Hydrogen Bonding and Transfer in the Excited State
Hydrogen Bonding and Transfer in the Excited State, Volume I & II ©...
31 downloads
1006 Views
76MB Size
Report
This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!
Report copyright / DMCA form
Hydrogen Bonding and Transfer in the Excited State
Hydrogen Bonding and Transfer in the Excited State, Volume I & II © 2011 John Wiley & Sons, Ltd. ISBN: 978-0-470-66677-7
Edited by Ke-Li Han and Guang-Jiu Zhao
Hydrogen Bonding and Transfer in the Excited State Volume I
Editors
Ke-Li Han State Key Laboratory of Molecular Reaction Dynamics, Dalian Institute of Chemical Physics, Chinese Academy of Sciences, Dalian, China
Guang-Jiu Zhao State Key Laboratory of Molecular Reaction Dynamics, Dalian Institute of Chemical Physics, Chinese Academy of Sciences, Dalian, China
Hydrogen Bonding and Transfer in the Excited State Volume II
Editors
Ke-Li Han State Key Laboratory of Molecular Reaction Dynamics, Dalian Institute of Chemical Physics, Chinese Academy of Sciences, Dalian, China
Guang-Jiu Zhao State Key Laboratory of Molecular Reaction Dynamics, Dalian Institute of Chemical Physics, Chinese Academy of Sciences, Dalian, China
This edition first published 2011 Ó 2011 John Wiley & Sons Ltd Registered office John Wiley & Sons, Ltd, The Atrium, Southern Gate, Chichester, West Sussex PO19 8SQ, United Kingdom For details of our global editorial offices, for customer services and for information about how to apply for permission to reuse the copyright material in this book, please see our website at www.wiley.com The right of the author to be identified as the author of this work has been asserted in accordance with the Copyright, Designs and Patents Act 1988. All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, electronic, mechanical, photocopying, recording, or otherwise, except as permitted by the UK Copyright, Designs and Patents Act 1988, without the prior permission of the publisher. Wiley also publishes its books in a variety of electronic formats. Some content that appears in print may not be available in electronic books. Designations used by companies to distinguish their products are often claimed as trademarks. All brand names and product names used in this book are trade names, service marks, trademarks or registered trademarks of their respective owners. The publisher is not associated with any product or vendor mentioned in this book. This publication is designed to provide accurate and authoritative information with regard to the subject matter covered. It is sold on the understanding that the publisher is not engaged in rendering professional services. If professional advice or other expert assistance is required, the services of a competent professional should be sought. The publisher and the authors make no representations or warranties with respect to the accuracy or completeness of the contents of this work and specifically disclaim all warranties, including, without limitation, any implied warranties of fitness for a particular purpose. This work is sold with the understanding that the publisher is not engaged in rendering professional services. The advice and strategies contained herein may not be suitable for every situation. In view of ongoing research, equipment modifications, changes in governmental regulations, and the constant flow of information relating to the use of experimental reagents, equipment, and devices, the reader is urged to review and evaluate the information provided in the package insert or instructions for each chemical, piece of equipment, reagent, or device for, among other things, any changes in the instructions or indication of usage and for added warnings and precautions. The fact that an organization or website is referred to in this work as a citation and/or a potential source of further information does not mean that the author or the publisher endorses the information the organization or website may provide or recommendations it may make. Further, readers should be aware that Internet websites listed in this work may have changed or disappeared between when this work was written and when it is read. No warranty may be created or extended by any promotional statements for this work. Neither the publisher nor the authors shall be liable for any damages arising herefrom. Library of Congress Cataloging-in-Publication Data Hydrogen bonding and transfer in the excited state / editors, Ke-Li Han, Guang-Jiu Zhao. p. cm. Includes bibliographical references and index. ISBN 978-0-470-66677-7 (cloth) 1. Hydrogen bonding. I. Han, Ke-Li. II. Zhao, Guang-Jiu. QP517.H93E93 2010 572’.33–dc22 2010015107 A catalogue record for this book is available from the British Library. ISBN 9780470666777 Set in 10/12pt, Times by Thomson Digital, Noida, India Printed in Great Britain by Antony Rowe Ltd, Chippenham, Wiltshire
Contents
Editors’ Biographies
xv
Reviewer Comments
xvii
List of Contributors
xxxiii
Preface
xxxix
Volume I 1
2
3
Vibrational Dynamics of the Double Hydrogen Bonds in Nucleic Acid Base Pairs Yun-an Yan and Oliver Ku¨hn 1.1 Introduction 1.2 Hydrogen Bonding and Nonlinear Infrared Spectroscopy 1.3 Correlated Vibrational Dynamics of an Adenine–Uracil Derivative in Solution 1.4 Conclusion Acknowledgement Appendix References Vibrational Energy Relaxation Dynamics of XH Stretching Vibrations of Aromatic Molecules in the Electronic Excited State Takayuki Ebata 2.1 Introduction 2.2 IR Spectra of 2-Naphthol and its H-Bonded Clusters in S1 2.3 VER Dynamics of Bare 2-Naphthol 2.4 VER Dynamics of H-Bonded Clusters of 2-Naphthol 2.5 Comparison of the cis ! trans Barrier Height Between S0 and S1 2.6 Conclusion References Hydrogen Bond Basicity in the Excited State: Concept and Applications Attila Demeter 3.1 Introduction 3.2 Experiment
1 1 3 9 22 23 23 24
29 29 30 31 31 36 37 37 39 39 40
vi Contents
3.3 Results and Discussion 3.4 Summary Acknowledgements References 4
5
6
7
Solute–Solvent Hydrogen Bond Formation in the Excited State. Experimental and Theoretical Evidence Iulia Matei, Sorana Ionescu and Mihaela Hillebrand 4.1 Introduction 4.2 The Prerequisite Conditions for Hydrogen Bond Formation 4.3 Diagnosis Criteria and Quantitative Treatment of Hydrogen Bonds 4.4 Design of the Experiments 4.5 Theoretical Modelling of the H-Bonds 4.6 Conclusions References Electronic-Excited-State Structures and Properties of Hydrated DNA Bases and Base Pairs Manoj K. Shukla and Jerzy Leszczynski 5.1 Introduction 5.2 Ground-State Structures of Nucleic Acid Bases and Base Pairs 5.3 Excited-State Structures of Nucleic Acid Bases 5.4 Excited States of Base Pairs 5.5 Excited-State Dynamics and Non-Radiative Decays 5.6 Conclusions Acknowledgements References Insight from Singlet into Triplet Excited-State Hydrogen Bonding Dynamics in Solution Guang-Jiu Zhao and Ke-Li Han 6.1 Introduction 6.2 Theoretical Methods 6.3 Results and Discussion 6.4 Conclusion Acknowledgements References Probing Dynamic Heterogeneity in Nanoconfined Systems: the Femtosecond Excitation Wavelength Dependence and Fluorescence Correlation Spectroscopy Shantanu Dey, Ujjwal Mandal, Aniruddha Adhikari, Subhadip Ghosh and Kankan Bhattacharyya 7.1 Introduction 7.2 Solvation Dynamics in Nanoconfined Systems 7.3 Fluorescence Resonance Energy Transfer (FRET): lex Dependence
41 76 77 77
79 79 80 82 98 104 117 119
125 125 128 129 138 142 143 143 144
149 149 151 151 156 156 156
159
159 160 166
Contents
7.4 Excited-State Proton Transfer (ESPT) 7.5 Diffusion of Organic Dyes by Fluorescence Correlation Spectroscopy (FCS) 7.6 Conclusions Acknowledgements References 8
9
10
11
12
Fluorescence Studies of the Hydrogen Bonding of Excited-State Molecules Within Supramolecular Host–Guest Inclusion Complexes Brian D. Wagner 8.1 Introduction 8.2 Hydrogen Bonding Involving Excited States of Fluorescent Probes in Solution 8.3 Hydrogen Bonding of Excited States of Included Guests 8.4 Conclusions References Hydrogen Bonding on Photoexcitation Debarati Dey, Manas Kumar Sarangi and Samita Basu 9.1 Introduction 9.2 Intermolecular Excited-State Hydrogen Bonding 9.3 Concluding Remarks References Effect of Intramolecular H-Bond-Type Interactions on the Photochemistry of Aza-Stilbene-Like Molecules Giampiero Bartocci, Ugo Mazzucato and Anna Spalletti 10.1 Introduction 10.2 Control of the Conformational Equilibria in the Ground State 10.3 Control of Radiative and Reactive Relaxation 10.4 Unusual Adiabatic Photoisomerization in the E ! Z Direction References Hydrogen Bonding Barrier-Crossing Dynamics at Biomimicking Surfaces Rajib Kumar Mitra, Pramod Kumar Verma, Debapriya Banerjee and Samir Kumar Pal 11.1 Introduction 11.2 Materials and Methods 11.3 Results and Discussion 11.4 Conclusion Acknowledgements References Intermolecular Hydrogen Bonding in the Fluorescence Excited State of Organic Luminophores Containing Both Carbonyl and Amino Groups Ilijana Timcheva and Peter Nikolov 12.1 Introduction 12.2 Experimental 12.3 Results and Discussion 12.4 Conclusion References
vii
167 170 172 172 172
175 175 177 180 187 188 193 193 194 202 202 205 205 206 210 211 214 217 217 222 237 259 260 260 269 269 270 270 284 284
viii
Contents
13
Hydrogen-Bonding Effects on Excited States of Para-Hydroxyphenacyl Compounds David Lee Phillips 13.1 Introduction 13.2 Experimental and Computational Methods 13.3 Hydrogen-Bonding Effects on the Excited States of Selected Phenacyl Model Compounds 13.4 Hydrogen-Bonding Effects on the Excited States of Selected Para-Hydroxyphenacyl Ester Phototriggers and the Role of Water in the Deprotection and Subsequent Reactions References
14
15
16
17
Hydrogen-Bonding Effects on Intramolecular Charge Transfer Govindarajan Krishnamoorthy 14.1 Introduction 14.2 Polarity and Viscosity 14.3 Hydrogen Bonding with the Donor Moiety 14.4 Hydrogen Bonding with the Acceptor Moiety 14.5 Conclusion Acknowledgements References Chemical Dynamics in Room-Temperature Ionic Liquids: the Role of Hydrogen Bonding Souravi Sarkar, Rajib Pramanik and Nilmoni Sarkar 15.1 Photoinduced Electron Transfer in a Room-Temperature Ionic Liquid 15.2 Dynamics of Solvent Relaxation in Room-Temperature Ionic Liquids Containing Mixed Solvents Acknowledgements References Vibrational Spectroscopy for Studying Hydrogen Bonding in Imidazolium Ionic Liquids and their Mixtures with Cosolvents Johannes Kiefer 16.1 Introduction 16.2 Experimental Approaches 16.3 Hydrogen Bonding in Ionic Liquids 16.4 Potential, Challenges and Future Applications Acknowledgements References Intramolecular H-Bond Formation Mediated De-Excitation of Curcuminoids: a Time-Resolved Fluorescence Study Luca Nardo, Alessandra Andreoni and Hanne Hjorth Tønnesen 17.1 Introduction 17.2 Experimental Set-Up and Data Analysis Methods 17.3 Results and Discussion 17.4 Conclusions References
287 287 288 289
302 310 313 313 317 318 320 327 327 327 331 331 335 339 339
341 341 342 345 349 349 350
353 353 363 366 373 373
Contents
18
19
20
21
Hydrogen Bonds of Protein-Bound Water Molecules in Rhodopsins Hideki Kandori 18.1 Introduction 18.2 Detection of Water Under Strongly Hydrogen-Bonded Conditions in Bacteriorhodopsin 18.3 Hydration Switch Model as a Proton Transfer Mechanism in the Schiff Base Region of Bacteriorhodopsin 18.4 Time-Resolved IR Study of Water Structural Changes in Bacteriorhodopsin at Room Temperature 18.5 Role of the Water Hydrogen Bond in a Chloride-Ion Pump 18.6 Strongly Hydrogen-Bonded Water Molecules and Functional Correlation with the Proton-Pump Activity 18.7 Conclusion Acknowledgements References Ground- and Excited-State Hydrogen Bonding in the Diazaromatic Betacarboline Derivatives Carmen Carmona, Manuel Balo´n, Marı´a Asuncio´n Mun˜oz, Antonio Sanchez-Coronilla, Jose Hidalgo and Emilio Garcı´a-Fern andez 19.1 Introduction 19.2 MBC–HFIP and MHN–HFIP 19.3 BCA–HFIP 19.4 BC–HFIP 19.5 BC–BC and BC–PY 19.6 Concluding Remarks Acknowledgements References
ix
377 377 379 380 382 384 386 388 388 389
393
393 396 403 406 409 415 416 416
Effect of H-bonding on the Photophysical Behaviour of Coumarin Dyes Sukhendu Nath, Manoj Kumbhakar and Haridas Pal 20.1 Introduction 20.2 Effect of Intermolecular H-bonding 20.3 Effect of Intramolecular H-bonding on ICT to TICT Conversion 20.4 Summary References
419
Role of Hydrogen Bonds in Photosynthetic Water Splitting Gernot Renger 21.1 Introduction 21.2 Photosystem II: Overall Reaction Pattern and Cofactor Arrangement 21.3 Hydrogen Bonds and the Thermal Stability of PS II 21.4 Reaction Sequences of PS II and the Role of Hydrogen Bonds 21.5 Concluding Remarks and Future Perspectives Acknowledgements References
433
419 421 426 429 430
433 434 436 437 452 452 452
x
Contents
Volume II 22
23
24
25
26
Proton Transfer Reactions in the Excited Electronic State Vladimir I. Tomin 22.1 Introduction 22.2 ESIPT in 3-Hydroxyflavones and Some Related Compounds 22.3 Dynamic Quenching of Fluorescence as a Simple Test for Study of Photochemical Reaction Character 22.4 Use of Dynamic Quenching of Fluorescence for Study of Reactions from Higher Excited States 22.5 ESIPT from the S2 Singlet State in 3-Hydroxyflavone 22.6 Concluding Remarks Acknowledgements References Controlling Excited-State H-Atom Transfer Along Hydrogen-Bonded Wires Carine Tanner Manca, Christian Tanner and Samuel Leutwyler 23.1 Introduction 23.2 Prototype System 23.3 What Favours/Prevents ESHAT 23.4 Conclusion Acknowledgements References Excited-State Proton Transfer via Hydrogen-Bonded Dimers and Complexes in Condensed Phase Cheng-Chih Hsieh, Chang-Ming Jiang and Pi-Tai Chou 24.1 Introduction 24.2 Biprotonic Transfer Within Doubly H-bonded Homo- and Heterodimers 24.3 Proton Transfer Through Host/Guest Types of Hydrogen-Bonded Complexes 24.4 Solvation Dynamics Coupled into the Proton Transfer Reaction 24.5 Conclusions References QM/MM Study of Excited-State Solvation Dynamics of Biomolecules Tetsuya Taketsugu, Daisuke Kina, Akira Nakayama, Takeshi Noro and Mark S. Gordon 25.1 Introduction 25.2 Applications 25.3 Concluding Remarks Acknowledgements References Excited-State Intramolecular Proton Transfer Processes on Some Isomeric Naphthalene Derivatives: A Density Functional Theory Based Computational Study Sankar Prasad De and Ajay Misra 26.1 Introduction 26.2 Theoretical Calculations
463 463 467 475 483 509 518 520 520 525 525 527 540 551 551 552
555 555 556 563 567 574 575 579 579 580 587 587 587
589 589 591
Contents
27
28
29
30
xi
26.3 Results and Discussion 26.4 Conclusions Acknowledgements References
591 606 607 607
Conformational Switching Between Acids and Their Anions by Hydrogen Bonding Taka-aki Okamura, Hitoshi Yamamoto and Norikazu Ueyama 27.1 Introduction 27.2 pKa Shift of Acids by Neighbouring Amide NH 27.3 Coordination of Anion Ligand to Metal Ion 27.4 Conclusions References
609
Charge Transfer in Excited States: ab initio Molecular Dynamics Simulations Maciej Kolaski, Anupriya Kumar, Han Myoung Lee and Kwang S. Kim 28.1 Introduction 28.2 Charge-Transfer-to-Solvent-Driven Dissolution Dynamics of I(H2O)2–5 Upon Excitation 28.3 Dynamics of Water Photolysis: Excited-State and Born–Oppenheimer Molecular Dynamics Study 28.4 Photodissociation of Hydrated Hydrogen Iodide Clusters: ab initio Molecular Dynamics Simulations 28.5 Excited-State Dynamics of Pyrrole–Water Complexes: ab initio Excited-State Molecular Dynamics Simulations 28.6 Conclusions References
627
Competitive ESIPT in o-Hydroxy Carbonyl Compounds: Perturbation Through Solvent Modulation and Internal Torsion Sivaprasad Mitra 29.1 Excited-State Proton Transfer: An Overview 29.2 Excited-State Intramolecular Proton Transfer (ESIPT) 29.3 ESIPT in o-Hydroxy Carbonyl Compounds 29.4 Concluding Remarks Acknowledgements References Excited-State Double Hydrogen Bonding Induced by Charge Transfer in Isomeric Bifunctional Azaaromatic Compounds Dolores Reyman and Cristina Dı´az-Oliva 30.1 Introduction 30.2 Pyrrolo-Quinoline Derivatives (PQ, DPC, TPC) 30.3 Methylene-Bridged 2-(20 -Pyridyl)indoles and Pyrido[2,3-a]carbazole (PC) 30.4 Fluorescence Quenching by Electron Transfer in Pyrroloquinolines and PyIn-n 30.5 Betacarboline Derivatives 30.6 Conclusions References
609 610 615 623 625
627 628 630 633 633 636 638
641 641 646 650 657 658 658
661 661 662 673 678 680 705 705
xii
Contents
31
Hydrogen-Bonded Large Molecular Aggregates of Charged Amphiphiles and Unusual Rheology: Photochemistry and Photophysics of Hydroxyaromatic Dopants Moazzam Ali and Swapan K. Saha 31.1 Introduction 31.2 Microstructural Transition of Micelles in the Presence of Inorganic and Organic Salts 31.3 Microstructural Transition of Micelles in the Presence of Neutral Aromatic Dopants 31.4 Photochemistry and Photophysics of Hydroxyaromatic Compounds 31.5 Excited-State Proton Transfer (ESPT) of Hydroxyaromatic Compounds 31.6 ESPT of Hydroxyaromatic Compounds in Organized Media and Some Unusual Emission Phenomena 31.7 Perspectives Acknowledgements References
32
33
34
35
Excited-State Intramolecular Proton Transfer in 2-(20 -Hydroxyphenyl)benzoxazole Derivatives Yi Pang and Weihua Chen 32.1 Introduction 32.2 Intramolecular Proton Transfer in 2,5-bis(20 -hydroxyphenyl)benzoxazole Derivatives 32.3 Summary and Future Prospects References Ultrafast Dynamics of the Excited States of Hydrogen-Bonded Complexes and Solvation Dipak K. Palit 33.1 Introduction 33.2 Identification and Characterization of Hydrogen-Bonded Complex 33.3 Vibrational Dynamics of the C¼O Stretching Mode of Fluorenone 33.4 Dynamics of the Excited States of Hydrogen-Bonded Complex 33.5 Summary and Conclusion Acknowledgement References Volume Changes Associated with Solute–Solvent Reorganization Following Photoinduced Proton Transfer in Aqueous Solutions of 6-Methoxyquinoline Stefania Abbruzzetti and Cristiano Viappiani 34.1 Introduction 34.2 Materials and Methods 34.3 Results and Discussion References Molecular Recognition and Chemical Sensing of Anions Utilizing Excited-State Hydrogen-Bonding Interaction Ashutosh S. Singh and Shih-Sheng Sun 35.1 Introduction 35.2 Recognition and Sensing of Anions by Intramolecular Hydrogen Bonding in Excited States
711 711 712 716 730 735 737 743 743 743
747 747 756 758 759 761 761 762 772 775 790 792 792
797 797 798 799 802
805 805 806
Contents
Recognition and Sensing of Anions by Intermolecular Hydrogen Bonding in Excited States 35.4 Recognition and Sensing of Anions by Conjugated Polymers through ESIPT 35.5 Concluding Remarks References
xiii
35.3
36
37
38
39
Theoretical Studies of Green and Red Fluorescent Proteins Hong Zhang, Qiao Sun, Sufan Wang, Seth Olsen and Sean C. Smith 36.1 Introduction 36.2 Method of Calculation 36.3 Results and Discussion 36.4 Conclusions and Future Work Acknowledgements References Changes in Active Site Hydrogen Bonding upon Formation of the Electronically Excited State of Photoactive Yellow Protein Wouter D. Hoff, Zhouyang Kang, Masato Kumauchi and Aihua Xie 37.1 Central Importance of Light in Biology 37.2 Possible Importance of Excited State Hydrogen Bonding in Photoreceptors 37.3 Introduction to Photoactive Yellow Protein 37.4 Hydrogen Bonding in the Initial State of PYP 37.5 Assignment of Vibrational Modes in PYP 37.6 Identification of Vibrational Structural Markers 37.7 Changes in Hydrogen Bonding During the Initial Stages of the PYP Photocycle 37.8 Sub-Picosecond Time-Resolved Transient Spectroscopy of PYP 37.9 Changes in Active Site Hydrogen Bonding upon the Formation of the S1 State of PYP 37.10 Excited State Proton Transfer in the Y42F Mutant of PYP Acknowledgements References Excited State Dynamics in the Light-Driven Enzyme Protochlorophyllide Oxidoreductase (POR) Marie Louise Groot and Derren James Heyes 38.1 Introduction 38.2 Protochlorophyllide Oxidoreductase (POR) 38.3 Catalytic Mechanism of POR 38.4 Ultrafast Catalytic Processes of the Isolated Pchlide Species 38.5 Ultrafast Catalytic Processes of the Enzyme-Bound Pchlide Species 38.6 Conclusions References Photodissociation of Molecules in Pure and Doped Water and in Nitrogen Heterocyclic Clusters in the Excited State Michal F arnı´k, Petr Slavı´cek and Udo Buck
808 810 813 813 815 815 820 824 834 835 835
839 839 840 840 841 843 843 844 846 848 850 851 851
857 857 858 859 860 861 862 863
865
xiv
Contents
39.1 Introduction 39.2 Experiment 39.3 Aqueous Photochemistry from the Cluster Perspective 39.4 Hydrogen Bonded Clusters of Nitrogen Heterocycles 39.5 General Conclusions and Outlook Acknowledgements References Index
865 866 868 880 888 889 889 893
Editors’ Biographies
Ke-Li Han was born in 1963 in Shandong Province, China. He received his doctorate in 1990 from the State Key Laboratory of Molecular Reaction Dynamics at the Dalian Institute of Chemical Physics and subsequently became an assistant professor at the Dalian Institute of Chemical Physics. He pursued postdoctoral studies at the Emory University and the University of California at Davis in the years 1993–1995. In 1995, he became a full professor of Chemical Physics at the State Key Laboratory of Molecular Reaction Dynamics at the Dalian Institute of Chemical Physics. He was also an adjunct professor at the Dalian University of Technology and Shandong University and a visiting professor at the University of Melbourne, the City University of Hong Kong, the National University of Singapore, the University of California at Berkeley, New York University, the University of Bristol, and so on. Professor Han received the Outstanding Young Scientist award from the National Natural Science Foundation of China in 1998 and the Natural Science Prize (first class) of the Chinese Academy of Sciences and the Young Chemist Prize of the Chinese Chemical Society in 1999, as well as the Natural Science Prize (first class) of Liaoning Province in 2005. His own achievements have been published in over 300 publications. Professor Han’s current research interests involve experimental and theoretical chemical dynamics, including non-adiabatic reaction dynamics of small molecules, the photodissociation dynamics of gas-phase molecules, the excited-state hydrogen-bonding dynamics of large molecules in solution, biochemical reaction mechanisms and dynamics catalysed by enzymes.
Guang-Jiu Zhao was born in 1980 in Hebei Province, China. He received his bachelor’s degree in Material Engineering in 2003 at the Dalian University of Technology. He received his doctorate in Chemical Physics in 2008 from the State Key Laboratory of Molecular Reaction Dynamics at the Dalian Institute of Chemical Physics. Subsequently, he became an assistant professor at the Dalian Institute of Chemical Physics. In 2009, he was promoted to associate professor at the Dalian Institute of Chemical Physics. He has won the Chinese Academy of Sciences Director Award in 2009, the Natural Sciences Research Award of Liaoning Province in 2008, the Lu-Jiaxi Award for Chinese Excellent Graduate Student in 2007, and so on. His research interests are focused on excited-state hydrogen bonding and hydrogen transfer in photophysics, photochemistry and photobiology by the use of combined experimental and theoretical methods.
Reviewer Comments
Professor Richard N. Zare Chair of the Department of Chemistry, Stanford University, USA Hydrogen bonding has always been a bit of a mystery to me, it having the character of directionality but being an order of magnitude or more weaker than a typical covalent bond. Hydrogen bonding can occur between molecules or between different parts of the same molecule. At last, we have a compilation of studies concerning hydrogen bonding and hydrogen transfer reaction in excited-state species, a most welcome addition to the literature on this important topic. I commend the reading of this monograph to all chemists. Professor Donald G. Truhlar Associate Editor of the Journal of the American Chemical Society, Regents Professor of Chemistry, Chemical Physics, Nanoparticle Science and Engineering, and Scientific Computation, Department of Chemistry, University of Minnesota, USA I have just completed looking at the preface, contents and abstracts of the new book on excited-state hydrogen bonding. Although this area is very important in both biological and technological chemistry, the field has been hampered by the lack of a monograph. The book you have assembled is very impressive, with contributions from a remarkably broad set of groups working in this kind of research. I was especially pleased to see that the coverage includes both standard topics and unusual ones, such as hydrogen bonding in triplet states, which is a very interesting subject, and hydrogen bonding in ionic liquids. The book is sure to become a classic in the field. Professor Wolfgang Domcke Chief Editor of Chemical Physics, Chair of Theoretical Chemistry, Department of Chemistry, Technical University of Munich, D-85747 Garching, Germany I have read the tables of contents and the abstracts of the book chapters. This book gives an impressively broad overview of the current research on excited-state hydrogen bonding in chemistry. Numerous organic chromophores and their intramolecular as well as intermolecular hydrogen and/or proton transfer dynamics are discussed in detail. DNA bases, base pairs and photoactive proteins are considered, as well as basic features of the photosynthetic reaction centre and of the photochemistry of water itself. Interesting aspects that are somewhat underrepresented are the role of hydrogen bonds in the excited-state dynamics of peptides and of protonated peptides, the zwitterionic forms of amino acids in water, as well as hydrogen transfer reactions in hydrogen-bonded chromophore–solvent clusters in supersonic jets. Overall, the book represents a good balance of experimental and computational research. The book provides an excellent introduction to an important contemporary research topic for graduate students as well as for experienced researchers.
xviii
Reviewer Comments
Professor Andrjez Sobolweski Institute of Physics, Polish Academy of Science, Poland Thank you very much for your invitation to review the book. As I was in touch with Wolfgang Domcke at the time he was reviewing this book, I am already familiar with this proposal, and my opinion is in line with his comments, including his reservations. Generally, I think the book represents a really good introduction to the topic for a broader readership. Professor C.N.R. Rao Jawaharlal Nehru Centre for Advanced Scientific Research, Bangalore, India It is nice that we have a much-needed book on excited-state hydrogen bonding. This is most welcome and will be useful to workers in the field. Professor Kankan Bhattacharyya Senior Editor of The Journal of Physical Chemistry, Director of the Indian Association for the Cultivation of Science, Kolkata, India This is a comprehensive text that summarizes the latest developments in hydrogen bonding and its role in many fundamental issues. I liked the wide range of topics covered. The 39 chapters spread over nearly 1200 pages dealt with many systems that range from proteins, ionic liquids and micelles to ultracold vapour in supersonic jets. Many spectroscopic (electronic and vibrational) and microscopic techniques with very high temporal and spectral resolution have been used. The primary aim of these volumes is to focus on hydrogen bonding. Implications of this in many issues, such as solvation dynamics, proton/charge transfer and FRET, have been discussed. This will be an excellent textbook and reference material for graduate students and research scientists. Professor Jun Zeng Guest Professor, Sichuan University, China, and Chief Scientific Officer, Qubist Molecular Design, Australia This is probably the first book that presents comprehensive reviews on the recent theoretical and experimental investigations on the nature of excited-state hydrogen bonding and hydrogen transfers and their influences on many aspects of photophysics, photochemistry and photobiology. From this book, readers will gain much insightful information on the structure, dynamics and spectroscopic properties of hydrogen bonding in the excited states of many important chemical and biological systems. A very useful reference book! Professor Steven D. Schwartz Biophysics and Biochemistry, Albert Einstein College of Medicine, USA This volume promises to be of significant value. Proton transfers are ubiquitous in both complex condensed phases as hydrogen bonds and in biological systems both as hydrogen bonds and as (one of) the chemical steps in enzymatic reactions in biology. In addition, biotechnology through such reactants as GFP is critically dependent on hydrogen transfer. This volume, containing both experiment and theory, promises many useful reviews and new results. In addition, for a western audience, the volume has the advantage of including authors well known to the American audience and others whose work will be new. Professor Hans Lischka Institute for Theoretical Chemistry, University of Vienna, Waehringerstrasse 17, A-1090 Vienna, Austria I have read the Table of Contents of your book Hydrogen Bonding and Transfer in the Excited State with great interest. It contains excellent chapters written by leading world scientists. I am sure that the book will be a great success.
Reviewer Comments
xix
Professor Chang-Guo Zhan Department of Pharmaceutical Sciences, University of Kentucky, USA I am pleased to read your detailed plan for a book entitled Hydrogen Bonding and Transfer In the Excited State. I think this will be an important book that covers all aspects of hydrogen bonding and hydrogen transfer in excited states. The book will be very interesting for all scientists in the field of chemistry, biochemistry and biophysics who are interested in hydrogen bonding or hydrogen transfer in excited states. I hope this book will be published as soon as possible. Professor Jeffrey R. Reimers ARC Professorial Research Fellow, School of Chemistry F11, The University of Sydney, Sydney NSW 2006, Australia Biochemical structure and function always involve a delicately controlled balance of hydrophobic and hydrophilic forces. While the hydrophobic force is non-specific and always present, the hydrogen-bonding interactions that empower the hydrophilic forces are specific and directed. They are critical to molecular recognition, driving the secondary structure of proteins and the helix formation of DNA. But life is more than biological structure – it is dynamics and motion, metabolism and vitality. What happens to hydrogen bonds in systems with excess energy? Can molecular recognition be modified and a cascade of biological processes ensue? How are proteins and DNA modified when molecules absorb light? Sometimes a change just happens from one possible tautomeric form to another, sometimes whole new motifs like strong hydrogen bonding to aromatics occurs. How quickly do these processes occur, how quickly is the energy dissipated and how quickly does the system return to normal? This is the first book to review excited-state hydrogen bonding, detailing the great variety of consequences found. It provides new insights into the very nature of the forces that create secondary structure in chemistry and biology. Professor Zhi-Ru Li State Key Laboratory of Theoretical and Computational Chemistry, Jilin University, China Hydrogen bonding plays an important role in chemistry, biology and physics. Research on excited-state hydrogen bonding and hydrogen transfer is a novel field. Excited-state hydrogen bonding and hydrogen transfer play significant roles in many photophysical processes and photochemical reactions. This book includes 39 chapters covering various frontier areas of excited-state hydrogen bonding. The contents of this book are very rich. This is very beneficial for researchers and graduate students who are interested in the fields of molecular and supramolecular photochemistry, photobiology and photophysics. Professor Shengli Zou Chemistry Department, University of Central Florida, USA Hydrogen bonding is one of the most important and complex interactions between different molecules or different groups in a big molecule, especially a biomolecule. Understanding the roles of hydrogen bonding in chemical reactions, proton transfer and charge transfer is crucial in revealing the mechanism of these processes. The authors address hydrogen-bonding-induced charge transfer, conformational switching between acids and their anions and controlled intramolecular proton transfer. The importance of hydrogen bonding in photosynthetic water splitting and green fluorescence protein is also discussed. The investigation of hydrogen bonding involving electronically excited molecules is a substantial challenge both experimentally and theoretically. There are few books focusing on hydrogen bonding of molecules in excited states owing to the complexity of the system, especially for theoretical calculations. The proposed book will be a helpful
xx
Reviewer Comments
reference book for research groups interested in understanding hydrogen bonding in different environments and processes. The book is highly recommended for publication. Professor Anna Spalletti Dipartimento di Chimica, Universit a di Perugia, 06123 Perugia, Italy Thank you for the information about the new book on hydrogen bonding. My coauthors and I congratulate you on your success in collecting, in a relatively short time, such abundant material (39 contributions!) on a variety of aspects of HB effects on the spectral, photophysical and photochemical properties of so many different organic compounds. From a glance through the abstracts we did not notice any omission. Some discrepancies (for example, in the length of the chapters) and repetitions will certainly be present, but this is bound to happen in such a large review work. Best wishes for the success of the book. Professor Noam Agmon Institute of Chemistry, Hebrew University, Jerusalem The skeleton of the new book looks very impressive in its scope: 39 chapters by world experts covering different aspects of excited-state dynamics within hydrogen-bonded systems. At this stage, when only abstracts are available, it is hard to say more, but I am definitely waiting eagerly for this project to appear in print, as I believe it will be an important milestone for those working in the field and those considering doing so. Professor Gernot Renger Max-Volmer-Laboratorium f€ ur Biophysikalische Chemie, Technische Universit€at Berlin, Strasse des 17. Juni 135, 10623 Berlin, Germany Hydrogen bonds are the most important structural determinants in nature. Striking examples of the paramount role of hydrogen bonds are the unique properties of water and the structure of DNA and proteins. The functional relevance of hydrogen bonds is clearly illustrated by their participation in proton transfer mechanisms (e.g. the Grotthus mechanism, proton-transfer-coupled electron transfer, etc.). Of special interest in basic research are the properties of hydrogen bonds in electronically and vibronically excited molecules. This book is an excellent summary of our current stage of knowledge on the different facets of hydrogen bonding which plays a central role for the interaction between molecules. It covers, in 39 chapters, a wide field of topics ranging from the basic properties of hydrogen bonds in comparatively simple electronically excited pigments to the role in complicated biological systems like rhodopsins and photosynthetic water splitting. This book will find a broad audience. It is of great value for scientists working on various aspects of hydrogen bonding. It provides, in a single publication, a nice overview of such a wide field of different topics. Professor Jingwen Chen Key Laboratory of Industrial Ecology and Environmental Engineering, Department of Environmental Science and Technology, Dalian University of Technology, Linggong Road 2, Dalian 116024, China Hydrogen bonding determines the properties and activities of many compounds, which is of great importance in chemistry, biology, physics and environmental science. Electronically excited-state hydrogen bonding and hydrogen transfer play an increasingly important role in many photophysical processes and photochemical reactions. In the field of environmental science, studies in recent decades have proved that photodegradation is an important transformation or degradation pathway for toxic organic compounds in aquatic and atmospheric environments, and environmental media have great effects on the photodegradation kinetics and pathways. In some cases, environmental media were observed to influence the photodegradation via hydrogen bonding or hydrogen transfer. Excited-state hydrogen bonding and hydrogen transfer may determine the indirect/direct
Reviewer Comments
xxi
photodegradation kinetics and pathways of many organic pollutants, including halogenated aromatic compounds (e.g. polychlorinated dibenzo-p-dioxin/dibenzofurans, polychlorinated biphenyls, polybrominated diphenyl ethers), pesticides, pharmaceutical and personal care products, etc. Excited-state hydrogen bonding and hydrogen transfer may also have great impacts on the photoinduced toxicities of organic pollutants. This monograph will be the first to deal with hydrogen bonding in excited states, presenting an extensive description of the research progress on excited-state hydrogen bonding and hydrogen transfer in recent years. Both experimental and theoretical investigations on excited-state hydrogen-bonding structures and dynamics of many organic and biological chromophores are included. There are also several chapters describing the influences of excited-state hydrogen bonding and hydrogen transfer on various photophysical processes and photochemical reactions. Thus, this book will be very helpful in understanding the nature of hydrogen bonding in relevant areas and in understanding the photochemical transformation/photoinduced toxicity of environmental organic pollutants. Professor Brian D. Wagner 3M Canada National Teaching Fellow, Department of Chemistry, University of Prince Edward Island, Charlottetown, PE C1A 4P3, Canada I am very impressed by the comprehensive coverage represented by the many chapters in this two-volume set. All of the major topics and considerations involving hydrogen bonding of excited states have been covered. This will be a very useful set of books for a wide range of researchers. I am proud to have been able to make a contribution to the book. Professor Hiroshi Sekiya Department of Chemistry, Faculty of Science, Kyushu Unviversity, 6-10-1 Hakozaki, Higashi-ku, Fukuoka 812-8581, Japan This book covers a very wide range of topics on hydrogen-bonding and excited-state proton/hydrogen transfer reactions in various molecules and molecular clusters developed by spectroscopic meaasurements and theoretical studies. Many of the results are quite new and interesting for physists, chemists and biologists. I would like to recommend this book for many young and senior researchers interested in the intriguing field of hydrogen bonds and proton/hydrogen transfer reactions. Professor James C. Crabbe Professor of Biochemistry, Dean of the Faculty of Creative Arts, Technologies and Science, University of Bedfordshire, Park Square, Luton LU1 3JU, UK This is an exciting new publication on one of the key elements of life – the hydrogen bond. The authors have produced an array of exciting chapters on hydrogen bonding and hydrogen transfer, covering many aspects of chemistry and biochemistry. This will be an important reference work for many years to come. Professor Takayuki Ebata Department of Chemistry, Graduate School of Science, Hiroshima University, Higashi-Hiroshima 739-8526, Japan This book covers recent experimental and theoretical studies on the dynamics of H-bonded systems from the simple aromatic molecules to real biomolecules. An interesting point is that it concentrates on the topic of the electronic excited state, which is different from other books published so far. In this sense, I think (and hope) this book will attract people in a variety of fields.
xxii
Reviewer Comments
Professor Jianzhang Zhao State Key Laboratory of Fine Chemicals, Dalian University of Technology, Dalian 116024, Liaoning, China This book focuses on excited-state hydrogen bonding and proton transfer. The chapters cover a wide range from the formation of the hydrogen bond in the excited state to the fate of the hydrogen bond in the excited state, such as ESIPT (excited-state intramolecular proton transfer) and vibration dissipation of excited-state energy. Experimental as well as theoretical methods are employed to elucidate hydrogen bonding in the excited state, such as time-resolved vibrational spectra and ab initio or DFT calculations. The subjects involved in the discussion are very diverse, ranging from small organic molecules (such as fluorescent dyes) to biological systems (such as DNA). Therefore, I believe this book addresses most of the research topics of excited-state hydrogen bonding, and the publication of the book will be of significance for the scientific community. Professor Mihaela Hillebrand Department of Physical Chemistry, Faculty of Chemistry, University of Bucharest, Bd. Regina Elisabeta 4-12, Bucharest, Romania The book encompasses the latest achievements in excited-state hydrogen-bonded systems by means of experimental and computational methods. The papers collected provide a good insight into how advances in ultrafast spectroscopic techniques and state-of-the-art quantum chemical calculations have opened up new perspectives on excited-state processes, namely hydrogen bond formation and hydrogen bond transfer in a wide range of chemical and biochemical hydrogen-bonded systems, from molecules, clusters or complexes to biopolymers. It is the first monograph devoted to this subject, and its publication is worthwhile from two points of view – the overall subject and the content. Firstly, considering the importance of hydrogen bond formation in many chemical and biochemical processes and the difficulties related to a good understanding of the excitedstate photophysics, a comprehensive treatment of the subject is necessary. Secondly, the book covers most of the aspects of the topic and is characterized by a good balance between a review of up-to-date literature data and some new results. The book benefits from contributions by renowned scientists with acknowledged results in the field. The editors, remarkably, have succeeded in putting together theoretical aspects involved in excitedstate hydrogen photodynamics and possible applications. The book Hydrogen Bonding and Transfer in the Excited State will be a good tool both for researchers in the field and for graduate students. Professor Samir Kumar Pal Unit for Nano Science & Technology, Department of Chemical, Biological & Macromolecular Sciences, S.N. Bose National Centre for Basic Sciences, Block JD, Sector III, Salt Lake, Kolkata 700098, India The book proposal, entitled Hydrogen Bonding and Transfer in the Excited State and edited by Ke-Li Han and Guang-Jiu Zhao, consists of 39 chapters contributed by eminent scientists in this field from all over the world. The contributions embodied here are mostly based on experimental results, along with six papers based on theoretical calculations and simulation results. The theme of the proposed book is very interesting, as many of the fundamental processes in photophysics and photobiology occur in the excited states and involve formation and/or rupture of the hydrogen bonds (HB). A very popular example of such a process is the water splitting in photosynthesis. Ultrafast proton transfer (PT) also serves as the key reaction in many important processes. Unfortunately, there has been no such monograph in the present literature that discusses the various aspects of HB and PT in the excited state. In this regard, this attempt to gather important information on HB and PTwithin a single cover is very encouraging. The proposed book mainly consists of experimental results obtained from steady-state and time-resolved fluorescence studies, as this technique extracts the maximum valuable information on the excited state. Other experimental techniques dealt within the book are UV-IR double-
Reviewer Comments
xxiii
resonance excitation, time-resolved resonance Raman spectroscopy, time-resolved FTIR, etc. The related changes in basicity, solvation, hydrogen bond dynamics and other fundamental photophysical and photochemical properties of fluoroprobes upon excitation are discussed and reviewed in many chapters (e.g. Chapters 3, 5, 6, 7, 20, 31, etc.). Intermolecular charge transfer (ICT) is discussed in Chapters 4 and 14. The excited-state photochemistry and photophysics of many organic molecules are discussed in Chapters 12, 13, 19, 25, 27, 30 and 32. Excited-state PT and energy transfer (ET) in different molecules and solvents are discussed in Chapters 10, 17, 19, 22, 23, 24, 27, 29, 32, 33 and 34. Chapters 15 and 16 deal with the hydrogenbonding dynamics in room-temperature ionic liquids (IL). HB barrier-crossing dynamics in nanoconfinement is discussed in Chapter 11. Excited-state HB in biologically important molecules like nucleic acids (Chapters 1 and 9), bacteriorhodopsin (Chapter 18), green fluorescent protein (Chapter 34) and photoconductive yellow protein (Chapter 35), as well as in some complex systems like host–guest complexes (Chapter 8), worm-like micelles (Chapter 29) and clusters (Chapters 1, 2, 26 and 37), is also discussed. Some very interesting topics involving excited-state HB and PT, like the water splitting process in photosynthesis (Chapter 21), excitedstate H-atom transfer (Chapter 24), phototautomerization (Chapter 28) and the catalytic process in light-driven enzymes (Chapter 36), are also included in this monograph. All the contributions embodied in this proposed book are supported by state-of-the-art experimental and computational results, and the topics cover the wide range of diversity in this field. In my opinion this monograph will serve as a very fundamental tool for understanding excited-state HB and PT processes for researchers in the field of photophysics, photochemistry and photobiology. I strongly recommend the publication of this monograph. Professor Soo Young Park School of Materials Science and Engineering, Seoul National University, Korea Congratulations on your excellent publication. It seems that your book covers all aspects of excited-state H-bonding and H-transfer. This book will draw the attention of scientists in many different disciplines such as the organic, physical, as well as materials chemistry fields. Professor Swapan K. Saha Department of Chemistry, University of North Bengal, Darjeeling-734 013, India Thank you for the mail and the attachments. You have done a great job! Congratulations! The coverage of the proposed book is wide and impressive. The authors are mostly of international standing and the topics covered are up to date and relevant to current interest. Professor Weiqun Zhou College of Chemistry and Chemical Engineering, Soochow University, Suzhou 215123, China The studies on the hydrogen bond have been one of the most important research areas in materials chemistry, chemistry and bioscience. The hydrogen bond has a special significance for biomacromolecules; it is part of the reason why protein and level II, level III and level IV nucleic acid can be stable. The excited-state hydrogen bond structure and dynamics play an important part in many chemical, physical and biological procedures. The fluorescence emission behaviours of organic and biological chromophores are often influenced by the interactions of hydrogen bonds between chromophore molecules and protic solvents or biological surroundings. The ultrafast deactivation processes of the photoexcited molecules and the supramolecular systems are also likely to be easier under the influence of excited-state hydrogen bonding. The hydrogen bond in the excited state and hydrogen transfer are becoming an increasingly important subject in the realm of photochemical and photophysical reactions. The publishing of this book will help us to learn more comprehensively the characteristics of the hydrogen bond and also help us to realize the
xxiv
Reviewer Comments
importance of the hydrogen bond in photochemistry, photobiology and photophysics. We sincerely hope that this book, with systematic description of the hydrogen bond in its excited state and hydrogen transfer, can be published soon. Professors Sean C. Smith and Hong Zhang Centre of Computational Molecular Sciences, University of Queensland, Australian Institute of Bioengineering & Nanotechnology, ARC Ctr Funct Nanomat, Brisbane, Qld 4072 Australia The publication of the book Hydrogen Bonding and Transfer in the Excited State is a timely landmark contribution to the field, drawing together a wide range of theoretical and experimental contributions that collectively provide a comprehensive picture of recent advances in the field. It covers the recent important work of the experts in this field from all over the world and coherently links the theoretical studies with the experimental developments in this important area. The 39 contributing chapters are well written and thematically organized. The book is of high quality and will no doubt become a mandatory part of library and personal collections for institutions and individuals – researchers and students alike – engaged in this fascinating area of molecular science. We look forward to its publication as soon as possible. Professor Attila Demeter Institute of Materials and Environmental Chemistry, Chemical Research Centre of Hungarian Academy of Sciences, 1525 Budapest, P.O. Box 17, Hungary The proposed book Hydrogen Bonding and Transfer in the Excited State, edited by Ke-Li Han and Guang-Jiu Zhao, is a stop-gap issue that may reckon with considerable interest in the field. There is no really well-known monograph on this discipline from the classical books of Pimentel (1960) and Vinogradov (1971), although the subject is widely studied. The 39 studies cover a very wide area, indicating that the understanding of the influence of hydrogen bonds on photoprocesses is crucial almost everywhere. Most expert readers will find half a dozen studies touching upon their close interests; however, the book will be a valuable tool for obtaining knowledge on topics slightly further afield. One rarely has time to collect such scientific studies from journals, and it is certainly valuable to have them gathered together by expert editors. Professor H. H. Limbach Institut f € ur Chemie und Biochemie, Freie Universit€at Berlin, Takustr. 3, 14195, Berlin This is a timely book in two volumes and 39 chapters, to which many well-known authors from all over the world have contributed. The systems studied are dyes, water wires, ionic liquids and nanoconfined and self-assembled systems up to biomolecules and large proteins. The appetite of the reader is whetted by chapters covering excited-state phenomena such as vibrational dynamics, acid–base interactions, proton and charge transfers, H-bond-induced conformational switching and molecular recognition, as well as the function of complex proteins. The experimental and theoretical techniques used are adapted to the systems and phenomena studied. It will be an important piece in the canon of books on hydrogen transfer and bonding. Professor Johannes Kiefer University Erlangen Nurnberg, LTT, Weichselgarten 8, D-91058 Erlangen, Germany I very much like the fact that a broad range of aspects is discussed in this book concerning both the analytical methods and the systems under investigation. Therefore, it will be of interest for a large readership in the classical fields of physics and chemistry, but also for rather new areas like life science and biophysics.
Reviewer Comments
xxv
Professor Giuseppe Buemi Dipartimento di Scienze Chimiche, Universit a di Catania, Viale A. Doria nr. 6, 95125 Catania, Italy I have read the summary of the papers enclosed in the new book you have coedited with Prof. Han. Even if I have little experience with excited states, I think that such a book could be very interesting and very useful for collegues working in this field, and so I think the book must be published. Professor Dipak K. Palit Radiation & Photochemistry Division, Bhabha Atomic Research Centre, Mumbai 400 085, India This monograph, presented in two volumes, provides a very timely update on the recent developments in the field of hydrogen-bonding interactions in the excited states of different kinds of molecular system in both homogeneous solutions as well as heterogeneous media, including micelles, vesicles and ionic liquids. As rightly mentioned by the editors in the preface, hydrogen-bonding structures and dynamics in the excited states of molecules play important roles in determining many chemical, physical and biochemical processes. In addition to the most popular fluorescence spectroscopic techniques, recent developments of ultrafast, both linear and nonlinear, time-resolved infrared spectroscopic techniques have provided a great opportunity to understand the microscopic structures and functions in many complex hydrogen-bonded systems. While there are quite a good number of monographs published describing the hydrogen-bonding interactions in the ground state of molecules, to the best of my knowledge there is none to deal with the same aspect exclusively in the excited states of molecules. This book presents an extensive review of the progress of research, both experimental and theoretical, on hydrogen bonding and hydrogen transfer, both intramolecular and intermolecular, in the excited states of a wide variety of molecular systems. This book comprises 39 chapters, most of which are written by experts and provide authoritative overviews of each area. Overall, the editors have fulfilled their primary objective of creating a reference volume valuable to both experts and beginners or students who are engaged in investigation of the dynamics of hydrogen-bonding interactions in the excited states of molecular systems forming hydrogen-bonded complexes. This book will provide an excellent entry to the literature of hydrogen bonding and hydrogen transfers in the excited state. Professor Andong Xia The State Key Laboratory of Molecular Reaction Dynamics, Institute of Chemistry, Chinese Academy of Sciences, P.R. China The book entitled Hydrogen Bonding and Transfer in the Excited State, edited by Keli Han and Guangjiu Zhao, is a timely and important work for researchers working on the excited-state hydrogen-bonding structure and dynamics. Reading this book will help the reader understand the basic concepts of complex excited-state hydrogen-bonding processes. There are at least three advantages of this book: 1.
2. 3.
The topics covered are extensive and comprehensive. Volume I introduces the structure and dynamics in excited-state hydrogen-bonding systems, and the influences of excited-state hydrogen bonding on photophysical and photochemical processes. Volume II then focuses the attention on the dynamics and control of the excited-state hydrogen proton transfer process. A series of organic chromophores and biomacromolecues in different systems, as well as their inter- and intramolecular hydrogen proton transfer dynamics, are discussed in detail. It has a ‘handbook’ character to some extent, and it is easy to understand the basic concepts of hydrogen bonding for systems specific to researchers. It is very authoritative. The contributed authors are distinguished scholars in this field. Their studies are sufficiently representative of the current overall level in this field. It pays attention to both experimental and theoretical studies. This will be helpful and welcome to experimental researchers seeking theoretical support, and vice versa.
xxvi
Reviewer Comments
Professor Laszlo Biczok Hungarian Academy Sciences, Chemistry Research Centre, POB 17, H-1525 Budapest, Hungary This book provides a unique comprehensive overview of the photoinduced processes of hydrogen-bonded systems. The chapters, written by internationally recognized experts, cover the latest developments and fundamental aspects of this rapidly evolving research field. In view of the ubiquitous nature of hydrogenbonding and light-initiated processes, this excellent reference book is likely to be of interest to members of a wide scientific community. It serves as a valuable source of information and inspiration for newcomers and experienced researchers alike. Professor G. Krishnamoorthy Department of Chemistry, IIT Guwahati, Guwahati 781039, India Hydrogen bonding is a fundamental phenomenon that plays a key role in various chemical and biological processes. The hydrogen-bonding effects may be altered significantly upon molecular excitation owing to redistribution of electron density in the excited state. This will have a drastic effect on the photochemistry and photophysics of the system. Thus, excited-state hydrogen bonding and hydrogen transfer are very important subjects of study. In this book, recent advances on both experimental and theoretical studies of hydrogen bonding in photochemistry and photophysics are reviewed. The effect of hydrogen bonding on various phenomena, such as proton transfer, charge transfer, isomerization and photodissociation by conventional solvents to complex proteins, clusters and ionic liquids, are discussed. The book will be a useful reference to active researchers and graduate students. Professor Luca Nardo University of Insubria, Dept. of Physics and Mathematics & C.N.I.S.M.-C.N.R., U.d.R. Como Via Valleggio 11, 22100 Como, Italy Concerning the book as a whole, we sincerely think that it has resulted in a really successful editorial project and are proud of having taken part in it. The book is a detailed and comprehensive compendium on all the principal aspects of H-bonding and H-transfer in the excited state, both from the experimental and from the theoretical point of view. The approach of the book sounds intriguingly interdisciplinary. In this regard, we believe that the readership could be quite broad, and that it could be helpful to divide the book into parts/sections, each section collecting chapters on similar topics and being opened by the chapter offering the most complete introduction on both the theory and the experimental techniques eviscerated in the section itself. This should simplify the task of finding specific information within the very wide spectrum of contents. Moreover, the readers would best appreciate the unity and the development of the arguments, and the monographic character of the work. To this purpose, we also suggest that the editors write a short summary/abstract for each section. We see that the order in which the chapters are presented already seems to match these suggestions, and only wish to highlight the opportunity to take up these suggestions. Professor Oliver Kuhn Institut f € ur Physik, Universit€ at Rostock, D-18051 Rostock, Germany As regards the overall impression of the book you have compiled (based on the Table of Contents), I was very excited. You have managed to bring together a very broad range of people essentially covering many of the fascinating subjects of this field. Congratulations!
Reviewer Comments
xxvii
Professor Pi-Tai Chou Department of Chemistry, National Taiwan University, Taipei, Taiwan 106 I thank the editors, Dr Ke-Li Han and Dr Guang-Jiu Zhao, for the invitation to contribute a chapter to this new book. The book extensively covers the hot topics of hydrogen bonding and the associated proton (hydrogen) transfer reaction, from theoretical and experimental approaches to the fundamentals to current-interest biological and material applications. I found that this book may be particularly suited to those readers who are at the stage of initiating proton (hydrogen) transfer research and need a broad spectrum of current/ previous progress in various aspects. The reader can specifically pick a few chapters for his/her own interest and treat other chapters as key references. This would make the reading more convenient and comfortable. Evidently, the contents of this book are very rich and provide an in-depth discussion of various theories on spectroscopy and dynamics. The reader will be able to discern differences, for example in applications, between similar topics, and, conversely, similarities, for example in theory, between different topics. I thus believe that by reading this book the reader will gain deep and broad insights into hydrogen-bonding phenomena and the associated excited-state proton (hydrogen) transfer reaction and perhaps latent applications in several cutting-edge areas. As for the contemporary research progress in hydrogen-bonding studies, the book is indeed a significant milestone for studying excited-state hydrogen bonding and/or hydrogen transfer reactions. Professor Ricard Gelabert Unitat de Quimica Fisica, Departament de Quimica, Edifici Cn Universitat Autonoma de Barcelona, 08193 Bellaterra, Spain First of all, I have not had access to the contents of the book, either in full or in part, except for the index, the foreword and the abstracts of the different chapters, and as such all I can provide is a general overview or impression of the coverage of the topics related to excited-state proton transfer, as far as my knowledge of the field permits, but not a view on the quality of each of the chapters. The book is made up of 39 chapters, each authored by an author or group of authors active in the general field of hydrogen bonding and hydrogen/proton transfer in excited states. The list of chapters is quite extensive and covers several current topics in ESIPT and hydrogen bonding. In a broad sense, the publication of a book devoted to excited-state hydrogen transfer/bonding is good news, as there has been a certain lack of specialized texts on the state of research. This is a collection of specialized papers on current research in this area, and may be of help in particular to researchers in the field, as a guide to the state of the topic at the beginning of this century. The Table of Contents shows the contributions from experimental and theoretical researchers, which should provide a balanced view of research in an area where synergy between theory and experiment has proven to be paramount. Even though the detailed contents of each chapter are unavailable to me, a list of sections and authors is. Thus, a rather balanced and wide spectrum of topics greets the eye. The interested reader will find chapters dealing with vibrational dynamics aspects of hydrogen bonding, acid–base character, solvent effects, supramolecular chemistry, hydrogen bonding in ionic liquids, properties and behaviour of many landmark systems, including fluorescent proteins, rhodopsin and others. Last but not least, a number of the contributions are written by well-known researchers in their field, as is the case, for instance, with no pretence of completeness, with the chapters devoted to GFP/RFP and the vibrational analysis of hydrogen bonds in nucleic base pairs. As a downside, I have found a certain omission in the chapters, namely systems displaying intramolecular double-proton transfers (or multiple ones). These systems have proved to be more complex than expected, as different mechanisms can be envisaged, usually involving zwitterionic intermediates. I could only find one chapter where homodimers of 7-azaindole are described. Another issue that might have deserved specific coverage is that of nonradiative pathways, ubiquitous in excited-state proton transfer and usually a highly competitive process to
xxviii
Reviewer Comments
fluorescence. Certainly, the broad topics covered in the book make it necessary to limit the scope covered within the limited confines of the book. Professor Samita Basu Chemical Sciences Division, Saha Institute of Nuclear Physics, 1/AF, Bidhannagar, Kolkata 700 064, India This book is an attempt to meld together theory and experiment in the fundamental aspect of hydrogen bonding, especially in the excited state. It is a collaborative effort, made possible by the willingness of a group of leading figures in the subject to write about their own areas of expertise. It contains 39 chapters, the themes of which are summarized very elegantly in the Preface. This book will aid those who wish to understand the structural, physical, chemical and biological basis of hydrogen bonding and plan their own experiments. It should be of value for graduate students and advanced undergraduates. Professor Satoshi Minakata Department of Applied Chemistry, Graduate School of Engineering, Osaka University, Yamadaoka 2-1, Suita, Osaka 565-0871 Japan The book deals with the recent achievements in exited-state hydrogen bonding and hydrogen transfer. Hydrogen bonding and transfer are very important interactions not only in chemistry but also in biology and physics. In particular, the studies of hydrogen bonding in the exited state should contribute to the development of these fields. This book should have a beneficial effect on the development of various investigations and provide valuable information to research scientists as well as graduate students. Professor Rafael Escribano Inst. de Estructura de la Materia, CSIC Serrano 123, 28006 Madrid, Spain This book presents a very comprehensive collection of interesting contributions to the subject of hydrogen bonding in excited-state molecules, plus a number of chapters dealing with hydrogen transfer processes. Whereas there have been many articles and books on the general subject of hydrogen bonding over the years, it is true that the number of scientific communications on the specific area of excited-state hydrogen bonding is small. In this regard, this book fulfils an important task in providing high-level, state-of-the-art contributions to this particular topic. The chapters cover a wide range of scientific communications on different applications that have in common the existence of this type of bonding in the molecular systems under study. Many of the subjects considered deal with biological species, which are of course of high interest in current science. Perhaps an introductory chapter on the general aspects of hydrogen bonding, such as the spectroscopic effects that this kind of bonding has on bandwidths, bandshifts and intensities, and a basic description on the computational techniques applied to its study, even at the ground-state level, could be added at the beginning of the book. Similarly, there do not seem to be any contributions dealing with hydrogen bonding in solid-state systems, which are of course much less common than in the liquid phase. On the whole, the book is excellent and can be wholeheartedly recommended for researchers and university teachers interested in this field. Professor Shih-Sheng Sun Institute of Chemistry, Academia Sinica, 128 Sec2 Academia Road, Taipei, Taiwan I think you have done a fabulous job in bringing together all these topics related to excited-state H-bonding. The subtopics cover many different aspects of H-bonding properties in the excited states. It will be a nice contribution to this area as a valuable reference book.
Reviewer Comments
xxix
Professor Chong Rae Park School of Materials Science and Engineering, Seoul National University, San 56-1, Shinlim-dong, Kwanak-ku, Seoul 151-744, Korea Hydrogen bonding is perhaps the most interesting and complex bonding that prevails in the material world, ranging from natural lives to man-made organic and/or inorganic materials. In spite of the many excellent review papers and monographs already available, there still remain many phenomena due to the presence of hydrogen bonds that require deeper understanding. Apart from hydrogen bonds themselves, understanding these bonds in the excited state and related phenomena is even more intriguing and absolutely necessary if we are to deepen our understanding of the mysterious materials world and phenomena that are created by excited-state hydrogen bonding. In this sense, the book will definitely be a good guide to the yet unknown world of lives and materials. As mentioned in the Preface, this book would be a good reference and/or textbook for postgraduates and experts in materials science who are dreaming of designing new functional materials. Professor Sivaprasad Mitra Department of Chemistry, NEHU, Shillong 793 022, India I appreciate your idea of editing a whole new book (of two volumes) on this topic. The effect of HB is well known and has been studied in detail in the ground state over the years. However, recent developments in ultrafast laser spectroscopy, as well as theoretical chemistry, have helped scientists to explore this phenomenon in the excited state with more certainty. In spite of the vigorous activity (more than 1 150 000 hits resulted from a search in google scholar with hydrogen bond + excited state!), the results are scattered and some compilation was absolutely necessary. I think and sincerely hope that your effort (with the compilation of 39 chapters from different parts of the world) will help immensely in this regard. Professor Mikhail V. Vener Department of Quantum Chemistry, Mendeleev, University of Chemical Technology, Moscow 125047, Russia I was impressed by the Table of Contents of the book. It contains the results of experimental investigations of proton/hydrogen atom transfer in excited-state hydrogen-bonding systems of different type and theoretical studies of the proton-transfer dynamics in biomimicking molecules. The book bridges the gap between the well-developed concepts of the hydrogen bond phenomenon in the ground electronic state and hydrogen bonding in the excited electronic states. Summing up, the book provides valuable insights into the nature of hydrogen bonding and proton dynamics in supramolecular photochemistry and photobiology. Professor Jim Jr-Min Lin Institute of Atomic and Molecular Sciences, Academia Sinica, Taipei, Taiwan 10617 It is a comprehensive book. There would be no life without hydrogen bonding. Although focusing on hydrogen bonding, the broad coverage of this book makes it also a nice reference book for studying reactions and dynamics, especially for those who are interested in excited-state chemical dynamics. Professor Satinder Kumar Sikka Off. Principal Sci. Adviser Govt India, Vigyan Bhawan Annexe, New Delhi, India The book is perhaps the first one on the novel topic of excited states of hydrogen bonds, a field that has applications ranging from biology to climate change. It will serve as an excellent resource book for researchers in this field. Personally, I learnt a lot of things about this subject by reading through the abstracts.
xxx
Reviewer Comments
Professor Tetsuya Taketsugu Division of Chemistry, Graduate School of Science, Hokkaido University, Sapporo 060-0810, Japan This book is well organized, and a lot of good scientists have contributed to the chapters. I believe this book should play a significant role in this field. Professor Sergio Trasatti University of Milan, Dept Phys Chem & Electrochem, Via Venezian 21, I-20133 Milan, Italy The editors of this new book have to be congratulated for the impressive amount of information gathered from internationally renowned groups, which fills a gap in the dedicated literature. The inclusion of recently developed topics such as ionic liquids is particularly interesting. Hydrogen bonding is relevant also to the field of electrochemistry, where heterogeneous electron and ion transfer are influenced by interfacial hydrogen bonding. The reaction of heterogeneous proton transfer is one of the more studied topics in electrochemistry. Professor Udo Buck Max-Planck-Institut fuer Dynamik und Selbstorganisation, delivery: Bunsenstr. 10, 37073 Goettingen, Germany The new book is an important contribution to the rapidly evolving field of processes in the excited state of hydrogen-bonded systems, with crucial applications to biological systems. These include photophysical and photochemical reactions, the behaviour of many chromophores and hydrogen transfer reactions in the excited state. Most of the presented examples take place in solutions. But some of them also deal with specially prepared nanoconfined surfaces and clusters of different sizes. These systems allow for a microscopic description of some of the key processes and form an important source for the understanding of what is going on in these systems. Professor Yi Pang Department of Chemistry & Maurice Morton Institute of Polymer Science, University of Akron, Akron, OH 44325, USA This much needed monograph places the subject of H-bonding and H-transfer in the excited state in its most modern context. With a broad range of applications in chemistry, biology and material sciences, this authoritative resource will certainly benefit researchers and graduate students who wish to acquire an insight into H-bonding in the excited state. Professor Alenka Luzar Department of Chemistry, Virginia Commonwealth University, 1001 West Main St., Richmond, VA 232842006, USA I have just finished reading the Preface, Table of Contents and the 39 abstracts of your new book proposal. My first impression is that you have gathered an impressive cohort of researchers working in exited-state hydrogen bonding and proton transfer and covering a variety of applications and approaches. I could not notice any obvious omission. There is a proper mix of experimental and theoretical abstracts. The book, when it comes out, will be an excellent source of information gathered in one place for all researchers working in this very important topic of hydrogen bonding and hydrogen transfer in excited states. Congratulations!
Reviewer Comments
xxxi
Professor Peter Hamm Institute of Chemical Physics, University of Zurich, Winterthurerstr 190, CH-8057 Zurich, Switzerland This is a very impressive collection of articles on hydrogen bonding and hydrogen transfer in excited states. The book is impressive in terms of the number of contributions that have been collected, and in particular in terms of the breadth of themes and molecular systems. Hydrogen bonding and hydrogen transfer (or proton transfer), per se, play a pivotal role in many chemical and in particular biological processes. Studying them in excited states adds a twist to the story; for example, it opens up an opportunity to study these processes in a time-resolved manner. For sure this will become an important collection for people working in the field. Professor Laurence Noirez CNRS Research Director, Laboratoire Leon Brillouin (CEA-CNRS), CE-Saclay 91400 Gif-sur-Yvette Cedex, France Finally – a book that gathers and sums up the current state of knowledge on hydrogen bonding and its role in various media. This fascinating and rapidly changing field expands considerably from year to year. This book meets the difficult challenge of upgrading our knowledge of hydrogen bonding. The contents and the abstracts show that this new book covers an impressively broad domain of studies describing the dynamic properties of hydrogen bonding, from the nano to macroscopic length scales, gathering investigations by experts in domains as different as rheology, nonlinear IR spectroscopy, photoexcitation and acoustic and densitometry studies. This book offers a very complete approach to the state-of-the-art in H-bonding research (39 chapters). Such a contribution has become essential because of the necessary specialization of the current research. It is an invitation to open our minds to complementary methods, to multiscale properties, making it possible to establish a synthetic view so essential and yet so often missing. Professor Igor Pugliesi Univ Munich, Lehrstuhl BioMol Opt, D-80538 Munich, Germany I find that the contributions are interesting and cover many of the areas and topics of excited-state hydrogen bonding and hydrogen transfer. I also find it good that you have given a voice to up-and-coming researchers in Asia. However, I think the book would benefit from contributions by authors who are well known in this field, to create a balance between new and established, between east and west.
List of Contributors
Stefania Abbruzzetti, Dipartimento di Fisica, Universita` degli Studi di Parma, Parma and Dipartimento di Biotecnologie, Universita di Verona, Verona, Italy Aniruddha Adhikari, Department of Physical Chemistry, Indian Association for the Cultivation of Science, Jadavpur, Kolkata 700 032, India Moazzam Ali, Department of Chemistry, University of North Bengal, Darjeeling 734 013, India Alessandra Andreoni, University of Insubria, Department of Physics and Mathematics and C.N.I.S.M.C.N.R., U.d.R. Como Via Valleggio 11, 22100 Como, Italy Manuel Balo´n, Department of Physical Chemistry, Faculty of Pharmacy, University of Seville, Spain Debapriya Banerjee, Unit for Nano Science and Technology, Department of Chemical, Biological and Macromolecular Sciences. S. N. Bose National Centre for Basic Sciences, Block JD, Sector III, Salt Lake, Kolkata 700098, India Giampiero Bartocci, Dipartimento di Chimica, Universita` di Perugia, 06123 Perugia, Italy Samita Basu, Chemical Sciences Division, Saha Institute of Nuclear Physics, 1/AF, Bidhannagar, Kolkata 700 064, India Kankan Bhattacharyya, Department of Physical Chemistry, Indian Association for the Cultivation of Science, Jadavpur, Kolkata 700 032, India Udo Buck, Max-Planck-Institut fu¨r Dynamik und Selbstorganisation, Bunsenstrasse 10, D-37073 Go¨ttingen, Germany Carmen Carmona, Department of Physical Chemistry, Faculty of Pharmacy, University of Seville, Spain Weihua Chen, Department of Chemistry and Maurice Morton Institute of Polymer Science, The University of Akron, Akron, OH 44325, USA Pi-Tai Chou, Department of Chemistry, National Taiwan University, Taipei 106, Taiwan, R.O.C. Attila Demeter, Institute of Materials and Environmental Chemistry, Chemical Research Centre, Hungarian Academy of Sciences, 1025 Budapest, Pusztaszeri u. 59-67, Hungary Debarati Dey, Department of Chemistry and Environment, Heritage Institute of Technology, Chowbaga Road, Anandapur, P.O. East Kolkata Township, Kolkata 700 107, India
xxxiv
List of Contributors
Shantanu Dey, Department of Physical Chemistry, Indian Association for the Cultivation of Science, Jadavpur, Kolkata 700 032, India Cristina Dı´az-Oliva, Departamento de Quı´mica-Fı´sica Aplicada, Facultad de Ciencias, Universidad Auto´noma de Madrid, Cantoblanco, 28049 Madrid, Spain Takayuki Ebata, Department of Chemistry, Graduate School of Science, Hiroshima University, HigashiHiroshima 739-8526, Japan Michal Fa´rnı´k, J. Heyrovsky´ Institute of Physical Chemistry, Academy of Sciences, Dolejsˇkova 3, 182 23 Prague 8, Czech Republic Emilio Garcı´a-Ferna´ndez, Department of Physical Chemistry, Faculty of Pharmacy, University of Seville, Spain Subhadip Ghosh, Department of Physical Chemistry, Indian Association for the Cultivation of Science, Jadavpur, Kolkata 700 032, India Mark S. Gordon, Department of Chemistry, Iowa State University, Ames, Iowa 50011, USA Marie Louise Groot, Department of Physics and Astronomy, Faculty of Sciences, Vrije Universiteit, De Boelelaan 1081, 1081 HV Amsterdam, The Netherlands Ke-Li Han, State Key Laboratory of Molecular Reaction Dynamics, Dalian Institute of Chemical Physics, Chinese Academy of Sciences, 457 Zhongshan Road, Dalian 116023, China Derren James Heyes, Manchester Interdisciplinary Biocentre, University of Manchester, 131 Princess Street, Manchester M1 7DN, UK Jose Hidalgo, Department of Physical Chemistry, Faculty of Pharmacy, University of Seville, Spain Mihaela Hillebrand, Department of Physical Chemistry, Faculty of Chemistry, University of Bucharest, Bd. Regina Elisabeta 4-12, Bucharest, Romania Wouter D. Hoff, Department of Microbiology and Molecular Genetics, Oklahoma State University, Stillwater, OK 74078, USA Cheng-Chih Hsieh, Department of Chemistry, National Taiwan University, Taipei 106, Taiwan, R.O.C. Sorana Ionescu, Department of Physical Chemistry, Faculty of Chemistry, University of Bucharest, Bd. Regina Elisabeta 4-12, Bucharest, Romania Chang-Ming Jiang, Department of Chemistry, National Taiwan University, Taipei 106, Taiwan, R.O.C. Hideki Kandori, Department of Frontier Materials, Nagoya Institute of Technology, Showa-ku, Nagoya 466-8555, Japan Zhouyang Kang, Department of Physics, Oklahoma State University, Stillwater, Oklahoma 74078, USA Johannes Kiefer, Lehrstuhl fu¨r Technische Thermodynamik (LTT) and Erlangen Graduate School in Advanced Optical Technologies (SAOT), Universita¨t Erlangen-Nu¨rnberg, Am Weichselgarten 8, 91058, Erlangen, Germany. Present address: School of Engineering, University of Aberdeen, Fraser Noble Building, King’s College, Aberdeen AB24 3UE, Scotland, UK Kwang S. Kim, Centre for Superfunctional Materials, Department of Chemistry, Pohang University of Science and Technology, San 31, Hyojadong, Namgu, Pohang 790-784, Korea
List of Contributors
xxxv
Daisuke Kina, Division of Chemistry, Graduate School of Science, Hokkaido University, Sapporo 060-0810, Japan Maciej Kolaski, Centre for Superfunctional Materials, Department of Chemistry, Pohang University of Science and Technology, San 31, Hyojadong, Namgu, Pohang 790-784, Korea, and Department of Theoretical Chemistry, Institute of Chemistry, University of Silesia, 9 Szkolna Street, 40-006 Katowice, Poland Govindarajan Krishnamoorthy, Department of Chemistry, Indian Institute of Technology Guwahati, Guwahati 781039, India Oliver Ku¨hn, Institut fu¨r Physik, Universita¨t Rostock, D-18051 Rostock, Germany Anupriya Kumar, Centre for Superfunctional Materials, Department of Chemistry, Pohang University of Science and Technology, San 31, Hyojadong, Namgu, Pohang 790-784, Korea Masato Kumauchi, Department of Microbiology and Molecular Genetics, Oklahoma State University, Stillwater, OK 74078, USA Manoj Kumbhakar, Radiation and Photochemistry Division, Bhabha Atomic Research Centre, Trombay, Mumbai 400085, India Han Myoung Lee, Centre for Superfunctional Materials, Department of Chemistry, Pohang University of Science and Technology, San 31, Hyojadong, Namgu, Pohang 790-784, Korea Jerzy Leszczynski, NSF CREST Interdisciplinary Nanotoxicity Centre, Department of Chemistry and Biochemistry, Jackson State University, Jackson, MS 39217, USA Samuel Leutwyler, Department fu¨r Chemie und Biochemie, Universita¨t Bern, Freiestrasse 3, 3012 Bern, Switzerland Carine Tanner Manca, Laboratorium fu¨r Physikalische Chemie, ETH Zu¨rich, CH-8093 Zu¨rich, Switzerland Ujjwal Mandal, Department of Physical Chemistry, Indian Association for the Cultivation of Science, Jadavpur, Kolkata 700 032, India Iulia Matei, Department of Physical Chemistry, Faculty of Chemistry, University of Bucharest, Bd. Regina Elisabeta 4-12, Bucharest, Romania Ugo Mazzucato, Dipartimento di Chimica, Universita` di Perugia, 06123 Perugia, Italy Ajay Misra, Department of Chemistry and Chemical Technology, Vidyasagar University, Midnapore 721 102, WB, India Rajib Kumar Mitra, Unit for Nano Science and Technology, Department of Chemical, Biological and Macromolecular Sciences, S. N. Bose National Centre for Basic Sciences, Block JD, Sector III, Salt Lake, Kolkata 700098, India Sivaprasad Mitra, Department of Chemistry, North-Eastern Hill University, Permanent Campus, Shillong 793022, India Marı´a Asuncio´n Mun˜oz, Department of Physical Chemistry, Faculty of Pharmacy, University of Seville, Spain Akira Nakayama, Division of Chemistry, Graduate School of Science, Hokkaido University, Sapporo 060-0810, Japan
xxxvi
List of Contributors
Luca Nardo, University of Insubria, Department of Physics and Mathematics and C.N.I.S.M.-C.N.R., U.d.R. Como Via Valleggio 11, 22100 Como, Italy Sukhendu Nath, Radiation and Photochemistry Division, Bhabha Atomic Research Centre, Trombay, Mumbai 400085, India Peter Nikolov, Institute of Organic Chemistry with Centre of Phytochemistry, Bulgarian Academy of Sciences, 1113 Sofia, Bulgaria Takeshi Noro, Division of Chemistry, Graduate School of Science, Hokkaido University, Sapporo 060-0810, Japan Taka-aki Okamura, Department of Macromolecular Science, Graduate School of Science, Osaka University, Toyonaka, Osaka 560-0043, Japan Seth Olsen, The University of Queensland, Australian Institute for Bioengineering and Nanotechnology, Centre for Computational Molecular Science, QLD 4072, Brisbane, Australia Haridas Pal, Radiation and Photochemistry Division, Bhabha Atomic Research Centre, Trombay, Mumbai 400085, India Samir Kumar Pal, Unit for Nano Science and Technology, Department of Chemical, Biological and Macromolecular Sciences, S. N. Bose National Centre for Basic Sciences, Block JD, Sector III, Salt Lake, Kolkata 700098, India Dipak K. Palit, Radiation and Photochemistry Division, Bhabha Atomic Research Centre, Mumbai 400 085, India Yi Pang, Department of Chemistry and Maurice Morton Institute of Polymer Science, The University of Akron, Akron, OH 44325, USA David Lee Phillips, Department of Chemistry, University of Hong Kong, Pokfulam Road, Hong Kong Rajib Pramanik, Department of Chemistry, Indian Institute of Technology, Kharagpur, PIN-721302, WB, India Sankar Prasad De, Department of Chemistry and Chemical Technology, Vidyasagar University, Midnapore 721 102, WB, India Gernot Renger, Max-Volmer-Laboratorium fu¨r Biophysikalische Chemie, Technische Universita¨t Berlin, Strasse des 17. Juni 135, 10623 Berlin, Germany Dolores Reyman, Departamento de Quı´mica-Fı´sica Aplicada, Facultad de Ciencias, Universidad Auto´noma de Madrid, Cantoblanco, 28049 Madrid, Spain Swapan K. Saha, Department of Chemistry, University of North Bengal, Darjeeling 734 013, India Antonio Sa´nchez-Coronilla, Department of Physical Chemistry, Faculty of Pharmacy, University of Seville, Spain Manas Kumar Sarangi, Chemical Sciences Division, Saha Institute of Nuclear Physics, 1/AF, Bidhannagar, Kolkata 700 064, India Nilmoni Sarkar, Department of Chemistry, Indian Institute of Technology, Kharagpur, PIN-721302, WB, India
List of Contributors
xxxvii
Souravi Sarkar, Department of Chemistry, Indian Institute of Technology, Kharagpur, PIN-721302, WB, India Manoj K. Shukla, NSF CREST Interdisciplinary Nanotoxicity Centre, Department of Chemistry and Biochemistry, Jackson State University, Jackson, MS 39217, USA Ashutosh S. Singh, Institute of Chemistry, Academia Sinica, Taipei 115, Taiwan, ROC Petr Slavı´cˇek, Department of Physical Chemistry, Institute of Chemical Technology, Technicka´ 5, Prague 6, Czech Republic Sean C. Smith, The University of Queensland, Australian Institute for Bioengineering and Nanotechnology, Centre for Computational Molecular Science, QLD 4072, Brisbane, Australia Anna Spalletti, Dipartimento di Chimica, Universita` di Perugia, 06123 Perugia, Italy Shih-Sheng Sun, Institute of Chemistry, Academia Sinica, Taipei 115, Taiwan, ROC Qiao Sun, The University of Queensland, Australian Institute for Bioengineering and Nanotechnology, Centre for Computational Molecular Science, QLD 4072, Brisbane, Australia Tetsuya Taketsugu, Division of Chemistry, Graduate School of Science, Hokkaido University, Sapporo 0600810, Japan, and Department of Theoretical and Computational Molecular Science, Institute for Molecular Science, Okazaki 444-8585, Japan Christian Tanner, TOFWERK AG, Uttigenstrasse 22, CH-3600 Thun, Switzerland Ilijana Timcheva, Institute of Organic Chemistry with Centre of Phytochemistry, Bulgarian Academy of Sciences, 1113 Sofia, Bulgaria Vladimir I. Tomin, Institute of Physics, Pomeranian University, 76-200, Słupsk, Poland Hanne Hjorth Tønnesen, School of Pharmacy, University of Oslo, PO Box 1068 Blindern, 0316 Oslo, Norway Norikazu Ueyama, Department of Macromolecular Science, Graduate School of Science, Osaka University, Toyonaka, Osaka 560-0043, Japan Pramod Kumar Verma, Unit for Nano Science and Technology, Department of Chemical, Biological and Macromolecular Sciences, S. N. Bose National Centre for Basic Sciences, Block JD, Sector III, Salt Lake, Kolkata 700098, India Cristiano Viappiani, Dipartimento di Fisica, Universita` degli Studi di Parma and NEST Istituto NanoscienzeCNR, Parma, Italy Brian D. Wagner, Department of Chemistry, University of Prince Edward Island, Charlottetown, PE, C1A 1Z5 Canada Sufan Wang, The University of Queensland, Australian Institute for Bioengineering and Nanotechnology, Centre for Computational Molecular Science, QLD 4072, Brisbane, Australia Aihua Xie, Department of Physics, Oklahoma State University, Stillwater, OK 74078, USA Hitoshi Yamamoto, Department of Macromolecular Science, Graduate School of Science, Osaka University, Toyonaka, Osaka 560-0043, Japan
xxxviii
List of Contributors
Yun-an Yan, Institut fu¨r Physik, Universita¨t Rostock, D-18051 Rostock, Germany Hong Zhang, The University of Queensland, Australian Institute for Bioengineering and Nanotechnology, Centre for Computational Molecular Science, QLD 4072, Brisbane, Australia Guang-Jiu Zhao, State Key Laboratory of Molecular Reaction Dynamics, Dalian Institute of Chemical Physics, Chinese Academy of Sciences, 457 Zhongshan Road, Dalian 116023, China
Preface
Hydrogen bonding is of universal importance in chemistry, biology and physics. Hydrogen bonding is central to the understanding of the microscopic structures and functions in many complex systems, for example hydrogen-bonded water or alcohol networks, organic compounds in solution, crystal engineering, selfassembled supramolecular architectures, proteins and DNA building blocks of life. Moreover, hydrogenbonding structures and dynamics in the excited states also play important roles in determining many chemical, physical and biochemical processes. In general, fluorescence emission behaviours of organic and biological chromophores can be significantly influenced by the intermolecular hydrogen-bonding interactions between chromophores and protic solvents or biological surroundings. Furthermore, ultrafast deactivation processes of photoexcited molecular and supramolecular systems can be remarkably facilitated by excited-state hydrogen bonding. In particular, excited-state hydrogen transfer is closely related to excited-state hydrogen-bonding structures and dynamics. Hydrogen bonding in the ground state for various types of molecular and supramolecular systems has been described systematically in an abundance of published monographs. However, to the best of our knowledge, there are no monographs on hydrogen bonding in excited states until now. It has been found that excited-state hydrogen bonding and hydrogen transfer are playing an increasingly important role in many photophysical processes and photochemical reactions. Therefore, a new book presenting hydrogen bonding and hydrogen transfer in excited states is urgently needed. This scientific book will be very helpful in extensively understanding the nature of hydrogen bonding and its key roles in photochemistry, photobiology and photophysics. This book gives an extensive description of research progress on excited-state hydrogen bonding and hydrogen transfer in recent years. First of all, both experimental and theoretical investigations on excited-state hydrogen-bonding structures and dynamics of many organic and biological chromophores are presented: coumarin and its derivates, fluorenone and its derivates, diazines, quinones, b-carbolines, harmane derivatives, substituted phthalimides, 4-aminoindandiones, hydroxyphenacyl compounds, diazaromatic betacarboline derivatives, imidazolium ionic liquids, supramolecular host–guest complexes, nanoconfined systems, rhodopsins, hydrated DNA bases and base pairs, and so on. After that, several chapters will describe the influences of excited-state hydrogen bonding on various photophysical processes and photochemical reactions. For example, the effects of hydrogen bonding on fluorescence emission behaviours and photoisomerization; intramolecular-hydrogen-bond-formation-mediated de-excitation of curcuminoids; the role of hydrogen bonding in photosynthetic water splitting; ultrafast catalytic processes in the light-driven enzyme protochlorophyllide oxidoreductase (POR); photoinduced electron transfer and solvation dynamics in roomtemperature ionic liquids; hydrogen-bonding barrier crossing dynamics at biomimicking surfaces; the effects of hydrogen bonding on intramolecular charge transfer, vibrational energy relaxation and the ICT-to-TICT conversion; hydrogen bond basicity in excited states and dynamic heterogeneity. Finally, in the final chapters
xxiv
Preface
we will focus our attention on excited-state hydrogen transfer. Some experimental and theoretical studies on excited-state hydrogen transfer in some isomeric naphthalene derivatives, benzoxazole derivatives, pyridoindole and pyrrolo-quinoline derivatives, green and red fluorescent proteins, hydrated halides, o-hydroxy carbonyl compounds, hydroxyl aromatic dopants and photoactive yellow protein will be presented. Moreover, investigations on controlling excited-state hydrogen transfer along hydrogen-bonded wires, excited-state double-proton transfer, ab initio QM/MM excited-state molecular dynamics, the reaction volume for photoinduced proton transfer in aqueous solutions of 6-methoxyquinoline, conformational switching between acids and their anions by hydrogen bonding, molecular recognition and chemical sensing of anions utilizing excited-state hydrogen bonding, photodissociation of hydrogen-bonded clusters and proton transfer reactions for dynamic quenching of fluorescence are also reported. The readers of this book will be faculties and researchers in the fields of molecular and supramolecular photochemistry, photobiology and photophysics. It will also serve as a good reference book for graduate students for the study of recent topics and progress in excited-state hydrogen bonding and hydrogen transfer. Finally, we would like to express our sincere thanks to all the authors who have contributed with their excellent chapters to the realization of this monograph. We greatly acknowledge the financial support by NSFC (Nos 20833008 and 20903094) and NKBRSF (Nos 2007CB815202 and 2009CB220010) and the assistance of co-workers from the group in the Dalian Institute of Chemical Physics (DICP) in the editorial process. Also, we thank the team at John Wiley and Sons Ltd, in particular Mr Paul Deards and Mr Mingxin Hou, for their helpful guidance during the entire project. Ke-Li Han Guang-Jiu Zhao State Key Laboratory of Molecular Reaction Dynamics, Dalian Institute of Chemical Physics, Chinese Academy of Sciences, 457 Zhongshan Road, Dalian 116023, China November 2010
Plate 1
The power spectrum of the velocity ACF for the atoms in the QM region of the A–U pair
Plate 2 A sequence of 2D IR photon echo spectra given by the real part of equation (1.8) of a 0.1 M solution of A–U in CDCl3 in the NH stretching region for different population times T. Each panel has been separately normalized to the maximum value of the signal
15
0.8
10
0.6
5 0.75
0.80
0.85
0.90
0.95
(1-A /A)/[HFIP]
absorbance / arbitrary
1.0
1.00
A /A
0.4
0.2
0.0 15000
20000
25000
30000
35000
wavenumber / cm-1
Plate 3 Absorption (full line) and fluorescence spectra (dotted line) of I (black line) and of the complex of I with HFIP (red line) in n-hexane. (At 25 000 cm1 the increasing absorbance corresponds to 0, 0.0027, 0.0054, 0.011, 0.019, 0.027, 0.040, 0.053 and 0.080 mol dm3 [HFIP], and the derived complex species spectrum.) Inset: linearized plot (in accordance with equation (3.3)) of the 400 nm absorbance results. Reprinted with permission from [6]. Copyright 2005 American Chemical Society
Hydrogen Bonding and Transfer in the Excited State, Volume I & II © 2011 John Wiley & Sons, Ltd. ISBN: 978-0-470-66677-7
Edited by Ke-Li Han and Guang-Jiu Zhao
0.8
3
absorbance
0.6
0.4
h
i
j
o n m l k
g f e
0.2
[HFIP] / mol dm (a) 0.0 (b) 2.2E-4 (c) 4.4E-4 (d) 0.0011 (e) 0.0022 (f) 0.0033 (g) 0.0054 (h) 0.01 (i) 0.016 (j) 0.023 (k) 0.044 (l) 0.066 (m) 0.087 (n) 0.108 (o) 0.129
d
0.0 300
c b a
310
320
330
340
350
360
370
380
390
400
wavelength / nm
Plate 4 Absorption spectra of DMPN (8 105 mol dm3) with various [HFIP] in n-hexane. Reprinted with permission from [8]. Copyright 2004 American Chemical Society
absorbance
2.0
1.5
N
2.0
absorbance
N
1.5 1.0 0.5 0.0 200
220
240
260
280
300
wavelength / nm 1.0
0.5
0.0 200
220
240
260
280
300
wavelength / nm
Plate 5 Absorption spectra of DMAP (6.14 105 mol dm3) with and without HFIP additive in n-hexane. (At 280 nm, the increasing absorbance corresponds to 0, 0.00006, 0.00013, 0.00026, 0.00052, 0.0009, 0.0013, 0.0019, 0.0033, 0.0059, 0.011, 0.018, 0.028, 0.039, 0.050, 0.062, 0.076, 0.090, 0.105, 0.121, 0.137 and 0.154 mol dm3 HFIP concentrations.) Inset: the resolved absorption spectra of the uncomplexed (black line), singly (red line) and doubly (blue line) complexed species. Reprinted with permission from [17]. Copyright 2007 American Chemical Society
2.0
A absorbance
3.5 3.0 2.5
1.5
N
1.0 0.5
2.0
0.0 200
220
240
260
280
300
320
wavelength / nm
1.5 1.0
0.0 3.5 3.0 2.5
1.5
B
N absorbance
absorbance
0.5
1.0
0.5
2.0 0.0 200
1.5
220
240
260
280
300
320
wavelength / nm
1.0 0.5 0.0 200
220
240
260
280
300
320
wavelength / nm
Plate 6 Absorption spectra of pyridine (4.3 104 mol dm3) (A) and N,N-dimethyl aniline (9.3 105 mol dm3) (B) with and without HFIP additive in n-hexane. (The increasing absorbance corresponds to 0, 0.0034, 0.0069, 0.010, 0.017, 0.025, 0.036, 0.052, 0.069, 0.104, 0.175, 0.25, 0.36 and 0.54 mol dm3 [HFIP] at 250 nm (A), and to 0, 0.0033, 0.0065, 0.010, 0.016, 0.023, 0.033, 0.043, 0.052, 0.066, 0.082, 0.099, 0.13, 0.17, 0.20 and 0.27 mol dm3 [HFIP] at 275 nm (B).) Insets: the resolved absorption spectra of the uncomplexed (black line), singly (red line) and doubly (blue line) complexed species. Reprinted with permission from [17]. Copyright 2007 American Chemical Society
Plate 7 The calculated IR spectra of fluorenone in both the S0 and T1 states. The ground-state IR spectrum of the hydrogen-bonded FNMeOH complex is also shown for comparison
Plate 8 Calculated IR spectra of the hydrogen-bonded FNMeOH complex in both the S0 and T1 states
O. D.
fl. intensity (a. u.)
2.0x106
1.5x106
1.0x106
wavelength (nm)
5.0x105
0.0 390
420
450
480
510
540
wavelength (nm)
Plate 9 Steady-state fluorescence spectra of DBPZ (1 105 M) (lex ¼ 350 nm) in cyclohexane (—) and in MeCN (----). The inset shows absorption spectra of DBPZ (1 105 M) in MeCN (black), 63% (v/v) water–MeCN mixture (red), 2% MeCN–water mixture (blue) and MeCN–HClO4 mixture (green). (The inset is Reprinted with permission from [70]. Copyright 2007 American Chemical Society)
Plate 10 Geometrically optimized structures of DBPZ–1H2O (Ort 1) and two possible structures of DBPZ–2H2O (Ort 2 and Ort 3) systems. Colour code: red ¼ oxygen; blue ¼ nitrogen; yellow ¼ carbon; white ¼ hydrogen. Reprinted with permission from [70]. Copyright 2007 American Chemical Society
1599
1357
1481
#
1167
363
808
(b) 400ns 200ns 100ns 80ns 60ns 40ns
Intensity (arb. units)
20ns
10ns
(a) 6000ps 3000ps 1000ps 500ps 100ps 50ps 10ps 5ps 2ps 0ps
400
600
800
1000 1200 1400 -1 Raman Shift (cm )
1600
1800
Plate 11 Picosecond-TR3 (a) and nanosecond-TR3 (b) spectra of HA in water solution obtained at various time delays with 267 nm pump and 400 nm probe wavelengths for the picosecond spectra and 266 nm pump and 416 nm probe wavelengths for the nanosecond spectra. The band labelled by # is due to a stray laser line. Reprinted with permission from [26]. Copyright 2005 American Chemical Society
(b)
# 500ns 300ns 200ns 100ns 50ns
Intensity (arb. units)
20ns 10ns
(a)
6000ps 3000ps 1000ps 500ps 100ps 50ps 10ps 5ps 2ps 0ps
400
600
800
1000
1200
1400
1600
1800
-1
Raman Shift (cm )
Plate 12 Picosecond-TR3 (a) and nanosecond-TR3 (b) spectra of HA in buffered water solution with pH ¼ 9.0, obtained at various time delays with 267 nm pump and 400 nm probe excitation wavelengths for the picosecond spectra and 266 nm pump and 416 nm probe wavelengths for the nanosecond spectra. The band labelled by # is due to a stray laser line. Reprinted with permission from [26]. Copyright 2005 American Chemical Society
1527 1567 1617
1074
843
705
367
700ns 500ns
Intensity (arb. units)
300ns
150ns 100ns 60ns 40ns 20ns 10ns 0ns
400
600
800 1000 1200 1400 1600 1800 -1
Raman Shift (cm )
Plate 13 Nanosecond-TR3 spectra of HA in water solution obtained at various time delays with 267 nm pump and 341.5 nm probe wavelengths. Features in blue are due to the neutral HA triplet state. See the text for details of the attributions of other features. Reprinted with permission from [26]. Copyright 2005 American Chemical Society
1527 1302 1322
1168
843 367
1617
1074
705
1567
(a)
Intensity (arb. units)
(b)
(c)
400
600
800
1000 1200 1400 -1 Raman Shift (cm )
1600
1800
Plate 14 Comparison of the ns-TR3 spectrum (a) obtained at a 100 ns time delay in Figure 13.7 with the 341.5 nm resonance Raman spectrum of the ground-state HA anion (b) obtained in water/NaOH (0.1 M) solution. B3LYP/6311G(d, p) DFT-calculated normal Raman spectrum of the HA anion ground state (c) is also shown to compare and help with the assignment of the experimental spectra. Features in black shown in spectrum (a) are due to the unidentified species ‘X’ (see the text for details). Reprinted with permission from [26]. Copyright 2005 American Chemical Society
Intensity (a.u.)
4
Time Delay (ps)
(b)
100 70 49 35 24 17 12 8 6 4 3
(a)
3
2
1
0
(c)
*
#
Intensity (a.u.)
(d)
*
#
(e)
* #
*
(f)
# 200
250
300
350
400
450
500
550
Wavelength (nm)
Plate 15 (a) Femtosecond-KTRF contour of HPA obtained with 267 nm excitation in MeCN. (b) Normalized fluorescence decay at 330 nm (circles in blue) and 440 nm (circles in red) for HPA in MeCN. The solid lines show two exponential fittings to the experimental data; the dotted line is the instrumental response function (see Ref. [24] for details). (c) to (f) steady-state absorption spectrum (in black) and typical fluorescence profile of the blue (in blue) and red (in red) fluorescence for HPA in MeCN (c), THF (d), MeOH (e) and 90% H2O/10% MeCN mixed solvent (f). The absorption spectra were normalized to the blue fluorescence spectra. Sharp features indicated by # and are the solvent Raman band and the second harmonic generation of the 800 nm gating pulse from the Kerr medium respectively. Reprinted with permission from [24]. Copyright 2005 American Chemical Society
Viscosity (cP)
70 (a) RTIL-acetonitrile mixture RTIL-methanol mixture 60 50 40 30 20 0.10
0.15
0.20
0.25
0.30
0.35
0.40
τ avs(ns)
80
RTIL-acetonitrile mixture R=0.99171 RTIL-methanol mixture R=0.95902
70
(b)
Viscosity(cP)
60 50 40 30 20 1.0
1.5
2.0
2.5
3.0
3.5
τrot(ns)
Plate 16 The plot of (a) average solvation time versus bulk viscosity of RTIL þ cosolvents mixtures and (b) average rotational relaxation time versus bulk viscosity of RTIL þ cosolvents mixtures. Reprinted with permission from the American Chemical Society. Copyright 2009 0
10
420 nm 532 nm -1
Intensity
10
-2
10
-3
10
0
500
1000
time (ps)
Plate 17
Instrumental pulse response at 420 nm (violet) and 532 nm (green)
10000 CURC in Ethyl Acetate: 1000
Experimental decay pattern Single exponential fit
Counts
100
Residuals
10
1 200 100 0 -100 -200
0
5000
10000
15000
20000
Plate 18 Upper panel: experimental fluorescence decay pattern of CURC in ethyl acetate (black dots) and single exponential fitting curve (red line; t ¼ 468 ps). Lower panel: residual plot
Plate 19 The K minus BR (A), L minus BR (B), M minus BR (C) and N minus BR (D) difference infrared spectra in the 2750–1930 cm1 region. The spectra are compared between hydration with D2O (red lines) and D218O (blue lines). The K, L, M and N intermediates were produced by illuminating BR films at 77, 170, 230 and 270 K respectively. The grey curve in the 2700–2000 cm1 region represents OD stretching vibrations of D2O. Greenlabelled frequencies correspond to those identified as OD stretching vibrations of water. Purple-coloured tags represent OD stretch of Thr89 [22, 23], while the underlined tags represent ND stretch of the Schiff base [24, 53]. Reproduced by permission of The Royal Society of Chemistry (RSC) for the European Society for Photobiology, the European Photochemistry Association, and the RSC
L-BR
SIMILAR!!
artifactual cytoplasmic water cavity at 170 K
3700
3600
No water cluster deprotonation at 230K
293K 230K strongly weakly H-bonded H-bonded water water
293K 170K 3800
M-BR
3500 -1
wavenumber / cm
3400
3800
3600
2800
2600 -1
wavenumber / cm
293K 230K 2200
2000
1800 -1
wavenumber / cm
Plate 20 Spectral comparison of water signals in the L minus BR (left) and M minus BR (middle, right) difference FTIR spectra between room temperature (red lines) and low temperature (blue lines). Modified with permission from [25]. Copyright 2008 American Chemical Society
Plate 21 (A) The K minus D85S difference infrared spectra of the absent halide (a) and containing Cl (b), Br (c) or I (d) bound form in the 2700–2000 cm1 region. The sample was hydrated with D2O (red curves) or D218O (blue curves), and spectra were measured at 130 K. Green-labeled frequencies correspond to those identified as OD stretching vibrations of water. Modified with permission from [37]. Copyright 2006 American Chemical Society. (B) The K minus BR difference infrared spectra of the wild-type (a), halide-free (b), Cl-bound (c), Brbound (d) and I-bound (e) D212N in the 2380–2020 cm1 region. The samples were hydrated with D2O (red curves) or D218O (blue curves), and spectra were measured at 77 K. Green-labeled frequencies correspond to those identified as OD stretching vibrations of water. Underlined frequency (2171 cm1) in the wild type also contains the ND stretch of the Schiff base [24]. Modified with permission from [39]. Copyright 2007 American Chemical Society
Rhodopsins Having strongly H-bonded water D85N and D212N BR
Having proton-pump activity Bacteriorhodopsin (BR) various BR mutants including D212N(Cl)
D85S(Cl) BR 13-cis,15-syn BR Halorhodopsin (HR)
Azide-bound Halorhodopsin Anabaena Sensory Rhodopsin (ASR)
Salinibacter Sensory Rhodopsin I Sensory Rhodopsin II (SRII) without HtrII
Sensory Rhodopsin II (SRII) with HtrII
Neurospora Rhodopsin (NR)
Proteorhodopsin (PR) Leptosphaeria Rhodopsin (LR)
Visual Rhodopsins
Plate 22 Various rhodopsins are classified in view of (i) proton-pump activity and (ii) whether they have strongly hydrogen-bonded water molecules (OD stretch in D2O at 1 corresponds to solvents such as dimethylsulfoxide (DMSO) and water, able to promote such stabilization. Complementary to the p scale are the a and b scales, describing the Br€ onsted acid/base character of the solvent respectively. Thus, a shows the ability of the solvent to donate a proton in a solvent-to-solute H-bond (HBD ability), while b shows the solvent ability to accept a proton in a solute-to-solvent H-bond (HBA) [49]. The p, a and b coefficients are interpreted as solute properties and measure the effect of each process on XYZ. XYZ0 is the spectral observable independent of solvent effects (taken in the reference solvent, CHX, for which p , a and b are 0). d is a polarizability correction term for aromatic (d ¼ 1.0) and chlorinated aliphatic (d ¼ 0.5) solvents [50]. The validity of the Kamlet–Taft approach can be tested by plotting the XYZ values estimated by the model versus the XYZ values determined experimentally. Good correlations are generally observed, attesting that the model can indeed account for the experimental trends. Within the framework of the Kamlet–Taft model, protic solvents can provide the proton to form an H-bond with the HBA functions in the solute, but can also accept a proton to form an H-bond with HBD solute moieties. The greater the a and/or b values, the stronger are the H-bond(s) formed. As regards aprotic solvents, they can be cast into three categories: HBAs like CCl4, dioxane, ethyl acetate, acetone, tetrahydrofurane (THF), N,Ndimethylformamide (DMF) and DMSO (low a, high b values); HBDs like dichloromethane and chloroform (low b, high a values); solvents like acetonitrile (ACN), characterized by high a and b values. A collection of a, b and p values for different solvents as well as for some solvent mixtures can be found in [47]. Another advantage of the Kamlet–Taft model worth mentioning is that, as opposed to the e(«, n) and ET(30) parameters described above, the p , a and b parameters are obtained by averaging solvent effects on a variety of indicators rather than on a single probe [50], thus being much more reliable. The p scale was first based on the solvent-induced shifts of the longest wavelength p–p absorption band of seven nitroaromatic indicators, but was then improved by multiple least-squares correlations with more solvatochromic probes [39]. A similar model was proposed by Catalan, who developed his solvent polarity (SPP) [16], solvent acidity (SA) [51] and solvent basicity (SB) [52] scales on the basis of the solvatochromic behaviour of fluorophores and their homomorphs, i.e. compounds possessing similar but not identical structure to the fluorophores, thus allowing cancellation of many of the spurious effects involved in measurements of solvent properties. The SPP scale was obtained from the solvatochromic shifts undergone by the longest-wavelength absorption maximum of two indicators: 2-(dimethylamino)-7-nitrofluorene (DMANF) and its homomorph, 2-fluoro-7-nitrofluorene (FNF).
86
Hydrogen Bonding and Transfer in the Excited State
Similarly, the SB and SA scales are based on the probe–homomorph couples 5-nitroindoline and 1-methyl-5nitroindoline, and o-tert-butylstilbazolium (TBSB) and o,o0 -di-tert-butylstilbazolium (DTBSB) betaine dyes respectively. The TBSB/DTBSB pair is unsuitable for evaluation of solvents more acidic than methanol (MeOH) owing to the protonation of the indicator, and therefore 3,6-diethyltetrazine is employed instead. The dependence of XYZ on the SPP, SA and SB parameters is given by XYZ ¼ XYZ0 þ pSPP þ aSA þ bSB
ð4:6Þ
More recently, Catalan parameters for mixed solvents became available as well [53]. Mancini et al. determined the microscopic solvent properties – dipolarity/polarizability SPP, basicity SB and acidity SA – for binary mixtures of ethyl acetate with chloroform, ACN or MeOH as cosolvent, where specific intersolvent interactions by H-bonds are involved [54]. The results were then correlated to the Kamlet–Taft dipolarity/ polarizability p, acidity a and basicity b scales in order to evaluate the concordance between the two models. The good correlation showed that the Catalan scale is also appropriate for interpreting the effects of the various processes occurring in mixed solvents. These models provide an insight into the solute–solvent interactions in the ground and first excited singlet states. The magnitudes and signs of the Kamlet–Taft p, a and b parameters are indicative of the relative stabilization or destabilization effect of the respective solvent property on the spectral feature of the solute. The magnitude of the coefficients, sometimes expressed as percentage contributions, allows us to identify the dominant solvent effect in S0 and S1 states. As regards the signs, positive p coefficients for the absorption data mean that increasing the solvent polarity/polarizability produces a blue-shift in na, i.e. a stabilization of the ground state of the solute as compared with the FC state. By contrast, for the fluorescence data, negative p coefficients correlate with the bathochromic shift of nf upon increasing solvent polarity, indicating that the relaxed excited state is more stabilized in polar solvents. The relative magnitudes of a and b coefficients reflect the dominant contribution in H-bond interactions. The literature contains numerous investigations that make use of the Kamlet–Taft and/or Catalan models in order to explain the effects of pure solvents and solvent mixtures on the photophysical properties of molecules and to demonstrate the role of excited-state H-bonds [55–61]. Some examples of studies employing the Kamlet–Taft model are presented in Table 4.2. Aaron et al. [66] correlated the values/signs of the estimated Kamlet–Taft coefficients with the trends of the experimental spectra of a series of fused benzothiophenes. It can be seen that for BTT (Table 4.2), while the solute–solvent interaction in S0 is governed by solvent polarity (p ¼ 67%), in S1 the HBA of the solvent becomes the predominant contribution (b ¼ 79%), indicating the formation of H-bonds. The positive sign for p correlates well with the hypsochromic shift experimentally observed for the absorption spectra in polar media, while the negative p sign for the excited state reflects the bathochromic shift of the fluorescence maximum. A similar correlation was performed for F, which resulted in negative values for the p and b coefficients, showing the increase in F with solvent polarity and HBA ability, as observed experimentally (Fmax ¼ 0.15 was found in DMSO). The F values of 4-hexylresorcinol are strongly influenced by the nature of the solvent: the highest values are recorded in alcohols (0.8), while the smallest one is in CHX (0.05) [44]. A plot of F versus ETN was found to be linear for protic solvents (r ¼ 0.88), and non-linear for aprotic solvents. The Kamlet–Taft analysis suggested that the solvent’s HBA is responsible for the F enhancement. This was explained in terms of an H-bond donated by the hydroxylic hydrogen atom of the dye to the solvent, which becomes stronger upon excitation. The best Kamlet–Taft correlation for F was obtained when only protic solvents were employed (r ¼ 0.99). As can be seen from Table 4.2, DMAC presents solvent-dependent photophysical properties, explained by polar as well as by specific H-bond interactions: 66% polar and 30% HBD contribution in S0; 63% polarity effect and 21% HBA in S1 [68]. Correlations of the absorption and emission energies, Ea and Ef, reinforce the
O3S
-
Dye
Table 4.2
N+
I
NHR
Rose Bengal
O
O
SO3Na
OH
Heptamethine cyanine dye
I
I
COO-Na+
Cl
HO
2-Naphthol
1-Naphthol-5-sulfonate
Na-O
+
I
Cl
Cl
Cl
N
SO3-
na ¼ na,0 70p þ 270a 510b (r ¼ 0.98) nf ¼ nf,0 450p þ 0a 800b
nf ¼ 26 325 þ 0p 1198a þ 471b (r ¼ 0.95)
nf ¼ 12831 þ 410p þ 612a 236b (r ¼ 0.99)
na ¼ 19 262 1458p þ 267a 1012b
Kamlet–Taft LSER
HBA (64%)
HBA (60%)
HBD (72%)
HBD (49%)
Polarity (53%), HBA (37%)
[65]
[64]
[63]
[62]
Ref.
ðcontinuedÞ
Predominant solvent effect (%)
Influence of the solvent polarity and HBD and/or HBA ability on the spectral features of various dyes in S0 and S1 states
Solute–Solvent Hydrogen Bond Formation in the Excited State 87
S
O
O
H
H
N
Benzothieno[3,2b]-thiophene (BTT)
S
OH
OH
CH3
CH3
N
N
N
N
1,4-Bis[ -(2-quinoxalyl)-vinyl]benzene
N
1-(2-Pyridyl)-5-(4-dimethylaminophenyl)penta-2,4-diene-1-one (DMAC)
N
O
3-[4-Di(2-hydroxyethyl) amino]phenyl-l-(2-furyl)-2-propene-1one
Dye
Table 4.2 (Continued)
Fprotic ¼ 1.305 0.927p 0.161a (r ¼ 0.98) Faprotic ¼ 0.334 þ 0.335p þ 0.195a (r ¼ 0.99)
Dna,1/2 ¼ 3799 þ 757p þ 967a (r ¼ 0.88) Ea ¼ 67.97 5.09p 2.34a 0.31b (r ¼ 0.91) Ef ¼ 59.58 12.48p 3.11a 4.16b (r ¼ 0.92)
na ¼ 25071 1315p 679a 715b (r ¼ 0.98) nf ¼ 22988 2750p 2251a 2366b (r ¼ 0.92)
na ¼ 37037 þ 233p 19a þ 97b (r ¼ 0.91) nf ¼ 30309 185p 1946a 342b (r ¼ 0.90)
Kamlet–Taft LSER
Polarity (63%)
Polarity (85%)
Polarity (63%), HBA (21%)
Polarity (66%), HBD (30%)
HBD (56%)
HBD (31%), HBA (32%)
Polarity (49%)
HBA (79%)
Polarity (67%)
Predominant solvent effect (%)
[69]
[68]
[67]
[66]
Ref.
88 Hydrogen Bonding and Transfer in the Excited State
Solute–Solvent Hydrogen Bond Formation in the Excited State 89
HBD solvent effect in S0 and show an HBA effect in S1, which means a different stability of the H-bonds in the ground and excited states. Other important observations are an increased width of the emission band, decreased fluorescence quantum yield and occurrence of dual fluorescence in polar protic solvents (BuOH, PrOH, EtOH, MeOH). Moreover, the comparison of the excitation and fluorescence spectra of DMAC led to the assumption that two emitting species, one non-H-bonded and another H-bonded, are present in alcohols, the solventcomplexed molecule emitting at longer wavelength. As we will see in the next sections, these are quasi-general observations for potential H-bond formatting probes in protic solvents and can be used as diagnosis criteria. Excited-state H-bond formation often means an increase in the energy of the H-bonds already present in the ground state owing to the change in the charge distribution upon excitation. 7-Amino coumarins are particularly interesting for studying H-bonding interactions in ground and excited states [70–72]. These compounds undergo electron transfer upon excitation, from the amino to the carbonyl group. Moreover, they are capable of forming H-bonds by participation of either the amino nitrogen (type A in Figure 4.3) or the carbonyl oxygen (type B) with HBD solvents, or by participation of the amino hydrogens (type C) with HBA solvents. Das et al. reported a study on the influence of the substituent at the 7-amino group on H-bond formation ability [4]. A Kamlet–Taft analysis was performed on three structurally related coumarins: C151 is the unsubstituted compound, while C500 and C35 are monoethyl and diethyl N-substituted respectively. The polarity contribution to S1 was found to be similar for all dyes (50%). The probability of H-bonding between the carbonyl oxygen or the lone pair of the amino nitrogen and the solvent (type A and B) increases in S1 compared with S0, as revealed by the a values, C151 showing the most significant effect (a is 1% in S0 and 20% in S1). The efficiency of the H-bonding of the amino hydrogens (type C) decreases upon excitation (for C151 the value of the b coefficient decreases from 73% in S0 to 33% in S1). Moreover, as the number of amino hydrogens decreases on going from C151 to C35, so does the magnitude of the interaction. The results show that, although the HBD and HBA solvent polarities make different contributions to the stabilization of the ground state, in excited state their effects become comparable. A comparison between the results of Kamlet–Taft and Catalan treatments can be found in the study of Brooker’s merocyanine, 1-methyl-4(40 -hydroxystyryl) pyridinium betaine, a compound characterized by strong hypsochromic absorption energy shifts (of 16.63 kcal mol1) and moderate hypsochromic fluorescence energy shifts (of 4.57 kcal mol1) [73]. Equations (4.7) and (4.8) present the correlations of the energy of the fluorescence maxima with the p and a Kamlet–Taft parameters and with the Catalan SA: f DEmax ðkcalÞ ¼ 44:75 þ 2:31p* þ 3:33a f DEmax ðkcalÞ ¼ 46:89 þ 4:31SA
ðr ¼ 0:97Þ
ðr ¼ 0:98Þ
ð4:7Þ ð4:8Þ
The polarity/polarizability of the solvents stabilizes the dipolar solute by electrostatic interactions, while the protic solvents interact through H-bonds to stabilize the negative oxygen atom of the dye. The a coefficient is slightly higher than p, suggesting that the H-bonding factor might play the main role in the stabilization process in S1. The two models yield similar conclusions, as the good Catalan correlation also shows the importance of the oxygen atom of the dye as a strong basic centre, easily forming an H-bond with protic solvents. A special type of H-bonding must also be mentioned, which involves the system of p-electrons of the aromatic rings that can act as H-bond acceptors in the presence of strong proton donors. This is the case of indole derivatives, including here the tryptophan-containing compounds. Experiments showed that the photophysical processes of these derivatives in protic solvents are rather complex in nature [74]. Two hypotheses were given for the mechanism of H-bonding of indole in protic solvents, assuming either the implication of the pyrrolic hydrogen or that of the p-electronic cloud of the heteroring. In order to discard the first hypothesis, experiments were performed on the methyl derivative. Its spectral properties were studied in n-hexane in the presence of increasing amounts of trifluoroethanol (TFE). The fluorescence spectra show the dependence on the TFE
90
Hydrogen Bonding and Transfer in the Excited State
concentration. At low alcohol concentration (104–102 M) the effect is insignificant, whereas at higher concentration a red-shift of the maxima, a broadening of the band and the presence of an isoemissive point can be noted, reflecting the formation of a new emitting species. The deconvolution of the spectra allows us to determine the emission maxima of the two species, at 310 nm for indole and for the ground-state H-bonded complex (HBC), and at about 330 nm for the new emitting species. This new species was assigned to an exciplex between the solute and TFE, further assigned to a proton-transfer H-bonded complex (PTC). Another experimental observation was the fluorescence quenching at large concentrations of TFE, as in the previously discussed case of DMAC. Moreover, lifetime measurements have shown that PTC is formed from the excited HBC, the decay of fluorescence being analysed in terms of triexponential functions, one of them with a negative pre-exponential factor, suggesting the formation of new fluorescent species from an excited precursor. This brings us to the complementary data of quantum yield and lifetime measurements and how they can be used as other diagnosis criteria in H-bond formation. For various fluorescence probes prone to specific interaction with the solvent, important changes in the emission efficiency, together with a decrease in the lifetime, were observed in protic solvents [75]. Performing time-resolved experiments, the occurrence of excited-state H-bonds can be attested by the analysis of fluorescence decays, bi- or multiexponential, correlated with the presence of multiple emitting species, solvated species, exciplexes, tautomers, etc. As stated above, a negative value for the pre-exponential factor in the fluorescence decay analysis indicates the presence of a species formed from another excited species. In protic solvents, in the case of biexponential decays, one of the lifetimes has a larger value and is usually assigned to the H-bonded species [76]. Correlating the experimental data on the excited-state lifetime (tf), the fluorescence quantum yield (w) and, when possible, the intersystem crossing quantum yield (wISC), the deactivation constants kf, kIC and kISC can be calculated using the formulae: ð4:9Þ kf ¼ w=tf kISC ¼ wISC =tf kIC ¼ ð1wwISC Þ=tf
ð4:10Þ ð4:11Þ
Linear dependencies of w, tf, knr and kf on the solvent polarity function were found, with different behaviour in protic and aprotic solvents [36, 77, 78]. The conclusion of such studies was that, for the majority of fluorophores, H-bond formation modifies the rate of the non-radiative deactivation processes through different mechanisms, which can lead either to a decrease or to an increase in w. Literature data attest that H-bond formation can either facilitate the ICT process, increasing the fluorescence emission from the ICT state [79, 80], or can quench this emission [81]. Possible mechanisms for the quenching or enhancement of the fluorescence quantum yield assisted by excited-state H-bond interactions were described, some of them correlated with the observation of dual fluorescence. They are summarized in Table 4.3 and will be extensively discussed in the following. Table 4.3 Mechanisms by which the intermolecular H-bond formation can influence the fluorescence quantum yield Fluorescence quenching Internal conversion due to the high-frequency accepting mode of H; most common Changes in excited state geometry Fluorescence enhancement Reversal of 1n–p , 1p–p states Suppression of the ISC process Increase in the energy barrier to conical intersection Inhibition of other deactivation channels (PET)
Solute–Solvent Hydrogen Bond Formation in the Excited State 91
4.3.2 Hydrogen-bond-assisted fluorescence quenching In most reported cases, a decrease in the emission intensity in protic solvents was found that was due to the predominant effect of the non-radiative deactivation processes [82–87]. Although the protic solvents can also act through non-specific polarity interactions, the main process leading to low quantum yields is the internal conversion (IC) induced by H-bonding. Owing to the high vibrational frequency of the OH group, the H-bond formation in alcoholic media can act as an effective accepting mode of non-radiative deactivation. Although the quantum yield of trans-ethyl-p-(dimethylamino)cinnamate is positively influenced by the polarity of the solvents, it decreases drastically in protic solvents, i.e. w ¼ 0.003 in water versus 0.010 in benzene and 0.022 in ACN [88]. The deviation from linearity in the Lippert–Mataga plots and the lifetime values support the correlation of the influence of protic solvents with the occurrence of a new deactivation channel due to H-bond formation. Another example is the highly solvent-sensitive fluorescence quantum yield of a quinoxalyl vinyl benzene dye [69]. In aprotic solvents, w increases strongly with increasing solvent polarity owing to a decrease in the vibronic coupling between the lowest 1 p–p and 1 n–p states. By contrast, in protic solvents, w decreases with increasing polarity, an effect rationalized in terms of an efficient IC by extensive vibronic mixing of the close-lying n–p and p–p states, enhanced by H-bonding with the solvent. The corresponding Kamlet–Taft fits are given in Table 4.2. It must be pointed out that no significant correlation of w was obtained considering all solvents (protic and aprotic, r ¼ 0.26). Similarly, one of the mechanisms proposed for explaining the experimental steady-state and lifetime measurements on cyano-substituted indolines in protic solvents assumes a deactivation channel due to Hbonds. For 5-cyano-N-methylindoline (CMI), unlike for N-methylindoline (MI), a short lifetime in EtOH was observed (0.82 ns versus 3.6 ns in ACN and 7.1 ns for MI in EtOH), but the most striking feature is the more than 10 times higher IC rate: 95 107 s1 in EtOH versus 9.4 107 s1 in ACN and 4.0 107 s1 for MI in EtOH. A similar behaviour is exhibited by 5-cyanoindoline (CI). As none of the other rates of the photophysical processes changed to a great extent in protic solvents, it was proposed that H-bond formation favours IC and is the main reason for lifetime shortening. Moreover, both CMI and CI have the same behaviour, so that it is not the NH group that is involved in this process, but the CN group [89]. This mechanism is also supported by theoretical calculations (see Section 4.4). As shown in Figure 4.3, more than one H-bond can be formed when the solute molecule presents two or more H-bond donor and/or acceptor groups. Furthermore, their geometry can be influenced by the conformation of the solute molecule, in which case more than one H-bonded excited species can be formed. An interesting process of fluorescence quenching in alcohols is reported for 7-(30 -pyridyl)indole and 7-(40 -pyridyl)indole, for which there is no possibility of intramolecular H-bond formation [90]. It was found that the rapid IC in these compounds is due to the simultaneous formation of two H-bonds implicating the indolic and pyridinic nitrogen atoms. For 7-(30 -pyridyl)indole, which can be present as anti and syn rotamers, characterized by the position of the pyridinic nitrogen on the opposite side or on the same side of the indole NH group respectively, the two H-bonds are either separated or form a bridged structure through the water molecule (Figure 4.6). Although a mechanism described in many biological systems, the photoinduced electron transfer (PET) favoured by H-bond formation is less known in organic molecules. It was found for an oxazine derivative [91] whose emission quenching in protic solvents can be explained by this mechanism. The different pattern of the H-bonds in the S1 state can also be due to a change in the molecular geometry upon excitation. Thus, it was found that the fluorescence of monocyanoanilines is practically quenched in water and TFE, although, in other solvents, w can reach values of about 0.34 [92]. The decrease in the quantum yield is associated with a decrease in lifetime, the non-radiative deactivation constant being very large, 2.3 1010 s1. The similitude of the behaviour in water and in the strong HBD solvent TFE supports the explanation that the formation of H-bonds involving the solute amine group is the cause of fluorescence quenching. In the case of 3-cyanoaniline, the variation in knr with temperature allows us to estimate an
92
Hydrogen Bonding and Transfer in the Excited State 7-(3'-pyridy1)indole syn
anti
O H R
N pyridyl
indole NH
pyridyl N
O H
indole NH
H
R
H
O
R
O R
Two separated H bonds
Cyclic structure of two H bonds
Figure 4.6 Hydrogen bond formation for the anti and syn conformations of 7-(30 -pyridyl)indole. Adapted with permission from [90]. Copyright Elsevier
activation energy for H-bond formation of about 11 kJ mol1. It was found that the non-radiative constant correlates with the difference, Du, in the pyramidal angle of the amine group in S0 and S1 states. The changes in geometry upon excitation are reduced for the dicyanoanilines, explaining the reduced fluorescence quenching in water. In the study of several anthraquinone, fluorenone, phthalimide and coumarin derivatives, Inoue et al. have observed a different behaviour on increasing the amount of ethanol (EtOH) in benzene [93]. The anthraquinone and fluorenone derivatives presented only a quenching of the fluorescence, for the phthalimides both a quenching and a shift of the maximum position were observed, while for the coumarin derivatives no effect was noticed. In order to explain these differences, the authors introduced a concept similar to that of hard and soft acid/base, considering as hard anions the compounds in which the negative charge is significantly localized on the carbonylic oxygen in the excited state, and as soft anions the compounds in which the charge is delocalized on the whole molecule. The anthraquinone and fluorenone derivatives for which a maximum effect of the alcohol was obtained behave as hard anions, the carbonylic oxygens having a significant negative charge to allow a strong H-bonding interaction. The coumarin and benzoxazine derivatives are characterized as soft anions, the negative charge on oxygens being low and not sufficient to ensure formation of a strong H-bond. As regards the phthalimide derivatives, the authors conclude that, although they can be classified as strong anions, other factors must be considered as well for explaining the weak fluorescence quenching by alcohol. 4.3.3 Hydrogen-bond-assisted fluorescence enhancement There are some cases in which the formation of H-bonds produces an enhancement of the fluorescence emission [30, 94, 95]. The most encountered mechanism consists in reducing either the ISC or IC non-radiative deactivation channels by modifying the order and/or the energy gaps between the first two excited states, S1 and S2, and the corresponding triplets. In aromatic compounds containing pyridinic nitrogens or carbonyl groups, the relative position of the 1 n–p and 1 p–p excited states (S1, S2) is strongly influenced by the substituents and by the solvents. A low S1–S2 energy gap determines, by the ‘close proximity effect’, a vibronic coupling that enhances the IC process. In cases in which the lowest singlet is 1 n–p , the quantum yield is rather low. The protic solvents determine, by both non-specific and specific interactions, a stabilization of the 1p–p state, as well as an increase in the energy of the 1 n–p state, reversing their energetic order and leading to a larger energy gap between them, hence causing a fluorescence enhancement. A qualitative scheme of the energy levels in non-polar and polar protic solvents is given in Figure 4.7. Several cases are reported underlying this effect of the reversal of the n–p and p–p excited states by increasing the energy of the weak emissive n–p state. In the case of 3-chloro-7-methoxy-4-methylcoumarin, the increase in w on going from hydrocarbon solvents to alcohols (0.10 in hydrocarbons versus 0.80 in
Solute–Solvent Hydrogen Bond Formation in the Excited State 93 E (S2) - *
1
3
-
*
(S2) n- *
1
n- * (S1)
1 1 3
n-
nonpolar
*
- * (S1)
3
n-
3
* *
polar (protic)
Figure 4.7 Qualitative scheme of the excited-state energy levels in non-polar and polar protic solvents for compounds containing both n and p electrons
alcohols and water) is due to a decrease in the non-radiative deactivation constant knr, kf being quasiconstant [95]. Another example is the comparative study of 4-phenoxy-N-methyl-1,8-naphthalimide and the unsubstituted N-methyl-1,8-naphthalimide [96]. The experimental data on 4-phenoxy-N-methyl-1,8-naphthalimide showed, on increasing the solvent polarity, a bathochromic shift of the fluorescence maximum, an increase in the bandwidth and a decrease in the emission. A Lippert–Mataga plot was quasi-linear if the points in the protic solvents, MeOH and EtOH, were excluded. All these effects attested the presence of an ICT excited state and of specific interactions in protic solvents. To gain a better understanding of the solvent influence, experiments were performed in dioxane in the presence of increasing amounts of water. Surprisingly, in the case of the unsubstituted naphthalimide, the presence of water determined an enhancement of the emission owing to a suppression of the ISC deactivation pathway, while for the 4-phenoxy derivative the same quenching as in pure solvents was noted [97–103]. Han et al. reported the synthesis of new fused phenothiazine derivatives with high solvent-dependent sensitivity (Figure 4.8, compounds a and b) [30]. Compound a is characterized by very low fluorescence quantum yields in protic solvents, i.e. in MeOH and water w < 0.0001. In spite of the very similar structure, compound b is highly fluorescent in alcohols and water as compared with the other solvents, i.e. w ¼ 0.048 (hexane), 0.132 (DMF, ACN) versus 0.481 (MeOH) and 0.554 (water). The authors rationalized the different behaviour by the change in the sequence of the first excited singlets. In aprotic solvents, the transition is of n–p nature, leading to low w. Owing to the close-lying n–p and p–p states of the dye, increasing the HBD ability of the solvent modifies the transition from an n–p type to a p–p type. In the case of a, the H-bonds formed in the excited state increase the probability of non-radiative deactivation, leading to quenching of the fluorescence in protic solvents. This behaviour is very specific, and the presence of a double bond in compound b of the series changes this ‘reversed polarity sensitivity’. Kamlet–Taft and Catalan treatment of the data supported this explanation. The experimental studies of the photophysical properties of substituted alloxazines (Figure 4.8, compound c) in ACN, dichloroethane, MeOH and water show a different behaviour in the first two solvents as compared with the protic solvents [104]. In protic solvents, beside a red-shift of the fluorescence band, an enhancement of the emission was noticed, the larger value being obtained in water. As the alloxazines have several centres that could be involved in H-bonding with proton donors, it is assumed that the behaviour in protic solvents is due to specific interactions. As it was previously shown that H-bonding to the pyrrolic NH group is of little importance, there remain as potential H bonding centres the pyridinic nitrogen atoms in the median rings and the two carbonyl groups. Taking into account the position of the fluorescence band in protic solvents, the
94
Hydrogen Bonding and Transfer in the Excited State S
S
N
N O
O (b)
(a)
10 N
HN
NH
O NH
N
NH
N
N
O
O
O (c)
(d)
R O
O
N
R'
R N
O H
H
H
R' (e)
Figure 4.8 Selected compounds showing fluorescence enhancement upon hydrogen bond formation
authors considered that the H-bond at N10 has a major influence on the p system, leading to a flavin-like structure. Although some experimental and theoretical data attest the p–p nature of the first excited singlet state, the low quantum yield of the alloxazines was explained by the energetically close n–p and p–p states. Considering the ‘proximity effect’, it can be assumed that the vibronic coupling between these states is strongly solvent dependent, the gap between the two states being increased in protic solvents. The n–p character of S1 in alloxazines as compared with the p–p nature of S1 in isoalloxazines (Figure 4.8, compound d) explains the lower quantum yields and the larger non-radiative deactivation constants of the former [105]. 1-Hydroxyfluorenone allows us to study three types of H-bond: an intramolecular bond between the OH group and the carbonylic oxygen, an intermolecular bond of the OH with the protic solvents and an intermolecular H-bond between the hydroxylic hydrogen and an H-bond acceptor [106]. The fluorescence maximum is red-shifted on going from CHX to more polar solvents. The results show a different behaviour in non-protic solvents as compared with that in the protic solvents; thus, as reflected by the deactivation constants kisc and kic, the predominant deactivation pathway changes from the ISC process in the former solvents to IC in alcohols (Table 4.4). The important role of the ISC process was explained by the close value of the energies of the S1 (p–p ) and T3 (n–p ) states. Any factor that modifies the difference in energy between these states, solvent or substituent, influences the kisc value [107]. Another explanation for the increase in fluorescence emission in protic media is the increase in these solvents of the barrier to reach a conical intersection (CI) that favours an IC deactivation process.
Solute–Solvent Hydrogen Bond Formation in the Excited State 95 Table 4.4 The different deactivation channels of 1-hydroxyfluorenone in aprotic and protic solvents. Data from Ref. [106] Solvent Cyclohexane CF3CH2–OH (TFE)
kIC (107 s1)
kISC (107 s1)
72 14 2 0.5
18 9 89 9
Unfortunately, this explanation requires elaborate calculations, an example being the theoretical study of 2-aminopurine [108]. The understanding of the photophysical properties of 2-aminopurine represents a good model for the behaviour of purine, a constitutional isomer of adenine. The experimental data show that, unlike adenine, the fluorescence quantum yield of 2-aminopurine is strongly enhanced in water as compared with CHX. The theoretical model used to explain this behaviour showed that the water molecules surrounding the solute increase the barrier of the transition state, leading to a CI that is an efficient IC funnel. The calculated statistical number of water–solute H-bonds is 0.81, 0.94 and 0.86 for the excited-state minimum, the transition state and the CI respectively, mainly implicating N3 and N7. Besides IC, other non-radiative deactivation channels can be suppressed by the formation of H-bonds, i.e. intramolecular photoinduced electron transfer (PET). N-[2-(2-hydroxylethylamino)-ethyl]-1,8-naphthalimide (HEAN) (Figure 4.8, compound e) presents potential H-bond acceptor groups [109]. Its photophysical properties are strongly solvent dependent. In non-polar or low-polarity solvents, or in aprotic polar solvents such as DMF, ACN and ethylacetate, the fluorescence quantum yield is very low, at the limit of detection. A strong enhancement of the emission was observed in polar protic solvents, maximum w being found in water (0.453). A plot of w versus the solvent polarity function (Lippert–Mataga model) or ETN parameter resulted in a regular curve. These experimental observations show a dependence of the photophysical properties on the proticity of the solvent. The special behaviour in protic solvents was explained by the possibility of H-bond formation between the proton of the solvent and the lone electron pairs of the strong acceptors of the molecules, the alkoxide oxygen and the secondary amine nitrogen. The formation of these intermolecular H-bonds suppresses a deactivation channel via a PET mechanism from 2-hydroxyethylamino-N to the dicarboximide group. 4.3.4 Dual fluorescence In certain conditions, the fluorescence emission of some compounds is characterized by the occurrence of two bands, more or less resolved. The band at shorter wavelength (SW) is usually ascribed to the FC excitation leading to a locally excited state (LE), and it is also referred to as ‘normal’ fluorescence, Fn. The band located at longer wavelength (LW) is considered as ‘abnormal’ fluorescence, Fa, being correlated with relaxed ICT or TICT excited states, or with the presence of new emitting states, H-bonded complexes, tautomers, zwitterionic structures and exciplexes. The positions and the intensities of these bands are strongly solvent dependent, and the phenomenon can be enhanced by interactions with the protic solvents [110]. An interesting case is represented by three related compounds, 2-(40 -N,N-dimethylaminophenyl)imidazo [4,5-b]pyridine (DMAPIP-b), 2-(40 -N,N-dimethylaminophenyl)imidazo[4,5-c]pyridine (DMAPIP-c) and 2-(40 -N,N-dimethylaminophenyl)imidazo[4,5-d]pyridine (DMAPIP-d), which differ only in the position of the pyridine nitrogen (Figure 4.9). All compounds show dual fluorescence in protic solvents, with the most significant effect being observed in the case of DMAPIP-b (more red-shifted and better-resolved LW band) [111, 112]. The LW emission has been assigned to the formation of a TICT state assisted by Hbonding. This explanation is supported by several facts: (i) the dependence of the ratio of the two band intensities on the HBD ability of the solvent; (ii) the decrease in or even the lack of an LW band in viscous
96
Hydrogen Bonding and Transfer in the Excited State CH3
N
CH3
N
N NH
N
N N
CH3
NH
DMAPIP-b
H3C
H 3C
O
DMAPIP-c
H3C
NH
CH3
N
CH3
O
O
H3C
N
O N
CH3
MAPAEE
MDMANA
O
CH3
DMANAN
H O
N
O
H
3-HQ-Bf
Figure 4.9 Selected compounds presenting dual fluorescence in protic solvents
solvents which reduce the possibility of intramolecular rotation of the electron donor and acceptor moieties; (iii) a biexponential fluorescence decay in protic solvents. From the three nitrogen centres of the molecule, the experimental data attest implication of the pyridine nitrogen in the H-bond. In an acidic medium, only a single emission band is observed, as the nitrogen from the imidazole ring is protonated. Furthermore, H-bondinduced TICT is not observed in the related compound (N,N-dimethylaminophenyl) benzimidazole, where the pyridine nitrogen is absent. The H-bond between the solvent and the electron acceptor fragment determines a better planarity of the acceptor group with the aromatic cycle and favours charge transfer towards the electron donor part. In water, the strongest H-bond donor, the TICT state is more stabilized, and another non-radiative deactivation channel towards the ground or the proximal triplet states becomes active. Three related compounds containing a secondary or tertiary amine as a donor group and an ester or nitrile as an acceptor group all present dual fluorescence in protic solvents. The presence of a double bond between the acceptor and the aromatic ring ensures, for all compounds, a supplementary p-conjugation. For (E)-3-(4methylamino-phenyl)-acrylic acid ethyl ester (MAPAEE in Figure 4.9), three possible positions for H-bond formation in protic solvents can be assumed: two sites involving the oxygen atoms of the acceptor and one implicating the lone pair of the nitrogen which can fix a hydrogen from the solvent [83]. In non-polar solvents
Solute–Solvent Hydrogen Bond Formation in the Excited State 97
such as methylcyclohexane (MCHX), a single band is observed, assigned to the LE state. Adding increasing quantities of MeOH to MCHX determines the appearance of a new band, bathochromically shifted, and a slight decrease in the intensity of the former band; the family of spectra shows a clear isoemissive point. The new band was assigned to an ICT solvated state due to the solute–MeOH H-bonded complex involving the nitrogen lone pair. The decrease in the quantum yield in MeOH as compared with the value in an aprotic solvent of similar polarity, ACN, together with the linear dependence of the fluorescence band maximum on the a parameter, supports the role of H-bond formation in the overall photophysical properties. Similarly, the methyl ester of N,N-dimethylaminonaphthyl-(acrylic)-acid (MDMANA) also presents dual emission in polar and protic solvents and a single fluorescence band in non-polar solvents [113, 114]. The LW band shifts from 461 nm in CCl4 to 521 nm in water and was ascribed to a CT band. The addition of EtOH to MCHX shifts continuously the band from 440 nm (MCHX) to 480 nm at a content of 70% EtOH. The linear dependence of nf versus a shows the H-bond formation in protic solvents. On the other hand, the lack of mirror symmetry of the absorption and fluorescence spectra attests a change in the geometry upon excitation. The experimental data supported by DFT calculations are rationalized in terms of three coexisting excited states: an LE state, a TICT state and an H-bonded state. Unlike MAPAEE and MDMANA, N,N-dimethylaminonaphthyl-(acrylo)-nitrile (DMANAN) can be implicated in only one type of H-bond, between the amino nitrogen lone pair and the hydrogen of the alcohol [115]. The LE maximum is located at about 425 nm, and the LW (CT) band is in the range 473–498 nm. The involvement of the H-bond in the LW band is evidenced by the linear dependence of nf on a. Additional support for the assumption that the H-bond implicates the nitrogen lone pair was obtained from the experiments in acidic medium. Thus, the protonation of the diethylamino group and consequently the fixation of the nitrogen lone pair hinder the charge transfer and determine the decrease in the LW (CT) band. N,N-dimethylbenzodiazepine exhibits a single fluorescence band in aprotic solvents (normal fluorescence, Fn band) and two bands in protic solvents (an abnormal band, Fa, and the normal one, Fn) [116]. The abnormal fluorescence is assigned to an intermolecular interaction between the solute and solvent molecules in the excited state, with the formation of an exciplex. Plotting the Stokes shift for both bands, Fa and Fn, against the solvent polarity function gives two lines with different slopes, allowing us to estimate the variation in the dipole moment upon excitation for the two emitting states. In protic solvents, fluorenone also presents a dual fluorescence owing to the presence of two excited forms, a free form and an H-bonded complex, with different S1 photophysical features. In the strongly HBD solvent TFE, only the H-bonded complex is evidenced [117]. Some experiments were conducted in mixtures of solvents with similar polarity (close « values) but different basicity. An example is the study of 3-hydroxyquinolones in mixtures of ACN (« ¼ 35.7, b ¼ 0.32) and DMF (« ¼ 37.2, b ¼ 0.74) [110]. 2-Benzofuryl–3-hydroxy–4(1H)-quinolone (3-HQ-Bf) (Figure 4.9) presents a dual fluorescence with good resolved bands and a separation of the maxima in the range 2700–4200 cm1. It was found that the solvent effect is better reflected by the intensities of the two bands observed, IN /IT, than by the changes in the maximum positions. This ratio was found to be linearly dependent on the b parameter (log(IN /IT ) ¼ 1.59 þ 1.62b, r2 ¼ 0.9). Time-resolved experiments were also performed on 3-HQ-Bf in MeOH as protic solvent and in aprotic solvents of different basicity. The fluorescence decays monitored on the two bands, N and T, were biexponential, with similar lifetimes but different pre-exponential factors. The negative sign for the T factor attests its formation from N by an excited-state intramolecular proton transfer (ESIPT) mechanism. The overall mechanism is characterized by the following steps. Firstly, the molecule in the ground state is solvated at the OH group by a basic solvent; upon excitation, the charge densities are changed and the carbonylic oxygen, now with a high negative density, competes with the solvent for H-bonding with the hydroxylic hydrogen, generating the N species. This species goes to the T form through the ESIPT mechanism. The T state is further deactivated by radiation emission. The overall mechanism is presented in Figure 4.10.
98
Hydrogen Bonding and Transfer in the Excited State C-O-
C-OO-H
solvent
C-O-H
H
ESIPT
+
+
S1-solvated N state
+
N*state
C=O N* fluorescence O-H
O-
O
*
T state
T* fluorescence
solvent
S0-solvated N state
S0-N state
S0 -T state
Figure 4.10 Schematic representation of the ESIPT mechanism. Adapted with kind permission from [110]. Copyright 2009 Springer Science þ Business Media
4.4 Design of the Experiments For a good understanding of solvent effects, and especially for separating the non-specific and specific interactions involved, the experimental conditions must be carefully chosen. As was previously shown, the classification of solvents into polar and non-polar is made in terms of the polarity/polarizability functions, which depend on the dielectric constant « and refractive index n. The polar solvents can be further separated into protic and non-protic ones. The Kamlet–Taft parameters, a and b, characterize the ability of the solvents to act as HBD or HBA respectively. A survey of the conditions required for performing reliable experiments leads to the following: . . .
selection of the fluorescence probes; selection of the solvents and mixtures of solvents; use of isotopic effects.
4.4.1 Selection of fluorescence probes The fluorescence probes used for describing solvent effects and excited-state deactivation processes due to H-bond formation must be selected in such a way as to avoid the occurrence of other non-radiative pathways such as ISC and the close positioning of the 1 p–p , 1 n–p , 3 n–p excited states [118, 119]. A convenient structural parameter in such probes is a rigid molecular skeleton that does not favour any conformational or configurational relaxation processes [120]. The most important fluorescence probes for investigating H-bonding interactions in the excited state were presented in Figure 4.2. Performing experiments on related compounds differing only in a single structural element or in a substituent brings about valuable information by comparison of the results [5, 107, 118]. A special case is the substitution of the hydrogen by a methyl radical in compounds containing several atoms that can participate in H-bond formation, i.e. pyrrolic or amine nitrogen/hydrogen (see, for instance, the previously discussed case of CMI versus CI, which proves that NH is not implicated in the H-bond decisive for the
Solute–Solvent Hydrogen Bond Formation in the Excited State 99
deactivation processes [89]. Another case in which N-methyl substitution allows us to identify the H-bond centre is with some pyrrole derivatives. The photophysical properties of non-methylated compounds are strongly influenced by the presence of pyridine as a solvent, unlike those of the N-methylated derivatives. The suggested mechanism comprises, in the first step, the formation, in the ground state, of an H-bond between the NH group as H-bond donor and the pyridine nitrogen as H-bond acceptor [121]. In compounds containing an amine group that can be involved in an intramolecular H-bond, the progressive alkyl substitution of the amino group is useful in differentiating intermolecular H-bonds from intramolecular ones. Thus, the solvatochromism of o-nitroaniline and N-alkyl-o-nitroaniline probes was studied in solvent mixtures involving an ‘inert’ non-polar cosolvent (CHX) and an HBA solvent (THF) [122, 123]. An interesting study was performed on a symmetrical molecule that has no permanent dipole moment in the ground state, 2,6-diaminoanthraquinone [124]. The observed experimental effects in such compounds are due only to intermolecular H-bonds with the solvents; the presence of two carbonyls that can act as independent dipoles explains this behaviour. The linear dependence of DnSt against the b solvent parameter confirms this explanation. The presence of a solvated species is also attested by the hypsochromic shift of the emission maximum in EtOH at increasing excitation wavelength. 4.4.2 Selection of solvents Some general rules for the selection of the most appropriate solvents can be extracted from a survey of the literature data. The experiments must firstly be performed in non-polar solvents or solvents with low polarity in which specific interactions are not expected. Quantitative treatment in such solvents through the Lippert– Mataga or Dimroth–Reichardt analysis allows us to estimate the change in the dipole moment upon excitation. Starting with this result, the use of polar and/or protic solvents permits the identification of the nature of the excited state and the contribution of the specific interactions to the overall solvent effect. In order to identify the role of H-bond formation in the general pattern of the solvation processes, the use of solvents with similar polarity but different H-bonding capabilities is recommended. The selection of solvents also depends on the expected type of H-bonds. Valuable results are obtained by using mixtures of solvents with different characteristics. Several examples will be discussed in the following. A good fluorescence probe for analysing the role of H-bonds in the excited state is 4-aminophthalimide, studied by Krystkowiak et al. [34]. Containing both an electron donor and an electron acceptor group, this compound represents a good model for the change in the electron density pattern upon excitation and, consequently, for the change in the solute–solvent interactions in the S1 state. However, with more than one group prone to form H-bonds, choosing the solvent is very important for evidencing the various possible effects. The authors consider that the best solvents for evidencing only the non-specific interactions are the 1-chloro-n-alkanes. Although they cover only a small range of e(«, n) values (0.11–0.23), both their HBD (a) and HBA (b) capabilities in the Kamlet–Taft model are zero. The second condition was to choose solvents that can form only a single type of H-bond. Thus, for studying the H-bonds of the NH2 group, the authors used DMSO, characterized by b ¼ 0.76 and a ¼ 0. For investigating the H-bonds of the oxygen atom of the carbonyl group, solvents with a 0 and b ¼ 0, i.e. polyfluorinated alcohols (hexafluoroisopropanol (HFIP), a ¼ 1.96, b ¼ 0) were used. Comparing the absorption, emission and excitation spectra in all these solvents, the authors succeeded in evaluating the energy of the H-bonds formed by the different groups in the S0 and S1 states, and to establish the excited species present in each case and the main deactivation pathways. In some cases the alcohols do not have sufficient H-bonding capability to induce H-bond formation, and strong HBD solvents are necessary. In the case of fluorenone, the presence in alcohols of H-bonds implicating the oxygen of the carbonyl group was extensively reported [60, 84–86, 125–127]. However, the related ketone benzo[b]fluorenone was not quenched by alcohols [128], and H-bond formation was observed only in the strong HBD solvent TFE. This was explained by the structural differences between the two compounds, and
100 Hydrogen Bonding and Transfer in the Excited State
especially by the changes in the overall aromaticity that determine different charges on the carbonyl group. In the case of benzo[b]fluorenone, the fact that its emission spectrum has been found to be independent of the excitation wavelength indicates that the emission occurs from a single excited species. The use of mixtures of solvents allows extension of the properties of the media. The separation of the different kinds of interaction can be done using increasing concentrations of a potential HBD or HBA solvent in a non-polar, non-protic solvent [129]. The choice of the particular solvent mixtures depends on the fluorescence probe used. An interesting case is the class of b-carbolines [76, 130, 131]. They present very attractive photophysical properties, dependent on the solvent features. Owing to their specific structure, the simultaneous presence of the pyridinic and pyrrolic nitrogen, b-carbolines can act as both HBD and HBA. The studies were performed in CHX, in the presence of increasing amounts of HBD solvents, i.e. HFIP, 2-chloroethanol (2-ClEtOH), tert-butyl alcohol (t-BuOH) and HBA solvents, i.e. THF, DMF and hexamethyl-phosphoramide. Differing from the 1- and 1,4-disubstituted anthraquinones [132, 133] which can be involved in both intraand intermolecular H-bonds, the 2- or 2,6-disubstituted anthraquinones are only involved in intermolecular H-bonds in protic media [124]. The solvent influence on the excited-state molecular interactions was studied using either pure solvents or several solvent mixtures combining a polar solvent, DMF, with a non-polar (benzene), a protic (EtOH) and two other polar solvents with similar polarities but different b values, DMSO (HBA) and ACN (HBA and HBD). In all mixtures, a deviation from the ideal behaviour was found, the most interesting results being obtained for the system DMF–EtOH. It was found that the solute–solvent H-bonding interactions can break the H-bonds between the self-associated alcohol molecules. The presence of solute–solvent H-bonds was attested by the linearity of the plot DnSt versus b and by the dependence of the emission band in EtOH on the excitation wavelength. A hypsochromic shift of the fluorescence spectrum was obtained for an increase in the excitation wavelength from 458 to 515 nm; such behaviour was not observed in the polar aprotic DMF. The differences in the photophysical properties in mixtures of solvents as compared with an ideal solution are due to preferential solvation in one of the solvents. Some models were described [134, 135] to treat this behaviour. A good example of their applications is the study of fluorenone and 4-hydroxyfluorenone using two mixtures of different solvents, non-polar–polar aprotic (CHX–THF) and non-polar–protic (CHX–EtOH). The last system allows the detection of possible H-bonding effects [136, 137]. The plot of the fluorescence shift against the molar fraction of one component shows a non-linear dependence. According to Bakshiev [23], the averaged molar fraction of the polar component in the first solvation shell of the solute is given by the formula B «eff «n xp ¼ «p «n
ð4:12Þ
where represents the effective dielectric constant «eff ¼ «n xBn þ «p xBp
ð4:13Þ
The constant can be estimated from the shift of the absorption (d~nA ) or fluorescence (d~nF ) maxima in the binary mixtures with respect to the non-polar solvent, as follows: d~nA;F
! «eff 1 n2 1 2n2 þ 1 nþp n ¼ D~nA;F D~nA;F ¼ m1;2 2 n þ2 «eff þ 2 n2 þ 2
ð4:14Þ
Solute–Solvent Hydrogen Bond Formation in the Excited State 101
where m1 and m2 depend on the ground- and excited-state dipole moments, mg and me, and on the Onsager cavity radius of the solute, a: m1 ¼
~ m g ð~ m e ~ mgÞ ; 2p«0 hca3
m2 ¼
~ m e ð~ m e ~ mgÞ 2p«0 hca3
ð4:15Þ
The plot versus x p deviates from linearity, indicating a preferential solvation. The preferential index of solvation, Z, is calculated using the equation Z¼
1 Cm2 M Df ð«np Þ 4p«0 RTdr6ss
ð4:16Þ
where C is estimated as 3=8p, considering the solute and the solvent as spherical molecules, De(«np) ¼ e(«n) e(«p), the solvent functions being calculated for both types of solvent by the usual formula, considering the respective dielectric constants «n or «p: eð«Þ ¼
2ð«1Þ 2« þ 1
ð4:17Þ
M represents the mean molecular mass for the two solvents, d is the mean molecular density of the polar and non-polar solvents and r is the solvent–solute distance (rs–s ¼ a þ r). Coumarin 153 was studied in toluene–ACN and toluene–MeOH mixtures in order to evidence the difference between the polar (ACN) and the protic (MeOH) solvent [87]. Although in the ground state C153 does not show preferential solvation, a non-ideal dependence of the photophysical properties on the solvent composition was found in the excited state. The behaviour observed in the presence of MeOH was explained by H-bond formation. The interaction of harmane with strong HBD solvents leads to two species in the S0 state, both implicating the pyridinic nitrogen: an H-bonded complex (HBC) and a second species, labelled as the proton transfer complex (PTC), in which a proton is transferred to the nitrogen via the assistance of a second molecule of solvent [76]. In fact, the excitation process determines the excitation of the free harmane molecule and of these two H-bonded species. In the presence of HFIP, the excited PTC species undergoes other processes leading to a cation-like and a zwitterionic species. It was found that the best conditions to evidence all these species were when a mixture of CHX–toluene (90:10 v/v) was used. Steady-state and time-resolved experiments in the presence of increasing amounts of HFIP allow us to estimate the kinetic aspects of the overall process. In order to isolate the H-bond donor of the heteroring, the 9-methyl derivative of harmane was used. The fluorescence decays in alcoholic solvents were fitted to a biexponential function; the shortest lifetime, t1 ¼ 2.1 ns, similar to that in pure CHX, was assigned to the free molecule, and the longer lifetimes, t2 ¼ 3.4 ns (ClEtOH) and t2 ¼ 3.7 ns (HFIP), to the H-bond-complexed species. Starting with the non-methylated derivative, a similar behaviour was firstly observed, followed by the appearance in ClEtOH and HFIP of a new band assigned to a zwitterionic structure. A separation of the polarity and H-bond interaction by using mixtures of THF and CHX (non-polar), ACN (polar, non-protic), trichloroacetic acid (TCAA) and several alcohols, MeOH, EtOH, PrOH and BuOH (polar, protic), is described by Yamamoto et al. for 4-phenyl-1-N,N-dimethylaminobutane (PDAB) [138, 139]. The fluorescence spectrum of PDAB consists of three bands labelled in order of their increasing wavelength as A (285 nm), B (343 nm) and C (385 nm), and assigned to the phenyl chromophore, to the excited amine group and to an intramolecular exciplex respectively. The last two bands are overlapped and can be separated only by temperature effects. Considering the polar solvents, ACN only has a polar effect, and TCAA only an
102 Hydrogen Bonding and Transfer in the Excited State
H-bonding effect, whereas the alcohols provide both non-specific and specific interactions. The presence of increasing amounts of ACN determines a decrease in bands B and C, without a change in the intensity of band A. In the THF–BuOH mixtures, the addition of the protic solvent determines an increase in the intensity of band A, at the expense of B and C intensities. The same effect, but more pronounced, was noticed in the presence of TCAA. In this case it can be assumed that the changes in the fluorescence spectrum are due exclusively to H-bond formation, i.e. to specific interactions. Defining the ratios X and Y, equations (4.18) and (4.19), and plotting their values against the percentage of BuOH in the mixture, we can separate the two effects, because the X versus BuOH% dependence reflects only the contribution of H-bonding to the change in the fluorescence emission: X¼
Y¼
IA IA0 ½TCAA ¼ 0 excess IA IA ½PDAB 0 IB;C IB;C 0 IB;C
¼
ð4:18Þ
½TCAA ½PDAB
ð4:19Þ
The dependence of the ratio of the H-bonding effect to the polar effect versus the volume percentage of alcohol is linear for BuOH and presents a maximum deviation from linearity for MeOH. For up to 10% vol. MeOH the effect is predominantly due to H-bonding interaction; the plot shows saturation at a value of 0.8 for the H-bonding to polar effect ratio. 4.4.3 Use of deuterated alcohols One of the experimental methods for checking the non-radiative deactivation pathway via H-bond formation is the use of deuterated alcohols, in which case H-bond formation increases the IC rate via a vibrational mechanism. The reduction in the vibrational frequency upon deuteration and the lower ability of the deuterated compounds to be involved in H-bonds determine a decrease in the non-radiative deactivation constant. The fluorescence emission of amino-substituted fluorenones, except 1-aminofluorenone, is strongly quenched in polar protic solvents. Performing experiments in deuterated alcohols, a decrease in the nonradiative deactivation constant was found in going from EtOH to EtOD, i.e. a ratio knr(EtOH)=knr(EtOD) ¼ 1.2, whereas for other aminofluorenones this ratio reaches values as large as 1.9 (Table 4.5) [118]. Some explanations for this behaviour are the presence of an intramolecular H-bond between the carbonylic oxygen and the amine hydrogen, an enhanced decay by an ISC mechanism or the formation of a TICT state. The other compounds in the series, 3- and 4-aminofluorenones, are also strongly quenched in the presence of alcohols, the ratio of the non-radiative deactivation constants in EtOH to those in benzene being in the range between 125 (4Table 4.5 Comparative effects of water, alcohols and the corresponding deuterated species on the ratio of the non-radiative deactivation constants knr(R–OH)/knr(R–OD) knr(Et–OH)/knr(Et–OD) Compound Ref.
F 1.9
3MAF 1.3 [118]
knr(H–OH)/knr(H–OD)
2-PAQ
4-AP
Cum
R6G
R128
R101
RB
1.7 [140]
5.7 [34]
1.1 [36]
5.3
2.9
3.3
1.3
[141]
a F: fluorenone; 3-MAF: 3-(methylamino)-9-fluorenone; 2-PAQ: 2-piperidino-9,10-anthraquinone; 4-AP: 4-aminophthalimide; Cum: coumarin derivatives; R: rhodamine.
Solute–Solvent Hydrogen Bond Formation in the Excited State 103
aminofluorenone) and 43 (3-aminofluorenone). The same behaviour has been previously discussed for some aminoanthraquinones, for which this ratio reaches larger values (158 for 2-aminoanthraquinone and 263 for 2piperidine-anthraquinone (2-PAQ)) [140]. Aromatic thioketones present fluorescence from the S2 excited state [142], the non-radiative deactivation mechanism in fluorinated hydrocarbon solvents being S2–T1. The fluorescence from the S1 state is not observed owing to the more rapid ISC process S1–T1 than the IC S2–S1. The same evolution of the excited states was also postulated in solvents presenting interaction with the solute, such as hydrocarbons and ACN, for which a shortening of the lifetime was observed. The shortest lifetime was obtained in water (1 ps), reflecting the de-excitation of a benzopyranthione–water complex [143]. Support for this assumption was obtained by repeating the measurements in deuterated water. It was found that t(S2) in D2O is twice the value in water. There are also cases in which the use of deuterated water rules out the hypothesis of the H-bonds influencing the solute photophysics in protic solvents. Thus, in a recent study on coumarin 7, Pal et al. found that in MeOH–water and MeOH–deuterated water mixtures, the values for the fluorescence quantum yield and the lifetime are practically unchanged [36]. The lack of any isotope effect reflects the fact that the peculiar photophysical properties of the compound in protic media are not due to H-bond formation. 4.4.4 Other experimental parameters (temperature, excitation wavelength) Emission experiments can be performed in some other experimental conditions as well, such as different temperatures or excitation wavelengths. Increasing the temperature, H-bond formation is disfavoured, so that the rate of the processes it determines should be modified. Steady-state fluorescence and lifetime determinations at variable temperatures can provide useful information about the different processes occurring in the excited states. Although the fluorescence emission is deeply influenced by temperature, it was stated that the effect is merely due to changes in the non-radiative deactivation constants, the radiative processes being practically uninfluenced. Temperature effects were useful in evidencing the presence of excited-state H-bonds. Two kinds of experiment are reported in the literature data: measurements in an interval of temperatures around room temperature [36, 115, 134, 144–147] and measurements in glassy solutions at 77 K [29, 83, 120, 148]. The measurements of fluorescence lifetimes at variable temperatures in short interval ranges around 298 K in protic solvents was used to differentiate between some possible mechanisms such as electron ejection [89], participation of TICT states in the deactivation pathway, flip-flap motion of the amine group in non-polar and polar solvents, 1-N-methylamino- and 1-N,N-dimethylamino-9,10-anthraquinone dyes and coumarin [36, 115, 144, 145], the ESIPT mechanism [146], etc. Experimental determinations were also used for estimation of the activation energy of the various processes and for estimation of the rotational diffusion time [147]. Measurements at several temperatures can thus be used as additional support in identifying excited-state Hbonds. Experiments performed for DMANAN [115] in ACN and MeOH at several temperatures in the range 273–323 K showed a different behaviour in the two solvents. With increasing temperature, in MeOH, both SW and LW bands decreased in intensity, while in ACN only the LW band was affected (strongly decreased). These observations are explained assuming different mechanisms of non-radiative deactivation: in ACN, the normal effect of temperature on the non-radiative process is operating, while, in MeOH, increasing temperature produces the breaking of H-bonds. The influence of temperature on H-bond formation can also be outlined by lifetime measurements, i.e. by investigating the temperature effect on tf. It was found that, in protic solvents with e(«, n) < 0.17, the tf of a coumarin derivative increases with temperature, while, in solvents with e(«, n) > 0.17, tf decreases with temperature [144]. In order to explain this behaviour, H-bonding interactions with the formation of a solute–protic solvent complex were inferred. As depicted in Figure 4.3, the coumarin derivatives can be involved in three types of H-bond (A, B and C) with protic solvents. It was assumed that the H-bonds
104 Hydrogen Bonding and Transfer in the Excited State
correspond to the B type. Increasing the temperature, the equilibrium is shifted towards the free, uncomplexed dye, and tf increases. Experiments at 77 K, sometimes correlated with measurements in viscous media, are performed for compounds presenting dual fluorescence at room temperature, for reinforcing the assignment of the LW band to a solute–H-bonded species. It is assumed that, in glassy solutions, the solvent molecules have reduced the possibility of undergoing a reorientation and the spectra reflect the predominant species in the system. As was previously discussed, the widely used fluorescence probe fluorenone [120] exhibits dual fluorescence (LW and SW bands), at room temperature, in all solvents, but only the SW band in MeOH and the LW band in TFE at 77 K. The experimental data at room temperature were rationalized in terms of an equilibrium between the free fluorenone and fluorenone–H-bonded species, equilibrium dependent on the proton donor capability of the solvent. In MeOH the equilibrium is shifted towards the free fluorenone (SW band), and in the strong proton donor, TFE, towards the fluorenone–H-bonded species. The comparison of these spectra supports the assignment of the LW band to a fluorene–H-bonded excited species, the main species in TFE. A similar case is presented for 4-N,N-dimethylamino cinnamaldehyde [29], which presents at room temperature a single fluorescence band in hydrocarbon solvents but distinct dual fluorescence in both polar aprotic and protic solvents. The main difference between the spectra in aprotic and protic solvents consists in a larger Stokes shift for the LW band in the latter solvents. This observation, together with a good linearity of the plot DnSt versus a HBD parameter, attests the presence of H-bonded species. Additional support for H-bond formation was obtained from experiments in EtOH:water mixtures; it was found that the addition of small water quantities produces significant modifications, the spectrum being similar to that in pure water. On the other hand, the emission spectra do not represent the mirror image of the absorption spectrum, pointing to a different geometry of the S1 state, i.e. to the presence of a possible TICT excited state. To support the assignment of the band to an H-bonded species, measurements at 77 K were performed, assuming in this condition a possible reduction in molecular flexibility. The fluorescence spectrum consists of a single band in hydrocarbon glasses, but of two bands in hydrocarbon glasses with traces of water or in EtOH glasses. At the same time, as compared with the room-temperature results, a hypsochromic shift of the LW band and a fluorescence enhancement were noticed. As the TICT states present phosphorescence at low temperature, the presence of the LW band in EtOH glasses supports the assumption concerning the role of H-bonds against TICT processes. Experimental measurements at different excitation wavelengths can indicate if there are more than one excited species in the system. When the shape of the spectrum depends on the excitation wavelength, multiple species are present. Several excitation wavelengths, lex, are usually selected, characteristic to the absorption of two or more possible species in the ground state, in order to obtain different populations of these species in the excited state, leading to different emission spectra. An example in which the use of several lex chosen towards the red edge of the absorption spectrum allows us to identify the excited species formed from coexistent H-bonded and non-H-bonded ground state species is reported in [149]. Thus, the fluorescence spectra of 40 -dialkylamino-3-hydroxyflavones obtained at different lex can provide a measure of the ground-state distribution of H-bonded and non-H-bonded species, making the compounds useful as sensors for detecting the H-bonding potential of the environment. The sequence of experimental determinations that can be used to evidence the formation of new species in the excited state and can also be applied to investigate H-bond formation, together with the most important information they provide, are summarized in Table 4.6.
4.5 Theoretical Modelling of the H-Bonds As already discussed, it is difficult to gain direct experimental insight into the processes that take place in the excited states of molecules. This is why many authors rely on quantum chemical data as a complementary
Solute–Solvent Hydrogen Bond Formation in the Excited State 105 Table 4.6 Experimental steps in the study of excited-state H-bonds Method
Information
Recording of absorption spectra
Characterization of the electronic transitions (n–p , p–p )
Recording of fluorescence spectra
Solvent influence on: emission maximum (maxima for dual fluorescence) bandwidth fluorescence quantum yield
* * *
Analysis of experimental data: Lippert–Mataga plot Reichardt–Dimroth plot Kamlet–Taft or Catalan models
Change in the dipole moment upon excitation Separation of non-specific and specific interactions Quantification of HBD and HBA effects
Deconvolution of the emission band
Number and position of the more or less overlapped bands
Recording the fluorescence spectrum using several excitation wavelengths, lex (at the red edge of the absorption spectrum)
Evidence of the number of excited species: lack of influence of lex ! single species spectrum modifications dependent on lex ! multiple species
Recording of the excitation spectrum, comparison with the absorption spectrum
Perfect match of the spectra ! same species in S1 as in S0 Differences in the two spectra ! new species in S1
Time-resolved experiments
* *
Parameters of fluorescence decay: lifetime amplitude of the exponential function(s) determination of kr, knr
* * *
Variable temperature determinations
Breaking of hydrogen bonds Determination of activation energy
Comparison of absorption, emission and excitation spectra
Evaluation of the energies of hydrogen bonds
source of information. Different models have been used to characterize the H-bond formation between the solvent and the solute in its excited state. In what follows we will span the recent literature concerning this aspect. It is worth mentioning that the theoretically studied molecules are usually model molecules, with a relatively small number of atoms, in order to use state-of-the-art quantum chemical methods capable of yielding accurate results. This section focuses on the application of theoretical models to excited-state solute–solvent H-bond formation, without a detailed description of the computational methods used, which can be found elsewhere [150, 151]. 4.5.1 Modelling the system For non-H-bond formatting solvents, where no specific interactions with the solute molecule can occur, a continuum model constitutes a reasonable idea with good correlation with the experiment [152, 153], although there are studies that show a better estimation of some spectral features by explicitly taking into account the solvent molecules [154]. The problem of an excited-state solute that forms H-bonds with the solvent molecules comprises a complex model system formed by the solute molecule in its first excited singlet and a number of accessible/relevant solvent molecules adequately oriented in order to form H-bonds, surrounded by the bulk solvent molecules, which are not implicated in the specific interactions. In quantum chemical terms, such a system has at least two components, the solute–solvent molecule(s) complex, which
106 Hydrogen Bonding and Transfer in the Excited State
treats the relevant solvent molecules explicitly, and the bulk solvent, usually a global component in which the solvent is treated implicitly, by means of some macroscopic parameters derived from experimental data, such as the dielectric constant, density, molar volume, etc. Another way of treating the solvent is by statistical methods. The accuracy of the results depends on the adequacy of the model and, of course, on its complexity. It was implied that quantitative correlation with the experimental data can be acquired if all the solvent molecules in the first solvation shell are treated explicitly by an ab initio quantum chemical method and the bulk solvent effect is taken into account within the framework of a continuum solvation model [155–159]. Continuum solvation models alone are scarcely used and account poorly for the specific interactions [155, 160], while the simplified model of a solute–solvent complex gives good results. No matter whether the polarizable continuum is part of the model, when the complex is modelled, it should be noted that the solute can either play the role of H-bond acceptor or act as a donor, or both, if it comprises different heteroatomic groups of H-bond acceptor and donor character, such as OH or NH2. On the other hand, the solvent itself, be it water or MeOH, to give but two examples, is prone to H-donor or acceptor behaviour, depending on its orientation relative to the solute molecule and/or the geometrical parameters of the latter. For a molecule that contains only a carbonyl group, an H-bond will be formed with one solvent molecule (Figure 4.11(a)). When a carbonyl and an amino group, i.e. both an H-donor and an acceptor, are present, and furthermore the amino group can form up to two H-bonds (as in Figure 4.11(c)), there are several solute–water complexes to be considered, from 1:1 to 1:3, in all possible combinations, but quantitative results are obtained for the 1:3 complex [161]. As regards the position of the solvent molecule relative to the plane of the solute, there are two possible geometries to be considered: in-plane and out-of-plane (see Section 4.2 for a thorough discussion). Note that, in Figures 4.11(a) and (b), in-plane H-bonds are formed, whereas in Figure 4.11(d) they are out-of-plane. Thus, choosing the number and orientation of solvent molecules implicated in the specific interactions is not straightforward and should be made on a case-by-case basis. Starting from these considerations, building of the theoretical model has to take some factors into account, such as the electronic structure of the solute and solvent molecules, the nature, number and topology of the H-bond acceptor or donor groups, the possible relative orientations of the solute and solvent molecules in order to find the optimal H-bond donor–acceptor ‘matches’, the choice of the quantum chemical method to treat the solute–explicit solvent subsystem and the choice of the continuum or statistical model for the bulk solvent. The final model should be a balance between a good correlation with the experimental available data and the computational resources needed. The literature comprises a multitude of solvation models, some of which consider the solvent explicitly, on a molecule-by-molecule basis, while others do it implicitly, as a polarizable continuum characterized by some global parameters. Here are some of the possibilities: QM treatment of a solute–solvent molecule(s) complex [162–166]; QM for the solute molecule with a continuum model for the solvent [155–157, 160]; QM for the solute which is placed in an electrostatic potential generated by the solvent [167, 168]; QM for a solute–solvent complex, combined with a continuum model for the bulk solvent [156, 157, 169, 170]; a QM/MM approach, in which the solute molecule or the solute–solvent complex is treated at the QM level and the bulk solvent within the framework of MM; MD for the solute–solvent complex [171] or in a solvent molecule box/droplet [172, 173]. Although the continuous solvation models take into account the polarity of the solvent, treating the solute–solvent specific interaction explicitly yields more complex and reliable results. Barone et al. [160] proved that the solvent shift for the FC state can be theoretically predicted with good accuracy for acetone in water only if a combined explicit/implicit model for the solvent is considered. If only the continuum model is used, the cyclohexane to water shift is 1184 cm1. When two water molecules are added to acetone in two different conformations, values of 2812 and 1550 cm1 are obtained, which are closer to the experimental value of 2000 cm1. It must be noted, however, that the explicit solvent molecules add to the system treated by a quantum chemical method, so their number is limited, especially for large solute molecules. The largest model includes the first solvation shell or all those molecules directly involved in the
Solute–Solvent Hydrogen Bond Formation in the Excited State 107 uracil 6
5 8
O
fluorenone
N
O
O
O
H O
2
3
H
H H
1N
4
H
H
O
H
H
O H
H
H O
CH3
(b)
(a) H3C
CF3
O H
H O H
H3C
N
O
O H
H
N
O CH3
in-plane H bonds
out-of-plane H bond
(c)
O
H
H O
H
H O H H
N
N N N H H
O
dibenzo[a,c]phenazine (d)
Figure 4.11 Models of some solute–solvent complexes: (a) 1:1 fluorenone–MeOH [163]; (b) 1:4 uracil–water [156]; (c) 1:3 C151–MeOH [161], the rectangle represents the plane of the solute; (d) 1:2 dibenzophenazine–water [219]. Solvent molecules in bold
specific interaction with the solute molecule [155–157, 170]. An example in which four water molecules are considered is given in Figure 4.11(b) for uracil [156]. The respective number and configuration of these molecules were assigned on the grounds of experimental data from techniques such as NMR, laser-induced fluorescence or theoretical MD results, which had indicated that no water molecules interact with C5 and C6 and that O7 and O8 interact with two and one water molecules respectively. 4.5.2 Computational methods The quantum mechanical part of the system is treated within the framework of methods usually used for excited states: the configuration interaction (CI) formalism [174], with single (CIS) or single and double
108 Hydrogen Bonding and Transfer in the Excited State
excitations (CISD); the complete active space self-consistent field (CASSCF) [175, 176] method, as such or combined either with a second-order perturbative correction, i.e. CASPT2 [177], or with a configuration interaction scheme, i.e. MRCI [178]; the time-dependent density functional theory (TDDFT) method [179]; coupled cluster-based methods (CC) [180]. As regards the bulk solvent, it can be mimicked by continuum solvation models as the polarizable continuum model (PCM) of Tomasi and Barone [152] and the conductorlike screening model (COSMO) [181]. A different approach is by statistical methods, such as MD or MC, which consider the bulk solvent explicitly, as a number (usually hundreds) of molecules, treated at a less accurate level of theory, MM [182]. The other is an intermediate between the continuum and MD models, which mimics the bulk solvent as an electrostatic potential generated by the atomic charges, i.e. the reference–interaction site model self-consistent field (RISM-SCF) method [183, 184]. The CIS method represents an acceptable compromise between the cost and accuracy for the excited-state geometry optimization, but no quantitative energies can be obtained. In recent years, the commonly used ab initio method for excited states has been the CASSCF method, for which solvation models are available [185]. CASPT2 and MRCI can treat transition energies with quantitative accuracy, and the first can be used to locate CIs [186–188]. The CC method was also found to give good results on transition energies for a large set of molecules, but, statistically speaking, the best theoretical estimate and closest to the experimental values remain the CASPT2 results [189]. In the last decade, the TDDFT method has gained tremendous importance in theoretical studies of excited states. It is an accurate and relatively fast method developed by Runge and Gross [179] and for which continuum models were adapted by the group of Barone [190]. As regards the excited states, both FC and relaxed geometries can be calculated, along with the above-mentioned molecular parameters. From the many functionals present in the literature, the PBE0 [191] functional gives the most reliable results for spectroscopic properties [192]. There are also documented shortcomings of TDDFT in the description of long-range interactions, charge transfer processes [193] and CI [194]. Nevertheless, being a relatively low-resourcedemanding method, it allowed the most complex models of the excited-state H-bond formation process to be constructed, comprising a large number of solvent molecules, up to the number found in the first solvation shell, which were treated quantum mechanically, to which the bulk solvent effect was added within the framework of the PCM model [156–158]. For thorough tests on the performance of different computational methods, the reader is referred to Refs [189], [195] and [196]. After the model and the method have been chosen, there are some parameters of interest for an H-bonded excited state that can be calculated: excitation energies, geometric parameters, atom charges, nature of the excited state, dipole moment, emission spectrum, oscillator strength, vibrational spectrum, H-bond length, H-bond energy. We will focus here only on the electronic and energetic properties of the excited singlet states, which can be correlated with the experimental electronic spectra. They characterize the excited-state photophysical properties or the H-bond and can be used to explain and complete the experimental data. There are two excited states that can be calculated: the FC and the fully relaxed S1 states. Although the latter is the one implied in the emission properties, there are authors that study only the FC state (especially for large systems) and draw conclusions on the emission properties, based on the different electron distribution in the excited state and its effect on the solute–solvent H-bond. Relaxation does not modify the electron distribution to a great extent, but the nature of the excited state does. Such results are not quantitative, but give an insight into the excited-state properties. Once the model and the method have been chosen, there are some computational procedures to be followed, depending on the complexity of the results needed. Let us suppose that the molecule is directly excited to the first excited singlet. Subsequent to excitation, the molecule arrives in the FC state, with the same geometry as the ground state. We can compute the vertical transition energy, denoted by Ea in Figure 4.12, at the optimized geometry of the ground state. It can be correlated with the absorption maximum. Another parameter is the
Solute–Solvent Hydrogen Bond Formation in the Excited State 109 S1 TS
E FC
S0/S1 CI
S1 Ea
Ef
S0
Figure 4.12 Schematic representation of the various electronic states important in the photophysical processes and of the radiative (solid arrows) and non-radiative (dashed arrows) deactivation pathways
emission transition energy, Ef, computed at the optimized geometry of the excited state and correlated with the emission maximum. In order to acquire some information on the decay processes, more features of the potential energy surface (PES) should be known. These include the critical points involved and the minimum energy path (MEP) that connects them. Apart from state minima, they are transition states (TS) on S1 and S1/S0 CIs, which are responsible for ultrafast non-radiative decay by IC to S0, as in Figure 4.12, where the nonradiative path is shown by dashed arrows. Various relative positions of the critical points on the PES are possible (e.g. S1 minimum on the path to the CI, no barrier, etc.). 4.5.3 H-bond-induced changes in the excited-state properties Excited-state optimization of the solute–solvent complex constitutes the simplest model for treating the H-bond explicitly. It is, in fact, the model most often used. As early as the 1970s, del Bene developed an MO theory of the H-bond in a series of some 20 articles (see, for instance, [197–199]), some being devoted to the solute–solvent H-bond in the excited state, but only as far as the FC state is concerned [198]. The author treated the solvent explicitly and calculated the vertical transition energies of some amides and their complexes with one or two water molecules at the CIS/STO-3G level. When the N–H group is the proton donor, the change in the transition energy is not significant. The blue-shift of the n–p band determined experimentally in protic solvents could be explained only by an H-bond formation in which the oxygen atom acts as a proton acceptor. A second H-bond formed with another water molecule, with nitrogen as a proton donor, determines an additive effect on the band shift. In this paper, the information on the excited-state H-bond is somehow indirect. As the blue-shift is very close in value to the calculated ground-state H-bond energy, the author concluded that its energy in the excited state is very small. Another argument for the weakening of the H-bond subsequent to excitation is the calculated atomic charge on the oxygen atom, which decreases owing to the electron redistribution during the n–p transition. We would like to conclude here that the net atomic charge on the atoms implicated and the bond energy are useful in characterising the specific interaction. On the other hand, when the solute molecule has multiple H-bond donor and/or acceptor centres, care must be taken as to the relative orientation of the solute and solvent molecule(s), in order to find the significant geometries. The presence of H-bonds can modify the nature of the excited state and its properties, as, for instance, the geometry. Dahiya et al. [75] considered the case of a diaminoanthraquinone for which the experimental Lippert–Mataga plots of Stokes shift, lifetime, fluorescence quantum yield and radiative and non-radiative constants indicated a different behaviour for polar solvents than for non-polar solvents. They modelled the polar solvent as eight water molecules located out-of-plane, in the vicinity of the two carbonyls and two amino
110 Hydrogen Bonding and Transfer in the Excited State
groups, and arranged two by two, above and below the plane of the solute. Eight H-bonds are thus formed. No details are presented as to why the exact number and relative geometry of the water molecules were chosen. By optimizing the solute in vacuo and in the presence of the water molecules, the authors find different planarity and charge distribution of the respective states, indicating the presence of a non-planar LE state in vacuo/nonpolar solvents and a planar ICT state in water/polar solvents. On the other hand, when the in-plane type of H-bond is considered [161] for C151 (see Figure 4.11 for the molecular structure), the geometry of the amino fragment in the excited state becomes more pyramidal owing to H-bond formation with MeOH (17.7 versus 5.7 in the isolated molecule for the angle that the NH2 fragment makes with that of the aromatic system). In the case of fluorenone in MeOH, where the carbonylic oxygen acts as an H-bond acceptor for one MeOH molecule, the length of the C¼O bond changes from 1.250 to 1.259 A [163]. The effect of solvent molecules on the geometry of the excited states can be more dramatic and even influence some photodissociation processes [171] or intramolecular proton transfer [200]. This is, for instance, the case for pyrrole in water, where N–H dissociation in the excited state is inhibited by the presence of water molecules, as experimental data indicate [201]. Franck and Damianos [171] studied pyrrole clusters formed with up to six water molecules by means of an MD method. They found theoretical proof that, owing to an electron transfer from pyrrole to the solvent (what is called a solvated electron) in the first excited state, the N–H distance is kept at a quasi-normal value of 1.1 A up to a simulation time of 0.4 ps, whereas for the isolated pyrrole the dissociation occurs at about 0.2 ps after the excitation. Moreover, the excited-state PES presents a minimum for the same 1.1 A N–H distance only in the presence of a water molecule. It seems that this stabilization constitutes a general mechanism, as was found for indole as well. As stated in the previous sections, for electron donor–acceptor molecular systems, an ICT excited state can be attained, which will be more polarized and by consequence stabilized in polar solvents. The extent of polarization can be monitored on the grounds of calculated dipole moments, but also by counting the location of frontier molecular orbitals implicated in the respective transition on different fragments of the molecule. We recall here C153 (see Figure 4.1), for which a difference between the dipole moment in excited and ground states of 6.0 D was found from the Lippert–Mataga plot. ATDDFT/PCM study [156] found a very good value of 6.1 D in CHX and 6.6 D in DMSO for the same property. Secondly, the first electronic transition has an essentially HOMO–LUMO character. While the HOMO is delocalized on the whole molecule, with significant contribution by the orbitals of the ‘central’ benzenic ring and of the nitrogen atom, the LUMO is mainly localized on the ‘quinone-like’ terminal ring, with a significant contribution of the p orbital of the carbonyl group. As a consequence, the S0 ! S1 transition has a partial intramolecular charge transfer character – from the nitrogen atom to the carbonyl group – and S1 has a partial zwitterionic character, with the nitrogen atom and the oxygen of the carbonyl group bearing formal positive and negative charges respectively. This results in a significant solvatochromic red-shift: 0.21 eV from CHX to DMSO, which correlates well with the experimental value. In the case of a 9-oxo-imidazopurine derivative containing a pyridine fragment as well (see Figure 4.13), a red-shift of the emission band was observed, along with a decrease in the fluorescence quantum yield and lifetime in alcohols [202]. In the meantime, if a phenyl replaces the pyridine, there are no significant changes in alcoholic media. Theoretical studies by means of the TDDFT method revealed that HOMO is located on the imidazopurine fragment, which is the electron donor, while LUMO is located mostly on the pyridine ring, the acceptor (Figure 4.13). That determines an ICT character for the first excited singlet, which is further proved by the dipole moment difference between the ground and excited states, Dm ¼ 15.2 D (in good correlation with the 11.6 D value found experimentally). The solute–solvent specific interaction was modelled as a 1:1 complex with one MeOH molecule located near the C¼O, N1, N4 and the pyridine N respectively. The bulk solvent effect was considered within the conductor-like PCM [203] model. By calculating the energy needed for such an H-bonded complex to be formed, both in the ground and excited state, the authors found that the pyridine N HO bond is the most stable and the only one that stabilizes in the excited state. This is not surprising
Solute–Solvent Hydrogen Bond Formation in the Excited State 111 O 1
N
9
N N
3
N H
5
N H
N4
O
O
N
N
N
N
N
N N H
N H
N H
N
N H
N
LUMO
HOMO
Figure 4.13 Molecular structure and frontier orbital location for a 9-oxo-imidazopurine derivative. Adapted with permission from [202]. Copyright 2006 American Chemical Society
considering the purine-to-pyridine charge transfer, resulting in an increase in the electronic density for the pyridine N, thus increasing its H-bond acceptor ability. So the solute–solvent H-bonding interaction in the first excited singlet state is responsible for the fast radiationless decay rates determined in protic solvents. Moreover, the nitrogen atom of the pyridine moiety is involved in the interaction, explaining the different behaviour of the phenyl-substituted analogue. Let us now consider the case of a merocyanine dye [169] (see Figure 4.14), which possesses an indolenine fragment as electron D and a barbituric acid fragment as A, linked by an alternant hydrocarbon bridge. It can be proved on theoretical grounds that it is an atypical D–A molecule, in which the p–p HOMO–LUMO transition decreases the polarization of the molecule. In MeOH solution, the localization of the electron in the HOMO is shifted towards the acceptor fragment compared with benzene. Not only the electron in the HOMO but also the distribution of other electrons vary with solvent polarity; therefore, the dipole moment of the S0 state is substantially increased with increasing solvent polarity from benzene (19.78 D) to MeOH (34.63). Similarly, LUMO migrates to the donor part. Consequently, the dipole moment in S1 decreases, i.e. 17.27 D in benzene and 24.16 D in MeOH. The more prominent decrease in MeOH can explain the blue-shift of the emission band in protic solvents found experimentally (0.155 eV) and theoretically (0.091 eV). A different type of CT state, more seldom studied, is the intermolecular CT state between a solute and a solvent molecule. The PET process is assisted by solute–solvent H-bond formation and results in a severe
-
O
11
N 13 9 6
22
14 7
23
21
20
17
12
4
16
N
5 26
24
N+
18
2
3
8
O
25 19
1
Figure 4.14
Merocyanine dye
10
O
15
112 Hydrogen Bonding and Transfer in the Excited State
fluorescence quenching. This was proved in the case of an oxazine on the grounds of electron transition contributions to the excited states and MO location for the 1:2 solute–EtOH complex [165]. This trimer is first excited to the S2 state (LE in nature) owing to the zero oscillator strength of S1 and then passes to S1, an intermolecular CT state, corresponding to a transition between an occupied MO located on the EtOH and a vacant MO located on the oxazine molecule. This functions as a ‘donor state’ for the ET process. It is a state that corresponds only to the H-bonded complex, and cannot be found for the solute molecule, which means that the ET process is indeed assisted by the H-bond formation. Specific interaction with solvent molecules determines changes in the energy gap between electronic states and sometimes even state interchanges. This is correlated with the greater stabilization of a strong H-bonded state versus a weak or non-H-bonded state. This state interchange was experimentally suggested for many compounds that possess non-bonding electron pairs and first excited singlets of n–p character in non-polar solvents [30, 104, 107]. This was proved for uracil [156], for which S1 has an n–p character in the gas phase, but, owing to the strong interaction with the water molecules, the p–p state lowers in energy and in the FC region becomes S1. Considering both four water molecules directly linked to the solute (see Figure 4.11) and the bulk solvent effect, it is possible to account for the Stokes shift (0.6 versus 0.8 eV found experimentally), the fluorescence energy (31 357 versus the experimental value of 32 051 cm1) and the state interchange in uracil. Neither by considering up to six water molecules [204] nor by using a PCM model alone [156] was this state interchange found. As regards the excitation energy, the most dramatic effect of specific interaction with the solvent is for the n–p state. In this type of interaction, the dipole moment value plays an important role. Because for the n–p state the dipole moment is lower than in the ground state, there will be a blue-shift of S1 in protic solvents, whereas S2 will be lowered in energy by the interaction with the solvent [204]. Another case is that of the TICT-forming compounds [83, 196, 205, 206]. Here again, the charge transfer was proved by monitoring the location of the frontier orbitals and the dipole moments in the ground and excited states. In Figure 4.15 the HOMO and LUMO orbitals of methyl p-dimethylaminobenzoate at the planar and twisted geometry are presented [207]. The charge separation appears only for the twisted conformation, confirmed by the large dipole moment in the TICT state (12.43 D versus 6.16 D in the LE state and 4.00 D in S0). Although there are many papers that deal with solvent effects on these twisted excited states, the great majority only consider polar aprotic solvents such as ACN, and in an implicit manner in any case [113, 115]. They were capable of showing that the excited state has a minimum on the PES corresponding to the twisted conformation (D–A torsion dihedral of 90 ) and that the shape of the PES is modified in polar solvents owing to greater relative stabilization of this twisted state, characterized by D–A charge separation and a large dipole moment compared with the LE state. However, studies of the well-known TICT-forming compound dimethylaminobenzonitrile (DMABN) and its complex with MeOH [208] revealed that H-bond formation may be involved in the photophysics of these systems. Comparing the calculated bond lengths and energies in the ground and excited states, we can see that the LE and TICT states behave differently, although in both cases the complexes are of the in-plane type. While the H-bond in the LE state becomes slightly weaker than in the ground state (21.12 versus 22.76 kJ mol1 respectively), it strengthens in the TICT state to 29.48 kJ mol1. This means that the TICT state is relatively stabilized by the H-bond formation, and in the meantime that the IC non-radiative deactivation process between the TICT and ground state is thus favoured, in accordance with the experimental data for this and other compounds [208]. The spectral features of H-bonded systems can be explained on the grounds of relative H-bond strength in different electronic states. Zhao and coworkers [209] studied two related thiocarbonyl compounds, i.e. thiocoumarin (TC) and 4H-1-benzopyrane-4-thione (BPT). They proved that for TC the H-bond is weakened in all the excited states, whereas for BPT it is weakened, too, except for the S2 state. At the same time, in the calculated emission spectrum, a blue-shift relative to the isolated molecule was observed for all the electronic states, with a weaker H-bond compared with the ground state, i.e. S1, S2, T1 for TC and S1 and T1 for BPT,
Solute–Solvent Hydrogen Bond Formation in the Excited State 113
N
N
O
O
O
O
LUMO
N
O
N
O
O
O
HOMO
Figure 4.15 Schematic representation of the frontier molecular orbitals for p-dimethylaminobenzoate: planar conformation (left), twisted conformation (right). Adapted with permission from [207]. Copyright Elsevier
except for the S2 state of BPT, for which the shift is to the red edge of the spectrum. Hence, a strengthening in the H-bond upon excitation determines a lowering of the excitation energy and thus a spectral red-shift, while H-bond weakening induces an increase in the excitation energy and thus a spectral blue-shift. Moreover, the energy gap between different electronic states with different H-bond strengths depends on the relative H-bond energy, so it was inferred that both photophysical and photochemical processes for this type of compound can be tuned by the intermolecular H-bond formation. The presence of solvent molecules can change both the photochemistry and the photophysics of the excited state. This can be best evidenced by calculating other critical points on the excited-state PES besides the minimum, as, for instance, TS. A somewhat exotic feature of excited states, but a feature that has gained increasing importance in elucidating their dynamics, is localization of CIs [186, 210, 211]. They are considered to provide an ultrafast non-radiative decay path to the ground state. Quite recently, conical intersections for solute–water complexes started to be studied [155–157, 212]. They can give important information on the mechanism of excited-state deactivation by non-radiative processes, in conjunction with
114 Hydrogen Bonding and Transfer in the Excited State
the calculation of the MEP followed by the molecule starting with the FC state, passing through the S1 minimum, a TS (if this is the case) and then deactivating to S0 through the CI. Sobolewski and Domcke [213] discussed generic mechanisms of excited-state deactivation via hydrogen atom dynamics in some model systems of isolated aromatic chromophores (indole), their complexes with amphoteric solvent molecules (indole–ammonia, pyridine–ammonia), H-bonded pairs of aromatic chromophores (indole–pyridine) and bifunctional intramolecularly H-bonded aromatic systems. The paper shows that solute–solvent H-bond formation in an ICT electronic state facilitates access to the S1/S0 CI, which determines ultrafast decay kinetics of the excited state. This mechanism may be implied in the concerted decrease in the fluorescence quantum yield and lifetime found experimentally in protic solvents for a large pool of CT states. 2-Aminopurine, an adenine analogue, has a great importance as a fluorescence probe in DNA. Unlike adenine, it presents high emission efficiency, with a fluorescence quantum yield varying from 0.01 in non-polar CHX up to 0.68 in water. Here also, the S0/S1 CI is responsible for its photophysical properties [108, 214]. To prove that, it is necessary to build the MEP in the first excited state, pp in character, which is directly accessed by the molecule subsequent to excitation [214]. Starting from the FC sate, the relaxation process to S1 minimum is barrierless, but, in order to reach the S0/S1 CI, the molecule passes through a TS, as in Figure 4.16. The calculated activation energy at the CASPT2 level in the gas phase is 2.4 kcal mol1 [108], whereas in aqueous solution, within the framework of the QM/MC method, it is 5.5 kcal mol1, over 3 kcal mol1 higher. So the solvent effect on the various critical points on the first excited singlet surface, implicated in the photophysical processes, resides in a greater stabilization of the S1 minimum compared with the TS, which leads to an increase in the energy barrier. In this way, reaching the CI is unfavoured by solvation, and thus the fluorescence quantum yield increases. A similar theoretical treatment can be applied to uracil and some of its derivatives [156, 157, 170, 212], which have a different experimental behaviour, i.e. short lifetimes in water and protic solvents, which were correlated with the H-bond formation ability of the solvent. The difference between uracil and 2-aminopurine is that S1 has an n–p character, so the initial excitation takes place to the bright S2 state, which has to be computed as well. The models used to study the effect of H-bond formation were: (1) different 1:1 complexes with water, for all possible types of H bond in uracil, at the MRCI level [212]; (2) 1:4 uracil–water complex embedded in a continuum, at the TDDFT level [156, 157, 170] (see Figure 4.11). The less demanding TDDFT method allowed for a more complex model to be used and for entire 1D and 2D regions of the hypersurfaces for the electronic states involved to be calculated. The results of the two models are in good correlation with one another and with experimental data. Here, the IC process is favoured by the presence of water owing to the TS
E
5.5
2.4
S0/S1 CI MIN
geometry change
Figure 4.16 Schematic representation of the relative energy of the S1 critical points in gas phase (dotted line) and in aqueous solution (continuous line) for 2-aminopurine. Adapted with permission from [108]. Copyright Elsevier. Activation energy in kcal mol1
Solute–Solvent Hydrogen Bond Formation in the Excited State 115
decrease in the barrier to the TS that links the S1 minimum with the S1/S0 CI. The second model confirmed the state interchange in the FC region in water compared with the gas phase, previously stated on the grounds of vertical transition energy and emission transition energy [156]. The PES shape is modified by specific interaction with water. Whereas uracil in the gas phase and ACN is excited to the p–p state, which crosses the np state in a region far from the FC state and then decays to S0, in water the crossing is in the FC region, as the p–p state is stabilized in water by solvation and the gap between the two states is very small. Nonetheless, of equally great importance in tuning the photophysical properties by solvation are the effects of substituents, in this case the substituent in position 5 [170]. Pyramidalization and out-of-plane motion of C5 is the predominant vibrational mode in passing from the S1 minimum to the CI, and in this way inductive and/or hyperconjugative effects of the substituent determine the barrier energy on this path. 4.5.4 Characterizing the H-bond The calculated molecular parameters that can give information on the H-bond are the relative solute–solvent geometry, the bond length, the charge distribution and the bond energy. Starting from the model(s) chosen, an optimization of the respective complex should be made in order to find the minimum conformations. We have already mentioned the possible types of H-bond theoretically found: (1) in-plane with the solute molecule, what would be called the ‘classical’, quasi-linear type; (2) out-of-plane. After optimization, for the in-plane geometry there are two conformations that correspond to a minimum: the all-planar complex and/or the complex in which the O–H fragment is coplanar to the solute molecule, but the rest of the solvent molecule is out-of-plane (see Figure 4.11). The two of them are quasi-isoenergetic, the most stable being the second one, as, for instance, in the case of fluorenone, where the energy difference is 1.7 kJ mol1 in the excited state, with an even larger difference in the ground state, 15.1 kJ mol1 [120]. The out-of-plane complex type predicted from experimental data [5, 6] was found on the grounds of theoretical results. This is the case of the indole–(H2O)2 complex. Fang [162] studied the ground- and excitedstate complexes of indole with one or two water molecules. In the 1:1 complex, the N–H group of indole is the H-bond donor, while the oxygen atom from the water molecule acts as the acceptor. It is an in-plane H-bond. In the 1:2 complex, besides this bond, a second bond is formed between the H-atom of a second water molecule and the p-electron cloud of the benzene ring in indole. Another case of an out-of-plane complex found to be a minimum in the first excited state is with thiocarbonyls such as TC and BPT [209]. Although in-plane conformers in which, again, only the OH group of MeOH is coplanar to the solute molecule are found in other electronic states (S0, S2, T1), S1 is characterized by a non-planar solute–solvent complex. The most important geometrical parameter is the length of the H-bond. The H-bond strength increases with decreasing bond length. It also increases with energy for the H-bond formation. This can be calculated by means of the equation Eb ðS1 Þ ¼ Ecomplex ðS1 ÞðEsolute ðS1 Þ þ Esolvent ðS0 ÞÞ
ð4:20Þ
where Eb is the energy required for H-bond formation, S1 stands for the first excited state and S0 for the ground state, Ecomplex(S1) is the calculated total energy of the optimized S1 state of the complex, Esolute(S1) is the energy of the optimized excited state of the isolated solute molecule and Esolvent(S0) represents the total energy of the equilibrium conformation of the isolated solvent molecule in its ground state. This formula refers to the case of locally excited states of the complex, where the solvent electronic structure is not perturbed by the excitation, which is entirely located on the solute. Whether this hypothesis is real or not can be decided on the grounds of the frontier molecular orbitals, which should both be located entirely on the solute and have no participation from the solvent. See Ref. [163] for a useful discussion. An alternative method is to calculate Eb as the dissociation energy of the complex [8, 195].
116 Hydrogen Bonding and Transfer in the Excited State Table 4.7 Calculated molecular parameters for the H-bond in the excited state. Ground-state values in parentheses
Compound/solvent
Stoichiometry
H-bond D/A groupa
H-bond typeb
Distance (A)
Energy (kJ mol1)
Method
Ref.
Formamide/water
1:1
C¼O
1
— (—)
— (26.94)
CIS
[198]
Fluorenone/MeOH
1:1
C¼O
1
1.903 (1.909) 1.802 (1.906)
42.3 (38.3) 42.62 (27.85)
CIS TDDFT
[120] [163]
Fluorenone/TFE
1:1
C¼O
1
1.803 (1.909)
— (47.5)
CIS
[120]
Dibenzophenazine/ water
1:1 1:2
N¼C (ring)
1 2
2.05 (2.19) 2.05 (2.21)
26.94 (20.29) 52.97 (39.20)
CASSCF
[217]
C151/MeOH
1:3
C¼O NH2 NH2 (D)
1 2 1
1.888 (1.908) 2.060 (2.017) 1.834 (1.925)
32.68 (23.95) 9.89 (11.17) 36.67 (27.34)
TDDFT
[161]
Indole/water
1:1 1:2
N–H (D) p-cloud
1 2
2.077 (2.087) 2.601 (2.719)
— (—) — (—)
CASSCF
[162]
TC/MeOH
1:1
C¼S
2
2.849 (2.442)
14.6 (24.35)
TDDFT
[210]
Cyano-Nmethylindoline/ water
1:1 1:1
CN N (ring)
1 2
2.24 (2.23) —c (2.16)
12.6 (12.5) —c (5.6)
CIS
[89]
a
H-bond acceptor if not otherwise (D) stated. 1 ¼ in-plane; 2 ¼ out-of-plane. c Unstable. b
H-bond distances and energies are presented in Table 4.7. In general, based on comparing the values of bond length and energy in ground and excited states, the most recent literature states that in-plane H-bonds with different types of acceptor or donor group strengthen upon excitation [120, 161, 163, 216]. In this way, the equilibrium in the excited state is shifted towards formation of the H-bonded species. The explanation of this strengthening is that the electronic distribution changes in the excited state, usually shifting towards the H-bond acceptor. So the charge density on the donor or acceptor heteroatom is very useful as a parameter to characterize the H-bond. Even if the most widely used solvents are water and MeOH, there are papers that consider some other solvents such as phenol [163] or TFE [120]. They deal with a coumarin derivative, C102, and fluorenone, respectively, both having a carbonyl as H-bond acceptor in the molecule. The type of conformation is in-plane, with the solvent molecule, other than the OH group, non-coplanar to the solute molecule. In both cases, again, the H-bond is strengthened in the excited state. So it could be concluded that this may be a general hypothesis for the in-plane solute–solvent H-bond complexes, at least as far as the carbonyl functions as the H-bond acceptor are concerned. Furthermore, it was demonstrated that the H-bond in complexes with TFE is stronger than that in the complex with MeOH, as found experimentally [120]. Regarding the out-of-plane complex, literature data indicate an opposite trend, the H-bond being weakened upon excitation, as can be seen in Table 4.7 from the values of bond length in the ground and excited states.
Solute–Solvent Hydrogen Bond Formation in the Excited State 117
Compare 2.060 versus 2.017 A for C151 in MeOH and 2.601 versus 2.719 A for indole in water, to take but two examples. Moreover, the H-bond energy is generally lowered in S1 compared with the ground state. For instance, it decreases from a value of 24.35 to 14.65 kJ mol1 upon excitation for TC in MeOH. Information on the bond length can be acquired by QM/MD combined studies in a different way. It is characterized by the distribution function of the H atom around the acceptor centre, as a function of the donorto-hydrogen distance. A pyrimidine in water MCSCF/MD study [218], which considered the pyrimidine molecule at the CASSCF level and more than 200 solvent molecules treated by MD, revealed that the N H distance is 1.9–2.0 A. Furthermore, it was possible to illustrate the polarization of the solute molecule by the surrounding medium, by way of calculating the induced dipole moment during the MD run, i.e. about 1 D. One of the advantages of this method is that it takes into account the statistically averaged possible conformations. This is also the case for another statistical method, QM/MC, which gives an average O H distance of 1.9–2.0 A for the acetone–450 water molecule system [219]. In the case of the 1,2,3-triazine–water system, all three N atoms are implicated in H-bonds, both in the ground and excited state [220, 221]. The distribution function of the H atom around all three N atoms presents a sharp maximum at around 1.8 and 2.0 A in the ground state. In the excited state there is a shallow peak for the N2–H distance of 2.6 A, meaning a very weak H-bond and a sharp peak for the H-bond located at the symmetric N1, N3 atoms at 1.8 A, indicating a strong, well-structured H bond. It is thus proved that both methods can describe H-bond formation in the ground and excited states of solute molecules in liquid media. A second geometrical issue found in the literature is that the in-plane bond length is smaller than the out-ofplane bond length. Considering the indole–water complexes again, it can be seen from Table 4.7 that the in-plane H-bond is 2.077 A in length, and the out-of-plane H-bond 2.601 A in length. The trend is maintained also for C151 and TC in MeOH. For cyano-N-methylindoline (CMI) in water and MeOH, both conformations are possible. Moreover, there are also in-plane stable complexes with non-linear geometry (see Ref. [89]), but weaker than the linear geometry both in the ground and excited state. Only the in-plane complexes are stable in the excited state, the out-of-plane bond being broken. On the basis of the respective H-bond energies, as well as bond length, the inplane conformation is obviously more stable for all the compounds. This points to the stronger interaction of the H atom with an H-bond acceptor on a quasi-linear geometry than with an n- or p-cloud on a nonlinear geometry. It was inferred that S1 ! S0 IC is favoured by coupling to the vibrational modes of the H-bond. Calculated values of different possible vibrations for CMI are relatively low (20–100 cm1) [89], thus increasing the density of states and consequently the probability of vibronic coupling between the ground and excited states. Another argument is the comparison with N-methylindoline, for which ultrafast non-radiative decay was not observed experimentally. The quantum chemical calculations show that the only H-bond possible is the weak out-of-plane one, which is broken in the excited state. Therefore, a direct correlation can be made between the formation of a stable H-bond in the excited state (CN OH in this case) and an effective IC deactivation pathway. To summarize, theoretical computations can explain experimental data and many times confirm hypotheses made by experimentalists. It is possible nowadays thoroughly to characterize various electronic states of molecules and the regions of the PES implied in the photophysical and photochemical processes, even for complex systems such as a solvated molecule involved in specific interactions with the solvent.
4.6 Conclusions A survey of literature data on excited-state H-bonds allows us to draw some general conclusions. As regards experimental data, two aspects can characterize the continuously increasing number of papers devoted to this subject. One aspect refers to H-bond studies performed on newly synthesized compounds. These papers contain steady-state fluorescence or time-resolved experiments with the aim of characterizing the photochemical
118 Hydrogen Bonding and Transfer in the Excited State Experimental part Solvent selection Recording of the absorption, excitation and emission spectra
Analysis of the data Identification of the H bonds; Separation of nonspecific and specific interaction
Molecular modelling Calculations of the optimized S0, S1 structures of the isolated molecule Analysis of the MO features Calculations for the electronic spectra Calculations of solute–solvent clusters Characterization of the H bonded species
Correlation of experimental and theoretical results
Change in the experimental conditions Change in the theoretical part
Figure 4.17 interaction
Main steps in designing a combined experimental and theoretical study on excited-state specific
properties and to evidence the presence of excited-state H-bonds. The main data used for assuming the presence of H-bonds are the spectral modifications in protic solvents as compared with non-polar or aprotic solvents, bathochromic shift of the emission band, changes in the lifetime and quantum yields. One of the problems in obtaining reliable experiments is a good design of the experiment, comprising the selection of the probes, solvents or mixtures of solvents, experimental excitation and emission conditions and temperature, in order to distinguish the H-bonding process from all the other possible processes occurring in the excited states. The other aspect is correlated with more precise studies on the already well-described fluorescence probes, aiming to obtain a deeper understanding of excited-state behaviour. Generally, the new studies are correlated with the dynamical aspects of the excited-state solvation process or with a more quantitative analysis of the excited-state kinetics. An increasing interest is being shown in the correlation of experimental data with molecular modelling results. In the last part of our chapter, a description is given of how computational complex methods that combine a QM treatment for the solute and the solvent molecules involved in specific interactions with continuum or statistical models to address the bulk solvent effect are applied to H-bond formation in different electronic states of solvated molecules. These theoretical calculations aim to obtain characterization of the excited-state geometry and nature of the solute–solvent ensemble and the H-bond energy, along with the energetic of the electronic states, which provide an insight into the photophysical processes that take place subsequent to excitation. As a general conclusion, the main steps reflecting the interplay between the experimental and theoretical approaches in the study of excited-state HB interactions are displayed in Figure 4.17.
Solute–Solvent Hydrogen Bond Formation in the Excited State 119
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. 38. 39. 40. 41. 42. 43. 44. 45. 46.
Z. R. Grabowski, K. Rotkiewicz, A. Siemarczuk et al., Nouv. J. Chim., 3, 443 (1979). W. Rettig, Angew. Chem. Int., 25, 971–988 (1986). W. Rettig and B. Zietz, Chem. Phys. Lett., 317, 187–196 (2000). K. Das, B. Jain and H. S. Patel, J. Phys. Chem. A, 110, 1698–1704 (2006). T. Yatsuhashi and H. Inoue, J. Phys. Chem., 101, 8166–8167 (1997). A. Morimoto, T. Yatsuhashi, T. Shimada et al., J. Phys. Chem. A, 2001, 105, 8840–8849. S. J. Grabowski, J. Phys. Chem., 105, 10 739–10 746 (2001), and references therein. Z. L. Cai and J. R. Reimers, J. Phys. Chem. A, 111, 954–962 (2007). J. G. Kirkwood, J. Chem. Phys., 2, 351–361 (1934); 7, 911–919 (1939). L. Onsager, J. Am. Chem. Soc., 58, 1486–1493 (1936). E. M. Kosower, J. Am. Chem. Soc., 80, 3253–3260 (1958). L. G. S. Brooker, A. C. Craig, D. W. Heseltine et al., J. Am. Chem. Soc., 87, 2443–2450 (1966). M. J. Kamlet, J. L. Abboud and R. W. Taft, J. Am. Chem. Soc., 99, 6027–6038 (1977). D. C. Dong and M. A. Winnik, Can. J. Chem., 62, 2560–2565 (1984). R. S. Drago, J. Chem. Soc. Perkin Trans., 2, 1827–1838 (1992). J. Catalan, V. Lo´pez, P. Perez et al., Liebigs Ann. 241–252 (1995). B. Siddlingeshwar and S. M. Hanagodimath, Spectrochim. Acta A, 72, 490–495 (2009). U. S. Raikar, C. G. Renuka, Y. F. Nadaf et al., Spectrochim. Acta A, 65, 673–677 (2006). M. Ravi, A. Samanta and T. P. Radhakrishnan, J. Phys. Chem., 98, 9133–9136 (1994). A. Kawski, Z. Naturforsch., 57a, 255–262 (2002). E. Lippert, Z. Naturforsch., 10, 541–546 (1955); Z. Electrochem., 61, 962–975 (1957). N. Mataga, Y. Kaifu and M. Koizumi, Bull. Chem. Soc. Jpn., 28, 690–691 (1955); 29, 465–470 (1956); N. Mataga and T. Kubota, Molecular Interactions and Electronic Spectra, Marcel Dekker, New York, NY, 1970. N. G. Bakshiev, Opt. Spectrosk. (USSR), 12, 564 (1962); 16, 821–832 (1964). E. G. McRae, J. Phys. Chem., 61, 562–572 (1957). A. Kawski, Acta Phys. Polon., 25, 258 (1964); A. Kawski and B. Pasztor, Acta Phys. Polon., 29, 187–193 (1966). A. Chamma, P. Viallet and C. R. Hebd, Seances Acad. Sci. Ser. France, 270, 1901 (1970). L. Bilot and A. Kawski, Z. Naturforsch., 18A, 621 (1962). P. Suppan, J. Lumin., 33, 29–32 (1985); J. Chem. Soc. Faraday Trans., 83, 495–509 (1987). P. R. Bangal, S. Panja and S. Chakravorti, J. Photochem. Photobiol. A: Chem., 139, 5–16 (2001). F. Han, L. Chi and W. Wu, J. Photochem. Photobiol. A: Chem., 196, 10–23 (2008). W. Verbouwe, L. Viaene, M. Van der Auweraer and F. C. De Schryver, J. Phys. Chem., 101, 8157–8168 (1997). S. Ionescu, D. Gavriliu, O. Maior and M. Hillebrand, J. Photochem. Photobiol. A: Chem., 124, 67–73 (1999). S. Nad and H. Pal, J. Phys. Chem. A, 105, 1097–1106 (2001). E. Krystkowiak, K. Dobek and A. Maciejewski, J. Photochem. Photobiol. A: Chem., 184, 250–264 (2006). M. Takara and A. S. Ito, J. Fluoresc., 15, 171–177 (2005). A. K. Satpati, M. Kumbhakar, S. Nath and H. Pal, Photochem. Photobiol., 85, 119–129 (2009). K. Dimroth, C. Reichardt, T. Siepmann and F. Bohlmann, Liebigs Ann. Chem., 661, 1–37 (1963); 727, 93–105 (1969). C. Reichardt, Solvents and Solvent Effects in Organic Chemistry. VCH Publishers, Weinheim, Germany (1988). C. Reichardt, Chem. Rev., 94, 2319–2358 (1994). C. Reichardt and E. Harbusch-Gornert, Liebigs Ann. Chem., 721–743 (1983); Z. B. Maksimovi, C. Reichardt and A. Spirt, Z. Anal. Chem., 270, 100–104 (1974). J. Dey and S. K. Dogra, J. Phys. Chem., 98, 9133–9136 (1994). J. Zou, Q. Yu and Z. Shang, J. Chem. Soc. Perkin Trans., 2, 1439–1443 (2001). A. Vlahovici, H. Ofenberg, L. Cires et al., J. Luminesc., 62, 227–236 (1994). S. P. Olivares, S. Risso and M. I. Gutierrez, Spectrochim. Acta A, 71, 336–339 (2008). I. Timtcheva, P. Nikolov, N. Stojanov and S. Minchev, J. Photochem. Photobiol. A Chem., 101, 145–150 (1996). P. Nikolov and I. Timtcheva, J. Photochem. Photobiol. A Chem., 131, 23–26 (2000).
120 Hydrogen Bonding and Transfer in the Excited State 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. 79. 80. 81. 82. 83. 84. 85. 86. 87. 88. 89. 90. 91. 92. 93. 94. 95. 96.
M. J. Kamlet, J. M. Abboud, M. H. Abraham and R. W. Taft, J. Org. Chem., 48, 2877–2887 (1983). J. Catalan, Handbook of Solvents, ed. by G. Wypych. ChemTech Publishing, Toronto, ON, p. 583 (2001). M. J. Kamlet and R. W. Taft, J. Am. Chem. Soc., 98, 377–386 (1976); 98, 2886–2894 (1976). B. Valeur, Molecular Fluorescence. Wiley-VCH, Weinheim, Germany, p. 204 (2002). J. Catalan and C. Dıaz, Liebigs Ann., 1941–1949 (1997). J. Catalan, C. Dıaz, V. Lo´pez et al., Liebigs Ann., 1785–1794 (1996). J. Catalan, C. Dıaz and F. Garcıa-Blanco, J. Org. Chem., 65, 9226–9229 (2000); 66, 5846–5852 (2001); Org. Biomol. Chem., 1, 575–580 (2003). P. M. Mancini, A. Bock, C. Adam et al., ARKIVOC, 373–381 (2003). H. R. Park, C. H. Oh, H. C. Lee et al., Photochem. Photobiol., 75, 237–248 (2002). G. J. F. Demets, E. R. Triboni, E. B. Alvarez et al., J. Spectrochim. Acta A, 63, 220–226 (2006). M. A. Mun˜oz, C. Carmona and M. Balo´n, Chem. Phys., 335, 43–48 (2007). Y. Altun, J. Sol. Chem., 33, 479–497 (2004). H. Demirelli, J. Sol. Chem., 34, 1283–1295 (2005). M. Jozefowicz, J. R. Heldt and J. Heldt, J. Chem. Phys., 323, 617–621 (2006). M. Umadevi, P. Vanelle, T. Terme et al., J. Fluoresc., 18, 1139–1149 (2008). M. A. Rauf, J. P. Graham, S. B. Bukallah and M. A. S. Al-Saedi, Spectrochim. Acta A, 72, 133–137 (2009). L. C. Zhou, G. J. Zhao, J. F. Liu et al., J. Photochem. Photobiol. A: Chem., 187, 305–310 (2007). A. A. Abdel-Shafi and S. S. Al-Shihry, Spectrochim. Acta A, 72, 533–537 (2009). K. M. Solntsev and D. Huppert, J. Phys. Chem. A, 102, 9599–9606 (1998). J. J. Aaron, Z. Mechbal, A. Adenier et al., J. Fluoresc., 12, 231–239 (2002). M. El-Sayed, T. Blaudeck, F. Cichos and S. Spange, J. Photochem. Photobiol. A: Chem., 185, 44–50 (2000). T. A. Fayed, Chem. Phys., 324, 631–638 (2006). T. A. Fayed, S. A. El-Daly, and S. E. H. Etaiw, J. Photochem. Photobiol. A: Chem., 121, 183–190 (1999). T. Gustavsson, L. Cassara, V. Guilbinas et al., J. Phys. Chem. A, 102, 4229–4245 (1998). R. S. Moog, D. L. Bankert and M. Maroncelli, J. Phys. Chem., 97, 1496–1501 (1993). T. Molotsky and D. Huppert, J. Phys. Chem. A, 106, 8525–8530 (2002). S. T. Abdel-Halim and M. K. Awad, J. Mol. Struct., 754, 16–24 (2005). M. A. Mun˜oz, C. Carmona and M. Balo´n, Chem. Phys. Lett., 393, 217–221 (2004). P. Dahiya, S. D. Choudhury, D. K. Maity et al., Spectrochim. Acta A, 69, 134–141 (2008). C. Carmona, M. Balo´n, M. Galan et al., Photochem. Photobiol., 76, 239–246 (2002). H. Pal, S. Nad and M. Kumbhakar, J. Chem. Phys., 119, 443–452 (2003). A. Barik, S. Nath and H. Pal, J. Chem. Phys., 119, 10 202–10 208 (2003). C. Cazeau-Dubroca, S. A. Lyazidi, P. Canbou et al., J. Phys. Chem., 93, 2347–2358 (1989). Y. H. Kim, H. W. Cho, M. Yoon et al., Chem. Phys. Lett., 264, 673–679 (1997). Y. Hirata, T. Okada and T. Nomoto, J. Phys. Chem. A, 102, 6585–6589 (1998). R. Chen, G. Zhao, X. Yang et al., J. Mol. Struct., 876, 102–109 (2008). A. Chakraborty, S. Ghosh, S. Kar et al., J. Mol. Struct., 917, 148–157 (2009). T. Fujii, M. Sano, S. Mishima and H. Hiratsuka, Bull. Chem. Soc. Jpn., 69, 1833–1839 (1996). M. Jozefowicz, J. R. Heldt, J. Karolczak, and J. Heldt, Z. Naturforsch., 58a, 144–156 (2003). L. Biczo´k, T. Berces and H. Linschitz, J. Am. Chem. Soc., 119, 11 071–11 077 (1997). R. Krolicki, W. Jarzeba, M. Mostafavi and I. Lampre, J. Phys. Chem. A, 106, 1708–1713 (2002). T. S. Singh and S. Mitra, J. Lumin., 127, 508–514 (2007). K. Pal, M. Kallay, G. K€ohler et al., Chem. Phys. Chem., 8, 2627–2635 (2007). G. Wiosna, I. Petkova, M. S. Mudadu et al., Chem. Phys. Lett., 400, 379–383 (2004). G. J. Zhao, J. Y. Liu, L. C. Zhou and K. L. Han, J. Phys Chem. B, 111, 8940–8945 (2007). J. Oshima, T. Yoshihara and S. Tobita, Chem. Phys. Lett., 423, 306–311 (2006). A. Morimoito, T. Yatsuhashi, T. Shimada et al., J. Phys. Chem. A, 105, 10 488–10 496 (2001). S. Uchiyama, K. Takehira, T. Yoshihara et al., Org. Lett., 8, 5869–5872 (2006). J. Seixas de Melo, R. S. Becker, F. Elisei and A. L. Macanita, J. Chem. Phys., 107, 6062–6069 (1997). J. L. Magalhaes, R. V. Pereira, E. R. Triboni et al., J. Photochem. Photobiol. A: Chem., 183, 165–170 (2006).
Solute–Solvent Hydrogen Bond Formation in the Excited State 121 97. 98. 99. 100. 101. 102. 103. 104. 105. 106. 107. 108. 109. 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. 138. 139. 140. 141. 142. 143. 144. 145. 146. 147.
E. Martyn, J. L. G. Coronado, J. J. Camacho and A. Pardo, J. Photochem. Photobiol. A: Chem., 175, 1–7 (2005). V. Bojinov, J. Photochem. Photobiol. A: Chem., 162, 207–212 (2004). T. Soujanya, T. R. S. Krishna and A. Samanta, J. Phys. Chem., 96, 8544–8548 (1992). M. E. Vazquez, J. B. Blanco and B. Imperiali, J. Am. Chem. Soc., 127, 1300–1306 (2005). T. C. Barros, G. R. Molinari, P. Berci Filho et al., J. Photochem. Photobiol. A: Chem., 76, 55–60 (1993). B. Ramachandram, G. Saroja, N. B. Sankaran and A. Samanta, J. Phys. Chem. B, 104, 11 824–11 832 (2000). A. P. Silva, H. Q. N. Gunaratne, T. Gunnlaugsson et al., Chem. Rev., 97, 1515–1566 (1997). E. Sikorska, I. V. Khmelinskii, J. L. Bourdelande et al., Chem. Phys., 301, 95–103 (2004). E. Sikorska, I. V. Khmelinskii, D. R. Worrall et al., J. Fluoresc., 14, 57–64 (2004). K. Sebok-Nagy, L. Biczok, A. Morimoto et al., Photochem. Photobol., 80, 119–126 (2004). L. Biczo´k, T. Berces and H. Inoue, J. Phys. Chem. A, 103, 3837–3842 (1999). V. Ludwig, M. Serrou do Amaral, Z. M. da Costa et al., Chem. Phys. Lett., 463, 201–205 (2008). J. L. Chen, R. Tang, Z. Luo and C. Yang, J. Mol. Struct., 917, 170–175 (2009). M. D. Bilokin, V. V. Shvadchak, D. A. Yushchenko et al., J. Fluoresc., 19, 545–553 (2009). N. Dash, F. A. S. Chipem, R. Swaminathan and G. Krishnamoorthy, Chem. Phys. Lett., 460, 119–124 (2008). G. Krishnamoorthy and S. K. Dogra, Spectrochim. Acta A, 55, 2647–2658 (1999). S. Mahanta, R. B. Singh, S. Kar and N. Guchhait, J. Photochem. Photobiol. A: Chem., 194, 318–326 (2008). R. B. Singh, S. Mahanta, S. Kar and N. Guchhait, Chem. Phys., 342, 33–42 (2007). R. B. Singh, S. Mahanta, S. Kar and N. Guchhait, J. Lumin., 128, 1421–1430 (2008). A. Elassyry, B. Benali, Z. Lazar et al., J. Mol. Liquids, 128, 46–49 (2006). S. Paudel, P. Nandhikonda and M. D. Heagy, J. Fluoresc., 128, 681–691 (2008). T. Yatsuhashi, Y. Nakajima, T, Shimada and H. Inoue, J. Phys. Chem. A, 102, 3018–3024 (1998). S. Songa, D. Ju, J. Li et al., Talanta, 77, 1707–1714 (2009). V. Samant, A. K. Singh, G. Ramakrishna et al., J. Phys. Chem. A, 109, 8693–8704 (2005). J. Herbich, M. Kijak, A. Zielinska et al., J. Phys. Chem. A, 106, 2158–2163 (2002). R. Cattana, J. Silber and J. Anunziata, Can. J. Chem., 70, 2677–2682 (1992). H. Shimada, A. Nakamura, T. Yoshihara and S. Tobita, Photochem. Photobiol. Sci., 4, 367–375 (2005). V. Sasirekha and V. Ramakrishnan, Spectrochim. Acta A, 70, 626–633 (2008). J. R. Heldt, J. Heldt, M. Jozefowicz and J. Kaminski, J. Fluoresc., 11, 65–73 (2001). M. Jozefowicz, J. R. Heldt and J. Heldt, Z. Naturforsch., 57a, 787–796 (2002). M. Jozefowicz, Spectrochim. Acta A, 67, 444–449 (2007). J. L. Jamison, L. Davenport and B. W. Williams, Chem.Phys. Lett., 422, 30–35 (2006). M. K. Nayak and S. K. Dogra, J. Photochem. Photobiol. A: Chem., 161, 169–183 (2004). J. Hidalgo, A. Sanchez-Coronilla, M. A. Mun˜oz et al., J. Lumin., 127, 671–677 (2007). A. Sanchez-Coronilla, M. Balo´n, M. A. Mun˜oz et al., Chem. Phys., 351, 27–32 (2008). P. Dahiya, M. Kumbhakar, T. Mukherjee and H. Pal, J. Mol. Struct., 798, 40–48 (2006). P. Dahiya, M. Kumbhakar, D. K. Maity et al., J. Photochem. Photobiol. A: Chem., 181, 338–347 (2006). S. Senthilkumar, S. Nath and H. Pal, Photochem. Photobiol., 80, 104–111 (2004). B. R. Gayathri, J. R. Mannekutla and S. R. Inamdar, J. Mol. Struct., 889, 383–393 (2008). M. Jozefowicz and J. R. Heldt, Chem. Phys., 294, 105–116 (2003). M. Jozefowicz, Spectrochim. Acta A, 71, 537–542 (2008). S. Yamamoto, T. Guobin, M. Nakamura and Y. Sueishi, Chem. Lett., 31, 760–761 (2002). G. Xie, M. Nakamura, Y. Sueishi and S. Yamamoto, J. Photochem. Photobiol. A: Chem., 162, 73–79 (2004). H. Inoue, M. Hida, N. Nakashima and K. Yoshihara, J. Phys. Chem., 86, 3184–3188 (1982). J. A. B. Ferreira and S. M. B. Costa, Chem. Phys., 321, 197–208 (2006). G. Burdzinski, A. Maciejewski, G. Buntinx et al., Chem Phys. Lett., 368, 745–753 (2003). G. Burdzinski, A. Maciejewski, G. Buntinx et al., Chem. Phys. Lett., 393, 102–106 (2004). P. Dahiya, M. Kumbhakar, T. Mukherjee and H. Pal, Chem. Phys. Lett., 414, 148–154 (2005). P. Dahiya, D. K. Maity, S. K. Nayak et al., J. Photochem. Photobiol. A: Chem., 186, 218–228 (2007). L. Wis´niewski, A. Urbanowicz, L. Jerzykiewicz et al., J. Mol. Struct., 920, 45–51 (2009). U. S. Raikar, C. G. Renuka, Y. F. Nadaf et al., J. Fluoresc., 16, 847–854 (2006).
122 Hydrogen Bonding and Transfer in the Excited State 148. M. Aleksiejew, J. Heldt and J. R. Heldt, J. Lumin., 129, 208–220 (2009). 149. V. V. Shynkar, A. S. Klymchenko, E. Pimont et al., J. Phys. Chem. A, 108, 8151–8159 (2004). 150. C. J. Cramer, Essentials of Computational Chemistry: Theories and Models. John Wiley & Sons, Ltd, Chichester, UK (2004). 151. F. Jensen, Introduction to Computational Chemistry. John Wiley & Sons, Ltd, Chichester, UK (1999). 152. J. Tomasi, B. Mennucci and R. Cammi, Chem. Rev., 105, 2999–3093 (2005) and references cited therein. 153. C. J. Cramer and D. G. Truhlar, Chem. Rev., 99, 2161–2200 (1999). 154. K. Sen, S. Basu and D. Bhattacharyya, Int. J. Quant. Chem., 102, 368–378 (2005). 155. S. Millefiori and A. Alparone, Chem. Phys., 303, 27–36 (2004). 156. R. Improta and V. Barone, J. Am. Chem. Soc., 126, 14 320–14 321 (2004). 157. F. Santoro, V. Barone, T. Gustavsson and R. Improta, J. Am. Chem. Soc., 128, 16 312–16 322 (2006). 158. V. Karunakaran, K. Kleinermanns, R. Improta and S. A. Kovalenko, J. Am. Chem. Soc., 131, 5839–5850 (2009). 159. I. Ciofini, C. Adamo and V. Barone, J. Chem. Phys., 121(14), 6710–6718 (2004). 160. M. Cossi and V. Barone, J. Chem. Phys., 112(5), 2427–2435 (2000). 161. Y. Liu, J. Ding, R. Liu et al., J. Photochem. Photobiol. A, 201, 203–207 (2009). 162. W. H. Fang, J. Chem. Phys., 111(12), 5361–5367 (1999). 163. G. J. Zhao and K. L. Han, J. Phys. Chem. A, 111(38), 9218–9223 (2007). 164. G. J. Zhao and K. L. Han, Biophys. J., 94, 38–46 (2008). 165. G. J. Zhao, J. Y. Liu, L. C. Zhou and K. L. Han, J. Phys. Chem. B, 111, 8940–8945 (2007). 166. K. R. F. Somers and A. Ceulemans, J. Phys. Chem. A, 108, 7577–7583 (2004). 167. T. Ishida and P. J. Rossky, J. Phys. Chem. A, 105, 558–565 (2001). 168. N. Minezawa and S. Kato, J. Phys. Chem. A, 109, 5445–5453 (2005). 169. A. Toutchkine, W. G. Han, M. Ullmann et al., J. Phys. Chem. A, 111(42), 10 849–10 860 (2007). 170. T. Gustavsson, A. Banyasz, E. Lazzarotto et al., J. Am. Chem. Soc., 128(2), 607–619 (2006). 171. I. Frank and K. Damianos, Chem. Phys., 343, 347–352 (2008). 172. Y. Mochizuki, Y. Komeiji, T. Ishikawa et al., Chem. Phys. Lett., 437, 66–72 (2007). 173. A. Cembran and J. Gao, Theor. Chem. Acc., 118, 211–218 (2007). 174. J. B. Foresman, M. Head-Gordon, J. A. Pople and M. J. Frisch, J. Phys. Chem., 96, 135–149 (1992). 175. B. O. Roos, Adv. Chem. Phys., 69, 339–445 (1987). 176. M. W. Schmidt and M. S. Gordon, Ann. Rev. Phys. Chem., 49, 233–266 (1998). 177. B. O. Roos, K. Andersson, M. P. F€ulscher et al., Advances in Chemical Physics: New Methods in Computational Quantum Mechanics. Vol. XCIII, ed. by I. Prigogine and S. A. Rice. John Wiley & Sons, Inc., New York, NY, pp. 219–331 (1996). 178. R. J. Buenker, S. D. Peyerimhoff and W. Butscher, Mol. Phys., 35, 771–791 (1978). 179. E. Runge and E. K. U. Gross, Phys. Rev. Lett., 52, 997–1000 (1984). 180. J. F. Stanton and R. J. Bartlett, J. Chem. Phys., 98(9), 7029–7039 (1993). 181. A. Klamt and G. Sch€u€urmann, J. Chem. Soc., Perkin Trans., 2, 799–805 (1993). 182. D. Frenkel and B. Smit, Understanding Molecular Simulation; Computational Science 1. Academic Press, London, UK (2002). 183. S. Ten-no, F. Hirata and S. Kato, Chem. Phys. Lett., 214, 391–396 (1993); H. Sato, F. Hirata and S. Kato, J. Chem. Phys., 105(4), 1546–1551 (1996). 184. F. Hirata (ed.), Molecular Theory of Solvation, in Understanding Chemical Reactivity Series, Vol. 24. Kluwer Academic Publishers, Dordrecht, The Netherlands (2004). 185. M. Cossi, V. Barone and M. A. Robb, J. Chem. Phys., 111(12), 5295–5302 (1999). 186. A. Sinicropi, T. Andruniow, L. De Vico et al., Pure Appl. Chem., 77(6), 977–993 (2005). 187. L. M. Frutos, T. Andruniow, F. Santoro et al., Proc. Natl Acad. Sci. USA, 104(19), 7764–7769 (2007). 188. L. Serrano-Andres, M. Merchan and A. C. Borin, Proc. Natl Acad. Sci. USA, 103(23), 8691–8696 (2006). 189. M. Schreiber, M. R. Silva, Jr, S. P. A. Sauer and W. Thiel, J. Chem. Phys., 128, 13 4110–13 4134 (2008). 190. R. Improta, V. Barone, G. Scalmani and M. J. Frisch, J. Chem. Phys., 125, 54 103–54 112 (2006). 191. J. P. Perdew, K. Burke and M. Ernzerhof, Phys. Rev. Lett., 77, 3865–3868 (1996); 78, 1396–1396 (1997). 192. C. Adamo, G. E. Scuseria and V. Barone, J. Chem. Phys., 111(7), 2889–2899 (1999).
Solute–Solvent Hydrogen Bond Formation in the Excited State 123 193. W. Koch and M. C. Holthausen, A Chemist’s Guide to Density Functional Theory, 2nd edition. Wiley-VCH Verlag GmbH, Weinheim, Germany (2001). 194. B. G. Levine, C. Ko, J. Quenneville and T. Martinez, J. Mol. Phys., 104, 1039 (2006). 195. Z. L. Cai and J. R. Reimers, J. Phys. Chem. A, 106(37), 8769–8778 (2002). 196. C. Jamorski, J. B. Foresman, C. Thilgen and H. P. Luthi, J. Chem. Phys., 116(20), 8761–8771 (2002). 197. J. E. Del Bene, Chem. Phys., 15(3), 463–472 (1976). 198. J. E. Del Bene, J. Am. Chem. Soc., 100(5), 1395–1398 (1978). 199. J. E. Del Bene, J. Comput. Chem., 2(2), 200–206 (1981). 200. A. Kyrychenko and J. Waluk, J. Phys. Chem. A, 110, 11 958–11 967 (2006). 201. H. Lippert, V. Stert, L. Hesse et al., Chem. Phys. Lett., 376(1–2), 40–48 (2003). 202. G. Wenska, J. Koput, T. Pedzinski et al., J. Phys. Chem. A, 110(38), 11 025–11 033 (2006). 203. V. Barone and M. Cossi, J. Phys. Chem. A, 102, 1995–2001 (1998). 204. C. M. Marian, F. Schneider, M. Kleinschmidt and J. Tatchen, Eur. Phys. J. D, 20, 357–367 (2002). 205. C. Jamorski Jodicke and H. P. Luthi, J. Chem. Phys., 117(9), 4146–4156 (2002). 206. C. Jamorski Jodicke and H. P. Luthi, J. Chem. Phys., 117(9), 4157–4167 (2002). 207. M. Aleksiejew, J. Heldt and J. R. Heldt, J. Lumin., 129, 208–220 (2009). 208. G. J. Zhao and K. L. Han, J. Comput. Chem., 29(12), 2010–2017 (2008). 209. Z. R. Grabowski and K. Rotkiewicz, Chem. Rev., 103, 3899–4031 (2003) and references cited therein. 210. G. J. Zhao and K. L. Han, ChemPhysChem, 9, 1842–1846 (2008). 211. B. G. Levine and T. Martınez, J. Annu. Rev. Phys. Chem., 58, 613–634 (2007). 212. A. L. Sobolewski and W. Domcke, Eur. Phys. J. D, 20, 369–374 (2002). 213. A. Yoshikawa and S. Matsika, Chem. Phys., 347, 393–404 (2008). 214. A. L. Sobolewski and W. Domcke, J. Phys. Chem. A, 111(46), 11 725–11 735 (2007). 215. L. Serrano-Andres, M. Merchan and A. C. Borin, Proc. Natl Acad. Sci. USA, 103, 8691 (2006). 216. S. Coussan, A. Bach and S. Leutwyler, J. Phys. Chem. A, 104, 9864–9873 (2000). 217. D. Dey, A. Bose, D. Bhattacharyya et al., J. Phys. Chem. A, 111, 10 500–10 506 (2007). 218. M. E. Martin, M. L. Sanchez, M. A. Aguilar and F. J. Olivares del Valle, J. Mol. Struct. (Theochem.), 537, 213–222 (2001). 219. K. Coutinho and S. Canuto, J. Mol. Struct. (Theochem.), 632, 235–246 (2003). 220. D. Xie, X. Ma and J. Zeng, Chem. Phys. Lett., 368, 377–383 (2003). 221. J. Zeng and D. Xie, J. Comput. Chem., 25(6), 813–822 (2004).
5 Electronic-Excited-State Structures and Properties of Hydrated DNA Bases and Base Pairs Manoj K. Shukla and Jerzy Leszczynski NSF CREST Interdisciplinary Nanotoxicity Center, Department of Chemistry and Biochemistry, Jackson State University Jackson, Mississippi 39217, USA
5.1 Introduction Computational chemistry methods have been established as an attractive and reliable alternative to the costly and time-consuming experiments used to determine structures, properties and reactivities of molecular species. Although the modelling of the exact environmental conditions at the high theoretical level is still very expensive and may not be possible in several cases, computational methods are capable of providing reliable predictions in many ways, which can be useful to experimentalists and to the general scientific community [1, 2]. Theoretical methods are especially attractive in areas such as the determination of excitedstate geometries of complex molecules, where experiments are not yet possible. Another example of the role of computational studies is the quantitative prediction of amino group pyramidalization of nucleic acid bases. The neutron diffraction study of adenine crystals has suggested a non-planar amino group [3]. Using the ab initio quantum chemical method, the amino groups of bases were suggested to be non-planar more than a decade ago [4, 5]. However, only recently, an experimental method has verified such non-planarity in the gas phase of adenine and cytosine, using the measurement of the vibrational transition moment angles [6]. We strongly believe that theoretical and experimental methods are complementary to each other, and a judicious decision is needed for their efficient applications to determine the structures and properties of different systems. Hydrogen bonding is ubiquitous. It plays an important role in different aspects of all living organisms. Hydrogen bonds can be classified as weak, moderate and strong, depending upon their energies [7]. Hydrogen
Hydrogen Bonding and Transfer in the Excited State, Volume I & II © 2011 John Wiley & Sons, Ltd. ISBN: 978-0-470-66677-7
Edited by Ke-Li Han and Guang-Jiu Zhao
126 Hydrogen Bonding and Transfer in the Excited State
bonds can also express unique features that allow them to be classified as blue-shifted hydrogen bonds and dihydrogen bonds [8–11]. In blue-shifted hydrogen bonds, the stretching vibration frequencies associated with CH or NH vibrations are blue-shifted with respect to the corresponding isolated species after hydrogen bond formation. In dihydrogen bonds the two hydrogen atoms are hydrogen bonded to each other. The hydrogen bonds formed between the complementary purine (adenine and guanine) and pyrimidine (thymine and cytosine) bases (see Figure 5.1 for the structures of bases and base pairs) are responsible for the double helical structure of deoxyribonucleic acid (DNA). Importantly, the specific sequences of these hydrogen-bonding patterns in DNA define the genetic code, the carrier of heredity. Alteration in DNA structure may lead to mutation by producing a permanent change in the genetic code. The exact cause of mutation is not known, but several factors, e.g. environment, irradiation, etc., may contribute towards such phenomena. The formation of pyrimidine dimers between adjacent thymine bases on the same strand results in the most common UVinduced DNA damage. Based on a femtosecond time-resolved IR spectroscopic study of thymine oligodeoxynucleotide (dT)18 and thymidine 50 -monophosphate (TMP), Kohler and coworkers [12] have recently shown that thymine dimerization is an ultrafast process that usually occurs in the femtosecond timescale, where the formation of photodimer from the initially excited singlet pp state of thymine is barrierless. However, proper geometrical orientation of stacked thymine pairs is the necessary requirement for the formation of the photodimer. H62
H61
O6
N6
N7
N1
C5
C2
C4
N1 C8 H8
H2
C6
H1
C6
C8
H21 C2
C4
N2
N9
H8 N9
N3
N3 H9
N7
C5
H22
H9
Adenine (A)
Guanine (G) H42
H41
O4
N4
C4
H3 N3
R5 C4
C5 N3
C2 O2
C6
H6
C5
C2
N1 O2
H5
C6 N1
H6
H1 H1
Thymine (T)/Uracil (U)
Cytosine (C)
Adenine-Thymine (Uracil) Base Pair
Guanine-Cytosine Base Pair
R5’
Figure 5.1 Structures and atomic numbering schemes of nucleic acid bases and Watson–Crick base pairs. In thymine, R5/R50 ¼ CH3, and in uracil, R5/R50 ¼ H
Electronic-Excited-State Structures and Properties of Hydrated DNA Bases and Base Pairs
127
Nucleic acid bases in DNA exist in their respective canonical tautomeric forms. However, minor tautomers, when formed for some reason, will induce mispairing among bases, and this can cause mutation if left unrepaired. In nucleic acids, the number of minor tautomers is reduced owing to the presence of sugar at the N9 and N1 sites of purines and pyrimidines respectively. However, the possibilities of keto–enol and amino–imino tautomerism remain there. The tautomeric analysis of nucleic acid bases, especially guanine, is further complicated by recent experimental and theoretical results where comparatively less stable imino tautomers of guanine have been assigned in the supersonic jet-cooled spectra [13–22]. Water is one of the most important ingredients of our planet and an essential part of our life. In vivo, DNA is heavily hydrated, and the percentage relative humidity determines the degree of hydration. Further, it plays a prominent role in determining the three-dimensional structures and functions of nucleic acids [23, 24]. Nuclear magnetic resonance (NMR) investigations have suggested that the relative humidity and temperature control the movement of the backbone as well as the bases in nucleic acids [25–27]. Schneider et al. [28] have performed an extensive analysis of crystallographic data of hydration of DNA bases. They found that, in general, the degree of hydration depends upon the types of DNA, the location of the minor and major grooves and the bases. Some minor tautomers of nucleic acid bases were found to become more stable under hydration [29–32]. Further, water molecules were also found to increase stacking interaction in base pairs [33– 35]. Our group has shown that the presence of a water molecule in the proton transfer reaction path of the keto–enol tautomerization reaction of nucleic acid bases and related molecules reduces the proton transfer barrier height significantly [29, 36–39]. The transition states of such water-assisted proton transfer reactions have been found to have a zwitterionic structure [37–39]. Transfer of a proton corresponding to the keto–enol tautomerization of such hydrated species is characterized by the collective process. The investigation of polyhydration of adenine, uracil, thymine and cytosine with 11–16 water molecules in the ground state shows significant geometrical deformation of bases [40–42]. Korter et al. [43] have performed an experimental investigation on the interaction of a water molecule with the N–H site of indole, which could be considered as a simplified model of a DNA base, in the ground and the lowest singlet excited state, and found that, consequent to electronic excitation, the position and orientation of the water molecule were changed with respect to the ground state. Nucleic acid bases have an ultrashort lifetime, of the order of subpicoseconds [44]. Although they absorb ultraviolet (UV) radiation efficiently, nucleic acid bases have a very poor radiative quantum yield. Most of the absorbed energy is dissipated in the form of heat mainly through the conical intersection of the groundand excited-state potential energy surfaces [44–48]. Several theoretical calculations supplemented with experimental data have suggested that structural non-planarity in bases facilitates the conical intersection among different electronic states, thus providing paths for efficient non-radiative deactivation [44, 45]. Recently, we have found that the electronic singlet excited-state geometry of guanine significantly depends on the mode and degree of hydration [49, 50]. It was also concluded that excited-state dynamics of guanine will also depend upon the mode and degree of hydration. In another recent combined experimental and theoretical study, the dynamics of isolated thymine was revealed to be different from that of the hydrated form [51]. Recently, several review articles have described the excited-state structures, properties and mechanism of non-radiative deactivations in nucleic acid fragments using both experimental and theoretical techniques [44–46, 52–57]. An excellent collection of lucid descriptions of excited-state phenomena in nucleic acid fragments, including recent developments of theoretical methods to study the excited states of a variety of molecules, can be found in a recent book entitled Radiation Induced Molecular Phenomena in Nucleic Acids [1]. This chapter describes the recent investigation of excited-state structures of nucleic acid fragments under a hydrogen-bonding environment. Brief information on ground- and excited-state properties of isolated nucleic acid bases is also presented.
128 Hydrogen Bonding and Transfer in the Excited State
5.2 Ground-State Structures of Nucleic Acid Bases and Base Pairs Nucleic acid bases can exist in various tautomeric forms, and such tautomeric distribution strongly depends upon the environment. In nucleosides and nucleotides, the N9 $ N7 prototropic tautomerism in purines (adenine and guanine) and the N1 $ N3 prototropic tautomerism in pyrimidine (cytosine) is blocked. However, bases can still show keto–enol and amino–imino tautomerism. The scope of the present review is not to discuss the ground-state properties of DNA fragments in detail. Therefore, ground-state properties are only very briefly discussed here. For detailed discussion of ground-state structures and properties of nucleic acid fragments, the reader is referred to review articles in this area of research [2, 8, 29, 58, 59]. Theoretically, it has been found that the six-membered ring of NABs has significantly large conformational flexibility [60, 61]. It is well known that the amino groups of NABs are non-planar, and such pyramidalization occurs owing to the partial sp3 hybridization of the amino nitrogen. Among NABs, guanine exhibits the largest degree of pyramidalization [4, 29, 58]. Dong and Miller [6] have indicated experimentally the pyramidal nature of the amino group in adenine and cytosine in the gas phase. Recent experimental and theoretical analysis has really complicated our understanding of tautomeric distributions of guanine in particular, and therefore nucleic acid bases in general, under different environmental conditions [13–22]. These complications arise owing to the presence of relatively less stable imino tautomers of guanine under the supersonic jet-cooled condition [16]. Initially, based upon the results of resonance-enhanced multiphoton ionization spectroscopy, the existence of the keto-N9H, keto-N7H, enolN9H and enol-N7H tautomeric forms of guanine has been suggested [13, 14]. However, Choi and Miller [15] have assigned the keto-N9H, keto-N7H and cis and trans forms of the enol-N9H tautomer of guanine trapped in helium droplets. This assignment was based on the comparison of the infrared (IR) data of guanine trapped in helium droplets with theoretically computed vibrational frequencies of guanine tautomers at the MP2 level using the 6-311 þþ G(d, p) and aug-cc-pVDZ basis sets. Mons et al. [16] reassigned their previous R2PI data and found that the enol-N9H-trans, enol-N7H and two rotamers of the keto-N7H-imino tautomers of guanine are present in the supersonic jet-cooled beam. The imino tautomers of guanine are about 8.0 kcal mol1 less stable than the most stable keto-N7H tautomer in the gas phase at the MP2/6-311 þþ G(d, p)//B3LYP/6311 þþ G(d, p) level [17]. Recent experimental and theoretical investigations suggested the presence of three tautomers, namely N9H, N7H and N3H, of adenine in the dimethylsulfoxide solution, with N9H being the major tautomer, while N7H and N3H are the minor tautomeric forms [62]. However, in earlier experimental investigations, the presence of only N9H and N7H tautomers of adenine has been suggested [63–65]. The N9H form was the major tautomer, while the relative population of the minor N7H form was found to be environmentally dependent [63–65]. Recent theoretical study shows that the N7H and N3H tautomers of adenine have a similar stability, and the N9H tautomer represents the global minimum [66, 67]. Cytosine is the only pyrimidine base that has several tautomeric forms. For example, it is present as a mixture of amino-oxo (N1H) and amino-hydroxy forms in the argon and nitrogen matrices, but tautomeric equilibrium is shifted towards the latter form [68, 69]. Microwave spectroscopic investigation has shown the presence of three (amino-oxo, imino-oxo and amino-hydroxy) tautomers of cytosine [70]. However, in water solution only amino-oxo forms (N1H and N3H) have been revealed [71]. Nir et al. [72] have shown the existence of keto and enol tautomers of jet-cooled cytosine. The imino-oxo tautomer has been suggested for 1-methyl and 5-methylcytosine in a matrix isolation study [73, 74]. However, in the crystal environment, only the amino-oxo-N1H form is present [75]. Theoretically, various methods up to the CCSD(T) level of theory have been used to determine the relative stability among different tautomers of cytosine [31, 76]. Thymine and uracil exist mainly in the oxo-tautomeric form [29, 45, 58, 77]. The Watson–Crick (WC) base pairs are found to be planar at the HF and DFT levels, including the amino group [29, 58, 78–81]. However, amino groups of the WC AT and GC base pairs are found to be pyramidal at the electron-correlated MP2 level [80, 82]. Interestingly, the amino group of the WC AT base pair at the MP2
Electronic-Excited-State Structures and Properties of Hydrated DNA Bases and Base Pairs
129
level was found to be non-planar only at the lower basis set, but planar at the larger basis set. The non-planarity in the GC base pair has been suggested to increase the stacking interaction and may increase the stability of the helix [82]. The geometries of some reverse Watson–Crick (RWC), Hoogsteen (H) and reverse Hoogsteen (RH) base pairs have also been computed to be non-planar [5, 29, 58]. In addition, the energetics of hydrogenbonded and stacked base pairs were studied up to the CCSD(T) level of theory [83–87]. Several theoretical and experimental investigations were also performed to determine the electron affinity, ionization potential and protonation and deprotonation properties of purine and pyrimidine bases and base pairs [88–103]. Based on the experimental and theoretical data, the adiabatic valence electron affinity for pyrimidine bases has been estimated to be in the range 0–0.2 eV, while those for adenine and guanine amount to about 0.35 and 0.75 eV respectively [88]. In general, purines have lower and pyrimidines have higher ionization potentials [89–94]. Guanine has the lowest ionization potential among NABs. Therefore, guanine is the most susceptible among NABs to one electron oxidation under irradiation. Podolyan et al. [95] computed proton affinities of all nucleic acid bases up to the MP4(SDTQ) level and found that the computed proton affinities are very close to the experimental data. Schaefer and coworkers [101] have recently investigated the structures and properties of deprotonated GC base pairs. Kumar et al. [102, 103] have recently investigated the adiabatic electron affinities of GC, AT and hypoxanthine–cytosine base pairs at the DFT level. These authors have found that polyhydration leads to significant increase in the electron affinity of the AT base pair.
5.3 Excited-State Structures of Nucleic Acid Bases 5.3.1 Electronic transitions in nucleic acid bases In general, the absorption bands of purines near 260 and 200 nm are composed of two transitions whose transition dipole moments are non-parallel [48, 104, 105]. Usually, guanine shows five transitions near 4.51, 4.96, 5.51, 6.08 and 6.59 eV (275, 250, 225, 204 and 188 nm) in the UV region [47, 48, 106–108]. Clark [107] has tentatively suggested the existence of the np transitions near 5.21, 6.32 and 7.08 eV (238, 196, and 175 nm) in guanine. The spectral origin of the first electronic singlet pp transition of guanine tautomers is measured using R2PI spectroscopy in the laser-desorbed jet-cooled beam [13, 14, 16]. The reassigned R2PI spectra [16] have suggested the spectral origin of enol-N9H-trans, keto-N7H-imino-cis, keto-N7H-imino and enol-N7H tautomers [17] at 4.31, 4.20, 4.12 and 4.07 eV respectively. Adenine also shows several electronic transitions in the UV region. A photoacoustic spectroscopic study has shown four electronic transitions in the region 180–300 nm [109]. The main absorption peak of adenine at 4.77 eV (260 nm) generally shows two components – the stronger one at 4.75 eV (261 nm) is short-axis polarized, while a weak shoulder near 4.64 eV (267 nm) is long-axis polarized [110]. The existence of np transitions near 5.08 and 6.08 eV was suggested in the crystal of 20 -deoxyadenosine [111] and near 5.38 eV in the stretched polymer film of 9-methyladenine [112]. Based upon an REMPI investigation, Kim et al. [113] have suggested that the first electronic transition of adenine with the spectral origin at 35 503 cm1 (281.7 nm, 4.40 eV) has an np character, while the second transition with the spectral origin at 36 108 cm1 (276.9 nm, 4.48 eV) has a pp character. However, the np assignment was not supported by Luhrs et al. [114], and these authors have speculated the involvement of some other tautomer of adenine. Luhrs et al. [114] have predicted that the spectral origins of the first pp transition of adenine and 9MA are located at 36 105 cm1 (277 nm, 4.48 eV) and 36 136 cm1 (276.7 nm, 4.48 eV) respectively, and these results are in accordance with the observation made by Kim et al. [113]. The R2PI study by Nir et al. [115] on laser-desorbed adenine has also provided similar results. The absorption spectrum of cytosine is generally solvent dependent and shows peaks or shoulders near 4.66, 5.39, 5.85 and 6.29 eV [48, 116–123]. There is significant experimental and theoretical evidence for the existence of an np transition near 5.3 eV (232 nm) in cytosine [45]. Zaloudek et al. [120] have suggested the existence of
130 Hydrogen Bonding and Transfer in the Excited State
another np transition near the 5.6 eV (220 nm). The spectral features of uracil and thymine are generally similar, and both bases have absorption bands near 4.77, 6.05 and 6.89 eV (260, 205 and 180 nm respectively) [44–48]. Several investigations have suggested the presence of an np transition within the first absorption envelope of uracil, thymine and their analogues [48, 109, 124, 125]. In the gas phase and in an aprotic solvent, the np state is the lowest, but in the protic environment it has higher energy than that of the pp state [48, 124, 125]. Transition moment directions in nucleic acid bases have also been assigned by Clark and other researchers in series of investigations [107, 112, 120, 126–128]. Various theoretical methods such as CASPT2/CASSCF [129–131], TDDFT [132–143], coupled cluster [144–147] and CIS [47, 138, 148–152] have been used to compute electronic transition energies of nucleic acid bases, base pairs and their stacked complexes. In one of the TDDFT calculations, several sets of diffuse functions were also used [139]. Computed transition energies were generally found to be in good agreement with the corresponding experimental data. Detailed analysis of experimental and theoretical electronic transitions of nucleic acid bases can be found in a recent review article [45]. Ritze et al. [153] have studied the effect of base stacking on the electronic transitions at the SAC-CI and RI-CC2 levels of theory by considering cytosine–cytosine and thymine–thymine stacked dimers in the A- and B-DNA configuration. It was predicted that the spectral splitting in thymine–thymine stacked dimers in the A-DNA is significantly (6 times) larger than that in the B-DNA. 5.3.2 Excited-state geometries of nucleic acid bases The N9H and N7H tautomers of adenine have planar ground-state geometry. However, the amino groups of both tautomers show pyramidal character, and such pyramidalization is more pronounced for the N7H form. The N9H tautomer has an almost planar structure in the electronic lowest singlet pp excited state, while the N7H tautomer has non-planar geometry in the same state. However, the amino group has been predicted to be pyramidal for both tautomers in the excited state [47]. In the electronic lowest singlet np excited state, the N9H tautomer has a non-planar structure, and such non-planarity is localized around the N1C2N3 fragment of the six-membered ring. For the N7H tautomer in the electronic lowest singlet np excited state, the molecular geometry is reminiscent of twisted intramolecular charge transfer states [151, 154, 155]. The geometry of the molecule has Cs symmetry, and amino hydrogens make dihedral angles of 61 with respect to the ring plane. However, no significant intramolecular charge transfer was revealed for this tautomer in the considered state [47]. The geometries of both tautomers were also optimized in the ground and electronic lowest singlet pp excited state under hydrated conditions where three water molecules were in the first solvation shell. It was revealed that water molecules induce planarity in the system. Consequently, the ground- and electronic lowest singlet pp excited-state geometries of both tautomers were found to be almost planar, including the amino group [47]. Significant non-planarity around the C6N1C2N3 fragment was revealed for the keto-N9H tautomer of guanine in the electronic lowest singlet pp excited state [47]. In the electronic lowest singlet np excited state, which is characterized by the excitation of the carbonyl group lone pair electron, the C6O6 bond length was predicted to be increased by about 0.1 A with respect to the ground-state value. Further, in this state the O6 and H1 atoms were found to be displaced away from the ring plane and located opposite to each other. The geometrical distortions in the keto-N7H tautomer in the excited states were computed similarly to the keto-N9H tautomer, but the amount of distortion was generally smaller than that in the keto-N9H tautomer [47]. The excited-state hydration of guanine will be discussed in detail in the next section. Ground state geometry of the keto-N1H tautomer of cytosine is planar, while that in the electronic lowest singlet pp excited state was found to be non-planar, mainly around the N1C6C5C4 fragment. The amino group is pyramidal in both states. In the electronic lowest singlet np excited state, the amino group was revealed to be considerably rotated, and the N3 atom was located appreciably out-of-plane [148, 152]. The structural deformation in the np state was attributed to the excitation of the N3 lone pair electron. The rotation of the
Electronic-Excited-State Structures and Properties of Hydrated DNA Bases and Base Pairs
131
amino group was speculated to be due to the partial contribution of orbitals due to the amino lone pair electron in the electronic excitation. Under hydration with three water molecules, the amino group was found to be more planar. The mode of hydration in the electronic lowest singlet np excited state was predicted to be completely modified. The N3 site, the lone pair electron of which was responsible for the np excitation, was revealed to provide a repulsive potential to hydrogen bonding [152]. The ground-state amino group pyramidalization in the keto-N3H tautomer of cytosine was revealed in a more pronounced manner compared with that in the keto-N1H tautomer, while theground-state ring geometry was found to be planar. The electronic lowest singlet pp excitedstate geometry of the keto-N3H tautomer was found to have a boat-type structure where N1C2C4C5 atoms were approximately in one plane and the N3 and C6 atoms were out of this plane. A twisted structure around the N1C2 bond was revealed in the electronic lowest singlet np excited state of the keto-N3H tautomer. The ground-state amino group pyramidalization of the keto-N3H tautomer was found to be decreased, while the ring geometry was found to be slightly non-planar after hydration with three water molecules. In the np excited state, the hydrated structure was revealed to be completely modified. Such changes in the hydrated structures in the np state are in accordancewith the established fact that hydrogen bonds are largelydestabilized under np excitations [43, 156– 158]. The amino group of the enol tautomer of cytosine is found to be more pyramidal than the keto-N1H tautomer, while it shows a lower degree of pyramidization than the keto-N3H tautomer in the ground state. The electronic lowest singlet pp excited-state geometry of the enol tautomer was found to be largely distorted, especially around the C5C6 bond, and length of this bond was found to be increased by about 0.092 A. Hydration of the enol tautomer significantly reduced the amino group pyramidal character in the ground state. Further, the geometrical deformation of the hydrated enol form in the electronic lowest singlet pp excited state was predicted to be similar to that in the isolated tautomer [152]. The ground-state geometries of uracil and thymine were found to be planar, while the electronic lowest singlet np excited-state geometries were revealed to be slightly non-planar [159]. Further, in the electronic lowest singlet np excited state, the excitation being characterized by the promotion of the C4O lone pair electron to the antibonding p orbital, the length of the C4O bond was increased by about 0.1 A with respect to the ground-state value. Further, the C4O bond was found to be appreciably out of the approximate ring plane. Thymine in the lowest singlet pp excited state has been predicted to adopt a boat-type structure with the N1, C2, C4 and C5 atoms being approximately in one plane, while the N3 and C6 atoms are out of this plane. The geometries of the hydrated forms of thymine and uracil in which three water molecules were considered in the first solvation shell were also optimized in the ground and singlet excited states. It was revealed that hydration generally induces planarity in the system. However, structural deformation in the excited state was generally found to be similar to that in isolated species. Detailed discussion about hydrated species can be found in our earlier paper [159]. In jet-cooled studies, the geometrical deformation of the pp excited states of thymine and uracil has been suggested to be responsible for the diffuseness of spectra of these compounds [160, 161]. 5.3.3 Hydration of guanine 5.3.3.1 Structure We have recently performed systematic computational studies of the interaction of water molecules with guanine in the ground and the corresponding electronic lowest singlet pp excited state [49, 50]. In these investigations, 1, 3 and 5–13 water molecules were considered in the solvation shells of guanine. Ground-state geometries were optimized at the HF level, while excited-state geometries were optimized at the CIS level and the 6-311G(d, p) basis set was used in all calculations. The nature of the potential energy surfaces was ascertained through harmonic vibrational frequency analysis; all computed geometries were revealed minima at the respective potential energy surfaces. The selected calculated geometrical parameters are shown in Table 5.1. The ground-state geometries of the isolated and hydrated guanine were found to be planar, except for
119.1 120.5 119.8 0.6 46.5 10.2 26.5 28.0 7.1 177.3 10.9 178.2
H21N2C2 H22N2C2 H21N2H22 360-SHNH C6N1C2N3 N1C2N3C4 C2N3C4C5 N3C4C5C6 N1C6C5C4 N2C2N3C4 H21N2C2N1 H22N2C2N1
119.1 120.2 120.0 0.7 34.9 0.9 29.4 27.6 5.0 175.9 1.7 172.6
G þ 8W
117.1 114.3 113.8 14.8 64.7 39.2 2.0 18.1 7.3 159.7 31.0 167.9
G þ 1W
a
Values in parentheses correspond to CASSCF [18] and TDDFT [20] results.
G þ 7W3
115.3 (153.3, 117.1) 112.7 (113.3, 115.4) 111.4 (113.0, 115.7) 20.6 (18.4,11.8) 64.0 (70.3, 64.9) 44.2 (65.1, 67.3) 2.4 (14.4, 20.6) 18.5 (26.9, 21.8) 0.6 (23.2, 27.1) 161.4 (151.9, 139.5) 42.3 (54.9, 45.8) 171.8 (172.6, 172.7)
H21N2C2 H22N2C2 H21N2H22 360-SHNH C6N1C2N3 N1C2N3C4 C2N3C4C5 N3C4C5C6 N1C6C5C4 N2C2N3C4 H21N2C2N1 H22N2C2N1
Parameters
Ga
Parameters
122.8 118.6 116.4 2.2 32.6 0.2 31.4 32.5 1.2 174.9 21.8 175.8
G þ 9W
116.7 114.4 113.3 15.6 64.8 42.4 0.0 18.8 3.3 160.5 34.8 170.4
G þ 3W
122.4 118.6 116.1 2.9 32.4 0.6 31.8 32.7 1.1 174.6 23.1 177.0
G þ 10W
119.2 120.4 119.8 0.6 38.1 1.7 32.2 32.2 2.6 177.3 10.2 178.7
G þ 5W
114.5 114.2 114.2 17.1 67.2 51.4 6.9 20.9 5.9 164.0 32.2 166.6
G þ 11W1
119.2 120.4 119.9 0.5 43.0 6.5 29.0 29.7 5.4 179.7 10.6 177.4
G þ 6W
115.7 114.4 112.5 17.4 65.4 35.4 6.5 15.3 15.5 161.8 41.9 175.1
G þ 11W2
119.2 120.1 119.9 0.8 37.1 2.3 30.6 31.0 2.2 177.2 14.8 175.8
G þ 7W1
114.7 114.3 114.3 16.7 67.3 51.2 6.8 20.7 5.4 164.1 31.6 166.4
G þ 12W
119.1 120.2 119.9 0.8 43.9 7.3 28.6 29.5 5.8 179.1 11.2 178.5
G þ 7W2
116.9 116.1 113.4 13.6 63.7 31.5 9.7 16.6 15.9 162.9 33.2 171.4
G þ 13W
Table 5.1 Selected dihedral angles (deg) and amino group angles (deg) of guanine and different hydrated complexes in the lowest singlet pp excited state obtained at the CIS/6-311G(d, p) level. Reprinted with permission from [50]. Copyright 2008 American Chemical Society
132 Hydrogen Bonding and Transfer in the Excited State
Electronic-Excited-State Structures and Properties of Hydrated DNA Bases and Base Pairs
133
the amino group, which was found to be pyramidal in the isolated and some hydrated forms of guanine. The water molecules, when involved in direct interaction with amino hydrogens, were found to induce planarity in the amino group. Electronic excitation of guanine and hydrated guanine complexes to the electronic lowest singlet pp excited state (S1(pp )) was revealed to be characterized by the HOMO ! LUMO configuration. The geometry of the isolated guanine in the S1(pp ) excited state was found to be significantly non-planar. The C6N1C2N3, N1C2N3C4 and N3C4C5C6 dihedral angles were predicted to deviate by 64.0 , 44.2 and 18.5 , respectively, from the planarity at the CIS/6-311G(d, p) level. Mennucci et al. [162] have also optimized the lowest singlet pp excited-state geometries of the keto-N9H and keto-N7H tautomers in the gas phase and in water solution at the CIS/cc-pVDZ level; the aqueous environment was considered using the integral equation formalism continuum model. A significant solvent effect was revealed in the excited-state geometries, mainly affecting bond lengths and bond angles of the molecules. The excited-state properties of guanine tautomers have also been investigated at the CASSCF [18, 19] and TDDFT [20] levels. Selected excited-state geometrical parameters of guanine obtained at the CASSCF and TDDFT levels are shown in Table 5.1. It is clear from this table that the CIS/6-311G(d, p) predicted excitedstate geometry of guanine is reasonably similar to that obtained from the CASSCF and TDDFT methods. The major difference is obtained for the N1C6C5C4 dihedral angle, which has been predicted to deviate about 23 from planarity at the CASSCF level, while it is planar at the CIS level. Further, the C2 atom is also displaced significantly more out-of-plane at the CASSCF level than at the CIS level. On the other hand, the TDDFT predicted geometry for guanine is qualitatively similar to that obtained from the CASSCF method. However, it should be noted that, for the keto-N7H tautomer of guanine, the TDDFT method predicted significantly less non-planar excited-state geometry [20]. On the other hand, the corresponding geometry at the CASSCF level was found to be significantly non-planar, similar to that obtained for the keto-N9H tautomer [18, 19]. Nevertheless, the CIS method also predicted significantly non-planar excited-state geometry for the keto-N7H tautomer, although the amount of non-planarity was smaller than that of the keto-N9H tautomer [47]. The plots containing the variations in selected important dihedral angles, namely C6N1C2N3, N1C2N3C4, C2N3C4C5, N3C4C5C6 and N1C6C5C4, of guanine with respect to the degree of hydration in the electronic lowest singlet pp excited state are presented in Figure 5.2. Based upon structural data and the dihedral angle plot, the excited-state geometries were arranged in three groups [49, 50]. Hydrated complexes of guanine with 5–10 water molecules were in group I, the isolated guanine and its hydrated complexes with 1, 3, 11 (one configuration – the G þ 11W1 complex) and 12 water molecules were in group II and complexes with 11 (other configuration – the G þ 11W2 complex) and 13 water molecules were in group III. The structures of selected species as a representative of each group in the ground and excited state are shown in Figure 5.3. Excited-state geometries of guanine revealed remarkable differences among these three groups. For example, the complexes belonging to group I had a planar N1C2N3C4 dihedral angle. However, this angle is about 50 and 35 away from the planarity for complexes belonging to groups II and III respectively. The C6N1C2N3 dihedral angle is found to be about 30 for complexes belonging to group I, but for complexes belonging to groups II and III the magnitude of the same angle is predicted to be about 65 . Significant variations, though smaller in magnitude, were also revealed for other dihedral angles. Generally, the excited-state structural deformation of guanine was found to be similar for groups II and III, except for the N1C2N3C4 and N1C6C5C4 dihedral angles. The magnitude of these two dihedral angles is found to be about 15 smaller and 10 larger, respectively, for the group III complexes. The HOMO and LUMO for guanine and hydrated complexes representing each of the groups are presented in Figure 5.4. It is evident that orbital contamination from water is not present in hydrated guanine complexes. The relaxed electronic lowest singlet pp excited state of guanine complexes was revealed to be characterized by the HOMO ! LUMO configuration. Further, the HOMO is p-type and has a similar nature for both the isolated and hydrated guanine. The LUMO has been found to be localized mainly on the six-membered ring and revealed to have a p -type nature. The distributions of LUMO orbitals among the three groups were found
134 Hydrogen Bonding and Transfer in the Excited State
Figure 5.2 Variation in selected important dihedral angles of guanine with the degree of hydration in the electronic lowest singlet pp excited state [49, 50]. Reprinted with permission from [50]. Copyright 2008 American Chemical Society
to be significantly different, as shown in Figure 5.4. Thus, the change in the nature of LUMO orbitals clearly depends upon the mode and degree of hydration of guanine. It has been suggested that the difference in the LUMO distribution is responsible for the remarkably different excited-state structural deformation of guanine in the electronic lowest singlet pp excited state [49, 50]. 5.3.3.2 Ground- and Excited-State Stretching Vibrational Frequencies Table 5.2 shows the computed ground and electronic lowest singlet pp excited-state stretching vibrational frequencies, at the HF/6-311G(d, p) and CIS/6-311G(d, p) levels respectively, corresponding to the N9H (nN9H), N1H (nN1H), symmetric NH2 (nsymNH2) and asymmetric NH2 (nasymNH2) vibrational modes (symmetric and asymmetric NH2 assignments are approximate owing to the structural non-planarity) of guanine in isolated and hydrated forms. The variation in these vibrational modes with respect to the degree and mode of hydration is depicted in Figure 5.5. It has been revealed that stretching vibrations of different protondonating sites of guanine are affected by the degree of hydration and the structural non-planarity in the excited state. As ab initio computed vibrational frequencies need to be scaled for comparison with the corresponding experimental data, a scaling of 0.9051 [163] has been used for ground-state vibrations. The same scaling factor was used for the excited-state vibrations – this selection was justified because the CIS method is the HF analogue for the excited state. A significant spectral shift consequent to the hydration of guanine and that of the electronic excitation was revealed. Further, it was inferred that generally the arrangement of water molecules around the hydrogen-bond-donating sites of guanine in hydrated complexes has an important role in the corresponding NH stretching vibrations. However, in the electronic lowest singlet pp excited state, the nN9H vibrational frequencies of guanine in isolated and hydrated forms are generally similar to those of
Electronic-Excited-State Structures and Properties of Hydrated DNA Bases and Base Pairs
135
Figure 5.3 Ground- (S0) and electronic lowest singlet pp excited-state (S1(pp )) geometries of complexes of isolated guanine and its hydrated complexes with ten and 13 water molecules. Reprinted with permission from [49]. Copyright 2005 American Chemical Society
the corresponding ground-state values. Similarity among ground- and excited-state nN9H vibrational frequencies of guanine and hydrated complexes was suggested owing to the fact that the five-membered guanine fragment remains planar in the excited state. 5.3.4 Ground- and excited-state proton transfer in guanine The ground- and electronic singlet pp excited-state proton transfer barrier height of the guanine corresponding to keto–enol tautomerism has also been investigated [39]. The effect of a water molecule in the proton
136 Hydrogen Bonding and Transfer in the Excited State
Figure 5.4 HOMO and LUMO orbitals corresponding to the excited-state optimized geometry of selected hydrated complexes of guanine. Reprinted with permission from [49]. Copyright 2005 American Chemical Society
transfer reaction path was also studied. In this investigation, the ground-state geometries, including transition states, were also optimized at the B3LYP/6-311 þþ G(d, p) level, and geometries including the transition state in the electronic lowest singlet pp excited state were optimized at the CIS/6-311G(d, p) level. The TDB3LYP/6-311 þþ G(d, p) approach was used to compute the vertical transition energies using the CIS/6-311G (d, p) optimized geometries. The effect of bulk water solution was investigated using the PCM solvation model. The computed proton transfer barrier heights are shown in Table 5.3. It is clear that the ground- and excited-state proton transfer barrier height has been predicted to be significantly large both in the gas phase and in water solution. However, in the presence of a water molecule in the proton transfer reaction path, the barrier heights are significantly reduced compared with the unhydrated species. Interestingly, the excited-state barrier heights are generally found to be slightly larger than the corresponding ground-state values. Theoretical calculations suggested that the singlet electronic excitation of guanine may not facilitate the keto–enol tautomerization either in the gas phase or in the water solution. The geometries of the transition states were also found to be significantly non-planar. Geometries of the hydrated transition states in the ground and lowest singlet pp excited states were found to adopt a zwitterionic form in which the water molecule is in the form of a hydronium cation (H3Oþ ) and the guanine is in the anionic form, except for the N9H form in the excited state, where the water molecule is in hydroxyl anionic form (OH) and the guanine is in cationic form.
3758 3750
3702
3700 3781 3845 3748 3839
G þ 7W3 G þ 8W
G þ 9W
G þ 10W G þ 11W1 G þ 11W2 G þ 12W G þ 13W
3349 3422 3480 3392 3475
3351
3401 3394
3479 3386 3366 3391 3398 3401 3400
B
3733 3675 3815 3676 3814
3734
3793 3773
3809 3748 3738 3783 3790 3788 3790
A
B
3379 3326 3453 3327 3452
3380
3433 3415
3448 3392 3383 3424 3430 3429 3430
S1(pp )
3704 3684 3744 3690 3766
3731
3718 3717
3900 3900 3799 3884 3715 3712 3823
A
S0
3352 3334 3389 3340 3409
3377
3365 3364
3530 3530 3438 3515 3362 3360 3460
B
A
3683 3637 3718 3643 3753
3752
3727 3722
B
3333 3292 3365 3297 3397
3396
3373 3369
3530 3533 3435 3497 3371 3372 3458
S1(pp )
3900 3903 3795 3864 3724 3726 3821
nN9H
a A represents unscaled and B represents scaled (scaling factor 0.9051, Ref. [163]) frequencies. *Bold numbers show significant mixing with other stretching vibrations.
3844 3741 3719 3746 3754 3758 3757
G G þ 1W G þ 3W G þ 5W G þ 6W G þ 7W1 G þ 7W2
A
S0
nN1H
3648 3644 3762 3629 3730
3662
3735 3699
3805 3819 3817 3714 3731 3730 3734
A
S0
3302 3298 3405 3285 3376
3314
3381 3348
3444 3457 3455 3362 3377 3376 3380
B
3654 3726 3724 3729 3697
3663
3714 3700
B
3307 3372 3371 3375 3346
3315
3362 3349
3288 3353 3333 3327 3354 3350 3358
S1(pp )
3633 3705 3682 3676 3706 3701 3710
A
nsymNH2
3888 3911 3871 3922 3861
3894 3822 3830 3861
3919 3940 3937 3885 3890 3873 3890
A
S0
3519 3540 3504 3550 3495
3524 3459 3467 3495
3547 3566 3563 3516 3521 3505 3521
B
3838 3854 3859 3848 3826 3850 3824
B
3488 3470 3493 3483 3463 3485 3461
3486 3513 3508 3468 3479 3469 3480 3485 3486 3470
S1(pp )
3852 3881 3876 3832 3844 3833 3845 3850 3851 3834
A
nasymNH2
Table 5.2 Selected stretching vibrational frequencies (cm1) of guanine and hydrated guanine in the ground and excited state, obtained at the HF/6-311G(d, p) and CIS/6-311G(d, p) levels respectivelya . Reprinted with permission from [50]. Copyright 2008 American Chemical Society
Electronic-Excited-State Structures and Properties of Hydrated DNA Bases and Base Pairs 137
138 Hydrogen Bonding and Transfer in the Excited State
Figure 5.5 Variation in different stretching vibrational frequencies (unscaled) of guanine with the degree of hydration in the ground and excited state. Reprinted with permission from [50]. Copyright 2008 American Chemical Society
5.4 Excited States of Base Pairs The electronic singlet excited-state properties of the Watson–Crick GC, AT and AU base pairs were evaluated theoretically [78, 79]. Ground-state geometries of base pairs were optimized at the HF level, and electronic singlet excited states were computed at the CIS level. These calculations were performed under the assumption of a planar symmetry, and the 6-31 þþ G(d, p) basis set was used in the study. Computed vertical singlet pp and np transitions of base pairs suggested that electronic excitations were due to the contribution of molecular Table 5.3 Computed ground- and excited-state barrier height (kcal mol1) for the keto–enol guanine tautomerism at the B3LYP/6-311 þþ G(d, p) and TD-B3LYP/6-311 þþ G(d, p)//CIS/6-311G(d, p) level, respectively, in the gas phase and in water solution. Reprinted with permission from [39]. Copyright 2005 American Chemical Society Species
Keto-N9H ! TS-N9H Enol-N9H ! TS-N9H Keto-N7H ! TS-N7H Enol-N7H ! TS-N7H Keto-N9HH2O ! TS-N9HH2O Enol-N9HH2O ! TS-N9HH2O Keto-N7HH2O ! TS-N7HH2O Enol-N7HH2O ! TS-N7HH2O
Ground state
Excited state
Gas
Water
Gas
Water
37.5 36.3 40.6 35.9 15.9 12.8 17.3 12.0
45.2 38.2 46.5 38.5 16.7 10.4 17.1 10.2
42.9 36.9 36.8 36.8 19.8 12.8 13.9 13.4
45.7 40.8 41.4 39.3 18.3 13.5 13.5 11.2
Electronic-Excited-State Structures and Properties of Hydrated DNA Bases and Base Pairs
139
orbitals localized on one or other monomeric unit. These results are supported by the experimental observation of AT and GC polymers and natural DNA where the electronic transitions were assigned to the corresponding monomer bases [164]. Some charge-transfer-type states, characterized by the excitation of electrons from the occupied orbitals of one base to the virtual orbitals of the complementary base of the base pair, were also revealed, and these states are located slightly higher in energy. A significant effect on the singlet pp transition energies of monomers consequent to the base pair formation was not revealed. However, the np singlet states of monomers were found to be significantly blue-shifted as a result of base pair formation. Such a blue-shift in np states is in agreement with the experimental findings that such transitions are blue-shifted in hydrogenbonding environments [48, 165]. The excited-state geometrical deformation (under planar symmetry) was found to be localized at the excited monomer under base pair excitation. The interaction energies of base pair formation in the ground and electronic singlet excited states were also calculated. The interaction energy between two interacting monomers X and Y can be defined as Eint ¼ EðXYÞEðXÞEðYÞ
ð5:1Þ
where E(XY) is the total energy of the XY dimer, and E(X) and E(Y) are the energies of the interacting monomers. However, this type of interaction energy calculation is contaminated by the basis set superposition error caused by the different number of basis functions used to describe the monomers and dimer. The basis set superposition error (BSSE)-corrected interaction energies of nucleic acid base pairs were estimated using the Boys–Bernardi counterpoise correction scheme [166]. The interaction energy (Eint) in the ground state was calculated using the formula Eint ¼ EðXYÞEðXXY ÞEðYXY Þ
ð5:2Þ
where E(XY) is the total energy of the XY base pair in the ground state, and E(XXY) and E(YXY) are the total energies of the X (adenine in the case of the AT (AU) base pair and guanine in the case of the GC base pair) and the Y (thymine (uracil) in the case of the AT (AU) base pair and cytosine in the case of the GC base pair) monomeric moieties using the optimized XY base pair geometry and ghost atoms in place of the complementary base. It should be noted that the deformation energy of monomers (Edef ¼ E(X) E0 (XXY); E0 (XXY) is the energy of X monomer in the dimer geometry) is ignored in the above equation. However, in the case where deformation energy is significant, it should be added to the equation. We have utilized the Boys–Bernardi counterpoise correction schemes developed for the ground state [166] to compute the BSSE-corrected interaction energy for excited states. The interaction energy in the excited state ðnÞ (Eint ), where the excitation is localized at the X monomeric moiety, was estimated using the formula ðnÞ
0
Eint ¼ EðnÞ ðXYÞEðn Þ ðXXY ÞEð0Þ ðYXY Þ
ð5:3Þ
For the excited state where the excitation is localized at the Y monomeric moiety, the interaction energy was estimated using the formula ðnÞ
0
Eint ¼ EðnÞ ðXYÞEð0Þ ðXXY ÞEðn Þ ðYXY Þ
ð5:4Þ
In equations (5.3) and (5.4), E(n)(XY) is the total energy of the XY base pair in the nth excited state, 0 E (XXY) and E(n )(YXY) are the total energies of the X and Y monomeric moieties, respectively, in the n0 th excited state, which corresponds to the nth state of the XY base pair (as the nth state of the XY base pair may not necessarily correspond to the nth state of excited X or Y [78, 79]), and E(0)(XXY) and E(0)(YXY) are the (n0 )
140 Hydrogen Bonding and Transfer in the Excited State
ground-state total energies of the X and Y monomeric moieties respectively. In these calculations, the geometries of the X and Y monomeric moieties were those of the bases in the optimized geometry of the XY base pair in the nth excited state, while the ghost atoms were added in place of the complementary base. The hydrogen bond parameters and interaction energies of the base pairs in the ground and excited states are shown in Table 5.4. As expected, the hydrogen bond parameters and interaction energies for the AT and AU base pairs are similar. The analysis of the hydrogen bond angles shown in Table 5.4 clearly suggest that hydrogen bonds are more nonlinear in the singlet np states than those in the ground and singlet pp states. Further, it is also clear from the values of the interaction energy of base pairs in the ground and excited states that base pairs would have a similar stability in the ground and electronic singlet pp states, but would be significantly destabilized in the np states. The geometries of the guanine–cytosine (GC) and guanine–guanine (GG) base pairs in the ground and the electronic lowest singlet pp excited state were also investigated at the HF and CIS levels using the 6-311G(d, p) basis set. The computed results were compared with those for isolated guanine obtained at the same level of
Table 5.4 Hydrogen bond lengths (A), hydrogen bond angles (deg) and interaction energies (Eint, kcal mol1) of the AT, AU and GC base pairs in the ground and different singlet excited statesa. Reprinted with permission from [79 & 78]. Copyright 2002 American Chemical Society AT base pair N6. . .O40 N1. . .N30 H61(N6). . .O40 N1. . .H30 ffO40 H61N6 ffN30 H30 N1 Eint AU base pair N6. . .O40 N1. . .N30 H61(N6). . .O40 N1. . .H30 ffO40 H61N6 ffN30 H30 N1 Eint GC base pair O6. . .N40 N1. . .N30 N2. . .O20 O6. . .H410 (N40 ) H1(N1). . .N30 H21(N2). . .O20 ffO6H410 N40 ffN1H1N30 ffN2H21O20 Eint
S0
S2(p–p )
S3(p–p )
S6(n–p )
S10(n–p )
3.083 3.023 2.090 2.010 172.6 177.7 9.9
3.011 3.071 2.016 2.063 171.1 174.8 10.0
3.027 3.025 2.029 2.014 174.5 177.2 10.6
3.778 3.090 2.778 2.086 161.8 183.2 5.8
3.094 3.152 2.106 2.149 170.2 174.3 7.1
S0
S2(p–p )
S4(p–p )
S5(n–p )
S10(n–p )
3.082 3.019 2.088 2.007 172.6 177.6 10.1
3.011 3.066 2.016 2.059 171.0 174.6 10.2
3.029 3.022 2.032 2.010 174.2 177.0 10.7
3.742 3.087 2.786 2.083 161.9 183.0 5.9
3.092 3.146 2.104 2.143 170.2 174.1 7.3
S0
S3(p–p )
S4(p–p )
2.932 3.053 3.028 1.926 2.048 2.028 175.7 175.3 176.9 24.8
2.937 3.023 3.037 1.929 2.021 2.034 177.7 176.3 177.4 22.9
2.958 3.066 3.098 1.954 2.059 2.100 177.3 176.4 177.7 20.4
States are given in ascending energy order (see Refs [78] and [79]).
a
Electronic-Excited-State Structures and Properties of Hydrated DNA Bases and Base Pairs
141
Figure 5.6 Molecular geometries in the electronic lowest singlet pp excited state of GC, GG17 and GG16 base pairs. The top indices correspond to the ground state obtained at the HF/6-311G(d, p) level, and the bottom indices correspond to the excited state obtained at the CIS/6-311G(d, p) level. Reprinted with permission from [167]. Copyright Elsevier
theory [167]. For the GG base pair, two different hydrogen-bonding configurations, namely GG16 and GG17, were studied. The GG16 structure was obtained by the formation of the cyclic and symmetrical hydrogen bonds in between the N1H site of the first guanine monomer and the carbonyl group of the second guanine monomer, and vice versa. On the other hand, the GG17 base pair was obtained by the formation of hydrogen bonds between the amino group and the N1H site of the first guanine monomer acting as hydrogen bond donors (GD) with the carbonyl group and the N7 site of the second monomer acting as the hydrogen bond acceptor (GA). The electronic excitation to the lowest singlet pp excited state of the GC base pair was predicted to be dominated by orbitals mainly localized at the guanine moiety. In the GG17 base pair, electronic excitation was found to be localized at the GA monomer. In the case of the GG16 base pair, the orbitals involved in the lowest singlet pp electronic excitation were found to be delocalized [167]. The optimized ground- and excited-state base pair geometries are depicted in Figure 5.6, and selected geometrical parameters are shown in Table 5.5. It was found that the amino groups of guanine belonging to the GC and GG16 base pairs are almost planar, but the corresponding amino group in the GG17 base pair was revealed to be significantly pyramidal both in the ground and in the electronic lowest singlet pp excited state. The excited-state geometries of the isolated Table 5.5 Selected parameters of the guanine monomer in the GC, GG16 and GG17 base pairs and GC þ 5W complex in the ground and lowest singlet pp excited state obtained at the HF/6-311G(d, p) and CIS/6-311G(d, p) levels respectivelya. Reprinted with permission from [167]. Copyright Elsevier GC S0 H21N2C2 H22N2C2 H21N2H22 360-SHNH C6N1C2N3 N1C2N3C4 C2N3C4C5 N3C4C5C6 N1C6C5C4 N2C2N3C4 H21N2C2N1 H22N2C2N1
122.8 117.1 120.1 0.0 0.0 0.0 0.0 0.0 0.0 180.0 0.0 180.0
GG16 pp
122.8 117.4 119.7 0.1 36.4 3.3 28.8 29.6 2.6 177.7 3.7 178.5
S0 120.5 117.4 119.3 2.8 0.1 0.4 0.3 0.1 0.5 178.5 10.9 171.7
GG17 pp
120.7 118.1 119.9 1.3 3.1 0.7 1.3 1.1 1.0 179.8 8.2 175.2
S0 118.4 114.2 115.3 12.1 0.5 0.7 1.1 1.3 0.9 177.4 29.3 170.1
GC þ 5W pp
116.6 113.5 112.2 17.7 64.7 44.4 2.2 18.1 1.9 156.9 41.6 174.3
S0 122.0 117.1 119.0 1.9 0.1 0.7 1.0 0.4 0.5 179.4 8.0 172.0
pp 122.8 118.0 119.0 0.2 37.0 4.8 27.1 27.4 4.1 178.4 7.1 177.9
In the GG17 base pair, the data correspond to the GA guanine moiety which acts as a hydrogen bond acceptor, while in the GG16 base pair both monomers have similar geometry owing to the symmetry.
a
142 Hydrogen Bonding and Transfer in the Excited State
Figure 5.7 Ground- (S0) and electronic lowest singlet pp excited-state (S1(pp )) geometries of the complex of the GC base pair with five water molecules
guanine and those of the GC and GG17 base pairs were found to be non-planar. The predicted structural nonplanarity in these base pairs was located at the excited guanine monomer. A remarkable difference was revealed in the mode of non-planarity, and it was found to be significantly influenced by the hydrogen bonding in the base pair. The geometrical deformation of isolated guanine and the GA monomer of the GG17 base pair in the excited state were similar but found to be significantly different to the structural deformation of guanine monomer of the GC base pair in the excited state. The geometry of the GG16 base pair was predicted to be almost planar in the excited state, except that both guanine monomers are folded with respect to each other. It was inferred that hydrogen bonding environments will have significant influence on excited-state dynamics of bases and DNA. Effects of specific water solvation of the Watson–Crick GC base pair in the ground and electronic lowest singlet pp excited state were studied at the HF/6-311G(d, p) and the CIS/6-311G(d, p) levels respectively. In this investigation, five water molecules were used to hydrate the guanine, and the resulting structure hereafter will be referred to as GC þ 5W. The optimized geometries of the GC þ 5W complex in the ground and in the electronic lowest singlet pp excited state are shown in Figure 5.7, and selected geometrical parameters are shown in Table 5.5. The harmonic vibrational frequency analysis suggested that these structures correspond to minima at the respective potential energy surfaces. Similarly to the isolated GC base pair, the electronic lowest singlet pp excited state of the GC þ 5W complex is found to be dominated by orbitals mainly localized at the guanine moiety. Further, no orbital contamination from water molecules was revealed in the electronic excitation to the lowest singlet pp excited state of the complex. In addition, the ground- and excited-state geometrical distortion of the GC þ 5W complex was found to be similar to the isolated GC base pair in the respective electronic states.
5.5 Excited-State Dynamics and Non-Radiative Decays It is unclear as to what was the basis of the selection of purine and pyrimidine bases (adenine, guanine, cytosine, thymine and uracil) as genetic material by nature and whether their selection was absolute or they have evolved through some natural selection processes over the period of time. However, it is certain that these molecules are highly photostable, and this extraordinary photostability stems from the ultrashort excited-state lifetime [44, 52]. Consequently, the excitation energies are released in the form of non-radiative deactivation processes. Fortunately, the non-radiative relaxation abilities of nucleic acid bases are generally intact in nucleic acids also, although the excited-state dimers formed owing to the presence of stacking of bases also
Electronic-Excited-State Structures and Properties of Hydrated DNA Bases and Base Pairs
143
show decay of radiation in a longer time domain [53]. Here it should be noted that molecular environments for nucleic acid bases are significantly different in the nucleic acids compared with those of isolated species. The non-canonical forms of NAB have a significantly longer excited-state lifetime than the canonical forms [53]. Long back it was suggested that the N7H tautomeric forms of guanine and adenine are fluorescent species, and this conclusion was based upon comparison of the fluorescence spectra of these molecules with the respective methylated forms [48, 168]. The isomeric forms of bases also show an entirely different photostable character. One of the most cited examples is 2-aminopurine, an isomeric form of adenine (6aminopurine). 2-Aminopurine has very strong fluorescence [169], and in fact this property is utilized for a fluorescence probe in DNA dynamics by substituting it in place of adenine [170]. Substitutions on these bases also generally significantly affect photostability, with the exception of some methylated forms, which generally have similar characteristics to unmethylated canonical nucleic acid bases. The most glaring example is uracil (which is present in RNA) and its 5-methylated form, thymine, which is present in DNA, and both species are highly photostable. Recent state-of the-art theoretical and experimental work has convincingly shown that conical intersections between the excited-state and ground-state potential energy surfaces are responsible for non-radiative deactivation in nucleic acid bases [1, 18–22, 44, 45, 51–57, 171–179]. Thus, excited-state structural non-planarity plays an import role in the photophysics of nucleic acid bases, and the molecule at the point of conical intersection has a biradical character [178, 179]. Neither can the role of dark state in conical intersections in pyrimidines be ruled out [52]. As it is not the main objective of the present article to discuss the details of non-radiative deactivation processes, interested readers can find more detailed information in recently published review articles [52–54]. Following the previous discussions, it can be concluded that the mode and degree of hydration significantly influence the excited-state geometrical deformation of guanine [49, 50], and therefore it can be argued that excited-state dynamics of guanine and other bases will exhibit a dependency on hydration.
5.6 Conclusions Ground-state geometries of nucleic acid bases and base pairs in isolated and hydrated forms are generally almost planar, except for the amino groups, which usually have a pyramidal character. However, electronic singlet excited-state geometries of bases and base pairs are very often strongly non-planar, and such nonplanarity is localized in the six-membered ring. The advance of superfast computers and state-of-the-art computational algorithms have enabled researchers to study complicated excited-state phenomena of complex molecules such as nucleic acid bases and base pairs at the multireference theoretical level. Coherent efforts by both experimentalists and theoreticians have convincingly shown that conical intersections between excitedand ground-state potential energy surfaces are responsible for the ultrafast non-radiative decays in nucleic acid bases. Although most of these theoretical calculations have been performed in the isolated condition, we believe that in the near future it will be possible to consider the effect of solvents in such calculations. Further, it will also be possible to consider a larger system where hydrogen bonding and stacking interaction are both present. As real DNA is extremely complex, consideration of such a system will provide better information related to the photophysical and photochemical properties of the genetic system.
Acknowledgements The authors are grateful for financial support from NSF-CREST grant No. HRD-0833178 and NSF EPSCoR grant No. 440900362427-02. They are also grateful to the Mississippi Center for Supercomputing Research (MCSR) for generously providing computer facilities.
144 Hydrogen Bonding and Transfer in the Excited State
References 1. M. K. Shukla and J. Leszczynski (eds), Radiation Induced Molecular Phenomena in Nucleic Acids, Vol. 5 in the book series Challenges and Advances in Computational Chemistry and Physics, ed. by J. Leszczynski. Springer Science þ Business Media B.V. (2008). 2. J. Sponer and F. Lankas (eds), Computational Studies of RNA and DNA, Vol. 2 in the book series Challenges and Advances in Computational Chemistry and Physics, ed. by J. Leszczynski. Springer, Dordrecht, The Netherlands (2006). 3. R. K. McMullan, P. Benci and B. M. Craven, Acta Crystal B, 36, 1424 (1980). 4. J. Leszczynski, Int. J. Quantum Chem., 19, 43 (1992). 5. J. Sponer, J. Florian, P. Hobza and J. Leszczynski, J. Biomol. Struct. Dynam., 13, 827 (1996). 6. F. Dong and R. E. Miller, Science, 298, 1227 (2002). 7. G. A. Jeffrey, An Introduction to Hydrogen Bonding. Oxford University Press, New York, NY (1997). 8. P. Hobza and Z. Havlas, Chem. Rev., 100, 4253 (2000) and references cited therein. 9. S. J. Grabowski (ed.), Hydrogen Bonding, Vol. 3 in the series Challenges and Advances in Computational Chemistry and Physics, ed. by J. Leszczynski. Springer, Dordrecht, The Netherlands (2006). 10. J. Wessel, J. C. Lee, E. Peris et al., Angew. Chem. Int. Ed. Engl., 34, 2507 (1995). 11. R. Custelcean and J. E. Jackson, Chem. Rev., 101, 1963 (2001). 12. W. J. Schreier, T. E. Schrader, F. O. Koller et al., Science, 315, 625 (2007). 13. E. Nir, Ch. Janzen, P. Imhof et al., J. Chem. Phys., 115, 4604 (2001). 14. M. Mons, I. Dimicoli, F. Piuzzi et al., J. Phys. Chem. A, 106, 5088 (2002). 15. M. Y. Choi and R. E. Miller, J. Am. Chem. Soc., 128, 7320 (2006). 16. M. Mons, F. Piuzzi, I. Dimicoli et al., J. Phys. Chem. A, 110, 10921 (2006). 17. M. K. Shukla and J. Leszczynski, Chem. Phys. Lett., 429, 261 (2006). 18. H. Chen and S. Li, J. Phys. Chem. A, 110, 12360 (2006). 19. H. Chen and S. Li, J. Chem. Phys., 124, 154315 (2006). 20. C. M. Marian, J. Phys. Chem. A, 111, 1545 (2007). 21. S. Yamazaki and W. Domcke, J. Phys. Chem. A, 112, 7090 (2008). 22. S. Yamazaki, W. Domcke and A. L. Sobolewski, J. Phys. Chem. A, 112, 11965 (2008). 23. R. E. Franklin and R. G. Gosling, Acta Cryst., 6, 673 (1953). 24. V. I. Poltev, A. V. Teplukhin and G. G. Malenkov, Int. J. Quantum Chem., 42, 1499 (1992). 25. R. Brandes, A. Rupprecht and D. R. Kearns, J. Biophys., 56, 683 (1989). 26. J. A. DiVerdi and S. J. Opella, J. Mol. Biol., 149, 307 (1981). 27. H. Shindo, Y. Hiyama, S. Roy et al., Bull. Chem. Soc. Jpn, 60, 1631 (1987). 28. B. Schneider, D. Cohen and H. M. Berman, Biopolymers, 32, 725 (1992). 29. J. Leszczynski, in Advances in Molecular Structure Research, Vol. 6, ed. by M. Hargittai and I. Hargittai. JAI Press, Stamford, CT, p. 209 (2000). 30. G. G. Sheina, S. G. Stepanian, E. D. Radchenko and Yu. P. Blagoi, J. Mol. Struct., 158, 275 (1987). 31. S. A. Trygubenko, T. V. Bogdan, M. Rueda et al., Phys. Chem. Chem. Phys., 4, 4192 (2002). 32. M. Hanus, F. Ryjacek, M. Kabelac et al., J. Am. Chem. Soc., 125, 7678 (2003). 33. M. Kabelac and P. Hobza, Chem.-Eur. J., 7, 2067 (2001). 34. M. Kabelac, F. Ryjacek and P. Hobza, Phys. Chem. Chem. Phys., 2, 4906 (2000). 35. D. Sivanesan, I. Sumathi and W. J. Welsh, Chem. Phys. Lett., 367, 351 (2003). 36. L. Gorb and J. Leszczynski, J. Am. Chem. Soc., 120, 5024 (1998). 37. M. K. Shukla and J. Leszczynski, J. Mol. Struct. (Theochem.), 529, 99 (2000). 38. M. K. Shukla and J. Leszczynski, J. Phys. Chem. A, 104, 3021 (2000). 39. M. K. Shukla and J. Leszczynski, J. Phys. Chem. A, 109, 7775 (2005). 40. O. V. Shishkin, L. Gorb and J. Leszczynski, Int. J. Mol. Sci., 1, 17 (2000). 41. O. V. Shishkin, L. Gorb and J. Leszczynski, J. Phys. Chem. B, 104, 5357 (2000). 42. O. S. Sukhanov, O. V. Shishkin, L. Gorb et al., J. Phys. Chem. B, 107, 2846 (2003). 43. T. M. Korter, D. W. Pratt and J. Kupper, J. Phys. Chem. A, 102, 7211 (1998).
Electronic-Excited-State Structures and Properties of Hydrated DNA Bases and Base Pairs
145
44. C. E. Crespo-Hernandez, B. Cohen, P. M. Hare and B. Kohler, Chem. Rev., 104, 1977 (2004). 45. M. K. Shukla and J. Leszczynski, J. Biomol. Struct. Dynam., 25, 93 (2007). 46. M. K. Shukla and J. Leszczynski, in Computational Studies of RNA and DNA, ed. by J. Sponer and F. Lankas,in the book series Challenges and Advances in Computational Chemistry and Physics, ed. by J. Leszczynski. Springer, Dordrecht, The Netherlands, p. 433 (2006). 47. M. K. Shukla and J. Leszczynski, in Computational Chemistry: Reviews of Current Trends. Vol. 8, ed. by J. Leszczynski. World Scientific, Singapore, p. 249 (2003). 48. P. R. Callis, Ann. Rev. Phys. Chem., 34, 329 (1983). 49. M. K. Shukla and J. Leszczynski, J. Phys. Chem. B, 109, 17333 (2005). 50. M. K. Shukla and J. Leszczynski, J. Phys. Chem. B, 112, 5139 (2008). 51. J. Gonzalez-Vazquez, L. Gonzalez, E. Samoylova and T. Schultz, Phys. Chem. Chem. Phys., 11, 3927 (2009). 52. C. T. Middleton, K. de La Harpe, C. Su et al., Ann. Rev. Phys. Chem., 60, 217 (2009). 53. L. Serrano-Andres and M. Merchan, J. Photochem. Photobiol. C: Photochemistry Reviews, 10, 21 (2009). 54. C. Z. Bisgaard, H. Satzger, S. Ullrich and A. Stolow, ChemPhysChem, 10, 101 (2009). 55. D. Markovitsi, T. Gustavsson and F. Talbot, Photochem. Photobiol. Sci., 6, 717 (2007). 56. M. S. de Vries and P. Hobza, Ann. Rev. Phys. Chem., 58, 585 (2007). 57. H. Saigusa, J. Photochem. Photobiol. C: Photochem. Rev., 7, 197 (2006). 58. J. Sponer, J. Leszczynski and P. Hobza, Biopolymers (Nucl. Acid Sci.) 61, 3 (2002). 59. J. Sponer, P. Hobza and J. Leszczynski, in Computational Chemistry: Reviews of Current Trends. Vol. 5, ed. by J. Leszczynski. World Scientific, Singapore, p. 171 (2000). 60. O. V. Shishkin, L. Gorb and J. Leszczynski, Chem. Phys. Lett., 330, 603 (2000). 61. O. V. Shishkin, L. Gorb, P. Hobza and J. Leszczynski, Int. J. Quant. Chem., 80, 1116 (2000). 62. A. Laxer, D. T. Major, H. E. Gottlieb and B. Fischer, J. Org. Chem., 66, 5463 (2001). 63. M. T. Chenon, R. J. Pugmire, D. M. Grant et al., J. Am. Chem. Soc., 97, 4636 (1975). 64. J. Lin, C. Yu, S. Peng et al., J. Am. Chem. Soc., 102, 4627 (1980). 65. M. J. Nowak, H. Rostkowska, L. Lapinski et al., J. Phys. Chem., 98, 2813 (1994). 66. M. Hanus, M. Kabelac, J. Rejnek et al., J. Phys. Chem. B, 108, 2087 (2004). 67. C. F. Guerra, F. M. Bickelhaupt, S. Saha and F. Wang, J. Phys. Chem. A, 110, 4012 (2006). 68. M. Szczesniak, K. Szczepaniak, J. S. Kwiatowski et al., J. Am. Chem. Soc., 110, 8319 (1998). 69. J. S. Kwiatkowski and J. Leszczynski, J. Phys. Chem., 100, 941 (1996). 70. R. D. Brown, P. D. Godfrey, D. McNaughton and A. P. Pierlot, J. Am. Chem. Soc., 111, 2308 (1989). 71. M. Drefus, O. Bensaude, G. Dodin and J. E. Dubois, J. Am. Chem. Soc., 98, 6338 (1976). 72. E. Nir, M. Muller, L. I. Grace and M. S. de Vries, Chem. Phys. Lett., 355, 59 (2002). 73. J. Smets, L. Adamowicz and G. Maes, J. Phys. Chem., 100, 6434 (1996). 74. L. Lapinski, M. J. Nowak, J. Fulara et al., J. Phys. Chem., 94, 6555 (1990). 75. R. J. McClure and B. M. Craven, Acta Crystallogr., 29B, 1234 (1973). 76. R. Kobayashi, J. Phys. Chem. A, 102, 10 813 (1998). 77. J. Rejnek, M. Hanus, M. Kabelac et al., Phys. Chem. Chem. Phys., 7, 2006 (2005). 78. M. K. Shukla and J. Leszczynski, J. Phys. Chem. A, 106, 1011 (2002). 79. M. K. Shukla and J. Leszczynski, J. Phys. Chem. A, 106, 4709 (2002). 80. L. Gorb, Y. Podolyan, P. Dziekonski et al., J. Am. Chem. Soc., 126, 10119 (2004). 81. Y. Podolyan, M. J. Nowak, L. Lapinski and J. Leszczynski, J. Mol. Struct., 744, 19 (2005). 82. N. Kurita, V. I. Danilov and V. M. Anisimov, Chem. Phys. Lett., 404, 164 (2005). 83. I. Dabkowska, P. Jurecka and P. Hobza, J. Chem. Phys., 122, 204322 (2005). 84. J. Sponer, P. Jurecka and P. Hobza, J. Am. Chem. Soc., 126, 10142 (2004). 85. P. Jurecka, J. Sponer, J. Cerny and P. Hobza, Phys. Chem. Chem. Phys., 8, 1985 (2006). 86. J. Sponer, P. Jurecka, I. Marchan et al., Chem. Eur. J., 12, 2854 (2006). 87. L. Zendlova, P. Hobza and M. Kabelac, ChemPhysChem, 7, 439 (2006). 88. X. Li, Z. Cai and M. D. Sevilla, J. Phys. Chem. A, 106, 1596 (2002). 89. J. Lin, C. Yu, S. Peng et al., J. Am. Chem. Soc., 102, 4627 (1980). 90. X. Yang, X.-B. Wang, E. R. Vorpagel and L.-S. Wang, Proc. Natl Acad. Sci. USA, 101, 17588 (2004).
146 Hydrogen Bonding and Transfer in the Excited State 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. 119. 120. 121. 122. 123. 124. 125. 126. 127. 128. 129. 130. 131. 132. 133. 134. 135. 136. 137. 138. 139. 140. 141.
A. B. Trofimov, J. Schirmer, V. B. Kobychev et al., J. Phys. B: At. Mol. Opt. Phys., 39, 305 (2006). D. M. Close, J. Phys. Chem. A, 108, 10376 (2004). C. E. Crespo-Hernandez, R. Arce, Y. Ishikawa et al., J. Phys. Chem. A, 108, 6373 (2004). E. Cauet, D. Dehareng and J. Lievin, J. Phys. Chem. A, 110, 9200 (2006). Y. Podolyan, L. Gorb and J. Leszczynski, J. Phys. Chem. A, 104, 7346 (2000). Y. Huang and H. Kenttamaa, J. Phys. Chem. A, 108, 4485 (2004). E. S. Kryachko, M. T. Nguyen and T. Zeegers-Huyskenes, J. Phys. Chem. A, 105, 1288 (2001). S. Ganguly and K. K. Kundu, Can. J. Chem., 72, 1120 (1994). R. Benoit and M. Frechette, Can. J. Chem., 64, 2348 (1986). F. Greco, A. Liguori, G. Sindona and N. Uccella, J. Am. Chem. Soc., 112, 9092 (1990). M. C. Lind, P. P. Bera, N. A. Richardson et al., Proc. Natl Acad. Sci. USA, 103, 7554 (2006). A. Kumar, P. C. Mishra and S. Suhai, J. Phys. Chem. A, 109, 3971 (2005). A. Kumar, M. Knapp-Mohammady, P. C. Mishra and S. Suhai, J. Comput. Chem., 25, 1047 (2004). J. C. Sutherland and K. Griffin, Biopolymers, 23, 2715 (1984). W. Voelter, R. Records, E. Bunnenberg and C. Djerassi, J. Am. Chem. Soc., 90, 6163 (1968). L. B. Clark, J. Am. Chem. Soc., 99, 3934 (1977). L. B. Clark, J. Am. Chem. Soc., 116, 5265 (1994). C. Santhosh and P. C. Mishra, J. Mol. Struct., 198, 327 (1989). T. Inagaki, A. Ito, K. Heida and T. Ho, Photochem. Photobiol., 44, 303 (1986). R. F. Stewart and J. Davidson, J. Chem. Phys., 39, 255 (1963). L. B. Clark, J. Phys. Chem., 99, 4466 (1995). A. Holmen, A. Broo, B. Albinsson and B. Norden, J. Am. Chem. Soc., 119, 12240 (1997). N. J. Kim, G. Jeong, Y. S. Kim et al., J. Chem. Phys., 113, 10051 (2000). D. C. Luhrs, J. Viallon and I. Fischer, Phys. Chem. Chem. Phys., 3, 1827 (2001). E. Nir, K. Kleinermanns, L. Grace and M. S. de vries, J. Phys. Chem. A, 105, 5106 (2001). C. A. Sprecher and W. C. Johnson, Jr, Biopolymers, 16, 2243 (1977). D. Voet, W. B. Gratzer, R. A. Cox and P. Doty, Biopolymers, 1, 193 (1963). L. B. Clark and I. Tinoco, J. Am. Chem. Soc., 87, 11 (1965). T. Yamada and H. Fukutome, Biopolymers, 6, 43 (1968). F. Zaloudek, J. S. Novros and L. B. Clark, J. Am. Chem. Soc., 107, 7344 (1985). W. C. Johnson, Jr, P. M. Vipond and J. C. Girod, Biopolymers, 10, 923 (1971). A. Kaito, M. Hatano, T. Ueda and S. Shibuya, Bull. Chem. Soc. Jpn, 53, 3073 (1980). K. Raksany and I. Foldvary, Biopolymers, 17, 887 (1978). R. S. Becker and G. Kogan, Photochem. Photobiol., 31, 5 (1980). M. Fujii, T. Tamura, N. Mikami and M. Ito, Chem. Phys. Lett., 126, 583 (1986). L. B. Clark, J. Phys. Chem., 93, 5345 (1989). L. B. Clark, J. Phys. Chem., 94, 2873 (1990). J. S. Novros and L. B. Clark, J. Phys. Chem., 90, 5666 (1986). M. P. Fulscher, L. Serrano-Andres and B. O. Roos, J. Am. Chem. Soc., 119, 6168 (1997). J. Lorentzon, M. P. Fulscher and B. O. Roos, J. Am. Chem. Soc., 117, 9265 (1995). M. P. Fulscher and B. O. Roos, J. Am. Chem. Soc., 117, 2089 (1995). A. W. Lange, M. A. Rohrdanz and J. M. Herbert, J. Phys. Chem. B, 112, 6304 (2008). S. Tonzani and G. C. Schatz, J. Am. Chem. Soc., 130, 7607 (2008). F. Santoro, V. Barone and R. Improta, J. Comput. Chem., 29, 957 (2008). R. Improta and V. Barone, Theoret. Chem. Acc., 120, 491 (2008). A. W. Lange and J. M. Herbert, J. Am. Chem. Soc., 131, 3913 (2009). P. R. N. Kamya and H. M. Muchall, J. Phys. Chem. A, 112, 13691 (2008). B. Mennucci, A. Toniolo and J. Tomasi, J. Phys. Chem. A, 105, 4749 (2001). M. K. Shukla and J. Leszczynski, J. Comput. Chem., 25, 768 (2004). A. Tsolakidis and E. Kaxiras, J. Phys. Chem. A, 109, 2373 (2005). D. Varsano, R. D. Felice, M. A. L. Marques and A. Rubio, J. Phys. Chem. B, 110, 7129 (2006).
Electronic-Excited-State Structures and Properties of Hydrated DNA Bases and Base Pairs 142. 143. 144. 145. 146. 147. 148. 149. 150. 151. 152. 153. 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.
147
J. Cerny, V. Spirko, M. Mons et al., Phys. Chem. Chem. Phys., 8, 3059 (2006). T. Gustavsson, A. Banyasz, E. Lazzarotto et al., J. Am. Chem. Soc., 128, 607 (2006). T. Fleig, S. Knecht and C. Hattig, J. Phys. Chem. A, 111, 5482 (2007). A. Tajti, G. Fogarasi and P. G. Szalay, ChemPhysChem, 10, 1603 (2009). K. Kowalski and M. Valiev, J. Phys. Chem. A, 112, 5538 (2008). M. Schreiber, M. R. Silva-Junior, S. P. A. Sauer and W. Thiel, J. Chem. Phys., 128, 134 110–134111 (2008). M. K. Shukla and P. C. Mishra, Chem. Phys., 240, 319 (1999). M. K. Shukla, S. K. Mishra, A. Kumar and P. C. Mishra, J. Comput. Chem., 21, 826 (2000). S. K. Mishra, M. K. Shukla and P. C. Mishra, Spectrochim. Acta, 56A, 1355 (2000). A. Broo, J. Phys. Chem. A, 102, 526 (1998). M. K. Shukla and J. Leszczynski, J. Phys. Chem. A, 106, 11338 (2002). H.-H. Ritze, P. Hobza and D. Nachtigallova, Phys. Chem. Chem. Phys., 9, 1672 (2007). J. Andreasson, A. Holmen and B. Albinsson, J. Phys. Chem. B, 103, 9782 (1999). B. Mennucci, A. Toniolo and J. Tomasi, J. Am. Chem. Soc., 122, 10621 (2000). N. J. Kim, H. Kang, G. Jeong et al., J. Phys. Chem. A, 104, 6552 (2000). H. Kang, K. T. Lee and S. K. Kim, Chem. Phys. Lett., 359, 213 (2002). M. K. Shukla and P. C. Mishra, Chem. Phys., 230, 187 (1998). M. K. Shukla and J. Leszczynski, J. Phys. Chem. A, 106, 8642 (2002). B. B. Brady, L. A. Peteanu and D. H. Levy, Chem. Phys. Lett., 147, 538 (1988). L. Chinsky, A. Laigle, L. Peticolas and P.-Y. Turpin, J. Chem. Phys., 76, 1 (1982). B. Mennucci, A. Toniolo and J. Tomasi, J. Phys. Chem. A, 105, 7126 (2001). A. P. Scott and L. Radom, J. Phys. Chem. A, 100, 16502 (1996). P.-J. Chou and W. C. Johnson, Jr, J. Am. Chem. Soc., 115, 1205 (1993). G. J. Brealey and M. Kasha, J. Am. Chem. Soc., 77, 4462 (1955). S. F. Boys and F. Bernardi, Mol. Phys., 19, 553 (1970). M. K. Shukla and J. Leszczynski, Chem. Phys. Lett., 414, 92 (2005). R. W. Wilson and P. R. Callis, Photochem. Photobiol., 31, 323 (1980). C. Santhosh and P. C. Mishra, Spectrochim. Acta A, 47, 1685 (1991). M. A. O’Neill and J. K. Barton, J. Am. Chem. Soc., 124, 13053 (2002). Z. Lan, E. Fabiano and W. Thiel, ChemPhysChem, 10, 1225 (2009). E. Mburu and S. Matsika, J. Phys. Chem. A, 112, 12485 (2008). G. Olaso-Gonzalez, M. Merchan and L. Serrano-Andres, J. Am. Chem. Soc., 131, 4368 (2009). Z. Lan, E. Fabiano and W. Thiel, J. Phys. Chem. B, 113, 3548 (2009). T. Gustavsson, A. Banyasz, N. Sarkar et al., Chem. Phys., 350, 186 (2008). G. Zechmann and M. Barbatti, J. Phys. Chem. A, 112, 8273 (2008). M. Barbatti and H. Lischka, J. Am. Chem. Soc., 130, 6831 (2008). M. Z. Zgierski, S. Patchkovskii and E. C. Lim, J. Chem. Phys., 123, 81101 (2005). M. Z. Zgierski, S. Patchkovskii, T. Fujiwara and E. C. Lim, J. Phys. Chem. A, 109, 9384 (2005).
6 Insight from Singlet into Triplet Excited-State Hydrogen Bonding Dynamics in Solution Guang-Jiu Zhao and Ke-Li Han State Key Laboratory of Molecular Reaction Dynamics, Dalian Institute of Chemical Physics, Chinese Academy of Sciences, 457 Zhongshan Road, Dalian 116023, China
6.1 Introduction Solute–solvent interactions play a fundamental role in the photochemistry of organic and biological chromophores in solution [1–14]. In addition to non-specific dielectric interactions between solute and solvent, site-specific intermolecular hydrogen bonding interaction between hydrogen donor and acceptor molecules is another important type of solute–solvent interaction and is central to the understanding of the microscopic structure and function in many molecular systems [15–29]. Moreover, the dynamical behaviour of intermolecular hydrogen bonds in electronic excited states plays an important role in determining the rates of many chemical, physical and biochemical processes that occur in hydrogen-bonded surroundings [30–35]. Therefore, investigation of the hydrogen bonding dynamics of photoexcited chromophores in hydrogenbonded surroundings is very valuable and helpful in understanding the interesting photophysical and photochemical behaviours of these chromophores, as well as in designing and synthesizing organic functional materials by using hydrogen bonding interactions [36–51]. In previous studies we have theoretically investigated the electronic-excited-state structures and dynamics of hydrogen bonding for a variety of organic and biological chromophores, such as coumarin, fluorenone, oxazine, thioketones, novel D p A systems, dihydrogen-bonded phenol-BTMA, protochlorophyllide a, etc. [52–59]. The ground state and electronic excited states were investigated using the density functional theory (DFT) and the time-dependent density functional theory (TDDFT) methods respectively. It has been theoretically demonstrated for the first time that the intermolecular hydrogen bonds formed between these chromophores and the solvents can be significantly strengthened or weakened in the electronic excited states of
Hydrogen Bonding and Transfer in the Excited State, Volume I & II © 2011 John Wiley & Sons, Ltd. ISBN: 978-0-470-66677-7
Edited by Ke-Li Han and Guang-Jiu Zhao
150 Hydrogen Bonding and Transfer in the Excited State
chromophores [52–59]. At the same time, we also confirmed that excited-state dynamical behaviours of intermolecular hydrogen bonds contribute importantly to the photophysics and photochemistry of these chromophores [52–59]. It should be noted that all the investigations of electronic-excited-state hydrogen bonding dynamics only relate to the singlet electronic excited states. However, as we know, intersystem crossing (ISC) is a very important non-radiative transition process in many organic compounds and functional materials. Hence, the triplet electronic excited states that can be reached after the ISC process would be very important for understanding the photochemistry of these molecules. How does intermolecular hydrogen bonding influence the triplet electronic excited states of these chromophores? The answer to this question will be closely associated with the triplet electronic-excited-state hydrogen bonding dynamics. On the other hand, it is very difficult in experiments to monitor the hydrogen bonding dynamics in triplet excited states using timeresolved spectroscopic techniques owing to the relatively weak signals for triplet excited states. Therefore, it is a good choice to study the intermolecular hydrogen bonding in the triplet electronic excited states of chromophores by using theoretical methods. As we know, some spectral shifts of the characteristic vibrational modes involved in the formation of intermolecular hydrogen bonds can be induced by changes in intermolecular hydrogen bonding interactions [60–63]. Thus, femtosecond time-resolved vibrational spectroscopy has shown the potential to give good insight into the microscopic dynamics and provide information on local structures [52]. In addition, the time-dependent density functional theory (TDDFT) method has been demonstrated to be a reliable tool for calculating the vibrational spectra in electronic excited states [52–59, 64–69]. Consequently, in the present work, the TDDFT method will be used to study triplet excited-state hydrogen bonding dynamics by monitoring the spectral shifts of some characterized vibrational modes involved in the formation of intermolecular hydrogen bonds in different electronic states. The photophysical and photochemical properties of fluorenone (FN) and its derivatives have recently attracted considerable attention because they are excellent model compounds for investigating the molecular structures in both the ground state and electronic excited states, microscopic solvation dynamics and the ultrafast radiationless deactivation mechanism for photoexcited molecules [70–85]. To identify the nature of the electronic excited states of chromophores, several criteria such as absorption intensity, transition energy shift with increasing solvent polarity, direction of transition moment, vibrational structures, fluorescence and phosphorescence lifetime and quantum yields, singlet–triplet splitting, heavy atom effect on the S0 ! T1 transition and the electron spin resonance have been proposed [70–85]. It has been revealed by experimental investigation of the above criteria that the lowest S1 state of the fluorenone chromophore and its many derivatives in polar solvents is of pp nature. Very recently we have also theoretically demonstrated, by analysing the frontier molecular orbitals (MOs) of the hydrogen-bonded FN MeOH complex [53], that the S1 state of fluorenone in polar alcoholic solvents is of a distinct pp character. In addition, during the p ! p transition, the electron is delocalized from the benzene rings to the carbonyl group of fluorenone [53]. The electronegativity of the carbonyl group in the S1 state of the hydrogen-bonded FN MeOH complex would be greatly increased. Therefore, it has been demonstrated in our previous work [52–54] that the intermolecular hydrogen bond C¼O H O between fluorenone and methanol molecules is significantly strengthened in the S1 state upon photoexcitation of the hydrogen-bonded FN MeOH complex. Furthermore, hydrogen bond strengthening in the S1 state upon photoexcitation of fluorenone in the alcoholic solvent can be used to explain well all the spectral features of fluorenone chromophore in alcoholic solvents. At the same time, we have demonstrated that the radiationless deactivation of the fluorescent state via internal conversion (IC) can be facilitated by hydrogen bond strengthening in the S1 state [53]. So the IC from the fluorescent state to the ground state becomes the most important dissipative process for the S1 state of fluorenone in polar alcoholic solvents [53]. As a result, the dynamic behaviour of intermolecular hydrogen bonding in the triplet electronic excited state is difficult to monitor by experimental techniques, although the intersystem crossing (ISC) from the S1 state to the T1 state is also an important radiationless deactivation channel [70–75]. Therefore, theoretical study of the intermolecular hydrogen bonding in the T1 state of the hydrogen-bonded FN MeOH
Insight from Singlet into Triplet Excited-State Hydrogen Bonding Dynamics in Solution 151
complex will be helpful for understanding the deactivation of the populated triplet electronic states after the ISC process. The nature of the triplet electronic state can also be elucidated by vibrational spectroscopy [82–85]. The different nature of the T1 state for benzophenone and 4-phenylbenzophenone has been revealed by transient resonance Raman spectral studies. It has also been demonstrated that, for the T1(np ) state of benzophenone, the carbonyl group has a single-bond character, while for the T1(pp ) state of 4-phenylbenzophenone the carbonyl group retains a double-bond character [82–85]. Moreover, the time-resolved infrared absorption spectra of fluorenone and its carbonyl 18O-substituted analogues have also been studied to determine the nature of the T1 state of fluorenone in different solvents [85]. The stretching frequencies of the T1 state in both polar and nonpolar solvents were found to be located in the double-bond region, indicating that the T1 state of fluorenone is of pp character [85]. Thus, it could be expected that the intermolecular hydrogen bond C¼O H O between fluorenone and methanol molecules in the T1 state of the hydrogen-bonded FN MeOH complex should also be stronger than that in the ground state, as the T1 state is of the same nature as the S1 state for fluorenone in polar alcoholic solvents. Moreover, it has also been reported that the degree of electronic delocalization between the benzene rings and the carbonyl group for fluorenone in acetonitrile-d3 is smaller in the T1(pp ) state than in the S1(pp ) state [85]. So it could also be expected that the intermolecular hydrogen bonding in the T1 state of the hydrogen-bonded FN MeOH complex may be weaker than that in the S1 state. In the present work, to delineate the detailed aspects of triplet excited-state hydrogen bonding dynamics, the hydrogen-bonded FN MeOH complex and the isolated fluorenone chromophore have been theoretically studied using the DFT and TDDFT methods. Only the solvent molecules in the inner solvation shell can be attributed to the hydrogen bonding dynamics occurring in the ultrafast timescale. The hydrogen-bonded complex proposed here is a good model for studying the ultrafast hydrogen bonding dynamics in solutions [53]. Moreover, the vibrational absorption spectra of the hydrogen-bonded complex in the triplet excited states have also been calculated by the TDDFT method. It has been demonstrated that the intermolecular hydrogen bond is significantly strengthened in both the S1 and T1 states of the fluorenone chromoophore.
6.2 Theoretical Methods Ground-state geometry optimizations of isolated monomers and the hydrogen-bonded FN MeOH complex were performed using density functional theory (DFT) with Becke’s three-parameter hybrid exchange function with the Lee–Yang–Parr gradient-corrected correlation functional (B3-LYP functional) [86–89]. The triple-z valence quality with one set of polarization functions (TZVP) was chosen as the basis set throughout. The excited-state electronic structures were calculated using time-dependent density functional theory (TD-DFT) with the B3-LYP hybrid functional and the TZVP basis set. Fine quadrature grids 4 were also employed. Both the convergence thresholds for the ground-state and excited-state optimization were reset to 108 (default settings are 106). The excited-state Hessian was obtained by numerical differentiation of analytical gradients using central differences and default displacements of 0.02 Bohr [90–92]. The infrared intensities were determined from the gradients of the dipole moment. All the electronic structure calculations were carried out using the Turbomole program suite [86–92].
6.3 Results and Discussion The electronic excitation energies for the triplet electronic excited states of the isolated fluorenone and the hydrogen-bonded FN MeOH complex are calculated using the TDDFT method and listed in Table 6.1. In addition, the electronic excitation energies for the singlet electronic excited states are also listed for
152 Hydrogen Bonding and Transfer in the Excited State Table 6.1 Calculated electronic excitation energies (in nm) and the corresponding oscillator strengths of the singlet and triplet excited states for isolated fluorenone and the hydrogen-bonded FN MeOH complex FN
State 1 State State State State State State State State
2 3 4 5 6 7 8 9
FN MeOH
Singlet
Triplet
Singlet
Triplet
392 (0.004) H ! L 98.5% — 390 (0.000) 305 (0.016) 276 (0.032) 255 (0.000) 249 (0.869) 242 (0.043) 233 (0.005) 226 (0.003)
497 (0.004) H ! L 84.4% H ! L þ 1 6.0% 458 (0.000) 381 (0.045) 358 (0.023) 342 (0.136) 299 (0.007) 282 (0.162) 280 (0.030) 258 (0.000)
411 (0.003) H ! L 98.7% — 375 (0.000) 345 (0.003) 314 (0.036) 281 (0.034) 250 (0.137) 250 (0.719) 247 (0.058) 239 (0.000)
518 (0.029) H ! L 86.6% H ! L þ 1 4.9% 430 (0.000) 386 (0.044) 372 (0.045) 348 (0.005) 345 (0.104) 306 (0.018) 282 (0.138) 279 (0.044)
comparison. Note that the electronic excitation energy for the T1 state of the hydrogen-bonded FN MeOH complex is shifted to the lower energy in comparison with that of the isolated fluorenone. Thus, the energy level of the T1 state of the fluorenone chromophore can be induced to undergo a red-shift by intermolecular hydrogen bonding, which is similar to the hydrogen bonding effects on the singlet electronic excited states of the fluorenone chromophore [53]. As we know, it has been demonstrated that, if hydrogen bonding interaction induces an electronic spectral shift to the red, the intermolecular hydrogen bond will be strengthened in comparison with that in the ground state [59]. Hence, the intermolecular hydrogen bond may also be strengthened in the T1 state of the fluorenone chromophore in alcoholic solvents. It is also noted that the T1 state of both the isolated fluorenone and its hydrogen-bonded complex is lower in energy than the corresponding S1 state, which is consistent with the energy levels for the T1 and S1 states obtained in spectral studies [70–85]. Moreover, the contributions of orbital transition to the T1 and S1 states are also listed in Table 6.1. It has been demonstrated in a previous study [53] that the transition from HOMO to LUMO is the dominant orbital transition for the S1 state of both the isolated fluorenone and the hydrogen-bonded FN MeOH complex. It can be established that the T1 state of both the isolated fluorenone and the hydrogen-bonded FN MeOH complex is also dominantly contributed to by the orbital transition from HOMO to LUMO. In addition, the orbital transition from HOMO to LUMO þ 1 also contributes to the T1 state. Thus, three orbitals will be considered to discuss the nature of the T1 state. The three frontier molecular orbitals (HOMO, LUMO and LUMO þ 1) of both the isolated fluorenone and the hydrogen-bonded FN MeOH complex are shown in Figure 6.1. It is clear that the HOMO orbital of the isolated fluorenone is of p character, while both the LUMO and LUMO þ 1 orbitals of the isolated fluorenone are of p nature. Consequently, the T1 state of the isolated fluorenone, which corresponds to the orbital transition from HOMO to LUMO and LUMO þ 1, is demonstrated to be of distinct pp nature. Moreover, it can be noted that the nature of these three orbitals is nearly unchanged by intermolecular hydrogen bonding. The HOMO of the hydrogen-bonded FN MeOH complex is also of p character, while both LUMO and LUMO þ 1 orbitals of the hydrogen-bonded FN MeOH complex are again of p nature. Thus, the pp nature of the T1 state for the hydrogen-bonded FN MeOH complex is also elucidated. From frontier molecular orbital analysis for the isolated fluorenone and the hydrogen-bonded FN MeOH complex it can be theoretically confirmed that the T1 state of the fluorenone chromophore in both polar and non-polar solvents is of pp nature, which is consistent with previous studies [70–85].
Insight from Singlet into Triplet Excited-State Hydrogen Bonding Dynamics in Solution 153
Figure 6.1 The frontier molecular orbitals HOMO, LUMO and LUMO þ 1 of both the isolated fluorenone and the hydrogen-bonded FN MeOH complex
Furthermore, it is clear that the electron density of the LUMO orbital is strongly localized on the carbonyl group, while that of the LUMO þ 1 orbital is strongly localized on the benzene rings. Thus, the degree of electron delocalization between the benzeneringsand the carbonylgroup for the orbital transition from HOMO to LUMO is much stronger than that for the orbital transition from HOMO to LUMO þ 1. As a result, it can also be confirmed that the degree of electron delocalization from the benzene rings to the carbonyl group for fluorenone in alcoholic solvents is smaller in the T1(pp ) state than in the S1(pp ) state. So the electronic dipole moment of both the isolated fluorenone and the hydrogen-bonded FN MeOH complex in the T1 state should be smaller than that in the S1 state [53]. In addition, the electronegativity of the carbonyl group in the T1 state should be weaker than that in the S1 state. Therefore, it could be expected that the intermolecular hydrogen bonding in the T1 state of the hydrogen-bonded FN MeOH complex may be weaker than that in the S1 state. Calculated IR spectra of fluorenone in both the S0 and T1 states are shown in Figure 6.2. In addition, the ground-state IR spectrum of the hydrogen-bonded FN MeOH complex is also shown for comparison. Note that the C¼O stretching vibrational mode is calculated to be 1780 cm1 for the isolated fluorenone in the ground state, while the calculated C¼O stretching mode is 1756 cm1 for the hydrogen-bonded FN MeOH complex in the ground state. Thus, the formation of the intermolecular hydrogen bond C O H O can induce a slight red-shift of 24 cm1 for the C¼O stretching vibrational mode. Moreover, it is noted that the C¼O stretching mode of the isolated fluorenone is drastically red-shifted to 1678 cm1 in the T1 state from 1780 cm1 in the ground state. It is evident that the vibrational frequency of the C¼O stretching mode in the T1 state of fluorenone is located in the double-bond region. So the pp character of the T1 state for fluorenone can be confirmed again [82–85]. At the same time it can be noted that both the electronic excitation to the triplet states and intermolecular hydrogen bonding interactions can shift the stretching vibrational mode of the C¼O group to the red. However, the electronic excitation from the ground state to the T1 state can induce a larger redshift for the C¼O stretching mode than the intermolecular hydrogen bonding interaction. Therefore, the stretching mode of the C¼O group is not a sensitive vibrational mode for monitoring intermolecular hydrogen bonding [52, 53]. However, it has been demonstrated that the stretching mode of the O H group is very sensitive to the intermolecular hydrogen bonding interaction [53]. On the other hand, the methanol moiety remains in its electronic ground state upon photoexcitation to the T1 state of the hydrogen-bonded FN MeOH complex. So the stretching vibrational mode of the O H group cannot be strongly influenced by the electronic excitation. Thus, changes in intermolecular hydrogen bonding can be distinctly reflected by monitoring the stretching vibrational mode of the O H group in different electronic states.
154 Hydrogen Bonding and Transfer in the Excited State
Figure 6.2 The calculated IR spectra of fluorenone in both the S0 and T1 states. The ground-state IR spectrum of the hydrogen-bonded FN MeOH complex is also shown for comparison (See Plate 7)
The calculated IR spectra of the hydrogen-bonded FN MeOH complex in both the S0 and T1 states are shown in Figure 6.3. For comparison, the free O H stretching mode of the ground-state methanol molecule is also shown here. It can be noted that the O H stretching mode of the hydrogen-bonded FN MeOH complex in the ground state is significantly red-shifted from 3817 to 3641 cm1 owing to the formation of the intermolecular hydrogen bond C¼O H O. In the T1 state of the hydrogen-bonded FN MeOH complex, the O H stretching mode is further red-shifted from 3641 to 3541 cm1. The larger spectral red-shift of the O H stretching mode in the T1 state of the hydrogen-bonded FN MeOH complex in comparison with that in the S0 state indicates that the intermolecular hydrogen bond C¼O H O is significantly strengthened in the T1 state of the fluorenone chromophore. Therefore, we have demonstrated that the intermolecular hydrogen bond C¼O H O between fluorenone chromophore and methanol solvents can also be strengthened in the
Figure 6.3 Calculated IR spectra of the hydrogen-bonded FN MeOH complex in both the S0 and T1 states (See Plate 8)
Insight from Singlet into Triplet Excited-State Hydrogen Bonding Dynamics in Solution 155
Table 6.2 The calculated hydrogen bond binding energies Eb (in kJ mol1) and the related bond lengths L (in A), as MeOH well as the dipole moment m (in D) in the ground state and S1 and T1 states of the hydrogen-bonded FN complex and isolated fluorenone FN MeOH
S0 S1 T1
FN
MeOH
Eb
LC¼O
LO H
LH O
m
LC¼O
m
LH O
27.85 42.62 37.74
1.219 1.259 1.244
1.906 1.802 1.833
0.972 0.981 0.978
4.061 6.759 5.653
1.212 1.250 1.234
3.520 5.982 4.892
0.963 — —
triplet electronic excited states. It should be noted that the O H stretching vibrational frequency is calculated to be 3482 cm1 in the S1 state of the hydrogen-bonded FN MeOH complex [53]. Thus, the O H stretching mode in the T1 state is blue-shifted in comparison with that in the S1 state of the hydrogen-bonded FN MeOH complex. That is to say, the intermolecular hydrogen bond C¼O H O in the T1 state may be weaker than that in the S1 state. The calculated hydrogen bond binding energies and the related bond lengths as well as the dipole moment in the T1 state of the hydrogen-bonded FN MeOH complex and isolated fluorenone are listed in Table 6.2. The corresponding values in both the S0 and S1 states are also listed here for comparison. The bond length of the C¼O group in the T1 state of isolated fluorenone is calculated to be 1.234 A. The calculated C¼O bond length in the T1 state of the hydrogen-bonded FN MeOH complex is slightly lengthened to 1.244 A. It can be noted that the bond length of the C¼O group in the T1 state of the isolated fluorenone and the hydrogen-bonded complex are of a double-bond character. It can also be noted that the hydrogen bond binding energy for the intermolecular hydrogen bond C¼O H O in the T1 state of the hydrogen-bonded FN MeOH complex is calculated to be 37.74 kJ mol, which is significantly larger than that in the ground state. Thus, it is clear that the intermolecular hydrogen bond C¼O H O in the T1 state is significantly strengthened in comparison with that in the ground state. Thus, intermolecular hydrogen bond strengthening in the triplet electronic excited states is evidently confirmed. At the same time, the intermolecular hydrogen bond length is shortened to 1.833 A in the T1 state from 1.906 A in the ground state. Both the hydrogen-bonded C¼O and O H groups are correspondingly lengthened in the T1 state. On the other hand, it can also be noted that the hydrogen bond binding energy of the intermolecular hydrogen bond C¼O H O in the T1 state is smaller than that in the S1 state. Thus, it is reconfirmed that the intermolecular hydrogen bond C¼O H O in the T1 state of the hydrogen-bonded FN MeOH complex is weaker than that in the S1 state. In addition, it can also be noted that the electronic dipole moments of both the isolated fluorenone and the hydrogen-bonded FN MeOH complex in the T1 state are smaller than those in the S1 state. As discussed above, intermolecular hydrogen bonds formed between fluorenone and alcoholic solvents can be dynamically changed following important photophysical processes. Upon photoabsorption, fluorenone is initially photoexcited to the S1 state, in which the intermolecular hydrogen bond is significantly strengthened in comparison with that in the ground state. After internal conversion (IC) from the S1 state to the ground state, the intermolecular hydrogen bond will be weakened in the ground state. On the other hand, the fluorenone molecules in the S1 state can also get to the T1 state through intersystem crossing (ISC). After the ISC process, the intermolecular hydrogen bond is also weakened in comparison with that in the S1 state. The different changes in intermolecular hydrogen bond in the triplet and singlet electronic excited states may be closely associated with the photophysics and photochemistry of fluorenone in alcoholic solvents. More experimental and theoretical studies will be needed to gain a good understanding of the relationship between the photophysics of fluorenone and the hydrogen bonding dynamics in the electronic excited state.
156 Hydrogen Bonding and Transfer in the Excited State
6.4 Conclusion The time-dependent density functional theory (TDDFT) method was used to investigate the triplet electronicexcited-state hydrogen bonding dynamics of the fluorenone chromophore in alcoholic solvents. In the present work, the geometric structures of the hydrogen-bonded FN MeOH complex as well as the isolated fluorenone in the first triplet excited state (T1) were fully optimized using the TDDFT method. Moreover, the vibrational absorption spectra of the hydrogen-bonded FN MeOH complex in the triplet excited states were also calculated by the TDDFT method. At the same time, the calculated triplet electronic-excited-state IR spectra were compared with those in the ground state and singlet electronic excited states. Consequently, the triplet excited-state hydrogen bonding dynamics of fluorenone in alcoholic solvents was investigated for the first time by monitoring the spectral shifts of some characterized vibrational modes involved in the formation of intermolecular hydrogen bonds in different electronic states. As a result, it was demonstrated that the intermolecular hydrogen bond can be significantly strengthened in both the S1 and T1 states of the fluorenone chromoophore in comparison with that in the ground state. In addition, we also found that the intermolecular hydrogen bond in the T1 state is weaker than that in the S1 state. Therefore, the strengthened intermolecular hydrogen bond in the S1 state will become weakened in the T1 state after the intersystem crossing (ISC) process from the S1 state to the T1 state. The hydrogen bond strengthening in the triplet excited state of fluorenone is consistent with the triplet excited-state spectral red-shift due to the formation of an intermolecular hydrogen bond. Furthermore, frontier molecular orbital (MO) analysis also confirmed that the intermolecular hydrogen bond can be strengthened in the T1 state in comparison with that in the ground state. The excited-state hydrogen bonding dynamics in the T1 and S1 states of the fluorenone chromophore in alcoholic solvents may play an important role in the competition between IC and ISC processes during S1 state deactivation.
Acknowledgements This work was supported by NSFC (Nos 20903094 and 20833008) and NKBRSF (Nos 2007CB815202 and 2009CB220010).
References 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15.
J. D. Simon, Acc. Chem. Res., 21, 128 (1988). R. Jimenez, G. R. Fleming, P. V. Kumar and M. Maroncelli, Nature, 369, 471 (1994). E. Pines, D. Pines, Y. Z. Ma and G. R. Fleming, ChemPhysChem, 5, 1315 (2004). M. Glasbeek and H. Zhang, Chem. Rev., 104, 1929 (2004). W. P. de Boeij, M. S. Pshenichnikov and D. A. Wiersma, Annu. Rev. Phys. Chem., 49, 99 (1998). G. J. Kearley, F. Fillaux, M. H. Baron et al., Science, 264, 1285 (1994). A. Douhal, S. H. Kim and A. H. Zewail, Nature, 378, 260 (1995). R. Coquet, M. Tada and Y. Iwasawa, Phys. Chem. Chem. Phys., 9, 6040 (2007). M. Tada, N. Kojima, Y. Izumi et al., J. Phys. Chem. B, 109, 9905 (2005). M. Nagasaka, H. Kondoh, K. Amemiya et al., Phys. Rev. Lett., 100, 106101 (2008). G.-J. Zhao, Y.-H. Liu, K.-L. Han and Y. Dou, Chem. Phys. Lett., 453, 29 (2008). G.-J. Zhao, K.-L. Han, Y.-B. Lei and Y. Dou, J. Chem. Phys., 127, 094307 (2007). W. J. D. Beenken, M. Sun, G.-J. Zhao and T. Pullerits, Phys. Status Solidi B, 245, 849 (2008). S.-B. Wan, S.-S. Liu, G.-J. Zhao et al., Biophys. Chem., 129, 218 (2007). L.-C. Zhou, G.-J. Zhao, J.-Y. Liu et al., J. Photochem. Photobiol. A: Chem., 187, 305 (2007).
Insight from Singlet into Triplet Excited-State Hydrogen Bonding Dynamics in Solution 157 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55. 56. 57. 58. 59. 60. 61. 62. 63. 64. 65. 66.
L.-C. Zhou, J.-Y. Liu, G.-J. Zhao et al., Chem. Phys., 333, 179 (2007). R.-K. Chen, G.-J. Zhao, X.-C. Yang et al., J. Mol. Struct., 876, 102 (2008). D. J. Nesbitt and F. Dong, Phys. Chem. Chem. Phys., 10, 2113 (2008). D. J. Nesbitt and R. W. Field, J. Phys. Chem., 100, 12735 (1996). M. Yoshiteru, T. Ebata and N. Mikami, J. Phys. Chem. A, 106, 5591 (2002). M. Rueda, F. J. Luque, J. M. Lo´pez and M. Orozco, J. Phys. Chem. A, 105, 6575 (2001). M. Fores, M. Duran, M. Sola et al., J. Phys. Chem. A, 103, 4525 (1999). E. Cubero, M. Orozco and F. J. Luque, Chem. Phys. Lett., 310, 445 (1999). A. Dkhissi, L. Houben, J. Smets et al., J. Mol. Struct., 484, 215 (1999). L. Lapinski, M. J. Nowak, J. Fulara et al., J. Phys. Chem., 96, 6250 (1992). P. Zielke and M. A. Suhm, Phys. Chem. Chem. Phys., 8, 2826 (2006). N. Borho, M. A. Suhm, K. Le Barbu-Debus and A. Zehnacker, Phys. Chem. Chem. Phys., 8, 4449 (2006). T. N. Wassermann, C. A. Rice, M. A. Suhm and D. Luckhaus, J. Chem. Phys., 127, 234309 (2007). P. Zielke and M. A. Suhm, Phys. Chem. Chem. Phys., 9, 4528 (2007). O. David, C. Dedonder-Lardeux, C. Jouvet et al., J. Chem. Phys., 120, 10101 (2004). N. Tsuji, S. Ishiuchi, M. Sakai et al., Phys. Chem. Chem. Phys., 8, 114 (2006). K. C. Hunter, L. R. Rutledge and S. D. Wetmore, J. Phys. Chem. A, 109, 9554 (2005). F. L. Gervasio, P. Procacci, G. Cardini et al., J. Phys. Chem. B, 104, 1108 (2000). M. Pagliai, G. Cardini, R. Righini and V. Schettino, J. Chem. Phys., 119, 6655 (2003). R. Chelli, V. Schettino and P. Procacci, J. Chem. Phys., 122, 234107 (2005). P. Li and Y. X. Bu, J. Chem. Phys., 121, 9971 (2004). X. H. Chen and Y. X. Bu, J. Am. Chem. Soc., 129, 9713 (2007). P. Li and Y. X. Bu, J. Comput. Chem., 26, 552 (2005). L. B. Zhang and Y. X. Bu, J. Phys. Chem. B, 112, 10723 (2008). H.-C. Chang, J.-C. Jiang, W.-W. Lai et al., J. Phys. Chem. B, 109, 23103 (2005). P. Pyykko and F. Elmi, Phys. Chem. Chem. Phys., 10, 3867 (2008). P. Pyykko, Chem. Rev., 97, 597 (1997). N. Balucani, D. Skouteris, G. Capozza et al., Phys. Chem. Chem. Phys., 6, 5007 (2004). N. Balucani, G. Capozza, F. Leonori et al., Int. Rev. Phys. Chem., 25, 109 (2006). Y. Komoto, K. Sakota and H. Sekiya, Chem. Phys. Lett., 406, 15 (2005). H. Ishikawa, K. Iwata and H. Hamaguchi, J. Phys. Chem. A, 106, 2305 (2002). I. Petkova, M. S. Mudadu, A. Singh et al., J. Phys. Chem. A, 111, 11400 (2007). M. Y. Ballester and A. J. C. Varandas, Phys. Chem. Chem. Phys., 7, 2305 (2005). A. Alijah and A. J. C. Varandas, J. Chem. Phys., 129, 034303 (2008). M. Moreno, A. Douhal, J. M. Lluch et al., J. Phys. Chem. A, 105, 3887 (2001). E. Cubero, F. J. Luque and M. Orozco, J. Am. Chem. Soc., 123, 12018 (2001). G.-J. Zhao and K.-L. Han, J. Phys. Chem. A, 111, 2469 (2007). G.-J. Zhao and K.-L. Han, J. Phys. Chem. A, 111, 9128 (2007). G.-J. Zhao, J.-Y. Liu, L.-C. Zhou and K.-L. Han, J. Phys. Chem. B, 111, 8940 (2007). G.-J. Zhao and K.-L. Han, J. Chem. Phys., 127, 024306 (2007). G.-J. Zhao and K.-L. Han, Biophys. J., 94, 38 (2008). G.-J. Zhao and K.-L. Han, J. Comput. Chem., 29, 2010 (2008). G.-J. Zhao, R.-K. Chen, M.-T. Sun et al., Chem. Eur. J., 14, 6935 (2008). G.-J. Zhao and K.-L. Han, ChemPhysChem, 9, 1842 (2008). T. S. Zwier, Annu. Rev. Phys. Chem., 47, 205 (1996). C. A. Southern, D. H. Levy, G. M. Florio et al., J. Phys. Chem. A, 107, 4032 (2003). Y. Liu, J. Ding, D. Shi and J. Sun, J. Phys. Chem. A, 112, 6244 (2008). Y. F. Liu, J. X. Ding, R. Q. Liu et al., J. Comput. Chem., 30, 2723 (2009). Y. F. Liu, J. X. Ding, R. Q. Liu et al., J. Photochem. Photobiol. A: Chem., 201, 203 (2009). G.-J. Zhao and K.-L. Han, J. Phys. Chem. A, 113, 4788 (2009). G.-J. Zhao and K.-L. Han, J. Phys. Chem. A, 113, 14329 (2009).
158 Hydrogen Bonding and Transfer in the Excited State 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.
G.-J. Zhao, K.-L. Han and P. J. Stang, J. Chem. Theory Comput., 5, 1955 (2009). S. Chai, G.-J. Zhao, P. Song et al., Phys. Chem. Chem. Phys., 11, 4385 (2009). A. L. Sobolewski and W. Domcke, J. Phys. Chem. A, 111, 11725 (2007). L. Biczok, T. Berces, T. Yatsuhashi et al., Phys. Chem. Chem. Phys., 3, 980 (2001). J. L. Jamison, L. Davenport and B. W. Williams, Chem. Phys. Lett., 422, 30 (2006). L. Biczok, T. Berces and H. Inoue, J. Phys. Chem. A, 103, 3837 (1999). M. Y. Odoi, N. I. Hammer, H. P. Rathnayake et al., ChemPhysChem, 8, 1481 (2007). V. Samant, A. K. Singh, G. Ramakrishna et al., J. Phys. Chem. A, 109, 8693 (2005). Y. Matsushita, T. Ichimura and T. Hikida, Chem. Phys. Lett., 360, 65 (2002). L. Biczok, T. Berces and H. Linschitz, J. Am. Chem. Soc., 119, 11071 (1997). M. K. Nayak and S. K. Dogra, J. Photochem. Photobiol. A: Chem., 169, 79 (2005). A. Morimoto, L. Biczok, T. Yatsuhashi et al., J. Phys. Chem. A, 106, 10089 (2002). H. Kawabata and H. Tachikawa, Phys. Chem. Chem. Phys., 5, 3587 (2003). G. A. Zalesskaya, D. L. Yakovlev, E. G. Sambor and N. N. Bely, Phys. Chem. Chem. Phys., 4, 5634 (2002). S. I. Druzhinin, Y. B. Jiang, A. Demeter and K. A. Zachariasse, Phys. Chem. Chem. Phys., 3, 5213 (2001). T. Tahara, H. Hamaguchi and M. Tasumi, Chem. Phys. Lett., 152, 5875 (1987). T. Tahara, H. Hamaguchi and M. Tasumi, J. Phys. Chem., 91, 135 (1988). T. Tahara, H. Hamaguchi and M. Tasumi, J. Phys. Chem., 94, 171 (1990). S. Tanaka, C. Kato, K. Horie and H. Hamaguchi, Chem. Phys. Lett., 381, 385 (2003). O. Treutler and R. Ahlrichs, J. Chem. Phys., 102, 346 (1995). M. V. Arnim and R. Ahlrichs, J. Comput. Chem., 19, 1746 (1998). R. Ahlrichs, M. B€ar, M. H€aser et al., Chem. Phys. Lett., 162, 165 (1989). A. Sch€afer, C. Huber and R. Ahlrichs, J. Chem. Phys., 100, 5829 (1994). J. A. Aquino, H. Lischka and C. H€attig, J. Phys. Chem. A, 109, 3201 (2005). S. Grimme, F. Furche and R. Ahlrichs, Chem. Phys. Lett., 361, 321 (2002). F. Furche and R. Ahlrichs, J. Chem. Phys., 117, 7433 (2002).
7 Probing Dynamic Heterogeneity in Nanoconfined Systems: the Femtosecond Excitation Wavelength Dependence and Fluorescence Correlation Spectroscopy Shantanu Dey, Ujjwal Mandal, Aniruddha Adhikari, Subhadip Ghosh and Kankan Bhattacharyya Department of Physical Chemistry, Indian Association for the Cultivation of Science, Jadavpur, Kolkata 700 032, India
7.1 Introduction Liquids confined in a nanocavity play a key role in many natural processes. Prominent examples of nanoconfined liquids are water in the hydrophobic pocket of a protein or in a biological cell or many liquids in a nanoporous catalyst and in other materials. In recent years, femtosecond spectroscopy and computer simulations have generated a considerable amount of new knowledge on the dynamics of liquids confined in a nanocavity [1]. The nanoassemblies are heterogeneous on a molecular length scale. As a result, the spectra and dynamics of a fluorescent probe in different regions of such a nanoassembly are distinctly different. In an attempt spatially to resolve ultrafast dynamics, we have recently applied the excitation wavelength (lex) dependence. This method is based on the simple fact that molecules in different environments are spectrally distinct, and, as a result, at different lex, different subsets of molecules are excited. This is the basis of the so-called red-edge excitation shift (REES) which is observed in many organized assemblies [2–4]. Using this strategy, we have been able to delineate the difference in several ultrafast phenomena in different regions of a micelle, reverse micelle, gel and lipid vesicle. The most interesting observation is that the lex dependence shows even a neat ionic liquid to be heterogeneous. The spatial resolution of a microscope is l/2, i.e. 200 nm for 400 nm excitation. As a fluorescent probe of 1 nm dimension probes its immediate neighbourhood, the spatial resolution achieved by lex variation exceeds the spatial resolution of a microscope. Hydrogen Bonding and Transfer in the Excited State, Volume I & II © 2011 John Wiley & Sons, Ltd. ISBN: 978-0-470-66677-7
Edited by Ke-Li Han and Guang-Jiu Zhao
160 Hydrogen Bonding and Transfer in the Excited State
In this review, we will concentrate on ultrafast solvation dynamics, excited-state proton transfer (ESPT), fluorescence resonance energy transfer (FRET) and fluorescence correlation spectroscopy (FCS) in several nanoconfined systems. The latter include protein, micelles, ionic liquids, a cyclodextrin host–guest complex and others.
7.2 Solvation Dynamics in Nanoconfined Systems Solvation in a nanocavity is fundamental in many biological and catalytic processes. We begin with solvation dynamics in a few nanoconfined assemblies.
7.2.1 Solvation dynamics: basic features Solvation dynamics probes the time evolution of interaction between a solute dipole with polar solvent molecules. On excitation of certain solutes by an ultrafast laser, a dipole is created suddenly. With the passage of time, as the solvent molecules reorganize, the energy of the solute dipole decreases. This causes a gradual shift of the fluorescence maximum to lower energy, i.e. towards a longer wavelength. This is known as the dynamic Stokes shift (DSS). The gradual change in solvation (i.e. solvation dynamics) is monitored by the decay of the solvation time correlation function, C(t), which is defined as [5] CðtÞ ¼
nðtÞnð1Þ nð0Þnð1Þ
ð7:1Þ
where n(0), n(t), and n(1) are the observed emission frequencies at time zero, t and infinity respectively. Solvation dynamics in bulk water and several other polar liquids (e.g. methanol, acetonitrile, etc.) may be described by a major component on a 0.1 ps (100 fs) timescale and a minor component of 1 ps [6]. Interestingly, the solvation dynamics of these polar liquids confined in many organized and biological assemblies displays a component on a 100–1000 ps timescale [1]. Many aspects of the ultraslow component have been reviewed earlier [1]. In this account, we focus only on the spatial or lex variation. 7.2.2 Solvation dynamics in ionic liquid A room-temperature ionic liquid (RTIL) consists of an extended positive ion with a large organic moiety and a relatively smaller negative ion. The steric hindrance between the ions frustrates the formation of crystals and consequently makes the melting point quite low, while the interionic attraction raises the boiling point and lowers the vapour pressure. Recent experimental studies and computer simulations suggest that an ionic liquid forms nanostructural aggregates if the alkyl group is sufficiently large [7, 8]. Very recent small-angle X-ray scattering and optical Kerr effect studies have indicated the formation of nanosized aggregates (1.3–2.7 nm) in imidazolium ionic liquids containing an alkyl group with more than three carbon atoms [9]. In such an aggregate there is a clear segregation of the polar (ionic) domain and the non-polar (alkyl group) domains. The structure and dynamics of the bulky ions in an RTIL play a key role in polar chemical reactions. The solvation dynamics in ionic liquids has been studied by ultrafast laser spectroscopy and large-scale computer simulation [10–13]. The early studies [10–13] focused mainly on the role of the counterions on solvation dynamics, but did not consider the presence of spatial heterogeneity in an RTIL.
Probing Dynamic Heterogeneity in Nanoconfined Systems
161
7.2.2.1 Neat Ionic Liquid: kex Dependence and Dynamic Heterogeneity Adhikari et al. have recently studied the lex dependence of solvation dynamics and anisotropy decay in a neat room-temperature ionic liquid (RTIL), 1-pentyl-3-methyl-imidazolium tetrafluoroborate ([pmim][BF4]) [4a]. In neat IL, heterogeneity may arise from clustering of the pentyl groups surrounded by a network of the cationic head group and anions. Using the excitation wavelength dependence, they spatially resolved the dynamics in different regions of neat [pmim][BF4]. The solvation dynamics of C480 in neat [pmim] [BF4] was found to depend on lex (Figure 7.1). For lex ¼ 375 nm, the decay of the solvent correlation function, C(t), exhibits an average solvation time of 860 ps. However, with increase in lex the average solvation time decreases by a factor of 6 to 135 ps at lex ¼ 435 nm. The faster solvation dynamics at long lex may arise from the polar region containing the BF4 anion, while the slower solvation at short lex may be ascribed to the nonpolar domain around the alkyl chains. A similar lex dependence of solvation dynamics was also reported by Maroncelli and coworkers [11a]. In contrast to solvation dynamics, the fluorescence anisotropy decay does not exhibit any lex dependence in neat [pmim][BF4]. In neat [pmim][BF4], C480 exhibits a very slow rotational dynamics with a single exponential decay of time constant 3800 ps. It may be noted that in bulk water the time constant of fluorescence anisotropy decay of C480 is much faster, 70 ps [14]. The slow anisotropy decay in neat ionic liquid may obviously be ascribed to the high viscosity of [pmim][BF4]. The rotational diffusion of the large (1 nm) probe molecule sweeps almost the entire ([pmim][BF4]) aggregate (1.5 nm in size). During this motion, the probe experiences an average friction, and hence there is no lex or location dependence in neat RTIL. 7.2.2.2 RTIL Microemulsion An ionic liquid may be sequestered inside a reverse micelle in the form of a ‘pool’. Using small-angle neutron scattering studies (SANS), Eastoe and coworkers showed that the ionic liquid pool is ellipsoidal in shape with a semi-minor radius of 2.4 nm and a length 11 nm for an equimolar ratio of ([bmim][BF4]) and the surfactant
Neat [pmim][BF4]
0.8
λ ex
375 nm 405 nm 435 nm
C(t)
0.6
0.4
0.2 0
50
100
150
200
Time (ps)
Figure 7.1 Decay of the solvent response function, C(t), of C480 in neat [pmim][BF4] for lex ¼ 375 nm (D), lex ¼ 405 nm (*) and lex ¼ 435 nm (&). The points denote the actual values of C(t), and the solid lines denote the best fit. Reprinted with permission from [4a]. Copyright 2007 American Chemical Society
162 Hydrogen Bonding and Transfer in the Excited State
(TX-100) [15a]. Gao et al. reported that as much as 6 wt% water may be encapsulated in the polar nanodomain of such a microemulsion [15b]. Solvation dynamics of C480 in the [pmim][BF4]/TX-100/benzene microemulsion is found to exhibit a much stronger lex dependence compared with neat RTIL [pmim][BF4]. For all lex, the decay of C(t) in this microemulsion may be described by a triexponential decay – a very slow (2200 ps), a slow (200 ps) and a very fast (2 ps) component [4a]. Apart from this, there is an ultrafast component ( NH proton of the imidazole ring to form a monoanion. In monoanonic form, the TICT R R
O O
H H
H
N
N
N R
O
N
CH3 CH3
H R
Figure 14.11
O H
Hydrated 2-(40 -N,N-dimethylaminophenyl)pyrido[3,4-d]imidazole
Hydrogen-Bonding Effects on Intramolecular Charge Transfer CH3
N
N
323
N N H
CH3
Neutral (N) N
CH3
N
N
N N
N H
CH3
Monoanion (MA) H
N
N N H
CH3
N
N
CH3
N
N H
CH3
Monocation 2 (MC2)
CH3
Monocation 3 (MC3)
N*
MC2*
(dual emission in protic solvents)
(normal emission)
hv hv
H CH3
Monocation 1 (MC1)
H N
CH3
N
N
Biprotonicphototautomerism
MC1 + MC3
H+
hv
N - H+
MC3* (TICT emission)
MA
hv MA* (dual emission with enhanced TICT)
Figure 14.12
2-(40 -N,N-dimethylaminophenyl)pyrido[3,4-d]imidazole and its ionic forms
emission was enhanced substantially compared with the neutral form. A similar increase in TICT emission was also reported for dimethylaminobenzoic acid [80]. AM1 calculations predicted that in both DMAPPI and dimethylaminobenzoic acid the acceptor becomes more planar in the anionic form than in the neutral form. From this it was gleaned that increase in planarity between the acceptor and the phenyl ring is responsible for the enhanced TICT emission. At pH 4.0, DMAPPI forms two kinds of monocation: (i) protonated at the dimethylamino nitrogen (MC1); (ii) protonated at the pyridine nitrogen (MC3) in the ground state. Upon excitation, the monocation formed by the protonation of dimethylamino nitrogen is deprotonated and undergoes biprotonic phototautomerism to protonate the imidazole nitrogen (MC2). However, the other monocation, formed by the protonation of pyridine nitrogen, was also observed in the excited state. However, it emits from the TICT state and not from the LE state. Again based on prediction of semi-empirical calculation, the argument concerning planarity between the acceptor and the phenyl ring promoting TICT formation was extended to the monocation also. However, all three dications and trication (Figure 14.13) show only normal fluorescence. Since solvatochromic and prototropic studies of DMAPPI did not yield a definite conclusion concerning the role of hydrogen bonding of the solvent with the donor of DMAPPI, the studies were extended to
324 Hydrogen Bonding and Transfer in the Excited State H N
N
CH3 N
HN
N H H N
H CH3 CH3
N
HN
N H
CH3
N
CH3 N
HN
N H H N
H CH3
CH3 N
N H
H CH3
Figure 14.13 Dications and trication of 2-(40 -N,N-dimethylaminophenyl)pyrido[3,4-d]imidazole
cyclodextrins, with an expectation that encapsulation of either donor or acceptor by cyclodextrins may lead to a conclusion concerning the role of hydrogen bonding of the solvent with the donor in the formation of the TICT state (Table 14.1) [81]. DMAPPI forms a 1:1 inclusion complex with cyclodextrins. The red-shift observed in the absorption spectra in cyclodextrins clearly indicates that DMAPPI enters the cavity by breaking the hydrogen bond between the dimethylamino nitrogen and water (Figure 14.14). Owing to its size, DMAPPI was only partially encapsulated in cyclodextrins. The pyridoimidazole ring of DMAPPI was outside the cavity. The water molecules form hydrogen bonds with acidic and basic centres of the pyridoimidazole ring. Such hydrogen bonding ensured the presence of different solvated structures and was confirmed by the red-shift in the fluorescence excitation spectra with lem. There is an enormous increase in TICT emission of DMAPPI in cyclodextrins, for instance it was enhanced by a factor of 45 in b-cyclodextrin with respect to an aqueous medium. The destabilization of the TICT state in the non-polar cyclodextrin cavity increased the energy gap between the TICT state and the low-lying states, thereby lowering the non-radiative rates. Blue-shift in the normal and TICT emissions band maxima substantiates this. Observation of TICT emission in the cyclodextrin inclusion complexes suggests that it is the hydrogen bonding of protic solvents with the pyridoimidazole ring, i.e. the acceptor, that is responsible for its protic-solvent-induced TICT emission. The donor (the dimethylamino group) was buried inside the cyclodextrin cavity and was not able to form hydrogen bonds either with the protons of the hydroxyl rim of the cavity or with water molecules at the interface. This rules out the role of hydrogen bonding of solvent with donor in the formation of the TICT state in DMAPPI. Destabilization of the TICT state inside the cyclodextrin cavity indicates that the donor plays a major role in the dipole–dipole
H H N
H
N N H N H
Figure 14.14
H H
2-(40 -N,N-dimethylaminophenyl)pyrido[3,4-d]imidazole b-cyclodextrin inclusion complex
Hydrogen-Bonding Effects on Intramolecular Charge Transfer
325
interaction solvation stabilization of the TICT state. The conclusions were further substantiated by the observed enhancement of TICT emission of MC3, the monocation formed by the protonation of pyridine nitrogen of DMAPPI in cyclodextrins. There, too, the donor was buried inside the cavity. One interesting aspect about DMAPPI is that, owing to charged interactions, its TICT species has different binding constants with ionic micelles compared with normal species, and the binding constants are the same with non-ionic micelles [82]. Different lifetimes observed for TICT and normal species of DMAPPI suggest that the equilibrium was not established between the two states. This may be due to greater stabilization of the TICT state by H-bonding with the acceptor, which should increase the activation energy barrier for the reverse process. Like DMAPPI, 2-(40 -N,N-dimethylaminophenyl)imidazo[4,5-b]pyridine (DMAPIP-b) (Figure 14.15) also shows longer-wavelength emission only in protic solvents [83]. The ratio of normal emission to TICT emission increases with increasing H-donating capacity of the solvent. However, in water the TICT emission is almost completely quenched. Lifetime data suggested that equilibrium is not established between the two emitting states (Table 14.2). The intensity ratio of longer-wavelength emission to shorter-wavelength emission decreases with increase in viscosity of the protic solvent, suggesting the presence of a viscosity-dependent barrier in the formation, i.e. common for TICT emission. On decrease in pH, unlike DMAPPI, only one fluorescence band is observed for DMAPIP-b in water at pH 4.0. This is due to the normal emission of a monocation formed by protonation on imidazole nitrogen of DMAPIP-b. The fluorescence spectra of DMAPIP-b are more red-shifted than those of DMAPPI, indicating a better charge transfer interaction when the position of the nitrogen changes. This is supported by the larger change in the dipole moment between the excited state and ground state, calculated by a Lippert–Mataga plot [84] for DMAPIP-b (7.1 D), than that calculated for DMAPPI (5.3 D) [77]. The longer-wavelength emission is also shifted to red in DMAPIP-b. The fluorescence intensity of the band is also enhanced enormously. For example, in contrast to DMAPPI, in DMAPIP-b the longer-wavelength emission in 1-butanol, 1-propanol, 2-propanol and ethanol appears as a clear shoulder. This shows that the long-wavelength emission is also strongly influenced by the position of the pyridine nitrogen. This was further substantiated by ab initio calculations showing that the electron density on
H N
CH3 N
N H (Normal emission)
N H
+
CH3
CH3
N N N
N H (Normal emission)
CH3 ROH
CH3
N N N HOR
Figure 14.15
N H
CH3 (Normal and TICT emission)
2-(40 -N,N-dimethylaminophenyl)imidazo[4,5-b]pyridine and its monocation
326 Hydrogen Bonding and Transfer in the Excited State Table 14.2 Absorption band maximum (lab, nm), fluorescence band maximum (lf, nm) and lifetimes (t, ns) of DMAPIP-b in different solventsa Solvent
lab
lf
t
Cyclohexane Ethyl acetate Acetonitrile Dimethylformamide 1-Butanol
336, 352 338 (4.49) 345 (4.50) 346 (4.50) 348 (4.47)
1-Propanol
349 (4.48)
Ethanol
350 (4.48)
Methanol
350 (4.50)
Glycol
358 (4.50)
Glycerol Water
360 345
359, 379, 398 390 407 428 407 475 410 486 413 494 414 506 428 526 446 451
1.99 1.28 1.50 1.49 0.98 (42.48) 2.45 (57.52) 0.90 (24.22) 2.27 (75.78) 0.71 (24.22) 1.95 (75.78) 0.29 (67.70) 1.12 (32.30) 0.48 (78.17) 1.11 (21.83) 0.16 (98.75) 2.13 (01.25)
From Ref. [83].
a
dimethylamino nitrogen is less in DMAPIP-b than in DMAPPI and on pyridine nitrogen is more in DMAPIP-b than in DMAPPI (Table 14.3). The ab initio calculation performed predicted that both in DMAPPI and DMAPIP-b the donor (dimethylamino group), the phenyl ring and the acceptor (pyridoimidazole moiety) are coplanar in the ground state [83]. DFT calculations performed later for DMAPIP-b also predicted that the phenyl ring and the pyridoimidazole moiety are in same molecular plane [85]. This virtually rules out the earlier prediction based on the semi-empirical calculations that hydrogen bonding of solvent with acceptor makes it more planar with the phenyl ring, increasing the charge flow from phenyl ring to acceptor and being responsible for the formation of TICT emission in these molecules. Nonetheless, the hydrogen-bond-induced TICT emission in DMAPPI and DMAPIP-b can be explained as follows: hydrogen bonding of protic solvents with pyridine nitrogen enhances the electron affinity of the acceptors, thereby favouring the formation of the TICT state in DMAPPI and DMAPIP-b. The following evidence substantiates the explanation: (i) H-bonding-induced TICT emission is observed only in DMAPIP-b and in DMAPPI, but not in (N,Ndimethylaminophenyl)benzimidazole, where pyridine nitrogen is absent; (ii) when imidazole nitrogen is protonated, no TICT emission is observed in DMAPIP-b or in DMAPPI. On the other hand, the monocations formed by the protonation of pyridine nitrogen of DMAPPI and DMAPIP-b emit only TICT emission, and (iii) Table 14.3 Theoretical parameters obtained by ab initio calculation on DMAPPI and DMAPIP-b
Phenyl ring/NMe2 dihedral angle Phenyl/PI dihedral angle Charge on pyridine nitrogen Charge on dimethylamino nitrogen mg (D)
DMAPPI
DMAPIP-b
0.0 0.0 0.716 0.844 4.5
0.0 0.0 0.690 0.955 7.2
Hydrogen-Bonding Effects on Intramolecular Charge Transfer
327
huge enhancement in TICT is observed in cyclodextrin inclusion complexes of DMAPPI, where pyridine nitrogen is involved in hydrogen bonding with water [77, 79, 81, 83, 85]. Interestingly, unlike DMAPPI, DMAPIP-b emits dual emission only in sodium dodecylsulfate and Triton X-100 micelles, and not in cetyltrimethyl ammonium bromide micelle, which has a relatively poor hydrogen-bond-donating capacity [86]. Thus, the suggestion by Fansi et al. that twisting of the acceptor by hydrogen of the solvent is a mechanism for protic-solvent-induced TICT emission was not supported by theoretical calculation. Although semiempirical calculation predicted that hydrogen bonding of the solvent with the acceptor makes it planar with the benzene ring, this is not supported by ab initio calculations. Most experimental evidence suggests that hydrogen bonding of the solvent with the acceptor enhances its electron affinity, thereby inducing TICT emission.
14.5 Conclusion Like general interaction, hydrogen bonding also plays a major role in the formation and stabilization of ICT states in many molecules. The hydrogen bonding of protic solvents with acceptor groups of the molecules plays a major role in the formation and stabilization of ICT states. Hydrogen bonding of the solvent with the acceptor moiety increases the electron affinity of the acceptor group, thereby favouring ICT emission. In some molecules, such hydrogen bonding is essential for lowering the energy level of the ICT state. The hydrogen bond between protic solvent and donor breaks in the excited state to form the ICT state. Most experimental evidence suggests that it does not play an important role in the formation of ICT emission in solution phase. This was substantiated by the enormous increase in longer-wavelength emission of molecules like DMAPPI in cyclodextrins, where the donor is buried inside the cavity and is not available for hydrogen bonding. However, dipole–dipole interaction between donor and polar solvent is also one of the important factors that is responsible for solvent stabilization of ICT states. In some cases, equilibrium is not established between the LE state and the ICT state in protic solvents. This is due to greater stabilization of the ICT state by hydrogen bonding of protic solvents with acceptor moieties, which increases the barrier for the ICT ! LE reverse process.
Acknowledgements The author expresses his gratitude to Prof. S. K. Dogra who introduced him to this exciting research area. The author thanks the Department of Science and Technology (DST), New Delhi, and the Council for Scientific and Industrial Research (CSIR), New Delhi, for financial support. The author also thanks Prof. A. K. Mishra and Dr U. Subuddhi for their valuable suggestions and comments.
References 1. W. Rettig, Angew. Chem. Int. Ed. Engl., 25, 971 (1986). 2. Z. R. Grabowski, K. Rotkiewiczand W. Rettig, Chem. Rev., 103, 3899 (2003). 3. E. Lippert, W. L€uder and H. Boos, in Advances in Molecular Spectroscopy, ed. by A. Mangini. Pergamon Press, Oxford, UK, p 443 (1962). 4. B. Valuer, Molecular Fluorescence Principles and Applications. Wiley-VCH, Weinheim, Germany, p. 62 (2002). 5. W. Rettig, Topics in Current Chemistry, Vol. 169, Electron Transfer I, ed. by J. Mattay. Springer-Verlag, Berlin, Germany, p. 253 (1994).
328 Hydrogen Bonding and Transfer in the Excited State 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. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53.
T. Soujanya, R. W. Fesseanden and A. Samanta, J. Phys. Chem., 100, 3507 (1996). V. I. Stsiapura, A. A. Maskevich, V. A. Kuzmitsky et al., J. Phys. Chem., A, 111, 4829 (2007). K. Rotkiewicz, K. H. Grellmann and Z. R. Grabowski, Chem. Phys. Lett., 19, 315; 21, 212 (1973). K. Rotkiewicz, Z. R. Grabowski, A. Krowczynski, and W. Kuhnle, J. Lumin., 12–13, 877 (1976). W. Rettig, K. Rotkiewicz and W. Rubaszewska, Spectrochim. Acta, A, 40, 241 (1984). G. Wermuth and W. Rettig, J. Phys. Chem., 88, 2729 (1984). W. Rettig and R. Gleiter, J. Phys. Chem., 89, 4676 (1985). Z. R. Grabowski, K. Rotkiewicz, A. Siemiarczuk et al., Nouv. J. Chim., 3, 443 (1979). K. Rotkiewicz and W. Rubaszewska, Chem. Phys. Lett., 70, 444 (1980). W. Rettig and F. Marschner, Nouv. J. Chim., 7, 425 (1983). H. Bischof, W. Baumann, N. Detzer and K. Rotkiewicz, Chem. Phys. Lett., 116, 180 (1985). U. Leinhos, W. Kuhnle and K. A. Zachariasse, J. Phys. Chem., 95, 2013 (1991). K. A. Zachariasse, T. von der Haar, A. Hebecker et al., Pure Appl. Chem., 65, 1745 (1993). W. Schuddeboom, S. A. Jonker, J. M. Warman et al., J. Phys. Chem., 96, 10 809 (1992). A. L. Sobolewski and W. Domcke, Chem. Phys. Lett., 250, 428 (1996). K. A. Zachariasse, M. Grobys and E. Tauer, Chem. Phys. Lett., 274, 372 (1997). K. A. Zachariasse, Chem. Phys. Lett., 320, 8 (2000). J. Dobkowski, J. Wojcik, W. Kozminski et al., J. Am. Chem. Soc., 124, 2406 (2002). T. Yoshihara, S. I. Druzhinin and K. A. Zachariasse, J. Am. Chem. Soc., 126, 8535 (2004). S. Cogan, S. Zilberg and Y. Haas, J. Am. Chem. Soc., 128, 3335 (2006). E. Lippert, W. Luder and H. Boos, in Advances in Molecular Spectroscopy, ed. by A. Mangini. Pergamon Press, Oxford, UK, p. 443 (1962). W. Rettig, J. Mol. Struct., 84, 303 (1982). P. Suppan, Chem. Phys. Lett., 128, 160 (1986). C. Cazeau-Dubroca, A. Peirigua, S. Ait Lyazidi and G. Nouchi, Chem. Phys. Lett., 98, 511 (1982). D. Pilloud, P. Suppan and L. van Haelst, Chem. Phys. Lett., 137, 130 (1987). W. Baumann, H. Bischof, J. C. Frohling et al., J. Photochem. Photobiol. A: Chem., 64, 49 (1992). M. J. Kamlet, J. L. M. Abboud and R. W. Taft, Prog. Phys. Org. Chem., 13, 485 (1981). M. Kupfer and W. Abraham, J. Prakt. Chem., 325, 95 (1983). A. Muller, H. Ratajczak, W. Junge and E. Diemann, Electron and Proton Transfer in Chemistry and Biology. Elsevier, Amsterdam, The Netherlands (1992). R. L. Cukier, J. Phys. Chem., 98, 2377 (1994) and references therein. Z. R. Grabowski, K. Rotkiewicz, A. Siemiarczuk et al., Nouv. J. Chim., 3, 443 (1979). M. Zander and W. Rettig, Chem. Phys. Lett., 110, 602 (1984). K. A. Zachariasse, T. V. D. Haar, A. Hebecker et al., Pure Appl. Chem., 65, 1745 (1993). J. Hicks, M. Vandersall, Z. Babarogic and K. E. Eisenthal, Chem. Phys. Lett., 116, 18 (1985). N. Chattopadhyay, M. V. Auweaer and F. C. De Schryver, Chem. Phys. Lett., 279, 303 (1997). D. Majumdar, R. Sen, K. Bhattacharyya and S. P. Bhattacharyya, J. Phys. Chem., 95, 4324 (1991). A. Siemercuk, Z. R. Grabowski, M. Asher and M. Ottalenghi, Chem. Phys. Lett., 51, 315 (1977). J. Hicks, M. Vandersall, Z. Baraogic and K. B. Eisenthal, Chem. Phys. Lett., 116, 180 (1985). K. B. Eisenthal, Topics in Applied Physics, Vol. 60, ed. by W. Kaiser. Springer-Verlag, New York, NY, p. 319 (1988). R. Hayashi, S. Tazuke and C. M. Frank, Macromolecules, 20, 983 (1987). K. A. Al-Hassan and T. C. Aumi, Chem. Phys. Lett., 121, 146 (1988). S. Tazuke and R. K. Guo, Macromolecules, 23, 719 (1990). S. Tazuke, R. K. Guo and T. Ikeda, J. Phys. Chem., 94, 1408 (1990). J. Paczkowski and D. C. Neckers, Macromolecules, 24, 3013 (1991). P. Changenet, P. Plaza, M. M. Martin and Y. H. Meyer, J. Phys. Chem., A, 101, 8186 (1997). W. Rettig and R. Lapouyade, in Topics in Fluorescence Spectroscopy, Vol. 4, Probe Design and Chemical Sensing, ed. by J. R. Lakowicz. Plenum Press, New York. NY, p. 109 (1994). A. Levy, D. Avnir and M. Ottolenghi, Chem. Phys. Lett., 121, 233 (1985). S. Kundu, S. Maity, S. C. Bera and N. Chattopadhyay, J. Mol. Struct., 405, 231 (1997).
Hydrogen-Bonding Effects on Intramolecular Charge Transfer 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. 79. 80. 81. 82. 83. 84. 85. 86.
329
G. Krishnamoorthy and S. K. Dogra, Chem. Phys. Lett., 104, 2542 (2000). G. S. Cox, P. J. Hauptmann and N. J. Turro, Photochem. Photobiol., 39, 597 (1984). A. Nag and K. Bhattacharyya, Chem. Phys. Lett., 151, 474 (1988). K. Bhattacharyya and M. Chowdhury, Chem. Rev., 93, 507 (1993). C. Cazeau-Dubroca, S. Ait Lyazidi, P. Cambou et al., J. Phys. Chem., 93, 2347 (1989). C. Cazeau-Dubroca, A. Peirigua, M. Ben Brahim et al., Chem. Phys. Lett., 157, 393 (1989). C. Cazeau-Dubroca, A. Peirigua, M. Ben Brahim et al., Proc. Indian Acad. Sci. (Chem. Sci.), 104, 209 (1989). C. Cazeau-Dubroca, G. Nouchi, M. Ben Brahim et al., J. Photochem. Photobiol. A: Chem., 80, 125 (1994). C. Cazeau-Dubroca, A. Peirigua, S. Ait-Lyazidi et al., Chem. Phys. Lett., 124, 110 (1986). K. A. Al-Hassan and T. Azumi, Chem. Phys. Lett., 146, 121 (1988). T. Kobayashi, M. Futakami and O. Kajimoto, Chem. Phys. Lett., 130, 63 (1986). R. B. Singh, S. Mahanta, S. Kar and N. Guchhait, Chem. Phys., 324, 33 (2007). S. Mahanta, R. B. Singh, S. Kar and N. Guchhait, J. Photochem. Photobiol. A: Chem., 194, 318 (2008). J. Herbich, Z. R. Grabowski, H. Wo´jtowicz and K. Golankiewicz, J. Phys. Chem., 93, 3439 (1989). J. Herbich, J. Karpiuk, Z. R. Grabowski et al., J. Lumin., 54, 165 (1992). D. Huppert, S. D. Rand, P. M. Rentzepis et al., J. Chem. Phys., 75, 5714 (1981). Y. Wang and K. B. Eisenthal, J. Chem. Phys., 77, 6076 (1982). A. Declemy, C. Rulliere and Ph. Kottis, Chem. Phys. Lett., 101, 401 (1983). G.-J. Zhao and K.-L. Han, J. Comput. Chem., 29, 2010 (2008). M. Zakharov, O. Krauss, Y. Nosenko et al., J. Am. Chem. Soc., 131, 461 (2009). E. Fasani, A. Albini, P. Savarino et al., J. Heterocycl. Chem., 30, 1041 (1993). Y. H. Kim, D. W. Cho, M. Yoon and D. Kim, J. Phys. Chem., 100, 15 670 (1996). Y. Kim, H. W. Cheon, M. Yoon et al., Chem. Phys. Lett., 264, 673 (1997). G. Krishnamoorthy and S. K. Dogra, Spectrochim. Acta, 55A, 2647 (1999). G. Krishnamoorthy and S. K. Dogra, J. Colloid Interface Sci., 228, 335 (2000). G. Krishnamoorthy and S. K. Dogra, J. Org. Chem., 64, 6566 (1999). Y. B. Jiang, J. Photochem. Photobiol., 78, 2 (1994). G. Krishnamoorthy and S. K. Dogra, J. Phys. Chem., 104A, 2542 (2000). G. Krishnamoorthy and S. K. Dogra, Chem. Phys. Lett., 323, 234 (2000). N. Dash, F. A. S. Chipem, R. Swaminathan and G. Krishnamoorthy, Chem. Phys. Lett., 460, 119 (2008). E. Lippert, Z. Electrochem., 61, 962 (1957). N. Dash, F. A. S. Chipem and G. Krishnamoorthy, Photochem. Photobiol. Sci., 8, 1708 (2009). N. Dash and G. Krishnamoorthy, J. Fluoresc., 20, 135 (2010).
15 Chemical Dynamics in Room-Temperature Ionic Liquids: the Role of Hydrogen Bonding Souravi Sarkar, Rajib Pramanik and Nilmoni Sarkar Department of Chemistry, Indian Institute of Technology, Kharagpur, PIN-721302, WB, India
15.1 Photoinduced Electron Transfer in a Room-Temperature Ionic Liquid The term ‘ionic liquid’ has come to indicate a class of molten organic salts that are liquid at room temperature and may contain bulky aromatic moieties such as imidazolium or heterocyclic pyridinium as cation and an inorganic anion such as PF6, BF4, [(CF3SO2)2N], NO3, etc. The room-temperature ionic liquids (RTILs) have some unique properties such as good electrical conductivity, negligiblevapour pressure, high ionic mobility, excellent electrochemical and thermal stability and a wide liquidus temperature range (96 to 300 C) [1–4]. Various photophysical, theoretical and ultrafast spectroscopic studies have been done on RTILs [5–13]. Photoinduced electron transfer (PET) is one of the most fundamental reactions in biological, physical, inorganic and organic chemical systems. The theory regarding electron transfer was first proposed by Marcus [14], and the electron transfer rate can be written as "
kET
ðDGo þ lÞ2 ¼ n exp 4lkB T
# ð15:1Þ
where l is the energy required structurally to reorganize the donor and acceptor, kET is the electron transfer (ET) rate constant, n is the frequency of the motion in the reactant potential well and T and kB are the temperature and Boltzman constant respectively. From this theory it is clear that, with increase in free energy change (DGo), the ET rate initially increases, reaches a maximum at DGo ¼ l and then decreases when DGo > l. The experimental evidence for the bell-shaped dependence of rate constant on DGo has been established only for a first-order reaction having fixed donor–acceptor separation. However, for bimolecular electron
Hydrogen Bonding and Transfer in the Excited State, Volume I & II © 2011 John Wiley & Sons, Ltd. ISBN: 978-0-470-66677-7
Edited by Ke-Li Han and Guang-Jiu Zhao
332 Hydrogen Bonding and Transfer in the Excited State
transfer reactions, the observed rate has the form of a consecutive reaction mechanism consisting of the diffusion (kdiff) and activated rate constant (kact) for electron transfer: kET ¼
kact kdiff kdiff þ kact
ð15:2Þ
The electron transfer process is extremely fast at DGo ¼ l. The rate constant for a bimolecular ET reaction shows an increase with increase in the free energy, reaches a maximum and finally becomes independent of the free energy of the reaction. Recently there have been some simulation studies of model systems of two ionic liquids and comparison with acetonitrile to judge the applicability of Marcus’ theory in ionic liquids [15]. Our group has already studied PET in several confined systems [16]. In this study our aim is to understand how the dynamics of PET is affected in RTILs owing to its unique feature. In order to investigate the dynamics, mechanism and how the PET process is affected by viscosity, polarity and ionic constituents of the RTILs, we have used N,N-dimethyl ethanol ammonium formate (DAF) as a protic non-aromatic room-temperature ionic liquid. We have used several coumarin dyes as the acceptor and dimethyl aniline (DMA) as the donor for PET studies in DAF using steady-state (SS) and time-resolved (TR) fluorescence spectroscopy. We have carried out fluorescence quenching studies by gradual increase of the DMA concentration in the solution of coumarin dyes in DAF. It is observed that, with the addition of DMA, both steady-state fluorescence intensity and the fluorescence lifetime of coumarin dyes in DAF are quenched (Figure 15.1). The fluorescence-quenching constant is determined by the well-known Stern–Volmer equation I0 t0 ¼ 1 þ KSV ½Q ¼ 1 þ kq t0 ½Q ¼ I t
ð15:3Þ
where I0 and I and t0 and t are the fluorescence intensity and lifetime of the coumarin dyes in the absence and in the presence of the quencher, Ksv is the Stern–Volmer constant and [Q] is the quencher concentration. The I0/I versus [Q] and t0/t versus [Q] plots for different probes in the presence of different concentrations of DMA are shown in Figures 15.2(a) and (b). For all coumarin–amine pairs, the plots are linear in nature. The fluorescence quenching constant kq values are determined by dividing the Stern–Volmer constant by the lifetime of the coumarin dyes in the absence of quencher. The measured kq values for different coumarin– amine systems in DAF from steady-state fluorescence are listed in Table 15.1. In this case we can see that the kq values obtained from time-resolved measurement are somewhat smaller than the kq values obtained from steady-state measurement, which indicates a contribution of static quenching to the kq values obtained from steady-state measurement, even though there is no indication of ground-state complex formation from the absorption spectra; also, there is an ultrafast decay component that remains undetected in our time-resolved set-up using picosecond (90 ps) time resolution. The ET rate depends on the free energy change of the system. So we have calculated DGo for each coumarin–amine system. The common expression for calculating DGo is given by the famous Rehm–Weller equation [17] DGo ¼ EðD=D þ ÞEðA=A ÞEIPS E00
ð15:4Þ
where E(D/Dþ ) and E(A/A) denote the oxidation potential of the donor and the reduction potential of the acceptor respectively, EIPS is the Coulombic attraction term and E00 is the energy difference between the S0 and S1 states. The respective DGo value for all the systems are listed in Table 15.1 and shown in Figure 3(b).
Chemical Dynamics in Room-Temperature Ionic Liquids: the Role of Hydrogen Bonding
Fluorescence intensity (normalized)
1.0
(a)
(i)
0.8
333
(ii)
0.6
(iii) (iv)
0.4
(v) (vi) (vii)
0.2 0.0
450
500
550
600
Fluorescence intensity (Counts)
Wavelength (nm)
10000
(b) (ii)
8000
(iii)
6000 4000
(iv)
2000 (i) 0
4
8 Time (ns)
12
Figure 15.1 (a) Steady-state fluorescence spectra of C-151 in the ionic liquid DAF in the presence of total DMA concentration: (i) 0 mM; (ii) 12.62 mM; (iii) 25.25 mM; (iv) 37.87 mM; (v) 50.49 mM; (vi) 63.11 mM; (vii) fluorescence spectra of neat DAF. (b) Time-resolved fluorescence decays of C-153 in the presence of DMA concentration: (i) lamp profile; (ii) 0 mM; (iii) 41.02 mM; (iv) 78.89 mM. Reprinted with permission from the American Chemical Society. Copyright 2009
In a homogeneous solvent, the electron transfer process is mainly governed by the diffusive motion of the reactants, as diffusion is the rate-determining step; in a higher free energy region, the electron transfer rate ultimately levels off with the diffusion rate. We have determined the diffusion constant by the Smoluchowski equation: 8RT ð15:4Þ kdiff ¼ 3000h where R is the gas constant (J K1 mol1), T is the temperature (K) and h is the viscosity of the medium (P). In our case the kdiff value for the ionic liquid DAF is approximately 0.9 108 m1 s1 and the kq values are 10 times higher than the kdiff value. This difference is due to the existence of voids in RTIL, and the solute
334 Hydrogen Bonding and Transfer in the Excited State 1.3
2.5
1.2
τ0/τ
2.0 I0/I
(b)
(a)
1.5
1.1 1.0
1.0
0.9 0
20
40
60
80
0
20
[Q] (mM)
40
60
80
[Q] (mM)
Figure 15.2 (a) Stern–Volmer plots for C-152A (*), C-151 (~) and C-522 (&) systems in DAF with increasing quencher concentration using steady-state intensity. (b) C-152 (~), C-480 (*) and C-522 (&) using time-resolved data. Reprinted with permission from the American Chemical Society. Copyright 2009 Table 15.1 Lifetime, quenching constants, redox potentials, E00 values and DGo for different coumarin–amine systems studied in neat DAF System
t0 (ns)
SS kq (109M1 s1)
TR kq (109 M1 s1)
E(A/A) V/vs SCE
E00 (eV)
E(D/Dþ ) V/vs SCE
DGo (eV)
3.59 2.94 2.72 0.84 0.76 1.72
1.621 3.946 3.765 13.607 11.987 11.645
0.549 1.303 1.162 2.952 3.026 3.087
2.10 1.685 1.653 1.660 1.626 1.565
2.920 2.586 2.657 2.734 2.756 2.846
0.756
0.254 0.315 0.428 0.498 0.554 0.705
C-480 þ DMA C-153 þ DMA C-522 þ DMA C-152A þ DMA C-152 þ DMA C-151 þ DMA 0.4
(b)
(a) 3.00E+009
2.00E+009
0.2
kq
r(t)
0.3
0.1
0.0 0.0
1.00E+009
2.5
5.0
Time (ns)
7.5
10.0
-0.7
-0.6
-0.5
-0.4
-0.3
-0.2
0
ΔG (eV)
Figure 15.3 (a) Decays of fluorescence anisotropy r(t) of C-522 (*) and C-153 (!) in DAF. (b) The plot of kq versus DGo for several coumarin–DMA systems in DAF. Reprinted with permission from the American Chemical Society. Copyright 2009
molecules diffuse rapidly from one void to another through the movement of the segment of the ions. We have measured the viscosity of neat DAF, and also, by using the rotational relaxation time of the probe molecules, the microviscosity h of the surrounding environment in the ionic liquid DAF was estimated by using the Stokes–Einstein relation [18]
Chemical Dynamics in Room-Temperature Ionic Liquids: the Role of Hydrogen Bonding
hV trot ¼ kT
335
ð15:6Þ
where h is the viscosity, V is the volume of the fluorophore, k is the Boltzmann constant, T is the absolute temperature and trot is the rotational relaxation time. We have observed that the measured viscosity of neat DAF is different from the calculated microviscosity of the probe molecules. This also supports the notion that the microviscosity sensed by the solute molecules is different from the macroscopic viscosity. The macroscopic viscosity depends on the movement of the whole ion [19]. We have compared the electron transfer rate constants of the two probes C-153 and C-522 in CH3CN with the rate constant in DAF (Table 15.1) [20]. It is found that the rate constant decreases by an order of magnitude from CH3CN to the ionic liquid DAF. This is a real effect of the ionic liquid on the rate of the electron transfer reaction and is not due to the limiting effect of diffusion. We have plotted kq values obtained from time-resolved measurement with DGo; here, kq increases with increasing DGo and approaches the diffusion limit. So, in this PET reaction in ionic liquid DAF we have observed Rehm–Weller [16] behaviour like that of other homogeneous solvents such as acetonitrile. In the PET reaction, solvent polarity has a major effect on the free energy of activation and solvent reorganization energy. The dynamic solvent effect, i.e. the friction between the reactants and polar medium, play a major role in the rate of PET. Therefore, the rate of PET is dependent on the solvent relaxation time. We have reported the solvent relaxation time of C-153 in neat DAF [8]. The solvent relaxation is very fast in DAF, and, using our TCSPC set-up, we are unable to capture the solvation dynamic at 298 K owing to the limited time resolution [8]. We have reported the solvent relaxation time of C-153 in neat DAF at different low temperatures (278, 283 and 288 K). From this study, the rate constant of solvation dynamics was calculated at 2 109 s1. In the present work we have observed that the rate constant of the PET reaction is of the order of 109 s1. Therefore, in the present system, the solvation dynamics and PET are competitive in nature. Although the appearance of dynamic heterogeneity in ionic liquid is well demonstrated [6, 9, 13], the ET rate (kq) versus DGo plot shows saturation to a diffusion-controlled value in the higher exergonicity region, which is observed in other neat solvents [20]. In the present work, the photoinduced electron transfer rate in the ionic liquid DAF decreases by an order of magnitude compared with the conventional solvent acetonitrile. Moreover, the electron transfer rate constants in DAF are significantly higher than the values predicted from the measured viscosity on the basis of the Smoluchowski equation. This may possibly be due to the fact that the microviscosity sensed by the solute molecules is different from the macroscopic viscosity. We have observed a saturation that shows a Rehm–Weller type of behaviour in the electron transfer rate in the correlation of the quenching constant (kq) with the free energy change (DGo) of the reaction.
15.2 Dynamics of Solvent Relaxation in Room-Temperature Ionic Liquids Containing Mixed Solvents The dynamics of the solvent and rotational relaxation of 153 (C-153) has been investigated in N,N, N–trimethyl-N-propylammonium bis(trifluoro-methanesulfonyl) imide (abbreviated as [N3111][Tf2N]) as an RTIL. Methanol and acetonitrile were used as the primary polar cosolvents to study the effect on solvation dynamics in the RTIL. [N3111][Tf2N] is an aprotic ionic liquid. Furthermore, for [N3111][Tf2N] the optical density at 410 nm and the fluorescence emission are very low compared with other aromatic RTILs. For these reasons, we have chosen [N3111][Tf2N] for this solvation and rotational dynamics study. Moreover, we optimized the geometry of [N3111][Tf2N]–methanol by density functional theory (DFT) methods [21]. C-153 in neat [N3111][Tf2N] shows an absorption peak at 425 nm. With the gradual addition of methanol, the emission
336 Hydrogen Bonding and Transfer in the Excited State
peak is red-shifted and finally reaches a value of 530 nm after the addition of 0.4 mol fraction of methanol in neat [N3111][Tf2N]. Similarly, the addition of 0.4 mol fraction of acetonitrile in neat [N3111][Tf2N] leads to a red-shift of the emission maximum to 525 nm. There were several studies on solvation dynamics in neat RTILs [22–24] after the first report of Karmakar and Samanta [25]. The emission obtained from neat [N3111][Tf2N] is only 1500 V) before breakdown is reached, and hence a greater operating gain. Doped APDs are therefore more sensitive, but also more fragile compared with other semiconductor photodiodes. Single-photon avalanche diodes (SPADs) are a particular class of APD that are not only able to detect extremely low intensity signals (down to the single photon) but also to signal the time of the photon arrival with high temporal resolution (a few tens of picoseconds) [81–86]. The SPADs, like any other APD, exploit the photon-triggered avalanche current of a reverse biased p–n junction to detect an incident radiation. The fundamental difference between SPADs and APDs is that SPADs are specifically designed to operate with a reverse bias voltage well above the breakdown voltage (whereas APDs operate at a bias voltage slightly below the breakdown voltage). This kind of operation is also called the Geiger mode in the literature, by analogy with the Geiger counter. At this bias, the electric field is so high (>3 105 V cm1) that a single charge carrier injected in the depletion layer can trigger a self-sustaining avalanche. The current rises swiftly (subnanosecond rise time) to a macroscopic steady level in the milliampere range. If the primary carrier is photogenerated, the leading edge of the avalanche pulse marks (with a picosecond time jitter) the arrival time of the detected photon. The current continues to flow until the avalanche is quenched by lowering the bias voltage down to, or below, the breakdown value: the lower electric field is no longer able to accelerate the carriers to impact-ionize with lattice atoms, and therefore the current ceases. In order to be able to detect another photon, the bias voltage must be raised again above breakdown. These operations require a suitable circuit, which has to sense the leading edge of the avalanche current, generate a standard output pulse synchronous with the avalanche build-up, quench the avalanche by lowering the bias down to the breakdown voltage and restore the photodiode to the operative level. This circuit is usually referred to as a quenching circuit. The simplest quenching circuit is commonly called the passive quenching circuit and is composed of a single resistor in series with the SPAD. This experimental set-up has been employed since the early studies on the avalanche breakdown in junctions. The avalanche current self-quenches simply because it develops a voltage drop across a high-value resistance (>100 kW). After quenching of the avalanche current, the SPAD bias slowly recovers to the operational value, and therefore the detector is ready to be ignited again. A more advanced quenching scheme is called active quenching. In this case a fast discriminator senses the steep onset of the avalanche current across a 50 W resistor and provides a digital output pulse synchronous with the photon arrival time. Any detector must be shielded, and the light to be analysed must be suitably filtered in order to eliminate non-time-correlated stray light from the source or the environment, and to depress the probability of more than one photon per excitation pulse impinging on the detector. In fact, as anticipated, none of the existing single-photon detectors can reset instantaneously after detecting one photon. The characteristic time during which a detector remains blind after revealing a photon
362 Hydrogen Bonding and Transfer in the Excited State
is called ‘dead time’. The function of the detector in the TCSPC system is to convert each photon impinging on its sensitive area into a photoelectron, and subsequently to amplify the microscopic photoelectric current in order to obtain a macroscopic current pulse. To accomplish this task, the detector must feature the highest possible ‘detection quantum efficiency’, which is the fraction of impinging photons converted into measurable current pulses. Moreover, the current pulse generated by a detected photon should be much higher than the parasitic currents intrinsic to the detector circuitry, in order to avoid ‘fictitious photon’ detection. In other words, the detector ‘operational gain’ should be as high as possible. Finally, in any singlephoton detector there are thermal and/or electronic processes leading to dark count events. Dark counts are processed as real photodetection events. This causes the detector to be blind for a whole dead time period following each false photon detection event. Moreover, detection of dark counts diminishes the signal-tonoise ratio, especially when very weak light pulses are to be analysed. For these reasons it is desirable for the detector of an efficient TCSPC system to show a low dark count rate. The properties that a detector must have in order to assure a high time resolution in TCSPC measurements will be discussed after explaining the operating principles of the time-to-amplitude converter (TAC). Before triggering the TAC, both the START and the STOP signals are evaluated by a discriminator. The discriminator is a key element in any TCSPC system, as it is the unit that makes TCSPC essentially insensitive to the noise problems that plague analogical methods of light analysis. A discriminator is an electronic circuit designed to differentiate between current intensity levels. If the input signal to the discriminator is below a specified threshold level, the signal is completely ignored; conversely, an input signal greater than the threshold level is recognized, causing the discriminator to produce a very neat output pulse which in turn is delivered to the TAC. By setting the discriminator threshold level to a current value substantially greater than the detector mean parasitic current level but smaller than the detector photocurrent pulse amplitude, the detector noise can be removed from the data. The central element in a TCSPC system is the TAC. A TAC can be viewed as a very precise stopwatch, with the light source providing the START signal and the emitted light providing the STOP. When it receives the START signal, the TAC begins to charge the plates of a capacitor by means of a precisely controlled constant current. When it receives the STOP signal, the TAC suddenly stops the current flow through the capacitor plates and generates an analogical voltage pulse whose amplitude is exactly equal to the potential difference DV between the capacitor plates. The latter is proportional to the time lapse between the START and the STOP signal. In other words (hence the instrument name), the TAC associates a voltage pulse of welldefined amplitude to any START/STOP time interval. As the TAC current starts flowing with a fixed delay from production of the START pulse by the START discriminator and stops flowing with a fixed delay from production of the STOP pulse by the STOP discriminator, the precision with which a photon is timed does not depend on the actual pulse width of the detector photocurrent pulse, as would happen in any analogical technique, but rather on the time jitter between detection of a photon and emission of the corresponding photocurrent pulse. This jitter must be the smallest possible to assure good timing performances in TCSPC experiments. The timing accuracy can be up to 10 times better than the half-width of the detector pulse response. It should be noted here that TCSPC provides a differential time measurement that is virtually unaffected by drifts and instabilities in the light source pulse period. The analog-to-digital converter (ADC) measures the amplitude of the voltage pulse coming from the TAC to determine into which slot of the histogram approximating the detected photon temporal distribution a particular detected photon should be recorded. It sends that ‘time slot information’ to the multichannel analyser (MCA) in the form of a digital channel number. Upon receiving the channel number, which is really just a memory address, the MCA adds 1 to the contents of that memory cell to record the fact that a photon was just detected with that specific START/STOP time lag. This process, which typically overall takes only a few microseconds, or even less, is then repeated over and over again until the events being recorded yield a reliable approximation of the actual temporal distribution to be studied.
Intramolecular H-Bond Formation Mediated De-Excitation of Curcuminoids 363
17.2 Experimental Set-Up and Data Analysis Methods 17.2.1 Curcuminoids and solvents Curcumin (CURC) and dicinnamoylmethane (DCMeth) were synthesized as previously described [65, 87]. Their chemical structures are reported in Figures 17.3 and 17.6 respectively. Studies on the crystal structures [10, 88, 89] of the two compounds show that they display different chemical affinities with respect to the formation of KEHB in the solid state. In CURC, the enol proton is localized with equal probability around either of the two carbonyl oxygens [88], indicating the formation of a fairly weak KEHB, while in the DCMeth crystal structure the enol proton is equidistant from the two carbonyl oxygens [89], indicating complete charge delocalization and the formation of a semi-aromatic ring at the diketo structure, which is necessarily mediated by an extremely tight KEBH. The solvents used in the present study were divided into: non-polar (cyclohexane), polar non-H-bonding (chloroform, ethyl acetate, acetone and acetonitrile), H-bond-accepting (dimethylformamide (DMFA) and dimethylsulfoxide (DMSO)) and alcohols (isopropanol, ethanol and methanol). Note that the alcohols display both H-bond-donating and H-bond-accepting properties. To assess the polarity, we use the dielectric constant. The dielectric constants of the above solvents, together with the hydrogen-bonding donor parameter of the H-bond donors (acidity parameter a) and the hydrogen-bonding acceptor parameter (basicity parameter b) of the H-bond acceptors, are reported in Table 17.1. All the solvents were of 99.5% purity and used as received, except ethyl acetate, which was dried over sodium sulfate. Samples in the solvents were prepared the same day they were used for measurements.
O
O
Figure 17.6 Chemical structure of the dicinnamoylmethane molecule
Table 17.1 Solvent properties: hydrogen-bonding donor parameter a; hydrogen-bonding acceptor parameter b; dielectric constant at 25 C « Solvent Non-polar Polar non-H-bonding
H-bond acceptors Alcohols
Cyclohexane Chloroform Ethyl acetate Acetone Acetonitrile DMFA DMSO Isopropanol Ethanol Methanol
«
a
b
2.02 4.81 6.02 20.60 38.8 37.6 48.9 19.92 25.07 33.62
0 0.44 0 0.08 0.19 0 0 0.78 0.83 0.93
0 0 0.45 0.48 0.31 0.69 0.76 0.95 0.77 0.62
364 Hydrogen Bonding and Transfer in the Excited State
17.2.2 Our TCSPC set-up Our TCSPC experimental set-up is sketched in Figure 17.7. For the measurements presented here, as the light source we used a continuous-wave SESAM-mode-locked Ti:sapphire laser (fundamental emission wavelength 840 nm) emitting pulses at 48 MHz repetition rate with a 3.9 ps pulse width (Tiger-ps SHG; Time Bandwidth Products, Zurich, Switzerland). The samples were excited at 420 nm by the Ti:sapphire built-in second harmonic. The light exiting the excitation source is partially reflected by means of a microscopy cover glass. The principal part of the beam is transmitted through the glass and conveyed to the sample, while the small fraction of reflected light is delivered to the detector via a multimode optical fibre (Fsync), without interacting with the sample, to provide a time reference [90–93] stable against electronic drifts. In fact, the start pulses due to photons reaching the SPAD through the Fsync fibre give rise to two sharp peaks which are separated in time by exactly a pulse period. By selecting a temporal acquisition window slightly longer than the pulse period, they can be positioned at the first and last channels of the MCA, as depicted in Figure 17.7 (bottom right), by adjusting the delay (see below). The time distance between these peaks is extremely stable over time (it was measured to be constant within one channel over several hours of monitoring), while the absolute peak positions drift by as much as 100 channels. Superimposing the reference peaks of two different
Figure 17.7 Our TCSPC set-up. BS: beam splitter; ND: neutral density filters; Fsync: multimode fibre conveying the reference peaks to the detector; L: collimating lens; PIN: fast photodiode for synchronization with the excitation pulse; CFD: constant-fraction discriminator; TAC: time-to-amplitude converter; MCA: multichannel analyser; OBJ: 40 microscope objective; SPAD: single-photon avalanche diode; AQC: active quenching circuit
Intramolecular H-Bond Formation Mediated De-Excitation of Curcuminoids 365
fluorescence decay patterns provides a common absolute timescale for the distributions to be analysed, as the difference in time between detection of photons conveyed to the reference branch of the TCSPC apparatus (sent to Fsync) and impingement onto the sample cell of photon pulses conveyed to the sample branch is constant. Collection of stray light by Fsync was prevented by inserting a black plastic tube (3 cm diameter) in front of the Fsync input end. A neutral density filter (OD ¼ 4) was attached to the tube entrance. Our TCSPC system features a passive electronic delay line for rough positioning of both the reference peaks and the temporal distribution of the photon emerging from the sample inside the temporal window. However, in order to be able to shift the reference peaks with respect to the measured distribution so as to obtain time distributions perfectly centred into the time window and reference peaks at its extremes, an optical delay has been put on the reference branch. The emitted light is collected at 90 with respect to the excitation beam and selected from excitation stray light by means of suitable cut-off (long-wavelength pass) filters. It is focused by means of a 40 microscope objective on the sensitive area of the detector. This is a SPAD (PDM50, Micro Photon Devices, Bolzano, Italy) featuring the following technical properties: 50 mm diameter of the sensitive area, 3,5CNAI > 3CNAI 7AI. The trend of the observed rate of the ESPT process is in good correlation with the proposed solvent-polarity-induced barrier resulting from the difference in the
572
Hydrogen Bonding and Transfer in the Excited State
3,5CNAI
ΔG+ 5CNAI
ΔG+ Energy
DiCNAI
ΔG+
R1
R1
R2
R2 N
N
N*
N
N
H
3CNAI R1=CN, R2=H 3,5CNAI R1=CN, R2=CN 5CNAI R1=H, R2=CN DiCNAI R 1=H, R2=
H
T*
eq
CN CN
eq
Solvent Coordinate
Figure 24.14 The proposed mechanism of ESPT incorporating a solvent-polarity-induced barrier in protic solvents following optical charge transfer and solvent relaxation. Reprinted with permission from [131]. Copyright 2008 Wiley-VCH Verlag GmbH & Co. KGaA
N
N
*
C
C
hυ N H
N H
O R
adiabatic ESCT
N
*
C
ESPT N H
N H
H
O R
N
N
H O R
normal emission
tautomer emission
Figure 24.15 The proposed ESPT/ESCT coupled mechanism of 5CNAI in protic solvents
changes in dipole moments between the equilibrium polarization of the normal (Neq ) and tautomer species (Teq ) in the solvent coordinate (Figure 24.13) [131]. Moreover, as shown in Figure 24.14, it is obvious that the barrier is increased upon increase in the difference in dipole moment (either magnitude or direction) between the normal and tautomer forms, consistent with the theoretical prediction regarding horizontal displacement of dipole separation described in Section 24.4.1.
Excited-State Proton Transfer via Hydrogen-Bonded Dimers and Complexes in Condensed Phase
573
Table 24.1 Photophysical properties of 7AI and its correlated cyano analogues in methanol. Reprinted with permission from [131]. Copyright 2008 Wiley-VCH Verlag GmbH & Co. KGaA labs (nm)
lem (nm) (F)a
7AIc
288
N: 374 T: 503 (0.07)
3CNAI
285
N: 343 T: 480 (0.02)
3,5CNAI
294
N: 377 T: 515 (0.03)
5CNAI DiCNAI
297 351
N: 395 (0.20) N: 600 (0.0014)
t (ns)b t: 0.146 t1: 0.134 (0.44) t2: 0.654 (0.56) t: 0.23 t1: 0.24 (0.49) t2: 5.88 (0.51) t: 0.69 t1: 0.71 (0.52) t2: 1.13 (0.48) t: 4.8 t: 0.20
a
The reported F is the sum of the normal (N) and tautomer (T) emission bands. Data in parentheses are the fitted pre-exponential factors. The photophysical properties of 7AI are taken from Ref. [17].
b c
The above 7AI analogues serve as one of a few experimental proofs for the solvent-induced barrier in proton transfer reaction. It is thus believed that, through other ingenious design, the systematic investigation of the ESCT/ESPT coupled reaction becomes possible in protic solvents, which may be crucial to gaining fundamental insights into the current research fields regarding, for example, proton-coupled electron transfer in a living system. 24.4.3 Excited-State proton transfer coupled with charge transfer Both theoretical and experimental progress elaborated in Sections 24.4.1 and 24.4.2 has fundamental importance in that it clearly addresses the role of solvent polarity channelling into the proton transfer dynamics. Unfortunately, up to this stage, for most experimental model systems applied, the ESCT/ESPT dynamics is ascribed to case (A) in which ESCT takes place prior to ESPT. Thus, upon Franck–Condon excitation, the associated reaction dynamics has been complicated by competitive solvent relaxation to reach equilibrium polarization. As such, the study of early ESCT/ESPTreaction dynamics is commonly limited to the rate of solvent relaxation. To overcome this hurdle, it is of great fundamental interest to seek an ideal system to probe ESCT/ESPT coupling reactions free from early solvent relaxation process. In theory, an ideal case in point stems from case (B), for which a molecule is designed such that it undergoes ESPT prior to the ESCTreaction. Recently, via the strategic design and synthesis of molecule 2-((2-(2-hydroxyphenyl) benzo[d]oxazol-6-yl)methylene)malononitrile (diCN–HBO) [132], the study of ESPT-coupled ESCT, i.e. process (B), becomes feasible. From the molecular structure point of view, for diCNHBO, the lone-pair electrons of the benzo nitrogen atom are intrinsically involved in p electron resonance to establish aromaticity, such that its electron-donating strength, compared with that of alkyl and aryl amines, is negligibly weak. Thus, upon Franck–Condon excitation of diCNHBO, the degree of charge transfer should be negligible. On the other hand, similarly to its parent molecule 2-(20 -hydroxyphenyl)benzoxazole (HBO) [133–135], ESPT is expected to take place from the hydroxyl proton to the N1 nitrogen, resulting in a proton transfer tautomer, i.e. a keto form. Once the proton transfer tautomer is formed, the N1 nitrogen atom becomes the secondary alkyl amino nitrogen and thus should act as a good electron donor (Figure 24.16). Unlike most of the ESCT/ESPT systems, in which ESCT takes place prior to ESPT, diCNHBO undergoes ESPT, concomitantly accompanied with the charge transfer process, such that the ESPT reaction dynamics is directly coupled with solvent polarization effects. The long-range solvent polarization interactions result in
574
Hydrogen Bonding and Transfer in the Excited State H
H
O
O
N
N
hv ESPT
O
O
HBO
H 1
H
O
N
O
N
hv ESPT
O
O
NC
CN
diCN-HBO
NC
CN
adiabatic ESCT
proton motion coupled with charge transfer
H
O
+δ N
O
NC
-δ
CN
Figure 24.16 Proposed ESPT reaction for HBO and ESPT/ESCT coupled reaction for diCNHBO. Reprinted with permission from [136]. Copyright 2008 American Chemical Society
a solvent-induced barrier that affects the overall proton transfer reaction rate. According to the spectroscopic measurements, dual emission was observed, and the proton transfer tautomer emission peak moves drastically in solvents bearing different polarities. In cyclohexane, the rate constant of ESPT of diCNHBO was determined to be 1.1 ps, which is apparently slower than the 150 fs for the parent molecule HBO. Upon increase in solvent polarity, the ESPT rate constants were also determined to be 1.00 0.13 ps in benzene, 0.60 0.05 ps in CH2Cl2 and 0.31 0.03 ps in CH3CN, values revealing that an increase in solvent polarity tends to produce an increase in the rate of ESPT [136]. The overall reaction dynamics can be described by a mechanism incorporating both solvent polarization and proton- transfer reaction coordinates (Figure 24.17). In the solvent coordinate, the proton transfer tautomer possessing a high degree of charge transfer character is obviously stabilized upon increase in solvent polarity, while the PC state is not influenced, and thus the corresponding solvent-induced barrier is reduced. This diCNHBO system serves as the first ESPT/ESCT example in which ESPT occurs prior to the ESCT process. And, again, the existence and importance of the solvent-induced barrier are revealed.
24.5 Conclusions In this chapter, we have succinctly reviewed three pivotal topics relevant to the excited-state proton transfer reaction. In the first section, on biprotonic transfer within doubly H-bonded homo- and heterodimers, we have
Excited-State Proton Transfer via Hydrogen-Bonded Dimers and Complexes in Condensed Phase
H
575
O
N H O
O
CN
NC
Free Energy
ESPT/CT
PC*
O
N
NC
CN
solvent relaxation
C6H12 CH2Cl2 CH3CN
increasing solvent polarity CPT*
PC Solvent polarization coordinate P
Figure 24.17 Proposed ESPT/ESCT reaction/relaxation dynamics using diCNHBO as a model. Reprinted with permission from [136]. Copyright 2008 American Chemical Society
described the concerted versus stepwise type of excited-state double-proton transfer for hydrogen-bonded dimers. For the case of the 7-AI dimer, although a definitive mechanism of double proton transfer is still debatable at the time of writing, both pro and con research groups present their viewpoints, all based on fundamental and state-of-the-art techniques, from which the readers should gain profound knowledge. Thus, this issue, although disputable, should attract a broad spectrum of readership. Section 24.3 fully extends the dimer or heterodimer hydrogen bonding concept (in Section 24.2) to any host/guest types of hydrogen-bonded complex. Their corresponding proton transfer reaction may thus be either dynamically or thermodynamically controlled, depending on the tautomerism for host, guest or both. Thus, the proton transfer reaction can be fine-tuned on the basis of chemical structures designed for both host and guest molecules. Last but not least, once in the polar solvents, owing to possible changes in the dipole moment during the proton transfer reaction, solvent polarity is expected to play a crucial role in channelling into the reaction dynamics. In this section, theory on solvent-polarity-induced reaction dynamics is first elaborated, followed by its verification based on two reaction types, namely the charge-transfer-induced proton transfer reaction (case A) and the protontransfer-induced charge transfer reaction (case B). Both prove to be prototypical models for proof of the concept. We thus hope that, through reading this chapter, the reader can gain a more fundamental knowledge of the excited-state proton transfer reaction and perhaps its potential in future applications.
References 1. For example, see A. M€uller, H. Ratajack, W. Junge and E. Diemann, Studies in Physical and Theoretical Chemistry, Vol. 78, Electron and Proton Transfer in Chemistry and Biology. Elsevier, Amsterdam (1992). 2. J. Waluk, Conformational Analysis of Molecules in Excited States. Wiley-VCH (2000). 3. T. H. Elsaesser and H. J. Bakker, Ultrafast Hydrogen Bonding Dynamics and Proton Transfer Processes in the Condensed Phase. Springer, Netherlands (2002).
576 Hydrogen Bonding and Transfer in the Excited State 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. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52.
M. Kasha, J. Chem. Soc., Faraday Trans. 2, 82, 2379 (1986). P. T. Chou, J. Chin. Chem. Soc., 48, 651 (2001). L. M. Tolbert and K. M. Solntsev, Acc. Chem. Res., 35, 19 (2002). J. Waluk, Acc. Chem. Res., 36, 832 (2003). T. E. Dermota, Q. Zhong and A. W. Castleman, Jr, Chem. Rev., 104, 1861 (2004). In general it is difficult to distinguish the mechanism between proton transfer and hydrogen atom transfer reactions. In this review article we have simply adopted the conventional term ‘proton transfer’ throughout the text, which is particularly suitable when describing the sequential double-proton transfer reaction in the later sections. M. Rini, B. Z. Magnes, E. Pines and E. T. J. Nibbering, Science, 301, 349 (2003). S. J. Bankovic and S. Hammes-Schiffer, Science, 301, 1196 (2003). N. Agmon, J. Phys. Chem. A, 109, 13 (2005). S. Y. Park, Y. S. Lee and D. J. Jang, Phys. Chem. Chem. Phys., 10, 6703 (2008). T. Hirano, Y. Hasumi, K. Ohtsuka et al., J. Am. Chem. Soc., 131, 2385 (2009). J. D. Watson and F. H. C. Crick, Nature, 171, 737 (1953). C. A. Taylor, M. A. El-Bayoumi and M. Kasha, Proc. Natl Acad. Sci. USA, 63, 253 (1969). M. Negrerie, F. Gai, S. M. Bellefeuille and J. W. Petrich, J. Phys. Chem., 95, 8663 (1991). R. S. Moog and M. Maroncelli, J. Phys. Chem., 95, 10 359 (1991). Y. Chen, R. L. Rich, F. Gai and J. W. Petrich, J. Phys. Chem., 97, 1770 (1993). Y. Chen, F. Gai and J. W. Petrich, J. Am. Chem. Soc., 115, 10 158 (1993). R. L. Schowen, Angew. Chem. Int. Ed. Engl., 36, 1434 (1997). S. Mente and M. Maroncelli, J. Phys. Chem. A, 102, 3860 (1998). D. E. Folmer, E. S. Wisniewski, J. R. Stairs and A. W. Castleman, Jr, J. Phys. Chem. A, 104, 10 545 (2000). O. H. Kwon, Y. S. Lee, H. J. Park et al., Angew. Chem. Int. Ed., 43, 5792 (2004). W. M. Hetherington, III, R. H. Micheels and K. B. Eisenthal, Chem. Phys. Lett., 66, 230 (1979). P. Share, M. Pereira, M. Sarisky et al., J. Lumin., 48/49, 204 (1991). A. Douhal, S. K. Kim and A. H. Zewail, Nature, 378, 260 (1995). D. E. Folmer, L. Poth, E. S. Wisniewski and A. W. Castleman, Jr, Chem. Phys. Lett., 287, 1 (1998). M. Chachisvilis, T. Fiebig, A. Douhal and A. H. Zewail, J. Phys. Chem. A, 102, 669 (1998). T. Fiebig, M. Chachisvilis, M. Manger et al., J. Phys. Chem. A, 103, 7419 (1999). D. E. Folmer, E. S. Wisniewski and A. W. Castleman, Jr, Chem. Phys. Lett., 318, 637 (2000). L. Serrano-Andres and M. Merchan, Chem. Phys. Lett., 418, 569 (2006). O. H. Kwon and A. H. Zewail, Proc. Natl Acad. Sci. USA, 104, 8703 (2007). O. H. Kwon and A. H. Zewail, Proc. Natl Acad. Sci. USA, 105, E79 (2008). S. Takeuchi and T. Tahara, J. Phys. Chem. A, 102, 7740 (1998). S. Takeuchi and T. Tahara, Chem. Phys. Lett., 347, 108 (2001). J. Catalan, P. Perez, J. C. del Valle et al., Proc. Natl Acad. Sci. USA, 99, 5799 (2002). J. Catalan, P. Perez, J. C. del Valle et al., Proc. Natl Acad. Sci. USA, 101, 419 (2004). K. Sakota, A. Hara and H. Sekiya, Phys. Chem. Chem. Phys., 6, 32 (2004). H. Y. Chen and I. Chao, ChemPhysChem, 5, 1855 (2004). K. Sakota and H. Sekiya, J. Phys. Chem. A, 109, 2718 (2005). K. Sakota, C. Okabe, N. Nishi and H. Sekiya, J. Phys. Chem. A, 109, 5245 (2005). J. Catalan and J. L. G. de Paz, J. Chem. Phys., 123, 114 302 (2005). S. Takeuchi and T. Tahara, Proc. Natl Acad. Sci. USA, 104, 5285 (2007). J. Catalan, Proc. Natl Acad. Sci. USA, 105, E78 (2008). K. C. Ingham and M. A. El-Bayoumi, J. Am. Chem. Soc., 96, 1674 (1974). H. Bulska and A. Chodkowska, J. Am. Chem. Soc., 102, 3259 (1980). J. Catalan and M. Kasha, J. Phys. Chem. A, 104, 10 812 (2000). J. Waluk, J. Herbich, D. Oelkrug and S. Uhl, J. Phys. Chem., 90, 3866 (1986). T. Sekikawa, T. Kobayashi and T. Inabe, J. Phys. Chem. A, 101, 644 (1997). P. Dufour, Y. Dartiguenave, M. Dartiguenave et al., Can. J. Chem., 68, 193 (1990). W. S. Yu, C. C. Cheng, C. P. Chang et al., J. Phys. Chem. A, 106, 8006 (2002).
Excited-State Proton Transfer via Hydrogen-Bonded Dimers and Complexes in Condensed Phase 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. 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.
577
C. Shi, Q. Zhang and K. K. Wang, J. Org. Chem., 64, 925 (1999). W. T. Hsieh, C. C. Hsieh, C. H. Lai et al., ChemPhysChem, 9, 293 (2008). C. C. Hsieh, C. C. Wang, G. H. Lee and P. T. Chou, unpublished results. Y. Komoto, K. Sakota and H. Sekiya, Chem. Phys. Lett., 406, 15 (2005). J. Catalan, C. Dıaz and J. L. G. Paz, Chem. Phys. Lett., 419, 164 (2006). J. Catalan, C. Dıaz, P. Perez and J. L. G. Paz, J. Phys. Chem. A, 110, 9116 (2006). C. Chang, N. Shabestary and M. A. El-Bayoumi, Chem. Phys. Lett., 75, 107 (1980). J. Waluk, J. Herbich, D. Oelkrug and S. Uhl, J. Phys. Chem., 90, 3866 (1986). K. Fuke, K. Tsukamoto, F. Misaizu and K. Kaya, J. Chem. Phys., 95, 4074 (1991). P. T. Chou, W. S. Yu, C. Y. Wei et al., J. Am. Chem. Soc., 123, 3599 (2001). E. Nir, K. Kleinermanns and M. S. de Vries, Nature, 408, 949 (2000). C. E. Crespo-Hernandez, B. Cohen, P. M. Hare and B. Kohler, Chem. Rev., 104, 1977 (2004). T. Schultz, E. Samoylova, W. Radloff et al., Science, 306, 1765 (2004). C. E. Crespo-Hernandez, B. Cohen and B. Kohler, Nature, 436, 1141 (2005). S. Perun, A. L. Sobolewski and W. Domcke, J. Phys. Chem. A, 110, 9031 (2006). P. J. Thistlethwaite and P. J. Corkill, Chem. Phys. Lett., 85, 317 (1982). M. Itoh, T. Adachi and K. Tokumura, J. Am. Chem. Soc., 106, 850 (1984). J. Konijnenberg, G. B. Ekelmans, A. H. Huizer and C. A. G. O. Varma, J. Chem. Soc., Faraday Trans. 2, 85, 39 (1989). T. Nakagawa, S. Kohtani and M. Itoh, J. Am. Chem. Soc., 117, 7952 (1995). K. Tokumura, M. Natsume, T. Nakagawa et al., Chem. Phys. Lett., 271, 320 (1997). P. T. Chou, C. Y. Wei, C. R. C. Wang et al., J. Phys. Chem. A, 103, 1939 (1999). S. Kohtani, A. Tagami and R. Nakagaki, Chem. Phys. Lett., 316, 88 (2000). C. Tanner, C. Manca and S. Leutwyler, Science, 302, 1736 (2003). O. H. Kwon, Y. S. Lee, B. K. Yoo and D. J. Jang, Angew. Chem. Int. Ed. Engl., 45, 415 (2006). M. Balo´n, J. Hidalgo, P. Guardado et al., J. Chem. Soc. Perkin Trans. 2, 99 (1993). C. Carmona, M. Galan, G. Angulo et al., Phys. Chem. Chem. Phys., 2, 5076 (2000). P. T. Chou, Y. I. Liu, G. R. Wu et al., J. Phys. Chem. B, 105, 10 674 (2001). C. Carmona, M. Balo´n, A. Sanchez-Coronilla and M. A. Munˇoz, J. Phys. Chem. A, 108, 1910 (2004). D. Reyman, F. Hallwass, S. M. da Cruz Goncalves and J. J. Camacho, Magn. Reson. Chem., 45, 830 (2007). J. C. del Valle, E. Domınguez and M. Kasha, J. Phys. Chem. A, 103, 2467 (1999). A. Kyrychenko, J. Herbich, F. Wu et al., J. Am. Chem. Soc., 122, 2818 (2000). J. Herbich, M. Kijak, A. Zielin´ska et al., J. Phys. Chem. A, 106, 2158 (2002). M. S. Mudadu, A. Singh and R. P. Thummel, J. Org. Chem., 71, 7611 (2006). Y. Nosenko, G. Wiosna-Sałyga, M. Kunitski et al., Angew. Chem. Int. Ed., 47, 6037 (2008). P. T. Chou and C. Y. Wei, J. Phys. Chem. B, 101, 9119 (1997). C. P. Chang, W. C. Hwang, M. S. Kuo et al., J. Phys. Chem., 98, 8801 (1994). P. T. Chou, C. Y. Wei, C. P. Chang and C. H. Chiu, J. Am. Chem. Soc., 117, 7259 (1995). P. T. Chou, W. S. Yu, Y. C. Chen et al., J. Am. Chem. Soc., 120, 12 927 (1998). P. T. Chou, G. R. Wu, C. Y. Wei et al., J. Phys. Chem. B, 104, 7818 (2000). A. P. Davis and R. S. Wareham, Angew. Chem. Int. Ed., 38, 2978 (1999). J. Chin, S. Chung and D. H. Kim, J. Am. Chem. Soc., 124, 10 948 (2002). K. Yoshimoto, S. Nishizawa, M. Minagawa and N. Teramae, J. Am. Chem. Soc., 125, 8982 (2003). M. Mazik, M. Kuschel and W. Sicking, Org. Lett., 8, 855 (2006). C. R. van den Brom, I. Arfaoui, T. Cren et al., Adv. Funct. Mater., 17, 2045 (2007). W. Yang, H. He and D. G. Drueckhammer, Angew. Chem. Int. Ed., 40, 1714 (2001). A. Ajayaghosh, E. Arunkumar and J. Daub, Angew. Chem. Int. Ed., 41, 1766 (2002). V. Hegde, P. Madhukar, J. D. Madura and R. P. Thummel, J. Am. Chem. Soc., 112, 4549 (1990). V. Hedge, C. Y. Hung, P. Madhukar et al., J. Am. Chem. Soc., 115, 872 (1993). H. C. Chou, C. H. Hsu, Y. M. Cheng et al., J. Am. Chem. Soc., 126, 1650 (2004). D. Borgis and J. T. Hynes, J. Phys. Chem., 100, 1118 (1996). D. Borgis and J. T. Hynes, J. Chem. Phys., 94, 3619 (1991).
578 Hydrogen Bonding and Transfer in the Excited State 104. 105. 106. 107. 108. 109. 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.
S. Lee and J. T. Hynes, J. Chem. Phys., 93, 1783 (1996). P. M. Kiefer and J. T. Hynes, J. Phys. Chem. A, 106, 1834 (2002). P. M. Kiefer and J. T. Hynes, J. Phys. Chem. A, 106, 1850 (2002). J. T. Hynes, T. H. Tran-Thi and G. Grunucci, J. Photochem. Photobiol. A: Chem., 154, 3 (2002). E. D. German and A. M. Kuznetsov, J. Chem. Soc., Faraday Trans. 1, 77, 397 (1981). E. D. German, A. M. Kuznetsov and R. R. Dogonadze, J. Chem. Soc., Faraday Trans. 2, 76, 1128 (1980). M. Morillo and R. I. Cukier, J. Chem. Phys., 92, 4833 (1990). S. Lee and J. T. Hynes, J. Chem. Phys., 93, 1783 (1996). Y. M. Cheng, S. C. Pu, C. J. Hsu et al., ChemPhysChem, 7, 1372 (2006). Y. M. Cheng, S. C. Pu, Y. C. Yu et al., J. Phys. Chem. A, 109, 11 696 (2005). P. T. Chou, W. S. Yu, Y. M. Cheng et al., J. Phys. Chem. A, 108, 6487 (2004). D. Gormin and M. Kasha, Chem. Phys. Lett., 153, 574 (1988). J. Heldt, D. Gormin and M. Kasha, Chem. Phys., 136, 321 (1989). V. V. Shynkar, Y. Mely, G. Duportail et al., J. Phys. Chem. A, 107, 9522 (2003). F. Parsapour and D. F. Kelley, J. Phys. Chem., 100, 2791 (1996). P. T. Chou, M. L. Martinez and J. H. Clements, J. Phys. Chem., 97, 2618 (1993). A. D. Roshal, J. A. Organero and A. Douhal, Chem. Phys. Lett., 379, 53 (2003). S. Ameer-Beg, S. M. Ormson, X. Poteau et al., J. Phys. Chem. A, 108, 6938 (2004). P. T. Chou, M. L. Martinez and J. H. Clements, Chem. Phys. Lett., 204, 395 (1993). T. C. Swinney and D. F. Kelley, J. Chem. Phys., 99, 211 (1993). F. Parsapour and D. F. Kelley, J. Phys. Chem., 100, 2791 (1996). S. M. Ormson, R. G. Brown, F. Vollmer and W. Rettig, J. Photochem. Photobiol., 81, 65 (1994). P. T. Chou, C. H. Huang, S. C. Pu et al., J. Phys. Chem. A, 108, 6452 (2004). A. M. Napper, N. J. Head, A. M. Oliver et al., J. Am. Chem. Soc., 124, 10 171 (2002). M. J. Shephard, M. N. Paddon-Row and K. D. Jordan, Chem. Phys., 176, 289 (1993). M. N. Paddon-Row and M. J. Shephard, J. Am. Chem. Soc., 119, 5355 (1997). C. H. Lai, E. Y. Li, K. Y. Chen et al., J. Chem. Theory Comput., 2, 1078 (2006). C. C. Hsieh, K. Y. Chen, W. T. Hsieh et al., ChemPhysChem, 9, 2221 (2008). J. Seo, S. Kim and S. Y. Park, J. Am. Chem. Soc., 126, 11 154 (2004). T. Arthen-Engeland, T. Bultmann, N. P. Ernsting et al., Chem. Phys., 163, 43, (1992). O. K. Abou-Zied, R. Jimenez, E. H. Z. Thompson et al., J. Phys. Chem. A, 106, 3665 (2002). H. Wang, H. Zhang, O. K. Abou-Zied et al., Chem. Phys. Lett., 367, 599 (2003). C. C. Hsieh, Y. M. Cheng, C. J. Hsu et al., J. Phys. Chem. A, 112, 8323 (2008).
25 QM/MM Study of Excited-State Solvation Dynamics of Biomolecules Tetsuya Taketsugu1,2, Daisuke Kina1, Akira Nakayama1, Takeshi Noro1 and Mark S. Gordon3 1
Division of Chemistry, Graduate School of Science, Hokkaido University, Sapporo 060-0810, Japan Department of Theoretical and Computational Molecular Science, Institute for Molecular Science, Okazaki 444-8585, Japan 3 Department of Chemistry, Iowa State University, Ames, Iowa 50011, USA
2
25.1 Introduction Excited-state hydrogen transfer (ESHT) is a fundamental reaction that plays an important role in a variety of biological processes [1–3]. The microscopic mechanism and dynamics of ESHT can be elucidated only through a combined knowledge of experimental and theoretical studies. The theoretical approach can be separated into a static examination of excited-state potential energy surfaces (PES) and dynamics simulations on excited-state PES. The excited-state PES can be obtained through ab initio electronic structure or density functional theory (DFT) calculations. Taking into account the applicability to reaction processes accompanying bond cleavage and formation, one should employ, at least, the multiconfigurational self-consistent field (MCSCF) method. In dynamics simulations, an ab initio molecular dynamics (AIMD) approach has proven to be a powerful tool [4]. AIMD is a classical trajectory method in which the force acting on each atom is calculated ‘on the fly’ by ab initio electronic structure methods and enables one to perform a more reliable molecular dynamics simulation. Recently, this methodology has been extended to dynamical processes in electronic excited states [5–7] by including a non-adiabatic surface-hopping scheme, and it can be employed to predict physicochemical properties in the condensed phase by combining QM calculations with some appropriate treatment of solvent effects. For chemical processes in the condensed phase, a hybrid quantum mechanics/molecular mechanics (QM/ MM) method [8] has been developed. If solute–solvent interactions are strong in solution, they could have a large influence on the electronic structure of the solute molecule, including the excitation spectrum [9]. In a
Hydrogen Bonding and Transfer in the Excited State, Volume I & II © 2011 John Wiley & Sons, Ltd. ISBN: 978-0-470-66677-7
Edited by Ke-Li Han and Guang-Jiu Zhao
580 Hydrogen Bonding and Transfer in the Excited State
QM/MM approach, each solvent molecule is represented explicitly by classical MM potentials. The solvent molecules that participate directly in the chemical process should be included in the QM part, while the environmental solvent molecules can be included in the MM part. The quality of the QM/MM calculations depends on the model potential used in the MM part, and a number of sophisticated model potentials have been developed. The effective fragment potential (EFP) method [10, 11] is a very sophisticated QM/MM approach that has been implemented in the program package GAMESS [12, 13]. The method accounts for three important solute–solvent and solvent–solvent interaction terms: (a) Coulomb interactions; (b) polarization interactions; (c) exchange repulsion þ charge transfer interactions. In the presence of an ab initio solute, these terms are added as one-electron terms to the ab initio Hamiltonian. The Coulomb term is treated with a distributed multipolar analysis (DMA) of the solvent molecule, expanded through octopoles. The polarization term is treated by a self-consistent distributed finite field model with localized molecular orbital (LMO) polarizability tensors. These two terms have been calculated for many points on the water dimer potential energy surface and subtracted from the total Hartree–Fock interaction potential. Then, this remaining (exchange repulsion þ charge transfer) interaction is fitted to a functional form. In GAMESS, the EFP option can be used with a state-specific MCSCF wave function, where the electron density used to determine EFP-induced dipoles is obtained from the orbitals and CI coefficients of the selected MCSCF state [14]. Thus, the modelling of excited-state dynamics in solution is feasible at the MCSCF-EFP level. Very recently we have developed a general AIMD code for excited-state dynamics in solution using the EFP code in GAMESS, and reported two applications of solvated excited-state dynamics, i.e. ESHT of 7-azaindole (7AI) plus a water cluster [15] and excited-state dynamics of coumarin 151 [16]. In this review, we introduce these two applications to provide insight into the excited-state dynamics in solution, including ESHT of a DNA-base model molecule.
25.2 Applications 25.2.1 Excited-state H-transfer in 7-azaindole-(H2O)n Tautomerization of 7AI, accompanying the ESHT from the five-member ring to the six-member ring, has been studied with much attention both experimentally [17–25] and theoretically [26–29], as 7AI can be regarded as a simple model for a DNA base. The 7AI molecule in its ground electronic state S0 has a delocalized (‘aromatic’) six-member ring attached to a partially saturated five-member ring that contains an N–H bond. In the higher-energy tautomer, the N–H hydrogen is transferred to the six-member ring, thereby breaking the strong delocalization in that ring. In the first excited S1 state, the relative energies of the two isomers is reversed, since the excitation is essentially a p–p excitation in the six-member ring. Sakota et al. measured the dispersed fluorescence (DF) spectra and resonance-enhanced multiphoton ionization (REMPI) spectra for the 7AI–(H2O)n (n ¼ 2, 3) cluster at low temperature in the gas phase and found that the tautomerization of 7AI–(H2O)n (n ¼ 2, 3) occurs in the S1 state [24]. It was also reported [22, 23] that excited-state triple-H transfer occurs in the 7AI–(MeOH)2 cluster, based on the measurement of the DF and REMPI spectra. They pointed out that ESHT in 7AI–(MeOH)2 proceeds via a tunnelling mechanism in their experimental conditions. An early theoretical effort on the ESHT reaction of 7AI–H2O was performed by Chaban and Gordon [26, 27], in which the intrinsic reaction coordinate was calculated for the tautomerization in an isolated 7AI molecule and the 7AI–H2O complex in the singlet ground (S0) and first-excited (S1) states at the complete active space MCSCF (CASSCF) level of theory, with improved energetics obtained using multireference second-order perturbation theory (MCQDPT2) [30]. It was shown that the normal form of 7AI is more stable than the tautomer in the S0 state, while the relative energies are reversed in the S1 state. The activation energy for tautomerization in 7AI is significantly reduced by the complexation with water, because the H transfer occurs via the 7AI–H2O hydrogen-bonded network. In addition to experimental observation in
QM/MM Study of Excited-State Solvation Dynamics of Biomolecules
581
the gas phase, tautomerization in 7AI was also observed in the condensed phase in alcohol or water solution [17–19], where the tautomerization exhibits a strong dependence on the water concentration in mixtures of water and aprotic solvents [19]. It is possible that ESHTwould exhibit different mechanisms for the reactions in the gas phase and in the condensed phase. Spectroscopic experiments in the gas phase have been carried out at very low temperatures, so quantum-mechanical tunnelling could contribute significantly to the ESHT reaction. On the other hand, the experiments in solution have been carried out at room temperature, where ESHT reactions could occur classically, because there is sufficient energy in the system to overcome the small activation barrier. In our previous study [15], geometry optimizations and normal mode analyses were carried out for 7AI and 7AI–(H2O)n (n ¼ 1, 2) in the ground and first excited singlet states by the state-specific CASSCF method with the segmented DZP basis set [31–34], using GAMESS [12, 13]. The CASSCF active space includes ten electrons in nine orbitals, which comprise all of the 7AI p orbitals and electrons. Following the examination of the energetics of the reactions, AIMD simulations were performed for the ESHT processes in 7AI–(H2O)n (n ¼ 1, 2). The AIMD code used here is based on the leapfrog algorithm [35]. The initial conditions for the dynamics simulations were the equilibrium structures in the ground state, with the atomic velocities given in the directions of two normal modes of vibration related to the H transfer where the kinetic energy corresponds to three quantum numbers of the respective modes. The time step was taken as 0.5 fs throughout the simulation. Figure 25.1 shows the calculated equilibrium geometries for the normal and tautomer forms of 7AI–H2O and the transition state (TS) geometries for tautomerization in both the S0 and S1 states, where the interatomic
Figure 25.1 Equilibrium geometries of the normal form, the tautomer form and the transition state (TS) in the S0 and S1 states of 7AI–H2O. Interatomic distances are given in A . Relative CASSCF energies are also shown, where the numbers in parentheses are vibrational zero-point corrected values. Reprinted with permission from [15]. Copyright 2008 American Chemical Society
582 Hydrogen Bonding and Transfer in the Excited State
distances and relative energies are given. Comparing the S0 and S1 equilibrium structures, it is observed that most CC and CN bond distances increase upon photoexcitation. As a result, vibrational excitation of the 7AI ring skeleton may be induced by photoexcitation. The normal form is lower in energy than the tautomer by ca 13 kcal mol1 in the S0 state, while the tautomer is lower in energy than the normal form by ca 32 kcal mol1 in the S1 state. The activation barriers relative to the normal isomer are 39.7 and 16.9 kcal mol1, respectively, in the S0 and S1 states, so the H transfer reaction occurs more easily in the S1 state. The energy profiles for the tautomerization in 7A–(H2O)2 are almost the same as those in 7AI–H2O. The AIMD trajectory simulations were started on the S1 excited state from the ground-state equilibrium structure for the normal form of 7AI–(H2O)n. Figure 25.2 displays the time evolution of the relative energy between the S0 and S1 states ((a) 7AI–H2O and (b) 7AI–(H2O)2) and the interatomic distances related to the H transfer ((c) 7AI–H2O and (d) 7AI–(H2O)2) along the AIMD trajectory. In the case of 7AI–H2O, the molecule stays in the normal form region of the potential energy surface during the initial 40 fs, with fluctuations in the excitation energy with a period of 10 fs (Figure 25.2(a)). The excitation energy in Figure 25.2(a) decreases rapidly at 40–60 fs, and the hydrogen bond distances N H and O H decrease to covalent bond lengths, indicating that ESHT occurs through the hydrogen-bonded network. The N H and O H hydrogen bond lengths initially exhibit in-phase vibrational motions, but their relative phases gradually change, and, just before the H transfer at t ¼ 40 fs, they become completely out of phase, indicating an asynchronous H transfer. The H is transferred from H2O to 7AI via the N H hydrogen bond, resulting in 7AI–Hþ OH around t ¼ 50 fs. The net charge on OH is 0.49 using charges based on the electrostatic potential (ESP) fitting method. So, the transferring H carries a significant positive charge, but it is not a true proton transfer. Following
Figure 25.2 Time evolution of the relative energy of S0 and S1 states ((a) 7AI–H2O and (b) 7AI–(H2O)2) and the interatomic distances related to the proton transfer ((c) 7AI–H2O and (d) 7AI–(H2O)2) along the AIMD simulation. Reprinted with permission from [15]. Copyright 2008 American Chemical Society (See Plate 34)
QM/MM Study of Excited-State Solvation Dynamics of Biomolecules
583
this process, the second H transfer occurs promptly from the NH on the five-member ring to OH, completing the tautomerization of 7AI. After the H transfer, the molecular system acquires considerable energy owing to the stabilization of the tautomer in the S1 state, and the 7AI and H2O fragments separate, breaking the two hydrogen bonds. This dynamical behaviour indicates that the ESHT reaction in 7AI–H2O proceeds in a concerted manner via the TS structure shown in Figure 25.1, but rather asynchronously. It is important to consider the experimental study of Takeuchi and Tahara [25] for the double-proton transfer in the 7AI dimer in solution, in which the 7AI dimer converts to the tautomer configuration in the excited state. They investigated this process by excitation wavelength dependence in steady-state and femtosecond timeresolved fluorescence spectroscopy, and concluded that it proceeds in a concerted manner. They noted, however, that this conclusion does not necessarily translate to a synchronous motion of the two protons. A concerted mechanism does not imply strict simultaneity, only that the motions of the two protons are correlated. In the case of 7AI–(H2O)2, changes in the interatomic distances (Figure 25.2(d)) indicate that tripleH-transfer relays occur in the time range t ¼ 40–60 fs, accompanying the gradual decrease in the excitation energy as shown in Figure 25.2(b). A detailed picture of this process is summarized in the following three steps: (1) H2 moves from N1 to O3; (2) H4 moves from O3 to O5; (3) H6 moves from O5 to N7. In the first step, the sum of the net atomic charges on the H5O2 moiety is þ 0.50 according to the ESP population analysis. The reaction mechanism is a concerted asynchronous process. As shown in Figure 25.2(b), the adiabatic energies of the S0 and S1 states approach each other in the tautomer region, suggesting the possibility of a non-radiative decay through a conical intersection of the potential energy surfaces. In order to examine this possibility, the minimum energy conical intersection (MECI) point between the S0 and S1 states was determined for an isolated 7AI molecule. In the MECI structure, the participating H atom attached to the six-member ring is out of the molecular plane. Therefore, when an H is transferred from the out-of-plane side to the six-member-ring N atom, 7AI could easily reach this conical intersection point, suggesting a non-radiative decay process after the phototautomerization in the 7AI–(H2O)n (n ¼ 1, 2) cluster. The activation barrier for tautomerization in an isolated 7AI molecule is much higher than that in 7AI–H2O because, without a bridge of water molecules, an H needs to be directly transferred from the five-member-ring N atom to the six-member-ring N atom [27]. Note that the addition of dynamic correlation via CASPT2 significantly reduces the activation barrier for the reaction in both the S0 and S1 states. In particular, the CASPT2 activation barrier for 7AI–H2O in the S1 state is predicted to be only 2.5 kcal mol1 at the state-averaged CASSCF geometries. This small activation barrier indicates that, at room or higher temperatures, the H transfer reaction does not require quantum tunnelling to proceed. Now, consider aqueous solvation effects on the ESHT dynamics of 7AI–H2O and 7AI–(H2O)2. To simulate the ESHT processes in water, 100 Hartree–Fock (HF)-based effective fragment potential (EFP1/HF) water molecules were distributed around 7AI–(H2O)n. The RHF/STO-3G-AIMD simulations for 7AI–(H2O)n–EFP were performed in the ground state for 10 ps at a constant temperature of 300 K with a time step of 1 fs, where velocity scaling was applied for a time interval of 0.25 ps, to control the temperature of EFP–waters at 300 K, and the geometry of the solute 7AI–(H2O)n was fixed. The temperature undergoes a small fluctuation around 300 K for t > 1.0 ps. Twenty different configurations and velocities of EFPs were chosen from the above AIMD trajectories and used as the initial conditions for the AIMD simulations in the excited state. The AIMD simulations for phototautomerizations in 7AI–(H2O)n–EFP were performed at the CASSCF level of theory with the segmented DZP basis set. During the simulations, each EFP–water maintains a hydrogen bond with other EFP or ab initio waters. When one hydrogen bond breaks, a new hydrogen bond forms. In the AIMD simulations for 7AI–H2O–EFP, CASSCF does not converge in one trajectory; among the remaining 19 trajectories, the ESHT reaction occurs in 15 trajectories, while no reaction occurs in three trajectories. In one trajectory, the ESHT reaction occurs once, but promptly the system returns to the
584 Hydrogen Bonding and Transfer in the Excited State
normal-form region, which is an example of transition-state recrossing. This suggests that the ESHT of the 7AI–H2O cluster takes place asynchronously in both the gas phase and in solution. In most of the reactive trajectories, the ESHT occurs faster compared with that in the gas phase by one period of the N H bondlength oscillation, presumably owing to the thermal energy of the surrounding solvent molecules. ESHT was not observed in three trajectories, suggesting that the solvent water molecules could diminish the probability of ESHT, depending on the configuration of the surrounding water molecules. AIMD simulations were also carried out for 7AI–(H2O)2–EFP, starting from 20 different initial conditions. The triple-proton transfer occurs in 14 trajectories, while no reaction occurs in five trajectories. Recrossing behaviour is seen in one trajectory. Unlike 7AI–H2O–EFP, the surrounding solvent waters affect the ESHT dynamics more significantly, resulting in different proton transfer mechanisms, depending on the initial solvent configurations. In 12 trajectories, the ESHT dynamics proceeds asynchronously via an H2O–OHd 7AI-Hdþ species in nearly half of the trajectories, while the other half of the trajectories exhibit an H2O–Hdþ –H2O 7AId species. In two trajectories, an almost-synchronous proton transfer is observed. This work reports classical dynamics simulations that do not include tunnelling contributions, which could play an important role in the ESHT process. One needs a quantum dynamical approach to include such quantum effects in the excited-state dynamics simulation. In solution at standard thermodynamic conditions, however, the total system consists of a large number of solute and solvent molecules, and is expected to have sufficient energy to surpass the activation barrier upon photoexcitation. Although the present results are obtained from classical dynamics simulations, they nonetheless provide useful information related to ESHT reaction dynamics. 25.2.2 Excited-state dynamics of coumarin 151 7-Aminocoumarins are a well-known important group of chromophores that emit in the blue-green spectral region, with their fluorescence quantum yields often close to unity [36–38]. As the structures contain an electron-donating amino group at the 7-position and an electron-withdrawing carbonyl group, 7-aminocoumarins are polarized in the ground state. For most of the 7-aminocoumarins there is a large change in dipole moment upon photoexcitation to the first singlet state [36–38]. All of the 7-aminocoumarins show very large Stokes shifts between their absorption and fluorescence maxima. These Stokes shifts are again very sensitive to the solvent polarities [38–40]. In addition, the 7-aminocoumarins are capable of forming hydrogen bonds with solvent molecules; this can influence the solvation dynamics. Owing to these interesting properties, many 7aminocoumarins have been widely used as probes to elucidate a variety of physicochemical processes in the condensed phase. These properties include solvatochromic behaviour, polarities of different environments and measurement of solvent relaxation times using the dynamic Stokes shift method [39, 40]. Although many theoretical [41–44] and experimental [45, 46] studies have been carried out on photophysical properties of 7-aminocoumarins, there remains unusual behaviour that has not yet been understood. Rechthaler and K€ ohler [45] investigated the photophysical properties of several 7-aminocoumarins in diverse organic solvents and found that, unlike other 7-aminocoumarins, the fluorescence quantum yields are very low for coumarin 120 (C120) and coumarin 151 (C151) in non-polar solvents, such as hexane and heptane. Later, Nad and Pal [46] examined this process using picosecond laser flash photolysis and pulse radiolysis techniques, and suggested that the non-radiative decay process of C151 is attributed not to transition to the triplet state but to the ground state (S0) via the umbrella motion of the 7-amino group. Cave et al. [41, 42] performed a detailed theoretical study on the ground and first-excited singlet states of C120 and C151. They found that time-dependent (TD) DFT gives good agreement with the experimental gas phase S0–S1 excitation energies; however, TDDFT combined with a dielectric continuum model overestimates the excitation energies in protic solvents, suggesting that an explicit description of solute–solvent interactions is important in these systems. In another survey with hybrid QM/MM TDDFT-MD simulations [43, 44] the red-shifts of three
QM/MM Study of Excited-State Solvation Dynamics of Biomolecules
585
aminocoumarins (C151, C35 and C153) caused by water and acetonitrile solvents were successfully reproduced. In these studies, however, the unusually strong quenching of C120 and C151 in non-polar solvents is not well understood. In our previous study [16], geometry optimizations were carried out for C151 in the ground and first excited singlet (S0 and S1) states by the state-specific CASSCF method with segmented DZP basis sets [31–34] using GAMESS [12, 13]. Preliminary calculations determined the CASSCF active space to be six electrons in six orbitals of C151. RHF/DZP-AIMD simulations were performed for C151 in the S0 state for 10 ps at 300 K with a time step of 1 fs. Then, eight different sets of atomic positions and velocities were taken from the above trajectory with a fixed time interval and used as the initial conditions for the subsequent AIMD simulations in the S1 state. CASSCF/DZP AIMD simulations were performed for isolated C151 in the S1 state, with a time step of 0.5 fs. To simulate the solvation processes of C151 in water, 150 EFP1/HF water molecules were distributed around C151. The preliminary RHF/DZP-AIMD simulations were performed for the C151–EFP system in the ground state for 10 ps at 300 K with a time step of 1 fs, and eight different configurations and velocities were chosen as the initial conditions in the same way as in the simulations of the isolated C151 molecule. The CASSCF/DZP AIMD simulations were performed for excited-state solvation processes of C151 in water, and the solute–solvent interactions were investigated from a dynamical point of view. The time step was taken as 0.5 fs throughout the simulation. Figure 25.3 shows the equilibrium geometry for C151 in the S0 and S1 states with geometrical parameters and dipole moments, calculated by the state-specific CASSCF method. In the S0 state, C151 has an almost planar structure, except for the amino group. Owing to the electron-donating amino group and electronwithdrawing carbonyl group, C151 has a relatively large dipole moment of 5.0 D in the S0 state. The CASSCF geometry optimization located two minimum energy structures on the S1 potential energy surface, denoted as I and II. In structure I, the amino group stays almost in the molecular plane while the six-member ring with the carbonyl group is deformed from the plane. In structure II, the six-member ring with the carbonyl group maintains planarity, but the amino group rotates relative to the molecular plane. Structures I and II have similar energies relative to the S0 minimum, while the S0 energy at the two excited-state structures is
Figure 25.3 Equilibrium structures of C151 in the S0 and S1 states. Bond lengths (in A ), the distance of N1 atom relative to the C5H1H2 plane, r(N1–C5H1H2), the dihedral angle d(C1C2C3O1), dipole moments (indicated by the arrow) and relative energies to the S0 equilibrium point are given. Reprinted with permission from [16]. Copyright 2009 Wiley Periodicals, Inc., A Wiley Company
586 Hydrogen Bonding and Transfer in the Excited State
76.0 kcal mol1 (structure I) and 38.2 kcal mol1 (structure II). This indicates that the energies of the S1 and S0 states are close to each other near structure I. Dipole moments were calculated to be 11.1 and 8.3 D for structures I and II respectively. Their orientations are almost in parallel to that of the dipole moment in the S0 state. In both structures, the S1 state has a HOMO–LUMO single-electron excitation as the dominant configuration. The electron density distributions in the HOMO and LUMO are consistent with the direction of the dipole moment in the S1 state. AIMD simulations were performed for an isolated C151 in the S1 state, starting from eight different initial conditions. These simulations may be regarded as approximate excited-state dynamics simulations of C151 in the presence of non-polar solvents. Analyses of the AIMD simulations show that trajectories can be classified into two types: in three trajectories, C151 exhibits vibrational motion around structure I and reaches the crossing point of the S0 and S1 states, while in five other trajectories, C151 remains near structure II with vibrational motions. Figure 25.4 illustrates the time evolution of the energy of the S0 and S1 states along the AIMD trajectories for (a) C151 and (b) C151–EFP. Figure 25.4(a) illustrates that the adiabatic energies of the S0 and S1 states approach each other around t 340 fs, indicating the possibility of a non-radiative decay through a crossing point of the two potential energy surfaces. The energy of the S1 state is almost constant, with a small fluctuation, while the energy of the S0 state increases by ca 40 kcal mol1 in the initial stage, and again starts to increase around t 220 fs, reaching the crossing point. Here, the S0 energy is estimated using the statespecific CASSCF wave function that was obtained for the S1 state. State-specific CASSCF calculations for the S0 state were also performed at selected points along the S0 trajectory, and the energies obtained in this manner are depicted in the same figure. As expected, the S0 energy obtained in this manner is lower than the ones obtained using the state-specific CASSCF wave function for the S1 state, but the energies for the two calculations show a similar trend in changes along the trajectory. The dipole moment in the S1 state remains close to the large value of 10 D, and then starts to decrease near the crossing point. As shown in Figure 25.4(b), the variations in the S1 and S0 energies are essentially parallel to each other, and this is very different from the AIMD simulations on isolated C151 shown in Figure 25.4(a). The energies of the two states become close to each other occasionally, but no crossing point is observed during the 500 fs of the trajectory. The C151–EFP S1 potential energy exhibits strong fluctuations compared with those in Figure 25.4(a), suggesting a strong electrostatic interaction between C151 and the surrounding EFP waters. The S1 dipole moment is very large owing to the strong polarization effects from the solvent, which could be responsible for the large fluctuations of the S1 potential energy. The solvated S0 dipole moment is also large, 12 D, pointing in almost the same direction as the S1 dipole moment. This finding suggests that the electronic
Figure 25.4 A time evolution of the energy of S0 and S1 states relative to the S0 energy at the equilibrium geometry along the AIMD trajectory for (a) C151 and (b) C151 þ EFP. The S0 energies calculated by the state-specific CASSCF method for the S0 state are shown by the dashed line (S00 ). Reprinted with permission from [16]. Copyright 2009 Wiley Periodicals, Inc., A Wiley Company (See Plate 35)
QM/MM Study of Excited-State Solvation Dynamics of Biomolecules
587
wave functions of the S1 and S0 states are similar to each other in this highly polar solvent. This could partly explain the parallel variation in the potential energies.
25.3 Concluding Remarks In this review we have described two applications of an AIMD approach with a combination of the EFP method, which can be used to examine excited-state reaction mechanisms and dynamics in solution. In both 7AI–H2O and 7AI–(H2O)2 clusters, an asynchronous hydrogen transfer occurs via water molecules at t 50 fs in a concerted manner, after the photoexcitation. While the ESHT mechanism for 7AI–H2O in water does not change appreciably compared with that in the gas phase, AIMD simulations on 7AI–(H2O)2 in water solution exhibit two different mechanisms. Using the results of the AIMD trajectories, the minimum energy conical intersection point in the tautomer region has also been located. In AIMD simulations for isolated C151, two patterns are observed for the dynamics: (a) C151 decays from S1 to S0 via a crossing point near structure I (charge transfer state); (b) C151 vibrates on the S1 state. In AIMD simulations for C151 in the presence of 150 EFP waters, the S1 and S0 energy variations are very similar in all of the trajectories, so no crossing point is observed. Assuming that the gas-phase simulation does correspond to a non-polar solvent, this indicates that C151 in a polar solvent is likely to remain on the S1 potential energy surface for a longer time than in a nonpolar solvent. This is in agreement with experimental observations.
Acknowledgements This work was supported in part by a grant-in-aid for scientific research from the Ministry of Education, Culture, Sports, Science and Technology, and in part by a grant from the US Department of Energy via the Ames Laboratory.
References 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13.
14. 15. 16. 17.
L.G. Arnaut and S.J. Formosinho, J. Photochem. Photobiol. A, 75, 1 (1993). S.J. Formosinho and L.G. Arnaut, J. Photochem. Photobiol. A, 75, 21 (1993). A. Douhal, F. Lahmani and A.H. Zewail, Chem. Phys., 207, 477 (1996). M. S. Gordon, G. Chaban and T. Taketsugu, J. Phys. Chem., 100, 11512 (1996). T. Taketsugu, A. Tajima, K. Ishii and T. Hirano, Astrophys. J., 608, 323 (2004). M. Kayanuma, T. Taketsugu and K. Ishii, Chem. Phys. Lett., 418, 511 (2006). M. Kayanuma, T. Taketsugu and K. Ishii, Theor. Chem. Acc., 120, 191 (2008). A. Warshel and M. Levitt, J. Mol. Biol., 103, 227 (1976). S. Yoo, F. Zahariev, S. Sok and M.S. Gordon, J. Chem. Phys., 129, 144 112 (2008). P.N. Day, J.H. Jensen, M.S. Gordon, et al. J. Chem. Phys., 105, 1968 (1996). M.S. Gordon, M.A. Freitag, P. Bandyopadhyay, et al. J. Phys. Chem. A, 105, 293 (2001). M.W. Schmidt, K.K. Baldridge, J.A. Boatz, et al. J. Comp. Chem., 14, 1347 (1993). M.S. Gordon and M.W. Schmidt, “Advances in Electronic Structure Theory: GAMESS a Decade Later” Theory and Applications of Computational Chemistry, Ch. 41, C.E. Dykstra; G. Frenking; K.S. Kim; G.E. Scuseria Eds., Elsevier, 2005. M. Krauss and S.P. Webb, J. Chem. Phys., 107, 5771 (1997). D. Kina, A. Nakayama, T. Noro, et al. J. Phys. Chem. A, 112, 9675 (2008). D. Kina, P. Arora, A. Nakayama, et al. Int. J. Quantum Chem., 109, 2308 (2009). C.A. Taylor, M.A. El-Bayoumi and M. Kasha, Proc. Natl. Acad. Sci. USA, 63, 253 (1969).
588 Hydrogen Bonding and Transfer in the Excited State 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31.
32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46.
K.C. Ingham and M.A. El-Bayoumi, J. Am. Chem. Soc., 96, 1674 (1974). P.-T. Chou, M.L. Martinez, W.C. Cooper, et al. J. Phys. Chem., 96, 5203 (1992). A. Douhal, S.K. Kim and A.H. Zewail, Nature, 378, 260 (1995). A. Nakajima, M. Hirano, R. Hasumi, et al. J. Phys. Chem. A, 101, 392 (1997). K. Sakota, Y. Komoto, M. Nakagaki, et al. Chem. Phys. Lett., 435, 1 (2007). K. Sakota, N. Inoue, Y. Komoto and H. Sekiya, J. Phys. Chem., 111, 4596 (2007). H. Sekiya, private communication. S. Takeuchi and T. Tahara, Proc. Natl. Acad. Sci. USA, 104, 5285 (2007). M.S. Gordon, J. Phys. Chem., 100, 3974 (1996). M.G. Chaban and M.S. Gordon, J. Phys. Chem. A, 103, 185 (1999). R. Casadesu´s, M. Moreno and J.M. Lluch, Chem. Phys., 290, 319 (2003). T. Taketsugu, K. Yagi and M.S. Gordon, Int. J. Quant. Chem., 104, 758 (2005). H. Nakano, J. Chem. Phys., 99, 7983 (1993). This basis set consists of a Hartree-Fock set (¼ Tatewaki-Koga’s all-electron, non-relativistic, segmented contraction] plus a correlating set (¼ natural orbital based segmented cgtf, including valence correlation only] that is referred to as TK/NOSeC-V-DZP. See http://setani.sci.hokudai.ac.jp/sapporo/ H. Yamamoto and O. Matsuoka, Bull. Univ. Electro. Comm., 5, 23 (1992). T. Noro, M Sekiya and T. Koga, Theor. Chem. Acc., 98, 25 (1997). T. Noro, M. Sekiya and T. Koga, Theor. Chem. Acc., 109, 85 (2003). R.W. Hockney and J.W. Eastwood, Computer Simulations Using Particles, McGraw-Hill, New York, 1981. E.J. Schimitschek, J.A. Trias, P.R. Hammond, et al. Opt. Commun., 16, 313 (1976). R.L. Atkins and D.E. Bliss, J. Org. Chem., 43, 1975 (1978). Jones G., II, W.R. Jackson and A.M. Halpern, Chem. Phys., 72, 391 (1980). K. Tominaga and G.C. Walker, J. Photochem. Photobiol. A, 87, 127 (1995). J.A. Gardecki and M. Maroncelli, J. Phys. Chem., 103, 1187 (1999). R.J. Cave, K. Burke and Castner E.W. J. Phys. Chem. A, 106, 9294 (2002). R.J. Cave and Castner E.W., J. Phys. Chem. A, 106, 12117 (2002). M. Sulpizi, P. Carloni, J. Hutter and U. Rothlisberger, Phys. Chem. Chem. Phys, 5 4798 (2003). M. Sulpizi, U.F. R€ohrig, J. Hutter and U. Rothlisberger, Int. J. Quantum Chem., 101, 671 (2005). K. Rechthaler and G. K€ohler, Chem. Phys., 189, 99 (1994). S. Nad and H. Pal, J. Phys. Chem. A, 105, 1097 (2001).
26 Excited-State Intramolecular Proton Transfer Processes on Some Isomeric Naphthalene Derivatives: A Density Functional Theory Based Computational Study Sankar Prasad De and Ajay Misra Department of Chemistry and Chemical Technology, Vidyasagar University, Midnapore 721 102, W.B, India
26.1 Introduction The proton transfer reaction has been found to occur extensively in both chemical and biological processes. Proton transfer may be of intra- or intermolecular nature, and each intra- and intermolecular process can occur either in the ground or in the excited state. Among the various types of proton transfer process, excited-state intramolecular proton transfer (ESIPT) has received much attention owing to its importance in both chemical and biological processes. Numerous ESIPT molecules have been strategically designed and synthesized with the aim of shedding light on the fundamentals of the proton transfer mechanism and/or exploring their potential applications. ESIPT reactions are of great scientific and technological interest. Since its introduction, the photoinduced excited-state intramolecular proton (or hydrogen) transfer reaction, which generally incorporates transfer of a hydroxyl (or amino) proton to the carbonyl oxygen (imine nitrogen) through a pre-existing intramolecular hydrogen bonding (IMHB) configuration, has received considerable attention because it has led to a wide range of applications, such as laser dyes [1, 2], polymer stabilizers [3, 4], environmental probes in biomolecules [5], etc. The main requirement of the ESIPT reaction is that the molecule must have acid and basic groups and at the same time a strong intramolecular hydrogen bond between the two groups. The large number of molecules belonging to this class include, for example, o-hydroxybenzoyl [6–16], o-hydroxy Schiff bases [17–20] and so on. Hydrogen Bonding and Transfer in the Excited State, Volume I & II © 2011 John Wiley & Sons, Ltd. ISBN: 978-0-470-66677-7
Edited by Ke-Li Han and Guang-Jiu Zhao
590
Hydrogen Bonding and Transfer in the Excited State
Weller [21, 22], in his pioneering work, had pointed out the dual emission in the fluorescence spectra of salicylic acid and methyl salicylate and attributed it to the asymmetric double well potential arising from proton transfer in the ground as well as in the excited state (Scheme 26.1). The two wells in the ground-state potential energy curve represent the primary (N) and tautomeric (T) forms, and the two wells in the excitedstate curve represent the corresponding excited states N and T respectively. It is clear from Scheme 26.1 that the N form is the most stable in the ground state, and T in the excited state. Since the original work of Weller [21] on the ESIPT of methyl salicylate (MS), a large number of experimental [23–30] and theoretical [31–37] studies on the ESIPT of a variety of systems have been reported. Among them, MS [38–44] and its related compounds o-hydroxyacetophenone (OHAP) [45–47] and o-hydroxybenzaldehyde (OHBA) [48–51] have been well studied as prototypes of molecules showing the ESIPT process. Although theoretical studies on ESIPT on MS and related compounds are quite abundant, similar information about their naphthalene analogue is scant. Evidence for the existence of an intramolecular hydrogen bond in methyl-2-hydroxy-3-naphthoate (MHN23) has been reported by Bergmann et al. [52]. The dual emission of MHN23 was first reported by Naboikin et al. [53]. MHN23 possesses a strong IMHB in the ground electronic state and also shows ESIPT upon excitation to its excited singlet states. On the other hand, the lack of ESIPT emission from methyl-1-hydroxy-2-naphthoate (MHN12) and methyl-2-hydroxy-1naphthoate (MHN21) had been explained by Catalan et al. [54] in terms of the non-radiative dynamics of their respective normal tautomers. Shizuka et al. [55] carried out a comprehensive study on ESIPT of 1-hydroxy-2-acetonaphthone (1H2AN), 1-hydroxy-2-naphthaldehyde (1H2NA) and methyl-1-hydroxy2-naphthoate (1H2MN) by means of laser photolysis, time-resolved thermal lensing and the fluorometry method. They found that both 1H2NA and 1H2AN show ESIPT, whereas 1H2MN gives no ESIPT emission. They also observed a structured ESIPT emission band (lem ¼ 476 nm) for 1H2NA with very little Stokes shift. They showed that distinct relaxation properties of 1H2NA, 1H2AN and 1H2MN are responsible for the relative stabilities between the parent enol (N) form and tautomeric keto (T) form in the lowest excited singlet states of these compounds. On the other hand, laser-induced fluorescence studies by Wu et al. [56] showed a weak tautomeric emission band maximized at 715 nm for 2-hydroxy-3-naphthaldehyde (2H3NA) in cyclohexane at 298 K, with the fluorescence quantum yield as low as 6.8 106. 1H2NA and 2H3NA differ only in the relative
N*
T*
P.E.
T N
Proton transfer coordinate
Scheme 26.1
Excited-State Intramolecular Proton Transfer Processes 591
position of the OH and CHO groups, but they show a wide difference in their fluorescence properties. This wide difference in fluorescence properties of these two compounds motivates us to carry out the present extensive quantum mechanical investigations in order to gain some insight into their electronic structure. Hybrid HF/DFT methods have been proposed as a reliable tool for electronic computation in a general protocol for studying the static and dynamic properties of hydrogen-bonded systems [54, 57, 58]. One such method, B3LYP [59], predicts molecular data that match with the available experimental data as well as with results obtained using the highest post-HF method [60]. In view of its widespread success for the calculation of large molecules [58], we decided to choose the density functional approach for the present calculation of the excited-state proton transfer process in 1H2NA and 2H3NA.
26.2 Theoretical Calculations All ab initio calculations reported in this chapter were carried out using the Gaussian 03 suite of programs [60]. We compared the results for a number of methods and basis sets and found the DFT-based calculations using a hybrid functional (B3LYP) with the 6-31G basis set to be optimal in terms of price–performance ratio for carrying out elaborate electronic structure calculations within our limited computational resources. Analytic vibrational frequency computations at the optimized structure were done to confirm the optimized structure to be an energy minimum or a transition structure. The strength of the intramolecular hydrogen bond (IMHB) of each molecule studied was evaluated as the difference between the energy of the fully optimized structure of the non-hydrogen-bonded form (hydroxyl group rotated by 180 , i.e. open-form Scheme 26.2) and the energy of the N tautomer. Ground-state intramolecular proton transfer (GSIPT) curves were calculated with energies of the B3LYP/6 31G fully optimized structures at fixed OH distances over the 0.8–2.0 A range. Information on the ESIPT mechanism was obtained by calculating the Franck–Condon (FC) transition energies for the DFT (B3LYP)/631G ground-state structures at the TD-DFT(B3LYP)/6-31G level. The Franck–Condon (FC) curves for the proton transfer processes were obtained by adding the TD-DFT(B3LYP)/6-31G excitation energies to the corresponding GSIPT curves. Free energy calculations on the optimized ground-state N and T forms of both compounds were done using RHF/6-31G(d) level of theory.
26.3 Results and Discussion The ground-state optimized structure of both 1H2NA and 2H3NA shows that the enol (N) form is the stable structure having strong intramolecular hydrogen bonding (Scheme 26.2).
1' 1.005
H
O
O 1.409
4'
C 1.442
O
1.370
6'
1.266
1.360
3'
2'
1.705
2'
1.439
5'
H
3'
4' 5'
C H
1H2NA (enol form)
2H3NA (enol form)
Scheme 26.2
1'
1.783 1.451
H
0.995
O 6' 1.258
592 Hydrogen Bonding and Transfer in the Excited State
Table 26.1 Calculated bond length (in A), bond angle (in deg) and dihedral angle (in deg) at the DFT-B3LYP(631G) level for the IMHB ring (Scheme 26.2) of the N form of 1H2NA 1H2NA Bond length H(1)O(2) 1.005 O(2)C(3) 1.360 C(3)C(4) 1.409 C(4)C(5) 1.442 C(5)O(6) 1.266 O(6)H(1) 1.705
Bond angle
Dihedral angle
H(1)O(2)C(3) 109.262 O(2)C(3)C(4) 121.570 C(3)C(4)C(5) 119.846 C(4)C(5)O(6) 123.926 C(5)O(6)H(1) 100.979
H(1)O(2)C(3)C(4) 0.000 C(3)C(4)C(5)O(6) 0.000 C(5)O(6)H(1)O(2) 0.012
Bond length, bond angle and dihedral angle of the six-member ring system containing an intramolecular hydrogen bond of both 1H2NA and 2H3NA, as shown in Tables 26.1 and 26.2, give some idea about the ground-state geometry of these two compounds. The ground-state bond angle and dihedral angle data given in Tables 26.1 and 26.2 suggest that the sixmember rings formed by IMHB for 1H2NA and 2H3NA are planar and are in the same plane as the naphthalene ring. A critical analysis of the bond length data shows that the C(30 )C(40 ) and O(60 )H(10 ) bonds are much shorter in 1H2NA than in 2H3NA. A shorter O(60 )H(10 ) distance is a measure of stronger IMHB in 1H2NA. Again the double-bond character at C(30 )C(40 ) in 1H2NA and the single-bond character at C(30 )C(40 ) in 2H3NA support the fixed double-bond character of the naphthalene ring. In other words, IMHB in 1H2NAwill be more conjugated, and hence the strength of hydrogen bonding will be greater compared with 2H3NA. In order to get some idea about the relative strength of IMHB in 1H2NA and 2H3NA, we compare the C¼O and OH stretching frequencies of these two compounds with some model compounds like 1-naphthol, 2-naphthol and 2-naphthaldehyde (Table 26.3). We used B3LYP/6-31G(d) level of theory for frequency calculations, and 0.9613 was used as the scale factor for frequencies, as given in Ref. [61]. Table 26.3 shows that our methodology for calculation of vibrational frequency works nicely, as our calculated C¼O and OH stretching frequencies agree well with the experimental results. It is reasonable to infer from both the experimental as well as theoretical calculations that the position of OH absorption in these monosubstituted compounds (1-naphthol and 2-naphthol) is independent of substitution. Hunsberger [62] showed that, for the disubstituted compounds (1H2NA and 2H3NA), the displacement of the OH band (DnOH) from its average
Table 26.2 Calculated bond length (in A), bond angle (in deg) and dihedral angle (in deg) at the DFT-B3LYP (6-31G) level for the IMHB ring (Scheme 26.2) of the N form of 2H3NA 2H3NA Bond length H(1)O(2) 0.995 O(2)C(3) 1.370 C(3)C(4) 1.439 C(4)C(5) 1.451 C(5)O(6) 1.258 O(6)H(1) 1.783
Bond angle
Dihedral angle
H(1)O(2)C(3) 109.981 O(2)C(3)C(4) 121.146 C(3)C(4)C(5) 120.696 C(4)C(5)O(6) 124.094 C(5)O(6)H(1) 100.965
H(1)O(2)C(3)C(4) 0.000 C(3)C(4)C(5)O(6) 0.010 C(5)O(6)H(1)O(2) 0.014
Excited-State Intramolecular Proton Transfer Processes 593 Table 26.3 Theoretical and experimental carbonyl and hydroxyl stretching frequency values of 1H2NA, 2H3NA and some model compounds Compounds
Expa.
Theor. a-naphthol b-naphthol 2-Naphthaldehyde 1H2NA 2H3NA
O–H stretching frequencies (incm1)
C¼O stretching frequencies (cm1)
1729 1654 1675
1702 1637 1670
Theor.
Exp.a
3608 3609
3618 3618
3097 3280
3178 3249
DnC¼O
DnO-H
Theor.
Exp.a
Theor.
Exp.a
75 54
64 31
511 329
440 369
a
Experimental values of IR frequencies (in 0.02 molal CCl4) are obtained from Ref. 62.
position in 1-naphthol and 2-naphthol and the displacement of the C¼O band (DnC¼O) from its average position in the corresponding monosubstituted compound, i.e 2-naphthaldehyde, are taken as a quantitative measure of the strength of IMHB in the disubstituted compounds. Greater red-shift of the band positions from its monosubstituted compound, i.e. larger values of DnOH and DnC¼O, are taken as evidence for corresponding stronger IMHB. Both the experimental and theoretical data of DnOH and DnC¼O in Table 26.1 suggest that 1H2NA has stronger IMHB than 2H3NA. The strength of the intramolecular hydrogen bond of the enol form (N) was calculated by rotating the phenolic OH group out of the hydrogen-bonded conformation and computing the difference in energy between the closed and open form for 1H2NA and 2H3NA and is shown in Figures 26.1 and 26.2 respectively. The corresponding calculated values are given in Table 26.4. The calculated IMHB strength for 1H2NA and 2H3NA are found to be 17.14 and 12.34 kcal mol1 respectively. Figures 26.1 and 26.2 also show that the barrier for phenolic OH rotation is 19.08 kcal mol1 in 1H2NA and 15.25 kcal mol1 in 2H3NA. Thus, the relative strength of IMHB is 5 kcal mol1 greater in 1H2NA than in 2H3NA.
Energy / Kcal mol - 1
20
15
10
5
0 0
20
40
60
80
100 120 140 160 180 200
Dihedral angle (H1'-O2'-C3'-C4') / Degree
Figure 26.1 Energetics of the transformation from IMHB from (N) to the non-hydrogen-bonded form (hydroxyl group rotated by 180 ) of 1H2NA. For each value of the dihedral angle (H(10 )O(20 )C(30 )C(40 )), the geometry has been optimized using the DFT-B3LYP(6-31G) level of theory. Reprinted with permission from [65]. Copyright Elsevier
594
Hydrogen Bonding and Transfer in the Excited State 18
Energy / Kcal mol - 1
16 14 12 10 8 6 4 2 0 0
20
40
60
80
100
120
140
160
180
200
Dihedral angle (H1'-O2'-C3'-C4') / Degree
Figure 26.2 Energetics of the transformation from IMHB from (N) to the non-hydrogen-bonded form (hydroxyl group rotated by 180 ) of 2H3NA. For each value of the dihedral angle (H(10 )O(20 )C(30 )C(40 )), the geometry has been optimized using the DFT-B3LYP(6-31G) level of theory. Reprinted with permission from [65]. Copyright Elsevier
Table 26.4 Energy (in kcal mol1) at various dihedral angles (H10 O20 C30 C40 ) of 1H2NA and 2H3NA, as obtained using DFT-B3LYP(6-31G) level of theory 1H2NA Dihedral angle (H10 O20 C30 C40 ) (deg) 0.0000 8.0429 16.2478 24.8341 34.1085 44.6711 56.1993 68.2001 80.2491 92.0792 103.5461 114.5939 124.9997 134.7501 144.0421 153.0322 161.9035 170.8967 180.0000
2H3NA Energy (kcal mol1)
Dihedral angle (H10 O20 C30 C40 ) (deg)
Energy (kcal mol1)
0.00000 0.47561 1.85288 3.99800 6.69213 9.62155 12.42010 14.81030 16.68163 17.99987 18.78274 19.08348 18.98950 18.62970 18.15252 17.70268 17.37548 17.19588 17.14143
0.0000 8.8077 17.7134 26.9077 36.5782 46.9904 58.1623 69.8019 81.4787 92.9985 104.3356 115.2993 125.8265 135.8144 145.3008 154.3749 163.1061 171.6157 180.0000
0.00000 0.36711 1.43833 3.11904 5.24561 7.59349 9.90706 11.95480 13.60097 14.79426 15.50799 15.75039 15.57133 15.05793 14.34672 13.58978 12.93528 12.49629 12.34152
Excited-State Intramolecular Proton Transfer Processes 595
The conversion from N to T in the ground electronic state can be thought of as arising from proton transfer from Od(20 ) to Oa(60 ) with simultaneous redistribution of electron density within the six-member hydrogenbonded ring. Some authors have considered the OdOa distance as fixed and have varied the OdH bond distance to get an idea about the potential energy curve for both GSIPT and ESIPT processes. A plot of Od(20 )Oa(60 ) distance as a function of rOd _H of 1H2NA, as shown in Figure 26.3 (corresponding data are given in Table 26.5), reveals that, as the proton shifts from Od to Oa, the Od—Oa distance changes significantly. At smaller OdH distance it increases slowly. In the close vicinity of a stable OdH distance, the OdOa distance falls sharply. However, as the proton shifts further from Od, it decreases sharply, passes through a minimum and then enlarges to a distance comparable with that in the T form. This variation is almost identical in the case of 2H3NA (Figure 26.4). Figure 26.5 shows the variation in OdHOa angle as a function of rOH distance. At smaller OdH distance, it increases slowly. The OdHOa angle increases sharply in the near vicinity of the stable OdH distance. It increases with increase in rOH distance, reaches a maximum and then shows a sudden fall with further increase in rOH distance. We obtained similar variation in OdHOa angle with rOd --H in the case of 2H3NA (Figure 26.6). The corresponding data are given in Table 26.6. Therefore it becomes clear that, by freezing the geometry or by fixing the OdOa distance at a particular value, one ends up introducing artificial constraints on the system and hence a barrier for the enol (N) to keto (T) conversion. In this chapter we used the ‘distinguished coordinate’ approach as proposed by Sobolewski et al. [63], where the OdH bond distance is varied and the rest of the structural parameters are allowed to relax for each choice of rOH. Maheswari et al. [64] conducted an extensive theoretical study on salicylic acid and showed that the variation in OdH bond length can be used as a reaction coordinate in order to get some idea about the PE curve for the ground-state as well as for the excited-state proton transfer processes. Catalan et al. [54] used a similar reaction coordinate (rOH) for the ESIPT processes of some naphthalene derivatives.
2.65
2.60
2.50
d
a
rO -O / A0
2.55
2.45
2.40 0.50
0.75
1.00
1.25
1.50
1.75
0
rO-H / A
Figure 26.3 Variation in the Od(20 )Oa(60 ) distance of 1H2NA with rOd --Ha , as obtained from DFT-B3LYP(6-31G) level of theory. Reprinted with permission from [65]. Copyright Elsevier
596
Hydrogen Bonding and Transfer in the Excited State
Table 26.5 Value of OdOa bond distance at various OdH distances (in A) of 1H2NA and 2H3NA, as obtained using DFT-B3LYP(6-31G) level of theory
rOH (A)
rOd --Oa (A)
0.50 0.75 0.80 0.85 0.90 0.95 1.00 1.05 1.10 1.15 1.20 1.25 1.30 1.40 1.50 1.70 1.75 2.00
1H2NA
2H3NA
2.6000 2.6358 2.6373 2.6350 2.6281 2.6146 2.5928 2.5608 2.5161 2.4643 2.4212 2.3969 2.3919 2.4149 2.4608 2.5777 2.6090
2.6237 2.6676 2.6706 2.6713 2.6682 2.6606 2.6460 2.6237 2.5914 2.5477 2.4978 2.4567 2.4355 2.4404 2.4780 2.5887 2.6190 2.7771
26.3.1 Ground- and excited-state potential energy curve of 1H2NA The ground-state potential energy curve, as shown in Figure 26.7, the corresponding data of which are given in Table 26.7, reveals that the enol (N) form is the most stable one. Surprisingly, there is no shallow minimum for the keto form. The barrier for the enol (N) to keto (T) form is about 6.26 kcal mol1, and this is large enough to
2.75
2.70
2.60
d
a
r0 -O (A0)
2.65
2.55
2.50
2.45 0.6
0.8
1.0
1.2
1.4
1.6
1.8
2.0
2.2
rO -H(A0) d
0
0
Figure 26.4 Variation in the Od(2 )-Oa(6 ) distance of 2H3NA with rOd --Ha , as obtained from DFT-B3LYP(6-31G) level of theory
Excited-State Intramolecular Proton Transfer Processes 597 160
150
Od -H-Oa angle / 0
140
130
120
110
100
90 0.50
0.75
1.00
1.25
1.50
1.75
2.00
2.25
2.50
0
rO -H / A d
Figure 26.5 Variation in the Od—H—Oa angle of 1H2NA with rOd --Ha , as obtained from DFT-B3LYP(6-31G) level of theory. Reprinted with permission from [65]. Copyright Elsevier
make any GSIPT under thermal conditions. Again, the calculated free energy change for the keto–enol tautomerization of 1H2NA gives a large positive value (9.201 kcal mol1), and the equilibrium constant obtained from the free energy change, i.e 1.7 106, suggests that the above equilibrium lies towards the enol form. On the basis of the equilibrium constant, the population ratio in the gas phase for the enol form versus 152 150
Od-H-Oa (Degree)
148 146 144 142 140 138 136 134 0.6
0.8
1.0
1.2
1.4
1.6
1.8
2.0
0
rO -H (A ) d
Figure 26.6 Variation in the Od—H—Oa angle of 2H3NA with rOd --Ha , as obtained from DFT-B3LYP(6-31G) level of theory
598
Hydrogen Bonding and Transfer in the Excited State
Table 26.6 Value of the OdHOa angle (in deg) at various OdH distances (in A) of 1H2NA and 2H3NA, as obtained using DFT-B3LYP(6-31G) level of theory
rOH (A)
OdHOa angle (deg)
0.50 0.75 0.80 0.85 0.90 0.95 1.00 1.05 1.10 1.15 1.20 1.25 1.30 1.40 1.50 1.70 1.75 2.41
1H2NA
2H3NA
139.11581 139.82649 140.34876 140.99463 141.82576 142.91872 144.22973 145.94132 147.98119 150.06545 151.67221 152.36214 152.25716 150.76805 148.30341 142.28409 140.70802 94.33674
139.20331 139.73050 140.21231 140.72809 141.39689 142.21755 143.24496 144.53978 146.12801 148.00281 149.88770 151.64820 151.41013 148.50879 142.74887 141.16524 133.41114
keto form in the ground state is 6 106:1. This again confirms that there is hardly any possibility of GSIPT at normal temperature in 1H2NA. The GSIPT curve for the keto (T) form is almost flat, which implies that the proton transfer fluorescence may not show a broad structureless band. Interestingly, in their experimental work, Tobita et al. [55] observed a structured and somewhat less broad emission band of 1H2NA in cyclohexane having lem 476 nm. They also 100
Energy (kcal./mol)
95 90 85
S2
80
S1
14 12 10 8 6 4 2 0
S0
0.9
1.0
1.1
1.2
1.3
1.4
1.5
1.6
1.7
1. 8
0
r(O -H)(A ) d
Figure 26.7 GSIPT curve (S0) and ESIPT Franck–Condon curves of 1H2NA, as obtained from DFT-B3LYP(6-31G) and TDDFT-B3LYP(6-31G) levels of theory. Reprinted with permission from [65]. Copyright Elsevier
Excited-State Intramolecular Proton Transfer Processes 599
Table 26.7 Values of relative energy (in kcal mol1) at various OdH distances (in A) of 1H2NA, as obtained using DFT-B3LYP(6-31G) level of theory
Relative energy (kcal mol1)
rOH (A)
0.50 0.75 0.80 0.85 0.90 0.95 1.00 1.05 1.10 1.15 1.20 1.25 1.30 1.40 1.50 1.70 1.75
S0 state
S1 state
S2 state
506.51003 53.38376 29.33885 14.28628 5.58148 1.29799 0.00000 0.55610 2.11823 3.83657 5.16336 5.93043 6.25985 6.27911 6.15437 6.69000 7.01953
589.91244 135.56460 111.17397 95.69962 86.47393 81.53817 79.40583 78.94320 79.28839 79.83127 80.29374 80.57219 80.67804 80.44608 80.26372 80.81779 81.17037
590.29273 137.91782 114.00660 98.75121 89.86663 85.34574 83.71355 83.83173 84.82227 85.87451 86.65506 86.76064 86.51616 85.68725 84.96556 84.53086 84.68983
observed that the red-shifted absorption band of 1H2NA is p–p in nature, with lmax ¼ 368 nm in cyclohexane. On the other hand, our gas-phase calculation shows lmax at 360 nm for the enol (N) form of 1H2NA. Thus, our calculated lmax is in good agreement with the experimental findings of Tobita et al. [55]. 26.3.2 Ground- and excited-state potential energy curve of 2H3NA For the calculation of the ground-state potential energy curve, we used the same technique as before, i.e the ‘distinguished coordinate’ approach as proposed by Sobolewski et al. [63], and obtained a minimum in the PE curve at an rOH distance of about 1 A, which is due to the N form of 2H3NA. Surprisingly, there is no shallow minimum for the T form; rather, the ground-state potential energy curve (Figure 26.8) increases steadily as the rOH distance increases from 1 to 2 A. The corresponding data are given in Table 26.8. The FC potential energy curve for the S1 state shows two minima, one at rOH 1 A and the other, which is much lower in energy, at rOH 1.5 A. The former minimum is due to the excited enol form (N ), and the latter minimum is due to the excited keto tautomer (T ). The repulsive nature of the GSIPT curve and the energy gap between the S0 and S1 curves at the keto tautomer position (rOH 1.5 A) implies that the keto tautomer emission will be broad, structureless and largely red-shifted. In their laser-induced fluorescence measurements, Wu et al. [56] showed a large Stokes-shifted, extremely weak emission band of 2H3NA maximized at 715 nm for the keto tautomer resulting from ESIPT. We believe that the nanosecond UV pulse excites the normal species to a vibronic S1 state via a Franck–Condon transition, where the nuclear coordinates remain unchanged. The rapid charge distribution after excitation results in an electronic potential surface possessing a substantial gradient and an energy minimum shifted towards the proton position in the keto tautomer. The potential energy surface of the S1 state shows that some of the normal vibrational modes are displaced from the equilibrium position. As a result, the system begins to evolve along these normal coordinates on the excited-state energy surface towards a new equilibrium position. This motion may be conceived as
600
Hydrogen Bonding and Transfer in the Excited State
Energy (kcal./mol)
90 S2
80
70
S1
20
S0
10
0 0.9 1.0 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9 2.0 0
r(O -H)(A ) d
Figure 26.8 GSIPT curve (S0) and ESIPT Franck–Condon curves of 2H3NA as obtained from DFT-B3LYP(6-31G) and TDDFT-B3LYP(6-31G) levels of theory. Reprinted with permission from [65]. Copyright Elsevier
the propagation of a wave packet made up of the superposition of wave functions of various vibronic states that are involved in the temporal development. In particular, those vibrational coordinates coupled with the proton displacement should deviate significantly from the equilibrium position, which is eventually reached by the formation of the excited keto tautomer.
Table 26.8 Values of relative energy (in kcal mol1) at various OdH distances (in A) of 2H3NA, as obtained using DFT-B3LYP(6-31G) level of theory
Relative energy (kcal mol1)
rOH (A)
0.50 0.75 0.80 0.85 0.90 0.95 1.00 1.05 1.10 1.15 1.20 1.25 1.30 1.35 1.40 1.50 1.70 1.75 2.00
S0 state
S1 state
S2 state
505.29500 52.38250 28.44140 13.52560 5.00210 0.96190 0.00000 1.06360 3.32270 6.08320 8.73930 10.86930 12.38530 13.43310 14.18480 15.28630 17.38900 18.00060 21.58020
581.90969 126.57021 101.97224 86.30045 76.87577 71.73155 69.37985 68.66643 68.63453 68.53705 68.04706 67.44354 67.04423 66.81515 66.70254 66.73230 67.84623 68.32646 71.57647
586.08373 132.33688 108.19295 93.05359 84.27425 80.10498 78.75818 86.05172 80.99530 82.96755 85.05206 86.95388 87.28750 85.66171 84.36672 82.58719 81.26722 81.23347 82.32847
Excited-State Intramolecular Proton Transfer Processes 601
The repulsive nature of the ground-state PE curve outright ruled out the possibility of ground-state intramolecular proton transfer in 2H3NA. Again, our free energy calculation for the ground-state enol–keto equilibria of 2H3NA gives a positive free energy change (DG ¼ 29.70 kcal mol1), and the calculated equilibrium constant is 1.40 1022. On the basis of the equilibrium constant, the population ratio in the gas phase for the enol vs. keto form in the ground state is 7 1021:1. This clearly explains that the GSIPT is thermodynamically unfavourable. Experimental investigations of Wu et al. [56] shows that lmax of 2H3NA in cyclohexane is nearly 390 nm. On the other hand, our calculated value of lmax in the gas phase is about 404 nm, and this is in good agreement with their experimental findings. 26.3.3 Comparison of IMHB and ESIPT of 1H2NA and 2H3NA Both compounds 1H2NA and 2H3NA contain a naphthalene ring and two other chromophore OH and CHO. However, they differ only in the relative position of the OH and CHO groups. Our calculation suggests that, for both these molecules, the N form is the most stable in their ground state. The strength of IMHB is nearly 5 kcal mol1 greater in 1H2NA than in 2H3NA. As far as the ring skeleton of the N form is concerned, 1H2NA and 2H3NA resemble phenanthrene and anthracene respectively. We then decided to compare the energy of phenanthrene and anthracene using the same quantum mechanical method (DFT/ B3LYP with 631G basis), and, to our complete surprise, we found that, energetically, phenanthrene is nearly 5 kcal mol1 more stable than anthracene. As significant characteristic differences are not observed between the S2 states of 1H2NA and 2H3NA, this state is not shown in the potential energy surfaces, although the corresponding values are given in Tables 26.7 and 26.8. Figure 26.7 shows that the excited singlet state (i.e 1(pp )) potential energy curve of 1H2NA has a minimum at the equilibrium distance of the N tautomer and a wide minimum near the keto tautomer position. Figure 26.7 also shows that the barrier of the N tautomer and T tautomer is very small. If the proton transfer process or any other non-radiative deactivation channel is quite fast compared with the lifetime of the excited N tautomer, as observed by Tobita et al. [55] for 1H2NA, it will be very difficult to observe emission from the N tautomer. On the other hand, owing to the faster formation rate, the population of the T tautomer will be high enough to observe emission. Whereas the 1(pp ) potential energy curve of 2H3NA exhibits two minima and an exothermal behaviour of the potential energy curve of the T tautomer explains the occurrence of proton transfer fluorescence. Figures 26.9(a), (b) and (c) show the PES with simultaneous variation in OdH and OdOa distances in the ground (S0) state, the first excited singlet (S1) state and the second excited singlet (S2) state, respectively, of compound 1H2NA. The contour levels are given in relative energies and expressed in kcal mol1. In the S0 state (Figure 26.9(a)), the presence of a single valley corresponding to the enol (N) form indicates that in the ground state the enol form is the most stable one. By contrast, in the S1 (Figure 26.9(b)) and S2 (Figure 26.9(c)) states, two valleys are observed, corresponding to the excited enol (N ) and excited keto (T ) tautomers. Figures 26.9(b) and (c) also illustrate a narrow potential well for N and a much wider and deeper well for T, thus indicating a greater possibility of its existence over the N form in the S1 as well as in the S2 excited state. As S1 is the lowest excited state, proton transfer emission will come from the S1 (according to Kasha’s rule) state only. An almost identical contour curve is obtained in the case of 2H3NA, as shown in Figures 26.10(a), (b) and (c), which represent the S0, S1 and S2 states respectively. In Figures 26.11(a) and (b), the oscillator strength f for the S0–S1 transition increases with increase in OH bond distance in the case of both 1H2NA and 2H3NA. It indicates that the red-shifted emission can be substantial. The corresponding data are given in Table 26.9. A detailed analysis of the electron density of HOMO and LUMO of these two compounds can throw some light on ground- and excited-state electron transfer processes. Both HOMO and LUMO are of the p type, but their phases are quite different in 1H2NA and 2H3NA. The HOMO orbital on the IMHB ring of 1H2NA
602
Hydrogen Bonding and Transfer in the Excited State
(a)
1.8
1.8
(b) 2.632
72.22
1.6
1.6
-10.53 1.4
1.4
61.11
2.632
15.79 42.11
0.8
68.42 81.58 94.74
55.26
94.44
d
1.0
0.4
0
1.2
28.95
r(O -H) / A
d
r(O -H) / A0
1.2
0.6
72.22
83.33
107.9 121.1 160.5 226.3 265.8 134.2 147.4 278.9 239.5 292.1 357.9 186.8 213.2 305.3 384.2 331.6 173.7 252.6 200.0 2.40
2.45
2.50
2.55
1.0
0.8
0.6
0.4
105.6 127.8 138.9 116.7 161.1 150.0 172.2 194.4 183.3 238.9 283.3 316.7 205.6 216.7 305.6 350.0 416.7 338.9 261.1 294.4 227.8 383.3 461.1 250.0 272.2 327.8 361.1 427.8 2.40
2.60
2.45
0
2.60
r(O -O )/A
a
(c)
2.55 0
r(O -O )/A d
2.50 d
a
1.8
1.6
83.33
d
r(O -H) / A0
1.4
72.22
1.2
105.6 94.44 1.0
172.2
0.8
0.6
0.4
127.8 116.7 150.0 138.9 161.1 183.3 194.4
238.9
205.6 283.3
327.8 316.7 294.4 216.7 227.8 350.0 427.8 272.2 305.6 338.9 383.3 261.1 461.1 250.0 372.2 416.7 2.40
2.45
2.50
2.55
2.60
0
r(O -O )/A d
a
Figure 26.9 2D PES for the (a) ground (S0) state, (b) the first excited singlet (S1) state and (c) the second excited singlet (S2) state of 1H2NA, plotted as a function of OdH and OdOa distances. The contour levels are the relative energies of the respective states (in kcal mol1)
(Figure 26.12) is primarily of bonding character over the C(30 )C(40 )C(50 ) atoms, whereas C(30 )O(20 ) and C(50 )O(60 ) show antibonding character. Both the hydroxyl oxygen and aldehyde oxygen have bonding character, with a larger electron density over the hydroxyl oxygen. Analysis of the HOMO electron density after ground-state electron transfer (Figure 26.13) still shows a larger density on the hydroxyl oxygen and also a shift of electron density over the C(40 )C(50 ) bond. The HOMO electron density around the IMHB ring (Figure 26.14) of the N tautomer of 2H3NA shows a node at C(40 ). Again, there is no electron density contribution over the aldehyde group of 2H3NA. The HOMO electron density of the ground-state proton
Excited-State Intramolecular Proton Transfer Processes 603 (a)
(b)
2.0
1.8
18.75
1.8
1.6
1.6
5.000
r(O -H) / A0
1.4
1.2
0.8 0.6 0.4 2.40
87.50
101.3
142.5 156.3
73.75 60.00 128.8 115.0
121.4
2.50
2.55
1.0
92.86 78.57
64.29
135.7
107.1
178.6 192.9 0.8
170.0 197.5 238.8 183.8 211.3 293.8 266.3 321.3 225.0 307.5 376.3 335.0 252.5 280.0 2.45
1.2
d
46.25 32.50
d
r(O -H) / A0
1.4
1.0
2.0
2.60
2.65
2.75
0.4 2.40
2.80
207.1
164.3
235.7 250.0 278.6 321.4 364.3 264.3 292.9 335.7 407.1 378.6 450.0 307.1 392.9 350.0 221.4
0.6
2.70
150.0
2.45
2.50
2.55
0
d
2.60
2.65
2.70
2.75
2.80
0
r(O -O )/A
r(O -O )/A
a
d
a
2.0
(c)
1.8
78.57
1.6
d
r(O -H) / A0
1.4 1.2
135.7
150.0
1.0
192.9 0.8 0.6 0.4 2.40
78.57 107.1 92.86 121.4 164.3
207.1 221.4
178.6
250.0 264.3 292.9 335.7 278.6 307.1 378.6 350.0 421.4 321.4 392.9 407.1 464.3 364.3 235.7
2.45
2.50
2.55
2.60
2.65
2.70
2.75
2.80
0
r(O -O )/A d
a
Figure 26.10 2D PES for the (a) ground (S0) state, (b) the first excited singlet (S1) state and (c) the second excited singlet (S2) state of 2H3NA, plotted as a function of the OdH and OdOa distances. The contour levels are the relative energies of the respective states (in kcal mol1)
transfer form of 2H3NA (Figure 26.15) shows less electron density on C(40 )C(50 ) and aldehyde oxygen than that of 1H2NA. Thus, the lower conjugation through the IMHB ring in 2H3NA weakens its IMHB strength. Again, the less effective electron transfer along the proton transfer coordinate makes the GSIPT process less probable. The opposite effect was found for 1H2NA, thereby strengthening the IMHB ring system and stabilizing the GSIPT potential energy curves with respect to 2H3NA. Nevertheless, this stabilization is not sufficient to produce a GSIPT process. For both 1H2NA (Figure 26.12) and 2H3NA (Figure 26.14), LUMO is p in nature. If we look into the electronic charge distribution of LUMO within the IMHB ring (N tautomer) for both 1H2NA and 2H3NA, the C(40 )C(50 ) position has a bonding character, whereas the C(30 )(20 ), C(30 )C(40 ) and C(50 )O(60 ) positions have an antibonding character. The LUMO of the enol form of 2H3NA possesses a high electron density on the O(60 )
604
Hydrogen Bonding and Transfer in the Excited State 0.105
(a)
Oscillator strength(f)
0.100
0.095
0.090
0.085
0.080
0.4
0.6
0.8
1.0
1.2 (Od -H)
0.048
1.4
1.6
1.8
0
r
(A )
(b)
Oscillator strength(f)
0.046 0.044 0.042 0.040 0.038 0.036 0.034 0.032 0.4
0.6
0.8
1.0
1.2
r
(Od -H)
1.4
1.6
1.8
2.0
2.2
0
(A )
Figure 26.11 Variation in the oscillator strength with OdH distance in the S0–S1 transition of 1H2NA (a) and 2H3NA (b) computed using DFT (B3LYP)/6-31G level of theory
atom, and there is no electron distribution on the O(20 ) atom. On the other hand, the LUMO of 1H2NA possesses high electron density at O(60 ) and comparatively lower electron density at O(20 ) than that of the corresponding HOMO. After tautomerization, the LUMO of the keto tautomer of 2H3NA still shows high electron density on the O(60 ) atom and much lower electron density on the O(20 ) atom. Thus, it favours the
Excited-State Intramolecular Proton Transfer Processes 605
Table 26.9 Values of oscillator strength f at various OdH distances (in A) of 1H2NA and 2H3NA, as obtained using DFT-B3LYP(6-31G) level of theory
rOH (A)
0.50 0.75 0.80 0.85 0.90 0.95 1.00 1.05 1.10 1.15 1.20 1.25 1.30 1.40 1.50 1.70 1.75 2.00
Oscillator strength f 1H2NA
2H3NA
0.0797 0.0818 0.0823 0.0828 0.0833 0.0841 0.0849 0.0858 0.0873 0.0894 0.0921 0.0941 0.0949 0.0979 0.0992 0.1006 0.1009
0.0353 0.0349 0.0347 0.0346 0.0344 0.0341 0.0338 0.0335 0.0332 0.0331 0.0336 0.0348 0.0362 0.0389 0.0409 0.0438 0.0444 0.0471
Figure 26.12 HOMO and LUMO orbitals of 1H2NA (enol form), as obtained with DFT-B3LYP(6-31G). Reprinted with permission from [65]. Copyright Elsevier
transfer of a proton from O(20 ) to O(60 ). However, in the case of the keto tautomer of 1H2NA, O(20 ) and O(60 ) have comparable electron density with that of the HOMO. So, the ESIPT process in 1H2NA is less favourable compared with that in 2H3NA. Our PES calculations along the proton transfer coordinate also suggest the exothermal behaviour of ESIPT processes. Our PES calculations of 1H2NA suggest that the N and T forms have comparable energy in the first excited singlet state. We believe that, owing to the faster rate of formation of the excited T form and also the presence of other faster non-radiative deactivation channels from the excited singlet form of the N tautomer, it is very difficult to observe normal emission in 1H2NA. This type of extensive study on the intramolecular proton transfer process has been conducted by the present authors – see Refs [65] to [67].
606 Hydrogen Bonding and Transfer in the Excited State
Figure 26.13 HOMO and LUMO orbitals of 1H2NA (keto form), as obtained with DFT-B3LYP(6-31G). Reprinted with permission from [65]. Copyright Elsevier
Figure 26.14 HOMO and LUMO orbitals of 2H3NA (enol form), as obtained with DFT-B3LYP(6-31G). Reprinted with permission from [65]. Copyright Elsevier
Figure 26.15 HOMO and LUMO orbitals of 2H3NA (keto form), as obtained with DFT-B3LYP(6-31G). Reprinted with permission from [65]. Copyright Elsevier
26.4 Conclusions The relative position of the OH and CHO groups in 1H2NA and 2H3NA determine the strength of the intramolecular hydrogen bond and the nature of ESIPT emission. In the ground state, the strength of IMHB in 1H2NA is greater than in 2H3NA. Free energy calculation, HOMO electron density, ground-state PE calculation and also contour diagrams for both 1H2NA and 2H3NA support the non-viability of GSIPT processes. On the other hand, excited-state potential energy, LUMO electron density and oscillator strength calculations support the ESIPT for both 1H2NA and 2H3NA. The nature of the ground- and excited-state
Excited-State Intramolecular Proton Transfer Processes 607
potential energy curves nicely explains the red-shifted broad structureless emission band of 2H3NA, and also the less broad and less Stokes-shifted ESIPT band of 1H2NA. Analysis of the LUMO electron density suggests that ESIPT is more favourable in 2H3NA. As all the calculations have been carried out by the DFT method with hybrid functionals (B3LYP/6-31G), they again support the potential of the DFT method for calculations of ESIPT processes.
Acknowledgements We gratefully acknowledge the financial support received from DSTand UGC, New Delhi, for carrying out this research work.
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.
P. T. Chou, D. McMorrow, T. J. Aartsma and M. Kasha, J. Phys. Chem., 88, 4596 (1984). J. Catalan, J. L. Paz, J. C. D. Valle and M. Kasha, J. Phys. Chem. A, 198, 5284 (1997). J. Keck, M. Roessler, C. Schroeder et al., J. Phys. Chem. A, 102, 6975 (1998). M. L. Martinej, W. C. Cooper and T. T. Chou, Chem. Phys. Lett., 193, 151 (1992). R. W. Munn, Chem. Br., 517 (1989). A. Sytnik and M. Kasha, Proc. Natl Acad. Sci. USA, 80, 8627 (1980). J. Catalan, J. C. Valle, J. Palomar et al., J. Phys. Chem. A, 103, 10 921 (1999). T. Nishiya, S. Yamauchi, N. Hirota et al., J. Am. Chem. Soc., 108, 3880 (1986). C. Lu, R. M. Hsieh, L. R. Lee and P. Y. Cheng, Chem. Phys. Lett., 310, 103 (1999). J. A. Organero, M. Moreno, L. Santos et al., Chem. Phys. Lett., 328, 83 (2000). J. A. Organero, M. Moreno, L. Santos et al., J. Phys. Chem. A, 104, 8424 (2000). S. Maheshwari, A. Chowdhury, N. Sathyamurthy et al., J. Phys. Chem. A, 103, 6257 (1999). J. Catalan, J. Palomar and J. L. G. Paz, J. Phys. Chem. A, 101, 7914 (1997). S. Nagaoka, Y. Shinde, K. Mukal and U. Nagashima, J. Phys. Chem. A, 101, 3061 (1997). E. Orton, M. A. Morgan and G. C. Pimentel, J. Phys. Chem., 94, 7936 (1990). P. T. Chou, M. L. Martinez and S. L. Studer, J. Phys. Chem., 95, 10 306 (1991). V. Vargas and L. Amigo, J. Chem. Soc., Perkin Trans II, 1124 (2001). L. Antonov, W. M. F. Fabian, D. Nedltcheva and F. S. Kamounah, J. Chem. Soc., Perkin Trans II, 1173 (2000). S. H. Alarcon, A. C. Olivieri, D. Sanz et al., J. Mol. Struct., 705, 1 (2004). I. K. Starzomska, A. Filarowski, M. Rospenk et al., J. Phys. Chem. A, 108, 2131 (2004). A. Weller, Z. Elektrochem., 60, 1144 (1956). A. Weller, Prog. React. Kinet., 1, 187 (1961). F. Lahmani and A. Zehnacker-Rentien, J. Phys. Chem. A, 101, 6141 (1997). D. Gormin, A. Sytnik and M. Kasha, J. Phys. Chem. A, 101, 672 (1997). P. F. McGarry, S. Jockusch, Y. Fujiwara et al., J. Phys. Chem. A, 101, 764 (1977). V. Guallar, M. Moreno, J. M. Lluch et al., J. Phys. Chem., 100, 19 789 (1996). J. Keck, H. E. A. Kramer, H. Port et al., J. Phys. Chem., 100, 14 468 (1996). F. Parsapour and D. F. Kelley, J. Phys. Chem., 100, 2791 (1996). R. M. Tarkka and S. A. Jenekhe, Chem. Phys. Lett., 260, 533 (1996). H. Zhang, P. van der Meulen and M. Glasbeek, Chem. Phys. Lett., 253, 97 (1996). S. Mitra, R. Das, S. P. Bhattacharyya and S. Mukherjee, J. Phys. Chem. A, 101, 293 (1997). P. G. Yi, Y. H. Liang and C. Z. Cao, Chem. Phys., 315, 297 (2005). S. Nagaoka and U. Nagashima, J. Phys. Chem., 94, 1425 (1990). S. Nagaoka and U. Nagashima, Chem. Phys., 136, 153 (1989). M. V. Verner and S. Scheiner, J. Phys. Chem., 99, 642 (1995).
608 Hydrogen Bonding and Transfer in the Excited State 36. 37. 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.
A. L. Sobolewski and W. Domcke, Chem. Phys., 184, 115 (1994). J. Catalan, J. Palomar and J. L. G. De Paz, J. Phys. Chem. A, 101, 7914 (1997). K.-Y. Law and J. Shoham, J. Phys. Chem., 98, 3114 (1994). K.-Y. Law and J. Shoham, J. Phys. Chem., 99, 12 103 (1995). A. U. Acuna, F. Toribio, F. Amat-Guerri and J. Catalan, J. Photochem., 30, 339 (1985). F. Toribio, J. Catalan, F. Amat and A. U. Acuna, J. Phys. Chem., 87, 817 (1983). A. U. Acuna, F. Amat-Guerri, J. Catalan and F. Gonzalez-Tablas, J. Phys. Chem., 84, 629 (1980). J. Goodman and L. E. Brus, J. Am. Chem. Soc., 100, 7472 (1978). K. K. Smith and K. J. Kaufmann, J. Phys. Chem., 82, 2286 (1978). S. Nagaoka, N. Hirota, M. Sumitani and K. Yoshihara, J. Am. Chem. Soc., 105, 4220 (1983). T. Nishiya, S. Yamauchi, N. Hirota et al., J. Phys. Chem., 90, 5730 (1986). L. A. Peteanu and R. A. Mathies, J. Phys. Chem., 96, 6910 (1992). M. A. Morgan, E. Orton and G. C. Pimentel, J. Phys. Chem., 94, 7927 (1990). S. Nagaoka, N. Hirota, M. Sumitani et al., J. Am. Chem. Soc., 106, 6913 (1984). S. Nagaoka, U. Nagashima, N. Ohta et al., J. Phys. Chem., 92, 166 (1988). J. Catalan, F. Toribio and A. U. Acuna, J. Phys. Chem., 86, 303 (1982). E. D. Bergmann, Y. Hirshberg and S. Pinchas, J. Chem. Soc., 2351 (1950). U. V. Naboikin, B. A. Zadorozhnyi and E. N. Pavlova, Opt. Spectr.(Eng. Transl.), 6, 312 (1959). J. Catalan, J. C. Delvalle, J. Palomar et al., J. Phys. Chem. A, 103, 10 921 (1999). S. Tobita, M. Yamamoto, N. Kurahayashi et al., J. Phys. Chem. A, 102, 5206 (1998). K. C. Wu, Y. M. Cheng, Y. S. Lin et al., Chem. Phys. Lett., 384, 203 (2004). A. Bouchy, D. Rinaldi and J.-L. Rivali, Int. J. Quan. Chem., 96, 273 (2004). A. Dreuw and M. H. Gordon, Chem. Rev., 105, 4009 (2005). A. D. Becke, J. Chem. Phys., 98, 5648 (1993); C. Lee, W. Yang and R. G. Parr, Phys. Rev. B, 37, 785 (1988). V. Barone and C. Adamo, J. Phys. Chem., 99, 15 062 (1995). M. J. Frisch, G. W. Trucks, H. B. Schlegel et al., Gaussian 03, Revision C.02. Gaussian, Inc., Wallingford, CT (2004). I. M. Hunsberger, J. Am. Chem. Soc., 72, 5626 (1950). A. L. Sobolewski and W. Domcke, Chem. Phys., 232, 257 (1998). S. Maheshwari, A. Chowdhury, N. Sathyamurthy et al., J. Phys. Chem. A, 103, 6257 (1999). S. P. De, S. Ash, S. Dalai and A. Misra, J. Mol. Struct. (Theochem), 807, 33 (2007). S. P. De, S. Ash, H. Bar et al., J. Mol. Struct. (Theochem), 824, 8 (2007). S. P. De, S. Ash, H. K. Bar et al., Spectrochim. Acta Part A: Mol. Biomol. Spectr., 71, 1728 (2009).
27 Conformational Switching Between Acids and Their Anions by Hydrogen Bonding Taka-aki Okamura, Hitoshi Yamamoto and Norikazu Ueyama Department of Macromolecular Science, Graduate School of Science, Osaka University, Toyonaka, Osaka 560-0043, Japan
27.1 Introduction In general, pKa shifts for acids, such as thiols, carboxylic acids, phenols and phosphoric acid monoanions, are considered to occur under homogeneous conditions and change with dielectric constant. Theoretical calculations have been performed for homogeneous environments [1–4], and some improved theoretical studies have considered anisotropic and heterogeneous environments [4–8] to deal with the particularly large shift for thiols, phenols and carboxylic acids located inside proteins. This paper discusses pKa shifts for acids having neighbouring amide NHs. A relationship between the pKa shift and the formation constant is observed for metal complexes in hydrophobic environments. The influence of intramolecular NH X hydrogen bonds on the metal–X bond character is also observed for metal complexes. The presence of NH S hydrogen bonds in iron–sulfur proteins observed by crystallographic studies has been suggested by Adman and coworkers [9]. Various synthetic metal–thiolate complexes having chelating Cys-containing oligopeptides and synthetic simple intramolecularly NH S hydrogen-bonded thiolate ligands exist. NH S hydrogen bonds in metal–thiolate complexes play an important role in the following functions. Intramolecular hydrogen bonds contribute to the following: increase in the stabilization constant, e.g. Fe(II), Zn(II), Cd(II), Hg(II) and Ca(II) complexes; positive shift of the redox potential, e.g. Fe(II), Fe(III), Mo(IV) and W(IV) complexes; and increase in air stabilization, e.g. Fe(II) and Fe(III) complexes. In addition, the NH S hydrogen bond increases the formation constant for metal–sulfur complexes formed by substitution reaction between non-hydrogen-bonded and hydrogen-bonded thiolates [10]. The large formation constant is due not only to the pKa-lowering shift of thiol but also to a pp–dp interaction by the NH S hydrogen bond. Hydrogen Bonding and Transfer in the Excited State, Volume I & II © 2011 John Wiley & Sons, Ltd. ISBN: 978-0-470-66677-7
Edited by Ke-Li Han and Guang-Jiu Zhao
610 Hydrogen Bonding and Transfer in the Excited State
27.2 pKa Shift of Acids by Neighbouring Amide NH 27.2.1 pKa shift of thiol, phenol and carboxylic acid derivatives by prelocated hydrogen bond Thiol, phenol, carboxylic acid and phosphate monoanions exist as relatively weak acids under hydrophobic and neutral conditions, although their deprotonated anion states are strong anions that coordinate to metal ions. When the anion has an amide NH in the neighbourhood, it forms a strong NH X (X¼S, O) hydrogen bond (Figure 27.1). The extent of deprotonation in these acids reflects the lowering of pKa by the prelocated amide NHs. Crystallographic analyses of thiol (RSH) and its thiolate anion (RS) in the solid state, and 1H NMR and FT-IR analyses in solution, indicate that the S atom of the SH group in an amidated thiophenol derivative does not interact with the neighbouring amide NH in the ground state (Figure 27.2(a)), but the thiolate anion does this strongly. If the thiophenolate has an amide NH at the p-position, the thiolate anion strongly interacts with other thiolate molecules, intermolecularly (Figure 27.2(b)). Therefore, when a solvent has an amide NH, the amide NH in the solvent strongly interacts with the thiolate anion. Similarly, the S atom in the disulfide does not interact with the neighbouring amide NH in bis[2,6-di(pivaloylamino)] phenyl disulfide (Figure 27.2(c)), as shown by crystallographic analysis in the solid state and 1H NMR and IR spectroscopic analyses in solution [11].
R''
R
H
S
H
R
R'
O
N
O
R''
R'' S
S
H
H
R
R'
O
N
+ H+
H N R'
Figure 27.1 Deprotonation of RSH assisted by prelocated amide NH
(a)
H
(c) S
H
-H
N O
H
S
+
R
R
O
+ H+
R
R
N
R R
O
O
H N N H
(b)
H
S
O
S
- H+
S
H
S
N
H N N H
H N
O
O
O
+ H+ S H N
O
R R R
R
R
R
No NH---S interaction (R = H, Me, Ph)
Figure 27.2 Deprotonation processes: (a) 2-t-BuCONH-C6H4SH; (b) 4-t-BuCONH-C6H4SH. After deprotonation, the thiolate anion intermolecularly interacts with amide NH. (c) Disulfide structure determined by crystallographic analysis. No interaction is observed between amide NH and sulfur
Conformational Switching Between Acids and Their Anions by Hydrogen Bonding
611
The pKa shifts by prelocated amide NH in thiols were examined in aqueous micellar solution using bulky thiol. The shift for RSH is known to occur because of the electron-withdrawing character of the R group. Electronic effects for the amide group on the phenyl ring are negligible by Hammet’s parameter (sp ¼ 0). The shift for the N-methylated derivative, which cannot form hydrogen bonds, is slight, while that for 2,6-(t-BuCONH)2C6H3SH with double amide NHs is large (pKa ¼ 4.4) (Table 27.1). Intermolecular NH S hydrogen bond formation is detected for the anion form of 4-t-BuCONHC6H4SH, which has a relatively high
Table 27.1 pKa values for various thiols in Triton X-100 aqueous micellar solution at room temperature Thiol
pKa 8.0
5.4
5.7
6.1
4.9
4.4
4.9
9.6 8.8(Cys1), 10.0 9.3(Cys1), 9.8(Cys2)
612 Hydrogen Bonding and Transfer in the Excited State
pKa (7.6) compared with that for 2-BuCONHC6H4SH. It turns out that the high value is due to a lack of interaction with the remote amide NH even at the transition state because of the lower encounter probability in intermolecular collisions between thiolate and amide NH. The pKa value for Cys-containing oligopeptides in aqueous micellar solution depends on the location of the Cys residue in the peptide chain [12]. For a short Cys-containing tripeptide, pKa is 9.6 in aqueous micellar solution. This peptide in the thiolate state forms a strong NH S hydrogen bond, detected by the IR and 1 H NMR shifts of amide NH not only in chloroform but also in aqueous micellar solution. For Z-Cys(1)-ProLeu-Cys(2)-OMe, the values for Cys(1) and Cys(2) are pKa ¼ 8.8 and 10.0 respectively. For Z-Cys(1)-Ala-ProCys(2)-OMe, the values for Cys(1) and Cys(2) are high, pKa ¼ 9.3 and 9.8 respectively. The solution structures determined by 1H NMR indicate that the former tetrapeptide has a hairpin turn structure, which readily forms an NH S hydrogen bond after easy deprotonation from Cys(1) thiol, whereas the latter tetrapeptide has an extended structure that yields similar pKa values for Cys(1) and Cys(2) after deprotonation. The pKa shift with the neighbouring amide NH is due to the lowering of the energy barrier for the transition state during deprotonation. In the ground state the thiol is not affected by the neighbouring amide NH, whereas in the transition state it interacts with the adjacent amide NH. The pKa shift corresponds to the extent of energy lowering for deprotonation. The pKa values for phenol [13–15] and carboxylic acid derivatives [16] show a similar tendency to the prelocated amide NH in aqueous micellar solution. Singly and doubly hydrogen-bonded phenol and benzoic acid derivatives exhibit a clear pKa shift. For example, although phenolic OH oxygen is known to have a weak negative charge, the O atom of the OH group does not interact with the neighbouring amide NH, even being closely located to the O atom in the ground state. The NH group begins to interact with the O atom of the phenol under the transition state for deprotonation (Figure 27.3(a)). This effect contributes to the pKa shift through deprotonation [13]. Thus, only the prelocated neighbouring amide NH affects the pKa shift. When the amide NH is in the p-position of the phenol, the pKa does not shift. Of course, solvent having an amide NH also does not significantly affect the pKa shift. The intramolecularly prelocated amide NH increases the probability that the amide NH will be near the OH group of phenol.
(a)
H
(b) H
O
O
O
N H O
R
O N H
pKa lowering
R conformational change
O H
O
H
O
H R
N O H
O
H N
R O
prelocated amide NH
H O
O
H N
R O
H O
R
N H non-prelocated amide NH
O
N
R O
Figure 27.3 Deprotonation reaction coordinates: (a) for phenol and 2-acylaminophenol with prelocated amide NH; (b) for 2-carbamoylphenol, accompanied with conformational change
Conformational Switching Between Acids and Their Anions by Hydrogen Bonding
613
In contrast, the carbamoyl derivative forms an intramolecular hydrogen bond between the phenolic OH and amide CO in the ground state (Figure 27.3(b)). The amide NH does not prelocate near the O atom of phenolic OH. Therefore, this phenol has a large pKa. When deprotonation occurs, the phenolate anion obtained is unstable without the NH O hydrogen bond and becomes an energetically stable conformer with amide NH by rotation of the amide plane. However, in biological systems, it has been proposed that the pKa also decreases when the SH group undergoes RSH base interaction [17].
27.2.2 Proton-driven conformational switching in asp- oligopeptide and model complexes We mentioned previously that a prelocated amide NH decreases the pKa value for weak acids, such as thiols, phenols, carboxylic acids, and phosphoric acid monoanions, and destabilizes their anions. In contrast, a remote amide NH around thiol and phenol does not cause prompt deprotonation of their acids (Table 27.1; Figure 27.2), although the anion form interacts intermolecularly with amide NHs in other molecules. If the conformationally stabilized structure of the anion differs from that of thiols, phenol or carboxylic acids, preferable conformational switching occurs intramolecularly between conformers of the acid and the anion such as 2-acylaminophenol (Figure 27.3(b)). Such intramolecular conformational change is relevant to the construction of switching devices. Various conformationally restricted molecules have been synthesized using oligopeptides, Kemp’s acids, salicylamide and maleic acid derivatives. As a simple conformationally switching compound, doubly amidated Kemp’s acid derivative r-1,c-3,c-5(CH3)3-3,5-(Ph2CHNHCO)2C6H6-1-COOH has been synthesized [18]. A chair form of the carboxylic acid converts to a twist-boat form with deprotonation (Figure 27.4). This transformation is observed over a relatively high energy barrier, approximately 40–80 kJ mol1, in acetonitrile at room temperature. The twistboat form is extremely stabilized by the two NH O hydrogen bonds between the carboxylate anion and the two amide NHs in the solid state and in acetonitrile solution. A proton-driven conformational change in Asp-containing oligopeptides occurs from the carboxylic acid to the carboxylate anion. An Asp-containing tripeptide, AdCO-Asp(COOH)-Val-Gly-NHCH2Ph (Ad ¼ adamantyl), has an inverse g-turn structure in the acid state, whereas the tripeptide carboxylate anion converts to a b-turn-like conformer with intramolecular NH O hydrogen bonds in chloroform or aqueous micellar solution (Figure 27.5) [19]. These structures were determined by 1H NMR. Any conformational change in the Asp-containing dipeptide is not detected between AdCO-Asp(COOH)- Val-NHAd and its anion state, AdCO-Asp(COO)-Val-NHAd. The model peptide study suggests that the hydrogen bond stabilizes
H H
N O
H
H O
H N
H
O H
O
- H+ +
+H
Carboxylic acid
O N
NH O H
O
O
Carboxylate anion
Figure 27.4 Proton-driven conformational change between an amidated Kemp’s acid, r-1,c-3,c-5-(CH3)3-3,5(Ph2CHNHCO)2C6H6-1-COOH and its anion form
614 Hydrogen Bonding and Transfer in the Excited State (a)
O
O
Asp
O
H
Val
O
H N
H
N H
O
Gly H
N
-H+
O
N O
N H
O H H
O
+H+
H N
N
O
O
Carboxylic acid
N
O
Carboxylate anion
(b)
Val O
O
O
N H
N
H
H Asp
O O
H
O Gly -H+
N
+H+
H
Carboxylic acid
N
O
N H
O
O
N O
H H
O
H
N
N
O
Carboxylate anion
Figure 27.5 Proton-driven conformational changes of AdCO-Asp(COOH)-Val- Gly-NHCH2Ph and its anion form: (a) in chloroform; (b) in aqueous micellar solution
the anion state and thus decreases the basicity of the carboxylate anion, presumably resulting in decreased nucleophilicity. An Asp-containing peptide, benzyloxycarbonyl-Phe-Asp(COO)-Thr-Gly-Ser- Ala-NHCy (Cy ¼ cyclohexyl) anion, has a hairpin-turn structure in acetonitrile. This fragment has been reported to function as a nucleophile in the active centre of proteases, such as pepsin, which contains a –Phe-Asp- Thr-Gly-Ser-Ser– fragment as well as the invariant amino acid fragment –Asp-Thr-Gly– in HIV-1 proteases [20, 21]. Crystallographic analyses of these proteins in the resting state indicate that the invariant Asp-containing peptide fragments of Asp-X-Gly (X ¼ Thr, Ser) form a similar hairpin turn with NH O hydrogen bonds. It is likely that conformational switching of this fragment is related to the formation of a strong nucleophile. Conformational switching is observed in unsymmetrically linked phenolic oligoamides having an intramolecularly NH O hydrogen-bonded structure. Crystallographic analysis indicates that the salicylamide unit changes the conformation around the amide plane (Figure 27.3(b)) [13]. Deprotonation of this molecule causes a linear-to-turn conformational switch (Figure 27.6) [14, 15]. In general, oligophenols have a complicated pKa because an increase in neighbouring phenolate anions prevents the residuary phenols from deprotonating. The decrease in basicity of the anion due to the NH O hydrogen bond allows accumulated phenolate anions to move close to each other. In fact, p-nitrocalix[4]arene is known to have a wide pKa range (2.9–13) because of increasing anionic repulsion [22, 23]. Diarylazomethine carboxylic acid and its carboxylate have an amide group linked to an azomethine moiety, which introduces photoinduced switching supported by the intramolecular NH O hydrogen bond. The ciscarboxylate compound forms a stronger intramolecular NH O hydrogen bond than does the cis-carboxylic acid compound (Figure 27.7(a)) [24]. (E)-3-(2-Pivaloylaminophenol) acrylic acid switches the intramolecular
Conformational Switching Between Acids and Their Anions by Hydrogen Bonding t-Bu
t-Bu H N
t-Bu
O
H
H N
O N H
O
O
615
O
H
N H
O H
H N
O O
O
t-Bu
H
t-Bu
t-Bu
4H+ t-Bu O O t-Bu
N
O
N H
H
t-Bu O H N
O
O H N
t-Bu
H O
t-Bu
N O
O t-Bu
Figure 27.6 Linear-to-turn conformational switching induced by the deprotonation of unsymmetrically linked phenolic tetraamide
distance between amide group and carboxylate O atoms by E/Z photoisomerization of the cinnamate framework (Figure 27.7(b)). An intramolecular NH O hydrogen bond is formed predominantly in the Z-carboxylate form not only in solution but also in the solid state. The pKa value for the carboxylic acid is lowered because of E/Z photoisomerization [25]. ortho-Coumaric acid derivatives with an amide group linked to an olefin moiety also show photoinduced switching accompanied with intramolecular hydrogen bonding [26]. Another type of intramolecular OH O¼C hydrogen bond in the Z-phenol compound switches to an intramolecular NH O hydrogen bond in the Z-phenolate state by deprotonation (Figure 27.7(c)). The pKa value for the Z-phenol derivative is lower than for the E-phenol derivative. Thus, a new photocycle system involving protonation and deprotonation processes has been achieved using carboxylic acid and phenol derivatives with remote amide NHs.
27.3 Coordination of Anion Ligand to Metal Ion 27.3.1 Increase in the stabilization constant by pka shift in metal complexes The lowering of pKa for thiols, phenols, carboxylic acids and phosphoric acid monoanions as precursor acids for a ligand anion is significant for complexation, especially for soft metal ions. A hydrophobic layer prevents ionic ML (M ¼ metal ion; L ¼ ligand anion) interactions in metalloproteins [27]. The conventional
616 Hydrogen Bonding and Transfer in the Excited State
Figure 27.7 Photoisomerizations of (a) diarylazomethines with thermal reversion, (b) cinnamic acid derivatives and (c) ortho-coumaric acid derivatives
complexation formation constant b is described by equation (27.1) under hydrophobic conditions, where K is the equilibrium constant. When the L anion is a strong base, i.e. when LH has a high pKa, a newly proposed formation constant b0 depends on the pKa value for LH, as in equation (27.2): log b ¼ log K
ð27:1Þ
log b0 ¼ log KpKa
ð27:2Þ
Conformational Switching Between Acids and Their Anions by Hydrogen Bonding
617
In general, a high anion basicity is preferable for complexation, although it promotes hydrolysis with water in aqueous solution. The ML bond depends on the pKa value for LH, and the L anion competes with the water molecule. Protection of the ML bond from hydrolysis by water has been demonstrated for NH S hydrogen-bonded ferredoxin model complexes, such as (Et4N)2[Fe4S4(S-2-cholylNHC6H4)4], in aqueous micellar solution. The low pKa value for thiol prevents hydrolysis and air stability owing to the positive shift in redox potential [28]. Similarly, dissociation of the HgS bond was observed by monitoring Hg(0) formation during reduction of Hg(II) complexes having NH S hydrogen-bonded ligands [29]. HgS dissociation is shown to affect the pKa value for thiol but not the covalency of the HgS bond in aqueous micellar solution. The formation of NH S hydrogen bonds can be employed for conventional synthesis of NH S hydrogenbonded metal complexes. For example, reaction (27.3) proceeds quantitatively: ½FeðSPhÞ4 2- þ 4 2-t-BuCONHC6 H4 SH ! ½FeðS-2-t-BuCONHC6 H4 Þ4 2- þ 4 PhSH ½FeðSPhÞ4 2- þ 2 ð2-t-BuCONHC6 H4 SÞ2 ! ½FeðS-2-t-BuCONHC6 H4 Þ4 2- þ 2 PhSSPh
ð27:3Þ
The addition of 2-t-BuCONHC6H4S anion to [Fe(SPh)4]2 results in the formation of mixed-ligand complexes in equilibrium [10]. The complete reaction (27.3) proceeds over the high energy barrier for the deprotonation of thiol and for redox reaction of disulfide. Similarly, for synthesis of Ca, Zn and Cd complexes of carboxylate with intramolecular NH O hydrogen bonds [30–32], the carboxylate ligand with a hydrogen bond promotes ligand-exchange reaction (27.4). Ligand-exchange reaction (27.5) between a phosphate Ca(II) complex and NH O hydrogen-bonded phosphoric acid also proceeds quantitatively [33]: MII ðOCO-2; 4; 6-Me3 C6 H2 Þn ðOH2 Þ4-n þ 2; 6-ðt-BuCONHÞ2 C6 H3 COOH ! MII ðOCOC6 H3 2; 6-ðt-BuCONHÞ2 Þn ðOH2 Þ4-n þ 2; 4; 6-Me3 C6 H2 COOH
ð27:4Þ
(where M ¼ Ca(II), Zn(II), Cd(II)) CaII ðO2 PðOHÞO-2; 6-i-Pr2 C6 H3 Þn ðOH2 Þ4-n þ 2; 6-ðt-BuCONHÞ2 C6 H3 OPO3 H2 ! CaII ðO2 PðOHÞOC6 H3 -2; 6-ðt-BuCONHÞ2 Þn ðOH2 Þ4-n þ 2; 6-i-Pr2 C6 H3 OPO3 H2
ð27:5Þ
Sulfonate ligands like RSO3 are known to have a low pKa owing to their highly conjugated structure. Then, NH O hydrogen-bonded sulfate (PPh4)(SO3-2-t-BuCO-NHC6H4) has a weak NH O hydrogen bond, as determined by IR and 1H NMR analyses in solution and by crystallographic analysis in the solid state [34]. The amide NH does not direct towards one of the sulfonate O atoms; rather, it directs to the middle position between the two O atoms. The strength of the NH O hydrogen bond depends on the basicity of the anion state, and increases in the following order: phenoxides, thiolates, carboxylates, phosphates and sulfates. Thus, the low energy barrier for the deprotonation of these acids lowers the pKa value of the precursor acid for metal ligands. When the coordinating anion form has a high pKa value, the formation and dissociation of a metal–ligand bond depend on that value. Dissociation is prevented by the NH O hydrogen bonds from coordinating carboxylate groups, because these NH O hydrogen bonds lower the pKa value for the corresponding carboxylic acid [16, 30, 31, 35]. The hydrogen bonds involving the coordinating O atom of phenolate [36] or thiolate function similarly [37, 38].
618 Hydrogen Bonding and Transfer in the Excited State
Figure 27.8 Molecular structures: (a) [CaII{O2C-C6H3-2,6-(NHCO-t-Bu)2}4]2; (b) CaII{OCO-2,6-(t-BuCONH)2 C6H3}2(H2O)2
27.3.2 pp–dp covalent interaction affected by hydrogen bonds A tetrakis carboxylate Ca(II) complex, Ca{OCO-2,6-(t-BuCONH)2C6H3}42, has a regular CaO bond distance of 2.4 A with an acute CaOC angle in the bidentate mode (Figure 27.8(a)). Thus, a totally anionic Ca(II) complex has a relatively long CaO bond distance, and the bond is presumably weak because of the typical cationic bond character. A neutral Ca(II) complex, Ca{OCO-2,6- (t-BuCONH)2C6H3}2(H2O)2, has a relativelyshorterCaO bonddistance of2.278(2) Awithan obtuseCaOC angle of164.4(2) intheunidentate mode (Figure 8(b)). The covalency of the CaO bond has been discussed by Einspahr [39]. The relationship between the CaO bond distances and the CaOP bond angles for various Ca(II) complexes with O2POR, O2P(OH)R, O3PR and O2P(OH)R ligands suggests that the covalent CaO bond characters are similar [33]. Our theoretical analysis of the relationship between the CaO bond distance and the CaOP bond angle in Ca phosphate complexes supports the belief that the obtuse CaOP angle increases the d-orbital occupation number for Ca(II) to form a bonding pp–dp interaction with oxygen pp [33]. The NH O hydrogen bond increases the bonding orbital in the CaO bond. NH O hydrogen bonds to the coordinated O atoms prevent the CaO bonds from dissociating by lowering not only the pKa value of the ligands but also the CaO covalent bond strength. Stabilization of metal–oxygen bonds by NH O hydrogen bonds was established for a Tb(III) complex, the ionic radius of which is similar to that of Ca(II) [40]. A Tb(III) complex of 2,6-bis(acetylamino)benzoate exhibits a higher emission intensity than does non-substituted benzoate in aqueous solution. In general, the Tb3þ ion exists as an aqua complex [Tb(OH2)9]3þ in the absence of anion ligands [41, 42]. These coordinated water molecules efficiently quench intrinsic Tb(III) luminescence; the emission intensity of an aqueous terbium solution is normally very weak under these conditions. In the presence of 2,6-bis(acetylamino) benzoate ligand, the carboxylate ligand forms a stable Tb(III) complex assisted by the NH O hydrogen bond and displaces water molecules. Benzenesulfonate Ca(II) complexes have a common ionic CaO bond character without any correlation between CaO bond strength and oxy anion basicity. Intermolecular NH O hydrogen bonds to the sulfonate O atom have been observed in the solid state and in solution [43, 44]. Intramolecular NH O hydrogen bonds between the amide NH and SO have been observed for the Ca(II) arylsulfonate complex, [Ca2(SO3-2-t-BuCONHC6H4)2(H2O)4]n(2-t-Bu-CONHC6H4SO3)2n, sulfonate anion, (HNEt3)(SO3-2-tBuCONHC6H4), (PPh4)(SO3-2-t-BuCONHC6H4), (n-Bu4N)(SO3-2-t-BuCONHC6H4) and the sulfonic
Conformational Switching Between Acids and Their Anions by Hydrogen Bonding
619
acid, 2-t-BuCONHC6H4SO3H by IR and 1H NMR analyses both in the solid state and in solution [34]. The sulfonic acid, the sulfonate anion and its Ca(II) complex have a substantially weak (electrostatic) intramolecular NH O hydrogen bond between the sulfonate O atom and the amide NH. A weak NH O hydrogen bond forms between them because of the strong conjugation of the sulfonate group which decreases the basicity of the oxy anion. In this case, the CaO bond possesses an ionic character. Thus, carboxylates and phosphate monoanions, which have high basicity, differ from sulfonates in the properties of their hydrogen bonds. Systematic investigation reveals that strong NH O hydrogen bonds form only with an oxy anion atom in the high basicity of a ligand coordinating to the metal ions, namely that the ligand is a weak acid such as thiol, phenol, carboxylic acid or phosphoric acid monoanion. 27.3.3 Regulation of thiolate and phenolate ligands in metal complexes by NH X hydrogen bonds The chemical functions of intramolecular NH S hydrogen bonds between thiolate and neighbouring amide NH have been elucidated using simple thiolate and Cys-containing peptide complexes. The hydrogen bond provides several clear functions because the sulfur pp -orbital lobe on the thiolate is remarkably large compared with that on carboxylate oxygen. When the MS bond has a covalent character due to sulfur pp electrons, the NH S bond is extremely weak even if the amide NH is near the thiolate S atom. The amide NH IR band for Hg(S-t-BuCONHC6H4)2 in the solid state shows the absence of NH S hydrogen bonding owing to the consumption of sulfur pp electrons for the covalent HgS bond. A tetrakis Hg(II) complex, (Et4N)2[Hg(S-2-CH3NHCOC6H4)4], has a clear NH S hydrogen bond (Figure 27.8(a)) [45]. Thiolate in a tetrahedral Hg(II) complex can form a relatively strong NH S hydrogen bond with a large amide NH (IR shift Dn ¼ 215 cm1). Formation of the hydrogen bond has been proposed by solution structure analysis of a trigonal complex, Hg(S-t-Bu)(t-buty-oxycarbonyl-cys-Pro-Leu-cysOMe) [46].The NH S hydrogen bond presumably involves geometrical switching between linear Hg(II) and tetrahedral Hg(II) structures. The tetrakis Cd(II) complex has a similar electron configuration in (Et4N)2[Cd(S-2-t-BuCONHC6H4)4], which forms a relatively strong NH S hydrogen bond with shortening CdS bond distance. Theoretical calculations suggest that the NH S hydrogen bond decreases the sulfur pp electron density in the HOMO and thus weakens the related CdS antibonding pp–dp interaction. However, Hg(II) does not exhibit shortening of the HgS bond with NH S hydrogen bonding because of its strong covalency. 199 Hg and 113Cd NMR analyses show a stabilized four-thiolate-coordinated structure for the tetrakis complexes. The chemical shifts show the influence of the NH S hydrogen bonds on pp(Hg)–pp(S) interactions. NH stretching in the IR bands for amide NH in Cd(II) and Hg(II) complexes shows that the NH S hydrogen bonds are stronger than in the corresponding Zn complex. Experimental and theoretical results suggest that the NH S hydrogen bond influences the efficient capture of toxic Cd and Hg ions by metallothioneins [47]. [PtII(bpy)(S-t-BuCONHC6H4)] (bpy ¼ 2,20 -bipyridine) [48] shows the presence of relatively weak NH S hydrogen bonds owing to strong Pt–S covalency. Mononuclear metal–thiolate complexes including Cu(I) [49], Mo(IV) [50], Fe(II) [51] and Co(II) ions [10], exhibit shortening of the MS bond distances with NH S hydrogen bonding when compared with the corresponding thiophenolate complexes without hydrogen bonding. For example, a Co(II)–thiolate complex has a tetrahedral geometry with a strong intramolecular NH S hydrogen bond (Figure 27.8(c)). [MoIVO(S-2t-BuCONHC6H4)4]2 also has a weak NH S hydrogen bond, as detected by IR shift. The MoSe bond distance in [MoIVO(Se-2-t-BuCONHC6H4)4]2 is almost the same as the MoS distance in [MoIVO(S-2-tBuCONHC6H4)4]2. A clear COSY cross-peak is observed in the 77Se1H COSY spectrum (J(77Se1H) ¼ 5.4 Hz). The coupling constant indicates that the covalency of the NHSe bond in the NH Se hydrogen bond is approximately 10% of that of the SeH single bond in CH3SeH (J(77Se1H) ¼ 41.7 Hz) [52].
620 Hydrogen Bonding and Transfer in the Excited State
Figure 27.9
Molecular structures of intramolecularly hydrogen-bonded metal–thiolate complexes
Mo(IV), W(IV), Mo(VI) and W(VI) complexes having intramolecular NH S hydrogen bonds, for example, [MoIVO{S2C2(CONH2)2}]2 and [MVIO2{3,6- (Ph3CCONH)2-1,2-bdt} 2]2 (M ¼ Mo, W) with a 3,6-diamide-1,2-dithiolene ligand, were synthesized as models of the M(IV) and M(VI) states in molybdoand tungsten enzymes (Figure 27.9(a)) [53, 54]. In the WVIO2 complexes, the NH S hydrogen bond trans to the oxo ligand is stronger than that cis to the ligand. The hydrogen bond stabilizes the ligand by trans influence and regulates O-atom transfer in molybdo- and tungsten enzymes. Probably, the hydrogen bond on the trans S atom is a crucial trigger for activation of M¼O transfer. In contrast, the FeS bond distance in P450 model porphyrin complexes slightly elongates with NH S hydrogen bonding, as it does for the similar Ga(III) complex [55, 56]. IR shifts for [FeIII(OEP)(S-tBuCONHC6H4)] (OEP ¼ octaethylphorphinato) and [GaIII(OEP)(S-t-BuCONHC6H4)] indicate the presence of a relatively weak NH S hydrogen bond. Thus, the NH S hydrogen bond controls the properties of the MS bond, which affects the reactivity of a trans-coordinating substrate by mutual trans influence [57]. Such regulation is significant for stabilization and activation of the metal centre towards redox reactions in metalloproteins, metalloenzymes and their model complexes. In transition metal complexes, the NH S hydrogen bond shifts the redox potential towards the positive side [50, 58–60]. Hydrogen bonding causes easy reduction of these electron-rich metal–thiolate complexes by a mild reductant. A similar hydrogen bond is found to the NH O hydrogen bond in metal complexes of phenolate with neighbouring amide groups. The axial tyrosine of heme in catalase is known to have double NH O hydrogen bonds with a neighbouring arginine residue fixed by hydrogen-bonding networks [61, 62]. The crystal structures of [FeIII(TPP){O-2,6-(CF3CONH)2C6H3}] and [FeIII(TPP) (O-2-CF3CONHC6H4)] suggest that NH O hydrogen bonds contribute to the positive shift in redox potential of Fe(III)/Fe(II) [63]. OEP Fe(III) complexes with these ligands at the axial position, [FeIII(OEP)(O-2-CF3CONHC6H4)], [FeIII(OEP)(O-2,6(CF3CONH)2C6H3)], [FeIII(OEP)(OPh)] and [FeIII(OEP) (O-2,6-(i-Pr)2C6H3)], were synthesized and compared with the corresponding TPP Fe(III) complexes [36]. The NH O hydrogen bond elongates the FeO
Conformational Switching Between Acids and Their Anions by Hydrogen Bonding
621
bond distance and widens the FeOC bond angle compared with those in [FeIII(OEP)(OPh)] without the hydrogen bond. Perturbation by the hydrogen bond towards the S atom, presumably associated with a bonding FeO in the LUMO, weakens the FeO bond and decreases the dp–pp overlap. 27.3.4 Switching by an NH S hydrogen bond during the catalytic cycle in p450 and CPO Non-hydrogen-bonded model Fe(III) complexes for the resting state of native P450 and chloroperoxidase (CPO), for example, [FeIII(OEP)(SMe)], are known to be unstable, whereas NH S hydrogen-bonded model Fe(III) complexes are thermodynamically stable. However, reduction of these model complexes results in decomposition. A reported Raman spectral analysis of the complex between P450ox and oxidized [2Fe–2S] putidaredoxin suggests a shortening of the FeS bond upon complexation [64]. Probably, suitable regulation of the FeS bond occurs during the redox reaction. Cys-containing oligopeptide Fe(III) model complexes are also stable owing to the formation of some NH S hydrogen bonds in less polar solvents such as acetonitrile. Reduction of the peptide complex yields a relatively stable Fe(II) form. The NH S hydrogen bond can be controlled by rotation of the amide plane, because breaking the bond stabilizes the Fe(II) species. The orientation angle of the amide planes is affected by other hydrogen bonds, depending on the amide C¼O in the same plane. The addition of multiamide compounds, such as t-butoxycarbonyl-Gly-m-anthranyl methyl ester, to [FeIII(OEP)(Z-cys-Leu- Gly-LeuOMe)] and [FeIII(OEP)(Z-cys-Pro-Ala-Leu-OMe)] in acetonitrile results in a negative shift in redox potential for the former complex and decomposition for the latter complex. Complexation leads to a negative shift in redox potential (Figure 27.10). It is likely that the peptide ligand in the former is flexible and in the latter rigid. The ligand in the former can be used to tune the structure so as to be stable for the reduced state upon oneelectron reduction. Reduction of [FeIII(OEP)(Z-cys-Pro-Ala-Leu-OMe)] having a rigid ligand also results in decomposition. It is consistent that the active centre in CPO does not involve the Fe(II) state. In contrast, reduction of [FeIII(CO)(OEP)(Z-cys-Leu- Gly-Leu-OMe)] with (Et4N)(BH4) under carbon monoxide yields [FeII(CO) (OEP) (Z-cys-Leu-Gly-Leu-OMe)]. The tetrapeptide ligand can switch the structures during both redox states. Solution structures of 6-coordinated ruthenium complexes show that the invariant fragments maintain a b-III turn-like structure and then form weak NH S hydrogen bonds between Cys S and NH (third and fourth amino acid residues) that differ from those in 5-coordinated [GaIII(OEP)(cys-peptide)]. Thus, for the ruthenium complexes, elongation of the N S distance in the NH S hydrogen bond is induced by steric hindrance between porphyrin ring and peptide side chain. This fact suggests that dynamic conformational
Figure 27.10 Molecular structures: (a) [WVIO2{3,6-(CH3CONH)2-1,2-bdt}2]2; (b) [MII{3,6-(CH3CONH)2-1,2bdt}2]2 (M ¼ Zn, Cd, Hg)
622 Hydrogen Bonding and Transfer in the Excited State
switching of the peptide chain induces a change in coordination geometry, which regulates reactivity through the FeS bond character during the catalytic cycle in the native enzymes. However, a Cys-containing hairpin-turn fragment, Cys-Leu-Gly-Leu, supported by the subsequent a-helix, is not able to support an Fe(II) state by the reduction. Presumably, the rigid peptide structure cannot convert to a suitable conformer for the Fe(II) state over a high energy barrier. A simple switching model ligand, such as t-BuNHCOC6H4SH, can provide both states, although even the NH S hydrogen-bonded Fe(III) form gradually decomposes, because the rotation energy barrier between aromatic and amide planes is not high. Furthermore, P450 porphinate model complexes with Cys-containing a-helical peptides have shown that FeS bond properties can be regulated by the NH S bonds supported by the hairpin-turn conformation and the successive NH O¼C hydrogen bond networks in the helix [65]. The above-mentioned hairpin turn structure involving NH S hydrogen bonds is supported by the subsequent a-helix fragment with the same direction of amide dipoles. Actually, the peptide complex [FeIII(OEP)(Ac-LcPAF-LLLLL- ALFL-OMe)] has a larger positive shift (D 130 mV) than does the corresponding tripeptide complex, as the NH S hydrogen bond is stabilized by the a-helix [65]. The solution structure of the peptide model ligand in Fe(III) complexes for P-450 and CPO was determined using [GaIII(OEP)(Ac-LcPAF-LLLLL- ALFL-OMe)] in chloroform-d. The structure has a shorter NH S distance than does the corresponding Cys-containing tetrapeptide complex (Figure 27.11). Thus, an a-helix following the coordinating cysteinyl residue strengthens the NH S hydrogen bond. In contrast, [FeIII(OEP)(AcLcLAF-LLLLL-ALFL-OMe)] has a positive shift of only 70 mV because the hydrogen bond of the Cys-Leu-Ala fragment is not stabilized by the a-helix. The redox reactions seem to be regulated by the cooperating effect of the a-helix and the NH S hydrogen-bonded invariant fragment in the native enzymes. Thus, the NH S hydrogen bond positively shifts the redox potential, which actually contributes to easy reduction of these Fe(III) complexes to Fe(II) complexes with a mild reductant such as (Et4N)(BH4) in acetonitrile. This suggests that the local conformational change of a peptide ligand is crucial for functional switching of the metal centre. 27.3.5 Structural transformation of calcium phosphate clusters involving rearrangement of inter- and intramolecularly hydrogen-bonding networks Ca(II) complexes of phosphate (ROPO3) and phosphonate ligands (RPO3) such as Ca(O3POCH2CH2NH3) have a linear structure [66]. Others, such as Ca(O3PMe), have layered structures [67]. Small ligands predominantly form polymeric structures. On the other hand, extremely bulky and less bulky amide ligands can control coordination geometry. A phosphate monoanion complex, (NMe4)[CaII{O2P(OH)OC6H3-2,6-
Figure 27.11 reduction
On-off switching of NH S hydrogen bonding in [FeIII(OEP)(Z-cys- Leu-Gly-Leu-OMe)] upon
Conformational Switching Between Acids and Their Anions by Hydrogen Bonding
Figure 27.12 1 H NMR
623
Solution structure of [GaIII(OEP)(Ac-LcPAF-LLLLL-ALFL-OMe)] in chloroform-d, determined by
(NHCOCPh3)2}3(NCMe)3], and a phosphate dianion complex, [CaII{O3POC6H3-2,6-(NHCOCPh3)2} (H2O)3(MeOH)2], with an extremely bulky triphenylacylamino group are forced by steric congestion to form a mononuclear Ca(II) core. Less bulky ligands enable restriction of coordination to the side of the Ca cluster. A less-bulky phosphoric acid, 2,6-(PhCONH)2C6H3OPO3H2, gives three novel polynuclear Ca(II)–phosphate and Na(I)–phosphate complexes [68]. The first is the zigzag-chain Ca cluster [CaII{O3POC6H3-2,6-(NHCOPh)2}(H2O)4(EtOH)]n. The second is the cyclic octanuclear form [CaII8{O3POC6H3-2,6-(NHCOPh)2}8(O¼CHNMe2)8(H2O)12]. The third is the hexanuclear complex (NHEt3)[Na3{O3POC6H3-2,6-(NHCOPh)2}2(H2O)(MeOH)7. Crystallographic structures reveal that all have an unsymmetric ligand position owing to the less bulky amide groups. Dynamic transformation of the zigzag-chain Ca structure to the cyclic octanuclear Ca complex is induced by the addition of N,N-dimethylformamide (DMF) owing to the coordination of the DMF molecules (Figure 27.12(a)) [69]. Transformation occurs with reorganization of the intermolecularly and intramolecularly hydrogen-bonding networks. DMF coordination breaks one of the hydrogen bonds to rearrange the network (Figure 27.12(b)). Biopolymer and synthetic polymer ligands successively connected as the less bulky phosphate and carboxylate ligands can precisely coordinate to the surface of Ca clusters, calcium carbonate or hydroxyapatite, which has an unsymmetrical surface coordination geometry [35, 70–72].
27.4 Conclusions Amide NH, prelocated towards the S atom of thiols or the O atom of carboxylic acids, phenols or phosphoric acid monoanions, lowers the pKa value by deprotonation under hydrophobic conditions. The anion easily forms a strong intramolecular NH X (X¼S, O) hydrogen bond that is thermodynamically stabilized. When thiols, phenols or carboxylic acids do not have a prelocated amide NH but have a conformationally switchable remote amide NH, proton-driven conformational switching can be achieved. We have demonstrated
624 Hydrogen Bonding and Transfer in the Excited State
Figure 27.13 (a) Transformation of a Ca(II)–phosphate cluster from zigzag to octanuclear by the addition of DMF. (b) Rearrangement of hydrogen-bond networks by the addition of DMF
twist-boat-to-chair switching on a partially amidated Kemp’s acid derivative, linear-to-turn conformational switching on an unsymmetrically linked phenolic tetraamide, extended-form-to-turn switching on Aspcontaining oligopeptides and photoisomerization switching on azomethine, cinnamic and ortho-coumine carboxylic acid derivatives. These switchings involve a crucial intramolecular hydrogen bond between amide NH and carboxylate O atom. These hydrogen bonds are associated not only with an increase in the formation constant for metal–thiolate, metal–carboxylate, metal–phenolate and metal–phosphate complexes by pKa shift but also with an increase in the covalent character of the metal ligand. The NH S hydrogen bond in P450 model complexes contributes to stabilization of the Fe(III) state, but not of the Fe(II) state. This switching is presumably related to the conformational switching of the Cys-containing peptide chain in P450. Structural transformation occurs on the calcium phosphate cluster between the cyclic octanuclear Ca complex and the zigzag-chain Ca complex, involving rearrangement of inter- and intramolecular hydrogen-bond networks.
Conformational Switching Between Acids and Their Anions by Hydrogen Bonding
625
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. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47.
W. L. Jorgensen, J. M. Briggs and J. Gao, J. Am. Chem. Soc., 109, 6857 (1987). C. O. de Silva, E. C. de Silva and M. A. C. Nascimento, J. Phys. Chem. A, 103, 11 194–11 199 (1999). M. D. Liptak and G. C. Shields, J. Am. Chem. Soc., 123, 7314 (2001). F. Ding, J. M. Smith and H. Wang, J. Org. Chem., 74, 2679 (2009). P. Claverie, J. P. Daudey, J. Langlet et al., J. Phys. Chem., 82, 405 (1978). L. C. G. Freitas, R. L. Longo and A. M. Simas, J. Chem. Soc., Faraday Trans., 88, 189 (1992). J. R. J. Pliego and J. M. Riveros, J. Phys. Chem., 106, 7434 (2002). C. Schutz and A. Warshel, Proteins: Struct., Funct., Genet., 44, 400 (2001). E. Adman, K. D. Watenpaugh and L. H. Jensen, Proc. Natl Acad. Sci. USA, 72, 4854 (1975). T. Okamura, S. Takamizawa, N. Ueyama and A. Nakamura, Inorg. Chem., 37, 18 (1998). N. Ueyama, T. Okamura, Y. Yamada and A. Nakamura, J. Org. Chem., 60, 4893 (1995). N. Ueyama, S. Moriyama, T. Ueno and A. Nakamura, pKa shift by NH S hydrogen bond in the hair-pin turn structure of Cys-containing oligopeptides, in Peptide Science: Present and Future Proceedings of the International Peptide Symposium. Kluwer, Kyoto, Japan, p. 288. (1999). D. Kanamori, A. Furukawa, T. Okamura et al., Org. Biomol. Chem., 3, 1453 (2005). D. Kanamori, T.-A. Okamura, H. Yamamoto and N. Ueyama, Angew. Chem. Int. Ed., 44, 969 (2005). D. Kanamori, T. Okamura, H. Yamamoto et al., Bull. Chem. Soc. Jpn, 77, 2057 (2004). A. Onoda, Y. Yamada, J. Takeda et al., Bull. Chem. Soc. Jpn, 77, 321 (2004). H. Wang, G. M. Vath, K. J. Gleason et al., Biochemistry, 43, 8234 (2004). A. Onoda, H. Haruna, H. Yamamoto et al., Eur. J. Org. Chem., 641 (2005). A. Onoda, A. Onoda, H. Yamamoto et al., Peptide Sci., 80, 233 (2005). K. Suguna, R. R. Bott, E. A. Padlan et al., J. Mol. Biol., 196, 877 (1987). A. Wlodawer, M. Miller, M. Jasko’lski et al., Science, 245, 616 (1989). S. Shinkai, K. Araki, P. D. J. Grootenhuis and D. N. Reinhoudt, J. Chem. Soc., Perkin Trans., 1883 (1991). P. D. J. Grootenhuis, P. A. Kollman, L. C. Groenen et al., J. Am. Chem. Soc., 112, 4165 (1990). T. Matsuhira, H. Yamamoto, A. Onoda et al., Org. Biomol. Chem., 4, 1338 (2006). T. Matsuhira, H. Yamamoto, T. Okamura and N. Ueyama, Org. Biomol. Chem., 6, 1926 (2008). T. Matsuhira, K. Tsuchihashi, H. Yamamoto et al., Org. Biomol. Chem., 6, 3118 (2008). M. M. Yamashita, L. Wesson, G. Eisenman and D. Eisenberg, Proc. Natl Acad. Sci. USA, 87, 5648 (1990). N. Ueyama, M. Inohara, T. Ueno et al., Polymer J., 29, 949 (1997). N. Ueyama, M. Inohara, A. Onoda et al., Inorg. Chem., 38, 4028 (1999). N. Ueyama, Y. Yamada, J. Takeda et al., J. Chem. Soc., Chem. Commun., 1377 (1996). N. Ueyama, J. Takeda, Y. Yamada et al., Inorg. Chem., 38, 475 (1999). A. Onoda, Y. Yamada, M. Doi et al., Inorg. Chem., 40, 516 (2001). A. Onoda, Y. Yamada, T. Okamura et al., Inorg. Chem., 41, 6038 (2002). A. Onoda, Y. Yamada, M. Doi et al., Inorg. Chem., 40, 516 (2001). N. Ueyama, T. Hosoi, Y. Yamada et al., Macromolecules, 31, 7119 (1998). D. Kanamori, Y. Yamada, A. Onoda et al., Inorg. Chim. Acta, 358, 331 (2005). N. Ueyama, K. Taniuchi, T. Okamura et al., Inorg. Chem., 35, 1945 (1996). N. Ueyama, T. Okamura and A. Nakamura, J. Chem. Soc., Chem. Commun., 1019 (1992). H. Einspahr and C. E. Bugg, Acta Crystallogr., Sect. B: Struct. Sci., 57, 1044 (1981). A. Onoda, Y. Yamada, Y. Nakayama et al., Inorg. Chem., 43, 4447 (2004). A. Chatterjee, E. N. Maslen and K. J. Watoson, Acta Crystallogr., Sect. B: Struct. Sci., 44, 386 (1988). A. Chatterjee, E. N. Maslen and K. J. Watoson, Acta Crystallogr., Sect. B: Struct. Sci., 44, 381 (1988). T. Balasubramanian and P. T. Muthuah, Acta Crystallogr. Sect. C, C52, 2072 (1996). V. A. Russell, C. C. Evans, W. Li and M. D. Ward, Science, 276, 575 (1997). M. Kato, K. Kojima, T.-A. Okamura et al., Inorg. Chem., 44, 4037 (2005). T. Yamamura, T. Watanabe, A. Kikuchi et al., J. Phys. Chem., 99, 5525 (1995). K. Baba, T. Okamura, H. Yamamoto et al., Inorg. Chem., 47, 2837–2848 (2008)
626 Hydrogen Bonding and Transfer in the Excited State 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.
M. Kato, T. Okamura, H. Yamamoto and N. Ueyama, Inorg. Chem., 44, 1966 (2005). T. Okamura, N. Ueyama, A. Nakamura et al., J. Chem. Soc. Chem. Commun., 1685 (1993). N. Ueyama, T. Okamura and A. Nakamura, J. Am. Chem. Soc., 114, 8129 (1992). N. Ueyama, T. Okamura and A. Nakamura, Chem. Commun., 1019 (1992). T. Okamura, K. Taniuchi, K. Lee et al., Inorg. Chem., 45, 9374 (2006). K. Baba, T. Okamura, C. Suzuki et al., Inorg. Chem., 45, 894 (2006). K. Baba, T. Okamura, H. Yamamoto et al., Inorg. Chem., 45, 8365 (2006). N. Ueyama, N. Nishikawa, Y. Yamada et al., J. Am. Chem. Soc., 118, 12 826 (1996). T. Okamura, N. Ueyama and A. Nakamura, Chem. Lett., 199 (1998). A. Dey, T. Okamura, N. Ueyama et al., J. Am. Chem. Soc., 127, 12 046 (2005). N. Ueyama, T. Terakawa, M. Nakata and A. Nakamura, J. Am. Chem. Soc., 105, 7098 (1983). N. Ueyama, M. Nakata, M. Fuji et al., Inorg. Chem., 24, 2190 (1985). N. Ueyama, Y. Yamada, T. Okamura et al., Inorg. Chem., 35, 6473 (1994). M. R. N. Murthy, T. J. ReidIII, A. Sicignano et al., J. Mol. Biol., 152, 465 (1981). I. Fita and M. G. Rossmann, J. Mol. Biol., 185, 21 (1985). N. Ueyama, N. Nishikawa, T. Okamura et al., Inorg. Chem., 37, 2415 (1998). M. Unno, J. F. Christian, D. E. Benson et al., J. Am. Chem. Soc., 119, 6614 (1997). T. Ueno, Y. Kousumi, K. Yoshizawa-Kumagaye et al., J. Am. Chem. Soc., 120, 12 264 (1998). P. Bissinger, O. Kumberger and A. Schier, Chem. Ber., 124, 509 (1990). G. Cao, V. M. Lynch, S. Swinnea and T. E. Mallouk, Inorg. Chem., 29, 2112 (1990). A. Onoda, Y. Yamada, T. Okamura et al., J. Am. Chem. Soc., 124, 1052 (2001). A. Onoda, Y. Yamada, T. Okamura et al., J. Am. Chem. Soc., 124, 1052 (2002). N. Ueyama, H. Kozuki, M. Doi et al., Macromolecules, 34, 2607 (2001). K. Takahashi, H. Kozuki, A. Onoda et al., J. Inorg. Organomet. Polym., 12, 99 (2002). K. Takahashi, M. Doi, A. Kobayashi et al., Chem. Lett., 33, 192 (2004).
28 Charge Transfer in Excited States: ab initio Molecular Dynamics Simulations Maciej Kolaski1,2, Anupriya Kumar1, Han Myoung Lee1 and Kwang S. Kim1 1
Center for Superfunctional Materials, Department of Chemistry, Pohang University of Science and Technology, San 31, Hyojadong, Namgu, Pohang 790-784, Korea 2 Department of Theoretical Chemistry, Institute of Chemistry, University of Silesia, 9 Szkolna Street, 40-006, Katowice, Poland
28.1 Introduction Dissociation of a chemical compound by light is called photodissociation or photofragmentation, which plays an important role in many chemical [1, 2] and atmospheric phenomena [3]. In the earth’s atmosphere, photolysis occurs as a part of a series of reactions in which primary pollutants such as hydrocarbons and nitrogen oxides react to generate secondary pollutants [4]. In the stratosphere, ozone is formed through photolysis by ultraviolet (UV) light [5]. Chlorofluorocarbons (CFCs) are broken down by photolysis in the uppermost atmosphere to form halogen-free radicals, which are responsible for destroying the ozone layer [6]. Ion–water cluster interactions are important for understanding solvation/desolvation phenomena in chemical processes, designing efficient ionophores for biological molecular recognition and engineering of selfassembled nanomaterials. Photoexcitation of halide anion–water clusters causes the charge transfer of an electron from the halogen anion to the solvent (water), i.e. the charge-transfer-to-solvent (CTTS) phenomenon. This photoexcitation results in the dissociation of hydrated halide anions into very active free halide radicals and an excess electron–water cluster [e(H2O)n] in which the excess electron is stabilized by the electron dipole interaction with the water network. The photolysis of water is essential in neutron irradiation to the cooling water in nuclear reactors and in the destruction of living cells caused by radiation. A possible fuel source may be obtained via the photolysis of water to hydrogen and oxygen gases. The mechanism of water photolysis is still not well understood. The experimental ionization energy of water in the gas phase is around 12.6 eV, which is significantly larger than
Hydrogen Bonding and Transfer in the Excited State, Volume I & II © 2011 John Wiley & Sons, Ltd. ISBN: 978-0-470-66677-7
Edited by Ke-Li Han and Guang-Jiu Zhao
628 Hydrogen Bonding and Transfer in the Excited State
the corresponding excitation energies in the UV/VIS spectrum. Thus, if the photon energy is above the ionization threshold, the mechanism of dissociation is driven by the ionization process. Photoexcitation of hydrated iodic acids breaks the HI bonds and releases both hydrogen and iodine radicals, which leads to the possible utilization of hydrogen radicals in generating hydrogen by dissociating water with hydrogen iodide. Pyrroles are one of the important compounds for studying aromatic systems, as they are the components of various functionally significant biological molecules, such as amino acids, nucleotide bases and aromatic rings such as porphyrins of heme, chlorins/bacteriochlorins of chlorophyll and corrin rings of vitamin B12. In photochemical reactions in water, a universal solvent and quencher, the electron transfer process from aromatic compounds, such as optically active amino acid tryptophane and indole, has been confirmed to trigger the formation of a hydrated electron. In spite of the importance of photolytic reactions in environmental chemistry and biology, as well as its potential role in harnessing nature’s energy source, the excited-state dynamics of hydrated ions, radicals and aromatic compounds is hardly understood. Here, by using excited-state ab initio molecular dynamics (ES-AIMD) simulations based on complete-active-space self-consistent field (CASSCF) formalism, we have investigated the difference in the photoexcitation dynamics between I (H2O)n¼2–4 and I(H2O)n¼5 clusters, water photolysis, the photoinduced charge transfer process of hydrated hydrogen iodide [(H2O)nHI (n ¼ 1–6)] and the photoexcitation of pyrrole–water clusters. However, the lack of information available for reliable excited-state geometries, as well as the detailed experimental data, has made the study challenging.
28.2 Charge-Transfer-to-Solvent-Driven Dissolution Dynamics of I(H2O)2–5 Upon Excitation The excitation of hydrated halides results in the transfer of an excess electron from the anionic precursor to the water cluster [7–12] which is then stabilized by the network of water molecules. This causes the dissociation of hydrated halides into halogen radicals and wet electron–water clusters. In this review, we report the CTTSdriven dissolution dynamics for I(H2O)n¼2–5 complexes using ES-AIMD simulations employing the CASSCF method. This analysis shows that, after the iodine radical is detached from the I(H2O)n¼2–5 complex, a simple population decay is observed for small water clusters (n ¼ 2–4), while a substantial reorganization of the water network to form an entropy-driven structure is observed for n ¼ 5. These results are in very good agreement with ultrafast pump–probe experiments [13]. In this work, O and H atoms are treated with the aug-cc-pVDZ þ (2s2p/2s) basis set [14], where the extra diffuse 2s2p and 2s functions are added to all oxygen and hydrogen atoms in order to properly stabilize the excess electron. In the case of hydrated iodine anion clusters, the extended basis sets are necessary to study the excited states, and the set of highly diffuse functions are needed for the study of the dissociation of the anionic species, including excess-electron–water clusters. As the aug-cc-pVDZ basis set is not available for iodine, the CRENBL effective core pseudopotential (ECP) basis set was employed [14]. An active space of six electrons and six active orbitals was used (CAS[6, 6]) for clusters with up to three water molecules. As ES-AIMD simulations are computationally very demanding for clusters comprising four and five water molecules, the optimum choice of the active space for the reduction of computational resources is required, even though numerical accuracy is slightly sacrificed. Thus, a reduced active space (CAS[4,4]) was used for the practical study of larger complexes (n ¼ 4, 5). The ES-AIMD simulations of the I(H2O)n¼2–5 complexes were carried out for 800 fs (n ¼ 2–4) and 600 fs (n ¼ 5) with a time step of 0.2 fs [15]. The ES-AIMD simulations for an excited state of I(H2O)2–4 exhibit
Charge Transfer in Excited States: ab initio Molecular Dynamics Simulations
629
Figure 28.1 Snapshots of the evolving process of the I(H2O)n¼2,5 clusters upon excitation, taken from ab initio molecular dynamics simulations. Reprinted with permission from [15b]. Copyright 2008 American Chemical Society
simple population decay, while that of I–(H2O)5 shows drastic rearrangement of the water network. Snapshots of the time evolution of the structures of I(H2O)n¼2,5 clusters upon excitation are shown in Figure 28.1. In the ground state, the most stable structure of the water pentamer cluster in the I(H2O)5 complex is a planar cyclic water tetramer ring with a water molecule attached through an H-bond. For the I(H2O)5 complex, five different ES-AIMD simulations with KE0 ¼ 0, 100, 200, 300, and 400 K were performed. The simulation results are almost identical, but there is a substantial difference in structures and electronic properties between ES-AIMD simulations carried out at 0 K and at higher temperatures. In the latter, the pentamer has a quasi-linear structure due to the entropy effect (which favours more flexible structures), while in the former it has a deformed quasi-tetragonal ring stabilized by an additional hydrogen bond. In this regard, all the simulations are almost identical, except for KE0 ¼ 0 K, regardless of the initial geometry and initial velocities. We find that this is due to the fact that, above 100 K, the linear structure of e(H2O)5 is more stable in terms of the free energy than the quasi-tetragonal ring structure. As compared with the tetragonal ring structure, the quasi-linear structure is 1.0 kcal mol1 less stable at 0 K, but becomes more stable above 100 K. For instance, at 298 K, it is 4.2 kcal mol1 more stable at the B3LYP/6-311þþG level of theory. The experimental VDEs of I(H2O)n are evaluated to be 0.03, 0.08 and 0.16 eV for clusters n ¼ 2, 3 and 4 respectively [16]. The abnormal shift from 0.15 to 0.4 eV in the VDE of I(H2O)5 was observed in the pump–probe experiment to be around 600 fs after excitation. This is demonstrated by the computed VDEs, which increase from 0.15 to 0.38 eV during 600 fs of ES-AIMD simulations. This indicates complex dynamics for the cluster n ¼ 5, which involves significant reorganization of the water cluster to more effectively accommodate the excess electron. In contrast, smaller clusters exhibit a simple population decay. Our study clearly shows how the detachment process of I(H2O)n¼2–5 evolves upon excitation and why the evolution process of I(H2O)5 is significantly different from the other smaller-sized clusters I(H2O)2–4. To reproduce experimental VDE values for I(H2O)5, it was necessary to perform ES-AIMD simulations above 100 K
630 Hydrogen Bonding and Transfer in the Excited State 0.50 IW2 IW3 IW4 IW5 300K
0.45 0.40
VDE [eV]
0.35 0.30 0.25 0.20 0.15 0.10 0.05 0.00 0
200
400
600
800
Time [fs]
Figure 28.2 Time evolution of the vertical detachment energy (VDE) for the I(H2O)n¼2–5 clusters. Reprinted with permission from [15b]. Copyright 2008 American Chemical Society
(Figure 28.2). The ES-AIMD study gives an insight into the process of the dehydration/dissolution of halogen atoms by photoexcitation phenomenon and the rearrangement mechanism of the excess-electron–water clusters.
28.3 Dynamics of Water Photolysis: Excited-State and Born–Oppenheimer Molecular Dynamics Study In spite of the importance of water photolysis in atmospheric chemistry, the mechanism of photodissociation is still not well understood [17]. Two competitive mechanisms of water photolysis have been suggested: *
*
(i) Upon excitation, a water molecule is defragmented into two radicals: OH and H , and the excess energy is transferred mainly to the H radical. Consequently, the H radical has extremely large kinetic energy and promptly reacts with a water molecule in the first hydration shell, forming H3Oþ and a hydrated electron [, 18–20]: *
*
H2 O þ hn ! H2 O* ! H þ OH *
*
H ðhotÞ þ H2 O ! H3 O þ þ e ðhydratedÞ *
(ii) It has been proposed that ionization above the Born–Oppenheimer threshold leads to the creation of H3Oþ , an OH radical and a hydrated electron [19, 21–24]: *
H2 O þ hn ! H2 O þ þ e ðhydratedÞ H2 O þ þ H2 O ! H3 O þ þ OH
*
Both proposed schemes lead to identical products, in spite of their significantly different reaction mechanisms.
Charge Transfer in Excited States: ab initio Molecular Dynamics Simulations
631
Figure 28.3 Snapshots of the evolving process of the (H2O)n¼1,2 clusters upon excitation (ES-AIMD simulations). Reprinted with permission from [26]. Copyright 2008 American Chemical Society
In this study, we report on the photoexcitation-driven mechanism by employing ES-AIMD simulations based on the complete-active-space self-consistent field (CASSCF) approach, and on the photoionizationdriven mechanism by using the ground-state unrestricted Møller–Plesset second-order perturbation theory (UMP2) based on Born–Oppenheimer molecular dynamics (BOMD) simulations [25]. The CAS[6, 6] active space was used for the water molecule and the CAS[12, 12] active space for the water dimer. In ES-AIMD simulations, the molecular orbitals involved in the excitation should be present in the active space. In this ES-AIMD simulation, we considered the first singlet excited state (i.e. HOMO–LUMO excitation) [26]. The initial structures having the ground-state minimum-energy geometry (optimized at the CASSCF/augcc-pVDZ level of theory) were vertically excited. The ES-AIMD simulations of the water clusters were carried out for 40 fs with a time step of 0.1 fs. In this case, it was necessary to decrease the time step because of convergence problems. We also performed longer ES-AIMD simulations, but all important changes take place during the first 20 fs of the MD run. Figure 28.3 shows the time evolution of the conformational changes of the water clusters (H2O)n¼1,2. At the beginning of the ES-AIMD simulation, the dynamics of a single water molecule differs from the dynamics of larger water clusters, showing very strong OH stretching motions. Both OH bonds start to break at 5 fs. At 10 fs, one hydroxyl group is formed again, and the hydrogen radical is released. The detached hydrogen radical has extremely large kinetic energy (it is very hot), which increases up to 55 kcal mol1, and immediately reacts with a water molecule from the first hydration shell. This is consistent with the photodissociation mechanism of water photolysis. For the water dimer, the OH bond starts to stretch at 4 fs, and subsequently the hydrogen radical is released from the molecular system with very large kinetic energy. While the OH bond breaks, the kinetic energy of the released hydrogen grows very rapidly. After 10 fs, the kinetic energy of the hydrogen radical increases to 55 kcal mol1 and remains constant until the end of the molecular dynamics run. When the hydrogen radical is finally detached, the hydroxyl group and the adjacent water molecule form a stable structure. In the case of UMP2-BOMD simulations for ionized water clusters (Figure 28.4), the initial UMP2optimized minimum energy structures of neutral clusters were employed. We set the initial KE to zero, as in the former case. The UMP2-BOMD simulations were carried out for 1000 fs with a time step of 0.2 fs. The proton strongly oscillates between the hydroxyl group and the neighbouring water molecule in the ionized water dimer. The mobile proton has a relatively large kinetic energy; however, it does not exceed 10 kcal mol1, which holds the proton between the two oxygen atoms. The time evolution of OH distances clearly shows that the Eigen (H3Oþ -like water cluster) and Zundel (H2O. . .Hþ . . .OH2-like water cluster) forms compete [27]. After 500 fs of UMP2-BOMD simulations, an Eigen form is finally created, which is
632 Hydrogen Bonding and Transfer in the Excited State
þ Figure 28.4 Snapshots of the evolving process of the (H2O)n¼2 clusters upon photoionization of (H2O)n¼2 (UMP2-BOMD simulations). Reprinted with permission from [26]. Copyright 2008 American Chemical Society
3.6
-152.09
+
(H2O)2
3.2
+
(H2O)2 -152.10
r(A...B) [A]
2.8
H
-152.11
2.4
2
H-O1 H-O2 O1-O2
2.0
1
-152.12
1.6 -152.13
1.2 0.8
-152.14 0
200
400
600
Time [fs]
800
1000
0
200
400
600
800
1000
Time [fs]
Figure 28.5 Time evolution of distances and potential energy for the ionized water dimer. Reprinted with permission from [26]. Copyright 2008 American Chemical Society
energetically more favourable. While the Eigen conformer is formed, the kinetic energy of the mobile proton significantly decreases. Figure 28.5 shows that the distance between O1 and O2 atoms oscillates highly at the beginning of the simulations, but, after 500 fs, it does not change considerably. The time evolution of potential energy, presented in Figure 28.5, for an ionized water dimer proves that the resulting structure is the minimum energy conformer. This study reported the first ab initio excited-state molecular dynamics results to explain the mechanism of water photolysis. The first mechanism is driven by the photoexcitation process. By carrying out ES-AIMD simulations based on the CASSCF approach, we proved that, upon excitation, water clusters release very hot hydrogen radicals and (hydrated) hydroxyl radicals within 15 fs. In the case of water photolysis driven by ionization, all structural changes are much slower in comparison with the dynamics controlled by the photoexcitation phenomenon. The hydrogen atoms that are responsible for hydrogen bond formation determine the structural rearrangement of ionized water clusters. At the beginning of UMP2-BOMD simulations, these mobile protons have relatively large kinetic energy (although much smaller in comparison with that in ES-AIMD simulations), and thus Eigen and Zundel forms compete. After 500 fs, one proton is attached to the adjacent water molecule, and the hydronium and OH radical are finally created. *
Charge Transfer in Excited States: ab initio Molecular Dynamics Simulations
633
28.4 Photodissociation of Hydrated Hydrogen Iodide Clusters: ab initio Molecular Dynamics Simulations In various chemical, biochemical and atmospherical systems, it is important to note that photochemical reactions often produce hydrogen radicals [28]. As the addition of an electron to a proton, which involves hydrogen transfer, proton transfer [29] and solvent-driven charge transfer [30], can form a hydrogen radical, the charge-transfer-to-solvent-driven hydrogen radical chemistry is an interesting subject. Hydrogen halide acids produce hydrogen and halogen radicals through the photodissociation phenomenon [31]. In spite of its importance in widespread environmental and biological issues, as well as in the possibility of harnessing nature’s energy source, the hydrogen radical reaction mechanism and dynamics are not well understood. As one of the important issues in atmospheric chemistry, we have studied the photoinduced excitation dynamics of hydrated hydrogen iodide [32]. To investigate the dynamic behaviour of the dissociation of hydrogen iodide and the release of radicals from the cluster, we show in Figure 28.6 the structural changes and NBO charge evolution from the ES-AIMD simulations for HI(H2O) and HI(H2O)2 complexes with initial kinetic energies KE0 of 0 and 200 K. In the simulations, the transition state of the proton between HI and H2O was observed, and the kinetic energy of the released H atom was very large (35 kcal mol1). The r(IH) distance monotonically increases, showing the release of iodine atom at both KE0 ¼ 0 K and KE0 ¼ 200 K for HI(H2O). The H radical in HI(H2O) was released at the KE0 of 200 K, but not at the KE0 of 0 K. For HI(H2O), the NBO charge of H in HI is þ 0.09 au before excitation and þ 0.08 au at the time (t ¼ 0) of excitation, but þ 0.34 au after 60 fs (at 0 K) [0.0 au after 70 fs for 200 K]. Then, the formation of the neutralized H3O is noted along with the complete release of the iodine radical. To study the release of the hydrogen radical from the neutralized H3O, ground-state AIMD simulations for H3O and H5O2 were carried out. In this case, the H radical was easily released at the KE0 above 1000 K (for H3O) and 2500 K (for H5O2) to cross over the small activation energy barrier (0.1 eV) (Figure 28.7). However, it was not released below the KE0 of 1000 K. Even though the energy barrier from H3O to H þ H2O is relatively small (0.1 eV ¼ 2 kcal mol1), a high KE0 is required for the H radical to get over the barrier towards the dissociation process. Nevertheless, it should be emphasized that, even at very low temperatures, the release of the H radical occurs because in a system consisting of molecules of the order of Avogadro’s number (1024), a large number of molecules would have enough KE0 (according to the Maxwell–Boltzmann distribution) to release H radicals. Thus, hydrated iodic acids, upon excitation, release hydrogen radicals or hydrogen molecules as well as iodine radicals. This photoexcitation process involving CTTS takes place in four steps: (i) hydration of the acid; (ii) charge transfer to water upon excitation of hydrated acid; (iii) release of the neutral iodine atom; (iv) detachment of the hydrogen radical [26]. The detachment of iodine from excited hydrated hydro-iodic acids is highly exothermic. The detachment of hydrogen radicals from hydrated hydronium radicals is spontaneous. It occurs only if the initial kinetic energy of the cluster is large enough to cross over the small activation energy barrier.
28.5 Excited-State Dynamics of Pyrrole–Water Complexes: ab initio Excited-State Molecular Dynamics Simulations Investigation of hydration- and photoexcitation-driven charge transfer phenomena [33, 34] is essential for understanding the solvation phenomena. Studies of solvated clusters involving hydrogen bonds, as well as
634 Hydrogen Bonding and Transfer in the Excited State
Figure 28.6 ES-AIMD simulations of HI(H2O)1,2 with an initial kinetic energy of 0 or 200 K (CASSCF calculations with aug-cc-pVDZ and CRENBL ECP basis sets). The time evolutions of the HI, HO and IO distances and the NBO charges in the dissociation process are depicted. Reprinted with permission from [32]. Copyright 2008 Wiley-VCH Verlag GmbH & Co. KGaA
Charge Transfer in Excited States: ab initio Molecular Dynamics Simulations
635
Figure 28.7 Ground-state AIMD simulations of H3O and H5O2 (CASSCF with the aug-cc-pVDZ basis set). The time evolutions of interatomic distances and the NBO charges in the dissociation process are depicted. Reprinted with permission from [32]. Copyright 2008 Wiley-VCH Verlag GmbH & Co. KGaA
those involving weak van der Waals interactions with p systems, have been useful for mimicking and understanding solute–solvent interactions and s/p hydrogen bonding networks with p systems [35–38]. For the pyrrole–(H2O) cluster, an ES-AIMD simulation was carried out with an initial KE0 of 0 K [39]. For the pyrrole–(H2O)2 cluster, it was necessary to perform ES-AIMD simulations with KE0 values of 0 and 300 K.
636 Hydrogen Bonding and Transfer in the Excited State
The initial structures having the ground-state minimum-energy geometry (optimized at the CASSCF/aug-ccpVDZ þ (2s2p/2s) level of theory) were vertically excited at 0 fs. The experiments were performed at very low temperatures, so we set up the initial kinetic energy of the systems to 0 K for both pyrrole–(H2O) and pyrrole–(H2O)2 complexes. For the ES-AIMD simulations of pyrrole–(H2O)2 at 300 K, the initial velocities were set according to the Maxwell–Boltzmann distribution. The ES-AIMD simulations demand enormous computing time, so it is not possible to do more than one or two trajectories. Even a single trajectory in an ES-AIMD simulation performed at 300 K is likely to represent what would be observed near the peak in the distribution of trajectory results; therefore, simulations with a few different initial KE0 values should provide a reasonable understanding of the dynamics. The results of ES-AIMD simulations for pyrrole–(H2O) and pyrrole–(H2O)2 are presented in Figure 28.8. It shows the changes in the structures of pyrrole–(H2O) (KE0 ¼ 0 K) from the 400 fs ES-AIMD and those of pyrrole–(H2O)2 (KE0 ¼ 0 and 300 K) from the 200 fs ES-AIMD simulation trajectories. In the case of pyrrole–(H2O), the KE of the pyrrole increases up to 7 kcal mol1 at the beginning of ES-AIMD simulations (5 fs). On the other hand, the KE of the water molecule smoothly increases up to 2 kcal mol1 around 40–80 fs, and then starts to fluctuate. Pyrrole–(H2O) undergoes charge transfer upon excitation and forms a highly polarized state during the ES-AIMD run. The CTTS is maximized around 100 fs, after which it strongly fluctuates and reaches another maximum around 300 fs. In the case of a pyrrole–(H2O)2 cluster with KE0 ¼ 0 K, the KE of pyrrole increases up to 5 kcal mol1 at the beginning of ES-AIMD simulations. The second water molecule (denoted as W2) involved in the p–H interaction picks up the KE up to 3 kcal mol1, while the first water molecule (denoted as W1) s H-bonded to the pyrrole molecule picks up the KE up to 2 kcal mol1. Therefore, the water molecule involved in the p H-bond formation is more accessible to excitation than the other water molecule involved in the s H-bond. The distance between oxygen atoms (r(O1. . .O2)) increases after 100 fs, and the hydrogen bond between two water molecules is broken after 150 fs of simulations. In the meantime, the electron cloud is completely reorganized (the total charge is localized on one water molecule). At 0 K, a substantial charge transfer occurs from pyrrole to W2 through space, while W1 is neutral, i.e. charge transfer is not observed for the s H-bond. During ES-AIMD simulations, the negative charge begins to build up even for W1. At 100 fs, the net charges localized on pyrrole and W2 are 0.53 and 0.29 au respectively, which exhibits the maximum charge transfer during simulations. Consequently, the cluster was kept as a highly polarized state complex, showing dynamics very similar to the pyrrole–H2O system (as W2 is isolated from the complex). The simulation results of pyrrole–(H2O)2 with KE0 ¼ 300 K show the dissociation of the cluster into pyrrole and two completely isolated water molecules. For the latter simulations, the charge transfer takes place at the beginning of simulations, however, and almost disappears as time elapses, because the pyrrole–(H2O)2 cluster turns into pyrrole and two isolated water molecules. In the excited state, at t ¼ 0 fs, the charge density is almost completely transferred to the farthest water molecule, which is repelled by the depleted p-electron cloud of the aromatic ring. During ES-AIMD simulations, the electron charge density is distributed over all the water molecules (but more on the nearest water molecule). In the excited-state dynamics of pyrrole–(H2O) and pyrrole–(H2O)2, the charge transfer occurs through a hydrogen bond between pyrrole and one water molecule upon excitation, and the charge transfer is maximized at 100 fs, resulting in a significantly polarized state of the complex. However, a CTTS complex is never formed. As the temperature increases, the charge transfer becomes less important.
28.6 Conclusions Excited-state molecular dynamics based on the complete-active-space self-consistent field approach is a valuable tool for studying the charge transfer phenomena in various molecular systems. Our study
Charge Transfer in Excited States: ab initio Molecular Dynamics Simulations
637
Figure 28.8 Time evolution for the first excited states of a pyrrole–(H2O) complex with a KE0 of 0 K (a) and of a pyrrole–(H2O)2 complex with a KE0 of 0 K (b) and 300 K (c) in CASSCF[6,6]/aug-cc-pVDZ ES-AIMD simulations. Here, both the ground and the excited states at 0 fs have the same geometry. (d) Time evolution of NBO (natural bond orbital) charges localized on the pyrrole and water molecules for pyrrole–(H2O) and pyrrole–(H2O)2 with a KE0 of 0 K. Reprinted with permission from [39]. Copyright 2008, American Institute of Physics
638 Hydrogen Bonding and Transfer in the Excited State
demonstrates how the detachment process of I(H2O)n¼2–5 evolves upon excitation and why the evolution process of I(H2O)5 is different from the other smaller-sized clusters I(H2O)2–4. The I(H2O)n dynamics is important for the design of novel dynamic receptors that can selectively bind ions and then release them by the CTTS mechanism upon excitation. This concept could open a new field of the dynamic host–guest chemistry involved in the capture–transport–release mechanism of smart receptors. We reported the first ab initio excited-state molecular dynamics results to unravel the mechanism of water photolysis. We studied the photoexcitation mechanism by using ES-AIMD simulations based on the CASSCF approach and the photoionization mechanism by using the ground-state unrestricted Møller–Plesset second-order perturbation theory based on BOMD simulations. We performed ES-AIMD simulations for HI(H2O)n¼1,2 clusters. This study clearly shows how the hydrogen and halogen radicals can be dissociated and released from their hydrated acids. Indeed, we were able to demonstrate the predicted process from very simple experiments. H radicals released from iodic acid in water by UV light are useful as a reducing agent. In the excited-state dynamics of pyrrole with one and two water molecules, charge transfer occurs through a hydrogen bond between the pyrrole molecule and one water molecule upon excitation, which is maximized at 100 fs, resulting in a significantly polarized state of the complex. As the temperature increases, the charge transfer becomes much less significant.
References 1. (a) M. N. R. Ashfold and J. E. Baggott (eds), Molecular Photodissociation Dynamics; Advances in Gas-Phase Photochemistry and Kinetics. Royal Society of Chemistry, London, UK (1987); (b) P. Pichat, Water Sci. Technol., 55, 167 (2007). 2. (a) S. M. Hurley, T. E. Dermota, D. P. Hydutsky and A. W. Castleman, J. Chem. Phys., 118, 9272 (2003); (b) T. Ikeda, S. Fujiyoshi, H. Kato et al., J. Phys. Chem. B, 110, 7883 (2006); (c) R. Abe, K. Sayama and H. Sugihara, J. Phys. Chem. B, 109, 16 052 (2005). 3. (a) B. P. Bonev, M. J. Mumma, M. A. DiSanti et al., Astrophys. J., 653, 774 (2006); (b) B. P. Bonev, M. J. Mumma, N. Dello Russo et al., Astrophys. J., 615, 1048 (2004); (c) B. J. Sandor and R. T. Clancy, J. Geophys. Res., 108, 4463 (2003);(d) Y. L. Yung and W. B. DeMore, Photochemistry of Planetary Atmospheres. Oxford University Press, New York, NY (1999). 4. C. N. Hewitt, Reactive Hydrocarbons in the Atmosphere. Academic Press, San Diego, CA (1999). 5. C. S. Zerefos, I. S. A. Isaksen and I. Ziomas (eds), Chemistry and Radiation Changes in the Ozone Layer. NATO ASI Series, 557. Kluwer Academic Publishers, Dordrecht, The Netherlands (2000). 6. S. Cagin, Between Earth and Sky: How CFCs Changed Our World and Endangered the Ozone Layer. Pantheon, New York, NY (1993). 7. (a) J. A. Kloepfer, V. H. Vilchiz, V. A. Lenchenkov et al., J. Chem. Phys., 113, 6288 (2000); (b) J.-H. Choi, K. T. Kuwata, Y.-B. Cao and M. Okumura, J. Phys. Chem. A, 102, 503 (1998). 8. (a) P. Ayotte, G. H. Weddle, J. Kim et al., J. Phys. Chem. A, 103, 443 (1999); (b) P. Ayotte, G. H. Weddle, J. Kim and M. A. Johnson, Chem. Phys., 239, 485 (1998). 9. M. Roeselova, M. Mucha, B. Schmidt and P. Jungwirth, J. Phys. Chem. A, 106, 12 229 (2002). 10. (a) W. H. Thompson and J. T. Hynes, J. Am. Chem. Soc., 122, 6278 (2000); (b) O. M. Cabarcos, C. J. Weinheimer, J. M. Lisy and S. S. Xantheas, J. Chem. Phys., 110, 5 (1999). 11. (a) D. Borgis and A. Staib, J. Chem. Phys., 104, 4776 (1996); (b) P. Chaudhury, R. Saha and S. P. Bhattacharyya, Chem. Phys., 270, 277 (2001); (c) S. Roszak, M. Kowal, R. W. Gora and J. Leszczynski, J. Chem. Phys., 115, 3469 (2001). 12. (a) D. Majumdar, J. Kim and K. S. Kim, J. Chem. Phys., 112, 101 (2000); (b) J. Kim, H. M. Lee, S. B. Suh et al., J. Chem. Phys., 113, 5259 (2000); (c) H. M. Lee and K. S. Kim, J. Chem. Phys., 114, 4461 (2001). 13. (a) L. Lehr, M. T. Zanni, C. Frischkorn et al., Science, 284, 635 (1999); (b) A. V. Davis, M. T. Zanni, R. Weinkauf and D. M. Neumark, Chem. Phys. Lett., 353, 455 (2002); (c) D. E. Szpunar, K. E. Kammrath, A. E. Faulhaber and D. M. Neumark, J. Chem. Phys., 124, 054318 (2006).
Charge Transfer in Excited States: ab initio Molecular Dynamics Simulations
639
14. Basis sets were obtained from the Extensible Computational Chemistry Environmental Basis Set Database, developed and distributed by the Molecular Science Computing Facility, Environmental and Molecular Sciences Laboratory, Pacific Northwest Laboratory, PO Box 999, Richland, WA 99352 ( https://bse.pnl.gov/bse/portal). 15. (a) M. Kolaski, H. M. Lee, C. Pak et al., J. Phys. Chem. A, 109, 9419 (2005); (b) M. Kolaski, H. M. Lee, C. Pak and K. S. Kim, J. Am. Chem. Soc., 130, 103 (2008). 16. (a) H. M. Lee, S. Lee and K. S. Kim, J. Chem. Phys., 119, 187 (2003); (b) H. M. Lee, S. B. Suh, P. Tarakeshwar and K. S. Kim, J. Chem. Phys., 122, 044309 (2005). 17. C. L. Thomson, D. Madsen, S. R. Keiding et al., J. Chem. Phys., 110, 3453 (1999). 18. H. Okabe, Photochemistry of Small Molecules. Wiley, New York, NY (1978). 19. R. A. Crowell and D. M. Bartels, J. Phys. Chem., 100, 17940 (1996). 20. (a) P. Han and D. M. Bartels, J. Phys. Chem., 96, 4899 (1992); (b) C. G. Elles, I. A. Shkrob, R. A. Crowell and S. E. Bradforth, J. Chem. Phys., 126, 164 503 (2007). 21. E. J. Hart and M. Anbar, The Hydrated Electron. Wiley-Interscience, New York, NY (1971). 22. A. J. Swallow, Radiation Chemistry. An Introduction. Longman, London, UK (1973). 23. K. J. Hartig and N. Getoff, J. Photochem., 18, 29 (1982). 24. (a) D. N. Nikogosyan, A. A. Oraevsky and V. I. Rupasov, Chem. Phys., 77, 131 (1983); (b) V. Birkedal, E. S. Y. Madsen, C. Petersen et al., J. Chem. Phys., 328, 119 (2006). 25. T. Helgaker, E. Uggerud, H. J. A. Jensen, Chem. Phys. Lett., 173, 145 (1990). 26. A. Kumar, M. Kolaski, H. M. Lee and K. S. Kim, J. Phys. Chem. A, 112, 5502 (2008). 27. (a) M. Eigen and L. D. Maeyer, Proc. R. Soc. Lond., Ser. A, 247, 505 (1958); (b) G. Zundel and H. Metzer, Z. Phys. Chem. (Munich), 58, 225 (1968); (c) I. Shin, M. Park, S. K. Min et al., J. Chem. Phys., 125, 234 305 (2006); (d) N. J. Singh, M. Park, S. K. Min et al., Angew. Chem. Int. Ed., 45, 3795 (2006); (e) J. C. Jiang, Y. S. Wang, H. C. Chang et al., J. Am. Chem. Soc., 122, 1398 (2000); (f) P. Tarakeshwar, H. S. Choi, K. S. Kim et al., J. Chem. Phys., 114, 4016 (2001); (g) J. M. Headrick, E. G. Diken, R. S. Walters et al., Science, 308, 1765 (2005). 28. (a) P. O. Wennberg, T. F. Hanisco, L. Jaegle et al., Science, 279, 49 (1998); (b) J. G. Anderson, Ann. Rev. Phys. Chem., 38, 489 (1987); (c) A. L. Sobolewski and W. Domcke, J. Phys. Chem. A, 111, 11 726 (2007); (d) H. Langer, N. L. Doltsinis and D. Marx, ChemPhysChem, 6, 1734 (2005). 29. (a) C. Tanner, C. Manca and S. Leutwyler, Science, 302, 1736 (2003); (b) O. F. Mohammed, D. Pines, J. Dreyer et al., Science, 310, 83–86 (2005); (c) A. Marcinek, J. Adamus, K. Huben et al., J. Am. Chem. Soc., 122, 437 (2000); (d) G. Niedner-Schatteburg and V. E. Bondybey, Chem. Rev., 100, 4059 (2000). 30. (a) E. M. Knipping, M. J. Lakin, K. L. Foster et al., Science, 288, 301 (2000); (b) D. Wei and D. R. Salahub, J. Chem. Phys., 101, 7633 (1994). 31. (a) T. P. Rakitzis, P. C. Samartzis, R. L. Toomes et al., Science, 300, 1936 (2003); (b) A. Brown, G. G. Balint-Kurti and O. S. Vasyutinskii, J. Phys. Chem. A, 108, 7790 (2004); (c) A. G. Smolin, O. S. Vasyutinskii, G. G Balint-Kurti and A. Brown, J. Phys. Chem. A, 110, 5371 (2006). 32. H. M. Lee, M. Kolaski and K. S. Kim, ChemPhysChem, 9, 567 (2008). 33. D. Bonhommeau, P. T. Lake, C. Le Quiniou et al., J. Chem. Phys., 126, 051104 (2007). 34. P. Nuernberger, G. Vogt, G. Gerber et al., J. Chem. Phys., 125, 044512 (2006). 35. P. Hobza, H. L. Selzle and E. W. Schlag, Chem. Rev., 94, 1767 (1994). 36. E. C. Lee, D. Kim, P. Jureeka et al., J. Phys. Chem. A, 111, 3446 (2007). 37. K. S. Kim, P. Tarakeshwar and J. Y. Lee, Chem. Rev., 100, 4145 (2000). 38. M. O. Sinnokrot and C. D. and Sherrill, J. Phys. Chem., 110, 10656 (2006). 39. A. Kumar, M. Kolaski and K. S. Kim, J. Chem. Phys., 128, 034304 (2008).
29 Competitive ESIPT in o-Hydroxy Carbonyl Compounds: Perturbation Through Solvent Modulation and Internal Torsion Sivaprasad Mitra Department of Chemistry, North-Eastern Hill University, Permanent Campus, Shillong 793022, India
29.1 Excited-State Proton Transfer: An Overview The transfer of a proton from one group to another is probably one of the most fundamental reactions in chemistry. Owing to the light mass of the proton, the transfer process usually occurs on an ultrafast timescale. As experimental techniques have advanced up to the femtosecond timescale, it has now been possible to study the different aspects of proton transfer (PT) processes that were previously inaccessible. The molecules undergoing excited-state proton transfer (ESPT) have attracted considerable attention in recent times owing to their potential application in chemistry and biology [1, 2]. In a typical ESPT reaction, AHþ þ B ! A þ BHþ , where A and B denote different molecular species for intermolecular (ESIerPT) or different sites of a single molecular species for intramolecular (ESIPT) proton transfer process (Figure 29.1). These processes depend strongly on the nature of the probe molecules and also the surroundings in the given experimental situation. In the following sections we give a brief review on different aspects of the ESPT reaction, followed by a detailed discussion of the ESIPT process in specific cases and its perturbation through solvent modulation as well as internal torsion. 29.1.1 General background In 1931, Weber [3] first reported the difference in acid–base equilibrium constant (Ka) of organic photoacids in the ground and excited states. In 1949, F€ orster [4] proposed the correct explanation for the observation that an
Hydrogen Bonding and Transfer in the Excited State, Volume I & II © 2011 John Wiley & Sons, Ltd. ISBN: 978-0-470-66677-7
Edited by Ke-Li Han and Guang-Jiu Zhao
642 Hydrogen Bonding and Transfer in the Excited State
Figure 29.1
Schematic diagram showing intermolecular (i) and intramolecular (ii) proton transfer
organic acid having a large pKa difference in the ground and excited states may undergo protonation and deprotonation in the excited state, thus initiating the field of ESIerPT. The excited prototropic species thus produced relaxes to the ground state, with a subsequent reverse process (deprotonation or reprotonation), completing a four-level proton transfer cycle, the so-called F€orster cycle (Figure 29.2). On the other hand, Weller [5] in 1955 first proposed the formation of a proton transferred isomer (Ia) in the excited state, to explain the unusually large Stokes-shifted fluorescence for intramolecularly hydrogen-bonded methyl salicylate (I) (Figure 29.3). He also showed that only UV fluorescence with a mirror-image relationship with its absorption counterpart appears when the phenolic proton of I is methylated. With this pioneering work, the field of ESIPT was studied extensively.
A* ΔH*
ΔEA
(AH+)* ΔEAH+
A ΔH
AH+
Figure 29.2
A schematic representation of the F€ orster cycle for the acid–base equilibrium AHþ $ A þ Hþ
OCH3
OCH3
C
C O
O
H
H O
O I
Figure 29.3
Ia
Tautomerization in methyl salicylate (I)
Competitive ESIPT in o-Hydroxy Carbonyl Compounds
643
29.1.2 Description of ESPT in terms of the potential energy (PE) diagram In most cases, ESPT reactions comprise the following steps: (a) excitation of the normal photoacid; (b) rapid, non-radiative isomerization for ESIPT (or, ionization in the case of ESIerPT) in the excited state; (c) radiative decay of the newly formed excited species; (d) reverse isomerization (or reprotonation) in the ground-state surface. Thus, the entire process is cyclic and consists of a double-minimum potential in both the ground and excited states. Depending on the stability of different species formed during the ESPT process, the nature of the double-minimum potential can vary from symmetric to quasi-symmetric to fully asymmetric, both in the ground- and excited-state surfaces [6, 7]. 29.1.2.1 Symmetric and Slightly Asymmetric Double Minima If the isomerized structure following proton transfer is exactly identical to the initial photoacid, then this process must be characterized by a symmetric double-minimum potential, as in the cases of 9-hydroxy phenalenone and tropolone (structures II and III) (Figure 29.4). In these cases, the proton-transferred products IIa and IIIa are exactly identical to II and III respectively, and there is no physical property that can monitor the isomerization in real time. However, in the case of 2-(20 -hydroxy phenyl) benzothiazole (IV), the isomerized product (IVa) is a physically different molecule and the rate can be monitored by the fluorescence spectrum. The potential energy diagram is an asymmetric double minimum assuming that both the chemical structures are local minima. The other examples of this type of molecule are V and VI (Figure 29.5). In the asymmetric case, two possibilities, namely the deepest minimum in (a) the same atom (common asymmetry) or (b) the opposite atom (reverse asymmetry) may occur (Figure 29.6).
Figure 29.4
Examples of systems undergoing ESIPT in a symmetric double-minimum potential
644 Hydrogen Bonding and Transfer in the Excited State
Figure 29.5 Examples of systems undergoing ESIPT in an asymmetric double-minimum potential
Figure 29.6 Three schematic possibilities of the potential energy curves for proton transfer in the ground and excited states
29.1.2.2 Strongly Asymmetric Double Minima There are numerous examples of ESIPT systems that do not possess substantial amplitude in either local well in the excited singlet surface above the lowest vibrational level. The classical example that shows strongly reverse asymmetry is the case of methyl salicylate (I). 29.1.3 Thermodynamics of excited-state proton transfer The direct thermodynamic approach to measuring the electron density distribution in different energy states is to measure the respective chemical acidity constants (Ka). Most studies use the general method proposed by
Competitive ESIPT in o-Hydroxy Carbonyl Compounds
645
F€orster [4], which involves a thermodynamic cycle (the F€orster cycle) (Figure 29.2) combining thermodynamic and spectroscopic data. If DH and DH are the respective enthalpies of proton dissociation in the ground and the excited states, then DH ¼ DG þ TDS
ð29:1Þ
DH * ¼ DG* þ TDS*
ð29:2Þ
If DS ffi DS , then DHDH * ¼ DGDG* ¼ RTðln Ka ln Ka* Þ where Ka and Ka* are the acid–base equilibrium constants in the ground and excited states respectively. On rearrangement of the above equation, we have DpKa ¼ pKa pK*a ¼
DHDH * 2:303RT
ð29:3Þ
From Figure 29.2 we can write DEAH þ þ DH * ¼ DH þ DEA or DHDH * ¼ DEAH þ DEA ¼ NA hcðn0 AH þ n0 A Þ Then we can write DpKa ¼ pKa pK*a ¼
NA hcDn0 2:303RT
ð29:4Þ
where, Dn0 is the frequency difference (in wave numbers) of the lowest absorption bands between the acid and the basic forms. This method, although very successful in interpreting the acid–base properties of several organic compounds, involves several assumptions and limitations [8–10]. A more direct use of spectroscopic data to determine the pKa values of the organic acids was proposed by Weller [11, 12]. In this method, the relative fluorescence quantum yields (wf) of the excited-state species are plotted against the pH of the solution. However, a major disadvantage is that this method requires prototropic equilibrium to be established in the excited state, and also makes some assumptions regarding the fluorescence lifetime of the excited state. 29.1.4 Kinetics of excited-state proton transfer Although proton transfer reactions seem to be very simple at first sight, the kinetics of these reactions often becomes complex with the involvement of solvent molecules during the transfer process. With certain approximations, a number of models have been proposed to account for such a fast process, such as (a) the
646 Hydrogen Bonding and Transfer in the Excited State
Eigen model [13], (b) the bond energy bond order (BEBO) model proposed by Johnston et al. [14], (c) the Agmon–Levine model [15], (d) the intersecting state model [16] and (e) even Marcus theory, which was originally developed to interpret the rates of electron transfer reactions [17]. Although detailed discussion of all these models is out beyond the scope of this review, interested readers are advised to go through the original papers mentioned above to get a complete picture of the developments in this fascinating field of research.
29.2 Excited-State Intramolecular Proton Transfer (ESIPT) Weller [5, 11, 12] was the first to propose an intramolecular proton transfer mechanism in the excited state to account for the unusually large Stokes shift in the fluorescence properties of the methyl salicylate parent molecule. Since this first observation and its subsequent explanation, the field of ESIPT has expanded rapidly owing to its potential application in different areas of science and technology [18–21] and has been extensively reviewed by several authors [22–25]. In this section we will briefly discuss some of the aspects of ESIPT reactions. ESIPT usually involves systems having six-membered rings with an intramolecular hydrogen bond. The proton moves in between two highly electronegative heteroatoms of the type –O–H. . .O, –N–H. . .O, –S–H. . .O, etc. The five-membered intramolecularly hydrogen-bonded ring systems also undergo ESIPT. Excitation with an ultrashort laser pulse causes rapid redistribution of charge within the molecule, resulting in ultrafast ESIPT occurring within a 1012 s (ps)–1015 s (fs) timescale. 29.2.1 Different types of ESIPT Kasha [26] was the first to describe different types of ESIPT systematically. In this section we will give a brief description of these types: (a)
The symmetrical intramolecular proton transfer process as described earlier is observed in the cases of tropolone, 9-hydroxy phenalenone etc. In these cases, the transfer rates are usually very high and explained in terms of the tunnelling mechanism. (b) The molecules possessing a five- or six-membered intramolecular hydrogen bond (including the donor as well as the acceptor groups) usually undergo highly asymmetric proton transfer to show unusually large Stokes-shifted fluorescence. Depending on the strength of the internal hydrogen bond, more than one species (tautomer, different rotamers as well as proton-dissociated anions, etc.) may generate in the solution to give solvent-dependent fluorescence spectra. This particular type of ESIPT is intrinsic in nature. There are numerous examples of this category in the literature, but the most important are 3-hydroxy flavones (VII), o-hydroxy benzaldehyde (VIII) (Figure 29.7) and methyl salicylate (I). (c) Some of the systems, which may not have an internal hydrogen bond in suitable form to undergo intrinsic ESIPT, can undergo concerted double-proton transfer in the dimer form. The intensity of the longwavelength tautomer emission in these cases is found to increase with increase in solute concentration or with lowering of experimental temperature. The examples include 7-azaindole (IX) and N-phenyl benzamide (X) (Figure 29.7) [27]. Another type of ESIPT process, which may be termed solvent assisted intramolecular proton transfer, involves relay of the proton from one part of the molecule to the other, mediated by solvent bridges. Examples are 7-hydroxy quinoline (with two methanol molecules) and 3-hydroxy xanthone (with three methanol molecules). Some of the recent literature reports the importance of this type of ESIPT phenomenon, particularly in biological systems [28].
Competitive ESIPT in o-Hydroxy Carbonyl Compounds
647
Figure 29.7 Structures of the compounds 3-hydroxy flavone (VII), o-hydroxy benzaldehyde (VIII), 7-azaindole (IX) and N-phenyl benzamide (X)
29.2.2 Mechanism of ESIPT: tunnelling or vibrational relaxation? ESIPT reactions are characterized by a double-well potential energy surface having the reactant on one side and the product (phototautomer) on the other. Depending on the barrier height between these stationary points, ESIPT reactions may proceed either as proton tunnelling or as part of an intramolecular vibrational redistribution (IVR) in a barrierless adiabatic potential. Indication of proton tunnelling can be observed by the kinetic isotope effect (KIE) during the proton transfer process [29]. The extent of KIE dependence for the proton tunnelling is limited to certain factors. These can be either the degree to which the reaction is nonadiabatic and characterized by tunnelling through the potential barrier or if the reaction occurs by means of IVR, then the role of vibrational motions other than the O–H stretch becomes important. There is even precedence for the ESIPT process not to exhibit the KIE [30]. In these cases, excitation of the initial molecules causes a vibationally hot state, from which intramolecular relaxation caused by the change in geometrical parameters leads to the formation of the product. Hynes et al. [31] have presented a theory of proton transfer in both the adiabatic and the non-adiabatic limit. The theory assumes that three coordinates play the key role during ESIPT; the coordinate for the proton motion itself, the intramolecular separation of the two heavy atoms between which the proton is transferred and a collective solvent coordinate. Here, the electrons are always treated adiabatically, and the proton transfer process is considered to be adiabatic or non-adiabatic depending on the separation of two heavy atoms. For typical O–H. . .O proton transfer reactions, an O. . .O distance greater than 2.7 A causes the wave function for the proton to be localized about one of the oxygen atoms and a large barrier to exist that proton transfer must overcome through a non-adiabatic tunnelling process. Here, the rate of tunnelling is modulated by the O. . .O separation and the solvent fluctuations. If, however, the separation is less than 2.7 A, then the barrier of the proton transfer is greatly decreased and IVR becomes dominant over tunnelling in the PT mechanism. 29.2.3 Some examples of ESIPT As an active area of contemporary research over the last three decades, a large number of molecules have been found to undergo ESIPT that are so complex and so diverse that detailed discussion of all of them is beyond the scope of this review. Here we give a brief description of some of the well-studied examples of
648 Hydrogen Bonding and Transfer in the Excited State
ESIPT systems that have found scientific and technological applications. However, the most important cases of ESIPT in o-hydroxy carbonyl compounds and its perturbation are described separately and not included in this section. 29.2.3.1 3-Hydroxy Flavones 3-Hydroxy flavones (VII) are one of the most interesting and widely studied systems among ESIPT molecules. Since the first observation of ESIPT in 3HF by Sengupta and Kasha [32], numerous studies have been devoted to unravelling the intricacies of ESIPT in these molecules. In highly purified hydrocarbon solvents, VII undergoes very fast ESIPT at room temperature to give green fluorescence emission of the tautomer (lmax 520 nm), with a rise time of about 250 fs (k 1012 s1). The shift of the tautomer band with respect to the absorption onset is about 8500 cm1. The fluorescence lifetime of the excited tautomer is approximately 4 ns. After fluorescence, ground-state recovery of the normal structure occurs within 40–60 fs [33, 34]. The OCCOH moiety of VII forms a five-membered ring and thus has very weak internal hydrogen bond strength. This bond can be easily broken by interacting solvents such as water, ether and/or alcohol. Detailed discussion on the ESIPT of VII in different environments is available in the literature [35]. 29.2.3.2 2-(20 -Hydroxy Phenyl) Benzothiazole Cohen and Flavian [36] first proposed the large Stokes-shifted fluorescence for 2-(20 -hydroxy phenyl) benzothiazole (IV) towards the formation of proton-transferred tautomer. Since then, a number of studies have been devoted to this system [37]. It was suggested that the ground-state stable conformer of IV had the proton predominantly on the phenolic oxygen and the excited state had the proton on the nitrogen atom. The rate of ESIPT was found to be very high (6 1012 s1), faster than vibrational and torsional relaxation. Again, the absence of a significant isotope effect indicates that the proton transfer is essentially a barrierless process rather than tunnelling [38]. 29.2.4 Perturbation to ESIPT Ultrafast excited-state proton transfer can be perturbed in many ways. The most common causes of perturbation arise from (i) solvent interaction, (ii) the structure of the molecule undergoing ESIPT and (iii) competition with other possible non-reactive deactivation channels. In this section we will describe all these points in brief, with some representative examples. 29.2.4.1 Solvent Perturbation As ESIPT is a direct manifestation of a preformed hydrogen bond between donor and acceptor atoms bridging through the transferring proton, any solvent that ruptures this bond can cause perturbation to ESIPT. Purely intrinsic ESIPT can be achieved only in solutions of analytically pure samples with dried hydrocarbon solvents. The time dependence of ESIPT dynamics has frequently been found to vary systematically with the nature of the solvent [39, 40]. In general, the ESIPT rate decreases to a large extent in hydrogen-bonding solvents owing to the formation of intermolecular hydrogen bonding with the solvent and consequent increase in the barrier height for proton transfer. Hochstrasser et al. [41] showed a good correlation of the proton transfer rate in 2(20 -hydroxy-50 -t-octyl phenyl) benzotriazole (HOPB) with a longitudinal relaxation time (tL) of the 1-alkanol solvent series. These
Competitive ESIPT in o-Hydroxy Carbonyl Compounds
649
data are particularly intriguing because, in spite of this dependence on tL, no time-dependent Stokes shift is observed in this system. The ESIPT time of 240 fs in the case of 3HF in hydrocarbon solvents changes to more than 10 ps in hydrogen-bonding alcohol solvents. Petrich and coworkers [42] reported that the ESIPT time in hypericin, a naturally occurring polycyclic quinone that has received interest for its ability to deactivate the HIV virus, is about 10 ps, which is independent of the nature of the solvent. They argued that the O–H. . .O intramolecular hydrogen bonding in hypericin is much stronger than any potential hydrogen bonding in solution. Sytnik et al. [43] reported that 4-hydroxy-5-azaphenanthrene (HAP) exhibits ESIPTwhich appears to be unaffected by hydrogen-bonding solvent perturbation, and fluorescence emission is found to appear from the proton-transferred form even in ethanol. 29.2.4.2 Structural Perturbation The presence of only an internal hydrogen bond is not sufficient for ESIPT to occur. The nature of the product formed in the excited state, tautomer emission, quantum yield and PT rate are strongly dependent upon the structure of the molecule concerned. Nagaoka et al. [44] showed that ESIPT in VIII occurs only in the S1(p) state but not in the S0 and S2(p) states. They explained this behaviour of OHBA and other related molecules (Figure 29.8) such as methyl salicylate (I), o-hydroxy benzophenone (XI) and 2-(20 -hydroxy phenyl) benzothiazole (IV) in terms of the nodal pattern of the wave function in different states. The idea of the nodal plane dependence of ESIPT has also been extended to other systems. For example, methyl-3-hydroxy 2-naphthoate (XIII) is found to show ESIPT in the S1(pp ) state; however, phenyl-1-hydroxy-2-naphthoate does not show any proton transfer [45]. Furthermore, in the case of 2,5-bis(2-benzoxazolyl) hydroquinone (XIV), only one proton is transferred in the excited state, although this molecule has two intramolecular hydrogen-bonded sites [46]. 29.2.4.3 Competition with Other Non-Radiative Processes The low quantum yield of proton-transferred tautomer emission as observed in most of the ESIPT cases is mainly due to the competition among different deactivating channels other than proton transfer. A number of processes may be involved, such as intersystem crossing (ISC) from the initially excited state of the normal
Figure 29.8 Structures of the compounds o-hydroxy benzophenone (XI), methyl-3-hydroxy-2-naphthoate (XII), phenyl-1-hydroxy-2-naphthoate (XIII) and 2,5-bis(2-benzoxazolyl) hydroquinone (XIV)
650 Hydrogen Bonding and Transfer in the Excited State
conformer to the triplet, twisted intramolecular charge transfer (TICT), the formation of different rotamers in the excited state, etc., which results in multiple fluorescence bands. Among all these processes, competition of ESIPT with TICT processes was studied in great detail by Kasha and coworkers [47]. 29.2.5 Applications of ESIPT As an active area of contemporary research for the last three decades, ESIPT has found a large number of applications in the scientific and technological arena. Prototype examples are the development of laser dyes [48], molecular scintillators [49], special Raman filters [50], switches for dye laser pulse shortening [51], fluorescence probes for biomolecules [52], the conversion of solar into electrical energy [53] and the design of new bistable units in optical digital memory storage devices [54]. Again, as most of the important ESIPT reactions occur in solutions and are restricted to the subpicosecond timescale, ESIPT probes are widely used for novel formulations of solvation dynamics [55].
29.3 ESIPT in o-Hydroxy Carbonyl Compounds 29.3.1 o-Hydroxy benzaldehyde and its derivatives o-Hydroxy benzaldehyde (VIII) or salicylaldehyde is the simplest compound with six-membered internal hydrogen bonding in the ground state that undergoes intrinsic ESIPT. A detailed study of the spectroscopic properties of VIII and its derivatives was conducted by Nagaoka and coworkers [44]. The absorption spectra of the closed form of VIII covers the range 375–280 nm in non-polar solvents. The large Stokes-shifted fluorescence emission due to the formation of OHBA tautomer arises with a maximum at ca 510 nm. Nagaoka and Nagashima [44c] also performed ab initio calculations on different states of VIII and explained the cause of ESIPT only in the S1(p) state in terms of the nodal pattern of the wave function. The observed behaviour concerning proton transfer in other hydrogen-bonded molecules was explained similarly. However, in polar protic solvents such as ethanol, VIII gives dual fluorescence, one in the 520 nm region and the other at 420 nm, arising from different ground-state species, as confirmed by excitation spectra corresponding to these two emission bands, indicating the existence of two ground-state conformers. The emission peak at 420 nm is assigned to an intermolecularly solvent-mediated hydrogen-bonded open conformer (Figure 29.9). 29.3.2 Symmetrically substituted o-hydroxy carbonyl compounds In spite of all the observations for classic ESIPT systems such as o-hydroxy benzaldehyde, o-hydroxy acetophenone, o-hydroxy propiophenone and methyl salicylate described in the previous sections, there remains some questions regarding the nature of the fluorescent species, external control of the final photoproduct, etc. For the past few years we have been making a systematic effort to understand some of these questions in symmetrically substituted o-hydroxy carbonyl compounds. The representative structure of the systems investigated is given in Figure 29.10. 29.3.2.1 ESIPT and Solvent Participation in Excited-State Photophysical Processes The absorption spectra of XV in non-polar hydrocarbon solvents show a single broad band with a maxima at 350 nm due to pp (S1 S0) transition from the normal enol structure. On the other hand, the fluorescence spectrum exhibits a large Stokes-shifted emission in the 440–680 nm regions, with a
Competitive ESIPT in o-Hydroxy Carbonyl Compounds
651
Figure 29.9 Possible structural forms during proton transfer of o-hydroxy benzaldehyde (VIII) and its derivatives
maximum at 535 nm corresponding to the proton-transferred keto structure. However, the emission spectra for XV at 77 K consist of an additional band in the 460–470 nm region owing to the formation of open conformers (Figure 29.11), along with 535 nm tautomer emission. Both these open conformer structures can give rise to 3np -type phosphorescence in non-polar solvents because this spectrum is characterized by a progression of the C¼O stretching mode in other benzaldehyde types of molecule [56]. This indicates that the formation of an open conformer competes effectively with proton transfer in XV at 77 K in non-polar solvents. However, in moderately interactive solvents such as 1,4-dioxane and tetrahydrofuran, the open conformer emission arises even at room temperature. The nature of the excitation spectra for 535 and 460 nm emissions are different and appear at 360 and 420 nm respectively, which conclusively points to the formation of different conformers responsible for these emission bands. The steady-state spectral properties at different temperatures, along with the fluorescence lifetime of the excited species, are summarized in Table 29.1; the structures of the corresponding species having specific emission properties are also given in Figure 29.11. However, the solvent-dependent fluorescence spectra of XVI [57] indicate that the properties of intramolecularly hydrogen-bonded compounds change considerably, depending upon the substitution in the vicinity of intramolecularly hydrogen-bonded sites (Table 29.2). While the formyl group in XV rotates freely to give 3np -type phosphorescence at lower temperature, this is not observed in XVI owing to the restricted rotation of the acetyl group. The temperature and added electron donor act in a similar way on the decay properties of XVI, as in XV. The activation energy for the non-radiative processes is dependent on the group attached to the carbonyl carbon and indicates the importance of carbonyl torsion in the non-radiative decay path. The proton transfer dynamics of XVI is mainly controlled by the proton-accepting ability or basicity of the solvent, but orientational motion of the solvent molecule has a minor effect on it [57c]. The ESIPT dynamics of XV and XVI was studied in detail by femtosecond transient experiments as well as by semi-empirical and ab initio level of theory [58]. Ultrafast
652 Hydrogen Bonding and Transfer in the Excited State CH3
CH3
H
H C O
O
H3C
CH3
C
C
O
O
C O
XV
XVI CH3
CH3
OH
HO C O
O H
H
O
H3CO
OCH3
C
C
O
O
C O
XVII
O H
H XVIII
Figure 29.10 Structure of the compounds having a symmetrically substituted proton transfer site: 4-methyl-2, 6-diformyl phenol (XV), 4-methyl-2,6-diacetyl phenol (XVI), 3-methyl-6-hydroxy-m-phthalic acid (XVII) and 4-methyl-2,6-dicarbomethoxy phenol (XVII)
ESIPT occurs at about 200 fs in XV, followed by an IVR component of about 2.8 ps before undergoing fluorescence decay from the excited-state proton-transferred keto structure (Figure 29.12). The corresponding time constants for XVI are in the range of 250 fs and 1.5 ps respectively. The construction of the potential energy surface (PES) for ESIPT, as well as the estimation of different vibrational levels and the corresponding eigenfunctions supported at the ground and excited PESs using the Fourier grid Hamiltonian (FGH) method, showed that Franck–Condon excitation from the S0 state would take the system almost over the barrier and eventually into the potential well representing the tautomeric form; however, ESIPT through tunnelling could not be ruled out completely [58a]. In fact, FGH-based complex coordinate rotation calculation estimates a tunnelling rate constant (ktunnel) 1011 s1, which is almost an order of magnitude slower than the experimentally measured rate constant, indicating the importance of the over-barrier mechanism in these systems. On the other hand, the ESIPT systems XVII and XVIII are symmetrically substituted analogues of salicylic acid and methyl salicylate respectively. These systems can exist in two different hydrogen-bonded conformers in the ground state (Figure 29.13); one of these is capable of undergoing ESIPT, leading to tautomer emission, and the other will show normal emission. Extensive studies of these systems indeed indicate that there exist at least two conformers in the ground state, whereas in the excited state at least three species are formed. Moreover, the species responsible for tautomer and open conformer emission originate directly from the respective ground-state structures. Further, as observed in the other cases, the non-radiative decay process dominates again in the excited-state photophysics [59].
Competitive ESIPT in o-Hydroxy Carbonyl Compounds
Figure 29.11
653
Different structural forms of XV produced in the excited state
29.3.2.2 Modulation of Excited-State Phenomena by Solvent Dielectric Interaction As ESIPT involves movement of a hydrogen atom over a very small distance, in most cases the solvent dielectric does not play a major role in determining the excited-state phenomena. However, judicious substitution in and around the PT site can have a detrimental effect on the fluorescence properties of these systems. A recent example of this type of system, where the ESIPT phenomenon is modulated by internal torsion as well as by the solvent dielectric, has been the case of substituted o-hydroxy acetophenone derivatives (Figure 29.14) [60]. The absorption spectrum of both XIX and XX in non-polar solvent shows a single peak at around 355 nm. However, the fluorescence emission spectrum corresponding to this absorption consists of two bands. The high-energy, structured emission at around 430 nm has less Stokes shift (DnSS 5000 cm1) compared with the broad, low-energy, unstructured emission at 505 nm which has unusually large Stokes shift (DnSS 8300 cm1). The excitation spectra corresponding to both these emissions match very closely the ground-state absorption spectra and appear at about 375 and 360 nm respectively. In accordance with previous studies of ESIPT in o-hydroxy acetophenone derivatives, the largely Stokes-shifted emission arises owing to proton
654 Hydrogen Bonding and Transfer in the Excited State Table 29.1 Summary of absorption (labs), emission (lem) and excitation (lexc) spectral maxima, along with the fluorescence quantum yield (wf) and excited-state lifetime (tf), of different ground- and excited-state species of XV (for structure, see Figure 29.11) Solvent
labs(nm)
Cyclohexane CCl4 1,4-Dioxane
350 350 350
THF Acetone Acetonitrile
350 350 350
DMF
350 460d 350 480 350 350
DMSO Methanol Water
a
lem(nm)
lexc(nm)
wf
tf (ns)
535 535 535 460c 460 460 460 520d 460 520 460 520 520 520
360 360 360 400 400 430 400 440 400 460 400 480 440 440
0.10 0.12 0.18 0.20 0.10 0.20 0.24 0.30 0.24 0.32 0.40 0.48 0.35 0.26
1.2 1.3 1.4 4.2 3.9 2.2 4.7 4.9 3.9 4.1 4.6 5.0 4.5 4.8
b
Corresponding to structure XV. Corresponding to structure XVa. Corresponding to structure XVc. d Corresponding to a proton-dissociated anionic structure. a
b c
Table 29.2 Summary of fluorescence quantum yield (wf) and excited-state lifetime (tf) of XVI (for structure, see Figure 29.11) in different solvents at room temperature (298 K) and 77 K Solvent
3-Methyl pentane CCl4 DMSO DMF Acetonitrile
Fluorescence yield (wf)
Fluorescence lifetime (tf,ns)
298 K
77 K
298 K
77 K
0.04 0.06 0.28 0.21 0.08
0.32 0.34 0.82 0.76 0.18
0.40 0.40 4.3 3.3 0.3
4.1 4.5 6.8 5.1 3.2
transfer (ESIPT fluorescence) in the S1(pp ) state of the primary enol (E) form, where the intramolecular hydrogen bond involves the phenolic hydrogen and carbonyl oxygen of the acetyl group in the ortho position (Figure 29.15, structure (a)) to form the keto (K) conformer (structure (b)). On the other hand, the structured emission with a lower DnSS value may arise from another conformer of the enol structure where the hydrogen bonding partners involve the phenolic hydrogen and oxygen of the nitro group in the ortho position. As shown in Figure 29.15, two structural forms can be presumed for this conformer (E1 and E2, structures (c) and (d) respectively). However, from the geometry of these two conformers, it can be seen that in E1 the proximity of oxygen atoms in neighbouring acetyl and phenolic groups may cause additional non-bonding interaction. Therefore, the relative abundance of E2 should be greater in the solution compared with E1. The results of theoretical calculation indeed predict the same. On the other hand, the relative energy difference, as obtained from DFT calculation for XX, between E and E2 structures varies from 0.5 to 0.02 kcal mol1 in the different solvents studied here. The very small energy difference in solution indicates that there is equilibrium among
Competitive ESIPT in o-Hydroxy Carbonyl Compounds
655
Figure 29.12 Schematic views of ESIPT phenomena and corresponding time parameters for XV, as obtained by femtosecond transient absorption (TrA) experiments. The shaded arrows indicate the origin of TrA at different time delays (adapted from Ref. 58c)
different enol conformers in the ground state, and ESIPT emission from E is in competition with normal fluorescence from E2. The calculated excited-state energy difference between E and E2 is 5.6 kcal mol1 in CCl4 and decreases substantially to 3.5 kcal mol1 in polar solvents such as ethanol or acetonitrile. The large energy difference in non-polar solvents ensures that ground-state equilibrium between these structures is perturbed in the excited state, and respective emission appears from individual conformers. However, the situation is different in polar solvents and demands careful attention. The interaction of excited fluorophore and solvent dipole reduces the energy difference substantially. Furthermore, ESIPT is known to be a very efficient non-radiative deactivation channel for intramolecularly hydrogen-bonded compounds and occurs on an ultrafast timescale. Also, the keto
CH3
CH3
HO
OH C O
O
HO
O
C
C
O
O
C O H
H
Figure 29.13
O
Possible hydrogen-bonded structural forms of XVII
H
656 Hydrogen Bonding and Transfer in the Excited State
Figure 29.14 phenol (XX)
(a)
Structural representation of 2-acetyl-4-methyl-6-nitrophenol (XIX) and 2-acetyl-4-chloro-6-nitro-
O
(b)
O
O
N
O N
O
O H
H
O
O
C
R
C
R CH3
(c)
O
CH3
(d)
O N
O
O N
H
H O
O
CH3
O R
C
CH3
R
C
O
Figure 29.15 Structure of possible conformers of XIX (R ¼ CH3). The enol (a), keto (b) and non-ESIPT enol structures (c) and (d) are represented as E, K, E1 and E2 in the text (adapted from Ref. 60)
Competitive ESIPT in o-Hydroxy Carbonyl Compounds
657
Fluo. Intensity (arb. units)
250
200
8 150
1
100
50
0 425
450
475
500
525
550
575
600
Wavelength /nm
Figure 29.16 Fluorescence emission spectra of XX with increasing concentration of b-cyclodextrin in aqueous solution. [b-CD] (mM) ¼ 1 (0.0), 2 (2.2), 3 (3.6), 4 (4.9), 5 (6.8), 6 (9.0), 7 (17.6) and 8 (22.2) respectively (adapted from Ref. 61a)
conformer (K) is the most stable geometrical entity in the excited state. Therefore, it was argued that ESIPT acts both as a thermodynamic and as a kinetic sink in the excited-state potential energy surface, and the equilibrium is tilted heavily towards the pro-ESIPT conformer (E) in these solvents for both XIX and XX. As a result, the fluorescence spectra show only ESIPT emission at around 510 nm in these solvents. The argument concerning the difference in photophysical properties of XX in polar solvents to those in non-polar solvents like CCl4 was further substantiated by observing the fluorescence spectra of a probe encapsulated in cyclodextrin nanocavities. As the probe experiences a relatively more hydrophobic environment inside the host cavity, it is expected that the ESIPT fluorescence of XX in pure water will be changed to both ESIPT and non-ESIPT fluorescence under encapsulation with cyclodextrin (CD). As shown in Figure 29.16, this is indeed the case in the presence of CDs [61].
29.4 Concluding Remarks In this chapter, different aspects of ESIPT phenomena have been reviewed, with a special emphasis on the nature of participation of solvent molecules to cause specific perturbations in intrinsic ESIPT. Examples were presented to emphasize the most common cause of perturbation in ESIPT through direct participation of solvent to rupture the internal hydrogen bond; however, thermodynamically favourable internal torsion and solvent dielectrics can also play a major role in determining the nature and number of fluorescing species arising from these systems. Some of the concepts are well established; others are more speculative, and a welldocumented conclusion is still lacking. However, the present vigorous research activity in the field of ESIPT with ultrafast laser experiments, as well as high-level theoretical calculation, will undoubtedly unravel many more examples of this type and the associated mechanisms. The present results address fundamental issues of competitive ESIPT and solvent-modulated excited-state photophysics from structurally similar ground-state
658 Hydrogen Bonding and Transfer in the Excited State
conformers resulting from torsional motion. These motions are very important for the construction, stability and function of macromolecular architecture.
Acknowledgements Thanks are due to Professors S. Mukherjee and S. P. Bhattacharyya, both from the Department of Physical Chemistry, the Indian Association for the Cultivation of Science, Jadavpur, Calcutta, and to Professor N. Tamai, Department of Chemistry, Kwansei Gakuin University, School of Science, Sanda, Japan, for their guidance, help and constant enthusiasm on this topic. The author also expresses heartfelt thanks to all those who have collaborated on this problem at different times. Partial financial support from the University Grants Commission (UGC), Government of India, on ‘Spectroscopy and Dynamics of Photoinduced Processes in Homogeneous and Biomimetic Environments’ (34-299/2008 (SR)) is gratefully acknowledged.
References 1. (a) R. P. Bell, The Proton in Chemistry, 2nd edition. Chapman and Hall, London (1973). (b) J. T. Hynes, J. P. Klinman, H.-H. Limbach and R. L. Schowen (eds), Hydrogen Transfer Reactions. Wiley-VCH Verlag GmbH & Co. KGaA (2007). 2. (a) R. A. Copeland and S. I. Chan, Ann. Rev. Phys. Chem., 40, 671 (1989); (b) I. P. Pozdnyakov, A. Pigliucci et al., J. Phys. Org. Chem., 22, 449 (2009). 3. K. Weber, Z. Phys. Chem. B, 15, 18 (1931). 4. T. F€orster, Naturwiss, 36, 186 (1949). Z. Electrochem., 54, 43 (1950); Pure Appl. Chem., 24, 443 (1970). 5. A. Weller, Naturwiss, 42, 175 (1955). 6. P. F. Barbara, P. K. Walsh and L. E. Brus, J. Phys. Chem., 93, 29 (1989). 7. S. J. Formosinho and L. G. Arnaut, J. Photochem. Photobiol. A: Chem., 75, 21 (1993). 8. H. H. Jaffe and H. Jones, J. Org. Chem., 30, 964 (1965). 9. T. C. Warner and D. M. Hercules, J. Phys. Chem., 73, 2005 (1969). 10. Z. R. Grabowski and A. Grabowska, Z. Phys. Chem., 101, 197 (1976). 11. A. Weller, Prog. React. Kinetics, 1, 189 (1961). 12. A. Weller, Z. Electrochem., 56, 662 (1952); 61, 956 (1957). 13. M. Eigen, Angew. Chem. Int. Ed. in Eng., 3, 1 (1964); Z. Physik. Chem., 1, 176 (1954). 14. (a) H. S. Johnston, Adv. Chem. Phys., 3, 131 (1960); (b) H. S. Johnston and P. Goldfinger, J. Chem. Phys., 37, 700 (1962); (c) H. S. Johnston and C. Parr, J. Am. Chem. Soc., 85, 2544 (1963). 15. (a) N. Agmon and R. D. Levine, Chem. Phys. Lett., 52, 197 (1977); J. Chem. Phys., 71, 3034 (1979). 16. A. J. C. Varandas and S. J. Formosinho, J. Chem. Soc. Chem. Commun., 163 (1986); J. Chem. Soc., Faraday Trans. II, 82, 953 (1986). 17. (a) S. J. Formosinho, Pure Appl. Chem., 58, 163 (1986); (b) H. D. Burrows and S. J. Formosinho, J. Chem. Soc., Faraday Trans. II, 82, 1563 (1986); (c) L. G. Arnaut and S. J. Formosinho, J. Photochem. Photobiol. A: Chem., 75, 1 (1993). 18. (a) D. Zhong, A. Douhal and A. H. Zewail, Proc. Natl Acad. Sci. USA, 97, 14 056 (2000); (b) S. Park, O.-H. Kwon, Y.-S. Lee et al., J. Phys. Chem. A, 111, 9649 (2007) and references cited therein. 19. P. Miskovosky, Int. J. Photoenergy, 4, 45 (2002). 20. R. C. Haddon and F. H. Stillinger, In: Molecular Electronic Devices, ed. by F. L. Carter. Marcel Dekker, New York, NY (1987). 21. O. P. Bondar, V. G. Pivovarenko and E. S. Rowe, Biochim. Biophys. Acta, 1369 (1998). 22. P. F. Barbara and H. P. Trommsdorff (eds.), Special issue on Spectroscopy and Dynamics of Elementary Proton Transfer in Polyatomic Systems. Chem. Phys., 136, 153 (1989). 23. M. Kasha Fetschrift, J. Phys. Chem., 95, 10 215 (1991).
Competitive ESIPT in o-Hydroxy Carbonyl Compounds
659
24. For a review on the biologically relevant excited-state double-proton transfer process, see P. T. Chou, J. Chin. Chem. Soc., 48, 651 (2001). 25. S. M. Ormson and R. G. Brown, Prog. Reac. Kinet., 19, 45 (1994). 26. (a) M. Kasha, J. Chem. Soc., Faraday Trans. II, 82, 2379 (1986); (b) J. Heldt, G. Gormin and M. Kasha, Chem. Phys., 136, 321 (1989). 27. (a) C. A. Taylor, M. A. El-Bayoumi and M. Kasha, Proc. Natl Acad. Sci. USA, 63, 253 (1969); (b) G.-Q. Tang, J. McInnis and M. Kasha, J. Am. Chem. Soc., 109, 2531 (1987). 28. (a) E. J. de Berker, J. D. Geerlings and C. A. G. O. Varma, J. Phys. Chem. A, 104, 5916 (2000); (b) M. K. Nayak and P. Wan, Photochem. Photobiol. Sci., 7, 1544 (2008). 29. (a) T. Carrington and W. H. Miller, J. Chem. Phys., 84, 4364 (1986); (b) N. Shida, P. F. Barbara and J. Alml€ of, J. Chem. Phys., 92, 4061 (1989); (c) M. Kijack, Y. Nosenko, A. Singh et al., J. Am. Chem. Soc., 129, 2738 (2007). 30. (a) W. Frey, F. Laermer and T. Elsaesser, J. Phys. Chem., 95, 10391 (1991); (b) B. J. Schwartz, L. A. Peteanu and C. B. Harris, J. Phys. Chem., 96, 3591 (1992); (c) S. Lochbrunner, K. Stock, C. Schriever and E. Riedle, Ultrafast phenomena XIV, Part VI. Springer Ser. Chem. Phys., 79, 491 (2005). 31. (a) J. T. Hynes, Ann. Rev. Phys. Chem., 36, 573 (1985); (b) D. Bogris and J. T. Hynes, J. Chem. Phys., 94, 3619 (1991); (c) H. Azzouz and D. Bogris, J. Chem. Phys., 98, 7361 (1993). 32. P. K. Sengupta and M. Kasha, Chem. Phys. Lett., 68, 382 (1979). 33. (a) P. M. Felker, W. R. Lambert and A. H. Zewail, J. Chem. Phys., 77, 1603 (1982); (b) D. McMorrow and M. Kasha, Proc. Natl Acad. Sci. USA, 81, 3375 (1984); (c) P. T. Chou, M. L. Martinez and J. H. Clements, J. Phys. Chem., 97, 2618 (1993); (d) K.-Y. Chen, Y.-M. Cheng, C.-H. Lai et al., J. Am. Chem. Soc., 129, 4534 (2007). 34. (a) M. Itoh and Y. Fujiwara, J. Phys. Chem., 87, 4558 (1983); (b) M. Itoh, Y. Tanimoto and K. Tokomura, J. Am. Chem. Soc., 104, 4164 (1982); (c) R. de Vivie-Riedle, V. D. Waele, L. Kurtz and E. Riedle, J. Phys. Chem. A, 107, 10591 (2003). 35. (a) P.-T. Chou, C.-H. Huang, S.-C. Pu et al., J. Phys. Chem. A, 108, 6452 (2004); (b) A. N. Bader, F. Ariese and C. Gooijer, J. Phys. Chem. A, 106, 2844 (2002); (c) S. Ameer-Beg, S. M. Ormson, X. Poteau et al., J. Phys. Chem. A, 108, 6938 (2004). 36. M. D. Cohen and S. J. Flavian, J. Chem. Soc. B, 321 (1967). 37. (a) W. Al-Soufi, K. H. Grellmann and B. Nickel, Chem. Phys. Lett., 174, 609 (1990); (b) F. Laermer, T. Elsaesser and W. Kaiser, Chem. Phys. Lett., 148, 119 (1988); (c) S. Lochbrunner, A. J. Wurzer and E. Riedle, J. Phys. Chem. A, 107, 10580 (2003). 38. P. F. Barbara, P. M. Rentzepis and L. E. Brus, J. Am. Chem. Soc., 102, 2786 (1980). 39. (a) A. U. Acun˜a, F. Amat-Guerri, J. Catalan and F. G. Tables, J. Phys. Chem., 84, 629 (1980); (b) M. Itoh, Y. Fujiwara, M. Sumitani and K. Yoshihara, J. Phys. Chem., 90, 5672 (1986); (c) S. Nagaoka, M. Fujita, T. Takemura and H. Baba, Chem. Phys. Lett., 123, 489 (1986). 40. (a) O. K. Abou-Zied, R. Jimenez, E. H. Z. Thompson et al., J. Phys. Chem. A, 106, 3665 (2002); (b) S. J. Schmidtke, D. F. Underwood and D. A. Blank, J. Am. Chem. Soc., 126, 8620 (2004). 41. Y. R. Kim, J. T. Yardley and R. M. Hochstrasser, Chem. Phys., 136, 311 (1989). 42. F. Gai, M. J. Fehr and J. W. Petrich, J. Phys. Chem., 98, 8352 (1994). 43. (a) A. Sytnik and J. C. D. Valle, J. Phys. Chem., 99, 13028 (1995); (b) A. Sytnik and M. Kasha, Proc. Natl Acad. Sci. USA, 91, 8627 (1994). 44. (a) S. Nagaoka, U. Nagashima, N. Ohta et al., J. Phys. Chem., 92, 166 (1988); (b) S. Nagaoka, K. Sawada, Y. Fukumoto et al., J. Phys. Chem., 96, 6663 (1992); (c) S. Nagaoka and U. Nagashima, Chem. Phys., 136, 153 (1989). 45. G. J. Woolfe and P. J. Thistlethwaite, J. Am. Chem. Soc., 103, 3849 (1981). 46. A. Mordzin´ski, A. Grabowska, W. Kuˆhnle and A. Kro´wezyn´ski, Chem. Phys. Lett., 101, 291 (1983). 47. (a) J. Heldt, D. Gormin and M. Kasha, Chem. Phys., 136, 321 (1989); (b) C. A. S. Potter, R. G. Brown, F. Vollmer and W. Rettig, J. Chem. Soc., Faraday Trans. I, 90, 59 (1994). 48. (a) A. U. Khan and M. Kasha, Proc. Natl Acad. Sci. USA, 80, 1767 (1983); (b) P. T. Chou, D. McMorrow, T. J. Aartsmma and M. Kasha, J. Phys. Chem., 88, 4596 (1984); (c) A. U. Acun˜a, A. Costela and J. M. Mun˜oz, J. Phys. Chem., 90, 2807 (1986). 49. A. Sytnik and M. Kasha, Radiat. Phys. Chem., 41, 331 (1993). 50. P. T. Chou, S. L. Studer and M. L. Martinez, Appl. Spectr., 145, 513 (1991).
660 Hydrogen Bonding and Transfer in the Excited State 51. N. P. Ernsting and B. Nikolaus, Appl. Phys. B, 39, 155 (1986). 52. (a) A. Sytnik, D. Gormin and M. Kasha, Proc. Natl Acad. Sci. USA, 91, 11698 (1994); (b) M. Gutman and D. Huppert, J. Am. Chem. Soc., 103, 3709 (1981). 53. F. Vollmer and W. Rettig, J. Photochem. Photobiol. A: Chem., 95, 143 (1996). 54. R. W. Munn, Chem. Br., 517 (1984). 55. (a) M. Maroncelli, E. W. Castner, B. Bagchi and G. R. Fleming, Faraday Disc. Chem. Soc., 85, 199 (1988); (b) P. J. Rossky and J. D. Simon, Nature, 370, 263 (1994). 56. (a) S. Mitra, R. Das and S. Mukherjee, Chem. Phys. Lett., 202, 549 (1993); (b) S. Mitra, R. Das and S. Mukherjee, Spectrochim. Acta, 50A, 549 (1994); (c) S. Mitra, R. Das and S. Mukherjee, J. Photochem. Photobiol. A: Chem., 79, 49 (1994); (d) R. Das, S. Mitra and S. Mukherjee, J. Photochem. Photobiol. A: Chem., 76, 33 (1993); (e) R. Das, S. Mitra and S. Mukherjee, Bull. Chem. Soc. Jpn, 66, 2492 (1993). 57. (a) S. Mitra, R. Das and S. Mukherjee, Chem. Phys. Lett., 228, 393 (1994); (b) A. Mandal, D. Guha, R. Das et al., J. Chem. Phys., 114, 1336 (2001); (c) D. Guha, A. Mandal, R. Das et al., Isr. J. Chem., 39, 375 (1999). 58. (a) S. Mitra, R. Das, S. P. Bhattacharyya and S. Mukherjee, J. Phys. Chem. A, 101, 293 (1997); (b) S. Mitra and S. Mukherjee, J. Lumin., 118, 1 (2006); (c) S. Mitra, N. Tamai and S. Mukherjee, J. Photochem. Photobiol. A: Chem., 178, 76 (2006). 59. (a) R. Das, S. Mitra, D. Guha and S. Mukherjee, J. Lumin., 81, 61 (1999); (b) A. Mandal, S. Mitra, D. Banerjee et al., J. Chem. Phys., 118, 3154 (2003); (c) A. Mandal, D. N. Nath, S. Mukherjee et al., J. Chem. Phys., 117, 5280 (2002). 60. S. Mitra, T. S. Singh, A. Mandal and S. Mukherjee, Chem. Phys., 342, 309 (2007). 61. (a) S. Mitra, ISRAPS Bull., 20, 23 (2008); (b) S. Mitra and T. S. Singh, unpublished results.
30 Excited-State Double Hydrogen Bonding Induced by Charge Transfer in Isomeric Bifunctional Azaaromatic Compounds Dolores Reyman and Cristina Dı´ıaz-Oliva Departamento de Quı´mica-Fı´sica Aplicada, Facultad de Ciencias, Universidad Auto´noma de Madrid, Cantoblanco, 28049 Madrid, Spain
30.1 Introduction Proton transfer is one of the most important processes involved in both chemical reactions and biological systems. Consequently, a large number of studies on hydrogen bonding and proton transfer processes in ground and excited states have been carried out [1]. A special class within the systems presenting proton transfer is represented by heteroaromatic molecules containing both hydrogen-bonding donor and hydrogen-bonding acceptor groups. The reaction may occur both as an intramolecular or intermolecular process. Examples of the former are provided by numerous molecules such as salicylic acid and its derivatives [2–6], 2-hydroxybenzophenone [7], 5-hydroxy-flavone [8, 9], 1,5-dihydroxyxanthraquinone [10], 2-hydroxy-4,5-naphthotropone [11], 2-(20 -hydroxyphenyl)benzoxazole, 2-(20 -hydroxyphenyl)-benzothiazole [12–14] and 2-(20 hydroxy-50 -methylphenyl)benzotriazole [15–18]. Several reviews are based on these molecules [19–22]. Conversely, in another class of molecules, the possibility of an internal H-bond is hindered either for geometrical reasons or because of a large spatial separation between proton-donating and proton-accepting sites. The prototype of these molecules is exemplified by 7-azaindole (7AI), which represents the first documented example of dual-proton transfer in dimers and alcohol complexes [23]. However, in spite of decades of research on 7AI [23–44], the phototautomerization process continues to arouse much controversy. Based on time-resolved experiments, it was proposed that the transfer occurs in a stepwise fashion [25, 27, 45–51]. Some authors, however, postulate a concerted simultaneous double-proton transfer mechanism [52–57].
Hydrogen Bonding and Transfer in the Excited State, Volume I & II © 2011 John Wiley & Sons, Ltd. ISBN: 978-0-470-66677-7
Edited by Ke-Li Han and Guang-Jiu Zhao
662 Hydrogen Bonding and Transfer in the Excited State
7AI belongs to N-heteroaromatic compounds, which possess two or more nitrogen atoms in their structure, and are of great interest [58] owing to their being ubiquitous in nature and comprising the basic elements of most macromolecules of biological interest. Thus, some are components of DNA (such as purine bases), others are prototypes of photoprotectors (e.g. Tinuvin P) and others are parts of the active sites of all heme proteins involved in the biological functions of oxygen transport (hemoglobin), electron transport (e.g. cytochrome in respiration) and biocatalysis (e.g. enzymes such as catalases and peroxidases, and cytochrome P450). Usually, these heteroaromatic molecules can be structurally viewed as condensation products of simpler individual Nheteroaromatic rings, a p-deficient pyridine ring fused to a p-excessive indole ring. As a general rule, the acidity/ basicity of the nitrogen atoms in these condensed molecules greatly differs from that observed in the individual rings [59]. The changes in molecular structures and electronic charge distribution accompanying the annelation process can greatly affect the intrinsic acidity/basicity of these nitrogen centres [60]. These molecules possessing both hydrogen-bond donor and hydrogen-bond acceptor groups are able to form dimers or complexes with hydrogen-bonding solvents. In many cases, the excited-state properties of such intermolecular hydrogen-bonded species are totally different from those of non-bonded molecules in non-polar or polar aprotic solvents leading to tautomerization processes [19, 61–65]. The driving force for excited-state proton transfer is the electron density redistribution occurring upon excitation. This redistribution causes an increase in the excited-state pKa of the proton acceptor, a decrease in the excited-state pKa of the proton donor or a combination of both. This review focuses on analysing ground- and excited-state effects induced by hydrogen-bonding compounds in two series of isomeric bifunctional azaaromatic chromophores based on pyrrolo-quinoline, such as pyrido[3,2-g]indole (PQ), dipyrido[2,3-a:30 ,20 -i]carbazole (DPC), 7,8,9,10-tetrahydropyrido[2,3-a]carbazole (TPC) and pyrido[2,3-a]carbazole (PC), and on pyrido-indole, such as methylene-bridged 2-(20 -pyridyl) indoles (PyIn-n) and 9H-pyrido[3,4-b]indole (BC) (or norharmane), 1-methyl-9H-pyrido[3,4-b]indole (HN) (or harmane), N9-methyl-harmane (MHN) and N2-methylharmane (T-MeHN). In the betacarbolinic derivatives (BCD) the pyridine and pyrrol rings are fused, while in the others they are separated by a benzo ring spacer, such as in pyrrolo-quinolines, or by a single bond and a methylene bridge, such as in 2-(20 -pyridyl)indoles.
30.2 Pyrrolo-Quinoline Derivatives (PQ, DPC, TPC) 1-H-pyrrolo[3,2-h]quinoline (or pyrido[3,2-g]indole) (PQ) and related compounds (Chart 30.1), can be considered as the 7AI modified by a benzo ring spacer, which separates the pyrido and the pyrrolo rings. The geometry of these molecules does not enable an internal H-bond from the pyrrolo-H to the aza-N atom. PQ has been cited in numerous places in the literature in connection with potentially valuable chemical and biomedical applications. In chemical applications, some derivatives of 1-methyl-pyrrolo[3,2-h]quinoline have been shown to be good stabilizers for polymers [23]. The molecule PQ has also been proposed as a host molecule in molecular recognition ab initio studies of ring motions [66]. Biomedical reports propose its use as a potential anticancer drug [67]. It exhibits tuberculostatic activity [68] at 4 mg mL1 against tubercule strain H-37, and has also been tested as an antimalarial drug [69]. Furthermore, DPC has been considered as a probe of hydrophilic/hydrophobic surface character owing to its sensitivity towards hydroxilic groups [70]. It is thus of substantial interest to understand the electronic structural characteristics and consequent physical and chemical behaviour of these compounds in order to gain a full understanding of their properties [71]. 30.2.1 Steady-state measurements Absorption spectra for these compounds show a red-shift of all transitions upon increasing solvent polarity and a distinct band on the red energy side. The addition of small amounts of alcohol (c < 102 M) to a solution
Excited-State Double Hydrogen Bonding
N
N
N
N
PQ
N
N
N
H
H
663
H
DPC
TPC (CH2)n
N
N
N
N H
H PC
PyIn-n
N N
R2
R1 BC HN MHN
Chart 30.1
R1, R2 = H R1 = H, R2 = Me R1, R2 = Me
Formulae and acronyms of the compounds
of these compounds in n-hexane (non-polar solvent) leads to the appearance of isosbestic points, which are indicative of the equilibrium between two ground-state species (Figure 30.1). Initially, one of the two species was assigned to the alcohol complex and the other to the ‘bare’ molecule (uncomplexed) [72]. The equilibrium constant, K, for a reaction involving n alcohol molecules B þ nA ¼ BAn
K ¼ ½BAn =ð½B½An Þ
may be expressed as K¼
ðOD OD0 Þ ðOD¥ ODÞ ½An
ð30:1Þ
where OD0 and OD¥ denote the optical density measured when only the uncomplexed or complexed forms are present, and OD is the optical density measured at an alcohol concentration [A]. From the plot of ln [(OD OD0)/[(OD¥ OD)] versus ln[A], the number of alcohol molecules in a complex can be obtained. The values extracted from this plot point to the 1:1 stoichiometry at a low concentration of alcohols [72, 73] (see Figure 30.1). In emission, all the molecules considered reveal intense fluorescence both in non-polar and polar aprotic solvents [74]. The spectra consist of a single emission, labelled as F1, which has been assigned as the radiative decay of the initially excited state [71, 72]. The change from an aprotic to a protic solvent causes a decrease in the fluorescence intensities of this band, F1, and the appearance of a new red-shifted fluorescence band, labelled as F2 (Figure 30.1). As can be seen in Figure 30.2, the spectral locations and shapes of these
664 Hydrogen Bonding and Transfer in the Excited State
Figure 30.1 Titration of the solution of PQ in n-hexane with 1-butanol at 293 K. Top, changes in absorption; bottom, evolution of both fluorescence bands. The arrows show spectral changes accompanying the addition of alcohol. The alcohol concentration varied from 3.0 104 to 2.2 102 M. The insert in the top part shows the determination of the equilibrium constant and stoichiometry of the complex from the absorption data recorded at 28011 cm1. Reprinted with permission from [73]. Copyright 1999 American Chemical Society
low-energy emissions (F2) coincide almost exactly with the fluorescence recorded for tautomeric model structures (Chart 30.2) [73]. These findings leave no doubt that low-energy fluorescence originates in the tautomeric species, obtained as a result of a double-proton transfer reaction occurring in complexes of the photoexcited chromophore with alcohol. However, the fact that the decrease in the intensity of the F1 fluorescence was much greater than the concomitant increase in the tautomeric emission led the model that assumes the presence of only two species, and thus only one form of the alcohol solvate, to be considered as probably excessively simple [73]. This finding has been explained by the presence of at least two different forms of alcohol complexes. A cyclic, doubly hydrogen-bonded 1:1 complex that can undergo phototautomerization, and a non-cyclic 1:n complex (n being the number of alcohol molecules associated with the substrate) that is also efficiently deactivated, but by a pathway that is different from proton transfer. Another argument for the presence of 1:1 cyclic, doubly hydrogen-bonded solvates in the ground state is the fact that the F2 band can be still observed in rigid glasses at 77 K for PQ and DPC [74, 75]. Lowering the temperature leads to an increase in both F1 and F2 emission intensities, where F1 is the most affected. For instance, the quantum yield for DPC, in n-butanol, increases by an order of magnitude between 293 and 183 K (from 0.0005 to 0.005), while the corresponding values for F2 are 0.002 and 0.007. Moreover, the F1 emission becomes structured and very similar to the emission observed at 293 K in non-polar solvents. However, in spite of the recovery of the fluorescence intensities, the sum of F1 and F2 quantum yields in alcohols at low
Excited-State Double Hydrogen Bonding
665
Figure 30.2 Fluorescence of PQ, TPC, PC and PyIn-2 (solid lines, a) and of the corresponding tautomeric model structures (dotted lines, b). The spectra were recorded in 1-butanol at 293 K. Reprinted with permission from [73]. Copyright 1999 American Chemical Society
temperature is still much lower than the fluorescence quantum yield at 293 K in non-polar or polar aprotic solvents [72]. For these compounds, phosphorescence was also detected, and was located between the F1 and the F2 fluorescence bands. The excitation spectra taken at 77 K showed that the ‘normal’ F1 emission and phosphorescence have the same precursor, whereas the tautomeric emission occurs from another type of complex [72] (contrary to the behaviour observed at room temperature, where the fluorescence excitation spectra of F1 and F2 coincide with each other and with the absorption spectrum). Once more, these findings show that, in the ground state, at least two different forms of the solvates exist, only one of which is capable of undergoing phototautomerization at low temperatures (see Scheme 30.1). The F2 emission has been attributed to the cyclic 1:1 complexes, and the high-energy fluorescence and phosphorescence to ‘incorrectly solvated’ species, or ‘open’ form, with only one intermolecular hydrogen bond, which requires structural rearrangement prior to tautomerization. Such rearrangement should obviously be a function of viscosity and temperature, and therefore should be blocked at low temperatures. The behaviour of these compounds in alcohols differs from that observed in 7AI and 1-azacarbazole (1AC). In the latter two molecules, lowering the temperature causes the disappearance of the tautomeric emission band and the recovery of the radiative properties (a strong
666 Hydrogen Bonding and Transfer in the Excited State
N
N
N
N
Me
Me T-MePQ
T-MeTPC (CH2)n
N
N
N
N
Me
Me
T-MePC
T-MePyIn-n
N Me N Me T-MeHN
Chart 30.2
Chemical models for the tautomers
increase in F1 fluorescence). This has been interpreted as the DPC and PQ solvates being ‘better prepared’ for photoinduced tautomerization [72]. The driving force for excited-state proton transfer is the electron density redistribution occurring upon excitation. This redistribution leads to an increase in the excited-state pKa of the proton acceptor and a decrease in the excited-state pKa of the proton donor. Changes in pKa in the excited state have been obtained from the ‘F€ orster cycle’ [76, 77], employing the formula nA nB00 DpKa ¼ pK*a pKG a 0:00207 ~ 00 ~
ð30:2Þ
where pK*a and pKG a denote values in the excited and ground state respectively; the electronic transition energies in the acid ~nA nB00 forms are expressed in wave numbers. Thus, for PQ it has been shown 00 and the base ~ that, upon excitation, the pyridine nitrogen becomes much more basic (DpKa ¼ þ9.6), and the acidity of the N–H group is also strongly enhanced (DpKa ¼ 6) [73]. Similar results have also been predicted by calculation using as reactivity indices [78, 79] the effective valence electron potentials [79–82]. 30.2.2 Time-resolved kinetics studies Subsequently, time-resolved kinetics studies led to the detection of a faster non-radiative process. Photophysical parameters for PQ, DPC and TPC measured at room temperature are set out in Tables 30.1 to 30.3. The time dependence of the F1 and F2 band emissions was measured, and the behaviour of the two bands was found to be quite different [83]. For instance, for detection within the F1 band, the fluorescence transient of PQ consisted of an instantaneous rise followed by a decay on the picosecond timescale. The transients were fitted to a triexponential decay function. Transients detected within the F2 emission band exhibited initially a
Excited-State Double Hydrogen Bonding
N
N
N
H H O
N
N
O H O R
H
H O
O
R
N
N
H
H
O R H
R
N H
ESDPT
H
H O R
R
N
N
R
IC
667
H
H O R
F2 h~
F'1
h~
F1''
h~
F1'
F1''
h~
N
N BDPT
H
H O R
N
N
N
H O R
O H O R
N
N
H
H
O R H
N H
H
H R
O
H O
N
N
H
H O R
R
R
Scheme 30.1 Equilibria between various types of complex for PQ, and related compounds, with protic solvents, and excited-state deactivation channels. IC: internal conversion; ESDPT: excited-state double-proton transfer; BDPT: back double-proton transfer
biexponential rise, with time constants equal to the decay constant of the F1 band, followed by a decay of several hundred picoseconds. Evidently, the time t1 and t2 are somehow related to the proton transfer process. On the other hand, the slower decay component of F1 had no counterpart in the F2 rise. The discovery of these two time parameters (t1 and t2) have made it possible to establish that the double-proton transfer process for PQ and related compounds does not occur in a simple one-step or two-step mechanism. Both mechanisms would give rise to a single decay step for the initially excited species and thus lead to a single exponential decay of the F1 band emission. These findings supported the proposal of two distinct solute–solvent species emitting the ‘normal’ F1 fluorescence. One of these is capable of undergoing tautomerization (the cyclic complex), while the other is a ‘blocked’ complex that can suffer both tautomerization and radiationless deactivation (the non-cyclic one). In other words, the excited complex, which was already cyclic in the ground state, avoids the region of rapid deactivation and undergoes tautomerization at a rate approximately an order of magnitude faster than internal conversion. However, for the non-cyclic species, an efficient radiationless deactivation channel was located on the excited-state energy surface, along the path leading from non-cyclic to cyclic species. The comparison of the photophysical parameters obtained for the phototautomer with those of its N-methylated chemical model provided more arguments for the presence of two kinds of alcohol complex deactivated via different channels (see Table 30.4). For instance, for PQ, in spite of the room temperature, the
668 Hydrogen Bonding and Transfer in the Excited State Table 30.1 Photophysical parameters for PQ in various solvents, measured at room temperature: absorption (~ nabs ) and fluorescence (~ nfl ) maxima, fluorescence quantum yields (wfl ) and decay times (t) [83]. Pre-exponential factors in parentheses n~abs (cm1)a
Solvent Methanol
d
Ethanold
30 100
1-Propanold 1-Butanola
29 700
Decanold H2Ob,e,h ACN þ H2Od,f,i Et2O þ H2Od,g n-Hexanek DMSOk ACNb Butyronitrilek Pyridinek
28 900 30 500
F1 F2 F1 F2 F1 F2 F1 F2 F1 F2 F1 F2 F1 F2 F1 F2
n~fl (cm1)b
wfl b;c
25 300 16 000 25 500d 17 200d 25 800 16 250 25 850 16 250
0.0003 0.0008
23 300 15 900 25 200 16 000 26 000 16 000 27 300a
0.00085 0.0003 0.0025 0.00055 0.01 0.001 0.25k 0.23k 0.16 0.16 0.014
25 850
0.00035 0.0011 0.0004 0.0014
t1 (ps) 0.6 0.6 0.7 0.7 0.7 0.7
(0.14) (0.31) (0.19) (0.62) (0.16) (0.46)
0.9 (0.37) 0.9 (0.66)
60e,f,g 290e,f C¼O group and alcohols or amines, the frequencies of the stretching modes for both the O--H or N--H bonds as well as the C¼O bonds are shifted to the lower energy (or frequency) region and the amount of frequency shift with respect to that of the stretching mode of the free molecule (or uncomplexed) is indicative of the strength of the hydrogen bond formed between the solute and the solvent [51, 53]. C102 has been considered as an ideal probe molecule to study hydrogen bond dynamics, since the C¼O group is the only site to serve as the acceptor of hydrogen bond from alcohols, phenols and amines [51, 53, 72, 74, 97]. The dynamics of the hydrogen-bonded complex between C102 and the hydrogen-bond donating phenol have been investigated both theoretically by Zhao et al. [72, 73] and experimentally by Elsaesser, Nibbering and coworkers [51a,b]. The stretching mode of the free carbonyl (C¼O) group of C102 dissolved in non-hydrogen bonding solvent, C2Cl4, appears at around 1735 cm1 (Figure 33.7a). Addition of phenol leads to a downshift of the band to 1695 cm1. This redshift by about 40 cm1 was assigned to the formation of a strong C¼O H--O hydrogen bond between the C¼O group and an O--H group of phenol, with a binding energy on the order of 60 kJ mol1. Formation of the C¼O H--O hydrogen bond strongly affects the O--H stretching bands of the phenol molecules. The spectrum of phenol in C2Cl4 displays a stretching band at 3610 cm1, which has been assigned to free O--H groups, not being part of a hydrogen bond. However, a broader band in the 3450–3550 cm1 region, which arises predominantly due to 1 : 1 phenol–phenol complexes, because of formation of intermolecular O H--O hydrogen bond between two phenol molecules appears, too
770 Hydrogen Bonding and Transfer in the Excited State
Figure 33.7 (a) Ground-state C¼O stretching bands of C102 in pure C2Cl4 (dashed line) and of C102–(phenol)n complexes in C2Cl4 (solid line). The C102 and phenol concentrations were 5 and 40 mM, respectively. (b) Steady-state O–H stretching bands of phenol dissolved in C2Cl4 (concentration 30 mM). (c) Ground-state infrared spectrum of C102–(phenol)n complexes [51a]. Adapted with permission from [53]. Copyright 2003 American Chemical Society
(Figure 33.7b). Addition of C102 leads to the formation of an additional broad band that is strongly redshifted to around 3380 cm1. This band is assigned to the O--H group in the C¼O H--O hydrogen bonds (Figure 33.7c). We have studied the hydrogen-bonding interaction between C102 and aniline [53]. Curve “a” in Figure 33.8 shows the steady state FTIR spectrum of a solution of C102 (1.5 102 mol dm3) in tetrachloroethylene (TCE). It shows a very strong absorption band having the maximum at ca 1738 cm1 due to the stretching vibration of the free C¼O group. The vibrational spectrum of C102 recorded in neat aniline has an absorption maximum due to C¼O group at 1698 cm1 (curve b). The appearance of the new absorption band indicates the formation of an association complex between C102 and aniline via formation of a hydrogen bond between the C¼O group of C102 and the H--N group of aniline (C¼O H--N). For several reasons (vide infra), fluorenone is another ideal probe molecule for studying the dynamics of hydrogen-bonded complex and has been investigated both theoretically and experimentally [21, 55, 59, 98, 99]. Tominaga and coworkers have recently studied the hydrogen-bonding interaction between fluorenone and primary alcohols (Figure 33.9) [59]. Figure 33.10 (A) below shows the IR absorption spectrum of fluorenone in cyclohexane, a non-hydrogen bonding solvent. The CO stretching mode of fluorenone shows a sharp band with a
Ultrafast Dynamics of the Excited States of Hydrogen-Bonded Complexes and Solvation
1.5
771
a
Absorbance
b 1.0
0.5 1680
1700
1720
1740
-1 Wavenumber (cm )
Figure 33.8 Steady state FTIR spectra of C102 (1.5 102 mol dm3) in tetrachloroethylene (TCE) (curve a) and in neat aniline (curve b) [53]. Reprinted with permission from [59]. Copyright 2007 Elsevier
peak wavenumber of 1725 cm1 and a full-width at half maximum (fwhm) of 4.5 cm1. This figure also shows the IR absorption spectrum of fluorenone in 1-octanol. The spectrum shows multiple bands with two maxima at around 1713 cm1 (a) and 1721 cm1 (b), and a shoulder at around 1700 cm1 (c). The oxygen atom has two lone pairs, which may act as hydrogen-bonding sites to form two possible hydrogen-bonding complexes with one and two methanol molecules. Optimization of geometries and normal mode coordinates using density functional theory provided the frequency of the CO stretching mode of the free fluorenone (FL) as 1784 cm1, and those for the hydrogen-bonded complexes formed with methanol, i.e., FL : MeOH and FL : (MeOH)2, appear at 1761 and 1736 cm1, respectively (Figure 33.9). By correlating this with the fact that the peak frequency of the mode shifts to a lower frequency with increasing number of hydrogen bonds, the observed bands in the IR spectrum have been assigned to different complexes of fluorenone and solvent molecules, that is, the (a), (b) and (c) bands correspond to free fluorenone, a fluorenone complex with one alcohol molecule, and a complex with two alcohol molecules, respectively. Figure 33.10(B) displays the temperature dependence of the IR spectrum of fluorenone in 1-octanol from 293 to 323 K. The relative intensities of the three bands at 1713, 1721 and 1700 cm1 depend on the temperature. This indicates that the three bands result from three different species in equilibrium and not from an intramolecular effect, such as Fermi resonance, and therefore support the spectral assignment.
Figure 33.9
Structures of fluorenone–methanol complexes
772 Hydrogen Bonding and Transfer in the Excited State
Figure 33.10 (A) Absorption spectra of fluorenone in cyclohexane and in 1-octanol. The concentrations are 25 mM, and the optical path length is 0.5 mm. (a) Free fluorenone, (b) fluorenone complex with one alcohol molecule and (c) complex with two alcohol molecules. (B) Temperature dependence of the absorption spectrum of 9-fluorenone in 1-octanol [59]. Reprinted with permission from [59]. Copyright 2007 Elsevier (See Plate 39)
33.3 Vibrational Dynamics of the C¼O Stretching Mode of Fluorenone In hydrogen-bonded complexes, various dynamical properties, such as reactivity and energy relaxation, are strongly influenced by intramolecular as well as intermolecular hydrogen bonds [24–29, 59]. The effects of the OH stretching mode of alcohols or water on vibrational energy relaxation have been investigated by timeresolved infrared (IR) spectroscopy extensively; the vibrational energy relaxation of the OH stretching mode is accelerated by more than one order of magnitude [24, 25, 100–105]. To understand the effects of hydrogen bonds on the vibrational dynamics such as vibrational frequency fluctuations or excitation energy transfer, Tominaga and coworkers have investigated the vibrational energy relaxation of the CO stretching mode of the fluorenone molecule in alcohol solvents by sub-picosecond IR pump–probe spectroscopy [59]. IR pump–IR probe spectroscopy measures transient changes in the vibrational populations, as schematically shown in Figure 33.11. It involves the use of two ultrashort pulses of IR light; the pump pulse causes a reduction in the ground state population and an increase in the excited state emission, while the probe pulse monitors the evolution of these changes in population as a function of time (i.e., the delay between the two pulses).
Ultrafast Dynamics of the Excited States of Hydrogen-Bonded Complexes and Solvation
773
Figure 33.11 Schematic depiction of the single color pump–probe spectroscopy. The pump produces a ground state bleach and excited state emission. For systems with large anharmonicities (D), no excited state absorption is observed. The “molecule” in the excited state and the “hole” in the ground state result in the amplification (from stimulated emission) and reduced absorption of the probe pulse, respectively. As the excited state decays, the absorption of the probe increases, resulting in a reduced pump–probe signal. Reprinted with permission from [59]. Copyright 2007 Elsevier (See Plate 40)
Figure 33.12(a) displays the IR pump–IR probe signal of fluorenone in cyclohexane at 1727 cm1 which is close to the peak wavenumber. The signal corresponds to the recovery of the ground state bleach and the decay of the stimulated emission. The signal shows a sharp spike at around t ¼ 0 ps, owing to a coherent artifact, which persists up to about 0.3 ps. Hence the pump–probe signals beyond 0.3 ps delay-time were fitted with a monoexponential decay with a time constant of 4.07 0.07 ps. Lim and Hochstrasser have studied the CO stretch of methyl acetate in carbon tetrachloride and found that the vibrational energy relaxation of the CO stretching mode is biexponential with time constants of 0.23 and 8.2 ps [106]. In this study, it was not possible to resolve the ultrafast component. A pump–probe signal at 1723 cm1, which is close to the peak of the CO stretching vibration of free fluorenone (Figure 33.12b), decays exponentially with a time constant of 4.3 0.1 ps. The time constant is similar to that in cyclohexane. Additionally, Figure 33.12(c) displays a pump–probe signal at 1712 cm1, close to the peak of the CO stretching vibration of the 1 : 1 complex of fuorenone–1-octanol. The signal decays exponentially with a time constant of 1.6 0.1 ps. Figure 33.13 displays the time-resolved differential absorption spectra of fluorenone in 1-octanol constructed at the delay-times of 0.45, 1.05 and 5.10 ps. Considering the three bands in the static IR spectrum of fluorenone in 1-octanol, which have been attributed to free fluorenone, the 1 : 1 complex and the 1 : 2 complex, it becomes evident that more than one species gives rise to signal intensities in the IR spectrum. The fact that
774 Hydrogen Bonding and Transfer in the Excited State
Figure 33.12 IR pump–probe signals of 9-fluorenone in (a) cyclohexane at 1727 cm1, (b) 1-octanol at 1723 cm1 and (c) 1-octanol at 1712 cm1. Reprinted with permission from [54a]. Copyright 1993 Elsevier (See Plate 41)
the signals at both 1723 and 1712 cm1 can be reproduced well with a single exponential decay suggests that at these wavenumbers a single species, that is, the free fluorenone and 1 : 1 complex, respectively, is the dominant contributor to the signal. However, the positive absorption band in the 1680–1710 cm1 region may have a contribution from all three species. Hence, assuming that the decay time constant of the 1 : 2 complex is faster than those of the free fluorenone and 1 : 1 complex, each spectral component was extracted using a global fitting analysis with the decay time constants 4.7 0.1, 2.3 0.1 and 0.27 0.02 ps. Therefore, the fast component may be due to a mixture of the coherent artifact and the 1 : 2 complex. The major factors determining the vibrational relaxation rate are the vibrational density of states, strength of couplings with overtone of the bending modes, and energy gap [107, 108]. Formation of a hydrogen bond could
Ultrafast Dynamics of the Excited States of Hydrogen-Bonded Complexes and Solvation
775
Figure 33.13 Time-resolved differential spectra obtained at 0.45, 1.05 and 5.10 ps of fluorenone in 1-octanol. The green and purple lines are at 1712 and 1723 cm1, corresponding to the peak wavenumbers of the 1 : 1 complex and the free fluorenone, respectively [59]. Adapted with permission from [55]. Copyright 2005 American Chemical Society
influence all these factors. Since the energy shift due to formation of a hydrogen bond in this case is about only 10 cm1, it is unlikely that the energy gap change is a major factor for the increased rate of vibrational relaxation time in the hydrogen-bonded complex. However, the increase in density of states may play an important role in the vibrational relaxation process. Theoretical calculation using density functional theory reveals the presence of intermolecular vibrational modes with frequencies of 15, 37 and 49 cm1 for the fluorine–methanol complex and these modes may be accepting modes in the relaxation process. Another important point to note is that the time constants of the free fluorenone and the 1 : 1 complex have different values. This suggests that interconversion between the two conformers is not faster than vibrational relaxation of the two complexes.
33.4 Dynamics of the Excited States of Hydrogen-Bonded Complex 33.4.1 Electronic spectroscopy Despite the widespread importance of hydrogen bonding in both biological and non-biological systems, surprisingly little was known about the dynamics of hydrogen-bond formation and breakage until Berg and coworkers introduced a new approach that allowed the direct, time-resolved measurement of the lifetime of
776 Hydrogen Bonding and Transfer in the Excited State
Figure 33.14 Chemical structure of resorufin. Adapted with permission from [55]. Copyright 2005 American Chemical Society
hydrogen bonds between various solvents and resorufin (Figure 33.14), a hydrogen-bond accepting dye molecule [54]. This paper reported the first systematic study of the hydrogen-bond breaking and forming process. The results demonstrated that the rate of bond breaking was strongly affected by solvent dynamics and led to a multistep model for the bond-breaking process. This report has inspired the examination of several chemical systems to study the dynamics of hydrogen bonds using visible pump–probe spectroscopy. 33.4.1.1 Resorufin–Alcohols (Ref. [54]) Comparison of the ground state electronic absorption spectra of resorufin in a polar aprotic solvent, say propylene carbonate, and in protic solvents of varying hydrogen-bond donating ability has revealed the formation of two distinct kinds of hydrogen-bonded complexes, having absorption maxima at 583.5 and 592 nm, which have been assigned to two different forms of the molecule, namely “B” (blue) and “R” (red), respectively. These two forms remain in equilibrium in solution. In weaker hydrogen bond-donating solvents, say ethanol, the equilibrium shifts in favor of the B form and in a very strong hydrogen-bond donating solvent (TFE) only the B form is seen. However, the fluorescence spectra are only weakly affected [54a]. The interconversion of R and B forms has been studied using time-resolved measurements of the absorption change following photoexcitation of resorufin (Figure 33.15). In the polar aprotic solvent, DMSO, the transient spectra recorded at 1.5 and 300 ps are virtually identical (Figure 33.15a), suggesting that the polar solvation of resorufin is very weak, because of a small change in dipole moment on excitation, between 0.2 and 0.4 D. However, following photoexcitation of resorufin in ethanol using 577 nm light, the short-wavelength portion of the transient spectrum does not change, indicating that the absorption bleach does not change. However, the excited-state emission on the long-wavelength side of the spectrum undergoes a shift to longer wavelengths with time. At 577 nm, the R and B forms are excited in approximately the proportion found in the ground state at equilibrium. Thus, there is little rearrangement of the ground state populations after excitation. In the excited state, on the other hand, the populations are out of equilibrium. The B form converts into R to restore equilibrium, and the change in the emission frequency is seen in the transient spectrum. The existence of an isosbestic point indicates that the relaxation is between two distinct species, and is inconsistent with the continuous shift that results from polar solvation. Spectral evolution represents the conversion of the B form into the R form. Figure 33.15(c) shows the transient spectra after excitation at 590 nm; the spectral evolution suggests the conversion of the R form into the B form. Excitation at 590 nm will bleach primarily the R form out of the ground state, disturbing the ground state equilibrium. As in the ground state B converts into R to restore the equilibrium, and the frequency of the absorption bleach should shift from the frequency of R to the frequency of B. Assuming that the process of interconversion between the R and B forms follows first-order kinetics, the equilibration time in ethanol excited at 577 nm, where the excited-state dynamics predominate (Figure 33.15b), or that in ethanol excited at 590 nm, where ground state dynamics predominate (Figure 33.15c), is nearly the same and the ground-state equilibration rate constant in ethanol has been determined to be 42 ps. The equilibrium constant in the ground state has been estimated to be 1.86 and the hydrogen-bond formation time and the lifetime of hydrogen bond have been determined to be 65 and 120 ps, respectively.
Ultrafast Dynamics of the Excited States of Hydrogen-Bonded Complexes and Solvation
777
Figure 33.15 Transient spectra of resorufin solutions. (a) Spectra are the sum of absorption and fluorescence components (dotted). In the polar aprotic solvent DMSO there is little spectral change between 1.5 ps (solid) and 300 ps (dashed). In the hydrogen-bonding solvent ethanol [(b) excited at 577 nm, (c) excited at 590 nm], spectral evolution with an isosbestic point occurs: 1.5 ps (solid), 20 ps (long-dashed), 50 ps (dotted), 300 ps (short-dashed). The vertical lines indicate the integration regions used in determining rates [54a]. Adapted with permission from [55]. Copyright 2005 American Chemical Society
The lifetime is similar to the reported 250 ps lifetime of hydrogen bonds in ethanol oligomers isolated in a non-polar solvent [109], but is much longer than the subpicosecond lifetime reported in pure water [110]. The lack of a strong viscosity dependence of the equilibration time suggested that the bond formation/breaking event was a very local motion that did not involve motion of the center of mass of the solvent molecule or largescale rearrangement of the solvent hydrogen-bond network but, rather, that only a simple rotation around the O--C bond might be involved. 33.4.1.2 Fluorenone–Alcohols (Ref. [55]) Fluorenone has been considered to be an ideal molecule for studying the hydrogen bond dynamics for several reasons. First, the primary photophysics of fluorenone has been well studied [82–85]. In fluorenone, like all other molecules with an active C¼O group, hydrogen bonding and polarity are the key factors in controlling the pathways of energy dissipation following electronic excitation. Second, as discussed in Section 33.2, fluorenone forms intermolecular hydrogen-bonded complex with alcohols in the ground state and both the S1 and T1 states are dynamically quenched due to hydrogen-bonding interaction with the alcoholic solvent molecules [73, 85]. Third, fluorenone is a planar molecule with a rigid framework, and, hence, no other
778 Hydrogen Bonding and Transfer in the Excited State
relaxation process, such as conformational or configurational relaxation, than the relaxation process arising due to solvent motions is important in the S1 state. Fourth, the change of dipole moment (Dm ¼ 2.2 D) upon photoexcitation of fluorenone to its S1 state is not very large [43]. Hence, the difference in spectroscopic properties between the Franck–Condon (FC) state and the relaxed excited state following dipolar solvation is expected to not be very significant. However, we may expect a significant change in the spectroscopic properties of the excited state due to hydrogen-bonding interaction between the carbonyl group and the molecules of the protic solvents. Finally, since fluorenone is very weakly fluorescent, the transient absorption spectroscopic technique should be a very useful tool to reveal the role of hydrogen bond dynamics in the excited-state relaxation processes of fluorenone [17]. Figure 33.16 shows the time-resolved absorption spectra of the transient species formed upon photoexcitation of fluorenone in acetonitrile and DMSO using 400 nm laser pulses of 70 fs duration. In each of these solvents, the time-resolved transient absorption spectra recorded in the sub-5 ps time domain show two distinct absorption bands in the 470–530 and 530–700 nm regions. Within this time-domain, evolution of spectral characteristics is insignificant but a slight decrease of absorbance occurs in the 570–700 nm region. Each of the temporal profiles recorded in this wavelength region consists of an ultrafast decay component, followed by another very long-lived component, which arises as a residual absorption in sub-500 ps time domain and can be assigned to the S1 state. Two such typical temporal profiles recorded at 630 nm in acetonitrile and at 610 nm in DMSO are shown in the insets of Figure 33.16. The lifetimes of this short component are 1.4 and 2.1 ps in acetonitrile and DMSO, respectively, and have been assigned to the vibrational relaxation process. Acetonitrile
20
0.15 ps 0.8 ps 2 ps 10 ps 50 ps
15 5
630 nm τ1 (d)= 1.4 ps
10
Δ Absorbance (mOD)
5
τ2 (d) = long
0
0 0 15
5 10 Time, ps
15
6
610 nm τ1(d)= 2.1 ps
3
τ2(d) = long
0
10
DMSO 5
0 500
0.2 ps 2 ps 5 ps 10 ps 50 ps 100 ps
0
5
550 600 Wavelength (nm)
10
15
Time, ps
650
Figure 33.16 Time-resolved transient absorption spectra of fluorenone in acetonitrile and DMSO constructed for different delay times following photoexcitation with 400 nm laser pulses. Insets: temporal absorption profiles recorded at 630 and 610 nm following photoexcitation of fluorenone in acetonitrile and DMSO, respectively. Solid lines represent the best-fitted dual exponential functions. The lifetimes of the shorter component, which are given in the figure, are assigned to the vibrational energy relaxation process happening in the S1 state and the long component arises due to long lifetime (19 ns) of the S1 state of fluorenone in these solvents [55]. Reprinted with permission from [111]. Copyright 2005 American Chemical Society
Ultrafast Dynamics of the Excited States of Hydrogen-Bonded Complexes and Solvation
779
16 0.15 ps 0.4 ps 5 ps
12 Δ Absorbance (mOD)
15 ps 30 ps 100 ps a 8
b
1 4
0 500
550
600
650
0 500
550
600
650
Wavelength (nm)
Figure 33.17 Time-resolved transient absorption spectra of fluorenone in 1-propanol, constructed for different delay times following photoexcitation using 400 nm laser pulses. Inset: comparison of transient spectra constructed at 0.15 ps delay time following photoexcitation of fluorenone in acetonitrile (a) and 1-propanol (b) [55]. Reprinted with permission from [111]. Copyright 2005 American Chemical Society (See Plate 42)
The spectroscopic and dynamical features of the transient species have been investigated in normal alcohols as well as in ethylene glycol. The spectroscopic and the dynamical features of the transient species have been observed to be very similar in these solvents. As a typical example, Figure 33.17 shows time-resolved spectra of the transient species produced upon photoexcitation of fluorenone in 1-propanol using 400 nm laser pulses. The transient spectrum constructed for the 0.15 ps delay time, that is, immediately after photoexcitation, has features that are very similar to those of the transient spectrum recorded in acetonitrile (inset of Figure 33.17), although the relative intensities of two bands in these spectra are somewhat different. However, evolution of the transient spectra with increasing delay time is more significant as compared to that in acetonitrile, suggesting a stronger solute–solvent interaction. Despite the fact that 1-propanol (dielectric constant, « 21.1) is less polar than acetonitrile (« 37), on comparing the features of the time-resolved spectra recorded in acetonitrile and 1-propanol the significant difference in evolution of the spectral characteristics of the transient species in these two solvents cannot be assigned to the simple dipolar solvation but, obviously, to the hydrogen-bonding interaction between the excited state of fluorenone and the solvent. The temporal dynamics of transient absorption have been seen to be wavelength dependent, because of overlapping of absorption bands due to more than one differently hydrogen-bonded species (vide infra). In a series of normal alcoholic solvents, the viscosities of the solvents increase as the length of the linear hydrocarbon chain increases from methanol to 1-pentanol. The spectral and temporal characteristics of the transient species generated in these solvents as well as in ethylene glycol (EG) have been seen to be very similar to those
780 Hydrogen Bonding and Transfer in the Excited State
6
0.15 ps 0.4 ps 1 ps 2 ps 5 ps 10 ps 20 ps
(A)
4 a'
Δ Absorbance (mOD)
2
4
a 0
0
500
550
600
650 6
(B) 6
3
4
0
b' b
20 ps 40 ps 60 ps 100 ps
2
0
500
550
500
600
550
600
650
Wavelength (nm)
Figure 33.18 Time-resolved transient absorption spectra constructed for different delay times following photoexcitation of fluorenone in TFE using 400 nm laser pulses. (A) Time-resolved spectra in sub-20 ps time domain. Inset of (A): comparison of the transient spectra constructed for 0.15 ps delay-time in TFE (a) and in 1-propanol (a0 ). (B) Time-resolved spectra in post-20 ps time-domain. Inset of (B): comparison of the transient spectra constructed for 100 ps delay-time in TFE (b) and in 1-propanol (b0 ) [55]. Reprinted with permission from [111]. Copyright 2005 American Chemical Society (See Plate 43)
in 1-propanol. Hence the dynamics of solute–solvent interaction is not controlled by the viscosity of the solvents. Relaxation dynamics of the S1 state of fluorenone in TFE have also been investigated (Figure 33.18). The features of the spectrum recorded at 0.15 ps are very similar to those of the spectra recorded in acetonitrile and 1-propanol at the same delay-time. Following the spectral evolution in sub-20 ps time domain, the transient spectrum constructed at 20 ps delay-time is seen to consist of an ESA band with maximum at 550 nm and a shoulder at 510 nm. At longer delay times, beyond 20 ps, this entire band decays without any further evolution of the spectral features. The features of the transient spectrum recorded at 20 ps delay time have been compared to those of the transient spectrum recorded at 100 ps delay time in 1-propanol (inset of Figure 33.18B). This shows that while the transient spectrum recorded in 1-propanol consists of two ESA bands with maxima at 520 and 570 nm, the transient spectrum recorded in TFE has only a single band with maximum at ca 540 nm and the band with maximum at 570 nm is missing in the latter. Both steady-state electronic and IR spectroscopic study have clearly revealed that in solutions of fluorenone in normal alcoholic solvents both the free fluorenone molecule and the hydrogen-bonded complex exist in equilibrium, although the equilibrium remains in favor of the free molecule [55, 59, 75]. However, in TFE, which is a strong hydrogen bond-donating solvent, the equilibrium shifts in favor of the hydrogen-bonded complex. Hence, photoexcitation of fluorenone molecules in normal alcoholic solvents using 400-nm laser light creates the excited states of both the free fluorenone molecule as well as its hydrogen-bonded complex.
Ultrafast Dynamics of the Excited States of Hydrogen-Bonded Complexes and Solvation
781
On the other hand, in TFE, the excited state of only the hydrogen-bonded complex is created. Considering the similar positions of the maxima in the transient spectra recorded at 0.15 ps delay time in these solvents (insets of Figures 33.17 and 33.18) the transient spectrum in1-propanol or TFE may be assigned to the excited state of the free form of fluorenone. This might have been formed either upon photoexcitation of the free fluorenone molecule existing in the ground state (in 1-propanol) or by ultrafast photodissociation of the hydrogen bond in the excited state of the hydrogen-bonded complex (in 1-propanol and TFE). However, the solvent molecules cannot reorganize rapidly to incorporate the newly released alcohol molecule into its two-dimensional hydrogen bond network structure; that is, there is a “dangling” hydrogen bond still present [54b]. This non-equilibrated state of the excited fluorenone molecule is associated with a completely non-hydrogen-bonded solvent molecule, or a solvent molecule bonded into a chain to form a branch point, or some other unfavorable hydrogen bond configuration. This is a poorly solvated state, which is sufficiently unstable and tends to undergo geminate reformation of fluorenone–alcohol hydrogen bond with high probability. This reformation process is accompanied by the requisite reorganization of the hydrogen bond structure of the solvent to fully equilibrate and incorporate the dangling hydrogen bond into the hydrogen bond network structure of the solvent. With increasing delay time, the evolution of the time-resolved absorption spectra, which is associated with the rapid decrease of transient absorption in the 590–700 nm region and concomitant increase in absorption in the 510–590 nm region, can be assigned to the geminate reformation of the hydrogen bond, possibly with a new equilibrium geometry, accompanied by equilibration of the hydrogen bond network structure of the solvent. The significant difference in evolution of the spectral characteristics of the transient species in these two solvents cannot be assigned to the simple dipolar solvation but obviously to the hydrogen bonding interaction between the excited state of fluorenone and the solvent. In other words, the hydrogen bond dynamics are responsible for the observed evolution of the time-resolved spectra of the S1 state of fluorenone in alcohols. However, wavelength-dependent dynamics might have arisen due to both hydrogen reorganization process and the presence of more than one kind of hydrogen-bonded species [59]. Zhao and Han have demonstrated that the intermolecular hydrogen bond between fluorenone and methanol is significantly strengthened in the S1 state of the hydrogen-bonded complex [72]. This conclusion is quite reasonable and justified considering the fact that the excited state of fluorenone has an ICT character and, because of the increased charge density on the oxygen atom, the hydrogen bonded complex reformed following hydrogen-bond reorganization process is expected to have stronger hydrogen bond in the excited state. 33.4.1.3 Ketocyanine Dyes–Alcohol Systems (Refs [111,112]) The pronounced solvent effects in both absorption and emission spectra of the isosbestic dyes 2,5-bis[(2,3dihydroindolyl)propylene]cyclopentanone (KCD)and 2,5-bis(N-methyl N-1,3-propdienylaniline)cyclopentanone (MPAC) (Figure 33.19) make them promising probes for monitoring micro-polarity and hydrogenbond donating interactions. Bagchi and his coworkers have made a systematic study on the solvation characteristics of several ketocyanine dyes using both the steady state absorption and fluorescence and the time-resolved fluorescence measurements [113]. They showed that while non-specific dipolar solvation is O
O
N
N
N H 3C
I
N CH 3
II
Figure 33.19 Chemical structures of KCD (I) and MPAC (II). Adapted with permission from [55], and reprinted with permission from [111]. Copyright 2005 American Chemical Society
782 Hydrogen Bonding and Transfer in the Excited State
responsible for the solvatochromic properties of these dyes in aprotic solvents, specific interaction involving formation of intermolecular hydrogen-bond provides stronger solvatochromic behavior of these dyes in protic solvents. Workers from different groups have also shown that the ketocyanine dyes form strong hydrogenbonded complexes both in the ground state and in the excited state [113–116]. However, the fluorescence efficiency of these dyes is seen to be reduced significantly in protic solvents due to hydrogen-bonding interaction with the solvent. One very interesting result regarding the photophysics of KCD reported by Bagchi’s group is that both the fluorescence quantum yield (ff) and the fluorescence lifetime (tf) of KCD and MPAC, unlike in the case of many other ketocyanine dyes, increase with increasing solvent polarity. Additionally, in protic solvents, both ff and tf are remarkably higher than in aprotic solvents of comparable polarity. This is in contrast to the properties of other ketocyanine dyes [113]. The change in dipole moment (Dm) between the ground state and the fluorescing S1 state of KCD has been determined to be 3.6 D [105]. The lowest energy absorption (S1 S0) band arises due to p–p transition, involving an intramolecular charge transfer (ICT) from the electron donating indolyl group to the electron accepting carbonyl group through the intervening conjugated system [113–116]. The cause of increased ff and tf of the S1 state of these dyes could be predicted by correlating the fluorescence energy with the various measures of the solvent parameters (Figure 33.20). We find very poor correlation between the fluorescence energy and the reaction field parameter of the solvent, F. The correlation between the fluorescence energy and the Kamlet–Taft solvatochromic parameters, a and p , of the solvents are also very poor. In contrast, the fluorescence energies show good correlation with the ET(N) values of the solvents. These facts suggested formation of a hydrogen-bonded complex in the S1 state of these molecules. 20 C. C. = 0.84
C. C. = 0.79
Fluorescence Energy (103 cm-1)
18
18
16 16 0.0
0.1
0.2
0.3
0.0 0.4 0.8 1.2 1.6
Δ F( ε ,n)
20
α -Scale
20 C. C. = 0.59
C. C = 0.99
18
18
16
16
14
14 0.0
0.4 0.8 π *-Scale
0.0
0.3
0.6
0.9
ET(N) Scale
Figure 33.20 Correlation of fluorescence energy (corresponding to the maximum of fluorescence spectrum) with different solvent parameters. The fluorescence energy shows good linear correlation with ET(N) values of the solvents [111]. Adapted with permission from [51a]. Copyright 1999 American Chemical Society
Ultrafast Dynamics of the Excited States of Hydrogen-Bonded Complexes and Solvation
783
a
(B) g
0 g
-3 a
-6
-9
500
Delay times (ps): (a) 0.2, (b) 0.5, (c) 1, (d) 2, (e) 10, (f) 20 and (g) 30
600 700 Wavelength (nm)
2
Δ Absorbance (mOD)
Δ Absorbance (mOD)
3
(A)
Delay-times (ps) : 0.15, 0.3, 0.5, 1, 5, 20, 40, 80 and 150 ps
0
-2
500 550 600 650 700 750 800
Wavelength (nm)
Figure 33.21 Time-resolved transient absorption spectra constructed at different delay times following photoexcitation of KCD in DMSO (a) and 1-propanol (b) using 400 nm laser pulses [111]. Adapted with permission from [53]. Copyright 2003 American Chemical Society (See Plate 44)
In Figure 33.21, we compare the time-resolved differential transient absorption spectra recorded in DMSO and 1-propanol following photoexcitation of KCD using 400 nm excitation. Since, 400 nm light excites the KCD molecules in the S2 state in both these solvents, early time dynamics have been assigned to that of the S2 state and will not be discussed here in detail. Lifetimes of the S2 state have been determined to be about 0.5 and 0.33 ps in DMSO and 1-propanol, respectively. Decay of the negative absorption band in the 470–530 nm region in the early time-domain could be assigned to the recovery of the ground state bleaching and decay of the stimulated emission associated with the S2 state. Simultaneous to the decay of the S2 state in DMSO, the growth of the SE band at 550 nm indicates the formation of the S1 state, which is long lived with a life time of about a few nanoseconds [113]. However, the nature of the time-resolved spectra and the dynamical behavior of the transient species in 1-propanol are significantly different from that observed in DMSO. At longer delay times, ranging from 1 to 150 ps, we observe the development of the stimulated emission band in the 570–800 nm region and an ESA band in the 530–570 nm region as well as the reduction of the negative absorption band in the 490–530 nm region. On increasing the delay-time, the maximum of the stimulated emission band is shifted from 590 nm (observed at 1 ps delay time) to 650 nm (observed at 150 ps delay time). The dynamic Stokes shift of the maximum of the time-resolved stimulated emission band from 590 nm to 650 nm could not be assigned merely to solvation, since the stimulated emission band also grows simultaneously with increasing delay time. This leads to the presumption of involvement of a precursor state, which has been designated as the S10 state, prior to formation of the long-lived S1 state. The dynamics of the excited states of KCD in straight-chain alcohols with varying length of the hydrocarbon chain, for example, methanol to 1-octanol, as well as in ethylene glycol (EG) have been investigated. The spectral and temporal characteristics of the transient species produced in these alcoholic solvents have been observed to be very similar to those in 1-propanol. Figure 33.22 shows time-resolved spectra of the transient species produced due to photoexcitation of KCD in TFE with 400 nm laser pulses. The spectrum constructed for 0.15 ps delay-time consists of two negative absorption bands in the 490–530 nm (bleaching band) and 530–620 nm regions (SE band), and an ESA band in the 620–800 nm region. All these features are characteristics of the S2 state of KCD. The time-resolved spectra constructed at longer delay-times reveal the rapid decay of the SE band in the 490–630 nm region and the
784 Hydrogen Bonding and Transfer in the Excited State 4 2
Delay-times (ps) (a) 0.15, (b) 0.3, (c) 0.5, (d) 1 & (e) 2
(A)
0 a f -2
Δ absorbance (mOD)
f
a
-4 -6
Delay-times (ps)
2
(B)
(f) 4, (g) 10, (h) 20 & (i) 50
0 f -2 -4 -6
500
550
600 650 700 Wavelength (nm)
750
Figure 33.22 Time-resolved transient absorption spectra constructed in sub-5 ps (A) and sub-50 ps (B) time domains following photoexcitation of KCD in TFE [111]. Copyright American Chemical Society, reproduced with permission
development of another stimulated emission band with a maximum at 600 nm. Following the line of assignments of the different transient species in other solvents, the emission bands with maximum at ca 0 0 500, 600 and 650 nm have been assigned to the S2, S1 and S1 states, respectively. The lifetimes of the S2 and S1 states could be assigned as 0.5 0.1 and 9.0 0.5 ps, respectively. As in other solvents, the lifetime of the S1 state is too long to measure here. 33.4.1.4 Hydrogen Bond Dynamics Versus Solvation (Refs [54, 55, 111, 112]) The standard picture of non-specific or dipolar solvation dynamics assumes relatively weak intermolecular interactions between the excited solute and the surrounding solvent molecules [10–15]. The nature of interaction between the solute and each of the large number of solvent molecules is equally important, and the dynamics are associated with the motion of many solvent molecules along a collective coordinate. The barriers to these motions are small compared to thermal energies. However, although the energy barrier for the hydrogen bond breaking or forming is low, it is not negligible as compared to the thermal energies. In reality, hydrogen-bond dynamics are intermediate between non-specific solvation and covalent bond breaking. Additionally, interaction of the solute with the single hydrogen-bonded solvent molecule is stronger than its interaction with other solvent molecules. However, the strength and dynamics of this interaction may not differ too much from the hydrogen bonding interaction between the solvent molecules, and hence the latter may
Ultrafast Dynamics of the Excited States of Hydrogen-Bonded Complexes and Solvation
785
influence the relaxation dynamics of the excited solute molecule in a significant way [54]. Finally, hydrogen bonding, which has been described here as a “specific” interaction, is similar to non-specific interactions such as dipole–dipole or van der Waals interactions [117]. From this viewpoint hydrogen-bond making or breaking is just a change in solute–solvent interaction resulting from the reorganization of the solvent, that is, it is a type of solvation dynamics. The possible reason behind the fact that dielectric relaxation of alcoholic solvents has attracted more attention than the dynamics of specific interaction may be the choice of the probe molecules, which do not display much difference in fluorescence behavior in protic and aprotic solvents because of weak hydrogenbonding interaction between the solute and the solvent. Morimoto et al. have studied the quenching of fluorescence of a series of aromatic carbonyl compounds by alcoholic solvents [20]. This work revealed that the S1 states of 3-amino- and 4-aminofluorenones are more efficiently quenched by alcohols than those of two coumarin dyes, namely, coumarin 153 and coumarin 151. These dyes are popular fluorescent probes for studying the dipolar solvation dynamics [118]. To explain this difference in behavior, Morimoto et al. have invoked the concept of “hard and soft acid–base” (HSAB) behaviors of the solvent and the excited state, respectively. Both the amino and the carbonyl groups are present in both the kinds of molecules and the values of Dm of the aminofluorenones are comparable to those of the coumarin dyes (Dm 5.8 and 3.1 D for C153 and C151, respectively). However, theoretical calculations have revealed that in the S1 state with ICT character the negative charge is mostly localized on the carbonyl oxygen atom in the case of aminofluorenones, whereas in the case of the coumarin compounds the negative charge is substantially delocalized over the whole aromatic moiety [20]. Following the creation of the S1 state, the local charge density on the amino group is decreased by about 21% in both kinds of molecules but the charge density on the carbonyl oxygen atom is increased by about 9% in the case of the aminofluorenones, whereas it is increased by only about 2% in the case of the coumarin molecules. Hence, the S1 states of the aminofluorenones have been considered as hard anions (or bases), since the negatively charged oxygen atom interacts strongly with the hydroxyl hydrogen of the alcoholic solvent, which has been considered as a hard cation (or acid). On the other hand, the S1 states of the coumarin molecules are soft anions (or weak bases) and do not have appreciable interaction with the hydroxyl hydrogen of the alcohols. This is why the coumarin dyes have proven to be ideal probes for studying the dynamics of polar solvation in both aprotic and protic solvents without any quenching of fluorescence of the probe via specific interaction between the probe and the solvent [118]. In cases of the chemical systems described above, while the hydrogen bond dynamics investigated in TFE can be thought of as bond-breaking and bond-making chemical reactions rather than a solvation process, in normal alcohols the wavelength-dependent dynamics of the hydrogen bond lead us to consider it as merely a solvation process. In hydrogen-bonding solvents, the Debye dielectric relaxation process follows multiexponential dynamics and the longest component is generally assumed to be connected with the rate of hydrogen bond reorganization in the solvent [9]. The most common and conventional method of determining the lifetime of dipolar solvation process is to follow the time-dependent redshift of the emission maximum of the emission or SE spectrum or blue-shift of the ESA spectrum during the relaxation of the excited state of a probe molecule [10–15, 118]. The timeevolution of the fluorescence spectrum due to solvation is quantified by the time correlation function, C(t), for the dynamic shift of the emission maximum observed in the time-resolved fluorescence experiment, and is expressed by equation (33.4): CðtÞ ¼
nðtÞnð1Þ nð0Þnð1Þ
ð33:4Þ
where n(0) is the optical frequency of the maximum of the emission of the probe at zero time (i.e., just after excitation), n(1) at infinite time (when the solvent dipoles have relaxed to equilibrium) and n(t) at
786 Hydrogen Bonding and Transfer in the Excited State
intermediate time (t) (i.e., during the solvent relaxation). However, unfortunately, in the present cases, the values of C(t) could not be determined with reasonable accuracy, because of significant mutual overlap between the ESA and SE (or fluorescence) bands [111, 112] and/or involvement of multiple excited states [54, 55]. However, a solvent relaxation process has clearly been revealed by the more rapid growth and decay of the intensity of SE band at the blue-edge of the emission spectrum and slower growth and decay of the same on the red-edge of the emission spectrum [111, 112]. Hence, considering these facts, application of equation (33.4) in determining the lifetime of the solvent reorganization process does not seem to be justified and may lead to erroneous results. Therefore, we have adopted the “single or linear wavelength” method to determine the average solvation time [119, 120]. We observe that, in protic solvents, both the growth and decay lifetimes of the intensity of the SE band increase as the monitoring wavelength approaches the lower energy region of the SE band. The largest value of this lifetime has been obtained at wavelengths in the lower energy region beyond the maximum wavelength of the SE band. This value has been considered here as the average lifetime of the solvent reorganization process, tR. In the case of solvent-controlled barrier crossing reactions, in which the reactant and the solvent are strongly coupled, dynamical solvent effects are usually discussed in terms of “friction”, z, to account for the effect of the solvent on the lifetime (t) of the reactant [121–123]. Berg and coworkers conceived the fact that the dynamics of a specific interaction affecting hydrogen bond reorganization time of the solvent are well correlated with the dielectric relaxation time, and the hydrogen bond lifetimes are very similar to the longitudinal relaxation time, tL [54]. Based on these arguments, equation (33.5) could be rewritten by equating the solvent friction, j, to the longitudinal relaxation time (tL) of the solvent as [equations (33.6) and (33.7)]: t ¼ Aj expðDH=RTÞ
ð33:5Þ
t ¼ AðtL ÞexpðDH=RTÞ
ð33:6Þ
lnðtÞ ¼ lnðtL Þ þ lnðAÞ þ DH=RT
ð33:7Þ
Assuming that the enthalpy of the transition state, DH, does not vary significantly due to a change of solvents belonging to the same class, for example, primary alcohols, we expect a linear correlation between ln(tR) and ln(tL) [equation (33.7)]. As shown in Figure 33.23, we find a perfect linear correlation between these two parameters in the cases of fluorenone and KCD in normal alcohols and reconfirm the postulation that hydrogen bond reorganization around the excited solute molecule is solely responsible for the dynamics of the relaxation process of the S1 state in the alcoholic solvents observed here. 33.4.2 Time-Resolved IR absorption spectroscopic technique The examples presented above show that the electronic transitions are strongly broadened owing to coupling with the fluctuating solvent and are relatively featureless due to overlapping of different transitions as well. Spectral shifting and reshaping caused by solvent reorganization leads to wavelength-dependent dynamics, and ensemble-averaged time-correlation functions for liquid motion provide very limited information regarding microscopic solvent structure. However, despite the many disadvantages of probing electronic transitions of the transient species, valuable information could be obtained regarding the dynamics of chemical reactions by detailed analyses of the transient spectra and the dynamics monitored at different wavelengths using ultrafast visible spectroscopy. The following two examples show that time-resolved vibrational spectroscopy is a more powerful technique for monitoring hydrogen-bond dynamics in real time. The vibrational spectroscopic technique also has the added advantage that it is possible to observe changes in distinct functional groups involved in hydrogen bond formation, which provides site-specific insight into local dynamics.
Ultrafast Dynamics of the Excited States of Hydrogen-Bonded Complexes and Solvation
787
4.0 3.5
A. Fluorenone C. C. = 0.95
2.5
R
)
3.0
ln(τ
2
3
4
5
4 2 B. KCD
0
C. C. = 0.97 -2 -2
-1
0
1
2
3
4
5
ln (τL)
Figure 33.23 Correlation between the solvent relaxation time, tR, and the longitudinal relaxation times, tL, of the alcoholic solvents [55, 111]. Copyright American Chemical Society, reproduced with permission
33.4.2.1 C102–Phenol (Ref. [51]) In Section 33.3, we discussed the hydrogen-bonding interaction and formation of a hydrogen-bonded complex between C102 and phenol in the ground electronic state. To investigate the ultrafast response of the complexes to an electronic dipole, the C102 chromophore was excited by a 100 fs pulse at 400 nm, resonant to the purely electronic S0–S1 transition, whereas the phenol molecules remain in their electronic ground state. The resulting changes of vibrational absorption of both the C102 and phenol molecules are monitored by mid-infrared probe pulses tunable in the range of the C¼O and O--H stretching bands. The time-resolved data gave the first direct insight into the microscopic dynamics of the complexes and revealed two distinct time scales of structural response, relating to the different types of hydrogen bonds in the system. Upon photoexcitation of C102, the stretching band of the C¼O group of C102 at 1695 cm1 disappears and is replaced by a weaker broad band with a maximum at about 1740 cm1 (Figure 33.24a). This position is characteristic of a free C¼O group for both the ground state (Figure 33.7a) and the S1 state of C102 [51b]. Since these changes take place within the time resolution of the experiment of 200 fs, these observations lead to the conclusion that the C¼O H--O hydrogen bond in both the C102-phenol complexes broke within 200 fs after excitation of C102. The fast cleavage of the C¼O. . .H--O hydrogen bond is manifested in the step-like increase of free O--H absorption at 3610 cm1 (Figure 33.24b). The cleavage of this bond has been explained by the changes of the local charge distribution of C102. In particular, the polarity, and thus the hydrogen affinity, of the C¼O group decreases in the S1 state of C102, as is suggested by semi-empirical calculations of the S1 charge distribution [42]. Cleavage of the hydrogen bond could be an instantaneous process owing to a rearrangement of electronic density upon excitation. It could also involve nuclear motion on a time scale set by the period of low-frequency vibrations of the hydrogen bonded groups in the 100–200 cm1 range. Such frequencies translate into time constants < 200 fs, in agreement with the dynamics found here. Significant changes of the broad O--H band between 3200 and 3550 cm1 has also been found upon electronic excitation of C102 for delay times up to 5 ps. There is a strong increase of infrared absorption
788 Hydrogen Bonding and Transfer in the Excited State
Figure 33.24 (a) Transient C¼O stretching band of C102 in the complexes after excitation to the S1 state of C102. The spectra are shown for different time delays after excitation. (b) Transient vibrational spectra of the (phenol)n units in the complexes after excitation of C102, recorded at the same delays as shown in (a) [51a]. Copyright American Chemical Society, reproduced with permission
in this frequency range, followed by a reshaping of the spectrum. The reshaping of the spectrum occurs with a characteristic time constant of 800 fs. The vibrational band found after about 5 ps stays unchanged for even later times and is very close to the stretching band of the hydrogen bonded O--H group in the phenol dimer. The absorption changes disappear on a time scale of nanoseconds with the decay of the S1 state of C102. Evolution of the transient absorption spectra in the sub-5 ps time-domain presented in Figure 33.24 reflects the non-equilibrium geometry of the complexes immediately after cleavage of the C¼O. . .H--O hydrogen bond. Cleavage is accompanied by a redistribution of electron density in the phenol molecule close to C102, and it acts back on the neighboring C¼O group of C102, thereby changing the strength of the C¼O band. However, the hydrogen bond with the other phenol molecules in C102–(phenol)n complex remains intact and hence the cleavage leads to changes in hydrogen bond strength and the vibrational transition moments of the bridged OHgroups. Subsequently, the released (phenol)n moiety reorganizes itself to a new equilibrium configuration, mainly by reorientation of the phenol units relative to each other and to the excited C102 molecule. This structural reorganization of (phenol)n with a time constant of 800 fs has been directly monitored in this study (Figure 33.24b). The new band between 3400 and 3550 cm1 has a shape very close to the infrared absorption of 1 : 1 phenol–phenol complexes (Figure 33.7b), leading to the conclusion that the contribution of larger complexes (n > 2) towards the observed dynamics is not significant. The reorientation time of 800 fs corresponds to a low-frequency motion of about 40 cm1, a frequency typical for librational motions that could well represent the relative tilting of the rings in (phenol)2. Importantly, this technique has provided an
Ultrafast Dynamics of the Excited States of Hydrogen-Bonded Complexes and Solvation
789
insight into the dynamics involving the phenol unit not directly bonded to C102, demonstrating the potential of femtosecond vibrational spectroscopy for probing dynamics in hydrogen bonding networks at a finite distance from the directly excited groups. Wells et al. have recently reported the intermolecular response of solvent molecules following photoexcitation of C102 in acetonitrile–water binary mixtures using a technique measuring the time-domain Raman response [74]. At low water concentrations, the solvent response was consistent with a dipolar solvation process. However, with increasing water concentration, an additional response was found subsequent to dipolar solvation, exhibited as a rapid gain in the solvent’s polarizability on the 250 fs time-scale. Simulation studies indicated that the probability of the C102 solute being hydrogen bound with two water molecules simultaneously, both as donors at the carbonyl site, increases in a correlated fashion with the amplitude of the additional response in the measurements, leading to the conclusion that the excitation of C102 simultaneously weakens and strengthens hydrogen bonding in complexes with two inequivalently bound waters. 33.4.2.2 C102–Aniline (Ref. [53]) Time-resolved infrared absorption spectroscopic studies have been performed with solutions containing C102 in neat aniline. Upon photo-excitation of C102–amine systems by ultrashort (duration of about 150 fs) laser pulses of 400 nm, which is resonant to the electronic S1 S0 transition of C102, only the C102 chromophore is excited. Aniline molecules do not absorb at 400 nm and, hence, remain in the ground electronic state. Figure 33.25 presents the temporal profile of the transient absorption monitored at 1742 cm1. The hydrogenbonded complex has no absorption at 1742 cm1 but the free C¼O group absorbs strongly at this frequency. The probe is resonant to C¼O stretching and causes the v ¼ 0 to v ¼ 1 transition in the S1 state of C102. We observe the instrument response time-limited rise of transient absorption followed by a tri-exponential decay. The lifetimes of the two components decaying in the early time domain were determined to be 0.5 0.1 and 6.8 0.4 ps and the third component is very long and has a lifetime of about a few nanoseconds.
Absorbance, mOD
1.5
C-102 in neat AN ν probe
= 1742 cm-1
τ 1 (d) = 0.5 ps τ 2 (d) = 6.8 ps
1.0
τ 3 (d) = long
0.5
0.0 -5
0
5
10
15
20
25
Time, ps
Figure 33.25 Time-resolved change of vibrational absorption monitored at 1742 cm1 in neat aniline. The lifetimes obtained by two exponential fittings of the data are given [53]. Copyright American Chemical Society, reproduced with permission
790 Hydrogen Bonding and Transfer in the Excited State
Since the lifetime of the S1 state of C102 in neat aniline is 1.4 ns, the time dependent absorbance changes, as shown in Figure 33.25, indicate the role of a hydrogen bond in the excited state dynamics of C102 in aniline. The appearance of a transient absorption signal at 1742 cm1, which indicates the formation of free C¼O group, immediately after the electronic excitation of the C102 chromophore in C102–aniline hydrogenbonded complex, indicates the instantaneous dissociation ( pTsO. Furthermore, the same excitation spectra for both emission bands confirmed that the two excited states originate from the same ground state. The correlation of the second emission band with bound anion basicity suggested that excited-state proton transfer was occurring with the increased acidity of the fluorophores in the CT excited state upon anion binding.
35.4 Recognition and Sensing of Anions by Conjugated Polymers through ESIPT Conjugated polymers possess unique electrical and optical properties due to their extended p-conjugation in the main chain. With appropriately functionalized receptors incorporated into conjugated polymers, it is possible to detect, transduce and amplify chemical information into an optical signal after analyte–receptor interactions by the so-called “molecular-wire effect” [22]. There are several examples reported in the literature of anion sensing by conjugated polymers [23]. Recently, an interesting example has been reported by Pang et al., showing the application of a conjugated polymer incorporating 2-(2-hydroxyphenyl)-1,3-benzoxazole (HBO) derivative for anion sensing through H-bonding in excited state followed by an ESIPT process [24]. It is known that HBO undergoes ESIPT with transformation of enol into keto form (dependent upon solvent [25] and temperature [26]). HBO exists in two rotamers (4a and 4b) in a 1 : 1 ratio in the crystalline state: one with the –OH group oriented towards N-atom and the other with the –OH group towards the O-atom of the oxazole ring [27]. However, only one conformer (4a), in which the –OH group is oriented towards N-atom, undergoes the ESIPT process to give emission with a large Stokes shift [28]. For sensitive performance by HBO-type moiety in sensing terms, it is highly desirable to reduce the formation of conformer 4b. In this regard, the 2,5-bis(benzoxazol-2-yl)benzene-1,4-diol derivative (5) is a suitable system to study in which two benzoxazole units remain in the same plane as the central para-catechol moiety through intramolecular hydrogen bonding in the ground state (Scheme 35.2) [24]. With a series of polymers and corresponding monomers based on the 2,5-bis(benzoxazol-2-yl)benzene-1,4-diol derivative, Pang et al. have shown that polymer 7, obtained after de-alkylation of parent polymer 6, exhibited much weaker fluorescence
Scheme 35.2 Schematic representation of the ESIPT process in bis(HBO). Reproduced with permission from Ref. [24]. Copyright 2007 The American Chemical Society
Molecular Recognition and Chemical Sensing of Anions Utilizing Excited-State Hydrogen-Bonding Interaction 811
than polymer 6. A large redshift (200 nm) of the emission band, however, was observed at lmax 619 nm for 7, despite the bathochromic shift (40 nm) in absorption spectra compared to 6 (Figure 35.6a). This large Stokes shift (200 nm) in polymer 7 was attributed to an ESIPT process through bis(HBO) moiety, which was not possible in polymer 6.
The absorption spectra of polymer 7 shows a disappearance of the bands at 400 and 421 nm along with the appearance of a new band at 510–540 nm (Figure 35.6b) upon addition of certain anions. The large shift of 120 nm in absorption spectra after addition of anions shows significant perturbation in the ground state. After
Figure 35.6 Absorption (solid line) and emission (dotted line) spectra (a) of polymers 6 and 7, in anhydrous THF, and UV–vis (solid line) and fluorescence (broken line) spectra (b) of polymer 7 and its anion complexes. Reprinted with permission from [24]. Copyright 2007 American Chemical Society (See Plate 45)
812 Hydrogen Bonding and Transfer in the Excited State
Figure 35.7 Titration spectra of compound 8 with Bu4NOH in THF/EtOH (30 : 1). Reprinted with permission from [24]. Copyright 2007 American Chemical Society (See Plate 46)
addition of fluoride or acetate anion to polymer 7 in THF/ethanol (50 : 1), it turned from a very weakly fluorescent to a strong fluorescent solution and the result clearly showed the inhibition of ESIPT process because of the deprotonation of the –OH functional group. Because of the poor solubility of compound 7, the authors have employed a monomeric derivative 8 to study in more detail. The titration spectra of 8 with Bu4NOH in mixture of THF and ethanol show the appearance of new band at 496 nm with disappearance of bands at 400 and 410 nm (Figure 35.7). With a gradual increase in concentration of Bu4NOH, the intensity of the band at 496 nm increases and shows a characteristic isosbestic point at 432 nm and no new band appears, which suggests the formation of monoanion species after deprotonation of the bis(HBO) moiety. This was also confirmed through 1 H NMR titration and ESI-MS spectra. A similar observation was made in the titration spectra of 8 with Bu4NF (Figure 35.8a); the Benesi– Hildebrand plot (Figure 35.8b) showed that complex formation involved two fluoride anions and it was
Figure 35.8 Titration spectra (a) and Benesi–Hildebrand plot (b) of 8 with Bu4NF in THF/EtOH (100 : 1). Reprinted with permission from [24]. Copyright 2007 American Chemical Society (See Plate 47)
Molecular Recognition and Chemical Sensing of Anions Utilizing Excited-State Hydrogen-Bonding Interaction 813
suggested that fluoride anions first interact with 8 to form H-bonding in the ground state to form 9 followed by second fluoride anion induced deprotonation to form 10 with elimination of [F H F] anion.
35.5 Concluding Remarks In conclusion, for anion sensing through ESIPT process the receptors should possess the following characteristics: (i) the receptor should form intramolecular hydrogen bonding in the ground state within neighboring proton donor and acceptor groups, (ii) the proton donor should be acidic enough to become deprotonated upon interaction with basic anions but should be resistant to solvent disturbance and (iii) the proton donor should be strong for fast ESIPT process and increase the quantum yield of tautomer emission. The ESIPT process provides a new platform for a fast and selective route for anion sensing. Notably, some literature examples have reported anion sensing based on an ESIPT mechanism but, instead, they were simply ground-state acid–base reactions. It is, therefore, crucial to carefully judge all the experimental evidence before claiming the anion recognition and sensing mechanism. Although the number of examples in the literature for anion sensing based on the ESIPT mechanism is still very low, the judicious choice of molecular receptor with appropriate complementary structures for anions and incorporation of a suitable moiety for ESIPT process would surely open up the way for detection of biologically and environmentally important anions in the future.
References 1. (a) P. Schuster, G. Zundel and C. Sandorfy, The Hydrogen Bond - Recent Developments in Theory and Experiments. II. Structure and Spectroscopy, North Holland, Amsterdam (1976); (b) G. A. Jeffrey, An Introduction to Hydrogen Bonding, OUP, New York, (1997); (c) G. R. Desiraju and T. Steiner, The Weak Hydrogen Bond in Structural Chemistry and Biology, OUP, Chichester, (1999); (d) N. J. Turro, Modern Molecular Photochemistry, University Science Books, Mill Valley, CA, p. 132 (1991). 2. (a) A. Jarczewskia and C. D. Hubbard, J. Mol. Struct., 649, 287 (2003); (b) L. M. Tolbert and K. M. Solntsev, Acc. Chem. Res., 35, 19 and references cited therein (2002); (c) J. T. Hynes, T.-H. Tran-Thi and G. Granucci, J. Photochem. Photobiol. A: Chem., 154, 3 (2002); (d) J. Waluk, Conformational aspects of intra- and intermolecular proton transfer, in Conformational Analysis of Molecules in Excited States, ed. J. Waluk, Methods in Stereochemical Analysis Series, Wiley-VCH, New York, Chapter 2, pp. 57–111 (2000); (e) L. G. Arnaut and S. J. Formosinho, J. Photochem. Photobiol. A: Chem., 75, 1 (1993); (f) S. J. Formosinho and L. G. Arnaut, J. Photochem. Photobiol. A: Chem., 75, 21 (1993). 3. (a) A. Weller, Z. Elektrochem., 60, 1144 (1956); (b) A. Weller, Prog. React. Kinet., 1, 188 (1961). 4. (a) A. Morimoito, T. Yatsuhashi, T. Shimada et al., J. Phys. Chem. A, 105, 10488 (2001); (b) H. Inoue, M. Hida, N. Nakashima and K. Yoshihara, J. Phys. Chem., 86, 3184 (1982); (c) M. Maroncelli, J. Macinnis, G. R. Fleming, Science, 243, 1674 (1989). 5. (a) H. Shirota, H. Pal, K. Tominaga and K. Yoshihara J. Phys. Chem., 100, 14575 (1996); (b) M. Sugita, T. Shimada, T. Tachibana and H. Inoue, Phys. Chem. Chem. Phys., 3, 2012 (2001). 6. (a) K. Choi and A. D. Hamilton, Angew. Chem., Int. Ed., 40, 3912 (2001); (b) H. Tong, G. Zhou, L. Wang et al., Tetrahedron Lett., 44, 131 (2003); (c) X. Zhang, L. Guo, F. Y. Wu and Y. B. Jiang, Org. Lett., 5, 2667 (2003). 7. (a) T. Schultz, E. Samoylova, W. Radloff et al., Science, 306, 1765 (2004); (b) A. L. Sobolewski and W. Domcke, Europhys. News, 37, 20 (2006).
814 Hydrogen Bonding and Transfer in the Excited State 8. (a) J. Keck, H. E. A. Kramer, H. Port et al., J. Phys. Chem., 100, 14468 (1996); (b) P. G. Parejo, M. Zayat and D. Levy, J. Mater. Chem., 16, 2165 (2006); (c) P. T. Chou and M. L. Martinez, Radiat. Phys. Chem., 41, 373 (1993); (d) A. Sytnik and J. C. Delvalle, J. Phys. Chem., 99, 13028 (1995). 9. (a) Z. C. Wen and Y. B. Jiang, Tetrahedron, 60, 11109 (2004); (b) K. Thomas and J. A. George, Neuron, 27, 447 (2000); (c) S. Nishizawa, Y. Kato and N. Teramae, J. Am. Chem. Soc., 121, 9463 (1999); (d) B. Ramachandram, R. L. Joseph and D. G. Chris, Anal. Biochem., 327, 82 (2004); (e) B. Yann, J. C. Martin, P. David and S. Rachel, Chem. Commun., 1930 (2002); (f) S. Jayaraman, J. Biwersi and A. S. Verkman, Am. J. Physiol., 276, 747 (1999); (g) C.-L. Chen, Y.-H. Chen, C.-Y. Chen and S.-S. Sun, Org. Lett., 8, 5053 (2006); (h) X. Peng, Y. Xu, S. Sun et al., J. Org. Biomol. Chem., 5, 226 (2007). 10. H.-J. Schneider and A. K. Yatsimirsky, Principles and Methods in Supramolecular Chemistry, John Wiley & Sons, Ltd., Chichester (2000). 11. B. A. Moyer and P. V. Bonnesen, Supramolecular Chemistry of Anions, ed. by A. Bianchi, K. Bowman-James and E. Garcıa-Espan˜a, Wiley-VCH, New York, Chapter 1, pp. 1–44 (1997). 12. (a) C.-Y. Chen, T.-P. Lin, C.-K. Chen et al., J. Org. Chem., 73, 900 (2008); (b) X. Zhang, L. Guo, F.-Y. Wu and Y.-B. Jiang, Org. Lett., 5, 2667 (2003). 13. (a) X. Peng, Y. Wu, J. Fan et al., J. Org. Chem., 70, 10524 (2005); (b) Y. Cui, H.-J. Mo, J.-C. Chen et al., Inorg. Chem., 46, 6427 (2007); (c) M. M. Henary, Y. Wu, J. Cody et al., J. Org. Chem., 72, 4784 (2007). 14. (a) B. Liu and H. Tian, J. Mater. Chem., 15, 2681 (2005); (b) S. K. Kim, J. H. Bok, R. A. Bartsch et al., Org. Lett., 7, 4839 (2005); (c) P. A. Gale, Chem. Commun., 3761 (2005); (d) S. J. Brooks, L. S. Evans, P. A. Gale et al., Chem. Commun., 734 (2005); (e) S. Camiolo, P. A. Gale, M. B. Hursthouse et al., Chem. Commun., 758 (2002). 15. (a) C. B. Black, B. Andrioletti, A. C. Try et al., J. Am. Chem. Soc., 121, 10438 (1999); (b) H. Miyaji, H.-K. Kim, E.-K. Sim et al., J. Am. Chem. Soc., 127, 12510 (2005); (c) R. Nishiyabu and P. Anzenbacher, Jr., Org. Lett., 8, 359 (2006); (d) P. Anzenbacher, Jr., M. A. Palacios, K. Jursıkova and M. Marauez, Org. Lett., 7, 5027 (2005). 16. (a) V. Amendola, E. D. Go´mez, L. Fabbrizzi and M. Licchelli, Acc. Chem. Res., 39, 343 (2006); (b) A. B. Descalzo, K. Rurack, H. Weisshoff et al., J. Am. Chem. Soc., 127, 184 (2005); (c) E. J. Cho, J. W. Moon, S. W. Ko et al., J. Am. Chem. Soc., 125, 12376 (2003). 17. Y. Qian, S. Li, G. Zhang et al., J. Phys. Chem. B, 111, 5861 (2007). 18. (a) C. A. Taylor, M. A. EI-Bayoumi and M. Kasha, Proc. Natl. Acad. Sci. USA, 63, 253 (1969); (b) P. Avouris, L. L. Yang and M. A. EI-Bayoumi, Photochem. Photobiol., 24, 211 (1976). 19. (a) C. Chang, N. Shabestary, M. A. EI-Bayoumi, Chem. Phys. Lett., 75, 107 (1980); (b) J. Waluk, S. J. Komorowski and J. Herbich, J. Phys. Chem., 90, 3868 (1986); (c) J. Waluk, J. Herbich, D. Oelkrug and S. Uhl, J. Phys. Chem., 90, 3866 (1986). 20. (a) M. Boiocchi, L. D. Boca, D. E. Go´mez et al., J. Am. Chem. Soc., 126, 16507 (2004); (b) D. E. Go´mez, L. Fabbrizzi and M. Licchelli, J. Org. Chem., 70, 5717 (2005); (c) D. E. Go´mez, L. Fabbrizzi, M. Licchelli and E. Monzani, Org. Biomol. Chem., 3, 1495 (2005); (d) V. Amendola, M. Boiocchi, L. Fabbrizzi and A. Palchetti, Chem. Eur. J., 11, 120 (2005); (e) M. Boiocchi, L. D. Boca, D. E. Go´mez et al., Chem. Eur. J., 11, 3097 (2005). 21. Y. Wu, X. Peng, J. Fan et al., J. Org. Chem., 72, 62 (2007). 22. (a) D. T. McQuade, A. E. Pullen and T. M. Swager, Chem. Rev., 100, 2537 (2000); (b) U. H. F. Bunz, Chem. Rev., 100, 1605 (2000); (c) L. Chen, D. W. McBranch, H. L. Wang et al., Proc. Natl. Acad. Sci. USA, 96, 12287 (1999); (d) M. Leclerc, Adv. Mater., 11, 1491 (1999); (e) T. M. Swager, Acc. Chem. Res., 31, 201 (1998). 23. (a) C.-Y. Wu, M.-S. Chen, C.-A. Lin et al., Chem. Eur. J., 12, 2263 (2006); (b) G. Zhou, Y. Cheng, L. Wang et al., Macromolecules, 38, 2148 (2005); (c) D. Aldakov and P. Anzenbacher, J. Am. Chem. Soc., 126, 4752 (2004); (d) J. K. Lee, J. Na, T. H. Kim et al., Mater. Sci. Eng., C, 24, 261 (2004). 24. Q. Chu, D. A. Medvetz and Y. Pang, Chem. Mater., 19, 6421 (2007). 25. (a) J. Seo, S. Kim and S. Y. Park, J. Am. Chem. Soc., 126, 11154 (2004); (b) O. K. Abou-Zied, R. Jimenez, E. H. Z. Thompson et al., J. Phys. Chem. A, 106, 3665 (2002). 26. K. Das, N. Sarkar, A. K. Ghosh et al., J. Phys. Chem., 98, 9126 (1994). 27. Y. P. Tong, Acta Crystallogr., Sect. E, 61, o3076 (2005). 28. (a) K. Das, N. Sarkar, D. Majumdar and K. Bhattacharyya, Chem. Phys. Lett., 198, 443 (1992); (b) G. J. Woolfe, M. Melzig, S. Schneider and F. Doerr, Chem. Phys., 72, 213 (1983).
36 Theoretical Studies of Green and Red Fluorescent Proteins Hong Zhang, Qiao Sun, Sufan Wang, Seth Olsen and Sean C. Smith The University of Queensland, Australian Institute for Bioengineering and Nanotechnology, Centre for Computational Molecular Science, QLD 4072, Brisbane, Australia
36.1 Introduction The green fluorescent protein (GFP) of the Aequorea victoria jellyfish (and its structural analogues, e.g., red fluorescent protein) has emerged as a unique fluorescent label and has evolved into an extraordinarily important platform for biotechnological and cell biology applications due to its amazing ability to generate a highly visible, efficiently emitting internal fluorophore (for an overview, see review articles [1, 2]). Its unique photophysical properties are nowadays being commonly exploited for imaging studies of protein folding, gene expression, protein trafficking and cell development, since the gene in GFP contains all the information necessary for the post-translational synthesis of the chromophore and expression of the gene in other organisms can create fluorescence. Structurally, GFP is an 11-stranded b-barrel, which is nearly a perfect cylinder, threaded by an a-helix running up the axis of the cylinder (Figure 36.1) [3, 4]. The chromophore is attached to the a-helix and is buried almost perfectly in the centre of the cylinder. The chromophore (p-hydroxybenzylideneimidazolinone) is auto-catalytically generated by the post-translational modification of a three-amino-acid sequence. This characteristic structure of GFP can both constrain the motions of the chromophore and shield it from the surrounding water solvent that would otherwise quench its fluorescence. A surprising number of polar groups and structured water molecules are buried adjacent to the chromophore and appear to play a crucial role in the functionality of the system through their participation in hydrogen bonding network. Owing to the importance of GFP as a bio-imaging reagent, there is substantial interest in understanding the photophysics of the embedded chromophore and in particular in determining the structural changes and the dynamics in and around the chromophore following the photo-excitation. Various
Hydrogen Bonding and Transfer in the Excited State, Volume I & II © 2011 John Wiley & Sons, Ltd. ISBN: 978-0-470-66677-7
Edited by Ke-Li Han and Guang-Jiu Zhao
816 Hydrogen Bonding and Transfer in the Excited State
Figure 36.1 Structure of green fluorescent protein and its chromophore. Reprinted with permission from RCSB PDB
spectroscopic techniques as well as theoretical tools have been employed to elucidate the fluorescent mechanisms in progressively greater detail [3, 5–15]. The wild-type (wt) GFP has been widely regarded to exist under ambient conditions in two forms that are related by deprotonation of the chromophore through a proton shuttle mechanism that favors the neutral form in vivo. The absorption spectrum of wtGFP is characterized by two bands at 395 nm (band A) and at 475 nm (band B) [16]. Excitation into either band results in highly efficient generation of green fluorescence. The excited state dynamics of wtGFP have been studied using ultrafast fluorescence and absorption spectroscopies [16–18]. Based upon their picosecond spectroscopy studies, Boxer et al. concluded that light-driven conversion between the neutral (A state) and anionic (B state) chromophore proceeds via excited state proton transfer (ESPT) and passes through an intermediate state I [18]. Using hole-burning spectroscopy, Creemers et al. have characterized the A, B and I states in wtGFP and demonstrated that they have distinct photophysical properties [19]. Thus, the generally accepted picture about wild-type GFP [18] is that upon excitation at 395 nm there is proton transfer on the excited A state to form an intermediate excited state I of GFP, from which state fluorescence of 508 nm (the characteristic intense green fluorescence) is emitted and GFP returns to the ground intermediate state I. Figure 36.2 shows one mechanism for the photo-isomerization of GFP in which proton transfer processes play a vital role [3]. The three-proton relay includes chromophore (in green), water W22, and Ser205 and Glu222 residues, which is connected using the arrow symbols. Upon excitation at 475 nm, there is no proton transfer on the excited state, and the emission of 503 nm (at room temperature) is from the B to B state of GFP. The I state is electronically very similar to B state the (chromophore is in its anionic state), but environmentally very similar to A state (i.e., structurally not relaxed). During most light absorption/emission cycles, the proton transfer eventually reverses on the ground state. However, occasionally the proton does not return to the chromophore, so the neutral chromophore is photo-isomerized to the anionic form, which involves a slower structural relaxation (e.g., side chain of Thr203 rotates to solvate and stabilize the phenolate oxyanions [3]). Of course, it is still arguable about how the proton moves and how many intermediate (I) states exist [20–25]. As pointed by Boxer et al. [18], there are several sites within the chromophore that could accept or donate a proton, and the transfer could occur within the chromophore or between the chromophore and a nearby residue of the protein. The mechanism in Figure 36.2, as proposed by Brejc et al. [3], is supported by most experimental evidence, and in the following discussions we will focus on this mechanism, although other mechanisms derived from the experimental evidence have also been proposed [21, 23–25]. Evidence of the A to I proton pathway and the nature of the I state have been confirmed by transient infrared absorption spectroscopy, which has allowed the detection of protonated Glu222 after photoconversion [26–28]. For the
Theoretical Studies of Green and Red Fluorescent Proteins
817
Figure 36.2 (a) Hydrogen bonding network and the three-proton relay mechanism surrounding the chromophore. (Reprinted with permission from [3]. Copyright 1997 National Academy of Sciences). The three-proton transfer chain includes chromophore, water W22, and Ser205 and Glu222 residues. (b) Energetics for ground and excited state proton transfers in GFP. Reprinted with permission from [35]. Copyright 2002 National Academy of Sciences
electronic ground state, Kennis et al. have found that there exist two distinct anionic ground intermediate states (denoted as I1 and I2) [7]. They absorb at 500 and 497 nm, respectively, and interconvert on a picosecond time scale. The I2 intermediate has a lifetime of 400 ps, corresponding to a proton back-transfer process that regenerates the neutral ground state. Hydrogen/deuterium exchange of the protein leads to a significant increase of I1 and I2 life time, indicating that proton motion underlies their dynamics. More recently there is even evidence pointing to an extended proton wire linking Glu222 to Glu5 on the protein surface [23]. In addition, based upon the asymptotic behavior of the fluorescence, Agmon et al. have proposed a mechanism involving a conformational change enabling the rotation of Thr203, which eventually allows the proton, via the backbone carbonyl of His148, to escape to the exterior solution [24, 25]. Theoretically, very important work has been performed for GFP, which includes electronic structure calculations of the GFP chromophore in different protonation states (some including its immediate surrounding residues) [29–31], mixed quantum mechanics/molecular mechanics calculations [8, 10] and molecular dynamics (MD) simulations [32–35]. For example, Langhoff and coworkers [31] have carried out exploratory ab initio calculations to investigate the mechanism of internal conversion via torsional motions within the chromophore. Patnaik et al. have [36] studied the relationship between molecular structure and the redshift in absorption spectra of S65G and S65T green fluorescent protein through a combination of molecular dynamics simulations with time-dependent density functional theory (TDDFT) method. Also notable is the development of a CHARMM force field for representation of the chromophore in GFP by Thiel and coworkers [34], which will help in the implementation of more reliable molecular dynamics studies. While much interest has been placed on the possible rapid deactivation pathways of the excited state due to transit through non-adiabatic crossings in GFP and its mutants, specific studies on the proton relay energetics and dynamics are rare. A dynamical simulation of the operation of the proton wire would provide insight into matters such as feasibility of the operation of this mechanism, its rate and nature (e.g., whether is concerted or sequential, and in the latter case what the order of the motion is). On the ground state, our group [12, 13] and others [37] have performed quantum DFT calculations and explored several cluster models for the ground-state proton chain transfer pathways. Mechanistic aspects of the proton transfers have been revealed, indicating a combination of both concerted and sequential aspects of the proton motions. Despite the overall “concerted” nature of the
818 Hydrogen Bonding and Transfer in the Excited State
potential profile, the configurational evolution along the reaction coordinate involves sequential movement of the protons. On the excited state, Vendrell et al. [14] have explored the overall topology of the PES for the proton wire in a sightly different cluster model for both the S0 ground state and 1pp excited state, utilizing complete active space self-consistent field (CASSCF) and complete active space with second-order perturbation theory (CASPT2) methods. Importantly, they found that the most energetically favourable pathways on both the ground and excited states follow the same ordering of movement of the protons as illustrated in Refs [12, 13, 37]. However, despite the qualitative consistency there are quantitative discrepancies in terms of the barrier height and overall endothermicity, which might be due to different methods used, different models and different symmetry constraints. While important, proper dynamical simulation is extremely complicated in GFP, since the basic system is very large and non-adiabatic transition on the excited state cannot be disregarded. In addition, the quantum effects for proton transfers are strong due to the lightness of the protons involved. Thus only very limited dynamical studies have been reported so far. Zhang and Smith [38] used a simplified quantum dynamical model to investigate the photo-absorption and excited state proton transfer processes in wild-type GFP recently. Lill and Helms [35] have implemented molecular dynamics together with a stochastic jump mechanism for incorporating the quantum proton transfer rates in an approximate manner. In their simulations it was suggested that proton-wire operation is triggered by the phenolic proton transfer from chromophore to the captive water molecule, and afterward the remaining proton transfers would occur very rapidly. Recently, Vendrell et al. [15] have performed a six-dimensional variational multi-configurational time-dependent Hartree (MCTDH) quantum dynamics investigation for the proton transfers in GFP using an empirical valence bond (EVB) fitted potential energy surface based upon 338 ab initio data [39]. Their ab initio calculations were performed at CASSCF/CASPT2 theory, assuming a planar geometry (Cs symmetry). From analysis of the energetics it seems that the first transfer of protons is from Ser205 to Glu222, which is opposite to the assumption of the phenolic proton transfer movement. Their quantum dynamics (QD) simulations indicate that proton motion in the wire is essentially concerted due to the very small potential barrier, led (as noted above) by movement of the last proton onto the Glu222 acceptor group and overall very fast, with multiple wavepacket recurrences apparent on the picosecond timescale. In a very recent study, we adopted a contrasting quantum mechanical approach from the work of Vendrell et al. [15] towards modeling the nuclear dynamics of proton chain transfer. As noted by those authors, the multiple and persistent wavepacket recurrences observed on the ps timescale in their simulations may indeed be due to the lack of any dissipative component in their model. Similarly, over tens of picoseconds the PES experienced by the protons would be expected to modulate due to relaxation effects of other modes in the chromophore and immediate surroundings. In light of these considerations, we adopt a quantum kinetic framework for the ground state I ! A proton shuttle in which the cluster is assumed to be thermalized and we compute directly the cumulative reaction probability for “first passage” multidimensional transmission from the I state to the A state. We have extended our earlier ab initio DFT calculations [12] to complete a full 3D relaxed potential energy surface for the ground state proton transfer. Dissipative conditions are explicitly incorporated in the form of absorbing boundary conditions (i.e., a complex absorbing potential) on both the I-state side and the A-state side of the proton transfer barrier. In addition to the dynamical motivations outlined above for our adoption of the present dissipative time-independent quantum framework, the approach also facilitates direct calculation of the kinetic isotope effect (KIE) for the ground state I ! A proton shuttle, a quantity of direct experimental interest that has been measured recently by Kennis et al. [7] In their femtosecond multipulse control spectroscopy study, intermediate ground states in the GFP photocycle were uncovered for the first time and the relevant kinetic lifetimes measured for both normal and deuterated proteins as mentioned above. Two distinct anionic ground state intermediates, I1 and I2, were unveiled (“anionic” here referring to the charge state of the chromophore). The KIE on the ground state for the I1 ! I2 conversion was 2, implying a structural rearrangement of some sort, while the KIE for the subsequent I2 ! A conversion was 12.
Theoretical Studies of Green and Red Fluorescent Proteins
819
The large KIE for the I2 ! A conversion suggests that this stage intimately involves the proton chain transfer as a rate-determining step. Hence, our model quantum mechanical calculations address directly the measured KIE for the I2 ! A conversion, and we will compare our calculated KIEs from this work with the measured ones later in this chapter. Following the success of GFP as a fluorescent marker, there has been a concerted research effort worldwide into engineering new fluorescent proteins (FPs) with enhanced properties for various types of applications, particularly those with far-red emission [40]. Red fluorescent proteins (RFP) provide unique opportunities for noninvasive labeling and tracking of specific cell types in living organisms in real time. Together with the development of new systems for whole-body imaging, red fluorescent proteins allow visualization of changes in target-gene promoter activity, tracking cellular movement in embryogenesis and inflammatory processes, monitoring migration of small parasites within a host, and studying important aspects of cancer, such as tumor cell trafficking, invasion, metastasis and angiogenesis [41, 42]. Furthermore, photoactivatable fluorescent proteins [43, 44] and fluorescent timers [45] expand the scope of temporally and spatially controlled tracking experiments. Thus, as a complement to the emission color palette of GFPs, intrinsic red fluorescent proteins (RFPs) with origins in Anthozoa species potentially have very significant advantages. Firstly, far-red FPs provide an additional color for multi-color labeling, as their emission is well separated from the green-yellow autofluorescence of cells. Secondly, the reduced light scattering at longer wavelengths facilitates imaging of thick tissues because animal tissues are almost translucent to far-red light. These optical properties make RFPs potentially invaluable tools for deep tissue in vivo imaging in biomedical research. However, the application of naturally occurring RFPs is complicated by oligomerization, slow maturation and low quantum yield. As a result, much experimental work has been put forward to generate improved RFPs [46]. Far-red fluorescent proteins whose emission maxima reach the 650-nm barrier have been developed, such as HcRed [47], mPlum [48] and AQ143 [49]. In particular, a recently reported far-red fluorescent protein, named Katushka, which is seven- to tenfold brighter than the spectrally close HcRed or mPlum, is characterized by fast maturation and a high pH-stability and photostability [50]. Our group is at the forefront of efforts to theoretically characterize RFP chromophore and their electronic states [51–54]. Prior to our work, the number of theoretical studies on isolated RFP chromophores is very limited [55]. There are currently few whole-protein modeling studies of RFPs present in the literature, in particular for the excited state studies. Thus RFP physics on the molecular level is poorly understood. In recent years we have calculated the bond rotation profiles for model DsRed chromophore using DFT method [52], and have performed TDDFT calculations on model RFP Rtms5H146S chromophore [51]. We also employed complete active space self-consistent field (CASSCF) and multireference Rayleigh–Schr€odinger second-order perturbation theory (MRPT2) methods to characterize the bridge photoisomerization pathways of a model red fluorescent protein (RFP) chromophore (see, for example, Figure 36.3 for the structure of DsRed) [53]. Going beyond the cluster model, it is essential to incorporate the solvated protein environment to gain the most realistic representation of the potential energy surface (PES). Very recently we have initiated quantum mechanical/molecular mechanical (QM/MM) modeling for RFP through collaboration with Thiel’s group, which will help gain a better fundamental understanding of how RFPs’ fluorescent properties are controlled and mediated by the protein environment. To develop insights into the mechanistic aspects that govern the function of promising FP candidates, we have also been collaborating with structural biologists for the fluorescent protein and related studies. This has led to a series of publications [51–54, 56–58] that have elaborated both key structural and mechanistic issues relating to the RFP HcRed [57] and pocilloporin Rtms5H146S [59, 60] in recent years, and provides an important experimental context within which the knowledge gained from our calculations can aid in furthering experimental development of new engineered proteins. In this regard simulations and modeling can add fundamental understanding to the direction aspect of a directed evolution approach, since the physical interactions and mechanisms introduced by mutations that yield improved RFPs are unknown, and
820 Hydrogen Bonding and Transfer in the Excited State
Figure 36.3 Structure of red fluorescent protein (DsRed) and its chromophore. Reprinted with permission from [53]. Copyright 2007 American Chemical Society
mutagenesis studies alone cannot reveal them. Thus an understanding of these interactions would streamline development and lower the associated cost. For example, our computational studies have tackled the basic question of how structural rigidity of the chromophore varies between the GFP and RFPs and which residues impact most on this, and the recent benchmark studies from our collaborative team provide strong evidence that this is a crucial factor in controlling fluorescent quantum yield. The rest of this chapter is arranged as follows: the description of the method of calculation is given in Section 36.2, and a discussion of the results follows in Section 36.3. Conclusions and future outlook are summarized in Section 36.4.
36.2 Method of Calculation We investigate the fluorescence mechanism in green fluorescent protein (GFP) and in red fluorescent protein (RFP) through a combination of computational methods. For GFP, most of our studies are based on cluster models that consider only the residues close to the chromophore. For RFP, we focus on the truncated chromophore model initially; extension to include the whole protein environment using QM/MM method is currently in progress. 36.2.1 Computational methodology for GFP proton wire 36.2.1.1 Cluster Model for GFP Proton Wire Three-proton transfer models have been built and one example is shown Figure 36.4, where the water and methanol have been chosen as the bridge molecules and acetic ion as the terminal acceptor. The model is based on the GFP protein data file 1emb.pdb [3], and all other residues and water molecules have been removed. The cluster model was employed to mimic the proton chain transfer within GFP as shown in Figure 36.2, in which the methanol represents the residue Ser205, whereas the acetic ion represents residue Glu222. The acetate moiety as acceptor group is more basic in character than the more heavily truncated formate, and thus more representative for the residue Glu222. The geometry optimization of the model system has been carried out for the electronic ground state using the cc-pVDZ basis set at the hybrid density-functional theory (DFT) level of Becke and Lee, Yang and Parr (B3LYP) [61–63]. The three O–H bond lengths are labeled as r1 for O–H bond of phenol in chromophore, r2 for O–H bond of water and r3 for O–H bond of methanol, which represent the proton coordinates in the proton transfer chain. We noticed that there are several different types of cluster models being explored for GFP, and we focus on two of them, namely, fully relaxed model and rigid model, since we have performed calculations using the two
Theoretical Studies of Green and Red Fluorescent Proteins
821
Figure 36.4 One cluster model for the GFP chromophore and proton wire as used in quantum dynamics simulations [82]. Reproduced by permission of the PCCP Owner Societies
kinds of cluster models in our work. In the fully relaxed model, all other degrees of freedom except three O–H bond lengths are fully relaxed, whereas in the rigid model the geometry of the model system is held fixed at their values in the minimum in the ground electronic state when moving the three protons between donor and acceptor O atoms in order to construct three-dimensional (3D) PESs. The rigid model has also been employed by Vendrell et al. [14] in a slightly different context (see below for comparison). They also explored a partially relaxed model very recently in which only three O–H bond lengths and the three donor–acceptor distances are allowed to move (six-dimensional model) [15]. Such model calculations are necessary and practical in a quantum mechanical sense, given the sheer size of the system. 36.2.1.2 Ground State DFT Calculations in Fully Relaxed Model In our calculations of the three-dimensional PES grid data for ground state proton transfers, the grid points are defined by constraining three O–H bond lengths along the chain, namely, the chromophore phenolic O–H bond (labeled as r1), the breaking/forming O–H bond of water (labeled as r2) and the O–H bond of methanol (labeled as r3). All other degrees of freedom in our model are fully relaxed. This relaxation of all other modes in the cluster most likely leads to some degree of underestimation of the effective proton transfer barrier, whereas the “frozen” approach in which no other modes are relaxed [14, 15] most likely overestimates it (see also discussion below). Density function theory (DFT) has been used to complete the 3D potential energy surface scan for the ground electronic state, as implemented in the Gaussian 03 package [64]. This level of theory has been discussed and compared against HF and MP2 methods in relation to a two-step model of the GFP proton chain transfer [12, 13]. We carried out the DFT calculations for three-dimensional PES in order to construct a realistic PES for the quantum dynamics calculations for the three proton transfer reactions, which need to span a much larger configuration space. The constraint geometries have been optimized under B3LYP/ 6-31g theory level followed by the single point energy calculation under B3LYP/6-31 þ g level. The step size for the scan is 0.05 A with 16 steps for each O–H bond, and the initial bond length is 0.95 A and the final length is 1.70 A. In total 16 16 16 (¼ 4096) ab initio potential energy points have been obtained. We have not found it necessary to fit this PES with analytic forms. Rather we evaluate the potential as necessary for the construction of the discrete variable representation (DVR) Hamiltonian using cubic spline interpolation.
822 Hydrogen Bonding and Transfer in the Excited State
36.2.1.3 Excited State ab initio Calculations in a Rigid Model Utilizing the cluster model for the proton wire for determination of the excited state PES for proton chain transfer, we have performed high-level ab initio calculations to model the GFP excited state proton transfer (ESPT), which constitutes the first stage of the photocycle after excitation of the chromophore. To determine the molecular orbitals (MOs), complete active space self-consistent field (CASSCF) calculations have been carried out. After determining the MOs, we have performed multi-reference (MR) configuration interaction (CI) calculations to obtain low-lying potential energy curves. In this method, multi-reference (MR) single- and double-excitation (SD) CI is employed, in which the configuration state functions (CSFs) were generated by single and double excitations with respect to the reference configuration used in the CASSCF calculations. Since the ab initio MRCI calculations for a model of this size (Figure 36.4) are quite challenging, we have used a moderate basis set, that is, Dunning’s cc-pVDZ (correlation consistent, polarized valence, double zeta) basis set [65]. As for CASSCF active space, six MOs from 83a to 87a should be included in the active space, since they are important to describe the low-lying electronic states and should be occupied by six electrons in CASSCF (6 MOs/6 electrons). The MOLPRO 2002.6 program package [66] was used to obtain the potential energy curves and surfaces of electronic ground and excited states. Our ab initio calculations span a larger configuration space, to construct a 3D PES for the quantum dynamics calculations for the three proton transfer reactions later. The step size for the scan is 0.1 A for most data points with 12 steps for each O–H bond, and the initial bond length is 0.7 A and the final length is 1.70 A. In total 1728 ab initio potential energy points have been obtained. 36.2.1.4 Quantum Dynamics (QD) Calculations Two kinds of QD methods have been employed in our simulations, namely, time-independent (TI) and timedependent (TD) QD methods. In the former method, we solve an eigenvalue problem to obtain the cumulative reaction probabilities for proton transfer reactions, from which the rate constants and thus kinetic isotope effects (KIE) can be extracted. In the latter method, we use the TD wavepacket split-operator propagation technique (the Hamiltonian itself is time-dependent when laser field is explicitly treated) to simulate the real time proton transfer dynamics, which is especially useful for simulating the femtosecond pump–probe spectroscopy. While TI and TD methods have their own advantages and disadvantages in terms of describing proton transfer reactions, we focus on the TI method in this chapter. The TI method we have implemented in the current modeling of the kinetic isotope effect (KIE) for the GFP proton chain transfer is based on a three-dimensional model, which includes the line coordinates for the three protons moving between the donor and acceptor oxygen atoms (e.g., Figure 36.4). We calculated a fully relaxed potential energy surface as a function of these three specific coordinates as mentioned above, which feeds into the Hamiltonian for the quantum dynamics calculation. Then we compute the cumulative reaction probability as a function of energy, N(E), for this system and then Boltzmann average to obtain the temperature-dependent equivalent, N(T). An important component of our model Hamiltonian is the use of a complex absorbing potential (CAP) in order to impose dissipative boundary conditions on either side of the proton transfer barrier. This allows us to directly access the rate of proton chain transfer connecting the neutral and anionic chromophore states without the complication of wavepacket recurrences. Within our formulation, then, the rate constants and the KIE are given by:
kH=D ðT Þ ¼ 2p hQH=D ðT Þ
1
ð þ¥ ¥
dEeE=kB T NH=D ðEÞ ¼
NH=D ðT Þ 2phQH=D ðT Þ
ð36:1Þ
Theoretical Studies of Green and Red Fluorescent Proteins
823
and: k H ðT Þ QD ðT Þ NH ðT Þ KIEðT Þ ¼ ¼ k D ðT Þ QH ðT Þ ND ðT Þ
ð36:2Þ
Our modelling shows that it is crucial to adopt a direct numerical approach to calculation of the partition functions also, since the PES for the proton wire is highly anharmonic and this has a very important impact on the predicted KIE. We emphasize that the key components in our approach involve exact three-dimensional quantum dynamical calculation of the energy-dependent reaction probability [67], given by equation (36.3), for the triple-proton chain transfer in the cluster model followed by Boltzmann averaging to obtain the thermal quantities, including full quantum evaluation of the three-dimensional reactant partition functions: 1= 1 1 ^ «E 1 ^«p H ^ þ i^«E 1 ^«r=2 N ðEÞ ¼ Tr « ^r 2 Hi^ 4
ð36:3Þ
36.2.2 Computational methodology for model chromophores in FP 36.2.2.1 Ground State Calculations We employ truncated models for FP chromophores (e.g., DsRed, Rtms5H146S). For ground state calculations, the models were optimized with B3LYP density functional theory using a 6-31 þ G basis set. For coordinatedriving potential scans, we optimized the models under the constraint of a constant angle of the dihedral. At each of the optimized geometries we calculated the energy using B3LYP DFT and a 6-31 þþ G basis set. All calculations were performed using the GAUSSIAN03 package [64] 36.2.2.2 Excited State Calculations For the excited state calculations for model chromophores in several fluorescent proteins (e.g., green fluorescent protein-GFP, kindling fluorescent protein-KFP and red fluorescent protein-RFP), we also use a model of the FP chromophore that is truncated to include only the chromophore. Connections that would be made to the protein backbone are terminated by hydrogen atoms. For excited state calculations, we have analysed the electronic structure of the S0 and S1 states of more substantial models and determined that truncation does not change their nature. The structures of these chromophores are considered in their Z and E configurations. The reader should note that the Z isomer is often referred to as cis in the fluorescent protein literature, and the E isomer as trans. Our electronic wave functions were generated via dynamically correlated multistate electronic structure computations built upon a state-averaged [68] complete active space self-consistent field [69] (SA-CASSCF) reference wave function, using geometries optimized at the SA-CASSCF level. To evaluate energies at these geometries, we make use of multireference Rayleigh–Schr€odinger second-order perturbation theory (MRPT2) [70]. In this perturbation theory, the reference space is remixed as the perturbation is applied, and the states are obtained by diagonalization of a symmetrized perturbed effective Hamiltonian. MR-MS-RSPT2 provides size-consistent energies. Our choice of orbitals was guided by orbital energies and occupation numbers that were obtained in a preliminary battery of self-consistent field calculations. We are interested in bridge torsion. To obtain a broader view of the photoisomerization, we calculated energies and properties along coordinate-driven slices through the potential energy surfaces. Coordinatedriving potential surface scans were generated by fixing one (or both) of the two bridge dihedrals (the driven
824 Hydrogen Bonding and Transfer in the Excited State
coordinates) of the model chromophore and minimizing all other degrees of freedom subject to this constraint. All of the results for the excited state calculations were obtained with the MOLPRO program [66]. 36.2.3 QM/MM method for RFP 36.2.3.1 SCC-DFTB/MM Molecular Dynamics
The crystal structure of HcRed protein (protomer B) with a resolution of 2.1 A provided the starting point of our calculations [57]. The addition of missing hydrogen atoms was determined using the HBUILD facility in CHARMM at pH 7 and the definition of protonation states of titratable residues were performed using the PROPKA method [71, 72]. The systems consisting of protein and crystallographic water molecules were solvated in a sphere of radius 30 A formed of TIP3P water molecules [73] and the water molecules too close to existing atoms were deleted. We performed 12 hydration cycles until the numbers of water molecules were approximately constant. Finally, the MD production runs without restrains were performed for 500 ps to complete the preparation of the complexes. The systems were heated from 50 to 300 K through increasing the temperature by 0.1 K each step. In the MD simulations the chromophore was described by the semi-empirical SCC-DFTB (self-consistent charge density-functional tight-binding) method [74], while the protein environment has been treated using the CHARMM force field [75]. The QM/MM boundary at the two covalent bonds was treated by generalized hybrid orbital (GHO) method [76]. 36.2.3.2 DFT/MM Calculations Four snapshots for the initial structures for QM/MM optimization were randomly taken from the 500 ps MD trajectories of cis chromophore in HcRed. The structures with the trans conformation of chromophore were derived from the optimized cis isomer by a manual rotation of the hydroxyphenyl group and complete reoptimization of the geometry parameters at the same calculational level. In the QM/MM calculations, the QM part was treated by the B3LYP density functional method with the basis sets of SV(P) and TZVP, and the MM part was described by the CHARMM force field. An electronic embedding scheme was adopted in the QM/MM calculations [77]. Hydrogen linker atoms with the charge shift model were employed to treat the QM/MM boundary. The TURBOMOLE program was used for the QM treatment in the QM/MM as well as in the pure QM calculations. The CHARMM force field was run through the DL_POLY program to handle the MM part of the systems. The QM/MM calculations were performed with the ChemShell package [78, 79] that integrates the TURBOMOLE and DL_POLY programs and also performs geometry optimization with the HDLC optimizer [80].
36.3 Results and Discussion 36.3.1 Proton transfers in GFP Our work in this regard focuses on the proton transfer dynamics in green fluorescent protein (GFP) using rigorous quantum mechanical methods, based on several cluster models. We performed density functional theory (DFT) calculations for the ground-state proton chain transfer pathway in model GFP [12, 13, 81]. The mechanistic conclusions arising from studies in our laboratory [12, 13] and others [14, 15, 37] may be summarized as follows: (i) Based on all the cluster models explored so far, proton transfer on the ground state as well as on the excited state is predicted to occur via a single barrier, implying a concerted mechanism. (ii) Despite the presence of a single barrier, implying a concerted kinetic process, analysis of the minimum energy pathway (MEP) showed clear signatures of largely sequential movement of the protons. The ordering of
Theoretical Studies of Green and Red Fluorescent Proteins
825
movement of the protons along the reaction pathway corresponds to a “pulling” rather than a “pushing” mechanism – for example, for the neutral to anionic proton chain transfer, the first proton movement is from the bridging Ser 205 moiety to the accepting Glu 222 group, followed by the second proton moving from the bridging water to Ser 205, and the phenolic proton on the chromophore being the last one in the chain to move. In other words, the mechanism of proton chain transfer is impacted heavily by the presence of the charged acceptor group at the end of the chain. The mechanistic picture arising from analysis of the MEPs is hence a blend of the classic “concerted” and “stepwise” mechanisms. As highlighted in the sections above, the actual dynamical processes in GFP are very complicated, and quantum-mediated kinetics of proton chain transfer have been found to play a crucial role in the photocycle of the GFP. Thus, moving beyond the static reaction pathway picture, to explore the true proton transfer dynamics we have performed nuclear quantum dynamics (QD) simulations, which is the only reliable and accurate way to describe the ultrafast dynamics involved. We have used both time-dependent split-operator propagations to simulate the ultrafast femtosecond pump–probe dynamics [38] and time-independent methods to address the thermal rates and KIEs for proton chain transfer on the longer timescale of 1 ps to 1 ns. The TD method can directly probe in real time the primary proton transfer processes within the chromophore and its immediate environment. In a recent preliminary study [38], we explored the time-dependent idea for the one-dimensional excited state proton transfer case in GFP. This work is also partly motivated by the femtosecond experiments performed for GFP [6, 7] since femtosecond experiments can provide new insights into the microscopic mechanisms of the photodynamics in GFP. In the frequency-resolved femtosecond pump–probe experiment in wt-GFP, V€ ohringer et al. [5] studied the microscopic origin of the dispersive kinetics and the molecular mechanism of the primary events involved in the excited state proton transfer (ESPT) dynamics. They proposed the energetic scheme with additional two higher-lying electronic configurations, upon which we have built the model potential energy surfaces for quantum dynamics calculations. While our model quantum dynamics calculations can explain the origin of the early-time stimulated emission observed in experiment, the predicted KIE for the ESPT is too small (about 1.25) compared with the experimental one of about 6.0 [18]. Following the TD quantum dynamics work, we have very recently performed TI quantum dynamics simulations to predict the rate constants of the ground state proton transfer reactions and thus the KIEs. Our main goal is to compare the simulations with experimental results (in particular KIE) in order to understand the underlying reaction mechanisms. We use a three-dimensional potential energy surface for the ground state proton movements generated using DFT. Using the TI approach we have achieved a remarkable nearquantitative agreement with the very large experimental KIE of 12 at room temperature for the ground state proton chain transfer (Figure 36.5) [82]. Ours and others previous attempts to model the KIE have generally predicted KIEs in the range 1–3 (see Table 36.1 for a comparison). For example, in a very recent report [15], the true movements of the proton transfers have been explored through MCTDH quantum dynamical calculations. In the MCTDH simulations, within the initial phase of short-time dynamics, a clear isotope effect of about 1.5 appears for the H/D substitution on the excited state. However, an inverse KIE in the long time range was found, which we believe is artificial due to the reflections of the wave-packets without appropriate absorptions in the product region. Thus our latest result suggests that our reduced-dimensional time-independent quantum dynamics model offers considerable predictive power not only for the GFP but potentially also for other important biological proton wires. While the KIE predicted by our first implementation of this TI approach for the GFP is in very good accord with the experimental measurement (and is the only model to have successfully achieved this), the absolute rate constants for proton and deuteron transfer are still two orders of magnitude too fast. This implies that our PES, which allows all other degrees of freedom in the cluster to relax as a function of the proton coordinates, is predicting absolute barriers that are too low (although the overall shape and anharmonicity of the surface are apparently reasonably well represented). In practice, it is unlikely that slower modes will be able to fully relax on the timescale of operation of the proton wire. However, we know already that freezing all other coordinates
826 Hydrogen Bonding and Transfer in the Excited State
Figure 36.5 Predicted KIE for GFP ground state proton chain transfer. Quantum model results: solid line. Experimental result: circle. Standard translational model: triangle [82]. Reproduced by permission of the PCCP Owner Societies
Table 36.1 Comparison of kinetic isotope effects (KIEs) from both calculations and experiments. MCTDH result is from Ref. [15], whereas the experimental data is from Table 1 in Ref. [7]. Other data are from our work TI QD Ground state (I2 to A) Ground state (I1 to I2) Excited state
MCTDH
TD QD
Exp
1.25
12 2 6
15.5 1.5 (early)/inverse (later)
yields barriers that are much too high (see below for the results from rigid model). This implies that a static averaging approach – whereby the proton PES is calculated with all other modes frozen and then averaged by configuration sampling over the other modes – might result in barriers that are too large. Hence, an intermediate approach involving only partial relaxation of other modes will be necessary – and this strategy must necessarily be informed by dynamical studies. Our proposal is to implement ab initio molecular dynamics calculations for the cluster model in order to explore this question of how to determine the most appropriate way of computing an effective PES for the proton motions in a biological proton wire (using GFP as our prototype). This will be an essential prerequisite for the quantitative calculation of absolute rates of proton chain transfer in such systems. To illustrate this more clearly, we have very recently performed a series of constrained quantum chemical minimum energy pathway calculations for the ground and first excited electronic states of a cluster model of the GFP photocycle; each MEP corresponding to a different assumption as to which modes are relaxing along the reaction coordinate. Figure 36.6 presents the ground state energy profiles computed by density functional theory (DFT) at the B3LYP/cc-pVDZ level for a set of four different relaxation models: (i) in which all coordinates of the cluster model have been allowed to relax, yielding the true (fully relaxed) MEP for this cluster model; (ii) in which all cluster modes except the internal modes of the chromophore are allowed to relax (the chromophore is constrained to remain in a configuration corresponding to the optimized neutral geometry – excepting of course the phenolic proton which is transferred to W22); (iii) in which only six coordinates are relaxed, corresponding to the donor–acceptor oxygen atoms of the water, Ser205 and Glu222 and the three protons (all other modes being fixed at the optimized neutral geometry); and (iv) in which only the
Theoretical Studies of Green and Red Fluorescent Proteins 48.9
50
Relative energy (KJ/mol)
827
43.9 iv
40
34.7
33.7
iii
30 20
8.4
10 0
0.0
4.9
4.2
ii
neutral -11.8
-10
i
-20
Reaction Path
Figure 36.6 Energy profiles for the ground state proton transfer computed at the B3LYP/cc-pVDZ level for the cluster model
three protons (linear coordinates along the lines between donor and acceptor oxygens) are allowed to relax. The fully relaxed calculation (i) in Figure 36.6 corresponds to plot (1c) in Figure 2 of our earlier work [12], with the only difference being a somewhat larger basis in the present calculations. Figure 36.7 presents an analogous set of energy profiles for the first excited state proton transfer computed at the modest level of CIS/cc-pVDZ by constrained optimizations to locate the saddle-points and final anionic geometries. Clearly, while the absolute values of the barriers would likely decrease substantially with a higher level of theory and a larger basis, a similar trend of sensitivity in the barrier height to the details of the relaxation model is seen as for the ground state results of Figure 36.6. Thus the effective barrier for proton chain transfer is found to be markedly sensitive to the choice of the closely dynamically coupled group of degrees of freedom. The complexity of this choice is a particularly vexatious feature of proton wire systems, as exemplified by the GFP. These results demonstrate that the choice of closely coupled degrees of freedom that will be explicitly incorporated in quantum simulations (or tunneling-corrected transition state theory calculations) must be carefully made and will need to be informed by dynamical studies in appropriate cluster models. This group of
Relative energy (KJ/mol)
120
110.9
100
101.8
80
80.9
66.0 70.6
60
35.3
40
36.6
iv iii
ii
20 0
0.0 neutral
-20
-36.5 -40
i
Reaction Path
Figure 36.7 Energy profiles for the first excited state proton transfer computed at the CIS/cc-pVDZ level for the cluster model
828 Hydrogen Bonding and Transfer in the Excited State
Figure 36.8
Molecular orbitals (MOs): 86a and 87a
closely dynamically coupled modes will then be treated explicitly in the quantum simulations, based on a computed ab initio potential energy surface (ideally with subsequent configurational averaging over the “environmental” modes). This is currently under investigation. Lately, we have also extended our investigations into the excited state proton transfers using high level ab initio methods and exact nuclear quantum dynamics simulations. We employ rigid cluster model based on the optimized structure for ground state A with neutral chromophore as shown in Figure 36.4. We performed high level complete active space self-consistent field (CASSCF) and multi-reference configuration interaction (MRCI) calculations to generate three-dimensional potential energy surfaces for the three proton transfers. In total 12 12 12 ab initio data points are used for the construction of the 3D PES. Our CASSCF active space consists of six molecular orbitals (MOs): 83a to 88a, with 87a being HOMO, and 88a being LUMO. As examples, in Figure 36.8 we have plotted some selected molecular orbitals (MOs) corresponding to optimized state A structure with neutral chromophore. The electronic structure is characterized by the 86a, 87a and 88a MOs, and in particular the 86a and 87a MOs feature the p complex conjugate system. Analysis of the molecular orbitals indicate that the first excited state is of 1pp character, which is the photoactive state. If further excitations (e.g., 1ps) are of interest, we will need to include more MOs in our calculations. Figure 36.9 presents some selected 2D contour plots of the 3D potential energy surfaces for three proton transfers in the cluster model for the excited state. The plots correspond to the proton transfer leading coordinate r3 ¼ 1.5, 1.2 and 1.0 A for (a)–(c), respectively. Following the pictures from reactant state to product state from (a) to (c), we can see that as the third proton moves from the reactant state to transitional state and finally to product state, the well depths of the product state are becoming deeper and deeper, thus making it easier for other two protons to transfer from reactant state to product state. The results of 3D potential energy
(b) r3=1.2 angstrom
1.4
1.2
r2 (ang
1.0
strom )
0.8 0.8
4 1.6
1.4
1.2
r2 (ang
1.0
strom
)
0.8 0.8
4 6 8 10 12 14
14 12 10 8 1.8 1.6 1.4 1.2 1.0
6
m)
Ve(eV) m)
m)
tro
1.6
an gs
4
1.8 1.6 1.4 1.2 1.0
6
tro
1.8 1.6 1.4 1.2 1.0
6
8
an gs
8
10
r1 (
Ve(eV)
12
10
r1 (
Ve(eV)
12
4 6 8 10 12 14
14
tro
14
4 1.6
1.4
an gs
4 6 8 10 12 14
(c) r3=1.5 angstrom
1.2
r2 (ang
1.0
strom
0.8
r1 (
(a) r3=1.0 angstrom
0.8
)
Figure 36.9 Potential energy surfaces for three proton transfers in the electronic excited state of the cluster model: r3 ¼ (a) 1.0, (b) 1.2 and (c) 1.5 A
Theoretical Studies of Green and Red Fluorescent Proteins
829
Table 36.2 Barrier heights DH and relative energies DE in eV; indicates excited state. CASSCF and CASPT2 results are from Ref. [14], and experimental estimation is from Ref. [6] MRCI CASSCF CASPT2 DFT EXP (GFP)
DE ¼ EI – EA
DH ¼ ETS – EA
DE ¼ EI – EA
DH ¼ ETS–EA
0.28 0.73 0.90 0.18 0.07
0.59
0.33 0.24 0.34
0.48
0.24
0.12
0.197 0.24
surface show that the correlations of movement between protons are very important, and the transfer of the more energetically favorable proton (e.g., proton 3) facilitates the transfers of the more energetically unfavorable protons (e.g., proton 2 and 1). The results of the 3D MRCI potential energy surface also further support the mechanistic picture of the concerted proton transfer processes with a single barrier; however, the sequential movement of the protons can be identified within the concerted picture. Our observations in this work agree with previous principle mechanistic conclusions arising from literature of ours [12, 13] and others [14, 15, 37]. This rigid cluster model generates energetics that qualitatively agree with the experimental findings [6], and in Table 36.2 we have listed the barrier heights DH and relative energies DE from different sets of calculations and from some available experimental estimations. In this table the unit is eV, and represents the excited state. From this table we can see that MRCI calculations produced the relative energy of 0.28 eV between anionic ground state I and neutral ground state A and the barrier height of 0.59 eV for the ground state, both of which are higher than the experimental estimation (the experimental result is roughly 0.24 eV [6] for barrier height and is roughly 0.07 eV for the relative energy [6]). For the excited state a similar trend is found; for example, the calculated barrier height and relative energy are higher than the experimental ones, which might be due to the rigid model employed in which no relaxation is allowed in our MRCI calculations. The experimental estimations for the energetic terms are based upon some available but limited experimental data for the ground and excited state, mainly from three experimental results, that is, picosecond spectroscopy of Boxer et al. [18], hole-burning spectroscopy of Creemers et al. [19] and frequency-resolved femtosecond pump–probe spectroscopy [6]. Based on these experimental results, illustrative PESs have been given by V€ohringer et al. [6], with which our comparisons are made. In this table we also list the relative energies for both ground and excited state from CASSCF and CASPT2 calculation [14], in which a sightly different model and a co-planar Cs symmetry constraint are employed. For the excited state, CASPT2 predicts a fairly similar relative energy with MRCI calculation, whereas CASSCF predicts that the energy in the anionic state I is higher than the neutral state A, which disagrees with other calculations and the experimental estimation. For the ground state, both CASSCF and CASPT2 produced larger relative energies than MRCI calculations. We also list the DFT results from our 3D PES calculations with constrained relaxation for the ground state [82], and, as we can expect, it predicts a barrier height that is lower than the MRCI calculation and the experimental estimation. 36.3.2 Internal conversion mechanism in FP chromophores The preferred optical window for deep tissue imaging is in the far-red and near-infrared (650–1100 nm) because this window largely avoids the (optical) absorptions of melanin and hemoglobin at shorter wavelengths and vibrational absorptions of water at longer wavelengths. Thus, following the success of GFP as a unique fluorescent label, a search has been going on for new fluorescent proteins, particularly those with far-red emission, which could enable such in vivo deep tissue imaging. RFPs have been discovered in
830 Hydrogen Bonding and Transfer in the Excited State
numerous coral species and it is now known that RFPs and their homologues are responsible for much of the coloration seen on coral reefs [83]. Current evidence points to a possible photoprotective role for these proteins [84]. Red fluorescent proteins possess a chromophore similar to that of GFP, but with additional chemical modifications that occur auto-catalytically during a complicated maturation process (Figure 36.3). At present, the detailed origin of the redshift of these proteins is not completely understood, and our goal of the RFP modeling project is to further the understanding of the molecular physics of red fluorescent proteins. Through an understanding of the mechanisms leading to the redshift, the design of further redshifted variants may be possible, leading to applications in deep-tissue biomedical imaging. Furthermore, the application of RFPs is quite often hampered by a low quantum yield. Through an understanding of the deactivation processes that compete with fluorescence in these systems, brighter variants may be designed. The tools we used for RFP include ground and excited state electronic structure methods coupled with molecular dynamics methods for multiple states. In our early study [52], we carried out the first comparative examination of internal rotation barriers in acylimine (model R0) and peptide (model G0) substituents in model red fluorescent protein chromophores in DsRed. Model G0 (G-green) and model R0 (R-red) represent the immature and mature DsRed chromophore, respectively. Figure 36.10 shows the results from the potential energy surface scans generated by coordinate driving of the uCNCO dihedral in R0 and G0 models. The definition of the uCNCO dihedral is given on the righthand side. The highest calculated energy for R0 is 5.5 kcal mol1 (uCNCO ¼ 105 ) relative to the cis acylimine optimized geometry and the highest energy for G0 is 16.3 kcal mol1 (uCNCO ¼ 90 ) relative to the trans peptide optimized geometry. These energies represent lower bounds to the true transition state energy. Our results indicate that the barrier to trans–cis isomerization of the substituent is much lower if it is an acylimine instead of a peptide, providing prima-facie evidence that acylimine formation precedes trans–cis isomerization in DsRed chromophores. Our study provides partial evidence for the emerging picture that RFPs with low quantum yield do not maintain the chromophore in a structurally rigid conformation as does the GFP – thus facilitating bridge twisting and internal conversion via conical intersections and a low fluorescence quantum yield. We also performed TDDFT calculations on model RFP Rtms5H146S chromophore [51], which provided strong supporting evidence for the assignment of the protonation state of the chromophore of the yellow form of Rtms5H146S at low pH.
Figure 36.10 Internal rotation profile of models R0 and G0. The internal rotation potential scans about the central dihedral of the acylimine (model R0, filled upward triangles/solid line) or peptide (model G0, open downward triangles/dashed line). Energies were evaluated using B3LYP DFT and a 6-31 þþ G basis set at geometries optimized with B3LYP DFT and a 6-31 þ G basis set under the constraint of constant uCNCO. Energies are in kcal mol1 and angles are in degrees. All energies are referenced to the lowest energy conformer for each model. Peak heights for each curve are indicated. Solid (R0) and dashed (G0) lines are simple spline curves. Reprinted with permission from [52]. Copyright 2006 Elsevier
Theoretical Studies of Green and Red Fluorescent Proteins
831
Figure 36.11 Characterization of energies and charge distributions in isomerizing RFP chromophores. Reprinted with permission from [53]. Copyright 2007 American Chemical Society
In a more recent study [53], we used CASSCF and MRPT2 methods to characterize the bridge photoisomerization pathways of the model red fluorescent protein (RFP) chromophore, DsRed. Figure 36.11 shows energy profiles and charge distribution profiles on the anionic RFP chromophore as a function of the key isomerization coordinates that lead to internal conversion via twisted conical intersections. The study indicates that photoisomerization of the imidazolinone-bridge bond is suppressed by the induction of a high barrier on the S1 surface, whereas photoisomerization of the phenoxy-bridge bond is favored via stabilization of the S1 pathway and the convergence of the pathway with an S0/S1 conical intersection seam at intermediate values of the bond torsion. These effects are due to the action of the strongly electronegative acylimine on the twisted intermolecular charge-transfer (TICT) states that are encountered along the pathways. Our results are further evidence of the importance of TICT states and charge-transfer intersections in the control of photoisomerization processes in fluorescent protein chromophores. Indeed, bridge photoisomerization is an important internal conversion mechanism in other fluorescent proteins, as very recently reported by our studies in the anionic green fluorescent protein and kindling fluorescent protein chromophore models [54]. In this report [54], the ground and excited state electronic structures and the potential energy surfaces of two model chromophores [representing green fluorescent protein (GFP) and kindling fluorescent protein (KFP), respectively] have been characterized. The two fluorescent protein chromophores differing by a single substitution demonstrated qualitative differences in the potential energy surfaces that indicate inversion of bond selection in the photoisomerization reaction. Bond selection is also modulated by whether the reaction proceeds from a Z or an E conformation. These configurations correspond to fluorescent and non-fluorescent states of structurally characterized FPs, including some that can be reversibly switched by specific illumination regimes. We explain the difference in bond selectivity via substituent stabilization effects on a common set of charge-localized chemical structures, as different combinations of these structures give rise to both optically active (planar) and twisted intramolecular charge-transfer (TICT) states of the molecules, and offer an experimental proposal to test our hypothesis.
832 Hydrogen Bonding and Transfer in the Excited State
Theory and molecular simulation together with structural characterization have a uniquely powerful role to play in achieving insights into the mechanistic aspects that govern the function of promising FP candidates. Without this knowledge to guide design principles, the strategy for directing evolutionary approaches utilizing random mutagenesis must remain primitive – based on measured properties rather than insight. The recent determination of structure of the RFP HcRed [57] – one of only a handful of RFPs that have been structurally characterized to date – was achieved by a combination of X-ray structure and optical spectroscopy aided by our quantum chemical modeling. More recently, a remarkable enhancement in fluorescence efficiency at high pH of the weakly fluorescent RFP Rtms5 has been reported, structurally characterized and mechanistically rationalized by the same collaborative team [60]. The mechanistic picture that emerges hinges on the protonation state of the chromophore and its surrounding residues, which not only impacts the electronic structure, and consequently the absorption and emission wavelengths, but also the structural stability of its more highly fluorescent isomeric forms. Following the early quantum chemical modeling of these chromophores in our laboratory that lent support to the experimental interpretations, we have performed wholeprotein QM/MM calculations to elaborate in detail the structural properties controlling these protonation states and the isomerization propensity of the chromophore. The insights accruing from this theoretical work have significant potential for aiding future design efforts. 36.3.3 QM/MM studies in RFP Very recently we have investigated the far-red fluorescent protein HcRed using molecular dynamics (MD) and QM/MM calculations, and the preliminary outcomes of this study are summarized in Figure 36.12. Figure 36.12(a) shows the monomeric version of HcRed, embedded within a water cell in silico, which is then relaxed and equilibrated. Figure 36.12(b) shows the hydrogen bonding network surrounding the chromophore, embedded within the protein cavity of HcRed as determined by our calculations. Figures 36.12(c) and (d) show, respectively, snapshots of the two different isomeric structures of the chromophore (cis and trans) as obtained from our QM/MM molecular dynamics simulations. The results of our QM/MM calculations, implemented with density functional theory (DFT) for the QM part and the CHARMM forcefield for the MM part, demonstrate that different protonation states of glutamines (Glu214 and Glu146) nearby the chromophore are crucial in determining the stability of cis and trans isomers, with important ramifications for the fluorescent properties of HcRed [85]. In more detail, firstly, we carried out SCC-DFTB/MM MD simulations for the anionic form chromophore in which Glu214 is protonated and Glu146 is deprotonated (defined as model B). Table 36.3 presents the
Figure 36.12 (a) The far red protein HcRed in solvent, (b) hydrogen network around the cis conformation of chromophore of MD runs; hydrogen network at DFT/CHARMM level of (c) cis and (d) trans conformations in HcRed
Theoretical Studies of Green and Red Fluorescent Proteins
833
Table 36.3 Important dihedral angle ( ) of QM region (SCC-DFTB method) and bond distances (A) around the cis conformation of chromophore with protonated Glu214 and deprotonated Glu146 of HcRed of the average 500-ps MD production runs
MDa (A)
Dihedral angle ( ) N2_CA2_CB2_CG2 CA2_CB2_CG2_CD1 Bond distance (A) O_NE2(Gln107) O2_NH2(Arg93) OH_OG(Ser144) N2_OE2(Glu214) N2_NE2(Gln40)
Dm0 a (A)
6.4 6.2
Exp [57] (A)
6.4 2.2
3.054 2.676 2.856 3.447 3.231
0.0 8.4
0.035 0.514 0.255 0.481 0.048
3.019 3.190 2.601 2.966 3.183
a MD ¼ average MD values from the average 500-ps MD production runs; Dm0 ¼ standard deviations between average MD values and x-ray 1YZW pdb
important dihedral angles ( ) and bond distances (A) obtained from model B. The MD results match well with the available crystallographic data: (1) Most bonds in the QM region are well reproduced and the rootmean-square (rms) deviation between experimental bond distances and average MD data is 0.079 A. (2) For the tyrosyl moiety of the chromophore, the average dihedral angles (i.e., N2_CA2_CB2_CG2 and CA2_CB2_CG2_CD1) obtained from the average 500-ps MD runs are 6.4 and 6.2 . Smaller dihedral angles demonstrate that the trajectories of the cis chromophore during MD runs are nearly co-planar, which is consistent with the experimental observation. (3) The rms deviations of five hydrogen bonds (bond distance is measured between the two heavy atoms) around the chromophore are quite small, and are in the range 0.035–0.514 A. We also performed SCC-DFTB/MM MD simulations with the chromophores of HcRed in different protonation states for the residues of Glu214 and Glu146, which are defined as model A and C, respectively. We then performed QM/MM (DFT/MM) optimizations with the structures, starting with arbitrarily selected snapshots from the MD calculations. Table 36.4 lists the QM energies (E(QM,MM)), MM energies (E(MM,QM)), total energies (Etotal) and relative energies (DE, relative to the energy of the cis isomer) of four snapshots of cis and trans chromophore of model B (i.e., Glu214 is protonated and Glu146 is deprotonated) at the DFT/CHARMM level. The results of the E(QM,MM) of the four snapshots show that cis-conformations are always more stable than the trans-isomers. The relative QM energies (DEQ) of trans Table 36.4 QM energies, MM energies, total energies and relative energies (relative to cis conformation) of four snapshots of cis and trans isomers of model B (Glu214 is protonated and Glu146 is deprotonated) at the DFT(B3LTP/ SV(P))/MM level Snapshot 1 2 3 4
cis trans cis trans Cis trans cis trans
E(QM,MM) (a.u.)
DEQ (kcal mol1)
E(MM,QM) (a.u.)
DEM (kcal mol1)
Etotal (a.u.)
DEt (kcal mol1)
1478.94658 1478.90565 1478.95365 1478.92237 1478.94117 1478.90826 1478.95541 1478.92339
0 25.7 0 19.6 0 20.7 0 20.1
54.24006 54.26645 54.21931 54.23319 53.84091 53.85320 53.67451 53.68634
0 16.6 0 8.7 0 7.7 0 7.4
1533.18664 1533.17210 1533.17296 1533.15555 1532.78208 1532.76147 1532.62992 1532.60973
0 9.1 0 10.9 0 12.9 0 12.7
834 Hydrogen Bonding and Transfer in the Excited State
chromophore [relative to E(QM,MM) of cis conformation] in HcRed are in the range 19.6–25.7 kcal mol1. In contrast, E(MM,QM) energies of cis conformations of the corresponding four snapshots are higher than those of trans isomers, and the relative MM energies (DEM) of the trans isomers (relative to E(MM,QM) of cis conformation) are from 16.6 to 7.4 kcal mol1. Hence, the total energies DEt of cis chromophore are lower than those of trans isomer by about 9.1–12.9 kcal mol1, which means the cis conformation is more stable than the trans counterpart. Moreover, in the case of model C where both Glu214 and Glu146 are deprotonated, the cis is much more stable than the trans, by about 12.4–19.9 kcal mol1. However, in model A, where both Glu214 and Glu146 are protonated, the stability of the cis and trans chromophores of HcRed is reversed. Hence, the different protonation states of Glu214 and Glu146 nearby chromophore control the stability of cis and trans isomers, which can further influence the fluorescent properties of HcRed. The study gains insight into the experimental phenomena that some fluorescent proteins such as mKate and Rtms5 show bright fluorescence at high pH, which might be due to the deprotonation of residues near the chromophores. Therefore, it provides a simple and useful manner to tune the photochemical properties of fluorescent proteins by modifying the protonation state of existing residues near chromophores, for example, by pH-induction.
36.4 Conclusions and Future Work The proton chain transfer event in GFP is crucial to its photophysical functions, through which the neutral chromophore undergoes fast proton transfer on the excited state to yield the green fluorescent anionic chromophore. After fluorescence, the reverse proton transfers on the ground state PES regenerate the neutral chromophore. We have extensively studied structural, energetic, dynamic and spectral properties of GFP chromophores on the basis of several cluster models. Our computational investigations have revealed some mechanistic aspects in green fluorescent protein (GFP) that indicate both concerted and sequential nature for the proton motions. Despite the overall “concerted” nature of the potential profile, the configurational evolution along the reaction coordinate involves sequential movement of the protons. Moving beyond the static reaction pathway picture, we performed nuclear quantum dynamics (QD) simulations using both TI and TD methods. In particular, the time-independent quantum dynamical approach based on a minimal quantum chemical cluster model designed to represent the ground state proton transfer between the neutral and anionic states of the GFP chromophore is very encouraging. The calculated KIE value for the I ! A transition of 15.5 at 300 K is in remarkable agreement with the experimentally measured value of 12, suggesting that this quantum dynamical approach has considerable promise for future investigation of proton chain transfer kinetics in the GFP as well as other biomolecular systems. A key conclusion, which comes out of comparison of the present exact 3D quantum calculations with more simplistic models that would treat the proton coordinates as limiting harmonic or translational modes, is that these limiting approximations fail utterly to approach the experimentally measured KIE – suggesting that correct incorporation of anharmonic features of the PES is crucial to reliably model this complex proton transfer process. Despite the remarkable early success of this new TI approach to modeling biological proton wires, there remain many challenges to a more complete and general development of this strategy, which we discuss below. Very recently we have extended our quantum mechanical studies to complete full 3D potential energy surfaces for the proton transfers on the excited state in a minimal quantum mechanical cluster model for GFP. We employ the MO MRCI methods to compute the ab initio potential energy surfaces for both the ground and the first excited state. 12 12 12 ab initio data points were used to construct the potential energy surfaces, and no symmetry restriction is assumed. Comparison of the energetic calculated from current MRCI calculations using rigid model with available experimental estimations for GFP in the protein environment indicates that while the barrier heights and relative energies on both the ground and excited states are in qualitative agreement, the calculated excitation wavelength and emission wavelength are smaller than the
Theoretical Studies of Green and Red Fluorescent Proteins
835
experimental ones for GFP. These discrepancies might be due to the rigid model employed and the neglect of the protein environment effects. Currently we are exploring the excited state proton transfer nuclear dynamics through the time-independent (TI) quantum dynamics method, in which the dynamic properties such as rate constants and KIEs will be computed using the high-level 3D PESs and will be compared with the available experimental data. For red fluorescent protein (RFP) chromophores, the bridge photoisomerization pathways of the model chromophores have been characterized using electronic structure methods. Our work, carried out in collaboration with experimental studies, has revealed that the embedded chromophore of red fluorescent protein takes both cis-coplanar and trans-non-coplanar conformations due to a degree of inherent mobility that is not displayed in the GFP. The cis-coplanar conformation is suggested to exhibit bright fluorescence whereas the trans conformation is non-fluorescent, and cis–trans isomerization might play a key role for the fluorescence mechanism in RFP. Our work using truncated chromophore models has laid the theoretical groundwork and established the key mechanistic questions that need to be targeted in subsequent wholeprotein computational studies, which will seek to understand the critical role of protein environment on the fluorescent properties. Owing to the truncation of the models and the neglect of environmental considerations, cluster models need to be extended to include the protein environmental effects, and more reliable QD methods need to be developed to describe the proton nuclear dynamics in a more realistic way in the future. In this direction, some preliminary QM/MM work is currently undertaken through collaboration between our laboratory and Thiel (PI)’s group in Germany. For example, combined quantum mechanics/molecular mechanics (QM/MM) calculations are being performed to gain insight into the function of the far-red fluorescent protein (FP) and to seek an understanding of which residues within the protein influence the fluorescent properties of the embedded chromophore. In addition, reliable mixed quantum dynamics/molecular dynamics (QD/MD) methods are being explored to study the proton transfers on both the excited and ground states in green fluorescent protein. Given the lightness of proton and the corresponding quantum effects involved, these proton transfer dynamics should be considered within the framework of quantum mechanics. Thus the full description of the reaction mechanisms in GFP should combine the quantum-dynamical model for the protons with MD simulations for all other atoms. This is one direction that needs further developments.
Acknowledgements We are grateful to the Australian Research Council and The University of Queensland for supporting this work. We also acknowledge generous grants of high performance computer time from both The University of Queensland (the Computational Molecular Science cluster computing facility) and the Australian National Computational Infrastructure (NCI) Facility.
References 1. 2. 3. 4. 5. 6. 7. 8.
R. Y. Tsien, Annu. Rev. Biochem., 67, 509 (1998). M. Zimmer, Chem. Rev., 102, 759 (2002). K, Brejc, T. K. Sixma, P. A. Kitts et al., Proc. Natl. Acad. Sci. USA, 94, 2306 (1997). M. Ormo, A. B. Cubitt, K. Kallio et al., Science, 273, 1392 (1996). R. A. G. Cinelli, V. Tozzini, V. Pellegrini et al., Phys. Rev. Lett., 86, 3439 (2001). K. Winkler, J. Lindner, V. Subramaniam et al., Phys. Chem. Chem. Phys., 4, 1072 (2002). J. T. M. Kennis, D. S. Larsen, I. H. M. Stokkum et al., Proc. Natl. Acad. Sci. USA, 101, 17988 (2004). A. Toniolo, S. Olsen, L. Manohar and T. J. Martinez, Faraday Discuss., 127, 149 (2004).
836 Hydrogen Bonding and Transfer in the Excited State 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. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55. 56. 57. 58. 59.
N. Y. A. Baffour-Awuah and M. Zimmer, Chem. Phys., 303, 7 (2004). M. A. L. Marques, X. L. Varsano, D. Castro and A. Rubio, Phys. Rev. Lett., 90, 258101 (2003). S. S. Patnaik, S. Trohalaki and R. Pachter, Biopolymers, 75, 441 (2004). S. Wang and S. C. Smith, J. Phys. Chem. B, 110, 5084 (2006). S. Wang and S. C. Smith, Phys. Chem. Chem. Phys., 9, 452 (2007). O. Vendrell, R. Gelabert, M. Moreno and J. M. Lluch, J. Am. Chem. Soc., 128, 3564 (2006). O. Vendrell, R. Gelabert, M. Moreno and J. M. Lluch, J. Phys. Chem. B, 112, 5500 (2008). R. Heim, D. C. Prasher and R. Y. Tsien, Proc. Natl. Acad. Sci. USA, 91, 12501 (1994). A. B. Cubitt, R. Heim, S. R. Adams et al., Trends Biochem. Sci., 20, 448 (1995). M. Chattoraj, B. A. King, G. U. Bublitz and S. G. Boxer, Proc. Natl. Acad. Sci. USA, 93, 8362 (1996). T. M. H. Creemers, A. J. Lock, V. Subramaniam et al., Nat. Struct. Biol., 6, 557 (1999). H. Lossau, A. Kummer, R. Heinecke et al., Chem. Phys., 213, 1 (1996). J. J. v Thor, T. Gensch, K. J. Hellingwerf and L. N. Johnson, Nat. Struct. Biol., 9, 37 (2002). C. Seebacher, F. W. Deeg, C. Br€auchle et al., J. Phys. Chem. B, 103, 7728 (1999). N. Agmon, Biophys. J., 88, 2452 (2005). P. Leiderman, D. Huppert and N. Agmon, Biophys. J., 90, 1009 (2006). N. Agmon, J. Phys. Chem. B, 111, 7870 (2007). J. J. v. Thor, G. Zanetti, K. L. Ronayne and M. Towrie, J. Phys. Chem. B, 109, 16099 (2005). D. Stoner-Ma, A. A. Jaye, P. Matousek et al., J. Am. Chem. Soc., 127, 2864 (2005). D. Stoner-Ma, E. H. Melief, J. Nappa et al., J. Phys. Chem. B, 110, 22009 (2006). H.-Y. Yoo, J. A. Boatz, V. Helms et al., J. Phys. Chem. B, 105, 2850 (2001). T. Laino, R. Nifosı` and V. Tozzini, Chem. Phys., 298, 17 (2004). W. Weber, V. Helms, J. A. McCammon and P. W. Langhoff, Proc. Natl. Acad. Sci. USA, 96, 6177 (1999). V. Helms, T. P. Straatsma and J. A. McCammon, J. Phys. Chem. B, 103, 3263 (1999). V. Tozzini and R. Nifosı`, J. Phys. Chem. B, 105, 5797 (2001). N. Reuter, H. Lin and W. Thiel, J. Phys. Chem. B, 106, 6310 (2002). M. A. Lill and V. Helms, Proc. Natl. Acad. Sci. USA, 99, 2778 (2002). S. S. Patnaik, S. Trohalaki, R. R. Naik et al., Biopolymers, 85, 253 (2007). R. Zhang, M. T. Nguyen and A. Ceulemans, Chem. Phys. Lett., 404, 250 (2005). H. Zhang and S. C. Smith, J. Theor. Comput. Chem., 6, 789 (2007). O. Vendrell, R. Gelabert, M. Moreno and J. M. J. Lluch, Chem. Theory Comput., 4, 1138 (2008). J. R. Wiedenmann, S. Ivanchenko, F. Oswald and G. U. Nienhaus, Marine Biotechnol., 6, 270 (2004). R. M. Hoffman, J. Biomed. Opt., 10, 41202 (2005). Y. J. Passamaneck, A. Di-Gregorio, V. E. Papaioannou and A. K. Hadjantonakis, Microsc. Res. Tech., 69, 160 (2006). S. A. Wacker, F. Oswald, J. Wiedenmann and W. Knochel, Dev. Dyn., 236, 473 (2007). K. A. Lukyanov, D. M. Chudakov, S. Lukyanov and V. V. Verkhusha, Nat. Rev. Mol. Cell. Biol., 6, 885 (2005). R. C. Mirabella, C. Franken, G. N. van-der-Krogt et al., Plant Physiol., 135, 1879 (2004). N. C. Shaner, R. E. Campbell, P. A. Steinbach et al., Nat. Biotechnol., 22 (2004). N. G. Gurskaya, F. Fradkov, A. Terskikh et al., FEBS Lett., 507, 16 (2001). L. Wang, W. C. Jackson, P.A. Steinbach and R. Y. Tsien, Proc. Natl. Acad. Sci. USA, 101, 16745 (2004). M. A. Shkrob, Y. G. Yanushevich, D. M. Chudakov et al., Biochem. J., 392, 649 (2005). D. Shcherbo, E. M. Merzlyak, T. V. Chepurnykh et al., Nat. Methods, 4, 741 (2007). S. Olsen, M. Prescott, P. Wilmann et al., Chem. Phys. Lett., 420, 507 (2006). S. Olsen and S. C. Smith, Chem. Phys. Lett., 426, 159 (2006). S. Olsen and S. C. Smith, J. Am. Chem. Soc., 129, 2054 (2007). S. Olsen and S. C. Smith, J. Am. Chem. Soc., 130, 8677 (2008). L. A. Gross, G. S. Baird, R. C. Hoffman et al., Proc. Natl. Acad. Sci. USA, 97, 11990 (2000). S. Olsen, J. Riesz, I. Mahadevan et al., J. Am. Chem. Soc., 129, 6672 (2007). P. G. Wilmann, J. Petersen, A. Pettikiriarachchi et al., J. Mol. Biol., 349, 223 (2005). P. G. Wilmann, J. Battad, T. Beddoe et al., Photochem. Photobiol., 82, 359 (2006). P. G. Wilmann, K. Turcic, J. M. Battad et al., J. Mol. Biol., 364, 213 (2006).
Theoretical Studies of Green and Red Fluorescent Proteins 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.
837
J. M. Battad, P. G. Wilmann, S. Olsen et al., J. Mol. Biol., 368, 998 (2007). A. D. Becke, J. Chem. Phys., 98, 5648 (1993). C. Lee, W. Yang and R. G. Parr, Phys. Rev. B, 37, 785 (1988). M. J. Frisch, J. A. Pople and J. S. Binkley, J. Chem. Phys., 80, 3265 (1984). M. J. Frisch, G. W. Trucks, H. B. Schlegel et al., Gaussian 03, Gaussian, Inc., Wallingford, CT (2004). D. E. Woon and T. H. Dunning, J. Chem Phys., 98, 1358 (1993). H.-J. Werner, P. J. Knowles, R. Lindh et al., MOLPRO version 2006. 1, a package of ab initio programs, see http:// www.molpro.net. U. Manthe and W. H. Miller, J. Chem. Phys., 99, 3411 (1993). K. K. Docken and J. Hinze, J. Chem. Phys., 57, 4928 (1972). B. O. Roos, P. R. Taylor and P. E. M. Siegbahn, Chem. Phys., 48, 157 (1980). P. Celani and H.-J. Werner, J. Chem. Phys., 112, 5546 (2000). A. T. Brunger and M. Karplus, Proteins: Struct. Funct. Genet., 4, 148 (1988). H. Li, A. D. Robertson and J. H. Jensen, Proteins: Struct., Funct. Bioinformat., 61, 704 (2005). W. L. Jorgensen, J. Chandrasekhar, J. D. Madura et al., J. Chem. Phys., 79, 926 (1983). M. Elstner, D. Porezag, G. Jungnickel et al., Phys. Rev. B, 58, 7260 (1998). A. D. MacKerell, D. Bashford, M. Bellott et al., J. Phys. Chem. B, 102, 3586 (1998). J. Z. Pu, J. L. Gao and D. G. Truhlar, J. Phys. Chem. A, 108, 5454 (2004). D. Bakowies and W. Thiel, J. Phys. Chem., 100, 10580 (1996). ChemShell, a Computational Chemistry Shell, see www.chemshell.org. P. Sherwood, A. H. de Vries, M. F. Guest et al., C. J. Mol. Struct.-Theochem., 632, 1 (2003). S. R. Billeter, A. J. Turner and W. Thiel, Phys. Chem. Chem. Phys., 2, 2177 (2000). S. Wang and S. C. Smith, Chem. Phys., 326, 204 (2006). H. Zhang, S. Wang, Q. Sun and S. C. Smith, Phys. Chem. Chem. Phys., 11, 8422 (2009). S. G. Dove, O. Hoegh-Guldberg and S. Ranganathan, Coral Reefs, 19, 197 (2001). A. Salih, A. Larkum, G. Cox et al., Nature, 408, 850 (2000). Q. Sun, M. Doerr, Z. Li, S. C. Smith and T. Thiel, QM/MM studies of structural and energetic properties of the far-red fluorescent protein HcRed, Phys. Chem. Chem. Phys., 12, 2450 (2010).
37 Changes in Active Site Hydrogen Bonding upon Formation of the Electronically Excited State of Photoactive Yellow Protein Wouter D. Hoff,1 Zhouyang Kang,2 Masato Kumauchi,1 and Aihua Xie2 1
Department of Microbiology and Molecular Genetics, Oklahoma State University, Stillwater, OK 74078, USA 2 Department of Physics, Oklahoma State University, Stillwater, OK 74078, USA
37.1 Central Importance of Light in Biology Light plays three main roles in biology. First, light from the sun provides the main source of energy driving the biosphere through chlorophyll-based electron transfer in photosynthetic reaction centers and retinal-based proton pumping in rhodopsins [1]. Second, light is a source of information about the environment for many organisms using photosensory proteins, including animals, plants, fungi and bacteria [2, 3]. Third, light can cause damage to biological systems. Light from the near-UV region can damage DNA [4], and excess visible light can damage the photosynthetic machinery in photoinhibition [5]. Thus, light is of great importance for a wide range of biological processes. Proteins that perform photosynthesis or light sensing consist of an apoprotein and a bound chromophore. Only a small number of chromophores account for most known processes in photobiology: chlorophyll and linear tetrapyrroles, caretoids and retinal, flavins and p-coumaric acid [2]. In many cases the chromophore has strong and functionally important interactions with the protein binding pocket, particularly charge–charge interactions and hydrogen bonding interactions. While many different proteins interact with light in a wide range of organisms, almost all of these systems are based on key types of photochemical processes, including electron transfer, C¼C double bond isomerization and the formation of chemical bonds [2, 6]. These photochemical events often trigger a cascade of thermal reactions in the protein that extends to the millisecond and minute range, and that result in a biologically relevant output. If the cascade of thermal reactions results in the re-formation of the initial state, the process is referred to as a photocycle. It is the initial ultrafast
Hydrogen Bonding and Transfer in the Excited State, Volume I & II © 2011 John Wiley & Sons, Ltd. ISBN: 978-0-470-66677-7
Edited by Ke-Li Han and Guang-Jiu Zhao
840 Hydrogen Bonding and Transfer in the Excited State
photochemical event that drives the subsequent thermal events in a photocycle through strong chromophore– protein interactions. In this chapter we examine current understanding of ultrafast changes in hydrogen bonding upon photoexcitation of photoactive yellow protein (PYP), a bacterial photosensor.
37.2 Possible Importance of Excited State Hydrogen Bonding in Photoreceptors In many photosensory proteins the primary photochemical event involves isomerization of double and single bonds. These isomerization events can greatly alter hydrogen bonding interactions between the chromophore and protein binding pocket, including the disruption of hydrogen bonds and the formation of new hydrogen bonds. These changes in hydrogen bonding can have important consequences for subsequent functionally important conformational changes in the protein. Evidence for functionally important changes in hydrogen bonding during the initial ultrafast transitions of the photocycle have been reported not only for PYP (see below), but also for other photoreceptors, such as the BLUF domain [7]. In addition, it is possible that hydrogen bonding in the electronically excited state is important in guiding the initial steps of the photocycle. Such functional effects of excited state hydrogen bonding in photosensory proteins have only just been started to be explored, and raise the question as to how these hydrogen bonds are altered in the electronically excited state and how this affects [8] the earliest steps in the function of the protein. Here we review knowledge regarding hydrogen bonding in the electronically excited chromophore in the active site of PYP.
37.3 Introduction to Photoactive Yellow Protein Photoactive yellow protein [9] is a bacterial blue-light receptor. It is a highly accessible model system that exhibits proton transfer between active site groups with shifted pKa values during its light-triggered functional cycle [9, 10]. The PYP from Halorhodospira halophila (Hh PYP) is a highly studied [11–13] member of a growing family of bacterial blue-light receptors [14]. A rich body of information is available on functionally important active site hydrogen bonds and proton transfer events in the PYP from H. halophila, providing an attractive and tractable system to examine these ubiquitous and central aspects of protein function. PYP was first discovered in the extremely halophilic anoxygenic photosynthetic proteobacterium H. halophila [9, 15], where it functions as a photoreceptor for negative phototaxis [16]. It exhibits a light-triggered photocycle [9, 17–19] based on its p-coumaric acid (pCA) chromophore [20, 21]. The pCA is covalently linked to Cys69 in the protein via a thioester bond (Figure 37.1) [22, 23]. Three key structural features of the pCA are that its phenolic oxygen can undergo protonation/deprotonation, its central C7 ¼ C8 double bond can undergo photoisomerization, and both its phenolic and carbonyl oxygen can form hydrogen bonds. In the initial state of PYP the pCA is deprotonated and in the trans configuration. The crystal structure of PYP consists of a central antiparallel six-stranded b-sheet flanked by five a-helices (Figure 37.1a) [24]. PYP shares its fold with a large superfamily of proteins called the PAS domain [25–27]. Absorption of a blue photon by the pCA chromophore embedded in the protein initiates the light sensing process of PYP by chromophore trans to cis photoisomerization [10, 28–31]. This ultrafast isomerization reaction converts the initial pG state into the redshifted pR state via a complex series of ultrafast processes. The initial steps of the PYP photocycle (Figure 37.2) have been studied extensively by sub-ps time-resolved spectroscopy [32], and will be discussed below. The pR state thermally decays to the blue-shifted pB0 intermediate in a sub-millisecond process that involves protonation of the pCA chromophore [21] by proton transfer from Glu46 to the pCA [10, 33, 34]. Large conformational changes [33, 35–43] convert the pB0 state into the pB state on the millisecond time scale. The destabilizing buried negative charge on Glu46 has been identified as a key factor in driving these conformational changes [33, 44]. The large structural changes that
Changes in Active Site Hydrogen Bonding upon Formation of the Electronically Excited
841
Figure 37.1 Crystal structure of PYP. (a) Overall fold of PYP showing the pCA, Glu46, Tyr42 and Cys69 (BDP ID 1NZW). (b) Active site of PYP with three key hydrogen bonds. (c) The thioester linked pCA chromophore in PYP with numbering of its atoms
occur upon pB formation can be described as a protein quake driven by the electrostatic epicenter formed by the negative charge of Glu46 [33]. The pB state is long-lived and is thought to be the active signaling state of PYP that interacts with an as yet unidentified downstream signaling partner. It thermally recovers to the initial pG state in a few hundred milliseconds.
37.4 Hydrogen Bonding in the Initial State of PYP Extensive crystallographic studies have shown that the phenolic oxygen of the negatively charged pCA chromophore of PYP interacts with active site residues Glu46 and Tyr42 via a forked ionic hydrogen bond. The
842 Hydrogen Bonding and Transfer in the Excited State
Figure 37.2 Schematic representation of the PYP photocycle. The pCA isomerization state in the various intermediates is indicated as superscripts and their absorbance maxima as subscripts. The wavy line indicates a photochemical reaction, straight lines indicate thermal transitions. The structure of the pCA and its hydrogen bonds with Tyr42, Glu46 and Cys69 in the intermediates are indicated schematically
side chains of Glu46 and Tyr42 form unusually short hydrogen bonds with the pCA (Figure 37.1b): 2.59 and 2.50 A, respectively [45]. For comparison, the average length of hydrogen bonds in proteins involving O and N atoms is 3.0 A [46]. Recently the pCA–Glu46 interaction was identified as a low-barrier hydrogen bond based on neutron crystallography [47]. However, an independent neutron diffraction study did not reveal this, and discussed the partial occupancy of deuterions in the hydrogen bonding network at the phenolic oxygen of the pCA [48]. Direct spectroscopic evidence for hydrogen bonding between Glu46 and the pCA has been obtained from FTIR studies (see below). The C¼O stretching mode of the side chain of Glu46 in Hh PYP is at 1736 cm1 in H2O [10]. This frequency corresponds to a single strong hydrogen bond to the Glu46 side chain [49]. In D2O this mode is downshifted by 10 cm1 to 1726 cm1 [10], a typical shift for H/D exchange of a protonated carboxylic acid. Several studies have shown that the hydrogen bond between Glu46 and the pCA is of great functional importance. First, this hydrogen bond is an important contributor to the strong shifts in the pKa values of both the pCA chromophore and Glu46 in PYP compared to their values in solution. The pKa of the thioester-linked pCA is downshifted from 8.8 in solution [50] to 2.8 in PYP [15, 51], while the pKa of Glu46 is strongly upshifted from 4.25 in solution [52] to >10 in PYP. Owing to this “pKa inversion” the ionized [28] pCA (normally a base) is hydrogen bonded to the protonated [10, 33] side chain of Glu46 (normally an acid). Mutations in both Glu46 and Tyr42 significantly shift the pKa of the pCA in the direction of its value in solution [53–55]. This indicates the importance of the hydrogen bonds of Glu46 and Tyr42 to the pCA in downshifting its pKa. The pKa inversion of these Glu46 and the pCA in PYP is of central functional importance since proton transfer between these two groups is a key event during the light-triggered photocycle in PYP [10, 33]. A second functionally important aspect of active site hydrogen bonding in PYP is that the strength of the hydrogen bond between Glu46 and the pCA is a main factor in regulating the absorbance spectrum of PYP. Deprotonated thioester-linked pCA absorbs near 400 nm in solution [50], which is strongly blue-shifted from the absorbance maximum of PYP at 446 nm. Interactions between the pCA and its protein binding pocket cause a 46 nm redshift in its absorbance maximum. In the E46Q mutant, in which the hydrogen bond between the pCA and residue 46 is weakened and lengthened by 0.3 A [45, 56] the absorbance maximum is redshifted by 16 nm [57, 58], presumably by the reduced localization of p-electrons by the hydrogen bonding interaction.
Changes in Active Site Hydrogen Bonding upon Formation of the Electronically Excited
843
These results illustrate the functionally central role of the hydrogen bond between Glu46 and the pCA chromophore in PYP.
37.5 Assignment of Vibrational Modes in PYP The molecular events that occur during the PYP photocycle have been studied with both resonance Raman and infrared difference spectroscopy. Vibrational signals originating from the pCA chromophore can be identified by resonance Raman spectroscopy; infrared spectra contain contributions from both the protein and the chromophore. In principle such vibrational spectra contain rich information on changes in structure, protonation state and hydrogen bonding interactions during the PYP photocycle. A key first step in extracting such information from the spectra is the assignment of specific vibrational modes to molecular groups in PYP. Three strategies have been used for the assignment of the vibrational modes of the pCA chromophore in PYP. First, the vibrational spectra of model compounds of the pCA chromophore, such as the p-coumaryl phenyl thioester [28], have been studied and compared to the vibrational spectra of PYP. Studies of these model compounds in different states, particularly with a neutral or ionized phenolic oxygen or in the trans and cis states, have yielded valuable information on which vibrational modes are sensitive to which molecular events in the PYP photocycle [28, 59]. Second, computational methods for calculating the vibrational spectrum of molecules are now sufficiently accurate to be a highly valuable tool for assigning vibrational modes, and this strategy has been applied to PYP [60–62]. In such calculations it is critical to have sufficient experimental data to determine if the calculations are reliable. No vibrational calculations on pCA in the electronically excited state have been reported yet. Third, various derivatives of pCA labeled with stable isotopes have been incorporated into PYP [59, 60, 62–64]. This can be a highly valuable tool for the assignment of vibrational signals in PYP to specific vibrational modes in the pCA chromophore. In this approach it is critical to design isotopically labeled pCA derivatives that affect only a small number of vibrational modes. This is a significant issue, since some isotopic substitutions affect many vibrational modes, defeating the original purpose of the isotopic labeling. The C¼O stretching mode of functionally important active site residue Glu46 in light-induced FTIR difference spectra was assigned based on (i) the unusual protonation state of this residue, which is the sole carboxylic acid in PYP that is not exposed to solvent, and (ii) the sensitivity of this signal to pCA photoisomerization at 80 K, indicating its immediate vicinity to the pCA [10]. Subsequent FTIR studies on the E46Q mutant of PYP confirmed this assignment [34]. This still remains the only amino acid side chain for which a firm assignment of signals in FTIR difference spectra of PYP is available. Extensive assignments have been performed for the vibrational modes of the pCA as observed by resonance Raman spectroscopy. These assignments are particularly well established for the pG state [62] and the longlived pB state [63, 64]. Some key pCA vibrational bands in the pR state have also been studied [60]. These assignments were based on strategically chosen isotopic labeling of the pCA in combination with high-level ab initio vibrational calculations. Signals that were found to be structurally most informative (see below) are the C¼O stretching mode at 1631 cm1, the C8–C9 stretching mode at 1052 cm1 and the in-plane pCA ring CH bending mode at 1164 cm1 (stated values are for the pG state in H2O). In FTIR difference spectra a strong and structurally informative band at 1498 cm1 has been assigned to a pCA ring C¼C stretching vibration [33].
37.6 Identification of Vibrational Structural Markers The interpretation of vibrational signals in terms of molecular structure requires the identification of modes that serve as reliable indicators of specific molecular processes. A reliable vibrational structural marker is a vibrational mode whose frequency depends only (or largely) on a single structural feature, for example,
844 Hydrogen Bonding and Transfer in the Excited State Table 37.1
Vibrational structural markers for structural changes during the PYP photocycle. All values are in cm1
Structural change pCA trans to cis isomerization pCA protonation from O to OH pCA O C¼O loss of hydrogen bonding pCA OH C¼O loss of hydrogen bonding Glu46 loss of hydrogen bonding Glu46 deprotonation
Mode
Before event
After event
C8–C9 stretching ring CH bending ring C¼C stretching C¼O stretching C¼O stretching COOH C¼O stretching COO(H) C¼O stretching
1050 1165 1498 1640 1665 1736 (1726)a 1736 (1726)a
1000 1174 1515 1673 1692 1765 (1755)a 60(5–19%) 38–69(10–25%) — — 40 —
426 0.22 0.15 0.3 0.11 0.19 0.18
1300 440 420 410 410 400 400
50 11 25 65 25 75
3.6 — — 7.1 6.3 3.6
— 220 — 1300c 220 18
0.2 3–6 0.1 0.17 0.1 1.2
400/464 452 395 390 400/485 428
15 22 15 12 70 75
2.8
40
0.12
395
270
9 — 6.2
950c — 700c
0.18 0.23 0.2
475 445 475
30 60 60
8
—
2.8
446/460
11
a
Signal to noise ratios were estimated from figures in cited references. Values depend on detection wavelength, with faster rates observed at the blue and red flanks of the emission band. These values indicate the kinetics of pR formation from I0. d The somewhat slower rate of this process compared to the 0.7 ps component observed in other studies may be due to the time resolution of this experiment. A fourth time constant of 463 ps was also observed in this study b c
X-ray crystallography of the cryotrapped photoproduct state [30, 31]. However, the small wavelength dependence for the kinetics of dump pulse-induced fluorescence intensity depletion observed by ultrafast fluorescence pump–dump spectroscopic measurements indicates that the electronically excited state is homogeneous with respect to emission transition energy [76]. Since only part of the electronically excited state is converted into the primary photoproduct, a parallel pathway is the decay of the excited state back to the initial pG dark state. The quantum yield of the forward reaction was initially estimated as 0.64 [35], but a later study adjusted this value down to 0.35 [88]. More recent ultrafast pump–dump spectroscopy revealed that the pathway from the excited state to the pG state proceeds via a ground state intermediate (GSI), which decays with a time constant of 3.6 ps [82]. Another complicating factor is that the initial photoproduct has been reported to partly decay back to the initial pG state in a thermal branching reaction [70, 82] (Figure 37.2). The formation of the primary photoproduct containing cis-pCA has been reported to occur on a time scale of 2 ps [68], mainly based on the photoproduct signal at 1289 cm1 [59]. A more recent study used the wellestablished cis-pCA isomerization marker [60] at 1000 cm1 to derive a time constant for the formation of the cis photoproduct as 3 1 ps [69]. The decay of the Glu46 signal at 1747 cm1 follows the same kinetics. The finding that the time constant for the formation of the cis-pCA electronic ground state primary
848 Hydrogen Bonding and Transfer in the Excited State
photoproduct appears to follow the slower 4 ps phase in excited state decay matches an interpretation in which the 0.5 ps time constant for the excited state is caused by its vibrational cooling, not by its decay to the primary photoproduct. In this interpretation the 0.5 ps process is barrierless transition on the excited state energy surface, while the 4 ps process is the conversion of the electronically excited state into the primary photoproduct that proceeds over a barrier on the excited state energy surface of 1.9 kcal mol1 [77]. The occurrence of ultrafast double bond isomerization of the pCA was also concluded based on changes in visible anisotropy, indicating a 24 degree change in transition dipole moment upon the formation of the primary photoproduct [83].
37.9 Changes in Active Site Hydrogen Bonding upon the Formation of the S1 State of PYP
963
1069 1041
1163
1498
1631 1755 1747
0 1438
1740 1725
-2
-4
1800
1557
Amplitude (mOD)
2
1288
Since the fastest time constant in the reported sub-ps infrared spectroscopic studies is a 2 ps time constant associated with the decay of the excited state, positive signals in the infrared difference spectra taken at sub-picosecond time delays are dominated by contributions from the electronically excited state. Figure 37.4 compares the calculated species associated difference spectrum for the formation of the electronically excited state measured in H2O [68] with the difference spectrum at 0.4 ps time delay taken under polarization anisotropy free conditions measured in D2O [69]. For comparison, the steady state resonance Raman spectrum of the pG dark state in H2O is shown to aid the identification of pG state pCA
1600
1400
1200
1000
Wavelength (cm-1)
Figure 37.4 Ultrafast mid-infrared difference spectroscopy for formation of the electronically excited state of PYP. The species-associated difference spectrum for formation of the electronically excited state of PYP in H2O (blue; from Ref. [68]) is shown together with the transient absorbance difference spectrum at 0.4 ps time delay for PYP in D2O obtained under polarization anisotropy free conditions (black, from Ref. [69]). For comparison, the steady state resonance Raman spectrum of the pG dark state of PYP is also shown (green; from Ref. [62]). Solid vertical lines indicate vibrational modes in the pG state that correspond well to negative signals in the transient infrared absorbance difference spectra; dotted vertical lines indicate modes where the match between the Raman and infrared data is less good. For the infrared difference spectra the vertical scale indicates the amplitude of the signals in milliOD
Changes in Active Site Hydrogen Bonding upon Formation of the Electronically Excited
849
signals in the infrared spectra. Below, these two FTIR difference spectra will be interpreted using the assumption that they represent the difference spectrum for the formation of the excited state from the pG ground state. A highly informative pair of positive and negative signals is observed at 1740/1755 cm1 in H2O and 1725/ 1747 cm1 on D2O (Figure 37.4). Based on steady state difference spectra with a high signal to noise ratio [10] the negative peaks corresponding to the pG state would have been expected at 1736 and 1726 cm1 in H2O and D2O, respectively. These data reveal that the side chain of Glu46 in the electronically excited state of PYP is located at 1755 and 1747 cm1 in H2O and D2O, respectively. Thus, the data show that the C¼O stretching mode of the side chain of Glu46 in the excited state of PYP exhibits a typical downshift upon H/D exchange. In general a very good correlation exists between the frequency of the C¼O stretching frequency of protonated acidic side chains and their hydrogen bonding strength [49]. Since the side chain of Glu46 remains in the electronic ground state, this published vibrational spectral marker can be applied to the signals in the electronically excited state of PYP. The value of 1736 cm1 in the pG state indicates the presence of a fairly strong single hydrogen bond, while the value of 1755 cm1 indicates either the absence of hydrogen bonding or a very weak single hydrogen bond. The above correlation between C¼O stretching frequency and hydrogen bonding applies to steady state conformations of proteins. In the case of functional intermediates the correlation is less strong, presumably because of structural perturbations [49]. It has been proposed that a large change in charge distribution in the electronically excited state of the pCA is a major factor in causing the upshift in Glu46 C¼O stretching frequency in the excited state, particularly because a frequency downshift in this mode occurs upon the formation of the primary photoproduct [68, 69]. A large change in permanent dipole moment upon formation of the electronically excited state of PYP has been detected by Stark spectroscopy [89]. Its value of 26 debye corresponds to the movement of one electron over 5 A, and has been attributed to a displacement of the negative charge on the phenolic oxygen of the pCA towards the thioester bond [89]. However, the extent of the change in Glu46 C¼O stretching mode expected to be caused by this change in charge distribution has not been quantitatively discussed. An additional argument in favor of the continued presence of the Glu46-pCA hydrogen bond in the excited state was obtained from ultrafast polarized infrared measurements, indicating that the orientation of the C¼O group of Glu46 remains the same in the excited state [69]. These data confirm that the phenolic ring of the chromophore does not undergo significant structural changes, but that it is the pCA C¼O group that rotates upon photoexcitation. A second structurally informative vibrational signal is the pCA C¼O stretching mode, which undergoes a 35 cm1 upshift upon disruption of the pCA-Cys69 hydrogen bond [60, 61]. In the primary photoproduct this mode is located at 1663 cm1 in H2O [68] and at 1669 cm1 in D2O [69]. This is clearly upshifted from its position at 1631 cm1 in the pG state, providing direct evidence for the disruption of the pCA–Cys69 hydrogen bond in the primary photoproduct. As explained below, the interpretation of the pCA C¼O stretching in the electronically excited state is not as straightforward. The absence of a strong positive band near 1665 cm1 in the electronically excited state of PYP has been proposed as evidence that the hydrogen bond between the pCA C¼O and Cys69 remains intact in the excited state [70]. The absence of this same band in the ground state intermediate indicates that in PYP molecules that are unsuccessful in entering the PYP photocycle the hydrogen bond between the pCA C¼O and Cys69 remains intact [70]. Thus, the disruption of this hydrogen bond has been proposed to be a key step for the successful entry into the PYP photocycle. A signal at 1640 cm1 has been assigned to the C¼O stretching mode of the electronically excited state, and has been presented as evidence for the presence of the pCA C¼O–Cys69 NH hydrogen bond [70]. The rotation of the pCA C¼O group is a key reaction coordinate during the initial stage of the photocycle, and causes the disruption of this hydrogen bond in the primary photoproduct. Thus, the status of this hydrogen bond in the electronically excited state is of considerable interest. The above assignment depends on several assumptions. First, the signal at 1640 cm1 is present in the difference spectrum between the electronically excited state and
850 Hydrogen Bonding and Transfer in the Excited State
the ground state that was extracted from the time-resolved data by mathematical modeling. Since the calculation of this species-associated difference spectrum is model-dependent and the correct kinetic model for the initial steps in the PYP photocycle has not yet been firmly established, this introduces an additional uncertainty. In addition, since an electron is excited from a bonding p orbital to an anti-bonding p orbital of the pCA upon formation of the electronically excited state, the location of the C¼O stretching mode may be significantly altered. Its definitive identification could be performed experimentally using appropriate isotopic labeling of the pCA. A third issue regards the structural interpretation of the frequency of the C¼O stretching mode in the electronic excited state. The reported interpretation of this mode in terms of hydrogen bonding [60] applies to the electronic ground state. High-level calculations are required to establish the frequency of this mode in the electronically excited state with and without hydrogen bonding. This can yield definitive information regarding the status of the hydrogen bond between the pCA C¼O and the Cys69 NH group in the electronically excited state. Two additional observations can be made based on Figure 37.4. First, the resonance Raman signals for the pG state at 1557, 1438, 1163, 1068, 1041 and 963 cm1 closely match negative signals in the sub-picosecond time-resolved infrared data. This is strong confirmation of the high experimental accuracy of the time-resolved data. However, three negative signals in the infrared spectra do not match well with the pG resonance Raman spectrum. These are located at 1631, 1498 and 1288 cm1. These discrepancies may be due to overlap with signals originating from vibrations in the protein. The resonance Raman signal at 1631 cm1 of the C¼O stretching mode of the pCA [60] is a key part of deriving changes in pCA hydrogen bonding during the pCA photocycle. The slight mismatch between the resonance Raman signal and the negative signal in the infrared difference spectra increases the need for assignment of this important mode by isotopic labeling of the pCA. Similarly, it is not clear from the difference spectra in Figure 37.4 if the C¼O stretching signal is slightly upshifted or downshifted in the electronically excited state. A second observation is that some infrared difference signals show a clear sensitivity to H/D exchange. Even though the pCA chromophore in the pG state does not contain exchangeable protons, the vibrational spectrum of the pCA in PYP is altered upon H/D exchange. The most striking change is that in D2O the doublet of signals at 1059 and 1043 cm1 merges into a single band at 1059 cm1 [28, 62]. Presumably, this effect is caused by vibrational coupling between the pCA chromophore and H/D exchanging groups in the pCA binding pocket, particularly the side chains of Tyr42 and Glu46, and the amide backbone of Cys69. Inspection of Figure 37.4 reveals that a strong negative signal at 1490 cm1 in H2O is shifted to 1510 cm1 in D2O. The direction of this 20 cm1 shift is opposite from what one would expect for H/D exchange, and has not been interpreted. Vibrational signals for probing the third hydrogen bond of the pCA chromophore (to Tyr42) have not yet been identified in infrared difference spectra. However, a recent picosecond time-resolved ultraviolet resonance Raman spectroscopic study reported Tyr signals that were tentatively assigned to Tyr42 [90]. The loss of intensity of a Tyr ring vibration near 1605 cm1 was interpreted to indicate the ultrafast strengthening of the hydrogen bond between Tyr42 and the pCA in the electronically excited state of PYP, and the subsequent slight weakening of this hydrogen bond with an 8 ps time constant [90]. Identification of this signal in ultrafast infrared difference spectra of PYP (Figure 37.4) will likely require isotope editing.
37.10 Excited State Proton Transfer in the Y42F Mutant of PYP Very recently it was reported that the Y42F mutant of PYP exhibits excited state proton transfer [94]. In this mutant the hydrogen bond between active site residue 42 and the pCA (Figure 37.1) is disrupted, while that between Glu46 and the pCA is strengthened. This results in a novel spectroscopic species that is in thermal equilibrium with the pG state. This species absorbs near 390 nm, and was concluded to contain a protonated
Changes in Active Site Hydrogen Bonding upon Formation of the Electronically Excited
851
pCA. Since a very similar photocycle is observed upon excitation of the pG state with an ionized pCA and the 390 nm species with a protonated pCA, it was concluded that photoexcitation of the 390 nm species most likely results in excited state proton transfer [94]. This is in contrast to the situation in wild-type PYP, where photoexcitation triggers pCA photoisomerization, followed by much slower proton transfer on the electronic ground state energy surface. In summary, a range of sub-ps spectroscopic studies have been performed on the electronically excited state of PYP and the initial stages of the PYP photocycle. These studies have fully confirmed a model of PYP isomerization that involves the rotation of the pCA C¼O group, while the phenolic group of the pCA does not significantly move [10]. The data indicate that in the excited state all three hydrogen bonds of the pCA remain intact, while disruption of the pCA C¼O–Cys69 hydrogen bond is a key step for successful photocycle entry. Further studies will be needed to confirm and extend this model and to determine the origin of the multiphasic decay of the electronically excited state. In addition, ultrafast spectroscopic studies on PYP mutants promise to uncover rich information on how the protein binding pocket tunes events in the electronically excited state of the pCA chromophore.
Acknowledgements W. D. H. gratefully acknowledges support from NIH grant GM063805 and OCAST grant HR07-135S, and from startup funds provided by Oklahoma State University. A. X. was supported by funds from OCAST (HR02-137R) and from the Oklahoma State Regents for Higher Education. The authors thank Marie Louise Groot and Masashi Unno for providing ASCI data of their vibrational measurements on PYP, and Delmar Larsen for helpful discussions.
References 1. J. K. Lanyi, Bacteriorhodopsin, Annu. Rev. Physiol., 66, 665–688 (2004). 2. M. A. van der Horst and K. J. Hellingwerf, Photoreceptor proteins, “star actors of modern times”: a review of the functional dynamics in the structure of representative members of six different photoreceptor families, Acc. Chem. Res., 37, 13–20 (2004). 3. M. A. van der Horst, J. Key and K. J. Hellingwerf, Photosensing in chemotrophic, non-phototrophic bacteria: let there be light sensing too, Trends Microbiol., 15, 554–562 (2007). 4. A. Sancar, Structure and function of DNA photolyase and cryptochrome blue-light photoreceptors, Chem. Rev., 103, 2203–2237 (2003). 5. A. Telfer, Too much light? How beta-carotene protects the photosystem II reaction centre, Photochem. Photobiol. Sci., 4, 950–956 (2005). 6. J. T. M. Kennis and M. L. Groot, Ultrafast spectroscopy of biological photoreceptors, Curr. Opin. Struct. Biol., 17, 623–630 (2007). 7. M. Gauden, I. H. M. van Stokkum, J. M. Key et al., Hydrogen-bond switching through a radical pair mechanism in a flavin-binding photoreceptor, Proc. Natl. Acad. Sci. USA, 103, 10895–10900 (2006). 8. G. J. Zhao and K. L. Han, Effects of hydrogen bonding on tuning photochemistry: concerted hydrogen-bond strengthening and weakening, ChemPhysChem, 9, 1842–1846 (2008). 9. T. E. Meyer, E. Yakali, M. A. Cusanovich and G. Tollin, Properties of a water-soluble, yellow protein isolated from a halophilic phototrophic bacterium that has photochemical activity analogous to sensory rhodopsin, Biochemistry, 26, 418–423 (1987). 10. A. Xie, W. D. Hoff, A. R. Kroon and K. J. Hellingwerf, Glu46 donates a proton to the 4-hydroxycinnamate anion chromophore during the photocycle of photoactive yellow protein, Biochemistry, 35, 14671–14678 (1996).
852 Hydrogen Bonding and Transfer in the Excited State 11. M. A. Cusanovich and T. E. Meyer, Photoactive yellow protein: a prototypic PAS domain sensory protein and development of a common signaling mechanism, Biochemistry, 42, 4759–4770 (2003). 12. K. Hellingwerf, J. Hendriks and T. Gensch, Photoactive Yellow Protein, a new type of photoreceptor protein: Will this “yellow lab” bring us where we want to go? J. Phys. Chem. A, 107, 1082–1094 (2003). 13. Y. Imamoto and M. Kataoka, Structure and photoreaction of photoactive yellow protein, a structural prototype of the PAS domain superfamily, Photochem. Photobiol., 83, 40–49 (2007). 14. M. Kumauchi, M. Hara, P. Stalcup et al., Identification of six new photoactive yellow proteins: diversity and structurefunction relationships in a bacterial blue light photoreceptor, Photochem. Photobiol., 84, 956–969 (2008). 15. T. E. Meyer, Isolation and characterization of soluble cytochromes, ferredoxins and other chromophoric proteins from the halophilic phototrophic bacterium Ectothiorhodospira halophila, Biochim. Biophys. Acta, 806, 175–183 (1985). 16. W. W. Sprenger, W. D. Hoff, J. P. Armitage and K. J. Hellingwerf, The eubacterium Ectothiorhodospira halophila is negatively phototactic, with a wavelength dependence that fits the absorption spectrum of the photoactive yellow protein, J. Bacteriol., 175, 3096–3104 (1993). 17. W. D. Hoff, I. H. van Stokkum, H. J. van Ramesdonk et al., Measurement and global analysis of the absorbance changes in the photocycle of the photoactive yellow protein from Ectothiorhodospira halophila, Biophys. J., 67, 1691–1705 (1994). 18. J. Hendriks, I. H. van Stokkum and K. J. Hellingwerf, Deuterium isotope effects in the photocycle transitions of the photoactive yellow protein, Biophys. J., 84, 1180–1191 (2003). 19. B. Borucki, C. P. Joshi, H. Otto et al., The transient accumulation of the signaling state of photoactive yellow protein is controlled by the external pH, Biophys. J., 91, 2991–3001 (2006). 20. W. D. Hoff, P. Dux, K. Hard et al., Thiol ester-linked p-coumaric acid as a new photoactive prosthetic group in a protein with rhodopsin-like photochemistry, Biochemistry, 33, 13959–13962 (1994). 21. M. Baca, G. E. Borgstahl, M. Boissinot et al., Complete chemical structure of photoactive yellow protein: novel thioester-linked 4-hydroxycinnamyl chromophore and photocycle chemistry, Biochemistry, 33, 14369–14377 (1994). 22. J. J. van Beeumen, B. V. Devreese, S. M. van Bun et al., Primary structure of a photoactive yellow protein from the phototrophic bacterium Ectothiorhodospira halophila, with evidence for the mass and the binding site of the chromophore, Protein Sci., 2, 1114–1125 (1993). 23. W. D. Hoff, B. Devreese, R. Fokkens et al., Chemical reactivity and spectroscopy of the thiol ester-linked p-coumaric acid chromophore in the photoactive yellow protein from Ectothiorhodospira halophila, Biochemistry, 35, 1274–1281 (1996). 24. G. E. Borgstahl, D. R. Williams and E. D. Getzoff, 1.4 A structure of photoactive yellow protein, a cytosolic photoreceptor: unusual fold, active site, and chromophore, Biochemistry, 34, 6278–6287 (1995). 25. J. L. Pellequer, K. A. Wager-Smith, S. A. Kay and E. D. Getzoff, Photoactive yellow protein: a structural prototype for the three-dimensional fold of the PAS domain superfamily, Proc. Natl. Acad. Sci. USA, 95, 5884–5890 (1998). 26. B. L. Taylor and I. B. Zhulin, PAS domains: internal sensors of oxygen, redox potential, and light, Microbiol. Mol. Biol. Rev., 63, 479–506 (1999). 27. M. H. Hefti, K. J. Francoijs, S. C. de Vries et al., The PAS fold – a redefinition of the PAS domain based upon structural prediction, Eur. J. Biochem., 271, 1198–1208 (2004). 28. M. Kim, R. A. Mathies, W. D. Hoff and K. J. Hellingwerf, Resonance Raman evidence that the thioester-linked 4hydroxycinnamyl chromophore of photoactive yellow protein is deprotonated, Biochemistry, 34, 12669–12672 (1995). 29. R. Kort, H. Vonk, X. Xu et al., Evidence for trans-cis isomerization of the p-coumaric acid chromophore as the photochemical basis of the photocycle of photoactive yellow protein, FEBS Lett., 382, 73–78 (1996). 30. R. Kort, K. J. Hellingwerf and R. B. Ravelli, Initial events in the photocycle of photoactive yellow protein, J. Biol. Chem., 279, 26417–26424 (2004). 31. S. Anderson, V. Srajer and K. Moffat, Structural heterogeneity of cryotrapped intermediates in the bacterial blue light photoreceptor, photoactive yellow protein, Photochem. Photobiol., 80, 7–14 (2004). 32. D. S. Larsen and R. van Grondelle, Initial photoinduced dynamics of the photoactive yellow protein, ChemPhysChem, 6, 828–837 (2005). 33. A. Xie, L. Kelemen, J. Hendriks et al., Formation of a new buried charge drives a large-amplitude protein quake in photoreceptor activation, Biochemistry, 40, 1510–1517 (2001).
Changes in Active Site Hydrogen Bonding upon Formation of the Electronically Excited
853
34. Y. Imamoto, K. Mihara, O. Hisatomi et al., Evidence for proton transfer from Glu-46 to the chromophore during the photocycle of photoactive yellow protein, J. Biol. Chem., 272, 12905–12908 (1997). 35. T. E. Meyer, G. Tollin, J. H. Hazzard and M. A. Cusanovich, Photoactive yellow protein from the purple phototrophic bacterium, Ectothiorhodospira halophila. Quantum yield of photobleaching and effects of temperature, alcohols, glycerol, and sucrose on kinetics of photobleaching and recovery, Biophys. J., 56, 559–564 (1989). 36. W. D. Hoff, A. Xie, I. H. van Stokkum et al., Global conformational changes upon receptor stimulation in photoactive yellow protein, Biochemistry, 38, 1009–1017 (1999). 37. B. C. Lee, A. Pandit, P. A. Croonquist and W. D. Hoff, Folding and signaling share the same pathway in a photoreceptor, Proc. Natl. Acad. Sci. USA, 98, 9062–9067 (2001). 38. B. C. Lee, P. A. Croonquist, T. R. Sosnick and W. D. Hoff, PAS domain receptor photoactive yellow protein is converted to a molten globule state upon activation, J. Biol. Chem., 276, 20821–20823 (2001). 39. Y. Imamoto, H. Kamikubo, M. Harigai et al., Light-induced global conformational change of photoactive yellow protein in solution, Biochemistry, 41, 13595–13601 (2002). 40. J. M. Zhao, H. Lee, R. A. Nome et al., Single-molecule detection of structural changes during Per-Arnt-Sim (PAS) domain activation, Proc. Natl. Acad. Sci. USA, 103, 11561–11566 (2006). 41. G. Rubinstenn, G. W. Vuister, F. A. Mulder et al., Structural and dynamic changes of photoactive yellow protein during its photocycle in solution, Nat. Struct. Biol., 5, 568–570 (1998). 42. C. Bernard, K. Houben, N. M. Derix et al., The solution structure of a transient photoreceptor intermediate: Delta25 photoactive yellow protein, Structure, 13, 953–962 (2005). 43. M. E. van Brederode, W. D. Hoff, I. H. van Stokkum et al., Protein folding thermodynamics applied to the photocycle of the photoactive yellow protein, Biophys. J., 71, 365–380 (1996). 44. N. Derix, R. Wechselberger, M. A. van der Horst et al., Lack of negative charge in the E46Q mutant of photoactive yellow protein prevents partial unfolding of the blue-shifted intermediate, Biochemistry, 42, 14501–14506 (2003). 45. S. Anderson, S. Crosson, K. Moffat, Short hydrogen bonds in photoactive yellow protein, Acta Crystallogr. D Biol. Crystallogr., 60, 1008–1016 (2004). 46. T. Kortemme, A. V. Morozov, D. Baker, An orientation-dependent hydrogen bonding potential improves prediction of specificity and structure for proteins and protein-protein complexes, J. Mol. Biol., 326, 1239–1259 (2003). 47. S. Yamaguchi, H. Kamikubo, K. Kurihara et al., Low-barrier hydrogen bond in photoactive yellow protein, Proc. Natl. Acad. Sci. USA, 106, 440–444 (2009). 48. S. Z. Fisher, S. Anderson, R. Henning et al., Neutron and X-ray structural studies of short hydrogen bonds in photoactive yellow protein (PYP), Acta Crystallogr. D Biol. Crystallogr., 63, 1178–1184 (2007). 49. B. Nie, J. Stutzman and A. Xie, A vibrational spectral marker for probing the hydrogen bonding status of protonated Asp and Glu residues, Biophys. J., 88, 2833–2847 (2005). 50. A. R. Kroon, W. D. Hoff, H. P. Fennema et al., Spectral tuning, fluorescence, and photoactivity in hybrids of photoactive yellow protein, reconstituted with native or modified chromophores, J. Biol. Chem., 271, 31949–31956 (1996). 51. W. D. Hoff, I. H. M. van Stokkum, J. Gural and K. J. Hellingwerf, Comparison of acid denaturation and light activation in the eubacterial blue-light receptor photoactive yellow protein, Biochim. Biophys. Acta, 1322, 151–162 (1997). 52. R. L. Thurlkill, G. R. Grimsley, J. M. Scholtz and C. N. Pace, pK values of the ionizable groups of proteins, Protein Sci., 15, 1214–1218 (2006). 53. T. E. Meyer, S. Devanathan, T. T. Woo et al., Site-specific mutations provide new insights into the origin of pH effects and alternative spectral forms in the photoactive yellow protein from Halorhodospira halophilia, Biochemistry, 42, 3319–3325 (2003). 54. Y. Imamoto, H. Koshimizu, K. Mihara et al., Roles of amino acid residues near the chromophore of photoactive yellow protein, Biochemistry, 40, 4679–4685 (2001). 55. A. F. Philip, K. T. Eisenman, G. A. Papadantonakis and W. D. Hoff, Functional tuning of photoactive yellow protein by active site residue 46, Biochemistry, 47, 13800–13810 (2008). 56. M. Sugishima, N. Tanimoto, K. Soda et al., Structure of photoactive yellow protein (PYP) E46Q mutant at 1.2 A resolution suggests how Glu46 controls the spectroscopic and kinetic characteristics of PYP, Acta Crystallogr. D Biol. Crystallogr., 60, 2305–2309 (2004).
854 Hydrogen Bonding and Transfer in the Excited State 57. U. K. Genick, S. Devanathan, T. E. Meyer et al., Active site mutants implicate key residues for control of color and light cycle kinetics of photoactive yellow protein, Biochemistry, 36, 8–14 (1997). 58. K. Mihara, O. Hisatomi, Y. Imamoto et al., Functional expression and site-directed mutagenesis of photoactive yellow protein, J. Biochem., 121, 876–880 (1997). 59. Y. Imamoto, Y. Shirahige, F. Tokunaga et al., Low-temperature Fourier transform infrared spectroscopy of photoactive yellow protein, Biochemistry, 40, 8997–9004 (2001). 60. M. Unno, M. Kumauchi, J. Sasaki et al., Resonance Raman spectroscopy and quantum chemical calculations reveal structural changes in the active site of photoactive yellow protein, Biochemistry, 41, 5668–5674 (2002). 61. D. H. Pan, A. Philip, W. D. Hoff and R. A. Mathies, Time-resolved resonance Raman structural studies of the pB0 intermediate in the photocycle of photoactive yellow protein, Biophys. J., 86, 2374–2382 (2004). 62. M. Unno, M. Kumauchi, F. Tokunaga and S. Yamauchi, Vibrational assignment of the 4-hydroxycinnamyl chromophore in photoactive yellow protein, J. Phys. Chem. B, 111, 2719–2726 (2007). 63. M. Unno, M. Kumauchi, J. Sasaki et al., Evidence for a protonated and cis configuration chromophore in the photobleached intermediate of photoactive yellow protein, J. Am. Chem. Soc., 122, 4233–4234 (2000). 64. M. Unno, M. Kumauchi, J. Sasaki et al., Assignment of resonance Raman spectrum of photoactive yellow protein in its long-lived blue-shifted intermediate, J. Phys. Chem. B, 107, 2837–2845 (2003). 65. W. D. Hoff, S. L. S. Kwa, R. van Grondelle and K. J. Hellingwerf, Low temperature absorbance and fluorescence spectroscopy of the photoactive yellow protein from Ectothiorhodospira halophila, Photochem. Photobiol., 56, 529–539 (1992). 66. R. Brudler, R. Rammelsberg, T. T. Woo et al., Structure of the I1 early intermediate of photoactive yellow protein by FTIR spectroscopy, Nat. Struct. Biol., 8, 265–270 (2001). 67. D. Hoersch, H. Otto, M. A. Cusanovich and M. P. Heyn, Distinguishing chromophore structures of photocycle intermediates of the photoreceptor PYP by transient fluorescence and energy transfer, J. Phys. Chem. B, 112, 9118–9125 (2008). 68. L. M. Groot, L. J. G. W. van Wilderen, D. S. Larsen et al., Initial steps of signal generation in photoactive yellow protein revealed with femtosecond mid-infrared spectroscopy, Biochemistry, 42, 10054–10059 (2003). 69. K. Heyne, O. F. Mohammed, A. Usman et al., Structural evolution of the chromophore in the primary stages of trans/cis isomerization in photoactive yellow protein, J. Am. Chem. Soc., 127, 18100–18106 (2005). 70. L. van Wilderen, M. A. van der Horst and I. H. M. van Stokkum et al., Ultrafast infrared spectroscopy reveals a key step for successful entry into the photocycle for photoactive yellow protein, Proc. Natl. Acad. Sci. USA, 103, 15050–15055 (2006). 71. Chosrowjan H., Mataga N., Nakashima N. et al., Femtosecond-picosecond fluorescence studies on excited state dynamics of photoactive yellow protein from Ectothiorhodospira halophila, Chem. Phys. Lett., 270, 267–272 (1997). 72. M. Vengris, M. A. van der Horst, G. Zgrablic et al., Contrasting the excited-state dynamics of the photoactive yellow protein chromophore: protein versus solvent environments, Biophys. J., 87, 1848–1857 (2004). 73. P. Changenet, H. Zhang, M. J. van der Meer et al., Subpicosecond fluorescence upconversion measurements of primary events in yellow proteins, Chem. Phys. Lett., 282, 276–282 (1998). 74. H. Hanada, Y. Kanematsu, S. Kinoshita et al., Ultrafast fluorescence spectroscopy of photoactive yellow protein, J. Lumin., 94–95 593–596 (2001). 75. N. Mataga, H., Y. Shibata et al., Ultrafast photoinduced reaction dynamics of photoactive yellow protein (PYP): observation of coherent oscillations in the femtosecond fluorescence decay dynamics, Chem. Phys. Lett., 352, 220–225 (2002). 76. R. Nakamura, N. Hamada, H. Ichida et al., Coherent oscillations in ultrafast fluorescence of photoactive yellow protein, J. Chem. Phys., 127, 215102 (2007). 77. N. Mataga, H. Chosrowjan, Y. Shibata et al., Effects of modification of protein nanospace structure and change of temperature on the femtosecond to picosecond fluorescence dynamics of photoactive yellow protein, J. Phys. Chem., 104, 5191–5199 (2000). 78. N. Mataga, H. Chosrowjan, S. Taniguchi et al., Ultrafast photoreactions in protein nanospaces as revealed by fs fluorescence dynamics measurements on photoactive yellow protein and related systems, Phys. Chem. Chem. Phys., 5, 2454–2460 (2003).
Changes in Active Site Hydrogen Bonding upon Formation of the Electronically Excited
855
79. R. Nakamura, Y. Kanematsu, M. Kumauchi et al., Photo-induced protein dynamics measured by femtosecond timeresolved luminescence, J. Lumin., 102, 21–26 (2003). 80. H. Chosrowjan, S. Taniguchi, N. Mataga et al., Low-frequency vibrations and their role in ultrafast photoisomerization reaction dynamics of photoactive yellow protein, J. Phys. Chem. B, 108, 2686–2698 (2004). 81. A. Xie, L. Meer, W. Hoff and R. H. Austin, Long-lived amide I vibrational modes in myoglobin, Phys. Rev. Lett., 84, 5435–5438 (2000). 82. D. S. Larsen, I. H. van Stokkum, M. Vengris et al., Incoherent manipulation of the photoactive yellow protein photocycle with dispersed pump-dump-probe spectroscopy, Biophys. J., 87, 1858–1872 (2004). 83. T. Gensch, C. C. Gradinaru, I. H. M. van Stokkum et al., The primary photoreaction of photoactive yellow protein (PYP): anisotropy changes and excitation wavelength dependence, Chem. Phys. Lett., 356, 347–354 (2002). 84. Y. Imamoto, M. Kataoka, F. Tokunaga et al., Primary photoreaction of photoactive yellow protein studied by subpicosecond-nanosecond spectroscopy, Biochemistry, 40, 6047–6052 (2001). 85. S. Devanathan, A. Pacheco, L. Ujj et al., Femtosecond spectroscopic observations of initial intermediates in the photocycle of the photoactive yellow protein from Ectothiorhodospira halophila, Biophys. J., 77, 1017–1023 (1999). 86. A. Baltusˇka, I. H. M. van Stokkum, A. Kroon et al., The primary events in the photoactivation of yellow protein, Chem. Phys. Lett., 270, 263–266 (1997). 87. F. Gai, K. C. Hasson, J. C. McDonald and P. A. Anfinrud, Chemical dynamics in proteins: The photoisomerization of retinal in bacteriorhodopsin, Science, 279, 1886–1891 (1998). 88. M. E. van Brederode, T. Gensch, W. D. Hoff et al., Photoinduced volume change and energy storage associated with the early transformations of the photoactive yellow protein from Ectothiorhodospira halophila, Biophys. J., 68, 1101–1109 (1995). 89. L. L. Premvardhan, M. A. van der Horst, K. J. Hellingwerf and R. van Grondelle, Stark spectroscopy on photoactive yellow protein, E46Q, and a nonisomerizing derivative, probes photo-induced charge motion, Biophys. J., 84, 3226–3239 (2003). 90. M. Mizuno, N. Hamada, F. Tokunaga and Y. Mizutani, Picosecond protein response to the chromophore isomerization of photoactive yellow protein: selective observation of tyrosine and tryptophan residues by time-resolved ultraviolet resonance Raman spectroscopy, J. Phys. Chem. B, 111, 6293–6296 (2007). 91. T. E. Meyer, G. Tollin, P. Causgrove et al., Picosecond decay kinetics and quantum yield of fluorescence of the photoactive yellow protein from the halophilic purple phototrophic bacterium, Ectothiorhodospira halophila, Biophys. J., 59, 988–991 (1991). 92. L. Ujj, S. Devanathan, T. E. Meyer et al., New photocycle intermediates in the photoactive yellow protein from Ectothiorhodospira halophila: picosecond transient absorption spectroscopy, Biophys. J., 75, 406–412 (1998). 93. P. Changenet-Barret, P. Plaza, M. M. Martin et al., Role of arginine 52 on the primary photoinduced events in the PYP photocycle, Chem. Phys. Lett., 434, 320–325 (2007). 94. C. P. Joshi, H. Otto, D. Hoersch et al., Strong hydrogen bond between glutamic acid 46 and chromophore leads to the intermediate spectral form and excited state proton transfer in the Y42F mutant of the photoreceptor photoactive yellow protein, Biochemistry, 48, 9980–9993 (2009).
38 Excited State Dynamics in the Light-Driven Enzyme Protochlorophyllide Oxidoreductase (POR) Marie Louise Groot1 and Derren James Heyes2 1
Department of Physics and Astronomy, Faculty of Sciences, Vrije Universiteit, De Boelelaan 1081, 1081 HV Amsterdam, The Netherlands 2 Manchester Interdisciplinary Biocentre, University of Manchester, 131 Princess Street, Manchester M1 7DN, UK
38.1 Introduction Enzymes can catalyse chemical reactions with rate enhancements of up to 1017 compared to the equivalent reaction in solution [1]. However, understanding exactly how enzymes achieve their huge catalytic power remains challenging [1–3] as catalysis is generally limited by diffusion-associated processes, such as the binding of substrates and cofactors, or by conformational changes of the enzyme. Hence, for most enzymes that require mixing strategies to initiate the reaction it is impossible to directly study the real time formation of catalytic intermediates. However, this possibility is afforded in light-driven enzymes, such as photosynthetic reaction centres [4–6], DNA photolyase [7, 8] and protochlorophyllide oxidoreductase [9–11], where the reaction can be initiated with a trigger that is more rapid (i.e. a laser pulse) than the fastest dynamics involved. Reaction centres are transmembrane pigment–protein complexes that are responsible for catalysing lightdriven charge separation and represent one of only a very few systems where electron transfer between redox centres can be monitored with femtosecond time resolution. The initial steps in the electron-transfer reaction take place on a picosecond timescale, leading eventually to the two-electron reduction of a bound quinone molecule on a millisecond timescale, after the subsequent absorption of two photons [4–6]. DNA-photolyase induces the repair of UV-induced lesions in DNA by scission of covalent bonds between neighbouring
Hydrogen Bonding and Transfer in the Excited State, Volume I & II © 2011 John Wiley & Sons, Ltd. ISBN: 978-0-470-66677-7
Edited by Ke-Li Han and Guang-Jiu Zhao
858 Hydrogen Bonding and Transfer in the Excited State
pyrimidines. Photoactivation of an FAD cofactor occurs via a fast (picosecond timescale) multistep electron transfer through a chain of three tryptophan residues [7, 8]. However, in the present chapter we focus on another light-driven enzyme, protochlorophyllide oxidoreductase (POR), which also provides an opportunity to study initial ultrafast processes that are required for catalysis. Detailed reviews on POR can be found in Refs [9] to [11] but in the present chapter we give a limited overview of the enzyme and provide a detailed description of recent time-resolved transient absorption and fluorescence experiments.
38.2 Protochlorophyllide Oxidoreductase (POR) Chlorophyll is the most abundant pigment on Earth and is essential for life, directly or indirectly, as the cofactor for the photosynthetic proteins that harvest sunlight and convert it into photochemical energy for the cell. POR catalyses one of the latter steps in the chlorophyll biosynthesis pathway, the light-dependent trans addition of hydrogen across the C17–C18 double bond of the D-ring of protochlorophyllide (Pchlide) to produce chlorophyllide (Chlide) (Figure 38.1) [9–11]. In a biological context, the light-driven reaction catalysed by this enzyme provokes a profound change in the morphological development of the plant that is visualized at the ultrastructural level as a modification and reorganization of plastid membranes [9, 10]. In etiolated plants POR is mainly found to exist as a ternary complex in highly organized networks of tubular membranes termed prolamellar bodies [12, 13]. After illumination and subsequent formation of Chlide, there is a disintegration of the POR-pigment aggregates, resulting in the dispersion of the prolamellar bodies [14]. In addition to POR, non-flowering land plants, algae and cyanobacteria possess a light-independent Pchlide reductase, which consists of three separate subunits and allows these organisms to produce Chlide in the dark [15, 16].
Figure 38.1 Light-driven reduction of protochlorophyllide (Pchlide) catalysed by protochlorophyllide oxidoreductase (POR). The trans addition of hydrogen across the C17–C18 double bond of Pchlide to form chlorophyllide (Chlide), catalysed by the light-driven enzyme POR, is a key regulatory step within the chlorophyll biosynthetic pathway. The enzyme requires NADPH as a cofactor and the dashed box indicates the double bond that is reduced during the reaction
Excited State Dynamics in the Light-Driven Enzyme Protochlorophyllide Oxidoreductase (POR) 859
Figure 38.2 Catalytic cycle of POR. An overall scheme for the catalytic cycle of POR showing the stepwise formation of the reaction intermediates together with the rate constants that have been calculated previously [18, 26, 29]
38.3 Catalytic Mechanism of POR The POR enzyme requires NADPH as a cofactor and in the dark it is found in a ternary complex with the two substrates [17]. Various experimental approaches [18–28] have been used to solve the catalytic mechanism of POR and have resulted in a detailed understanding of the reaction cycle (Figure 38.2). The formation of the ternary enzyme–substrate complex is the rate-limiting step in the overall catalytic cycle and involves multiple steps, which are controlled by slow conformational changes in the protein [18]. During the reaction, the Pchlide molecule performs the function of the photoreceptor and upon illumination it is proposed that a hydride is transferred from the pro-S face of NADPH to the C17 position of the Pchlide molecule [19, 20] and a conserved Tyr residue donates a proton to the C18 position [21]. By using low temperature spectroscopy to trap catalytic intermediates the reaction cycle has been shown to consist of an initial light-driven reaction [22] followed by a series of subsequent (slower) dark reactions [23, 24]. The initial light-driven step involves hydride transfer from the NADPH molecule to form a charge-transfer complex, which facilitates the subsequent protonation of the C18 position of Pchlide during the first of the ‘dark’ reactions [25]. Recent laser photoexcitation studies have revealed that these two enzymatic H-transfer reactions occur in a sequential mechanism on the microsecond timescale [26]. By combining studies of the temperature and isotopic dependence of the rate of Pchlide reduction it was shown that both H-transfer reactions proceed by using quantum mechanical tunnelling that is coupled to specific motions (vibrations) in the protein [26]. Further studies on the role of the bulk solvent on catalysis suggested that solvent slaved motions control proton tunnelling but not hydride tunnelling, implying that a long-range ‘dynamic network’ from solvent to the enzyme active site facilitates proton transfer [27, 28]. Furthermore, a series of recent studies, aimed at resolving the remaining ‘dark’ steps in the reaction, have shown that these latter catalytic events represent a series of ordered product release and coenzyme binding processes [23, 24]. Thus, following the rapid formation of the ternary enzyme–product complex, NADPþ is first released from the enzyme and is then replaced by the NADPH coenzyme. This is followed by the release of the Chlide product and subsequent binding of the Pchlide substrate to complete the catalytic cycle [24]. Laser photoexcitation experiments have been used to follow the interconversion of these various bound/ unbound Chlide product species on the millisecond to second timescales, which suggested that conformational changes and/or reorganization of the protein were required to facilitate the release of the products and substrate rebinding [29]. This is also in agreement with previous cryogenic measurements, which revealed that these stages of the reaction could only proceed above the ‘glass transition’ temperature of proteins [23, 24].
860 Hydrogen Bonding and Transfer in the Excited State
38.4 Ultrafast Catalytic Processes of the Isolated Pchlide Species Although the actual chemical steps in the POR-catalysed reaction proceed on a much slower timescale [26], catalysis is completely dependent on excited state processes in the Pchlide molecule, which occur on an ultrafast picosecond timescale. Several recent studies have attempted to characterize some of these short-lived Pchlide species and we now have a more detailed description of the excited-state dynamics that precede catalysis [30–36]. To clarify the complex photochemical reactions that occur in the POR enzyme Dietzek et al. have used time-resolved absorption and fluorescence spectroscopy to study the excited-state chemistry of the isolated Pchlide substrate [30–32]. To mimic the different environmental conditions in the enzyme complex, various solvents were used, which demonstrated that the excited state dynamics of Pchlide strongly depends on the solvent polarity [30, 31]. In polar solvents, such as methanol and acetonitrile, the excited state relaxation dynamics are multiexponential with three distinguishable timescales of 4.0–4.5 ps for vibrational relaxation and vibrational energy redistribution of the initially excited S1 state; 22–27 ps for the formation of an intermediate state, most likely with a charge transfer character; and 200 ps for the decay of this intermediate state back to the ground state [31]. However, in the non-polar solvent cyclohexane, only the 4.5 ps relaxational process can be observed [31]. In addition, by increasing the viscosity of the solvent by adding glycerol to a methanolic solution of Pchlide, two of the former relaxation processes are found to be decelerated [30, 31]. This suggests that not only is the vibrational cooling of the S1 state slowed in the more viscous surrounding, but the formation rate of the intermediate state with charge transfer character can also be reduced. Hence, it is likely that nuclear motions along the reaction coordinate are involved in the charge transfer and that the formation of the intermediate state may be related to the dynamic solvation process of Pchlide in the S1 state [30–32]. Consequently, the site-specific solvation of photoexcited Pchlide in methanol has been investigated by using time-dependent density functional theory [33]. It was shown that intermolecular site-specific coordination and hydrogen-bonding interactions between the Pchlide and methanol molecules play a very important role in the steady-state and time-resolved spectra of the pigment. A coordinated and hydrogen-bonded Pchlide-(MeOH)4 complex was proposed to represent the intermediate state that is formed in 22–27 ps in the time-resolved spectroscopic studies [33]. Moreover, the intermolecular coordination and hydrogen bonds between the Pchlide and methanol molecules can be strengthened in the electronically excited state of Pchlide, which, in turn, induces the site-specific solvation of the photoexcited Pchlide molecule. It was also shown that all of the steady-state and time-resolved spectral features of Pchlide in different solvents can be explained by the formation of this hydrogen-bonded intermediate state after the site-specific solvation [33]. A more recent pump–probe study in the visible and near-IR regions has extended the current model for the excited-state dynamics of Pchlide into the sub-ps and ns-time domains (Figure 38.3) [34]. Following excitation, an initial ultrafast 450 fs process is suggested to represent the motion out of the Franck–Condon region on the excited state surface. In addition, a long-lived photointermediate with absorption bands at approximately 530 and 890 nm builds up in 3.5 ns. This long-lived photointermediate is ascribed to a low-lying excited triplet state and is formed directly from the fully equilibrated excited S1 state [34]. The data on the isolated Pchlide pigment have provided important insights into the mechanism of lightactivation in the enzyme–substrate complex and suggest that the enzyme-catalysed photoreduction of Pchlide is strongly influenced by the protein environment in the substrate binding pocket. It has been proposed that specific chromophore–protein interactions in the enzyme–substrate complex may quench the non-productive and high-risk formation of a triplet state by efficiently promoting the biologically relevant photoreduction reaction [34]. It is likely that the polarity of the protein environment controls the reaction and that the formation of an intramolecular charge transfer complex plays a significant role in the excited-state reaction dynamics of the enzyme-bound Pchlide [37]. The dipolar nature of the intramolecular charge-transfer state is expected to result from the presence of the electron-withdrawing carbonyl group on the D-ring of the Pchlide molecule.
Excited State Dynamics in the Light-Driven Enzyme Protochlorophyllide Oxidoreductase (POR) 861
Figure 38.3 Schematic model of the light-induced relaxation processes in Pchlide. The excited state relaxation processes have been proposed by studies on Pchlide in solution [30–32, 34]. FC refers to the Franck–Condon region, SX denotes a secondary S1 excited state on the S1 potential energy surface and SICT an intramolecular charge transfer state
As the catalytic photoreduction of the C17–C18 double bond involves the initial transfer of a hydride ion to the C18 position of Pchlide [19, 20, 25, 26], a decreased local electron density at this position would enhance this nucleophilic attack process. Hence, it is interesting that the excited-state charge-localization observed in the isolated Pchlide molecule might lead to an electronic configuration that favours the reduction of the C17–C18 double bond in the enzyme active site [37].
38.5 Ultrafast Catalytic Processes of the Enzyme-Bound Pchlide Species To investigate the initial ultrafast steps that are directly associated with the formation of Chlide we have recently used femtosecond transient absorption measurements to study the excited state processes in the enzyme-bound Pchlide species [35, 36]. Immediately after excitation a negative signal was observed due to the bleached absorption and stimulated emission of Pchlide, peaking at approximately 640 nm. After a further few picoseconds a negative signal, corresponding to stimulated emission, appeared at approximately 674 nm, which occurred with two time constants of 3 and 400 ps [35]. The absence of this negative signal in measurements on Pchlide in solution and in a mutant (Y189F) in which the putative proton donor Tyr189 was replaced by a Phe, led to the conclusion that this was a catalytic product [35]. Unfortunately, the exact nature of this excited state species, now referred to as I675, could not be fully determined at this stage. However, with the recent theoretical studies on the Pchlide excited state [33] and the identification of the hydride and proton transfer reactions [26] it is likely that I675 is a precursor species in which Pchlide forms a strongly hydrogen-bonded complex with residues in its direct environment and/or NADPH that is essential for the subsequent hydride and proton transfer steps to proceed on a microsecond timescale [26]. In a subsequent study [36], the ultrafast reaction dynamics of the enzyme–substrate complex were analysed under single pulse conditions. By using a Lissajous sample scanner, in combination with very high detection sensitivity, reaction rates and quantum yields were measured as a function of the total number of laser shots previously seen by the sample. The dynamics of the POR–substrate complex proved to be very strongly dependent on the number of pulses applied to the sample [36]. Following a single laser pulse only minor dynamics in the Pchlide region were observed with no I675 formation, but after further laser pulses stimulated emission from I675 appeared on the same timescale as reported previously [35]. To analyse the data in more
862 Hydrogen Bonding and Transfer in the Excited State
Figure 38.4 Model for the ultrafast catalytic reactions in POR [36]. The laser induced dynamics in complexes that had not been excited before results in the kinetic scheme for inactive enzymes, showing only Pchlide photochemistry (from Pchlide I to Pchlide II to Pchlide III). In complexes that had been previously excited, the intermediate I675 appears on a picosecond timescale and results in the kinetic scheme for active enzymes, including the intrinsic photochemistry from the Chlide product accumulated in previous scans (Chlideaccum )
detail the full set of spectra were fitted to a model in which the POR enzymes were divided into two populations: one that is inactive and shows intrinsic Pchlide photochemistry but does not lead to product, and an active population that also shows the formation of I675 (Figure 38.4). The fraction of active enzymes was found to be dependent on the number of laser pulses and demonstrates that the rate and quantum yield of formation of the I675 intermediate is significantly enhanced after the Pchlide substrate has cycled through the excited state at least once [36]. Therefore, it appears that a first single photon is needed to activate the POR complex, whereas a second photon then induces catalysis. This remarkable effect was suggested to arise from a more favourable catalytic configuration of the enzyme–substrate complex, caused by the changed electron distribution in the Pchlide excited state [36]. Consequently, rapid scan FTIR spectroscopy was used to investigate whether conformational changes in the enzyme are triggered upon the absorption of a photon. The corresponding spectral changes in the mid-infrared showed that initial laser photoexcitation induced absorption changes associated mainly with protein bands, amide I and II, whereas subsequent laser shots induced absorption changes that could be assigned to the disappearance of NADPH and Pchlide together with the appearance of NADPþ and Chlide [36]. This suggested that indeed the first photon produces a conformational change in the POR enzyme that switches it from inactive to active, and that when in the active state a second photon can induce catalysis that leads to Chlide formation with a quantum yield of 0.3.
38.6 Conclusions Light-driven enzymes, such as POR, provide ideal model systems for studying excited state processes that are linked to enzymatic catalysis. In POR it is likely that several excited state Pchlide species are formed within a few nanoseconds [30–37], which are essential for the subsequent chemical steps to occur on the microsecond timescale [26]. An intramolecular charge transfer complex [30–32] and a hydrogen bonded intermediate [33, 36] are proposed to play a significant role in the excited-state reaction dynamics of the enzyme-bound
Excited State Dynamics in the Light-Driven Enzyme Protochlorophyllide Oxidoreductase (POR) 863
Pchlide and are thought to be linked to conformational changes in the protein [36]. We expect that further experiments on POR will lead to a full identification of the reaction pathway at room temperature and the identification of the initial intermediate state(s) involved, together with a better understanding of the structural changes that lie at the origin of the activation process. The resolution of the structure of POR, either by X-ray diffraction or by NMR techniques, will undoubtedly be important in this process.
References 1. S. J. Benkovic and S. Hammes-Schiffer, A perspective on enzyme catalysis, Science, 301, 1196–1202 (2003). 2. J. Villa and A. Warshel, Energetics and dynamics of enzymatic reactions, J. Phys. Chem. B, 105, 7887–7907 (2001). 3. E. Z. Eisenmesser, O. Millet, W. Labeikovsky, et al., Intrinsic dynamics of an enzyme underlies catalysis, Nature, 438, 117–121 (2005). 4. M. Y. Okamura, M. L. Paddock, M. S. Graige and G. Feher, Proton and electron transfer in bacterial reaction centers, Biochim. Biophys. Acta-Bioenerg., 1458, 148–163 (2000). 5. M. L. Paddock, G. Feher and M. Y. Okamura, Proton transfer pathways and mechanism in bacterial reaction centers, FEBS Lett., 555, 45–50 (2003). 6. C. A. Wraight, Proton and electron transfer in the acceptor quinone complex of photosynthetic reaction centers from Rhodobacter sphaeroides, Frontiers Biosci., 9, 309–337 (2004). 7. C. Aubert, M. H. Vos, P. Mathis et al., Intraprotein radical transfer during photoactivation of DNA photolyase, Nature, 405, 586–590 (2000). 8. A. Mees, T. Klar, P. Gnau et al., Crystal structure of a photolyase bound to a CPD-like DNA lesion after in situ repair, Science, 306, 1789–1793 (2004). 9. N. Lebedev and M. P. Timko, Protochlorophyllide photoreduction, Photosynth. Res., 58, 5–23 (1998). 10. T. Masuda and K. Takamiya, Novel insights into the enzymology, regulation and physiological functions of lightdependent protochlorophyllide oxidoreductase in angiosperms, Photosynth. Res., 81, 1–29 (2004). 11. D. J. Heyes and C. N. Hunter, Making light work of enzyme catalysis: protochlorophyllide oxidoreductase, Trends Biochem. Sci., 30, 642–649 (2005). 12. C. Sundqvist and C. Dahlin, With chlorophyll pigments from prolamellar bodies to light-harvesting complexes, Physiol. Plantarum, 100, 748–759 (1997). 13. F. Franck, U. Sperling, G. Frick et al., Regulation of etioplast pigment-protein complexes, inner membrane architecture, and protochlorophyllide a chemical heterogeneity by light-dependent NADPH:protochlorophyllide oxidoreductases A and B Plant Physiol., 124, 1678–1696 (2000). 14. L.B. Zhong, B. Wiktorsson, M. Ryberg and C. Sundqvist, The Shibata shift: effects of in vitro conditions on the spectral blue-shift of chlorophyllide in irradiated isolated prolamellar bodies, J. Photochem. Photobiol. B-Biol., 36, 263–270 (1996). 15. Y. Fujita and C. E. Bauer, Reconstitution of light-independent protochlorophyllide reductase from purified BchL and BchN-BchB subunits - in vitro confirmation of nitrogenase-like features of a bacteriochlorophyll biosynthesis enzyme, J. Biol. Chem., 275, 23583–23588 (2000). 16. M. J. Br€ocker, S. Virus, S. Ganskow et al., ATP-driven reduction by dark-operative protochlorophyllide oxidoreductase from Chlorobium tepidum mechanistically resembles nitrogenase catalysis, J. Biol. Chem., 283, 10559–10567 (2008). 17. W. T. Griffiths, Reconstitution of chlorophyllide formation by isolated etioplast membranes, Biochem. J., 174, 681–692 (1978). 18. D. J. Heyes, B. R. K. Menon, M. Sakuma and N. S. Scrutton, Conformational events during ternary enzyme-substrate complex formation are rate limiting in the catalytic cycle of the light-driven enzyme protochlorophyllide oxidoreductase, Biochemistry, 47, 10991–10998 (2008). 19. V. Valera, M. Fung, A. N. Wessler and W. R. Richards, Synthesis of 4R- and 4S-tritium labeled NADPH for the determination of the coenzyme stereospecificity of NADPH: protochlorophyllide oxidoreductase, Biochem. Biophys. Res. Commun., 148, 515–520 (1987).
864 Hydrogen Bonding and Transfer in the Excited State 20. T. P. Begley and H. Young, Protochlorophyllide reductase. 1. Determination of the regiochemistry and the stereochemistry of the reduction of protochlorophyllide to chlorophyllide, J. Am. Chem. Soc., 111, 3095–3096 (1989). 21. H. M. Wilks and M. P. Timko, A light-dependent complementation system for analysis of NADPH:protochlorophyllide oxidoreductase. Identification and mutagenesis of two conserved residues that are essential for enzyme activity, Proc. Natl. Acad. Sci. USA, 92, 724–728 (1995). 22. D. J. Heyes, A. V. Ruban, H. M. Wilks and C. N. Hunter, Enzymology below 200 K: the kinetics and thermodynamics of the photochemistry catalyzed by protochlorophyllide oxidoreductase, Proc. Natl. Acad. Sci. USA, 99, 11145–11150 (2002). 23. D. J. Heyes, A. V. Ruban and C. N. Hunter, Protochlorophyllide oxidoreductase: “dark” reactions of a light-driven enzyme”, Biochemistry, 42, 523–528 (2003). 24. D. J. Heyes and C. N. Hunter, Identification and characterization of the product release steps within the catalytic cycle of protochlorophyllide oxidoreductase, Biochemistry, 43, 8265–8271 (2004). 25. D. J. Heyes, P. Heathcote, S. E. J. Rigby et al., The first catalytic step of the light-driven enzyme protochlorophyllide oxidoreductase proceeds via a charge transfer complex, J. Biol. Chem., 281, 26847–26853 (2006). 26. D. J. Heyes, M. Sakuma, S. De Visser and N. S. Scrutton, Nuclear quantum tunneling in the light-activated enzyme protochlorophyllide oxidoreductase, J. Biol. Chem., 284, 3762–3767 (2008). 27. G. Durin, A. Delaunay, C. Darnault et al., Simultaneous measurements of solvent dynamics and functional kinetics in a light-activated enzyme, Biophys. J., 96, 1902–1910 (2009). 28. D. J. Heyes, M. Sakuma and N. S. Scrutton, Solvent slaved protein motions accompany proton but not hydride tunneling in light-activated protochlorophyllide oxidoreductase, Angew. Chem. Int. Ed., 48, 3850–3853 (2009). 29. D. J. Heyes, M. Sakuma and N. S. Scrutton, Laser excitation studies of the product release steps in the catalytic cycle of the light-driven enzyme, protochlorophyllide oxidoreductase, J. Biol. Chem., 282, 32015–32020 (2007). 30. B. Dietzek, R. Maksimenka, T. Siebert et al., Excited-state processes in protochlorophyllide a: a femtosecond timeresolved absorption study, Chem. Phys. Lett., 397, 110–115 (2004). 31. B. Dietzek, W. Kiefer, J. Popp et al., Solvent effects on the excited-state processes of protochlorophyllide: a femtosecond time-resolved absorption study, J. Phys. Chem. B., 110, 4399–4406 (2006). 32. B. Dietzek, S. Tschierlei, G. Hermann et al., The excited-state chemistry of protochlorophyllide a: a time-resolved fluorescence study, ChemPhysChem, 7, 1727–1733 (2006). 33. G. J. Zhao and K. L. Han, Site-specific solvation of the photoexcited protochlorophyllide a in methanol: formation of the hydrogen-bonded intermediate state induced by hydrogen-bond strengthening, Biophys. J., 94, 38–46 (2008). 34. B. Dietzek, W. Kiefer, A. Yartsev et al., Protochlorophyllide a: a comprehensive photophysical picture, ChemPhysChem, 10, 144–150 (2009). 35. D. J. Heyes, C. N. Hunter, I. H. M. van Stokkum et al., Ultrafast enzymatic reaction dynamics in protochlorophyllide oxidoreductase, Nat. Struct. Biol., 10, 491–492 (2003). 36. O. A. Sytina, D. J. Heyes, C. N. Hunter et al., Conformational changes in an ultrafast light-driven enzyme determine catalytic activity, Nature, 456, 1001–1004 (2008). 37. M. Schmitt, B. Dietzek, G. Hermann and J. Popp, Femtosecond time-resolved spectroscopy on biological photoreceptor chromophores, Laser Photonics Rev., 1, 57–78 (2007).
39 Photodissociation of Molecules in Pure and Doped Water and in Nitrogen Heterocyclic Clusters in the Excited State Michal F arnık1, Petr Slavıcek2 and Udo Buck3 1
J. Heyrovsky Institute of Physical Chemistry, Academy of Sciences, Dolejsˇkova 3, 182 23 Prague 8, Czech Republic 2 Department of Physical Chemistry, Institute of Chemical Technology, Technick a 5, Prague 6, Czech Republic 3 Max-Planck-Institut f€ur Dynamik und Selbstorganisation, Bunsenstrasse 10, D-37073 G€ ottingen, Germany
39.1 Introduction The role of hydrogen bonding in the ground state chemistry is widely acknowledged. The most prominent examples of its significance are the structure of water and ice, the pairing of base pairs in nucleic acids, molecular recognition or enzyme catalysis, and so on. All of the above examples are related to essential conditions of life. For this reason, hydrogen bonding phenomena became a subject of extensive studies. Photoinduced processes in hydrogen bonded systems are explored to a much lesser extent. While photochemistry of systems with intramolecular hydrogen bonds have been studied for a relatively long time, for example, in a context of sunscreen protection or fluorescence spectroscopy, photochemistry of intermolecular hydrogen bonds has only recently become a major research subject [1]. There are interesting questions to be addressed: Is there any photochemistry specific for hydrogen bonded systems? And if the answer is yes: Is the different photochemistry related to biophysically relevant processes? The answer to the first question is affirmative, photoinduced processes are controlled by the ground state structure and the X-H Y structural motif implies ultrafast excited state hydrogen (or proton) transfer as a typical reaction of hydrogen bonded systems.
Hydrogen Bonding and Transfer in the Excited State, Volume I & II © 2011 John Wiley & Sons, Ltd. ISBN: 978-0-470-66677-7
Edited by Ke-Li Han and Guang-Jiu Zhao
866 Hydrogen Bonding and Transfer in the Excited State
The hydrogen bond was suggested to be a key component in protecting nucleic acids against radiation damage [2–4]. Recently, it was also shown that an analogical mechanism can protect another class of biomolecules-peptides [5]. Experimental studies of hydrogen bonding in the ground state are relatively straightforward, for example, using vibrational spectroscopy. Experimental investigation of hydrogen bonding in the excited state is more complicated since the absorbed photon triggers ultrafast dynamical processes. Various time-resolved experiments, based on photoabsorption, fluorescence or photoionization, can then be used to study the excited state dynamics [6]. The dynamical processes occur often on the edge of the time resolution of the respective experiments. In our studies, we have concentrated on experiments in the energy domain. In particular, we detect the kinetic energy distribution of the atomic fragments released from the studied system. The natural choice for the study of hydrogen bonded systems is to detect the kinetic energy of the hydrogen atoms. If the detected hydrogen atom was released from the molecule participating in the hydrogen bonding, we obtain information about the character of this bond both in the ground and in the excited states. Such experiments on the photochemistry of hydrogen bonded systems are a direct continuation of our previous studies of the photodissociation in cluster environments [7]. In our experiment, we observe photoinduced processes in molecular clusters. The cluster approach has many advantages [7, 8]. First, quantities only hardly accessible in the bulk can be monitored (e.g., kinetic energy of the fragments). Second, some control over the size of the particles under study can be imposed. Third, one can study both bulk and surface phenomena within the molecular clusters. Finally, results obtained on molecular clusters can be directly compared with theoretical calculations. In this chapter, we present results of our studies on two types of hydrogen bonded systems: (i) water and aqueous solutions and (ii) hydrogen bonded heterocycles. Both topics are related to the subject of radiation damage of biomolecules. Biological molecules (which are often of heterocyclic character) can be damaged directly by a photon; they can, for example, react in the excited state, fragment or ionize. Biomolecules can be also damaged indirectly: photons can either excite or ionize water, resulting into the production of radicals that then attack the biomolecules. The photochemistry of doped water particles is also interesting from the perspective of atmospheric science. The chapter is organized as follows. First we briefly overview our experimental technique and methods employed. Then we discuss the photochemistry of aqueous systems, that is, first pure water clusters and then water clusters with hydrogen halide molecules. In the second part of our contribution the photochemistry of small heteroaromatic molecules relevant to biological systems is studied. Finally, general conclusions and an outlook are provided.
39.2 Experiment The photodissociation experiments were carried out in a molecular beam apparatus that contains a cluster source, a buffer chamber for manipulating the beams, a laser port for the photoexcitation, and a time-of-flight mass spectrometer for the detection of the dissociated fragments, in this case the H atoms. In addition there is another chamber with a quadrupole mass spectrometer and an electron impact ionizer to mass analyse the cluster beam composition. For details we refer the reader to review articles [7, 9, 10]. The host clusters are produced by isentropic expansions through conical nozzles. By varying pressure and temperature of the source the average size is shifted from n ¼ 2 to more than 1000. The large clusters in this size range usually follow a log–normal distribution that is nowadays directly measured by fragmentation free [11] methods, namely by doping with Na atoms. The resulting average sizes can be correlated with the source parameters based on the ideas of Hagena [12]. The resulting parameters are available for rare gases [13], water and ammonia [11] clusters. The size distributions of small clusters are analysed by a scattering experiment in a
Photodissociation of Molecules in Pure and Doped Water and in Nitrogen Heterocyclic Clusters 867
crossed-molecular beams arrangement [14–16]. The well-defined cluster beam is scattered by a beam of He atoms. In elastic collisions the clusters of different sizes are scattered into different laboratory angles. By measuring the angular and velocity distributions of the scattered clusters the sizes are selected independent from their detection method. To prepare embedded or adsorbed molecules with these clusters two different techniques are applied. The preparation of molecules that are adsorbed on the surface is realized by the so-called pick-up technique introduced by Scoles and coworkers [17]. The cluster is passed through a small scattering cell filled with the molecular vapor with variable pressure. The number of molecules captured depends sensitively on this pressure and follows a Poisson distribution [18]. By a suitable choice of the source pressure, one can easily arrange conditions at which only one molecule is adsorbed on the surface of the cluster. The probability of penetrating inside the cluster depends on the minimum distance and the well depth of the local interaction. From a series of new calculations [19] one can estimate that, for example, HBr on Arn stays near the surface in the first and second shell. To place the molecule inside the clusters a co-expansion with the host gas is used. Since hydrogen bonded molecules form, because of the higher binding energy, much more easily clusters with themselves than with the rare gas atoms, one can either generate the pure hydrogen bonded molecular clusters this way or one can go to more dilute mixtures to generate a small molecular cluster in the rare gas cluster. The biomolecules that are treated in Section 39.4 are produced in this way. By analysis of the measured mass spectra, the laboratory angular and velocity distributions for various fragments, the mean neutral cluster sizes are obtained. The molecules in or on the cluster beam are dissociated by a focussed pulsed laser beam of 243.07 nm and/or 193 nm and a pulse duration between 10 and 20 ns. The ionization takes place with the 243.07 nm laser pulse in a (2 þ 1)-resonant enhanced multiphoton ionization (REMPI) scheme. The ions are extracted into a two-stage time-of-flight mass spectrometer (TOFMS) of the Wiley–McLaren type [20], which is typically used in the low-field mode to detect the protons and measure their TOF distribution. By applying a small electric field we extract those ions already flying in the direction of the detector and turn around those ones that start in the opposite direction. In this way also H atoms with small and even zero velocity are detected. They give rise to a single peak centered between the peaks of the fast fragments. The way to extract from this TOF spectra the kinetic energy distribution is described in detail in Refs [21, 22]. We note that in our experimental arrangement the detection probability is extremely enhanced at small kinetic energies so that we are, in particular, sensitive to the caged atoms [21]. The different contributions that can be derived from the measured kinetic energy of the H atom, Ekin(H), are best discussed by the energy balance of the process. We can illustrate this with an example of HBr molecule dissociation in clusters: hn þ Eint ðHBrÞ ¼ D0 þ Eint ðBrÞ þ Ekin ðBrÞ þ Ekin ðHÞ þ Eclu
ð39:1Þ
where the excitation wavelength hn and the dissociation energy D0 of HBr are known, and Ekin(H) is measured. By conservation of momentum, the kinetic energy of the Br atoms, Ekin(Br), is also known. The excitation of the spin–orbit state Br* in the Br product channel is presented by Eint(Br) and is measured as energy loss in the kinetic energy of the H atom, Ekin(H). These effects would also appear in the dissociation of HBr monomers and indicate the direct cage exit. The influence of the cluster is expressed by the continuous energy loss Eclu of the H atoms caused by the collisions with the cage. This leads, depending on the position, to delayed cage exit or complete caging. The internal excitation of the HBr molecule before the dissociation, Eint(HBr), is observed as energy gain. In general, the molecules will be in the ground state after the expansion. There are, however, various possibilities to observe internally excited molecules. One is
868 Hydrogen Bonding and Transfer in the Excited State
Figure 39.1 Photodissociation of HBr molecules on Arn clusters at two different wavelengths. The spin–orbit states of the Br atom are well resolved in the cage exit contributions
the existence of recombined molecules that are quite hot. The other might occur by the vibrational excitation in collisions within the cluster. A typical example of experimental kinetic energy distributions (KED) is shown in Figure 39.1 for the systems HBr–Ar159 dissociated at two different wavelengths [23]. The distributions exhibit mainly three peaks: one very sharp one at zero energy, which is caused by the completely caged H atoms, and two others that originate from the direct cage exits of H atoms dissociated into the two different spin–orbit states of the Br atom. The higher energy at 193 nm leads to the higher kinetic energy of the H atoms.
39.3 Aqueous Photochemistry from the Cluster Perspective In this section we present the results of our photodissociation studies of pure water clusters [24] and water clusters doped with the hydrogen halide molecules HBr and HCl [25, 26]. 39.3.1 Photoinduced processes in isolated water molecules We start with a description of what is known about the photodissociation of the bare molecule. The photodissociation of the water molecule is a well-studied subject both experimentally and theoretically [27]. A single H2O molecule can undergo two fundamental processes after being irradiated with ultraviolet (UV) ~ state is at photons: photoionization and photodissociation. The maximum of the first absorption band of the A 7.4 eV (168 nm). This state is repulsive with respect to the OH bond, leading to fragmentation into OH and H radicals [28]: ~ ! H þ OH H2 O þ hn ! H2 OðAÞ
ð39:2Þ
Photodissociation on this state has been studied both at 193 nm [29] and 157 nm [30, 31] and found to result in a very low internal excitation of the OH radical product. This process on a single repulsive surface leads to fast and direct dissociation with only a weak excitation of the vibration or rotation of the resulting OH radical.
Photodissociation of Molecules in Pure and Doped Water and in Nitrogen Heterocyclic Clusters 869
~ with the absorption maximum at 9.6 eV At higher energies, water is excited to another dissociative state, B, ~ and B ~ absorption bands are rather broad (full-width at half-maximum, FWHM, 1 eV). (129 nm). Both the A ~ state lead to a far more complex behavior, resulting in a highly rotationally excited OH The dynamics on the B ~ state and the ground X ~ state, which introduces a large fragment. The reason is a conical intersection with the A torque on the OH fragment that is then left in a highly excited rotational state. Most of the population is found in ~ state: the ground X ~ ! H2 OðXÞ ~ ! H þ OH H2 O þ hn ! H2 OðBÞ
ð39:3Þ
~ or The kinetic energy of the resulting H atoms is therefore quite different depending on the excitation to the A ~ state. In the first case they carry nearly all the available energy, while in the latter case the maximum is found B around 1.3 eVafter excitation at 10.2 eV [32, 33]. At higher photon energies, above 12.6 eV, water ionization to H2Oþ starts to compete with the excitation leading to the dissociation. 39.3.2 Photoinduced processes in hydrogen-bonded water: bulk and cluster perspectives The electronic structure and role of hydrogen bonds in irradiated liquid water is also an unresolved issue. The solvation apparently deeply influences both the photodissociation and photoionization [25, 34, 35]. The maximum of the first absorption band of water shifts to higher excitation energies of 8.2 and 8.4 eV for liquid ~ water [36] and ice [37], respectively. The blue shift appears because of the partially Rydberg character of the A state, but also a small red tail has been observed in the absorption spectrum of small water clusters [38]. This suggests a surface preference for the photon absorption at low energies. The effect of solvation on the ionization processes is even more pronounced, leading to dissociation of the protonated water molecule along the hydrogen bond coordinate. In this way the solvated electron and the hydronium cation are formed: 2H2 OðliqÞ þ hn ! H3 O þ þ OH þ e ðaqÞ
ð39:4Þ
The photoelectron spectrum shifts by 1.5 eV (vertical ionization potential) to lower energies upon bulk solvation while the width of the spectrum increases [34]. Thus the ionization and dissociation processes start to overlap below 10 eV. Actually, the two processes might not be distinguished at all in hydrogen bonded systems. The H atom released in the photodissociation can be scavenged by a solvating water molecule to generate the H3O radical. It has been proposed [1, 39, 40] that this species can be interpreted as the solvated electron–hydronium cation pair. Stated differently, the hydronium radical can serve as a cluster model of the solvated electron. The barrier to the H þ H2O limit is only 0.12 eV. However, the H3O radical is stabilized by the solvation with water molecules. This is illustrated in Figure 39.2, which shows the minimum energy path of the H3O (H2O)3 cluster for the H atom detachment. The barrier increased to about 0.3 eV and the released energy is 0.4 eV. We argue here that the hydrated hydronium radical (H3O)aq plays a central role in the hydrogen bonded water systems in general. And on top of that it also plays a central role in the photochemistry of the mixed hydrogen halide–water clusters, as we will demonstrate in Section 39.3.4. We have studied the photodissociation of the pure (H2O)n clusters at 243 nm (5.1 eV) in the range of the the mean sizes between n ¼ 85 and 670. Figure 39.3 shows a typical example of the KED of the ejected H atoms. Evidently, hydrogen atom fragments observed in our experiment come predominantly from two-photon processes. Neither gas-phase nor bulk water absorbs at 5.10 eV. There are, however, neutral states that are 10.2
870 Hydrogen Bonding and Transfer in the Excited State
Å
Figure 39.2 Potential energy function along the minimum energy path for H atom detachment from the hydronium solvated in three water molecules H3O(H2O)3 [39]. Reproduced by permission of the PCCP Owner Societies
Figure 39.3 Photodissociation of water clusters for n ¼ 670. For comparison a spectrum of water clusters for n ¼ 400 doped with HBr molecules is also displayed [24]. Reproduced by permission of the PCCP Owner Societies
above the ground state, to which the system can be promoted by absorption of two photons. The spectrum is composed of a major contribution from very slow fragments with a maximum around 0.05 eV and kinetic energies up to 0.4 eV and a smaller, broader contribution of faster fragments with energies up to 1.5 eV. The spectral shape reflects various processes taking place upon excitation in the clusters. After the hydrogen atom is formed by the photodissociation, it can follow several scenarios: the hydrogen can leave the cluster without being disturbed (cage exit) and the atom can be trapped within the cluster (caging). Finally, the hydrogen atom can also react with the surrounding water molecules to form either a metastable H3O radical or the OH radical and the hydrogen molecule (H2). Let us first discuss the long tail of the spectrum of fast H atoms, effectively ending at 1.5 eV. The overall excess energy available after the two-photon excitation at 243 nm (5.1 eV) is much higher. The ground state OH bond dissociation energy is 5.1 eV [41], therefore the maximum available energy for hydrogen is also ~ state about 5.1 eV. However, it has been shown previously for the bare water molecule that, when exciting the B in this energy range, most of the excess energy is deposited into the rotational energy of the OH fragment during the internal conversion process [32, 33]. The H-fragment KED of the bare molecule has therefore a maximum at about 1.2–1.4 eV. In addition, this part of the spectrum is generated by surface molecules, a position that favors direct cage exit processes. Thus the experimental findings point to the direct dissociation ~ state. processes originating from the B An explanation of the peak at slow energies that dominates the spectrum is not as straightforward as that for the fast peak. An obvious possibility would be trapped or slowed down H atoms induced by the cluster
Photodissociation of Molecules in Pure and Doped Water and in Nitrogen Heterocyclic Clusters 871
environment. However, direct comparison with experiments of hydrogen halide molecules in rare gas clusters reveals characteristic differences. The peak intensity for the caged H-fragments is always observed at zero energy [7, 23, 42] (see also Figure 39.1). This result is also confirmed by accompanying calculations. This suggests that another mechanism is responsible for the slow part of the KED. It is proposed to be the formation of the H3O radical and the subsequent decay into H þ H2O. In a first step the metastable H3O radical is formed after the photodissociation. This is energetically possible for hydrogen atoms with a kinetic energy of at least 0.8 eV, depending on the environment [39, 43, 44] (see also Figure 39.2). The hydronium molecule then decays within an activated process H3O ! H þ H2O, and H atoms with an energy of at most 0.8 eV are produced. These atoms can then lose their energy again by collisions within the clusters and be detected after emerging from it. Thus, generation of the H3O molecule acts as an effective way of energy thermalization, which leads to the prevailing generation of the slow fragments. This interpretation is supported by direct comparison with the experimental result obtained for water clusters doped with hydrogen halide molecules, as is also shown in Figure 39.3, and will be further discussed in Section 39.3.4. There is a striking similarity between the slow component of the present water spectra up to 0.4 eVand the HBr(H2O)n spectrum. For the HX(H2O)n clusters, the H3O arrangement is already present in the system before the photodissociation takes place, due to an acidic dissociation in the ground state of the system to H3Oþ and Cl. The excitation then proceeds via a charge-transfer-to-solvent process, in which the H3O radical is formed directly. Since the acidic dissociation of pure water (i.e., water autoionization) does not occur to a significant extent, the metastable H3O radical can only be formed after the corresponding photodissociation and reaction. The mechanisms that lead to the measured KED are summarized in Figure 39.4. As for the slow peak the ~ state where they dissociate into H water molecules in the cluster are excited by a two-photon process to the B atoms with moderate velocities. They are further slowed down in the cluster cage and are finally trapped by another water molecule, forming H3O. This radical then decays with a maximum energy release of 0.8 eV. For the energetics we refer to Figure 39.2. The faster fragments originate from the direct photolysis of H2O molecules in the cluster, where most of the 5.1 eV of the available energy is transferred into the internal excitation of the system. This process is more probable for the surface molecules as we argue in the next paragraph.
Figure 39.4
Mechanism of H2O photodissociation in clusters
872 Hydrogen Bonding and Transfer in the Excited State
Figure 39.5 Pure water clusters: relative fraction of slow fragments as a function of the mean cluster size. The dashed line indicates the fraction of interior molecules [24]. Reproduced by permission of the PCCP Owner Societies
The size of the cluster affects the processes triggered by the photon absorption. If the excited water molecule is embedded inside of the cluster, there is a high probablity of hydronium radical formation. With decreasing cluster size the relative contribution of the faster component increases in intensity. This trend is illustrated in Figure 39.5, which shows the relative fraction of the slow fragments in dependence of the mean cluster size. The fraction has been obtained as an integral of the corresponding KED between 0 and 0.4 eV relative to the total KED integral from 0 to 1.5 eV. The dashed line in Figure 39.5 indicates the fraction of the molecules inside the cluster volume compared to the total number of molecules in the cluster as a function of cluster size, obtained assuming an icosahedral cluster structure. This comparison suggests that the slow fragments originate from the photodissociation of molecules inside the cluster volume, while the fast fragments originate from the cluster surface. Our experiments were performed for large clusters. Here, we discuss how the photochemical processes converge with respect to clusters size. The surface effects are seen already for the excitation process, that is, for the absorption spectra. We have calculated the absorption cross section of the first band for the water clusters (H2O)n (n ¼ 1–5) using the TDDFT/BHandHLYP/6-31þþg approach. We were able to estimate both the positions of the peaks and their widths. The latter quantity was obtained using the reflection principle method. Figure 39.6 shows the results. The absorption maxima shift to larger energies with increasing size of the clusters. This can be rationalized due to a Coulombic repulsion between a positive charge of the excited water oxygen and a positive charge of a corresponding ground state water hydrogen [38]. The increasing width of the first absorption band is related to a vibrational delocalization induced by the lowering of the OH bond vibrational frequency and also to the fact that more states are involved, caused by the excitation to orbitals localized at different molecules. Figure 39.6 also shows the experimental absorption spectra for an isolated water molecule [45], liquid water [36] and ice [37]. There is satisfactory agreement between our calculated monomer absorption spectrum and the measured one. Minor discrepancies can be assigned mostly to the electronic structure method used. The absorption spectrum then evolves in the direction of ice and liquid water. Clearly, the ice (bulk water) limit is not fully reached within five molecules. However, the maximum has shifted from 170 nm in the gas phase to 150 nm for (H2O)5. The last question to be discussed here is the dependence of the observed processes on the excitation energy. We have also measured the photodissociation in the clusters at 193 nm (6.43 eV). Figure 39.7 shows the KED from the photodissociation of water clusters for n ¼ 430 at this wavelength. The top spectrum (a) corresponds to the two overlapping laser pulses: 193 nm for molecule dissociation and 243 nm for H fragment REMPI ionization. The middle spectrum (b) corresponds to the two-photon dissociation with the 243 nm laser only, and the difference spectrum (c) is then due to the action of the 193 nm photon only. Notably, due to the lower
Photodissociation of Molecules in Pure and Doped Water and in Nitrogen Heterocyclic Clusters 873
Intensity (arb. units)
4 3 2 1 0 140
160
180
Wavelength λ (nm)
~ in comparison with the experimental Figure 39.6 Absorption cross section of H2O clusters in the first band (A) results for the bare molecule and the condensed phase of ice and water. Reprinted with permission from [25]. Copyright 2008 American Chemical Society
Intensity (arb. units)
2
1
0 0.0
0.5 1.0 Kinetic Energy (eV)
1.5
¼ 430 at 193 nm. Top spectrum (a) corresponds to the two Figure 39.7 Photodissociation of water clusters for n overlapping laser pulses: 193 nm for molecule dissociation and 243 nm for H fragment REMPI ionization, middle spectrum (b) corresponds to the two-photon dissociation with 243 nm laser only, and the difference spectrum (c) is then due to the action of the 193 nm photon only [24]. Reproduced by permission of the PCCP Owner Societies
laser intensity and less tight beam focus the 193 nm spectrum corresponds to single-photon processes. Increasing the laser intensity did not change the character and the intensity of the spectra, suggesting that for the two 193 nm photons (12.86 eV) the competing ionization channel prevails, adding no intensity to the Hfragment dissociation channel. Surprisingly, upon excitation at lower energy of 6.4 eV (versus 10.2 eV), the slow H-fragments are absent in the blue spectrum. The 1.2 eV peak is in good agreement with the available energy after a direct dissociation process, namely 6.4–5.1 ¼ 1.3 eV. This peak and the missing slow component ~ state spectrum, we preferentially excite surface water molecules. suggest that, in the extreme red tail of the A 39.3.3 Photoinduced processes of isolated hydrogen halides ~ state represents another example of direct photodissociation process. Hydrogen halides excitation to the A The reaction is very similar to the photodissociation of water described above. The major differences arise for
874 Hydrogen Bonding and Transfer in the Excited State
the heavier hydrogen halides due to the increasing importance of the spin–orbit interaction. This opens new dissociation channels, leading to two spin–orbit states of the halogen atom, which in turn results in Hfragments with two different kinetic energies in the KEDs (see Figure 39.1). The HCl molecule starts absorbing below approximately 190 nm, while the absorption spectra of the heavier hydrogen halides are shifted to the red [46]. These chromophores were used for studies of the rudimentary manifestations of the solvation in the rare gas clusters, the so-called cage effect [47]. Apparently, electronic processes remain unchanged in the rare gas clusters and the observed effects have only mechanical character. Therefore, these systems can serve as a baseline for observations of more elaborate solvation phenomena such as hydrogen bonding in the water clusters. 39.3.4 Photoinduced processes of hydrogen halides in water The photochemistry of hydrogen halides attached to water particles is significantly more complex than the photochemistry of pure water clusters. The reason stems from the fact that the photochemical channels are controlled by the structural features of the clusters in the ground state, which, however, are not straightforward. Hydrogen halides can form hydrogen bonds with water (XH OH2). Hydrogen halides can also acidically dissociate, forming ionic bonds of the X H3Oþ type. In the latter case, we can further distinguish contact ion pairs (the hydrogen halide anion is closely adjacent to the hydronium cation) and solvent separated pairs (the ionic components are separated by a number of water molecules). Which type of structure is present in the systems is controlled by the halogen and by the temperature and structure of the water particles. Despite huge effort both on the experimental and theoretical side, the character of adsorbed hydrogen halides under different conditions remains a somewhat controversial issue. Thus, when we began our experimental study, we did not know in which form the hydrogen halide was present on the cluster. However, as we argue below, the photodissociation experiments also answer this question. First, we briefly review the current status on the structure of hydrogen halides on water/ice surfaces. For small clusters of the type HX(H2O)n a series of theoretical investigations predicted that the formation of stable solvated ions H3Oþ(H2O)n1X first occurs for n ¼ 4 [48–52]. Experiments that explore this transition are still quite rare. In matrix isolation spectroscopy such a transition was claimed, although the assignment of the bands to the correct size is difficult [53]. A pump–probe photoionization experiment with HBr(H2O)n clusters confirmed the transition for n ¼ 5 [54]. On the other hand, vibrational spectroscopy of free HCl(H2O)n clusters could only unambiguously identify covalently bound species up to n ¼ 2 [55, 56]. The situation changes when large particles are considered. The HX dissociation has been theoretically studied on model ice surfaces [57– 62], and it has also been the subject of extensive experimental investigations using temperature controlled desorption, X-ray absorption spectroscopy, reactive Csþ ion scattering, low energy sputtering and Fouriertransform infrared spectroscopy [63–71]. The process of dissociation is strongly temperature dependent and requires collective action of an extended hydrogen bonded network [63]. Despite the partially contradictory results, the following general picture emerges. The amount of intact neutrals decreases with increasing temperature, while that of the dissociated ions increases. The crossover takes place between about 80 and 120 K. Similarly as for water particles, also for the HX–(H2O)n systems the hydronium radical is a central intermediate. In water, this particle is formed in the excited state. This mechanism can again be operational for the doped water clusters. However, the H3O moiety can be also produced by excitation of the acidically dissociated ground state. Two different ways of photochemical production of the hydronium molecule are proposed (Figure 39.8). The first mechanism (1) starts with the acidic dissociation of the HX molecule, forming a zwitterionic structure with X and H3Oþ ions. The resulting species of the H3OþX(H2O)n1 type is then excited by the 193 nm laser pulse. The excited state of this structure is of the charge-transfer-to-solvent
Photodissociation of Molecules in Pure and Doped Water and in Nitrogen Heterocyclic Clusters 875
Figure 39.8 Mechanism of HX photodissociation on H2O clusters. Modified with permission from [26]. Copyright 2007, American Institute of Physics
(CTTS) character. In the S1 state, the system then relaxes into a biradical minimum [40]. The biradical can ultimately decay into X and H radicals and water, which means that dissociation of the H3O is observed [39, 43]. The second possible mechanism (2) of hydronium production starts with exciting an intact HX molecule into its dissociative state. The released high energy hydrogen atom penetrates into the cluster. After it loses part of its energy by inelastic collisions it can form the H3O molecule. This molecule then again decomposes into water and hydrogen. This second mechanism has been proposed above as a way of producing hydronium in water. Below we provide evidence that the two proposed reaction channels can be distinguished based on the different character of the KED spectra. We arrive at the conclusion that the acidic dissociation in the ground state takes place in our clusters followed by excitation into the CTTS state, which than relaxes to the H3O radical and subsequently dissociates to H2O and H. Three arguments are given below to support this claim. 39.3.4.1 Shape and Intensity of the KED Spectrum The first experiments were carried out for HBr(H2O)n clusters in the size range n ¼ 400–500. The mixed clusters are produced by pick-up and photolysed by laser light of 193 nm. The produced H atoms are then detected by photoionization at 243.07 nm. The result of the kinetic energy spectrum is shown in the lower part of Figure 39.3. The spectrum consists of one peak only at low kinetic energies. Compared to the spectrum of pure water clusters, the long tail for larger kinetic energies is completely missing. Actually, if comparable laser beam intensities were used in both experiments, the H-fragment signal from the HBr(H2O)n system is more than a factor of ten higher. In other words, the H-atom signal from pure water clusters represents only an order of magnitude weaker background in these experiments. However, it ought to be mentioned that this signal could be increased by tuning experimental parameters, namely by increasing the 243 nm laser intensity, since it is caused by two-photon processes as discussed above. Thus the signals measured in the experiments with the pure water clusters, discussed in Section 39.3.2, were actually only slightly smaller than the present ones. In comparison with the results obtained for HBrArn clusters of Figure 39.1 both the fast exit contributions and also the completely caged H atom at zero energy are missing. The spectrum resembles the low energy part of KED obtained for pure water clusters (Figure 39.3). This can be rationalized by the fact that in the mixed
876 Hydrogen Bonding and Transfer in the Excited State
clusters the hydrogen atoms come entirely from the H3O radical. This would be consistent with acidic dissociation in the ground state and the subsequent CTTS excitation. Direct HX dissociation can be excluded in this reaction channel. 39.3.4.2 Double Isotope Substitution This argument provides unambiguous experimental evidence for the hypothesis that the detected H-atom originates from the H3O species. We measured, aside from the just shown HBr(H2O)n clusters, also the isotopic variants DBr(H2O)n and HBr(D2O)n. The results for the H atom fragment time-of-flight (TOF) distributions are displayed in the right-hand side of Figure 39.9. The background signal and also signals due to the photodissociation of pure (H2O)n clusters were subtracted from the spectra in the figure. We also note that the (2 þ 1) REMPI process for D atoms occurs at 243.00 nm [72], well outside the 0.04 nm bandwidth of the ionizing laser. Besides, the heavier D fragments would arrive at the detector approximately 2 ms later than the lighter H fragments, that is, in a TOF-spectra region where we do not observe any signal. Thus only H atoms are detected. The TOF spectra for the different isotope variations are very similar in shape. Their one peak structure leads to the type of KED displayed in Figure 39.3. The intensity ratio is 3.0 : 2.1 : 1.0. Since we are only able to detect H atoms, the compelling explanation is that we detect in the three experiments H atoms from H3O, H2DO and HD2O species. These processes are pictorially demonstrated on the left-hand side of Figure 39.9. The origin of the signals from pure water clusters and also from HBr can be ruled out. Indeed, the pure (H2O)n produce signals more than an order of magnitude lower and with a slightly different shape. The signal also cannot be simply due to the direct photolysis of the HX molecule on the cluster, since the pick-up of DBr molecules on (H2O)n also produces an intense Hfragment signal. At the same time, H atoms are also produced by the photodissociation of HBr on (D2O)n clusters. This indicates an exchange of hydrogen atoms between the hydrogen halide and the water cluster.
Figure 39.9 Photodissociation of HBr molecules on water clusters. Deuteration experiments pointing to the generation and dissociation of the H3O radical (See Plate 48)
Photodissociation of Molecules in Pure and Doped Water and in Nitrogen Heterocyclic Clusters 877
Figure 39.10
Photodissociation of HX molecules on water clusters at 243 nm
The same result is obtained when we replace the HBr molecule with HCl. The results for the KEDs are shown in Figure 39.10. We observe the same shape of the distribution within the experimental error bars, but the intensities differ by about a factor of seven. This intensity ratio is a fingerprint of the acidic dissociation, as will be discussed below, but more importantly at this point the similarity of the KED shape for both HBr(H2O)n and HCl(H2O)n clusters suggests that the H-fragments originate from the same species in both systems. This characteristic shape results, as was already discussed for the pure water clusters, from the decay of the hydronium radical according to H3O ! H þ H2O. However, we note that in the present case the generation of H3O is completely different from the result for pure water clusters. Contact Ion Pair or Solvant Separated Structure The estimated internal temperature of 100–130 K of our clusters [73] is beyond the onset of the acid dissociation proposed by the various studies mentioned above. The missing fast H fragments in our measurements also suggest the acid dissociation in the ground state of our clusters. Therefore, the excitation (1) in Figure 39.8 of some form of the zwitterionic species H3Oþ–Cl is the more likely route of H3O generation. The finally considered zwitterionic structure can be either of a contact ion type (ClH3Oþ(H2O)n1) or the ions can be separated by solvent water molecules (Cl (H2O)n1 H3Oþ). The latter option would imply a higher proton mobility leading to a rapid isotopic dilution in the experiment with D2O. The ratio of the H-fragment signal from HBr(D2O)n to HBr(H2O)n clusters with n ¼ 500 would be expected to be 1/1000 if the dilution would take place rather than the measured 1/3. Since the isotopic dilution is not observed, the formation of a contact ion pair is suggested. Proton mobility on ice surfaces has been studied by Park et al. [74]. After adding the HCl, the H/D exchange for the surface layer became almost complete in 10 min. Our experiment provides complementary evidence that the ions do not separate during the 0.65 ms between pick-up and photolysis. 39.3.4.3 CTTS and Direct Excitation Cross Sections Which mechanism is indeed operating was confirmed by comparison of the measured signal intensity ratio from HBr(H2O)n and HCl(H2O)n clusters with calculations on the structure and the absorption cross section of these clusters. We will discuss these options in detail. As discussed earlier, there are three major structural motifs for HX on water clusters. The first structural type consists of the hydrogen halide remaining in a
878 Hydrogen Bonding and Transfer in the Excited State
Figure 39.11 Structures of small HX–(H2O)n clusters, X ¼ Cl,Br, and n ¼ 1–5. Different structural motifs appear for n 4: (a) intact, (b) contact ion pair and (c) solvent separated ion pair. Reprinted with permission from [25]. Copyright 2008 American Chemical Society
covalent state HX(H2O)n, further referred to as “intact.” Upon the acidic dissociation, the oxonium cation and halogenide anion can generate either the contact ion pair or the solvent separated structure. Figure 39.11 shows the optimized structures for HX(H2O)n, X ¼ Cl,Br and n ¼ 4,5. Optimization was performed at the MP2/6-31þþg level, at which the structures seem to be converged from both energetical and structural perspectives. For clusters with n ¼ 1–3, the only local minima found are the intact structures. For clusters with n ¼ 4,5, all three types of structural motifs are present: intact structure (a), contact ions (b) and solvent separated ions (c). From the relative energies at several levels of theory, it is found that HCl and HBr clusters are predicted to support dissociated states starting from n ¼ 4 onwards. The general picture of the HX dissociation process is in agreement with previously reported calculations [50, 75, 76]. Based on the results of the electronic excitation we have calculated the electronic absorption spectra for HX(H2O)4 clusters with X ¼ Br,Cl in both the covalent state and the ionic states. Figure 39.12 shows the results for HCl(H2O)4 and HBr(H2O)4 clusters. The absorption spectra were calculated with the TDDFT/BHandHLYP/6-31þþg method employing the reflection principle. The absorption is significantly redshifted upon the acidic dissociation. The position of the maximum shifts from 153 nm to approximately 180 nm for HCl (H2O)4 and to 210 nm for HBr(H2O)4. Importantly, this shift leads to a huge enhancement of the photoabsorption cross section at 193 nm for HCl. The photoabsorption cross section at 193 nm increases also for HBr but to a lesser extent than in the case of HCl. Chloride and bromide anions are directly involved in the photoexcitation process. For the intact structure, the first excited state is mostly connected with excitations within the HX molecule, after the acidic dissociation the excitation adopts the CTTS character; that is, an electron is promoted from the X moiety to orbitals of the water solvent. This CTTS band falls into the spectral range of our photodissociation experiment and can thus be probed. The major findings of our calculations can be summarized as follows: (i) The addition of HCl, HBr or of another water molecule to the water cluster leads to a slight blue-shift in the electronic absorption spectrum. (ii) The acidic dissociation of HCl and HBr results in a significant redshift in the absorption spectrum; in fact, a new CTTS band appears. Thus the CTTS band provides an important spectroscopically observable clue of the acidic dissociation. There are now two major interconnected questions to be discussed: (1) Are the calculations
Photodissociation of Molecules in Pure and Doped Water and in Nitrogen Heterocyclic Clusters 879 6
Intensity (arb. units)
4 2
0
160
180
200 220 240
260
160
180
200 220 240
260
5 4 3 2 1 0
Wavelength l (nm)
Figure 39.12 Absorption cross sections of HX(H2O)4 clusters for intact, contact ion pairs and solvent separated structures. The geometries of the clusters are shown in Figure 39.11. The long vertical arrows mark the excitation wavelength. Reprinted with permission from [25]. Copyright 2008 American Chemical Society
supported by our recent photodissociation experiments? (2) Are HCl and HBr acidically dissociated under the experimental conditions? The quantity for which the theoretical prediction can be directly compared with the photodissociation experiment is the ratio of hydrogen atom yields produced from HCl and HBr in different environments. These ratios are compared with the values calculated for the model HX(H2O)4 clusters in Table 39.1. The measurement for the bare molecule, which gives 23 5, agrees within experimental errors with the known absorption cross sections [77, 78]. The ratio does not change very much when the halogen halides are complexed with argon, leading to 20 5. The same ratio, however, measured for hydrogen halides deposited on the water clusters gives 7.4 1. The calculations have to be carried out for: jðHBrÞ ½s193 ðHBrÞ þ s243 ðHBrÞ ¼ jðHClÞ ½s193 ðHClÞ þ s243 ðHClÞ
ð39:5Þ
since, in the experiment, photons of the two wavelengths are actually present. The result for water complexes yields a theoretically predicted ratio of the photolysis of about 2.7 for the contact ion and 1.8 for the solvent separated ion structure (Table 39.1). A theoretical estimate of this ratio for the intact structure (as well as for bare HX from our calculations) is difficult because even at the 193 nm we sample only the very red tail of the HCl spectrum. Nevertheless, by extrapolation one can conclude that the ratio exceeds 30 for the bare molecule and 45 upon the complexation with water. Therefore, the comparison strongly suggests that acidic dissociation takes place. The measured ratio still somewhat exceeds the theoretical prediction of the j(HBr)=j(HCl) ratio. The most likely explanation is that the precise quantitative agreement is not achieved due to the comparison of theoretical calculations performed for small water clusters with the experiment conducted on much
880 Hydrogen Bonding and Transfer in the Excited State Table 39.1 Comparison of calculated and experimental ratios of the photolysis rate of HCl to HBr, j(HBr)=j(HCl), in different environments and dissociation states Environment Bare molecule HX HX–Ar100 HX–(H2O)400
Measured 23 5 20 5 7.4 1.0
Motif
Calculated
Bare molecule
30
HX(H2O)4 dissociated HX(H2O)4 intact
2.7 (1.8)a 45
a
Number in parenthesis refers to the solvent separated ions.
larger clusters. In addition, the TDDFT calculations may introduce some errors. The conclusion that the j(HBr)=j(HCl) ratio drops upon an acidic dissociation is, however, quite robust. Summarizing, we state that the photodissociation mainly follows the mechanism presented on the righthand side of Figure 39.8 [indicated by (1)]. In the first step the intact hydrogen halide molecule undergoes acidic dissociation to the zwitterionic state. Then this state is excited by the 193 nm laser pulse to the CTTS state, which relaxes to the biradical state with H3O and Br=Cl. The H3O then decays into H2O and H, which is detected by photoionization. The interaction between the HBr and the HCl molecules with the water molecules of the cluster is of a very local nature within the experimental time scale (200 nm penetrating into the stratosphere than the HCl molecule [80]. We have shown that the acidic dissociation itself is sufficient to significantly enhance Cl radical production from HCl on ice nanoparticles. The acidic dissociation shifts the electronic absorption spectrum to the red, so that it can be expected that the photolysis rate will be much enhanced when the hydrogen halide lands on a water cluster and acidically dissociates. To quantify this issue, we have calculated the photolysis rate using an actinic flux measured at an altitude of 50 km. The photolysis rate increased by four orders of magnitude upon the HCl uptake and dissociation on (H2O)4 cluster; the enhancement for HBr was two orders of magnitude. Notably, our dynamical calculations show that the Cl radical is released directly from the ice nanoparticle. Ice nanoparticles thus might catalyse HCl photolysis in a way that can potentially play a non-negligible role in the stratospheric chlorine budget.
39.4 Hydrogen Bonded Clusters of Nitrogen Heterocycles To investigate the influence of different hydrogen bonding patterns on the photostability of molecules, the photodissociation of pyrrole (Py), imidazole (Im) and pyrazole (Pz) molecules in clusters has been studied. All three molecules have a very similar five-membered heteroaromatic ring structure (Figure 39.13), which can be found in many biological compounds. Pyrrole is one of the simplest molecules with biological relevance: the
Photodissociation of Molecules in Pure and Doped Water and in Nitrogen Heterocyclic Clusters 881
Figure 39.13
Structures of the studied molecules
pyrrole structure is present in, for example, hemes and chlorophylls. The imidazole structure can be found in purine, histidine and so on. Purine is a molecular skeletal building block of the nucleic acid bases adenine and guanine. Histidine is a naturally occurring amino acid and imidazole can be found in its side chain. Imidazole is, for example, also an important ligand towards transition metal ions in vitamin B12. Pyrazole is an isomer of imidazole, which is rare in nature; however, due to the different position of the nitrogen atom in the heteroaromatic ring, it generates clusters with hydrogen bond motifs different from those of the imidazole clusters. Therefore, pyrazole clusters have been investigated for the purpose of comparison. The different hydrogen bond motifs for the dimers are schematically represented in Figure 39.14. These different structural motifs can lead to different pathways in photodissociation. In the pyrrole clusters the NH bond of one molecule binds to the p-electron cloud of the neighboring molecule, NH p. This type of interaction represents the simplest type of solvation, in which the chromophore does not react with the solvent. If, however, the NH bond is involved in hydrogen bonding, the hydrogen atom of the donor molecule can migrate to the acceptor molecule and new phenomena can occur, for example, hydrogen or proton transfer in the excited state. This is the case for pyrrole ammonia [81], pyrrole water [82], phenol ammonia [83, 84] or pyrrole pyridine complexes [85]. In the imidazole clusters the “normal” NH N hydrogen bond is observed, and in the pyrazole dimer the NH N double-bond is present. The latter structure closely resembles the bonding pattern between DNA base pairs. Therefore it is interesting to investigate how these different bonding motifs can influence the photochemistry following a UV excitation and the stability of the molecules in relation to the question of stability and radiation damage of larger biomolecules. The UV photochemistry of isolated nitrogen heterocycles in the gas phase has been studied extensively both experimentally [86–89] and theoretically [90–93]. Photochemical pathways in these compounds are controlled by an interplay between the dissociation channel on the ps* state and ground state recovery via different mechanisms. Especially, the role of the ps* states in the photodissociation of these and similar molecules (phenol) has recently attracted much attention [86]. The potential energy surfaces (PES) for all three molecules Py, Im, and Pz possess qualitatively very similar features, involving low lying excited ps* and pp* states and conical intersections between them and between the ground S0 state. Figure 39.15 shows, schematically, cuts
Figure 39.14
Schematic picture of the different hydrogen bonding motifs represented by the Py, Im and Pz dimers
882 Hydrogen Bonding and Transfer in the Excited State
Figure 39.15 Schematic picture of general PES typical for the studied species based on ab initio calculations for Py molecule. The cuts through PES along the NH stretching and ring-deformation coordinates are shown. The arrows that follow the cuts indicate the possible dissociation pathways. Circles indicate the CI ps* =S0 , pp* =ps* and pp* =S0
along the NH stretching coordinate and the ring deformation coordinate through a general PES for such systems (based on the real calculated PES for the pyrrole molecule [94]). The general picture of the photochemistry, which emerged from the numerous gas phase studies, is schematically represented in Figure 39.16. Upon low energy excitation, the photodynamics is dominated by the ps* state. This state is (asymptotically) dissociative and, as a result, the hydrogen atom is released (hydrogen dissociation, HD, channel). At elongated NH distances, the ps* =S0 intersection occurs. It is therefore possible that frustrated dissociation (FD) takes place, and the molecular ground state is recreated. At higher photon energies, the pp* state is populated. A molecular ring distortion (MRD) reaction channel is thus opened. Subsequently, the molecule quenches into the vibrationally hot ground state where again the hydrogen atom can dissociate or other molecular fragments can be formed. As pointed out above, the role of the ps* states in biomolecules has lately been much discussed for the gas phase [39]. The functionality of biomolecules is, however, controlled by their environment, either by a solvent or by specific interactions with other molecular units (such as hydrogen bonding in DNA base pairs).
Figure 39.16 Schematic picture of photodissociation processes possible in the molecules Py, Im and Pz. Excitation of the ps* state can lead to direct hydrogen dissociation (HD) and fast H-atom production. Alternatively, frustrated dissociation (FD) can occur via a ps* =S0 conical intersection at elongated NH distances. At higher photon energies, the pp* state is populated and molecular ring distortion (MRD) can occur, quenching the molecule into the vibrationally hot ground state, where it can again dissociate, yielding the slow H-fragments or other products. Modified with permission from [102]. Copyright 2009 American Chemical Society
Photodissociation of Molecules in Pure and Doped Water and in Nitrogen Heterocyclic Clusters 883
Therefore, the photochemistry in solvated systems can be quite different. This section summarizes the various effects the solvation and hydrogen bonding in clusters can have on the photodissociation dynamics and stability of these molecules. 39.4.1 Experimental results We have investigated the photodissociation of Py, Im and Pz clusters at two different wavelengths, 243 and 193 nm, and for various mean cluster sizes. Figure 39.17 shows examples of the measured H-fragment KEDs from Py [left-hand panels (a) and (b)] and Im [right-hand panels (c) and (d)] clusters at 243 nm. The top panels (a) and (c) correspond to the clusters generated in expansions with He as a buffer gas, resulting in mean cluster sizes of n 3 for both Py and Im clusters. The bottom panels (b) and (d) correspond to the larger clusters generated in expansions with Ar as a buffer gas. Our mass spectrometric and scattering experiments have 8 for pyrrole [95], while pure Imn shown that these result in mixed PynArm clusters with n 4 and m clusters with n 6 were generated for imidazole. No spectra for pyrazole are presented in Figure 39.17 since we did not observe any measurable photodissociation signal at 243 nm for pyrazole at any exploited expansion conditions. It should be mentioned that the KEDs in Figure 39.17 are normalized and the corresponding measured TOF spectra intensity exhibits a strong dependence on the mean cluster size, that is, the signal decreases significantly with the cluster size, as suggested by the increasing error bars on the data in Figure 39.17. All the spectra in Figure 39.17 exhibit a bimodal character with a narrower peak of faster fragments at approximately 0.8 eV, and a slower component with a broad distribution peaking below 0.4 eV. The spectra have been analysed and deconvoluted to the contributions of the two components (red and blue lines in Figure 39.17). Similar spectra have also been measured for the bare pyrrole [96–101] and imidazole [87] molecules. The relative ratio of the two contributions changes with the cluster size. Already from Figure 39.17 it is obvious that the fast component decreases in intensity relative to the slow one with increasing mean cluster size. Figure 39.18 shows this trend quantitatively: here the ratio of fast to slow (F/S) component obtained by integrating the corresponding peaks in the measured KEDs for Py and Im clusters at 243 nm is plotted as a function of the mean cluster size.
Figure 39.17 Measured KEDs of Py (a, b) and Im (c, d) clusters at 243 nm. The top panels correspond to small 3, while the bottom spectra correspond to mixed PynArm clusters with n 4 and m 8 clusters of mean size n 6 (d). The spectra are analysed for a slow and fast component (See Plate 49) (b) and pure Imn clusters with n
884 Hydrogen Bonding and Transfer in the Excited State
Figure 39.18 Fast-to-slow fragment ratios evaluated from the measured KEDs for Py and Im clusters at 243 nm plotted as a function of the cluster mean size. The F/S ratio obtained from Figure 39.17(b) for Py is plotted at n ¼ 12, which corresponds to the total mean size of the mixed PynArm cluster (See Plate 50)
Figure 39.19 shows the KEDs for Py, Im and Pz clusters of the mean cluster size n 3 measured at 193 nm. The spectra for Py and Im closely resemble the spectra measured for bare molecules in other experiments [87, 96–101]. Again, the spectra could be deconvoluted to the contributions of the fast and slow components. Also the dependence on the mean cluster size has been investigated at the various expansion conditions; however, since the relative contribution of the fast component is very small at 193 nm both for the bare molecule (see the references cited above) and for clusters (Figure 39.19) a reliable quantitative F/S ratio dependence is difficult to obtain. Nevertheless, the evaluated F/S ratios for Py, Im and Pz clusters at 193 nm are shown in Figure 39.20. Notably, here at 193 nm, unlike at 243 nm, a good H-fragment signal has been measured for Pz clusters, comparable to the signals from Py and Im clusters at similar conditions. As mentioned above, the comparable error bars on the three normalized spectra in Figure 39.19 suggest comparable signal intensities.
3 at 193 nm. The spectra are analysed Figure 39.19 Measured KEDs of Py (a), Im (b) and Pz (c) clusters with n for a slow and fast component (See Plate 51)
Fast/Slow H-fragment ratio
Photodissociation of Molecules in Pure and Doped Water and in Nitrogen Heterocyclic Clusters 885
0.3
Pz Py
0.2
Im 0.1 0.0
0
5 Mean cluster size n
10
Figure 39.20 Fast-to-slow fragment ratios evaluated from the measured KEDs for Py, Im and Pz clusters at 193 nm (See Plate 52)
39.4.2 Photochemistry of isolated and solvated heterocycles The two peaks in the KED spectra can be attributed to two distinct processes: (i) The slow peak corresponds to the hydrogen dissociation from the vibrationally excited hot ground state. (ii) The fast peak results from direct dissociation on the ps* state. The shape of the first peak can be modeled using statistical theory as was outlined in Ref. [102]. Note, as a technical point, that the photodissociation resulting in the slow component in our experiment corresponds to processes in which two photons were subsequently absorbed by the cluster within a single laser pulse, resulting in a statistical decay of the system with an energy content corresponding to almost twice the photon energy. Notably, the statistical decay of the imidazole molecule after single photon absorption resulted in a substantially narrower energy distribution of the H-fragment peaking at somewhat lower energies [87, 102]. The remaining part of the signal can be then attributed to the direct dissociation channel. The ratio between these two peaks therefore depends on the relative importance of the different dissociation channels. What we are trying to establish here is how the photodissociation dynamics of the molecule outlined in Figure 39.16 changes in clusters. Alternatively, one can approach this problem as, how the potential energy surfaces in Figure 39.15 change upon solvation. Phrased this way, the theoretical calculations can provide a significant insight. We start with the pyrrole clusters bound with the NH p bond. The influence of a solvent on the photochemistry of the pyrrole molecule is shown schematically in Figure 39.21. We have extended the photochemical mechanism presented by Barbatti et al. [103] in a theoretical study to situations where the pyrrole molecule is solvated. The upper part of Figure 39.21 represents the cuts through the PES of pyrrole molecule along the NH stretch (right) and ring-deformation (left) coordinates. This includes the calculation of the ps* =S0 conical intersection along the NH stretch coordinate and the pp* =S0 and ps* =pp* conical intersections of the ring deformation. The influence of the solvent was simulated by the presence of an Ar atom at a fixed distance in the plane of the carbon atoms, which resembles the NH p binding motif of the pyrrole dimer. The presence of the Ar atom has a strong influence on the conical intersection in the NH stretch coordinate (see the lower part of Figure 39.21). It is expected that the crossing between the ps* and the ground state will increase in energy due to the repulsive Pauli interaction with the argon atom. The effect is, however, even stronger – the conical intersection will cease to exist. The reason is as follows. While the S0 and the excited valence states are virtually unchanged, the highly delocalized ps* states, which are of a Rydberg character, are shifted by 0.5 eV in the Franck–Condon region and the interaction between the Rydberg state and argon becomes stronger with increasing NH distance. As a result the ps* =S0 conical intersection is no longer present and the channel for the fast H atoms is closed. The changes in the deformation coordinate upon solvation with Ar are, on the other hand, almost negligible. The pp* =S0 intersection remains open and the population of the slow peak is not hindered. To account for
886 Hydrogen Bonding and Transfer in the Excited State
Figure 39.21 Schematic picture showing the influence of a solvent on the PES of pyrrole molecule based on ab initio calculations. The upper panel shows the bare Py molecule, while the lower panel illustrates the effects of solvation in the NH bond stretching and ring deformation coordinates. Reprinted with permission from [94]. Copyright 2007, American Institute of Physics
solvation effects other than the Pauli repulsion, we have also performed the calculations for the pyrrole dimer. However, the above conclusions hold true even when the argon atom is replaced by another pyrrole molecule. Note that a very similar effect has also been observed at the same time in the group of Kitsopoulos [100, 101] by solvating the pyrrole molecule with a single Xe atom. The results of our photodissociation experiments can be understood in terms of this mechanism. We observe many more H atoms in the slow peak in clusters with respect to the molecule, because the route to the direct dissociation into the fast peak is closed. Furthermore, even when the molecule reaches the ps* =S0 intersection, the solvent can prevent direct dissociation and the system quenches to the bound ground state [104]. In addition, the out-of-plane mode is also operating in clusters and plays an important role, possibly even at 243 nm. The increase of the slow component relative intensity for 193 nm with respect to 243 nm is traced back to the direct excitation of the pp* state. The pp* =S0 conical intersection is in this case diabatically directly accessible and this does not change with complexation. An important message that can be drawn from these results is that conclusions about the photochemical behavior in a confined environment (e.g., in biological systems) that are solely based on gas-phase data may be misleading. In imidazole clusters a somewhat different picture arises from the calculations. Owing to the NH N bonding, the hydrogen transfer (HT) between the imidazole units can play a role in the excited imidazole cluster. This is illustrated by our calculations for the imidazole dimer in Figure 39.22, which shows the energy profile along the interpolation coordinate between the optimized ground state structure (Franck–Condon point) (right) and the hydrogen transferred structure (left) going through a transition state (center), where the hydrogen is shared in the middle between the two imidazole units. The calculations performed at CAS-PT2 level of theory based on the CAS-SCF wavefunction with six electrons in six active orbitals are shown. The bottom curve always corresponds to the ground state S0 while the lower excited curve corresponds to the ps* state and the upper one to the pp* state, and the conical intersections are indicated. The HT mechanism is opened upon the excitation to the pp* state (at 193 nm in our experiment). Barriereless transition of the H-atom towards the hydrogen bond acceptor unit occurs on the PES until the system reaches the conical intersection pp* =S0 on the left-hand side boundary of Figure 39.22. There the electronic population can be funnelled to the ground state S0, where the H-atom moves back to the former hydrogen bond donor molecule. In this process the excitation energy dissipates between the two participating imidazole molecules and the imidazole dissociation is then much less probable. Thus the HT process in the cluster has a stabilizing
Photodissociation of Molecules in Pure and Doped Water and in Nitrogen Heterocyclic Clusters 887
Figure 39.22 Schematic picture showing the effect of the hydrogen transfer on the photochemistry of imidazole dimer. The CAS-PT2 method based on the CAS-SCF wavefunction with six electrons in six active orbitals was used. The lowest curve corresponds to the ground state S0 while the lower excited curve corresponds to the ps* state and the upper one to the pp* state. The conical intersections are indicated
effect on the imidazole molecule. At the longer wavelength of 243 nm, the HT is still a barriereless process on the ps* state PES, yet the driving force for it is negligible (the PES is essentially flat). Therefore, the excitation energy stays localized on the one imidazole molecule within one complex, which may lead to its fragmentation. It ought to be mentioned that in the dimer all the processes observed for the bare imidazole molecule are still possible with the free NH bond of the hydrogen bond acceptor unit. However, larger species, starting from trimer, generate cyclic structures without the free NH bonds available. In our experiments, we probed the cluster size distributions with n 3 and found some contributions of dimers and monomers. Therefore, the contribution from processes analogical to the processes in the bare Im molecule can still be non-negligible for n 3. However, the contribution from these processes decreases with increasing mean cluster size, that is, for n 6. Figure 39.23 summarizes suggested photochemical pathways for imidazole clusters (compare to Figure 39.16 summarizing the processes possible in the molecule). A similar effect can be expected in pyrazole clusters. It is now interesting to compare the photodissociation mechanism in all the studied clusters: At 243 nm the measured F/S ratio decreases with the complexation for both Py and Im clusters (Figure 39.18). We have shown that the change in PES due to the solvation, which closes the dissociation channel leading to the fast fragments, is responsible for this dependence in pyrrole clusters. In the imidazole clusters the EFD process is suggested to cause the F/S ratio decrease. No photodissociation was observed for pyrazole clusters at this wavelength, presumably due to their larger NH bond stability (the dissociation energies are 4.07, 4.12 and 4.6 eV for Py, Im and Pz, respectively). At the shorter wavelength of 193 nm the fast fragments are much less populated, therefore the F/S ratio trends are difficult to asses. Yet, it seems that this ratio is independent of the cluster size for pyrrole clusters, while it decreases for imidazole and pyrazole (Figure 39.20). This would suggest that a different mechanism operates in Py clusters than in the Im and Pz clusters. In Py the dissociation at 193 nm proceeds mainly along the ring-deformation coordinate, which remains unchanged by the solvation, and therefore the F/S ratio remains independent of cluster size. On the other hand, in imidazole clusters the dissociation at 193 nm is driven to the HT channel, which funnels the population to the ground state and thus decreases the probability for generating the fast fragments. This is indeed not possible for the bare molecule, which leads to the F/S ratio decrease with complexation. A similar situation can be expected in pyrazole
888 Hydrogen Bonding and Transfer in the Excited State
Figure 39.23 Schematic picture of photodissociation processes possible in imidazole clusters (dimer): At 243 nm the ps* state of an imidazole molecule is excited, leading to the direct free hydrogen dissociation. However, the frustrated dissociation may be enhanced by the cluster environment (EFD). At higher photon energies, the pp* molecular state is populated. In addition to the molecular ring distortion (MRD) channel present for the molecule the hydrogen transfer (HT) channel can also occur in the clusters. Modified with permission from [102]. Copyright 2009 American Chemical Society
clusters, which are also bound by the NH N bonds, allowing for the HT process to occur. Thus this is tentative evidence for operation of the HT process in the clusters bound by the NH N bonds (Im, Pz) as opposed to the NH p bound Py clusters.
39.5 General Conclusions and Outlook Generally, by the above examples of aqueous systems and clusters of heteroaromatic ring molecules we have illustrated that the various hydrogen bonded systems provide a large playground for various processes to occur in the excited states: opening new reaction channels such as the generation of the H3O radical or the solvated electron, on the one hand, and closing some dissociation channels upon solvation, on the other hand, either by electronic interaction with the solvent or by hydrogen transfer reactions and subsequent energy dissipation. Now, as the next step, we would like to merge the two areas of our research presented here and investigate the biologically relevant heteroaromatic molecules in the aqueous clusters. In this way, for example, the role of the solvated electron from the solvent environment on the photochemistry of the heteroaromatic molecules can be investigated. The ultimate goal of such research is to shed some light onto the molecular level mechanism of such important and complicated processes as DNA radiation damage. More specifically, we have provided some evidence that the H3O radical is a central species in the photochemistry of the aqueous systems. Two mechanisms of H3O generation were suggested: In the multiphoton excitation of the pure (H2O)n clusters the H3O is generated in the excited state by a reaction of hydrogen atom fragments with water molecules. The H3O radical is stabilized by solvating water molecules and ultimately it decays to the observed H-atom. A completely different mechanism of H3O radical production was found in the HX(H2O)n clusters. Here, the precursor ion H3Oþ is generated by the acidic dissociation already in the ground state, which is then excited to generate the H3O radical via a charge-transfer-to-solvent (CTTS) process. This mechanism seems to be a general pattern in the systems, where acidic dissociation can occur. Thus a logical next step of our research after HBr and HCl molecules on water clusters is to prove experimentally that the mechanism also operates for the analogical system with HI molecules. A very
Photodissociation of Molecules in Pure and Doped Water and in Nitrogen Heterocyclic Clusters 889
interesting case is the photochemistry of the much less acidic HF molecule. As a weak acid, hydrogen fluoride should not undergo the acidic dissociation and the CTTS mechanism should therefore play no role. Recent experiments have, however, suggested [105, 106] that the actual structure of these species can be best described as a proton shared structure. From a more practical point of view, and of relevance to the atmospheric chemistry, an extension towards nitrogen- and chlorine-containing molecules (HNO3, NOx, OClO) in aqueous clusters is of great interest. For the heteroaromatic molecules we have demonstrated how profound an effect the solvation can have on the photochemistry of these species. The major message from these studies is that one should cautiously examine the relevance of gas-phase data when making conclusions about the photochemistry of these molecules solvated in biological systems. In particular, we have demonstrated that the dissociation channel can be closed upon solvation by electronic interaction in pyrrole clusters. On the other hand, in imidazole and possibly also in pyrazole clusters the closing of the fast dissociation channel is suggested to be due to the hydrogen transfer process and subsequent energy dissipation in the system. Further studies of the present molecules in water clusters, which is the natural solvent in biological systems, are planned, as well as the extension of our experiments towards larger biomolecules.
Acknowledgements Support by the special program “Nanotechnology for society” of the Czech Academy of Sciences via grant Nr. KAN400400651, grants Nr. 203/09/0422 and 203/07/P449 of the Grant Agency of the Czech Republic are acknowledged. M. F. acknowledges a special J. E. Purkyneˇ fellowship of the Czech Academy of Sciences.
References 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22.
A. L. Sobolewski and W. G. Domcke, J. Phys. Chem. A, 111, 11725 (2007). T. Schultz, E. Samoylova, W. Radloff et al. Science, 306, 1765 (2004). P. Markwick and N. Doltsinis, J. Chem. Phys., 126, 175102 (2007). N. Schwalb and F. Temps, J. Am. Chem. Soc., 129, 9272 (2007). D. Shemesh, A. L. Sobolewski and W. Domcke, J. Am. Chem. Soc., 131, 1374 (2009). T. Dermota, Q. Zhong and A. Castleman, Chem. Rev., 104, 1861 (2004). U. Buck, J. Phys. Chem. A, 106, 10049 (2002). P. Jena and A. W. Castleman, Proc. Nat. Acad. Sci. USA, 103, 10560 (2006). U. Buck and M. Farnık, Int. Rev. Phys. Chem., 25, 583 (2006). M. Farnık and U. Buck, Phys. Scrip.: Commun. At. Mol. Opt. Phys., 76, 73 (2007). S. Sch€utte and U. Buck, Int. J. Mass Spectrom., 220, 183 (2002). O. F. Hagena, Surf. Sci., 106, 101 (1981). U. Buck and R. Krohne, J. Chem. Phys., 105, 5408 (1996). U. Buck and H. Meyer, Phys. Rev. Lett., 52, 109 (1986). U. Buck and H. Meyer, J. Chem. Phys., 84, 4854 (1986). U. Buck, C. Lauenstein, H. Meyer and R. Sroka, J. Phys. Chem., 92, 1916 (1988). T. E. Gough, M. Mengel, P. A. Rowntree and G. Scoles, J. Chem. Phys., 83, 4958 (1985). M. Behrens, R. Fr€ochtenicht, M. Hartmann et al., J. Chem. Phys., 111, 2436 (1999). N. H. Nahler, R. Baumfalk, U. Buck et al., Phys. Chem. Chem. Phys., 5, 3394 (1999). W. C. Wiley and I. H. McLaren, Rev. Sci. Instrum., 26, 1150 (1955). R. Baumfalk, N. H. Nahler, U. Buck et al., J. Chem. Phys., 113, 329 (2000). R. Baumfalk, U. Buck, C. Frischkorn et al., J. Chem. Phys., 111, 2595 (1999).
890 Hydrogen Bonding and Transfer in the Excited State 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 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.
M. Farnık, N. H. Nahler, U. Buck et al., Chem. Phys., 315, 161 (2005). V. Poterya, M. Farnık, M. Oncak and P. Slavıcek, Phys. Chem. Chem. Phys., 10, 4835 (2008). M. Oncak, P. Slavıcek, V. Poterya et al., J. Phys. Chem. A, 112, 5344 (2008). V. Poterya, M. Farnık, P. Slavıcek et al., J. Chem. Phys., 126, 071101 (2007). R. Schinke, Photodissociation Dynamics, Cambridge University Press, Cambridge (1993). R. Schinke, V. Engel and V. Staemmler, Chem. Phys. Lett., 116, 165 (1985). D. F. Plusquellic, O. Votava and D. J. Nesbitt, J. Chem. Phys., 109, 6631 (1998). P. Andresen, G. S. Ondrey and B. Titze, Phys. Rev. Lett., 50, 486 (1983). X. F. Yang, D. W. Huang, J. J. Lin and X. Yang, J. Chem. Phys., 113, 10597 (2000). H. J. Krautwald, L. Schnieder, K. H. Welge and M. N. Ashfold, Faraday Discuss., 82, 99 (1986). S. A. Harich, D. W. H. Hwang, X. Yang et al., J. Chem. Phys., 113, 10073 (2000). B. Winter, R. Weber, W. Widdra et al., J. Phys. Chem. A, 108, 2625 (2004). L. Belau, K. R. Wilson, S. R. Leone and M. Ahmed, J. Phys. Chem. A, 111, 10075 (2007). J. M. Heller, R. N. Hamm, R. D. Birkhoff and L. R. Painter, J. Chem. Phys., 60, 3483 (1974). K. Kobayashi, J. Phys. Chem., 87, 4317 (1983). J. N. Harvey, J. O. Jung and R. B. Gerber, J. Chem. Phys., 109, 8747 (1998). A. L. Sobolewski and W. Domcke, Phys. Chem. Chem. Phys., 4, 4 (2002). A. L. Sobolewski and W. Domcke, J. Phys. Chem. A, 107, 1557 (2003). B. Ruscic, A. F. Wagner, L. B. Harding et al., J. Phys. Chem. A, 106, 2727 (2002). P. Slavıcek, P. Jungwirth, M. Lewerenz et al., J. Phys. Chem. A, 107, 7743 (2003). A. L. Sobolewski and W. Domcke, J. Chem. Phys., 122, 184320 (2005). A. L. Sobolewski and W. Domcke, Phys. Chem. Chem. Phys., 9, 3818 (2007). R. Mota, R. Parafita, A. Giuliani et al., Chem. Phys. Lett., 416, 152 (2005). H. Okabe, Photochemistry of Small Molecules, John Wiley & Sons, Inc., New York (1978). J. Franck and E. Rabinowitch, Trans. Faraday Soc., 30, 120 (1934). M. J. Packer and D. C. Clary, J. Phys. Chem., 99, 14323 (1995). C. Lee, C. Sosa, M. Planas and J. J. Novoa, J. Chem. Phys., 104, 7081 (1996). S. Re, Y. Osamura, Y. Suzuki and H. F. Schaefer, J. Chem. Phys., 109, 973 (1998). B. Gertner, G. Peslherbe and J. Hynes, Isr. J. Chem., 39, 273 (1999). A. Milet, C. Struniewicz, R. Moszynski and P. E. S. Wormer, J. Chem. Phys., 115, 349 (2001). C. Amirand and D. Maillard, J. Mol. Struct., 176, 181 (1988). S. M. Hurley, T. E. Dermota, D. P. Hydutsky and A. W. Castleman, J. Chem. Phys., 118, 9272 (2003). M. Weimann, M. Farnık and M. A. Suhm, Phys. Chem. Chem. Phys., 4, 3933 (2002). M. Farnık, M. Weimann and M. A. Suhm, J. Chem. Phys., 118, 10120 (2003). K. Bolton and J. Pettersson, J. Am. Chem. Soc., 123, 7360 (2001). K. Bolton, J. Mol. Struct. -Theochem, 632, 145 (2003). A. Al-Halabi, R. Bianco and J. Hynes, J. Phys. Chem. A, 106, 7639 (2002). L. Wang and D. C. Clary, J. Chem. Phys., 104, 5663 (1996). B. J. Gertner and J. T. Hynes, Faraday Discuss., 110, 301 (1998). M. Svanberg, J. B. C. Pettersson and K. Bolton, J. Phys. Chem. A, 104, 5787 (2000). V. Buch, J. Sadlej, N. Aytemiz-Uras and J. P. Devlin, J. Phys. Chem. A, 106, 9374 (2002). J. Graham and J. Roberts, J. Phys. Chem., 98, 5974 (1994). H. Kang, T. H. Shin, S. C. Park et al., J. Am. Chem. Soc., 122, 9842 (2000). S. C. Park and H. Kang, J. Phys. Chem. B, 109, 5124 (2005). J. P. Devlin, N. Uras, J. Sadlej and V. Buch, Nature, 417, 269 (2002). F. Bournel, C. Mangeney, M. Tronc et al., Phys. Rev. B, 65 (20), 201404 (2002). F. Bournel, C. Mangeney, M. Tronc et al., Surf. Sci., 528, 224 (2003). P. Parent and C. Laffon, J. Phys. Chem. B, 109, 1547 (2005). M. Kondo, H. Kawanowa, Y. Gotoh and R. Souda, J. Chem. Phys., 121, 8589 (2004). S. W. Downey and R. S. Hozack, Opt. Lett., 14, 15 (1989). J. Brudermann, U. Buck and V. Buch, J. Phys. Chem., 106, 453 (2002).
Photodissociation of Molecules in Pure and Doped Water and in Nitrogen Heterocyclic Clusters 891 74. 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.
S. C. Park, K. H. Jung and H. Kang, J. Chem. Phys., 121, 2765 (2004). E. M. Cabaleiro-Lago, J. M. Hermida-Ramon and J. Rodriguez-Otero, J. Chem. Phys., 117, 3160 (2002). S. Odde, B. J. Mhin, S. Lee et al., J. Chem. Phys., 120, 9524 (2004). J. B. Nee, M. Suto, L. C. Lee, J. Chem. Phys., 85, 4919 (1986). B. M. Cheng, C. Y. Chung, M. Bahou et al., J. Chem. Phys., 117, 4293 (2002). B. J. Finlayson-Pitts and J. N. Pitts, Jr., Chemistry of the Upper and Lower Atmosphere, Academic Press, San Diego (2000). M. J. Molina, T. L. Tso, L. T. Molina and F. C. Y. Wang, Science, 238, 1253 (1987). O. David, C. Dedonder-Lardeux, C. Jouvet et al., J. Chem. Phys., 120, 10101 (2004). A. Kumar, M. Kolaski and K. Kim, J. Chem. Phys., 128, 034304 (2008). O. David, C. Dedonder-Lardeux and C. Jouvet, Int. Rev. Phys. Chem., 21, 499 (2002). A. Carrera, I. B. Nielsen, P. Carcabal et al., J. Chem. Phys., 130, 024302 (2009). L. M. Frutos, A. Markmann, A. L. Sobolewski and W. Domcke, J. Phys. Chem. B, 111, 6110 (2007). M. N. R. Ashfold, B. Cronin, A. L. Devine et al., Science, 312, 1637 (2006). A. L. Devine, B. Cronin, M. G. D. Nix and M. N. R. Ashfold, J. Chem. Phys., 125, 184302 (2006). M. Schwell, H.-W. Jochims, H. Baumg€artel and S. Leach, Chem. Phys., 353, 145 (2008). B. Jagoda-Cwiklik, P. Slavıcek, L. Cwiklik et al., J. Phys. Chem. A, 112, 3499 (2008). F. B. C. Machado and E. R. Davidson, J. Chem. Phys., 97, 1881 (1992). L. Serrano-Andres, M. P. Fulscher, B. O. Roos and M. Merchan, J. Phys. Chem., 100, 6484 (1996). M. D. Su, J. Phys. Chem. A, 111, 1567 (2007). M. Barbatti, H. Lischka, S. Salzmann and C. M. Marian, J. Chem. Phys., 130, 034305 (2009). V. Poterya, V. Profant, M. Farnık et al., J. Chem. Phys., 127, 064307 (2007). V. Profant, V. Poterya, M. Farnık et al., J. Phys. Chem. A, 111, 12477 (2007). J. Wei, A. Kuczmann, J. Riedel et al., Phys. Chem. Chem. Phys., 5, 315 (2003). J. Wei, J. Riedel, A. Kuczmann et al., Faraday Discuss., 127, 267 (2004). B. Cronin, M. G. D. Nix, R. H. Qadiri and M. N. R. Ashfold, Phys. Chem. Chem. Phys., 6, 5031 (2004). V. Papadakis and T. N. Kitsopoulos, Rev. Sci. Instrum., 77, 083101 (2006). L. Rubio-Lago, D. Zaouris, Y. Sakellariou et al., J. Chem. Phys., 127, 064306 (2007). M. L. Lipciuc, F. Wang, X. Yang et al., ChemPhysChem, 9, 1838 (2008). V. Poterya, V. Profant, M. Farnık et al., J. Phys. Chem. A, 113, 14583 (2009). M. Barbatti, M. Vazdar, A. J. A. Aquino et al., J. Chem. Phys., 125, 164323 (2006). M. Farnık, V. Poterya, O. Votava et al., J. Phys. Chem. A, 113, 7322 (2009). P. Ayotte, S. Plessis and P. Marchand, Phys. Chem. Chem. Phys., 10, 4785 (2008). V. Thomas and R. Iftimie, J. Phys. Chem. B, 113, 4152 (2009).
Index Bold page numbers indicate tables. Italic numbers indicate figures. ab initio molecular dynamics (AIMD) approach, 579, 580, 587 ABPH, see 5-amino-3-arylmethylene-1(3H)isobenzofuranones Acaryochloris marina, 445 acceptor rehybridization model, 315 acetic acid (AcOH), interactions with betacarboline derivatives, 699–704 acetone-HFIP systems, 47 acetone-PFTB systems, 47 2-acetyl-4-chloro-6-nitrophenol, 653–7, 656 2-acetyl-4-methyl-6-nitrophenol, 653–7, 656 acridanones, 280–3 acridine, 737, 747–8 acridine orange (AO), 247 activation energy barrier-crossing transition, see Arrhenius-type energy-barrier-crossing model 2-acylaminophenol, 613 ADC, see analog-to-digital converter adenine, 126, 128, 129 adenine-thymine base pairs, 1–3, 15, 126 excited-state properties, 138–140 ground state structures, 128–9 adenine-uracil base pairs, 2, 3, 9, 22–3 correlations between transitions 0 ! 1 and 1 ! 2, 17–18 excited-state properties, 138–140 frequency fluctuation correlation function, 15–17 geometric correlations, 9–11 IR lineshape, 20, 23 NH stretching-HB length correlation, 13–15 photon echo spectra, 21–2 pump-probe spectra, 20–1 quantum correction for the bath modes, 18–19 velocity autocorrelation function, 12–13 adenosine, 200
Hydrogen Bonding and Transfer in the Excited State, Volume I & II © 2011 John Wiley & Sons, Ltd. ISBN: 978-0-470-66677-7
adiabatic compressibility, 236–7 adiabatic photoisomerization, 211–14 adiabatic/nonadiabatic proton transfer reaction, 737, 738 Agmon-Levine model, 646 7AI, see 7-azaindole AACID, see 5-amino-2-aryl-2-carboxymethylindan-1,3diones AID, see 5-aminoindan-1,3-diones AIMD approach, see ab initio molecular dynamics approach alcohols deuterated, 102–3 hydrogen-bonding interactions with fluorenone, 770–5, 777–81, 786, 787 hydrogen-bonding interactions with ketocyanine dyes, 781–4 hydrogen-bonding interactions with resorufin, 776–7 involvement in hydrogen bond formation, 80, 81 alloxazines, 93–4 amide NH prelocated, 609–13 remote, 613–15 5-amino-2-aryl-2-carboxymethylindan-1,3-diones (AID), 270–4 5-amino-3-arylmethylene-1(3H)-isobenzofuranones (ABPH), 269, 278–80 aminoanthraquinones, 81, 82, 103 aminobenzylidenaphthalides, see 5-amino-3-arylmethylene-1(3H )-isobenzofuranones (ABPH) 7-aminocoumarins, 89, 178, 420, 584 ICT to TICT conversion 424–6 aminofluorenones, 81, 102–3, 766, 785 5-aminoindan-1,3-diones (AID), 269, 270–4 4-aminophthalimide (4-AP), 220, 243 4-aminophthalimide, 99, 420 2-aminopurine, 95, 114, 143
Edited by Ke-Li Han and Guang-Jiu Zhao
894 Index amphiphilic molecules, 711 analog-to-digital converter (ADC), 362 aniline, complex with coumarin 102 (C102), 770, 771, 789–90 1-anilino-8-naphthalenesulfonic acid (ANS), 221, 224, 225, 251–6 1-anilinonaphthalene-8-sulfonate (1,8-ANS), 179, 180 anion sensing by conjugated polymers, 810–13 through ESIPT process, 806–8, 810–13 through ESPT process, 808–10 anisotropy, 231–6 anoxygenic photosynthetic bacteria, 434, 437, 451, 452 see also purple bacteria reaction centres (PBRCs) 1,8-ANS, see 1-anilinonaphthalene-8-sulfonate ANS, see 1-anilino-8-naphthalenesulfonic acid 9,10-anthraquinone (AQ), 200, 201 anthraquinone derivatives, 81, 92, 100 AO, see acridine orange AOT, see bis(2-ethylhexyl)sulfosuccinate sodium salt APDs, see avalanche photodiodes AQ, see 9,10-anthraquinone AQ143 (fluorescent protein), 819 aromatic molecules, 29 Arrhenius-type energy-barrier-crossing model, 219–20, 237 electrostatic attachment of ligand molecules, 256–9 in micellar systems, 237–44 in mixed reverse micellar systems, 250–6 in reverse micellar systems, 244–50 3-arylmethylene-1(3H)-isobenzofuranones (BPH), 275–80 Asp-containing oligopeptides, 613–14 ATR, see attenuated total reflection attenuated total reflection (ATR), 343 avalanche photodiodes (APDs), 361 7-azaindole (7AI), 465, 556, 646, 647, 661–2 conjugated dual hydrogen bonding (CDHB) formation, 563 crystal form, 559–61 cyano analogues of, 571–3 ESDPT in analogue homodimers, 558–61 ESDPT in dimers, 556–8 ESDPT in heterodimers, 561–2 ESDPT in host/guest-type HB complexes, 563–7 hydrogen-bonding geometries, 671 7-azaindole-(H2O)n clusters, 580–4 7-azaindoline (7AZD), 565–6 7AZD, see 7-azaindoline azines, 194 bacteriorhodopsin (BR), 378, 379, 525 detection of water stretching vibrations in, 379–80 hydration switch model, 380–2, 383 role of strong hydrogen bond of water, 379–80, 386–7 structural changes of water in, 382–4
Badger’s rule, 11 basis set superposition error (BSSE), 139 BC, see betacarboline BCA, see N2-methyl-9H-pyrido[3,4-b]indole BEBO model, see bond energy bond order Beer-Lambert law, 684 Beer-Lambert relation, 343 Benesi-Hildebrand equation, 684, 694–5 benzenoid-p bases, 682–5 benzimidazoles, 225, 754–5 benzo[b]fluorenone, 99–100 2-benzofuryl-3-hydroxy-4(1H)-quinolone (3-HQ-Bf), 97 4H-1-benzopyrane-4-thione (BPT), 112–13, 182, 184 benzopyridinic bases, 687–92 2,5-bis(2-benzoxazolyl) hydroquinone, 649 benzylidene phtalides, see 3-arylmethylene-1(3H)isobenzofuranones beta-turn structure, 613, 614 betacarboline (BC), 394, 663 BC-AcOH system, 699–704 BC-BC system, 409–14 BC-HFIP system, 395, 406–9 BC-pyridine system, 405–14, 689–92 interactions with benzenoid-p bases, 682–5 interactions with benzopyridinic bases, 689–92 interactions with methylbenzene bases, 685–6 betacarboline derivatives, 394–5, 680–2 BCA-HFIP system, 403–6 HN-AcOH system, 702–4 HN-HFIP system, 695–8 HN-pyridine system, 687–8 interactions with hydrogen-bond acceptors, 682–92 interactions with hydrogen-bond donors, 692–9 interactions with hydrogen-bonding donoracceptors, 699–705 MBC-HFIP system, 396–403 MHN-t-BuOH system, 693–4 MHN-CIEtOH system, 693–4 MHN-HFIP system, 396–403, 693–4 MHN-pyridine system, 687 bile salts, 163–5, 167 ‘biological water’, 218 biomimicking systems, 218–22, 259–60 Arrhenius model in micellar systems, 237–44 Arrhenius model in mixed reverse micellar systems, 250–6 Arrhenius model in reverse micellar systems, 244–50 self-organized assemblies, 222–3 biomolecules, radiation damage, 866 biorecognition, 566 biprotonic transfer reactions, 556–62 2,5-bis(benzoxazol-2-yl)benzene-1,4-diol (bis (HBO), 756–8 2,5-bis[(2,3-dihydroindolyl)propylene]cyclopentanone (KCD), 781–4, 786, 787 bis[2,6-di(pivaloylamino)] phenyl disulfide, 610
Index bis(2-ethylhexyl)sulfosuccinate sodium salt (AOT), 218, 221, 222, 223, 224, 244–50 in anion sensing, 810–11 bis(HBO), see 2,5-bis(benzoxazol-2-yl)benzene-1,4-diol 2,5-bis(N-methyl N-1,3-propdienylaniline)cyclopentanone (MPAC), 781–2 bis(2,4,6-trihydroxy-phenyl), 182 blue-shifted hydrogen bonds, 126 bond energy bond order (BEBO) model, 646 Boys-Bernardi counterpoise correction scheme, 139 BPH, see 3-arylmethylene-1(3H)-isobenzofuranones BPheo, 439–40 BPT, see 4H-1-benzopyrane-4-thione BR, see bacteriorhodopsin Brønsted-Lowry definition, 463 BSSE, see basis set superposition error t-BuOH, see tert-butyl alcohol tert-butyl alcohol (t-BuOH), interactions with betacarboline derivatives, 692–4 2-butylamino-6-methyl-4-nitropyridine-N-oxide, 509 Ca(II) complexes, 618–19, 622–3, 624 calcium phosphate clusters, 622–3 calixarenes, 185–6 Car-Parrinello molecular dynamics (CPMD), 2 carbazole, 179, 180, 182–3 b-carbolines, 100, 194 carboxylic acid derivatives, 612–15 CARS, see coherent anti-Stokes Raman scattering CASPT2 method, 108 CASSCF method, 108 Catalan model, 85–6, 89 ‘cation-like exciplex’ (CL ), 695–8 CBs, see cucurbiturils CC methods, 108 Cd(II) complexes, 619 CDHB effect, see conjugated dual hydrogen bonding effect CDs, see cyclodextrins cetylpyridinum chloride, 713–14 cetyltrimethylammonium bromide (CTAB) CTAB-methoxynaphthalene systems, 723–5, 726–30 CTAB-naphthol systems, 719–30 CTAB-NaSal systems, 714 micelles, 169, 739, 742 CFCs, see chlorofluorocarbons chair structure, 613 chalcones, synthetic, 283–4 charge-transfer-to-solvent (CTTS), 627, 628–30, 633 Chlide, see chlorophyllide chloride-ion pump, 384–5, 387 chlorine radicals, 880 1-chloro-n-alkanes, 99 2-chloroethanol (CIEtOH), interactions with betacarboline derivatives, 692–5 chlorofluorocarbons (CFCs), 627 chloroperoxidase (CPO), 621, 622
895
chlorophyllide (Chlide), 858–9, 861–2 CHT, see chymotrypsin chymotrypsin (CHT), 225, 443 CI, see configuration interaction CIEtOH, see 2-chloroethanol CIS method, 107–8, 534 CISD method, 108 CL , see ‘cation-like exciplex’ ‘close proximity effect’, 92, 94 CMC, see critical micelle concentration Co(II)-thiolate complexes, 619 coherent anti-Stokes Raman scattering (CARS), 342, 345 conductor-like screening model (COSMO), 108 configuration interaction (CI), 107, 108, 109, 113–15 confined environments, 737–8 see also nanoconfined systems conformational switching, 613–15, 621–2 conjugated dual hydrogen bonding (CDHB) effect, 563 conjugated polymers, 810–13 COSMO, see conductor-like screening model p-coumaric acid (pCA) chromophore, 840–51 coumarin 1 (C1), 178, 424, 425–6 coumarin 7, 103, 426–9 coumarin 30 (C30), 426–9 coumarin 102 (C102), 178–9, 194–5 C102-aniline complexes, 770, 771, 789–90 C102-phenol complexes, 102, 741, 769–70, 787–9 chemical structure, 764 in hydrogen-bonded complexes, 769–70, 771, 787–9 coumarin 120 (C120), 421–4, 584, 585 coumarin 151 (C151), 107, 110, 220, 333, 334, 421, 424, 785 excited-state dynamics of, 584–7 coumarin 152 (C152), 178, 334, 424, 426 coumarin 153 (C153), 101, 110, 178, 220, 334, 334 335, 785 solvation dynamics in [N3[1]][Tf2N] ionic liquid, 335–9 coumarin 343 (C343), 245–6 coumarin 480 (C480), 247, 334 coumarin 481 (C481), 424, 426 coumarin 500 (C500), 221, 224, 225, 244–56 coumarin 522 (C522), 334, 334, 335 coumarins, 81, 92, 420 as fluorescent probes, 178–9 ICT to TICT conversion, 424–9 intermolecular hydrogen bonding, 194–5, 421–6 intramolecular hydrogen bonding, 426–9 PET reaction in coumarin-amine systems, 332–5 temperature-dependent solvation, 220 CPB-naphthol systems, 719–21 CPMD, see Car-Parrinello molecular dynamics CPO, see chloroperoxidase CPP, see critical packing parameter cresol, 716 critical micelle concentration (CMC), 711
896 Index critical packing parameter (CPP), 712–13 CTAB, see cetyltrimethylammonium bromide CTTS, see charge-transfer-to-solvent cucurbiturils, 186–7 CURC, see curcumin curcumin as photosensitizer, 357–8 chemical structure, 357 ESIPT, 368–9 hydrogen-bond-mediated de-excitation, 366–9 KEHB formation, 363 cyano-substituted indolines, 91 3-cyanoaniline, 91 cyclodextrins (CDs), 180–5 chemical structure, 181 in host-guest inclusion complexes, 169, 180–5, 324–5 ESIPT, 185, 657 ESPT, 185, 738 Cys-containing oligopeptides, 612 cytosine, 126, 128, 129 DAF, see N,N-dimethyl ethanol ammonium formate ‘dangling’ hydrogen bonds, 781, 790 DBO, see 2,3-diazabicyclo[2.2.2]oct-2-ene DBPZ, see dibenzo[a,c]phenazine DCM, see 4-(dicyanomethylene)-2-methyl-6(p-dimethylamino-styryl) 4H-pyran DCMeth, see dicinnamoylmethane deuterated alcohols, 102–3 deuterated water, 103 DHBQ, 407, 409, 415, 696, 698–9 DHBZ, 696, 698–9 2,3-DHN, see 2,3-dihydroxynaphthalene 40 -dialkylamino-3-hydroxyflavones, 104 4-dialkylaminopyrimidines, 320, 321 diarylazomethines, 614, 616 2,3-diazabicyclo[2.2.2]oct-2-ene (DBO), 183 diazines, 194, 689 dibenzo[a,c]phenazine (DBPZ), 196–9 dicinnamoylmethane (DCMeth) chemical structure, 363 ESIPT, 371 hydrogen-bond-mediated de-excitation, 369–71 KEHB formation, 363, 369, 371–3 diCN-HBO system, 573–4 4-(dicyanomethylene)-2-methyl-6(p-dimethylamino-styryl) 4H-pyran, (DCM), 220, 224–5, 239–43 p-(N,N-diethylamino)benzoic acid (DMABA), 184, 321 diffusion, 170–2 6-diformyl phenol, 650–2, 653, 654, 655 dihydrogen bonds, 126 2,3-dihydroxynaphthalene (2,3-DHN), 716, 717 1,2-di(30 -isoquinolyl)ethene ((3IQ)2E), 208–10, 211 b-diketones, 356–8 dimethyl aniline (DMA), 332 N,N-dimethyl ethanol ammonium formate (DAF), 332–5
4-N,N-dimethylamino cinnamaldehyde, 104 4-(dimethylamino)-pyridine (DMAP) absorption properties, 51–6 DMAP-HFIP complexes, 51–2, 54–6, 70–2, 74 hydrogen bond basicity, 63 solvatochromism, 69–72 triplet formation yield, 74 4-dimethylaminobenzethyne, 315 p-dimethylaminobenzoate, 112, 113 dimethylaminobenzonitrile (DMABN), 70–2, 112 dual fluorescence, 313–16 role of polarity and viscosity in ICT emission, 317–18 TICT emission, 183–4, 318–20 N,N-dimethylaminonaphthyl acrylates, 319 N,N-dimethylaminonaphthyl-(acrylic)-acid (MDMANA), 97 N,N-dimethylaminonaphthyl-(acrylo)-nitrile (DMANAN), 97, 103 2-(40 -N,N-dimethylaminophenyl)imidazo[4,5-b]pyridine (DMAPIP-b), 95, 325–7 2-(40 -N,N-dimethylaminophenyl)pyrido[3,4-d]imidazole (DMAPPI), 121–7 inclusion complex with cyclodextrins, 324–5 N,N-dimethylaniline (DMAN), 52–3 N,N-dimethylbenzodiazepine, 97 N-(2,6-dimethylphenyl)2,3-naphtalimide (DMPN) absorption spectra , 43, 44–5, complexation, 68–9 DMPN-HFIP-n-hexane systems, 43, 44–9, 56–60, 75–6 DMPN-PFTB-n-hexane systems, 47 hydrogen bond basicity, 56–9 Dimroth-Reichard model, 84–5 1,2-di(20 -naphthyl)ethene ((2N)2E), 208 dipolar solvation, 761 dipyrido[2,3-a:30 ,20 -i]carbazole (DPC), 662–9 1,2-di(30 -quinolyl)ethene ((3Q)2E), 208, 211 ‘distinguished coordinate’ approach, 595, 599 DMA, see dimethyl aniline DMABA, see p-(N,N-diethylamino)benzoic acid DMABN, see dimethylaminobenzonitrile DMAC, see 1-(2-pyridyl)-5-(4-dimethylaminophenyl)penta-2,4-diene-1-one DMAN, see N,N-dimethylaniline DMANAN, see N,N-dimethylaminonaphthyl-(acrylo)nitrile DMAP, see 4-(dimethylamino)-pyridine DMAPIP-b, see 2-(40 -N,N-dimethylaminophenyl)imidazo [4,5-b]pyridine DMAPPI, see 2-(40 -N,N-dimethylaminophenyl)pyrido [3,4-d]imidazole DMPN, see N-(2,6-dimethylphenyl)2,3-naphtalimide DNA, 126–7, 199–201 see also nucleic acid base pairs; nucleic acid bases DNA photolyase, 858 dodecyltrimethylammonium bromide (DTAB), 739–40 donor rehybridization model, 315
Index double hydrogen bond complexes, 407–9 double-proton transfer, 661, 664, 667, 676 see also excited-state double proton transfer (ESDPT) DPC, see dipyrido[2,3-a:30 ,20 -i]carbazole drugs, fluorescence decay studies, 354–5 DsRed, 819, 820, 830, 831 DSS, see dynamic Stokes shift DTAB, see dodecyltrimethylammonium bromide dual fluorescence, 464, 465, 590 applications, 518–19 betacarboline derivatives, 403, 407, 414 effect of excitation frequency on, 506–9 3-hydroxyflavones, 467–9, 505, 516–17 solute-solvent hydrogen bond formation, 95–8 TICT mechanism, 313–16 dual-luminescent napthalimides, 48–51 dynamic quenching of fluorescence, 475–509 effect of excitation frequency on dual fluorescence, 506–9 effect on excited-state depopulation rates, 476–8 effect on fluorescence intensities, 489–91 effect on populations of excited states, 484–91 effect on proton transfer rate, 503–4 effect on steady-state spectra, 478–80 in 3-hydroxyflavone, 480–1, 504–6 in 3-hydroxyflavone analogues, 481–3 reactions from high-lying excited states, 483–509 Stern-Volmer constants, 491–4, 499–501 temperature effect on proton transfer rate, 501–3 temperature quenching, 496–9 two-state excited-state reaction, 475 under different physical conditions, 494–99 dynamic solvation, 82 dynamic Stokes shift (DSS), 160 Edward’s method, 71 EET, see excitation energy transfer effective fragment potential (EFP) method, 580 effective medium theory, 237 EFP method, see effective fragment potential method Eigen model, 646 electron nuclear double resonance (ENDOR), 440 ENDOR, see electron nuclear double resonance enzymes, light-driven 858–9 see also protochlorophyllide oxidoreductase (POR) ES-AIMD simulations, 628–9, 631–2, 635–6 ESCT, see excited-state charge transfer ESDPT, see excited-state double proton transfer ESHAT, see excited-state H-atom transfer ESHT, see excited-state hydrogen transfer ESIPT, see excited-state intramolecular proton transfer ESPT, see excited-state proton transfer excitation energy transfer (EET), 437 excitation wavelength dependence, 159–66 excited-state charge transfer (ESCT), ESPT-coupled 567–74, 569–73
897
excited-state double proton transfer (ESDPT), 465 catalytic versus noncatalytic reaction, 563–4, 566 in 7-azaindole analogue homodimers, 558–61 in 7-azaindole dimers 556–8 in 7-azaindole heterodimers, 561–2 in host/guest types of hydrogen-bonded complexes, 563–7 in multiple-hydrogen bonding systems, 566–7 p-electron conjugation tuning, 565–6 excited-state H-atom transfer (ESHAT), 526 chromophore local solvation, 546–8 in green fluorescent protein, 546–8 mode selectivity, 543–6 reaction path, 536–40 solvent effect, 548–51 wire solvation, 540–2 excited-state hydrogen transfer (ESHT), 579–80 in 7-azaindole-(H2O)n system, 580–4 excited-state intramolecular proton transfer (ESIPT), 355–6, 589–90, 646 anion sensing based on, 806–8, 810–13 applications, 650 compared in 1H2NA and 2H3NA, 601–5 four-level model, 472–3, 475, 642 from the Sn singlet state, 484, 509–18 in curcuminoids, 368–9, 371 in b-diketones, 356–7, 358 in o-hydroxy carbonyl compounds, 650–7 in 3-hydroxyflavones, 467–75, 509–18, 647, 648 in 2-(20 -hydroxyphenyl)benzothiazole, 648 in 2-(2-hydroxyphenyl)benzoxazole, 748–50, 755 in 2-(2-hydroxyphenyl)benzoxazole derivatives, 750–59 in naphthalene derivatives, 589–607 internal torsion and solvent dielectric, 653–7 kinetic type of reaction, 473, 474, 479 mechanism, 97, 98, 647 perturbation to, 648–50 solvent assisted, 646 thermodynamic type of reaction, 474, 480 types of, 646–7 within cyclodextrin cavities, 185 see also dynamic quenching of fluorescence excited-state proton transfer (ESPT), 464–5, 526–7, 641–2 adiabatic/nonadiabatic, 737, 738 anion sensing based on, 808–10 excited-state charge-transfer-coupled (ESCT/ESPT system), 569–73, 569–73 in nanoconfined systems, 167–70 in Y42F mutant, 850–1 intermolecular/intramolecular, 641, 642 kinetics of, 645–5 of hydroxyaromatic compounds, 736–7 of hydroxyaromatic compounds in organized media, 737–41
898
Index
potential energy diagram, 643–4 thermodynamics of, 644–5 within cyclodextrin cavities, 185, 738 see also excited-state double proton transfer (ESDPT); excited-state intramolecular proton transfer (ESIPT) excited-state reactions, characteristics of, 736–7 F dye, 467, 481–2 F2 dye, 467, 481–2 F127, 171–2 FA dye, 481–2 FCS, see fluorescence correlation spectroscopy Fe(II)) complexes, 622 Fe(III) complexes, 621–2 fluorenone, 92, 100, 104 as fluorescent probe, 179, 182 chemical structure, 764 CO stretching mode, 772–5 fluorenone-methanol complexes, 151–6, 741–2, 771, 781 interactions with alcohols, 741–2, 763–8, 770–5, 777–81, 786, 787 triplet excited states, 150–6 see also aminofluorenones fluorescence definition, 176 magic angle emission, 234 see also dual fluorescence fluorescence anisotropy, 231–6 fluorescence correlation spectroscopy (FCS) 170–2 fluorescence emission, 528, 529 fluorescence enhancement, 90, 92–5 fluorescence lifetime, 90, 176 fluorescence probes, for studying solvation dynamics, 81, 98–9 fluorescence quantum yield, 90, 176 fluorescence quenching, 90, 91–2 by hydrogen-bond strengthening, 741–2 by solvent, 498 in betacarboline derivatives, 411–3 in pyridylindoles, 678–79, 680 in pyrroloquinolines, 678–9, 680 fluorescence-quenching constant, 332 fluorescence resonance energy transfer (FRET), 166–7 fluorescence spectroscopy, 176, 177, 528, 530 and electronic absorption, 763–8 fluorescent proteins (FP), 819 computational methodology, 823–4 internal conversion mechanism, 829–32 TICT states in, 831 see also green fluorescent protein (GFP), red fluorescent proteins (RFPs) N-(3-fluorophenyl)-2,3-naphthalimide (MFPN), 48–51 F€ orster cycle, 464, 642, 645, 666 F€ orster equation, 687 four-wave mixing (IR) spectroscopy, 3–6
FP, see fluorescent proteins FRET, see fluorescence resonance energy transfer FTIR spectroscopy, 379–80, 388 G-factor, 235 Geiger mode, 361 Gibbs free energy, 433–4, 437, 761 green fluorecent protein (GFP), 526–7, 815–19 chromophore local solvation, 546–8 cluster models, 820–22 computational methodology, 820–3 photoisomerization mechanism, 816–17 pp and ps states, 546–8 proton transfer in, 816–18, 820–1, 824–9, 834–5 quantum dynamics calculations, 822–3, 834 structure of, 815–16 wild-type, 816, 825 Green’s function, 227, 229 Grotthus mechanism, 525–6 ground-state intramolecular proton transfer (GSIPT) 1H2NA, 596–9 2H3NA, 599–601 GSIPT, see ground-state intramolecular proton transfer guanine, 126, 128, 129, 200 hydration of, 127, 131–8 guanine-cytosine base pairs, 126 excited-state properties, 138–142 ground state structures, 128–9 hydration, 142 guanine-guanine base pairs, 140–2 guanosine 50 -monophosphates, 201 H258, see 20 -(4-hydroxyphenyl)-5-[5-(4-methylpiperazine-1-yl)benzimidazo-2-yl HA, see p-hydroxyacetophenone HA-13C, 288, 295 HAB, see 3-hetarylmethylene-1(3H)-isobenzofuranones HA-D4, 288, 295 hairpin-turn structure, 614, 622 halide anion-water clusters, 627 halorhodopsin (HR), 378, 379, 387, 388 role of water hydrogen bond, 384–5 Halorhodospira halophila, 840 1H2AN, see 1-hydroxy-2-acetonaphthone HAP, see 4-hydroxy-5-azaphenanthrene hard and soft acid-base behaviour (HSAB), 785 harmane, 101, 662 see also 1-methyl-9H-pyrido[3,4-b]indole (HN) HBC, see hydrogen bond complexes HBI, see 2-(2-hydroxyphenyl)benzimidazole HBO, see 2-(2-hydroxyphenyl)benzoxazole HBO derivatives, see 2-(2-hydroxyphenyl)benzoxazole derivatives Hbr-Arn systems, 867–8, 875 HBT, see 2-(2-hydroxyphenyl)benzothiazole HcRed, 819, 832–4
Index HEAN, see N-[2-(2-hydroxylethylamino)-ethyl]-1,8naphthalimide Henderson-Hasselbalch equation, 801 3-hetarylmethylene-1(3H)-isobenzofuranones (HAB), 274–5, 276, 277 1,1,1,3,3,3-hexafluoro-propan-2-ol (HFIP) complexes with betacarboline, 395, 406–9 complexes with betacarboline derivatives, 395–406, 692–8 complexes with DMAP, 51–2, 54–6, 70–2, 74 complexes with DMPN, 43, 44–9, 56–60, 75–6 complexes with isoindolo[2,1-a]indole-6-one, 41–4, 61–2, 73 hexanol, 718 4-hexylresorcinol, 85 HFIP, see 1,1,1,3,3,3-hexafluoro-propan-2-ol 3HFs, see 3-hydroxyflavones Hg(II) complexes, 619, 620 high-lying excited states, 483–4 6HIQ, see 11-propyl-6H-indolo-[2,3-b]quinoline 1H2MN, see methyl-1-hydroxy-2-naphthoate HN, see 1-methyl-9H-pyrido[3,4-b]indole 1H2NA, see 1-hydroxy-2-naphthaldehyde 2H3NA, see 2-hydroxy-3-naphthaldehyde Hoogsteen base pairs, 129 HOPB, see 2(20 -hydroxy-50 -t-octylphenyl) benzotriazole host-guest inclusion complexes, 175–7, 187–8 calixarenes, 185–6 cucurbiturils, 186–7 cyclodextrins, 180–5 supramolecular, 175, 177, 188 pHP, see p-hydroxyphenacyl HPA, see p-hydroxyphenacyl acetate HPAA, see p-hydroxyphenylacetic acid HPDP, see p-hydroxyphenacyl diphosphate HPPP, 288, 307 HPTS, see 8-hydroxypyrene-1,3,6-trisulfonate 7HQ, see 7-hydroxyquinoline 3-HQ-Bf, see 2-benzofuryl-3-hydroxy-4(1H)-quinolone 7HQ(NH3)n clusters, 530–2, 534, 541 ESHAT, 537–4, 548–51, 549 ESPT reaction path, 537 7HQ(NH3)n(H2O)m clusters, 548, 549 HR, see halorhodopsin HSAB behaviour, see hard and soft acid-base behaviour human serum albumin, 472 hydration number, 237–9 hydration switch model 380–2, 383, 384 hydrogen bond basicity of DMAP 63 of DMPN and DMPN-HFIP systems, 56–60 of isoindolo[2,1-a]indole-6-one, 61–2 hydrogen bond complexation, 39–40 complexation mechanism, 41, 45 effect on absorption and fluorescence spectra, 41–56 effect on triplet-state properties, 73–6
899
reaction rate, 64–9 solvatochromism, 69–72 hydrogen bond strengthening, 741–2 hydrogen-bonded complexes, 761–2 electronic spectroscopy studies, 775–90 formation of, 761–2 hydrogen bond dynamics versus solvation, 784–6 of betacarbolines, 395–415 of coumarin 102 (C102), 769–70, 771, 787–90 of fluorenone, 763–8, 770–5 spectroscopic properties of, 763–71 time-resolved vibrational spectroscopy studies, 786–90 hydrogen-bonded wires, 525–8, 550, 551 see also excited-state H-atom transfer (ESHAT); 7HQ (NH3)n clusters hydrogen-bonding effects, 288–9 p-hydroxyacetophenone (HA), 289–301 p-hydroxyphenacyl acetate (HPA), 301–4 p-hydroxyphenacyl diphosphate (HPDP) p-methoxyacetophenone (MAP), 290, 300 on intramolecular charge transfer, 313–16, 318–27 hydrogen bromide, see Hbr-Arn systems hydrogen chloride, 880 hydrogen fluoride, 889 hydrogen halide-water clusters, 870, 871, 874–5 CTTS and excitation cross sections, 877–80 double isotope substitution, 876–7 KED spectrum, 875–6 hydrogen halides, 871, 873–5, 880 see also hydrogen halide-water clusters hydrogen iodide, 633, 634 hydrogen radicals, 633 hydronium radical (H3O) in doped water clusters, 874–5, 888 in pure water clusters, 869–72, 888 1-hydroxy-2-acetonaphthone (1H2AN), 590 4-hydroxy-5-azaphenanthrene (HAP), 649 o-hydroxy benzaldehyde, 646, 647, 650, 651 o-hydroxy-benzophenone, 649 o-hydroxy carbonyl compounds ESIPT in, 650–7 o-hydroxy benzaldehyde, 650 symmetrically substituted compounds, 650–7 1-hydroxy-2-naphthaldehyde (1H2NA), 590–6, 598 ESIPT, 601–5 HOMO and LUMO orbitals, 601–6 intramolecular hydrogen bonding, 601–3 potential energy curve, 596–9 structure, 591 2-hydroxy-3-naphthaldehyde (2H3NA), 590–6, 598 ESIPT, 601–5 HOMO and LUMO orbitals, 601–6 intramolecular hydrogen bonding, 601–3 potential energy curve, 599–601 structure, 591
900
Index
2(20 -hydroxy-50 -t-octylphenyl) benzotriazole (HOPB), 648 2(20 -hydroxy phenyl)benzothiazole, 648, 649 7-hydroxy quinoline, 646 3-hydroxy xanthone, 646 o-hydroxyacetophenone (OHAP), 590 p-hydroxyacetophenone (HA), 288–301 hydroxyaromatic compounds ESPT, 736–41 ESPT in organized media, 737–41 fluorescence quenching by H-bond strengthening, 741–2 photochemistry of, 730–6 photophysics of, 730–6 role of interfacial H-bonding, 742–3 o-hydroxybenzaldehyde (OHBA), 590, 649 3-hydroxybenzofuranochromones, 481 3-hydroxychromones, 467, 468, 475, 481 3-hydroxyflavones (3HFs), 465–6 chemical structure, 467, 647 dual fluorescence, 467–9, 505–9, 516–17 dynamic quenching of fluorescence, 480–3, 504–6 dynamic quenching under different physical conditions, 494–99 effect of excitation frequency on dual fluorescence, 506–9 ESIPT, 467–75, 509–18, 647, 648 excited-state transformations, 472–5 ground-state anionic form, 469–72 novel analogues, 481–3 S2 fluorescence, 469 separation of a weak fluorescence signal, 504–6 Stern-Volmer constants, 499–501 temperature effect on proton transfer, 501–3 1-hydroxyfluorenone, 94–5 4-hydroxyfluorenone, 100 N-[2-(2-hydroxylethylamino)-ethyl]-1,8-naphthalimide (HEAN), 95 3-hydroxynaphthalene-2-carboxylate (SHNC), 725 p-hydroxyphenacyl (pHP), 288 p-hydroxyphenacyl acetate (HPA), 301–4, 307 p-hydroxyphenacyl diphosphate (HPDP), 301–9 20 -(4-hydroxyphenyl)-5-[5-(4-methylpiperazine-1-yl)]benzimidazo-2-yl (H258), 224, 225, 256–9, 260 p-hydroxyphenylacetic acid (HPAA), 288, 307–8 2-(2-hydroxyphenyl)benzimidazole (HBI), 754 2-(2-hydroxyphenyl)benzothiazole (HBT), 648, 754 2-(2-hydroxyphenyl)benzoxazole (HBO) ESIPT in, 748–50, 755 2-(2-hydroxyphenyl)benzoxazole (HBO) derivatives, 758–9 effect of heteroatom substitution on ESIPT, 754–5 effect of temperature on ESIPT, 755–6 ESIPT in bis(HBO), 756 in anion sensing, 810 8-hydroxypyrene-1,3,6-trisulfonate (HPTS), 168–70
7-hydroxyquinoline (7HQ), 527–8, 532–3, 535–8, 540, 544–5 see also 7HQ(NH3)n clusters; 7HQ(NH3)n(H2O)m clusters 3-hydroxyquinolones, 97, 465 hypericin, 649 IBW-1 molecules, 243 IBW-2 molecules, 243 ice, 872, 873, 874, 880 ICT, see intramolecular charge transfer IFW molecules, 243 IHB, see intramolecular hydrogen bond I(H2O)2–5 complexes, 628–30 imidazole clusters, 880–8 imidazolium ionic liquids, 341–2, 346–9 imidazolium ring, 341–2 in-plane bonds, 80, 81, 115, 117 indoles, 179, 182–3 indole-(H2O)2 complex, 115, 117 indolines, cyano-substituted, 91 infrared (IR) spectroscopy, 342–3 2D, 21–2, 23, 792 characterizing hydrogen-bonded complexes, 768–72 for studying hydrogen bonding in ionic liquids, 342 four-wave mixing, 3–6 linear/nonlinear, 1–2, 3–8 pump-probe, 4–5, 772–3 second-order cumulant expansion, 6–8 intermolecular charge transfer, 355 intermolecular excited-state hydrogen bonding 269–70 acridanones, 280–3 5-amino-2-aryl-2-carboxymethylindan-1,3diones, 270–4 3-arylmethylene-1(3H)-isobenzofuranones, 275–80 b-carbolines, 194 coumarins, 194–5 diazines, 195–9 3-hetarylmethylene-1(3H)-isobenzofuranones, 274–5, 276, 277 quinones, 199–202 synthetic chalcones, 283–4 intersecting state model, 646 intersystem crossing (ISC), 150, 176, 649 intramolecular charge transfer (ICT), 355 conversion to TICT state, 424–9 formation of, 80, 81 hydrogen-bonding effects, 313–16, 318–27 hydrogen-bonding with acceptor moiety, 320–7 hydrogen-bonding with donor moiety, 318–20 planar, 315 rehybridization by, 315 role of polarity and viscosity, 317–18 wagging, 315 see also twisted intramolecular charge transfer (TICT) intramolecular hydrogen bond (IHB), 205–6
Index in stilbene-like molecules 206–14 intramolecular vibrational redistribution (IVR), 647 iodic acids, 628, 633 ion-water clusters, 627 ionic liquid/cosolvent mixtures, 348 ionic liquids, 331, 341–2 FRET in, 166–7 imidazolium, 341–2, 346–9 vibrational spectroscopic studies, 341–9 see also room-temperature ionic liquids (RTILs) (3IQ)2E, see 1,2-di(30 -isoquinolyl)ethane IR spectroscopy, see Infrared (IR) spectroscopy ISC, see intersystem crossing isoindolo[2,1-a]indole-6-one, 41–4 hydrogen bond basicity, 61–2 kinetics of complexation reaction, 64–8 triplet formation yield, 73 isoindolo[2,1-a]indole-6-one-HFIP-n-hexane systems, 41–4, 61–2, 73 IVR, see intramolecular vibrational redistribution Kamlet-Taft model, 85, 87–8, 89 Kamlet-Taft parameters, 98 Kamlet-Taft solvatochromic scale, 767, 768 Kasha rule, 469, 508 Katushka, 819 KCD, see 2,5-bis[(2,3-dihydroindolyl)propylene] cyclopentanone KEHB, see keto-enol H-bond Kemp’s acids, 613, 624 keto-enol H-bond (KEHB), 356, 357, 363, 368–9, 371–3 ketocyanine dyes-alcohol systems, 781–4 7-ketoquinoline (7KQ), 527–8 KFP, see kindling fluorescent protein KIE, see kinetic isotope effect kindling fluorescent protein (KFP), 831 kinetic isotope effect (KIE), 448, 647 Kok cycle, 440–1, 442, 447, 449 Kosower Z values, 735–6 7KQ, see 7-ketoquinoline Laplace’s equation, 236 LBHB, see low-barrier hydrogen bond lecithin, 221, 224, 250–6 light-induced charge separation effects of hydrogen bonding, 438–40 type II reaction centres, 437–8 linear solvation energy relationship (LSER), 85 linear-to-turn conformational switch, 614, 615 Lippert-Mataga equation, 177–8, 419, 423, 731, 766 Lippert-Mataga model, 83–5, 731 low-barrier hydrogen bond (LBHB), 443 LSER, see linear solvation energy relationship 3MAI, see 3-methyl-7-azaindole MAP, see p-methoxyacetophenone
901
MAPAEE, see (E)-3(4-methylamino-phenyl)-acrylic acid ethyl ester Marcus theory, 443, 646 Marquardt’s algorithm, 45, 75 MBC, see N9-methyl-9H-pyrido[3,4-b]indole MCA, see multichannel analyser MCSCF method, 579, 580 MCTDH method, 818, 825 MDMANA, see N,N-dimethylaminonaphthyl-(acrylic)acid menadione, 199–201 merocyanine dye, 111 metal complexes effect of pKa shifts, 615–18 NH S hydrogen bonds switching, 621–2 pp-dp interaction, 618–19 rearrangement of hydrogen-bond networks, 622–3 regulation of thiolate and phenolate ligands, 619–21 metal-sulfur complexes, 609 metal-thiolate complexes, 609, 619–20 methanol complex with Phlide, 860 complexes with fluorenone, 741–2, 771, 781 p-methoxyacetophenone (MAP), 290, 300 methoxynaphtalene-CTAB systems, 723–5, 726–30 N-(4-methoxyphenyl)-2,3-naphthalimide, (PMPN), 48–51 6-methoxyquinoline (6MQ), 798–802 3-methyl-7-azaindole (3MAI), 559–61 4-methyl-2,6-diacetyl phenol, 651–2, 654 4-methyl-2,6-dicarbomethoxy phenol, 652, 655 N9-methyl-harmane, 662 see also N9-methyl-1-methyl-9H-pyrido[3,4-b]indole (MHN) methyl-1-hydroxy-2-naphthoate (1H2MN), 590 methyl-1-hydroxy-2-naphthoate (MHN12), 590 methyl-2-hydroxy-1-naphthoate (MHN21), 590 methyl-2-hydroxy-3-naphthoate (MHN23), 590 methyl-3-hydroxy-2-naphthoate, 649 3-methyl-6-hydroxy-m-phthalic acid, 652 1-methyl-4-(40-hydroxystyryl) pyridinium betaine, 89 2-(N-methyl-N-isopropylamino)-5-cyanopyridine, 315, 316 N9-methyl-1-methyl-9H-pyrido[3,4-b]indole (MHN), 394, 395, 663 interactions with benzopyridinic bases, 687–9 interactions with hydrogen-bond donors, 692–4 MHN-CIEtOH system, 693–4 MHN-HFIP system, 396–403, 693–4 MHN-pyridine system, 687 MHN-t-BuOH system, 693–4 N-methyl-1,8-naphthalimide, 93 2-methyl 1,4-naphthoquinone (MQ), 199–201 1-methyl-9H-pyrido[3,4-b]indole (HN), 662 chemical structure, 663 HN-AcOH system, 702–4
902 Index HN-HFIP system, 695–8 HN-pyridine system, 687–8 interactions with acetic acid, 702–4 interactions with benzenoid-p bases, 682, 684–5 interactions with benzopyridinic bases, 687–9 interactions with hydrogen-bond donors, 692, 695–8 interactions with methylbenzene bases, 686 interactions with non-aromatic acceptors, 692 N2-methyl-9H-pyrido[3,4-b]indole (BCA) 394, 395 BCA-HFIP system, 403–6 N9-methyl-9H-pyrido[3,4-b]indole (MBC), 394, 395 MBC-HFIP system, 396–403 methyl salicylate (MS), 355, 590, 644, 646, 649 tautomerization, 642 (E)-3(4-methylamino-phenyl)-acrylic acid ethyl ester (MAPAEE), 96 methylbenzene bases, 685–6 N2-methylcarboline, 698 N2-methylharmane (T-MeHN), 666, 681, 692 MFPN, see N-(3-fluorophenyl)-2,3-naphthalimide MHN, see N9-methyl-1-methyl-9H-pyrido[3,4-b]indole MHN12, see methyl-1-hydroxy-2-naphthoate MHN21, see methyl-2-hydroxy-1-naphthoate MHN23, see methyl-2-hydroxy-3-naphthoate micelles, 218, 711–12 Arrhenius model at interface, 237–44 CTAB, 739, 742 DTAB, 739–40 effect of neutral aromatic dopants, 716–30 effect of salt anions, 712–15 interactions with dopant, 726–30 micelle-to-vesicle transition, 714–15, 716–18 polymer-like, 713–15 role of OH group of dopants, 725 SDS, 220, 222, 237–44, 739, 741 shear-induced viscoelasticity, 722–5 shear-induced viscosity, 719–21 solvation dynamics in, 738 structure of, 222, 223 vesicle-to-micelle transition, 715 worm-like 712, 714–15, 716–8, 725–6 mKate, 834 Mn4OxCa cluster, 446–50 molecular aggregates, see micelles molecular bridged compounds, 314 molecular hydrodynamic approach, 227 molecular recognition, 566–7, 805 see also anion sensing molecular-wire effect, 810 MPAC, see 2,5-bis(N-methyl N-1,3-propdienylaniline) cyclopentanone MPDP, 307 mPlum, 819 MQ, see 2-methyl 1,4-naphthoquinone 6MQ, see 6-methoxyquinoline MRCI method, 108 MRPT2 method, 819, 823
MS, see methyl salicylate MS-PET, see multiple-site electron and proton transfer multichannel analyser (MCA), 362 multiple-site electron and proton transfer (MS-PET), 443 mutation, 126–7, 556 (2N)2E, see 1,2-di(20 -naphthyl)ethane NaDC, see sodium deoxycholate NADPH, 859, 861, 862 nanoconfined systems, 159–60 diffusion of organic dyes, 170–2 excited-state proton transfer (ESPT), 167–70 fluorescence resonance energy transfer (FRET), 166–7 solvation dynamics, 160–6 2-naphthaldehyde 592–3 naphthalimides, 48–51 1-naphthol C¼O and O-H stretching frequencies, 592–3 ESPT, 736, 737–41 spectral shifts, 730–6 transition energies, 735 volume changes, 802 2-naphthol cis ! trans barrier height, 36 ESPT, 737–41 H-bonded clusters of, 30–6 IR spectra, 30–1 isomers of, 30 spectral shifts, 730–6 transition energies, 735 VER dynamics, 31–6 volume changes, 802 naphthol-CPB systems, 719–21 naphthol-CTAB systems, 719–30 naproxen, 182, 183 NaSal, 713, 714, 716 NET, see nonadiabatic electron transfer 2-nitrobenzaldehyde, 798 nonadiabatic electron transfer (NET), 443 [N3[1]][Tf2N], see N,N,N-trimethyl-N-propylammonium bis(trifluoro-methanesulfonyl)imide nucleic acid base pairs, 1–3 excited-state properties, 138–42 ground-state structures, 128–9 hydration, 142 structures of 126 see also adenine-thymine base pairs; adenine-uracil base pairs; guanine-cytosine base pairs nucleic acid bases, 125, 127 electronic transitions in, 129–30 geometries of, 130–1, 140–2 ground-state structures of, 128 hydration, 127, 131–8 mutation, 125–7 non-radiative deactivation, 142–3 structures of 126
Index tautomerism, 127, 562 nucleoside 50 -monophosphates, 201 OHAP, see o-hydroxyacetophenone OHBA, see o-hydroxybenzaldehyde oligopeptides Asp-containing, 613–14 Cys-containing, 612 Onsager’s reaction field theory, 83 out-of-plane bonds, 80, 81, 115, 117 oxazines, 91, 111–12, 741 oxidative water splitting general reaction pattern, 440–1 hydrogen bonding of YD, 445–6 oxidation of WOC, 445–51 oxidation of YZ, 441–6 9-oxo-imidazopurine derivative, 110–11 ozone destruction, 627, 880 P123 triblock copolymer FRET in, 167 gel, 165, 171 micelles, 171 solvation dynamics in, 165–6 P123-CTAC aggregate, 166, 169–70 P123-SDS aggregate, 165–6 P450, 620, 621–2 P680, 435, 439, 441–6 PBRCs, see purple bacteria reaction centres PCET, see proton-coupled electron transfer PCM, see polarizable continuum model PDAB, see 4-phenyl-1-N,N-dimethylaminobutane PEO-PPO-PEO triblock copolymers, 162–3, 165–6, 170 perfluoro-tert-butanol (PFTB), 47, 49, 50 perfluoro-tert-butyl alcohol (PFTB), 763 PES, see potential energy surfaces PET, see photoinduced electron transfer PFTB, see perfluoro-tert-butanol PFTBA, see perfluoro-tert-butyl alcohol phenolate-metal complexes, 620–1 phenols change of properties in excited state, 747–8 conformational switching, 613–15 effect of dopants on micelle microstructure, 716 ESPT, 737 incorporation in surfactants, 717, 718 interactions with coumarin 102, 741, 769–70, 787–9 pKa shifts by prelocated hydrogen bond, 612–13 phenothiazine derivatives, 93 N-phenyl 1–2-aminonaphthalene, 419–20 N-phenyl-benzamide, 646, 647 4-phenyl-1-N,N-dimethylaminobutane (PDAB), 101–2 phenyl-1-hydroxy-2-naphthoate, 649 Pheo, 439–40 Phlide, see protochlorophyllide photoactive yellow protein (PYP), 840–1
903
crystal structure, 840, 841 electronically excited state of, 848–50 initial stages of the photocycle, 840–3, 844–8, sub-picosecond time-resolved transient spectroscopy, 846–8 vibrational modes, 843 vibrational structural markers, 843–4 Y42F mutant of 850–1 photocycle, 839–40 photodeprotection reaction, 288, 307–9 photodissociation, 627 of bare water molecule, 868–9 of doped water clusters, 870, 871, 874–80 of hydrated hydrogen iodide clusters, 633, 634 of hydrogen halide-water clusters, 870, 871, 874–80 of isolated hydrogen halides, 873–4 of nitrogen heterocycles, 880–8 of pure water clusters, 869–73 photodynamic therapy, 355 photoexcitation, 193–4 see also intermolecular excited-state hydrogen bonding photoinduced electron transfer (PET) in HEAN, 95 in oxazine 91, 111–12 in room-temperature ionic liquids, 331–5 photoinduced processes, 865–6 photoionization, 869 photoisomerization, adiabatic, 211–14 photolysis, 627 of hydrogen chloride, 880 of water, 627–8, 630–2 photomultiplier tubes (PMTs), 361 photoreactivity, of drugs, 354–5 photoreceptors, 840 see also photoactive yellow protein (PYP) photosensitizers, 355–8 photosynthetic reaction centres, 858 photosynthetic water splitting, 433–4 see also photosystem II (PS II) photosystem I (PS I), 434 photosystem II (PS II), 434 cofactor arrangement, 434–5 light-induced charge separation sequence, 437–40 overall reaction pattern, 434–5 oxidative water splitting sequence, 440–50 plastoquinol formation sequence, 451–2 thermal stability of, 436–7 phthalimide derivatives, 81, 92 pick-up technique, 867 PICT, see planar ICT pigment-protein complexes, 434, 858 PJT coupling, see pseudo-Jahn-Teller coupling pKa shifts, 609 and stabilization constant in metal complexes, 615–18 of carboxylic acid derivatives, 612 of phenols, 612–13
904
Index
of thiol derivatives, 610–12 planar ICT (PICT), 315 plastoquinol (PQH2), 434, 451–2 plastoquinone (PQ), 440, 451 PMPN, see N-(4-methoxyphenyl)-2,3-naphthalimide PMTs, see photomultiplier tubes polarizable continuum model (PCM), 108 poly(vinylpyrolidone)-SDS complexes, 741 POR, see protochlorophyllide oxidoreductase porphyrin complexes, 620 potential energy surfaces (PES), 579 PQ, see plastoquinone 1HPQ, see 1-H-pyrrolo[3,2-h]quinoline PQH2, see plastoquinol PRG, see proton release group PROPKA method, 824 11-propyl-6H-indolo-[2,3-b]quinoline (6HIQ), 559, 560–1 6HIQ/7AI heterodimer, 562 protochlorophyllide (Phlide) 858–9 enzyme-bound species, 861–2 isolated species, 860–1 Phlide-methanol complex, 860 protochlorophyllide oxidoreductase (POR), 857–8 catalytic mechanism of, 859 see also chlorophyllide (Chlide); protochlorophyllide (Phlide) proton-coupled electron transfer (PCET), 443, 447–8 proton-driven conformational switching, 613–15 proton-pumping, in rhodopsins, 380, 384–5, 386–8 proton release group (PRG), 384 proton transfer complex (PTC), betacarbolines 395, 415, 689–92, 695–8 ‘proton transfer fluorescence’, 185 proton transfer reactions, 463–7, 465–6, 555 adiabatic/nonadiabatic, 737, 738 and volume changes, 797–802 see also excited-state double proton transfer (ESDPT); excited-state intramolecular proton transfer (ESIPT); excited-state proton transfer (ESPT) proton tunnelling, 647 ‘proton wires’, 525 pseudo-Jahn-Teller (PJT) coupling, 315 PT reactions, see proton transfer reactions PTC, see proton transfer complex pump-probe spectroscopy, 4–5, 772–3 purple bacteria reaction centres (PBRCs), 437–40 PyIn-n, see 2-(20 -pyridyl)indoles PYP, see photoactive yellow protein 2PyPhBu, see 2-pyridylphenylbutadiene pyrazole clusters, 880–8 pyridine hydrogen-bond complexation, 52–3 interactions with betacarbolines, 405–14, 687–92 interactions with pyridylindoles, 678–9 interactions with pyrrolinoquinolines, 678–9 pyrido[2,3-a]carbazole (PC), 663, 665, 670, 678
pyrido[3,2-g]indole, 662–72 9H-pyrido[3,4-b]indole, see betacarboline 1-(2-pyridyl)-5-(4-dimethylaminophenyl)-penta-2,4-diene-1-one (DMAC) 86–9 7-(30 -pyridyl)indole, 91, 92 2-(20 -pyridyl)indoles (PyIn-n), 663, 665, 670, 673–9 fluorescence quenching by electron transfer, 678–80 2-pyridylphenylbutadiene (2PyPhBu), 212–14 pyrimidine, 117 pyrrole-water complexes, 633–6, 637 pyrroles, 110, 628 clusters, 880–8 1-H-pyrrolo[3,2-h]quinoline, 662 pyrroloquinolines, 662–73 fluorescence quenching by electron transfer, 678–80 tautomerization, 664–8, 670 QA molecules, 440 (3Q)2E, see 1,2-di(30 -quinolyl)ethane QM/MM, see quantum mechanics/molecular mechanics method quantum correction factors, 2, 19 quantum mechanics/molecular mechanics (QM/MM) method, 3, 579–80 for studying fluorescent proteins, 819, 824, 832–4 quenching circuit, 361 quinones, 199–202 radiation damage, 866 Raman spectroscopy, 342, 343–4 for studying hydrogen bonding in ionic liquids, 342 Stokes/anti-Stokes, 343–4 see also coherent anti-Stokes Raman scattering (CARS) reactive oxygen species (ROS), 358 red-edge excitation shift (REES), 159 red fluorescent proteins (RFPs), 819–20, 835 internal conversion mechanism, 830–2 QM/MM studies, 824, 832–4 redox reactions, in photosynthetic water splitting, 433–4 REES, see red-edge excitation shift Rehm-Weller behaviour, 335 Rehm-Weller equation, 332 rehybridization, 315 rehybridization by ICT (RICT), 315 resorufin, 776–7 reverse micelles (RMs), 218, 222–3 Arrhenius model at interface, 244–59 solvation dynamics, 220–1 structure of, 222–3 RFPs, see red fluorescent proteins rheopexy, 714 Rhodobacter (Rb.) sphaeroides, 438, 452, 525 rhodopsins, 378–9 proton-pump activity, 380, 384–5, 386–8 role of strong hydrogen bond of water, 379–80, 386–8 see also bacteriorhodopsin; halorhodopsin
Index RICT, see rehybridization by ICT RISM-SCF method 108 RMs, see reverse micelles room-temperature ionic liquids (RTILs), 160 microemulsion, 161–2, 166–7 mixed micelle, 162–3 neat, 161 solvation dynamics in, 160–3, 335–9 photoinduced electron transfer (PET) in, 331–5 room-temperature phosphorescence (RTP), 185 ROS, see reactive oxygen species RTILs, see room-temperature ionic liquids Rtms5, 819, 830, 832, 834 RTP, see room-temperature phosphorescence ruthenium complexes, 621 SCC-DFTB method, 824 SDS, see sodium dodecyl sulfate SED equation, see Stokes-Einstein-Debye equation serine-type protease, 443 serum albumin, 472 SFG, see sum-frequency generation SHNC, see 3-hydroxynaphthalene-2-carboxylate Silica encapsulation, 717, 718 single-photon avalanche diodes (SPADs), 361 single photon counting (SPC), 359 see also time-correlated single-photon counting (TCSPC) single-photon detectors, 361–2 Smoluchowski equation, 333 sodium deoxycholate (NaDC) FRET, 167 solvation dynamics, 163–5 sodium dodecyl sulfate (SDS), 220, 222, 224, 237–44, 739, 741 solute-solvent complexes, 105–7 solute-solvent hydrogen bond formation, 79–80, 82 changes in excited-state-properties, 109–15 characterizing hydrogen bonds, 115–17 combined experimental and theoretical approaches, 118 design of experiments, 98–104, 105 dual fluorescence, 95–8 fluorescence enhancement, 92–5 fluorescence quenching, 91–2 prerequisite conditions for, 80–2 solvatochromic analysis, 83–90 theoretical modelling, 104–17 solute-solvent interactions, 149, 177–8, 761–2 specific/nonspecific, 79–80, 761, 784–5 see also hydrogen-bonded complexes; solute-solvent hydrogen bond formation solvation, 761–2 ‘dipolar’, 761 versus hydrogen bond dynamics, 784–6 solvation dynamics, 160, 225–31
905
coupled into ESPT reaction, 567–74 in bile salt aggregate, 163–5 in biomimicking systems, 219–21 in confined systems, 738 in ionic liquids, 335–9 in nanocofined systems, 160–6 in P123 triblock copolymer, 165–6 in room-temperature ionic liquids, 160–3, 335–9 solvation energy, 226–7, 761 solvation interactions, see solute-solvent interactions solvatochromic effect, 79 solvent effects, 83–4, 177–8, 730–1 solvent friction, 786 solvent shift method, 83 solvents, 99–102 sound velocity, 236, 237 SPADs, see single-photon avalanche diodes SPC, see single photon counting spinach, 437, 445, 450 spontaneous Raman scattering, see Raman spectroscopy 3St-2AP, see 3-styryl-2-azaphenanthrene 3St-7AP, see 3-styryl-7-azaphenanthrene sterically hindered compounds, 314 Stern-Volmer constant, 489, 768 for excited states of photoproduct, 491–4 for 3-hydroxyflavones, 499–501, 503, 505 Stern-Volmer equation, 332, 489, 491, 499, 686, 687–8 stilbene-like molecules adiabatic photoisomerization, 211–14 IHB effects on conformational equilibria, 206–10 IHB effects on radiative and reactive relaxation, 210–11 Stokes-Einstein-Debye equation (SED equation), 236, 256, 257, 334–5 Stokes shift, 176 3StP, see 3-styrylphenanthrene 4StQ, see 4-styrylquinoline 8StQ, see 8-styrylquinoline streak cameras, 359 3-styryl-2-azaphenanthrene (3St-2AP), 206–7 3-styryl-7-azaphenanthrene (3St-7AP), 206–7 3-styrylphenanthrene (3StP), 206–7 4-styrylquinoline (4StQ), 210, 211 8-styrylquinoline (8StQ), 210, 211 sub-picosecond time-resolved transient spectroscopy, 846–8 sulfonamide, 807 sum-frequency generation (SFG), 343, 345, 349 supersonic expansions, 529 Synechocystis sp. PCC 6803, 449 TAC, see time-to-amplitude converter Tb(III) complexes, 618 TC, see thiocoumarin TCAA, see trichloroacetic acid TCSPC, see time-correlated single-photon counting
906
Index
TDDFT, see time-dependent density functional theory method TEMPO, see 2,2,6,6-tetramethylpiperidine 1-oxyl tetrahydrobetacarboline, 182, 183 3,4,5,6-tetrahydrobis(pyrido[3,2-g]indolo)[2,3-a:30 ,20 -j] acridine (TPIA), 566–7 7,8,9,10-tetrahydropyrido[2,3-a]carbazole (TPC) 663, 665, 666, 668, 670 2,2,6,6-tetramethylpiperidine 1-oxyl (TEMPO), 475, 480, 481, 482, 494–7, 499 TFE, see trifluoroethanol Thermosynechococcus (T.) elongatus, 445, 447, 451–2 thiocoumarin (TC), 112 thioketones, 103 thiols conformational switching, 613 pKa shifts, 610–12 thylakoids, 436–7, 445 thymidine, 200 thymine, 126, 127, 128, 130, 143 TICT, see twisted intramolecular charge transfer time-correlated single-photon counting (TCSPC), 359–62, 364–6 time-dependent density functional theory (TDDFT) method, 108, 149–51, 156 time-resolved emission spectra (TRES), 230–1 time-resolved fluorescence evaluation of drug photoreactivity, 354–7 techniques, 358–62 time-resolved IR absorption spectroscopy, 786–90 ultrafast, 762, 790, 791–2 time-to-amplitude converter (TAC), 360, 362 T-MeHN, see N2-methylharmane TNS, see 2,6-toluidinonaphthalene sulfonate 2,6-toluidinonaphthalene sulfonate (TNS), 248 TPC, see 7,8,9,10-tetrahydropyrido[2,3-a]carbazole TPIA, see 3,4,5,6-tetrahydrobis(pyrido[3,2-g]indolo)[2,3a:30 ,20 -j]acridine TRES, see time-resolved emission spectra 1,2,3-triazine-water complex, 117 triblock copolymers interaction with ionic liquids, 162–3 solvation dynamics in, 165–6 trichloroacetic acid (TCAA), 101–2 trifluoroethanol (TFE), 763 interactions with betacarbolines, 692 interactions with fluorenone, 780–1 N,N,N-trimethyl-N-propylammonium bis(trifluoromethanesulfonyl) imide ([N3[1]][Tf2N]), 335–9 triplet electronic excited-states, 73–6, 149–56 tryptophan, 89–90, 179 turmeric, 357 twist-boat structure, 613 twisted intramolecular charge transfer (TICT) and dual fluorescence, 313–16 formation of 80, 81, 176, 183–4
in coumarins, 424–9 in DMABN, 318–20 in DMAPPI, 321–7 in fluorescent proteins, 831 TICT-forming compounds, 112 two-colour resonant two-photon ionization (2CR2PI), 528–9 tyrosine D, 445–6 tyrosine Z 441–5 ubiquinol, 451, 452 ubiquinone, 440 UMP2-BOMD simulations, 631–2 uracil, 112, 114–15, 130, 143 see also adenine-uracil base pairs urea, 807–8 recognition, 567 UV-IR DR spectroscopy, 29–30, 37 UV-UV depletation, 528, 529 UV-UV hole burning, 528, 529 van’t Hoff equation, 42 Vavilov’s rule, 469 VER, see vibrational energy relaxation vesicles, 712, 714–15, 716–18 vibrational energy relaxation (VER), 29–30, 37, 786–90 bare 2-naphthol, 31 H-bonded clusters of 2-naphthol, 31–6 ionic liquids, 341–9 viscosity and solvation in ionic liquids, 336–8, 339 shear-induced, 719–21 volume changes, 797–802 wagging ICT (WICT), 315 water ‘biological’, 218 deuterated, 103 dynamics at biological interfaces, 217–8, 221–2 in biomimicking systems, 218–23, 237–60 in rhodopsins, 378–88 ion solvation mechanism, 226 photolysis of, 627–8, 630–2 pure, 217–18 solvation time correlation function, 229 see also photodissociation; photosynthetic water splitting water-oxidizing complexes (WOCs), 441 effect on hydrogen bonding of YD in PS II, 445 effect on hydrogen bonding of YZ in PS II, 442–5 oxidation of, 446–51 Watson-Crick base pairs, 2, 9, 126, 128 excited-state properties, 138–42 reverse, 129 see also nucleic acid base pairs WICT, see wagging ICT
Index wobbling-in-cone analysis, 235–6 WOCs, see water-oxidizing complexes wtGFP, see green fluorescent protein (GFP), wild type
YD, 445–6 YZ, 441–5 Y42F, 850–1
xanthone, 184–5
zigzag-chain structure, 623, 624
907