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!
!
l
Ii
.~ >1';'" I ,~~i__ li~:;~~;,:
instruments (S and X). The deviations across the entire range ofvarying conditions are systematically very small. The same ap-
plies correspondingly for the BPTs with electrical signal acquisition, where the difal 45 65 ference is, however, somewhat larger. This 60 />+' . r; 0.9942 is also true for BPTs with manual readout of 55 . , ' , : the same type, but with large differences (up 55 60 65 70 75 60 65 90 95 to 9K) for BPTs ofdiffering design (data not BPT el. A In °C shown). In view of such BPT temperature difFigure 13. Three BPT under varied conditions of exposure, ferences and the indicated differences between BSTs and BPTs, it is apparent that, for these differing measurement systems, "the same indicated temperature" cannot be equated with "identical surface temperature" of exposed (black) samples, which is one of the essential prerequisites for reproducible adjustment of conditions in devices. Gi
t
70
/,', ,,.'1 ! " " . i ,~' ,.~' I
..
Q
SURFACE TEMPERATURES OF EXPOSED MATERIAL SYSTEMS While the adjustment of identical black standard temperatures is a prerequisite for the creation ofadequately similar surface temperatures on exposed samples under natural conditions as well as in various simulation devices, such an adjustment alone is not adequate. For this,
116
Weathering of Plastics
00
.•.
~ ~ ' ". ".
'" \.
--- .. --- --- ..f - - - - f - - - j - - - - j
CT
30 i
o
'" 60 lrrad.,nce (300"'00 om) in W/m'
eo
""
ST - surface temperature of different coloured samples
., I
""; tW< "" I
I
;'!lui
I
(Ia:l
!
~
:~-r~i
•
i
.. ~
VI
,
4
v
immer sion at S·C
y
lmmersion at 30·e
-4 0
50
100
time
150
200
250
300
(hours)
Figure 6. Development of residual stress in an AA coating 252 Jlm thick during wet (5°C)/dry (room ten1perature) cycling (96 h period).
Results similar to those for AB are shown in Figure 5 for a bi-Iayer of 256 Jlm AB over 179 Jlm AA. The main difference between Figures 4 and 5 is the stress scale, which is expanded for Figure 5. Note that for the bi-Iayers it is assumed that the application ofthe AB top coat does not change the characteristics of the AA and that the change in curvature of the AB+AA+substrate combination is caused by stress changes in the top coat only. Changes in stress in AA caused by absorption of solvent from AB are ignored. Detailed differences occurred in the stresses observed for different coating thicknesses and, for bi-Iayers, different combinations of coating thicknesses. 2 Tensile stresses of nearly 2 MPa were observed during the drying out phase of the second and third cycle of an AB coating 293 f.lm thick. In bi-Iayers the stress after several cycles depended on the relative thickness ofthe two components and could be either tensile (generally when AA thickness was greater) or compressive (generally when AB thickness was greater).2
TEMPERATURE CYCLING Cooling samples to 5°C produced large tensile stresses which relaxed significantly during the cold dwell (Figures 6-8). In AB the stress reversed on returning to 30°C and the stress changes were repeated each temperature cycle (Figure 7). In the AA coating there was a progressive drift to higher (tensile) stresses (Figure 6). Bi-Iayers showed behavior closer to AB than to AA (Figure 8).
157
Residual Stress Development
AS
v
immersion at S·C
,.
immersion at
J
bi-Iayer
30·C
AS on AA
2 ~
c..
~
'1 1
$I)~
... 4)
'-
r"IUOWfW
-2
0
"
\
~
en
T
-1 T
-4
~........L......~......l-.l.~~~~~~~
t....1-::......;.iI..w...J........ "(...............
o
50
100
time
150
200
250
300
(hours)
Figure 7. Development of residual stress in an AB coating 235 J1m thick during \vet (5°C)/dry (roon1 temperature) cycling (96 h period).
v
immersion at S*C
,.
immersion at 30·C
-2 0
50
100
time
150
200
250
300
(hours)
Figure 8. Development of residual stress in bi-layer coating consisting of264 J1m of AB on top of 142 J1m of AA during wet (5°C)/dry (room temperature) cycling (96 h period).
DISCUSSION The residual stress development in AA coatings was similar to that observed by Cro1l 3A who also found that the residual stress in thermoplastic coatings reached an equilibrium value that was independent of the thickness and that the thickest coatings took the longest time to reach equilibrium. Tensile stresses form as the result of the volumetric shrinkage that accompanies the loss of solvent. During the early part of this process the coating is still fluid and stresses begin to form only when sufficient solvent has been lost for the coating to develop some energy elastic resistance to deformation. The time dependence of stress build up is determined by the diffusion of solvent through the coating and by the relaxation processes in the coating. The concentration profile will be dependent on the coating thickness and the relaxation rate will depend on the concentration. It is thus curious that the final stress level should be independent of coating thickness. The residual stresses in AB thermoset coatings were also tensile but showed greater scatter in magnitude and did not always approach a steady value even after 22 days. Crol15 also investigated thermoset coatings but used a solventless amine-cured epoxy. In his studies the coatings developed compressive stresses when thin «55 Jlm, thinner than any of the coatings investigated in the current work) and tensile stresses when in the range of thicknesses used here. Croll could not use solvent evaporation to explain stress development and he attributed the tensile stress to structural changes during the curing process. He SUlTIlised that compres-
158
Weathering of Plastics
sive stresses were caused by swelling due to water absorption (from the atmosphere). No attempt was made to control the humidity in the experiments repolted here and the small lack of consistency between different runs with AB coatings may have been caused by different contributions from this source. In the case of thermoset coatings the diffusion of solvent becomes progressively more difficult as the polymer network develops and release of solvent may proceed for an extended period of time. When AB was overcoated on top of a dry AA coating, solvent release from AB was not only into the air at the free surface but also into AA at the interface between the two coatings. Solvent entering AA will cause swelling giving an increment ofcompressive stress so that the overall build up of stress was much slower than for a similar AB coating applied direct to the substrate and the increment of stress due to the AB coating was much less than that obtained with an AB coating alone. 2 The behavior of the coatings when immersed in water and on subsequent drying out requires careful consideration. The initial tensile stress observed in AA coatings has not been explained with certainty. It is speculated that water may plasticize the coating, assisting the escape of residual solvent (or some other minor component). Subsequent changes in stress on dry/wet cycling are small but the sense of the changes are opposite to those which would be caused by water swelling during immersion and reversal of this effect during drying out. It is as if water has occupied the free volume and provided attractive forces to draw the molecules closer together. After water immersion the measured Young's n10dulus ofAA was higher than after solvent evaporation and it increased still further if allowed to dry out partially. This could be explained if water acted both to plasticize the polymer and to provide stronger intermolecular bonds and if the water participating in plasticization was less tightly bound (and more easily lost on drying out) than that providing intermolecular bonding. An initial increment of tensile stress was also observed in AB coatings on water immersion, possibly caused by a similar mechanism to that in AA. After about half an hour this effect reversed and subsequently for all phases of the wet/dry cycling the changes in stress were consistent with swelling by water (giving compression) with reversal during desorption of water. The overall drift in stress in the tensile direction could be due to further solvent evaporation (assisted by water plasticization of the coating). Broadly similar results were obtained by Negele and Funke 6 using a simpler epoxy coating. Of perhaps greatest interest here are the results obtained with AB coatings on top of AA coatings. The results are explainable qualitatively in terms of water diffusing through the AB coating and on into the AA coating during immersion and then this process reversing during drying out. The concentration gradients will be complex and will cause significant inertia in the time signature of the changes. As a result of the different stress responses of AA and AB coatings to water the sense ofstress in the bi-Iayer coatings depended on the relative thickness
Residual Stress Development
159
ofthe two layers, with smallest stresses occurring when their thicknesses were approximately equal. The largest stresses were obtained during the temperature cycling experiments. Differential thermal contraction is believed to be responsible for the generation oftensile stresses of the order of 4 MPa in AB coatings on immersion into water at 5°C. Partial relaxation of this stress then occurred and this caused the formation ofcompressive stress when the sample was restored to a higher temperature. The behavior of AA was basically similar but with a drift towards a permanent tensile stress. AB on top of AA showed behavior sinlilar to that of AB.
CONCLUSIONS The highest residual stresses observed in this study were caused by differential thermal contraction between coating and substrate. A temperature change silnilar to that between a dry dock in a warm climate and the open sea gave stresses of 4 MPa and more, a significant fraction of the failure strength. Other sources of residual stress are complex and are probably highly specific to the coating composition. When using bi-Iayered coatings the changes in stresses were moderated somewhat and it appears that a significant and beneficial reduction in the stress magnitude can be achieved by appropriate combination of thicknesses of the two layers.
ACKNOWLEDGMENTS The authors acknowledge Courtaulds Coatings for providing the materials used in this study and for the provision of a strain gauge signal conditioning unit. We are grateful to M Buhaenko for advice and for stitnulating discussions throughout the project.
REFERENCES 1. 2. 3. 4. 5. 6.
E M Corcoran, 1.Paint Technol., 41 (1969) 635. Van Gu, MPhil thesis, University of Newcastle upon Tyne (1997). S G Croll, 1. Coatings Techno!., 50 (638) (1978) 33. S G Croll, 1. Appl. PO(1'11l. Sci., 23 (1979) 847. S G Croll, 1. Coatings Technol., 51 (659) (1979) 49. 0 Negele and W Funke, Progl: Org. Coatings, 28 (1996) 285.
Balancing the Color and Physical Property Retention of Polyolefins Through the Use of High Performance Stabilizer Systems
M. J. Paterna, A. H. Wagner and S. B. Samuels C)Jtec Industries, Research & Developnlent, 1937 West Main Street, P.O. Box 60, Staniford, CT 06904-0060, USA
INTRODUCTION Polyolefin usage is growing in many n1arkets, including construction, farming, consun1er goods, toys and automotive parts. Unfortunately, polyolefin atiicles will degrade and undergo loss of physical properties and change in appearance unless adequately stabilized. UV stabilizers are added to inhibit degradation during outdoor exposure. To combat degradation during processing and fabrication, polyolefins usually contain phenolic antioxidants (AO), potent radical scavengers, and one or more hydroperoxide decomposing secondary antioxidants (thioesters, phosphites). Several factors must be balanced when designing a stabilization package for polyolefins. The package must be cost effective and must maintain part aesthetics on aging. In addition, the package must ensure that the resin will process well and that the fabricated part will meet its targeted service life in the intended application. Since stabilization packages typically contain several additive components, the potential interactions, chemical and functional, of the additives cannot be ignored. For example, the additives in a stabilization package may interact synergistically,1,2 as in the case of primary and secondary antioxidants. Negative interactions between additives are also possible, and when unanticipated, these can lead to premature product failure and legal liability. An example of adverse additive interactions is the reduction in color strength that occurs for certain combinations of pigments and hindered amine light stabilizers (HALS).3
162
Weathering of Plastics
When a stabilizer package is exposed to environmental agents (ultraviolet light, acid rain, gaseous byproducts of fuel combustion, smog), additional complex additive interactions are possible which may adversely affect the article's appearance or retention ofphysical properties. For example, upon exposure to exhaust gases (which contain a high concentration of NO x), resins containing certain hindered phenolic antioxidants; will discolor. This phenomenon, known commonly as "gas fading", can occur during warehouse storage prior to or after fabrication or at anytilne during the part's service life. Samuels et al. 4 studied the effect of exhaust fumes on a series of HALS and antioxidant packages. They found that exposure to exhaust fumes greatly increased the rate of discoloration ofmost HALS/AO packages. The rate ofdiscoloration upon NO x exposure was found to be primarily dependent on antioxidant structure, but the HALS can also influence the discoloration rate. In order to avoid gas fading, it is possible to use a very low pKa HALS, like HALS-l, with an antioxidant prone to gas fading since the rate of discoloration with this blend is very low. However, this combination will result in the sacrifice of physical property retention since HALS-l is not a high performance HALS. Formulations containing high performance HALS and a gas fade resistant antioxidant will not discolor upon NO x exposure. With care in formulating, it is possible to achieve excellent UV performance without encountering gas fade discoloration. An exan1ple is the combination ofHALS-2 and the gas fade resistant antioxidant, AO-l, a 1,3,5-tris(4-tert-butyl-3-hydroxy-2,6-dimethylbenzyl)-1 ,3,5-triazine2,4,6-(IH,3H,5H)-trione. The latter also affords excellent processing protection. 4 In many applications, adequate protection of the resin can be provided by a single UV stabilizer. In some systems, however, it is advantageous to use UV stabilizers with complementary mechanisms. UV absorbers competitively absorb the radiation, reducing the an10unt reaching the chromophores (impurities, microstructural features) in the polymer, thus reducing the photoinitiation rate. Hindered alnines are multifunctional as well, and will trap radicals and decompose peroxides at use temperatures. It has been suggested that HALS will also quench excited state complexes. Stabilizer packages containing antioxidants, HALS and an UV absorber are commonly used. The current study builds upon the previous work4 by investigating the interactions ofUV absorbers with antioxidants and HALS. To elucidate the relative contributions of the HALS, UV absorber, and antioxidant components to gas fade color formation, studies were conducted to determine the relative rate of color development in polypropylene (PP) formulations prepared with systematically varied HALS/UV absorber/antioxidant combinations. The formulations were exposed to the fumes of methane combustion. These results were compared with those results of additive saturated filter papers.
Color and Physical Property Retention
163
EXPERIMENTAL FILTER PAPER Five percent (wt/wt) solutions of the additives were prepared in methylene chloride. Volumes ofeach solution were mixed to achieve the correct additive ratios. Cellulose filter papers were allowed to soak in the blended solutions for five minutes before being allowed to air dry. Paper color was determined with a Macbeth Color Eye Colorimeter under Lab conditions with illuminate C, 2° observer, specular component excluded, and UV component included. Filter papers were exposed in a United States Testing Co. Atmospheric Fume Chamber (Model 8727) with custom temperature control. The charrlber was maintained between 57-60°C. The papers were exposed for a total of24 hours. PLAQUES Solid additives were weighed into polymer powder and dry blended for five minutes. The blended material was melt-n1ixed in a Haake torque rheometer base equipped with a 0.75 inch 25: 1 single screw extruder. The polymer was processed at 50 RPM and 220°C melt temperature. Plaques 2 x 2.5 x 0.100" were prepared by compression molding at 275°C on a PI-II Press. Plaques were exposed in an United States Testing Co. Atmospheric Fume Chamber (Model 8727). The chamber was maintained between 57-60°C. The plaques were exposed for a total of 48 hours. MATERIALS A variety of UV absorbers (Figure 1) and HALS (Figure 2) were tested with and without AO-2, a 1:2 blend of tetrakis[methylene (3,5-di-tet1-butyl-4-hydroxy-hydrocinnanlate)] methane and tris(2,4-di-tert-butylphenyl)phosphite. Three classes of UV absorbers were tested: hydroxytriazines, hydroxybenzophenones, and hydroxybenzotriazoles. Within the benzotriazole class, three different absorbers were evaluated (UVA-2, UVA-3, and UVA-4). Five different HALS were evaluated. The HALS varied in basicity with pKa values fronl 9.0 to 5.7. 4 Mantel's Profax 6501 unstabilized polypropylene was employed in this study.
RESULTS AND DISCUSSION FILTER PAPER SCREENING TESTS As part ofthis investigation, a rapid screening method was used to predict the relative propensity of different stabilizer components and additive mixtures for color formation. The test involves impregnating paper filters with solutions of the additives. The impregnated papers are
Weathering of Plastics
164
then exposed in a gas fade chamber and color development is monitored as a function of time over a 24 hour period. UVA·1 UVA·2 The advantages of this method UVA·3 include its simplicity, flexibility, speed, and cost when compared to evaluations involving resin extrusion. Although the filter paper method is not without its shortcomings,4 it does allow the rapid UVA·S UVA·4 determination ofwhether UV absorbFigure I. UV absorbers. ers are susceptible to gas fade discoloration, the effect of UV absorber structure on gas fade 1;\ MIXTURE HALSo1 HALs-4 01' HAl.$. uw.s.-s discoloration, and the effect ofHALS HALSo2 HAL~ on the gas fade discoloration of UV absorbers. .• Gt«lPJi The discoloration of filters impregnated with either UV absorbers HALS-3 HALS-6 or HALS alone was quite low, as ilR..Vanety(l( lustrated in Figure 3. Although their eJrooPl structures contain phenolic moieties, hydroxybenzophenone, hydroxybenzotriazole and hydroxytriazine UV absorbers do not gas fade at a signifiFigure 2. Hindered amine light stabilizers. cant rate. In contrast, filters containing a high pKa HALS with a UV absorber discolored to a greater extent (Figure 4). A low pKa HALS and a UV absorber combination did not discolor under these conditions (Figures 5). With a given UV absorber, HALS structure will determine the degree of gas fade in UV absorber/HALS blends. As illustrated in Figure 4, the UV absorber structure can influence gas fade in UV absorber/HALS blends. UVA-l, representative of the benzophenone class of UV absorbers, exhibited the greatest discoloration of all the UV absorbers tested. Within the benzotriazole family, substitution on the phenolic ring in the ortho position will decrease the degree of gas fade discoloration. When chlorine is located in the 5 position, the initial color is increased but the rate ofdiscoloration is similar to that for the 5-H-disubstituted benzotriazole, UVA-3. The
W---J..-,..Jr
R.V~';el)'O(
(\lIu~dH:
C'C~ac..."1J~,
~~. ~~
. =t-4,.
165
Color and Physical Property Retention
3On-------;::==~ .Oht Z S · . j . f - - - - - - - - - - - - - l m5 hr
i
~
.U~
j;: ..
20.u-------------=--'-l
30.,..,-;:===;------------, 110 hr
..
25
mShr 1 - - - - - - - .24 hr
.::; 20
.s
~
15 J-I----.....,..
'"
!l:
~... 10
;..
;;..
5
5
o UVA-4
UVA·2
UVA-3
UVA·!
o
UVA·S
Figure 3. NO, mediated discoloration ofUV absorbers.
UVA·4
UVA·3
UVA·l
UVA·5
Figure 4. Effect of HALS on NO, mediated discoloration of UV absorbers.
ImHALS.31
1k - - - - - - - - - .HALS-4 ~ "tl
0.07
j;:
0.06
~
0.05
!
0.04 0.03 0.02 0.01
o With AO·l
Without AO-l
Figure 5. Effect of HALS on NO, mediated discoloration of Figure 6. Rate (dYUdt) of NO, mediated discoloration for PP UVabsorbers. formulations containing HALS. 0.15% HALS, 0.04% AO-2, 0.05% CaSt, 0.08% TBPP, 100 mil.
triazine UV absorber/HALS blend discolored at a slightly slower rate than the disubstituted benzotriazoles.
PP PLAQUES As demonstrated in the previous study4 in LLDPE, in a PP resin matrix, HALS alone (without phenolic antioxidant) do not discolor appreciably upon exposure to NO x. As shown in Figure 6, the rate of discoloration is quite low and appears dependent on the substituent on the piperidinyl nitrogen. However, when an NO x sensitive antioxidant is introduced into the for-
Weathering of Plastics
166
•.05
0.05
U4 0.04
~
:;; 0.03
;:::;
....
..
~ ~
,:a:;
~ 0.63
....
0.&2
~
0.01
~
lIl. R and North American commercial clearcoats; are reproduced in Figure 3. The results were used to rank these materials as likely candidates. But these are not full systems and full l~lut~.n •• system results do not always conespond Cltt\l\·~ closely to the isolated film results. With the development of the "L'1Trans.,Expos." term for It • films as a measure of oxidation product accu.;u • t,: 1·ju,. (lI'lI'Jlh ,11 llt,un I mulation as measured in transmission, a Ii! MII)y l'ltlI«firlfl similar procedure can be applied to the PAS-IR data on complete systems. As the Figure 3. The accumulation of photo-decomposition products portion of the sample examined is limited to ("~trans,5K hrsBoco,Boro"x4), film loss (JIm x4) and evaluation ofUV the top layer and since the spectrum is most protection in coatings after 5,000 hrs of accelerated aging as representative of the outer most portion measured in isolated films by transmission. thereof,14 the oxidation product accumulation will be highlighted even more than in the transmission experiments. Since the intensity observed in the region used to calculate "L'1" is not overwhelming, the problem of PAS saturation is minimal. Confounding the fundamental understanding ofthese data is the gradient nature ofthe photooxidation process. As illustrated in Figure 1C, the accumulation of oxidation products is apparent, but at present, there is no easily applied model for analysis of the distribution of products in non-uniform samples. However, the L'1[(vNH,\OH)arealvCH area ]PAS,Aus. for this system proceeds from 0.72 to 1.3 to 3.25 for 1,2 and 4 years exposure in Australia. "L'1 PAS" will have different values compared to "L'1Trans " but rankings made on the basis of"L'1 PAs " are similar and agree with field experience. There are other intriguing insights obtained by the combination ofthese two techniques. In Figure 4, the spectral data obtained for new urethane system and aged samples of two colors are presented. The coating in the new state has not undergone all of the isocyanate chemistry possible. After weathering, the samples have lost the isocyanate band and developed a "L'1 PAS,Fla.3y" of O. 73 for the red sample, but the silver sample has a"L'1 PAS ,Fla.3y" of 1.78. The evidence of a color dependent photooxidation is reflected in the UV results. The UVA in the red sample has developed a concentration gradient through half of the clem'coat. The silver sample, however, has lost considerably more of the UVA and the gradient includes the r--+--+---I '.. due to defects that generated i" If 1 \1 ,.. 1 in the sample. Therefore, ~ embrittlement of matrix can " 1/,/ \_~.~~~-\. ~_._-- -_... .. be controlled by two associ1f/+--+----t---1,ir---t .. ated processes: reduction of molecular weight and in1, .1-, '"':-_~ ...~_"",_'--_~_l-_~_-l-. __..J. ...._ _...I_'--_....I_.. creased crystallinity.18 As a 'Ii
/,,1
I"
\
J
...
result of morphological changes in the material's structure, chemical mechanisms are also changed. The degradation of the specimen leads to the formation of monoclinic forms and a reduction in crystallite size (Table 4 and 5). Morphological changes affect the diffusivity as well as other physical properties. This detailed study on nanocomposite and pure PBT shows that total crystallinity is not the only parameter that should be considered when interpreting structural data. Crystallite size, the type of crystalline structure, and their effect on conformation also need to be considered. Polymer morphology may, in some cases, be a greater influence on degradation than its chemical structure. 19 This study also shows that during the aging process some reorganization in structure occurs. The crystallite size of the PBT matrix first rose, then remained approxiFigure 8. Degree of crystallinity (I) and crystallite size (2) versus exposure time for UV exposed pure PBT samples on the non-radiated surface.
208
Weathering of Plastics
mately constant, and finally dropped. For PBT matrix, the crystallite size decreased starting at 1200 hrs of exposure and continued until 1700 hrs of exposure. At the same time, mechanical properties decreased. It is suggested that decreasing crystallite size is a criteria of long-term degradation and full reorganization of structure. We can also predict that if the crystallite size decreases mechanical properties will show a corresponding decrease (strain at break, maximum stress, impact strength).
CONCLUSIONS 1. 2. 3.
4. 5.
6.
7.
It was found that the depth of UV light penetration was limited to approximately 25 IJm for the nanocomposite and to 50 IJm for pure PBT. The filler serves as a screen to block the penetration ofUV light into nanocomposites and therefore improves weatherability. Chemi-crystallization probably occurs in the presence of filler by mechanisms similar to those that occur in the pure polymer, involving sholt-range motions of the molecules. UV radiation increases the degree of crystallinity of exposed surfaces. The degree ofcrystallinity is not the only parameter that should be considered for the reorganization of structure. Clystallite size and type of crystal structure are also important. During the aging process, some reorganization in structure occurs. The crystallite size of the PBT matrix initially rose, remained constant, and finally dropped. We suggest that decreasing crystallite size is an important criteria of long-tenn degradation of PBT matrix. Our results indicate that the service life ofthe polymer can be significantly increased by using clay (montmorillonite) as a filler.
REFERENCES 1. 2. 3.
4. 5. 6. 7. 8. 9. 10. 11.
Bank, L.C., Gentry, T.R., Barkatt, A., and Reinf, 1., Plastic Composites, 14, 559 (1995). Morgan, R., American Society of Composites- Second Technical Conference, Delaware, 250 (1987). Toennaelae, P., Suokas, E., Paeaekkoenen, E., and laervelae, P.K., Translation [rOtTI KunststojJe German Plastics, 76, 9 (1986). Bryk, M.T., Degradation of Filled Polymers, High Temperature, and Thermal-Oxidative Processes, Ellis Honrood, Chichester, (1991). Rabello, M.S. and White, l.R., Po(vmer Composites, 17, 5 (1996). Davidson, D. and Stewart, P., Society of Plastics Engineers ANTEC Tech. Papers, 31, 977 (1985). Hu, X., Xu, H., and Zhang, T., Polymer Degradation Stability, 43,225 (1994). Nikolova, M. and Mateev, M. Polymer Degradation Stability, 43, 977 (1985). Narisawa,1. and Kuriyama, T., AngelV Macromol. Chemistl)!, 216,87 (1994). Casu, A. and Gardette J., Polymel; 36, 4005 (1995). Goldman, A.Ya, Sorokin, A., Eisenhour, D., Barajas, A., Montes, J.A., and Beall, G., Intern't Conference on Polymer Characterization, Univ. of North Texas, Jan. (1997).
Effect of Aging
12. 13. 14. 15. 16. 17. 18. 19.
Murthy, N.S. and Minor, H., PoZrmeJ; 31,996 (1990). Cerius User Guide. Molecular Simulations Inc., San Diego, (1997). Peak Solve-Peak Fitting for Windo\vs, User's Guide. Galactic Industries Corp. (1991-95). Garbauskas, M.R., LeGrand, D.G., and Goehner, R.P., Advances in X-Ray Analysis, 36, 373 (1993). Wypych, G., Handbook of l\1aterial \Veathering, 2nd edition. Chem Tec Publishing, Toronto (1995). Gooden, R., Davis, D.D., Helln1an, M.Y., Lovinger, A.J., and Winslow, F.H., Macromolecules, 21, 1212 (1988). Minkova, L. and Nikolova, M., Polymer Degradation Stability, 25, 49 (1989). Schurz, J., Zipper, P., and Lenz, S., 1. Macromol Sci., Pure Appl. Chem., A30, 603 (1993).
209
The Influence of Degraded, Recycled PP on Incompatible Blends
Chiudia M. C. Bonelli, Agnes F. Martins and Eloisa B. Mano Instituto de Macronl0leculas Professora Eloisa Mano, Federal University ofRio de Janeiro, 21945-970 Rio de Janeiro, RJ, Brazil
Charles L. Beatty Departnlent ofMaterials Science and Engineering, Universit)) ofFlorida, Gainesville, FL 32611, USA
INTRODUCTION The recycling of post-consumer polyolefins has attracted much interest since these resins low density polyethylene (LDPE), high density polyethylene (HDPE) and polypropylene (PP) - are some of the most common polymers in the domestic plastic waste stream. 1,2 About half the weight of the total polYlner produced in the world is composed of polyolefins. Those are the cheapest plastics and are largely used for packaging. Their impact on the environment is considerable due to the low density and the hollow shape of the one-way packaging items, like bottles, containers, bags, etc, which contribute for them to en1erge either in waters or landfills. The large volulne they occupy makes them more conspicuous as nature pollutants than other waste products of equivalent weight. Recycled plastic can be obtained by two different approaches: a two-step process, involving plastics fractionation and processing of the separated plastics fractions; and a single-step process, using directly the mixture of plastic residues. Reprocessing mixed polyolefin waste can lead to products with lowered mechanical properties, since these polymer mixtures are usually incompatible. 3 Due to its favorable characteristics of price, density and versatility, PP is gradually replacing some materials. Its incompatibility with LDPE and HDPE causes loss of the mechanical properties. For example, HDPE or LDPE items may contain PP caps as contaminants, which are difficult to completely separate from the other polyolefins due to their similar densities and physical properties. 4,5
212
Weathering of Plastics
The use of compatibilizing agents may overcome this problem to a certain extent. Compatibilization ofthe Inultiple composition mixtures offers the possibility of reversing the deterioration of properties as less sorting occurs. 6 It can be commonly attained in principle through melt processing techniques, using in situ fonned copolymers, or adding copolymers, or low molecular weight con1patibilizing compounds. The CopolYlners have Segn1entS capable of specific interactions and/or chemical reactions with the blend components. 7 In this paper, we investigated the compatibilizing action of molecularly modified, recycled PP on the mechanical properties of SO/50 PP/I-IDPE blend, as suggested by preliminary experiments carried out in this laboratory.
EXPERIMENTAL MATERIALS The raw materials used in this work were: • PP, supplied by PPH Companhia Industrial de Polipropileno, Rio Grande do SuI, Brazil; type H503; specific gravity, 0.90 g/cm 3; MFI, 2.9 gilD min; • HDPE, supplied by Polialden Petroquimica, Bahia, Brazil; type BT 003; specific gravity, 0.95 g/cm3; MFI, 0.3 gilD min; • Post-consumer rigid plastic waste, supplied by the Municipal Company of Urban Solid Waste - COMLURB, Rio de Janeiro, Brazil- 50 kg.
METHODS The post-consumer raw n1aterial for the preparation of recycled PP (PPrl) was submitted to cutting, washing with water and drying in industrial equipment. A representative sample of the resulting flakes was taken for use in this work. Fraglnents ofPP were separated from other polymers by floating successively in water and hydroalcoholic solution (sp. gr., 0.91 g/cm 3) in tanks of200 liters and dried at room temperature (30°C).2,3 Binary and ternary blends were prepared in a Brabender single-screw extruder, model GN F 106/2, with L/D=25 and screw diameter 19 mm; screw rotation speed, 100 rpm at 190,200,210 and 215°C. The extrudates were cooled at 25-30°C and reduced to particles under 2.7 mm length. For the ternary blends, recycled PP was incorporated on a basic 50/50 PP/HDPE mixture. PPrl was ground and extracted by methyl-ethyl-ketone for 60 hours. The extracted recycled material (PPr2) was also incorporated to the binary mixture. Molecular weights (Mw ) were determined by GPC in a Waters 510 equipment, with differential refractometlY 410 detector, using trichlorobenzene for polymers and chloroform for extract as effluents. IR spectra were taken in FTIR Perkin-Elmer 1720 spectrometer. Solid-state NMR spectra were performed in a CP-MAS, Varian VXR 300 equipment, frequency 75.4 MHz, pulse 90°. DSC data were obtained in a Perkin-Elmer model DSC-7
Recycled PP
213
Table 1. Physical, thermal, and mechanical analyses of virgin and recycled polyolefins Test M\\ M\)M n Tm,oC Tc,oC Tonseb °c Tensile strength at Stress, MPa yield Elongation, % Tensile strength at Stress, MPa rupture Elongation, 0/0
PP
HDPE
PPrl
PPr2
172,500 5.0 166 117 440 36 13 35 640
277,000 8.9 139 121 470 29 11 21 630
38,500 3.5 163 128 443 28 14 27 523
45,500 2.2 163 128 436 29 9 29 601
equipment, using IOo/2°C/min heating/cooling rate. The melting temperature (Tm) and crystallization temperature (Tc ) analyses were run from 30 to 200°C. TGA analyses were carried out in a Perkin-Elmer 7 Series Thermal Analysis System, using 1DOC /min heating rate under nitrogen, from 100 to 550°C. Melt flow index (MFI) tests were performed according to ASTM D1238, procedure A, conditions E and L, in a Emic equipment IFT-315. Specific gravity measurements were taken according to ASTM D792. Tensile tests were carried out according to ASTM D1708 in an Instron tensile tester, model 4204, 1 kN cell, cross-head speed of 1 cm/min, gauge length of 2.225 cm. Samples were cut from 0.1 x 15.0 x 15.0 cm plates, molded in a Carver press at 200°C and 22.2 kN, for 5 minutes.
RESULTS AND DISCUSSION The total loss in the industrial grinding, washing and drying was about 10% in weight. PP fraction, obtained by sink-float procedure, represented 10% of the total fragments of polyolefin residues. The extraction of polar contaminants from recycled PP (PPr1), cOIning from unremoved food residues (oils and fats), resulted in only 1.5 % waxy extract ( Mw = 725, Mw/M n =1.2), remaining apparently unaffected powder residue (PPr2). The extract shows monodispersity, which could be expected for non-polymeric nlaterial. Table 1 presents the results of the physical, thermal and mechanical tests perfolmed on virgin and recycled polyolefins. The nl01ecular weight ofPPr1 was lower than the virgin PP, since the recycled n1aterial may have been exposed to environmental and thermal degradation involving mainly n1acromolecular chain cleavage, probably with oxidation to some extent. Consequently, there was an increase in the melt flow index (14.8 g/l 0 min) and a decrease in the tensile strength at
214
Weathering of Plastics
~
I
Ii I'
I~
I I .
I .
I
.. ' t'
,",.
14'
III
~"" ..,,_ , "._,v,_.'. ,., 141
U
h~
" ,,' .f ,', "" ".u." ..
~
"
U
l
'"
,...•.. , •
,,,
.,.
,
. ~
-0'
,., "", • .,
.
'
'
.. ·.··"1·· .. '··· " t ' ' ' ' , •. l'
••
;~
1)1\11
j\J '. V\ ,__,
"·····'.~··
_._
" ..•:••' .... "; ..-, .. ':;.
_" ." -tt
Figure 2. NMR spectrum ofPPr2.
Figure 1. NMR spectrum of PPr1. lO'
I
i
- r - - - - - - - - - - - - -.. . .
-,.'
. -------------.....,
,
r .,:. 1 M"
Of-
OJ
.
n·
... .
o. _
'
It
..
.. Figure 3. IR spectrum ofPPrl.
, ~,M
ute
Jot. u"
ltQ•
'
Figure 4. IR spectmm of PPr2.
yield (28 MPa) ofPPrl, as compared to virgin PP (2.9 gilD min and 36 MPa, respectively). PPr2 did not exhibit significant differences in the molecular weight and in the tensile strength, as compared to PPr 1. T m and Tc confinned the composition of PPr 1. The temperatures of degradation (Tonser) ofPPrl and PPr2 were the same as for virgin PP. HDPE data showed that the higher degree of crystallinity was parallel associated to the higher Tonser.
Recycled PP
215
Table 2. Thermal and mechanical analyses of polyolefin blends PP/HDPE/PPr2
PP/HDPE/PPrl
Test
50/50/0
50/50/1
Tm,oC
165; 136
165; 135
164; 135
165; 135
163; 134
164; 135
Tc,oC
122; 128
122; 125
122; 125
121; 125
122; 124
122; 124
Tonseh °c
50/50/2
50/50/5
50/50/2
50/50/5
360
452
444
456
454
450
Tensile strength at Stress, MPa yield Elongation, 0/0
-
-
-
-
28
28
-
-
-
14
11
Tensile strength at Stress, MPa rupture Elongation, %
21
22
30
30
24
27
4
5
6
8
21
12
C~n I, leiA " II ~~hll ~B8 5o-p1# }l"'~t, •• A~9
!ted DI:~ 'll I~, 4l*~ 1997
Figures 1 and 2 show NMR spectra of PPrl and PPr2, while ' t c50·S(1..2X ! I ISO.O Figures 3 and 4 present IR specI ~ ....." .... " 1 i .. • 3~/~P,r(~-50-1 I .ISf't: Xl I I tra of PPrl and PPr2. These 0 ~~~~.(3;;~U-3. spectra showed the characteris~.J._._ 1. ,~.m~~U~.,·w2) "Q 100.0 .;~ ~ hdpaIPF"'ht tic peaks of PP, as expected. On -5' - ~o~.~> '~ the other hand, NMR spectra of \ PPrl and PPr2 exhibited an adi\ ditional peak at 33 ppm, \ 2:5. 0 associated with unsaturated ~:~'f C.D ethylenic, probably vinyl termiI I nal group and/or vinyline units, 1 200.0 100.0 300.0 '00.0 at the middle of the chain. IR N, ~~~: ~D:S ~ ~l=l: spectra of the recycled materials did not show any carbonyl abFigure 5. TGA analyses of polyolefin blends. sorption, which evidences that the PP degradation was not oxidative, with chain cleavage and unsaturation which did not change the apolar character of the PP molecule and kept its affinity to other polyolefins. 8 Table 2 shows the results of the thermal and tnechanical analyses performed on polyolefin blends. Concerning thermal data from the virgin polymers and their binary/ternary blends, DSC measurements indicate that the differences in Tc were larger than in T m. HDPE, which has higher degree of crystallinity than PP, was more affected by the presence of PP. TGA performed on polyolefin blends showed results presented on Figure 5. It can be inferred that there was an increase of 100°C in Tonset with addition of 1 % of PPr1 and/or PPr2 to 50/50 PP/HDPE blends, indicating some compatibilizing action of the recycled materials on the blends. "'9
P"f~lJ!('O:SO)
*11"'/~f50/~
..,,,.,t.:
(y\ •
.:>
J ___
,_
i:'p/~cP'l*'~
~e.
14!~.
t
r='=
1--.-
I
\
\\',
'~'i i 1\'"
I
~o.o
216
Weathering of Plastics
•
l"J'lfitwE~~
•
lP'ti(W£1l!"tl
:·U·~~I
P,til»t1'f"tl
~!4'2
..
"'H09"E"P'rI~J4"
PI' HOX PM ~o..~..oJ f>PHWf.:''Pfl!~ ~~,~~
(;
10
,u
·tu
Figure 6. Tensile strength tests of polyolefin blends.
The mechanical prope11ies of the polyolefin blends are shown on Figure 6. It is possible to see the incompatibility of 50/50 PP/HDPE blend, causing the premature break of the test samples before reaching the yield point. The increase of tensile strength and elongation at rupture of ternary blends as conlpared to binary blend indicated some compatibilizing action ofPPrl on the system conlponents. The ternary blends broke before the yield point. Small quantities ofPPrl were enough to produce better results on the mechanical properties of the blends under investigation. The addition ofPPr2 on binary blends was more effective, as far as compatibilizing action was concerned. The ternary blends reached and surpassed the yield point before breaking.
CONCLUSIONS The experimental results suggest that the degradation which occurred in PP molded, post-consumer artifacts after exposition to natural, uncontrolled outdoor conditions provided spontaneous, non-oxidative chemical modifications on PP molecules which brought a certain degree of compatibilization action towards polyolefin residues.
ACKNOWLEDGMENTS The Authors thank the Army Research Institute, Rio de Janeiro, RJ, Brazil, for the DSC analyses.
REFERENCES 1 2 3 4 5 6 7 8
c-s. Ha, H-D. Park, Y. Kim, et aI., Polymers for Advanced Technologies,
7, 483-492 (1996). C.M.C. Bonelli, "Recuperacao Secundaria de Plasticos Provenientes de Residuos So 1idos Urbanos do Rio de Janeiro", M. Sc. Thesis, Instituto de Macromoleculas, Federal University of Rio de Janeiro, Brazil (1993). E.B. Mano, C.M.C. Bonelli, M.A. Guadagnini, and S.J. Moyses-Luiz, Polimeros: Ciencia e Tecllologia, 4 (3) 19-24 (1994). R.S. Stein, "Miscibility in Polymer Recycling", in Emerging Technologies in Plastics Recycling, 513, 3948, ACS Symposiuln Series, Washington (1992) C.L. Beatty, Proceedings of SPE Annual Technical Conference, 3032-3033 (1994). K.C. Johnson, Proceedings of SPE Annual Technical Conference, 3732-3737 (1995). A.L. Bisio and M. Xanthos, "How to Manage Plastics Waste - Technology and Market Opportunities", Hanser/Gardner, New York (1994). Y. Long, B.E. Tiganis and R.A. Shanks, J. Appl. Polym. Sci., 58, 527-535 (1995).
Interactions of Hindered Amine Stabilizers in Acidic and Alkaline Environments
K. Keck-Antoine and D. Scharf Specialty Che111icals Group, BU Additives, Hoechst Celanese Co/po Charlotte, NC 28217, USA H. Koch R&D Departnlent, BU Additives, Hoechst AG; Augsburg, Gerl1zany
INTRODUCTION Hindered Amine Stabilizers (HAS) are very effective UV-stabilizers that outperfonn all other types ofUV-stabilizers mainly in polyolefins. In addition some HAS are known to offer outstanding long tem1 thermal stability. This high efficiency is based on a radical scavenging mechanism. However, cases have been reported where the performance of HAS was significantly lower than expected. In the majority of these cases, HAS stabilized polyolefin films had been in contact with reactive chemicals and subsequently failed prematurely. While the chemical reactivity of HAS is needed for their outstanding performance it can cause antagonistic interactions in the presence ofother reactive chemicals. These interactions can significantly decrease the UV-performance of HAS. In addition, interactions between HAS and reactive chemicals can also influence the long tem1 thermal stability, processing stability and discoloration effects of polyolefins.
ALKALINITY OF HAS Hindered Amine Stabilizers (HAS) are basically radical scavengers which require a certain level of chemical activity. As a result of their amine chemistry, they can be expected to be more or less alkaline. Very often the pKa-value is used to characterize the alkalinity of HAS (Table 1).
218
Weathering of Plastics
Table 1. Alkalinity of HAS (HMW = high moOne possible reaction scheme of lecular weight, LMW low molecular weight) acid-HAS interactions describes the salt fonnation as a result of an acidpKa base reaction. 4 Such a salt fonnation HAS Type [1] [2] [3] would deactivate the functional HAS-l HMW 9.7 9.2 8.6 group of the HAS and consequently HAS-2 9.2 HMW. 9.1 limit its perfolmance. HMW HAS-3 6.5 6.5 5.5 During processing, storage and HMW HAS-4 9.6 use, HAS-stabilized polymers may HMW HAS-5 6.7 be exposed to more or less strong acLMW 9.0 HAS-6 9.3 ids or (more general) reactive LMW 9.2 HAS-7 chemicals which can migrate into the polymer. Further, acids or reactive Table 2. Influence of acid exposure on chemicals can fonn in the polymer HAS-stabilized LOPE films matrix or can already be present due to other additives or ingredients.
=
Acid
UV-Stabilization improvement factor
none
none
1.0
HAS-l >15
HAS-5 >15
HN0 2
1.1
6.8
10.3
H 2S03
0.6
1.4
2.1
san1ple: 300 micron blown film; stabilization: LDPE-l + 3000 ppm HAS; treatment: dipped each 100 h for 16 h in 0.1 n acid, washed with deionized water and dried at room temperature; criterion: exposure tin1e until ~CO=O.3; weathering: X 150 xenon-arc (standard conditions).
UV·STABILITY OF POLYOLEFIN FILMS A typical example are agriculture PE films used for crop enhancement. These films are in contact with reactive chemicals 4 and often show "un-
explained early degradation" under field conditions. In a model experiment, LDPE films containing HAS with significantly different alkalinity were brought into contact with two different acids (Table 2). As predicted from the acid-base reaction, the stronger acid, H2S0 3 (pKa == 1.92) had a more negative impact on the film perfonnance versus HN0 2 (pKa == 3.34). The film with the more alkaline HAS-l (pKa==8.6)3 was significantly more affected by either acid compared to the less alkaline HAS-5 (pKa==6.7).3 Without acid contact, both films revealed a comparable lifetime. To confinn salt formation as one possible mechanism the experimental set-up was repeated and, additionally, the accumulation of selected trace elements in the films was measured. 4 Trace elements were sulfur for H2 S0 3 and Metham Sodium respectively chlorine for Sumi(sc)lex (Tables 3 and 4). In all three cases a correlation was found between the perfonnance of HAS and the accumulation of ce11ain trace elements in the film. The films with the less alkaline
219
Interactions of Hindered Amine Stabilizers
Table 3. Activity of HAS after contact with reactive sulfur containing chemicals
Chemical
H2S03 (0.1 mol/I)
Metham sodium (3% solution)
UV
Stabilization 5000 ppm HAS-l 2500 ppm UVA-l 5000 ppm HAS-5 2500 ppm UVA-I 5000 ppm HAS-I 2500 ppnl UVA-l 5000 ppIn HAS-5 2500 ppm UVA-l
Retained relative elongation [%] after after 2000 h 74
Sulfur content [ppm] after 2000 h
Sulfur increase (linear regression)
Exposure tiole [h] until 1000 ppm sulfur in the film
750
y=0.24x+274.2
3050
95
444
y=O.16x+122.8
5462
73
1640
y=O.80x+31.1
1204
88
737
y=0.34x+45.8
2762
sample: 200 micron blown film; stabilization: LDPE-l + HAS + UVA-l; treatment: each 144 h contact for 24 h with chemical; dried at room temperature; criterion: retained relative elongation at break [%] sulfur content [ppm]; weathering: X 450 xenon-arc; standard conditions (no rain cycle)
Table 4. Activity of HAS after contact with reactive chlorine containing chemicals
Chemical
Sumisclex
UV
Stabilization 1500 ppnl HAS-l 1500 ppm HAS-5
Retained relative elongation [%] after after 1000 h 14 68
Sulfur content [ppm] after 1000 h
Sulfur increase (linear regression)
Exposure time [h] until 1000 ppm sulfur in the film
335 279
y=0.32x+ 15.0 y=0.24x+39.1
3078 4004
sample: 200 micron blown film; stabilization: LDPE-l + HAS; treatment: each 125 h contact for 24 h \vith 0.05 Procymidon solution; dried at room temperature; criterion: retained relative elongation at break [0/0] sulfur content [ppm]; weathering: X 1200 xenon-arc; standard conditions (no rain cycle)
HAS-5 showed a longer lifetime and accumulated less trace elements. This means that the exposure time to reach a threshold trace element level was significantly longer. The trace element accumulation showed linear behavior.
220
Weathering of Plastics
As reported earlier5 there was further evidence of the potential deactivation of HAS due to salt formation. The in-situ formation of [HAS-1 ]sulfite showed an IR-absorption at 2480 cm- I •3 Additionally, films containing in situ created [HAS-1 ]sulfite showed no peaks at 1565 and 1530 cm- I compared to films containing "standard" HAS- 1. Shachar et at. found similar phenomena. 6 Although acid-HAS (base) reactions are a significant part ofpotential antagonistic interactions, it seems that other mechanism may occur as well with complex chemicals.
"LONG TERM THERMAL STABILITY" OF HOPE GEOMEMBRANES Most of the work related to acid-HAS in-
Table 5. Onset of Oxidation (OIT) of teractions has been focused on thin-walled HOPE plaques after contact with reactive
chemicals (OIT = Oxidation Induction applications (films) and the performance
of HAS as UV-stabilizer. Some of the interactions appear to require the presence ofUV-light. 3 OIT, min However, they may also occur in none HAS-l HAS-5 thick-walled polyolefin applications, without acid 36 63 67 without excessive UV-light and may afwith acid 29 56 63 fect the long tem1 them1al performance. sample: 1 mm injection molded plaques; stabilization: HDPE-2 + The lifetime of HDPE plaques was meaAO (proprietary) + carbon black + 900 ppm HAS; treatment: sured by DSC (OIT value). When no acid dipped once into 0.25 mol/l H 2S03 for 48 h; criterion: OIT at 210°C [min] is present HAS serve as excellent long term thermal stabilizers. However, the contact with H2 S03 decreases the OIT value. This decrease is more pronounced with the higher alkaline HAS-l than the less alkaline HAS-5 (Table 5). Time)
PROCESSING OF POST CONSUMER POLYOLEFIN RESINS An interesting effect was observed when processing post consumer polypropylene battery case resin into multifilament yam. Objective of the study ofG. Coy7,S was to efficiently produce quality fiber from recycled raw materials of different sources. The maximum take-up speed was found to correlate closely with observations made during actual processing and can be used as criterion for spinnability. The addition of an antioxidant package increased the spinnability of the compound. Further addition of UV-stabilizers of the HAS type (in particular HAS-I) either maintained the spinnability level or even increased it.
Interactions of Hindered Amine Stabilizers
221
Table 6. Spinning stability of post-consumer polypropylene This result (Taresin. 7,a Maximum take-up speed [m/min] as function of the ble 6) contradicts additional stabilizer package experitnents with a different set-up as described elsewhere. 9 The addition of HAS-l decreased the spinnability beCar battery resin 299 340 273 583 low the level of the waste control sample stabilization: PP-l (50%) + post-consumer PP (50%); criterion: "spinnability", maximunl without any additake-up speed [m/min]; equipment: Research Spin Unit (RSU) [customer results] tional stabilizer. The battery case resin contained (despite cleaning) about 200 ppm of sulfur; indicating acid contamination. Consequently, an interaction between the acid and the highly alkaline HAS-l was expected to take place. This was confirmed by repeating the experiment with the low alkaline HAS-5, which resulted in a strong improvement of spinnability. The poor improvement in UV-stability of the formulation containing the highly alkaline HAS-l provided further evidence of acid-HAS interactions and a corresponding deactivation of the HAS-structure. "Spinnability" Take-up speed as function of stabilizer package 0.04% AO-I, 0.16% P-l, 0.04% AO-I 0.20/0 HAS none O.16%P-l HAS-l HAS-5 Bottle resin waste 207 251 not tested 211 rm/n1inl
DISCOLORATION OF ADDITIVE CONCENTRATES IN ALKALINE ENVIRONMENTS Table 7. Discoloration of additive con- Certain additive concentrates, like those typicentrates. Color deviation after 375 cally used for LDPE agriculture film, may days of storage at room temperature contain polyn1eric HALS, UV absorbers and HAS conlponent in Concentrate
Total color change ~E after 375 days
HAS-l
HAS-5
42.9
4.1
formulation: 81.0% LDPE-l + 12.0% HAS + 6.0% UVA-l + 1.0% AO-2
phenolic antioxidants. Concentrates containing UV absorbers are yellow from the beginning due to the absorption of the benzophenone (or benzotriazole) structure. In contrast to concentrates containing HAS-5, which show little to no discoloration, concentrates based on the highly alkaline HAS-l discolor strongly (Table 7). This continuous discoloration with HAS-l (and
222
Weathering of Plastics
oo~--------------,
I
HAS*11
• •• • ··HAS-S
Q.J-_ _..........AIIIil:....-.,...a=--IIIIIIC:l:;.;----i 2CO
2&J
~
~
4CX)
wave length [nm] Figure 8. Light transmittance ofLDPE films. Sample: approx. 80 micron compression molded film; formulation: see Table 7. FilIns were manufactured from discolored additive concentrates (Table 7).
mostly all highly alkaline HAS) is less pronounced in the absence of either the benzophenone structure (UVA-I) or a hindered phenolic structure (AO-2). The color shift does not occur in the absence of both, the UV absorber and the hindered phenolic structure. 3 In this case it appears, that the HAS structure is not directly involved in chemical interactions. Rather it seems that more alkaline HAS structures create a sufficiently alkaline environment which favors the oxidation of phenolic structures thus creating
I
d
.
co ore qUInone structures. Films compression molded from the concentrates (no letdown) did not indicate a lack of active benzophenone. This is probably due to the very high overall amount ofbenzophenone in the film. However, analysis of the active benzophenone content gave indications of a loss in the range of 10% (Figure I). On the other hand it was previously reported 10 that a difference in light transmittance in the range of 350 - 400 run has a significant impact on the quantity of light available for the crop and consequently the corresponding crop yield.
CONCLUSIONS The high alkalinity of son1e HAS structures is responsible for interactions that can reduce the performance of HAS. Evidence was found that this is not only true for UV-stability, but for long teon thermal stability and processing stability as well. Additionally, certain discoloration phenomena appear to be caused by the alkalinity of HAS. Consequently, the selection of the HAS structure should take into consideration the perfonnance under non-ideal, e.g. acid exposure conditions and should not only focus on the UV aspect. In particular in applications where interactions are quite obvious, the selection must focus on the perfonnance ofHAS and the potential risk of premature failure due to HAS deactivation.
REFERENCES 2 3
Horsey D., Leggio A., Reinicker R.; Hindered amine light stabilizers (HAS)/pigment interactions -HAS structural effects on color strength; published by SPE (Effects in Plastics); 1993. Gray R.; A novel non-reactive HALS boosts polyolefin stability; Plastics Engineering; June 1991. Hoechst AG, inteolal data.
Interactions of Hindered Amine Stabilizers 4 5 6 7 8 9 10
223
Keck-Antoine K.; Stabilization ofagriculture filnls by polynleric HALS with particular emphasis on possible interactions \vith agro - chelnicals; ANTEC '95; Boston, MA. Carlsson D., Can Z., Wiles D.; Polypropylene photostabilization by hindered amines in the presence of acidic species; Journal ofApplied Po(rmer Science, Vol. 33,875-884, 1987. Shachar R., SteIman R., Shai E., Efrat B., Ashkenazi Y., Asenheirn D.; HALS stabilized LDPE agrifilms under the influence of elenlental sulfur, 1996. Coy G.; Processing post-consumer poly-propylene resin; Sunlmer Intern, Virginia Polytechnic Institute and State University; 1995. Coy G.; The properties, morphology and stability ofmultifilament polypropylene yam containing post-consumer recycled resin; Virginia Polytechnic Institute and State University, 1994. Keck-Antoine K.; Interactions of hindered amine stabilizers - During processing and manufacturing; Additive '97; New Orleans; February 1997. Lagier 1., Rooze K., Moens F.; COlnparative agronomical experiment on greenhouse filnls stabilized with HALS and nickel quenchers; Plasticultllre, #96; 1992.
CODE
TRADENAMF,
SUPPIJER
HAS~l
1t) Chimassorb ~.44
elBA Additives
Chimassorb 119
elBA Additives
HAS-J
® Tinuvin 622
elBA Additives
HAs....
® Cyasorb 3346
STRl;C'fURE
Weathering of Plastics
224
eoof.
TR.-\DENAME
SUPPLIER
HAS-5
~
Hocchst AG
Ho!t3\'in NjO
l
STRUCTURE
CH. CH,
P
O
t-
CH.-tCH"'l I CH.
H-N
J
C-N-R
CH, CH, ~
HAS-6
Tin",in 770
elBA Addilivcs
n
H~O-C-ICH,lrc -O~NH
>.:....J
II
II 0
o
\....,('
HoC-tCH ,» HAS-7
Hoslavin N20
CH, CH,
Hooch$! AG
ty
j
I o-_ .... -CH.
H··- N
C····N-H CH, CHl I!
° UVA·l
Chillla,som 81
ellA Adduivcs
Hostavin ARO 8
H~hSl
AG
(( ~
A().I
li)
A().2
Irgano, 1076
1'·1
1')
Irganox 1010
IrCi'fos
J6~
)\. HostallO.\. PAR24
'----
~,&OH
I
I
~
OCaHu
elDA Addih"<S
elBA Addili,·cs
ClIlA Addlll\'CS
Hocch~l AG
V -
HO
~
/,
o
I' CH,CH,COC"H..
rL4 ~"\- o_l, _f)· _.
P
I
I _.---.J
Appendix I. Agrochemicals used in the study
Trade name
Type
Application
KMetham Sodium
sodium methyldithiocarbamate
soil fung,icide, nematicide, herbicide
KSumisclex (Sumilex)
N·(3,S·dichlorophenyl)·I,2-dimethylcyclopropane·I,2dicarboximide
fungicide
Interactions of Pesticides and Stabilizers in PE Films for Agricultural Use
Edina Epacher and Bela Pukanszky Technical University ofBudapest, Depart111ent ofPlastics and Rubber Technology, Institute ofChe111istlY, Che111ical Research Centel; Hungarian Acadelny ofSciences
INTRODUCTION Traditionally Hungary is an agricultural country. In recent years the use ofPE films for greenhouses became widespread, the production of such films increased significantly. Proper stabilization of films used for such a purpose is an important financial issue, a more efficient stabilizer package extends the lifetime of the film, but increases its price. On the other hand, the cost of installing and dismantling the houses, as well as that of the disposal of waste films decrease if the film lasts several seasons. However, the stabilization of agricultural films is a serious technical challenge. During their use, the films are exposed to the effect of oxygen, tTIoisture, summer heat, and UV radiation, among which the last has the strongest influence on lifetime. As a consequence, the most important, or even the sole aspect of stabilization in this field is the development ofan appropriate light stabilizer package. PE grades used for the production of agricultural films practically always contain a hindered amine light stabilizer (HALS) and often also an UV absorber. Properly stabilized films survive two, sometimes three agricultural seasons. However, antagonistic interaction of light stabilizers and phenolic antioxidants was observed sometimes, which may decrease the efficiency of the stabilizer package. 1,2 Further interactions are expected in the presence ofpesticides which are used for the protection of the crop grown in the greenhouse. Usually pesticides have complicated formulations, they contain a number of compounds beside the active component. It is a well known fact that films are destroyed prematurely when certain pesticides are used indicating an antagonistic interaction of the formulation and the stabilizer package. The pesticide must react with the stabilizer decreasing its effect or conlpletely destroying it. The practical importance of the problem is obvious, thus the goal of our study was to identify the pesticide
226
Weathering of Plastics
formulations or active components which decrease the lifetime ofPE films, on the one hand, and to grade the stabilizer packages tested, on the other. Furthermore, an attempt was made to explain the mechanism of interaction in the presence of harmful formulations.
EXPERIMENTAL
MATERIALS The same PE grade (Tipolen FA 2210, TVK, Hungary) was used throughout the experiments. The performance of three stabilizer packages was compared in a film with an anticipated lifetinle of 1 year. Stabilizer package S, the standard system ofthe producer of the film, contained a combination ofTinuvin 622 and Chimassorb 81 UV. The experimental package A consisted ofHostavin N30 and Hostavin ARO 8, while package B ofTinuvin 622 and Chimassorb 944. Tinuvin 622, Hostavin N30 and Chimassorb 944 are HALS compounds and Chimassorb 81, which corresponds to Hostavin ARO, is an UV absorber. An attempt was made to use the widest possible range of pesticides. The formulations used in the largest quantity in Hungary were all included into the study and practically the complete range ofactive cOlnponents were represented among the investigated products including organic phosphorous and sulphurous compounds, halogenides, organometallic compounds, etc.
MEASUREMENTS The selected 24 pesticides were diluted with water to a concentration recommended by the supplier for the user. The films were soaked in these solutions for 1 hour, 1 day and 1 week. The first corresponds to a weak, the second to a moderate, while the third to a strong exposure to the effect of the pesticide. Weathering experiments were carried out under dry conditions for 300 and 600 hours with films treated for 1 day. Tensile properties were measured on small dumbbell specimens cut parallel to the extrusion direction. The measurements were carried out on a Zwick 1445 machine with 100 mm/min cross head speed. Thermooxidative stability ofthe films was characterized by the initial temperature of degradation (Td) measured in nonisothermal degradation experiments at 1O°C/min heating rate on a Perkin Elmer DSC-2 apparatus. FTIR spectra were recorded on a Mattson Galaxy 3020 apparatus in the range of 4000 and 400 cm- 1 wavenumbers, while UV spectra on a Hewlett Packard HPUV 8452 equipment between 200 and 800 run wavelengths.
RESULTS All c01l1binatiolls of24 pesticides, 3 stabilizer packages, and 3 soaking times represent a large number of experiments and measurements, thus some screening tests were carried out in a first step. In latter stages the effect of those pesticides was studied only, which considerably influenced the properties of the films. Our attention is focused mainly on sulphur containing
Interactions of Pesticides and Stabilizers
............. .-- ---,--- --I D 100'..>1
.14000
M: at 2500 kJ 22.5 1.8
All samples contain: O. 1 % calcium stearate, 150/0 talc; Pigment: 1.6% Mixed Blue
Table 1 and Table 2 show that the stabilizer system based on the new dialkylhydroxylamine type process stabilizer was far more weatherable than a systen1 based on the traditional hindered phenol/phosphite system (i.e. LS A system). Visually, the differences in performance are even more spectacular in the red pigmented TPa compositions than what Table 1 suggests. The LS A stabilized sample exhibited severe chalking, while the sample stabilized with the hydroxylamine based system (LS B) showed only a small color change
244
Weathering of Plastics
after 2500 kilojoules of weathering, and no chalking until 4000 kilojoules. In the blue pigmented TPO compositions, the sample containing the hydroxylamine based system showed virtually no color change at 2500 kilojoules, and no evidence of chalking even after 4000 kilojoules of weathering! In comparison, the traditional LS A sample began to chalk at 1920 kJ and was severely discolored at 2500 kJ. The data also indicates that it was much more difficult to stabilize the red pigmented TPO composition than the blue TPO system.
Comparison of Light Stabilizers in PP/EPDM The previous tables showed the dramatic improvement in light stability that can be achieved by substituting a dialky1hydroxylamine process stabilizer for a hindered phenolic/phosphite based system. The weatherability of the hydroxylamine based system can be further enhanced by using more effective light stabilizers than the current state-of-the-art system. This is shown by the gloss retention results in the following graphs.
l.o.1%I.$A
.0."%1.$6
.M5%I.$C
OO.M%LSO
I
GO~============;::L-.. 60,----:===========~ ~ f---------.§ 50 f - - - - - - - - - ;;:: ; .ot----~~~- ~ 40 -1----------;; 'al IX 30 t - = ; : - - - - - - - t - - IX 30 . 1 - - - - - -
a
=
.2
20 t-'=~--
Cl
I
10
f--:::-+--:--2500 kJ Xenon Weathering (SAE J 1960)
Figure I. UV stability of red pigmented IPa (PP/EPDM).
.. 20+-----~
c:; ~
10
.1-------
0..----2500 kJ Xenon Weathering (SAE J 1960)
Figure 2. UV stability of blue pigmented IPa (PP/EPDM).
Figures 1 and 2 show that stabilizer systems containing the dialky 1hydroxylamine were superior to the hindered phenolic/phosphite based system. In addition, the two light stabilizer systems based on HALS 3 (LS C and LS D), a new monomeric NOR type HALS, gave significantly higher gloss readings compared to the system based on the more well known HALS 1.
PP/SSC PLASTOMER Comparison of Impact Modifiers EPDM and EPR rubber have been used as impact modifiers for polypropylene for quite some time. More recently TPO suppliers have begun to use new impact modifiers that have been produced using single site constrained geometry catalysts. The weatherability of TPO compositions containing these new plastomers has not been well documented in the trade litera-
245
New High Performance Light Stabilizer
ture nor has their photostability been compared to the traditional systems. One of the objectives of this project was to compare the weatherability of PP/EPDM and pp/sse Plastomer. The following are the results of this comparison.
oor;::==============::;--
60
so W
40
~
30
c!l
30
.0.7 lS A wlt~ 0.25'" Fled 3B .0.65% lS 8 wit~ 0.25% Fled 38 • 0.7% lS A with I.S·;' Mixed Blue o 0.65'Y.lS B with 1.6% Mixed Blue
g 70
-
.. a::
~60 c; so '"
'o"
4Q
.0.7 LS A wlt~ 0.25% Fled 38 .0.65'1. LS 8
aO.6S~/.lS B
a 2O-f---
'0
"I- '0 t-:=---" o PPlPlaslomer
2500 kJ Xenon Weathering (SAE J 1960)
Figure 3. Comparison of impact modifiers on the UV stability ofTPO.
0.25% Red 38
1.6~\
Mixed Blue
with 1.0". Mixed Blue
30
20
PPIEPDM
wlt~
.0.7% lS A with
PPIEPDM
PPlPlastomer
2500 kJ Xenon Weathering (SAE J 1960)
Figure 4. Comparison of effect of impact modifiers on the UV stability ofTPO.
Based on the reduced discoloration and the higher gloss retention, Figures 3 and 4 show that in both pigment systems the PP/Plastomer TPO compositions were considerably more weatherable than the PP/EPDM compositions. Again, the hydroxylamine based system LS B was superior to the hindered phenolic/phosphite stabilizer system (LS A), and the red compositions showed the largest shifts in color and gloss.
Comparison of Light Stabilizers in PP/Plastomer The weatherability of sse Plastomer modified TPOs can be further improved by the use of the appropriate light stabilizers. This is demonstrated in Figures 5 and 6. Figure 5 shows that the stabilizer systems based on the new hydroxylamine process stabilizer (LS B, LS C, LS D) were much better than the hindered phenolic /phosphite system (LS A) in the red pigmented TPO composition. Although there was little differentiation in hydroxylamine LS systems based on color retention examination of the gloss retention data reveals some interesting results. Figure 6 clearly shows that LS C or LS D which were based on the NOR type HALS outperformed LS B which was based on the conventional secondary amine HALS, HALS 1. Developmental High Molecular Weight Hindered Amines in PP/Plastomer As higher demands are being placed on the UV stability of plastic automotive components, the industry's requirement for new stabilizers to meet the need intensifies. Several new high molecular weight hindered amine stabilizers were examined in the most difficult to stabilize
Weathering of Plastics
246
E
..
._---_._----.~._._----
25
LS A
.O.MY, LS B
.0.65% LS C
OO,65,aS 0
I
96); this seems to be clear evidence for crosslinking. Thus even though the molecular weight averages for the material in the interior after 49 weeks exposure are not very different from the undegraded (reference) value, the molecular weight distribution is different to that for the undegraded material and it is deduced that SOllle reaction has occurred. In the stabilized grade the average molecular weight fell progressively with exposure (Figure 6). No high molecular weight tail is evident at any depth with this lllaterial (Figures 3 and 4). The whole molecular weight distribution shifted to the left at all depths. The amount of shift was quite similar at all depths. Near the surface of the sample the fall in molecular weight was much less than that observed in the unstabilized samples, indicating that the stabilizer system is very effective in reducing chain scission. In the interior, the fall in average molecular weight was greater than that observed in unstabilized material. This is believed to be the consequence of continued availability of oxygen in the interior. In the stabilized polymer the oxidation rate is much slower than that in the unstabilized PP so that oxygen can diffuse into the interior without being consumed, replenishing that lost by reaction and allowing reactions requiring oxygen to continue. The absence of a high molecular weight tail means that there is no evidence for large scale crosslinking in the stabilized polymer, which is consistent with the suggestion made above that crosslinking is favored when the concentration of oxygen is very low. Crosslinks occur when long chain radicals react together and the likelihood of this happening will inevitably be enhanced if there is no oxygen available for competing reactions and reduced in the presence of a radical scavenger. For the X-EPF 30U polypropylene the addition ofTi0 2 pigment caused a large reduction in chain scission at all depths (Figure 7). After 16 weeks the molecular weight average had dropped only about 10% in the surface zone (0-0.1 mIn) and by much less at other depths. In the presence of pigment the extra addition of stabilizer did not make much difference in the measurements lllade here. It is deduced that the pigment has limited the penetration ofUV radiation and that this has been responsible for the reduction in chain scission. On comparing salnples exposed for 32 weeks with on/off cycles with a 12 hour period with those exposed for 16 weeks continuously it is noted that in the interior the fall in 11101ecu-
268
Weathering of Plastics
lar weight was greater in the cyclically exposed samples (Figure 8). Diffusion during the dark periods will have replenished the oxygen levels in the interior and the observation of lower average molecular weights is to be expected if a higher oxygen concentration favors chain scission, as suggested above. This does not explain why the molecular weight fell to a lower level near the surface when the radiation was uninterrupted than when it was cyclic. Small radicals and hydroperoxides are formed during polymer photodegradation and play an important role in the chain reactions involved. 12 They are produced predominantly near to the surface under the exposure conditions applied here, and under continuous exposure they will sustain the reaction in this locality. During periods ofdarkness it is possible that they migrate away from the surface so that when the UV radiation is turned on again the reaction will not be as rapid as it was immediately before the radiation was switched off. Conversely those reagents that migrate into deeper zones within the salnples may promote greater reaction there than happens when continuous exposure is applied, providing another reason for observing lower Mw values in the interior ofcyclically exposed samples, additional to the oxygen concentration explanation given above.
CONCLUSIONS Oxygen starvation limits molecular degradation in the interior of unstabilized polypropylene except in thin sections (say 2'1'·2:
on polycarbonate melt viscosity was determined by measuring apparent viscosity of two commercial cap-grade resins over a range of shear rates with a capillary rheometer at 270°C. Results are summarized in Figure 6. HPT-I at 4.5 % showed a slightly greater plasticizing effect than BZT-I at 7.0%. Thus the use of a slightly lower melt temperature may be advisable when HPT-I is used in the cap layer. When the melt temperature was reduced to 265°C, the cap resin containing 4.5 % HPT-I gave a viscosity vs. shear-rate profile comparable to cap resin containing 7.0% BZT-I at 270°C.
H1>T·l:
CONCLUSIONS Compared to BZT-I, the new UV absorber HPT-I exhibits stronger absorbance at wavelengths where polycarbonate is most sensitive and improved photostability. As a result, HPT-I provides superior weatherability to twin-wall coextruded polycarbonate sheet. Sheet stabilized with
276
Weathering of Plastics
3.5% HPT-1 in the cap layer exhibits weatherability comparable to or better than sheet stabiIized with 7.0% BZT-1. HPT-1 also features low-volatility and has only a minimal effect on melt viscosity.
ACKNOWLEDGMENTS The authors would like to thank Mr. Guy Jordy, Ms. Emerald Collins, and the Additives Analytical Research Department for their excellent laboratory work in support ofthis project, and to Ciba Specialty Chemicals Corporation for pem1ission to publish this paper.
REFERENCES 1 2 3 4 5 6 7
Press Release PR #41-98, GE Structured Products, September 4, 1998. H. Hahnsen, W. Nising, T. Scholl, H.-J. Buysch, and U. Grigo (Bayer AG), U.S. Patent 5,108,835; 1992. P. A. Mullen and N. Z. Searle, J. Appl. Polym. Sci., 1970, 14, 765-776. A. L. Andrady, K. Fueki, and A. Torikai,1. Appl. Polym. Sci., 1991, 42, 2105-2107. R. C. Hirt, N. Z. Searle, and R. G. Schlnitt, SPE Trans., 1961,1,26-30. D. R. Bauer, 1. Coatings Tech., 1997, 69,85-95. 1. E. Pickett, "Pemlanence ofUV Absorbers in Plastics and Coatings", presented at 7th Annual ESD Advanced Coatings Technology Conference and Exposition, Detroit, MI, September 1998.
Ultraviolet Light Resistance of Vinyl Miniblinds Part 2. Reaction Products Formed by Lead in Air
Richard F. Grossman Halstab
BACKGROUND Exposure of vinyl compositions to sunlight or to laboratory sources of ultraviolet light normally does not result in exudation of the heat stabilizer, whether based on lead, tin, or other metals. In particular, the resistance of lead stabilizers to nligration is well known. 1 Previously, samples of a typical rigid profile extrusion compound were exposed to UV-A and UV-B irradiation for 1500 hours in a Q-Panel QUV accelerated weathering apparatus. 2 Some of these samples were lead stabilized; others contained a tin mercaptide stabilizer. In no case did surface lead increase above the error in detection by the atonlic absorption procedure used, 0.01 ~lg/cm2. On the other hand, the same exposure of lead stabilized vinyl miniblinds led to detectable quantities of surface lead, leveling offat 0.1-0.2 ~g/cn12. This difference in behavior may be a reflection of the very high filler loadings (as lTIuch as 80 phI' CaC0 3 ) used by sonle miniblind manufacturers. The lead compound that exudes to the surface appears to be the reaction product of the stabilizer, tribasic lead sulfate, with HCI, that is, mono- or dichlorotribasic lead sulfate. Lead stabilizers and their HCI reaction products are highly insoluble in dilute hydrochloric acid. 2 Concentrated nitric acid was required to extract the lead from the miniblind surface after UV light exposure. It is interesting to note that in their study, the Consumer Safety Protection Agency (CPSC) found complete equivalence using conc. HN0 3 or dilute aqueous HC1. 3 The dilute HCI, at 37°C for 6 hours, was intended to mimic human digestion. To permit samples to be run more rapidly, conc. HN0 3 was substituted. This yielded the same results. What they dissolved for analysis was, therefore, primarily not insoluble lead stabilizer or its similarly insoluble reaction products, but some readily HCI-soluble lead compound. There are such compounds widely available. Basic lead carbonate, the comnlon constituent of airborne lead-containing dust,4 is quite soluble in dilute HCI (as well as in conc. HN0 3).
278
Weathering of Plastics
Many vinyl miniblinds in the field have been found with surface lead exceeding 0.1-0.2 Jlg/cm2.5 Levels of 1-10 Jlg/cn12 are common, with some specimens having levels as high as 40-80. Indeed, CPSC found surface lead levels well above 0.1-0.2 flg/cm2 after exposure of clean n1iniblind surfaces to UV light (using, of course, conc. HN0 3 to dissolve the product). Their technique, however, involved exposure for a length of time, cleaning the surface, and re-exposure, with their results summing the lead found after each successive step. There is no way that CPSC could have foreseen that such treatment would remove the protective surface layer that the previous exposure generated. Nor is it likely they would have been aware of the insolubility of lead stabilizers in dilute hydrochloric acid.
EXPERIMENTAL An unhighlighted but intriguing feature of the CPSC data was that one ofthe highest levels of miniblind surface lead found in the field (in North Carolina) came from a site where the vinyl contained no detectable lead. 2 The obvious conclusion, that the miniblind was not lead stabilized, was rejected by North Carolina Dept of Health & Natural Resources (NC-DEHNR) in favor of: (A) all the lead must have exuded to the surface, therefore none was left in the miniblind; or (B), maybe it was a different miniblind than the one whose surface was analyzed. 5 It was, therefore, with some relief that a genuinely tin-stabilized miniblind was found (in NW Indiana) that apparently had not been cleaned in 2-3 years. Slats from this miniblind contained no detectable lead (by AA) but about 0.2 wt % tin. (It is thus likely that they were manufactured in North America). The surface, however, had 40-50 Jlg/cm2 lead. SEM-XRF indicated that most of the surface coating was calcium carbonate. Interspersed were large rhombic crystals that appeared to grow from the vinyl surface. These were large enough (5-10 11m) for detailed analysis and proved to contain Pb, Ca, and CI in the ratio of 1: 1:3, and did not correspond to any previously known compound. 6 A test compound was prepared, corresponding to rigid conduit: PVC 100, CaC0 3 40, impact modifier 3, processing aid 2, ester lubricant 2, stabilized with 2 phr of 65/35 di- to monobutyltin isooctyl thioglycolate. Strips of 1.5 mm thickness were exposed to UV-A radiation at 50°C in a continuous moving stream of air (circa 5 l/min) that was previously passed over finely divided (2-5 11m) basic lead carbonate. Under these conditions, in 300 hours of exposure, 50-65 I1g/cm2 of surface lead developed. This coating proved soluble both in dilute hydrochloric and conc. nitric acids. It again contained Pb, Ca, and CI in the ratio of 1: 1:3. Equimolar quantities ofCaC0 3 and basic lead carbonate were dissolved in hot IN HCl. Slow cooling yielded large crystals (1-3 lllin) of the above compound (containing Pb, Ca, and CI in the ratio of 1: 1:3).
279
Ultraviolet Light Resistance
Similar experiments were carried out using the same compound as above, but without CaC0 3 filler. One phr Ti0 2 was used instead to provide opacity. After 300 hours ofUV-A exposure, 50-65 f.1g/cm2 of surface lead also developed (again despite the absence of a lead stabilizer). This product was also soluble in dilute hydrochloric acid, contained Pb and CI in a 1: 1 ratio, and appeared identical to reference samples of basic lead chloride.
DISCUSSION The most common naturally occurring fonn of basic lead carbonate, hydrocerussite, [2PbC0 3 .Pb(OH)2], corresponds to:
o
II O-C-O- Pb-OH Pb/
"O-C-O- Pb-OH II
o Reaction with HCI generates CO 2 and basic lead chloride [Pb(OH)CI]. The latter is reasonably light stable, at least in comparison to PbCI 2, which rapidly loses Cl 2 to leave colloidal lead, much like AgCl. Thus basic lead chloride as a stable end product from the settling, or static attraction of lead dust, in air is not surprising. It is well known that PbCI2, although not a strong Lewis acid, forms double compounds with alkali metal and alkaline ealth halides, e.g., CaCI2.PbCI2.7 This is apparently also the case with basic lead chloride. For example, we find a 1: 1 addition compound with hexa-chloro-l,3-butadiene [C 4CI 6 .Pb(OH)CI] that appears highly resistant to UV light degradation at 50°C. The reaction product in the presence of CaC0 3 appears to be the double compound CaCI2.Pb(OH)Cl. This is quite interesting since CaCl 2 does not otherwise appear, despite the prevalence of CaC0 3 at the degraded vinyl surface. Certainly one must suspect that the presence of the lead salt is involved, and thus consider generally the extent to which stabilizers may be able to transfer chloride to receptive "filler".
CONCLUSIONS It is likely that most, if not all, of the lead-containing detritus found on the surfaces of vinyl miniblinds results from the conversion of lead dust in air to chlorinated products, principally basic lead chloride, from the HCI produced by UV-light assisted degradation of vinyl. These
280
Weathering of Plastics
conversion reactions probably serve to retain lead-containing compounds accumulated from the air, as compared to chemically neutral surfaces (e.g., painted or coated wood or aluminum equivalents). There seems to be no need for the presence of lead stabilizer in the compound for the surface lead accumulation reactions to occur. Although there is no requirement in terms of desired properties to use lead stabilizers in vinyl miniblinds, the widely publicized conclusion that such stabilization adds to the hazard of heavy metal exposure is simply not justified by experiment, and probably resulted from consideration ofdata from the field using too narrow a technical base.
REFERENCES 1. 2. 3. 4. 5. 6. 7.
The Environmental Itnpact of Lead Stabilizers, Nordic Plastic Pipe Association, Stockholm, Jan. 1995. R.F. Grossman and D. Krausnick, JVAT, in press. W.K. Porter, CPSC Report, Miniblind Lead Investigation, Sept. 18, 1996. Ter Haar and Bayard, Nature, 232, 553 (1971). Private comluunication, North Carolina Dept. of Environment, Health and Natural Resources. J.\V. Mellor, Inorganic and Theoretical Chemistry, Vol. VII, Longmans, Green & Co. D. Greninger et al., Lead Chemicals, ILZRO, New York, 1975
Case Studies of Inadvertent Interactions Between Polymers and Devices in Field Applications
Joseph H. Groeger, Jeffrey D. Nicoll, Joyce M. Riley, Peter T. Wronski Altran Materials Engineering, a Division ofAltran Corporation, Canlbridge, MA, USA
INTRODUCTION Polymeric compounds are selected for a wide range of applications by technical persons with a variety of backgrounds. Initial choices may be moderated by other specialists who are often unaware of the potential pitfalls and adverse interactions associated with the use of cost-effective or inappropriate alternate materials. Manufacturers who provide subcomponents may not be included in the design reviews of finished products into which their components are being used. Additionally, suppliers of commercial polymeric n1aterials may be unaware of how their materials are being applied. As a result of these and other considerations, materials selections may be made based on a review limited to basic engineering properties. Considerations oflong-tenn perfoffi1ance and response to specific operating conditions requires a degree of attention and insight that may be overlooked. Several case histories are cited in which some aspect of materials selection and design were deficient in the application. A thennally activated electrical switch fonnerly made with a phenol formaldehyde thermoset resin was redesigned to include a thennoplastic resin. Localized heat associated with the arcing activity of the switch contacts caused thermal erosion of the housing, releasing reactive sulfur compounds which then reacted with the electrical contact faces, causing irregular performance and eventual contact welding. A pressure relief device in a consumer product was found to have highly variable performance as a result ofextensive processing aid additions to the base polymer, selection of poor quality raw materials, and no attention to a root cause analysis with a review of the compound. Plasticizers released from PVC wire insulation at elevated operating temperatures wicked along the conductor strands and onto relay contacts, resulting in a power plant shutdown. Components from a pharmaceutical product container were found to be exuding phthalate compounds which were not expected based on an initial review of the raw materials.
282
Weathering of Plastics
These cases are presented as constructive examples for those seeking to maximize the performance and useful life of devices making use of polymeric components through an integrated materials selection and design approach.
DECOMPOSITION OF THERMAL SWITCH A thermal limit switch used in a number of domestic and commercial appliances was historically manufactured with either a ceramic or a crosslinked phenol formaldehyde, providing many years of reliable service. A change in materials had been implemented to facilitate processing, resulting in a housing made with thermoplastic polyphenylene sulfide (PPS). The housing contained silver-laminated bronze electrical contacts, one of which was mounted on a bimetallic arm to provide thermally-controlled switching action. Failures of this switch were encountered wherein the contacts were found to weld together, resulting in a thermal runaway condition caused by a failure to interrupt current to the heater that the switch was intended to control. A foren'. sic review of representative failed switches was undertaken. Figure I presents a scanning electron micrograph of the surface of a contact reFigure I. Surface of contact showing raised areas where moved from a failed switch. On the surface, welding occurred, 150x. many melted areas are clearly visible. Some ofthese are flat, showing the previously molten condition of the metal contacts. Metallographic crosssections through such a contact showed severe localized melting. Elemental analysis of the contact surface indicated that silver sulfide was present. This compound produced an insulating layer on the surface of the contact, resulting in erratic current flow and localized heating due to limitation of the available contact surface area. Switches in various stages of degradation were operated with thermocouples placed on the contacts and housing. Measurements indicated significant resistive heating, merely due to flow of the rated current. Chemical analysis ofthe polymer heated to the as-found level, using gas chromatography and mass spectrometry (GC/MS), confirmed formation of hydrogen sulfide, carbonyl sulfide, sulfur dioxide, hydrogen, and methane. Examination of the switch housing interior surfaces surrounding the contacts revealed significant erosion of the polymer as shown by the light colored oval region in Figure 2. Closer examination revealed the glass and mineral reinforcement particles within the PPS
Case Studies
283 compound standing in relief, due to polymer pyrolysis. This damage was due to the intense localized heating produced by arcing as electrical contact between the switch contacts was established then broken during normal operation. The combined evidence of contact melting and PPS pyrolysis suggested short-term temperatures in excess of 600°C.
INCONSISTENT PRESSURE RELIEF MEMBRANE A pressure relief membrane used in a consumer product was found to exhibit erratic performance both in quality assurance testing and in the consumer market. The pressure relief device was a critical component and played an integral role in product function and safety. The device was manufactured using a compounded thennoplastic polypropylene which was injection molded into the necessary form. As can be seen in Figure 3 the molded part is quite complex in design; consisting of numerous ribs, radial formations and most importantly, the thin membrane which acts as a pressure rupture diaphragm. Figure 3. Top view of pressure relief device. The latter is coined in the injection molding process. Investigation of the device revealed many areas of misapplied designs and a general focus on processing performance instead of functionality. The thermoplastic compound which was used to fabricate the units made use of a fairly complicated fonnulation. The original base resin was dropped from the supplier's product line and alternates were substituted. In conjunction with these changes, increased device anomalies and difficulties controlling the burst pressure range were experienced. After a preliminary materials investigation ofthe disclosed formulation, interactions of the materials being used were identified as being inordinately complex and in some cases inappropriate for this application. Figure 2. Interior surface of switch housing showing polymer, erosion, 9x.
284
Weathering of Plastics
Organic chemical analyses of representative devices were conducted using GC/MS. This method was selected to confirm the identity of the organic ingredients and processing aids in the questionable formulation. GC/MS analysis of the seals revealed significant formulation variations between different lots ofmaterial. It was determined that the use of additives such as the antioxidants, antiblocking agents, internal lubricants, and other processing aids was inconsistent. The most significant variations were among materials not specified in the formulation. Processing aids such as silicones (used as internal lubricants to modify flow behavior), plasticizers (typically used for increasing impact resistance and adding flexibility), and waxes (used as lubricants and flow modifiers) were noted to be present in many of the device lots. These components appeared at random and were not used consistently. It was suspected that they were added as on-line processing aids to assist with mixing by the compounding operators and/or to achieve a target n1elt flow index. The formulation suffered from years of incremental modification for performance and processing issues which often suppressed the symptoms but never addressed the root causes. For exan1ple, there were three agents listed in the formulation which served as antioxidants. Due to the nature of their chemical functionality, these materials did not enjoy a positive synergy. Instead they competed in the formulation causing none of these materials to offer as much protection to the resin and other organic components in corrlbination as they would when used individually. The antioxidant package was further complicated when a review of their functional characteristics was completed. Originally, the molded pressure relief device suffered from a reaction with copper within the contacting unit surfaces. A metal deactivating antioxidant was added to the formulation to correct this problem. A review ofthe formulation clearly indicated that the original antioxidant was an amine (nitrogen-hydrogen) compound. This antioxidant sustained limited thermal decomposition during processing, leading to the production of amine compounds. These reacted with copper, leading to the formation of blue-colored copper compounds. While the addition of the metal-deactivator was successful in reducing this occurrence, the original antioxidant was left in place. The replacement and original antioxidants were not chemically compatible, nor was the amine antioxidant stable with respect to the antioxidant supplied in the base polypropylene resin. A third antioxidant was then added to improve oxidative stability. A different problem was noted when a scanning electron n1icroscope (SEM) was used to examine selected areas ofrepresentative seals. The high magnification ofthe SEM provided a view of the relative size ofthe individual filler particles and their alignment in key areas such as the diaphragm. Examination revealed that the filler materials had a tendency to agglon1erate in this region and that the overall filler concentration in the diaphragm area was inconsistent throughout many devices. As shown in Figure 4, the talc particles were quite large when compared with the overall thickness ofthe diaphragm. As illustrated in this micrograph, the particles aligned in the plane of the n1errlbrane and created a stacking effect. In this
285
Case Studies
Figure 4. Micrograph of diaphragm cross-section, 605x.
Figure 5. Micrograph of diaphragm comer, 226x.
case, the shape of the particles was inappropriate due to the flow mechanics in the mold cavity. Figure 5 shows the comer at the edge of a representative diaphragm. The filler particles in this area were also dramatically aligned along the curvature of the diaphragm. This suggested that the resin flow in this area during molding was restricted by the presence of the talc particles. This caused the residual stresses in the diaphragm area to be quite high and the particle size of the talc to vary depending on the level of flow restriction during injection. The effect of the talc particle size variation on the inconsistent performance of the seals was significant. This characteristic mainly affected the flow rheology ofthe compound under high shear conditions during injection molding. The talc particle size, in comparison to the diaphragm thickness, also lead to an erratic influence on the tear characteristics during product performance. Talc agglomeration and absence of bonding with the base polymer further contributed to poor performance. Inconsistent diaphragm burst performance was caused by a combination of chemical, physical and rheological phenomena. The lots of devices which exhibited a particularly high burst pressure were the result of a very fine particle size talc in conjunction with a low concentration of processing aids. The increased strength of the base resin and lack of large talc particles for burst initiation necessitated high burst pressures. The devices which exhibited lower diaphragm burst pressures suffered from a combination of large talc particles and an absence of lower molecular weight polymer to assist with the flow and wetting of the filler. This resulted in high orientation effects which led to very high residual stresses causing premature failure. These anomalies illustrate the combined effects of the uncontrolled chemical additives and random talc particle size on the consistent performance ofthe compound. In this formulation, even if extreme care were taken in manufacturing, the number of materials involved and the inherent variability and performance of the talc made it virtually impossible to produce a consistent product.
286
Weathering of Plastics
PLASTICIZER BLOOM FROM PVC CABLE JACKETS During inspections at a nuclear power plant, green liquid deposits were found concentrated on the surface of selected low voltage cables, at their terminations as well as in the instrument panel in which these cables ended at connections. The cables were rated at 600 volts and incorporated a cross-linked polyethylene (XLPE) insulation with a polyvinylchloride (PVC) jacket. The estimated age of the cables was 20 years. The green liquid deposits were determined to be non-drying, with a high viscosity, and good lubricity. Analysis of this liquid by Fourier Transform Infra-Red Spectroscopy (FTIR) confirmed that it was mostly adipic acid diethyl ester. This compound is a common plasticizer for PVC and is typically yellowish in color. An FTIR absorption peak unaccounted for by adipic acid diethyl ester was assigned to a silicone fluid (diphenylsilane). This may be attributed to a second plasticizer used in these cable jackets. Samples of the liquid were pyrolized and the residue was analyzed with energy dispersive X-Ray analysis (EDX). This revealed the presence of copper with traces of aluminum, silicon, calcium, iron, and lead. The presence ofcopper salts in the fluid was responsible for the noted green color. The presence of these green fluid deposits closely followed a record 'heat wave' in this particular region. It was deduced that this elevated regional temperature caused the sudden appearance of these exuding plasticizer compounds from the PVC cable jackets. These compounds can cause severe consequences in electrical systelns due to their insulating properties. If these compounds were allowed to nligrate into electrical switches, relays, or meters they would inhibit proper performance. In this particular case, the plasticizer impinged on the jackets of adjacent cables, causing them to swell then split. In another identical occurrence, a plant shutdown resulted when plasticizer crept onto the surface of electrical contacts used for a punlp motor relay.
EXTRACTS FOUND IN PHARMACEUTICALS The presence oftwo plasticizers, dioctyl phthalate (DOP) and diisooctyl phthalate (DIOP), in a drug formulation caused significant concern to the pharmaceutical companies since aromatics of this type are under regulatory scrutiny. Investigation into the origins of these contaminants led to analytical review of elastomeric components of the product container. Extensive GUMS analysis isolated the source of the DIOP as being the elastomer raw material. Further research indicated that the supplier of this elastomer was adding DIOP during manufacture to act as a melt-flow modifier to control the Mooney Index of the final product. The DOP, however, was traced to contamination from the polymer compounding equipment. Frequently, oils used to lubricate mixing equipment exude into the compound being produced through dust seals, for example. Knowing this, manufacturers will often utilize lu-
Case Studies
287
brication products which are compatible with the polymer products that are produced. In this case, however, the oil utilized contained DOP which would be acceptable in many thennoplastic compounding applications, not slated for medicinal use. The resultant extraction of DOP from components of the product container, however, was not acceptable.
DISCUSSION Development of thermoplastic and thermoset polymer cOlnpounds is a mature science that continues to grow with the developlnent of new types of additives, changing regulatory requirements, and proprietary considerations. The selection of all materials that are incorporated into a cOlnpound may follow lines that are not always clear. Some ingredients may be outdated. Others may have been added for a customer-specific end use and the compound later became available for the general market. A very wide range of off-the-shelf compounds are available for engineering applications. Many will fit into the existing requirements or designs and/or processes may be altered to accomnl0date the compound that best fits the needs. These choices, though, are often limited to the general engineering/technical properties without sufficient detailed consideration of the materials in context of the application. An ideal situation is one in which the end-use nlanufacturer has available the equipment necessary to develop a polymer compound specifically suited to an individual application. In this clean sheet approach, each ingredient may be carefully considered in context ofthe application, aging characteristics, processing effects, and synergy with other formulation conlponents. Conlpounding facilities need not be directly available; contract organizations are available and many of the commercial polymer compound suppliers offer custom compounding services. Analyses of plastics failures and contamination issues often indicate that it is necessary to return to the basics and re-examine the material in context of the application. With this approach, a polymer would be formulated using a minimum number of ingredients, each of which would be the most appropriate and efficient for the end use. By reducing the nUluber of ingredients, the controls necessary for each supplier are greatly silnplified and the potential for adverse interactions reduced. Many raw materials are more complex than may be apparent and, in some cases, the 'hidden' ingredients may be detrimental to an application. Virtually all commercial elastomers are supplied with an antioxidant already included and the type may change periodically. Masterbatching agents and processing aids, such as calcium stearate, may be used when adding antioxidants to a raw polymer. Crosslinking additives and their synergists are another source of antioxidants and other cOlnpounds. Crosslinking is a chemically challenging process in which thermal decomposition of a reactive peroxide is typically used to provide free radicals. This requires an additional antioxidant to protect the polymer,
288
Weathering of Plastics
while reaction products including acetophenone, cumyl alcohol, and acetic acid become available to interact with the other raw materials or additives. In the case of the thermal limit switch, the choice of materials for the housing inherently led to a reduction in the useful life of the device. The stability and useful life of the switch could be readily enhanced through the use of a polymeric housing that does not produce reactive gaseous products. Many thermoset materials are available, as are ceramics. While the near-term economy of using a thermoplastic material may have appeared attractive, the long-term effect on performance may not have been readily apparent when a material substitution was made. In the second situation corrective measures were implemented so that predictable and consistent performance of the pressure relief device could be attained. Compound reformulation took place which included the careful selection of a clean homopolymer base resin, a specially designed and compounded antioxidant and a low aspect ratio, small particle size reinforcement. The compound simplification, in combination with highly functional components, allowed for exceptional performance and reliability. In the example of plasticizer bloom from a set of cables, it is interesting that the simple loss of a compounding ingredient could lead to such indirect, but major consequences. In this case, exposures to long-term conditions of elevated temperature could be surmised, based on the application and service environment. Grafted plasticizers are available. Alternatively, though, a complete reconsideration of the material in this environment would have been beneficial. A polymer compound that is inherently flexible would not involve a plasticizer and the potential adverse effects of its loss. Finally for the pharmaceutical container component example, reformulation of the raw polymer compound, as well as substitution of machine lubricant with a food-grade aliphatic mineral oil was necessary, followed by substitution of increased purity raw materials, before use of this material could be continued.
CONCLUSIONS It is important that the total life cycle ofpolymeric compounds be considered in context of the end-use application. Some basic guidelines can be developed from a review of situations in which the process was not optimized. • The application should be well understood in terms of stresses (thermal, chemical, physical, radiation, etc.). Near- and long-term exposures must be considered. • Review the candidate or existing material with a fresh perspective and careful attention to all raw n1aterials, their quality, and roles.
Case Studies
289
• Simplicity of design facilitates processing, cost control longevity, and quality assurance. This requires that raw materials be inherently suited for the compound, rather than placing a strong reliance on complex additive packages. • The perfonnance of a compound cannot be limited by processing or post-processing handling. Attention to detail is necessary to assure that the intended and realized formulations are identical. • While some of these suggestions may seem tedious, it is often the case where short-tenn economy and lack of application-specific insight may lead to significant losses when a poorly selected material fails in service.
REFERENCES 1 2 3 4 5 6 7 8 9
Hoffman, Werner (1989). Rubber Technology Handbook. New York, NY. Hanser Publishers. Bhowmick, Anil and Howard Stephens, Eds (1988). Handbook of Elastomers. New York, NY. Marcel Dekkel; Inc. Schnaebel, Wolfram (1981). Polymer Degradation: Principles and Practical Applications. New York, NY Hanser Publishers. Sekutowski, Dennis (1992). "Inorganic Additives". in Plastics Additives and l\1odifiers Handbook, Jesse Edenbaum, Ed. New York, NY. Van Nostrand Reinhold. Gachter, R. and H. Muller (1993) Plastics Additives Handbook. Cincinnati, OH. Hanser Gardner Publications. Charrier, JM (1990). Polymeric Materials and Processing. New York, NY. Hanser Publishers. Barth, H. and Mays, J. (1991). Modem Methods of Polynler Characterization. New York, NY. John Wiley Publishers. Engineering Plastics and Composites (1990). Metals Park, OH, ASM International. Rauwendaal, C. (l991).l\lixing In Polymer Processing, New York, NY. Marcel Dekkel:
Automotive Clearcoats
George Wypych Che111Tec Laboratories, Inc., Toronto, Canada
Fred Lee Atlas Electric Devices Co 11 Ipa n)l, Chicago, USA
INTRODUCTION Preceding chapter indicated the need for specific infonnation required to design experiment of material weathering. The aim of this paper is twofold: • generate and systematize infonnation on degradation behavior of automotive coatIngs • provide an example of data selection in preparation for weathering studies The first reason is driven by the fact that such review of technology was not presented so far in spite of the fact that clear coats are of interest of many groups in industry, testing, and university research, including: automobile, motorcycle, bicycle, manufacturers; manufacturers of coatings for repairs; Inanufacturers of exterior metal parts; manufacturers of exterior plastic parts; manufacturers of polymer blends for auto1110tive applications; compounders of plastics; niche markets for clear coats (office furniture, shelving, lighting fixtures, tool boxes, doors); raw material suppliers for coating manufacturers (polymers, curatives, stabilizers, catalysts, initiators, rheological additives, pigments); research institutes (development ofnew products, methods of testing, raw materials used for coatings); national testing institutes; standardization organizations; commercial testing laboratories; university research (development of new products, methods of testing, raw materials for coatings); environmental institutes (studies on environmental impact of degradation products); corrosion protection (research, 111anufacturers of protective chemicals); consultants in the area of weathering and ISO 14,000; military (research, engineering, quality control); aerospace (all aspects of exterior applications of coatings and plastics); others working in the similar fields. This long list shows that the number ofpeople and institutions involved is very large thus a comprehensive review ofinfonnation that is currently scattered is required. As a long list of
292
Weathering of Plastics
references shows, the currently available information is available in many sources - some of which are difficult to obtain. The information provided in this chapter should be updated in the next two years concerning that a very broad research on powder coatings is under way which will affect provided here list of materials used and the list of potential mechanisms of degradation. For the second purpose of this chapter, it is important to mention that the choice of automotive industry is ~dequate because it size warrants a large number of quality research and thus data. This allows to review all aspects of data required prior to weathering testing. It is also important to note that automotive coatings were recently developed from prone to failure technology to robust process which yields durable products. This successful conversion occurred in spite of the fact that the process was complicated by additional needs to eliminate or limit use of solvents which imposed many restrictions on the development process. It s also impoliant to note that there are still large gaps in understanding which this contribution tries to point out to generate required research.
APPLICABLE STANDARDS EXPOSURE IN LABORATORY DEVICES
Table 1. Automotive exterior coatings • applicable standards for the laboratory testing Standard
Equipment
lrradiance, W1m2
Temperature °C
RH,%
SAE J1647
HID chanlber
80
38-47
50
SAE J1960
Xenon-arc (water)
0.55 @340
38 and 70
95 and 50
SAE J2019 SAE J2020
Xenon-arc (air)
80 @300-400
38 and 47
95 and 50
Fluorescent UV
0.43 ~310
VDA 621-4 (Gennan)
Xenon-arc
70 UV/SO dark 63 UV/I0 dark
LP-463PB-16-0 1 (Chrysler) LP-463PB-9-0 1 (Chrysler) BO 101-1 (Ford) GM9125P (GM)
Carbon-arc Humidity chamber Carbon-arc Carbon-arc Fluorescent UV
63-71 none
37.2-38.4 60-65 (BP)
98-100
60±2 70 UV/SO condo
Xenon-arc MO 135 (Nissan) BS AU 148 (British) JIS D0205 (Japanese)
Carbon-arc Xenon-arc Mercury lanlp Carbon-arc Xenon-arc
63±3 or 83±3
50 and 90
89±3 and 38±3 63±3 or 83±3 63±3 or 83±3
50±5 50±5
Automotive Clearcoats
293
OUTDOOR EXPOSURE SAE J 1976 applies to outdoor exposures of automotive coatings and other exterior materials. Coating systems are exposed in panel racks (unbacked exposure) and black boxes.
SOLAR FRESNEL REFLECTOR APPARATUS SAE J 1961 applies to the use of concentrated radiation for exposure of automotive samples including coatings. Apparatus should be operated in dry, sunny climates receiving 3000-4000 h of sunshine. In addition to exposure during the day, specimens are sprayed in the night for 3 min in each 15 min. Two types of exposure are used: non-insulated and insulated (backed with plywood). In insulated exposure, the insulation is only used between November 1 and March 31.
SUMMARY It is interesting to note that the national standards are not playing an essential role in testing of automotive coatings. Only Britain and Japan have national standards. The British standard is old (1969) and probably not frequently used. The laboratory testing is mostly based on SAE standards which allow for the use of all three weathering devices (carbon-arc, xenon-arc, and fluorescent UV). It is important to note that only Xenon-arc device offers full control of all weathering parameters (irradiance, temperature, and humidity) which are specified in the SAE standards.
GENERAL DISCUSSION OF TRENDS Quality of automotive finishes, legal requirements, and environmental concerns were the driving factors for changes in automotive coatings. 1 During 1950-1970, oven-dried alkyd-melamine, lTIOnOcoat, straight-shade, coatings were in the common use. In the period of 1970-1990, the evolution of paint technology was gradually leading toward a more complex systen1 of autolTIotive finishes which eventually included low-solids, solvent-borne basecoats and alkyd-melamine clear coats, later replaced by high-solids basecoats and acrylic-melamine clear coat with UV/HALS stabilizers. These systems included metallic basecoat. During the 1110st recent times, several new solutions were introduced, including water-borne basecoats with urethane clear coats. Even more recently, water-borne basecoats were combined with powder clear coats. The above short introduction indicates three major trends:
294
Weathering of Plastics
Period
Action
Drivers
1950-1970 development of new technology of coatings
quality
1970-1990 development of clear coat technology
quality, appearance, durability
1990-
envirOlunent, legislation
development of water-borne and powder coatings
The period of 1970-1990 was especially instructive in stressing importance of testing with a special emphasis on weathering testing. During this time, many failures occurred, indicating that both long-term perfonnance predictions and quality control must include weathering testing, considering that failure is very expensive.
PERFORMANCE CONDITIONS Automotive coatings meet variable environmental conditions due to the widespread use of cars in different climatic conditions. Table 2 gives a list of essential parameters.
Table 2. Typical parameters of performance of automotive coatings. Parameter UV radiation
Average value wavelength: 295-380 nm irradiance: 0.35 W/m 2 @340 nm
Maior influences photochemical conversions photooxidation degradation of metallic effect
Telnperature as a function of air -60 -:- 100°C (up to 115°C) telnperature, IR, and color
conlbined degradation activity increased rate of reactions caused by other parameters
Humidity
stress due to thermal movement hydrolysis
10 -:- 100%
non-oxidative photodegradation mar (acid etch) stress due to change of volume Wetness
1-40% total time
extraction hydrolysis penneation to interface
Pollutants and fog
pH of fog as low as 2
surface erosion mar (acid etch) hydrolysis crack initiation
Acid rain (dew)
pH as low as 1 pH of de\v as low as 2
surface erosion lnar (acid etch) hydrolysis crack initiation deposition of salts into clearcoat
295
Automotive Clearcoats
Parameter
Average value
Major influences
Dust particles
widely variable
absorb moisture and acids embedment into clearcoat
Salt (deicing, coastal)
surface etching delatllination corrosion shrinkage
Evaporation of volatile C0111pOnents
surface roughening crack initiation
Pancreatine (bird droppings)
surface etching
MODES OF FAILURE Table 3 gives a list and analysis of modes of failure. Table 3. Modes of failure of automotive coatings in relationship to causes and essential parameters of weathering involved in the failure. Mode of failure Causes Parameters Gloss loss photoxidative processes caused by combination of parameters; correlation UV wavelength (18 months in Florida)38 strongly depends on the control and simulation of conditions of degradation;2o,21 irradiance level initial loss is due evaporation of volatiles 22 (1700 h Xenon arc)38 temperature humidity shrinkage Yello\ving (2500 h Xenon arc i 8 Adhesion loss (2 years in Florida)21
chemical conversion of certain chemical groups in some resins; sonle hardeners UV radiation increase probability;38 more visible with lighter (white basecoat) colors temperature partially attributable to photochemical processes but becomes visible due to UV radiation stress causing delanlination (sources of stress - variable temperature and temperature tnoisture intake) moisture pH
Cracking
see adhesion, water spots, and surface erosion
see adhesion loss
(2 years in Floridai l
Mar (a few months )26
fort11ation of fine scratches due to the environtnental effects (associated defects: UV radiation defonnation and spotting); car washing, in-plant polishing and exposure are precipitation (pH) main causes; typical reasons are photochetnical damage, droplet's swelling, and abrasion solid particle deposits 19,26 H 20 concentration
Water spots
occurs due to deposition of inorganic salts into the surface of clear coats (initial acid rain (pH) UV exposure under dry and cool conditions limits the process );25 fonnation of hydrolysis microscopic blisters and clear coat cracking fonns the so-called defect of UV radiation "unrenlovable water spots,,17
temperature
296
Weathering of Plastics
Mode of failure Surface erosion
Causes Parameters acid rain in combination with dust collection (dust absorbs pollutants) and UV radiation photooxidation; pancreatine related surface damage mostly occurs \vith freshly oxygen produced cars 1 dissolved acids pancreatine
Oil staining
polluted lnotor oil containing carbonaceous products of degradation 38
used oil
Substrate con-osion
loss of barrier properties, transport of ions to interface with nletal
deicing salt salnvater particles
The above list of modes of failures indicates that failure is generally caused by a con1bination of factors which sets the important criteria of testing: • parameters of exposure must precisely imitate conditions of performance • reproduction of conditions depends on the precise control of several parameters (not just UV radiation) • method of exposed specimen testing determines result. The length of time to encounter failure is given as a general infolmation to illustrate premature failures of selected formulations.
CHEMICAL COMPOSITION Automotive coatings are applied for two groups of substrates: metal and plastic. The following diagrams best explain component elements of the coating systems:
Clearcoat Basecoat Primer Electrocoat Phosphate METAL Phosphate Electrocoat
Clearcoat Basecoat Primer PLASTIC
It is easy to predict that the clearcoat must be designed to withstand environmental impact (effect of parameters of performance). For this reason, the emphasis is given to the clearcoat in this report. The general literature l lists currently used clearcoat systems, which include: one-component acrylic-melamine, one-component polyurethane, two-component polyurethane, one-component waterborne, one-component powder. Powder coatings are still on the stage of development and extensive testing thus some data should be updated in future.
Automotive Clearcoats
297
The weathering performance (durability) depends on the chemical composition which must include all components of the mixture since each component, even used in very small quantity, may contribute to photochen1ical degradation. In order to describe composition, recent patents 2- 14 obtained by the major manufacturers of these materials were analyzed to construct a list of individual cOlnponents given below. Components of automotive clearcoats: Polymeric materials:
Powder coatings: 47,48,55,58 • copolymer of methacrylate, ll1ethyl & butyl methacrylate, and styrene with epoxy functionality cured with diacid or uretdione (HDI, IPDI) • polyacrylate polyol (MMA, esters of acrylic & methacrylic acid, styrene) OH group functionality polyester polyol (dialcohol + diacid) cured with aliphatic or cycloaliphatic ketone (ketoxill1e) polyisocyanate or isocyanurate • polyester (hydroxymethacrylate, n-butylacrylate, MMA, neopentyl glycol, and dicarboxylic acid) with OH functionality, polyacrylate containing hydroxyl group cured with HMDI blocked with 1,2,4-triazole • acrylic copolymer (styrene, methacrylic acid, butyl & methyl methacrylates cured with crosslinker of carboxylic groups (epoxides or oxazolines) Solve 11 t-conta i11 i11g: 47-54,56,57,59,60 • acrylic resin OH terminated alkoxysilyl group-containing copolymer (urethane or siloxane bonding) cured by reaction of hydroxyl group from acrylic resin with alkoxysilyl • acrylic resin with OH functional groups cured with aminoplast (condensate of formaldehyde and urea, thiourea or melamine) Resimene 755 from Monsanto or Cymel 1130 (methylate melan1ine-foffi1aldehyde cond.) • acrylic polymer with OH groups microgel based on acrylic cured with aminoplast or polyisocyanate (2-colnponent system) • organosilane polymer (styrene, methacryloxy propyltrimethoxy silane, and trin1ethylcyclohexyl n1ethacrylate) acrylic polyol (styrene, alkyl methacrylate, hydroxy alkyl acrylate) - macrogel urea by reaction of Resimene 755 from Monsanto or Cyn1el 1130 (methylate melamine-foffi1aldehyde cond.) • polyol (caprolactone copolymerized with 1,4-cyclohexanedimethanol) star polymer (ehyleneglycol dimethacrylate, methyle, benzyl and 2-hydroxyethyl methacrylates) cured with isocyanurate or aminoplast (Cymel 1133) • acrylic polymer (styrene, alkyl methacrylate, hydroxy alkyl methacrylate) with OH functionality polyol (caprolactone copolymerized with 1,4-cyclohexanedimethanol) cured with isocyanate (triphenyhnethane triisocyanate or trimer of hexamethylene diisocyanate
298
Weathering of Plastics
• acrylic polymer (styrene, MMA, alkyl methacrylate, alkyl acrylate) crosslinking acrylic (the same but containing glycidyl) • acrylic resin aminoplast (Cymel 1130) • acrylic polymer (hydroxypropyl acrylate, styrene, butyl acrylate, butyl methacrylate, acrylic acid) cured with aminoplast (CymeI 1130) In SUlTIlnary, the following polymeric materials will be analyzed in the section discussing chemical mechanisms of degradation: • acrylic polymers and copolymers • polyurethanes • aminoplasts The importance of this analysis is to include typical chemical groups in order to predict potential products of degradation.
Solvents • • • • • • • • • • • • •
xylene Solvesso 100 n1ethanol butanol, iso-butanol mineral spirits heptane butyl acetate ethyl acetate methyl ethyl ketone acetone dipropyleneglycol monomethylether methyl amyl ketone hexyl acetate
Initiators various initiators used in polymerization of acrylic resins
UV stabilizers • HALS (Tinuvin 144,292) • UV absorbers (Tinuvin 400, 900, 1130)
Catalysts • tin (most frequently DBTL) • amIne
299
Automotive Clearcoats
Flow/rheology modifiers • Perenol F30 and F45 - polyacrylates • Modaflow PIlI - polybutyl acrylate • polydimethyl siloxane oil • Byk 361, 323 & 325 - polyacrylates • BYK 306 - polyether modified dimethyl polysiloxane Other components • fume silica • phosphites
EFFECTS OF PROCESSING Processing effects are given in Table 4.
Table 4. Process parameters, their potential effects, and induced modes of failure. Process parameter Altered composition Production in spring and sumnler
Potential effect Induced mode of failure durability, Quality of finish all modes of failure possible increased acid etch 25 which can be compensated by cracking exposure to UV under dry, cool conditions delamination mar
Reduced rotation speed of spraying random orientation of metal flakes, orange pee1 28 ,36 bells popping, fuzziness, wrinkling, poor gloss28
Residual moisture in the basecoat Dust in plant3o ,43
cracking
craters; cars need to be repainted with different paint potential corrosion (more initiator) faster degradation
Higher temperature of baking 30 Lower film thickness
lower durability cracking
degradation products
35
chromophores
in solvent-base paints shorter life, in powder paints corrosion uneven finish (particle size too close to filnl cracking thickness)
Particle size 35
surface defects
cracking
HUlnidity43
gloss (lower durability)
delanlination
mar Spray gun orientation
43
Paint volume output vs. line speed
43
thickness uniformity
cracking
thickness
corrosion cracking
300
Weathering of Plastics
NUlnerous effects can be induced by the method of processing and precision of equipment operation. At the same time, it should be considered that probabilities of these inconsistencies in production are very low because automotive companies have invested in very sophisticated equiplnent which prevents such artifacts. It is very essential to note that many of these failures are related to film uniformity and that film uniformity can completely change coating perforn1ance in relationship to its durability.61 These effects are discussed further in the next section. Similarly, errors in composition may seem very remote since paint manufacturers are very experienced. At the same time, present coatings (clear coat/base coat) are very risky in automotive applications because of their weathering properties. Previously used coatings deteriorated in a gradual process initiated by a loss of gloss. It was therefore possible to obtain early waluing that particular paint (batch) does not work. In the case of lnodern paints, this warning does not exist, only catastrophic failure (cracking, peeling) suddenly occurs without much detectable difference in perfonnance prior to the failure. Under these circumstances, precise control of coatings prior to their application makes good business practice, considering that in-field failure is very expensive.
MECHANISMS OF FAILURE Many aspects of degradation must be analyzed to reach expected understanding which allows one to pinpoint chemical changes contributing to the modes of failure included in the Table 3 and to find candidate n1ethods which can predict failure. Some of these data can be found in the existing literature l ,17-26,46,61-76 and some mechanisms are still not fully understood. First, we need to analyze the mechanisms of degradation of individual polymers which are used for the production of clear coats as listed above. These polymers include: acrylic polymers and copolymers, polyurethanes, and aminoplasts. The analysis is performed to select the most important reactions which determine durability of automotive coatings. Figure 1 shows typical reactions of acrylic resins. These reactions are ituportant for all three types of resins used in automotive clear coats because they all contain acrylic backbone but differ in the method of chain extension (cure). Acrylic resins are UV stable. They are only degraded because of presence of photoinitiators fron1 impurities. The initial step of photochemical degradation consists of macroradical formation. This first step opens numerous possibilities such as chain scission, crosslinking, formation of hydroperoxides, and formation of carbonyl groups. It is impoliant to mention that there is a general agreement that these changes take place but the kinetics of these changes varies. For example, one research group presents data indicating decrease in carbonyl group formation. 73 In other paper,77 there is an experimental evidence of a reverse trend. This information is very essential to follow degradation because
Automotive Clearcoats
301
Formation of radicals CH3
CH3
I
-
I
CH,- C - CH, -
I
.
COOH
cI
This reaction may lead to formation of hydroperoxides \vhich may
COOH
either decompose producing carbonyl groups or becollle precursors of further chain scission. The examples below show two reactions CH3
CH3
-CH1-~-CHz-fCOOH
that affect molecular weight:
CH3
I I -CH2-~-CH2-1-
+ CH3
+ t:OOH
COOH
COOH
chain scission (molecular weight reduction): -CH.,-~H-CH.,tO~H
-----+
-C H2 - CH =C H2 +
-r.
H
COOH
crosslinking (molecular weight increase)
Formation ofmacroradicals and subsequent reactions affecting molecular weight occur also (in similar sequence of reactions) in esters: -CH.,-CH-
I
COOR
-eo I
-CH2-CH-
or
-CH2-CH-
+
or
+
The above reactions give exarnples offomlation ofmacroradicals which only occur due to abstraction ofa side group with a help ofphotoinitiators rather than by a direct action ofUV itself(bonds involved are UV stable to sunlight radiation). These reactions also show that carbonyl groups are lost in the process of photolytic degradation, although they can be also formed from decomposition of hydroperoxide as represented by the following reaction:
-CH.,-CH-
I
OOH
Figure 1. Typical reactions of acrylic resins.
302
Weathering of Plastics
this is one of the basic measurements. Since there is an agreement among most research groups that correlation (between, for example, natural and laboratory exposures) requires verification ofmechanism, it is difficult to reconcile this statement with the fact that such different estimations of fundamental product of degradation exist. More comments on this subject are included below. Hydroperoxide concentration depends on two competing reactions: oxidation of macroradicals and decomposition of hydroperoxide by heat or UV. Here is one important parameter of weathering - temperature - which plays an essential role in the studies of these materials. Depending on temperature, reaction may take different course. Figure 2 shows another potential anomaly in the course of degradation which occurs when wrong wavelength of light is used in the studies. During such reaction main chain scission occurs which never happens in the outdoor environment. CH3 I
CH3 I
-CH2-C-CH2-CI
0= COCH3
-hv -.
I
COOCH3
Figure 2. Reaction not typical of outdoor exposures of acrylic resins.
In polyurethane clear coats, urethane linkages are formed. Figure 3 shows two potential reactions which may take place at urethane linkage. Both reactions have low probability which is most likely the most important reason for which urethane coatings are used more frequently than aminoplasts. Especially, in regard to acid etching, polyurethanes are superior to other coatings which have either ether or ester linkages. 26 Cleavage ofC-N bond: -O-C-NHII
o
---+
.
o-cII a
Hydrolysis:
Figure 3. Typical reactions of urethane bonding. 62
The mechanism of acid etching of melamine cured systems is given in Figure 4. Presence of water and acid causes hydrolysis of ether linkage which changes molecular weight and thus physico-mechanical properties of coating. 26
Automotive Clearcoats
303
OH
OH
OH
+
HO OH
Figure 4. The mechanism of acid etching. 26
These changes prompted some research groups20,23,25 to conduct extended studies especially in connection with field observations that cars produced during fall or winter have more durable paint than those produced in spring and summer. Figure 5 explains perceived mechanism. If car is painted in winter, the coating cures at dry, cool conditions which ultimately leads to the last compound to the left in the 2 nd row. These changes do not cause a change in molecular weight ofpolymer forming coat. If the hydroperoxide (compound at the right ofthe 2 nd row) is decomposed by UV or heat then changes eventually lead to hydrolysis which weakens coating (last formula at the left of the 3rd row). Similar coating protection can be achieved by controlled exposure of coating to UV. The proposed mechanism helps to understand some problems with melamine coatings. In addition, it indicates importance of other parameters of weathering such as temperature, humidity, and acid rain. In summary, one may observe that some progress was n1ade in qualitative understanding ofchemistry ofautoillotive degradation. At the same time, there is still deficiency in quantitative data - necessary to control mechanisms during an experiment (outdoor, laboratory, or correlation of both).
INTERRELATIONS BETWEEN THE PERFORMANCE CO-NDITIONS, THE MODES OF FAILURE, AND THE CHEMICAL MECHANISMS OF DEGRADATION Table 5 lists these interrelationships for the modes offailure from Table 3, typical parameters of performance from Table 2, and information included in the literature on the mechanisms of chemical degradation in relationship to failure modes.
Weathering of Plastics
304
,
,
/ N
N.J-. N
H
\ I,.. II N~N~N
A.
)
/
hv
-.
\ N
0
~
t
~
/
\
N .... H
N
II
N~N
)
/
o
o
\
\
R
/ N
R
~
N
0
~
~ ·,H N
\A)l
/N
N
N
0
,,0·
A
0
0) \
R
! this branch applies to \vinter production
!
this branch applies to summer production Figure 5. Photooxidation mechanisms ofmelamine. 25
hv or heat
~
Automotive Clearcoats
305
Table 5. Mode of failure versus parameters involved and chemistry of changes. Mode of failure
Parameters involved
Chemical changes involved
Gloss
UV radiation
loss of amide and urethane (PU), loss of ether and triazine not well resolved (n1elamine),23 carbonyl decrease/ 3 carbonyl increase and crosslink scission correlates with hydroperoxide concentration,77
irradiance level
increased irradiance does not always accelerates degradation
temperature
increase in carbonyl & decrease in triazine on temp. increase by lO°C 23
hun1idity
melamine photooxidation rate increase with humidity increasing l8
sulphuric acid shrinkage
loss of anlide (PU), loss triazine (n1elanline)23 loss of volatiles 23
Yellowing
UV radiation temperature
no specific data
Adhesion loss
UV radiation telTIperature moisture pH
oxidation of lower layer (basecoat),18
Cracking
UV radiation temperature
Mar
generally related to photooxidation but no data and correlation with gloss decrease 24 no specific data no specific data see UV radiation
UV radiation
affects crosslinking loss (no specific data)
precipitation (pH)
accumulation of dust particles helps to retain moisture and acid 26 car washing resistance correlates with Taber test 26 and stress-strain19 concentration of water in film depends on hydrophobicity of film 26
acid rain (pH) particle embedding lTIoisture UV radiation
Surface erosion
no specific data no specific data accelerated by cOITlbined action of UV and pH (decreasing) 17
moisture pH
abrasion H 20 absorbed temperature Water spots
no specific data; affected by weathering equipment (QUV different than W_O_M)74
increases water penneability; coating may have higher temp. than T pH affects surface etching rate,17,26 several acids in composition17 no specific data no specific data effect confirnled, 17 no specific data
UV radiation
effect confirmed, 17 no specific data
oxygen
no specific data increases with pH decreasing, 17,26 new paint mostly affected, 1 no specific data
dissolved acids pancreatine
Q
Oil staining
used oil
staining due to carbonaceous materials,38 no specific data
Substrate corrosion
deicing salt
n10st severe cases are due to the loss of environn1ental protection due to the damage of coating: mechanical or photochemical 46
saltwater particles
306
Weathering of Plastics
There are numerous publications available which deal with the subject (36 publications references during 2 two years) and extensive infonnation available on qualitative reasons for automotive coating degradation. The quantitative data are still very scarce. From the above list, it is easy to note that only a few chemical changes can be selected as a base for quantitative measuren1ent of the degradation rate (based on existing data). Gloss change is the most investigated mode offailure and perhaps there is a possibility to select methods of chetnical analysis which may correlate with gloss. At the same time, experts 18 ,24 in the field clearly indicate that gloss loss is not the major problem of clear coat/base coat systems. Moreover, it is indicated24 that good gloss retention cannot preclude catastrophic failure of coating which occurs by peeling and cracking. Frequently, these last two failures are described as "unpredictable". This seems to signalize the nature of the problem of the lack of correlation which is discussed in more detail below. Sitnilar systems are used for coating plastic parts of an automobile. They also include clear coat/base coat system. Several current publications deal with this subject. 42 ,65,67,68 Two directions are taken into consideration: development of directly paintable and adherable polyolefin compounds and preparation of TPO for painting. If the first direction prevails in future (more novel solution) then weathering aspects will be described by similar relationships. If the second n1ethod prevails then preparation method of a surface must be included. These methods include: chemical oxidation, corona discharge, flame treatment, plasma treatment, UV/benzophenone surface degradation, and adhesion promoters. Except for the last method, all methods used affect photodegradation since all these methods induce potential defects which may initiate further degradation which must be accounted for. Finally, the above discussion included only chemistry of degradation. At the same time, it is well known that the structure offilm (unevenness, defects, orange peel, problems offlow, problems with sintering ofpowder coatings) has essential bearing on its degradation. There is no data to report on this matter and thus there is a clear need for extensive research in this area, considering that initial defects in the film surface can alone ruin chances to obtain correlation in experimental work.
SPECIMEN TESTING Some existing standards define testing method which should be used for the evaluation ofexposed specimens. These methods include: • description of changes to appearance7, 11,13 • comparison with origina1 7 • testing according to material specification5 • color change 3,4,13 • glossll,l3
Automotive Clearcoats
307
• physical properties 11 • mechanical properties] 1,13 The above methods are important for the final product evaluation but do not have any predictive value which can be used in the design of weathering experitnent which may help to establish correlation. There are several analytic methods which are used to follow a degradation rate: • FTIR to determine carbonyl, melamine crosslink density, and amide II in PU 77 ,78 • photoinitiation rate (PIR) based on ESR measurement 79 • hydroperoxide tiration80 • surface composition by XPS 39 • orange peel by image analysis 37 All the above methods are suitable and they can eventually contribute to the selection of laboratory weathering conditions. At the same time, the methods have some ilnportant deficiencies. ESR measurement is time-consuming and expensive. FTIR and XPS methods are affected by the surface contamination of a specitnen which is especially important in outdoor exposures. Carbonyl determination does not allow to distinguish between carbonyl loss, due to degradation of carboxyl and ester groups, and carbonyl gain due to hydroperoxide decomposition. Hydroperoxide titration gives reliable data but there are always two competing processes during degradation: hydroperoxide formation and hydroperoxide decomposition. It is therefore difficult to determine extent of photochemical reaction. Fronl the above, a clear need for a further search of chemical analytic method is needed. In addition to the study of selective chemical change there is a need to assess distribution of changes. The so-called "catastrophic (without wanling) failure" clearly indicates that a part of a mechanism of cracking must be related to the changes in crystalline structure which Inakes material increasingly non-uniform to cause sudden (unexpected) crack. Several other opportunities still exist for monitoring the degradation. One is described in the separate chapter of this book. Clearcoats retain their properties due to a high addition of UV stabilizers. Therefore the method of monitoring the concentration of active stabilizers is another useful approach. Recent developments in image analysis allow for simultaneous monitoring of gloss, color change, and surface changes such as formation of haze, microcracking, delamination, etc. This methods tested for sealants applications 81 have proven to be very efficient in durability assessment. If there is a choice between direct determination of defects and indirect (such as chemical analysis) the direct method should always be selected since it provides information on changes directly responsible for perceived failure. The chemical analysis is still very useful because it allows to confirm mechanisms ofchange - useful in remediation of the problem.
308
Weathering of Plastics
EXPECTED LIFETIME For the lack of cOlTelation with studies conducted in laboratory the only requirement used for OEM coatings is that of 5 years Florida exposure without failure. All other methods are still auxiliary techniques used more to accumulate data and experience than as a screening procedure. There is a clear need to develop an expected lifetime in Xenon-arc Weather-a-Meter and EMMAQUA, even ifbased on energy assumption as a starting point. If such standard is not clearly stated (and results not compared with) false expectations regarding laboratory exposures will always exist. Bauer21 recently suggested a new approach to the prediction of durability of a painted car and in view of these considerations such standard is essential.
NATURAL EXPOSURE The precise guidelines can be found in SAE standard. 14 Coating systems are exposed in the panel exposure racks and black boxes. Alten1ative method of outdoor exposure includes the use of solar Fresnel reflector apparatus. 15 Environmental data include: total solar radiation, total UV, optionally selected wavelength radiation, and time of wetness. In Fresnel reflector exposures, it is necessary to determine radiant exposure, elapsed exposure time, black panel temperature, and spray cycle.
LABORATORY EXPOSURE The summary of standardized laboratory methods of exposure is given in Table 1. It can be additionally mentioned that there is an interest in extending laboratory methods to include the effect of acid rain on weathered coatings. Interesting modification of SAE J1960 is re23 ported. Panels were removed for 1h three times a week and sprayed with solution (pH=3.2) of mixture of sulphuric, nitric, and hydrochloric acids in proportions 1:0.3 :0.17.
CORRELATIONS The situation is well characterized by two statements included in Bauer's paper: 18 "Given the complex photodegradation chemistries that occur in these coating systems, a lack of correlation between outdoor exposures and conventional accelerated tests, which employ harsh exposure conditions, is not surprising."
Automotive Clearcoats
309
"It is clear that the predicting free radical photooxidation requires measurement ofboth K (reaction constant) and hydroperoxide concentration." These two statements include several important messages: • harsh exposure conditions • complex photodegradation chemistry • n1easurement Further discussion concentrates on these subjects. It is absolutely certain that the industry needs to accelerate testing. Otherwise, product improvement will be slow. There are two options which can be exploited to achieve this goal: • increase values of quantities involved in photodegradation • find "magnifying glass" The first option was tried for many years. Various equipments and sets of parameters were tested and, since correlation is still not available, failed. Most researchers in the field of durability of materials agree today that acceleration of testing cannot be done by modifying test environment. Also, the reason is clear - complex photodegradation chemistry does not allow to predict what such changes in parameters will affect. It is thus clear that one has to simulate in laboratory conditions typical of natural environment. There is no particular barrier in equipment which would not allow to achieve consistent control of • radiation wavelength • radiation intensity • tenlperature (composite of air temperature, infrared, and specimen color) • humidity The above are the main parameters controlling photodegradation and they can be controlled with precision (see two other chapter in this book on application of different equipment to studies of automotive coatings). The current developments in weathering devices allow one to run any complex program, such as for example, close simulation of seasonal effects. There are two environmental parameters which are not currently simulated in weathering devices. These are stress and pollutants but at the same time there are many methods to include them using the existing equipment (one example was discussed in the previous section). The key to the further development is to find nlethods which allow to verify if the chosen program of weathering conditions allows to follow degradation in the outdoor environnlent. In order to achieve this, the future work should concentrate on the understanding of degradation mechanisms rather than looking for universal new machines for testing as discussed in one recent publication. 24
310
Weathering of Plastics
In order to use the second method (nicknamed "magnifying glass") designed to shorten testing time (or tinle of decision point), two directions of studies are needed: • understanding a chemical mechanism of degradation • establishing a consistent indicator of degradation useful in measuring kinetics of degradation. There were made comments on this subject in the recent paper: 24 "There is little reason to suspect that comparable composition changes should have comparable physical repercussions in coatings from different chemical faluilies." "Weathering tests based on chemical composition change rates provide no information about the physical repercussions of the chemical changes. Therefore, these tests can make no comment on the physical tolerance of clearcoats to the chemical composition changes they undergo, leading to possible erroneous conclusions regarding their durability." The above comments suggest that the fact of detecting a certain concentration of, for example, carbonyl groups does not nlean that a coating, regardless of its formulation, is bound to fail. At the same time, it is possible to observe that the particular coating fails when it attains a particular composition of carbonyl groups, providing that the conditions of degradation (determining the mechanisms of photochemical changes) where the same. This sets the goals for experiment design which may offer correlation: • prior to the experiment the chemical mechanisms of degradation were sufficiently understood to select a measurable quantity which allows to check that mechanisms of degradation are the same in two correlated environments • physical parameters are chosen to have close proximity of exposure conditions • a measurable quantity allows to detect early changes which have been found to signalize particular failure. In automotive coatings, this stage was not reached yet in spite of extensive effort. One reason is that, in most studies, goals too difficult to achieve were set. In the most extensive studies, attempts were made to find universal method, whereas there are no universal changes for, say, polyurethanes and melamine cured acrylics. On a surface, they produce the same hydroperoxides, carbonyls but these concentrations "can make no comment on physical tolerance of clearcoats". There are many examples in automotive coatings which show that focusing on a particular problem helps to solve it. When difference between summer and winter products was observed, the problem was solved as described in Figure 5. When the hydrolysis of aminoplasts was discovered as described by Figure 4, polyurethane coatings gained markets. Many years ago, coatings were degrading because some undesirable solvents
Automotive Clearcoats
311
were used which where then eliminated. UV stabilizers were not giving protection and this problem was eliminated because new stabilizers were introduced to assure their lower volatility during coating baking. The first powder coating was developed long time ag0 55 and it did not perform because today's rheological additives and UV stabilizers were not available to support idea. Present powder coatings are close to the required perfonnance. All these examples show that well focused effort can produce results. The second reason can be related to very rapid changes in automotive coatings which did not allow to stabilize situation. Before any required mechanisms were found, a new range of products was introduced and work had to be repeated. The third reason is related to the fact that too many exploratory research works were needed to scrutinize the test n1ethods which can be useful. It seems that it is a matter of time when proper correlations will be developed. In order for this to happen, the approach must include fundamental analysis of the problem which can be narrowed down to • exposure should simulate conditions found in environment of material performance • verification of these conditions should be established by the use analytical factor which confirms that chemical changes are the same in compared studies • the modes of failures of interest should be related to the chemical changes which can be easily measured.
REFERENCES 1.
2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16.
A. Jurgetz, ~o/fetal Finishing, 93, 10, 53-5 (1995). Plastic tnaterials and coatings for use in or optical parts such as lenses and reflectors of high-intensity discharge forward lighting devices used in motor vehicles, SAE J1647 MAR95. Accelerated exposure of automotive exterior materials using a controlled irradiance water-cooled xenon arc apparatus, SAE J1960 JUN89. Accelerated exposure ofautonlotive exterior nlaterials using a controlled irradiance air-cooled xenon arc apparatus, SAE J2019 JAN94. Accelerated exposure of autonlotive exterior materials using a fluorescent UV and condensation apparatus, SAE J2020 MAY95. Testing the crack resistance of clear coats in two-coat nletallic finishes, VOA 621-4, German Motor Manufacturer's Association. Weather-Ometer test, LP-463 PB-16-0 1, Chrysler Corporation. Condensing humidity resistance test, LP-463PB-9-0 1, Chrysler Corporation. Resistance to artificial weathering, BO 101-1 , Ford. Procedures for laboratory accelerated exposure of automotive tnaterials, GM9125P, General Motors. Weatherability and light resistance test methods for synthetic resin parts, MO 135: 1990, Nissan. Methods of testing for motor vehicle paints. Part 12: Resistance to accelerated weathering. BS AU 148: 1969. Test nlethod of weatherability for automotive parts, JIS 00205. Outdoor weathering of exterior tnaterials, SAE J1976 FEB94. Accelerated exposure of automotive exterior materials using a solar Fresnel Reflector apparatus, SAE J 1961 JUN94. W.J. Putman, M. McGreer, and O. Pekara, STP 1294, ASTM, 1996.
312
17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 50. 51. 52. 53. 54. 55. 56. 57. 58. 59. 60. 61. 62. 63. 64. 65. 66. 67. 68. 69.
Weathering of Plastics
U. Schulz and and P. Trubiroha, STP 1294, ASTM, 1996. D. Bauer, J. Coat. Techno!., 66,835,57 (1994). 1.L. Courter, J Coat. Techno!., 69,866,57 (1997). K.M. Wemstahl and B. Carlsson, J Coat. Technol., 69, 865, 69 (1997). D. Bauer, J Coat. Tec/mol., 69,864,85 (1997). 1.H. Braun and D.P. Cobranchi, J. Coat. Technol., 67, 851, 55 (1995). K.M. Wernstahl, Po(vm. Deg. Stab., 54, 57 (1996). M.E. Nichols, 1.L. Gerlock, and C.A. Slnith, PO~l'm. Deg. Stab., 56, 81 (1997). P.H. Lamers, B.K. Johnston, and W.H. Tyger, Polym. Deg. Stab., 55, 309 (1997). B.V. Gregorovich and 1. Hazan, Prog. Org. Coat., 24, 131 (1994). U. Pitture Vernici, 71, 20, 16 (1995). T. Triplett, Ind. Paint POlvdel~ 71, 7, 20 (1995). E.A. Praschan, ASTM Standardization News, 10,40, 1995. J.M. Bailey, Ind. Paint Powdel~ 71, 7, 14 (1995). 1. Schrantz, Ind. Finishing, 67, 10, 20 (1991). J.M. Bailey, Ind. Finishing, 67, 4, 30 (1991). R. Jaeger and S. Kernaghan, Obeljlaeche/JOT, 35, 9, 42 (1995). Anon., Obeljlaeche/JOT, 36,9, 16 (1996). J.M. Bailey, Ind. Paint POlvder, 70, 12, 10 (1994). J.T. Guthrie and A.P. Weakley, Sillf Coat., Int., 79, 2, 58 (1996). R.T. Quazi, S.N. Bhattacharya, E. Kosior, and R.A. Shanks, Sillf Coat. Int., 79, 2, 63 (1996). H. Schmidt and D. Fink, Sillf Coat. Int., 79, 2, 66 (1996). T.E Barton, D.C.W. Siew, and S.E. Werner, Slllf Coat. Australia, 33, 4,18 (1996). H. Schmidt, Paint Ink Int., 9, 3, 1994. S.L. Kiefer, Paint Ink Int., 8, 12 (1995). E. Lau and D. Edge, ANTEC'93, 2487. B.A. Graves, Products Finishing, 59, 10,48 (1995). B.A. Graves, Products Finishing, 55, 10, 42 (1991). U. Biskup, PUture Vernici, 72, 1, 13 (1996). A. Amirundin and D. Thierry, Prog. Org. Coat., 28, 59 (1996). US Pat. 5,508,337, Bayer Aktiengesellschaft, Gennany, 1996 US Pat. 5,492,955, Bayer Aktiengesellschaft, Gennany, 1996. US Pat. 5,283,084, BASF Corp., USA, 1994. US Pat. 5,354,797, E.!. Du Pont de Nemours and Company, USA, 1994. US Pat. 5,244,696, E.!. Du Pont de Nemours and Company, USA, 1993. US Pat. 5,159,047, E.!. Du Pont de Nen10urs and Company, USA, 1992. US Pat. 4,937,281, E.!. Du Pont de Nemours and Company, USA, 1990. US Pat. 4,402,983, E.r. Du Pont de Nemours and Company, USA, 1983. US Pat. 4,808,656, PPG Industries, Inc., USA, 1989. US Pat. 4,680,204, PPG Industries, Inc., USA, 1987. US Pat. 5,580,660, DSM N.V., Netherlands, 1996. US Pat. 4,728,543, Nippon Paint, Co. Ltd., Japan, 1988. US Pat. 5,063,114, Kanegafuchi Kagaku Kogyo Kabushiki Kaisha, Japan, 1991. M.A. Grolitzer and D.E. Erickson, Waterborne, Higher-solids, and Powder Coatings Symposium, 1994. G. Wypych, Handbook of l\laterial Weathering, 2nd edition, Chem Tec Publishing, Toronto, 1995. U. Schultz and R.-D. Schulze, Oberjlaeche/JOT, 35, 9, 62 (1995). K. Gaszner, M. Heinrich, and T. Schuler, Aluminum, 71, 6, 751 (1995). M. Hartung, H. Hintze-Bruening, and H.-J. Oslowski, Metallobeljlaeche, 50, 6,494 (1996). T. Suzuki, T. Tsujita, and S. Okamoto, Eur. Coat. J, 3, 118 (1996). C. Daniels, Products Finishing, 56, 11,64 (1992). A.S. Wimolkiatisak, A.S. Scheibelhoffer, D. Chundury, and P.M. Mokay, ANTEC'92, 296. L.W. Hill, H.M. Korzeniowski, M. Ojunga-Andrew, and R.C. Wilson, Prog. Org. Coat., 24, 147 (1994).
Automotive Clearcoats 70. 71. 72. 73. 74. 75. 76. 77. 78. 79. 80. 81.
313
G. Linger and E. Hess, S10! Coat. Int., 79, 2, 66 (1996). lL. Gerlock, W. Tang, t\-1.A. Dearth, and TJ. Komiski, Polym. Deg. Stab., 48,121 (1995). J.L. Gerlock, T.J. Prater, S.L. Kaberline, and J.E. deVries, Polym. Deg. Stab., 47,405 (1995). N.S. Allen, MJ. Parker, CJ. Regan, R.B. McInture, and W.A.E. Dunk, Polym. Deg. Stab., 47, 117 (1995). C. Gopsill and P.W. Griggs, Sla! Coat. Int., 76, 6,247 (1993). D.S. Allan, N.L. Maecker, D.B. Priddy, and NJ. Schrock, Macromolecules, 27, 7621 (1994). B.L. Rytov, V.B. Ivanov, V.V. Ivanov, and V.M. Anisimov, Polymel~ 37, 25, 5695 (1996). D.R. Bauer, D.F. Mielewski, and lL. Gerlock, Polym. Deg. Stab., 38,57 (1992). D.R. Bauer, lL. Gerlock, and D.F. Mielewski, Po~vnl. Deg. Stab., 27, 271 (1990). lL. Gerlock, D.F. Mielewski, and D.R. Bauer, Polym. Deg. Stab., 20,123 (1988). D.F. Mielewski, D.R. Bauer, and lL. Gerlock, Polym. Deg. Stab., 33, 93 (1991). G. Wypych, F. Lee, B. Pourdeyhimi, Comparative study of sealants durability. Surface changes, RILEM Symposium 2000.
Index
A ABS 61 accelerated electrons 178 accelerated tests 10 acceleration factor 12 acid etching 303 acid rain 162 acids 218 acrylic-melamine 296 acrylics 297 activation energy 173 additives 253 agriculture 218, 225 alkalinity 217 alkyd-lnelalnine 293 alninoplasts 298 alnorphous 77, 183, 199 antagonistic interaction 225 antioxidants 169, 179, 225, 234 antistatics 5 appliances 99 Arizona 5, 19,72,77,93 Arrhenius activation energy 169 equation 171 plot 180 ASA61 AIR 78,186 autoillotive 29, 43, 72, 161, 185, 241 autolnotive coatings 293
B bags 211 basecoats 293
benchtop instrulnents 10 bioburden 178 biocides 5 blends 212 blistering 151 bond breaking energy 62 cleavage 70 bottles 211 Brabender 212 branching 81 brittle layer 178 brittleness 63 buildings 15, 70, 133
C calciuln stearate 84 carbon black 127 carbon fiber 99 carbon-arc 7, 16 carbonyl groups 78, 228, 261, 301 catastrophic failure 307 chain cleavage 78, 215 scission 141, 170, 301 chenlical resistance 237 chelni-crystallization 141, 149 chemiluminescence 170 chromophores 97, 228 ClRA8 clearcoat 185, 296 clilnates 17, 261 coatings 186 coextrusion 93, 271 color 61, 96, 134
Index
316 compatibilization 212 cOlnpressive stress 155 COlnpton secondary electrons 178 condensation 5 conductive cooling 137 construction 161 Inaterials 1 consumer goods 161 containers 211 corrosion 151 CPE 61 CPVC 61 crack depth 182 cracking 151 cracks 102 crosslinking 301 crosslinks 267 crystalline regions 199 crystallinity 141, 197, 214 change 142 crystallite size 197 crystallites 170 crystallization temperature 213
D dark colors 138 daylight irradiance 4 debonding 102 degradation mechanism 10 - 11 rate 1 telnperature 214 degrading parameters 11 delalnination 151, 307 design life 2 diffraction 197 diffusion 229, 253 DIN 113 discoloration 162 disposal 2 DSC 169, 177
durability 169 durability testing 2
E early degradation 218 elongation 70, 230, 233 elnbrittlelnent 173, 233 end-groups 77 EPDM 170 epoxy resin 99 EPR 170,178 equatorial tracking 24 equipment 7, 50, 105 ESR 186,307 ethylene-propylene CopolYlner 262
F factors of aging 69 failure 1, 2, 178, 188, 295 criteria 137 fiber 233 fiber-reinforced plastics 99 fibrils 173, 182 films 180, 218, 225 filter system 34 flaking 151 flame retardants 71, 238 Florida 5, 18 - 19, 72, 93, 236, 273, 308 fluorescent lamps 7, 16 Fresnel-reflecting mirrors 16 FTIR83, 186,212,229,286 fuel combustion products 162
G gas fading 162, 248 Gaussian distribution 199 GC 282 glass fiber 103, 195 glass transition 183 gloss 306 GPe 78,94,212,263
317
Index
greenhouses 225, 271
H HALS 161, 164, 185,225,233 deactivation 222 haze 96,307 heat buildup 133 deflection telnperature 136 HIPS 61 HMDI297 hot water 99 HPLC 177 hUlnidity 309 hydrolysis 5, 61 hydroperoxide titration 186 hydroperoxides 170,227,268, 300
I imnlersion 155 tests 99 il11pact strength 61, 233 inductive coupled plasllla 84 infrared energy 4 heating 121 initiation 227 injection molding 141 installation 2 interface 103, 170 international organizations 15 IPDI297 IR 107 reflective piglnents 134 irradiance 3 - 4, 47, 95 irradiation 105 ISO 113 isocyanurate 297
L latitude 22
layer rellloval procedure 141 lead stabilizer 277 life prediction 15 light penetration 203 lightfastness 29 long-tenn data 10 Lorentz 199 low-volatility 272
M macroradicals 301 maintenance 2 marine coatings 151 Inaterial degradation paralneters 2 mathematical weighing process 135 Inatrix 102 Ineasureillent geometry 135 Inechanical stress 83 melaluine 304 Inelting telnperature 213 nletallocene 69 methane cOlnbustion 162 luethods of measureillent 12 microcracking 307 Inicrocracks 233, 274 migration 72 milling 262 111iniblinds 277 M n 79 l110de of failure 306 lTIodulus 141 lTIoisture 5 monoclinic fonl1 207 Mz/M n 81
N nanoconlposites 196 NMR 186,212 Norrish 206 NO x 162
318
o OIT 69, 169, 177 optical properties 134 outdoor exposure 15 oxidation 78 oxygen 261 diffusion 182
p paint 185 parafocus 199 PAS 186 performance criteria 10 - 11 pesticides 225 pH 177 phenolic antioxidants 161 photohydrolysis 97 photoinitiators 300 photons 178 photo-oxidation 141 pigtnents 127, 134,237, 265 pKa-value 217 plasticizers 5, 286 plastics 1 PMMA61 polarization 199 polishing 200 pollutants 5 polyacrylate 297 polyalnide 5, 61 polybutyleneterephthalate 172, 195 polycarbonate 3, 5, 271 polyester 5, 170, 297 polyethers 61 polyethylene 61, 170, 178, 211, 218, 225, 261, 286 polyethyleneterephthalate 77, 93 polyol297 polyolefins 211, 21 7 polyoxymethylene 99, 129 polyphenylene sulfide 282
Index
polyphenyleneether 99 polyphenylenesulfide 99 polypropylene 61, 71, 144, 170, 177, 211, 233, 261 polystyrene 56, 61 polyurethane 39, 296 polyvinylchloride 61, 83, 107, 277, 286 post application 2 post-irradiation 169 failure 184 powder coating 296 prediction tnethods 174 preliminary studies 10, 13 product temperature 4 propagation 182 pUlnps 99
Q quenchers 70 QUV 84, 262, 277
R radiant energy 133 radiation 7, 177 global 107 simulation 46 radical trapping 71 radicals 77, 170, 181, 231, 268 Ralnan 186 reaction kinetics 108 reactive chetnicals 218 recycling 211 reference Inaterial 35 reflectance 135 relaxation 141 relief melnbrane 283 renewable resources 2 replacen1ent 2 residual stress 141,147 - 148, 151 rigid vinyl 136 routine testing 11
319
Index
S salt formation 218 scattering 199 scission 263 secondary crystallization 141 SEM 99, 173, 182,284 sensors 99 shelf-life 169 shrinkage 151 siding 83 skylights 271 slnog 162 solar absorptance 136 solar cut-on 3 - 4 solar spectfUln 133 - 134 solidification 141 solubility 253 spectral irradiance 106 spectfUln 8 spinning stability 221 stabilization 70 stabilizers 233, 267 package 225 standard reference Inaterial43 standards 49 sterilization 177 strain 143 stress 6, 143 build up 154 distribution 141 intensity 183 surface crazing 233 swelling 141 switch 282
test fixtures 21 testing conditions 69 textiles 1 Tg 109 TGA 213 thennal conductivity 48 thennolneters 112 thickness 156 through-translnission 121 tie molecules 170 tilt angles 21 titning 24 titaniuln dioxide 84 toys 161 TPO 241 transit shelters 271 translucent polymers 124 transmittance 135, 222 transparent polYlners 124
U unsaturations 215, 228 urethane 302 UV absorbers 70, 77, 164, 185, 233 cycling 265 radiation 3 spectra 228 spectroscopy 186 stabilizers 5
V valves 99 viscosity 275 volatility 253
T temperature 4, 8,49,72, 105, 124, 169, 171,309 gradient 141 tensile 233 strength 12,73, 99, 230, 233 terrestrial body 106
W walkways 271 warping 146 water uptake 155 waterborne 296
320 wavelength 3, 135 Weather-Ometer 9, 72, 94, 242 Wien's Law 127 welding 127 window profile 83
X xenon-arc 7, 16, 43 Xenotest 30, 228 X-ray 197
y yellowness index 273 yield point 216 Young's modulus 146, 151
Z zinc stearate 84
Index
