Advances in Photochemistry, Volume 29

ADVANCES IN PHOTOCHEMISTRY Volume 29 Editors DOUGLAS C. NECKERS Center for Photochemical Sciences, Bowling Green State ...

Author: Douglas C. Neckers | William S. Jenks | Thomas Wolff

54 downloads 809 Views 4MB Size Report

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!
Report copyright / DMCA form

DOWNLOAD PDF

ADVANCES IN PHOTOCHEMISTRY Volume 29 Editors

DOUGLAS C. NECKERS Center for Photochemical Sciences, Bowling Green State University, Bowling Green, Ohio

WILLIAM S. JENKS Department of Chemistry, Iowa State University, Ames, Iowa

THOMAS WOLFF Technische Universita¨t Dresden, Institut fu¨r Physikalische Chimie und Elektrochimie, Dresden, Germany

WILEY-INTERSCIENCE A JOHN WILEY & SONS, INC., PUBLICATION

ADVANCES IN PHOTOCHEMISTRY Volume 29

ADVANCES IN PHOTOCHEMISTRY Volume 29 Editors

DOUGLAS C. NECKERS Center for Photochemical Sciences, Bowling Green State University, Bowling Green, Ohio

WILLIAM S. JENKS Department of Chemistry, Iowa State University, Ames, Iowa

THOMAS WOLFF Technische Universita¨t Dresden, Institut fu¨r Physikalische Chimie und Elektrochimie, Dresden, Germany

WILEY-INTERSCIENCE A JOHN WILEY & SONS, INC., PUBLICATION

Copyright # 2007 by John Wiley & Sons, Inc. All rights reserved Published by John Wiley & Sons, Inc., Hoboken, New Jersey Published simultaneously in Canada No part of this publication may be reproduced, stored in a retrieval system, or transmitted in any form or by any means, electronic, mechanical, photocopying, recording, scanning, or otherwise, except as permitted under Section 107 or 108 of the 1976 United States Copyright Act, without either the prior written permission of the Publisher, or authorization through payment of the appropriate per-copy fee to the Copyright Clearance Center, Inc., 222 Rosewood Drive, Danvers, MA 01923, (978) 750-8400, fax (978) 750-4470, or on the web at www.copyright.com. Requests to the Publisher for permission should be addressed to the Permissions Department, John Wiley & Sons, Inc., 111 River Street, Hoboken, NJ 07030, (201) 748-6011, fax (201) 748-6008, or online at http:// www.wiley.com/go/permission. Limit of Liability/Disclaimer of Warranty: While the publisher and author have used their best efforts in preparing this book, they make no representations or warranties with respect to the accuracy or completeness of the contents of this book and specifically disclaim any implied warranties of merchantability or fitness for a particular purpose. No warranty may be created or extended by sales representatives or written sales materials. The advice and strategies contained herein may not be suitable for your situation. You should consult with a professional where appropriate. Neither the publisher nor author shall be liable for any loss of profit or any other commercial damages, including but not limited to special, incidental, consequential, or other damages. For general information on our other products and services or for technical support, please contact our Customer Care Department within the United States at (800) 762-2974, outside the United States at (317) 572-3993 or fax (317) 572-4002. Wiley also publishes its books in a variety of electronic formats. Some content that appears in print may not be available in electronic formats For more information about Wiley products, visit our web site at www.wiley.com. Library of Congress Cataloging-in-Publication Data: Library of Congress Catalog Card Number: 63-13592 ISBN 13: 978-0-471-68240-0 ISBN 10: 0-471-68240-3

Printed in the United States of America 10 9

8 7 6

5 4 3 2

1

CONTRIBUTORS

F. C. De Schryver KULeuven Department of Chemistry Celestijnenelaan 200F Heverlee B-3001 Belgium Mamoru Fujitsuka The Institute of Scientific and Industrial Research Osaka University Mihogaoka 8-1 Ibaraki, Osaka 567-0047 Japan J. Hofkens KULeuven Department of Chemistry Celestijnenelaan 200F Heverlee B-3001 Belgium M. Lor KULeuven Department of Chemistry Celestijnenelaan 200F Heverlee B-3001 Belgium

Tetsuro Majima The Institute of Scientific and Industrial Research Osaka University Mihogaoka 8-1 Ibaraki, Osaka 567-0047 Japan G. Schweitzer KULeuven Department of Chemistry Celestijnenelaan 200F Heverlee B-3001 Belgium Bernd Strehmel Kodak Polychrome Graphics Research and Development Division An der Bahn 80 D-37520 Osterode Germany Veronika Strehmel University of Potsdam Applied Polymer Chemistry Karl-Liebknecht Str. 24/25 D-14476 Golm Germany

v

vi

CONTRIBUTORS

M. van der Auweraer KULeuven Department of Chemistry Celestijnenelaan 200F Heverlee B-3001 Belgium

PREFACE

Volume 1 of Advances in Photochemistry appeared in 1963. The stated purpose of the series was to explore the frontiers of photochemistry through the eyes of experts and pioneers. As editors we have solicited articles from scientists who have strong identifications with the work presented, and strong points of view. Photochemistry has expanded enormously since those first days. A serious percentage of the papers in any single volume of the Journal of the American Chemical Society, for instance, can rightly fall in its purview. The emergence of the laser and the evolution of theoretical methods strongly influenced research. With new computational methodology almost no intermediate lives too short a time to be detected and its dynamics characterized. The fundamental objective of our field, elucidation of the history of a molecule that absorbs radiation, is now within reach in even the most complicated cases. We hope that the series continues to reflect the frontiers of photochemistry as it evolves into the future. We report, sadly, that one of the founding Editors, George S. Hammond, passed away in Portland, Oregon on October 5, 2005. George will be sorely missed. With the publication of this volume, Douglas Neckers will be leaving the post of Senior Editor. He has served, first as Associate Editor and more recently as Editor, for nearly half of the volumes in the Advances series. He wishes to express his appreciation for all of the cooperation he has received from everyone involved in the series. He will remain as a consultant, and Pavel Anzenbacher will take over as Editor beginning with Volume 30. Bowling Green, Ohio, USA Dresden, Germany Ames, Iowa, USA

Douglas C. Neckers Thomas Wolff William S. Jenks

vii

CONTENTS

Ensemble Photophysics of Rigid Polyphenylene Based Dendritic Structures M. LOR, G. SCHWEITZER, M. VAN DER AUWERAER, J. HOFKENS, AND F. C. DE SCHRYVER Photochemistry of Short-Lived Species Using Multibeam Irradiation MAMORU FUJITSUKA AND TETSURO MAJIMA

1

53

Two-Photon Physical, Organic, and Polymer Chemistry: Theory, Techniques, Chromophore Design, and Applications BERND STREHMEL AND VERONIKA STREHMEL

111

Index

355

Cumulative Index Volumes 1–29

379

ix

ENSEMBLE PHOTOPHYSICS OF RIGID POLYPHENYLENE BASED DENDRITIC STRUCTURES M. Lor, G. Schweitzer, M. van der Auweraer, J. Hofkens, and F. C. De Schryver KULeuven Department of Chemistry, Celestijnenelaan 200F, Heverlee B-3001, Belgium

CONTENTS I. II. III. IV.

V. VI.

Introduction Electronic Excitation Transfer Stationary Measurements Single-Photon Timing Measurements A. Time-Resolved Fluorescence Measurements Performed Under Magic Angle Polarization Condition 1. Para-substituted Carbon Core Dendrimers 2. Meta-substituted First Generation Carbon Core Dendrimers B. Time-Resolved Fluorescence Polarization Measurements 1. Meta-substituted First Generation Carbon Core Dendrimers 2. Para-substituted First Generation Carbon Core Dendrimers Femtosecond Fluorescence Upconversion Measurements Femtosecond Transient Absorption Measurements A. p-C1P1 and m-C1P1 B. p-C1P3 and m-C1P3

Advances in Photochemistry, Volume 29 Edited by Douglas C. Neckers, William S. Jenks, and Thomas Wolff # 2007 John Wiley & Sons, Inc.

1

2

ENSEMBLE PHOTOPHYSICS OF POLYPHENYLENE BASED STRUCTURES

C. m-C1P3 D. p-C2P1 and p-C2P4 VII. Conclusions Acknowledgments References

I. INTRODUCTION Excited state processes in multichromophoric systems have attracted attention for a long time [1], since these processes are of great importance in biological and material science. Indeed, in the light harvesting system as well as in polyconjugated polymers, multiple chromophores are present and the efficiency of the system in the energy cascade to the reaction center or in the efficiency of the charge generation is influenced by excitation and electron transport along the multichromophoric system. Because of the controllable incorporation of various functional groups in different parts of their structure, dendrimers have attracted much attention recently as model systems for the study of photoinduced intramolecular energy and electron transfer. Dendrimers can act as scaffolds that tether the donor and acceptor chromophores [2], providing versatility such that additional features can easily be introduced by simply changing the various components of the dendrimer. Alternatively, the dendrimer backbone itself can concurrently be used as the energy donor or acceptor. Several types of chromophoric dendrimer backbones such as poly(phenylacetylene) [3], poly(phenylene) [4], and poly(benzylether) [5] have been used as light absorbers, and the energy was efficiently transferred to the core acceptor. While most of these systems have high energy transfer efficiencies, they still suffer from a weak fluorescence or a low fluorescence quantum yield. However, polyphenylene dendrimers composed of tens or hundreds of out-of-plane twisted phenyl units can be used as chromophoric backbones [6] carrying highly luminescent dyes at the periphery. The earliest work on intramolecular energy transfer in dendritic macromolecules originates from Moore and co-workers [7], who synthesized dendritic structures based on phenylacetylene units with perylene in the center. The excitation of the phenylacetylene units at the rim at a wavelength of 310 nm leads to fluorescence emitted by the center perylene unit, indicating intramolecular excitation energy transfer. A significant increase in the rate of excitation energy transfer was achieved by modifying the dendrimer skeleton. This was done in such a way that additional phenylacetylene units with lower excited state energy and larger conjugation length toward the core were introduced near the perylene unit. Recently, Bardeen and co-workers

3

INTRODUCTION

1-H:

R¢ R

1-TMS:

R = H, R¢ = H R = Si(CH3)3, R¢ = H

1-Ph:

R = 3,5-di-t-butylphenyl, R¢ = t-butyl

R¢ R R

R

R

R 2-H:

R=H

2-TMS: R = Si(CH3)3 2-Ph:

R = 3,5-di-t-butylphenyl

3-H:

R=H

3-TMS:

R = Si(CH3)3

3-Ph:

R = 3,5-di-t-butylphenyl

Figure 1.1. Building blocks of phenylacetylene dendrimers studied by Bardeen and co-workers [8].

reported the role of Fo¨rster, Dexter, and charge transfer interactions in phenylacetylene dendrimers [8]. They demonstrated by steady state spectroscopy, picosecond time-resolved emission and anisotropy measurements, and ab initio calculations that while the subunits of polyacetylene dendrimers (Fig. 1.1) are weakly coupled in their equilibrium ground state geometry, they can become strongly coupled in the excited state. This geometry-dependent electronic coupling will affect the modeling of energy transfer in these molecules. They found that the variation of the electronic coupling V with molecular geometry is due to the throughbond or charge transfer type of interaction rather than due to variation of the more familiar dipole–dipole and Dexter terms. These dendritic structures are rigid systems in which the branches are also the absorbers and the Bardeen study underlines the complexity of these systems in terms of excitation transfer. Most of the dendritic molecules investigated for excitation transfer between chromophores attached at the periphery belong to a class in which the arms are rather flexible. This of course leads to data related to excitation transfer, which are averaged over all the possible branch conformations leading to a distribution in distances between donor and acceptor. Balzani et al. [9] reported metal-containing dendrimers, where the core and branching unit are built up from ruthenium complexes of a polypyridine

4


O

O

O

O

O O

O

O

O

O

O

O

O

O O

N

N

O

O

O Ru

N

2+

N

O

N

O O

N

O O O

O

O

O

O

O

O

O

O O

O

O

O

Figure 1.2. Molecular structure of a metal-containing dendrimer investigated by Balzani and co-workers [10].

ligand serving as core and branching units. By varying the ligands and metals used, different directional excitation energy transfer processes were observed, either from the center to the rim or from the rim to the core [10]. The molecular structure of such a dendrimer with a ruthenium complex in the center is depicted in Figure 1.2. Recently, Balzani and co-workers published results on dendrimers consisting of a benzophenone core and branches containing four and eight naphthalene units (Fig. 1.3) [11]. In both dendrimers, excitation of the peripheral naphthalene units is followed by fast singlet–singlet energy transfer to the benzophenone core; but on a longer time scale a back energy transfer takes place from the triplet state of the benzophenone core to the triplet state of the

5

INTRODUCTION

O

O

O

(a)

O O O

O

O

O

O

(b) O

O

O O

O

O

O O O

O

O O

Figure 1.3. Molecular structures of dendrimers with 4 (a) and 8 (b) peripheral naphthalene units and a benzophenone core investigated by Balzani and co-workers [11].

peripheral naphthalene units. Selective excitation of the benzophenone unit is followed by intersystem crossing and triplet–triplet energy transfer to the peripheral naphthalene units, which could be observed by nanosecond transient absorption. Using a similar type of branch, developed by the Frećhet group, they have published extensively on chromophore labeled dendrimers [12]. The dendrimers possessing coumarin-2 dyes at the periphery, and either coumarin-343 (Fig. 1.4) or a heptathiophene dye at the core, were studied by time-resolved fluorescence and transient absorption spectroscopy. It was revealed that upon excitation of the rim chromophores almost no direct fluorescence occurred from these initially excited chromophores. Instead, only the center

6

ENSEMBLE PHOTOPHYSICS OF POLYPHENYLENE BASED STRUCTURES O

O O

O O

O O

N

O

N N N O O O

O

O

O

N O

N

O O

O

N O O O

O

N O

O

O O

N

Figure 1.4. Molecular structure of a third generation dendrimer with coumarin-343 at the center investigated by Frećhet and co-workers [12].

chromophore showed emission; thus proving efficient excitation energy transfer within this dendrimer. The efficiency of the excitation energy transfer decreased by increasing the generation number from 3 to 4. This comes from the fact that increasing the generation number increases the average distance between the chromophores and thus the overall efficiency of excitation energy transfer decreases. Recently, Frećhet and co-workers reported intramolecular energy transfer in dendritic systems containing one or more two-photon absorbing chromophores at the periphery, which act as energy donors, and a Nile Red chromophore at the core that acts as energy acceptor as well as fluorescence emitter [13]. The two-photon energy absorbed by the chromophores at the periphery was transfered to the core, where the core’s emission was strongly enhanced. The emission from the core chromophore in these dendritic systems was significantly greater than the emission from the core itself when the core was not connected to the donor chromophores. This increased emission arises from the much larger two-photon absorbing cross section of the donor chromophores compared to the core acceptor at the excitation wavelength.

INTRODUCTION

7

N

Figure 1.5. Molecular structure of a second generation triaryl dendrimer investigated by Goodson and co-workers [15].

Meijer and co-workers investigated the dynamics of excitation energy transfer for a series of spherical porphyrin arrays based on different generations of poly(propylene-imine) dendrimers using time-resolved fluorescence anisotropy measurements in a glass environment [14]. They demonstrated that the multiporphyrin functionalized dendrimers were able to absorb light and efficiently distribute the excitation energy by hopping over the chromophore arrays with minimal loss during the energy migration process. Goodson and co-workers investigated excitation energy transfer processes in nitrogen cored distyrylbenzene and triarylamine dendrimer systems (Fig. 1.5) by photon echo and polarized fluorescence upconversion spectroscopy. Observed components of less than 1 ps were attributed to a coherent energy transport mechanism. The contributions from his group were recently summarized [15]. De Cola and co-workers recently published [16] a study of the photophysical properties of a molecular system consisting of a bay-functionalized perylene

8


O

O O

O

O O

O O

N

O

O

N

N

O

N

O O O

O O

O O

Figure 1.6. Molecular structure of a bichromophoric pyrene–perylene bisimide system investigated by De Cola and co-workers [16].

bisimide, containing four appended pyrene and two coordinating pyridine units (Fig. 1.6) using steady state, time-resolved emission and femtosecond transient absorption spectroscopy. Analysis of the data showed the presence of a fast intramolecular photoinduced energy transfer process from pyrene*–perylene to pyrene–perylene* (ken 6:2 109 s1 ) with a high yield (>90%), followed by efficient intramolecular electron transfer from pyrene–perylene* to pyrene.þ–perylene. (70%, ket 6:6 109 s1 ). Both processes occur from the pyrene unit to the perylene moiety. The Fo¨rster distance was calculated to be 3.4 nm and the corresponding donor–acceptor distance was calculated from the energy transfer rate as 0.9 nm. No indications for energy hopping between different pyrene moieties were observed. Similarly, a number of terrylenediimide core dendrimers with semiflexible arms were investigated by our research group at the ensemble [17] and at the single molecule level [18]. Different generations of a polyphenyl dendrimer containing a terrylenediimide core with peryleneimide chromophores at

9

INTRODUCTION

O O

N

N

O

O O

N

O

O

O

O

O

O

N

O

O N

O N

O O

Figure 1.7. First generation polyphenylene dendrimer with terrylene as a luminescent core.

the periphery (first generation depicted in Fig. 1.7) have been studied with respect to intramolecular energy transfer processes. Excitation of the peryleneimide at 480 nm resulted in fluorescence of the terrylenediimide chromophore at 700 nm with an almost complete disappearance of the fluorescence of the peryleneimide chromophore at 550 nm, indicating a very efficient energy transfer process between the peryleneimide and terrylenediimide chromophore. Single molecule data measured at room temperature indicated that a distri bution of excitation transfer rate constants could be observed [18], while Basche´ and co-workers [19] showed, studying linewidths at low temperature, that the observed rates are larger than expected from classical Fo¨rster excitation transfer theory and suggested that in these systems through-bond interaction might play a role. Similarly, phenylacetylene based dendrimers [7, 8] and those investigated by Goodson and co-workers [15] show substantial coupling between the branches while all others discussed above, due to flexibility of the connecting arms, have an undefined three-dimensional structure and hence variable donor– acceptor distances.

10

ENSEMBLE PHOTOPHYSICS OF POLYPHENYLENE BASED STRUCTURES R1 R1

O N O

R2

O

C

C

N

R2

O

R3

R3

O N O

p -C1Px

m -C1Px

PI 1

p-C1P1 p-C1P3 p-C1P4

R H PI PI

2

R H PI PI

1

3

R H H PI

m-C1P1 m-C1P2 m-C1P3 m-C1P4

R H PI PI PI

2

R H H PI PI

3

R H H H PI

Figure 1.8. Molecular structures of p-C1Px ðx ¼ 1; 2; 4Þ, para-substituted first generation dendrimers, and m-C1Px ðx ¼ 1; 2; 3; 4Þ, meta-substituted first generation dendrimers; PI, peryleneimide chromophore.

In the present contribution we want to focus on rigid dendritic structures in which the coupling between the chromophores is weak and in which the distance between the chromophores involved is fixed in space. To achieve this goal together with the Mu¨llen group (MPI Mainz), a series of molecules was developed based on the general structure in Figure 1.8. Besides these first generation dendrimers, second generation dendritic structures p-C2Pn were also investigated (p-C2P1, p-C2P2, p-C2P3, p-C2P4) (see Fig. 1.9).

II.

ELECTRONIC EXCITATION TRANSFER

One of the basic mechanisms in multichromophoric systems, electronic excitation transfer has been in the past and still is in many studies largely described using Fo¨rster theory. As stated by Fo¨rster [20], this model is developed for the weak coupling limit as it is based on an equilibrium Fermi Golden Rule

11


O

N

O

O O

N

C

O

N O

O

N

O

Figure 1.9. Molecular structure of p-C2P4.

approach and the derived Fo¨rster equation is valid provided a number of conditions are fullfilled as recently discussed by Scholes [21]: ‘‘(a) A dipole–dipole (or convergent multipole–multipole) approximation for the electronic coupling can be employed appropriately for the donor–acceptor interaction. (b) Neither the donor fluorescence lifetime, emission line shape, acceptor absorption line shape, nor oscillator strength is perturbed because of interactions among donors or acceptors, respectively. (c) Static disorder (inhomogeneous line broadening) is absent in the donor and acceptor line shapes. (d) the energy transfer dynamics are incoherent.’’ Different complicating factors led to the development of a more generalized approach [22, 23] in which the Coulomb interaction is now considered in terms of local interactions between donor and acceptor transition densities. This is

12


(a)

(c) (b)

Figure 1.10. (a) Chemical structure of p-C1P1. (b) Two-dimensional (2D) representation of where chromophores can be attached to the dendrimer and (c) three-dimensional (3D) representation of isomer 2A2B of p-C1P2. The arrows indicate the possible substitution patterns.

particularly important when the donor–acceptor ‘‘chromophores’’ are large compared to their center-to-center separation. To verify if the above-mentioned boundary conditions are valid for the molecular structures reported in Figure 1.8, electronic coupling constants were calculated. Doing this, one needs to take into account that, as a result of the asymmetric building blocks used in the Diels–Alder cycloaddition in the course of the reaction, the attachment of the chromophores leads to structural isomers. Therefore, if multiple chromophores are present, small differences can occur in the efficiencies of photophysical properties among different isomers. An example of possible structural isomers (2D picture) and one example of a 3D isomer of p-C1P2 (2A2B) are given in Figure 1.10. As depicted in Figure 1.10b, there are four attachment places for the chromophores and this normally results in four possible isomers for p-C1P2. However, there is an asymmetry in the four polyphenyl branches resulting in two possible ways in which the two chromophores can be attached. The arrows indicate the possible substitution patterns of the chromophores in the different structural isomers. The positions where a chromophore can be attached are A2, A3, B2, B3, C2, C3, D2, and D3, where A, B, C, and D represent the different branches and 2 and 3 the second or third phenyl group within each branch where a chromophore can be attached. For p-C1P1, however, the two different structural isomers that can be formed will show similar photophysical behavior. Also, for p-C1P4 there are a number of possible structural isomers as can be seen in Figure 1.11a, b. These two


13

Figure 1.11. (a, b) Two-dimensional representation of the two structural isomers p-C1P4A and p-C1P4B of p-C1P4. ða0 ; b0 Þ Three-dimensional representation of the two structural isomers of p-C1P4.

minimized structures were obtained using a molecular mechanics optimization method (Merck molecular force field) present in SPARTAN1. Geometry optimization of p-C1P4A shows that the center-to-center distance between the chromophores is on average 3.17 nm. For two structural isomers of this compound (p-C1P4A and isomer p-C1P4B, respectively), the difference in interaction between the chromophores in each isomer was calculated. Calculations of the electronic transitions of the two depicted structural isomers of p-C1P4 were done by using the CEO-INDO/S procedure [24]. Besides revealing the energy of the electronic transitions, this method allows for the calculation of the electronic coupling constants between the transition dipole moments of the chromophores. All reported values apply to a molecule in vacuum at 0 K. CEO calculations were performed on two isomers (p-C1P4A and p-C1P4B) of p-C1P4,

14


obtained by energy minimization (see Fig. 1.11a and 1.11b). The results of the CEO calculations on both isomers show an average value for this coupling of the chromophores of p-C1P4A to be 22.6 cm1. In p-C1P4B, the average distance between the chromophores is 3.3 nm except for pair 1–4, where the distance is only 1.7 nm. The average value for the coupling constants is 21.22 cm1, except for pair 1–4 for which a value of 62.6 cm1 is obtained. However, one needs to take into account that all the calculations are done assuming a temperature of 0 K, and hence at room temperature these couplings will be minimal. Furthermore, in collaboration with Beljonne and co-workers, transition densities were calculated [25] for excitation transfer between two peryleneimide chromophores coupled by a fluorene trimer (separation 3.4 nm) and found to be in line with the Fo¨rster approximations.

III.

STATIONARY MEASUREMENTS

The steady state absorption and fluorescence spectra of all first generation dendrimers in toluene are depicted in Figure 1.12. Within experimental error, the former ones are identical for all compounds. In the emission spectra, however,

Figure 1.12. Steady state absorption and emission spectra of the first generation dendrimers in toluene: p-C1P1, p-C1P3, (solid lines,—), m-C1P1 (short dashes, - - -), and m-C1P3 (long dashes, – – – ).

SINGLE-PHOTON TIMING MEASUREMENTS

15

a small shift and broadening of the meta-substituted compounds spectra relative to the ones of the para-substituted compounds can be seen. Moreover, a change in the intensity ratio between the two vibronic maxima is also visible. For the meta compounds, the vibronic maximum at 595 nm is relatively more pronounced as compared to the one for the para compounds. The para coupling allows for a better conjugation of the p-electrons of the peryleneimide over the aromatic phenyl ring of the branch. As this effect is more important in the excited state than in the ground state, it will alter the perpendicular orientation of the neighboring phenyls in the excited state compared to the ground state. The width of the fluorescence band at half maximum (FWHM) increases slightly with the number of chromophores from 2680 cm1 for m-C1P1 to 2750 cm1 for m-C1P4. The fluorescence spectra of the first generation para-substituted dendrimers p-C1Px (x ¼ 1--4) are independent on the number of PI chromophores. Similarly, the absorption and emission spectra of the second generation rigid dendrimers (p-C2P1, p-C2P2, p-C2P3, p-C2P4) were found to be independent of the number of chromophores present in the dendrimers. The fluorescence quantum yield (f ) is calculated to be 0.98 0.05 and is identical within experimental error for all compounds. The similarity of the fluorescence properties of all the para-substituted dendrimers in terms of spectral shape, fluorescence maxima, and fluorescence quantum yield suggests that the emission occurs from the same state in all the dendrimers. Triplet formation is very inefficient in these chromophores: the rate constant of intersystem crossing could be measured using single molecule spectroscopy and was found to be equal to 7 103 s1 [25, 29].

IV. SINGLE-PHOTON TIMING MEASUREMENTS A. Time-Resolved Fluorescence Measurements Performed Under Magic Angle Polarization Condition In order to examine the properties of the fluorescent states for the dendrimers more closely, fluorescence decay times for all first generation dendrimers were determined in toluene by single-photon timing detecting the emission under magic angle condition. 1. Para-substituted Carbon Core Dendrimers Table 1.1 shows that the lifetimes of p-C1P1, p-C1P3, and p-C1P4 are identical with the fluorescence decay measured for an adequate model containing a peryleneimide chromophore. A representative plot of the fluorescence decay of the first generation para-substituted dendrimers is given in Figure 1.13 for p-C1P4.

16


TABLE 1.1 Fit Parameters of the Fluorescence Magic Angle and Anisotropy Decays Measured for p-C1Px (x ¼ 1; 3; 4) in Toluene with kexc ¼ 488 nm and kflu ¼ 600 nm and Average Peryleneimide–Peryleneimide Distances (dDA) Compound p-C1P1 p-C1P3 p-C1P4

t (ns)

r0

y1 (ns)

y2 (ps)

b1

b2

4.2 4.2 4.2

0.34 0.31 0.34

1.4 1.6 2.0

— 70 50

0.34 0.09 0.07

— 0.33 0.37

b2/r0 (%) dDA (nm) — 71 79

— 2.7 2.7

The corresponding decay parameters are collected in Table 1.1. Similarly, the decays of the second generation dendrimers were measured and all decays could be fitted globally by a single exponential with a time constant of 4.2 ns (Table 1.2).

Figure 1.13. Time-resolved fluorescence decays of p-C1P4 with fits at 600 nm and 700 nm detection wavelengths. The upper panel shows the weighted distribution of residuals (Ri) and the lower panel represents the autocorrelation (ac) function for the decays. Inset reports on a shorter time scale.

17


TABLE 1.2 Fit Parameters of the Fluorescence Magic Angle and Anisotropy Decays Measured for p-C2Px (x ¼ 1; 2; 3; 4) in Toluene with kexc ¼ 488 nm and kflu ¼ 600 nm and Average Peryleneimide–Peryleneimide Distances (dDA) Compound p-C2P1 p-C2P2 p-C2P3 p-C2P4

t (ns)

r0

y1 (ns)

4.2 4.2 4.2 4.2

0.32 0.36 0.36 0.35

2.7 3.1 2.7 3.0

y2 (ps) — 410 310 280

b1

b2

0.32 0.18 0.14 0.08

— 0.18 0.22 0.27

b2/r0 (%) dDA (nm) — 50 61 77

— 3.6 3.7 3.8

2. Meta-substituted First Generation Carbon Core Dendrimers The corresponding decay parameters are collected in Table 1.3. The fluorescence intensity of the dendrimer having only one chromophore (m-C1P1) decays single exponentially with a decay time of 4.25 0.05 ns. However, as the number of chromophores is increased in the dendrimer, a small contribution of an additional long decay component of 7.4 0.6 ns is found essential to fit the experimental data. It has to be noted, however, that the amplitude of this long decay component is very small in m-C1P2 and m-C1P3. Thus, in order to minimize the error in the fit procedure, an additional component with a fixed decay time of 7.4 ns, as obtained for m-C1P4, was introduced in the analysis of the fluorescence decays of m-C1P2 and m-C1P3 to allow a better comparison of the corresponding amplitudes. It was furthermore observed that the relative amplitude of the longer decay time is larger at the red edge of the fluorescence spectrum for all multichromophoric dendrimers as shown in Table 1.3 by the comparison of results obtained at 600 nm and 725 nm emission. From the small difference in the spectral width (vide supra), the assumption of an excited state excimer-like (or dimer) chromophore–chromophore interaction is possible but not conclusive. Better insight into the extent of excimer-like emission is obtained from the fluorescence decays, where only for the multichromophoric dendrimers is a long decay component of 7.4 ns observed along

TABLE 1.3 Fluorescence Decay Times (si ) and Associated Relative Amplitudes (ai ) for m-C1Px (x ¼ 1–4) Measured in Toluene at Room Temperature Using kexc ¼ 488 nm Compound m-C1P1 m-C1P2 m-C1P3 m-C1P4

t1 (ns)

t2 (ns)

4.25 4.25 4.25 4.25

0.32 0.36 0.36 0.35

a1-600 (%) 100.0 99.2 98.1 96.0

a2-600 (%) 0.0 0.8 1.9 4.0

a1-725 (%) — 98.7 96.2 93.7

a2-725 (%) — 1.3 3.8 6.3

18


with the typical peryleneimide fluorescence decay time of 4.25 ns as obtained for the monochromophoric model compound m-C1P1. The attribution of this long time constant can be made to an ‘‘excimer-like’’ species as the decay time is similar to that reported for the higher generation dendrimers having a flexible biphenyl core [26]. Further evidence for this assignment can be derived from the dependence of the amplitude a2 connected with the 7.4 ns component on the number of chromophores and the dependence on the emission wavelength (lflu), respectively. As reported in Table 1.3, this amplitude is 0.8% for m-C1P2 and increases to 4% for m-C1P4 at lflu ¼ 600 nm. This is reasonable as the probability of formation of the ‘‘excimer-like’’ entity increases as the number of chromophores in the dendrimer increases. By detecting at lflu ¼ 725 nm, a2 increases to 1.3% for m-C1P2 and to 6.3% for m-C1P4. The larger contribution of that component at longer emission wavelengths is also consistent with a red-shifted fluorescence from ‘‘excimerlike’’ entities. This suggests that a fraction of the molecules have a substitution pattern in which two of the PI chromophores are relatively close in space. No such long decay component of 7.4 ns is observed for para-substituted dendritic structures p-CnPn. The absence of the long decay component is therefore due to the different position of substitution leading to a better spatial separation of the individual chromophores. This is also supported by a comparison of the molecular structures of the para- and meta-substituted dendrimers obtained from molecular modeling, since the average center-to-center distance among the chromophores is 2.9 nm for the para series but only 2.6 nm for the meta series in the first generation series.

B. Time-Resolved Fluorescence Polarization Measurements From time-resolved fluorescence depolarization measurements, the anisotropy decay times () and the associated anisotropy (b) have been determined for all first generation dendrimers using Eq.(1): X X bi ð1Þ rðtÞ ¼ bi expðt=i Þ with r0 ¼ i

The sum of all bi is called the limiting anisotropy r0. 1. Meta-substituted First Generation Carbon Core Dendrimers For the monochromophoric meta-substituted dendrimer (m-C1P1), a monoexponential fit of the anisotropy decay function is sufficient, which gives a relaxation time of 1 ¼ 950 30 ps with b1 ¼ r0 ¼ 0:38 (Table 1.4). However, the anisotropy decay functions for the meta-substituted dendrimers having more than one chromophore (m-C1P2 to m-C1P4) can only be fitted with two exponential decay

19


TABLE 1.4 Fit Parameters of the Fluorescence Anisotropy Decays Measured for m-C1Px (x ¼ 1–4) in Toluene with kexc ¼ 488 and kflu ¼ 600 nm at Which There Is Only Monomer Emission and Average Chromophore–Chromophore Distances (dDA) Compound m-C1P1 m-C1P2 m-C1P3 m-C1P4

r0 0.38 0.31 0.28 0.24

y1 (ns) 0.9 1.1 1.2 1.3

y2 (ps) — 200 130 110

b1

b2

0.38 0.16 0.10 0.08

— 0.15 0.18 0.16

b2/r0 (%) — 48 63 66

dDA (nm) — 2.6 2.6 2.7

functions (Table 1.4). The long depolarization time constant is similar to that obtained for m-C1P1. The value of this long time component increases with the number of chromophores from 1:1 0:04 ns for m-C1P2 to 1.3 0.07 ns for m-C1P4, while the value of the fast component (2) changes from 200 30 ps for m-C1P2 to 110 20 ps for m-C1P4 (Table 1.4). The sum of the bi for m-C1P3 and m-C1P4 is substantially smaller than the limiting anisotropy. This strongly suggests that, at a time shorter than the resolution of single-photon timing, there is already a process leading to loss of fluorescence polarization in the meta-substituted dendritic systems. 2. Para-substituted First Generation Carbon Core Dendrimers For the para-substituted dendrimer with one chromophore (p-C1P1), a monoexponential function is found to be sufficient to fit the anisotropy decay trace, which can be related to the relaxation time of 1 ¼ 1:4 ns 30 ps with b1 ¼ r0 ¼ 0:34 0:04 (Table 1.1). However, the anisotropy decay traces for the dendrimers having more than one chromophore (p-C1P3 and p-C1P4) can only be fitted with two exponential decay functions (Table 1.1). The amplitude of the component with a long depolarization time increases with the number of chromophores increasing from 1 to 4, while the value of the fast component (2) changes from 70 ps for p-C1P3 to 50 ps for p-C1P4 (Table 1.1). From fluorescence depolarization measurements, anisotropy relaxation times and the associated anisotropy values have been determined for p-C2P1, p-C2P2, p-C2P3, and p-C2P4. For the dendrimers with more than one chromophore, a two-exponential function was found to be necessary to fit the experimental anisotropy decay traces (Table 1.2). The multichromophoric dendrimers present two-exponential decays in the anisotropy traces. The fast component (410 ps to 280 ps) of the anisotropy decay (Table 1.2) is found to decrease from p-C2P2 to p-C2P4. Contrary to the meta-substituted dendrimers m-C1Pn, the sum of the bi is now always close to the limiting value of the anisotropy even if n is larger than one. Energy transfer processes can be revealed by time-resolved anisotropy data. The large value for the limiting anisotropy (r0) of p-C1P1, p-C2P1, and m-C1P1

20


(Tables 1.1, 1.2, and 1.4, respectively) confirms the parallel orientation of the absorption and emission transition dipole moment for a single chromophore. In contrast to the fluorescence anisotropy decay of m-C1P1, which contains only one peryleneimide chromophore, an additional shorter picosecond anisotropy decay component is observed in the multichromophoric dendrimers p-C1Pn , p-C2Pn , and m-C1Pn ðn > 1Þ. Therefore, this fast depolarization process can unambiguously be related to excitation energy hopping among the identical chromophores. The time scale of a hundred picoseconds for these processes suggests that the observed energy hopping occurs in terms of fluorescence resonance energy transfer. Within the framework of the Fo¨rster formulation [20, 21, 27], a rate constant for excitation transfer between donor D and acceptor A can be expressed as kET ¼

6 R0 kD R

ð2Þ

where R is the interchromophoric distance, kD is the inverse of the decay time of the donor, and R0 (the Fo¨rster radius) is the distance at which the efficiency equals 50%, that is, the distance at which an equal probability exists for the excited chromophore to relax to the ground state or to undergo energy transfer. R0 depends on the relative orientation of the transition dipoles toward each other (k2), the spectral overlap (J(l)) of the absorption spectrum of the acceptor, and the normalized emission spectrum of the donor, D, that represents the quantum yield of fluorescence of the donor, and n that represents the refractive index of the solvent as can be seen in Eq. (3): R0 ¼ 0:211ðê2 n4 D Jð€eÞÞ1=6

ð3Þ

The calculated value of J ¼ 2:5 1014 M1 cm3 and R0 ¼ 3:8 nm using the spectral data (e ¼ 38; 000 M1 cm1 , F ¼ 0:98) of the monochromophoric m-C1P1 model compound have the typical order of magnitude. The efficiency as a function of relative distance, E, is E¼

R60 R60 þ r 6

ð4Þ

Information about the rate constant of hopping (khopp) through excitation energy transfer can be derived from the fast anisotropy decay time (2). In order to take into account the possibility of multiple energy transfer channels in the case of a multichromophoric system containing identical chromophores, among which efficient dipole–dipole interactions occur, the measured anisotropy decay time 2 can be related to khopp by Eq. (5), where the value of i represents the number


21

of chromophores fully interacting in both forward and backward directions [28]. khopp ¼

1 1 i2 i1

ð5Þ

If we take m-C1P2 as the model system for energy hopping between two peryleneimide chromophores and further assume that the energy transfer occurs in both directions, then the rate constant (khopp) calculated from Eq. (5) with i ¼ 2 results in khopp ¼ 2:0 ns1 . Using this value for khopp, we can calculate the expected anisotropy decay time (2) for the case of equally distributed and interacting chromophores in m-C1P3 and m-C1P4, which gives 133 ps and 67 ps, respectively. These results are in good agreement with the experimentally observed 2 ¼ 130 ps of m-C1P3 and 2 ¼ 110 ps of m-C1P4. This indicates the suitability of the proposed model of energy hopping among all chromophores. Within the Fo¨rster formulation, the donor–acceptor distances (dDA) can be calculated by Eq. (6) and are listed in Table 1.4: 6 ¼ dDA

R60 kET tD

ð6Þ

where tD is the fluorescence decay time of the donor chromophore. All calculated values of dDA for the peryleneimide dendrimers are on the order of 2.6 nm, which is in good agreement with the average distance between two chromophores in different conformations obtained from molecular modeling results [28, 29]. This agreement further substantiates the suitability of the above proposed model for an energy hopping mechanism in the present dendrimers. However, with an increasing number of chromophores r0 decreases from 0.38 in m-C1P1 to 0.31 in m-C1P2 and to 0.24 in m-C1P4. This means that an additional fast depolarization process on a time scale below the time resolution of 30 ps for the time-correlated single-photon counting anisotropy experiments takes place. This loss in initial anisotropy can be explained by the occurrence of ultrafast energy hopping between neighboring chromophores, which can approach one another to distances on the order of 1 nm or by dimer formation within the temporal resolution, also observed in transient absorption anisotropy (vide infra). Hence, the Fo¨rster approach for m-C1Pn is a first approximation, which is not fully adequate for that fraction of molecules in the ensemble of the constitutional isomers for which the closer distance leads to electronic coupling values no longer negligible and cannot be described with a weak coupling model. This restriction, however, is not applicable for the p-C1Pn series. In contrast to the monoexponential anisotropy trace of monochromophoric p-C1P1, the corresponding traces of the multichromophoric dendrimers p-C1P3 and p-C1P4 reveal a second and fast anisotropy decay component on the order of 50–80 ps (Table 1.1). Within the framework of the Fo¨rster

22


formulation, an effective interaction radius (R0) can be calculated from the steady state spectra and the fluorescence quantum yield of the donor chromophore (fD), yielding a value of R0 ¼ 3:8 nm. On the basis of Eq. (6), a value of around 4:6 ns1 is obtained for khopp of p-C1P3 and p-C1P4. It might seem surprising that this value is more than twice as large as that (khopp ¼ 2 ns1 ) obtained for the meta-substituted dendrimers, even though the interchromophoric distances are 0.2 nm larger in the para series. In fact, by employing the excited state lifetime tD ¼ 4 ns, the above derived values of R0 ¼ 3:8 nm, and khopp, the calculation of the distance between two chromophores by Eq. (6) yields too small values of dDA. The obtained interchromophoric distances dDA for the peryleneimide chromophores are on the order of 2.3 nm in spite of the expected 2.8 nm from molecular modeling structures. The only reason for this large discrepancy can be the wrongly estimated value of R0 due to the too simplified assumption of the dipole–dipole orientation factor k2 value of 2/3, which is strictly valid only for a random orientation of the chromophores. Here, this assumption is not true anymore because of the attachment of chromophores into the dendrimer backbone. The real values of k2 are described by Eq. (7) k ¼ sinðdD Þ sinðdA Þ cosðjDA Þ 2 cosðdD Þ cosðdA Þ

ð7Þ

where jDA is the azimuthal angle between the involved transition dipole moment directions of the energy donor D and acceptor A, and dD and dA are the angles between the corresponding dipole directions of D and A with the internuclear D–A axis, respectively. For m-C1Pn, the values of k2 have been calculated using geometrical data derived from a 3D molecular mechanics calculation and leading to average values of around 0.8 for the chromophore orientations in the meta-substituted dendrimers, confirming that the approximation of k2 ¼ 2=3 (vide infra) was reasonable. However, for the para-substituted dendrimers p-C1Pn, the average k2 is determined as 2.1 and thus is much larger. The ratio of the calculated k2 values for the para versus those of the meta series is about 2.6. This value is in good agreement with the respective ratio of the experimentally determined hopping rate constants being about khopp(para)/khopp(meta) ¼ 2.3 or, if the slightly different interchromophoric distances (d ffi 2.6 nm for meta and d ffi 2.8 nm for para) are taken into account, with the ratios khoppd 6(para)/ khoppd 6(meta) ¼ 3.5. Consequently, the faster energy hopping kinetics in the para series can directly be traced back to a better orientation of the peryleneimide chromophores toward each other, yielding a much larger Fo¨rster interaction radius R0 of 4.4 nm than in the meta series. Using this value of R0 in Eq. (6) indeed leads to values of dfret ¼ 2:7--2:8 nm, which are in good agreement with the average interchromophoric distances found in molecular mechanics modeling.

23


50

S1

S0

S0

S1 Fluorescence Intensity (a.u.)

ε (103/ M cm)

40

energy transfer

30

20

10

0 350

400

450

500

550

600

650

700

750

Wavelength (nm)

Figure 1.14. Stationary absorption and emission spectra of p-C1P4 in toluene. The spectral overlap is depicted in gray. Inset: The scheme representing singlet–singlet excitation hopping.

Energy hopping is a Fo¨rster-type process that is present in the multichromophoric dendrimer such as p-C1P4 and can be related to the spectral overlap as depicted in Figure 1.14. Using the values of dDA and R0 mentioned above for p-C1P4, efficiencies of 97.5% are obtained for energy hopping. The efficiency of energy hopping and singlet–singlet annihilation in p-C1P4 as a function of distance is shown in Figure 1.15. The figure clearly indicates that 50% efficiency is reached for a distance of 4.5 nm. It also allows us to see where in this three-dimensional picture the p-C2Pn series is situated. As the attachment of the chromophores to the dendrimer backbone in p-C2Pn cannot be taken as random, the value of k has been calculated from the threedimensional molecular structure using Eq. (7). The average value of about 2.7 has been found for the dendrimers where the chromophores are at large distances from each other (Fig 1.16a, a0 ). However, for the isomer of p-C2P4 with a short distance pair of chromophores (Fig. 1.16b, b0 ), the average k2 for all couplings between pairs of two chromophores is obtained as 1.5. The presence of two sites in each branch where the formation of different constitutional isomers is possible will lead to a much broader relative distribution of the distances and angles between the chromophores compared to p-C1P4. Hence, the hopping rate constant

24


1.0 Energy hopping 0.8

Efficiency

0.6

0.4

0.2

0.0 0

2

4

6

8 10 Distance (nm)

12

14

16

Figure 1.15. Schematic representation of the efficiency of the energy hopping process present in p-C1P4 as a function of distances expressed in Eq. (4).

khopp obtained from experimental results should be considered as an average hopping for the different possible constitutional isomers in the dendrimer. Taking into account the possibility of random hopping in the multichromophoric systems containing identical chromophores, an average hopping rate constant (khopp) according to the energy hopping model is given by Eq. (5), where 1 and 2 are the experimental extracted decay times and the value of i represents the number of chromophores. Using Eq. (5), a value of 0.85 ns1 for khopp is obtained for these dendrimers. This value is more than five times smaller than that of the corresponding first generation dendrimers. Based on the excited state lifetime (tD), the derived values of R0 (4.5 nm) with a value of 2.7 for k2 and khopp the distance between the two chromophores have been calculated from Eq. (6). This yields a value of d ¼ 3:7 nm, which is in good agreement with the average interchromophoric distance obtained from molecular modeling. From the sixth power dependence of khopp on the average interchromophoric distance (dDA) and using the ratio of the values of khopp for first and second generation series, an average for dDA for the second generation series is found to be 3.7 nm. As shown in Figure 1.15, the multichromophoric second generation dendrimer is still inside the active sphere in which energy hopping can take place with high efficiencies. The decrease in hopping rate constant in these molecules thus scales with the sixth power of the distance difference as expected within the Fo¨rster model.

FEMTOSECOND FLUORESCENCE UPCONVERSION MEASUREMENTS

(a)

O

N

25

(a¢)

O

O N

O C

O

N O

Isomer p-C2P4

O

N

O

(b)

(b¢)

O N O O N O

O N C O

Isomer p-C2P4

O

N

O

Figure 1.16. Molecular structures of p-C2P4 isomers: (a) isomer with a long distance pair of chromophores; ða0 Þ 3D structure of isomer with a long distance pair of chromophores; (b) isomer with a short distance pair of chromophores; ðb0 Þ 3D structure of isomer with a short distance pair of chromophores.

V. FEMTOSECOND FLUORESCENCE UPCONVERSION MEASUREMENTS To reveal possible ultrafast processes occurring on a time scale less than 30 ps, femtosecond fluorescence upconversion experiments were performed [30] in toluene under magic angle polarization. To extract complete information of the decay times and their amplitudes in function of detection wavelength, the measurements were performed in three time windows of 5 ps, 50 ps, and 420 ps. In order to reveal properties that are independent of potential chromophore– chromophore interactions, p-C1P1 was investigated in a first series of measurements as a model compound, since it contains only one chromophore. Figure 1.17a shows a typical result for p-C1P1 at two different detection

26


Normalized Fluorescence Intensity

(a)

620 nm

540 nm

0 0

10

(b)

20 30 Delay (ps)

40

50

Normalized Intensity

1.0 590 nm

0.8

p -C1P4

0.6

p -C2P4

0.4 0.2

(a)

0.0 0

100

200

300

400

Normalized Intensity

1.0 0.8

p -C1P1 p -C2P1

590 nm

0.6 0.4 0.2

(b)

0.0 0

100

200 Delay (ps)

300

400

Figure 1.17. (a) Time-resolved fluorescence intensity of p-C1P1 detected at 540 nm and 620 nm as indicated. (b) Comparison of the time-resolved fluorescence intensity recorded at 590 nm. (a) Multichromophoric first generation p-C1P4 and multichromophoric second generation p-C2P4. (b) Monochromophoric first generation p-C1P1 and monochromophoric second generation p-C2P1.


27

TABLE 1.5 Decay Times Resulting from Global Analysis for All Dendrimers Investigated in Toluene Compound

t1 (ps)

t2 (ps)

t3 (ps)

t4 (ns)

p-C1P1 p-C1P3 p-C1P4 m-C1P1 m-C1P3 m-C1P4 p-C2P1 p-C2P4

0.5–2.0 0.5–2.0 0.5–2.0 0.5–2.0 0.5–2.0 0.5–2.0 0.5–2.0 0.5–2.0

6.3 4.6 4.0 10.0 8.0 7.5 6.0 5.8

110 45 45 188 137 83 50 40

4.2 4.2 4.2 4.2 4.2 4.2 4.2 4.2

wavelengths excited at 495 nm, showing a clear wavelength dependence of the fluorescence decay and a complex multiexponential decay consisting of several components. Especially in the first few picoseconds pronounced wavelength dependence is evident. These data are then compared to those of dendrimers containing 3 and 4 peryleneimide chromophores at the rim. By global analysis, four decay components were revealed in both first and second generation compounds. Decay traces from second and first generation dendrimers p-C1P4, p-C2P4, p-C1P1, and p-C2P1 are compared in Figure 1.17b. The resulting time constants obtained by the global analysis procedure for the various compounds are summarized in Table 1.5. The values for t1 are not constant at different analysis wavelengths throughout the spectrum, so these decay times could not be linked globally. The second component (t2, a2) exhibits a fast time constant on the order of a few picoseconds for all compounds and represents 15–40% of the total amplitude, depending on the wavelength and the compound. The third component contributes at most 10%, and in most cases even less to the total amplitude, but is found necessary to obtain good fits. The largest part of the amplitude, however, is found in the nanosecond component 4 (t4, a4) for all compounds. Figure 1.18a shows the partial amplitudes for p-C1Px for the ultrafast decay component 1 as a function of the detection wavelength. The related decay component t1 is the only decay time out of the four resolved in our analysis that is wavelength dependent as shown in Figure 1.18c for p-C1P3. It clearly demonstrates the increase of the decay time with increasing fluorescence detection wavelength. This shortest time constant is measured at the shortest detection wavelength and has a value of 500 fs evolving as shown in Figure 1.18c to 2 ps from shorter to longer wavelengths. While this decay time remains more

28


(a) 0.2

a1

0.0

–0.2

520

540

560

580

600

620

640

660

680

540

560

580

600

620

640

660

680

600

620

640

660

680

(b)

a1

0.0

–0.2

520 (c)

p-C1P3

τ1/ps

2.0

1.5

1.0

0.5 520

540

560

580

Wavelength / nm

Figure 1.18. Dependence of the intramolecular vibrational reorganization process amplitude a1 on the detection wavelength, (a) for the p-C1Px dendrimers (p-C1P1 [&], pC1P3[*], p-C1P4 [~]) and (a) comparison of p-C1P1 [&] and m-C1P1 [&]. (c) Time constant t1 as a function of the detection wavelength for p-C1P3.


29

or less constant between 500 and 700 ps between 540 and 640 nm, it increases rapidly at longer wavelengths. Thus, the time constant could not be kept constant in the global analysis, although the values obtained for the partial amplitudes are still the result of the global analysis procedure in which the three other decay times were linked. A second observation that can be made for this decay component is that it has negative partial amplitudes at all detection wavelengths above 540 nm. This means a growing-in of the decay curves at the early times after excitation due to the population of a fluorescing state from the initially populated vibronic level. For all compounds p-C1P1, p-C1P3, and p-C1P4, a similar behavior with respect to partial amplitudes and decay times could be observed at the measured fluorescence wavelengths (Fig. 1.17a). This behavior of the negative partial amplitudes, the order of magnitude and change in decay time depending on detection wavelength is typical for an intramolecular vibrational reorganization process in the electronically excited state of the chromophore [31]. This component is found in first and second generation dendrimers discussed here as well in the mono- and multichromophoric ones and is a combination of various processes resulting from the static and dynamic response of the environment of the chromophore [32]. Also, a fast relaxation of vibrationally excited levels (max. 2000 cm1 ) of the first singlet excited state in the peryleneimide cannot totally be excluded as a part of this component [32]. The second decay component that could be found in the para-substituted dendrimers has a value of 6.3 ps to 4 ps, depending on the compound (Table 1.5). Figure 1.19a shows the partial amplitudes of p-C1Px for this component as a function of the detection wavelength. First, considering only the monochromophoric compound p-C1P1 (Fig. 1.19a [&]) with a t2 of 6.3 ps, a change of sign of the partial amplitude can be observed. Taking into account the shape and the positive/negative behavior of this kinetic component, it is attributed to a vibrational relaxation in the electronically excited state of the peryleneimide chromophore. This process is coupled to a relaxation of the solvation shell around the chromophore, as the solvent molecules have to accommodate for the newly populated S1 state of the peryleneimide [33]. At fluorescence detection wavelengths close to the excitation, this will be seen as a fast decay component, whereas at longer wavelengths the fluorescence is detected from a state that first has to be populated with the time constant resolved. In the kinetic analysis, this is found as a rise term with the corresponding time constant. Thus, it can be concluded that this kinetic component is related to the single chromophore itself and its interaction with the surrounding solvent toluene molecules. The finding of a 6.3 ps component and its attribution is in line with literature, where an ultrafast stimulated transient absorption spectroscopy setup [34] was used to determine a vibrational population relaxation time in the same order of magnitude for

30


(a) 0.6

a2

0.4 0.2 0.0 –0.2 520

540

560

580

600

620

640

660

680

540

560

580

600

620

640

660

680

(b) 0.6

a2

0.4 0.2 0.0 –0.2 520

Wavelength (nm)

Figure 1.19. Wavelength dependence of the amplitude a2 of the second component in toluene (a) for the p-C1Px dendrimers (p-C1P1 [&], p-C1P3 [*], p-C1P4 [~]) and (b) comparison of p-C1P1 [&] versus m-C1P1 [&] and p-C1P4 [~] versus m-C1P4 [~].

molecules such as perylene in toluene solution. In many other investigations [35], time constants of a few picoseconds were found and attributed to a vibrational relaxation process for various chromophores in toluene and other solvents. To study the influence of the number of chromophores attached to the dendrimer on this second component, the multichromophoric compounds p-C1P3 and p-C1P4 were also studied. As can be seen in Figure 1.19a, the typical shape and wavelength dependence of the partial amplitude is persistent for all three dendrimers, but an additional positive shift can clearly be observed which increases with the number of chromophores. This clearly indicates the contribution of more than one process to this second kinetic component, meaning a more complex attribution compared to the one in the monochromophoric compound. Thus, for the interpretation of these results, two different contributions 2a and 2b to this component are assumed, which are related to different kinetic processes,


31

both of which exhibit time constants that are very close to each other and because of this cannot be separated by the global analysis. First, contribution 2a is present in all compounds and can be attributed to the vibrational and solvent relaxation, responsible for the typical shape and wavelength dependence and appearing purely only in the analysis of the monochromophoric compound. Contribution b of component 2 (2b) can only be observed if more than one chromophore is present and hence if there is an intramolecular interaction possible between two or more chromophores. It is superimposed on the wavelengthdependent contribution a of time constant 2 (2a) and is almost wavelength independent and increasing in amplitude with the number of chromophores. If more than one chromophore per molecule can get excited, a singlet–singlet annihilation process could occur, eventually resulting in a first excited singlet state and a ground state chromophore [30]. When the photon flux available in the laser focus at the sample position is calculated, a value of several tens of photons per chromophore and per laser pulse can be found; hence, intensity dependence of the photophysics can be expected. The estimated distance obtained by means of molecular modeling between two chromophores is about 2.9 nm [28]. The combination of this estimated distance and the fact that more chromophores can get excited simultaneously in one molecule yields the possibility for an intramolecular singlet–singlet annihilation of two excited chromophores resulting eventually in a first excited state and a ground state [36]. This annihilation process has been reported [36, 37] and has been the experimental topic in different investigations in various systems such as pigment–protein complexes [38] and J aggregates [39]. Thus, it is assumed that contribution 2a can be attributed to vibrational and solvent relaxation whereas 2b, only present in multichromophoric dendrimers, is attributed to singlet–singlet annihilation. In order to distinguish and separate these two kinetic decay channels, an excitation energy-dependent study was performed on the mono- and multichromophoric para compounds. The excitation energy imposed onto the sample was systematically varied between 20 and 400 nJ, corresponding to several tens and several hundreds of photons per laser pulse and per chromophore, respectively, and a clear dependence of the amplitude of the total second component (2a þ 2b) could be observed. Figure 1.20 shows the decay curves for p-C1P4 at two different well chosen wavelengths, namely, 590 nm (a) and 630 nm (b) at two different excitation energies. These detection wavelengths were selected because of the values of the amplitude of the vibrational relaxation process observed in the data obtained for p-C1P1 (Fig. 1.19a). At 590 nm, it is close to zero, while at 630 nm it has a clear negative value. In contrast, for the multichromophoric compounds, the amplitudes are positive and substantially larger at these selected wavelengths, which is due to the admixture of annihilation process 2b. Its partial amplitude should decrease as the excitation energy diminishes, and hence the partial amplitudes as a function of wavelength of component 2 of the

32


Figure 1.20. Comparison of the time-resolved fluorescence intensity I recorded at low and high excitation energy (as indicated). (a) Multichromophoric compound p-C1P4 detected at 590 nm. (b) Multichromophoric compound p-C1P4 detected at 630 nm. (c) Monochromophoric compound p-C1P1 detected at 590 nm.

multichromophoric compounds should converge to those observed for p-C1P1 (component 2a) at low excitation energies. This means that at 630 nm detection the total amplitude of this second decay component (2a þ 2b) should turn from a positive into a negative value with decreasing excitation energy. This is exactly what is observed (Figure 1.20b). At 590 nm, there is a clear decrease of the contribution of the annihilation process 2b upon lowering the excitation energy. Since the partial amplitude of the vibrational relaxation at this wavelength is also zero (vide infra), the total amplitude (2a þ 2b) of component 2 will vanish, yielding a decay consisting


33

of a nanosecond component that only appears as a constant in this short time window (Figure 1.20b). In order to cross-check these findings, a similar energy series has also been performed for the monochromophoric compound p-C1P1. The results shown in Figure 1.20c show no detectable excitation energy dependence within the measured range as expected and in contrast with the multichromophoric compound measured at this detection wavelength. Thus, at all excitation intensities, the partial amplitudes a2 are constant, which is a clear indication that in this monochromophoric compound the contribution b of time constant 2 is nonexistent. Figure 1.21 shows the partial amplitude a2 for p-C1P1 (&) and p-C1P4 () at the two selected detection wavelengths (Fig. 1.21a at 590 nm and Fig. 1.21b at 630 nm) as a function of the excitation energy. The data of the multichromophoric dendrimers ( p-C1P4 shown) contain the typical dependence of an annihilation process, while those of the monochromophoric p-C1P1 do not exhibit excitation energy dependence variations (Fig. 1.21a, b).

Figure 1.21. Dependence of the partial amplitude a2 of the second component from the laser excitation energy for the para-substituted p-C1Px (p-C1P1 [&], p-C1P4 [~]) and meta-substituted peryleneimide dendrimers m-C1Px (m-C1P1 [&], m-C1P4 [~]). (a) Detection done at 580 nm (meta compounds) and 590 nm (para compounds). (b) Detection done at 620 nm (meta compounds) and 630 nm (para compounds).

34


This is a clear indication that in this monochromophoric compound the second contribution b of kinetic component 2 is not present and that this amplitude spectrum is showing only the contribution a of kinetic component 2 (related time constant 6.3 ps), which is attributed to the vibrational/solvent relaxation of the molecule. Looking at decay times for this component 2 (Table 1.5), one can observe a decrease in decay time upon increasing number of chromophores. This can be explained by the fact that the decay time determined by the analysis is a weighted combination of these two separate decay times t2a and t2b of the vibrational/solvent relaxation and the annihilation process as shown in the previous paragraph. Since the relaxation process is the slower of these two processes, the more chromophores present, the more important the annihilation becomes and the shorter the overall decay times. Similar observations were made for the meta-substituted dendrimers m-C1Pn, when n is larger than 1. In the second generation dendrimers, the second decay component that could be recovered has a decay time of 6 ps (Table 1.5). Figure 1.22 shows the partial amplitudes for the monochromophoric second generation compound p-C2P1 as a function of the detection wavelength. The positive offset of the partial amplitude curves of p-C2P4 (Fig. 1.19c) compared to p-C2P1 indicates that more than one process is contributing to the apparent component 2 of the multichromophoric compound as also observed in the first generation dendrimers. The first contribution (2a) to this process in both compounds has been attributed to a relaxation process (vide supra). The second contribution (2b) to this process is again an intramolecular singlet–singlet annihilation process that is independent of detection wavelength and exists only in compounds with multiple chromophores. It is, however, clearly less important than in the first generation p-C1P4. To further underpin the hypothesis formulated for p-C2P1 and p-C2P4 and to be able to separate these two processes discussed above, an excitation energy0.6

a2

0.4 0.2 0.0 –0.2 520

540

560

580

600

620

640

660

680

Figure 1.22. Wavelength dependence of the amplitude a2 of the second component for the compounds p-C2P4 [&], p-C1P4 [&], p-C2P1 [~], and p-C1P1 [].


35

dependent study was also performed on both compounds. By varying the excitation energy impinging on the sample between 20 and 420 nJ, a clear dependence of the amplitude of the 6 ps component could be observed. This energydependent study was performed at the strategically chosen detection wavelength of 590 nm (vide supra). For the monochromophoric compound (p-C2P1), the partial amplitude for the component 2 is close to zero. However, for the multichromophoric compound (p-C2P4), the amplitude of the apparent component 2 becomes positive. Thus, all intensity dependence observed at 590 nm detection wavelength can be attributed to the intramolecular singlet–singlet annihilation process. The partial amplitudes for the multichromophoric compound (p-C2P4) are shifted to a higher value over the entire detection wavelength range. In Figure 1.23, the decays recorded at the 590 nm detection wavelength and at different excitation energies are depicted for the multichromophoric p-C2P4. Because the relative importance of the annihilation process should increase as the excitation energy increases, the partial amplitudes as a function of the detection wavelength of the 6 ps component of the multichromophoric compound should at low excitation energy resemble the one for the monochromophoric compound. This is exactly what is observed.

Figure 1.23. Dependence of the partial amplitude a2 of the second component from the laser excitation energy for the compounds p-C2P4 [&], p-C2P1 [~], p-C1P4 [&], and p-C1P1 [~] at detection wavelengths (a) 630 nm and (b) 590 nm.

36


The dependence of the partial amplitude a2 of this 6 ps component on the incident laser energy is shown in Figure 1.23a, b for p-C2P4 and p-C2P1 at 630 nm and 590 nm detection wavelength, respectively. This is a clear indication that in p-C2P1 the annihilation process is absent and that the amplitude spectrum is only showing the vibrational/solvent relaxation of the chromophore itself. The positive amplitude offset of the multichromophoric compounds with respect to the monochromophoric compounds is more pronounced for the first generation p-C1P4 than for the second generation p-C2P4, as seen in Figure 1.23 where the amplitude a2 is displayed as a function of the excitation energy at the detection wavelengths 630 nm (Fig. 1.23a) and 590 nm (Fig. 1.23b). At both detection wavelengths, the curve for the first generation compound p-C1P4 has larger partial positive amplitude compared to the second generation compound p-C2P4. The third kinetic component that could be recovered for all para-substituted peryleneimide dendrimers p-C1Px at all detection wavelengths has a time constant on the order of 100 ps and a relatively low partial amplitude. By checking a possible concentration effect between 105 M and 106 M on the different kinetic components by diluting the samples, this was the only component that was found to be dependent on the concentration. As a time constant on the order of 100 ps was also retrieved in SPT measurements performed on concentrated solutions of p-C1Px, this component can be attributed to an intermolecular process. For the second generation p-C2P4, however, this partial amplitude is intensity dependent as can be seen in Figure 1.24. Figure 1.24 depicts the partial amplitude a3 for p-C2P1, p-C2P4, and p-C1P1 at 590 nm as a function of the excitation energy. The monochromophoric p-C2P1 and p-C1P1 show no dependence on the excitation energy at the selected

Partial Amplitude α3

0.3 0.2 0.1 0.0

–0.1 0

50

100

150 200 250 Excitation Energy (nJ)

300

350

Figure 1.24. Dependence of the amplitude a3 of component 3 from the laser excitation energy for the compounds p-C2P4 [&], p-C2P1 [~], and p-C1P1 [&] at detection wavelength 590 nm.

FEMTOSECOND TRANSIENT ABSORPTION MEASUREMENTS

37

detection wavelength while p-C2P4 clearly shows a dependence at these wavelengths (Fig. 1.24). In view of the typical power dependence, the 40 ps component of p-C2P4 can also be attributed to an annihilation process. Because of the dependence of a2 and a3 on the excitation energy in p-C2P4, both the 5 and 40 ps components can be attributed to singlet–singlet annihilation processes. Since this is a Fo¨rster allowed excitation energy transfer process, it has to be distance dependent. The appearance of two annihilation processes in p-C2P4 probably relates to the presence of constitutional isomers, which gives a broader distribution of distances between neighboring chromophores compared to that of p-C1P4. As a result, besides a fast (5 ps) annihilation process occurring between chromophores at short distances, similar to p-C1P4 but less important in p-C2P4, an additional annihilation process (50 ps) is resolved, which can be attributed to interactions between chromophores at longer distance. Two possible structures for isomers with a short- and a long-distance pair of chromophores are depicted in Figure 1.16a, a0 and 1.16b, b0 , respectively. The relative contribution of the short annihilation process indicates approximately 10–15% of isomers where the two chromophores are at shorter distances. The fourth and the longest component (t4, a4) is in the range of a few nanoseconds and thus cannot be determined precisely in the time windows used here. Instead, the actual values were taken from measurements performed using a single-photon timing detection setup and is attributed to the intrinsic fluorescence lifetime of the peryleneimide chromophore equal to 4.2 ns.

VI.


So far, the photophysical properties of the dendrimers were investigated using the fluorescence signal either by single-photon timing (SPT) or fluorescence upconversion. Transient absorption is used to validate the presence of the annihilation process, to allow quantifying the spectral overlap between emission and absorption of the S1 state, the basis of singlet–singlet annihilation, and to evaluate the influence of the substitution pattern and the number of PI chromophores on the transient absorption properties of these dendrimers [40].

A. p-C1P1 and m-C1P1 The wavelength-dependent absorption changes are presented in Figure 1.25 for a number of different delay times after excitation. At positive times, two different parts in the transient spectrum can be seen: a negative signal extending from

38


Figure 1.25. Transient absorption spectra of p-C1P1 at different delay times: 10 ps (&), 1 ps (*), 2 ps (~), 5 ps (q), 10 ps (^), and 30 ps (3). Inset: Detailed display of the 520–580 nm region.

450 to 600 nm and a positive signal beyond 600 nm with a maximum approximately at 660 nm. In first approximation, both features can be seen instantaneously after excitation and decay on a nanosecond time scale. Since the signal in the transient absorption spectrum above 600 nm is positive, it can predominantly be attributed to an excited state absorption (ESA) process. From previous studies, it is known that p-C1P1 has a fluorescence quantum yield of almost unity and a fluorescence lifetime of 4.2 ns; thus, the ESA found here can be attributed to S1–Sn absorption within the peryleneimide chromophore. As the steady state absorption spectrum shows no intensity above 560 nm while the fluorescence spectrum extends from 510 to 750 nm, the negative signal in the transient spectrum cannot solely be attributed to ground state bleaching. It seems reasonable to assume that ground state bleaching dominates the signal between 450 and 510 nm. Above 510 nm, both ground state bleaching and stimulated emission are responsible for the negative signal, whereas the signal in the range between 560 and 600 nm is dominated by stimulated emission. There is no reason to assume that stimulated emission would only occur in the very blue part of the fluorescence spectrum, so it must be considered that


39

there also is a contribution of stimulated emission above 600 nm. However, as the absolute value of the cross section for excited state absorption at these wavelengths exceeds the stimulated emission, the net result of transient absorption and stimulated emission is a large positive signal in this wavelength range. As stated earlier, the transient signal mainly decays on a nanosecond time scale. However, a detailed decay analysis of the transient absorption intensities as a function of delay time for different wavelengths reveals an additional picosecond relaxation process, which can most clearly be seen in the inset of Figure 1.25. Within the first 20 ps, the transient absorption intensity drops at a wavelength of 530 nm, at about 555 nm the intensity remains the same, while at 570 nm it rises. This relaxation process has been described before in detail [41] and is interpreted as a combination of vibrational and solvent relaxation. This feature, where the intensity decays at a given wavelength and rises at another wavelength with an identical time constant (6.3 ps), had been found (vide supra) in fluorescence upconversion experiments [30]. In the transient absorption data discussed, the same feature can be observed; however, the signs of the amplitudes are of course reversed. The results of the measurements for m-C1P1 are very similar to those of the para compound and the data sets can be interpreted identically. The obtained time constant of this vibrational/solvent relaxation process is 10 ps as found previously in fluorescence upconversion experiments [30]. For the compound m-C1P1, the maximum of the positive transient absorption band attributed to the S1–Sn absorption is shifted about 5 nm to the blue and also the zero crossing point is shifted from 610 nm in the case of p-C1P1 to 602 nm for m-C1P1.

B. p-C1P3 and m-C1P3 Another series of experiments was performed on p-C1P3, which contains three peryleneimides at the rim. Comparing the transient absorption spectra of this compound (see Fig. 1.26 top) to those of p-C1P1 (Fig. 1.26, bottom), one can see that the general shape is identical. Since the same chromophore is involved, the attribution of the signals in p-C1P3 can be the same as for p-C1P1. However, the transient absorption signal of p-C1P3 is two times lower in intensity than that of p-C1P1. This suggests the occurrence of an additional decay channel, which in view of the results discussed earlier can be attributed to an ultrafast singlet–singlet annihilation process. The temporal evolution of the transient spectra of p-C1P3 and p-C1P1 is grossly different. It seems that the signal in the multichromophoric dendrimer at 530 and 650 nm decays faster when compared to p-C1P1. This feature is demonstrated in Figure 1.27, where the transient absorption intensity as a function of time is plotted for p-C1P3 and p-C1P1 at detection wavelengths of 530 nm (top) and 650 nm (bottom). In

40


Figure 1.26. Transient absorption spectra of p-C1P3 (top) and p-C1P1 (bottom) at different delay times: 10 ps (&), 1 ps (*), 2 ps (~), 5 ps (q), 10 ps (^), and 30 ps (3).

accordance with previous findings [25, 30], this feature is attributed to singlet– singlet annihilation between two excited states within one dendrimer, leading to a first excited state and a ground state. In order to confirm this attribution, an additional series of experiments was performed in which, in analogy to the upconversion experiments, only the excitation intensity impinging on the sample was decreased by a factor of 5. These measurements were performed at 530 nm (maximum of the negative part of the transient signal) and 650 nm (maximum of the positive part of the transient spectrum). The results of these measurements are also collected in Figure 1.27. The decays of p-C1P1 are independent of the excitation intensity in contrast to the ones of p-C1P3. This is further strong support for the earlier made assumption of singlet–singlet annihilation. Thus, at 530 nm the annihilation process reduces the number of excited peryleneimide chromophores, leading to a decrease in both stimulated emission and ground state bleaching. At 650 nm it can be understood as a decrease in the amount of peryleneimides in the excited state. This singlet–singlet annihilation process is the additional decay channel in the transient absorption measurements of p-C1P3 compared to p-C1P1. The relative decrease of the signal due to the singlet–singlet


41

Figure 1.27. Excitation intensity-dependent plot of the normalized transient absorption signals as a function of time at high (*, &, q, ~) and low (*, & ,s, ~) excitation power recorded at 530 nm (top) and at 650 nm (bottom) for the dendrimers p-C1P3 (*, *) and p-C1P1 (&, &).

42


annihilation at longer times is, at high excitation intensity, only 70% of the signal observed for p-C1P1.

C. m-C1P3 A similar series of experiments were performed on m-C1P3, which contains three peryleneimides (PI) connected in meta position to the outer phenyl ring, leading to exactly the same picture as derived from the comparison of p-C1P3. This includes the same general attribution of the transient bleaching and absorption signals as in m-C1P1 and the occurrence of singlet–singlet annihilation in m-C1P3, which is also evidenced by an additional excitation intensity-dependent study at a detection wavelength of 650 nm. The differences in the photophysical properties due to the different substitution patterns can be determined by comparing the compounds (m-C1P1, 3) with the compounds (p-C1P1,3). Already in the emission spectra, a bathochromic shift of the latter can be observed while the ground state absorption spectra are identical. They also show a less pronounced vibrational structure of the emission spectra. Differences can also be seen in the transient behavior of these compounds. The maximum of the positive transient absorption band for the compound m-C1P1 is shifted about 5 nm to the blue and also the zero crossing point is shifted from 610 nm in the case of p-C1P1 to 602 nm in the case of m-C1P1. The influence of the different substitution pattern upon the fluorescence dynamics of these dendrimers was discussed in detail using SPT and fluorescence upconversion detection (vide supra). The transient absorption measurements reported here show very similar features, thus confirming the above interpretation. The para coupling leads to a better conjugation of the p-electrons of the peryleneimide over the aromatic phenyl ring of the branch. This better conjugation lowers the excited state energy, leading to a bathochromic shift of the emission spectrum. This is illustrated by the 5 nm shift of the zero crossing point of the transient absorption spectrum from m-C1P1 compared to p-C1P1. In the meta-substituted compounds m-C1Pn, steric hindrance between the hydrogens of PI and the 2,6-phenyl rings on the second phenyl of the dendritic arm will disrupt the conjugation between PI and its 9-phenyl ring (Fig. 1.8).

D. p-C2P1 and p-C2P4 In order to reveal the influence of the generation number on the photophysical properties, the compound p-C2P1 was studied. It is a monochromophoric second generation dendrimer consisting of an interior building block and one


43

Figure 1.28. Transient absorption spectra of p-C2P4 (top) and p-C2P1 (bottom) at different delay times: 10 ps (&), 1 ps (*), 5 ps (~), 20 ps (q), 50 ps (^), and 400 ps (3).

peryleneimide chromophore attached in a para position to the outer phenyl ring at the rim. The time-dependent transient absorption spectra are shown in Figure 1.28 (bottom). At positive times, two different parts in the transient spectrum can be seen: a negative signal that reaches from 450 to 600 nm with different maxima at about 500 nm, and a positive signal beyond 600 nm with a maximum at approximately 660 nm. Both features appear instantaneously after excitation and decay on a nanosecond time scale. Another series of experiments were performed on p-C2P4, which contains four peryleneimides at the rim. The time-dependent transient absorption spectra are displayed in Figure 1.28 (top ). Comparing the transient absorption spectra of p-C2P4 to those of p-C2P1 (Figure 1.28, bottom), one can see that the general shape is identical; thus, the attribution of the signals in p-C2P4 can be the same as in p-C2P1. The initial drop in transient absorption signal for p-C2P4 compared to p-C2P1 is smaller than in their first generation counterpart. This can be explained by the smaller relative contribution of the singlet–singlet annihilation process in p-C2P4 compared to p-C1P4. These general features are exactly the same as those observed for the first generation monochromophoric dendrimer p-C1P1. Since the chromophore involved, the steady state spectra, and quantum

44


yields are identical, the above information leads to the identical attributions of the negative signal to ground state bleaching/stimulated emission and of the positive signal to S1–Sn absorption as in the case of p-C1P1. In the short wavelength part of the signal related to ground state bleaching, however, we find a more intense signal as compared to the transient absorption spectrum of p-C1P1. This negative spectrum also resembles the ground state absorption spectrum in a more precise way, which seems reasonable assuming there is no excited state absorption in this spectral region. The temporal evolution of the spectra, however, is different. The change is completely analogous to the one reported above for the comparison of first generation mono- to multichromophoric dendrimers. In fluorescence upconversion (vide supra), two intensity-dependent annihilation processes—one with a time constant of 40 ps and by far the main component and a second, minor amplitude process with a time constant of about 5 ps—were observed. In the transient absorption measurements reported here, only two different excitation intensities were investigated from which we can neither exclude nor claim the presence of a second minor annihilation component. The occurrence of more than one annihilation process might relate to the presence of different isomers resulting in a broad distribution of rate constants, which under certain conditions would be analyzed as two annihilation processes. The influence of the generation number can be deduced by comparing the results of the second generation dendrimers (p-C2P1, 4) to those of the first generation dendrimers (p-C1P1, 3). While in the monochromophoric compounds no difference can be observed between the first and second generation dendrimer, in the multichromophoric dendrimers a clear dependence of the annihilation process on the generation number can be observed. Although this process is seen in both generations, the corresponding time scales are grossly different: while in p-C1P3 the annihilation process relates to a decay time of 4.2 ps, it corresponds to a decay time of 53 ps for p-C2P4. These two decay times result from global analysis of the transient absorption decays obtained at different probe wavelengths. For p-C1P3 and p-C2P4, the decay times of the annihilation processes determined by fluorescence upconversion were 4.6 ps and 40 ps, respectively [30, 42], which is in good agreement with the transition absorption data. Why is singlet–singlet annihilation faster than energy hopping? Since the distribution of distance is identical for both processes, one can visualize the difference based on the overlap between the emission and absorption. The transient absorption data allow extracting the absorption spectrum of the S1 to Sn transition and if we compare the spectral overlap for this transition with the emission (Fig. 1.29) with that for the S0 to S1 transition (Fig. 1.14), one immediately sees that the overlap integral is substantially larger for the annihilation process, hence leading to larger rate constants. As due to residual induced emission at longer wavelength, the extinction coefficient of the S1–Sn absorption can only

45

CONCLUSIONS

60 energy

S1

transfer

internal

S0

S2

conversion

S0

S1

Fluorescence Intensity (a.u.)

S1

50

30

ε (103/Mcm)

40

20

10

350

400

450

500

550

600

650

700

0 750

Wavelength (nm)

Figure 1.29. Overlap between the fluorescence spectrum of p-C1P1 and its transient absorption spectrum of the S1–Sn transition. Inset: The singlet–singlet annihilation process.

be underestimated; the rate for singlet–singlet annihilation can even exceed the values estimated here.

VII. CONCLUSIONS Ensemble photophysics of two series of rigid dendrimers with an identical rigid central sp3 core and substituted with peryleneimide chromophores at the meta (m-C1Px) and para (p-CnPx) position of the outer phenyl ring have been investigated by steady state and nanosecond to femtosecond time-resolved spectroscopic techniques. This series of molecules were synthesized to investigate chromophore–chromophore interactions and to validate models describing such processes. A complicating factor in the synthesis due to two different modes of Diels–Alder addition led to a mixture of constitutional isomers, which could not be separated. This means that even if all prerequisites for the application of the Fo¨rster model are fulfilled, the resulting rate constants will be average values. Similar limitations will exist for all nonrigid dendritic structures, where in

46


solution a distribution of conformations present in a bulk experiment will lead to a distribution of distances. This was observed in directional excitation transfer in the semirigid dendrimers represented in Figure 1.7 [17]. This distribution could be resolved, however, in single molecule experiments [18]. In the time-resolved single-photon counting measurements, the meta-substituted dendrimers showed a contribution of an emission from an ‘‘excimer-like’’ species resulting from chromophore–chromophore interaction with a decay time of 7.4 ns, beside the emission of the individual phenyl-substituted peryleneimide chromophores with 4.2 ns, whereas for the para-substituted ones no such state has been observed. This suggests that a fraction of the molecules containing more than one chromophore show a stronger interaction and consequently can not be described using the weak coupling condition. Study of the same molecules at the single molecule level [43] underpins this conclusion. The excitation of the peryleneimide chromophore results in excitation energy hopping among similar chromophores for both dendrimer series. In the first generation para-substituted dendrimers, this energy hopping takes place among all peryleneimide chromophores with a hopping rate constant experimentally determined to be khopp ¼ 4:6 ns1 . This value is in accordance with rate constants theoretically derived on the basis of molecular modeling structures. By comparing to polyphenylene dendrimers, where the peryleneimide chromophores are attached in meta instead of para position, the importance of the dipole orientation factor k2 could experimentally be demonstrated in excellent agreement with the theoretical Fo¨rster equation. While the value of k2 ¼ 0:8 in the meta series yields a hopping rate constant of khopp ¼ 2 ns1 , the improved orientation of peryleneimide chromophores in the para series yields a larger k2 value of about 2.1, leading to a more than two times faster hopping dynamics in spite of a larger average distance between the chromophores. To determine the influence of the distance between the chromophores in these dendrimers on intramolecular energy hopping, a series of second generation para-substituted peryleneimide dendrimers with rigid tetrahedral core (p-C2Px) were investigated. The energy transfer process could be explained in terms of Fo¨rster-type energy transfer and the average values obtained for khopp scale properly with the sixth power of the distance ratio between first and second generation. The short time scale dynamics have been studied by means of femtosecond fluorescence upconversion. For all dendrimers these measurements revealed size-independent kinetic processes related to an internal vibrational redistribution, a vibrational/solvent relaxation. Singlet–singlet annihilation, only present in the multichromophoric compounds, was established by an excitation energydependent study. It has been shown that this type of process contributes to a larger extent in the para-substituted dendrimers compared to the meta-substituted ones. These differences between the meta- and para-substituted dendrimers

REFERENCES

47

demonstrate the important role of the spatial distribution of the chromophores at the periphery in the dynamics of the photophysical processes involved. Moreover, in the multichromophoric second generation p-C2P4, a dual annihilation process was observed. The fast annihilation process occurs between a short distance pair of chromophores comparable in distance to the one in p-C1P4, while the longer time annihilation process occurs among the more prevalent pair of chromophores at longer distance. The origin of this can be traced back to the distribution of constitutional isomers as a result of the synthesis as mentioned earlier. The presence of a generation-dependent annihilation process and the influence of the substitution pattern have been validated by femtosecond timeresolved transient absorption measurements. Two other Fo¨rster allowed processes can occur in multichromophoric systems under condition of multiple excitations, namely, singlet–triplet quenching and singlet ion/radical quenching if either the triplet or ion/radical absorption spectra do overlap with the fluorescence spectrum of the donor. These processes were not observed for these systems at the ensemble level because of the low probability of formation of these species resulting in a small relative abundance at the ensemble level. However, at the single molecule level they could be visualized [25, 44].

ACKNOWLEDGMENTS All compounds discussed were synthesized in the research group of Prof. K. Mu¨llen to whom we are greatly indebted. This fruitful and exciting collaboration was made possible through a Max Planck Research Award to FDS and a IAPV-03 grant by the Fedral Science Policy Agency. We are also indebted to many co-workers whose names are mentioned in the references to the original papers and to D. Beljonne and S. Mukamel for computational support.

REFERENCES 1. (a) F. C. De Schryver, N. Boens, J. Huybrechts, J. Daemen, and M. De Brackeleire, Pure Appl. Chem. 1977, 49, 237–247. (b) F. C. De Schryver, N. Boens, and J. Huybrechts, In: Advances in Photochemistry, Vol. 10, J. N. Pitts, G. S. Hammond, and K. Gollnick, Eds., John Wiley & Sons, Hoboken, NJ, pp. 359–465. 2. (a) S. Hecht and J. M. J. Frećhet, Angew. Chem. Int. Ed. 2001, 40, 74. (b) A. P. H. J. Schenning, E. Peeters, and E. W. Meijer, J. Am. Chem. Soc. 2000, 122, 4489.

48

3.

4.

5.

6. 7.

8.

9.

10.

11. 12.


(c) M.-S. Choi, T. Aida, T. Yamazaki, and I. Yamazaki, Chem. Eur. J. 2002, 8, 2668. (d) L. F. Lee, A. Adronov, R. D. Schaller, J. M. J. Frećhet, and R. J. Saykally, J. Am. Chem. Soc. 2003, 125, 536. (e) V. Balzani, S. Campagna, G. Denti, A. Juris, S. Serroni, and M. Venturi, Acc. Chem. Res. 1998, 31, 26. (f) G. M. Stewart and M. A. Fox, J. Am. Chem. Soc. 1996, 118, 4354. (g) R. Gronheid, J. Hofkens, F. Ko¨hn, T. Weil, E. Reuther, K. Mu¨llen, and F. C. De Schryver, J. Am. Chem. Soc. 2002, 124, 2418. (h) T. Weil, E. Reuther, and K. Mu¨llen, Angew. Chem. Int. Ed. 2002, 41, 1900. (a) Z. Xu and J. S. Moore, Acta Polym. 1994, 45, 83. (b) J. S. Melinger, Y. Pan, V. D. Kleiman, Z. Peng, B. L. Davis, D. McMorrow, and M. Lu, J. Am. Chem. Soc. 2002, 124, 12002. (a) M.-H. Xu, J. Lin, Q.-S. Hu, and L. Pu, J. Am. Chem. Soc. 2002, 124, 14239. (b) M. Kimura, T. Shiba, T. Muto, K. Hanabusa, and H. Shirai, Macromolecules 1999, 32, 8237. (a) E. M. Harth, S. Hecht, B. Helms, E. E. Malmstrom, J. M. J. Frećhet, and C. J. Hawker, J. Am. Chem. Soc. 2002, 124, 3926. (b) D.-L. Jiang and T. Aida, Nature 1997, 388, 454. D. Liu, S. De Feyter, M. Cotlet, A. Stefan, U.-M. Wiesler, A. Herrmann, D. GrebelKoehler, J. Qu, K. Mu¨llen, and F. C. De Schryver, Macromolecules 2003, 36, 5918. (a) R. Kopelman, M. R. Shortreed, Z.-Y. Shi, W. Tan, Z. Xu, and J. S. Moore, Phys. Rev. Lett. 1997, 78, 1239. (b) M. R. Shortreed, S. F. Swallen, Z.-Y. Shi, W. Tan, Z. Xu, C. Devadoss, J. S. Moore, and R. Kopelman, J. Phys. Chem. B 1997, 101, 6318. (c) S. F. Swallen, Z.-Y. Shi, W. Tan, Z. Xu, J. S. Moore, and R. Kopelman, J. Luminescence 1998, 76=77, 193. (d) S. F. Swallen, Z. Zhu, J. S. Moore, and R. Kopelman, J. Phys. Chem. B 2000, 104, 3988. (a) K. M. Gaab, A. L. Thompson, J. Xu, T. J. Martinez, and C. J. Bardeen, J. Am. Chem. Soc. 2003, 125, 9288. (b) A. Thompson, K. M. Gaab, J. Xu, C. J. Bardeen, and T. J. Martinez, J. Phys. Chem. A 2004, 108, 671. (a) V. Balzani and F. Scandola, Supramolecular Chemistry, Ellis Horwood, New York, 1991. (b) V. Balzani, S. Campagna, G. Denti, A. Juris, S. Serroni, and M. Venturi, Acc. Chem. Res. 1998, 31, 26. (c) V. Balzani, P. Ceroni, S. Gestermann, C. Kauffmann, M. Gorka, and F. Vo¨gtle, Chem. Commun. 2000, 10, 853. (d) F. Vo¨gtle, S. Gestermann, C. Kauffmann, P. Ceroni, V. Vicinelli, and V. Balzani, J. Am. Chem. Soc. 2000, 122, 10938. (e) M. Venturi, S. Serroni, A. Juris, S. Campagna, and V. Balzani, Top. Curr. Chem. 1998, 197, 193. (a) S. Campagna, S. Serroni, M. Ciano, A. Juris, and V. Balzani, Inorg. Chem. 1992, 31, 2982. (b) V. Balzani, A. Juris, M. Venturi, S. Campagna, and S. Serroni, Chem. Rev. 1996, 96, 759. G. Bergamini, P. Ceroni, M. Maestri, V. Balzani, S.-K. Lee, and F. Vo¨gtle, Photochem. Photobiol. Sci. 2004, 3, 898. (a) S. Gilat, A. Adronov, and J. M. J. Frećhet, Angew. Chem. Int. Ed. 1999, 38, 1422. (b) A. Adronov, S. L. Gilat, J. M. J. Frećhet, K. Ohta, F. V. R. Neuwahl, and G. R. Fleming, J. Am. Chem. Soc. 2000, 122, 1175. (c) F. V. R. Neuwahl, R. Righini, A. Adronov, P. R. L. Malenfant, and J. M. J. Frećhet, J. Phys. Chem. B 2001, 105, 1307.

REFERENCES

49

13. G. S. He, T.-C. Lin, Y. Cui, P. N. Prasad, D. W. Brousmiche, J. M. Serin, and J. M. J. Frećhet, Opt. Lett. 2003, 28, 768. 14. E. K. L. Yeow, K. P. Ghiggino, J. N. H. Reek, M. J. Crossley, A. W. Bosman, A. P. H. J. Schenning, and E. W. Meijer, J. Phys. Chem. B 2000, 104, 2596. 15. (a) O. Varnavski, G. Menkir, and T. Goodson III, Appl. Phys. Lett. 2000, 77, 1120. (b) T. Goodson III, Acc. Chem. Res. 2005, 38, 99. 16. B. K. Kaletas, R. Dobrawa, A. Sautter, F. Wu¨rthner, M. Zimine, L. De Cola, and R. M. Williams, J. Phys. Chem. A 2004, 108, 1900. 17. (a) S. Jordens, G. De Belder, M. Lor, G. Schweitzer, M. van der Auweraer, T. Weil, E. Reuther, K. Mu¨llen, and F. C. De Schryver, Photochem. Photobiol. Sci. 2003, 2, 177. (b) G. Schweitzer, R. Gronheid, S. Jordens, M. Lor, G. De Belder, T. Weil, E. Reuther, K. Mu¨llen, and F. C. De Schryver, J. Phys. Chem. A 2003, 107, 3199. 18. M. Cotlet, R. Gronheid, S. Habuchi, A. Stefan, A. Barbafina, K. Mu¨llen, J. Hofkens, and F. C. De Schryver, J. Am. Chem. Soc. 2003, 125, 13609. 19. (a) M. Lippitz, C. G. Hu¨bner, Th. Christ, H. Eichner, P. Bordat, A. Herrmann, K. Mu¨llen, and Th. Basche´, Phys. Rev. Lett. 2004, 92, 103001. (b) Th. Basche´, Nanotech Insights Abstract 2005, p. 53. 20. Th. Fo¨rster, Ann. Phys. 1948, 2, 55. 21. G. D. Scholes, Annu. Rev. Phys. Chem. 2003, 54, 57. 22. B. P. Krueger, G. D. Scholes, and G. R. Fleming, J. Phys. Chem. B 1998, 102, 5378. 23. D. Beljonne, J. Cornil, R. Silbey, P. Millie´, and J. L. Bre´das J. Chem. Phys. 2000, 112, 474. 24. (a) S. Tretiak, V. Chernyak, and S. Mukamel, J. Am. Chem. Soc. 1997, 119, 11408. (b) S. Tretiak, C. Middleton, V. Chernyak, and S. Mukamel, J. Phys. Chem. B 2000, 104, 9540. (c) S. Tretiak, W. M. Zhang, V. Chernyak, and S. Mukamel, Proc. Nat. Acad. Sci. U.S.A. 1999, 96, 13003. (d) Y. Zhao, T. Meier, W. M. Zhang, V. Chernyak, and S. Mukamel, J. Phys. Chem. B 1999, 103, 3954. 25. (a) J. Hofkens, M. Cotlet., T. Vosch, P. Tinnefeld, K. D. Weston, C. Ego, A. Grimsdale, K. Mu¨llen, D. Beljonne, J.-L. Bre´das, S. Jordens, G. Schweitzer, M. Sauer, and F. C. De Schryver, Proc. Natl. Acad. Sci. U.S.A. 2003, 100, 13146. (b) T. Vosch, M. Cotlet, J. Hofkens, K. Van Der Biest, M. Lor, K. Weston, Ph. Tinnefeld., M. Sauer, L. I. Latterin, K. Mu¨llen, and F. C. De Schryver, J. Phys. Chem. 2003, 107, 36,6920. 26. Y. Karni, S. Jordens., G. De Belder, G. Schweitzer, J. Hofkens, T. Gensch, M. Maus, F. C. De Schryver, A. Herrmann, and K. Mu¨llen, Chem. Phys. Lett. 1999, 310, 73. 27. (a) D. L. Andrews and A. D. Demidov, Eds., Resonance Energy Transfer, John Wiley & Sons, Hoboken, NJ, 1999. (b) H. C. Cheung, In: Topics in Fluorescence Spectroscopy, Vol. 2, J. R. Lakowicz, Ed., Plenum Press, New York, 1991, p. 127. 28. M. Maus, R. De, M. Lor, T. Weil, S. Mitra. U.-M. Wiesler, A. Herrmann, J. Hofkens, T. Vosch, K. Mu¨llen, and F. C. De Schryver, J. Am. Chem. Soc. 2001, 123, 7668. 29. M. Maus, S. Mitra, M. Lor, J. Hofkens, T. Weil, A. Herrmann, K. Mu¨llen, and F. C. De Schryver, J. Phys. Chem. A 2001, 105, 3961. 30. (a) G. De Belder, G. Schweitzer, S. Jordens, M. Lor, S. Mitra, J. Hofkens, S. De Feyter, M. van der Auweraer, A. Herrmann, T. Weil, K. Mu¨llen, and F. C. De Schryver, Chem.

50

31. 32.

33. 34. 35.

36. 37. 38.

39.

40.

41. 42.


Phys. Chem. 2001, 1, 49. (b) G. De Belder, S. Jordens., M. Lor, G. Schweitzer, R. De, T. Weil, A. Herrmann, U.-M. Wiesler, K. Mu¨llen, and F. C. De Schryver, J. Photochem. Photobiol. A: Chem. 2001, 145, 61. (a) J. S. Baskin, L. Banares, S. Pedersen, and A. H. Zewail, J. Phys. Chem. 1996, 100, 11920. (b) D. J. Nesbitt and R. W. Field, J. Phys. Chem. 1996, 100, 12735. (a) M. L. Hornig, J. A. Gardecki, A. Papazyan, and M. Maroncelli, J. Phys. Chem. 1995, 99, 17311. (b) T. Nakabayashi, H. Okamoto, M. Tasumi, J. Phys. Chem. A 1997, 101, 3494. R. M. Stratt and M. Maroncelli, J. Phys. Chem. 1996, 100, 12981. P. K. McCarthy and G. J. Blanchard, J. Phys. Chem. 1996, 100, 14592. (a) P. K. McCarthy and G. J. Blanchard, J. Phys. Chem. 1996, 100, 14592. (b) T. Gustavsson, G. Baldacchino, J. C. Mialocq, and S. Reekmans, Chem. Phys. Lett. 1995, 236, 587. (c) W. Jarzeba, G. C. Walker, A. E. Johnson, M. A. Kahlow, and P. F. Barbara, J. Phys. Chem. 1998, 92, 7039. (d) Y. Kimura, J. C. Alfano, P. K. Walhout, and P. F. Barbara, J. Phys. Chem. 1994, 98, 3450. (e) L. Reynolds, J. A. Gardecki, S. J. V. Frankland, M. L. Horng, and M. Maroncelli, J. Phys. Chem. 1996, 100, 10337. (f) P. Changenet, P. Plaza, M. M. Martin, and Y. H. Meyer, J. Phys. Chem. A 1997, 101, 8186. (g) P. Changenet, H. Zhang, M. J. Van der Meer, K. J. Hellingwerf, and M. Glasbeek, Chem. Phys. Lett. 1998, 282, 276. R. D. Harcourt, K. P. Ghiggino, G. D. Scholes, and R. P. Steer, J. Chem. Phys. 1998, 109, 1310. G. Paillotin, C. E. Swenberg, J. Breton, and N. E. Geacintov, Biophys. J. 1979, 25, 513. (a) V. Gulbinas, L. Valkunas, D. Kuciauskas, E. Katilius, V. Liuolia, W. L. Zhou, and R. E. Blankenship, J. Phys. Chem. 1996, 100, 17950. (b) L. Valkunas and V. Gulbinas, Photochem. Photobiol. 1997, 66, 628. (c) V. Barzda, G. Garab, V. Gulbinas, and L. Valkunas, Biochim. Biophys. Act. Bioenerg. 1996, 1273, 231. (d) V. Barzda, G. Garab, V. Gulbinas, and L. Valkunas, Biochim. Biophys. Act. Bioenerg. 1996, 1273, 231. (e) W. H. J. Westerhuis, M. Vos, R. Van Grondelle, J. Amesz, and R. A. Niederman, Biochim. Biophys. Act. Bioenerg. 1998, 1366, 317. (f) A. Ruseckas, M. Theander, L. Valkunas, M. R. Andersson, O. Inganas, and V. Sundstrom, J. Lumin. 1998, 76=77, 474. (g) A. Ruseckas, M. Theander, L. Valkunas, M. R. Andersson, O. Inganas, and V. Sundstrom, J. Lumin. 1998, 76=77, 474. (a) V. Sundstro¨m, T. Gillbro, R. A. Gadonas, and A. Piskarskas, J. Phys. Chem. 1988, 89, 2754. (b) G. Scheblykin, O. P. Varnavsky, M. M. Bataiev, O. Sliusarenko, M. van der Auweraer, and A. G. Vitukhnovsky, Chem. Phys. Lett. 1998, 298, 341. S. Jordens, G. De Belder, M. Lor, G. Schweitzer, M. van der Auweraer, T. Weil, A. Herrmann, U.-W. Wiesler, K. Mu¨llen, and F. C. De Schryver, Photochem. Photobiol. Sci. 2003, 2, 1118. S. Jordens, G. De Belder, M. Lor, G. Schweitzer, M. van der Auweraer, T. Weil, E. Reuther, K. Mu¨llen, and F. C. De Schryver, Photochem. Photobiol. Sci. 2003, 2, 177. M. Lor, R. De, S. Jordens, G. De Belder, G. Schweitzer, M. Cotlet, J. Hofkens, T. Weil, A. Herrmann, K. Mu¨llen, M. van der Auweraer, and F. C. De Schryver, J. Phys. Chem. A 2002, 106, 2083.

REFERENCES

51

43. T. Vosch, J. Hofkens, M. Cotlet, F. Ko¨hn, H. Fujiwara, R. Gronheid, K. Van Der Biest, T. Weil, A. Herrmann, K. Mu¨llen, S. Mukamel, M. van der Auweraer, and F. C. De Schryver, Angew. Chem. Int. Ed. 2001, 40, 4643. 44. F. C. De Schryver, T. Vosch, M. Cotlet, M. van der Auweraer, K. Mu¨llen, and J. Hofkens, Acc. Chem. Res. 2005, 38, 514–522.

PHOTOCHEMISTRY OF SHORT-LIVED SPECIES USING MULTIBEAM IRRADIATION Mamoru Fujitsuka and Tetsuro Majima The Institute of Scientific and Industrial Research (SANKEN), Osaka University, Mihogaoka 8-1, Ibaraki, Osaka 567-0047, Japan

CONTENTS I. Introduction II. Pulse Radiolysis–Laser Flash Photolysis A. Excited Radical Cations B. Fluorescence from Excited Radical Cations C. Excited Radical Anions III. Two-Color Two-Laser Flash Photolysis A. Higher Triplet Excited States ðTn Þ 1. Energy Transfer from the Tn State 2. Energy Gap Law 3. Substituent Effect on the Tn State Lifetimes 4. Bond Dissociation from the Tn State 5. Electron Transfer from the Tn State 6. Direct Observation of the Tn State B. Ketyl Radicals in the Excited State C. Excited Radical Cations D. Other Reactive Intermediates

Advances in Photochemistry, Volume 29 Edited by Douglas C. Neckers, William S. Jenks, and Thomas Wolff # 2007 John Wiley & Sons, Inc.

53

54

PHOTOCHEMISTRY OF SHORT-LIVED SPECIES USING MULTIBEAM IRRADIATION

1. 2.

Two-Color Two-Laser DNA Damaging Two-Color Laser Photolysis for Determination of the Rate Constant from the Product Analysis 3. Two-Color Laser Control of Photocatalytic Reaction on TiO2 Surface IV. Three-Color Three-Laser Flash Photolysis A. Three-Laser Control of Intermediate Population B. Stepwise Bond Cleavage of Two C—O Bonds Via the Tn State V. Conclusions Acknowledgments References

I. INTRODUCTION One important technology development is the laser. Today, we see various lasers in use everywhere. In science, pulse lasers have been used for the generation of various chemical intermediates by photoexcitation [1]. From laser flash photolysis experiments, various interesting reactions have been revealed. Recent improvements in the stability and pulse duration of lasers have made experiments quite easy to perform. For example, pulse duration decreased almost six orders of magnitude, going from the nanosecond to femtosecond regime. Furthermore, the pump and probe method, using a stable femtosecond laser, gives reliable transient absorption spectra even when spectral change is quite subtle, on the order of 103 absorbance. Because operation of pulse lasers is quite easy, scientists in various fields can use laser systems with the expectation of high performance. An electron beam from a linear accelerator is another powerful tool to generate reactive intermediates, such as radical cations and radical anions [2]. By selecting appropriate reaction conditions, including the solvent and gas atmosphere, selective generation of an intermediate can be achieved, since the reaction pathways generating these intermediates are well established. Furthermore, concentration of generated intermediates can be increased in relatively large volume even when the intermediate is difficult to generate with other methods. These points are quite useful; however, use of pulse radiolysis is not common among scientists. By combination of these pulse techniques (multibeam irradiation), multistep excitation of intermediates can be achieved. For example, laser excitation of an intermediate generated during pulse radiolysis can be realized. For years, chemistry based on multibeam irradiation methods has been investigated [3]. There are several advantages for the multibeam irradiation method.

55

INTRODUCTION

LUMO HOMO

M(S0)

M(S1)

M(Tn )

M(T1)

M.+(D0)

M.+(D1)

Figure 2.1. Electronic structures of M(S0), M(S1), M(T1), MðTn Þ, M þ(D0), and M þ*(D1).

First, by employing multibeam irradiation, higher excited states, which cannot be accessed by the single-pulse excitation method, can be generated. Excited doublet and higher triplet excited states are examples of such intermediates (Fig. 2.1). For these higher excited states, various reactions, which do not proceed from the lowest excited states, are expected (Fig. 2.2). Isomerization, bond dissociation, rearrangement, and ionization are expected as intramolecular reactions via the higher excited states [4–14]. As for intermolecular reactions via the higher excited states, energy transfer, electron transfer, and hole transfer processes have been investigated [15–17]. Second, it should be pointed out that yield and selectivity of these photoinduced processes caused by the multibeam irradiation depend largely on the delay time of the second beam irradiation with respect to the first beam. Because an intermediate generated by the first beam excitation has a finite lifetime, concentration of the targeted intermediate depends on the delay time of the second laser irradiation after the first beam excitation. That is, chemical reaction induced by the multibeam irradiation is a ‘‘time-selective’’ process (Fig. 2.3). This feature is

3

M**

h v2 1

M*

Products x x

3

h v1

M*

M

Figure 2.2. Schematic illustration of a reaction caused by multibeam excitation. Reaction from the higher triplet excited state generated during two-color two-laser flash photolysis was representative.

56


hv2

hv1 deactivation

hv2 B

C

B

low

A* concentration

A

C

selectivity

high

0 Second beam delay from first beam A*

First beam irradiation

time

Figure 2.3. Schematic illustration describing the ‘‘time-selective’’ process of multibeam chemistry. When the second beam produces different products B and C from A and A*, respectively, as indicated in the scheme on the left, selectivity depends on the second beam delay from the first beam irradiation.

one of the unique points of the chemical reactions induced by multibeam irradiation and can be regarded as a distinct advantage. Third, since a chemical reaction induced by multibeam irradiation only proceeds at the position where both beams overlap, a chemical process induced by multibeam irradiation is a ‘‘site-selective’’ process. This third advantage is important in photodynamic therapy, in which damage to healthy cells must be avoided. One of the most advanced techniques is to perform selective damage of cancer tissues deep beneath the skin surface (Fig. 2.4). Molecular memory is also possible when taking this feature into account. Because various reactive

Figure 2.4. Schematic illustration describing the ‘‘site-selective’’ process of multibeam chemistry. The damaged area caused by the multibeam laser irradiation can be limited to the overlap of two beams.

PULSE RADIOLYSIS–LASER FLASH PHOTOLYSIS

57

intermediates such as radicals have strong absorption bands in the visible region, bleaching and reappearance of the visible absorption band by the multibeam irradiation are equivalent to chromism control, which provides for information storage and read-out. From these characteristics, multibeam irradiation using various excitation sources can be applied to various fields, not only to the basic chemistry of the higher excited states. For years, various groups have investigated the reactions induced by multibeam irradiation [3] and various compounds, including basic molecules, supramolecules for molecular devices, semiconductor nanoparticles, and biomolecules like DNA. Because of their availability, nanosecond lasers have been combined to achieve multibeam irradiation, although excited intermediates have lifetimes usually shorter than nanoseconds. Thus, the properties determined by the nanosecond technique are indirect ones, including various ambiguities. Furthermore, the nanosecond technique does not allow investigation of photoswitching faster than the nanosecond regime. For detailed investigations, use of ultrashort pulse lasers is important. A series of studies on molecular switching by Wasielewski and co-workers [18–22] and a study of the higher excited states of thin films of conjugated polymers by Masuhara and co-workers [23] are examples of utilizing multibeam irradiation with ultrashort pulse lasers. These examples indicate the importance of multibeam irradiation in the picosecond regime, since the estimated kinetic parameters directly facilitate design of functional molecules. Studies using multibeam irradiation with ultrashort pulse lasers will become more important in the near future. In this chapter, we summarize recent progress in the photochemistry of shortlived species by use of multibeam excitation, including our recent achievements in this field. Our research group has employed various multibeam irradiation methods to reveal reaction processes of various excited intermediates, including basic molecules and biomolecules. We also achieved direct observation of shortlived species utilizing ultrashort pulse lasers. These results are interesting recent examples of reactions induced by multibeam irradiation. Based on the excitation method, this chapter is divided into the following sections: Pulse Radiolysis– Laser Flash Photolysis (Section II), Two-Color Two-Laser Flash Photolysis (Section III), and Three-Color Three-Laser Flash Photolysis (Section IV). Each section is further divided into subsections based on the topics.

II. PULSE RADIOLYSIS–LASER FLASH PHOTOLYSIS Pulse radiolysis is a powerful tool for generating various kinds of intermediates, such as radical cations and radical anions. Since an electron and a hole are

58


generated from the initial radiolytic reaction during the pulse radiolysis of solutions, the electron or hole can be selectively trapped by solvents. Therefore, a solute radical cation is generated during the pulse radiolysis in alkylhalides, such as 1,2-dichloroethane or butyl chloride, and a solute radical anion is generated in basic solvents such as N,N-dimethylformamide and 2-methyltetrahydrofuran [24]. Selective generation of radical ions is quite useful for investigating the excited states of radical ions on the time scale of nanoseconds to 100 ms, because of the easy subsequent excitation. In the present section, we introduce some examples of studies on the excited states of radical ions using the pulse radiolysis–laser flash photolysis combined method [25–29].

A. Excited Radical Cations Upon photoexcitation, several organic compounds undergo isomerization. It is well known that cis (c)–trans (t) isomerization of stilbene (St) occurs via twist C double bond in the singlet or triplet excited states ing about the central C upon irradiation with UV light [30]. Lewis and co-workers [31] have reported that a St radical cation (St þ) undergoes thermal c–t one-way isomerization via the St dimer radical cation (St2 þ) as an intermediate. On the other hand, photochemical c–t one-way isomerization of c-St þ to t-St þ occurs in rigid matrices at 77 K [32] and in solution at room temperature based on the laser flash photolysis of c-St þ formed during pulse radiolysis in 1,2-dichloroethane or secondary electron transfer (ELT) in acetonitrile [14, 33]. The photochemical c–t isomerization has been reported to take place in the second doublet excited (D2) state but not in the lowest doublet excited (D1) state of c-St þ[14]. Characterizations of St þ in the D2 state (St þ*) are necessary to elucidate the isomerization mechanism. c-St þ generated during pulse radiolysis of c-St in 1,2-dichloroethane showed D0 and D1 D0 transitions at 515 and the absorption bands due to the D2 780 nm, respectively. In the case of t-St þ, the corresponding peaks appeared at 480 and 760 nm [14, 34, 35]. The excitation of St þ(D0) at 532 nm produces St þ* with excitation energies of 50 and 53 kcal mol1 for c-St þ* and t-St þ*, respectively, that were calculated from the red edges of the absorption bands at 420–580 nm. The irradiation of t-St þ with a laser flash at 532 nm exhibited no change in the transient absorption spectra and time profiles of O.D.480, where t-St þ shows an absorption peak (Fig. 2.5a). Therefore, t-St þ* does not isomerize to c-St þ in the ground (D0) state with the irradiation but decays to t-St þ in the D0 state with the rate constant of internal conversion (IC) of tSt þ* (Scheme 2.1). The irradiation of t-St þ in the presence of anisole (ANS) caused a decrease in the O.D.480 immediately after the flash (Fig. 2.5b). The O.D.480 increased

59


Figure 2.5. Kinetic traces of O.D.480 during the pulse radiolysis–laser flash photolysis experiment of t-St (5 103 M) in the absence (a) and presence (b) of ANS (1.0 M) in Ar-saturated 1,2-dichloroethane.

with increasing concentration of ANS ([ANS]). It is obvious that t-St þ* is quenched by ANS via hole transfer quenching to give t-St and ANS þ with the bimolecular rate constant of kA (Scheme 2.1). The rise of O.D.480 after the laser flash corresponds to the hole transfer from ANS þ to t-St to produce ANS and t-St þ [35], which occurs at the rate constant of 7.8 109 M1 s1, equivalent to the diffusion-controlled rate (kdiff) in 1,2-dichloroethane. The chemical yield of [t-St þ]disapp ðYt Þ in the presence of ANS is represented by Eq. (1):

Yt ¼ ð½t-Stþ disapp =½t-Stþ 0 Þ ¼ I0 kA ½ANS =ðt1 t þ kA ½ANS Þ

ð1Þ

60


t-St * (τt = 240 ps) kA[ANS] hν

kdt

t-St + ANS

t-St

Scheme 2.1. Photochemistry of t-St þ involving the internal conversion at the rate constant of kdt and hole transfer quenching of t-St þ* by ANS at the bimolecular rate constant of kA.

where [t-St þ]0 and [t-St þ]disapp are the concentrations of t-St þ before irradiation of the laser flash and t-St þ disappeared immediately after the laser flash, I0 is the efficiency of the formation of t-St þ*, and t1 is the reciprocal of the t 1 versus lifetime of t-St þ*. According to Eq. (1), the Stem–Volmer plots of Yt 1 1 [A] produced a linear line with an intercept of I0 and a slope of ðI0 kA tt Þ1 . Consequently, tt ¼ 240 50 ps was obtained. Contrary to t-St þ*, c–t one-way isomerization of c-St þ* to t-St þ* was observed within a laser flash. The chemical yield of t-St þ from c-St þ* is approximately (75 15)% per flash, and isomerization of c-St þ to t-St þ proceeds as the main process. Because only the decay of c-St þ and formation of t-St þ were observed in Figure 2.6, the remaining (25 15)% of c-St þ that disappeared is considered to convert to c-St, t-St, or a radical cation as a product

Figure 2.6. (a) Transient absorption spectra recorded before the laser flash (open circles) and immediately after (filled circles) and 200 ns (open triangles) and 1 ms (filled triangles) after the laser flash during pulse radiolysis–laser flash photolysis of c-St (5 103 M) in Ar-saturated 1,2-dichloroethane. (b, c) Kinetic traces of O.D.480 and O.D.515, respectively, as a function of time after the electron pulse.


c-St * (τc = 120 ps)

kp

product

61

kA[ANS] ki

c

hν

kd

c-St + ANS

c-St

t-St

Scheme 2.2. Photochemistry of c-St þ involving the internal conversion at the rate constant of kdc , c–t isomerization to t-St þ at the rate constant of ki, product formation at the rate constant of kp, and hole transfer quenching of c-St þ* by ANS at the bimolecular rate constant of kA.

showing little or weak absorption obscured by the strong absorption bands of t-St þ and c-St þ over the range of 400 –700 nm. An electrocyclic product such as dihydrophenanthrene is assumed as an intermediate of the phenanthrene that is formed as an oxidation product in the photolysis of c-St [36]. In order to determine the lifetime of c-St þ* ðtc Þ, c-St þ was irradiated in the presence of ANS. The transient phenomena of c-St þ* involving the isomerization and hole transfer quenching of c-St þ* with ANS are shown in Scheme 2.2. According to the Stem–Volmer plots, tc was calculated to be approximately 120 30 ps and found to be one-half of tt . The measurement of the lifetime of the D2 state using the hole transfer quenching was also examined for 1,2-diphenylcyclobutene radical cation in the D2 state (CB þ*). Because CB has a rigid planar structure with the c-St chromophore structurally constrained by the cyclobutene ring and is stable for geometrical isomerizations, CB þ is expected to have the same rigid planar structure as CB. The irradiation of CB þ with a laser flash at 532 nm exhibited no change in the transient absorption spectra and the time profile of O.D.480. Therefore, CB þ* is deactivated to CB þ in the D0 state within the laser flash, which is similar to t-St þ* (Scheme 2.1). The lifetime of CB þ* was estimated to be 380 30 ps. The shorter lifetime of c-St þ* is attributed to isomerization and conversion to another product via twisting about the central C C double bond. The analogous process in CB þ* is severely hindered by structural constraints.

B. Fluorescence from Excited Radical Cations Fluorescence is a quite sensitive probe widely used in various fields. Although most radical cations and anions are nonemissive, there are several exceptions. In this section, we introduce the fluorescence detection of radical cations generated during pulse radiolysis. Fluorescence of the radical cations gave us unique information on their reactivities [26, 27].

62


Figure 2.7. Transient absorption spectra observed at 0 ns, 100 ns, and 1 ms after the 8-ns electron pulse, and transient fluorescence spectrum of TMB þ* observed at 300 ns after the electron pulse during pulse radiolysis–laser flash photolysis of TMB (1 102 M) in Ar-saturated 1,2-dichloroethane. Excitation wavelength, 532 nm. Laser pulse energy, 140 mJ pulse1.

1,3,5-Trimethoxybenzene radical cation (TMB þ), which is generated during pulse radiolysis of TMB in 1,2-dichloroethane, gave a transient fluorescence spectrum around 600–750 nm upon photoexcitation at 532 nm with the second harmonic pulse of a Nd:YAG laser as shown in Figure 2.7. The transient fluorescence spectrum of TMB þ in the doublet excited state (TMB þ*) showed a mirror image symmetry to the absorption spectrum of free TMB þ [37]. Since the duration of the fluorescence was almost the same as that of the laser pulse, the fluorescence lifetime of TMB þ* is shorter than 1 ns. The fluorescence intensity monitored at 620 nm decreased with the increase of the delay time of the laser pulse to the 8-ns electron pulse, similar to the temporal profile of the transient absorption monitored at 590 nm (Fig. 2.8). However, the decay observed around 532 nm was much slower than that at 590 nm because of the ion pair formation ([TMB þ Cl]). When TMB þ was excited with the 532-nm laser pulse, a permanent depletion of the absorption was observed. Although the fluorescence spectra did not change with changing the delay time, the depleted absorption spectrum changed with the delay time. Immediately after the laser photolysis at an early stage shorter than 100 ns after pulse radiolysis, the depletion of the transient absorption of free TMB þ around 590 nm was small (less than 10%). However, greater depletion of the transient absorption of the ion pair at 520 nm (15–25%) was observed at a later stage longer than 100 ns after

63


Figure 2.8. Kinetic trace of O.D. at 590 nm of TMB þ during the pulse radiolysis followed by the consecutive irradiation of a 532-nm laser pulse. Fluorescence intensity (open circle) as a function of the delay time of the 532-nm laser pulse relative to the electron pulse is superimposed on the decay curve.

pulse radiolysis. These results suggest that TMB þ* is quenched by Cl within the ion pair. From the quenching experiment with tetrabutylammonium chloride, the transient absorption spectrum after 30 ns was no longer assigned to free TMB þ. The estimated fluorescence quantum yield (f) was 3.1 105 and was constant in the time range longer than 50 ns. Since the lifetime of TMB þ* seems to be too short for the quenching of TMB þ* by quencher molecules at a diffusion-controlled rate, the estimated f is considered to be TMB þ* in the ion pair. On the other hand, f for free TMB was estimated from the experiment involving intermolecular hole transfer. Pulse radiolysis of a 1,2-dichloroethane solution containing biphenyl and TMB gave the radical cation of biphenyl þ immediately after the electron pulse. Since the oxidation potential of biphenyl is higher than TMB, biphenyl þ is quenched by the hole transfer to TMB at a nearly diffusion-controlled rate. Although the transient absorption of TMB þ showed a rise corresponding to the hole transfer, f decreased monotonously with the increase of the delay time. The estimated f value was 1.1 103 at 10 ns. At this stage, TMB þ can be considered a free ion because of the low yield of TMB þ and Cl (approximately 105 M), although the rate constant of collision between TMB þ and Cl is quite large (2.6 1011 M1 s1). Therefore, this value was considered to be the f value for free TMB þ*.

64


The radical cation of 3,5-dimethoxyphenol showed fluorescence in a similar manner [27]. The f value and lifetime were estimated to be (2 0.3) 103 and 350 ps, respectively.

C. Excited Radical Anions Radical anions of organic compounds can be generated selectively during pulse radiolysis in solvents such as N,N-dimethylformamide. Generated radical anions can be excited selectively by employing a laser at an appropriate wavelength. The c–t isomerization is also reported for the radical anion of St (St ). It is unclear whether or not the photoisomerization of St occurs via a mechanism similar to that for St þ. In order to characterize St in the D2 state (St *), we have investigated the selective ELT quenching of isomeric St * using biphenyl as an electron acceptor and estimated the lifetimes of St * using the pulse radiolysis–laser flash photolysis combined method [28]. From the selective ELT from St * to biphenyl, kbiphenyl t ¼ 10:7 and 17.8 M1 were obtained for c-St and t-St , respectively, where kbiphenyl and t denote the rate constants of the ELT and the lifetimes of St *, respectively. The t values are estimated to be approximately 1.5 0.4 ns and 2.5 0.7 ns for c-St * and t-St *, respectively, from kbiphenylt assuming the diffusion-controlled rate constant for kbiphenyl. The shorter t of c-St * is attributed to the c–t isomerization * and via twisting about the central C C double bond. The t values of c-St * þ* þ* t-St are one order of magnitude larger than those of c-St and t-St [25]. The selective ELT quenching of the radical anions of dicyanoanthracene, phenazine, and anthraquinones in the higher doublet excited ðDn Þ state by electron quenchers such as fumaronitrile or dicyanobenzene is also investigated in N,N-dimethylformamide at room temperature using the pulse radiolysis–laser flash photolysis combined method [29]. The radical anions generated during the pulse radiolysis do not change upon irradiation with a laser flash at 532 nm. The radical anions in the Dn state decay into the D0 state within the laser flash (5 ns). Lifetimes of approximately 4 ns are estimated for three radical anions in the Dn state assuming a diffusion-controlled rate constant for the ELT quenching. The shorter lifetimes of 1.0–1.4 ns for methyl and chloro substituents on anthraquinone can be explained in terms of IC from the Dn to the D0 state of the radical anions accelerated by rotation of the substituents. The energy gap between the Dn and D0 states of the radical anions is a significant factor for the rate of IC. The quencher radical anion–neutral molecule pair is suggested as an intermediate in the ELT quenching of the radical anions in the Dn state by the electron quencher. In the present section, we introduced some examples of excited radical ions generated during pulse radiolysis. As pointed out previously, pulse radiolysis is a

65

TWO-COLOR TWO-LASER FLASH PHOTOLYSIS hν

N

M

M

M

hν

+ e–

– e–

N Δ

Q

N

hν

Δ

hν

Δ

Δ

hν

D

Q

Scheme 2.3. Reaction pathways applicable to the multibit molecular memory utilizing the multibeam chemistry of a molecule (M), in which isomerization of M to N and charge transfer from M to quencher (Q) play important roles to realize the multibit memory.

powerful tool for generating radical ions selectively and in a high yield. Furthermore, an oxidant or a reductant generated during pulse radiolysis as the initial process reacts with various kinds of compounds even though their oxidation or reduction is difficult with other methods. At the present stage, a nanosecond laser has been employed to excite radical ions generated during pulse radiolysis. Therefore, estimation of the properties of the excited radical ions was carried out in a rather indirect manner. For direct observation, detection and excitation systems with picosecond resolution should be developed. We summarized the present section from the viewpoint of a basic study to reveal the properties of excited radical ions. On the other hand, the present reaction system employing pulse radiolysis and laser photolysis gives a basis of a multibit molecular memory composed of various charged states and isomeric structures. Reaction pathways applicable to the multibit molecular memory utilizing multibeam chemistry are indicated in Scheme 2.3, in which isomerization and charge transfer to a quencher play important roles in achieving the multibit memory. It should be pointed out that a rather fast response is expected for multibit molecular memories utilizing an adequate chromophore-linked system or solid-state support for these chromophores. Furthermore, various intermediates, such as a highly charged state, are also applicable to this approach.

III. TWO-COLOR TWO-LASER FLASH PHOTOLYSIS A pulse laser is another excitation source for generating various reactive intermediates efficiently. In the case of pulse lasers, selective excitation of the ground and excited states is easy because laser pulses with various wavelengths can be obtained by harmonic generation with nonlinear crystals. Recent development of a laser utilizing optical parametric oscillation, which emits a variable wavelength, enlarged the scope of study. Thus, two-color two-laser flash photolysis has been adapted to a wide variety of fields [3]. Furthermore, utilization of

66


ultrashort pulse lasers is beneficial for direct observation of higher excited states with quite short lifetimes. We recently succeeded in the synchronization of nanosecond and picosecond lasers. In this section, we summarize recent results of two-color two-laser flash photolysis studies on the higher triplet excited states, excited radicals, excited radical ions, and other reactive intermediates.

A. Higher Triplet Excited States (Tn) 1. Energy Transfer from the Tn State Upon the photoexcitation of a chromophore with a doubly occupied highest molecular orbital, almost all chromophores generate the lowest triplet excited (T1) state through the intersystem crossing (ISC) process (Fig. 2.1). Since the lifetime of the T1 state is usually on the order of a microsecond or millisecond, the excitation to the higher triplet excited ðTn Þ state is feasible by combining nanosecond lasers. By employing adequate timing circuits, synchronization of two nanosecond lasers can be achieved rather easily. Thus, explorations of the Tn state have been carried out for years. In the late 1960s, Liu and co-workers demonstrated the energy transfer (ENT) process from the Tn state of compounds such as anthracene based on the product analysis, which can be obtained only from the Tn states [38–42]. They determined the lifetime of anthracene(T2) using high-concentration benzene, which acts as the triplet energy quencher and triplet energy carrier to endo-dicyclopentadiene, giving norbornene from the T1 state. Saltiel et al. [43] reported a similar approach in which the sensitized photoisomerization of St or 2,4-hexadiene is used as a probe of the triplet energy quenching of anthracene(T2). Kokubun and co-workers measured fluorescence from the S1 state after ISC from the Tn state [44–46]. The research groups of Scaiano and McGimpsey revealed various reactions for the Tn states [7, 9–11, 15, 16, 47, 48]. For the systematic understanding of the Tn state, we have studied intermolecular ENT processes of several aromatic hydrocarbons in the Tn states. In this section, studies on the Tn state properties of naphthalene (Np), a fundamental molecule, are described [49]. Since the quantum yield of fluorescence of Np (f ¼ 0.19) is not negligible [50], Np in the lowest triplet excited state (Np(T1)) was obtained by the triplet sensitization [51, 52]. As a triplet sensitizer, we employed benzophenone (BP), which yields the T1 state quantitatively with higher triplet energy than that of Np(T1). The generation of Np(T1) was confirmed by the growth of the transient absorption band at 415 nm (Fig. 2.9). Np(Tn) was obtained by the excitation of Np(T1) with irradiation of laser light at 425 nm. (Scheme 2.4 shows an energy diagram involving Np(S0), Np(T1), Np(Tn), Q(S0), and Q(T1).) No change of the transient absorption of Np(T1) was observed during the second laser irradiation. It is

67

TWO-COLOR TWO-LASER FLASH PHOTOLYSIS

0.4

0.2 [CCl4] = 0M 0.60 M 0.0 0

100

200

300

Time (ns)

Figure 2.9. Kinetic traces of O.D.415 during two-color two-laser flash photolysis of Np in the absence and presence of CCl4— 0.25, 0.35, 0.50, and 0.60 M—in Ar-saturated acetonitrile solution at room temperature.

suggested that the fast IC of the Tn !T1 transition occurs within the laser flash duration of 5 ns. However, entirely different consequences were observed when the solution included CCl4. In the presence of CCl4, the second laser irradiation caused bleaching of the transient absorption of Np(T1). Furthermore, the bleaching increased with increasing concentration of CCl4. No change of transient absorption of Np(T1) was observed under the irradiation by one laser at 355 or 425 nm Sn

Tn ENT

IC

S1

T1

IC

hν2

ISC

reaction

T1 T1

ENT

hν1

ENT

ISC ISC

S0

S0

S0 BP

Np

Q

Scheme 2.4. Energy diagram of BP(S0), BP(S1), BP(T1), Np(S0), Np(T1), Np(Tn), Q(S0), and Q(T1) involving the triplet ENT from NpðTn Þ to Q(S0) giving Np(S0) and Q((T1), and from Q(T1) to Np(S0) giving Q(S0) and Np(T1).

68


in the presence of CCl4. The bleaching was observed only in the presence of CCl4 and with irradiation of two lasers as shown in Figure 2.9. Similar experimental results were observed in the presence of CH2Cl2. On the other hand, the formation of Cl–benzene complex was confirmed by detection of the transient absorption with a peak at 490 nm in the presence of benzene in CCl4 solution [53, 54], indicating that NpðTn Þ donated its energy to CCl4, causing the cleavage of the C—Cl bond of CCl4, that is, CCl4* ! CCl3 þ Cl. In order to elucidate the mechanism involving these phenomena, several quenchers such as dichlorobenzene (DCB) and dicyanobenzene (DCNB) were used. The bleaching and recovery of the transient absorption of Np(T1) were observed in the presence of these quenchers. The bleaching increased with increasing concentration of the quenchers (0.3 < [Q] < 1.0 M) as shown in Figure 2.10. The recovery was accomplished in 100% yield without formation of a new peak. If DCNB could act as an electron acceptor for the quenching of NpðTn Þ, Np radical cation and DCNB radical anion would be observed. Therefore, no ELT quenching occurred, but ENT quenching did. The absence of the radical cation and anion also indicated that ISC from the Tn to S1 (or Sn ) can also be neglected in the present processes, since it is known that Np(S1) causes the ELT with DCNB. It is well established that the triplet ENT occurs at the diffusion-controlled rate (kdiff) when ET of the triplet energy donor is 13 kJ mol1 higher than that of the triplet energy acceptor [41, 52]. In the present case, the ET values of DCB and DCNB in the T1 state (335 and 305 kJ mol1, respectively) are much higher

0.4

0.40

[DCB]= 0M 0.50 M 0.70 M 0.90 M

.D.415

0.2

0.35 100.0

150.0

200.0

Time (ns)

0.0 0

100

200

300

Time (ns)

Figure 2.10. Kinetic traces of O.D.415 during two-color two-laser flash photolysis of Np in the absence and presence of DCB—0.50, 0.70, and 0.90 M—in Ar-saturated acetonitrile at room temperature.

69


than ET of BP(T1) and Np(T1) (289 and 253 kJ mol1, respectively) [50], and the recovery rate of transient absorption of Np(T1) is almost independent of the acceptor concentration ([DCB] or [DCNB]). Therefore, it can be concluded that the triplet ENT from quencher(T1) to Np occurred at kdiff to give Np(T1) and quencher(S0). Since the formation and decay of NpðTn Þ occurred within the duration of the 425-nm laser flash, the transient phenomena of NpðTn Þ cannot be monitored directly using nanosecond lasers. The bleaching of O.D.415 upon laser irradiation (O.D.415 ¼ O.D.before O.D.after) in the presence of CCl4 resulted from the ENT from NpðTn Þ to CCl4 and increases with an increase of the concentration of CCl4. The inverse of O.D.415 is represented by concentration of CCl4 as in Eq. (2) [25, 28]: ðO:D:415 Þ1 ¼ b þ bðkENT t½CCl4 Þ1

ð2Þ

where b is a constant depending on the experimental condition, kENT is the rate constant of the triplet ENT from NpðTn Þ to CCl4, and t is the lifetime of NpðTn Þ. According to Eq. (2), the Stern–Volmer plots of (O.D.415)1 versus [CCl4]1 gave a linear line with an intercept of b and slope of b(kENTt)1 as shown in Figure 2.11. From ET of Np(T1), 253 kJ mol1 [50], ET of NpðTn Þ is estimated to be 534 kJ mol1 under the 425-nm laser excitation in our experiments. ET of NpðTn Þ is much higher than those of Qs(T1). Therefore, it is reasonable to suggest that the ENT from NpðTn Þ to Qs(S0) occurs at kdiff. However, the ENT kinetics from NpðTn Þ to Q(S0) is different from that from Q(T1) to Np(S0). The concentration of Q (0.3 < [Q] < 1.0 M) is high and t is short, while the

25

20

15

10 1

2

3 [CCl4]–1 (M –1 )

4

Figure 2.11. Plots of (O.D.415)1 versus [CCl4]1 in two-color two-laser photolysis of Np in the presence of CCl4.

70


concentration of Np (6 mM) is low and the lifetime of Q(T1) is long (several hundred microseconds [50]). Consequently, lifetime-dependent quenching should be considered in the case of quenching of short-lived species such as NpðTn Þ [16, 55]. In such cases, kENT can be expressed by the lifetime-independent and lifetime-dependent terms shown in Eq. (3): kENT ¼ kdiff þ kdiff s0 =ðpDtÞ0:5

ð3Þ

where kdiff ¼ 4pNs0 D, N is Avogadro’s number, s0 is the reaction distance, and D is the sum of the diffusion coefficients for the excited molecule and quencher molecule. Therefore, t of NpðTn Þ was found to be 4.5 ps at s0 ¼ 0.6 nm and D ¼ 2.0 107 dm2s1 [16]. Similar results were obtained when CCl4 was replaced by DCB or DCNB. This is the first report on t ¼ 4.5 ps of NpðTn Þ. 2. Energy Gap Law For other aromatic hydrocarbons (AH) such as chrysene (CHR) [56] and dibenzanthracene (DBA) [57], lifetimes of the Tn states were estimated. For anthracene, the Tn state lifetime was reported by Scaiano and co-workers [16]. Employing these estimated values, the factors governing the lifetime were discussed [58, 59]. Since neither unimolecular reaction nor luminescence of AHðTn Þ was observed, AHðTn Þ is suggested to be deactivated through IC to AH(T1). Therefore, the rate constant (kIC) of IC of AHðTn Þ is defined to be the inverse of t. Generally, kIC depends on the energy gap between two states, following the energy gap law for IC as shown in Eq. (4) [52]: t1 ¼ kIC 1013 expða EÞ

ð4Þ

where a (eV1) is a constant and E (eV) is the energy gap between the zero point vibrational levels of the states undergoing IC. The a value is usually smaller than 5 eV1 and does not change so much for rigid AHs [16, 52, 60]. For example, a ¼ 3.3 eV1 was calculated from the data of anthracene(T2), which decays through radiationless processes with the lifetime of the T2 state (tT2) of 11 ps and E value between the T1 and Tn states (ET2-T1) of 1.39 eV [16]. With a ¼ 3.3 eV1 and t of AH ðTn Þ calculated from the experimental results, the E values between the T1 and Tn states ðETnT1 Þ of Np, DBA, and CHR were calculated to be 1.15, 1.54, and 1.94 eV, respectively. The ET2-T1 values between the T1 and T2 states for Np and CHR are theoretically calculated to be 1.17 and 1.75 eV, while ET3-T1 values of Np and CHR are 1.29 and 2.03 eV, respectively [61–63]. Therefore, NpðTn Þ and CHRðTn Þ are assigned to Np(T2) and CHR(T2), respectively. Since no theoretical study has been reported on DBAðTn Þ, the triplet manifolds of DBAðTn Þ cannot be determined. However, from the results of Np, anthracene, and CHR, DBAðTn Þ is tentatively assigned to be DBA(T2). It is suggested that the AHðTn Þ initially generated decays through fast IC to a lower triplet excited state, the T2 state, which has

71


the longest lifetime among the Tn states. NpðTn Þ having a large transition dipole moment was found to be Np(T10) according to theoretical calculation T1 (2 < n < 10) tranusing the QCFF/PI þ CISD method [61]. All other Tn sitions are forbidden or have very small oscillator strengths. Therefore, the T–T T1 transition. In other words, absorption spectrum is assigned to the T10 NAP(T10) is generated by the 425-nm laser excitation of Np(T1) and decays through fast IC within picoseconds to Np(T2) with t of 4.5 ps. The t values of Np(T2), DBA(T2), and CHR(T2) calculated from the ENT quenching experiments increased in the order of Np(T2) (4.5 ps) < DBA(T2) (16 ps) < CHR(T2) (60 ps). This order is consistent with the energy gap law for the transition from AH(T2) to AH(T1). Therefore, AH(T2) with the longest lifetime among AHðTn Þ is responsible for the ENT quenching. 3. Substituent Effect on the Tn State Lifetimes By employing a method similar to the one described in Section III.A.2, the lifetime of the T2 state of BP was estimated to be 450 ps [64]. For a series of BP derivatives, the T2 state lifetimes were estimated as listed in Table 2.1 [65]. The introduction of the substituent tends to decrease the T2 state lifetime. In the case of Np derivatives (NpD), the electron-donating substituent increases the t value, while the electron-withdrawing substituent decreases t [66]. The t values of NpD(T2) in cyclohexane increase in the order of OCH3 > CH2CH3 > CH3 > H ¼ CH2CN > CN. This is the same order as the Hammet constant sp . As shown in Figure 2.12, the ET1 values of NpD(T1) are almost constant. In contrast, the ET2 values of NpD(T2) depend on sp significantly. The T2 state (B2u) consists mainly of two electronic configurations, namely, HOMO 1! LUMO and HOMO ! LUMO þ 1 [61]. The electron-withdrawing group is expected to reduce the antibonding character of the LUMO þ 1 reducing ET2. Therefore, substitution of the electron-withdrawing group seems to decrease E, leading to shorter t.

TABLE 2.1 Lifetimes of Benzophenone (BP) and Substituted Benzophenone (BPD) in the T1 and Tn States (sT1 and sTn , respectively), Triplet Excited State Energies of BP(T1) and BPD(T1) (ET1), and the Hammett Constants of the para Substituents (rp) for BP and BPD (4-X-C6H4COC6H4-Y-40 ) X Y

H H

CH3O H

CH3 H

F H

Cl H

CF3 H

CN H

CH3O CH3 CH3O CH3

ET1(kJ mol1) sp tT1 (ms) tTn (ps)

289

290

290

292

288

285

280

292

0.00 0.71 450

0.27 0.23 240

0.17 0.06 0.27 0.21 250 260

0.23 0.54 0.66 0.50 0.41 0.44 280 140 110

290

F F

CN CN

294 276

0.27 0.17 0.06 0.66 0.19 0.26 0.72 0.67 350 300 300 250

72


4.5

E (eV)

4.0

3.5

3.0

2.5 –0.2

0.0

0.2

0.4

0.6

p

Figure 2.12. Plots of ET1 (closed triangle) and ET2 (closed circle) of 1-substituted naphthalene and ET1 (open triangle) and ET2 (open circle) of 2-substituted naphthalene versus the Hammett constant sp in cyclohexane.

4. Bond Dissociation from the Tn State For the Tn state, reaction that cannot be observed for the T1 states can be expected. A bond dissociation process is one of the reactions, which can be expected for the Tn states. Here, we introduce the formation of naphthylmethyl radical from the Tn states [67]. Naphthylmethyl radical (NpCH2) is a typical organic radical and has been studied extensively [68–74]. It is well established that NpCH2 is produced from the photolysis of naphthylmethylhalides through the cleavage of the naphthylmethyl–halogen bond [69, 71]. The cleavage also occurs in other naphthylmethyl compounds under the laser irradiation. Steenken and co-workers also found that the C—O bond cleavage occurred in the S1 state but not in the T1 state of 1-[(4-benzoylphenoxy)methyl]naphthalene (1-NpCH2-OBP) [74]. They assumed that the C—O bond cleavage could occur if the Np moiety is excited to the Tn state through further photon excitation of the T1 state. When 1-NpCH2-OBP was irradiated at 355 nm using a Nd:YAG laser in cyclohexane, the transient absorption spectrum observed immediately after the laser flash (Fig. 2.13a) was coincident to that of Np(T1) with a peak at 420 nm [72, 74]. Because the Np chromophore has no absorption at 355 nm, the first 355-nm photon is absorbed by the BP chromophore to give BP(S1), from which ISC occurs to give BP(T1) in quantum yield of 1.0 [50]. The intramolecular triplet ENT from BP(T1) to the Np chromophore occurs to give Np(T1) within


73

Figure 2.13. Transient absorption spectra obtained at 1.1 ms after the 355-nm first laser irradiation (a) (broken line) and at 1 ms after the second 430-nm laser irradiation (b) (solid line) during two-color two-laser flash photolysis of 1-NpCH2-OBP in Ar-saturated cyclohexane at room temperature. The delay time of the second laser after the first laser was 100 ns. Inset: Spectrum obtained by (b)(a) in the region of 350–380 nm.

a laser flash of 5 ns. No appearance of the peak at 365 nm assigned to 1-NpCH2 indicates that the C—O bond cleavage did not occur from 1-Np(T1)CH2-OBP [74]. This observation is adequate since the triplet energy of Np(T1) (254 kJ mol1) is lower than the dissociation energy of the C—O bond (285 kJ mol1) [74]. The absorption peak at 365 nm assigned to 1-NpCH2 was observed with the bleaching of the 420-nm peak within a laser flash, when the second 430-nm OPO laser flash was irradiated to 1-Np(T1)CH2-OBP with a delay time of 100 ns after the first 355-nm laser flash (Fig. 2.13b). Because only 1-Np(T1)CH2-OBP has an absorption at 430 nm, 1-Np(T1)CH2-OBP can be excited to 1-NpCH2-OBPðTn Þ by the second laser. The 430-nm photon (278 kJ mol1) supplies sufficiently high energy into 1-Np(T1)CH2-OBP, giving 1-NpCH2-OBPðTn Þ, from which a rapid C—O bond cleavage occurred within the laser flash (Scheme 2.5). The quantum yield () of 1-Np(T1)CH2-OBP disappeared and 1-NpCH2 formed was calculated to be 0.042 0.004 from the slopes of the linear lines of the plots of j O.D.420 j and j O.D.365 j versus 430-nm laser power. The small value indicates that IC from 1-NpCH2-OBPðTn Þ to 1-Np(T1)CH2-OBP is the predominant process (95.8% yield). Similar experimental results to those of 1-NpCH2-OBPðTn Þ were obtained for 2NpCH2-OBPðTn Þ. The bleaching of the transient absorption of 2-Np(T1) CH2-OBP

74


O

CH2

O

hν355

CH2

< 100 ps ISC

O

O

S1 O

CH2

O

< 5 ns

CH2

hν430

ENT

IC

O

O T1

T1 O

CH2

O

< 5 ns

H2C +

O

O Tn

Scheme 2.5. Two-color two-laser photochemistry of 1- and 2-NpCH2-OBP involving intramolecular triplet ENT, selective excitation of Np(T1)CH2-OBP to NpCH2-OBPðTn Þ, and cleavage of the C—O bond from NpCH2-OBPðTn Þ. Dotted square shows the excitation energy delocalization.

at 420 nm and the formation at 380 nm assigned to 2-NpCH2 were observed during the two-color two-laser flash photolysis of 2-NpCH2-OBP [75, 76]. However, the bleaching of the transient absorption at 420 nm and the formation at 380 nm observed in 2-Np(T1)CH2-OBP were less than those in 1-Np(T1)CH2-OBP. The value of 2-Np(T1)CH2-OBP disappeared and 2-NpCH2 formed was calculated to be 0.020 0.002, which was almost half of that of 1-Np(T1)CH2-OBP. This result indicates that the C—O bond cleavage of 1-NpCH2-OBPðTn Þ occurs more efficiently than that of 2-NpCH2-OBPðTn Þ. Although the T1 state is localized on the Np chromophore of 1- and 2-Np(T1)CH2-OBP, the Tn state could be delocalized in 1- and 2-NpCH2-OBPðTn Þ including the C—O bond. It is expected that the extent of the delocalization is more prominent in 1-NpCH2-OBPðTn Þ than in 2-NpCH2OBPðTn Þ. Similar bond cleavage was observed for the Tn state of the other compounds. For example, p-phenoxymethylbenzophenone (BPCH2OPh) and p-methoxymethylbenzophenone (BPCH2OCH3) undergo bond cleavage to generate benzoylbenzyl radical and phenoxy or methoxy radical through the Tn state of BP during the two-color two-laser flash photolysis [77]. The cleavage yield of BPCH2OPh was higher than that of BPCH2OCH3. We also found that the C—Si bond is another target of the bond cleavage from the Tn state generated by the two-color two-laser photolysis [78]. The


75

two-color two-laser flash photolysis of p-trimethylsilylmethylacetophenone generated p-acetylbenzyl radical, indicating the bond cleavage from the Tn state of acetophenone. On the other hand, p-trimethylsilylmethylbenzophenone did not generate the bond cleavage products, in spite of the higher Tn state energy than the C—Si bond dissociation energy. These results indicate that the existence of a bond cleavage crossing between the potential surfaces of the Tn state and a dissociative state of the C—Si bond is also an important factor for the bond dissociation in addition to the energetic consideration. 5. Electron Transfer from the Tn State Photoinduced ELT, one of the fundamental processes in physical, chemical, and biological aspects, is an attractive subject and has received much attention. Usually the lowest singlet and triplet excited states participate in the photoinduced ELT. Although the ELT from the higher excited states is also energetically possible, studies are limited. Intermolecular ELT from the higher singlet excited (S2) state of zinctetraphenylporphyrin to dichloromethane was reported by Okada and co-workers [79]. LeGourrie´rec et al. [80] reported the intramolecular ELT from the S2 state of porphyrin in a covalently linked zinc porphyrin–ruthenium(II) tris-bipyridine dyad. Systematic study of the intramolecular ELT from the S2 state of porphyrins was carried out by Mataga et al. [81–83]. The dependence of the intramolecular ELT rate from the S2 state on the free energy change and solvent was confirmed. Intermolecular ELT quenching of the S2 states of azulene, benz[a]azulene, and xanthone by several electron donors was investigated by Muller and Vauthey [84]. Furthermore, intermolecular ELT from St þ in the D2 state was reported by our group [25, 28]. These studies clearly demonstrate that the ELT from the higher excited states is possible even when the lifetime of the higher excited state is as short as a few picoseconds. Therefore, intermolecular ELT from the higher triplet excited states ðTn , n 2) seems to be also possible. Because large excitation energy of the Tn state affords larger driving force of the ELT even when the ELT from the T1 state is energetically unfavorable, the ELT is expected to occur from the Tn state. Limited numbers of studies have been reported on the properties of molecules in the Tn states. Two-color two-laser flash photolysis can be applied to study photoinduced reactions from the Tn states. It has been reported that the main reaction path from the Tn states is the triplet ENT to the triplet quenchers. To the best of our knowledge, there has been only one report on the ELT from the Tn state. Wang et al. [48] reported the ELT from anthracene(T2) to ethyl bromoacetate. However, no detailed mechanism of the ELT from the Tn state has been reported. In this section, we summarize our recent systematic study of the intermolecular ELT from a series of substituted naphthalenes (NpD) in the Tn state to electron acceptors [85, 86].

76


NpD(T1) was generated from the triplet sensitized reaction during the first laser (355 nm, 3 mJ pulse1) irradiation to a mixture of BP (7.0 103 M) and NpD (7.0 103 M) in Ar-saturated acetonitrile at room temperature. NpDðTn Þ was generated by the excitation of NpD(T1) with the second laser (425 nm, 9 mJ pulse1) at 100 or 150 ns after the first laser. In the presence of CCl4, bleaching of the transient absorption of 1-methoxynaphthalene(T1) and growth of new transient absorption peaks at 385 and 702 nm were observed immediately after the second laser irradiation as shown in Figure 2.14. The new absorption bands were assigned to 1-methoxynaphthalene radical cation [35, 87]. The inset of Figure 2.14 shows the kinetic traces of O.D. at 440 and 702 nm. Trace a shows the second laser-induced bleaching of the absorption of 1-methoxynaphthalene(T1) at 440 nm within the second 425-nm laser flash duration of 5 ns. This bleaching indicates that 1-methoxynaphthaleneðTn Þ generated with the second laser irradiation did not reproduce 1-methoxynaphthalene(T1). Trace b shows the growth of the radical cation absorption at 702 nm within the laser pulse duration. Thus, it is clearly indicated that the ELT from 1-methoxynaphthaleneðTn Þ to CCl4 occurred to give the 1-methoxynaphthalene 0.15 0.12 second laser fire 0.08

a 440 nm

0.10 0.04

b 702 nm

0.00 0

0.05

100

200 300 Time (ns)

400

500

0.00 350

400

450

500 550 600 Wavelength (nm)

650

700

750

Figure 2.14. Transient absorption spectra obtained during two-laser (first 355-nm and second 425-nm) excitation (filled circles) and one-laser (355-nm) excitation (open circles) of BP (7.0 103 M) with 1-methoxynaphthalene (7.0 103 M) in Ar-saturated acetonitrile in the presence of CCl4 (1.0 M) at room temperature. The second laser was irradiated at 200 ns after the first laser pulse. Inset: Kinetic traces of O.D. at 440 (a) and 702 nm (b) with and without the second 425-nm laser irradiation. The initial growth of O.D in the 50-ns time scale corresponds to the formation of NpD(T1) by the ENT from BP(T1).

77


radical cation within the laser flash duration. The ELT from the Tn state was also observed for other NpD such as 1-methylnaphthalene and 1-ethylnaphthalene. It should be noted that the radical cation was not observed without the second laser irradiation even in the presence of CCl4, indicating the contribution of the Tn state to the occurrence of the ELT. Since no fluorescence from NpD(S1) was observed after the second laser irradiation in the absence of CCl4, back ISC (Tn ! S1) can be neglected. Therefore, NpD(S1) is not involved during the second laser irradiation although the ELT from NpD(S1) to CCl4 was reported [88]. The efficiency of the ELT from NpD(T2) to CCl4 obviously depends on the properties of NpD(T2). Both ENT and IC are involved in the NpD(T2) decay process [49]. Therefore, the efficiency of the ELT (fELT) from NpD(T2) is represented by fELT ¼ O.D.702/ (eNPDþ [NpD(T2)]0), where O.D.702 is the difference of O.D. at 702 nm with and without the second laser irradiation in the presence of CCl4 (1.0 M), eNPDþ is the molar absorption coefficient of NpD radical cation, and [NPD(T2)]0 is the initial concentration of NpD(T2). The [NpD(T2)]0 value can be estimated from the rate constant of IC and the bleaching of transient absorption of NpD(T1) in CCl4 upon the second laser irradiation. The estimated fELT values for NpD(T2) are listed in Table 2.2. The rate of the ELT depends on the driving force (GELT). It is necessary to estimate GELT to explain the difference of fELT. CCl4 undergoes the C—Cl bond cleavage following the one-electron reduction [89–93]. According to the ELT mechanism involving the formation of NpD radical cation and the

TABLE 2.2 Lifetimes of Naphthalene and Substituted Naphthalenes in the T2 State (NpD(T2)) (sT2), Energies of the T1 and T2 States (ET1 and ET2, respectively), the Driving Force of the ELT (GELT), Half-Wave Oxidation Potentials (Eox) in Acetonitrile, Efficiencies of the ELT from NpD(T2) to CCl4 ( fELT), and Calculated Efficiencies ( fELT(calcd)) NpD Naphthalene 1-Methyl1-Ethyl1-Isopropyl1-Methoxy2-Methyl2-Ethyl2-Methoxy-

tT2(ps)

ET1 (eV)

ET2 (eV)

9.4 2.0 9.5 0.8 19 4.3 45 1.2 61 8.6 34 6.5 36 9.2 63 13

2.65 2.67 2.59 2.58 2.66 2.62 2.62 2.62

3.8 3.9 4.0 4.2 4.3 4.2 4.2 4.3

Eox (V) GELT vs SCE (eV) 1.78 1.63 1.63 1.63 1.38 1.65 1.64 1.41

1.3 1.5 1.6 1.8 2.2 1.8 1.8 2.2

fELT

fELT (calcd)

0 0.11 0.1 0.16 0.1 0.31 0.1 0.50 0.1 0.26 0.1 0.29 0.1 0.59 0.1

0.06 0.14 0.21 0.31 0.34 0.28 0.29 0.34

78


dissociative electron attachment to CCl4 leading to the C—Cl bond cleavage [89–93], GELT is represented by Eq. (5): GELT ¼ Eox ERCl=R þCl wp ET2

ð5Þ

where Eox is the oxidation potential of NpD, ERCl=R þCl is the reduction potential of the RCl/R þ Cl couple (for CCl4, ERCl=R þCl ¼ 0:825 V vs. SCE) [93], wp is a Coulombic energy (0.06 eV), and ET2 is the energy level of the T2 state given by the sum of the energy of the T1 state (ET1) and E. The estimated GELT values are summarized in Table 2.2. As shown in Figure 2.15, the fELT value increases with an increase of the GELT value. From the theoretical calculation (vide infra), the faster ELT is expected for larger GELT in this system. Since the ELT from NpD(T2) occurs competitively with other fast processes such as IC (T2 ! T1) and ENT, the large GELT value is necessary for the occurrence of the ELT. Since the ELT to CCl4 leads to the concerted C—Cl bond cleavage, contribution of the bond breaking should be considered. The dissociative ELT model, in which the Morse potential curve is employed, has been developed to describe

0.8 0.7 0.6

f ELT

0.5 0.4 0.3 0.2 0.1 0.0 1.2

1.4

1.6 1.8 –ΔGELT (eV)

2.0

2.2

Figure 2.15. Plots of fELT versus GELT for the ELT from NpDðTn Þ to CCl4 during the 355- and 425-nm two-color two-laser flash photolysis of BP with NpD in Ar-saturated acetonitrile in the presence of CCl4.


79

such ELT [89–93]. According to the ‘‘sticky’’ dissociative ELT model, the activation energy (G*) can be represented by Eq. (6) [91–93]: " #2 pffiffiffiffiffiffi pffiffiffiffiffiffi ð DR Dp Þ2 þ l0 GELT Dp 1 þ pffiffiffiffiffiffi pffiffiffiffiffiffi 2 G ¼ 4 ð DR Dp Þ þ l0

ð6Þ

where DR is a bond dissociation energy of reactant RX, Dp is an interaction energy of the radical ion pair, and l0 is the solvent reorganization energy independent of bond breaking. In the present work, the following parameters were employed: DR ¼ 2.99 eV, Dp ¼ 0.161 eV [93], and l0 ¼ 1.48 eV [92]. In simplified form, the activation-energy-controlled ELT rate constant can be expressed by Eq. (7): G kELT ¼ n exp ð7Þ RT where v is the frequency factor. Here, v is assumed to be 5.0 1013 M1 s1 [90]. For the bimolecular reaction, the ELT rate constant (k0 ELT) can be given by Eq. (8) [94]: 1 1 1 ¼ þ 0 kELT kdiff kELT

ð8Þ

taking the formation of an encounter complex into account. Because of the high concentration of CCl4 and short lifetime of the Tn state, the ELT rate constant 00 (kELT ) from NpDðTn Þ to CCl4 can be expressed by Eq. (3). According to the literature, D0 ¼ 2 105 cm2 s1 [50, 55]. The efficiency of the ELT from NpDðTn Þ (fELT(calcd)) can be calculated by Eq. (9): fELT ðcalcdÞ ¼

00 ðkELT

00 kELT ½CCl4 þ kENT Þ½CCl4 þ kIC

ð9Þ

where kIC is the IC rate constant given by the reciprocal of t, and [CCl4] is the concentration of CCl4 as a quencher ([CCl4] ¼ 1.0 M). The estimated fELT(calcd) values are summarized in Table 2.2. As shown in Figure 2.16, the plots of fELT versus fELT(calcd) showed a good correlation. It is suggested that the efficient ELT from NpD(T2) to CCl4 and the inefficient ELT from NpD(T1) to CCl4 are explained qualitatively by the ‘‘sticky’’ dissociative ELT model [91–93]. The rate constant of the ELT from 1-ethylnaphthalene(T2) to CCl4 is calculated to be 5.9 1010 s1, while the rate constant of the ELT from 1-ethylnaphthalene(T1) to CCl4 is calculated to be 660 s1. Thus, the dissociative ELT from the triplet excited NpD to CCl4 occurs only from the T2 state.

80


Figure 2.16. Plots of fELT versus fELT(calcd) for the ELT from NpDðTn Þ to CCl4 during the 355- and 425-nm two-color two-laser flash photolysis of BP with NpD in Ar-saturated acetonitrile in the presence of CCl4.

When some polychlorobenzene, such as chlorobenzene, 1,4-dichlorobenzene, and 1,2,4-trichlorobenzene, was employed as the electron acceptor, it was found that the ELT proceeds from the NpDðTn Þ, in which the dissociative ELT is not operative. Since the C—Cl bond dissociation rate of chlorobenzene radical anion was reported to be 1.8 109 to >2 1010 s1, the radical ion pair should be formed after the ELT from NpD(T2) [95, 96]. For the stepwise mechanism (Scheme 2.6), the dissociative ELT theory for the concerted dissociation cannot be applied [55]. The Marcus theory is adequate for the present system. According to the Marcus ELT theory, the free energy change for the ELT can be given by Eq. (10): GELT ¼ Eox Ered wp ET2

Stepwise mechanism D(Tn) + RX D·+ + RX·– Ion pair Concerted mechanism D·+ + R· + X– D(Tn) + RX

ð10Þ

D·+ + R· + X–

Scheme 2.6. Stepwise and concerted mechanisms for the dissociative ELT from DðTn Þ to RX to give D þ, R , and X.


The activation energy (G*) can be represented by Eq. (11) [94, 97]: l G 2 G ¼ 1þ 4 l

81

ð11Þ

where l is an intrinsic barrier corresponding to the bond length change and solvent reorganization. According to the literature, we employed l ¼ 1.9 eV [98]. By employing Eqs. (3) and (7)–(9), the fELT(calcd) values are estimated. The plots of fELT versus fELT(calcd) showed good correlation, indicating that the efficiency of the ELT from NpD(T2) to chlorobenzenes is explained qualitatively by the Marcus ELT theory. The ELT from the Tn state was inefficient in several NpD–polychlorobenzene pairs, where no NpD radical cation was detected, although bleaching and recovery of NpD(T1) were observed upon the irradiation of the second laser. Figure 2.17a shows the transient spectra obtained during the two-color two-laser experiment of the Np–1,2,4-trichlorobenzene system. The absorption band around 350 nm is reasonably assigned to the corresponding Cl—CHD and chlorine atom [99, 100]. Figure 2.17b shows simultaneous growth of Cl—CHD at 350 nm. Thus, it is suggested that Cl—CHD was generated from the homolytic C—Cl bond cleavage after the ENT from NpðTn Þ. It should be noted that the homolytic C—Cl bond dissociation energies of 1,2,4-trichlorobenzene, 1,2-dichlorobenzene, and chlorobenzene were reported to be 3.7, 3.8, and 3.8 eV, respectively, which are higher than the energy levels of chlorobenzenes(T1) (3.4–3.5 eV) [101]. Thus, it is concluded that the C—Cl bond dissociation takes place from 1,2,4-trichlorobenzeneðTn Þ (n > 1) after the ENT from NpðTn Þ (Scheme 2.7). This is the first example of the intermolecular ENT from the molecules in the Tn state to quenchers giving the quenchers in the Tn state. 6. Direct Observation of the Tn State The existence of the Tn states has been indirectly indicated by the bimolecular ENT from the Tn states to acceptors with the T1 energies higher than the T1 energy of the donor. Direct observation of the Tn states is indispensable for the study of the photoinduced reactions specific to the Tn state which has a much larger excitation energy than the T1 state. Since the lifetimes of the Tn states are reported to be on the order of picoseconds, utilization of short-pulse lasers is necessary to detect them. In this section, we present examples of direct observations of the Tn states using two-color two-laser flash photolysis employing a picosecond laser. For the detection of the Tn state, we employed the pump and probe method to attain better time resolution [102–104]. Oligothiophenes have been selected as the target molecule, since their photoexcitation processes are well established and they are known as electron donors with large extinction coefficients of the T1 state and radical cation in the D0 state [105–108]. The S0 and T1 states of the trimer, tetramer, and

82


(a)

0.20

0.15

0.10

0.05

0.00 340

360

380

400

420

440

Wavelength (nm) (b)

0.08

0.06

second laser fire

0.04

0.02

0.00 0

200

400 600 Time (ns)

800

1000

Figure 2.17. (a) Transient absorption spectra obtained during the one-laser (355-nm) excitation (solid line) and the two-laser (first 355-nm and second 425-nm) excitation (broken line) of BP (7.0 103 M) with naphthalene (7.0 103 M) in Ar-saturated acetonitrile in the presence of 1,2,4-trichlorobenzene (1.0 M). The second laser irradiation was at 150 ns after the first laser pulse. These spectra were obtained at 200 ns after the first laser. (b) The kinetic traces of O.D. at 350 nm with (dotted line) and without (solid line) the second laser irradiation.

pentamer of thiophene (3T, 4T, and 5T, respectively) can be excited selectively with the 355-nm nanosecond laser and the 532-nm picosecond laser, respectively. Figure 2.18a shows the transient absorption spectra of 4T in toluene at 40 ps before and at 20 and 150 ps after the second laser excitation;

83


NpD(Tn)+A IC

.

NpD + Cl-CHD +Cl

NpD+A(Tn)

.

IC

NpD(T2)+A

ENT1

NpD+A(T1)

ELT

(NpD·+ A· –)3

IC ENT2

NpD(T1)+A

NpD(S0)+A

A: triplet energy acceptor

Scheme 2.7. Schematic energy diagram for ELT from NpDðTn Þ to polychlorobenzene (A). As a competitive process to ELT, ENT generating AðTn Þ was confirmed.

Figure 2.18. (a) Transient absorption spectra observed at 40 ps before and 20 and 150 ps after the laser flash during two-color two-laser flash photolysis of 4T in toluene employing a nanosecond YAG laser (355 nm, FWHM 5 ns, 7 mJ pulse1) and a picosecond YAG laser (532 nm, FWHM 30 ps, 21 mJ pulse1). (b) Difference spectra of transient absorption spectra at 20 and 150 ps. (c) Kinetic traces of O.D. at 650 and 600 nm. Thick lines are fitted curves.

84


in the present study, the probe time was indicated with respect to the second laser excitation. At 40 ps, transient absorption peaks at 614 and 580 nm indicated the generation of 4T(T1) by the first laser excitation [105–107]. At 20 ps, the absorption bands of 4T(T1) became low and structureless and a new absorption band appeared around 650 nm, as shown in Figure 2.18b. The new absorption band at 650 nm with an absorption tail extending to around 800 nm can be attributed to the Tn state of 4T. In the present case, generation of 4T þ can be excluded since 4T þ showed a rather sharp absorption band at 640 nm [105–108]. Thermal effects can also be excluded from the quenching of Tn by the ENT. It should be noted that at 150 ps, the transient absorption spectrum had almost recovered to that before the second laser irradiation, indicating the quantitative relaxation to 4T(T1) from the Tn state by IC. The kinetic traces of O.D. at 650 and 600 nm (Fig. 2.18c) indicate that 4TðTn Þ (650 nm) decayed within 100 ps after the second laser irradiation with the concomitant recovery of 4T(T1) (600 nm). From the deconvolution fitting of a single exponential function to these kinetic traces, the rate constant of IC to 4T(T1) (kIC) was estimated to be (2.6 0.8) 1010 s1, which corresponds to a 38-ps lifetime of 4TðTn Þ. In the cases of 5T and 3T in toluene, the lifetimes of the Tn states were estimated to be 31 and 38 ps, respectively. Rentsh et al. [109] reported that the T2–T1 gaps of 3T and 4T were 1.48 and 1.37 eV, respectively. Although a comparison of oligothiophenes with ArH is difficult due to their different molecular structures, the estimated Tn state lifetimes of the oligothiophenes are close to those of ArH(T2) described earlier when taking the T2–T1 gaps into account. Thus, IC between T2–T1 is the ratedetermining step for the oligothiophenes. The solvent polarity effect was examined by employing acetonitrile as the solvent. The difference spectrum of 4T in acetonitrile at 20 ps showed a broad absorption band at 650 nm, which is essentially the same as that observed in toluene. At 150 ps, on the other hand, a sharp absorption band was confirmed at 640 nm, which can be attributed to 4T þ, indicating the ionization from 4TðTn Þ [105–108]. In the kinetic trace of O.D. at 600 nm, bleaching of 4T(T1) showed incomplete recovery, indicating that 38% of 4TðTn Þ changed to 4T þ. The radical cation formation in acetonitrile seems to be an adequate process since polar solvents stabilize the radical ion. The generation of a radical cation from the Tn state in acetonitrile was also confirmed for 5T in acetonitrile at 700 nm [105–108]. The Tn states of oligo(p-phenylenevinylene)s (OPVn, n denotes the number of phenyl rings, n ¼ 3, 4) were also investigated by means of the nanosecond– picosecond two-color two-laser flash photolysis. The lifetimes of the Tn states were estimated to be 35 and 30 ps for OPV3 and OPV4, respectively. Based on the ENT from OPVnðTn Þ to a series of triplet energy quenchers, the T2–T1 energy gaps of OPV3 and OPV4 were estimated to be 1.3 and 1.1 eV, respectively [103].


85

The T2 state lifetime was also estimated for CHR(T2). By combining the estimated lifetime (45 ps) and the ENT rate from the T2 state to a triplet energy quencher, the reaction distance of the bimolecular ENT was estimated to be ˚ in cyclohexane and acetonitrile, respectively, indicating that the 3.8 and 3.7 A collision between CHR(T2) and quenchers occurs in accordance with the exchange mechanism [104]. Direct measurement of the Tn state lifetime by nanosecond–picosecond twocolor two-laser flash photolysis was applied to BP derivatives with various substituents [110]. The Tn state lifetime of BP was estimated to be 37 ps, which is shorter than the value estimated by the energy transfer process discussed earlier, indicating the oversimplification of the assumption and the importance of the direct measurements. On the other hand, the tendency that the introduction of the substituents decreases the Tn state lifetime was also confirmed in the direct measurements. Furthermore, the ENT process from the Tn state to the quencher was investigated by analyzing the kinetic traces during the second picosecondlaser flash, from which lifetime-independent and lifetime-dependent processes were distinguished. It was revealed that the contribution of the lifetimedependent term on the ENT rate became larger as the size of the energy quencher increased.

B. Ketyl Radicals in the Excited State The nanosecond–picosecond two-color two-laser flash photolysis method is also useful to study the excited state of radicals, that is, the D1 state. We applied nanosecond–picosecond two-color two-laser flash photolysis to detect the absorption and fluorescence spectra of the ketyl radical of benzophenone and its derivatives (BPH and BPDH ) in the excited state in the UV–visible region [111], since BPH and BPDH are well investigated radicals in various fields. Since BPH and BPDH are generated from irreversible ways, such as photoionization, we employed a streak camera to realize the single-shot detection of these intermediates. BPH and BPDH were generated from the photoreduction of BP and BPD in cyclohexane. BP and BPD in the lowest triplet excited states (BP(T1) and BPD(T1)) decayed through the hydrogen abstraction from cyclohexane to produce BPH and BPDH after the first 266-nm nanosecond-laser irradiation. The generated BPH was excited at the visible absorption band using the second laser (532 nm, 30 ps FWHM) with the delay time of 1 ms after the first laser. Upon the excitation, BPH showed fluorescence with a peak at 564 nm. Similarly, fluorescence of BPDH was observed with the second laser irradiation. The measured tf of BPH is 2.0 ns, which is close to the reported value [112–114]. 4Chloro-, 4,40 -dichloro-, 4-bromo-, 4,40 -dibromo-, and 4-trifluoromethyl-substituted

86


(a)

0.15 c

a 0.10 0.05 b (b)

0.00 0.08 0.04 0.00 –0.04

350

400

450

500

550

600

Wavelength (nm)

Figure 2.19. (a) Transient absorption spectra observed at 0.5 (broken line a) and 20 (dotted line b) ns after the second laser irradiation during two-color two-laser photolysis (266 and 532 nm) and the spectrum observed during one-laser photolysis (266 nm, solid line c) of BP (1.0 104 M) in Ar-saturated cyclohexane. The second laser irradiation was at 1 ms after the first laser pulse. The transient absorption spectrum of BPH (D1) (b) was given by subtracting spectrum c from spectrum a. The blank around 532 nm in the spectra is due to the residual SHG of the Nd3þ:YAG laser.

BPDH showed larger tf values than others. It is generally admitted that tf increased with the decrease of the Stokes shift [115]. Immediately after the second laser irradiation, the bleaching of the absorption of BPH and growth of new transient absorption peaks at 350 and 480 nm were observed as shown in Figure 2.19a. The spectral shape of the new transient species was given by subtracting the spectrum observed before the second laser irradiation from that observed at 0.5 ns after the irradiation (Fig. 2.19b). The t estimated from the absorption decay was essentially the same as tf estimated from the fluorescence decay of BPH . Therefore, these transient absorption D1 transition (Scheme 2.8). Similar spectral bands can be attributed to the Dn changes were observed for other BPDH . The intermolecular reaction of BPH (D1) with the solvent molecules and the unimolecular cleavage of the O—H ketyl bond of BPH (D1) yielding BP and a hydrogen atom have been observed in the microsecond time scale [112–114]. Thus, the decay of BPH(D1) can be attributed to the combination of a chemical reaction and nonradiative and radiative transition processes


87

3.5 eV Dm 2.4 eV

Dn D2

D1

2.2 eV

kC hν532

kIC

kf

Product D0

OH

Scheme 2.8. Deactivation processes of BPH (D1) including radiative and nonradiative deactivation processes and chemical reactions.

(Scheme 2.8). Because the chemical reaction does not regenerate BPDH (D0), reaction rate constants can be determined based on the lifetime and yield of the recovery of BPDH(D0). Generally, the lifetime increased with the decrease of the Stokes shift (Table 2.3). The E(D1 D0) values of BPDH are similar to each other, indicating that the lifetime was not affected by the E(D1 D0) values. Therefore, it is suggested that BPDH (D1) such as 4-fluoro-, 4,40 -difluoro-, 4-methoxy-, and 4,40 -dimethoxy-substituted BPDH (D1) have shorter lifetimes and larger Stokes shift, showing higher reactivity with large kC. The distorted conformation in the D1 state may enhance the reactivity of such BPDH (D1). A similar relationship between the Stokes shift and the lifetime of the D1 state was confirmed for other BP analogues such as 4-benzoylbiphenyl and bis(biphenyl-4-yl)methanone [116]. In the case of ketyl radical of 4,40 -dimethoxybenzophenone, generation of the bis(4-methoxyphenyl)methanol cation and 4,40 -dimethoxybenzophenone radical anion was confirmed upon excitation of ketyl radical [117]. The generation of cation and radical anions indicates the ELT between the excited ketyl radical and parent molecule in the ground state. For the ELT mechanism, the following two pathways were indicated: (1) two-photon ionization of the ketyl radical followed by electron capture by the parent molecule and (2) an intermolecular collisional ELT process between the excited ketyl radical and

88


TABLE 2.3 Stokes Shifts for BPDH (m SS), Lifetimes of Transient Absorption (s) and Fluorescence (sf), Rate Constants of the Chemical Reaction (kC), Rate Constants of the Radiative (kf) and Nonradiative (kIC) Relaxation Processes, and the Energy Gaps of BPDH Between the D1 and D0 States (E(D1 D0)) Ketyl Radical

t nSS (103cm1) (ns)

Benzophenone 4-Fluoro 4,40 -Difluoro4-Chloro4,40 -Dichloro4-Bromo4,40 -Dibromo4-Trifluoro4-Methyl4-Methoxy4,40 -Dimethyl4,40 -Dimethoxy

0.75 0.96 1.5 0.75 0.72 0.78 0.69 0.73 0.79 1.2 0.78 1.7

a

2.0 0.1 1.4 0.47 3.3 0.1 3.4 0.4 2.1 3.5 0.1 4.3 0.3 1.9 0.1 0.86 0.04 1.7 0.1 0.34 0.02

tf (ns) 2.0 0.1 1.3 0.1 0.47 3.2 0.1 3.4 0.1 2.1 3.6 0.1 4.3 0.3 1.8 0.89 1.6 0.35

kC (108 s1)

kf þ kIC (108 s1)

1.7 0.1

3.4 0.1

a

a

a

a

a

a

1.4 0.1 2.2 0.1 1.8 0.1

1.5 0.1 2.8 0.1 1.1 0.1

a

a

a

a

a

a

a

a

b

b

Not determined because BPDH (D0) was not observed due to the overlap of the Dn The recovery was not observed.

E (D1D0)(eV) 2.20 2.18 2.12 2.16 2.11 2.16 2.10 2.15 2.18 2.11 2.14 2.06 D1 absorption.

b

parent molecule (Scheme 2.9). The two-photon ionization of the ketyl radical in process (1) was confirmed by the laser power dependence of the absorption band of the cation. Furthermore, a fluorescence decay rate that depends on the concentration of the parent molecule supports the ELT process (2) at the diffusion-limiting rate. These dual ELT pathways were also confirmed for BP [116]. The D1 state properties were also examined with xanthone ketyl radical (XnH ). The absorption and fluorescence of XnH (D1) were observed for the first time by using the nanosecond–picosecond two-color two-laser photolysis [118]. Several factors governing the deactivation processes of XnH(D1) such as interaction and reaction with solvent molecules were pointed out. The remarkable change of reactivity of XnH (D1) compared with that in the ground state (XnH (D0)) was indicated from the experimental results. The rapid halogen abstraction of XnH (D1) from some halogen donors, such as carbon tetrachloride (CCl4), was found to occur. The halogen abstraction occurred more efficiently in the polar solvents than in the nonpolar solvents. It is suggested that the polar solvents promote the spin distribution of XnH (D1) of the phenyl ring favorable to the halogen abstraction. Time-resolved absorption and fluorescence spectra of azaxanthone (AX) ketyl radical (AXH ) in the excited state (AXH ðDn Þ (n ¼ 1 or 2)) were also


89

Scheme 2.9. Dual ELT pathways of ketyl radical of 4,40 -dimethoxybenzophenone (1 in this scheme).

observed during nanosecond–picosecond two-color two-laser flash photolysis [119]. AXH showed dual fluorescence peaks at 460 and 645 nm, which were assigned to the D2 ! D0 and D1 ! D0 transitions, respectively (Fig. 2.20). The lifetime of the D2 ! D0 fluorescence (1.0 ns) was longer than that of the D1 ! D0 fluorescence (0.4 ns). Fluorescence quantum yields (f) of the D1 ! D0 and D2 ! D0 fluorescence were estimated to be 0.0008 0.0002 and 0.05 0.02, respectively. These anomalous emitting properties can be attributed to the pyridine ring in AX. As discussed earlier, the excited radical has an enhanced reactivity. Especially, excited ketyl radicals can be used as excellent reducing agents, which can be generated by the two-laser excitation. Thus, we employed excited ketyl radicals as reducing agents to generate metal nanoparticles [120]. In order to generate gold nanoparticles (AuNps) in a polymer matrix, a two-color laser was irradiated to the poly(vinyl alcohol) (PVA) film including BP and AuCl 4 . The first laser irradiation generated the BPH , which was confirmed by the transient absorption spectrum. Upon excitation of BPH with the second laser, BPH showed fluorescence, although the fluorescence intensity became weak in the presence of AuCl 4 , indicating the reduction of AuCl4 by the excited ketyl radical. Actually, the PVA film including BP and AuCl4 changed color upon the two-color laser irradiation, indicating the surface plasmon band due to the formation of AuNps, while one laser irradiation did not change the color of the film. The formation of AuNps with a 2.5–4 nm diameter was

90


0.10 OH

0.08

N

460 nm

O

AXH Fluor. Int.

0.06 645 nm 0.04

0

2 4 Time (ns)

6

0.02

0.00 400

500 600 Wavelength (nm)

700

800

Figure 2.20. Absorption (black line) and fluorescence spectra of the AXH in Ar-saturated cyclohexane at room temperature. Fluorescence spectra were obtained during the 266- and 355-nm (dark gray line) or 266- and 532-nm (light gray line) two-color two-laser flash photolysis. The absorption spectrum was obtained during onelaser photolysis (266-nm, black line) of AX (4.0 104 M). The second laser irradiation was at 1 ms after the first laser pulse. All the fluorescence spectra of AXH were normalized with the corresponding absorption peaks. Inset: Kinetic traces of the fluorescence intensity of AXH at 460 and 645 nm.

also confirmed by TEM (Fig. 2.21). This fact indicates that the enhanced reducing power of the excited BPH efficiently reduced AuCl 4 generating AuNps. Since the reaction of the excited radical can be limited to a rather small region by controlling the overlapping volume of the two-laser light,

Figure 2.21. (a) TEM image of AuNps in HAuCl4-doped PVA film containing BP. (b) Distribution of diameter of AuNps formed in PVA films.

91


three-dimensional control of the AuNps fabrication can be achieved using the present method.

C. Excited Radical Cations Fluorescence of radical cation gives important information during various reaction processes. For example, distance between the ion pair can be estimated based on the Fo¨rster-type ENT theory under the condition where a pair of radical ions generated from the photoinduced ELT are fluorescent and there is a good spectral overlap of fluorescence and absorption for the pair. Here, we introduce an example in which two-color two-laser flash photolysis was employed to estimate the distance between TMB þ and the radical anion of 1,4-dicyanonaphthalene (DCN ) [121], since TMB þ* shows fluorescence as indicated earlier [26]. Fluorescence intensity of TMB þ generated during the laser flash photolysis of a N2-purged acetonitrile solution containing TMB and DCN with a XeCl laser pulse (308 nm, 30 ns FWHM) was measured with various delay times of the second laser pulse (532 nm, 5 ns FWHM). The fluorescence intensity of TMB þ immediately after the first pulse was unexpectedly weak, although its concentration was at the maximum as observed in the trace of the transient absorption of TMB þ. The fluorescence intensity increased with the increase of the delay time and reached maximum at approximately 220 ns, and decreased in a second-order kinetics. On the other hand, the fluorescence in an aerated solution did not show the initial increase and decreased monotonously in a second-order kinetics, in accordance with the decay of the transient absorption of TMB þ. Since DCN was readily quenched by oxygen within 50-ns transient absorption measurements, the rise of the fluorescence intensity corresponds to the decrease of the ENT quenching of TMB þ* by DCN with the increase of the distance between TMB þ* and DCN . The transient absorption due to TMB þ showed no depletion by the 532-nm excitation, indicating not ELT quenching but Fo¨rster-type ENT quenching of TMB þ* by DCN . From the time-dependent fluorescence intensity, it was revealed that the distance between ions increased from 0.6 nm at 50 ns to 1.7 nm at 200 ns. From these values, the apparent diffusion constant was estimated to be 8.5 108 cm2 s1, which is much smaller than the typical value, 2 105 cm2 s1, for acetonitrile. It could be attributed to the contribution of the Coulombic interaction between the radical ions.

D.

Other Reactive Intermediates

1. Two-Color Two-Laser DNA Damaging Photodynamic therapy (PDT) is a promising treatment for cancer based on the photosensitized oxidative reaction at the diseased tissues producing cell death, and DNA is considered as a potential target [122]. Compared with surgery and chemotherapy, the

92


Scheme 2.10. Mechanism of DNA damage by two-color two-laser irradiation. The first laser irradiation causes charge separation between sensitizer (S) and DNA. The second laser causes photo ejection from S , making charge recombination impossible. Then, chemical reaction of DNA þ occurs leading to DNA damage.

combination of a photosensitizer (S) uptake in malignant tissues and selective light delivery offers the advantage of a selective method of destroying diseased tissues without damaging surrounding healthy tissues. Excitation of DNAbound sensitizers produces the S / DNA þ charge-separated state through photoinduced ELT. However, the efficiency of producing photosensitized DNA damage is low because the charge recombination rate is usually much faster than the process leading to DNA damage, such as the reaction of G þ with water [123–126]. Here, we introduce the first study of nanosecond-laser DNA damaging, using a combination of two-color two-laser pulses as a promising new strategy to reach a high DNA damaging efficiency. The first laser pulse was applied for the production of S and DNA þ, and the second laser pulse for the electron ejection from S , making the reaction irreversible (Scheme 2.10) [127]. In this study, naphthaldiimide (NDI) was selected as an S that can be excited with the first laser at a wavelength of 355 nm [128–132]. First, to assess the feasibility of electron ejection from S bound to DNA, the pulse radiolysis–laser flash photolysis of NDI-conjugated oligodeoxynucleotide (NDI-ODN) was performed (Fig. 2.22) [133]. NDI with a maximum absorption peak at 495 nm [129] was generated from the electron attachment during pulse radiolysis of NDI-ODN. Since S often absorbs light at a longer wavelength compared with its nonreduced form, laser pulses with a longer wavelength can be used for the excitation of S , and a 532-nm laser was applied as the second laser. Irradiation of NDI in NDI-ODN with a 532-nm laser pulse caused a decrease in O.D. of NDI and the formation of absorption at 630 nm assigned to a solvated electron (eaq) immediately after the flash (Fig. 2.22, inset), demonstrating the successful electron ejection from NDI to the solvent water. Figure 2.23 shows the time profile of NDI in the one-color laser photolysis of NDI-ODN. Upon the first laser excitation, hole transfer via consecutive fast

93


Figure 2.22. Electron ejection from NDI promoted by a 532-nm laser pulse. NDI was generated from the electron attachment during pulse radiolysis of NDI-ODN (NDIAAAAAAGTGCGC/TTTTTTCACGCG) (gray), and photoirradiated with a 532-nm laser pulse at 2 ms after the electron pulse (black). Inset: Formation and decay of the solvated electron monitored at 630 nm.

adenine hopping leads to a charge-separated state within the laser duration (5 ns), and the charge recombination proceeds by the single-step superexchange from the ˚ away from NDI with a lifetime of guanine (G) radical cation (G þ) about 14 A 240 ns [123, 134, 135]. Figure 2.23 also shows the consumption of G as a function

Figure 2.23. Formation and decay of NDI and the effect of the delay time between two laser pulses on the consumption of G during the laser flash photolysis of NDI-ODN (NDITTTCGCGCTT/AAAGCGCGAA).The transient absorption of NDI was monitored at 495 nm following the 355-nm excitation (left axis). The consumption of G is plotted as a function of the delay of the 532-nm pulse with respect to the 355-nm pulse ( , right axis). The dashed line shows the consumption of G in the absence of the 532-nm pulse.

94


of the delay time of the second laser pulse in the time-delayed two-color photolysis. The delay time dependence of the consumption of G agreed well with the decay of the transient absorption of NDI obtained in one-color laser photolysis. Thus, the acceleration caused by the second laser is clearly based on the excitation of NDI . The experiments were performed under single-hit (i.e., low-conversion) conditions where, on average, each duplex reacts once or not at all, and the consumption of G was linearly correlated with the irradiation time and the power of the second laser in the present experimental arrangement.

2. Two-Color Laser Photolysis for Determination of the Rate Constant from the Product Analysis o-Quinodimethane is one of the extensively studied intermediates. Cycloaddition of o-quinodimethane with alkenes and alkynes is one of the well-investigated fields [136–138], but the reported kinetic studies are mainly of some substituted o-quinodimethanes [139], and, to the best of our knowledge, the rate constant of the cycloaddition of parent o-quinodimethane in room-temperature solutions has not been reported so far. The lack of such important and basic kinetic data seems to be due to the difficulty in conducting the experiments by spectroscopic means. In the cycloaddition reactions, it is possible to observe the decay of o-quinodimethane that is generated from conventional precursors by the flash photolysis technique because in most cases the absorption of o-quinodimethane appears at a longer wavelength than the precursors. However, the decay does not simply reflect the formation of the cycloadduct because o-quinodimethane also gives other products, such as its dimers, oligomers, and polymers [140, 141]. There are also difficulties in tracing the formation of the cycloadduct spectroscopically because the absorption of the cycloadduct appears at the same wavelength region as the precursors (a large amount of the precursor remains intact even after the laser pulse irradiation due to the low photochemical efficiency) and the above-mentioned dimers, oligomers, and polymers. The strategy used in our experiment for determination of the rate constant of the cycloaddition of o-quinodimethane and maleic anhydride in room-temperature solutions was the fast and efficient generation of o-quinodimethane during the first laser pulse irradiation and the quenching of the reaction by the decomposition of remaining o-quinodimethane during the second laser pulse irradiation (Scheme 2.11) [142]. When o-quinodimethane is generated during the first KrF laser pulse irradiation in the presence of maleic anhydride, the cycloaddition of o-quinodimethane with maleic anhydride gives cis-1,2,3,4-tetrahydro-2,3-naphthalenedicarboxylic anhydride (path 1 with the rate constant of k1) together with thermal products of o-quinodimethane [140, 141], which are formed by second-order kinetics [138] (path 2 with the bulk rate constant of k2). The reaction is quenched by the second XeCl laser pulse after a particular time in the course of the reaction to decompose the remaining o-quinodimethane, partly forming benzocyclobutene [bicyclo(4.2.0)-octa-1,3,5-

95


O O O cis-1,2,3,4-tetrahydro-2,3-naphthalenedicarboxylic anhydride O O

Δ

path 1; k1

O maleic anhydride

SePh

KrF excimer laser 2hν -2PhSe

SePh

Δ

path 2; k2

o-quinodimethane + hν

other photochemical products

XeCl excimer laser

other thermal products

+ benzocyclobutene

Scheme 2.11. Determination of the rate constants of the cycloaddition reaction of o-quinodimethane based on the product analysis depending on the delay time between the first 248-nm laser irradiation and second 308-nm laser irradiation. o-Quinodimethane was generated by the first 248-nm laser pulse irradiation. The cycloaddition reaction with maleic anhydride (path 1 with the rate constant of k1) and the dimerization (path 2 with the bulk rate constant of k2) were quenched by the photochemical conversion of o-quinodimethane into benzocyclobutene with the second 308-nm laser pulse irradiation.

triene] as a photochemical product. By analyzing the dependence of delay time between the first and second lasers (0–0.1 s) on the yield of tetrahydronaphthalenedicarboxylic anhydride, the rate constant of the cycloaddition of o-quinodimethane and maleic anhydride (path 1: o-quinodimethane þ maleic anhydride ! tetrahydronaphthalenedicarboxylic anhydride) was determined to be 2.1 105 M1 s1. To the best of our knowledge, this is the first report on the determination

96


of the rate constant of the cycloaddition of parent o-quinodimethane with an alkene. Therefore, this is an interesting example of the application of two-color laser photolysis to determine the rate constant even under the condition where the spectroscopic analysis is difficult [143]. 3. Two-Color Laser Control of Photocatalytic Reaction on TiO2 Surface Photocatalytic reactions on the semiconductor surface attract quite a lot of attentions today. Especially, the photocatalytic reaction on the TiO2 surface has been widely used in areas such as environmental purification, hydrogen production, gas sensors, and dye-sensitized solar cells. Typically, the reaction processes are initiated by the band-gap excitation of TiO2 particles with UV irradiation to generate reactive intermediates such as hole, electron, and various oxygen-related species. The reactivity and lifetime of these oxidizing species play an important role in controlling the overall kinetics of the oxidative processes. Recently, Keggin-type polyoxometalates (POMs) have been applied to TiO2 photocatalytic systems as electron scavengers to retard fast charge recombination between the hole and electron to enhance the reactivity of the hole. In addition, POM, which is generated by a one-electron reduction of POM by the conduction band electron of TiO2 nanoparticles, absorbs visible light to form the excited POM i.e., POM*, which synergistically catalyzes the reduction process of solute in the solution [144]. Thus, by employing two-color two-laser flash photolysis, the reduction process via POM* can be examined. By employing methylviologen (MV2þ) as an indicator of the reduction via POM*, we studied two-color laser control of the photocatalytic reaction of the TiO2/POM/MV2þ ternary system [145]. The electron transfer from the conduction band of TiO2 to POM was confirmed by the increase of the visible absorption band during the laser flash photolysis of TiO2/POM colloidal solution. The efficiency of the electron transfer from TiO2 to POM was on the order of H2W12O406 < SiW12O406 < PW12O403, depending on the reduction potential of the POMs. Electron injection from PW12O404* to the conduction band of TiO2 was clearly observed as the bleaching of the absorption band due to PW12O404 upon excitation of PW12O404 with the second laser (Fig. 2.24). In the presence of MV2þ, the extent of the bleaching decreased with an increase in the concentration of MV2þ. Direct electron transfer from TiO2 to MV2þ is negligible under the present condition, while complex formation is possible between POM and MV2þ. Thus, the present observation can be attributed to the electron transfer from PW12O404* to MV2þ generating POM and MV þ. Because the reduction potential of MV2þ is more negative than that of PW12O403, cascade electron transfer from MV þ to PW12O403 is possible, which is in accordance with the absence of the absorption band of MV þ in Figure 2.24. Scheme 2.12 summarizes the energy diagram for the TiO2/POM photocatalytic redox process. By employing two-color two-laser flash photolysis, electron


97

Figure 2.24. (a) Transient absorption spectra obtained at 0.1 ms (solid squares) before and 0.3 ms (open circles), 4 ms (solid triangles), and 30 ms (open inverted triangles) after a second 532-nm laser flash following the first 355-nm laser flash with a delay time of 1 ms during two-color two-laser flash photolysis of an argon-saturated TiO2/PW12O403 colloidal solution in the presence of MV2þ. (b) Time traces observed at 605 nm during two-color two-laser flash photolysis of TiO2/PW12O403 colloidal solution in the absence and the presence of MV2þ.

transfer from POM* became possible. This process is interesting in terms of mimicking the ‘‘Z-scheme’’ of the natural photosynthesis system. In this section, we introduced several examples of studies employing twocolor two-laser flash photolysis. In many cases, the second laser has been employed to excite intermediates, such as triplet excited states, radicals, and radical ions. Thus, two-color two-laser flash photolysis can be regarded as a powerful tool to produce the higher excited intermediates, which cannot be accessed by the single laser flash. As indicated earlier, the higher excited

98


Scheme 2.12. Energy diagram for the TiO2/POM photocatalytic redox processes. Dotted and broken arrows represent the deactivation and charge recombination processes, respectively.

intermediates are generally short-lived species. Thus, investigation in the picosecond regime is important to elucidate the properties of the higher excited state. To detect such short-lived species, we introduced two methods—streak camera detection and the pump and probe method. The detection method using a streak camera is useful for detecting an intermediate generated from an irreversible process. Fluorescence detection is also possible by using a streak camera. Thus, we employed streak camera detection for studies on the excited ketyl radicals. On the other hand, the pump and probe method gives kinetic parameters of the reversible processes with better time resolution than the streak camera detection. We employed it for studying the Tn state. By employing a femtosecond laser, dynamic properties of higher excited states in the subpicosecond regime will become clear in the near future. It should be stressed that two-color two-laser flash photolysis is not limited to the study of higher excited states. For example, the study on DNA cleavage described earlier is one such example. Two-color two-laser flash photolysis is an advantageous DNA cleavage method, which realizes site selectivity to avoid damage to healthy cells and to perform selective damage of cancer tissue deep beneath the skin surface. Application of two-laser photolysis to photodynamic therapy is also possible by other approaches, since two-laser photolysis generates site selectivity for highly reactive intermediates, such as radicals, which are expected to induce damage on various diseased tissues. Generation of highly reactive intermediates by the second laser excitation is also beneficial in the decomposition of harmful pollutants in the air. Therefore, application of two-laser photolysis to semiconductor catalytic reactions is another interesting approach. Two-color two-laser photolysis is also an interesting approach to realize ultrafast control over the function of molecular devices.

99

THREE-COLOR THREE-LASER FLASH PHOTOLYSIS

IV. THREE-COLOR THREE-LASER FLASH PHOTOLYSIS Combination of pulse lasers can be achieved rather easily by using delay circuits as indicated earlier. We have already reported the studies using three-color threelaser photolysis. In this section we briefly introduce the results to show the availability of the multibeam irradiation method.

A. Three-Laser Control of Intermediate Population Three-color three-laser photochemistry of di(p-methoxyphenyl)methyl chloride ((p-CH3OC6H4)2CHCl ¼ An2CHCl) was studied by three-step excitation using 308-, 355-, and 495-nm lasers with delay times of 100 ns to 3 ms (Scheme 2.13) [146]. Di(p-methoxyphenyl)methyl radical (An2CH ) was produced together with An2CH in the excited state (An2CH*) and di(p-methoxyphenyl)methyl cation (An2CHþ) in the quantum yields of 0.09, 0.12, and 0.12, respectively, after a laser flash during the 308-nm laser (first laser) photolysis of An2CHCl in acetonitrile. Excitation of An2CH with the 355-nm laser (second laser) resulted in the formation of transient absorption of An2CH* and An2CHþ and fluorescence of An2CH* with a peak at 550 nm. The formation of An2CHþ from An2CH requires two-photon energy at 355 nm and proceeds by the resonant two-photon ionization (RTPI) of An2CH through sequential excitation of An2CH *. Excitation of An2CHþ with the 495-nm laser (third laser) produced fluorescence with D0

S1 za

hν495

ni

io

hν355

I. P.

tio n

Dn S1

hνf560

S0

IC

hν308

e

ag av

cle

hν355

D1 hνf550

D0 S0 An2CHCl

An2CH .

An2CH+

Scheme 2.13. First laser generates di(p-methoxyphenyl)methyl radical (An2CH ) together with An2CH in the excited state (An2CH *) and di(p-methoxyphenyl)methyl cation (An2CHþ). Second laser excitation of An2CH resulted in the formation An2CH *, which shows fluorescence at 550 nm ðhnf 550 Þ, and An2CHþ. Excitation of An2CHþ with the 495-nm laser (third laser) produced fluorescence with a peak at 560 nm ðhnf 560 Þ.

100


a peak at 560 nm. Although the fluorescence of An2CHþ was also observed without the second laser excitation because of the initial formation of An2CHþ during the first 308-nm laser photolysis, the fluorescence intensity of An2CHþ increased approximately 1.2 times with the second 355-nm laser excitation of An2CH . Therefore, the second laser excitation can perform the conversion of An2CH to An2CHþ through RTPI within the laser flash duration, and the fluorescence intensity of An2CHþ can be controlled by the second laser irradiation.

B. Stepwise Bond Cleavage of Two C—O Bonds Via the Tn State As described previously, (4-benzoylphenoxy)methylnaphthalene exhibited C—O bond dissociation via the Tn state. The multiple laser excitation technique allows us to study stepwise cleavage of two equivalent bonds in a molecule. To provide clear evidence for the stepwise photocleavage of two equivalent bonds in a molecule, we introduce here two C—O bond cleavages of 1,8-bis[ (4-benzoylphenoxy)methyl]naphthalene (1,8-(BPO-CH2)2Np) to give acenaphthene with the three-step excitation using three-color three-laser flash photolysis (Scheme 2.14) [147].

O

O

O

O

ENT

O

O

O

O hν430

T1

hν308

+ H2C D0

< 5 ns, 8.5%

< 20 ns

O

hν355

O

O

O H2C

CH2

H2C < 5 ns

< 5 ns

O

CH2 + O

1,8-(BPO-CH2)2Np

acenaphthene

Scheme 2.14. Three-color three-laser photochemistry of 1,8-(BPO-CH2)2Np involving intramolecular triplet ENT, selective excitation of 1,8-(BPO-CH2)2Np(T1) to 1,8-(BPOCH2)2NpðTn Þ, the C—O bond cleavage from 1,8-(BPO-CH2)2NpðTn Þ to give 1-(BPOCH2)NpCH2 , and selective excitation and C—O bond cleavage of 1-(BPO-CH2)NpCH2 to give acenaphthene as the stable product through 1,8-( CH2)Np. Dotted square shows the excitation energy delocalization.

THREE-COLOR THREE-LASER FLASH PHOTOLYSIS

101

Figure 2.25. Transient absorption spectra obtained at 300 ns after the first (308 nm) laser (a, broken line) and at 100 ns after the second (430 nm) laser (b, solid line) during laser flash photolysis of 1,8-(BPO-CH2)2Np in Ar-saturated acetonitrile. The delay time of the two lasers was 200 ns. Top Inset: Time profiles of transient absorptions obtained at 415 and 345 nm during one-laser irradiation at 308 nm (c and e) and two-laser irradiation at 308 and 430 nm (d and f). The growth of the transient absorption in the time scale of a few tens of nanoseconds was due to the formation of Np(T1) through intramolecular triplet ENT from BP(T1) to Np. Bottom Inset: Plots of O.D.345 of 1,8-(BPOCH2)2Np versus 430-nm laser intensity (I430).

A transient absorption spectrum was observed during the 308-nm laser irradiation of 1,8-(BPO-CH2)2Np (6.8 105 M) in acetonitrile (Fig. 2.25a). Absorption peaks at 400 and 425 nm were assigned to 1,8-(BPO-CH2)2Np(T1), whose energy was localized at the Np moiety because of an intramolecular triplet ENT from BP(T1) to Np(S0) producing Np(T1). Immediately after the second 430-nm laser (7 mJ pulse1) flash, the formation of a transient absorption band around 345 nm and bleaching of the peaks of 1,8-(BPO-CH2)2Np(T1) at 400 and 425 nm were observed (Fig. 2.25b). Because 1,8-(BPO-CH2)2Np(S0) has no absorption at 430 nm, only 1,8-(BPO-CH2)2Np(T1) was excited to give 1,8-(BPO-CH2)2NpðTn Þ during the second 430-nm laser irradiation. The absorption around 345 nm was assigned to 1-(BPO-CH2)NpCH2 [67, 74]. These results reasonably show that the Tn state energy is higher than the C—O bond dissociation energy and it delocalizes in the molecule including the C—O bonds. The quantum yield () of the formation of 1-(BPO-CH2)NpCH2 from the photoreaction of 1,8-(BPO-CH2)2Np(T1) was calculated to be 0.085 0.004. Because 1-(BPO-CH2)NpCH2 has an absorption at 355 nm, 1-(BPOCH2)NpCH2 was irradiated by the third 355-nm laser (10 mJ pulse1) at 200 ns after the second 430-nm laser irradiation. The results of the three-color three-laser photolysis of 1,8-(BPO-CH2)2Np are shown in Figure 2.26. Bleaching

102


Figure 2.26. Three-color three-laser photolysis of 1,8-(BPO-CH2)2Np (a, spectra; b, time profiles detected at 345 nm) in Ar-saturated acetonitrile. The transient spectra observed during the irradiation of the 308-nm laser (at 500 ns after the laser pulse) (a), successive irradiation with the 308- and 430-nm lasers (at 300 ns after the second laser pulse; delay time between the two lasers: 200 ns) (b), successive irradiation of the 308- and 355-nm lasers (at 100 ns after the second laser pulse; delay time between the two lasers: 400 ns) (c), successive irradiation of the 308-, 430-, and 355-nm lasers (at 100 ns after the third laser pulse; delay time between the lasers: 200 and 200 ns) (d), and irradiation of the 355nm laser (at 100 ns after the laser pulse) (e). The inset in (a) shows the spectra b a (f) and d c (g). In panel (b), 1, 2, and 3 refer to the irradiation sequence order of the 308-, 430-, and 355-nm lasers, respectively.

of the transient absorption at 345 nm during the third 355-nm laser irradiation was clearly observed, indicating the C—O bond cleavage from 1-(BPO-CH2)NpCH2 ðDn Þ. It is reported that the molecular orbital of the Dn state is delocalized not only on the Np chromophore but also on the C—O s orbital. Formation of acenaphthene from 1-(BPO-CH2)NpCH2 ðDn Þ may be explained by the radical backside attack mechanism or the second C—O bond cleavage to give the 1,8( CH2)2Np biradical, which rapidly cyclizes to form acenaphthene. However, since similar results were observed for 1,4-(BPO-CH2)2Np [148], the formation of acenaphthene is most likely explained by the 1,8-(CH2)2Np biradical mechanism (Scheme 2.14). The absorption of 1,8-(CH2)2Np, which was reported to appear at 500 nm [149], was not observed probably due to the small absorption coefficient, the formation of 1,8-(CH2)2Np at a low concentration, and fast cyclization of 1,8-( CH2)2Np that occurred within the laser duration (5 ns). Similar stepwise bond cleavage was confirmed for 1,8-bis(phenoxymethyl)naphthalene (1,8-(PhOCH2)2Np) and 1,4-(PhOCH2)2Np, while not with 1,8bis(hydroxymethyl)naphthalene (1,8-(HOCH2)2Np) and 1,4-(HOCH2)2Np. Furthermore, the cleavage yields of 1,8-(PhOCH2)2Np and 1,4-(PhOCH2)2Np were larger than those of 1,8-(BPOCH2)2Np and 1,4-(BPOCH2)2Np. In addition, cleavage yields with 1,8-substituted compounds were larger that those of 1,4subsituted compounds, when the compounds bear the same substituents. These

ACKNOWLEDGMENTS

103

facts indicate that the cleavage yield depends on the types and position of the substituents [148]. We introduced two examples of three-color three-laser photolysis. The former example employed the third laser as an excitation source to evaluate the amount of intermediates generated by the first and second lasers irradiation. On the other hand, the latter example used the second and third lasers to promote bond dissociation from the respective Tn and Dn states. The role of each laser is quite different. These examples indicate that one can control reactions by selecting laser wavelength and delay time based on the properties of each intermediate. Combination of pulse radiolysis and two-laser flash photolysis will be another interesting subject of three-beam excitation chemistry to be investigated in the near future. In this case, radical ions generated during pulse radiolysis will be excited by the successive two lasers. That is, one can introduce additional pathways to the reaction scheme indicated in Scheme 2.3. Utilization of a dyad or triad molecular system will realize fast intramolecular reaction systems applicable to the multibit molecular memory.

V.

CONCLUSIONS

In this chapter, several examples of photochemistry of short-lived species by multibeam irradiation are introduced. In many cases, the properties of excited intermediates have been investigated by using nanosecond lasers. Since the lifetime of excited intermediates is usually quite short, investigations employing ultrashort laser pulses are intrinsically important. Various properties estimated by the direct manner will appear in the near future. Recent progress of ultrashort pulse lasers and detection systems will make this possible. On the other hand, utilization of multiple excitations is not limited to the basic study of excited intermediates. These are applicable to biological and environmental fields. Further fruitful results are expected for these explorations.

ACKNOWLEDGMENTS We thank our collaborators, particularly Dr. Xichen Cai, Dr. Masanori Sakamoto, Dr. Kiyohiko Kawai, Dr. Takashi Tachikawa, Dr. Akihiko Ouchi, Dr. Minoru Yamaji, Dr. Nobuyuki Ichinose, Dr. Michihiro Hara, and Mrs. Sachiko Tojo and Mr. Yosuke Oseki, for their contribution to this work, as well as the members of the Radiation Laboratory of SANKEN, Osaka University, for running the linear accelerator. This work has been partly supported by a Grant-in-Aid for Scientific Research (Project 17105005, Priority Area (417), 21st Century COE

104


Research, and others) from the Ministry of Education, Culture, Sports, Science and Technology (MEXT) of the Japanese Government.

REFERENCES 1. J. C. Scaiano, Acc. Chem. Res. 1983, 16, 234. 2. Handbook of Radiation Chemistry, CRC Press, Boca Raton, FL, 1991. 3. J. C. Scaiano, L. J. Johnston, W. G. McGimpsey, and D. Weir, Acc. Chem. Res. 1988, 21, 22. 4. J. C. Scaiano and P. J. Wagner, J. Am. Chem. Soc. 1984, 106, 4626. 5. L. J. Johnston and J. C. Scaiano, J. Am. Chem. Soc. 1986, 108, 2349. 6. J. C. Scaiano, W. G. McGimpsey, and H. L. Casal, J. Am. Chem. Soc. 1985, 107, 7204. 7. T. Gannon and W. G. McGimpsey, J. Org. Chem. 1993, 58, 5639. 8. L. J. Johnston and R. W. Redmond, J. Phys. Chem. A 1997, 101, 4660. 9. W. G. McGimpsey and J. C. Scaiano, Chem. Phys. Lett. 1987, 138, 13. 10. Z. Wang and W. G. McGimpsey, J. Phys. Chem. 1993, 97, 9668. 11. G. A. Smith and W. G. McGimpsey, J. Phys. Chem. 1994, 98, 2923. 12. Y. Kuriyama, F. Hashimoto, M. Tsuchiya, H. Sakuragi, and K. Tokumaru, Chem. Lett. 1994, 1371. 13. T. W. Ebbesen, R. Akaba, K. Tokumaru, M. Washio, S. Tagawa, and Y. Tabata, J. Am. Chem. Soc. 1988, 110, 2147. 14. T. W. Ebbesen, J. Phys. Chem. 1988, 92, 4581. 15. W. G. McGimpsey and J. C. Scaiano, J. Am. Chem. Soc. 1988, 110, 2299. 16. C. Bohne, S. R. Kennedy, R. Boch, F. Negri, G. Orlandi, W. Siebrand, and J. C. Scaiano, J. Phys. Chem. 1991, 95, 10300. 17. Z. Wang and W. G. McGimpsey, J. Phys. Chem. 1993, 97, 5054. 18. M. P. Debreczeny, W. A. Svec, E. M. Marsh, and M. R. Wasielewski, J. Am. Chem. Soc. 1996, 118, 8174. 19. D. Gosztola, M. P. Niemczyk, and M. R. Wasielewski, J. Am. Chem. Soc. 1998, 120, 5118. 20. R. T. Hayes, M. R. Wasielewski, and D. Gosztola, J. Am. Chem. Soc. 2000, 122, 5563. 21. A. S. Lukas, P. J. Bushard, and M. R. Wasielewski, J. Am. Chem. Soc. 2001, 123, 2440. 22. M. Andersson, L. E. Sinks, R. T. Hayes, Y. Zhao, and M. R. Wasielewski, Angew. Chem. Int. Ed. 2003, 42, 3139. 23. K. Watanabe, T. Asahi, H. Fukumura, H. Masuhara, K. Hamano, and T. Kurata, J. Phys. Chem. B 1998, 102, 1182. 24. A. Mozumder and Y. Hatano, Charged Particle and Photon Interactions with Matter: Chemical, Physicochemical, and Biological Consequences with Applications, Marcel Dekker, New York, 2004.

REFERENCES

105

25. A. Ishida, M. Fukui, H. Ogawa, S. Tojo, T. Majima, and S. Takamuku, J. Phys. Chem. 1995, 99, 10808. 26. N. Ichinose and T. Majima, Chem. Phys. Lett. 2000, 322, 15. 27. N. Ichinose, S. Tojo, and T. Majima, Chem. Lett. 2000, 1126. 28. T. Majima, M. Fukui, A. Ishida, and S. Takamuku, J. Phys. Chem. 1996, 100, 8913. 29. M. Fujita, A. Ishida, T. Majima, and S. Takamuku, J. Phys. Chem. 1996, 100, 5382. 30. J. Saltiel, J. D’Agostino, E. D. Megarity, L. Metts, K. R. Neuberger, M. Wrighton, and O. C. Zafiriou, Organic Photochemistry, Vol. 3, Marcel Dekker, New York, 1973. 31. F. D. Lewis, A. M. Bedell, R. E. Dykstra, J. E. Elbert, I. R. Gould, and S. Farid, J. Am. Chem. Soc. 1990, 112, 8055. 32. T. Shida and W. H. Hamill, J. Chem. Phys. 1966, 44, 2375. 33. Y. Kuriyama, F. Hashimoto, M. Tsuchiya, H. Sakuragi, and K. Tokumaru, Bull. Chem. Soc. Jpn 1993, 66, 1622. 34. S. Takamuku, S. Komitsu, and S. Toki, Radiat. Phys. Chem. 1989, 34, 553. 35. T. Shida, Electronic Absorption Spectra of Radical Ions, Elsevier, New York, 1988. 36. E. V. Blackburn and C. J. Timmons, Q. Rev. 1969, 23, 482. 37. L. C. T. Shoute and P. Neta, J. Phys. Chem. 1990, 94, 2447. 38. R. S. H. Liu and D. M. Dale, J. Am. Chem. Soc. 1968, 90, 1897. 39. R. S. H. Liu, J. Am. Chem. Soc. 1968, 90, 1899. 40. K. S. Y. Lau, R. O. Campbell, and R. S. H. Liu, Mol. Photochem. 1972, 4, 315. 41. R. S. H. Liu and J. R. Edman, J. Am. Chem. Soc. 1969, 1969, 1492. 42. R. S. H. Liu and R. E. Kellogg, J. Am. Chem. Soc. 1969, 91, 250. 43. J. Saltiel, D. E. Townsend, and A. Sykes, J. Am. Chem. Soc. 1983, 105, 2530. 44. S. Kobayashi, K. Kikuchi, and H. Kokubun, Chem. Phys. Lett. 1976, 42, 494. 45. H. Fukumura, K. Kikuchi, and H. Kokubun, Chem. Phys. Lett. 1982, 92, 29. 46. K. Kikuchi, H. Fukumura, and H. Kokubun, Chem. Phys. Lett. 1986, 123, 226. 47. W. G. McGimpsey and J. C. Scaiano, J. Am. Chem. Soc. 1989, 111, 335. 48. Z. Wang, S. J. Weininger, and W. G. McGimpsey, J. Phys. Chem. 1993, 97, 374. 49. X. Cai, M. Hara, K. Kawai, S. Tojo, and T. Majima, Chem. Commun. 2003, 222. 50. S. L. Murov, I. Carmichael, and G. L. Hug, Handbook of Photochemistry, 2nd ed., Marcel Dekker, New York, 1993. 51. A. Terenin and V. Ermolaev, Trans. Faraday Soc. 1956, 52, 1042. 52. N. J. Turro, Modern Molecular Photochemitry, Benjamin/Cummings, Menlo Park, CA, 1978. 53. R. E. Bu¨hler and M. Ebert, Nature 1967, 214, 1220. 54. K. D. Raner, J. Lusztyk, and K. U. Ingold, J. Phys. Chem. 1989, 93, 564. 55. J. C. Andre, M. Niclause, and W. R. Ware, Chem. Phys. 1978, 28, 371. 56. X. Cai, M. Hara, K. Kawai, S. Tojo, and T. Majima, Chem. Phys. Lett. 2002, 368, 365. 57. X. Cai, M. Hara, K. Kawai, S. Tojo, M. Fujitsuka, and T. Majima, Tetrahedron Lett. 2003, 44, 6117.

106


58. X. Cai, M. Sakamoto, M. Hara, S. Tojo, K. Kawai, M. Endo, M. Fujitsuka, and T. Majima, Phys.Chem. Chem. Phys. 2004, 6, 1735. 59. X. Cai, M. Sakamoto, M. Hara, S. Tojo, K. Kawai, M. Endo, M. Fujitsuka, and T. Majima, J. Phys. Chem. A 2004, 108, 9361. 60. P. Brodard, A. Sarbach, J.-C. Gumy, T. Bally, and E. Vauthey, J. Phys. Chem. A 2001, 105, 6594. 61. M. Z. Zgierski, J. Chem. Phys. 1997, 107, 7685. 62. S. Koshihara and T. Kobayashi, J. Chem. Phys. 1986, 85, 1211. 63. S. Koshihara, T. Kobayashi, and S. Iwashima, Chem. Phys. Lett. 1986, 124, 331. 64. X. Cai, M. Hara, K. Kawai, S. Tojo, and T. Majima, Chem. Phys. Lett. 2003, 371, 68. 65. X. Cai, M. Sakamoto, M. Hara, A. Sugimoto, S. Tojo, K. Kawai, M. Endo, M. Fujitsuka, and T. Majima, Photochem. Photobiol. Sci. 2003, 2, 1209. 66. M. Sakamoto, X. Cai, M. Hara, M. Fujitsuka, and T. Majima, J. Phys. Chem. A 2005, 109, 4657. 67. X. Cai, M. Sakamoto, M. Hara, S. Tojo, M. Fujitsuka, A. Ouchi, and T. Majima, Chem. Commun. 2003, 2604. 68. G. H. Slocum, K. Kaufmann, and G. B. Schuster, J. Am. Chem. Soc. 1981, 103, 4625. 69. L. J. Johnston and J. C. Scaiano, J. Am. Chem. Soc. 1985, 107, 6368. 70. L. J. Johnston, Chem. Rev. 1993, 93, 251. 71. O. Ito, M. M. Alam, Y. Koga, and A. Ouchi, J. Photochem. Photobiol. A 1996, 97, 19. 72. A. Ouchi, Y. Koga, M. M. Alam, and O. Ito, J. Chem. Soc. Perkin Trans. 2 1996, 1705. 73. A. Ouchi, Y. Koga, and W. Adam, J. Am. Chem. Soc. 1997, 119, 592. 74. W. Adam, K. Schneider, and S. Steenken, J. Org. Chem. 1997, 62, 3727. 75. E. F. Hilinski, D. Huppert, D. F. Kelley, S. V. Milton, and P. M. Rentzepis, J. Am. Chem. Soc. 1984, 106, 1951. 76. S. Watanabe, M. Nagano, T. Suzuki, and T. Ichimura, J. Photochem. Photobiol. A Chem. 2000, 137, 125. 77. X. Cai, M. Sakamoto, M. Yamaji, M. Fujitsuka, and T. Majima, J. Phys. Chem. A. 2005, 109, 5989. 78. X. Cai, M. Sakamoto, M. Hara, S. Inomata, M. Yamaji, S. Tojo, K. Kawai, M. Endo, M. Fujitsuka, and T. Majima, Chem. Phys. Lett. 2005, 407, 402. 79. H. Chosrowjan, T. S., T. Okada, S. Takagi, T. Arai, and K. Tokumaru, Chem. Phys. Lett. 1995, 242, 644. 80. D. LeGourrierec, M. Andersson, J. Davidson, E. Mukhtar, L. Sun, and L. Hammerstrom, J. Phys. Chem. A 1999, 103, 557. 81. N. Mataga, H. Chosrowjan, Y. Shibata, N. Yoshida, A. Osuka, T. Kikuzawa, and T. Okada, J. Am. Chem. Soc. 2001, 123, 12422. 82. N. Mataga, H. Chosrowjan, S. Taniguchi, Y. Shibata, N. Yoshida, A. Osuka, T. Kikuzawa, and T. Okada, J. Phys. Chem. A 2002, 106, 12191. 83. N. Mataga, S. Taniguchi, H. Chosrowjan, A. Osuka, and N. Yoshida, Chem. Phys. 2003, 295, 215.

REFERENCES

107

84. P.-A. Muller and E. Vauthey, J. Phys. Chem. A 2001, 105, 5994. 85. M. Sakamoto, X. Cai, M. Hara, M. Fujitsuka, and T. Majima, J. Am. Chem. Soc. 2004, 126, 9709. 86. M. Sakamoto, X. Cai, M. Hara, S. Tojo, M. Fujitsuka, and T. Majima, J. Phys. Chem. A 2004, 108, 10941. 87. M. Yamaji, T. Sekiguchi, M. Hoshino, and H. Shizuka, J. Phys. Chem. 1992, 96, 9353. 88. F. Bosca, N. Canudas, M. L. Marin, and M. A. Miranda, Photochem. Photobiol. 2000, 71, 173. 89. J.-M. Saveant, J. Am. Chem. Soc. 1987, 109, 6788. 90. J. Mohanty, H. Pal, and A. V. Sapre, J. Chem. Phys. 2002, 116, 8006. 91. L. Pause, M. Robert, and J.-M. Saveant, J. Am. Chem. Soc. 2000, 122, 9829. 92. A. Cardinale, A. A. Isse, A. Gennaro, M. Robert, and J.-M. Saveant, J. Am. Chem. Soc. 2002, 125, 13533. 93. C. Costentin, M. Robert, and J.-M. Saveant, J. Am. Chem. Soc. 2003, 125, 10729. 94. G. J. Kavarnos and N. J. Turro, Chem. Rev. 1986, 86, 401. 95. I. V. Beregovaya and L. N. Shchegoleva, Chem. Phys. Lett. 2001, 348, 501. 96. J. S. Jaworski and P. Leszczynski, J. Chem. Soc. Faraday Trans. 1997, 93, 3529. 97. R. A. Marcus and N. Sutin, Biochim. Biophys. Acta 1985, 811, 265. 98. I. R. Gould, D. Ege, J. E. Moser, and S. Farid, J. Am. Chem. Soc. 1990, 112, 4290. 99. D. O. Martire, J. A. Rosso, S. Bertolotti, G. C. Le Roux, A. M. Braun, and M. C. Gonzalez, J. Phys. Chem. A 2001, 105, 5385. 100. J. W. Taylor, G. Ehlker, H.-H. Castensen, L. Ruslen, R. W. Field, and W. H. Green, J. Phys. Chem. A 2004, 108, 7193. 101. J. Cioslowski, G. Liu, and D. Moncrieff, J. Phys. Chem. A 1997, 101, 957. 102. M. Fujitsuka, Y. Oseki, M. Hara, X. Cai, A. Sugimoto, and T. Majima, ChemPhysChem 2004, 5, 1240. 103. Y. Oseki, M. Fujitsuka, M. Hara, X. Cai, A. Sugimoto, and T. Majima, J. Phys. Chem. B 2004, 108, 16727. 104. X. Cai, M. Sakamoto, M. Hara, S. Tojo, K. Kawai, M. Endo, M. Fujitsuka, and T. Majima, J. Phys. Chem. A 2004, 108, 7147. 105. R. S. Becker, J. S. de Melo, A. L. Macanita, and F. Elisei, J. Phys. Chem. 1996, 100, 18683. 106. V. Wintgens, P. Valat, and F. Garnier, J. Phys. Chem. 1994, 98, 228. 107. D. V. Lap, D. Grebner, and S. Rentsch, J. Phys. Chem. A 1997, 101, 107. 108. D. Fichou, G. Horowitz, B. Xu, and F. Garnier, Synth. Met. 1990, 39, 243. 109. S. Rentsch, J. P. Yang, W. Paa, E. Birckner, J. Schiedt, and R. Weinkauf, Phys. Chem. Chem. Phys. 1999, 1, 1707. 110. X. Cai, M. Sakamoto, M. Fujitsuka, and T. Majima, Chem. Eur. J. 2005, 11, 6471. 111. M. Sakamoto, X. Cai, M. Hara, S. Tojo, M. Fujitsuka, and T. Majima, J. Phys. Chem. A 2004, 108, 8147.

108


112. V. Nagarajan and R. W. Fessenden, Chem. Phys. Lett. 1984, 103, 497. 113. L. J. Johnston, D. J. Lougnot, V. Wintgens, and J. C. Scaiano, J. Am. Chem. Soc. 1988, 110, 518. 114. R. W. Redmond, J. C. Scaiano, and L. J. Johnston, J. Am. Chem. Soc. 1992, 114, 9768. 115. H. Hiratsuka, T. Yamazaki, Y. Maekawa, T. Hikida, and Y. Mori, J. Phys. Chem. 1986, 90, 774. 116. M. Sakamoto, X. Cai, M. Fujitsuka, and T. Majima, Chem. Eur. J. 2006, 12, 1610. 117. M. Sakamoto, X. Cai, M. Fujitsuka, and T. Majima, J. Phys. Chem. A. 2005, 6830. 118. M. Sakamoto, X. Cai, M. Hara, M. Fujitsuka, and T. Majima, J. Phys. Chem. A. 2005, 109, 2452. 119. M. Sakamoto, X. Cai, M. Hara, S. Tojo, M. Fujitsuka, and T. Majima, J. Am. Chem. Soc. 2005, 127, 3702. 120. M. Sakamoto, T. Tachikawa, M. Fujitsuka, and T. Majima, Chem. Phys. Lett. 2006, 420, 90. 121. N. Ichinose, T. Tanaka, S. Kawanishi, and T. Majima, Chem. Phys. Lett. 2000, 326, 293. 122. D. E. J. G. J. Dolmans, D. Fukumura, and R. K. Jain, Nat. Rev. Cancer 2003, 3, 380. 123. F. D. Lewis, J. Liu, X. Zuo, R. T. Hayes, and M. R. Wasielewski, J. Am. Chem. Soc. 2003, 125, 4850. 124. F. D. Lewis, X. Liu, J. Liu, S. E. Miller, R. T. Hayes, and M. R. Wasielewski, Nature 2000, 406, 51. 125. B. Giese and M. Spichty, ChemPhysChem 2000, 1, 195. 126. K. Kawai, T. Takada, S. Tojo, N. Ichinose, and T. Majima, J. Am. Chem. Soc. 2001, 123, 12688. 127. M. Goez, V. Zubarev, and G. Eckert, J. Am. Chem. Soc. 1998, 120, 5347. 128. J. E. Rogers and L. A. Kelly, J. Am. Chem. Soc. 1999, 121, 3854. 129. J. E. Rogers, S. J. Weiss, and L. A. Kelly, J. Am. Chem. Soc. 2000, 122, 427. 130. F. D. Lewis, R. S. Kalgutkar, Y. S. Wu, X. Y. Liu, J. Liu, R. T. Hayes, S. E. Miller, and M. R. Wasielewski, J. Am. Chem. Soc. 2000, 122, 12346. 131. K. Kawai, Y. Wata, M. Hara, S. Tojo, and T. Majima, J. Am. Chem. Soc. 2002, 124, 3586. 132. K. Kawai, Y. Wata, N. Ichinose, and T. Majima, Angew. Chem. Int. Ed. 2000, 39, 4327. 133. K. Kawai, T. Takada, S. Tojo, and T. Majima, J. Am. Chem. Soc. 2003, 125, 6842. 134. K. Kawai, T. Takada, T. Nagai, X. Cai, A. Sugimoto, M. Fujitsuka, and T. Majima, J. Am. Chem. Soc. 2003, 125, 16198. 135. T. Takada, K. Kawai, X. Cai, A. Sugimoto, M. Fujitsuka, and T. Majima, J. Am. Chem. Soc. 2004, 126, 1125. 136. P. G. Sammes, Tetrahedron 1976, 32, 405. 137. W. Oppolzer, Angew. Chem. Int. Ed. Engl. 1977, 16, 10. 138. W. S. Trahanovsky and J. R. Macias, J. Am. Chem. Soc. 1986, 108, 6820.

REFERENCES

109

139. Y. Takahashi, K. Miyamoto, K. Sakai, H. Ikeda, T. Miyashi, T. Ito, and K. Tabohashi, Tetrahedron Lett. 1996, 37, 5547. 140. M. P. Cava and A. A. Deana, J. Am. Chem. Soc. 1959, 81, 4266. 141. L. A. Errede, J. Am. Chem. Soc. 1961, 83, 949. 142. A. Ouchi and Y. Koga, J. Org. Chem. 1998, 63, 6780. 143. A. Ouchi, Z. Li, M. Sakuragi, and T. Majima, J. Am. Chem. Soc. 2003, 125, 1104. 144. M. Yoon, J. A. Chang, Y. Kim, J. R. Choi, K. Kim, and S. J. Lee, J. Phys. Chem. B 2001, 105, 2539. 145. T. Tachikawa, S. Tojo, M. Fujitsuka, and T. Majima, Chem. Eur. J. 2006, 12, 3124. 146. M. Hara, S. Tojo, and T. Majima, J. Phys. Chem. A 2003, 107, 4778. 147. X. Cai, M. Sakamoto, M. Hara, S. Tojo, A. Ouchi, K. Kawai, M. Endo, M. Fujitsuka, and T. Majima, J. Am. Chem. Soc. 2004, 126, 7432. 148. X. Cai, M. Sakamoto, M. Hara, S. Tojo, A. Ouchi, A. Sugimoto, K. Kawai, M. Endo, M. Fujitsuka, and T. Majima, J. Phys. Chem. A. 2005, 109, 3797. 149. M. C. Biewer, M. S. Platz, M. Roth, and J. Wirtz, J. Am. Chem. Soc. 1991, 113, 8069.

TWO-PHOTON PHYSICAL, ORGANIC, AND POLYMER CHEMISTRY: THEORY, TECHNIQUES, CHROMOPHORE DESIGN, AND APPLICATIONS Bernd Strehmel Kodak Polychrome Graphics GmbH, Research and Development, An der Bahn 80, D-37520 Osterode, Germany Veronika Strehmel University of Potsdam, Applied Polymer Chemistry, Karl-Liebknecht Str. 24/25, D-14476 Potsdam-Golm, Germany

CONTENTS I. Introduction II. Two-Photon Absorption: Theory, Mechanism, and Quantification A. Theoretical Background 1. Relation Between Two-Photon Absorption and Nonlinear Optical Parameters 2. Theoretical Methods for Description of Two-Photon Absorption 3. Symmetry Considerations 4. Bond Length Alternation B. Mechanistic Consideration 1. Two-Level Model Advances in Photochemistry, Volume 29 Edited by Douglas C. Neckers, William S. Jenks, and Thomas Wolff # 2007 John Wiley & Sons, Inc.

111

112

TWO-PHOTON PHYSICAL, ORGANIC, AND POLYMER CHEMISTRY

2. Three-Level Model 3. Vibrational Contributions 4. Surface Plasmons C. Evaluation of Two-Photon Absorbing Materials 1. Experimental Techniques in Nonlinear Absorption 2. Absolute Evaluation of Two-Photon Absorbing Materials 3. Relative Evaluation of Two-Photon Absorbing Materials by Two-Photon Excited Fluorescence D. Pulse Propagation III. Chromophore Design and Optimization of Two-Photon Absorption A. Chromophores with Large p-Systems 1. General p-Structures 2. Conjugated Polymers and Oligomers B. Dipolar Chromophores 1. Neutral Donor-p-Acceptor Compounds 2. Ionic Donor-p-Acceptor Compounds C. Symmetric Chromophores with Large Two-Photon Absorptivities 1. Donor-p-Donor Compounds 2. Acceptor-p-Acceptor Compounds 3. Donor-Acceptor-Donor Compounds D. Octupolar/‘‘Propeller Shaped’’ Chromophores with Large Two-Photon Absorptivity E. Polymers with Nonlinear Absorbing Chromophors IV. Two-Photon Excited Fluorescence V. Organic Reactions Upon Two-Photon Excitation A. Isomerization B. Cycloaddition C. Singlet Oxygen VI. Simultaneous Two-Photon Initiated Polymerization and Crosslinking A. Radical Two-Photon Initiated Polymerization B. Cationic Two-Photon Initiated Polymerization VII. Applications A. Three-Dimensional Micro- and Nanofabrication B. Three-Dimensional Data Storage C. Optical Band-Gap Materials D. Waveguide Materials E. Fluorescence Imaging F. Two-Photon Medical Applications G. Upconverted Lasing H. Optical Power Limiting VIII. Outlook for Two-Photon Photosciences References

INTRODUCTION

113

I. INTRODUCTION The first theoretical prediction of a simultaneous two-photon absorption (TPA) event was reported by Maria Go¨ppert-Mayer in 1931 [1]. The development of high-energy pulsed lasers experimentally allowed confirmation of TPA a few decades later [2] because simultaneous multiphoton excitation requires high peak intensities. Such large photon densities can be achieved by (1) use of a collimating lens and (2) a laser providing large photon flux. At least 1028 photons/ (cm2 s) are needed to access the excited state by simultaneous multiphoton excitation. Since the experimental discovery of TPA, multiphoton excitation has become a popular tool in the photochemical sciences to determine the excitation energy of states with parity forbidden transition [3–54]. Transitions that are parity forbidden by one-photon (OP) excitation can thus become allowed by two-photon (TP) excitation. TP excitation spectroscopy localizes the energetic position of TP excited states, which cannot be observed by OP excitation. These pioneering works confirmed many quantum chemical studies predicting the existence of TP excited states and therefore experimentally completed the pattern of electronic transitions in organic compounds. In general, TP excitation had been mainly limited to academic interest until the end of the 1980s [2–24, 26–45, 47–52, 55–69]. Modern photochemistry based on multiphoton excitation uses femtosecond NIR-lasers, resulting in population of an excited state absorbing either in the UV or visible spectral range as described by an upconversion mechanism [16, 23, 31, 41, 70–73]. However, practical applications require the use of inexpensive laser sources, which are in most cases diode lasers emitting continuous wave (cw) light around 800 nm. The use of such laser sources requires the development of TP absorbing materials with extraordinarily large nonlinear absorptivity. In this chapter we introduce possible ways to accomplish these requirements. Perhaps future research will advance even in TP photosciences if inexpensive and miniaturized pulsed femtosecond (fs) lasers become available for industrial applications. Thus, large TP absorptivity of the chromophore and improved efficient excitation sources may spur interest even in industrial applications that need a high spatial resolution. TP excitation results in a significantly better spatial resolution, compared to OP excitation, because excitation occurs with high accuracy at the focal point, and the efficiency of TPA decreases with increase of the distance from the focal point by the power of 4. This clearly shows an advantage of TP excitation. It is attractive for applications requiring high spatial resolution, which is not accomplished by OP excitation. Thus, a spatial resolution significantly below the diffraction limit of the excitation light can be achieved by TP excitation. This spatial resolution is >120 nm and results in higher storage

114


densities needed for some industrial applications [70, 72–76]. Furthermore, no significant attenuation of the excitation light occurs by its traveling through even thick samples if the linear absorption coefficient vanishes at the excitation wavelength. This allows selective excitation of a small volume in thick samples to store/read information more densely with higher accuracy than possible using OP excitation. Since the early 1990s, development of fs lasers [77] has opened new directions for technologies, which can be based on TPA. These light sources offer significantly higher photon densities in a short pulse, in comparison with traditional high-energy nanosecond (ns) or cw lasers. This allows the excitation power of the fs laser to be maintained at a level where material destruction does not occur, while the high photon density in a short laser pulse does not significantly damage the chromophore. This can be seen as an additional important benefit of using fs excitation. However, high cost and the complicated setup, compared to inexpensive cw laser sources (diode lasers), have limited fs lasers as excitation sources mostly to the academic landscape. Perhaps future laser development will result in less expensive and more compact fs lasers even for industrial uses in TP applications. TP excited materials have been successfully applied for imaging applications [78–112], holography [70, 113–117], data storage [118, 119], optical power limiters [53, 120–130], three-dimensional (3D) submicro- and nanofabrication [74–76, 88, 113, 114, 116, 131–146], optical band gap materials [145, 147–155], and upconverted lasing materials [126, 156–178]. Optical power limiters are important for development of eye protecting materials [127]. This is caused by the rapid development of new laser systems. There is currently a need to protect eyes against any laser frequency that may result in an accident. Thus, materials are needed that are transparent under ambient light conditions but whose transparency rapidly changes in 50 GM. Although the data in Table 3.3 clearly exhibit a gradual increase of d by systematic extension of the conjugated system, the overall data do not exhibit large TPA cross sections. This can be seen, for example, by comparing data of benzene, naphthalene, and anthracene (Table 3.3). Details regarding multiphoton absorption of aromatic hydrocarbons have been reported [5, 8–23, 25, 28–32, 35–39, 43, 45, 49–51, 53, 54, 58, 67–69, 379]. In addition, comparison of data obtained in crystals with data obtained in solution shows 55 times more TPA in the solid state [5]. This can be seen in Figure 3.22. The TPA of the 1A1g ! 1B2u transition, which is forbidden for TP excitation, is vibronically induced by a not totally symmetric vibration of the frequency 1200 400 cm1. Section II.B.3 combines the necessary relationships explaining the increase of d as a result of vibronic coupling. In crystals, one can observe an additional TP state with g-symmetry. This has no counterpart in solution. This state exhibits a charge transfer state formed in the crystal due to the packing. In particular, this charge transfer state was believed to be responsible for increasing the TPA cross section. Furthermore, systematic increase of the TPA cross section was found in the series of trans-diphenylpolyenes 14–17 [29]. Then larger TPA may result from the changes in the electronic nature of the p-moiety from an aromatic to a polyene as in 14–17, resulting finally in an increase of d. This may be explained by BLA, which was discussed as a tuning factor for the TPA of polyenes in Section II.A.4 [131, 132, 233, 283–285, 318, 320–322, 325]. Furthermore, experiments focusing on energetic ordering of the OP and TP excited states indicate a change of the order for the odd (OP) and even (TP) states with increasing chain length of the polyene. While the S1 state is almost the OP excited state in 14 and 15, the TP excited state becomes the lowest excited state in 16 and 17, that is, n 3 [52]. This has been one of the minor examples in which state S1 is directly accessible by TPA if the chromophore exhibits a symmetric pattern. No electronic coupling is necessary between S1 and higher excited states according to the

158


TABLE 3.3

Compilation of TPA Data of Several Types of Conjugated p-Sytemsa

Compound

Matrix

l (nm)

d (GM)

Method Reference

Substituted Benzenes Benzene Toluene o-Xylene m-Xylene p-Xylene Mesitylene Aniline Fluorobenzene Chlorobenzene Bromobenzene Phenol

Neat Vapor Neat Neat Neat Neat Neat Vapor Vapor Vapor Vapor Vapor

532 0.00025 525 0 532 0.0036 532 0.028 532 0.035 532 0.096 532 0.096 588 50 528 0.5 538 5 540 7 550 7 Multiple Phenyl Rings

TPF TPF TPF TPF TPF TPF TPF TPF TPF TPF TPF TPF

[51] [37] [51] [51] [51] [51] [51] [37] [37] [37] [37] [37]

Biphenyl Carbazol Dibenzofuran Dibenzothiophen Fluorene Difluorenyl (2,2)-Paracyclophane

Crystal Crystal Crystal Crystal Crystal Crystal Crystal

Qualitative spectroscopic studies; no quantitative data reported presumably due to low signal

TPF TPF TPF TPF TPF TPF TPF

[8, 13, 22] [13] [13] [13] [14] [14] [17]

PTL TPF TPF TPF TPF TPF TPF TPF TPF TPF TPF TPF TPF TPF TPF TPF TPF

[42] [15] [15] [15] [15] [15] [15] [5] [5] [5] [7] [7] [7] [21] [13] [14] [6] [54]

Condensed Aromatic Rings Naphthalene

Anthracene

Azulene Phenanthrene 2,20 -Binaphthyl Chrysene Pyrene

CCl4 606 0.4 Crystal 635 2 Crystal 613 10 Crystal 493 40 Cyclohexane 635 0.02 Cyclohexane 611 0.11 Cyclohexane 597 0.45 Crystal 696 0.3 Crystal 597 7.6 Crystal 576 370 Cyclohexane 641 0.05 Cyclohexane 602 0.45 Cyclohexane 578 0.65 Spectroscopic studies Crystal Crystal Crystal 0.01 - 0.001 694 0.22

CHROMOPHORE DESIGN AND OPTIMIZATION OF TWO-PHOTON ABSORPTION

TABLE 3.3

159

(Continued )

Compound

l (nm)

Matrix

d (GM)

Method Reference

Heteroaromatics Pyridine Acridine

Neat

Stilbene (14)

Cyclohexane CHCl3 Benzene

514 693 608

12.1 50 14.4

TWM [29] NLA [10] TWM [29]

Benzene

608

43.3

TWM [18, 25, 29]

Benzene

608

61.0

TWM [18, 29]

694 Polyenes

Diphenylbutadiene (15) Diphenylhexatriene (16) Diphenyloctatriene (17) all-trans-Retinol

EPA (77 K)

18

Cyclohexane

704 Distyrylbenzenes 600

0.1 2

TL

[26] [54]

20

TPF

[24]

62

TPF

[49, 380]

a

l, wavelength related to the TPA cross section d; TPF, two-photon excited fluorescence; PTL, photothermal lensing; TWM, three-wave mixing; TL, thermal lensing; NLA, nonlinear absorption; EPA, solvent mixture made of ether:isopentane:ethanol ¼ 5:5:2.

mechanism depicted in Figure 3.2a because TP excitation results directly in population of the first excited singlet state.

n n = 1: 14: n = 2: 15: n = 3: 16: n = 4: 17

p-Bis(o-methylstyryl)benzene (18) was investigated as a further standard for TPA measurements between 537 and 694 nm [49, 380]. It is a phenylenevinylene chromophore exhibiting a TPA cross section of 62 GM at 600 nm [380]. Though this value is slightly increased compared to stilbene (14), the larger p-system of 18 does not have the desired impact on d.

18

160


TWO-PHOTON ENERGY (eV)

ABSORPTION CROSS SECTION (CM4 SEC PHOTON–1)

3.2

3.6

4.0

4.4

4.8

10–48

10–49

10–50

10–51

10–52

10–53

23

27

31

35

39

–1 TWO-PHOTON ENERGY (103 CM )

Figure 3.22. TPA spectra of anthracene in crystal (;&) and solution (); more details can be found in Ref. [5]. (From Ref. [5] with permission of the American Institute of Physics.)

2. Conjugated Polymers and Oligomers Conjugated polymers bearing distinct unsaturated segments have evoked a lot of interest in TP photoscience [27, 33, 381–394]. Table 3.4 summarizes TPA data of some TP materials [27, 33, 296, 381–394]. These materials exhibit interesting p-moieties and the data in Table 3.4 determine if these substructures are responsible for the increase of d. These materials exhibit a significantly larger p-system in comparison with the compounds in Table 3.3 and should therefore have larger nonlinear hyperpolarizability as well [129]. Conjugated polymers have been known to exhibit large nonlinear hyperpolarizability g [71, 129, 388, 395–399]. Therefore, these materials are expected to exhibit a large d as well, because a linear relationship exists between d and g according to Eq. (17). Thus, two-photon spectroscopy has gained importance in determining localization of TP allowed transitions that

161

459 777

394 844

394 790

394 841

3332

34,000 34,000 44,000 44,000

100 fs 100 fs 35 ps

Film Film Film CHCl3

8 ns 80 fs 100 fs 8 ns 80 fs 8 ns 80 fs 100 fs

100 fs 150 fs

cw

2.6 ns

100 fs

1

1

1

12

12

[381] [387] [389]

[381]

[381]

[384] [384] [381] [384] [384] [384] [384] [381]

[381] [386]

[382]

[383]

Pulse Length Reference

Film

TCE TCE Film TCE TCE TCE TCE Film

CHCl3 CHCl3

451 718

CHCl3

TOL

19,000

458 769

387 625

d b lexc l01 l02 M01 M12 Mn (GM) (cm/GW) (nm) (nm) (nm) (D) (D) f (g mol1) Solventb nc

Summary of Nonlinear Absorption Data of Conjugated Oligomers and Polymersa

19 20,000 625 R1 ¼ R2 ¼ hexyl R1 ¼ R2 ¼ ethyl-hexyl 20 72,000 R1 ¼ C6H13, R2 ¼ C6H13; R3 ¼ CH3; R4 ¼ CH3; R5 ¼ p-C10H21-Phenyl; R6 ¼ H 21 19 5 800 11 R1 ¼ C6H13; R2 ¼ CH3; R3 ¼ p-C10H21-Phenyl 22a 10,798 1.1 810 R1 ¼ ethylhexyl, R2 ¼ CH3O 42 796 675 180 800 22b 16,443 1.7 810 R1 ¼ ethylhexyl, R2 ¼ 9-phenylanthryl 49 796 22c 2945 0.3 810 R1 ¼ H, R2 ¼ 9-phenylanthryl 18 796 22d 300 80 800 R1 ¼ H, R2 ¼ H 22e 94–300 25–80 800 R1 ¼ CH3O, R2 ¼ CH3O 22f 150 40 800 R1 ¼ C8H17, R2 ¼ C8H17 22g 165 44 800 R1 ¼ OC8H17, R2 ¼ OC8H17 14 1064

Material

TABLE 3.4

162

(Continued)

525 625 726 726 726

710 812

0.2

420 10,000 55 15 130 229 215 109

532 800

3.5 26,520 70,000

500

444 760 708 18 438 710 0.58 534 803 0.01 332 530 525 625 401 726 5.7 5.8 0.74 4255 397 726 6.2 5.7 0.97 5680 390 726 4.2 4.6 0.5 3310

293 460

435 758 619 1059

Film CHCl3 THF CHCl3 Benzene CHCl3 DMSO CHCl3 TOL 1 TOL 1.6 TOL 1

CHCl3

CHCl3 Crystal

Crystal

Mn d b lexc l01 l02 M01 M12 (GM) (cm/GW) (nm) (nm) (nm) (D) (D) f (g mol1) Solventb nc

ns ns 100 fs 100 fs 100 fs

fs 10 ns

100 fs

7 ns

25 ns

[388] [389] [393] [394] [385] [296, 383] [296] [296] [392] [392] [392]

[390]

[389] [27, 33]

[391]

Pulse Length Reference

a d, TPA cross section; b, nonlinear absorption coefficient; lexc , excitation wavelength for d; l01, wavelength for maximum of OPA; l02 , wavelength for maximum of TP excitation; M01 , transition dipole moment for S0 ! S1 optical transition; M12 , transition dipole moment for the optical S1 ! STP transition; f, fluorescence quantum yield; Mn, average number of molecular weight; nc, effective conjugation length used for quantification of d; Pulse Length, discloses the pulse length of the laser used for excitation. b Solvent, discloses the solvent/matrix used for TP experiment; CHCl3, chloroform; TOL, toluene; TCE, 1,1,2,2-tetrachloroethane; THF, tetrahydrofurane; DMSO, dimethylsulfoxide.

27 (sulfate) 28a (R1 ¼ H; R2 ¼ C6H13) 28b (R1 ¼ H; R2 ¼ C8H17) 32 33 35 36 37 38 39 40

22h R1 ¼ OC8H17 R2 ¼ carbazolyl 23 24 R ¼ -CH2SO3-phenylene-CH3 26

Material

TABLE 3.4


163

are forbidden by OP excitation for conjugated polymers. Various conjugated polymers and oligomers have been investigated regarding their two-photon properties: polyfluorenes (19) [383, 400], ladder-type conjugated poly(phenylenes) (20) [382, 386, 400], which are comparable with poly(indenofluorenes) (21) [381], polyphenylenevinylenes (22) [384, 387–389, 391, 401], polyphenyleneacetylene (23) [389], cumulene-containing polymers (24) [402], polydiacetylenes (25) [33, 403], s-p alternating polymers (26) [390], polyaniline (27) [388], fivemembered heteroaromatic oligomers such as polythiophenes (28) [389, 393], oligothiophene (29), oligopyrrole (30), and oligofurane (31) [305], and rigid polymer consisting of an unsaturated carbon moiety and a heterocyclic group (32) [394]. Polymers 19–32 [27, 33, 296, 381–394] represent some examples used in nonlinear absorption studies. Investigations show that large d can be obtained if the unsaturated chain is stiff. This keeps distortion of the p-system as small as possible, and becomes more clear by comparing the data of the stiff laddertype polymer 20 with the more flexible PPV 22. In other words, the more planar the entire p-system, the larger the amplitude of the TPA (Table 3.4). TPA into the TP allowed excited state of substituted poly(p-phenylene acetylene) 23, poly(p-phenylene vinylene) 22, and poly(thiophene) 28 required more excitation energy (about 0.7–0.8 eV) in comparison with the OPA into the S1. Thus, excitation conditions were comparable with the mechanism depicted in Figure 3.2a. TPA of a methyl-substituted ladder-type poly(p-phenylene) 20 needed the same excitation frequency when using either the two-photon excited fluorescence method or the two-photon excitation pump and probe method [386]. The TP excited state exhibited a much broader spectral feature without resolved vibronic structure. Femtosecond transient absorption into the TP excited state occurs in the time frame of the excitation pulse (140 fs). The TP excited state decayed rapidly into the S1 state, which in this case is almost the OP excited state. State S1 exhibited a fluorescence decay of about 600 fs caused by the large transition dipole moment M01 typical for large conjugated systems with negligible distortion of the p-system. Furthermore, the broad absorption feature observed for the TP excited state was similar to that of electroabsorption. Intensity borrowing of state S1 from the energetically higher TP excited state is achieved by coupling between both states. Coupling can occur by intensity borrowing from higher vibrational modes as shown in Eq. (32). The ratio of the excitation energy for the TP excited state (mAg) with respect to the lowest OP excited state (1Bu) is about 1.25 for 20. This agrees with results found for other conjugated polymers [386]. Figure 3.23 shows the fluorescence spectrum of a film made of the methyl-substituted laddertype poly(para-phenylene) 21 obtained by simultaneous two-photon excitation. A similar spectral pattern was observed for OP excitation, showing that the emission belongs to the same excited state, which is S1. These findings are in agreement with the excitation pattern depicted in Figure 3.2a.

164


R4

R1 R1

R5

R3

R2 n

R3

R2 R6

19 R1

R6 R2

R2

R5

nR

1

R4

20 R3

R1

R1

R1

R1 R2

R2

R1

R3

n

n

n R1

R1

21

R2 22

23

R1

R C C C C n

24

R

25

R1

N

n

N

R

R

Si

Si N

R

N

m Ru2+

N N

R m yn

N N

2 PF6xn

26 R1

R2

R1

R1

N R

n

27 C10H21

Z

S n 28

C10H21 N

S

S

N n

32

n 29: Z = S; 30: Z = NH; 31: Z = O


165

Figure 3.23. Absorption spectrum (solid line) and photoluminescence spectrum obtained after simultaneous TPA (dashed line) of methyl-substituted ladder-type poly(paraphenylene) film at room temperature under vacuum of 12, use of Eq. (49) was suggested to explain the large TPA of 19. dðnc Þ ¼

nm dðnc ¼ 12Þ 12

ð49Þ

This relation contains the TPA cross section for nc ¼ 12 (d(nc ¼ 12)) and the number of monomer units in the conjugated polymer nm. The quantity d(nc ¼ 12) is calculated in Eq. (48) and using this result in Eq. (49) results in a d(nc) of 10,000 GM for 19 with nm of 60. This agrees well with the experimental value of 20,000 GM [383]. These considerations clearly demonstrate the importance of a reliable determination of the effective conjugation length, since steric factors may seriously

177


affect nc. A detailed description of the determination of effective conjugation length of conjugated polymers was shown for the poly(arylenevinylene)s containing a biphenyl moiety (38–40), which are called oligomers [392]. The Strickler– Berg relationship, Eq. (50), was applied for evaluation of effective values (nc) contributing to OPA [338]. Thus, Eq. (50) was used to calculate the effective extinction coefficient (eeff max ) at the maximum for the optical S0 !S1 transition, resulting in the rate constant for fluorescence, kf. The latter can be determined by taking both fluorescence quantum yield (f) and fluorescence decay time (tf). ð 9 2 2 eff kf ¼ f =tf ¼ 2:88 10 n nmax emax erel ðnÞ dn ð50Þ where kf ¼ rate constant for fluorescence, n ¼ refractive index of the solvent, eeff max ¼ effective molar extinction coefficient at the maximum for the S0 !S1 optical transition, nmax ¼ absorption energy at the maximum for the S0 !S1 optical transition Ðin cm1, f ¼ fluorescence quantum yield, tf ¼ fluorescence decay time, and erel ðnÞdn ¼ integral of the normalized absorption band with erel ¼ 1. OC6H13 OC6H13

OC6H13 C6H13O

OC6H13 OC6H13

OC6H13

n

38

OC6H13

n

39 OC6H13

OC6H13

n

40

The effective photophysical parameter eeff max is used in a modified Lambert– Beer relationship (Eq. (51)), where m ¼ mass of the sample, ODmax ¼ extinction at the maximum for the optical S0 !S1 transition, V ¼ volume, and d ¼ thickness. M eff ¼

eeff max m d ODmax V

ð51Þ

This equation incorporates the effective molecular weight (M eff ) contributing to OP photonic properties. M eff is often significantly smaller than the number

178


average molecular weight of the sample (M n ). In particular, an increased flexibility of connecting single bonds and thus a lowered planarity are the main factors affecting effective conjugation length [410, 411]. M eff was used to calculate the effective emitter concentration needed for quantification of d. M eff was also taken for determination of the TP quantum yield for irradiation abs [392]. The use of average molecular weights determined by chromatographic or mass spectroscopic methods would result in unreliably large photonic data because of the above-mentioned difference between M n and M eff . Determination of M eff from photophysical measurements permits rough evaluation of nc as shown in Eq. (52): nc ¼

Meff Mr

ð52Þ

where Mr corresponds to the molecular weight of one repeating unit. The results obtained for 38–40 from the OP experiments for nc are included in Table 3.4. The transition dipole for S0 ! S1 (M01 ) was calculated by Eq. (53): sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi 9:19 103 kf ð53Þ M01 ¼ 2:88 109 n3max n3 Interestingly, nc was approximately unity for 38 and 40. This was notable and presumably due to twisting about the biphenyl moiety. The slightly larger nc of 39 (nc ¼ 1.6) may be caused by its specific alkoxy substitution pattern compared to the other systems. Similarly short effective conjugation lengths and the influence of alkoxy substitution were also reported for poly(2,20 -bipyridine-5,50 -diylethynylene[2,5di(2-ethylhexyl)oxy-1,4-phenylene]ethynylene) [410, 411]. Additional evidence for a small effective conjugation length was supported by the UV–Vis spectra of different molecular weight fractions of 39 using semipreparative GPC. Each fraction exhibits essentially the same absorption maximum, indicating no significant dependence of UV–Vis absorption with molecular weight. By comparison, a series of fractionated oligofluorenes (19) showed a significant dependence of photonic properties on molecular weight [404] and data resulted in nc 12. Two-photon excitation spectra obtained for 38–40 are plotted in Figure 3.32 [392]. The maximum is located at approximately the same excitation wavelength (726 nm) for all three compounds. However, there is a significant difference in the amplitudes. Systems 38 and 39 possess similar d of 229 and 215 GM, respectively, while d of 40 approaches a value about half as large. The better torsional mobility of the biphenyl moiety may cause the smaller d of 40 compared to 38 and 39. This agrees with OP results.ÐFurthermore, the fluorescence signal generated by simultaneous TP excitation ( IðlÞdl) is proportional to the square of the light intensity of excitation (I0) for 38–40. The slope of a


179

δ / GM

200 150 100 50 0 750

800

200

850

900

850

900

λ / nm

δ / GM

150 100 50 0 750

800

100

λ / nm

δ / GM

80 60 40 20 0 750

800

850

900

λ / nm

Figure 3.32. TP excitation spectra of 38 (&), 39 (), and 40 (!) obtained in toluene. (Adapted From Ref. [392].)

Ð lnð IðlÞdl) !ln(I0) plot is 2.17 for 38, 1.95 for 39, and 2.16 for 40. This demonstrates the occurrence of a nonlinear absorption process requiring two photons. Simultaneous three-photon excitation does not efficiently occur according to the optical setup chosen [392]. The d values for 38–40 are fairly large (Table 3.4), although the effective conjugation length does not significantly exceed more than one repeating unit. The model based on molecular aggregates [64, 408, 409] was applied to discuss the enhancement of optical transitions. Accordingly, each repetitive unit generates locally excited states resulting in an n-fold band of levels with discrete energy levels for each optical transition (Fig. 3.33). The bandwidth for each excited state depends on the number of coupling monomer units. It can become large,

180 E


S2

M 12 S1 xn

M 01 S0

monomer unit

conjugated sample

θ

θ

Figure 3.33. Band formation from locally excited states of repeat units in 38–40. (Adapted from Ref. [392].)

that is, >1000 cm1. The bandwidth is furthermore proportional to the intensity of the optical transition, explaining the enhancement of transition probability. Thus, the higher the number of coupling monomer units, the broader the bandwidth and therefore the higher the transition dipole moment for the optical transition. This has a particularly strong impact on the transition dipole moments between the S2 (TP excited) state and S1 (OP excited) state, M12 , as well as for the transition dipole moment for the S0 !S1 transition, M01 . Both transition dipole moments strongly affect d as shown by Eq. (27). The angle dependence between monomer units is important to maximize the transition dipole moment Mij between two excited states. Equation (54) shows the relation between the angle y of the coupling monomer unit (Fig. 3.33), the transition dipole of the respective monomer unit Mijm , and the number of coupling units N [64, 409]. In a stiff conjugated sample, the angle y can be small depending on molecular constitution. Therefore, Mij often decreases according to Eq. (54) if y slightly becomes larger than zero. pffiffiffiffi Mij ¼ N ðcos yÞMijm ð54Þ The values obtained for M12 according to Eq. (27) are large (Table 3.4) [392]. They support the model of efficient electronic coupling between the OP and TP excited states despite the fact that 38–40 possess short net conjugation lengths. According to Eq. (27), d depends quadratically on the transition dipole moments M01 and M12 . Thus, putting Eq. (54) into Eq. (27) for each transition dipole moment shows that d / N2 (N ¼ number of coupling units). This demonstrates


181

250

δ / GM

200 150

1.0 0.8 0.6

100

0.4

50

0.2

0

absorption/ a. u.

Excitation: two-photon one-photon

0.0 360

380 400

420 440 λ / nm

460

480

500

Figure 3.34. Absorption spectra obtained for TP (left axis) and OP (right axis) excitation of 38 in toluene at room temperature. (Adapted from Ref. [392].)

that d can become large in conjugated molecules if coupling between monomers occurs despite the fact that the net conjugation length is short. These results eff , which is demonstrate also the importance of a reliable determination of M needed for determination of d. The latter requires knowledge of concentration and therefore molecular weight. Comparison of OP and TP excitation spectra is shown for 38 in Figure 3.34. For this class of oligomers, the TP excited state displays a higher excitation energy in comparison to the lowest OP excited state. The results for 38–40 were essentially the same. Thus, the energetic relations depicted in Figure 3.33 are justified. This has a strong impact on the photochemistry, particularly for 40. Either OP or TP excitation results in the same photochemically active state, S1, and therefore the same photochemical pathways. Most TPA data for conjugated polymers suffer from a poor consideration of the effective conjugation length needed for a reliable determination of d for a single molecule. Therefore, TPA data for conjugated polymers/oligomers needs to be carefully considered if distinct structural patterns are to be compared in order to understand the relation between d and structure. Nevertheless, the macroscopic nonlinear optical parameters of conjugated polymers related to nonlinear absorption, such as the nonlinear refractive index (n2) and third-order nonlinear susceptibility (w3), are available in the literature and are important for applications in optoelectronic and photonic devices [35, 71, 129, 398]. The nonlinear refractive index is proportional to the real part of the third-order nonlinear optical susceptibility (w(3)), and the TPA coefficient b as well as the TPA cross section d are proportional to the imaginary part of the third-order nonlinear optical susceptibility (w(3)) (Eq. (15)). Large w(3) values were reported for conjugated polymers/oligomers if a large conjugation length exists for the molecule.

182


Sometimes it was claimed that w(3) changed as a function of the energy gap between the HOMO (highest occupied molecular orbital) and the LUMO (lowest unoccupied molecular orbital). However, the data summarized in Table 3.4 show that the dependencies of the energy gap between the OP and TP excited states and the third-order nonlinear optical susceptibility are rather more complex. Nevertheless, the energy gap between HOMO and LUMO is influenced by different contributions. This can be aromatic resonance, degree of bond length alternation, contributions of the inductive or mesomeric electronic effects of substituents grafted on the conjugated main chain, contributions of interchain interactions discussed in some cases for neat films, and contributions of geometry. The aromatic resonance and the degree of bond length alternation are characteristic of the conjugated backbone of the polymer. The nonlinear optical parameters for the polymer in solution and the polymer in the solid phase can be considerably different because of the difference in local-field factors, density differences between the polymer and the solvent, and different orientation of the polymer molecules in the liquid and the solid state. These aspects need careful consideration in the quantitative evaluation of TPA properties. Other authors give information about the second hyperpolarizability (g) examined, for example, by the z-scan technique [402]. In the case of cumulene-containing polymers, g was intensity independent [402]. The nonlinear refractive indexes show similar values for the conjugated polymers obtained by z-scan or degenerate four-wave mixing. A more detailed discussion is difficult because different conditions were used concerning the wavelength, the repetition rate of the pulses, and the energy of the laser beam. The z-scan measurements on thin polymer films are quite difficult because of the relatively high light intensities necessary to obtain a measurable signal on a short propagation path.

B. Dipolar Chromophores 1. Neutral Donor-p-Acceptor Compounds Both D-A substitution of unsaturated compounds and large M01 result in an increase of TPA according to Eq. (24) because d / (m01M01)2. Thus, substitution of a polarizable p-system with appropriate donor and acceptor groups results in materials with intramolecular charge transfer (ICT) showing solvatochromism in distinct dielectric surroundings if the ICT formed radiatively deactivates [224]. Solvatochromism can indicate a large TPA in the case of dipolar chromophores as long as M01 is large [219, 224, 234, 337, 412–414]. The strength of ICT formation is tuned by the oxidation potential for the donor (Eox ) and the reduction potential (Ered ) for the acceptor group. The lower Eox and the higher Ered, the more efficient is ICT formation. TPA data of the neutral D-p-A chromophores 41–52 are compiled in Table 3.5 [88, 90, 224, 230, 234, 237, 415–431].

183

n ¼ 1, R1 ¼ NO2, R2-R5 ¼ H, R6-R7 ¼ CH3

n ¼ 2, R1 ¼ NO2, R2-R5 ¼ H, R6-R7 ¼ CH3 n ¼ 1, R1 ¼ SO2(CH2)6OH, R2-R5 ¼ H, R6-R7 ¼ Ph R1 ¼ NO2, R2 =H, R3 ¼ H

R1 ¼ NO2, R2 ¼ H, R3 ¼ CH2CH2OH, R4 ¼ C2H5

R1 ¼ NO2, R2 ¼ Cl, R3 ¼ CH2CH2OH, R4 ¼ C2H5 R1 ¼ NO2, R2 ¼ H, R3 ¼ R4 ¼ CH3 R1 ¼ NO2, R2 ¼ H, R3 ¼ R4 ¼ H

41g

41h 41i

42b

42c

42d 42e

42a

n ¼ 1, R1-R5 ¼ F, R6-R7 ¼ CH3 n ¼ 2, R1-R5 ¼ F, R6-R7 ¼ CH3 n ¼ 3, R1-R5 ¼ F, R6-R7 ¼ CH3 n ¼ 1, R1 ¼ CN, R2-R5 ¼ H, R6-R7 ¼ CH3 n ¼ 2, R1 ¼ CN, R2-R5 ¼ H, R6-R7 ¼ CH3 n ¼ 1, R1 ¼ NO2, R2-R5 ¼ H, R6-R7 ¼ C4H9

41a 41b 41c 41d 41e 41f

Substituents

15.5

17

38,000

33,100 26,800

19 16

19 20 12

41,000 28,000

26,900 44,000

6.1 6.4 7.6

32,500 45,100 52,100

1

750 825 850

180 1010 410 750 178 973 118 944

501 478

535

490

750 850

185 102

485 512

800

428 452

370 396 412 386

50 980 100 750 110 1000

744

895 920 88 1020 191 904

170

120 300 500

e m01 lTP lOP max 1 (M cm ) (D) d(GM) (nm) (nm)

0.1

0.02 0.03 0.03 0.06 0.14

f

DMSO DMSO DMSO DMSO

DMSO THF

DMSO DMSO DMSO

DMSO THF CHCl3

THF THF THF

TOL TOL TOL

fs fs fs

z-Scan, z-Scan, z-Scan, z-Scan,

NLT NLT NLT NLT

z-Scan, NLT Kerr ellipsometry

z-Scan, NLT z-Scan, NLT z-Scan, NLT

TPEF

z-Scan

fs fs fs fs

fs

fs fs fs

ns

fs

[430] [430] [431] [431]

[430] [424]

[430] [430] [430]

[224] [224] [224] [429] [422] [424] [420] [237] [431] [422] [90]

Excitation Sourcec Reference

Kerr ellipsometry ps z-Scan, NLT fs

TPEF TPEF TPEF

Solventa Methodb

TABLE 3.5 Compilation of Extinction Coefficient (e), Change of State Dipole Moment (l01 ), Maximum for Two-Photon Excitation (kTP), Two-Photon Absorption Cross Section (d) at kTP, Maximum for One-Photon Absorption (kOP max ), Fluorescence Quantum Yield (f), and the Light Source Used for Determination of TPA Data

184

R1 ¼ C2H5, R2 ¼ OP(OC2H5)2 R1 ¼ C2H5, R2 ¼ NO2 R1 ¼ C2H5 R1 ¼ C6H13 R1 ¼ C10H21

24,500

R1 ¼ C10H21; n ¼ 0

R1 ¼ C10H21; n ¼ 1

48b

13,300

48a

R1 ¼ C10H21

R1 ¼ C2H5 R1 ¼ C2H5

40,000

29

800 800 800 800 615

605 670 800 800 800 796

6710 800 8450 800 104 800 7940 800 131 800 1163 1064 (880)d 830 850 210 780 250 990 370 780 350 990

3920 2270 9770 30 820

650 1300 9700 10,600 11,600 22

0.9

0.9

f

492 0.45

370 386 389 383 345 500 408 0.39 411 0.04 497 0.46

378

403 389 392

383 414 388 389 390

lTP lOP e m01 max (M1cm1) (D) d(GM) (nm) (nm)

15,000 36,700 50,000 10,200

44 45 46 47

43f 43g

43c R1 ¼ C2H5 43d R1 ¼ C2H5 43e1 R1 ¼ C2H5 R1 ¼ C10H21 43e2

43a1 43a2 43b1 43b2 43b3

(Continued)

Substituents

TABLE 3.5

THF Hexane THF THF THF THF THF THF TOL DMF TOL TOL TOL TOL

CH3CN CH3CN THF THF THF THF Hexane THF THF THF

TPEF TPEF TPEF TPEF

NLT NLT NLT NLT NLT NLT

WLC

NLT NLT NLT

WLC WLC NLT NLT NLT NLT

Solventa Methodb

ns ns ns ns

ns ns ns fs fs fs ns ns fs ns ns ns

fs fs ns ns ns fs

[416] [416] [415] [415] [234, 415] [418] [417] [230] [230] [230] [418] [88, 419] [417] [230] [230] [421] [234, 415] [415] [423] [426] [426] [425] [425] [425] [425]


185

R1 ¼ CH3, R2 ¼ Mesityl R1 ¼ C2H5, R2 ¼ Mesityl R1 ¼ Ph, R2 ¼ Mesityl N(R1)2 ¼ N-carbazolyl, R2 ¼ Mesityl R1 ¼ CH3, R2 ¼ Mesityl R1 ¼ C2H5, R2 ¼ Mesityl R1 ¼ Ph, R2 ¼ Mesityl N(R1)2 ¼ N-carbazolyl, R2 ¼ Mesityl R2 ¼ Mesityl R ¼ C2H5 31,500

188 194 300 212 74 93 119 123 239 240 499

403 414 402 359 431 444 428 397 417 418

0.55 0.60 0.91 0.79 0.35 0.35 0.82 0.84 0.65 0.03

b

TOL, toluene; THF, tetrahydrofurane; DMSO, dimethylsulfoxide; CH3CN, acetonitrile; CHCl3, chloroform. TPEF, two-photon excited fluorescence; WLC, white light continuum; NLT, nonlinear transmission. c fs, femtosecond; ps, picosencond; ns, nanosecond. d From quantum chemical calculations.

a

49a 49b 49c 49d 50a 50b 50c 50d 51 52

THF THF THF THF THF THF THF THF THF CHCl3

TPEF TPEF TPEF TPEF TPEF TPEF TPEF TPEF TPEF TPEF

fs fs fs fs fs fs fs fs fs fs

[427] [427] [427] [427] [427] [427] [427] [427] [427] [428]

186


Charge Separation

Excitation D

A

D

A Δμ01

M 01

Solvatochromism

Figure 3.35. Sketch about the relationship between chromophore size and both transition dipole moment M01 and change of the state dipole moment m01 .

TPA of dipolar chromophores depends on the same parameters tuning OPA. The optical S0 !S1 transition is therefore tunable by the same parameters (M01 and m01 ), showing both OPA and TPA of dipolar chromophores. Size of the psystem and planarity of the conjugated system affect M01 . Planarity can be adjusted by a stiff planar molecular pattern resulting in optimal conditions for the conjugated p-system. Furthermore, m01 depends on the distance of both photon-induced partial charges for donor and acceptor, respectively, and the amplitude of each induced partial charge. The higher the strength of donor and acceptor, the larger the size of the charges generated upon ICT (Fig. 3.35). However, the low fluorescence quantum yield f of many dipolar compounds diminishes the utility of such chromophores for TP applications based on fluorescence imaging [92–112]. f is important to scale the efficiency for TPEF Ztp f . This quantity is called TPA cross section action for fluorescence is a product of the intrinsic TPA cross section d and f (Eq. (55)). [224]. Ztp f Ztp f ¼ f d

ð55Þ

Donor–acceptor substituted polyenes (41) have been well investigated regarding ICT [219] and nonlinear optical properties [429, 432–438]. These materials show solvatochromism upon change of the dielectric surrounding [337]. Data compiled in Table 3.5 demonstrate the relationship between chromophore size, dipole moment change, and TPA. Examples 41a–c, which are equal to 1–3, show the feasibility of increasing d by enhancing the size of the p-system [224]. It appears from the data that the molecular extinction coefficient, which is related to M01 , has a greater impact on d than m01 . The pentafluorobenzene moiety in 41a–c exhibits an interesting electron-deficient moiety having a similar electron deficiency compared to benzonitrile, as concluded from electrochemical data [341, 342, 439]. Introduction of fluorinated phenyl rings in unsaturated p-systems results in solvatochromism if an electron-rich unit is


187

additionally inserted at the chromophore [219, 224, 412, 413, 440]. This was first reported in 1999 [412]. Furthermore, the chemical reactivity of fluorine is small in comparison with many other electron-withdrawing substituents having nucleophilic properties. Therefore, chromophores with fluorinated moieties can be used in the manufacture of fluorescent probes because they possess a better chemical stability in materials having either electrophilic or nucleophilic properties [219, 337, 412, 413]. R3 R6 N R7

R2 R1

n R5

R4

41

Azobenzenes with D-A pattern (42) have been well investigated for NLO purposes [430, 431]. Thus, it is desirable to have TPA data for these chromophores. Examples 42a–c show an increase of d if strength of the donor increases by comparison of 42a and 42b (Table 3.5). Inserting an additional electron-withdrawing substituent at the azobenzene chromophore results in a further increase of d due the increased m01 of 42c. Furthermore, a higher TP excited state was observed for the general azobenzene 42 as shown for the TP excitation data measured at 750 nm. The peak observed at around 1000 nm corresponds to twice the wavelength needed for OPA. This supports the observations made for 41 that S1 is almost accessible by both OP and TP excitation. Moreover, TPA of 42b is smaller in comparison with the corresponding stilbene 41g (Table 3.5). These studies are complementary to 42d–e (Table 3.5) [431]. R4 N R3

N

R1

N R2

42

Fluorenes with D-A substitution (43) are important in TP photosciences as well. Thus, change of strength for the electron-withdrawing group affects TPA if a diphenylamino group is used as the donor. The differences measured for TPA between 43a1 and 43a2 (Table 3.5) are mainly caused by the different strengths of the acceptor group R2. The nitro group, according to Hammett’s constant (sp ¼ 0:78; sm ¼ 0:71½441 ), is a stronger acceptor in comparison to the phosphonate (sp ¼ 0:6; sm ¼ 0:55½441 ). Furthermore, TPA of dipolar compounds can be tuned by the introduction of heterocyclic moieties with electron-withdrawing (A) properties resulting in 43b–f [230, 415, 417–419, 421]. For this purpose, fluorene chromophores bearing a distinct D-p-A substitution pattern were investigated for TPA. Although 43b–f comprise different acceptor units with distinct acceptor strengths, there is no

188


R1

R1

N

R

N

N R1

R:

N N

43a

43b

43c

O 43d

N N

N

S O 43e

43f

43g

clear correlation between d and acceptor strength as concluded from Hammett’s substitution constants [442]. Furthermore, nonlinear transmission was applied to explore TPA of 43b–f. This may cause an extremely large value for the absorption because the excited state populated by TPA absorbs additional intensity during the lifetime of the pulse that is not needed for TPA (compare Section II.D). Thus, one observes a higher value for the absorption of excitation intensity than needed for TPA. This becomes clearer by considering the data measured by using a fs laser. These values may become smaller compared to data measured by a ns laser [418]. Therefore, we recommend comparing only those data obtained by a similar physical method. This was confirmed by an additional study in which results of several methods were compared [443]. Fluorescence of 43b3 (Table 3.5) was investigated in solvents of different solvent polarity [234]. A significant shift to lower emission energy was reported upon increase of the solvent polarity (Fig. 3.36). The energy of the fluorescence decreases by 0.7–0.8 eV upon change of the surrounding from hexane to ethanol. Treatment of the solvatochromic data according to Eq. (25b) results in a straight line. However, authors reported this equation did not work when comparing the larger chromophore 44 bearing the same substituents as 43b3 but exhibiting a significantly larger chromophore due to incorporation of two fluorene units. Not being able to use Eq. (25b) to compare solvatochromic data for 44 and 43b3 can be seen in the Onsager model [336], which may fail for long molecules [444]. This theory requires point dipoles. However, the structure of these large


189

Figure 3.36. Solvent-dependent fluorescence of 43b3 in various solvents. (From Ref. [234] with permission of the American Chemical Society.)

molecules is far away from a point dipole, showing reduced reliability for application of this model.

N N 44

Chromophore 45 bears a thiophene moiety instead of the diphenylamino group. The thiophene moiety is known to have electron-donating properties as well. TP experiments show that this group does not have the desired impact on TPA. Data observed for 45 are significantly smaller in comparison with 43b. Although the latter bears an additional polarizable double bond, we believe that the higher d of 43b is caused by the diphenylamino group. R1

R1

N S

45

190


Chromophore 46 can be considered a nitrobenzene bearing in the para position a substituted pyrazoline moiety. TPA of this chromophore is large and presumably caused by the strong acceptor group and the electron-donating pyrazoline unit resulting finally in a large m01 .

N N

NO2

46

The TP chromophore 47 was investigated from a theoretical point of view [445]. The time-dependent DFT/B3LYP method resulted in the transition dipole moment M01 , excitation energy, dipole moment for the ground state, and dipole moment for the excited state needed for calculation of TPA. Furthermore, by using a finite field approach on the excitation energy, one can get a direct measure of the dipole moment change between the ground and excited state S1 (m01 ). d was calculated from these data and the theoretical solvatochromic studies showed a dependence on the dielectric surrounding. Calculations screened an excess of electron density at the donor. This implies that 47 is able to donate an electron into the surrounding; this is essential to function as a TP photoinitiator. Furthermore, the quantum chemical calculations of optical properties were also confirmed by OPA experiments [426]. Solvatochromic studies in different solvents showed a huge change in m01 (29 D). Nearly full charge separation over the long distance of the chromophore causes this huge change of m01 . The fluorescence quantum yields in even polar solvents are significantly large compared with other dipolar chromophores compiled in Table 3.5. The TPA cross section was equal to 880 GM. The highly delocalized p-system and the separated charges are the main factors initiating TP polymerization by TP excitation. C4H9 N C4H9 N 47

D-A chromophores bearing triple bonds were investigated to understand the relationship between chromophore size and both nonlinear absorption [425] and nonlinear hyperpolarizablity [446]. In particular, 48 [425] comprises a chromophore bearing at one side a dialkylamino group as donor and at the opposite side

191


a cyano group, connected through an alternating anthryl-ynyl p-bridge. TPA data were larger compared to 41a, due to the larger size of the p-system affecting in particular the transition dipole moment M01 . Several TP excited states were reported for 48. One TP excited state is S1, exhibiting similar excitation energy as that needed for OP excitation. A further TP excited state exhibiting a larger d at higher TP excitation energy was reported for this chromophore as well.

R1 N

CN

R1 n

48

Synthesis, structure, and fluorescence properties were disclosed for a series of D-p-A compounds bearing trivalent boron as the electron-deficient group as the acceptor on one side, and distinct donor moieties on the other side of the chromophore [427, 447]. Structure 49 is one possibility. Incorporation of the dimesityl boron unit results in an increase in the strength of the electron-withdrawing part. The acceptor strength of dimesityl boron is located between a cyano group and a nitro group, which was concluded from spectroscopic measurements [448]. These organic boron compounds are stable and exhibit intense fluorescence, large TPEF cross-section action, and huge red-shift of fluorescence upon increase of solvent polarity (Fig. 3.37). The spectral shift of the emission was about 100 nm by switching the surrounding from a nonpolar (toluene) to a polar surrounding (acetonitrile). R2 R1 N R1

B R2 49

As mentioned earlier, the B(mesityl)2 group causes the strong bathochromic shift. Intramolecular charge transfer from donor to acceptor greatly enhances the dipole moment of the excited state. Thus, change of the dipole moment with respect to the ground state is large, as documented by the large slopes of the Lippert–Mataga equation (Eq. (25b)). Both 49a and 49c (Table 3.5) exhibit a slope of about 13,000 cm1 [427], which is again large for a substituted stilbene [337]. Thus, the huge m01 obtained from solvatochromic experiments also explains the large d observed for 49a–d, showing the different strengths of the donating groups according to the relationship in Eq. (24). The Hammett substituent constant sH is a useful parameter to obtain an estimate of the electronwithdrawing ability of an acceptor group. Based on a spectroscopic method, a

192


0.5

580

49a

401

0.4 0.3

538

403

0.2 403

Toluene THF CH3CN

592

0.1

Absorbance or Intensity (a.u.)

0.0

0.5

476 402 404

522

0.4 574

399

0.3

49c Toluene THF CH3CN

0.2 0.1 0.0 300

400

500

600

700

800

λ/nm

Figure 3.37. Linear absorption spectra and fluorescence spectra of 49a and 49c in three solvents. (From Ref. [427] with permission of Wiley-VCH.)

value of sH ¼ 0:65 was spectroscopically determined for a dimesitylboryl group [448]. Absorption spectra obtained in this study show that the dimesityl-boron group has an acceptor strength between the cyano and the nitro group as shown in the following series: Br < CN < B(Mes)2 < NO2. Replacement of the phenyl group by a thiophene unit results in structure 50. These chromophores exhibit less TPA amplitude compared to the similar structures of 49. d of 50 decreases to about half that for comparable compounds 49, as shown by the data in Table 3.5. Thus, incorporation of thiophene in the p-bridge results in less efficient chromophores compared to corresponding compounds bearing stilbene in the main molecular skeleton. These findings agree well with the results disclosed for 45.


193

R2

S R1 N R1

B R2 50

However, d again increases if a second thiophene is placed at the opposite side, resulting in 51. TPA of this chromophore is comparable to that of 49. Thus, the thiophene unit at the left side is an appropriate donor in this series. It functions similarly as the anilino group in 49. R2

S B

R2 S

51

TP chromophore 52 bears a triazine as electron-withdrawing moiety, which is also known as a strong acceptor [449, 450]. This is concluded from electrochemical measurements [451]. d is comparable to the TPA of 41f, 49a, and 49b bearing strong electron-withdrawing groups. R R

N N N

N 52

2. Ionic Donor-p-Acceptor Compounds Compounds with the general structure 53 have been well investigated with OPA. They are well known as chargeshift probes [452–458] in which the charge of the chromophore moves from one side to the opposite, resulting in a change of the dipole moment. This can be seen by a blue shift of fluorescence upon increase of solvent polarity [337], which is called negative solvatochromism [335], in contrast to the neutral fluorophores of the previous section. In other words, the ground state possesses a fairly large dipole moment, while upon excitation the dipole moment of the excited state decreases with respect to the S0 state as a result of charge shift. Photonic data of the TP chromophores 53–56 are compiled in Table 3.6 [122, 156, 160, 161, 172, 173, 176, 229, 370, 459–462]. R2

N R1 N

R3

53

194

Substituents

R1: CH3 ; R2 CH3; R3: C2H4OH

R1: CH3; NR2R3: pyrrolidinyl

R1: CH3; R2 C2H5; R3: C2H4OH R1: CH3; R2 Ph; R3: Ph X ¼ S; R1 ¼ n-butyl, R2 ¼ CH3, R3 ¼ C2H4OH

Compound

53a

53b

53c 53d 54a

43,500d

1

12

3.4

8

10

e m01 (M cm1) (D)

1064 920 1064 1064 930 1064 1064 1064 1064

1064 1064 1064 1064

1300 1010

700 1010 47000 312 930 570 470 880 920

600 730 540 540

d lTP (GM) (nm) f

530

484

482 482

0.01

478 0.009 0.007

474

lOP max (nm)

NLT NLT

NLT NLT NLT NLT

Methodb

DMF

NLT

Epoxy matrix DMF NLT DMF NLT DMF NLT DMF NLT DMF NLT DMF NLT

DMF BzOH HEMA P-HEMA CH3CN EtOH DMF

Solventa

ps

ps ps ns ps ps ps ns ps ps

ps ns ns ns

[460]

[161] [160] [160] [160] [173] [176] [229] [460] [156] [459] [459] [229] [172] [229] [122]


TABLE 3.6 Compilation of Extinction Coefficient (e), Change of State Dipole Moment (l01 ), Two-Photon Absorption Cross Section (d), Maximum for Two-Photon Excitation (kTP), Maximum for One-Photon Excitation (kOP max ), Fluorescence Quantum Yield (f), and the Light Source Used for Determination of TPA Data for a Series of Dipolar Chromophores with the General Structures Taken in Various Solvents Using Different Methods

195

X ¼ S; R1 ¼ CH3, R2 ¼ C2H5, R3 ¼ C2H4OH R1: CH3; R2 C2H5; R3: C2H4OH R1: CH3; R2 Ph; R3: Ph 51,400

11 12

880 1064 1090 1064 119 790

1064

559 0.004

531

DMF DMF DMSO

DMF

NLT NLT NLT

NLT

ps ps fs

ps

[229] [229] [370]

[467]

a DMF, dimethylformamide; BzOH, benzylalkohol; HEMA, hydroxyethylmethacrylate; P-HEMA, poly(hydroxyethylmethacrylate); EtOH, ethanol; DMSO, dimethylsulfoxide; CH3CN, acetonitile. b TPEF, two-photon excited fluorescence; WLC, white light continuum; NLT, nonlinear transmission. c fs, femtosecond; ps, picosecond; ns, nanosecond. d Dependent on gegen ion.

55a 55b 56

54b

196


Charge Shift

R2 N R3

N R1

S0

hν

* R2 N R3

N R1

R2 N R3

N R1

S1

Figure 3.38. Change of bond order upon charge shift for the probe 53.

A general sketch in Figure 3.38 shows the change of bond order upon charge shift. Thus, a single bond becomes a double bond and a double bond returns to a single bond, agreeing well with the model of bond length alternation discussed in Section II.A.4. The structure on the left side is similar to a stilbenoid pattern while the structure of S1 exhibits a chinoid-like pattern. Interestingly, trans–cis photoisomerization of 53 is of minor importance [214, 412, 463]. Thus, these chromophores are useful as fluorescent probes [452–458, 463, 464]. TPA of 53, compiled in Table 3. 6 for a series of different substituents, is larger compared to that of the neutral stilbene 41a (Table 3.5). This may be caused by the mechanism depicted in Figure 3.38, showing the acceptance of charge by an electron-withdrawing group. The pyridinium ring in 53 is a better electron acceptor than pentafluorobenzene 41a according to the higher reduction potential (Ered (pyridinium) ¼ 1.4 V [465]; Ered (pentafluorobenzene) ¼ 2.4 V [342]). Therefore, d of 53a must be significantly larger compared to that of 41a [161–163, 172, 173, 176, 459, 460]. The easier shift of the charge can be seen as one main reason for the higher TPA of 53 in comparison with neutral D-A chromophores. The boron-bearing chromophore 49 contains a strong electron-withdrawing group and the solvatochromic response is even stronger compared to 53. Hence, a large m01 is a prerequisite to obtain large d values. This was concluded from solvatochromic studies in which 53 exhibited a smaller response [337]. Thus, the capability to shift the charge within the molecular skeleton is important to obtain large d values according to Eq. (24). This occurs presumably with higher efficiency in the charged chromophore, resulting in a high twophoton absorbing material, which is receiving increased interest for use as two-photon pumped lasing material [126, 160–163, 165–168, 171–173, 176]. Similar results were also obtained for 54 and 55, which bear the benzothiazol and quinolinium group as unsaturated moiety. Ered of these units is about the same order as that of pyridinium [466]. According to Eq. (24), d is proportional to the square of M01 and m01 , respectively. Chromophores compiled in Table 3.6 exhibit significantly larger m01 values that may explain the larger TPA in comparison with 41a. However, the small fluorescence quantum yield of the ionic chromophores compiled in Table 3.6 diminishes the TPA cross section action for fluorescence (Ztp f ), which

197


is important for imaging applications. The dependence of Ztp f on f was shown by Eq. (55). R1 N R2 N R3

R2 N R3

X

N R1

55

54

Recently, chromophore 56 was the subject of TPA studies (Table 3.6) [370]. The TPA cross section is smaller in comparison to that of 53–55 although d for 56 was not taken at the peak maximum of the TPA spectrum. The maximum for TPA is far away from the data reported [370]. Data obtained are larger compared with 43b, which was extensively investigated in TPA studies. This shows again that dipolar ionic chromophores are more appropriate for TPA applications because of the larger d. The fluorescence quantum yield is also low and comparable with that of 56. H3C

N

CH3

N

N

S N

S

56

C. Symmetric Chromophores with Large Two-Photon Absorptivities 1. Donor-p-Donor Compounds TPA of symmetric donor-p-donor compounds is significantly larger compared to chromophores having a similar size of the p-system with no donor groups at the ends. This becomes more clear by comparison of stilbene (14) with the donor substituted 4,40 -amino-stilbenes 57a and 57b (Table 3.7). Thus, the end capped amino groups result in a better spreading of excitation density from the ends to the chromophore center. This is schematically shown in Figure 3.39. Furthermore, extension of the p-system results in an increase of the transition dipole moment M01 as shown for the polyenes 57b–g (Table 3.7). TPA of all symmetric compounds discussed in this section can be described with the mechanism depicted in Figure 3.2a showing that the TP excited state is energetically higher compared to the S1 state. Thus,

198 65,000 81,000 96,000 103,000 47,000

R1 ¼ C4H9, n ¼ 1

R1 ¼ CH3, n ¼ 2

R1 ¼ CH3, n ¼ 3

R1 ¼ CH3, n ¼ 4

R1 ¼ CH3, n ¼ 5 R1 ¼ CH3, n ¼ 6 N(R1)2 ¼ N-carbazolyl, n ¼ 1 N(R1)2 ¼ N-carbazolyl, n ¼ 3 R1 ¼ C4H9, R2 ¼ H, n ¼ 1

R1 ¼ C6H13, R2 ¼ H, n ¼ 1 R1 ¼ C6H13, R2 ¼ H, n ¼ 3

R1 ¼ C4H9, R2 ¼ CH3O, n ¼ 1 R1 ¼ C4H9, R2 ¼ CH3O, n ¼ 3

57b

57c

57d

57e

57f 57g 57h 57i 58a

58b 58c

58b 58c

67,000 111,000

74,000 60,000 82,800 127,000

48,000 47,000 46,000

R1 ¼ Phenyl, n ¼1

57a

Structure Substitution 75 190 240 200 230 260 340 320 410 425 1300 190 250 950 995 900 870 1450 1230 900 1420

e d (M1cm1) (GM) 720 690 605 600 640 640 695 710 695 730 730 690 640 695 730

Advances in Photochemistry, Volume 29

ADVANCES IN GEOPHYSICS VOLUME 29, Volume 29

Advances in Photochemistry. Volume 15

Advances in Photochemistry (Volume 28)

Advances in Photochemistry, Volume 18

Advances in Photochemistry (Volume 14)

Volume 20, Advances in Photochemistry

ADVANCES IN IMMUNOLOGY VOLUME 29, Volume 29 (v. 29)

Advances in Catalysis, Volume 29

Advances in Agronomy, Volume 29

Advances in Parasitology Volume 29

Advances in Genetics Volume 29

Advances in Computers, Volume 29