PHOTOCHROMISM Molecules and Syste.ms Revised Edition
This Page Intentionally Left Blank
PHOTOCHROMISM Molecules and Systems Revised Edition Edited by
Heinz Diirr Organische Chemie, Uniuersitlit des Saarlandes, Fachrichtung 8.12 OC, 0-66041, Saarbrucken, Germany and
Henri Bouas-Lament Laboratoire de Chimie Organique et Organomdtallique, Groupe de Photochimie, Uniuersitd Bordeaux I , F-33405, Talence Cedex, France
2003
ELSEVIER AMSTERDAM - BOSTON - LONDON - NEW YORK - OXFORD - PARIS SAN DIEGO - SAN FKANCISCO - SINGAPORE - SYDNEY - TOKYO
ELSEVIER B.V. Radarweg 29 P.O. Box 211,lOOOAE Amsterdam,The Netherlands
ELSEVIER Ltd The Boulevard Langford Lane, Kidlington, Oxford OX5 IGB, UK
ELSEVIER Inc. 525 B Street Suite 1900, San Diego CA 92101-4495, USA
ELSEVIER Ltd 84 Theobalds Road London WClX 8RR UK
02003 Elsevier BY All rights reserved. This work is protected under copyright by Elsevier Ltd, and the following terms and conditions apply to its use: Photocopying Single photocopies of single chapters may be made for personal use as allowed by national copyright laws. Permission of the Publisher and payment of a fee is required for all other photocopying, including multiple or systematic copying, copying for advertising or promotional purposes, resale, and all forms of document delivery. Special rates are available for educational institutions that wish to make photocopies for non-profit educational classroom use. Permissions may be sought directly from Elsevier’s Rights Department in Oxford, UK: phone: (+44) 1865 843830, fax: (+44)1865 853333, e-mail:
[email protected]. Requests may also be completed on-line via the Elsevier homepage (http://www.elsevier.com/locate/permissions).
In the USA, users may clear permissions and make payments through the Copyright Clearance Center, Inc., 222 Rosewood Drive, Danvers, MA 01923, USA; phone: ( + I ) (978) 7508400, fax: (+ 1) (978) 7504744, and in the UK through the Copyright Licensing Agency Rapid Clearance Service (CLARCS), 90 Tottenham Court Road, London W I P OLE UK; phone: (+44)20 7631 5555; fax: (+44)20 7631 5500. Other countries may have a local reprographic rights agency for payments. Derivative Works Tables of contents may be reproduced for internal circulation, but permission of the Publisher is required for external resale or distribution of such material. Permission of the Publisher is required for all other derivative works, including compilations and translations. Electronic Storage or Usage Permission of the Publisher is required to store or use electronically any material contained in this work, including any chapter or part of a chapter. Except as outlined above, no part of this work may be reproduced, stored in a retrieval system or transmitted in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, without prior written permission of the Publisher. Address permissions requests to: Elsevier’s Rights Department, at the fax and e-mail addresses noted above. Notice No responsibility is assumed by the Publisher for any injury and/or damage to persons or property as a matter ofproducts liability, negligence or otherwise, or from any use or operation of any methods, products, instructions or ideas contained in the material herein. Because of rapid advances in the medical sciences, in particular, independent verification of diagnoses and drug dosages should be made. First edition 2003 Reprinted 2006 British Library Cataloguing in Publication Data A catalogue record is available from the British Library. Library of Congress Cataloging in Publication Data A catalog record is available from the Library of Congress. ISBN: 0-444-51322-1
1
Working together to grow libraries in developing countries www.elsevier.com 1 www.bookaid.org
I
www,sabre.org
I
@ The paper used in this publication meets the requirements of ANSUNIS0 239.48-1 992 (Permanence of Paper). Printed in Great Britain, by MPG Books Ltd., Bodmin, Cornwall
Preface (1990 Edition) The book “Photochromism - Molecules and Systems” was conceived to cover the field, which has developed enormously in the last decade. Since 1971, when the outstanding book by G Brown “Photochromism” appeared, a considerable amount of theoretical and experimental results in this research area have been published. Details on presentation of new data are given in the chapter “Organization of the Book”. A considerable number of experts in their field have contributed to the book. The literature in the different chapters covers the years up to 1988 and, in some cases, even up to 1989. New developments in the field have been included when they appeared most promising. Our thanks go to all colleagues who were willing to collaborate on this book and who were patient enough to endure the difficult steps in making the final edition possible. We are deeply indebted to Professor Michael Ashworth for dealing with the linguistic problems of the various manuscripts in a gentle way. Our thanks go also to Sabine Kenziora and Hildrun Steuer (Saarbriicken) as well as Sophie Lafont (Bordeaux) who were heavily involved in the typing for or connected with this book. We also appreciate the help of our coworkers in the delicate task of reading: H. Kilburg, R. Johnn, J. Hoffman, V. Port, K. Zwer, G. Cortellaro, and we are grateful to U. Karrenbauer for technical assistance.
The Editors
This Page Intentionally Left Blank
VII
Foreword (2003 Revised Edition) Photochromism is both of great basic interest and of high application potential. It brings together scientists from a variety of domains in chemistry, physics and engineering. It bridges molecular, supramolecular and solid state chemistry, as well as organic, inorganic and physical chemistry. It thus requires a multidisciplinary approach and a broad presentation. The first publication, in 1990, of the book "Photochromism: Molecules and Systems" appeared timely at a period of revival of the field, especially for the organic substrates. It was the second comprehensive treatise, after the Brown volume in 1971, collecting together the contributions of a great number of specialists. It became rapidly out of print. Because of the constant development of the research on photochromism in universities and in industry, a second printing of the Diirr and Bouas-Laurent book will meet the demand. The present version includes updated lists of references to publications and patents, which attest to the intense research activity in the field, as emphasized by the Editors in the preface. It will be an important source of information and inspiration for a wide public. It provides a wide picture, illustrates the state of the art research and indicates perspectives for future development.
Jean-Marie Lehn 6 Juin 2002
This Page Intentionally Left Blank
Preface (2003 Revised Edition) "Photochromism" is simply defined as the light induced reversible change of color. The word is not as popular as photocopy or photograph but the public are aware of ophthalmic lenses, which darken in the sun and bleach in dim light. The field of photochromism developed considerablyover the last century. In 1971, appeared the outstanding first treatise entitled "Photochromism", edited by G . Brown. The second such book "Photochromism: Molecules and Systems" was written up to cover the large gap in the field from 1971 to 1990. The first printing was not meant to be comprehensive but an enormous amount of new concepts and data was offered in a single volume such as the presentation of the main families based on the pericyclic reaction mechanism, the review of new families (e.g. the dihydroindolizines), some bimolecular photocycloadditions (singlet oxygen, polynuclear aromatic hydrocarbons) and some new promising systems (e.g. hole burning, photochromism by orientation), a survey of some biological systems and some potential new applications. That the field is blooming is demonstrated by the fact that fiom more than 9600 references on Photochromism almost 60% appeared in the last decade. The research is driven by the attempts to improve the established materials and to discover new devices for applications. Some of these are directly related to the change in absorption or emission spectra such as variable transmission optical materials applied to ophthalmic lenses or camera filters, optical information storage, cosmetics, novelty items, authentication, fluid flow visualization, optical power-limiting substances. Other applications are based on the photocontrol of other physical properties such as dielectric constant, refiactive index, phase transition, solubility, electric conductivity, ion capture, viscosity, surface wettability, magnetic behavior etc. to be used for optoelectronic or reversible holographic systems, nonlinear optical switches, photoswitchable biologic systems, etc. Some substances displaying also thermochromic or electrochromic properties are more versatile as they can respond to various stimuli. Despite their advantages, organic materials undergo chemical degradation leading to a decrease of performance called "fatigue"; this is the genetic disease of organic photochromic substances. However, some families were found to be more fatigue-resistant and their lifetime can be increased when they are embedded (or chemically grafted) in rigid matrices such as polymers, liquid crystalline materials, silica prepared by sol-gel processes, molecular sieves or if they are used as single crystals. In order to highlight the progress in the field, we have arbitrarily selected two significant examples: 1. ophthalmic lenses; 2. thermally bistable and fatigue resistant systems. 1. Ophthalmic lenses The first photochromic glass (Photogray0 lens) was introduced into the market by "Corning Glass" in 1966. The photochromic substance is made of small aggregates of copper-doped silver halide embedded in a borosilicate glass. The darkening-bleaching cycle can be repeated indefinitely and it was an immediate success. In parallel, the use of plastic (non photochromic) lenses developed to a marked extent at the expense of glass, heavier and breakable. Lens manufacturers asked PPG for a photochromic version of their successful CR39@polymer known for its mechanical strength and optical clarity. But the first photochromic plastic lenses, named PhotoliteO were introduced in 1982 by "American Optical"; however the color was blue and that hindered the development. Ophthalmic lenses exhibiting a true
x gray color (Transitions@plus) were produced in 1992 by "Transition Optical", a joint venture of PPG and Essilor International. The dye was constituted of a mixture of spirooxazines and chromenes. The commercial success came immediately. Further, to overcome the difficulty of color matching of different families, a new "all-chromene dye package" was introduced in 1996, under the name of Transitions@111, accompanied with an improvement of the optical performances (light transmission and darkening) (ref 1). 2. Thermally bistable and fatigue resistant systems Application to ophthalmic lenses implies that the colored form be thermally unstable at ambient temperature and return to the more stable colorless form in the dark. But this property prevents employment for reversible photon-mode recording devices. To that end, thermal stability of both chemical forms is especially required, in addition to fatigue resistance, rapid response etc. A new class of molecules named diarylethenes, exhibiting such characteristics appeared in 1988 (ref 2). The mechanism is based on the well known (4n + 2) electrocylization of stilbene to phenanthrene but the novelty is that the aryl subunits are heterocyclic. The thiophene derivatives linked to hexafluorocyclopentene (ref 3) such A/B represented below show outstanding thermal stability (up to 300 "C) and fatigue resistance (more than 104 cycles without degradation). In this particular case, A/B were found to undergo more than 13.000 cycles, even in aerated hexane, the opening and closing processes occurring in less than 10 ps.
A (colorless)
B (colored)
Technical applications require dispersion of the compounds into solid matrices; the ideal solid matrices are single crystals but photochromic molecules, which show performance in the crystalline phase, are rare (ref. 4). The reversible reaction of dithienylethenes was found to be very efficient from single crystal to single crystal (ref 5). A recent paper (ref 6) reports that a crystalline face (loo), flat and colorless, became blue and wrinkled after irradiation at 366 nm and again flat and colorless following irradiation at more than 450 nm. The single crystal was not destroyed, displaying a remarkable ability to heal in response to an external stimulus. The photoinduced coloration-decoloration cycles of the crystals could be repeated more than 104 times while maintaining the shape of the single crystals and the photogenerated colored states were found to be stable even at 100 "C. An indication of the outgrowth of academic and industrial interest in the field is the
organization of three successful international meetings, named I.S.O.P. (International Symposium on Photochromism). ISOP-93 was held in France, on Embiez Island near Marseilles (Sept 12-16, 1993; Chairman: R. Guglielmetti), ISOP-96 in the USA, in Clearwater, Florida (Sept 8-12, 1996; Chairmen: J. Crano and R. Bertelson), ISOP-99 in Japan, in Fukuoka (Nov 14-18,1999; Chairman: M. Irie).
XI
The publication of other books and important review articles attests also to the worldwide interest in photochromism: "Organic photochromes" by El'tsov (edited by Whittal) in 1990; "Applied Photochromic Polymer Systems" edited by McArdle in 1992. Several chapters of the "Handbook of Organic Photochemistry and Photobiology", edited by Horspool and Pill Soon Song in 1995, were devoted to some important aspects such as those about fulgides, photochromic nitrogen containing compounds, phytochromes, bacteriorhodopsin and rhodopsin. "Organic Photochromic and Thermochromic Compounds" edited by Crano and Guglielmetti in two volumes appeared in 1999. More recently, a thematic issue of Chemical Reviews (Chem. Rev. 2000, 100, No 5, guest editor: M. Irie) was devoted to "Memories and Switches". None of these books and reviews covers all the fundamental aspects or the real and potential applications of photochromism because the field is extremely wide and rapidly developing. These books constitute different pieces of an encyclopedia to be consulted at will.
The great demand for "Photochromism: molecules and systems" encouraged us to publish a second printing. The main chapters were maintained as presented in the chapter "Organization of the book", to give a basic entry into the different subjects thus covering the literature up to 1989. Updated lists of references (1989-2001) have been added to deal with some selected families (ref 7) thus allowing the reader to have immediate access to recent developments. In addition, a short IUPAC review, entitled "Organic Photochemistry, a technical report" was included as a recent survey of definitions and concepts; it constitutes a separate chapter and has its own subject index. It is a nice complement to the "Glossary of terms used in photochemistry", the last chapter of the first printing. The general subject index was given a new format and completed with the updated lists of references, It is our hope that this second printing will be helpful for an efficient entry into this flourishing field.
1. 2. 3. 4. 5. 6. 7.
B. Van Gemert "The Commercialization of Plastic Photochromic lenses: a tribute to John Crano", Mol. Cryst. Liq. Cryst., 2000,344,57-62. M. Irie and M. Mohri, J. Org. Chem., 1988,53,803-808. H. Hanazawa, R. Sumiya, Y. Horikawa and M. Irie, J.C.S. Chem. Commun., 1992,206-207. The performance of some acridizinium salts (Tomlinson and Chandross) should be stressed see this book pp 596-598. M. Irie in "Solid State and Surface Photochemistry", V. Ramamurthy and K.S. Schanze eds., M. Dekker, Inc., New York, 2000, pp 195-225. M. Irie, S. Kobatake, M. Horichi, Science, 2001,291, 1769-1772. In the literature survey, emphasis has been put on the established families, which experienced a large development and some others. In order to maintain the lists within reasonable limits, the publications selected (including the patents) are in English and the abstracts or proceedings were excluded. The updated literature is collected at the end of the book after Chapter 30. The additional pages are given A numbers to clearly distinguish them from the old text. The number of references (2001-1988) (patents (P) inclusive) are given below. Spiropyrans 370 (52P); Chromenes 100 (33P); Naphthopyrans 144 (96P); Spiro(o)xazines 147(29); Fulgides 109 Fulgimides 34, Diarylethenes 134 (2P); Azo compounds 258 (1P) A d s 62 (9P); H-transfer 6, Anthracenes 45 (1P) Triarylmethanes 11; Dihydrospiroindolizines 35 (4 P); Quinones 30 (3P); Viologenes 36; Perimidines 11.
The Editors
This Page Intentionally Left Blank
XI11
See Literature Survey Update (1989 - 2001) Additional Literature Chapter 4 - Azo-compounds Additional Literature Chapter 6 - Dihydro-indolizines
A1 A19
Additional Literature Chapter 7 - Diarylethanes
A23
Additional Literature Chapter 8 - Chromenes
A33
Additional Literature Chapter 8 - Naphthopyrans
A4 1
Additional Literature Chapter 8 - Spiropyrans
A5 1
Additional Literature Chapter 9 - Fulgides
A75 A87
Additional Literature Chapter 9 - Fulgimides
Additional Literature Chapter 10 - Spirooxazines
A89
Additional Literature Chapter 10 - Spiroxazines
A97
Additional Literature Chapter 14 - Anthracenes Additional Literature Chapter 16 - Hydrogen Transfer
A99 A103
Literature Survey on Spiroperimidines- Chapter 16
A105
Additional Literature on Anils - Chapter 17
A107
Literature on Group Transfer Photochromism of Quinones - Chapter 17
a111
Literature Survey for Photochromism based on Electron Transfer of Bipyridinium-salts (Viologenes) Chapter 17 Literature Survey on Photochromism of Triarylmethanes- Chapter 18
A115 A119
This Page Intentionally Left Blank
xv
Table of Contents Preface (1990 Edition) Foreword (2003 Revised Edition) Preface (2003 Revised Edition) See Literature Survey Updates (1989 - 2001) General Defmitions ORGANIC PHOTOCHROMISM A TECHINICALREPORT ON CLASSES AND TERMS USED IN PHOTOCHROMISM by H. Bouas-Laurent and H. Diin List of Contributors
V VII
IX
xi11 XXVII LV
General Introduction Chapter 1 GENERAL INTRODUCTION by H. Diirr 1. Organization of the Book 2. Brief Historical Survey of Photochromism 3. Definition of Photochromism 4. Outlook and Future Developments References
1 3 5
10 14
Photophysical, Photochemical and Photokinetic Properties of Photochromic Systems Chapter 2 PHOTOPHYSICAL,PHOTOCHEMICAL AND PHOTOKINETICPROPERTIES OF PHOTOCHROMICSYSTEMS by G. Gauglitz PhotophysicalProperties 1. 1.1 Interaction between radiation and matter 1.2 Jablonski diagram 1.3 Energy transfer Photochemical properties 2. 2.1 Photochemicalprimary processes 2.2 Bimolecular processes 2.3 Diabatic and adiabatic processes 3. PhotokineticsApplied to Photochromism 3.1 Principles of photokinetic examinations Amount of absorbed light 3.2 3.3 Quantum yields 3.4 Derivation of differential equations 3.5 Absorption measurement 37 3.6 Examination of the mechanism 3.7 Determination of reaction constants 3.8 Competitive thermal backreactiodphotoreactionsand differential equations 3.9 Photochromic systems embedded in polymers 3.10 Combination of UV spectroscopy with other physical methods 4. Concluding Remarks References Addendum Photochromism based on triplet-triplet absorption
15 15 17 19 21 21 22 22 24 24 28 30 32
40 45
53 55 58 58 61
63
Photochromism Based on “E-ZUIsomerhation of Double Bonds Chapter 3 CIS-TRANS ISOMERIZATIONOF C=C DOUBLE BONDS by J. Saltiel and Y. -P. Sun 1. 2. 3. 3.1 3.2 3.3 4. 4.1 4.2 5.
Introduction Generalizations Stilbene The lowest excited singlet potential energy surface Ultrafast laser spectroscopy Medium effects Diphenylpolyenes Diphenylhexatrieneand diphenyloctatetraene Diphenylbutadiene Rotamerism
5.1 trans-l-Phenyl-2-(2-naphthyl)ethene 5.2 trans-1,2-Di(2-naphthyl)ethene 6. Concluding Remarks Acknowledgement References Addendum Photochromismbased on eis-trans isomerizatiodindigo derivatives
64 66 68 68 74 90 114 115 130 143 143 149 153 153 154 163
Chapter 4 Azoc o m o u N D s by H. Rau 1. Introduction 2. Special Features 2.1 Compounds of the azobenzene type 2.2 Compounds of the aminoazobenzenetype 2.3 Pseudo stilbenes 3. Isomerization 3.1 Compounds of the azobenzene type 3.2 Compounds of the aminoazobenzenetype 3.3 Pseudo stilbenes 4. Application of Photocbromism of Azo Compounb 5. summary References
165 166 166 169 170 172 172 183 185 186 187 188
Photochromism Based on Pericyclic Reactions: Electrocyclhation Reactions Chapter 5 4n SYSTEMS BASED ON 1,3-ELECTROCYCLIZATION by C. Schulz and H. Diirr 1. 2. 3. 3.1 3.2 4. 4.1 4.2 4.3
Introduction Stereochemistry in 1,3-ElectroyclicReactions Oxiranes Monocyclic aryloxiranes Bicyclic and tricyclic oxiranes Aziridines Monocyclic aziridines Bicyclic and tricyclic aziridines Photochromic oxalic acid salts of bicyclic aziridines
193 194 195 195 197 199 199 202 204
5. Nitrones 6. Azomethinimines Applications of Photochromic Three-memberedHeterocycles 7. 8. Concluding Remarks References
205 207 208 208 209
Chapter 6 4n+2 SYSTEMS BASED ON 1,s-ELECTROCYCLIZATION by H. Diirr 1. 2. 3. 3.1 3.2 3.3 3.4 3.5 3.6 3.7 4. 4.1 4.2 4.3 5. 5.1 5.2 5.3 5.4 5.5 5.6 6. 6.1 6.2 7. 8. 8.1 8.2 8.3 9.
Introduction Theoretical Studies of 1,5-Electrocyclization Photochromism Based on Pentadienyl Anions with One Heteroatom Type-2-systems Regioselectivity of 1,5-electrocyclization Stereoselectivity Type-3-systems Spectra of the colorless and colored forms Solvent effects on colored forms Conversion rate of colorless forms and coloration efficiency Thermal Reactions in Solution Structure-fadingrate relations Hammett-studies (LFER-linear free energy relationship) of thermal 1,5-electrocyclization Thermodynamicproperties .activationparameters for 1,5-electrocyclization Photoreactions/DeactivationChannels of Type-2-System Luminescenceproperties Multiplicity studies Quantum yield of photocoloration and photobleaching Kinetics of photoreactions Mechanism of photocoloration Mechanism of fatigue/photostability Photochromic Systems Based on Pentadienyl Anions with Two Heteroatoms Type-1,2-systems Type-2,3-systems Photochromic Systems Based on Pentadienyl Anions with Three Heteroatoms Environmental Effects and Application Photochromic systems in the adsorbed state Photochromic Systems in liquid crystallinephase Photochromic systems in polymers Syntheses of Photochromic Molecules Based on the 1,5-Electrocyclic Reaction References
210 213 217 217 22 1 222 226 229 229 230 232 232 234 236 236 238 238 239 240 242 244 245 245 248 253 255 255 255 257 260 266
Chapter 7 4n+2 SYSTEMS: MOLECULES DERIVED FROM Z-HEXA-1,3,5-TRIENE/ CYCLOHEXA- 1,3-DIENE by W.H.Laarhoven 1. Introduction 2. All-Carbon Systems 2.1 Dihydronaphthalenes 2.2 1,2-Diarylethylenes 2.3 Tetraarylethylenes, dianthrylidenes 3. Systems with Heteroatoms 3.1 Hexatrienes with one heteroatom 3.2 Hexatrienes with more than one heteroatom References
270 279 279 282 295 300 301 305 309
Chapter 8 4n+2 SYSTEMS: SPIROPYFWNS by R. Guglielmetti Introduction Historical survey Definition and presentation of the spiropyrans and related compounds Spectrokinetic Properties of Photochromic Interconversions in Solution Spectrokinetic and thermodynamicparameters Open form spectra and thermal fading ratedstructure correlations (solvent effects) Coloration efficiency or “colorability”; quantum yield Structural and Theoretical Studies; Coordination Chemistry Characterization, reactivity and conformational studies by ‘Hand ”C NMR spectroscopy X-ray dimaction structures of spiropyrans and merocyanines Coordination chemistry of the photochromic equilibrium or separately of the spiropyran and the permanent merocyanine Dipole moments of spiropyrans and solvatochromy of soluble model-merocyanines 3.4 3.5 Identification of compounds and study of the fragmentation by mass spectroscopy Characterization and structural aspects by infrared and Raman spectrometry 3.6 Conformational calculations on the open forms by different mechanical and 3.7 quantum methods. Influence of steric hindrance, correlation with thermal kinetics 3.8 Calculation of electric charge distribution and of electronic transitions (compared to experimental values) by quantitative methods 3.9 X-ray photoelectron spectroscopy of benzoxazoline spiropyran and derived permanently stable heterocyclic merocyanines 4. Photochemical and Photophysical Studies of Photochromic Inter-conversion, Photocoloration and Photodegradation 4.1 Nature of the electronic states of spiropyrans and products of their photochromic reactions 4.2 Mechanism of the primary photophysical process of coloration 4.3 The photodegradation Synthetic Routes to Spiropyrans or Derived Compounds 5. General synthetic routes for nitrogen-containingspiropyrans 5.1 5.2 By-products discovered in the spiropyran condensation 5.3 On some reported abnormal reactions during the preparation of azaheterocyclic s p i r o p ~ n s General synthesis methods for non-nitrogen-containingspiropyrans 5.4 Synthesis of ortho hydroxylated aromatic aldehydes 5.5 Synthesisof thio analogs of spiropyrans 5.6 5.7 Synthesis of bifunctional benzothiazolino- and indolino-spiropyrans Concluding Remarks 6. Acknowledgement References
1. 1.1 1.2 2. 2.1 2.2 2.3 3. 3.1 3.2 3.3
314 314 315 323 326 335 355 360 360 365 37 1 375 380 383 389 395 396 398 399 407 414 417 418 43 1 438 438 446 446 448 452 454 455
Chapter 9 4n+2 SYSTEMS: FULGIDES by J. Whittal 1. 1.1 1.2 1.3 1.4 1.5 1.6 2. 2.1 2.2 2.3
Introduction Definition Stereochemistry Mechanism of photochromism in fulgides Photochromism in fulgides Chromophore structure and photochemistry of fulgides Color and constitution of fulgides and their photochromes Phenyl Fulgides Methoxyaryl fulgides Fulgides containing the fluorenylidene group Polyphotochromicfulgides
467 467 468 469 471 47 1 474 475 475 479 481
XIX 3. Fatigue Resistance 3.1 Fatigue resistant photochromic fulgides 3.2 Fury1 hlgides 3.3 Actinometry 4. Steric Effects 4.1 Adamantylidene llgides 5. Concluding Remarks References
48 1 481 483 484 486 4 490 491
Chapter 10 4n+2 SYSTEMS: SPIROOXAZINES byN.Y.C. Chu Introduction Spectral Properties Photochromism in solution 2.2 Absorption spectrum of the colorless form 2.3 Absorption spectrum of the colored form 3. Thermochromism of Spirooxazine Compounds 4. Kinetics of the Thermal Fading 5. Photochemical Properties 6. Synthetic Methods I. Addenda References
1. 2. 2.1
493 495 495 496 498 502 505 505 506 508 508
Chapter 11 4n AND 4n+2 SYSTEMS (102) BASED ON 1,7- AND 1,lO-ELECTROCYCLIZATION by H. Diirr 1. Introduction 2. 4n Systems (02): 1,7-Electrocyclizations 3. 4n+2 Systems (n>2): 1,lO-Electrocyclizations References
510 510 511 513
Photochromism Based on Pericyclic Reactions: Cycloaddition Reactions Chapter 12 CYCLOADDITIONREACTIONS INVOLVING 4n ELECTRONS (2+2) CYCLOADDITION, PHOTOCHEMICALENERGY STORAGE SYSTEMS BASED ON REVERSIBLE VALENCE PHOTO1 SOMERIZATION by G. Jones I1 Introduction (2+2) Addition: Direct and Triplet Sensitized Photorearrangement (2+2) Photoaddition of substituted norbornadienes Features of mechanism for direct and triplet sensitized isomerization (2+2) Addition: Pathways Involving Electron Transfer or Excited Complexes 3.1 Photorearrangementvia radical-ions or exciplexes 3.2 Photorearrangementvia transition metal complexes 4. (2+2) Reversion: Reactivity of Small Ring Radical Cations 4.1 Thermal (catalyzed)cycloreversionof small rings 4.2 Photosensitized cycloreversion via electron transfer or electron donor-acceptor complexes 5. Concluding remarks: Future Directions Acknowledgements References 1. 2. 2.1 2.2 3.
514 518 518 520 522 522 521 528 528 529 532 536 536
Chapter 13 CYCLOADDITIONRE.AC"JS INVOLVING4n ELECTRONS: (2+2) CYCLOADDmOW MOLECULES WITH MULTIPLE BONDS INGORPOUTED IN OR LINKED 10 AROMATIC SYSTEMS by J-P. Desvergne and H. Bow-Lament Introduction 1. Intermolecular (2+2) Photocycloaddition 2. 2.1 Phenanthrenes 2.2 Acenaphthylenes 2.3 Cinnamic acids and related compounds 2.4 Miscellaneous Intramolecular (2+2) Photoaddition 3. 3.1 Bisphenanthrenes 3.2 Bisacenaphthylenes 3.3 Biscinnamates and related compounds Miscellaneous 3.4 Summary and Conclusion 4. Acknowledgements References
539 542 542 544 545 547 55 1 551
553 554 555 557 557 558
Chapter 14 CYCLOADDITION REACTIONS INVOLVING 4n ELECTRONS: (4+4) CYCLOADDITION REACTIONS BETWEEN UNSATURATED CONJUGATEDSYSTEMS by H. Bouas-Laurent and J-P. Desvergne 1.
Introduction Theoretical considerations 561 Overview of the (4+4) cycloaddition reaction 2. Mechanism of Photodimerimtionand Cycloreversion 2.1 Photodimerization of the anthracene ring 2.2 Photodissociation and thermal cycloreversion 3. Intermolecular (4+4) Photocycloaddition 3.1 Anthracene derivatives 3.2 Naphthalene derivatives and other acenes 3.3 Crossed cycloadditions 3.4 Heterocyclic compounds 4. Intramolecular (4+4) Photocycloaddition 4.1 Bichromophores Cyclophanes incorporating two aromatic rings 4.2 5. Summary and Conclusion Acknowledgements References 1.1 1.2
561 562 564 564 571 571 573 580 594 593 598 599 613 62 1 622 622
Chapter 15 CYCLOADDITION REACTIONS INVOLVING4n+2 ELECTRONS. PHOTOCHROMISM BASED ON THE REVERSIBLE REAGTION OF SINGLET OXYGEN WITH AROMATIC COMPOUNDS by H. -D. Brauer and R. Schmidt 1. 2. 3. 4. 4.1 4.2 4.3 4.4 4.5
Introduction Results of Theoretical Investigations on the Formation Reaction and the Chemical Behaviour of Endoperoxides Structural Formulas Experimental Results on the Chemistry of Endoperoxides Formation by photooxygenation Photochemical cycloreversion Photochemical rearrangement Thermal cycloreversion Thermal rearrangement
631 632 635 635 635 637 638 639 639
XXI 4.6 5.
Comparison of quantum yields and thermal yields of cycloreversion and rearrangement Photochromic Systems of High Reversibility Based on the Reversible Photooxygenationof Aromatic Compounds 5.1 Concept for the molecular structure of endoperoxides with small yields of rearrangement Spectral, photochemical and thermal data of bridged endoperoxides 5.2 5.3 Photochromic properties of the new Systems AR + O&RPO 5.4 Applications 6. Conclusions References
640 64 1 642 643 648 651 65 1 653
Photochromism Based on Tautomerism (Hydrogen Transfer) Chapter 16 TAUTOMERISMBY HYDROGEN TRANSFER IN SALICYLATES,TRIAZOLES AND OXAZOLES by H.E.A. Kramer 1. 2. 2.1 2.2 3. 3.1 4. 5. 5.1 5.2 5.3 5.4 5.5
Introduction Fiirster Cycle Proton-induced quenching pK of triplets @KI) Methyl Salicylate Kinetic experiments TunnelEffect 2-(2'-Hydroxyphenyl) benzotriazole and Related Compounds Introduction X-ray crystal structure determination of TIN Absorption and emission spectra Tautomerization equilibria in the ground and excited state Long wavelength fluorescence (A438 nm): quantum yields and decay times 5.6 Tinuvin in polar medium: TIN (inter) References
654 655 658 659 659 662 663 665 665 666 666 670 67 1 676 681
Chapter 17 TAUTOMERISM BY HYDROGEN TRANSFER IN ANILS, ACI-NITRO AND RELATED COMPOUNDS by E. Hadjoudis 1.
Anil tautomerism Anils of Salicylaldehydes 1.2 Heterocyclic anils 1.3 Picosecond flash photolysis 1.4 Effect of crystal structure 2. Aci-Nitro Phototautomerisrn 3. Other H-transfer Phototautomerism 3.1 Metal dithizonates 3.2 Ortho-alkyl aromatic imines 3.3 Ortho-nitrobenzylidene-acylhydrazides Acknowledgement References 1.1
685 685 693 695 699 702 706 706 707 708 710 710
XXII
Photochromism Based of Dissociation Processes Chapter 18 PHOTOCHROMISM BASED ON DISSOCIATIONPROCESSES by R. Aldag
1. Introduction Photochromismbased on Homolytic Dissociation Processes 2. 2.1 Cleavage of C-N bonds: triarylimidazole dimers Cleavage of C-CI bonds 2.2 Cleavage of N-N bonds: nitroso dimers 2.3 2.4 Cleavage of C-S bonds 2.5 Homolysis of C-C bonds Photochromism Based on Heterolytic Dissociation Processes 3. 3.1 Triarylmethanes 3.2 Related Systems References
713 714 714 717 724 725 726 727 727 73 1 734
Photochromism in Biological Systems Chapter 19 PHYTOCHROME by S.E. Braslavsky
1. Introduction 2. Phytochrome The molecule and its function 2.1 2.2 Differences between P, and P,+ 2.3 Kinetics of the P,-P,+- phototransformation The P,++P,-phototransformation 2.4 3. Concluding Remarks Acknowledgements References
738 739 739 741 743 752 752 753 754
Chapter 20
RETINAL PROTEINS by F. Siebert
1. 2.
Introduction The Chromophore:Retinal, Unprotonated and Protonated Retinylidene Schiff Base 3. Visual Pigments 3.1 Rhodopsin 3.2 Other vertebrate visual pigments 3.3 Invertebrate pigments 4. Retinal Proteins of Halobacterium Halobium 4.1 Bacteriorhodopsin 4.2 Halorhodopsin 4.3 Sensory rhodopsins of halobacterium halobium 5. Concluding Remarks References
756 759 762 762 768 769 772 773 780 783 784 785
XXIII
Environmental Effects on Organic Photochromic Systems Chapter 2 1 ENVIRONMENTALEFFECTS ON ORGANIC PHOTOCHROMICSYSTEMS by V.A. Krongauz 1. Introduction 2. Indolinobenzospiropyrans Spiropyrans dissolved in a polymer matrix 2.1 Spiropyran incorporated covalently in a macromolecule 2.2 3. Aromatic Azo Compounds 3.1 Kinetics of cis-trans isomerization 3.2 Photoinduced viscosity change 3.3 Photoinduced CD change 3.4 Photomechanical effect 3.5 Liquid crystal polymers 4. Conclusion References
793 793 793 801 815 815 816 817 817 818 818 820
The Use of Silver Salts for Photochromic Glasses Chapter 22 THE USE OF SILVER SALTS FOR PHOTOCHROMICGLASSES by H.J. Hoffmann 1.Introduction 2. Fundamental Technical Properties of Photochromic Glasses doped with Silver Halides 3. Compositions of Photochromic Glasses 4. Origin of the Induced Absorption Effect 5. Reaction Kinetics of the Induced Absorption Coefficient 6. Thermal and Optical Regeneration 7. Implications and Discussion of the Reaction Kinetics 8. An Atomic Model of the Darkening and Regeneration Kinetics of Photochromic Glasses and Open Questions 9. Summary and Concluding Remarks Acknowledgements References
822 823 826 832 835 842 845 848 852 852 853
Applications Chapter 23 SPIROPYRANSAND RELATED COMPOUNDS by R. Guglielmetti 1. 1.1 1.2 1.3 1.4 1.5 2. 2.1 2.2 213 2.4
Introduction and Classification of Applications Applications depending on sensitivity to W radiation Applications depending upon reversibility Applications depending upon specific color changes Applications depending upon thermal, chemical or physical properties biological applications Applications depending on environmental effects An Autoprocessor Reprography System Using Photochromic Spiropyrans or Merocyanines Stabilization of photomerocyaninesby polymer resins Reprography process by photocoloration of spiropyrans Reprography process by "themophotodegradation" of spiropyrans Themophotodegradatipn of permanent merocyanine dyes
855 855 856 858 858 860 864 864 864
865 868
2.5 Conclusion A potential Application of Spiropyran Derivatives in Thermographic Recording Process 3. 3.1 Principle Results and optimization of the process 3.2 3.3 summary References
871 871 872 874 874 876
Chapter 24 sPIRooxAzINEs by N.Y.C. Chu 1. Light Filters Ophthalmic and sunglass lens applications 1.1 1.2 Other light filter applications Applications Other than Light Filters 2. References
879 879 881 881 881
Chapter 25 ACTINOMETRY by G. Gauglitz 1. Comparison to Physical Measurements 2. Kinetic Principles in Actinometry 2.1 Examples of actinometers at partial absorption Photochromic Systems Embedded in Polymers 3. References
883 885 892 896 901
Chapter 26 PHOTOCHROMIC MATERIALS AND PHOTORESISTS by K. Ichimura 1. 2. 3. 4.
Introduction Photochromic Micelles and Vesicles Photochromic Liquid Crystals Photochromic Polymer Liquid Crystals 5. Two-photon Radical Photoinitiators 6. Dual Mode Photoresists 7. Contrast Enhanced Layer (CEL) 8. Conclusion References
903 904 907 910 912 913 915 917 917
New Developments Highly Promising for Applications Chapter 27 PHOTOCHROMISM BY ORIENTATION by J. Michl 1. 2. 2.1 2.2 2.3 3. 3.1 3.2 4. 4.1
Introduction The Principles The mechanism of light absorption Absorption intensity and polarization Quantitative description of alignment in uniaxial samples Photoinduced Orientation "Destructive" photoorientation 'Won-destructive" photoorientation Examples of Photochromismby Orientation Octaethylporphinein solid solutions
919 920 920 92 1 922 924 924 926 927 927
XXV 928 928 929 929
4.2 FA center in a Na-doped KCI single crystal Potential for erasable optical information storage 4.3 Acknowledgement References
Chapter 28 S PECTRAL HOLE .BURNING by U.P. Wild and A. Renn 930 932 932 936 938 938 939 943 943 943 944 947 947 948 948 948 950 95 1
Introduction Principles of Spectral Hole-burning Homogeneous and inhomogeneousbandwidth Photochemical and photophysical hole-burning Detection of Spectral Holes Transmission and fluorescence detection Holographic detection Other detection methods Properties of Spectral Holes Temperature dependence Electrical field effects Magnetic field effects Pressure effects Optical Information Storage 5. 5.1 Data storage in the frequency and the electric field domain 5.2 Holographic image storage Conclusions 6. References
1. 2. 2.1 2.2 3. 3.1 3.2 3.3 4. 4.1 4.2 4.3 4.4
Chapter 29 BACTERIORHODOPSINAND ITS FUNCTIONALVARIANTS: APPLICATIONSIN MODERN OPTICS by N. Hampp and C. Brauchle 1. Introduction Structure and Function of Bacteriorhodopsin 2. 3. BacteriorhodopsinVariants Optical Applications of BR-Wild Type and BR-326 4. 4.1 BR-WT and BR-326 in dynamic holography 4.2 Non-holographic optical techniques Acknowledgements References
POTENTIAL
954 955 958 960 960 968 972 973
Glossary of Terms Chapter 30 GLOSSARY OF TERMS USED IN PHOTOCHEMISTRY by S.E.Braslavsky and K.N. Houk
976
Appendix of Literature Updates Additional Literature Chapter 4 - Azocompounds Additional Literature Chapter 6 - Dihydro-indolizines Additional Literature Chapter 7 - Diarylethanes
A1 A19
A23
Additional Literature Chapter 8 - Chromenes Additional Literature Chapter 8 - Naphthopyrans
A33 A4 1
Additional Literature Chapter 8 - Spiropyrans Additional Literature Chapter 9 - Fulgides Additional Literature Chapter 9 - Fulgirnides Additional Literature Chapter 10 - Spirooxazines Additional Literature Chapter 10 - Spiroxazines
A5 1 A75 A87 A89 A97
Additional Literature Chapter 14 - Anthracenes
Additional Literature Chapter 16 - Hydrogen Transfer Literature Survey on Spiroperimidines- Chapter 16 Additional Literature on Anils -Chapter 17 Literature on Group Transfer Photochromismof Quinones - Chapter 17
A99 A103 A105 A107 a111
Literature Survey for Photochromismbased on Electron Transfer of Bipyriddinium-salts (Viologenes) Chapter 17 Literature Survey on Photochromismof Triarylmethanes- Chapter 18
Subject Index
A1 15 A1 19
1033
XXVII Pure Appl. Chem., Vol. 73,No. 4,pp. 639-665,2001, 02001 IUPAC INTERNATIONAL UNION OF PURE AND A P P L E D CHEMISTRY ORGANIC CHEMISTRY DIVISION COMMISSION O N PHOTOCHEMISTRY*
ORGANIC PHOTOCHROMISM (IUPAC Technical Report) Prepared for publication by
HENRI BOUAS-LAUREN"' AND HEINZ DURR'
'Luboratoire de Chimie Organique et Organom&allique, Groupe de Photochimie Organique, CNRS UMR5802, Universite'Bordeaux I , F-33405 Talence Cedex, France; 'FR 11.2 Organische Chemie, Universitat des Saarlandes, Postfach 151150, 0-66041Saarbriicken, Germany
*Membership of the Commission during the preparation of this report (1996-99) was as follows:
Chairman: J. R. Bolton (Canada); Secretary: R. G. Weiss (USA): Titular Members: J. R. Bolton (Canada); H. Bouas-Laurent (France); R. G. Weiss (USA); J. Win (Switzerland);Associate Members: A. U. Acuiia (Spain); H. Diirr (Germany); H. Masuhara (Japan); N. Serpone (Canada); National Representatives: S. J. Formosinho (Portugal);P. Hrdlovic (Slovakia); B. S. Martincigh (S.Africa); U. K. Pandit (The Netherlands);B. Pandey (India);
F. C. De Schryver (Belgium); S. C. Shim (Korea); V. G. Toscano (Brazil); C. H, Tung (China); C. Wentrup (Australia); I. Willner (Israel); Working Party for Photochromism: H. Bouas-Laurent; R. Guglielmetti; H. G. Heller; M. Irie; J. C. Micheau; J. L. Pozzo; A. Samat.
Republication or reproduction of this report or its storage and/or dissemination by electronic means is permitted without the need for formal IUPAC permission on condition that an acknowledgment,withfull reference to the source along with use of the copyright symbol 4 the name IUPAC, ond the year of publicotion. are prominently visible. Publication of a translation into another language is subject to the additional condition of prior approval from the relevant IUPAC National Adhering Organization.
XXVIII
Organic photochromism (IUPAC Technical Report) Abstracf: This technical report is a general introduction to organic photochromism. The definition of photochromism (PC) is given together with that of words with the ending “chromism”, such as thermo-, electro-, piezo-, and tribochromism. Important concepts such as two-photon, gated, dual-mode PC and chirochromism are illustrated. The concept of fatigue (chemical degradation) and the determination of the main photochromic parameters (number of cycles, cyclability, half-life), and the spectrokinetic and mechanistic aspects are discussed. The main families of PC (organic compounds and biological receptors) are illustrated with chemical formulae, and the different types of reactions involved in the photochromic processes (pericyclic reactions, En isomerization, group transfer, etc.) are listed. Some examples of applications to “optical power limiting” substances, photoresponsive materials, and photoswitchable biomaterials are considered.
1. INTRODUCTION “Photochromism” is simply defined as a lighf-inducedreversible change of color (a more precise definition is given in Section IU);it has become a common name because many people wear photochromic spectacles that darken in the sun and recover their transparency in diffuse light. The first commercial glasses were made of glass lenses impregnated with inorganic (mainly silver) salts but in recent years, organic photochromic lenses, which are lighter and therefore more comfortable to wear despite their limited lifetime, have made an important breakthrough in the world market. Moreover, the fact that some chemical species can undergo reversible photochemical reactions goes beyond the domain of variable optical transmission and includes a number of reversible physical phenomena such as optical memories and switches, variable electrical current, ion transport through membranes, variable wettability, etc. For this purpose, organic photochromic compounds are often incorporated in polymers, liquid crystalline materials, or other matrices. Given the increasing use of bistable systems in materials sciences, a survey of the basic concepts and language in this developing field would be useful. This article includes the following sections:
II.
Brief historical survey
III. General definitions and concepts IV. V.
Families of organic photochromic compounds Photochromic biological receptors VI. Chemical processes involved in organic photochromism VII. Mechanistic aspects VIII. Applications IX. Acronyms X. Technical vocabulary for ophthalmic lenses XI. Books on photochromism XII. References Xm. Subject index
Q 2001 IUPAC, Pure and Applied Chemistry 73,639-665
XXIX
II. BRIEF HISTORICAL SURVEY 11.1 From the first examples to the name coinage
Fritzsche reported in 1867 [l] the bleaching of an orange-colored solution of tetracene in the daylight and the regeneration of the color in the dark, Later, ter Meer [2] found a change of color of the potassium salt of dinitroethane in the solid state (yellow in the dark; red in the daylight). Another early example was published by Phipson [3], who noted that a painted gate post appeared black all day and white all night (due to a zinc pigment, probably lithopone). In 1899, Markwald studied the reversible change of color of 2,3,4,4-tetrachloronaphthalen-l(4H)-one (P-TCDHN) in the solid state [4]. He believed it was a purely physical phenomenon, naming it “phototropy”. Although used in that period, that term is not proper and should be avoided because it is akin to phototropism, which denotes biological phenomena (vide infra). hv colorless orange 7 heat
tetracene
Interest in photochromism was continuous but limited until the 1940-1960 period, which saw an increase of mechanistic and synthetic studies, particularly in the research groups of Hirshberg and Fischer in Israel. In 1950, Hirshberg [5] suggested the term “photochromism” [from the Greek words: phos (light) and chroma (color)] to describe the phenomenon. This is the name used today. However, it is not limited to colored compounds; it applies to systems absorbing from the far UV to the IR, and to very rapid or very slow reactions.
11.2 DEVELOPMENT OFTHE FIELD
Photochromism expanded during the 1960s in parallel with the development of physical methods (IR,NMR, X-ray, UV, time-resolved and flash spectroscopy) and organic synthesis. Photochromic glasses became available at that period and further stimulated research. Applications, such as the photochromic micro image (PCMI) process, which showed the possibility of reducing the 1245 pages of a Bible to about 6 cm2, attracted considerable interest. An important book was published in 1971 [6]. However, it appeared that the photodegradation of the known families of organic photochromes limited their potential for applications [7]. A revival of activity started in the 1980s,essentially because of the development of fatigue-resistant spirooxazine and chromene derivatives. They triggered the fabrication and commercial application of photochromic ophthalmic lenses. Since then, other commercial systems have been developed, and new photochromic systems have been discovered and explored. In parallel, several books have been printed (vide infra); a flow of articles in scientificjournals has appeared and International Symposia on Photochromism (ISOP) have been organized. Q 2001 IUPAC, Pure and Appried Chemistry 73,639-665
111. GENERAL DEFINITIONS AND CONCEPTS 111.1 Definition of photochromism
"Photochromism is a reversible transformation of a chemical species induced in one or both directions by absorption of electromagnetic radiation between two forms, A and B, having differenta absorption spectra".
The thermodynamically stable form A is transformed by irradiation into form B. The back reaction can occur thermally (Photochromism of type r ) or photochemically (Photochmrnisrnof type P). The most prevalent organic photochromic systems involve unimolecular reactions: the most common photochromic molecules have a colorless or pale yellow form A and a colored form B (e.g., red or blue). This phenomenon is referred to as positive photochromism. Other systems are bimolecular, such as those involving photocycloaddition reactions. When &,=(A) > &,=(B), photochromism is negative or inverse. The unimolecular processes are encountered, for example, with spiropyrans, a family of molecules that has been studied extensively. Solid photochromic spiropyrans or solutions (in ethanol, toluene, ether, ketones, esters, etc.) are colorless or weakly colored. Upon UV irradiation, they become colored. The colored solutions fade rhermally to their original state; in many cases, they can also be decolorized (bleached) by visible light. A few spiropyrans display negative photochromism. They are colored in the dark and bleached by UV light. Many spiropyrans are also thermochromic (see definition below), and spectra of the colored forms are identical to those produced photochemically. 111.2 One-photon and two-photon systems
In general, the photochromic processes involve a one-photon mechanism. B is formed from the singlet ('A*) or triplet (3A*) excited states or both. B, the photoproduct, may also be formed from an upper excited state populated by absorption of two photons.
Two-phofon photochromism The transition probability to populate the final state (hence to obtain the photoproduct) depends on the product of the photon irradiances EMl) and EM*)of the two exciting beams. It is, therefore, advantageousto utilize lasers emitting high photon irradiance. such as those generating picosecond or subpicosecond pulses. Two absorption processes may be distinguished: a) b)
simultaneous absorption of two photons via a virtual level. stepwise (or sequential) two-photon absorption where the second photon absorption takes place from a real level.
'The difference in spectrometric (optical) properties is accompanied by a difference of other physical properties.
0 2001 IUPAC, Pure and Applied Chemistry73,639-665
snt--B XXXI
B'
hv 2
virtual level
i
SO hv'
real level
I
so
--
stepwise two-photon absorption leading to product B'
simultaneous two-photon absorption leading to product B
The simultaneous process (a) has been successfully used for exciting photochromic molecules at specific positions inside a volume for 3D memory systems (writing process). A two-photon absorption process was also used to excite the written molecules that emit fluorescence (reading process) [8,9]. The excitation process can also proceed through a metastable intermediate (process b) as with the dinaphthopyran derivative (1).It was found to isomerize to the bicyclohexene derivative (2) via an intermediate X (not isolated). The authors used two 405-nm photons and observed that the quantum is proportional to the square of the photon irradiance. The reverse reaction 2 + 1was found yield to proceed at 334 nm [lo].
S1
r' -
SO
1
\
A
*L 2
stepwise two-photon photochromic reaction
Q 2001 IUPAC, Pure and Applied Chemistry 73,639-665
111.3 Photochromic compounds
Photochromic compounds are chemical species having photochromic properties. The following terms, sometimes used as synonyms, should be avoided: “Photochromics” is colloquial; “photochromes” are proofs in color photography; “photochromy”is a former technique of color photography. 111.4 Phototropism
Phototropism, observed in plants, refers to a light-induced growth response directed toward or away from the light. 111.5 Chromism
Chromism, as a suffix, means “reversible change of color” and, by extension, a reversible change of other physical properties. The prefix indicates the phenomenon inducing the change [e.g., chromism induced by light (or electromagnetic radiation) is photochromism]. The following are related to photochromism. 111.6 Heliochromism
Heliochromism was coined by Heller [l 11. Heliochromic compounds have a high efficiency for coloring with near UV radiation and a low efficiency for bleaching with visible light, but a moderate efficiency for thermal fading at ambient temperatures. They are activated by unfiltered sunlight and deactivated under diffuse daylight conditions. Therefore, they are suitable for sun lens applications.
111.7 Electrochromism Electrochromism is the reversible change of absorption spectra between two forms, A and B, resulting from electrochemical (oxidatiodreduction) reactions [121. 111.8 Thermochromism
Thermochromism is a thermally induced reversible color change. A large variety of substrates, such as organic, inorganic, organometallic, and macromolecular systems (e.g., polythiophenes) or supramolecular systems (such as liquid crystals) exhibit this phenomenon. Spiroheterocycles (spiropyrans, spirooxazines), Schiff bases, and bianthrones are well-known thermochmmic organic compounds. When the thermochromism of molecular systems results from association with another chemical species such as a metal ion or proton or from modification of the medium by a thermal effect, the phenomenon is called “thermosolvatochromism”[131. 9.9-Bixanthenylidene, a typical thennochromic molecule, is colorless at liquid nitrogen temperature, yellow-green at room temperature and dark-blue when melted or heated in boiling mesitylene tl41.
0 O E 9,9-Bixanthenylidene
/ /
/ / Q 2001 IUPAC, Pure and Applied Chemistry 73,639-665
XXXIII 111.9 Piezochromism
Piezochromism is the phenomenon when crystals undergo a major change of color due to mechanical grinding. The induced color reverts to the original color when the fractured crystals are kept in the dark or dissolved in an organic solvent. The grinding generates a metastable form (see tribochromism). An example of a piezochromic organic compound is diphenylflavylene [151. 111.10Tribochromism
Tribochromism is the phenomenon when crystals undergo a major change of color during mechanical grinding (as is the case of piezochromism), but the induced color change does not revert to the original color when the fractured crystals are kept in the dark or dissolved in an organic solvent. The crystals, prior to fracture, are in a metastable state [16].
H5c6)f4 0
R, R’ = alkyl cycloalkylidene or cycloalkyl
H5C6
R?
R
General formula of some tribochromic compounds
111.11 Solvatochromism
Solvatochromismis the reversible variation of the electronic spectroscopicproperties (absorption,emission) of a chemical species, induced by solvents [17,18].
111.12 Halosolvatochromism Halosolvatochromismis a color change upon increasing ionic strength of the medium without a chemical change of the chromophore.
2,6-diphenyl-4-(2.4,6-triphenyl-l -pyridinio)phenoxide
Examples are solutions of the betaine dye shown above. It undergoes a very large solvatochromic effect, which is the basis of an empirical parameter of solvent polarity, the E,.(30) scale. Furthermore, the addition of salts such as KI, Ca(SCN),, or Mg(C10& to solutions of the betaine dye in acetonitrile also induces a hypsochromic shift of the electronic absorption spectrum that increases with the charge density of the cations. The combination of both properties is termed “halosolvatochromism”[191.
Q 2001
IUPAC,Pure and Applied Chemistry 73, 639-665
XXXN 111.13 Gated photochromism
Gated photochromism is a special type of photochromism in which one or both forms of the photochromic system are transformed (chemically or electrochemically) reversibly into a nonphotochromic form. The photochromic process is controlled like the flow through a gate. The opening or closing of the gate may depend on external stimuli such as protonation, oxido-reduction, solvation, and temperature. Gated photochromism is illustrated with the following example [20]. The conrotatory photocyclition of diarylethene 3 proceeds only from the antiparallel form (3a) in ethanol. The cyclization is completely inhibited in cyclohexane or decalin, where the parallel form (3b) is maintained by H-bond formation. Heating at 100 "C(or adding EtOH) disrupts the intramolecular H-bonds and allows the photochromic reaction to occur between 3a and 3c.
cydohexane HOOC
d
-
ethanol or heat antiparallel open form
b-H---d'
3a
parallelopen form
A
hv
hv' 1
COOH
HOOC closed form (colored)
3c adapted from [20] with permission; copyright 1992 American Chemical Society
111.14 Dual-mode photochromism
Dual-mode photochromism occurs in complex systems triggered alternativelyby two different external stimuli, such as light and an electric current. In such a case, photochromism and electrochromism are mutually regulated. The following substrates have been proposed as suitable for a dual-mode optoelectrical molecular switching device [21]. The device may be reversibly converted among states 4.5, and 6.which are all thermally stable and exhibit very distinct absorption spectra (shown below). (D 2001
IUPAC, Pure and Applied Chemistry73,639-665
Is 11w
adapted from [21], by permission of the Royal Society of Chemistry
Compound 4 may be reversibly transformed to 5 photochemically, 4 being electrically inert and 5 being active and reversibly oxidized to 6 within the -1 to + l V range.
HO
4
4
between - l V and + 1V
+2e
+2H+
A
-2e-2H’
Vis (>600nm)
0
Further, the system can be electrochemically interconverted between the reduced form 5 (which is photochromic) and the quinoid species 6, which is photochemically stable. Therefore, after writing with UV light, the information may be safeguarded by an electrochemical process (5 + 6) and readout at ca 600 nm; after electrochemical reduction (6 + 5),the information may be erased with visible light. A dual-mode photoswitching of luminescence has been described [22]. (See also chirochromism). 111.15 Acidichromism
In acidichromism, the protonated form and the conjugate base of some compounds may have distinctly different absorption spectra. This phenomenon is well known for phenols and aromatic amines. It can occur in addition to photochromism, e.g., for spirooxazines (SO) which generate merocyanines (MC) 1231 as illustrated in the following scheme: Q 2001 IUPAC, Pure and Applied Chemistry 73,639-665
so Photochromism
-
/ oAisl1U.v. \ MC
-
..
H+
OH'
*
1
*
or vis
H+
OH-
*
+
SOH-
UV . ,.
\ Photochromism
MCH+
In some cases, it is possible to take advantage of acidichromism to develop nondestructive readout systems in which one of the forms can be used for readout and the others for writing and erasing ~41.
111.16 lonochromism Ionochromism is a reversible change of color by addition of salts. It can occur in addition to photochromism and may trigger a modulation of conductivity. (See scheme for acidichromism and replace H+by M').
111.17 Halochromism Halochromism,a term introduced by von Baeyer [25], denotes the trivial color change of a dye on addition of acids or bases. It is caused by the formation of a new chromophore, as in acid-base indicators (see acidichromism).
Example :@,C-Cl + AlCl, colorless
2 @,C+ AlC1,yellow
111.18 Chirochromism 111.18.1 Chirochromism Generally, chirochromism is a reversible change of rotation of the plane of polarized light between two chiral diastereomers of a photochromic system (vide infra). It should be noted that the interconversion of two enantiomers of a photochromic compound which have, by definition, identical absorption spectra in nonchiral media, does not pertain to photochromism; however, these enantiomers might exhibit different absorption spectra in chiral media, especially in solid matrices. 111.18.2Diastereoselective photochromism (diastereophotochromism) Diastereoselective photochromism (diastereophotochromism) is a photoinduced reversible change of absorption spectra between two diastereomers A and B. The diastereomeric excess (Ede) at the photostationary state (pss), from irradiation of a mixture of A and B at the wavelength L with nonpolarized light, depends on the molar absorption coefficients (E) and the interconversion quantum yields, @-(A + B) and %A(B + A), according to the following equation [ 261:
Q 2001 IUPAC, Pure and Applied Chemisfty73,639-665
The mixture (A,B) will result in an enrichment of B at Al and of A at 4. If A and B contain chiral subunits, their interconversion will result in different effects on plane polarized light (which can be measured by circular dichroism); this reversible interconversion is referred to as “chirochromism”. Enantiomerically pure, sterically overcrowded heterocycles, 7 (cis) and 8 (trans),exhibit a stereospecific photochemical isomerization of 7 to 8 and 8 to 7 (diastereophotochromism); the two diastereoisomers have been shown to be thermally stable, and no isomerization was detected at ambithe ent temperature in the dark. The reaction is also thermally controlled (diastereothermochri~m); isomerization is accompanied by a simultaneous reversal of helicity and can be followed by circular dichroism. Such a system can be a chiroptical molecular switch [27].
3 ‘’ I
s
I,
\
CH,
hv or heat hvor heat
7 (cis)
8 (trans)
Related chiral photochromic molecules [28] were shown to display different fluorescence emission spectra which are regulated by reversible protonation. It is another case of gared phorochromism (see definition). 111.19 Colorability Colorability is the ability of a colorless or a slightly colored (pale yellow) photochromic material to develop coloration. In dilute solutions, the initial absorbance “Ao(&” immediately after photolysis is proportional (proportionality constant k includes the incident photon flux) to Qc0, (coloration quantum yield), EB ( molar absorption coefficient of the colored form), and cA(concentration of colorless form) at a given irradiation wavelength [29]. A&) = k
(9 2001
Qc0i
EBCA
IUPAC, Pure and Applied Chemistry73,639-665
XXXVIII 111.20 Fatigue Photochromism is a nondestructive process, but side reactions can occur. The loss of performance over time, due to chemical degradation of a material, is termed “fatigue”. Usually, the major cause of damage to photochromic substances is oxidation [30]. The following are examples of particularly fatigue-resistant substances: Single crystals of methyl-substituteddithienylperfluorocyclopentenesare stable at 100 “C and can be cycled more than lo5 times without loss of their shape. Because of physical damage due to surface reconstruction during photoisomerization, transmittance of the single crystals started to decrease after lo4cycles [31]. Bacteriorhodopsin from the purple membrane of “Halobacteriumhalobium” is reported to undergo more than lo5 photochemical cycles without damage [32].
a)
b)
111.21 Number of cycles The number of cycles that a system can undergo under well-defined conditions (solution, matrix, concentration, temperature) is an important experimental parameter. In a cycle, a system A is transformed (photochemically)into system B which returns to A (thermally or photochemically);the terms “switch on” and “switch off are used. Ideally, the yields of the two reactions are quantitative, but byproducts are actually formed. If the degree of degradation in a cycle is x , the nondegraded fraction y after n cycles will be:
y = (1-x)”
For very small x and very large n, this expression can be approximated as: y=l-nx Thus, for x = 0.001 (yield = 99.9%). after Id cycles, 63% of A will be lost, and after lo4cycles virtually no A will be present.
111.22 Cyclability (&) is the number of cycles required to reduce the initial absorbance at a specific wavelength by 50%
WI.
111.23 Half-life (T,n) T I Ris the time necessary for thermal bleaching to half of the absorbance of the colored form at a specific wavelength during one cycle. (See cyclability and thermal fade rate.)
111.24 Readout number
For application to ROM (read only memory) devices, a useful parameter is the number of readings under continuous irradiation. Assuming the reading time to be =lo p,there can be up to 8 x lo9 readings per day. IV. FAMILIES OF ORGANIC PHOTOCHROMIC COMPOUNDS
In all cases, hv2< hv,.
0 2001 IUPAC, Pure and Applied Chemistty 73,639-665
IV.1 Spiropyrans
Open Form (Merocyanine)
Closed Form
(Quinonic form)
IV.2 Spirooxazines
A or hv2
Open Form (Merocyanine)
Closed Form
(Quinonic form)
V.3 Chromenes hv, ___)
c -
A or hv,
p \
0
Open Form
Closed Form
IV.4 Fulgides and fulgimides The name “fulgides” (from the Latin fulgere, to glisten) was given by their first investigator, Stobbe, because they were isolated as fine glittering crystals. Their photochromic properties have been studied extensively by Heller [34]. X = 0 (fulgides) X = NR (fulglmldes)
Open Form
Closed Form
IV.5 Diarylethenes and related compounds
Open Form
0 2001 IUPAC, Pure and Applied Chemistry 73,639-665
Closed Form
IV.6 Spirodihydroindolizines H,CO&
Closed Form
Open Form
IV.7 Azo compounds
trans (anti)
cis (syn)
IV.8 Polycyclic aromatic compounds
IV.9 Anils and related compounds (hydrogen transfer)
c--
A
A
N B Aldehydes react with primary amines to form imines (Schiff bases). When the m i n e is aniline, the imine is known as an “mil”.
0 2001 IUPAC, Pure and Applied Chemistry 73,639-665
XLI IV.10 Polycyclic quinones (periaryloxyquinones)
IV.11 .Perimidinespirocyclohexadienones
IV.13 Triarylmethanes
V. PHOTOCHROMIC BIOLOGICAL RECEPTORS
Many biological systems are photochromic, but few remain so when isolated from the living cell of which they are part [35]. V.l Retinal proteins
Rhodopsin: the chromophore, retinal, is bound to a protein via a lysine through a protonated Schiffbase. The complex photochromic cycle is outlined below:
:
hv
rhodopsin
all-trans-retinal
Bacteriorhodopsin (BR): the retinal is also bound to a lysine fragment of a protein present in the purple membrane (PM) of halobacterium halobium; a simplified model of the BR photocycle involves forms B and M as follows:
0 2001 IUPAC, Pure and Applied Chemistry 73,639-665
XLII
V.2 Phytochrome (P)
Phytochrome (P) controls the photomorphogenesisof plants. +Cyst
fin (321)4-
I=red. absorbing in the red
fr= far-red, absorbing in the far-red structure of the Pr form (660nrn)
VI. CHEMICAL PROCESSES INVOLVED IN ORGANIC PHOTOCHROMISM VI.1 Pericyclic reactions
Electrocyclizations, whether concerted or not concerted, are 6n 6 atom processes for spiropyrans, spirooxazines, chromenes, hexa- 1,3,5-triene, diheteroarylethenes, and cyclohexa-1.3-diene systems, and 6x 5 atom processes for spirodihydroindolizinesand other pyrazoline based systems. Cycloadditionsare found in (2+2) cycloadditionsbased on valence isomerizationsor in molecules with multiple bonds incorporated in or linked to aromatic systems. ( 4 4 ) Cycloadditions are found mainly in polycyclic aromatic hydrocarbons. (4+2) Cycloadditions are found, for example, in additions of singlet oxygen to aromatic compounds. V1.2 Cis-trans (uz)isomeritations
Cis-trans (HZ) isomerizations occur in stilbenes, azo compounds, azines, thioindigoids, etc., as well as some photochromic biological receptors that are part of living systems.
V1.3 Intramolecular hydrogen transfer
Intramolecular hydrogen transfer is found in anils, benzylpyridines, aci-nitro and related compounds, salicylates, triazoles, oxazoles, metal dithizonates, and perimidinespirohexadienones. V1.4 Intramolecular group transfers
Intramolecular group transfers operate in polycyclic quinones (periaryloxyparaquinones).
0 2001 IUPAC, Pure and Applied Chemistry 73,639-665
XLIII V1.5 Dissociation processes
Heterolytic bond cleavages occur in triarylmethanes and related systems. Homolytic bond cleavages are found in triarylimidazole dimers, tetrachloronaphthalenes,perchlorotoluene, nitrosodimers, hydrazines, etc.
V1.6 Electron transfers (oxido-reduction) Electron transfers (oxido-reduction)are photoinduced in viologens and related systems. The same compounds can also undergo electrochromism.
VII. MECHANISTIC ASPECTS VII.1 Kinetics of photochromic compounds The determination of the photochromic parameters, such as the number, nature, and kinetic and spectral properties of the transient species formed under irradiation, is not a trivial task because the photoproducts are too labile to be isolated in many cases. As an illustration, the kinetic behavior of the unimolecular systems is considered (e.g., spiropyrans, spiroxazines, dihydroindolizines, which are of major importance for applications to ophthalmic lenses). It can be accommodated in the following scheme:
It involves A, the noncolored “closed” form (see IV.1,2,6),M,a short-lived transient species (a singlet andor triplet excited state or a very labile photoisomer), and B,the long-lived but not isolable colored “open” form. The analysis can be performed using either pulsed or continuous irradiation methods. a) b)
Pulsed irradiation: Immediately after a light pulse, only M has accumulated. Continuous irradiation: Using a low-power photon flux, a product such as B (often a photoisomer) can accumulate over time scales of 1 9 - 1 0 + ~ s.
In order to estimate some relevant parameters related to photoisomer B (quantum yields as well as UVhisible spectra), specially designed kinetic experiments must be carried out. Under continuous monochromatic irradiation, a photochromicsystem can be considered to be at nonequilibrium and open. The evolution of the concentrations of the reacting species (starting compounds, photoisomers, and degradation products) can be described by an appropriate set of differential equations. The only simplifying hypothesis that is used for their establishment is that the well-stirred mixture obeys Beer’s law. UVhisible multiwavelength analysis of absorbance vs. time curves recorded under continuous monochromatic irradiation provides information about the evolution of the comsponding concentrations [36].Simulated curves from numerical integration of the differential equations generated from the kinetic scheme are compared with the experimental curves using curve-fitting procedures. An example of a typical kinetic analysis of a unimolecular photochromic system under continuous monochromatic irradiation and following the above kinetic scheme (1) is given in Fig. 1.
Q 2001 IUPAC, Pure and Applied Chemistry 73, 639-665
0.6 0.5
8 0.4
B 8 a P
0.3
-t 0.2 0.1
0
0
2
4 6 time I min
8
10
a,
Fig. 1 Evolution of the absorbance at two wavelengths of a unimolecular photochromic system under continuous of the open form, while a', and b', are at the irradiation irradiation. Curves al and b, are monitored at the wavelength &). w irradiation has been stopped at time t,. The thermal rate constant kBAis extracted from the relaxation processes (b, and b', from t , --f -) in the dark, while mABand % are extracted from the kinetic curves under irradiation (a, and a', corresponding to the irradiation periods 0-1,).Knowledge of the evolution of the. absorbance at the irradiation wavelength (&) allows the fraction of incident photon flux that is really absorbed by the photochromic starting compound A to be calculated.
More complex situations that include photobleaching and photodegradation or the presence of several interconverting photoisomers (as in the case of chromenes) can also be treated by similar methods [37].
V11.2 Photochromic behavior and spectrokinetic properties The coloration efficiency (see "colorability") is given by the absorbance A,,(& at the maximum wavelength of the colored form immediately after a pulse of radiation (t = 0). This parameter obtained under standard conditions (concentration of the closed form ca. 2.5 x lo-' M, in toluene solution, at 25 "C) for a series of photochromic colorless compounds such as spiropyrans, spirooxazines, chromenes, etc. (together with the maximum wavelength of their visible absorption spectra), is an indicator of their relative photochromic behavior. Other important data are. the first-order thermal decay rate constants (kh, and the time (t*obd), necessary to obtain half of the initial colorability after continuous irradiation (test of fatigue). Table 1 reports values obtained for some examples of typical photochromic compounds in toluene solution [38]. It is useful for selecting the suitable derivatives for a particular purpose, but the photochromic properties may be different in polymers and vary according to the nature and the oxygen content of the matrix.
V11.3 Photochromic parameters measurements A computer-controlled apparatus [30,38] has been constructed to determine the main parameters describing some photochromic properties such as colorability A&), thermal bleaching rate constant (kh,, and the time necessary to reach half of the initial absorbance (t*cAo12,, reflecting fatigue resistance). The apparatus can be operated in three modes:
0 2001 IUPAC, Pure and Applied Chemistry 73,639-665
XLV Pulsed irradiation: Sequencesof several excitation pulses are followed by a detection pulse; nR is the number of pulses necessary to reach the A, the half value of the colorability A&%). Cyclic mimicking a daylight exposure: Sequences of successive bright continuous irradiation and dark periods are carried out until the A , value is obtained. Continuous: Sequences are.the same as in the cyclic mode but the dark period is reduced to 1 min to stabilize the photomultiplier tube. Table 1. Photochromic parameters in diluted toluene solutions. Compound
(2.5x10~5M)
in toluene
ColoredForm A,/m
A,(,%) r-0 4.6
HlCo
bniho-garthoxv BPS
1.08
0.84
VIII. APPLICATIONS
VIII.l General applications General applications of photochromism can be divided into two categories: a) b)
those directly related to the change in absorption or emission spectra such as variable transmission optical materials, optical information storage, cosmetics, authentication systems, and flowfield visualization those related to other physical or chemical property changes such as refractive index, dielectric constant, electric conductivity, phase transitions, solubility, viscosity, and surface wettability
The most famous application is sunglassses; some other potential applications are described below.
V111.2 Actinometry “An actinometer is a chemical system or a physical device which determines the number of photons in a beam integrally or by unit time.. .”as defined in the “Glossary of Terms Used in Photochemistry”
WI.
0 2001 IUPAC, Pure and Applied Chemistry 73,639-665
XLVI Chemical actinometry in solution has the advantage over physical actinometry in that the former can be used under conditions similar to those of the photoreaction to be studied [40]. In addition, photochromic actinometers can be used repeatedly, thus obviating the need for a fresh sample for each measurement. Some thermally stable photochromic compounds, such as the following, fulfil this requirement. AberchromeTM540is well suited for chemical actinometry within the range 310-370 nm (coloring reaction) and 435-545 nm (decoloration) domains [41a]. It should be noted that it is useful at 365 nm where several other actinometers are inappropriate.
-
W (310 370 nm) a
WsiMs (bleaching)
The quantum yield for coloring is temperature-independent(5-55 “C). It can be used repeatedly mol dm” toluene solution is employed. A drop in quantum yield is found on repeated when a 5 x use at lower concentrations [41b-e]. Azobenzene >6 x lo4 mol dm-3 in methanol can be used in the 254-334 nm range [42a,b]. The absorption spectra of the trans and cis isomers differ sufficiently to produce significant changes in absorbance during the photoisomerization. Heterocoerdianthrone endoperoxide (HECDPO) is a suitable actinometer in the 248-334 nm region. It is formed from an air saturated solution of heterocoerdianthrone(HECD) in methylene chloride that has been in sunlight with a 420480 nm band pass filter and stored in the dark [43]. Between 253 and 302 nm, the reaction efficiency does not depend on irradiation wavelength.
sunlight
0
0
HECDPO (colorless)
HECD (red)
V111.3 Optical power-limiting substances Optical power-limiting substahces are used to protect the human eye or optical sensors from the damaging effects of intense flashes of light. An ideal limiter becomes suddenly opaque during exposure to a laser burst and immediately transparent again at the end of the pulse. Compounds that are potential optical limiters are fullerenes, indanthrones, porphyrins, mixed metal clusters, and phthalocyanines (especially,chloroindiumphthalocyanine) [44,451. The phenomenon is caused by the promotion of a very large fraction of the molecule from the ground state to an excited state, which absorbs photons more strongly than the ground state. Q 2001 IUPAC, Pure and Applied Chemistry 73,639-665
XLVII
I
s,
V111.4 Photoresponsive materials Reversible photoinduced changes of physical and chemical properties can be transferred to the microenvironment by a photochromic molecule incorporated in the system (Fig. 2).
Fig. 2 Photoinduced cyclic variation of a physical propelty in a photoresponsive system.
Some examples are given below: Photomechanical effects: Reversible photomechanical deformation could be observed using Brewster angle optical microscopy with monolayers of polyvinyl alcohol bearing an azobenzene side chain, “PVA-AzB” derivatives,on water surface [46]. Chiroptical molecular switches: Photoisomerization of photochromic units (azobenzenes, spiropyrans, etc.) in poly(a-amino acid)s is able to trigger a random coil to a-helix transition. This primary photochemical event occurring in the side chains is amplified and transduced by the structural variations of the macromolecularmain chains. The latter are accompanied by large and reversible variations of optical activity [47]. Because of their reversibility, these systems can act as “chiroptical molecular switches” (see also [27]). Q 2001 IUPAC, Pure and Applied Chemistry 73, 639-665
XLVIII Sol-gel transition temperature: 1% Aqueous solutions of poly(N-isopropylacrylamide) can form thermoreversible gels [48]. Pendant azobenzene groups (2.7 mol %) have been shown to change reversibly the gelation temperature. At 750 nm, the solution is transparent, whereas the gel is opaque as shown in the graph below (Fig. 3). The cis form of azobenzene maintains a high percent transmittance between 20 "C and 26 "C the trans isomer induces a sol-gel transition temperature around 20 "C. These transition temperatures are controlled by light (A, or 4) which triggers cis-trans isomerization, so that the transmittance can be tuned from 80% to 0% (or 0% to 80%) by irradiation in the 20-26 "C temperature range [48]. (tsol.gel = 30 "C), whereas
100
sol
t I'C
Fig. 3 Transmittance variation at 750 nm of the photochromic solution (see text) vs. temperature; 0 in the dark; o upon photoirradiation. Irradiation at 350 nm < .2,c 410 nm solubilizes the polymer, and the solution becomes transparent; visible irradiation at 1,> 410 nm decreases the solubility, and the polymer leads to phase separation. (Adapted from [48] with permission; copyright 1990 Springer Verlag.)
V111.5 Photoswitchable biomaterials The combmation'of biomaterials and photochromic compounds is the basis for the design of optobioelectronic devices. The following example, describing a photoreversible immunosensor electrode (see Fig. 4). illustrates the contribution of photochromism to the development of biomaterials science
WI.
The antigen (Atg) functionalized gold electrode, modified by a self-assembled monolayer (a), yields an amperometric signal in the presence of a redox probe (R+/R). Association of the antibody (Ab) to the monolayer (b) insulates the electrode towards the solubilized redox probe resulting in the absence of an electrical signal. Owing to the tight "Atg-Ab" association, such electrodes are limited to a single analysis. But the antigen can be chemically modified by a photochromic component (A) which does not impede the amperometric sensing of the antibody. The antigen monolayer will be perturbed by photoisomerization of A into B and will release the antibody (c). The active Atg monolayer (a) is regenerated by the reverse photoisomerization B + A. Further analyses are thus possible.
Q 2001 IUPAC, Pure and Applied Chemistry73,639-665
XLIX
Fig. 4 Schematic assembly of a photoreversible immunosensor electrode. Reproduced from [49] with permission; copyright 1997 American Chemical Society.
IX. ACRONYMS ARPO
gomatic endoBerQxide
BIPS
“Benzo Indolino Pyrano Spiran”
B P S are derivatives of l’,3’,3’ trimethylspiro (2H-l-benzopyran-2,2’-indoline) 6-nitro BIPS is one of the most popular Spiropyrans
-
BISO Benzo Indolino Spiro Oxazine
BR DHI DHPP DNE DPB DPH DPO HR MC
bacteriorhodopsin spirodihydroindolizines dihy dropyrazolo-pyridine dinaphthylethylene 1,Cdiphenylbuta-1,3-diene 1,6-diphenylhexa- 1,3,5-triene 1,I-diphenylocta- 1,3,5,7-tetraene halorhodopsin merocyanine (open form of SP, SO, etc.)
(y$-/p-(y!$+ pcb I
N
0
quinonic form
zwitterionic form
Q 2001 IUPAC, Pure and Applied Chemistry 73,639-665
L
NPE Pr Pfr
naphthylphenylethylene phytochrome (“red” absorbing) phytochrome (“far red” absorbing)
so
spirooxazines (such as BISO, NISO,QISO, etc.) spiropyrans
SP
X. TECHNICAL VOCABULARY FOR OPHTHALMIC LENSES X.1 Luminous transmittance The darkening efficiency is given by the transmittance variation at a given temperature.
X.2 Activation time Activation time is the time necessary to achieve a given luminous transmittance by exposure to UV at a given temperature. X.3 Thermal fade time
Thermal fade (also termed fade-back) time is the time necessary to bleach (indoors) a colored form to one half its original absorbance at a given temperature. X.4 Bleached state Bleached state is the colorless form of a photochromic system. X.5 Activated state
Activated state is the colored form of a photochromic system.
XI. BOOKS ON PHOTOCHROMISM Photochromism, G. H. Brown (Ed.) (Techniques of Chemistry Vol. m),Wiley-Interscience, New York, 1971 (853 pp.). Organic Photochromes, A. V. El’tsov (Ed.), (translation edited by J. Whittal), Consultants Bureau, New York and London, 1990 (280 pp.). Photochromism, Molecules, and Systems, H. Durn and H. Bouas-Laurent (Eds.) (Studies in Organic Chemistry 40),Elsevier, Amsterdam, 1990 (1068 pp.). Applied Photochromic Polymer Systems, C. B. McArdle (Ed.), Blackie, Glasgow and London, 1992, (published in the USA by Chapman and Hall), New York (255 pp.).
Q 2001 IUPAC, Pure and Applied Chemistry73,639-665
LI Handbook of Organic Photochemistry and Photobiology, W. M. Horspool and P. S. Song, (Eds.), CRC Press, Boca Raton, FL, 1995 (1636 pp). Part I: Ch. 13, “Fulgides and related systems”, H. G. Heller. Ch. 83, “Photochromic nitrogen containing compounds”, H. Durr. Part 11: Ch. 20, “Phototropism”, K. L. Poff and R. Kongevic. Ch. 21, “Phytochromes”, W. Parker and P. S. Song. Ch. 28, “Bacteriorhodopsin and Rhodopsin”, R. Needleman. Organic Photochromic and Thermochromic Compounds (Topics in Applied Chemistry), J. C. Crano and R. Guglielmetti (Eds.), Plenum, New York, 1999, Vol. 1 (376 pp); Vol. 2 (473 pp).
ACKNOWLEDGMENTS We are deeply indebted to our colleagues of the Working party on Photochromism for their contributions and very helpful suggestions. We also thank the members of the W A C Photochemistry Commission, particularly J. Bolton, A. U. Acuiia, and J. Wirz, for their encouragements, critical reading and perceptive comments. Special thanks are due to R. G. Weiss, J. L. Pozzo, and J.-P. Desvergne for assistance.
XII. REFERENCES 1. 2. 3. 4. 5. 6. 7. 8. 9.
10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21.
22.
J. Fritzsche. Comptes Rendus Acad. Sci.,Paris, 69, 1035 (1867). E. ter Meer. Ann. Chem. 181, 1 (1876). (a) T. L. Phipson. Chem. News 43,283 (1881); (b) J. B. Om.Chem. News 44, 12 (1881). W. Markwald. Z. Phys. Chem. 30, 140 (1899). Y.Hirshberg. Compt. Rend. Acad. Sci., Paris, 231,903 (1950). Photochromism, G. H. Brown (Ed.), Wiley-Intersciences, New York (1971). R. C. Bertelson. Mol. Cryst. Liq. Cryst. 246, 1 (1994) (period 1955-1993). D. A. Parthenopoulos and P. M. Rentzepis. Science 245,843 (1989). A. S. Dvornikov, S. E. Esener, l? M. Rentzepis. Optical Computing Hardware, Ch. 11, AT&T and Acad. Press (1994). M. Uchida and M. Irie. J. Am. Chem. SOC.115,6442 (1993). H. G. Heller. In Photochromicsfor the Future Electronic Materials, L. S . Miller and J. B. Mullin (Eds.), Plenum, New York (1991). P. M. S. Monk, R. J. Mortimer, D. R. Rosseinsky. Electrochromism: Fundamentals and Applications, VCH, Weinheim (1995). A. Samat and V. Lokshin. “Thermochromism of Organic Compounds” in Organic Photochromic and Thermochromic Compounds, J. C Crano and R. J. Guglielmetti (Eds.), Vol. 2, Ch. 10, Plenum, New York (1999). A. Ault, R. Kopet, A. Serianz. J. Chem. Ed. 48,410 (1971). A. SchiSnberg, M. Elkaschef, M. Nosseir, M. M. Sidky. J. Am. Chem. SOC.80,6312 (1958). “Tribochromic Compounds and their Applications”, H. G. Heller and A. M. Asiri, PCT, Int. Appl., WO 94 26,729 (1994). C. Reichardt. Solvents and Solvent Effects in Organic Chemistry, pp. 3, 285, 286, VCH, Weinheim (1990). P. Suppan and N. Ghoneim. Solvatochromism, Roy. SOC.Chem., Cambridge (1997). C. Reichardt. Solvents and Solvent Effects in Organic Chemistry, pp. 288,375, VCH, Weinheim (1990). M. Irie, 0.Miyatake, K. Uchida. J. Am. Chem. SOC. 114,8715 (1992). S. H. Kawai, S.L. Gilat, J.-M. Lehn. J. Chem. SOC.,Chem. Commun. 1011 (1994). N. P. M. Huck and B. L. Feringa. J. Chem. SOC.,Chem. Commun. 1095 (1995).
Q 2001 IUPAC, Pure and Applied Chemistry 73,639-665
LII 23. X. D Sun, M. G Fan, X. J. Meng, E. T Knobbe. J. Photochem. Photobiol. A. Chem, 102, 213 (1996). 24. (a) Y. Yokoyama, T. Yamane, Y. Kurita. J. Chem SOC., Chem. Commun. 1722 (1991) (Fulgides); (b) F. Pina, M. J. Melo, M. Maestri, R. Ballardini, V. Balzani. J. Am. Chem. SOC.119,5556 (1997) (Flavylium salts); (c) C. Weber, F. Rustemeyer, H. Dun. Adv. Mate,: 10, 1348 (1998) (Spirodihydroindolizines,DHIs). 25. A. von. Baeyer and V. Villiger. Ber. Dtsch. Chem. Ges. 35, 1189 (1902). 26. M. Zhang and G. B. Schuster. J. Am. Chem SOC.116,4852 (1994). 27. B. L. Feringa, W. F. Jager, B. de Lange. J. Chem. Soc., Chem. Commun. 288 (1993). 28. N. F? M. Huck and B. L. Feringa. J. Chem. Soc., Chem. Commun. 1095 (1995). 29. P. Appriou, F. Gamier, R. Guglielmetti. J. Phorochem. 8, 145 (1978). 30. V. Malatesta. “Photodegradation of Oxganic Photochromes” in. Organic Photochromic and Thermochromic Compounds, J. C. Crano and R. J. Guglielmetti (Eds.), Vol. 2, Ch. 2. Plenum, New York (1999). 31. M. hie and K. Uchida. Bull. Chem SOC. Jpn 71,985 (1998). 32. N. Hampp and C. Brauchle. Photochromism, Molecules and Systems, H. Diirr and H. BouasLaurent (Eds.), Ch. 29, Elsevier, Amsterdam (1990). 33. (a) J. Epperlein, B. Hoffmann, K. S. Topperka. J. SigMlaufzeichnungsmater 3, 173 (1975); (b) J. Epperlein, B. Hoffmann, K. S. Topperka. J. SigMlaufieichnungsmater 4, 155 (1976). 34. H. G. Heller. “Fulgides and Related Systems” in Handbook of Organic Photochemistry and Photobiology, W. M. Horspool and Pill-Soon Song (Eds.), Ch. 13, CRC, Boca Raton, FL (1995). 35. See Handbook of Organic Photochemistry and Photobiology, W. M. Horspool and Pill-Soon Song (Eds.), Chs. 21-33, CRC, Boca Raton, FL (1995), and Photochromism Molecules and Systems, H. Diirr and H. Bouas-Laurent (Eds.), Chs. 19 and 20, Elsevier, Amsterdam (1990). 36. V. Pimienta, D. Lavabre, G. Levy, A. Samat, R. Guglielmetti, J. C. Micheau. J. Phys. Chem 100, 4485 (1996). 37. M. H. Deniel, D. Lavabre, J. C. Micheau. In Organic Photochromic and Thermochromic Compounds, J. C . Crano and R. J. Guglielmetti (Eds.), Vol. 2, Ch. 3, Plenum, New York (1999). 38. R. Dubest, F? Levoir, J. J. Meyer, J. Aubard, G. Baillet, G. Giusti, R. Guglielmetti. Rev. Sci. Instrum. 64,1803 (1993). 39. J. W. Verhoeven. “Glossary of Terms used in Photochemistry”, Pure Appl. Chem. 68, 2223 (1996). 40. A. Braun, M. T. Maurette, E. Oliveros. Photochemical Technology, Wiley, Chichester (1991). 41. (a) H. G. Heller and J. R. Langan. J. Chem SOC.Perkin Trans 2, 341 (1981); (b) Y. Yokohama, H. Hayata, H. Ito, Y. Kurita. Bull. Chem. SOC. Jpn. 63, 1607 (1990); (c) P. Boule and J. F. Pilichowski. J. Phorochern Photobiol. A: Chem. 71,51 (1993); (d) P. Boule and J. F. Pilichowski. EPA Newsletter 47, 42 (1993);(e) H. G. Heller. EPA Newsletter 47,44 (1993). 42. (a) G. Gauglitz and S. Hubig. J. Photochem. 30,121 (1985); (b) G. Persy and J. Win. EPA Newsletter 29.45 (1987). 43. H. D. Brauer and R. Schmidt. Photochem. Photobiol. 37,587 (1983). 44. Non Linear Optics of Organic Molecules and Polymers, H. S . Nalwa and S. Miyata (Eds.), CRC Press, Boca Raton, FL (1997). 45. C. W. Spangler. J. Mate,: Chem. 9, 2013 (1999). 46. T. Seki, H. Sekizawa, R. Fukuda, T. Tamaki, M. Yokoi, K. Ichimura. Polym. J. 28,613 (1996) and references therein. 47. A. Fissi, 0. Pieroni, G. Ruggeri, F. Ciardelli. Macromol. 28,302 (1995) and references therein. 48. M. Irie. Adv. Polym. Sci., H. Fujita (Ed.), pp. 27-67. Springer Verlag, Berlin (1990). 49. 1. Willner. Acc. Chem. Res. 30, 347 (1997).
Q 2001 IUPAC, Pure and Applied Chemistry 73,639-665
LIII
XIII. SUBJECT INDEX Aberchrome, VI11.2 Acidichromism, 111.15 Acronyms, IX Actinometry, VIII.2 Activation time, X.2 Activated state, X.5 Anils, IV.9 Anthracene, IV.8 Applications. VlII Applied photochromic polymer system, XI.4 Azobenzene, VIII.2, VIII.4 A m compounds, IV.7, V1.2 Bacteriorhodopsin, 111.20, V.1, IX, X1.5 Biological photoreceptors, V Biomaterials (photoswitchable). VIII.5 BIPS, IX Bleached, 111.1 Bleached state, X.4 Books on photochromism, XI Chemical processes involved in organic photochromism, VI Chirochromism, 111.18 Chiroptical molecular switches, II1.18, V111.4 Chromenes, 11.2, IV.3, VII.1, VII.2 Chromism (ending), 111.5 Cis-transisomerization, V1.2 Colorability, 111.19, V11.2, VI1.3 Coloring reaction, VIII.2 Cyclability, 111.22 Decolorization (decolorstion), VIII.2 Degradation (chemical), II1.20,111.21, VII. 1 DHI, IV.6, VI.l, IX Diarylethenes, 111.13, IV.5 Diastereophotochromism, III. 18.2 Diastereothermochromism, 111.18.2 Dihydroindolizines (DHls), 1V.6, VI.l, IX Dissociation processes, V1.5 Dithienylcycloperfluorocyclopentenes,IIl.20 Dual-mode photochromism, 111.14 Electrochromism, 111.7,111.14 Electmn transfer, VI.6 Erasing, 111.14 Families of organic photochromic compounds, IV Fatigue, IIl.20, V11.2, VII.3 Fluorescence, 111.2 Fulgides, IV.4, XI Fulgimides, IV.4 Gated photochromism, 111.13, 111.18 Group transfer, V1.4 Half-life, 111.23 Halochromism, 111.17 Halosolvatochromism, 111.12 Heliochromic compounds, 111.6 Heterocoerdianthrone endoperoxide (HECDPO), IV.9, VIII.2 Hydrogen transfer, IV.9, VI.3 Inverse photochromism, 111.1 Ionochromism, III.16 Kinetic analysis (of photochromic systems), VII.1 Kinetics (of photochromic systems), VII.1 Luminescence, 111.14 Luminous transmittance, X.1
Q 2001 IUPAC, Pure and Applied
Mechanistic aspects, VII Memory system (3D). 111.2 Merocyanines, IILIS, IV.1, W.2,IX Negative photochromism, 111.1 Number of cycles, 111.21 One-photon photochromism, 111.2 Ophthalmic lenses, 11.2 Opto-electrical switch, 111.14 Opto-bioelectronic devices, VIII.5 Optical power limiting substances, V111.3 Oxidation, 111.20 Periaryloxyparaquinones,IV.10. V1.4 Pericyclic reactions. VL1 Perimidinespimcyclohexadienones, 1V.1I Photobleaching. VII. 1 Photochrome (photochromes). 111.3 Photochromic compounds, 111.3, XI Photochromic nitrogen containing compounds, XI Photochromic parameters measurements, VII.2, VII.3 Photochromic polymer system (applied), X1.4 Photochromics, 111.3 Photochromism. I, Il.1, 11.2.111.1,111.6. XI Photochromism (definition), I, 111.1 Photochromism (brief history), 11.1.11.2 Photochromism: type. P, type T., 111.1 Photochromy, 111.3 Photodegradation, VII.1 Photomechanical effects, VII1.4 Photoresponsive materials. VI11.4 Photoreversible immunosensor, electrode, VIII.5 Photoswitchable biomaterials, VIII.5 Phototropy, 11.1 Phototropism, ILl,I11.4, XIS Phytochrome, V.2, XI.5 Piezochromism, 111.9 Polycyclic aromatic compounds, IV.8 Polycyclic quinones, IV.10. V1.4 Positive photochromism, 111.1 Readout (reading process), 111.2, I11.14 Readout number, 111.24 Retinal proteins, V.1 Rhodopsin, V.l, XI.5 Simultaneous absorption (of two photons), 111.2 Solvatochromism, 111.1 1 Sol-gel transition (photoinduced), VII1.4 Spectrokinetic properties, VII.2 SO (Spirooxazines), 11.2.111.8, I11.15. IV.2, V1.I. IX SP (Spiropyrans), III.1, III.8, IV.l, VLI, VI11.4 Stepwise absorption (of two photons), 111.2 Sun lens applications, 111.6 Switch (on and off), 111.21 Tetracene, 11.1 Thermal decay bleaching rate constanf VI1.2. V11.3 Thermal fade time, X.3 Thermochromic, III.l,II1.8, XI.6 Thermochromism. III.l,111.8. XI.6 Thermosolvatochromism, 111.8 Triarylmethanes, IV.13 Tribochromism, 111.10 Two-photon photochromism, 111.2 Viologens, IV. 12 Writing ( rocess), II1.2,111.l4 z,1 n . b
Chemistry 73,639-665
This Page Intentionally Left Blank
1
Chapter 7
General Introduction H. Durr
1 Oraanizationof the Book
Photochromism, dealingwith photochemicalreactionswhichare thermallyor photochemicallyreversible,is a part of photochemistry.It has receivedconsiderableattention ever since its discoveryin 1876 and stillis an activefield of researchmainlybecauseof itsactual and potentialapplicationsand for its paramountimportancein biologicalphenomena. The book reflectsthe state of affairsin the differentareas of researchin bothoraanic and inoraaniccomrJoundsand svstemswith a largercontribution for organicmolecules. Specialemphasisis givento photochromic bioloaicalsvstemsand the environmentalor supramolecularrole on the photochromic propertiesof organiccompounds.. A Q Q ! ~ ~ tions are coveredin the last section.Organicphotochromism,beingdefinedas a molecularproperty,makes it temptingto organizeits presentationaccordingto the different classesof moleculesbut each classundergoesvarious types of reactionsand this wouldlead unavoidablyto a great numberof repetitions.A classificationbased on reactiontypes, withthe pericyclicreactionsas the mainguidewheneverit is possible, seems more adequateto preventredundancy.In addition,we believethat it helpsto give a moregeneralunderstandingand appraisalof the fieldthan the formerapproach. Nevertheless,it may be usefulto quicklyfind out whichphotochromic reactionsa given moleculeor a certainclassof compoundscan undergo; here the subjectindexshould be helpful. The bookon "Photochromism-Molecules and Systems"beginswith an introductory Chapter (2) giving the photophysical backgroundof the phenomenon"photochromism". Photochromicsystemsbasedon cis-transisomerismsuch as simpleor functionalized olefinsand azocompoundsare presentedin Chapter3 and 4. These chaptersdeal with systemsin whichcis-transisomerizationleadsto differentlyabsorbingspecies which revertreversiblyto the startingmaterial. Electrocyclicreactionsof olefinessuchas stilbenes are treatedin Chapter 7 . Not treated inthis bookare cis-transisomerizationsof indigoderivatives.A listof recentreferencesis given. Pericyclicreactionsare collectedin the followingChapters(5-15). Here, heterocyclic and carbocycliccompoundsas well as the open ring moleculeswill be combinedin the sectionon electrocyclicreactions.Spiropyransbelongingto one of the most important classes of photochromic systemsare describedin thiscontext.It shouldbe mentionedhere however,whetherthe basic ringopeningprocessin spiropyranesoccursin a concertedor non concerted(heterolytic)fashion.Thus the generalpatternof the reactionsand the relationshipof the electronicreorganizationin the variousmoleculeswill becomeevident. Hence not the natureof the bondcleavagebut ratherthe complete Qlactronic situationof a moleculewill dominateits reactions.
The cycloadditionsleadingto photochromicsystemsin an intra-or intermolecularreactionare mentionedin Chapters12 to 15. Hydrogentransferreactionscan be the basicprocessesin photochromic systems. These transformations(Chapters16,17) may, in a very formalsense, be regardedas sigmatropicreactions. In thischapter, they are classifiedin detail accordingto the differenttypes of compounds. Dissociationprocesses followin Chapter 18 normallyproducingfragmentseitherof ionicor radicalnature. Biologicallyimportantphotochromicprocessesare coveredin Chapters19 and 20. Here phytochromesand the visualprocessare center points.The photosynthesis will be excludedsincethistopicdoes not meet the definitionof photochromismin a strictsense. The importanceof the environmentalor supramoleculareffectson or in photochromic systemsis reportedon in Chapter 21. Here the importanceof modifyingthe properties of a photochromic systemespeciallyfor potentialuses is demonstrated. Inorganicphotochromic systemsare presentedin Chapter 22, also with regardto the fundamentalprocesses.The applicationsof bothorganicand inorganicphotochromic systemsare describedin Chapters23 to 29. The Chapterwill informthe reader in whichfield and for whichpurposephotochromicsystemscan be, and are, employed.It stressesthe classicaluses of photochromicsystemsin the field of ophthalmiclensesor photocoloration,photoimagingand others. New actualand potentialapplicationssuch as actinometry,reusable informationstorage, i.e. chemicalswitchesand molecularelectronicsfor computers,data displays,holography,non linearopticsand other novelapplicationswill be surveyedin the last section. Photochromism based on triplet-triplet absorptionhas not been explicitelyincludedin thisbook, but a listof recentreferencesis givenafter Chapter 2. Novel phenomena beinglinkedto photochromism withoutmajorstructuralchangessuchas orientation inducedphotochromism "energyselection"photochromism by holeburningand a new promisingphotochromic systembased on bacteriorhodopsin are includedin this book.
3
2 BriefHistoricalSurvevof Photochromism
Photochromism was reportedfor the firsttime by Fritschein 1867 who observedthat tetracenewith air and lightproduceda colorlessmaterialwhichregeneratedtetraceneon heating(ref. la). Shortlyafter, in 1876, ter Meer describedthe same phenomenonwith potassiumsalt of dinitroethane,in the followingway (1b): "EsbildetKrystallevon reingelberFarbe mit starkemGlanz, die jedochan der Lufttriibe werden und sichrothen. Im Dunkelnverschwindetauffallendetweisedie rotheFarbe wieder." Phipsonobservedthat a gate postpaintedwith a zinc pigment(probablysome kindof a lithophone)was blackwhen exposedto sun butwhiteduringthe night(ref. 2). Marckwald,when investigatingin 1899 the behaviorof benzo-l-naphthyridine and tetrachloro-l,2-keto-naphthalenone in light,recognizedthat the resultingcolorchangesare in fact due to a new phenomenon.He coinedthe term "ghototroov" (ref. 3) for thisexperimentalfact. "Mankenntnun Stoffe,welche unter der Einwirkungvon Lichteine Zustandsanderung erleiden.... diese Lichteinwirkungen sinddadurchcharakterisiert,daB die Zustandsanderungnachder Belichtungschnelleroder langsamerwieder verschwindet... fur solche Zustandsanderungenschlageichden Begriff"Phototropie" vor." However,todayghototropvis understoodas the light-inducedinteractionsoccurring in biologicalsystems,includingthe effectsof lighton the nutritionalsystemsof plantsor microoraanisms. whereas .(3hototropismis the tendencyof plantsto turntowardsthe sourceof light (ref. 4). Thereforethe phenomenon"ghotochromism" was suggestedby Hirshbergin 1950 (ref. 5) in the followingway: "Siune solutionde bianthroneest irradi6epar la raie 3650 f h 60' C, sa couleurverte se change en rouge brunstre.L'effet est tres net d6jh apres une minuted'irradiation.II est d'un caracteretransitoireet reversible,car en rkhauffantla solutiona la temperature ordinaireon recuperela bianthroneinalter6e." This definitionof photochromism whichuses the Greek words = lightand xpoj.ca = coloris more appropriateto describea reversiblechargeof color; it is usedtoday almostexclusivelyand thus is employedin this book. However,this definitionis not fullysufficientand has to be enlarged (videinfra). Almostat the same time as the firstreporton photochromism, other compoundswere shownto be photochromicas well. Wislicenus(ref. 6) observedthat benzalphenylhydrazone, Biltz (ref. 7) that benzalphenylhydrazones and osazones,were photochromic.The photochromism studiesuntil1921 dealt mainly with the phenomenonin a more practical or descriptiveway. Thus synthesisof photochromicmolecules,the selectionof exci-
4
tingradiationand speed of excitation,bleachingand fatiguewere the mainpointsof interest. The literaturewas reviewedin the papersof Stobbe(ref. 8), Chalkley(ref. 9), Bhatanagaret al. (ref. lo), van Overbeck(ref. 11) and Brown(refs. 12, 13). The literature until1921 was collectedin ref. 8 and the later literatureuntil1961 in refs. 12, 13. Afterthe first20 years in the periodfrom 1900-1920 only limitedattentionwas devoted to the effectof photochromismin the 1930’s; Let us emphasizefor example,the examinationof malachitegreen (refs. 14, 15) and semicarbazone(refs. 16-18). From 1940 the interestin photochromismrose again. Numerousinvestigations were carriedout to gain insightintothe mechanismsof the photochromicprocesses,the structureof the products,the intermediatesformed,the mechanismsinvolvedfor photo chromismand fatiguereactions.More and more sophisticatedphysicaltoolsavailable, suchas nmr, ir , esr and x-ray, were used. It is worthmentioning,amongothers,the pioneeringwork by Hirshberg’steam in Israel (refs. 5, 19, 20). A milestonethe patentingand manufacturingof photochromic sunglasses(CorningGlassWorks) in the sixties. With Porter ’s (ref. 21) discoveryof time-resolvedor flashspectroscopy,a new area for the studiesof transients,lifetimesand propertiesof excitedstatesemerged.With the aid of thistechniquewhichhas recentlybeen extendedfrom detectionmodesby uv-absorptionto emission,ir, Raman, nmr and esr, a very refinedinsightintothe natureof transientsand excitedstatesand the reactivitiesof photochromicmoleculeshas been, and stillis possible,fromabout 1960-1988 whichcouldnot have been imagined20 years ago. This fosteredprogressparticularlyin biologicalphotochromicsubstances.In recentyears also, considerableeffectswere devotedto a systematicinvestigationof establishedfamiliesof photochromic compoundssuchas fulgides,spiropyransand relatedheterocycliccompounds.New photochromic systemswere also discoverede.g. the photoaddition of singletoxygento aromaticcompoundsand the spiroindolizines. Incorporationof photochromic moleculesin polymersor molecularassembliesis found, to be of increasingimportancefor the modulationof physical properties of materials. Finallythe adventof laser photophysics allowsan expansionof the field. A meetingof the researchersin the fieldof photochromism was held some years ago in lnterlakenin 1984. This meetingbroughttogetherall those interestedin photochromism fromthe whole worldand stimulatedexchangeof ideas and activityin this area. Today, it appearsthat the field of photochromismis not dead but muchalive and.exciting.
5
3 Definitionof Photochromism It has been mentionedabovethat photochromism was observedin the early periodas a color formedin sunlightduringthe day, fadingaway in the night.Scientificallyspoken thisphenomenonmust'be extendedto all electromagneticradiation,thus leadingto a very generaldefinitionof photochromism extendingthe scope of the early observations (refs. 1-5): "Photochromism is a reversibletransformationof a singlechemicalspeciesbeing inducedin one or bothdirectionsby electromagneticradiationbetweentwo states havingdifferentdistinguishable absorptionspectra."The radiationchangesmay be inducedby uv, visibleand ir radiation.Reversibiliiis the importantcriterion.All irreversiblereactionsare normalphotochemistry and are not includedin the book. Most photochromic systemsare based on unimolecularreactions.
/? I
1
c
I
A
hP
A-
!Lp unimolecular
The startingmaterialA (educt)undergoesformationof productP, inducedby electromagneticradiation.The back reactionP -- > A can occurthermally(T-type) or photochemically (P-type). (The termsT-type and P-typephotochromism are usedin some chapters of this book).The typicalpatternsof the absorbanceprior,duringand after irradiation of a photochromic moleculeare givenin Fig. 1. (adaptedfromref. 19).
!iv-\
6
EXCITING RADIATION ON
STEADY STATE
C$UCENTRATlON
oW n L c *
pa
,
11
I
EXCITING OFFRADIATION
I
TIYE-
12
( A 2 P) Fig. 1 Typicalanalysiscurvefor a photochromicsystem(Absorbancevs time at a givenwavelength)taken from ref. 19. The definitionof Hirshberggivenabove does not, however,cover bimolecularreactions. Such bimolecularreactionsare reversiblephotocycloadditions or electrontransferprocessesof the type:
A
+
B ->
P (bimolecularprocess)
Dependingon the system,mono-or bimolecularback reactionscan occur (vide infra).
Cyclicreactionsof the type A ---> B B + C ---> A + D are excluded. However, fwo .DhotonDrocesseswill be treatedin this book. If a light-induced process occursfrom an upper excitedstate populatedby simultaneousabsorptionof two light quanta,the whole systembecomesa nonlinearone. Such systemshave been prepared for opticalrecordingsystems (ref. 22; see Chapter26); two Al, A2 as well as three. Al, A2, A3 energy levelsystemshave been suggestedfor these two photon processesA four energy leveldiagramand systembecomepossibleiftripletstates (BI, 82) in additionto the singletstates (Al, A2) are involved.If B1 is a groundstate molecule the populationof the 82 statecan risevia a two photonabsorption.Non linear systemsof thistype are appliedin holography.Such systemsare truly photochrornic.
7
"37-
=z
Fig.2 Two-photonphotochemistry (a) A two-levelone-photonsystem (b) a three-leveltwo-photonsystemand (c) a four-leveltwo-photonsystem ( adapted,fromIchimura,chapter26)
The importantchangesthat occurin photochromic systemsare in the absorptionand emissionspectraof A and P in the quantumyield cp A --> P, in the refractiveindexand in the dielectricconstant.These changesare intrinsicto photochromic phenomena.The enthalpychangesinvolvedin the transformationA -- > P are responsiblefor the thermal (or photochemical)back reaction.The side reactionscausethe fatigueof a photochromic molecule.These changesat the molecularlevelinduceconformationalmodificationsof the environmentor the surroundingmatrix(solvent,solidor liquidphase, polymers). The differencesof physicalor chemicalpropertiescan be transferredto the microenvironmentor supramolecularstructureby photochromic molecules.Such effectswere studiedfor instancewith cis-transisomerizationsor electrocyclicreactions. These differentlight-inducedpropertychangesin more complexassembliescan be regardedas supramolecularsystems.The photochromicsupramolecularunitshave also been called photoresponsive materials(see Chapter 26). A surveyof potentialmodificationsto be expectedin supramolecularassembliesis givenin Table 1.
8
Table 1 Reversiblechangesof propertiesin photochromic compoundsand materials(adapted fromK.lchimura, see Chapter26).
Properties optical
Reversiblechangein Molecularstructure Supramolecularstructureof matrices absorptionspectrum emissionspectrum refractiveindex dielectricconstant
absorptionspectrum emissionspectrum refractiveindex dielectricconstant light scattering birefringence opticalrotatorypower reflectivity
chemical
chelateformation iondissociation enthalpy
electrical
bulk
chelateformation iondissociation enthalpy catalysis enzyme activity membranepermeability conductivity photoconductivity capacitance membranepotential
phasetransition solubility
phasetransition solubility viscosity wettability density elasticity
9
Summary: Photochromicsystemscan be classifiedin severalgroupsaccordingto the natureof the photochemically inducedprimarystep: 1) photoreversiblesystems,in whichthe coloredformP undergoesa light-induced reactionbackto the formA (unimolecularprocess)or A and 6 (bimolecularprocess).It is understoodthat the initialform absorbsat shorterwavelengththanthe photoproduct. 2) thermoreversiblesystems,in whichthe colorvariant P revertsthermallyto A (or A and 6) 3) systemswhichare bothphoto-and thermoreversible 4) inversephotochromic systems,in whichthe initialform A (or A and 6)can absorb at longerwavelengthsand form P at shorterwavelengths. 5) multiphotochromicsystems,in whichmorethan two formsundergophotochemical and/orthermochemicalinterconversion.
10 4. Outlookand FutureDeVelODmentS
Studiesof the simplecis-transisomerizationin olefinswiththe aid of moderntime resolved techniquesleadsto a deep understandingof this process.These investigations with one of the mostsimplechromophoresare extremelyimportantsincethey serve as modelsfor the more complicatedsystemsin visualpigments,and photoreceptors suchas phytochromes (Chaps. 19,20) as well as far cis-transisomerizationof urocanic acidin the epidermisand of the vitaminD synthesis.
Azo compounds,beingcommercialdyes belongingmostlyto the class of "pseudostil-
benes"(see below), possessa highlightfastness.They are already and may be used to a larger extentin the future in supramolecularassembliessuchas polymers,crownethers, liquidcrystalsor vesiclesto triggerenvironmentalchangesin opticalrecording systems. A new photochromic systemhas been developedbeingbased on the 1,delectrocyclization of mono-or poly-aza-pentadienes showinga large potentialwhichhas not been fullyexplored.An openfield is that of 1,8electrocyclicreactionsof type-i systems(derivativeof i-amino-1,3butadiene)whichhave not been studiedso far.
interconversion is one of the mostimportantreactions The hexatriene/cyclohexadiene givingriseto photochromic systems.Althoughthe basic unsubstitutedsystemis not truly photochromic,the substitutedderivatives,such as aryl-substitutedhexatrienes,bianthrones,spiropyransand fulgides,are certainlythe mostimportantphotochromic molecules.With the developmentof spirooxazines,new very photostable(lightfast) photochromicswere developed, extremelyinterestingfor application(Chap. 10 and 24). Spiropyranshave been and stillare probablythe most intensivelystudiedphotochromic molecules.The knownsystemhas been enlargedincludinga numberof heterocyclic ringsas partialstructuresof spiropyrans.Several systemsusingphotochromic spiropyransfor variousapplicationssuch as read and writesystemsin the fieldof information storagehave been developed(Chaps. 23-29) and show interestingpropertiesin application. New photochromic moleculesare the spirooxazines.They were discoveredto be photochromicin 1961 but only in the ? 9eigthiesthey receiveda broaderinterest.Spirooxazines are more fatigueresistentor fatiguefree than the structurallyrelatedspiropyrans. The synthesisof thisclassof compoundshas been describedas well as some of their
11
properties.The basicphotophysical and photochemicalprocesseshave been studied only scarcely. Similarremarkscan be made with regardto the fulgideswhichare based on work of Stobbe. But recentlyfatigueresistantderivativeshave been developed, whichhave a highpotentialfor variousapplications. 1,7-, 1,8- and 1,9- electrocyclization has been shownto be in some cases reversible. However,no reallyphotochromic systemhas been reported, based on these reactions. There seems to be stillsome for furtherwork to be done in thisfield. (2 + 2) or (4 +4) Cycloadditionscan be a structuralbasisfor a bimolecularphotochro-
+
createsonly smallcolor mic system. Howeverin simplemolecules,(2 2)-cycloaddition changes,interestingare new systemsusingeither energytransferor electrontransfer processesthus producingmore importantcolordifferences.Some (4 + 4) intramolecular cycloadditions in bichromophoric systemsexhibitlarge changesof physicalproperties whichcan be of advantage.Practicalphotochromicsystemsbased on cycloadditions are more promisingin the solidstate or in polymer-bound assembliesas recentwork demonstrates. The (4 + 2) cycloaddition of singletoxygen' 0 2 to aromaticcompoundsis a new and highlyefficientphotochromic system. Pointsto be improvedare solubilityand photostability.These problemsmightbe solvedby introducing solubilizinggroupsand structuralmodificationsin the molecules. The fundamentalprocessesin moleculesundergoingphotoinduced protontransferare understoodnow quitewell. The protontransferoccursin the excitedstate and is followed by fast radiationlessdeactivationprocessesresultingin vibrationsor librationsof the molecule.Picosecondlaser studiescouldgive in the future more insighton state distributions and a more detailedpictureof the decay modesof excitedstatesinvolved. Anilsof salicylicaldehydesand their heteroanaloguesshow bothphoto-and thermochromism.Picosecondlaser studiesrevealedthe detailsof the photophysical processes in these systems.Solutionand solidstate studiesare stillgivingriseto interesting results. Studieswith triarylmethanesindicatethat photoinitiated heterolysisas well as homolysis is possible.The mechanismof the photoreaction of the old classof tetrachloronaphthalenes is stilldiscussedin controversy.In the fieldof cationicpolymethinedyes, researchactivityhas been very limited.An exceptionis Stenhouse'ssalt - an aryl polymethine dye - showinginversephotochromism.
12
Biologicalphotoreceptorsa) triggerphotomovement,b) controlplantgrowthand c) are responsiblefor the visualprocess.The secondclassof biologicalmoleculesincludes energyconvertingsystemssuchas bacteriorhodopsin and the proteinchlorophyll complexes. The biologicalphotoreceptors are photochromic.However,the pigmentsisolatedfrom these living systemspossessthis propertyin a few cases only. So here the living complexsystemmustbe regardedas photochromic.With regardto this restrictionphotosynthesis- the energyconversionin plants- has beenexcludedfromthisbook. The visualpigmentsof molluscsand mammals,althoughbeingvery differentfromeach other, are basicallyvery similar,namely,chromoproteins usingretinalsas photosensitive moiety.The resultsobtainedin recentyears are mainlydue to the developmentof time resolvedspectroscopy.The studiesundertakentoday may even lead to an understandingof how these photoreceptors developed.
Photochromism of spiro-and azo-compoundsin polymericenvironmentrevealsthat these supramolecularaggregatesshow propertiesdependingon the highviscosityand non-uniformdistribution of free volumein the polymermatrix.Thus conformational changesin the macromolecules can be induced,leadingto crosslinking,aggregationor crystallizationof the macromolecules.These photomechanical or mechano-optical effectscan be exploitedin application. Orientationinducedphotochromism,as in octaethylporphyrin, solidmatricesor NadopedKCI-singlecrystalsis a novelphotochromic systemwithoutundergoingstructural transformationsin molecules.It is a photo-induced changein the alignmentproducing lineardichroism, associatedwith birefringence.The systems foundso far exhibitthis phenomenononly at low temperatureand show a nottoo longlivedmemoryeffect. These two propertiesshouldbe furtherimprovedin futuresystems. The applicationsof photochromic systemshave been intensivelyinvestigatedin the last 20 years. The classicaluses of photochromic systemsare to a largeextent based on intrinsicmolecularpropertiessuchas changesin color, refractiveindex,dielectricconstant,enthalphyand so on. The systemsdevelopedrecentlyinvolvesupramolecularor polymericassemblieswhere the photoinduced alterationscan triggerchangesin micro environment.Thus in additionto the chemicalconversionsmodificationsin the physical propertiesof suchassembliesin initiated.These new directionsof basicand appliedresearchseems to be very promisingin the future.
13
Glassesdopedwith silverhalidesand cuprousions can becomephotochromic after thermalactivitation. The colorchangesoccurringcan extendfromthe uv to the near ir regionthus covering a rather broadpart of the spectrum.Althoughthe photochemicalbasisof the processes responsiblecan be understoodon an energy banddiagram,this modelshouldbe testedin furtherstudies.A photochemicalreverseor bleachingeffect is also observed in these glasses, howevera definitemechanismfor thisreactionis stillmissing. A new phenomenonwhichcan be classifiedas photochromism,is the spectralhole burningtechnique.At cryogenictemperatures,this new type of "energyselective"photochromismis observedwith photostablemolecules.Narrowbandopticalexcitationmay changethe.micro-environment of a guest moleculein crystallinematricesor in amorphoushost. The propertychangesinducedby hole burningopenthe way to application in the fieldof data storageor holographicimagestorage. Thus one can s,ummarizethe progressin the fieldfor the period1970-1988 as follows: - the knownphotochromicsystemsare understoodbetterand improvedby employing new techniquesof investigation;Some of these systemshave been extended. - new photochromicmoleculeshave been discovered - supramolecularphotochromismbeinga rapidlyexpandingfield allowsnew and promisingapplications.
14
REFERENCES
la M.Fritsche, Comp.Rend., 69 (1867) 1035. I b E. ter Meer, Ann.Chem., 181 (1876) 1. 2 T.L.Phipson, Chem.News, 43 (1881) 283. J.B.Orr, Chem.News, 44 (1881) 12. 3 W.Marckwald, Z.phys.Chem., 30 (1899) 140. 4 O.A.Neumiiller,Basis-Rbmmp,FranckhscheVerlagshandlung,Stuttgart, 1977. 5 Y.Hirshberg, Compt.Rend., 231 (1950) 903. 6 W.Wislicenus,Ann.Chem., 277 (1893). 7 H.Biltz, Ann.Chem., 305 (1899) 170; 2.Phys.Chem. 30 (1899) 527. H.Biltz, A.Wienands, Ann.Chem., 308 (1899) 1. 8 H.Stobbe, Verhandl. sachs. Akad. d. Wiss., Leipzig,74 (1922) 161; Chem.Abstr., 17 (1923) 3020. 9 L.Chalkley,Chem.Rev., 6.(1929) 217. 10 S.S.Bhatanagar, P.L.Kapur, M.S.Hashmi, J.lndianChem.Soc. 15 (1938) 573. 11 J. van Overbeck, Botan.Rev., 5 (1939) 655. 12 G.H.Brown, WrightAir DevelopmentCenter, Techn. Report, 59 (1959). 13 G.H.Brown, W.G.Shaw, Rev.Pure AppLChem., 11 (1961) 2. 14 L.Harris,J.Kaminsky,R.G.Simard, J.Amer.Chem.Soc., 57 (1935) 1151. 15 J.G.Calvert, H.J.Rechen, J.Amer.Chem.Soc., 74 (1952) 2101. 16 C.V.Georghiu, V.Matei, Bull.Soc.Chim., 6 (1939) 1324. 17 C.V.Georghiu, B.Arrventien,Bull.Soc.Chim.,47 (1930) 105. 18 C.V.Georghiu, Bull.Soc.Chim., 1 (1934) 97. 19 R.Dessauer, J.P.Paris, Advancesin Photochemistry,Vol. 1. W.A.Noyes, G.S.Hammond, J.N.Pitts, Eds., P. 275ff, Interscience,New York, 1963. 20 G.H.Brown, Ed. Photochromism,TechniquesChem., Wiley Interscience, New York, 1971. 21 G.Porter, Angew.Chem., 80 (1968), 882; Int. Ed. 7 (1968). R.G.W. Norrishand G.Porter, Nature (London),164 (1949) 658; G.Porter, Proc.Roy.Soc. (London)Ser. A. 200 (1950) 284. 22 C.Br&uchle,U.P.Wild, D.W.Barland, G.C.Bjorklund,D.C.Alvarrez, IBM Res.Develop.,26 (1982) 217.
15
Chapter 2
1
Photophysical, Photochemical and Photokinetic Properties of Photochromic Systems G.Gauglitz
PHOTOPHYSICAL PROPERTIES
Electromagnetic radiation can be considered as a wave as well as a particle according to de Broglie’s theorem of dualism. Either of these can be used to explsjn the specific behaviour of the radiation interacting.with molecules or optical systems. By use of both the models, the linear momentum p and the wavelength X can be related according to
h X
p = - = m . c,
(Planck’s constant h = 6.63. 10-34Js)
m being the mass in kg and c the velocity of light in m s-’. Using Einstein’s equation
the energy E in J can be correhted to the frequency v (s-’, Hz),the wavelength A (nm), and the wavenumber fi = 1 / X (cm-l) (refs. 1,2). In principle, photochemical reactions can be caused by visible (380-780 nm) or ultraviolet (200-380 nm) radiation. Frequently mercury arc lamps are used in this wavelength range (ref. 3). 1.1 Interaction between radiation and matter
In any case, electromagnetic radiation passing a medium shows dispersion, that means the index of refraction depends on wavelength (refs. 4 - 6). The electromagnetic wave polarizes the molecule. In molecules without dipole moment the incident vector E of the electromagnetic wave can induce a momentum, which alternates with the frequency of the radiation. Molecules with permanent dipole moment are oriented opposite to the incident vector B of the electromagnetic wave. It is diminished in the medium by both the effects. Therefore the index of refraction is larger than 1 (ref. 7). At certain wavelengths (resonance: frequency of incident light fits with the distance of the energy levels) the electron density distribution can be permanently changed by the interaction of radiation with the molecules. The molecules are excited and the intensity of radiation is reduced (absorption). In organic molecules T--A* or n-T*-transitions are normal. a-a*-Transitions cause breaking of the bond between C-atoms (dissociation) (refs. 2,5,8). A change of solvent influences these energy levels differently (ref. 2). Increasing interaction between the orbitals and the solvent molecules lowers the energy of the states of the molecule. In the following sequence the r - , ~ * and - n-orbitals give increasing interaction with the solvent by their larger extension in the surrounding area. The more polar the solvent is the more it influences these electron density distributions and decreases the electronic energy of the states (refs. 2,7). For this reason polar solvents cause a red shift (bathochromic) for r - T*- and a blue shift (hypsochromic) for n - r*-transitions.
16
Fig. 1: Interaction between electromagnetic radiation and molecules, (a) causing a decrease in amplitude of the radiation (intensity), (b) changing the electron density distribution, (c) giving a new electron distribution between the two states of energy.
The latter are less intense (absorptivities E,-,.: only 100- 10000) than x - x'-transitions (E,-,. : 10 000 - 100 000). The reason is the smaller probability for electronic transitions between n- and A*- orbitals in comparison to x- and x*- orbitals, because of their different properties of orbital symmetry relative to each other.
As demonstrated in Fig. 1 an interaction between the resonance frequency of electromagnetic waves and these orbitals causes a new distribution of electron density (refs. 2,9). The energy absorbed is taken from the electromagnetic wave. Its amplitude is damped by the sample. Furthermore any transition between the two states of energy causes a relative change in distribution of molecules in the ground and the electronically excited state. Each photon absorbed can excite one molecule at maximum, the number of particles in the electronic ground state decreases, but increases in the excited state (refs. 2,4,6,10). The stationary energy of an orbital (specific electron density distribution) can be calculated by a quantum mechanical linear equation, the Schrodinger equation. These stationary states of energy are of great interest in photochemistry. In principle the excited states are only quasi-stationary, since their lifetime is not infinite. The longer this lifetime the better defined is its energy. This correlation between lifetime and uncertainty of energy is given by Heisenberg's law Ii
AEaAt 2 2
.
(3)
The more a particle in an excited state can contact with others or dissipate internally its electronic energy to translational, vibrational or rotational energy, the shorter it lives in this excited state. Instead of a distinct energy a broad band will be obtained. This uncertainty of energy therefore correlates with the uncertainty of the lifetime of this excited state (refs. 1,2,10,11). The time-dependencehas to be introduced in the Schrodinger equation, if these transitions between different states of energy have to be described.
All time-dependent processes starting at this excitrd electronic state (emission, radiationless deactivation, intersystem crossing to other electronic configurations) are concurrent to the photochemical process. For this rea.son qwmtum yields smaller than one are found
17
3.3). Therefore an understanding of these concurrent photophysical deactivation pathways is necessary to obtain high photochemical quantum yields and selective production of photoproducts, that means it is necessary to guide the reaction.
(s. section
Spontaneous as well as induced emission are processes opposite to absorption. In both cases the electron density distribution of the ground state orbital is obtained, a photon is emitted, and the relative distribution between those molecules in the excited and the ground state is changed. However, there is a very significant difference between the processes: In the case of spontaneous emission the photon is emitted statistically in any direction at any time within the lifetime of the excited state (Fig. 2).
1-
8
E = hu
Fig. 2: Spontaneous emission, statistically in any direction and at any time within the lifetime of the excited state: (a) change in electron density, (b) emitting dipoles, (c) spatial distribution of emission, (d) transition between states.
In the case of induced emission the inducing photon of the incident electromagnetic wave causes an excited molecule to deactivate by emission of a second photon. The two photons couple to each other (Fig. 3). The result is an amplification of the incident intensity of radiation, the emitted photon having the same direction and phase as the incident one. This type of process is the fundamental of a LASER (Light Amplification by Stimulated Emission of Radiation) (refs. 2,7,11). 1.2 Jablonski Diagram
The actual energy of a molecule is the sum of the following contributions 0 0
the electron density distribution (electronic energy), the vibration with respect to the center of gravity (vibrational energy), and/or the rotation of the molecule with respect to this center (rotational energy).
Many vibrational and rota,tional levels belong to each electronic level (Fig. 4) (refs. 2,ll). The f i s t ones are marked by v, the latter ones by J . Two term systems are to be distinguished:
18
Fig. 3: Induced emission of radiation, coherent induction of a transition of a molecule from the excited state by electromagnetic radiation, (a) change in electron density, (b) emitting dipoles, (c) amplification of radiation, (d) both photons being coherent with respect to direction and frequency of emission. 0
0
The singlet system (S), which is represented by antiparallel spins (t 1)in bonding electronic levels. According to Pauli's principle (ref. ll), these two spins must have opposite directions. Normal organic molecules have this singlet state as their electronic ground state. It is symbolized by So. The triplet system (T), which is characterized by parallel spins (It). According to Pauli's principle both these spins have to be distributed into different states of energy. In general, organic molecules do not have degenerate levels of energy and no triplet ground state To exists. The triplet state 2 ' 1 is lower in general than the corresponding singlet state S,, because of the smaller overlap of the parallel spins. Quite a few inorganic molecules have degenerate electron energy levels. These have to be filled at first by one electron each according to Hund's law causing a very low energy (ref. 11). This system is a triplet TOone and represents the electronic ground state, by which inorganic transition metal complexes are characterized.
In Fig. 4 singlet and triplet terms are drawn side by side for graphical convenience. In reality, this type of term diagram contains only one axis, the energy axis. In the diagram a possible transition to an excited state by absorption of radiation in the UV or visible yields an energy term marked A'. Excitation by irradiation (a) lasts s and correlates to the frequency Y of the electromagnetic radiation. The obtained state A' is very unstable and deactivates thermally very fast to the ground state of vibration in the excited electronic state S1 (A'). In principle two processes of thermal deactivation have to be distinguished (refs. 2,4,5): 1. Energy deactivation within the same electron density distribution ( b , d ) , which is called vibrational relaxation (thermal equilibration) and the 2. isoenergetic energy conversion by change of the electron density distribution called internal conversion, IC (c, e, m), which results in a reorganization of the electron density distribution giving the electronic ground state, but combined with a very highly excited vibrational state.
19
Fig. 4: Jablonski diagram of rotational, vibrational, and electronic states of energy for two spin systems drawn schematically (rotation energy levels do not have equal distance).
-
In contrast to A' and Sz the level S1 has a relatively long lifetime (lo-" t o 10-7~).For this reason four concurrent processes can take place, starting from it: 1. further radiationless deactivation (f) (vibrational relaxation), 2. emission of radiation (luminescence. In the case of S1 + So : fluorescence (g)), 3. conversion to a triplet (intersystem crossing, ISC (h)) with successive deactivation ( i ) to TI (A"), and subsequent deactivation (m, u) or luminescence (phosphorescence (l)), 4. a photoreaction by change of the molecular structure, starting from the singlet S1 as well as after intersystem crossing from the triplet TI.
In general the process with the relatively highest rate will take place in dependence on the environment of the molecule (solvent, catalyst/surface, tenside aggregation, temperature, pressure). If the photoreaction is fast enough to overcome the other photophysical processes it must happen within the lifetime of the excited singlet or triplet state. 1.3 Energy transfer
Besides the mentioned processes of energy transfcr by conversion of translational, vibrational and rotational energy &one molecule, those bimolecular processes are very
20
essential in photochemistry which include the transfer of electronic energy from donor to acceptor molecules. In general, numerous mechanisms are discussed, but they can be summarized under two principle energy transfer processes (refs. 4,5,12):
1. Coulomb interaction: Is discussed as an induced dipole mechanism. The changing dipole of the donor induces another dipole in the acceptor without physical contact between the molecules. This process of induced resonance takes place over larger distances (up to 10 nm). A classical analogy to interaction between charged particles is possible. Therefore it is sometimes called “transmitter/antenna-mechanism” (ref. 5). Although the coulombic interaction does not need a collision, Forster demonstrated that the dipoledipole interaction depends on the oscillator strength for the radiative transitions in the donor and acceptor molecules (refs. 13,14). Thus small Fkank-Condon-factors or changes in multiplicity lead to a small interaction energy. The energy transfer is proportional to a) the square of the transition dipole moment of donor as well as acceptor and b) the inverse of the sixth power of the distance between excited donor and the ground state of the acceptor. Further examination shows that the rate of energy transfer depends on the overlap integral of the experimental absorption and emission curves taking into account the fluorescence intensity and the absorptivity in the related region of wavelength. Furthermore, since dipole-dipole interaction depends on the relative orientation, in rigid solvents this orientation has to be considered, too.
2. Exchange interaction: Implies a collision between the molecules, the overlap of the electronic density distributions gives a kind of transfer of momentum. According to a theory by Dexter (ref. 15) an exponential dependence of energy transfer is assumed on the ratio of the donor/acceptor separation r relative to their Van-der-Waals radii 1. The constant of proportionality is given by the specific orbital interactions and the normalized overlap integral. In contrast to coulombic interaction the rate is predicted to be independent of the absorption characteristics of the acceptor. Comparing the theories of FGrster and Dexter one c a n predict the following differences:
1. The rate of dipole/dipole-induced energy transfer decreases with a distance of
T - ~
in comparison to exchange induced transfer which decreases with exp(-r/l). This exchange drops to negligibly small values within one or two diameters of the molecule (500 pm - 1 nm at maximum). The dipole induced transfer depends on both the oscillator strengths of donor and acceptor radiative transitions in contrast to the rate of the exchange transfer, which is independent of them. 2. The efficiency of an energy transfer by dipole mechanism depends to a large extent on the fluorescence quantum yield of the donor. This is not the case for the exchange interaction. Since the collisional exchange interaction is a diffusion-controlled process, it depends on the viscosity.of the medium. One-step and several-step mechanisms are discussed. Both cause the exothermic or endot,hermicenergetics to influence the transfer rate appreciably.
21
Four types of spin-distribution of donor and acceptor for energy transfer can be considered: either singlet/singlet, or triplet/triplet, or triplet/singlet or singlet/triplet energy transfer reactions. They are allowed to a different extent for the two mechanisms of COUlomb or exchange interaction. Only in the case of exchange interaction all four types of transition are possible. In any case only singlet/singlet interaction happens to a large extent. The values of distance are in general less than 1.5 nm. Since a triplet/triplet transition is a forbidden dipole transition, it is also spin-allowed only for an exchange mechanism. That means the only relevant energy transfer via the Coulomb-interaction dipole-dipole mechanism is the singiet/singlet energy transfer. Both the mechanisms are relevant in photosensitization and in quenching experiments.
2
PHOTOCHEMICAL PROPERTIES
2.1 Photochemical primary processes
Molecules in the excited energy states S, or Tl can show a large number of different photochemical primary processes (refs. 4,16,17): A
+ hv
4
photoproducts
A* + A" -+
In general photochemical processes are to be considered as pseudo first order reactions. The molecules of one species are excited by the photons. The rate law of the first step of the photochemical processes, following the photophysical processes, is considered to be first order. After absorption of energy the molecule deactivates very fast from the higher excited state A* to the first excited electronic states (either Sl or 2'1) according to the photophysically most preferable pathway. The subsequent primary photochemical processes may be classified into four different groups: 1. Spontaneous monomolecular deactivation processes 0
into two radicals A" -+ B. e.g. Brz -+ Br . Br .
+
+ C.,
into two stable molecules A" e.g. CHz0" CO Hz
+
0
0
-+
B
+ C,
+
into two ions A" -+ A+ C-, e.g. leuconitrile of triarylmethanes into a positive ion and an electron A" e.g. triphenylamine in n-hexane
2. Rearrangements
A"-+B, e.g. trans-cis-photoisomerization
3. Reactions with other reactants
+
A" B(A) -+ C, e.g. photoaddition, photodimerization
-+
A+
+ e-,
22
+
A" RH2 + A H + RH, e.g. photoreduction
4. Photopolymerization A"
+
A2 + AS.
Which type of reaction takes place is determined by preceding photophysical processes, the partner of the reaction, the excited molecule itself, and the solvent. In general the photoreaction takes different pathways, depending on whether the reaction starts from the singlet or the triplet system (Sl, Tl).
2.2 Bimolecular processes The above mentioned bimolecular reactions (types 3 or 4) can take place either from the singlet or from the triplet state. The reaction only happens if the excited molecule A" collides with the non-excited molecule within its lifetime. Assuming that this collisional process is diffusion-controlled, for which a rate constant of k = 10" 1 mol-' s-* can be expected at maximum, one is able to calculate the probability for both the processes to take place. The lifetime of the excited singlet state A' is of the order of magnitude of lo-' s. Within this time the partners of collision have to meet with the above mentioned rate constant. For this reason concentrations of at least mol I-' can be calculated for the collision partner. If the excited molecule has a triplet energy electron density distribution (A"), its 3). Therefore a smaller necessary concentration can lifetime is longer (approximately be calculated for the partner of collision B of only lo-' mol I-'. The shorter the lifetime of the excited states the higher the concentration of the partner has to be. Otherwise diffusion carries too few partners close to the excited molecule.
2.3 Diabatic and adiabatic processes In general photophysical processes cause a change in the electronic density distribution (section 1.2). In contrast the primary photochemical processes cause a change of the relative arrangement of the atoms in the molecules (section 2.1). The changes in atomic distances can be correlated to a potential energy. The two-dimensional graph visualizes a specific 'reaction coordinate' as part of the 'potential energy surface'. Whereas thermal reactions stay on the ground potential energy surface, photochemical reactions can take place from higher ones in different ways. In Fig. 5 three classes of photochemical reactions are plotted. Class I and I11 are diabatic reactions and class I1 is an adiabatic one (ref. 5 ) . The mentioned terms adiabatic and diabatic depend on the pathway the reaction takes. If the photoreaction completely remains on one surface, it is called adiabatic. If the energy surface is changed during the reaction pathway, this reaction is called diabatic (ref. 18). The pathway taken during the photoreaction greatly depends on the form of both the energy surfaces and on the energy, the molecule has started with. A special case is the
23
I
II
m
Fig. 5: Class I: Diabatic photoreactions. Class 11: Adiabatic photoreactions. Class Ill: “Hot” ground state photoreactions.
“hot” ground state photochemistry given by class I11 in Fig. 5. This type of photoreaction preferably occurs in gas phase reactions at lower pressure. In this case the excited molecule “jumps”to the ground surface with a large excess of thermal energy. This is enough to overcome the potential energy barrier between the reactant and the product.
If the potential energy surfaces in the ground and the excited states approach each other at any relative atomic arrangement the energy difference between both the states is relatively small and a “jump” between these two surfaces becomes possible. This diabatic type of photochemical reaction is the most common one in organic photoreactions. Depending on the molecular properties the potential energy surfaces can cross or avoid crossing. The non-crossing rule is less restrictive for polyatomic molecules than for diatomic ones. For this reason in normal organic molecules the potential energy diagrams actually look more like states with avoided crossing than intersections. Which one of the three types of photoreaction discussed takes place depends on 0
0 0
the locations of the minima and maxima of both the potential energy surfaces given by either So, S, or TI states; the separation of the two energy surfaces, which gives more or less interaction, and the geometries at those internuclear distances, where the surfaces come close to each other.
It is evident that a very high energy in the excited state allows the molecule to take internal conversion back to the ground state before overcoming this barrier. In this case we find the typical procedure of a “hot” ground state class I11 photoreaction which is called diabatic or adiabatic depending on the point of view of the authors. If one only considers that part of the photoreaction that takes place after the photophysical processes, just a ground state pathway can be taken and the process can be called adiabatic (ref.
24
18). On the other hand other authors consider the overall process. They take the internal conversion with the rearrangement of the electron density distribution and the “jump” from the higher to the lower potential energy surface as part of the total photochemical process. In this case “ h ~ t ground ” state photoreactions can be taken as diabatic (ref. 5).
Some of the conditions for an adiabatic class I1 photoreaction seem to be more or less trivial. This reaction occurs adiabatically on the higher potential energy surface without cross-over to another one. This higher surface should have a minimum at the product conformation which is lower or at least not essentially higher than that at reactant configuration - if there is one. Besides, the energy barrier between both minima should not be higher than some kT. To make sure that no cross-over takes place, the potential surfaces must be unfavorable for interconversion in the region between reactant and product configuration. This can be avoided if the energy barrier between the two configurations is not too flat or even does not allow intermediate states. In this case the time spent in this region is too short to allow interconversion. The larger the distance between the two energy surfaces the less likely is this interconversion. This implies that the energy barrier in the ground state is relatively low, too. Hence such types of photoreaction are accompanied by a relatively fast thermal back-reaction. Typically photoisomerizations are discussed in the literature with respect to either an adiabatic or a diabatic pathway of reaction. Especially in the case of stilbene a long discussion has taken place. The different possible photochemical pathways decide about the properties of the photochromic systems - especially whether pure photochemical or superimposed thermal reactions occur (refs. 19 - 21). PHOTOKINETICS APPLIED TO PHOTOCHROMISM
3
3.1 Principles of photokinetic examinations
In principle photochemical reactions aim at the following: 1. as much turnover as possible, 2. as much yield as possible, 3. as short time for the reaction as possible, and 4. as few by-products as possible.
Usually it is very difficult to achieve all these aims. For this reason the reaction procedure has to be examined and optimized with regard to different points of view. The most essential ones are discussed in the following (ref. 22): 0
Dependence on the absorption of the initial product Photochemical reactions only can take place when the reactant absorbs photons. Hence the absorption spectrum of the initial product has to be known, to find an optimal wavelength of irradiation in different solvents. Since excitation of ?r - a’ or n - a* transitions can cause different pathways of photoreaction the nature of
25
the absorption bands has to be determined. Furthermore it is necessary to know absorption spectra of intermediates or final products, because otherwise a bad choice of the wavelength of radiation causes degradation or unwanted side reactions. Choice of the solvent The solvent has to be transparent at the wavelength of irradiation. Furthermore any additional products have to be avoided, formed by reactions between the solvent and initial or final products as well as intermediates. Since the pathways of reaction can depend on the polarity of the solvent, the choice has to be careful. 0
Choice of the light source In general immersion lamps are used. Irradiation has not to be monochromatic, since in solution it is unnecessary to select a single wavelength from the continuum. The interaction between the solvent and the solute will cause broad bands of absorption. If arc lamps are used, the different lines can be selected by use of liquid or glass filter material. Frequently only the UV-part of the radiation is cut off. This could be done by use of Duran-glass which absorbs radiation at wavelength X 5 300 nm. TOavoid thermal effects, the lamps are frequently surrounded by a second glass cylinder with a filter solution. By this means, distinct wavelength ranges can be selected. In some cases, it is necessary to define the wavelength of irradiation very exactly. Photoisomerizations are a good example. Their equilibrium is influenced by the wavelength of irradiation. Even though photochemical quantum yields do not usually depend on wavelengths within one band of absorption, the amount of absorbed light is responsible for the rate of the reaction. Therefore absorption coefficients are essential factors. They differ for both the isomers in dependence on wavelength. For this reason the rate of the approach of the photostationary state (equilibrium between the two photoisomers) as well as the overall yield depend on the irradiation wavelength. One industrial application of monochromatic irradiation is known for the vitamin D synthesis. In such cases the use of LASE& instead of mercury arc lamps is discussed (ref. 23). Choice of concentration According to Lambert-Beer's law the number of photons absorbed by a sample is proportional to a) the optical pathway d and the number of absorbing molecules (concentration a ) and b) the constant of proportionality for each molecule.
6
(absorptivity), where t is a property specific
One finds an exponential dependence for the transmitted light I d relatively to the incident light intensity I,,(section 3.2). If the concentration of the sample is chosen too low, only a part of the incident light is absorbed. The remainder is lost for the photoreaction. On the other hand, if too high a concentration is chosen, the light is totally absorbed in a very small layer next to the light source. In this layer the photoproducts are locally enriched causing unwanted degradation reactions. In
26
some cases even the glass of the immersion lamp becomes coated by products of polymerization or tar. Furthermore different concentrations may yield different pathways of reaction. For example high concentrations support the formation of dimers. Hence a small concentration at high pathlengths must sometimes be chosen to avoid this formation of dimers. 0
0
Monitoring of the reaction In any photokinetic treatment the reaction procedure has to be monitored very efficiently. The method chosen must be very fast as well as exact. Usually absorption spectroscopy in the ultraviolet and visible meets both prerequisites. It is therefore the method of choice in photokinetics (section 3.5). Procedure of the reaction To avoid local inhomogenities in the sample effectivestirring is necessary. Otherwise local concentrations of photoproducts could become too high. Unexpected consecutive products can result. In the case of solids, irradiation of photochromic systems in polymers, or irradiation by LASER light sources yields a gradient of concentration of the different compounds. It has to be taken into account in the evaluation ( s . section 3.9). The extent of this gradient depends on the absorption at the irradiation wavelength. Using a LASER it is frequently advisable to reduce the intensity of the radiation to avoid either multiphoton processes or interactions between excited molecules. Furthermore the following points have to be considered: - The reactor has to be temperature-controlled to avoid overheating. - In the case of sensitized reactions quenching has to be excluded. That means in
some cases the reaction has to be performed under exclusion of oxygen. - Irradiation should be stopped before total conversion to avoid coating on the light source or consecutive products, if molecules tend to photodegrade.
Any photokinetic examination has two aims, a qualitative and a quantitative one. The proposed procedure of examination allows unwanted by-products to be found, steps of the process to be modified, the reaction pathway to be optimized, and the values of the rate constants of the reaction to be determined. These give the partial yields of the different steps and the overall yield for the total reaction. The qualitative part of the examination of the reaction is a combination of the intuition of the chemist and some methods of graphical and/or numerical determination of the number of steps, the number of participating products and their reaction mechanisms. This means that a simple determination of the final products and an assumption of the overall mechanism is not enough. Physicochemical methods such as UV-, CD-, IR-, NMR-, fluorescence and mass spectroscopy are necessary to follow the pathway of the reaction. Further, the complex combination of reactants has to be separated by methods such as gas or liquid chromatography, to give information about the “history1’ of the different reactants. In Fig. 6 a
27
kinetic analysis:
start:
mechanism reactionconstants
numberof steps numberof reactants thermalrate constants partialphotochemical quantum yield
[ unknown reaction } L
I
recordingof a reactionspectrum absorptionfluorescence
reactionconstants and standarddeviations \
simulated
E t -curves -error limited
by use of simulated '*theoretical data"
-
mechanismonlyas longvalida s no better has been found
Fig. 6: Scheme of photokineticanalysisto obtain mechanisticinformationand reaction constants from experimentaldata by modellingand simulation.
28
general block scheme demonstrates the considerations and the method of the examinations for elucidating the mechanism of the reaction and calculating quantitative constants (refs. 24,26). In general the most efficient form of analytical method has to be chosen. But it has to be considered that exact values of concentration at as many as possible reaction times have to be obtained for a quantitative evaluation. Both the qualitative as well as the quantitative aspect is discussed in later chapters in more detail.
3.2 Amount of absorbed light Knowledge about the amount of absorbed light is the prerequisite of all quantitative examinations in photokinetics. It can be obtained from Lambert-Beer’s law. In Fig. 7 the given exponential dependence between the incident and the transmitted intensity is plotted schematically for liquids (ref. 2). The derivation is based on the assumptions, that the
0
d
X --I
-1
dx
Fig. 7: Exponential decay of light intensity in absorbing matter
light incident to the sample 0
is parallel and falls homogeneously onto the total front area of the sample.
29
Selecting a small volume element in the sample A with concentration a and with the pathlength d x at the position x , the intensity decreases in this volume element according to (refs. .2,7)
This equation gives the intensity of light I, incident on the front area of this volume element. K. is the natural absorptivity (a proportionality factor) and a the concentration of the absorbing particles. By integration of this differential equation within the limits of the pathlengths 0 and d either Id = 1 0 . e-*ad
with natural absorptivity
(5)
IE
or
Id = 10 . lo-'
with decadic absorptivity
ad
E
(6)
are obtained. This derivation applies to one component absorbing in the solution. Since the proportionality coefficient is a value which depends on the wavelength, the LambertBeer law is only valid for monochromatic radiation. Besides, this law is limited to diluted mol P ) . solutions (normally up to The decrease in intensity is caused by all components absorbing in the solution. For this reason the amount of absorbed light in the solution by all particles is given by
ItbS= I0 - I, = I,. (1 - 10-8')
(7)
in the units mol photons crn-*s-l, where E' defines the absorbance as a sum of all components' absorbances (n different absorbing samples, including the solvent). This absorbance at the wavelength of irradiation A' is given by
i=l
Normally only one component is photophysically excited and starts the photochemical process. Therefore its absorbance ( d . ~ :. a ) has to be taken relative to the total absorbance (E'). One finds the following relationship for the amount of absorbed light by component A, which starts the photoreaction, to be:
I L A
=
lo. (1 - 10-E') . d . E L . a d.
c&: .
i=n
ni
i=l
(9)
30
In this equation the intensity of irradiation is used in the units mol photons per area and second. The unit “mol of photons” is sometimes called Einstein. Assuming a homogeneods1y stirred solution the exponential decay of irradiation intensity can be averaged for each volume element of the solution by an additional factor l / d . Thus I,&A is expressed . s-’ as by units mol
taking into account that the amount of light is absorbed per volume. Since usually the concentrations a; are taken in units of mol I-’ and the irradiation intensity in mol cm-2 s-’ a factor of 1000 cm3I-’ has to be used to combine these two units. Therefore the “corrected intensity”
I
= 1000.I0
mol photons . cm . s-’. I-’
(11)
is obtained, which effects all photochemical reactions. In the equation derived above, the expression
F(t)=
1 - 10-E’ E’
is called the “photokinetic factor”. It takes into account that only some part of the totally absorbed amount of light induces photoreaction. Using LASE& as light sources or working in polymers the derived amount of light absorbed can depend on the volume element (varies with penetration depth 5 in the sample). This fact has to be taken into account in the calculation of quantum yields. It will be discussed later in detail (section 3.9). Next, solutions are be considered which can be perfectly homogenized by stirring during irradiation with respect to all the volume elements (refs. 16,26).
3.3 Quantum yields In preparative photochemistry the quantum yield usually is understood as the product yield of a reaction. It is calculated by the number of reacting moles of reactant per unit time. This type of quantum yield has been preferred, since it can be measured with small expenditure. Its disadvantage is that it depends on the time section of the reaction in which it is determined. For this reason it can be considered to be an apparent quantum yield. The problem becomes evident when the reaction consists of several steps. Under these conditions the determined apparent quantum yields certainly depend on the time at which they are determined during the reaction. Recently photochemical quantum yields have been redefined by the IUPAC (ref. 27) with respect to above considerations. An integral and differential photochemical quantum yield is distinguished. The general symbol Q, is used. The differential quantum yield of interest is
31
where d [ s ] / d tis the rate of change of a measurable quantity, and n the amount of photons (mol or its equivalent Einsteins) absorbed per unit time. In photokinetics photochemical quantum yields have to be discussed in more detail. It has been derived in eq. (10) that in many cases only one of the absorbing reactants undergoes the interesting photophysical excitation which starts the photochemical reaction. Therefore the photokinetic factor has been defined. For this reason the following discussion seems necessary. In contrast to apparent photochemical quantum yields a true one depends neither on the moment at which it is determined nor on the absorption of the solution (ref. 16). It is obvious that such true quantum yields have to be determined for each step, reflecting the ideas mentioned in the chapter above about the dependence of a reaction rate on the amount of light absorbed. ! h e and apparent quantum yields differ depending on momentary absorption of the solution and the amount of the absorption of the reactant, which starts the photoreaction. Another distinction can be seen between the so-called integral and differential quantum yields. The first ones, integrating the number of photons absorbed by the solution over a longer time give a large change in concentration. On the other hand the differential ones assume an infinitesimally small change in concentration. In general the quantum yield is a ratio between the change of concentration of the reactant of interest and the number of photons incident on and absorbed by the reactant during the measured time. Considering the discussion in the section above, only those photons cause a reaction which are absorbed by the reactant starting the photoprimary step. Hence any definition using apparent quantum yields gives wrong and time-dependent results. In general, according to the literature, five different types of photochemical quantum yields cap be discussed (ref. 16): 1. Apparent Integral Quantum Yield
2. True Integral Quantum Yield
3. Apparent Differential Quantum Yield
&=
f &/IT
4. True Differential Quantum Yield
32
5. Partial Quantum Yield
In these equations N gives the number of reacting molecules and N h v the number of photons emitted by the light source. The index T used for IT gives the total number of photons absorbed, whereas means the amount of light absorbed (only the absorbed photons count) by the reactant, that undergoes the photoprimary step. The partial (true differential photochemical) quantum yield is defined for each linear independent step of the reaction. It is expected that only one reactant starts the photoreaction. In the case of complicated photoreactions one can define for each of the k steps one partial photochemical quantum yield each with respect to the degree of advancement xk of this partial reaction. The degree of advancement (refs. 1,16,25) corresponds to the stoichiometric change of concentration with respect to any reactant. This last type of a partial photochemical quantum yield is the only one which quantitatively and correctly describes reactions with a complex mechanism. It is independent of the moment at which it is measured, and can be used together with the absorptivities at the wavelength of irradiation to give a very good prediction of the turnover, even in complex reactions. In principle, in any case it has to be recognizable which type of photochemical quanturn yield has been determined. Otherwise the published numerical values are meaningless. Certainly a general overall turnover for the reaction can be easily determined by measurement of product formation with time. But the correct definition of partial photochemical quantum yields involves some experimental and numeric expenditure (refs. 25,26). Nevertheless in the next chapters an idea will be given about methods to determine these partial photochemical quantum yields. 3.4 Derivation of differential equations
In photochemical reactions usually a pseud-first order is assumed. Besides, in each step of the reaction only one product is considered to start the photoprimary step. As discussed in section 1.2 a specific molecule is excited to an electronic, vibrational and/or rotational highly excited state A*, which can be the starting-point for the different photochemical primary processes discussed in section 2.1. As treated in the derivation of the amount of absorbed light and in the discussions on quantum yields, the rates of the photoreaction are proportional to the amount of light absorbed and the quantum yield of the distinct step. Photochromic systems are considered to be equilibrium reactions between two products, whereby the photoreaction causes a colour change in the solution. 1n.principle one can distinguish four types of such photochromic reactions:
33
a)
A%B
b)
Ah"'B B -%A,
c)
Ah"'.B B k3. A,
d)
A%B B-%A
BAA
The problem is that all these different types of reaction appear as a single linear independent step of a photochemical reaction from a mathematical point of view. The reason is that the steps of reaction given above depend of each other by the law of conservation of mass. Consequently these four types cannot be distinguished formally nor the different rate constants (thermal and photochemical) can be calculated in a simple way. Using the above mechanistic assumptions four types of rate law can be derived (refs. 16,26):
These four equations can be classified into two categories: 0
Those with photochemical reaction steps only and
0
those with thermal reactions superimposed on photoreactions.
The latter involve much more complicated kinetic evaluation. Even though in the equations given above the concentration of the product B has to be considered in some cases, the use of the law of conservation of mass
34 a(t)
+ b(t) = a ( 0 )
(23)
allows the substitution of b ( t ) by the initial concentration a(0) at reaction time t = 0 and the concentration of the educt a(t) (ref. 25). Therefore the following rearrangements of eq. (20) can be made: U=
- I * Q ( t ) * ( & ~ '+ c p&~; * c p f ) .
1 - 10-E'
E'
+I.&;.cp;.a(O)
.
1 - 10-E'
E'
424)
TABLE 1: Abbreviations in photokinetics. ~
F(t) =
I =
the photokinetic factor, relating the absorbed amount of light to the reactant, which starts the photoreaction
1000
I0
the intensity of the light source in molphotons cm * I-'s-l
R1 = Rz =
. pf, EL.~ p f '
R
R1
=
+ RZ
the product of partial photochemical quantum yield and the absorptivity the pseudo quantum yield
Therefore in purely reversible photoreactions the rate law can be described for the rate of concentrations' change of component A by the following equation using the abbreviations given in Table 1: Q
= -I
. R . Q ( t ). F ( t )
+ I .Rz
* Q
(0) .F ( t ) .
(25)
This equation cannot be solved in a closed form. In order to simplify this expression the condition for the equilibrium in the photostationary state (at infinite reaction time t = m) is used. Under these conditions the above equation can be rewritten to
At equilibrium there is no absolute change in the concentrations, therefore a ( m ) = 0, and I as well as F (m) can be eliminated in the above equation (refs. 16,25,26). It can be solved with respect to ~(0): a(0) =
R -
R2
*
a(..)
35
This value can be put in eq. (25), and the final equation is obtained (refs. 4,16,25,26):
This expression allows the determination of quantum yields by use of the least squares method applied to concentrations experimentally obtained. If the thermal backreaction has to be taken into account, the eqs. (21) and (22) have to be taken:
ir = - I . R 1 . a ( t ) . F ( t ) + k3(u(0) - a ( t ) )
(29)
or u =
-I.R.a(t).F(t)
+ I.Rz.a(O).F(t)+ k3(u(0) - a ( t ) ) .
(30)
In both cases the rate at equilibrium cannot be used to set up a new equation, which helps to substitute the concentration a ( 0 ) . Therefore the solution of the equation becomes more complex. For pure photochemical reactions the rate equation can be rewritten in a more generalized form using matrix and vector expressions:
The elements of the &matrix are the kinetic constants characterizing the different steps of the reaction. In this matrix representation it has been taken as certain that there is no linear relation between the different steps of the elementary reactions. Otherwise the number of equations has to be reduced by the law of conservation of mass as has been done according to eq. (23) above. Two trivial types of linear dependences between different photochemical steps of reaction exist: 0
One typical example is the reversible photoisomerization. That means that both dependent steps
A%B
and
B%A
have to be combined to one linear independent step. The consequence is - as given in line 2 of Table 2 - that the reaction constants for the two linear dependent steps form together one linear independent equation with one reaction constant, the so-called pseudo quantum yield R. 0
Another example is a pmdlel photoreaction, both the reactions having the same order of reaction:
36
In Table 2 some further examples are given. The so-called Jacobi matrices for thermal and photochemical reactions are compared. Reactions with more than one linear independent step of reaction are included in the table, because in many cases photochromic systems not only show the one step process from the educt to the photochromic product or the reverse photoreaction, but can give additional photodegradation. This problem will be considered in a later chapter.
TABLE 2: Comparison between rate constants of thermal and photochemical reactions :YP=
4 + B A=: B
. b
A - B
A+B-+C A*B-+C A-+B+C
. b
A = B
The use of matrix equations allows pure photochemical equations like eq. (31) to be solved by simple matrix operations. The measurement of the concentration with time, oi the light intensity, and of the photokinetic factor enables an evaluation of the Jacobi matrix. The given equation represents a system of linear differential equations. Formation of the so-called normal equation and taking the inverse, yields the &matrix. The evaluation gives the elements of this matrix, called the Jacobi matrix. The procedure will be shown in a later chapter (s. section 3.7). The only problem is that usually the concentrations cannot be determined directly. They have to be obtained indirectly by the measurement of a signal, which is expected to be proportional to the concentration. As discussed above a method has to be found, which allows a very exact determination of the signal and a very fast measurement. It has been stated that spectroscopy in the visible and ultraviolet is the method of choice.
37 3.5
Absorption measurement
Spectroscopy in the ultraviolet and visible wavelength range is very exact and relatively fast. Both conditions are prerequisites for using this method in photokinetics. UV/Visspectroscopy is a standard method in modern analytical applications. By use of LambertBeer’s law according to eq. (8) a linear dependence exists between the measured signal, the absorbance, and the concentration ai of the component i . However, the use of this relationship can cause two problems: the equation is only valid, if the light source of the spectrometer is monochromatic and samples containing more than one component gives an absorbance which is the sum of all the absorptions of the different reactants. That means that eq. (8) has to be rewritten for every wavelength of measurement in the form
The summation has to be taken over all different components in the solution, each having an absorptivity at the wavelength A. The mea.sured absorbance depends on the used optical pathlength d of the cell. Without going into details, deviations from this relationship can be caused 0
either by bad spectrometric conditions (broad band of measurement wavelength) or by interactions between the different components (association, dissociation, or charge transfer complexes).
In any case, knowing the components, in classical analysis a so-called multicomponent analysis can be carried out, doing in a first step a calibration by measurement of exact concentrations of the different components. This yields the absorptivities of all the components. In a second step the absorptivities obtained are used to calculate the concentrations of the different components in the unknown analyte. To do so it is necessary to know the components and have them separated. In classical analysis this method is quite successful up to three or four components. But in kinetics its successful use is more difficult, since 0
usually one does not know the spectral properties of intermediates or of degradation products, the relative concentrations permanently change during reaction.
Therefore a so-called dynamic multicomponent analysis has to be done, which turns out to be more difficult than the stationary case. Neglecting the photodegradation process in photochromic systems it could be shown that the large number of dependent steps of reaction could be reduced to a single linear independent step. In principle measurement at one single wavelength should be sufficient
38
to follow the course of this reaction. The reason is, that any decrease in educt concentration gives an equivalent increase in product concentration. Using the law of conservation of mass the measured absorbance can be taken to determine the change in concentration of one of the photochromic partners, provided their absorptivities are known.
The use of the spectra instead of one single wavelength has the advantage of yielding more information during one experiment. By theory this information has to be the same at the different measured wavelengths. This fact gives the chance to compare the different information, to make statistics, and to calculate the reaction constants with their relative standard deviations. The larger the change in absorbance during the reaction, the better the signal to noise ratio. Measurement of the spectral change gives the chance to select the best wavelengths for the successive kinetic evaluations. It is a principle of statistics that many measurements give a better result than a single one, even though in this case the different wavelengths of the spectrum are not really independent. They have, however, a different signal to noise ratio and, assuming no systematic error, differing statistical noises. For this reason at many reaction times not only the absorbances at a single wavelength are recorded, but a total spectrum of the significant wavelength range is taken. This overlay of the spectra is called a “reaction spectrum”. It visualizes the spectral changes during the reaction. By its use areas of large or small changes in absorbance can be found. Qualitatively this knowledge can be used to select the best wavelengths for a later quantitative evaluation. Such a type of reaction spectrum is presented in Fig. 8 for a photochromic system. It demonstrates the spectral changes of 1,8a-dihydro-2’,3’-dimethoxycarbonyl-l1ldiphenyl-9,l1-pyrrolo-[1,2-b]pyridazine (a dihydroindolizine derivative, DHI 294) during the photoreaction in dichloromethane. The spectrum contains one characteristic point at 395 nm, at which there is no visible change in absorbance during reaction. According to eq. (32) two reactants cause an absorbance
at any time of the reaction. Using eq. (23) above relationship can be rearranged to
Therefore the change in absorbance during reaction is given by
Since concentration a changes with time, absorbance only stays constant if the absorptivities of both the reactants at a specific wavelength are equal ( E X A = E A B ) . Then the result is
EDA(t) = 0
.
(36)
39
I
8.88--
8.60--
8.40-
0.20--
0 00-
258
310
378
438
490
wavelength
cnnl
Fig, 8: Reaction spectrum of a [1,8]-dihydro-indolizine derivative (DHI 294, for detailed formula see text) in CH2C12, irradiated a t 436 nm, transmittance plotted versus wavelength.
If the spectra of the two reactants superimpose, a wavelength can be found at which both their absorptivities are equal, and no change of absorbance happens during reaction. This characteristic feature in the reaction spectrum is called m “isosbestic point”. These spectral peculiarities give a first hint - but no proof - that the reaction is simple. This problem will be discussed later. In some cases the changes in absorbance are very small and not very significant, especially if the reaction is very complex. Under these conditions it has proved advantageous to visualize the relative spectral changes in photokinetics, too (ref. 28). Therefore the absorbance signal is differentiated with respect to the wavelength. But it has been shown that neither higher orders of numerical differentiation (high order derivative spectra) nor electronic differentiations give results with the necessary quality. Besides, the differentiation of a signal increases the noise in contrast to integration. For this reason any differentiated spectrum will be worse than the original crude data. On the other hand by forming the derivative even small changes can be visualized. Therefore the calculation of a derivative reaction spectrum gives the experimenter the chance to find areas in the wavelength range where the information is better than in others. This can be seen by comparing two figures one giving the reaction spectrum of a diphenylhexatriene in a micellar solution. It turns out to be a multistep process one would not expect from regarding the reaction spectrum (s. Fig. 9). The other figure gives the first derivative of this spectrum, which can he used to select the best waveIengths for evaluation (s. Fig. 10). The derivative spectrum also contains
40
T
I
0.80--
0.60--
0.40-
0.20--
260
280
300
320
340
360 Uovelength
380
CnmJ
Fig. 9: 2,6-Diphenyl-1,3,5-hexatriene in Cremophor-EL (5) micellar solution, irradiated at nm, showing a reaction spectrum, recorded in transmittance.
365
some more information. The spectral crossings of the different spectra do not lie at zero absorption, but somewhat above at wavelengths less than 320 nm. This can be explained by a non-changing-basicabsorption, caused by the turbidity of the micellar solution. This scattering is eliminated by differentiation of the spectrum and becomes only evident by a shift of the “zero”-line of the derivative spectra. Now the original crude data of the measurement of absorbances at selected best wavelengths are used for further evaluation. It contains two parts:
first, the determination of the mechanism and 0
second, the quantitative evaluation of kinetic constants.
3.6 Examination of the mechanism Since the form of the differential kinetic equation depends markedly on the mechanism, the first step of any kinetic examination is the determimtion of the mechanism. As has been shown in section 3.1 this procedure is a combination of chemical intuition and information, obtained during different analytical processes (refs. 24,26). Mass spectroscopy, NMRspectroscopy or even IR-spectroscopy give information about the chemical structure of intermediates or the photoproducts. The data obtained during UV-spectroscopy can be
41
dT/dA
lo-’
!
0.60
-
0.40.-
-
0.20
0.00-
-
-0.20--
!
-0.40 260
280
300
320
340
360 Wavelength
380 Cnm3
-
Fig. 10: Derivative reaction spectrum of 2,6-diphenyl-1,3,5-hexatriene in micellar solution (Cremophor-EL ( 5 ) ) , irradiated at 365 nm; first derivative of the transmittance signal VS. wavelength.
used on the other hand to determine the number of linear independent steps of reaction. This information is necessary to set up the differential kinetic equations. As explained above, this number of linear independent steps of reaction correlates neither with the order or molecularity of the reaction nor with the number of elementary steps taking place during reaction procedure. Moreover it gives the rank of the Jacobi matrix, which is determined by the number of differential equations necessary to describe the total reaction, but excluding any linear dependences between different product concentrations or kinetic constants. This rank will usually be smaller than the number of kinetic equations first set up to describe all different steps of the reaction and dependencies of the concentrations of reactants taking part. There are two possibilities to determine the rank of the Jacobi matrix, either a numerical or a graphical one. In the numerical determination computer algebra is used to reduce the starting-matrix to the final Jacobi matrix by a numerical algorithm. The quality of this procedure depends on the quality of the measured data and on the relative rates of the different steps. Since the crude data have a statistical error superimposed, the numerical determination is limited by this. In many cases it will be somewhat difficult to decide between different numbers of indcpendent steps, since the standard deviation can amount to f0.5. Especially in such problematic cases the graphical methods are advantageous. It
42
1 .OO 0.aa 0.80 0.40 0.20
1 .OO
2.00
3.00
4.00
time CSI
Fig. 11: Absorbance-time diagram (Et-) of “DHI 294” in CH2C12, irradiated at 436 nm; only the time dependence observed at 469 nm is given.
is more easy to decide from a graph whether a curve differs from a straight line influenced by systematic deviation or by a statistical one. Besides, even in the case of very noisy crude spectroscopic data, one will better distinguish graphical data than trust a numerical value. In Fig. 11 an absorbance-time graph is given. The absorbance values for the different measured wavelengths are taken from the reaction spectrum of the photochromic dihydroindolizine (DHI) at time intervals. Therefore each curve in the diagram represents the change in absorbance during reaction at one specific wavelength of measurement. Each two combinations of these wavelengths are used to construct a sc-called absorbance diagram (Ediagram, s. Fig. 12). The ahsorbances used are taken for the same irradiation times and plotted versus each other. The result is a curve in this type of diagram. Its form gives the possibility to decide between a single or more than one linear independent steps of reaction. The E-diagram in Fig. 12 shows a straight line for some typical combinations of absorbances at different wavelengths even for noisy data. Therefore the number of linear independent steps is one (s = 1). Such a straight line should be obtained in the case of photochromic systems (DHI’s), since all the different elementary reactions mentioned depend on each other. If a straight line is obtained one calls the type of the reaction a uniform one. The straight line is formally obtained according to:
43
E469
4
1.28
..e
...... ......
. ..... ... .. ... . .......
$
0.88
-
0.60
-
0.20
. . ... .
.
..l..
. .. ....
-
.
.*. .. . .. ...
.: ..
i'
I
.......'j,
..2
...
..1
..C.
..
x:nn
1 :520 2 t 490 3 s 436 4 :405 5:334 8 :280
.. ../.....
..... ,. /,.. .:':
. . ......' ../.'
0. .a
..
. . .. ............. .... . . . . ... . . . ....
1 . 0 ~-
0.40
4
D
a
i
i
Fig. 12: Linear absorbance diagram (E-diagram) of "DHI 294" in CH2C12, irradiated at 436 nm. The absorbance at 469 nm is plotted vs. wavelength, marked at th'e right top of the figure by the symbols 1 to 6.
Ez(t) =
(EZA
-
+ Ezo
E Z B ) ~ (*~d )
(38)
By substitution of the concentration a in eq. (37)by eq. (38)one obtains the equation
E O
which determines the linear relationship between the absorbances at the two wavelengths
(ref. 24).
A further example of a photochemical reaction of the photochromic dihydroindolizine system is given in Fig. 13. One can observe that the data do not fit to the straight line and show a systematic curvature. The diagram c m be interpreted by two linear sections. It proves that the photostationary state can react further in a consecutive photodegradation process. In Fig. 14 the corresponding reaction spectrum is given. Whereas at longer
44
wavelengths no principal difference can be observed, in the short wavelength range the isosbestic point disappears.
. .': .
E 520 ) 0.40
.. . . ..
-
0.35 0.30
0.20
0'25 0.15
... . ..
-
t
x'
. .
.. .
0.101
0.05
..
..
. . . . . . . .
. .. *
I
0.20
0.40
J
i
EX
r
0.60
0.80
Fig. 13: Curved E-diagram of "DHI 294" in n-hexane, irradiated a t 302 nm. Absorbance E520 plotted vs. absorbances at two other wavelengths 385nm).
Ed15
and E385 respectively
(A: 415, *:
To distinguish between two or more linear independent steps of the reaction, diagrams of higher order have to be plotted. By forming the difference between the absorbances at two times during the reaction for the wavelengths XI and A 2 one obtains
instead of eq. (39). Zero point straight lines are found in the diagrams plotting this difference for one wavelength versus the difference at another wavelength in the caSe of s = 1. That means the information of such ED-diagrams turns out to be equivalent to that of E-diagrams. Forming the quotient of such differences and using the information at three different wavelengths, a linear relationship is obtained in the case of two linear independent steps during the reaction. This type of higher order diagram is called an 'absorbance-differential quotient diagram of order 2' (EDQP-diagram). The relationship is given by
45
Fig. 14: Reaction spectrum of "DHI 294 in n-hexane, irradiated at 302 nm, representing two independent steps of reaction.
whereby c1 and c2 are constants containing combinations of the absorptivities at all the three wavelengths. In Fig. 15 such an EDQZdiagrun is plotted for the photoreaction of dihydroindolizine including the photodegradation step (3 = 2). The straight line proves that a ring-opening and a consecutive degradation step take place at chosen conditions.
3.7 Determination of reaction constants In Section 3.4 the rate laws for .different types of photochromic systems were given in dependence on concentration and time. As discussed above, UV/Vis-spectroscopy is used to observe these changes in concentration. Hence equations like (21) have to be transformed from concentration to the measurable absorbances. Lambert-Beer's law (eq. 34) can be rewritten in a general form as
46 nm
A : 365 6:313 C:302
3.58
v.aa 2.58 '
2.00
I .sa
;f
.r:
..
r&&$ '. ... ... 0, G..
i '
.... .... ... ... . . ... ...
.
.
'a... '=....
I .aa
-4.08
-2.88
6.80
2.88
'0 4.80
8.ea
8.138
Fig. 15: Linear absorbance difference quotient diagram (EDQZ) of "DHI 294" in n-hexane, irradiated a t 302 nrn. Quotient of absorbance differences E D Q 2 g plotted vs. EDQ2L0, where represents wavelengths (A: 365 nrn), (8:313 nrn) or (C: 302 nrn).
*
and in its dependence on time as
By reducing the matrix Q -' to the minimal rank without linear dependences (by applying the eq. ( 25) for the conservation of mass), the Q'-matrix becomes a regular one.
-
The same statement was valid for the &- or 2-matrix, called the Jacobi matrix. If the reaction is given by the mechanism A -+ B + C , meaning that three different reactants appear during the reaction procedure, the Q-matrix is in the reduced form: -
(2) = E = ( \
d(EiA d(E2A
- E i c ) d(E1B - 61C) - E Z C ) d(EZB - EZC) Q v
), ( ) ( -k
E1C €
By combination of eqn. (31)and (43)one obtains ( d = 1cm assumed) whereby
.40)
2 *~ Q(0)
)
(44)
41
and
a(m) =
Q-'(E(m) -E(O)) .
(47)
Therefore the following equations finally result:
&=Q - . 2 Q-' - E(t) I *
*
*
*
F(t) - 4.2 * Q-' . E(m) * I * F ( t )
or
By defining
and
one gets
and
In the subsequent part the procedure is discussed how to solve the presented photokinetic differential equations. Two specific mechanisms are used to explain the elements of the different matrices given, 0
either the photoconsecutive mechanism
A % B % C
(53) containing two linear independent steps of reaction, which can take place during photodegradation of photochromic systems, 0
or the reversible photoreaction of the photochromic molecule, which has the rank one:
A%B, B % A
(54)
48
In the latter case the g-as well as the 2-matrix are reduced to a vector or a constant, respectively. For their determination only a single wavelength has to be observed. In the fist case more than one wavelength have to be employed. The more general approach, which is independent of the rank, can be demonstrated by the mentioned consecutive mechanism. In principle the higher the rank of the Jacobi matrix is the more wavelengths have to be measured. Quations like eq. (48) have to be set up for at least as many wavelengths as linear independent steps of the reaction have to be determined. Such equations form a set of coupled differential equations (refs. 16,29). Therefore E and E depend on time and wavelength as well:
The absorbances at two wavelengths (All A,) have to be measured at many times ( t ) of the reaction. Therefore E and E become the matrices &Al t) and H(A,t ) , respectively. The elements zlrn contain complex relations between partial photochemical quantum yields and absorptivities (those at the two wavelengths of measurement as well as those at the wavelength of irradiation A'). For example the element 201 becomes
Because of the photokinetic factor F ( t ) equations like (55) cannot be solved in a closed form, even not for the most simple mechanism A B. For this reason Mauser introduced a new method, called the 'formalintegration' (ref. 30). Eq. (51) is rearranged in such a way that the integrals obtained can be calculated numerically (refs. 16,26,28,29):
J r d E ( A , t ) = z . I . / " E11( A , t ) F ( t ) d t - z . I . E ( A , m ) / F ( t ) d t
.
(60)
This equation is valid for the purely reversible photoreaction (rank one). But, the consecutive mechanism requires a coupled set of differential equations l i e eq. (60). The &matrix contains elements such as for example zll. One can generalize these formulae and use them in matrix notation:
49
In this equation the matrix
contains the values, obtained for at least two wavelengths
of measurement at many times of the reaction by numerical integration of the changes in
absorbance. The integral matrix m ( X , t ) becomes more complex. It contains elements like
INqj(X) =
[J"Ex(t) F ( t ) dt - E x ( w ) J1' F ( t )dt] . ti
(62)
ti
Thus the matrices
and
NTi(1) u= ( IINT,l(2)
INTi(1)
.. .
INT,"(l)
INTZ(2)
...
INT,"(2)
are obtained. This matrix equation can be solved by the following procedure to obtain the elements of the g-matrix: 0
0 0
J
J
First the integrals Ex(t) F ( t ) dt, Ex(..) F ( t )dt, and/ dE(X,t ) are calculated from the data measured by numerical integration. Next, these values are put into eq. (61). The g-elements are calculated by use of the over-determined equation (more times of reaction measured than necessary). This can be done by matrix calculus.
1. In the first step eq. (61) is multiplied by the transposal of matrix the right
D E *m T= I*g.INT-INT'= . _
I*g.g.
uT
from
(65)
The multiplication of a matrix by its transposed gives the so-called. normal matrix &. The new equation is called the normal equation, familiar from common least squares procedures in error analysis.
2. In a second step this normal matrix is multiplied by its inverse (the product gives the unity matrix):
D E .I N T ~ . ~ -='
r.g.&.g-'
= I . H .
(66)
Thus the multiplication of the three matrices on the left yields the elements of the &matrix.
50
The problem is, that only the absorbances can be measured. Therefore the complex elements of the g-matrix are obtained instead of the expected elements of the Jacobi matrix 2. It is evident that the elements of the g-matrix contain the reaction constants (photochemical quantum yields or thermal rate constants) and some combinations of the absorptivities in a very complex manner. Therefore, in these elements of the &matrix the reaction constants can only be obtained if all the absorptivities are known. The result is, that in many kinetic analyses the results obtained are limited in practice. But in such and g can be used to get further cases a characteristic quality of both the matrices information. Both these matrices are similar, regular matrices. Therefore they have the same eigenvalues. Their trace and determinant are equal. This information can be used for the determination of reaction constants in many cases (ref. 17). It was mentioned that the first step of the photoreaction of a photochromic system has the rank one. Therefore only a single wavelength is necessary for evaluation. The assumption of the purely reversible photochemical step (no further photodegradation, no superimposed thermal reaction) reduces the Q-and 2-matrices to one element each:
-
Thus one can set up eqs. (60) and (61) in vector notation
whereby
or
That means the assumed simple mechanism allows direct calculation of ~ 1 1since , qll . q;1' gives 1. The difference between eq. (61) and eq. (70)is, that in the latter the matrix is reduced to the vector
m,.
51
But, to obtain ‘pf and ‘pf (the partial photochemical quantum yields), the absorptivities of both the reactants at the wavelength of irradiation have to be known. The solution of the over-determined equations yields the two constants 21 and 2 2 . Hence E(w) can be calculated by division of 22 by zl. Any other mechanism results in more complex matrices. Only exceptional cases givc partial photochemical quantum yields directly. Therefore preferably concentrations should be observed during reaction. This method is discussed later in section 3.10. The knowledge of the absorptivities allows the reaction constants to be calcula.tctl by use of the eigenvalues. All experimental data are affected by noise. Therefore the quality of the calculation discussed above depends on the wavelength of measurement used for the evaluation. Consequently formal integration yields reaction constants with a standa.rrl deviation. Therefore the normal procedure of kinetic analysis is extended by another step: the eigenvalues of the Jacobi matrix are calculated for some wavelengths. Their mea.n valucs are put into the rate equation. This procedure allows recalculation of absorba.nces at different times of reaction procedure. These calculated (‘theoretical’) absorbance-tiinc values can be compared with the experimental ones. The result is given in Fig. 1G. The symbols represent the measured experimental data set as absorbances for cliffcrciit wavelengths of measurement (every 5th value is marked by a symbol). In addition to thc data at each wavelength three lines q e plotted. The average one of those is obtained by the above mentioned procedure of recalculating ‘theoretical’ data sets from the avera.ged reaction constants. This procedure is called a Runge-Kutta-algorithm ‘simulation’ (ref. 31). Beginning at absorbances for time t = 0 (refs. 4,8,30)in a step by step wa.y nc\v absorbances are calculated. The mathematical formalism ensures that the simulated viilucs fit in an optimal manner. In general a mis-postulated mechanism causes a systematic deviation between 111Cil.sured and simulated absorbances. But, it turns out that the algorithm can be not scnsitivc enough to respond to wrong mechanistical assumptions under certain experimental conditions (extreme differences in reaction constants for the different steps, poor spectroscopic conditions, small changes in absorbances). Therefore a third step of evaluation is acldccl. The above mentioned standard deviation from the calculated reaction constants is usctl to superimpose an upper and a lower error limit to the averaged values of the reaction constants. Consequently beside the averaged reaction constants those with a superiniposcd extreme positive and extreme negative deviation are used for simulation, too. These t\ro additional calculation steps yield the other two curves next to the experimental clnttn sct.s as can be seen in Fig. 16. The more “noisy” the experimental data set is, the larger will I x the difference between the three curves for each wavelength of measurement. Because of the shape of the three curves, they can be called an ‘error trumpet’. Significant deviatioiis of the experimental data set and the graph of the ‘error trumpet’ demand a thorough reconsideration of the proposed mechanism.
As a conclusion it should be mentioned that the quality of the quantum yields or rei1.ction constants obtained depend on the exactness of the measurement of the absorptivities
52 302 540 520 490 X ' 436 A 415 385 334 280 254
0Q
1
0 80
B 60
nm nm
nm nm
nm nm
nm nm nrn
nm
0 20
I
Fig. 16: Evaluation by Runge-Kutta-simulation of the consecutive photoreaction of "DHI 294" in n-hexane, irradiated at 302 nm, absorbance versus time, experimental (symbols) and recalculated (lines) data sets for some wavelengths of measurement.
and on the determination of the intensity I of the irradiation source. The best procedure to obtain this intensity is chemical actinometry explained in a later chapter. The determination of the absorptivities needs detailed experiments. A prerequisite is the knowledge of the identity of all the reactants and to get significant calibration curves. In any case, the quality of the determination of absorptivities is limited. This property and the problem that spectroscopically measured data are noisy limits the distinction of different mechanisms and the quality of the determined rate constants. Even though photokinetic equations cannot be solved in a closed form, a transformation of the time a x i s will allow the use of exponential fit techniques. The general photokinetic equations (19 to 22) can be rewritten by use of a 'pseudo-time'
8 =I
J F(t)dt
.
Therefore da(@) -
d@
- -&
with the solution
*
a(@)
+ R2 . b(@)
(73)
53
is obtained (ref. 33). By use of Lambert-Beer’s law a general relationship for pure photochemical reactions is derived:
E x ( @ ) = PI . ,-P2Q
+ P3 .
(75)
The constants Pl, Pz,and P3 contain combinations of the reaction constants (R1, R2) and the absorbance at the end of the reaction E ( m ) . This equation can be fitted by typical exponential functions. The parameter found during the curve fitting techniques is used to determine the reaction constants and E(oo). Up to now, only pure photochemical reactions have been considered with respect to photokinetical examinations. If the photoreaction is superimposed on a thermal backreaction, the evaluation is more complex. 3.8 Competitive thermal backreaction / photoreactions and differential equations
It has been mentioned in the previous section that the superposition of thermal backreactions increases the task of evaluation. In this case eq. (21) is no longer valid. For a purely photochemical reversible reaction another concentration-time law has to be taken (eq. 22) and transformed to absorbance-time differential equations. In the case of a thermal backreaction the use of the photostationary state (u(m) = 0) cannot be employed to simplify the relationship of eq. (25). Consequently one obtains by formal integration
J
~ = E
21
./E(t)
*
F(t)dt
+ z 2 / F ( t )d t + r3 /[E(O) - E(t)]d t .
(76)
The use of this equation takes into account the influence of the thermal reaction. But in comparison to the differential equation for pure photochemical reactions the Jacobi matrix contains three constants. Consequently numerical problems increase during evaluation. Whereas exact numerical data sets, calculated with assumed rate constants, intensities, and absorptivities, are without any “numerical noise”, the experimental data, obtained by spectroscopy during photoreaction, in contrast to these results are dispersed. Therefore the latter data cause more numerical instabilities in the solution of the differential equations. Because of these numerical instabilities, the evaluation becomes more complex where three reaction constants have to be determined instead of two. Therefore this reason one tries to avoid too many reaction constants and equations like eq. (76). Two approximations are possible: 0
0
If the equations for the purely photochemical reaction (eq. 20) are used, one has to take into account that wrong reaction constants are obtained. If another approximation is used, which assumes that the photokinetic factor F ( t ) does not change during photoreaction (irradiation wavelength close to an isosbestic point), then F ( t ) becomes constant (F). It has not to be included in the integrals and therefore eq. (76) can be rearranged to give
54
Both approximations reduce the number of constants in the equations to two. But it is obvious that both the approximated equations are only valid under certain conditions. A systematic examination of the two approximated equations (eq. (77) and eq. (21)) and the exact eq. (76) proved that the correctness of the obtained constants depends on the photochemical quantum yields relative to the thermal rate constant. The ratio of constants 21 and 23 takes into account this dependency, because one can find that
To get an idea of the limitations of the approximations in eq. (60) and (77), respectively, instead of experimental data numerically calculated ones have to be used. The reason is that the uncertainty of the experimental data does not allow a clear distinction between a good and a bad evaluation result. Therefore a number of data sets are simulated for a large variety of assumed photochemical quantum yields, absorptivities, light intensities, and thermal reaction constants. In Fig. 17 the influence on the error in the constant z1 in dependence on the amount of the thermal reaction (given by z3)is shown. The smaller the ratio of 23 : 2 1 becomes, the better the assumption of purely photochemical reaction will fit the calculated theoretical data set. The result is an error less than a few per cent, when 23 amounts to less than 5 % of 21. Furthermore it can be seen that the exact eq. (76) gives a very small deviation from the previously assumed values for all ratios 23 : zl. That means that using numerical exact data sets the calculation results in reaction constants with small standard deviations (refs. 32,33). The other question is, how much a change in the photokinetic factor will influence the quality of the evaluation. That means how close the irradiation wavelength has to be to the isosbestic point in the reaction spectrum. Therefore the two constants z1 and 2 2 derived from the eqns. (76) and (77) are tabulated in comparison to the exact assumed values. As can be seen from table 3, even minimal changes in the absorbance at the wavelength of irradiation caiise the constants z1 and z2 to be sometimes no longer of the same order of magnitude as the assumed ones. That means that even a small deviation of the photokinetic factor from constancy causes large errors. In general the following conclusions can be drawn, when a photochemical reaction is superimposed on a thermal one:
In contrast to the purely photochemical reactions, those with superimposed thermal reactions require very critical evaluation procedures. 0
The approximation of the “constant” photokinetic factor is only applicable if an isosbestic point is given at the wavelength of irradiation or the change in absorbance A E’ remains smaller than 0.05 units during reaction. When the thermal reaction rate constants Ic3 are smaller than 5% of 21, the “wrong” evaluation by the purely photochemical equation leads to better results for 21 and 22 than the correct equation with three constants. The reason is the better numerical stability of this equation.
55
GO
1
Fig. 17: Plot of the error deviation of Jacobi matrix element z1 vs. the ratio of thermal (23) to photochemical (2,) elements. Curve ( 0 - - 0 ) gives the correct eq.( 76), (0 . .O)the purely photochemical approximation by eq.
( 20)
The correct equation provides elements z1 and 2 2 with errors of less than 5% or 0.5%, respectively, for different ratios of 21 : 2 2 : 23. Moreover the error in 23 (the thermal reaction constant) increases for small thermal back-reactions and noisy data sets (ref. 33). 0
With the exception of these special conditions, photoreactions with superimposed thermal reactions need ax evaluation according to eq. (76). In any case, thermal reaction rate constants should be determined separately according to common procedures in thermal kinetics, but not by eq. (76). Later on the determined values of, 23 can be inserted during the evaluation procedure. In this case, the significance of the total evaluation increases.
3.9 Photochromic systems embedded in polymers The derivations of rate laws discussed have one strict prerequisite: the sample has to be homogenized by stirring during irradiation. Otherwise, the concentration in the rate laws would depend on time and on the volume element as well. Using the symbols I and Ri R2 = R for purely photochemical reactions, the rate law has to be written as a
+
56
TABLE 3: Change in absorbance at wavelength of irradiation causes a n error in according to eq. ( 76) and ( 77)
I
AE‘ 0
0.117
21 22
21 22
0.286
21 22
0.516
21 22
21
and
22
% deviation from correct assumed value eq. ( 76) eq. (77) 0.60 0.36 0.35 0.64 0.46 19.23 0.24 0.59
0.27 0.28 8.83 4.08 0.25 9.52 29.10 17.90
partial differential equation:
-a+, t ) --I. R1 . a ( i , t ) . e r p [ - - / 6 ( . : , . a ( z , t ) + . ’ B . b ( z , t ) ) d z ]
at
+ I . R2
*
b (2, t ) . e z p
[- [(
K:
. a (z,t ) + K’B .b (2, t ) )d z]
a q-z , t ) at
- --aa (2, t ) a t .
(79)
(80)
62 and K)B (= 2.303-6‘)are the natural absorptivities at the wavelength of irradiation. In analogy to the time transformation used in eq. (72)a transformed time axis Q is introduced in the above equation, too. Therefore one is able to write the differential equation in a closed form instead of using the time-dependent photokinetic factor F(t). The result is a pseudo linear differential equation. The transformation for Q is given in this case by Q =
it [
I
ezp
-
[(62 - a ( z , t ) + K)B - b ( z , t ) )d z ] dt
.
(81)
As a consequence, eq. (79)can be rearranged to d a(Q) -
dQ
- -R1*a ( Q ) + Rz .
The transformation of the time axis yields new equations for the concentrations (refs.
34,35):
a(Q) =
4 0 ) (R2 + R1 . e - R e ) R
(84)
57
These equations are comparable to those derived for thermal reactions. The two concentrations a ( z , t ) and b ( z , t ) can be inserted in the equation for the transformed time 0. Then 0 is given in dependence on time and volume element by
0 = I J t e z p [ - ~ ( C l + C ~ . e - R e ) d ds t] ,
(86)
where the two constants Cl and C2 can be calculated to be
In the next step 0 is first differentiated with respect to time, a0 = I e z p - l ( C 1 C2 . e-RQ) d z ]
[
-
at
+
and then with respect to volume element
This second differentiation is somewhat more difficult, since an integral with given limits z and 0(z ,t ) in the exponent has to be solved.
By integration of the above equation with respect to time t the integration constant f(z), which depends on the variable z, can be determined by use of the initial conditions for the time t = 0:
a0 _ az
4
1
. 0 + -ecR2
-R.e
+ f(z)
At this time there is no gradient of concentration. Therefore
d Q-( z , O ) -0 dr and Q(z,O) = 0
.
(93)
axe valid. Insertion of these two conditions in the above equation, rearrangement and formal integration between the limits 0 and z give the desired z-value and a time-dependent constant g ( t ) . This can be calculated by use of the initial condition z = 0. The final result is the integral
58
This integral cannot be solved explicitly. Consequently first O-values are calculated by a Newton-Raphson-iterationmethod. These 0-values allow the determination of the concentrations of a and b according to eqns. (84,85). Knowing these two concentrations, one is able to calculate absorbances at any time of the reaction. These calculated absorbancetime curves can be compared to the experimental values. As shown before, this comparison between experimental and calculated data (‘simulated’) can be used to verify the mechanism and the estimated rate constants. Recently the embedding of photochromic systems in polymer blocks or matrices has found great interest, since such systems can be used for information storage or as actinometers. Recently dihydroindolizines as well as azobenzene have been embedded in such polymer blocks. It has been shown that a detailed kinetic examination of such systems requires the transformation of the rate law from the solution case to the more complex one in viscous matter. The application of such rate equations will be discussed in detail in the chapter on actinometry.
3.10 Combination of UV spectroscopy with other physical methods The evaluation of rate laws by use of absorbance measurements has proved that in many cases these data alone are not sufficient to obtain the rate constants and photochemical quantum yields directly (s. section 3.7). Obviously additional information was necessary. Should some of the reactants during the reaction pathway fluoresce, the measured intensity of fluorescence can be used to monitor them selectively. Such combined measurements of fluorescence and absorbance were used to obtain all three partial photochemical quantum yields of a consecutive photoreaction in the case of LASER dyes (ref. 36). Even though the first step is a reversible one and no absorption coefficients of the consecutive photoproducts were known, the additional information by fluorescence allowed a photokinetic analysis (refs. 36,37,38).
Another possibility gives the combination of separation methods, as high performance liquid chromatography (HPLC)with absorbance measurements. Since chromatography allows the reactants to be separated and yields rather good quantitative values for peak area or peak height, these data can be put into the differential concentration-time laws. This procedure has the advantage that the &-or the &matrices can be calculated directly. On the other hand chromatographic detection is not as accurate as direct absorbance measurements. Nevertheless chromatography gives mechanistic information as well as
59
the change in concentrations separately and fairly qimntitatively. With some effort, by process controlled time-dependent chromatograms, a so-called reaction chromatogram can be measured, which is comparable to the absorbance reaction spectra mentioned. In Fig. 18 an example of such an reaction chromatogram is demonstrated. The model photoreaction of trans-stilbene via cis-stilbene to phenanthrene is presented (refs. 3941).
absorbanae
6.68 7 . 2 0 7 . 8 0 8.d8
9.0a
Xn’
time Is1
Fig. 18: Reaction chromatogram recorded by HPLC: Signal versus retention time measured at various times during photoreaction (plotted to the back of the diagram). Peak sequences: A: trans-stilbene; B: cis-stilbene; C: phenanthrene.
At first glance, the reaction chromatogram allows three reactants to be distinguished. The concentration of the first one (A) decreases relatively fast, t h e second concentration (B) passes through a maximum and the third one (C) is built up in the end of the reaction. Improvement of the apparatus, process control, and the data acquisition system, allow relatively good peak areas to be obtained. Iterations of experimental and calculated concentration values according to Newton-Raphson algorithm allows the rate constants to be determined directly. The method introduced has been used recently with success in a number of cases: 0
in the photoreaction of phenylpropene (ref. 42)
60 0
in the photoreaction of stilbene (refs. 39,40) mentioned above,
0
in the mechanistic determination of photoreactions of LASER dyes (ref. 43) and
0
4
the determination of mechanistic information in complex photodegradation reactions of dihydro-indolizines (ref. 44).
CONCLUDING REMARKS
The principles of photokinetics have been described. Possibilities and problems of the calculations of reaction constants have been discussed. The usual rate laws have been expanded to be usable for the evaluation of photoreactions in polymeric systems. Some new methods have been outlined which combine spectroscopic and chromatographic data to increase information obtainable during reactions. All these considerations are a prerequisite for obtaining information beyond overall yields and possible photoproducts. Any determination of quantitative values demands laborious studies. It was shown that, taking the trouble of such examinations, it allows thermal rate constants and partial photochemical quantum yields to be calculated with enough precision to compare them for different conditions of the reaction. This result offers the chance to control and optimize the reaction pathways of photochromic systems.
61
REFERENCES
P. W. Atkins, “Physical Chemistry”, 3rded., Oxford Universtity Press, Oxford, 1986. 2.
G. Gauglitz, “Praktische Spektroskopie”, Attempto Verlag, Tubingen, 1983.
3.
J. Kiefer (ed.), “Ultraviolette Strahlen”, Walter de Gruyter, Berlin, 1977.
4.
G. Gauglitz, “Photochemie in der Leiterplattenfertigug, Grundlagen und Anwendungsmoglichkeiten”, Verlag Simanowski, Gomaringen, 1987.
5.
N. J. Turro, “Modern Molecular Photochemistry”, The Benjamin/Comings Publishing Company Co., Amsterdam, 1978.
6.
G. M. Barrow, “Introduction to Molecular Spectroscopy”, McGraw Hill, New York, 1962.
7.
G. Gauglitz: “UV/VIS - Spektroskopie”, Verlag Chemie, Weinheim, in preparation.
8.
R. Frank, G. Gauglitz, Chemie, Anlagen and Verfahren, p. 19 ff, July 1978. G. Gauglitz, “Wechselwirkung zwischen Strahlung und Materie”, in “Untersuchungsmethoden in der Chemie”, H. Naumer, W. Heller (eds.), Thieme Verlag, Stuttgart, 1987.
9.
10.
G. Gauglitz: “Elektronenspektroskopie”,in “Untersuchungsmethoden in der Chemie”, H. Naumer, w. Heller (eds.), Thieme Verlag, Stuttgart, 1987.
11.
G. Wedler, “Lehrbuch der Physikalischen Chemie”, Verlag Chemie, Weinheim, 1972.
12.
G. Biinau, T. Wolff, “Photochemie”, VCH Verlagsgesellschaft, Weinheim, 1987.
13.
T. Forster, “Fluoreszenz Organischer Verbindungen”, Gottingen, van den Hoeck 8z Ruprecht, 1951.
T. Forster, Disc. Faraday SOC.,2 (1959) 7. 15. D. L. Dexter, J. Chem. Phys., a (1953) 836. 16. H. Mauser, “Formale Kinetik”, Vieweg Verlag, Diisseldorf, 1974. 14.
17.
H. Mauser, “Photoreaktionen”, in Ullmann’s Enzyklopadie der Technischen Chemie 3. ed., 16. vol. p. 440 ff., Munchen, 1965.
18.
Th. Forster, Pure and Appl. Chem., 24 (1970) 443
19.
K.A. Muszkat, E. Fischer, J. Chem. SOC.,(B 1967) 662.
20.
G. Orlandi, W. Siebrand, Chem. Phys. Lett.,
J. ‘hoe, Chem. Phys. Lett., 114 (1985) 241. 22. GDCh-Kurs “Photochemie”, Tubingen, 1970. 21. 23.
.
a (1975) 352.
V. Malatesta, C. Willis, P. A. Hackett, J. Am. Chem. SOC.,
(1981) 6781.
62
24. G. Gauglitz, GIT Fachz. Lab., 26 (1982)597. 25. G. Gauglitz, GIT Fachz. Lab., 26 (1982)205.
26. G. Gauglitz, GIT Fachz. Lab., 29 (1985)186.
27. S. E.Braslavsky, K. N. Houk (eds.) Pure & Appl. Chem., !X! (1988)1055. 28. G. Gauglitz, T. Klink, Z. Phys. Chem. (N.F.), m(198l) 177. 29. V. Starrock, Thesis, Tubingen, 1974. 30. H. Mauser, U. Hezel, Z. Naturforsch.,
(1971)203.
31. R. Zurmuhl, “Praktische Mathematik”, Springer Verlag, Heidelberg 1961 32. R. Bar, Thesis, Tubingen, 1987.
33. G. Gauglitz, R. B&, J. Photochem. Photobiol., 46 (A 1989) 15. 34. D.Frohlich, masters thesis, Tubingen, 1986.
35. G. Gauglitz, D.Frohlich, M. Guther, in preparation. 36. W. StooB, Thesis, Tubingen, 1987.
37. G. Gauglitz, R Goes, W. StooB, R. h u e , Z. Naturforsch., 4(La (1985)317. 38. J. Riedt, masters thesis, Tubingen, 1985.
39. T. Klink, Thesis, Tubingen 1984. 40. G.Gauglitz, T. I
63
Addendum to chapter 2 PHOTOCHROMISM BASED ON TRIPLET-TRIPLET ABSORPTION List of References: It seems appropriate to present an updated list of references, concerning Triplet-Triplet Absorption after G.Gauglitz' chapter as it gives rise to a photochromic system based on a photophysical process (see General Introduction).
J.L. Kropp; M.W. Windsor, Triplet to triplet absorption in organic molecules for use in photochromic devices, U.S.Govt.Res.Deve1op.Resp. 1968, a ( 2 2 ) (1968) 56.
J.L. Kropp; M.W. Windsor, Triplet to triplet absorption in organic molecules for use in photochromic devices, U.S.Govt.Rcs.Deve1op.Resp. 1970, 7(i(10) (1969) 58. V.A. Murin; V.F. Mandzhikov; V.A. Barachevskii, Study of the triplet-triplet absorption of photochromic spiropyran by the method of laser photoexictation, 0pt.Spektrosk. E ( 6 ) ( 1 9 7 6 ) 1084. Lisyutenko; V.A. Bai achevskii, Quantum-chemical triplettriplet absorption in photochromic indoline spiropyrans, Theor. Eksp.Kh5.m. 24(3) (1988) 291.
V.N.
64
Chaprer 3
Cis-Trans lsomerization of C=C Double Bonds
J. Saltiel and Y .-P. Sun 1 INTRODUCTION
The cis-trans isomerization of olefins involves 180' rotation about a C,C double bond. Except in strained cyclic olefins, the reaction is usually
A,
'8
c=c
D
-
'
'.E
A
A,
'8
c =cD ,
E /
highly activated as a thermal process in the absence of catalysts, but readily occurs on the lowest excited singlet state surface or on the triplet state surface. Depending on the olefin, reversible photoisomerization following direct excitation can occur on either or both of these surfaces, but can generally be expected to proceed on the triplet surface when the energy is delivered to the olefin by triplet excitation transfer. Since cis and-trans isomers usually have significantly different absorption spectra the isomerization is a photochromic process though a change in color is perceived only when the changes occur in the visible part of the spectra. Comprehensive reviews on photochromic olefins were published in 1971 (ref. 1) and in 1978 (ref. 2) and, more generally, on the topic of cis-trans photoisomerization in 1973 (ref. 3) and in 1980 (ref. 4). This chapter, reflecting our research fnterests, i s written in the spirit of the last two reviews, and focuses on methanistic insights uncovered since then. Because of space and time 1 imitations the coverage is 1 imited to processes occurring in the lowest excited singlet states of olefins. Though much of this work involves simple olefins, such as the stilbenes, which serve as model systems, it is inspired, in large measure, by the desire to understand this "simple" photoisomerization reaction in biological molecules. Important examples are (a) the isomerization of rhodopsin, the Schiff base of 11-cis-retinal (1) with the protein opsin, to the all-trans isomer, prelumirhodopsin (2) in the
65
1
h-opsin
2
retina (refs. 5, 6), (b) the trans -,cis isomerization of urocanic acid (3) in the epidermis (ref. 7), (c) the isomerization in model compounds of the plant photosensor pigment phytochrome (ref. 8 ) , and (d) the interconversion of
tachysterol (4) and previtamin D (5) (ref. 9). The reader will recognize that
in each of these equations, the reaction is presented in a deceptively simple form. For instance, trans-urocanic acid, aside from its anionic forms, can exist as an equilibrium mixture of two tautomers, each of which can be present as a mixture of four rotamers. The other two examples neglect the presence of ground state s-trans 3 s-cis equilibria. It can be seen, that strictly speaking, rotation about the central double bond in 4 would lead not to 5 but to a more sterically congested conformer. Precisely such considerations led Havinga to postulate the NEER (non-equilibration of excited rotamers) principle in photochemistry (ref. 9) which has inspired many recent
66
photochemical and photophysical i n v e s t i g a t i o n s on model o l e f i n s .
A better
understanding o f c i s - t r a n s isomerization o f photochromic o l e f i n s has ensued through a p p l i c a t i o n o f approaches developed f o r e l u c i d a t i o n o f mechanism i n model systems.
2
GENERALIZATIONS To a f i r s t approximationt h e k i n e t i c s of photoisomerizationand competing
processes can be based on Mulliken's p o t e n t i a l energy curves f o r ethylene ( r e f s . 10, 11). Following e x c i t a t i o n , t h e bond order o f t h e o l e f i n i c C,C bond i s s u b s t a n t i a l l y reduced and t h e e x c i t e d o l e f i n relaxes by t w i s t i n g t o a perpendicular o r phantom e x c i t e d state, the p* state, whose geometry i s roughly equivalent t o t h e geometry of t h e t r a n s i t i o n s t a t e f o r thermal isomerization on t h e ground s t a t e surface. Radiationless decay from p* gives
ground s t a t e c i s and t r a n s isomers.
The r a t e o f t w i s t i n g varies from o l e f i n
t o o l e f i n since s u b s t i t u t i o n of H w i t h groups which extend t h e x system changes t h e d e t a i l s o f t h e p o t e n t i a l energy curves and tends t o s t a b i l i z e transoid, t*, and c i s o i d , c*, geometries o f r e l e v a n t e x c i t e d states. I n 1,2d i s u b s t i t u t e do l e f i n s , t h i s e f f e c t i s o f t e n more pronounced on t h e t r a n s side due t o s t e r i c hindrance i n c i s o i d geometries. Mechanisms f o r photoisomerization are c l a s s i f i e d as s i n g l e t o r t r i p l e t according t o t h e m u l t i p l i c i t y o f the e x c i t e d s t a t e i n which t w i s t i n g occurs. A simple example o f a s i n g l e t mechanism i s shown i n Scheme 1, where the numeral i n d i c a t e s m u l t i p l i c i t y and as i s the f r a c t i o n o f l p * which decays t o It.
It can be seen by inspection t h a t quantum y i e l d s f o r t r a n s
-
-, c i s
and
c i s -,t r a n s isomerization are given by #tc = (1 a s ) and #ct = a s , r e s p e c t i v e l y , and t h a t , f o r e x c i t a t i o n w i t h monochromatic l i g h t , the photoequi1ib r ium [ [ l tI/[ b]) = (+q)[as/(l-as)3 , where (+/q), e x c i t a t i o n r a t i o , i s the r a t i o o f molar a b s o r p t i v i t y c o e f f i c i e n t s o f t h e two isomers a t Xexc and [as/(l-a,)]i s the decay r a t i o . When It*, k* and, o r l p * are s u f f i c i e n t l y l o n g - l i v e d , o r when perturbations e x i s t which enhance spin-
67
Scheme 1 Singlet Mechanism for Olefin Photoisomerization.
orbit coup1 ing, intersystem crossing can compete with other decay pathways and photoisomerization on the triplet potential energy surface can also become important. Photoisomerization quantum yields are diminished when fluorescence or other photoreactions a1 so occur. F1 uorescence most often contributes to the decay of It* but is usually observed from k* only at low temperatures in rigid media. Photoisomerization cleanly through a triplet pathway occurs when olefins are excited by triplet excitation transfer from suitable energy donors, 3D*, (sensitizers) as was first demonstrated by Hammond and coworkers (ref. 12), Scheme 2. By analogy with the singlet mechanism, isomerization yields are given by q5tc = (1 - at) and &-t = at, and photostationary states by ([ltl/[lc]) = (k,/kt)[at/(l-at)] where kc/kt, the ratio of excitation transfer rate constants to the two isomers, depends on the triplet energy of the donor. In arylolefins the relative energy of the three triplet intermediates in Scheme 2 has been shown to depend sensibly on substituents (e.g., steric effects favor 3p* by destabilizing planar geometries) and on the triplet energy of the parent arene corresponding to the aryl substituent. Significant progress has been made in the latter category by Tokumaru and coworkers. Extensive studies, especially on anthryl substituted olefins, demonstrate the functioning of one way cis trans triplet pathway photoisomerization in olefins where the energy of 3t* is sufficiently lower than the energies of 3p* and 3c* (refs. 13-15). Such olefins, though providing important mechanistic information, are not photochromic at least through the triplet pathway. For other important contributions which, regretfully, will not be included in this chapter the reader is referred to the thorough and excellent work by Gorner
-.
68
Scheme 2 Triplet Mechanism for Sensitized Olefin Photoisomerization.
I
3t kt
“tl
L~D*I
’D’]
‘t
and coworkers (ref. 16 and other papers from this laboratory) and by Caldwell and coworkers (refs. 17-19 and other papers in this series). The above generalizations stem primarily from extensive investigations of the photophysical and photochemical behavior of stil bene and related arylethenes. The direct photoisomerization of stil benes, the prototypic reaction in this area, is considered in some detail in the next section. 3
STILBENES
Through vigorous interplay of the efforts of photochemists, spectroscopists and theoreticians, the trans s?. cis photoisomerization of
stilbene has achieved the status of being one of the best understood photochemical reactions. 3.1
The lowest excited sinqlet Dotential enerav surface
The complementarity of It* fluorescence and photoisomerization was established by pioneering studies, primarily in Fischer‘s laboratory, of the
69
temperature dependence o f fluorescence quantum y i e l d s , $f, and o f $t-,c(refs. 20-26) and w i t h t h e advent of l a s e r pulse e x c i t a t i o n , rf (ref. 27). Comparative quenching studies of e x c i t e d s t i l b e n e s t a t e s obtained by d i r e c t e x c i t a t i o n and by t r i p l e t e x c i t a t i o n t r a n s f e r established t h a t intersystem crossing from It* i s a t best a minor process a t ambient temperatures (refs.
-
12, 28-30) and t h a t t h e a c t i v a t e d process l e a d i n g t o photoisomerization involves t w i s t i n g about t h e c e n t r a l bond It* l p * ( r e f . 31). I n c l u d i n g kf and k i s t from It* i n Scheme 1 gives
where ktp
-
Atpexp(-Etp/RT)
i s t h e Arrhenius expression f o r t w i s t i n g and k i s t
confines intersystem crossing t o t r a n s o i d geometry.
Selected data i n
hydrocarbon solvents are p l o t t e d i n Figure 1 ( r e f . 4). The curves are c a l c u l a t e d using best fit parameters l o g A t p = 12.6 f 0.1, Etp = 3.53 f 0.10 kcal/mol, k i s t = (3.86 f 2.3) x 107 s-1 and kf = (5.89 f 0.25) x lo8 s - l , i n eqs. 1-3 ( r e f . 4). The small value of k i s t i s c o n s i s t e n tw i t h t h e n e g l i g i b l e
c o n t r i b u t i o no f t h e t r i p l e t pathway f o r d i r e c t t r a n s -,c i s photoisomerization of t h e stilbenes. Though t h e u n c e r t a i n t y i n kist i s large, i t s value agrees w i t h d i r e c t l y measured k i s values i n styrenes (ref. 32).
0.w
++-c
0.2!
800
A
O.O(
Fig. 1
r
h
100
-
I
200
'.
'%? 300
95
Temperature dependence o f $f (0) ( r e f . 24), r f ( * ) ( r e f . 27), and
( r e f . 20) f o r t r a n s - s t i l b e n e [curves c a l c u l a t e d using eqs. 1-31. Reprinted from r e f . 4 w i t h permission o f the copyright holder, Academic Press.
Qt+c (A)
70
In addition to the excitation transfer studies mentioned above, Saltiel’s assignment of the activated process to It* -. 1p* torsional displacement in the lowest excited singlet state, Figure 2a, was supported by (a) the value of Atp which is several orders of magnitude too high to be associated with intersystem crossing (b) the failure to detect stilbene triplets flash spectroscopically following direct excitation of the stilbenes in fluid solution (ref. 33), (c) enhancement of Atp and Etp parameters in solvents of higher viscosity leading to higher values (refs. 24, 25) and (d) qjf values close to unity for electronic models of trans-stilbene (6) (ref. 34) and c i s stilbene (7) (ref. 35). Assuming c2h molecular symnetry, the ground state of
of
7 stilbene is of As symmetry and consequently single photon excitation leads to excited singlet states of 8, symmetry. The possible role of a doubly excited state Ag in stilbene photochemistry was recognized (ref. 36) soon after such a state was shown to be the lowest excited singlet state of all-trans-1,8diphenyl-1,3,5,7-octatetraene (refs. 37-39). Orlandi and Siebrand (ref. 36) proposed that S2, the second excited singlet state of stilbene, is a doubly excited state with symmetry which correlates with the ground state, So. Twisting about the central bond stabilizes S2 and destabilizes So. Configuration interaction leads to an avoided crossing between these two states at the p geometry, creating the barrier for twisting in So and an energy minimum, lp**, in the surface of the doubly excited state. The barrier for twisting in S1 was assigned to the crossing between the 8, (S1) and Ag (S2) states, Figure 2b. Earlier molecular orbital calculations failed to predict the involvement of doubly excited states in the photoisomerization because they neglected to include doubly excited configurations in configuration interaction (refs. 40-42). Their inclusion in later more sophisticated MO calculations led to results qualitatively in agreement with
71
'75
'25-
O
O
P
(c)
75
t
P
C
P
t
ANGLE OF TWIST-
Fig. 2 The barrier for twisting in the lowest excited singlet state of stilbene. Figures 2a-2d adapted from the literature as follows: (a) SaltSel proposal (ref. 3 1 ) , (b) Orlandi and Siebrand B,-Ag crossing (ref. 361, (C) Birks Bu-Ag avoided crossing (ref. 4 5 ) , (d) Hohlneicher and Dick 4A-2A avoided crossing (ref. 48). Most energies shown are arbitrary.
72
Orlandi and Siebrand’s proposal, though predicted excited state energies were generally far from satisfactory (refs. 43, 44). An avoided crossing between Bu and Ag excited state potentials could arise due to asymmetric perturbations and Birks suggested torsional vibrations about non-central bonds, induced by collisions with solvent molecules as likely candidates, Figure 2c (ref. 45). Consideration of theoretical calculations and their relationship to two-photon absorption spectra (refs. 46-48) led Hohlneicher and Dick to propose another source for the torsional barrier in the It* -. 1p** conversion. These authors assigned the first two relatively weak bands in the two-photon fluorescence 308 and 282 nm, to the two lowest excited lA excitation spectrum, Amax states, ZIA and 3lA, respectively (ref. 48). These two states were considered not suitable candidates for the Orlandi and Siebrand S2 state because according to the calculations, their energies increase with twisting about the central bond (refs. 43, 44, 49). The appropriate lA state, 4lA, was assigned to the most intense two-photon absorption band: 247 nm in the fluorescence excitation spectrum (refs. 46-48), 234 nm in the thermal lensing spectrum (ref. 47). The latter value was assumed to reflect the true energy of this state because the two-photon fluorescence quantum yield diminishes sharply (ref. 47) in the wavelength region corresponding to the third excited lA state. The excitation wavelength effect on the fluorescence quantum yield suggests that twisting directly to 1p** provides a viable decay pathway for the third excited lA state which by-passes the formation of the fluorescent lBU* state. Photoisomerization from the lBU* state would require an equilibration step with the first excited lAg state and passage over a torsional barrier created by an avoided crossing between ZIA and 4lA potential energy surfaces, Figure 2d. A clear choice between the mechanisms in Figure 2 is not possible at this time. However, the excellent agreement of the theoretical kf value for the lBu* -, lAg transition (ref. 50) and the experimental kf, eq. 1, at all temperatures (refs. 4, 27, 51) places the constraint on the mechanism in Figure 2d that the equilibrium population of molecules in ZIA be no larger than a few % at the highest temperatures employed. Larger populations of 21A would lead to smaller apparent kf values as found for all -trans-1,6-diphenyl-1,3,5-hexatriene, DPH, and all-trans-1,8diphenyl-1,3,5,7-octatetraene, DPO (refs. 37-39, 52). As pointed out by Hohlneicher and Dick, if the ZIA state lies about 6 kcal/mol above the state, as estimated roughly from the low resolution spectra, the barrier in the ZIA state would be too high to contribute to the photoisomerization mechanism and the other mechanisms in Figure 2 involving Bu -. crossing
-
73
would still apply (ref. 48). It will be shown below, that an A state relationship of the type shown in Figure 2d lends itself better to the understanding of the behavior of diphenylpolyenes. Recent modified neglect of di fferenti a1 over1 ap-configuration interaction calculations for sti 1 bene by Troe and Weitzel suggest that photoisomerization involves adiabatic rotation about the central bond of the initially formed IB excited state (ref. 53). The small energy barrier along this pathway is described as "CI-induced" because the calculations trace its origin to a larger degree of configuration interaction in the trans than in the perpendicular geometry (ref. 53). While these calculations bring us back full circle to the Saltiel mechanism in Figure 2a, a definite preference for any of the mechanisms in Figure 2 cannot be based on experiment at this time. Relatively less is known about the excited states of cis-stilbene because they are very ihort-lived. The lack of fluorescence in fluid solutions was an early indication of the short lifetime of 1c* (ref. 54). Only in rigid media at low temperatures is cis -* trans photoisomerization rendered inefficient (refs. 23, 24) and fluorescence, exhibiting a large Stokes red shift (- 4.5 kcal/mol), is observed from 1c* (refs. 55-58). In the absence o f constraints imposed by the medium, &+t is temperature independent (refs. 22, 23) indicating that 1c* -* 1p** twisting occurs without an internal energy barrier. Another very fast reaction pathway from 1c* is cyclization to dihydrophenanthrene (DHP) (refs. 59-61), a photochromic reaction discussed in
chapter 7 of this book.This process too is inhibited by low temperature and rigid media (ref. 58). An Sn +- S1 transient absorption spectrum, Xmax = 660 nm, observed following 266 nm pulsed excitation of lc in 3-methylpentane at 77 K has been assigned to lc* (ref. 62). The fluorescence lifetime of lc* was 4.7 ns under these conditions (ref. 62). Probably because only with the advent of fs laser pulses has direct investigation of the dynamics of lc* recently become possible in the absence o f medium imposed constraints (see Section 3.2), theory has not been successfully applied to the cis side of stilbene's potential energy diagram. In fact, Orlandi and Siebrand's schematic presentation of excited state stilbene potential energy curves
74
incorrectly predicts an internal barrier for 1c* -, 1p** twisting, Figure 2a (ref. 36). Others have shied away from extending their calculations to the cis side of the potential energy diagram (e.g., ref. 43), and the most advanced calculations applied to date (ref. 44) predict a barrier to lc* lp** isomerization due to an avoided crossing between IB excited states. In the past some liberty has been taken in utili'zing these calculations so that, consistent with experiment, the predicted potential energy curves would show no barrier for lc* -,lp** rotation (ref. 4).
-.
3.2 Ultrafast laser soectroscoDv 3.2.1
Solution studies
Since the late seventies ps and fs laser pulses have been used to investigate the behavior o f excited stilbene at different points on the singlet energy surface. The study of the temperature dependence o f the fluorescence lifetime of trans-stilbene by Sumitani et al. (ref 27) was especially important. It refuted experimentally earlier claims by Birch and Birks that at high temperatures (253 - 333 K) in fluid hydrocarbon solutions the lifetime of I t * fluorescence increased to the ns time range because it was dominated by the decay rate of 1p** from which it formed reversibly (ref. 63). This claim, though also inconsistent with earlier fluorescence and isomerization quantum yield measurements (ref. 51), stimulated many of the earlier pulsed excitation solution studies. It is now well established that lifetimes in the ns time range observed by Birch and Birks (ref. 63) and in later studies (refs. 64-67) were caused either by impurities or instrumental difficulties. The parameters obtained by fitting quantum yields and 1 ifetimes to eqs. 1-3 predict 7 = 97 ps for It* in 3:2 methylcyclohexane/isohexane at 298 K (ref. 4), which can be compared with the reported value of 108 ps (ref. 27). Single exponential decay of It* fluorescence has yielded 7 = 75 ps under the same conditions (ref. 68) and others have measured values in the 68 - 130 ps range at room temperature in a variety of hydrocarbon solvents (refs. 6972). Since the lifetime is strongly solvent dependent some of these differences may not be due to experimental error (see below). The transient fluorescence studies have been complemented by several excited transient absorption and ground state recovery investigations. The Sn S1 transition in It* has been assigned to a transient spectrum generated immediately upon laser pulse (- 10 ps) excitation of It*: Xmax = 585 nm, 7 = 90 f 5 ps in n-hexane, 296 K, ,Iexc = 265 nm (refs. 73, 74); Xmax = 580 nm, 7 = 110 f 10 ps in 3-methylpentane, 297 K, Xexc = 266 nm (ref. 62). The narrowing +
75 o f t h i s spectrum over t h e f i r s t 25 ps has been a t t r i b u t e d t o r e l a x a t i o n of excess v i b r a t i o n a l energy (refs. 73, 74). Using t h e onset o f t h i s absorption a t 650 nm ( r e f s . 73, 74) leads t o the conclusion t h a t Sn i s
-
10 kcal/mol above t h e o r i g i n o f t h e 4lA s t a t e i n the two-photon absorption spectrum (refs.
413-
47, 10
48). Nonetheless, i t i s tempting t o assign t h i s spectrum t o t h e lBU t r a n s i t i o n o f It*, because i t has been shown t h a t e x c i t a t i o n o f M t r a n s - s t i l b e n e i n 3:2 methylcyclohexane/2-methylpentane solvent a t 266
-.
nm, 11 ps pulse width, followed, a f t e r a 90 ps delay, by a second pulse a t 532 nm leads t o a sharp drop i n It* fluorescence and enhances t c isomerization ( r e f . 75). Furthermore, enhanced photoisomerizationby t h e second photon i s s t a r t s t o decrease o n l y a t T < 100 K) under observed a t low temperatures ($tdc c o n d i t i o n s which i n h i b i t ordinary s i n g l e photon isomerization v i a t h e lBU s t a t e ( r e f . 75). These observations are c o n s i s t e n t w i t h t h e decrease i n fluorescence i n t e n s i t y which was noted upon d i r e c t two-photon e x c i t a t i o n of t h e llAg ground s t a t e t o t h e 41Ag e x c i t e d s t a t e ( r e f s . 47, 48). They suggest t h a t decay from t h e 41Ag s t a t e may l e a d d i r e c t l y t o lp** v i a a sequence of events predicted, a t l e a s t i n part, by t h e p o t e n t i a l energy curves i n Figure 2d ( r e f s . 47, 48). Experimental demonstrationsof t h e s h o r t 1i f e t i m e o f t h e phantom s i n g l e t state, lp**, have thus f a r been i n d i r e c t . Laser pulse e x c i t a t i o n , 266 nm, of c i s - s t i l b e n e , 10-4 M i n n-hexane, was followed by d e t e c t i o n of It* fluorescence produced by a second 266 nm pulse ( r e f . 76). An exponential 20 ps was observed f o r n e t It* fluorescence increase i n d i c a t i n g r i s e time o f t h a t t h e sequence I c e -. 1p** -. o l t t ( 1 - a ) l c i s complete w i t h i n t h e time
-
r e s o l u t i o n o f t h e apparatus.
This conclusion was a l s o reached by monitoring
ground s t a t e absorbance changes i n t h e 240-250 nm ( t C> e t ) and t h e 310-320 nm ( € t> cC) regions f o l l o w i n g e x c i t a t i o n of e i t h e r trans- o r c i s - s t i l b e n e ,
M i n n-hexane, w i t h an 11 ps 266 nm l a s e r pulse ( r e f . 71). S t a r t i n g w i t h t h e t r a n s isomer, c i s - s t i l b e n e appears w i t h a r i s e time o f 85 k 10 ps a t 295 K which i s w i t h i n experimental e r r o r o f the It* fluorescence l i f e t i m e , 75 f 5 ps, under i d e n t i c a l conditions ( r e f . 71). Starting with cis-stilbene, formation o f t h e t r a n s isomer occurs w i t h i n the 15-20 ps instrumental response time l e a d i n g t o t h e conclusion t h a t t w i s t i n g t d lp** i s r a t e determining f o r isomerization i n both d i r e c t i o n s ( r e f . 71). Attempts t o observe the Sn + S 1 t r a n s i t i o n of c i s - s t i l b e n e i n f l u i d s o l u t i o n w i t h 10-20 ps time r e s o l u t i o n f a i l e d ( r e f s . 62, 70, 71, 73). However, Greene, Hochstrasser and Wiseman were able t o estimate an upper l i m i t
-
1 ps f o r the l i f e t i m e o f l c * i n n-hexane a t 296 k 1 K, thus r e f u t i n g r e s u l t s from an e a r l i e r i n v e s t i g a t i o n ( r e f . 65) employing subpicosecondpulses
Of
16
which led to the assignment of a significantly longer lived, 7 = 7 2 1 ps, absorption to lc* in hexane (refs. 70, 73). Direct observation of lc* decay was achieved following excitation in 1c* in n-hexane with subpicosecond 312.5 nm laser pulses (refs. 77, 78). In the first of these investigations 5 x M cis-stilbene (97%) at room temperature was employed and excitation with a 0.19 ps 312.5 nm pulse was followed by detection of a time dependent ion current caused by subsequent absorption by excited states of two 625 nm photons (ref. 77). The ion current increases to a maximum value immediately within the duration of the first laser pulse and decays biexponentially. The fast decay corresponding to 7 = 1.2 f 0.2 ps was assigned to lc* and the slower decay, r = 80 ps, was assumed to be due to It* formed by excitation of It present as an impurity in the 1c sample (ref. 77). The latter signal is substantial (i.e. its area exceeds that of the short component) and, since It can represent no more than 3% of the initial total stilbene concentration the corollary to this interpretation is that two photon excitation of It* at 625 nm must have a much higher cross section than the same event in lc*. The opposite holds for single photon excitation at 625 nm (refs. 62, 73). In the second investigation a 1.0 x M cis-stilbene solution at 295 K was excited with a 0.15 ps 312.5 nm pulse and transient absorption at 625 nm or 312.5 nm was monitored as a function of time (ref. 78). Both absorption signals exhibit a pulsewidth limited rise ( < 100 fs), however the signal at 625 nm, where only is .expected to absorb significantly, decays exponentially, r = 1.35 ps, while that at 312.5 nm where both ground state stilbenes also absorb
Fig. 3 Time resolved transient absorption of cis-stilbene probed and pumped at 312.5 nm; smooth curve calculated with 7p = 3 ps. Reprinted from ref. 78 with permission of the copyright holder, Elsevier Science Publishers B . V .
77
(ct = lo€,) exhibits a more complex behavior, Figure 3. The transient absorption features a pulsewidth limited rise time, a rapid decay during the first 2 ps (i.e., formation and decay of k*),and a slower rise time attributed to the formation of lp* and its subsequent decay to lc and It, Scheme 1. Using this model good fits were obtained only by assuming that k*, lp*, lc and It all absorb light at 312.5 nm, Figure 3. The lifetime of lp*, 7 = 3 2 2 ps, was obtained as a parameter in the fitting. The treatment neglects possible contributions in the absorption signal of lDHP* and b H P . This may be a serious omission since a quantum yield of 0.1 has been estimated for lc -, lDHP conversion following 313 nm excitation of lc (ref. 60). Furthermore the molar absorptivity of lDHP is about 6.2 times larger than that of lc at 313 nm, having a Xmax at 310 nm, 6 = 2.22 x lo4 cm-l M-l, in methylcyclohexane isohexane solvent (ref. 60). Nothing is known about the absorption spectrum of 1DHP* or its lifetime but its involvement in lDHP formation has been suggested (ref. 26).
-
Direct spectroscopic observation of the phantom singlet state, lp**, as an intermediate in the pathway to trans -,cis photoisomerization has, thus far, been achieved only in the case of the rigid stilbene analogue, trans-1,l'biindanylidene (9) (ref. 79). The five-membered rings inhibit rotation about
the 1-a and 1'-a' bonds in 9, which has been dubbed a stiff-stilbene in the literature (ref. 79) and this profoundly affects the UV absorption spectra of trans- and cis-9 (refs. 3, 25, 79-82), Figure 4. The lack of structure in the spectrum of cis-stilbene and the diffuse nature of the vibronic progression in the S1 So transition of trans-stilbene at room temperature had been associated by Platt (ref. 83) to deviation of the stilbenes from planar geometries due to twisting of the phenyl groups about the 1-a and l',a' bonds. In agreement with this interpretation the relatively well resolved vibronic structure o f trans-9 at room temperature is very similar to that observed in the spectrum of trans-stilbene at 77 K, where the latter is confined to more planar geometries by the rigidity of the glassy medium (refs. 3, 25, 79, 82). The origin of the spectrum of c i s - 9 i s significantly to the red o f trans-9 +
78
3.61
Fig. 4 Ultraviolet absorption spectra of trans- (-) benzene (ref. 80).
and cis-9 (- - -) in
(refs. 79-81), probably due to destabilization of So cis-9 because the severe steric interaction between the methylene groups of the five-membered rings can be relieved mainly by twisting about the C,C double bond. The short lifetime of It* in hydrocarbon solvents and the lower internal barrier for rotation, 1.5 kcal/mol, to lp** were inferred from a study of the temperature dependence of its fluorescence quantum yield (ref. 25) and were confirmed by direct measurement of the fluorescence decay lifetime of It* (A, = 265 nm, 10 ps pulse, Xem = 430 nm) in n-hexadecane, 26 ps, at 295 k 3 K (ref. 84). Lifetimes in less viscous solvents were estimated using the ratio of q5f values and the lifetime in n-hexadecane, e.g., 7 = 4.5 ps in n-hexane (ref. 84). The decay of a transient absorption spectrum, in the 450-600 nm region, blue shifted but similar to that of It* in trans-stilbene, was monitored at 527 nm following pulsed excitation at 264 nm and gave 7 = 26 ps and 3.1 ps in nhexadecane and n-hexane, respectively, at 295 K, in reasonable agreement with the fluorescence lifetimes (ref. 79); it was accordingly assigned to the Sn S1 transition in It* of 9. The short lifetime of It* afforded the opportunity to observe the formation of lp** as a bottleneck on the way to lc. This was accomplished by monitoring the time evolution of transient absorption at 351 nm following 264 nm pulsed excitation of 5 x M 9 in a series of nalkane solvents (ref. 79). An initial, medium dependent rise in absorbance was observed to decay within the first 50 ps to a constant absorbance which
-
79
was assigned to lc of 9, Figure 4. The crucial observation which leads to assignment of the short lived transients at 351 nm to lp** is that the amplitude o f its absorption increases with a decrease in medium viscosity, Figure 5 (ref. 79). The opposite trend is observed for the transient at 527
DODECANE
HEXADECANE
____________--------0
20
40
60
80
100
DELAY TIME (ps)
0
20
40
60
80
100
DELAY TIME ( ps)
Time evolution of the transient absorption at 351 nm of 5 X M trans-9 in (a) hexane (b) nonane, (c) dodecane, and (d) hexadecane. Solid circles are experimental points taken at 295 K and open circles taken at 370 K. Solid and dashed lines are calculated. Reprinted from ref. 79 with Fig. 5
permission of the copyright holder, the American Institute of Physics.
nm which was assigned to It* since at lower viscosities substantial decay occurs within the initial excitation pulse (ref. 7 9 ) . Since the It* decay is the feeding rate for 1p** formation, more pronounced bottleneck behavior (i.e., higher lp** absorbance) is observed in the lower viscosity solvents at constant temperature, and at higher temperature in the same solvent where It* decay is faster (ref. 79). The kinetics analysis of the evolution of the 351 nm absorption was consistent with a medium independent lifetime of lp**, 7p = 10 f 3 ps (ref. 79). Unfortunately, the demonstration (ref. 81) of serious aggregation phenomena for trans-9 at even lower concentrations than that employed in ref. 79 has cast a shadow on the interpretation of the observations in Figure 5. The shorter transient absorption lifetime of It* of 9 (ref. 79) from that inferred from the fluorescence measurements (ref. 84) may also reflect a difference in trans-9 concentration.
80 A recent report by Schilling and Hilinski of a strong dependence of the
lifetime of lp* for tetraphenylethylene (10) on solvent polarity is consistent
10 with a zwitterionic description for this species (ref. 85). The fluorescence of 10 shifts from green, Xmax 525 nm, to blue Xmax 460 nm, as the temperature of the medium is lowered and the fluid solution undergoes a glass forming transition (refs. 86, 87). The short fluorescence lifetime of the green emission of 10, inferred from fluorescence depolarization measurements (ref. 86) and confirmed later by direct measurement of fluorescence decay, 6 ps, in 3-methylpentane at 293 K (ref. 87), was attributed to a medium dependent torsional relaxation. At lower temperatures the blue emission from vibrationally relaxed molecules, which are torsionally held by the viscous media at a ground state-like geometry, was shown to decay with time constants equal to those of the rise times of the green emission from molecules which have undergone some torsional relaxation. Excitation of 10 in n-hexane at 305 nm with a 0.5 ps laser pulse gave transient(s) with absorption bands centered at 423 and 630 nm (ref. 88). Greene assigned the 630 nm band, 7 = 5 f 1 ps, to the partially torsionally relaxed S1 state of 10 (green emission), and the 423 nm band, which shifted to 417 nm as the 630 nm band decayed, to lp* the fully torsionally relaxed excited state of 10 (ref. 88). The 417 nm band had a surprisingly long lifetime in n-hexane, 3.0 f 0.5 ns, at 295 f 2 K (ref. 88). It is this band that was observed to decay faster with increasing medium polarity, e.g., 7 = 1.4 f 0.2 ns and 15 f 8 ps in cyclohexane and acetonitrile, respectively (ref. 85), and it was reasoned that stabilization of the zwitterionic lp* by the more polar solvent reduces the lp to lp* energy gap and increases the rate constant for 1p* lp radiationless decay. The presence o f strong absorption bands for the diphenylmethyl cation and anion in the 630 nm region supports the assignment of this band to a zwitterionic intermediate (ref. 85). Some peculiarities of the observations are still unexplained. Firstly, the 423 nm absorbance is present immediately following
-
-
-.
81
the 0.5 ps pulse; only a shift in X but no rise in absorbance is reported as the 630 nm transient decays (ref. 85). This would suggest that the relatively planar green emitting species also absorbs at 423 nm. Secondly, the lifetime o f the 420 nm species is reported to increase by more than a factor of 2 as the intensity of the initial 355 nm pulse is increased beyond the linear range of transient absorbance response to intensity (ref. 85).
-
The notion that twisted excited olefins are zwitterionic in nature was first proposed by Dauben (refs. 89, 90) and supported by theoretical considerations soon thereafter (refs. 91-94). While on the ground state surface lp is predicted to behave as a biradical (structure I for ethylene), on the excited state surface lp* is predicted to be a highly polarizable species corresponding to the mixing of zwitterionic structures I1 and 111.
H
rc H. *,-=
\a
H
\
I
H
H
\+
H
c-c'
/
n
H.
H
\
H
-3
H
3H
+.' c-c \H /
\-
IR
Although the zwitterionic nature of lp* in ethylene is theoretically well established and Salem's sudden polarization effect at perpendicular geometries with respect to twisting about the olefinic bond is expected to lead to zwitterionic *p* states in substituted olefins, the involvement of these states in cis-trans photoisomerization generally has been challenged. Valence bond theory considerations have led Malrieu et al., to the conclusion that, at the twisted geometry of polyenes, a neutral doubly excited state, which does not exist in ethylene, is lower in energy than lp* zwitterionic states and is the more likely candidate for the photoisomerisation intermediate !p** (ref. 95). If one considers the twisted molecule to be composed o f two radical parts A and B, the argument can be expressed simply in terms o f the orbital occupancies shown on the following page. The prediction which, with the possible exception o f 1,3-butadiene, is supported by MO calculations is that combinations o f type V and VI, i.e., mixing of A- B+ and At B-, leading to zwitterionic states are higher in energy than combinations of type VII and VIII, i.e., mixing o f (A'* B' + A'*' E m ) , and (A' B'* + A' B'*') which lead to nonpolar intermediates (refs. 95-97). The argument is applicable to phenyl substituted,ethylenes. For styrene, the prediction that the lowest twisted excited singlet state is zwitterionic (ref.
82
98) was not borne out by results of more advanced MO calculations (refs. 96, 99-102) which indicated the neutral biradical doubly excited state to be lowest in energy at the perpendicular geometry. Experimental justification for the latter result was provided by comparison of calculated spectral transitions for twisted styrene with observed transitions in the benzyl radical (refs. 96, 101, 102). Especially important with respect to the mechanism of photoisomerization is Nebot-Gil and Malrieu's demonstration that the lowest twisted zwitterionic state correlates with the lBU state as stilbene-1 ike molecules twist toward planar geometries, while the neutral twisted state correlates with a doubly excited vertical lAg state (ref. 96). This picture is consistent with the Orlandi-Siebrand and Hohlneicher-Dick explanations for the barrier to trans cis stilbene photoisomerization, Figures 2c and 2d, respectively, provided that the excited twisted intermediate is nonpolar in character. Presumably the analogous neutral twisted singlet excited state should exist in tetraphenylethylene and its energy should be relatively lower than in stilbene because increasing conjugation is expected to stabilize neutral biradical statgs more than ionic states (refs. 44, 95, 96). In view of these considerations, Schilling and Hi1 inski 'S observations assume increased importance, because, if correctly interpreted, they indicate that contrary to theoretical prediction, torsional relaxation in 10 occurs adiabatically by twisting in the lBU state directly to a zwitterionic singly excited Ip* intermediate. Accordingly the phantom excited singlet intermediate will be designated as 'p*(*) in the following. Obviously, in spite of all of the new knowledge obtained with advanced instrumentation, important details of the mechanism of cis-trans photoisomerization are still in question.
-.
83
3.2.2
Gas Dhase studies
It is clear from the above that torsional motion along the olefinic bond of It* and lc* isomers provides the major radiationless deactivation pathway for olefins which isomerize within the singlet manifold. This viscosity dependent motion leads to Ip*(*) and is the key step in the photoisomerization sequence. For stilbene, it was inferred from solution studies that the It* -* 'p*(*) process has an internal energy barrier of 3.5 kcal/mol while the k* -* 'p*(*) process is barrierless. Before discussing medium viscosity effects on this motions we describe the results of gas phase investigations which concern the behavior of isolated excited stilbene molecules. These studies generally confirm the above picture of stilbene photoisomerization and reveal a wealth of new details on the dynamics of vibrational and torsional relaxation.
-
Subnanosecond time scale transient absorption measurements in dilute stilbene vapors were first reported by Hochstrasser's research group (refs. 103, 104). The short lifetimes of stilbene excited singlets ensure that measurements were carried out under collision free conditions. The total energy content of excited molecules remains constant over the measurement time period and changes reflect internal reorganization of this energy amongst internal degrees of freedom (refs. 103, 104). The diffuse transient absorption spectrum obtained at 390 K 10 ps after excitation of trans-stilbene at 265 nm, 7 = 15 f 4 ps, was measurably blue-shifted from that obtained by 520 nm, were excitation at 287 nm, 7 = 55 f 5 ps, and both spectra, Xmax significantly blue-shifted from the thermalized solution spectra described in the preceding section. A slightly narrower spectrum was obtained under identical conditions from trans-stilbene-dl2 excited at 265 nm, 7 = 19 f 3 ps. Relative fluorescence quantum yields for It*-do and lt*-d12, calculated by dividing the fluorescence signal by the relative extinction coefficient for different Xexc were fitted to the measured 1 ifetimes to estimate their dependence on excess vibrational energies up to 6,000 cm-I (ref. 104). Though it was' concluded that the increase of the lifetime following 265 nm excitation upon perdeuteration is less than 15% at any energy of excitation in the 32,000 to 39,500 cm-l range, examination of Figure 8 in reference 104 indicates up to a factor of 2 increase. The large decrease in fluorescence quantum yield with excess vibrational energy is consistent with the change in transient absorption lifetime, for 265 and 287 nm excitation. Decay rate constants for It*-do and ltx-d12 increase with excess vibrational energy, the increase becoming steeper beyond 1,500 - 2,000 cm-1. These observations established the presence o f a significant potential energy barrier on the isomerization
84
pathway in isolated It* molecules, without defining well the magnitude of the barrier height (ref. 104). In the absence of excess vibrational energy the decay rate constant approaches the radiative rate constant, k f . The absence of transient absorption following 265 nm excitation of cis-stilbene vapor led to an upper limit of 1 ps for the lifetime o f k* in the gas phase (ref. 104). Much better definition of the internal barrier height for twisting in It* was achieved by investigations of the spectroscopy and fluorescence decay of trans-stilbene vapor cooled in supersonic jet expansions. Under these collision-free conditions molecules can be excited into specific vibrational modes of the excited state, and it is possible to determine the rate of intramolecular vibrational energy redistribution ( I V R ) which is crucial to photoisomerization. The initial important results by Zewail and coworkers of time-resolved fluorescence spectra and fluorescence lifetimes as a function of excitation wavelength (ref. 105) have been refined, complemented and extended in a series of subsequent publications by the same group (refs. 106-108) and by Jortner and coworkers (refs. 109-111). Fluorescence and fluorescence excitation spectra were well resolved and allowed assignment of specific frequencies to vibrational modes (ref. 106) with theory as guide (ref. 112) and without perturbing complications of medium or lattice interactions which were present in previous condensed phase studies (refs. 21, 113-118). Especially significant in this connection is the observation for the first time of low frequency torsional modes which were easily perturbed by the medium (ref. 106). In agreement with previous solution and vapor phase studies (refs. 21, 114, 115, 118, 119) it was concluded that ground state trans-stilbene, which has a propeller-like geometry due to out-of-phase rotation of the phenyl groups out o f the molecular plane, undergoes significant geometry change upon excitation, presumably due to a shift of double bond character from the central bond to the adjacent C-phenyl bonds (ref. 106). Relevant to the photoisomerization are (a} the observation of pronounced broadening of dispersed fluorescence spectra with increasing excess S1 vibrational energy, Ex, until the spectra become totally diffuse for Ex 2 1200 cm-l, close to the barrier height for isomerization (ref. 107), and (b) the rapid decrease in fluorescence intensity for Ex > 1500 cm-l revealing the competition of isomerization with radiative decay (ref. 107). Historically, following Dyck and McClure’s pioneering study (ref. 21) there has been a tendency to assign a large frequency to the olefinic C,C stretching mode. The initial assignment, 1599 cm-l, led to erroneous conclusion that the central bond retained its high bond order following excitation so that isomerization in It* would have a prohibitively high activation energy (ref. 21). Work with deuterated stilbenes led to a lower frequency assignment (ref. 120). No
85
such high frequency band was prominent in either the absorption (ref. 109) or the fluorescence excitation spectra (ref. 106) of supersonic jets of transstilbene, although a very high frequency was tentatively assigned based on the latter (ref. 106). Applying the corollary of the Dyck and McClure argument (i .e., the low torsional barrier reflects a low stretching frequency) requires that this assignment and a similar one based on the Raman spectrum of It* (refs. 121, 122) must be incorrect. Elegant demonstrations of the presence of the internal barrier to the It* -, lp*(*) radiationless transition are the variations of fluorescence decay rate constants (refs. 105, 107, 108, 110) and fluorescence quantum yields (refs. 109, 123) for jet-cooled trans-stilbene with E x , the energy in excess to 0,O excitation. Under collision-free conditions and in the absence of thermally populated higher ground state vibrational levels, Ex provides the only energy source for overcoming the barrier toward radiationless It* lp*(*) transition. Decay rate constants for lt*-dO and lt*-d12 vs. Ex, Figure 6, are independent of Ex for Ex 5 1200 cm-1 (3.4 kcal/mol) but increase +
0
0 0 0 0
+* + *
o+*+pQ*o
1000
+
0,
I
2000
3( 30
Fig 6.
Measured fluorescence decay rates of trans-stilbene-do, 0, and trans-stilbene-dip, t, as a function of excess energy in S1. Reprinted from ref. 108 with permission of the copyright holder, the American Institute of Physics.
monotonically as Ex is increased beyond this threshold value (refs. 105, 107110). The correspondence of this energy with the activation energy for It* -, 'p*(*) twisting in low viscosity hydrocarbon solutions (refs. 4, 20-27) is striking and not likely to be coincidental. Below the threshold energy,
86
the decay is essentially all due to radiation and corresponds to the radiative lifetime (l/kf) = 2.7 f 0.1 ns (refs. 105, 107, 108) or 2.50 t 0.10 ns (ref. 110) for lt*-do, and 2 . 5 f 0.1 ns for lt*-d12 (ref. 108) consistent with the observation of unity fluorescence quantum yields for fx L 900 cm-l (refs. 109, 123). These radiative lifetimes of isolated molecules are longer than the 1.7 ns (refs. 4, 27, 50), of lt*-do in hydrocarbon radiative lifetime, solvents due, at least in part, to the dependence of kf on the refractive index of the medium (refs. 4, 51, 124). A slight decrease in $f in the 9001300 cm-l Ex region is followed by a sharp decrease at higher Ex values (6f-l linearly dependent on Ex) (ref. 109) which is, within experimental uncertainty, in quantitative agreement with the dependence of 7f-l on Ex (refs. 105, 107, 108, 110). The beginning of the intermediate Ex region, 900 cm-l, coincides with an apparent onset of I V R which is reflected in the emergence of a broad emission continuum superposed on the resolved fluorescence spectrum and is related to the increase in the density of accepting modes with increasing Ex (refs. 107-110). Broad spectra are also observed at lower Ex when S1 levels involving combinations with low frequency modes are excited (ref. 107). In the high Ex region decay rate constants for lt*-d12 are slower than those for It*-dO by a factor of 2, Figure 6 (refs. 107, 108).
-
-
Subtraction of the radiative rate constants kf from the decay rate constants in Figure 6 gives the dependence of ktp, the rate constant for It* -,'p*(*) transition, on Ex which is amenable to treatment by statistical theories on rate constants for unimolecular reactions. Both the classical RRK s- 1
(4)
formula where Eo is the barrier height, s is the number of effective oscillators and A s a frequency factor (refs 109, 125), and the nonclassical RRKM formula (ref. 126)
where h is Planck's constant, E I ~ ( E x - E 0 ) is the number of vibrational states in the transition state having energies less than or equal to Ex-€, and p(fx) is the density of vibrational states of the molecule at energy Ex (refs. 107, 108, 125, 127, 128) have been used. It is now recognized that though good fits can be based on eq. 4, they generally underestimate both A and Eo. Although details of the application of eq. 5 are beyond the scope of this Chapter, we will describe the results from the two approaches employed.
87
Zewail and coworkers selected Eo = 1200 cm-l based on the visual threshold energy in Figure 6 and Vr = 400 cm-l for the reaction coordinate frequency based primarily on Warshel's calculation (ref. 112) of the relevant torsional fundamental frequency (refs. 107, 108). Furthermore, they assume that, with the exception of Vr, the frequencies of all the normal modes do not change in going from It* to the transition state for twisting to lp*(*). Although calculated rate constants predict the deuterium isotope effect nicely, they overestimate the absolute magnitude of ktp by nearly an order of magnitude (refs. 107, 108, 125). The discrepancy was accounted for by introducing a nonadiabatic transmission coefficient in eq. 5, which would be expected if the mechanism in Figure 2b applied and the ZIA*/lB* crossing were not completely avoided (ref. 108). Alternatively, the discrepancy could be accounted for in terms of the Z1A*/1B* equilibrium constant in Figure 2d. The second approach, employed by Troe, is based on an optimized RRKM fit in which Vr = 88 cm-l is used, and Eo and a scalar factor F which adjusts activated complex vibrational frequencies relative to It* frequencies are treated as adjustable parameters (refs. 127, 128). An excellent fit for the lt*-do ktp's was obtained for Eo = 1300 cm-l, V p = 88 cm-l and F = 1.2. The choice of Vr = 88 cm-l, based on an assumed value for cis-stilbene by Greene and Farrow (ref. 129), has been questioned (ref. 108). Frequencies of 403 cm-l and 560 cm-l have been observed in resonance Raman spectra of cis-stilbene in cyclohexane and have been assigned to out of plane torsional modes involving the olefinic double bond since they are close to calculated frequencies for these modes (ref. 130). Since Eo and F are not strictly independent parameters, an uncertainty of f 50 cm-1 was estimated for Eo (ref. 127). The Troe approach eliminates the need for seeking a mechanism which accounts for a less than unity transmission coefficient in eq. 5, but introduces parameters which may or may not have physical significance. A fundamental assumption of RRKM theory is that vibrational energy is rapidly and statistically partitioned among all accessible vibrational states, of the molecule following excitation (ref. 126). Though time resolved fluorescence spectra demonstrate that IVR occurs on a time scale of 20 - 50 ps for Ex as low as 1250 cm-1, they do not establish whether the energy flow statistically accesses all available vibrational states (ref. 108). That vibrations other than CC torsional modes may play a key role along the lp*(*) reaction coordinate is illustrated by the observation by Caldwell It* et al., of a "quasiprimary" deuterium isotope effect for vinyl H substitution in the thermal trans cis isomerization of trans-1-phenylcyclohexene in heptane solution, k ~ / k =~ 2.0 (ref. 131). This unusual effect was rationalized by assuming that the entire Ce-H (Ce-0) out of plane bending
-
-.
88
frequency is lost in reaching the transition state for the trans -,cis reaction (ref. 131). Analysis of the vibrational structure of fluorescence excitation and dispersed fluorescence spectra of jet-cooled trans-stilbene and trans-stilbene-a-dl indicates essentially vibrationless molecules to be planar in both So and S1 states (ref. 132). However, vibrationally excited molecules are expected to assume propeller-1 ike geometries due to excitation of torsional modes of the floppy phenyl groups about the ethylenic carbon-phenyl bonds in the ground state. It could therefore be argued that for propellerlike trans-stilbene and non-planar cis-stilbene (ref. 130), following excitation, the principal displacement 1 eadi ng to i somerizati on involves the ethylenic hydrogens coming out of the CICe=Ce' (or C@e'CI') plane and moving toward the plane of the phenyl rings. If this motion were sufficient to reach the transition state for twisting in the gas phase, a much larger Vr would have to be employed in the RRKM calculations, than was employed by either Troe's or Zewail's groups (see above). Furthermore, the differences in It* 'p*(*) rate constants obtained upon perdeuteration may reflect, at least in part, a somewhat higher Eo value for lt*-d12 due to the difference in zero point energies. A recent study has indeed revealed a striking positional dependence of the deuterium isotope effect on It* fluorescence lifetimes measured in a supersonic expansion, in the thermal gas phase, and in solution (ref. 133). In the jet, almost all, and in thermal gas samples or in solution, all the effect of full deuteration is obtained by deuteration of only the two ethylenic positions. For example, in n-hexane or methanol solution, the decay rate constants of It*-dO are indistinguishable from those of lt*-dlO at several temperatures, but are systematically 1 arger than those for lt*-d2 and lt*-d12 by a factor of 1.5 (ref. 133). The results for jetcooled stilbenes suggest a significantly higher threshold energy for isomerization of It*-d2 than for It*-do or Itf-dlo (ref. 133); consistent with the conclusion that the change of zero-point energy of It* is reflected in an increased barrier height only when olefinic positions are deuteriated. This concl usion is tentative, however, because Arrheni us activation energies for twisting rate constants in solution appear to be independent of deuterium substitution within experimental uncertainty (ref. 133). The special role of the olefinic hydrogens in overcoming the torsional barrier is reminiscent of their role in promoting the radiationless decay of twisted stilbene triplets, 3p* (refs. 134, 135).
-.
-
As a test of the assumption of rapid and complete energy randomization over all accessible vibrational modes, the dependence of fluorescence decay rates on Ex was studied for supersonic jets of trans-stilbenes substituted at the 4 position with methyl, ethyl and n-propyl groups (refs. 110, 136). It
89
was reasoned that, according to RRKM theory, the introduction of alkyl groups should markedly decrease ktp (35% for methyl) due to the increase in the total number of vibrational degrees of freedom. In the first of these studies, ktp for the methyl derivative was observed to begin to contribute at the same threshold Ex, but contrary to RRKM theory, it was generally 15% larger than corresponding ktp values for trans-stilbene (ref. 110). This was interpreted as an invariance of ktp with methyl substitution, and it was concluded that "the I V R pertinent to the nonradiative process does not involve the vibrational degrees of freedom of the distant CH3 group'' (ref. 110). Neglecting the systematic increase o f ktp, it was suggested that RRKM theory, or any other statistical theory, could still be applied by considering only the subset o f the vibrational degrees of freedom of the trans-stilbene moiety (ref. 110). This conclusion was criticized by Troe, who showed that predicted values in good agreement with the observed ktp for the methyl derivative could be obtained by including all vibrational degrees of freedom but reducing the scaling factor paramater f from 1.2, the best fit value for trans-stilbene, to 1.0 (ref. 127). Extension of this work to the ethyl and n-propyl derivatives yielded even more striking deviations from theoretical expectation. Nearly identical ktp values were obtained for the two derivatives starting at the somewhat lower threshold Ex = 1000 cm-1 value, which again, contrary to theory, are substantially larger (- 70%) than corresponding values for transstilbene (ref. 136). In interpreting these data the approach of Troe was adopted in which all vibrational modes are utilized in the RRKM treatment, but the data are fit by using E0 and F as adjustable parameters (ref. 136). Good fits were obtained for f = 1.2 and Eo = 1100 f 100 cm-l for trans-4ethylstilbene and fo = 1000 f: 100 cm-1 for trans-4-n-propylstilbene, and it was concluded that the observations cannot be taken as evidence of a breakdown of RRKM theory (ref. 136). We suspect that this result illustrates the weakness of the adjustable parameter RRKM approach since it does not seem reasonable that the torsional barrier should be so sensitive to such minor substituent changes at a position considerably remote from the reaction site. It should be noted that Hammett u values are essentially identical for the alkyl groups employed (ref. 137, 138). If Eo were sensitive to substituent changes in the 4 position, a much larger effect would have been expected for chlorine substitution, yet kpt values for trans-4-chlorostilbene fall, within experimental uncertainty, on the same line as those of trans-stilbene (ref. 110). The only difference obtained for 4-chloro substitution appears to be the opening of an additional Ex independent decay channel which is attributed to heavy-atom induced intersystem crossing and is evident in the reduction of df from unity to 0.65 t 0.03 for Ex below the threshold energy (refs. 110, 111).
-
90
There has been one sub-picosecond time resolved study on cis-stilbene vapor at room temperature, Torr, in which 1c* was generated with a 312.5 nm (corresponding to Ex =: 50 cm-l), 0.25 ps pulse and its decay monitored by multiphoton ionization employing a 625 nm probe pulse (ref. 129). A single exponential decay, 7 = 0.32 ps, was observed which was associated with motion along the torsional coordinate, unimpeded by any appreciable barrier (ref. 129). A small residual ion current which remained constant over a 10 ps time period could be due to trans-stilbene present in the sample as an impurity (ref. 129). The possible contributions of dihydrophenanthrene (refs. 59-61) and phenanthrene formation (ref. 139) as decay channels for lc* were not considered. 3.3
Medium effects
The data presented in the preceding sections suggest that in hydrocarbon solvents of low viscosity the Arrhenius activation energy for rotation in It*, Etp, eqs. 2 and 3, coincides with the intrinsic barrier height, E o , eqs. 4 and 5, established for isolated It* in the vapor phase (Etp = Eo t RT, see ref. 125). In more viscous media the slower rate at which the solvent cage is capable of rearranging imposes an additional barrier to twisting in It* SO that the effective Arrhenius activation energy fobsd is greater than Etp. When the polarity, or polarizability of the medium is changed, possible changes in the relative energies of the relevant excited states may affect the barrier height for rotation and must also be considered. Several studies on trans-stilbene which attempt to separate these effects are reviewed below. Also described are the relatively few observations concerning the effect of the medium on the decay of the much shorter lived k*. 3.3.1 The df values of flexible molecules which can undergo facile torsional displacements in the lowest excited singlet state generally increase strongly with increasing medium viscosity. The presence of a medium imposed barrier on the potential energy surface for twisting about the central bond in It* was proposed by Becker and Kasha (ref. 140) but systematic studies of the medium dependence of Of in stilbenes were first reported by Fischer and coworkers (refs. 22-24). We begin by comparing some of the approaches which have been used to account quantitatively for the pronounced medium effects. The model employing a medium enhanced barrier to rotation, Figure 7, was
91
Fig. 7 Apparent enhancement of the It* -, 1p* torsional barrier for stilbene by a viscous medium. Adapted from ref. 25. first used to treat fluorescence data by Saltiel and D'Agostino (ref. 25). A slightly modified form of this treatment follows. In high viscosity regions where the slow motions of a medium restrict motions associated with radiationless decay of the excited solute, the medium imposes a characteristic barrier Ev on the motion of the solute. The observed rate constant under such conditions, kobsd, can then be expressed as kobsd
=
avkte
- Ev/R7
where a, is also medium-characteristic. At "zero" or low viscosities EV and a, approach zero and unity, respectively, and kobsd = kt, the inherent rate cqnstant for the unimpeded motion which may or may not.be associated with an intrinsic barrier, Et. For the specific case of trans-stilbene, kt is given by ktp in eqs. 2 and 3 and substituting Etp = Et and Atp = At gives This equation accounts for the adherence o f kobs to the Arrhenius equation and ,Eobsd = Et t Ev. For example, in the viscous medium, giving Aobsd = glycerol, plots o f ln(l/Qf-ti) vs. T-l, where ti = 1 + kis/kf, give fobsd Values of 9.7 and 12.3 kcal/mol and log Aobsd values of 16.3 and 18.5 for transWhen stilbene and trans-1,l'-biindanylidene, 9, respectively (ref. 25).
92
medium v i s c o s i t y i s v a r i e d by varying t h e temperature i n a s i n g l e solvent and t h e shear v i s c o s i t y , q s , obeys t h e Andrade equation I n qs = I n Arl
+ frl/RT
(8)
where Eq i s the a c t i v a t i o n energy f o r viscous flow, eq. 8 can be used t o express kobsd as a f u n c t i o n o f qS
I n kobsd = I n Aobsd t (Eobsd/Eg)ln Aq
-
(Eobsd/Eq)ln ‘Is
(9)
When medium v i s c o s i t y i s v a r i e d a t constant T by varying t h e medium (e.g., r e f . 25) o r t h e pressure (refs. 141, 142) p l o t s o f l n kobsd vs. I n qs a r e n o t expected t o be l i n e a r over a l a r g e v i s c o s i t y range since such changes are expected t o vary Eq and A. which, i n turn, a f f e c t the magnitude of av and Ev (ref.
25).
However, assuming t h a t k t remains constant f o r small medium
changes, t h e r e 1a t i o n s h i p I n kobsd = I n a v k t + (Ev/Erl)ln Arl
-
(€,/Eq)ln
rls
(10)
can be used t a n g e n t i a l l y t o estimate E, f o r a s p e c i f i c medium. The successful a p p l i c a t i o n o f eq. 10 was demonstrated f o r t r a n s - s t i l b e n e and f o r 9 by changing t h e v i s c o s i t y o f g l y c e r o l through a d d i t i o n o f small amounts o f water (up t o 5%) ( r e f . 25). Theoretical j u s t i f i c a t i o nf o r t h e medium modified p o t e n t i a l energy diagram i n Figure 7 has been based on t h e Born-Oppenheimer approximation f o r the s o l u t e and i t s solvent cage ( r e f . 143). The parameters ZV and a, can be r e l a t e d t o thermodynamic p r o p e r t i e s of t h e so!ute/solvent cage system ( r e f .
144). We assume t h a t solvent organization around t h e substrate, lt*, i s random w i t h respect t o t h e geometric requirements o f t h e t w i s t i n g process, and
t h a t a d i s t i n c t population o f I t * i n solvent cages, S’, which do n o t r e s t r i c t r o t a t i o n , i s i n e q u i l i b r i u mw i t h It* molecules which occupy solvent cages, S, unfavorable t o isomerization. The o v e r a l l t w i s t i n g process can accordingly be represented by KS
( l t * S ) 3 (‘t*S’)
kt -b
and t h e e f f e c t i v e t w i s t i n g r a t e constant i s kobsd = Kskt. It f o l l o w s t h a t aV = e and Ev = AHv, where ASv and AHv are t h e entropy and enthalpy changes f o r t h e solvent cages t o achieve c o n f i g u r a t i o n sfavorable t o t h e geometry change r e q u i r e d f o r r a d i a t i o n l e s s decay ( r e f . 144).
It should be noted t h a t
t h e S a l t i e l and O’Agostino approach does n o t r e q u i r e t h e use o f t h e Andrade equation i n d e r i v i n g t h e dependence of kobsd on q s , eq. 9. Other empirical
93
relationships, e.g., the Litovitz equation (ref. 145). log qs
=
A
t
B/T3
(12)
could be used to substitute for T in the Arrhenius equation. The approach of Fischer and coworkers (refs. 23, 24) is based on the dependence of qs on Vf, the mean free volume of the solvent (refs. 146, 147) 7
where Vo is the critical free volume for a motion of the solvent. By assuming that the additional free volume required of the solvent which allows the rearrangement of a solute is a fraction a of Vo, a 1, the rate constant for a medium dependent rearrangement is given by k
=
koe-avO/vf
(14)
which combined with eq. 13 gives In k
=
In ko
t
aln qo - aln qs
=
B
- aln
qs
Equation 15 has the same form as eqs. 9 and 10 and, by analogy, the Fischer parameter a corresponds to the Ev/Eq parameter in the Saltiel and D’Agostino equations. What was not recognized in the Gegiou et al., paper (ref. 23) was that since q s was varied by varying the temperature, the slope of the In k VS. In q s plot gives (Et t Ev)/Eq, see eq. 9, which reduces to the desired parameter a only when the rearrangement of the solute has no inherent activation barrier, i . e . , Et = 0, and is strictly controlled by the medium. It is probably for this reason that Sharafi and Muszkat limited the application of eq. 15 to cis-stilbene and sterically congested trans-stilbenes whose negligible fluorescence quantum yields in media of low viscosity suggest that the Et = 0 condition is fulfilled (ref. 24). A variation of eq. 15 which factors out the temperature dependence of ko, the intrinsic rate constant in the absence o f medium constraints, for Et # 0 has been applied by Velsko and Fleming (ref. 148). Approaches which treat the viscosity dependence of the diffusional motion of part o f the molecule have also been employed. A relationship proposed by Oster and Nishijima is based on the Stokes-Einstein equation and assumes a 7 / q s dependence for the intramolecular solute motion (ref. 149). A theoretical derivation o f this behavior has also been presented (ref. 150). The limited success o f this expression (ref. 25) is probably mainly due to the
94
A restriction of €,,/,FV in eqs. 9 and 10 and a in eq. 15 to unity. new relationship, based on a model in which parts of the molecule rotate to equilibrium positions after excitation at a rate constrolled by Stokes- ike damping, was derived by Forster and Hoffman (ref. 151). It predicts an vS 2/3 dependence for l/bf of molecules whose fluorescence is due entirely to medium imposed frictional hindrance of motion and thus restricts and a to a value of 2/3. This theory a1 so predicts nonexponential fluorescence decay contrary to experimental observations (refs. 141, 150) and has received little use (refs. 150, 151, 153). The most popular theory in recent years i s due to Kramers (refs. 154, 155) and was first applied to stilbene It* twisting by Hochstrasser (ref. 70). The Kramers expression for diffusive barrier crossing is
where w and w' are the frequency of the initial well and the curvature at the top of the barrier, respectively, E0 is the height of the barrier and 8 , the reduced friction coefficient is, in accord with the hydrodynamic model, frequently taken as proportional to the shear viscosity of the solvent. At the limit of sufficiently high viscosity, i.e. for very large B , eq. 16 reduces to its Smoluchowski limit (see, e.g., refs. 84, 156, 157)
The Kramers model adopts the concept o f Brownian motion in describing the motion of a molecule, in a one-dimensional potential well, subject to friction and fluctuating forces caused by interaction with the medium. The solute molecule, randomly acted upon by the surrounding solvent molecules, can cross an energy barrier along the reaction coordinate only when fluctuations in the solvent create a net force in that direction. In the case of stilbene, the rotation of the phenyl group about the central bond axis is taken as the onedimensional motion. The different models employed to account for viscosity effects are conceptually similar. The motions of the solute are restricted by a solvent cage and the magnitude of the effect depends on the requirements of the geometry change of the solute and on the rigidity of the solvent. The free volume model focuses on the volume required for the isomerization to occur. The ease with which solute molecules can attain that critical volume is determined by the ability of the solvent to provide it, that is, by the free volume o f the solvent which is related to its viscosity. The equilibrium model is also based on the assumption that only a fraction of solute molecules
95
can escape solvent restrictions along the reaction coordinate. However, instead of limiting itself to critical volume considerations, this model, more generally defines the fraction of the molecules which can proceed along the reaction coordinate in terms of thermodynamic equilibrium. In Kramers' model the isomerization rate is determined by the fraction of solute molecules which, in their random interaction with the solvent, experience a sufficient net force to overcome the "friction" of the solvent along the reaction coordinate. It is not surprising therefore that there are strong similarities between the resulting relationships when certain boundary conditions are fulfilled. Most recent studies of medium effects on the torsional relaxation of excited stilbenes and related molecules have been carried out in n-alkanes (refs. 84, 148, 158-164) or in the homologous series of primary alcohols (refs. 67, 148, 156, 157, 161, 162, 165-170). Viscosity was changed either by changing temperature (refs. 23-27, 148, 157, 158, 162, 163, 166, 170) or pressure (refs. 141, 142) in a single solvent or solvent mixture, or by varying the solvent at constant temperature and pressure (refs. 25, 84, 156, 157, 164, 166, 167, 169, 170). Arrhenius behavior was observed in most cases where the temperature was changed without changing the solvent in accord with eq. 7. Clearly, in those viscosity regions for which the Andrade equation holds, eq. 9 is also obeyed. Since in the case of alcohol solvents changes in polarity and polarizability appear to affect the height o f the intrinsic thermal barrier to rotation, Eo or Et, those results will be considered in the next section. We consider here studies in alkane solutions and studies in the vapor phase in the presence of alkanes which attempt to bridge the gap between the isolated excited state molecule and the same molecule in an alkane solvent (refs. 158, 171-175). Arrhenius plots of ktp for It* of trans-stilbene in n-alkane solutions (refs. 26, 158-161, 163) and in hydrocarbon mixtures (refs. 4, 72) give Eobsd values in the 3.5 - 4 . 9 kcal/mol range. Assuming that Eobsd = Et t aEV, i.e., letting Ev = aEV in the Saltiel and D'Agostino model, Sundstrom and Gillbro employed a limited and somewhat imprecise set of data to reach the erroneous conclusion that Et decreases as the chain length of the n-alkane solvent is increased (ref. 159). Application o f transition state theory on a combined set of all available data from several laboratories gives A d O b s d values for internal rotation in the n-alkane series which adhere closely to (c2-c16)
mfobsd where
=
Ado =
t dEq
2.85
t
(18) 0.04
kcal/mol and a
=
0.39 f 0.02, Figure 8 (ref. 144).
96
t
a
h
5.0
I (a)
4.51
-
E
\
L )
2
0.0
2 1.0
2.0
.
3.0
E, (kcal/mol)-
5
4.0
2 5.0
Fig. 8 AHH+,bsd and ~ ~ + , b ~dependence d on E,, for n-alkanes; from ref. 144 with permission of the copyright holder, the American Chemical Society. This Ado value is also obtained more precisely from isomicroviscosity transition-state-equation plots (see below) and represents a better estimate of €, the intrinsic potential energy barrier for twisting in It*. The value of Eo = 3.5 kcal/mol used in most previous treatments is based on the Arrhenius activation energy and differs from Ado, and hence E0, by RT. Consequently, experimental preexponential factors in he 1 iterature are overestimated by the factor e. A plot of ln[(xk/h)ave ~ 5 ’ 0 , vs. E,, is also shown in Figure 8. This plot which is analogous to the isokinetic plot in Figu e 8 of ref. 25, allows extrapolation to E,, = 0 and a, = 1 and gives &eAS d R =0.47, where IC is the transmission coefficient. It will be shown below that this value defines the preexponential term for the maximum intrinsic twisting rate constant in thermally equilibrated molecules whose motions are not impeded by the medium (ref. 144). Since RRKM fits described above have placed Eo for the jet-cooled isolated stilbene molecule in the narrow range of 3.4 kcal/mol (refs. 107, 108) to 3.7 kcal/mol (refs. 127, 128), passing from the vapor phase into the n-alkane medium appears to cause a significant decrease of 0.5 - 0.8 kcal/mol in fo. The likelihood that such a change occurs was inferred by Troe and coworkers (ref. 172) from the observation o f a substantial (- 1200 cm-1) red shift in the UV absorption
c
97
spectrum of trans-stilbene in passing from the medium of dilute gaseous ethane to liquid ethane. Adopting the Orlandi and Siebrand model in Figure 2b, these authors proposed that preferential stabilization of the I€!,* state lowers the energy of lBU*/lAg* crossing and results in a lower barrier to twisting (ref. 172). Changes in the nonradiative decay rates of It* of trans-stilbene and of 9 at 20 - 23 "C in the n-alkane (C4 to c16) solvent series served as an important early test for the applicability of Kramers' equation (ref. 84). The reduced friction coefficient B which is equal to the ratio of (r, the friction coefficient for the twisting of the aryl group, to Z, i t s moment of inertia, was related to the solvent's shear viscosity using the hydrodynamic re1 ationshi p
= +,
-
or
where Q = 4 for the slip boundary condition (usually employed) and Q = 6 for the stick boundary condition (refs. 176, 177). Equation 19 is based on the assumption that the twisting motion of the stilbene molecule is equivalent,to translational diffusion o f a molecular group o f radius r along a curved pathway defined by the radius of gyration rg. Accordingly the friction coefficient for twisting, cr, equals rg2ct where {t is the friction coefficient for translation and since Z = mrg2, where m is the mass of the twisting group, it follows that the reduced friction coefficient for rotation is assumed equal to the reduced friction coefficient for translation, i.e., Br = bt (ref. 178). Substitution of eq. 19 into eq. 16 gives where A
=
(4271) and B
=
(2mo'/arr) and substitution into eq. 17 gives
98
kobsd
[(AB)/(29s)le
-Eo/RT
(21)
the Smoluchowski-Stokes-Einstein limit of Kramers' equation. Equation 21 has the same form as eq. 10 provided that EV = Eq. Thus the Smoluchowski-StokesEinstein 1 imit of Kramers' equation requires that the entire activation energy for viscous flow of the medium, EV, be added to the intrinsic barrier height, Eo, rather than a part of it as is usually observed. It was found that for trans-stil bene the nonradiative decay rate constants (the small contribution of kis was neglected) in the C4 to c16 alkane series could not be fitted by either eq. 20 or eq. 21 (ref. 84). Use of eq. 20 to fit the rate constants in the lower viscosity alkanes, Figure 9, leads to underestimation of the
Fig. 9 The Kramers hydrodynamic approximation fit, eq. 20, of the photoisomerization rate constants of trans-stilbene in n-alkane vs. shear viscosity. Reprinted from ref. 160 with permission of the copyright holder, the American Institute of Physics. observed rate constants for the higher viscosity alkanes (ref. 84). The nonradiative rate constants for trans-1,l'-biindanylidene, 9, on the other hand (no doubt fortuitously) seemed to follow eq. 21. The derivation of Kramers' equation is based on the assumption that the time scale of the motion of the solvent molecules is much faster than the time scale of the change in configuration of the solute (ref. 84). This model has been modified by Grote and Hynes by including correlated solvent response to the motion of the solute which i s expected when the time scale for barrier crossing by the solute i s comparable to the time between collisions o f the
99
solute with the solvent (refs. 84, 178-181). The modification leads to a frequency dependent friction coefficient. Although the trans-stilbene are fit well by this model, the resulting radiationless decay rate constants parameters make no physical sense (ref. 84). Other refinements of Kramers' model have included extension to the multidimensional case (refs. 182, 183). As with the original Kramers' equation, application of the refined models to specific cases requires the assumption of a hydrodynamic model for solvent friction (refs. 84, 156, 158, 184, 185), and leads to equations having the same form as eqs. 9 and 10, or as more usually expressed (ref. 158)
Though, empirically, eq. 22 fits the stilbene rate constants in the alkane series at constant T well (ref. 158), comparison with eq. 10, which has the same form, shows that B includes solvent specific parameters En, AV and the related a, and E, which change from solvent to solvent (for EV, AV values in alkanes see, e.g., ref. 186). In fact, it is clear that any treatment which utilizes shear viscosity as a measure of microscopic friction in accounting for the diffusion of the rotating group in different media is doomed to failure as has been well established for translational diffusion in bimolecular diffusion-controlled reactions (refs. 186, 187). On the other hand, the success of the Smoluchowski relationship in accounting for bimolecular rate constants when empirical diffusion coefficients are employed suggests that proper tests of Kramers' equations 16 and 17 must also be based on empirical measures of microvisdosity rather than on Stokes-Einstein type approximations of the friction coefficient, i.e., eqs. 20 and 21. A corollary to this conclusion is that the expected and demonstrated failures of eqs. 20 and. 21 to account for isomerization rate constants in a solvent series invalidate the isoviscosity Arrhenius plot approach which is often used to estimate fo (e.g., refs. 157-160, 166, 167). Reasoning that rotational and isomerization motions ought to feel frictions which are proportional to each other, Velsko, Waldeck and Fleming first used rotational reorientation times as a measure of microscopic friction in Kramers' equation in an attempt to account for isomerizarion rate constants of a cyanine dye (3,3'-diethyloxadicarbocyanine iodide) in the n-alcohol series (ref. 166). Rotational reorientation times, 'Trot, in the diffusion controlled limit are related to the medium friction coefficient,
100
Assuming that the friction coefficient for rotational motion, Tor, is proportional to the friction coefficient for isomerization, Tr, Tor
=
Podr
(24)
where Por is the proportionality constant, eliminates the need of applying the hydrodynamic approximation for the frictional coefficients, giving directly the Kramers-Hubbard re1 at ionshi p
where A and C are related to w and w', respectively (refs. 160, 163, 166). Fleming and coworkers have proposed that the definition of A , namely A = w/Zn, be modified to A = wQf/2nQ', where Q' and Qf are partition functions for reactant and transition state, respectively (ref. 158). In its first application eq. 25 failed to give a satisfactory fit to the data for the cyanine dye because Tor being inversely proportional to shear viscosity, exhibited Stokes-Einstein behavior. This failure of eq. 25 (ref. 166) was thus equivalent to the failure of eq. 20 (ref. 157). Since the proportionality in eq. 24 did not hold, it was concluded that molecular tumbling and intramolecular motion leading to radiationless decay experience different frictional influences. It was suggested that inclusion of the frequency dependence of the friction coefficient (refs. 176, 178) would probably account for the isomerization results (ref. 166). The proportionality in eq. 24 appears to be much more nearly correct in the case of stilbene, since the Kramers-Hubbard equation leads to significantly improved fits of It* radiationless decay rate constants in the n-alkane series at constant T ; compare Figures 9 and 10 (refs. 160, 161). The two independent experimental tests of eq. 25 led to remarkably similar parameters: w = 3.69 x 1013 s-1 and (w'/por) = 3.42 x 1013 s-l (ref. 160) and w = 3.64 x 1013 s-l and (w'/por) = 3.43 x 1013 s-l (ref. 161). Although these parameters were considered to be much more physically acceptable than those obtained by applying the hydrodynamic approximation for B together with the Grote and Hynes theory of frequency dependence for the friction coefficient (ref. 84), their meaning is somewhat unclear since w may include and possibly a less than unity transmission coefficient, K , while the magnitude of por in the second parameter is unknown. As B 0 the preexponential term in Kramers' equation, eq. 16, reduces to w/2n which
d/Q'
-.
101
Fig. 10 The Kramers-Hubbard fit, eq. 25, of the photoisomerization rate constants o f trans-stilbene in n-alkanes vs. Bar, the reduced friction coefficient for rotational motion. Reprinted from ref. 160 with permission of the copyright holder, the American Institute of Physics. corresponds to the friction-independent rate preexponential term predicted by transition state iheory (see, e.g., refs. 84, 156-158). It follows that IR, except that in the present case, since the Eo value (o/2r) = (nkT/h)e employed was too high by approximately RT, the relationship is (o/2x) = ( ~ k T / h ) e ( ’ ~ ” ’ / ~ )(ref. 144). Substitut‘on of the o values for refs. 160 and 161 into this expression gives rce As’/R = 0 . 3 5 and 0.34, respectively, in good agreement with the value obtained using the Saltiel and D’Agostino model and transition state theory directly (ref. 144). The corrolary to this analysis is that literature o values must be adjusted downward to o = 1.36 x lOI3 s - l and 1.34 x 1013 s - l for refs. 160 and 161, respectively. It is noteworthy that the Kramers-Hubbard equation gives a less satisfactory fit for the radiationless decay rate constants of the “stiff”-diphenylbutadiene, 11,
*’
11 than for trans-stilbene (ref. 160).
Furthermore, the ( o ‘ / p o r ) parameter
102 obtained f o r 11 i s 1 1 - f o l d smaller than t h a t f o r s t i l b e n e leading t o t h e l e s s than s a t i s f y i n g conclusion t h a t t h e p r o p o r t i o n a l i t y constant Por can vary by an order o f magnitude from molecule t o molecule. Use o f 70)" as an empirical measure o f t h e f r i c t i o n c o e f f i c i e n t demonstrates t h e inadequacy o f t h e StokesE i n s t e i n approximation, b u t leaves t h e question o f the importance o f i n c l u d i n g frequency dependent f r i c t i o n u n s e t t l e d ( r e f s . 160, 161). Ihe above studies on t r a n s - s t i l b e n e were extended r e c e n t l y t o include measurements a t d i f f e r e n t temperatures (ref. 163). Though t h e f i t i s improved when eq. 25 i s used instead of eq. 20, t h e t h e o r e t i c a l curve shows l e s s curvature than t h e experimental p o i n t s and t h e f i t i s n o t as s a t i s f a c t o r y as was obtained f o r the constant T data sets, e.g.,
Figure 10 ( r e f . 163).
The
r e o r i e n t a t i o n times, ror, f o l l o w 70r = S ( q / T ) i n each alkane solvent w i t h S d i m i n i s h i n gby almost a f a c t o r of two as t h e s i z e o f t h e alkane i s decreased from n-pentane t o n-hexadecane ( r e f . 163). It was concluded t h a t coupling of s o l u t e w i t h solvent decreases w i t h increasing so1vent:solute s i z e r a t i o . The changes i n Tor w i t h changes i n alkane solvent and T are e n t i r e l y analogous t o p r e v i o u s l yobserved changes i n t r a n s l a t i o n a l d i f f u s i o nc o e f f i c i e n t s , D, and i n bimolecular d i f f u s i o n - c o n t r o l l e dr a t e constants ( r e f s . 186, 187). I n view of t h e importance o f r e l a t i v e s o l u t e size, we were l e d t o examine t h e p o s s i b i l i i t y t h a t t h e e f f e c t s of medium on t h e t r a n s l a t i o n a l d i f f u s i o n o f a benzyl r a d i c a l , or, toluene, would p a r a l l e l those on t h e intramolecular r o t a t i o n i n I t * b e t t e r than the e f f e c t s on t h e tumbling o f t h e e n t i r e transs t i l b e n e molecule ( r e f . 178). The comprehensive studies o f Fischer and coworkers on t h e effects of solvent and T on r a d i c a l s e l f - t e r m i n a t i o n r a t e constants (refs. 186, 187, 189) have provided ample v e r i f i c a t i o n of t h e semiempirical formula of Spernol and W i r t z ( r e f . 190) which r e l a t e s d e v i a t i o n s o f experimental D's from t h e Stokes-Einstein value, DSE, t o s o l u t e (r) and solvent ( r L ) molecular-radius r a t i o s: fl-
kT
6rqrDexptl
= -
%E Dexptl
f = (0.16 t 0.4 r/rL)(0.9 t 0.4 TrL
-
0.25 Tr)
where f i s t h e m i ' c r o f r i c t i o n f a c t o r f o r t r a n s l a t i o n , and t h e parameters TrL and Tr are reduced solvent and s o l u t e temperatures, r e s p e c t i v e l y :
103
where 7 is the experimental temperature and Tbp and Tmp are the boiling and melting points of solvent for T,L and of solute for Tr. The reduced temperature term makes f less solvent- and sol Ute-dependent. Mol ecul ar radii in A are estimated from molar volumes, V , in cm3, using
Spernol and Wirtz diffusion coefficients, Dsw, for isobutane or n-butane in the C7 to c16 n-alkanes have been critically compared with calculated diffusion coefficients based on eight other empirical and semiempirical methods (ref. 186). Noteworthy is the good agreement between Dsw and calculated values (refs. 186, 187), Demp, from the strictly empirical equations developed by van Geet and Adamson (ref. 191), but using more recent experimental n-a1 kane self-diffusion coefficients (ref. 192). The van Geet and Adamson method was developed with n-alkane data and its application to the branched a1 kane, i sobutane, represents somewhat of an extrapol ation. However, since ratios of experimental D ' s in n-hexane at 25 "C for the pairs pentane, isopentane and octane, isooctane are 1.04 and 1.03, respectively (Table 7 in ref. 186), no large errors should be introduced by this approximation. For the C7 to C12 n-alkanes Dsw values are within 5% of Demp in the 295-365 K range. For the C14 and c16 n-alkanes the ratio Dsw/Demp increases systematically from 0.8 to L.0 as the temperature is raised from 300 to 365 K (refs. 186, 187). A direct comparison between experimental D ' s for toluene in toluene or cyclohexane with Dsw values in the 248 - 331 K range reveals maximum deviations of 16% (ref. 193). Accordingly, the Spernol and Wirtz equation was employed to calculated D's for toluene in the n-alkanes at different T ' s . In addition a set of Demp was estimated at 294.5 K by adjusting Dsw values for toluene using factors Demp/Dsw calculated for isobutane. The latter procedure assumes that Dsw values for toluene will deviate from experimental D ' s by the same factors that they deviate for isobutane. Since the factor f corrects the macroscopic friction coefficient to a microscopic one, it can also be used to convert the shear viscosity qs to microviscosity qc( QC( = f q s
(30)
Theoretical justification for this approach is provided by the microfriction theory of Gierer and Wirtz (ref. 194) and Dote et al. (ref. 195). Substitution of r , for ~ ~ q s in eq. 20 is equivalent to using the
104
viscosity experienced by the solute in the Stokes-Einstein relationship. An excellent fit of the kobsd values in n-alkanes at 21 - 23 'C (ref. 160) to eq 31 is obtained with best fit parameters o = 1.35 x 1013 s-l and o' = 1.59 x 1013 s-l (ref. 178), Figure 11. The o value is identical to that obtained using the Kramers-Hubbard equation with stilbene reorientation times as the measure of microfriction (refs. 160, 161). The ratio (U'/por) from ref. 160 and the o' value from the microviscosity treatment of the same data gives Por = 2.2 which probably reflects mainly the difference in the moments of inertia for rotational reorientation of the entire molecule and for twisting of only half o f it (ref. 178).
4 20.0
0.2
0.4
0.6
0.0
1.0
Fig. 11 The Kramers fit of the photoisomerization rate constants of transstilbene in n-alkanes employing microviscosity, qp in eq. 20 instead of shear viscosity (ref. 178). As pointed out earlier, the failure of eq. 20 to account for the data is due primarily to the failure of the Stokes-Einstein relationship (refs. 160, 161) to account for specific solvent-solute interactions. Furthermore, the failure of eq. 20 demonstrates the inappropriateness of using isoviscosity Arrhenius plots to obtain the barrier height Eo. A valid approach to Eo must employ isomicroviscosity Arrhenius plots as was recognized recently by Zeglinski and Waldeck (ref. 162) in a study of medium effects on radi ationl ess decay rate constants for tran~-4,4'-dimethoxystil bene. Isomicroviscosity plots, based on transition-state theory, are shown for trans-stilbene in n-alkanes in Figure 12 for five different constant qfi = f q s values. The slopes of the resulting parallel lines give d o= Eo = 2.89 f 0.07 kcal/mol, in excellent agreement with the value obtained independently using the medium enhanced barrier model (refs, 144, 178). This value of Eo
105
was used in fitting kobsd values to eq. 31. The alternative approach proposed by Zeglinski and Waldeck that employs solvent rotational self-diffusion coefficients as a measure of microfriction failed to lead to an improved fit to Kramers’ equation probably because these coefficients do not take into account differences in solvent-solute molecular dimensions or any other specific solvent-solute molecular interactions (ref. 162). Use of solvent rotational sel f -di ffusion coefficients to define i somi crofri cti on conditions for Arrhenius or transition-state treatments leads to no improvement over the use of shear viscosity for the same reasons (refs. 162, 178).
t
e
n
W
n
n *O
-
U
CT
15.21 2.4
I
3.0
r-‘ x
I
3.6
I
4.2
103(~-9-
\
d
B
Fig. 12 Isomicroviscosity transition state plots for trans-stilbene in nalkanes (ref. 178). Values next to lines are q p ’ s in cP. The upper line represents extrapolation of all other points to qp = 0. An empirical approach for obtaining good fits to the Kramers equation has been proposed by Lee, Zhu and Robinson (ref. 196). An effective viscosity q* is defined
where parameters a and b are specific for each solute in each solvent series. Substitution of q* for qs in eq. 20 yields a three-adjustable-parameter
106
Kramers equation which generally gives very good fits for the viscosity dependence of rotational isomerization rate constants of excited molecules (ref. 196). The empirical microviscosity q* approaches ( q s / a ) at low viscosities and a limiting value b-1 at high viscosities. It can thus be made to mimic the behavior of empirical diffusion coefficients and "correct" for deviations from the Stokes-Einstein relationship. This approach ignores differences in specific solute-solvent interactions as the solvent is changed within a solvent series, and requires the isomerization rate constants in order to determine r)* through fitting. In this respect approaches based on 701" or Dt which use molecular probes to independently define microviscosity are preferable. Demonstration of the self-consistency of the latter two approaches to the treatment of trans-stilbene rate constants was achieved by plotting Tor vs. vp for trans-stilbene in the n-alkane solvent series (ref. 178). In contrast to the analogous vS plot (ref. 160), the vp plot is linear with slope and intercept in excellent agreement with theory (ref. 178). Since the failure of eq. 20 to fit the data is due primarily to the introduction of qs via the hydrodynamic model, attempts to make eq. 20 conform to the observations by using Eo as an adjustable parameter lead to incorrect conclusions (refs. 144, 172). The results in Figure 8 reveal no variation in Eo within the uncertainty limits of the activation parameters for each individual n-alkane. It follows that at best such changes must be minor; a sensible conclusion in view of the minor solvent changes involved. In this section we have, thus far, considered results from studies for which the isomerization of It* can be regarded as a diffusion-controlled torsional motion of the solute. The medium viscosity region where this occurs is known as the diffusion-controlled regime of Kramers' equation, which as we have seen describes variations in the rate constant well when microviscosity is employed. In section 3.2.2, the other extreme of jet-cooled stilbene photoisomerization was described and it was established that the rate constant for twisting is essentially zero when the excess excitation energy is less than the threshold energy Eo and can be fit by RRKM theory at higher energies. Introduction of an alkane gas as a low viscosity medium provides a thermal bath for excited molecules. Collisions between the initially excited stilbene, It*, and bath molecules can supply the energy needed to overcome the intrinsic barrier to twisting and thus promote the isomerization process. This low viscosity region is known as the energy-controlled regime of Kramers' equation (ref. 84). Since the medium effect is opposite in the two regimes, a cross-over, where the rate constant for twisting reaches a maximum value, i s expected as the system passes from the diffusion-controlled regime to the
107
energy-controlled regime. The viscosity region between the two Kramers' equation regimes is known as the Kramers turnover region. Several investigations whose purpose was the definition of conditions where the Kramers turnover region occurs have been described (refs. 158, 171175, 197-200). Extension of the study of stilbene fluorescence lifetimes to the smaller liquid alkanes, propane and ethane, at room temperature failed to reveal the turnover region. The twisting rate constants continued to increase monotonically with decreasing medium viscosity (ref. 158). On the other hand, the rate constant for twisting increased with increasing methane (refs. 174, 175) or ethane pressure (refs. 172, 197) in the gas phase suggesting that the barrier crossing i s energy-controlled in thermally equilibrated vapor and that the turnover region lies at higher pressures of vapor phase ethane (refs. 158, 172). A direct view o f the turnover region was provided by the elegant study by Lee, Holtom and Hochstrasser of trans-stilbene fluorescence lifetimes measured as a function of ethane pressure (0 to 170 atm) at 350 K (ref. 173). By carrying out the reaction above ethane's critical temperature passing through a gas-liquid transition was avoided, the solvent remaining a uniform A plot of ktp vs. shear viscosity reveals the Kramers isotropic fluid. turnover at vs 0.03 cP, Figure 13, corresponding to ktp = 3.45 x lolo s-l at 120 atm ethane pressure (ref. 173). This maximum ktp value is very close to ktp = 3.37 x 1O1O s-l obtained using transition s ate theory with parameters $ defined from the liquid alkane results: KeAs IR= 0.47 and AH+, = 2.85
-
-
Fig. 13 Fluorescence decay rate constants of trans-stilbene vs. the measured pressure of ethane at 350 K. Pc indicates the critical pressure of ethane. Reprinted from ref. 173 with permission of the copyright holder, Elsevier Science Publishers, B.V.
108
kcal/mol ( r e f . 144), and i s a l s o close t o ktp = 4.00 x 1O1O s-l pred’cted AS+/R = using parameters obtained by f i t t i n g t h e kobsd values t o eq. 31: rce 0.36, Eo = 2.89 kcal/mol (ref. 178). However, t h e isomerization r a t e constants o f Lee e t a l . ( r e f . 173) are s y s t e m a t i c a l l y somewhat lower than those i n f e r r e d from fluorescence l i f e t i m e s by Fleming and coworkers (refs.
158,
175),
and s i g n i f i c a n t l y smaller than values based on t r a n s i e n t absorption measurements by Troe and coworkers ( r e f . 172). While t h e r e may be s u b t l e changes i n d oi n passing from t h e i s o l a t e d e x c i t e d molecule t o t h e alkane medium i n vapor c l u s t e r s o r i n solution, t h e d r a s t i c changes proposed by Troe’s group can be discounted, for reasons given i n d e t a i l above, ( r e f s .
144, 172, 175). The p o s s i b i l i t y t h a t t h e degree o f a d i a b a t i c i t y o f t h e b a r r i e r crossing changes somewhat on passing from t h e vapor t o t h e l i q u i d phase has been proposed ( r e f s . 106, 107).
The short l i f e t i m e , 7 = 0.32 ps, assigned t o the lowest e x c i t e d s i n g l e t s t a t e o f c i s - s t i l b e n e i n the gas phase a t room temperature ( r e f . 129) i s
-
thought to be associated mainly with t h e b a r r i e r l e s s lc* lp* torsional displacement. Study o f v i s c o s i t y effects on such a process i s o f considerable t h e o r e t i c a l i n t e r e s t because Kramers’ equation was derived f o r cases f o r which an i n t e r n a l energy b a r r i e r e x i s t s .
The three- t o f i v e - f o l d increase i n
t h e decay l i f e t i m e o f t h e t r a n s i e n t 1c* absorption on passing from t h e vapor phase i n t o n-hexane s o l u t i o n a t room temperature, 7 = 1.0 1.35 ps ( r e f s . 77,
-
78, 161), no doubt stems from a small medium imposed b a r r i e r . Such a b a r r i e r
i s e s p e c i a l l y evident i n glassy media a t low temperature where t h e decrease i n t h e c -, t quantum y i e l d ( r e f s . 22, 23) i s accompanied by an increase i n l c * fluorescence ( r e f s . 24, 55-58). A study o f t h e temperature dependence o f qjf i n a h y d r o c a r b o n medium i s a v a i l a b l e f o r 1:l methylcyclopentane/methylcyclohexane (MCP-MCH) solution (ref. 24). Quantitative
treatment o f these 1c* qjf data i s hampered because values obtained from Figures 4 and 8 i n r e f . 24 are n o t s e l f - c o n s i s t e n t . The f o l l o w i n g conclusions
are based on the data i n Figure 8 using qjf = 0.75, the l a r g e s t experimentally observed fluorescence quantum y i e l d , as the l i m i t i n g value a t very low T. The
t r a n s i t i o n s t a t e p l o t i s l i n e a r r e s u l t i n g i n a solvent-induced b a r r i e r AHv =
3.9 kcal/mol.
Use o f a more r e a l i s t i c , l a r g e r qjf l i m i t i n g value leads t o an
upward curvature i n t h e t r a n s i t i o n s t a t e p l o t a t t h e lower T range suggesting a decrease i n AHv w i t h increasing v i s c o s i t y . I n contrast, t h e Andrade p l o t o f vS f o r t h e medium shows pronounced upward curvature w i t h decreasing T w i t h .€v increasing sharply as T i s decreased, EB a T2 ( r e f . 24). The inescapable conclusion, t h a t would n o t be a f f e c t e d i f data from Figure 4 i n r e f . 24 were employed instead, i s t h a t a = AHv/EB diminishes w i t h increasing medium v i s c o s i t y . This v i s c o s i t y s a t u r a t i o n e f f e c t can be observed d i r e c t l y by
109
plotting log (kobsd/T) vs. log q S . The plot is linear in the lo3 to lo6 CP range, (AHv/Ea) = 0.59, but shows upward curvature at higher viscosities, Figure 14 (ref. 178). The saturation effect is not expected to persist if alr were employed instead of qs in Figure 14, however, since the Spernol and Wirtz relationship, eqs. 26 and 27, was developed with data at much lower viscosities, qF values are presently not available. The behavior in Figure 14 is consistent with Lee, Zhu and Robinson's effective viscosity proposal, eq. 32. Pronounced viscosity effects on the fluorescence quantum yields of cis-
8.8 1
7.2
6.4 5.6
2 .o
'
0 0
I
4.0
I
I
8.0
6.0
IC 3
Fig. 14 The effect of shear viscosity on cis-stilbene photoisomerization rate constants, kisom = kf[($fO-$f)/$f$fO], in MCP-MCH with kf = 4 x lo8 s-l and $fo = 0.75, 0 ; $fo = 0.90, 0 . Data from Figure 8 in ref. 24 as treated in ref. 178.
-
stilbene have also been observed in decalin, 100 - 170 K, and in a 1:l mixture of methylcyc1ohexane/3-methylpentane, 80 - 120 K, but the data are too scantily presented to allow quantitative treatment (ref. 201).
-
In conflict with the above analysis of lc* fluorescence is the interpretation of recent observations of fluorescence from cis-stilbene in inert gas clusters (ref. 202). Supersonic beams of cis-stilbene, at 80 "C, seeded in He, Ne, Ar, Kr, N2 or C2H4 exhibit fluorescence and fluorescence excitation spectra whose intensity shows a strong dependence on the identity and pressure of the seeding gas. A featureless emission spectrum was obtained which closely resembles the previously observed spectra in viscous solution that were considered in the preceding paragraph. The unexpected increase in
110
fluorescence quantum yield with increasing stagnation pressures (up to 1600 Torr) and increasing noble gas atomic mass was proposed to originate from the trapping of vibrationally relaxed 1c* in an inherent minimum on the S1 surface, from where it decays, with the observed 20 ns lifetime (ref. 202). The problem with this interpretation is that it predicts too low a radiative rate constant for k*: kf = 5 x 107 s-1, in the unlikely event that unity fluorescence quantum yield were attained. A much more reasonable kf = 2 x lo8 s-l value can be based on the 4.7 ns lifetime of 1c* (ref. 62) observed at 77 K in 3-methylpentane glass under conditions for which fluorescence quantum yields close to unity have been reported (refs. 24, 201). The comparable, large Stokes shifts and similar shapes of fluorescence spectra obtained in inert gas clusters and in rigid hydrocarbon glass media (cf., also, ref. 203) suggest that the emissions have a common electronic state origin and reflect similar geometry changes between ground and excited states of cis-stilbene. It seems likely, therefore, that the 20 ns lifetime in the gas phase represents some form of delayed fluorescence and is thus not indicative of an abnormally low kf value. A possible mechanism involves reversible formation of electronically excited DHP which through inert gas cluster or collisional stabilization contributes a delayed k* fluorescence component. Adiabatic lDHP* formation in solution has been considered in accounting for the azulene quenching of c t photoisomerization of bromostilbenes (ref. 26). However, involvement of a lowest lAg* state could also account for the small kf. +
A recent extensive spectroscopic study employing a series of 1,2diphenylcycloalkenes, cyclo-cg to CyClO-c6, as models of cis-stilbene has shown that the first ionization potential in the ultraviolet photoelectron spectra, and the blue shift of the first absorption band and the Stokes shift in the optical spectra increase with increasing ring size. Theoretical calculations and X-ray crystal structural data reveal an increasing loss of ground state planarity, due primarily to twisting about the phenyl-vinyl bonds, with increasing ring size, that provides a logical explanation for the spectral changes (ref. 203). A twist angle of 43.2 t 10 * for ground state cis-stilbene in the gas phase had been deduced from an electron diffraction study (ref. 204). 3.3.2
Polarity effects.
An early indication that alcoholic solvents lower the intrinsic barrier for It* -,'p*(*) was a 26% decrease in t$f for trans-stilbene on changing the solvent from n-pentane to tert-butyl alcohol at 30 'C despite a 15-fold increase in medium viscosity (ref. 30). Sundstrom and Gillbro used ps
111
absorption recovery measurements at 305 nm to measure It* fluorescence lifetimes n the n-alcohol solvent series. A monotonic increase of kobsd-l from 42 ps in ethanol to 115 ps in 1-octanol at 296 t 1 K was observed with increasing n-alkyl group size that was nearly linearly dependent on qs (ref. 167). On the other hand, no systematic change in the nonradiative lifetime could be discerned when the qs was kept constant at 2.55 CP by varying the temperature in the C3 to Cg n-alcohol series. Based on an apparent zero slope for the Arrhenius isoviscosity plot it was concluded that the internal barrier for twisting of It* in these solvents is probably much lower than 1 kcal/mol (ref. 167). Consequently, since the Kramers theory was derived for systems with significant internal barriers, its applicability to the It* twisting motion of stilbene in alcohol solvents was questioned (ref. 167). Extension of this work to It* lifetime measurements at more temperatures led to observations which appeared to strengthen the notion that the twisting motion is barrierless in alcohols (ref. 169). The theoretical models of Rapp (ref. 205) and of Baghi, Fleming and Oxtoby (ref. 150) that were designed to account for medium effects on barrierless processes were therefore used to The reduction of A d o on going from interpret the data (refs. 169, 206). alkane to alcohol solvents was attributed to a polarity effect on the energetics for twisting (refs. 167, 169). A similar conclusion was reached in a later study in which It* fluorescence lifetimes and fluorescence anisotropy decay times were measured in the C1 to C1o series of n-alcohols (ref. 161). No attempt was made to apply the Kramers-Hubbard relationship on 71" and 70)" values obtained due to unpublished observations which suggested polarity induced changes in the barrier height for the twisting process with changes in solvent and temperature (ref. 161).
-
A more extensive application o f the isoviscosity Arrhenius plot approach
in evaluating the effects of changes in alcohol solvent, and of temperature in a given alcohol was described recently by Hicks, Vandersall, Sitzmann and Eisenthal (ref. 170). Employing five isoviscosities in the 1.5 to 10.7 CP range, it was found that slopes of Arrhenius plots decrease with increasing isoviscosity leading to apparent d&/ values in the 1.2 to 2.6 kcal/mol range (ref. 170). These values, though significantly smaller than those obtained in n-a1 kanes are nonetheless in conflict with the earlier conclusion that It* twisting is barrierless in'n-alcohols (refs. 167, 169). A parallel study on the photoi someri zati on of t r a n s - 4 , 4 ' -dimethoxysti 1 bene a1 so indicates a decrease in barrier height with increasing isoviscosity in n-alcohols (ref. 162). These results on stilbene and its dimethoxy derivative are somewhat paradoxical since higher viscosities are attained by lengthening the alkyl group of the alcohol, which is equivalent to decreasing solvent polarity
112
towards that of the corresponding n-alkane for which a larger barrier applies, Figures 8 and 12 (refs. 144, 178). Closer examination of Figure 3 in ref. 170 reveals that the higher slopes obtained in the Arrhenius plots for 1.5 and 3.0 CP are almost entirely determined by the 7r values for ethanol. Removal of the ethanol points leads to an €t value for trans-stilbene which is relatively independent of shear viscosity in the C3 to C1o alcohol series. The same criticism applies to the observations of Zeglinski and Waldeck on the dimethoxy derivative (ref. 162), suggesting that the influence of the shorter alcohols methanol, ethanol and, possibly, n-propanol on the twisting process should be treated separately from that of the higher homologs. Based on the data available in the literature (refs. 67, 161, 167, 169, 170), and private communications of unpublished results (refs. 207, 208), transition state plots yield Adobsd values ranging from 3.15 kcal/mol in 1butanol to 4.53 kcal/mol in 1-decanol (ref. 209). The value for ethanol, Inclusion of the AH+,bsd = 3.95 kcal/mol, is clearly anomalously high. ethanol points from refs. 167 and 169 diminishes the linearity of the transition state plot and further increases d o b s d . Also anomalously high are the values for methanol and 1-propanol (ref. 208). These high values may reflect the stronger intermolecular hydrogen bonding between smaller alcohols which diminishes solvent-solute interactions. The anomalously high AHjobsd = 4.13 kcal/mol for n-hexanol is associated with a correspondingly high AS+,bsd value and is probably due to experimental error. The plot of AH+obsd Vs. Ev, Figure 15, shows pronounced upward curvature, in contrast to
t
-
4.5-
I
{ -
4.0-
0
0
Y
Y
0
u)
3.5-
m
So
4 3.03
The dependence of experimental activation enthalpies on Ev for the n-alcohol solvent series (see text). Fig. 15
113
the corresponding plot for n-a1 kane solutions, Figure 8. Clearly, application of the medium enhanced barrier model is not straightforward in the alcohol case because the increase in AH+,bsd with E~ contains, in addition to a viscosity induced AHv increment, an unknown and variable increase in AHSO as the medium becomes more alkane-like. The inescapable conclusion from Figure 15 is that the hydroxyl group in the solvent significantly lowers A d o , the intrinsic barrier for twisting. If we assume that the relationship AHv = 0.39E4, established for n-alkanes and glycerol (ref. 144), holds in the alcohols, Ado values are obtained that increase monotonically from 1.37 kcal/mol in 1-butanol to 1.98 kcal/mol in 1-decanol. A considerably larger AH), of 2.41 kcal/mol is obtained in ethanol. Though the estimated uncertainties in these values is larger than 1 kcal/mol, neglecting the ethanol results, a rather pleasing picture emerges of AH), approaching its value in n-alkanes as the size of the alkyl group is increased (ref. 209). In none of the alcohols do the results support the conclusion (ref. 169) that the twisting process experiences no internal barrier. In contrast to the above interpretation, the internal barriers inferred from isoviscosity plots (refs. 170, 208), lead to the unreasonable conclusion that, instead of approaching the hydrocarbon value, the intrinsic barrier decreases as the a1 kyl chain length is increased. The significance of these plots is questionable on at least two grounds. Firstly, as was demonstrated for the n-alkanes, the proper plot is an isomicroviscosity plot, and not one based on shear viscosity. Secondly, since AHSO varies from alcohol to alcohol, even isomicroviscosity transition state plots are not expected to be linear. The complex behavior of alcohol solvents is apparently due to a large measure to their highly associated structure via hydrogen bonding rather than simply to their increased polarity, since a recent parallel study of kobsd values in the homologous n-alkanenitriles reveals no dependence of AH+, on alkyl chain length (ref. 210). Isoviscosity plots give AH+, = 2.0 kcal/mol for the C3-Cio alkanenitriles (ref. 210). The modest reduction of the intrinsic barrier in passing from alkane solvents to the more polar alcohols and nitriles is attributed to a greater stabilization of the more polar transition state o f the It* -.* 1p* process than of the initially formed transoid IBu* state. Since considerable charge separation is expected in the lBU* state this interpretation is more consistent with formation, at least initially, of a highly charge separated lp* intermediate. systematic study of solvent polarity effects on cis-stilbene fluorescence has not yet been reported. However, a sharp increase in Of has been observed for solutions of cis-stilbene in ethanol-water (6.67%) and in A
114
1-propanol as the temperature is lowered from 133 to 93 K (ref. 24). A plot of qlf vs. log q s reveals a somewhat steeper increase in the former solvent. O f special interest is the observation that at the very large vS values attained in 1-propanol, i.e., qs > 1018 CP at 93 K, qlf has leveled off and 0.4 (ref. 24). This is in sharp appears to approach a limiting value of contrast to observations described above for hydrocarbon solvents and is indicative of a significant, relatively T and q independent, radiationless decay channel which functions in competition with twisting and radiative decay in hydroxylic media. Whether this process is associated with enhanced formation of DHP remains to be established.
-
4
DIPHENYLPOLYENES
The goal o f developing model molecules whose properties better approximate those of the retinyl polyenes, which are related to vitamin A and the visual pigments, has given rise to numerous studies of the a,o-diphenylpolyenes, 12, of which stilbene can be regarded as the initial (n = 1) member. These studies have been rich in their rewards and have been the subject of several
12 reviews (refs. 39, 211, 212). The scope o f this section will be limited to a comparison o f the properties of the three higher members of the series with those of stilbene. In contrast to the stilbenes, photochemical results lag far behind the spectroscopic observations and ambiguities still exist in assigning possible mechanisms to the radiationless processes leading to the decay of the lowest excited singlet states of the diphenylpolyenes. There is sufficient photochemical evidence, however, to support the position that the usual assignment of radiationless decay processes to twisting about one of the former double bonds, by analogy with the st-ilbenes, is not at all secure. The large difference in behavior between the singlet excited states of the stilbenes and those of the polyenes is due, in large measure, to a change in the relative energies of the lowest excited Bu and Ag states. In the foll owing, we wi 1 1 begin with a1 1 -trans-l,6-diphenyl-l,3,5-hexatriene (DPH) , 12, n = 3, and all-trans-l,8-diphenyl-1,3,5,7-octatetraene (DPO), 12, n = 4, because in these and higher polyenes, there exists a large body of evidence indicating that the doubly excited lAg state lies below the dipole allowed lBU
115
state, the reverse order from that found in stilbene. The section will be completed with a discussion of trans, trans-1,4-diphenyl-l,3-butadiene (DPB), 12, n = 2, for which these two states are thought to be nearly degenerate.
4.1
Diohenvl hexatriene and diDhenvloctatetraene
An excellent brief review of theoretical calculations in the 1960’s that described the properties and possible role of a low-lying A state in polyene photochemistry and spectroscopy is available (ref. 213). The existence of an excited state was first unambiguously demonstrated by Hudson and Kohler for DPO (refs. 32-39). Assignment of the fluorescence spectrum of DPO to the 2IAg llAg transition, where the doubly excited 21Ag state is identified as the lowest polyene excited singlet state, was confirmed by numerous experimental (refs. 214-218) and theoretical (refs. 43, 219-222) investigations. The spectroscopic evidence also supports the 21Ag assignment for the lowest excited singlet state of DPH. Though these assignments seem well founded (ref. 211), the. consequences of the 11Bu/21Ag state reversal, relative to trans-stilbene, to the photochemical and photophysical properties of the polyenes are still not well understood. +
In this section we limit ourselves to those observations which are crucial for developing a mechanism for the trans -, cis photoisomerization of the lowest excited singlet states of DPH and DPO. As in stilbene, intersystem crossing has been shown to be rather inefficient in these polyenes, $is S 0.02, and furthermore triplets formed by sensitized excitation have been shown not to undergo trans -, cis photoisomerization (ref. 223, 224). The fluorescence emissions of DPH and DPO display several unusual features: (a) Experimental radiative rate constants, kfobsd, calculated from fluorescence quantum yields and lifetimes, eq. 1, are significantly smaller than theoretical radiative rate constants, kfth, based on the Strickler-Berg or Birks-Dyson equations (refs. 225, 226) applied to the first dipole allowed electronic absorption bands. Ratios of (kfth/kfobsd) of 5.4 - 9.9 and 1721, depending on the solvent, have been reported for DPH and DPO, respectively, and more than 10-fold higher ratios have been found for higher polyenes (ref. 211). (b) Fluorescence quantum yields and lifetimes are strongly solvent dependent (refs. 211, 227, 228). Lifetimes are essentially temperature independent in nonpolar solvents, but decrease with increasing temperature in polar solvents. Fluorescence quantum yields, on the other hand, decrease with increasing temperature in all solvents. (c) The absorption and fluorescence excitation spectra shift to the red as solvent
116
polarizability is increased, while the position of the fluorescence spectra is insensitive to solvent changes (refs. 38, 211, 228, 229). (d) The intensity of thermally activated fluorescence, observed in DPH, its derivatives, and in DPO as a weak band on the blue edge of the main spectrum, changes with solvent and increases with increasing temperature (refs. 230-233). Many of these features are readily explained if it is assumed (ref. 37-39) that the lowest DPH or DPO singlet emitting states and those of higher polyenes are 2lAS states which are observed in absorption as weak features lying to the low energy side of the origin of the llBu llAg transition in the single photon absorption spectrum, and as relatively strong allowed Z1% llAg transitions in two-photon fluorescence excitation spectra (ref. 211). Other models proposed to account for the fluorescence properties of polyenes have generally been discredited. They include the early proposal by Birks and Dyson (ref. 226), later elaborated upon by Berlman (ref. 234), that fluorescence originates from a relaxed lBu* state with significantly different molecular geometry than the initially formed Franck-Condon allowed excited state. Also unsatisfactory is a model proposed for DPH by Cehelnik and coworkers (ref. 228). This mechanism incorporates geometry change prior to emission, as proposed by Birks and Dyson (ref. 226) and by Berlman (ref. 234), but assumes that the initial thermalized twisted state lDPHp* relaxes further to a new excited state lDPHp*' whose emission is presumably associated with a fast radiative rate constant and is therefore dipole allowed (ref. 228). Though the mechanism is clearly at odds with unambiguous evidence requiring that the emitting state be the 21% state, it incorporates several attractive kinetics features and deserves more detailed consideration. The Cehelnik model is illustrated in Scheme 3. +
+
Scheme 3 Solvent Perturbation Mechanism for Relaxation of Polyene Singlets.
+
1DPH$
h'
t
~DPH
'DPH
+
hv
products
Rapid relaxation of the initially formed Franck-Condon excited state, ~DPHFc*, involving solvent reorientation and twisting of the molecule, gives thermal ized state lDPHp* which undergoes "specific solvent-induced
117
perturbations.”t o g i v e e m i t t i n g s t a t e lDPHp*’.
4f
= (-
k’
k’+knr
I (
k f’ kf‘+knr’
The mechanism p r e d i c t s
1
(33)
where t h e f i r s t term i s t h e e f f i c i e n c y o f formation of t h e e m i t t i n g s t a t e . BY assuming f u r t h e r t h a t i n nonpolar solvents, k‘ << k n r ’ o r kf‘, t h e fluorescence l i f e t i m e i s given by (l/k’). I n p o l a r solvents knr becomes important leading t o diminished, temperature dependent, fluorescence 1ifetimes
“ff
l/(k‘+knr)] and quantum y i e l d s (ref. 228). That k f ‘ i s assumed t o represent a r a d i a t i v e t r a n s i t i o n from a IB, s t a t e i s i m p l i c i t i n t h e assumption t h a t i t s value, 6.4 x 108 s-1, can be based on DPH’s llBu llAg =
+
absorption spectrum ( r e f s .
225,
226,
It was suggested,
234).
as
an
afterthought, t h a t i f the 21Ag s t a t e were involved i n t h e Scheme i t would have t o be assigned as t h e e s s e n t i a l l y nonfluorescent 1DPH; state. A fatal drawback o f t h e Cehelnick model i s t h a t i t f a i l s t o account f o r t h e very d i f f e r e n t solvent p o l a r i z a b i il t y dependence o f polyene absorption and fluorescence spectra (refs. 211, 229-231). I t s importance l i e s i n t h e r e c o g n i t i o n t h a t t h e mechanism should be able t o decouple fluorescence It i s t h i s f e a t u r e of t h e l i f e t i m e s from fluorescence quantum y i e l d s . mechanismwhich we propose t o r e t a i n i n modified form (see below). Widely accepted as t h e mechanism accounting f o r t h e i n t r i g u i n g fluorescence behavior of t h e polyenes i s t h e borrowed i n t e n s i t y model t h a t appears t o be a l o g i c a l consequence of the assignment o f t h e b u l k o f t h e
Sl
A
V
Fig. 16 235).
V
I
!
The Hudson and Kohler Model f o r polyene. fluorescence (refs. 38, 211,
118
fluorescence spectrum to the lowest lying 21Ag excited state. Initially proposed by Hudson and Kohler (ref. 38), this model has been elaborated upon by Birks and coworkers (refs. 45, 52, 229) and by Kohler, Hudson and coworkers, Figure 16 (refs. 211, 235). lAg transition is symmetry forbidden in the single The lAg* (or 2l$) photon absorption spectrum and allowed in the two photon absorption spectrum; the converse applies for the IB,* 1% transition. This accounts for the relatively long experimental radiative lifetimes of DPH, DPO and higher polyenes, since the calculated theoretical radiative rate constant corresponds to kfb while the experimentally observed radiative rate constant corresponds to kfa, Figure 16. That kfa is not even smaller than it is, is attributed to borrowed probability from the allowed lBu* -c lAg transition through mixing of the lAg* and lBU* states. Though there is disagreement as to whether this mixing is induced by lBu* - lAg* potential interaction, via a bu vibrational mode (ref. 229) or by vibronic coupling of the two states (ref. 235), both approaches lead to the conclusion that +
.-
where Y is the coupling matrix element and A€ba is the lB,* - '$* energy gap. The relative dependence of fb and of €a on solvent is determined experimentally from absorption and fluorescence data, respectively, e.g., Figure 17. The dependence of the transition energies on the solvent is given by the approximate theoretical relationship
where i = a or b, P i is a collection of constants whose magnitude reflects the size of the transition dipole for excitation from ground state to state i , € j and € i o are the transition energies for excitation in a specific medium and in the gas phase, respectively, and (I
=
(36)
(n2-1)/(2+2)
is related to medium polarizability where n is the index of refraction (refs. 211, 235). Plots of fib vs. Q for 12 with n = 1 - 8 show good linearity, extrapolating to the gas phase value for those cases, n = 1 - 4, where fibo is known (refs. 211, 233, 235, 236). The slopes of the lines give Pb = 10,000 cm-l. In contrast €a is nearly independent of Q, presumably reflecting the small transition dipole for the lAg* lAg transition. It follows that AEba decreases as solvent polarizability, and hence a , is increased +
119
I
I
Fig. 17 Absorption spectral shifts of the diphenylpolyenes 12 with n = 14. Reprinted, in part, from ref. 236 with permission of the copyright holder, the American Chemical Society. providing a rationalization for the dependence of polyene fluorescence properties on solvent. In particular a decrease in AEba leads to an increase in kfa, eq. 34, and improves the likelihood o f thermal repopulation of the lBU* state and delayed lBU* fluorescence (refs. 211, 233). Though the intensity-borrowing model succeeds qua1 itatively in accounting for many of the photophysical observations and in some instances appears to fit the data quantitatively (refs. 229, 233, 235), it falls short from providing internal consistency in many important details. For instance, in DPH a plot of AEba vs. a gives AEbao = 10.0 kcal/mol (3500 2 66 Cm-l) and APba = 25.3 f 2.7 kcal/mol (8830 f 950 cm-1) (ref. 235) which predicts AEba = 4.2 kcal/mol in n-hexane at room temperature, in reasonable agreement with the observed spectral energy gap (ref. 233). On the other hand, analysis of the temperature dependence of thermally activated fluorescence, presumably from 'BU*, from DPH in n-hexane gives AEba = 1.8 kcal/mol (ref. 233) which is significantly smaller than the spectroscopic value. Another challenge to the intensity-borrowing model is that it supposedly enhances &fa without there being a corresponding enhancement in the reverse lAg* IAg transition. Again
-
120
using DPH as the example, changing the solvent from n-perfluorohexane to benzene at room temperature lowers the energy of lBu* sufficiently to produce a drop in AEba from 2100 cm-l (6.0 kcal/mol) to 900 cm-l (2.6 kcal/mol) (ref. 235), thereby incre'asing the experimental kfa value from 1.6 x lo7 s-l to 1.2 x lo8 s-l (ref. 227). Such large apparent kfa values require that state mixing be sufficiently strong as to confer upon the symmetry forbidden process up to 20% of the probability o f the allowed lBu* lAg transition. It seems surprising, therefore, that the same mixing is not reflected in a transition in the single photon absorption correspondingly strong IAg* spectra whose oscillator strength is similarly solvent dependent. It should be noted further that recognition that fluorescence spectra are composed in part of thermally activated lBu* emission (refs. 230-233) invalidates application of eq. 1 to the calculation of experimental kfa values.
.-
Perhaps the major weakness of the presently accepted intensity-borrowing model concerns the way it has been used to account for radiationless decay leading to trans cis photoisomerization. Extending the Orlandi and Siebrand model for stilbene photoisomerization to the polyenes, Figure 26 with IAg* below it is generally assumed that twisting occurs, in competition with fluorescence, in the lowest excited singlet state, namely 21Ag. Since in the original proposal the pure 21% state is postulated to experience no barrier to twisting, one is left with the dilemma of accounting for relatively long fluorescence lifetimes, which, at least in nonpolar solvents, are surprisingly insensitive to temperature changes (ref. 227). Recognizing that, within the Orlandi and Siebrand model, reversal o f state energy order eliminates the possibility of assigning the potential energy barrier to the crossing o f lBU* and lAg* states along the twisting coordinates, Birks and coworkers postulated that the closer proximity of these two states at the planar geometry causes the mixing between 1B,* and lAg* potential energy surfaces to diminish with motion along the twisting coordinate (refs. 45, 52, 229). Consequently, the greater stabilization of the lAg* state at the planar geometry is assumed to be responsible for the barrier to twisting. While at first glance this suggestion appears attractive in view of the strong mixing postulated to account for the large kfa values, it is inconsistent with spectroscopic observations indicating that Ea and Eb are independent of AEba (refs. 211, 229, 235). If these energies reflected potential energy repulsion due to mixing, this repulsion should become more pronounced as AEba -,0 and plots of Eb vs. a for 12, n = 1 - 8, should not have generated the observed series of nearly parallel straight lines, nor should E , have remained independent of AEba ( r e f s . 211, 229, 233, 236). It should be noted that AEba appears to change regularly with increasing number of double bonds in the +
121
polyene, starting at a negative value for n = 1, passing through nearly zero for n = 2 and becoming progressively more positive for higher n (ref. 211). The puzzle of how the strong mixing between IAg* and lBU* allows for dramatic increases in &fa and produces a barrier to twisting, without affecting the energies of the mixed states (ref. 230) is not satisfactorily answered by the accepted model.
-.
The trans cis photoisomerization of DPH was first reported by Lunde and Zechmeister who isolated, but did not identify, four isomers containing cis double bonds (ref. 237). Isomerization is accompanied by other irreversible photoreactions, including probably the photocyclization of cis isomers (ref. 238), which are especially efficient in hydroxylic solvents (ref. 239). Gorner in a more recent study measured UV absorption spectral changes for trans,trans-1,4-diphenyl-1,3-butadiene, DPB, DPH and DPO under direct and triplet sensitized excitation conditions and was able to conclude that trans cis photoisomerization occurred only in singlet excited states (ref. 224). Although relative isomerization efficiencies were determined under different conditions of temperature and solvent, no isomerization quantum yields were reported. Thus, the assignment of radiationless decay in DPH and DPO solely to twisting leading to trans -,cis photoisomerization is risky since it is not experimentally founded as in trans-stilbene, Figure 1. Nonetheless, this assumption was applied by Birks and coworkers (ref. 229) in interpreting the temperature and solvent dependencies of qjf and 7f reported by Cehelnic and coworkers (ref. 227). Using the potential interaction barrier model, twisting was assumed to occur entirely in the lowest excited lAg* state. The fact that in nonpolar solvents 7f is nearly temperature independent, or slightly decreases, while qjf generally increases with decreasing temperature, was ingeniously attributed to a compensation between radiative, kfa, and nonradiative, ktp, decay constants (ref. 229). The temperature dependence of kfa Was shown to adhere to eq. 34 by assuming AEba proportional to a and kfb = n2kfb0 where kfbo = 2.23 x 108 s-l for DPH is the radiative Constant of the 'BU* state in the absence of solvent (ref. 229). Activation parameters for the twisting process were then obtained from the temperature dependence of ktp = (7f)-I - kfa = (l-qjf)(~f)-~. While an apparently self-consistent picture emerges, it seems highly unlikely that a nearly exact compensation of changes in ktp and &fa could occur in each nonpolar solvent employed; note that temperature independent 7f's vary from 32 ns in n-perfluorohexane to 6.1 ns in benzene (ref. 227). Aside from the requirement of a fortuitous cancelation of temperature effect on two such disparate decay processes, the mechanism fails to account for several key observations. The first concerns the absence of an oxygen quenching effect on the rate of DPH disappearance despite a pronounced +
122
oxygen quenching of DPH and DPO fluorescence (ref. 240). In both benzene and acetonitrile the presence of air appears to enhance photoisomerization 25% (ref. 224) instead of quenching it as would be expected if efficiency by both fluorescence and isomerization were properties of the same transoid lAg* state. Enhancement of photoisomerization via 02 promoted intersystem crossing (refs. 240-242) appears to be unlikely in this case in view of the reported failure of DPH triplets to undergo trans cis photoisomerization. The Birks version of the Orlandi-Siebrand mechanism also fails to provide a satisfactory explanation for the absence of a viscosity effect on the lifetime of DPH fluorescence; if anything, the lifetime of DPH singlets decreases on going from n-hexane ( 7 = 11.2 ns) to n-hexadecane solution (7 = 10.8 ns) despite a substantial increase in $f (ref. 243). Along the same lines, the mechanism leads to the absurd conclusion that substitution of bulky substituents at the para positions of the phenyl groups of DPH, 13, n = 3, dramatically diminishes the barrier to twisting in methylcyclohexane from 4.5 kcal/mol in 12, n = 3, to 0.2 kcal/mol in 13, n = 3 (ref. 243).
-
-.
Finally, even if the mechanism were correct, its application is flawed since it bases &fa values on eq. 1 without taking into account considerable temperature dependent contributions of lBU* emission in the fluorescence spectra. Since little, if any, difference is discerned between shapes of fluorescence spectra for DPH in argon- or oxygen-saturated benzene or methylcyclohexane solutions (ref. 244), prompt fluorescence from the lBU* state appears to be negligible. A kinetics model for emission from two fully equilibrated states predicts (refs. 51, 52, 244)
123
where subscript e indicates the full equilibration condition, Kba is the equilibrium constant for lBU* lAg*, and ktb and kta are radiationless decay processes, including twisting, from the two states, respectively. Examination of eq. 40 shows that the procedure used routinely to obtain kfa actually yields an effective kf which is a combination of fractional contributions from kfa and kfb. Since kfb > kfa at all T and Kba decreases with increasing T , an increase in the effective kf with increasing T is expected which is contrary to experimental observations in nonpolar solvents (ref. 227). Thus, the need to postulate a decrease in &fa with increasing T is retained. Furthermore, examination of eq. 3 8 shows that the full equilibration condition accommodates twisting in either the lBU* or the lAg* state, but again requires fortuitous cancellation of changes in kfa by changes in kta and/or in ktb. It follows that the fuller treatment of the intensityborrowing mechanism avoids none of the shortcomings mentioned in the preceding paragraph. 4.1.1
Time-resolved sriectroscooic studies
A time-resolved fluorescence study of DPH and DPO in n-hexane solutions at room temperature has been reported that employed 10 ps 354.7 nm excitation pulses (ref. 245). Fluorescence signals were detected using the frequencyconversion gating technique. Zero time corresponded to zero displacement between excitation and probe pulses. Very similar fluorescence spectra were obtained for DPH at 3 ps and 416 ps (only the 370 - 450 nm onset of the spectrum was probed), with the shorter time spectrum showing a subtle increase in intensity in the 380 - 400 nm region. This weak excess fluorescence in the region assigned to lBU* -,lAg emission (refs. 230 -233) was observed to decay to 70% of its initial value within about 30 ps (ref. 245). This evidence suggests that the initially formed I8,* molecules relax to the equilibrium distribution of lBu* and lAg* states with a rate constant exceeding 3 x lolo s-l, consistent with the absence o f significant contributions of prompt fluorescence in continuous excitation fluorescence spectra. Under the same conditions, or in benzene solution, pronounced, obvious differences are observed between short- (-3 ps) and long-time (86 ps) DPO fluorescence spectra, but the features at the onset of the spectrum, which are probably due to prompt lBU* -t lAg emission, decay with a pulse-duration limited rate constant ( > 1 x 10l1 s-l) to the steady-state spectrum (ref. 245). Assignment of the short-time features to lBU* emission is supported by the fact that they undergo nearly the same red shift in going from n-hexane to benzene as is observed in the lBU* lAg absorption spectra (ref. 245).
-
+
124
-
Transient Sn S 1 absorption spectra in the ns time region have revealed strong unstructured absorptions at 450 nm and somewhat weaker absorptions at u- 650 nm for both DPH and DPO in cyclohexane solution which were initially assigned to nlBu 2’b transitions (refs. 246, 247). The absence of triplettriplet absorption in the spectra was presumably established by showing that transient absorption and fluorescence spectra exhibit identical time response profiles (ref. 246). It may be significant, however, that triplet-triplet absorptions have been observed in benzene solutions of DPH, Xmax = 425 nm, and DPO, Xmax = 444 nm, following direct excitation of the polyenes (ref. 224, 240). Although small intersystem crossing yields are involved (refs. 223, 224, 240), these absorptions are close to the Xmax of the ns transients and may contribute in accounting for the large discrepancy between the ns time scale transient absorptions and those obtained in the ps time scale (see below) (ref. 248).
-
+-
The ps transient absorption study of DPH and DPO in methylcyclohexane (MCH), 3-methylpentane (MP) and ethanol solutions employed 355 nm excitation
pulses and probed absorption in the 430 - 800 nm region in 10 ps steps during the lifetime of the transients (ref. 248). The spectra obtained in methylcyclohexane bear a rough similarity with those reported by Kliger and co-workers (ref. 246) but the relative absorbance at 480 nm to that at 650 nm is much smaller in the ps time scale spectra, especially for DPH for which the relative intensity of the two bands is reversed. The ratio, r , of intensities, 1480/1650, was shown to decrease with time, reaching a plateau within about 100 ps. Furthermore the value o f r for the plateau was shown to decrease with decreasing 7. Based on these observations, the 480 nm band was assigned to the nlAg llBu transition, the 650 nm band was assigned to nlg, 2lA9 transition and r, at t = 500 ps following excitation, was assumed to be a measure of the equlibrium [lBu*]/[lAg*] ratio (ref. 248). van’t Hoff plots of the variation of r with 7 yielded even smaller AH =. AEO values than were obtained from the 7 dependence of the dual fluorescence ratio (ref. 233). Specifically, a value of A@ 180 f 30 cm-l (or 0.51 kcal/mol) in 1:l MCH/MP from the transient absorption measurements (ref. 248) can be compared with 630 cm-l (1.8 kcal/mol) in n-hexane from the dual fluorescence measurements (ref. 233), and both values are significantly smaller than AEO = 1000 cm-l (extrapolated value at 0 K) based on the spectral origins of lAg* and lBU* emissions (ref. 233). Parallel measurements for DPO show similar discrepancies (refs. 233, 248).
-
-
-
+
-
It has been proposed that a larger AEba is associated with ground state conformations than with relaxed, solvent stabilized excited state
125
conformations and that this accounts for the difference between AEba based on the spectra and those based on the van't Hoff plots (ref. 248). However, the spectroscopic difference cited is taken from fluorescence spectra and should already reflect excited state relaxation events. It seems likely that the low AEba values from the van't Hoff plots are a consequence of improper separation of strongly overlapping absorption and emission spectra of two equilibrating states. For instance comparison of DPH spectra 1, 2 and 5 in Figure 1 of ref. 248 suggests strongly that band A at 650 nm includes contributions of absorption from both lAg* and lBu* states instead of just lAg* as assumed in the treatment. Similar difficulties had been encountered in the van't Hoff treatment o f benzophenone' s strongly over1 appi ng delayed fluorescence and phosphorescence spectra (ref. 249). The temperature independence of r for DPH in ethanol has been interpreted as reflecting a temperature independent [lBu*]/[lAg*] ratio due to AEba = 0 in that solvent (ref. 248). Unfortunately, transient Sn + S1 absorption spectra in ethanol were not reported (ref. 248), but, in view of our conclusion that both short and long wavelength bands in hydrocarbon solvents include contributions from lBU* absorption, an alternative explanation for the 7 independence of r is that only the lBu* state i s observed in ethanol. A solvent induced state order reversal could account for the relatively normal behavior of q5f and 7f values in polar solvents. 4.1.2
Excited sinqlets of isolated DPH
The fluorescence spectrum, one photon fluorescence excitation spectrum, and single vibrational level decay kinetics have been measured for DPH seeded in helium in a supersonic free jet (ref. 250). Consistent with observations from two photon excitation spectroscopy in EPA glass at 77 K (ref. 217), absorption to the 2lAS state is observed as a sharp progression of bands starting at 25,741.8 cm-1 above the ground state (ref. 250). Fluorescence excitation intensity from the lowest observed band, assigned as a vibronically promoted false origin, is smaller than that of the IIBu llAg fluorescence excitation origin at 29,157 cm-l by at least a factor of lo4 (ref. 250). The relatively small shifts to 25,050 cm-l in EPA at 77 K (ref. 217) and to 25,187 Cm-' in n-hexane at 4.2 K (ref. 214) for the weak 21Ag llAg transition can be contrasted with the 3200 cm-l shift to 25,973 cm-l in n-hexane at 4.2 K (ref. 214) experienced by the strong llBu llAg transition, and are consistent with the differential effects of medium polarizability change discussed above. +
+
+
126
Especially informative is the broad shape of the fluorescence spectrum obtained upon excitation of the 0-0 band of the llBu llAg transition (ref. 250). The spectrum bears a strong resemblance to solution spectra and shows l%* relaxation in the isolated molecule is fast in the ns time that lBU* scale. The fluorescence lifetime gradually decreases with excitation frequency from 90.7 ns at 25,742 cm-1 to 39.1 ns at 30,771 crn-l. Excitation in the 0-0 band of the lBU* + lAg transition gives 7 = 48.1 ns (ref. 250). Lifetimes were measured by treating decay curves 7 ns following maximum fluorescence intensity attainment, so no kinetics information was obtained for shorter times when rapid IVR changes must be occurring. However, the absence of prompt lBU* fluorescence shows that relaxation to the lAg* state is rapid 9.7 in the absence of solvent effects, despite the much larger AEba kcal/mol value. It is unknown at this time to what degree, if any, conformational changes leading to trans + cis photoisomerization contribute as radiationless decay channels. In the absence of fluorescence quantum yield information the lifetimes obtained provide upper bounds for effective radiative rate constants, and clearly establish the forbidden nature of the radiative transit ion.
-.
+
-
A study of fluorescence and fluorescence excitation spectra from static vapors of DPH (368 K) and DPO (377 K) bridges the gap between the solution work and the above study of the jet-cooled isolated molecule. Though DPH and DPO were reported to behave similarly, results for the former were presented in greater detail (ref. 251). Emission from both was assigned to the 21Ag state. For DPH, fluorescence intensity was shown to increase with increasing buffer gas pressure (perfluorohexane 0 - 250 Torr) and to decrease slightly with increasing excitation energy (310 - 340 nm). The pressure dependence of 4f was fitted by assuming kfa = 4.2 x lo7 s-1, independent of excitation wavelength and buffer gas pressure. This value seems rather high in view of the large Afba in the vapor phase. It corresponds to a radiative lifetime of 24 ns which is significantly smaller than actual lifetimes observed for jetThe identity of the radiationless channel(s) cooled DPH (ref. 250). responsible for the relatively low fluorescence quantum yields at low buffer gas pressure remains to be defined. 4.1.3
Concernins the Dhotoisomerization mechanism
In the absence of quantum yield information on the photoisomerization of DPH and DPO, proposals concerning the photoisomerization mechanism must be viewed with a healthy touch o f skepticism. The early work of Zechmeister does suggest one bond photoisomerization following direct excitation into singlet
127
excited states (ref. 237). Interpretations of photophysical behavior of these excited states are either explicitly or imp1 icitly based on the assumption that only excitation of the all-s-trans conformations need be considered. The independence of spectroscopic properties on Xexc tends to support this assumption. We may cite here the Xexc independent triplet-triplet absorption spectra (ref. 224) and the Xexc independent fluorescence spectra and fluorescence lifetimes (ref. 227). However, small but significant contributions from s-cis conformers could lead to subtle, easily overlooked, changes and we must not lose sight of the fact that Havinga’s NEER principle was initially proposed to account for the photochemical behavior of a triene (ref. 9). An example of Xexc dependent photoisomerization which has been attributed to selective excitation of different conformers concerns the 5phenyl-1,3,5-heptatriene, 14 (ref. 252). In I4 as in alloocimene (2,6-
dimethyl-2,4,6-octatriene), 15, direct excitation leads to trans
-, cis
photoisomerization about only one of the double bonds (ref. 253). In the latter case triplet excited states were shown to undergo two bond photoisomerization following a single excitation events (refs. 253, 254). The difference in behavior between triplet and singlet excited states is reminiscent of that found for the 2,4-hexadienes (refs. 3, 4, 255, 256) and provides early indirect evidence for the absence of intersystem crossing as a significant pathway following direct excitation. The fact that photoisomerization occurs from these nonfluorescent, presumably very shortlived, excited polyene singlet states shows that the process can be very fast.
128
The photochemistry of the parent 1,3,5-hexatrienes and of mainly alkyl substituted conjugated trienes has been reviewed (ref. 257). The major photochemical process of trans-1,3,5-hexatriene, 16, is isomerization of the central double bond to give the cis isomer, 17, 6tdc = 0.034 (refs. 257, 258). The reverse reaction occurs with a somewhat lower quantum yield, Qc+t = 0.016, in competition with cyclization to 1,3-cyclohexadiene, 18 (refs. 257-259). Other valence isomerization products, 19 and 20, are obtained upon prolonged irradiation of the photoequilibrium mixture o f 16, 17, and 18. In addition the vinylallene, 21, is obtained upon gas phase irradiation of 1,3,5hexatriene (refs. 257, 260, 261). Product distribution shifts with changes in Xexc and with judicious placement of alkyl substituents have been attributed to selective excitation of conformational isomers, which do not equilibrate in the excited state (refs. 257, 262-264). Some of these generalizations applied to the parent trienes are summarized in Scheme 4 (refs. 257-264). Within the limitations of present knowledge we tentatively postulate an alternative mechanism for the photoisomerization o f DPH and possibly DPO (ref. Scheme 4 The NEER Principle and 1,3,5-Hexatriene Photochemistry.
I7
II
18 16
=tJ 19
9 II. 20
u 21
129
244). Starting with the notion that only the all-s-trans conformer need be considered, we propose that the radiationless decay process leading to isomerization occurs primarily as a prompt event in the initially formed lBU* state in competition with rapid relaxation, probably along a different torsional coordinate, to a geometry more favorable to mixing of the lB,* and lAg* states. The equilibrating pair of mixed states, 'Am*, 'Bm*, could give rise to the observed dual fluorescence and contribute little to the photoisomerization pathway, Scheme 5. The barrier to rotation in lBU* could Scheme 5 A1 ternative Photoisomerization Mechanism for Polyene Singlets.
products
'Ag +hub
'Ag+hua
be inherent in that state as in Figure 2a or could arise from crossing with a higher lAg state, analogous to the 41Ag state for trans-stilbene, Figure 2d. This mechanism retains the advantage of the Cehelnik proposal, Scheme 3, of uncoupling fluorescence quantum yields from fluorescence lifetimes, by introducing a temperature and viscosity dependent efficiency factor, e f , in eq. 39 without affecting the lifetime expression in eq. 38. The mechanism will also account for biexponential decay behavior under conditions which inhibit torsional changes in the initially formed lBU* state and allow observation of prompt fluorescence from that state (refs. 227, 243). The prediction is that fluorescence .and isomerization quantum yields will reflect the temperature, medium and substituent dependence of the decay pathways of lBU*, while the insensitivity of fluorescence lifetimes to these variables reflects the absence of torsional decay pathways from the equilibrated mixed states, lB,* and lA,*. The absence of prompt fluorescence from lBU* is consistent with a very short lifetime for that state and accounts for the apparent inability of 02 to quench the photoisomerization (ref. 224). Interaction of the mixed fluorescing states with 02, on the other hand, leads to pronounced fluorescence quenching without affecting the isomerization quantum yields (refs. 224, 240). The major advantage of this mechanism is that it does not require cancellation of 7 effects on twisting and fluorescence rate constants; A comparative study of fluorescence and isomerization quantum yield quenching by the electronic energy acceptors
130
azulene and ferrocene, which may have provided an experimental test of the mechanism, was not reported in sufficient detail to be used for this purpose, though it did help to eliminate triplets as potential intermediates in the photoisomerization (ref. 224). Observations with DPO are more sketchy than those for DPH but generally suggest that the same mechanism applies for both systems (ref. 224). 4.2 Diohenvlbutadiene Theoretical calculations concerning state ordering in polyenes generally place the 21Ag state above the llBu state for stilbene in a transoid excited state geometry and predict the reverse order for DPH, DPO and higher polyenes (refs. 211, 212) trans-trans-l,4, -Diphenyl - 1,3- butadiene, DPB, i s general 1 y predicted to have 21Ag - llBu relative energies intermediate between transstilbene and DPH. The question, therefore, of whether the photochemical and photophysical behavior of DPB is stilbene-like, DPH-like,-or resemblgsneither because of a near degeneracy of the lowest two excited singlet states has attracted a great deal of attention. The answer, though somewhat elusive, depends strongly on whether DPB is studied in condensed media or in the vapor phase. As with DPH and DPO a paucity of photochemical investigations hampers detailed interpretations of the numerous excellent spectroscopic studies. In this section, solution and vapor phase spectroscopic results on DPB will be presented in turn, and some of the assumptions on which interpretations were based will be evaluated in light of fragmentary photochemical observations.
.
4.2.1
2
In almost every respect spectroscopic observations for DPB in solution parallel those for trans-stilbene and strongly suggest that the lowest excited state of DPB observed in fluorescence is the same lBu* state that is formed upon single photon excitation o f the ground state. An early indication o f this fact was the near identity between experimental and theoretical radiative rate constants, r = 1.2 (ref. 226). This initial result has been borne out by detailed studies of the T dependence of df and 7f in both nonpolar and polar solvents. Application of eq. 1 gives kf values that are nearly T and solvent independent. For instance, in the mixed hydrocarbon solvent methylcyclohexane/isohexane, 3:2, df increases from 0.22 at 50 "C to 0.91 at -50 'C, but the corresponding increase in 7f gives kf = (8.5 f 0.7) x lo8 s-l independent of 7 (ref. 265). Over the same temperature range df and 7f increase much more steeply in the more polar and viscous solvent, 1,2propanediol, (0.06-to 0.98 for df), but kf = (7.9 f 0.5) x lo8 s-l, remains
131
unaltered (ref. 265). The same result holds in the n-alkane solvent series, C5 - c16, for which kf = (7.7 f 0.4) x lo8 s-l (ref. 148). Especially significant in establishing the identity of the lowest excited singlet state as predominantly lBU* in character is the strong and similar dependence of Xmax'S for absorption and emission spectra on the solvent polarizability measure 0 , eq. 35, P = 6975 f 1300 cm-l and 6338 f 1000 cm-l for absorption and emission spectra, respectively (ref. 148). This contrasts the behavior of DPH and DPO for which emission spectra from predominantly l%* lowest excited states are independent of 0. Results from ns (refs. 266, 267) and ps (ref. 268) time scale lowest excited singlet state Raman spectroscopic studies are also consistent with the lBU* assignment for these states. A convincing case for this conclusion has been presented by Gustafson, Palmer and Roberts who applied ps time scale resonance Raman spectroscopy to obtain the spectrum of 0.025 M DPB in tetrahydrofuran (ref. 268). It was based on comparison with the spectrum of trans-stilbene in n-hexane, and with theoretical calculations of vibrational band frequencies (ref. 269), and reversed the earlier assignment (ref. 266) of lAg* character for the lowest excited singlet state of DPB. The 21Ag llAg transition of DPB in solution has been investigated by two photon fluorescence excitation spectroscopy (refs. 215, 270, 271). Employing a rather concentrated DPB solution, 0.01 M, in cyclohexane, a low resolution progression of broad vibronic bands was obtained whose origin appeared to be at 28,259 cm-I, somewhat below the first well-defined vibronic band maximum of the one-photon llBu llAg absorption spectrum at 28,800 cm-l(ref. 270). However, if the onsets of the two-photon and one-photon spectra were compared instead, the opposite conclusion would result, namely that the llBu state lies at lower energy than the ZIAg state (ref. 270). This becomes especially clear when the two-photon thermal lensing spectrum is compared to the one-photon absorption spectrum (ref. 271). An assignment of the 21Ag origin at 27,900 k 20 cm-l in EPA glass at 77 K, about 130 cm-1 below the llBu origin (ref. 272) similarly seems highly suspect. One-photon absorption and fluorescence spectra under these conditions show well-defined vibronic structure and reasonable mirror symmetry with lowest bands at 356.7 nm in absorption and 359.0 nm in emission (ref. 272). Instead of attributing these two spectra to formation and decay of the same llBu state with a small Stokes shift accounting for the apparent non-coincidence of 0-0 bands, Bennett and Birge attributed the absorption spectrum to the llBu + llAg transition and the emission to the ZIAg -* llAg transition (ref. 272). In making this bold assignment they re1 ied on a two-photon fluorescence excitation spectrum o f DPB +
-
-
132
in EPA whose origin at 27,900 f 20 cm-1 appeared to coincide with the origin of the fluorescence spectrum (ref. 272). Once again a highly concentrated solution, 5 x 10-3 M in EPA, was employed for the two-photon spectrum and the possible contribution of DPB aggregates was not considered, despite the fact that crystallization of DPB was observed at 10-2 M under the conditions of the experiment. The authors went so far as to conclude that DPB fluorescence originates from the 2lAS state even in cyclohexane solution, contrary to the evidence in the preceding paragraphs that unequivocally establishes its llBu nature. In rejecting the interpretation in ref. 272 we note that an analysis based on eq. 39 readily shows that if the 2lAS state were 130 cm-l below the llBu state Kba 11 can be calculated for 77 K and since kfb >> kfa emission from llBu would still be expected to be dominant in the fluorescence spectrum. It will be shown below that the conclusion that the llBu state lies below the ZIAg state of DPB in condensed media is consistent with spectroscopic observations for DPB in supersonic jets (refs. 273-275).
-
As with DPH and DPO, intersystem crossing to triplet states of DPB does not contribute significantly to the radiationless decay of DPB singlet states (refs. 223, 224, 240, 276). Essentially negligible intersystem crossing rate constants of 3.3 x lo7 s-l, 1 x lo7 5-1 and < 2.3 x lo7 s-l have been estimated for DPB at 295 K in cyclohexane, benzene and methanol, respectively (ref. 240). In keeping with the Birks and Birch extension (ref. 52) of the Orlandi and Siebrand photoisomerization mechanism (ref. 36), radiationless decay competing with fluorescence has usually been assigned to twisting about one of the olefinic bonds as the first step to photoisomerization (refs. 148, 165, 171, 174, 198, 265). Unfortunately there are several reasons for doubting this interpretation, most prominent of which is the observation of relatively small trans, trans cis, trans quantum yields, $tt-.ct. Following the initial report of trans -. cis photoisomerization of DPB in n-hexane by Zechmeister and co-workers (ref. 277), quantum yields for the direct photoisomerization were measured by Whitten et al. in benzene, $tt-.ct = 0.25, bt+t = 0.19 (ref. 278), and later by Yee et al. in cyclohexane, $tt-.,-t = 0.11, $,-t-.tt = 0.04 and $cc-ct = 0.20 (ref. 276). Taken together with df = 0.42 in cyclohexane, the low isomerization quantum yields strongly suggest the functioning of radiationless decay pathways which do not produce fully twisted intermediates in the diene portion of DPB (ref. 276). Also significant in this context is the demonstration that suppression of the photoisomerization pathway in the structural analog of DPB, 1,5-diphenyl-2,3,4,6,7,8hexahydronaphthalene, 22, increases rather than reduces the contribution
-
133
22 of radiationless decay to the deactivation of its lowest singlet excited state (ref. 279). Based on the mirror image symmetry of absorption and emission spectra, an experimental kf value greater than 6 x lo8 s-l, and the similarity of its Sn S1 spectrum to that of DPB, the lowest excited singlet state of 22 was identified as l*Bu (ref. 279). Absorption and fluorescence spectra of 22 in cyclohexane at room temperature are devoid o f vibronic structure, show a large Stokes shift, and thus bear a strong resemblance to those of cisstilbene, suggesting a large equilibrium geometry change between So and S 1 states (ref. 279). In view of the relative rigidity of the polyene unit, and of results from theoretical calculations, it was concluded that the significant change involves twists of the phenyl rings from a nonplanar geometry in So (phenyl-vinyl dihedral angles of 74 to an essentially planar S1 geometry (ref. 279). The phenyl ring motions appear to induce both S1 So radiationless decay [compare Qf = 0.06, Tf < 100 ps for 22 with Qf = 0.42, rf = 600 ps for DPB in cyclohexane (ref. 279)], and increase the rate constant for S1 -. Ti intersystem crossing [compare Qis = 0.12 for 22, with Qis = 0.02 for DPB in cyclohexane (ref. 279)]. Furthermore, Arrhenius plots of decay rates of DPB and 22 in 3-methylpentane are claimed to have virtually the same slope, leading to a rather low En, = 2.4 kcal/mol (see below), and bringing into question the need to postulate a radiationless decay pathway in DP8 involving large twisting motions in the olefinic portion of the molecule (ref. 279). +
-.
-
O )
-
It should be clear from the above that in the absence o f photochemical studies concerning the temperature dependence of trans cis photoisomerization quantum yields, detailed interpretation of the temperature and solvent dependence of the overall radiationless decay rate constants for DPB must be considered with some reservations. More than one radiationless decay pathway, each involving different geometry changes, is available to S 1 of this molecule, and may be subject to different inherent and/or medium imposed barriers. Furthermore, torsional motions involving the phenyl groups and the diene unit are probably coupled as suggested in a recent study of the
134
ps evolution of the Sn S1 spectrum of DPB (ref. 280). Using a ps continuum as a probe beam extending from 400 - 1000 nm, Rulliere et al. observed a broad S1 spectrum in methylcyclohexane at room temperature initially (30 ps) Sn centered at 625 nm, which at 100 ps had developed a shoulder at about 690 nm. The spectrum was generally consistent with the observation of a single band at 650 nm in cyclohexane in the ns time scale (ref. 246) but failed to reveal 440 nm (ref. the strong absorption reported by Chattopadhyay and Das at 247). The intensity of the band at 625 nm showed a rise time of 80 - 100 ps, while that of the shoulder at 690 nm had a slightly slower rise time of 100 ps following which both bands decayed at identical rates of 700 f 100 ps (ref. 2801, consistent with DPB's fluorescence lifetime. More pronounced spectral changes were observed at -70 'C, with the two bands being nearly identical in intensity at 10 ps and with only the band at 690 nm remaining prominent at t > 50 ps. The decay of the absorption gave 7 = 1.2 ns at -70 'C (ref. 280) consistent with measurements o f fluorescence lifetimes at low temperatures (refs. 148, 265). In accounting for these observations and for an enhancement of the absorption at 690 nm with increasing excitation pulse intensity Rulliere et al. advocate a more explicit version of a mechanism proposed by Goldbeck et al. (ref. 246). They suggest that the equilibrium ground state geometry corresponds to large phenyl -vlnyl bond torsional angles that stabilize the Z1% state and destabilize the llBu state. This leads to a sequence of events following initial formation of the nonplanar llBu state at the ground state equilibrium geometry, of twisting of the phenyl group to achieve a planar geometry at which the 116, state replaces the 2'% state as the lowest excited singlet state, and emission from the planar IIBu state in competition with twisting about the olefinic bonds through an Orlandi-Slebrand type mechanism, Figure 2b, c (ref. 280). In the rigid medium of EPA at 77 K twisting of the phenyl group is suppressed and the llBu state presumably decays to the Z1% state from where fluorescence occurs. This mechanism for all seasons is based on the interpretation of the EPA fluorescence spectra by Bennett and Birge (ref. 272) which was rejected above. It also suffers by not associating a radiationless S1 -. SO pathway with the motion of the phenyl groups as clearly suggested by the observations of Yee and co-workers (ref. 279) and by the low photoisomerization quantum yields (refs. 276, 278). Nonetheless, the main feature of the mechanism, namely the temperature dependent relaxation to an equilibrium mixture of conformers in the IIBu state in the ps time scale appears to be well founded in the experimental observations. Stabilization of the llBu state by torsional motion of the phenyl groups may indeed be the process that converts lBu* to l$* in Scheme 5 for the higher polyenes. +
+
-
-
135
Also relevant to this section are fluorescence lifetime measurements obtained at 20.0 'C in the n-a1 kane series for €,€-ditetrahydronaphthylideneethane, 11 (ref.,l60). This molecule was dubbed stiff-DPB and appears to have been synthesized with the purpose of determining the effect of phenyl motion inhibition on the decay characteristics of DPB (ref. 160). The observation of a nearly 10-fold increase in the radiationless decay rate constants of 11 relative to DPB [compare 7f = 54 and 288 ps for 11 (ref. 160) with 7f = 450 and 755 ps for DPB (ref. 148) in n-C5 and n-C15 alkanes, respectivel~l is reminiscent of the change obtained in going from the stiff-stilbene, 9, to trans-stilbene (ref. 84), and appears to lead to conclusions which contradict those obtained from the study o f 22. However, a six-membered ring is not nearly as stiff as a five-membered ring, and phenyl motions in 11 are much less hampered than phenyl motions in 9. Thts becomes especlally clear when absorption and fluorescence spectra of trans-stil bene analog 23 are compared
with those of trans-stilbene and of 9 (refs. 25, 281). In contrast to 9 whose spectra show much better defined vibronic structure than those of transstilbene under identical conditions (refs. 25, 281), the spectra of 23 reveal considerable structural flexibility and larger llAg, llBu equilibrium geometry difference by being less structured than those of trans-stilbene (ref. 281). Unfortunately, the spectra of 11 have not been published, but a calculated value of the radiative rate constant kf = 6.2 x lo8 s-l based on these spectra, suggests a less allowed transition in 11 than in DPB (ref. 160). Since structural models of DPB, that are intended to selectively restrict torsional motions, unavoidably introduce alkyl substitutions on the chromophore, it is important to evaluate the results for the possibility that substitution may differentially affect relative llBu and 21Ag energies. The conclusion seems clear that in both 11 and 22 the same order o f these two states that applies to unsubstituted DPB is preserved. Interestingly, initial
136
results with the DPB derivative 24 (ref. 243) suggest that alkyl substitution at the para positions of DPB profoundly increases the interaction between llBu and $'2 states. The evidence consists of (1) a decrease by a factor of two of the shift of the Xmax of the fluorescence spectrum of 24 relative to the shift in its absorption spectrum, and relative to the shifts of both absorption and emission spectra of DPB on changing the solvent from n-hexane to n-hexadecane, (2) the appearance of a second significant and longer lived component in the fluorescence decay of 24, and (3) a sharp decrease of the Arrhenius activation energy for knr in cyclohexane from 5.3 kcal/mol in DPB to 1.6 kcal/mol in 24 (ref. 243). These changes strongly suggest that the subtle substitution in 24 has narrowed the energy gap between llBu and 21Ag states. A study of the consequences of this change on the photoisomerization of 24 would therefore be highly desirable. In view of the diversity of decay pathways available to DPB and its structural analogs, the significance of treatments, based on the Kramers equation, o f medium effects on the overall radiationless decay rate constants of these molecules cannot be evaluated with confidence. Studies of the temperature dependence of isomerization quantum yields should, in time, allow the separation of radiationless decay processes into rate constants associated with geometry changes leading to photoisomerization and rate constants of radiationless decay processes associated with photochemically unproductive geometry changes. At this point it should suffice to state that medium effects on the overall knr values parallel those observed for trans-stilbene. Arrhenius treatments have yielded activation energies, Enr in the 5 - 6 kcal/mol range in hydrocarbon solvents (refs. 148, 174, 243, 265), and 4 - 7 kcal/mol in alcohols (refs. 165, 265). Attempts to obtain intrinsic barriers to radiationless decay have been based on isoviscosity Arrhenius plots and are subject to the limitations of this method (see Sections 3.3.1 and 3.3.2). Values of 4.7 f 0.5 kcal/mol (ref. 148) and 3.3 kcal/mol (ref. 160) were obtained for DPB and for 1 1 in n-alkanes, respectively, and 0.7 -+ 0.3 kcal/mol for DPB in n-alcohols (ref. 165). The decrease in barrier in the more polar alcohol solvents was taken to signify a higher polarity of the transition state for twisting than of the planar llBu state, and was based on the
137
Orlandi-Siebrand proposal of crossing between llBu and 21Ag states as in the case of stilbene (ref. 165). Evidence against this mechanism for stilbene has been discussed in detail, and applies to DPB with equal force. It is likely that, to the extent that twisting of the olefinic portion of the molecule contributes to knr, rotation occurs adiabatically in the lBU* state possibly leading directly to a zwitterionic twisted state, or diabatically by crossing into a higher lAg* state, analogous to 41Ag in stilbene, Figure 2d. Adding to the complexity of interpreting the DPB data are the very low fluorescence quantum yields that have been measured in polar solvents, Qf = 0.4 and 0.3 in ethanol and 2-methyltetrahydrofuran at -196 'C, respectively, under conditions where photoisomerization should be totally inhibited (ref. 224). No doubt, agreement of knr values in alcohols with the hydrodynamic approximation of Kramers equation, eq. 20, (ref. 165) is fortuitous. Application of the Kramers-Hubbard relation, eq. 25, succeeds in fitting changes in knr for 11 (ref. 160) in the n-alkane solvent series, but evaluation of the significance of this result must await further study. 4.2.3
VaDor DhaSe sDectroscoDic studies
Several investigations demonstrate conclusively that the lowest excited singlet state of jet-cooled DPB or of hot DPB vapors is the 21Ag state (refs. 251, 273-275). The lowest energy vibronic band observed for this transition was at 29,623 cm-l in the two-photon fluorescence excitation spectrum of jetcooled DPB reported by Shepanski, Keelan and Zewail (ref. 274). The ZIAg llAg origin is also observed in the one-photon fluorescence excitation spectrum of jet-cooled DPB as a weak progression o f vibronic bands with a vibronic coupling induced false origin placed at 29,653 cm-l (ref. 2731, 29,657 cm-1 (ref. 274) or 29,673 cm-1 (ref. 275), well below the origin of the dipole allowed llBu llAg transition at 30,700 k 50 cm-l (refs. 273-275). We note first that AEba = 1100 f 50 cm-l in the isolated molecule is much smaller than the energy gap in DPH and DPO, and that this fact alone would lead us to expect an inversion in the order of the 21Ag and llBu states on passing from the isolated molecule to the condensed phase. Using eq. 35, a red shift of about 1780 cm-l can be calculated for the llBu llAg transition upon taking vapor phase DPB into cyclohexane solution (ref. 148) and that would place the llBu state some 700 cm-l below the p o l a r i z a b i l i t y - i n s e n s i t i v e ZIAg state which should remain close to 29,600 cm-l. The inconsistency between this conclusion and the 27,900 cm-1 assignment for 21Ag llAg transition in EPA at 77 K (ref. 272) was recognized by Kohler and co-workers who proposed that different lAg states are accessed by two-photon fluorescence excitation spectroscopy in EPA glass and in the isolated molecule (ref. 273), +
.-
+
+
138
a suggestion that was abandoned later (ref. 251). consider the EPA assignment suspect.
As discussed above we
F1 uorescence 1 i fet imes (refs. 273 , 274) and quantum yields (ref. 275) under jet-cooled conditions vary with excess excitation energy in a manner consistent with emission from a single electronic state with radiative constant kf = (1.54 f 0.11) x 107 s-1 (ref. 275). The long radiative lifetime associated with this transition establishes its identity as the lowest lAg* state with borrowed transition probability from the nearby lBU* state. Experimental fluorescence lifetimes vary from as high as 63 ns (ref. 274) at zero excess energy (ref. 273 gives 52.8 ns) to as low as 426 ps at Ex = 1823 cm-l. The drop in lifetime is gradual up to Ex = 1050 cm-l, 7 = 31 ns (ref. 274), 7 = 27.9 ns (ref. 273), and much faster at higher Ex. Correspondingly, q+ begins at unity for Ex = 0 and drops in the same way as 7f at higher Ex (ref. 275). In the absence of additional information, discussions of the identity of the radiationless decay pathways competing with fluorescence must be considered speculative. The results do seem to suggest that the onset of fast radiationless decay coincides with the origin of the llBu llAg transition, and seem to discount isomerization with a small barrier in the ZIAg state as an important decay pathway. Kohler and coworkers attributed the sudden drop in 7f for Ex > 1050 cm-l to direct excitation into the llBu state (ref. 273). However, the demonstration by Jortner and coworkers that the radiative rate constant is independent of Ex shows that even when excitation is at sufficiently high energies to access the llBu state the emission of the isolated molecule is from the 21% state (ref. 275). They suggest that this is due to the higher density of vibrational states in the manifold of the lowest excited state which favors it in statistical distribution. The possibility exists that, as Ex increases, a high density of torsional states in the llBu allows torsional radiationless decay pathways to compete with the llBu -,21Ag process. Whatever the fate of the llBu state, its decay must be too fast to allow observation of its fluorescence from isolated DPB. The high qjf values at low Ex eliminate the .possibility (ref. 273) that a lower energy exists which is observable in condensed phase, but not in non-fluorescent the vapor phase. +
'As*
Evidence has been provided concerning the behavior of DPB excited singlet states under conditions intermediate between the isolated molecule and solution. In a study of DPB fluorescence, Itoh and Kohler measured fluorescence and fluorescence excitation spectra of DPB vapors at 368 K as a function of buffer gas pressure (perfluorohexane 0 - 260 Torr) and monitoring wavelength (ref. 251). In the absence of buffer gas a weak, diffuse and
139
structureless fluorescence spectrum, Amax = 350 nm is observed for excitation well into the lBU* region, Xexc = 290 nm. As Xexc is shifted to lower energies the fluorescence Amax shifts to 390 nm, the spectrum develops some structure and generally resembles the spectrum obtained from vapor seeded into a supersonic He expansion for similar excitation wavelength. Emission under these conditions is assumed to originate from unequilibrated 21Ag states (ref. 251). As the buffer gas pressure is increased the fluorescence spectrum eventually becomes almost independent of Xexc, increases in intensity by a factor of 40, resembles in appearance the fluorescence spectrum in solution, and is consequently assigned to the llBu state. The high buffer gas pressure provides a means for 11Bu/21Ag equilibration to be achieved through collisional transfer of population from the 21Ag to the llBu state and 'BU emission dominates due to kfb >> kfa, eq. 39 (ref. 251). Actually, based on Eba = 1050 cm-l, Kba = 40 can be calculated for 368 K. Assuming that the radiative lifetime in the free jet defines kfa gives (kfblkfa) = 25 which, according to eq. 39, would still require a considerable contribution of ZIAg fluorescence under equilibrium conditions. It seems that either the (kfdkfa) ratio was underestimated, or that the entropy difference favors Z1BU and leads to a much smaller Kba value.
In attempting to approach gas phase conditions in solution Courtney and Fleming measured DPB fluore-scence lifetimes as a function of 7 in liquid ethane, 41 atm, and liquid propane, 8.2 atm (ref. 171). The lowest viscosity achieved in this work, qs = 0.039 cP in ethane at 24 "C, was only twice that of air at 25 'C and gave the shortest 7f = 150 ps observed for DPB in alkane solution (ref. 171). Nonetheless, Arrhenius activation energies obtained, 5.3 f 0.9 kcal/mol and 5.1 f 0.4 kcal/mol for ethane and propane, respectively, were very similar to the values for the higher alkanes (ref. 148) and the new data showed no departure from eq. 22. It was concluded that since, even at the lowest viscosities, radiationless decay rate constants showed a continuous increase with decreasing viscosity, the Kramers theory turnover region had not been attained (ref. 171). Rate constants in ethane and propane solution have been compared with rate constants for radiationless decay of jet-cooled DPB (refs. 174, 198, 282). Comparison of collision rates with 7f led to the conlusion that a Boltzmann distribution characterized by the 7 should give an adequate description of vibrational populations (ref. 198). The average vibrational energy content, Ev, for DPB in solution, calculated using the normal mode vibrational frequencies of Pierce and Birge (ref. 269), was assumed equal to Ex for jet-cooled DPB, and isolated molecule and solution phase data were plotted together, Figure 18 (ref. 198). Though rate constants for ethane solution and for the jet-cooled molecule fall on the same line,
140
t
n
Fig. 18 Radiationless decay rate constants of DPB as a function of vibrational energy, 0 - jet results, X - in liquid ethane, t - in liquid propane; the vertical line shows the influence of solvent viscosity at 24 "C, Ev = 2270 cm-1. Reprinted in part from ref. 198 with permission of the copyright holder, the American Institute of Physics. this is probably fortuitous. In contrast to trans-stilbene, for which the llBu state remains lowest in energy in solution and gas phase, the spectroscopic evidence shows that the 21Ag state of DPB replaces the llB, state as lowest excited singlet state in the gas phase. Thus DPB radiationless rate constants are for the lBu* state in solution and for the lAg* state in the gas phase and are not expected to be related in any simple way. 4.2.4
Photoi somerization of 1.3-dienes
There is an obvious need for quantitative photochemical investigations of the fate of DPB excited states in solution and in the gas phase. Without such studies, the degree to which torsional motions about the olefinic portion of DPB contribute to its radiationless decay remains open to speculation. This is especially true in view of observations with DPB analog 22 which suggest that phenyl -vinyl torsional motions provide efficient radiationless decay pathways (ref. 279). Presently available observations (ref. 277-279) reveal strong similarities between the photochemical responses of DPB and those o f
141
the somewhat more thoroughly studied a1 kyl-substituted 1,3-butadienes (refs. 3, 4). These similarities are highlighted briefly in this section. AS with the 2,4-hexadienes (ref. 256) and the 1,3-pentadieneJ (refs. 256,
283) direct excitation of the DPB's in solution leads to photoisomerization about only one of the two double bonds (refs. 276-278). In all cases the photoisomerization quantum yields are sufficiently small to suggest some inefficiency in the formation of an allylmethylene excited state intermediate. No direct evidence for the formation of such an intermediate is available, nor is there any conclusive evidence that would lead to a preference for a diradical vs. a zwitterionic twisted species, though a case for the latter has been based on the observation of highly selective photoisomerization favoring the d-substituted double bond of cis-l-deuteriotrans-1,3-pentadiene (ref. 283). S1 + T I intersystem crossing is undetectable in the alkyl-substituted dienes (refs. 3, 4, 255, 256, 284) and makes a small contribution in the decay of DPB, 4is I 0.02 (refs. 223, 224, 240, 276). A pronounced decrease in the photoisomerization quantum yields of the 1,3pentadienes on changing the excitation wavelength from 254 nm (4t+c 0.083, Qc-.t = 0.100) to 229 nm (#t+c = 0.022, &,t = 0.025) (ref. 286) has been attributed to a higher preference for rotation about the 1,2 bond in the S1 state of the s-trans diene conformer which is preferentially excited at 229 nm (ref. 283). However, while this explanation seems plausible in accounting for the observations in the trans cis direction, it seems less SO for the cis trans direction since cis-1,3-pentadiene should be essentially devoid of s-cis conformers. In this context it is important to note that the quantum yield for cyclobutene formation, which is expected to have an excited s-cis conformer precursor, is about ten times larger from trans-1,3-pentadiene than from the cis isomer (refs. 286, 287). Also, indicating that excitation of scis conformers is of photochemical significance is the observation that cyclobutene formation is observed only from the tt isomer of the 2,4hexadienes (refs. 256, 288). Thus far, no evidence has been presented for the participation of s-cis-DPB in the photophysics or photochemistry of DPB. In fact the excitation wavelength independence of $f for DPB in solution has been used as a criterion for purity (ref. 148).
-
-.
-+
indicated earlier, DPB triplets have been studied using triplet excitation donors (refs. 224, 247, 276). The report that DPB triplets do not undergo trans -., cis photoisomerization (ref. 224) has been shown to be a consequence of the high DPB concentration employed (ref. 276). The functioning of a quantum chain mechanism, analogous to one found for the 1,3pentadienes and the 2,4-hexadienes (ref. 289), leads to very high effective As
142
tc-tt quantum yields which increase with increasing DPB concentration and give extremely tt-DPB-rich photostationary mixtures (ref. 276). The transfer of triplet excitation from transoid DPB triplets to tc-DPB occurs at much lower diene concentrations than the corresponding process in the 1,3-pentadienes and 2,4-hexadienes mainly because of the much longer lifetime of DPB triplets (refs. 223, 224, 247, 276, 284, 285). Generally, the long triplet lifetimes allow an equilibrium to be established between planar and allylmethylene triplet conformations. Excitation transfer occurs from planar conformations (large T1 - So energy gap) and radiationless decay to ground state isomers occurs from allylmethylene conformations. In fluid solutions common triplet states are formed independent of which diene isomer is employed and two-bond photoisomerization is observed following a single excitation event. In the case of the 2.4-hexadienes the temperature dependencies of one-bond to twobond isomerization ratios starting from tt and cc isomers are consistent with facile equilibration between trans-twisted, 3tp* and cis-twisted 3cp*,
t
--.
T tp* 3
-
3cp*
t
allylmethylene triplets, which is inhibited at lower temperatures owing to medium and possibly inherent barriers (ref. 290). This accounts for the shifting o f the mechanism from two-bond isomerization in fluid solution to one-bond isomerization in viscous media (ref. 290). Initial observations with DPB suggest a similar mechanism in fluid solution except that transoid double bonds are energetically more favored than in the 2,4-hexadienes (ref. 276). Accordingly, following triplet excitation transfer to any of the three DPB isomers the major triplet diene species in equilibrium in solution are 3tt* and 3tp* and, in contrast to the 2,4-hexadienes, two-bond isomerization is observed starting with the cc isomer but not starting with the tt isomer (ref. 276). Work to date provides no evidence for the formation of s-cis DPB triplets. The search for such evidence could be worthwhile since in the 2,4hexadienes and the 1,3-pentadienes s-trans and s-cis triplets have been shown to exhibit different trans/cis decay ratios (ref. 291). In fact in the 1,3pentadienes increased selectivity for s-cis conformers in the excitation transfer step with decreasing sensitizer triplet energy is a primary cause for the dependence of trans/cis photostationary ratios on the energy of the sensitizer (ref. 291).
143
5
ROTAMERISM
Molecular orbital calculations at all levels predict that upon excitation, flexible molecules containing conjugated double bonds will undergo reversal of single/double bond order. Consequently, conformational changes which occur freely in the ground state are often not energetically feasible in singlet and triplet lowest excited states and vice versa. It follows that conformations that are in equilibrium in the ground state become distinct noninterconverting excited molecules having different 1 ifetimes and favoring different radiationless decay paths. This is the basis o f Havingas' NEER principle (refs. 9, 257) to which we have alluded in different parts o f this Chapter and of which the different behavior of s-cis and s-trans polyene excited states addressed in the last section are prime examples. The excitation wavelength dependence of photoproduct distributions reflect differences in the absorption spectra of conformers (rotamers) which account for variations of compositions of excited rotamers with .,,A The phenomenon of rotamerism has been especially well documented by spectroscopic studies for diarylethenes and reviews of early work are available (refs. 292-295). In these olefins, mixtures of interconverting ground state conformers arise via rotation about 1-0 essential single bonds when the aryl groups involved are unsymmetrically substituted. A1 though there are now numerous examples of molecules for which complex photophysical behavior has been attributed to the presence of nearly isoenergetic interconverting ground state conformers, in this section we will focus on the two that have been most thoroughly studied: trans-l-phenyl-2-(2naphthyl)ethene, NPE, and trans-1 ,2-di- (2-naphthyl)ethene, DNE. 5.1 5.1.1
trans-l-Phenvl--2-/2-naDhthvl)ethene
SDectroscoDic studies
In the ground state NPE exists as an equilibrium mixture of two nearly isoenergetic conformers, NPE-A and NPE-B. Since the C1 - C2 bond in naphthalenes is substantially shorter than the C2 - C3 bond r1.381 A in DNE-A X-ray structure (ref. 296)], the structures of NPE-A and NPE-B can be regarded as analogous to s-cis and s-trans structures of polyenes, respectively. The analogy extends to the relative planarity of the NPE conformers. Based on the X-ray structure of DNE-A (see below) we can expect NPE-B to be nearly planar. The shorter H-H nonbonded distance, H3 - H!, = 2.16 A in NPE-B in the planar conformation leads to a small repulsion which
144
NPE-A
NPE-B
can be accommodated by a slight torsion (ca. 8 in DNE-A) about the C2 - C, bond. The corresponding H1 - H, interaction in NPE-B is more severe, since it occurs over the shorter C1 - C2 bond, and forces a larger departure from planarity. Experimentally, the two conformer equilibrium in NPE is reflected in Xexc dependent fluorescence spectra and biexponenti a1 fluorescence decay curves arising from the superposition of emissions from the non-equilibrating excited conformers (refs. 299, 300). Fischer‘s group sought to obtain the pure fluorescence spectra by selective excitation of NPE in a hydrocarbon glass at 88 K (ref. 299). Excitation at most X resulted in a spectrum with a sharply defined progression of vibronic bands which was assigned to the more extended and more stable conformer NPE-B while excitation at the red edge of NPE absorption gave a less resolved spectrum with 0-0 band at lower energy which was assigned to NPE-A. The possibility that the second spectrum is due to aggregates of NPE-B was not excluded, however, this is rendered unlikely by the observation that fluorescence spectra from aggregates of diarylethenes obtained by slow cooling of saturated hydrocarbon solvents are substantially broader and less structured (ref. 301). Birks et al. calculated approximate NPE pure conformer absorption and emission spectra via point by point analysis of the Xexc and Xem dependence of the preexponential parameters from biexponential fits to fluorescence decay curves (ref. 300). Principal component analysis with self-modeling, PCA-SM, a method proposed by Lawton and
145
Sylvestre ( r e f . 302) was shown by S a l t i e l and Eaker t o be equal t o the challenge of obtaining pure component fluorescence spectra d i r e c t l y by analysis of experimental NPE fluorescence spectra containing different contributions from the two conformers (ref. 303). The i n i t i a l application of PCA-SM was based on a small spectral input matrix and though the pure conformer spectra obtained were i n good agreement with Fischer's low T spectra (ref. 299) they were a t variance w i t h spectra obtained by Bartocci, Mazzucato and coworkers ( r e f . 304) through a more refined application of the Birks approach ( r e f . 300). PCA-SM analysis of a much larger spectral matrix consisting of fluorescence spectra of NPE in methylcyclohexane a t 30 'C f o r different Xexc and 02 concentrations has yielded the pure component spectra Structural assignments t o the spectra, shown i n Figure 19a ( r e f . 305).
X (nm)
X (nm)
F i g . 19 (a) Pure component NPE-A and NPE-B fluorescence spectra i n methylcyclohexane a t 30 'C from PCA-SM analysis. (b) Fluorescence Spectra of 1-methyl and 3-methyl NPE i n methylcyclohexane a t 30 "C (ref. 305).
consistent with those of previous workers ( r e f s . 297-299) were achieved by comparison with spectra of NPE analogs 25 - 27 which are structurally or
25
26
27
146
sterically constrained to a single conformation (refs. 303, 305). The fluorescence spectra of 25 and 26, Figure 19b, though slightly red shifted and somewhat broader, bear an unmistakable resemblance to the two pure component NPE spectra which are accordingly assigned as NPE-A and NPE-B. For each pair of spectra in Figure 19, the more planar structure corresponds to the spectrum with the better defined vibrational progression. The spectrum of the more planar analog, 27, (ref. 303) i s similar to that of 26 but shows peak to valley ratios which are intermediate between those in 26 and in NPE-B. Based on analysis o f fluorescence decay curves of NPE in methylcyclohexane lifetimes of 4 ns and 23 ns at 30 'C are associated with NPE-A and NPE-B, respectively (refs. 299, 304, 306), consistent with a Stern-Volmer constant ratio for quenching by 0 2 of 5, obtained independently from the spectral matrix leading to the spectra in Figure 19a. The ability to extract individual conformer Stern-Volmer constants from steady-state fluorescence spectra adds a new dimension to the study of photochemical and photophysical properties of flexible molecules, because it affords conformer specific rate constants for i ntermol ecul ar interactions
-
-
-
.
The Birks treatment o f biexponential decay parameters obtained by monitoring the decay at an isoemissive Xf, has allowed the decomposition of the NPE absorption spectrum in n-hexane (ref. 300) and in ethanol (ref. 306) into NPE-A and NPE-B contributions. Moreover, by assuming that the ratio of the molar absorptivity coefficients of the two conformers is equal to the ratio of the respective radiative rate constants, the ratio of their concentrations was obtained (refs. 300, 306). The more extended, s-trans conformer, NPE-B is more abundant in both n-hexane, 76% at 293 K, and in ethanol, 80% at 293 K, and the 7 dependence of the concentration ratio shows it to have the lower enthalpy content, AH = 0.68 kcal/mol in n-hexane and 0.72 kcal/mol in ethanol (ref. 306). An attempt to obtain pure component absorption spectra by PCA-SM treatment of absorption spectra at different 7's failed because broadening and polarizability induced shifts of the spectra of the individual conformers dominate the overall temperature effect (ref. 307). It is therefore highly doubtful that this approach succeeds in the case o f trans-l-phenyl-Z-(Z-anthyl)ethene as has been claimed (ref. 308). Simulation of NPE spectra using the appropriately shifted absorption spectra of analogs 25 and 27, on the other hand, gives very satisfactory results and predicts ([NPE-B]/[NPE-A]) = 3.0 in methylcyclohexane at 30 'C (ref. 307) in excellent agreement with the independently derived value (ref. 306).
-
Sn S1 NPE absorption observed in the 450 - 550 nm region following a 337 nm laser pulse changes with time (refs. 309, 310). The decay of this
147
absorption, measured in n-heptane at 300 I(, is biexponential with TA < 12 ns 20 ns (refs. 309, 310). Pulse excitation at the red edge of NPE's and T B S1 absorption in absorption spectrum, 352 nm, has yielded a rather broad Sn the same spectral regions whose decay time of 2 ns in air-saturated n-hexane at 298 K confirms its assignment to NPE-A (ref. 311). +
Steady state observations on the triplet sensitized cis-trans photoisomerization of NPE in benzene at 298 K led to the conclusion that the lifetimes of NPE triplets were in the 100 ns range (ref. 310). This was confirmed by a ns laser pulse study in which triplet energy donors were employed to produce NPE triplets and record Tn + T1 absorption spectra and decay times (ref. 312). The observations were consistent with the presence of 400 and 500 nm and two NPE triplet transients with absorption maxima at lifetimes of 150 ns and 80 ns, respectively (ref. 312, cf. also ref. 313). These triplets are not observed following direct excitation of NPE at 353 nm, indicating inefficient inetersystem crossing under these conditions (ref. 312). However, triplet transients are observed following direct excitation of NPE at lower T ' s (ref. 306, 314).
-
5.1.2
-
Photoi someri zation
Although no clear mechanistic picture emerged, many of the salient features of the NPE problem were considered by Hammond et al. in the first study of NPE photoisomerization (ref. 315). Wavelength dependent photoisomerization quantum yields for the direct photoisomerization were reported and the possible involvement of NPE-A and NPE-6 excited states was considered (ref. 315). It was also pointed out that the low isomerization quantum yields could reflect localized excitation in the naphthalene unit and retention in the excited state of a larger barrier to rotation about the Wavelength dependent quantum olefinic bond than in stilbene (ref. 315). yields were also reported by Fischer and coworkers (ref. 316) and later by Kovalenko et a1 ., who first concluded that they reflected different photoreactivities of the two conformers (ref. 317). The viability of a triplet mechanism for the photoisomerization was established by the observation of relatively high quantum yields, $t-c = 0.64, Qc4t = 0.12, with benzophenone as the triplet energy donor (ref. 315). Also, the quenching of NPE fluorescence by 02 has been shown to be accompanied by increased isomerization quantum yields that are attributed to 02 enhanced intersystem crossing (refs. 241, 242, 317, 318). Nonetheless, the absence of the long-lived triplet transients at room temperature following direct NPE
148
excitation (refs. 309, 311-313), led Saltiel and Eaker to propose a singlet pathway for NPE photoisomerization (ref. 310). This mechanism has been modified by Mazzucato and coworkers (ref. 306) who presented evidence suggesting that a mixed triplet and singlet mechanism operates with the singlet mechanism dominating at T ' s higher than room temperature and the triplet mechanism dominating at lower 7's. According to Mazzucato and in disagreement with earlier results from Fischer's group (ref. 319), q5f for NPE increases as T is lowered from 347 K to limiting values of 0.58 for methylcyclohexane, and 0.73 for ethanol at T I 265 K (ref. 306). Correspondingly, 7f for the two conformers increases to limiting values of 3.9 and 23.0 ns in methylcyclohexane, and 4.1 and 20.2 ns in ethanol, for NPE-A and NPE-6, respectively, Figure 20 (ref. 306). Arrhenius plots of the
-
-
17
2.8
I
3.0
.y..
-
3.2
PK 103(~-')
1
3.4
I
Fig. 20 Variation of fluorescence lifetime with T for (a) NPE-A and (b) NPEB in (-) methylcyclohexane and (- - -) ethanol. The corresponding Arrhenius plots are shown on the right. Reprinted from ref. 306 with permission of the copyright holder, the Royal Society of Chemistry. temperature dependence of 7f give similar activation parameters for the two conformers Etp = 10.5 and 10.3 kcal/mol in methylcyclohexane, 8.1 and 8.2 kcal/mol in ethanol and logAtp = 14.3 and 13.8 in methylcyclohexane and 13.8
149
and 12.8 in ethanol, for NPE-A and NPE-B, respectively. The high frequency factors rule out intersystem crossing as the temperature dependent radiationless process competing with fluorescence and the process is accordingly assigned to It*+lp* twisting (ref. 306), consistent with the large isomerization yields reported by Fischer and coworkers at high T ’ s , 4t+c 2 : 0.50 at 366 nm, 0.35 at 334 nm, both measured at 363 K in methylcyclohexane (ref. 316). The T independent process competing with fluorescence at lower T ‘ s has been assigned to intersystem crossing (ref. 306) and is associated with progressively decreasing $t+c values observed at lower T ’ s (ref. 316). In both n-hexane and ethanol the radiative rate constant of NPE-A is significantly larger, 1.6 x 108 and 1.1 x lo8 s-l, than that of NPE-B 1.8 x lo7 and 3 x lo7 s-1, respectively (ref. 306). Based on M.O. calculations it was concluded that S1 for NPE-B shows primarily naphthalenic character whereas that for NPE-A is a mixed state exhibiting both ethylenic and naphthalenic characteristics (ref. 306). Although the NPE data are sparce the solvent dependencies of 7f and kf are reminiscent of observations on DPH (ref. 318). The major effect of replacement of the phenyl group of stilbene with the 2-naphthyl group appears to be a substantial increase in the internal barriers for lt*+lp* in both conformers, leading to diminished isomerization and increased fluorescence quantum yields (ref. 306).
-
5.2
trans-l,2-Di (2-na~hthvlIethene
Rotation of the two naphthyl groups about the essential single bonds connecting them to the ethene moiety leads to three DNE conformers that exist in dynamic equilibrium in the ground state. The three structures are analogous to those expected for DPH and other symmetrically 1,6-disubstituted 1,3,5-trienes. X-ray crystallographic measurements show that the molecular structure of DNE in the crystal corresponds to ONE-A (ref. 296). As discussed for the NPE conformers, due to the unequal bond lengths in the naphthalene units 1.381 and 1.427 A for C1 - Cp and C2 - C3 bonds, respectively (ref. 296) nonbonded H-H interactions worsen as the system departs from DNE-A, the most extended conformation, and assumes the less extended conformations DNE-B and DNE-C (ref. 298). These interactions are relieved primarily by torsions about the aryl-vinyl bonds and thus the equilibrium geometries of the three DNE conformers are expected to become progressively less planar in the order A, B, c
L.
The presence of three DNE conformers in fluid solution at room temperature was first demonstrated by selective excitation of fluorescence with different Xexc (ref. 294) and spectra consistent with this interpretation were given for
150
A
DNE selectively excited in a polyethylene film at -100 'C (ref. 293). The film was thought to allow the freezing in of higher T equilibrium conformer distributions and thus to favor the higher energy conformers B and C (ref. 293). Similar results have been obtained in a polymethylmethacrylate matrix (ref. 292). Even larger nonequilibrium contributions of DNE-C fluorescence have been achieved by exciting trans-DNE obtained by cis -, trans photoisomerization of the cis isomer in frozen ethanol or decalin at -175 and -145 'C, respectively (ref. 320 cf. also ref. 321). The presence of two components, B and C, in DNE fluorescence obtained by excitation of the long wavelength onset of ONE absorption had also been inferred from the differential quenching of this fluorescence by CCl4 (ref. 322). Pulsed excitation fluorescence studies, on the other hand, could only establish the presence of a minimum of two DNE emitting species. Biexponential decay was observed in argon-flushed methylcyclohexane, 71 = 2.1 - 2.3 ns, 72 = 7.2 - 7.5 ns (ref. 299) or in degassed cyclohexane, 71 = 1.92 f 0.04 ns, 72 = 8.6 f 0.7 ns (ref. 323), and the short-lived component could be observed exclusively following excitation in the red edge of DNE absorption, 71 = 2.1 and 1.7 ns for Xexc = 369 and 373 nm, respectively (ref. 299).
151
Time-resolved emission spectra in cyclohexane at 20 'C, Xexc = 297 nm, yielded a long-lived species with Xmax at roughly 359, 381, 399 and 425(sh) nm (30 ns delay time, 3 ns gate) and a broader spectrum for the short-lived species with new Xmax at approximately 371, 391, 415(sh), and 450(sh) nm (0 ns delay time, 3 ns gate) (ref. 323). Evidence presented below has established that the spectrum of the long-lived species is due to DNE-A, whereas the early spectrum shows mainly spectral features of DNE-B and a small contribution from DNE-A and possibly from DNE-C (refs. 324-326). The inability to time-resolve the spectra of DNE-B and DNE-C has been traced to their nearly identical lifetimes. It is important to note that all transients in both NPE and DNE form promptly, and no evidence for a grow in of long-lived components at the expense of the short-lived components has been observed even in the ps time scale (ref. 323). This establishes that the NEER principle condition is fulfilled for these compounds, in contrast to observations on 2vinylanthracenes (refs. 327-330). Steric hindrance involving the 3-methyl group(s) and the d - H ( s ) in DNE-A and DNE-B restricts DNE(3m) to an equilibrium mixtures of conformers B and C, and DNE(3,3'm2) to the single conformer C. Approximate resolution and structural assignment of three components of DNE fluorescence was initially achieved by allowing the spectra (appropriately shifted) of the methyl derivatives, DNE(3m) and DNE(3,3'm2) to establish the limits for DNE-B and DNE-C in the PCA-SM treatment of a matrix o f fluorescence spectra of all three DNE compounds at different Xexc and [02] (ref. 324). Improved solutions for pure component fluorescence spectra for DNE-B and DNE-C were later achieved without relying on the methyl derivative spectra by applying PCA-SM on a matrix of fluorescence spectra obtained for Xexc = 366 nm and 0 s [CC14] I 1 M in methylcyclohexane (ref. 325). This result took advantage of Fischer's finding that excitation at the onset of DNE absorption gives only the shortlived component (ref. 299), Ghiggino's report that CCl4 differentially quenches the resulting DNE-B and DNE-C fluorescence (ref. 322), and our extension of Lawton and Sylvestre's PCA-SM method to three components (ref. 325). The need for the.three component analysis was due to the presence of variable amounts of scattered excitation light in the experimental spectra. Finally, application of the PCA-SM method separately to two matrices consisting of fluorescence' spectra for different Xexc and quencher, CCl4 or 02, concentrations gave nearly identical pure component spectra, Figure 21 (ref. 326). The choice of pure component spectra was refined by imposing the additional constraint that the Stern-Volmer quenching constants, KSV, for each quencher be independent of Xexc (ref. 326). Individual Ksv values obtained for each conformer quantitatively demonstrate preferential quenching of DNE-B
152
X(nm)-
Fig. 21 Pure component fluorescence spectra (corrected for nonlinearity of instrumental response) and pure component fluorescence excitation spectra (uncorrected for nonlinearity of instrumental response) for the DNE/CC14 system: DNE-A (- - -), DNE-B (-), DNE-C (- -). Fluorescence spectra reprinted from ref. 326 (note correction) with permission of the copyright holder, the American Chemical Society; fluorescence excitation spectra from ref. 331. fluorescence by CCl4, and show that DNE-B and DNE-C have identical Ksv values for 0 2 quenching, as expected for the diffusion-controlled quenching of excited states with nearly identical lifetimes (ref. 326). The use of methyl derivatives as guides to pure component spectra was evaluated quantitatively and shown to provide a viable alternative approach for obtaining reasonable approximations of such spectra when more direct methods fail (ref. 326). Also shown in Figure 21 are fluorescence excitation spectra (uncorrected for nonlinear instrumental response) for the three DNE conformers. These were obtained by applying PCA-SM on a matrix of fluorescence spectra obtained by systematically changing Xexc so that columns of the spectral input matrix correspond to DNE fluorescence excitation spectra at specific Xf (ref. 331). The excitation spectra arise naturally from the solution of this square emission-excitation matrix since they are given by the X dependence of the fractional contributions of the respective pure component fluorescence spectrum to the experimental spectra (ref. 331).
153
6
CONCLUDING REMARKS
In the above we have attempted to present a comprehensive and critical account of what is known concerning the mechanisms of the singlet pathways of olefin cis-trans photoisomerization. In approaching this enormous task we have opted for detail and thoroughness of presentation of a few well-studied systems rather than for broadness of coverage. It is hoped that we have succeeded in doing justice to the remarkable progress that has been made since the 1980 review (ref. 4 ) , especially by the application of laser spectroscopy to the problem, and in providing a critical and fresh evaluation of the work that suggests directions for future research efforts. We certainly found the preparation of this Chapter most stimulating and highly rewarding to our own research in this area. It is fitting to conclude with a word of gratitude to the editors of this book for not taking no for an answer when they first proposed this task. Acknowledgment. CHE-8713093.
The preparation of this Chapter was supported by NSF Grant
154 REFERENCES 1
2 3 4 5 6
7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37
D. L. Ross and J. Blanc, Photochromismby Cis-Trans Photoisomerization, i n : G. H. Brown (Ed.), Photochromism, Wiley, New York, 1971, p. 471. 6. Haucke and R. Paetzold, Nova Acta Leopold. Suppl. 11 (1978). J. S a l t i e l , J. D'Agostino, E. D. Megarity, L. Metts, K. R. Neuberger, M. Wrighton and 0. C. Z a f i r i o u , Org. Photochem., 3 (1973) 1. J. S a l t i e l and J. L. Charlton, Cis-Trans Isomerization of Olefins, i n : P. de Mayo (Ed.), Rearrangements i n Ground and Excited States, Vol. 3, Academic Press, New York, 1980, p. 25. R. R. Birge, Ann. Rev. Biophys. Bioeng., 10 (1981) 315. F. Siebert, S t a t i c and Time-Resolved I n f r a r e d D i f f e r e n c e Spectroscopy Applied t o Rhodopsin and Bacteriorhodopsin, i n : C. Sandorfy and T. Theophanides (Eds.), Spectroscopy of B i o l o g i c a l Molecules, D. Reidel Publishing Co., Dordrecht, Holland, 1984, 347. H. Morrison and R. M. Deibel, Photochem. Photobiol., 43 (1986) 663. S. E. Braslavsky, Pure Appl. Chem., 56 (1984) 1153. H. J. C. Jacobs and E. Havinga, Adv. Photochem., 11 (1979) 305. R. S. M u l l i k e n and C. C. J. Roothaan, Chem. Rev., 41 (1947) 219. A. J. Meyer and R. S. Mulliken, Chem. Rev., 63 (1969) 639. 6. S. Hammond, J. S a l t i e l , A. A. Lamola, N. J. Turro, J. S. Bradshaw, D. 0. Cowan, R. C. Counsell, V. Vogt and C. Dalton, J. Am. Chem. SOC., 86 (1964) 3197. T. A r a i , T. Karatsu, H. Sakuragi, and K. Tokumaru, Tetrahedron Lett., 24 (1983) 2873. T. Karatsu, T. Arai, H. Sakuragi, and K. Tokumaru, Chem. Phys. Lett., 115 (1985) 9. H. Hamaguchi, M. Tasumi, T. Karatsu, T. Arai, and K. Tokumaru, J. Am. Chem. SOC., 108 (1986) 1698. H. Gorner, J. Photochem. Photobiol. A, 40 (1987) 325. R. A. Caldwell, Pure Appl. Chem., 56 (1984) 1196. R. A. Caldwell, L. Carlacci, C. E. Doubleday, Jr., T. R. F u r l a n i , H. F. King, and J. W. McIver, Jr., J. Am. Chem. SOC., 110 (1988) 6901. T. Ni, R. A. Caldwell, and L. A. Melton, J. Am. Chem. SOC., 111 (1989) 457. S. Malkin and E. Fischer, J. Phys. Chem., 68 (1964) 1153, and e a r l i e r papers i n t h i s series. R. H. Dyck and 0. S. McClure, J. Chem. Phys., 36 (1962) 2336. K. A. Muszkat, D. Gegiou, and E. Fischer, J. Am. Chem. SOC., 89 (1967) 4814. D. .^Gegiou, K. A. Muszkat, and E. Fischer, J. Am. Chem. SOC., 90 (1968), IZ.
S. Sharafi and K. A. Muszkat, J. Am. Chem. SOC., 93 (1971) 4119. J. S a l t i e l and 3. T. D'Agostino, J. Am. Chem. SOC., 94 (1972) 6445.
J. S a l t i e l , A. Marinari, 0. W. L. Chang, J. C. Mitchener, and E. D. Megarity, J. Am. Chem. SOC., 101 (1979) 2982. M. Sumitani, N. Nakashima, K. Yoshihara, and S. Nagikura, Chem. Phys. L e t t . , 51 (1977) 183. J. S a l t i e l , E. D. Megarity, and K. 6. Kneipp, J. Am. Chem. SOC., 88 (1966) 2336. J. S a l t i e l and E. D. Megarity, J. Am. Chem. SOC., 91 (1969) 1265. J. S a l t i e l and E. D. Megarity, J. Am. Chem. SOC., 94 (1972) 2742. J. S a l t i e l , J. Am. Chem: SOC., 89 (1967) 1036; 90 (1968) 6394. R. Bonneau, J. Am. Chem. SOC., 104 (1982) 2921. 0. V. Bent and D. Schulte-Frohlinde, J. Phys. Chem., 78 (1974) 446; 78 (1974) 451. J. S a l t i e l , 0. C. Z a f i r i o u , E. 0. Megarity, and A. A. Lamola, J. Am. Chem. SOC., 90 (1968) 4759. C. D. DeBoer and R. H. Schlessinger, J. Am. Chem. SOC., 90 (1968) 803. G. Orlandi and W. Siebrand, Chem. Phys. L e t t . , 30 (1975) 352. B. S. Hudson and B. E. Kohler, Chem. Phys. L e t t . , 14 (1972) 299.
155 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78
B. S. Hudson and B. E. Kohler, J. Chem. Phys., 59 (1973) 4984. B. S. Hudson and B. E. Kohler, Ann. Rev. Phys. Chem., 25 (1974) 437. P. B o r r e l l and H. H. Greenwood, Proc. Roy. SOC. London Ser. A, 298 (1967) 453. C.-H. Ting and D. S. McClure, J. Chin. Chem. SOC. (Taipei), 18 (1971) 95. F. Momicchioli, M. C. Bruni, I . Baraldi, and 6. R. Corradini, J. Chem. SOC. Faraday Trans. 11, 70 (1974) 1325. P. Tavan and K. Schulten, Chem. Phys. L e t t . , 56 (1978) 200. 6. Orlandi, P. Palmieri, and G. Poggi, J. Am. Chem. SOC., 101 (1979) 3492. J. B. B i r k s , Chem. Phys. L e t t . , 54 (1978) 430. T. M. Stachelek, T. A. Pazoha, W. M. McClain, and R. P. Drucker, J. Chem. Phys., 66 (1977) 4540. K. Fuke, S. A. Sakamoto, M. Ueda, and M. Itoh, Chem. Phys. Lett., 74 (1980) 546. G. Hohlneicher and B. Dick, J. Photochem., 27 (1984) 215. 6. Olbrich, Ber. Bunsenges. Phys. Chem., 86 (1982) 209. A. M a r i n a r i and J. S a l t i e l , Mol. Photochem., 7 (1976) 225. J. L. Charlton and J. S a l t i e l , J. Phys. Chem., 8 1 (1977) 1940. J. B. B i r k s and D. J. S. Birch, Chem. Phys. Lett., 31 (1975) 608. J. Troe and K.-M. Weitzel, J. Chem. Phys., 88 (1988) 7030. G. N. Lewis, T. T. Magel, and D. L i p k i n , J. Am. Chem. SOC., 62 (1940) 2973. H. Stegemeyer and H. H. Perkampus, Z. Phys. Chem. ( F r a n k f u r t am Main) [N. S.], 39 (1961) 125. F. Aurich, M. Hauser, E. Lippert, and H. Stegemeyer, Z. Phys. Chem. (Frankfurt am Main) [N. S.], 42 (1964) 123. E. L i p p e r t , Z. Phys. Chem. (Frank,furt am Main) [N. S.], 42 (1964) 125. A. A. Lamola, G. S. Hammond, and F. B. Mallory, Photochem. Photobiol., 4 (1965) 259. F. B. Mallory, C. S. Wood, and J. T. Gordon, J. Am. Chem. SOC., 86 (1964) 3094. K. A. Muszkat and E. Fischer, J. Chem. SOC. B, (1967) 662. F. B. M a l l o r y and C. W. Mallory, Organic Reactions, 30 (1984) 1. K. Yoshihara, A. Namiki, M. Sumitani, and N. Nakashima, J. Chem. Phys., 71 (1979) 2892. D. J. S. B i r c h and J. B. Birks, Chem. Phys. Lett., 38 (1976) 432. E. Heumann, W. T r i b e l , R. Uhlmann, and B. Wilhelmi, Chem. Phys. Lett., 45 (1977) 425. 0. Teschke, E. P. Ippen, and G. R. Holtom, Chem. Phys. L e t t . , 42 (1977) 233. J. R. Taylor, M. C. Adams, and W. Sibbett, Appl. Phys. Lett., 35 (1979) 590. J. R. Taylor, M. C. Adams, and W. Sibbett, J. Photochem., 12 (1980) 127. F. Heisel, J. A. Miehe, and B. Sipp, Chem. Phys. Lett., 61 (1979) 115. U. Steiner, M. H. Abdel-Kader, P. Fischer, and H. E. A. Kramer, J. Am. Chem. SOC., 100 (1978) 3190. R. M. Hochstrasser, Pure Appl. Chem., 52 (1980) 2683. M. Sumitani and K. Yoshihara, B u l l Chem. SOC. Jpn., 55 (1982) 85. H. P. Good, U. P. Wild, E. Haas, E. Fischer, E.-P. Resewitz, and E. L i p p e r t , Ber. Bunsenges. Phys. Chem., 86 (1982) 126. 8. I . Greene, R. M. Hochstrasser, and R. B. Weisman, Chem. Phys. L e t t . , 62 (1979) 427. B. I . Greene, R. M. Hochstrasser, and R. B. Weisman, J. Chem. Phys., 70 (1979) 1247. M. Sumitani and K. Yoshihara, J. Chem. Phys., 76 (1982) 738. M. Sumitani, N. Nakashima and K. Yoshihara, Chem. Phys. L e t t . , 68 (1979) 255. B. I . Greene and T. W. Scott, Chem. Phys. L e t t . , 106 (1984) 399. F. E. Doany, R. M. Hochstrasser, B. I . Greene, and R. R. M i l l a r d , Chem. Phys. L e t t . , 118 (1985) 1.
156 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99
100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118
F. E. Doany, E. J. H e i l w e i l , R. Moore, and R. M. Hochstrasser, J. Chem. Phys., 80 (1984) 201. J. S a l t i e l , 0. C. Z a f i r i o u , and Y.-P. Sun, unpublishedr e s u l t s . J. Vogel, S. Schneider, F. Dorr, P. Lenmen, and 0. Lenoir, Chem. Phys., 90 (1984) 387: K. Ogawa, H. Suzuki, and M. Futakami, J. Chem. SOC. Perkin Trans. I1 (1988) 41. R. N. Beal and E. M. F. Roe, J. Chem. SOC. (1953) 275. G. Rothenberger, 0. K. Negus, and R. M. Hochstrasser, J. Chem. Phys., 79 (1983) 5360. C. L. S c h i l l i n g and E. F. H i l i n s k i , J. Am. Chem. SOC., 110 (1988) 2296. H. Stegemeyer, Ber. Bunsenges. Phys. Chem., 72 (1968) 335. P. F. Barbara, S. 0. Rand, and P. M. Rentzepis, J. Am. Chem. SOC., 103 (1981) 2156. B. I.Greene, Chem. Phys. L e t t . , 79 (1981) 51. W. 6. Dauben and J. S. Ritscher, 3. Am. Chem. SOC., 92 (1970) 2925. W. G. Dauben, M. S. Kellog, J. I. Seeman, N. 0. Wietmeyer, and P. H. Wendschuh, Pure Appl. Chem., 33 (1973) 197. V. BonaciC-Kouteckj, P. Bruckman, P. Hiberty, J. Kouteckj, C. Leforestier, and L. Salem, Agnew. Chem. I n t . Ed. Engl., 14 (1975) 575. L. Salem, Acc. Chem. Res., 12 (1979) 87. W. R e t t i g , Angew. Chem. I n t . Ed. Eng., 25 (1986) 971. V. Bonacid-Kouteckj, J. Kouteckj, and J. Michl, Angew. Chem. I n t . Ed. Engl., 26 (1987) 170. J. P. Malrieu, I. Nebot-Gil, and J. Sinchez-Marin, Pure Appl. Chem., 56 (1984) 1241. I.Nebot-Gil and J. P. Malrieu, J. Am. Chem. SOC., 104 (1982) 3320. V. Bonacid-Kouteckl, M. Persico. D. Dohnert. and A. Sevin, J. Am. Chem. SOC., 104 (1982) 6900. M. C. Bruni, F. Momicchioli, and 1. Baraldi, Chem. Phys. Lett., 36 (1975) 484. G. L. Bendazzoli, 6. Orlandi, P. Palmieri, and 6. Poggi, J. Am. Chem. SOC., 100 (1978) 392. 6. Orlandi, P. Palmieri, and 6. Poggi, J. Chem. SOC. Faraday Trans. 11, 77 (1981) 71. I . Nebot-Gil and J. P. Malrieu, Chem. Phys. Lett., 84 (1981) 571. M. Said and J. P. Malrieu, Chem. Phys. L e t t . , 102 (1983) 312. B. I.Greene, R. M. Hochstrasser, and R. B. Weisman, J. Chem. Phys., 71 (1979) 544. B. I. Greene, R. M. Hochstrasser, and R. B. Weisman, Chem. Phys., 48 (1980) 289. J. A. Syage, Wm. R. Lambert, P. M. Felker, A. H. Zewail, and R. M. Hochstrasser, Chem. Phys. Lett., 88 (1982) 266. J. A. Syage, P. M. Felker, and A. H. Zewail, J. Chem. Phys., 81 (1984) 4685. J. A. Syage, P. M. Felker, and A. H. Zewail, J. Chem. Phys., 81 (1984) 4706. P. M. Felker and A. H. Zewail, J. Phys. Chem., 89 (1985) 5402. A. Amirav and J. Jortner, Chem. Phys. Lett., 95 (1983) 295. T. J. Majors, U. Even, and J. Jortner, J. Chem. Phys., 8 1 (1984) 2330. A. Amirav, M. Sonnenschein, and J. Jortner, Chem. Phys., 88 (1984) 199. A. Warshel, J. Chem. Phys., 62 (1975) 214. Z. MeiC and H. Gusten, Spectrochim. Acta Part A, 34 (1978) 2326. M. Edelson and A. Bree, Chem. Phys. L e t t . , 41 (1976) 562. A. Bree and M. Edelson, Chem. Phys., 51 (1980) 77. A. Bree, R. Zwarich, and M. Edelson, Spectrochim. Acta P a r t A, 38 (1982) 719. W. M. Hetherington I11 and B. S. Hudson, Chem. Phys. Lett., 65 (1979) 261. M. Traetterberg, E. B. Frantsen, F. C. M i j l h o f f , and A. Koekstra, J. Mol. Struct., 26 (1975) 57.
157
119 T. Kobayashi, H. Suzuki, and K. Ogawa, B u l l Chem. SOC. Jpn., 55 (1982) 1734. 120 C. H. Ting, Ph.D. Thesis, U n i v e r s i t y o f Chicago, Chicago, I L (1965). 121 T. L. Gustafson, D. M. Roberts, and D. A. Chernoff, J. Chem. Phys., 79 (1983) 1559. 122 H. Hamaguchi, T. Urano, and M. Tasumi, Chem. Phys. L e t t . , 106 (1984) 153. I23 M. Sonnenshein, A. Amirav, and J. Jortner, J. Phys. Chem., 88 (1984) 4214. 124 T. Shibuya, Chem. Phys. L e t t . , 103 (1983) 46. 125 L. R. Khundkar, R. A. Marcus, and A. H. Zewail, J. Phys. Chem., 87 (1983) 2473. 126 R. A. Marcus, J. Chem. Phys., 20 (1952) 359. 127 J. Troe, Chem. Phys. Lett., 114 (1985) 241. 128 J. Schroeder and J. Troe, J. Phys. Chem., 90 (1986) 4216. 129 B. I. Greene and R. C. Farrow, J. Chem. Phys., 78 (1983) 3336. 130 A. B. Myers and R. A. Mathies, J. Chem. Phys., 81 (1984) 1552. 131 R. A. Caldwell, H. Misawa, E. F. Healy, and M. J. S. Dewar, J. Am. Chem. SOC., 109 (1987) 6869. 132 L. H. Spangler, R. D. van Zee, S. C. Blankespoor, and T. S. Zwier, J. Phys. Chem., 91 (1987) 6077. 133 S. H. Courtney, M. W. Balk, L. A. P h i l i p s , S. P. Webb, D. Yang, D. H. Levy, and G. R. Fleming, J. Chem. Phys., 89 (1988) 6697. 134 J. S a l t i e l , J. T. D'Agostino, W. 6. Herkstroeter, 6. Saint-Ruf, and N. P. Buu-Hoi', J. Am. Chem. SOC., 95 (1973) 2543. 135 R. A. Caldwell, Pure Appl. Chem., 56 (1984) 1167. 136 K. Rademann, U. Even, S. Rozen, and J. Jortner, Chem. Phys. Lett., 125 (1986) 5. 137 D. H. McDaniel and H. C. Brown, J. Org. Chem., 23 (1958) 420. 138 0. Exner, A C r i t i c a l Compilation o f Substituent Constants, in: N. B. Chapman and J. Shorter (Eds.), C o r r e l a t i o n Analysis i n Chemistry, Plenum, New York, 1978, p. 439. 139 R. Srinivasan and J. C. Powers, Jr., J. Am. Chem. SOC., 85 (1963) 1355. 140 R. S. Becker and M. Kasha, Luminescence and Spectroscopy o f Molecules and t h e Photosynthetic System, i n : F. H. Johnson (Ed.), Luminescence of B i o l o g i c a l Systems, AAAS, Washington, D.C., 1955, p. 25. 141 L. A. Brey, G. B. Schuster, and H. G. Drickamer, J. Am. Chem. SOC., 101
(1979) 129. 142 L. A. Brey, G. B. Schuster, and H. G. Drickamer, J. Chem. Phys., 71 (1979) 129. 143 B. D e l l i n g e r and M. Kasha, Chem. Phys. L e t t . , 38 (1976) 9. 144 J. S a l t i e l and Y.-P. Sun, J. Phys. Chem., 93 (1989) 6246. 145 T. A. L i t o w i t z , J. Chem. Phys., 20 (1952) 1088. 146 A. K. D o o l i t t l e , J. Appl. Phys., 22 (1951) 1471. 147 M. H. Cohen and D. Turnbull, J. Chem. Phys., 31 (1959) 1164. 148 S. P. Velsko and G. R. Fleming. J. Chem. Phys., 76 (1982) 3553. 149 6. Oster and Y. Nishijima, J. Am. Chem. SOC., 78 (1956) 1581. 150 B. Bagchi, G. R. Fleming, and 0. W. Oxtoby, J. Chem. Phys., 78 (1983) 7375. 151 T. Forster and G. Hoffman, Z. Phys. Chem. ( F r a n k f u r t am Main), 76 (1971) 63. 152 L. A. Brey, G. B. Schuster, and H. 6. Drickamer, J. Chem. Phys., 67 (1977) 2648. 153 M. D. Hirsch and H. Mahr, Chem. Phys. Lett., 60 (1979) 299. 154 H. A. Kramers, Physica, 7 (1940) 284. 155 S. Chandrasekhar, Rev. Mod. Phys., 15 (1943) 1. 156 D. P. M i l l a r and K. B. Eisenthal, J. Chem. Phys., 83 (1985) 5076. 157 S. P. Velsko and 6. R. Fleming, Chem. Phys., 65 (1982) 59. 158 S. H. Courtney and G. R. Fleming, J. Chem. Phys., 83 (1985) 215. 159 V. Sundstrom and T. G i l l b r o , Ber. Bunsenges. Phys. Chem., 89 (1985) 222. 160 M. Lee, A. J. Bain, P. J. McCarthy, C. H. Han, J. N. Haseltine, A. B. Smith, and R. M. Hochstrasser, J. Chem. Phys., 85 (1986) 4341.
158 161 S. H. Courtney, S. K. Kim, S. Canonica, and G. R. Fleming, J. Chem. SOC., Faraday Trans. 11, 82 (1986) 2065. 162 D. M. Z e g l i n s k i and D. H. Waldeck, J. Phys. Chem., 92 (1988) 692. 163 S. K. K i m and G. R. Fleming, J. Phys. Chem., 92 (1988) 2168. 164 6. R. Fleming, D. H. Waldeck, K. M. Keery, and S. P. Velsko, Isomerization Viewed as a Model f o r Activated B a r r i e r Crossing i n Solution, i n : K. B. Eisenthal (Ed.), A p p l i c a t i o n s o f Picosecond Spectroscopy t o Chemistry, Reidel, Dordecht, 1984, p. 67. 165 K. M. Keery and G. R. Fleming, Chem. Phys. Lett., 93 (1982) 322. 166 S. P. Velsko, D. H. Waldeck, and 6. R. Fleming, J. Chem. Phys., 78 (1983) 249. 167 V. Sundstrom and T. G i l l b r o , Chem. Phys. Lett., 109 (1984) 538. 168 Z. Smedarchina, Chem. Phys. Lett., 116 (1985) 538. 169 E. Akesson, H. Bergstrom, V. Sundstrom, and T. G i l l b r o , Chem. Phys, L e t t . , 126 (1986) 385. 170 J . M. Hicks, M. T. Vandersall. E. V. Sitzmann. and K. 6. Eisenthal, Chem. Phys. Lett., 135 (1987) 413. 171 S. H. Courtney and G. R. Fleming, Chem. Phys. Lett., 103 (1984) 443. 172 G. Maneke, J. Schroeder, J . Troe, and F. Voss, Ber. Bunsenges. Phys. Chem., 89 (1985) 896. 173 M. Lee, G. R. Holtom, and R. M. Hochstrasser, Chem. Phys. Lett., 118 (1985) 359. 174 G. R. Fleming, S. H. Courtney, M. W. Balk, J. Stat. Phys., 42 (1986) 83. 175 M. W. Balk and G. R. Fleming, J . Phys. Chem., 90 (1986) 3975. 176 J. McCaskill and R. G. G i l b e r t , Chem. Phys., 44 (1979) 389. 177 R. F. Grote and J. T. Hynes, J. Chem. Phys., 73 (1980) 2715. 178 Y.-P. Sun and J. S a l t i e l , unpublishedr e s u l t s . 179 J. T. Hynes, i n : M. Baer (Ed.), The Theory o f Chemical Reaction Dynamics, Chemical Rubber, Boca Raton, F l o r i d a (1985). 180 S. A. Adelman, Advances i n Chemical Physics, Wiley, New York, Vol. L I I I (1983) p. 61. 181 6. Carmeli and A. Nitzan, J. Chem. Phys., 79 (1983) 393. 182 R. F. Grote and J. T. Hynes, J. Chem. Phys., 74 (1981) 4465. 183 G. van der Zwan and J. T. Hynes, J. Chem. Phys., 77 (1982) 1295. 184 B. Bagchi and D. W. Oxtoby, J . Chem. Phys., 78 (1983) 2735. 185 A. 6. Zawadzki and J. T. Hynes, Chem. Phys. Lett., 113 (1985) 476. 186 H.-H. Schuh and H. Fischer, Helv. Chim. Acta, 61 (1978) 2130. 187 J . S a l t i e l and B. W. Atwater, Adv. Photochem., 14 (1988) 1, and references c i t e d . 188 P. S. Hubbard, Phys. Rev., 131 (1963) 1155. 189 H. Fischer and H. Paul, Acc. Chem. Res., 20 (1987) 200. 190 A. Spernol and K. Wirtz, Z. Naturforsch, A, 8 (1953) 522. 191 A. L. Van Geet and A. W. Adamson, Ind. Eng. Chem., 57 (1965) 62. 192 H. E r t l and F. A. L. Dullien, Amer. I n s t . Chem. Eng. J., 19 (1973) 1215. 193 M. Lehni, H. Schuh, and H. Fischer, I n t e r n a t i o n a l J . Chem. Kin., 11 (1979) 705. 194 A. Gierer and K. Wirtz, Z. Naturforsch, A, 8 (1953) 532. 195 J. L. Dote, D. Kivelson, and R. N. Schwartz, J. Phys. Chem., 85 (1981) 2169. 196 J . Lee, S.-B. Zhu, and 6. W. Robinson, J. Phys. Chem., 91 (1987) 4273. 197 6. Maneke, J. Schroeder, J. Troe, and F. Voss, Springer Proc. Phys., 4 (1985) 156. 198 S. H. Courtney, G. R. Fleming, L. R. Khundkar, and A. H. Zewail, J . Chem. Phys., 80 (1984) 4559. 199 J. Schroeder and J. Troe, Chem. Phys. Lett., 116 (1985) 453. 200 J. Troe, J. Phys. Chem., 90 (1986) 357. 201 G. Fischer, G. Seger, K. A. Muszkat, and E. Fischer, J. Chem. SOC. Perkin I1 (1975) 1569. 202 H. Petek, Y. Fujiwara, D. Kim, and K. Yoshihara, J . Am. Chem. SOC., 110 (1988) 6269.
159 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241
6. Hohlneicher, M. Muller, M. Demmer, J. Lex, J. H. Pen, L. Gan, and P. 0. Loesel, J. Am. Chem. SOC., 110 (1988) 4483. M. Traetteberg and W. B. Frantsen, J. Mol. Struct., 26 (1975) 69. W. Rapp, Chem. Phys. Lett., 27 (1974) 187. B. Baghi, Chem. Phys. L e t t . , 115 (1985) 209. K. B. Eisenthal and E. V. Sitzmann, p r i v a t e communicationo f unpublished results. S. K. Kim, S. H. Courtney and 6. R. Fleming, submitted f o r p u b l i c a t i o n ; S. K. Kim, p r i v a t e communication. Y.-P. Sun and J . S a l t i e l , unpublishedr e s u l t s . N. Sivakumar, E. A. Hoburg, and D. H. Waldeck, J. Chem. Phys., 90 (1989) 2305. B. S. Hudson, B. E. Kohler, and K. Schulten, Linear Polyene E l e c t r o n i c S t r u c t u r e and P o t e n t i a l Surfaces, i n : E. C. Lim (Ed.), Excited States, Academic Press, New York, 6 (1982) 1. M. T. A l l e n and D. G. Whitten, Chem. Rev., i n press. W. F r o l i c h , H. J. Dewey, H. Deger, B. Dick, K. A. Klingensmith, W. Puttmann, E. Vogel, G. Hohlneicher, and J. Michl, J. Am. Chem. SOC., 105 (1983) 6211. A. N. N i k i t i n a , N. A. Ponomareva, L. A. Yanovskaya, V. A. Dombrovskii, and V. F. Kucherov, Opt. Spectrosc. 40 (1976) 144. R. J. M. Anderson, G. R. Holtom, and W. M. McClain, J . Chem. Phys., 70 (1979) 4310. H. L.-B. Fang, R. J. Thrash, and 6. E. Leroi, J. Chem. Phys., 67 (1977) 3389. H. L.-B. Fang, R. J . Thrash, and 6. E. Leroi, Chem. Phys. L e t t . , 57 (1978) 59. L. A. Brey, G. 6. Schuster, and H. 6. Drickamer, J. Chem. Phys., 71 (1979) 2765. K. Schulten and M. Karplus, Chem. Phys. Lett., 14 (1972) 305. K. Schulten, I . Ohmine, and M. Karplus, J. Chem. Phys., 64 (1976) 4422. T. Ikeyama and T. Azumi, J . Chem. Phys., 76 (1982) 5672. W. H. Henneker, W. Siebrand, and M. Z. Zgierski, J . Chem. Phys., 79 (1983) 2495. R. Bensasson, J. Lafferty, E. J. Land, R. S. S i n c l a i r , and T. 6. Truscott, Chem. Phys. Lett., 41 ( (1976) 333. H. Gorner, J. Photochem., 19 (1982) 343. S. J. S t r i c k l e r and R. A. Berg, J. Chem. Phys., 37 (1962) 814. J. B. B i r k s and D. J. Dyson, Proc. Roy. SOC. London, Ser. A, 275 (1963) 135. E. D. Cehelnik, R. B. Cundall, J. R. Lockwood, and T. F. Palmer, Chem. Phys. L e t t . , 27 (1974) 586. E. D. Cehelnik, R. B. Cundall, J. R. Lockwood, and T. F. Palmer, J. Chem. Phys., 79 (1975) 1369. J. B. Birks, 6. N. R. T r i p a t h i , and M. D. Lumb, Chem. Phys., 33 (1978) 185. P. C. A l f o r d and T. F. Palmer, Chem. Phys. Lett., 86 (1982) 248. P. C. A l f o r d and T. F. Palmer, J. Chem. SOC. Faraday Trans. 11, 79 (1983) 433. 6. R. Jones and R. B. Cundall, Chem. Phys. L e t t . , 126 (1986) 129. T. I t o h and B. E. Kohler, J. Phys. Chem., 9 1 (1987) 1760. I. B. Berlman, J. Phys. Chem., 74 (1970) 3085. J. R. Andrews and B. S. Hudson, J. Chem. Phys., 68 (1978) 4587. L. A. Sklar, B. Hudson, M. Petersen, and J. Diamond, Biochem., 16 (1977) 813. K. Lunde and L. Zechmeister, J. Am. Chem. SOC., 76 (1954) 2308. F. B. M a l l o r y and C. W. Mallory, Drg. React., 30 (1984) 1. R. Mason and E. D. Cehelnik, J. Photochem., 9 (1978) 219. S. K. Chattopadhyay, P. K. Das, and 6. L. Hug, J. Am. Chem. SOC., 104 (1982) 4507. P. Bortolus and G. Galiazzo, J. Photochem., 2 (1974) 361.
160
242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281
G. Fischer and E. Fischer, Mol. Photochem., 6 (1974) 463. M. T. Allen, L. Miola, and 0. G. Whitten, J. Phys. Chem., 9 1 (1987) 6099. Y.-P. Sun and J. S a l t i e l , unpublished. T. C. Felder, K.-J. Choi, and M. R. Topp, Chem. Phys., 63 (1982) 175. R. A. Goldbeck, A. J. Twarowski, E. L. Russell, J. K. Rice, R. R. Birge, E. Switkes, and.D. S. K l i g e r , J. Chem. Phys., 77 (1982) 3319. S. K. Chattopadhyay and P. K. Das, Chem. Phys. Lett., 87 (1982) 145. C. R u l l i e r e and A. Declemy, Chem. Phys. L e t t . , 135 (1987) 213. Y.-P. Sun, D. F. Sears, Jr., and J. S a l t i e l , J. Am. Chem. SOC., 111 (1989) 706. B. E. Kohler and T. S p i g l i a n i n , J. Chem. Phys., 80 (1984) 5465. T. I t o h and B. E. Kohler, J. Phys. Chem., 92 (1988) 1807. P. Courtot and J . Salaun, Tetrahedron L e t t . , (1979) 1871. R. S. H. L i u and Y. Butt, J. Am. Chem. SOC., 93 (1971) 1532. Y. C. C. Butt, A. K. Singh, B. H. Baretz, and R. S. H. Liu, J. Phys. Chem., 85 (1981) 2091. J. S a l t i e l , L. Metts, and M. Wrighton, J. Am. Chem. SOC., 9 1 (1969) 5684. J. S a l t i e l , .L. Metts, and M. Wrighton, J. Am. Chem. SOC., 92 (1970) 3227. W. 6. Dauben, E. L. McInnis, and D. M. Michno, Photochemical Rearrangements i n Trienes, in: P. de Mayo (Ed.), Rearrangements i n Ground and Excited States, Vol. 3, Academic Press, New York, 1980, p. 91. P. J. Vroegop, J. Lugtenburg, and E. Havinga, Tetrahedron, 29 (1973) 1393. J. Meinwald and P. H. Mazzochi, J. Am. Chem. SOC., 88 (1966) 2850. R. Srinivasan, J. Am. Chem. SOC., 83 (1961) 2807. R. Srinivasan, J. Chem. Phys., 38 (1963) 1039. J. W. J. Gielen, H. J. C. Jacobs and E. Havinga, Tetrahedron L e t t . , (1976) 3751. A. M. Brouwer, J. Cornelisse, and H. J. C. Jacobs, Tetrahedron, 43 (1987) 435. A. M. Brouwer, Ph.D. Thesis, U n i v e r s i t y o f Leiden, Leiden, The Netherlands, 1987. D. J. S. B i r c h and R. E. Imhof, Chem. Phys. L e t t . , 88 (1982) 243. R. Wilbrandt, N.-H. Jensen, and F. W. Langkilde, Chem. Phys. Lett., 111 (1984) 123. A. Kasama, M. Taya, T. Kamisuki, Y. Adachi, and S. Maeda, i n : A. Laubereau and M. Stockburger (Eds.), Time-resolved V i b r a t i o n a l Spectroscopy, Springer, B e r l i n , 1985, p. 166. T. L. Gustafson, J. F. Palmer, and D. M. Roberts, Chem. Phys. Lett., 127 (1986) 505. B. M. Pierce and R. R. Birge, J. Phys. Chem., 86 (1982) 2651. R. L. Swofford and W. M. McClain, J. Chem. Phys., 59 (1973) 5740. H. L. Fang, T. L. Gustafson, and R. L. Swofford, J . Chem. Phys., 78 (1983) 1663. J . A. Bennett and R. R. Birge, J. Chem. Phys., 73 .(1980) 4234. L. A. Heimbrook, B. E. Kohler, and T. A. Spiglanin, Proc. N a t l . Acad. Sci , USA, 80 (1983) 4580. J. F. Shepanski, B. W. Keelan and A. H. Zewail, Chem. Phys. L e t t . , 103 (1983) 9. A. Amirav, M. Sonnenschein, and J. Jortner, Chem. Phys., 102 (1986) 305. W. A. Yee, S. J. Hug, and D. S. K l i g e r , J. Am. Chem. SOC., 110 (1988) 2164. J. H. Pinckard, B. Wille, and L. Zechmeister, J. Am. Chem. SOC., 70 (1948) 1938. L. R. Eastman, Jr., 8. M. Zarnegar, J. M. B u t l e r , and D. 6. Whitten, J. Am. Chem. SOC., 96 (1974) 2281. W. A. Yee, J. S. Horowitz, R. A. Goldbeck, C. M. Einterz, and D. S. K l i g e r , J. Phys. Chem., 87 (1983) 380. C. R u l l i e r e , A. Declemy, and P. K o t t i s , Laser Chem., 5 (1985) 185. K. Ogawa, H. Suzuki, and M. Futakami, J . Chem. SOC. Faraday Trans. 11, 84
.
161
282 J. Troe, A. Amirav, and J. Jortner, Chem. Phys. Lett., 115 (1985) 245. 283 M. E. S q u i l l a c o t e and T. C. Semple, J. Am. Chem. SOC., 109 (1987) 892. 284 A. A. Gorman, I . R. Gould, and I . Hamblett, J. Am. Chem. SOC., 103 (1981) 4553. 285 R. A. Caldwell and M. Singh, J. Am. Chem. SOC., 104 (1982) 6121. 286 P. Vanderlinden and S. Bou6, J. C. S. Chem. Comm., (1975) 932. 287 S. Bout! and R. Srinivasan, J. Am. Chem. SOC., 92 (1970) 3226. 288 R. Srinivasan, J. Am. Chem. SOC., 90 (1968) 4498. 289 J. S a l t i e l , D. E. Townsend, and A. Sykes, J. Am. Chem. SoC., 95 ( 973) 5968. 290 J. S a l t i e l , A. D. Rousseau, and A. Sykes, J. Am. Chem. SOC., 94 ( 972) 5903. 291 J. S a l t i e l . L. Metts. A. Svkes. . , , and M. Wriahton, J. Am. Chem. SOC , 93 (1971) 5302. 292 Y. B. Scheck, N. P. Kovalenko, M. V. Alfimov, J. Lumin., 15 (1977) 157. 293 E. Fischer, J. Photochem., 17 (1981) 331. 294 G. Fisher and E. Fischer, J. Phys. Chem., 85 (1981) . 2611. 295 U. Mazzucato, Pure Appl .'Chem. ,-54 (1982) 1705. 296 Z. Shakked, unpublishedr e s u l t s quoted i n r e f s . 293, 297, 298. 297 K. A. Muszkat and T. Wismontski-Knittel, Chem. Phys. L e t t . , 83 (1981) 87. 298 K. A. Muszkat and T. Wismontski-Knittel, J. Phys. Chem., 85 (1981) 3427. 299 E. Haas, 6. Fischer, and E. Fischer, J. Phys. Chem., 82 (1978) 1638. 300 J. B. Birks, G. Bartocci, G. G. A l o i s i ; S. Dellonte, and F. B a r i g e l l e t t i , Chem. Phys., 51 (1980) 113. 301 F. Masetti, G. Bartocci, U. Mazzucato and E. Fischer, J. Chem. SOc., Perkin Trans. 11, (1983) 797. 302 W. H. Lawton and E. A. Sylvestre, Technometrics, 13 (1971) 617. 303 J. S a l t i e l and D. W. Eaker, 3. Am. Chem. SOC., 106 (1984) 7624. 304 6. Bartocci, U. Mazzucato, F. Masetti, G. G. A l o i s i , Chem. Phys., 101 (1986) 461. 305 J. S a l t i e l , D. F. Sears, Jr., J.-0. Choi, Y.-P. Sun, F. B. Mallory, and
-
C. W. Mallory, unpublishedr e s u l t s .
306 G. Bartocci, F. Masetti, U. Mazzucato, and G. Marconi,. J. Chem. SOc., Faraday Trans. 11, 80 (1984) 1093. 307 J.-0. Choi, Masters Thesis, F l o r i d a State U n i v e r s i t y , Tallahassee, FL 32306. 308 K. P. Ghiggino, P. F. S k i l t o n , and E. Fischer, J. Am. Chem. SOC., 108 (1986) 1146. 309 M. Sumitani, S. Nagakura, and K. Yoshihara, Chem. Phys. L e t t . , 29 (1974) 410. 310 J. S a l t i e l and D. W. Eaker, Chem. Phys. Lett., 75 (1980) 209. 311 S. A. Krysanov and M. V. Alfimov, Chem. Phys. L e t t . , 98 (1983) 176. 312 H. Garner, D. W. Eaker, and J. S a l t i e l , J, Am. Chem. S O C . , 103 (1981) 7164. 313 T. Wismontski-Knittel and P. K. Das, J. Phys. Chem., 88 (1984) 2803. 314 E. Fischer and N. Castel, J. Mol. Struct., 145 (1986) 367. 315 G. S. Hammond, S. C. Shim, and S. P. Van, Mol. Photochem., 1 (1969) 89. 316 M. Kaganowich, G. Fischer, E. Fischer, Ch. Goedicke, and H. Stegemeyer, Z. Phys. Chem. N. Folge, 76 (1971) 79. 317 N. P. Kovalenko, M. V. Alfimov, A. Abdukadirov, and Yu. B. Shekk, B u l l . Acad. Sci. USSR Ser. Chem. (Eng. Translation) (1979) 1164. 318 G. 6. A l o i s i , U. Mazzucato. J . B. Birks, and L. M i n u t i , J. Am. Chem. SOC., 99 (1977) 6340. 319 T. Wismontski-Knittel, G. Fischer, and E. Fischer, J. Chem. SOC., Perkin Trans. 11, (1974) 1930. 320 N. Castel-and E.'Fischer, J. Mol. Struct., 127 (1985) 159. 321 M. F. Alfimov, V. F. Razumov, A. G. Rachinksy, V. N. Listvan, and Yu. B. Shekk, Chem. Phys L e t t . , 101 (1983) 593. 322 A. C. Matthews, R. Sakurovs, and K. P. Ghiggino, J. Photochem., 19 (1982) 235.
162 323 K. P. Ghiggino, J. Photochem., 12 (1980) 173. 324 J. S a l t i e l , D. F. Sears, Jr., F. B. Mallory, C. W. Mallory, and C. A. Buser, J . Am. Chem. SOC., 108 (1986) 1688. 325 Y.-P. Sun, D. F. Sears, Jr., and J . S a l t i e l , Anal. Chem., 59 (1987) 2515. 326 Y.-P. Sun, 0. F. Sears, Jr., J. S a l t i e l , F. B. Mallory, C. W. Mallory, and C. A. Buser, J . Am. Chem. SOC., 110 (1988) 6974. 327 A. S. Cherkasov, Doklady Acad. Sci. USSR, 146 (1962) 852. 328 A. M. Brearley, A. J. 6. Strandjord, S. R. Flom, and P. F. Barbara, Chem. Phys. Lett., 113 (1985) 43. 329 S. R. Flom, V. Nagarajan, and P. F. Barbara, J. Phys. Chem., 90 (1986) 2085. 330 A. M. Brearley, S. R. Flom, V. Nagarajan, and P. F. Barbara, J. Phys. Chem., 90 (1986) 2092. 331 Y.-P. Sun, D. F. Sears, Jr., 3. S a l t i e l , F. 8. Mallory, C. W. Mallory, and C. A. Buser, unpublished results.
163
Addendum to chapter 3 PHOTOCHROMISM BASED ON CIS-TRANS ISOMERIZATIONlINDIGO DERIVATIVES (List of references) Here after is an updated list of references on Indigo Derivatives following J.Saltiel's chapter as they are photochromic systems based on cis-trans isomerization of a double bond. C.R. Giuliano; L.D. Hess; J.D. Margerum, Cis-trans isomerization and pulsed laser studies of substituted indigo dyes, J.Am.Chem. SOC. w ( 9 ) (1968) 587. D.L. Ross, Photochromic indigoids. 111. Photochromic element based on the cis-trans photoisomerization of a thioindigo dye, Appl.Opt., u ( 3 ) (1971) 571. Z.V. Zhidkova, Photochromic properties of films dyed with perinaphthothioindigo and 2-perinaphthopenthiophene-2'-(5'-methylthionaphthene)-indigo, Zh.Prikl.Spektrosk., l3(2) (1972) 325, Chem.Abstr. 71 (1972) 12267r. M.A. Mostoslavskii; L . F . Gorbas; E.B. Goergieva, Correction of: earlier abstract Synthesis, spectral, and photochromic properties Of 7,7'-disulfoperinaphthioindigo, Khim.Geterotsikl.Soedin., (51, (1973) 637. Z.V. Zhidkova, Photochromic properties of solutions and films, colored by the dyes 2-perinaphthpenthiophene-2'-thionaphtheneindig0 and 2-perinaphthpenthiophene-2'-(ethoxythionaphthene)indigo, Zh.Prikl.Spektrosk., a ( 5 ) (1975) 861, Chem.Abstr. 83 ( 1 9 7 6 ) 81174~. Z.V. Zhidkova, Photochromic properties of solutions and films colored by 2-perinaphthopenthiophene-2'-(5',6'-benzo-7'-chlorothionaphthene)indigo and 2-perinaphthopenthiophene-2'-(4',5'benzothionaphthene)indigo, Zh.Prikl.Spektrosk., 25(2) (1976) 315 Chem-Abstr. 86 (1977) 197888a
164
T. Kobayashi; P.M. Rentzepis, O n t h e picosecond k i n e t i c s and photostability of indigo and 6,6’-dimethoxyindigo, J.Chem.Phys., 70(2) (1979) 886. M. Irie and M. Kato, Photoresponsive Molecular tweezers. Photoregulated ion capture and release u s i n g t h i o i n d i g o d e r i v a t i v e s having ethylenedioxy side groups. J.Am.Chem.Soc. 107 (1985) 1024. M.A. Mostoslavskii; V.G. Nazar’ko, Nitrobenzene prevention of the catalytic effect of trace impurities produced during photoisomerization of benzene solutions of’thioindigo, Zh.Prikl.Khim. (Leningrad) 59(6) (1986) 1374. Chem.Abstr. 105 (1986) 105576g A.V. Veniaminov; G.I. Lashkov, Effect of polymer medium on photochemical isomerization of thioindigo compounds in poly(methy1 methacrylate), Vysokomol.Soedin., Ser.A g ( 4 ) (1986) 861, Chem. Abstr. 105 (1986) 1 0 5 5 7 ~ . J.G. Victor; J.M. Torkelson, On measuring the distribution of local free volume in glassy polymers by photochromic and fluorescence techniques, Macromolecules, a ( 9 ) (1987) 2241.
165
Chapter 4
Azo Compounds H. Rau
1. INTRODUCTION The azo group -N=N- has given its name to a large group of compounds. There are the families of organic (ref. l), inorganic (ref. 2 ) and organometallic (ref. 3 ) azo compounds. In this chapter we concentrate on the organic family which has several branches depending on the res2dues attached to the azo group: aliphatic, aromatic and mixed azo compounds and we confine ourselves to the aromatic species which absorb visible light and are therefore colored materials. They meet the requirement of photochromism in the visible part of the electromagnetic spectrum. The azo group is isosteric with the ethylene group. Two stable configurations thus can be anticipated (Fig. 1) which have been called trans and cis. Indeed, in 1937 Hartley isolated cis azobenzene and characterized it (ref. 4), and in 1964 Hutton and Steel prepared the first aliphatic cis azo compound (ref. 5).
Fig. 1.
Isomerization of azobenzene
&
The nomenclatureof azo compounds has been adjusted to the needs of modern requirements. The historic term of azo compounds is still generally understood, but it has been replaced by the systematic one of "diazene". Azobenzene is listed in Chemical Abstracts as "diazene, diphenyl-". The cis compounds are now called "2" and the trans."E" forms. There are two reactions of azo compounds that meet the
166
requirements of the definition of photochromism. For azobenzene and most substituted azobenzenes the E and Z forms can be isomerized reversibly by light and heat (Fig. 11, and ortho- and parahydroxy or -amino substituted azo aromatics may show photochromism by intramolecular proton transfer. The latter reaction is dealt with in Chapters 16 and 1 7 . Molecular photochromism requires a molecular transformation. This is a reaction which lives in competition with other possible photoreactions. It is therefore necessary to consider (i) the spectroscopy which is an indicator of photochromism and gives information about the reacting state and its generation, and (ii) the mechanism of molecular transformation which may give clues to possible branching points in the photochromic reaction. There are several reviews of the photochromic or isomerization properties of azobenzenes. Wyman has covered the literature up to 1954 (ref. 6 1 , Ross and Blanc up to 1969 (ref. 7 ) . The spectroscopy was reviewed in 1973 by Rau (ref. 8 ) , and the chemistry and characteristics of azocompounds were treated extensively by Zollinger (ref. 1l.A survey ofthe different photoreactions of azobenzenes including isomerization, photocyclization, photoreduction and fading of azo dyes was given by Griffiths (ref. 9). This article starts from the basis of the treatise of R o s s and Blanc. From the newer work three points are considered most important: (i) the differentiation between azobenzene type, aminoazobenzene type and pseudo stilbene type molecules on the basis of the relative order of (n,n*)- and (n,n*)-states, (ii) the mechanistic aspects of isomerization and (iii) the use of azobenzene photoisomerisation for photocontrol of material properties. 2. SPECTRAL FEATURES 2.1. Compounds of the azobenzene type Azo compounds of the azobenzene type generally are yellow to red. The main feature of their absorption spectra i s a relatively weak long wavelength band well separated from the shorter wavelength band systems (refs. 10-15). Thus the azobenzene type is characterized by a low-lying '(,,,*)-state and a large energy gap Fig. 2. shows between this and the next higher '(n,n*)-state. typical spectra. The low energy n + n * band near 22 000 cm-l (440 nm) is without vibrational structure. Indeed, there was never observed any structure on the n + TI*band of any noncyclic azo compound.
167
30 000
20 000
40 000
b
OD
30000
20 000
40 000 ij/C*’’
Fig. 2.
( a ) A b s o r p t i o n s p e c t r a o f E-azobenzene
2-azobenzene
-------
s p e c t r a o f E-& ( r e f . 8).
a t 2 9 8 K i n EtOH ( r e f . 1 2 ) . and 2 -------, 2-4 -
-.-.-.-
and ( b ) Absorption a t 7 7 K i n EtOH
b a n d s o f c a r b o n y l compounds, t h i s azo band -1 , f o r Zis r a t h e r i n t e n s i v e : f o r E-azobenzene E~~~ = 405 M-lcrn azobenzene E = 1250 M-lcm-’ ( r e f . l o ) . The d i f f e r e n c e i n i n t e n 440 s i t y i s d u e t o s e l e c t i o n r u l e s : i n t h e p l a n a r C Z h a n d CaV symmetr i e s t h e n + 1 ~ * t r a n s i t i o n i s f o r b i d d e n f o r t h e E, b u t a l l o w e d f o r
Compared t o t h e n
-+
TT*
168
the Z form. Under this aspect the intensity of the band in the spectrum of the E form is very high. This has been attributed to non-planar distortions of the molecule and to vibrational coupling (ref. 16). The absorption spectra of the shorter wavelength band systems are very similar to those of stilbenes, but the bands are red shifted by some 2000 cm-l. In the spectra of the E-azobenzene these bands are nearly without vibrational structure at room temperature (Fig. la). But they are clearlystructured, though less than those of stilbene, in a rigid solvent matrix at 77 K (Fig. 2b) or when azobenzene is a guest in a dibenzyl crystal at 20 K (ref. 17). In the spectra of 2-azobenzenes the IT + n* bands generally do not show vibrational features (Fig. 2b). However, the influence of molecular structure on the n .+ n* bands is demonstrated by the - is more planar spectra of o,o'-azodiphenylmethane (Fig. 2b): 2 (red shift) and more rigid (vibrational structure) than Z-azobenzene. The spectra are rather insensitive to solvent polarity: E-azobenzene has a n + n* maximum at 442 nm in hydrocarbon and at 437 nm in ethanol solution, E-4,4'-dimethylazobenzene at 443 nm and 438 nm, respectively (ref. 15). 0- and p-Alkyl, halogen or similar substituents shift both the n + TI*and II .+ P* bands (refs. 14,15,18). 2,2',4,4',6,6'-hexaphenylazobenzene 1 0 is blue and shows the largest shift of the n .+ TI*band to 520 nm (ref. 18). However, oand p-amino substituents have a much more severe effect on the TI .+ TI*bands, shifting these to longer wavelengths than the n + n* band. This class of azo compounds will be treated in sections 2.2. and 2.3.
169
Noncyclic azo compounds are generally non-emitting. There are very few reports on n + n* emission. Struve detected an extremely weak short lived ( 2 5 ps) emission of azobenzene (ref. 19) and later even an emission from the second excited state which is of v,n* character (ref. 2 0 ) . Bisle and Rau reported low intensity fluorescence in frozen solutions of highly hindered azobenzene derivatives (ref. 2 1 ) . Shinomura and Kinotake found fluorescence in azobenzene-containing bilayer membranes which can be quenched by a cyanin Neither phosphorescence nor a short lived transient dye (ref. 2 2 ) . have ever been detected. The triplet states of azo compounds have only been located by triplet quenching experiments (refs. 23-26). There are several semiempirical calculations of the electronic states of azobenzene. The calculations generally are in agreement with the experimental data (refs. 2 7 - 3 3 ) . Monti, Orlandi and Palmieri have obtained theoretical data on the spectroscopy of azobenzene by an ab initio calculation (ref. 34). This short review of selected spectroscopic data suggests intensive communication of the electronic states in azobenzene and related compounds. The electronic states may be coupled by molecular vibrations, but one important molecular feature seems to be a certain floppiness of the molecular structure allowing large scale molecular distortions up to the limit of reactive deactivation by isomerization. 2.2. Compounds of the aminoazobenzene type Substitution of azobenzene by 0- or p-amino groups changes the spectra dramatically. The n IT* band is shifted by several thousand cm-l to longer wavelengths. Thus the compounds of the aminoazobenzene type are characterized by close energetic proximity of the (n,n*)- and (~,a*)-states. Fig. 3 shows'a typical spectrum. The absorption bands have less vibrational structure than those of the molecules of the azobenzene type. They are relatively sensitive to solvent polarity (ref. 15). Gerson and Heilbronner showed that the azo group is more ready to interact with hydrogenbonding or protonating solvents than the amino group (ref. 3 5 1 , probably with reordering of the states on the energy scale. This type of compounds may appear as a true azo compound in hydrocarbon, but a pseudo stilbene (vide infra) in alcohol and acidic solution. Azo compounds of the aminoazobenzene type may show fluorescence. Bisle et al. (ref. 15) and Rau (ref. 3 6 ) report on the low temperature emission properties in glassy solvents. From the Position of the fluorescence band on the red side of the n + T * -f
170
E
3
E *.lo-&
A
10-4
2
I
1
300
LO 0
500 nm
Fig. 3. Absorption spectra of 4-dimethylamino-azobenzene at 298 K in n-hexane and EtOH
-------.
absorption they assigned the emission to n + TI* fluorescence. In ethanol the emission is blue shifted and a mirror image of the red shifted n + IT*absorption band. Although there is no phosphorescence it has been possible to locate the lowest triplet state by energy transfer experiments. Monti et al. (refs. 26,37) found the triplet states of 4-(C2H5j2Nazobenzene, 4-(C2H5)2N-4'-OCH3-azobenzene and 4-N02-azobenzene near 140 kJ mol-l. They conclude that the triplet molecules are distorted in comparison to the E configuration which lowers their energy by about 48 kJ mol-l, and they assign the lowest triplet state to the (n, **)-state contrary to what would be expected from the usually much higher singlet-triplet split for (r,r*)-states. Their assignment is, however, supported by the fact that no triplettriplet absorption could be detected for these molecules in contrast to the case of the pseudo stilbene 4-(C2H5)2N-4'-N02-a~obenzene (ref. 38). 2.3. Pseudo stilbenes Pseudo stilbenes are characterized by a low-lying '(TI,IT*>state. The sequence of the electronic states of azobenzenes can be rearranged to that of pseudo stilbenes by either raising the energy of the (n,n*)-state or by lowering the energy of the (*,IT*>state. The former is achieved by protonation of azobenzenes (refs. 39,401 - Haselbach and Heilbronner have shown that protonation is asymmetric at one N-atom (refs. 41,42) - or complexing the nelectrons (refs. 43,44), the latter by substitution of azobenzenes with electron donating and accepting groups. This kind of substitution as in 4-(CH3)2N-4'-N02-azo-
171
benzene 4 shifts the R + TI* bands to still longer wavelengths than in aminoazobenzenes (Fig. 4) due to increased charge transfer character of the transition. The weak n + R * band is buried in the n + R * absorption. Little vibrational structure is observed in the spectra at room temperature. It is not surprising that the polarity of the solvent influences the band position in the spectrum of the donorlacceptor substituted compounds. Most of the commercial azo dyes are pseudo stilbenes rather than azo compounds.
300
LOO
500
600 nm
Fig. 4. Absorption spectrum of 4-dimethylamino-4'-nitro-azobenzene at 2 9 8 K in n-hexane. Some pseudo stilbenes emit fluorescence. Protonated E-azobenzene, though non-fluorescent at room temperature, fluoresces strongly in rigid sulfuric acid at 7 7 K as do most other protonated Donorlacceptor Subazo compounds and pseudo stilbenes (ref. 4 5 ) . stituted azobenzenes emit weakly at low temperatures (ref. 15). $ has been shown to fluoresce when adsorbed to cellulose, polyamide or NaCl at room temperature (ref. 36), and it has been known for a long time that azo dyes adsorbed to fibers may fluoresce (ref. 46). Still, the emission of donorlacceptor substituted azobenzenes, contrary to that of stilbenes, is weak even at low temperature in rigid solvents. Phosphorescence of pseudo stilbenes is generally not detected. However, the triplet states can in some cases be observed by transient absorption spectroscopy. Gorner, Gruen and SchulteFrohlinde (ref. 3 8 ) have found that the energy of the triplet state of 4-nitro-4'-diethylaminoazobenzene is > 165 kJ mol-' and they have given detailed information about the population and the decay of the triplet states as a function of temperature (and concomitantly of solvent viscosity). Monti et al. (ref. 37) investigated
172 the same molecule in triplet quenching experiments and assigned the lowest triplet state to be of (n,r*) character. They located this state at 140 kJ mol" and found a small distortion energy of about 10 kJ mol-l. 3. ISOMERIZATION 3.1. Compounds of the azobenzene type The isomerization of azobenzene according to Fig.1. can be E) and by heat (Z E). effected by means of light (E -+ Z, Z The E form of azobenzenes is thermodynamically more stable For azobenzene Schulze et al. than the Z form (refs. 47-49). found heats of formation of AH&8 (E) = 311 kJ mol-l, (Z) = 367 kJ mol" (ref. 4 8 ) , thus the energy difference is 56 kJ mol-l. The Z form, however, is stabilized kinetically by an activation energy Ea of isomerization. Talaty and Fargo (ref. 50) determined Ea-values between 85 and 100 kJ m 1 - l (in agreement with other authors) and pre-exponential factors between 10l2 and 3.10 13 .-1 -+
-+
for a series of substituted ground state azobenzenes in solution. Ortruba and Weiss reported the same values in liquid crystals (ref. 51). The activation energy is a bit higher in the melt (ref. 471, 112 kJ mol-' in the vapor phase (ref. 52) but nearly twice as high in the solid : 223 kJ mol-' (ref. 47). From the linear plot of AH* vs. AS* data from all three phases Andersson et al. (ref. 52) conclude that the mechanism of thermal isomerization is the same in all cases (inversion, vide infra). The thermal isomerization can be catalyzed by iodine (refs. 53-55) and electron donors and acceptors (ref. 55). The photochromism of azo compounds appears as a deepening of the color of thermally equilibrated samples on irradiation at a wavelength which is absorbed more strongly by the E form, e.g. at 366 nm in the case of azobenzene. Restoration of the initial color is reached by heating or by selection of an irradiation wavelength preferentially absorbed by the Z form, e.g. 436 nm. The photoisomerization of azobenzene is one of the cleanest photoreactions known. Its deaereated hydrocarbon solution can be irradiated with visible or near UV radiation for days without a change of absorption (ref. 56). When oxygen is not excluded a very slow photooxidation to give predominantly azoxybenzene is the only side reaction (ref. 56). But the presence of oxygen does not influence the measured values of isomerization quantum yields. Photoreactions can be analyzed by means of the powerful diagnos-
173 TABLE 1
I s o m e r i z a t i o n quantum y i e l d s f o r
0.20
0.11 0.09 0.10
0.10
0.10 0.11 0.12
0.42 0.4
0.42
0.44
0.44 0.40
0.27
'TI
-
0.36 0.27 0.25 0.28
0.45 0.75 0.4
0.20
0.68 0.55 0.56
0.20
0.25 0.24
0.55 0.68
>'TI*
and n ->m*
Aa HCb
HC HC HC HC HC HC
HC
excitation
135 59 65 23
136 137 63
67 62
a A l c o h o l (EtOH). bHydrocarbon ( u s u a l l y n-hexane).
t i c s d e v e l o p e d by Alauser ( r e f s . 57,58). M a u s e r ' s A-diagrams f o r t h e a z o b e n z e n e p h o t o i s o m e r i z a t i o n are p e r f e c t l y l i n e a r i n d i c a t i n g only one independent r e a c t i o n i n t h e i r r a d i a t i o n system. T h i s j u s t i f i e s t h e u s e of a z o b e n z e n e i s o m e r i z a t i o n as a c o n v e n i e n t a c t i n o m e t e r ( c f . t h e c h a p t e r i n t h i s volume by G a u g l i t z ) . The k i n e t i c s of t h e p h o t o i s o m e r i z a t i o n c a n b e e v a l u a t e d t o g i v e v a l u e s f o r t h e quantum y i e l d s Q~ f o r t h e E -> Z and $z f o r t h e Z -> E r e a c t i o n s . Zimmerman e t a l . ( r e f . 59) h a v e made u s e o f t h e f a c t t h a t t h e p h o t o s t a t i o n a r y s t a t e is a p p r o a c h e d a s a f i r s t order reaction. T h i s allows t h e d e 6 e r m i n a t i o n o f t h e quantum y i e l d s e v e n when t h e i r r a d i a t i o n is s t a r t e d w i t h s a m p l e s t h a t are n o t p u r e Z o r E. The e v a l u a t i o n o f t h e r e a c t i o n , a n d e s p e c i a l l y t h e v a l u e s of 'bZ, a r e v e r y s e n s i t i v e t o e r r o r s i n t h e € - v a l u e s . T h i s may be t h e r e a s o n f o r t h e v a r i a t i o n s i n t h e numbers f o r t h e quantum y i e l d s d e t e r m i n e d b y d i f f e r e n t a u t h o r s ( T a b l e 1). A method w h i c h r e q u i r e s o n l y t h e s h a p e o f t h e s p e c t r u m b u t n o t t h e € - v a l u e s ( r e f s . 60,61) h e l p s i n t h i s respect. I t is e v i d e n t f r o m T a b l e 1 t h a t t h e y i e l d s $E a n d 9, d o n o t a d d up t o u n i t y . However, t h e most s t a r t l i n g i n f o r m a t i o n i n T a b l e 1 and i n F i g u r e 5 is t h e f a c t t h a t $E is l a r g e r when t h e l o w -
Q, 0.L
-
+
+
+
+
+
OZ
+
+
0.2 -
1
1
I
I
I
Fig. 5. Dependence of the isomerization yields of azobenzene 1 on the wavelength of irradiation + ref. 59, o ref. 62. lying l(n,n*)-state is excited than when the higher (n,n*)-states are populated. This is a violation of Kasha's rule of total deactivation of higher to the lowest excited state. The dependence Of $E on excitation wavelength has been investigated by Zimmerman et al. (ref. 59) and by Bihler and Rau (ref. 62): excitation within the spectral region of the n ->n* and of the n ->n* bands gives constant $E values of 0.24 and 0.12, respectively (Fig. 5). A similar appearance of the $Z-values is stated by Siampiringue et al. (ref. 63): $z is 0.55 for n ->n* and 0.40 for n->n* excitation.
0
-40 -80 -120 -160%
Fip. 6 . Dependence of E+Z quantum yields on temperature and excitation wavelength for (a) azobenzene, (b) 2,2'-azonaphthalene 3 (after ref. 64). -
175 The temperature dependence of the quantum yields of azobenzene (Fig. 6) and the azonaphthalenes has been determined by Fischer and his group (refs. 64-66). The $E values for azobenzene on n ->a* and T - > T * excitation decrease nearly linearly with decreasing temperature but they are significantly different at all temperatures. For the azonaphthalenes these $ E values approach one another at low temperatures. The Z -> E yield, however, is high (>lo%). The Z -> E isomerization cannot be frozen out nor stopped by high viscosity. Bortolus and Monti (ref. 67) find a pronounced increase of the E -> Z yields but a decrease of the Z -> E yields with increasing solvent polarity regardless which state of azobenzene is populated by irradiation (Table 2 ) . Mono- or di-substitution does not lead to considerable changes in the yields (ref. 65).
V When,however, substitution causes steric hindrance leading to reduction or elimination of the floppiness of the molecule a new feature appears. Luddecke and Rau (ref. 681, Rau (ref. 61) and Rau and Shen (ref. 18) have investipated the compounds 5 to &Q and found.the quantum yields of Table 3 . Under these special conditions the E -> 2 yields are equal for excitation of the (n,n*)- and (a,a*)-states. This has implications for the mechanism of isomerization. (ref. 81, A. (ref. 691, 42 (ref. There are a few azobenzenes 70)) that are “locked” in. the 2 form and cannot isomerize. For m 6 in 12 the most relaxed structure is reached when the phenyl rings are perpendicular to the plane of the azo group and facing one another, as is concluded from ring current effects in the ‘H-NMR spectra (ref. 70). This structural feature can also be recognized in the X-ray structure of Z-2,2‘,6,G‘-tetra-iso-
(a
176 N
-4z--s--7
N=N
TABLE 2
S o l v e n t d e p e n d e n c e o f p h o t o i s o m e r i z a t i o n y i e l d s o f a z o b e n z e n e (67).
+E
Solvent
n ->
n- h e x a n e ethyl iodide e t h y l bromide ethanol a c e t o n itr i l e
HZO/EtOH 80/20v:v
+Z II
TI*
0.25
->
TI*
0.11
0.24 0.26
0.11
0.28
0.15
0.31 0.35
0.15 0.21
n ->v*
n ->n*
0.56 0.69
0.27
0.51
0.24
0.58
0.25
0.46 0.41
0.21 0.15
TABLE 3
P h o t o i s o m e r i z a t i o n quantum y i e l d s of non r o t a t i n g a z o b e n z e n e s . +E
ComDound ~
5
I
8-
-9
as
~~
n -> ~~~
~
0.24 0.29 0.25 0.19
0.05
+Z
n*
n
-> v*
~~
n ->v*
n->*
Ref.
0.53
67 60
0.6
18
~
0.24 0.29
0.56
0.04
0.4
0.19 0.18
0.5 0.55
0.6 0.55
18 18
p r o p y l a z o b e n z e n e Z-&Q (71) ( P i g . 7) a n d t h e most r e l a x e d D r e i d i n g -
model of 2-g. L i t t l e i s known a b o u t t h e p h o t o c h e m i s t r y o f t h e s e l o c k e d Z-azo compounds; t h e t o t a l d e s t r u c t i o n of t h e azo g r o u p of t h e m o l e c u l e s 12 w i t h m < 6, where t h e s p e c i f i c s p e c t r u m of t h e a z o b e n z e n e u n i t d i s a p p e a r s on i r r a d i a t i o n ( r e f . 701, h i n t s a t a n a l t e r n a t i v e h i t h e r t o unknown r e a c t i o n c h a n n e l f o r Z compounds.
177 C12h
Fig. 7. ORTEP d i a g r a m o f Z-2,2',6,6'-tetraisopropylazobenzene ( r e f . 71). However, i f i s o m e r i z a t i o n is p o s s i b l e t h i s c h a n n e l or o t h e r s are t o t a l l y unimportant.
I s o m e r i z a t i o n o f azo compounds is n o t o n l y i n d u c e d b y d i r e c t
e x c i t a t i o n b u t a l s o by t r i p l e t s e n s i t i z a t i o n . T h i s was first d e m o n s t r a t e d f o r a z o b e n z e n e b y J o n e s a n d Hamrnond ( r e f . 7 2 ) who
d e t e r m i n e d a s t a t i o n a r y c o n c e n t r a t i o n o f a b o u t 2% o f Z f o r m . F i s c h e r ( r e f . 7 3 ) f o u n d 25%. T h i s h a s i n i t i a t e d f u r t h e r i n v e s t i g a t i o n s by L e m a i r e a n d c o w o r k e r s ( r e f s . 2 3 , 2 4 ) who f o u n d t h a t t w o t r i p l e t s t a t e s Ta a n d T B of t h e Z ( a t a b o u t 1 9 5 kJ mol-'
and 180 k J m o l - l ) a n d E f o r m s ( a t a b o u t 192 k J m o l e ' a n d b e t w e e n 130 a n d 180 k J m o l - l ) are i n v o l v e d i n t h e i s o m e r i z a t i o n . P r a p s t u s e d t r i p l e t s e n s i t i z e r s p r o d u c e d i n t h e r e c o m b i n a t i o n o f electroc h e m i c a l l y created r a d i c a l s a n d f o u n d 1 7 0 + 1 0 k J m o l - ' (ref.25). Monti e t a l . ( r e f . 2 6 ) r e a l i z e d t h a t t r i p l e t e n e r g y t r a n s f e r t o E a n d Z a z o b e n z e n e o n l y s l o w l y d e v i a t e s from t h e d i f f u s i o n l i m i t when ET o f t h e s e n s i t i z e r s i s l o w e r e d . They c o n c l u d e t h a t t h e a c c e p t o r a z o b e n z e n e may be d i s t o r t e d a n d t h u s b e h a v e "nonc l a s s i c a l l y " . They c a l c u l a t e a d i s t o r t i o n e n e r g y o f 4700 cm-' for t h e E f o r m a n d 6400 cm-' f o r t h e Z form. I n a n o t h e r e x p e r i m e n t B o r t o l u s a n d Monti ( r e f . 6 7 ) f o u n d t h a t t r i p l e t s e n s i t i z e r s w i t h ET > 190 kJ mol" t r a n s f e r t h e energy i n e v e r y e n c o u n t e r ( d i f f u s i o n l i m i t e d ) t o b o t h E- and Z-azobenzene, b u t o n l y few E m o l e c u l e s isomerize: i n a g r e e m e n t w i t h J o n e s a n d Hammond ( r e f . 7 2 ) t h e y e s t a b l i s h a v e r y l o w $E o f 0 . 0 1 5 f o r t h e i s o m e r i z i n g s t a t e , b u t $z = 1 . 0 . I n c o n t r a s t t o s t i l b e n e ( r e f . 74) n o a z u l e n e e f f e c t h a s b e e n o b s e r v e d ( r e f . 68) i n t h e s e n s i t i z e d i s o m e r i z a t i o n o f 2.
178
01 *
u
f
/
t.
(-jN-N \
\r
Fig. 8 .
inv.
\ dN% f
R o t a t i o n a n d i n v e r s i o n p a t h w a y s of a z o b e n z e n e .
F o r b i s a z o aromatic compounds c o n s e c u t i v e ( r e f s . 6 8 , 7 5 , 7 6 ) and s i m u l t a n e o u s ( r e f . 7 7 ) i s o m e r i z a t i o n h a s b e e n r e p o r t e d . One o f t h e m o s t i n t e r e s t i n g i s s u e s of a z o b e n z e n e s p e c t r o s c o p y a n d p h o t o c h e m i s t r y is t h e q u e s t i o n o f t h e mechanism o f isomerizat i o n . Only a f e w y e a r s a f t e r H a r t l e y h a d f o u n d t h e E -> Z isomeriz a t i o n of a z o b e n z e n e Magee, Shand a n d E y r i n g ( r e f . 7 8 ) a r t i c u l a t e d t h e idea t h a t t h e r e m i c h t b e an i s o m e r i z a t i o n mechanism a c t i v e which is d i f f e r e n t from t h a t o f s t i l b e n e . They p r o p o s e d a p l a n a r t r a n s i t i o n s t a t e w i t h b r o k e n d o u b l e bond i n t h e azo p r o u p . The
" l a t e r a l s h i f t mechanism" ( i n v e r s i o n ) c o n c e p t w a s i n t r o d u c e d by C u r t i n , Grubbs a n d McCarthy i n 1966 ( r e f . 7 9 ) i n a n a l o g y t o t h e g r o u n d s t a t e i s o m e r i z a t i o n mechanism o f i m i n e s . The t r a n s i t i o n s t a t e s h o u l d h a v e s p - h y b r i d i z e d g e o m e t r y a t t h e N-atom, t h e n-system would be l i t t l e i n v o l v e d . R o s s a n d B l a n c w r o t e i n 1971: " T h i s " r e h y b r i d i z a t i o n " mechanism now a p p e a r s t o be w i d e l y
accepted
....However,
179 t h i s d o e s n o t mean t h a t t h i s mechanism i s
o p e r a t i v e i n t h e p h o t o i s o r n e r i z a t i o n s and t h e r e is, i n f a c t , circ u m s t a n t i a l e v i d e n c e t o t h e c o n t r a r y " . When w e keep i n mind t o d i s c r i m i n a t e between a z o compounds and pseudo s t i l b e n e s a c l o s e r l o o k a t t h e new i s o m e r i z a t i o n r e s u l t s r e v e a l s t h a t t h i s s t a t e m e n t should be reconsidered. The r o t a t i o n v s . i n v e r s i o n c o n t r o v e r s y h a s c a l l e d f o r t h e o r e t i c a l t r e a t m e n t . O l d e r c a l c u l a t i o n s ( r e f . 80) s u f f e r e d from t h e o m i s s i o n of t h e doubly e x c i t e d s t a t e t h a t O r l a n d i and S i e b r a n d ( r e f . 8 1 ) c o n s i d e r e d first f o r s t i l b e n e . A b i n i t i o calc u l a t i o n s f o r t h e model compounds d i i m i n e and azomethane ( r e f s . 82-85) have shown t h a t a l i n e a r t r a n s i t i o n s t a t e i n t h e i n v e r s i o n mechanism which may be s e e n t o d e v e l o p a l o n g a v i b r a t i o n a l normal c o o r d i n a t e , is much h i g h e r i n e n e r g y t h a n a s e m i l i n e a r t r a n s i t i o n s t a t e w i t h i n v e r s i o n of t h e bond a n g l e a t o n l y one n i t r o g e n atom ( F i g . 8 ) . Thus a l i n e a r t r a n s i t i o n s t a t e is o u t of d i s c u s s i o n today. Some c a l c u l a t i o n s i n d i c a t e t h a t r o t a t i o n s h o u l d b e p r e f e r r e d i n t h e ( n , n * ) s t a t e ( r e f . 8 2 ) , o t h e r s a l l o w b o t h pathways i n b o t h ( n , n * ) - and ( n , n * ) - s t a t e s ( r e f s . 8 3 , 8 4 ) . For azobenzene t h e r e a r e CNDO c a l c u l a t i o n s ( r e f . 86) and an a b i n i t i o c a l c u l a t i o n of t h e e n e r g y of f o u r r e p r e s e n t a t i v e p o i n t s ( E , Z , 90' t w i s t and sernil i n e a r ) i n t h e c o n f i p u r a t i o n s p a c e ( r e f . 8 7 ) . The l a t t e r i n d i c a t e s t h a t t h e t w i s t e d ( n , n * ) - s t a t e is h i g h e r i n e n e r g y t h a n E and Z, whereas t h e s e m i l i n e a r c o n f i g u r a t i o n is lower t h a n 2, b u t h i g h e r t h a n E. Monti, O r l a n d i and P a l m i e r i ( r e f . 34) have e v a l u a t e d t h e i r e a r l i e r c a l c u l a t i o n s ( r e f . 87) f u r t h e r and proposed a p o t e n t i a l e n e r g y diagram ( v i d e i n f r a ) . A l l c a l c u l a t i o n s a g r e e t h a t i n t h e t w i s t e d as w e l l as i n t h e s e m i l i n e a r c o n f i g u r a t i o n s t h e e x c i t e d s t a t e and ground s t a t e e n e r g i e s are n o t f a r a p a r t which f a v o u r s v e r y f a s t r a d i a t i o n l e s s deactivation. From t h e e v a l u a t i o n of e x p e r i m e n t a l d a t a Ross and B l a n c ( r e f . 7) and o t h e r s ( r e f . 24,64,88) have assumed t h a t p a r t i a l i n t e r s y s t e m c r o s s i n g of t h e e x c i t e d E-azobenzene m o l e c u l e t o t h e t r i p l e t s t a t e and i t s d e a c t i v a t i o n t o t h e ground s t a t e s h o u l d b e t h e r e a s o n for t h e o b s e r v e d l a c k of f l u o r e s c e n c e and t h e p e c u l i a r i t i e s i n p h o t o i s o m e r i z a t i o n . D i f f e r e n c e s from s t i l b e n e s h o u l d be due t o t h e p a r t i c i p a t i o n of t h e ( n , n * ) - s t a t e which changes t h e s i n g l e t - t r i p l e t t r a n s i t i o n p r o b a b i l i t i e s a c c o r d i n g t o
180
the rules of El-Sayed (ref. 89). Indeed, the non-Arrhenius behavior of the temperature dependence of the quantum yields and the falling short of the sum of $E and $z from unity are facts that hint at the participation of more than one electronic state in the photoisomerization process. Fischer and coworkers have constructed an energy level scheme in order to rationalize the experimental results available up to 1968 (ref. 64). This scheme has been completed by Ronayette et al. (ref. 24) by including proposed transition efficiencies (Fig. 9a). However, the new information that azobenzene molecules which cannot rotate around the -N=N-bond for structural reasons do isomerize (ref. 68) and that for them Q E is equal on n - > n * and a->n* excitation (refs. 18,68) requires a modification of the proposed isomerization mechanism. Important information is provided by the absorption spectra. The absence of any structure from the n - > n * bands of all noncyclic azo compounds indicates that the Franck-Condon state reached by excitation is situated on a slope of the potential energy surface of the '(n,n*)-state. This holds also for the spectra of molecules 5 and 2. We therefore have to infer that the relevant molecular coordinate is that of inversion. This is in accord with the simple one-electron orbital scheme of Walsh (ref. 90) for diimine and does not contradict more recent calculations for azomethane. n->n* excitation of the E form populates a state analoguous to the (n,n*)-state of stilbene. It is known that Estilbene rotates with a small activation to the 9 0 " singlet "phantom" state in about 7 0 ps (ref. 91) and that the lightinduced Z -> E isomerization is complete in less than 20 ps (ref. 92) (cf. Chapter 3 ) . The weaker vibrational structure of the a -> P* bands of azobenzene compared to that ofstilbeneindicatesa very low barrier towards rotation. This barrier is higher for more rigid structures, e.g., Ag, and highly viscous solvents (Fig. 2b). In the spectra of the Z form the n - > a * and n->n* bands are continuous. Both excited Franck-Condon states are on slopes of the '(n, a * ) and the ( a , **)-state, respectively. In order to rationalize this difference in the band characteristics and the factor 2 in the isomerization yields on (n,n*> and ( * , * * I excitation Rau has developed the concept of different singlet isomerization mechanisms in different states (ref. 61). The crucial experiments for this concept are those with azobenzenes without the option of rotation (refs. 18,61,68).
ta J
azobenzene ozobenzene tmns
cis TRANS
p=SW
P
C
5
cis
Fig. 8 .
trans
Energy diagrams of isomerization o f azobenzene. ( a ) Jablonski type diagram according to r e f .
181
24, (b) Potential energy diagram according to ref. 61, ( c ) Potential energy of rotation according to ref. 34, ( d ) Potential enerey diapram of inversion according t o ref. 34.
182 The o b s e r v e d e q u a l i t y of t h e quantum y i e l d s of E -> zation f o r the (n,r*)-
and ( * , n * ) - e x c i t e d
Z isomeri-
m o l e c u l e s of t h e s e
compounds i n d i c a t e s t h a t a l l m o l e c u l e s r e g a r d l e s s of which s t a t e is e x c i t e d i n i t i a l l y r e a c h t h e l o w e s t e l e c t r o n i c s t a t e and i s o m e r i z e by t h e same ( i n v e r s i o n ) mechanism. On t h e o t h e r hand, t h e f a c t t h a t t h e v a l u e s of t h e E -> Z y i e l d of t h e s e m o l e c u l e s and o f azobenzene are t h e same i n d i c a t e t h a t i t is t h e i n v e r s i o n mechanism which is ac t i v e i n t h e ( n , n * ) - s t a t e
$E of 0.20 t o 0.25.
o f azobenzene w i t h
I n azobenzene which i s w i t h o u t r e s t r i c t i o n , a r o t a t i o n a l f e a t u r e is o p e r a t i v e i n t h e ( * , * * ) - s t a t e which
c a u s e s OE t o d r o p t o 0.10 t o 0.15. A p o t e n t i a l e n e r p y diagram h a s been c o n s t r u c t e d ( F i g . 9 b . ) t h a t c o m p r i s e s most of t h e e x p e r i m e n t a l e v i d e n c e , c a l c u l a t i o n s and t h e o r e t i c a l c o n s i d e r a t i o n s . The proposed d o u b l e or mixed mechanism r e q u i r e s a v e r y low i n t e r n a l c o n v e r s i o n r a t e from t h e E - ' ( r , n * ) t o t h e E- 1( n , r * ) s t a t e . I t is r e a s o n a b l e t o e x p e c t t h i s as t h e r e is a n u n u s u a l l y
large e n e r g y gap between t h e s e s t a t e s . The e n e r g y d i s t a n c e of t h e Franck-Condon s t a t e s is 6000 t o 8000 c m - l ( F i g . l ) , t h a t o f t h e r e l a x e d '(n,n*)- and t h e E - l ( n , r * ) - s t a t e s may be n e a r 10 0 0 0 c m - l . Small e l e c t r o n i c s t a t e c o u p l i n g g i v e s t h e m o l e c u l e s i n t h e l(n,n*)-state a chance t o c r o s s t h e r o t a t i o n a l b a r r i e r t o t h e "phantom" s t a t e . An i n c r e a s e i n s o l v e n t v i s c o s i t y ( a s is e f f e c t e d by low t e m p e r a t u r e ) s h o u l d f a v o r i n t e r n a l c o n v e r s i o n and t h u s p a r t i c i p a t i o n o f t h e i n v e r s i o n mechanism o f i s o m e r i z a t i o n i n t h e l o w e s t e x c i t e d s t a t e . T h i s would e x p l a i n t h e non-Arrhenius b e h a v i o r of azobenzene. However, t h e n t h e y i e l d s of 313 nm and 436 nm e x c i t a t i o n s h o u l d become e q u a l a t 7 7 K i n g l a s s y s o l v e n t s as h a s been o b s e r v e d f o r t h e a z o - n a p h t h a l e n e s b u t n o t f o r azobenzene ( r e f . 6 4 ) . Monti, O r l a n d i and P a l m i e r i ( r e f . 3 4 ) have e v a l u a t e d t h e same d a t a and d e r i v e d s t a t e c o r r e l a t i o n s on t h e b a s i s of t h e i r c a l c u l a t i o n o f t h e 4 prominent p o i n t s i n c o n f i g u r a t i o n s p a c e (E, Z , 90' t w i s t and s e m i l i n e a r ) . They have a l s o i n c l u d e d t h e t r i p l e t s t a t e s . T h e i r well-founded p o t e n t i a l e n e r g y d i a g r a m ( F i g . 9c, 9 d ) d i f f e r s from t h e o n e of Rau i n two r e s p e c t s . ( i ) The l ( n , r * ) - s t a t e i s r a i s e d above t h e ' ( r , n * ) - s t a t e on r o t a t i o n and ( i i ) t h e r e is no s h a l l o w minimum of t h e l ( n , n * ) - s t a t e a t t h e E c o n f i g u r a t i o n . According t o this diagram n - > n * e x c i t a t i o n l e a d s t o i n v e r s i o n , *->TI * e x c i t a t i o n c a u s e s a t w i s t t o an e n e r g y minimum. T h i s minimum c o n f i g u r a t i o n i s c o u p l e d t o t h e ' ( n , r * > s t a t e and t h e ground s t a t e c a u s i n g an assumed 50% b i f u r c a t i o n
183 p r o b a b i l i t y . The l ( n , r * > m o l e c u l e s s h o u l d go back t o i n c r e a s i n g P l a n a r i t y and start t o i s o m e r i z e by i n v e r s i o n . The Z -> E quantum y i e l d d a t a which are less a c c u r a t e , are of l i m i t e d h e l p i n t h e d i s c u s s i o n o f t h e i s o m e r i z a t i o n mechanism. S i a m p i r i n g u e e t a l . ( r e f . 6 3 ) i n t e r p r e t t h e i r d a t a i n terms o f t w o d i f f e r e n t Z -> E y i e l d s on e x c i t a t i o n of t h e '(r,r*)and
' ( n , r * ) - s t a t e s of 0 . 4 and 0 . 5 5 , r e s p e c t i v e l y . The d a t a on Z -> E i s o m e r i z a t i o n are n o t i n d i s a g r e e m e n t w i t h t h e proposed
mechanisms. Both mechanisms have t h e i r weakness, n e i t h e r o f them c a n r a t i o n a l i z e t h e t e m p e r a t u r e dependence of t h e quantum y i e l d s of azobenzene. But a t p r e s e n t t h e r e i s no e x p e r i m e n t a l way of d i s t i n g u i s h i n g between them. Both models t r e a t azobenzene i s o m e r i z a t i o n by d i r e c t e x c i t a t i o n as a s e q u e n c e o f s i n g l e t p r o c e s s e s f o r two r e a s o n s : ( i ) I t seems t h a t t h e motion o f t h e m o l e c u l e s on t h e s u r f a c e s is t o o f a s t f o r i n t e r - s y s t e m c r o s s i n g , oxygen or heavy atoms have v i r t u a l l y no i n f l u e n c e on t h e i s o m e r i z a t i o n y i e l d s ( r e f . 6 7 ) . ( i i ) The o n l y p i e c e s of e x p e r i m e n t a l i n f o r m a t i o n on t h e t r i p l e t
pathway are t h a t i s o m e r i z a t i o n i n t h e t r i p l e t works ( r e f s . 2 3 , 2 6 , 6 7 , 7 2 , 7 3 ) , even i n t h e azobenzenophanes ( r e f . 68) and t h a t t h e r e a r e t w o s t a t e s i n v o l v e d ( r e f s . 2 3 , 2 4 , 6 3 ) . T h e r e may b e some l e a k i n e from t h e e x c i t e d s i n g l e t t o t r i p l e t s t a t e s which are v e r y short-lived.
But B o r t o l u s and Monti c o n c l u d e from t h e d i f f e r e n t
Z -> E quantum y i e l d s of t h e i s o m e r i z i n g s t a t e s p o p u l a t e d by
d i r e c t e x c i t a t i o n and by e n e r g y t r a n s f e r , r e s p e c t i v e l y , t h a t t h e " d i r e c t p h o t o i s o m e r i z a t i o n of azobenzene o c c u r s i n t h e s i n g l e t s t a t e " ( r e f . 67). I n c o n c l u s i o n , t h e r e seems t o emerge t h e c o n c e p t of independent s i n g l e t and t r i p l e t i s o m e r i z a t i o n pathways. The i n v e r s i o n mechanism i n t h e l ( n , n * ) - s t a t e is g a i n i n g s u p p o r t and t h e r e i s a f e e l i n g t h a t t h e d e a c t i v a t i o n o f t h e l ( r , r * ) - s t a t e at l e a s t b e p i n s w i t h a r o t a t i o n a l motion. But a l t h o u g h new e x p e r i m e n t a l material h a s been f e d i n t o t h e d i s c u s s i o n t h e i s o m e r i z a t i o n mechanism is n o t y e t f u l l y understood. 3.2 Compounds o f t h e aminoazobenzene t y p e I r r a d i a t i o n o f amino or hydroxy s u b s t i t u t e d azobenzenes d o e s
n o t l e a d t o a p e r s i s t e n t c o l o r change, t h e t h e r m a l r e c o v e r y of
t h e i n i t i a l s p e c t r u m is f a s t compared w i t h azobenzene t y p e molec u l e s . Z i s o m e r s are formed by i r r a d i a t i o n which h a s been d e t e c t e d by Brode, Gould and Wyman ( r e f . 93) u s i n g t h e r o t a t i n g s h u t t e r
184
300
LOO
Fig. 10. E-form
500 nm
P h o t o i s o m e r i z a t i o n s p e c t r a o f 4-dimethylamino-azobenzene.
-, Z-form
(extrapolated)
------
( r e f . 93).
t e c h n i q u e ( F i g . 1 0 ) and by F i s c h e r a n d F r e i i n l o w t e m p e r a t u r e e x p e r i m e n t s ( r e f . 9 4 ) . H a l f l i v e s o f t h e Z f o r m o f 220 min a t 298 K i n t o l u e n e ( r e f . 9 4 ) a n d 52 min a t 313.5 K i n b e n z e n e / p i p e r i d i n e ( r e f . 55) h a v e b e e n r e p o r t e d f o r d i m e t h y l a m i n o a z o b e n z e n e . The a c t i v a t i o n e n e r g i e s f o r t h i s g r o u p are somewhat lower t h a n , b u t n o t v e r y d i f f e r e n t f r o m t h o s e o f a z o b e n z e n e t y p e
m o l e c u l e s : t h e y r a n g e b e t w e e n 75 a n d 88 kJ m o l - l , w i t h s t e r i c h i n d r a n c e b y o r t h o s u b s t i t u t i o n b e t w e e n 88 a n d 100 kJ m o l - l . A H a m m e t t r e l a t i o n s h i p b e t w e e n r a t e s of i s o m e r i z a t i o n a n d s u b s t i t u e n t s ( r e f . 9 5 ) a n d a l i n e a r r e l a t i o n o f I n kisom
vs. solvent p o l a r i t y
e x p r e s s e d as T a f t n* p a r a m e t e r h a v e b e e n p o i n t e d o u t by N i s h i m u r a e t a l . ( r e f . 96). A l b i n i , F a s a n i and P i e t r a h a v e m e a s u r e d t h e E -> Z quantum
y i e l d s o f 4-diethylamino-azobenzene and 4 - d i e t h y l a m i n o - 4 ' methoxyazobenzene a t d i f f e r e n t w a v e l e n g t h s ( r e f . 9 7 ) . F o r t h e l a t t e r t h e y f i n d i n d e a e r a t e d c y c l o h e x a n e s o l u t i o n s v a l u e s of
0.72 ( 4 3 4 nm), 0 . 2 1 ( 3 6 6 nm), 0 . 2 5 ( 3 1 3 nm) and 0.23 ( 2 5 4 nm) a n d f o r t h e f o r m e r 0 . 8 4 , 0.27, 0.31 a n d 0 . 3 4 , r e s p e c t i v e l y . T h i s is t h e same p a t t e r n a s i n a z o b e n z e n e : h i g h y i e l d f o r l o w e n e r g y e x c i t a t i o n , which may be due t o p r e f e r e n t i a l e x c i t a t i o n o f t h e
a t 434 nm. The quantum y i e l d s o f Z -> E i s o m e r i z a t i o n c a n n o t b e d e t e r mined a t room t e m p e r a t u r e b e c a u s e o f t h e p a r a l l e l t h e r m a l r e a c t i o n . T h e r e are no e f f o r t s r e p o r t e d i n t h e l i t e r a t u r e t o d i s c u s s t h e i s o m e r i z a t i o n mechanism. I n c o n t r a s t t o a z o b e n z e n e t h e a m i n o a z o b e n z e n e s may s u f f e r a n (n,r*)-state
185
irreversible photoreduction when excited to higher states. Albini et al. (ref. 97) have shown that the same reaction is induced by hiFh energy sensitizers and conclude that this photoreduction is a reaction of a triplet state which may be populated from a higher singlet or by a triplet donor. 3.3. Pseudo stilbenes The two types of pseudo stilbenes are different in terms of the lifetimes of the Z forms. Protonated Z-azobenzene isomerizes thermally with rates and activation energies that depend on the acid content of the solvent (refs. 39,981. In 66% sulfuric acid Ea iS > 100 kJ mol-' and the photoisomerization reaches a stationary state. Mauser, Francis and Niemann determined the quantum yields of isomerization $E = 0.27 and 4, = 0.25 (ref. 99). From the stationary state the p h o t o c y c l o d e h y d r o g e n a t i o n reaction starts to give benzo[c]cinnoline which was observed first by Lewis (refs. 39,40) quite in analogy to the phenanthrene formation by stilbene. The Z forms of donorlacceptor substituted azobenzenes isomerize quickly at room temperature. Flash experiments (refs. 38,100) reveal lifetimes in the ms range. The mechanism of this thermal reaction has been much in debate. In 1971 Wildes, Pacifici, Irick and Whitten investigated the relaxation of a series of 4nitr0-4'-alkylamino-azobenzenes following flash photolysis (ref. 100). From the drop of Ea with solvent polarity they infer a rotational mechanism, which was reconfirmed by Schanze, Mattox and Whitten in 1982 (ref. 101) and 1983 (ref. 102). From the activation volume Asano et al. (refs. 103,104) infer a change of mechanism, inversion in hexane, rotation in benzene and later (refs. 105,106) a competition of mechanisms. Nishimura et al. (ref. 96) come to favor again the inversion mechanism to explain their finding that the volumes of activation and reaction are nearly equal in Z -> E isomerization of $. The photoisomerization yields have been determined at low temperatures <160 K for $ (ref. 107). Of course, the $E are small: 0.02 at 313 nm and 0.027 at 436 nm for 163 K, but they are comparable, a feature which is expected for pseudo stilbenes. The QZ values are 0.62 and 0.71, respectively. The mechanism of photoisomerization has not been a major issue in discussion, probably because the difference between azo compounds and pseudo stilbenes has not been appreciated and because of the paucity of relevant experimental data. The weakness of
186
fluorescence indicates a fundamental difference in the deactivation compared to that of stilbenes. If the (r,a*)-state is below the (n,n*)-state at the E configuration then the potential energy diagrams of Ponti, Orlando and Palmieri (ref. 34) and of Rau (refs. 61,681 merge. In the donor/acceptor substituted azo compounds the ' ( n ,n*)- and l(a,n*)-states will be vibrationally coupled at and statically mixed near the E configuration because the two types of states are comparable in energy, and molecular distortions are easy. Then both rotation in a predominantly (n,n*)-state and inversion in a predominantly (n,n*)-state should be possible. In the triplet system, however, rotation should be the preferred mode of relaxation. This is similar to the stilbene case and, indeed, a rather long lived state has been observed in viscous media (ref. 38). This mechanism for the isomerization of pseudo stilbenes rationalizes the fact that the lightfast azo dyes are pseudo stilbenes, whereas stilbenes are used as optical brighteners. It also allows for the observed fading of azo dyes by self-sensitation (refs. 108-110) of lo2 induced oxidation. Degradation of azo dyes may also be effected by reductive processes (refs. 111-114). Photooxidation and photoreduction are "side exits" from the isomerization system and have quantum yields below 0.1%.
4. APPLICATION OF PHOTOCHROMISM OF A20 COMPOUNDS Azo compounds and pseudo stilbenes are easily isomerized reversibly with W and visible radiation and thus are candidates for information storage devices. However, the fast to slow thermal recovery of the E form limits the stability of the information deposited and eliminates these compounds as promising candidates for information storage materials. Another disadvantage is the fact that not a color but just an intensity change constitutes the information. Photoisomerization, however, is a switch for a number of properties different from color. The geometrical transformation of azobenzene units in larger molecules causes structural changes in these or in supramolecular systems. These changes can be maintained by continuous irradiation with near UV irradiation, and they can be switched off again with visible light. Photoisomerization of azobenzenes can also be m e d to trigger an irreversible sequence of structural events. Some examples will be given.
187 Azobenzene units in polymers have been used for diagnostic purposes to investigate the free volume (ref. 115) and for the study of dynamics of configurational changes (refs. 116,117) and of aging (ref. 118). Polymers with azobenzene units in the backbone and with pending azobenzenes have been prepared. Photoisomerization of these units influence swelling properties (ref. 119>, wettability (ref. 120), membrane properties (refs. 121-1251, viscosity (refs. 126,1271, and solubility (ref. 128). Photoresponsive crown ethers, like 2, whose complexing properties towards different ions can be modified by irradiation, have been designed for light-regulated ion extraction and transport (ref. 129). Binding abilities of azobenzene-containinp cyclodextrines have been influenced by Light (refs. 130,131). The catalytic properties of micelles of azobenzene-modified surfactants may be affected by irradiation (ref. 132) and liquid crystal properties can be changed by irradiation of embedded azobenzene units (ref. 133). A light-powered hydrogen ion pump has been devised which utilizes the differences in the waterloil distribution coefficients and in acidity constants of E- and 2-o-hydroxyazobenzenes (ref. 134). There are and will be more utilizations of the different properties of E and 2 azobenzene. 5.
SUMMARY
In this paper, the ensemble of organic aromatic azo compounds has been grouped in three types according to the relative energetic position of the l(n,a*)- and the '(r,r*)-states. It is possible to order the experimental data accordingly. The azobenzene type molecules are characterized by a large (a,a*) - (n,r*) state gap, which may cause restricted intersystem crossing and fast geometrical changes after excitation. The E -> Z isomerization yields are lower on high energy excitation than on excitation of the lowest excited state. The inversion mechanism in the pround and '(n,a*)-states seems to be accepted now. Still in discussion is the activity of the rotation mechanism in the (r,n*)-state. Increasing evidence appears for singlet isomerization on direct and triplet isomerization for sensitized excitation with little intercommunication of singlet and triplet states. Extremely weak fluorescence but no photophorescence has been observed, The aminoazobenzene type molecules are characterized by close-lying l(n,a*)- and '(a ,a*)-states, the solvent determines
188
which one is the lowest excited singlet state. In hydrocarbon solvents the E -> Z quantum yield is higher for low energy irradiation which is the same as for the azobenzene type molecules. The thermal back reaction is accelerated compared to the azobenzene type, inversion is assumed to be the mechanism. Weak fluorescence at 77 K but no phosphorescence has been observed. The pseudo stilbene type molecules are characterized by a low-lyinp '(IT,**)-state. Two groups are observed: donorlacceptor substituted azobenzenes and azobenzenes protonated at the azo group. (i) For the first group the quantum yields of low and high energy excitation are equal. Little is known about the isomerization mechanism in the excited states. Some suggestions are made in this paper. The thermal Z -> E isomerization is very fast, its mechanism still being debated, however with a bias towards inversion. Weak fluorescence at 77 K is recorded but no phosphorescence. In flash experiments, transient triplet states are observed in viscous media. (ii) The second group is so stilbenelike that not much work has been dedicated to it. The thermal Z -> E isomerization rate is dependent on the acid strength, the phokoisomerization yield about equal in both directions, and from the photostationary state photocyclization to benzolclcinnolin occurs. There is intense fluorescence but no isomerization at 77 K in sulphuric acid. Although photoisomerization is the main reaction of all types of azo compounds side reactions occur with quantum yields below They are photoreduction, photooxidation and photocyclization. For practical applications the most important type of azo compounds are still the pseudo stilbenes as most of the commercial dyes belong to this type. Their high lightfastness is based on the ease of giving up electronic energy to heat by configurational distortion (combined with tautomerism if free OH groups are in a favorable position). Azo compounds are less suited for optical storage devices, but they may have a future for triggering features other than color in tailor-made systems like polymers, crown ethers or vesicles. REFERENCES 1 H. Zollinger, "AZO and Diazo Chemistry", Interscience, New York and London, 1961; "Colour Chemistry. Synthesis, Properties and Applications of Organic Dyes", VCH, Weinheim 1987. 2 H. Bock, G. Rudolph, E. Baltin and J. Kroner, Angew. Chem., ?I (1955) 469-484; Angew. Chem. Intl. Ed. Engl., 4 (1955) 457.
189 3
M.J. Bruce, Angew. Chem., 89 (1977) 75-89; Angew. Chem. Intl. Ed. Engl., 16 (1977) 73-87, G.S.Hartley,Nature, 140 (1937) 281; J. Chem. SOC., 1938, 633-642.
R.F. Hutton and C. Steel, J. Am. Chem. SOC., 86 (1964) 745-746 G. Wyman, Chem. Rev., 55 (1955) 625-657. D.L. Ross and J. Blanc in ltPhotochromism",G.H. Brown, ed. Wiley-Interscience, New York, 1971. 8 H. Rau, Angew. Chem., 85 (1973) 248-258; Angew. Chem. Intl. Ed. Engl., 12 (1973) 224-235. 9 J. Griffiths, Chem. SOC. Rev., 1 (1972) 481-493. 10 DMS.UV-Atlas, Verlag Chemie, Wehheim 1967, spectra C7/5,
11
c7/17.
H. Suzuki "Electronic Absorption Spectra and Geometry of Organic Molecules", Academic Press, New York 1967. 12 H.H. Jaffe and M. Orchin "Theory and Application of Ultraviolet Spectroscopy", Wiley, New York, London 1962. 13 P.P. Birnbaum, J.H. Linford and D.W.G. Style, Trans. Faraday SOC., 49 (1953) 735-744. 14 P.H. Gore and D.W. Wheeler, J. O r e . Chem., 26 (1961) 3295-3299 15 H. Bisle, M. Romer and H. Rau, Ber. Bunsenges. Physik. Chem., 16 17
18 19 20 21 22 23
80 (1976)
301-305.
H. Rau, Bunsenges. Physik. Chem., 72 (1968) 408-414. R.H. Dyck and D.S. McClure, J. Chem. Phys., 36 (1969) 23262345.
H. Rau and Y.Q. Shen, J. Photochem. Photobiol. A Chemistry,
42 (1988)
321-327.
W.S. Struve, Chem. Phys. Lett., 46 (1977) 15-19. C.G. Morrante and W.S. Struve, C s m . Phys. Lett.,
267-271.
68
(1979)
H. Bisle and H. Rau, Chem. Phys. Lett., 31 (1975) 264-266. M. Shinornura and T. Kunitake, J. Am. Chem. S O C . , 109 (1987) 5175-5183.
24
J. Ronayette, R. Arnaud, P. Lebourgeois and J. Lemaire, Can. J. Chem., 52 (1974) 1848-1857. J. Ronayette, R. Araaud and J. Lemaire, Can. J. Chem., 52
30
M.A. El-Bayoumi and F.M. Abdel Halim, J. Chem. Phys.,
(1974) 1858-1867.
F. Pragst, 2. phys. Chem., (Leipzig), 261 (1980) 791-792. S. Monti, E. Gardini, P. Bortolus and E. Amouyal, Chem. Phys. Lett., 77 (1981) 115-119. 27 ref. 11, chapt. 19, 20 and 23. 28 G. Hohlneicher and W. Sanger in "Quantum Aspects of Heterocyclic Compounds in Chemistry and Biochemistry", E.D. Bergmann and B. Pullman. eds. Academic Press, New York 1970, p.193. 29 D.L. Beveridge and H.H. Jaffe, J. Am. Chem. SOC., 88 (1966) 25 26
31 32 33 34 35 36 37
1948-1953.
(1968) 2536-2541.
48
D.R. Kearns and M.A. El-Bayoumi, J. Chem. Phys., 2 (1963) 1508-1517.
E.V. Brown and G.R. Granneman, J. Am. Chem. Soc.,
91 (1975)
621-627. S. Ljunggren and G. Wettermark, Acta Chem. Scand., 1599-1606.
S. Monti, G. Orlandi and P. Palmieri, Chem. Phys.,
87-99.
F. Gerson and E. Heilbronner, Helv. Chim. Acta, 59.
45
2
(1971)
71 (1982)
(1962) 51-
H. Rau, Ber. Bunsenges. Physik. Chem., 75 (1971) 1343-1347. S. Monti, S. Dellonte and P. Bortolus, J. Photochem., 23 (1983) 249-256.
190 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74
H. G o r n e r , H. Gruen and D. S c h u l t e - F r o h l i n d e , J. Phys. Chem., (1980) 3031-3039. G.E.Lewis, J. Org. Chem., 25 (1960) 2193; T e t r a h e d . L e t t . , 1960 no. 9, 12. G.M. Badger, R . J . Drewer and G.E. L e w i s , Aust. J. Chem., 16 (1963) 1042. E. H a s e l b a c h , Helv. Chim. Acta. 53 (1970) 1526-1543. E. H a s e l b a c h and E. H e i l b r o n n e r , T e t r a h e d r o n L e t t . , 1967 4531-4535. G.E. L e w i s and R . J . M a y f i e l d , Aust. J. Chem., 19 (1966) 1445; i b i d . , 20 (1967) 1899; i b i d . , 2 (1968) 1601-1609. V. Gutmann and A. S t e i n i n g e r , Montsh. Chem., (1965) 11731182. H. Rau, B e r . Bunsenges. P h y s i k . Chem., 2 (1967) 48-53; i b i d . , 72 (1968) 408-414. P. P r i n g s h e i m , " F l u o r e s c e n c e and P h o s p h o r e s c e n c e " , N e w York 1949, p.422. E. Wolf and H.K. Camenga, Z . Phys. Chem., (Wiesbaden) NF, 107 (1977) 21-38. F.W. S c h u l z e , H . J . D e t r i k , H.K. Camenga and H. K l i n g e , Z. Phys. Chem., (Wiesbaden) NF, 107 (1977) 1-19. A.A. Adamson, A. V o g l e r , H. Kunkely and R. Wachter, J. Am. Chem. SOC., 100 (1978) 1298-1300. E.R. T a l a t y and J.C. F a r g o , J. Chem. SOC. Chem. COmmUn., 1967 , 65-66. J . P . O r t r u b a I11 and R.G. Weiss, J. Org. Chem., 48 (1983) 34483453. J . A . Andersson, R. P e t t e r s o n and L. T e g n e r , J . Photochem., 20 (1982) 17-32. S. Yamashita, B u l l . Chem. SOC. J p n . , 24 (1961) 842. R. Arnaud and J. L e m a i r e , Can. J. Chem., 52 (1974) 1869-1871. D. S c h u l t e - F r o h l i n d e , L i e b i g s Ann. Chem., 612 (1958) 131-138. H. Mauser, p e r s o n a l communication. H, Mauser, "Formale K i n e t i k " , B e r t e l s m a n n U n i v e r s i t a t s v e r l a g , D u s s e l d o r f 1974. H. Mauser, 2. N a t u r f o r s c h . , 23B (1968) 1025-1030. G. Zimmerman, L.Y. Chow and U.3. P a i k , J. Am. Chem. SOC., So (1958) 3528-3531. H. Rau, EPA N e w s l e t t e r , 2 (1984) 31-34. H. Rau, J. Photochem., 26 (1984) 221-225. 0. B i h l e r , wiss.Arb.f.H%.Lehramt, Univ. Hohenheim 1976. N. S i a m p i r i n g u e , C. Guyot, S. Monti and P. B o r t o l u s , J. Photochem., 37 (1987) 185-188. S. Malkin a n r E . F i s c h e r , J. Phys. Chem., 66 (1962) 2482-2486. D. Gegiou, K.A. Muszkat a n d E. F i s c h e r , J . Am. Chem. SOC., 9 0 (1968) 12-18. D. Gegiou, K.A. Muszkat and E. F i s c h e r , J. Am. Chem. SOC., 9 0 (1968) 3907-3918. (1979) 648-652. P. B o r t o l u s and..S. Monti, J . Phys. Chem., H. Rau and E. Luddecke, J. Am. Chem. SOC., 104 (1982) 16161620. B.P. H a l l a m and M.L._Pauson, J. Chem. SOC., 1958,642-645. U. Funke and H.F. Grtltzmacher, T e t r a h e d r o n , 43 (1987) 37873785. N . J . Bunce, G. F e r g u s o n , C.L. F o r b e r and G . J . S t a c h n y k , J. 0%. Chem., 52 (1987) 394-398. L.B. J o n e s and G.S. Hammond, J. Am. Chem. SOC., (1965) 42194220. E. F i s c h e r , J. Am. Chem. S O C . , 90 (1968) 796. J. S a l t i e l and E.D. P e g a r i t y , J. Am. Chem. S O C . , 94 (1972) 2742-2749.
84
191 75 76 77 78 79
80 81
82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109
A.H. Cook and D.G. J o n e s , J . Chem. S O C . , 1939 1309-1315. A.H. Cook, D.G. J o n e s and J.B. P o l y a , J. Chem. SOC., 1939 1315. K . I t o h o , T. Masuda, M. T a k e i , Y. S u k a r i and M. Nishigarni, J. Chem. SOC. Chem. Commun., 1986 1028-1030. J . L . Magee, W. Shand Jr. and n y r i n r , J . Am. Chem. SOC., 63 (1941) 677-688. D.Y. C u r t i n , E . J . Grubbs and C.G. McCarthy, J. Am. Chem. SOC., 88 (1966) 2775-2786. D.R. K e a r n s , J. Phys. Chem., 69 (1965) 1062-1065. G. O r l a n d i and W. S i e b r a n d , C G m . Phys. L e t t . , 30 (1975) 352354. N.C. B a i r d and J . R . Swanson, Can. J. Chem., (1973) 30973101. R.N. Camp, I.R. E p s t e i n and C. S t e e l , J . Am. Chem. Soc., 99 (1977) 2453-2459. G. O l b r i c h , Chem. P h y s . , 21 (1978) 117-125. B.M. Gimarc, J. Am. Chem. SOC., 92 (1970) 266-275. A. G o u r s o t , P. J a q u e s and J. F a u r e , J. Chim. P h y s . , 5 (1976) 694-695. S. Monti, G. O r l a n d i and P. P a l m i e r i , V I I I IUPAC Syrnposium on P h o t o c h e m i s t r y , S e e f e l d ( A u s t r i a ) 1980, Book o f A b s t r a c t s p a p e r No. 119. H. Rau, J. Luminescence, (1969) 191-199. M.A. El-Sayed, J . Chem. P h y s . , 36 (1962) 573; i b i d . , 38 (1963) 2434-2438. A.D. Walsh, J . Chem. SOC., 1953 2288-2296. R.M. H o c h s t r a s s e r , P u r e Appl. Chem., 52 (1980) 2683-2691. M. S u m i t a n i , N. Nakashima and K. Y o s h i h a r a , Chem. Phys. L e t t . , 68 (1979) 255-257. W.R. Brode, J . H . Gould and G. Wyman, J . Am. Chem. Soc., 74 (1952) 4641-4646. E. F i s c h e r and Y. F r e i , J . Chem. P h y s . , 27 (1957) 328. N. Nishimura, T. S u e y o s h i , H. Yamanaka, E. Imai, S. Yamamoto and S. Hasegawa, B u l l . Chem. SOC. J p n . , 49 (1976) 1381-1387. N. Nishimura, T. Tanaka, M. Asano and Y. S u e i s h i , J . Chem. SOC., P e r k . 11, 1986 1839-1845. A. A l b i n i , E. F a s a n i and S. P i e t r a , J . Chem. S O C . , P e r k . 1 1 , 1983 1021-1024. n a u , A.D. Crosby, A. S c h a u f f l e r and R. F r a n k , 2. Naturf o r s c h . , 36A (1981) 1180-1186. H. Mauser, D.J. F r a n c i s and H . J . Niemann, 2. P h y s i k . Chem. (Wiesbaden) NF, 3 (1972) 318-333. P.D. W i l d e s , J . G . P a c i f i c i , G . I . I r i c k and D.G. W h i t t e n , J . Am. Chem. SOC., 93 (1971) 2004-2008. K . S . Schanze, T.F. Mattox and D.G. W h i t t e n , J . Am. Chem. SOC., 104 (1982) 1733-1735. K . S . Schanze, T.F. Mattox and D.G. W h i t t e n , J . Org. Chem., 48 (1983) 2808-2811. T. Asano, T. Okada, S . S h i n k a i , K . S h i g e m a t s u , Y. Kusano and 0. Manabe, J. Am. Chem. S O C . , 103 (1981) 5161-5165. T. Asano, J. Am. Chem. S O C . , 102 (1980) 1205-1206. T. Asano, T. Yano and T. Okada, J . Am. Chem. SOC., 104 (1982) 4900-4904. T. Asano and T. Okada, J. Org. Chem., 49 (1984) 4387-4391. G. Gabor and E. F i s c h e r , J. Phys. Chem., 75 (1971) 581-584. J . G r i f f i t h s and C. Hawkins, J . Chem. SOC. P e r k 11, 1977 747-752. H . Gruen, H. S t e f f e n and D. S c h u l t e - F r o h l i n d e , J . SOC. Dyers C o l o r i s t s , 91 (1981) 430.
-
-
192 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137
H. M u r t r o p h , 2. Chem., 23 (1983) 61.
G. I r i c k a n d J . G . P a c i f i c i , T e t r a h e d . L e t t . , 1969 NO. 17. 1303-1306. A. A l b i n i . E. F a s a n i a n d S. P i e t r a , J. Chem. SOC. P e r k . 1 1 , 1982 1393-1395. B.E. B l a i s d e l l , J. SOC. D y e r s C o l o u r i s t s , 65 (1949) 618. V. Rehak. F. Novak and I . CeDcanskv. C o l l . Czech. Chem. Commun., 38 (1973) 697-705. e . g . C.D. E i s e n b a c h . B e r . Bunsenges. P h y s i k . Chem., (1980) 680-690. M. I r i e and W. S c h n a b e l . M a c r o m o l e c u l e s , (1985) 394-398. A. Yamamoto, Macromolecules, 19 (1986) 2472-2476. L. L a m a r r e a n d C.S.P. Sung, M a c r o m o l e c u l e s , 16 (1983) 17291736. K. I s h i h a r a , N. Namada, S. Kato a n d I . S h i n o h a r a , J. P o l y m e r . S c i . , Polym. Chem. Ed., 22 (1984) 121-128. K. I s h i h a r a , N. Namada, S. Kato a n d I . S h i n o h a r a , J. P o l y m e r Sci.Polym. Chem. Ed., 3 (1983) 1551-1555. W . J . D e a l , B.F. E r l a n g e r and D. Nachmansohn, P r o c . Natl. Acad. S c i . , U S . 64 (1969) 1230. D. Balasubramaman, S. Subramani a n d C. Kumar, N a t u r e , 254 (1975) 252-254. T . I . A n z a n i , H. S a b a k i , A. Ueno and T. O s a , J. Chem. soc. P e r k . 11, 1985 903-907. Y. O k a h a t a , H - J . Lim a n d S. H a c h i y a , J. Chem. SOC., P e r k . 1 1 , 1984 989-994. n k a h a t a , S. F u j i t a a n d N. I i z u k a , Angew. Chem., 98 (1986) 723-725; Angew. Chem. I n t l . Ed E n g l . , 25 (1986) 7 5 1 7 M. I r i e , Y. H i r a n o , S . Hashimoto a n d K. H a y a s h i , Macrom o l e c u l e s , g (1981) 262-267. R. L o v r i e n , Proc. N a t l . Acad. S c i . , US, 57 (1967) 236. M. I r i e and €IT.a n a k a , M a c r o m o l e c u l e s , c ( 1 9 8 3 ) 210-214. S. S h i n k a i a n d 0. Manabe, Top. C u r r . Chem., 121 (1984) 67-104. A. Ueneo, H. Yoshimura and T. O s a , J . Am. Chem. S o c . , 101 (1979) 2779-2780. P . B o r t o l u s and S. M o n t i , J . P h y s . Chem., 91 (1987) 5046. S. S h i n k a i , K. Matsuo, A . Harada and 0. Manabe, J. Chem. SOC., P e r k . 1 1 , 1982 1261-1265. F. H e n t z e , Z. Chem., 17 (1977) 294, P. H a b e r f i e l d , J. Am.Chem. SOC., 109 (1987.) 6177-6178; i b i d . , 109 (1987) 6178-6179. P.P. Birnbaum a n d D.W. S t y l e , T r a n s . F a r a d a y SOC., 50 (1954) 1192. S. Y a m a s h i t a , H. Ono and 0. Toyama, B u l l . Chem. SOC. J p n . , 35 (1962) 1849. H. S t e g e m e y e r , J. Phys. Chem., (1962) 2555-2560. -
I
-
See Additional Literature (1989
- 2001): Azocornpounds, PA1
193
Chapter 5
4n.Systems Based on 1.3Electrocyclization C. Schulz and H. Durr
1 INTRODUCTION
The photochromic behavior of many classes of compounds is based on electrocyclizations. In this chapter, photochromic systems with four electrons which undergo 1,3-electrocyclic reactions are dealt with. 1,5-Electrocyclizations involving six electrons are described elsewhere. In scheme 1, a survey of electrocyclic processes in 4n-systems (n = 1,2,...) is shown. Scheme 1: Electrocyclic Processes in In-Systems (n = 1,2,...) hv or A
cycles n=l
n=l
n=2
-
open ring form
type of electrocyclization
__c
___)
Important photochromic heterocycles and open ring forms i S 0 electronic with the cyclopropyl and ally1 anion, respectively, are given in scheme 2. The cyclobutene-butadiene interconversion, where a wealth of derivatives has been studied both theoretically and experimentally, has not given rise to important photochromic systems. This reaction is therefore not treated in this book. The reader is
194
referred to earlier literature (ref. 1). Similar arguments hold for the cyclooctatriene-octatetraene interconversion which is excluded from this book as well. Scheme 2: Photochromism based on 1,3-electrocyclization (4 electrons) heterocycles hv or A
openrhgfom
VpOf bond cleavage
2 STEREOCHEMISTRY IN 1,3-ELECTROCYCLIC REACTIONS
An outstanding feature of electrocyclic reactions is their stereospecificity. The orbital symmetry conservation principle in the form of the Woodward-Hoffmann rules (ref. 2) predicts the following behavior for symmetrical hydrocarbons: When the number of interacting electrons in the cyclic array is 4n (n = 1,2,..) the thermal electrocyclic reactions proceed via conrotatory pathways, the electrocyclic reactions which occur as primary photochemical processes via disrotatory pathways. The rules may be applied to slightly disturbed systems. The question which arises concerns the application of the WoodwardHoffmann rules to unsymmetrical substances including both highly substituted hydrocarbons as well as compounds containing hetero-
195
atoms. A paper by Snyder (ref. 3) describes the CNDO-calculated electrocyclization of the open ring form l, isoelectronic with the ally1 anion, to the corresponding three membered ring 2. 700' 0
loor@ Ox\
___)
y n z
4--
1
X e C H , NH
Y = CH2, NH, 0 Z e CH2, NH, 0
Y-2 2
The results of these calculations show that heteroatom substitution of I, 2 leads to three types of ring closures: 1. Thermal conrotation is allowed: X = NH, Y = Z = CH2; Z = CH2; Lr 2; X = CHI Y X = CH, Y = CH2, Z = NH. 2. Thermal conrotation and disrotation are allowed: 21 2: X '2 NH, Y = CH2: X = NH, Y = CH2, Z = 0. 3. Conrotation and disrotation are thermally forbidden but photochemically allowed: 1, 2; X = CH, Y = CH2, Z = 0; X = CH, Y = Z = NH; X-CH, Y=N?I, Z = O ; X=CH, Y = Z = O .
3
OXIRANES
3.1 Monocvclic Arvloxiranes
The literature on oxiranes has been reviewed (refs. 4-6). It has been shown that room temperature photolysis of aryloxiranes (incorporating 3) causes cycloelimination to give arylcarbenes and carbonyl compounds.
3
5
4
Photolysis (77 K) in rigid glasses produces highly colored intermediates (ref. 7), which are stable at low temperatures, but
6
7
196
are bleached by warming to 25 OC or upon irradiation with visible light. Color formation is attributed to C-C-bond cleavage with formation of carbonyl ylides 2. Thap DoMinh, Trozzolo and Griffin (ref. 8) investigated the photochromism of cis- and trans-stilbene oxide. In the present review these compounds are treated as typical examples of other photochromic monocyclic oxiranes. Irradiation of trans-stilbene oxide in ethanol glass at 77 K produces an orange material = 490 nm) along with small amounts of benzaldehyde, phenylmethylene and desoxybenzoin. Irradiation of the cis isomer gives similar products, but the colored intermediate is a deep red compound (Amax = 510 nm). At 140 K, both colors disappear (the bleaching occurring somewhat faster in the trans than in the cis isomer) giving benzaldehyde and phenylmethylene. The amount of fragmentation products formed by this photolysis-warm-up-procedure is estimated to be 20-25 times more than the amount originally produced by photolysis. Irradiation in the visible causes rapid fading and regenerates the original oxirane together with small amounts of fragmentation products. The process may be envisaged as an electrocyclic reaction which interconverts oxiranes and open chain carbonyl ylides by a conrotatory or disrotatory mode. Three possible ylides may be derived from stilbene oxides:
8
9
10
The order of stability is 2 > s >> lo. From the relative stabilities and absorption spectra it is possible to assign cis-oxoylide 2 to the more stable red shifted intermediate from cisstilbene oxide, and trans-ylide & to the one from trans-stilbene oxide. Thus the electrocyclic reaction very probably involves a disrotatory course and in the case of cis-stilbene oxide an outward disrotation. This is supported by the results of Huisgen and coworkers on theisoelectronicaziridine-azomethine ylide system (ref. 9, see also this review, 4.1) and additionally by the photochromic be-
197 havior of bicyclic oxiranes whose special geometry permits an outward disrotatory course (ref. 10).
only
hv ___,
t
disrotation
H
Ph
-
11
hv
t--
Ph
Ph 12
disrotation
Ph
H 8
0
phYoYph H
H
9
3.2 Bicvclic and Tricvclic Oxiranes 3.2.1 5-0xabicvclo~2.1.01~1entane Stabilization of carbonyl ylides may be achieved by incorporation of this moiety into a cyclic structure. Solutions of 13 in diglyme or benzene become colored (Amax 5 4 0 nm, purple), when heated to 100 OC or when irradiated (253.7 nm) at room temperature. The visible absorption spectrum of the colored species is essentially the same whether generated thermally or photochemically and is undoubtedly due to the same species. The colored material is stable toward return to 13 (but reacts as a typical 1,3-dipole). The carbonyl ylide structure 14 is assigned to the colored species by analogy to similar systems N
Me
13
14
and by absence of an esr signal from the products obtained by irradiation of 13. Furthermore trapping reactions of the irradiation product (with cyanobenzene, p-methylcyanobenzene, dimethylacetylenedicarboxylate, fumaronitrile and maleonitrile) are interpreted as involving the cycloaddition of the carbonyl ylide 14 to the dipolarophiles (ref. 10).
198
3.2.2 CvcloDentadienone and Indenone Oxides Stable ylides have been generated also from substrates incorporating aryl-substituted cyclopentadienone or indenone oxides. Photochromism and thermochromism of these compounds have been studied in particular by Ullman and coworkers and reviewed earlier (especially the chemistry of indenone oxide u ) (ref. 1). Therefore, in this section, only the essential points are summarized in the following schemes.
R4
R4
k4
I
0
0
15 16 Rl, R4 = H, alkyl or aryl groups, C02H, C02CH3, CONH2 R2, R3 = H, aryl or the atoms necessary to complete an aromatic ring Ph
Ph
I
I
\\
0
17
18
Evidence for is based on nmr and ir spectra as well as the absence of esr signals and on trapping experiments: The colored species reacts as a 1,3-dipole with dipolarophiles such as dimethylacetylenedicarboxylate or norbornadiene to give the corresponding adducts 19 and a,respectively.
19
0
20
199
~OxvdiDhenylsuccinirnide.DiDhenvlepoxvmaleic Anhvdride Griffin and coworkers prepared a new class of stable cyclic carbonyl ylides (ref. 11). Colors are generated when the bicyclic oxides 21 and a,which incorporate vicinal diary1 oxirane moieties, are irradiated at 77 K in rigid matrices such as 2methyltetrahydrofuran (Amax = 520 nm and 541 nm, respectively). These colors persist after the matrix is warmed up, softens and appears to become fluid (130-140 K). The mesoionic ylides a and 24 are responsible for the photochromic behavior. 3.2.3
I
H
H
21
22b
22a
hv
c--
23
24
Unfortunately color formation is not visually or spectroscopically detectable upon irradiation of a and 23 at ambient temperature in fluid solution in 2-methyltetrahydrofuran. Ylides 22 and 24 are stable also in the solid state, even at room temperature. 4
AZIRIDINES
4.1 Monocvclic Aziridines
The photochromism of arylaziridines was discovered some time ago by Cromwell and coworkers (ref. 12). Padwa and Hamilton (ref. 13) investigated the photochromism in the arylaziridine System 25, 26, 22. Intensely colored solutions are obtained by irradiation of 25 and 26 in glassy solution at 77 K. on warming slowly, the colors fade away and.the aziridine is regenerated. All attempts to generate colored species by carrying out the
200
irradiation in solution fail as well as trapping experiments with an unsaturated system like dimethylacetylenedicarboxylate.
I
H
R
H
25
27
ri 26
Trozzolo and Thap DoMinh presented evidence that the colored species from aziridines are best described in terms of an azomethine ylide structure (ref. 14). It has been found too that the presence of either a phenyl or benzoyl substituent on both ring carbons of the aziridine constitutes the structural requirement for photochromism in the solid state or in a glassy solution (ref. 5). When groups such as nitro and methoxy are introduced into the para-position of the aromatic substituents of aziridine a the Y
x
28
201
absorption of the colored intermediate is red-shifted stability is improved markedly (ref. 5).
and
its
Table 1: hmax and life time of azomethine ylides from
28
X
Y
Amax (nm)
life time
a (cis) H b OMe C H
H H OMe
475 485
T
> several min.
T
> several min.
OMe NO2
NO2 OMe
d e f
9
H NO2
NO2 H
490
485, 520 475 490, 540
T
(room temp.)
r > 2 4 h
600
T
> 15 min.
The p-methoxy group provides considerable stabilization and the nitro group exerts a pronounced effect on the absorption spectrum of the intermediate when it is present in either phenyl group attached to the aziridine ring. The combined effect of these two groups is quite spectacular and leads to systems such and 28s. In 28Q the coloration can be achieved by warming as as well as by irradiation and, in addition, the photothermochromic behavior is observable even in solution.
29f
28f
Huisgen and Mader evaluated thermodynamic and kinetic data of their isomeric azomethinylides (a, (ref. 15).
l-(p-methoxyphenyl)azirine-2,3-dicarboxylates (30, 31) and
u)
202
Irradiation of and 31 with uv light induces disrotatory ring opening to 33 and 22, respectively, whereas the thermal ring closure occurs in a conrotatory manner (ref. 9).
Ar
Ar
4.2 Bicvclic and Tricvclic Aziridines
Bicyclic aziridines were desribed by Heine and coworkers (ref. 16), Thap DoMinh and Trozzolo (refs. 14, 17, 5). Typical of the compounds investigated is 34. The crystals of and its glassy solutions at 77 K upon exposure to light ( A < 450 nm) rapidly develop an intense blue color. The quantum efficiency for color formation in 2-methyltetrahydrofuran at 77 K is estimated to be 0.85. The color can be erased by irradiation in the visible region ( A > 550 rim) or by heat. Ph
34
Ijo, 35
The stability of the colored intermediates is strongly influenced by both electronic and steric changes in the structure
203
of aziridines: 1. Removal of the nitro-group, or shifting it to a meta-position, markedly reduces the photochromic sensitivity of the aziridines and blue-shifts the absorption spectra of the colored species. 2. A second fused ring appears to stabilize the ylide relative to the aziridine. In solution, the lifetime of the colored intermediate is much shorter (several minutes at room temperature) and its formation is not entirely reversible: Irradiation of an argon-purged solution of 34 in benzene at room temperature (10 s ) produces a bright red color (Amax = 565 nm) which fades away after several minutes to yellow. Continued irradiation regenerates the red Ph
#MeT
)GN
I
(room temp.)
Me
Nb2 34
NO2 35
36
Ph
N Me
34
Me
Fh
37
38
204
colored species 35, but after 3 h, the reaction mixture becomes quite yellow and most of the aziridine has reacted to give enediimine a. The nature of the red-colored intermediate was revealed by trapping experiments. Consequently, the photoinduced ring opening appears to involve a conrotatory motion (ref. 17) in contrast to a disrotatory mode described by Huisgen and coworkers for the simpler aziridineazomethine ylide system 0 - 33 (ref. 9). Possibly, the reaction might be a ground state analogous to the one suggested by Ullman and Henderson for the indenone-pyrylium oxide system (ref. 19). Irradiated single crystals of aziridines, including 34, were found to be highly dichroic. The blue monoclinic crystals absorb strongly along one axis and are essentially transparent in the perpendicular direction when observed under a polarizing microscope. The anisotropy presumably reflects a highly stereospecific ring opening of the aziridines (ref. 14). Another class of photochromic aziridines was investigated by Lown and Matsumoto (ref. 19) as well as Padwa and Vega (ref. 20). Ph or
NO
N-C6H11
‘‘6Hll
39
40
4.3 photochromic oxalic Acid Salts of Bicvclic Aziridines
The crystals of the oxalic acid salts of 4 were found to give a reversible photochromic reaction (ref. 5). Upon irradiation, a red coloration is produced. In the dark, this red color lasts for more then three weeks at room temperature. The coloration-erasure cycle can be repeated as many times as desired. When a stream of ammonia gas is passed over the red intermediate, it instantly becomes blue colored to give the azomethine ylide s. This reaction supports the hypothesis that the red color is due to the protonated azomethine ylide.
205 5 NITRONES
The nitrones (Q), considered as azomethine.oxides, are structurally related to azomethine ylides. In contrast to the azomethine ylides, the azomethine oxides are thermodynamically favored as against their cyclic form (ref. 15). The photoisomerization of nitrones has been shown to proceed via the excited singlet leading to oxaziridines (42) which on heating regenerate the nitrones. Unless the oxaziridine is stabilized by appropriate substituents, further reactions (either thermal or photochemical) may take place yielding amides which cannot revert to nitrones (ref. 21).
42
41
The tetracyclic nitrone fi was reported to isomerize quantitatively to an oxaziridine 44 (ref. 22).
A
43
&*%A Me
hv
__t
c--
A
HO
Me
Me
Me
R
e
H 44
The quantum yield of the oxaziridine formation in methanol with monochromatic light ( A = 285 nm) is found to be 0.17. When the irradiated methanolic solution is set aside at room temperature in the dark, oxaziridine 44 gradually reverts to nitrone 43
206
and after 18 hours the transformation is complete. Orbital symmetry control in the nitrone-oxaziridine system has been investigated by Splitter (ref. 23). In contrast to the azomethine-ylide-aziridine system, the stereochemistry of the nitrone photocyclization to oxaziridine can only be observed on carbon and nitrogen atoms because the oxygen atom has lone-pair electrons instead of substituents. For the reverse thermal ring opening of oxaziridine to the nitrone, the two possible conrotatory motions of C-0 bond cleavage should result in a mixture of the cis and trans isomers of the nitrone, in contrast to the photoclosure which should be stereospecific.
45
a
c
46
R~ = ( c H ~ ) ~ ~2 N = C H ~ R1 = NO2
R2 = CH3
c R 1 = H d
R2=
CH3 R2 = CH2CH3
R1 = NO2
Results of photolysis of several nitrones
are shown below:
a) Long time irradiation of trans- and cis 45a-d at -60 OC: trans 45a.b trans 45c.d
hv / -60°C
CD30Dl CDC13 hv / -6OOC CDC13
> trans-oxaziridines 46a.b exclusively
> 69% cis-, 31% transoxaziridines 46c.d
b) Short time irradiation of trans- and cis trans
cis
hv / -6OOC
hv / 25OC
m:
> 37% cis-, 63% trans 45d (photo-stationary state at 15% conversion to oxaziridines) > trans
45d (rapidly)
207
c) Quantum yields: cis trans cis trans
45d 45d
45d 45d
-> trans 45d -> cis 45d -> m -1 -
0 = 0.46
-
0 = 0.18 0 = 0.087 0 = 0.016
The ratio of cis- and trans 45d at the photostationary state is very different from the ratio of cis- and trans formed at the end of the photocyclization reaction. From the quantum yields of the photoreactions the calculated ratio of 46d isomers after the photostationary state is reached is 6 8 % cis and 32% trans. These results indicate stereospecifity in the nitrone photocyclization reaction giving an oxaziridine with the same configuration as the initial nitrone. 6 AZOMETHINIMINES
An interesting example of a photochromic system based on azomethinimines is ffz (which is the thermodynamically more stable form).
k
k
47
a
48
b
C
d
Irradiation of 47a-d with monochromatic light (333-436 nm) in dioxane solutions leads in a clean photoreaction to diaziridines 48a-d. Thermally 48 reverts to 47 (ref. 24). The compounds 48b-d on monochromatic irradiation afford the dipole 47 also in a photochemical reaction. The quantum yield for the 47d to 48d conversion is 0 = 0.21. The backreaction has a quantum yield of the same order. The spectra of the diaziridines reflect the aromatic substituent linked to the system. The difference in the
208
a
maxima of and gs comes up to 100 nm. Thus in optimal conjugation between the aromatic ring and the 1,3-dipole moiety is possible. According to these results, suitably substituted azomethinimines/aziridines may be used as photochromic systems. 7 APPLICATIONS OF PHOTOCHROMIC THREE-MEMBERED HETEROCYCLES
The isomerization of nitrones to oxaziridines stores up to 100 kJ mol”. Therefore, these compounds could find application as energy storage material (ref. 21). Image formation has been claimed possible with aziridines. Exclusion of oxygen by introducing an efficient 02-barrier increases the stability of the color by a factor of about 1000 (ref. 25). Oriented crystals of aziridines and oxiranes are photochromic in an unusual way (ref. 26). The effect of the excitation radiation and the strength of the resulting absorption band responsible for the appearance of color depend on the polarization of the radiation. Articles have been described which utilize this photochemical property of these heterocyclic compounds. Typical articles are - windshields or glass panes which darken on being exposed to glare sunglasses - switching devices which exist in two states; one state characterized by absorption of certain radiation over a wavelength range and the other state by high transparency.
-
8 CONCLUDING REMARKS
The irradiation of three-membered heterocycles produces colored ylide intermediates. The stability of these intermediates can be controlled by stereoelectronic factors as well as by solid phase constraints. Combining these two criteria, interesting mainly in solid matrix - can be made which photochromic systems have been suggested for energy storage and for the classical applications of photochromes.
-
209
REFERENCES 1 2
3 4 5 6 7 8 9
G.H.Brown, Photochromism, Techniques of Photochemistry,
Vo1.3, Wiley-Interscience, New York, 1971. R.B.Woodward, R.Hoffmann, Angew.Chem., 21 (1969) 797; Angew. Chem.Int.Ed.Eng1. , 8 (1969) 781. B.Schilling, J.P.Snyder, J.Am.Chem.Soc., 97 (1975) 4422. see also: R.Huisgen, Angew.Chem., 92 (1980) 979; Angew.Chem. Int.Ed.Engl., 19 (1980) 947. G.W.Griffin, Angew.Chem. 16 (1971) 604; Angew.Chem.Int.Ed. Engl., 10 (1971) 597.
A.M.Trozzolo, A.S.Sarpotdar, T.M.Leslie, Mol.Cryst.Liq. Cryst., 50 (1979) 201 and references therein. N.R.Bertoniere, G.W.Griffin, Photochemistry of Three-Membered Heterocycles, in: 0.L.Chapman (Ed.), Organic Photochemistry, Vo1.3, Marcel Dekker, New York, 1973, p. 115. R.S.Becker, R.O.Bost, J.Kolc, N.R.Bertoniere, R.L.Smith, G.W.Griffin, J.Am.Chem.Soc., 92 (1970) 1302. Thap DoMinh, A.M.Trozzolo, G.W.Griffin, J.Am.Chem.Soc., 92 (1970) 1402.
(a) R.Huisgen, W.Scheer, H.Huber, J.Am.Chem.Soc., 89 (1967) 1753. (b) R.Huisgen, W.Scheer, H.Mader, E.Brunn, Angew.Chem.,
16 (1969) 619; Angew.Chem.Int.Ed.Engl.,
8 (1969) 604.
D.R.Arnold, A.Karnischky, J.Am.Chem.Soc., 92 (1970) 1404. 11 G.W.Griffin, K-Nishiyama, K.Ishikawa, J.Org.Chem., 42 (1977)
10
12 13 14 15 16
180.
(a) N.H.Cromwel1, J.A.Caughlan, J.Am.Chem.Soc., 67 (1945) 2235. (b) N.H.Cromwel1, H.Hoeksema, ibid., 71 (1949) 708. A.Padwa, L.Hamilton, J.Heterocycl.Chem., 4 (1967) 118. Thap DoMinh, A.M.Trozzolo, J.Am.Chem.Soc., 94 (1972) 4046. H.Hermann, R-Huisgen, H.Mader, J.Am.Chem.Soc., 93 (1971) 1779.
H.W.Heine, R.H.Weese, R.A.Cooper, A.J.Durbetaki, J.Org.Chem., 32 (1967) 2708.
17 18 19 20 21 22
Thap DoMinh, A.Trozzolo, J.Am.Chem.Soc., 92 (1970) 6997. E.F.Ullman, W.A.Henderson, J.Am.Chem.Soc., 86 (1966) 5050. J.W.Lown, K.Matsumoto, J.Org.Chem., 36 (1971) 1405. A.Padwa, E.Vega, J.Org.Chem., 40 (1975) 175. T.Laird, Chem.Ind.(London), 1978, 186 and references therein. H.Suginome, T.Mizuguchi, T.Masamune, Bull.Chem.Soc.Japan, 50
23
J.S.Splitter, T.-M.Su, H.Onol M.Calvin, J.Am.Chem.Soc.,
24 25 26
(1977) 987.
(1971), 4075.
93
G.Tomaschewski, G.Geissler, G.Schauer, J.Prakt.Chem., 322 (1980) 623.
Minnesota Mining and Manufacturing Co. (D.L.Fanselow, Inv.), D.O.S. 28 28 965 (January 18, 1979). (a) Bell Telephone Laboratories, Inc. (A.M.Trozzolo, Inv.), US Pat. 3,984,177 (October 5, 1976). (b) Bell Telephone Laboratories, Inc. (A.M.Trozzolo, Inv.), US Pat. 3,964,823 (June 22, 1976). (c) Western Electric Co., Inc., (A.M.Trozzolo, InV.), D.O.S. 25 10 836 (September 25, 1975).
210
Chapter 6
4n+2 Systems Based on 1 5 Electrocyclization
H.Diirr
1 INTRODUCTION
n ReAn electrocyclic reactionis the formationof a crbondbetweenthe terminiof a fullyconjugatedneutralor ionica-systemor its reverseprocess.It is thereforea type of intramolecularcycloaddition.Electrocyclicreactionsare the basisof many photochromic systems.In the followingscheme, a surveyof the electrocyclicreactionsis shown where it has so far been possibleto prepareinterestingphotochromic systems. Scheme 1 ElectrocyclicProcessin 4n educts
0
+ 2 - Systems
hv or A
products
type of reaction
211
In this chapter1,5-electrocyclization will be dealt with. This reactionconstitutesa basis for a new classof photochromics (ref. 1). This processis relatedto the electrocyclic ringclosureof hexatrienesand its reversalgivingriseto a wealthof knownphotochromics (ref.2). These moleculeshave the essentialsystemembeddedin more complex structuressuchas stilbenesor fulgides,bianthronesand spiropyranes(see chapters3, 8, 9 1.
Systemspossessing10 P-electronshave only recentlybeen shownto be photochromic (see chapter 11). For the 1,5-electrocyclization of the pentadienylanion Huisgen(ref. 3) has pointedout that for heteroanaloguesof pentadienylanions,three cases (1, 3 and 5) may be distinguished,substitutingone of the carbonatoms isoelectronically by a heteroatom. Type-I systems(heteroatomin position1) possessa formallycharge-freeresonance formulasuchas 2a. The cyclicstructurecan only be representedby chargedformulas la,b. This is the reasonwhy the more stablespeciesin thiscase is thereforethe openring2. In type-2systemsthe heteroatomis in position2 allowingan unchargedresonance formula3a for the cyclicmoleculeto be drawn,thusfavoringthe latter. In type-3systemsonlythe open-ring6b can be writtenin a nonpolarform, which becomeshere the most probablestructurein the equilibriumshown. Scheme 2
3a
b
4a
b
C
5a
b
6a
b
C
212
Usingthe conceptof 1,5-electrocyclization new photochromicsystemshave been developedas is demonstratedin Table. 1 . No type-I-systemhas been discoveredso far to be photochromic.However,type-2, type9 and the mixedsystemstype-1,2 , type-2,3 and type-I,2,3 clearlypossessphotochromicpropertieswhen substitutedproperly. The differentclassesof photochromicmoleculesare describedin thischapter. Table. 1: Categoriesof 1,5-electrocyclizations in new photochromicsystems . .. .
categoryof 1,5-electrocyclization type-1
type-2
dipolar structure
G-
cyclic structure
Q Q
A*-pyrroline
A'
type3 I
I
a'
-pyrroline
isoxazoline
I
type23
Firstthe spectralpropertiesof colorlessand coloredformsof the photochromic systems are described. The thermaland photochemicalreactionsare treated. Environmentaleffects,synthesis and applicationare mentionedlast. The nomenclaturewill adhere normallyto the IUPAC rules. In mostcases abbreviations are usedto give a shorterdescriptionof the molecules.
213
2 THEORETICALSTUDIES OF I,5-ELECTROCYCLIZATION Usingthe Woodward-Hoffmann formalismfor the conservationof orbitalsymmetry (ref. 4) in pericyclicreactions1,5-electrocyclization can be analyzedas follows: The simplestdescriptionfor the stereochemicalcourseof thisreactioncan be deduced froma simpleorbitalcorrelationdiagram. Fig. 1 showsthe differentorbitalsof the pentadienyl-and the cyclopentadienyl-anions (refs. 5,6).
A
Q
u*-A\
al-A
Q
\ A-
x-
u-s
S-
i-"" Q
-Y x Q -A
S-Ss
4 3
A-Z2
-A
S-$
C
-S
\ s-u
Q
Fig.1: Orbitalcorrelationdiagramfor the conrotatory(photochemical)and disrotatory (thermal) 1,5-electrocyclization of the pentadienylanion.
214
Usingthe frontierorbitalapproachit is obviousthat the photochromic reactionstarting from d * o f the pentadienyl-anion mustbe conrotatory,whereasthe thermalprocess mustbe disrotatory.A simplifiedstate diagramaccordingto Turro (ref. 6 ) givesa visual representationof this rule for 4n + 2-systems.In the type-1to type-3 systemsthe same stereochemicalconsequencesshouldprevail.Thus substituentsin decisivepositions, do play an importantrole in boththe photochemicaland the thermalreactions.
4n+2
I-
s,
pathway
pathway
SO
0 R BITA L S Y M ME T R Y FORBIDDEN GROUND STATE REACTION
0 RB ITA L SYMMETRY ALLOWED GROUND STATE REACTION
Fig. 2: Simplifiedsingletsurfaces(accordingto Turro)(ref.G).
G E N E R AL RULE
215
With the type-2systemssemiempiricalMIND0/3 and MNDO calculationshave been carriedoutfor the groundstate molecules(ref.7). In an extensionto approachthe new photochromic systemsof the [1,8a]-dihydroindolizine type moreclearlya spiroring (7, 8) and a pyridinering(9, 10) have been attachedto the simplepyrroline2.
Scheme3
7
CN
CN
9
The parametersobtainedfor the cyclicstructure10 agree well withthe geometryof compound15 (vide infra)determinedby an X-ray analysis(ref.7). The resultsfor the modelof the coloredform, e.g. 16 are collectedin Fig. 3. They are givenfor the energeticallymost relaxedstructurehavingthe heterocyclicringperpendicularto the plane of the C-I, C-2,C-3unit:
216
-qote
0,038
I
-0p27
\
/
/
y2
-O , l 3 3
\
0,054
*
\ 0,HJ
- 0,201
liiiszl
Fig.3: Resultsof a MIND0/3 calculationof 10 representingthe coloredform 16. 1) The E- and Z-Formof 16 are almostisoenergetic.This meansthat no barriersexist betweenthese forms. E- and Z- isomersare conformers. 2) The bondlengthC-2'3 is half way betweena singleand a doublebond. In other wordsrotationaroundthis bondshouldbe fairlyeasy. The most relaxedstructureof the coloredform has no full conjugationbetweencyclopentadieneand heterocyclicrings. Electrondensitycalculationsshowthat the positivechargein 16 residesmainlyat the Natom. The negativechargeis primarilylocatedat C-3 and to a muchsmaller extent at
c-I.
217 3 PHOTOCHROMISM BASED ON PENTADIENYLANIONS WITH ONE HETEROATOM
In developingphotochromic systemson the basisof a 1,5-electrocyclization, molecules mustbe envisagedinwhichthe open-ringand the cyclicstructurehave a similarenergy content. This problemhas been solvedby the synthesisof moleculesundergoingreversible13electrocyclization. The simplepentadienylanionshowscyclizationneitherthermallynor photochemically (ref. 8a-d). With the carbocyclicsystemsonly irreversibleringclosurein 8-memberedringsor ring openingin 7-memberedringshas been observed,dependingon the substituentsor ringstrainas in 14 (refs. 8 e,9. Scheme4 25OC
11
I2
13
14
3.1. Tvlse-3_S v s m Incorporation of a A2-pyrrolineringintoa more complexmoleculehas led to a very efficientnew classof photochromic compounds(ref. 1). Scheme6 R3
R
15
R
17
16
218
Irradiationof the spiro-[l,8a]dihydroindolizines(DHI) 15 with longwavelengthUV or visiblelightaffordsthe coloredbetaines16 (refs.9-19). Typicalspectraare givenbelow:The mitterions16 can in principleexist in E- or Z-configuration.However,as has been shownby semiempiricalcalculations(videsupra)the energy barrierbetweenE-17and 2-16is very smallso that only one formis observed (vide infra). For 16 and 17 resonanceformulasare possiblewhere bondC-2/C-3 is infacta singlebond. In the new photochromic systems15,16 basedon the ringopeningof a hetereocyclopenteneringand its reversal,the basic process is the breakingor reformingof a 0bond(see scheme6). This reactioncan be inducedphotochemically (or, lessfrequently,thermally).The product 16,17 formedin thisconversioncan only be writtenin a mitterionicform. This is the basicdifferenceto spiropyraneswhichundergoalso a 6T-electrocyclic reaction, yielding, however,an electro-neutralspecies. Furtherthe zwitterion16,17 may exist as Eor Z- isomer,whichaccordingto MIND0/3 results,possessonlyvery slightlydiffering AHf- values. Typicalspectrafor 2 selectedexamplesfor a simpleDHI 15,and an aza-DHI 15 are givenin Figs. 4a,b. Spectraldata are collectedin Table 2. Table 2a: UV-visibledata of Spiro[l,8a]dihydroindolizines(DHI) 15 (cyclopropeneroute a) (see text) Solvent: CHnCldEther(refs. 9,ll) (R' =C02CH3) and half life (t112) at 293 K.
X a b C
-
-
d H,H
e f g
h
i j
-
-
-
k 1 C=O
R CH=CH
CH=CH CH=CH CH-CH CH=CH CH=CH
Y CH CD CH CH CH CCH3
R2
R3
R4
R5
H
H
H
CH=CH H
H D H H H H
H
CH3
H
H
D
H H H H
D OCH3 CH=CH CH=CH H
CH=CH N H H CH-CH CH=CH CH=CH H CH=CH CH=CH CH=CH H CH=CH
Ph
CH=CH
CH CH CH
H H H
H H
H
D H CH-CH
H H
H H
H H
max [nm]
'max
384
586
376 385
552 572
15
383
378
383
389 360
H
363 376
H
388
392
ti
[nml [s1.10-~ 16
586
4.88 3.4
11.0 1.2
600
40
505 724
14 2.7
570
726
694
629
2.21
0.1
832
27.7
219
Table 2b: UV-visible data of DHI 15 (pyrazole route b) (refs. 10, 14) (R=-(CH =CH)-2) at 293 K.
-
m n
-
-
0
-
P 4 r U
CN CN CN CN CN
-
CF3
-b
t
N
-
-a
S
CN
C
CN
- CF3 C=O CN
V
W
C=O CN
X
Y
C=O CN C=O CN
z
S
E S E H,H E H,H E
CH
N
H
H
H
H
CCH3 H
H
CH CH CH CH N CH CH
C02CH3
440 (4.6 lo3)
440 (4.0 lo3) 420 (6.3 lo3)
391 (2.3 lo3) 415 (1.5 104)
N(CH3)2
H H
CH=CH-CH=CH 338 (3.3 lo2) H H 418 (5.2 lo2)
H H
CH=CH-CH=CH 420 (6.8 lo3) N(CH3)Z H
H
H
H
H
H
410 (6.7 lo3)
395 (9.9 103)
H
H
327 (3.4 lo2)
H
392 (8.5 lo3)
C H H H H CH H CH-CH-CH-CH CH=CH-CH=CH H H CH H CH=CH-CH=CH
a) only here R5 = CH3 =
H
CCH3 H CCH3 H CCH3 H CCH3 H CH=CH-CH-CH H H CH H CH=CH-CH-CH
b) stable betaine
E
H
H
376 380 355 376
(9.1 (1.0 (2.1 (4.3
lo4) lo4) lo4) lo3)
560 (3.3 lo4)
535 (3.2 lo4)
13800 3375
570 (1.1 lo4) 113760
565 (1.1 lo4) 194400 560 (1.9 lo4) 16800 673 (1.6 lo3) 597 (9.7 lo3)
280 12
540 (5.2 lo4) stable
685 (6.4 lo3) stable
500 (3.6 lo3) 600 726 (5.5 lo3) 572 (1.3 lo4)
2280
12540 578
220
hu
L
3-
't
16a ~571) /-' I \
I
I
I
I
I
/
I
\
\ \ \ \
/
\
\
200
300
400
500
600
\
\
'. 700
b
h tnml
Fig. 4a: Electronicabsorptionspectraof a simpleDHI 15 (-) and its coloredform 16 (--) 298 K in CH2C12; E 45 = 6.6. lo3, 16 = 36.9 . lo3).
Fig. 4b: Electronicabsorptionspectraof Aza-DHI 15 (-)and its coloredbetaineform 16 (---) 298 K in CH2C12, E = C02CH3.
221
3.2 Reaioselectivitv of 1.5-Electrocvclization
By introducing substituentsintoeither position6 or 8 of the betaine16, the 1,5-electrocyclizationto the DHI 15 can yieldtwo regioisomers.The DHI 15t' (-8) and 15t (-6) are possibleproducts.In contrastto a simpleworkinghypothesis,the more hinderedregioisomer15t (-8) predominatesin the reactionmixture.In the case of R = -0CH3, -CH3, -CI in 16 the 1,5-electrocyclization is even reaiosDecific.
16i
1st ( - 8 )
Fig. 5: Ratio of regioisomersformedon cyclizationof 16, 15.
Isomer r a t i o % R2/4
15t(-8)
OCH3
100
CH3
c1 Ph COzCH3 CN
1 5 t (-6)
100
100 40
58
75
60
42 25
1st
(-6)
222
3.3 Stereoselectivity As has been mentioned(videsupra)the 1,s-electrocyclization of betaine16 to DHI 15
generatesa chiralcenter position8a in 15. Ifringclosureof 16 followsthe rulesof orbitalsymmetryconservationone wouldexpect stereoselectivity due to a ’disrotatoryprocess. In the reaction16, 15 the geometryof 16 is important.Of the variouspossibleconformations,1611.and 1 6 ~are ’ essentialfor cyclization.We made the reasonableassumptionthat conformation1 6 is~the dominatingone (scheme7). 1,5-Electrocyclization of 16Wproducesthe syn-enantiomers15p (1R.8aR) and 15* (1S,8aR) as the predominantproductswhereas 16*’ affordsthe anti-enantiomers 15fi’ in a smallamount,bothformedin a disrotatorymode . An NOE study of the cyclizationproducts15. showsthat the syn- enantiomerssyn-15, are formedpreferentiallythe diastereomericexcess (d.e.) being = 84 - 88%. Carrying out severalcolorationand decolorationcycles,the ratioof diastereomersremainsconstant.The ringclosure16,15 can thus be regardedas beingdiastereoselective(ref. ISa). Scheme7
E=CO,Ch
RZ:
rvnlanti: d.e.
.Ph .. 94 : 8
Sty 9 2 : 8 Sty 94,3:5.7
80 84 80,6
223
The 1,5-electrocyclization of betaines16 to DHI 15 has been provedto be: a) regioselective,b) diastereoselective;the reactionc) showslow AH* and &S* values, d) can be nucleophilic or electrophilicand e) is controlledby the HOMO ( 93). Thus 1,5electrocyclization can be regardedwith highprobabilityas a pericyclicreaction.
Biphotochromic DHl's; a) If two h*-pyrazolinesare presentin one moleculebiphotochromic DHls becomeaccessible.Dependingon structurephotochromism may be observedin moleculessuch as 18' but not in 18. It is not clear if only one ringis openedin 18' or bothto afford 20.
(refs.18,19) b) With two isolatedor directlylinkedsubunits.Separatingthe two chromophoresinto two differentspiro-dihydroindolizine moietiesgivesa versatilefamilyof biphotochromic systems.Here symmetrical(R' = R") as well as unsymmetrical(R' = R") biphotochromicmoleculessuchas 21 can be made (Scheme9). Scheme8
r,hY,
19
18'
FI+ = fluorene
20
224
Table 3: UV-visibledata for 21 and 23 in CH2C12 (refs. 17,20).t112: half life of 23 at roomtemperature).
E Ac E E E E E E E
H H H
396
20.4
630
6.7
396
20.9
628
2.1
H
381
24.6
Me
H H H H
420 399
394
407 404
401
Scheme9
x
22'
21
13.6
575
1.7
623
17.8
15.1
702
4.6
22.3 23.8
590
9.0
14.5
16.7
596
585
650
6.9
2.5 1.3
225
As is shownbelowthe half life of the coloredform can be controlledso that molecules withtailormadepropertiesbecome available.A typicalspectrumand some data are givenin Table 3 and Fig. 6 (refs. 17'20).
Fig.6: Evolutionof uv-spectraduringi,5-electrocyclization of 23 (in CH2C12 at 293 K). c) ModifiedSystems Interrupting the conjugation in the DHI 15 creates a slightlymodifiedsystem,whichhas been termed spiro-tetrahydroindolizine (THI) 24. UV or visiblelightconverts24 into deep blue or bluF blue-greenbetaines25 (refs. 21,22). Scheme 10
E 25
25'
226
The coloredspeciescan exist in two conformersof whichwe favor 25. This is again a very powerful,new photochromic type-2-system;however,here the chromophoreis not a butadienylvinylamine (vide infra)as in 15 but rathera simpleenamine.The colored form is also a conjugatedsystemwhichexplainsthe limitationin color(blue or bluegreen). The maximumwavelengthof the betaine25 is onlyweakly affectedby substitution. TypicalUVNlS spectraare shownfor 24 and 25 in Fig. 7.
E
10000-
- 24 .--.-2 5
5000.
,
.*-.
--. \
0
\
\ \
0
/
\ I
Fig. 7: UV-andVIS-spectraof THI 24 and coloredform betaine25 in ethanolat 218 K. - 3 - _ terns Moleculeshavinga heteroatom in position3 have been studiedextensively.Electrocyclizationof these compoundshas been usedto preparecarbazoles(refs. 23-26), indoles(ref. 23), furanes (ref. 27) and thiophenes(ref. 28). The transformationof suitable precursorsgivesthe productsmentioned.If oxygenis excluded,however,reversible systemsare possiblein severalcases. Irradiationof diphenylamines26 affordsas primaryproducta dihydrocarbazole27 via a tripletstate in a conrotatory1 3 - electrocyclization(see S.W.Staley, ref. 3). 27 absorbsat 610 nm (R =CH3) or 640 nm (R = Et) and its lifetime is very short. Excludingoxygenand loweringtemperatureallowsa photochromic reaction26, 27. In competition withthe thermalringopening,a disproportionation leadsto the tetrahydrocarbazole29 and the carbazole30.The reactionrates 27, 26 and 29 , 30 are equalat 240 K. 30 is also producedirreversiblyif oxygenis present. On furtherirradiation27 (R = Et) affordsthe heterocycle28 (ref. 29).
227
Scheme 11
+ 30
I C”3
28
29
30
An oxygencontainingsystemwhichmay involvean intermediaterelatedto the dihydrocarbazole27 is possiblein the photolysisof the ether 31 (ref.30). This conversionalso proceedsstereoselectively to give onlythe trans-fusedisomer33. Scheme 12
Recentflash photolysisexperimentssubstantiatethe formationof the intermediate32 (ref. 31). The corresponding sulfurcompounds,suchas partiallyhydrogenatedor diaryl-thioethers,show a similarbehavioron irradiation(refs. 28,32).
228
Scheme 13
40
I
Ph 43
41
0
42
The photoreaction of thioethers34 and 37 as well as 40 proceedsvia an excitedtriplet state. The productsof 1,5-electrocyclization 35, 38 and 41, absorbat 590 620 nm. They are convertedin a moderatelyfast reactionwith oxygento the thiophenes36,39 and 42. It is not clear if a reversiblereactionoccursat all to affordthe open-ring speciesagain. The stereochemistry of the lightinduced1,5-electrocyclization is in good agreementwiththe Woodward-Hoffmannrulesfor a conrotatoryprocess(see S.W. Staley, ref. 3). This is borneout by the trappingof the mitterion41 to the adduct43 (ref. 28).
-
229
3.5 SDectra of the Colorlessand ColoredForms
The UV spectraof the spiro[l,8a]dihydroindolizines 15 (DHI) showtwo typicalmaxima one at 250-240the otherat 360-410 nm. The coloredform (betaine), 16, absorbsinthe range500-700 nm. Similarabsorptionrangesare observedfor the moleculesshownin Table 2. The DHls 15 are thereforealmostcolorlessor slightlyyellowishcompounds. The coloredform, however,dependingon substitution,absorbsin almostall regionsof the visiblespectrum. k i a n m e n tof Transitions:The differentabsorptionbandsin the DHI 15 can be assigned to the followingexcitations: The bandat 360-410 nm resultsfroma locallyexcitedv-v*- transition(LE).It is located in the butadienyl-vinyl-amine chromophoreof the DHI 15 (ref. 9,10,14). Accordingto MIND0/3 calculationsthe HOMO of a dicyano-DHI15 showsa similarwave functionas does the LUMO. Thus a LE transition(n-w*)is mostprobablyhiddenunderthis band. The assignmentof the secondbandof 15 is less clear. If the coloredformhas a relatively fast fading rate neitherthe maximanor the extinctioncoefficientcan be recorded correctly.The following proceduresare possiblein principle(ref. 2 ): 1) flashspectroscopy, 2) studiesof the photostationarystate usingside illuminationtechniques, 3) generatingthe coloredspeciese.g. by treatinga salt with base, 4) determinationof Amax (E) in the solidstate (in KBr pellets), 5) usingthe Fischerapproach(ref. 33) and 6) evaluationof spectraof free A or B if only one form fluoresces(ref. 34). Mainlyl), 3), and 5) have been employedto obtaincorrectabsorptionparameters. SDectra of ColoredForm% The assignmentof the betainebandsin 16 is more complicated.The longwavelength band is most probablya CT transition.This has been shownto be the case for the closelyrelatedpyridiniumcyclopentadienylylid studiedby Kosower(ref. 35). In the betaine16, and relatedmolecules,the CT transitionshouldoccurbetweenthe symmetrical MOs of the donor and acceptorregionsof a molecule.This is in goodagreement withrulespublishedby Fabian(ref. 36) for similarcompounds.
3.6SolventFffectson the ColoredForms
In the studyof the coloredforms16 in varioussolventsan interesting effecton the long wavelengthband is observed.Mostof the DHI 15 investigatedshow a negativesolvatochromiceffect. This is consistent withearlier reportson spiropyrans(ref.2). A typical example is givenfor the coloredform 16.
230 ET(301 O r 2 [kcallrnol]
70-
60-
50
40 -
I,
55
50
I1 ’
61
h%T*
65
-.
m hYCT2
[kC.llnlOIl
Fig. 8: Solvatochromic effecton betaine16 (R’ = C02CH3; Z-parameters(--), ET(3o)-values(xx). Usingsolventpolarityas a measurefor the Z-parametersof Kosower(ref. 35) and the ET(3O)-ValUeSof Dimroth(ref. 37) one obtainsa goodcorrelationbetweenthe hypsochromicshiftof 15 and these parameters.The positiveslope of the correlationis good evidencefor the CT characterof the two transitionsin 15 In contrastto these resultsfor ester-betaines16 (R’ = C02CH3) the dicyano-derivatieffect. ves (R’ = CN) show a positivesolvatochromic
.
3.7. ConversionRate of Colorless Formsand V The conversionrate 16 , 15 was determinedfor selectedexamples.This parameteris a measureof how efficientlylightconvertsthe photochromicmaterialA to productB: A +E3.
Fischerhas advanceda simplemethodto evaluatethe conversionrate at two different photostationarystates (see ref. 33). Fischer’sformulais as follows:
1+-
A1
€A2
-n(l-
A2 ) EA2
231
is the conversionrate after irradiationof A with lightof wavelenght~ 2 . EAI, EM and El, E2 are absorbancesof speciesA beforeand after irradiation.EBI and E B are ~ extinctionsof speciesB after irradiationwith AI or x2. and CQ
and n = EBI I E B ~
Usingthis procedurea conversionrate of 94% for dicyano15 (R1 = CN, Y = CH or N) is calculated.For the diester 15 (R1 = COnCH3)theconversionrate dropsto only 52%, in comparisonto fulgideswhere this parameteris 60% (ref. 38).This showsthat the dicyano-DHls15 are very efficientlycolored.
15
The Coloration effi c i wis another importantparameterfor characterizingphotochromics. It is relatedto the quantumyield of a photochromic compound. A comparisonof cyano-and ester- DHls 15 is shownin Fig. 9. (Determinedin CH2C12, Bauschand Lomb Monochromator,HBO 250 lamp at A m a of 15).
232
1
fl bFI CN
2
Fl
4
41
1
2
3
3
4
t (min)
4
E
Fig. 9: Colorationefficiencyof DHI-15 concentrations:1: 6.10” mol/l, 2: 3.10-5 mOl/l, 3 and 4: 9.10“ mol/l in CH2C12, at roomtemperature(Fi = fluorene). As Fig. 9 clearlydemonstrates,the cyano-DHls15 show a highercolorationefficiency than the ester-DHls15. This is consistentwiththe conversionrate for these compounds. 4 THERMAL REACTIONS IN SOLUTION
4.1 Structure - Fadina Rate Relatiom Irradiationof the photochromic DHls 15 givesriseto the coloredforms 16. The thermalbackor fade reaction16 15, i.e. the 1,5-electrocyclization,can be easily studiedkineticallyusingUV spectroscopy.The rate constant k thus determinedcan be convertedby the equation:ti12 = In 2/k (2) to the half livesof 16. Scheme 14 k2
k’
R’
.k-2
\
R’
15
16
16
A
61
82
233
The kineticschemefor analyzingthe thermalfade rate is givenby equation(2). Neglectinga photochemicalback reaction16 ,15 (vide infra)the systemcan be characterizedusingthe rate constantsof the thermalsteps k-I; k2 and k-2. All spectroscopic data (see below)indicatethat the equilibriumBI, B2 is very fast (k-ic k2 and ka) thus in a firstorder analysis,the thermalfade rate k shoulddependonly on k-1. The values obtainedexperimentallyfollow clearlyfirstorder kineticsand the parametersobtained by standardevaluationare shownin Table. 2 (Temp.: 293 K). In the case of regio-or stereoselectivereactionof 16 , 15 as well as in biphotochromic systems23 , 21 the kineticsbecomemorecomplicated(see Table 3; see ref. 17,19,20). The tin = ln2/k.l values in Table 2 can be usedto a roughorder for a structure-fade rate relationship.To analyzethe data in a simpleway, the DHI 15 is dividedintothe regionsA, B and C (ref. 39).
mJblAl
In goingfrom the anthronyl-,tetraphenyl-cyclopentadienyl, indenyl-,fluorenyl-to the diphenyl-group,t1/2 increasessteadily.
-
Substitutionof R‘ = hydrogenby electronattractorssuch as trifluoromethyl-, benzoyl-, acetyl-, methoxy-carbonyland cyano-leadsto increasingtr/2 - values. R’: CN > C02R > COR > CF3 > H
234
ReaionC; Introduction of the dihydroazinemoietyintoregionC beingderivedfrom benzo[c]quinoline, quinolinesor isoquinolines and azines increasesthe t112 valuesconstantly. In goingfrom left to rightthe positivechargein the betaine16 is stablizedmore efficient-
lY*
A
B
C
e2 Hammett-Studies[LFER = LinearFree Fnerav
of t w m a l 1 . 5 - F l e cyclization Mechanisticinsightintothe detailsof the electrocyclization of 16 15 can be gainedby linearfree energyrelationshipstudies(LFER). A seriesof electrondonatingor acceptingsubstituentsare introducedintothe moleculesto be studied. The DHI molecule15 is again dividedintothe 3 regions(videsupra).
ReaionA: Change of substituentin regionA clearlyaffectsthe electrocyclization 16, 15. A positive p-value ( p=0.32 and 0.75) has been foundand discussedin detail (refs. 18,40,41). The positive,-value showsthat 1,5-electrocycIization is faster with acceptorsubstituents in regionA.
&aion 6; A slightlyless goodcorrelationis obtainedwhen substituentswithdifferentelectron demandare introducedintoregion6.A negativerather highp-value(-6.82) is calculated (ref. 42). Substituentsin position2,3 (region6)obviouslyinteractwith a nucleophilic center (see Fig. 10).
235
-4-0*
0.8
U P - 1.0
0.9
Fig. 10: Hammettplotfor variationsin region6 for the 1,5 electrocyclization 16, 15 (log k = f (up); R1 = C02CH3, , COCH3, COPh, CN). maionC; Introduction of differentsubstituentsintoregionC again show linear Hammett-plots havinga positivep-values(16: (R2) = 1,9; 16 (R3): = 0,7). As Fig. 11 demonstratesfromthese data interactionof substituentswith a positive reaction center has to be assumed(ref. 18).
-0.2
0
0.5
Fig. 11: Hammett-correlation for the electrocyclicringclosure16 , 15 (variations:regionC). logk/ko as functionof up (16 R2) and urn (16 R3).
236
4.3 Thermodvnmic Properties-Activation Parameters for 1.5-Electrocvcli7atiorl Carryingout kineticstudiesat differenttemperaturespermitsevaluationof the activation parametersfor the 1,5-electrocyclization 16 ,15 to be determinedwiththe aid of the Eyringequation.Applyingthis standardtechnique,AG*and AS*were derivedand are compiledfor typicalexamplesin Table 4. The values of AG* vary between20-30kcal/mol.AH* is structure-dependent and showsa gradualchange, AS*values are negativeand characteristicfor an electrocyclicprocesswith highstericalrequirements.
Table 4: Activationparametersfor the cyclizationof betaines16 to DHls 15 (in CH2C12). AG*is givenat 293 K; e.u.= cat. mof' K'; E = C02CH3.
&*
DHI 16
@
8 F CN CN
CN CN
&*
M*
rkcal/moll
re.u. 1
=f
rsi
20,l
15,9
-14,2
14,2
23,4
17,8
-18,6
16800
22,4
15,3
-22,o
3375
31,6
23,2
-24,g
113760
5 PHOTOREACTIONS/DEACTIVATION CHANNELSOF TYPE-2-SYSTEM To obtaina deeper insightintothe photochemicalreactionDHI 15 to betaine16, it is necessaryto determinefirstlyits multiplicityand secondlyits kinetics.To do this, however, a quantitativeevaluationof the competingphotophysical processis needed, and the absorptionand emissionspectramustbe known.Table 5 givesa summaryof experimental data for selectedDHI derivatives15 (ref 43).
237
Table 5: Emissiondata and singletenergiesof DHI 15 (E = C02CH3).
1
495
3 60
67,6
62,2
3
460
410
65,O
64,O
520
450
2
4 54
4
508
5
6
7
8
450
10
497
9
67,6
68,2
67,6
66,5
340
(6517)~ 65,8
330
462
59,3
58,9
335
493
60,6
63,6
400
460
67,l
66,6
370
65,8
70,2
66,l
66,5
66,5
66,5
395 390
O
1
2
3
6
7
8
4
l
5
Br
CI
9
10
b fluorescencewavelenghtmaximumat 25OC (in CH2C12) c kX(max), excitationwavelength d fromemissionspectra e fromabsorptionspectra f at 77 K Fromthe UV spectra(see Tables 2 and 3) one can furtherderivethe naturallifetimeto and from the fluorescencespectra,the fluorescencequantumyield @F.
238
5.1 LuminescenceProDerties As Table 5 shows,the DHI derivatives15 can be dividedintothree groups: a) those with a spiro bridgecontaining6 carbonatomsand ester or cyanosubstituents at positions2 and 3 (R' = E,CN) b) thosewith a spirobridgeof 5 carbonatomsand CN substituentsat positions2 and 3. c) thosewith a spirobridgeof 5 carbonatomsand CN substituentsat positions2 and 3. On goingfromgroupB to groupA the UV absorptionand fluorescencemaxima undergobathochromicshifts.The fluorescencequantumyieldsare very small ( @F = 109. The singletenergiesfor groupsB and C are 65-68 kcal/mol,whereasgroupA showsESI valuesof 59-63 kcal/mol(ref. 43). The exact excitationof the DHI derivatives15, made up of orthogonalstructuralunits,is likelyto fall withinregionC.
Pariser-Parr-Poole calculationssupport thisassignment.It appears,therefore,that the effectivechromophorein 15 may be the butadienyl-vinyl unit. Intramolecularchargetransferexcitationsappear less likely,and the 'Ag state from whichemissionoccursis not identicalwiththe state by absorption. Fluorescencequantumyieldswere determinedby generaltechniques(ref.37) usingdiacetylor rhodamineas standards.The differentvalues givenin Table 7 are discussed below. -was seen in only one case (ref. 40). A clear emissionat ratherlongwavelengthcan be observedat 615 nm (77 K), whichis due to a triplet,the nitro-betaine16 beingthe emittingspecies.This is remarkablesince neither fluorescenceor phosphorescence of betaine16 has been detectedso far. This resultreceivesattentionsince6-nitrobips showsphosphorescenceat 620 nm. The nitro-groupbeingobviouslyimportantfor phosphorescence.
Quenchingexperimentsare employedto distinguishwhether singletor triplet is responsiblefor a photochemicalreaction.If one triesto quenchthe tripletstate of 15 by using knownquenchingagents, suchas diacetyl,anthraceneor oxygen,no quenchingof the reactionis foundto occurfor the conversion15, 16. When tetramethyl-l,2-dioxetane is decomposedthermally,tripletacetone is formedas a chemicallyinducedspecies(refs. 44,451). If thisexperimentis carriedout in the presenceof 15, one does notfind 16. Fromthese experimentsone mustconcludethat the excitedstate is a singlet(Or a fast reactingtriplet)whichdoes not undergobimolecularquenching(ref.43).
239
5.3 QuantumYield of Photocoloration and Photobleachinq The determinationof the quantumyieldfor ringopeningof 15 is not a trivialproblem. The generalschemecan be formulatedon the principlethat the reactionis not only thermallybut also photochemically reversible. This meansthat the reversephotochemicalreaction(rate constantk9-1) is superimposed on the thermalback reaction(rate constantk-1). The analysismust be carriedout usingnon-linearoptimizationmethods(s.1.c. approximationref.46).
<
*
hvjl kl
A
k:,
1''
hy/k'-,
A
k-1
62
k-2
Table 6. Quantumyieldsfor photochemicalringopeningof 15 (A) to 16 (6) (ref. 46); (R=CH=CH, R'=CN, R2-R4=H) R3
I0
A
k2
B
01A 3,6'10-11
4,5'10-11 €3
,0-10-10
8,0'10-10
3,8'10-4
. 1,7'10-4
2 I 1-10-4
2,1'10-4
incl 02B
1,32
OIA
. k' -1 a)
0,84+0,11
0,94
02B
a)
0,05 -10,12
1,11
0,02
0,81+0,09
0,015+0,11
0,85
0,003
0,86+0,13
O,O04kOl16
0,€36
0,87+0,09
0,002
0,002-10,lO
a) Corrected quantum yields for forward- (alA) and back-reaction
(02B)
15
,
16; k,l
2
0.
Table 6 liststhe exact valuesfor @A and CPB.Approximatedata for @R = @A neglecting the reversethermaland photochemicalreactionsdo not deviatedramaticallyfromthe exact parameters(see Table 6). The quantumyieldsare in some cases slightlywavelengthdependentwhereas in othersthey remainratherconstant(ref.40, see Fig. 12). No simpleexplanationhas so far been foundfor these facts.
240
I
A
1 350
4d0
B
\
460
ACnml
C
Fig. 12: Wavelengthdependenceof photoreaction 15, 16.
An importantpointis the fact that the CN-betainecan be reconvertedto the DHI 15 by light.These systemsare thereforephotochemically reversible.This may be usedin applications.
5.4 Kineticsof Photoreactions
In orderto evaluatethe completekineticsof a photoreaction a crucialparameteris the lifetime of the excitedstate. Sincea tripletstate has not been detectedin the photoreactionof type-2-systems15, 16 (one exception),it is assumendthat + ISC = 0. Thus the importantstate remainingfor the photoreaction15, 16 is the singlet.Similarfindings have been obtainedfor the bichromophoric DHl 21. Fromthe standardequations(refs. 46, 47) the naturaland the experimentalsingletlifetimescan be derived.They are connectedby the equation:
241
1 7s =
kF
+ kR + kic + kisc
TO = TO
natural Ilfetirne, = l/kF
TS
= experimentallifetime
where (PF, (PR,(Pic, and (Piw are the quantumyieldsfor fluorescence,reaction,internal conversionand intersystemcrossing,respectively,and kF, kisc, kic and kr the rate constants. SingletlifetimesT~ for 15 and 21 were measuredusinga) singlephotoncountingor b) phasefluorimetry.For the followingcompoundsthe Ts values weredeterminedand are collectedin Table 7 (ref. 48). Table 7: Photokineticdata of DHI 15
1
1,94
2
0,68
-
3
0,21
4
0,42
5
0,50
-
6
Q,15
1,O
7
0,43
2,3
-
0,43
2,2O.1Oa
0,67
3,41.108
0,35
5,12.10a
0,65
9,61.10a
3,s
1 ~ 4 . 1 0 7 0,63
3,00.109
0,36
1 ~ 4 . 1 0 ~
2,2
5,24.106
0,5a
1,3a.109
0,41
9,76.10a
0,80
1,63.109
0,20
4,04.108
3,12.106
0,7Q
4,67.109
0,29
1,93.109
5,40.106
0,56
1,30.109
0,43
1,00.109
-
242
5.5 Mechanismof Photocoloration I aser flash ohotolysisstudiesof the kineticsof the process15, 16 indicatethat, followingthe photoexcitation, the betaine16 appears immediatelyin the Z-form (ref.49). However,flash spectroscopic experimentson dibenzoylspirodihydroindolizine15 suggestthe formationof an isomerwhichundergoesa firstorder conversionto the Zform (ref. 49). The stereochemicalcourseof the electrocyclicringopening15, 16 has been demonstrated to be a disrotatoryprocess(ref.40). An interestingpointis also that the systemDHI-betaineis photoreversible.If regionB of the moleculecarriesCN groupsthe photoreaction can be effectedbothways.
lAHR I
DHI 15n
CN
15n
Betame 16n
Fig. 13: Jablonskidiagramof 15b Fig. 13 showsthe Jablonski-diagram of the photochemicalconversionof DHI, 15, to betaine, 16. The rate constantkr is of the orderof 0,3 - 2,6.109 s-’ in the whole series 15. In other wordsthe reactionof the excitedstate 15*(S1) is very fast. Internalconversioncompetes very efficientlyfrom S1 of 15, whereasQF is rather small. The relativerate constantsfor the photochemically inducedelectrocyclicringopeningof 15 are shownin figure13 (for a selectedcase) and Table 7 (ref. 48).
243
Scheme 15 RegionA:
kr[lO*/s]: 9,61
<
10,o ;
4,o
<
19,3
Region6:
\
/ kr[108/s]:17,4
/
CN
CN >
RegionC:
Fromthese data the followingrulesmay be derived,separatingagain the DHI molecule 15 intothree regionsA,B and C(vide supra). ReaionA: The spiro-bridgealmostdoes scarcelyaffectkr. Electronattractingsubstituentsin regionA make the ringopening15 , 16 more easily. Reaion 6: Electronacceptorsin the positions2 and 3 of 15 increasethe reactionrate.
244
Reaion C ; More electron-deficient bases in regionC favor electrocyclicringopeningof 15. Only benzoannulationin regionC retardskr.
5.6 Mechanism of Fatiaue/ Photostab ility When the spirodihydroindolizines, 15, are irradiatedwith lightit is foundthat the compoundsdecomposeafter sometime. However, if is excludedby workingwith degassedsamplesthese systemsare noticeably morestable. It is possiblethat in the presenceof oxygenthe betaine16 acts as a sensitizertowardssingletoxygen.We have triedto clarifythisproblemby experimentsin whichsingletoxygenwas deliberately generated,with hematoporphyrin as a sensitizer. Scheme 16
B
R
= HC:CH
C
245
Underthese conditionsthe reactionof '02 with 15,16 gave a base-freeproduct,which inthe case of the bis-ester15, was the methylfluorenylideneoxalylacetate A; and inthe case of the the bis-cyanocompoundwas the cyanoaceticacid derivativeA'. It is possible that the first stage is a photochemicalringopeningof the spirodihydroindolizine derivative15, with an electrocyclicreactiongivinga betaine16. Oxygencouldadd on to the betaine16 in a (2 + 2) cycloaddition,withformationof the dioxetaneB. Ring openingand eliminationof pyridine-N-oxideshouldthen affordthe productsA or A'. However,a possibilitywhichnowappears more likelyis the directactionof ' 0 2 on the betaine16 to form a peroxide30,whichthen decomposesin the mannerindicated, givingthe appropriateproduct28 or 28' (ref. 50). Recentlyit has been shownthat betaines16 can act as sensitizersthus producing'02. The singletoxygenwas trappedin thisexperimentby 9,lO-dimethylanthracene (ref. 51). A comparisonof the efficienciesof the new photochromic DHls 15 with knownphotochromics, showsthat the DHI 15 (oxygenexcluded)can be comparedto the well knownsalicylidenanils and the Aberochromes(see: Chapter 9 and 17). 6 PHOTOCHROMIC SYSTEMS BASED ON PENTADIENYLANIONS WITH TWO
HETEROATOMS Accordingto the definitiongivenin Table 1 pentadienylanionsmay containmorethan one heteroatom. Thusvarioussystemsbecomepossible(only N as heteroatomsand vicinalarrangementare shown).
293
394
495
a) Numbers1,2 etc. where chosenfor systematicreasonsand are not in accordwith IUPAC rules. Fromthese moleculesthe type-l,2 and -2,3 have provedto be relevantin contextwith so far.For the azoderivatestype-3,4, photochemicalextrusionof N2 is photochromisrn possible;thishas been widelyusedin makingcyclopropanes(refs. 5234). 6.1. TvDe-1.2-Svstems
Startingfrom a pyrazolinemoleculeof type 44 irradiationyieldsbetaine46. The colored species46 has only a shorthalf life undergoingvery fast electrocyclization to 44. The non-formation of isomer45 in the pericyclicprocessindicatesa regiospecificreaction. The reasonis easy to understand:benzoannulation forces46 to cyclizeselectivelyto 44 (refs. 55,56).
246
Scheme 17
A l3
r;1 RIA A
n
n
\
The colorlessformabsorbsbetween399 and 470 nm, the coloredformsshow absorptionbetween570 and 750 nm. Most remarkableis the longwavelengthabsorptionof the coloredspecies (750nm) whichshouldbe quite interestingfor potentialapplications. The lifetimeof the coloredspeciescomparedwith DHI 15 is, however, in the millisecondto secondtime scale. It couldonly be determinedusingflashspectroscopy.
4
Fig. 14: UV and VIS data of pyrazolines44 and-colored forms46 (in CH2C12, at room temperature).
247
Table 8: UV- and thermodynamicalparametersfor 44 and the conversion46, 44 in CH2CC. (R-R =fluorene, aR =C6H5) A max 44
44
R1
a
C02Me
405
(4.05)
C
COPh
413
(3.75)
b
d
e f
[nml (logE)
COMe
435
COMea
C02Meb CN
432
399 470
46
AH*
[nml
[kcal/mol]
660
15.2
0,73
475
16.2
0.224
15.3
(3.98)
750
(3.98)
540.685 730
(3.96) (3.94)
a) R1=COMe I R1 '=C02Me
tl/2 [e-u.1 [Sl
As*
13.3
-
382.605
15.0
0.02
0.077
-1.85
0.006
-3.38
-
1.2
-1.0
-
-7.94
b) R1=C02Me I R1 '=COMe
Thus we are dealingwith a very fast revertingsystem.The AH*and AS*-valuesfor the 1,5-electrocyclization of 46 to 44 are rathersmall, in goodagreementwiththe fast cyclization process. The betaine46 does not exist in a planar geometrybuta rathertiltedconformation whichseems to be the most relaxedstructure(ref. 56). A related 1,2-systemwhichmightshow reversibleelectrocyclicreactionsis the isoxazoline molecule47. However,here irradiation.doesnot resultin ringopening,in contrastto moderatewarming.The reaction47, 48 is the basisfor a thermochromicsystem;at slightlyhighertemperaturesit rearrangesirreversiblyto the oxazoline50 (ref. 57). Scheme 18
8
48
47
49
50
The mechanismof the photoreaction of type-l,2-systemshas not been studiedso far.
248
8.2.Tvee-2.3-Svstem~ Substitutionof two carbonsin position2,3 of the pentadienylanioncreatesa type-2,3system. Such moleculesare the dihydropyrazolo-pyridines (DHPP) 51 (refs. 5839). Scheme 19
R3
hr
Irradiating51 formsthe coloredspecies52, which(vide supra) can undergoa thermal 1,5-electrocyclization to 51. In unsymmetrical moleculesregioselectiveringclosuremay occur. Two typicalexamplesare represented: in case (a) stericreasonsmay be determiningwhereasin case (b) the avoided destructionof the aromaticsystemis the responsiblefor regioselectivii. A typicalUV spectrumof the colorchangesin the DHPP 51 in reactingto 52 is representedin Fig. 15. Scheme 20
5 1’
52
51
249
E (Absorbance)
.-.
' I
Q5 -
I
I
I
'\
\
I
\
, $
I
\\\
,
b \
\
\
\
\
\
\ \
\
\
\-
I
500
400
300
[nml
6oo
Fig. 15: UVWlS spectraof DHPP 51a and coloredform52b (in CH2C12 at 298 K). Table 9: DetailedUV-dataof substitutedDHPP 51 and their coloredforms52 (in CH2C12 at roomtemperature)
a b
E
f
E E E E E
g
E
C
i j k 1 m n
E E E E E P h CN
C C C
C
d
e
E
h
0 ~~~
a) Amax
H H H H H H
C H C H C Me C
H
C H C H C
H Ph H Me CN Bz
H
C
H H H H
H H H H
C
H
H
H H H H H
H H Benzo Benzo H H B H Benzo Benzo H H H H
Benzo H H Benzo Naphth
H N Benzo C H H C H H C K H ~
H
~~~
~~
-
H H
H
51
401
C
C
C C C C
C
C
N C N C C
~
in nm, b) low temp.,
406 409 399
406 414
365
3,95
3,96 3,90
3,96
3,96 3,81
384
3,80 4,19
400
4,18
394
400 388 437 366 ?
396
4,04
4,17 3,95 4,OO
3,98 ?
3,83
red violet red red violet violet
52
511
547 504 511 527
556
-
-
-
282 1443 367
-
165
-
-
135
594 <5 blue 0.8 purple 540 586 - <5 blue 586 - <5 blue red 502 - 1.1 lo5 violet 529 - 60 purple 534 <5 redb 54OC <1 red 498 4,5 3.8 lo5
-
~
c) them.
rsi
320
52 +
51 fast,
250
FromTable 9 it is obviousthat the colorlessspecies51 absorbsin the region390 - 410 nm. The chromophore,beinga hetero-butadienyl anion, is mostprobablyresponsible for a .rrP*-excitation of the longwavelengthband. Relatedto the photochromic 1,2-system,44, 52 are similarspirocompoundscontaining a heterocyclicspirobridgeand thus belongto the class 2,3-systems. Scheme 21
54
53
R
I)
h
GF 56
55
In contrastto the photochromics 51/52 (R = H or Ph) compound53 is ratherunstable and does not give 54. Compounds55 on the other hand rearrangethermallyto 56. The photochemistry of these moleculeshas notyet been studied(ref. 60).A reversiblephotochromic2,3-systemwas reportedin 1979 (ref. 61). Scheme22
57
58
59
251
The pyrazoline57 is photochemically converted(3h) quantitativelyinto58. The reaction proceedsmostprobablyvia two consecutive1,5-electrocyclizations throughan intermediate 59. On standingseveraldays 58 undergoesquantitativeelectrocyclization to 57. No furtherdetailsare givenin the paper (ref. 61).
B.2.1.Kineticsof the Thermal Back Reaction52, 51 The thermal1,5-electrocyclization of betaines52 to DHPPs 51 has been investigatedby UV spectroscopy.The conversionis a clear firstorder reaction.This provesthat the differentconformersof betaine52 mustequilibratefast and that the slow reactionis the bondformingstep to afford51. Plottingthe k-valuesfor the ringclosureversus Hammett-parameters resultsin a linear correlationwith a positivep-value= 0,532. This is goodevidencethat the 1,5-electrocyclization 52, 51 is favoredby electronattracting substituents.Selectedk-valuesand the Hammett-plotare shownin and Fig. 16.
40
t
Fig.16: Hammett-plotfor the cyclizationof 52 (R3= H, Ph, CH3, CN) in CH2C12 at room temperature;the equationfor the correlationis: lg (k/kH) = 0.5232 IJP + 0.031
252
8.2.2Photoreactionsof TvDe 3 3-Svstem
The DHPP 51 possessa ratherweak fluorescencewith @F lo9 in the region507 - 614 nm. This compareswell withthe DHI 15 havingan equallylow quantumyieldof fluorescence (see Table 10). Singlephotoncountingmeasurementswere carriedout to accessthe singletlifetime (ref. 65). Table 10: Emissiondata of 51 in CH2C12 at 297 K (E=C02CH3).
DHPP
Z
a h e
C
C C
R1
R2'
R2
R3
E
H H H
H H H
H H Benzo CN H
E E
R4
R5
Y
H H H
C C C
A
abs
[ml
401 394 437
A
em
b X %.lo3
507
106 202
[ml
596
614
[nml
177
1.6 2.5 7.6
In contrastto the findingswith DHI 15 for 51 a biexponentialdecay of the fluorescence intensityresults. The correlationof thesefactsgave two singletlifetimesof the order of =3 ns and 72 =8-9 ns. The excitedstate of the photoreaction of DHPP 51 to betaine52 is mostprobablya singlet(vide supra). 1) Groundand excitedstate hypersurfacesshowenergy minimaresultingfrom conformationalchangesand electrocyclicprocesses. 2) The moststableconformationshouldbe similiarto the one calculatedby MIND0/3 for DHI 15 (ref. 7).
Scheme 23
A
B
C
253
The groundstate conformationsfor the betaine52 are B and C. A corresponds to cyclicDHPP 51 . 1) The excitedstate hypersurfacesshouldbe rathersimilar.Photochemical excitationof 51 shouldproduceA* affordingB* in a fast reaction.It is not clear if B* is the betaine 52 in its excitedstate or only a conformerof A*. BothstatesA* and B* can emit, thusexplainingthe biexponentialdecay of the fluorescence intensity. 2) Anotherexplanationis the formationof B directlyin a higherelectronicalor vibrational excitedstate. Similarresultsand argumentshave been advancedfor spiro-pyranes and spiro-oxazines(refs.62-64). The ic step B*-,B leads again to the groundstate hypersurface.From B the most G Sn\
\A
C A
7 PHOTOCHROMIC SYSTEMS BASED ON PENTADIENYL ANIONS WiTH THREE
HETEROATOMS Substitutionof 3 carbonatomswith heteroatomsin the pentadienylanionleadsto the followingstructures(onlyvicinalN- substitution is considered): The resultingheterocycleis a triazolinering. Photochromicmoleculescontainingthis structuralelementhave been made only in two cases (refs. 59,65).
254
62 Here again the 1,5-electrocyclization of betaine61 is only possibleto 60. A formationof 62 is energeticallyunfavorablebecauseof destructionof the aromaticsystem. The UVNlS spectraof 60 and 61 are at 400 and 586 nm, respectively,with half livesof the coloredspeciesof ‘61 of t1/2 = 60 s. The thermalcyclizationof the betaine61 to 62 is unfavorablein the diazasystem52 , 51. Fluorescenceis observedat 614 nm with a ratherlow quantumyieldof QF = 7.6 . lo3, showinga similarvalue those observedfor DHI 15 and DHPP 51. The photochemistryof this new photochromic systemhas not been studiedin more detail. A reversible1,2,3-systemhas been detectedat lowtemperature(N2 matrix)with a sydnone.Thus illuminationof 4-phenyl-l,3,2- oxathiazolylio-5-oxide 63 in EPA (85 K, 420 nm) clearlyshowselectrocyclicringopeningto the ketene 64 and photochemical cleavageto benzonitrilevia 65. Scheme25 Ph
hw
63
Ph 64
255
In an N2 matrix(1OK) isolated63 givesalmostexclusivelyketene 64,whereas aggregates of 63 resultmainlyin fragmentation.The thermalreversereactionhas not been studiedmoreclosely(ref. 66).
8 ENVIRONMENTALEFFECTS AND APPLICATION moleculessuch as DHI 15 and relatedcompoundsare The propertiesof photochromic dependentnot only on structurebut to a large extent on their environment.Clear differencesare to be expectedif solutionor solidphasesof photochromic moleculesare comparedwith a) the absorbedstate, b) liquidcrystallinephasesor c) polymerscontainingphotochromic systems.It is necessaryto use the appropriatespectralmethodsfor the detectionof differentproperties.In additionto normalmethodsphotoacoustic spectroscopy(PAS), polarizedlightand othersmay be applied.
8.1 PhotochromicSvstemsin the AdsorbedState
of DHI 15 is normallyobservedin solutiononly. Ester-DHI-15does not Photochromism show photochromismin the solidstate. HoweverPAS studiesindicatethat cyano-DHls 15 are photochromic in solutionand in the solidstate. The DHI maximumis shiftedbathochromically.In additionin the solidstate photoreversiblephotochromism is obseived. If cyano-DHI-15is adsorbedon Si02 the systemis stillphotochromic. The spectrumof CN-DHI 15 on SiO2 compareswell withthat of the solidphase (ref.67).
3 3 PhotochromicSvstems in LiguidCrvsblline Phase If the photochromic moleculesare especiallytailoredto containlongalkylgroupsthey
may be orientedin suitablesolventssuch as liquidcrystals.In DHI-15 alkylchainsmay be introducedin regionA, B or C. Appropriatesubstitutionof regionB and C has been achievedin the moleculesindicatedbelow.
256
Scheme 26
Dissolvingthese moleculesin the liquidcrystalsolvent(ZLI-2452: phase)was followed by orientingthem betweenappropriateglass plates. Polarizedlightshoweda different effectin that the photochromisrn of the DHI 15 dependedon the dominatingorientation. As Fig. 18 demonstrates,the electrocyclicringopeningof 15 to betaine16 is favored when the lightis polarizedparallelto the preferentialmode of orientation(ref. 68).
=el
Ell for DHI 15 is 554.This is a very As is evidentfrom Fig. 18 the dichroiticratioD large value. SimilarcompoundshavingC7 alkylchainsshowa smalleffect. The optical order parameterS value is 0,34 whichis about50% of knownsystems(ref.69).
This phenomenonmay be usedin a guest hostdisplayin electroopticaldisplays.There an electricalfieldis usedto effectthe changesin the display(refs. 70,71,72) (see also chap. 26). Dependingon the liquidcrystallinephase employedthe other parametersof DHI 15 suchas ti12 and the activationparametersare different.The structureof the liquidcrystalsas hostsis the determiningfactor.
257
15
16
DHI----
Betaine
-
Fig 18: Dichroiticenvironmentaleffects in liquidcrystallinephases (ZLI-2452) ; parallel: a ; perpendicular: b
.
8.3 PhotochromicSvstemsin Polvmers
Photochromicmoleculesmay be a) dissolvedin, b) adsorbedon, or c) boundto a polymer.Dependingon these three differentenvironments,the propertiesof photochromicmoleculesmay change in a very subtleway. a) dissolvedin Polvmers The easiesttechniqueto establisha polymer/PC(photochromicmolecule)interactionis dissolvingthe photochromic in a polymersolutionfromwhichthe solventis evaporated afterwards.DHI 15 have been incorporated by thistechniqueintopolymethyl-or poly-(nbutyl)-methacrylate,vinylidenechloride-acrylonitrile (SARAN F), polycarbonate,polystyrenehbutadiene(Panarez). Two resultsare obtainedin this way: a) the half life of the coloredformcan be increased and b) the numberof cyclesis enormouslyaugmented.In the rigidpolymermatrix the coloredbetaines16 are less mobileto cyclize. Kineticsof thiscyclizationare frequentlymore complicatedthan in solution(ref. 73). The numberof cyclesin DHI 15 increasessinceoxygenis excludedwhich, after activationto ‘02, causesdestructionof the coloredbetaine16 (ref. 74). The numberof cyclesmeasurede.g. in SARAN exceeds600. Similarmatrixeffectshave been observedwith spiropyrans(ref. 75).
258
In a polyvinyl-butyral (TrosifolR)matrixthe numberof cyclesfor the DHI 15, 16 exceeds 5000 (ref. 18). Incorporation of DHI 15 intoa polymericmatrixand applicationof an electricalfield allowsDHI 15 to be oriented.This effectwas studiedwith longalkylchainscontaining DHI 15. The polymericmatricesobtainedby thistechniqueshowednonlinear behavior with LASER light.So 15 may be usedas frequencydoubler(ref. 76 see for comparison: refs. 77-79).
; A very easy PC/polymerinteraction may be studiedif the photochromic materialis adsorbedon the surfaceof a suitablepolymer.A polymerof this kindis the transparent CR 39 (diethyleneglycol-bis-alkyl-carbonate;ref. 78). It is used in industryin lightfilters suchas opticalglasses. The DHl‘s 15 havingappropriatesubstituentsis adsorbedin many cases on the CR 39 surface.The half livesof the coloredbetaines16 are thus normallyincreasedby a factor of 10. Dependingon the substitution patternof DHl’s 15, the photostabilities in adsorbedstate differsgreatly.This parameteris correlateddirectlyto the numberof cyclesfor the photochromicmolecule.The higherthe photostability(irradiation:Xe highpressurelamp, 140 000 LUX) the largerthe numberof cycles. A typicalfatiguecurvedependingon irradiationtime is given in Fig. 19.
I
20
-I ,
\
\2
,
,
1‘-~
2 3 4 5 6 [h]* Fig. 19 Typicalfatiguecurvesfor DHI adsorbedon CR 39;1: R215 = CH3, R2” = H; 3: I?” = CH3; * Durationof Xenon lamp irradiation. 1
f?*4 =
H; 2:
259 As one sees clearlyfrom Fig. 19 the photostabilityof the new photochromic molecules
based on a 1,5-electrocyclization decreasesin the followingorder, the THIS beingthe moststable molecules. The THI 24 and the DHI 15 are morefatigueresistentthan the classicalspiropyranes (ref. 2). The most stablecompoundson the markettoday are, however,spiroxazines (refs. 79,80).
c) in Dolvmer Specialeffectsare envisagedwhen photochromic moleculesare linkedto the polymer backbone.Such a materialwouldbe extremelyinterestingfor molecularinformation storagedevices(refs. 72, 81, 82). Firststeps in thisfield were made with compound15 in an attemptto copolymerizea monomericDHI 15 with methylmethacrylate (MMA). This experimentfailedCompletely. Scheme27 CH-CH-E
8
C PM
E
E
E= Co2CH3
CPM= MMA =Copolymer see text
FI
=;g
260 ~~
~
~~~
UV-Absorption A m[nm] 15 16
[min]
Kinetics A H* A G* [kcal/mol]
(e.u.)cal.K-’mol-’
527
7,4
11,35
20,48
-31,2
554
2,7
17,98
19,W
-6,6
t1l2
A
s*
Polymer: 390 Monomer: 378
a) Solvent: roomtemperature. Reacting,however,a poly-pyridazine withspirocyclopropene a polymericmaterial(PN15) with photochromicpendingside chainswas isolated. This materialwas clearlyphotochromic. The UV maximaof CPM-15 - and betaine CPM-16 did not changeto a largerextentcomparedwiththe monomers15 and 16. However,the half livesincreasedby a factor of 3, whichcan be explainedby the changein activationparameters. Environmentaleffectscan thus be employedto changethe propertiesof new photochromicaza-pentadienesystems.Structuraland environmentaleffectscan be usedto * . ta&xthe molecular.rameteCSto a specificvalue desirablefor. 9 SYNTHESES OF PHOTOCHROMIC MOLECULES BASED ON THE 1,5-ELECTROCYCLlC REACTION In developingphotochromicsystemson the basisof a 1,5-eIectrocyclization and its reverse, moleculesmust be envisagedin whichthe open ring as well as the cyclic structurehave similarenergycontents.This is the problemof the synthesisof molecules withtailormadepropertieswith regardto photochromism.
1s y m Photochromicexamplesare not knownso far.
3 svstem; Type-2 systemsare the basisof a new type of very potentphotochromics. Several routesto this systemhave been developedrecently.
261
CvcloprQge ne-Route (& Reactionsof the easily preparedspirocyclopropenes 63 (ref. 9,11,83) withazines suchas pyridines,pyridazines,quinolines andisoquinolines and others affordsspiro(1,8a) dihydroindolizines (DHI) 15 in goodyields(50-84 %). The intermediate in this synthesisis the coloredbetaine 16. Typicalexampleshave been made via this route (refs. 9,11,12,13). Scheme 28
x&fRR,
R
R
R R'
+ = # ~ I p ; ~ ]
65
R
R
66
R
R 67
Pvrazoleroutefbk In some cases when stronglyelectronattractinggroupssuchas CN (or CF3) in 15 are desirable,the cyclopropene routecannotbe taken. An alternativesynthesishas been developedwhere spiropyrazolesare photolyzeddirectlyin an azine/CHnClnmixture.Intermediatesin this processare the vinyldiazoalkanes66 and carbenes67 (refs. 10,14). Overallyieldsare between40-70 %.
262
Retro-1.5-electrocvclization Id and Diazo-routeId) Route (a) and @) give access onlyto DHls 15 bearingidenticalsubstituentsR'. However,it was shownabovethat the propertiesof DHls 15 dependvery stronglyon the substitution pattern'ofdoublebond2,3.Thereforetwo additionalrouteswere workedout. Route (c) startingfromdiazoalkane66 whichis formedof 66 via intermediate 65. The vinyl-diazoalkane 66 is convertedto 15 in the same way as in the pyrazole route (refs. 15,18).The diazo-routeuses a classicalentry via the fulvene69 intothe diazoalkane series66 (ref. 84). Identicalconversionslead to the DHls 15 in 30-50% yields (see scheme25). Recentlynew compoundsof type 15 belongingto moleculeswith supramolecularpropertieswere prepared(ref. 84). Scheme 29
Route c
16
15
Route d
BiDhotochromic Svstems; The preparationof bichromophoric systemscan be achievedin principlein two ways: additionof 1) monofunctional spireneto base or 2) bifunctional spireneto base.
263
Route 1:The additionof spirenes63 to 1,3-or 1,4-diazinesgivesthe potentialphotochromicbisadducts17 or 19 (ref. 17). However only 19 showsphotochromism. A separationof the two nucleophiliccentresin the azine leadsto moleculesof the bipyridine type A. Reactionof these bis-basesaffordsthe photochromic 1:l adducts 15 and the bichromophoric 2:l adducts21 dependingon the reactionconditions. By controlling the molar ratio63/diazineor reacting15 with 63, symmetricaland unsymmetricalbichromophoric DHI compounds21 are accessibleselectively(overallyields4070 %; refs. 17,lQ). Scheme30
+ 63 RQR2
x
x
hu
-Befaine
22,23
21
R'= C02CH3
56
MixedSvstems-
-
-
The use of dibenzoannulated1,Bdiazinessuch as benzo[c]cinnoline in the cyclopropene route (a) affordsinteresting1,2- systemsof type 44 (refs. 5536).
264
TVDe-1.2-aza-oxa-svstems; It shouldbe possibleto prepareisoxazolesfrom nitronesand alkynesby 1,3 dipolarcycloaddition.It has in fact been foundthat the appropriatespiroxazolinescan be preparedfromdiazo-fluoreneand alkynes(ref. 57). These are thermochromic(not photochromic)at moderatetemperatures.At highertemperatureshoweverthey undergoirreversiblesecondaryreactionsto 30 via the intermediate29 (see scheme 18).
Tv~e-7 3-diaza-s~ stems: The 1,3 dipolarcycloaddition reactionwhichis so effectiveas a methodfor synthesizing 5-memberedringscan also be usedto preparespiro-heterocyclic compoundsof the 2,3-type. If a suitablebromo-substituted dibenzopentafulvene is reactedwith pyridine the pyrazolo/pyridine51 is formedin 30-90 % yields.Analogeousfytreating69 with70 as a 1,3 dipole57 is obtained.
Scheme31
R2'
+ 69
\
R2
.
265
ConcludinaRemarks 1,5-Electrocyclization is a potentreactionwhichhas been employedin creatingnew photochromic systems.It fits in the frame of knownelectrocyclic processeswhichhave been usedin variousphotochromic systems(see Table 1). The main problemis to preparecolorlessand coloredformshavingalmostequal energy contents.1,8Electrocyclizationof the coloredbetaineto the colorlesscyclicstructurecan be controlleddramaticallyby the substitution patternof the molecules.This basicconcepthas been extendedlately from monoaza-(DHI),to diaza- (DHPP) and finallyto triaza-&ringheterocycles. Normallythe formationof a C-N bondin the electrocyclization is notfavorable.It can be forcedin this direction,however,by annulatingaromaticringsin suitablepositions. The characteristicsof the photochromic systemsbased on 1,5-electrocyclization are: a) a neutralmoleculeas educt givesa zwitterionicspeciesas primaryproduct,and b) reversibilityis possibleonly if secondaryreactionsleadingto heteroaromaticcompoundsare suppressed.Typicalstructuralelementsare thereforespiro-(orgem-) substitutionto avoid rearomatization. The photoreaction of the colorlessform to the betainesdoes involvea singletspeciesin all cases studiedso far. Ringopeningof the photochromic 5-ringheterocyclescan be inducedonly photochemically. The reverseprocess,however,can be broughtaboutthermallyor photochemically. Type 3 systemsshow in principlephotochromism.Howeverthe systemsstudiedso far carry hydrogenatomsin the new carbonbondformed. In many cases consecutivereactionsoccurbeingnormallyirreversible.This systemmightbe optimizedby introducingnon hydrogensubstituents.This is a pointwhichneedsfurtherwork.
266
REFERENCES 1 H. Diirr,Angew. Chem., 101 (1989) 427;Int. Ed. Engl. 28 (.l989) 413. H.DOrr, Wiss. Zeitschr.TH Leuna-Merseburg,26
(1984)664. Techniquesof Chemistry,Vol. 111, Wiley Interscience, 2 H.G. Brown, Photochromism, New York, 1971. 3 R. Huisgen,Angew. Chem., 92 (1980) 979;Int. Ed. 19 (1980) 947. Review; E. C. Taylor, 1. J. Turchi,Chem. Rev., 79 (1979)181. for furtherreferenceson 1,5-electrocyclizationsee: J.Elguero, Bull. SOC.Chim. Fr., (1971)1925. Reviews:G. Coispeau, J. Elguero,Bull.SOC.Chim. Fr., (1979)2717. R. Huisgen,Angew. Chem., 72 (1960) 359.H. Reimlinger,Chem. Ber., 103 (1970)1900. S.W. Staley in PericyclicReactions, R. E. Lehr and A.P.Marchand, Eds. Vol. 1, p.199, AcademicPress, New York, 1977. 4 R.B. Woodwardand R. Hoffmann,Angew. Chem., 81 (1969)797;Int. Ed. 8 (1969) 781;K.Fukui, ACC.Chem. Res., 4 (1971)57. 5 T.L. Gilchristand R.C. Storr, OrganicReactionsand OrbitalSymmetry, Cambridge, Univ. Press, 1979. 6 N.J.Turro, ModernMolecularPhotochemistry,Benjamin Cummings,Publ. Co., Menlo Park, 1978. 7 a) C. Dorweiler,P. Spang, H. Diirr, K. Peters, H.G.v. Schnering,Isr. J. Chem., 25 (1983)241.b) N.A. Garcia, G. Rossbroich,S. E. Braslavsky,H. Diirr,C. Dorweiler, J. Photochem.,31(1985) 297;c) C.Dorweiler,H. Diirr,H.P. Jdnsson,X IUPAC Symposium on Photochemistry,Interlaken,Abst., (1984)359.d) C. Dorweiler,Thesis,Universitatdes Saarlandes,Saarbriicken,1986. 8 a) E. Marvell,Thermal ElectrocyclicReactions,AcademicPress, New York, 1980. b) R.Bates,D.W. Grosselink,J.A.Kaczynski,TetrahedronLett., (1967)205.c) H. Kloosterziel,J.A.A. van Drunen, Rec.Trav.Chim.Pays-Bas,89 (1970)270. d) S.Brenner, J.Klein, 1sr.J. Chem., 7 (1969)735.e) R.B. Bates, D.A. Combs, TetrahdronLett., (1969)977; 9 R. Sustmannand H.J. Dern, Chem. Ber., 116 (1983)2958. 9 G. Hauck and H. Diirr,Angew. Chem, 91 (1979)1010,Int. Ed., 18 (1979)945. 10 H. Grot3 and H. Diirr, Angew. Chem., 94 (1982)204,Int. Ed., 21 (1982)216. 1 1 H. Diirrand G. Hauck, D.O.S. 29 06 193,1980. 12 H. Diirr,H. Grot3 and K.D. Zils, D.O.S. 32 20 257 A 1, 1983. 13 H. Diirr, H. Grot3and G. Hauck, Chem. Ber., 116 (1983)856. H.DUrr,W. Gomblerand R. Sergio, Angew. Chem., 84 (1972) 215; Int.Ed. 1 1 (1972)224. 14 H. GroO, Thesis, Universitiitdes Saarlandes,Saarbriicken,1982. 15 H. Dijrr, H.P. Jdnsson,P. Scheidhauer,T. Miinzmayand P. Spang, D.0.S P 3521432.5,1985. 16 H. Diirr, P. Scheidhauerand J. Besse, XI IUPAC Symposiumon Photochemistry, Lisbon,(1986)202. 17.H. Durr, P. Spang, Angew. Chem., 96 (1984)227;Int.Ed. 23 (1984)241.
267
18 H.P. Jonsson,Thesis,Universit& des Saarlandes,Saarbrucken, 1985; P. Spang, Thesis, Universitatdes Saarlandes,Saarbrucken,1985. 19 T.Munzmay, P. Spang, M. Holderbaum,H. Durr, E. Rabe, C. Kruger, Chem. Ber., 121 (1988) 843. 20 H. Durr and P. Spang, D.O.S, P 32 20 2571 (1984). 21 H.Durr,H.GroB, K.D. Zils, G. Hauck, G. Klauck. H. Hermann,Chem. Ber., 116 (1983) 3915. 22 T. Miinzmay,Thesis, Universitatdes Saarlandes,Saarbrucken,1987. 23 O.L.Chapman, G.L. Eian, A. Bloom,J. Clardy, J. Am. Chem. SOC.,93 (1971) 2918. 24 K.H.Grellmann,G.M. Sherman, H. Linschitz,J. Am. Chern. SOC.,85 (1963) 1881. 25 H. Linschitz,K.H. Grellmann,J. Am. Chem. SOC.,86 (1964) 303. 26 K.H. Grellmann,W. Kuhnle,H. Weller, T. Wolff, J. Am. Chem. SOC., 103 (1981) 6889; and earlier lit.. 27 R.D.Yousefyeh,M. Weisz, J. Am. Chern. SOC.,96 (1974) 315; and earlier lit.citedtherein. 28 A.G. Schultz,Acc. Chern. Res., 16 (1983) 210; and earlier lit. citedtherein. 29 K.H.Grellmann,U.Schmitt,H.Weller, J. Chem. SOC.Comm., 1982, 591. 30 A.G. Schultz,R.D. Lucci,J. Org. Chem., 40 (1975) 1371. A.G. Schultz,W.Y. Fu, R. D. Lucci,B.G. Kurr, K.M. Lo, M. Boxer,J. Am. Chem. SOC., 100 (1978) 2140. 31 T. Wolff, J. Org. Chem., 46 (1981) 978. 32 T. Wolff, J. Am. Chem. SOC.,100 (1978) 6157. 33 E. Fischer,J. Phys. Chem., 71 (1967) 3704. 34 J. Blanc, D.L. Ross, J. Phys. Chem., 72 (1968) 2817. 35 E.M. Kosower,J. Am. Chem. SOC.,80 (1958) 3259. see a1so:E.M. Kosower,An Introduction to PhysicalOrganicChern., J. Wiley and Sons, New York, 1968. 36 J.Fabian,H. Hartmann,LightAbsorptionof OrganicColorants,SpringerVerlag, Berlin,1980. 37 K.Dimroth,C.Reichhardt,T.Siepmann, F.Bohlmann,LiebigsAnn. Chem., 661 (1963) 1. 38 P.C. Darcy, R.J. Hart, H.G.Heller, J. Chem. SOC.PerkinI, (1978) 571. 39 H. Durr, see I.c. 1). 40 T. Munzmay,masterthesis, Universitatdes Saarlandes, Saarbrucken,1985, and I.C. 21). 41 M. Holderbaum,masterthesis, Universitatdes Saarlandes, Saarbrucken,1986. 42 a) G. Klauck,Thesis, Universitatdes Saarlandes,Saarbrucken, 1983. b) A. Berkessel,masterthesis, Universitatdes Saarlandes, Saarbrucken,1982. 43 H. GroB, H. Durr, W. Rettig,J. Photochem.,26 (1984) 165.
268
44 N.J. Turro, P. Lechtken,N.E.Schore, G. Schuster, H. Steinmetzerand A. Yekta, ACC.Chem. Res., 7 (1974) 97, 45 W. Adam, Adv. Heterocycl.Chem., 21 (1977) 437. 46 G. Bar, G. Gauglitz,R. Benz, J. Polster,P. Spang, H. Diirr, Z. Naturforsch.,39a (1984) 662. 47 see ref. 6). 48 P. Scheidhauer,Thesis,Universitat des Saarlandes, Saarbrkken, 1987. 49 a) U. Steiner, UniversitgtKonstanz,PrivateCommunication.b) see ref. 12). 50 H. Diirrand H. GroO, TetrahedronLett., (1981) 4679. des Saarlandes, Saarbrkken,1989. 51 T. Wegmann,masterthesisJJniversit8t 52 H. DW, TopicsCurr. Chem., 40 (1973) 103. 53 W. Rundel, Houben-Weyl,Vol. 4/5b, 1131, ThiemeVerlag, Stuttgart,1975. 54 H.DCrr, Houben-Weyl,Vol. 4/5b, 1158, Thieme Verlag Stuttgart,1975. R.A.Moss, M.Jones, Carbenes, Vol. 1,11, Wiley and Sons, New York, 1973. 55 A. Thome, masterthesis, Universitatdes Saarlandes,Saarbrccken,1985. 56 H. DCrr,A. Thome,U. Steiner, T. Ulrich,C. Kriiger, E. Raabe, J. Chem. SOC.Chem. Comm., (1988) 338. 57 V. Bach,masterthesis,Universitatdes Saarlandes, Saarbrkken,1984. 58 H. Dijrr, C. Schommer,T. MOnzmay, Angew. Chem., 98 (1986) 565, Int. Ed., 25 (1986) 572; s. a. ref. 40). 59 C. Schommer,Thesis,Universitatdes Saarlandes, Saarbrkken, 1987. 60 A. Kakehi,.S. Ito, B. Wada, Bull. SOC.Chim. Jap., 57 (1984) 893. 61 K. Burger,C.Zettl, F. Hein, H. Schickaneder,Chem. Ber., 112 (1979) 2620. 62 P.Appriou,RGuglielmetti,J.Metzger, Helv.Chim.Acta,55 (1972) 1782. 63 J.Kolc and R.S.Becker, Photochem.and Photobiol.,12 (1970) 383. C.Lenobleand R.S. Becker,J.Phys.Chem., 90 (1986) 62. M.Gehrtz, C.Brauchle,J.VoitlBnder,J.Am.Chem.Soc. 104 (1982) 1340. 64 M.Melzig, D.O.S, P 3345624; 3345325 (1985). 65 A. Thome, Thesis, Universitatdes Saarlandes,Saarbrkken, 1988. 66 a) I.R. Dunkin,M. Poliakoff,J.J. Turner, TetrahedronLett. (1978) 873. b) A. Holm, N. Harrit, N.H. Tourro,J. Am. Chem. SOC., 97 (1975) 6167. 67 H.Grol3, Thesis, Universitgtdes Saarlandes,Saarbrficken,1982. 68 V. Bach,Thesis,Universitatdes Saarlandes,Saarbrkken,1987. 69 T. Uchida, M. Wada, Mol. Cryst. Uq. Cryst., 63 (1981) 19. 70 R. Eidenschink,Kontakte(Darmstadt),2 (1984) 25. 71 V. Krongauz,Nature, 43 (1984) 271.
269
72 Kkhimura,see chapter 10. 73 G. Smets, PolymerScience50, 18, SpringerVerlag Heidelberg,1983. G. Smets Pure Appl. Chem., 30 (1972) 1. 74 H. Dijrr, H. GroR, see ref. 50). 75 R. Gautron,Bull. SOC.Chim. Fr., (1968) 3190. 76 H. Dijrr,V. Bach, D.O.S:P3710889.1 (1987). 77 a) G.R. Meredith,D.J. Williams,S.N. Fishrnan,E.S. Goldbrecht,V.A. Krongauz,
J. Phys. Chem., 87 (1983) 1697. b) G.R. Meredith,V. A. Krongauz,D.J. Williams,
Chem. Phys. Lett., 87 (1982) 289. c) V.A. Krongauz,Isr. J. Chem., 18 (1979) 304. 78 H.W. Starkweather,A. Acidoff,F.R. Eirich, Ind. Eng. Chem., 47 (1955) 302. 79 R.J. Hovey,C.H. Fuchsmann,P. G. Piusz, D.0.S 29 36255 (1980). 80 a) R.J.Hurditch, PrivateCommunication,Olin Hunt, East Providence,Rhode Island,USA.b) see also ref. 64). 81 a) H. Sixl,Umschau,16 (1983) 474. b) W.P. Su,JRSchrieffer,A.J. Heeger, Phys. Rev. Lett., 42 (1978) 1698. 82 K. Menke, S. Roth, Chemie in unsererZeit, 20 (1986) 1. C.C.Ku, R.Liepius, ElectricalProperties of Polymers,Hanser Publ., Munchen,1987. 83 H. Dijrrand R. Gleiter,Angew. Chem., 90 (1978) 591; Int. Ed. 17 (1978) 84 H.Durr, K.P.Janzen, A.Thome, B.Braun, DEOS P 35 21 432.5 (1988). 85 M. Regitz, Angew. Chem., 78 (1966) 684, Int. Ed., 5 (1966) 681. A19 See AdditionalLiterature(1 989 - 2001): Dihydro-indolizines,
210
Chapter 7
4n+2 Systems: Molecules Derived from 2-Hexa-I ,3,5-Triene/Cyclohexa-I ,3-Diene W. H. Laarhoven
INTRODUCI'ION
1
According to the definition of photochromism given in chapter 1 the title compounds of this chapter do not present a photochromic system. Though irradiation of cyclohexa-1,3-diene (2) yields Z-hexa-13,s-tiene (l),the reversed reaction is not observed in the parent system, neither thermally nor photochemically, or only to a very small extent (ref. 1-3). In Scheme 1, the photoproducts from (1) are given. The most efficient photoreaction of (1) is cis-trans isomerization into the E-isomer (3). The other photoproducts are formed by a [2+21 or a [4+21
0
hv
1
Scheme 1
2
1'
4
5
6
3
intramolecular cycloaddition (cornpounds (4) and (3, respectively) or by a [1,5] hydrogen shift giving (6). These types of reaction are commonly also encountered with derivatives of (1). More efficient ring closures are observed with derivatives of (1) substituted at C2 and C5 as in Z-2,S-dimethylhexa-l.3,5-tiene(7).Irradiation of (7)at 313 nm gives the cyclized compound (10) as the main product (Scheme 2). However, upon changing the wavelength of irradiation to 254 nm, cis-tram isomerization to (8) becomes again the main photoreaction (ref.4-6). Compound (13) is only formed on irradiation of E-(7). The example shows both the influence of substituents and of the wavelength of irradiation on the photoproduct composition. From extensive studies of these phenomena the importance of the conformations in the ground state of the substrates for the formation of the photoproducts became evident, resulting in the formulation of the NEER (Non Equilibration of Excited
271
Scheme 2
7 313 nm 254 nm
8 1.13 8.05
9
0.01 0.93
10 8.85 0.75
12
0.01 0.27
11
0.05 0.05
13
--
--
Photoproduct composition (in %) of 2-2,5-dimethylhexa-1,3,5-mene after 10% conversion on irradiation at 313 and 254 nm. Roamers) principle (ref. 2). (For a recent review see ref. 7). When the expression photochrome is taken literally, alkyl-substituted hexatrienes, which give cyclohexadienes in high yield upon irradiation, do not form a photochromic system with their cyclization products, even not when the reversed reaction occurs too, because the participating compounds do not absorb in the visible part of the spectrum. Substitution of hexatriene with phenyl groups increases the wavelength of absorption considerably. However, on irradiation of these derivatives valence isomerization products (derivatives of 5) are formed which do not give the reverse reaction (ref. 8,9). This can be ascribed to the geometries of the ground state conformations, which are unfavorable for electrocyclization of these compounds. Good examples of the influence of auxochromic groups together with a favorable, fixed conformation are presented by some derivatives of dihydroxanthenones as reported by Ullman and Huffman (ref.10) (Scheme 3). Solutions of (14) in non basic solvents are colorless to faintly red and are thermochromic. On irradiation they become bright red by the formation of the conjugated ring-opened product (15). In table 1 the wavelengths of the lowest energy absorption maxima of (14) and (15) are given. Solutions of (14a-d) in basic solvents are already colored by the presence of (15a-d). Extended irradiation of (14) leads to the irreversible formation of (16) without the intermediacy of (15). Presumably (16) is formed by a di-n-methane rearrangement involving a CN-group (ref.27a). It is of interest to note that (15) with R2 = H does not give (14), neither thermally nor photochemically. This demonstrates the importance of a substituent at this position for the
272
Scheme 3
14
R2
16
R2
TABLE 1. longest wavelength of maximum absorption
substituents
a
R1
R2
H
c6H5
Ar
(14) (solvent)
(15) (benzene)
c6H5
312 (CsHi2)
532
conformation required for ring opening. From the above mentioned considerations, the following conditions for a potentially photochromic hexatriene 1 cyclohexadiene system arise: i) The preferred conformation in the ground state of the hexatriene should be the c,Z,c conformation (see Scheme 4) to ensure a maximally enhanced rate of @hoto)cyclizationinto the corresponding cyclohexadiene. ii) The photoreaction has to be accompanied by an extension of the conjugation in the resulting product, in order to achieve extension of the absorption spectrum of the system into the visible region. Such an extension of the number of double bonds occurs on photochemical (or thermal) ring opening of (2). With (1) this condition is only fulfilled when at least one of the double bonds is part of an aromatic system, Photocyclization will then cause loss of aromaticity, thereby extending the conjugation in the product with at least two double bonds. The loss of aromaticity in the photoproduct is of course a destabilizing factor. As a consequence the product might be thermodynamically much less stable than the starting material. When no additional stabilizing factors are present the photoproduct will usually revert to the original compound without the need of much energy of activation. This makes the molecule more suited as a
273
=c
7
Scheme 4
t tzt
tEt
tzc
czt
czc
r
J tEc
cEt
cEc
c = cisoid conformation over a single bond t = transoid conformation over a single bond photochromic compound. An additional, favourable effect of the introduction of aromatic rings is their influence on the conformation. Substitution of a terminal double bond has the effect of a substituent at the C2 ( C5)position of (l), whereas substitution of the central double bond by a
g
Scheme 5
20
21
2
22
8 \
23
benzene ring causes a rigid Z configuration. According to condition i) both effects increase the potency of the compound to act as a photochromic compound. Systematic substitution of each double bond in (1) or (2) by a benzene ring leads to the
274
compounds given in Scheme 5. In this introduction, a short survey of their photochemical properties and an indication of their potential ability as a photochmic compound will be given. It will appear that not all of them are photochromic compounds, and the photochromic behavior of some of them has not been fully investigated. Those compounds for which photochromic properties have been established will be treated in more detail in this or a subsequent paragraph. On irradiation of 1-phenylbutadiene (17) in apolar solvents only phenylcyclobutene is
fl a 5 Scheme 6
24
\
26
17
H 25 22
27
28
R
29
formed in low yield (ref.11). However, on irradiation of a solution of (17) in the presence of iodine as an oxidizing agent naphthalene (26) is obtained in good yield (ref.12). This indicates the intermediacy of 1,9-dihydronaphthalene(25) which is apparently unable to rearrange into 1,2dhydronaphthalene (22); the rate of a thermal or photochemical 1,5-hydrogen shift is obviously much slower than the rate of ring opening. Attempts to trap (25) at low temperature have not been reported. Aryl-substituted derivatives of (17) (e.g. (27)) behave similarly (ref.13). NO investigation of the same reaction without iodine at low temperature is known. Structurally related to 1,4-diarylbutadienes are the 1.4-diarylbut-l-en3-ynes (e.g.(30)). On irrradiation they yield arylaromatics but in this case a dehydrogenation step is not necessary (ref.14). The reaction is catalyzed by amines, oxygen. alcohols, BF3 a.o., all of which react with the short living intermediate (31) (Scheme 7) in a protonation-deprotonationreaction (ref.14b). A colored product is not expected in this reaction.
275
Scheme 7
30
31
1
32
Another example of a system representing a derivative of (17) is the en01 form of 2-methylbenmphenone (34) (Scheme 8) which has an absorption maximum at 400 nm. Irradiation of (33) forms the cyclization product (35) absorbing at 383 nm. The latter compound
&&& Scheme 8
\
/
\
/
/
hv or A
CH3
33
- h\ v + &
34
35
is sensitive to oxidation and yields anthrone (ref.15). This kind of photochomism, based on hydrogen aansfer, is treated in chapter 5. Related to the compounds (17) and (27) are the fulgides and analogous compounds in which the configuration around the central single bond of the butadiene moiety is made rigid by a three-atom bridge (see (36)). Such molecules possess the preferred configuration for ring
R 36 closure and the ring-closed photoproducts may be more or less stabilized by suitable atoms at
276
the positions a and c. The photochromism of fulgides and analogues compounds is treated in section 4.2.4 by H.G.Heller. The primary photoproduct (37) of 1,2-divinylbenzene (18) (Scheme 9) is formed by a [4+2] photocycloaddition, analogous to the conversion of (1) into (5). A subsequent vinyl-cyclopropane rearrangement leads to the stable benzobicycl0[3.1 .O]hex-2ene (38) (ref. 16).Hence no photochromic behavior can be expected with this compound. Introduction of a terminal phenyl group into (18) leads to 2-vinylstilbene (39); its photochemistry and that of its derivatives have been investigated comprehensively (ref. 17). Irradiation of (39) and derivatives substituted at the vinyl group affords (substituted) 5-phenylbenzobicyclo[2.1.l]hex-2- enes (41) via a radical pathway (ref.17a). Even in the presence of iodine as an oxidizing agent stilbene Scheme 9
18
37
39
39
38
40
a. R,
#
H, R2 = H
41
38a
b. R1= H, R*+H (39) gives (41) as the main photoproduct. The photoproduct formation is dictated by the NEER-principle: Derivatives substituted at Ca of the central double bond, e.g. (39a) give rise to derivatives of (38) (ref.17b). whereas substitution at Cf3 (39b) leads to photostability under anaerobic conditions (ref. 17d). However, under circumstances in which conformations of the hexatriene moiety of (39) having a c.Z.c-arrangement possess lifetimes long enough to allow their excitation, naphthalene derivatives are found as photoproducts. This is the case when (39) adsorbed on silicagel is irradiated (ref.18). Substitution of (18) at both terminal positions with phenyl groups leads to (42, R,=H). which yields only dimers on irradiation, even at very low concentrations. Derivatives of (42) with R,+H give photoproducts analogous to those of (39)
277
(ref.19). This is rationalized by the existence of two independently absorbing chromophores in these formally conjugated molecules. Photochromism is not expected in these cases; only when R1 in (39a) or R, in (42) is an aromatic group the compound might react analogously to stilbene derivatives and yield colored, primary cyclization products (ref. 19)(see below). Irradiation of 2-vinylbiphenyl (19) (Scheme 10) gives rise to 9.10-dihydrophenanthrene (23a), which must be formed via Ba,9-&hydrophenmthrene(43) by a 19-hydrogen shift (ref.20). Scheme 10
19
a.R=H b. R = C ~ H S
43
23a
&J &) +g 23b
/
46
\
19b
\ /
/
44
\
/
45
Coloration by (43) has not yet been observed, at low temperature (19)/(43) might exhibit true
278
photochromic properties. The same holds for the phenyl-substituted derivatives of (19). Both cis- and trans-2-styrylbiphenyl(19b) give 9-phenyl-9.10-dihydmphenanthrene(23b) (ref.2la,b). In this case, however, it is expected that at lower temperature two different, colored, primary cyclization products will be present, as on irradiation of (19b) in the presence of an oxidizing agent besides 9-phenylphenanthrene (44)the 1-isomer (45) is also obtained (ref.21). It is of interest that in contrast to the cyclization of (19b) the cyclization of (46),into (23b) also occurs from its triplet state (ref.2lc,d). The photochromic properties of stilbene (20) and more generally of 1,2-diaryl- ,triaryl- and tetraaryl ethylenes will be treated in section 2.2. Ortho terphenyl (21) forms triphenylene (47) on irradiation in the presence of iodine. Results of low temperature measurements concerning the reversible formation of a colored cyclization product have not been reported. However, on irradiation of a degassed solution of
(g
Scheme 11
hv
\
21
47
50
the polyphenelene (48) a stable red-orange colored cyclization product (49) precipitate (scheme 11). On standing overnight. or on heating, the solid dissolves and the solution becomes colorless. On irradiation of (48) in the presence of iodine as an oxidant, (SO) is formed (ref. 22). Several analogous polyphenylenes do not form cyclization products (ref. 22b). The photochromic behavior of 1,2-dihy&naphthalene (22) will be outlined in section 2.1. The compound obtained by substitution of the double bond in (22) by a benzene ring is
279
9,lO-dihydrophenathrene (23). It is photostable under the usual experimental irradiation conditions. 2
ALLCARBON SYSTEMS
2.1
1,2-Dihvdronaphthalenes
1,2-Dihydronaphthalene (22) is a cyclohexa-l,3diene in which one double bond has been replaced by a benzene ring. The ring-opened product (51) which is rapidly formed on irradiation can be considered as a hexa-1,3,5-triene having a single bond in a fixed s-cis configuration. As outlined in the introduction this ring-opened product has the tendency to restore the aromaticity of the ring system. This can occur by a thermal as well as a photochemical reversion into (22), or by a secondary thermal or photochemical reaction of (51). Scheme 12
22
52
38
51
CH3
H3 c
H
7
19
(3)
+ c2H5
19
EGH 9
Numbers in parenthesis are yields (in %) at - 100' C. numbers without parenthesis are yields at +25OC
280
The latter reaction requires excitation of (51) at a long wavelength (X > 410 nm) (ref.23) because of its conjugated pentaene system (Scheme 12). The main photoproduct of (22) under these conditions is benzobicyclo[3.1.0]hexene (38) (ref.23). Especially when (22) is substituted with alkyl groups at C(l) or C(2) as in (52), products can also be formed by hydrogen shifts, restoring the aromaticity (ref.24). As substituents at C( 1) or C(2) can take two different positions viz. a pseudoaxial (pa) or a pseudoequatorial (pe), the concerted ring opening can occur in two different modes,governed by the principle of least motion (ref.25). In Scheme 12 this is illustrated for cis-l,2-dimethyl-l,2dihydronaphthalene (52).The yields (in S)of products are given at two different temperatures, viz. at 25OC and at -1OOOC (ref.24~).The data show that at lower temperatures the hydrogen shifts are inhibited. In (22). the position of an alkyl substituent determines the type of photoproduct obtained, as well as the course of the reaction. This can be understood by realizing the degm of stereochemical interaction in the ring-opened primary photoproduct. Thus, irradiation of 3-methyl-l,2-dihydronaphthalene (53) at room temperature gives rise only to 2-methylenetetralin(54). At lower temperatures increasing amounts of reaction products formed from the ring-opened isomer (55)are observed (ref.26). indicating the high rate of ring closure of (55) into (53) at room temperature (Scheme 13). To prevent the hydrogen shifts in the pentaene intermediates, derivatives of (22) substituted with phenyl groups have been investigated (ref.27). However, besides intramolecular photo Diels-Alder products also photocyclization products are formed when the phenyl groups are present at a terminalposition of the pentaene, viz. from 1- and 2-phenyl-l,24hydronaphthalene ((58) and (60),respectively),(see Scheme 14). The cyclization product of (58). Scheme 13
55
cis-dibenzo[3.3.0locta-2,7-diene(59), is formed even at 254 nm, the wavelength of ring opening Of (58) (ref.27~).With (60) a competition between the photochemical ring opening reaction and a di-tc-methane rearrangement is found (ref.27a). These examples illustrate that the photochromic system (22)/(51) can exist at lower temperatures, using low wavelengths of irradiation but that in several cases side reactions occur, which interfere with the reversibility of
281
Scheme 14
58 pe
+t
the system. The reactions described above are observed in apolar solvents. In methanol as the reaction medium the dihydronaphthalenes form methoxy compounds on irradiation (ref.28), both by a direct addition of methanol to the starting compound and to some extent by addition of methanol to the primary ring-opened product. In the presence of amine (Scheme 15) several Scheme 15 hv
22 63 1,2-dihydronaphthalenesform 1,4-dihydronaphthalenes(ref.29). Only in a small number of cases the Occurrence of a colored pentaene has been demonstrated, Kleinhuis (ref.24a) observed an absorption maximum at 410 nm (C: ca 9500, ref.24b) on irradiation of l-methyl-1-phenyl-l,2-dihydronaphthalene at -165' C. which persisted for several hours but disappeared on irradiation at that temperature with a broad spectrum lamp or by warming the solution. No long wavelength absorption was observed with 1,l-dimethyl- or
282
I-methyl-l-ethyl-l,2-dihydronaphthalene.Heimgartner et al.(ref.24d) observed an intermediate on irradiation of I,l-dirnethyl-4-phenyl-l,2dihydronaphthaleneat - 181’ C at 402 nm. The ring opened products of I-phenyl-1,2dihydronaphthalene, which consist of two geomemc isomers, absorb at 435 nm.(qn= 30 min.at 191’ K). The thermodynamic data for the thermal ring closure were estimated as AH+=13f 2 KJ/ mol.; AS*= 8 f 2 J/K.mol.(ref.ZS). It is of interest to note that neither 3-phenyl-l.2-dihydronaphthalene (ref.27b) nor 1,2and 3,4-dihydrophenantne (ref.12) give rise to ring opening on irradiation. 2.2 1.2-Diarvlethylenes The photocyclodehydrogenation of diarylethylenes is one of the most comprehensively investigated photoreactions. (Ref.30 gives some recent reviews). In Scheme 16, this type of reaction is pictured for stilbene (20), the simplest representative of this class of compounds. The reaction starts with the reversible isomerization of E and 2 stilbene, which occurs from both the singlet and the triplet excited state of the molecule. (See ref. 31 for a review).(The photochromisrn due to photoisomerization is treated in chapter 3.) Irradiation of 2 2 0 gives rise also to the formation of 4a,4bdihydrophenanthrene (64).This compound and derivatives arisen from analogous reactions will be indicated as DHP’s. as is usual in the literature. For a review Scheme 16
E-20
2-20
about DHF”s see ref.32. Like other DHP’s (64)is an unstable compound. It cannot be isolated in a pure state but can only be investigated in dilute solution. In general, the photocyclization of Zdiarylethylenes occurs from the first excited singlet state (ref.33) via a comtatory process into trans DHP (ref.34). (A cyclization via an excited mplet state has only been found in rare cases, e.g. from pentahelicene (dibenzo[c,g]phenanthrene)(ref.35) and from an analog of (20) containing a hetero atom (ref.36).) Thermally and photochemically the DHP reverts to stilbene or it forms phenanthrene (65) by dehydrogenation. From the thermal as well as the photochemical ring opening cis-stilbene is formed exclusively. The thermal process is symmetry-forbidden and needs some activation energy, notwithstanding the high energy of the DHP as compared to the diarylethylene. The symmetry-allowed photochemical ring opening usually requires no activation energy. The quantum yield of the process is high, about 0.5 or more. It will only be smaller when fluorescence of the DHP occurs. The dehydrogenation reaction can take place more efficiently by addition of various dehydrogenating or oxidizing compounds like iodine or TCNE (ref.30). The stability of DHP’s of higher diarylethylenes is
283
discussed below. Because of their extended conjugation in a non-aromatic x system DHP’s are colored compound,s. Therefore, the couple Z-diarylethylene / DHP presents a photochromic system when the dehydrogenation of the DHP is prevented by rigorous exclusion of air from the solution. It has to be realized, however, that other competing reactions are possible, which prevent either the formation of a DHP from a diarylethylene, or the ring opening of the DHP. Scheme 17
4-
EY hv
66
PhCH2-CH-CH-CH2Ph I I Ph Ph 67
+
Ph-CH2-CH2Ph 20 68
Examples of the first category are found in diarylethylenes, containing a substituent which enhances singlet-mplet intersystem crossing e.g. a nitro group or a substituent which induces another more efficient photoreaction (e.g.in 2-vinylstilbene (39), see above). Also external conditions can prevent (in part) the photocyclization reaction. Thus, when the stilbene concentration is higher than 0.01 mom cyclobutane derivatives may be formed by cyclodimerization; cycloaddition products can also arise when another olefiiic compound is present in the solution (ref.37). When stilbene is irradiated in polar solvents in the presence of a tertiary amine an amino-adduct (66) i s formed besides reduction- and dimeric (67) products (ref.38), (Scheme 17). Methanol also adds to excited stilbenes, but with a very small quantum yield (Scheme 17) (ref.39). To the second category, viz. reactions which prevent the ring opening of DHP’s, belong all those reactions which involve the acid protons 4a and/or 4b or which change the conjugated system in such a way that aromatization cannot occur anymore by ring opening. In Scheme 18 some examples are given. Irradiation of diethylstilbestrol (69) leads to DHP (70) which is stabilized by keto-enol tautomerism as the diketone (71) when the reaction is performed in a buffered alcoholic solution (ref.40,34a). Irradiation of stilbene dissolved in a primary amine or in a basic, alcoholic solution gives rise to a mixture of 9.10- (72) and ’
k44ihydrophenanthrene (73) via deprotonation-protonationreactions of the DHP; 1,Zand 3,4dihydrophenanthrenes are formed as intermediates in this reaction (ref.29b). Irradiation of stilbenes substituted with enolizable groups at the double bond form normal DHP’s in apolar solvents, but in alcoholic solution these primarily formed DHP’s are easily enolized using the 4a-hydrogen, thereby restoring the aromaticity in one ring. Eventually, a substituted
284
Scheme 18
69
70
OH
$: q: th v,
71
0
___)
/
\
\
76
77
\
78
9,lO-dihydrophenanthrene(75) is formed via a radical step (ref.41). The formation of the 9,lO-dihydrophenanthrene (78) from dicyanostilbene (76) takes place even in apolar solvents (ref.42) and, in part, also in the solid state (ref.43). In all these cases the reversibility of the photochromic process depends on the ratio of the rates of ring opening of the DHP and the competing reactions, and this depends on temperature, solvent, pH etc..To obtain maximal conversion into the DHP all rates of decompositions of the DHP, including its thermal and photochemical ring opening have to be minimized. In general the reactions pictured in Scheme 18 may also occur in higher diarylethylenes. In these cases, however, the reaction sequence is usually more complex due to the increased number of cyclization pathways. The influence of substituents and annelation is indicated in tables 2 and 3 with data of properties of DHP's.
285
2.2.1
Stilbenes
The absorption of the DHP's of stilbenes in the visible part of the spectrum consists of a usually structureless, broad band, which implies that the position of the maximum is in some cases approximative. Beside this band, other more structured bands (Jl and In) are present, which are usually hidden by absorptions of the trans and cis stilbene. For the DHP of unsubstituted stilbene these absorption bands are at: (II) 310 nm (e=22.200), 297 nm (~=20700); (III) 237 nm (~=15200)(ref.33a). The effect of substituents on the visible band is small; it increases with the number of substituents, and is larger when the substituent is at a terminal position of the conjugated system (C4). (In Tables 2 and 3 the wavelength of maximum absorption of several substituted DHP's, derived from stilbenes and diphenylcyclopentenes are given, together with other physical data.) In some cases, two absorption maxima are given for one compound in Table 2 (f,t). One of them belongs to a labile conformer, which is formed only at low temperature. So the '475'-isomer of 9,lO- dicyano-DHP (f) is formed above -4OoC, TABLE 2.
Wavelength of the visible absorption maximum (Lax, in nm), half life-time (T), activation energy (Ea, in kJ/mol) and activation entropy (ASf, in J/mol."K) of thermal ring opening of DHP's derived from stilbenes (20) in apolar solvents. T*
450 (6750)
b
3-Br
d e
3-OMe 3-N(Me)2 9,lO-diCN
f
9,lO-diF 3-Me 9,lO-&Me j 9-Me k 1,3,4a-triMe 1 1,3,4a,4b,6, 8-hexaMe m 2,4-diMe n 2,4,5,7tetraMe
454 450 a)475 b)510 430 447 455 452 460
g
h i
o 9-COOMe p
4 r
s
t
4-OMe 2-OMe 4-mz 2-m, 3-CN
* in min.,
I
("C)
96 (25) 106 (25) 90 (25) 9 (-40) 57 (-40)
475 (3100) 458
33h (25) 2h (-31)
525
6 (-120) 43 (-1602 7 (28)* 2.4s (28)***
450 485 510 497 a1442
unless given otherwise;
9
stable (0)
** in hexane; *** in methanol.
286
Lax(%l,)
substituent a b C
d e f
f i
jk 1
460 cl500) 455 455 455 455 455 455 455 464 483 460 485
H 3-CF3 3-OMe 3-F 3-cl 3-Me 3-OH 3-Br 2-Br 4-Br 2-OMe 4-OMe
("C)
Ea
ASf
ref.
23 (25) 35 (20) 44 (19) 43 (19) 43 (20) 61 (20) 2 (22) 28 (19) 97 (11) 38 (-16) 96 (11) 14 (11)
45.2 51.9
-164.4 -142.2
60.0 36.4
-116.3 -172.4
60.2 27.2 61.9 32.5
-114.2 -213.0 -109.2 -199.6
47 47 47 47 47 47 47 47 47 47 47 47
7
Scheme 19
Jgi
l ( 79 )*
II
R
fl
hv
80
1
oxid.
( 79 III
81 79
R=Me R=OMe
-25O C +197O c -25O C +197O C
81 : 83 1.1 8.0 1.o 3.0
83 R
287
whereas the '51O'-isomer predominates at -8OOC. Apparently there is no effect of substituents on the quantum yield of DHP formation from diphenylcyclopentenes, as can be concluded from the quantum yield of cyclopentenophenanthrene formation under optimal conditions for oxidation, viz. at low temperature to enhance the lifetime of the DHP, and using a high concentration of iodine. Under these conditions, the yield ratios of 2- and 4-substituted phenanthrene derivatives from meta- substituted diphenylcyclopentenes are unity (see Scheme 19)(ref.47). With stilbenes such a substituent independency of the quantum yield of cyclization is not observed. This may be caused by differences in the decay rates of the excited singlet states of stilbenes and diphenylcyclopentenes. For the stilbenes this process may involve a twisted configuration which can convert into the Z- and E- ground states. Such a twisted configuration is not well attainable for the cyclopentene derivatives. As can be concluded from the half lifetimes in Tables 2 and 3 the rate of the thermal ring opening of DHP's is influenced by substituents. A Hammett relation is not found, but an isokinetic relationship between activation energies and entropies appears to exist. For the thermal ring opening of DHP's derived from stilbenes in a methylcyclohexane-ishexane solution a linear correlation with a slope of 2.93 .lo" deg-' is calculated (ref.33 a), whereas for the reaction of the cyclopenteno-DHP's in methanol a slope of 2.44 .lo" deg' is observed (ref.47).The substituent dependency of the activation energy of the thermal ring opening , which is unexpected from a steric point of view, might be ascribed to a change in the vibrational modes of the DJ3P or the polarity of the substituent. The existence of an isokinetic relationship indicates a common structure of the transition state of all DHP's investigated. The influence of steric effects is not reflected only in a decrease of the quantum yield of cyclization but also in the shorter lifetimes of DHP's with a substituent at C4, or at C4 and C5, and in the absence of a statistical yield-ratio of the three or four possible phenanthrenes from stilbenes having a meta substituent in both phenyl rings (ref.48,49). This may be a reflection of the ratio of the conformers present in the ground state. [2.2]Metacyclophanenes,e.g. (84), are related to Z-stilbene (Scheme 20).The oxidation to dihydropyrene (86) is very rapid, even when a methyl group is present at C15 (84b). Some propemes are given in Table 4. The [2.2]-metacyclophanedienes like (87) have a more distant resemblance to the stilbenes. They have received a lot of attention, because in these compounds the ring-closed valence isomer is the thermodynamically more stable form. The latter absorbs at a longer wavelength than the ring-opened isomer of these dihydropyrenes. Several derivatives have been synthesized by Boekelheide (ref.52). Some data are included in Table 4. More recently several benzo-annelated derivatives of (86) have been synthesized, to investigate the influence of the annelation position on the stability of the aromatic 14-n system of the dihydropyrenes and on the reversibility of the reaction (ref.54). Some of the results are also included in Table 4. Compound (90b) could not be isolated, and thermal or photochemical ring opening of (90a and b) was not observed, indicating the stability of the Hueckel aromatic 14-x system. Irradiation of (94) (Scheme 21) cannot give a trans fused ring closure, because of the 15,16-methylene bridge. Below -30 C (94) yields the excited triplet of (95). This rearranges into (97) which is in thermal equilibrium with (96). Excitation of (96) leads to (99) which forms the
A@
288
@0
\
84
a. R , = R 2 = H
Scheme 20
oxid.
I R2 0
85 b. R 1 = H ; R 2 = M e
87
@ 0
0
86 c. R , = R 2 = M e
88
a. R = H b. R = Me
89
R = Me,2-NO2
($ ($ c.
d. R = M e , 4 - B r
hv
0
R/
___)
\
0
91
9o
a. R = H
92
a. R = R, = R 2 = H
b. R = Me
93
c. R = M e , R , = M e , R2=.H
b. R = M e , R 1 = R2= H d. R = M e , R , = COMe, R2=H e. R = M e , R , = H , R2= Me
289
TABLE 4.
Wavelength of the visible absorption maximum (La, in nm), half life-time (z) and activation energy (Ea, in kJ/mol) of thermal ring opening of some 15,16-dihydropyrenes -
Compound 85a 85b 8% 88a 88b 88c 88d 91a 91b 93a 93b 93c 93d 93e
500 (3500) 500 (3200) 520 (3000)
497,466,440 630,600,508 634,531,502
z ("C)
Ea
ref.
61 h (25) 56.9 h (25) 916 h (25)
85 79 63
50,51 50,51 50,51
11.5 m (30) 0.17m (30) 20 m(30)
96 82
53
105
26 m (22) 0.2 m (22) 0.8 m (22) 27 m(22)
53 53
53 54a 54a 54b 54b 54b 54b 54b
Scheme 21
94
95
96
97
100 98 99 thermally stable (98). Excitation of this product leads to (100) and eventually to pyrene and methylene. At room temperature the rearrangement (94)-(100) proceeds with a single excitation step by-passing the ground state intermediate (98) (ref.55).
2.2.2 Higher Diarylethylenes In principle, higher diarylethylenes can form 1, 2, 3, or 4 different cyclization products,
03
290
Scheme 22 1
101 z
0
0
t
g
lo9
/
107
106
105
1
[1,5]-
H shift, induced by 0 2
'
' '
111
108
110
291
fl Scheme 23
1’
m=1.10
1 \
\
/
113
t’” I
0
116
&
\
ZF*= l.l&
\
0
\
\
$> Fx 0
117
118 Scheme 24
hv
123
\
124
0 \
I
0
119
292
depending on the substitution pattern. as is mentioned above for di-meta substituted stilbenes. However, unlike the latter compounds the former show a high selectivity in ring closure. Examples are given in Scheme 22. Several reaction parameters have been applied to predict the preferred cyclization route, and some 'rules' of highly predictive value, based on these parameters have been formulated.Thus, using the simple Hueckel theory, calculation of the sum of the free valence numbers of the atoms involved in the cyclization in the excited state, (ZF*,), provides the basis for the rule that photocyclization is possible only when this sum is larger than 1.0 (ref.56). This rule is especially useful for diarylethylenes. More generally applicable (e.g. also for triphenylenes and pentahelicenes) are the Mulliken overlap populations, (AQ, which are based on extended Hueckel M.O.calculations (ref.57). No cyclization will occur when An** and/or n*, have negative values or are very small (see for a review ref.30a and 32). In Schemes 22-24 some numerical values are given. As already pointed out above the lietimes of the DHP's depend on the rates of thermal and photochemical ring opening and of the dehydrogenation reaction. In some cases, however, 1.5-hydrogen shifts have been demonstrated. These too decrease the lifetime of the colored isomer. Thus the DHP of 3-styrylbenzo[c]phenanthrene (125) gives rise to 6aJ6d-dihydrohexahelbne (127) (Scheme 25) (ref.58) whereas 3-styrylphenanthrene (108) undergoes such a hydrogen shift, forming Scheme 25
hv +
127
f-
hv, A
125
126
128 6aJ6d-dihydropentahelicene (Ill), only in the presence of oxygen (Scheme 22) (ref.59). Apart from the possibility of E-Zisomerization of the diarylethylene and the possibility of other cyclization modes the quantum yields of DHP formation depend on the temperature. Fischer and his group in Israel have investigated the relations between the quantum yields of isomerization, ring closure, ring opening and fluorescence (of the isomers as well as of the DHP) and the reaction temperature for several diarylethylenes (ref.58a,60). Figure 1 presents an example of these relations for a rather simple case, viz. 1,2-di(2-naphthyl)cyclopentene (129) (Scheme 26)
293
Scheme 26
-
+
hv
-
\ / 129
& -
\ /H
130
131
(ref.60). Most remarkable is the sharp decrease of the quantum yield of photocyclization with the temperature, and the difference in temperature sensitivity of the two isomers. This allows the preparation of (131) as the only product at low temperature. Analogous curves are found with other compounds (ref 60). In table 5 physical data of several DHP’s from higher diarylethylenes have been collected. For an easy comparison the data of compounds with a close resemblance have been placed together.The compounds (121) and (124) form the same Scheme 27
Rl
132
a. R = R , = H
133
b. R = M e ; R 1 = H
c. R = R , = Me
hv
R
134
R a. R = H b. R = M e
135
polycyclic aromatic compound on dehydrogenation, as do the pairs (102), (109) and (113), (118). On the other hand (130) and (131) are the cyclopenteno analogs of (102) and (103). It is
294
PIC
Fig. 1 Quantum yields of the photocyclization of 1,2-di(2-naphthyl)cyclopentene (DNCP, 129) into (DHPA, 130) and (DHPB, 131). the photo induced ring opening and the fluorescenceof DHPA in the temperature range +90°C to -18OOC. Solutions in a 2 1 mixture of methylcyclohexane and isohexane, except for measurements in the range +25O to +90°C where methylcyclohexane served as the solvent. (Reprinted by permission of Rec.Trav.Chim.) striking that the DHP’s derived from the styrylphenanthrenes (109). (118), (124) have a much shorter lifetime and absorb at a longer wavelength than the comparable DHP’sderived from the dinaphthylethylenes (102). (113). (121). Mmover the former can be obtained even by irradiation at -120 C, at which temperature photocyclization of the dinaphthylethylenesdoes not occur. (See fig. 1). The same holds for the photoerasure at this temperature. The differences may be due to the higher energy content of the DHP’s derived from the styrylphenanthrenes(ref.59). Introduction of a three carbon bridge over the DHP’s (102) and (103) causes a substantial decrease in the lifetimes of the substituted DHP’s obtained in this way. This, as well as the shift of the UV-maximum to longer wavelengths, may be caused by strain. due to the forced more planar conformation of the central six-membered ring. Introduction of a three-carbon bridge from an orthoposition to the a-position of the double bond usually inhibits the photocyclization. A bridge of four C-atoms does not prohibit the photocyclization.
295
TABLE 5. Quantum yield of DHP formation in an apolar solvent at room temperature (&), wavelength of the visible absorption maximum (Lax, in nm) ,half life-time (T), activation energy of thermal ring opening (Ea, in kJ/mol) and quantum yield of photochemical ring opening (00). ~~
DHP
cpc
Lax
102 109 130 103 131 113 118 121 124 126 133d 133b 133c 135a' 135b
0.02
448.4 530 466 572 585 410 470 410
0.17 0.06 0.14 0.06 0.23 0.15 0.3 0.2
445
530 423.7 436.7 434 446.4 487.8
~~
36.8 d 0.1 m 7 d 1 h o.Oo05s 3 d 30 m 11 d 320 m 5.8 s 3.6 h 2.0 h 55 h 1 h 39.8 m
~~
0.008
121 50
42 77 (84) 63 42(76) 51 53.3 65 98 86.5
0.03 0.001 1
0.7
60b 59
6oc
60b
6oc
60a 59 60b,61 59 58a 60a.61 61 61 59,60a,61 61
a) measured at 25'C; d = day, h = hour, m = min. b)values in parenthesis were obtained by flash photolysis. *) see Scheme 27.
2.3. Tetraarylethylenes,Dianthrylidenes Irradiation of tetraphenylethylene in the presence of an oxidant forms 9,10-diphenylphenanthrene;a second cyclization does not take place, obviously for steric reasons. By connecting the ortho atoms of the phenyl groups, bonded at the same ethylenic carbon atom, via a carbon atom a class of compounds is formed, known as dianthrylidenes (Scheme 28). Bridging via 0, NR, or S leads to compounds (138), (140) and (141). The thermochromism of dianthrone (137) was already published in 1909 by Meyer (ref.631, whereas the coloration induced by light was described in 1950 by Hirshberg who also introduced the name photochromism for this phenomenon (ref.64). The photochromic behavior turned out to be very complex. Several reviews have appeared about these compounds (ref.65). A recent survey has been written by Fischer, who contributed much to the understanding of the mechanism of the photochromism phenomena (ref.65~).In the description of the behavior of the compounds on irradiation the different colored forms are distinguished by letters. By irradiation of dianthrone (137a) (Scheme 29) at low temperature or at a very high viscosity no coloration is observed. On irradiation at increasing temperature (decreasing viscosity) a colored compound (B) is formed, which has an absorption at 650-730 nm. Its yield increases with increasing temperature until a constant value is reached. Irradiation with light absorbed by B neither prevents the coloration
296
Scheme 28 10’
10
X = CHR :Dianthrylidene (136)
X = CO :Dianthrone (137)
X = S :Dithioxanthylidene(141) nor decreases the color intensity at any temperature. The smcture of B was deduced from the UV and NMR spectra and by comparison of these data with those calculated with the aid of force-field calculations. The conformation of the parent molecule (A) can be described as a combination of two tricyclic structures, both consisting of a cyclohexa-1,rl-diene moiety in a boat conformation and annelated by two benzo p u p s and connected by a double bond in such a way that the two boat forms are situated in opposite directions. In figure 2(A) the Newman projection is given. The formation of B can be ascribed to twisting of the central double bond of A over 50 to 60 degrees. In B the two halves are close to planar (fig.Z(B)). At higher temperatures the dianthrones not substituted at the 1 and 8’ position undergo irreversible photoreactions even under anaerobic conditions (see below). Dianthrones substituted at the indicated positions form only B, but at higher temperatures also the temperature dependent reversed reaction (B-A) occurs; the equilibrium is in favour of A. Analogous observations have been made with dixanthylidene (138) and its derivatives. Irradiations, however, of (136),(139) and (140) did not show photocoloration at any temperature; they were assumed to be nonphotochromic. It turned out, however, that these compounds on irradiation form a species with an absorption maximum at a shorter wavelength than the parent compound A. This modification is denoted “ E . It is light-stable, no photoreversion to A takes place. As with the formation of B a mechanism via a triplet is plausible. Contrary. however, to the situation with the B modification, the E modification cannot be obtained by a thermal reaction, which indicates a larger difference in energy between A and E in solution than between A and B. By the same kinds of argument as used for the elucidation of the structure of B, the smcture of E was deduced and described as a molecule in which both non-planar halves are bent to the same side (fig. 2 E). From diacridylidene, methyl- substituted at the positions 1 and 8’ (14Ob) photoproduct F is formed instead of E . It differs from E by the absence of a hypsochromic shift. Its configuration can be described as that of E but twisted over the central
291
Scheme 29
0
136 a R=Rl=H b R=OH,Rl=H c R=OH,R1=Me
137 a R,=R2=R3=H b Rl=R3=H,R2=Me
@
d R2=H,Rl=R3=Me
0
\
c R2=R3=H,Rl=Me
\
I
k2
0
I
0
138
a Rl=R2=H
139
b R,=R2=Me c Ri=H,R2=OMe
I
Me 140 a Rl = R 2 = H b Rl=Me,R2=H
c Rl=H,R2=Br
141
298
B
A
A
B
E
E
F
F
Fig. 2 Steric formula and Newman projections (perpendicular to the 9-9’ bond) of the modifications A, B, E and F of dianthrylidenes and analogous compounds. The straight lines represent the projections of the phenyl groups. double bond. Of interest are the compounds (138~)and (139) both of which form the two modifications B and F. In both cases B is formed only at -18OOC; it reverts to F at -15OoC, at which temperature F is the only photoproduct. Besides the colored products mentioned the dianthrylidenes and analogous compounds can form photocyclization products, denoted C and comparable with the DHP’s of diarylethylenes. These products are also colored, showing wavelengths of longest absorption at about 800 nm. The long-wavelength absorption is at the same position as that of the B modification described above. Though this similarity and also the thermal reversibility of B and C may be confusing, the two isomers can be distinguished by some properties. So formation of C is, like that of other DHP’s, photoreversible; ring opening occurs by irradiation with visible light, B is unaffected by such irradiation. The photoformation of B is sensitized by triplet sensitizers; that of C cannot be sensitized. As with the formation of DHP’s the quantum yield of formation of C drops on cooling, while that for B is temperature-independent (for dianthrones) or less dependent (for dixanthylidenes). Provided the compounds C are unsubstituted at their cyclization positions they can be dehydrogenated to phenanthrene derivatives. In table 6 the more important photochromic properties of several dianthrylidenes are given. The ratio between the products B and C depends on the temperature and on the polarity of the solvent. Usually polar solvents and higher temperature favour the product C. Dianthrylidenes substituted in both halves of the molecule are supposed to form, in principle, more than one cyclization product. Using UV and N M R specmscopic techniques for the analysis of the photocyclization products from Z- and E- (137c) it was concluded that only (142) is formed and not (143) (ref.65a). (For thermal and photochemical isomerization of these compounds see ref.74). As mentioned above, dianthrones not substituted at the 1 and 8’ positions undergo, even under anaerobic conditions, irreversible photoreactions. It has been
299
TABLE 6. Photochromic data of dianthrylidenes (Scheme 29), wavelength of longest maximum absorption ( &, in nm) of the parent compound (A), @e modifications B, E or F, and of the photo-
cyclization compound (C),with half-life time ('5) and activation energy of ring opening (kJ/mol). Compound Parent A Lax
136a 136b 136c 137a 137b 137c 137d 138a 138b 138c
325 325
300 390 380 380 340 340
139
390
140a 140b 140c 141
420 390 455 360
I
~~~
photo-isomer mdficationB,Eor F
photocyclizationproduct
C
type ("C)
~max
Lax
E (-160) E (-160)
265 350
B (-160)
700
B B*(-160) B (-180)
660 650 650
(-190) (-150) (-180) (-150) (-180) (-180) (-180) (-100)
625 360 675 380 360 390 388 315
500 505 560 600 600 650 675 500 610
B F B F E F E E
0 142
540
'5
("C)
Ea
0.92 s (-20) 1.6 s (-20) 0.52 s (-20) 0.08 s (-75)
52.7 61.5 62.7
2.1 h (-58) 58 mini-50) 13.9 s (55) 4.1 h (-70) . .
58.5 58.5 46 50.2
ref.
66 67 67 65c,68 65a,73 62 65~73 69 70 70 71 71
510
0.7 s (0)
62.7
72
0 143
shown (ref.75) that the endproducts are a phenanthrene-like cyclization product, helianthrone (146). and bianthrol (147). The mechanism of formation of these compounds was shown to PmCeed via the reaction sequence given in Scheme 30. The first step is the photocyclization into (1441, which could only be observed as a transient in the flash-photolysis experiment at -75°C. BYa complex disproportionation mechanism between (145), the en01 form of (144), and (137a) the redox reaction is completed. As can be expected, compounds without a carbonyl group form only phenanthrene-likecompounds on irradiation in the presence of an oxidant.
300
Scheme 30
137a
-
hv
144
0
145
OH
% OH
0
0
0
OH
147
3
SYSTEMS WITH HETERO ATOMS
Substitution of one carbon atom in hexa-l,3,5-triene (1) by a hetem atom (0.Nor S) leads to seven different molecules; substitution of two carbon atoms in (1) by hetero atoms increases the number of possible molecules to about 50 (Scheme 31).Hence, the number of such compounds suited to act as a photochromic substance may be a multiple of that in the all-carbon series.There are, however, restrictions in the number of theoretical possibilities, due to the change of physico-chemical properties, induced by the h e m atoms. In general the introduction of a hetero atom in (1) or (2) will induce a new absorption band, due to the n,z*-transition of a lone-pair on the hetero atom. Compared to the corresponding all-carbon compound this band will usually be shifted towards the visible region. In general the corresponding molar absorption coefficient is small. A consequence of the n,n*-transition may be that excitation of the Compound does not lead to a cyclization reaction of the aiene system, for which a z,z*-transition is required, and another effect may be that due to spin-orbit coupling; the first excited singlet state crosses to a triplet state from which other reactions than cyclization (or ring opening) occur. Several compounds belonging to this class yield hetemyclic five-membered rings on irradiation (ref.76). A difference from the analogous all-carbon systems can also be caused by a difference in
301
bonding energy of the carbon-hetero atom bond compared with a carbon-carbon bond. The Scheme 31
X = 0, N, S Y=N,S effect of this difference may be more important for thermal than for light-induced reactions. Apart from these differences, the prerequisites for a compound to be photochromic, described in the introduction of this chapter, remain equally valid for a hexatriene/cyclohexaene system containing hetero atoms. In view of all these restrictions it is not surprising that many systems with hetero atoms give complex reaction mixtures on irradiation. Moreover, in several cases, photocoloration Can only be observed at very low temperatures. In fact, only a few of the many theoretical possibilities mentioned above have been described as photochromic systems. In contrast to the investigations with all-carbon systems studies on photochromic hetero-hexatrienes were published mainly before 1971 (ref.65b,77). In the following sections a short review will be given. 3.1
Hexatrienes with one hetero atom
To the compounds having a dienone moiety and examined for photochromic properties belong the derivatives of o-vinylbenzophenone (Scheme 32). Huffman and Ullman (ref. 78) demonstrated the orange-coloration of these compounds on irradiation in a glass and at low temperature in several solvents. 'Bleaching occurs on heating and on irradiation with visible light. The authors could prove the presence of the cyclization product (149). Extended irradiation of (148) in benzene resulted in rearranged compounds, whereas in methanol a~ the solvent also methanol adducts were formed. On flash photolysis of (150) a transient showing a maximum absorption at about 445 nm was observed. (The lifetime in deaerated cyclohexane was 18 ms, in ethanol 27 ms (ref.79)). Many more reports have appeared on the photochromism
302
-4 Scheme 32
Ph
I
@
e *\ \
148
150
/
Me
/
Ph
- % Mhve
\
\
149
Me
151
Me Me
Ph
Me
of the reverse reaction viz. the ring opening of 2H-pyranes and 2H-chromenes (Scheme 33). In Table 7 some of the results are given. Strongly related structures are found in the spiropyranes (168) (Scheme 34). Their ring-opened structures are stabilized by resonance. The photochromism of spiropyranes will be treated in chapter 8 . Also related are the isochromenes (169). Investigations on the photochromism of these compounds are not known, though some photochemical reactions have been described (ref. 92). Scheme 33a Ph
Ph
152
153
Some general conclusions can be drawn from the data of Table 7. Benzo-annulation in (154) at the positions 5,6 or 7,8 forming (156) and (158), respectively, increases the stability of the colored form. It is of interest to note again that the related compounds 1,2- and 3.4-dihydrophenanthrene do not show photoreactions caused by ring opening. This may be ascribed to differences in electronic structure and excitation levels. The open form of (154) is also stabilized by substituents at C2, which are conjugated with the n-system of the ring-opened isomer (154f-j). Benzo-annulation at the C6-C7 bond of (154) decreases the stability of the colored form (157). Flash photolysis experiments of some chmmenes (154e,157,159) in ethanol at room temperature and in 2-methyltetrahydrofuraneat 78'K reveal that the formation of a colored product from
303
TABLE 7. Photochromic data of some derivatives of pyranes and chromenes (Scheme 33). ~~~
Compound 152a 152b 154a 154b 154c 154d 154e 154f 154g 154h 154i 154j 156a 156b 156c 156d 156e 157a 157b 158a 158b 158c 159 160 161 163a 163b 165 166a 166b 166c 166d 167
* hl
R Me CHZPh Me Me Me Et Ph p-(Me)&”P p-(W2NPh p-(MehNPh p-(Me)Wh p-OMePh
H Ph Me H H
Rl
k
Me Me H H H H Me Me H H Me Ph (MehNPh H Me NHPh p-RPh
H Ph CN COPh C02Et COzEt
OH CN CN OEt
R2
H H 6-OMe H H H
hl 320 340 318 28 1 28 1 310 310
6-NO2
6-NO2 6-NO2, 8-Br H H Me Ph (Me12NPh
358 360 360
NHR
OCH3
Me Me
363 357 346 345 320
A 2
390 390 no color 340 360 590 500 bluea, violetb greena,blueb violeta, redb 460 430 bluea, violetb 347 490 460 590 570 green blue, red green blue brown, red orange rose yellow
indicates the wavelength of maximum absorption of the starting compound;
ref. 79 79 80 80 80 81 79 82 82 82 82 82 83 84 79 82 85 86 79 87 88 88 79 79 89 89 89 90 91 91 91 91 91
the
wavelength of maximum absorption of the colored product, for compounds showing a very broad absorption band the color is indicated. a in ethanol ,bin toluene as the solvent.
(154e and 157) occurs from the first excited state, from (159) mainly from the triplet state. From the difference in quantum yield of fluorescence at both temperatures it is concluded that some activation energy is required for ring opening (ref.93). The formation of the ring-opened product of (154e) is within 1 ns, that from (157) within 400 ns; their life-times are 3-4 and 400 p, respectively. Chromenes with electron withdrawing substituents give strongly colored isomers on irradiation, in contrast to chromenes containing electron donating groups. Whereas most colored substances are obtained at low temperatures as their life-times at room temperature are too low, some chromenes such as lapachinol (158a) form stable products at ambient temperature. CND0/2S calculations suggest that the multiplicity of the excited state responsible
304
Scheme 33b hv
R 154
155
Rl R2
156
157
158
0
159
Me
160
hv Phh
ph
p CHzPh h 161
Ph
CH2Ph
162
a
R
163
R b-
hv
Q
R 164
Se
1
CN
COPh 165
166
167
for the ring opening depends on the substituents (ref.94,95). Moreover, the role played by the
305
Scheme 34
168
R
169
170
excited states of naturally occurring chromenes in biochemical processes is discussed (ref.95). The solutions of (1%-i) in ethanol and toluene are colorless. Irradiation of (1%-h), at -10°C or lower, results in an intensive color. The compounds (154h-i) are only photochromic at -100°C. in accordance with the effect of substituents at C3 and C4. In general, derivatives of (156) are colorless compounds in the solid state. However, when an akylamino group is present at C2 the colored open ring structure is present in the solid phase. Derivatives of (158) in apolar solvents are present in the chromene form, in the solid state they possess the chinoid form. In polar solvents an equilibrium is present between both isomeric forms. The photochromic behavior then depends on the wavelength of irradiation and the temperature. Exchange of the oxygen atom of pyrane or chromene by a sulfur atom gives thiopyrane and thiochromene respectively. The photochromic properties of these compounds are analogous to those of the oxygen analogs, but there is only a small red shift in the colorless form, and a considerable red shift in the colored product. Flash photolysis experiments (ref.93) with (161) give results comparable with those of (152). The selenochromenes (165) are also photochromic. Again the incorporation of the Se-atom has only a small effect on the absorption spectra of the starting compounds, but a large effect on those of the open isomers: substitution of oxygen by sulfur results in a bathochromic shift of about 100 nm, substitution of sulfur by selene results in a red shift of about 90 nm. Bleaching of the colored form of (165) occurs by warming the solution at about 150°K. Sensitizing or quenching by triplet sensitizers has no effect on the photochromic behavior. The 1,2-dihydroquinolines (166) have photochromic properties comparable to those of the chromenes. The reversibility of the ring opening in compound (166a) is, however, less complete as a consequence of the formation of an aldehyde from the enolic group present in the quinoid structure. The colored form of (167) is stable for a short time at room temperature. 3.2
Hexatrienes with more than one hetero atom
Only two of the five possible combinations of a hexatriene containing an oxygen and a nitrogen atom (Scheme 31) have been described as photochromic systems, viz. Z-$-nitrostyrene (171) (Scheme 35) (or more generally 1-aryl-2-nitroalkenes) and 2H-1,4-oxazines (179). The photochromism of the former class of compounds was investigated extensively some 20 years ago (ref. 96). Though the reversible Z,E-isomerization is the main photoreaction, it could be shown that only the Z-isomer gives rise to a colored product on uv irradiation (Scheme 35).
306
Scheme 35
H
171 R3
R2 a. b. C.
173
d. e. f.
172
Rl
R2
R3
H H Ph Ph H
Me H H
H OMe H H
Ph
Fl?
Ph
H H
Substituents at both ortho-positions of the phenyl group prevent the photochromkm, probably by steric interference. From the results of the spectroscopic and kinetic investigations by flash photolysis with eight derivatives of (171) viz. (173a-h) (ref.96b) the following picture arises. In ethanol as the solvent, the maximum of the single absorption band in the visible part of the spectrum is, dependent on the substituents, between 410 and 440 NU, in cyclohexane between 420 and 450 nm. Bleaching of the colored compound occurs in a first order reaction (ref.96a). ~ ethanol; in cyclohexane, this value is The half-life times at 15OC vary from 10-30 5 ~ 1 0s-in smaller (except for 173d). The corresponding activation energy of ring opening varies, independently of the solvent, from 46 to 66 KJ/mol. More recently, the photochemistry and photochrornism of E-and Z-2-nitro-l-(9-phenanthry1)prop-1-ene (174) has been investigated (ref.97). It was expected that the reduced aromatic character of the phenanthrenic 9,lO-double bond should lead to rather equal stability of (174) and of the ring-closed product (175), so that the lifetime of the latter should be long enough to be investigated by NMR spectoscopy. Extended irradiation of (174) in dioxan gave, in addition to Z,E-isomerization, 9-phenanthryl aldehyde (176) and the furan derivative (177). both in small quantum yields (Scheme 36). However, by flash photolysis photochromkm could be observed. The structure (175) was assigned to the colored product, having an absorption maximum at 405 nm. In ethanol the half-life time at 3OoC was 6.1 s, the activation energy for ring opening 37 KJ/mol; in cyclohexane, these values are 12.9 s and 73 KJ/mol. respectively. By irradiation of a sample of (174) at low temperature in the NMR probe, a change of the position of the methyl peak was observed. Though the structure of (175) could not be elucidated unequivocally from the NMR spectrum, the position of the methyl group of the photoproduct at 6 = 2.2 ppm is in
307
Scheme 36
PMe* \
A
O"Lo
0
175
+ 176
177
agreement with this structure. It is not in accordance with compound (178), which can be obtained by ring closure between the oxygen of the nitro group and the a-C atom of (174) and presented as the product in an alternative mechanism for the photochromism of the P-nitrostyrenes (ref.98). However, compound (178) may be a precursor in the formation of (176). The second class of compounds containing oxygen and nitrogen and showing photochromic properties is formed by the 2H-1,4-oxazines. The representatives (179) and (181) and some related spiro-compounds have been investigated with the aid of flash photolysis in the temperature range of -40°C to -lO°C (Scheme 37) (ref.99). In Table 8, some characteristic spectroscopic and kinetic data of the reverse reaction are given. Upon flash photolysis about 1% of the absorption produced per flash persists. The solutions can be bleached completely by irradiation with red light at -40°C. Obviously, an isomer of the open form exists, which is more stable than the primarily formed product, which can be isomerized by red light. The thermal relaxation of the ring-opened isomers (180) and (182) can be described as a bi-exponential decay (the kinetic data in the table refer to the main term), which may be a consequence of E,Z-isomerization of the open form. The 3-methyl-substituted compounds (179b) and (181b) do
308
Scheme 37
hv
179
hv
8
X" Ar Ar
/
182
181
a. R = H b. R=Me not show photochromism in aprotic solvents; these compounds undergo more photodegradation than the other compounds. These results are in accordance with the general rules given before TABLE 8 Photochromic properties of some 2H- 1,4-oxazines. Wavelength of maximum absorption of the starting compound (L1.in nm); wavelength of maximum absorption (h), and half life time (T, in s) of the open isomer and activation energy (Ea) of the reverse reaction.
k2
Compound
T
ethanol
Ea benzene
ethanol
benzene
53.8
179a
265
610
7 . 8 ~ 1 0 - ~ 0.32
45.4
179b
265
610
0.47
39.1
181a
260
610
5 . 5 ~ 1 0 - ~ 5.5
39.1
181b
260
610
0.56
39.8
for a hexatriene/cyclohexadiene couple to act as a good photochmic system.
58.1
309
REFERENCES. 1 2 3
8 9
a. b. C.
10 a. b. C.
11 a. b. C.
12 13 a. b. 14 a. b.
15 16 a. b. C.
d. 17 a. b. C.
P.J.Vroegop. J.Lugtenburg and E.Havinga, Tetrahedron, 29 (1973) 1393. H.J.C.Jacobs and E.Havinga, Adv. Photochem. 11 (1979) 305. W.G.Dauben, E.L.McInnis and D.M.Michno, Photochemical rearrangements in menes, in Ed P. de Mayo, Rearrangements in ground and excited states, Vol. 3 (1980) 91, Academic Press. J.W.J.Gielen, H.J.C.Jacobs and E.Havinga, Tetrahedron Lett., (1976) 3751. A.M.Brouwer and H.J.C.Jacobs, Tetrahedron Lett., 27 (1986) 1395. A.M.Brouwer,J.Comelisse and H.J.C.Jacobs, Tetrahedron, 43 (1987) 435. W.H.Laarhoven, Photocyclizations and intramolecular cycloadditions of conjugated Olefins, in A.Padwa (Ed.) Organic Photochemistry, Vol. 9, p.129, M.Dekker, New YO* 1987. A.Padwa, L.Brodsky and S.Clough, J.Am.Chem.Soc.,94 (1972) 6767. P.Courtot, R.Rumin and J.Y.Salaun, Recl.Trav.Chim.(Pays Bash98 (1979) 192. P.Courtot and J.Y.Salaun, Tetrahedron Lett.,(1979) 1851. P.Courtot and J.Auffret, J.Chem.Res.(S), (1981) 304. K.R.Huffman, M.Loy, Wm.A.Henderson,Jr. and E.F.Ullman, Tetrahedron Lett.,(1967) 931. E.F.Ullman, Wm.A.Henderson,Jr. and K.R.Huffman, Tetrahedron Lett.41967) 935. K.R.Huffman, M.Loy, W.A.Henderson Jr.and E.F.UUman, J.Org.Chem., 33 (1968) 3469. M.Pomerantz and G.W.Gruber, J.Am.Chem.Soc.,93 (1971) 6615. P.J.Baldry, J.Chem.Soc.Perkin 1, (1975) 1913. P.J.Baldry, J.Chem.Soc.Perkin 2, (1980) 805. W.H.Laarhoven, unpublished results. G.J.Fonken, Chemmd Ind., (1962) 1327. J.Y.Wong, C.Manning and C.C.Leznoff, Angew.Chem.Int.Ed., 13 (1974) 666, and references cited there. A.H.A.Tinnemans and W.H.Laarhoven, J.Chem.Soc. Perkin 2, (1976) 1115. R.J.F.M.van Arendonk, Ph.Fomier de Violet and W.H.Laarhoven, Recl.Trav.Chim.(Pays Bas), 100 (1981) 256. K.R.Huffman, M.Loy and E.F.Ullman, J.Am.Chem.Soc., 87 (1965) 5417. M.Pomerantz, J.Am.Chem.Soc.,89 (1967) 694. J.Meinwald and R.P.Mazmchi, J.Am.ChemSoc., 89 (1967) 696. J.Meinwald and D.A.SeeIey, Tetrahedron Lett.,( 1970) 3739. L.Ulrich, H.J.Hansen and H.Schmid, Helv.Chim.Acta, 53 (1970) 1323. M.Sindler Kulyk and W.H.Laarhoven, J.Am.Chem.Soc., 100 (1978) 3819. P.M.op den Brouw and W.H.Laarhoven, J.Chem.Soc.,Perkh 2, (1983) 1015. P.M.op den Brouw, P.de Zeeuw and W.H.Laarhoven, J.Photochem. 27 (1984) 327.
310
d. W.H.Laarhoven and Th.J.H.M.Cuppen, J.Photochem. 32 (1986) 105. 18 A.J.W.To1 and W.H.Laarhoven, J.Org.Chem.,51(1986) 1663. 19 P.M.op den Brouw and W.H.Laarhoven, J.Org.Chem., 47 (1982) 1546. 20 a. S.W.Horgan, D.D.Morgan and M.Orchin, J.Org.Chem.38 (1973) 3801. b. P.H.G.op het Veld, J.C.Langendam and W.H.Laarhoven, Tetrahedron Lett.,(1975) 23 1. C. R.Lapouyade,R.Kousini and J.C.Rayez, J.Chem.Soc.,Chem.&m., (1975) 676. d. S.Lazare,R.Lapouyade and R.Bonneau, J.Am.Chem.Soc..lO7 (1985) 6604. 21 a. P.H.G.op het Veld and W.H.Laarhoven, J.Chem.Soc.Perkin 2, (1978) 915. b. P.H.G.op het Veld and W.H.Laarhoven, ibid (1978) 922. C. P.Fornier de Violet, R.Bonneau. R.Lapouyade, R.Koussini and W.R.Wm, J.Am.Chem.Soc.,lOO (1978) 6683. d. S.Lazare, R.Lapouyade and R.Bonneau, J.Amer.Chem.Soc. 107 (1985) 6604. 22 a. T.Sato and T.Morita, Bull.Chem.Soc.Jpn., 45 (1972) 1548. b. Y.Fujioka, Bull.Chem.Soc.Jpn., 58 (1985) 481. 23 a. K.Salisbury, Tetrahedron Lett.,(1971) 737. b. R.C.bkson, S.M.de B.Costa and J.Hudec, J.Chem.Soc. Chem. Comm., (1969) 1272. D.A.Seeley, J.Am.Chem.Soc.. 94 (1972) 4378. C. 24 a. H.Kleinhuis, R.L.C.Wijting and E.Havinga, Tetrahedron Lett., (1971) 255. b. H.Heimgartner, L.Ulrich, H.J.Hansen and H.Schmid, Helv.Chim.Acta, 54 (1971) 2313. W.Sieber, H.Heimgartner, H.J.Hansen and H.Schmid, ibid., 55 (1972) 3005. C. d. U.Widmer, H.Heimgartner and H.Schmid, ibid..58 (1975) 2210. 25 P.Cowtot, R.Rumin and J.Y.Salaun, Pure and Appl.Chem., 49 (1977) 317. and references cited there. 26 W.H.Laarhoven and N.Berendsen, Rec.Trav.Chim.(Pays Bas), 105 (1986) 367. 27 a. J.J.M.Lamberts and W.H.Laarhoven. J.Am.Chem.Soc.,l06 (1984) 1736. b. J.J.M.Lamberts and W.H.Laarhoven, Rec.Trav.Chim.(Pays Bas). 103 (1984) 131. C. W.H.Laarhoven. F.A.T.Lijten and J.M.M.Smits, J.Org.Chem., 50 (1985) 3208. J.Woning, F.A.T.Lijten and W.H.Laarhoven, Book of abstracts, XIIth International 28 Conference on Photochemistry, Tokyo (1985) 348. 29 a. J.B.M.Somers, A.Couture, A.Lablache Combier and W.H.Laarhoven, J.Am.Chem.Soc., 107 (1985) 1387. b. J.Woning and W.H.Laarhoven, Abstracts ~01.11. Xmth Internatinal Conference on Photochemistry, Budapest (1987) 335. W.H.Laarhoven, Recl.Trav.Chin(Pays Bas),l02 (1983) 185. ibid, 102 (1983) 241. 30 a. b. F.B.Mallory and C.W.Mallory, Org.Reactions, 30 (1984) 1. C. W.H.Laarhoven, Photocyclizations and intramolecular cycloadditions of conjugated aryl-olefines. in A.Padwa (Ed.) Organic Photochemistry. Vol 10. Marcel Dekker, New York (1989). J.Saltie1, J.D’Agostino. E.Dennis Megarity, L.Metts, K.R.Neuberger, M.Wrighton 31 and 0.C.Zafirion. The cis-tram isomerization of olefins, in 0 .L.Chapman (Ed.), Organic Photochemistry Vol3, (1963) 1; M.Dekker, New York.
311
32 33 a. b. 34 a. b. 35 36 37 a. b. 38 39 40 a. b. 41 a. b. c. 42 43
44 45
46 47 a. b. 48 49 50 51 52
53 54 a. b. 55 56 a. b. 57
K.A.Muszkat, Topics Curr.Chem.,88 (1980) 89. K.A.Muszkat and E.Fischer, J.Chem.Soc.B,( 1967) 662. G.Seger and T.A.Pakkanen, J.Chem.Soc.,Perkin 2, (1975) 1515. Th.J.H.M.Cuppen and W.H.Laarhoven, J.Am.ChemSoc.,94 (1972) 5914. Ch.Goedicke and HStegemeyer, Chem.Phys.Lett. 17 (1972) 492. K.H.Grellmann, P.Hentzsche1, T.Wismontski-Knittel and E.Fischer, J.Photochem., 11 (1979) 197. E.W.Forster,K.H.Grellmann and H.Linschitz, J.Am.Chem.Soc., 95 (1973) 3108. B.S.Green and.Rejto, J.Org.Chem.,39 (1974) 3284. G.KauppandH.Stark,Angew. Chem. Int.Ed., 17 (1978)718. F.D.Lewis, Adv.Photochem.,l3 (1986) 165. W.H.Laarhoven,A.Oudenampsen and J.Woning, Abstract Vol.1, XIIIth International Conference on Photochemistry, Budapest, (1987) 119. T.D.Doyle, N.Filipescu, W.R.Benson and D.Banes, J.Am.Chem.Soc., 92 (1970) 6371. T.D.Doyle, W.R.Benson and N.Filipescu, J.Amer.Chem.Soc., 98 (1976) 3262. P.H.G.op het Veld and W.H.Laarhoven, J .Am.Chem.Soc., 99 (1977) 7221. M.V.Sargent and C.J.Timmons, J.Chem.Soc., (1964) 5544. G.Rio and J.C.Hardy, Bull.Soc Chim.Fr., (1970) 3578. T.Wismontski-Knittel, K.A.Muszkat and E.Fischer, Mol.Photochem., 9 (1979) 217. K.Ichimura and S.Watanabe, Bull. Chem. Soc. Japan, 49 (1976) 2220. A.Bromberg, K.A.Muszkat and E.Fischer, Israel.J.Chem., 10 ( 1972) 765. K.A.Muszkat, D.Gegiou and E.Fischer, J.Chern.Soc.Chem.Com., (1965) 447. K.A.Muszkat, H.Kesse1 and S.Shardi-Ozeri, Israel.J.Chem., 16 (1977) 291. W.H.Laarhoven, Pure and Applied Chem., 56 (1984) 1225. J.B.M.Somers and W.H.Laarhoven, J.Photochem.,Photobiol.,A Chemistry, 40 (1987) 125. F.B.Mallory and C.W.Mallory, J.Am.Chem.Soc., 94 (1972) 6041. B. Oudenampsen, W.H.Laarhoven, unpublished results. R.Naef and E.Fischer, Helv.Chim.Acta, 57 (1974) 2224. C.E.Ramey and V.Boekelheide, J.Am.Chem.Soc., 92 (1970) 3681. H.R.Blattmann, V.Boekelheide, E.Heilbronner and J.P.Weber, Helv.Chim.Acta, 50 (1967) 68.,and ref. cited there. H.R.Blattmann and W.Schmidt, Tetrahedron, 26 (1970) 5885. R.H.Mitchel1, R.J.Carruthers, L.Mazuch and T.W.Dingle, J.Am.Chem.Soc.,l04 (1982) 2544. R.H.Mitchel1, J.S.H.Yan andT.W.Dingle, J.Am.Chem.Soc.,l04 (1982) 2551. J.Wirz, G.Persy, E.Romme1, 1.Murata and K.Nakasuji, Helv.Chim.Acta 67 (1984) 305. W.H.Laarhoven, Th.J.H.M.Cuppen and R.J.F.Nivard, Rec.Trav. Chim. (Pays Bas), 87 (1968 ) 687. W.H.Laarhoven, Th.J.H.M.Cuppen and R.J.F.Nivard, Tetrahedron, 26 (1970) 4865. K.A.Muszkat, G.Seger and Sharafi-Ozeri, J.Chem.Soc. Faraday Trans. 11, 71 (1975)
312
58 a. b. 59 60 a. b. C.
61 62 63
64 a. b. 65 a. b. C.
66 67 68 69 70 71 72 73 a. b. 74 75 76 77 78 79 80 81 a. b. 82 83
1529. T.Wismontski and E.Fischer, Mol.Photochem., 9 (1978/79) 67. W.H.Laarhoven, ThJ.H.M.Cuppen and RH.K.Brinkhof. Tetrahedron, 38 (1982) 3179. W.H.Laarhoven, Th.J.H.M.Cuppen, N.Caste1 and E.Fischer, J.Photochem., 33 (1986) 297. T.Wismontski-Knittel,G.Rscherand E.Fischer, J.Chem.Soc. Perkin II, (1974) 1930 . T.Wismontski-Knitteland E.Fischer, J.Chem.Soc. Perkin 11, (1979) 449. T.Wismontski-Knittel, M.Kaganowitch, G.Seger and E.Fischer, Rec.Trav.Chim.(Pays Bas), 98 (1979) 114. E.V.Blackburn. C.E.Loader and C.J.Timmons,J.Chem.Soc.(C),(1970) 163. E.Fischer, W.H.Laarhoven and W.Luettke, Abstracts Vol.II, XDXth International Conference on Photochemistry, Budapest,(1987) 3 10. H.Meyer, Ber.42 (1909) 143. Y.Hirshberg and E.Fischer, J.Chem.Soc., (1953) 629. Y.Hirshberg, J.Am.Chem.Soc.,78 (1956) 2304. T.Bercovici, R.Korenstein and E.Fischer, Pure Appl.Chea, 24 (1970) 531. J.D.Margerum and L.J.Miller, Photochromic processes by tautomerism. in G.H.Brown (Ed.), Photochromism. Chapter V. Techniques in Chemistry, Vol. 3 (1971). E.Fischer, Rev.Chem.Intermed.. 5 (1984) 393. R.Korenstein, K.A.Muszkat and G.Seger, J.Chem.Soc.Perkin 11, (1976 ) 1536. R.Korenstein, K.A.Muszkat and E.Fischer. J.Chem.Soc. Perkin II. (1977) 564. R.Korenstein, K.A.Muszkat and E.Fischer. Helv. Chim.Acta, 59 (1976) 1826. R.Korenstein, K.A.Muszkat, M.A.Slifkin and E.Fischer, J.Chem.Soc. Perkin II, (1976) 438. R.Korenstein, K.A.Muszkat and E.Fischer, J.Photochem., 5 (1976) 447. R.Korenstein, G.Seger, K.A.Muszkat and E.Fischer, J.Chem.Soc. Perkin 11, (1977) 550. R.Korenstein, K.A.Muszkat and E.Fischer, J.Photochem. 5 (1976) 345. G.Kortuem and G.Bayer, Angew. Chem. 75 (1963) 96. G.Kortuem, Ber.Bunsenges.Physik.Chem., 67 (1963) 24. Y.Tapuhi, 0.Kalisky and LAgranat, J.Org.Chem.,44 (1979) 1949. R.Korenstein, K.A.Muszkat and E.Fischer. Helv.Chim.Acta, 59 (1976) 1826. M.V.George, A.Mitra and K.B.Sukumaran, Angew. Chem. Int. Ed.,19 (1980) 973. R.C.Bertelson, in G.H.Brown Ed., Techniques of chemistry Vol III, Photochromism, Wiley Sons, New York, 1971. K.R.Huffma~~ and E.F.Ullman, J.Am.Chem.Soc., 89 (1967) 5629. C.Lenoble and R.S.Becker, J.Photochem., 33 (1986) 187. E.Davin, C.Balny and R.Guglielmetti, Comt.Rend.Acad.Sci.Paris,ser.C. (1972) 79. RSBecker and J.Michl, J.Amer.Chem.Soc., 88 (1966) 5931. R.S.Becker, E.Dolan andD.E.Balke, J.Chem.Phys., 50 (1969) 239. D.J.Zwanenburg and Th.A.M.M.Maas, Rec.Trav.Chim.(Pays Bas), 95 (1976) 97. D.R.Keams and W.A.Case, J.Amer.Chem.Soc.. 88 (1966) 5087.
313
84 85
86 87 88
89 90 91 92 93 94 95 a. b. 96 a. b. 97 98 99
C.Balny, M.Hannez.0 and A.Hinnen, J.Chim.Phys., 64 (1967) 1815. I.M.Andreeva, E.A.Bondarenko, N.V.Volbushko, M.I.Knyazhanskii, E.A.Medyantseva, A.V.Metelitsa,V.I.Minkin and B.Ya.Simkin, Khim.Geterotsikl.Soedin., 8 (1980) 1035; Chem.Abstr., 94 (1981) 29655~. LIwai and J.Ide, Chem.Pharm.Bull., 10 (1962) 926. J.Kolc and R.S.Becker, Photochem.Photobiol., 12 (1970) 383. LM.Andreeva, O.M.Babeshko, E.M.Bondarenko, N.V.Volbushk0, M.LKnyazhanskii, E.A.Medyantseva, A.V.Metelitsa and V.I.Minkin, Khim.Getemtsikl.Soedin.,1983) ( 1031; Chem.Abstr., 100 (1984) 102508~. R.S.Becker and J.Kolc, J.Phys.Chem,72 (1970) 997. B.S.Lukjanow, M.I.Knjazschanski, J.W.ReWinski, L.E.Niwor0zschk.h and W.I.Minkin, Tetrahedron Lett.,(1973) 2007. J.Kolc and R.S.Becker, J.Amer.Chem.Soc., 91 (1969) 6513. A.Padwa, A.Au, G.A.Lee and W.Owens, J.Org.Chem.,40 (1975) 1142. C.Lenoble and R.S.Becker, J.Photochem., 33 (1986) 187. B.Tinland, RGuglielmetti and O.Chalvet, Tetrahedron, 29 (1973) 665. M . B b , Czechoslov.Chem.Commun.,43 (1978) 416. M.B&ta, Coll. Czechoslov. Chem. Comm., 43 (1978) 3339. J.A.Sousa and J.Weinstein, J.Org.Chem., 27 (1962) 3155. J.A.Sousa, J.Weinstein and A.L.Bluhm, J.Org.Chem.,34 (1969) 3320. R.A.Humphry-Baker, KSalisbury and G.P.Wood, J.Chem.Soc.Perkin a,( 1978) 659. J.T.Pinhey and E.Rizzardo, Tetrahedron Lett. (1973) 4057. U.W.Grummt, M.Reichenbaecher and R.Paetzold, Tetrahedron Lett.,22 (1981) 3945.
See AdditionalLiterature(1989 - 2001): Diarylethanes,A23
314
Chapter 8
4n+2 Systems: Spiropyrans
R. Guglielmetti
1. INTRODUCTION
1.1. Historical survey Although the thermochromism of spiropyrans was noted in 1921, their photochromic properties (photocoloration in the indoline series) were not known prior to the observation by Fischer and Hirshberg, in 1952 (ref. 1) ; the same authors, and, independently, Chaud6 and Rumpf (ref, 2 ) discovered the photochemical reverse reaction. A few years later, Hirshberg's conception of a photochemical binary element for a computer memory (refs. 3, 4 ) and the possibility of getting a variable density optical shutter initiated an intense research activity in this field in industrial and academic laboratories world-wide (refs. 5, 6). The spiropyrans or spirochromenes and related compounds (e.g. spirooxazines) continue to arise a strong interest in connection with their applications in various fields such as : non silver high resolution photography (ref. 7), optical devices and variable transmission materials (refs. 8, 9 ) photovoltaic (ref. 10) and holographic (ref. 1 1 ) systems.
Some important articles on the thermochromism or photochroof spiropyran systems appeared between 1948 and 1970 (refs. 12-24) but the most comprehensive and important review was written by Bertelson in the book "Photochromism" edited by Brown (ref. 25) in 1971. The purpose of this chapter is to be complementary to Bertelson's account (ref. 25) and to emphasize the new insights and developments in the field since that date. mism
315
Definition and presentation of the spiropyrans and related comnounds The spiropyrans or spirochromenes present two heterocyclic parts linked together by a common tetrahedral sp3 carbon atom (Fig. 1 ) . 1.2.
I
Fig. 1. Sketch of a spirochromene showing the two orthogonal parts of the molecule in planes P an PI. The left part is depicted by H (Heterocyclic). The two halves of the molecule are in two orthogonal planes (P and PI). The benzopyran or 2H-chromene part is the common structure to all spiropyran compounds, essentially f o r synthetic : indeed salicylic aldehydes or orthohydroxylated reasons aromatic aldehydes are easily accessible synthetic intermediates. The H heterocyclic part is variable and often is built upon mono or bi-heteroatomic azaheterocycles saturated or benzofused. In solution, the spiropyrans present an absorption spectrum in the UV range 200 - 400 nm with an actinic band mostly situated near 320 to 380 nm. The absorption in this range leads, by cleavage of the carbon-oxygen bond, to the formation of colored isomers called "open form" ( O , F , ) as opposed to the "closed form" (C.F.) which is colorless (Scheme 1).
316
R3
R'
C.F. spiropyran
O.F. photomerocyanine
Scheme 1 , Photochromic interconversion in the spiropyran series. Then a conjugation between the 2 halves of the molecule is made possible which results in a shift of the U.V. absorption in the visible region. The "open form" isomer called photomerocyanine because of its transoYd structure, generally the most stable one, is quite similar to that of merocyanine dyes. A spiropyran may be defined as a potential dye generated by photochemical induction. The reversion to the closed uncolored form may be a fast process under thermal conditions (kA) or by photochemical path either by visible absorption (hY2) or also by U.V. absorption (hY1) when the photochromic system is under continuous irradiation (ref. 2 6 ) . Besides, a very long irradiation can lead to photodegradation products. Atmospheric oxygen facilitates this photodegradation by radical processes (ref. 2 7 ) . Use of the flash photolysia technique for the study of photocoloration-photobleaching phenomenon is appropriate as it reduces the secondary effects of fatigue. The photochromism of spiropyrans has no s y s tematic basis but rather depends on : - the structure of the compounds (nature of the heteroatom and substituents in different positions of the molecule and some positions are fundamental and characteristic for the results) - the medium (solvent, viscosity...) - the temperature, the photolysis energy and the range of absorption of the open form. The appearance of photochromism for an observer depends also on the coloration intensity (optical density of the colored form
317
which is a function of the concentration). These parameters are quantitatively observable by spectrophotometric methods. The quantitative determination of spectro-kinetic or photo-
chemical parameters is generally achieved in solution but many applications require use of plastic or other solid supports such
as paper, films, semi-conductors or polymeric materials. The formation of spiropyran structures is also linked to stereo-electronic parameters (nature of heteroatoms and substituents) (ref. 2 8 ) . Indeed, we can obtain, by direct synthesis, permanent merocyanine dyes absorbing in the visible region, without help of a photochemical process, or we can obtain also such tricyclic compounds corresponding to the bicondensation of intermediate substrates and depending upon the type of substitution and of the reaction conditions. Scheme 1 gives a very simple picture of the photochromic equilibrium but the real mechanism is more complex. For a complete understanding of the photochromic equilibrium, it is necessary to take into account the electronic structures of the open forms (depending on the molecular structure and the medium) as well as their geometries (different stereoisomers of cisoid configuration just after the breaking of the C-0 bond and of transoYd configuration after isomerization). The trans configuration appears to be more stable because it minimizes the non-bonding interactions. Finally we briefly discuss some fundamental studies related to this problem. The electronic distribution in transoid photomerocyanines is often symbolized by a delocalized repartition of n electrons with an excess of negative charge on the phenolic oxygen and an excess of positive charge on the heterocycle H or more particularly on the heteroatoms. This representation which corresponds to the experimental situation seems the most satisfactory. In such cases the 2 mesomeric forms (Scheme 2 ) - dipolar zwitterionic with localized charges - apolar polyenic or quinonic
may have an important contribution.
318
f
auinonic
dipolar
Scheme 2 . Electronic distribution on different mesomeric forms of the transoxd photomerocyanine. The real struoture is a hybrid of (A) and ( B ) and has partial charges 8 + on the spiran carbon and 6- on the phenolic oxygen (see scheme 3 ) . At last in the case of an unsymmetrical heterocyclic system that is generally the case, four "gauche" or twisted cisoid stereoisomers (the heterocyclic H and the phenolate planes showing a certain dihedral angle) and four pseudoplanar transoid stereoisomers could be written (cf. Scheme 3 ) .
H,
The planar cis stereoisomers are not energetically equal.
319 Scheme
3.
Representation
of
photomerocyanine stereoisomers.
all
possible cisoid and transoid
06-
320
The main spiropyran structures which have been synthesized are represented in Table 1 , but many other "sophisticated" series have been prepared a s mentioned by Bertelson (ref. 25). Our classification is based on the nature of the "left" heterocyclic part (H) divided into the azaheterocyclic series (I) and non-azaheterocyclic series (11). TABLE 1. Basic structures of spiropyran compounds,
Heterocyclic (H)-Spiro-Benzopyran or Naphthopyran
( I ) Azaheterocyclic part
- Indoline
:
-
Benzothiazoline :
-
Benzoxazoline :
-
Benzoselenazoline :
I
I
321
-
1,3-Thiazolidine :
-
l13-Oxazolidine :
-
Pyrrolidine :
-
1,S-Thiazine :
-
1,4-Thiazine :
-
l13-0xazine :
-
Piperidine :
-
Phenanthridine :
-
Quinoline :
Q I
G I
@$k I
322 (
I 1 1 Non-azaheterocuclic gart.
-
Benzodithiole :
-
Beneosathiole :
-
Benzopyran or 2H-Chromene
S
;
o r Naphthopyran :
-
Xanthene :
-
Dithiole :
This list does not exhaust of the possible the “left“ part o f spirochromenes formulae.
( t )
BIPS
= acronym
structures
f o r 1’,3’,3’-trimethylspiro 12H-11
ben~opyran-2,2~-indoline
for
a;c.)EJ CH3,
,CH3
CH,
323 2. SPECTROKINETIC PROPERTIES OF PHOTOCHROMIC INTERCONVERSIONS IN SOLUTION
The spiropyrans are usually not photochromic in the solid state but show this phenomenon in solution. In a practical sense,
the term "solution" includes not only dilute fluid solutions, but also solutes in more rigid media such as gels, plasticized resins, films, and bulk plastic solids, for example polymethylmethacrylate. The special case of solvent glasses at extremely low temperatures is discussed particularly in Bertelson's review (refs. 25, 29). Spiropyrans, according to their structure, exhibit photochromism in solvents of any polarity from water to hexane. Most solution studies have been carried out in toluene, ethanol, dioxane or acetonitrile. The solubility of relatively simple spiropyrans is poor in alcohols and aliphatic hydrocarbons, rather good in aromatic and intermediate in ethers, ketones and esters, Of course, groups such as sulfonate, carboxylate, quaternary ammonium or long chain alkyl can be incorporated into the molecule to alter the solubility as desired (ref. 30). Typically, a photochromic spiropyran is a nicely crystalline colorless or pale yellow solid. Its solutions are colorless or may become weakly colored by sunlight and, upon irradiation with UV light, develop color or become more intensely colored. The colored solutions fade thermally back to their original state. The colored solution in polar solvents can often be bleached by visible light. The photocoloration-thermobleaching or photocoloration-photobleaching cycles can be repeated many times typically around 100 or 1000, occasionally less than 5 or more than 10000. A few spiropyrans, especially those bearing free hydroxy, carboxy or amino groups on either ring, exhibit the so-called "negative" photochromism, i.e. they give in the dark moderately colored solutions, that are reversibly bleached with UV-light (refs. 31-36). This effect is sometimes strongly temperaturedependent. For example, ethanolic solutions of 6-hydroxyindolinospiropyran are bleached with visible light and recolored with UV-
light at -88'C but do not undergo the reaction at room temperature (ref. 36). Many spiropyrans are observably thermochromic, that is their solutions become colored upon heating. When such thermally
colored solutions are photobleached, the color reappears when the
324
irradiation stops. The colored material produced thermally appears to be spectroscopically identical with that produced photochemically. These generalizations are most readily explained by assuming that an interconversion exists between the colorless and colored forms of the spiropyran and that the reaction in either direction can occur both photochemically and thermally.
Spiropyran Colorless or Closed form (C.F.)
hYi (UV),kid b 4
hYz (visible)
Photomerocyanine Colored or Open form (O.F.)
kz,
This suggests that the behavior of a spiropyran solution can be partly described by the following parameters : the absorption spectra of the two forms, the quantum yield or the "colorability" (which will be defined later) as a function of wavelength for the reaction in each direction, the rate of the thermal reaction in each direction and the thermodynamic parameters. The real situation, of course, is much more complex as we pointed out previously. One approach in spiropyran research has been to prepare and test a large number of compounds in order to discover empirical correlations between the structure and the photochromic parameters as well as environmental effects (nature of the solvent, presence of oxygen or acidity of solution). Another approach has been the detailed study of the photochemistry and molecular spectroscopy of a few spiropyrans in order to get a better understanding of the processes involved (cf. chapter 2, Gauglitz). But prediction of photochromic behavior is often risky because we may find compounds which invalidate the correlations. The spectrokinetic properties of the photochromic interconversion in the ground state are generally investigated using a flash-photolysis apparatus equipped with an excitation lifetime in the range lms-lns. Classical discharge tubes may be used having a wide UV-visible excitation spectrum ; it is possible also to select band passes with appropriate filters ; high monochromatic energy laser photolysis may also be used with a
325
brief impulse time. The thermobleaching curve is represented as f o l J . o w s (Scheme4 ) . The maximum absorbance Ao is independent of the bleaching kinetics and is spectrophotometrically observable.
I
t
irradiation
Scheme 4. General thermobleaching curve of a photomerocyanine. Under continuous irradiation, a photostationary state is reached with a maximal absorption (Am) which is a function of the thermal bleaching kinetics and of photodegradation (Scheme 5).
4
Absorbance
I
eauilibrium
4I
uv irradiation
coloralion
4 I
decoloralion
switching off irradiation
Scheme 5 . Photochromic photostationary equilibrium.
326
2.1 Spectrokinetic and thermodynamic parameters 2.1.1 Kinetics of the thermal reactions in solution
The complexity and unpredictability of the spiropyrans behavior is strikingly demonstrated by kinetic studies of thermal fading in solution. The kinetics observed shortly after maximum photocoloration is entirely different from that observed at longer times after coloration, Changes in solvent polarity or viscosity have drastic and opposite effects on certain aspects of the observed behavior and conclusions true for one kind of spiropyran may be false for another one. Two general types of studies have been made about the thermal fading of a colored solution of a spiropyran. One type concerns kinetic behavior on a "long" time scale and has been performed on fluid solutions near room temperature, by measuring with an ordinary spectrophotometer, absorbance versus time over periods extending from tenths of seconds to many hours or days, in other words, simple classical kinetio experiments. The second general type of kinetic study has been performed at low temperatures in highly viscous media, where absorbance changes during times less than 1 sec were followed by the techniques of flash spectroscopy. Such experiments show kinetic behavior on a "short" time scale. Unfortunately, few compounds have been examined by both methods, and rarely has a single research group made both types of study. The results from these two types of studies suggest that spiropyrans exhibit a complex "fast" behavior owing to a rapid series of interrelated photo or thermal reactions immediately after coloration, and a fairly simple "slow" behavior at longer times (Scheme 6 ) . The relative rates of the reaction involved in the "fast" and "slow" phenomena are generally such that, to a first approximation, each type of behavior can be ignored when studying the other. The thermal kinetics over "long" times have been generally the most studied because they take into account the structure of the compounds and correspond to the potential applications of photochromism. These thermal fading rates are in agreement with a first order kinetic law.
327
A
0
-----
Scheme 6 . Typical curve for a spiropyran solution fast kinetic process followed by a slow process.
exhibiting
a
detailed study of thermal fading rates corresponding to the conversion of photomerocyanines into spiropyrans over "long" times and over "short" times was developed by Bertelson (ref. 2 5 ) involving complex equations taking into account many parameters. Different rate equations are postulated according to the hypothesis that various entities e.g. the closed and open forms are in equilibrium. Low temperature studies particularly, lead to a range of mechanistic interpretations depending on the research group but also on the polarity or viscosity of the medium (refs. 2 9 , 3 2 , A
37-54).
-
The thermobleaching constant (ka) and the maximum wavelength open form (photomerocyanine) can be determined by the flash-photolysis technique which largely avoids secondary photodegradation phenomena, readily observed when longer times of irradiation are used,
( A s ) of the
Principle of the "spectrokinetic" method If the photochromic cycle is represented by simplified scheme : 2.1.1.1
the
following
1
C . F . -c O.F. ,El )
2
(A2 J E 2
)
Reaction 1. corresponds to the photocoloration. Reaction 2 is the bleaching reaction which is mainly of thermal origin ( k b ) .
328
When the spiropyran is used at low concentrations (lo-4 to M ) the thermal bleaching reaction is of 1st order towards the substrate and the kinetic law has the form : dCoF/dt = -kp, C O F (1) C O F = concentration of Open Form (colored) = time t The .variation of C O F may be followed by measuring the optical density (O.D.) of the solution at the maximum of the absorption wavelength of the open form. In the usual concentration range used for investigat ( 1 0 - 4 to 1 0 - 5 M ) the Beer-Lambert law is valid and equation is transformed into equation (2) : dD/dt =-kAD Or Log D/DO = - kp,t 2.1.1.2 Determination of rate constants and hmax of the oDen form Plotting LogD = f(t) leads to the determination of the thermal bleaching rate ( k A ) which may be obtained graphically or by linear regression using a computer. The half lifes may be expressed directly from rate constants by the relation : 7112 = o.69/kA (3) The absorption maximum wavelength of the photomerocyanine is taken from the visible spectrum obtained with a fast spectrophotometer. The photochromic sample is irradiated by two flash lamps which produce a discharge with an energy of about 500 J in a few microseconds, but with more recent apparatus,energies in the 50 J to 100 J range can be used. The light beam from the source passes through the sample in a thermostated cell and by reflexions is directed either towards a rapid scanning spectrophotometer g. Warner and Swasey type (refs. 55-57) for determination of the visible absorption spectrum of photomerocyanine or to a monochromator-photomultiplier-oscilloscope system for kinetic analysis at constant wavelength (cf. Scheme 7). Another type of spectrometer equipped for determination of
hmax and rate constant was described by McQuain et al. (refs. 37, 5 8 ) . Simultaneously a combined system was also described by St Gobain’s group (refs. 26, 27, 39, 41, 48) and by Guglielmetti and Mosse (ref. 5 9 ) .
329
F---kE-
L.S: : Light Source F. : Flash Lamp Reaction Cell C. : C.C. : Capacitor
-w.s’.
D.U. : Delay Unit W.S.
0. :
: Warner and Swasey Spectrometer
Oscilloscope
Scheme 7 . General scheme for a kinetic apparatus. They used a kinetic bench for the determination of rate constant at fixed Amax. The UV-visible spectrum was obtained with a monochromator and using a film and a microdensito meter. 2.1.2 Thermodynamic and kinetic parameters
The activation energy (Ea) of the thermal fading reaction is derived from the Arrhenius equation : kA = A exp(-Ea/RT) or lOgkA = logA -Ea/RT andAH# = Ea-RT (4) The activation energy is obtained in the usual way by determining the rate constants at different temperatures. Because of the complexity of the reaction process involved in the reformation of spiropyran, the activation energy obtained is called the apparent activation energy (cf. section 2 ) . The values determined are between 14 and 30 kcal. mole-1 depending on the structure and of the type of solvent (ref. 25). The activation enthalpies and entropies are obtained by application of transition state theory and the Eyring equation : log kA/T = AHf.10s/4.57 x 1/T t AS#/4.57 t 10.319 (5) AH+ in kcal.mole-1 ; A S # in e.u. Access to thermodynamic parameters of the photochromic equilibrium ( A G O , AHO AS0 ) is not easy because equilibrium constants
330
(Keq = k decol./k col.) are not known accurately and are dependent on the extinction coefficients of the colored forms (ref.25). 2.1.3 Spectral properties
2.1.3.1 Spectra of colorless form In the colorless form of a spiropyran, the spirocarbon atom is assumed to prevent direct conjugation between the two halves
of the molecule. To a first approximation, then, the UV spectrum of a spiropyran should be the sum of those two portions of the molecule. In fact this is not quite true (Fig. 2). 0.6
0.4
.
0)
rn c
U
%
3 0.2 0 200
250
Wavelength (nm)
300
i0
Fig. 2. Ultraviolet absorption spectra of BIPS (solid curve) 3x10-5 M , and of an equimolar mixture of 2,2-dimethy1-2H-lbenzopyran and 1,2,3,3-tetramethylindoline (dashed curve) in ethanol solution. From Fox (ref. 32) taken from R.C. Bertelson "Photochromism", G.H. Brown edit. Chap. 111. J. Wiley and Sons Inc. Copyright owner. N.Y. 1971. The poor agreement is probably caused principally by the following three factors : (a) The spiro atom may not provide complete electronic insulation of the chromophores of the two halves of the molecule so that the two chromophores may interact. ( b ) The solvation effects on the spiro compound may be different from those on the components and (c) Direct electronic interaction may occur between the indoline portion and the benzene ring of the pyran portion , which, though nonbonded, is geometrically constrained to lie near the N-CH1 group. Similar spectral results are found for benzothiazolinospiropyran (ref. 60).
331
Generally the spectrum of the colorless spiropyran form bet-
ween 220 and 375 nm is quite independent of the nature of the solvent. At temperatures of 77'K, the spectra of colorless spiropyrans in E.P.A. (Ether/5-pentane/5-alcohol/2) or 3-methyl pentane glasses, are somewhat different from those observed at room temperature. The bands are often better resolved and transitions are a little red shifted. Becker (ref. 6 1 ) assigns the transitions at 312 nm to indoline and those at 325, 340 nm to benzopyran. 2.1.3.2 Spectra of colored form
Obtaining a true absorption spectrum of a pure colored form of a spiropyran in an ordinary spectrophotometer is usually a very difficult task. In solution,most colored forms thermally fade at a rate that is significant compared to the time required to record the spectrum. Flash spectroscopic techniques and the development of rapid scanning spectrophotometers have overcome this problem somewhat. A Beckman DK-1 spectrophotometer, modified for measurements of photochromic solutions is described .in Chapter I11 of Bertelson review (refs. 25, 58). Currently, a convenient method for registering the spectrum of the colored form is to use a rapid scanning spectrophotometer such as the Warner and Swasey type (ref. 5 5 ) interfaced with a computer. One method for the evaluation of molar absorptivity E uses 5,7-dich1oro-6-nitro-1,3"-trimethy1spiro[2~-11benzopyran-2,2'
indoline and the formation of its colored form by U.V. irradiation in a non polar solvent and then dissolution in a polar solvent such as ethanol. An average value for the extinction coefficient was found to be 39,300 f 9 0 0 l.mo1-1 cm-1 at A = 531 nm. In toluene at A = 5 9 4 nm the value of E was found 48,200 f: 1000 l.mol-1 cm-1. A second method for obtaining the molar absorptivity of the colored species consists of liberating the pure colored form by reaction of the hydrochloride salt of the "open spiropyran" with an organic base (refs. 29, 62).
A third method for determining the extinction coefficient of an unstable colored form is to analyze the solids by IR spectroscopy. Successively Fischer (ref. 6 3 ) and Blanc and R o s s (ref. 6 4 ) have described procedures for finding the absorption spectra of both uncolored and colored isomers in the photorever-
332
sible system if only one of the isomers fluoresces. By the various techniques described above, then, it is passible to obtain a reasonably accurate absorption spectrum of a pure colored form in the visible region. Absorption of the colorless and colored forms present at equilibrium is shown i n Figure 3 . 5
-
iq q /fl.\,
,
\
i i
-
i
i
J \.
i
‘.
\
\
1
!
C
00
300
1
600 Wavelength (nrn)
400
I
\..
\.
600
I
0
Fig. 3 . Absorption spectra of colorless (solid curve) and colored (broken curve) solutions of 5,7-dichloro-6-nitro BIPS in ethanol at 20’C. From Bowman et al. (ref. 37) taken from R.C. Bertelson “Photochromism, G.H. Brown edit. Chap. 111. J. Wiley and Sons Inc. copyright owner. N.Y. 1971. Generally, the colorless form absorbs more strongly than the colored form at wavelengths shorter than about 250 nm ; the spectra of both forms: are similar from about 250 to 320 nm and from about 3 2 0 to 400 nm the colored form usually‘ absorbs more strongly. Spiropyran spectra change markedly in the presence of acid ; this results from the formation of the salt of the open
form. The absorption spectra of various 0-hydroxyarylvinyl pyrylium, quinolinium, pyridinium and benzothiazolium salts have been recorded by Schiele and coworkers (refs. 6 5 - 7 2 ) . The spectrum of a colored form is quite sensitive to other environmental influences (temperature, solvent, concentration).
333 2 . 1 . 3 . 3 Temperature effects on spectra
Hirshberg and Fischer (ref. 7 3 ) noted that the absorption spectra of several quinonoYd merocyanines in alcoholic solution changed markedly as the temperature varied. Generally the absorption band at a long wavelength diminished in intensity as the temperature decreased from 20 to -15O'C and a band at a shorter wavelength increased concomitantly. In an extended study performed by Knott (ref, 7 4 ) they showed that merocyanines could be classified into different groups according to their visible absorption spectra in ethanol. All these reversible temperature effects are attributable to a change in the position of the thermal equilibrium between various colored species. The colored species may include stereoisomers as well as dimers and higher aggregates. Low-temperature studies (to -196.C) have interested many investigators, particularly Israeli workers. Results taking into account the nature of the solvent (polarity and viscosity) are numerous and it is difficult to make a simple analysis which depends also on the structure of the spiropyrans. Investigations at low temperatures were coupled with flash photolysis and these techniques were first applied to spiropyrans in 1962 by Windsor and coworkers (ref, 49). Further work was carried out at the Weizmann Institute (ref. 29), and also by Russian workers (ref. 5 0 ) and French workers (Cie de St-Gobain) (refs. 3 9 , 41, 4 8 ) . These studies have led to some contradictory observations and interpretations of intermediates involved in the complex photochromic phenomenona (ref. 25). 2.1.3.4
Solvent effects on spectra The effects of solvents upon the absorption spectra of dye solutions have been extensively studied both in theory and exThe colored form of a spiropyran may be periment (refs. 7 5 - 7 8 ) . considered as a merocyanine dye of high intrinsic polarity and it behaves as one would expect. In general as the polarity of the solvent is increased, the visible absorption maximum shifts to shorter wavelengths, tne extinction coefficient decreases and the half-width of the band increases. In addition to the pure solvatochromic effect on hmax and Emax a change in solvent may also affect the positions of equilibria between the closed form and the various stereoisomers of
334
the open form, and thereby alter the observed spectrum. Flannery (ref. 79) states that both the Amax and the shape of the visible absorption band for 6-nitro BIPS, 6-nitro, 8-bromo BIPS and 6-nitro-8-methoxy BIPS change significantly with solvent. Unfortunately he did not give the Amax values for the 25 pure solvents in which he measured the thermal fading rates of these compounds. Fox (ref. 3 2 ) reported Amax values for 6-nitro BIPS in 23 solvents, these ranged from 523 nm (methanol) to 616 nm (methylcyclohexane). A number of correlations describing the effect of solvent on the electronic spectra of organic compounds have been proposed. The absorption maxima of the 5,7-dichloro-6-nitro BIPS expressed in wavenumbers, or as transition energies, give a reasonably linear graph when plotted against Brownstein's S values (ref.80), the correlation is less good when Kosower's Z (ref. 81),Brooher's X B (ref. 76) or Zhmyreva's S values (ref. 82) are used as the solvent polarity indicators. The UV spectra of the colorless forms of the spiropyran are generally independent of the solvent. 2.1.3.5 Concentration effects on spectra Colored forms of spiropyrans behave like merocyanine and cyanine dyes in showing deviations from Beer's law attributable
to dimerization and higher aggregate formation. The spectral, kinetic and thermodynamic behavior show changes with concentration parallel to those of dyes for which aggregation has been established. Flannery (ref. 79) examined the spectra in the visible region of colorless 6-nitro BIPS at very high concentration in benzene and ethanol. Benzene solution showed bands at 596 and 555 nm attributed to two different monomeric species, since irradiating solely in the longer wavelength band caused a relative decrease in intensity. At higher concentrations, the intensity of an absorption band at 4 9 0 nm increased, with concurrent decrease of the 596 and 555 nm bands. The concentration-dependent 4 9 0 nm band was attributed to a dimeric species, and a shoulder on the short wavelength side of this band suggested the presence of yet higher aggregates. Ethanolic solutions of 6-nitro BIPS gave a visible absorption spectrum that remained almost unchanged in shape as the concentration increased from 4 . l O - s M to I.lO-*M, although h a x shifted from 540 to 532 nm.
335 2.2 Ooen form soectra and thermal fadinn rates/structure correlations (solvent effects) The ready availability of compounds of the BIPS series with a wide variety of substituents in each half of the molecule per-
mits the deduction of various empirical or semi-empirical structure-behavior relations. In particular the “slow“ thermal fading kinetics would be expected to follow a Hammett type correlation. Berman, Fox and Thomson (ref. 83) examined the first order fading of 12 BIPS in ethanol at 6‘C. All the compounds were substituted only in the pyran portion, the 1’,3’,3’ trimethylindoline moiety being lcept constant. When Brown’s u+ values were used for the meta-substituents (i.e. in the 5 and 7 positions) and para (6)-substituents and Taft’s a * values were used for the ortho (8)-substituents, the measured rate constant k correlated fairly well with the equation (6) log k - log ko = ~ E o .
Scheme 8. Ring closure of photomerocyanine to spiropyran. The value of p w a s -3.45 and the intercept (i.e. log ko) had the standard error was 0.249 and the correlathe value - 1 . 6 4 9 , tion coefficient r = 0 , 9 7 9 ,
If the decoloration reaction is considered as an attack of
a
positively charged portion by a phenolate anion (Scheme 8 ) , one would expect to use a- values for the correlation. Conversely, if the substituents on the indoline portion are varied, keeping the pyran substitution constant one might expect to use u + values for the correlation, considering the reaction as an attack on a posi-
tively charged center. Figure 4 shows the Hammett-type plots for the substituted BIPS derived from three different salicylaldehydes, where ui represents the inductive parameter.
336
,
8
.
44.6
1.2
1.0 '
-0.3-02-0.1 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 Fig. 4. Hammett-type plots f o r the thermal fade rates at 2O'C of substituted BIPS derived from three different salicylaldehydes : curve 1. 5-bromo-8-methoxy-6-nitro-substituted BIPS in ethanol, -log k = -2.25 Coi t 4.11 ( r = 0.968); curve 2.6-nitro-substituted BIPS in ethanol, -log k = -1.85 Coi t 3.61 (r= 0.937); curve & 8methoxy-6-nitro-substituted BIPS in ethanol, -log k = -2.00 Coi + 2.49 r= 0,980) curve 5-bromo-8-methoxy-6-nitro-substituted BIPS n toluene -log k = -2.02 E o i + 2.08 ( r = 0.966) (ref. 25) taken from R.C. Bertelson "Photochromism ", G.H. Brown edit. Chap. 111. J. W ley and Sons Inc. copyright owner. N.Y. 1971. In
the
indoline series the thermal fading rates in toluene
are not very different from one substituent to another (ref. 841, the contrast is better in the benzothiazoline series for instance (refs. 28, 59). The electron-donor groups stabilize the open form by decreasing the thermal fading rate.
331
Nitrogen-containing heterocycles 2 . 2 . 1 . 1 Nitrogen-containing heterocycles with one nitrogen atom The most stable transold or cisold colored species are those with a NO2 in position 6 and a OCHJ in position 8 (Scheme 9 ) . 2.2.1
I13
n Scheme 9. systems.
\ I
OkH,
Photochromic
. u
equilibrium
for
nitrogen-containing
These colored photomerocyanines have an important polar character in the ground state (negative solvatochromism) and are stabilized by polar solvents (protic or aprotic), i.e. the thermal fading rates strongly decrease. The electronic distribution look5 like a dipolar or zwitterionic one, the dipolar mesomeric form having an important contribution in the ground state. The open colored form is strongly conjugated and there is an electronic interaction between the two parts of the molecule. The substituents have also an important role in the thermal fading rates and on the absorptionhmax of the colored form. They have a direct action through the electronic effects (inductive through the a-bonds, mesomeric through the rr-bonds or non-bonding heteroatomic electrons) and also through their steric hindrance depending on their position on the skeleton of the pseudo-planar photomerocyanine obtained by UV irradiation of spiropyran. The linear or ramified alkyl substituents on heterocyclic nitrogen atom or in 3 position on the dimethine bridge have a direct influence on the thermal fading rates because of nonbonding interactions in the transoPd pseudo-planar photomerocyanine (scheme 10). This steric effect depends upon the planarity and rigidity of the system, it is more important in the case of benzo-fused systems.
338
Scheme 10. Main non-bonding interactions in the benzothiazoline photomerocyanine skeleton. The steric hindrance of substituents in 3-position and in 5position plays also an important role on the visible absorption spectrum of the open form and on the preferential geometry of the transold or cisoPd open form (ref. 85). It will be interesting to compare some values for the thermal fading rate and absorptionhmax of colored species in a nonpolar solvent such as toluene by choosing appropriate substituents showing directly electronic or steric influence. Table 2 gives the spectrokinetic parameters for some characteristic compounds through the nitrogen-containing systems. TABLE 2 .
Spectrokinetic values for some characteristic substituted compounds through nitrogen-containing systems.
F
kA(s-1) 25'C (OF-CF) toluene
Xmax O.F. (nm
refs.
OCH,
(*)Indoline ( A )
242
610 625
27,86 27,86
339
TABLE 2 continued
600
2 8 ,5 6
23.2
635
28,56,87
680
610
28,56,87
CHI
2.6
Cr H5 iC3
I
HI I
612
610
56 87
cs H5
6.4
635
28,56,87
OCH3 OCs H5
9.5.10-
640
5 6 ,8 7
625
56 I 87
c6
CH3
H7
16.2.10-3
3
Benzothiazoline(B)
Benzoxazoline(C)
0.82
600
88,89
6.10-2
620
88 I 89
3.30
600
88,89
S
H=
CH3JNX CH36H3
CH3
17
610
88,89
OCH~
1.8.10-2
600
8 8 ,8 9
Thiazolidine ( D )
Oxazolidine ( E )
CH3 Cz Hs Pyrrolidine ( F )
15
410,580
90
4 1 0 585
90
340
TABLE 2 continued
H = CH3
CH3 CH3&H3
1,3-Thiazine ( C )
CH3&H3
1,3-Oxazine ( H )
CsHs
OCH3
138 13.i
420 ,600 420 , 590
88 88
1.3.10-2 2.3.10-3
420 , 530 412 , 538
88 88
590
90
420,575 440
90 90
ZcH3
s
CH3
H = CH37( N
CH3 k H 3
CHI
50
H
0.40 0.13
1,4-Thiazine (I)
CH3 (*)
CH3
PiDeridine ( J )
( X ) In these series, steric hindrance is sufficient for obtaining a spiropyran structure directly by synthesis. If R 3 = H a permanent merocyanine form (or a bicondensed adduct) is formed as a result of the preparation.
The basicity of the ,eterocyclic nucleus (as described by Brooker) is also important. The benzimidazolium salts lead to the formation of merocyanines whatever the substituents (ref. 91). For the same series, for instance the benzothiazoline one, when the pair of substituents in 6 and 8 positions are inverted i.e. 8-NO2 6-OCH3 the photomerocyanine is considerably more unstable and the absorption hmax of the colored form undergoes a bathochromic shift (Table 3) (ref. 92).
-
341
TABLE 3 . Spectrokinetic values zopyran compounds.
kA (25'C) X max
= =
2.6 S - 1 610 nm
in toluene for 2 benzothiazolinospiroben-
k A (25'C) max
= 3035 S-' = 645 nm
2.2.1.2 Benzofused spiropyrans with quasi-planar photomerocyanines (A,B,C) Correlations have been made using electronic and steric effects : they will be illustrated on the benzothiazoline series which is particularly suited for these investigations (ref. 56).
Three types of substituents in the 3-position have been investigated : alkyl or alicyclic, para-substituted aryl and functional groups. The polar effect of alkyl substituent, is considered first by using Taft ol: parameters ; the compounds are sensitive to this
: the fading rate varies by a factor of 10 between CH3 (ol: = 0 . 0 0 ) and CZHS (ol: = -0.10) substituents ; with an isopropyl group (08 = - 0 . 1 9 ) the rate increases by a factor 30. This important variation is due to the intervention of the steric effect (Es = - 0 . 4 7 for i-C3H,) whereas Es values are comparable for CH3 and CZH5 (tO.00 and - U . O 7 respectively). This steric effect appears still more clearly with the cyclohexyl substituent (Es = - 0 . 7 9 ) which has a rate similar to that of isopropyl but iDs ul: coefficient (-0.13) lies between ul: of iPr and Ethyl substituents. These results show that steric effects are crucial in the rates of ring closure of the photomerocyanine form ; the rate increases with the bulk of groups in position 3. This phenomenon was confirmed by a theoretical evaluation of non-bonding interactions energies (ref. 93). To determine the electronic contribution to the whole substituent effect it is necessary to keep constant the large steric interaction. For this investigation, para substituted aryl groups CsHd with 2 = H, CH3 , OCH3 , OH, C1, Br and F for which steric effects
342
effects may be considered as largely invariant have been selected (Scheme 1 1 ) .
P Scheme 11. Representation of the delocalized electronic system. It appears that the stability of the photomerocyanine is governed by inductive a n d . resonance effects of the parasubstituents which are transmitted to the carbon atom in the aposition of the aryl group.
linear relationship (correlation coefficient r = 0.973) has been obtained using the Hammett CJ para-constants (Fig. 5 ) . A
1
I
- 0.3
- 0.1 I
I
0.1
6
P
3-
Fig. 5. Hammett plot for the thermal fading rate ( k 4 of J’-methyl
benzothiazolino-spiro-3-para-substituted-phenyl-6-nitro-8-methoxy benzopyran compounds in toluene at 2 5 ’ C ) .
343 The slope ( p = 1.21) of this relation indicates that a positive charge density is located on atom 3 , this has to be compared to a u p + relationship for 6'-substitution (refs. 28, 59). However, the steric hindrance leads to the rotation of the $-substituted para-aryl group out of the plane defined by atoms C2, C3 and C 4 . The degree of conjugation of the 3-aryl substituents can be estimated by the Yukawa-Tsuno relation (ref.
94) : 1og k/ko = p(a+RAa) (7) wherep is the reaction parameter, a is the inductive constant of the substituent, U + - U is the resonance contribution of the substituent effect as defined by Brown and Okamoto (ref. 9 5 ) and R is the degree of resonance interaction between the substituent and the reactive center. The calculation of this relation with a multiple regression program gives logk/ko = 1.50 ( a + 0.27 A a ) with a correlation coefficient r = 0.997. The value obtained for R(0.27) confirms the weak degree of the resonance interaction of the para substituent, which is to be related to the rotation of aryl groups with respect to the plane of the delocalized system. The colored form of methoxy and phenoxy derivatives is largely stabilized when compared to their methylthio, phenylthio or methyl homologs. Further to the electronic and steric effects, an additional factor of stabilization may be proposed for oxygen compared to sulfur : an intramolecular chelation by H-bonding.
When the substituent in position 3 is hydrogen, a permanently stable merocyanine is formed : this infinite stability must be related to the absence of steric effect and hence to an optimal conjugation (ref. 93).
344
A multiparameter correlation involving inductive, resonance and steric hindrance aspects was proposed by Charton (refs. 9698) : log k A = aui t Bur t $Jr'v t h (8) ui and ur are inductive and resonance parameters such asOp = ui t O r .
up are those chosen by Mc Daniel and Brown (ref. 99) ;
come from the compilation of Charton (ref. 100) ; rv is a steric parameter defined as an average of van der Waals radii of atoms or groups of atoms (refs. 96, 101, 1021, the reference being the hydrogen atom (rev = 1.20 A ) For a substituent R 3 , r O ~ n 3= r n 3 ~- rpv a, a , $ and h are constants for the reaction considered. For twelve values of R 3 = CH3 , CzH5, iC3Hr , CsHi i I OCHJ , OCzHs I 0nC3 H.r , 0-nC4HS , OCSH5 , SCHI, SC6H5 , CSHS , the triparameter correlation was found satisfactory : r = 0.987, a = 10 ?: 3 ; I3 = 12 f 3 ; 3/ 22 t 3 and h = -8.5. The participations of inductive and resonance factors are quite identical and the most important control of the ring closure is of steric origin. The non-bonding interactions concerning the position 5 have also an influence on the thermal bleaching rate and the absorption of the photomerocyanine, as observed by Smets and Vandewijer in the benzothiazoline series (ref. 103) (Table 4).
UI
TABLE 4 . Thermal bleaching constants and series.
iC3 H7
CH3
1180
absorption
in
496
benzothiazoline
345 '2.2.1.3
m o p y r a n s with non-totally planar photornerocyanines
( D i E i Fi G, Hi I , J ) 2.2.1.3.1 Five-membered heterocycles (thiazolidine, oxazolidine,
pyrrolidine series) The substitution 6-NO2,8-OCH3 is an ,interesting pattern inducing good photochromic properties,like for benzofused systems. by its steric and The substituent R 3 influences kA electronic effects, the former of which is prominent. However the steric effect is less important than for benzofused systems because the cycle is flexible and not s o highly sensitive to the
non-bonding interactions. When R 3 = alkoxy or aryloxy, the open forms are considerably stabilized. Their coloring ability is also very good. Generally, accumulation of methyl groups on the N-containing heterocycle greatly favors an increase in the life-time of the photomerocyanines as does the use of a polar solvent (ethanol). Moreover, a relatively good correlation is obtained using a Hammett relationship of the type log kA = P C O i (CH3) + 6 , especially for the oxazolidine series, log ka = 2.31 Coi ( C H 3 1 - 0 . 7 8 4 ( r = 0.979)
*
The results f o r the thiazolidine and oxazolidine series confirm and complement those already nbserved for benzofused Ncontaining series. The oxazolidine series is particularly interesting with respect to the benzoxazoline one in so far as it makes possible an extension of the visible absorption of photomerocyanines. The pyrrolidine photomerocyanines in toluene show 2 absorptions in the visible region near 4 2 0 nm (weak) and near 580 nm (strong) ; their thermal stability is quite similar to that of the thiazolidines. The importance of the conjugation brought by the benzene nucleus for the stabilization may be evaluated by comparing the thermal fading rate in toluene at 25'C with the two spirochromenes shown below (Table 5 ) .
346
TABLE 5. Compared kinetic bleaching values in toluene at 25’C.
4’,5’-tetramethylenethiazolidine
benzothiazoline k A = 2.6 s - 1
k& = 9 8 . 8
S-’
The solvent effect (from non polar toluene to polar and protic ethanol) on absorption spectra and thermal fading rate is more important than in benzofused systems (Table 6). The absorption of the C.F. is nearly identical in toluene and ethanol. TABLE 6 . Influence of structural features and solvent on the fading-rate of the Open Form (O.F.) and the Xmax of both Closed Form (C.F.) and O.F.
356
450-460
610
toluene ethanol
356
590
toluene ethanol
355.3
14.5
356
465
347 TABLE 6 continued
5.35.10-* 1.6.10-6
352.5 352
545 450
toluene ethano 1
540 442
toluene ethanol
2.2.1.3.2 Saturated six-membered heterocycles (l13-thiazine, l14-thiazine, 1,3-oxazine, piperidine series) The photomerocyanines obtained by UV-irradiation are generally characterized by a visible absorption spectrum consisting in 2 bands, one near 420 nm and the other near 470-530 nm for the oxygen-containing series or 580-600 nm for sulphurcontaining molecules by contrast with the benzofused or fivemembered ring saturated systems for which the band at 420 nm was a shoulder of weak intensity. The piperidine series synthesized by Petillon (ref. 9 0 1 , shows interesting features : the visible spectra have a single maximum wa.velength, very sensitive to the steric hindrance of the group in position 3 : R3 = H , Amax (O.F.) = 580 nm ; R3 = C H J , hmax (O.F.) = 440 nm. In the 1,4-thiazine series all the spirochromenes have a hydrogen atom in 3 position and only one absorption near 600 nm (in toluene) is observed ; the thermal fading rate is generally rapid and depends on the substitution in the benzopyran part (ref. 9 0 ) . For the 1,3-thiazine and l13-oxazine series the thermal fading kinetics (OF -P CF) was studied at the two wavelengths of the visible absorption. It was found 1 ' 'order whatever thehmax of analysis.
Generally one observes : - A decrease in the stability of the photomerocyanine when position 3 is sterically hindered. - The inductive effect of methyl groups, as in five-member ring systems, stabilizes the open form. - A better conjugation of the phenyl group in 3 position especially in the 1,3-thiazine series compared to the five-member ring saturated series. In an attempt to explain this spectacular spectral behavior of six-member ring saturated systems and particularly the piperidine one, various experiments concerning the temperature effect (between 20'C to 55'C), the solvent effect (apolar to polar solvents) and also the addition of acid or base for examining the influence of ammonium salt, formation or elimination were carried out (ref. 104). Finally several hypotheses formulated concerning the absorption band near 420 nm have been discarded one after the other : - Formation of a "salt" (refs. 4, 25, 38, 44, 62) - Triplet-triplet absorption (refs. 29, 105, 106) - Dimerisation or aggregation of photomerocyanines (refs. 79, 107) improbable under these experimental conditions (c = 5.10-sM).
- Charge transfer complexes between spirochromene and photomerocyanine (refs. 108-113) - Formation of degradation compounds from apirochromenes or photomerocyanines. The interconversion between twisted non polar cisofd "X" photomerocyanine and planar polar transord photomerocyanine was clearly shown. The relative stability of each depends on the geometry of the six-membered ring system and the piperidine series is a good illustration of this case with the role of the substituent in position 3. The demonstration of the existence of a twisted cisofd form stable at room temperature was shown for the first time. Moreover this study was supplemented by an investigation on some permanently stable merocyanines models (ref. 114). 2.2.2 Non Nitrogen-containing heterocycles (sulphur or oxygen) The pair of substituents leading to the most stable colored species is that having OCHJ in 6 position and & in 8 position.
349
An inversion of the effect is observed, electronic distribution whose representative quinonic or polyenic.
due to a non polar weight is mainly
OCH3 A marked effect concerning kinetics and absorption is observed depending on the type of substitution in 6 and 8 positions and also 3 position (see Tables 7 and 8). TABLE 7 Benzodithiole and benzoxathiole series. Effect of the on the fading-rate and the hmax of the Open Form.
a3
R3
Series
benzodithiolane
kA s - 1 Xmax (25'13) (nm) R S Z C H ~ 194
640
R3=CH3
39.6
635
R3=CsHs 19.5
645
R3=CsH5
0.66
610
91
640
.-
0.37
620
12
640
R3
=Cs Hs
R=CH, H R & o
benzoxathiolane
structure
RS=CsHs R=COCH,
t
id-
- id -
7.8.10- 2
620
92
115 92
115
92
116
350
TABLE 8.
Bis-spirobenzopyran series. Effect of the fading-rate and the hmax of the Open Form.
H H H CH3
5,6
benzo 5,6
benzo 5,6
benzo H
structure
on
the
H
H
8
522
t
H
H
44
560
I
I I
H
CH3
94
560
H
CH3
315
620
92
117 I I
CHa
CH3
4.5
620
I
I I
CHI
H
H
CH3
cs H5
Cs Hs
30
580
CHI
2
580
I I
I I
1
H
H
CS HS
58
620
I I
I
cs H5 c6 H5
CH3
H
H
c s H5
0.51
620
C6 HS
1
600
C s H5
not photochromic weak absorption
I I I
I I I
1
351
The preceeding Tables deserve some comments : the substituents in 3 position are known to have mainly two kinds of effect : polar and steric. In a previous work (ref. 86) on the benzothiazoline series, it was demonstrated that the steric hindrance of R1 and RJ' groups has a direct influence on the thermal stability of the photomerocyanine but no direct relation 'exists between logk4and Taft constants 0 % (Fig. 6). On the contrary, for the beneodithiole series, the absence of the R3' group decreases the non-bonding interactions and, in the case of 8-NO26-OCH3 substitution type, a good correlation between logk,,and O X is observed (Fig. 6). log k s s = -3.Ou* t 0.53 (r = 0.970) (9) (kss
=
Thermal fading rate for saturated substituents in 3poe ition )
Fig, 6 . log kss= f(o*) ; ( 0 ) Beneodithiole series (6-OCH3,8-N021 ( 0 ) Beneothiazoline series (6-NOz,8-OCH3 ) However fox the 6-NOz 8-OCH3 substitution type, the relation log kA= f ( u s ) is not linear ( r = 0.830) and shows a little sensitivity to the polar effect ( p * = -1.6). A phenyl group in 3 position plays an important role in the stabilization of the photomerocyanine in comparison with the benzothiazoline series. Substituents in p-position on the phenyl group have very little influence on the thermal bleaching rate, indicating that the electronic charge d.eveloped at carbon atom 3 is not significant. Concerning the UV-VISIBLE absorption spectra (Table 9 ) 1 an impor-
352
tant bathochromic shift is observed for the open f o r m s of spiropyrans (14 to 2 7 nm) from the "6-NOz , 8-OCH3" to the "8-NO2, 6-OCH3 '' systems.
TABLE 9.
hmax for Closed Form (C.F.) and Open Form (O.F.) o f heterocyclic structure (in toluene).
Heteroatoms
of
spiropyrans
x=s
CF 3 3 3
360
t
Y=NCHj
OF 6 1 0
645
t 35
x=o
CF 3 4 6
360
t 14
Y=NCHs
OF 6 1 0
670
t 60
x=s
CF 3 4 1
366
t 25
OF 6 4 0
635
-
Y=S
27
5
F o r the open forms the bathochromic shift is stronger (AX=35 to 60 nm) in the nitrogen-containing series but for benzodithiole spiropyrans the shift is hypsochromic ( P A = -5nm). For the benzodithiole series the substituents in position 5 ' have also a different influence on the thermal bleaching kinetics. In contrast to the benzothiamline series, electron withdrawing groups have a positive effect on the stabilization of the photomerocyanine.
353
The same is true for the benzoxathiole series (ref. 116). All these resu1t.s obtained by effect of solvents on thermal bleaching rates and absorption spectra (refs. 115, 116) (Tables 10 and 1 1 ) are in agreement with a different electronic distribution mostly of quinonic or polyenic type. TABLE 10. Effect of solvent on the thermal fading-rate ka Xmax of the Open Form (O.F.) (nm)
(s-1
at 25'C) and
ET and XR are empirical parameters of solvent polarity determined for reference substrates, such as a rnethylpyridinium iodide (ref, 81) and a merocyanine-like dye respectively (ref. 76).
XR = ET =
50 31.2
Xn = 48.7 ET = 3 2 . 5
XR = ET =
4%
kb 1.38
577
47.2 33.9
16.5
600
93
640
0.41
595
2.56
620
354
TABLE 1 1 . Effect of solvent polarity on the thermal fading-rate k A and hmax of the Open Form (O.F.) of a benzoxathiole derivative.
~
~
~~
~~
solvent
cyclohexane
CC14
toluene
kA
3 . 6 10-5
3 . 1 5 10-2
3 . 7 10-1
S-I
(25'C)
Xmax (o.F.) (nm)
530
552
620
All the results concerning the influence of substituents in positions 3 , 6 , 8 , on the thermal decoloration kinetics and absorption spectra are found also with spirobibenzopyrans. - In the case of disymmetric spirobibenzopyrans the ring opening arises probably in the benzopyran part not substituted by a methyl group. For an identical substitution in positions 61 8 , 6 ' and 8 ' a phenyl group in 3 position stabilizes the photomerocyanine. This result is in keeping with those observed in the benzodithioles (ref. 1151, benzoxathioles(ref. 116) and also dithioles (ref. 118) (Table 8 ) . Indeed the photomerocyanine obtained by U V irradiation of a spirobibenzopyran is more stabilized for a "8-NO2 6-OCHa" than for a "6-NO2 8-OCHs" substitution (Table 8 ) .
-
-
SO the electronic representation of the photomerocyanine of spirobibenzopyran oan be given as an important contribution of quinonic structure. All these assumptions are verified by structural studies on appropriate heterocyclic merocyanine models by determination of dipole moment in the ground state (ref. 119) and evaluation of dipole moment in the excited state in agreement with the solvatochromy of these compounds (ref. 120).
355
2.3 Coloration efficiency or "colorability", quantum yield A proper evaluation of coloration efficiencies would require knowledge of the quantum yield for dye formation for every activating wavelength used. For practical applications, one often simply wishes to find the material that gives the maximum absorbance under fixed experimental conditions. Such comparative evaluations provide a rapid determination of relative coloration efficiency Usually, the results from one laboratory are not directly comparable with those from anothqr, as the experimental conditions are never quite identical. The measurement of maximum absorbance depends on the thermal fading of the material ; a fast fading material will give lower absorbance than a more slowly fading one under the same experimental conditions ; the fluidity or viscosity and the temperature of the material are also important. With dilute fluid solutions at room temperature, experience shows that the 6-NOz18-OCH3 BIPS compounds are generally notably more sensitive than all other types of spiropyrans for the same range of concentration. Some substituents in the benzene nucleus may also influence the coloration intensity. There are very few data about the quantum yields for coloration or decoloration determined with monochromatic light because of the difficulties to analyze a photochromic system in which a reversible reaction proceeds in both directions thermally and photochemically at the irradiating wavelength, and in which reactant and product both strongly absorb the irradiating light. However different methods were described by Bertelson (refs. 2 5 , 3 1 1 , National Cash Register Company scientists (refs. 3 7 , 58, 1 2 1 ) ) Bach and Calvert (ref. 1 2 2 ) , and, some Russian authors (refs. 4 2 , 54, 1 2 3 ) have reported many values of quantum yields determined by different spectroscopic routes. The coloration Quantum yields in non-polar solvents are generally about 0 . 5 to 0.7 mole,einstein-1 for compounds such as 5'-chloro-6-nitro-BIPS and 5'-methoxy-5-bromo-6-nitro-8-methoxyBIPS. The quantum yields in ethanol are somewhat lower. Lashkov and Shablya (ref. 53) examined the direct photocoloration of 6nitro BIPS in several solvents and found that the quantum yield of photocoloration decreased as the solvent polarity increased. Photobleachinc quantum yields are rather small and solvent dependent and there are few available data.
.
356
Samat et al.
(ref. 124) have estimated the quantum yield of
a 3 , 8-dimethoxy-6-nitro-benzothiazolinospiropyranat hmax = 535 nm in EtOH, recording spectra of apiropyran and photoprecipitated photomerocyanine in EtOH and determining the isosbestic point (312.5 nm) : B c o i t 0.15 +- 0.03 ; this result is in agreement with coloration quantum yields found for indoline
spiropyrans (ref. 2 5 ) . The notion of "colorability" of spiropyrans was introduced to give a standard comparison between different series or between different compounds of the same series. .€OF (refs. This notion was defined by the expression 92, 221). A quantitative approach to the reaction is to measure directly the initial absorbance of the spiropyran solution immediately after the photolysis :
SCOI
AO
= (dDoF/dt)t=o
1
= cell length
(10) Using the analysis of Arnaud (ref. 1251, if dncr and d n o ~ represent the number of uncolored molecules (spiropyran) and colored molecules (photomerocyanine), the kinetics of coloration can be expressed : -dnc~/dt = dno~/dt The quantum yield at the irradiating wave ength is given dno F = ---/Iabs dt Iabs = 10 - I 1 = I~.~O-ECFQI.CCF 10
=
incident light intensity in ph0tons.s-1
I = transmitted light intensity
dnoF/dt = gc0l.Io.(l-lo-EcF.l.Gc~) The Beer-Lambert law is valid in the range of 10-5 to 1 0 - 3 M. D O F = E O F . ~ . C O F ; C O F = no~./N.v N = Avogadro number : v = volume of sample cell
DOF =
Eo~.l.no~/N.v dDor/dt = $ c o i . € o F . I o .(l-lO-EcF-l,CcF)/N~ ~
(13) concentration
357
When E C F . ~ . C C is F very small (weak concentration) the following expression can be derived : AO = (dDop/dt)tZo = $ c O l . E ~ F . I o . ~ . ~ ~ F ~ . C C F . ~ . ~ / N ~ ( 1 6 ) [lo-"
If I< AO =
2
1-2.3U ; ( 1 - 1 0 - " )
= 2.3.10.1*.€c~/Nv
The absorbance Cgcoi
OF.^) C C F
%
2.3U ;
U
E
ECF.1.CCFI
(17)
Plotting A o t o ~ , = f ( C C F )gives theoretically a straight line with a slope = gc0i. E O F . ~ (Fig. 7 ) . For more theoretical approaches refer to Gauglitz chapters 2 and 25.
Fig. 7 . Variation of absorbance in function of concentration and type of heterocyclic series, for the same experimental conditions. A maximum for Ao is reached at a different concentration for each series of spiropyrans. This maximum is due to excited states annihilation and aggregation phenomena as the concentration increases. For interesting comparisons between the series of spiropyrans having a similar substitution it is convenient to choose the linear part of the curves and generally at low concentration. It is possible to check the relative "coloration" for different compounds under the same experimental conditions of
358
photolysis. The practical interest of this determination is evident. For example, the experimental coloration Ao evaluated for different spiropyran structures in toluene at 2 5 ’ C at a concentration of 5 . 1 0 - 5 M , for a photolysis energy of 800 Joules and in a 3 c m cell is given in Table 12. TABLE 12. (ref. 92)
Compared coloration ( A o ) of different spiropyrans under the following standard conditions. Solvent : toluene (20 ppm HzO) ; temperature : 25‘c ; concentration : 5 . 1 0 - 5 M .
~
Ienzothitrol Ine,
cn,
“02
‘6%
NO2
@%
“%
0.25
e.12
359
Both the photocoloration and photobleaching can be sensitized. Bach and Calvert (ref. 1 2 2 ) examined the sensitized (and quenched) photocoloration of 5,7-dichloro-6-nitro BIPS in acetonitrile at 2 4 ' C under 3 6 6 nm irradiation and proposed a mathematical expression. These authors found that benzophenone and 2-acetonaphthone sensitize the formation of the colored form. As it is generally the case the photocoloration of this spiropyran proceeds through both the singlet and the triplet states,
360 3. STRUCTURAL AND THEORETICAL STUDIES ; COORDINATION CHEMISTRY 3.1 Characterization. reactivity and conformational studies by lH and 13C NMR spectrometry The NMR spectrometry is an interesting method for structure investigation of spiropyrans. It allows rapid and clear verification of the structure of a compound (ref. 123) after synthesis, by characteristic chemical shifts, coupling constants and number of different kinds of protons. In the spiropyran series prepared from 3-methoxy-5-nitro-salicyl aldehyde and benzothiazolium quaternary salts, the chemical shifts of the nitrogen-methyl group may be correlated with regard to the 6'-substituent by a Hammett equation : 6 = 8.34 Co t 181.98 ( r = 0.871) (18) but this correlation is not s o good (ref. 127). The rotation around the N-C bond is not unhindered for isopropyl or ethyl groups in contrast to such groups in 3 position of the skeleton and the signals are doubled. At 1OO'C we observe a normal NMR spectrum. The first results concerning a structural study of the equilibrium between spiropyran and photomerocyanine were published by Flannery (ref. 79) on a 6-N0a BIPS. The chemical shift of gem methyl groups in 3' position appeared surprising for the photomerocyanine and we verified (ref. 128), in agreement with Dzhaparidze (ref. 129), that chemical shifts for the photomerocyanine are larger taking the increase of polarity into account (Table 13). Benzothiazoline spiropyrans derived from 3-methoxy-5-nitrosalicylaldehyde and having an oxygen group in 3 position ( r e f s . 56, 87) are characterized by the formation of a highly stable photomerocyanine. These spiropyrans are available to give photoprecipitated photomerocyanines and 'H NMR spectra have been registered in DMSOd6 (Figs. 8a and 8b).
361
TABLE 13. 1H N.M.R. chemical shifts of 6-NOzBIPS and its photoprecipitated in toluene or hexane.
photomerocyanine
~~~
Compound
6(10)-6
our results
(Ref.79) (Ref.129) and(Ref.128)
.................................................................
a.-$ /‘@3
CH3
6 P*)
7
CDa OD 1.15(3H) 1.25(3H)
1.22366H)
solvents CD3 OD
CDCl3
1*18(3H) 1.2863H)
1.21(3H) 1.29(3H)
1.83(611)
1.80(6H)
3
1
0 Fig. spectrum of the 3’-methyl benzothiazolinospiro-3~~-dimethoxy-6-nitro-benzopyran ( C . F . ) in DMSOda 6 5 4 8a. Fourier transform ‘€I N.M.R.
2
1
.
They are significant and the results are quite interesting (ref. 130) and show quantitatively the polarity of open forms (photomerocyamines or permanent merocyanines). The chemical shift of the N-methyl group is comparable to that of heterocycloimmonium salts.
362
'pm) 11
7
* 6
5
4
I
3
I
2
1
I
C
Fig. 8b. Fourier transform 1H N.M.R. spectrum of the photomerocyanine (O.F.) corresponding to the 3'-methyl-benzothiazolinospiro-3,8-dimethoxy-6-nitro-benzopyran in DMSOd6. ,Their trans-configuration, the alternation of the electronic density on the dimethine bridge between the "benzo-thiazoline" and "phenolate" parts of these molecules are observed. A favored configuration is possible when photomerocyanines have an alkoxy or aryloxy group on the 3 position. 1 S C NMR spectra have been registered for the first time (Fig. 9a). Without any problem but for reasons of solubility, 1 3 C spectra have been registered on permanently stable merocyanines having a C I O chain on the nitrogen atom (Fig. 9b). A 1 s C NMR study of different heterocyclic spiropyrans has been carried out (..ref. 131). We have attempted to describe the specific influence of the heteroatoms on the chemical shift of the neighboring atoms.
363
I
120
.
I
90
,
I
60
I
30
I
0
Fig. 9a. 13C NMR spectrum of the 3’-methyl-benzothiazolinospiro 3-methoxy-benzopyran in CDCla.
6 Fig. 9b. 1 J C NMR spectrum of the Cl6H33-nitrogen merocyanine cor-
responding to a 8-methoxy-6-nitro-benzothiazolino-spiropyran, CDCla ,
in
364 Carbon13 chemical shifts and coupling constants have been also obtained for compounds similar to both parts of the spiropyran molecule : the heterocycle and the benzopyran. An important series of permanent a-0x0 benzimidazoline merocyanines differently substituted on the “phenolate“ part and bearing a paraffin chain on one heterocyclic nitrogen has been specifically studied by 1 3 1 2 NMR technique (ref. 132). Electron-donating substituents bring about an increase both of the electronic density on the 3-methine carbon and the alternation of electron density on the 3,4 dimethine bridge. Furthermore, the low field resonance of the 3 methine proton is probably due to a specific interaction between this proton and the “phenolate“ oxygen atom. This implies the existence of a favored trans-configuration for merocyanine dyes. The synthesis of oxa- and 1,3-thiazolidine spirochromenes with an asymmetric carbon atom in an a position to oxygen or sulfur leads generally to a mixture of two diastereoisomers for which the 2 configuration is predominant (Fig. 10).
Fig. 10.
Z
E
The diastereoisomers have been evaluated by 2 5 0 MHz 1H NMR spectra. Examination of the Dreiding models and application of the Karplus equation allow a good solution of this problem. On the other hand only one isomer exists in the 2,4,5,6-tetrahydro1,3-oxa or thiazine homologs compounds (ref. 133). The study of the geometry of the aza ring proves then that the configuration may be Z or E according to whether one of the asymmetric carbon atoms is in an a of R position to oxygen or sulfur. Finally heterocycloimmonium salts generally react with 3methoxy-5-nitro-salicylaldehyde to give spirochromenes or merocyanines. Only the latter compounds are obtained in the thiazoline and the benzimidazoline series.
365 The linking of a paraffin ring to the merocyanines facilitates their 13C NMR study. The comparison of their spectra with those of heterocycloimmonium salts reveals a hyperconjugative effect in the thiazoline and the benzimidaeoline series. The benzoxazine series appears to be in an intermediate position : the thermodynamic stabilities of spirochromenes and merocyanines are quite similar. The limiting value of the chemical shift of the carbon atom involved in the spiroannellation of merocyanines can be approximately determined (ref. 134). 3.2 X-ray diffraction structures of spiropyrans and merocsanines 3.2.1 X-ray molecular structure of spirovyrans The X-ray diffraction data (refs. 135-1421 show that, for almost all the spiropyrans investigated, the angle between the 2
heterocyclic parts is 9 0 ' . However, the heterocycles themselves are non-planar : the angle of the bend in the planes of the indoline ring systems along the C(3I)-N(l8) line is 23-30' and the angle in the bend in the benzopyran ring system along the C(3)O(1) line is 11-17' (ref. 136) depending on the ring substituents. To obtain a better insight into the correlation between molecular substitution, conformation and photochemical reactivity, the molecular structure of a photochromic 3-ethyl-8-methoxy6-nitro [2H-11 benzopyran-2-spiro-2' (3'-methyl thiazolidine) (I) (ref. 89) has been established (ref. 142a) and compared to the 8-bromo-6-nitro-indolinospirobenzopyran (11) conformation of (ref. 140) and of the 8-nitro derivative (111) (ref. 141).
k0,
The spiro carbon is involved in two long bonds to S [ C Z Z ~ - S =I ~1.852 A and C Z Z , - 01 = 1.469 A ]
and
0.
366
The thiazolidine ring is in an almost perfect half-chair conformation. The conformation of the pyran ring of thiasolidine and indoline-spiropyrans is intermeaiate between "sofa" and "boat" conformations and could best be defined as 1,3 diplanar (ref. 142b) or screw-boat (ref. 142c). Its puckering and torsion energy increase through introduction of such bulky substituenta in positions 3 and 8 of the benzopyran that cannot avoid short intramolecular contacts through steric accommodations. Molecules are piled along the b axis, alternately reversed in orientation and nearly equally spaced. Both enantiomers are present in the crystalline state as implied by the centro-symmetric space group (ref. 141). A perspective view with atomic numbering, bond lengths ( A ) and bond angles ( ' 1 is given on the following figures (Figs. lla,
(4
(a) Perspective view and atomic numbering of the compound (I). (b) Bond lengths ( A ) . (c) Bond angles ( * ) ,
Fig.
11.
367
The mo.lecular packing of the thiazolidino-spiro-benzopyran ( I ) viewed along a (distances in A ) is given on Fig. 12.
Fig. 12. Molecule (Distances in A ) .
packing
of
the compound (I) viewed along a
3.2.2 Molecular structure of a permanently stable merocvanine (IV) in the benzothiazoline series. A model compound for a photomerocyanine In the solid state, the polar zwitterionic form of the merocyanine (IV) is dominant according to results from solvatochromy (ref. 119) and NMR spectra (ref. 130).
I cans
t
I
The molecular geometry of the thiazole ring approaches that of a thiazolium cation. The phenolate ring is the negatively charged moiety of the molecule (ref. 143). In the crystal lattice, the molecules associate in pairs
368 across centers of inversion. The slip angle is 55'. The Ca-CI bond has also an important double bond character. Bond lengths ( A ) and bond angles ( * ) are given on Figs. 13a, b).
Fig. 13. X-ray parameters of compound (IV). ( a ) Bond lengths ( A ) . (b) Bond angles ( * ) .
369
3.2.3
Crystal
and
molecular
benzothiazolino-sr)iroDyran
structure
of Cobalt comvlex of a
The structure of the complex CoC12 ( 0 , 5 acetone) 18-methoxy-
3-methyl-6-nitro,[2H-l]benzopyran-2-spir0-2~(3~-methyl-ben~0thia-
zoline)] was solved by the multisolution technique and the atomic parameters refined by full-matrix] least squares calculations, to an R value of 0.09 for 25.44 observed reflexions. There are two molecules in the asymmetric unit: one in the form of dimer ( A ) and the other in the form of monomer (B). In both molecules the coordination sphere of the cobalt atom is composed of oxygen and chlorine atoms CoC1~02 for ( A ) and CoClz0~ for ( B ) (the oxygen of acetone solvent is involved) (ref. 144). The geometry of the coordination polyhedra is determined from the mean angular and dihedral angle deviation from idealized trigonal bipyramidal and tetragonal pyramidal models. It can be described as a more (in A ) or less (in B) distorted trigonal bipyramid. The geometry of the ligand could be associated with that of the hypothetical intermediate of the spiropyran t merocyanine photoconversion process. In the crystal lattice, molecular stacks of monomeric (along b) and dimeric complexes (along a) are formed and cross-linked through short van der Waals contacts. Perspective views with bond lengths ( A ) and bond angles ( * ) are given on the Fig. 14 for monomer and dimer structures. The bond C221-01 is near 2.65 A corresponding to the breaking of this bond and stabilization of the cisoid structure by cobalt chloride, 013)
370
371
(c)
S1'- C22'-N3': 122' *CZZ'-N3'-C9'= 116'
Fig. 14. (a) Perspective view and atomic numbering of the cobalt complex of benzothiazolino spiropyran (dimer and monomer) ; (b) Bond lengths ( A ) : (c) Bond angles ( ' ) Coordination chemistry of the photochromic eauilibrium or sewarately of the swiropyran and the vermanent merocvanine The colorless and colored form of a spiropyran are distinct chemical species and may undergo different chemical reactions. A selective chemical reaction of one form only is the basis of several methods of desensitization. One highly reactive function present in the colored form but not in the colorless form is the phenolate anion. This in conjunction with another appropriate groups can give a chelating moiety. Phillips and his coworkers (refs. 145, 146) have used 7-formyl-8-hydroxy-quinoline and its derivatives to give various spiropyrans ( 8 ) whose open forms obtained by irradiation contain the oxine function as shown in the following equation : 3.3
\
/
372
The well-known chelating properties of oxine were retained in (b) for treatment of the colored solution with metal ions and gave the deeply colored chelate (n). Experimentally a colorless solution of trimethyl (EL) in acetone at room temperature gave a red-purple coloration a few seconds after being treated with Cu2* or Fe3* ions ; at -78'C the solution gave no coloration for many minutes. When the colorless reagent solution was first irradiated it turned purple, and the subsquent addition of metal at -78'C ions produced an immediate deepening of the color. Finally, irradiation of the colorless reagent-metal ion solution at -78'C gave only the ultimate deep purple color. With-Niz+, Cop+ , ZnZ+ and Cd+2' ions the absorption spectra of the chelates ( c ) were negligibly or only slightly different from those of the unchelated dye (b). Different types of chelate groupings were described by Taylor (refs. 147, 148). These spiropyrans ( d ) possess, in the 8-position1 the groupings shown below :
C
R = CH20H, CH20R', CH2NR2
CH=NR, N=N-Ar
Experiments similar to those of Phillips demonstrated the ready formation of stable chelates with Zn", Cozt, NiZ* and Cuz+ ions. The chelates from d_ (R = CHzOH or CHZ N (CZHS)z) were generally orange or red, but the structure of the complexes was not firmly established. The application in reprography of the photochromic equilibrium of the spiropyrans would require selective stabilization of the spiropyran (Closed Form, C.F.) and of the photomerocyanine (Open Form, O.F.), to preserve the image obtained (refs. 86,149).
Spiropyran
Cyclized form (CF)
Photomerocyanine Open form (OF)'
373
We have studied the complexation (ref. 150) of merocyanines, in which the nitrogen-containing component is benzothiazoline, by MnII, C o I I and ZnII halides ; these merocyanines may or may not be stable in the open form. In the indoline series, the cyclized spiropyran can be opened by metal salts to give complexes similar to those isolated from benzothiazoline-containing merocyanines. The 1:l complexes are covalent, the spectra and magnetic properties of the cobalt complexes agree with a pseudo-tetrahedral environment of the metal ion. Proton N.M.R. of the zinc complexes and IR spectroscopy show a metal-oxygen bond to be present. However, there is no indication of a bond between the metal and the heterocyclic nitrogen atom. This agrees completely with the results obtained by Pommier et al. (ref. 151), Becker et al. (ref. 152) and Simkin et al. (ref. 153) showing that nitrogen heterocyclic merocyanines are largely dipolar, with a positive charge on the heterocycle and negative charge on 0 - 1 which is the preferred site of reaction. The presence of an ortho 8-OCH3 group encourages chelation of the metal ion by the oxygen atoms 1-0 and 8 - O C H a . Structural considerations, in particular related to steric hindrance (ref. 8 7 1 , led us to propose two geometrical structures for the metal complexes depending on the substitution at position 3 (Figs. 15 and 16).
Fig. 15. Probable structure of the complex when
R3
= H.
314
Fig. 16. Probable structure of the complex when R3 = O C b . The coordination chemistry of the first row transition metal Ni, Fe, Mn) and zinc halides and thiocyanates with benzothiazoline, thiazolidine and oxazolidine spiropyrans has been studied (ref. 1 5 4 ) . The molecular structures of different complexes have been investigated by physico-chemical methods such as IR, NMR, electronic absorption, magnetic moments and conductivity measurements, Debye-Scherrer diagrams and microanalysis, in solution and in the solid state. The methoxy substituent in position-8 plays an important role in the coordination of spiropyran, which becomes then a bidentate ligand. With metal halides either dimeric five-coordinated complexes are formed or monomers with a solvent molecule as the fifth ligand. Pseudo-tetrahedral complexes are isolated with spiropyrans unsubstituted in position-8. For metal thiocyanates, the physico-chemical data are in agreement with those observed for five-coordinated metal halogen complexes with 8-methoxy spiropyran ligands. The fivecoordination of each metal ion is established by two bridging and one terminal NCS group and by a bidentate spiropyran ligand. Thiocyanato complexes with spiropyrans not substituted in position-8 could not be separated. The X-ray molecular structure analysis of the CoClt benzothiazoline spiropyran complex contributed to the interpretation of the spectral characteristics. (Co,
375
3.4 Dipole moments of spiropyrans and solvatochromy of soluble model-merocsanines 3.4.1 Experimental and theoretical study of the dipole moments of a series of benzothiazoline spiropyrans For the conformational study of a series of benzothiazoline spiropyrans, experimental and theoretical (EHT method) dipole moments have been compared (ref. 155). The Hedestrand method (ref. 156) and the formulation of Cuggenheim (ref. 157) have been used for their experimental determination. The dipole moments have been measured for a series of benzothiazoline compounds (Table 13) and for the two halves of the spiropyran molecule : the 2,2,3-trimethyl-benzothiazoline and the 8-methoxy-6-nitro 12Hl chromene in benzene solution at 2 5 ‘ C ( A p = 2 0.03D).
= 2.40 D (pcalc. = 2.42 D)
p
= 4.02 D pcalc. = 3.71 D) p
For the spiropyran molecule two enantiomers I and I1 exist differing in the configuration of the asymmetric carbon atom. The two moieties are perpendicular to each other and so each enantiomer has 4 possible conformations (aa’, ba’, ab’, bb’) (Table 14.).
The energetic data of the semi-empirical calculations for the eight possible spiropyran conformations (four limiting conformations for each enantiomer) and the comparison between experimental and calculated dipole moments fit quite well, and allow an interesting approach to the conformation o f these molecules in spite of their complexity, The calculated (bb’) conformations correspond to the dipole moment which is nearest to the experimental values and the following correlation : [pexp = 1.00 pcalc. -0.16 (r = 0.97)] waa verified for 13 compounds studied.
(19)
376 TABLE 1 3 . D i p o l e m o m e n t s of b e n z o t h i a z o l i n e s p i r o p y r a n s .
R' '
R'
R'
R'
H
CHI
H
H
1.78
H
CHI
NO2
H
4.80
H
CH3
H
OCHI
1.98
H H H
CH3 CH3 CHs
NO2
OCHa
4.27
OCHJ
NOa
H
CHI
H H
CHI CHI
c1
NO2
H
CHI OCHa
H H
V(D)
............................................
c1 c1
4.58
H
2.56
c1
3.15
Nor
4.75
Br
Br
OCHI
3.86
H
H
2.73
OCHa OCHI
NO2
H
4.93
H
OCHa
2.63
H
OCHI
NOz
OCHa
4.18
H
0CisH1s
NOz
OCHa
4.28
OCs HI
NO:
OCHJ
3.75
Br
H
H
H H
OeCiOH7 Ca Hi i C s Hs
3.01
NO1
OCHa
3.65
N0z
OCHa
'4.16
NO?
OCHa
4.31
377
TABLE 14. One of the possible forms of spiropyran molecules and tions for the two spiropyran enantiomers (I), (11).
conforma-
I b,,'
I
I a,b'
1 a, b'
I blb'
I b,b'
3,4.2 Experimental determination of dipole moments in the ground state of heterocyclic permanent merocvanines The dipole moments of a-0x0 and 8-0x0-benzothiazoline, aoxobenzimidazoline and #-0x0-benzodithiole merocyanines (ref. 119) (Table 1 5 ) have been evaluated in benzene and in benzenechloroform mixture, The experimental difficulties due to very low solubility and intense coloration of some of these compounds have led us to use the methods of Halverstadt and Kumber (ref. 158) and also of Gilkerson and Srivastava (ref. 159) (Table 16).
378
TABLE
15.
Merocyanines studied. a-ox
8 -oxygen
ygen
benzothiazoline
X = S, Y = N-CisHii 11 : R6 = NO2 , Ra = H I I A : Re = N O z , RE = OCHs benzimidazoline I Y = N-CisHii In : Re = NO? I RE = H 11s : R6 = NO2 , RE = OCHi
X = N-CHa
benzothiazoline X = S, Y = N-CHi Ic : R3 = H benzodithiole
X = Y = S 10 : R' = CH3 IID : R' = C6H5
TABLE 1 6 .
Dipole moments of merocyanines determined at 25'C by the Halverstadt-Kumber method (a) and by the Gilkerson-Srivastava method (b). Compounds
Solvent
p ( D ) method (a)
p(D) method (b)
................................................................ cs Hs 8.3 f 0.2 CsH6 85 - CHCli 15 8.8 t 0.2 8.9 +- 0.2 I A
IIA
-
CHCli 15
-
CHCli 15
9.3 f 0.2 11.3 f. 0.3
11.5 f 0.3
cs Hs CrH6 85
CHCli 15 CHCli 15
12.8 f 0.3 13.2 t 0.3
12.9 f 0.3 13.1 -+ 0.3
-
CHClt 15
Cs H6 CsHs 85
-
CHClt 15
C6H6 85
In
cs Hs
IIB Ic
CeHs 85 CsHs 85
ID
IID
C6H6 85
10.4 2 0.3
2.7 f 0.2 2.8 ?r 0.2
4.5 +- 0.1 4.5 f 0,2
10.6 2 0.3
2.8 f 0.2
4.7
+- 0.2
379
The study of dipole moments revealed a slight polarization of the azaheterocyclic compounds. The polarization does not occur They present an in the case of benzodithiole substrates. electronic distribution of the quinoxdal or polyenic type. Moreover the strong polarizability of some solutes has been checked. 3.4.3 Solvatochromy of permanently stable merocyanines. Solutesolvent interactions
3.4.3.1 The absorption and fluorescence frequencies of lower energy transition of some a-0x0 benzothiazoline, a - 0 x 0 benzimidazoline and l(-oxo benzodithiole merocyanines have been measured in thirty one solvents of different polarity (ref. 120). The structure of these compounds, mainly for the a-0x0 derivatives is quite similar to the photomerocyanines derived through the photolysis of spiropyrans. The different terms of the MacRae theory (ref. 160) have been analyzed. The effect of electric dipolar interactions and of dispersion on the frequency shifts of lower energy transitions has been studied. The important part played by the solvent refraction index has been emphasized for the benzodithiole 1(-oxo merocyanines showing a positive solvatochromism. The difference between the solvent polarity parameters X R and XB proposed by Brooker (ref. 76) could be a-oxo interpreted. The benzothiazoline and benzimidazoline merocyanines have a negative solvatochromism. The unusual solvatochromism observed in the case of alcoholic solvents has been ascribed to a specific interaction (hydrogen bonding) between the solute and the solvent and has allowed an analysis of the basicity of those compounds (ref. 120). Finally, we have explained the inversion of solvatochromism observed for these s o l vents upon the fluorescence spectra. 3.4.3.2
By using the MacRae (ref. 160) and Kawski theories (refs. 161, 162) about Bolvent effects and taking into account the Stokes variation of ( v . - V r ) as a function of solvent polarity, the dipole moments of a-0x0-benzothiazoline, a-oxobenzimidazoline and E(-0x0-benzodithiole merocyanines have been determined, in their first singlet excited state (ref. 120). These compounds are similar to the photomerocyanines obtained from flash-photolysis of spiropyrans. The values of dipole mo-
380
ments (Table 17) are interesting for the interpretation of charge
distribution in the excited state of colored forms and a qualitative correlation may be made between the new electron structure and different experimental results concerning photo-bleaching and photodegradation phenomena. TABLE 17. Dipole moments in the excited state ( p e ) at 25'C
Compound
MacRae method
IA IIA Is 11s
1.8 4.6
IID
8.5
Kawski method 2.3 5.1 4.2 8.2
3.7
7.7
5.6
ID
(in D).
5.6
8.4
3.5 Identification of comwounds and study of the fragmentation by mass swectrometry Very few studies have been published in the literature concerning this field. Nevertheless
spiropyrans
the
mass
spectra
have been reported (ref,
of
163).
12
benzodithiole
Their fragmentation
pathways have been determined by exact mass and metastable transition measurements. The nature of the fragmentation was found to be strongly dependent upon the position and nature of the aubstituents.
381
TABLE 18.
List of benzodithiole spectrometry.
1 2 3 4 5
6
7
8
9 10 11 12
117
155 188 202 228
210 212 171 147 209 195 223
284
346 359 421 359 329 421 373 435 4 50 313 435
spiropyrans
investigated
H
CH3
H
H
Cs Hs
H
H
CHJ
OCHa
H
CS HS
OCHa
H
CHs
NO2
H
CH3
N0z
H
cs Hs
by
Mass
NO2
CH3
CHI
OCHa
CH3
Cs Hr
OCHa
H
Cs Hs
CH3
CH3
CHa
c6 HS
COCi Hs
NOz NO2
The mass spectra were registered on a VARIAN MAT-311 spectrometer at temperatures between 100 and 150 'C and at energy of 70 ev.
382 Two
examples
of
simplified
fragmentation
reproduced (Schemes 12 and 13).
q,H,,o,N
\
8%
4
0
are
s?'
/ /
schemes
/
\
/
\
\
: 1
\
/
3%
A0
NO2
\ 21%
f
Scheme 12. Electron beam (70 ev) fragmentation pathways of a benzodithiole 3-phenyl-6-methoxy-8-nitro
:1
CwHoOlh
spiropyran.
-$
CidhNO1Sz
Scheme 13. Electron beam (70 ev) fragmentation pathways of a benzodithiole 3-methyl-6-nitro-8-methoxy spiropyran.
383
The study of metastable transitions [DADI
(direct analysis and high resolution measures have allowed the different types of fragmentation mentioned on schemes 12 and 13 to be clarified. We observe the loss of the methyl group (for R3 substituent)
of daughter ions) and AVS (aooelerating voltage scan)],
the methoxy group (mainly for Ra substituent), the nitro group (mainly for RE substituent), the benzodithiole skeleton and the benzofuran skeleton.
3.6 Characterization and structural aspects by Infrared and Raman
spectrometry The number of studies concerning the normal vibrational modes in the spiropyran series is very limited (refs. 164-1671, T h e most interesting work was the one of Arnold and Paal (ref. 1 6 5 ) where these authors compare some IR vibrational bands of homologs such as indoline, benzothiazoline, benzoxazoline spiropyrans or spiro[l,4]oxazines. The difficulties met for direct interpretation of vibrational modes of beneothiazoline spiropyrans (ref, 168) have led us to make a preliminary analysis of IR and Raman speotra of the two constitutive halves of the spiropyran moleoules, We have synthesized some beneothiaeolines and some substituted 2H-chromenes as models of the two halves of beneothiasoline spiropyrans with the aim'of interpreting the infrared and Raman spectra of spiropyran structure (ref. 169).
While doing that, we have been able to improve the synthesis of chromenes from coumarins and have prepared new nitro-chromenes (cf. section 5 synthesis). By studying IR and Raman spectra of the two halves of spiropyran molecules, we are able to propose a rather satisfactory assignment of fundamental modes of vibration. This will be used as a base for the interpretation of IR and Raman spectra of a series of benzothiazoline spiropyrans. Arnold (ref. 164) in a general study on spirobibenzopyran derivatives correlates valence vibration of C spiro-0 bond to a
384
strong
band
in
the
910-960 cm-1 region.
In the unsubstituted
spiro bibenzopyran, Arnold locates this band at 952 cm-1. We observe 2 absorptions at 968 cm-1 ( E = 1538) and 952 cm-1 ( E = 324) this last one being linked to a T(CH). The IR spectra of 5 benzothiazoline spiropyrans (in the solid state and in solution) and Raman spectra.of 3 of them (in the solid state) have been registered from 4000 to 400 cm-1 and
have allowed an interpretation of molecular vibrations of these compounds (ref. 170) (Table 19). A simple analysis for the characterization of spiropyran structure based on IR spectroscopy has been realized with a series of twenty one compounds.
Moreover the infrared study has been developed by interpretation of the main spectral perturbations observed when a
spiropyran (closed form) is transformed into a merocyanine (open form) (ref. 170). Schiele and Arnold (ref. 171) and Dzhaparidze et al. (refs. 172, 173) have undertaken such IR studies about the opening of indoline spiropyrans. We have studied one permanent benzothiazoline merocyanine (R
= H) and two photomerocyanines (R = OCHJ and C H 3 ) generated from UV photolysis of benzothiazoline spiropyrans.
R : H ,CH,,OCH,
The photomerocyanine for which stable in the solid state.
R
= OCHI being
thermally
Two spectral regions (1700-1500 cm-1 and 1350-1200 cm-1) are particularly interesting for the analysis of opening of spiropyrans (Figs. 17 and 18).
TABLE 19. IR and Raman spectra of benzothiazoline spiropyrans : Assignment of different absorptions-
R3-CH3 R~=H R8=H
IR
C[m-']
R. I
c
-3088e 3070
359e -304Oe
--Mm
52
3070 --M& m3c
-
4
3032
3021e
3ooo 16S2 1610 1588 -IS&
14m 1476 1459 1448 1423 1378 1352
F[cm-']
R'-CH3 ~
43
20
27 F
57 F
181
m
c
284m
575 110 142
IR
87
82 f
164 m
47
1478
543
1474e
144s 1423 1377 1354
0
~
40 43 30
203
57=)
97
79 207
3074 3x1
-
3040
3099
34 21 20
302%
-3#k
1656 1615
1588 1580
R3=CH3 R6=N4 Rs= OCHs IR C[cm']
&
-3IDoc -3088~
2 8
-- 1463
4
qm-']
c
-3039 3026e 3005 -1652 1607~ 1588 1581
4
Rs= H
-309oc
-31ooc -307%
R3 CH3 R~-H R*= OCH3 IR
M
113 23$
m
-
3080e 3069
MS5 3038
-
c
K I
-Ms4c
--MS6e
41 37 28
3069
16SSfl 1620
1588
48 63 349
I AT
1530
566
1474
785
1476
I270 m
1449 1422
183 118
1344
l2SO
+4S2 1424 137% 1361 1339
49s f 240 el e -3083 F 1390 FF
1380
100
-
28 42 *
t(CH)
Mo8
63
I:
710
1528
R I
t
3034e
302%
3006
F[m-']
3101
33
Assignments
R3= OCH3 R6=N02 R8= OCH3 IR.
-
1478
-1 1452
1337
1189 m c
1
BT2dc4
304 m
BT3 d C s
1290 FF
i9401
w
%
TABLE 19 continued
-
1314 110 c 173 1301 1287~ 10s J 1276
- I246e 1232
-I
1227 l82C 1167 1160 1126 1116 1067
ION
1026 858 842 832 788 750 738
715
370
m
390
I70
I276
378
1259 123C
- 1256e -
-
1228
I191
79
42 m 314 1
310 I IS 145 f 147 197
1x2
F
IS0 f
84 58 f 5 10 528 f .95
- I127
II f 0 I161e
1300
191
1302
115 e e )
1288 I263
132 560
1’38
440
m
234
1241
540
445
554 83 I24
1226
235
1251 1238 12#k
1T fl
- I t26
98
1117 1063
282 267 106
I026 86I 837
I24 119 I 13
1050
1110 782
731 715
147
70
61 575 98
It68
116oC
I120 1069 1049 LO26 863 839 820 786
738 715
R is for Raman ; I R is f o r Infra Red.
353
191e) 108 f 32
132 328 286
-1188
I167 I160 I133 1 I23 1068 1050 1026
864 850
831
208
570
m
530
73s 714
93
520
I32
6 2 m
61 546
370 m 171:
- 1309
1300e 1292
829 m
1230
51s c
1I96
1I73 1158 1134 -1IUk
IOSl 1026
262m 238 m 123 m
-we
m / 85
553 FF
1246
168 I32 m
350
678 c
855
818 n4
736 712
281 lT 283 108 fl
691 c
f
1%
1
389 452 .c
I51 c 131 F 829 1
151
( a ) The wave numbers ( T I are generally taken in the spectra of compounds in solution and they are followed by the molar absorption coefficient ( E ) . ( b ) The wave number ( 7 ) taken in the liquid state are written in italic characters. (c) I = intensity : FF = very strong ; P = strong ; m = medium ; f = weak : ff; very weak ; e = shoulder. (d) The sequences noted BTi (i = 1 to 1 3 ) and C j ( j = 1 to 20) are respectively related to the modes of vibration involving the beneothiazoline and the beneopyran nuclei. ( e ) E caloulated after wings correction.
387
1700
I
I
1600
1500
I
1400
S(C
6’)
Fig. 17. Comparison of closed and open forms IR spectra between 1700 and 1450 cm-1 (solid state) [3,8-dimethoxy-6-nitro~ZH-llbenzopyran-2-spiro-Z’ (3’-methyl-benzothiazoline)] corresponding photomerocyanine ( T = transmission).
------
388
T
1400
1300
1200
1100 S(C
6'1
1000
Fig, 18. Evolution of IR spectrum in the rekion 1350 to 1100 cm-1 during the conversion of spiropyran into photomerocyanine. Example of ~3,8-di~ethoxy-6-nitro~2H-l]benzopyran-2-spiro-2i (3'-methyl-benzothiazoline)J. (1) Spiropyran ( C S Z solution) ; (2) (3) and (4) after subsequent irradiations ; ( 5 ) corresponding photomerocyanine (solid state, after photoprecipitation,, T = transmission).
389
Indeed, in the opening of spiropyran we observe an important
delocalization of the ethylene double bond and also a shift towards the high frequency of the VCO compatible with a localization of the C=O bond. These spectral data are in agreement with an electronic delocalization and a certain polarity of C-0 bond of open forms and also consonant with the infrared results obtained for thiazolidine spiropyrans and merocyanines (ref. 174). Benzothiazoline spiropyrans and merocyanines have been studied by IR spectroscopy, their structure (closed form or open form) characterized, their purity can be checked and also degradation products can be detected (ref. 175). The thermal stability of benzothiazoline spiropyrans has been also investigated on semi-conductors (TiO?, ZnO, A h 0 3 , SiOz) by the reflexion technique (ref. 176). Quantitative measurements were made directly with spiropyran layers on solid supports (polyester film) with polymeric resins (rhodopas or ethyl cellulose) under conditions of photographic application (ref. 177).
3.7. Conformational calculations on the oRen forms by different mechanical and auantum methods. Influence of steric hindrance, correlation with thermal kinetics The kinetic study of thermal fading is mostly a macroscopic description of the phenomenon ; the photomerocyanine form can take many possible configurations indeed. In the literature, some authors have made suggestions of four transoid stereoisomers for the open form not yet satisfactorily defined. Experimentally, the UV-visible spectra of the open form are variable in time after photolysis and are dependent on parameters such as temperature, viscosity, solvent, excitation wavelength, etc. Most authors have admitted three or four different spectroscopic species for the photomerocyanine without being sure of the cor,rect interpretation. Thus we have tried a theoretical approach o f configuration and conformation of the most stable open form of benzothiazolinospiropyran (refs. 178, 179). Chaude (ref. 38) firstly, then other authors (ref. 39) have made similar calculations, in the indoline series, taking into account only the electrostatic energy between N' and 0- dipoles.
390 Fully aware of the importance of steric effects in the thermal bleaching rate, we have estimated the non-bonding interactions by using an empirical method such as the Hill equation cal-
culating the van der Waals energies (refs. 180-182). E N D = B/r6 t Ae-cr (20) E in kcal; r in A ; A and B in kcal, depending upon the nature of the pair of atoms concerned ; C is homogeneous to l/r in A . (The calculations have been made with Symon’s program on an IBM 1130). The principal interactions evaluated are shown in Fig. 19.
Pig. 19. Representation of the principal types of steric interactions in benzothiazoline spiropyran series. We have used two types of geometry for the molecule (quinonic and ionic geometries). In the computation, we have assumed the planarity of the open form and have considered the four trans forms and the four cis forms ( P i g . 20).
391
A!!-. 0
- 0,
- 0. --.
'"
I
IV
' 0
Fig. 20. Trans- and cis-configurations of benzothiazoline photomerocyanine. These latter forms are not compatible because energy increases greatly with the internuclear distances. The non-bonding energy ( E N B )is growing with steric hindrance, in agreement with kinetic results (Table 20).
392
TABLE 20. Parallels between non-bonding energy for the trans-configurations and thermal fading rate of photomerocyanine of 3’-methyl benzothiazolino-spiro-6-nitro-8-methoxy-benzopyran. Non bonding energy
R = H
R = CHs
36.5
10.3
79.3
53.2
842.8
273.7
789.1
200.1
(506.5)
(316.4)
(434.5)
(244.4)
2.5
R = CZH5
245.4
23
R = i-CaH7
368.5
680
R = OCHI
48.2
76.9
R = SCHJ
338.5
203 3
9.5
.
10-3
76
The numbers between parentheses are energy values for quinonic geometry. In general, the stereoisomer (IV) is sterically the less strained and so the most stable except for H and OCHI substituents in 3-position for which the most stable is the stereoisomer (11). We have also evaluated electronic energy and stability of the heterocyclic system by using the extended Hiickel Theory (ref. 183). We have chosen the plane model (IV) with a quinonic geometry. The values of angles and energies are certainly overestimated by the Hoffmann method itself, nevertheless the most stable electronically open form seems to be twisted and not planar (Fig. 21).
393
-iPr
if5
-16.10
0.12 I
Fig. 2 1 . Curves& = f ( 0 ) for substituents (H, Me, iPr) in 3 position of a thiazoline merocyanine model. We have made calculations for hydrogen, methyl and isopropyl substituents in 3-position. Finally, we have used the latter two types of calculations : the Ifill equation ( E N D ) and a quantum method such as that of Pariser-Parr-Pople ( x electrons with interaction between nonbonded atoms). We added comparable energies and have made calculations only in the case of the less strained merocyanine (R = HI. For the four forms we have investigated all conformers and the comparison of curves of total energy as a function of 8 (dihedral angle between the two parts of the molecule) shows that the more stable conformation corresponds to the twisted trans form (11) for which 0 = 30' (Fig. 2 2 ) .
394
Fig. 22. Curves E = f(8) for the forms I, 11, 111, IV of a benzothiazoline merocyanine model. The precision of the twist angle is not so good because the rotational barrier is weak. The other minima observed at 8 = 60'
for forms I11 and IV and at 8 = 120' for all the four forms (corresponding to a cis twisted configuration 8 = 60") do not matter very much. As we have found in steric interaction same order of stability is verified. I1 > I )) IV > I11
calculations,
the
395
One should regard these calculations carefully as they contribute to a progressive understanding of the possible structure for the open form. The non-bonding interactions have confirmed kinetic values in benzothiazoline series and results obtained in a new series, the benzodithiole one (ref. 1 1 5 ) are also compatible with electronic and steric parameters. In conclusion, calculations of steric hindrance (van der Waals type) were checked on different planar "open" forms of stereoisomers of a benzothiazoline spiropyran, by choosing two types of geometry (dipolar and quinoidal structure) for the molecule. They have permitted classification of the various stereoisomers and identification of the sterically favored forms. The results in a homogeneous series have shown the importance of substitution in 3 position and established the firm correlation between steric hindrance and thermal fading kinetics. Furthermore, the theoretical conformational study by extended Hiickel Theory has been undertaken with some approximations and gives evidence for a twisted stable trans form : this result was confirmed by calculations involving electron x energy 184-186) and non-bonding (Pariser-Parr-Pople method) (refs. energy. 3.8 Calculation of electronic charge distribution and of electronic transitions (compared to experimental values) by quantitative methods
(i) The electronic structure and spectrum of a model compound of an indoline spiropyran were investigated using an all valenceelectrons CNDO-CI method (ref. 187). The Z+R* electronic excitations are localized on a given half of the molecule. The photochromic process is discussed on the basis of charge densities and bond orders (ref. 188). Previously Chaud6 and Masse (ref. 189) had used the simple HMO method only, (ii) The electronic structure of benzothiazoline spiropyrans and derived photomerocyanines was investigated by the CNDO/2 method using different appropriate structural models with a progression in their complexity (ref. 151). For closed forms, the results are in agreement with forecasts. For the isolated open forms
396
theoretically, a quinoidal structure was found in the isolated state, and energetically, twisted conformations are possible. This assumption was verified firstly by Becker et al. (ref. 190) with open forms of benzopyrans and then by Simkin et al. (ref. 191) on simplified open form models. These authors have compared the experimental electronic absorption spectra to the calculated transitions through the Pariser-Parr-Pople method and have found agreement with a dipolar structure. In solution, the aolvent has surely a great influence on the localization of electrons, in complete agreement with the other structural studies like dipole moments and interactions solutesolvent. (iii) The excited states of electronic transitions of some 2Hchromenes (considered reaponsible for the photochemical opening of photochromic spiropyrans) have been studied by the CNDO-S/CL method. The relative positions of nz* and nx* singlet and triplet energy levels of these compounds allow a satisfactory interpretation of their optical properties. Also highly localized electronic charge transfer is demonstrated in the case of 2Hchromenes substituted by a nitro group (ref. 192). The 2H-chromenes investigated were : the 2,2,3-trimethyl 2Hchromene, the 6-nitro-8-rnethoxy-2H-chromenel the 2,2,3-trimethyl6-nitro-8-methoxy-2H-chromene, the 2,2,3-trimethyl-6-methoxy 7nitro-2H-chromene, the 2,3,3-trimethyl-6-methoxy-8-nitro-2H-chro-
mene. The theoretical and experimental results are given and the relative position of nx$ and A X S singlet and triplet levels is assigned and emission properties are interpreted (ref. 192).
3.9 X-ray photoelectron spectroscopy of benzoxazoline sDiropyran and derived permanently stable heterocyclic merocyanines Since the chemical shifts observed in X-ray photoelectron spectroscopy (XPS) are directly related to the partial charges, this technique provides an interesting means of obtaining a better understanding of the electronic structure of the merocyanines (ref. 193) and has demonstrated the existence, in the solid state, of the zwitterionic form of these molecules.
397
On the basis of the measured binding energies, a classification of different heterocyclic merocyanines according to their polarity may be proposed (ref. 194). However, photomerocyanines produced under UV-irradiation are transient chemical species having half-lives in the millisecond to second range and can be studied only by using time-resolved spectroscopy such as time-resolved Raman spectroscopy.
398 4. PHOTOCHEMICAL AND PHOTOPHYSICAL STUDIES OF PHOTOCHROMIC INTERCONVERSION, PHOTOCOLORATION AND PHOTODEGRADATION
A recent review on the photochemistry and photophysics of spiropyrans has been published by Kholmanskii and Dyumaev (ref. I ) where the results of studies on the structure and nature of the electronic states of spiropyrans and the products of their photochromic reactions as well as the mechanism of the dissociation of the C spiro-0 bond and the formation of a colored form of the spiropyran are surveyed and analyzed. Also discussed are the mechanisms of the a-dissociation and predissociation of the bond, which make it possible to explain satisfactorily the dependence of the quantum yield of the photocoloring reaction on the structure of the spiropyran, the nature of the matrix, and temperature. Analysis of the experimental results has shown that the photochemical activity of spiropyrans and their luminescence characteristics are determined by the position of the energy levels and the orbital nature of the lowest electronically excited states. Furthermore, in order to understand the mechanism of the photodissociation of the bond in spiropyran, it is essential to take into account the degree of involvement of the electrons of the heteroatoms in the configuration of the electronically excited states and also the interaction of the electronic orbitals of the mutually orthogonal heterocycles in both ground and excited states. The spiropyran photocoloration reaction can be represented schematically as follows (ref. 195) :
SP Scheme
hv
'SP
ad
X-
Q>
PM
14.
where SP is the colorless initial spiropyran, lSP* its electronically singlet excited state : X the cis-cisold isomer of the colored form arising after the dissociation of the C-0 bond and
399 still preserving the orthogonality between the two heterocyclic halves : PM is the pseudo-planar colored form of the spiropyran (photomerocyanine) $ = quantum yield in the formation reactions of the PM form. $d = quantum yield in the bond dissociation process. The recombination reaction of the photoproducts with transition to the initial state and also the possible pathway leading to the formation of the PM form without formation of the intermediate X isomer are not shown in scheme 14. 4.1 Nature of the electronic states of spirouyrans and products of their photochromic reactions 4.1.1 Structure of spiropyrans and their photoproducts The X-ray diffraction structures of different spiropyrans and merocyanines have been, by now, investigated (refs. 135-144). These are greatly influenced by the ring substituents. It is of interest to compare the length of the dissociating bond g with the activation energies Ea for the thermal bond dissociation in spiropyrans and the quantum yield in the photocoloration reaction. As a rule, Ea falls with increase of 2 ; this correlation is a consequence of the fact that the dissociation of the bond in the thermal reaction proceeds via the ground electronic state of the vibrationally excited spiropyran. On the other hand at 313 or 365 there is no correlation between the quantum yield nm in different media (ref. 7) and c. This shows that the photodissociation of the bond in the spiropyran proceeds via the electronically excited state (probably IS*) of the spiropyran whose orbital nature determines the mechanism and the quantum yield $ of the spiropyran photocoloration. The geometry of the molecule and the orientation of substituents have their importance for the intensity of electronic transitions and for electronic interactions between the 2 heterocyclic halves. In order to determine the structure of the colored form of spiropyran, IR (refs. 170, 196), Raman (ref. 197) and photoelectron (ref. 193) spectroscopic methods were employed in addition to X-Ray diffraction analysis.
.
400
TABLE 21.
General structure of spiropyrans whose photocoloring process studied.
s7
*;@rR
A
& 'NO2
with different substituents
%I
was
on the indoline or benzopyran part and especially NOZ-group in 6-position.
401
TABLE 21 continued
+JO,
R = H and CH,
Luminescence. Spectroscovib characteristics of spiropyrans and models of halves The study of the orbital nature of the electronically excited states had, as its primary aim, the elucidation of the degree of involvement of the electrons of the heteroatoms in the electronic configuration of the lowest excited states of the x x * type. The orbital nature of the lowest electronically excited states of spiropyrans and their energy are determined by investigating the absorption and luminescence spectra using the theory of solvatochromism. Methods based on quantum chemical calculations for spiropyran or compounds modelling their fragments have also been used in order to achieve a qualitative justification (refs. 188, 198201) and estimate the charges in the excited states. A spiropyran with the nitro-group in the benzopyran ring ( A ) can be designated by DSA where D represent a heterocycle with a fairly low ionisation potential I’(indoline, xanthene, benzothiazoline, etc...) and S represents the spiro-carbon atom. It is known (ref. 2 0 2 ) that in molecules of the DSA type in which the D and A fragments are separated by three a-bonds, apart from the local n-m* transition in the A fragment, hyperconjugation via the system of a-bonds is responsible for the charge transfer transition from the D to the A fragment. 4.1.2
DSA
hv *
c@sRj
The positions and intensities of the absorption bands corresponding to the transitions in the A fragment and in the process described are determined alone by the value of I for the D fragment and the electron-affinity E A for the A fragment as
402
well as the energy of the interaction of the (DSAS) and (D+SA-)* configurations (ref. 202). In spiropyran the interactions with the x electrons of the D and A rings as a consequence of spiroconjugation (ref. 200) and the hyperconjugation of the n, a and x orbitals of the non-polar D and A rings in the ground state of the spiropyran can also lead to the appearance of a transition of the previous type (ref. 188).
S
-D
A
All the spiropyrans with a nitro group have emission of phosphorescence at 77’K and more particularly when the nitro group is in 6 position on the benzopyran ring (refs.25, 203-205); only fluorescence is observed for spiropyrans without a nitro group like BIPS or indolinospironaphthopyran (refs.61, 206, 207):
For some compounds we observe fluorescence and phosphorescence, and these luminescent properties will be explained by the relative positions of energy levels in function of the general heterocyclic structure and the nature of substituents. So different models of the benzopyran “right“ part such as 2H-chromenes and of the heterocyclic “left“ part such as _indolines or benzothiazolines have been synthesized and studied (refs. 117, 169, 2 0 5 , 208, 2 0 9 ) .
403
A good
approximation for (El
not synthesized till now
4
The absorption and emission (fluorescence, phosphorescence) characteristics of the three 2H-chromenes (B) ( C ) (D) with structures similar to those of good photochromic spiropyrans have been studied at room temperature ( t Z O ' C ) and at low temperature (-166'C). With compounds (B) or ( C ) by raising solvent polarity] a bathochromic shift of the upper band of the phosphorescence excitation spectrum occurred. That observation is in connection with results obtained in absorption : the lowest energy singlet is x x % type indeed. The difference of 27 nm noted on the S o + S l ( x x t ) band, between absorption spectrum and phosphorescence excitation spectrum in EPA, must be considered as a temperature effect, Similarly, Becker (ref. 61) has found a shift in absorption spectra by a comparative study of indolino-spiropyrans at room temperature and at 77'K. We note too a bathoohromic shift .of the (0-0) phosphorescence band, always by increasing solvent polarity which seems to prove that the triplet responsible for emission is also of A X * type (refs. 210, 211). Other arguments will support this suggestion : if we would suppose that the emitting triplet is a nx* type, it implies that the triplet ( x n * ) should be at higher energy, but, as we have seen before, the first singlet being of x x t type, the energy gap between singlet and triplet states would be wider for nx% than for x x % levels, so is in contrast to the general estimation about nxl: and x x % levels (ref. 212). Furthermore, if we consider the vibrational structure of the phosphorescence emission spectrum, the first two transitions
404 which are the most important ones, are separated in EPA by nearly 1100 cm-1. This vibrational spacing does not fit the symmetric 0N-0 stretching frequency as mentioned by Becker ref. 61) for the case of 6-nitro substituted indoline spiropyrans to which a nx* emission was assigned. This vibrational energy gap must be linked to a benzopyran nucleus vibration. Theoretical results obtained by the CNDO/S method (ref. 192) have also been used to propose a relative order of the lowest nn* and nn* singlet and triplet energy levels. The order of the energy levels plays an important role in the photocoloration of spiropyran (ref. 214). The relative disposition of lowest energy levels for 2Hchromenes (B) or (C) and (D) is illustrated in Schemes 1 5 and 16 and allows an interpretation of emission properties of such model compounds in agreement with Cowley’s observations (ref. 213) and 214) on indolinothe work of Russian researchers (ref. spiropyrans.
3,40 3,20
2,70 2,40
Scheme 15. Relative disposition of lowest energy levels for 2H-chromene ( B ) or (C).
405
I
Scheme 16. Relative disposition of lowest energy levels for 2H-chromene (D). The colored form (PM) exhibits only an intense fluorescence (refs. 25, 195, 215-217). In a recent study (ref. 218) of the photocoloring reactions of the 6-NOZ-BIPS at T\<4.2K, which stops at the state involving the formation of the cis-cisofd isomer X, the phosphorescence band of the X isomer was discovered (at 15000 cm-1 in trimethyl pentane and n-hexane). The formation of this isomer has been confirmed by the method involving the optically detectable magnetic resonance whose use together with theoreeical calculations made it possible to investigate the structure of the X isomer and to represent it in the following form :
Cis-cisord O . F .
406
It has been established by studying the photocoloring reactions of spiropyran at low temperatures and in the matrix comprising the compound itself that an absorption band whose frequency is close to the absorption band of the thermodynamically more stable trans-isomer of the PM form can be assigned to
the cis-cisoPd isomer of the PM form of spiropyranssuch as 6-NOz BIPS or spironaphthopyrans with and without the nitro group (ref. 219). 4.1.3 Study of the photochemical process involved in the oDeningl reaction of the chromene ring of Dhotochromio spiropyrans The analysis of absorption and emission spectra of different heterocyclic spiropyrans leads us to question the validity of the previously accepted general assumption of intramolecular energy energy transfer from the heterocyclic part to the chromene part (refs. 61, 205, 220) and to consider the concept of spiroconjugation (ref. 200). The singlet and triplet excited states responsible for the photochemical, opening of spiropyrans are of the X A * type with an intramolecular charge-transfer character towards the nitro group. The participation of singlet or triplet states depends essentially on the nature of the heterocyclic moiety and on the position and nature of substituents on the chromene moiety. In the indoline series the photochemical reaction proceeds nearly completely through the triplet state, whereas in the benzodithiole series it occurs almost completely in the singlet state (ref. 221). The "coloration" or "colorability" which was defined in section 4.1.2 (Spectrokinetic properties) permits an interesting comparison between different series of spiropyrans, concerning their power of coloration under some experimental conditions. Indeed, the "coloration" is directly proportional to the quantum yield of the opening reaction and to the molar absorptivity of photomerocyanine and was used as a means of evaluating the relative coloration yields through the singlet and triplet states in Stern-Volmer diagrams by testing a series of convenient quenchers or sensitizers under photostationary conditions (refs. 122, 222). Reeves and Wilkinson (ref. 223) have shown that in a rigid matrix the opening reaction of 6-nitro-BIPS involved essentially a photoreactive triplet, different from the photoemissive triplet responsible for the phosphorescence. Our investigations in
407
toluene solution (ref. 2 2 1 ) show that photoemissive triplets are the same state.
photoreactive
and
4.2 Mechanism of the rJrimary photophysical process of coloration
The principal question concerning the mechanism of the photo-dissociation of the C spiro-0 bond and the formation of the PM form is : 1 ) how the redistribution of electron density arising on excitation of the n-electron system of the spiropyran ensures the weakening of the C spiro-0 o bond and 2 ) what role the vibrational excitation of the spiropyran molecule plays in the bond dissociation. The intramolecular photophysical and photochemical processes are unambiguously determined by the orbital nature of the electronically excited state of the spiropyran which specifies the initial redistribution of electron density, regulating both the mechanism of the interaction of the electrons of the C spiro-0 u bond with other electrons of the molecule and the mechanism of the excitation of particular vibrations in the molecule. The orbital nature of the electronically excited state of the spiropyran is determined in its turn by its molecular structure. For this reason, the elucidation of the mechanism of the bond photodissociation makes it possible to relate directly the effectiveness of the photocoloring of the spiropyran to the structure, i.e. would allow the specific synthesis of spiropyran with particular photochromic properties.
q=flgo
4.2.1 The mechanism of the a-dissociation of the C spiro-0 bond The central part of the spiropyran can be represented as follows : nO
taken from A.S. Kholmanskii and K.M. Reviews, 56 (1987) 143 (ref. 7).
Dyumaev, Russian Chemical
Considering the close proximity of the nN orbital and the one can postulate the interaction of the electrons of these orbitals which consists in the mixing of the nN orbital and the unoccupied antibonding utC-0 orbital : the C spiro-0 bond is oC-0 bond,
408 weakened even in the ground state (ref. 7). The vibrational excitation of the spiro unit is also essential for the dissociation of the C-spiro-0 bond via the a-mechanism well known in the photochemistry of organic compounds (ref. 224) or a-dissociation of the bonds (type Norrish 1 reaction (refs. 225-227)). Both in spiropyrans having a nitro group o'n the benzopyran ring or on the indoline ring, and in nitro-2H chromenes appropriate model molecules for the benzopyran ring, we can assume that the ( D S S A * ) state of the spiropyran is likewise a photochemically active state and that the coloring reaction proceeds via the mechanisms (Scheme 17).
Scheme 17. The probability of bond dissociation on excitation in the photochemically active states of spiropyran is determined by the charges q on the oxygen and nitrogen atoms. The latter depend in turn on the change in the dipole moment of the molecule, which is correlated with the dipole moment of the transition / M I for transitions with intramolecular charge transfer (ref. 1 9 5 ) . It will be quite different for oxygenated spirocompounds such as benzofuranspirobenzopyran (ref. 228) where the coloring mechanism proceeds by an other pathway. The role of vibrational excitation and the mechanism of bond wedissociation in sDiroDyran Studies of the influence of the vibrational excitation of the electronically excited state on the kinetics of the photochemical reaction are of undoubted interest. For spiropyrans with the nitro-group, the photocoloring reactions proceed mainly.via the formation of the tridet state : the vibrational excitation of the photochemically active state is essential to ensure the effective interaction of the orbitals of the orthogonal rings, which constitutes the basis of the adissociation mechanism. The energy of the electronic excitation, 4.2.2
409
equal to the difference between the energies of the SI and TI states (5000-6000 cm-I), may be expended on this vibrational excitation (refs. 229, 230). This excitation is also necessary to ensure that the atoms whose bond is broken can move apart to a sufficient distance to avoid recombination (ref. 225). For spiropyrans without nitro-group and having an extended aromatic x system such as indolinospironaphthopyran the photocoloring reaction proceeds via the singlet state of the vibrationally excited molecule (ref. 7 ) . These results suggested that the bond disaociation in indolinospironaphthopyran proceeds via the predissociation mechanisms. The role of the dissociative level is then assumed by the electronic level of the cis-cisoPd isomer X, which interaeots the level of the photochemically aotive state in the vicinity of its first vibrational level. The formation of the isomer X leads to the isomerization mechanism for the degradation of the electronic excitation energy to vibrational energy (refs. 231, 232). It is possible to represent the path followed in the photochemical and thermal formation reactions of the isomer X in the PM form by potential energy curves and the transitions between them can be illustrated (Fig. 23). 4.2.3. The nature of the short-lived photowroducts arising in the flash or laser photolssis and pulse radiolysis of spiropyrans. The mechanism of the photocoloring reaction Numerous studies have been made of the classical photochromic reaction ;
SP
= .-hv
'SP'
-xQ)d
PM
by the method of kinetic pulse spectroscopy (especially laser spectroscopy (refs. 29, 233, 234-252)). The time resolution of pulse apparatus has been increased at the present time to the picosecond range (refs. 235, 241, 242, 250, 253). The values of hmax obtained in the above studies, characterising the absorption by short-lived photoproducts are generally near 430-450 nm and their lifetimes are in solution between 10-8 to 10-3 second.
410
€( kcal mol-'
reaction coordinate
Fig. 23. The potential energy surfaces for the ground and electronically excited states and the thermo and photocoloring reaction pathways of spiropyran.(taken from A . S . Kholmanskii and K.M. Dyumaev, Russian Chemical Reviews, 56 (1987) 145 (ref. 71.1 A large number of studies have been devoted to SPirOPYranS containing a nitro-group. The coloration of indolinospirobenzopyran (BIPS) and indolino-spironaphthopyran proceeds via the singlet state and only the absorption bands of the colored form PM but not the short-lived photoproducts are observed in the time range ) 10-7 second (ref. 7). The primary processes in the photoisomerization of 6-nitro 8-methoxy-2,2,3-trimethyl ZH-chromene (CHR) (a good model for the benzopyran part of the spiropyrans having a nitro-group) were studied in toluene (ref. 233) from -3 to t55'C by measuring the transient absorption changes produced by neodymium glass laser excitation at 353 nm (third harmonic) with a full width at half maximum of 6 ns. The results indicate that the formation of a long-lived (about 5 0 s at 24'C) colored isomer (I) occurs owing to C-0 bond opening, partly directly from the excited singlet state of chromene (CHR) and partly via the triDlet state. It is observed that bond opening leads to an isomer in the triplet state 3Z which subsequently transforms into the colored isomer (PM). The
411
lifetimes of 3SP and 3 X are about 80 ns and 4 5 0 ns, respectively, at room temperature. The following process may be Proposed [Scheme 18).
Scheme 18. Mechanism for the photochemical transformation of the 2,3,3-trirnethyl-6.-nitro-8-methoxy[2H]-chromene
After several excitations some photodegradation is'observed corresponding to a photochemical evolution of 2H-chromene by hydrogen shift into a phenol derivative (ref. 2 5 4 ) and the samples were renewed frequently duaing the investigations, The formation reaction of the colored form of spiropyrana having a nitro-group, by pulse radiolysis in benzene, proceeds via mechanisms similar to that described previously or more Precisely via the "triplet channel" (refs. 2 3 4 , 2 4 9 ) (Scheme 19).
SP
--
C6H6
3sP'
-
Scheme 1 9 . Photochemical pathway pulse radiolysis.
3x'-
for
SP
nitro-spiropyrans under
The presence of atmospheric oxygen accelerates the reaction increasing thereby the yield of PM. Two 6-nitro-8-methoxy-piperidine spiropyrans differently substituted in 3-position (R3 = H, R3 = C H a ) were studied by
3X-€JM
412
means of nanosecbnd laser photolysis (refs. 2 5 1 , 2 5 2 ) . In the case of the H-derivative, photoisomerization leads to an alltrans isomer (PM) which is formed almost entirely from the lowest triplet of SP (Scheme 20).
SP
+ hv
--jSP'
-3
sp*-3z
-3A-
pM'
- PM
Scheme 2 0 . Photochemical pathway for coloration of €I-derivative of piperidinospiro 6-nitro-8-methoxy-benzopyran. Four transients were detected along the triplet path. The first one 3SPS (t = 20 ns) is assigned to the lowest triplet state of piperidine spiropyran and the second one 3 Z ( T % 1 0 0 ns) to the triplet of the transient species obtained after the C-0 bond rupture. The third one S A ( T z 3 p s ) is suggested to be a triplet of transoxd structure, presumably the lowest triplet of the fourth transient, a trans isomer PM'(t % 8 p s , Xmax = 5 7 0 nm) which in turn leads to the final trans isomer PM (t 0 3 s ; max = 560 nm). Thus in the case of the H-derivative as in that of indolino-spiropyrans and other classes of spiropyrans, the ultimate stage of photoisomerization is the colored all-trans isomer PM (photomerocyanine). The striking feature in the photoisomerization of the methyl piDeridine derivative ( R 3 = CHI) is the lack of formation of the colored trans isomer, probably due to steric hindrance. The most stable isomer obtained in this case is assigned to the cis-cisoYd at 3 7 0 isomer X (t = 3s) characterized by absorption maxima and 4 4 0 nm. Such a cisoYd isomer X has been postulated to be an intermediate in the isomerization of spiropyrans - e.g. indoline, oxazine, thiazine (refs. 4 4 , 1 0 4 , 2 1 8 , 2 4 7 ) - but had never been isolated before at room temperature in fluid solution. This
413
isomer is formed exclusively via the first excited singlet of the methyl-derivative. However, two consecutive triplets were also detected on laser excitation : the first one ( r :: 25 ns) is assigned to 3SP* and the second one 3 Y ( z a 2 p s ) to the triplet Of an isomer formed after rupture of the C-0 bond but retaining a geometry close to that of the initial spiropyran. Both triplets relax to the ground state of the methyl-derivative instead Of leading to the X cisold isomer. A proposed Scheme 21 of photophysical process is thus given:
SP
- -3sP hv
'SP'
Scheme 2 1 . Photochemical pathway for coloration of J-meth~lderivative of piperidinospiro 6-nitro-8-methoxy-benzopyran. On the basis of analysis of different cases we can formulate the following general mechanism of the spiropyran photocoloring reaction (ref. 7 ) (scheme 2 2 ) .
hv
SP -'SP'
- %P - 3x
Scheme 2 2 . General photochemical pathway for coloration of spiropyrans. This simplified scheme does not show the recombination of the excited states SP1: and Xt to give SP and also the possible transitions 1X + 3 X and 3 X 4 X and the reverse dark reaction in the ground state PM -D SP. The occurrence of the photocoloring reaction via a particular channel is determined by the structure of the spiropyran and its geometry and also external conditions.
414
Photoreactions of photomerocyanines The all trans transient merocyanines of the H-derivative of piperidinospiropyran (Amax = 5 7 0 nm) and of the 6-nitro-8-methoxy trimethyl-2,2,3 PH-chromene (Amax = 5 0 0 nm) were generated by 3 5 5 nm laser excitation in toluene solution at 298'K and were excited selectively by a 530 nm laser pulse with an appropriate time delay between the two pulses (ref. 2 5 5 ) . A photoinduced isomer is formed on excitation of both photomerocyanines with a life time of 6 p s in the former case and 4 5 0 ns in the latter. The photoinduced isomer of pigeridino photomerocyanine (Amax = 5 8 5 nm) decays back towards the initial photomerocyanine. However the photoinduced isomer of chromene photomerocyanine (Amax = 4 9 0 nm) undergoes ring closure to regenerate the chromene. These photoisomers have not been deteoted in the photocoloration reaction of piperidinospiropyran (HSP) and chromene (CHR). The observed photoreactions of the photomerocyanines previously formed proaeed most probably via a singlet mechanism. The following Schemes (23a, b) are proposed to account for the experimental results :
4.2.4
CH R
I'
Schemes 2 3 . Phototranaformations of photomerocyanines. HSP : piperidinospiropyran CHR : chromene PM : photomerocyanine I : isomer stable obtained from CHR IPM*: first excited singlet state '1%: first excited singlet state PM' : a trans stereoisomer of PM I' : a photoinduced isomer of I The Dhotodegradation The photodegradation of indoline spiropyrans has been studied thoroughly by Gautron as a function of several parameters such as the substituents (on indoline and benzopyran parts), the medium (polar or apclar solvents), the ooncentration and the temperature (refs. 2 7 , 8 4 , 2 5 6 , 2 5 7 ) , The electron-donating groups have a tendency to increase the natural polarisation of,the C - 0 bond giving a negative charge on oxygen atom. The electron-withdrawing groups have an opposite 4.3
415
effect, they decrease the polarization of the bond. On UV irradiation the opening of the benzopyran cycle by breaking of the C - 0 bond can occur by 2 processes : - heterolytic with the creation of 2 charges : positive on carbon 2 , 2 ’ and negative on oxygen 1’ - homolytic leading to a diradical. depending on the nature of the substituents. The homolytic process is iargely responsible for the degradation of the spiropyrans and so their stability to light is higher when the cleavage is mostly heterolytic. The nature of the solvents has a direct influence on the polarizability of the C - 0 bond. A study was undertaken using mixtures of protic and aprotic solvents, and shows for the solvents having a high degree of solvatation that the spiropyrans have a better stability to light compared to their stability in toluene solutions (taken as reference). The technique used for investigation of photodegradation was
the UV photolysis by repeated flashes and counting of the number of flashes necessary for reaching a decrease to the half of maximum optical density measured after the first flash. The photodegradation of indolinospiropyrans in aerated solution OCcurs partially by oxidation and the identified products are carbon monooxide, carbon dioxide, substituted derivatives of salicylaldehyde, oxindole and polycondensates. A study with 1 4 C shows the formation of formaldehyde and oxidized derivatives of the solvent. Indolinospiropyrans have been also grafted to anionic living polymers which have both flexible and rigid sequences. Initial studies were carried out on homopolymers (polystyrene, polymethylmethacrylate, polyhexylmethacrylate and polyisoprene) and subsequently on block copolymers. The photodegradation results show that the polymer has no significant effect except in the case of polyisoprene for which degradation is linked to the unsaturation of the polymer chain (ref. 258). The photodegradation of benzothiazolinospiropyrans in toluene has been studied also by repeated flash-photolysis (ref. 259). The electron-donating groups (inductive and mesomeric) decrease photodegradation, which has been examined for 6-nitro 8 -
416
methoxy compounds substituted in - 3 , - 3 ’ and - 6 ’ positions. Hammett correlations involving inductive, mesomeric, and steric parameters have been established for a series of 3substituted spiropyrans ; the steric coefficient is the most important. Concentration, temperature and nature of the solvent influence the photodegradation ; the polar solvents tend to decrease it, as had been seen with indoline compounds. The photodegradation of saturated azaheterocyclic spiropyrans (ref. 260) showed that electron-donor substituents increase as in the benzoheterocyclic series, the fatigue properties towards UV light. The introduction of oxygen or sulfur in the 5- or 6- membered saturated azacycles or the substitution with an orthofused aromatic nucleus have practically no influence on photodegradation. In conclusion, from the azaheterocyclic spiropyrans, the indoline spiropyrans allow the greatest number of photochromic cycles before falling to half the initial absorbance under well defined conditions and identical substitution. Generally the values rank from 10 to 300 cycles. For nitrogen-containing heterocyclic systems having a nitrogroup, the photophysical coloration process involves almost entirely triplet states, the “coloration“ is very good but the photodegradation also fast. The nitrogen-containing heterocyclic systems without nitro-group have a better stability to W-light but a bad “coloration“, their photophysical coloration process proceeds via a singlet “channel“. We have observed the same photophysical mechanism and the same tendency to fatigue for nitrogen-free heterocyclic systems such as benzodithiole or benzoxathiole ones whose electronic structure of photomerocyanines are mostly not polar (ref. 2 6 1 ) .
417
5. SYNTHETIC ROUTES TO SPIROPYRANS OR DERIVED COMPOUNDS
Among the great number of synthesized series, the indoline one appears as that involving the most important part of research (publications and patents) of industrial and academic interests. This is linked to economic and practical reasons but also to spectacular results obtained in connection with photocoloration. Fischer's base or 1,3,3-trimethyl-2-methylene-indoline is commercially accessible from industry ; it is a convenient starting substrate.
Nevertheless, the indoline series does not lead to all the photochromic compounds necessary for different applications [depending on spectrokinetic properties and on the contrast effect between closed form and open form (colored)]. Through fundamental investigations both in the ground states and in the excited states, every series yields interesting information which is often complementary. The principal kinds of compounds which have been recently synthesized or those which have been reinvestigated for developing such specific fundamental or applied studies will be cited. The benzothiazolines have been prepared in Marseilles in the laboratory of Metzger (refs. 28, 262-264): then in Brest in our laboratory (refs. 56, 86). The benzoxazolines (ref. 265) and the benzoselenazolines (ref. 266) have been also studied. Numerous non nitrogen-containing compounds have been synthesized in Brest: the benzodithioles, the benzoxathioles and the dithioles (ref. 92) and for emphasizing the discussion upon the original results obtained with these new spiropyran models, homolog compounds have been synthesized in the bibenzopyran or bichromene series (ref. 117). 2H-Chromenes identical models for the common benzopyran Part have been prepared (refs. 169, 210) for the first time. New N-containing five-ring (refs. 88, 89) and six-ring (refs. 88, 89) saturated series have been synthesized for under-
418 standing the influence of
aromatic
n-systems
kinetic, photochemical chromic equilibrium.
and
tions
interpretation
on
the
spectro-
structural parameters in the photo-
The results of these studies, especially those dealing with spectroscopic properties are original and allow new interpretaconcerning
the
phenomenon to be proposed.
of
the
photochromic
In the benzimidaeoline (ref, 9 1 ) , thiazoline (ref. 8 8 ) and oxazoline (ref. 85) series, only permanent merocyanines have been isolated after synthesis, irrespective of the type of substituent. Concerning the synthetic methods we can simplify and distinguish essentially two principal cases :
-
Preparation of nitrogen-containing compounds. Preparation of nitrogen-free oompounds.
Considering the general struoture of a spiropyran, we have to build the heterocyclic part H through heterocyclic bases and also the benzopyran part obtained generally from orthohydroxylated aromatic aldehydes.
5.1 General svnthetio routes for nitrogen-containing spiropsrans The way is by now classical with azaheterocycles and consists in condensing a heterocyclic immonium salt dl in basic medium, onto a salicylaldehyde 2 substituted accordingly. As intermediate, an anhydrobase 3 is formed by abstraction of a hydrogen atom from the quaternary salt after basic attack. The bases chosen are generally piperidine or triethylarnine. The solvent is often anhydrous ethanol, or benzene when the substrates in the reaction (quaternary salts, anhydrobases or aldehydes) are labile or possess other reactive functions. The isolated anhydrobase 3 can react directly with an aromatic o-hydroxylated aldehyde to give the spiropyran 9 (Scheme 24).
However the reaction does not always give univocally the spiropyran but other substrates (merocyanine 5 or tricyclic compound 5 ) as a function of substitution of heterocycle or aldehyde.
The preparation of bicondensation compounds such as dihydropyran structures 5 occurs mainly in benzofused and particularly in the benzothiazoline and indoline series.
419
R?
CHO
q w $ 2 - R 3
1
1
-
(TS.-CH 3
-
1
R
generally R3>CH3 (R3=H for indoline series)
-R@OH 2
base -R3 R3zH
R3sCH3
and R' highly electron-attractive
RR3=H and RI= electron -donating
Scheme 2 4 .
In the latter case, whatever is the substitution of the aldehyde,
according to reaction medium conditions, the synthesis may be oriented towards the formatiqn of spiropyran or towards the formation of a tricyclic spirodihydropyran compound. Nevertheless this type of non-photochromic compound has interesting properties for application in the thermal registering ; this topic will be briefly treated in the part concerning the application of spiropyrans and its derivatives. In the indoline series, the anhydrobases are generally very stable and may be easily isolated. In other cases, an anhydrobase is implicitely obtained from the quaternary immonium salt in basic medium, The most probable general mechanism of formation of spiropyrans can be formulated as it follows (Scheme 2 5 ) .
420
Scheme 25. Some examples of series are shown below.
synthesis
in different azaheterocyclic
5 . 1 . 1 Indoline series
- 1,3,3-Trimethy1-2-methy1ene-indo1ine (Fischer's base) is the "simplest building block". - The 3,3-dimethyl-l-phenyl-2-methylene-indoline may be synthesized according to the description of Briinner (ref. 267) and used also by Gautron et al. (ref. 2 6 8 ) and the Company of St Gobain (ref. 269). Condensation of diphenylhydrazine with methyl isopropyl ketone leads to the corresponding di.phenylhydrazone.
Then cyclization is achieved in presence of zinc chloride or ace'tic acid in the Fischer reaction.
421
C6H5 \
N-N=C, &5
,CH3
CH3C02Hor ZnCldEtOH
yH-CH3
under dry N2 WH3)
CH3
-
c\H3/CH3
6
@ = = C H ,
The 3,3-dimethyl-l-isopropyl 2-methylene-indoline (refs., 268, 270) is prepared from the phenylhydraaone of isopropyl methyl ketone which is cyclieed in acetic acid according to the process of Fischer (ref. 271) and leads to 2,3,3-trimethylindolenine.
The quaternary salt is formed by alkylation with isopropyl iodide (under pressure at 12O'C).
Ring-closed spiropyrans may be prepared without any substituent in the 3-position. By far,the greatest number of known photochromic spiropyrans belong to the title class, the BIPS. Their synthesis proceeds from a 1,3,3-trisubstituted 2-alkylideneindoline and a aalicylaldehyde.
CHO
\ / x@chCH-R3
YR'
+
$&OH
R
=
\ / R3
R'
422
The ease of obtaining the starting materials accounts for the preponderance of BIPS. Generally, the condensation reaction between equimolar amounts of Fischer's base and salicylaldehyde occurs rapidly in boiling alcohol and the product is obtained in 70 to 98% yield as a crystalline solid easily purified to a sharp melting point. For these reasons, a methyleneindoline and a salicylaldehyde each serves as an excellent reagent for making a derivative of the other, and we suggest their use for this purpose. Especially valuable reagents are 5-chloro- or bromo derivatives of Fischer's These are base and 6-bromo-3-methoxy-5-nitrosalicylaldehyde. highly reactive and possess additional analytically determinable elements helpful in deriving an empirical formula. The reaction serves as a convenient analytical tool ; if'a solution becomes photochromic when treated with Fischer's base, it almost certainly contains an o-hydroxyaldehyde. This was a rapid and sensitive method of following the chromatographic separation of some isomeric bromo- and nitrohydroxynaphthaldehydes (ref. 272). The literature does not report the use of o-hydroxyketones, and all attempts to obtain a 4-substituted BIPS by this means have been unsuccessful. The condensation is very probably proceeds successivdly through an aldol-type condensa-tion, a dehydration, and a ringclosure. The intermediate carbinol has neither been isolated nor indeed detected by a spectrokinetic study. strong acid 5-nitrosalicylaldehyde
When the relatively
is mixed with Fischer's base, the instant formation of a deep orange color, which then turns purple as the reaction proceeds, suggests an initial proton transfer prior to the condensation, although the orange color could possibly be attributed to (I).
423 No experiments have ever been reported in which one methylene base competes for two different salicylaldehydes, or vice versa. A study of the relative reaction rates and equilibrium compositions in such systems would be interesting.
Salicvlaldehvde exchange in BIPS Condensation is reversible, for the salicylaldehyde moiety of a BIPS may be exchanged by refluxing a BIPS solution with a different salicylaldehyde. When 3,5-dinitrosalicylaldehyde is used, the reaction is driven essentially to completion because the resulting 6,8-dinitroBIPSl which exists almost entirely in the open form, is very insoluble in most solvents (Scheme 26).
Scheme 26. This reaction serves a practical purpose. Fischer’s base will create a b l o c k i n g g r o u p for the adjacent hydroxy and aldehyde functions. These two highly reactive functions are simultaneously protected in an aromatic ring while chemical transfor-
mations are carried out on substituent X or elsewhere in the benzene ring, and yet these functions are regenerated under extremely mild conditions (except that the dinitrosalicylaldehyde is a rather strong acid). For example, 6-formyl-BIPS(X=6-CHO) was converted to 6-CN, 6-CH=NOH1 and 6-carboxy-BIPS and condensed with nitromethane and various other active methylene compounds. After an exchange with dinitrosalicylaldehyde the corresponding 5-substituted salicylaldehydes were obtained. These could not be prepared in useful yield by carrying out the reactions directly
424
on 5-formylsalicylaldehyde. This novel blocking technique should be useful to those who synthesize or degrade the complex salicylaldehyde derivatives found in nature. Both Fischer's base and dinitrosalicylaldehyde, the blocking and regenerating reagents, are readily available. The more complex 2-alkylidene, 1,3,3-trisubstituted indolines with different alkyl groups could be prepared by the Fischer synthesis, the Bischler synthesis of an indole followed by alkylation or alkylation of a 2-methylene indoline followed by a Plancher rearrangement (refs. 273-275). The route chosen will depend in part on the availability of starting materials but some researchers favor the Bischler method because of the generally better overall yields and simpler procedure. When an N-unsubstituted aniline is used in the Bischler synthesis, alkylation of the resulting 1-unsubstituted indole gives a dialkylation product, a l12,3,3-tetra-substituted indolelenium iodide in which the 1- and one of the 3-substituents come from the alkylating agent used. When a N-substituted aniline is employed, monoalkylation occurs to introduce one of the 3substituents. This is shown in the reaction sequence (Scheme 27).
ZnCtl,
NH
I
R
(R = H or alkyl)
HO
CHI
acH CH3
I R
I
R'X
(X=I )
(R" = R' if R were H;otherwise R" = R) Scheme 2 7 . The methylene indolines can be alkylated on the methylene group, the resultant indolenium salt can be converted to the cor-
425
responding base or subjected to a Plancher rearrangement to interchange the alkylidene group with a methyl group in the 3-
position.
Although this reaction sequence works well with methyl
iodide to g ve ethylidene indolines and compounds with a
3-ethyl
substituent under the same reaction condit.ions, it proceeds very poorly with ethyl and higher alkyl iodides; the major product is the HI salt of the starting material (Scheme 28).
+ RI
m
-AH-
R'X
H
3
I
-CHR'
I
R R CHZR'
m
C
H
2
I R
R CH,R'
' OH-
a
C I H
3
R
Scheme 2 8 . The indoline nitrogen atom need not necessarily bear an alkyl or aryl group. 1-Dimethylamino- and l-methoxy-3-phenyl-3methyl-2-methylene-indoline are readily obtainable (ref. 276).
A new one "pot" method for
the
synthesis
of
2-methylene-
substituted indoline is available involving a p-substituted aniline with a ketone in presence of phosphoric anhydride and an alkyl alcohol (ref. 277).
5.1.2 In the benzothiazoline series, minimal steric hindrance ( R 3 CH3) in 3-position is necessary to get a closed spiropyran structure and not a permanently stable merocyanine. The preparation of intermediary benzothiazoline bases may be done according to two general routes (refs. 28, 278, 279). - Condensation, on o-aminothiophenol in chloroform, of an acid hydrochloride judiciously substituted €or getting a benzothiazolium hydrochloride that leads to the free heterocyclic base
2
426
after neutralization with dilute aqueous sodium hydroxyde.
+NaCl + 2H20
-
Condensation on the zinc salt of o-aminothiophenol acid anhydride in CHC13
of
an
The heterocyclic bases substituted in the aromatic nucleus may be obtained directly by an electrophilic substitution with a selective activation in position 6 , or the substituent may be initially fixed in para position to heterocyclic nitrogen and the
zinc salt may be prepared chloride refs 280, 281).
by
the
Hertz
method
using
sulfur
If there is no substituent in the p-position to nitrogen, this type of reaction leads to the electrophilic chlorination in this position. The benzothiazolium quaternary salts are prepared by alkylation of bases with alkyl iodides, tosylates or methosulfates with or without solvent in a sealed tube or by refluxing.
These salts counter-anion.
are
generally
stable
having
a
tosylate as
427
5.1.3 Benzoxazoline and benzoselenazoline series
These compounds are synthesized (refs. 265, 266) according to the same scheme as that used in the benzothiazoline series. Nevertheless, the relative instability of benzoxaeolium salts
which can undergo breaking of the benzoxazole ring (ref. 266) has to be noted. In this case, it is necessary to use, for the cyclization, a strictly anhydrous and apolar solvent such as
benzene, and to eliminate progressively the water of condensation by operating with a soxhlet and a dryer like MgSOi. Concerning
the steric hindrance in 3 position, the results are sensibly identical to those found for the benzothiazoline series ( R 3 2 CH3 )
.
5.1.4 Benzimidazoline series Whatever the type
part,
in
of
substituent
in
the
benzimidazoline
the 3 position (steric hindrance) and in the aldehydic
part, the synthesis leads to permanently open-ring merocyanines (ref. 91), in contrast to results published by some Soviet authors (refs. 282, 283).
It is presumably ascribable to the basicity of the benzimidazole nucleus according to the Brooker scale of basicity, the 2-carbon atom being not electrophilic enough.
5.1.5 Five-ring and six-ring saturated azaheterocycles
-
The preparation of 42-1,3-thiazoline (X=S),
a2-1,3-oxa-
zoline (X=O), 1-pyrroline ( X = C H z ) five-ring azaheterocycles
volved known methods (refs. 88, 90).
in-
Nevertheless the preparation of selective y substituted bases has
some methods. It is the same problem for hz-1,3-thiazines ( X = S ) ; A*-lI3-oxazines X = O ) ; A2-1,4-thiazines ( X = S ) and 1-piperidelnes (X=C(CH3)2) (refs. 88, 90). led
to
adjustment
of
428
Numerous substitutions by methylalkyl or other groups are pOSs i b l e in different positions. - Alkylation is performed with alkyl iodides in smoother conditions than for the benzo-fused seriesl generally at room temperature with quantitative yields. Nevertheless the quaternary salts are labile and must be kept in vacuo and protected from light. Intermediate - - - .- - - -heterocyclic - - - - - - -bases -- The general route to sulfur-containing 1,3-thiazolines uses N acylamino alcohols as starting materials (Wenker reaction - ref. 284). .
-
-
-
r
[
p4s,0
N 5/)tHIR3
Cis or t r a n s h2-1,3-thiazolines may be prepared from (ref. 285) by a stereospecific way.
a
thiirane
A cis thiirane affords a trans 42-thiazoline.
A trans thiirane affords a dD2-thiazoline. The general route to homologous oxygen-containing heterocycles such as 113 oxazolines involves dehydration of D-hydroxyamides (ref. 2 8 6 ) .
.
- Bromohydrine may be used also by (ref. 2 8 6 ) .
reaction with
nitriles
429
erythro
-
H
An oxirane may be employed.
R=H, CH3 Two modes of preparation have been used for L \ 2 - 1 , 3 (Scheme 29) (ref, 287). a
Scheme 29.
oxazines
430
The transformation of oxazine bases into thiazines is easily achieved by sulfuration with phosphorus pentasulfide with an adequate yield (ref. 288).
mixturetcis + trans)
cis or trans
-
The 1-pyrrolines are prepared with low yields (30%) using the following reaction (Scheme 30) (refs. 289, 290).
X(CH2J,CN f RMgBr ether
F
X(Ctl,),-$-R
xylem
NMgBr
or from 2-pyrrolidone (ref. 291).
n
k Scheme 30. The preparation of 1-pyrrolines substituted in 3,4 and 5 positions is not easy (ref. 294). The 3,4,5,6-tetrahydropyridine or 1-piperidehe bases, like 1-pyrrolines, are obtained by condensation of an organo magnesium halide on an a-halogenated nitrile but this method is limited considering the difficulties of preparation of W halogenated nitriles and the poor yields of cyclic compounds (refs. 290, 292, 293).
-
431
These compounds can exist in a tautomeric equilibrium :
The I-piperideines disubstituted in position 3 present an easier access, according to the following Scheme 31.
k2
Scheme 31,
- The 5,6-dihydro 2H ._1,4-thiazine bases have been prepared
by condensation (refs. 2 9 4 - 2 9 6 ) .
of
aziridines
on a-mercapto-isopropyl ketones
H
5 . 2 By-products discovered in the spiropyran condensation
Tricyclic compounds in the indoline series The reaction of condensation between indoline base and salicylaldehyde may also give a product resulting from the condensation of two molecules of the indoline with one molecule of the salicylaldehyde. Such products are generally formed in appreciable amounts when an electron-donating substituent such as 3-methoxy is present in the salicylaldehyde, In these cases,the use of a 30 to 5 0 percent excess of the aldehyde in the reaction greatly reduces the amount of dicondensed product. These products are much less soluble in ethanol than the corresponding 5.2.1
432
spiropyran, and are easily removed by recrystallization ; they are not photochromic. These materials, except those bearing a phenyl group on the indoline nitrogen, exhibit a strong fluorescence in the 365 to 370 nm region when excited by light in the 282 to 285 nm region (ref. 2 9 7 ) . This emission is not shown by the corresponding BIPS when completely pure. The disappearance of this emission is a convenient means of following the removal of traces of dicondensation product during the purification of the BIPS. Taking into account the spectroscopic data several types of structure have been suggested for the dicondensed products in the data and indoline series (refs. 298-301), but both N.M.R.
reasonable mechanisms are in agreement with a structure as it follows, resulting of a nucleophilic attack on carbon-4. The salicylaldehyde may condense with two molecules of Fischer’s base via the intermediate carbinol, or Fischer’s base may undergo a Michael addition to the open form of the BIPS :
+- - -
Such Michael additions can in fact be carried out, when the spiropyran is heated with excess Fischer’s base, the dicondensed product is formed. The reaction occurs with numerous different BIPS. Other examples of what appear to be Michael additions to a colored form have been observed (ref. 2 6 8 ) . These reactions are reversible, in that treatment of a dicondensed product with acetic acid readily gives the BIPS and Fischer’s base ; the base may be allowed to react with additional salicylaldehyde. In practice, then, the dicondensed product does not represent totally lost material, for it may be recycled to give the desired BIPS. It is also possible to convert a diconden-
433
sation product to the BIPS purely thermally, by strongly heating it as the bulk solid or in an inert solvent ; the Fischer's base literally can be distilled from the dicondensed material. 5 . 2 . 2 Tricyclic compounds in the
benzothiazoline series The condensation of 2,3-dimethylbenzothiazolium tosylate 2 with salicylaldehyde 88, ortho-vanillin &, or 2-hydroxy-lnaphthaldehyde 8c, by refluxing in ethanol in a basic medium, does not result in the expected spiropyran (refs. 60, 302, 303) or its open form (merocyanine), and it is impossible to separate or to identify the resultant mixture. On the other hand, when the condensation is made at room temperature, the product easily obtained is a bicondensed product to & different from the spiropyran.
blcondensed
The
indexes a to h correspond to : a : R = H ; b : R = 8-OCH3 ; c.: R = 5,6-benzo ; d : R = 6-NO2 ; e : R = 6-NO2 , 8-OCH3 : f : R = 8 - N O 2 , 6-OCH3 ; g : R = 6,8-diC1 ; h = 6-Br, 8-OCH3
Numbering of the aromatic part of the final product is the same as that used for the appropriate salicylaldehyde. However, when the salicylaldehyde contains an electronwithdrawing group as in structures & to i the reaction still
a,
434
m.
gives the same corresponding merocyanine to The formation of bicondensed products has already been observed in the indoline series 2 and also with the acridine com-
pounds by Soviet authors (ref. 304). Keller (ref. 305) obtained a similar product when he reacted 2-methylene-1,3,3trimethylindoline with 7-formylbenzimidazole 13. Recently, Zwanenburg and Maas (refs. 306-308) completed the earlier studies of Livingstone (ref. 309) in the chromene series 14.
We have described some properties of bicondensed benzothiazole products in acidic medium using thermal effects and also checking their degradation in solution. A mechanism is proposed to explain their evolution. Some examples of reactivity of 2methylbenzo-heterocycloammonium salts with salicylaldehydes are given. A very attractive utilization of bicondensed compounds in a thermographic recording process is reported below.
In order to determine the conditions which influence the formation of a product (spiropyran, merocyanine, or bicondensed compound) we have tried to modify the experimental conditions, the substituent groups, or the nature of the quaternary salt or aldehyde. We noticed, as an important result, the major role of the
435
nature of the aldehyde in the benzothiazoline series. Reactions conducted with aldehydes bearing one or more electron-withdrawing groups do not lead to bicondensed products because of the higher reactivity of such aldehydes towards the anhydrobase. In the case of an insufficiently activated aldehyde function, the anhydrobase can attack the electron-deficient 4-pOSition of the methine chain obtained in the first step. Work of Livingstone (ref. 309) and Zwanenburg (refs. 306308) demonstrated that bicondensed products could be the result of an addition reaction between 2-methylene-3-methylbenzothiazoline and the open form of the spiropyran. In the present case we do not need acid catalysis as required for the chromene series. The attack of a nucleophilic agent such as Z-methylene3-methylbenzothiazoline, would be on the 4-position of the merocyanine as expected due to the low electron density on this carbon. This low electron density has been shown by 1 3 C NMR data on two merocyanines (ref.. 130) and also by CNDO/Z (ref. 151) and EHT (ref. 1 5 5 ) calculations. A deficiency of positive charge localization on the 4Position will probably induce the merocyanine formation in the benzimidazoline series. The two nitrogen atoms of the heterocyclic ring have, as a result, the positive charge localized in this part of the molecule.
The 1H NMR spectra and their interpretation are given 24 and Table 22 ) .
I
9
I
8
I
7
I
6
I
5
I 4
I
3
I
1
2 IPDrn)
6
Fig. 24. NMR-spectrum of a bis-aza-merocyanine in C D C 1 3 ,
(Fig.
436
TABLE 2 2 . Comparison of the 13C and 1H shifts (ppm/TMS, solvent CDC13) of the 3,4-dimethine bridge in benzimidazoline rnerocyanines. 13C NMR
Substituents
A8(C+C-31
OCH3-6 96.2 no substituent 98,2 CI-6 97,2 Br-6 98,l NO,-6 101,8 Br-6 OCH3-8 Br-6 H:8 Br-6 Br-8 Br-6 NO,-8
95,4 99,l
51.4 50,5 50,7 49,3 46,O
1H
NMR
6H-3
6H-4
A6IH4-H-3)
8.0
7,8 7,7 7,4 7,3 7,4
-0,2 -0,5
-l,o -l,o
-1,2
103,5
52,2 49,3 47,5 42,4
7.4 7,3 72 72
NO2-6 (p-Om)
95,3
52,6
7,4
+1,0
NO,-6 OCH3-8 NO,-6 H-8 NO,-6 Br-8 NO,-6 NO,-8
100,7 101,8
47,6 46.0
7,4 7,4
-1.1 -1,2
CI-6 H-8 CI-6 CI-8 Ci-6 NO2-8
97,2 99,8 104,l
50,7 47,8 41,6
7,4 7,4
-l,o
-
-1,2
H-6 OCH3-8 Br-6 OCH3-8 NO,-6 OCH3-8
94,l 95,4 100.7
55.0 52,2 47,6
7.8 7.4 7,4
-0,4 -0,8 -1,l
100,l
slightly soluble
-0,8
-l,o -1,3 -1.3
-
-
-
5 . 2 . 3 By-products or degradation products resulting from synthesis of five-ring and six-ring saturated azaheterocycles CH3 and meroGenerally spiropyrans are formed when R S cyanines when R3 = H like in the azaheterocyclic benzo-fused series. The reaction mechanism is the same involving an intermediary anhydrobase. The secondary reactions (ref. 8 8 ) accompanying the synthesis of five or six-ring sulfur-containing spiropyrans or merocyanines show the formation of coumarins by hydrolysis of spiropyrans and also of heterocyclo-substituted phenols (ref. 310).
>/
437
In five and six-ring oxygen-containing series, by heating the quaternary salt in basic medium, an amino ester (ref. 310) is observed, resulting from the cleavage of the ring.
CH3A@5( CH3
P'PER'DINE
or NaOH
N CIi,R" I CH3 ,O
)c
R3CH2CO~c~H2 B
CH3 CI NHCH,
I
CH3
This amino ester is in equilibrium with a cyclic hemiorthoamide considered as a tetrahedral intermediate in a mechanism of condensation with salicylaldehydes (refs. 3 1 1 - 3 1 4 ) .
8
c
CH3-C
!
A
0 CH, CH,? H C H3
CH2CH2N-C0 I I CH3 OH
CHS
438
The secondary reactions afford the following compounds : coumarins, amino-esters and iminoestera.
The reactivity of saturated azaheterocyclic spirochromenes and merocyanines towards some nucleophilic agents (HzO, H z S , N a B H a ) has been studied and the structures of the corresponding adducts have been identified (ref. 315). 5 . 3 On some reported abnormal reactions during the weparation of azaheterocyclic spirowrans Many works have been reported by Schiele et al. (refs. 316317) and Wizinger et al. (ref. 318), concerning the condensation of salicylaldehydes on an activated or not activated methyl group of a heterocyclic compound (quinaldine, lepidine, 2methylbenzothiazole, a-picoline, 2(-picoline and their N-methyl quaternary salts).
Their results are largely seems the method of choice presence of a N-methyl group, mediately indicate whether the form.
contradictory and for determining and its chemical compound is an
NMR spectroscopy the absence or shift should imopen or closed
5 . 4 General synthesis methods for non-nitrogen-containing
spiropyrans 5.4.1 Benzodithiole series
If we accept the former scheme concerning the condensation in basic medium and apply it to the 1,3-benzodithio-lylium salts (perchlorates), we find out that according to their substitution some 1,3-benzodithiolylium salts (refs. 9 2 , 319, 320) lead only to the formation of monomer or dimer anhydrobases or to both. In some cases, anhydrobases and spiropyrans are isolated (the solvent having an influence on the relative proportions).
439
It has, to be mentioned that there is competition between the nucleophilic attack of the monomer anhydrobase formed as an intermediate on the benzodithiolylium salt itself and on the salicylaldehyde (Scheme 32).
-
0
C k R3 R
J
s 0Scheme 32.
R
In face of these difficulties, we have attempted to find new and more unequivocal ways. i ) The dipolar condensation between 4,5-benzo-1,2,3-thiadiaeole (refs. 321, 322) on 2-thiocoumarins follows the Scheme 33.
440
+ S'
0
200
-
210'
c Re
Scheme 3 3 . This method is interesting for getting substituted compounds, especially in 4-positionI the access to which is not possible with ketone derivatives, Nevertheless when an electron-
attracting group was placed in any position of the coumarin nucleus (ref. 323) (especially a nitro group), that method failed. The electron-attracting groups reduce the polarizability of the C = S bond of the thiocoumarin and so the reactivity in the dipolar addition. ii) The condensation of alkyl (or aryl) Z-methyl-1,3-benzothiolylium perchlorates on salicylaldehydes in acidic medium. A general process for synthesizing benzodithiole spiropyran has been conveniently modified. This method (refs. 324, 325) involves the preparation of o-benzenedithiols, substituted or not substituted, and finally the 1 , 3 benzodithiole-ylium perchlorates. The reaction sequence is described in Scheme 3 4 . The ortho-hydroxy styryl salts ( S ) may be formed, in some cases, from corresponding anhydrobases ( A ) . This type of substrate is easily synthesized, when R 3 is an aryl group, by condensation of a o-benzenedithiol and a benzyl nitrile.
441
R"
J base
Scheme 34. The presumed mechanism considers acetic acid as a pseudobase and is described in Scheme 35. The acid-base equilibrium is possible considering the anhydrobase and acetic acid as bases having relatively similar
strengths and also considering the acidity of the benzodithiolylium salt and,that of the conjugated acid of acetic acid A c ~ as H ~quite identical. The equilibrium 1 is probably slightly displaced towards benzodithiolylium salt but as the formed compounds are consumed in the condensation reaction (sequences 2 , 3 etc. . . ) I we can expect a displacement of that equilibrium slowly The R5' substituent is chosen in the starting towards the right benzothiadiazole (Scheme 34).
.
442
Scheme 35. 5.4.2 Benzoxathiole series Regarding the former results on the benzodithiole series, it was an obvious method to prepare benzoxathiolylium salts (refs.
92, 1 1 5 ) . Very few substrates of this type are described in the literature if we exclude the 2-aryl-1,3-benzoxathiolylium perchlorates (refs. 326, 327). Benzoxathiolylium salts may be prepared from p-hydroxy-ohydroxy-thiophenol according to the Scheme 36 (refs. 328, 329).
443
Scheme 3 6 These salts are very difficult to isolate because of their high instability and their tendency to explosive decomposition.
An attempt to generalize this method failed ; nevertheless we obtain
2-benzyl-5-hydroxy-lI3-benzoxathiolylium
isolated),
perchlorates (not
which have led to o-hydroxy styryl derivatives,
then
to spiropyrans, after reaction with salicylaldehydes in ether solution and in presence of HC1 gas (ref. 115). The spiropyran condensation mechanism is quite similar to that described for benzodithiole derivatives, ether having here the role of a pseudo-base. Besides, the benzoxathiole spiropyran having a hydroxy group in 5’position may be easily alkylated or acylated in a next step (ref. 1 1 5 ) . 5 . 4 . 3 1,3-Dithiole series
The synthesis of such substrates is linked to the preparation of 1,3-dithiolylium perchlorates by classic routes using
condensation processes previously described (ref. 3 3 0 ) .
5 . 4 . 4 Synthesis of 2H-chromenes (or benzopyrans) and of spiro-
bibenzopyrans (or bichromenes)
The easy routes to interesting 2H-ctromenes as models for the benzopyran part of spiropyran molecule are limited. From coumarins by action of organomagnesium halides (refs. 3 3 1 -
-
334) if the benzene nucleus has electron-attracting (Scheme 3 7 ) .
substituents
The direct nitration under mild conditions (nitric acid in acetic acid, T = 15’C) (ref. 169) of 2H-chromenes previously prepared is possible and has enabled, recently, interesting
444
models f o r photochemical studies to be obtained. From propargyl ethers (Scheme 38).
-
Scheme 37.
Scheme 38. In our case, we can isolate, depending on the substituents R 8 , either the wanted chromene or the benzofuran compound. These results may be discussed as sigmatropic reactions (ref. 335) * The nature of R6 and Re substituents direct the reactivity of the hydroxyl group of the intermediate IE), not isolated, and consequently the concerted process leading to 2H-chromene or the attack of allene carbon leading to 2-methylfuran.
Re and
445
- The synthesis of spirobibenzopyrans has been known for
a long time (ref. 25) but the reaction processes described in the literature are not always clear and have not allowed in such cases an easy access (ref. 3 3 6 ) . These molecules have been prepared with the aim of evaluating spectrokinetic parameters and their use as models f o r direct comparison with other developed series. Their preparation has been modified (ref. 117), ,the principal reaction sequences being collected in the Scheme 39.
/
CH$OzH,HCI
-3‘
03
/
\
pyridine
Scheme 39.
9as
I
5 . 5 Synthesis of ortho hydroxylated aromatic aldehydes
Ortho-hydroxyformyl compounds are almost invariably prepared
by formylation of the corresponding hydroxy compound. An exceptionally large number of formylation methods are known (refs. 337-342). Commonly used are reactions with chloroform and base (Reimer-Tiemann), with
hydrogen
chloride
and hydrogen cyanide
(Gatterman), chloromethylation and subsequent oxidation (Sommelet), with N, N’-diphenylformamidine, with ethyl N-phenylformamidate, and conversion of the ally1 ether to the o-ally1 hydroxy compound (Claisen rearrangement) followed by isomerization to the propenyl compound and oxidative cleavage with ozone
or
osmium
tetroxide-sodium periodate.
The route chosen depends
upon the compatibility of the various reaction conditions with substituents present in the starting hydroxy compound, and the possibilities for substitution in positions other than ortho to the hydroxy group. Hydroxy
compounds containing a nitro group,
nitronaphthols,
are best formylated by reaction
and especially
with
ethyl
N-
phenylformamidate. This reagent has always given better yields of the aldehyde anil than has done the more commonly used N, N’diphenylformamidine. The extreme simplicity of the procedure makes it a very convenient alternative to the Reimer-Tiemann route even with simple hydroxy compounds. The resulting anil may be used directly for spiropyran formation, if necessary. The use of a nitroso instead of the formyl group should give This procedure is the spiro- ( 1 , 4-oxazines) (chapters 10 and 2 4 ) . reasonably
successful
with nitrosonaphthols but gives very poor
results whzn nitrosophenols are used. In the latter case, one can reverse the functions and condense an aminophenol with an oximinomethyl quaternary heterocyclic salt .to give the desired oxazine. This is another example where the formyl group is as a derivative. 5.6
Synthesis of thio analogs of spiropyrans
used
- The thio BIPS can be prepared from 0-mercapto benzaldehyde, which is tedious to obtain. It is more convenient to convert a BIPS into its thio analog merely by treatment with phosphorus pentasulfide in pyridine or xylene. Nevertheless the product is difficult to purify, especially when nitro groups are present in the molecule.
447
Beclter and Kolc (ref. 3 4 2 ) have examined the photochromism of several thiobenzopyran derivatives and observed a bathochromic effect on t h e absorption wavelength of the colored form. - Novel photochromic indolinospirobenzothiopyrans having a nitrogroup in 6-position were prepared (refs. 3 4 3 - 3 4 6 ) and their properties in polymer films were examined. The absorption bands
of the colored lie around 100 nm deeper in the long-wave region of the visible spectrum than is the case with the common spiropyrans. The necessary 5-nitrothiosalicyladehydes were prepared by the following two methods : ( A ) : Conversion of a halogen atom of 0-chlorobenzaldehyde into an SH group (ref. 3 4 7 ) (Scheme 40) SH
CHo N32
--
1) N a S 2
+
2) NaOH
NO2
Scheme 4 0 . (B) : Conversion of an OH group of salicylaldehyde into an SH group (ref. 3 4 8 )
SH i02
NO2
NaH or
Lib Scheme 4 1
N02
5 . 7 . Synthesis of bifunctional benzothiazolino--and indolino-
spiropyrans Different types of symmetrical bifunctional indolino- and benzothialino-spiropyrans have been prepared, the two photochromic entities may be linked on the “left“ heterocycle part or on the “right“ benzopyran part. Some examples of such bifunctional structures are given below : 5 . 7 . 1 Bis-indolino-spirobenzopyrans and Bis-benzothiazolinospirobenzovyrans linked at the nitrogen atom by a paraffin chain having 2 to 12 carbon atoms. The general scheme of synthesis involves the preparation of a bis-quaternary salt (refs. 349-351) (Scheme 4 2 ) . H,C
CH,
v
CHO
Scheme 4 2 .
449 CHa
Scheme 42 (continued)
CHO I ,OH
pipcridine
__c
O?N @\OCHa
EtOH
5.7.2 B i s - i n d o l i n o - s p i r o b e n z o p y r a n s and Bis-benzothiazolinospirobenzopyrans linked at the G'position by means of a carbon bond or a , u diamide paraffin chain. - The general structures are the following (refs. 352-353).
X = H, OCH,
X = 1-1, OCH,
- The synthesis of diamide compounds scheme 43 (ref. 3 5 1 ) . [m = 4 and 81
'-
C,H, 4- CI
,NH-CO
fCH,+
"C
H,C,
!':a'
-
c
\ /
L,I
s-quaternary s n l t +
Scheme 43.
I
I
is
described
on
- C O f CHl k C 0 - CI
CO-NH,
in
salicylaldehyde
t
piperidine
the
450
Scheme 4 3 (continued)
The bleaching kinetic constants are near enough to those of the corresponding monofunctional compounds (ref. 3 5 1 ) 5.7.3 Bis-indolino-spirobenzopyrans nucleus of indoline heterocycle
mp =195-210'
:hax =
having
598 nm (toluene)
the
common
benzene
(ref. 354)
5.7.4 Bis-indolino-spirobenzopyrans linked on the benzopyran wart or having a "common" benzopyran nucleus. The following structures have been synthesized by Gautron (refs. 352, 353) or the N.C.R. (ref. 354)
(ref. 354)
X = H,CI
(refs. 352, 353)
451
R
R
CH, CH3 CH3 CbH5
I R
X
H
CI
(refs. 3 5 2 , 3 5 3 )
Br H
. H,C-N
(refs. 3 5 2 , 353)
352, 353)
d Me
Generally the purification of such compounds is tedious ; recrystallization is not easy and the silicagel column chromatography provides better results (ref. 3 5 1 ) .
452 6 . CONCLUDING REMARKS
In the field of photochromic spiropyrans, many fundamental studies have been carried out by academic and industrial teams and a lot of appplications have been realized. The most important fundamental results were harvested from 1960 to 1985. From 1971 (year of printing of the Bertelson’s review) until now ( 1 9 8 9 ) , numerous complementary investigations have led to a deeper insight into the structure, the spectroscopic properties, the kinetics, the fatigue behavior and the photophysical process of coloration of these systems. In addition a number of new syntheses have considerably extended the field.
on photochromism were conducted on the BIPS structure because of its ready synthetic availability and interesting properties of colorability and fatigue (as compared to similar series), but any heterocyclic series confers its own particularity to the structure and the photochemical coloration as well as to the photodegradation. A classification of the heterocyole series has been made, acoording to the type of reactivity and the spectrokinetic behavior in different solvents. The correlations between molecular structure or substituents and the spectrokinetics of the spiropyrans photoconversion, have allowed a better understanding of the role of different structural parameters (and also solvents) on the electronic distribution of the photomerocyanine ; therefore, the choice of a definite strucspecific properties is made easier. ture for reaching Nevertheless, it is often difficult to generalize such correlations. A great number of studies
The use of a combination of various techniques (spectrometry, dipole moments measurements, quantum calculations, evaluation of non-bonding interactions, conformational approach, coordination chemistry by metal salts...j has provided very important data for structural determinations.
453
In Bertelson's review (ref. 2 5 ) a lot of pertinent questions were asked about the mechanism of the photocloration process : nature of the chromophore, energy of absorbed photons, cleavage of the C - 0 bond, nature of the excited states involved, repartition of the excitation energy among light emission, photocoloration and photodegradation. Now, the majority of these questions have found an answer (although sometimes partially). We want to insist upon the fundamental role of the "right" benzopyran chromophore and the electronic or steric assistance of the "left" heterocycle part in the interconversion spiropyran-;rphotomerocyanine, for instance the importance of the geometry of the heterocycle part is illustrated in the behavior of the piperidinospirobenzopyran series.
The applications are varied in solution and in polymer materials ; the number of Patents is large. Nevertheless, the photodegradation is a limitation to some applications, especially when a large number of cycles is needed, e.g. for optical variable transmission materials. Many applications involving the solid state, spiropyrans evaporated (ref. 3 5 5 ) or linked to particular materials with photoconductive, semi-conductive, optical... properties, or to biological molecules or macromolecules deserve further investigations. A variety of substituents either on the heterocycle part or the benzopyran half, may be fixed on the spiropyran skeleton (for instance carboxy, carbomethoxy (ref. 3 5 6 ) or phenylazo (refs. 3 5 7 - 3 5 9 ) groups). The practical value of photochromic materials motivates a continuous search for new classes of substances related to spiropyrans. New photochromic molecules can be discovered. Recently, spiro-compounds such as spirooxazines (cf chapters 10 and 24) which present a good behavior to the U.V. light, have given a new impetue to research in this fascinating field.
.
454
ACKNOWLEDGEMENT
I am indebted to my supervisors P r o f s . J. Metzger and J.E. Duboia as well as to my main collaborators : Drs. F. Gamier, A. Samat, P. Appriou, M. Le Baccon, M. Maguet, 0 . Petillon, A. Kellmann, L. Lindqvist, F. Tfibel, J. Aubard, Y. Poirier, E. Miler-Srenger, M. Mosse, A. Botrel, A . Lebeuze, E. Davin, J. Kister, H. Pommier, C . Riou and P r o f . J. Robillard. I also gratefully acknowledge the financial support of different organisms, especially D.R.E.T., D.G.R.S.T., C.N.R.S., and Company Rh6ne-Poulenc La Cellophane,
-
455
REFERENCES 1 2 3
4 5
E. Fischer and Y. Hirshberg, J. Chem. SOC., (1952) 4522. 0. Chaude and P. Rumpf, C.R. Acad. Sci., 236 (1953) 697. Y. Hirshberg, Bull. Res. Council Israel, 5A (1956) 188. Y. Hirshberg, J. Am. Chem. Soc., 78 (1956) 2304. Industrial groups involved in photochromism : 1. National Cash Register Company, Polaroid Corporation, Xerox Corporation, American Cyanamide, Eastman Kodak(USA). 2 . Telefunken A.G., Kalle A.G., (West Germany)... 3. Fuji Photofilm, Toray Industries (Japan)... 4. Agfa Gevaert Company (Belgium).. 5. Compagnie St Gobain (France)... Academic laboratories having contributed largely to the development of the domain : The Weizmann Institute, Israel (H. Hirshberg, E. Fischer, R. Heiligman-Rim, T. Bercovici, ; The University of Houston USA (R.S. V. Krongauz...) Becker, J.G. Calvert...) ; The University of Leuven, Belgium (C. Smets...) ; The University of Paris, France (P. Douzau, C. Balny, 0. Chaude) ; The university of Strasbourg, France (C. Wippler) ; The Universities of Marseilles and Brest, France (J. Metzger, R. Guglielmetti) in collaboration with F. Garnier and J.E. Dubois (Paris) ; The team of Prof. E. Inoue, Tokyo, Japan ; and besides important teams of Soviet researchers (M.V. Sarostyanova, K.G. Dzhaparidze, A.V. Shablya, 13.1. Lashkov, V.I. Minkin, E.R. Zakhs, L.S. Efros, A . S . Kholmanskii, V.A. Kuzmin, V.A. Barachevsky...). A.S. Kholmanskii and V.M. Dymuaev, Russ. Chem Rev., 56 (1987) 136. Abstracts of the 14th Congress on High Speed Photography and Photonics. Moscow (1980). E. Fischer, Chemie Unserer. Zeit., 9 (1975) 85 and cited ref. M.E. Musser and S.C. Dahlberg, J. Chem. Phys., 72 (1980) 4084, T. Moriizumi and K. Kudo, Appl. Phys. Lett., 38 (1981) 85 K. Kudo and T. Moriizumi, Japan J. Appl. Phys., 20 (1981)
..
.
6
7
8
9
10
11
12 13 14 15 16 17 18 19 20 21
22
23 24
1553.
D.G. Milles Jr, P.D. Lamb, K.W. Rhee and C.S. Johnson Jr, J. Phys. Chem., 87 (1983) 4815. A. Mustafa, Chem. Rev., 43 (1948) 509. G.H. Brown and W.G. Shaw, Rev. Pure. Appl. Chem.,ll (1961) 2 P. Doueou and C. Wippler, J. Chim. Phys., 60 (1963) 1409. J.H. Day, Chem. Rev., 63 (1963) 65. A. Bertrand, Rev. Inst. Fr. Petrole, 21 (1966) 1009. R. Dessauer and J.P. Paris, "Photochromism" in W.A. Noyes Jr, G.S. Hammond and J.N. Pitts Jr, Ed'. Advances in Photochemistry vol. 1 Interscience, New York (1963) pp. 2882911 314-315. R. Exelby and R. Grintkr, Chem. Rev., 65 (1965) 247. M.V. Savostyanova, Opt. Mekh. Prom., 33(5) (1966) 31. E. Fischer, Fortschr. Chem. Forsch., 7 (1967) 605. L.D. Weis, T.R. Evans and P.H. Leermaker, J. Am. Chem. S O C . , 90 (1968) 6115. S. Daehne, Z. Wiss. Photogr. Photophys. Photochem., 62 (1969) 183. J. Vitry, Chim. and Ind., 102 (1969) 1333. J. Peyches, Chim. and Ind., 103 (1970) 2611.
456 25
26 27 28 29
30 31
32 33
34 35
36
37 38 39
40
41 42 43
44 45 46 47 48 49
50
51 52 53
54 55
R.C. Bertelson, "Photochromism". Edit. G.H. Brown chap. 111, J. Wiley and Sons Inc. New York (1971). J.C. Besnard, Thesis, University of Strasbourg (1966). R. Gautron, Thesis, .University of Strasbourg (1967)R. Guglielmetti, Thesis, University of Marseilles (1967). T. Bercovici, R. Heiligman-Rim and E. Fischer. Mol. Photochem., 1, (1969) 23. L.D. Taylor, (Polarofd Corporation) U.S. Patent 3,320,067 (1967). National Cash Register Laboratories Dayton (USA) results unpublished. R.E. Fox, "Research Reports as Test Items pertaining to Eye Protection of Air Crew personnel" Final Rept. on contract AF41 (657) 215 - April (1961) - AD440226. I. Shimidzu, H. Kokado and E. Inoue, Kogyo Kagaku Zasshi, 72 (1969) 171. I. Shimidzu, H. Kokado and E. Inoue, Bull. Chem Soc. Japan, 42 (1969) 1730. I. Shimidzu, H. Kokado and E. Inoue, Nippon Kagaku Zasshi, 88 (1967) 1127. I. Shimidzu, H. Kokado and B. Inoue, Nippon Kagaku Zasshi, 89 (1968) 755. R.E. Bowman, R.E. Fox, A.J. Kresch and E.G. Taylor, "Research and Reports on photochromics which may be used at Eye Protective Devices" April 1963 - AD 427601. 0. Chaude, Cahiers Phys. (France), 52 (1954) 3,39. 0. Chaudb, Cahiers Phys. (France), 51 (1953) 6. 0. Chaude, Cahiers Phys. (France), 50 (1952) 17. J. Arnaud, C. Wippler and F. Beaure d'Augeres, J. Chim. Phys., 64 (1967) 1165. P.H. Wandewijer and G. Smets. J. Polymer. Sci. Part.C 22 (1968) 231. J. Arnaud, M. Niclause and C. Wippler. J. Chim. Phys., 65 (1968) 2150. 0.1. Lashkov and A.V. Shablya, Opt. Specktrosk, 19 (1965) 821. R. Heiligman-Rim, Y. Hirschberg and E. Fischer, J. Phys. Chem., 66 (1962) 2465. R. Heiligman-Rim, Y. Hirshberg and E. Fischer, J. Phys. Chem., 66 (1962) 2470. Y. Hirshberg and E. Fischer, J. Chem. SOC., (1954) 3129. G.I. Lashkov, V.I. Brmolaev and A.V. Shablya, Opt. Spektrosk, 21 (1966) 546. Z.G. Gardlund, Polymer. Lett., 6 (1968) 57. J.C. Metras, M. Mosse and C. Wippler, J. Chim. Phys., 62 (1965) 659. M.W. Windsor, R . S . Moore and J.R. Novak, Spectrochim. Acta, 18 (1962) 1364. V.G. Dzhaparidee, E.G. Akhalkatsi and L.P. Shishkin, Soobshch Akad. Nauk Gruz S.S.R., 51 (1968) 323. T.A. Shakhverdov, Teor. Eksp. Khim, 4 (1968) 421. T.A. Shakhverdov and A.V. Shablya, Opt. Spektrosk, 24 (1968) 619. G.I. Lashkov and A.V. Shablya, lev. Akad. Nauk SSR. Ser. Fiz., 32 (1968) 1569. G.I. Laskhov, A.V. Shablya and D.N. Glebovskii, Opt. Spectrosk, 20 (1966) 176. R. Guglielmetti, J. Photog. Science, 22 (1979) 77.
-
457 56 57
58 59
60 61
62 63 64 65
66 67
68 69 70
71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93
(a) A. Samat, J. Metzger, F. Mentienne, F. Garnier, J.E. Dubois and R. Guglielmetti, J. Phys. Chem., 76 (1972) 3554. (b) F. Mentienne, A. Samat, R. Guglielmetti, F. Garnier, J.E. Dubois and J. Metzger. J. Chim. Phys., 70 (1973) 5440 J.A. Miller, P. Levoir, L.C. Fontaine, F. Garnier and J.E. Dubois, J. Anal. Chem., 47 (1) (1975) 29. D.H. McQuain, A.J. Kresh, R.E. Fox and R.E. Bowman, "Effort to evolve a method of Eye Protection from Flashblindness", Final Reports on Contract AF 41 (609)-2202 (Dec. 1964) A.D. 464983. R. Guglielmetti, M. Mossel J.C. Metras and J. Metzger, J. Chim. Phys., 65 (1968) 454. R, Guglielmetti and J. Metzger, Bull. SOC. Chim. Fr. (1967) 2824. N.W. Tyer and R.S. Becker, J. Am. Chem. Soc., 22 (1971) 1289, 1295. R. Heiligman-Rim, Y. Hirshberg and E. Fischer, J. Chem. S O C . (1961) 156. E. Fischer, J. Phys. Chem., 71 (1967) 3704. J. Blanc and D.L. ROSS, J. Phys. Chem., 72 (1968) 2817. C. Schiele, M. Ruch and D. Hendriks, Tetrahedron, 23 (1967) 3733. C. Schiele and D. Staudacker, Tetrahedron, 24 (1968) 471. C. Schiele, D. Staudacker, D. Hendriks and G. Arnold, Tetrahedron, 24 (1968) 5017. C. Schiele, D. Staudacker, K. Halfar and G. Arnold, Tetrahedron, 24 (1968) 5023. C. Schiele, A. Wilhelm, D. Hendricks, M. Stepec and G. Paal, Tetrahedron, 24 (1968) 5029. C. Schiele, G. Paal, 0. Arnold and M. Stepec, Tetrahedron, 24 (1968) 5037. C. Schiele and Stepec, Ann. Chem., 725 (1969) 226. C, Schiele and 0. Arnold, Z. Naturforsch, 23b (1968) 628. Y. Hirshberg and E. Fischer, J. Chem. Phys., 22 (1954) 572. E.B. Knott, J. Chem. SOC. (1951) 3038. L.G.S. Brooker et al., J. Am. Chem. SOC., 73 (1951) 5350. L.G.S. Brooker et al., J. Am. Chem. Soc., 87 (1965) 2443. K. Dimroth, Chimia, 15 (1961) 80. P. Suppan, J. Chem. SOC. (A) (1968) 3225. J.B. Flannery, J. Am. Chem. Soc,, 90 (1968) 5660. S. Brownstein, Canad. J. Chem., 38 (1960) 1590. E.M. Kosower, J. Am. Chem. S O C . , 80 (1958) 3253. 1.A. Zhmyreva, V.V. Zelinski, V.P. Kolobhov and N.D. Krasnitskaya, Dokl. Akad. Nauk SSR, 129 (1959) 1089. E. Berman, R.E. Fox and F.D. Thomson, J. Am. Chem. SOC., 81 (1959) 5605. R. Gautron, Bull. S O C . Chim. Fr. (1968) 3190. J. Kister, H.P. Pommier, J. Metzger and R. Guglielmetti, Can. J. Chem., 62 (19841, 2372. M. Le Baccon, Thesis, University of Brest (1978). A. Samat, Thesis, University of Brest (1976). M. Maguet, Thesis, University of Brest (1980). M. Maguet, F. Garnier and R. Guglielmetti, J. Chem. Res., S145 (1982) M 1519. G. Petillon, Thesis, University of Brest (1979). J.C. Le DUC, Thesis, University of Brest (1975). P o Appriou, Thesis, University of Brest (1977). A - Samat, R. Guglielmetti, Y. Ferre, H. Pommier and J. Metzger, J. Chim. Phys., 69 (1972) 1202.
458 94 95
96 97 98 99 100 101 102 103
104
Y.
Yukawa and Y.
965, 971.
Tsuno, Bull.
H.C. Brown and Y. Okamoto, J. 4979, 79 (1957) 1913.
Chem. SOC. Jap., 32 (1959
Am.
Chem.
SOC., 80 (1958
M. Charton, J. Am. Chem. SOC., 91 (1969) 615. M. Charton, J. Am. Chem. SOC., 91 (1969) 619, 624.
M. Charton and B.I. Charton, J. Org. Chem., 36 (1971) 260. D.H. McDaniel and H.C. Brown, J. Org. Chem., 23 (1958) 420. M , Charton, J. Org. Chem., 29 (1964) 1222. A. Biondi, J . Phys. Chem., 68 (1964) 441. M. Charton, private communication. P.H. Vandewijer, J . Hoefnagels and G. Smets, Tetrahedron, 25 (1969) 3251.
M.
Maguet, R. Guglielmetti and F. Garnier, New J. Chim., 6
(1982) 207.
105 106
M. Moss4 and J.C. Metras, J. Chim. Phys., 64 (1967) 691.
107 108
C. Milhaud, CNAM Thesis, University of Paris (1969). V.A. Krongauz and A.A. Parshutkin, Photochem. Photobiol., 15
109 110 111 112 113
A.Yu.
Kryukov and A.V.
(1978) 418.
Vannikov.,
High Energy
Chem.,
12
(1972) 503.
A.A. Parshutkin and V.A. Krongauz, Mol. Photochem., 6 (1974) 437.
V.A. Krongauz and E.S. Golburt, Nature, 271 (1978) 43. V.A. Krongauz, E.S. Golburt and S.N. Fishmann, J. Photochem., 9 (1978) 298. V.A. Krongauz, S.N. Fischmann and B . S . Goldburt, J. Phys. Chem., 82 (1978) 2469. V.A. Krongauz and E.S. Goldburt, Chem. Phys. Lett., 60 (1979) 251.
114
M. Maguet, R.
119
A. Le Beuze, A. Samat, A. Botrel, P. Appriou and R. Guglielmetti, J. Chim. Phys., 75 (1978) 46. (a) A. Le Beuze, A. Botrel, A. Samat, P. Appriou and R. Guglielmetti, J. Chim. Phys., 75 (1978) 255, 267. (b) M. Kryszewski and B. Nadolski, Adv. Mol. Rel. Processes,
Guglielmetti, F. Garnier, D. Grand and E. Amouyal, New J. Chim., 6 (1982) 215. 115 P. Appriou, J. Guillerez. F. Garnier and R. Guglielmetti, Helv. Chim. Acta, 58 (1975) 2553. 116 P. Appriou, F. Garnier and R. Guglielmetti, Helv. Chim. Acta, 58 (1975) 2563. 117 P. Appriou, C. Trebaul, J. Brelivet, F. Garnier and R. Guglielmetti, Bull. SOC. Chim. Fr,, (1976) 2039. 118 P. Appriou, F. Garnier and R. Guglielmetti, C.R. Acad. Sci., 120 121
279C (1974) 973.
5 (1973) 115.
J.G. Calvert and J.N. Pitts Jr., Photochemistry, J. Wiley Edit. New York (1966) p.487. 122 H. Bach and J.G. Calvert, J. Am. Chem. SOC., 92 (1970) 2608. 123 A.V. Shablya, K.D. Demidov and Yu N. Polyakov, Opt. Spek bosk, 20 (1966) 738. 124 A. Samat, J. Kister, R. Guglielmetti, F. Garnier and J. Metzger, Bull. SOC. Chim. Fr. (1975) 2627. 125 J. Arnaud, J. Chim. Phys. (1968) 159. 126 R. Guglielmetti, R. Meyer and.C. Dupuy, J. Chem. Educ., 50 127
(1973) 413.
Guglielmetti, E.J. Vincent and J. Chim. Fr., (1967) 4195. R.
Metzger,
Bull.
SOC.
459 128 129
130 131
132 133 134 135
136
137 138 139
140
141
142
143 144 145
146
147 148 149 150
151
152 153
A.
Samat, G. Martin and R. Guglielmetti, C.R. Acad. Sci.1 279c (1974) 573. K.G. Dzhaparidze, Y.Ya. Pavlenishvili, V.G. Tsitsishvili and D.P. Maisuradze, Bull. Acad. Sci. Georgian S.S.R., 70 (1973) 349, A.M. Samat, G.J, Martin and R.J, Guglielmetti, Org. Magn. Reson., 8 (1976) 62. P.Y. Appriou, A.M. Samat, R.J. Guglielmetti and J.Y. Le Gall, Org. Magn. Reson., 10 (1977) 39. J.P. Besse, A.M. Samat, J.Y. Le Gall, A.J. Le Beuze and R.J. Guglielmetti, Org. Magn. Reson., 1 1 (1978) 493. M. Maguet, Y. Poirier and R. Guglielmetti, Can. J. Chem., 60(21) (1982) 2635. M. Maguet, M. Le Baccon, Y. Poirier and R. Guglielmetti, Can. J. Chem., 60(21) (1982) 2644. R.P. Shibaeva and R.M. Lobkorskaya, Zh. Strukt. Khim., 20 (1979) 369. S.M. Aldoshin. Doctoral Thesis in Chemical Sciences, Moscow (1986). B.Ya. Simkin, S.P. Makarov, M.G. Furmanova, K.M. Karaev and V.1, Minkin, Khim. Geterotsikl. Soedin., (1984) 747. S.M. Aldoshin, A.S. Kholmanskii and L.O. Atovmyan, IZV. Akad. Nauk SSSR Ser. Khim., (1986) 576. (a) K.M. Karaev, N.G. Furmanova and N.V. Belov, Dobl. Akad. Nauk SSSR, 253 (1980) 338 ; 260 (1981) 1122. ( b ) K.M. Karaev and N.G. Furmanova, Zh. Strukt. Khim., 25 (1984) 185. S.M. Aldoshin, L.O. Atovmyan, O.A. D’Yachenko and M.A. Gal’Bershtam, Izv. Akad. Nauk SSSR, 12 (1981) 2720. R.P. Shibaeva, L.P. Rozenberg, A.S. Kholmanskii and A.V. Zukov, Dokl. Akad. Nauk SSSR, 226(6) (1976) 1374. (a) E. Miler-Srenger and R. Guglielmetti, Acta Crystallogr., C40 (1984) 2050. ( b ) R. Bucourt and D. Hainaut, Bull. SOC. Chim. Fr., (1965) 1366. (c) J.C.A. Boeyens, J. Cryst. Mol. Struct., 8 (1978) 317. E. Miler-Srenger and R. Guglielmetti, Acta Crystallogr., B38 (1982) 1980. E. Miler-Srenger and R. Guglielmetti, J. Chem, Soc. Perkin Trans. 11, (1987) 1413. F.. Przystal, T. Rudolph and J.P. Phillips, Anal. Chim. Acta, 41 (1968) 391. J.P. Phillips, A . Mueller and F. Przystal, J. Am. Chem. SOC. (1965) 4020. L.D. Taylor, J, Nicholson and R.B. Davis, Tetrahedron Lett. (1967) 1585. L.D. Taylor, Polaroid Corporation, U.S. Patent 3,299,079 (1967) c. Ceintrey, R. Guglielmetti, and M. Le Baccon, (Rhone La Cellophane) Fr. Patent Appl. N‘77.30027 (1977). Poulenc M. Le Baccon and R. Guglielmetti, J. Chem. Res., (1979) (S) 154, ( M ) 1801. H. Pommier, A . Samat, J. Metzger and R. Guglielmetti, J. Chim. Phys., 5 (1975) 589. L. Edwards, J. Kolc and R.S. Becker, Photochem. Photobiol., 13 (1971) 423. B.Y. Simkin, V . L . Minkin, and L.Y. Nivoroskin, Khim. Geterotsikl. Soedin., 1 (1976) 76. (CA 1974 80 107756 e).
.
-
460
154 155
156 157 158 159
160 161
162 163
164 165 166 167 168 169 170
171 172 173
174
175 176
177 178 179
180 181
182 183 184 185
186 187 188 189
190
Le Baccon and R. Guglielmetti, New J. Chem., 12 (1988) 825. A. Le Beuze, A. Botrel, A. Samat and R. Guglielmetti, J. Mol. Struct., 40 (1977) 77. G. Hedestrand, 2. Phys. Chem., B2 (1929) 428. E.A. Guggenheim, Trans. Faraday SOC., 45 (1949) 714. I.F. Halverstadt and W.D. Kumler, J. Am. Chem. Soc., 64 (1942) 2988. W.R. Gilkerson and K.K. Srivastava, J. Phys. Chem., 64 (1960) 1485. E.G. MacRae, J. Phys. Chem., 61 (1957) 562. A. Kawski, Acta Physica Polonica, 25 (1964) 285 and 29 (1966) 187. L. Bilot and A. Kawski, 2 . Naturforsch., 17a (1962) 621. J,L. Derocque, P. Appriou, R . Guglielmetti and H. Durr, Org. Mass. Spectrom., 12 (1977) 606. G. Arnold, 2. Naturforsch., 21b (1966) 291. G. Arnold and 0. Paal, Tetrahedron, 27 (1971) 1699. G. Arnold and C. Schiele, Z. Naturforsch., 22b (1967) 1228. 0. Arnold and G. Paal, Tetrahedron, 25 (1969) 5995. J.C. Panizzi, Thesis, University of Marseilles (1967)E. Davin, M. Guiliano, 0 , Mille, J. Chouteau and R . Guglielmetti, Helv. Chim. Acta, 60 (1977) 215. M. Guiliano, E. Davin-Pretelli, G. Mille, J. Chouteau and R. Guglielmetti, Helv. Chim. Acta, 61 (1978) 1072. C. Schiele and G. Arnold, Tetrahedron Lett., 13 (1967) 1191. K.G. Dzhaparidze, I.Ya. Pavlenishvili, M.T. Gugava and D.P. Maisuradze, 2. Fiz. Khim., 44 (1970) 582. V.G. Dzhaparidee, I. Ya. Pavleniavili, M.T. Gugava and D.P. Maisuradze, Soobshch. Akad. Nauk. SSR., 56 (1969) 577. M. Guiliano, M. Maguet, G. Mille and R. Guglielmetti, C.R. Acad. Sci., t 282 B (1976) 113. M. Guiliano, G. Mille, J. Chouteau, J. Kister and J. Metzger, Anal. Chim. Acta, 93 (1977) 33. J. Kister, E. Davin, M. Guiliano, G. Mille and J. Chouteau, Analusis, 1 1 (1983) 284. E. Davin, M. Guiliano, H. Reymond, G. Mille and J. Kister, Anal. Chim. Acta, 148 (1983) 271. A. Samat, R. Guglielmetti, Y. Perre, H. Pommier and J. Metzger, J. Chim. Phys., 70 (1973) 544. Y. Ferre, A. Samat, R. Guglielmetti, E.J. Vincent and J. Metzger, Tetrahedron, 30 (1974) 787. P. Geneste and G . Lamaty, Bull. SOC. Chim. (1967) 4456. N.L. Allinger, J.A. Hirsch, M.A. Miller and L.J. Tyminski, J. Am. Chem. SOC., 91 (1969) 338. T.L. Hill, J. Chem. Phys., 16 (1948) 399. R. Hoffmann, J. Chem. Phys., 39 (1963) 1397. J.A. Pople, Trans. Faraday SOC., 49 (1953) 1375. R. Pariser and R.G. Parr, J. Chem. Phys., 21 (1953) 466, 767. R.G. Parr and R . Pariser, J. Chem. Phys., 23 (1955) 711. J. Del Bene and H.H. Jaffe, J. Chem. Phys., 48 (1968) 1807. B. Tinland, R. Guglielmetti and 0. Chalvet, Tetrahedron, 29 (1973) 665. 0. Bloch-ChaudC and J.L. Masse, Bull. SOC. Chim. Fr. (1955) 625. L. Edwards, J. Kolo and R.S. Becker, Photochem. Photobiol., 13 (1971) 423.
M.
461
191
192
193 194
195 196
197 198 199
200 20 1
202 203
204 205
206
207 208
209
210
211
212 213 214 215 216 217
218 219
220
B.Ya, Simkin, V.I. Minkin and L.Y. Nivozoskin, Khim. Geterotsikl. Soedin. (1974) 76. A. Le Beuze, A. Botrel, P. Appriou and R. Guglielmetti, Tetrahedron, 35 (1979) 31. M. Delamar, J. Aubard, J.L. Albert and J.E. Dubois, J. Electron. Spectrosc. Rel. Phenom., 28 (1983) 289. M. Delamar, J.L. Albert, J. Aubard and J.E. Dubois, J. Electron. Spectrosc. Rel. Phenom., 32 (1983) 351. A.S. Kholmanskii, A.V. Zubkov and K.M. Dyumaev, Uspekhi. Khim., 50 (1981) 569 [Russ. Chem. Rev., (4) (1981). C. Mc.Ardle, €I.Blair, A. Barrand and A. Randel-Teixier, Thin Solid Films, 99 (1983) 181. J. Albert and J . Aubard, J. Raman Spectrosc., 14 (1983) 83. A.V. El'Tsov "Editor", Organichekie Fotokhromy (Organic Photochromic Compounds), Izd. Khimiya Leningrad (1982). V.M. Lisgutenko and V.A. Barachevskii, Teor. Eksper. Khim., 17 (1981) 485. H.E. Simmons and T. Fukunaga, J. Am. Chem. SOC., 89 (1967) 5208, V.M. Komarov and V.G. Plotnikov, Teor. Eksper. Khim., 10 (1974) 62. P. Pasman, J. Verhoeven and T. Devoer, Tetrahedron, 32 (1976) 2827. A.S. Kholmanskii and V.M. Dyumaev, Dokl. Akal. Nauk. SSSR, 261 (1981) 704. A.S. Kholmanskii, N.I. Kunavin, A.V. Zubkov and K.M. Dyumaev, Zh. Fiz. Khim., 56 (1982) 2794. [Russ. J. Phys. Chem., (11) (1982)l. C. Balny, R. Guglielmetti, M. Moss6 and J. Metzger, Photochem. Photobiol., 16 (1972) 69. A.S. Dvornikov, YaiN. Malkin and V.A. Kuz'Min, Izv. Akad. Nauk SSSR, Ser. Khim., 390 (1982) 1520. A.S. Dvornikov, Ya.N. Malkin, V.V. Mezheretskii, O.N. Zhukovskaya, Yu.V. Revinski and V.A. Kuz'Min, Izv. Akad. Nauk SSSR, Ser. Khim., 390 (1982) 2014. E. Davin-Pretelli, Thesis (3' cycle), University of Marseilles (1975). E. Davin, C. Balny and R. Guglielmetti, C.R. Acad. Sci., 275C (1972) 79, P. Appriou, J. Brelivet, C. Trebaul and R. Guglielmetti, J. Photochem., 6 (1976/77) 47. P. Appriou, A. Botrel, A, Le Beuze and R. Guglielmetti, J . Photochem., 8 (1978) 73. J.W. Sidman, Chem. Rev., 58 (1958) 689. D . J . Cowley, J. Chem. SOC. Perkin I1 (1975) 287, 1576. 1.L. Belaits and T.D. Platonova, Opt, Spektrosk, 35 (1973) 218. K.G. Dzhaparidze "Spirochromenes" Izd. Metsniereba, Tbilissi (1979). 0.1. Lashkov, M.V. Sevost'Yanova, A.V. Shablya and T.A. Shadhverdov, Symposium "Molecular Photonics", Izd. Nauka Leningrad (1970) 299. H. Suzuki, "Electronic absorption spectra and geometry of organic molecules". J Wiley intersciences, N.Y. (1967). M. Gehrtz, C. Brauchle and J. Voitlander, J. Am. Chem. SOC., 104 (1982) 2094. A.S. Kholmanskii and N.I. Kunavin, Zh. Fiz. Khim, 58 (1984) 991. [Russ. J. Phys. Chem., 4 (1984)l. C . Balny and P. Douzou, C . R . Acad. Sci., C 262 (1966) 1235.
462 221 222 223 224 225 226
227 228
229
230 231 232 233
234 235 236
237
238
239 240 241
242 243 244 245
246 247 248 249
P. Appriou, R. Guglielmetti and F. Garnier, J. Photochem., 8
(1978) 145. J.C. Dalton and N.J. Turro, Mol. Photochem., 2 (1970) 133. D.A. Reeves and F. Wilkinson, J. Chem. SOC. Faraday Trans 2, 9 (1973) 1381. A.S. Kholmanskii, I.P. Gloriozov, V.M. Mamaev and K.M. Dvumaev. Zh. Fiz. Khim.. 59 (1985) 934. [Russ. J. Phys. Chem., i4) (198511. D.M. Shigorin, V.G. Plotnikov and V.K. Potapov, Zh. Fie. Khim., 40 (1966) 192. [Russ. J. Phys. Chem., (1) (196611. J.G. Calvert and J.N. Pitts "Photochemistry" translated into Russian, Izd. Mir. Moscow (1968). J.A. Baltrop and J.D. Coyle, "Excited states in organic chemistry" (translated into Russian) Izd. Mir. MOSCOW, (1978). A.S. Kholmanskii, V.F. Tarasov and K.M. Dyumaev, Zh. Fiz. Khim., 57 (1983) 2659. [Russ. J. Phys. Chem., 1 1 (1983)l. D.N. Shigorin, Zh. Fiz. Khim., 44 (1970) 2681 ; 51 (1977) 2894 ; 54 (1980) 1905, 1936 ; 57 (1983) 822. [RUSS. J, Phrs. Chem., 11 (1970) ; 8 (1977) ; 8 (1980) ; 4 (198311. A.S. Kholmanskii, N.I. Kunavin, P.V. Zubkov and K.M. Dyumaev, Zh. Fiz. Khim., 56 (1982) 2794. [Russ. J. Phys. Chem., 1 1 (198211. G. Robinson and R. Frosch, J. Chem. Phys., 38 (1963) 1187. A. Murin, V.F. Mandzhikov and V.A. Barachevskii, Optika i Specktrosk, 40 (1976) 1084. A. Kellmann, L. Lindqvist, S . Monti, F. Tfibel and R. Guglielmetti, J. Photochem., 21 (1983) 223. A.S. Kholmanskii, V.M. Chudakov and A.K. Piraev, Khim. Fie. (1986) 1329. C. Lenoble and R.S. Becker, J. Phys. Chem., 96 (1986) 62. Y. Kalinsky and D. Williams, Chem. Phys. Lett., 86 (1982) 100 A.S. Kholmanskii, Ya.N. Malkin and V.A. Kuz'min, Izv. Akad. Nauk SSSR, Ser. Khim., (1984) 2124. V.A. Murin, V.F. Mandzhikov and V.A. Barachevskii, Optika i Spektrosk, 40 (1976) 1084. M. Mosse and J.C. Metras, J. Chim. Phys., 64 (1967) 691. S.G. Kuz'min, V.A. Barachevskii and V.F. Mandzhikov, Abstracts of reports at the XIV International Congress on high speed photography and photonics, Moscow (1980) p.354. S.A. Krysanov and M.V. Alfimov, Dokl. Akad. Nauk SSR, 172 (1983) 406. S.A. Krysanov and M.V. Alfimov, Laser Chem., 4 (1984) 129. T.A. Shakhverdov, Izv. Akad. Nauk SSSR, Ser. Kh'im., 32 (1968) 1564. T.A. Shakhverdov, and A.V. Shablya, Optika i Spektrosk, 24 (1968) 619. Yu.P. Strokach, V.A. Mandzhikov and V.A. Barachevskii, Khim. Fiz. (1982) 1144 ; (1983) 700. C. Lenoble and R.S. Becker, J. Photochem., 33 (1986) 187. V.A. Murin, V.F. Mandzhikov and V.A. Barachevskii, Optika i Spektrosk, 37 (1974) 378, 674. Yu.P. Strokach, V.F. Mandzhikov, V.A. Barachevskii, N.D. Dmitrieva and R.M. Liberzon, Optika i Spectrosk, 41 (1979) 997 ; 49 (1980) 1195. A.V. Vannikov and A.Yu. Kryukov, Khim. Vys. Energii., 334 (1978) 504. I
463 250 251
252 253
254 255 256 257 258 259 260 261 262 263 264 265 266
267 268 269 270 27 1
272
273 274 275 276 277 278 279 280 281
282
283
284 285
286
287
S. Krysanov and M.V. Alfimov, Chem, Phys. Lett., 91 (1982) 77 * A. Kellmann, L. Lindqvist, S. Monti, F. Tfibel and R. Guglielmetti, J. Photochem., 28 (1985) 547. A. Kellmann, L. Lindqvist, F. Tfibel and R. Guglielmetti, J. Photochem., 35 (1986) 155. Y. Kalinsky, T.E. Orlowski and D. Williams, J. Phys. Chem., 87 (1983) 5333. J. Padwa et al., J. Org. Chem., 40 (1975) 1142. F. Tfibel and A. Kellmann, J. Photoohem., 41 (1988) 299. R. Gautron, Bull. SOC. Chim. Fr. (1968) 3200. R. Gautron, Bull. S O C . Chim. Fr. (1970) 4255. D. Gaude, M. Le Baccon, R. Guglielmetti and R. Gautron, Bull. SOC. Chim. Fr. (1979) 489. D. Gaude, R. Gautron, R. Guglielmetti and J.C. Duffy, Bull. S O C . Chim. Fr. (1981) 14. D. Gaude, R. Gautron, M. Maguet and R. Guglielmetti, Helv. Chim. Acta, 66 (1983) 342. R. Gautron and R. Guglielmetti (work8 not published). H. Pommier, Thesis, University of Marseilles (1972). K. Kister, Thesis, University of Marseilles (1977). E . Davin-Pretelli, Thesis, University of Marseilles (1983). J. Rondon, R. Guglielmetti and J. Metzger, Bull. SOC. Chim. Fr., (1971) 2581. R. Guglielmetti, E. Davin-Pretelli and J. Metzger, Bull. SOC. Chim. Fr., (1971) 2039. K. Brunner, Chem. Ber., 31 (1898) 1943. A. Hinnen, C. Audic and R. Gautron, Bull. SOC. Chim. Fr. (1968) 2066. Cie de St Gobain S . A . , Fr. Patent 2,918,619 (1971) G. Plancher, Chem. Ber., 31 (1898) 1496. R.L. Hinman and E.B. Whipple, J. Am. Chem. Soc., 84 (1961) 2534. 0. Bloch-Claude, P. Rumpf and J. Sadet, C.R. Acad. Sci., 2 4 0 C , (1955) 1426. M. Nakazaki, Bull. Chem. Soo. Jap., 33 (1960) 461. M. Nakazaki, K. Yamamoto and K. Yamagami, Bull. Chem. SOC. Jap., 33 (1960) 466. M. Nakazaki, Bull. Chem. SOC. Jap., 33 (1960) 472. R.F. Meyer, J. Org. Chem., 30 (1965) 3451. J.J. Krutak, (Eastman Kodak) U.S. Patent 3,865,837 (1975). J. Metzger .and H. Plank, Bull. SOC. Chem. Fr. (1956) 1694. R. Guglielmetti, E. Pretelli and J. Metzger, Bull. Soo. Chim. Fr., 8 (1967) 2812, M.G. Ast and M.T. Bogert, J. Am. Chem. S O C . , 61 (1939) 2013. H.H. Fox and M.T. Bogert, Rec. Trav. Chim. Pays-Bas, 54 (1935) 917. N.K. Beresnova, E.R. Zakhs and L.S. Efros, Khim Geterotsikl. Soedin., 7 (1970) 1004. N.K. Bereanova, E.R. Zakhs and L.S. Efros, Khim Geterotsikl. Soedin., 8 (1971) 961. H. Wenker, J. Am. Chem. SOC., 57 (1935) 1079. M. Maguet and R. Guglielmetti, Bull. SOC. Chim. Fr. (1978) 539. M. Maguet, Y. Poirier and R. Guglielmetti, Bull. SOC, Chim. Fr. (1978) 550. M. Maguet, Y. Poirier and R. Guglielmetti, Bull. SOC. Chim. Fr. (1978) 561.
464
288
289 290
29 1 292 293
294
295 296
297
298
299 300 301 302
303 304 305 306
307 308
309
310 311
312 313 314 315
316 317 318 319
320 321 322
M. Maguet, Y. Poirier and R, Guglielmetti, Bull. SOC. Chim. Fr. (1978) 568. J.B. Cloke, J. Am. Chem. S o c . , 51 (1929) 1174. L.C. Craig, H. Bulbrook and R.M. Hixon, J. Am. Chem. S 0 c - 1 53 (1931) 1831. A. Etienne and Y. Correia, Bull. SOC. Chim. Fr. (1969) 3704. A. Debal, Thesis, University of Paris VI (1974)M. Larcheveque, A . Debal and T. Cuvigny, Bull. SOC. Chim. Fr. (1974) 1710. F. Asinger, F. Schmitz and 5 . Reichel, Ann. Chem., 652 (1962) 50. F. Asinger, H. Diem and W. Schafer, Monatsh. Chem., 95 (1964) 1335. F. Asinger, A. Saus, H. Offermanns, D. Neuray and K.H. Lim, Monatsh. Chem., 102 (1971) 321. C. Balny, H. Hinnen and M. Mosse, Tetrahedron Lett. (19681 3097. C.F. Koelsch and W.R. Workman, J. Am. Chem. S o c . , 74 (1952) 6288. G . Arnold and C. Schiele, Z. Naturforsch., 22b (1967) 1228. C. Schiele, 0. Arnold and A. Wilhelm, Tetrahedron Lett. (1966) 5775. C. Schiele and G. Arnold, Tetrahedron Lett. (1968) 245. R . Guglielmetti and J. Metzger, Bull. Soc. Chim. Fr. (1969) 3329. A. Samat, R. Guglielmetti and J. Metzger, Helv. Chim. Acta, 55 (1972) 1782. E.R. Zakhs, N.G. Leshekynk and L.S. Efros, Khim Geterotsikl. Soedin (1973) 539. A. Keller, Thesis of Bale University (1972). D.J. Zwanenburg and Th.A.M.M. Maas, Rec. Trav. Chim. Netherland, 94 (1975) 8. D.J. Zwanenburg and Th.A.M.M. Maas, Rec. Trav. Chim. Netherland, 94 (1975) 215, 218. D.J. Zwanenburg and Th.A.M.M. Maas, Rec. Trav. Chim. Netherland, 95 (3976) 97. E. Bradley, W.D. Cotterill, R. Livingstone and M. Walshaw, J. Chem. S o c . , 18 (1971) 3028. M. Maguet and R. Guglielmetti, Bull. Soo. Chim. Fr. (1982) 323. P. Deslongchamps, C. Lebreux and R.J. Taillefer, Can. J. Chem., 51 (1973) 1665. P. Deslongchamps, Tetrahedron, 31 (1975) 2463. P. Deslongchamps, S. Dube, C . Lebreux, D.R. Patterson and R.J. Taillefer, Can. J. Chem., 53 (1975) 2791. P. Deslongchamps and R.J. Taillefer, Can, J. Chem., 53 (1975) 3029. M. Maguet and R. Guglielmetti, J. Heterocycl. Chem., 15 (1978) 1439. C. Schiele and H.O. Kalinowski, Angew. Chem., 78 (1966) 389. C. Schiele and H.O. Kalinowski, Ann. Chem., 696 (1966) 81. R. Wizinger and H. Wenning, Helv. Chim. Acta, 23 (1940) 247. P. Appriou and R. Guglielmetti, Bull. Soo. Chim. Fr., (1975) 510.
R. Huisgen and V. Weberndorfer, Experiential 17 (1961) 556. J.M. Decrouin, Thesis, University of Caen (1968). P. Appriou and R. Guglielmetti, C.R. Acad. Sci., 27512 (1972) 57.
465 323
L. Soder and R. Wizinger, Helv. Chim. Acta,
325
L.Y. Nivorojkine, N.S. Loteva and W.I. Minkin, Khim Geterotsikl. Soedin., 3 (1972) 318. I. Degani, R. Fochi and P. Tundo, Ann. di Chim., 62 (1972) 570. L. Costa, I. Degani. R. Fochi and P. Tundo, J. Het. Chem., -
324 326
327 328
329
330 331 332
333 334 335
336
337 338 339 340 341
342 343 344 345 346
347 348 349 350 351
352
353 354
355
1779.
42 (1959) 1733,
1 1 (19741.943. R. Wizinger and D. Durr, Helv. Chim. Acta, 46 (1963) 2167. P. Appriou and R. Guglielmetti, C.R. Acad. Sci, 275C (1972) 1549;P. Appriou, R. Guglielmetti and F. Garnier, C.R. Acad. Sci., 279C (1974) 973. J. Houben, Chem. Ber., 37 (1904) 489. R.L. Shriver and A.G. Sharp, J. Org. Chem., 4 (1939) 575A. Abouassali, J. Royer and J, Dreux, C.R. Acad, Sci., 271C (1973) 887. R.W. Tickle, T. Melton and J.A. Elvidge, J. Chem. SOC. Perkin Trans. I, 5 (1974) 569. J. Zsindely and H. Schmid, Helv. Chim. Acta, 51 (1968) 1510R. Dickinson, J.M. Heilbron and F. O'Brien, J , Chem. SOC. (1928) 2077. 0. Bayer, in E. Mueller Ed, Methoden der Organischen Chemie, vol. 7 pt. 1, G. Thieme. Stuttgart (1954). E. Mosettig, Org. React., 8 (1954) 218. R.B. Wagner and H.D. Zook, Synthetic Organic Chemistry, Ed. Wiley, N.Y. (1953) p. 280-315. L.N. Ferguson, Chem. Rev., 38 (1946) 227. G.A. Olah and S.J. Kuhn, "Aldehyde Synthesis" in G.A. Olah, Ed. Friedel-Crafts and Related Reactions, vol. 3 pt. 2, Interscience New York (1964) chapter 38. S.R. Sandler and W. Karol "Organic Functional Group Preparation" Academic New York (1968) chapter 7. R.S. Becker and J. Kolc, J. Phys. Chem., 72 (1968) 997. S. Arakawa, H. Kondo and J. Seto, Chem. Lett., (1985) 1805. S. Tamura and J. Seto, (Sony Corp.) Eur. Patent 184,808 (1986) * Jap.. Patent 61,138,607 (1986) ; U.S. Patent 4 ,693,96B ( 1987). S. Tamura and J . Seto, (Sony Corp.) Jap. Patent 62,299,386 (1987) ' Jap. Patent 60,262,884 (1985) ; Jap. Patent 6 0 , 2 0 8 , 3 ~ 0(1985). A.S. Kholmanskii, Ya.N. Malkin, Zh. Fie. Khim., 60 (1986) 2522. C.C. Price and G.W. Stacy, J. Am. Chem. SOC., 68 (1946) 498. M.S. Newman and H.A. Karnes, J. Org. Chem., 31 (1966) 3980. G.A. Delzenne, G.J. Smets and J.A. Hoefnagels, Belg. Patent 2,012,029 (1969) ; Ger. Patent 1,919,129 (1970). C. Gaurat, C. Mercier and R. Dennilauler, Kodak Path6 (private communication), F. Ribes, R. Guglielmetti and J. Metzger, Bull. SOC. Chim. Fr., (1972) 143. R. Gautron, (Cie de St Gobain) French Patent 1,450,583 (1966). R. Gautron, (Cie de St Gobain) French Patent 1,451,583 (1966). National Cash Register Laboratories, Dayton (USA) results unpublished. J. Gervais and R. Guglielmetti, C.R. Acad. Sci., 271C (1970) 110.
466
356 357 358 359
G. Dumenil, P. Maldonado, R. Guglielmetti and J . Metzger, Bull. S O C . Chim. Fr., ( 1 9 6 9 ) 8 1 7 . J.C. Le Duc and R. Guglielmetti, C.R. Acad. Sci., 282C (1976) 691.
E.R.
Zakhs, C . A . Zvenigorodzkaya and L.S. Efros, Geterot. Soedin., 12 ( 1 9 7 3 ) 1 6 1 8 . J.C. Le DUC, Thesis, University of Brest, 1 9 7 5 .
See Additional Literature (1989 See Additional Literature (1 989 See Additional Literature (1989
- 2001): Chromenes, A33
- 2001): Naphthopyrans, A41 - 2001): Spiropyrans, A51
Khim.
467
Chapter 9
1. 1.I
4n+2 Systems: Fulgides J. Whittal
INTRODUCTION Definition
:xo
Fulgidesare derivativesof dimethylenesuccinicanhydridesas shown in formula(1).
R4
These were firstinvestigatedby Stobbe (ref. 1) aroundthe turnof the century. He reportedtheir synthesisby the reactionnow knownas the StobbeCondensation, whichwas extensivelyinvestigatedby Johnsonand his co-workerswho reviewedthe subjectin 1951 (ref. 2) (scheme 1). Stobbe named these succinicanhydride derivatives'fulgides'(fromthe Latinfulgere- to glistenand shine)becausethey were frequentlyobtainedas beautifulreflectivecolouredcrystals.
-
R'
[Stobbe Condensation]
R2
C02R
468
OHe
R4
Scheme 1. Synthesisof fulgidesvia Stobbe condensations. The mechanismof the Stobbecondensationwas elucidatedby Johnson& A (ref. 3) who demonstratedthe formationof an intermediatelactonicesterwhich subsequentlyundergoesan irreversiblebase inducedeliminationto give the halfester productas indicatedin scheme 1. Stereochemistw FulgidessubstitutedwithfourdifferentgroupsR1, R2, R3, R4 in (1)can exist as four geometricalisomers[(E,E), (E,Z), (Z,E), (Z,Z)]. The mostconvenientmethodof determiningthe stereochemistryof fulgidesis by protonn.m.r. spectroscopy.Some early workerson fulgideswere unableto acceptthe (E,E) configuration for bis-atyl fulgides(1) [R1 = R4 = H,R2, R3 = Ar] claimingstericovercrowdingwouldpreventthis isomerbeingformed. This causedsome incorrectassignationsof the protonn.m.r. signalsto be made (refs. 4-8). However, in 1970 Cohen and co-workers(ref. 9) demonstratedby X-ray crystallographythat bis (p-methoxyphenyl)fulgide(2) (preparedvia the Stobbe condensation)had the (E,E)-configuration. 1.2
469
The arylgroupswere not co-planarwith the anhydrideringbut were twistedby approximately300 fromthe plane of the anhydridering. Hellerand Hart (ref. 10) demonstratedthat the two most importantfactors affectingthe protonn.m.r. signalsin the n.m.r. spectraof fulgideswere the shieldingof all groupsunderlyingaryl ringstransto the carbonylgroupssuchthat signalsare at appreciablyhigherfieldsand the deshieldingof alkylor hydrogengroupscis to the carbonylgroupso that they appear at lowerfield. Carefuluse of these n.m.r. effects togetherwithcharacteristicsplittingpatternsfor vicinaland allylichydrogensallow unambiguousassignmentof stereochemistryto be made in mostcases. 1.3
Mechanismof ohotochromismin fulaids Stobbe observedphotochromicbehaviourin fulgideswith at least one aryl groupbut failedto explainthis phenomenonadequately,suggestingfirstlythat it was a crystaleffect (ref. 11) and laterthat it was due to E-2 isomenzations(ref. 12). In 1933, Schonberg(ref. 13) triedto rationalizean observationby Stobbe (publishedby Houbenin 1922) (ref. 14) that the tetraphenylfulgide (3)was converted into 1,2- and 1,4-dihydronaphthaIene derivatives(1,2- and 1,4-DHNs) (4) and (5) by suggestinga light-induced diradicalmechanism. There has been no evidencefrom e.s.r. studiesto supportthe presenceof diradicals.
ps m0+m O
* /
/
L
/
h
(3)
/
Ph (4)
0
Ph H
(5)
470
In 1968, Santiagoand Becker(ref. 15) suggestedthat the photochromismwas a molecularphenomenondue to a photochemicalring-closureto form a 1,8adihydronaphthalene(1,8a-DHN). This was inferredfrom oxidationwith molecular oxygenof the coloured1,8a-DHN (7) (formedon irradiationof the diphenylfulgide (6) in degassedmethanolat -77%) to give 1-phenylnaphthalene-2,3-dicarboxylicacid anhydride(8).
P Ph
h
/
z0- hu
\
\
H
0 -
H
/
/
0
H
(61 (7) (8) Hellerand Hart (ref. 10) had already establishedthat photochemical and thermalelectrocyclicring-closureof aryl fulgidesto 1,8a-DHN's occurredby conrotatoryand disrotatorymodes respectivelyin accordancewith the WoodwardHoffmanselectionrules(ref. 16).
Ph
[con.]
/ 0
0 -
\ Ph
0 Ph
This is exemplifiedby the (E)-fulgide(9) whichcyclizesphotochemicallyin a conrotatorymodeto givethe cis-l,8a-DHN (10). Thermalelectrocyclicring opening of (10) in a disrotatorymode givesthe (Z)-fulgide(11). The relativestereochemistryof these fulgideswas establishedby protonn.m.r. spectroscopy(ref. 10).
471
Photochromism in fulaides Typically,fulgidesare yellow or orange crystallinecompoundswhich change to orange, red or blue uponirradiationwith ultra-violetlight. Photochromismhas been observedin crystal,solution,polymersand glassesover a wide range of temperaturesand conditions.The phenomenonhas been shownto be subjectto a rangeof stericand electroniceffectswhich make the exact photochromic properties able to be modifiedby moleculardesignand tailoring. 1.4
ChrOmODhOreStructureand Dhotochemistrv . of fulaides Freudenbergand Kempermann(ref. 17) reportedthat the U.V. spectrumof dimethyl(E,E)-bisbenzylidenesuccinatewas similarto that of methyltrans-cinnamate and did not resemblea 1,6diphenylbutadiene type spectrum. This was taken as evidencethat the fulgidesystemhas two largelyindependentcinnamicacidtype chromophores(Fig. 1). 1.5
Fig. 1. Independentcinnamatetype chromophores(A) and (B). Furtherevidence of this independenceof chromophoreswas presentedby Hellerand Szewczyk(ref. 18) who showedthat the U.V. spectrumof (E,Z)-bis-aphenylethylidene succinicanhydride(12) closelyresemblesthe
combinedspectraof the (E,E) - and (Z,Z) isomers,as shownin Fig. 2.
472
300
200
400
Wavelength (nm)
Fig. 2. Comparisonof (E,Z) fulgideU.V. spectrumwith 1:1 mixtureof (E,E) and (Z,Z). These observationssuggestthat fulgidescan be viewedas two separate, a$unsaturatedacid chromophoresfor which electronicexcitationin one chromophore willthen allow cyclizationonto a suitablearomaticringin the other half of the molecule. In the relatedGlactone(ref. 19) derivativesshownin Fig. 3 photochromism is observedbut cyclizationis exclusivelyontothe aryl ringoppositeto the carbonyl group. It is likelythat thisphotoreactionoccursby a mechanismsimilarto that of the relatedfulgides.
n
Fig. 3.
Exclusivephotocyclization onto phenylringoppositeto carbonylgroup.
473
Paetzoldand llge (ref. 20) studiedthe photochemistry of a comprehensiverange and in the of phenylfulgidesand showedthese had typicalR + X* photoreactivity singletstate E - Z isomerizationsaboutone doublebondcompeteswith electrocyclic ringclosureto the colouredform but that the maindeactivationroutewas internal conversiondue to strongstericinteractionsin the molecularframework. This process increasesin the series mono-,tri- and tetra-phenylfulgides due to increasedsteric interactions.At roomtemperature,no luminescencewas observedin any of the fulgidesstudied. Intersystemcrossingwas not seen even with internaland external heavy atom effectexperiments.When tripletsensitizerswere used, the tripletstate energieswere foundto be in the range 220-250 kJrnol-1and simultaneousE - Z isomerizationsabout bothdoublebondswas the only observedphotochemicaltriplet reactivity. These resultsare summarisedin Scheme 2.
1,8a-DHN
Scheme 2. Photochemistry of phenyl fulgides (Paetzold and Ilge ref. 20).
474
1.6
Colourand Constitutionof Fulaidesand their photochromes
The carbonylgroupsin conjugationwith the double bondsare assumedto be responsiblefor the colourin bothfulgidesand their photochromes and changesin colourby substituenteffectshave been interpretedin simpleresonancetheory(ref. 21) termswith allowancesfor stericeffectsin some overcrowdedmolecules. The structuresin Fig. 4 showthe basicskeletonsfor mono-aryland monoheteroarylfulgidesand their respectivephotochromeswiththe alternateatoms starred. Electrondonatinggroupsin the starredpositionsare then in conjugationwith the main chromophorebut have littleeffectwhen they are in unstarredpositions.
0
*\ (13)
R'
*\ (14)
Fig. 4. Simple Resonance theory analysis of chromophoresin fulgides.
There are two important featuresin thisrepresentation,the firstbeingthat in the photochemical cyclizationthere is a change in sequenceof the starredcarbonsso that any electron-donating substituentthat is in conjugationwith a carbonylgroupin the fulgideis not in conjugationwith a carbonylin the photochromeand yice Vera and, secondly,in the photochromeof the heteroarylfulgide(16) there is only one carbonthat is starredand not quadri-valentso that there is only one positionavailable for addingelectron-donating substituentsthat will have a majoreffecton the colourof the photochrome.
475
2. 2.1
PHENYL FULGIDES Methoxvarvl . Fulaides One groupof fulgideswhich have been extensivelystudiedare the methoxyarylfulgides(refs.22, 23) whichclearly illustratethe above electroniceffects and show how stericeffectscan dramaticallyalterthe photochemistry. The fulgideswhich have been comparedare (17 20).
-
( Ua) R=H (18b) R=Me
(17a) R=H (17b) R=Me
0 OMe
\
R
Me
Me
Me
(20a) R=H (20b) R=MeO
(19a) R=H (19b) R=MeO
Some selecteddata of quantumyieldsfor the photochemistryof these fulgides are given in Table 1 (irradiationwavelength366 nm). Table 1:
Quantumyieldsfor the photoreactions of selectedfulgides at 366 nm. (Solvent- cyclohexane)after Paetzoldand llge (ref. 20).
Fulgide @E+Z Q, E + 1,8aDHN
(17a) 0.3
(im)
0.18
0.054
(18a) 0.25
10-4
(18b) (19a) 0.14 0.23 0.052 0.085
476
Table 2 givesselecteddata for the U.V. absorptionbandsof some of these fulgidesin toluenewhichshowsthe variousstericand electroniceffectson the absorptionmaximumof these fulgides. Table 2 : U.V. absorptiondata for methoxyarylfulgides. (198) (19b) (208) (20b) Fulgide hmaxnm 314 363 334 342 log&m 3.93 4.09 3.92 3.92
The methoxysubstituentaffectsthe absorptioncharacteristics of the compoundsby donationof the non-bondedp-electronsof the oxygenintothe unsaturatedsystem. Effectsare most markedwhen the methoxygroupis in conjugationwith the main chromophore.Methoxygroupsin the 2,4,or 6 positionof the phenylsubstituents causea bathochromicshiftof the longestwavelength absorptionbandof the fulgidebut have littleeffect on the visibleabsorptionbandof the 1,8a-DHN while methoxygroupsin the 3 or 5 positionshave littleeffecton the fulgideabsorptionbandsbut cause a majorbathochromicshift on the visiblebandof the 1,8a-DHN (ref. 24). The methoxygroupsin the 3 and/or 5 positionsalso increasesensitivityto colouringby ultravioletlightand it is thoughtthis is due to the reactionbeing mechanistically similarto the firststep in electrophilicsubstitution of afyl compounds as shownin scheme3 [i.e. the dipolarstate (21) can be writtento representthe excitedstate of the fulgideresponsiblefor photocyclization to coloured1,8a-DHN].
l*
477
etc]
Scheme 3. Mechanistic representation of photocyclization.
These electroniceffectsare alteredby stericeffectsas fulgideswith 2,4, or 6 methoxyphenyl substituents are only very weakly photochromicif there is a hydrogen in the positionp- to the carbonyl. For example (18a) has a quantumyieldfor and colouringof lessthan 1 x 10-4 (Table 1) and is practicallynon-photochromic compound (19b) (whichhas activating3,5-dimethoxysubstituents)is Only very weakly photochromic.However, when the hydrogenis replacedby an alkylgroup (fulgides(18b) (20b)) or the para-methoxysubstituentis removed(fulgide(19a) these compoundsare markedlyphotochromic. The differencein photochromicpropertiesis clearlydemonstratedby the spectrain Fig. 5 with the firstfor (20b) whichgivesa yellowto deep blue COlOUr changein lessthan three minutesirradiation,while the next for (19b) showsonly a smallcolourchange after morethan twentyfour minutesirradiationundersimilar conditions(ref. 22). The explanationfor this large effect is that when there is the hydrogenpresent coplanarityof the chromophorecontainingthe aryl groupis readilyobtainedand photoexcitation can give a dipolarstate representedby (22) very easily and thisstate cannotgive cyciizationto a 1,8a-DHN.
478 1.o
,*--.. I
#'
300
500
400 -(A)
300
400
,
(B)
'..
*
600
Wavelength (nm)
500 600 Wavelength (nm)
Fig. 5 . (A) Fulgide(20b)(solidline) and after 3 minutesirradiationat 366 n.m. (dashed line). (6) Fulgide(19b) (solidline) and after 24 minutesirradiationat 366 n.m. (dashedline).
However, when the hydrogenis replacedby an alkylgroup, stericinteractions with the adjacentcarbonylgroupof the anhydridemoietyresultsin a morefavoured excitationof the other chromophore[representedby dipolarstate (23)]whichthen allowsphotocyclization to give a colouredcompound.
479
Similarobservationsare made with the heterocyclicfulgides(24) whichare stronglyphotochromic when R is alkylbut not photochromic when R is hydrogen (ref. 25).
0
. . Slb[OLdescQ&jintna the Fluorenvlidene~lplco The previoussectiondiscussedelectroniceffectsmade by alteringthe substituentson the aryl ringontowhichthe photochemicalelectrocyclicringclosure was occurring.To make majorchangesin the rangeof activatingwavelengthsfor photo-coloration, it is necessaryto make changesin the structureof the grouping whichcyclizesontothe aryl ring. Unsaturatedchromophores whichhave terminal fluorenylgroupingshave large bathochromicshiftsrelativeto otheraryl groups. Stobbe had describedfluorenylidene(bisphenylmethylene) succinicanhydrideas deep red crystalswhile Goldschmidt(ref. 26) reportedthat bis(fluoreny1idene) Succinicanhydridewas a blackcrystallinecompound. Both fulgidesare nonphotochromic.Heller and Whittall(ref. 27) made the fluorenylfulgide(25) which was a deep red crystallinecompoundwhichgave a deep blue colourwhen irradiatedover a wide range of wavelengths,even givinga low conversionintothe blue 1,8a-DHN (26) when irradiatedat 546 nm. The u.v./visiblespectraof (25) and its photostationary state when irradiatedat 366 nm are shownin Fig. 6. 2.2
480
0.5
-
'\
0.0
350
,
450
'. *.-
550 650 Wavelen,@ (nm)
Fig. 6. Fulgide (25) (solid line) and its photostationary state at 366 nm irradiation (dashed line) [For a 0.73 X 10-4 M solution in toluene].
481
2.3
PolvDhotochromic Fuleides. When two methoxysubstituents are unsymmetricallypositionedon the phenyl ring(as in fulgide(27) cyclizationontothe two ortho-positions givestwo 1,8a-DHNs (28) and (29) one of which has a methoxy-groupin a starredpositionand is blue in colour(28) while the other has not and is red (29) (ref. 28). Thistype of fulgidehas been calledpolyphotochromic.The blue form (28) can be selectivelyremovedby photochemicalringopeningby irradiationat longerwavelengthswhere the red form (29) does not absorb.
O h'v/Me
mo Me Me
FATIGUE RESISTANCE w e ResistantPhotochromic Fulaids One of the majorproblemsassociatedwith organicphotochromic systemsis the irreversibleformationof non-photochromic productsby photochemical or thermal reactionsof the colouredform. In the fulgidesystemsdiscussedso far these major reactionsare due to a labilehydrogenat the site of cyclizationontothe aryl ring whichtendsto allow thermal, photochemicaland oxidativereactionsto occuras shownin scheme4. 3. 3.1
482
Scheme 4. Mechanisms for colour loss from fulgide photochromes. The above reactionsdo not involvereactiveradicals,ionsor dipolar intermediatesand it followsthat fatiguecan be virtuallyeliminatedby replacementof the angular8a-hydrogenin the colouredform by some othergroup, e.g. methyl. By introduction of methylgroupsin placeof the ortho-hydrogens,fatigue resistant photochromicfulgides(ref. 29) were obtainedi.e. (30)R=H, OMe, are pale yellow fulgideswhichgive thermallystable 1,8a-DHNs (whichare red for (31,R=H) and purplefor (31,R=OMe) on exposureto U.V. light.
R
Me
These 1,8a-DHNs do not undergomethylshifts,ethane eliminationor thermal disrotatoryringopeningup to 16OOC. The lack of thermalreversalof colouredforms (31)to fulgides(30)is attributedto stericinteractionsbetweenthe 1 and 8a methyl groupswhich must occurin the alloweddisrotatorymode. This stericinteractionis absent in the conrotatorymode and photochemical reversalof the 1,8a-DHNs (31 R=H, OMe) to fulgides(30)on exposureto white light,are efficientprocesses.
483
3.2
Fuwl fulaides For the fulgidesdescribedabove the extent of conversionto the colouredforms
at their photostationary stateswas low. Replacementof the phenylgroupby a 3-fury1
group(whichphotocyclizesexclusivelyontothe 2-positionand hence the 4-position does not need protectingwith a methylgroup)gives fulgideswhichconvertquantitativelyto the colouredform on irradiationwith 366 nm light. The firstexampleof this type of fulgideis (32)whichgives (33)on irradiation(ref. 30).
(33)
(32) The u.v/visiblespectraof (32)and (33)are shownin Fig. 7.
-1
---
\
---.
Wavelength (nm) Fig.7. Absorptionspectraof fulgide(32)(solidline)and its colouredform (33) (dashedline) [10-4Msolutionsin hexane].
A recentreportby Kuritaa d (ref. 31) has illustratedthe stericeffectson the quantumyieldsof colouringof fulgidesas shownin Table 3 for a range of fulgides (34a-e).
(34a) R = H
(34c) R = Et (348) R = iPr (34b) R = Me (34d) R = nPr
Table 3:
Stericeffectson quantumyieldsfor colouringof fulgides.
Fulgide CP (colouring) (at 366 nm)
(34a) (34b) (34c) (34d) 34e) -O* 0.19 0.34 0.45 0.62
(tsee Heller a
d ref. 25).
The variationof size of R in (34a-e) does not affectthe colourof the fulgideor the photochromeand the quantumyieldfor reversalto any extent.
3.3
The use of fulgide(32) as an actinometerfor the ultra-violetregionwas describedby Hellerand Langan(ref. 32). The quantumyieldfor colouringover a range of 313-366 nm was foundto be 0.20 and was independentof fulgide concentration and temperature. It was also unaffectedby recyclingso that the actinometercouldbe usedrepeatedlywithoutthe introduction of errors. (ref. 33) describesthe use of this A more recentreportby Scaiano actinometerfor 308 nm laser calibrationand concludesthat the quantumyieldsfor colouringare largelypowerindependentand that opticallydilutesolutionscan be used as long as the properproceduresin the 2alculationsare followed. This fulgidewas also usedfor actinr)metryin two-laserexperimentsinvolving the photolysisof short-livedintermediates.The extinctioncoefficientsof these could then be determinedby a techniquethat involvedfewer assumptionsthan previous methods.
485
The use of the reversereactionof thisphotochromicsystemas a visible actinometerhas also been investigated. The quantumyieldfor the changefrom coloured(33) to colourless(32)has been shownto exhibita lineardependenceon wavelength. The variationof quantumyieldwith wavelengthfor a toluenesolutionat 21% is given in equation1. CD = 0.179 -2.4 x 10-4 (hnm)
(1)
Heller (ref. 34) recommendthe usr!of the crystalline,7,la-dihydrobenzothiophene(35) as an actinometerin the vi::ible range 430 to 600 nm becauseit has a broaderabsorptionbandand a highermolarextinctioncoefficient(12,000 dm3 m0l-l cm-l in toluene)than (33).
Again a lineardependencyof CP (reversal)with respectto temperatureand wavelengthis reportedas shownin equation2 (for a toluenesolution).
-
-
= 0.0737 {[h(nm) 2.31 x T (OC)] x 1O-4}
The X-ray structureof (35) has been determinedand is shownin Fig. 8 (ref. 35).
Fig. 8. X-ray crystalstructureof (35).
(2)
486
4.
STERIC EFFECTS Adarnantvlidenefulaides The fatigueresistantfury1and thienylfulgidesdescribedfor visibleactinometry have quantumyieldsfor bleachingwhichare fairly low (0.02-0.06). One of the potentialuses for fatigue-resistant photochromics is in opticaldata Storage (first suggestedby Hirshberg(ref. 36) in 1956 whichinitiatedthe moderninterestiil the applicationsof photochromics).The methodfor recordingdata has beer; to use visiblelasersfor the writingprocessand hence the rate ar whichinformaticncan be recordedis directlyproportionalto the quantumyield for hieaching. Her!% meihods of increasimthe recordingefficiencywere sougnt. groupin iulgidef32j by the Hep!acementof the isopropylidene adaiiantylidenegroupgivesfulgide(36)(ref. 37;. The hu!Q and rigid and so the systemis frea f r m ackrnantylidenegrouphas all chairconformations anr;ie and conformational strain and so the fulgide(36)is stillthermallystableand pbc,mhemicalfatigue-resistant.
4.1
The adamantylidenegroupdoes not markedlyaffect the abso:piion characteristics of the fulgideor its colouredform. The quantumyieldio: colouringis only slightlyreducedbut there is a six-foldincteasein the quantumyieldfor the bleachingreactiondue to weakeningof the 7,7a-bondin the 7,7a-DHBF. These resultsare tabulatedbelow(Tables4 and 5). Table 4: Fulgide
(32) (36)
Spectroscopicdata for isopropylideneand adamantylidenefulgides.
1-
346 344
(nm)
Log 3. 81 3.78
O(colouring)(at 366 nm) 0.20
0.15
487
Table 5:
Spectroscopic data for colouredformsof isopropylideneand adamantylidenefulgides.
*DHBF
hmax(nm)
(33) (37)
494 520
LogEmax % Conversiont @ bleaching+ 100 0.047 3.89 3.10 93 0.296
* DHBF - dihydrobenzofuran tAt 366 nm photostationary state +At 546 nm in toluenesolutionat 21OC. The quantumefficiencyfor bleachingof 7,7a-DHBF (37)showsa similarlinear dependenceon wavelengthand temperaturedescribedfor 7,7a-DHBT (35). This relationshipis shownin equation3 (toluenesolutions) @ = 0.4326
- {[3.285h(nm)- 16.4 T (OC)]x I 0-4}
(3)
4.2
HeliochromicC m A bond-weakeningeffect of the bulky spiroadamantanegroupcan be seen in the 1,2-DHN derivativesfromfulgideswhichhave a hydrogenpresentin the position of photocyclization ontothe aryl ring. Photocyclization to coloured1,8a-DHNs which undergothermal[1,5]-hydrogenshiftsto form 1,2-DHNs (cf. Scheme 4). The 1,2carbon-carbonbondof the 1,2-DHN was sufficientlyweakenedby the bulky2spiroadamantanegroupthat photochemicalringopeningcan occurto give coloured orthoquinodimethane type derivatives. This reactionis not seen in similar compoundscontaininghydrogen,aryl, alkyl, or simplecycloalkylgroupsin the 2positionof the 1,2-DHN. Some examplesof this reactionare the methoxyphenyl (adamantylidene)fulgides(38)whichcolourto red or blue 1,8a-DHNs (39),typicalof the normalfulgidephotochromicresponsewhen exposedto U.V. irradiationbut the colourfades rapidly(typicallya few secondsfor toluenesolutionsat ambient temperature)to give the yellow 1,2-DHNs (40) which undergophotochemical electrocyclicring openingto give the colouredortho-quinodirnethane derivatives(41) on U.V. irradiation.This derivative,whichfaded rapidlyto the colourless benzocyclobutenes (42) was detectedby flash photolysis.The structuresof some of the benzocyclobutenes (42) were confirmedby X-ray crystallography (refs. 38, 39). *J-leliochromic corrlgpundsare pale yellowphotochromiccompoundswhichhave highefficiencyfor colouringwith near U.V. irrailiationwhile the colouredformshave a low efficiencyphotochemicalreverseactionbut have thermalreverseat ambient temperaturesso that these compoundscolourin sunlightand are suitablefor sunglassapplication(ref. 29).
(
M
e
O
)
,
,
,
q
(38) R=H,Me n = 0-3
.1 c
(39) R=H, Me n = 0-3
489
The flashphotolysisexperimentfor (40) R=Me n=3 showedthat (41) R=Me n-3 fadea with Iiwr order kineticswith a half-lifeof 630 f 50 pis at 23% in hexane (ref. 39). The rearrangementreaction(42) -$ (40) in tolueneat 78OC had half-lives that variedfrom a few minutesto severalhoursdependingon the natureof R and the numberof methoxygroups. While the ringclosureof oquinodimethanesinto benzocyclobutenes is Well known,the synthesisof furano-or thiopheno-cyclobutenes is much more diiircuk. Hence replacementof the phenylgroupsby five memberedheteroaromatic ringsallowsthe colourchangesto be seen at ambienttemperaturesand formsthe basisof a patentby Heller21?, a. (ref. 40) for the use of these photochromics in photoreactivelenses. The fulgides(43) are canvertedto the fused ringcompounds (45) by pholocyclization to (44) whichundergoa thermal[I ,5]-hydrogenshift.
ring openwith highquantum The coiourlesscompounds(45) photochemically efficiency(estimatedto be greaterthan 90%) (ref. 29) on exposureto U.V. lightto yield the highlycolouredbismethylenederivatives(46), whichshow remarkable photochemicalstabilityyet fade rapidlyat ambienttemperaturesby firstorder Kinetics !o give the parenicompound(45).
hu
I_)
A
c--
R
The absorptionspectrafor (45) and (46) (X = S, R1 = Me, R* = H,R3 = H) are shownin Fig. 9 and indicatesa yellowto deep blue colourchange.
490
0.
Wavelength (nm)
Fig. 9 Absorptionspectraof (45) in toluenebefore(solidline)and after (dottedline) exposureto a flashgun (ref. 29). A wide range of substituentshave been introducedonto the 3-fury1and 3-thienylgroupsof the fulgideprecursors(includingbenzannelationof the 4 3 positions)whichgive heliochromiccompoundswith a wide range of colourchanges fromorangeto deep blueand fade rateswhichvary from undera minuteto many hoursunder similarconditions.
CONCLUDING REMARKS The fulgides(ref. 41) one of the oldest knowngroupof organicphotochromic compounds,continueto attractmajorinterestbothacademicallyand commercially. This is due to their remarkablefatigueresistancecombinedwith photochromic propertieswhichcan be modifiedby moleculartailoring. This is exemplifiedby fulgide(30) which shows near ideal photochromic behaviouras its colouredform (31) only absorbslightweakly in the 300 nm to 370 nm regionand hence (30)showsa linearcolouringresponse(i.e. a minimalinternalfilter effect). Irradiationof fulgide(30)at 366 nm resultsin a quantitativeconversioninto the red 7,7a-DHBF (31). The fulgideis widelyusedas a chemicalactinometerfor the near U.V. region. The quantumefficiencyfor white lightbleachingis not good enough for opticaldata storageand so modification to give fulgide(34) whichshowsgood propertiesfor opticaldata storagewas carriedout. Modificationof (34) provides fulgidesof type (41) whichare precursorsto materialswhichare well suitedfor sunlightactivevariabledensityopticalfilters(sunglassesare an obviouscommercial 5.
491
applicationof this phenomenon). Other reportedapplicationsof these fulgidesare in spatiallightmodulation(ref. 42), opticalwaveguideconstruction (ref. 43) and nonlinearopticalswitching(ref. 44). This range of applications,togetherwith the convenientsynthesesof fulgides, indicatea revivalof interestin theirchemistryand application. 6.
REFERENCES H. Stobbe Ber., 37 (1904) 2236 W. S. Johnsonand G. H. Daub Org. Reactions,6 Ch. 1 (1951). D. A. Dunnigan,W. S. Johnsonand A. L. McClaskey,J. Am. Chem. SOC.,72 (1950) 514. G.Brunowand H. Tylli,Act. Chem. Scand., 22 (1968) 590. B. Weinstein,K. T. Wang and G. A. Swoboda,J. Chem. SOC.,(C) 1967,161. S. H. Harper, A. D. Kempand J. Tannock,J. Chem. SOC.,(C) 1970,626. S.M. Abdel-Wahhaband N. R. El-Rayyes,J. Chem. SOC.,(C) 1971, 3171. D.P. Chakrabarty,T. Sleigh, R. Stevenson,G. A. Swobodaand 6. Weinstein, J. Org. Chem., 31 (1966) 3342 . M. D. Cohen, H. W. Kaufrnann,D. Sunnreichand G. M. J. Schmidt,J. Chem. SOC., (6) 1970, 1035. R. J. Hart and H. G. Heller,J. Chern. SOC.,PerkinTrans. 1, 1972,1321. H. Stobbe Ann., 359 (1907) 1. H. Stobbe2. Elektrochem.,14 (1908) 473. A. Schonberg,Trans. Faraday SOC.,32 (1936) 514. J. Houben, Methodender OrganischenChemie, Vol. 111, p. 1033, Leipzig1922. A. Santiagoand R. S. Becker,J. Am. Chem. SOC.,90 (1968) 3654. P. J. Darcy, R. J. Hart and H. G. Heller,J. Chem. SOC.,PerkinTrans. I, 1978, 571. K. Freudenbergand T. Kempermann,Ann., 602 (1957) 184. H. G.Hellerand M. Szewczyk,J. Chem. SOC.,PerkinTrans. 1,1974, 1487. H. G. Heller, Personalcommunication. R. Paetzoldand H. D. Ilge, J. Prakt.Chem., 326 (5)(1974) 705. J. Griffiths,Rev. Prog. Coloration,11, (1981) 37. P. J. Darcy, H. G. Heller,S. Patharakorn,R. D. Piggottand J. Whittall,J. Chem. SOC.PerkinTrans. 1, 1986, 315 (and referencescitedwithin). H. D. Ilge, H. Langbein,M. Reichenbacherand R. Paetzold,J. Prakt. Chem., 323 (3), (1981) 367. R. D. Piggot,Ph.D. Thesis, Aberystwyth,1975. A. P. Glaze, S. A. Harris, H. G. Heller,W. Johncock,S. N. Oliver, P. J. Strydom and J. Whittall,J. Chem. SOC.,PerkinTrans. 11985, 957. S.Goldschmidt,R. Riedleand A. Reichardt,Ann, 604, (1957) 121. J. Whittall, Ph.D. Thesis, Aberystwyth,1979. 0. Cresente,Ph.D. Thesis,Aberystwyth,1976. H. G. Heller. 'Fine chemicalsfor the electronicsindustry',Ed. P. Bamfield,p. 120 (1986). P. J. Darcy,H. G. Heller, P. J. Strydomand J. Whittall,J. Chem. SOC.,Perkin Trans. I, 1981, 202. Y. Yokoyama,T. Goto,T. Inoue, M. Yokoyamaand Y. Kurita,Chem. Letts., 1988, 1049 H. G.Hellerand J. R. Langan,J. Chern. SOC.,PerkinTrans II, 1981, 341 V. Wintgens,L. J. Johnstonand J. C. Scaiano,J. Am. Chem. SOC.,110 1988, 51 1. H. G. Heller, A. P. Glaze, C. J. Morganand M. Rubin, Unpublishedresults.
492
H. G. Heller and M. Kaftory,Unpublishedresults. Y. Hirshberg,J. Am. Chem. SOC.,78, (1956) 2304. H. G. Heller and J. Whittall.To be publishedshortly. S. Patharakorn,Ph.D. Thesis, Aberystwyth,1979. W. Johncock,Ph.D. Thesis, Aberystwyth,1982. H. G. Heller, S. N. Oliver, J. Whittall,W. Johncock,P. Darcy and C. Trundle, G.B. 2146327A, 1985. H. G. Brown,Photochromism,Techniquesof Chemistry,Vol. Ill, Wiley Interscience, New York, 1971. C. J. Kirkbyand I. Bennion,IEE Proc., 1335, (1986) 98. I. Bennionand A. G. Hallam, Radio and ElectronicEngineer,53, (1983) 313. R. Cush, C. Trundle, C. J. G. Kirkbyand I. Bennion,Electron.Letters, 23, (1987) 419. See AdditionalLiterature(1989-2001): Fulgides,A75 See AdditionalLiterature( I 989-2001): Fulimides,A87 7.
LIST OF PATENTS RELEVANT TO THIS WORK
Country
Number
THERMALLY REVERSIBLE PHOTOCHROMICS Great Britain France Germany Holland Holland Hong Kong Italy Japan U.S.A. U.S.A.
1,442,628 74129508 P2441759.2174 74111587 77114607 243180 1030606 100791l74 4,145,536 4,182,629
THERMALLY IRREVERSIBLE PHOTOCHROMICS Great Britain France Germany Holland Holland Italy Japan U.S.A.
1,464,603 75123123 P2532224.1 7508987 77112693 1046887 89995175 4220708
ADAMANTANONE COMPOUNDS Great Britain Great Britain France Germany Holland Italy Japan U.S.A.
2002752A 205 18 13A 781215 19 P2831108.2 78107814 25815/78 893 14/78 4220708
USE OF THERMALLY IRREVERSIBLE PHOTOCHROMICS Great Britain
1600615
493
Chapter 70
1
4n+2 Systems: Spirooxazines N.Y.C. Chu
I NTRODUCT ION
I n t h i s chapter the term splrooxazine i s used t o denote a molecule containinga 3H oxazlne r i n g In which t h e number-3 carbon o f t h e r i n g i s involved i n a s p i r o linkage.
Furthermore, t h e molecule contains condensed r i n g
s t r u c t u r a l features such t h a t h e t e r o l y t i c cleavage and reformation of t h e carbon-oxygen s i n g l e bond of t h e oxazine r i n g i s feasible.
As I n t h e case of
sp iropyrans, t h i s heteroly f i c c I eavage of t h e carbon-oxygen bond r e s u l t s i n a change o f valence-bond s t r u c t u r e and conformationof t h e molecule. Consequently, the molecule absorbs i n d i f f e r e n t region o f t h e spectrum and, thus, changes i t s c o l o r appearance. This bond cleavage and reformation reaction can be represented by Scheme 1.
Scheme 1,
Generalized photochromicr e a c t i o n o f spirooxazines.
Since t h e chemical structures of spirooxazinesare c l o s e l y r e l a t e d t o those
of spiropyrans, t h e development of t h e photochromismof spirooxazines Is
mode!ed a f t e r spiropyrans. Thus many of the spirooxazinecounterparts t o spiropyrans have been synthesizedand t h e i r photochromicproperties investigated. The parent r i n g systems of t h e known spirooxazines are presented i n Table 1. The numbering and naming of spirooxazinesw i l l follow the IUPAC r u l e s of organic nomenclature. Thus, t h e names f o r 2,11-[3anaphth[2,
1-a[1,4]0xaz
ine],
1, 4, and 6
are spiro[indol ine-
sp i ro[ indol ine-2,2'-[2k!Iphenanthro[9,
[1,4]oxaz ine] and sp iro[ indoline-Z,3(-[3filpyr
ido[3,2-fl[1,4lbenzoxaz
10-L-il
ine],
respectively. These preferred names were used by t h e Chemical Abstracts f o r
the 1967-71 index period.
The present chemfcal abstracts index names are 1,3-
dIhydrospiro[2~-indole-2,3~-[3~naphth[2,l-~[l,4]oxazlne],
[2.k!-indoie-P,2'-C2~phenanthro
[2H-indole-2,3'-[3apyr
[9,10-~[1,4]oxaz
ido[3,2-f-J[1,4]benzoxaz
1,3-dihydrospiro
lne] and 1,3-di hydrospiro
ine] f o r 1 , 4 and
6,
respectlvely. Throughoixt t h i s sectionr 1, 4, 6 and so f o r t h w i l l be used to
494 TABLE 1 Parent r i n g systems o f known spirooxazines
replace t h e f u l l spirooxazine names as glven above. parent compound 1,3,3-trimethyl-l, -2,3'-[3Jjlnaphth[2,1
-a[1,410xaz
fi, r e f e r s
I ne].
Thus, f o r example, t h e
t o 1,3,3-trimethylspiro[indol ine
The photochromic phenomenon of spirooxazine compounds was f i r s t reported by
Fox ( 1 ) .
They syntheslzed t h e parent compound 1,3,3-
trlmethyl-1,
18,
and t h e
495 5-chIoro-1,3,3-trimettiy1
d e r i v a t i v e , and observed t h a t t h e t o l u e n e s o l u t i o n so f
these two compounds become b l u e upon i r r a d i a t i o n by u l t r a v i o l e t l i g h t a t 10
OC.
They attempted t o prepare t h e banzoxazine analogs from ortho-nitrosophenols
w i t h o u t success because of t h e i r i n s t a b i l i t y . Subsequently, o t h e r d e r i v a t i v e s such as 1,3,3,5-tetramethyI,
5-methoxy-1,3.3-trimethyI,
l-carboxyethyl-3,3
-dimethyi and l-cyanopropyl-3,3-dimel-hy1were r e p o r t e d t o be photochromic i n patents by Ono e t c i
(2, 31,
Ch!J and f i l e iaie Richard liovey ( 4 ) were f i r s t $0 recognize t h a t
.uand many
o t h e r spirooxazine d e r i v a t i v e s have a very low photodecompositionr a t e and thus, t h e i r pol-ep'rial commerciai appl ications, p s r t i c u l a r l yas I i y h t f i l t u r s .
They
expended +he synthesis of s p i r o o x a z i n scompot.indst o i n c l u d e d e r i v a t i v e s w i t h TM Subs~iPu~io i nn t h e naphthaierie r i n g ( 5 , 6). The i n t r o d u c t i o nof Orgaver sunglasses i n 1980 and PHOTOLITE TM p l a s t i c photochromic lenses i n 1982 made from these sp i r m x a z ine compotinds s t imuI ated considerabI e in t e r e s i from many i n d u s t r i a l l a b o r a t o r i e saround t h e w o r l d as evidenced from t h e p r o l i f i c numbers of p a t e n t s issued on d i f f e r e n t d e r i v a t i v e s o f
1
(7-10) and on
4 and d ( ! I ,
12)
and of t h e p l a s t i c photochromicslinglasses introducedt o t h e marketplace I n t h e past several years.
i n t h i s chapter, t h e photochromicp r o p e r t i e s o f v a r i o u s s p i r o o x a r i n e compounds having d i f f e r e n t - r i n g systems w i l l be reviewed.
Since most of t h e
data came from t h e p a t e n t l i t e r a t u r e , they tend t o be q u a l i t a t i v e i n nature. Up t o t h e present time, t h e major e f f o r t has been concentrated on t h e s y n t h e s i s of new d e r i v a t i v e s and on t h e p r a c t i c a l a p p l i c a t i o n s .
Fundamental research i n
t h e understandingo f t h e photophysical and photochemical processes as We1 I as f a t i g u e mechanism has been scarce.
The s i t u a t i o n may change as t h e p r a c t i c a l
a p p l i c a t i o n sof spirooxazine compoundsexpand i n scope and enjoy commercial success. The I i t e r a t u r e survey, b o t h s c l e n t i fi c and patent, has been performed t o cover t h e t i m e p e r i o d up t o September, 1987. 2 2.1
SPECTRAL PROPERT I ES P h o t o c h r a s m i n Solution Generally, spirooxazine compounds a r e n o t photochromic i n t h e s o l i d form,
b u t show photochromism
when d i s s o l v e d i n f l u i d s o l u t i o n s o r r i g i d
media such as gels, p l a s t i c resins, o r f i l m s .
Thus, a d i l u t e s o l u t i o n of
a in
t o l u e n e i s c o l o r l e s s and becomes i n t e n s e l y b l u e upon exposure t o u l t r a v i o l e t light. ceases.
The intense b l u e c o l o r disappears r a p i d l y when u l t r a v i o l e t i r r a d i a t i o n T h i s r e v e r s i b l e c o l o r change i s explained, as i n t h e case of
spiropyrans, by t h e t r a n s f o r m a t i o nof t h e tklosedll s p i r o s t r u c t u r e t o t h e "openvtmerocyanine dye s t r u c t u r e as i l l u s t r a t e d i n Scheme 2.
496
Scheme 2.
Photochromictransformationo f JA.
Typically, spirooxazinecompoundsare c o l o r l e s s or pale yellow solids. s o l u b i lI t y of spirooxazines i s r a t h e r low
The
i n alcohols and a l i p h a t l c
hydrocarbons, high i n aromatic hydrocarbons and intermediate i n ketones and halogenated hydrocarbons. Of course, t h e s o l u b i l i t ycan be a l t e r e d s i g n i f i c a n t l yby substituentssuch as sulfonate, carboxy o r long-chain a l k y l groups. The s o l u t i o n o f spircoxazine compounds i n a polar solvent normally shows b e t t e r c o l o r a t i o n upon i l l u m i n a t i o nby u l t r a v i o l e t l i g h t than i n a non-polar solvent.
I n p l a s t i c o r polymeric media, t h e p h o t o c o l o r a b i l i t yof spirooxazines
i s b e t t e r i n a s o f t host material. 2.2
-tion
srsstrum of t h e calorie- form
An ethanol s o l u t i o n of
lA displays three groups of
absorption bands i n t h e
u l t r a v i o l e t region of t h e spectrum as shown i n Fig. 1 (13).
The f i r s t group i n
t h e region above 280 nm has moderate absorption i n t e n s i t y and I s composed o f
t h r e e absorption peaks:
345 nm (€=4.76X103 M-l cm"),
cm-') and 297 nm (e=6.53X1O3 M-l
317 nm (€=6.78X103 M-l
The other two absorptiongroups i n
the f a r u l t r a v i o l e t region are very intense w i t h molar a b s o r p t i v i t y of 5.43X10 M-l cm-' and 4.41X103 M-l cm-',
4
respectively, f o r t h e absorptionmaxima a t 235
and 203 nm. Topologically, t h e s t r u c t u r e of JA can be separated I n t o two halves a t t h e s p i r o linkage and t h e two halves are e s s e n t i a l l y orthogonal. Thus, t h e absorption spectrum of JA may c o n s i s t of localized t r a n s i t i o n s belongingt o a p a r t i c u l a r h a l f o f The molecule w i t h l i t t l e i n t e r a c t i o n from t h e other half. I n comparing t h e absorptionspectrum of 1,3,3-trimethyl-spirolndolIno-
benzopyran w i t h those o f 1,3,3-trirnethyl-2-hydroxy-2-phenyI-indoline
2,2-diethyI-[2H]-chromene,
and
Tyler and Becker (14) assigned t h e two absorption
peaks of t h e spiropyran a t 312.5
and 324 nm t o t h e t r a n s i t i o n s e s s e n t i a l l y from
497
2.4
2.o
8 t
1.6
4
1.2
n. L (u
n
a
0.8 0.4
200
300
400
Wavelength(nm) Fig. 1. The absorption spectrum of t h e c o l o r l e s s form o f (3.26X10-5
M)
l@ i n ethanol
the benzopyran p a r t o f t h e molecule and t h e absorptfonmaxima a t 296 and 243.5 nm t o those from t h e l n d o l i n e p a r t of t h e molecule. Due t o t h e s t r u c t u r a l s i m i l a r i t y of spiropyranand spirooxazine, t h e two long wavelength absorption maxima a t 345 and 317 nm are I l k e l y t h e t r a n s i t i o n s e s s e n t i a l l y local ized I n t h e oxazine p a r t of t h e molecule, w h i l e t h e 297 and 235 nm absorptlon bands are from the indoline part. S u b s t i t u t i o nof hydrogens w i t h alkyls, alkoxys, o r halogens I n t h e indol ine p a r t of t h e molecule does not modify t h e absorptionspectrum much. However, a much larger spectral change is observed when t h e s u b s t i t u t l o nIs I n t h e oxazine p a r t of t h e molecule.
For example, as i l l u s t r a t e d i n Fig. 2, t h e two longest
wavelength peaks collapse i n t o one s i n g l e peak a t 331 rim f o r t h e 51-methoxyA carbomethoxy substltuent a t t h e 5' p o s i t i o n induces an even greater spectral s h i f t . Thus, t h e absorptionspectrum of t h e
1,3,3-trlmethyl
derivative.
51-carbomethoxy-1,3,3-trimethyl
d e r i v a t i v e shows a bathochromics h i f t of 29 nm
t o 360 nm as compared t o t h a t of t h e 5@-methoxy-1,3,3-trimethyl
derivatlve.
498
c
C
s:
2
0.6 0.4 0.2 300 350 400 Wavelength(nm)
Fig. 2. The absorption spectrum of t h e c o l o r l e s s form o f I:1,3,3,5-tetramethyi
-1, 2: 5 '-methoxy- 1,3,3-tr imethyI -1, and 3 :5 9-carbomethoxy-1,3,3-tr imethy1-1 in ethanol ( t h e absorbance scales f o r 2 and 3 are s h i f t e d upward by 0.2 and 0.4, respectively). The absorption spectra o f t h e c o l o r l e s s forms of 4 and
6
show a
bathochromics h i f t f o r t h e f i r s t absorption band as compared t o t h a t of .!J,
1to
2, produces t h e same . Figure 3 e f f e c t as t h a t of a methoxy group i n t h e naphthoxazlne p a r t of 1
Changing t h e indol ina p a r t of
benzoxazol Ine, I. e.,
i l l u s t r a t e s t h i s spectral s h i f t as t h e r i n g systems o f spirooxazinesare changed.
2.3
& s o r D t i o n soectrlrm o f the col-
I t i s extremely d i f f i c u l tt o obtain a t r u e absorption spectrum o f t h e colored form a t room temperature.
This i s because, f o r most spirooxazine
compounds i n solution, t h e colored form fades thermally a t a r a t e t h a t i s s i g n i f i c a n t compared t o t h e time required t o record t h e spectrum.
Furthermore,
because of t h i s r a p i d thermal fading rate, one cannot obtain a s o l u t i o n which contains the colored form only without t h e presence of t h e c o l o r l e s s form. However, t h e p o s i t i o n of t h e absorptionband of t h e colored form i n t h e v i s i b l e region p a r t i c u l a r l y i n a polar solvent, can be e a s i l y measured ( f o r techniques see Chapter 2 1.
This i s because t h e colored form e x i s t s i n thermal
499
1.0
0.8
z
t
2 0.6 L
0
cn
2.
0.4
0.2
300
350
400
Wavelength(nm) Fig. 3. The absorption spectrum of t h e c o l o r l e s s form o f 1:3,5-dimethyl-9'methoxy-2, 2:1,3,3,5-tetramethyI-6
-4 i n ethanol 0.8,
and 3:1,3,3,4,5-
and 1,3,3,5,6-pentamethyI
( t h e absorption scales f o r 2 and 3 are s h i f t e d upward by 0.4 and
respectively ) .
e q u i l i b r i u mw i t h t h e c o l o r l e s s form.
For many spirooxazine compounds, t h e
concentrationo f t h e colored form produced thermally i s s u f f i c i e n tt o impart c o l o r t o t h e s o l u t i o n and allow t h e absorption band, i n t h e v i s i b l e region, of the colored form t o be measured. Generally, the thermal fading r a t e of t h e colored form decreases as t h e temperature of t h e s o l u t i o n i s lowered, stopped a t low temperatures. temperature below -60
This thermal fading can be completely
For example, t h e thermal fading o f
The most promlnent absorption band of t h e colored form o f i n Fig. 4,
578 nm.
lA
stops a t
OC.
U, as i l l u s t r a t e d
i s i n t h e v i s i b l e region w i t h t h e absorptionmaxima a t 612 and
The molar a b s o r p t i v i t i e so f t h e peaks a t 612 and 578 nm are
8.1X104 M - l cm-' and 4.9X104 M-'
cm-',
respectively. Three additional
500
1
I
1
I
I
I
1 300
40 0
7
I
500
600
Wwelength(nrnq Fig. 4. The absorption spectrum of t h e colored form of
-75
OC
i n ethanol a t
(3.47X10-6 M I .
0.1 0
0.08 0.06
0.0 4 0.0 2 450
550 650 Wavelength(nm>
Fig. 5. The absorption spectrum of t h e colored form of 1:1,3,3,5-tetramethyI-l,
2:5~-carbomethoxy-l,3,3-trimethyl-~ and 3:5'-methoxy- 1,3,3-trimethyI-l
in
ethanol.
4 343 nm (c=2.7x1O4 M-l cm") absorption bands a t 410 nm (c=1.8XlO M-' cm-'1, 4 and 277 nm ( ~ 3 . 2 X 1 0 M-l cm") have also been measured. Alkyl and halogen substltuentshave very s l i g h t e f f e c t s on t h e absorption band of t h e colored form i n the v i s i b l e region as seen by s h i f t i n g i n t h e positionsof t h e absorption peaks by several nanometers (10, 15).
An alkoxy
substituenton e i t h e r side of t h e s p i r o linkage produces an additional absorption peak i n t h e spectral region between 450 t o 500 nm.
The 5I-carbo-
501
methoxy-1,3,3-trimethyI
d e r i v a t i v e induces a bathochromics h i f t o f t h e
absorption band by 20 nm (max. a t 630 nm), as i I l u s t r a t e d by Fig. 5. The p o l a r i t y o f t h e solvent has a strong e f f e c t on t h e absorptionspectrum o f t h e colored form i n t h e v i s i b l e region.
As t h e p o l a r i t y o f t h e solvent
decreases, a hypsochromics h i f t o f t h e absorptionband i s observed. example, t h e absorptionmaximuma t 612 nm o f and 555 nm f o r
u
For
i n ethanol i s s h i f t e d t o 590
JA i n toluene and i n cyclohexane, respectively.
Changing the indol ine p a r t of t h e molecule t o benzoxazoline s h i f t s t h e absorptiono f t h e colored form t o 578 nm from 612 nm. and 6 have s i m i l a r absorption spectra t o t h a t o f
0.1 0
U
c
The colored forms of
3
A
0.12
QI
1.
0.08
a
(II
0.04
0.02
500
700
600 Wavelength(nm)
Fig. 6. The absorption spectrum of t h e colored form of (6.85X10-3M)
a t d i f f e r e n t temperatures (0 t o
50C).
i n ethanol
4
502
3 THERMOCHROMISM OF SPIROOXAZINE COMPOUNDS A dilute solution of in ethanol M) is colorless. However, as the concentration of the solution increases, the solution becomes blulsh. The intensity of this blue color increases with concentration. Thls is because the existence of a thermal equllibrium between the ground state populations of the colorless and colored forms. Furthermore, at a fixed concentration the absorption intensity of the colored form Increases as the temperature of the solution Is increased. Thls phenomenon as illustrated in Fig. 6 for 18 is c m n among the the spirooxazlne compounds (7, 13, 16). In addition, spirooxazine compounds become highly colored when melted. The color Is generally purple or red. The thermal equillbrium between the ground state populatlons of the colorless and colored forms is influenced by the solvent polarlty as Well as the substituents on the molecule. Polar solvents such as ethanol and acetone promotes the formation of the colored form. Thus, polar solvents allow the absorption spectrum of the colored form to be measured at lower concentration than In non-polar solvents. Table 2 Effect of substltuent on the absorption Intensity of the colored form Concentration M)
i
Substituent*
++
Absorbance
at A max
H (i.e., lA) 5-Methy I 5'-Methoxy 5-Methyl-5I-Methoxy 5-Trifluoromethyl-5~-Methoxy 5-MethyI-5'-carbomethoxy 91-Methoxy
6.85 1.59 1.28 1.15 2.87 1.14 4.46
0.049 0.031 0.087 0.195 0.016 0.069 0.038
4,5- b 5,6-Dimethyl-4 5-Methyl-fi
3.69 5.82
1.118 0.208
*The first seven compounds are derivatives of 1. A1 I compounds have 1,3,3 -trimethyl substltutlon besides the substltuent(s) listed. + In ethanol *Measured at 20 O C
503 Alkyl and alkoxy substituentspromote formation o f t h e colored form w h i l e halogen and t r i f l u o r o m e t h y l substituentsdepress i t s formation.
This
substituent e f f e c t i s most profound when t h e s u b s t i t u t i o ni s a t t h e 5' position. As shown i n Table 2, absorption i n t e n s i t y of t h e colored form of t h e 1,3,3,5-tetramethyl
d e r i v a t i v e i s slightly higher than
a a t t h e same
s o l u t i o n temperature and concentration. However, the absorption i n t e n s i t y of t h e colored form of t h e 5-methoxy-l,3,3,5-tetr-amethyi almost 5-fold over t h a t of t h e 1,3,3,5-tetramethyl
d e r i v a t i v e increases derivative.
On t h e other
hand, t h e absorption i n t e n s i t y of t h e colored form of t h e 5-trifluoromethyl-5'methoxy-1,3,3-trimethyl
d e r i v a t i v e i s weak (absorbance 0.016)
a t 2.87X10-3
M.
The 5t-carbomethoxy substituent has t h e same e f f e c t of promotingt h e formation o f t h e colored form as t h e 5'-methoxy substituent, but t o a lesser degree. Higher a l k y l substituentsa t t h e 1 and 3 p o s i t i o n senhance t h e thermochromiSm Of
derivatives of
1 (7,
13, 16).
Derlvatives of
4 and 6
promote
10.0
0.1 0
0.01
3.2 3.4 3.6 3.8
-xT1
3
10
Fig. 7. P l o t s of t h e absorbance of the thermal e q u i l i b r a t e d colored form and the thermal decay r a t e of
fi versus
1/T.
504
thermal formation of the colored form more than the corresponding derivatives of 1 (especially derivatives of 4). Solutions of 4 are normally highly colored. Measurements of the absorption intensity of the colored form at several
temperatures provide data for calculating the standard enthalpy of the thermal equilibrium reaction between the colorless and the colored forms. According to the following equation:
a plot of the absorbance of the colored form against 1/T should result In a straight line as shown in Fig. 7 for
u.From
the least-square fit of the line,
the standard enthalphy of the thermal equl I lbrlum can be obtained by using eqn. Ell. Table 3 lists the values of the standard enthalphy of the thermal
equilibrium between the colorless and colored forms for some splrooxazine
Table 3
t
The kinetic and thermodynamic parameters of spirooxazine compounds
Ea(cm-l)
Compound*
H(i.e., 5-Methy I
5I-Methoxy 5-Methyl-5l-Methoxy
5-TrifIuoromethyI-5~-Methoxy 5-Methyl-5'-Carbomethoxy
9'-Methoxy
(7k3)X10-2 (1.9kO.1 )X103 3X 1 Oe2 1 .6X103 13X10-2 1ox10-2
2x10-2 15x10-2
5-Methyl-6
1 .3X103 1 .ox1o3
1 .5X103 1 .ox1o3 2.2~10~
(7.&0.4)XlO 10.0~1o3
3
7.7X 1 O3 10.4X 1 O3 9.6x1O3
5.9X103
i .3x103
*The first seven compounds are derivatives of 1. A l l compounds have 1,3,3trimethyl substitution besides the substituent(s1 listed.
+ In
ethanol and measured at 20 'C.
505
derivatives.
The p o l a r ! t y o f t h e s o l v e n t bas a l a r g e e f f e c t on t h e standard
As t h o p o l a r i t y of t h e s o l v e n t decreases, t h e v a l u e of t h o standard
enthalpy.
enthalpy insre.+ses.
For example, t h e enthalphy i s 2.4X103
for
Crn-'
cm-' i n ethanol. cyclohexane as compared t o ~l.9~0.1~X103 KINETICS OF THE THERMAL
4
JA i n
FADING
The colored form genereked by i t I m i n a t i o n w i t h u l t r a v i o l e t i i g l r t decays ?-+errnallyt o ?he c o l o r l e s s fnrm r e a d l l v at ambient tempemturn,
Gsnarally,
.+his -?'hermi F i , d i r \ p a ~ ~ a mRr t~. l~r :ij t tJr@er l i i r w t i c s i n soiuSlon a n d
*.>c~npIlcatei decz?; k.Ir!atics ii; 20;
poiymeric host rnaterle!
R
mere
Tho decay k i n e t i c s iri
u t i o n may r'eviate .iron; f i r c 6 Order a t Inwar* temperatvrss, f c ~ iex amp!^,
T=lOeC.
The teaiperatura a t which t h e deviation begins war ! e c xmewhal far each
spircmxazlne d e r i v a t i v e . By measuring t h e thermal decay r a t e o f the c o l o r e d form a t several
+emperaturos, t h e a c t i v a t i o n energy, Ea2 f w t h e decay t r a n s i t i o n o f t h e c o l o r e d form t o t h e colarle5.s Form can be c a l c u l a t e d by t h e Arrhenius equation:
n f=qii,
[?I
P 5s +he Li-wluwtcy f a c t o r and I(
a c t i v a t i o n t-wrgfez
Ey3
i s t h e thermal decay r s t s ,
The
of some spirooxazine d e r i v a t i v e s s a l c b l a t e d from a
Isasl-square f i t of t h e p l o t of t h e tnermal decay r a t e , k, versus 1/T a r e g i v e n i n Table 3.
5
PHOTOGHEMI CAL PROPEW IES I n c o n t r a s r t o s o i r a p v r a r s , sorne q)ironxa;l ine derivative; s r e except'ona!f y
s t s b l e towards u l + r a v i o l e t I i q h t I r r a d i a t i o n .
T h i s i s s u r p r i s i n gs i n c e +he
chemical s t r u c t u r e and many photophysical p r o p e r t i e s of spiropyran and spii'ooxazine a r e s i m i l a r .
However, t h i s exce: l e n t p h c t n s t a b l il-y l i s not
u n i v e r s a l t o a l l s p i r o o v a z i n ud e r i v a t i v e s . r u b s t l t u e n t q:an
Changing tha p : n g ys+erns or
a l t e r photostah11i t y d r a s t i c a lly.
For instance, repIacemsiit
of t h e i n d o l i n e r i n g w i t h t b e henzoxarolioe r i n g decreases the p h o t o s t a b i l l t y
of t h e s p l r o o m z i n e coapoirnd by, a t least, a f a c t o r of cne hundred (100) times. The photodecompasi+ionr a t s of i I l u m i n a t i o n i s 2X10-'
oyrans.
.u i n ethanol by u s i n g t h e 366
om Hg 1 ine f o r
which I s s i g n i f i c a n t l ysmaller than t h a t o f s p i r w
The 1 3 t t e r o f i e r show values I n the tenths,
Furthermore. t h e
photodecomposltionproducts do n o t I n t e r f e r e w i t h t h e photochromicf u n c t i o n of t h e remaining spirooxazine compound. T h i s is very important f o r p r a c t i c a l
a p p l i c a t i o n s s i n c e t h i s increases t h e u s e f u l l i f e o f a product s i g n i f l c a n t l y . The quantum y i e l d s o f t h e formation o f t h e c o l o r e d form of
M. i n ethanol
is
high(>0.90).
This high e f f i c i e n c y of photocolorationis consistentw i t h t h e
observationt h a t t h e ethanol s o l u t i o n o f 1pI prepared f r e s h l y from a well
p u r i f i e d sample does not show any detectable fluorescenceo r phosphorescence. The e l e c t r o n i c s t a t e from which t h e colored form o f JA i s derived i s
It i s
believed t o be t h e lowest excited s i n g l e t s t a t e of t h e c o l o r l e s s from.
l i k e l y t h a t t h e t r i p l e t s t a t e i s not involved i n t h e formation of t h e colored form.
I n photosensitizationof
JA i n a c e t o n i t r i l e by benzophenone, one cannot
detect any formation of t h e colored form. decompositionof
i s observed.
Instead, a r a p i d photosensitized
Thus, t h e e x c e l l e n t p h o t o s t a b i l i t yo f
U
may
be t h e r e s u l t o f a high e f f i c i e n c y of transformationt o t h e colored form from t h e excited s i n g l e t s t a t e and, thus, preventingt h e intersystemcrossing t o t h e t r i p l e t states from which photodecompositiontakes place. 6
SYNTHETIC METHODS
The most ccinmonmethod employed f o r the synthesis of spirooxazines i s t h e condensationo f an alkylldene heterocyclew i t h an ortho-nitroso aromatic alcohol I n a polar solvent such as methanol o r ethanol as depicted i n Scheme 3. The alkylidene heterocycles are generated from t h e correspondingquarternary s a l t s of nitrogen containingheterocycles. Typically, t h e y i e l d of t h e condensation r e a c t i o n i s about 30 t o 50 percent f o r those spirooxazlne compoundswhich are thermally stable.
An improvement I n y i e l d t o 70
- 75
percent has been reported when t h e alkylldene heterocycles, f o r example,
Scheme 3.
General synthetic scheme f o r splrooxazines.
substituted5-methyleneindoline derivatives, are generated from t h e corresponding indoleniniumiodides i n an a l k a l i n e aqueous s o l u t i o n and extracted w i t h a non-polar a l i p h a t i c hydrocarbon(17).
The alkylidene
heterocycle s o l u t i o n I n a non-polar solvent i s reacted w i t h an o r t h o - n i t r o s o aromatic alcohol i n methanol o r ethanol.
The water formed I n t h e condensation
r e a c t i o n i s removed as an azeotrope by vacuum d i s t i l l a t i o n . The o r t h o - n i t r o s oaromatic alcohols from which t h e s p i r m x a z i n ecompounds
507
have been successfully synthesizedare l-nitroso-2-naphtho1, 5-nitroso-6q u i n o l i n o l and 9-nitroso-10-phenanthrol.
No spirooxazine compoundsmade from
o r t h o - n i t r o s ophenol, 5-nitroso-5-quinolinol o r other ortho-nitroso aromatic alcohols have been reported.
So are t h e t h i o l analogues.
The quarternary lndoleniniumiodide s a l t s can be prepared by various synthetic methods f o r t h e indole r i n g system (18).
The most common procedures
used are t h e Fischer and t h e B i s c h l e r methods. Numerous quaternary indolenium iodide s a l t s have been successfully prepared. Synthesis of 2-naphthol and i t s d e r i v a t i v e s i s we1 I documentedsince they are used extensively as intermediates f o r manufacturingorganic dyes (19). Skraup method i s used f o r t h e synthesis of q u i n o l i n o l s (20).
The
The aromatic
alcohols are n i t r o s a t e dw i t h sodium n i t r i t e i n t h e aqueous s o l u t i o n t o obtain t h e ortho-nitroso d e r i v a t i v e s (21). The number of t h e spirooxazine compoundsknown i s f a r less than t h e correspondingspiropyrancompoundsbecause of d i f f i c u l t y i n synthesizingthe d e r i v a t i v e s of 1-nitroso-2-naphthol and the i n s t a b i l i t yof many spirooxazine compounds. Many condensationr e a c t i o n mixtures show photochromiceffects, b u t no spirooxazine compoundscan be Isolated and p u r i f i e d f o r characterization. Table 4 Additional parent r i n g systems o f known spirooxazines
7
508 7 ADDENDA
Since completion o f t h e manuscript, t h r e e a d d i t i o n a l spirooxazineso f
d i f f e r e n t parent r i n g systems o t h e r than those l i s t e d i n Table 1 have come t o t h e a u t h o r f s a t t e n t i o n . They a r e i l l u s t r a t e d i n Table 4. Compounds o f r i n g system 2 were prepared by r e a c t i n g p i p e r i d i n i u mi o d i d e s a l t s w i t h 1-nitroso-2-naphthol o r i t s s u b s t i t u t e d d e r i v a t i v e s i n anhydrous ethanoi i n t h e presence of t r i e t h y l a m i n e (22).
has an absorption max!mum a t 560 nm i n methanol.
580 nm !n toluene.
The 1,3,.3-trimethyl
derivative
The absorption peak s h i f t s t o
T h i s solvatochromics h i f t is o p p o s i t e t o t h a t of
shows a hypsochromics h i f t as t h e p o l a r i t y of t h e s o l v e n t decreases.
18. which
Preparations o f r i n g system 8 and 9 were r e p o r t e d by Yamamoto and Taniguchi and by Tanaka and Kida, r e s p e c t i v e l y (23, 24).
They a r e prepared by t h e
general s y n t h e t i c methods as discussed i n t h e preceding section.
The r e a c t i o n
y i e l d Is g e n e r a l l y smaller than t h e correspondingnaphthooxazines. Extending
t h e c o n j u g a t i o nfrom naphthalene t o anrhracene does n o t s h i f t t h e absorption maximumof t h e c o l o r e d form o f t h e d e r i v a t i v e s of 9. REFERENCES 1
R.E. Fox, Research Reports and T e s t Items P e r t a i n i n g t o Eye P r o t e c t i o no f
A i r Crew Personnel, F i n a l Rept. on Contract AF 41(657)-215, A p r i l , 1961, AD
440,226. 2 H. Ono and T. Osada, Photochromiccompound and cornposftionc o n t a i n i n g t h s same, U.S. Patent 3,562,172 (1971). 3 H. Ono, T. Osada and K. Kosuge, PhotochromicComPound, U.S. Patont 3,576,602 (1971). 4 N.Y.C. Chu and R.J. Hovey, unpublishedr e s u l t s . 5 R.J. Hovey, N.Y.C. Chu, P.G. Pigs2 and C.H. Fuchsman, Photochromic compounds, U.S. Patent 4,215,010 (1980). 6 R.J. Hovey, N.Y.C. Chu, P.G. Plusz and C.H. Fuchsman, Photochromic compounds, U.S. Patent 4.342.658 :1982). 7 M. M e l z i g and G. Martinuzzi, Photochromicsubstances, PCT I n t . Appl. WO 85 02,619 (1985). 8 S. Yamamoto and T. Tanigucni, Spirooxazine compounds and a photcchromlc shaped a r t i c l e , Eur. Pat. Appl. EP 171,909 (1986). 9 M. Hosoda, Photochromiccompounds, Eur. Pat. Appl. EP 186,364 (1986). 10 M. Ir i e and S. Maeda, 1 ,3-Dimethy Is p ir o cindol ir1e-Z,3~-naphth[Z, 1-uEl14I oxazine] d e r i v a t i v e s , Jpn. Kokai Tokkyo Koho JP 61,186,390 (1986). 1 1 M. Reichenbaecher, U. Grumt, R. Paetzold and J. Epperlein, Ger. (East) DD 153,690 (1982). 12 W. S. Kwak and R.J. Hurditch, Photochromiccompounds, Eur. Pat. Appl. EP 141,407 (1965). 13 N.Y.C. Chu, Can. J. Chem. 61 (1983) 300. 14 N.W. Tyer, Jr. and R.S. Becker, J. Am. Chem. Soc., 92 (1970) 1289. 15 N.Y.C. Chu, Photochromismof spiroindollnonaphthoxazine. 1 1 . Substituent e f f e c t s , in: Book of Abstracts, 194th ACS National Meeting, New Orleans, Louisiana, August 30-September 4, 1987. 16 S. Yamamoto and T. Taniguchi, Spirooxazine photochromico r thermochromic imaging system, Jpn. Kokai Tokkyo Koho JP 61 161,286 and 61 161,287 (1986) 17 C.H. Hoelscher and D.S. McBaIn, Method f o r preparing s p i r o ( i n d o 1 i n e ) - t y p e photochromiccompounds, U.S. Patent 4,634,767 (1987). 18 R.K. Brown, Synthesis of t h e i n d o l e nucleus, in: W.J. Houlihan(Ed.),
509
Indoles, Part 1, Wiley-lnterscience, New York 1972, pp. 227. 19 K. Venkataraman, The Chemistry of Synthetic Dyes, Vol. 1, Academic Press, New York, 1952. 20 H.T. Clarke and A.W. Davis, QuinolIne, in: H. Gilman and A.H. Blatt(Ed.), Organic Syntheses ColI. Vol. 1, John Wiley, New York, 1964, pp. 478-482. 21 C.S. Marvel and P.K. Porter, I-Nitroso-2-naphthol, in: H. Gilman and A.H.Blatt (Ed.), Organic Syntheses C o l I . Vol. 1, John Wlley, New York, 1964 pp. 411-413. 22 S. Kawauchi, S. Saeda and H. Yoshida, Spiropiperidinenaphthooxazine compounds, PCT Int. Appl. WO 87 03,874 (1987). 23 S. Yamamoto and T. Taniguchi, New spirooxazine compounds, Jpn. Kokai Tokkyo Koho JP 62 33,184 (1987). 24 T. Tanaka and Y. Kida, Photochromicmaterials, Jpn. Kokai Tokkyo Koho JP 62 72,778 !1987).
See Additional Literature (1989 See Additional Literature (1989
- 2001): Spirooxazines, A89 - 2001): Spiroxazines, A97
This family is referred to in the literature with two spellings: spiro-oxazines and spiroxazines. The references for these two terms are therefore separated accordingly.
510
Chapter 1 1
Qn and 4n+2 Systems (n2 2) Based on
l , 7 - and 4,lO-Electrocyclization H. Diirr
1.INTRODUCTION
The high potentialof electrocyclicreactionsin 4n and 4n + 2- systems(n= 1) has been extendedonly to a very limiteddegree to systemswith n =. 1. One problemmightbe the less we!! defined conforrna:ions~f ?he precursorfor an electrocyclicprocess.It appears tha! photachernistr,have PG exploitedthisfieldfurther. A few examplesa% !mwr?which michtbe used10 create photochromicsystems.Most of these reactionsare simple conversion8whichhave been provedto be reversible. 2. 43- SYSTEMS (n = 2): 1,7- Electrocyclizations The simplestmolecule studiedwas a cyclichexadienylanion1. This Bn-systemsmoothly undergoesa reversible1,7-electrocyclization. Accordingto the Woodward-Hoffmann rules a disrotatorymode is followed(ref. 1).
ratio: 1 : 2 = 113
On irradiationof cyclooctatrienone 3 a k k e t e n e4 is formedwhichcyclizesalready at - 80' in a very fast reaction to 3. 4 can be trappedwithsolventwhen the photolysisis carriedout in the presenceof methanol(ref. 2). Most of the other reportedelectrocyclizations have been carriedout onlythermallyso that no other moleculeshave been in studiedin detail photochemically. A diaza-system5 undergoesan electrocyclicringclosureto 6 and 7 in a 1,B- and 1 3 electrocyclization. The reversibilityof this reactionis an open problem(ref. 3). C H z N =N
hr
&-Ph S
____+
5
N 11
+6 ,Ph
7
511
3. 4n 2-SYSTEMS (n22): 1,lO-Electrocyclizations Photoinduced1,s-electrocyclizations beingreversibleseem not to be known. The II1O-electrocyclization on the other hand has been shownto lead to molecules whichmightbe used in photochromic systems. Suitabledihydrouulenes,whichdo not aromatizeimmediately,have been foundin the conversion8 , 9 (ref.4).
+
H
9
8
9’
A very efficientphotochromicmoleccjlewas reportedon recentlyin the reaction10, 11
11
10
R: NO2, COzMe, Br, H,
Me, OMe, NMe2, NH2
The 10T-electrocyclization of 11 shouldfollowa disrotatorymode. The basicstructural requirementsare obsiouslya) the presenceof a phenylgroup,b) electrondonorsR and c) electronacceptorsin the o-positionsto stabilizea negativechargein a dipolarresonancestructure. Irradiationof dihydroazulenes10 gives the deep red coloredspecies11. A typicalspectrum of 10 and 11 is shownin Fig. 1.
512
I
200
3 00
LOO
Fig.1 Photochromism of 10 in acetonitrile(c= 4.9 (---) start, (...) after 1 rnin, (-) after 7 min.
500
I
Xlnml 600
lo-’ mol/l), irradiationby daylight.
The coloredmoleculesdependingon substitutionabsorbin the regionbetween456 483 nm (lg 4) providingred coloredsolutions.Changingthe solventresultsin a positive solvatochromic shiftof the longwavelengthband of the coloredform. Photochromism of 10 was observedin solutionand in the solidstate. The thermalelectrocyclization of 11 to 10 is rather slowand needs severaldays at roomtemperature. The conversion10, 11 showsseveralcyclesof colorizationand decolorization.On longerirradiationtimessecondaryprocessescan occurin some of the substitutedderivativesof 10 (R = NOn), whereasthe derivatives(R = N((CH3)2) 10 and 11 are rather photostable. The thermalcyclizationof 11, producesa mixtureof diastereomers10. The stereochemistry of the electrocyclization has not been establishedyet. Howeverthe lossof chirality by irradiatingone of the diastereomerswas demonstrated. The photochemicalstep in the ringopeningwas investigatedfurther.A singletstate may responsiblefor the reaction10, 11. Higher r-systems were studiedto effect a photochemicalelectrocyclicprocess. Whereas the thermalreactionclearlyfollows: the Woodward-Hoffmannrulesno photochemicalreactionhas been observedso far (ref. 6).
513 12 w
14%
0 ‘I
6
12
hu: no reaction inert:direct sens.
n3
reaction inert: direct sens.
Summarv: Photochemicalelectrocyclicreactionsto give reversiblephotochromic systemsinvolvingmorethan &-electrons are ratherscarce. The limitednumberof examplesstudiedallowsno clear conclusionas to the potentialof this processfor designinga photochromic system. More work oughtto be done in this area.
References 1 S.W. Staley, N.J. Pearl, J. Am. Chem. SOC.,95 (1973)273. 2 L.L. Barber,O.L. Chapman,J.D.Lassila,J. Amer. Chem. SOC.,91 (1969)531. 3 D.P. Munroe,J.I. Sharp, J. Chem. SOC. PerkinTrans I, 1980,1718. 4 H. Prinzbach,H.J. Herr, Angew. Chem., 84 (1972)117;Int. Ed. 11 (1972)135. 5 J. Daub. T. Knochel,A. Mannschreck,Angew. Chem. , 96 (1984)980. Int. Ed. 23 (1984)960. J. Daub, S.Gierisch,U. Klement,T. Knochel,G. Maas, U. Seitz, Chem. Ber.,
119 (1986)2631. 6 H.Prinzbach,L.Knothe,Pure Appl. Chem., 58 (1986)25.
514
Chapter 72
Cycloaddition Reactions Involving 4n Electrons: (2+2) Cycloaddition; Photochemical Energy Storage Systems Based on ReversibleValence Photoisomerization G. Jones 11
1
INTRODUCTION
The cycloaddition of two w
bonds to yield a cyclobutane ring holds an
important place in the annals of photochemistry. This potentially reversible
transformation has become a celebrated example of a reaction which is kinetically inhibited as a thermally induced process due to orbital symmetry
or topology constraints (ref. 1). The kinetic stability of photogenerated small rings is an important feature insuring the synthetic viability of many strained ring systems and the potential for photochemical energy storage. The reduction in conjugation on photolysis of photochromism as well.
a substituted olefin leads to
The valence isomerization of norbornadiene
(1
+
z)
is the prototype
-
reversible (intramolecular)(2 + 2) cycloaddition. The first example of this Q, R = H) was important rearrangement involving the diacid derivative (3 reported by Cristol and Snell in 1954 (ref. 2), who named the small ring
product a derivative of
"quadricyclene" due to its olefin-like behavior.
Dauben and Cargill (ref. 3) followed with the report of photoisomerisation of noting also that isomer 2 is reverted to norbornadiene on thermolysis at
-1,
20O0.
Interestingly, the photoisomerisation of
1
is not
observed on uv
photolysis in the gas phase, in favor of fragmentation yielding (primarily)
cyclopentadiene and acetylene (ref. 4). That the rearrangement of 1 could be induced by widely different modes of photolysis was demonstrated early by Bammond and his coworkers (ref. S), qho showed that light absorption by ketone sensitizers is effective in driving isomerization under conditions in which
transfer of triplet energy to the diene is important. The norbornadiene rearrangement has achieved considerable generality] including examples displaying a variety of substituent groups (e.g., -CN, F) (ref. 6) which control to a degree the wavelengths of light suitable for
515
1
e
m
2
Z
hv
R
t
&co2R 4 CO2R
CO2R
3
5
hv b
O
R
6
0
8
0
R7
hv
9
516
photolysis and the level of
thermal stability in the quadricyclene product.
Among the more unusual structures which have
capable of "double isomerization'
moiety
@-I)
(ref. 7).
been studied is
5, which
is
The internal addition of an enone
has received some notoriety in terms of a critical ring forming
step in the cubane synthesis (ref. 8).
shift) which occurs on
The subtle photochromism (36 nm blue-
irradiation of
mechanisms for direct photolysis
s
was
identified early, along with
and sensitized rearrangement in this series
(ref. 9). Of historical importance is the first known isomerization of this type involving the transformation of carvone to carvone camphor @-QJ due to Ciamician and Silber (ref. 10).
from(2 + 2)addition as shown by
Access to other ring systems has resulted the internal photoaddition of myrcene (10-
(ref. 11) and the interconversion of lJ3-butadienes and A valuable review of early results regarding bicyclo [l 1 .O] butanes (ref. 12)
11)
.
.
intramolecular(2 Most (2
+
+
2)addition w a s provided by Dilling (ref. 13).
2) cycloadditions involving chroaophores which absorb in the
near ultraviolet are only modestly
photochromic.
Shifts generally smaller
than 50 nm are typical and most often associated with the reduction in chromophore length by one double bond. Further examples are provided by the interconversions of stilbenes
(12-g)(shifts
in
Amax,
(styrenes) and phenyl substituted cyclobutanes
300
- 260
nm)
(ref.
14).
The Paterno-Buchi
photocycloaddition of carbonyl compounds and alkenes constitutes yet another family of (2 + 2) additions displaying mildly photochromic behavior. In rare
cases thermal or metal catalyzed reversion of oxetane products (e.g., J 4 l have been studied (ref. 15). Other interesting chromophores participating in the
Paterno-Buchi reaction include phenones , quinones, and a-dicarbonyl compounds which undergo bleaching of absorption bands in the 350-450 nm range (ref. 16). The thione group is probably the most interesting in this regard. In several studies carried out by de Hayo (ref. 17), it was shown that the impressive
blue absorption (SOO-600 nm) of thiones is bleached on photolysis with alkenes leading to thietane products (e.g., l5). systems have not been studied in detail.
Reversions for these interesting
Attention w a s refocused in the 1970's on (2 + 2) additions of the norbornadiene type due to the potential of these systems for latent heat storage of radiant (solar) energy. Criteria for photochemical energy storage in this context were discussed by Jones and Ramachandran (ref. 18) in
connection with the reversible isomerization of Ifl. Later reviews summarized work on norbornadienes and other energy storing isomerizations (ref. lQ), and a number of preliminary feasibility studies regarding large scale systems have Recent papers (ref. 21, 22) have directed been carried out (ref. 20). attention to procedures for the optimization of the absorption and energy
5:
517
11
10
i
hv
hu
2
.
9
9
II
II
16
I
?I
hv
‘9
9
13
12
0
>i
R
R’ 14
17
518
storage capabi,-,Aes
The technique of
of norbornadienes
closely related polycyclic dienes.
or
photocalorimetry
has
been
review will
focus
used
to measure enthalpies of
isomerization of norbornadienes which are large, falling in the range of 14-25 kcal/mol (ref. 23). The present
photoreactions of the (2
+
2) type
encountered
(excited
take place
triplets,
the theme of
mechanisms
of
the photoreversion) which have some
The impressive variety of conditions in
potential for photochromic activity.
which photoisomerization will
on
(and
and the array of intermediates
exciplexes,
radical-ions,
donor-acceptor
complexes) have made these reversible systems a rich field for investigation.
2 2.1
(2
+
(2
2)ADDITION:
DIRECT AND TRIPLET SENSITIZED PBOTOREARRANGEY&NT
+ 2)Photoaddition of substituted norbornadienes The system which has received most attention in terms of intramolecular
addition of double bonds
Kaupp
and Prinzbach
is norbornadiene
(ref. 24) were
and
its derivatives (e.g., l-,3J.
the first to conduct
a survey of
norbornadiene derivatives which display significant absorption in the near uv.
The course of isomerisation of 18 (R = CgH5, R ’ = CH3) on photolysis of dilute solutions near 300 nm is shown in Fig. 1. Similar results were obtained for other substituent patterns with generally
high
quantum yield
for direct
photolysis (0 = 0.4-0.6). The efficient photolysis of norbornadienes is understood in part in terms of the proximity of double bonds which results in weak transannular interaction and alterations in properties such as uv A variety of norbornadienes absorption and ionization potential (ref. 25). have been produced in an effort
to enhance absorption by the diene moiety in
the visible. Yore absorbing derivatives 18 have included carboxylic acids which have the added feature of water solubility (ref. 26). Yoshida,
Yonemitsu and their coworkers (ref. 21, 22c, 27) have proposed use of electron or 20. With donor and acceptor groups for elaboration of derivatives as in this pattern of substitution, norbornadiene absorption is substantially red-
shifted (&,== 375-400 nm and weak absorption to 500 nm), the resultant photoisomers are less thermally stable, but quantum efficiencies for rearrangement can remain high (e.g., @ = 0.75 for Is, Ar = C6H5, B = C02Ye, C O 2 W , ref. 22c). The(2 + 2)photoaddition which is induced on excitation of an ancillary photosensitizer (S) (ref. (Scheme 1) (ref. 28).
On
5,
11)
requires
excitation
a mechanism
of energy transfer
of the photosensitizer, S, a relatively
long lived triplet excited state (3S*) is produced in high yield via intersystem crossing. Transfer of triplet energy to norbornadiene (N) follows in a bimolecular encounter, and excited triplet N partitions to isomer Q or
519
Figure 1.
Absorption of
of irradiation
time
5 (R
(300 nm,
= C&j1
R' = CE3) (Amax = 290 nm) as a function
acetonitrile
permission of Belvetica Chim. Acta)
Scheme 1
so
-
3S
+
3N
1s*
N
-
-
-.
3s* So
+
intersystem crossing
3N
triplet energy transfer triplet isomerization
N. The triplet energy required to drive the reaction of
kiallmol ,according to studies by Hautala
These
(ref. 24) (with the
sensitizer excitation and
Q (or N)
back to
solution)
included
ketones
immobilized
interfacial or heterogeneous (ref. 19c, 29) showed
triplet energies as
that
low
energy
3
could
on
as 46 kcal/mol,
visible absorption (e.g. , acridine
mechanism for employment of
polymer
transfer. be
ca. 70
supports which
allowed
Other sensitization studies
isomerized using sensitizers having
orange).
sensitizers
1 is
(ref 19a) using various sensitizers.
including
having
An
low
dyes displaying strong
important
feature of the
triplet energies is the
520
under circumstances in which step 2
temperature dependence of energy transfer
10 kcal/mol) giving rise to the concept
(Scheme 1) is endothermic (as much as
of "thermal upconversionn of isomerization quantum yields (ref. 19c,29,30). potentially complicating
valence isomers
feature
is the
photoaddition
sensitizers (ref. 31).
2.2
involving
irreversible
photolysis
A
of the
Q) observed for certain ketone
(to
Features of mechanism for direct and triplet sensitized isomerization
32) which connect N and Q isomers 2) addition. The model of Turro (ref. 28)
The potential energy surfaces (ref.
have become a prototype for (2 + (Fig. 2a) based on the
geometries of biradical species consists of
predicted
unsymmetrical surfaces for 91 and T1 states.
The partitioning which occurs at
manifold (sensitized photolysis)
efficient
(N
biradical geometries results in efficient isomerization and
less
singlet state (direct irradiation, observed
displaced
singlet
substituents in a
and
triplet
for
surfaces
subtle way.
acceptor substituted norbornadienes
The
has
is
efficient been
coworkers (ref. 33) with reference to the
1).
likely
8)
+
in the triplet
rearrangement for the
This partitioning along to
be
dependent
on
direct photolysis of donor-
rationalized
by Kutal and his
diagrams of Fig. 2. In their model
the singlet minimum will favor a more closed biradical geometry due to dipolar
contributions to biradical
structure.
observed partitioning
in
nature
wavelength
sensitized) photolysis of of
the
long
favor
of
20.
the
This
model
photoisomer
?r bond
is
consistent with the
for
(direct, but not
interaction and the charge-transfer
electronic
transition
for
donor-acceptor
substituted norbornadienes have been analyzed by Kobayashi (ref. 34). Although both singlet and
are expected to be
very
triplet
short
bonding, certain substitution
excited
lived
due
patterns
interrogated by flash kinetic
to
provide
techniques.
In rationalizing
the
phototransients
Thus,
convenient naphthalene chromophore, a triplet observed (ref. 35).
states of I_ and derivatives
the facility of transannular
of
3 which
for
lifetime,
7
that may be
displays the
= 5.6 p e c , was
flash results, a triplet common to
isomerization in both directions was proposed, as well as a dominant adiabatic
mechanism (excited state -excited returns the system to
triplet rearrangement
21.
S- 17
A
short-lived (ca. 0.1 nsec).
quantum yields that
triplet
2
22
and
formation
but partitioning
efficiency of isomerization.
36).
(ref. It
activation energy (4.7 kcal/mol)
biradical intermediate
state) for the triplet rearrangement which
quenching
of
was
the
the
from
provided
rate data for the
The triplet state of
possible
for the
show
of
study
rapid the
in
ring
indeed
closure leading to a
temperature
intermediate
16 is
this study to measure an
limits
independence of
the lifetime of
intermediate governs the overall quantum
521
so-
Du
NBD
-
Q
DQ
Figure 2. Postulated potential of norbornadiene
(1.
.-.
2)
the J. Amer. Chem. SOC.)
18
(a)
energy surfaces for the valence isomerization
and
20 (b)
(ref. 34) (with the permission of
20
19
21
22
522
3 3.1
+ 2)ADDITION: PATHWAYS INVOLVING ELECTRON TRANSFER OR EXCITED COMPLEXES
(2
Photorearrangement via radical-ions or exciplexes
Mechanisms of photosensitization which involve an electron transfer step
(EDA) interaction are well known (ref. in bimolecular(2 + 2)photoadditions has been The role of EDA exciplexes documented thoroughly in papers by Caldwell and Creed (ref. 42). For valence or bimolecular electron donor-acceptor
37).
isomerizations of the (2
+
2)type, mechanisms employing exciplexes, excited
EDA
complexes or radical ion intermediates have been proposed. The rearrangement of 23 has been studied in most detail. The important initial finding is that isomerization to 24 can be induced through bimolecular quenching of the singlet state of various sensitizers capable of acting as electron donors. In the table quantum yields are shown for rearrangement in polar (ref. 38) and
non-polar (ref. 39) solvents. Electron donors were selected in these studies for which it could be calculated, according to the Rehm-Weller formalism (ref.
40) that electron transfer between the moderately good acceptor 23 [B(1/2),,d V vs SCE, CE3CN, ref. 381 and the sensitizer is energetically favorable. Under these circumstances, the quenching of sensitizer
= -1.67
fluorescence is efficient (near diffusion limited rate constants, kq = 1-10 x
lo9 Y-l sec-I), and moderate concentrations of quencher 23 insure that a singlet sensitization mechanism must be operative. Although the pattern of quantum yield data (Tab. 1) showing a dependence
on sensitizer triplet enerm for the singlet donor sensitizers appears uniform (lower 0 for sensitizers having low lying triplet states), other experiments
suggest subtleties in mechanism. For sensitization in the polar solvent, acetonitrile, chemically induced dynamic nuclear polarization (CIDW) signals associated with the isomers are observed (emission associated with protons of 23 and enhanced absorption for 24 (ref. 38, 41). The CIDW results are consistent with a mechanism for quenching and isomerization involving electron
transfer and hyperf ine-induced intersystem crossing between nascent singlet and triplet radical ion-pairs (ref. 38). Spin polarization results from a partitioning in which singlet pair recombination favors return to 2 (with reduction in sensitization quantum yield) and triplet recombination of radical The principal steps are shown in Scheme 2. The ion-pairs leads to 24.
dependence on sensitizer triplet energy is then understood in terms of further
partitioning of triplet radical ion pairs (i.e., recombination can lead to a low-lying triplet of sensitizer or 23, or both, with population of the lower local triplet favored). The triplet recombination mechanism is consistent
with the failure of potential singlet sensitizers with low triplet energies
523
CO2CH3
C02CHs
hv
c0&h3
C02CH3
24
23
25
26
27
28
hv
29
-8 30
524
Table 1 Quantum efficiencies of isomerization, 23-24, Sensiti~er
sensitised by electron donorsa
ob
Triplet energy (kcal/mol)
triphenylene phenanthrene
66
0.39
62
0.16
2-methoxynaphthalene
62
chrysene
57
0.24
ca. 43
anthracene
0.20 0.28
48
pyrene
oc
<0.01
(0.01
<0.01d
aAr purged samples, excitation at 313 or 340-370 nm bAcetonitrile solvent CEydrocarbon solvent
d9,10-dimethyl- or lJ4-dimethoxyanthracene Scheme 2
23 23-.] 1[S+., 23-*] 3[S+-, g - 1 3[S+-,23-.] IS* + 1[S+.,
323* (3S*) 323*
-
-
-c
1[S+.
s
+
s
+
J
-
23-*] 23
singlet quenching singlet ion-pair recombination
323* 23
triplet ion-pair recombination
3[S+., 23-.] 3S*
+
3 (S)
24
ion-pair intersystem crossing
triplet decay
triplet isomerization
(e.g. , anthracenes, pyrene) to drive the isomerization or to afford CIDNP effects. Flash photolysis results show also that indeed the sensitirer triplet is populated on pyrene singlet quenching (ref. 38).
For sensitization of 23 in a non-polar solvent, the observations of sensitizer fluorescence quenching and the pattern of quantum yields of rearrangement are similar
(Table 1).
However,
CIDNP effects associated with
radical-ion pair intersystem crossing are not observed so that intermediates of somewhat different structure and lifetime are suggested. A mechanism
525
involving an excitation energy cascade and employing as a key feature exciplex
intersystem crossing has been proposed (ref. 39). The subtle distinction (vide infra) is that the cascade of exciplexes [singlet exciplex triplet exciplex -t
-norbornadiene triplet] need not deploy fully charge-separated radical ions (in non-polar solvent). Moreover, the mechanism of intersystem crossing for
an excited complex having both charge transfer and local excitation transfer character is more complicated, with spin-orbit interaction potentially of greater importance than hyperf ine interaction in dictating the rate of intersystem crossing to the triplet exciplex. The alternative deployment of a diene capable of internal addition as an
electron donor is illustrated in the photochemistry
of
25
(ref. 43).
This
less heralded relative of norbornadiene has a lower ionization potential (8.1
vs 8.6 eV for 1) due to a closer proximity of r bonds (0.29 nm separation). Direct irradiation (254 nm) of 25 gives the cage isomer S in high quantum yield (ca. 0.5), and energy transfer from triplet sensitizers with triplet energies > 60 kcal/mol (benzophenone, naphthalene) results in efficient rearrangement (@= ca. 1.0). The quenching by 25 of fluorescent sensitisers which are good electron also results in acceptors (9,10-dicyanoanthracene, 1-cyanonaphthalene) efficient conversion to the cage isomer. The results for singlet sensitisation of 25 contrast, however, in two interesting ways from the findings for 3. First, the low triplet energy sensitizers (e.g., anthracenes) which failed in electron transfer sensitization with 23 are capable of driving the rearrangement of 25 (e.g., limiting 0 = 0.14, cyclohexane) . An unusual mechanism of singlet-singlet energy transfer which is induced by
donor-acceptor
transformation (ref. 43).
interaction
The concept
has
been
invoked for this
is that collisional encounter with an
electron acceptor results in a withdrawal of electron density from the diene (perturbation of the HOMO of 25) which results in an increase in transannular bond order (ref. 19b). This entrance onto the potential surface for isomerization is further assisted by
the mixing
of
the configurations of
locally excited states associated with diene and sensitizer. Energy transfer
from the sensitiaer is feasible at or near the mid-point of rearrangement, a critical biradical geometry which is relatively low lying (cf. Fig. 2). The reaction coordinate diagram depicting this mechanism is shown in Fig. 3. A
critical feature of this mechanism
is the absence of an intersystem crossing step which is apparent in the sensitization data for 23.
Another novel feature for 25 is the finding of quantum yields > 1.0 for isomerization in a polar medium. Rearrangement under these conditions can be induced either by acceptor sensitiser fluorescence quenching or by irradiation
526
of ground state charge-transfer complexes of give rise to longer wavelength
25
and strong electron acceptors
(DDF) (ref. 43). The latter complexes CT absorption bands (Ama 300-400 nm) which are
such as diethyl 1,l-dicyanofumarate
active in photoisomerization (vide infra).
The observation of anomalously
high quantum yields for rearrangement is consistent with the formation of free
radical cations of the donor. at a glassy carbon electrode (ref. 43).
Indeed electrooxidation of
25
in acetonitrile
(2.0 V vs SCE) results in isomerization to
The relatively low yields of net electron transfer normally
S
P. E.
R
coordinate
D P through interaction with an electron acceptor sensitizer A (nonpolar medium) (ref. 43) (with the per miss^;)^:iJr .I Arc Chem. SOC.).
Figure 3. Potential energy surfaces f o r rearrangement of an electron donor to product
521
obrsL,ved for singlet quenching are amplified
Scheme 3 (S = electron acceptor
sensitizer).
by propagation steps as shown in
A similar pattern of amplified 44)
under
for rearrangement of conditions of acceptor
DDF) in polar medium, also consistent with the
known behavior of HldDB
quantum yields has been reported hexamethyl Dewar benzene (EMDB) (27
-
(ref. J 8 2
quenching or excitation of ground state complexes (e.g. , 27 with fumaronitrile
or
radical cations (ref. 45).
The
for 25 (unlike isomerization of exergonic (ref. 43),
chain sequence is energetically permissible
1)
since
the
overall
isomerization is
Scheme 3 + 25". -
26+. -
S--
-25 25 -
26+*
+
+
25+. -
S-.
4
25'-
+
26 -
S
+ +
electron transfer initiation radical ion rearrangement
25+* -
chain propagation chain termination
25
It should be noted that although diverse mechanisms for (2 + 2) photoaddition for diene systems appear to be available, for many candidate dienes a number of factors may conspire against an efficient internal addition.
Thus, 1,5-~yclooctadiene
(29), limonene , and
dicyclopentadisne are
fair quenchers of electron acceptor sensitizers, but photoisomerization (e .g. ,
does not result from singlet sensitization (ref. 19b). The absence of a low lying triplet or significant transannular interaction in the diene, along with competing diene-sensitiaer addition contribute to the poor
29 -30)
performance of some substrates.
3.2
Photorearrangement via transition metal complexes An assortment of photoreactions including isomerization and cycloaddition have been induced on irradiation of olefins in the presence of transition metal salts, the most
46).
The internal addition of.
notable examples involving Cu(1) salts (ref.
S
promoted
by uv photolysis of the diene in
the presence of cuprous chloride was first reported by Srinivasan (ref. 47). Reinvestigation of valence isomerization of and related alkene-Cu(1) systems (ref. 48) has shown that ground
olef in T system
are
important
and
result from formation of relatively bonds.
that
more
state
complexes
isomerization
reactive
trans
of the metal and
and other reactions
(medium ring) double
528
Kutal and his coworkers (ref. 49) extended the investigations of Cu(1) as a sensitizer for the isomerization of 1, showing that the metal plays a the diene which absorb just past 300 nm.
catalytic role via 1:l complexes of
Examples
of
isomerieation included
the
use
of
simple CuX salts and
Cu[HB(pz)3] (CO) , where BB(pz)3 is the hydrotris(1-pyrazoly1)borate ligand (@ up to 0.65). The mechanism of sensitization requires displacement by diene of a ligand in the Cu(1)
coordination sphere and the development of a charge-
transfer (olefin -.metal)
band.
transannular interaction in
the excited species to
2.
1 (ref. Other
CT photoexcitation leads to alteration of
19c, vide supra), and a partitioning of important features regarding the enhanced
efficiency for internal addition of the diene include single coordination of a
double bond by the metal (to inhibit olefin dimerization). the sensitization mechanism is further copper-phosphine
complexes
[CuX(PPh3)]
complexation of diene by the metal
quenching of
the excited metal
.
For this pathway ground state is no longer required, and a bimoleular
complex
quenching of typical organic triplet
-
Ligand control of
illustrated by the sensitization via
mechanism recently reported by Grutsch
is
important
sensitizers)
(more akin to the
(ref. 50). An interesting
and Kutal (ref. 51) for sensitization
of 1 2 by orthometalated iridium(II1) complexes involves energy transfer at a reaction mid-point, similar to the finding of isomerization of 25 induced by
singlet sensitizers (vide supra). This proposal followed experiments showing that the Ir(II1) sensitizer is energy insufficient for vertical energy
transfer, but which nonetheless is quenched at a moderate rate (kq = 1.4 x lo8 I4-l sec-l , acetonitrile) through an apparent EDA interaction.
4
(2
+
2)REVERSION: REACTIVITY OF SMALL RING RADICU CATIONS
4.1
Thermal (catalyzed) cycloreversion of small rings Although the conjugating and electron donating properties and general reactivity of fused small rings is well known (ref. 52), the products of internal(2 + 2)addition are found to be kinetically stable with respect to
For example, the return of 2 to norbornadiene proceeds slowly at 1800 in the gas phase in an aged reactor (ref. 53). Addition products of the (2 + 2) type are, however, susceptible to
thermally
induced cycloreversion.
ring opening including cycloreversion under a variety of oxidative conditions. For example, Cassar, Eaton, and Halpern (ref. 54) reported the clean, rapid transformation of cubanes to diene valence
isomers
in the presence of Rh(1)
catalysts (e.g. , [Rh(diene)C1]2, CCl4, 400). Hogeveen and Nusse (ref. 55) showed that reversion of occurred at temperatures as low as -260 in the
z
529
presence of Rh(1) or Pd(I1) catalysts.
ring compounds by transition
metals
Mechanisms of rearrangement for small
(including oxidative addition and ligand
insertion steps) have been reviewed by Bishop (ref. 56) The
cycloreversion
catalysts was discussed
of
by
completion of a photochemical energy
of
z to
norbornadiene
has
been
and
and in the presence of alumina (ref, classified
the
catalysts
quadricyclene as (1) square
metals which are electron
electron oxidizing agents.
planar
based
on
amide-linked
18, using
Cassman
and
quadricyclene and
his
are
complexes
Similarly, the reversion
have
polystyrene
small
showed
ring
with turnover numbers as
as
electrolyte, at 1 . 3
high
Yo,
W) (ref. 57)
successful
in
reaction
of
been
electrons) and (2) one-
described
(ref. 60) for
catalysts for cycloreversion which and
cobalt
tetraarylporphyrin.
an aqueous system for the water soluble
2, E(1/2)
subsequently
63)
mild conditions with a
of relatively large transition
valence
(15-16
electrochemical conditions (for and Kutal (ref.
under
of metal
in terms of
18)
Bautala, King, and Kutal (ref. 59)
anchored
coworkers
presence
(ref.
an anionic Co porphyrin (ref. 61). related
the
(Ni, Co(I1) porphyrins,
Conditions have also been extended to system,
out
58).
Procedures
preparation of effective polymer are
storage cycle.
which
deficient
in
Ramachandran
carried
variety of transition metal catalysts have
17
bis-homocubane
Jones
some
compounds
are
ago
(ref.
62) that
easily oxidized under
0 . 9 1 V vs SCE, CH3CN).
=
showed
years
Yasufuku
that electrooxidation is catalytic
11 for ring opening
of
2
(CH2C12, TBAP
V vs SCE). A chain reaction involving radical cations was
invoked, analogous to the mechanism of Scheme 3, but employing the propagation
step,
1'-+ 2
-. 1
+
2+.
in which the cage isomer is the preferred cation.
It has been also recently shown
that one-electron oxidation of a triarylamine
provides an efficient relay for electrocatalytic ring opening of
4.2
Photosensitized cycloreversion via electron transfer or electron donoracceptor complexes
Before the electron donor properties
fully appreciated, it was considered most
-1
z (ref. 64).
could
be
induced via
quenching of
of the quadricyclene molecule were
unusual that the rearrangement,
the
aromatic hydrocarbon sensitizers (ref. 65).
singlet
2
+
(fluorescent) state of
Solomon, Steel, and Weller (ref.
66) made the suggestion that a charge-transfer
interaction
between
2 and
sensitizer is important and is responsible for a correlation between quenching rate and the electron affinity of
sensitizers.
Taylor (ref. 67) showed that
the dependence of quenching ability of several small ring compounds is related
to ionization potential for the strained hydrocarbons. recently that direct photolysis of
2
It has been shown more
at short wavelengths (187 nm) also leads
to norbornadiene along with other products (ref. 68).
530
The photoelectron spectrum of 2 and the electronic structures for the hydrocarbon and its radical-cation have been of interest for some time (ref.
69).
Gamma radiolysis of
2
leads to the same radical-ion
(or
(Ama
1)
in an electron scavenging matrix at 77K A+* (ref. 70). The
= 635 nm) assigned to
apparently facile rearrangement of the radical-cation of
the calculated (]LINDO/I) potential surface (ref. 70b).
2 is predicted
from
The CIDW method has
been a valuable tool in the investigation of radical cation rearrangements of
this type (ref. 71).
In the case of photolysis of 2 in the presence of
electron acceptor sensitizers, the nuclear polarization results are consistent with the intermediacy of two discrete radical ions corresponding to valence
I
and 2. Also indicated is a mechanism of partitioning of intermediates involving cage recombination of the radical ion products of isomers,
electron transfer between quadricyclene and sensitizer, and cage escape of which is permitted to rearrange eventually leading to 1 (ref. 72).
2'.
In a series of investigations of valence isomerization induced by
electron donor-acceptor (EDA) interaction, Jones, Becker, and Chiang (ref. 43,73,75) have identified a variety of factors governing quantum yields of rearrangement, including donor structure, the degree of charge separation in excited
EDA complexes, and the wavelength of excitation with a certain mode of
photolysis
.
Under circumstances in which discrete radical-ions are less
likely intermediates for rearrangement of
perturbation of frontier molecular orbitals is important (ref. 19c, 43,
rearrangement of
2, 25,
73).
2
Moreover,
(and related structures), the (HOMO) by sensitizer interaction in these studies of sensitized
and 27, a persistent pattern of isomerization quantum
yield data was revealed (ref. 43-4, 7 3 4 , as follows.
(1) For quenching of acceptor sensitizer fluorescence (e.g., 1cyanonaphthalene) in non-polar solvent, rearrangement quantum yield is high, near unity.
(2) For quenching of acceptor sensitizers in polar medium (e.g., CH3CN), isomerieation yield (free radical ion yield on flash photolysis) is low (typically <0.1) [overall isomerization yield
for 25,
27
can remain high and >1.0
- due to a radical-ion chain reaction, vide
supra, Scheme 31.
-
irradiation of ground state CT complexes of 2 and strong acceptors (e.g., fumaronitrile) in non-polar solvent, yields are again low. (3) On
(4) On irradiation of CT complexes in a polar medium, quantum yields are variable, with an increase noted for excitation at shorter wavelengths within the charge-transfer band.
531
These results are consistent with the imposition of subtly different EDA intermediates for the various modes of sensitization. For sensitizer quenching in non-polar solvent, the polar (but not fully ionized) exciplex undergoes EDA induced structural changes which result in energy transfer at
biradicaloid geometries sensitization mechanism
(Fig. is
3)
(ref.
43).
thwarted,
however,
This highly efficient on
quenching acceptor
sensitizers in polar solvent. In this case, solvent separated ion-pairs are formed on sensitizer-substrate encounter; these fully ionized intermediates efficiently return to ground
state
(without isomerization)
through back
electron transfer (ref. 75). The excited states of ground state complexes (contact ion-pairs) suffer a similar fate except that on excitation at short wavelengths within a CT band, excess vibrational energy is utilized to evolve the contact pairs into solvent separated species, so that moderate yields of
A recent confirmation of isomerization are obtained (ref. 44,74). differential behavior for less polar exciplexes vs fully charge separated ion
pairs is apparent in the (2
+
pyren-1-ylcyclobutane for which important (ref. 76).
2) cycloreversion of 1- (dimethylaminophenyl)-2-
intramolecular donor-acceptor interaction is
Some small ring compounds
may
respond in their tendencies for
photoinduced ring opening in different ways
.
in terms of the degree of charge
separation imparted to exciplexes In their study of aryl-substituted cyclobutanes, Pac and Sakurai (ref. 77,78) have shown that a higher degree of charge separation (interaction with a more potent acceptor sensitizer) leads to a higher quantum yield of ring fragmentation in benzene solvent. They also
note the importance of
geminate
ion-pairs in sensitized cycloreversion of
diphenylcyclobutane in more polar media. Evidence is provided also for free radical-ions (and a chain component, vide supra) for ring opening of a more reducing dianisylcyclobutane derivative in acetonitrile (e .g., 9-cyano- or 9,lO-dichloroanthracene sensitizers) (ref. 77).
A multiplicity of modes of (2
+
2) cycloreversion is also demonstrated in
the studies of Mukai and his coworkers (ref. 79) on the bis-homocubanes 2. The cage compounds underwent clean ring opening to 32 on reaction with a very strong acceptor such as TCNE (in the ground state), alternatively via an excited state of the ground state complex, and on quenching a triplet quinone sensitizer (ref. 79). Also effective were cyanoaromatic acceptor sensitizers
and fluorescent pyrylium salts in driving the isomerization of 32 (ref. 80). For polar media a chain mechanism involving radical-cations (ala Scheme 3) was invoked (ref. 79).
An
extensive study
by Yiyashi
(ref. 81) which is also
related is the ahorno- (2 + 2 ) " or [3 + 21 cycloaddition which occurs on irradiation of CT complexes of methylenecyclopropanes, cyclopropanes,
532 aziridines, and oxiranes with
of the electron donors
In
TCNE.
and either TCNE
a novel profile of products, adducts (via geminate radical-ion pairs) or
molecular oxygen (via solvent separated ion-pairs) were obtained. Metal or semiconductor catalysts induce ring opening of
mechanisms that are
certainly of
quantum yield of isomerisation (102)
the
82) due to sensitiration by PdCl2(+-NBD)
EDA
2 by
way of
An especially high
type also.
is reported by Borsub and Kutal (ref.
[NED =
norbornadiene] and a radical-
ion chain process is implicated. Suspensions of seniconductor powders such as
CdS and Ti02 (n-type semiconductors) in cH2C12 or cH3CN solvents are effective in driving rearrangements of
5
2, 27, and 1 (ref.
83).
CONCLUDING REMARKS: FUTURE DIRECTIONS The evolution of
rather modestly with
photoinduced (2 +
2) cycloaddition mechanisms began
apparent requirements involving a photosensitizer for
transfer of triplet energy to an
olefin
.
(diene)
The process was assumed to
be irreversible on photolysis (although reversion is thermally induced) due to
a reduction in conjugation.
A
richer variety
appropriate for driving reactions of this
recent findings center around the interaction between
of mechanisms is now at hand,
type in either direction.
sensitizer and
(or between
substrate
partners) and the special role of strained-ring(2 donors. The more unusual mechanisms include the
norbornadiene
use
of
ionizing
(i
isomerization
rearrangement of
25
via
dimers by
+
-. 2)
(7-radiolysis)
been
The
related
(ref. 85).
photoreactivation flavoenzymes, which
from
23
the
The splitting of may
utilize
an
include triplet recombination of radical-
electron transfer between
isomerizable substrate such as
to drive
hemin-catalyzed
indirectly to the
electron transfer path, is also of current interest (ref. 86). ion pairs which result
cycloaddition
2)cycloadducts as electron
(ref. 84).
radical-ions has
The important new mechanisms
(EDA)
for (2 + 2)cycloaddition (reversion) now
radiation
mechanism of cytochrome oxidation (epoxidation) thymine
The more
importance of electron donor-acceptor
(Scheme 2).
a sensitieer and an
Exciplexes of lower polarity
(not fully ionieed) may employ either of two paths leading to 2 + 2 internal
adducts:
a
singlet-triplet exciplex
cascade involving exciplex intersystem
crossing and population of rearranging triplets, or (in the absence of a lowlying local
triplet state of substrate)
jbiradicaloid)
geometries following
sensitizer (Fig. 3).
EDA
an
energy
transfer at critical
interaction
of
substrate and
Acceptor sensitizers are also capable of producing
free
radical-cations which participate in rearrangment of small ring cycloadducts in polar media, often
initiating a
radical-ion chain cycloreversion (Scheme
533
31
33
Y’
35
32
34
dZvR 36
534
3). Finally, rearrangements may be induced by irradiation of ground state charge-transfer (CT) complexes of substrates. The key feature here is that resultant contact radical-ion pairs evolve into rearranging solvent-separated ion pairs depending on CT excitation energy. Wavelength dependent
photoisomerization via CT complexes is therefore a general result.
Practical photochromic systems will require adaptation of the solution
phase results for nearly all reported (2 + 2) additions to the solid state.
Recent progress in this area is illustrated by reported for ketones related to
which
quantitatively) on silica gel (ref. 87).
internal photoadditions
can be carried out (and analyzed
Also of interest in terms of new
media for cycloaddition reactions are examples of direct and sensitized norbornadiene rearrangement (e.g., 33 J 4 3 which can be observed on irradiation of
thin films of
-.
the polymer-bound system
(ref. 88).
The
photochromism of this system is shown in Fig. 4, including the reversion (34which is induced on treatment of (irradiated) films with a Co-porphyrin catalyst (section 4.1).
methods (ref. 64) for
Of practical interest also are electrochemical
'switching'
the direction of valence isoraerieations
(potentially the development of systems which are photochromic in one direction and electrochromic in the reverse).
Useful photochromic systems will also require substantial improvement in
the extent of wavelength alteration (color change) generally observed on photolysis of (2 + 2) systems. The bleaching of the blue colored thiones
-.
through(2 + 2)cycloaddition (e.g., J 5 L (ref. 17) is an exceptional example. Obviously, irradiations requiring sensitizers do not result in a bleaching of the principal light absorbing species. These limitations are in principle
solvable with the design of more interesting chromophores. Norbornadiene 2 which employs a cyanine-like chromophore (which would be lost on isomerization) is an example of a system with the desired level of pigmentation which has not been examined so far. One mode of sensitization (broadly defined) which
is highly photochromic involves CT complexation An attractive new design would employ an isomerizable substrate such as norbornadiene (N) with an attached triplet sensitizer (ref. 31b). The proposed structure is further modified by inclusion of a strong electron acceptor which will complex preferentially with the electron donor Q moiety following rearrangement. (section 4.2).
-N -
u
Sens
acceptor
triplet energy transfer
-
Sens
u
- Q -
acceptor
charge-transfer interaction
535
1.f
t 0 $ C
ul n
0.7
q
190
220
270
320
Wavelength (nm)
Figure 4. Absorption of 33 before and after irradiation of films (Hg lamp) and the progress of reversion (3-33) on reaction of irradiated films with a Co-porphyrin catalyst (ref. 88). (with the permission of Polymer J.) This proposed system could be
expected to shift absorption from 300 nm well
into the visible (ca. 500 nm) (ref. 74,89). The behavior of the interesting fused norbornadiene sytem 6 (R = S02C6H5) of Paquette and Kuenser (ref. 90) which participates in a Q-Q rearrangement due to CT interaction with an attached acceptor is suggestive of the concept.
with the development of the Q-acceptor complex
536
The(2
t
2)addition
due to the facility of
and
its reversion will continue to claim attention
forward and reverse processes and the stability
associated with novel polycyclic
(caged) structures.
These systems show an
extreme sensitivity to donor-acceptor interaction and have already served
admirably as mechanistic probes.
The author wishes to thank his coworkers who contributed to the research effort directed to valence photoisomer systems and electron transfer
photosensitization, and the U.
S. Department of Energy, Division of Basic
Energy Sciences, which provided financial support.
6
REFERENCES
1
b. R. Hoffman and R. B. Woodward, Acc. Chem. Res., 1 (1968) 17. S. J. Cristol and R. L. Snell, J. Am. Chem. SOC., 80 (1958) 1950; 76
2 3 4
5 6 7 8 9 10 11 12 13 14
15 16 17 18 19
a.
H. E. Zimmerman, Acc. Chem. Res.,
4 (1971) 272;
(1954) 5600.
W. G. Dauben and R. L. Cargill, Tetrahedron, 15 (1961) 197. B. C. Roquitte, J. Am. Chem. SOC., 85 (1963) 3700; J. Phys. Chem., 69 (1965) 2475.
G. S. Hammond, N. J. Turro, and A.
Fischer, J. Am. Chem. SOC., 83 (1961) 4674. a. C. W. Jefford, U. Burger, W. Broeckx, and J.-C.E. Gehret, Tetrahedron Lett. (1973) 5187; b. J. R. Edman, J. Org. Chem., 32 (1967) 2920. von H. Sauter, H. Hoerster, and E. Prinsbach, Angew. Chem., 85 (1973) 1106. P. E. Eaton, Acc. Chem. Res., 1 (1968) 50. U. Klinsmann, J. Gauthier, K. Schaffner, M. Pasternak, and B. Fuchs, Helv. Chim. Acta, 55 (1972) 2643. G. Ciamician and P. Silber, Ber., 41 (1908) 1928. R. S. H. Liu and C. S. Hanmond, J. Am. Chem. SOC., 89 (1967) 4936. W. G. Dauben and W. T. Wipke, Pure Appl. Chem., 9 (1964) 939. W. L. Dilling, Chem. Rev., 66 (1966) 373. a. H. Shizuka, I. Seki, T. Morita, and T. Iieuka, Bull. Chem. SOC. Jpn., 52 (1979) 2074. b. G. Kaupp and M. Stark, Chem. Ber., 110 (1977) 3084. c. G. Jones, I1 and V. L. Chow, J. Org. Chem., 39 (1974) 1447. a. G. Jones, I1 and J. C. Staires, Tetrahedron Lett. (1974) 2099. b. G. Adams, C. Bibby, and R. Grigg, J. C. S. Chem. Comm. (1972) 491. G. Jones, I1 in 'Organic Photochemistry," vol 5, A. Padwa, ed., Marcel1 Dekker, New York, 1981, chp 1. P. de Mayo, Acc. Chem. Res., 9 (1976) 52. G. Jones, I1 and B. R. Ramachandran, J. Org. Chem., 41 (1976) 798. a. R. R. Hautala, R. B. King, and C. Kutal, ed., 'Solar Energy: Chemical Conversion and Storage," The Humana Press, Clifton, New Jersey, 1979; b. G. Jones, 11, S. H. Chiang, and P. T. Xuan, J. Photochem., 10 (1979) 1. c. H.-D. Scharf, J. Fleischhauer, H. Leismann, I. Ressler, W. Schleker, and R. Weitz, Angew. Chem. Int. Ed. Engl., 18 (1979) 652.
537
20 C . Philippopoulos, D. Economou, C. Economou, and J . Marangozis, Ind. Eng. Chem. Prod. Res. Dev. 22 (1983) 627. 21 Z. Yoshida, J. Photochem., 29 (1985) 27. 22 a. T. Mukai and Y. Yamashita, Tetrahedron L e t t . (1978) 357. b. K. Yaruyama, T. Terada, Y. Yamamoto, J. Org. Chem., 46 (1981) 5294. c. H. H i r a o , A. Ando, T. Hamada, and 0. Yonemitsu, J. C. 5. Chem. Commun. , (1984) 300. 23 Y. Harel, A. W. Adamson, C. Kutal, P. A. Grutsch, and K. Yasufuku, J. Phys. Chem., 91 (1987) 901. 24 G . Kaupp and H. Prinzbach, Helv. Chim Acta, 52 (1969) 956. 25 H.-D. Martin and B. Mayer, Angew. Chem. I n t . Ed. Engl. , 22 (1983) 283. 26 K. Maruyama, H. Tamiaki, and S. Kawabata, J. Org. Chem., 50 (1985) 4742. 27 Z. Yoshida, Heterocycles, 21 (1984) 331. The Benjamin/Cummings 28 N. J. Turro, Vodern Molecular Photochemistry, Publishing Go., I n c . , Reading, Massachusetts, 1978, chp 8,9. 29 G. Jones, i n r e f . 19a. 30 G. Jones, I1 and R. J. B u t l e r , S o l a r Energy, 29 (1982) 579. 31 a. E. A. Fehnel and F. C. Brokaw, J. Org. Chem., 45 (1980) 578. b. A. J. 6. B a m i s e , A. A. Gorman, R. L. Leyland, C. T. Parekh, and P. G . Smith, Tetrahedron, 36 (1980) 397. 32 K. Raghavachari, R. C. Haddon, H. D. Roth, J. Am. Chem. SOC. , 105 (1983) 3110. 33 H. Ikezawa, C. Kutal, K. Yasufuku, and H. Yamarzaki, J. Am. Chem. SOC., 108 (1986) 1589. 34 T. Kobayashi, Z. Yoshida, Y. Asako, S. Miki, and S. Kato, i b i d . , 109 (1987) 5103. 35 6 . Nishino, S. Toki, and S. Takamuku, J. Am. Chem. SOC., 108 (1986) 5030. 36 G . Jones, I1 and B. R. Ramachandran, J. Photochem., 5 (1976) 341. 37 G. J. Kavarnos and N. J. Turro, Chem. Rev., 86 (1986) 401. 38 G. Jones, 11, W. Schwarz, and V. Malba, J. Phys. Chem., 86 (1982) 2286. 39 G . Jones, 11, P. T. Xuan, and W . Schwarz, Tetrahedron L e t t . (1982) 5505. 40 D. Rehm and A. Weller, Israel J. Chem., 8 (1970) 259. 41 W. Schwarz, K. M. Dangel, J. Bargon, and G. Jones, 11, J. Am. Chem. SOC., 104 (1982) 5686. 42 a. R. A . Caldwell and D. Creed, Acc. Chem. Res., 13 (1980) 45; b. R. A. Caldwell, K . Mizuno, P. E. Hansen, L. P. Vo, Y. Frentrup, and C. D. Ho, J. Am. Chem. SOC., 103 (1981) 7263. 43 G . Jones, 11, W. G. Becker, and S. H. Chiang, J. Am. Chem. SOC., 105 (1983) 1269. 44 G. Jones, I1 and W. G. Becker, J. Am. Chem. SOC., 105 (1983) 1276; 103 (1981) 4630. 45 T. R. Evans, B. W. Wake, M . M. S i f r a i n , Tetrahedron L e t t . (1973) 701. 46 C. Kutal and P. A. Grutsch, Adv. Chem. Ser. (1979) No. 173, 325. 47 R. Srinivasan, J. Am. Chem. SOC., 86 (1964) 3318. 48 a. E. Grobbelaar, C. K u t a l , and S . W. Orchard, Inorg. Chem., 21 (1982) 414; b. J. Th. Y . Evers and A. Mackor, Tetrahedron L e t t . (1978) 2321. 49 a. R. F. S t e r l i n g and C. Kutal, Inorg. Chem., 19 (1980) 1502; b. D. P. Schwendiman and C. K u t a l , J. Am. Chem. SOC., 99 (1977) 5677. 50 a. N. Borsub, S. Chang, and C. K u t a l , Inorg. Chem., 21 (1982) 538. b. D. J . F i f e , W. M. Moore, and K . W. Morse, J. Am. Chem. SOC., 107 (1985) 7077. 51 P. A. Grutsch and C. Kutal, J . Am. Chem. SOC., 108 (1986) 3108. 52 A. de Meijere, Angew. Chem. I n t . Ed. Engl., 18 (1979) 809. 53 H. M. Frey, J. Chem. SOC. (1964) 365. 54 L. Cassar, P. E. Eaton, and J. Halpern, J. Am. Chem. SOC., 92 (1970) 3515. 55 H . Hogeveen and B. J. Nusse, Tetrahedron L e t t . (1973) 3667. 56 K. C. Bishop, 111, Chem. Rev., 76 (1976) 461. 57 A. Sen and R. R. Thomas, Organometallics, 1 (1982) 1251. 58 D. N. B u t l e r and I. Gupta, Can. J. Chem., 60 (1982) 415.
538
59 60 61 62 63 64 65 66 67 68 69 70
R . R . Hautala, R. B. King, and C. Kutal, i n ref 19a, p 333. R. B. King and E. M. Sweet, J. Org. Chem. 44 (1979) 385. K. Maruyama and H. Tamiaki, J. Org. Chem., 51 (1986) 602. P. G. Gassman, R. Yamaguchi, 6. F. Koser, J. Org. Chem., 43 (1978) 4392. K. Yasufuku, K. Takahashi, and C. Kutal, Tetrahedron L e t t . (1984) 4893. P. G. Gassman and J. W. Hershberger, J. Org. Chem., 52 (1987) 1337. a. S. Murov and G. S. Hammond, J. Phys. Chem., 72 (1968) 3797. b. G. S. Bammond, P. Wyatt, C. D. DeBoer, and N. J. Turro, J. Am. Chem. Soc., 86 (1964) 2532. B. S. Solomon, C. S t e e l , and A. Weller, Chem. Comm. (1969) 927. G. N. Taylor, Chem. Phys. L e t t . 10 (1971) 355. R. Srinivasan, T. Baum, and 6 . Epling, J. C. 5. Chem. Comm. (1982) 437. H. -D. Martin, C. Heller, E. Haselbach, and 2. Lanyjova, Helv. Chim. Acta, 57 (1974) 465.
a. T. Shida, E. Haselbach, and T. Bally, Acc. Chem. Res., 17 (1984) 180. b. E. Haselbach, T. Bally, 2. Lanyiova, and P. Baertschi, Helv. Chim. Acta
62 (1979) 583. 71 H. D. Both, Acc. Cham. Res., 20 (1987) 343. 72 H. D. Roth, Y. L. Y. S c h i l l i n g , and G. Jones, 11, J. Am. Chem. SOC., 103 (1981) 1246. 73 a . G. Jones, 11, S. H. Chiang, W. G. Becker, and D. P. Greenberg, J. C. S. Chem. Comm. (1980) 681; b. G. Jones, 11, and S. H. Chiang, Tetrahedron, 33 (1981) 337. 74 G. Jones, I1 and W. 6. Becker, Chem. Phys. L e t t . , 85 (1982) 271. 75 G . Jones, 11, S. H. Chiang, W. G. Becker, and J. A. Welch, J. Phys. Chem., 85 (1982) 2805. 76 A . Tsuchida, M. Yamamoto, and Y. Nishijima, J. Chem. SOC. Perkin Trans. I1 (1987) 507. 77 C. Pac, Pure and Appl. Chem., 58 (1986) 1249. 78 a. C. Pac, T. Ohtsuki, Y. Shiota, S . Yanagida, and 8. Sakurai, Bull. Chem. Soc. Jpn: ! 59 (1986) 1133. b. T. Majima, C. Pac, and H. Sakurai, J. Am. Chem. SOC., 102 (1980) 5265. 79 T. Mukai, K. Sato, and Y. Yamashita, J. Am. Chem. Soc., 103 (1981) 670. 80 a. E. Hasegawa, K. Okada, and T. Mukai, J. Am. Chem. Soc., 106 (1984) 6852 ; b. K. Okada, K. Hisamitsu, T. Yiyashi, and T. Yukai, J. C. S. Chem. Comm., (1982) 974. 81 T. Miyashi, Y. Kamata, and T. Ilukai, J. Am. Chem. SOC., 109 (1987) 2780. 82 N . Borsub and C. Kutal, J. Am. Chem. SOC., 106 (1984) 4826. 83 a. H . Ikezawa and C. Kutal, J. Org. Chem., 52 (1987) 3299; b. K. Okada, K. Hisamitsu, and T. Yukai, J. C. S . Chem. Comm. (1980) 941; c. H. Al-Ekabi and P. de Mayo, J. Phys. Chem., 90 (1986) 4075. 84 a. M . J. Mirbach, M. F. Yirbach, R. Vartan-Boghossian, and A. Saus, Nouveau. J. Chim., 5 (1981) 113; b. J. W. F. van Ingen, C. H. C. van Tieghem, and W. A. Cramer, J. Chem. Phys., 53 (1970) 3665. 85 T. G . Traylor and A. R. Miksrtal, J. Am. Chem. SOC., 109 (1987) 2770. 86 C. Walsh, Acc. Chem. Res., 19 (1986) 216. 87 S. Lazare, P. de Mayo, and W. R. Ware, Photochem. and Photobiol., 34 (1981) 187. 88 T. Nishikubo, A. Sahara, and T. Shimokawa, Polym. J., 19 (1987) 991. 89 I. Tabushi, and K. Yamamura, Tetrahedron, 31 (1975) 1827. 90 L. A. Paquette and 8. Kuenrer, J. Am. Chem. Soc., 108 (1986) 7431.
539
Chapter 73
Cycloaddition Reactions Involving 4n Electrons: (2+2) Cycloaddition; Molecules with Multiple Bonds Incorporated in or Linked to Aromatic Systems
J-P. Desvergneand H. Bouas-Laurent 1
INTRODUCTION
Diverse aspects of (2+2) cycloaddition, on valence photoisomerisation of norbornadiene and similar compounds have been treated in the preceding chapter.
This complementary account describes the intermolecular m ' - I intramolecular photodimerisation of molecules having multiple bonds conjugated to aromatic rings, such as cinnamates, or incorporated in aromatic systems such as phenanthrenes or acenaphthylenes (Fig.1). A great deal of work has been devoted to these chemical species which have the advantage of absorbing light above 300 nm and present a potential for photochromic systems formation in solution and in the solid state (ref.1-3).
(2+2) and (4+4) Photocycloaddition presumably follow
similar pathways or proceed according to similar mechanisms which can be described by a common general model; the latter is detailed in the chapter on (4+4) photocycloaddition. An outline is given below. According to the Woodward-Hoffmann rules, the concerted (2+2) photocycloaddition is "allowed" in the excited state (ref.4). The same is true for the photodissociation. Other approaches due to Dewar (ref.51 and Zimmerman (ref.6) indicate that the reactions involving "4n" electrons are favoured in the excited state without any symmetry requirement. These reactions involve the Singlet state (S1); Caldwell (ref.7) has demonstrated for some of them, and inferred for the others, that there is an excimer or exciplex on the reaction pathway (even in the absence of fluores-
540
cence emission from these complexes) : the cycloadduct results from the "stereospecific collapse of the maximum overlap sandwich exciplex (or excimer)" (equation 1) :
This accounts for the stereoselectivity. Thus it is generally accepted, e.g. for phenanthrenes and acenaphthylenes, that the syn
photodimers
are
formed from the S1 state
R = H , alkyl
via
a
sandwich
Z = CN Y = OCH3
Z = CN, COOH
I I
'63
X
\ I
a
I
X = SiMepOSiMez
X = CH20CH2
Fig.1 Typical compounds undergoing ( 2 + 2) cycloaddition with a potential for photochromic systems formation; these molecules have a double bond conjugated to an aromatic ring or included in aromatic systems. a) molecules undergoing intermolecular cycloaddition b) bichromophoric systems leading to intramolecular cycloadducts.
541
-cu
excimer. (See & R1=R2=H)). But the regioselectivity does not always follow the frontier orbitals prediction (ref.4 b-c) as is shown in the photodimerisation of 9-cyanophenanthrene (5, R1=CN, R2=H) which generates the syn-head-to-tail (ht) photodimer (6, R1=CN, R2=H) from the $1 state instead of the theoretically expected head-to-head (hh) isomer. The same is true for the major photodimer of 1-cyanoacenaphthylene in its singlet excited state R1=CN, R2=H). No theoretical explanation seems to have been given to date. The cycloadducts can also be formed from a triplet excited state:
(u
then, a 1,4,diradical intermediate can appear on the reaction pathway and the stereospecificity is iost (equation 2) :
3A*
+
A
diradical
__L
A2
(2)
Thus, the sensitized photodimerisation of acenaphthylene is known to generate a mixture of anti (major) 9 (Rl=R2=H) and syn (lsl R1=R2=H) (photodimer) in agreement with the behaviour of a triplet biradical intermediate. The ratio anti:syn increases for 1-cyanoacenaphthylene as the cyano groups help stabilise the diradical intermediates, which results in the head-to-head (hh) anti photodimer (9,R1=CN, R2=H).
542
2
INTERMOLECULAR (2+2) PHOTOCYCLOADDITION 2.1 Phenanthrenes
Phenanthrene and derivatives (ref.7) absorb light in the near UV region ( xmaX = 360 nm); with activated olefins, they form cycloadducts absorbing at higher energies (biphenyl chromophores) than addends. As shown in the following reactions, the stereochemistry, the efficiency of the cycloaddition (+R) and the reverse process diss), are very sensitive to the nature of the substrate. Thus phenanthrene (1, R1 = H) and dimethyl fumarate (2, R2=R3=C02Et) give the two cyclobutanes 3a and &b with a low quantum yield (+R = 0.0013) and very efficiently regenerate the addends (+ diss 0.99) (ref.7) whereas anethole (2,R2 = p-anisyl, R3 = CH3) and 9cyanophenanthrene (&, R1 = CN) are readily transformed ($ R = 0.57) (refs.S,9) into the more hindered endo-phenyl adduct 4a which smoothly decomposes into the starting material (+diss = 0.39). In the latter case, Caldwell demonstrated for the first time the intermediacy of an exciplex in the photodimerisation process (ref-9).
(9
1 -
3a,b 4a,b
-2
.............................................................. R1 I R2 I R3 I @R I @ diss H CN
I
I
C02Et
I
I
p-anisyl
'I
C02Et
I I
I
I I
0.0013 I
I
CH3
I I
i
0.99
I
0.57
I
0.39
I I
The scientific literature provides other examples (refs.10-13) dealing with electron rich or electron poor olefins which show, except for 1,2-dichloromaleic anhydride (ref.l4), the non stereoselectivity of the reaction but no data are available concerning the reversible process.
543
If phenanthrene ( 5 , R1=R2=H) is photostable, the 9-cyan0 derivative (2,R1=CN, R2=H) was shown to photodimerise by Timmons (ref.15): later, its structure was established to be syn ht (refs.16,18,19). The 9,lO dicyano derivative (R1 = R2 = CN) also undergoes photodimerisation (ref.15). Similarly, 9-methoxy 10-cyanophenanthrene ( 5 ) which was shown to exhibit the first fluorescing excimer in the phenanthrene series, photodimerises easily (ref.17). Crossed photodimers between differently substituted phenanthrenes, such as (1)could be
5
.R1 = CN
R2
= OMe
R1 = CN
5
7 -
R2 = H
also obtained (ref.20). All of these dimers (which absorb higher energy than phenanthrene chromophore, Fig.2) could converted back to monomers by irradiation with light of <313 in a rigid glass at 77K. The photodimerisation quantum yield 9-cyano phenanthrene was measured in methanol and propanol different temperatures (ref.21).
at be nm of at
Fig. 2 Electronic absorption spectra of 9-cyanophenanthrene (---I (c = ~ X I O - ~ Mcyclohexane) , and its syn-l,3-photodimer (----) (c = 3x10-51, diethyl ether).
544
2.2 AcenaDhthvlenes Acenaphthylene and its derivatives, which are coloured, form, like phenanthrene, colourless cycloadducts with activated olefins
(ref.22). Moreover, these compounds easily photodimerise owing to the high reactivity of the 1,2 bond. The photodimerisation of acenaphthylene (8,R1 = R2 = H) has been comprehensively investigated (refs-23-28). R.,
R1 R2 CN
C02H No 2 H H
H H
H c1 H
8 -
9
10 (syn,
(anti, hh)
ht)
Photodimers of syn and anti configuration were isolated and found to be solvent and concentration dependent. The multiplicity of the excited state plays a major role, the syn isomers origina-
ting mainly in the singlet state, the anti isomers being generated via the triplet state. Acenaphthylene derivatives (&) (R1 = CN, C02H, NOZ, CONH2, R2 = H ; R1 = H , R2 = C1) behave similarly (refs.22,28,29). The photodimerisation (concentration dependent) generates colourless products (9)and (lo), there is a significant variation of the absorption spectrum (Fig.3).
ZtI 3
400
300
Electronic absorption spectra of
RI=CN, (---!
.-.,,,:4:
R2=M)
mi
1-cyanoacenaphthylene
i . . . ) and its anti 1 , 2 !hh! (-1
ptot.odimers
(c
L-
10-4M,
cyclohexane)
.
and s1;n 1 , 3
545
The photodimerisation of 1-cyanoacenaphthylene ( 8 ) (R1 = CN. R2 = H , as well as 1-acenaphthylenecarboxylic acid and its esters) leads to four isomers whose relative proportion depends on the nature of the medium (solvent, micelle solutions(refs.30.31) It is also worth noting that the photodimerisation of acenaphthylene in micelles of non ionic and anionic surfactants gives good yields ( > 90%) of photodimers, even with low concentrations at which no dimers are formed in benzene (ref.32); acenaphthylene was shown to photodimerise also when it is adsorbed on silica gel (ref.21).
-
TABLE 1 Photodimerization and reversion of acenaphthylene & (R1 = R2 = HI and its 1-cyano derivative & (R1 = CN,R2 = HI in solution (ri€s.
.
28,29,33,34) +7im concentration.
+
is the dirnerisation quantum yield at infinite
+H
H
0.33
83%
*CN
H
0.15
92%
deaerated cyclohexane, yields.
0.85
17%
0.74
8%
* degassed diethyl ether, ‘f preparative
The photodissociation of these interesting systems has not been studied in detail except for acenaphthylene itself ( 9 ,lo; R1 = R2 = H) (ref.33) : the process demonstrated to be adiabatic (ref.34) was found to be very efficient ( q d i s s = 0.8). The other photoproducts with (R = CN, C02H ) decompose readily especially the anti isomers into the starting materials, either upon heating or UV light irradiation.
...
2.3 Cinnamic acids and related comvounds $-trans-Cinnamic acids do not usually photodimerise in fluid solution, owing to their short excited lifetimes; only cis-trans isomerisation is known to occur. Photodimers were isolated from
546
the solid state (which can display three different crystalline modifications) in which the mutual arrangement of molecules controls the reactivity (topochemistry) (refs.35,36). Thus crystals of a-form && give rise only to a-truxillic acid 12 whereas The P-truxinic acid 13 is isolated from the p modification (110). $crystals are photostable owing to the large distance (>4A) between the reactive centres.
a-truxil1 i c acid
12 -
Although these solid state bimolecular photoadditions are well documented. information about the reverse process is scanty. Irradiation of neat liquid ethyl cinnamate 14 (ref.37) at room temperature and of ethyl cinnamate glasses at 8 0 ° C and - 196OC produces cyclodimers in high yields. Six of the eleven possible diastereoisomers have been identified in the products; the distribution of the cyclodimer stereoisomers is attributed to the immobilization of the reactants in the solid glass which causes the photoproducts to be determined by the mutual orientation of the reactant pair in the matrix.
-
thin film 14 -
547
Irradiation of the photodimers with 220 nm UV light back the corresponding starting monomers.
gives
2.4 Miscellaneous
Of interest, although the reverse process has not been mentioned, are the first examples of photodimerisation of an a, P unsaturated carbonyl 15, reported by Ciamician and Silber (ref.38). Direct irradiation (ref.39) of this compound in ethanol or in isopropanol/benzene solution yields cyclobutane although irradiation in the presence of uranyl chloride (refi40) results in cyclobutane s b .=a is suggested to be formed from a singlet excited state in agreement with theory (refs.4 b,c) and 16b from the excited triplet state.
-
9
hv
(y
+-CH=CH-CO-CH=CH-~ 15 -
COCH=CH@
o H c o C H = c16a H@
EtOH or
-
iPrOH/benzene
urany chloride
Among known reversible (2+2) cycloadditions, the following systems could be representative and potential candidates for photochromic applications. Some derivatives of thymine 17 (R1 = R2 = H) were recognized to photodimerise and restore the parent molecule under higher energy UV light.
280nm
I
R1
17 The quantum
yield
239nm
0
Rl
0
18 -
of dimerisation of
dimethylthymine =a
548
(R1 = R2 = CH3) is high at 80R ( = 1) in 1:l mixture of ethylene glycol and water, and 600 cycles (dimerisation - reversion) were performed without producing any detectable changes (ref.41). However, the reaction does not work well at room temperature. A reversible photodimerisation of the amphiphilic thymine derivative 17 (R1 = H and R2 = nC18H33) is observed in LangmuirBlodgett films. The reaction gives the trans-anti dimer in high yield (ref.42) Some ketosteroids Y.9, were also found as enones or dienones (r-fs.43,44)to dimerise under W light and the cyclobutanes were cleaved under heating.
.
a
&+ 0 ?eEt
0
"
19
20 -
Photodimerisation of 2-phenylbenzoxazole 21 produces a "ht" photodimer involving two carbon nitrogen double bonds (ref.45).
21 -
X = H,F,C1
H+,A or hv'
Ar 22 -
'ht"
The photodimer 22 is labile in fluid solution but thermally stable in the solid state at room temperature. This system has been proposed for light energy conversion, since the thermal reversion of the photodimer to the starting material 21. releases 116 kJ/mol.
549
Sakurai (ref.46), showed that irradiation of the following silacyclopfntadiene 3 , through 'Pyrex' gives. in tetrahydrofuran, the anti "ht" photodimer 24 which reverts to the monomer under 2 5 4 nm light. In benzene or ethanol, two other photodimers were isolated. hv, Pyrex
/ \
254nm
23 -
THF
I 24 -
@
hv
H
H
H
benzene or EtOH
Photopolymerisation, in the solid state, might provide a convenient source for photochromic systems. Olefinic compounds 25 of which the reactive centres are <4A apart were shown to form intermolecular [2+2] photocycloadducts 26 (ref.47).
25 -
26
c
The compounds can be converted into a linear high molecular weight crystalline polymer containing cyclobutane rings in the main chains; this type of polymerisation was called "four centre type polymerisation" (ref.47). The polymers are soluble only in strong acids. They all have high melting points and extremely high crystallinity. Some of them can be decomposed, both
550
thermally and photochemically .into their starting compounds. Irradiation of 5,5',1,4-phenylene-bis(2-cyano-2,4-pentadienOiC acid) (PCPA) derivatives 27 using near W light 340 nm < < 430 nm produced a bleaching of the yellow original material (film) (Fig.4). The photoproduct could be dissociated back to a yellow substrate at x < 310 nm (ref.48).
FN
RCO n
'C=CH-CH=CH=CI~=CH-CH=C, /
NC
COR
27 (yellow)
hv >340nm
hv1<310nm
CN
I
Abs.
11
Fig. 4 Change in UV spectrum of (PCPA-di-n-hexyl ester) 27 film irradiation with 358nm light, from ref.48, with after 0.5h ( - - - - ) permission.
551
Photosensitive poly(methacry1ates) 29 containing photodimerisable styrylpyridinium or styrylquinolinium groups were described by Ichimura (ref.49). Sensitivity was observed at 488nm. But the reverse process does not seem to have been reported. 3
INTRAMOLECULAR (2+2) PHOTOADDITION 3.1 BisphenanthrenEs
The phenanthrene ring was early involved (ref.50) in clean photoreversible intramolecular cycloadditions (30# 31 : 32 e 3 3 )
-
hv 366nm
31
30 -
L
hv 366nm A or hv'
33 -
32 -
hv (350nm)
hv ' (d 31 3nm)
or A
a. X = CO-0-Co
b. X = CH2-CH2 c. X = (CH2)3 d. X = SiMe2-O-SiMe2
e . X = CH2-O-CH2
-
(Ref. 16)
( ( ( (
I' 'I I' 'I
51) 52) 53) 54)
36 (syn h h ) (R = C02H,CH20H)
552
Later, to
independently, 9-phenanthryl anhydride (ref.16) was shown
photoisomerise in contrast to other bisphenanthrenes such
34 b-d.
as
The advantage of the short ether chain C H 2 O C H 2 has been demonstrated (ref.54) and here it allows a e to smoothly photocyclomerise into s e . The absorption spectra of =a-e (with a band tail down to 350 nm) are different from that of =a-e
(Fig. 5 )
and the photocyclomers can be photodissociated at 313 nm where their absorption is higher than that of the starting phenanthrene,leading to a photostationary state.
Fig. 5
Electronic absorption spectra of bis(9-phenanthry1)methyl
ether x e (-1 cyclohexane).
Photocyclomer
and its photoisomer s e (----)
=a
(c
=
was chemically transformed into the syn hh
CH2OH) which could be converted to cycloadducts 36 (R = C 0 2 H , the corresponding monomers by irradiation with light of X <313 nm in rigid glasses at 77K. In that connection, some 1,3-di(n,mphenanthryllpropanes Q c investigated by Zachariasse (ref.52) display intramolecular excimer fluorescence but they were found
to be photostable.
553
3.2 BisacenaDhtnylines At room temperature, deep yellow degassed (N2 bubbling solutions (cyclohexane)) of 1,3-bis-(l-acenaphthyl)-l,l,3,3-tetra methyldisiloxane =a, bleach under UV light ( X = 365 nm), owing to the formation of the photoisomer =a (ref.55).
300
400
Electronic absorption spectra of 1,3-bis-(l-acenaphthyl)(-) and its photoisomer %a (----) in cyclohexane, taken from G . FBlix, thesis, University of Bordeaux 1, 1980, with permission.
Fig. 6
1,1,3,3-tetramethyldisiloxane =a
UV light emitted by a low pressure mercury lamp (254 nm) readily cleaved the photocyclomer and a photostationary equilibrium was reached after 15h of exposure. Thermal reversion smoothly occurred at 265OC.
(gYb
365nm
0
\ / \ I a . Y =-Si-0-Si-
b. Y =-CO-O-CO-
0
254nm or A
37
-
38 -
(white powder)
Similarly, 1-acenaphthyloyl anhydride a b photoisomerised providing the syn hh cyclomer, subsequently hydrolysed into the corresponding diacid (ref-28). No mention was made of the reverse process although these photoproducts should decompose under heating (refs.28,29).
554
3.3 Biscinnamates and related conpounds Srinivasan(refs.56,57) demonstrated the potential ability for internal sensitized photocycloaddition of cinnamate chromophores
39,
yield
separated by several bonds (up to 35 bonds). The chemical of the closure reaction ranged from 40 to 50%. The effi-
cient reactivity of bis cinnamates was also used for designing photosensitive cation binding polymers 41 containing pendant pho-
0
todimerisable
groups (ref.58).
0
Irradiation of polymers is
fol-
lowed by an enhancement of their cation binding ability owing to the formation of cyclobutane rings which bring together the complexing units (benzo crown ethers).
555
The reverse reaction was not investigated on these materials. But the photocycloaddition and the photodissociation were studied on cyclophanes of related structure. With a a (n = 1) 92% of the p isomer {the photoadduct has a plane of symmetry) was formed with very few side reaction products; the presence of alkali metal cations was not found to improve the photocyclomer yield (ref.5 9 )
.
43 -
44 -
At 220 nm a steady state was attained within 2 hours but the cleavage reaction did not give back a high yield of the starting product. More sophisticated systems containing diazacrown ethers (Qb) were shown to display photoreversible reactions associated with net modifications of the extraction abilities towards alkali metal ions (ref.60), especially for L i ' . A reversible photoreaction was also observed with the following system 45 (ref.61). The main cycloadducts, 46, produced, with slow rates, were of the &-form which confers to the system a Poor cation complexing ability (ref.61).
45 n 3 . 4 Miscellaneous
= O,i,z
46 (&form) -
Bis-thymines, maleimides, coumarins, furocoumarins, pyrones, acyclic bisolefins (as cinnamates, vide supra), photoisomerise
556
under W light leading to cyclobutane rings (ref.62) which could undergo reverse processes (see § 2.4).
Thus, 1,l' linked bisthymines 47 yield exclusively the cissyn product @ - (ref.63). The photoreaction showed a marked dependence on the chain length, in relation with its dynamic properties (the propane chain is the most efficient). Reirradiation with light of 254 nm photodissociates the cycloadduct leading to a photoequilibrium state (ref. 64). 0
-
300nm
254nm
I O
r 0
-
CH3
47 n
= 2,3,4,6
48 -
Photochemistry of benzenes could also be expected from the following example :
of
interest as
OMe
49 -
f
OEt
neat
50
hv
180"C, 5mn
52 -
QOEt
51 -
When ortho-cyanoanisole 49 (ref.65) was irradiated in sunlight with vinyl ethyl ether, the initially produced intramolecular bicyclooctadiene 50, thermally rearranged into a photosensitive trisubstituted cyclooctatriene 51 which generated a strained
a.
a
557
bicyclooctadiene The conjugated cyclooctatriene was restored in five minutes when the last product 52 was heated at 180OC. It is interesting to note the intramolecular (6+6) (formally a "4n" process) photodimerisation of the benzene ring incorporated in a rigid system (ref.66); the adduct 54 can be photodissociated, a photostationary state being reached at 254nm.
4
SUMMARY AND CONCLUSION
A considerable number of (2+2) photocycloadditions is known. AS shown in this chapter, among them, some aromatic molecules and cinnamic like compounds give photoreversible or thermally reversible cycloadducts. The spectral variation between the addends and the cyclobutane cycloadducts can be modulated acc-ording to the selected chromophores within a range of about 350 nm (250-600 nm). The reactions were studied in organic solvents but also in polymers and in the crystalline state. A number of these reactions Were found to be clean, giving rise to a single product, in an intramolecular process which is of advantage to perform Cycles Of cycloaddition and dissociation in a solid matrix. Like in many other organic molecules, oxidation can be a limitation for their application in devices but it seems that the photochromic properties of these systems have not yet been systematically investigated. ACKNOWLEDGEMENTS The authors thank their coworkers cited in the references, in Particular Dr. R. Lapouyade, and Prof. A . Castellan who contributed to the research effort in the field o f phenanthrenes and acenaphthylenes, and the "Centre National de la Recherche Scientifique" together with the "MinistBre de 1'Education Natimale" for financial support.
558
REFERENCES 1 2 3
4
5 6 7 8
9 10
11 12 13 14 15 16 17 18 19 20
21 22 23 24 25 26 27 28 29
W.L. Dilling, Chem. Rev., 69 (1969) 845. G. Kaupp, Angew. Chem. Int. Ed. Engl., 13 (1974) 817. M.D. Cohen and G.M.J. Schmidt, "Reactivity of Solid", Elsevier, Amsterdam, (1980) 556. a/ R. Hoffmann and R.B. Woodward, Acc. Chem. Res., 1 (1968) 17. b/ I. Fleming "Frontier Orbitals and Organic Chemical Reactions" J. Wiley and Sons, London, 1976 pp.210-223. c/ W.C. Herndon, Chem. Rev., 72 (1972) 157. M.J.S. Dewar, Tetrahedron Suppl.8, Part I (1966) 75. H.E. Zimmerman, J. Amer. Chem. SOC., 88 (1966) 1764 and 1566. R.A. Caldwell and D. Creed, Acc. Chem. Res., 13 (1980) 45. S. Farid, J.C. Doty, J.L.R. Williams, J. Chem. S O C . Chem. Comm., (1972) 711. D. Creed and R.A. Caldwell, J. Amer. Chem. SOC., 96 (1974) 7369. T. Miyamoto, T. Mori, Y. Odaria, J. Chem. SOC. Chem. Como., (1970) 1598. K. Mizuno, C; Pac and H. Sakurai, J. Amer. Chem. Soc., 96 (1974) 2993. R.A. Caldwell and L. Smith, J. Amer. Chem. SOC., 96 (1974) 2994. R.A. Caldwell and T . S . Maw, J. Photochem., 11 (1979) 165. W. Ried, H. Schinzel, A.H. Schmidt, W. Schuckmann and H. Fuess, Chem. Ber., 113 (1980) 225. M.V. Sargent and C.J. Timmons, J. Chem. SOC., (1964) 5544. E.A. Chandross and H.T. Thomas, J. Amer. Chem. Soc., 94 (1972) 2421. R. Galante, R. Lapouyade, A. Castellan, J-P. Morand and H. Bouas-Laurent, C.R. Acad. Sci., 277C (1973) 837. C. Courseille, A. Castellan, B, Busetta, M. Hospital, Cryst. Struct. Comm., 4 (1975) 1. R.S. Harvey, B.V. McNally, C.J. Timmons, and S.C. Wallwork, Ibid, 3 (1974) 747. H. Bouas-Laurent, R. Lapouyade, A. Castellan, A. Nourmamode and E.A. Chandross, Zeitschrift fiir Phys. Chem. N.F., 101 (1976) 39. P. de Mayo, Pure Appl. Chem., 54 (1982) 1623. D.O. Cowan and R.L. Drisko, 'Elements of Organic Chemistry', Plenum Press, New-York and London, 1976 p.435. K. Dziewonski and C. Paschalski, Ber. 45 (1912) 2491 and 46 (19131 1986. G.W. Griffin and D.F. Veber, J. Amer. Chem. SOC., 82 (1960) 6417. E.J. Bowen and J.D.F. Marsh, J Chem. SOC., (1947) 109. R. Livingston and K.S. Wei, J. Phys. Chem., 71 (1967) 541. Amer. Chem. SOC., 98 (1976) J. Koziar and D.O. Cowan, J 1001 and ref. therein. A. Castellan, G. Dumartin, R Galante and H. Bouas-Laurent, Bull. SOC. chim. France, (1976 217. A. Castellan, G. Dumartin and H. Bouas-Laurent, Tetrahedron, 36 (1980) 97.
559 30
31 32 33 34 35 36 37
38 39
40
41 42 43 44
45 46 47 48 49
50
51 52 53 54 55 56 57 58 59
H. Mayer and J . Sauer, Tetrahedron Lstt., 24 (1983) 4091 and 4095. V. Ramesh and V. Ramamurthy, J. Photochem., 24 (1984) 395. Y. Nakamura, Y. Imakura, T. Kato and Y. Morita, J. Chem. SOC. Chem. Comm., (1977) 887. a/ R. Livingston and K.S. Wei, J. Phys. Chem., 71 (1967) 541. b/ N.Y.C. Chu and D.R. Kearns, J. Phys. Chem., 74 (1974) 1255. S. Yamamoto, M. Hoshino and M. Imamura, Chem. Phys. Lett., 98 (1983) 615. M.D. Cohen and B.S. Green, Chem. Brit., 9 (1973) 490. M . D . Cohen, Angew. Chem. Int. Ed. Engl., 14 (1975) 386. P.L. Egerton, E.M. Hyde, J. Trigg, A. Payne, P. Beynon, M.V. Mijovic and A.Reiser, J. Amer. Chem. S O C . , 103 (1981) 3859. G. Ciamician and P. Silber, Ber. Dtsch. Chem. Ges., 42 (1909) 1386. P. Praetorius and F. Korn, ibid., 43 (1910) 2744. G.W. Recktenwald, J.N. Pitts Jr. and R.L. Letsinger, J. Amer. Chem. SOC., 84 (1962) 2344. J. Eisinger and A.A. Larnola, Mol. Photochem., 1 (1969) 209. E. Yano, S . Tatsuura and K. Ikegami, Proceedings of the XIIth IUPAC Symposium on Photochemistry, Bologna, Italy, July 1722, 1988, Vol. of Abstracts p.232. M.B. Rubin, G.E. Hipps and D . Glover, J. Org. Chem., 29 (1964) 68. M.B. Rubin, D. Glover and R.G. Parker, Tetrahedron Lett., (1964) 1075. N. Paillous, S . Fery-Forgues, J. Jaud and J. Devillers, J. Chem. SOC. Chem. Comm., (1987) 578. Y. Nakadaira and H. Sakurai, Tetrahedron Lett., (1971) 1183. M. Hasegawa, Chem. Reviews, 83 (1983) 507. F. Nakanishi, K. Yamada and H. Nakanishi, J . Polym. SCi. Polym. Chem., 26 (1988) 329. M.K. Ichimura, J. Polym. Sci. Polym. Chem., 25 (1987) 3063. C.H. Krauch, S . Farid and D . Hess, Chem. Ber., 99 (1966) 1881. D.S. Tarbell and W.P. Wystroch, J. Amer. Chem. SOC., 65 (1943) 2149. K.A. Zachariasse, R. Busse, U. Schrader and W. Kiihnle, Chem. Phys. Letters, 89 (1982) 303. G. Felix, unpublished results. Nouveaux Derives Bis-Anthryles et Bis-Phenanthrylbs, R. Lapouyade, G. FBlix and H. Bouas-Laurent, F.P. 2,314,168 App1.75/18,559, 13 June 1975 (C.A. 87 (1977) 586, 1177299). M. Laguerre, G. Felix, J. Dunogues and R. Calas, J. Org. Chem., 44 (1979) 4275. J.A. O r s and R. Srinivasan, J. Chem. SOC. Chem. Comm., (1978) 400.
J.A. O r s and 315. M. Shirai, A. (1985) 501. S. Akabori, SOC. Jpn., 61
R.
Srinivasan, J. Amer. Chem. SOC., 100 (1978)
Ueda and M. Tanaka, Makromoleculare Chemie, 186
T.
Kumagai, Y. Habata and S. Sato, Bull. Chem. (1988) 2459.
560
S. Akabori. T. Kurnagai, Y. Habata and S. Sato, J. Chern. StC. Chem. Comm., (1988) 661. 61 S . Akabari, Y. Habata, M. Nakasawa, Y. Yamada, Y. Shindo. T. Sugimura and S . Sato, Bull. Chem. SOC. Jpn., 60 (1987) 3453. 62 F.C. De Schryver, N. Boens and J. Put, Adv. Photochem.; J.N. Pitts, G.S. Hammond and K. Gollnick Ed., Vol. 10, pp. 359-465. 63 N.J. Leonard, R.S. McCredie, M.W. Logne and R. Cundall, J. kmer. Chem. SOC., 95 (1973) 2320. 64 R. Beukers and W. Berends, Biochim. Biophys. Acta, 41 (1960) 550 ; 49 (1961) 181. 65 A. Gilbert and P. Heath, Tetrahedron Letters, 28 (1987) 5909. 66 G. Sedelmeier, W.D. Fessner, C. Grund, P.R. Spurr, H. Fritz and H. Prinzbach, Tetrahedron Lett., 27 (1986) 1277.
60
561
Chapter 74
Cycloaddition Reactions Involving 4n Electrons: (4+4) Cycloaddition Reactions between Unsaturated Conjugated Systems
H. Bouas-Laurentand J-P. Desvergne 1
INTRODUCTION
1.1 Theoretical considerations The ( 4 TTs + 4 l l s ) photocycloadditions are among the oldest known and,with the (2 TIs + 2 T l s ) cycloadditions, constitute an important group of photochemical reactions. They can be formally depicted by equation (1).
It says that melectrons are involved in the reaction, and " s " denotes "suprafacial" which means that bonds are made or broken on the same face of the reacting system ( 2 ) .
According to the Woodward-Hoffmann rules (ref.11, this reaction is allowed in the excited state and forbidden in the ground state. The rules are based on the conservation of orbital symmetry (ref.2) along the reaction path as indicated in the state correlation diagram (Fig.1); the scheme also applies to the reverse reaction, i.e. photodissociation of the cycloadduct; therefore these systems should be photochromic. Dewar (ref.3) used the "Perturbation of Molecular Orbitals" approach to show that the (4+4) cycloaddition is a particular case of reactions involving 4n (n = an integer) electrons in the transition state
562
termed
"antiaromatic"
which is of lower energy in
the
excited
state: this statement does not imply any symmetry requirement and is of a wider application. Independently, Zimmerman (ref.4) and others (ref.1) arrived at the same conclusion. Consequently,
the rules are valid for the derivatives of the
polyenes examined as well as for the parent compounds.
D
D
S
S G
G APDPNDS
CYCLOAODUCT
Fig.1 Simplified state correlation diagram for the (211s + all,) and (4nS+ 4 TIs) cycloaddition (after ref.1). G = Ground state: S = Singly excited state, D = Doubly excited state. At midpoint of
the reaction, there is an avoided crossing (----I between the two diabatic surfaces, generating a "pericyclic" minimum in the excited state surface and a maximum in the ground state surface (ref. 2 ) . 1.2 Overview of the (4+4) cycloaddition reaction Since the literature on (4+4) photocycloaddition is vast, it would be impossible to present a comprehensive survey of all the reactions reported. Several reviews have covered various aspects of cycloaddition reactions (refs.5-9). In this chapter, a selection has been made of cases relevant to actual and potential photochromic reactions.
Molecules known to date as being involved in such processes are mainly of the anthracene type. In fact, Fritzsche who disco-
vered (ref.10) the photodimerization of anthracene, also observed that the monomer could be quantitatively regenerated from photoproduct by heating; but nobody, prior to Chandross (ref.111, recognized that photodissociation could also be effective. A great den1 of work has been devoted to anthracene and anthracene derivatives sinct? the early days and the research in this field
563
is still active (ref.12). It includes other aromatic compounds such as naphthalene derivatives, benzanthracenes and some heterocyclic compounds (Fig.2). The reaction can either be intermolecular or intramolecular; in the latter case, although kinetically monomolecular, the fundamental process is basically a bimolecular photocycloaddition since the two reacting parts linked by a chain (bichromophores) have to diffuse together. For cyclophanes, such extensive molecular motion is not required usually. The CYCloadducts are then termed photocycloisomers, abbreviated "photocyclomers" (Fig.2). Inter or intramolecular crossed (or mixed) Photodimerizations between two different compounds are also considered. They should be termed "cophotodimers". Let us note that intermolecular processes are seldom involved in photochromic reactions. The mechanism of the cycloaddition and its reverse photochemical and thermal processes are best exemplified by that of anthracene in the next section.
Fig. 2 Examples of molecules undergoing ( 4 + 4 ) photocycloaddition : 1) intermolecular. 2) intramolecular processes a/ bichromophores b/ cyclophanes (see text).
564
MECHANISM OF PHOTODIMERIZATION AND CYCLOREVERSION 2.1 Photodimerization of the anthracene rinq The main photochemical reactions undergone by anthracene are photodimerization, photooxidation leading to an endoperoxide (see Brauer and Schmidt, this book) and single electron transfer, in the presence of an oxidant or a reducing agent generating respectively a cation-radical or an anion-radical (refs.13,14). The photodimerization of anthracene can be considered as the paradigm of the ( 4 + 4 ) photodimerization. The subject has been partially surveyed in several articles (refs.12,15) and in textbooks, especially that of Cowan and Drisko (ref.5). 2.
Fig. 3 The photodimerization of anthracene is a model of 41Ts) cycloaddition.
(4rTs+
In an aerated diluted solution, both endoperoxide and photodimer can be produced by irradiation through Pyrex. In degassed and concentrated solutions, the photodimer is the only product: the same is true, in general, for most of its derivatives in a variety of solvents and other media (crystals, molecular assem1. blies, polymers
...
,m,, 3
6
5
10
4
Fig.4 Numbering of the anthracene ring and, in parentheses, a (Or topological designation of the different positions : a, meso). The surface of the circles is proportional to the order of reactivity : 9 > 1 > 2 (ref.16).
b,r
Originally, it was assumed that anthracene reacted through the so-called 3 carbon atoms (ref.17) and the molecular weight of the photoproduct was first carefully determined by Orndorff and Cameron (ref.18) , who proposed the name of "dianthracene".
565
Hengstenberg and Palacios published, in 1932, (ref.l9), the first data on the crystallographic structure and proposed that the bonding between the two rings is effected between the 9,lO and 9',10' positions (see Fig.3). Coulson came .to the same conclusion in 1955 (ref.20) using UV-absorption spectrometry, demonstrating that the spectrum of dianthracene is made up of approximately the sum of 4 ortho-xylene units (Fig.5). Ehrenberg refined the X-ray structure analysis in 1966 (ref.21) and confirmed the structure of dianthracene represented in Fig.3. A great number of derivatives of anthracene are known (more than a hundred) to yield photodimers which are thermally stable at room temperature; these compounds are, in general, poorly soluble in organic solvents and do not show points, due to their being decomposed on heating.
sharp melting
log€
5' 4 -
32. I
I I
250
I
350
Xbm)
Fig. 5 Electronic absorption spectra of anthracene (A) (-) and dianthracene (A2) ( - - - - ) in cyclohexane. The first electronic transition (lLa) spans from z 290 to 370 nm and the second 'Lb (forbidden) is hidden under the lLa band; the third absorption ('Bb) peaks at = 255 nm. ~ , , b and ~,,b are Platt notations (ref.22) (taken from ref. 106, with permission). Anthracene smoothly photodimerizes upon irradiation (and not by a ground state reaction) in a variety of solvents such as aliphatic or aromatic hydrocarbons, alcohols, ethers, dichloromethane or chloroform and also in the solid state (ref.23) ; heavy atom solvents (CgHgBr, C ~ H S I , CS2 ... ) known to enhance intersystem crossing, were found to inhibit the photodimerization
...
566
(ref.24). Photophysical and kinetic studies of anthracene and some of its derivatives (refs.5,15a,d) led to the conclusion that the first singlet excited state is responsible for the reaction. The quantum yield is not dependent on the excitation wavelength (ref.25). Although one case is known of a triplet state dimerization (3A* + A 4A2) for a bisanthryl ketone (ref.26), the other observed facts are in agreement with the following kinetic scheme (Chart 1 : A = anthracene).
Chart 1
Kinetic scheme for anthracene photodimerization through the S1 state. kF, kIC, kISC, kdim, kCQ are rate constants for respectively fluorescence, internal conversion, intersystem crossing, dimerization and concentration quenching. r 1 = l/kl.
The Stern-Volmer relation for fluorescence follows :
+;/+F
= 1
(a)
k2 11 [A1 The dimerization quantum yield can be derived : +
+dim = kdim IAI/(kl + k2 [All At infinite concentration of A, +mdim = +mdim/+dim = 1 + l/k2~1[A]
kdim/k2
;
then
(b)
Evidence for a singlet state mechanism is given by the experimentally defermined slope of equation (a) to be equal to the inverse of the slope of relation (b) (ref-s. 12, 5). Some relevant data are given in TABLE 1.
567
TABLE 1 Kinetic data for the photodimerization of anthracene in different solvents at 298K. Second order '109M-1S-1
Solvent
i1 (ns)
+
-dim
Benzene
4.1
0.34
Toluene Cyclohexane
4
0.16 0.35
-
F kdim
kCQ
k2
ref.
2.3 1.0
4.3 4.0
6.6
27 24
2.25
-
11.8
-
5.0
14 10
27,28 29
The dimerization quantum yield was found I0.04 at 0.005 M and 0 0.07 at 0.01 M in benzene (ref.27). Livingston found the following empirical relation (c) for air-free diluted solutions (ref.30): +dim = 3.3 [Anthracenel
(C)
At the same time, the question of the intermediacy of an exciPer (1:l sandwich complex stable in the excited state and repulsive in the ground state) was raisaa. Although several anthracenf derivatives exhibitad excimer fluorescence in solution, no eXCimeK formation had been detected fcr the parent molecule under these conditions, prior to the report by Yang in 1976 (ref.31) who used concentrated ( ~ x ~ O - ~ chloroform M) solution. That anthracene is able to form an excimer, had been demonstrated by Chandross (ref.11) by photodissociating the photodimer in a rigid matrix at 7 7 which ~ led to two arithracene molecules being constrained in close proximity in the form of a sandwich emitting Pair. The first evidence of the intermediacy of an excimer in the formation of dianthracene came from Ferguson in 1974 (ref.32) who showed that, in a very rigid matrix (dianthracene single crystals), the decrease of excimer fluorescence is compensated by an increase of photodimerization quantum yield over a wide range of temperatures (20 - 180K). Cohen et al. (ref.33) demonstrated
568
that the same is true in solution (toluene-ethanol) by quenching the non-emitting excimer with a dye,leading to an estimation of its lifetime being ca 1.2 ns. Several other authors reached the same conclusion for anthracene derivatives (ref.34) especially for [2.2](1,4) (9,lO) anthracenophane (ref.35) and bisanthracenes (refs.36,37). The intermediacy of an excimer in other cases of anthracene photodimerization is now often postulated. It is in line with the accepted mechanism of the (211s + Z n S ) photoaddition reaction (ref.38). Accordingly, the kinetic scheme should be modified when an excimer is kinetically significant (Chart 2).
Chart
2
A + 'A* 1 (AA)*
(AA)* (AA)* (AA)*
DM
___3
____, ___)
L
___c
A +'A* 2A 2A
A2
(7)
~ M D ~ F D ~NRD ~ R D
Kinetic scheme of anthracene photodimerization through a singlet excimer state : l(AA)** kDMi kMDi kFDf kNRDt kRD are Excimer Formation, Dissociation, Fluorescence, Non Radiative Deactivation and Dimerization rate constants. Some kinetic data, listed in TABLE 2.
determined according to
Chart
2,
are
The photodimerization of anchracene (ET zz 43 kcal mo1-l) can also be sensitized by biacetyl (ET = 55 kcal mol-l) a s shown by BIckstrom in 1960 (ref.40); Saltiel was the first to propose (ref.41) that triplet-triplet annihilation could account for dianthracene formation. Considering again the correlation diagram of Fig.1, there are, theoretically, two pathways for photoaddition, which are indicated by arrows in Fig.6 : an adiabatic path following the singlet excited surface and a diabatic mechanism, suggested by Michl (ref.42) which passes from the singly into the doubly excited surface towards a pericyclic minimum and decaying through a "funnel" to the ground state cycloadduct and addends.
569
TABLE
2.
Estimated kinetic parameters for anthracene photodimerization through excimer in solution; the excimer formation rate is close to diffusion controlled (ref.39). second order (108M-I S-l) Solvent
+*FD
c$)mdim
Toluene
0.16
Chloroformb BenzeneC
0.35c
(a)
c
T
D
(
~
~
)
kRD
kID
~ F D
0.004
2.5
0.6
3.5
0.02
0.Olb
1.5b
2
6
0.07b
Dianthracene monocrystals
5d ~
~~~~~~~~~~~~
~
~
(a) Excimer fluorescence quantum yield extrapolated to infinite concentration of A. (b) Values taken from ref. 31, (c) taken from ref. 39, (d) calculated ar. 298K from activation parameters taken from ref. 32.
In the case of anthrackne, the doubly excited state was shown by Saltiel and Charlton (ref.43) to arise from interaction of two triplet states (TI) of anthracene 3A*, in a careful study of the D D
S
S
G ADDENDS
G CYCLOAPDUCT
Fig. 6 Theoretical state correlation diagram for "4n" cycloaddition (see Fig.1). +:adiabatic reaction diabatic pathway P : pericyclic minimum (common intermediate) (adapted from refs.1,2,42 and 43)
+
570
dependence of dianthracene quantum yields on light intensity for direct anthracene excitation. Therefore, the kinetic scheme should include additional processes which obtain in the absence of oxygen and at reasonably low concentration (Chart 3 ) . This experimental work supports the postulated doubly excited state as
Chart 3 3A*
+
3A*
~(AA)** 3A* + A
___)
___,
(i) among other products
a viable intermediate for photocycloaddition. An energy for anthracene dimerization is proposed in Fig. 7 .
profile
Fig. 7 Energy profile for the photodimerization of anthracene. The reaction coordinate is the distance between the meso carbons P : Pericyclic minimum. of the two reacting rings. E : Excimer The diagram is very approximately scaled. At the pericyclic minimum, the reactants should assume the structure of a common intermediate (C.1.); as will be discussed
571
later
with the photodissociation, let us stress that the sum of
dianthracene formation and photodissociation quantum yields was often found to amount to about unity. pointing to a common intermediate. Some authors, in particular Kaupp (ref.44) have long favoured a diradical structure, others a diradicaloid (or zwitterionic) complex as proposed by Michl (ref.42); as yet the intermediate has not been trapped, presumably because its lifetime is too short. Recently, a common intermediate (C.1.) was detected by picosecond laser flash spectroscopy of bisanthracenes (ref.45). The same spectra were observed by exciting the addends at 355 nm and the adducts at 266 nm. The electronic structure of the C.I. is different in polar vs non polar solvents; in particular, in CH3CN, it has some charge transfer character and in (C2H5)2Ot it could not be defined. More work is necessary to get a better understanding of these very fast processes. 2.2
Photodissociation Photodissociation
thermal cvcloreversion
--_-----------___
The photodimer of anthracene dissociates by irradiation at wavelengths shorter than B 300 nm (see Fig.5) in solution or in the solid state. But anthracene absorbs light in this wavelength range and the photocycloaddition occurs in competition; under continuous irradiation, a photostationary state can be established. Assuming that the dimerization quantum yield is the same at X <300 nm and at 365 nm (the usual excitation wavelength of anthracene) (see ref.25), it is possible to derive the dissociation quantum yield ( diss), which does not depend on concentration. Several data are listed in TABLE 3 . The mechanism of photodissociation was examined in detail by Grellmann and Yamamoto (ref.47a); it is strongly affected by temperature. The study was extended by the authors to other dimers such as di-9-methylanthracene and the cophotodimer of anthracene and naphthacene (ref.47b). The mechanisms are, in general, complex and the main following features emerge : the photodimers of anthracene and 9-methylanthracene photodissociate around 300 K mainly via the excited singlet state and around 77 K exclusively via the triplet state; at intermediate temperatures, the involvement of the triplet state becomes important. One of
512
the most interesting results is that photodissociation was found to be wavelength dependent between 248 and 292 nm. TABLE 3 Photodissociation quantum yield of dianthracene in solvents and at several temperatures. Solvent (temperature)
4 diss
Cyclohexane (283-343K) Acetonitrile (298K) Cyclohexane
(298K) Methyltetrahydrofuran * (300-1OOK) 93K 8 3K
Excitation wavelengthA(nm)
0.63
254
0.55
269
0.64
#
different
ref.
-
265
0.55 0.35 0 -22
*freeze-pump-thaw cycles degassed
'
+Tim
= 0-35
The cycloreversion of dianthracenes can also be sensitized, through formation of charge-transfer complexes or photoinduced bY electron transfer as shown by Roth (ref.48) and de 14ayo (ref.49) for dianthracene with chloranil or 9,lO-dicyanoanthracene; it was further studied by Kochi for dianthracene derivatives with tetracyanoethylene (ref.50); the dianthracene cation radical formed by
Z = H
Z = H,Me,CHO,CN MeO,OAc
S = 9,lO-di-CNA
re€.49
S =
ref.50
or chloranil tetracyanoethylene
573
electron transfer to the electron acceptor sensitizer, is readily decomposed into two ground state monomers; the chemical yield was found to be higher than 70% (ref.49) or quantitative (ref- 5 0 ) . Alternatively, de Mayo successfully used a suspension of cadmium sulfide for electron transfer photoinduced cleavage of dianthracene (ref. 51). Thermal cycloreversion
_____-______-_--__----
Very early on, it was recognized that dianthracene can be converted back to its monomer on heating it in solution or in the solid state (ref.10). Activation parameters were obtained in benzene by Greene (ref.52) : Ea = 37 kcal mo1-l (&H# = 36 kcal mo1-l) ; l o g A = 14 (AS# z 3 cal mol-lK-l). A similar Value for Ea ( = 35.5 kcal mol-l) was found by Bendig and Kreysig in cyclohexane (ref.46). 3
INTERMOLECULAR (4+4) PHOTOCYCLOADDITION 3.1 Anthracene derivatives Many
(mono, di and trisubstituted) anthracene derivatives are known to yield 9,9'-10,10' (classical) photodimers (see formulae below). Most of them are poorly soluble in organic solvents and are thermally stable at room temperature, especially in the solid state. Photoreactivity and thermal stability vary according to the position and the nature of substituents. Sf
9-Substituted derivatives
___-__-___---____________ An important question is the regioselectivity; in principle, two photodimers named head-to-head (hh) and head-to-tail (ht) can be formed :
-1
-z h h
-2 ht
574
During the sixties, the structures of isolated photodimers were studied by dipole moments, chemical correlation and OCCaSiOnallY by X-ray crystallography or NMR (refs.12,27). All the photoproducts examined were found to have a "ht" structure
(with the exception of Greene's product (ref.53) but this was recognized later (ref.52 as erroneous) and these results were considered to be general until the first NMR measurements of the mother solution obtained by extraction of the reaction mixture. The first observation of a "hh" photodirner was made by De Schryver (ref.54) from the NMR spectrum of the photoproduct of 9anthrylmethylacetate (hh 20%, ht 80%). Then Kaupp showed that 9methylanthracene, by irradiation in benzene, yields both isomers (hh 40%, ht 60%) (ref.55). Similar results were obtained in other laboratories (refs.12,561 Wolff increased the ratio of the "hh" dimer of 9-hydroxymethyl anthracene by irradiation in aqueous micellar solutions of sodium dodecylsulfate and studied other derivatives under these conditions (refs.57,58). The "hh" dimers are thermally less stable and, to our knowledge, none have been isolated pure so far, but,for some of them,the activation parameters for their thermal cycloreversion could be determined (TABLE 4).
.
TABLE 4 Activation parameters for the thermal cycloreversion of some 9substituted anthracene derivatives. The activation energy of the "ht" compounds are significantly higher than that of their "hh" isomers. (Ea: activation energy: A frequency factor: AH* activation enthalpy,AS? activation entropy) .SDS="Sodium Dodecyl Sulfate" CTAC="Cetyl Trimethyl Ammonium Chloride" Substituent
Medium
M f k c a l mol-I
Ea kcal mol-I
A$cal
mol K-l
ref.
A
b-1)
________________________________________---------------------
Me
(hh) " It
-
He CA20A
"
(ht (hh) I'
(ht) CHZOH (CH2)2C008 (hh) COOH
- coon
(hh)
(ht)
Cyclohexane Ether 0.21 SDS in A20 0.2M CTAC " " Tetralin Ether 0.211 SDS in F20 0-Dichlorobenzene Ether Ether 0-Dichlorobenzene
19.6 26.5 21.7 19.6 38.3 27 26.8 36.6 21.5 28
34.5
6x109 58 6x1Ol4 I, 3 ~ 1 0 ~ ~ 2x1010 I 1x1016 1x1015 77 4xd4 1.6~10'~ 4~10'~ I 7.9x1016 3.2~1014 ,I
575
Recently, 9-alkyloxyanthracenes with long chains ( 3 ) were prepared to obtain pure ht photodimers whose solubility in non polar organic solvents is significantly increased ( = 100 times in benzene and carbon tetrachloride) (ref.59).
OR
&I
-3
hv .Pyre& RT. R = C
H 10 21 1 6H33
Solubility : Benzene ^. 0 .3
CC14
M
= 0.1M
Not all the anthracene derivatives can be photodimerized. Steric crowding, e.g. for 9-tert-butyl and tert-pentylanthracene (s), is a major impediment. These two compounds, however, are photoisomerized (refs.60-63) into Dewar anthracenes ( 6 ) ; although this photovalence isomerization is not relevant to this R
R : -CMe3 : -CMe2Et i) Pyrex, OOC, in n-hexane;+, = 1 . 6 ~ 1 0 -at ~ 300K ii) AH#= 91 (heptane) 94 (EtOH); (kJ mol-l) A Sf: - 25 (heptane)-17 (EtOH) (J ~nol-~K-l)(ref.6la1, R = CMe3). chapter, the system is photochromic and was studied as a potential energy storage system (ref.63). The other anthracene derivative known to generate a similar Dewar photoisomer is decamethylanthracene (ref.64). Another crowded substituent, -SiMej, does not inhibit the photodimerization reaction (ref.65) in cyclohexane but 9-trimethylsilylanthracene undergoes protodesilylation by irradiation in alcohols (ref.66).
576
9,10 Disubstituted anthracenes
---------____________________
Meso-disubstituted derivatives photodimerize provided t h e substituents are not too bulky. Compounds 1 (refs. 67, 68) and a number of dissymmetrical photodimers ( 8 ) have been isolated (refs.67,72) and characterized. Most of these products are thermally unstable, in particular the 9-cyano-10-acetoxyanthracene (ref.70) and 9-methoxy-10-cyanoanthracene (ref.71) photodimers but the di-9-methyl-10-methoxyanthracene could be kept in the solid state without decomposition for a long time (ref.69).
-8
-7 2 = Me (ref.67)
F
(
"
68)
Z Me Et Me Me CN Me0 Me
Y CN
cI?
OCOMe OMe OCOMe CN CH20Me
ref. 69
,, t
70
51 72
derivatives and their cycloreversion.
....................................
The photoreactivity of several 9-(or 9,101 substituted derivatives has been studied and some quantitative useful data are collected in TABLE 5 ; the rate constants were obtained using Chart 1. The photodimerization quantum yield (+dim) of 2 4 anthracene derivatives was determined by Castellan (ref.73) in several solvents (benzene, ethanol, acetonitrile at room temperature, concentration ca. 10-2M). It is observed that+dim is higher in acetonitrile than in benzene; in the same solvent, the differences in +dim could be assigned to steric hindrance, charge dissymmetry in meso positions (ref.74) and heavy atom effect.
577
TABLE 5 Lifetimes Photodimerization quantum yields at infinite concentration ($"dim) and birnoiecular rate constants (see Chart 1) for 9 (Or 9.10) substituted anthracenes, at room temperature. a: conc :: 38 b: conc = 0.001M. Substituents Me CN Et n-Pr
n-Bu Me0 c1 Vinyl
9-Me-lO-CH20Mc
9,lO-diMe COOH
COONa COOEt CONH 2
COOAmyl . I
Solvents
11 +"dim (ns)
Toluene 9.5 Cyclohex. Glass Resin Benzene 13 Toluene 9.2
c
9.4 9.7 6.6
I
Benzene EtOH HZ0
Benzene THF Crystala PMMA~
10.0 12.5 12.5 4.5 1.5 12.9 8.0 1.7
1.9 -
0.24 0.115 0.067
1.3 1.1
0.16 0.3
0.164 0.07 0.05 0.05 0.004 0.02 0.2 0.1
-
'I
z
kdim
0.14
0.075
4.8 10.0
Second order M-1S-1x10-9
5.3 0.3
0.6 0.5 1.6 0.9 0.6 0.2 0.03 0.04 1.0 0.9 0.4 0.8 1.4
Kco
11.8
-
k2
13.7
-
4.1 8.1 7.8 5.9
4diss. 0.81 :: 0.006
5.4 9.2 8.4 6.4 8.2 9.8 9.9 11.8 10.8 11.4 3.8 4.0 c 7.6 7.6 1.8 = 1.85 4.4 5.4
8.4 1.7
1.1 15.7
27,72 46 76 27 27,72
0
27 Id
9.3
2.1 l.3 17.1
.
raf
t
75 z z
0.3 0.003
76
"
The ability of the photodimers to regenerate the nonomers by heating was verified many tines but kinetic measuremnts were performed on only a few dsrivatives !some data are listed in TABLE 4 ) . As expectad, activation e n e r g i e s strongly depend m tile hh 0'1' ht st.ructure and on the enviroiiirient (solvent, niicelles) , as shown by Wolff (refs.58b,77). Other derivatives There are about 40 known photodimers of mono-, di- or trisubstituted anthracenes on positions other than 9 or 9,lO. Ai!lorig them, let us emphasize 1,8-dichloro-9-methylanthracene (za) and its isomer 1,8-dichloro-l0 methylanthracene ( m a ) ; they readily generate their respective photodiners in solution with a higher quantum yield for 2 (ref.73) and this was attributed to the steric hindrance of the 9-methyl substituent between the two bulky chlorine atoms in positions 1 and 8 (termed per1 effect), which decreases the rate of dimerization (ref-781; these compounds (sa and m a ) are a150 easily photodirnerized in the crys-
578
talline
state
and
it
was
shown
that
the
CL
-9
a
b
occurs
CL
R
R = CH3 R = H
+dim (gal : 0 . 0 0 4 and ( u a ) : 0 . 0 0 8 (ref- 7 3 ) .
reaction
10
;
conc 7 . 5 ~ 1 0 - ~ in H ether
preferentially at defects and dislocation sites (ref.79). Simi ar (ref.80). observations were made on $& and Recently 2,6-didecyloxyanthracene (11)was shown to yield, by irradiation through Pyrex, a mixture of the classical 9,lO': 10 (=)and of 9,l' : 10,4' (=)photodimers with a low quantum Yield. This is the first intermolecular formation of bonds between the meso positions of one ring and the side positions of the other ring (ref.81). I-
11 -
Photopolymerization of anthracene derivatives
.............................................
It was anticipated that a molecule formed of two anthracene units linked by a chain in positions 9 and 9' ( g )could lead to polymers (15)by irradiation under suitable conditions and this was demonstrated by De Schryver (ref.54).
579
&
-
= 350nm,CH2C12
A) 180°C
1'4 (0.1-0.2M)
c
/
15 h t
Polymers
'-
15 hh
It is an example of a class of photopolymerization from the S1 state in which each propagation step is the result of a photochemical reaction, and not a chain process like in a photoinitiated polymerization. The molecular weights were determined for 14a (m = 4-7 Mn = 4,500-6,800 and m = 7-8, Mn = 12,000) and for 14b (p = 11, Mn = 52,000; p = 12, Mn = 28,000). In the case of 14b no "ht" photodimers were detected by NMR. The polymers are film forming and can be photochemically or thermally transformed back to monomers. Thus, 15 could be heated with a C02 laser and the irradiated spot was able to emit fluorescence at = 450 nm upon excitation with 365 nm light. These compounds were patented by Agfa-Gevaert A.- G. (ref.82). The generation of stable fluorescent products from non fluorescent precursors has been termed "photofluorescence" (ref.83). Modification of physical properties
-_____________________-------------
a/ In the preceding paragraph (Fig.
51,
the U.V.
absorption
spectra of anthracene and its photodimer show considerable differences. This is accompanied by a change in refractive index in spectral regions where there is negligible absorption, for instance in the visible region. Chandross and Tomlinson (ref.76) took advantage of this property to consider anthracene dimers as potential photochromic materials for holographic applications. In order to improve the cycling efficiency, the authors used a matrix such as " ~ o l y M _ e t h y l M _ e t h ~ c r y l a t e(PMMA) " to hold the two monomers in close proximity; indeed if the breaking efficiency of
580
the photodimer appears independent of the environment, remaking the dimer is markedly matrix dependent. To solubilize the dimer in PMMA, an ester was used, g-amylanthroate,which proved also more stable than anthracene under repeated cycling (cleavage at 3 0 3 nm, redimerizing at 365 nm); but the remaking efficiency was found P 0 . 3 % which means that the matrix does not maintain the pair of monomers in proper alignment for recyclization. Even "Glass Resin" was not found satisfactory. Finally, the photodimer of amylanthroate formed small crystals (a few millimeters) with a remaking efficiency of P 30% but the mechanical properties of these crystals made them unsuitable for applications. An effort to improve the quality of crystals led the authors to investigate dimers of materials which form salts as will be seen in section 3.4 (heterocyclic compounds). b/ Photochromic compounds can induce a variety of changes in physical properties. Wolff and von Biinau (ref.84) have recently observed that aqueous solutions containing 9-substituted anthracenes solubilized in cetyltrimethylammonium micelles show rheopectic, thixotropic and viscoelastic behaviour depending on the composition in monomer and photodimer. For instance, the viscosity of micellar solutions of 9-methylanthracene and of 9-anthracene carboxylic acid can be decreased by irradiation at X > 3 0 0 nm, the viscosity can be increased again by reirradiation at 249nm. 3.2 Naphthalene derivatives and other acenes Naphthalenes
----_-------
The first compound in the naphthalene series which was found to photodimerize is the 2-methoxy derivative, in 1963 (ref.85) I whereas naphthalene itself does not undergo this reaction. Other 2-alkoxy derivatives were shown later to give the same cycloaddition (refs.86.87). The highly insoluble photoproducts exhibit a centrosymmetric structure as determined by X-ray analysis for the methoxy derivative (16)(ref.88). Later, Sasse and his group discovered a second, much more soluble, photodimer to which they attributed the so-called "cis" structure (11) on the basis of its spectroscopic and chemical properties (ref.89). These substances regenerate their precursors on heating and by irradiation with Pyrex filtered U.V., in contrast to the cage dimers (ref.87).
581
(i) in organic solvents such as benzene; R = n-Bu; i-Pr; allyl. Such cage dimers (18)were isolated for the first time in 1972 and their structure was proposed on spectral and chemical grounds (ref. 90). The structure has been confirmed by X-ray analysis for the photodimer of methylnaphthalene-2-carboxylate (ref.91) and it was argued that molecules of type 18 were formed by irradiation of compounds of type 11 (refs.87,92).
A
z z
t: (.
1900C)
(i) in tetrachloroethylene 124OC (4h) or boiling DMF (12 mn). According to Sasse, the photodimerization is believed to proceed via the S1 state (but see poly(1-vinylnaphthalene) (ref. 93)) and to involve an excimer; polar factors such as electrostatic interactions in the ground state seem to be determinant (refs.87,92). The I-isomers have not been found to produce iS0lable dimers. The known photodimers are formed from 2-mOnOsubstituted (alkoxy, CN, COZR) and disubstituted (alkyl, alkoxy, C02R) naphthalenes. Little data on photodimerization and photo-
582
dissociation quantum yields are given in the literature but Sasse (ref.89b) has studied these compounds as potential energy storage systems. The 2-monosubstituted naphthalenes have (ref.94) U.V. absorp-
tion spectrum extending further to 320 nm (log E = 2 at 325 nm) and the photodimers have no significant absorption at wavelengths higher than 300 nm (Fig. 8 ) .
Fig. 8
Electronic absorption spectra of 2-methoxynaphthalene in
n-heptane (taken from Wilairat and Selinger (ref.941, with permission). Although,
in organic solvents, the major photodimer of the 2-
monosubstituted naphthalenes (substituent : OR, CN, C02Me) is
of
"trans" structure, in micellar solutions, the major photodimers were found by Ramamurthy to be of "cis" configuration (Ref.95); the micellar solutions enhance the reactivity by a local concentration effect and improve the regioselectivity of the process by preorientation of the monomers (as observed by Wolff in the anthracene series (refs.58,77,84). In brief, the intermolecular photodimerization of naphthalene derivatives, leads, in general, to two products, of the "trans" and "cis" structures, respectively; the latter can give rise to a
thermal Cope rearrangement or to a cage compound on prolonged irradiation; heating does not always regenerate the monomers and this seems to be a drawback for photochromic applications.
583
Naphthacene
---___-____
By irradiation, naphthacene (19) orange solutions (or suspensions) become colourless (refs.96,97). A centrosymmetric photodimer structure was attributed to the insoluble photoproduct. More recently, it was demonstrated that, in fact, two photodimers are formed : a centrosymmetric one (2511, only slightly soluble, and its planosymmetric isomer (2)more soluble in organic solvents (ref.98); the latter had thus escaped detection.
+ 19 -
20 -
21 -
a
easily revert to 19 in solution by gentle heating. Both and Photodissociation also regenerates the parent monomer. Livingston (ref. 30) determined the cycloaddition and dissociation quantum yields in dilute deoxygenated cyclohexane solutions and derived the following empirical formula :
'
436nm dim
= 2.2 Cnaphthacene]
The
same author also found, for the centrosymmetric photodimer 20, the following photodissociation quantum yield at 254 nm :
Owing
to
the weak solubility of
naphthacene, the
reactivity
(kdim) was only estimated with much approximation and found to be lower than that of anthracene (ref.27). An interesting feature is the important difference between the electronic absorption spectra of naphthacene and its photodimers (Fig.9). Consequently, the naphthacene ring appears to be a good basis f o r photochromic systems but the poor solubility of the parent molecule and its
584
extreme sensitivity to oxidation may limit the applications. 'Ogi
Fig. 9 Electronic absorption spectra of naphthacene 19 (-) in heptane and its photodimers 20 ( - - - - ) and 21 ( . - . . ) in diethylether at room temperature (taken, with permission, from ref.98 and from R.A. Friedel and M. Orchin, Ultraviolet Spectra of Aromatic Compounds, J. Wiley and Sons, New York, 1951). The photodimerization of benz[a]anthracene was reported by Schonberg (ref.99) and studied by Sandorfy (ref.97). It was later shown that the four expected isomeric photodimers are indeed formed (ref.100), but they were not isolated. To our knowledge, no attempt was made to investigate the photochromic behaviour of benz [a]anthracene.
. .
3 . 3 Crossed cvcloadditionq
It is possible to generate ( 4 + 4 ) cophotodimers (AB) between A and B if A or B is in the excited state. According to the mechanism developed in 2.1 (Chart 2 1 , the following reactions should occur : 1 ~ +* B lA* +
A
lB* +
B
-
7
-
~(A,B)* exciplex '(A,A)* __L excimer '(B,B)* excimer
-
AB A2 B2
585
A
competition is expected for the formation of the crossed dimer (AB) or the pure photodimers (A2 and B2). It depends on the respective absorption flux of A and B as well as on the rates of exciplex and excimer formation and of their collapse to the stable photoproducts. Moreover efficient singlet-singlet energy transfer can occur between lA* and B (or vice versa). Even under preparative conditions (i.e. polychromatic source of irradiation) crossed dimers are often produced in larger amount than are the pure photodimers; it seems especially to be the case when the difference of electronic affinity between A and B becomes important : for instance, if A and B are different molecules (anthracene and naphthacene) or are substituted by electron donating and withdrawing substituents, respectively, (A = 9-methoxyanthracene; B = 9-cyanoanthracene). Dipolar solvents, known to favour charge transfer interactions, can then induce electron transfer which leads to a pair of ion-radicals not conducive to such crossed dimers (equation d) : 1 ~ +* B
-
~(A,B)* ~t + B; dipolar solvent
(d)
However, in these solvents, reactions (b) or (c) are not prevented. Therefore, in order to prepare crossed dimers, the solvent has to be carefully chosen. In general, the cophotodimers regenerate the monomers
thermally or by irradiation. Having a lower degree of symmetry than the homodimers. the crossed dimers are generally found to be more soluble in organic solvents.
The first crossed dimers (22,241 were synthetic purpose.
mainly
Z = Me,n-Bu, C1, Br NH2, NHCOCH3, F , COOH CHO CN
prepared
for
(ref.101 (ref.lo21 (ref. 52) (ref.lo31
586
Among the monosubstituted cophotodimers, compound 22 (Z = CHO) was shown by Greene (ref.52) to undergo an acid catalyzed thermal cycloreversion; the same may be true for several other derivatives. The cyano derivative (22, Z = CN) was found to be thermally unstable like the "ht" photodimer of 9-cyanoanthracene. Albini (ref.l03), measured the cross dimerization quantum yield at infinite concentration (+"dim = 0.13 in methylcyclohexane and 0.45 in acetonitrile) which points to an interesting solvent
NC
23
(ref.34,711
24
Z = CH3, Z' = CH20CH3 (ref.104); Z = Z'= n-Pr (ref.104); Z = Z'= C1,Br (ref.101); 2 = Z'= CH3, OCH3 (ref.105); Z = Z'= n-C10H21,OC1oH21 (ref.106) effect:in this case, acetonitrile increases the yield in cophotodimer but when an equimolar mixture of 9-methoxyanthracene and 9-cyanoanthracene is irradiated in acetonitrile, the yield, in compound 23 is low ( = 18%), in contrast to the irradiation in ether which leads to the "hh" crossed dimer 23 in high yield ( = 8 7 % ) (refs.34,71); as previously mentioned, the dipolar character of acetonitrile is presumably the driving force of an ion-pair formation between electron donor (9-meth0xyanthracene)and electron acceptor (9-cyanoanthracene) molecules. Such a polar effect was also observed in the formation of 25 (Z' = CN) which, at 3OoC, is the almost exclusive photoproduct in ether (99.5%) but is produced in smaller amount in CH30H (65%); moreover,, in refluxing acetonitrile, only the thermally more stable di-9-cyanoanthracene (the pure photodimer) was isolated (ref.107). Of note is the preparation of cophotodimers 24 (Z = Z' = n-C10H21 and Z = Z ' = OC10H21) designed for having a high solubility in non protic organic media (0.7 to 1.0 M in C6H6; 0.5 to 1.3M in CCl4 (Compare the solubility of 24 (Z=Z'=CH-,): = 10-311 in C6H6 and ~ ~ 1 4 ) HOW.
587
Their ever, their solubility in EtOH is very poor (6.7 x 1 0 - 5 M ) . formation quantum yield in methylcyclohexane (at 400 or 406 nm conc = 10-3M) was found to be 0.11 ( 2 =C10H21) and 0.02 ( 2 = OCloH21) (ref.106).
HI
Z
25 2 = Me, 2 ’ = C1, Br, CN
(ref.lo51
Me Me
26 (ref. 108)
Finally, the irradiation of an equimolar mixture of 9.10dimethylanthracene and 9,lO-dimethoxyanthracene in ether under argon generates the crossed photodimer (26) as the only thermally stable photoproduct (ref.108) which is fully substituted at the meso positions. By heating, 26 regenerates the two monomers which are both sensitive to oxygen. The photochromic properties of these compounds were not further investigated.
and other acenes.
----------------
The preparation of these crossed photoproducts 27,28.29 was carried out essentially for mechanistic reasons. Generally, the crossed products are obtained smoothly especially from 9,lOdimethylanthracene and 9,lO-dimethoxyanthracene (refs.109,110). Benzo [a]anthracene was used in equimolar amounts (conc.= 2 ~ 1 0 - ~ M ) with anthracene or its derivatives in ether (ref.109); naphthacene was irradiated as a suspension in ether with stirring; the best yields were obtained with freeze-and-thaw degassed media (ref.110). Although intramolecular ( 4 + 4 ) crossed cycloaddition between the anthracene and naphthalene rings had been known (ref. 111) for several years, the intermolecular reaction (between 9cyanoanthracene and naphthalene or 2-methylnaphthalene) giving 9
588
was reported only recently (ref.103). A cross cycloadduct between 2-cyanonaphthalene and naphthalene was also obtained (ref. 112).
27 Z = H, Me, OMe (ref.1091
28 Z = H, Me, OMe (ref.110)
22 Z = H, Me (ref.103)
1,3-Dienes are known to be good triplet quenchers for a number triplet sensitizers (ref.2a,p.476). Hammond and coworkers (ref.113) showed that conjugated dienes quench also the flumescence of naphthalene and other aromatic hydrocarbons and suggested that exciplexes were the intermediates in the quenching Process. The chemical consequences of these deactivations were investigated by several research groups who discovered a host of new cycloaddition reactions between aromatic hydrocarbons (benzene, naphthalene, anthracene, dibenzanthracenes) and polyenes (acyclic and cyclic dienes, cycloheptatriene;. In most reactions, a (4ll+4ll) cycloadduct is formed. sometimes as the major photoadduct, but more often, a complex mixture of. products is obtained. The photodissociation, theoretically allowed, and the thermal cycloreversion have been studied in a few cases. The examples relevant to photochromism will be stressed. One of the earliest reports was the photoaddition of benzene to 1,3-butadiene which, in pentane at - 8 0 ° C , yields the stereospecific (4+4) strained cycloadduct with trans intracyclic double bond as the sole product (ref.114) but, at room temperature, a complex mixture results; the same is true for toluene and xylene; mixtures were also obtained with isoprene. Naphthalene (N) (0.01M) and cyclohexadiene (CHD) (1.OM) were shown to give tWO
of
589
adducts (disappearance quantum yield of N at 313 nsn:+2i3 = 0.23) 30 and 31 in the ratio 9:1. Compound 30 generates the cage compound 22. by photosensitization (ref.115). Naphthalene also reacts efficiently (C/32i3 = 0.42) in benzene-toluene with trans, trans2,l-hexadiene or 2,5-dimethyl-2,4-hexadiene (ref.116) to givebas the major product,the (4+4) cycloadduct (at 25OC and - 5OC, respectively) 3 3 , which undergoes a facile Cope rearrangement to 4
-
(respectively at 8OoC and 35OC).
AV
Xan thone
-
30 major
+
31 minor -
Similarly, naphthalene in the presence of dihydrophthalic anhydride, generates a ( 4 + 4 ) cycloadduct followed by Cope rearrangement (ref.1161, whereas octafluoronaphthalene produces (44-4) and (2+2) photocycloadducts with conjugated dienes (ref.117). Furan, known to have a diene character, gives clean photoadditions to 1cyano-(ref.118) and 2-cyanonaphthalenes (ref.119). The reversibility to the starting materials is achieved thermally ( 2 ) or photochemically (36).The same Japanese group isolated the ( 4 + 4 )
9,10-cycloadduct of furan with 9-cyanoanthracene (ref.120). Yang and Libman (ref.121) showed 'that cyclohexadiene adds efficiently to anthracene, a ( 4 + 4 ) cycloadduct being obtained in high chemical yield by irradiation in benzene.
590
&zN
CN
''/
35
1arge excess
mCN 03 +.
1 h3' (254nm)
This was extended to 9,10-dimethylanthracene. 9-cyano- and 9formylanthracene; the disappearance quantum yield (+-A) was found to be high for anthracenf (0.34) and 9-cyano anthracene (0.45) (ref.122). 9,10 Disubstitution by electron acceptor substituents
Z
k3 (Pyrex)
+ other products
benzene
A
z
H CN
CHD (excess)
37 -
(4+4 cycloadduct) major product
4-A
0.34 (in methylcyclohexane) 0.45 .(in benzene)
seems to affect the mode of photoaddition with 9,IO-dicyanoanthracene, cyclohexadiene (CHD) leads to a mixture of (4+4) and (4+2) adducts in benzene but, in CH2C12, diene dimers are dominant (ref.123). Saltiel and coworkers ref.124) isolated four cycloadducts by irradiation of CHD with 9,10-dichloroanthracene in benzene and their results strongly suggest the intermediacy of a non fluorescing exciplex. Under preparative conditions, 40 is not formed from 41 (although the Cope rearrangement readily occurs at 8OoC) and the (2+2) adduct is the major product.
591
*+o CI
CI
CHD benzene, argon R.T.
r
I(.@
38 ~ 3 1 %
\
39 -
<1%
41 -
~12%
\
cl
40 -
[DCA] = 2x10m3M [CHD] = 1.7M
+A
~57%
= 0.27
Compound 40 regenerates the starting materials by irradiation ( $ ! $ = 0.3 in CH3CN); compound 8 photodissociates with a high quantum yield ( = 1.0 in cyclohexane) but gives several products in addition to the starting compounds (ref.124b). The photoaddition of 9-phenylanthracene to CHD was also found complex (ref.55). Yang took advantage of the photoaddition of a specially designed derivative of CHD to prepare,in several steps, the formal ( 4 + 4 ) and ( 4 + 2 ) photoadducts of benzene to anthracene (ref.125); the efficient photochemical and thermal cycloreversion of 42 and 43 to anthracene and benzene has interesting (ref.125b) = 0.86; implications for orbital symmetry control (42 thermolysis : AH#= 33 kcal mol-l; ASs= 16 e.u. : 43 = 0-71; - 3 e.u.. All the data thermolysis : AH*= 24 kcal mol-I; were determined in n-octane).
43i54
Ad=
\
\
-
major in CH3CN
major i n
(X=Ph Y=H)
(X=Y=CH3)
# 42 -
:+A:! :+i::
CH2C12
t
43 -
592
Acyclic dienes add smoothly to anthracenes by irradiation. Thus, with trans,trans-2,4-hexadiene in excess (l.OM), anthracene ( ~ x ~ O - ~inM benzene ) gives a mixture of (4+2) and (4+4) photoadducts (44, 45, 46) and a thermally formed Diels-Alder adduct (47) (ref- 5 5 ) .
(not isol a ted )
46 _ -
A
A2 : dianthracene Stereospecific (4+2) photoadditions of the same diene to 9cyano and 9-formyl anthracenes were observed (ref.126) but Kaupp and coworkers showed that, for 9-cyanoanthracene, a (4+4) adduct originally formed between - 5oc and OOC, undergoes a 1,3 shift to a (4+2) adduct between 00 and 2OOC (ref.127). The overall reactions are efficient, the disappearance quantum yields of anthracene or 9-cyanoanthracene in methylcyclohexane or benzene being in the range 0.27-0.56 in the presence of an excess Of t,t-2,4 hexadiene or 2,5-dimethyl-2,4 hexadiene (ref.122). A careful mechanistic study by Saltiel (ref.128) led to a kinetic scheme being derived and the rate constants of the main processes being evaluated. The photoaddition of cyclopentadiene to anthracenes was studied by Kaupp and Gruter (ref.1291; they found that anthracene gives a (4+2) and a (4+4) cycloadduct but 9-chloro and especially 9-cyanoanthracene gives the ( 4 + 4 ) cycloadduct as
593
the only photoproduct;
the latter, heated at 186OC, produces
cyanoanthracene (80%) and the (4+2) adduct (20%).
9-
The photoaddition of cycloheptatriene to anthracene (A) was first reported by Sasaki and coworkers (ref.130) who, i n t e r
alia,
in
obtained a (4+4) cycloadduct with 25% yield by irradiation ethanol (+-A = 0.04); this adduct was photolyzed into its
starting materials (84% yield in anthracene). The reaction was also investigated later by Yang (ref.122) who reported the formation of ((6+4), ( 4 + 4 ) , and ( 4 + 2 ) photoadducts as well as a substitution
product for 9-cyanoanthracene.
The latter reaction
was reexamined by Raupp and coworkers (ref.127) who discovered a second (4+4) cycloadduct and dicycloheptatrienyl in addition.
With anthracene, these authors identified other products also (ref.127). Yang noticed that the relative yields of (4+4) adducts increase with the polarity of the solvent and also from cene to 9-cyanoanthracene (ref.122).
anthra-
Benzlalanthracene adds to 1,3- pentadienes to generate (4+2) adducts from the triplet manifold (ref.131) but cyclohexadiene
-
-
with benzo [a] dibenzo [a,c] and dibenzo [a,h] anthracenes lead to a mixture of ( 4 + 4 ) , (4+2) and (2+2) photoproducts (ref.132). ~
Analogous results were observed with cyclopentadiene (ref.133). 3.4 Heterocyclic ComDounds
Heterocyclic compounds such as pyridine, quinoline, isoquinoline are not generally reactive towards photoaddition reac-
tions, especially when the lowest excited state is n-R*. But, when the lowest excited state is of lT-ll* nature, a number of these derivatives can undergo (2+2) (e.g. thymine, coumarin ) and ( 4 + 4 ) cycloadditions, when they display a clear double bond or diene character, respectively.
...
Regarding the (4+4) photodimerization, the monocyclic Cornpounds known to give the cycloaddition display a diene-like structure (such as benzoisofuranl or a diene-like structure and an intramolecular charge-transfer character by incorporation Of an electron donating and an electron accepting group in the ring (see formulae below). The photoreactivity of azaanthracenes and acridizinium salts is more like that of the related polycyclic hydrocarbons.
594
tl 0
H
isobenzofuran
2-pyridone
a-pyrone
The photodimerization of 1,3-diphenylisobenzofurane (&&) was described in 1906 (ref.134) and further studied by Courtot in 1965 when
(ref.135);
the initial orange yellow
solution
&& is transformed into its colourless photodimer
can thermally revert to
g.
disappears (u)which
49 -
48 -
a-Pyrones can undergo (4+4) and (2+2) photodimerization but the (4+4) photodimers extrude C 0 2 on heating with formation of a cyclotetraene (refs.136,137).
The photochemistry of 2-pyridones has been extensively are rapidly explored. Taylor (ref.138) found that 50 and converted to solid photodimers and 53 whose structure Was demonstrated
q"
by two independent groups (refs.139.140).
2 solution
Q
A
-
0
55
o solution
50 51
R=H
54
( R=C2H5; nC3H7;a11y l ; i
R=CH3
52
R=H
A series
595
-
of derivatives of 2 (1H) pyridone (4) was also studied (refs.136,142). A more recent careful investigation of the photochemistry of N-methyl-2-pyridone (51)revealed that four iSOmeric ( 4 + 4 ) photodimers are formed (53, 5 6 , =) instead of a single product (53) (ref.143).
=,
2-Aminopyridine was investigated together with 2-pyridone, as well as a great number of derivatives and the structure of the photodimers was established by chemical correlation between the two series (refs.136,141,142); 2-aminopyridine was irradiated in concentrated hydrochloric acid and, so, reacted in its conjugate acid form ( 5 9 ) . The precipitated photodimer 60 (dichloride) was
59 -
60 -
then isolated. It could be dissolved in water and.reacted with phosphate or methanesulfonate salts in order to obtain the diphosphate or the dimethanesulfonate by precipitation (ref.76). The salts were used by Chandross and coworkers as photochromic materials for holographic devices; in these systems, to obtain a high density of active molecules which undergo a large change in light absorption, a crystal of the dimer is generally essential. The photodimers can be cleaved at short wavelengths into a Pair of monomers which are maintained in proper alignment for redimerization (remaking efficiency was found = 0.25 in the single crystals). The authors were able to write gratings but the mechanical failure of the crystals was thought to restrict the aPPlication of these crystals in that field (ref.76). Another related photocycloaddition reaction, occurring in the solid state, was
596
reported more recently (ref.144) for a derivative of
(61).
pyrazinone
A3
solid state
1-Aza (ref.145) and 2-aza (ref.146) anthracenes were shown by Etienne to photodimerize like anthracene. Later, Bendig and Kreysig (ref.46) reported that they have the same dimerization and
dissociation quantum yields as those of anthracene (see TABLE 3 ) . Etienne also photodimerized benz[b]acridine and dibenzlb,hlacridine (ref.147). The photodimerization of acridizinium bromide SJ
hv' (< 320m) ye1 1ow
63 -
64 -
an ionic heterocyclic derivative also related to anthracene, was first described by Bradsher et a1 (ref.148). The X-ray structures of the bromide and a series of other salts (I-, C1-, C104-,BFq-) containing water of crystallization, have been determined by Jones et al (ref.149), to examine in particular, the interesting single crystal (monomer) to single crystal (photodimer) conversion. The photochemical reactivity in both directions was determined by Bendig and Kreysig (ref.46) in alcoholic solution, for the parent molecule and some derivatives (TABLE 6).
597
TABLE 6
a0
Photodimerization at infinite concentration ( ($3 dim) and photodissociation (@ diss) of acridizinium derivatives and azaanthracenes at 298K (a) in MeOH (b) in EtOH (from ref.46).
+-dim
4 diss
0.23
0.74
0.23 0.22 0.16 0.10 0.02
0.76 0.76 0.83 0.75 0.72
(b)
0.33
0.65
(b)
0.34
0.67
(a)
&9
0
7
7-Me 9-Me 11-Me 9-c1 9-Br
Preparing the toluenesulfonate salt of acridizinium, Chandross and coworkers were able to grow single crystals (2mm x 2mm x 0.5mm) of good optical quality (ref.76) which could be used for holographic purposes (the UV spectra of 63 and 64 are given in Fig. lo), by writing and erasing gratings with high efficiency (dimer breaking at 313 nm z 0.3; dimer remaking at 365 nm 0.7); a great number of cycles (several thousands) was possible. Similar results were obtained in hard polymer matrices (ref.150). For a general and authoritative account of refractive-index image recording systems, please see ref.151. Benzacridizinium dimers (ref.152) were also studied, but unsuccessfully, by the same authors (ref.76).
598
log€
I
300
400
*m]
Fig. 10 Absorption spectra of acridizinium dimer monomer (-) after ref.76, with permission. 4
(----
)
and
INTRAMOLECULAR ( 4 + 4 ) PHOTOCYCLOADDITION
The linkage of two chromophores A and B by a non absorbing and flexible chain can be of great value for generating photochromic systems which are concentration independent (Scheme 1). According to its nature, the chain is also able to increase the solubility of these materials in various media. Moreover, as shown later, the dynamic properties of the link which strongly affect the photochemical behaviour of the chromophores have to be considered. Indeed, the intrinsic folding rate of the chain should be in the range of the decay rate of the excited chromophore and the preferred conformations should be compatible with the geometry imposed on the intramolecular encounter complex, and, subsequently, on the intermediate transition state (ref-153).
locally excited bichromophore
intramolecular "encounter complex"
intramolecular cycloadduct
Scheme 1 Simplified representation of the photochemical cycloaddition of a bichromophoric system.
599
4.1 Bichromophores 4.1.1 Bisanthracenes and bisacridiziniums
..................................
The intramolecular photocycloaddition of bisanthracenes has been extensively investigated (refs.12,153,155). For stereochemical considerations and as they are of an easier synthetic access, the 9-9' substituted derivatives were the most investigated; indeed, the lateral substitution of the anthracene ring is not always straightforward and several photoisomers can be formed either for statistical (ref.153) or electronic reasons (ref.81). Irradiation of the yellow coloured 9,9'-linked bisanthracene derivatives ( X > 400 nm) is usually accompanied, as observed for the intermolecular process, by a blue shift of the absorption spectrum of the starting material ( x < 300 nm) (Fig.5 and ll), owing to the saturation of the anthracene central ring in the photoproduct as shown in scheme 2.
65 (1)
65
-0)
Scheme 2 Photocyclomerization ("intramolecular dimerization") of bisanthracenes. For details see TABLES 7.8.9,lO. Particular emphasis was given to using short chains (Z bearing 1 to 4 atoms) in which a minimum of conformational transitions should be necessary for reaching the encounter complex. A consequence of the chain flexibility is the very short fluorescence lifetimes of the bichromophores, (see TABLE 7), compared to a reference monochromophoric derivative : 9-methylanthracene (Tf = 4.6 ns) (ref.162). TABLE 7 clearly shows that the photochemical closure is very sensitive to the natureof the substituents (X,Y) fixed on the 10 and 10' positions. A net decrease of the cyclomerization quantum yield is observed when X or Y is a bulky
600
TABLE 7 Quantum yields of isomerisation (+ 1 , 2 ), CYClOreVerSiOn ($2,1) and fluorescence emission of bisanthracenes bearing short chains ( 1 or 2 atoms) (X-A-Z-A-Y);TF denotes the fluorescence lifetimes and Ea (In A) are the activation energy and thg natural log. o f the preexponential factor, respectively,for thermal decomposition ( A = 9,lO-anthrylidene).
(4~)
2
x,y
H,H H,H H,OCH3
CH2 CHOH CHOH
+
1,2 (366nm) 0.15 0.29 0.05
+2,1 (<300nm) 0.76
0.81
-
O F 0.06
0.02 0.02
T ~ ( n s ) ref.
1.1 0.3 0.3
156 156 156
kcal /mol-l
- - - - - - - - - - - - - - - - _(In_ _A)_ _ _ ------ - - - - - -
CH2-CH2
H,H
0.26
CH2-CH2
H ICH3
0.26
CH2-CH2
CH3,CH3
0.04 0.14
CH2-CH2
H IOCH3
0.24
CH2-CH2
OCH3,OCH3
0.11
CH2-CH2
H,
CH2-CH2 OCH3
c‘
/
=
c
CH3 /
\
4
0.034
9, 9
<<0.015
HtH
0.20
34.2
(14.8) 31.0
(13.5)
-
24.3
(14.4) 32.3
(14.4) 25.9
(14.9) 29.3 (14.1) c
-
0.20
2.5
157
0.15
-
158
0.14 0.16
2.0
156 158
-
158
0.14
158
0.30
158
0.61
158
0.80 1.2,13 (dual) (dual)
158160
-
-
161
601
group (ref.158). With X = Y = +,the quantum yield of closure is minimum and the fluorescence intensity is maximum; this under-
lines the balance between the two processes (ref.159). The highest efficiency is always reached when X = Y = H.
Fig. 11 UV absorption spectra in MCH at room temperature of bis(9-anthryllmethylether (-) and its photoisomer ( - - - - ) . As expected from Hirayama's rule for excimer fluorescence (ref-1701, an efficient photocyclomerization occurs when a saturated three membered chain links the anthracenes (TABLE 8 ) ; by contrast, some links (carbonate, unsaturated groups) do not favour the photocyclomerization, presumably for electronic and conformational reasons. It is observed that the closure efficiency is greatly enhanced (by at least a factor 2) if one or two methylene units are replaced by oxygen ($R : 0.14 for (CH2)3 : 0.32 for CH20CH2 and 0.36 for OCH20). These differences have been
ascribed to the lower energy barrier of rotation about a C-0 bond than about a C-C bond (ref.171). These compounds in which oxygen plays the role of a hinge and in which the photocyclomerization and the reverse process mimic the closure and the opening of a Jaw were called "JAW PHOTOCHROMIC MATERIALS" or jaw compounds (refs.12,154,166).
602
TABLE 8 Quantum yields of isomerization ( 0 1 , 2 ) photodissociation (+2,1) and fluorescence emission ( + F ) of three and four 1
membered linked anthracenes, X-A-Z-A-Y (A = 9,lO-anthrylidene) in non polar solvents, at room temperature. For diasterereoisomeric a,a'-disubstituted ethers (see ref. 163).
Z
* in CH3CN.
bis(9-anthry1)methyl-
~ ~ ~ o (n P Fm
+ m.p.
were taken with a Kofler block.
iF(ns)
ref.
603
(h)
TABLE 9 Fluorescence emission +F and cyclomerization quantum yields of some bisanthracenes A-2-A (A = 9,10 -anthrylidene) bearing polymethylene and polyoxaalkane chains. kDM and kRD are the rate parameters corresponding to excimer and photoproduct.formation, respectively. Solvent (a): methylcyclohexane; (b): benzene. Fluorescence (d): dual, i.e. monomer and excimer type.
Z
DM
+R
(10-4)
(lO6s-I)
kRD (104s-1)
3,500
Thermal
diss. at RT
ref.
-of”), (b) m = 1 11
3
4
I,
5 6
2,200 2,600
0.22 0.12(d)
570
2,150 2,000
0.11(d) 0.18(d)
420 310
2,500
0.11
540
3,400 2,900
3,200
5,800
300 370
155 12,168 1
I
230
.............................................................................
CH 2-O+ (b)
O F -
1,900
(a) 0.04
3
169
604
Similar trends are observed for the four membered chains, namely a net enhancement of the reactivity is recorded when
oxygen is introduced between methylene units. The thermally stable photoproducts decompose to the open form only at fairly temperatures (TABLE 8). Although
some data are missing,
it emerges that the
high
afore-
mentioned photocyclomers were found to revert to the starting materials under short wavelengths (<300 nm) irradiation. The efficiency of the photodissociation @2,1 = 0.76 for Z = CH2, Y = H and
0.81 for Z = CHOH,
Y = H) is higher than that recorded
for
the photochemical closure (+1,2 < 0.4), except for Z = 2-propanone and X=Y=H for which the reaction is controlled by the tri-
plet state (refs.158,160). This makes these systems good candidates for photochromic applications. The thermal stability of the photoproduct is strongly X,Y substituent dependent as shown in TABLES 8 and 10. In some cases, the photocyclomer 652 is not
9,
CN stable enough to be stored (Z = CH20CH2, X,Y = CH3, CN or (TABLE 8 ) : Z = CH2, X,Y = OCH3 (TABLE 10) whereas several years
are necessary to decompose (6S2)Z = CH2, X,Y = H.H. Becker and Jones I1 have determined kinetic parameters (refs.158,164) for the thermal dissociation of some photocyclomers (TABLES 7,lO); it appears that alkoxy substituents (X,Y) destabilize the cyclomer probably for electronic reasons (antibonding interaction between
the oxygen lone pairs). Despite high activation energies (>2Okcal
mo1-l) the preexponential factor A helps reduce the "lifetimes" of the photocyclomer (for instance 65, , Z = CH2, X = Y = OCH3 : Ea = 22.3 kcal mol-l, A = 6.3 x 101ls-l and tM = 25 seconds). It has recently been observed for (652) 2 = CHzOCH2, X=Y=H
Z = OCH20, X = Y = H, that only a limited number of cycles 651?653 can be performed in solution at room temperature owing to the fatigue of the material (perhaps in relation with the breaking of the link (ref.167). This jaw photochromic compound 651 Z = CH20CH2, X = Y =: H ) has been recently used.in polymer solids, taking advantage of the refractive index change in the polymer matrix (which can be directly measured) for optical recording purposes (ref.172). In this connection, the four membered anthraldazines (66) dissolved in organic media or embedded in polymer matrices can be used as an optical filter and optical switch in the spectral region of 350-470nm (ref.173). and
605
CH=N-N=CH
TABLE 10 Kinetic parameters (activation energies Ea and frequency factors A) for the thermal cycloreversion of bisanthracene X-A-Z-A-Y photoisomers (refs.158,174). Solsent (a): o-dichlorobenzene, (b) : cyclohexane, (c):n-hexane. xty
Z = CH2
Solvent
HOCH3
a b a a
H,H
C
H,H OAc ,OAc
OCH3,OCH3
Ea(kcal/mol) 32.8
24.3 22.3
log A
tK(25OC)
ref.
13.6
528 years 6 min.
174
15.1 14.8
29.4
13.9
25.8
10.3
I,
25 sec.
359 days
I,
............................................................................. 2 = CH2CH2
H,H H ,3Me H ,142 Me,Me OMe,OI3e OEt,OEt
a a a
34.2
b
25.9 24.9
b
b
32.3
31,O 24.3
The photocyclomerization of a , W (CH2ln, X = Y = H) is extremely carbons separating the chromophores fluorescence (often dual) is always
14.8
14.4
13.5 14.4
14.9 15.2
385 years 39
"
38 33 min.
142 11
" "
32b
158,174 I,
(1
I,
II
I,
11
I,
I,
174
-bis-(9-anthryl)n-alkaneS
=
sensitive to the number of (refs.12,165). If an intense recorded from these materials
606
whatever the chain length, the closure efficiency rapidly decreases to a minimum .( = for n 2 5; even if k D M , which
represents the excimer formation rate constant,is relatively important ( > 1 0 7 ~ - 1 1 ,the excimer reactivity is weak (kRD s ~ x ~ O ~ Sowing, - ~ ) probably, to their geometry, inadequate to generate the photocyclomer (for rate parameters definition, see section 2 , chart 2 ) . An improvement of the reaction efficiency for long chains was observed with bichromophores linked at the 1,l' and 2,2' positions with CO-O-(CH2),-O-C0 chains (n = 2 ---+9) where& varies from 5 x to 1O-I). However, two photoproducts are formed (ref-175). It was anticipated that binding two anthracenes by polyethylene oxide links, would greatly facilitate photocyclomerization because of the recognized helicity and flexibility of these chains owing to their repeating "C-0-C" units. Indeed compounds 67m (m = 1,3-6) smoothly lead to their corresponding photocyclomers (TABLE 9), but the latter are thermally unstable (except For m = 1) at room temperature (refs.155,176,177). Thermal instability has been ascribed to electrostatic repulsion between oxygen lone.pairs and other intra ring interactions. Enlargement
Pol ar
sol vent
'- -67m 2
Scheme 3 Photochemical formation of crown ether with 4 to 7 oxygens (m = 3 6). Thermal stability is obtained by cation complexation (adapted from ref.177, with permission).
607
of the polyoxacyclanes (m = 3 --+m 2 4 ) increases the thermal stability (TABLE 9 ) . Addition of salts considerably enhances the lifetime of the so-called "PHOTOCROWNS" which become "cationlocked". Heating of the solid cation complexed photocrowns does not give back the open form. Unlocking is performed by using a polar solvent such as CH3CN or CH30H which extracts the cation from its complex. Important gains in stability of the photocrowns are obtained (ref.169) by preparing bisanthracenes @p (p = 2,3) (where a methylene group is directly substituted on anthracene) and especially the dissymmetrical bisanthracene 69 in which a methylene group is facing an oxygen in the photocyclomer (TABLE 9). However, it has been shown for compounds a p that the "normal" photocyclomer -2 is but the major product accompanied by a minor isomer which is thermally stable. The amount of this side photoproduct can be increased by repeating the number n of cycles -1 hv- -2 (see Fig.12). The structure of this compound has been tentatively assigned to the dissymmetrical photoproduct z p from its UV absorption spectrum.
-
-abs.
70 71 -
72
73
300
400
Fig. 12 Evolution of the UV absorption spectrum of a solution (MCH) of a p successively irradiated at 360nm and heated under reflux. (l), (5) and (9) denote the spectra obtained after 1,5 and 9 photo-thermocycles respectively. (Adapted from ref. 169,
with Permission). (MCH = methylcyclohexane).
Dissymmetrical photocyclomers 71-73 incorporating a naphthalene chromophore (Xmax = 340 nm), are the sole products obtained from bis anthracenes containing a silicon chain (refs.37, 65, 155, 178,179). This non classical stereochemistry was proposed to
608
be due to conformational constraints preventing the closure between the 9,9' and 10,lO' positions. The advantage of these compounds rests on the bathochromic shift (Fig.13) presented by the UV spectrum of the photocyclomer compared with those of the other bisanthracenes; this allows the photocyclomer to be cleaved by light of X = 290 nm as found with z?; (f$2,1 = 0 . 8 ) (ref.37).
'""f 5
Fig.13 UV absorption spectra in cyclohexane of the dissymmetrical (9,1'-10,4'photocyclomer) 71 (-1 and the symmetrical photocyclomer of A-CH20CH2-A (652 2 = CH2OCH2, X=Y=H) (----I. (Takan from ref.65, with permission). (4+2) Intramolecular photocycloadditions have also been observed with bisanthracenes 74 . These reactions, apparently originating from the triplet state (refs. 161,180,181). give photoproducts (75) which absorb light up to 390nm (Fig. 14). Cycloreversion is possible as mentioned by Becker (ref.182); for Z = (cis) -CH=CH- the quantum yield of the process in cyclohexane was reported to be 0 . 2 0 (Scheme 4 ) . ref. \ NR 161,180 2: c=c,
>400nm
___)
c-300nm
74 -
-co-
- CO-CH2- SiMe2-
181
Scheme 4 ( 4 + 2 ) Intramolecular photocycloaddition (9,lO : 1',2') of bisanthracenes 74.
609
0
Fig.14 Electron spectral absorption of 74 ( 2 = -?-) before (-) and after (---,I irradiation (4.4 x 10-6M) in degassed benzene irradiation > 410 nm) (adapted from ref.180, with permission).
(X
Although belonging to a different aromatic series, the photoisomerization of a,a-bis-(9-acridiziniumyl)-alkanes x n has been arbitrarily reported in this subsection owing to their similarity with bisanthxacenes.
At room temperature,En leads to anthracene like photocyclomers involving the saturation of the central rings; the spectral shifts are in the range of those recorded with bisanthracenes (ref.183). As listed in TABLE 11, the reverse process is efficient, and so,these materials - which are salts (perchlorates) could find photochromic applications in polar or ionic media (see section 3.4).
610
TABLE 11
Photocyclization &,
and
photodissociation
(&,I)
quantum yields of some a,W-bis-(9-acridiziniumyl)-alkanes methanol at room temperature.
9 1-2 n = 3 n = 4 n = 5
(cyclization. MeOH) ( 1 = 366nm) 0.03 0.08 0.10
in
(p2-1 reverse process ( x = 254nm)
0.65
0.33 0.33
Bis-naphthalenes were first shown to photoisomerize by Chan-
dross and Dempster (ref.185) in 1970 with 1,3-bis(a-naphthyl)propane 12 ( 2 = (CH2)3). The reaction conducted at X > 280 nm gives only the endo cyclomer 78, which can be split to the open form by irradiation at 254 nm or thermally transformed in boiling chloroform by a Cope rearrangement into compound 7 9 : the latter can be photodissociated to by irradiation with light of 5 254nm (Scheme 5 ) .
77
Scheme 5 Intramolecular photocycloadditions of bis a-naphthalenes.
611
When the propane chain is substituted ( 2 = CH2CHOH-CH2), an additional product (80)can be obtained, probably via 2, having a cubane-like geometry (ref.184); but 80 was not found to regenerate 77. No photoproduct was isolated with 1,3-bis ( 8 naphthy1)propane (ref.185).
Ether links (Z = CH20CH2) are also well suited for intramolecular closure of naphthalenes (ref.184). Photoproducts 78 endo, 7 3 exo and p - were isolated; they photodissociate by irradiation at 254 nm. The ultraviclet absorption spectra of these molecules are given in Fig.15 to illustrate the range of wavelengths covered by the bisnaphthalenes. Similar results are observed for bis-a-naphthalenes with a four-member chain ( 2 = CH2CH2-O-CH2) (ref.184b). log&
Fig. 15 UV absorption spectra of 77 (-), exo and 79 (----I in Ethanol at room temperature (taken from ref.l84d, with perrcission) . (.-.a)
4.1.3 Dissymmetrical bichromophores
-_--_--___-_________________
(4+4) Intramolecular cycloaddition has also been reported for dissymrnetrical bichromophores. 9-(l-anthryl)methoxymethylant.hracenr belongs to this category because of the different POsitions of the chain in the two rings (ref.65).
(a)
612
4:O.X
CHzOCH2 81 -
> 340nm
c . 292nm 4:0.66 (MCH ,RT)
The photoproduct 82 contains a naphthalene unit; the UV spectrum of Q is similar to that observed for compound 21 (Fig. 13). An
a
at a interesting feature is the possibility of dissociating wavelength > 280nm. Other dissymmetrical systems involving an anthracene derivative and another chromophore (83)were discovered to undergo a reversible photocycloisomerization (Scheme 6). The photoproducts ( 8 4 ) behave spectroscopically like anthracene photodimers: they absorb light at wavelengths shorter than 3GOnm and can be split back to
83 at 254nm.
EZ3-
-
I \
Scheme
83 -
Y Y
366nm
254nm o r g 84 -
\
Z
Y,Y
ref.
CH2OCH2
H,H
(CH2)3
lllb
H,H
llla
( ~ ~ 2 1 (3 ~ ~ 2 1186 4
6
(4+4) Intramolecular photocycloaddition was similarly observed between anthracene and furan; the photoproduct decomposes near its melting point ( = 19OOC), smoothly regenerating the parent bichromophore
m.
85 -
86 -
613
4.2 Cvclophanes incomoratina two aromatic rinss 4.2.1 Symmetrical cyclophanes
------_-___--_--_--__
Anthracenophanes 12-21 Benzenocyclophanes are known to possess special properties because of a strong interaction between the lT electrons in a direction perpendicular to the planes of the rings; this results, in particular, in an important bathochromic shift of the UV spectra (ref.187). With the anthracene ring, one thus expects the same behaviour and an enhanced photoreactivity. Indeed, bathochromic and hyperchromic shifts were observed for a series of 12.21 anthracenophanes (refs.188,189).
[2.21(9,10) Anthracenophane 87, the first cyclophane in the anthracene area, was prepared by Golden in 1961 (ref.190). It is an infusible orange crystalline compound, sparingly soluble in Organic solvents; the crystals rapidly fade in sunlight (ref.190) giving rise to a colourless solid isomer S (Scheme 7).
storage
d a v h v
Scheme 7 The latter slowly dissociates at room temperature to another crystalline form of 8 7 , termed the phane is dimorphic, the c x . By crystallisation in CHC13, =I3 is other form being denoted a transformed into o c iwhich reverts to J 7 8 on storage. The photoisomerization of 87 was carefully studied by Kaupp (ref.191) who showed that this poorly fluorescent anthracenophane < readily gives 8s in benzene with a high quantum yield 0 0.36) close tomi:+ (photodimerization at infinite concentration) of anthracene (TABLE 1); the photodissociation quantum yield was found to be 0.60.
m;
(4~ (4~
614
300
400
Fig.16 UV absorption spectra of 1,4-dimethylanthracene ( - - - - ) and syn (-1 in THF at room temperature. The curve Of 1,4dimethylanthracene is displaced downward by 0.5 log E Unit (adapted from ref.188 with permission). If 1,4-dimethylanthracene absorbs at wavelengths shorter than 400nm, the spectrum of 12.21 (1,4) anthracenophane CB) extends
to 480nm. Yts photoisomer (90) is readily split at 254nm or on heating to the yellow anthracenophane 89. which can isomerize into the anti fora at temperature > 240OC: the UV spectrum of 90 is blue shifted compared with that of 89 (refs.188.189).
0
R
anti 89
Scheme 8
-
254nm
or 220°c syn 89
90 -
Other anthracenophanes have been reported to photoisomerize; Ferguson studied in detail the behaviour of some (9,lO) anthracenophanes with dissymmetrical alkane chains. Thus [2.31, 12-43 but not 12.51 (9,lO) anthracenophanes are photoisomerizable (ref.192). An explanation was provided for the different photochemical behaviours, based on the calculation of conformational energies, spectroscopic studies at different temperatures and determination of the crystal structures.
615
In order to control the photochromic ability of bisanthracenes (in relation to their cation complexing properties), some (9,101 anthracenophanes incorporating polyethyleneoxide chains were prepared. Thus, the bright yellow polyoxa Ex.21 (9.10) anthracenophanes a n (n = 0 -2) (refs.12,193-19S) are rapidly
discoloured when irradiated with 356nm light, leading to the photoisomers a n which give back Qn at 40-50°C in the dark. No fatigue has been apparently observed with these compounds (refs.193,194).
h3
__t
A
The photoreaction shows no dependence on metal ions (refs.12, 193-1961, but the half life of thermal cycloreversion is prolonged in the presence of salts compared to that of ion-free solutions, especially for n = l; that was rationalized (refs.193, 194) by the formation of metal cation a n complexes. More sophisticated compounds ( s n ) were prepared by Misumi; they include four anthracene rings and two cation complexing links (ref.196). Irradiation was reported to be followed by the fornation of two photoisomers ( z n , E n ) which could be interconverted by light or temperature.
93n
(n = 0 , l )
94n -
95n -
A = 9,lO-anthrylidene Bouas-Laurent and Desvergne (ref.197) showed the
possibility
616
of
cation-assisted luminescence and photochemistry by
designing
the new class of cation complexing, fluorescent and photoreactive
macrocyclic (9,101 anthracenophanes s n . Thus, in the absence of salt, 961 yields photoisomer 97, whereas, in the presence of Na'
96 (Scheme 9). in excess, 98 is generated from the complex 2Na'C Of interest is the difference in the W spectra of the two photoproducts. Owing to their tedious synthesis, poor solubility and often weak stability (e.g. Golden's compound), the anthracenophanes known so far, do not appear to be presently of great interest for photochromic applications (except, p e r h a p s , a n )
-
hv( 366rim) ___z
0
97 -
11
96 -1
2Na'
hv(366nm)
96, 2Na' Scheme 9.
98 -
Adapted from ref. 197b. with permission.
617
Naphthalenophanes Like anthracenophanes, naphthalenophanes display ( 4 + 4 ) intramolecular photocycloaddition leading to dibenzotricyclododecatetraene skeletons absorbing at wavelengths shorter than 300 nm. L2.31 (1,4) naphthalenophane (99) exhibits two photoactive isomers i.e. the and anti forms which can yield the Endo and eXo cycloadducts as shown in Scheme 10 (ref.198).
99 syn -
100 endo -
100 exo -
Scheme 10 However, only anti [2.2](1,4) naphthalenophane 101 is photoreactive, giving, according to experimental conditions, two different photoadducts (Scheme 11) which can restore 101 Upon heating (refs.199,200).
Scheme 11 Similarly, anti [3.3](1,4) naphthalenophane was found to exhibit photochromic behaviour (ref.201). A series of chain substituted syn and anti [3.3] (1,4) naphthalenophanes (102) was shown by Inazu (refs.202,203) to be transformed into the corresponding photoisomers; the latter are reconverted to the original compounds when heated in the solid phase at 175OC. The variation of the UV spectra is shown in Fig.17.
618
I
I
200
300
qnm]
Fig.17
w absorption spectra of syn [3.3](1,4) naphthalenophane and its endo photocyclomer (-) in cyclohexane (taken from ref. 203, with permission).
(----)
Benzenophanes Benzene was shown to photodimerize in layered cyclophanes 103 (Scheme 12) the photoproduct gives back the open form under heating (ref.204).
-
h i g h pressure
mercury lamp THF (-78°C)
A 60°C
103 -
23
X
0 S
Se
a. b. c.
\
104 -
Scheme 12 The driving force for such reactions is due to the face-to-face stacking of two strained benzene nuclei. The range of wavelength covered by these systems is given in Fig.18.
619
Fig.18 UV absorption spectrum of m b ( - - - - ) and u CH2C12 (taken from ref. 204, with permission).
b
in
(-)
gR- Rx' 4.2.2. Dissymmetrical cyclophanes
-----------______-__------
[3.3] (1,4) Naphthaleno (9,lO) anthracenophane (105) is known to photoisomerize in anisole solution (Scheme 13); the reaction is thermally reversible (refs.202,205).
anisole
R
c
\
.//
R' 105
-
"=",
0.3)
___ic
0
-
a. R=R'=C02Et b. R=H R'=CO2nC&Hg
R'
106
Scheme 13
m.
The reversion is 40 times slower for than for Spectral modifications are similar to those observed for (9,lO) anthracenophanes. A particularly interesting case of photochromism is provided by paracyclo (9,lO) anthracenophanes (107) where a benzene ring photocyclizes with anthracene (Scheme 14). These cyclophanes present photoreversible intramolecular cycloaddition, both in solution and solid state (ref.206).
620
&gR (+ = 0.11)
R’
C? 420nk
6
colourless
greenish ye11OW 107 -
108 -
Scheme 14
o-*p-7Jpq
The reproducibility of absorbance changes of E repeated photo-thermo-cycles is shown in Fig. 19.
b
following
abs.
I
0
Fig.19
j
I
0.q
I
10
Absorbance change
,i
,SO
I
-990
t (min)
of m
i
I
330
b at 419nm
480
(in
at 420nm, followed by heating irradiation (-) (taken from ref. 206, with permission).
chloroform) (-----)
by
at 150°C
Compounds m a and m b were found to be stable to prolonged heating compared with 13.31 paracyclo (9,101 anthracenophane derivatives (ref.202).
Irradiation of [lo] (9,lO) anthracenopha-4,6 diyne 109 affords the colourless photodimer in quantitative yield. The reaction was shown to occur via a (4+4) process throughintermediate 111 (refs.207.208) which was trapped at 77K; 111 irradiated at shorter wavelength reverts to 109 quantitatively but it is easily
converted to
109
L
110 at
higher temperature:
111
I
110 -
621
The UV absorption spectra of in Fig. 20.
109, 110
and
111 are
reproduced
Fig. 20 UV absorption spectra of compounds 109 (-), 110 (----I and 111 in THF (taken from ref. 207, with permission). (.-.a),
5
SUMMARY AND CONCLUSION
In this chapter, a selection of molecules showing photochromic properties based on (4+4) photocycloaddition has been examined. Most of the photoreversible systems of this type, known to date, rest on the intermolecular and intramolecular, symmetrical and unsymmetrical, anthracene-like cycloaddition. The photochemistry of anthracenes and related compounds (other acenes, heterOCYCliC aromatic compounds) has been extensively explored over the last twenty years and is now much better understood. Among the large number and variety of molecules undergoing (4+4) photocycloaddition, only a few were studied as photochromic compounds, i.e. the cycloreversion (thermal or photochemical) was not often investigated. Several systems were patented : one based on the intermolecular photodimerization of acridizinium salts (ref.1501, others involving intramolecular cycloaddition in bichromophores (refs.82,154,173)or in cyclophanes (ref.194) : one of them (ref.150) was shown to be of interest for high resolution holography. One of the more recent systems described, Usui's
622
mixed anthraceno-benzenocyclophane (ref.206), exhibits a good reproducibility after several closure-opening cycles in solution and in the solid state and, consequently, a great potential for applications. These systems present attractive features : convenient wavelength range, large difference in absorption spectra between open and cyclized forms, thermal stability at room temperature of the cycloadducts, high quantum yield of photodissociation, slight change in molecular geometry in converting the photodimer into a Nevertheless, they often suffer from some pair of monomers drawbacks which can limit their use : poor solubility in organic media and liability to degradation, especially oxidation. The solubility in aprotic or protic solvents can be increased by chemical modification and the fatigue minimized by incorporating the photochromic units into rigid matrices or molecular assemblies with exclusion of air. Research along these lines is in progress in several laboratories.
...
ACKNOWLEDGEMENTS The authors are indebted to their coworkers cited in the references, and especially to Drs. R. Lapouyade, R. Lesclaux, J-C. Soulignac, Profs. J. Joussot-Dubien and A. Castellan for their contribution in anthracenes and naphthacene cycloadditions, and to the "Centre National de la Recherche Scientifique" and the "MinistBre de 1'Education Nationale" for financial support. REFERENCES 1
2
a/ R. Hoffnann and R.B. Woodward, Acc. Chem. Res., 1 (1968) 17. b/ R.B. Woodward and R. Hoffmann, Angew. Chem. Int. Ed.,Engl. 8 (1969) 781. c/ Nguyen Trong Anh, "Les RBgles de Woodward-Hoffmann", Ediscience, Paris, 1970. d/ D.O. Cowan and W.W. Schmiegel, J. Amer. Chem. Soc., 94 (1972) 6779. a/ N.J. Turro, Modern Molecular Photochemistry, The Benjamin Cummings Publishing Co., Inc., Menlo-Park (1978), Chapter 7. b/ A. Devaquet, Pure Appl. Chem., 41 (1975) 455.
623
3 4 5 6 7 8 9 10 11 12
13 14 15
16
17 18 19 20 21 22 23 24 25 26 27 28
M.J.S. Dewar and R.C. Dougherty, "The PMO Theory of Organic Chemistry", Plenum Press, New-York, (1975) p.367; M.J.S. Dewar, Tetrahedron Suppl.8, Part I, (1966) 75. H.E. Zimmerman, in "Pericyclic Reactions" A.P. Marchand and R.E. Lehr Ed., Acad. Press, Vol.1 (1977) pp.53-106. See also H.E. Zimmerman, J. Amer. Chem. SOC. 88, (1966) 1564 and 1566 D.O. Cowan and R.L. Drisko "Elements of Organic Photochemistry", Plenum Press, New-York, (1976) pp.388-480. W.C. Herndon, Topics Current Chem., 46 (1974) 141. W.L. Dilling, Chem. Rev., 66 (1966) 373; Ibid., 69 (1969) 845; Ibid., 83 (1983) 1. G. Kaupp, Angew. Chem. Int. Ed. Engl., 19 (1980) 243. J.J. McCullough, Chem. Rev., 87 (1987) 811. I. Fritzsche, J. prakt. Chem., 101 (1866) 337; 106 (1869) 274. E.A. Chandross, J. Chem. Phys., 43 (1965) 4175. H. Bouas-Laurent, A. Castellan and J-P. Desvergne, Pure Appl. Chem., 52 (1980) 2633 and references therein. C. Pac and H. Sakurai, Chem. Comm., (1975) 553. C. Pac and H. Sakurai, Tetrahedron Lett. (1969) 3829. a/ E.J. Bowen, Adv. Photochem., 1 (1963) 23. b/ A. Schonberg, "Preparative Organic Photochemistry", Springer Verlag, New-York, 1968, pp. 97-99. c/ D.J. Trecker in "Organic Photochemistry", O.L. Chapman Ed. Dekker, New-York, 2 (1969) pp. 63-116. a/ B and A. Pullman, "Les Theories Electroniques de la Chimie Organique, Masson, Paris, (1952) 586. b/ R.L. Flurry Jr., "Molecular Orbital Theories of Bonding in Organic Molecules", Dekker, New-York, (1968) 84. c/ K. Higasi, H. Baba and A. Rembaum, "Quantum Organic Chemistry", Interscience Publishers, J. Wiley, (1965) 276. C.E. Linebarger, Amer. Chem. J., 14 (1892) 597. W.R. Orndorff and F.K. Cameron, Amer. Chem. J., 17 (1895) 658. J. Hengstenberg and J. Palacios, Annales de la SOC. esp. de Fisica y Quimica, 30 (1932) 5 and Structurber., 11, (19281932) 916. C.A. Coulson, L.E. Orgel, W. Taylor and J. Weiss, J. Chem. SOC., (1955) 2961. M. Ehrenberg, Acta Cryst., 20 (1966) 177. H.H. Jaffe and M. Orchin, "Theory and Applications of Ultraviolet Spectroscopy", J. Wiley, New-York, (1962) pp.294-333. a/ R. Luther and F. Weigert, Z.Phys.Chem., (Leipzig) 51 (1905) 297 and 53 (1905) 25. b/ E.A. Chandross and J. Ferguson, J. Chem. Phys., 45 (1966) 3564 and ref. therein. R. Livingston, "Photochemistry in the Liquid and Solid States", J. Wiley, 1960, pp. 76-82. E.M. Pereira, M.F. Thomaz and C.L. Braga, Rev. Fis. Quim. Engenharia, 3 A (1971) 21. H.D. Becker and K. Anderson, 3. Org. Chem., 48 (1983) 4542. A. Castellan, R. Lapouyade and H. Bouas-Laurent, Bull. SOC. Chim. Fr., (1976) 201. T.M. Vember, T.V. Veselova, I.E. Obyknovennaya, A.S. Cherkasov and V.I. Shirokov, Izvest. Akad. Nauk. SSSR, Ser. Fiz., 37 (1973) 837.
624
29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55
J. Bendig and D. Kreysig, J. Prakt. Chem., 323 (1981) 471. Kei Sin Wei and R; Livingston, Photochem. Photobiol., 6
(1967) 229. J.K. McVey, D.M. Shold and N.C. Yang, J. Chem. Phys., 65 (1976) 3375. a/ J-Ferguson and A.W.H. Mau, Mol. Phys., 27 (1974) 377. b/ B.F. Anderson, J. Ferguson, M. Morita and G.B. Robertson, J. Amer. Chem. SOC., 101 (1979) 1832. M.D. Cohen, 2 . Ludmer and V. Yakhot, Chem. Phys. Lett., 38 (1976) 398. G.E. Berkovic and 2 . Ludmer, J. Amer. Chem. S O C . , 104 (1982) 4280. T. Hayashi, N. Mataga, Y. Sokata, s. Misumi, M. Morita and J. Tanaka, J. Amer. Chem. S O C . , 98 (1976) 5910. N. Boens, M. de Brakeleire, J. Huybrechts and F.C. De Schryver, 2. Phys. Chem. N.F., 101 (1976) 417. J. Ferguson, A. Castellan, J-P. Desvergne and H. BouasLaurent, Chem. Phys. Lett., 78 (1981) 446. R.A. Caldwell, Acc. Chem. Res., 13 (1980) 45. A. Castellan, R. Lapouyade and H. Bouas-Laurent, Bull. SOC Chim. Fr., (1976) 210. H.L.J. BBckstrom and K. Sandros, Acta Chem. Scand., 14 (1960 48. J. Saltiel, Survey Prcg. Chem., 2 (1964) 239. a/ J. Michl, Photochem. Photobiol., 25 (1977) 141 and ref therein. b/ W. Gerhartz, R.D. Poshusta and J. Michl, J. Amer. Chem. SOC., 98 (1976) 6427. J.L. Charlton, R. Dabestani and J. Saltiel, J . Amer. Chem. Soc., 105 (1983) 3473 ; J. Saltiel, S. Ganapathy and B.W. Atwater, J. Amer. Chem. S O C . , 109 (1987) 1209. G. Kaupp, Chimia, 25 (1971) 230 : G. Kaupp and H.W. Griiter, Angew. Chem. Int. Ed. Engl., 18 (1979) 881. L.E. Manring, K.S. Peters, G. Jones I1 and W.R. Bergmark, J. Amer. Chem. SOC., 107 (1985) 1485. J. Bendig, W. Buchwitz, J. Fischer and D . Kreysig, J . Prakt. Chem., 323 (1981) 485. a/ S. Yamamoto and K.H. Grellmann, Chem. Phys. Lett., 85 11982) 73. b/ Ibid., 92 (1982) 533. H.D. Roth and M.L.M. Schilling, IX IUPAC Symposium on Photochemistry, Pau, France, Vol. of Abstracts, (1982) 286. R.A. Barber, P. de Mayo, K. Okada and S.K. Wong, J. Amer. Chem. SOC., 104 (1982) 4995. J.M. Masnovi and J.K. Kochi, J. Amer. Chem: Soc., 107 (1985) 6781. R.A. Barber, P. de Mayo and K. Okada, J. Chem. SOC. Chem. Comm.. , (1982) 1073. F.D. Greene, Bull. Soc. Chim. Fr., (1960) 1356. F.D. Greene, S.L. Misrock and J.R. Wolfe, J. Amer. Chem. SOC., 77 (1955) 3852. F.C. De Schryver, L. Anand, G. Smets and J . Switten, POlYm. Letters, 9 (1971) 777. G. Kaupp and E. Teufel, Chem. Ber., 113 (1980) 3669.
625
56 57 58 59 60 61
62 63 64 65 66
67 68 69 70
71 72 73 74 75 76 77 78
79
J-P. Desvergne, A. Castellan and H. Bouas-Laurent, Tetrahedron Lett., 22 (1981) 3529. T. Wolff, J. Photochem., 16 (1981) 343. T. Wolff, N. Muller and G. von Bunau, J. Photochem., 22 (1983) 61. J-P. Desvergne, F. Fages, I. Frisch, H.C. Becker and H. Bouas-Laurent, unpublished results. H. Gusten, M. Mintas and L. Klasinc, J. Amer. Chem. SOC., 102 (1980) 7936. a/ H. Dreeskamp, B. Jahn and J. Pabst, 2 . Naturforschung 36a (1981) 665. b/ B. Jahn and H. Dreeskamp, Ibid., 37a (1982) 376. B. Jahn and H. Dreeskamp, Ber. Bunsenges. Phys. Chem., 88 (1984) 42. I. Kraljic, M. Mintas, L. Klasinc, F. Ranogajec and H. Gusten, Nouv. J. Chim., 7 (1983) 239. H. Hart and B. Ruge, Tetrahedron Lett., (1977) 3143. A. Castellan, J-P. Desvergne and H. Bouas-Laurent, N o w . J. Chim., 3 (1979) 231. J-P. Desvergne, H. Bouas-Laurent, A. Castellan, J. Kowalski, E. Yurek and A. de Haut de Sigy, J. Chem. Soc: Chem. Comm., (1986) 82. H. Bouas-Laurent and A . Castellan, J.Chem.Soc. Chem. Comm., (1970) 1648. a/ N.C. Yang, H. Shon, T. Wang and J. Masnovi, J. Am. Chem. SOC., 102 (1980) 6652. b/ D.A. Dougherty, C.S. Choi, G. Kaupp, A.B. Buda, J.M. Rudzinski and E. Osawa, J. Chein. SOC. Perkin 11, (1986) 1063. R. Calas and R. Lalande, Bull. SOC. Chim. Fr., (1959) 770. C. Dufraisse and J. Matthieu, Bull. SOC. Chim. Fr., (1947) 307. A. Castellan, R. Lapouyade, H. Bouas-Laurent and J.Y. Lallemand, Tetrahedron Lett., (1975) 2467. T.M. Vember, T.V. Veselova, I.E. Obyknovennaya, A.S. Cherkasov and V.I. Shirokov, Izvest. Akad. Nauk. SSSR, ser. Fiz., 37 (1973) 837. A. Castellan, C.R. Acad. Sci., 281C (1975) 221. H. Bouas-Laurent and C. Leibovici, Bull. S O C . Chim. Fr. (1967) 1847. R.S.L. Shon, D.O. Cowan and W.W. Schmiegel, J. Phys. Chem., 79 (1975) 2087. W.J. Tomlinson, E.A. Chandross, R.L. Fork, C.B. Pryde and A.A. Lamola, Applied Optics, 11 (1972) 543. T. Wolff, N. Muller and G. von Biinau, J. Photochem., 22 (1983) 61. a/ R. Lapouyade, J-P. Desvergne and H. Bouas-Laurent, Bull. SOC. Chim. Fr., (1975) 2137. b/ J-P. Desvergne, H. Bouas-Laurent, R. Lapouyade, J. Gaultier, C. Hauw and F. Dupuy, Mol. Cryst. Liq. Cryst., 19 (1972) 63. a/ J-P. Desvergne, J.M. Thomas, J.O. Williams and H. BouasLaurent, J. Chem. SOC. Perkin 11, (1974) 363. b/ J.M. Thomas, J.O. Williams, J-P. Desvergne, G. Guarini and H. Bouas-Laurent, J. Chem. SOC. Perkin 11, (1975) 84.
626
80 81 82
83 84 85 86 87
J-P. Desvergne, F. Chekpo and H. Bouas-Laurent, J. Chem. SOC. Perkin 11, (1978) 84. F. Fages, J-P. Desvergne, I. Frisch and H. Bouas-Laurent, J . Chem. SOC. Chem. Comm., (1988) 1413. Patent. "Photocycloaddition polymerization of bis(9-anthrylmethyl) diesters of alkylene dicarboxylic acids and alkylene diesters of 9-anthracene carboxylic acid.", De Schryver F.C., Delzenne G.A., (Agfa-Gevaert A.G.) Ger Offen 2,212,427 (C.A.,B (1973) 16755 w-16761j 1972 US 3, 892,642 (1971). G.A. Delzenne, Adv. Photochem., J.N. Pitts Jr., G.S. Hammond, K. Gollnick ed., 11 (1979) 1. N. Miiller, T. Wolff and G. von Bunau, J. Photochem. 24 (1984 37. J . S . Bradshaw and G.S. Hammond, 3. Am. Chem. S o c . , 85 (1963 3953. J . S . Bradshaw, N. Brent Nielsen and D.P. Rees, J. Org. Chem. 33 (1968) 259.
T. Teitei, D. Wells, T.H. Spurling and W.F.H. Sasse, Aust. J.
Chem., 31 (1978) 85. B.K. Selinger and M. Sterns, Chem. Comm., (1969) 978. a/ T. Teitei, D. Wells and W.H.F. Sasse, Tetrahedron Lett., (1974) 367. b/ W.H.F. Sasse in "Solar Power and Fuels", J . R . Bolton Ed., Acad. Press, New-York (1978) 90 P.J. Collin, D.B. Roberts, G. Sugowdz, D. Wells and W.H.F. Sasse, Tetrahedron Lett., (1972) 321. 91 C. Kowala, G. Sregowdz, W.H.F. Sasse and J.A. Wunderlich, Tetrahedron Lett., (1972) 4721. 92 T. Teitei, D. Wells and W.H.F. Sasse, Aust. J . Chem., 29 (1976) 1783. 93 T. Kamijo, M. Irie and H. Hayashi, Bull. Chem. SOC. Jpn., 51 (1978) 3286. 94 P. Wilairat and B.K. Selinger, Aust. J. Chem., 21 (1968) 733. 95 V. Ramesh and V. Ramamurthy, J . Org. Chem., 49 (1984) 536. 96 J.B. Birks, Photochem. Photobiol., 2 (1963) 493. 97 S. Singh and C. Sandorfy, Can. J. Chem., 47 (1969) 257. 98 R. Lapouyade, A. Nourmamode and H. Bouas-Laurent, Tetrahedron, 36 (1980) 2311. 99 A. Schonberg, A. Mustafa, M.Z. Barakat, N. Latif, R. Moubasher and Ak. Mustafa, J. Chem. SOC., (1948) 2126. 100 A. Nour-mamode,R. Lapouyade and H. Bouas-Laurent, unpublished results. R. Livingston and Kei Sin Wei, Photochem. Photobiol., 6 (1967) 229. 101 D.E. Applequist, R. Searle, M.D. Steinhardt, E.C. Friedrich and R.L. Litle, J. Org. Chem., 30 (1965) 2126 and references cited. 102 M.A. Lintner, Ph.D. thesis, University Of Illinois, 1967. 103 A. Albini, E. Fasani and D. Faiardi. J. Org. Chem., 52 (1987) 155. 104 T.M. Vember, Optika i Spectroscopiya, 20 (1966) 347. 105 H. Bouas-Laurent and R. Lapouyade, C.R. Acad. Sci., 264 (1967) 347.
88 89
627
106 F. Fages, J-P. Desvfrgne and H. Bouas-Laurent, Bull. SOC. Chim. Fr., (1985) 959. 107 H. Bouas-Laurent and R. Lapouyade, Chem. Comm., (1969) 817. 108 R. Lapouyade, A. Castellan and H. Bouas-Laurent, C.R. Acad. Sci., 268 (1969) 817. 109 R. Lapouyade, A. Castellan and H. Bouas-Laurent, Tetrahedron Lett., (1969) 3537. 110 H. Bouas-Laurent and A. Castellan, Tetrahedron Lett., (1969) 3537. 111 a/ E.A. Chandross and A.A. Schiebel, J. Amer. Chem. S O C . , 95 (1973) 611. b/ J-P. Desvergne, N. Bitit, A . Castellan and H. BouasLaurent, J. Chem. SOC. Perkin Trans. 11, (1983) 109. 112 A. Albini and L. Giannantonio, J. Org. Chem., 49 (1984) 3863. 113 L.M. Stephenson and G.s. Hammond, Angew. Chem. Int. E d . Engl., 8 (1969) 261. 114 K. Kraft and G. Koltzenburg, Tetrahedron Lett., (1967) 4357 and 4723. 115 N.C. Yang and J. Libman, J. Amer. Chem. S O C . , 94 (1972) 9228. 116 K.T. Mak, J. Srinivasachar, N.C. Yang, J. Chem. Soc. Chem. Comin., (1979) 1038. 117 J. Libman, J. Chem. S O C . Chem. Comm., (1976) 361 and 363. 118 C. Pac, T. Sugioka and H. Sakurai, Chen. Lett., (1972) 39. 119 T. Sugioka, C. Pac and H. Sakurai, Chem. Lett., (1972) 791. 120 K. Mizuno, C. Pac and H. Sakurai, Chem. Lett., (1973) 309. 121 N.C. Yang and J. Libman, J. Amer. Chem. SOC., 94 (1972) 1405. 122 N.C. Yang, R.L. Yates, J. Masnovi, D.M. Shold and W. Chiang, Pure Appl. Chem., 51 (1979) 173. 123 C.R. Jones, B.J. Alleman, A. Mooring and B. Spahic, J. Amer. Chem. SOC., 105 (1983) 652. 124 a) W.R. Smothers, M.C. Meyer and J. Saltiel, J. Amer. Chem. S O C . , 105 (1983) 545, b) W.K. Smothers and J. Saltiel, J.Amer. Chem. SOC., (1983) 2794. 125 a/ N.C. Yang, M.J. Chen, P. Chen and K.T. Mak, J. Amer. Chem. SOC., 104 (1982) 853. b l N.C. Yang, M.J. Chen and P . Chen, J.Amer. Chem. Soc., 106 (1984) 7310. 126 N.C. Yang, J. Libman, L. Barrett Jr., M.H. Hui and R.L. Loeschen, J. Amer. Chem. SOC., 94 (1972) 1406. 127 G. Kaupp, H.W. Gruter and E. Teufel, Chem. Ber., 116 (1983) 630. 128 J. Saltiel, R. Dabestani, K.S. Schanze, D . Trojan, D . E . Townsend and V.L. Goendken, J. Amer. Chem. SOC., 108 (1986) 2674. 129 G. Kaupp and H.W. Gruter, Angew. chem. Int. Ed. Engl., 18 (1979) 881. 140 T. Sasaki, K. Kanematsu and K. Hayakawa, J.' Amer. Chem. Sac., 95 (1973) 5632. 131 J. Saltiel, D.E. Townsend, L.L. Metts, M. Wrighton, W. Mueller and R.C. Rosanske, J. Chem. SOC. Chem. Comm., (1978) 588. 132 N.C. Yang, J. Masnovi and W. Chiany, J. Amer. Chem. S o c . , 101 (1979) 6465. 133 G. Kaupp and H.W. Gruter, Chem. Ber., 113 (1980) 1626. 134 A. Guyot and J. Catel, Bull. S O C . Chim. Fr., 35 (1906) 1124.
135 P. Courtot and D.H. Sachs, Bull. SOC. Chim. Fr., (1965) 2259. 136 R.O. Ran “Organic Photochemistry”, McGraw Hill, New-York (1966) p.178. 137 P. de Mayo and R.W. Yip, Proc. Chem. SOC., (1964) 84. 138 E.C. Taylor and W.W. Paudler, Tetrahedron Lett., 25 (1960) 1. 139 W.A. Ayer, R. Hayatsu, P. de Mayo, S.T. Reid and J.B. Stothers, Tetrahedron Lett., (1961) 648. 140 G. Slomp, F.A. MacKellar and L.A. Paquette, J. Amer. Chem. SOC., 83 (1961) 4472. 141 L.A. Paquette and G. Slomp, J. Amer. Chem. SOC., 85 (1963) 765. 142 E. Taylor and R.O. Kan, J. Amer. Chem. SOC., 85 (1963) 776. 143 Y. Nakamura, T. Kato and Y. Morita, J. Chem. SOC. Chem. Comm., (1978) 620. 144 T. Nishio, N. Nakajima and Y. Omote, Tetrahedron Lett., 21 (1980) 2529. 145 A. Etienne, C.R. Acad. Sci., 218 (1944) 841. 146 A. Etienne, Ann. Chim., 1 (1946) 5.
147 A. Etienne and A. Stachelin, Bull. S O C . Chim. Fr., (1954) 748. 148 C.R. Bradsher, L.E. Beavers and J.H. Jones, J. Org. Chem., 22 (1957) 1740. 149 Wang-Nang Wang and W. Jones, Tetrahedron, 43 (1987) 1273. 150 Optical Storage Devices (reversible photodimerization); E.A. Chandross, R.L. Fork, A.A. Lamola and W.J. Tomlinson 111, U.S. Patent 3,668,663, June 6, 1972. 151 W.J. Tomlinson and E. A. Chandross, Adv. in Photochem., J.N. Pitts, G.S. Hammond and G. Gollnick Eds (Wiley Interscience, New-York), 12 (1979) pp. 201-281. 152 C.K. Bradsher and T.W.G. Solomons, J. Amer. Chem. Soc., 82 (1960) 1808. 153 F.C. De Schryver, N. Boens and J. Put, Xdv. Photochem., 10 (1977) 359. 154 a/ Nouveaux bis-(anthryl-g)alcanes et leur prbparation. M. Daney, R. Lapouyade, G. Fblix and H. Bouas-Laurent, F.P. 2,314,165 App1.75/18,558, 13 June 1975 (C.A. 87 (1977) 690,134837h). b/ Nouveaux dbrivbs bis-anthryles et bis-phbnanthryles, R. Lapouyade, G. FBlix and H. Bouas-Laurent, F.P. 2,314,168 App1.75/i8,559, 13 June 1975 (C.A.87 (1977) 586, 11 72951). 155 J-P. Desvergne, N. Bitit, A. Castellan, M. Webb and H. BouasLaurent, J. Chem. SOC. Perkin 11, (1988) 1885 and e f erences therein. 156 W.R. Bergmark and G. Jones, Nouv. J. Chim., 1 (1977 271. 157 R. Livingston and Kei Sin Wei, J. Amer. Chem. Soc., 89 (1967) 3098. 158 H.D. Becker, T. Elebring and K. Sandros, J. Org. Chem., 41 (1982) 1064. 159 A. Castellan, J-P. Desvergne, R. Lesclaux and J-C. Soulignac, Chem. Phys. Lett., 106 (1984) 117. 160 H.D. Becker, Pure Appl. Chem., 54 (1982) 1589. 161 H.D. Becker and K . Andersson, J. Org. Chem., 52 (1987) 5205. 162 J.B. Birks, ‘Photophysics of Aromatic Molecules’, Wiley Interscience, London, 1970 p. 123.
629
163 H.D. Becker and K. Andersson, J. Org. Chem., 47 (1982) 354. 164 W.R. Bergmark, G. Jones 11, T.E. Reinhardt and A.H, Halpern, J. Amer. Chem. SOC., 100 (1978) 6665. 165 A. Castellan, J-P. Desvergne and H. Bouas-Laurent, Chem. Phys. Lett., 76 (1980) 390. 166 A. Castellan, J-M. Lacoste and H. Bouas-Laurent, J. Chem. SOC. Perkin 11, (1979) 411. 167 J-P. Desvergne and H. Bouas-Laurent, unpublished results. 168 J-P. Desvergne, A. Castellan and R. Lesclaux, Chem. PhYS. Lett., 71 (1980) 228. 169 J-P. Desvergne, N. Bitit and H. Bouas-Laurent, J. Chem. Res., (1984) ( s ) 214; ( M ) 1901. 170 F. Hirayama, J. Chem. Phys., 42 (1965) 3163. 171 J. Dale, Tetrahedron, 30 (1974) 1683. 172 Y. Yoshiszawa, K. Ashikaga, M. Yamamoto and Q. Tran-Cong, Polymer, 30 (1989) 534. 173 R. Paetzold, M. Reichenbaecher, Ger. (East) 108, 154 (C1.G 02b), 05 Sept. 1974, Appl. 175 267, 10 Dec. 1973; 2pp CA 82 (1975) 1628901~. 174 T. Elebring, Ph. D. Thesis, Goteborg (1985). 175 F.C. De Schryver, M. De Brakeleire, S . Toppet and M. Van Schoor, Tetrahedron Lett., (1973) 1253. 176 J-P. Desvergne and H. Bouas-Laurent, J. Chem. SOC. Chem. Comm., (1978) 403. 177 J-P. Desvergne and H. Bouas-Laurent, Israel J. Chem.. 18 (1979) 220. 178 G. Felix, R. Lapouyade, H. Bouas-Laurent and B. Clin. Tetrahedron Lett., (1976) 2277. 179 J-P. Desvergne, N. Bitit, J-P. Pillot and H. Bouas-Laurent, J. Chem. Res. ( s ) , (1989) 146. 180 H.D. Becker, L. Hansen, K. Andersson, J. Org. Chem., 51 (1986) 2956 and references therein. 181 M. Daney, c. Vanucci, J-P. Desvergne, A. Castellan and H. Bouas-Laurent, Tetrahedron Lett., 26 (1985) 1505. 182 H.D. Becker, K. Sandros and K. Andersson, Angew. Chem. SUPPI., (1983) 609. 183 J. Wagner, J. Bending and D. Kreysig, J. F. Prakt. Chemie, 326 (1984) 747 and 757. 184 a/ R. Todesco, J. Gelan, H. Martens, J. Put and F.C. De Schryver, Bull. SOC. Chim. Belg., 89 (1980) 521. b/ R.V. Todesco and J. Put, J. Photochem., 34 (1986) 305. c/ R.S. Davidson and T.D. Whelan, J. Chem. SOC. Chem. COmm., 11 (1977) 361. d/ R. Todesco, J. Gelan, H. Martens, J. Put, N. Boens and F.C. De Schryver, Tetrahedron Lett., (1978) 2815. e/ R. Todesco, J. Gelan, H. Martens, J. Put and F.C. De Schryver, J. Amer. Chem. S O C . , 103 (1981) 7304. 185 E.A. Chandross and C.J. Dempster, J. Amer. Chem. SOC.. 92 (1970) 703, 704 and 3587. 186 J. Ferguson, M. Puza and R.J. Robbins, J. Amer. Chem. SOC., 107 (1985) 1869. 187 E. Heilbronner and Zhong-Zhi Yang in "Cyclophanes II", Topics in Current Chemistry, Springer Verlag, Berlin, 115 (1983) 1.
630 188 A. 189 190 191 192 193 194 195 196 197
Iwama, T. Toyoda, M. Yoshida, T. Otsubo, Y. Sakata and S. Misumi, Bull. Chem. S O C . Jpn., 51 (1978) 2988. M. Morita, T. Kishi, M. Tanaka, J. Tanaka, J. Ferguson, Y. Sakata, S . Misumi, T. Hayashi and N. Mataga, Bull. Chem. S O C . Jpn., 51 (1978) 3449. J.H. Golden, J. Chem. S O C . , (1961 3741. G. Kaupp, Ann., (1973) 844. a/ A. Dunand, J. Ferguson, M. Puza and G.B. Robertson, J. Amer. Chem. SOC., 102 (1980) 3524 b/ A. Dunand, J. Ferguson, M. Puza and G.B. Robertson, Chem. Phys., 53 (1980) 225. I. Yamashita, M. Fuju, T. Kaneda and S. Misumi, Tetrahedron Lett., 21 (1980) 541. 1,4,7,10,10-pentaoxa [13.2](9,10) anthracenophane, Jpn. Kokai Tokyo Koho JP 81, 120, 682 Sept. 81.(C.A.96 (1982) 162760g). I. Tanigawa, Ph.D. Thesis, Osaka University (1984). S. Misumi, 43rd Annual Meeting of Japan Chemical Society, Tokyo, April 1981. a/ A. Castellan, M. Daney, J-P. Desvergne, M-H. Riffaud and H. Bouas-Laurent, Tetrahedron Lett., 24 (1983) 5215. b/ H. Bouas-Laurent, A. Castellan, M. Daney. J-P. Desvergne, G. Guinand, P. Marsau and M-H. Riffaud, J. Amer. Chem. S O C . ,
108 (1986) 315. 198 N.E. Blank and M.W. Haenel, Chem. Ber., 114 (1981) 1531. 199 H.H. Wasserman and P.M. Keehn, J. Amer. Chem. SOC., 89 (1967) 2770. 200 G.Kaupp and I. Zimmerman, Bngew. Chem., 88 (1976) 482; Angew. Chem. Int. Edit. Engl., 15 (1976) 441. 201 M. Yoshinaga, T. Otsubo, Y. Sakata and S. Misumi, Bull. Chem. SOC. Jpn., 52 (1979) 3759. 202 T. Shinmyozu, T. Inazu and T. Yoshino, Chem. Lett., (1978) 405. 203 T. Rawabata, T. Shinmyozu, T. Inazu and T. Yoshino, Chem. Letters, (1979) 315. 204 H. Higuchi, E. Kobayashi, Y. Sakata and S. Misumi, Tetrahedron, 42 (1986) 1731 and references therein. 205 S. Tazuke and H. Watanabe, Tetrahedron Lett., 23 (1982) 197. 206 M. Usui, T. Nishiwaki, K. Anda and M. Hida, Chem. Lett., (1984) 1561. 207 S. Misunii, Mem. Inst. Sci. Ind. Res., Osaka Univ., 36 (1979) 37. 208 S.T. Inoue, T. Kaneda and S. Misumi, Tetrahedron Lett., (1974) 2969. See Additional Literature (1989 - 2001): Anthracenes, A99
631
Chapter 7 5
Cycloaddition Reactions Involving 4n+2 Electrons. Photochromism Based on the Reversible Reaction of Singlet Oxygen with Aromatic Compounds H.-D. Brauer and R. Schmidt
1 INTRODUCTION Photooxygenationo f aromatic compoundsi s often accompaniedby c o lour changes. Because o f the r e s u l t i n gease of d e t e c t i o nsuch r e a c t i o n sfoundvery e a r l y p a r t i c u l a r i n t e r e s t . Perhaps the o l d e s t d e s c r i p t i o n of a photooxygenation was given by F r i t z s c h e i n 1867, who observed t h a t exposure of an orange s o l u t i o no f tetracene t o a i r and l i g h t produced a c o l o u r l e s s m a t e r i a l which regenerated tetracene upon heating ( r e f . 1). Thus already i n t h i s e a r l y study the thermal r e v e r s i b i l i t yof the photoreactionwas noted.
.
About 60 years l a t e r Moureu e t a1 i n v e s t i g a t e dthe photochemicaldecolourizat i o n o f rubrene s o l u t i o n s i n the presence o f a i r . They were the f i r s t who r e a l i z e d
t h a t oxygen (02) was bound chemically i n the photoproduct( r e f . 2).
This work was the s t a r t i n gp o i n t for numeroussystematic studies of the school of Dufraisse dealing w i t h the photooxygenationo f aromatic compounds. Dufraisse recognizedt h a t the photooxygenationproductsa r e transannula r endoperoxides( r e f
3). The r e a c t i v e intermediate i n photooxygenationwas f i r s t supposedby Kautsky t o being formed by energy transfer from e l e c t r o n i c a l l ye x c i t e d be s i n g l e t oxygen (bg) aromatic molecules o r s e n s i t i z e r moleculesand r e a c t i n gw i t h the aromatic molecule i n i t s ground s t a t e ( r e f . 4). Dufraisse r e a l i z e d t h a t cycloreversionl e a d i n g t o the regeneration of 02 and
aromatic compoundi s a general r e a c t i o npath i n thermal chemistryo f endoperoxides. Yields o f 02 i n thermolysiso f s o l i d endoperoxideshad been found t o depend strong-
l y on the nature o f endoperoxide ( r e f . 5 ) .
Obviously i r r e v e r s i b l e s i d e r e a c t i o n s compete w i t h thermal cycloreversion. P a r t i c u l a r l y the group o f Rigaudy i n v e s t i g a t e d i n d e t a i l the chemistry o f r e arrangement r e a c t i o n s o f endoperoxides ( r e f . 6).
Soon i t was observed t h a t most endoperoxides are a l s o photochemicallyr a t h e r unstable compounds, forming rearranged products. Rigaudy e t a1 found o u t t h a t i n the primary step o f photochemical rearrangement the same b i r a d i c a l i s formed as
.
632
in the thermal rearrangement reaction (re f. 7). The f i r s t detection of 1ight induced cycloreversion of an endoperoxide was already reported i n 1942 by Dufraisse and Mellier (ref. 8 ) . They noticed t h a t colourless crystals of the endoperoxide of heterocoerdianthrone obtain rapidly the violet colour of heterocoerdianthrone i n sunlight. However, this early observation on the photochemical splitting of an endoperoxide i n t o aromatic compound and 02 fell into oblivion. In 1978 the photocycloreversion reaction was rediscovered by Rigaudy e t al. for the endoperoxide of 9,lO-diphenylanthracene (re f. 9).Ashorttimelater i n 1979, independentlyviallet e t a1 ,Srinivasan e t a1 and Brauer e t a1 reported the photochemical cycloreversion of different endoperoxides (refs. 10 - 12). Thus rearrangement and cycloreversion cmpete in themchemistry as well as i n photochemistry of endoperoxides. Since photooxygenation of aromatic compounds i s accompanied by a distinct absorption s h i ft and since cycl oreversion of the corresponding endoperoxides can be achieved photochemically and/or thermally, such systems are i n principle photochrmic ( r e f. 13). In the following sections we sumnarize the essential findings on the photooxygenation of aromatic compounds and on the behaviour of endoperoxides w i t h respect to photochromism. From these data we shall derive a concept for the molecular structure of endoperoxides w i t h low yields of rearrangement. W e shall present some highly reversible photochromic systems and briefly discuss some applications.
.
.
.
2 RESULTS OF THEORETICAL INVESTIGATIONS ON THE FORMATION REACTION AND THE
CHEMICAL BEHAVIOUR OF ENDOPEROXIDES The formation reaction and the thermochemical and photochemical properties of the resulting endoperoxides were subjects of a theoretical investigation of Kearns and Kahn ( r e f . 14). Cyclopentadiene was chosen as model compound f o r the 4 + 2cycloaddition reaction with 02. A straightforward extension of the molecular orbital approach which Hoffman and Woodward had used so successfully to tr eat other ground and excited state electrocyclic reactions, however, could not be used f o r the 02 addition reaction, because of the degeneracy i n the highest f illed orbitals of 02. The degeneracy of the 3r,, the two l A g and the 11; states of 02 i s only removed when electron - electron interaction i s explicitely included. I n order to describe the course of the chemical reaction the authors constructed a state correlation diagram which was further refined by the additional consideration of an orbital correlation diagram. In their calculation Kearns and Kahn included thermochemical data t o locate the ground states of the reactants and products, spectroscopic data to locate the excited states and symnetry and spin selection rules t o determine the correlation of reactants and product states. The resulting state correlation diagram describing forward and backward reactions between 02cyclopentadiene adduct, endoperoxide and biradical i s given i n Fig. 1.
633
d
80 h
0 7 L
Q
6 6 a-- m - m 0-0
o=o
40
c Q
c m
-
-
v)
0 -
-
-40 Fig. 1. State c o r r e l a t i o n diagram d e p i c t i n g forward and backward r e a c t i o n s * IA rebetween 02-cyclopentadiene adduct, endoperoxide and b i r a d i c a l 1 ~ and present two degenerate e x c i t e d s t a t e s of the 02 adduct composed o f ground s t a t e cyclopentadiene and t h e two l A g s t a t e s of 02, 11 and 3 1 denote the adduct s t a t e s correspondingt o t h e 3x9 and 116 s t a t e s o f 02
.
From the c o r r e l a t i o ndiagram the f o l l o w i n g generalized conclusionscan be drawn:
1. Only s i n g l e t oxygen ( b g ) r e a c t s t o form an endoperoxide. 02 i n the 31s ground s t a t e and i n t h e e x c i t e d 1x6 s t a t e i s n o t r e a c t i v e i n photooxygenation.
2. Homolysiso f the peroxide bridge and s p l i t t i n go f f o f 02 can compete i n thermochemistry o f endoperoxides. It i s n o t possible t o p r e d i c t which r e a c t i o n w i l l
dominate. 02 may be released i n thermal cycloreversioni n the e x c i t e d bgs t a t e o r i n i t s 31s ground s t a t e depending on the behaviour a t t h e crossing p o i n t A.
3. E x c i t a t i o n o f t h e lowest s i n g l e t and t r i p l e t n;,
u&,
states, being l o c a l ex-
c i t e d s t a t e s o f the peroxide chromophore, leads o n l y t o the cleavage of the peroxide bridge
.
4. E x c i t a t i o no f upper e x c i t e d s i n g l e t s t a t e s o f the endoperoxidecan inducecycloreversion. Since one upper s i n g l e t s t a t e c o r r e l a t e s smoothly and exotherma l l y w i t h t h e lYr3(11) s t a t e o f t h e 02-cyclopentadiene adduct, photocycloreversionwould be expected t o produce the aromatic compoundi n i t s ground s t a t e and s i n g l e t oxygen (11;).
-
Whereas conclusions 1 3 describe already known experimental f a c t s one has t o p o i n t o u t t h a t conclusion4 was a t t h a t time a prediction, p o s t u l a t i n ga very strange type o f r e a c t i o n i n s o l u t i o n chemistry, i.e. an a d i a b a t i c photoreaction o r i g i n a t i n g from an upper e x c i t e d s i n g l e t state. We wish t o mention already a t t h i s stage t h a t t h i s p r e d i c t i o nwas v e r i f i e d experimentallyb y us about 10 years l a t e r ( r e f . 15). Thus t h e theory o f Kearns and Khan describes c o n s i s t e n t l yi n terms o f concerted r e a c t i o n sthe formation and the thermochemical and photochemical behaviour o f endoperoxides.
634
p 3
Q
CH3 A PO
DMAPO
DPAPO
Q T PAP0
T PO
DPT PO
BTPO
TBPPO
BDXPO
t H3 HOCDPO
QQ RUBPO
ADCPO Fig. 2
635
MDHPO
PPO
HECDPO Fig. 3
3 STRUCTURAL FORMULAS The s t r u c t u r a l formulas and abbreviationso f endoperoxides(ARPOS) of aromatic compounds (ARs) which are discussed i n t h i s review are l i s t e d i n Fig. 2 and 3. The meaning o f the abbreviations i s given below. APO: endoperoxide o f anthracene DMAPO: endoperoxide o f 9,lO-dimethylanthracene DPAPO: endoperoxide o f 9,lO-diphenyl anthracene TPAPO: endoperoxide of lY4,9,10-tetraphenyl anthracene TPO: endoperoxide o f tetracene DPTPO: endoperoxide o f 5,12-diphenyl tetracene RUBPO: endoperoxide of 5,6,11,12-tetraphenyl
tetracene (rubrene)
BTPO: endoperoxide o f benzoterrylene TBPPO: endoperoxide of tetrabenzopentacene ADCPO: endoperoxide o f anthradichromene BDXPO: endoperoxide of benzodixanthene HOCDPO: endoperoxide o f dimethylhomoocoerdianthrone HECDPO: endoperoxide o f heterocoerdianthrone PPO: endoperoxide o f pentacene
MDHPO: endoperoxide o f mesodiphenylhelianthrene
Abbreviations w i t h o u t t h e ending PO correspond t o t h e respective aromatic compounds( ARs)
4 4.1
.
EXPERIMENTAL RESULTS ON THE CHEMISTRY OF ENDOPEROXIDES Formation by photooxygenation
Photooxygenationi s a bimolecular process i n which b, s i n g l e t oxygen ( l 0 2 )
attacks AR.
The f o l l o w i n g s i m p l i f i e d r e a c t i o n scheme i s used t o discuss t h e
636
essential features of this reaction. AR*/S*
+ 02
+ AR
I02 + AR
-.
QA
AR/S + 102
kD
02
kQ
02 + AR
( 31
kR
ARPO
(4)
-
In sel f-sensitized photooxygenation 102 is produced by energy transfer from electronically excited AR* i n i t s S1- and/or TI-state (ref. 161, in sensitized photooxygenation by energy transfer frm an excited sensitizer S* to molecular 02. The overall quantum yield of 102 formation i s Q,, which depends on the individual compound AR or S and on 02 concentration. 102 can be deactivated physically by solvent molecules, by quenching by AR and by reaction w i t h AR w i t h rate constants k D , kp and k R , respectively. From the above scheme the quantum yield of photooxygenation may be expressed as
(6)
with
Thus i t follows t h a t Qp depends on CAR]. The value of the reactivity parameter p indicates the concentration of AR a t which Qp reaches half the maximum possible value. For [ A R I >> p Qp becomes independent of CAR].
In Table 1 the sum of the rate constants k R + kp and p-values for CgH6 and CS2 as solvents are l ist e d for several aromatic compounds as examples. k R + kQ depends very strongly on the nature of AR, however, i t i s almost solvent independent ( r e f . 19). Since p i s inversely proportional t o the lifetime of 102, which depends very strongly on solvent (re f. 20) distinct differences between p-values i n different solvents are observed. From the data of Table 1 i t can be seen that for most aromatic compounds experimental conditions can be chosen such that Qp is nearly CAR] independent under usual conditions ( [AR] > 10-4 M) Thus values 0.1 5 Qp 2 0.5 can
.
637
e a s i l y be met f o r most compoundsunder consideration. TABLE 1. Sum o f r a t e constants kR + kQ f o r several aromatic hydrocarbons. Data from r e f s . 17 and 18. p-values c a l c u l a t e d w i t h 102 l i f e t i m e s o f 30-10-6 s i n CgH6 and
34.10-3
s i n CS2. For abbreviations see s e c t i o n3. ~
~
~~
(kR + kp)
Compound
CM-l s-11
P (C&) [MI
p (CS2) [MI
A
0.015
2.2.10'1
2 .om10-4
DM4 DPA T
2.1 0.12 1.2
1.6*10'3 2.8 * 10-2 2.8 * 10-3
1.4.10'6 2.5.10-5 2.5 *
DPT RUB
0.37 4.2
9.0.10'3
7.9-
7.9.10-4
7.0.10-7
7.9.10-6
7.0.10-9
P
420
4.2 Photochemical c y c loreversion I n t h e e a r l y experimental r e p o r t s on t h e photocycloreversiono f endoperoxides t h i s r e a c t i o n was t r e a t e d o n l y q u a l i t a t i v e l y ( r e f s . 8-11). We were the f i r s t who studied photocycloreversion q u a n t i t a t i v e l y by means o f wavelength dependent measurements o f t h e quantum y i e l d o f cycloreversion Qc, b y i n v e s t i g a t i n g the e l e c t r o n i c product states, by studying the i n f l u e n c e o f heavy atoms on
Qc,
by
c a r r y i n g o u t time resolved measurements and v a r y i n g the ARPO molecular Structure (refs. 15,21-30). I n Fig. 4 the e s s e n t i a l r e s u l t s are s u m a r i z e d b r i e f l y . ARPOs are a t l e a s t bichromophoriccompoundsi n which the peroxide chromophorei s separated e l e c t r o n i c a l l y and i n space from the remaining aromatic chromophoresby both sp3 carbon atoms next t o the peroxide b r i d g e (compare Figure 2 ) . The l o c a l e x c i t e d .*,O
.*OO
s i n g l e t s t a t e o f t h e peroxide chromophorei s g e n e r a l l ythe S1-state of ARPOs (refs. 25,301. The upper e x c i t e d s i n g l e t s t a t e s sn ( n > 2) are u s u a l l y nn* s t a t e s o f t h e aromatic chromophores.
-
Cycloreversionoccurs from t h e lowest n** s i n g l e t s t a t e b u t can a l s o proceed frm Sn ( n > 3) m* s t a t e s ( r e f s . 24,26,30). The cycloreversionproducts a r e AR ( S O ) and formed q u a n t i t a t i v e l yi n the r a t i o 1 : 1 ( r e f s . 15,21,23,27,29,31). Cycloreversioni s thus a c t u a l l y an a d i a b a t i c photoreactiono r i g i n a t i n gfrom upper e x c i t e d s i n g l e t s t a t e s as p r e d i c t e d b y Kearns and Khan ( r e f . 14). It I S a r a p i d reaction. The r i s e time of ARs was determined by us and Eisenthal e t a l . t o range
638
Fig. 4. Singlet states and cycloreversion channels of a typical ARPO
between 10 and 75 ps (refs. 28,31). Although cycloreversion occurs from rather short lived excited states ( r e f . 28) QC-values up t o 0.35 have been determined ( r e f . 32). 4.3
Photochemical rearrangement
As was shown by Rigaudy e t a l . several thermal, partially parallel, partially consecutive reactions occur i n rearrangement of ARPOs (refs. 6,331. The photo-
chemical primary process i s the homolytic splitting of the peroxide bridge leading
DPAPO (Sol
R =
@; BR
DEO = bfradical
, DEO
= diepoxide
Fig. 5. Reaction p a t h s i n photochemical rearrangement.
639
v i a a b i r a d i c a l t o t h e formation o f a diepoxide as main intermediate product. I n Fig. 5 the main products o f the complex r e a c t i o n are given f o r t h e example of DPAPO. Photochemical rearrangement o f ARPOs occurs from t h e l o c a l e x c i t e d S 1 o r T1 x & cz0 s t a t e ( r e f s . 25,321. Quantumy i e l d s of rearrangement QR have o n l y been determined by us. They have been found t o vary, dependingon ARPO, between 0.08 and
1 ( r e f . 32). 4.4 Thermal cycloreversion Upon heating ARPOs regenerate AR and 02. Yields o f regeneration l i s t e d i n the 1i t e r a t u r e vary over a wide range ( r e f . 5 ) . However, the 1it e r a t u r e data have been determined under d i f f e r e n t experimental conditions. To g e t comparable data
we measured y i e l d s o f thermal c y c l o r e v e r s i o nAC i n m-xylene a t 139% f o r a series of ARPOs and found values 0.005 < AC < 0.94 depending on ARPO ( r e f . 32). Thermal cycloreversion demands a r a t h e r l a r g e a c t i v a t i o n energy o f u s u a l l y about 30 k c a l h o l ( r e f s . 13,15,34,35). From t h e a c t i v a t i o nparameters h a l f l i f e times w i t h respect t o cycloreversioncan be c a l c u l a t e d ranging f o r some ARPOs up t o hundredso f years. Thus although being i n p r i n c i p l et h e r m a l l yl a b i l e some ARPOs a r e thermally r a t h e r s t a b l e compounds. Already Wasserman e t a l . demonstrated t h a t i n cycloreversiono f DPAPO 102 i s released ( r e f . 36). For several ARPOs o f anthracened e r i v a t i v e sTurro e t a1 reported y i e l d s between 0.35 and 0.95 ( r e f . 34). Consequently thermal cycloreversion i s a chemiluminescentd i a b a t i c reaction. Indeed d u r i n g thermolysis o f ARPOs the phosphorescenceemission a t 1270 nm was r e c e n t l y observed by Chou and F r e i and by Wilson e t a l . ( r e f s . 37,381.
.
4.5 Thermal rearrangement Analyzing the productd i s t r i b u t i o no f thermolyzed and photolyzedARPO samples Rigaudy e t a l . recognized t h a t i n thermal and i n photochemical rearrangement of ARPOs very s i m i l a r products i n s i m i l a r amounts are formed ( r e f . 6). Obviously t h e i r r e v e r s i b l e thermal r e a c t i o n s of ARPOs proceed v i a the same b i r a d i c a l precursor
BR and diepoxide intermediate DEO as i n photochemicalrearrangement (compare Fig.
5 ) . Thus t h e primary step i n thermal rearrangementi s l i k e w i s et h e homolyticrupture Of t h e peroxide bridge.
Since t h i s i r r e v e r s i b l er e a c t i o n competes w i t h thermal cycloreversionone can express the o v e r a l l y i e l d o f rearrangementas & = 1 Ac. Therefore a l s o AR depends s t r o n g l y on the nature o f ARPO.
-
4.6
Comparison of quantum yields and thermal yields of cycloreversion and rearrangement Table 2 l i s t s for a series of ARpOs of the anthracene and tetracene series quantumyields and thermal yields of cycloreversion and rearrangement. An inspection of the d a t a reveal s t h e foll owing :
1. Qc(A) values of ARPOs of the anthracene series are larger t h a n those of the tetracene series. 2. Values of Q c ( A ) do not correlate w i t h AC values. 3. A rather good and approximately linear correlation exists between AR and Q R ( S ~ ) , the quantum yield of rearrangement upon excitation of S1. As has been shown by us i n a detailed analysis, the linear relation between Q R ( S ~and ) AR results from the f a c t t h a t the deactivation of the Sl(& c t 0 ) state occurs for ARPOs exclusively chemically on the repulsive potential curve of S1 or T i via the homolytic rupture of the 0-0 bond ( r e f . 32). The precursor BR can then form rearranged products via DEO or react back to form ARPO(So), as is the case i n the thermally induced reaction (see Fig. 5). Thus the same s t e r i c factors controlling thermal rearrangement become important i n photochemical rearrangement.
TABLE 2. Thermal yields Ac, AR and quantumyields Qc(A) and Q R ( S ~of) cycloreversion and rearrangement of ARPOs. Qc(A) determined a t wavelength A , QR(SI) calculated for excitation of S1. Data from ref. 32.
ARPO
APO DM4PO DPAPO TPAPO TPO DPTPO RUBPO
Qc(A) a
[ml
A
0.22 2 0.01 0.35 2 0.035 0.22 5 0.01 0.12 + 0.01 0.055 + 0.003 0.073 + 0.004
270 270 270 248 248 248 248 ~~
0.088
0.005
sc 0.01 0.54 0.94 0.94 0.005 0.40 0.75
QR(~I) 0.96 0.34 0.075 0.13 0.95 0.26 0.20
-+ O . l a -+ 0.05a -+ 0.015d -+ O.OZe -+ O.la 2 0.03’1
-+ 0.02e
ARC 0.990 2 0.003 0.46 + 0.05 0.06 It: 0.015 0.06 + 0.015 2 0.995 0.60 2 0.03 0.25 + 0.03
~~~~
a i n CH2C12 b same errors i n AC l i k e i n AR c in m-xylene a t 139OC
d i n toluene e i n m-xylene a t 200C
Q R ( S ~as ) well as AR decrease w i t h increasing volume of substituents i n 9,lO position of the anthracene frame and i n 5,12 position of the tetracene frame. Obviously the formation of DEO w i t h i t s two strongly strained three-membered rings i s s t e r i c a l l y hindered by large s u b s t i t u e n t s .
641
5
PHOTOCHROMIC SYSTEMS OF HIGH REVERSIBILITY BASE0 ON THE REVERSIBLE PHOTOOXYGENATION OF AROMATIC COMPOUNDS
Since enough 02 i s present i n an a i r c o n t a i n i n genvironment t o support photooxygenation and since t h e e s s e n t i a l p r o p e r t i e s such as r e v e r s i b i l i t y o f photor e a c t i o nand change i n absorptiona r e given, systemsAR + 02 photochromic:
+=ARPO are i n p r i n c i p l e
I n t h e following, an equation w i l l derived, p e r m i t t i n gan estimate o f the number n o f photoreversiblecycles by means o f t h e quantumyieldsQc and QR. It i s assumed t h a t s i d e r e a c t i o n s l e a d i n g t o an i r r e v e r s i b l e consumption o f the photochromic m a t e r i a l o n l y occur d u r i n gp h o t o l y s i so f ARPO b u t n o t d u r i n gphotooxygenation of AR. Thus one photochromicc y c l e i s represented by t h e scheme given below.
a~(A2)
The photoreversiblec y c l e s t a r t s w i t h i r r a d i a t i o n
o f a pure ARPO s o l u t i o n
of concentration CARPOI,.
I r r a d i a t i o n ends when a f r a c t i o n C of ARPO has been converted t o AR. Then t h e respective concentrationso f ARPO and AR amount t o : [ARPOI
=
(1
-
C ) CARP010
(8)
Afterwards AR i s completely converted t o ARPO by s e n s i t i z e d o r s e l f - s e n s i t i z e d photooxygenationa t wavelength h l . A f t e r completion o f the f i r s t cycle, t h e concentrationo f ARPO i s :
After
m cycles, the f r a c t i o n o f remaining ARPO amounts t o :
A f t e r n cycles [ARPO],
has decreased due t o i r r e v e r s i b l e side r e a c t i o n t o l / e
o f [ARPOlo ( e = Eulers number). Therefore equation (13) holds
With Q ~ ( h 2 )<< Qc(A2) and C
C<
1 one obtains the approximativeequation (14) f o r
the numbern o f photoreversiblec y c l e s u n t i l the photochromicm a t e r i a l i s decomposed t o l / e ( r e f . 13).
From eq. (14) i t f o l l o w s t h a t the r a t i o Q c ( A ~ ) / Q R determines (~~) the r e v e r s i b i r i t y of a photochromicsystem. From the compoundsl i s t e d i n Table 2 DPAPD w i t h Qc(270) = 0.22 and Q ~ ( 2 7 0 )= 0.06
( n o t i d e n t i c a l w i t h Q R ( S ~ ) )has the l a r g e s t r a t i o
Qc(270)/Q~(270) 3.5. Assuminga conversionof DPAPO o f C = 0.05 per photochromic c y c l e a r e v e r s i b i l i t ynumber o f n z 70 already r e s u l t s f o r DPAPO. Q
5.1
Concept f o r t h e molecular s t r u c t u r e of endoperoxidesw i t h small y i e l d s o f rearrangement I n order t o improve the p h o t o r e v e r s i b i l i t yone has t o increase Qc(A z ) / Q R ( ~ ~ )
accordingt o eq. (14) , which can be achievedby increasingQ&) and/or decreasing Q ~ ( x 2 ) .QC values are l a r g e s t f o r ARPOs of the anthracene series, b u t vary r a t h e r a r b i t r a r i l y w i t h ARPO structure. As has been shown i n Section 4.6 QR and AR, however, depend d e f i n i t e l yon the ARPOmolecular s t r u c t u r e . A1 thoughQR values are known o n l y f o r a r a t h e r small number of ARPOs, we can now extend our knowledge about the s t r u c t u r a l dependence o f QR b y making use o f the p r o p o r t i o n a l i t ybetween QR
and AR, since AR values have been determined f o r many ARPOs o f d i f f e r e n t structures. An i n s p e c t i o n o f the l i t e r a t u r e data ( r e f . 5 ) f o r ARPOs w i t h l o w y i e l d s AR demonstratest h a t a c t u a l l y ARPOs derived from 9,lO-diary1 s u b s t i t u t e danthracenes and HECDPO have the lowest tendency towards rearrangement. P a r t i c u l a r l y HECDPO, which can be regarded as a d e r i v a t i v e of WAPO i n which both phenyl s u b s t i t u e n t sare
connected t o the anthracene frame by two k e t o groups (compareFig. 2 and 31, appears t o be a promisingcandidate. The s t e r i c hindrance f o r the formation o f a diepoxide should be even stronger i n t h e case of HECDPO than f o r DPAPO, because of the r i g i d i t y o f i t s molecular s t r u c t u r e r e s u l t i n g from the CO-bridges. By thermolyzing HECDPO i n b o i l i n g m-xylene we measured AC = 0.995,
distinctly
643
l a r g e r than the l i t e r a t u r e value o f 0.95. I t f o l l o w s t h a t f o r HECDPO, AR '= 0.005, about one order o f magnitude smaller than the respective value f o r DPAPO.
1,5 connection
1,4 connection
Fig. 6. S t r u c t u r a l formulas o f bridged DPAPOs
We t h e r e f o r e concluded t h a t t h e f i x a t i o n o f the molecular s t r u c t u r e of DPAPO by -X-
bridges connecting t h e phenyl s u b s t i t u e n t si n 1,4 o r i n 1,5 p o s i t i o n w i t h
the anthracene frame (see Fig. 6 ) should l e a d g e n e r a l l y t o a reduction o f AR and
QR and thus t o b e t t e r p h o t o r e v e r s i b i l i t y( r e f s . 12,391. S u i t a b l e bridges -X- c o u l d be -CO-, -0-, -S- o r a r y l i c groups. 5.2
Spectral , photochemical and thermal data o f bridged endoperoxides Followingour concept we prepared and characterized i n a d d i t i o n t o HECDPO f o r
t h e f i r s t time the ARPOS, HOCDPO, BDXPO, ADCPO, TBPPO and BTPO Containing-co-, -0o r a r y l i c bridges -X- (see Fig. 2 and 3 f o r s t r u c t u r a l formulas). As i s shown i n Table 3 a c t u a l l y a reductionof AR and QR compared t o the values o f DPAPO resulted. However, a secondn o t l e s s importantimprovementwas achieved. Due t o the p l a n a r i t y o f molecular s t r u c t u r eo f the correspondingaromatic compoundsAR forced b y the bridges X, t h e i n t e r a c t i o n of the n - e l e c t r o n i c systems o f phenyl s u b s t i t u e n t s and anthracene increases strongly. This leads t o an enormous r e d s h i f t i n absorptiono f HECD, HOCD, BDX, ADC, TBP and BT comparedt o DPA. Consequentl y llcolour changes" which f o r DPAPO occur o n l y i n the uv region become now t r u l y v i s i b l e , i.e. colourless/red, c o l o u r l e s s / v i o l e t and colourless/blue. I n Figures 7 ARs are given.
- 12 the absorptionspectra o f the new ARPOs and the corresponding
Thermochemical and photochemicaldata r e l e v a n t f o r the photochromicp r o p e r t i e s of ARPOs are l i s t e d i n Table 3. Obviously the reduction i n QR and AR compared t o the respective values o f DPAPO i s about the same f o r 1,4 and 1 , 5 connected structures. However, i t depends on the QR decreases i n the -X- sequence a r y l i c bridge > ether bridge > nature of -X-. keto bridge. Since the Qc values o f Table 3 d i f f e r from t h e respective value f o r
Speclra 01
- AECD and
-0-0-
HECDPO in CH,CI,
4Qo
200
BM)
Navelength [nm] Fig. 7
Speclra of
Mo
Fig. 8
-HOCD and -*-.- HOCDPO in CH,CI,
400
600
Wavelength [nm]
645
Spectra o i -BDX and
-0-0-
BDXPO in CH,CI,
50
0 200
400
600
800
Wavelength [nm]
Fig. 9
Spectra of - ADC and
50
0
Fig. 10
200
-0-0-
ADCPO in CH2C12
600
400
Wavelength [nm]
800
646
* Spectra 01
50
200
-TBP and
4GQ
-0-0-
TBPPO in toluene
600
Wavelength [nm]
F i g . 11
Spectra of -BT and
+o-
BTPO in CH,CI,
SO
400
600
Wavelength [nm]
TABLE 3: Thermal y i e l d s A and photochemical quantum y i e l d s Q of c y c l o r e v e r s i o n (index C) and rearrangement
( i n d e x R) and p h o t o r e v e r s i b i l i t y number n o f b r i d g e d ARPOs. A c t i v a t i o n parameters and e x t r a p o l a t e d
half lifetimes
1 /2
T~~
a t 20° f o r thermal c y c l o r e v e r s i o n .
A R ~
A Cs-lle
70
0.97
0.03
1 . 5 ~ 1 0 ~ ~
25.5
0.02f
110
0.95
0.05
1.0~10~~
34.5
0.13d
0 *Old
260
0.99
0.01
4.0~1013
29.3
3.9
0. Nd
0.01d
360
0.99
0.01
3.2~1013
29.3
5.0
0.985
0.015
3.2~1014
32.9
240
0.995
0.005
1.6x1015
34.6
880
A [nml
5TPO
313
0.1oc
0.03c
TBPPO
313
0.145~
ADCPO
313
5DXPO
313
HOCDPO
313
HECDPO
302
QC(X)
0.29~~
na
0 . 0 0 4 ~ ~ 1300
Ea Ckcal/mle
a c a l c u l a t e d according t o eq. (14) w i t h conversion C = 0.05 p e r photochromic c y c l e i n b o i l i n g m-xylene i n dichloromethane d i n toluene e i n m-xylene
20 T1/2[~I
Acb
ARPO
1.6 120
648
DPAPO the improvement i n the p h o t o r e v e r s i b i l i t y i s n o t the same as i n QR. The number o f cycles ranges from moderate n = 70 f o r BTPO t o e x c e l l e n t n = 3200 for HOCDPO. AR and QR f i t again i n t o a roughly l i n e a r c o r r e l a t i o n and confirm the conclusions o f s e c t i o n 4.6 concerning the r e l a t i o n between photochemical and thermal rearrangement. From the a c t i v a t i o n parameters determined f o r the thermal cycloreversion, ha1f 20
l i f e t i m e s ~ 1 1 2with respect t o cycloreversion can be extrapolated, ranging from 1.6 years f o r BTPO t o about 900 years f o r HECDPO. Thus w i t h exception o f BTPO the new ARPOs are thermally very s t a b l e compounds. 5.3
Photochromic p r o p e r t i e s of t h e new systems AR + 02
-
ARPO
The values o f the photooxygenation r e a c t i v i t y parameter p o f the bridged ARs are about the same o r smaller than the one f o r RUB. Consequently already a t CAR]
z 8 ~ 1 0 -M~i n ~ 6 o ~r CAR] 6 z 7.10-7 M i n cs2, h a l f the maximum t h e o r e t i c a l possible quantum y i e l d o f photooxygenation i s achieved. Depending on the nature o f AR and on the solvent, quantum y i e l d s 0.05 < Qp < 0.25 can therefore e a s i l y be obtained i n air-saturated organic solvents i n self-sensitized experiments (compare a1 so s e c t i o n 4.1).
The o n l y exception i s HOCD which was found by us t o be unable t o s e n s i t i z e I02
since QA < 0.00006 ( r e f . 40). Although being very r e a c t i v e towards (P(H0CD) = 9.10-5 M i n toluene) HOCD can o n l y be photooxygenated by means o f a s e n s i t i z e r . It f o l l o w s t h a t the system HOCD + 02 e HOCDPO for i t s e l f i s n o t photochromic since reactions proceed o n l y from HOCDPD towards HOCD b u t n o t i n the reverse d i r e c t i o n . This p a r t i c u l a r behaviour o f HOCD enables the development o f a novel type o f photochromism besides the two conventional types. I n conventional photochromism a t l e a s t one r e a c t i o n i s induced by l i g h t . The reverse r e a c t i o n can be performed photochemically and/or thermally. I f the reverse colour r e a c t f on 1eadi ng t o t h e product absorb4 ng a t shorter wave1engths can be performed by long-wave i r r a d i a t i o n t h i s photoreversible type o f photochromic system has the advantage that, f o r example, a p l a s t i c f i l m information w r i t t e n by short-wave i r r a d i a t i o n can be cancelled o r corrected by long-wave i r r a d i a t i o n . The main disadvantage o f such a photoreversible photochromic system, however, i s the fading o f information during reading, as the absorbed l i g h t always i n i t i a t e s the photochromic reaction. Most photochromic compounds, among them a l l ARPOs o f Table 3 w i t h exception o f HOCOPO and the m a j o r i t y o f the spiropyranes ( r e f . 411, be1ong t o t h i s group.
If the reverse c o l o u r r e a c t i o n can o n l y be performed thermally t h i s type o f photochromic system i s advantageous i n having one 1i g h t - i n s e n s i t i v e component. Information stored i n f i l m s may t h e r e f o r e be read w i t h o u t fading i f no i r r e v e r s i b l e photochemical degradation occurs. The major disadvantages o f these systems are, however, t h a t a p a r t i a l c o r r e c t i o n o f information i s n e a r l y impossible by heating
-
649
u s u a l l y a l l the i n f o r m a t i o n i s cancelled o u t by t h i s method
- and t h a t f o r most o f
these systems thermal f a d i n g occurs already a t room temperature. spiropyranes belong t o t h i s photochromic group.
Some o f the
Based on the p h o t o s t a b i l i t y o f HOCD a novel type o f photochromic system combining t h e advantages o f t h e d i f f e r e n t conventional photochromic systems can be designed. This system can be switched between p h o t o s t a b i l i t y and photoreversible photochromism by using various i r r a d i a t i o n wavelengths A ( r e f . 40).
Upon a d d i t i o n o f a s u i t a b l e 102 s e n s i t i z e r (S), f o r example 9,10-dichloroanthracene (DCA), t o a b l u e s o l u t i o n o f HOCD we o b t a i n a r e a l l y photochromic system. As shown i n Fig. 13 t h e absorbance i n the l o n g wave region (spectrum 1) vanishes upon e x c i t a t i o n o f the s e n s i t i z e r w i t h A = 405 nm i r r a d i a t i o n and the s o l u t i o n becomes c o l o u r l e s s (spectrum 2). Afterwards, when t h e s o l u t i o n i s i r r a d i a t e d a t 313 nm where the photochemically produced HOCDPO absorbs, t h e b l u e c o l o u r comes back again (spectrum 3). By t h i s procedure a photoreversible photochromic c y c l e i s performed which can be repeated about 3000 times ( r e f . 40). I f the wavelength o f i r r a d i a t i o n i s switched t o A > 450 nm, however, our system becomes photostable. Thus i n f o r m a t i o n can be read w i t h o u t f a d i n g w i t h l i g h t o f 450 nm < A L
-
Fig. 13. Absorption spectra o f t h e photochromic system DCA DCA + HOCDPO I
To our knowledge the system S + HOCD + 02
5 750
nm!
+ HOCD + 02 ===
S + HOCDPO i s the f i r s t o f the
k i n d t h a t can be switched between p h o t o s t a b i l i ty and photoreversible photochromism by choice o f t h e i r r a d i a t i o n wavelength. Since r e v e r s i b i l i t y as w e l l as thermal s t a b i l i t y o f HOCDPO a r e excellent, as can be seen from Table 3, and since the system
650
works a1 so i n polymer matrices as do a l l systems o f Table 3 , i t s f u r t h e r i n v e s t i g a t i o n w i t h respect t o possible a p p l i c a t i o n s seems t o be very a t t r a c t i v e . F i n a l l y the f o r an a p p l i c a t i o n most important p r o p e r t i e s o f t h e novel
photochromic endoperoxides are compared i n Table 4 w i t h the respective p r o p e r t i e s o f t h e spiropyranes, which are s t i l l the b e s t i n v e s t i g a t e d c l a s s o f photochromic
compounds.
With respect t o the wavelength s h i f t A X
, the
quantum y i e l d s o f c o l o u r i z a t i o n
and d e c o l o u r i t a t i o n and the r e v e r s i b i l i t y , the photochromic ARPOs are j u s t as good as the spiropyranes. The factthatthecolouredcomponentsARoftheARPOsystemsfluoresce,represents an important advantage compared t o spiropyranes, since a1 ready small conversions o f ARPOs o f l e s s than 1 % can be detected q u a n t i t a t i v e l y by e x c i t i n g AR f l u o r e s cence. Using t h i s method the r e v e r s i b i l i t y can be increased very e a s i l y over the numbers given i n Table 4. TABLE 4. Comparison o f re1evant p r o p e r t i e s o f photochromic compounds measured i n solution
~~~~~~~
~
spiropyranes ( ref. 41) A A ( col oured/col our1ess , m)
200
f l uorescence o f col oured compound
no
Q (colourization) Q (decol ourization)
reversibil i t y thermal s t a b i l i t y a t 20% switching between
- 300
230 - 280
0.01 - 1, 0.1 - 0.5a 0 0.1, 0.01 0.05’ 5 - 17000,100 l@
-
-
endoperoxides
-
-
5s 5d not possible
photostabil ity/photoreversibil ity
Yes
- 0.6 0.05 - 0.25 70 - 3209 1.6 y - 880 y 0.1
possible with
S+tlOCDco2
* SMOCDPO
a t y p i c a l values
bcalculated according t o eq. (14) w i t h C = 0.05;
smaller conversions C l e a d t o
r e c i p r o c a l l y l a r g e r r e v e r s i b i l i t y numbers
A d e c i s i v e advantage i s the b e t t e r s t a b i l i t y o f the thermally l a b i l e component
-
of the ARPO systems, exceeding those o f t h e spiropyrane systems by f a c t o r s o f lo2 5.104. Therefore the hand1 i n g and storage o f photochromic f i l m s c o n t a i n i n g endoperoxides can occur a t room temperature whereas spiropyrane f i l m s have t o be
stored i n r e f r i g e r a t o r s i n order t o prevent l o s s o f information v i a thermal bleaching
.
A f u r t h e r important preference o f our photochromic system i s the p o s s i b i l i t y
of switching between p h o t o s t a b i l ity and photoreversible photochromism.
Reading
651
w i t h o u t simultaneous l o s s o f i n f o r m a t i o n i s accomplished v i a t h i s p o s s i b i l i t y . To our knowledge the system S + HOCD this possibility.
+ 02 e=s +
HOCDPO i s the f i r s t which o f f e r s
5.4 A p p l i c a t i o n s Up t o now t h e o n l y a p p l i c a t i o n o f photochromic systems AR + 02 s=ARPO occurs
i n the f i e l d o f actinometry. The system HECDPO =+HECD
+ 02 was developed by us as
a reusable chemical actinometer f o r the uv between 248 and 334 nm ( r e f s . 42,431. The development o f the r e d c o l o u r d u r i n g cycloreversion i s u t i l i z e d as actinometric reaction. A f t e r use d e c o l o u r i z a t i o n i s achieved behind a y e l l o w f i l t e r ( c u t o f f 455 nm). Between 253 and 302 nm actinometric factors do n o t depend on i r r a d i a t i o n wavelength and thus i t may be used as a polychromatic quantum counter. The accuracy, the s e n s i t i v i t y , the ease o f use and the h i g h degree o f r e u s a b i l i t y are the outstanding p r o p e r t i e s o f t h e actinometer Photon Technology I n t e r n a t i o n a l Inc.
R (248/334).
which i s commercially a v a i l a b l e from
(PTI)
under t h e designation Actinochrome
Three d i f f e r e n t , n o t reusable actinometers using o n l y t h e formation r e a c t i o n AR have been developed by us for
of ARPO by s e l f - s e n s i t i z e d photooxygenation o f
the wavelength ranges 334 nm < A < 395 nm (AR = DMA) 405 nm < A< 500 nm (AR = DPT)
- -
and 475 < X < 610 nm (AR = MDHl ( r e f s . 43
-
45).
The wavelength independence o f the actinometric f a c t o r s and quantum y i e l d s ,
comnon t o the photooxygenation actinometers, r e s u l t s from the f a c t t h a t Qd i s independent o f i r r a d i a t i o n
wavelength f o r the three compounds.
Therefore they
a l l can be used as polychromatic quantum counters. Accuracy, s e n s i t i v i t y and ease o f use must again be pointed out. The system developed e a r l i e r ( r e f s . 43,451, based on the photooxygenation o f MDH, i s commercially a v a i l a b l e from PTI under the designation Actinochrome N (475/610). The f o u r actinometers mentioned above complement one another and they j o i n t l y cover the broad wavelength range from 248 t o 610 nm. E f f o r t s are c u r r e n t l y being made t o extend t h i s range f u r t h e r i n t o the red. 6
CONCLUSIONS Endoperoxides were known u n t i l 1979 as thermally and photochemically r a t h e r
unstable compounds, w i t h a complex chemistry dominated by rearrangement reactions. By designing new endoperoxides undergoing 1i t t l e rearrangement,
the cyclo-
reversion r e a c t i o n became the major pathway which makes endoperoxides more a t t r a c t i v e as photochromic systems. Few o f them have been synthetized so f a r b u t they were i n v e s t i g a t e d i n d e t a i l . Among them, h i g h l y r e v e r s i b l e systems o f outstanding p r o p e r t i e s were found. I n the near f u t u r e our i n t e r e s t w i l l focus on a broader v a r i a t i o n o f c o l o u r range and a f u r t h e r
improvement o f r e v e r s i b i l i t y .
Both goals may be achieved by
652
introducing electron donating substituents and by testing new b r i d g i n g groups -X(-X- = -S- i s currently under investigation). With respect t o possible applications, the sometimes moderate solubility will be improved by suitable variation of molecular structures. Especially, the unique system permitting a switch between photostability and photoreversibility will be refined by optimizing the sensitizer and improving the solubility. Thus the topic i s currently an active research area with a great potential for applications and further progress i s expected.
653
7 REFERENCES
1. 2. 3. 4. 5.
.,
Fritzsche, C.R. Acad. Sci 64 (1867) 1035. Moureu, C. Dufraisse and P . r D e a n , C.R. Acad. Sci., 182 (1926) 1440. Dufraisse, B u l l . SOC. Chim. France, 53 (1933) 837. Kautsky, Trans. Faraday SOC., 35 (19x1 216. G o l l n i c k and G.O. Schenck i n "c4Cycloaddition Reactions", ed. J. Hamer, Academic Press, New York, 1967, chapter 10. 6. J. Rigaudy, M.-C. P e r l a t , D. Simon and N.K. Cuong, B u l l . SOC. Chim. France, M. C. C. H. K.
(1976) 493. 7. J. Rigaudy, N.C.
Cohen and N.K. Cuong, C.R. Acad. Sci., 264 (1967) 1851. 8. C. Dufraisse and M.-T. M e l l i e r , C.R. Acad. Sci., 215 (19X-J 541. 9. J. Rigaudy, C. B r e l i e r e and P. Scribe, T e t r a h e d r o n e t t . , 7 (1978) 687. 10. A. V i a l l e t , J. Rouger, H. Cheradame and A. Gandini, J. Pho€ochem., 2 (1979) 129. 11. R. Srinivasan, K.H. Brown, J.A. Ors, L.S. White and
