashi, M. Ochiai, J. Am. Oil Chem. Soc. 1989,66, 146. [138b] D. Balzer, Langmuir 1993, 9, 3375. [138c] G. Platz, C. Thunig, J. Policke, W. Kirchhoff, D. Nickel, Colloids and Surfaces A: Physicochemical and Engineering Aspects 1994, 88, 113.
392
VII
Lyotropic Surfactant Liquid Crystals
[138d] G. Platz, J. Policke, C. Thunig, R. Hofman, D. Nickel, W. Vonrybinski, Langmuir 1995, 11,4250. [139] P. Alexandridis, D. Zhou, A. Khan, Langmuir 1996,12,2690. [I401 D. Zhou, P. Alexandridis, A. Khan, J. Colloid fnt. Sci. 1996, 183, 339. [141] P. Alexandridis, U. Olsson, B. Lindman, Macromolecules 1995, 28, 7700. [I421 P. Alexandridis, U. Olsson, B. Lindman, J. Phys. Chem. 1996, 100, 280. [I431 P. Alexandridis, U. Olsson, B. Lindman, Langmuir 1996, 12, 1419. [ 1441 P. Alexandridis, U. Olsson, B. Lindman, Langmuir 1997, 13, 23. [ 1451 P. Holmqvist, P. Alexandridis, B. Lindman, Langmuir 1997,13,2471. [146] E. S. Lutton, J. Am. Oil Chem. Soc. 1966, 43, 28. [I471 H. Hoffmann, G. Oetter, B. Schwandner, Progr. Coll. Polym. Sci. 1987, 73, 95. [I481 C. La Mesa, B. Sesta, M. G. Bonicelli, G. F. Ceccaroni, Langmuir 1990, 6, 728. [I491 J. C. Blackburn, P. K. Kilpatrick, J. Colloidlnt. Sci. 1992, 149,450. [150] K. Rendall, G. J. T. Tiddy, M. A. Trevethan, J. Chem. Soc., Faraday Trans. 1 1983, 79, 637. [I511 E. S. Blackmore, G. J. T. Tiddy, J. Chem. Soc. Faraday Trans. 2 1988, 84, 1115. [I521 A. Khan, K. Fontell, G. Lindblom, J. Phys. Chem. 1982,86,383. [ 1531 R. G. Laughlin in Cationic Surfactants, Surfactant Science Series, Vol. 37 (Eds.: D. N. Rubingh, P. M. Holland), Marcel Dekker, New York, 1990. [I541 H. Hagslatt, 0. Soderman, B. Jonsson, L. B.A. Johansson, J. Phys. Chem. 1991,95, 1703. [I551 H. Hagulatt, 0. Soderman, B. Jonsson, Langmuir 1994, 10, 2177. [I561 S. Fuller, J. Hopwood, A. Rehman, N. N. Shinde, G. J. T. Tiddy, G. S. Attard, 0. Howell, S. Sproston, Liq. Cryst. 1992, 12, 521. [IS71 S. Fuller, Ph. D. Thesis, University of Salford 1995. [I581 S. Fuller, N. N. Shinde, G. J. T. Tiddy, G. S. Attard, 0. Howell, Langmuir 1996, 12, 1117. [I591 E. Alami, H. Levy, R. Zana, A. Skoulios, Langmuir 1993, 9,940. [I601 R. Zana, Y. Talmon, Nature 1993, 362, 228. [I611 S. Fuller, G. J. T. Tiddy, R. Zana, unpublished results, 1996. [162] L. Perez, J. L. Torres, A. Manresa, C. Solans, M. R. Infante, Langmuir 1996, 12, 5296. 11631 H. Hagslatt, K. Fontell, J. Colloid fnt. Sci. 1994, 165, 2,43 1. [I641 H. Wennerstrom, B. Jonsson, J. Phys. Chem. 1987, 91, 338. [165] J. A. Bouwstra, H. Jousma, M. M. van der Meulen, C. C. Vijverberg, G . S . Gooris, F.
Spies, H. E. Junginger, Coll. Polyrn. Sci. 1989, 267, 53 1. [166a] C. A. Barker, D. Saul, G. J. T. Tiddy, B. A. Wheeler, E. Willis, J . Chem. Soc. Faraday Trans. 11974, 70, 154. [166b] B. W. Ninham, V. Yaminski, Langmuir 1997, 13, 2097. [I671 J. W. McBain, A. J. Burnett, J. Chem. Soc. 1922,121, 1320. [I681 H. Gharabi, R. Palepu, G. J. T. Tiddy, D. G. Hall, E. Wyn-Jones, J. Chem. Soc. Chem. Commun. 1990,2, 115. [I691 D. F. Evans, A. Yamauchi, R. Roman, E. Z. Casassa, J. Coll. Inte$ Sci. 1982, 88, 89. [I701 D. F. Evans, A. Yamauchi, G. Wei, A. Bloomfield, J. Phys. Chem. 1983, 87, 3537. [I711 W. Binani-Limbele, R. Zana, Coll. Polym. Sci. 1989,267,440. [I721 M. Sjoberg, U. Henriksson, T. Warnheim, Langmuir 1990, 6, 1205. [I731 M. Jonstromer, M. Sjoberg, T. Warnheim, J. Phys. Chem. 1990,94,7549. [I741 I. Ricco, A. Lattes, J. Phys. Chem. 1986, 90, 5870. [175] A. Couper, G. P. Gladden, B. T. Ingram, Faraday Discuss. Chem. Soc. 1975,59,63. [176] A. Ray, J. Am. Chem. Soc. 1969,91,23,6511. [I771 P. D. I. Fletcher, P. J. Gilbert, J. Chem. Soc. Faraday Trans. 1 1989,85, 147. [I781 N. Moucharafieh, S. E. Friberg, Mol. Cryst. Liq. Cryst. 1979,49, 23 1. [I791 M. A. El Nokaly, L. D. Ford, S. E. Friberg, D. W. Larsen, J. Coll. Inter- Sci. 1981,84,228. [ 1801 L. Ganzui, M. A. El Nokaly, S . E. Friberg, Mol. Cryst. Liq. Cryst. 1982, 72, 183. [I811 S. E. Friberg, P. Laing, K. Lockwood, M. Tadros, J. Phys. Chem. 1984,88, 1045. [182] S. E. Friberg, C. Solans, L. Ganzuo, Mol. Cryst. Liq. Cryst. 1984, 109, 159. [I831 D. F. Evans, E. W. Kaler, W. J. Benton, J. Phys. Chem. 1983,87,533. [184] A. Balmajdoub, J. P. Marchal, D. Canet, I. Rico, A. Lattes, Nouv. J. Chem. 1987, 11,415. [ 1851 T. Warnheim, A. Jonssen, J. Colloid. lnterSci. 1988, 125, 627. [I861 X. Auvray, C. Petipas, R. Anthore, I. Rico, A. Lattes, J. Phys. Che.m. 1989, 93, 7458. [I871 Z. Lin, H. T. Davies, L. E. Scriven, Langmuir 1996,12,5489. [188] T. A. Bleasdale, G. J. T. Tiddy, E. Wyn-Jones, J. Phys. Chem. 1991,95,5385. [I891 T. A. Bleasdale, G. J. T. Tiddy in Organised Solutions (Eds.: S. E. Friberg, B. Lindman), Marcel Dekker, New York 1992. [I901 X. Auvray, C. Petipas, I. Rico, A. Lattes, Liq. Cryst. 1994, 17, 109. [I911 H. D. Dorfler, Z. Phys. Chem. 1994,187, 135. [ 1921 CLCR Synchrotron Radiation Department, Daresbury Laboratory Annual Report, 19951996.
Handbook ofLiquid Crystals D. Demus, J. Goodby, G. W. Gray, H.-W. Spiess, V. Vil Copyright 0WILEY-VCH Verlag GmbH. 1998
Chapter VIII Living Systems In memoriam Hans Kelker (21. 9. 22-25. 6. 92) and Horst Sackmann (3. 2. 21 -2. 11. 93), who pioneered the essentials of liquid crystal research and human coherences in a bipartite world. This chapter is also dedicated to the memory of my parents, of which, may father was a pupil of Daniel Vorlander f o r a short fortunate period before World War I ended all.
Siegfried Hoffmann
1 Introduction Living systems, like the universe from which they are descended, are mainly understandable as a process [ 1 - 81. They are characterized preferentially by their dynamics rather than by their statics. As “structure” and “phase” correspond in some way to “molecular biology” (Fig. 1) and “liquid crystals” (Fig. 2), the two fields will appear in our days as two scientific aspects of much more general duality phenomena (Fig. 3) that govern quite different stages and hierarchies of developmental processes within our universe. They both represent essentials in our somewhat inadequate attempts to cope with the complexity of life patterns, which can be only approximately - if at all - comprehended by the only partially adequate complexity of our different scientific approaches. The consideration of structure - phase dualities in pre-life transitions and in the development of life will thus appear as a continuation and amplification of basic quantum dualities - with their complementarities of “subject” and “relatedness” - into the transient order - disorder patterns of supra-
molecular organizations and their biomesogenic regulations, bringing them forward, at least in our spacetime, to the as-yet final achievements and destinations of the fields of self-awareness of our species within a creative universe.
Figure 1. “Structure”: Dickerson dodecamer [9] (left) and prealbumin dimer [lo] (right bottom) (Langridge’s first examples [ 111 of electrostatic potential visualizations in combination with skeletal and transparent CPK presentation) as molecular biological structural representatives of the nucleoprotein system, but also as biological representations of “individual” and “relatedness” within the frame of structure-phase transitions; familiarities between early universe structure computer simulations [ 121 and cholesteric texture paintings by Lehmann [ 13,141 (right top).
394
VIII
Living Systems
2 Biomesogens and the Grand Process Figure 2. “Phase”: from ammonium oleate (left), Lehmann’s [ 141 first “seemingly living crystals” (original photograph as a kind gift of Hans Kelker) [ 131, and life’s informational component’s liquid crystalline DNA texture [lS] (right), the present realization of Lehmann’s early visions.
Figure 3. “Duality”-phenomena in life onto- and phylogenesis [I -8,16,17,18] withinuniverse developments and continuations from quantum vacuum up to consciousness pattern creativities together with the intriguing background design modified from [ 161.
Biomesogens (Fig. 4) evolved as an early amphiphilic pattern from the interface dialectics between order and disorder on the primordial earth [5 d, 7 a, 17, 181.Their holographic molecular designs provided decisive prerequisites for the projection of individual molecular facilities into the structural and functional amplifications of cooperative and dynamic mesogen domain ensembles and their spatiotemporal coherences. By directional nonlinear dynamizations of the supramolecular organizations and their (bio)mesogenic operation modes, they arrived at a state of recognition and responsiveness. Competing individualizations drove autocatalytic propagation, self-reproduction, information processing, and optimization. Far from thermal equilibrium, the biomesogenic order - disorder patterns gained intelligence by reality adaptation and variation. Experiencing optimization strategies, they developed self-awareness and creativity.
Figure 4. “Biomesogen” introduction in the 1970s, mesogenic nucleic acid/protein interplays framed by cholesteric color variations of polydisperse evolutionary DNA [7 a, 17, 18, 19, 63a, 7.51.
2
Biomesogens and the Grand Process
395
2.1 Historical Dualities The dualities of supramolecular biomesogenic organizations [7, 17, 181 reemerge in the process of their scientific discovery. The history of “liquid crystals”, lovingly retraced by Kelker [ 131 (Fig. 5 ) , parallels the development of “molecular biology” [5 - 81. “Liquid crystals are beautiful and mysterious, I am fond of them for both reasons”, states de Gennes [ 191. And, indeed, how can we not feel attracted to what in the inanimate world seems most closely related to us? Notwithstanding the priority roles of lyotropics and amphotropics both in native standards and by their (decades) earlier detection [ 131, notwithstanding Virchow’s soft, floating substance from the nerve core that he named myelin in 1854 [20a], Mettenheimer’s discovery of the birefringence of this substance in 1857 [20c], Heintz’s statements in Halle in the early 1850s that “stearin becomes cloudy at 5 1 - 52 “C, opalizes at higher temperature, being opaque at 58 “C, and molten and clear at 62.5 “C” [20b], it was the “two melting points” of thermotropic cholesterol derivatives that started the scientific history of liquid crystals in the early “correspondence cooperation” between Reinitzer and Lehmann in 1888 [20-271. Reinitzer’s letter to Lehmann, dated March 14, 1888 [21, 221, which introduces from rather hidden beginnings the history of liquid crystals, is very impressed by their unique appearances. Reinitzer sends two substances “with the request to investigate more closely their physical isomerism”. He describes previously unnoticed and unknown phenomena: “Cholesteryl acetate melts to a clear liquid. On cooling the liquid down, a bright color variation moves through the liquid before crystallization. Bright violet, blue, green, yellow and
Figure 5. 100 years of liquid crystal research as depicted by Kelker [13].
orange-red colors appear. If one observes the molten compound with crossed nicols, on cooling suddenly starlike groups of crystals appear colored and brightly shining on a dark background. Thereby the rest of the substance remains liquid, which might be demonstrated by slightly touching the cover-slide, which makes the crystals flow”. Lehmann [ 141 investigates the preparations with a heated-stage polarization microscope constructed by himself. They remind him of his earlier studies on “iodinesilver”, and, even more closely, of his endeavors concerning “ammonium oleate hydrate”. He coins the term “liquid crystals”. The scientific development of “liquid crystals” or “cystalline liquids” (Fig. 6) - terms
396
VIII
Living Systems
Figure 7. Liquid crystal abstractions (left to right and top to bottom): thermotropics and lyotropics with structural visualizations of formerly alien discoid [28a, b], the well-known calamitic artificial liquid crystal l-[trans-4-(alk-3-en-l-yl)cyclohex-l-yl]-4cyanobenzene [28 c], and long disregarded native helical DNA [28d].
Figure 6. Historical views (top to bottom): Vorlander’s Crystallography of Liquids [24]beside Lehmann’s Seemingly Living Crystals - Presented by a Film [14]; Haeckel’s Artful Presentations of Nature 1231, related to Lehmann’s Liquid Crystals; his rather modem view of protein properties, illustrated by ribonuclease with a water cover at the active site 1291.
that will compete in the time following Lehmann and Vorlander [14, 241 in the search for more precise and sophisticated differentiations and that will ultimately be integrated into the more comprehensive “mesogen” concept of Kelker [13] - commences with contradictions similar to those that seem to be typical of their molecular patterns. Remarkably, it is the two biomolecules that, with the thermotropy of cholesterol and the
lyotropy of oleic acid derivatives, not only provoke a first dissent concerning priority but also seem to direct further research in quite different directions (Fig. 7) [25]. The thermotropics celebrated their first period of academic dominance in the school of Vorlander [24] in Halle. Within three decades at the beginning of the 20th century, more than 2000 synthesized liquid crystals were used to establish the basis of early systematizations and structure - function correlations. Much later, in the 1950s, Vorlander’s substance-inheritance - cigar-like abstracted molecules dedicated to posterity in cigarboxes - will mediate the historical “reentrant” phenomenon in the old Vorlander institute building [26, 271, preceded only by a forerunning reawakening of phase chemistry - a sleeping-beauty role - in the
2
search for a miscibility rule for imidazole derivatives [26 d - fl. Vorlander [24] anticipates many of the synthesis strategies used today on the basis of what might now be called stereoelectronic structure - function elucidations. Even his errors will offer intriguing suggestions. Much more recent polymer research [25] is foreseen in the “nature of the infinitely long molecules”; “star-, cross- and plateletshaped molecules” are offered for discussion. The “optically anisotropic building of packet-like stacked molecular platelets resembling a Volta-column” describes the stacks of today’s discoid arrangements [25, 281. In love, however, with his own hundredfold-proven rod model, Vorlander treats these ideas about liquid crystals with scepticism. Vorlander “does not see any possibility to use the crystalline-liquid substances for technical applications” and “feels himself much more attracted by pure scientific desire” for the “crystalline-liquid appearances” in which “even more pronounced than in the case of solid crystals, the original, sensitive and yet powerful nature of molecules and their constituents will be expressed”. He does not ignore the relations to life processes, but shrinks back from any interpretation. So he would not approve “the data of Lehmann as to the crystalline-liquid characteristics of ammonium oleate hydrate and similar soaps that - molten together with water - should yield crystalline liquids”, nor would he follow his general hypotheses. Lehmann’s “comparisons of liquid crystals with living beings” appear to him “unscientific” and “contadictory”, in their last stages even “mystical” and “degrading the very object”. Looking back, however, to the first liquid crystal empire, created by him and his school, Vorlander takes at last somewhat restrained pride “in having yet arrived at a state where crystalline-liquid substances might be synthesized
Biomesogens and the Grand Process - as
397
many and as different ones as might be expected or wanted”. And - going beyond all his own doubts - his “Nevertheless, the soft crystals of organic compounds are the housing of life; without soft crystals and colloids no living beings” forms a bridge towards de Gennes’s “soft matter” in particular and “life sciences” in general [24,25,28]. But a generation will pass before liquid crystals will be really wanted. Kelker’s “MBBA”, the first liquid crystal at ambient temperature, will provide the impetus for applicational research in liquid crystals ~31. Lehmann [ 141 in Karlsruhe, however, in close mental vicinity to and mutual stimulation by the “crystal souls” of Haeckel[23] sets out for the first grand romantic period of a general view of mesogenic behavior, with special emphasis also on its lyotropic aspects. “The analogies between the shaping and driving forces of both liquid crystals and living beings” had caused him “to characterize the myelin crystals of paraazoxycinnamic acid ester as seemingly living crystals”. Proposed by his own intuition, this early impression will intensify and, towards the end of his way, widen and amplify to general considerations on “liquid crystals and life theories”. Lehmann believes “that physiology once might be able to prove that those analogies are really caused by intimate relationship or even identity of the governing forces”. It “appears to him of outstanding interest that the so-called biocrystals of living substances will orientate themselves regularly to the shapes of organisms, so as if the molecules of living substances would, indeed, exert directional forces”. Lehmann who unravels the close similarities of his liquid crystals to the world of biologically relevant mono-, oligo- and polymers at that time, who enriches his horizons with their foreseen technical applications, and who uses -in a way comparable to quite
398
VIII Living Systems
recent approaches in molecular dynamics the simulative power of films to illustrate his results and derived implications [ 14h], this Lehmann leaves to us, like the life processes that he tried to reduce to their fundaments scientifically in his first early biomesogen approaches, a system that is open in all aspects. The ambitious attempt to bridge the gap between the phenomenology of liquid crystals and their biomolecular structural prerequisites ends, owing to the times, in nothingness. What the chemist Vorlander was able to elucidate scientifically in terms of the thermotropic liquid crystal empire erected by him and his pupils, the lone physicist Lehmann had to do without. His “biocrystals” are nebulous entities. Proteins also philosophically - are associated with life [5, 7 a]. Their structure, however, remains uncertain, although Haeckel’s intuitive perception [23] appears dramatically modern [5-8, 17, 18,29,30]. Miescher announced in a first prophetical view an early overture to molecular biology [29a]. But nobody will take his nucleic acids as informational components of life processes seriously. Lehmann had already outlined a lecithin model. The structure - function relationships veiled by membranes, nevertheless, will still have to await our insights into the nucleation of informational and functional components, before the compartmental components can be elucidated. Lehmann’s investigations, thoughts, and hypotheses will be valued by the future. And also not forgotten -within the creative Halle circle - will be Schenck’s [24d, el ingenious visions on liquid crystals, in which he attempted to create a modern synthesis of Lehmann and Vorlander’s contradictory theses and antitheses, and Staudinger’s pioneering path into the field of large molecules, which opened up to chemistry not only the first views of future prosperous fields, but established, morevoer, the scien-
tific basis for an understanding of life patterns. In the meantime, however, there is the lonely prophecy of Schroedinger [29 c] that the genetic material is an aperiodic crystal, which not only anticipates with its outstanding prediction the later - elucidated reality, but raises, moreover, questions about the meaning and content of terms such as order, disorder, entropy andnegentropy [ 1-8,171, and the philosophical categorizations, inferences and deductions around at a time that prefers merely static views. When Watson and Crick [29fl, following the laying of foundation stones in the field of protein geometries by Pauling [29 d], hypothesize their double helical matrix structure, it proves, with regard to its linguistics, indeed to be an aperiodic arrangement. With regard to its state, however, it appears to be a liquid rather then a proper solid crystal. And all biooligomers and biopolymers that have since been discovered likewise occupy fertile meso-positions between the sterile statics of solids and the vain dynamics of liquids. They will appear to be mesogens in the broad productive ranges between the extrema (Fig. 7) [5,7, 17, 18,28-301. The time had come to award Lehmann and his grand hypotheses their long-denied rehabilitation. But liquid crystal people on one hand and molecular biologists on the other were mainly concerned with their own problems and appeared, moreover, to be split within their own disciplines. The possible unifying (bio)mesogen view in the dialectical synthesis of the divergent approaches of Lehmann in Kalsruhe and Vorlander in Halle had to continue to await its realization.
2
2.2 “Genesis” and Phase Transitions A first grand phase transition from quantum vacuum to existence within an instability; probabilistics rather than deterministics in all the incomprehensible beginnings, irreversibilities and the time arrow of entropy guiding further developments; consecutive freezing of originally unified forces and symmetry breaking; “matter” developments between fermion - boson, particle - wave, and structure-phase dualities; interfacial routes to mesogens leading out of the contradictions - will these suggestions reveal valid pictures [I-8, 17, 18, 29-33]? “Somewhat forlorn between the glows of genesis and the shadows of decline: endangered, tentative structures at fluid borders, is such the origin of the long journey of life? Do the death rhythms of an extinguishing, cooling-down satellite of a middle-sized, middle-aged sun - somewhere in the intermediary populations of a nameless galaxy inspire functional and organizational patterns of new existences, do the basic matrices of future life patterns - influenced and imprinted by a chaos of interchanging surroundings - within a nearly unique life-expectant situation at these coordinates of a universe dare a completely new game, does a sequence of unique steps from first molecular coming together, forgetting, recognizing, not-completely forgetting, remembrance and discrimination, understanding and learning, of trial and error, failure and success, over the landmarks of cooperation, self-organization and self-reproduction, individualization, metabolism, cell-predecessors, cell-differentiation, the appearance of organism and species landscapes, up to the collective consciousness patterns of a thinking, speaking, abstracting, simulation power and creative activities developing
Biomesogens and the Grand Process
399
being lead to the present final outcome of a grand dynamic process, of ridge-climbing of selecting necessities over the endless plains of statistical coincidences, a principle, a plan, a movement of matter that optimizes itself in the growth and decay of its individual constituents, that gives the coming what it had to experience in the past, that suffers in each individual death the threatening end of its own existence and, nevertheless, sets out in each individual birth to new and unknown horizons?” [7 a]. The picture from the early 1970s is filled-in in our days. Monod’s “vagrant at the borders of a universe that remains silent to his complaints” [5 a] finds himself within the overwhelming changes of overall transitions. He channels his thoughts of loneliness and despair, the question of his destiny and his longing dreams of fellow existences within the universe into scientific approaches. He extends his probes to the depths of space and time, retraces the complexity of his ways, spells the codes that govern his existences and advances his capabilities in approaches to creating in partial simulations some sort of minima vita on his own. In his strange meso-position between elementary particles on one hand and cosmic dimensions on the other he tries to cover with quantum mechanics and general relativity theory the extremes, so far, however, being unable to formulate his desired unifying view of a final theory. Having changed his considerations from a more static, infinite universe to the more dynamic expectations of its beginning, developing, and ending, he questions again the laboriously achieved convictions by the more recent picture of grand fluctuating universes that combine dynamic developmental processes with the new - old quality of a forwarding eternal being [ 1- 81. He tries to answer the questions about the “before and behind of space and time” by
400
VIII
Living Systems
Figure 8. Universal processings versus processing of universes? [l-81. Background modified from [3].
Figure 9. Birth of universe - a first phase transition between nothingness and existence? [ l - 81.
the eternal dynamics of self-reproducing inflationary universes (Fig. 8). He tends to describe the birth of “his” universe in nonclassical versions as a phase transition between quantum vacuum and material existence, realizing instability-mediated possibilities of events, rather than in the former classical terms of an incomprehensible singularity obeying deterministic laws. He delights in the idea that all this might have been brought about by a first grand spontaneous fluctuation that induced by matter - gravitation compensation an almost-zero-energy transition (Fig. 9). Realizing his self-awareness and self-consciousness, he would look upon his maternal universe as a likewise unique individuality, and sometimes he indulges in the hybris of an anthropological
principle that favors his singularity as the final destination of a universe. His dreams, however, would prefer the intriguing suggestion that it is his self-consciousness that awakened a universe [ 1- 81. Anticipating at least the developmental aspects in a compromise between the tensions produced by these theses and antitheses, independently of whether they arose, together with space and time, in a first dramatic singularity, or in the new qualities of a grand fluctuation creating the universe as a spatiotemporal part of an eternally developing being, the grand process of our evolution - at least within our scope of space and time - seems to have endeavored for a period of 15-20 billion years to gain a certain consciousness and understanding of itself [l-81. Our roots thus reach back to the depths of the past. Together with the universe (or parts of it), life patterns originated from an alien phase transition in the incomprehensibilities of a first grand expansion. We were part of it at the very outset, and we will share the final destiny. Asymmetries of the developing patterns produced dynamic directionalities. Together with general amphiphilicities, they provided new qualities of dynamic order far from thermal equilibrium, advantages that constituted prerequisites for recognition, self-organization, self-reproduction, information processing and optimization [5-81. Leaving the mysteries of the first beginning of our universe fully intact, the following scenario might perhaps provide the framework for the evolutionary developments (Fig. 10). Partial freezing of originally unified forces, accompanied by corresponding symmetry breaking, seems to have transformed the primarily symmetric grand unifications of our universe into the diversifications of its present appearance [ 1-41. By the subsequent freezing of gravity, strong and elec-
2
Figure 10. Expansion physics and freezing of originally unified forces [2].
troweak interactions, the grand process evolved through the GUT, the electroweak, the quark, the plasma and elementary particle eras into the current period of atoms. A multitude of heavier atoms, burnt in the hellfires of stars and liberated in their catastrophes, engaged in quite different chemical interactions and in this way created hierarchical patterns of increasing complexity [ 18, 171. The chaos, however, appears to be predetermined by string inhomogeneities and a strange message from the inherencies of the universe. Among the four forces (gravity, strong, electromagnetic, and weak interactions) governing the world of elementary particles, the weak interaction and its unification with the electromagnetic interaction to the electroweak force exhibit a strange characteristic. Contrary to the other forces, which display parity-conserving
Biomesogens and the Grand Process
40 1
symmetries, the electroweak force - mediating by W+‘- and Zo bosons both weak charged and neutral currents - endows the whole process with a parity-violating asymmetric component [311. Not only atomic nuclei, but also atoms and molecules, as well as their multifarious aggregations, are sensitized to the special message. Static and dynamic states of enantiomeric species are distinguished from the beginning by a minute but systematic preference for one enantiomer and discrimination against its mirrorimage isomer. Amplification mechanisms traveling long evolutionary roads elaborated the first weak signals into dominant guiding patterns [7, 17, 181-11, 311. The freezing of strong chemical interactions at the interfaces of phase boundaries and by this also the freezing of the special characteristics of their spatiotemporal coherencies into the individualities of chirally affected mesophases - liberated the richness of their folds into the directionalities of dynamic order patterns (Fig. 11). By the freezing of self-reproduction and self-amplification conditions along water-mediated autocatalytic bundles of nucleation trajectories, the appearing systems - governed by Lehmann’s “shaping, driving and directional forces” - gained, far from thermal equilibrium, abilities of information generation, adaptation, storage, processing, transfer and, finally, optimization (Fig. 12). Based on the unique amphiphilic design of their constituents, the biomesogenic patterns evolving from there developed complex structure-motion linguistics and promoted their more and more homochirally based contents in synergetic regulations. Within their adaptational and spatiotemporal universalities, the evolving life patterns retraced the impetus of the early dynamics and reflected within the developing consciousness of their chirally instructed and determined organismic organizations the grand
402
VIII
Living Systems
2.3 Evolution of Amphiphilic Patterns
Figure 11. Evolution of amphiphilic patterns - end to the 1970s [7a, 17, 181.
Figure 12. Entropy - complexity - information relationships around the A = k In Z postulates, superimposed on evolutionary polydisperse DNA as elucidated by scanning force microscopy (SFM) investigations [l-8,751.
unifications that dominated their origins. Their creativity, however, somehow aims at beyond the boundaries.
Views derived from the structure -phase duality centering on solid - liquid-phase interfacial birth-zones and their contradictions. The liberation of blueprinting holographic constituents of amphiphilic transient order-disorder patterns and their mesophase coherences, order - disorder (rigidity -flexibility) gradients of molecular individuals and their amplification into phase/domain (transition) characteristics. Ways from molecules to macromolecules and from phases to microphases centering on supramolecular biomesogenic organizations. Supramolecular structure and systemic biomesogen relationships. Polyelectrolyte/water patterns competing for informational, functional, compartmental, and mediating components and combining individual contributions into integrative systems. The interchange of transiently acting multi-solvent - solute systems between non-classical chaos foundations and fundamental complexity goals, waiting for energy input to set all their enthalpy- and entropy-determined functional and informational richness “alive”. Entropy counting, complexity connecting, and information evaluating microstates. Guidelines between symmetry and asymmetry, ambivalence and directionality; autopoiesis and system evolution. Transition to life and spatiotemporal coherences. These examples show that it also seems to be possible to describe evolution in mesomorphic terms [5-8, 13, 17, 18, 30-351.
2
Understandable only as the as-yet last and most highly sophisticated derivatives of our universe, life patterns developed and displayed in their growing complexity all the abilities from which they had originated and that contributed to their further development. The predecessors of the highly advanced life patterns of our days learnt their first lessons while endeavoring to blueprint their maternal inorganic matrix systems [32] in .all their informative and functional richness, and they have been further educated in their attempts to follow the changes in their environment. Between the thesis of solid order and the antithesis of liquid disorder, capabilities of optimizable information and function processing seem to have been opened to amphiphilic systems that were able to develop flexible and adjustable structure - function correlation of their own by virtue of the multitude of their transient interchanging phase-domain organization. They gained individuality in the successful integrative handling of complex informational, functional, and compartmental patterns together with sensitive assistant water media. They reached developmental facilities by molecular recognition, intra- and inter-phase/domain relatedness, self-organization and self-reproduction, information processing and optimization far from thermal equilibria. Among the molecular species screened by evolution in a Darwinian selection for suitable constituents of first dynamic reality-adaptation and, later, reality-variation and creation patterns, amphiphiles with specific hydrophilic -hydrophobic and order disorder distributions - sensitized to the chiral message of the electroweak force were the preferred survivors of the grand process (Fig. 13) [17, 18, 331. At the beginning, a rather omnipotent biopolyelectrolyte pool, dependent on the phase dimensionalities of its outsets, pro-
Biomesogens and the Grand Process
403
Figure 13. Evolutionary strategies according to the dirnensionalities of arnphiphilic patterns (top to bottom): one-dimensional: information; two-dimensional: compartmentation; three-dimensional: function [5, 6, 7a, 17, 181.
vided informational, functional, and compartmental components. By preintelligently developing their individual molecular asymmetries into dynamic system directionalities, the mesogenic constituents of the developing chiral amphiphilic patterns succeeded in a fertile and creative synthesis. Avoiding hyperstatics and hyperdynamics, the disadvantages of the extreme states that contradicted their origins, they developed the creative meso-positions of ongoing dynamic order. Optimizable free-energy stategies on the basis of their molecularly imprinted affinity patterns selected, by preintelligently handling the entropic order - disorder gradients, patterns of chiral mesogenic backbone structures (Fig. 14) [ 171. It has been the exceptional usefulness of these side-chain backbone arrangements, the elite successors of the first amphiphilic blueprints of maternal inorganic matrices and their inherent informational and functional capabilities (Fig. 15), that allowed
404
VIII
Living Systems
Figure 14. Chiral biomesogenic backbone structures as survivors of a Darwinian selection for optimized homochiral transient order-disorder distributions: (top left) stacked nucleic acid single strand, (top right and bottom left) extended protein single strand, (bottom right) membrane component with a small but nevertheless chiral backbone [ S , 6, 7, 17, 18, 331.
Figure 15. Orientational presentation of liberation of solid-phase information content into self-replicating biomesogenic information patterns, neglecting complex intermediate states of highly condensed information channel designs: (left top to bottom) solid-phase space-partitioners: Ca,[AI2Si,,O,,], Li,,MgSi,, Ta,CI,, [32c]; (right) replicative DNA, including water cover and counterion clouds [33 a, c, p, q].
interactive division of labor, within mediating water swarms, into the specializations of information, function, and compartmentation, conserving, however, beneath the skin of their special adjustments, the continued primitive universalities of their origin [7a, 17, 181. Thus, while a first swift glance might connect the structural features of nucleic acids with information, that of proteins with (information realizing) function, and the remaining characteristics of membrane components with (information and function securing) compartmentation
[S-8, 17, 181, a closer look at the three major components of biopolymeric and biomesogenic organization reveals much broader ranges of different abilities. For instance nucleic acids exhibit functional capabilities in the widespread landscape of catalytic RNA species (341, proteins deal with information, not only in their instruction of certain old protein-production lines of their own [3S] but also in the skillful handling of diverse partner molecules and molecular organizations - including themselves [7a, 331, and compartmental membranes demonstrate additional informational and functional potentials in the complex tools available to them [7 a, 331. Their selforganizational and self-reproductive facilities, together with their ability to cooperate, created by nonlinear dynamics a rich variety of dissipative structures far from thermal equilibrium and caused - by symmetry breaking - the transition of the whole process from its mainly racemic prebiotic period into the optimizable biotic patterns of homochirality. They regained, as expressed within their final organisms on the basis of these homochiral molecular and supramolecular designs, the symmetry beauty and operation optimization of the whole world of bichirality (Fig. 16). The grand process, however, remained subject to the dialectics from which it originated, the general chiral approach between order and disorder, and the permanent renewal and achievement of selecting a forward-directed path out of these contradictions.
2.4 Transition to Life and (Bio)mesogenic Reflections Transitions to life and preintelligent operation modes; self-replication and self-reproduction between square-root laws and auto-
2
catalysis; transitions from hetero- to homochirality: dissipative structures and autopoieses; amphiphilicity and holographic designs; nonlinear dynamics and self-organizational forces: spatial order-disorder distributions and spatiotemporal coherences; phaseldomain transition strategies and cooperativity; reentrant phenomena. molecular hystereses and memory imprints, oscillations and rhythm generation; signal transduction and amplification; information processing and optimization; selfhon-self recognition and discrimination: structure - motion linguistics and system intelligence; the possibility of non-classical chaos and the emergence of reality-adaptation, variation, and creation patterns of critical, subcritical. and fundamental complexity: nucleoprotein system developments by metabolism, mutation, and selectivity; synergetics and system optimization; a transition to life almost as mysterious as the universes transition to existence - are such the development of the first blueprints of “Tala interface phenomena” to life‘? [ 5 - 8 , 13, 14, 17, 18, 23, 30-351. For a long time, science has tackled constructionist life processes with rather reductionist approaches. Our present investigations of pre-life states. transitions to life, and the developments of life patterns, unraveling pictures of utmost complexity, seem incapable of suitable elucidating the extensive coherences. “The whole being not only more than the sum of its parts, but exhibiting a superior new quality”, represents, amidst multiple guiding uncertainty principles, crucial but so far unsolved challenges of scientific endeavors [ 1, 5 -71. Chemistry, and especially organic chemistry, going against the long-held, preferentially reductionistic aims of a purely artificial carbon chemistry, and remembering its long-disregarded original intentions, has moved structures from the level of molecule
Bioniewgens and the Grand Process
405
Figure 16. Evolution of molecular and organismic chirality (top to bottom): L- and D-amino acid enantiomers [31]; valinomycin Kf-ion carrier [ 3 5 ] as a highly sophisticated outcome of independent amphiphilic bichirality protein developments; transitions from hetero- to homochirality via selection of enantiomers in building up suprachiral structures of RNAs and their ?elf-replication 1341; development of structural and functional asymmetric organismic bichirality on the basis of molecular homochirality [7 a, 17, 18,311.
to macromolecule to the creative fields of supramolecular organizations, where it has met the attempts of physics to elucidate idealized supreme phase characteristics in the underlying dynamics of microphase and
406
VIII
Living Systems
domain systems. Thus, the area of preintelligent self-organisational, self-informational, and self-reproducible pattern as a result of structure- (micro)phase coevolutions [7a, 171, anticipates with first scienfic excursions 136- 501 to “minimal replication” [44] and “minima vita” models 1491, within the complexity of their interacting molecular individual “parts”, presentiments of the “whole”. The biomesogenic approach [7 a, 171, proposed nearly two decades ago, tries to shed light on (pre)biologically relevant, highly condensed systems. Within the irreversibilities of non-equilibrium processes, supramolecularly structured and biomesogenically amplified sensitive, transient order- disorder patterns, due to their inherent entropy gradients, appear to be capable of properly utilizing energy dissipation for transformating entropic possibilities from a chaos of creative facilities into sense-giving information patterns. The apparent inconsistencies in the description of the animate and inanimate worlds uncover insufficiencies in our underlying categorizations. The current attempts to redefine terms such as entropy and information, chaos and complexity, matter and consciousness within equilibriudnon-equilibrium contractions might promote not only a better understanding of pre-life states, transitions to life, and life processes themselves, but might also found a more consistent picture of life singularities within the evolving creative universe. “Structure” rather than “phase”, however, maintained its leading role for scientific progress for a long time [ 1- 8, 17, 301.
2.4.1 Molecular developments Life-pattern developments have been promoted by structure -phase dualities. Struc-
tural contributions, however, rather than (bio)mesogenic approaches, constituted the “golden age of molecular biology”. Basic geometries of biopolymeric species, nevertheless, correspond to classical structural liquid-crystal motifs. Structural interplays, resulting from the earliest outcomes of proteins’ own information and production lines up to the nucleation of nucleoprotein systems and their further destinations, display common mesogenic features and mesomorphic operation modes [7 a, 171. Miescher’s prophetic view of his “Nuclein” as the genetic material [29a] and Virchow, Mettenheimer, and Heintz’s independent findings of behavioral abnormalities [20] could not have been valued at a time when proteins (IIpwzoy- the first) were the favored species and structure the ultimate goal of people engaged in “life sciences” [36]. And although some early intuitive perceptions of protein statics and dynamics (not to mention their intimate relatedness) appear convincingly modern [23], the real structural designs of the patterns of life remained obscure for nearly a century. Somehow indicative of the “optical disposition” of the species Homo sapiens, it was not the admirable achievements of the classical period of natural products chemistry [36], beginning, with Fischer [36a] and accompanying with its very subjects also the historical birth of liquid-crystal phase chemistry at the dawn of the century, it was not the highlighting work of Eschenmoser [36d] and Woodward [36c] half a century later and not the lonely prophecy of a physicist that an aperiodic crystal is the genetic material [29 c], and it was not even the mystic secret formulae of Chargaff [29e] concerning the base-pairing schemes of nucleic acids and anticipating the whole story to be further elucidated that caused the dramatic scientific “phase transition” in the 1950s. It was the perception of the two “ho-
2
ly” chiral structures: Pauling’s protein a-helix [29d] and Watson and Crick’s (as well as Wilkins and Franklin’s!) DNA double helix [29 fl that introduced a new age (Fig. 17). The first elucidations of basal chiral geometries of the nucleoprotein system and their immediate projection into the multifold molecular landscapes around was an event that irrevocably altered our views of life patterns in nearly all aspects. In these chiral structures and their biomesogenic relatedness molecular biology set out for far horizones, and the beacon of their structural beauty and informational and functional transparencies has enlightened scientific approaches up to our day. Since those times the landscape of nucleoprotein geometries has been increasingly enriched with quite different structural motifs, classified into the basal units of “secondary” structure, determined mainly by hydrogen-bond pattern, their “supersecondary” combination schemes, their “tertiary” adaptations to the stereoelectronic prerequisites of real three-dimensional entities and, finally, their aggreagations into supramolecular “quaternary” organizations and beyond [5-8, 29, 331. The realms of both nucleic acids and proteins (Figs. 17, 18) display a certain predilection for the flexible mesogenic filigree of helices - the dynamization of the supreme symmetry of a circle into the asymmetries of space and time, thus connecting both structural and developmental aspects. In the case of proteins, the helical designs vary from preferred right-handed single-stranded versions, via more restricted right- and left-handed single- and double-stranded arrays to special forms of left-handed helix triples. For nucleic acids, the structural landscape is dominated by right-handed antiparallel double-helix motifs, enriched by differently intertwisted helical triples and quadruples and contrasted by alien
Biomesogens and the Grand Process
407
Figure 17. Biomesogenic nucleic acid helix families (top to bottom): ds B-DNA; ds A-RNA; ts DNNRNA; ds Z-DNNRNA [7a, 33 a, c, f, p. q].
versions of left-handed antiparallel double helices. The antitheses of more rigid basic secondary-structure motifs appear in the chirally twisted parallel and antiparallel p sheets of proteins and will find some correspondence in the cylindrically wound-up double-helix design of the nucleic acid Afamilies as well as in the so far somewhat dubious versions of suprahelical Olson-type arrangements [5-8, 17, 18, 33, 5Ogl. In the rivalry between the D- and L-enantiomers of protein and nucleic acid pattern constituents as well as in their chiral amplifications into larger supramolecular motifs, the “less fortunate” enantiomers, energetically discriminated by the chiral instruction of the electroweak force, fought a losing battle from the beginning of the grand evolutionary competition [311. Ab initio
408
VIII
Living Systems
Figure 18. Biomesogenic peptide helices of a) POly(L-glutamic acid) and b) poly(L-lysine) with partial water cover, views along and perpendicular to helix axes [7a, 33g-1, p, q].
calculations of the preferred aqueous-solution conformations of some a-amino acids and glyceraldehyde, the classification-ancestor of all sugar-moieties, are indicative of a stabiliztion of L-a-amino acids and D-glyceraldehyde relative to their discriminated mirror-image enantiomers by some Jmol-’. The same seems to hold true
for the a-helical and Psheet patterns of La-amino acids in comparison to the structural mirror-images of their D-antipodes. Absolute weighting by the electroweak force, further selections by different amplification mechanisms, as well as the rejection of mirror-image species in building up supramolecular chiral arrangements, together with energy minimizations of the side-chain tools in both proteins (the 20-1 proteinogenic side-chain versions) and nucleic acids (the four-letter alphabet of the nucleobases) selected, finally, for the preferential appearance of L-a-amino acids’ right-handed protein-a-helices and p-sheets and favored (2’-deoxy)-~-ribosemoieties’ right-handed antiparallel double helices of nucleic acid families [7a, 17, 18, 311. And it is also due to these selection and optimization procedures that poly-L-proline backbones can be forced into left-handed helical versions, and special alternating sequences of nucleic acid designs tend to adopt the strange double-helical left-handed Z-motif. All these chiral structural standards and the corresponding families around them in proteins and nucleic acids represent, however, only the main building blocks upon which evolution has been operating. The colored loop- and knot-stretches and all the more disordered parts of protein and nucleic acid structural designs are of primary importance for the evolutionary aspects of informational and functional processings in our evolving life patterns. It fits nicely into the picture of dual structure-phase views of biomesogenic organizations that objects of “rod-like appearance”, for instance the little “world’ of the tobacco mosaic virus (which-its overall design reduced to a simple rod-like entity -became the starting point of Onsager’s theory [37]), as well as much simpler protein and nucleic acid helices are typical mesophase formers in the classical liquid-crystal phase
2
sense [7 a, 17,251.As the transient final “order” in self-recognition, self-organization, and self-reproduction is approachable only by means of mediating “disorder” steps, Fischer’s more static lock-and-key model [36a, d-f] turns by its own inherent “key”dynamics into Koshland’s “induced fit” [36g]. It is within the frame of its close connections with the preinstruction of biomesogenic species by the chiral message of the electroweak force that the transformation of chiral molecular units into the power of “supra-chiral” macromolecular patterns, and from here to the further extended chiral expression of supramolecular phase organizations and their autocatalytic self-reproductions, provided a useful sequence for the rescue and elaboration of an early message of our universe from the noise of its first hidden appearances up to the homochirality as a matter of course nowadays [7 a, 1 7 , 3 0 , 31, 471. The “central dogma” [8c, 29f, 33p, q], proposed by the creators of the DNA double-helix model in order to channel their considerations of the information flow in biological systems, was partially inverted in historical sequence. Its alpha and omega determined the evolutionary action scheme: variation in the information content of the genotype (DNA) and subsequent functional selection in the phenotype (protein). “Dynamic proteins” and “static nucleic acids” competed, as in the “chicken and egg” problem, for historical evolutionary priority [5 b]. But while the original alternative found its solution in both the informational and functional “hypercycle” [5 b], developing its space-symmetry beauties into timehelical creativities, and while dual structure-phase views - contrary to classical perceptions - foresaw much closer spatiotemporal coherencies for the evolving mesophase systems [ 7 , 171, while RNA turned out to be in some cases DNA-informative
Biomesogens and the Grand Process
409
[8 c, 33 p, q], the old chicken and egg problem presented a futher surprising aspect of the “meso”-positions of the so-far unfortunate and disregarded RNAs, a view of what might be called an early RNA world [34]. The evolutionary highways in the development of the DNA -RNA-protein triad, however, seem to have been preceded by a peculiar intermezzo. The ambitious attempt of proteins of cover self-consistently both informational and functional capacities has survived up to our days in some unusual and, despite all their structural restrictions, admirable molecular systems [35]. A group of depsipeptide and peptide carriers and channels as well as their possible subunit structures - produced on old, both informational and functional protein lines - display enchanting achievements in the skillful handling of biomesogenic operation modes [7 a, 17, 351. Interestingly, all this was brought about within any preference of a special homochirality. On the contrary, it is just the surprising selection of alternating different chirality patterns from the prebiotic racemic pool and the careful and intelligent order disorder design of rather small molecular entities with adjusted alternating chiral codes that made such beautiful arrangements as, for example, the “bracelet” of valinomycin work (Fig. 19) [ 181- n]. In the impossibility of bringing the giant information content not only functionally but also informationally alive, the ambitious enterprise of proteins, the “first” to create by the power of their own legislative and executive intelligent “bichiral” patterns, ended up in an evolutionary impasse - notwithstanding all the convincing achievements that match not only intriguingly later-to-be-developed operational modes of the DNA-RNA-protein triad but also our own intelligence. While proteins thus failed in their omnipotence, the apparently less qualified RNAs seem to have built up a preliminary, medi-
410
VIII Living Systems
Figure 19. Bacterial representatives of active iontransport through membranes, modeling phase-transition strategies of protein information function lines: stereo-presentation of a CPK-valinomycin “movie”, mediating by highly sophisticated biomesogenic interplays a K+-ion membrane passage [7a, 33p, q, 351.
ating, and progressing RNA world [34]. “A tRNA looks like a nucleic acid doing the job of a protein.” Crick’s pensive considerations became reality: mRNAs - the differentiating blueprints of the DNA information store, tRNAs - the structurally and phase/ domain disrupted mediators of nucleoproteinic interrelationships, and rRNAs - the long-blamed “structural support” of ribosome protein-synthesis machinery, not only
did they all develop a more and more interesting structure-phase “eigenleben”, RNAs also proved to be both informational and functional. With their special characteristic of molecular hysteresis [48], they were among the first to provide a basic understanding of memory records, oscillations, and rhythm generation in biological systems, and they supplied - starting with their catalytical facilities in self-splicing and ending up with general “ribozyme” characteristics - successful predecessors and competitors for proteinic enzyme functions in both basic and applied research. RNAs combining in their early forms both genoand phenotypic aspects - seem to have mediated the grand DNA - RNA - protein triad into its present expressions and survived up to now as an inherent principle of life patterns [17, 34, 39-44,48, 501. The final ways to utmost complexity in informational and functional processings are illustrated by impressive landmarks. The early prebiotic interactions of the shallow groove of sheet-like right-handed A-type RNAs with the right-handed twisted antiparallel /%sheet of a peptide partner might have established the first fertile intimacies between nucleic acids and proteins [38]. Within this hypothetical first productive organization, the archetypic protein-psheet, with its suitability for easily sorting polar - apolar and hydrophilic - hydrophobic distributions, as well as the archetype ARNA, with its partially complementary informational and functional matrix patterns, could “live” first vice versa polymerase capabilities and thus mutually catalyze early reproduction cycles, connected with chiral amplifications and informational and functional optimizations (Fig. 20) [7 a, 17, 18,38,43]. It proves possible to further extend this model to protein-psheets fitting into the minor groove of B-DNA, connecting by this geometry and functionality the
2
Figure 20. Nucleic acid and protein versus polymerase efficiencies. (top) RNA and (bottom) DNA with chirally twisted Psheets in the shallow and minor grooves, respectively [38], perhaps as an outcome of Katchalsky’s catalytic montmorillonite solid-phase patterns [7 a, 32 a].
A/B-alternatives of nucleic acid families by mediating P-protein design [38]. The last step to liberating all the inherent potential of these early nucleoproteinic systems of RNA and protein was found in the dual conformational abilities of the small 2’-deoxy-~-ribosecycle to account for a conformationally amplifying switch between A- and B-type nucleic acid versions [7 a, 8 c, 331. The discovery of the information store of DNA enabled the systems to separate the replication of a more densely packed DNA message from the informational and functional instruments needed in the hitherto existing nucleic-acid- protein interaction schemes and promoted by this
Biomesogens and the Grand Process
41 1
transcriptional and translational operation modes with far-reaching maintenance of so far elaborated RNA - protein cooperation. The new qualities were brought about by the structural deepening of the shallow RNA groove into the minor groove of DNA, which allowed the continuation of the successful, however, rather unspecific, RNA protein contacts together with newly established small effector regulations, and, still more important, by the opening of the sofar hidden huge information content of the RNA deep groove into the much more exposed major groove of B-type DNA, which admitted specific DNA -protein recognitions [7a, 17, 18, 331. All these newly developed and achieved operational capabilities brought about a dramatic transition from the unspecific contacts between proteins and nucleic acids in their shallow and minor grooves, respectively, to specific recognition and functional information processing, mostly exemplified by protein helices in the B-DNA major groove. The resulting DNA - RNA -protein triad - mediated by intelligently handled clusters of water molecules - promoted the development of universally applicable informational and functional patterns, which not only developed their order - disorder distributions into optimizable operational modes of biomesogenic systems, but responded, moreover, intelligently to the early chiral message from the inherent qualities of the universe and its translation into “molecular creativity” [7 a, 17, 30, 33, 431.
2.4.2 Molecular Matrices The projection of individual capabilities into collective cooperation modes; self-replication and self-reproduction in highly condensed states - suggestions for naturally occurring systems, simulations by artificial
412
VIII
Living Systems
Figure 21. Biomesogen regulations between structure and phase (left to right and top to bottom): visualization of biomesogen polyelectrolyte regulations exemplified by DNAlwater-shelI/(hydrated) counterioncloud pattern statics and dynamics; arbitrary DNA/ counterion-cloud/water-domain arrangements symbolizing operative biomesogen nucleations “between structure and phase” [7 a, 29, 33 a, c, f, p, q].
systems; square-root laws and cooperation of partners versus autocatalysis and decisive selections of competitors; inherent coherences and judging bifurcations - structure-
phase duality miracles abound in the pretransitions and transitions to life [5-8, 17, 31, 39-43]. It was the new quality of dynamic order of coherent matrix systems that established, within the transiently changing patterns between flexibility and rigidity, order and disorder, growth and decay, guidelines for self-organizational behavior and self-reproductive developments. The crucial step that had to be achieved by the candidate patterns of life, however, was the evolutionary development of self-reproduction systems that could not only propagate their dynamic order states within the surrounding disorder but would even make use of the creative possibilities of uncertainty principles within non-classical chaos for their own informational optimizations [7a, 431, a first decisive step that unravels as well all the other life developmental processes to follow that make use of mesogenic capacities for mediation and continuation of the grand process. “It has not escaped our notice that the specific pairing we have postulated immediately suggests a possible copying mechanism for the genetic material” - rarely has anything fascinated science as much as the “holy” structure of DNA, born in that classical understatement of Watson and Crick [29 fl, in which the prophecy of Schrodinger that an “aperiodic crystal” was the genetic material became reality (Fig. 21) [29 c]. The discovery of the structure of DNA irrevocably altered our view of life and represents one of the really major landmarks in the development of science. The prototype of a dynamic matrix became the beacon that illuminated research in the fields of molecular biology and redirected organic chemistry, which had, contrary to its first ambitions, developed mainly as a chemistry of artificial carbon compounds, back to its early aims. The duplication of the holy structure represented the key mechanism for prebiot-
2
ic systems on passing the boundary between the inanimate and animate worlds. The accompanying scientific views, however, elucidated mosaics of processing of forms rather than integrative visions of forms of processing. Todd [39a] sought an answer for the somewhat excluded organic chemistry in a landmark appeal: “The use of one molecule as a template to guide and facilitate the synthesis of another ... has not hitherto been attempted in laboratory synthesis, although it seems probable that it is common in living systems. It represents a challenge, which must, and surely can, be met by organic chemistry”. And this grand concept, indeed, appeared to be a good sign. Quite soon Schramm et al. [39 b] aroused enthusiasm with the “noenzymatic synthesis of polysaccharides, nucleotides and nucleic acids, and the origin of self-replicating systems.” Schramm’s experiments - interestingly, brought about in highly condensed states had to push the evolutionarily optimized standards back into the high error rates of their prebiotic beginning. Purely artificial template and matrix experiments that evaluated all the interactive possibilities of both covalent and non-covalent nature, however, were devoid of even this possible relationship to evolutionary developments. The first great “chemical” departure into the temptations of matrix reactions [39] ran aground on innumerable difficulties. Only a few special cases of matrix reactions found successful applications. The use of coordination matrices reminiscent of solid-phase informational and functional patterns, and already known from stereopolymerizations, advanced research on complicated natural products, for example, chlorophyll and vitamin B [50 a, c]. Nucleic-acid-derived recognitions enriched the informational and functional landscapes of mesogenic systems [17, 181. The main goal, however, to follow
,
Biomesogens and the Grand Process
413
evolution on the supposed way to make polymers on suitable templates in zipperlike reactions - ranging from covalent to noncovalent variants - failed due to the insurmountable complexity of the systems involved. In vain, all attempts to model nucleic acid self-replication properties by nucleic acid analogous polycondensation, polyaddition, polymerization, and polymer analog reactions [7a, 17, 18, 431. In vain, for instance, the once so intriguing idea of blueprinting natural patterns independently of natural polycondensations by the help of artificial matrix radical reactions of N-vinylnucleobases that linked together man-made polymerization groups and evolution’s recognition patterns [39f-i]. The dominant element in scientific progress turned out not to be the tremendous variety of chemical matrices and the distant aim of building up systems capable of selfreplication [39]. Rather, fascination with the seemingly effortless elegance of the native structure -phase prototypes stimulated the vanguards of chemistry and biology in their campaign of molecular penetration of biological systems. Rapid scientific breakthroughs introduced the “golden age” of molecular biology [40]. Oparin and Haldane’s [40b, c] heirs, Eigen and Kuhn [40f, h], gave these events a time perspective and with their “information” described the vector of the grand process of evolution. Along also with their hypercycle-view [5 b, 40h], a multitude of rival problems, which had been simmering in the collective subconsciousness, resolved themselves out of the complexity of the original goals of template experiments. Self-replication, mutation, and metabolism, as prerequisites for selection, made up the list of criteria, and through these, information and its origin, its evaluation and processing, and finally its optimization governed the evolutionary history of prebiotic and biotic systems. The two
414
VIII
Living Systems
ways of scientifically approaching the problems, the use of natural systems [40,41] and the “ab initio” chemical outsets [42], were to find their followers [43 -501. In connection with Spiegelman’s initial experiments [40e], Eigen and Schuster [5 b, 40h], Joyce, and others [41] attempted to “bring to life” the theoretical premises behind the laboratory realities of enzyme-catalyzed RNA replications and evolutionary experiments. Orgel’s group, by contrast, exerted itself of transform diverse matrix relationships into artificial enzyme-free nucleic acid formation [42, 43, 44 b]. When Altmann [34 a] and Cech [34 b, c] -confirming earlier speculations by Crick, Orgel and Woese [7a] - raised RNA to the throne of an archaic informational and functional omnipotence, it amounted to a late justification of the toil and trouble of this “nucleic acid first” [34, 421 route. Self-splicing RNA complexes, polymerase activities, ribozymes and hypothetical RNAsomes (views partially extendable also to DNA) afforded unusual insights into the genotypic and (!) phenotypic behavior of a single nucleic acid species [34 a - el. And with the supposedly greater understanding of this “RNA world” fresh impetus was given even to the purely chemical approaches, once so promosing and now almost forgotten. A rapidly growing number of artificial self-replication models, covering nucleic acids as well as their near and distant analogs, but also successors of the nearly forgotten Fox microspheres [40 d] in the field of membrane components, are representative of the wealth of variety of the creative outburst and clarify with their different degrees of abstraction one of the dynamic peripheries of our times [43]. Through elucidation of the static and dynamic principles of replicative information systems, all these models endeavor to gain insights into both the transitional stages
between chemical and biological evolution and the possibilities of artificial developments of their chemical simulations. After innumerable attempts to explain the matrix relationships of mono- and oligonucleotides to orientating and catalytic oligoand polynucleotide templates, two self-replicating nucleic acid models (a DNA-analogous hexamer system by von Kiedrowski et al. [44a, c-d, f, g-i] and an RNA-analogous tetrameric assembly from the Orgel group [44b]) opened new ways (Fig. 22) [44]. Von Kiedrowski’s first successful model - hexadeoxyribonucleotide duplex became a leitmotif in the detailed treatment of growth kinetics and anticipated later selfreplicating “minimal systems” in many ways. By using the ligation of oligonucleotides, he exploited their higher cooperativity to attain increased stability of the matrix duplex and the ternary formation complex. The choice of palindromic systems reduced the complexity of biological replication experiments to a simplified kinetic measurement of its identical (since self-complementary) matrix components. Demonstrating the autocatalytic behavior of his systems, he was able to obtain direct proof of self-replication. The surprising square root law of matrix growth kinetics (curiously even computer viruses seem to be in love with it) with its ideal case of a parabolic reaction course, derived from the hexadeoxynucleotide duplex, which was confirmed slightly later by the Zielinski/Orgel system [44 b] and, finally, also verified in the Rebek dimer assembly [45], was recognized as autocatalytic system behavior of self-replicating oligonucleotide templates under the constraints of isothermal conditions, where stability relationships of matrix reaction partners might exclude the expected exponential growth kinetics. Later functional and reactive refinements overcame certain shortcomings of autocatalytic reaction
2
Biomesogens and the Grand Process
415
channels that were too heavily weighted by spontaneous noninstructed syntheses. A hexameric system, constructed from two trimeric blocks by phosphoamidate coupling, revealed sigmoidal growth behavior and stood out on account of its extraordinary autocatalytic efficiency. Its present extension to a three-matrix block will not only enrich the informational content of the model system but seems for the first time to invite studies of selection behavior [40 h, k, 441. The patterns of autocatalysis with respect to parabolic and exponential reaction courses, which strongly affect the conclusions of Eigen’s evolution experiments concerning the decision criteria for mutant selection and coexistence [5 b, 40 h, k], can now be derived from the thermodynamic data of the matrix patterns and their reactivities, and offer quite new views, with autocatalytic cooperation between competitive species [40 k]. Separate from “enzyme-catalyzed” evolution experiments with RNA and DNA systems, basic questions of prebiotic behavior can for the first time become the subject of detailed experimental research [40 k, 431. While continuing their studies on more complex autocatalysis patterns, von Kiedrowski et al. diagnosed modulation of molecular recognition as an operational deficit of earlier artificial self-replicational nucleic acid systems with regard to exponential reaction courses, and identified it as an ideal aim for future models [44]. On its way to the nucleoprotein system, evolution must have Figure 22. (Bio)mesogens approaching pre-life states [7a, 17, 18, 431: a) von Kiedrowski’s and Orgel’s “minimal models of replication” on the basis of selfcomplementary oligonucleotide DNA and RNA systems [44 a-d, f- h]; b) distant nucleic acid strand-analogs as matrix reaction models [7a, 18, 19, 39f-i]; c) Rebek’s self-replicational and evolutionary nucleoside analog model [45]; d) von Kiedrowski’s self-replicational amidinium-carboxylate model, being suggestive of exponential growth kinetics [44e]; e) Lehn’s
successful mesomorphic system for demonstrating surprisingly efficient homochiral selection in building up supramolecular arrangements [47 a]; f) Lehn’s helicates, commemorative of early information transfers between inorganic matrices and organic ligands [47b-el; g) Luisi’s first “minima vita” approach of chemical autopoiesis, modeling self-replicational evolutionary states on the basis of compartmental and functional arrangements [49].
416
VIll
Living Systems
had a similar view of the problem, when, in Lucretius’s “external game” [40 a], it endowed nuclei acids, which are somewhat insufficient in this respect, with proteins, experienced in phase and domain regulation control, and thus achieved an ideal milieu for directing modulations of recognition [5d, 7 a , 17, 18m, n]. And, indeed, at this point something of what might have favored nucleoprotein systems so extremely in comparison to replicative states of nucleic acids themselves will become apparent. While, presumably, the both informational and functional RNAs achieved the first successful self-replications on their own 1431, they seem to have been outclassed i n future developments by the cooperative efforts of nucleic acids and proteins. The urgent need to establish suitable and reliable regimes of strand-recognition, annealing, separation, and reannealing, which had thus far only been easily brought about by drastic variations of reaction temperature, seem to have been accounted for under stringent isothermal conditions only by the complex patterns of nucleoprotein systems [7a, 17, 181. With no directing man-made artificial chemistry and physics at hand, the general evolutionary breakthroughs in self-replication had to await the play-educated abilities of complex biomesogenic organization. It was especially the impressive capabilities of proteins both functional and informational - which display with more then 20 side chains a great variety of affinity and entropy variations [7 a, 331, that could provide suitable systeminherent isothermal conditions. They did this, in close cooperation with their preferentially informational, but also functional, nucleic-acid matrix mates and mediating water patterns by the highly sophisticated transiently acting mixing of domain-modulated paths along desirable reaction coordinates. Only the integrative efforts of nucleic acids and proteins, which sub-
merged their structural individualities into the biomesogenic unifications of the functionally and informationally completely new characteristics of transiently acting order - disorder patterns of nucleoprotein systems, reached the desired optima in selforganization and self-replication, in general information-processing and optimizations [17]. Matrix growth kinetics as known for oligonucleotides were followed by a drastically abstracted artificial replication system developed by Rebek’s group [45]. This constitutes, in some sort of distant nucleic acid/protein analogy, with the two nucleicacid systems the pioneering triad i n this field, and seems to augur, after all that has gone before, a fresh departure for the far horizons of artificial chemical evolution systems. The dimer assembly developed from host-guest relationships combined the interactive and cognitive possibilities of native nucleic acid matrices with the more general biopolymer relationships of the amide bond formation involved i n the matrix reaction. “Kuhn divergences” [40 fl generate a tremendous number of new matrix and replication variations, and from this breadth develops an impressive range of impending convergences. 2-Pyridone derivatives [46c], already foreseen in mesogenic and polymeric nucleic acid analog approaches [46a, b], reoccur in new guises 1461. The successors to elite nucleic acid templates are surprisingly the apparently simpler amidinium-carboxylate matrices [44e]. Astoudingly, however, in the attempt to reduce generalizing principles as far as possible, desirable operational modes become accessible. The dramatically abstracted amidinium-carboxylate systems, which, nevertheless, cover certain essentials of complex nucleic acid - protein interactions, have proved to be susceptible to molecular recognition modulation and even seem to
2
display -to the delight of their examiners exponential growth kinetics. Beautiful chiral main-chain LC-polymers - built up as impressive supramolecular helical arrangements from bifunctional recognition units convincingly confirm in the hands of Lehn [47 a] the selection and discrimination capabilities of supramolecular organizations in transitions from hetero- to homochirality [7a, 17, 18, 311. Detailed aspects, such as the possibilities of coordination matrices in native nucleic acid assemblies, make themselves felt in helicates [47 b - d], whose structures reflect also relationships between our life process and its basic matrices [7 a, 321. Matrix studies of triplex systems [48a-q, t] model possible strategies of nucleic acid organization and by this detect not only protein-like behavior of RNA-Hoogsteen strands in reading informational DNAduplex patterns [48 1,0] but also bridge the gap of the basic hysteretic mechanism of information processing in more highly condensed systems [48a-q]. The paths of chemistry - from the chemical bond to the chemical system, from molecularity to supramolecularity, from statics to dynamics are united with the spatiotemporal coherences of phase and domain regulation strategies from physics in the asymmetries and nonlinearities of life sciences, affording processes that are stabilized by their own dynamics. From the earliest prelude to the unknown intimacies of nucleation of the nucleoprotein system up to today’s complex evolutionary character, only the to some extent adequate complexity of the scientific approaches might provide sufficient stimulation potential to approximate the grand process. It is fully within this context when, in addition to the nucleic acid-nucleic acid analog pioneering triad of self-replication systems, a fourth innovative approach adds to the “minimal systems” of preferentially
Biomesogens and the Grand Process
417
informational replicators [43, 441 the new view of a “minima vita” challenge [40 I] related mainly to functional and compartmental aspects. It appears somehow as a reincarnation of Fox’s microspheres [40d], when micelles, as representatives of the compartmental partner in life games, are proposed by Luisi et al. as first examples of “minimal life” models [49], where, independently of all historical outcomes, the chemical autopoiesis is taken as a minimum criterion for not only self-reproductive but also in some way life-bearing systems. With the quite recent implantation of informational nucleic acids and the future goal of a nucleoprotein core [49 fl, they might really develop as first intriguing artificial nucleations of primitive protocell design. An approach that might perhaps be able to realize Lucretius’s “quacumque inter se possent congressa creare” [40 a] of nucleic acids, proteins, and membrane components that governed - together with suitable mediating water networks the origin of life patterns. It is, indeed, just this complexity that for today’s chemistry provides provocation and stimulation, intimidation and temptation, love and hate and fate together. The present artificial systems still remain utterly outclassed by even the most primitive life forms such as RNA viruses [5 b, 17,18,40h, 431. The possibilities of describing natural selection behavior according to quasi-species distributions in the extreme multidimensionalities of sequence spaces are, for artificial systems, at best a very distant utopia. However, independent of the respective chemical character, basic patterns of autocatalytic activities once again make the original image of that “notational replicative” DNA double helix the center of hopeful expectations. With all its early primitiveness, but also with its promising inherent potential of “minimal models” of self-replication [43 -48, 501, and -just to follow -
418
VIII
Living Systems
Figure 23. Eschenmoser’s homo-DNA [50 c-g] playing the game of artificial evolution and simulating from nucleic acid viewpoints aspects of general carbohydrate codes [60, 611; (top) Olson-DNA and RNA arrangements [5Og] in comparison to comparable appearances of homo-DNA; bottom) Olson-RNA -peptide interactions modeling early intimacies in the nucleation of the nucleoprotein system [7 a, 17, 181 and by this excluding more rigid hexose-DNA/ RNAs from evolutionary trends [50].
“minima vita” models [49] of life, chemistry enters the fields of biological complexity in what should be the most decisive region of evolutionary formative processes, and in doing so is gaining new qualities. Qualities, however, that emphasize in continuation of Schramm’s starting point [39 b] the importance of still disregarded mesogenic reaction behaviors in particular and inherent “phase-chemistry” in general [7 a, 17, 181. And then there is finally -besides the artificial evolutions of selecting functional nucleic acids for specific interactions with cooperative mates, high-affinic recognitions, and catalytic activities from random pools - the “evolution” of an individual scientific life’s work [50], which itself follows decisive stages of the grand process: the exploration of early chemical requirements, the development of prebiotic ligand systems, the fixation into the ordered structures of informational inorganic matrix patterns
and, finally, the liberation of their inherent wealth of design and information into the order - disorder of today’s nucleoprotein system. Using basic hexoses - in some respects the successor molecules of evolution - a never attempted, or perhaps only forgotten, “evolutionary step” is taking place [50c-f, h] - once again in the area of replicative (homo)nucleic acid systems (Fig. 23), and intriguingly not without relevance to the so far unsolved problems of carbohydrate recognition codes [7 a, 8 c, 33 p, q]. But this is another whole story, and another great game. A game that is representative in all its loving utilizations and impressive manifestations of today’s chemistry standards and facilities for our future ways of modeling of what has created us without any chance, however, to renew artificially the “whole” on our own.
2.4.3 (Bio)mesogenic Order - Disorder Patterns Transient order - disorder coherences converting a chaotic variety of quantum-duality-derived structural prerequisites and mesogenic capabilities - far from thermal equilibrium - into complexity patterns fundamental to life; artificial systems in their independent enterprises and in the attempts to retrace evolution’s (bio)mesogenic developments; fundamental complexity irreducibilities contradicting reductionistic scientific approaches; quantitative limitations standing in contrast to qualitative comprehension; entropic gradients achieving equilibria self-organization and mediating nonequilibria spatiotemporal coherences into life-pattern developments - transiently acting (bio)mesogenic patterns intriguingly add to a consistent view of life-pattern developments within the grand process [7a, 17, 51-52].
2
Reflections of life’s biomesogenic modes of operation remained outside mainstream molecular biology for rather a long time. The rapidly enlarging range of thermotropic and lyotropic liquid crystals repeated in their second heyday the long journey of native mesogen developments from monomers to oligomers and polymers almost independently of molecular biological achievements (Figs. 24a-h) [7a, 17, 18,51,52]. The last few decades have brought about the end of simplicity by bizarre stereoelectronic variations and the accompanying difficulties of corresponding phase systematization [5 1, 521. The theses of rod-like thermotropics and head-tail designed lyotropies (Fig. 7), which governed liquid crystal developments for a century, have lost their uniqueness. Complex molecular interactions submerge the extreme conceptions of thermotropy and lyotropy. Dynamic order - disorder patterns are creating overlying interconnections. Artifical mesogens are approaching the complexity of their biomesogen standards and encountering all the advancedmaterial aspects evolution had already operationally integrated. But the same difficulties that Lehmann experienced are being met by the complex biomesogen systems [17, 18, 51-53]. The creative principles of “subject” and “relatedness”, having already traveled across the laborious plains of fermion boson and particle - wave dualities [ 1- 81, reach the complementaries of “structure” and “phase” [ 17, 181, culminating in the beloved species of “side-chain polymers” and their outstanding mesogenic capabilities [7a, 17, 18, 51,721. Biopolymeric and biomesogenic species seem further integrable into supreme guiding fractal motifs of organismic morphogenetic expressions [78] and even into quantum-field patterns of human consciousness fields [72 x-z, 79,801. From the discontinuities of rudimentary, but nev-
Biomesogens and the Grand Process
419
ertheless chiral backbone-stumps of membrane components up to the continuations of the beautiful organichnorganic strand-repetition of nucleic acids, the functional backbone-filigrees of proteins and the carbonsubstitution-framed water-cluster individualizations of carbohydrates, evolution has developed nearly infinite action mode capabilities from a limited selection of building and functional blocks within narrow operating temperature ranges [5-8, 17, 18, 29, 30,43, 51 -711. Impressive side-chain-pattern selections, adjusted to the individualities of quite different backbone variations, complete the energetic and entropic instruments for creating life as an outstanding new quality of universe accelerations. The primitive amide group - a vision of horror to all classical liquid-crystal scientists - embedded with all its hydrogen-bond capabilities into the exclusivities of heteroaromatic horizontal and vertical interaction modes and reappearing also in the intra- and interstrand recognition and stabilization patterns of proteins, has developed as an essential part of information origin, processing, reproduction functionalization, and optimization. Amides - partially modified by their imino substitutes - frame the first successful information channels and mediate the first replication schemes, probably of RNA species that gain additional functionality from the excess patterns of residual hydroxygroups within the conventional nucleic acid backbone designs. And amide groups also provide for the three-dimensional intraand interstrand “intelligence” of proteins that functionalize the amide sender- signal - channel -receiver patterns of the informational component. The developmental one-dimensional information patterns of nucleic acids, wound-up for purposes of information storage and transfer within the elegance of preferably double-helices and their suprahelical packages, find some
420
VIII
Living Systems
Figure 24. Biomesogenic structures: a) (Bio)mesogens displaying order -disorder distributions in CPKpresentation (left to right and top to bottom); hexa-nalkanoyl-oxybenzene discoid - Chandrasekar’s first non-rodlike liquid crystal [28 a, 51 c]; enantiomeric cholesteric estradiol- and estrone-derivatives [ 17 a, c, d, 26f, 51 a, s, u]; Reinitzer’s cholesterolbenzoate [21, 221 - together with the acetate the foundation stones of liquid crystal history [21, 221; Kelker’s MBBA first liquid crystal “fluid” at ambient temperature [ 13f, g]; Gray’s cyanobiphenyl nematics for electrooptic displays [25 a, 5 1el; lyotropic lecithin membrane component [7 a, 14, 27 d, 52 a] and valinomycin-K+membrane carrier [7 a, 351; thermotropic cholesterylside-chain-modified polysiloxanes with the combination of flexible main-chain and side-chain spacers [51 a, h]; thermotropic azoxybenzene polymers with flexible main-chain spacers [5 1 a]; thermotropic cya-
nobiphenyl-side-chain-modified polyacrylates with flexible side-chain spacers [52 b]; lyotropic stiff-chain polyamide [52 c, d]; lyotropic cellulose triacetate [52e]; smectic diacylhydrazines [17a, c, d, 26f, 51 t]; thermotropic multiple-phase long-chain vinamides displaying intra’ and intermolecular hydrogen bonds [52fl; n-alkoxybenzoic-acid dimer smectics [52g]; thermotropically designed 2-pyridone-dimer smectics [17a, c-g, 27g, 39g, 461; Lehn’s supramolecularhelix building-blocks on the basis of nucleic-acidbase-analog recognitions [47 a]; gramicidin-A-K+membrane channel [52 h]; (tentoxin)3-K+-membrane subunit carrier/channel/store [52 I]; lyotropic poly(ybenzyl-L-glutamate) [5 1d, 52 k]; poly(L-glutamic acid) [51 d, 5211; poly(L-lysine) [51 d, 5211; discoid metallophthalocyanine [5 1 0, 52 m]; (A),-stack as operating in messenger processings [52 n]; 2 -5(1),, as structural analog of the oligonucleotide regulides
2
2-5-A [520]; unusual pairings of ds (U), . (U)n, ds (C), . (CH+),, ds (AH+), . (AH+), [52n, u]; bisacridinyl/DNA bisintercalator complex as an advanced model for biomesogenic nucleic acidkmall ligand interactions [7a, 18, 5 2 ~ 1ds ; B-DNA as life’s informational component [7a, 17, 18,28d, 29f-h, 33a, c, f, p, q, 520-v]; ds Z-DNA as regulation facility within genome [33 a, c, f, p, q]; ds A-RNA with simulation potential for viruses and stimulators of cytokineinduction [7a, 17, 18, 33a, c, p, q, 520-v]; ts DNA/ RNAs as logic-blocks of hysteretic biomesogenic regulation systems [7a, 17, 18, 33a, c, f, p, q, 48a-q]; qs (G), with teleomeric genome regulation functions [7a, 17, 18, 33a, c, p, q, 52x1; ds RNA/analog semiplastic complex with interferon induction and reverse transcriptase inhibition activities [7 a, 18, 33 a, c, 52 w]; pcasomorphin messenger between opioid and cardiovascular-cardiotonic regulation programs [ 17, 18,52y]; substance P as messenger of central and peripheral neuronal programs [ 17, 18, 52 z]; (L-LYS),adaptation to B-DNA major and minor, and A-RNA
Biomesogens and the Grand Process
42 1
deep and shallow grooves [7 a, 17, 18, 33 a, c, p, q]; mismatched GGU-base-triple presumably mediating self-splicing of tetrahymena rRNA [ 18k]; specific recognition of a-helical Cro-repressor in the major BDNA groove of Cro-operator [33 m, n]; unspecific recognition of protein-psheet and B-DNA minor groove as modeling early polymerase efficiencies of nucleo-
422
VIII Living Systems
protein intimacies [38]; ds Olson-RNA/(L-Lys), interplay dynamics 117, 18, 50hl. b) Thermotropic artificial systems illustrating geometric and functional variabilities around the common motif of longitudinal and circular order-disorder patterns (top to bottom): hexa-n-alkanoxyloxybenzene discoid 128a, 5 1 c], cyanobiphenyl nematic [25 a, 51 el; diacylhydrazine smectic 117 a, c, d, 26f, 51 t]; enantiomeric cholesteric estradiol derivatives [I7 a, c, d, 26f, 51a, s, u]; vinamide smectics [52 fl; discoid metallophthalocyanine [5 10,52 m]. c) Thermotropic hydrogen-bond species partially modeling drastically abstracted nucleic acid and protein 0-sheet design (top to bottom): long-chain alkanolyoxybenzene substituted vinamides [52 fl; nalkoxybenzoic acid dimer smectic [52 g]; thermotropically designed 2-pyridone-dimer smectic [ 17 a, c-g, 39g, 461; diacylhydrazine smectic [17a, c, d, 26f, 5 1 t]. d) Polymer variations between natural and artificial systems, between side- and main-chain variations (top to bottom); thermotropic cholesteryl-sidechain-modified polysiloxane with the combination of
flexible main-chain and side-chain spacers [5 1 a, h]; thermotropic cyanobiphenyl-side-chain-modifiedpolyacrylate with flexible side-chain spacers [5 1 b]; thermotropic azoxybenzene polymer with flexible mainchain spacer [51a]; lyotropic stiff-chain aromatic polyamide 152c, d]; lyotropic cellulose triacetate 152el. e) Native lyotropics around membrane-component-headltail and peptide-helical and platelet motifs
2
Biomesogens and the Grand Process
423
structural correspondence in the fundamentally deformed, but nevertheless comparable individual expressions of two-dimensional ambiguous arrangements in membranes. On the way from nucleic acids to membranes, the precise coding of nucleic acid informational side-chain designs flexibilizes into the loose recognitional and functional interaction modes of long-chain aliphatics, while the geometrical beauties of nucleic acid backbone phosphates dissolve into the scattered mosaics of individualized head-group arrangements of flexible membrane components. Covalently bound counterion patterns of membrane components constitute an additional aid for tightly interwoven molecular carpets, while rigid nucleic acid backbones preserve their ability to regulate by easily shiftable counterion clouds. Two-dimensional membrane patterns rediscover their global geometries in the wound-up cylinders of nucleic acid Afamilies and in the chirally twisted bipolar adjustments of protein p-sheet designs. And it is within the pictures of long-favored proteidmembrane and long-disregarded nucleic acid/membrane intimacies that nucleic acid/membrane-passages, already suggested from the interferon-inducing facilities of ds RNAs and later used for artificial electric-field-pulse mediated nucleic acid
(top to bottom): lecithin [7 a, 14,27 d, 52 a]; poly(y-benzyl-L-glutamate) [5 1 d, 52 k]; poly(L-lysine) [5 1 d, 52 11; poly(L-glutamic acid) [51 d, 52 11; (tentoxin),K'-membrane-carrier subunit complex [52 i]. f) Lyotropic nucleic acid strand variations between cholesteric and columnarphase design (top to bottom): (U), . (U), duplex; ( C ) , . (CH+), duplex; (AH+), . (AH+), duplex; B-DNA duplex, A-RNA duplex; DNNRNA triplex; (G), quadruplex [7a, 17, 18, 33a, c, p, q, 520-w, 751. g) Mesogenic nucleoprotein models around helical motifs: (L-LYS),fit into RNA (top) and DNA (middle) shallow and minor grooves; (bottom) (L-L~s),-DNA/RNAmajor/deepgroovefits [7a, 17, 18,33a, c,p, q, 38,751. h) Low- and high-molecular-weight biomesogen inter-
relationships - simple artificial thermotropics modeling operation modes of complex native amphotropics: recognition and modulation in selfreplication modeled by mesophase behaviour of cis- and trans-amide patterns of 2-pyridones and 1,2-diacylhydrazines (left to right and top to bottom): hypothetical recognition and modulation within first vice-versa polymerase facilities between nucleic-acid DNA and RNA duplexes and protein P-sheets; isopotential space partitioners (red: 4.184, blue: 4 . 1 8 4 , yellow: 0 kJ/mol) visualizing prerequisites for cis-amide recognition in nucelobases and trans-amide modulation facilities in P-sheets; mesophase simulations by recognition systems of 2-pyridones and modulation systems of 1,2-diacylhydrazines.
424
VIII
Living Systems
membrane passages in genetics, are now opening prospering fields of “gene-immunization” on the basis of general membrane acceptances for nucleic acid patterns. Carbohydrates, their oligo- and poly saccharide variants as well as their extensive derivatizations, have developed - besides rudimentary polar- apolar distinctions - massive hydrogen-bond donor/acceptor capabilities of carbon-substitution-framed ordered water clusters into adjustable mediating patterns of widespread dynamic motifs, ranging from impressive helical designs to circles, holes and stretches, and emulating the mediating facilities of “water order -disorder patterns” in their emphathy roles of recognition and function. Finally, proteins have developed the giant set of instruments of at least 20 side-chain variants within nearly omnipotent threedimensional extensions of one-dimensional backbone stretches into a legislatively ungovernable, nearly infinite information content, but also into the executively incomparable functional universalities of intelligent playmates for everybody else. The inevitable problems of solubilization [53], which threatened to exile all these molecular achievements in the dead end of solid-phase immobilities - Fischer’s ingenuity ended at the level of around 20 residues in his first synthetic approaches to the kingdom of proteins at the beginning of the 20th century [36 a], and Guschlbauer, despite his farsightness [33a, b], still looked upon DNA as “a hopeless mess” at the beginning of the 1970s -were solved by the ability of these “side-chain-polymer polyelectrolytes” to occupy the fertile fields of mesophase solvent - solute solubilizations between the extrema of solid and liquid phases [7 a, 17,181. Water clusters not only liberated suitably designed biopolymeric species to structural mobility, but also sensitized them to the supreme control of liquid-crystalline states.
Helical motifs of informational and functionalnucleic acids [52n-y, 62-69,74,751 and peptides [52k, 1, 58, 59, 741, which might also be comprehended as wound-up discoids, display with the individual geometries of their constituents cholesteric phase variants of both calamitic and columnar arrangements and even present blue-phasetextures of cubic phase design [62]. While the informational and functional representatives are thus reminiscent of ancient thermotropic guiding motifs, the compartmental membranes, on the other hand, are easily reducible to the basal abstractions of primitivehead-taillyotropics [7 a, 17,18,54-571. Carbohydrates [52e, 60, 611, finally, have developed calamitic prerequisites into more sanidic designs of laterally and longitudinally extended main-chain arrangements. While water - and to a certain degree also other solvent partners and solvation aids - thus succeeds in the liberation of mesogenic side-chain polymers [7, 531, it forms by this, however, only sterile repetitions of overall phase symmetries. For the transition to life, the mediating abilities, of water patterns prove necessary, but not sufficient. The transition to life was not brought about by the mere existence of supramolecular organizations, nor by the phase beauties of banal basal repetition patterns. Approaches to life had to detect phase/domain regulation facilities within the nonlinear dynamics of supramolecular mesogenic organizations [ 18 n]. They had to develop the richness of informational and functional capabilities out of creative chaotic possibilities as well as the fleeting permanence of phase/domain transition strategies withing narrowing ranges of operation temperatures far from thermal equilibrium. Ways into the nucleation centers of the transition to life imply ways into fundamental complexity [6 b, 7 a, 171, the dominance of instabilities and probabilities rather than deterministic laws [I],
2
the subjection to irreversibilities along irreducible bundles of nucleation trajectories and the development of spatial order - disorder distributions into spatiotemporal coherences [ l - 8 , 17, 181. Ways of following them up scientifically imply the retreat from mere reductionistic approaches, with the shortcomings, however, of the dominance of qualitative inference rather than quantitative scientific precision. One of the real breakthroughs in transition-to-life capabilities seems to have been the establishing of connectivity lines of coherent solute - solvent relationships within the biomesogenic partners themselves [7 a, 17, 18, 53 b, 72r, s]. The decisive development of replicational nucleic acid systems with their square root laws of propagation, but also their inviting cooperativity of partners [40 h, k, 431 - had to await the overall assistance of proteins, displaying for exponential autocatalysis - and the decisive selection of competitors - not only suitable solvents but also the intelligent stereoelectronically mediating and promoting milieu [7a, 17, 181.Only the nucleation centers of nucleoprotein systems, individualized by early compartmental membrane walls, could provide all the enthalpic and entropic prerequisites to set the new qualities of the universe “alive” far from thermal equilibrium [l-8. 17, 181. The extreme coherences of such fundamental complex systems contradict our advanced theories of phase and phase-transition strategies [76]. The findings in the field of biomesogenic species reflect the situation [7 a, 17,18,5 1 -721. In the classical view only membranes and their components seem tolerable, with their close bilayer similarities to thermotropic smectics and headhail abstracted lyotropics [7a, 33p, q, 54-57]. While nucleic acids and proteins represent both structure and system individualities, membranes, as huge collectives of highly differentiated mem-
Biomesogens and the Grand Process
425
brane components with specifically adjusted lipid, protein, carbohydrate and perhaps even nucleic acid patterns, exhibit far more complex system hierarchies [54 b, c]. The slogan of the 1970s that “life is at membranes, in membranes and through membranes” [7 a, 48 b-e, r, s] characterizes a situation where the fundamental complexity of life has to meet up with the critical complexity of localized compartment organizations, which turn out to be completely irreducible to simple bilayer characteristics in structural design and mesogenic regulation capabilities. From the involvement of the fully expressed diversities of membrane components [7 a, 8 c, 33 p, q, 54 a], their asymmetric designs and nonlinear modes of operation, to different ligand, especially cholesterol, regulations [54 b, c], passive and especially active transport (to and from the system) by cooperative protein-carriers, pores, channels and complex subunit organizations [7 a, 33 p, q, 35, 52 h, i], the diversity of peripheral and integrative proteins [7a, 33p, q, 56, 58, 591, the cell-coat glycocalix made up of sugar side chains that are joined to inh-insic membrane glycoproteins and glycolipids as well as to glycoproteins and proteoglycans [7 a, 33 p, q, 54-59], carbohydrate-protein patterns in selectin and integrin adhesion molecules [59, 601, lockand-key glycoprotein - lectin interactions [6 b, 7 a], signal recognition, acceptance, transduction 156b] and amplification [7 a] in cellular selfhonself recognitions and discriminations [7a, 8 c , 17, 18,33p, q], cellular uptakes and propagations, cellular communications and differentiations [8c, 33 p, q], molecular linguistics vertically in genetic intraspecies information transfer and horizontally in the free-floating immunological extensions of individual CNS-peripheries, up to CNS as the operational basis of individual intelligence and the origin of con-
426
VIII
Living Systems
sciousness [5a, 8, 17, 18, 78-80] -all this constitutes, on the basis of highly differentiated membrane components, proteins, carbohydrates and nucleic acid cooperation partners, hierarchies of increasing complexity, resulting intracellularly in supreme organizational units and intercellularly in morphogenetic programs, organismic expressions, and system and species evolution. A picture that reveals in all its insufficiently sketched coherences not only the fundamental complexity of life patterns but also their irreducibilities [1-8, 17, 18, 51 -61, 781. Proteins present like problems. Completely neglecting the utmost structural and functional complexity involved [7 a, 8 c, 331, endeavors in the field of protein mesogen behavior have mainly concentrated on poly(L-glutamic acid) derivatives with their intimate relationships to rod-like helical designs [58, 591. More sophisticated than the usual cholesteric systems, poly(y-benzy1-Lglutamate) [5 1d, 52 x] responds to environmental influences. It covers within its very shape patterns of a “pre-Pauling” helix [7 a, 29 d, 5 1d] and anticipates with its aromatic stacks the later discoid arrangments [28 a, 51 c]. The replacement of chloroalkanesolubilized benzyl substituents by an alkyl disorder belt, mobilizing the helical-order cores, shifts the system from lyo- to thermotropic characteristics and, intriguingly, makes visible in this way the dominance of order-disorder distributions in the establishment of mesomorphic behavior [58 el. The protein-mesogen system - although primitive in comparison with native protein organizations - continues to be a source of inspiration, without so far being extendable to the real complexity of native protein patterns [51, 52, 58, 591. Later, together with carbohydrates [7 a, 60,611, the liquid crystals of the much more transparent nucleic acids [52 n- w] join
these circles. Nucleic acids, when treated under stringent conditions as idealized helices [62, 64-69] - and even small buildingblock composites fit into this picture [65] reveal impressive pictures of classical phase textures with relationships also to overlying, biomesogenic operation modes in chromatin (re)orientations [29g, h, 62,77 h] and mitosis [8 c]. Independently of intercalator, groove-binder or further nucleic-acid-ligand characteristics [7a, 17, 18, 52t, 671, small ligand-interaction capabilities up to smallpeptide partners [67-751 are integrable into the common picture of classical liquid-crystal design, and it was just within the field of biopolymer/ligand interaction that the “Vorlander effect” of directing mesophases by the “chiral infection” with small amounts of doping compounds [70] advanced as a model of hormonal signal direction and amplification (Fig. 25) [7a, 17, 181. However, also in these cases, the abundance of integrative biomesogenic operation modes of native standards will be insufficiently met by scientific progress in mesogenic abstractions. Even more interesting than the beautiful phases of idealized helical entities of nucleic acids, for instance, seem in this connection, the individual appearances of “unspecific” singular phase textures of polydisperse evolutionary DNAs and RNAs [73, 751, which “live” the whole complexity of mesogenic life-pattern possibilities. Carbohydrates [7a, 17, 52e, 60, 611, finally, so generally effective in phase stabilizations that even polyethyleneglycols, as strange artificial abstractions, promote classical nucleic acid phase expressions [52 t, 64-69], contribute with a spectrum of quite different approaches, under way in many laboratories, to a growing understanding of rapidly enlarging details, while so far only insufficiently coping with the urgent quest for generally attributable carbohydrate recognition codes [50, 60, 611.
2 Biomesogens and the Grand Process
Figure 25. Low-molecular-weight effectors directing mesophase patterns of high-molecular-weight biomesogen organizations (left to right and top to bottom): mesogenic ligand-biopolymer interactions Reinitzer’s chiral cholesterylbenzoate [2 1, 221 and Kelker’s MBBA [ 13 f, g ] , elucidating Vorlander’s “zirkulare Infektion” of nematic phases by chiral doping compounds [70], an early biomimetic thermotropic hormone model, simulating biomesogenic amplifications of molecular signals in biomesogenic surroundings [7a, 17, 181; native and artificial smallmolecule effectors of nucleic acid operation modes, modeling Vorlander’s “chiral infection” within complex bio-mesogen systems and reminding of their intriguing initiatory roles for the redetection of the Vorlander-effect of switching mesophase design in artificial and native systems.
Just like the traditional and obviously difficult to overcome scattered views of proteins, nucleic acids, membranes and carbohydrates as kingdoms of their own in molecular biology [7 a, 8 c, 33 p, q], the experimental findings in these new fields of liquid-crystal research constitute at present still a dispersed mosaic of details rather than the needed adequately complex and integrative picture [7 a, 17, 181. While liquid-crystal phase investigations remain preferably limited to helical aspects of only a few con-
427
tributing natural structures and their accompanying artificial derivatives, general questions 1721, such as microphase and domain tuning and regulation [7a, 17, 181, microstate distributions 172d] within interchanging organizations relating to entropy, information and complexity relationships 172d, el, complex energy/entropy compensations 133k, 72 k], general field extensions to quasicrystals 172b], glass and plastic states [72e, fl, as well as evolutionary relationships to spin-glasses 172d], intimate coherences in nucleic acid sequence patterns 172 s- w] and molecular dynamics 177b-d, k, n] within supramolecular biomesogenic organizations and their developmental aspects [7 a, 17, 771 still have to await unifying conceptions. Perhaps the previously offered “biomesogen” views [7 a, 17, 181 can provide integrative guiding patterns for all (potential) molecular biological meso(phase) “builders” - ranging from their individual statics and dynamics, their amplifications into interactions of highly cooperative (transient) domain organizations to the integrative (domain-modulated) phase transition strategies and regulation modes of classical liquid-crystalline states (Fig. 26). While so far the availability of computers has failed to extend individual molecular dynamics [77] into the developmental fields of supramolecular biomesogenic organizations, and phase/domain characterizations only succeed in the elucidation of uniform large molecular collectives [25, 5 1, 521, experiments following mesophase characteristics of self-organizational and self-reproductive processes down to the nanometer scale [73, 741 are receiving increasing interest and try to make up for the persisting difficulties of the lack of both farreaching and profound theoretical interpretations [76]. Molecular structure amplifications and behavior-determining phase reductions meet
428
VIII Living Systems
Figure 27. Visualization of biomesogenic nucleicacid selforganizations by molecular-resolving microscopies (top to bottom and left to right): nucleic acid strand patterns: B-DNA and A-RNA duplexes, DNA/ RNA triplex, G-quadruplex, DNMRNA-duplex/protein-p-sheet complexations; STM and SFM-investigations of nucleic acid organizational behavior: DNAplasmid [74 c]; RNA- and RNA/peptide-organizations: (U), . (A),-duplexes; (U), . (A), . (U),-triplexes; (G),-quadruplexes, (L-LYS),/(U), . (A),-complexes in 2D and 3D representation [7a, 17, 18, 33 a, c, f, P, q, 751.
Figure 26. (Bio)mesogenic order -disorder patterns rather than geometric anisotropies as general preconditions for the appearance of mesophases: a) classical thermotropic cyanobiphenyl nematic [25 a, 5 1 e] and thermotropically designed steroid hormone cholesteric [17a, c, d, 26f, 51 a, s, u]; b) protein-p-sheet simulative diacylhydrazine smectic [17 a, c, d, 26f, 51 t]; c) hexaalkanoyloxybenzene discoid [28 a, 5 1 c] ending the century of “rod-like’’ dominance; d) lyotropic poly(~-glutamicacid) [5 1 d, 52 k, 11 changing with an alkyl-disorder shell from lyotropic to thermotropic [58e]; e) ds A-RNA lyotropic [52n-w, 751; f) amphotropic (L-L~s),/RNArecognition complex [ 17, 18, 751.
within the creative fields of supramolecular biomesogenic organizations [ 17,301. Complex molecular recognition and information processing amplify the continuation of adequate phase/domain transition strategies into new qualities of intelligently developing self-organizational and self-reproductive systems that gain their transition to life and life-survival strategies by surprising entropy - information transformations far from thermal equilibrium [7 a, 171. Suitable scientific approaches suffer from the restrictions of methodologies that meet the requirements for solid- and liquid-phase investigations but are insufficient for the treatment of mesophase characteristics and developmental aspects of highly condensed systems, which play crucial roles in self-organization, self-reproduction, information processing, and optimization. While a multitude of experimental findings in the fields of polarization microscopy, chiroptical investigations, static and dynamic X-ray dif-
2
Biomesogens and the Grand Process
429
approaches and their scientific aims in direct nanometer-elucidations of the statics and dynamics of biomesogen nucleation patterns [74]. Adlayer arrangements provide direct visualization not only of supramolecular biomesogenic organizations, but also of their inherent dynamics (Fig. 27) [74, 751. Dimension reductions of threedimensional large scale bulk-phase repetition patterns into “two-dimensional mesophases” with their ability to survey multiple monolayer arrangements at the molecular (nanometer) level [73] display insights horizontally into the two-dimensional lateral extensions of phase relationships and vertically into the three-dimensional individualizations of creative nucleation centers of supramolecular biomesogenic developments (Fig. 28).
2.5 The Developed Biomesogenic Systems Figure 28. Complex biomesogenic organizations modeling evolutionary developments: (a) SFM visualization of polydisperse DNA [75];(b) nucleic acid textures (left to right): chicken-DNA, (U), . (A),, duplex (U),, . (A), . (U), triplex, (G), . (GI, . (G),, . (G), quadruplex, (U),, . (A),, . (L-LYS), complex [75]; ( c ) building up of liquid -protein -nucleic acid multilayer systems as two-dimensional simulations of the grand evolutionary triad - at least in partial cooperation with a mediating water milieu [7a, 17, 18, 33a, c, p, q, 63a, 751.
fraction techniques and nuclear magnetic resonance methods will provide - besides the often familiar and convenient texture patterns - a wealth of information on complex phase-domain organizations and their operation modes in leveling-off of statistic multitudes [5 1-72], the rapidly enlarging spectrum of molecular-resolution microscopies is receiving growing interest as to their
Emergence of new qualities of the universe; life spatiotemporal coherences between nonclassical chaos and fundamental complexity achievements; transiently interchanging supramolecular organizations and their biomesogenic regulation; amplifiable “parts” permanently qualifying changing “wholes”, developing “wholes” qualifying operational “parts”, structure - phase dualities inherent in the developments of nucleoprotein systems; stabilization within the dynamics; morphogenetic programs and regained operational symmetries on the basis of fundamental asymmetries - scientific approaches to life patterns (Fig. 29) unravel a complex of problems, fitting for the complexity of life patterns themselves [ 1-8,17, 18, 77-80]. After a century of mainly parallel and independent developments, artificial meso-
430
VIII
Living Systems
Figure 29. Biomesogens: (smaller pictures, top) transitions to life illustrated by the nucleation of nucleoprotein intimacies [7 a, 17, 181; (smaller pictures, bottom) replicative and informative nucleic acid triplexes [48 a-q, t] as forerunners, displaying molecular hysteresis as basic mechanisms of memory imprints and oscillation and rhythm generation [7a, 17, 18, 33a, c, f, 481; (top and bottom) the late followers’ ontogenesis, retracing biomesogen phylogenesis in “organismic first intimacies” [ 16, 231.
gens and their biomesogenic native standards (Fig. 30) seem, nowadays, capable of a creative dialectic synthesis of the thus far dominant thesedantitheses tensions. The unique spatiotemporal coherences of (bio)mesogen systems promote self-consistent evolutionary views and afford intriguing perceptions of the grand process. Even very simple and primitive mesogens differ from nonmesogenic species in that they bear some sort of rudimentary intelligence. At present, however, we appear to be faced with difficulties in treating this gleam of preintelligent behavior scientifically. Even our most advanced theoretical approaches prefer rather abstract and ethereal
shadows of real entities [76]. Changing from here to complex artificial or even native (bio)mesogen organizations amplifies the problems for quantitative treatments [7 a, 17, 18, 331. The constraints leading to only qualitative inferences -if not the temptation to speculate - are growing. Although artificial mesogens have been developed in terms of thermotropic and lyotropic characteristics, biomesogens both blur and dialectically combine these extreme positions within the cooperative network hierarchies of their interdependent chiral complexity patterns (Fig. 26) [7 a, 17, 181. Thermotropics resemble lyotropics in that their flexible, more disordered seg-
2
Figure 30. Anabolic and catabolic regulation hierarchies that channel biomesogenic information processing in a highly condensed phase [8 e, 18, 52 x, y].
ments serve the purpose of dynamization of the rigid, more ordered parts. Lyotropics, on the other hand, display, within their complex solvent- solute distributions, far more interactive coherencies than had originally been thought in classical views. While for lyotropics suitable solvents -especially water - provide a spectrum of entropy-driven organizational forces, thermotropics will profit by comparable entropy effects of order minimizations on the complex domain interfaces between rigid cores and flexible terminals. Within this picture, the solventlike labilizing areas of interacting biopolymer organizations appear as special expressions of more general mobility characteris-
Biomesogens and the Grand Process
43 1
tics of mesogens, which are able to build up within their dynamic chiral order - disorder distributions transiently functional and transiently acting chiral order- disorder patterns. Within the dynamics of water-mediated protein, nucleic acid and membrane organizations, dynamic chirally instructed parts - submerging the individualities of their respective partners -exert functions of partial solubilizations for the mobilization of rather static chiral solute areas, the preintelligent handling of which is a prerequisite for some new properties as, for instance, recognition, functional catalysis, semipermeable compartmentation, collective and cooperative operation modes, isothermal self-organizational behavior and selfreproduction of transiently acting chiral domain systems, information processing and optimizations. Our original ideas about mesogens, which had been attracted by the unifying principles of huge artificial and preferentially achiral molecular ensembles and the overwhelming symmetries of their mesophase relationships [7a, 13, 14, 17, 18, 24-28, 51, 521, are thus redirected into the limited and methodologically extremely difficult areas of cooperatively processed chiral phase/domain systems, which govern with increasing complexity a more and more precisely tuned and refined, highly sophisticated instrumentary of modulated phase/ domain transitions [7 a, 17, 18, 69, 7 1, 72, 771 (Fig. 31). The chiral mesogen individualities of the grand supra-chirally organized amphiphilic patterns of life display within their molecular imprints the prerequisites for the projection of individual molecular facilities into the structural and functional amplifications of cooperative dynamic mesogen domain ensembles. The chiral orderdisorder designed individual appears as a holographic image of the whole. By this, the classical views of interacting structural
432
VIII
Living Systems
Figure 31. Biomesogen operation modes in nucleoprotein systems: (a) FISlDNA recognition dynamics [71q]; DNA/protein unspecific (b) [38] and specific (c) [33 m, n] recognitions by Psheet and a-helix fits to DNA minor and major grooves, respectively; d, e) Zn-finger protein handling DNA [7 I fl.
Figure 32. Biomesogen dynamics: (top left) interleukin-4 dynamics as derived from multidimensional NMR investigations [77 k], depicting among others Iabilized loops; (top right) the more rigid designs of nucleic and triplexes [77 ll; (bottom) dependence of membrane dynamics (MD) [77 m, n] on structural prerequisites and temperature range, MD snapshots in pico-second range; dilauroylphosphatidylethanolamine (95 ps), dioleylphophatidylethanolamine (120 ps), dioleylphosphatidylcholine (100 ps).
individuals submerge into the new qualities of transiently acting mosaics of mutual domain cooperativities, where chirally instructed stereoelectronic patterns of individual representatives of the grand triad, together with the “liquid polymers” of mediating water-clusters, anneal into the spatiotemporal coherencies of newly achieved biomesogenic domain organizations, their self-reproduction, and informational and functional optimizations. It is within this dual view of biomesogens that both the structural and the phase/domain aspects will contribute intriguingly to a consistent picture of life patterns and their operation modes, and it had been, moreover, a prerequisite to establishing self-organizational, self-reproductive, and morphogenetic processes under nearly isothermal conditions [7a, 17, 181. At least partially reflected in modern theories of nonclassical chaos [6, 7 a , 72e, fl, the developed chiral amphiphilic patterns,
2
Biomesogens and the Grand Process
433
Figure 35. P-Casomorphin peptide messenger in addressing opioid and cardiovascular-cardiotonic (modulatory) subroutines of general psychoneuroimmunological programs, superimposed with organizational microstate [72d] distributions and simplified regulation schemes of heart muscle contractions [18, 36, 5 2 ~ZI. , Figure 33. Detailed biomesogen dynamics treatment: rotational isomerizations of Tyr-35 of bovine pancreatic trypsin inhibitor mediated by structural changes in the surrounding biomesogenic protein matrix overall view (top) and “film” in skeleton presentation (bottom) [77 b-d].
Figure 4. First insights into mesogenic chromatin regulations: superhelical DNA chromatin arrangement [29g, h, 33fl; winding up chromatin packages by Bonnet transformation operation-modes [77 h].
amphiphilic both in terms of their order -disorder dialectics and the spatiotemporal coherencies of their nearly isothermal processings, are acting according to meso-
gen strategies, ranging from the localized motions of molecular segments [77 b-d] (Fig. 32, 33), via their coupling to collective processes (Fig. 34) in the surroundings [72 b - d, n], to the interdependences of complex regulations [ 18, 5 1 y, z] (Figs. 35, 36). Statics and dynamics of asymmetrically liberated and directed multi-solvent-solute systems and their sensitive phase and domain-transition strategies characterize our picture of today’s biomesogen organizations. Synergetics [7 b] within complex solvent-solute subsystem hierarchies [7 a, 17, 18 v] and their mutual feedback found new aspects of chirally determined nonlinear dynamics with a number of preintelligent operation modes, such as reentrant phenomena [48 i], molecular hystereses and memory imprints, oscillations and rhythm generation [48 a-i, r, s], complex structuremotion linguistics and signal amplifications, selfhonself recognitions and discriminations, information processings and general optimization strategies [7 a, 17, 181 (Fig. 37).
434
VIII
Living Systems
Figure 36. Biomesogen patterns of immunorelevant structures: (top, left to right) mobility sites of myohaemoerythrin, engaged in antigen - antibody recognitions [77 el - molecular surface and a-carbon backbone color-coded: highly mobile, red and yellow, respectively; relatively well-ordered, blue; not studied, magenta; (center and bottom) polio virus exterior and interior surface patterns [71 a] reminiscent of the birefringent solutions of rodlike tobacco mosaic particles and their role in Onsager theory [37].
An almost infinite number of examples characterize complex motions of the grand process that are stabilized just in their dynamics; dynamics that continue asymmetrically the spatial order - disorder distributions of biomesogenic life patterns into the spatiotemporal coherencies of dynamic order [5b, 6b, 7a, 17, 181. The contradictions between crystalline order and liquid disorder, which rendered Lehmann and Vorlander’s “liquid crystals” [13, 14, 241 so extremely suspect to their contemporaries,
Figure 37. Chiral biomesogen dynamics in informational and functional processing (from top to bottom): complex structure - motion linguistics of a small molecule effector (pcasomorphin-5) in addressing neuroendocrine, cardiovascular-cardiotonic and immunomodulatory subroutines of general regulation programs [7a, 17, 52y, z]; cholesterol regulations in functional membranes [7 a, 54 b, c]; valinomycin interplays, mediating K + ions across membranes [7 a, 351; biomesogenic signal amplification in the chromatin: steroid hormone efficiencies in domain regulations [7a, 17, 18, 51 s]; molecular hystereses and memory imprints in polynucleotide triplexes [7 a, 17, 18, 33a, c, f, p, q, 48a-q, t]; intimacies of nucleic acids and proteins [7a, 17, 181 along nucleation trajectories of the nucleoproteinic system.
2
represent and rule in fact the survival principles of the grand process. Biomesogens, which display within their chiral molecular imprints and designs not only the richness of experienced affinity patterns but also the capabilities of their play-educated dynamics, link together entropy and information within a delicate mesogen balance (Fig. 38). In this connection, Schrodinger’s [29 c] question concerning the two so extremely different views of order appears to be dramatically relevant. We seem to be badly in need of new theoretical treatments of old terms that have changed their meaning somewhat on switching from statics to dynamics, from thermal equilibrium states to far from thermal equilibrium developments. What seems to be helpful are extensions of classical static descriptions to new selforganizing, self-stabilizing, and self-reproductive dynamic processes. The abstract and mathematically perfect order of an idealized crystal (Schrodinger’s “dull wallpaper” in all its symmetries) and the strange dynamic order of life patterns (Schrodinger’s “beautiful Raffael tapestry” - in all its asymmetries not completely matching the real process dynamics) are competitors for supreme roles. Covered by the general uncertainty principle - with the classical views of order and disorder as borderline cases - the new qualities and description capabilities of dynamic order unravel pattern developments from the coherence of life.
2.6 The Human Component “Subject” and “relatedness” - developments via fermion -boson, particle - wave, structure -phase, and processing -information dualities into the preliminary final destinations of matter and consciousness; ontogenesis retracting (universe) phylogenesis,
Biomesogens and the Grand Process
435
quantum mechanics and general relativity theory longing for “theories of everything”, self-consciousness and self-awareness, questions for quantum-selves within a creative universe - man at the interfaces to the universe, awakened by consciousness of himself [l-8, 17, 18, 76, 80]? “Will there ever be such an achievement as a computer simulation of our central nervous system (CNS), a simulation of any given complexity, but nevertheless deterministic? Or is this CNS already a new quality, from the most highly developed computer as far away as a quantum mechanical approach from some one of Newton mechanics? Is Bohr’s remark ‘that the search for a value equivalent of thought resembles the measurement effect on a quantum object’ more than only a formal analogy? Does our CNS work for the engrams of its information patterns with a hardly estimable diversity of the quantum fields of its stereoelectrically manipulated biopolyelectrolyte patterns? Is there more behind the hologram analogy of our brain storage? Does the whole giant information content of our brain equal the utmost complexity of collective quantum states of the macrostructures of the CNS? Is superconductivity of biopolymeric macrostructures a leading model? Are Josephson contacts perhaps relevant switches? Will the forecast of Kompaneyez as to the Bohr-analogy of thought being more suggestive of a giant continuous change of the collective states of our brain as a whole with its complex arrangements in the ‘enlightened spot of our consciousness’ one day become certain reality? Quantum theory so far has not approached such complex systems. There remain at present final speculative questions without any reasonably, competent answer” [7 a] - the early- 1970s visions match supreme themes of our time from the suggestions for supreme facilities of future quantum computers
436
VIII
Living Systems
BIOMESOGENS Amphiphilicity Holographic design
Nnn-linear dynamics
Molecule
Macromolecule
Supramolecular Organization Domain
Microphase
Phase
Trannen~order-disorder patterns Mu111 ~ o k e n ~ - w IS~Yi eS ~ C ~ E Phare-domain iransilion rualegier Phase-domaincoaprauvily Reentrantphenomena
Molecdar hyslercrer. memory ~mpnnls OIEIIIBIIU~S and rhythm genrratmn~ Signal LransdUCliOn and ampliflcaliun Inlormauon pra‘esnlngr and preintelliiem operation modes Spauo-temporalcoherences
COMPLEXITY
Figure 38. Supramolecular biomesogenic organizations between “structure” and “phase” (left to right and top to bottom): developing biomesogenic textures: ammonium oleate - Lehmann’s [ 141 first “scheinbar lebender Kristall” (apparently living crystal) (original photograph as a kind gift of Hans Kelker [ 13]), Lehmann’s wormlike features [14], the Halle rediscovery of phase chemistry by a miscibility rule for promesogenic imidazole derivatives in the 1950s [26d-f], cholesteric streak texture of thermotropically designed steroid hormone derivatives [5 1 t] (three images), textures of 146 bp DNA fragment (three images), cholesteric organization of Dinoflagellate chromosome [52 q, 621, cholesteric oily streak and chevron textures of high-molecular-mass chickenerythrocyte DNA [52 q, 621, cholesteric oily streak and chevron textures of high-molecular-weight chicken-erythrocyte DNA [52n, 0,s] (two images); dual structure-phase view of nucleic acid-protein interactions modeling supramolecular biomesogenic nucleation states of the nucleoprotein systems as visualized by RNA/(L-L~s),intimacies [7 a, 17,18,75]. These images have superimposed on them supramolecular domain organization between structure and phase: chiral secondary structure basic geometries [28 d], continued with water-shell cover [53] and counterion clouds [69 a], amplification of individual structures into order -disorder patterns of cholesteric phase arrangements with superhelical twist; suggestions for morphogenetic mesophase extensions, exemplified in mitosis of Huemanthus cells [8 c]; supramolecular biomesogenic organizations as approached from “structure’’ and “phase” [7a, 17, 181.
3 Outlook
437
Figure 39. Classical and nonclassical approaches to chaos theories [5-7, 72e, fl.
[80 b], the importance of Bose-Einstein condensates up to the last questions of quantum-self-constituted self-consciousness fields as some sort of independent quality in the universe [7, 80aI. An instability-induced grand transition from quantum vacuum to material existence might have created our universe. Energy and entropy gradients mediated duality developments and transitions between fermions and bosons, particles and waves, structure and phase into the process -information dualities of life patterns. At the interface to the universe, life patterns developed a preliminary finite duality between existent matter and self-consciousness fields. By this grand transition a facility of awareness beyond space and time originated that transformed its thesedantitheses tensions into creativity. Monod’s vagrant at the borders of “his” universe (Fig. 39) - realizing the extreme and unique exposition of his situation - becomes aware of the intriguing and inspiring motivations resulting from his borderline case. He begins to conceive of himself as an exciting creative periphery of matter and interfaces beyond, which adds to so far irrelevant evolutionary developments advancing criteria of conscience-controlled consciousness patterns feeding back to the grand process itself (Figs. 40, 41). He faces a future where his dreams of a universe will no long-
Figure 40. Order-disorder patterns in artificial mesogens and native biomesogens promoting spatial chiral order -disorder distributions into spatiotemporal coherences (top to bottom): amphotropic DNA/ prealbumin complex [lo], reminiscent in its orderdisorder distributions of the simple achiral thermotropic n-alkoxybenzoic acid dimers [52 fl; periodic and chaotic oscillations of multiple oscillatory states in glycolysis [78] - all this superimposed on the details of the “very” event - modeling in all its interfacial relationships early evolutionary origins of the nucleoproteinic system and its inherent structure-phase dualities [7a, 16, 17, 181.
er fade away in senselessness, but seem increasingly enriched by blessed capabilities of self-realization within developing evolutionary fields. Descartes’s proud “cogito ergo sum” is also - by descent and development, by creation and destination, by obligation and fairness - something like a first grand self-affirmation of our special spacetime within the universe. A maternal developing universe awakened in his consciousness, recognizing by himself also itself. The light of self-consciousness born in the dark - science seems to cope with biblical motifs.
438
VIII Living Systems
Figure 42. Evolutionary dynamic order patterns from the maternal galaxy to its transition to life in selfreplicational informational species [7 a, 17, 181.
Figure 41. Mesogen patterns (bottom to top): cholesteric DNA [63 a, 751 superimposed on early universe computer simulations [ 121 and Lehmann’s paintings of cholesteric phase [ 141; “Tala” mesogenic surface; human brain [79 a] with continued mesogenic patterns from bottom, reminiscent of mesomorphic relatedness of neuronal-activity patterns [80a] to mesogenic operation modes and quite recent quests for quantumcomputer capacities [gob].
3 Outlook Since that incomprehensible beginning, when the time-arrow of entropy marked the directionality of this outset, entropy seems
not only to have endeavored to make linear uninformative equilibria states vanish in senseless disorder, but has, from the very beginning, developed extraordinary facilities to endow nonlinear systems far from thermal equilibrium with the rapidly enlarging capabilities of growing dynamic order (Fig. 42). It was the very molecular order- disorder design of their constituents that rendered those patterns sensitive to the amazing influences of entropy. Counteracting these rather “intelligent” molecules in its usual eagerness and need for disorder, entropy compensated for their lost homelands between order and disorder with the gift of cooperativity, self-organization, and some sort of newborn directed dynamic order within supramolecular ensembles and growing organizations. And, like Mephisto in Goethe’s grand opus, who introduced himself as part of the power that always wants evil while always doing good, entropy enabled those beloved playmates to propagate in a dramatic outburst of directed dynamics over the deserts of order and the abyss of disorder as an endangered, but incomparably beautiful, grand mesogenic pattern, set out in a grand beginning for far unreachable (?) goals. Intimately interconnected with information and complexity within the framework of Boltzmann’s Htheorem, entropy valued and promoted the informational contents of complexity states and opened, with the grand bifurcation of
4
Figure 43. Scalar-field explosion and developmental universe: quantum vacuum mediating within a grand fluctuation existence. Order-disorder coherences forwarding far from thermal equilibria chaos offers to levels of complexity that originate on basis of indeterministic quantum dualities consciousness patterns - sharing the whole [3, 7a, 17, 18, 751.
the appearance of self-replicating systems, the ways for the development of what we now call life. Referring to an independent “what is life” definition, proposed by Eschenmoser as “life is a catalyst that equilibrates its surroundings” within a common-term discussion at Maratea in 1993 [5 d], it seems perhaps not without a certain elegance to conceive of living systems just in their skillful utilization of energy gradients provided by the universe by their own attracting entropy gradients. Abilities that give rise to kingdoms of transitoriness, transientness and dynamics that establish supreme fundamental complexity patterns from creative chaos within their order - disorder coherences and life in the extreme relatedness of quantumduality-based self-consciousness fields, a first recognition and perception of the grand process, a first creative quality beyond space and time (Fig. 43). “A sun system, probably even more, that - embedded within the boundlessness of further intra- and intergalactic relationships - created evolutionary fields that became capable of information generation, process-
References
439
ing and optimization within selection-determined processes of self-organization and self-reproduction. A dynamic movement seduced from the homing states of sterile rigid order hierarchies into the homelessness of dialectically stimulated action patterns, becoming increasingly capable of developing a simulative, and from there, within a multilevel orientating and interchanging process, also operational and creative consciousness of the dynamics of the grand process and its possible active promotion, a grand dialectically driven and enlarging consciousness patterns which escapes the static plains of persistence in dynamic actions between space and time. The statics of our ways so far, transformed, furthered and developed into the dynamics of the present and future - an admirable grand process amidst the finiteness of itself and its surroundings - set out in a billionfold lived and renewed forwarding effort of improvement and optimizations for far horizons: will that be our further way? A way that means for the transitoriness of our individualities, the finiteness of their endeavors the final goal, while the integrative vector of the grand process itself aims beyond into far inconceivabilities” [7 a]. Acknowledgements I cordially thank Hans Kelker for stimulation and encouragement and for unforgettable hours in Halle. The extensive computer searches by Christian Bohley, Frank Klimaszewsky and Werner Loch concerning “living systems” and the molecular modeling contributions of Christian Bohley and Maren Koban are gratefully acknowledged.
4 References [ l ] Transitions: a) I. Prigogine, 1. Stengers, Das Paradox der Zeit (Time,Chaos and the Quantum. Towards the Resolution of the Time Paradox), Piper, Munich 1993; Science 1978, 201, 777; T. Petrosky, I. Prigogine, Adv. Chem. Phys. 1997, 99, 1; Z. Naturjorschung 1997, 52, 37; b) C. F.
440
VIII
Living Systems
von Weizsacker, Die philosophische Interpretation der modernen Physik, Nova Acta Leopold. 1986, NF/207, 3712; Zeit und Wissen, Hanser, Munich 1992. [2] Universeprocessings: a) J. S . Trefil, The Moment of Creation, Scribner’s, New York 1983; b) S . Weinberg, Dreams qf Final Theory, Pantheon Books, New York 1993; The First Three Minutes, Basic Books, New York 1977; Sci. Am. 1994, 271 ( l o ) ,22; c) S . W. Hawking, A Brief History of Time,Bantam, New York 1988; Black Holes and Baby Universes, Bantam, New York 1993; Phys. Rev. D 1995, 52, 5681; R. Bousso, S. W. Hawking, Nucl. Phys. B 1997, 57, 201; S. W. Hawking, R. Penrose, Mercury 1997,26,33. [3] Versus processing of universes: A. Linde: Elementarteilchen und inflationarer Kosmos, Spektrurn, Heidelberg 1993; Spektrum Wiss. 1995, ( I ) , 32; Phys. Rev. D 1997, 56, 1841; A. Linde, D. Linde, A. Mezhlumian, Phys. Rev. D 1994, 49, 1783. [4] The arrow of evolution: a) E. Jantsch, Die Selbstorganisation des Universums, Hanser, Munich 1992; b) M. Gell-Mann, The Quark and the Jaguar, Freeman, New York 1994; c) R. Penrose, The Emperor’s New Mind Concerning Computers, Minds, and the Law of Physics, Oxford University Press, New York 1989; Computerdenken, Spektrum, Heidelberg 1991; Shadows and the Mind, Oxford University Press, New York 1994; Schatten des Geistes, Spektrum Heidelberg 1995; Astrophys. Space Sci. 1996, 244, 229; General Relativity Gravitation 1996, 28, 581; Intern. Stud. Philosophy Sci. 1997, 1 I , 7; S . W. Hawking, R. Penrose, C. Cutler, Am. J. Phys. 1997,65,676;d) F. J. Tipler, The Physics of Immortality, Doubleday, New York 1994; e) C. E. Shannon, W. Weaver, The Mathematical Theory of Communication, University of Illinois Press, Urbana, 11. 1964. [5] Transitions to life: a) J. Monod, L’Hazard et la ne‘cessite‘,Seuil, Paris 1970; b) M. Eigen, Stufen zum Leben, Piper, Munich 1987; Cold Spring Harbor Symp. Quant. Biol. 1987, LII, 307; Naturwissenschaften 1971,58,465;Trends Microbiol. 1996,4,216;c) C. de Duve, Blue-printfor a Cell, The Nature and Origin of Life, Patterson, Burlington 1991; Ursprung des Lebens, Spektrum, Heidelberg 1994; d) Reproduction of Supramolecular Structure, (Eds.: G. R. Fleischaker, S . Colonna, P. L. Luisi), NATO ASI Sel: Vol. 496, Kluwer, Dordrecht 1994. [6] The evolutionary field: a) R. Lewin, Complexity - Life at the Edge of Chaos, Macmillan, New York 1992; Die Komplexitatstheorie, Hoffmann & Campe, Hamburg 1993; b) F. Cramer: Chaos and Order, VCH, Weinheirn 1993; Der Zeitbaum, Insel, Franfurt a. M. 1993; Symphonie des Lebendigen, Insel, Frankfurt a. M. 1996.
[7] Coherences: a) S. Hoffmann, Molekulare MatriZen ( I Evolution, I1 Proteine, III Nucleinsauren, IVMembranen), Akademie-Verlag, Berlin 1978; b) H. Haken, Advanced Synergetics, Springer, Berlin 1983; Naturwissenschaften 1981, 68, 293; Rep. Progl: Phys. 1989, 52, 513; c) J. C. Eccles, Evolution of the Brain: Creation of the Self, Routledge, London 1989; d) D. Zohar, The Quantum Self, Quill/Morrow, New York 1990; Quantum Society, QuilllMorrow, New York 1994. [8] From physics to genetics: a) W. Ebeling, R. Feistel, Physik der Selbstorganisation und Evolution, Akademie-Verlag, Berlin 1982; H. Herzel, W. Ebeling, A. 0. Schmitt, M. A. JimCnez Montafio, in From Simplicity to Complexity in Chemistry - and Beyond (Eds.: A. Miiller, A. Dress, F. Vogtle), Vieweg, Braunschweig 1996, p. 13; b) C. Tudge, The Engineer in the Garden. Genes and Genetics. From the Idea of Heredity to the Creation of Life, Cape, London 1993; Wir Herren der Schiipfung, Spektrum, Heidelberg 1994; c) B. Alberts, D. Bray, J. Lewis, M. Raff, K. Roberts, J. D. Watson, Molecular Biology ofthe Cell, Garland, New York 1983. [9] X-ray proof of Watson-Crick: R. E. Dickerson, in Methods of Enzymology (Eds.: D. M. J. Lilley, J. E. Dahlberg), Academic Press, New York 1992, p. 67. [lo] C. C. F. Blake, Endeavour 1978,2, 137. [ 1 1 1 “Advanced’ molecular modeling: P. K. Weiner, R. Langridge, J. M. Blaney, R. Schaefer, P. A. Kollman, Proc. Natl. Acud. Sci. USA 1982, 79, 3754. [I21 J. 0. Burns, Kosmologie und Teilchenphysik, Spektrum, Heidelberg 1990. [I31 Liquid crystals - history and facts: H. Kelker, a) Mol. Cryst. Liq. Cryst. 1973,21, 1; b) in Zehn Arbeiten uber flussige Kristalle, 6. Fliissigkristallkonferenz sozialistischer Lander, Halle 1985 (Ed.: H. Sackmann), Wiss. Beitr. MartinLuther-Univ. 1986/52 (N 17), 193; c) Naturwiss. Rundschau 1986, 39, 239; d) Mol. Cryst. Liq. Cryst. 1988, 165, I ; e) Liq. Cryst. 1989, 5, 19; f) H. Kelker, R. Hatz, Handbook of Liquid Crystals, VCH, Weinheim 1980; g) H. Kelker, B. Scheurle, Angew. Chem. 1969, 81, 903; Angew. Chem. Int. Ed. Engl. 1969, 8, 884; h) H. Kelker, B. Scheurle, J. Sabel, J. Jainz, H. Winterscheidt, Mol. Cryst. Liq. Cryst. 1971, 12, 113; i) H. Trebin, Phys. Bl. 1988, 44, 221; k) H. Falk, P. Laggner, &err. Chem. Ztg. 1988, 9, 251; 1) V. Vill, Mul. Cryst. Liq. Cryst. 1992, 213,67; m) Condens. Matter News 1992, 1,25; n) V. Vill, J. Thiem, LIQCRYST - Datenbunk jlussigkristalliner Verbindungen, Springer, Heidelberg 1992; 0 ) R. Brauns, Flussige Kristalle und Lebewesen, Schweizerbart, Stuttgart 1931.
4 [ 141 Liquid crystals
- the romantic period: 0. Lehmann, a) Molekularphysik, Engelmann, Leipzig 1888/89; b) Plastizitut von ,fliissigen Kristallen sowie ihre Umlagerungen, Engelmann, Leipzig 1904; c ) Fliissige Kristalle und die Theorien des Lebens, 78. Versammlung deutscher Naturforscher in Stuttgart, 1906; d) Die scheinbar lehenden Kristalle, Schreiber, Esslingen 1907; e) Die neue Welt der fliissigen Kristalle uiid deren Bedeutung f u r Cheinie, Technik und Biologie, Akademische Verlagsgesellschaft, Leipzig 1911; f) Die Lehre von den fliissigen Kristallen und ihre Beziehungen zu den Problemen der Biologie, Bergmann, Wiesbaden 1918; g) Ergeb. Physiol. 1918,16,255;h) Fliissige Kristalle und ihr scheinbares Leben - Forschungsergebnisse dargestellt in einem Kinqfilm, Voss, Leipzig 1921. [IS] Phase beauties in liquid crystals: T. E. Strzelecka, M. W. Davidson, R. L. Rill, Nature 1988, 331,457. [I61 Continued in cellular beauties: P. Jaret, L. Nilsson, Natl. Geograph. Soc. USA 1986, 169, 702. [I71 Life sciences - biomesogen view: a) S. Hoffmann, W. Witkowski, in Mesomorphic Order in Polymers and Polymerization in Liquid Crystalline Media (Ed.: A. Blumstein), Am. Chem. Soc. Symp. Ser. 1978, 74, 178; Pnlym. Prepr. 1977, 18,45; b) S. Hoffmann, in Darwin Today (Eds.: E. Geissler, W. Scheler) Akademie-Verlag, Berlin 1983, 193; c) in Liquid Crystals and Ordered Fluids (Eds.: A. C. Griffin, J. F. Johnson, R. S. Porter), Mol. Cryst. Liq. Cryst. 1984, 110, 277; d) in Polymeric Liquid Crystals (Ed.: A. Blumstein), Plenum, New York 1985, 432; Polym. Prepc 1983,24,259;e) Z. Chem. 1987,27,395; f) Wiss. Z. Univ. Halle 1989, 38/H4, 3; ibid. 1992,41/H6, 37, 5 1 ; g) in Polymers and Biological Function (Ed.: D. Braun), Angew. Mukromol. Chem. 1989, 166/167, 81; h) see also: H. Kelker, in Zehn Arbeiten iiberjliissige Kristalle (Ed.: H. Sackmann), Wiss. Beitc MLU 1986/52, (N. 17), 193. [ 181 Liquid crystals - molecular biological aspects: a) S . Hoffmann, W. Witkowski, in Wirkungsmechanismen von Herbiciden und synthetischen Wachstumsregulatoren, RGW-Symposium Halle 1972 (Eds.: A. Barth, F. Jacob, G. Feyerabend), Fischer, Jena 1972/3, p. 29 1, 306; b) Nucleic Acids Res. 1975, SP I, 137; c) NiicleicAcidsSymp.Seu. 1978,4,22I;d)S.Hoffmann, W. Witkowski, S.-R. Waschke, A. Veckenstedt, Nucleic Acids Symp. Ser. 1981, 9, 239; e) S . Hoffmann, Wiss.Z. Univ. Halle 1983,32/H4, 5 I ; f) Nucleic Acids Symp. Ser. 1984, 14, 7; in Plenary Lectures - Symp. Chem. Heterocycl. Comp. (VIIIth)and Nucleic Acids Comp. (VIth)(Eds.: J. Beranek, A. Piskala), Czechoslovak. Acad. Sci., Prague 1985, p. 45; g) in Wirkstojforschung
[ 191
[20]
[21] 1221 [23]
1241
References
44 I
1Y80 (Ed.: H. Possin), Wissenschaftspubl. Martin-Luther-Univ., Halle 1981/2, 11, 35; h) S . Hoffmann, W. Witkowski, i n Wirkstojjfiirschung ’82 (Ed.: H. Posing), Wissenschaftspubl. Martin-Luther-Univ., Halle 1984, p. 102; i ) S . Hoffmann, in Wirkstojforschung (Eds.: A. Barth, H. Possin), Wiss. Beitr. Martin-Luther-Univ,, Halle 1988, 3(P32), 48; k) in 2nd Swedish-German Workshop on Modern Aspects of Chemistry and Biochemistry of Nucleic Acids (Ed.: H. Seliger), Nucleosides Nucleotides 1989, 7,555; 1) in Chirality - From Weak Bosons to the &Helix (Ed.: R. Janoschek), Springer, Heidelberg 1991, p. 205; m) Angew. Chem. 1992, 104, 1032; Angew. Chem. lnt. Ed. Engl. 1992, 31, 1013; n) in Reproduction of Supramolecular Structure (Eds.: G. R. Fleischaker, S . Colonna, P. L. Luisi), Kluwer, Dordrecht NATO AS1 See 1994, 496, 3. “Soft matter”: P. G. de Gennes, The Physics of Liquid Crystals, Oxford University Press, Oxford 1984; Angew. Chem. 1992, 104, 856; Angew. Chem. Int. Ed. Engl. 1992, 31, 842; P.-G. de Gennes, M. HCbert, R. Kant, Macromol. Symp. 1997, 113, 39; see also [24e]. The “predecessors”: a) W. Heintz, Jahresber. Chem. 1849, 342 (GieRen 1850); 1852, 506 (GieBen 1853); 1854,447 (GieBen 1855); b) .I. Prakt. Chem. 1855, 66, 1; c) P. Duffy, J. Chem. Soc. 1852, 5, 197; d) R. Virchow, Virchowk Archiv 1854, 6 , 571; e) M. Berthelot, C. R. Acad. Sci. 1855, 41,452; f) J. Prakt. Chem. 1856, 67, 235; f) C. Mettenheimer, Con-Blatt d. Verein.f: gem. Arbeit zur Forderung d. wiss. Heilkunde 1857,24, 33 1 ; g) L. Planer, Liebigs Ann. Chem. 1861, 118, 25; h) W. Loebisch, Bee Dtsch. Chem. Ges. 1872,5,510. The beginning of the history of liquid crystals: F. Reinitzer, Monatsh. Chem. 1988, 9, 421. P. M. Knoll, Fridericiana - Z. Univ. Karlsruhe 1981, 43. Last-century presentiments of inanimate- animate world coherences: E. Haeckel, a) Kristallseelen, Kroner, Leipzig 1917; b) Kunstformen der Natur, Bibl. Inst. Leipzig 1899-1900; c) F. Rinne, GrenZfragen des Lehens, Quelle & Meyer, Leipzig 1931; d) Ber. Naturforsch. 1930, 30, 1; e) Naturwissenschaften 1930, 18, 837. Liquid crystals - beginning of the academic period: D. Vorlander, a) Kristallin-jlussige SubstanZen, Enke, Stuttgart 1908; b) Chemisehe Kristullographie der Fliissigkeiten, Akademische Verlagsgesellschaft, Leipzig 1924; c) D. Vorlander, Ber. Dtsch. Chem. Ges. 1907, 40, 1970; ibid. 1908, 41, 2035; d) ibid. 1910, 43, 3120; e) Z. Phys. Chem. 1919, 105, 211; f) D. Vorlander, W. Selke, ibid. 1928, 229, 435; g) R. Schenck, Kristallinische Fliissigkeiten und fliissige Kristalle, Engelmann, Leipzig 1905; h) Ann. Phys.
442
VIII
Living Systems
1902, 9, 1053; i) Z. Elektrochem. 1905,50,951; k) F. Hoffmann, personal communication. [25] Liquid-crystal views: a) G. W. Gray, Molecular Structure and the Properties of Liquid Crystals, Academic Press, New York 1962; b) S. Chandrasekhar, Liquid Crystals, Cambridge University Press, New York 1977; c) J. D. Litster, J. Birgeneau, Phys. Today 1982, 35 (5),52. [26] Sleeping beauty’s reawakening: a) F. Sauerwald, Z. Metallkunde 1954,45, 257; b) H. Sackmann, Z. Elektrochem. 1955, 5Y, 880; c) H. Arnold, H. Sackmann, ibid. 1959,63, 117 1; Z. Phys. Chem. (Leipzig) 1960, 213, 137; d) S. Hoffmann, Diplomathesis, Halle University 1957; e ) G. Jaenecke, Diplomathesis, Halle University 1957; f) S. Hoffmann, G. Jaenecke, W. Brandt, W. Kumpf, W. Weissflog, G. Brezesinski, Z. Chem. 1986, 26, 284; g) H. Schubert, Wiss. Z. Univ. Halle 1981, 30, 126. [27] And animation: a) H. Sackmann, Nova Acta Leopoldina 1977, NF 47, 137; in Zehn Arbeiten iiber Piissige Kristalle (Ed.: H . Sackmann), Wiss. Beitr. Martin-Luther-Univ. 1986/52, (N 17), 193; b) D. Demus, H. Zaschke, Fliissige Kristalle in Tabellen I und 11, Deutscher Verlag fur Grundstoffindustrie, Leipzig 1974, 1984; c) D. Demus, S . Diele, S. Grande, H. Sackmann, Adv. Liq. Cryst. 1983, 6, 1; D. Demus, L. Richter, Textures of Liquid Crystals, Deutscher Verlag fur Grundstoffindustrie, Leipzig 1978;d) H.D. Dorfler, S. Diele, M. Diehl: Colloid Polym. Sci. 1984, 262, 139; e) H.-J. Deutscher, H.-M. Vorbrodt, H. Zaschke, Z. Chem. 1981,21,8;f) H. Zaschke, A. Isenberg, H.-M. Vorbrodt, in Liquid Crystals and Ordered Fluids (Ed.: A Griffin), Plenum, New York 1984,75; g) S . Hoffmann, Z. Chem. 1979, 19, 241, ibid. 1982, 22, 357; ibid. 1987,27, 395. [28] Strange discoid versus rod-like artificial classic and native alien: a) C. Destrade, H. Gasparoux, P. Foucher, N. H. Tinh, J. Malthete, J. Jacques, J. Chim. Phys. 1983, 80, 137; b) R. Blinc, Ber. Bunsenges. Phys. Chem. 1983, 87, 271; c) W. Schadt, M. Petrzila, P. R. Gerber, A. Villiger, Mol. Cryst. Liq. Cryst. 1985, 122, 241; d) R. Langridge, T. E. Ferrin, I. D. Kuntz, M. L. Conolly, Science 1981, 211, 661 - courtesy of Laurence H. Hurley. [29] Structure and phase - routes to mesogenic nucleoprotein systems: a) F. Miescher, “On the chemical composition of pyocytes”, in HoppeSeyler’s med. Untersuchungen 1971; in Die histochemischen und physiologischen Arbeiten (Ed.: F. C. W. Vogel), Leipzig 1897; b) R. Altmann, Arch. Anat. Phys. Phys. Abt. 1889; Die Elementarorganismen, Veit, Leipzig 1890; c) E. Schroedinger, What is Life?, Cambridge University Press, New York 1944; d) L. Pauling, R. B. Corey, H. R. Branson, Proc. Natl. Acad. Sci. USA
1951,37,205,735;L. Pauling, The Nature of the Chemical Bond, Cornell University Press, Ithaca, NY 1960; e) E. Chargaff, Experientia 1950, 6,201; ibid. 1970,26, 810; On some ofthe Biological Consequences of Base-Pairing in the Nucleic Acids in Development and Metabolic Control Mechanisms and Neoplasia, Williams and Wilkins, Baltimore, MD 1965; Nature 1974, 248,776; 0 J. D. Watson, F. H. C. Crick, Nature 1953, 177, 964; F. H. C. Crick, J. D. Watson, Proc. R. Soc. 1954,A233,80;g) R. A. Weinberg, Sci. Am. 1985, 253, 48; h) Die Molekiile des Lebens (Ed.: P. Sitte) Spektrum Wiss. 1988; see also [Sd, 7a, 17, 181. [30] Life sciences - the supramolecular view: a) J. M. Lehn, Science 1985, 227, 849; Angew. Chem. 1988, 100, 91; Angew. Chem. Int. Ed. Engl. 1988,27,89;ibid. 1990,102, 1347 (1990, 29, 1304); Chem. Biol. 1994, 1, XVIII; Supramolecular Chemistry, VCH, Weinheim 1995; b) R. Kramer, J.-M. Lehn, A. Delian, J. Fischer, Angew. Chem. 1993,105,764;Angew. Chem. Int. Ed. Engl. 1993, 32, 703; c) H. Ringsdorf, Supramol. Sci. 1994, I, 5 ; - “Order and Mobility” between “Life Sciences” and “Advanced Materials”; d) H. Ringsdorf, B. Schlarb, J. Venzmer, Angew. Chem. 1988, 100, 117; Angew. Chem. Int. Ed. Engl. 1988, 27, 113; e) M. Ahlers, W. Miiller, A. Reichert, H. Ringsdorf, J. Venzmer, ibid. 1990,102,1310 (1990,29,1269); f) S. Denzinger, A. Dittrich, W. Paulus, H. Ringsdorf, in From Simplicity to Complexity in Chemistry - and Beyond, (Eds.: A. Miiller, A. Dress, F. Vogtle), Vieweg, Braunschweig 1996, p. 63; g) K. Fujita, S. Kimura, Y. Imanishi, E. Rump, H. Ringsdorf, Adv.Biophys. 1997,34,127; h) D. Philp, J. F. Stoddart, Angew. Chem. 1996, 108, 1242; Angew. Chem. Int. Ed. Engl. 1996, 35, 1 154; Order-disorder coherences between chaos and complexity: see also [5 d, 7a, 17, 181. [31] Parity violation and chirality development: a) L. Pasteur, in LeFons de chimie professe‘es en 1860 par M M , 1886; L. Pasteur, A. A. T. Cahours, C. A. Wurtz, P. E. M. Berthelot, H. E. Sainte-Claire Deville, R. F. Banal, J.-B. A. Dumas, Sur la dissyme‘trie mole‘culaire, Collection Epistkme Paris Hachette Paris, 1861; b) R. Janoschek, Naturwiss. Rundschau 1986, 39, 327; in Chirality From WeakBosons to the a-Helix (Ed.: R. Janoschek), Springer, Berlin 1991, p. 18; c) H. Latal, ibid., p. 1 ; d) S. Weinberg, Science 1980, 210, 1212; e) S. Mason, Chem. BI: 1985, 21, 538; Trends Pharmacol Sci. 1986, (7,20; Nouv. J. Chim. 1986, 10, 739; f) G. E. Tranter, Nature 1985,318,172;Mol. Phys. 1985,56,825;Nachr. Chem. Techn. Lab. 1986,34,866; J. Theor. Biol. 1986,119,469; g) D. K. Kondepudi, G. W. Nelson, Nature 1985,314, 438; h) D. Z. Freedman, P. van Nieuwenhuizen, in Teilchen, Felder und
4 Symmetrien, Spektrum, Heidelberg 1988, p. 170; in Kosmologie und Teilchenphysik, Spektrum Heidelberg 1990, p. 120; i) M.-A. Bouchiat, L. Pottier, in Elementare Materie, Vakuumund Felder, Spektrum, Heidelberg 1988, p. 130; k) K. Mainzer, Chimia 1988, 42, 161; 1) Chirality (Eds.: I. W. Wainer, J. Caldwell, B. Testa), 1989, p. 1; m)M. Quack, Angew. Chem. 1989, 101, 578; Angew. Chem. Int. Ed. Engl. 1989,28, 959; n) R. A. Hegstrom, D. K. Kondepudi, Sci. Am. 1990, 262, 98; 0 ) G. Jung, Angew. Chem. 1992, 104, 1484;Angew. Chem. lnt. Ed. Engl. 1992,31, 1457; see also 118 1-111. [32] Evolutionary maternal inorganic matrices: a) M. Paecht-Horowitz, J. Berger, A. Katchalsky, Nature 1970,228,636; Angew. Chem. 1973,85, 422;Angew. Chem. lnt. Ed. Engl. 1973,12,349; b) A. G. Cairns-Smith, Genetic Takeover andthe Mineral Origins of Life, Cambridge University Press, New York 1982; c) H. G. von Schnering, R. Nesper, Angew. Chem. 1987, 99, 1087; Angew.Chem. lnt. Ed. Engl. 1987,26,1059;d) R. J. P. Williams, Biochem. Soc. Trans. 1990,18,689; e) L. Addadi, S. Weiner, Angew. Chem. 1992, 104, 159;Angew. Chem. lnt. Ed. Engl. 1992,31, 153; f) S. J. Lippard, J. M. Beng, Principles of Bioinorganic Chemistry, University Science Books, Mill Valley, CA 1994; Bioanorganische Chemie, Spektrum, Heidelberg 1994; g) D. E. Devos, D. L. Vanoppen, X.-Y. Li, S. Libbrecht, Y. Bruynserade, P. P. Knops-Gerrits, P. A. Jacobs, Chem. Eu,: J . 1995,1, 144; h) A. Fironzi, D. Kumar, L. M. Bull, T. Bresier, P. Singer, R. Huo, S. A. Walker, J. A. Zasadzinski, C. Glinka, J. Nicol, D. Margolese, G. D. Stucky, B. T. Chmelka, Science 1995,267, 1138. [33] Constituents of nucleoprotein system: a) W. Guschlbauer, Nucleic Acid Structure, Springer, Berlin 1976; in Encyclopedia of Polymer Science and Engineering, Wiley, New York 1988, Vol. 12, p. 699; b) Y. Mitsui, Y. Takeda, Adv. Biophys. 1979, 12, 1; c)Methods Enzymology (Eds: D. M. J. Lilley, J. E. Dahlberg), Academic, San Diego, CA 1992; d) R. W. Roberts, D. M. Crothers, Science 1992, 258, 1463; e) 0. Kennard, W.-N. Hunter, Angew. Chem. 1991, 103, 1280; Angew. Chem. lnt. Ed. Engl. 1991, 30, 1254; f) Structural Biology 1/11 (Eds.: R. H. Sarma, M. H. Sarma), Adenine Press, New York 1993 -with emphasis also on V. I. Ivanov’s concluding remarks: “Simplicity Lost”; g) R. E. Dickerson, I. Geis, The Structure and Action of Proteins 1971; Struktur und Funktion der Proteine, VCH, Weinheim 1971; h) R. E. Dickerson, M. L. Kopka, P. Pjura, Chem. Sc,: 1986, 26B, 139; i) H. A. Scheraga, K.-C. Chou, G. NCmethy, in Conformations in Biology (Eds.: R. Srinavasan, R. H. Sarma), Adenine Press, New York 1983, p. 1; k) R. Jaenicke, Prog. Biophys.
References
443
Mol. Biol. 1987, 49, 117; Angew. Chem. 1984, 96, 385;Angew. Chem. Int. Ed. Engl. 1984, 23, 395; Naturwissenschaften 1988, 75, 604; I ) M. Mutter, Angew. Chem. 1985, 97, 639; Angew. Chem. lnt. Ed. Engl. 1985, 24, 639; M. Mutter, S . Vuilleumier, ibid. 1989, 101, 551 and 1989, 28, 535; m) W. F. Anderson, J. H. Ohlendorf, Y. Takeda, B. W. Matthews, Nature 1981,290,754; D. H. Ohlendorf, W. F. Anderson, Y. Takeda, B. W. Matthews, J. Biomol. Struct. Dyn. 1983, I , 533; n) S.-H. Kim, in NucleicAcids Research: Future Developments (Eds.: K. Mizoguchi, I. Watanabe, J. D. Watson), Academic, New York 1983, p. 165; 0 ) E. Katchalski-Katzir, Makromol. Chem. Macromol. Symp. 1988, 19, 1; p) L. Stryer, Biochemistry, Freeman, New York 1988; Biochemie, Spektrum, Heidelberg 1990; q) D. Voet, J. G. Voet, Biochemistry, Wiley, New York 1990; Biochemie, VCH, Weinheim 1992. [34] Emergence of an “RNA world”: a) S. Altman, Angew. Chem. 1990,102,735;Angew. Chem. lnt. Ed. Engl. 1990, 29, 707; b) T. R. Cech, ibid. 1990, 102, 745 and 1990, 29, 716; Spektrum Wiss.1987, (I), 42; c) T. Inoue, F. Y. Sullivan, T. R. Cech, J. Mol. Biol. 1986,189, 143; d) R. B. Waring, P. Towner, S . J. Minter, R. W. Davies, Nature 1986, 321, 133; e) P. A. Sharp, Angew. Chem. 1994, 106, 1292; Angew. Chem. Int. Ed. Engl. 1994, 33, 1229. [35] Proteins’ own ambitions: a) Yu. A. Ovchinnikov, V. T. Ivanov, Tetrahedron 1975, 31, 2177; b) Yu. A. Ovchinnikov, Bioorganicheskaya chimya, Prosvyeschenye, Moscov 1987. [36] Highlighting classical organic chemistry: a) E. Fischer, Ber. Dtsch. Chem. Ges. 1894,27,2985, 3189; b) E. Fischer, B. Helferich, Liebigs Ann. Chem. 1911,383,68; c) E. Fischer, K. Freudenberg, Ber. Dtsch. Chern. Ges. 1915, 45, 915; d) W. Langenbeck, Lehrbuch der Organischen Chemie, Steinkopf, Dresden 1954; e) R. B. Woodward, in Perspectives in Organic Chemistry (Ed.: A. Todd), Interscience, New York 1956, p. 155; f) A. Eschenmoser, Angew. Chem. 1988, 100,S;Angew. Chem. lnt. Ed. Engl. 1988,27,5; g) A. Eschenmoser, ibid. 1994, 106, 2455 and 1994, 33, 2363; h) F. W. Lichtenthaler, ibid. 1992,104, 1577 and 1992,31, 1541; ibid. 1994, 106,2456 and 1994,33,2364;i) D. E. Koshland, ibid. 2468 and 2375; k) E. H. Fischer, ibid. 1993, 105, 1181 and 1993,32, 1130. [37] Complex “rod-like’’ virus: L. Onsager,Ann. N. K Acad. Sci. 1949,51,62Z. [38] Mesogenic nucleation capabilities of nucleoprotein system: a) C. W. Carter, J. Kraut, Proc. Natl. Acad. Sci. USA 1974, 71,283; b) G. M. Church, J. L. Sussman, S.-H. Kim, ibid. 1977, 74, 1458. [39] Omne vivum e matribus - highly condensed states and the mesogenic matrix problem: a) A. Todd, in Perspectives in Organic Chemistry (Ed.:
444
VIII
Living Systems
A. Todd), Interscience, New York 1956, p. 245; b) G. Schramm, H. Grotsch, W. Pollmann, Angew. Chem. 1962, 74.53;Angew. Chem. Int. Ed. Engl. 1962, I , 1; c) J. H. Winter, Angew. Chem. 1966,78, 887;Angew. Chem. Int. Ed. Engl. 1966, 5, 862; d) W. Kern, K. Kammerer, Chem. Ztg. 1967, 91,73; e) H. Kammerer, ibid. 1972,96,7; f) S . Hoffmann, Habilitationthesis Halle University 1968;Z. Chem. 1979,19,241;ibid. 1982,22, 357; g ) S. Hoffmann, H. Schubert, W. Witkowski, ibid. 1971,11,345,465; ibid. 1974,14, 154; h) D. Pfeiffer, L. Kutschabsky, S . Hoffmann, Cryst. Struct. Commun. 1982, 11, 1635; i) D. Scharfenberg-Pfeiffer, R.-G. Kretschmer, S. Hoffmann, Cryst. Res. Technol. 1988, 23, 881; k) G. Challa, Y. Y. Tan, Pure Appl. Chem. 1981, 53,627; I ) A. I. Scott, Angew. Chem. 1993, 105, 1281; Angew. Chem. Int. Ed. Engl. 1993, 32, 1223; see also [5d, 7a, 17, 181-n, 29, 301. [40] Evolutionists: a) T. C. Lucretius, De rerum natura, Weidmann, Leipzig 1795; b) A. I. Oparin, Origin of Life, Moskau 1924; The Origin of Life on Earth, Academic, New York 1957; c) J. B. S . Haldane, The Course of Evolution, Longman, New York 1932; d) S. W. Fox, Science 1960, 132, 200; The Origin of Prebiological Systems and of Their Molecular Matrices (Ed.: S . W. Fox), Academic, New York 1965; e) S. Spiegelman, Q. Rev. Biophys. 1971,2,213; f) H . Kuhn, Angew. Chem. 1972, 84, 838; Angew. Chem. Int. Ed. Engl. 1972,11,798;g) L. E. Orgel, The Origins of Life on Earth, Prentice Hall, Englewood Cliffs, NJ 1974; h) M. Eigen, P. Schuster, Naturwissenschaften 1977,64,541; ibid. 1978, 65, 7; M. Eigen, Chem. Scr 1986, 26B, 13; i) J. J. Wolken, in: Molecular Evolution and Protobiology (Eds.: K. Matsuno, K. Dose, K. Harada, D. L. Rohlfing), Plenum, New York 1984, 137; k) E. Szathmhry, in [5d], p. 65; J. M. Smith, E. Szathmhry, The Major Transitions in Evolution, Freeman, Oxford 1995; Evolution, Spektrum, Heidelberg 1996; 1) F. J. Varela, in [5d], p. 23; m) L. F. Orgel, Sci. Am. 1994,271(10),53;seealso[1-8,17,18,30]. [41] Approaches to directed artificial evolution: a) G. F. Joyce, Cold Spring Harbor Symp Quant Biol 1987, LII, 41; Nature 1989, 338, 217; b) A. A. Beaudry, G. F. Joyce, Science 1992, 257, 635; c) G. F. Joyce, Sci. Am. 1992, 269 (12), 49; d) R. R. Breaker, G. F. Joyce, Proc. Natl. Acad. Sci. USA 1994, 91, 6093; e) J. Tsang, G. F. Joyce, Biochemistry 1994, 33, 5966; f ) R . R . Breaker, G. F. Joyce, Chem. & B i d . 1994, I , 223; g) R. R. Breaker, G. F. Joyce, in [5d], 127; h) T. Li, Z. Zeng, V. A. Estevez, K. U. Baldenius, K. C. Nicolaou, G. F. Joyce, J. Am. Chem. Soc. 1994,116,3709; i) A. Pluckthun, L. Ge,Angew. Chem. 1991, 103, 301; Angew. Chem. Int. Ed. Engl. 1991,30,296;k) G. Jung, A. G. Beck-
Sickinger, ibid. 1992, 104, 375 (1992, 31, 367); 1) D. P. Bartel, J. W. Szostak, Science 1992,261, 1411; m) D. E. Huizenga, J. W. Szostak, Biochemistry 1995, 34, 656; n) J. W. Szostak, Protein Engin. 1995,8,3; Chem. Rev. 1997,97,347; 0)A. J. Hager, J. W. Szostak, Orig. Life 1996, 26, 269; p) J. R. Lorsch, J. W. Szostak, Acc. Chem. Res. 1996, 29, 103; q) J. S. McCaskill, Nachr Chem. Techn. Lab. 1995, 43, 199; r) G. von Kiedrowski, Angew. Chem. 1991, 103,839; Angew. Chem. Int. Ed. Engl. 1991, 30, 892; s) P. Burgstaller, M. Famulok, ibid. 1995, 107, 1303 (1995, 34, 1188); t) M. Egli, ibid. 1996, 108,486,2020 (1996,35,432,1894); ibid. 1997, 109, 494 (1997, 36, 480); u) H. Seliger, R. Bader, M. Hinz, B. Rotte, A. Astriab, M. Markiewicz, W. T. Markiewicz, Nucleosides Nucleotides 1997, 16,703; see also [5-7, 171. [42] The enzyme-free approach in nucleic acid matrix reactions: a) L. E. Orgel, R. Lohrmann, Acc. Chem. Res. 1974, 7,368; b) R. Lohrmann, L. E. Orgel, J. Mol. Biol. 1980,142,555; c) T. Inoue, L. E. Orgel, Science 1983,219,859;d) T. Inoue, G. F. Joyce, K. Grzeskowiak, L. E. Orgel, J. M. Brown, C. B. Reese, J. Mol. Biol. 1984,178,669; e) C. B. Chen, T. Inoue, L. E. Orgel, ibid. 1985, 181, 271; f) A. W. Schwartz, L. E. Orgel, Science 1985, 228, 585; g)O. L. Acevedo, L. E. Orgel, Nature 1986, 321, 790; h) L. E. Orgel, Cold Spring Harbor Symp. Quant. Biol. 1987, LII, 9; J. Theor Biol. 1986, 123, 127; Folia Biol. (Praha) 1983, 29, 65; J. Mol. Biol. 1995, 181, 271; Orig. Life 1996, 26, 261; i) J. P. Ferris, A. R. Hill Jr., R. Liu, L. E. Orgel, Nature 1996,381,59; k) B. C. F. Chu, L. E. Orgel, Bioconj. Chem. 1997,8, 103; see also [43,44]. [43] Self-replicative systems: a) L. E. Orgel, Nature 1992,358,203; b) S . Hoffmann, Angew. Chem. 1992, 104, 1032; Angew. Chem. Int. Ed. Engl. 1992, 31, 1013; in [5d], 3; c) J. Rebek, Jr., Sci. Am. 1994, 271 (7), 34; in [5d], 75; d) D. Sievers, T. Achilles, J. Burmeister, S. Jordan, A. Terford, G. von Kiedrowski, in [5d], 45; see also: e) P. Strazewski, C. Tamm, Angew. Chem. 1990,102,37;Angew. Chem. Int. Ed. Engl. 1990, 29, 36; see also [5-7, 17, 18, 301. [44] “Minimal models” of replication: a) G. von Kiedrowski, Angew. Chem. 1986, 93, 932; Angew. Chem. Int. Ed. Engl. 1986,25,932; in 40 Jahre Fonds der Chemischen Industrie 1950-90, VCI, Frankfurt 1990; Bioorg. Chem. Frontiers 1993, 3, 113; Nature 1996, 381, 20; b) W. S. Zielinski, L. E. Orgel, Nature 1987, 327, 346; J. Mol. Evol. 1989,29,281;c) G. von Kiedrowski, G. B. Wlotzka, J. Helbing, Angew. Chem. 1989, 102, 1259; Angew. Chem. Int. Ed. Engl. 1989, 28, 1235; d) G. von Kiedrowski, G. J. Helbing, B. Wlotzka, S. Jordan, M. Mathen, T. Achilles, D. Severs, A. Terfort, B. C . Kahrs, Nachr Chem.
4 Techn. Lab. 1992, 40, 578; e) A. G. Terfort, G. von Kiedrowski, Angew Chem. 1992,104,626; Angew. Chem. lnt. Ed. Engl. 1992,31,654;f) D. Severs, G. von Kiedrowski, Nature 1994, 369, 221; g) Th. Achilles, G. von Kiedrowski,Angew. Chem. 1993, 105, 1225; Angew. Chem. lnt. Ed. Engl. 1993,32, 1198; h) B. G. Bag, G. von Kiedrowski, Pure Appl. Chem. 1996, 68, 2145; i) J. Burmeister, G. von Kiedrowski, A. D. Ellington, Angew. Chem. 1997, 36, I321 ; Angew. Chem. lnt. Ed. Engl. 1997, 109, 1379; see also [S-7, 18, 39,42, 431. [45] “Extrabiotic replication and self-assembly”: a) J. Rebek Jr., Angew. Chem. 1990, 102, 261; Angew. Chem. Int. Ed. Engl. 1990,29,245; b) T. Tjivikua, P. Ballester, J. Rebek, Jr., J . Am. Chem. Soc. 1990, 112, 1249; c) T. K. Park, Q. Feng, J. Rebek Jr., ibid. 1992, 114, 4529; Science 1992, 256, 1179; d) M. Famulok, J. S. Nowick, J. Rebek Jr., Act. Chem. Scand. 1992,46,3 15; in [ S d], p. 75; e) R. J. Pieters, I. Hue, J. Rehek Jr., Chem. Eur. J. 1995,1, 183; f) J. Kang, J. Rebek Jr., Nature 1996,382, 239. [46] 2-Pyridones as drastically abstracted nucleic acid analogs: a) S. Hoffmann, W. Witkowski, G. Borrmann, H. Schubert, Z. Chem. 1974,14, 154; b) ibid. 1978,18,403;c) F. Persico, J. D. Wuest, J. Org. Chem. 1993, 58, 95. [47] From supramolecular chirality selectors to elementorganic matrices: a) T. Gulik-Krzywicki, C. Fouquey, J.-M. Lehn, Proc. Natl. Acad. Sci. USA 1993, 90, 163; b) J.-M. Lehn, A. Rigault, Angew. Chem. 1988, 100, 1121; Angew. Chem. lnt. Ed. Engl. 1988, 27, 1059; c) G. Bernardinelli, C. Piquet, A. F. Williams, ibid. 1992, 104, 1626 and 1992, 31, 1622; d) R. Kramer, J.-M. Lehn, A. DeCian, J. Fischer, ibid. 1993,105,764 and 1993,32,703;e) M. Kotera, J. M. Lehn, J. P. Vigneron, J. Chem. Soc., Chem. Commun. 1994, 197; see also [l-8, 181-n]. [48] Triplexes between molecular hystereses and self-replication: a) R. A. Cox, A. S. Jones, G. E. Marsh, A. R. Peacocke, Biochim. Biophys. Acta 1956, 21, 576; b) E. Neumann, A. Katchalsky, Ber. Bunsenges. Phys. Chem. 1970, 74, 868; c) Proc. Natl. Acad. Sci. USA 1972, 69, 993; d) A. Katchalsky, E. Neumann, lsr. J. Neuroscience 1972, 3, 175; e) E. Neumann, Angew. Chem. 1973, 85, 430; Angew. Chem. Int. Ed. Engl. 1973,12,356;f) W. Guschlbauer, D. Thiele, M.-Th. Sarvechi, Ch. Marck, in Dynamic Aspects of Conformation Changes in Biological Macromolecules (Ed.: C. Sadron), Reidel, Dordrecht 1973;g) S. Hoffmann, W. Witkowski, H.-H. Riittinger, Z. Chem. 1974, 14, 438; h) S. Hoffmann, W. Witkowski, ibid. 1975, 15, 149; ibid. 1976, 16, 442; i) S. Micciancio, G. Vassallo, Nuovo Cimento 1982,1, 121; k) R. W. Roberts, D. M. Crothers, Science 1992, 258, 1463;
References
445
I) N. T. Thuong, C. HClkne,Angew. Chem. 1993, 105,697;Angew. Chem. lnt. Ed. Engl. 1993,32, 666; m) J. F. Mouscadet, S. Carteau, H. Goulaveic, F. Subra, C. Anclair, J. Biol. Chem. 1994, 269, 21 635; n) T. Li, K. C. Nicolaou, Nature 1994,369,218;0) I. Radnakrishnan, D. J. Patel, Structure 1994, 2, 17; p) V. B. Zhurkin, G. Raghunathan, N. B. Ulyanov, R. D. Camerini-Otero, R. L. Jernigan in [33fl, p. 43; q) C. Y. Sekharudu, N. Yathindra, M. Sundaralingam, in [33 fl, p. 113; r) H. Trauble, H. Eibl, Proc. Natl. Acad. Sci. USA 1974, 71, 214; s ) D . Nachmansohn, ibid. 1976, 73, 82; t) V. N. Soyfer, V. N. Polaman, Triple-Helical Nucleic Acids, Springer, Heidelberg 1996, see also [17, 18, 520, s]. [49] “Minima vita” models: a) P. A. Backmann, P. Walde, P.-L. Luisi, J. Lang, J. Am. Chem. Soc. 1990, 112, 8200; ibid. 1991, 113, 8204; b) P. L. Luisi, F. J. Varela, Orig. Life 1990, 19, 633; c) P. A. Bachmann, P. L. Luisi, J. Lang, Nature 1992, 357, 57; d) P. L. Luisi, P. A. VermontBachmann, M. Fresta, P. Walde, E. Wehrli, J. Liposome Res. 1993, 3, 631; e) P. L. Luisi, in [Sd], 179; f) P. L. Luisi, personal communication; g) T. Oberholzer, R. Wick, P. L. Luisi, C. K. Biebricher, Biochem. Biophys. Res. Commun. 1995,207,250; h) P. L. Luisi, Adv. Chem. Phys. 1996, 92, 425; Orig. Life 1996, 26, 272; i) P. L. Luisi, P. Walde, ibid. 1996,26,429;k) D. Berti, L. Franchi, P. Baglioni, P. L. Luisi, Langmuir 1997, 13, 3438; 1) C. Cescato, P. Walde, P. L. Luisi, ibid. 1997, 13, 4480; see also [5-7, 401. [SO] Playing the game of evolution: a) A. Eschenmoser, Angew. Chem. 1988, 100, 5 ; Angew. Chem. Int. Ed. Engl. 1988,27,5; Nachr. Chem. Techn. Lab. 1991,39,795; Nova Acta Leopold. 1992, NF 67/281, 201; Chem. Biol. 1994, 1, IV; Orig. Life 1994, 24, 389; ibid. 1996, 26, 23.5; NATO ASI Series E 1996, 320, 293; b) A. Eschenmoser, M. Dobler, Helv. Chim. Acta 1992, 75, 218; c) A. Eschenmoser, E. Loewenthal, Chem. Soc. Rev. 1992, I ; d) J. Hunziker, H.-J. Roth, M. Bohringer, A. Giger, U. Diederichsen, M. Gobel, R. Krishnan, B. Jaun, C. Leumann, A. Eschenmoser, Helv. Chim. Acta 1993, 76, 259; e) S. Pitch, A. Eschenmoser, B. Gedulin, S . Hui, G. Arrhenius, Orig. Life 1995, 25, 297; f) A. Eschenmoser, M. V. Kisakiirek, Helv. Chim. Acta 1996, 79, 1249; R. Micura, M. Bolli, N. Windhab, A. Eschenmoser, Angew. Chem. 1997, 109,899; Angew. Chem. lnt. Ed. Engl. 1997,36, 870; g) W. K. Olson, Proc. Natl. Acad. Sci. USA 1977, 74, 1775; see also h) G . Quinkert, Nachr. Chem. Tech. Lab. 1991,39,788; see also [S-71. [SI] Liquid crystals - end of simplicity: a) A. Blumstein, L. Patel, Mol. Cryst. Liq. Cryst. 1978,48, 151; A. Blumstein, S. Vilasagar, S. Ponrathnam, S. B. Clough, R. B. Blumstein, G. Maret, J. Polym. Sci. Polym. Phys. Ed. 1982, 20, 877;
446
VIII
Living Systems
Polymeric Liquid Crystals (Ed.: A. Blumstein) Plenum, New York 1985; R. B. Blumstein, 0. Thomas, M. M. Gauthier, J. Asrar, A. Blumstein, in Polymeric Liquid Crystals (Ed.: A. Blumstein), Plenum,New York, 1985, p. 239; b) G. M. Brown, Am. Sci. 1972,60,64;J. Chem. Ed. 1983, 60,900; G. M. Brown, P. P. Crooke, Chem. Eng. News 1983,60, 24; c) S . Chandrasekhar, Liquid Crystals, Cambridge, University Press, New York 1977; S . Chandrasekhar, B. K. Sadashiva, K. A. Suresh, N. V. Madhusudana, S . Kumar, R. Shashidhar, G. Venkatesh, J. Phys. 1979,40, C3; d) E. T. Samulski, A. V. Tobolsky, in Liquid Crystals and Ordered Fluids (Eds. J. F. Johnson, R. S. Porter), Plenum, New York 1970, p. 1 1 1 ; Liquid Crystalline Order in Polymers (Ed.: E. T. Samulski), Academic Press, New York 1978; E. T. Samulski, Phys. Today 1982, 35 ( 5 ) ,40; Faraday Discuss. Chem. SOC.1985, 79,7; E. T. Samulski, M. M. Gauthier, R. B. Blumstein, A. Blumstein, Macromolecules 1984,17,479;E. T. Samulski, in Polymeric Liquid Crystals (Ed.: A. Blumstein), Plenum, New York 1985, p. 65; e) D. Demus, Mol. Cryst. Liq. Cryst. 1988,165, 45; f) G. W. Gray, A. J. Leadbetter, Phys. Bull. 1977,28, G. W. Gray, Philos. Trans. R. Soc. London 1983, A309, 77; Proc. R. Soc. London A 1985,402, 1; Mol. Cryst. Liq. Cryst., Lett. Sect. 1990, 7,47; Spec. Publ. - R. Soc. Chem. 1991, 88, 203; ibid. 1991, 96, 288; Mol. Cryst. Liq. Cryst. 1991,204,91; G. W. Gray, J. W. Goodby, Smectic Liquid Crystals, Leonard Hill, Glasgow 1984; g) A. C. Griffin, S. R. Vaidya, M. L. Steele, in Polymeric Liquid Crystals (Ed.: A. Blumstein), Plenum, New York 1985, p. 1; h) H. Finkelmann, H. Benthak, G. Rehage, J. Chim. Phys. 1983,80, 163; H. Finkelmann, G. Rehage, Adv. Polym. Sci. 1984,60, 100; H. Finkelmann, Angew. Chem. 1987,99,840; Angew. Chem. Int. Ed. Engl. 1987, 26, 816; i) P. S. Pershan, Structure of Liquid Crystal Phases, World Scientific, Singapore 1988; Phys. Today 1982, 35 (5), 34; k) H. Bader, K. Dorn, B. Hupfer, H. Ringsdorf, Adv. Polym. Sci. 1985,64,1;1) 6. Fliissigkristallkonferenz sozialistischer Landel; Halle 1985 (Ed.: H. Sackmann), Wiss. Beitr. Martin-LutherUniv. Halle 198937 (N 14); m) H. Stegemeyer, Angew. Chem., Adv. Muter. 1988, 100, 1640; n) M. Ballauf,Angew. Chem. 1989,101,260;Angew. Chem. Int. Ed. Engl. 1989,28,253;0 ) A. M. Giraud-Godquin, P. M. Maitlis, ibid. 1991,103, 370 and 1991,30,375; p) W. F. Brinkman, P. E. Cladis, Phys. Today 1982, 35 (5), 48; I. D. Litster, R. J. Birgeneau, Phys. Today 1982, 35 ( 5 ) , 26; J. D. Brock, R. J. Birgeneau, J. D. Litster, A. Aharony, Phys. Today 1989,42 (7), 52; q) M. KlBman, Rep. Prog. Phys. 1989,52,5 15; r) C. A. Angell, Science 1995, 267, 1924; s) S. Hoffmann, W. Weissflog, W. Kumpf, W. Brandt, W.
Witkowski, H. Schubert, G. Brezesinski, H.-D. Dorfler, Z. Chem. 1979, 19, 62; H.-D. Dorfler, G. Brezesinski, S . Hoffmann, Stud. Biophys. 1980, 80, 59; S. Hoffmann, Z. Chem. 1985, 25, 330; ibid. 1989, 29, 173, 449; S. Hoffmann, W. Kumpf, ibid. 1986, 26, 252, 293, 392; S. Hoffmann, W. Kumpf, W. Brandt, G. Brezesinski, W. Weissflog, ibid. 1986,26,253; S. Hoffmann, W. Brandt, H.-M. Vorbrodt, H. Zaschke, ibid. 1988, 28, 100; t) H. Schubert, S. Hoffmann, J. Hauschild, I. Marx, ibid. 1977,17,414;S. Hoffmann, H. Schubert, A. Kolbe, H. Krause, Z. Palacz, C. Neitsch, M. Herrmann, ibid. 1978, 18, 93; H. Schubert, J. Hauschild, D. Demus, S. Hoffmann, ibid. 1978, 18, 256; D. Pfeiffer, R.-G. Kretschmer, S . Hoffmann, Cryst. Res. Technol. 1986,21, 1321; u) S. Hoffmann, E. Gunther, W.-V. Meister; W. Witkowski, Wiss. Z. Univ. Halle 1988, 37/HI, 32; v) M. Descamps, C. Descamps, J. Phys. Lett. 1984,45, L-459; w) A. E. Elliot, E. J. Ambrose, Discuss. Faraday Soc. 1950, 9, 246; x) E. Sackmann, in Biophysics (Eds.: W. Hoppe, W. Lohmann, H. Markl, H. Ziegler), Springer, Berlin 1983, p. 425; y) J. W. Goodby, Mol. Cryst. Liq. Cryst. 1984, 110, 205, 221; z) H. Hoffmann, G. Ebert, Angew. Chem. 1988, 100, 933; Angew. Chem. In?. Ed. Engl. 1988,27,902; see also [5-7, 17, 18, 30, 521. [52] Mesomorphs and their order-disorder gradients: a) H. Eibl, Angew. Chem. 1984,96,247;Angew. Chem. Int. Ed. Engl. 1984,23,257;b) S. G. Kostromin, R. V. Talroze, V. P. Shibaev, N. A. Plate, Macromol. Chem. Rapid Commun. 1982,3,803; c ) W. G. Miller, Annu. Rev. Phys. Chem. 1978, 29, 519; d) S. L. Kwolek, P. W. Morgan, J. R. Schaefgen, L. W. Gulrich, Macromolecules 1977, 10, 1390; e) D. L. Patel, R. D. Gilbert, J. Polym. Sci., Polym. Phys. Ed. 1982, 20, 877; f) S . Hoffmann, N. thi Hanh, K. Mandl, G. Brezesinski, E. Gunther, Z. Chem. 1986, 26, 103; g) R. F. P. Bryan, P. Hartley, R. W. Miller, M.-S. Shen,Mol. Cryst. Liq. Cryst. 1980,62,281;h) D. W. Urry, Biochim. Biophys. Acta 1972,265,115; Proc. Natl. Acad. Sci. USA 1972,69, 1610; i) S. Hoffmann, M. G. Klotz, E. Muller, in 6th In?. Con$ on Surface Active Compounds (Ed.: F. Scheel), Akademie-Verlag, Berlin 1987, p. 545; k) C. Robinson, J. C. Ward, R. B. Beevers, Discuss. Faraday Soc. 1958, 25, 29; 1) N. Ise, T. Okubo, K. Yamamoto, H. Matsuoka, H. Kawai, T. Hashimoto, M. Fujimura, J. Chem. Phys. 1983, 78, 541; m) C. Piechocki, J. Simon, A. Skoulios, D. Guillon, P. Weber, J. Am. Chem. Soc. 1982, 104, 5245; n) I. Thondorf, 0. Lichtenberger, S. Hoffmann, 2. Chem. 1990,30,171; 0 ) W.-V. Meister, E. Birch-Hirschfeld, H. Reinert, H. Kahl, W. Witkowski, S. Hoffmann, ibid. 1990, 30, 322; p) S. Hoffmann, W. Witkowski, G. Borrmann, ibid. 1977, 17,291; S. Hoffmann,
4 R. Skolziger, W. Witkowski, ibid. 1986,26,331, 398; W.-V. Meister, R. Skolziger, G. Luck, Ch. Radtke, D. Munsche, W. Witkowski, S. Hoffmann, ibid. 1990, 30, 95; q) F. Livolant, Eur. J. Cell. Biol. 1984, 33, 300; r) R. L. Rill, Proc. Natl. Acad. Sci. USA 1986, 83, 342; s) W.-V. Meister, A.-M. Ladhoff, S. I. Kargov, D. Burckhardt, G. Luck, S. Hoffmann, Z. Chem. 1990,30, 213; t) Yu. M. Yevdokimov, V. I. Salyanov, A. T. Dembo, H. Berg, Biomed. Biochim. Acta 1983, 42, 855; u) E. Iizuka, Polym. J. 1978, 10, 235; 293; v) G. Damaschun, H. Damaschun, R. Misselwitz, D. Zirwer, J. J. Muller, I. A. Zalluskaya, V. I. Vorobev, T. L. Pyatigorskaya, Stud. Biophys. 1986, 112, 127; S . G. Skuridin, H. Damaschun, G. Damaschun, Yu. M. Yevdokimov, R. Misselwitz, ibid. 1986,112, 139; w) S. Hoffmann, W. Witkowski, A.-M. Ladhoff, Z. Chem. 1976, 16, 227; x) J. Feigon, F. W. Smith, R. F. Macaya, P. Schultze, in [33fl, p. 127; y) S . Hoffmann, Nucleic Acids Res. Symp. Ser. 1987, 18, 229; Z. Chem. 1990, 30, 94; S. Hoffmann, C. Liebmann, P. Mentz, K. Neubert, A. Barth, Win. Z. Univ. Halle 1992,41/H6,37; S. Hoffmann, M. Koban, C . Liebmann, P. Mentz, K. Neubert, A. Barth: in PCasomorphins andRelated Peptides: Recent Developments (Eds.: V. Brantl, M. Teschemacher), VCH, Weinheim 1993; z) S. Hoffmann, C. Liebmann, P. Mentz, K. Neubert, A. Barth, Z. Chem. 1989, 29, 210; see also [511. [53] Mediating water: a) E. Clementi, G. Corongin, J. Biol. Phys. 1983, 11, 33; b) M. U. Palma, in Structure and Dynamics of Nucleic Acids and Proteins (Eds.: E. Clementi, R. H. Sarma), Adenine Press, New York 1983,125;c) W. Saenger, W. N. Hunter, 0. Kennard, Nature 1986, 324, 385; d) W. Saenger, Annu. Rev. Biophys. Chem. 1987,16,93;e) N. Thantzi, J. M. Thoruton, J. M. Goodfellow, J. Mol. B i d . 1988, 202, 637; f ) E. Westlof, Annu. Rev. Biophys. Chem. 1988, 17, 125; g) W. Stoffel, Angew. Chem. 1990, 102, 987; Angew. Chern. Int. Ed. Engl. 1990, 29, 958; h) C. G. Worley, R. W. Linton, E. T. Samulski, Langmuir 1995, I I , 3805; A. F. Terzis, P. T. Snee, E. T. Samulski, Chem. Phys. Lett. 1997, 264, 481; see also [7a, 171. [54] Membranes - from bilayers and their different modeling aspects to biornembrane complexity: a) P. B. Hitchcook, R. Mason, J. Chem. Soc., Chem. Commun. 1974, 539; b ) M . Inbach, E. Shinitzky, Proc. Natl. Acad. Sci. USA 1974, 71, 4229; c) E. Shinitzky, Biomembranes 1-111, VCH, Weinheim 1994; d) H. Bader, K. Dorn, B. Hupfer, H. Ringsdorf, in Polymer Membranes, Springer, Heidelberg 1985; Adv. Polym. Sci. 1985, 64, I ; e) P. Laggner, in Topics in Current Chemistry, Vol. 145, Springer, Berlin 1988, p. 173; f) P. Laggner, M. Kriechbaum, A. Her-
[55]
[56]
[57]
[58]
References
447
metter, F. Paltauf, J. Hendrix, G. Rapp, Prog. Colloid Polym. Sci. 1989, 79,33; g) T. Kunitake, Angew. Chem. 1992, 104, 692; Angew. Chem. Int. Ed. Engl. 1992, 31, 709; h) L. F. Tietze, K. Boege, V. Vill, Chem. Ber. 1994, 127, 1065; i) D. D. Lasic, Angew. Chem. 1994, 106, 1765; Angew. Chem. Int. Ed. Engl. 1994,33, 1685; k) P. Huang, J. J. Perez, G. M. Loew, J. Biomol. Struct. Dyn. 1994, 11, 927; 1) P. Huang, E. Bertaccini, G. M. Loew, ibid. 1995, 12, 725; see also [7a, 17, 33, 51, 521. Membrane - carbohydrate systems: a) 0. Lockhoff, Angew. Chem. 1991, 103, 1639; Angew. Chem. Int. Ed. Engl. 1991, 30, 161 1 ; b) D. E. Devros, D. L. Vanoppen, X.-Y. Li, S. Libbrecht, Y. Bruynserade, P.-P. Knops-Gerrits, P. A. Jacobi, Chem. Eur. J. 1995,1, 144; see also [7a, 17, 33, 51, 521. Membrane-protein systems: a) H. R. Leuchtag, J. Theor. Biol. 1987, 127, 321, 341; b) H. Repke, C. Liebmann, Membranrezeptoren und ihre Effektorsysteme, Akademie-Verlag, Berlin 1987, c) W. Stoffel, Angew. Chem. 1990, 102, 987; Angew. Chem. Int. Ed. Engl. 1990,29,958; d) K. Eichmann, ibid. 1993, 105, 56 and 1993, 32, 54; e) T. E. Ramsdall, P. R. Andrews, E. C. Nice, FEBS Left. 1993, 333, 217; f) S.-E. Ryu, A. Truneh, R. W. Sweet, W. A. Hendrickson, Structure 1994,2,59;g) R. Smith, F. Separovic, T. J. Milne, A. Whittaker, F. M. Bennett, B. A. Cornell, A. Makriyannis, J. Mol. B i d . 1994,241, 456; h) A. Koiv, P. Mustonen, P. K. J. Kinnunen, Chem. Phys. Lipids 1994, 70, I; i) H. Chung, M. Caffrey, Nature 1994, 368, 224; k) A. P. Starling, J. M. East, A. G. Lee, Biochemistry 1995, 34, 3084; 1) Y.-P. Zhang, R. N. A. H. Lewis, R. S. Hodges, R. N. McElhaney, Biochemistry 1995, 34, 2362; m) Y. Fang, D. G. Dalgleish, Langmuir 1995, I I , 7 5 ; see also [7a, 17,33,51, 521. Membrane-nucleic acid systems: a) F. A. Manzoli, J. M. Muchmore, B. Bonara, S. Capitani, S. Bartoli, Biochem. Biophys. Acta 1974, 340, 1 ; b) H. Kitano, H. Ringsdorf, Bull. Chem. Soc. Jpn. 1985, 58, 2826; c) J.-P. Behr, Tetrahedrun Lett. 1986, 27, 5861; d) G. Gellissen, Biol. unserer Zeit 1987, 17, 15; e) K. Shirahama, K. Takashima, N. Tarisawa, Bull. Chem. Soc. Jpn. 1987, 60, 43; 0 Z. Khan, M. Ariatti, A. 0. Manutrey, Med. Sci. Rev. 1987, 15, 1189; g) G. M. Palleos, J. Michas, Liq. Cryst. 1992,1I,773; see also [7a, 17, 33, 51, 521. Liquid crystalline peptide systems between lyotropy and even thermotropy: a) E. Iizuka, Adv. Polym. Sci. 1976,20,80; b) E. Iizuka, Y. Kondo, Y. Ukai, Polym. J. 1977, 9, 135; c) H. Toriumi, Y. Kusumi, I. Uematsu, Y. Uematsu, ibid. 1979, 11, 863; d) N. Ise, T. Okubo, K. Yamamoto, H. Matsuoka, H. Kawai, T. Hashimoto, M. Fuji-
448
VIII
Living Systems
mura, J. Chem. Phys. 1983, 78, 54 1; e) J. Watanahe, Y. Fukuda, R. Gehani, T. Uetmatsu, Macromolecules 1984, 17, 1004; f) J. T. Yang, S. Kubota, in Microdomains in Polymer Solution (Ed.: P. Dubin), Plenum, New York 1985, p. 3 1 1 ; g) M. Mutter, R. Gassmann, U. Buttkus, K.-H. Altmann, Angew. Chem. 1991, 103, 1504; Angew. Chem. Int. Ed. Engl. 1991,30, 1514; h) T. Uehara, H. Hirata, H. Okahayashi, K. Taga, Y. Yoshida, M. Kojima, Colloid Polym. Sci. 1994, 272, 692; i) A. Schreckenhach, P. Wiinsche, Polymer 1994,35, 561 1; see also [7a, 17, 33, 51, 521. [59] Protein mesomorphs: a) G. H. Brown, R. J. Mishra, J. Agr. Food Chem. 1971, 19, 645; b) R. J. Hawkins, E. W. April, Adv. Liy. Cryst. 1983,6,243;c) R. E. Buxhaum, T. Dennerll, S . Weiss, S . R. Heidemann, Science 1987, 235, 15 1 1 ; d) R. Huber, Angew. Chem. 1988,100,79; Angew. Chem. lnt. Ed. Engl. 1988,27, 79; e) D. Oesterhelt, C. Brauchle, N. Hampp, Q. Rev. BiophyJ. 1991, 24,425; f) W. H. Daly, D. Poche, P. S . Russo, I. Negulescu, Polym. Prepc 1992, 33, 188; g ) J. J. Breen, G. W. Flynn, J. Phys. Chem. 1992, 96, 6825; h) D. W. Urry, Angew. Chem. 1993, 105, 859; Angew. Chem. lnt. Ed. Engl. 1993,32, 819; Sci. Am. 1995,272 (I), 44; i) C. La Rosa, D. Milardi, S. Fasone, D. Grasso, Calorim. Anal. Therm. 1993,24,233;k) A. Abe, in Macromolecules 1992, 34th IUPAC Int. Symp. on Macromolecules (Ed.: J. Kahovec), VSP, Utrecht 1993, p. 221; 1) W. H. Daly, I. I. Negulescu, R. M. Ottenbrite, E. C. Buruiana, Polym. Prepr. 1993, 34, 518; m) S. Kunugi, N. Tanaka, N. Itoh, in New Functional Materials (Ed.: T. Tsuruta), Elsevier, Amsterdam 1993, Vol. B., p. 125; n) R. Furukawa, R. Kundra, M. Fechheimer, Biochemistry 1993, 32, 12 346; 0 ) M. M. Giraud-Guille, Microsc. Res. Tech. 1994,27,420; p) Protein-protein mediators: H. Wiech, J. Buchner, R. Zimmermann, C. Jacob, Nature 1992, 358, 169; q) R. A. Stewart, D. M. Cyr, E. A. Craig, W. Neupert, Trends Biol. Sci. 1994,87;r) G. Fischer,Angew. Chem. 1994,106, 1479; Angew. Chem. Int. Ed. Engl. 1994, 33, 1479; see also [7a, 17, 33, 51, 521. [60] The difficulties around carbohydrates and their mesomorphic ambitions: a) G. A. Jeffrey, S. Bhattacharjee, Carbohydr. Res. 1983, 115, 53; G. A. Jeffrey, Acc. Chem. Res. 1986, 19, 168; J. Mol. Struct. 1994,322,21;P. M. Matias, G. A. Jeffrey, Carbohydr. Res. 1986, 153, 217; b) K. Praefcke, P. Psaras, A. Eckert, Liq. Cryst. 1993, 13, 551; K. Praefcke, D. Blunk, ibid. 1993, 14, 1 18 1; D. Blunk, K. Praefcke, G. Legler, ibid. 1994, 17, 841; R. Miethchen, H. Prade, J. Holz, K. Praefcke, D. Blunk, Chem. Ber. 1993, 126, 1707; K. Praefcke, D. Blunk, J. Hempel, Mol. Cryst. Liq. Cryst. Sci. Technol. Sect. A 1994,243, 323; c) V. Vill, H.-W. Tunger, H. Stegemeyer, K.
Diekmann, Tetrahedron 1994, 5, 2443; P. Stangier, V. Vill, S. Rohde, U. Jeschke, J. Thiem, Liq. Cryst. 1994, 17, 589; S. Fischer, H. Fischer, S. Diele, G. Pelzl, K. Jankowski, R. R. Schmidt, V. Vill, ibid. 1994, 17, 855; H. Fischer, V. Vill, C. Vogel, U. Jeschke, ibid. 1993, 15, 733; V. Vill, H.-W. Tunger, A. Borwitzky, Ferroelectrics 1996, 180, 227; P. Sakya, J. M. Seddon, V. Vill, Liq. Cryst. 1997, 23, 409; d) J. F. Stoddart, Angew. Chem. 1992, 104, 860; Angen: Chem. Int. Ed. Engl. 1992, 31, 846; e) M. Kunz, K. Ruck, ibid. 1993, 105, 355 (1993,32, 336); U. Sprengard, G. Kretzschmar, E. Bartnik, Ch. Hiils, H. Kunz, ibid. 1995, 107, 1104 (1995. 34, 990); f) G. Wenz, ibid. 1994,106,85 1 (1994,34,803); g) P. Zugenmaier, Cellul. Chem. Biochem. Mafer.Aspects 1993,105; W. V. Dahlhoff, K. Riehl, P. Zugenmaier, Liebigs Ann. Chem. 1993, 19, 1063; N. Aust, C. Derleth, P. Zugenmaier, Mucroniol. Chem. Phys. 1997, 198, 1363; h) U. Hinrichs, G. Buttner, M. Steifa, Ch. Betzel, V. Zabel, B. Pfannenziitter, W. Saenger, Science 1987, 238, 205; i) J. H. M. Willison, R. M. Abeysekera, J. Polym. Sci., Part C, Polym. Lett. 1988, 26, 71; k) C. A. A. van Boeckel, M. Petiton, Angew. Chem. 1993, 105, 1741; Angew. Chem. Int. Ed. Engl. 1993, 32, 1671; 1) C. Unverzagt, ibid. 1993,105, 1762 (1993,32, 1691); m) T. J. Martin, R. R. Schmidt, Tetrahedron Left. 1993, 34, 6733; n) H. Fischer, A. Keller, J. A. Odele, J. Appl. Polym. Sci. 1994,54, 1785; 0 ) T. Ogawa, Chem. Soc. Rev. 1994,397;p) T. P. Trouard, D. A. Mannock, G. Lindblom, L. Eilfors, M. Akiyama, R. N. McElhaney, Biophys. J. 1994,67, 1101; q) A.-C. Eliasson, Thermochim. Acta 1994, 246, 343; r) C.-H. Wong, R. L. Halcomb, Y. Ichikawa, T. Kajimoto, Angew. Chem. 1995,107,453,569;Angew.Chem. lnt. Ed. Engl. 1995,33,412,576;s) X. Auvray, C. Petipas, R. Anthore, I. Rico-Lattes, A. Lattes, Langmuir 1995,11, 433; t) D. E. Devros, D. L. Vanoppen, X.-Y. Li, S. Lihbrecht, Y. Bruynserade, P.-P. Knops-Gerrits, P. A. Jacobi, Chem. Eur. J. 1995, I, 144: see also [7a, 17, 33p, q, 51, 52el. [61] Carhohydrate-nucleic acid systems - the two sugar mates: a) T. Li, Z. Zeng, V. A. Estevez, K. U. Baldenius, K. C. Nicolaou, G. J. Joyce, J. Am. Chem. Soc. 1994,116,3709;b) K. C. Nicolaou, K. Ajito, H. Komatsu, B. M. Smith, F. Liun, M. G. Egan, H. Gomez-Paloma, Angew. Chem. 1995, 107, 614; Angew. Chem. Int. Ed. Engl. 1995, 34, 576; c) G. Pratviel, J. Bernadou, B. Mennier, ibid. 1995, 107, 819 (1995, 34, 746); see also [7a, 17, 33, 51, 521. [62] Nucleic acid liquid crystals - from phase elucidation to chromatin regulation: a) F. Livolant, Y. Bouligand, Chromosoma 1978, 68, 2 1 ; h) F. Livolant, Tissue 1984, 16, 535; c) Eur. J. Cell. Biol. 1984,33, 300; d) J. Phys. 1986,47, 1813;
4 ibid. 1986,47, 1605; ibid. 1987, 48, 1051; e) Y. Bouligand, F. Livolant, ihid. 1984.45, 1899;f) F. Livolant, M. F. Maestre, Biochemistry 1988,27, 3056: g) F. Livolant, Y. Bouligand, Mol. Cryst. Liq. Cryst. 1989, 166, 91; h) F. Livolant, A. M. Levelut, J . Doucet, J . P. Benoit, Nature 1989, 339,724; i) F. Livolant, J. M o l . B i d . 1991, 218, 165; k) D. Durand, J . Doucet, F. Livolant, J . Phys. (Paris) I1 1992, 2, 1769; 1) A. Leforestier, F. Livolant, ibid. 1992, 2. 1853; m) A. Leforestier, F. Livolant, Biophys. J. 1993, 65, 56: n) F. Livolant, A. Leforestier, D. Durant, J . Doucet, Lect. Notes Phys. 1993, 415, 33; o) A. Leforestier, F. Livolant, Liq. Cryst. 1994, 17, 651; p) J.-L. Sikorav. J . Pelta, F. Livolant, Biophys. J . 1994, 67, 1387; q) J . Pelta, J. L. Sikorav, F. Livolant, D. Durand. J . Doucet, J. Truce Microprobe Techn. 1995, 13, 401; r) J. Pelta, D. Durand, J. Doucet, F. Livolant, Biophys. J. 1996, 71,48; s) F. Livolant, A. Leforestier, Progr: Polym.Sci.1996,21, I115;seealso[7a,17,33,51, 52ql. [63] The “unspecific” textures of “evolutionary” polydisperse DNAs: a) W.-V. Meister, A,-M. Ladhoff, S . I. Kargov, G. Burckhardt, G. Luck, S . Hoffmann, Z. Chem. 1990,30,213; b) A. Leforestier, F. Livolant, “Phase Transition in a Chiral Liquid Crystalline Polymer”. Eur. Conf. Liq. Cryst. Sci. Techno]., Conf. Abstr., Films 1993, I - I Y , 24; see also [62]. [64] The beauties of nucleic acid textures - imaging phase behavior of adjusted “rod-likes”: a) R. L. Rill, P. R. Hilliand, G. C. Levy, J. Biol. Chem. 1983, 258, 250; b) R. L. Rill, Proc. Natl. Acad. Sci. USA1986,83,342:c) T. E. Strzelecka, R. L. Rill, J. Am. Chem. Soc. 1987,109.45 13; d) T. E. Strzelecka, M. W. Davidson, R. L. Rill, N(itnre 1988, 331. 457: e) R. L. Rill, F. Livolant, H. C . Aldrich, M. W. Davidson, Chromcrsomci 1989, 98, 280; f) D. M. van Winkle, M. W. Davidson, R. L. Rill, J . Chem. Phys. 1992, 97, 5641; g) K. H. Hecker. R. L. Rill, Anal. Biochern. 1997,244,67;h) D. H. Van Winkle, A. Chatterjee, R. Link, R. L. Rill, Plzys. Rev. E 1997, 55, 4354; see also [IS, 51, 52r]. 1651 Emergence of helical and suprahelical motifs from different nucleic acid composites: a) G. P. Spada, A. Carcuro, F. P. Colonna, A. Garbesi, G. Gottarelli, Liq. Cryst. 1988, 3, 651; b) G . P. Spada, P. Brigidi, G. Gottarelli, J. Chem. Soc. Chem. Comm. 1988,953;c) P. Mariani, C. Mazabard, A. Garbesi, G. P. Spada, J. Am. Chem. Soc. 1989, 111, 6369; d) S . Bonazzi, M. M. de Morais, G. Gottarelli, P. Mariani, G. P. Spada, Angew. Chem. 1993, 105, 251; Angew. Chem. Int. Ed. Engl. 1993, 105, 248; e) P. Mariani, M. M. De Morais, G. Gottarelli, G. P. Spada, H. Delacroix, L. Tondelli, Liq. Cryst. 1993, I S , 757; f) A. Garbesi, G. Gottarelli, P. Mariani, G. P.
References
449
Spada, Pure Appl. Chrm. 1993, 6.5, 641; g) F. Ciuchi, G. Di Nicola, H. Franz, G. Gottarelli, P. Mariani, M. G. Ponzi Bossi, G. P. Spada, J. Am. Chem. Soc. 1994. 116, 7064; h) A. J . MacDermott, L. D. Barron, A. Brack, T. Buhse, A. F. Drake, R. Emery, G. Gottarelli, J. M. Greenberg, R. Haberle, Planet. Space Sci. 1996, 44, 1441; i ) G. P. Spada, S . Bonzzi, A. Garbesi, S . Zanella, F. Ciuchi, P. Mariani, Liq. Cryst. 1997, 22, 341; k) F. M. H. deGroot, G. Gottarelli, S . Masiero, G. Proni, G. P. Spada, N. Dolci, Angew. Chem. 1997, 109, 990: Angew. Chon. Irit. Ed. Engl. 1997, 36, 954; I) K. H. Hecker, R. L. Rill, Annlyt. Biochem. 1997, 244, 67; m ) D. H. Van Winkle, A. Chatterjee, R. Link, R. L. Rill, Phys. Rev. E 1997,55.4354; see also [ S l , 521. [66] The detailed vision of helical liquid crystal motifs exemplified also in the case of nucleic acid systems: a) E. Iizuka, Y. Kondo. Y. Ukai, Polyrner J. 1977, 9, 135; b) E. Iizuka, ibid. 1978. 10, 293; c) ihitl. 1983. 15, 525; d) Adv. Biophys. 1988, 24, I ; e) E. Iizuka. J . T. Yang. in Liquid Crystuls and Ordered Fluids (Eds.: J . F. Johnson, R. s. Porter), Plenum, New York 1984, p. 197; see also [ S l , 5 2 ~ 1 . [67] Liquid crystal behavior of nucleic acids accompanying molecular biological aspects: a ) N . Akimenko, V. Kleinwachter, Yu. M. Yevdokimov. FEBS Lett. 1983, 156, 5 8 ; b) Yu. M. Yevdokimov, S. G. Skuridin, V. I. Salyanov. Liq. Cryst. 1988, 3, 1443; c) Yu. M. Yevdokimov. S. G. Skuridin; B. A. Chemukha, Biotekhnologya 1992, 5, 103; d) V. I. Salyanov, M. Palumbo, Yu. M. Yevdokimov, M o l . Biol. (Moscow) 1992, 26, 1036; e) V. 1. Salyanov, P. L. Lavreutier, B. A. Chernukha, Yu. M. Yevdokimov, ibid. 1994,28,1283;f) D. N. Nikogosyan, Yu. A. Repeyev, D. Yu. Yakovlev, Y. I. Salyanov, S. G. Skuridin, Yu. M. Yevdokirnov, Photochem. Photohiol. 1994. 59, 269; g) Yu. M. Yevdokirnov, V. I. Salyanov, E. Gedig, F. Spener, FERS Lett. 1996, 392, 269; h) Yu. M. Yevdokimov, V. I. Salyanov, S. V. Semenov, Biosens. Bioelectron. 1996, 11, 889; i) V. A. Belyakov, V. P. Orlov, S . V. Semenov. S . G. Skuridin, Yu. M. Yevdokimov, Liq. Cryst. 1996, 20, 777; see also 152 t]. [68] Nucleic acid liquid crystals by dynamic approaches: a) R. Brandes, D. R. Kearns, Biochemistv 1986, 25, 5890; b) R. Brandes, R. R. Vold, D. R. Klarus, A. Rupprecht. J. Mol. Biol. 1988, 202, 321; c) R. Brandes, D. R. Kearns, Biochemistry 1986, 25, 5890; d) J. Phys. Chem. 1988,92,6836;e ) D. R. Kearns, CRC Crit. Rev. 1988,15,237; t) V. L. Hsu, X. Jia, D. R. Kearns, Tbxicol. Left. 1996, 82, 577; see also [ S 1, 521. 1691 Complexity of further nucleic acid mesomorph approaches: a) G. L. Seibel, U. C. Singh, P. A . Kollman, Proc. Natl. Acad. Sci. USA 1985, 82,
450
VIII
Living Systems
6537; b) S. P. Williams, B. D. Athey, L. J. Mugliar, s. Schapp, A. H. Gough, J. P. Langmore, Biophys. J. 1986, 49, 233; c) Th. Ackermann, Angew. Chem. 1989, 101, 1005; Angew. Chem. Int. Ed. Engl. 1989, 28, 981; d) A. Fernandez, Naturwissenschaften 1989,76,469; e) Z. Reich, S. Levin-Zaidman, S. B. Gutman, T. Arad, A. Minsky, Biochemistry 1994, 33, 4177; f)Z. Reich, E. J. Wachtel, A. Minsky, Science 1994, 264, 1460; g) Z. Reich, 0. Schramm, V. Brumfeld, A. Minski, J. Am. Chem. Soc. 1996, 118, 6435; h) S . Levin-Zeitman, Z. Reich, E. J. Wachtel, A. Minski, Biochemistry 1996, 35, 2985; i) A. Kagemoto, T. Sumi, Y.Baba, Thermochim. Acta 1994, 242, 77; k) L. C. A. Groot, M. E. Kuil, J. C. Leyte, J. R. C. van der Maarel, Liq. Cryst. 1994, 17, 263; 1) A. Koiv, P. Mustonen, P. K. J. Kinnunen, Chem. Phys. Lipids 1994,70, 1; m) T. M. Bohanon, S. Denzinger, R. Fink, W. Paulus, H. Ringsdorf, M. Weck, Angew. Chem. 1995, 107, 102; Angew. Chem. Int. Ed. Engl. 1995,34,58;n) A. Chatterjee, D. H. van Winkle, Phys. Rev. E. 1994, 49, 1450; see also [7a, 17, 18, 51, 521. [70] Vorlander’s “chiral infection” modeling hormonal amplification: a) D. Vorlander, F. Jaenecke, Z. Phys. Chem. A 1913,85,697;b) S. Hoffmann, Z. Chem. 1989,29, 102; see also [7a, 17, 181. [7 11 Structure-phase problems around nucleic acidprotein interactions: a) J. H. Hogle, M. Chow, D. J. Filman, Science 1985, 229, 1358; b) D. L. Ollis, S. W. Weute, Chem. Rev. 1987, 87, 981; c) G. Dreyfuss, M. S. Swanson, S. Pinol-Roma, Trends Biochem. Sci. 1988, 13, 80; d) R. M. Evans, S. M. Hollenberg, Cell 1988, 52, 1; e) J. P. Leonetti, B. Rayner, M. Lemaitre, C. Gagnor, P. G. Milhand, J.-L. Imbach, B. Lebleu, Gene 1988, 72, 323; f ) T. J. Gibson, J. P. M. Postma, R. S. Brown, P. Argos, Protein Engl. 1988,2,209;g) R. G. Brennan, B. W. Matthews, J. Biol. Chem. 1989,264,1903;h) R. Bandziulis, M. S. Swanson, G. Dreyfuss, Genes Devel. 1989, 3,43 1 ; i) R. D. Klausner, J. B. Harford, Science 1989, 246, 870; k) T. A. Steitz, Q. Rev. Biophys. 1990, 23, 205; D. Wassaman, J. A. Steitz, Nature 1991, 349, 463; 1) M. Ptashne, A. A. F. Gann, ibid. 1990, 346, 329; m) D. Latchman, Gene Regulation, A Eukuryotic Perspective, Hyman, London 1990; n) N. P. Pavletich, C. 0. Pabo, Science 1991, 252, 809; 0 ) S. W. Ruby, J. Abelson, Trends Genet. 1991, 7, 79; p) C. Cuthrie, Science 1991,253,157;q) D. Kostrewa, J. Granzin, D. Stock, H. W. Choe, J. Lahban, W. I . Mol. Biol. 1992, 26, 209; W. Saenger, . Saenger,C. Sandmann, K.Theis,E. B. Starikow, G. Kostrewa, J. Laban, J. Granzin, in Nucleic Acids and Molecular Biology (Eds.: F. Eckstein, D. M. J. Lilly), Springer, Berlin 1993, 158; R. Giege, J. Drenth, A. Ducruix, A. McPherson,
W. Saenger, Progr. Cryst. Growth Charact. Muter. 1995, 30, 237; C. Sandmann, F. Cordes, W. Saenger, Proteins: Struct., Funct., Genet. 1996, 25, 486; W. Saenger, D. Kostrewa, J. Granzin, F. Cordes, C. Sandmann, C. Kisker, W. Hinrichs, in From Simplicity to Complexity in Chemistry - and Beyond (Eds.: A. Miiller, A. Dress, F. Vogtle), Vieweg, Braunschweig 1996, p. 51; r) D. G. Gorenstein, Chem. Rev. 1994,94, 1315; s) M. Famulok, D. Faulhammer, Angew. Chem. 1994,106, 191 I; Angew. Chem. lnt. Ed. Engl. 1994, 33, 1827; see also [7a, 17, 18, 33, 51, 521. [72] Selected problems around structure and phase general questions: a) G. SolladiC, R. G. Zimmermann, Angew. Chem. 1984,96,335;Angew. Chem. Int. Ed. Engl. 1984, 23, 348; b) S. Andersson, S. Lidin, M. Jacob, 0. Terasaki, ibid. 1991, 103, 771 (1991, 30, 754); c) A. Amann, W. Gans, ibid. 1989, 101, 277 (1989, 28, 268); d) H. Frauenfelder, Ann. NY Acad. Sci. 1986, 504, 151; e) Disordered Systems and Biological Models (Ed.: L. Peliti), CIF Series, World Scientific, Singapore 1989, 14; f) J. M. Ottino, F.4. Muzzio, M. Tjahjadi, J. G. Framjione, S. C. Jana, H. A. Kusch, Science 1992,257,754;g) D. Seebach,Angew. Chem. 1990,102,1362;Angew. Chem. Int. Ed. Engl. 1990, 29, 1320; h) F. M. Stillinger, Science 1995, 267, 1935; i) B. Frick, D. Richter, ibid. 1995,267,1939;k) R. Jaenicke, Angew. Chem. 1984,96,385;Angew. Chem. Int. Ed. Engl. 1984, 23, 395; 1) U. Sprengard, G. Kretzschmar, E. Bartnik, Ch. Huls, H. Kunz, ibid. 1995, 107, 1104 (1995, 34, 990); m) S. Friberg, Naturwissenschuften 1977, 64, 612; n) C. N. Paul, U. Heinemann, U. Mahn, W. Saenger, Angew. Chem. 1991, 103, 351; Angew. Chem. Int. Ed. Engl. 1991,30,343; 0 ) C. A. Angell, Science 1995,267, 1924; p) M. Quack, Angew. Chem. 1989, 101, 578; Angew. Chem. Int. Ed. Engl. 1989, 28, 959; q) A. Fernandez, Naturwissenschaften 1989, 76,469;r) K. Shinoda, Pure Appl. Chem. 1988, 60, 1493; s ) Nucleic acid frequence correlations and coherences?: C. A. Chatzidimitriou-Dreismann, Adv. Chem. Phys. 1991,80,201; Helv. Chem. Acta 1992,75, 2252; Nature 1993,361,212;Adv. Chem. Phys. 1997, 99, 393; D. Lavhammar, C. A. Chatzidimitriou-Dreismann, Nucleic Acids Res. 1993, 21, 5167; E. J. Brandas, C. A. ChatzidimitriouDreismann, Int. J. Quant. Chem. 1995, 53, 95; C. A. Chatzidimitriou-Dreismann, R. M. F. Streffer, D. Larhammar, Nucleic Acids Res. 1996,24, 1676; t) J. Maddox, Nature 1992,358, 103; u)O. Kennard, W.-N. Hunter, Angew. Chem. 1991, 103, 1280; Angew. Chem. Int. Ed. Engl. 1991,30,1254;v) R. E. Dickerson, J. Mol. Biol. 1989, 205, 787; w) E.-L. Winnacker, Angew. Chem. 1992,104, 1616;Angew. Chem. Int.
4 Ed. Engl. 1992, 31, 1578; x) Fractal motifs: H. Jurgens, H.-0. Peitgen, D. Saupe, Chaos und Fractale, Spektrum, Heidelberg 1989; y) A. S . Borovik, A. Y. Grosberg, M. D. Frank-Kamenetski, J. Biomol. Struct. Dyn. 1994, 12, 656; z) A. Blumen, M. Schnorer,Angew. Chem. 1990, 102, 158;Angew. Chem. Int. Ed. Engl. 1990,29, 126; z’) V. Percec, J. Heck, G. Johansson, D. Tomazos, M. Kawasumi, P. Chu, G. Ungar, Mol. Cryst. Liq. Cryst. Technol. Sect. A 1994, 254, 137; see also [7a, 17, 18, 19, 25, 33, 51, 521. [73] Layers and adlayers approaching nanometerscale-lipid-protein-systems: a) H. Ebato, J. N. Harron, W. Muller, Y. Okahata, H. Ringsdorf, P. Suci, Angew. Chem. 1992, 104, 1064; Angew. Chem. Int. Ed. Engl. 1992,31, 1078; b) W. Muller, H. Ringsdorf, E. Rump, X. Zhang, L. Augermaier, W. Knoll, M. Lilly, J. Spinke, Science 1993, 262, 1706; c) D. Vaknin, K. Kjaer, H. Ringsdorf, R. Blankenburg, M. Piepenstock, A. Diederich, M. Losche, Langmuir 1993,9, 1 17 1 ; d) M. Losche, C. Erdelen, E. Rump, H. Ringsdorf, K. Kjaer, D. Vaknin, Thin Solid Films 1994, 242, 112; e) R. Naumann, A. Jonczyk, C. Hampel, H. Ringsdorf, W. Knoll, N. Bunjes, P. Graeber, Bioelectrochem. Bioenerg. 1997,42, 241 ; Lipid-nucleic acid-systems: f) R. C. Ebersole, J. A. Miller, J. R. Moran, M. D. Ward, J . Am. Chem. Soc. 1990, 112, 3239; g) Y. Okahata, Y. Matsunobu; K. Ijiro, M. Mukae, A. Murakami, K. Makino, J. Am. Chem. Soc. 1992,114,8300; h) Y. Okahata, K. Ijiro, Y. Matsuzaki. Langmuir 1993, 9, 19; i) G. Decher, Nachr. Chem. Tech. Lab. 1993, 41, 793; k) A. U. Niu, L. Shain, W. Ling, J. Werenka, D. Larson, E. Henderson, SO Proc. SPIE lnt. Soc. Opt. Eng., Proc. Absorb. DNA Sequencing Technol. 1993,72; I) S. Yamaguchi, T. Shinomura, T. Tatsuma, N. Oyama, Anal. Chem. 1993,65, 1925; - General aspects: m) B. M. Robinson, N. C. Seeman, Protein Eng. 1987, I , 295; n ) N . C. Seeman, Y. Zhang, J. Chen, J. Vac.Sci. Technol. A 1994,12,1895;0 ) J. Economy, Angew. Chem. 1990, 102, 1301; Angew. Chem. Int. Ed. Engl. 1990, 29, 1256; p) F. M. Menger, ibid. 1991, 103, 1104 (1991, 30, 1086); q) K. Camman, U. Lemke, A. Rohrer, J. Sander, H. Wilken, B. Winter, ibid. 1991,103, 519 (1991, 30, 516); r) J. M. Wendorff, ibid. 1991, 103, 416 (1991, 30, 405); s) D. D. Lasic, ibid. 1994,108, 1765 (1994,33, 1685); t) J. Majewski, R. Edgar, R. Popovitz-Biro, K. Kjaer, W. G. Bouwman, J. Als-Noelsen, M. Lahav, L. Leiserowitz, ibid. 1995, 107, 707 (1995, 34, 649); see also [7a, 17, 18, 51, 521. [74] Molecular resolving microscopies: a) E. Spiess, R. Lurz, Methods Microbiol. 1988, 20, 293; b) J . Frommer, Angew. Chem. 1992, 104, 1325; Angew. Chem. lnt. Ed. Engl. 1992, 31, 1298; c) D. Jacquemain, S . G. Wolf, F. Leveiller, M.
References
45 1
Deutsch, K. Khaer, J. Als-Nielson, M. Lahav, L. Leiserowitz, ibid. 1992, 104, 134 (1992, 31, 130); d) Scanning Tunneling Microscopic and Related Methods (Ed.: R. J. Brehns), Kluwer, Dordrecht 1993, e) Scanning Tunneling Microscopy (Eds.: R. Wiesendanger, H.-J. Guntherodt), Springer, Berlin 1993; f) E. Delain, E. Lecane, A. Barbin-Asbogant, A. Fourcade, Microsc. Microanal. Microstruct. 1994,5,329;g ) T. Basche, Angew. Chem. 1994, 106, 1805; Angew. Chem. Int. Ed. Engl. 1994,33, 1723; h) H. Rohrer, Microsc. Microanal. Microstruct. 1994, 5, 237; i) E. F. Aust, M. Savodny, S. Ito, W. Knoll, Scanning 1994, 16, 353; approaches to self-organizing systems: k) C. D. Frisbie, L. F. Rozsnyai, A. Noy, M. S. Wrighton, C. M. Lieber, Science 1964, 264, 2071; 1) L. HauBling, B. Michel, H. Ringsdorf, H. Rohrer, Angew. Chem. 1991,103, 568; Angew. Chem. 1991, 103, 568; Angew. Chem. Int. Ed. Engl. 1991, 30, 569; m) J. J. Breen, G. W. Flynn, J. Phys. Chem. 1992, 96, 682.5; n) D. P. E. Smith, A. Bryant, C. F. Quate, J. P. Rabe, C. Gerber, J. D. Swallen, Proc. Nafl. Acad. Sci. USA 1987, 84, 969; 0 ) S. Buchholz, J. P. Rabe,Angew. Chem. 1992,104,88;Angew. Chem. lnt. Ed. Engl. 1992,31, 189; p) S. Lincotti, J. P. Rabe, Supramol. Sci. 1994, I , 7; q) H. G. Hansma, M. Bezanilla, F. Zerchansern, M. Adrian, R. L. Sinsheimer, Nucleic Acids Res. 1993,21,505;r) T. E. Ramsdale, P. R. Andrews, E. C. Nice, FEBSLett. 1993,333,217; s) B. Samori, c. Nigro, V. Armentanof, S. Cimieri, G. Zuccheri, C. Quagliariello, Angew. Chem. 1993, 105, 1482; Angew. Chem. Int. Ed. Engl. 1993, 32, 1461;t) T. Thundat, R. J. Warmack, D. P. Allison, K. B. Jacobson, Scanning Microsc. 1994, 8, 23; u) P. Pasero, C. Blettry, M. Marilley, B. Jordan, S. Granjeaud, M. Dayez, A. Humbert, J. Vac. Sci. Technol. B 1994, 12, 1521; v) G. Zuccheri, G. A. Ranieri, C. Nigro, B. Samori, J. Vac. Sci. Technol. B 1995, 13, 158; see also [S-7, 17, 18, 33, 51, 521. [75] Molecular resolving microscopies - ways to self-organizing nucleoprotein systems: a) C. Bohley, W.-V. Meister, S. Lindau, F. Klimaszewsky, J. Barthel, S . Hoffmann, unpublished; b) E. Birch-Hirschfeld, J. van Esch, S. Hoffmann, I. Kuniharu, H. Ringsdorf, W. Saenger, C. Sandmann, U. Schilken, M. Strube, I. Thondorf, unpublished; c) R. Heinz, J. P. Rabe, W. V. Meister, S. Hoffmann, Thin Solid Films 1995, 264, 246; d) C. Bohley, D. Matern, W.-V. Meister, S. Kargov, S. Lindau, J. Barthel, S. Hoffmann, Surface Interface Analysis 1997, 25, 614; e) C. Bohley, T. Martini, G. Bischoff, S. Lindau, E. Birch-Hirschfeld, S. Kargov, W.-V. Meister, J. Barthel, S . Hoffmann, Nucleosides Nucleotides 1997,16 (10); f ) C. M. Niemeyer, Angew. Chem. 1997, 109, 603; Angew. Chem. Int. Ed. Engl.
452
VIII
Living Systems
1997, 36, 585; g) S. C. Tsang, Z. Guo, Y. K. Chen,M. L. H. Green, H. A. 0. Hil1,T. N. Hambley, P. J. Sadler, ibid. 1997, 109, 2292 (1997, 36, 2198); see also [7a, 17, 18, 51, 521. [76] Leading theories in the constraints of reductionistic approaches: a) L. Onsager, Ann. N. I:Acad. Sci. 1949, 51, 622; b) P. J. Flory, Proc. R. Soc. London Sex A 1956, 234, 73; Adv. Polym. Sci. 1963, 61, 1; J. Polym. Sci. 1961,49, 104; c) W. Maier, A. Saupe, Z. Naturjorsch. A 1959, 14, 882; ibid. 1960,15,287; d) V. I. Ginzburg, Phys. Solid State 1960,2, 184; e) L. D. Landau, E. M. Lifshitz, Statistical Physics, Pergamon, New York 1980; f) I. W. Stewart, T. Carlsson, F. M. Leslie, Ferroelectrics 1993, 148, 41; g) F. M. Leslie, Liq. Cryst. 1993, 14, 121; h) B. Samori, S. Masiero, G. R. Luckhurst, S. K. Heeks, B. A. Timimi, P. Mariani, Liq. Cryst. 1993, 15, 217; i) G. R. Luckhurst, Ber: Bunsen-Ges. Phys. Chem. 1993, 97, 1169; G. R. Luckhurst, Mol. Phys. 1994,82, 1063;k) S. M. Fan, G. R. Luckhurst, s. J. Picken, J. Chem. Phys. 1994, 101, 3255; 1) A. P. J. Emerson,G. R. Luckhurst, S. G. Whatling, Mol. Phys. 1994,82,113;see also [ 17, 19, 25, 37, 51, 521. [77] Stabilization within the dynamics: a) S. L. Kwolek, P. W. Morgan, J. R. Schaefgen, L. W. Gulrich, Macromolecules 1977, 10, 1390; b) J. A. McCammon, C. Y. Lee, S. H. Northrup, J. Am. Chem. Soc. 1983,105,2232;c) M. Karplus, J. A. McCammon, Sci. Am. 1986, 254, 42; d) J. A. McCammon, S. C. Harvey, Dynamics of Proteins and Nucleic Acids, Cambridge University Press, New York 1987;e) J. A. Trainer, E. D. Getzoff, H. Alexander, R. A. Houghten, A. J. 01son, R. A. Lerner, Nature 1984, 312, 127;
J. Filman, Science 1985,229, 1358; g) R. A. Shalaby, M. A. Lauffer, Arch. Biochem. Biophys. 1985, 236, 300; h) Z. Blum, S. Lidin, Acta Chem. Scand. 1988, 342, 417; i) W. F. van Gunsteren, M. J. C. Berendsen, Angew. Chem. 1990,102,1020;Angew. Chem. Int. Ed. Engl. 1990,29,992; k) H. Oschkinat, T. Muller, T. Dieckmann, ibid. 1994,106, 284 (1994, 33, 277); 1) I. Radnakrishnan, D. J. Patel, Structure 1994, 2, 17; m) P. Huang, J. J. Perez, G. M. Loew, J. Biomol. Struct. Dyn. 1994, 12,927; n) P. Huang, E. Bertaccini, G. H. Loew, ibid. 1995, 12, 725; see also [l-8, 17, 18, 51-76]. [78] Structure-phase up to morphogenesis: a) B. Hess, M. Markus, Trends Biochem. Sci. 1987,12, 45; b) C. Nusslein-Volhard, Angew. Chem. 1996, 108, 2316; Angew. Chem. Int. Ed. Engl. 1996, 35,2176; c) E. Wieschaus, ibid. 1996,108,2330 (1996,35,2118);see also [l-8, 17, 18,51,52]. [79] Human brain between deterministics and probabilistics: a) Gehirn und Bewujtsein (Ed.: W. Singer), Spektrum, Heidelberg 1994; b) J. Travis, Science 1992,258,216;modeling by neuronal nets: c) D. Haarer, Angew. Chem. 1989,101, 1576; Angew. Chem. Int. Ed. Engl. 1989, 28, 1544;d) G. Schneider, P. Wrede, ibid. 1993,105, 1192 (1993,32, 1141);e) J. Gasteiger, J. Zupan, ibid. 1993, 105, 510 (1993, 32, 503); see also [l-8, 17, 181. [80] Human brain- the quantum view: a) K. J. Schulten, in Signalwandlung und Informationsverarbeitung (Ed.: W. Kohler), Nova Acta Leopoldinu 1995, NF 294/72, 9; b) A. Ekert, Quantum processings of information, ibid. 1995, 294/72, 14; see also [ I - & 17, 181. f) J. M. Magle, M. Chow, D.
Handbook ofLiquid Crystals D. Demus, J. Goodby, G. W. Gray, H.-W. Spiess, V. Vil Copyright 0WILEY-VCH Verlag GmbH. 1998
Chapter IX Cellulosic Liquid Crystals Peter Zugenmaier
1 Introduction Cellulose represents an important polymer, which is most abundant in nature, and serves as a renewable resource in many applications, e.g., fibers, films, paper, and as acomposite with other polysaccharides and lignin in wood. Cellulose derivatives will also be used as films and fibers, food additives, thermoplastics, and construction materials, to name just a few. Cellulose and cellulose derivatives have played an important role in the development of the macromolecular concept. So far, little use has been made of the fact that cellulose represents a chiral material except, e.g., in a rare case as stationary material in liquid chromatography for the separation of chiral compounds. Nature ifself uses the chirality of cellulose occasionally, and twisted structures of cellulose molecules are found in cell walls. Cellulose and some derivatives form liquid crystals (LC) and represent excellent materials for basic studies of this subject. A variety of different structures are formed, thermotropic and lyotropic LC phases, which exhibit some unusual behavior. Since chirality expresses itself on the configuration level of molecules as well as on the conformation level of helical structures of chain molecules, both elements will influence the twisting of the self-assembled supermolecular helicoidal structure formed in a mesophase. These supermolecular structures of chiral materials exhibit special optical properties as iridescent colors, and
may be used for special coloring effects such as crustaceans provide with the help of chitin.
2 The Structure of Cellulosics The structure and arrangement of cellulosic chains play an important role in the formation of liquid crystals. At present, neither the conformation of cellulosics nor the solvent bound to the chain in the case of a lyotropic mesophase are known for these liquid-crystalline systems. Nevertheless, these structural features form the basis for a discussion of structural and thermodynamic aspects. Information on cellulosics is available for the two borderline cases next to the LC state, i.e., for the solvent built-in solid state as well as for the pure solid state, obtained by X-ray, NMR, and potential energy analysis on one side, and for the semidilute state from light-scattering experiments on the other side. These data have to be evaluated for a discussion of possible structures and models in liquid-crystalline phases. Cellulose consists of P-D-anhydroglucosepyranose units linked by ( 1 -4)-glucosidic bonds in contrast to amylose in which the units are linked by a-(1-4) glucosidic bonds (Fig. 1). Cellulose is considered to form somewhat stiff chain molecules, whereas
454
IX Cellulosic Liquid Crystals
H
dHt),
Figure 1. Schematic representation of the configuration of (a) cellulose and (b) amylose with conventional atom labeling and the introduction of the torsion angles $ and y.
amylose has flexible ones. The solubility is quite different for these two polysaccharides. The monomer units are relatively stiff for cellulose, but more flexible for amylose, with some allowed rotations of the hydroxyl group at C6 leading to three distinct energy minima for both polysaccharides. The two dihedral or torsion angles @ and Y, involving the linkage of two adjacent residues, determine the chain conformation, which, especially for cellulosics, is quite limited by steric interactions between two adjacent residues and only allows extended chains as low energy systems. Nevertheless, sheet-like structures such as 2/1 helices or rod-like structures such as left-handed 3/2 or 8/5 helices are possible and found in cellulosics with a uniform distance of about 5 8, (0.5 nm) for one monomeric unit projected on the helix axis. The three hydroxyl groups of a monomeric unit form a dense hydrogen-bonded network in solid cellulose and provide sites for substitution by ester, ether, or carbamate
groups. Cellulose commonly exhibits a sheet-like structure such as a 2/1 helix in the solid state, and it was not until recently that the packing of natural cellulose was discovered to consist of two polymorphs of different ratio for the various kinds of natural cellulose. Strong interactions between cellulose and small molecules or ions, e.g., sodium hydroxide, may lead to a change in the cellulose chain conformation, and rod-like 3/2 helices in a solvent built-in crystalline state may be formed. Cellulose derivatives pack as sheet-like structures, for example, cellulose triacetate or butyrate as well as some ethers, e.g., trimethyl cellulose. Regio-selective derivatives of these mentioned esters or ethers placed at C2 and C3 with various other groups at C6 may belong to the same conformation class, that is, their conformation is determined by the substituent at positions 2 and 3. Many cellulose derivatives crystallize as rod-like helices of various kinds, e.g., 3/2,8/5, and these structures are also found in solvent built-in crystals. It should be mentioned that several cases are known where a change in conformation occurs if a polar solvent is built into the interstitial spaces between the polysaccharide chains. The sheet-like cellulose triacetate conformation converts to a rod-like 8/5 helix with nitromethane present. Sophisticated experimental methods allow the development of models for polymers in dilute and semi-dilute solutions. Chain stiffness may be represented by the Kuhn segment lengths and determined in dilute solution. Models for cellulose and cellulose derivatives have recently been published whose main features are the irreversible aggregation of chains, if hydrogen bonding is possible even in dilute solutions. Trisubstituted cellulose derivatives or cellulose in hydrogen bond breaking solvents exist as molecular dispersed chains. How-
3 The Chiral Nematic State
ever, at higher concentrations, in the semidilute regime, reversible cluster or association occur and, depending on the geometry of these particles, liquid crystals or gels are obtained. The association process is strongly influenced by the molecular mass, and above a certain length of the cellulosic chain, only gel formation is observed. Aggregation and association processes are comparable with micelles for amphiphilic molecules and lead to lyotropic liquid crystals upon increasing their concentration. Such a process of creating larger entities first, which then form lyotropic liquid crystals, has to be considered for cellulosics. The models on which theories on thermodynamics, flow behavior, and the structural features of a cholesteric phase depend have to take these novel ideas into account. In this respect, the term ‘lyotropic’ describes the feature of creating liquid crystals by forming clusters or other arrangements first, and in a further step overall order is introduced. ‘Thermotropic’ then means obtaining or changing the order through temperature.
3 The Chiral Nematic State 3.1 Methods for the Detection of Cellulosic Mesophases Lyotropic cellulosics mostly exhibit chiral nematic phases, although columnar phases have also been observed. The molecules in the thermotropic state also form chiral nematic order, but it is sometimes possible to align them in such a way that a helicoidal structure of a chiral nematic is excluded. Upon relaxation they show banded textures. Overviews on lyotropic LC cellulosics are
455
provided by Gray and co-workers [ 1,2], on thermotropics by Fukuda et al. [3], on polymer solvent interactions by Zugenmaier [4], and on LC cellulosics by Gilbert [ 5 ] . A first characterization of LC materials normally starts with the observation of textures in the polarizing microscope (see Fig. 2), followed by additional investigations evaluating other means of identifying the kind of phase present. Chiral nematics, also termed cholesteric phases, show unusual optical properties which are caused by a supermolecular structure by which the phase may be determined on applying spectroscopic means. These optical properties include a wave-guiding effect, selective reflection, optical rotation, and birefringence. Studies on the phase transition and the flow behavior add to our knowledge of chiral nematic phases. A survey on the structure and properties of cholesteric phases is given by Dierking and Stegemeyer [8], and theoretical models are discussed. The wave-guiding regime occurs for a pitch P s iland was investigated by Maugin [lo]. This effect plays an important role in display technology. The plane of polarization of light follows the twist of the nematic director defined within a nematic sheet. The wave-guiding regime has yet to be considered in the field of cellulosics. At the selective reflection wavelength, a component of linearly polarized light, e.g., right-handed or left-handed circularly polarized light, is totally reflected without a phase reversal by a twisted helicoidal structure at a wavelength correlated to the pitch P . An oblique incident of light is presumed onto the substrate plane or nematic sheet (see Fig. 3)
where il,is the selective reflection wavelength, E =(rzo,nem+ne,nem)/2, the mean re-
456
IX
Cellulosic Liquid Crystals
3
The Chiral Nematic State
457
Figure 2. Characteristic textures of cellulosics observed in the polarizing microscope at room temperature, if not otherwise noted (abbreviations: CTC cellulosetricarbanilate, cf. Table 4 for solvents): (a) Oily streaks of a planar oriented lyotropic LC solution of CTClMAA (c=0.7 glml, 1 unit of scale 17 pn, adapted from [ 6 ] ) . (b) Parabolic focal conic texture of CTClMAA (c=0.85 glml) found after annealing (1 unit of scale: 5.7 pm, adapted from [6]).(c) Planar and parabolic focal conic (at 323 K) textures of CTClDEMM at increasing temperatures (c=0.8 g/ml, DP= 125, adapted from [7]). (d) Planar textures of CTClDEME dependent on the molecular mass (narrow fractions, denoted by the degree of polymerization DP; c=0.8 glml, adapted from [S]). (e) Schlieren texture of a nematic-like LC of CTC/EMAc+EDAc (1: 1) (DP= 125, c=O.7 glml, adapted from [7]). (f) Planar and schlieren texture of LC CTC caused by a change of solvent from EMMAc to EMEAc to EMBAc, exhibiting in sequel a left-handed, nematic-like, right-handed structure (DP= SO, c = 0 . 8 glml, adapted from [S]). (8) Biphasic region of LC CTCIDEME; red areas remaining planar texture, black areas isotropic phase. Such polarization microscopic textures may serve for establishing phase diagrams (c = 0.65 g/ml, DP= 125, adapted from [S]). (h) Conoscopic pictures of CTClEME (a) without and (b) with , Iplate to establish the positive birefringence of CTClEME (adapted from [S]). (i) Fingerprint texture of CTClEMEAc (DP= 125, 1 unit of scale: 3 pm, adapted from [S]. (k) Cano rings of LC CTC/TRIME (c=0.8 glml, DP= 125, 1 unit of scale: 17 pm, adapted from [8]).
Figure 3. Schematic representation of the cholesteric liquid-crystalline structure of cellulosics; P = Lohi where P represents the pitch, 13, the reflection wavelength, and E the mean refractive index of a sheet. P>O for a right-handed twist, and P
