ELECTROCHEMISTRY OF FUNCTIONAL SUPRAMOLECULAR SYSTEMS Edited by
Paola Ceroni Alberto Credi Margherita Venturi
The Wiley Series on Electrocatalysis and Electrochemistry
ELECTROCHEMISTRY OF FUNCTIONAL SUPRAMOLECULAR SYSTEMS
WILEY SERIES ON ELECTROCATALYSIS AND ELECTROCHEMISTRY Andrzej Wieckowski, Series Editor
Fuel Cell Catalysis: A Surface Science Approach, Edited by Marc T. M. Koper Electrochemistry of Functional Supramolecular Systems, Paola Ceroni, Alberto Credi, and Margherita Venturi Catalysis in Electrochemistry: From Fundamentals to Strategies for Fuel Cell Development, Elizabeth Santos and Wolfgang Schmickler Fuel Cell Science: Theory, Fundamentals, and Biocatalysis, Andrzej Wieckowski and Jens Norskov
ELECTROCHEMISTRY OF FUNCTIONAL SUPRAMOLECULAR SYSTEMS Edited by
Paola Ceroni Alberto Credi Margherita Venturi
The Wiley Series on Electrocatalysis and Electrochemistry
Copyright 2010 by John Wiley & Sons, Inc. All rights reserved. Published by John Wiley & Sons, Inc., Hoboken, New Jersey Published simultaneously in Canada No part of this publication may be reproduced, stored in a retrieval system, or transmitted in any form or by any means, electronic, mechanical, photocopying, recording, scanning, or otherwise, except as permitted under Section 107 or 108 of the 1976 United States Copyright Act, without either the prior written permission of the Publisher, or authorization through payment of the appropriate per-copy fee to the Copyright Clearance Center, Inc., 222 Rosewood Drive, Danvers, MA 01923, (978) 750-8400, fax (978) 750-4470, or on the web at www.copyright.com. Requests to the Publisher for permission should be addressed to the Permissions Department, John Wiley & Sons, Inc., 111 River Street, Hoboken, NJ 07030, (201) 748-6011, fax (201) 748-6008, or online at http://www.wiley.com/go/permissions. Limit of Liability/Disclaimer of Warranty: While the publisher and author have used their best efforts in preparing this book, they make no representations or warranties with respect to the accuracy or completeness of the contents of this book and specifically disclaim any implied warranties of merchantability or fitness for a particular purpose. No warranty may be created or extended by sales representatives or written sales materials. The advice and strategies contained herein may not be suitable for your situation. You should consult with a professional where appropriate. Neither the publisher nor author shall be liable for any loss of profit or any other commercial damages, including but not limited to special, incidental, consequential, or other damages. For general information on our other products and services, or technical support, please contact our Customer Care Department within the United States at (800) 762-2974, outside the United States at (317) 572-3993 or fax (317) 572-4002. Wiley also publishes its books in a variety of electronic formats. Some content that appears in print may not be available in electronic books. For more information about Wiley products, visit our web site at www.wiley.com Library of Congress Cataloging-in-Publication Data: ISBN 9780470255575 Printed in the United States of America 10 9 8
7 6 5 4
3 2 1
CONTENTS
Preface to the Wiley Series on Electrocatalysis and Electrochemistry
vii
Foreword
ix
Preface
xi
Contributors 1. Electrochemically Controlled H-Bonding
xiii 1
Diane K. Smith
2. Molecular Motions Driven by Transition Metal Redox Couples: Ion Translocation and Assembling– Disassembling of Dinuclear Double-Strand Helicates
33
Valeria Amendola and Luigi Fabbrizzi
3. Molecular Encapsulation of Redox-Active Guests
59
Angel E. Kaifer
4. Dendritic Encapsulation of Redox-Active Units
87
Christopher B. Gorman
5. Redox-Active Metal–Polypyridine Dendrimers as Light-Harvesting Antennae
121
Fausto Puntoriero, Scolastica Serroni, Francesco Nastasi, and Sebastiano Campagna
6. Dendrimers as Multielectron Storage Devices
145
Paola Ceroni and Margherita Venturi
7. Self-assembled Monolayers and Multilayers of Electroactive Thiols
185
Ibrahim Yildiz, Fran¸cisco M. Raymo and Massimiliano Lamberto
v
vi
CONTENTS
8. Electrochemistry of Carbon Nanoparticles
201
Luis Echegoyen, Amit Palkar, and Frederic Melin
9. Molecular Devices Based on Fullerenes and Carbon Nanotubes
229
Matteo Iurlo, Demis Paolucci, Massimo Marcaccio, and Francesco Paolucci
10. Functional Electroactive Biomolecules
261
Xiaomin Bin, Piotr Michal Diakowski, Kagan Kerman, Heinz-Bernhard Kraatz
11. Functional Nanoparticles as Catalysts and Sensors
301
Brian J. Jordan, Chandramouleeswaran Subramani, and Vincent M. Rotello
12. Biohybrid Electrochemical Devices
333
Ran Tel-Vered, Bilha Willner, and Itamar Willner
13. Electroactive Rotaxanes and Catenanes
377
Alberto Credi and Margherita Venturi
14. Electrochemically Driven Molecular Machines Based on Transition-metal Complexed Catenanes and Rotaxanes
425
Jean-Paul Collin, Fabien Durola, and Jean-Pierre Sauvage
15. Electroactive Molecules and Supramolecules for Information Processing and Storage
447
Guanxin Zhang, Deqing Zhang, and Daoben Zhu
16. Electrochemiluminescent Systems as Devices and Sensors
477
Andrzej Kapturkiewicz
17. Recent Developments in the Design of Dye-Sensitized Solar Cell Components
523
Stefano Caramori and Carlo Alberto Bignozzi
Index
581
PREFACE to the Wiley Series on Electrocatalysis and Electrochemistry
This series covers recent advances in electrocatalysis and electrochemistry and depicts prospects for their contribution to the present and future of the industrial world. It illustrates the transition of electrochemical sciences from a solid chapter of physical electrochemistry (covering mainly electron transfer reactions, concepts of electrode potentials, and structure of the electrical double layer) to the field in which electrochemical reactivity is shown as a unique chapter of heterogeneous catalysis, is supported by high-level theory, connects to other areas of science, and includes focus on electrode surface structure, reaction environment, and interfacial spectroscopy. The scope of this series ranges from electrocatalysis (practice, theory, relevance to fuel cell science and technology) to electrochemical charge transfer reactions, biocatalysis, and photoelectrochemistry. While individual volumes may look quite diverse, the series promises updated and overall synergistic reports on insights into further the understanding of properties of electrified solid/liquid systems. Readers of the series will also find strong reference to theoretical approaches for predicting electrocatalytic reactivity by such high-level theories as DFT. Beyond the theoretical perspective, further vehicles for growth are the sound experimental background and demonstration of significance of such topics as energy storage, syntheses of catalytic materials via rational design, nanometer-scale technologies, prospects in electrosynthesis, new instrumentation, surface modifications in basic research on charge transfer, and related interfacial reactivity. In this context, readers will notice that new methods that are being developed for a specific field may be readily adapted for application in others. Electrochemistry has benefited from numerous monographs and review articles due to its unique character and significance in the practical world (including electroanalysis). Electrocatalysis has also been the subject of individual reviews and compilations. The Wiley Series on Electrocatalysis and Electrochemistry is dedicated to the current activity by focusing each volume on a specific topic of choice. The chapters also demonstrate electrochemistry’s connections to other areas of chemistry and physics, such as biochemistry, chemical engineering, quantum mechanics, chemical physics, surface science, and biology, and illustrate the wide range of literature that each topic contains. While the title of each volume informs of the specific focus chosen by the volume editors and chapter authors, the integral outcome offers a broad-based analysis of the total development of the field. The progress of the series will provide a global definition of what electrocatalysis and electrochemistry are concerned with now and how they evolve with time. The purpose is manifold, vii
viii
PREFACE TO THE WILEY SERIES ON ELECTROCATALYSIS AND ELECTROCHEMISTRY
mainly to provide a modern reference for graduate instruction and for active researchers in the two disciplines, as well as to document that electrocatalysis and electrochemistry are dynamic fields that expand rapidly and likewise rapidly change in their scientific profiles. Creation of each volume required the editors involvement, vision, enthusiasm, and time. The Series Editor thanks all Volume Editors who graciously accepted his invitations. Special thanks are for Ms. Anita Lekhwani, the Series Acquisition Editor, who extended the invitation to the Series Editor and is a wonderful help in the Series assembling process. ANDRZEJ WIECKOWSKI Series Editor
FOREWORD
Like the currently popular area, called “nanoscience”, the field of “supramolecular chemistry” has rather hazy boundaries. Indeed, both areas now share much common ground in terms of the types of systems that are considered. From the beginning, electrochemistry, which provides a powerful complement to spectroscopic techniques, has played an important role in characterizing such systems and this very useful book goes considerably beyond the volume on this same topic by Kaifer and Go´mezKaifer that was published about 10 years ago. Some of the “classic” supramolecular chemistry topics such as rotaxanes, catenanes, host–guest interactions, dendrimers, and self-assembled monolayers remain, but now with important extensions into the realms of fullerenes, carbon nanotubes, and biomolecules, like DNA. These topics lead to considerations of supramolecular devices, for example for use as sensors, and to molecular machines. Not only is electrochemistry an excellent way of characterizing such systems, for example, via cyclic voltammetry, but in the world of molecular machines, it is also the most straightforward approach to providing the energy to power such devices. These topics then naturally lead to consideration of the conversion of electricity to light (electrochemiluminescence) and light to electricity (dye-sensitized solar cells) via electrochemical devices. While the latter are not fundamentally supramolecular systems, their design could certainly benefit from the considerations in the very detailed and authoritative treatments in this volume. The idea of integrated chemical systems, based on nanoscience and nanotechnology, was proposed a little over 20 years ago and was the subject of my 1994 monograph, but, so far, few such systems have reached widespread practical utilization. Nevertheless, the principles of such systems, for example for synthesis, analysis, and perhaps computation remain of interest, and supramolecular electrochemistry can play a major role in their development. I hope this important volume will go a long way toward introducing such principles to a wide audience, especially to the young people who are less burdened by impressions of what is impossible. The University of Texas at Austin
Allen J. Bard
ix
PREFACE
Supramolecular chemistry is a highly interdisciplinary field that has been developed at an astonishingly fast rate during the last three decades. In a historical perspective, as pointed out by Jean-Marie Lehn, supramolecular chemistry originated from Paul Ehrlichs receptor idea, Alfred Werners coordination chemistry, and Emil Fischers lock-and-key image. It was only after 1970, however, that fundamental concepts such as molecular recognition, preorganization, self-assembly, and self-organization were introduced to chemistry; supramolecular chemistry then began to emerge as a welldefined discipline and was consecrated by the award of the Nobel Prize in Chemistry to Charles Pedersen, Donald Cram, and Jean-Marie Lehn in 1987. Supramolecular chemistry, according to its most popular definition, is “the chemistry beyond the molecule, bearing on organized entities of higher complexity that result from the association of two or more chemical species held together by intermolecular forces.” As the field developed, it became evident that a definition strictly based on the nature of the bond that links the components would be limiting. Many scientists, therefore, started to distinguish between what is molecular and what is supramolecular based on the degree of intercomponent interactions. In a general sense, one can say that with supramolecular chemistry there has been a shift in focus from molecules to molecular assemblies or multicomponent structures driven by the emergence of new functions. In the frame of research on supramolecular systems, the idea began to arise in a few laboratories that the concepts of “device” and “machine” could be applied at the molecular level. In other words, molecules might be used as building blocks for the assembly of multicomponent structures exhibiting novel and complex functions that arise from the cooperation of simpler functions performed by each component. This strategy, encouraged by a better understanding of biomolecular devices, has been implemented on a wide variety of chemical systems, leading to highly interesting results. As a matter of fact, the molecular bottom-up construction of nanoscale devices and machines has become one of the most stimulating challenges of nanoscience. Such achievements have been made possible because of the substantial progresses obtained in other areas of chemistry and physics—particularly concerning the synthesis and characterization of complex chemical systems, and the study of surfaces and interfaces. In this perspective, electrochemistry is a very powerful tool not only for characterizing a supramolecular system, but also for operating the device. Indeed, molecular devices, as their macroscopic counterparts, need energy to operate and signals to communicate with the operator. Electrochemistry can be an interesting xi
xii
PREFACE
answer to this dual requirement: it can be used to supply the energy needed to make the system work, and, by means of the various electrochemical techniques (e.g., voltammetry), it can also be used to “read” the state of the system, controlling and monitoring the operation performed by the device. Furthermore, electrodes represent one of the best ways to interface molecular-level systems to the macroscopic world, a feature that is important for future applications. Hence, it is not surprising that the marriage of electrochemistry and supramolecular chemistry has produced a wealth of very interesting devices and functions, thereby generating new scientific knowledge and raising expectations for practical applications in energy conversion, information and communication technologies, advanced materials, diagnostics, and medicine. Our aim with this book is to provide the reader with an overview of current electrochemical research applied to multicomponent chemical systems, with particular attention to properties and functions, and to strengthen the contacts between the electrochemical community and the researchers engaged in the field of nanoscience. Although the text covers a wide range of topics with contributions from leading authorities in their respective fields, it does not even attempt to be a comprehensive book on supramolecular electrochemistry. Rather, we would like to give the reader a flavor of the level of creativity and ingenuity reached by scientists working in this area. We hope that the book will be useful as a reference not only for experienced researchers, but also for graduate students and postdoctoral fellows who are interested in exploring electrochemistry at its frontiers with supramolecular chemistry, materials science, and biochemistry. It may also be a useful complement for students attending nanoscience and nanotechnology courses. The 17 chapters of the book are not grouped in sections but they are somehow logically ordered. The initial contributions, describing basic science investigations on systems in solution, are followed by chapters dealing with less conventional multicomponent architectures and/or environments. The final part contains contributions on devices and systems of high complexity and/or applicative interest. Although the book does not include introductory or tutorial sections, most chapters begin with a discussion of the basic concepts that are relevant for the presented topics. We hope that these sections will make the book comprehensible also to nonspecialists. We would like to express our gratitude to the distinguished colleagues and friends who contributed the chapters: their commitment is indeed a fundamental ingredient of this initiative. We also thank people at Wiley for their assistance during the various phases of the editorial work. Finally, we would like to thank our families because their love and patience are an invaluable support for our professional activity. Bologna, June 2009
PAOLA CERONI ALBERTO CREDI MARGHERITA VENTURI
CONTRIBUTORS
Valeria Amendola, Dipartimento di Chimica Generale, Universita di Pavia, Pavia, Italy Carlo Alberto Bignozzi, Dipartimento di Chimica, Universita di Ferrara, Ferrara, Italy Xiaomin Bin, Department of Chemistry, The University of Western Ontario, London, Ontario, Canada Sebastiano Campagna, Dipartimento di Chimica Inorganica, Chimica Analitica e Chimica Fisica, Universit a di Messina, Messina, Italy Stefano Caramori, Dipartimento di Chimica, Universita di Ferrara, Ferrara, Italy Paola Ceroni, Dipartimento di Chimica “G. Ciamician”, Alma Mater Studiorum Universit a di Bologna, Bologna, Italy Jean-Paul Collin, Laboratoire de Chimie Organo-Minerale, UMR 7177 du CNRS, Faculte de Chimie, Universite de Strasbourg, Strasbourg Cedex, France Alberto Credi, Dipartimento di Chimica “G. Ciamician”, Alma Mater Studiorum, Universit a di Bologna, Bologna, Italy Piotr Michal Diakowski, Department of Chemistry, The University of Western Ontario, London, Ontario, Canada Fabien Durola, Laboratoire de Chimie Organo-Minerale, UMR 7177 du CNRS, Faculte de Chimie, Universite de Strasbourg, Strasbourg Cedex, France Luis Echegoyen, Department of Chemistry, Clemson University, Clemson, SC, USA Luigi Fabbrizzi, Dipartimento di Chimica Generale, Universita di Pavia, Pavia, Italy Christopher B. Gorman, Department of Chemistry, North Carolina State University, Raleigh, NC, USA Matteo Iurlo, Dipartimento di Chimica “G. Ciamician”, Alma Mater Studiorum, Universit a di Bologna, Bologna, Italy Brian J. Jordan, Department of Chemistry, University of Massachusetts, Amherst, MA, USA xiii
xiv
CONTRIBUTORS
Angel E. Kaifer, Center for Supramolecular Science, Department of Chemistry, University of Miami, Coral Gables, FL, USA Andrzej Kapturkiewicz, Institute of Physical Chemistry, Polish Academy of Sciences, Warsaw, Poland and Institute of Chemistry, University of Podlasie, Siedlce, Poland Kagan Kerman, Department of Chemistry, The University of Western Ontario, London, Ontario, Canada Heinz-Bernhard Kraatz, Department of Chemistry, The University of Western Ontario, London, Ontario, Canada Massimiliano Lamberto, Department of Chemistry, Medical Technology and Physics, Monmouth University, West Long Branch, NJ, USA Massimo Marcaccio, Dipartimento di Chimica “G. Ciamician”, Alma Mater Studiorum, Universit a di Bologna, Bologna, Italy Frederic Melin, Department of Chemistry, Clemson University, Clemson, SC, USA Francesco Nastasi, Dipartimento di Chimica Inorganica, Chimica Analitica e Chimica Fisica, Universit a di Messina, Messina, Italy Amit Palkar, Department of Chemistry, Clemson University, Clemson, SC, USA Demis Paolucci, Dipartimento di Chimica “G. Ciamician”, Alma Mater Studiorum, Universit a di Bologna, Bologna, Italy Francesco Paolucci, Dipartimento di Chimica “G. Ciamician”, Alma Mater Studiorum, Universit a di Bologna, Bologna, Italy Fausto Puntoriero, Dipartimento di Chimica Inorganica, Chimica Analitica e Chimica Fisica, Universit a di Messina, Messina, Italy Fran¸cisco M. Raymo, Department of Chemistry, University of Miami, Coral Gables, FL, USA Vincent M. Rotello, Department of Chemistry, University of Massachusetts, Amherst, MA, USA Jean-Pierre Sauvage, Laboratoire de Chimie Organo-Minerale, UMR 7177 du CNRS, Faculte de Chimie, Universite de Strasbourg, Strasbourg Cedex, France Scolastica Serroni, Dipartimento di Chimica Inorganica, Chimica Analitica e Chimica Fisica, Universit a di Messina, Messina, Italy Diane K. Smith, Department of Chemistry and Biochemistry, San Diego State University, San Diego, CA, USA Chandramouleeswaran Subramani, Department of Chemistry, University of Massachusetts, Amherst, MA, USA
CONTRIBUTORS
xv
Margherita Venturi, Dipartimento di Chimica “G. Ciamician”, Alma Mater Studiorum, Universit a di Bologna, Bologna, Italy Ran Tel-Vered, Institute of Chemistry, The Hebrew University of Jerusalem, Jerusalem, Israel Bilha Willner, Institute of Chemistry, The Hebrew University of Jerusalem, Jerusalem, Israel Itamar Willner, Institute of Chemistry, The Hebrew University of Jerusalem, Jerusalem, Israel Ibrahim Yildiz, Department of Chemistry, University of Miami, Coral Gables, FL, USA Deqing Zhang, Beijing National Laboratory for Molecular Sciences, Organic Solids Laboratory, Institute of Chemistry, Chinese Academy of Sciences, Beijing, China Guanxin Zhang, Beijing National Laboratory for Molecular Sciences, Organic Solids Laboratory, Institute of Chemistry, Chinese Academy of Sciences, Beijing, China Daoben Zhu, Beijing National Laboratory for Molecular Sciences, Organic Solids Laboratory, Institute of Chemistry, Chinese Academy of Sciences, Beijing, China
CHAPTER 1
Electrochemically Controlled H-Bonding DIANE K. SMITH Department of Chemistry and Biochemistry, San Diego State University, San Diego, CA, USA
1.1
INTRODUCTION
Due to their strength and directionality, hydrogen bonds are one of the most important and useful types of intermolecular interactions available for the construction of supramolecular complexes. The iconic examples of DNA base pairing and the formation of secondary structure in proteins provide ample proof of their utility for the assembly of well-defined, functional structures. Examples1 of the use of hydrogen bonds in synthetic, solution-phase supramolecular chemistry range from H-bonded dimers2 held together by up to 6 H-bonds3 to large “rosette” assemblies constructed from up to 15 components and 72 H-bonds.4 A wide variety of open H-bonded structures have also been prepared, including those that self-assemble into capsules of various sizes and shapes5,6 and cyclic peptides that assemble into hollow tubes.7 Although, from a purely chemical point of view, learning how to create these complicated supramolecular structures has its own value, there are plenty of more practical reasons to investigate this chemistry. In the short term, these include catalysis and sensor applications, and in the long term, molecular electronics and molecular machines. With perhaps the exception of catalysis, all these applications will require some sort of signal transduction to allow for communication with the supramolecular device. This, of course, is one of the main reasons that electrochemistry is useful for supramolecular chemistry. Electron transfer provides a wellunderstood and very sensitive method to both communicate with supramolecular assemblies and control their structure.8 However, although electrochemistry can be used for the above, it will do so only if this functionality is designed into the structure. At minimum, two requirements Electrochemistry of Functional Supramolecular Systems. Edited by Paola Ceroni, Alberto Credi, and Margherita Venturi Copyright 2010 John Wiley & Sons, Inc.
1
2
ELECTROCHEMICALLY CONTROLLED H-BONDING
must be met. First, a reversible redox couple must be present as part of the structure. Reversible in this context means that both oxidation states are chemically stable under the experimental conditions used and that the electron transfer kinetics are reasonably fast. Second, reduction or oxidation of the redox couple must significantly perturb the strength of important binding interactions holding the assembly together. The most straightforward type of interaction to perturb electrochemically are ion–ion interactions, but the electrostatic nature of a hydrogen bond makes it a close second, while also allowing for neutral molecules as binding partners. In this chapter, the basic principles behind electrochemically controlled H-bonding will first be described, along with some simple, illustrative examples and a brief discussion of the use of cyclic voltammetry to characterize such systems. Next, some general considerations regarding the design of these systems are discussed: the properties of the redox couple, the structures of host and guest, the choice of solvent and electrolyte, and the possibility for proton transfer. Finally, a selection of electrochemically controlled H-bonding systems will be described, organized by the nature of the binding partner and the type of redox couple.
1.2
BASIC PRINCIPLES
A H-bond is a favorable interaction formed between a relatively positively charged hydrogen atom in a polar bond, XH, and either a lone pair on a relatively negatively charged atom, Y (Eq. 1.1), or a highly polarizable pi-bonding electron pair (Eq. 1.2).9 As this description implies, there is a high degree of electrostatic character associated with H-bonding, although only the very weakest H-bonds are purely electrostatic. Strong H-bonds range in strength from 15 to 40 kcal/mol and are considered to be mainly covalent in character. These bonds are characterized by close to linear XHY bond angles and XY distances that are substantially smaller than the sum of the van der Waals radii. These are typically formed from ionic species, for example, N þ H or O þ H as the H-donor and/or F, O, or N as the H-accepting atom. Moderate strength H-bonds range in strength from 4 to 15 kcal/mol and are mostly electrostatic in character. They can show a greater range in bond angles, from 130 to 180 , and will have bond lengths that are slightly smaller than the sum of the van der Waals radii. Typical examples are H-bonds formed from neutral oxygen and nitrogen functional groups, for example, NH and OH as the donor group and uncharged N and O as the accepting atoms. Weak H-bonds, those less than 4 kcal/mol in strength, are almost purely electrostatic in character and are characterized by a range of bond angles and XY distances that may be greater than the sum of the van der Waals radii. These are formed from the weaker H-donors such as CH and/or weaker acceptors such as pi electron pairs. +
X H
+
+
Y R
X H
Y R
ð1:1Þ
1.2
+
+
X H
1.2.1
+
+
R Y Y R
X H
3
BASIC PRINCIPLES
R Y Y R
ð1:2Þ
Direct Perturbation of Hydrogen Bonds
Although, from the above discussion, H-bonds are generally not purely electrostatic, a favorable electrostatic interaction plays an important role in all types of H-bonds. This provides a very simple way to think about how to use electron transfer to directly perturb the strength of H-bonds. There are two main ways to do this as shown in Scheme 1.1. The first is to make a H-acceptor a better acceptor by using a reduction reaction to increase the negative charge on the H-accepting atom (Scheme 1.1a). The second is to make the H-donor a better donor by using oxidation to increase the positive charge on a H-donating functional group (Scheme 1.1b). Alternatively, electron transfer can be used to weaken H-bonds through the opposite effects, making a H-donor a weaker donor through reduction or making a H-acceptor a weaker acceptor through oxidation. A simple example of reduction-based, electrochemically controlled H-bonding is provided by nitrobenzenes.10 Despite the standard Lewis structure that places a formal negative charge on one of the oxygens and a positive charge on the nitrogen, nitro groups are generally weak H-acceptors in solution because the NO bond is not very polar. However, reduction of an aromatic nitro compound to its radical anion greatly increases the negative charge on the oxygens, resulting in much stronger H-bonding to a H-donating guest. In the case of nitroaniline, 1 (X ¼ NH2), with 1,3-diphenylurea, 2 (Eq. 1.3), the equilibrium constant for H-bonding in 0.1 M NBu4PF6/DMF goes from <1 in the zero state to 8 104 M1 in the radical anion state.
O N O
X
O N O
+ e– X
O N O
+2 X
1
Ph H N O H N 2 Ph
ð1:3Þ
Y (a)
(b)
+ X H
+
+ e–
Y
– e–
X H
HX
HX
YR
YR
Scheme 1.1 Methods to increase H-bond strength using redox reactions.
4
ELECTROCHEMICALLY CONTROLLED H-BONDING
Oxidation reactions can also be used to control H-bonding. These have mainly been successful with anionic guests, but a recent example involving the dimethylaminophenylurea 3 gives a large binding enhancement with the cyclic diamide 4 (Eq. 1.4).11 In this case, reversible oxidation of the dimethylaminophenyl group increases the positive charge on one of the urea NHs, greatly increasing its H-donating ability to a good H-accepting guest such as 4. This results in the binding constant in 0.1 M NBu4B(C6F5)4/ CH2Cl2 increasing from 60 M1 in the reduced state to 2 105 M1 in the oxidized state. Me Me N
Me Me N
N H O
N H
– e– O
Ph 3
+4
N H
O
N H
O
Me N
ð1:4Þ
O
N H
1.2.2
Me Me N
N H Ph
N Me
Ph 4
Indirect Perturbation of Hydrogen Bonds
An alternative strategy to control H-bonding electrochemically does not rely on the electroactive unit playing a direct role in the H-bonding, but rather uses it to create additional favorable or unfavorable interactions that either strengthen or weaken an assembly held together through H-bonds. For example, electron transfer can be used to create charged sites that break apart a H-bonded dimer due to electrostatic repulsion. While the focus of this chapter will be on the direct perturbation methods, a well-characterized example of the indirect method will be described later on.
1.3 DETECTION AND CHARACTERIZATION OF ELECTROCHEMICALLY CONTROLLED H-BONDING The most important and useful technique for studying electrochemically controlled H-bonding and other forms of redox-dependent binding is cyclic voltammetry (CV). By observing how the voltammetry of the electroactive component changes in the presence of possible binding partners, one can readily determine whether redoxdependent binding is occurring and which oxidation state binds the strongest. More careful analysis of this type of data can yield the equilibrium constants for binding and possibly kinetic data as well. Typically in supramolecular chemistry, the term “host” refers to the larger and more structurally complex of two binding partners, while the term “guest” refers to the smaller, less complex binding partner. However, for the purpose of this discussion, the term “host” will be used to refer to the electroactive binding partner and “guest” to refer to the nonelectroactive binding partner, irrespective of their size or structural complexity.
1.3
DETECTION AND CHARACTERIZATION
5
+ e– H
Kox
H
EH +G
Kred
+G
+ e– HG
Scheme 1.2
EHG
HG
Equilibria involved in redox-dependent host–guest binding.
With the above definitions, the equilibria involved in the redox-dependent formation of a 1:1 host–guest complex can be described by the square shown in Scheme 1.2. In this representation, the electron transfer reaction of the host by itself is shown on the top horizontal axis and that of the host–guest complex on the bottom axis. Binding of the guest to the oxidized host is on the left vertical axis and binding to the reduced host is on the right. EH is the standard electrode potential for the host by itself and EHG is that of the host–guest complex. Kox is the binding constant of the guest to the host in its oxidized form and Kred is that in the reduced form. Qualitatively, if a guest binds more strongly to the oxidized form of the host, the oxidized host will be stabilized and it will be harder to reduce the host in the presence of the guest. This means that EHG is negative of EH. On the other hand, if the guest binds more strongly to the reduced form, it will be easier to reduce the host in the presence of the guest and EHG is positive of EH. Quantitatively, it is straightforward to show that if electron transfer and binding are fast and reversible, the four-membered square behaves as a one-electron redox couple with an E that depends on the true E values and the K values (Eq. 1.5). Note that in this equation [G] ¼ the concentration of the free (unbound) guest, which is not equal to the added guest if binding occurs. However, if [G] is greater than 10 times [H], it is reasonable to assume that [G] is approximately equal to the added guest concentration. When Kox/red[G] 1 (large [G] and/or large K), Equation 1.5 reduces to the more often seen Equation 1.6, which relates the maximum change in E to the ratio of binding constants or binding enhancement. A particularly handy form of this equation can be derived by switching to regular logarithms and filling in the constants to give Equation 1.7, which says that at 25 C each 60 mV shift in E corresponds to a 10-fold difference in binding strength between oxidation states. RT 1 þ Kred ½G ð1:5Þ ln Eobs ¼ EH þ F 1 þ Kox ½G RT Kred DEmax ¼ EHG EH ¼ ð1:6Þ ln Kox F 10DEmax =60 mV ¼
Kred Kox
at 25 C
ð1:7Þ
In general, there are two types of limiting voltammetric behavior observed for redox-dependent receptors.12 First, if there is strong binding in both oxidation
6
ELECTROCHEMICALLY CONTROLLED H-BONDING
states (Kox and Kred are both large), then addition of a half equivalent of the guest results in the current for the original CV wave decreasing by half and the appearance of a new CV wave of approximately equal height. The half-wave potential, E1/2, for the original CV wave should be approximately equal to EH, and that of the new wave approximately equal to EHG. Under these circumstances, after addition of 1 equivalent of the guest, only the new CV wave remains, now with a height equal to that of the original host-only wave. Further addition of guest produces no additional changes, unless greater than 1:1 guest–host binding is possible. In contrast, if the binding constants are small or there is only strong binding in one oxidation state, then generally one CV wave is observed even with less than 1 equivalent of guest. This is the behavior commonly observed for redox-dependent H-bonding systems. As the guest concentration is increased, the E1/2 of the CV wave will shift from EH toward EHG. An example of this typical behavior is shown in Fig. 1.1 with CVs of nitroaniline, 1 (X ¼ NH2), in the presence of increasing amounts of diphenylurea, 2.10 As discussed previously, diphenylurea is expected to bind more strongly to the reduced nitroaniline (Eq. 1.3), making it easier to reduce and resulting in a positive shift in the E1/2. This is indeed what is observed. Scan (a) is that of nitroaniline by itself. Addition of half equivalent of diphenylurea, scan (b), results in a broad wave, with a new shoulder appearing at more positive potentials. However, unlike the case in which strong binding is observed in both oxidation states, the position of the new shoulder does not correspond to the maximum shift. With 1 equivalent of diphenylurea, scan (c), the
Figure 1.1 CVs of p-nitroaniline in 0.1 M NBu4PF6/DMF in the presence of different amounts of 1,3-diphenylurea: (a) 0 mM urea, (b) 0.5 mM urea, (c) 1 mM urea, and (d) 10 mM urea. 500 mV/s scan rate.10
1.3
DETECTION AND CHARACTERIZATION
7
wave sharpens up and moves farther positive. Further additions of diphenylurea result in continued positive shifts in the CV wave. It is often assumed that if a significant excess of G is added (10–50-fold), then E1/2 of the CV wave will be EHG, and the binding enhancement can be calculated using Equation 1.6. However, this is not necessarily the case. At the very least, the assumption that saturation binding has been reached should be checked by doubling the concentration of guest and making sure no further change in E1/2 is observed. A more accurate method to determine the binding constants is to do a complete binding titration where the CVs are recorded at a range of guest concentrations as shown in Fig. 1.1. If the CVs are uncomplicated, a smoothly shifting wave with no significant change in wave shape, then it may be possible to obtain binding constants by plotting E1/2 versus [G] and fitting the curves to Equation 1.5 using a nonlinear least squares regression method. This can also be done for cases where more than 1:1 guest–host binding is possible.13,14 While the above technique can be used in many cases, it does require uncomplicated voltammetry and that significant E1/2 shifts are observed at guest concentrations >10 times the host concentration. If these conditions are not met, then an alternative strategy is needed. The most powerful is to use CV simulation software to fit the experimental CVs to the square scheme or a more complicated mechanism if necessary. This method allows determination of the thermodynamic parameters and possibly the kinetic parameters as well. An example of the use of CV simulation to determine binding constants and explain more complicated CV behavior is shown in Fig. 1.2. These are CVs observed
Figure 1.2 Experimental (lines) and simulated (dots) CVs of 1 mM 3 in 0.1 M NBu4B(C6F5)4/ CH2Cl2 in the presence of different amounts of 4: (a) 0 mM, (b) 0.5 mM, (c) 1 mM, and (d) 100 mM. 500 mV/s scan rate.11
8
ELECTROCHEMICALLY CONTROLLED H-BONDING
for oxidation of the electroactive urea, 3, discussed previously in the presence of the cyclic diamide, 4.11 Now the guest is expected to bind more strongly to the oxidized form (Eq. 1.4), resulting in a negative shift in E1/2 upon addition of guest, the opposite of the effect seen in Fig. 1.1. Note that under the experimental conditions used, the CV wave for the urea by itself, scan (a), is broad. With half equivalent of guest, scan (b), an even broader CV results, with a new shoulder negative of the original wave. With continued addition of guest, the wave now sharpens up and continues to move negative. In order to explain the original broad wave observed for the urea, it has been proposed that formation of dimer 5 between an unoxidized urea and an oxidized urea is quite favorable. This makes it easier to oxidize the first half of the urea (since the dimer will result) than the second half of the urea (which requires breakup of the dimer). The result is an unusually broad CV wave. Addition of greater than 1 equivalent of the guest breaks up the dimer resulting in a sharper wave. The viability of this mechanism is confirmed by showing that CVs simulated using this mechanism (dots in Fig. 1.2) give good fits to the experimental CVs (lines) with reasonable values for the different reaction parameters.
H
Ph N
Me2NPh
N
O
H
H
O 5
N
Ph H N PhNMe2
While cyclic voltammetry is clearly the most important technique for studying electrochemically controlled H-bonding, other physical methods, in particular spectroscopic techniques, can be quite helpful. These can be used to provide structural information on the host–guest complex, and also to provide another means to determine binding constants in at least one of the redox states. For H-bonded supramolecular complexes, the most commonly used technique is simply 1 H NMR, since the chemical shifts of hydrogens involved in the H-bonding will be very sensitive to the presence of the binding partner. However, a significant limitation for the use of NMR in electrochemical H-bonding studies is that it can only be used with diamagnetic systems, and in many cases the stronger binding state is paramagnetic. 1.4 GENERAL CONSIDERATIONS FOR THE DESIGN OF ELECTROCHEMICALLY CONTROLLED HYDROGEN-BONDED ASSEMBLIES 1.4.1
The Redox Couple
First and foremost, the redox couple must be reversible, meaning relatively fast electron transfer reactions and products that are stable in both oxidation states. As discussed previously, for systems in which the hydrogen bonds are to be directly perturbed, the redox reaction must affect the charge distribution on the atoms involved
1.4
GENERAL CONSIDERATIONS FOR THE DESIGN
9
TABLE 1.1 Shift in Half-Wave Potential, DE1/2, for Different para-Substituted Nitrobenzenes in the Presence of 1,3-Diphenylureaa Substituent NH2 CH3O CH3 H CF3
E1/2 (V) versus Fc
DE1/2 (mV)
1.870 1.698 1.635 1.582 1.362
197 164 156 153 93
a 1 mM nitrobenzene in 0.1 M NBu4PF6/DMF þ 50 mM 1,3-diphenylurea. Values are the averages of at least three independent measurements.
in one or more of the hydrogen bonds. This can be done through what can be described as either an inductive or a resonance effect. The nitrobenzene (Eq. 1.3)10 and dimethylaminophenylurea systems (Eq. 1.4)11 that were discussed in the previous sections are good examples of resonance effects. In these cases, the electron transfer directly places charge on the atoms involved in H-bonding, which are an integral part of the redox couple. One interesting possibility is to couple this with simple substituent effects in order to further “tune” the binding strength. This hypothesis has only been explored in two systems so far, with mixed results. One is the nitrobenzene system, where the strategy does appear to work, as shown by the data in Table 1.1.10 This table gives the observed E1/2 of the nitrobenzene 0/1 redox couple, along with the observed DE1/2 in the presence of 50 equivalents of 1,3-diphenylurea for five different p-substituted nitrobenzenes. As expected, the E1/2 values fall in order of the electron-donating/withdrawing strength of the substituents, with the nitrobenzene with the most electron-donating substituent, NH2, being the hardest to reduce and that with the most electron-withdrawing substituent, CF3, being the easiest to reduce. Interestingly, the DE1/2 values fall in the same order, with the NH2 derivative giving the largest shift and the CF3 the smallest. This is consistent with the electron-donating NH2 forcing more of the negative charge in the radical anion onto the oxygens leading to an increase in H-bond strength, and with the electron-withdrawing CF3 removing some of the negative charge from the oxygens leading to a decrease in H-bond strength. However, although tuning DE1/2 values with substituents appears to be quite effective in the nitrobenzene system, an even more in-depth substituent study for the flavin/amidopyridine H-bonding system,15 which will be discussed in detail later in the chapter, shows the expected E1/2 dependence, but no clear trend in the DE1/2 values. It therefore remains to be seen how generally useful this strategy may be. It is also possible to affect H-bond strength electrochemically without the H-bonding site being an integral part of the redox couple. Good examples of this are found in the many redox-dependent receptors that utilize metallocenes, primarily ferrocene and cobaltocenium, as the redox couple. These are primarily used in ion receptors, but examples of metallocene receptors that show a significant redox dependence with neutral guests are the cobaltocenium and ferrocene diamides,
10
ELECTROCHEMICALLY CONTROLLED H-BONDING
6 and 7. Both metallocenes undergo a reversible one-electron transfer reaction between 0 and þ 1 charge states. However, since cobalt has one more valence electron than iron, the cobalt derivative has the optimum 18 valence electron configuration and is therefore most stable in the þ 1 state, whereas the iron analog has the optimum 18 valence electron configuration and is most stable in the zero charge state. O N H O 6 M = Co+ 7 M = Fe
N H O (CH2)3
M O H N O
O H N
8
Tucker and coworkers have shown that both of these compounds bind glutaric acid, 8, through H-bonding between the carboxylic acids and the diamidopyridine.161 H NMR titrations in CDCl3/DMSO (0.5%) indicate that the cobaltocenium host binds 20 times more strongly than the ferrocene host (Kox of 6 ¼ 9.8 104 M1; Kred of 7 ¼ 4.6 103 M1). This is consistent with the positive charge on the cobaltocenium, which would be expected to further polarize the NH amide bond through an electron-withdrawing effect, resulting in stronger H-bonding to the carbonyl oxygen of the acid. In fact, crystal structures show that the amide H-bonds are shorter in the cobaltocenium derivative than the ferrocene, whereas the pyridine H-bonds are about the same. Reduction of the cobaltocenium to the zero state, and oxidation of the ferrocene to the þ 1 state would be expected to switch the binding preferences, and indeed the CV studies in CH2Cl2/DMSO (0.5%) indicate that this is the case. Addition of excess glutaric acid to both compounds results in the same maximum DE1/2 of 90 mV. This corresponds to a 30-fold difference in binding strength between oxidation states, with the þ 1 state being the stronger binding one in both cases. This indicates that the difference in binding between oxidation states for this system is simply due to the change in charge. Unlike the nitrobenzene example, it is independent of variation in structure (Co versus Fc) or actual E1/2 since the E1/2 of Co(Cp)2 þ 1/0 is quite negative of the E1/2 of Fe(Cp)2 þ 1/0. In comparing the two different ways the redox couple can perturb H-bonding strength, there are clearly pros and cons to both. The resonance approach, in which the H-bonding sites are an integral part of the redox couple, is likely to give a stronger redox perturbation than the inductive approach, where the redox couple is simply in conjugation with the H-bonding site. The existing examples bear this out. With neutral guests, where H-bonding is the main binding interaction, the electroactive hosts that have been shown to give the very large DE1/2 values are all those in which the H-bonding site is part of the redox couple. The disadvantage is that this really
1.4
GENERAL CONSIDERATIONS FOR THE DESIGN
11
puts severe limitations on host/guest combinations possible. The advantage of the inductive approach is that the binding site can be designed somewhat independent of the redox couple, making it, in principle, easier to tailor the host for different guests. The other advantage is that very inert redox couples such as the metallocenes can be used, so there is less chance of unwanted reactions taking place. Another factor to take into consideration when choosing a redox couple is the availability of additional electron transfer equilibria. The redox-dependent hosts considered so far, nitrobenzenes, 1, the dimethylaminophenylurea, 3, and the metallocene diamides, 6 and 7, just undergo one reversible electron transfer reaction. However, many organic redox couples undergo at least two successive electron transfers. The prototypical example are the quinones, 9, which in aprotic solvents undergo two reversible reductions, first to the radical anion and then to dianion (Eq. 1.8). p-Phenylenediamines, 10, provide a related example of a redox couple that undergoes two reversible oxidations, giving first a radical cation and then a dication (Eq. 1.9). (Note that the dimethylaminophenylurea, 3, is a phenylenediamine derivative, but the second oxidation is irreversible in this case.) In general, the more highly charged dianions and dications will be significantly more basic or acidic than the mono-ions, and therefore will H-bond more strongly to a given guest, resulting in much larger DE1/2 values in the 1/2 or þ 1/ þ 2 redox couple than the 0/1 or 0/ þ 1 redox couples. This phenomena has been particularly well studied with quinones.17 Addition of weak H-donors, such as alcohols, produces large positive shifts in the 1/2 redox potential with little change in the 0/1 potential. Similar results are observed with p-phenylenediamines and weak H-acceptors such as tertiary amides.18 O
O +
e–
O +
e–
ð1:7Þ O
O
O
NH2
NH2
9 NH2 – e–
– e–
ð1:8Þ NH2
NH2
NH2
10
Based on the above discussion, there is an advantage in using redox couples with multiple electron transfer equilibria since this makes incremental control of binding strength possible. However, caution must also be used, since the stronger H-bonding also means that the response will be much less selective, in that weaker H-donors/ acceptors could also cause a significant shift. In addition, because the dianions and dications will be more reactive, there is a greater likelihood of proton transfer and other chemical reactions taking place at the second electron transfer, which could
12
ELECTROCHEMICALLY CONTROLLED H-BONDING
compromise the reversibility of the system. These issues are problematic for practical sensor applications, where the redox response should be highly selective for a particular analyte in a range of different environments. However, this may not be as much an issue in other applications, where the environment can be controlled to prevent unwanted interactions with the more reactive oxidation states. 1.4.2
Host/Guest Structure
In order to achieve strong, selective binding with electroactive hosts, just as in other types of molecular recognition, there needs to be a high degree of complementarity between the host and the guest. This means that the binding site needs to have the proper shape, along with the correct positioning of functionality to provide multiple favorable contacts with the target guest. The principle of preorganization is also important, which says that binding strength will be maximized when the host is “preorganized” to fit the guest because this will minimize the entropy loss upon formation of the host–guest complex. For this reason, rigidity in both the host and if possible the guest is generally desirable. Of course, in order to produce a redox signal for binding, at least one of the binding interactions must be strongly perturbed by a reversible electron transfer reaction. As discussed in the previous section, a greater effect will generally be observed at the second electron transfer (if available), producing a larger DE1/2 and making the redox potential of the host sensitive to the guest over a larger range of concentrations. However, this will also likely make the response much less selective, so for sensor applications, the first electron transfer will probably need to be used. This means that a big change in binding strength is needed based on what will typically be a 0/1 or 0/ þ 1 redox couple. In looking at existing electrochemically controlled H-bonding systems in which significant E1/2 shifts have been observed at the first electron transfer, two minimum features appear necessary. First, there needs to be the possibility of at least two strong, almost linear H-bonds formed between host and guest, and, second, the electron transfer reaction needs to strongly perturb at least one and preferably two H-bonds between host and guest. Examination of the systems discussed so far in this chapter, as well as those yet to be discussed, shows that all obey these basic design principles. A nice demonstration of the above criteria is provided by a study of the redoxdependent H-bonding properties of 9,10-phenanthrenequinone, 11, and related compounds.19 As with nitrobenzenes, diarylureas such as 2 make excellent binding partners, providing appropriately positioned amide NHs to be able to simultaneously H-bond with both carbonyl oxygens. Also like nitrobenzene, reduction of phenanthrenequinone produces a radical anion, with enhanced negative charge on the oxygens. This should result in stronger H-bonding in the radical anion state (Eq. 1.10). Consistent with this, addition of 10 equivalents of 1,3-diphenylurea causes the E1/2 of the phenanthrenequinone 0/1 couple to shift positive by 200 mV in CH2Cl2. Significant positive E1/2 shifts are also observed in the more polar DMF. CV simulation of the latter gives Kox ¼ 1 M1 and Kred ¼ 905 M1 in 0.1 M NBu4PF6/DMF.
1.4
13
GENERAL CONSIDERATIONS FOR THE DESIGN
Ph O
O
+ e–
+2
O
H N
O
H N
O O
O
2 Ph 11
ð1:10Þ In order to investigate the structural requirements for the strong redox dependence, DE1/2 values were also measured for anthraquinone, 12, and benzyl, 13, in the presence of 5 equivalents of diphenylurea in DMF. Under these conditions, phenanthrenequinone gives a shift of 61 mV, whereas anthraquinone gives a shift of only 8 mVand benzil 5 mV. Unlike phenanthrenequinone, the urea can only H-bond to one carbonyl oxygen at a time with anthraquinone. Two bifurcated H-bonds are possible, but these together would be much weaker than the two close to linear H-bonds possible with o-quinones. A similar situation arises with benzil, since rotation about the central CC bond will be hindered in the radical anion and the favored conformation will have the oxygens trans due to electrostatic repulsion and steric effects. O O
O O 12
13
Another interesting comparison is the difference in DE1/2 observed for phenanthrenequinone with the two pyridylureas 14 and 15. The electronic character of both compounds should be similar, but with 14 a strong intramolecular H-bond will be formed between the 2-pyridyl N and one of the urea NHs. This means that only one NH will be available for H-bonding to the phenanthrenequinone radical anion. This has a huge effect on the DE1/2 values, with 5 equivalents of 14 producing only a 16 mV shift in DMF compared to an 85 mV shift with 15. The latter is actually larger than 1,3diphenylurea. This can be explained by the greater electronegativity of N compared to C, making the pyridyl group more electron withdrawing than a phenyl group. O H
N N N
N H
N H 14
1.4.3
O N H 15
Solvent and Electrolyte
In addition to the structures of host and guest, another important consideration for studies of electrochemically controlled H-bonding is the solvent system.
14
ELECTROCHEMICALLY CONTROLLED H-BONDING
Binding constants for supramolecular assemblies are always strongly solvent dependent because binding of the host to the guest will be in competition with solvation. Since H-bonding has a strong electrostatic component, more polar solvents will solvate the binding sites more effectively than less polar solvents, thereby decreasing the equilibrium constant for host–guest binding, whereas less polar solvents will tend to increase the binding strength. For nonelectrochemical studies, chloroform (more specifically, deuterated chloroform for 1 H NMR studies) seems to be the most common solvent. For electrochemical studies, a less resistive solvent is needed, with dichloromethane being the most common choice. Other aprotic solvents such as acetonitrile or dimethylformamide are also used. Sometimes, these can help to simplify the behavior by decreasing the amount of host–host or guest–guest interaction that occurs in the less polar dichloromethane. Another important consideration for electrochemical studies is the choice of electrolyte, the ionic compound that is added to maintain electroneutrality and provide a means of charge flow through solution. Because at least one of the oxidation states of the host will be charged, it is possible that ion–ion interactions will play a role in the observed electrochemistry. Indeed, this is useful if the objective is to design a redox-dependent ion receptor, but it can be an interference if the guest is neutral. The most common electrolyte used today for electrochemical studies in aprotic solvents is tetrabutylammonium hexafluorophosphate, NBu4PF6. The use of NBu4 þ is a good choice for reduction-based systems since it is a very large cation and is unlikely to interact significantly with the anionic form of the host. For oxidation-based systems, which generally involved cationic intermediates, it is the anion in the electrolyte that is of issue. PF6 has generally been considered to also be a large and noncoordinating ion. However, recent evidence suggests that it is not nearly as inert as once believed. In particular, electrochemical studies with organometallic cations show that PF6 and other “large” anions such as ClO4 and BF4 interact with þ 2 and þ 3 charged species enough to have a large effect on the observed E1/2 values, sometimes also causing distortion in the CV wave shapes and/or precipitation of the cationic products onto the electrode.20,21 If good H-donor groups are part of the redox couple, it might be expected that even þ 1 charge species could show strong interactions with these anions. This has recently been shown to be the case with the electroactive urea 3 discussed earlier (Eq. 1.4).11 In the results described previously (Fig. 1.2), the electrolyte was 0.1 M NBu4B(C6F5)4/ CH2Cl2. B(C6F5)4 is a very large anion, much larger than the more commonly used PF6. If the same experiments are run with NBu4PF6 or NBu4ClO4, much smaller E1/2 shifts are observed with the same guest. For example, addition of 1 equivalent of guest 4 to 3 gives DE1/2 values of 2 and 16 mV using NBu4ClO4 or NBu4PF6, respectively, but the same amount of guest gives 109 mV shift when NBu4B (C6F5)4 is used as the electrolyte. These results indicate that the smaller anions (which are present at much larger concentrations than the guest) interact strongly enough with the oxidized urea to block interactions with the guest. The much larger anion B(C6F5)4 does not do this, with the result that significantly larger shifts are observed in the presence of this anion. Based on these results, it would seem prudent
1.4
GENERAL CONSIDERATIONS FOR THE DESIGN
15
to at least test oxidation-based systems with these very large anion electrolytes, particularly if a less polar solvent such as CH2Cl2 is being used. 1.4.4
H-Bonding Versus Proton Transfer
Although it seems to be often pointed out that H-bonding and proton transfer are not the same thing (and obviously they are not), there is clearly a close relationship between the two, with stronger acids generally being better H-donors and stronger bases being better H-acceptors. As the pKa of a H-donor guest decreases, the E1/2 shift between the guest and the host will increase but so will the possibility of proton transfer. For organic 0/1 or 0/ þ 1 redox couples, the occurrence of proton transfer is usually quite obvious. Instead of seeing simply a shift in potential of the reversible, one-electron CV wave upon addition of the guest, the wave doubles in size and becomes irreversible. This behavior, which was originally observed upon addition of acids to aromatic anions, is due to an ECEC mechanism. (E stands for electron transfer reaction and C stands for a chemical reaction such as proton transfer.) In the case of reductions, the first step is electron transfer to form a radical anion (Eq. 1.11). Normally, addition of a second electron occurs at a more negative potential because it will be harder to add an electron to an already negatively charged species. However, protonation of the radical (Eq. 1.12) gives an uncharged radical that is typically easier to reduce than the starting species. The result is immediate addition of a second electron (Eq. 1.13) resulting in a two-electron CV wave. Since an acid strong enough to protonate the radical is usually strong enough to protonate the anion product, the final product will generally be the two-electron, two-proton reduced species, RH2 (Eq. 1.14). The reduction is irreversible because it will be much harder to oxidize RH2 than R. A similar mechanism can be observed with oxidations in the presence of bases capable of removing a proton from the radical cation intermediate. R þ e ! R
ð1:11Þ
R þ H þ ! RH
ð1:12Þ
RH þ e ! RH
ð1:13Þ
RH þ H þ ! RH2
ð1:14Þ
Since the products of the second electron transfers are typically more basic or more acidic than those of the first electron transfer, it is possible to add a guest that is only capable of proton transfer with the product of the second electron transfer. It is important to note that in this case the result of proton transfer may be very similar to that of H-bonding, with simply a positive E1/2 shift for the second wave in a reduction or a negative shift for an oxidation. In this case, information about pKa values is very helpful for trying to sort out whether it is H-bonding or proton transfer that causes the shift.
16
ELECTROCHEMICALLY CONTROLLED H-BONDING
1.5 EXAMPLES OF ELECTROCHEMICALLY CONTROLLED H-BONDING SYSTEMS WITH ANIONIC GUESTS Although the examples of electrochemically controlled H-bonding discussed so far in this chapter have been with neutral guests, the reality is that the majority of the work that has been done in this area has been with ionic, and in particular anionic guests. This work started in the late 1980s, several years before the first reports with neutral guests. It is still an incredibly active area, with numerous new receptors reported each year. Work through the 1990s has been covered in review articles.22,23 The first reported example of anion binding with a synthetic organic macrocycle was the protonated cryptand 16, reported by Park and Simmons in 1968.24 1 HNMR evidence was presented showing that this compound binds chloride by a combination of H-bonding and favorable electrostatic interaction. Although numerous other anion receptors have been reported since then, including those that utilize Lewis acid and hydrophobic interactions, the two binding interactions seen in this first example, ion–ion and H-bonding, have turned out to be the mainstays of anion recognition. This leads naturally to electrochemically controlled H-bonding since adding an electroactive group provides a simple way to perturb both electrostatic and H-bonding interactions. O (CH2)9 N H
Cl (CH2)9 (CH2)9 16
1.5.1
O O
O
Co+
Co+
H N
O
O
O
O 17
Cobaltocenium
The first redox-active receptor for anions was the bis-cobaltocenium macrocylic ester 17 reported by Beer and Keefe in 1989.25 FAB-MS and FT-IR evidence indicates that 17 can bind anions in the oxidized state. Reduction of the cobaltoceniums to the zero state would be expected to weaken this interaction, and indeed a modest 45 mV E1/2 shift is observed in acetonitrile upon addition of 4 equivalents of Br. In the above example, binding and redox dependence rely solely on perturbation of electrostatic interactions. A few years later, in 1992, Beer and coworkers reported similar or better redox dependence with halides using simpler acyclic cobaltoceniums.26,27 The key was the addition of an amide NH, providing a H-bonding site. For example, the simple diamide cobaltocene 18 gives a 60 mV shift in E1/2 of the cobaltocenium upon addition of 4 equivalents of Br in acetonitrile, corresponding to at least a 10-fold decrease in binding strength. The analogous NMe compound results in a smaller than 5 mV shift. A later crystal structure of 18 with Br clearly
1.5 EXAMPLES OF ELECTROCHEMICALLY CONTROLLED H-BONDING SYSTEMS
17
shows the H-bond with the amide NH.28 In addition, a significantly larger shift of 240 mV is observed with 18 and H2PO4, due to the greater H-bonding capability of this anion compared to simple halides.
O
O N H H N
Co+
18
1.5.2
Fe
19
N H H N
N H
H N
O
O
Ferrocene
The cobaltocenium anion receptors are examples of electrochemically controlled H-bonding systems in which the electron transfer decreases H-bonding strength. By switching to ferrocene as the redox-active group, anion receptors can be made in which electron transfer increases binding strength. This has proven to be a particular popular strategy and there are now numerous examples of ferrocene-based, redoxdependent anion receptors in the literature. A generally successful strategy, learned from the early cobaltocenium examples, is to simply attach good hydrogen donors, such as amides, secondary amines, guanidiniums, and so on, to one or both cyclopentadienes. A representative example is the diurea-substituted ferrocene, 19, reported by Molina and coworkers,29 which shows a significant redox-dependent response to F (DE1/2 ¼ 208 mV) and H2PO4 (DE1/2 ¼ 90 mV) in DMSO, but not to Cl, Br, AcO, NO3, and HSO4. This is an example of a receptor where strong binding is observed in both oxidation states, resulting in two CV waves being observed at less than 1 equivalent of guest. Most redox-dependent receptors have structural features similar to 17–19, which are either a macrocyclic or a cleft-type binding site with one or two redox-active groups directly attached. An interesting alternative, introduced by Astruc and coworkers, are the ferrocene-terminated dendrimers, such as 20.30,31 Although each dendrimer contains multiple ferrocenes, just one CV oxidation wave is generally observed, indicating that there is no significant interaction between the ferrocenes. Addition of a good H-bonding anion either causes this wave to shift negative or produces a new CV wave. One of the interesting results is that there is generally a “positive dendritic effect,” in that higher dendrimer generations show a larger maximum shift. For example, Fig. 1.3 shows the effect of different generations of amine dendrimers on the observed DE1/2 caused by addition of HSO4. (18-Fc is 20, which has 18 ferrocenes, 9-Fc is the previous generation with 9 ferrocenes, and so on.) The anions are believed to bind to the amide groups between dendrimer chains as
18
ELECTROCHEMICALLY CONTROLLED H-BONDING
Figure 1.3 Observed DE1/2 for the ferrocene 0/ þ 1 CV wave upon addition of NBu4HSO4 to different generations of Fc dendrimers in NBu4BF4/CH2Cl2. The x-axis indicates the number of equivalents of HSO4 added per ferrocene unit.30
shown in structure 21. As the dendrimer generation grows, the chains are forced closer together creating a tighter and more preorganized binding cleft that leads to stronger binding and larger potential shifts. Similar effects have been observed with a variety of different ferrocene dendrimers, as well as Au nanoparticles that have amidoferrocene-terminated alkyl thiols attached to their surfaces.32 These type of receptors, where the binding site is a wedge in between outward radiating chains, have been classified as “exo” receptors.
O
Fc NH
Fc O NH O Fc N H O Fc NH N O
Fc H N O Fc H N Fc N N O O O N H H N O O Fc N O Fc N O H H O N O Fc N N N Fc O O O H H N O O O N Fc O N Fc H N H N N O Fc N O 20 O H Fc HN N O HN H Fc Fc O Fc
+Fc N H O O S H O O O H N +Fc O
21
19
1.6 EXAMPLES OF REDUCTION-BASED
1.6 EXAMPLES OF REDUCTION-BASED, ELECTROCHEMICALLY CONTROLLED H-BONDING SYSTEMS WITH NEUTRAL GUESTS 1.6.1
Flavins
Although the idea of using H-bonding to control redox properties is fairly recent in man’s chemistry, nature has been doing this for quite a long time. A nice example is seen in flavoenzymes that use the redox-active flavin group, 21, to catalyze a wide variety of reactions, such as oxidation of amines to imines and the hydroxylation of aromatic compounds. That one cofactor can catalyze such a range of reactions is possible because the protein environment interacts with the flavin through a variety of intermolecular interactions, including H-bonding, so as to adjust the redox chemistry to be appropriate for each particular task. R N
N
O N
N
+ e–
H
R N
O N
+ H+
R N
+ e–
N H R N
H
O
R N N H
N
O N
O
H
ð1:15Þ
O
N
N
O N
N
O 21 R N
N
H
N H
N
O N
ð1:16Þ H
O
N
O N
ð1:17Þ H
O
The reactions most commonly involved in flavin redox chemistry are shown in Equations 1.15–1.17. One-electron reduction of the flavin (Eq. 1.15) produces a relatively stable radical anion. Protonation of the radical anion produces an unstable neutral radical (Eq. 1.16), which will be rapidly reduced by another electron (Eq. 1.17) to give the flavohydroquinone anion. In an effort to better understand how the redox behavior of flavins is influenced by hydrogen bonding, Rotello and coworkers reported a study in 1995 where they looked at the electrochemistry of flavin 22 (R1 ¼ R2 ¼ Me) in the presence of diamidopyridine 23 (R ¼ Et).33 This compound hydrogen bonds to the flavin as shown, mimicking the hydrogen-bonding pattern in the active site of a number of flavoproteins. The binding constant of the diamide to the flavin in the oxidized form is 537 M1 in CDCl3 as determined by 1 H NMR. Reduction of the flavin to the radical anion would be expected to strengthen binding further by increasing the negative charge on the carbonyl oxygens, and, indeed, addition of 5 equivalents of 23 to the flavin in NBu4ClO4/CH2Cl2 resulted in a þ 155 mV shift in the E1/2 of the flavin 0/1 potential (Fig. 1.4b), indicating strong stabilization of the radical anion through H-bonding. This result was a nice confirmation of the role that
20
ELECTROCHEMICALLY CONTROLLED H-BONDING
Figure 1.4 CVs of different flavins, 22, in 0.1 M NBu4ClO4/CH2Cl2 by themselves (solid line) and in the presence of 50 mM 23 (dashed line): (a) 22, R1 ¼ Me, R2 ¼ NMe2, (b) 22, R1 ¼ R2 ¼ Me, (c) 22, R1 ¼ R2 ¼ H, and (d) 22, R1 ¼ R2 ¼ Cl. 200 mV/s scan rate.15
H-bonding plays in the flavoenzymes, but, perhaps more importantly, it really showed how strong an effect intermolecular H-bonding with a neutral binding partner could have on the redox potential of an organic redox couple, and, as a result, ushered in all the work that has followed on electrochemically controlled H-bonding with neutral guests. R O O
H N
N N
N H N
R2 R1
22
O
N 23
H N O R
21
1.6 EXAMPLES OF REDUCTION-BASED
Since their initial report on flavins, Rotello and coworkers have reported a number of other detailed studies on this system.15,34–44 One of the interesting things about the flavin/diamidopyridine system is that not only does the guest substantially alter the redox potential, but it also changes the nature of the electrode reaction. As shown in Fig. 1.4, in the absence of the guest, the flavin electrochemistry is generally not completely reversible. The return oxidation peak is too small and there is a second oxidation peak that appears at more positive potentials. Addition of the guest both shifts the reduction positive and makes it reversible. A completely reversible wave is also observed, without the guest, when the imide N is methylated, indicating that the irreversibility is due to the presence of the relatively acidic imide NH. Rotello and coworkers have explained this by suggesting that the apparent one-electron reduction of the flavin in aprotic solvent actually corresponds to the two-electron, one-proton reduction to the flavohydroquinone anion (Eqs 1.15–1.17), where the proton source is an oxidized flavin coming in from bulk solution.45 Deprotonation of the oxidized flavin prevents its reduction, with the result that the overall process is still one electron/flavin (Eq. 1.18). The return oxidation peak at more positive potentials is then due to oxidation of the flavohydroquinone anion. Addition of the diamidopyridine increases the reversibility of the flavin reduction by preventing the proton transfer both through stabilization of the imide NH on the oxidized flavin and by alteration of the charge distribution in the radical anion.15
R2
R N
2 R1
O
N N
N O
H
+
2e–
R2 R1
R N N H
O
N N O
R2
R N
H
R1
N
O
N
+
N O
21
ð1:18Þ Other groups have also looked at the electrochemistry of flavins with various Hbonding partners. One of the most interesting studies is that by Yano and coworkers who designed the elaborate binding partner 24 for the azoflavin 25.46 The pendant guanidinium groups provide four H-bonds in addition to the three set up to bond to the imide portion of the flavin. This results in very strong H-bonding even in the oxidized state, with Kox ¼ 1.9 104 M1 in 20% ACN/CH2Cl2. Reduction of the flavin to the radical anion increased the binding strength even further, giving a DE1/2¼ 317 mV. This corresponds to a binding enhancement of 2.2 105 and a Kred ¼ 4.3 109 M1, certainly one of the largest binding constants that has been reported for these types of H-bonded complexes. Because of the very strong binding in both oxidation states, this is also one of the few examples where two CV waves, one at EH and one at EHG (Scheme 1.1), are observed at less than 1 equivalent of the guest.
22
ELECTROCHEMICALLY CONTROLLED H-BONDING
H H
H
N
N H
N H
H22C12 N N
H
N
N
H N
H N
H
1.6.2
O H
25
N
N
H N
O H H
N
N N
24
N
C H N 6 13 C6H13
Arylimides
With flavins, the redox-dependent H-bonding site is the imide functional group on one side of the molecule. A year after Rotello’s initial report on the flavin/ diamidopyridine system, Smith and coworkers reported that a similar type of redox-dependent system can be created by just using a simple aromatic imide such as 1,8-naphthalimide, 26, with the same type of diamidopyridine guests.47 Subsequently, Rotello’s group has used aromatic imides in a variety of redox-dependent binding studies.48–50 Like the flavins in aprotic solvents, the aromatic imides undergo reversible reductions to radical anions in aprotic solvents, which greatly increases the strength of binding to diamidopyridine or other similarly structured binding partners. R O O N H O
H N N 23
H N O
26 R
With the flavin and imide systems, the central NH provides an additional H-bond that helps to strengthen the interaction with guest in the oxidized form. However, this is a H-donating group while the carbonyls are H-acceptors, so while reduction would be expected to increase the H-accepting ability of the carbonyls and strengthen the H-bonds between imide carbonyl and amide, it would actually be expected to decrease the strength of the H-bond between the imide NH and the pyridine N. A comparison of the E1/2 values of a flavin and its N-methyl analog in solvents of varying H-acceptor strength, along with some computational work, supports this hypothesis.51
23
1.6 EXAMPLES OF REDUCTION-BASED
Since, from the above analysis, the imide NH actually reduces the redox-dependent binding effect, an alternative way to use imides in H-bonding systems is to eliminate the NH, replacing it with an N-alkyl group.52,53 Reversible reduction still results, with now just the carbonyl oxygens being involved in the H-bonding. Since it is so straightforward to alkylate at the imide NH, this also provides a simple means to connect additional functionality. Very nice examples of this are the H-bond-based molecular shuttles reported by Leigh, Paolucci, and coworkers.54,55 Molecular shuttles are constructed from at least two separate, but physically interconnected molecular units. The assembly, called a rotaxane, consists of a long linear molecule that is “threaded” through a macrocycle. Bulky groups on either end of the “thread” prevent the macrocycle from slipping off. For a shuttle, the thread will contain at least two binding sites for the macrocycle, with one of these being strongly preferred over the other in the initial state. This preference changes upon external stimulus, such as electron transfer. O Ph Ph
N H
H N
O N
7
O
O
O
O NH
HN
NH
HN
27
O Ph Ph
N H
H N
O N
9
O
O
O
O
Ph N
28 O
O
Ph 29
The threads of the H-bonding shuttles reported by Paolucci, Leigh, and coworkers contain a succinamide station connected to either an imide, 27, or a diimide, 28, station. 1 H NMR clearly shows that the amide macrocycle, 29, strongly prefers the succinamide station in the oxidized state. However, upon reduction of the imide/ diimide, the preference changes, with now the imide/diimide station being strongly preferred. This results in net movement of the macrocycle from the succinamide to the imide/diimide (Eq. 1.19). This behavior can be deduced from the difference in the CVs of the thread by itself and that of the macrocycle-threaded shuttle. With the naphthylimide shuttle, the succinamide station is preferred over the imide 106 to 1 in the oxidized state, but, in the reduced state, the imide is preferred 500 to 1. Switching from an imide to a diimide opens up an additional control element because the diimide now undergoes two reversible reductions, first to the radical anion and then to a dianion. Analysis of the electrochemical data for this system indicates a 200 to 1 preference for the succinamide station in the zero state. This switches to a 2 to 1 preference for the diimide in the 1 state and a 4000 to 1 preference for the diimide in the 2 state.
24
1.6.3
ELECTROCHEMICALLY CONTROLLED H-BONDING
o-Quinones
In addition to the arylimides, Smith and coworkers also introduced o-quinones, discussed earlier in this chapter, as another example of a redox-dependent H-bonding receptor for neutral guests (Eq. 1.9).19,47,56 This system differs from the flavins and the imides in that a strong interaction is only observed in the reduced state, making these receptors function as on/off switches. The groups of Rotello and Cooke have also published several papers exploring the redox-dependent H-bonding properties of o-quinones.36,57–59 Many of the applications that can be envisioned for electrochemically controlled H-bonding and redox-dependent binding in general will require that at least one of the components be attached to an electrode surface. It is therefore important to determine whether behavior observed in solution can still be observed when the host is anchored to a surface. This was first tested for a redox-dependent H-bonding receptor with the o-quinone system using the phenanthrenequinone pyrrole 30.56 Electrooxidation of the pyrrole unit results in the formation of a pyrrole polymer that coats the electrode surface as it is formed. The amount of polymer deposited can be controlled by the number of CV cycles into the pyrrole oxidation wave. With 30, thick polymer layers give broad CV waves in the quinone voltage region, but thinner layers produce a well-resolved wave for the quinone 0/1 reduction, which is reasonably stable when the electrodes are placed into fresh electrolyte solution with no 30. As in solution, addition of different urea derivatives causes this wave to shift positive. The relative magnitude of the shifts mirror that seen in solution. Furthermore, the E1/2 moves back to the original potential when the derivatized electrode is put back into a blank solution containing no urea. O
O
(CH2)4 N 30
Cooke and coworkers have reported preparation of flavin-modified electrodes using a similar electropolymerization procedure.34 They have also studied electrodes coated with self-assembled monolayers (SAMs) formed from both flavin39 and phenanthrenequinone disulfides.59 The monolayers are stable in CH2Cl2 solution, and, as with the electrodes formed from 30, show redox-dependent binding behavior similar to that seen in solution. Interestingly, the phenanthrenequinone SAM
1.6 EXAMPLES OF REDUCTION-BASED
25
electrodes were also studied in the presence of phenylurea-terminated dendrimers. Addition of excess dendrimer produces a þ 200 mV shift in the E1/2 of the phenanthrenequinone 0/1 reduction, which is significantly larger than that obtained with phenanthrenequinone dissolved in solution. 1.6.4
Nitrobenzenes
Nitrobenzenes, also introduced by Smith and coworkers10 and discussed previously, provide yet another example of a simple organic redox couple in which reduction increases negative charge on two convergent oxygens, in this case the nitro oxygens (Eq. 1.3). The use of this group has not been explored as much as the flavin, imide, and o-quinone systems. However, the simplicity of its structure and the ease of which it can be introduced into organic structures suggests it may prove to be useful in the construction of more elaborate supramolecular assemblies. An example of the possible utility of the nitroaromatic group in creating more complicated structures comes from the still very simple case of 1,4-dinitrobenzene, 31.60 Almost all of the redox-dependent H-bonding systems investigated so far involve 1:1 complex formation. The ferrocene dendrimers are one exception. Another is 1,4-dinitrobenzene, where cyclic voltammetry provides very strong evidence for the formation of the 1:2 complex with diphenylurea, 2, upon reduction to the dianion (Eq. 1.19). This dianion is unusually stable because the negative charge is spread over four oxygen atoms. It is possible that it could function as a simple linker unit that could tie together other units containing, for example, two urea groups, allowing larger and more complex structures to be assembled and disassembled under redox control. Ph
Ph O
O N O
N O
+ 2e–, + 2 2
N H
O
N H
O
N
O Ph
31
1.6.5
2
O N O
H N O H N
ð1:19Þ
Ph 2
Tetrazines
All of the examples discussed above involve an increase in negative charge on two convergent oxygen atoms. Nitrogen of course is also a good H-accepting atom, but a negatively charged nitrogen would generally be expected to be quite basic, resulting in greater problems with protonation. However, recently Rotello and coworkers have reported that the tetrazine derivatives, 32 and 33, act as redox-dependent H-bonding hosts for dialkylthioureas (Eq. 1.20).611 H NMR titrations in CDCl3 indicate that there is no significant interaction with the thioureas and the tetrazines in their oxidized state. However, addition of 50 equivalents of 1-ethyl-3-octylthiourea, 34, to solutions of the tetrazines in CH2Cl2 produced a þ 80 mV shift in the E1/2 of the 0/1 reduction of 32 and a þ 60 mV shift in that of 33, indicating fairly significant interaction
26
ELECTROCHEMICALLY CONTROLLED H-BONDING
between the thioureas and the tetrazine radical anions. Computational studies support the formation of H-bonds between the urea NHs and two adjacent nitrogens in the radical anion, as indicated in Equation 1.20. X N N
X N N
+
e–
X
N N
X N N
+ 34
X
32 X = O 33 X = S
N N
Et N N
X
H N S H N C8H17
ð1:20Þ
34
1.7 EXAMPLES OF OXIDATION-BASED, ELECTROCHEMICALLY CONTROLLED H-BONDING SYSTEMS WITH NEUTRAL GUESTS 1.7.1
Ferrocene and Phenylenediamines
While using oxidation, in particular the oxidation of ferrocene to ferrocenium, has been a generally successful strategy to increase the strength of H-bonds with anionic guests, it has not as of yet been as successful with neutral H-accepting guests. Undoubtedly, part of the reason for this, as discussed earlier, is that this is such a good strategy for anions that the commonly used electrolyte anions, PF6, BF4, and ClO4, will compete for the binding site, particularly in CH2Cl2. By far, the largest DE1/2 values for oxidation-based redox-dependent H-bonding with a neutral guest reported so far are with the dimethylaminophenylurea/diamide system (Eq. 1.4).11 However, this system was studied in the presence of a large anion electrolyte. Much smaller shifts are observed with electrolytes containing PF6 and ClO4. The next largest shifts that have been reported for oxidations with neutral guests are with the ferrocene diamide, 7, reported by Tucker and coworkers. Maximum shifts of 90 mV have been reported with dicarboxylic acid guests,16 and 60 mV with cyclic ureas.62 However, these studies were done in NBu4PF6/CH2Cl2. It is possible that much larger shifts would be observed in a large anion electrolyte. 1.7.2
Tetrathiofulvalene
In addition to ferrocene, the oxidative redox couple that has received the most attention in supramolecular chemistry is tetrathiofulvalene (TTF), 35. This compound undergoes two reversible one-electron oxidations, first to a radical cation and then to a dication (Eq. 1.21). TTF first came to prominence in the 1970s when it was discovered that the charge transfer complex between it and 7,7,8,8-tetracyanoquinonedimethane (TCNQ) shows metallic conductivity. As a result, a large variety of different TTF derivatives have been prepared and characterized. This rich synthetic chemistry, coupled with the electroactivity, has intrigued supramolecular chemists for some time, with the result that the TTF unit has been incorporated into a wide variety of
27
1.8 INDIRECT ELECTROCHEMICALLY CONTROLLED H-BONDING SYSTEMS
macrocycles and other supramolecular assemblies.63 S
S
S
S
– e–
S
S
S
S
– e–
S
S
S
S
ð1:21Þ
35
Like ferrocene, TTF does not inherently contain good H-bonding functionality, but it is possible to attach good H-bonding sites that will be in conjugation with the TTF. This allows for possible perturbation of H-bond strength through an inductive-type effect. Two examples of this strategy have appeared in the literature. The first is the imide-annealed TTF derivative 36 reported by Goldenberg and Neilands,64 which like other imides will H-bond with diamidopyridines. Oxidation of the TTF should decrease the H-bonding by removing electron density from the electron-donating imide carbonyls, and indeed a modest þ 30 mV E1/2 shift for the TTF 0/ þ 1 couple is observed in NBu4PF6/CH2Cl2, indicating a threefold decrease in binding strength upon oxidation. R H O S
S
S
S 36
N O
N N
N Bu
O
H H
N
O
O R
23
S
S
S
S 37
H H N
H N 38 H N
O
Another TTF derivative modified for H-bonding that has been studied is the amidopyridine TTF, 37, prepared by Cooke, Rotello, and coworkers.65 In this case, oxidation of the TTF should increase H-bonding strength by removing electron density from the amide NH, and, in fact, a negative DE1/2 of 33 mV is observed upon addition of amide 38. Interestingly, a computational study suggests that the hydrogen adjacent to the amide group on the TTF ring also likely participates in H-bonding to the carbonyl oxygen of the guest in the þ 1 state. Note that in both of the above cases, the observed E1/2 shifts are modest, but the experiments were done in 0.1 M NBu4PF6/CH2Cl2. Much larger effects might be observed if the work was repeated using a large anion electrolyte.
1.8 INDIRECT ELECTROCHEMICALLY CONTROLLED H-BONDING SYSTEMS In the previous examples, electrochemical control of supramolecular structure is achieved by using electron transfer to directly affect H-bond strength. An alternative strategy is to not attempt to directly perturb the H-bonds, but instead to use electron
28
ELECTROCHEMICALLY CONTROLLED H-BONDING
transfer to create an additional, non-H-bonding interaction that either breaks apart or helps form a H-bonded complex. Given the large number of H-bonded complexes that have been prepared, it seems surprising that there are not more examples of this strategy. Perhaps, one of the reasons is that in this case the voltammetry is unlikely to provide unambiguous evidence that the binding is changing, and therefore alternative proof must be provided. One nice example of indirect control, for which good supporting evidence exists, is the ferrocenylurea calix[4]arene, 39, reported by Moon and Kaifer.66 It has been well established that similar urea-substituted calixarenes form dimeric capsules through H-bonds between the ureas on the two halves in noncompetitive solvents such as chloroform. 1 H NMR evidence shows that 39 also forms dimers in mixed solutions containing up to 3:10 CD3CN/CDCl3. Cyclic voltammetry of the complex in 1:10 CH3CN/CHCl3 shows a single ferrocene wave indicating that the ferrocenes are noninteracting. The thought was that oxidation of the ferrocenes to ferroceniums should break apart the capsule due to the creation of very unfavorable electrostatic interactions. However, although the CV wave is distorted from an ideal one-electron reversible process, this is not enough to prove that the capsule is breaking apart upon oxidation. More convincing evidence is that the diffusion coefficient of the oxidized calixarene, as measured by an NMR technique, is twice as fast as that of the reduced species. Furthermore, FT-IR spectroscopy shows that the peak for the urea carbonyls goes from 1657 cm1 in the reduced species, consistent with H-bonded oxygens, to 1783 and 1698 cm1, in the oxidized species, consistent with nonH-bonded oxygens. O Fc O
Fc NH O NH
NH
HN
NHHN
Fc
OR OR OR RO 39
HN
OHN
N
Fc O
N
H N
H N
O
Fe
O
H 40
H
Fe
O
N H
N H
N
O
N O
While the above example shows that electrochemically created charge can be used to break apart a H-bonded complex, this is not always the case. This is shown by a study of another ferrocene-containing H-bonded dimer, also by Kaifer and coworkers.67 The monomer in this example is based on the well-studied ureidopyrimidine framework that provides a linear AADD (A ¼ H-acceptor; D ¼ H-donor) array capable of forming four linear H-bonds with itself. The version studied by Kaifer, 40, is modified to prevent the keto-enol tautomerization that complicates the chemistry of these systems. 1 H NMR indicates that, as expected, 40 exists as a dimer in CD2Cl2, but as a monomer in the more polar CD3CN. CVs of 40 in different ratios of CH2Cl2 and CH3CN are shown in Fig. 1.5. In pure CH2Cl2, scan (a),
1.9
CONCLUSIONS
29
Figure 1.5 CVs of 1 mM 40 in 0.1 M NBu4PF6 with different ratios of CH2Cl2 and CH3CN: (a) 100% CH2Cl2, (b) 50% CH2Cl2, 50% CH3CN, (c) 20% CH2Cl2, 80% CH3CN, and (d) 100% CH3CN. 100 mV/s scan rate.67 (See the color version of this figure in Color Plates section.)
two reversible CV waves separated by 390 mV are observed. When CH3CN is added, scans (b) and (c), the first oxidation grows at the expense of the second. In pure CH3CN, scan (d), only the first wave is observed at a potential close to the first wave in CH2Cl2. These results are consistent with the NMR, with the two-wave voltammogram being due to dimer and the one-wave voltammogram being due to the monomer. The observance of two well-separated, reversible CV waves for the dimer in this system is a remarkable result. There has to be a very high degree of electronic communication between the two ferrocenes in order to make it so much harder to oxidize the second ferrocene than the first. Given the spatial distance between the ferrocenes, which are on opposite sides of the dimer, this cannot be explained by a through-space interaction, so apparently this electronic interaction is going through the H-bonds. Another surprising feature is that the second wave is reversible. This means that the two monomers dimerize in CH2Cl2 even when both are positively charged. Evidently, for this system under these experimental conditions, the four Hbonds are strong enough to overcome the electrostatic repulsion between the monomers.
1.9
CONCLUSIONS
H-bonds, being one of the strongest intermolecular interactions, are used extensively in the supramolecular chemistry of both man and nature. Their high electrostatic character makes it straightforward to perturb their strength electrochemically. There are two main ways to do this. Reduction can be used to make a H-acceptor a better acceptor by increasing the negative charge on a H-accepting atom, or oxidation can be used to make a H-donor a better donor by increasing the positive
30
ELECTROCHEMICALLY CONTROLLED H-BONDING
charge on a H-donating atom. There are now a number of examples demonstrating the effectiveness of both strategies. In a noncompetitive solvent such as CH2Cl2, it appears possible to change binding strengths by factors of 102–104, and even 105 if additional interactions such as ion–ion can be invoked. The minimum elements required to achieve these large effects appear to be that there are two strong, almost linear H-bonds between host and guest and that at least one of these is strongly perturbed by the electron transfer reaction. Based on the existing examples, the largest effects are observed if the affected H-bonds are an integral part of the redox couple, but significant effects are also possible when the redox couple is simply attached through conjugation with the H-bonding site. Other factors to take into consideration are the solvent and the electrolyte. For reductions, electrolytes with NBu4 þ as the cation appear to be good choices. For oxidations, especially with neutral guests, the best choices may be electrolytes with very large anions such as B(C6F5)4, since the more common electrolyte anions such as PF6, ClO4, and BF4 may cause interference. Another factor to be aware of is the possibility of proton transfer between host and guest. Looking toward the future, some basic work in electrochemically controlled Hbonding still remains to be done. In particular, the role of electrolyte needs to be investigated further. The number of different redox couples that have been used successfully is also still rather limited. Nonetheless, it is also clear that much of the groundwork has been laid, and that the next major steps will be to couple what has been learned about how electrochemistry can be used to control the strength of Hbonds in simpler systems with advances in H-bond assembly in solution, in order to construct larger, more sophisticated supramolecular structures operating under electrochemical control. Leigh and Paolucci’s work with molecular shuttles and Kaifer’s work with redox-active capsules are steps in this direction, but this is really just the beginning.
REFERENCES L. J. Prins, D. N. Reinhoudt, P. Timmerman, Angew. Chem., Int. Ed. 2001, 40, 2382. R. P. Sijbesma, E. W. Meijer, Chem. Commun. 2003, 5. P. S. Corbin, S. C. Zimmerman, J. Am. Chem. Soc. 2000, 122, 3779. K. A. Jolliffe, P. Timmerman, D. N. Reinhoudt, Angew. Chem., Int. Ed. 1999, 38, 933. J. RebekJr., Angew. Chem., Int. Ed. 2005, 44, 2068. F. Hof, S. L. Craig, C. Nuckolls, J. RebekJr., Angew. Chem., Int. Ed. 2002, 41, 1488. D. T. Bong, T. D. Clark, J. R. Granja, M. R. Ghadiri, Angew. Chem., Int. Ed. 2001, 40, 988. C. A. Nijhuis, B. J. Ravoo, J. Huskens, D. N. Reinhoudt, Coord. Chem. Rev. 2007, 251, 1761. 9. G. A. Jeffrey, An Introduction to Hydrogen Bonding, Oxford University Press, New York, 1997. 10. J. Bu, N. D. Lilienthal, J. E. Woods, C. E. Nohrden, K. T. Hoang, D. Truong, D. K. Smith, J. Am. Chem. Soc. 2005, 127, 6423. 1. 2. 3. 4. 5. 6. 7. 8.
REFERENCES
31
11. J. E. Woods, Y. Ge, D. K. Smith, J. Am. Chem. Soc. 2008, 130, 10070. 12. S. R. Miller, D. A. Gustowski, Z. H. Chen, G. W. Gokel, L. Echegoyen, A. E. Kaifer, Anal. Chem. 1988, 60, 2021. 13. M. Gomez, F. J. Gonzalez, I. Gonzalez, Electroanalysis 2003, 15, 635. 14. M. Gomez, F. J. Gonzalez, I. Gonzalez, J. Electrochem. Soc. 2003, 150, E527. 15. Y.-M. Legrand, M. Gray, G. Cooke, V. M. Rotello, J. Am. Chem. Soc. 2003, 125, 15789. 16. J. D. Carr, S. J. Coles, M. B. Hursthouse, M. E. Light, J. H. R. Tucker, J. Westwood, Angew. Chem., Int. Ed. 2000, 39, 3296. 17. N. Gupta, H. Linschitz, J. Am. Chem. Soc. 1997, 119, 6384. 18. J. E. Woods, D. K. Smith, unpublished results. 19. Y. Ge, L. Miller, T. Ouimet, D. K. Smith, J. Org. Chem. 2000, 65, 8831. 20. R. J. LeSuer, W. E. Geiger, Angew. Chem., Int. Ed. 2000, 39, 248. 21. F. Barriere, W. E. Geiger, J. Am. Chem. Soc. 2006, 128, 3980. 22. P. D. Beer, Acc. Chem. Res. 1998, 31, 71. 23. P. D. Beer, P. A. Gale, Angew. Chem., Int. Ed. 2001, 40, 486. 24. C. H. Park, H. E. Simmons, J. Am. Chem. Soc. 1968, 90, 2431. 25. P. D. Beer, A. D. Keefe, J. Organomet. Chem. 1989, 375, C40. 26. P. D. Beer, D. Hesek, J. Hodacova, S. E. Stokes, J. Chem. Soc., Chem. Commun. 1992, 270. 27. P. D. Beer, C. Hazlewood, D. Hesek, J. Hodacova, S. E. Stokes, J. Chem. Soc., Dalton Trans. 1993, 1327. 28. P. D. Beer, M. G. B. Drew, A. R. Graydon, D. K. Smith, S. E. Stokes, J. Chem. Soc., Dalton Trans. 1995, 403. 29. F. Oton, A. Tarraga, A. Espinosa, M. D. Velasco, P. Molina, J. Org. Chem. 2006, 71, 4590. 30. C. Valerio, J.-L. Fillaut, J. Ruiz, J. Guittard, J.-C. Blais, D. Astruc, J. Am. Chem. Soc. 1997, 119, 2588. 31. D. Astruc, M.-C. Daniel, J. Ruiz, Chem. Commun. 2004, 2637. 32. M.-C. Daniel, J. Ruiz, S. Nlate, J.-C. Blais, D. Astruc, J. Am. Chem. Soc. 2003, 125, 2617. 33. E. Breinlinger, A. Niemz, V. M. Rotello, J. Am. Chem. Soc. 1995, 117, 5379. 34. G. Cooke, J. Garety, S. Mabruk, V. Rotello, G. Surpateanu, P. Woisel, Chem. Commun. 2004, 2722. 35. J. B. Carroll, G. Cooke, J. F. Garety, B. J. Jordan, S. Mabruk, V. M. Rotello, Chem. Commun. 2005, 3838. 36. G. Cooke, H. A. de Cremiers, F. M. A. Duclairoir, J. Leonardi, G. Rosair, V. M. Rotello, Tetrahedron 2003, 59, 3341. 37. E. C. Breinlinger, V. M. Rotello, J. Am. Chem. Soc. 1997, 119, 1165. 38. A. K. Boal, V. M. Rotello, J. Am. Chem. Soc. 1999, 121, 4914. 39. G. Cooke, F. M. A. Duclairoir, P. John, N. Polwart, V. M. Rotello, Chem. Commun. 2003, 2468. 40. G. Cooke, J. F. Garety, B. Jordan, N. Kryvokhyzha, A. Parkin, G. Rabani, V. M. Rotello, Org. Lett. 2006, 8, 2297. 41. C. Bourgel, A. S. F. Boyd, G. Cooke, H. Augier de Cremiers, F. M. A. Duclairoir, V. M. Rotello, Chem. Commun. 2001, 1954. 42. G. Cooke, Y.-M. Legrand, M. Rotello Vincent, Chem. Commun. (Camb.) 2004, 1088.
32
ELECTROCHEMICALLY CONTROLLED H-BONDING
43. M. Gray, A. J. Goodman, J. B. Carroll, K. Bardon, M. Markey, G. Cooke, V. M. Rotello, Org. Lett. 2004, 6, 385. 44. A. S. F. Boyd, J. B. Carroll, G. Cooke, J. F. Garety, B. J. Jordan, S. Mabruk, G. Rosair, V. M. Rotello, Chem. Commun. 2005, 2468. 45. A. Niemz, J. Imbriglio, V. M. Rotello, J. Am. Chem. Soc. 1997, 119, 887. 46. T. Kajiki, H. Moriya, K. Hoshino, T. Kuroi, S. I. Kondo, T. Nabeshima, Y. Yano, J. Org. Chem. 1999, 64, 9679. 47. Y. Ge, R. R. Lilienthal, D. K. Smith, J. Am. Chem. Soc. 1996, 118, 3976. 48. R. Deans, A. Niemz, E. C. Breinlinger, V. M. Rotello, J. Am. Chem. Soc. 1997, 119, 10863. 49. A. Niemz, V. M. Rotello, J. Am. Chem. Soc. 1997, 119, 6833. 50. M. Gray, A. O. Cuello, G. Cooke, V. M. Rotello, J. Am. Chem. Soc. 2003, 125, 7882. 51. A. O. Cuello, C. M. McIntosh, V. M. Rotello, J. Am. Chem. Soc. 2000, 122, 3517. 52. J. B. Carroll, M. Gray, K. A. McMenimen, D. G. Hamilton, V. M. Rotello, Org. Lett. 2003, 5, 3177. 53. S. I. Kato, T. Matsumoto, K. Ideta, T. Shimasaki, K. Goto, T. Shinmyozu, J. Org. Chem. 2006, 71, 4723. 54. A. Altieri, F. G. Gatti, E. R. Kay, D. A. Leigh, D. Martel, F. Paolucci, A. M. Z. Slawin, J. K. Y. Wong, J. Am. Chem. Soc. 2003, 125, 8644. 55. G. Fioravanti, N. Haraszkiewicz, E. R. Kay, S. M. Mendoza, C. Bruno, M. Marcaccio, P. G. Wiering, F. Paolucci, P. Rudolf, A. M. Brouwer, D. A. Leigh, J. Am. Chem. Soc. 2008, 130, 2593. 56. Y. Ge, D. K. Smith, Anal. Chem. 2000, 72, 1860. 57. G. Cooke, V. Sindelar, V. M. Rotello, Chem. Commun. 2003, 752. 58. J. B. Carroll, M. Gray, G. Cooke, V. M. Rotello, Chem. Commun. 2004, 442. 59. G. Cooke, J. Couet, J. F. Garety, C.-Q. Ma, S. Mabruk, G. Rabani, V. M. Rotello, V. Sindelar, P. Woisel, Tetrahedron Lett. 2006, 47, 3763. 60. C. Chan-Leonor, S. L. Martin, D. K. Smith, J. Org. Chem. 2005, 70, 10817. 61. B. J. Jordan, M. A. Pollier, L. A. Miller, C. Tiernan, G. Clavier, P. Audebert, V. M. Rotello, Org. Lett. 2007, 9, 2835. 62. J. Westwood, S. J. Coles, S. R. Collinson, G. Gasser, S. J. Green, M. B. Hursthouse, M. E. Light, J. H. R. Tucker, Organometallics 2004, 23, 946. 63. M. B. Nielsen, C. Lomholt, J. Becher, Chem. Soc. Rev. 2000, 29, 153. 64. L. M. Goldenberg, O. Neilands, J. Electroanal. Chem. 1999, 463, 212. 65. A. S. F. Boyd, G. Cooke, F. M. A. Duclairoir, V. M. Rotello, Tetrahedron Lett. 2003, 44, 303. 66. K. Moon, A. E. Kaifer, J. Am. Chem. Soc. 2004, 126, 15016. 67. H. Sun, J. Steeb, A. E. Kaifer, J. Am. Chem. Soc. 2006, 128, 2820.
CHAPTER 2
Molecular Motions Driven by Transition Metal Redox Couples: Ion Translocation and Assembling–Disassembling of Dinuclear Double-Strand Helicates VALERIA AMENDOLA and LUIGI FABBRIZZI Dipartimento di Chimica Generale, Universita di Pavia, Pavia, Italy
2.1
INTRODUCTION
There exists a current interest in the design of molecular systems capable of converting chemical energy into an intramolecular movement, thus producing mechanical work. These systems have been defined as molecular machines and their synthesis and investigation represent one of the most active and appealing areas of supramolecular chemistry.1 The energy necessary for promoting the motion is provided by an ancillary reaction: when this reaction has a redox nature, being carried out either chemically or electrochemically, we have the molecular equivalent of the electrical machines of the macroscopic world. In order to perform controlled and reversible movements and to behave as a machine, the envisaged molecular system should have a mobile and a fixed component: one of the components should be redox active and the oxidized and reduced states should have almost comparable stability and should be connected by a reversible, and possibly fast, electron transfer process. The two oxidation states should display a different topological affinity with respect to the other component, so that a redox change can induce a modification of the topology of the whole molecular system, generating an intramolecular motion. The occurrence of fast and reversible movements also requires that the interaction between the mobile and the fixed part is based on Electrochemistry of Functional Supramolecular Systems. Edited by Paola Ceroni, Alberto Credi, and Margherita Venturi Copyright 2010 John Wiley & Sons, Inc.
33
34
MOLECULAR MOTIONS DRIVEN BY TRANSITION METAL REDOX COUPLES
a labile interaction: electrostatic, hydrogen bonding, p–p donor–acceptor, and metal– ligand (involving labile metal centers). First redox-driven intramolecular movements (and first electrochemical machines) were reported in 1994 and involved a rotaxane held by p–p interactions2 and a catenane based on metal–ligand interactions.3 In the Stoddart’s rotaxane (1),2 the wheel (A, conventionally assumed as the mobile component) had a strong p-acceptor nature, while the axle (the fixed component) contained two unequivalent stations of varying p-donor properties: a 4,40 -benzidine (D1, more p-donor) and 4,4-dimethoxydiphenyl (D2, less p-donor). Under unperturbed conditions, A stays on D1. However, D1 is redox active, undergoing fully reversible one-electron oxidation to D1 þ . Thus, on chemical or electrochemical oxidation to D1 þ , the station loses its donor tendencies and the wheel A conveniently moves to D2. Then, on addition of a proper reducing agent or adjusting the working electrode to the appropriate cathodic potential, D1 þ reduces to D1 and A moves back to the more donating station, resetting the system. In this machine, the D1/D1 þ couple is the engine and the auxiliary redox reaction provides the fuel.
Another redox-driven intramolecular movement involved the half-turn of one ring of an asymmetric catenane.3 The Sauvage’s catenane consisted of two intertwined rings, one containing a phen fragment (2) and the other containing both a phen and a terpy fragment (3).
The two interlocked rings are held together through the coordinative interactions to a copper center. When copper is in the þ 1 oxidation state, it is coordinated by two phen subunits, one from each ring. Such a coordinative arrangement is determined by the preference of the posttransition cation CuI (d10 electronic configuration) toward four-coordination, according to a tetrahedral geometry. On one-electron oxidation, the transition cation CuII (electronic configuration d9) forms, which tends to increase its coordination number, to 5 in the present case, through the coordination of the phen subunit of ring 2 and the terpy subunit of ring 3. This promotes a half-turn of ring 3, thus
2.1
INTRODUCTION
35
Figure 2.1 A square scheme illustrating a redox-driven intramolecular motion. Species with an asterisk (Ox* and Red*) are metastable and tend to rearrange to their stable topological isomer (Ox and Red).
giving rise to a controlled intramolecular motion. On CuII-to-CuI reduction, the halfturn is reversed and the system resets. In the present case, the engine of the electrical machine is represented by the CuI/CuII couple, fueled by an auxiliary redox reduction. The pioneering papers by Stoddart and Sauvage have stimulated the design of a variety of movable rotaxanes and catenanes, whose controlled motion is promoted by a redox change. In all cases, the process of the redox-driven intramolecular motion can be described by a square scheme, as illustrated in Fig. 2.1. It is assumed that the species Red at the top left corner of the scheme is the stable form of the supramolecular system, under unperturbed conditions (e.g., rotaxane 1, with the wheel placed on the benzidine station).2 On one-electron oxidation (step (i)), a metastable species is generated, Ox* (e.g., 1 þ , with the wheel still located on the benzidinium radical cation fragment), which thermally evolves to the stable form Ox, according to the “translational” step (ii). On reduction of Ox (step (iii)), the metastable species Red* forms (1 with the wheel on the dimethoxydiphenyl station), which undergoes thermal rearrangement to Red, according to a reverse “translational” step. The direct movement consists of the consecutive steps (i) and (ii) and its rate is determined by the slowest step. Electron transfer processes are in general fast; thus, the rate of the intramolecular motion is in most cases controlled by the “translational” step (ii). This step is thermal in nature and is affected by a kinetic barrier, whose magnitude is related to the steric interactions between the mobile and the fixed component. Sliding of wheel A from D1 to D2 and the D2-to-D1 reverse motion are fast processes, probably due to the flexibility of the aliphatic segment connecting the two stations. On the other hand, in the Sauvage’s catenate,3 the translational step (ii), which follows the CuI-to-CuII oxidation, can take hours. This may be due to the fact that the half-turn of ring 3 primarily involves the dissociation of one or more CuIIN bonds, a rather endothermic process due to the loss of ligand field (LF) energy experienced by the transition cation CuII. On the other hand, the reverse movement is fast because the d10 metal center CuI does not lose LF energy on CuIN bond dissociation. The square scheme illustrated in Fig. 2.1 applies to any redox process followed by a conformational rearrangement and was first discussed in detail considering the redox-driven linkage isomerization of the sulfoxide ligand associated with the RuIIRuIII change in [RuII,III(NH3)5(sulfoxide)]2 þ /3 þ complexes (S-coordination to RuII and O-coordination to RuIII).4 There exist other types of redox-driven intramolecular motions that can be interpreted on the basis of the square scheme of Fig. 2.1 and are promoted by
36
MOLECULAR MOTIONS DRIVEN BY TRANSITION METAL REDOX COUPLES
a metal-centered redox couple. They include, among the others, (a) the translocation of ions (either metal ions or anions) within ditopic ligands and receptors and (b) the assembling and disassembling of double helices. Representative examples of these processes will be described in the following sections. 2.2 2.2.1
ION TRANSLOCATION Metal Translocation
In the most general situation, a redox-active metal ion is translocated from a given site to another site of the same molecular system, following a chemical (a redox reaction) or an electrochemical input. The redox-driven reversible translocation of a metal ion in a two-component molecular system is schematically sketched in Fig. 2.2. The box in Fig. 2.2 symbolizes a system containing two distinct and separate ligating compartments, A and B, which differ in their coordinating properties. In particular, A is a hard receptor and B a soft one. Given a metal M possessing two adjacent oxidation states of comparable stability, Mn þ of soft nature and M(n þ 1) þ hard in character, in a solution containing equimolecular amounts of the reduced metal Mn þ and the two-compartment system A–B, Mn þ will occupy the soft compartment B. But, if Mn þ is oxidized, chemically or electrochemically, the hard ion that forms, M(n þ 1) þ , will find it thermodynamically convenient to move from B to the nearby hard compartment A. Consecutive oxidation and reduction processes would make the metal center M shuttle back and forth, between A and B, along a defined route. The square scheme of Fig. 2.1 applies to the translocation process. In particular, the rates of the translocation processes should depend (i) on the nature of the coordinative bonds between M and receptors A and B, whether kinetically labile or inert, and (ii) on the feasibility of the conformational rearrangement associated with the metal displacement. In the present case, the metal center plays two roles: that of the mobile part and that of the engine, fueled by the reaction with an oxidizing or a reducing agent added to the solution. Two redox couples that fit well the prerequisites mentioned earlier are FeIII/FeII and CuII/CuI. 2.2.1.1 Metal Translocation Based on the FeIII/FeII Couple The first example of redox-driven translocation of a metal center was based on the FeIII/FeII couple and took place in ditopic ligands containing (i) a tris-hydroxamate compartment and (ii) a tris-(2,20 -bipyridine) compartment.5
Figure 2.2
The redox switched translocation of a metal ion within a two-compartment ligand.
2.2
37
ION TRANSLOCATION
The bidentate hydroxamate (4), containing two oxygen donor atoms and having a negative charge, can be considered a hard ligand and forms very stable 1 : 3 octahedral complexes with tripositive metal ions such as FeIII (d5 electronic configuration, highspin state). The log b3 value for the [FeIIIL] (L ¼ 4, R1 ¼ CH3, R2 ¼ H, acetohydroxamate) in water is 28.33.6 It is suggested that metal coordination stabilizes the mesomeric form 4b, in which both oxygen donor atoms retain a negative charge. On the other hand, the bpy ligand gives rise to especially stable [MII(bpy)3]2 þ complexes of octahedral geometry with the FeII cation (d6, low spin). The high stability is due to the extended back-donation from the filled dp orbitals of the metal to the low-energy p* molecular orbital of bpy molecules. The two-compartment ligand 5 consists of three strands, each containing a hydroxamate subunit and a bpy subunit, covalently linked to a tripodal platform. In the system illustrated in Fig. 2.3, the platform is provided by the tetramine tren.
N N
N
N N
N
N
N N
N
O
NH O
O
HN
O
O N
O
O
O O
+ e–
O
O
NH
HN
NH
O
O
–e – N
HO
N
HN N OH
HO
O NH
O
HN
N
5a (FeIII)
O
N N
O NH
O
O
NH
NH
HN
O
N
5b (FeIII)
Figure 2.3 Translocation of an iron center within a two-compartment ligand, driven by the FeIII/FeII redox couple. FeIII prefers the inner compartment, which provides six oxygen donor atoms and retains a triply negative charge; FeII chooses the peripheral compartment consisting of three bpy subunits. Consecutive chemical reduction and oxidation makes the metal move back and forth between the two compartments.
38
MOLECULAR MOTIONS DRIVEN BY TRANSITION METAL REDOX COUPLES
On addition of 1 equivalent of FeIII to a MeOH/H2O solution of the ditopic ligand 5, the solution takes a light brown color with formation of a rather intense band centered at 420 nm (e ¼ 2400 M1 cm1). These optical features pertain to FeIII tris-hydroxamate complexes and show that the metal ion has occupied the tris-hydroxamate compartment. On addition of ascorbic acid, the solution turns violet, while an intense band centered at 540 nm develops (e ¼ 4700 M1 cm1). Such a band results from a metal-to-ligand charge transfer transition, typically occurring in the [FeII(bpy)3]2 þ chromophore. Thus, the FeIII-to-FeII reduction process promotes the translocation of the iron center from the inner compartment (complex 5a) to the peripheral one (5b), as pictorially illustrated in Fig. 2.3. The 5a-to-5b translocation process has a lifetime t ¼ 22 s. The reverse process can be induced on oxidation of the FeII center with S2O82, but it is remarkably slower, taking place over a period of minutes, at 70 C! The sluggishness of the process must be probably ascribed to the complexity of the conformational changes experienced by the three strands. On the other hand, the extreme slowness of the 5a-to-5b translocation may be due to the intrinsic sluggishness of the redox processes involving peroxydisulfate ion as an oxidizing agent.
Iron translocation based on the FeIII/FeII redox couple has been later observed in the ditopic ligand 6,7 which is structurally similar to 5. In analogy with the ditopic ligand 5, system 6 consists of three equivalent strands, covalently linked to a tren platform. Each strand contains a salicylamide group and a bpy subunit. Salicylamide easily deprotonates to give a bidentate ligand whose oxygen donor atoms retain a negative charge, delocalized on both through a p mechanism (see mesomeric forms 7a and 7b).
2.2
ION TRANSLOCATION
39
Compartment A consists of three salicylamide groups: each group behaves as a negatively charged ligand and offers two oxygen donor atoms. Thus, compartment A exhibits a definitely hard nature and is expected to display a specific affinity toward high-spin FeIII. Compartment B, like in the previously considered system 5, consists of three 2,20 -bipyridine fragments and establishes strong coordinative interactions with low-spin FeII. The 1:1 complex of FeIII and 6, in a DMF/H2O solution, shows an orange color (lmax ¼ 460 nm, e ¼ 3200 M1 cm1), as expected for a tris-salicylamide FeIII complex, thus indicating that FeIII resides in compartment A. On addition of a reducing agent (e.g., ascorbic acid), the solution gradually turns violet, while an absorption band develops (lmax ¼ 574 nm, e ¼ 2300 M1 cm1, with a shoulder at 543, 2250). These spectral features are those expected for a [FeII(bpy)3]2 þ chromophore and demonstrate that, on reduction, the metal center has moved to compartment B. On addition, at room temperature, of the strong oxidizing agent H2O2, a violet solution of the FeII complex again takes an orange color, due to the formation of the FeIII tris-salicylaldimine complex. Hydrogen peroxide provides on reduction hydroxide ions (H2O2 þ 2e ¼ 2OH), which favor the deprotonation of the OH group of each salicylate moiety and consequent coordination of the phenolate oxygen atom to FeIII. Kinetic studies showed that translocation processes in either direction take place on the 10 s timescale. The higher rate of the A-to-B translocation process can be ascribed to the greater flexibility of system 6 with respect to 5. Moreover, the use of the very reactive oxidizing agent H2O2, compared to S2O82, may account for the much faster back-translocation occurring within the ditopic receptor 6 with respect to 5. A FeIII/FeII-driven translocation process of different nature is illustrated in Fig. 2.4.8 The ligand 8 consists of a 4-methylphenol platform, to which two different terdentate subunits have been appended in 2- and 6-positions. One appendance consists of a tertiary amine nitrogen atom and two phenolate oxygen atoms (deprotonation of all the phenolic groups of 8 is guaranteed by the presence of the base 2,4,6trimethylpyridine (collidine) in the MeCN solution). The other appendance possesses one tertiary amine nitrogen atom and two pyridine nitrogen atoms. When 1 equivalent of FeIII(ClO4)3 is added to a MeCN solution of 8, in the presence of collidine, the FeIII cation seeks for the coordination of the three available oxygen atoms, one from the platform and two from the NO2 appendance. Six-coordination, to give the preferred octahedral geometrical arrangement, probably distorted because of the steric constraints, is achieved by coordination of the nitrogen atom of the NO2 appendance and two MeCN molecules (indicated with S in formula 8a in Fig. 2.4). On the other hand, if stoichiometric FeII(ClO4)2 is added to a solution of 8 containing collidine, the FeII center looks for the coordination of the three nitrogen atoms of the N3 appendance, while the phenolate oxygen atom of the platform and two solvent molecules complete
40
MOLECULAR MOTIONS DRIVEN BY TRANSITION METAL REDOX COUPLES
Figure 2.4 The “pendular” motion of an iron center. The left compartment provides a donor set consisting of three negatively charged phenolate oxygen atoms, a tertiary amine nitrogen atom, and two solvent molecules (S ¼ MeCN) and hosts FeIII. The right compartment offers as donors a phenolate oxygen atom, one tertiary amine nitrogen atom, two pyridine nitrogen atoms, and two solvent molecules and is suitable for the coordination of FeII. The central phenolate oxygen atom is shared by the two compartments and acts as a pivot in the translocation process.
six-coordination. The above evidence provides the basis for the occurrence of a redox-driven translocation of the iron center between the two ligating compartments, which share the central phenolate oxygen atom. Indeed, translocation was carried out electrochemically and was investigated through cyclic voltammetry (CV) experiments at a platinum working electrode, in a MeCN solution made 0.1 M in [Bu4N]ClO4. The CV experiment (Fig. 2.5) starts at a potential of 800 mV (a), where the FeII ion stays in the right compartment (8b in Fig. 2.4), which provides the donor set [N3O þ 2S]. On increasing the potential, oxidation to FeIII takes place, with
b FeIII
Current
oxidation
–0.8
c
–0.4
0
Potential (V)
a
reduction
FeII d
Figure 2.5 Cyclic voltammogram of a MeCN solution equimolar (4 103 M) in 8, FeII(ClO4)2, and collidine. Working electrode: platinum disk (5 mm diameter); supporting electrolyte: [Bu4N]ClO4; scan rate: 0.1 V/s; reference electrode: AgNO3/Ag in MeCN. Diagram adapted from Ref. 8.
2.2
ION TRANSLOCATION
41
Figure 2.6 A square scheme illustrating the “pendular” motion of an iron center, driven by the FeII/FeIII redox couple. As judged from voltammetric experiment carried out at varying potential scan rate, the lifetime t for both translocation processes is <10 ms.
a current peak at ca. 200 mV (b). At þ 100 mV (c), the potential scan is reversed: in the presence of an electrochemically reversible process (Dp 60 mV), one would expect a reduction peak at ca. 260 mV. However, no peak is observed around this potential value. In fact, in the meantime, the FeIII ion has moved to the left compartment (8a in Fig. 2.4), which provides the more favorable [NO3 þ 2S]3 donor set. However, this donor set is less propitious for the FeII cation, so that the FeIII-to-FeII reduction is made more difficult and takes place at a more negative potential: reduction peak (d) at ca. 700 mV. After this point, FeII moves back to its preferred compartment with donor set [N3O þ 2S] and the system is ready for a further cycle. The CVexperiment provides an estimate of the time of the translocation. At a rate of 100 mV/s, the scan from c to d (700 mV) takes place in 7 s. Thus, the translocation of FeII from the [NO3 þ 2S]3 compartment to [N3O þ 2S] compartment takes place in less than 7 s. CV studies at high scan rates indicated that the lifetime t of the translocation is lower than 10 ms. The high translocation rate can be ascribed to the beneficial assistance of the central phenolate oxygen atom, which keeps the iron center coordinated over the course of the direct and reverse motion, thus reducing the energy of the transition state. The voltammetric behavior can therefore be accounted for on the basis of the square scheme in Fig. 2.6. 2.2.1.2 Metal Translocation Based on the Cu II/Cu I Couple On paper, the most favorable couple for promoting a redox-driven metal translocation should be CuII/CuI. In fact, the uptake/release of a single electron makes the metal cross the border between the rich and diversified realm of transition metals (CuII) and the monotonous and foreseeable land of posttransition metals (CuI). In particular, the CuII center (electronic configuration d9) is ready to adapt itself to any coordination number (4, 5, or 6) and geometry offered by the coordinating environment, in order to benefit most from ligand field stabilization energy (LFSE); a further energy gain is made possible by subtle structural distortions related to the Jahn–Teller effect. On the other hand, the CuI ion (d10) is inclined to one coordination geometry, the tetrahedral one, an obligatory choice dictated by interligand steric repulsions. Probably for the
42
MOLECULAR MOTIONS DRIVEN BY TRANSITION METAL REDOX COUPLES
Figure 2.7 The hypothesized translocation of a heteroditopic cylindrical ligand driven by the CuII/CuI redox couple.
above-mentioned reasons, the first “planned” redox-driven metal translocation was based on the CuII/CuI couple. In particular, Lehn anticipated that a cylindrical ditopic ligand of type 9 could provide the ideal condition for the redox-driven translocation of a copper center.9 The soft CuI cation is expected to prefer the coordination by the N2S2 tetradentate macrocycle, in view of its affinity toward thioethereal sulfur atoms (more precisely, the capability of the d10 center to back-donate electron density to the sulfur atom through a dp–dp mechanism). On the other hand, the CuII transition metal ion will prefer to be coordinated by the N2O3 quinquedentate macrocycle, in order to benefit from a higher coordination number: 5. Even if quite reasonable, the translocation process illustrated in Fig. 2.7 was never carried out. In any case, occurrence of the back and forth motion of the copper center would require that the two spacers linking the macrocycles have a rather flexible nature, in order to permit occasional binding of the metal center, either in the þ 1 or in the þ 2 oxidation state, to be partially coordinated by donor atoms of the two macrocycles. Rigid spacers would imply a transition state of extremely high energy and would make the “jumping” quite improbable. A fast and reversible copper translocation driven by the CuII/CuI couple was carried out within the flexible ditopic receptor 10.10 The translocation process, illustrated in Fig. 2.8, is fast and reversible and can be followed both visually and spectrophotometrically. In particular, a MeCN solution equimolar in both 10 and CuII is blue-violet (metal-centered absorption band: lmax ¼ 548 nm, e ¼ 120 M1 cm1), which indicates that the oxidized cation is located in the tetramine compartment, where it experiences a square coordination geometry (10a). On addition of a reducing agent (e.g., ascorbic acid), the color of the solution turns brick red, as observed with the model complex [CuI(bpy)2] þ , while an intense band develops at a lower wavelength (MLCT transition: lmax ¼ 430 nm, e ¼ 1450 M1 cm1). This clearly indicates that the CuII/CuI reduction process has taken place and that the reduced metal center has moved to the (bpy)2 compartment (10b). On addition of an oxidizing agent (e.g., H2O2), the solution again takes its original blue-violet color, indicating that, following the CuI-to-CuII redox change, the oxidized metal center has moved back to the tetramine compartment.
2.2
ION TRANSLOCATION
43
Figure 2.8 Redox-driven translocation of a copper center, based on the CuII/CuI change. The CuII ion stays in the tetramine compartment of the ditopic ligand 10, whereas the CuI ion prefers to occupy the bis-(2,20 -bipyridine) compartment. The translocation of the copper center between the two compartments is fast and reversible when carried out through the CuII-toCuI reduction with ascorbic acid and CuI-to-CuII oxidation with H2O2, in a MeCN solution.
The complete process can be interpreted on the basis of a square scheme, as shown in Fig. 2.1. Quickness of the “mechanical” steps (iv) and (ii) is ensured by the high flexibility of the molecular framework. Molecular movement associated with the CuII/CuI redox change has been described by Shanzer11 and Canary.12 However, in the described systems, a definite displacement of the metal center is not observed and the processes can be more properly considered a conformational change of the ligating framework, which reorganizes in order to fulfill the coordinative requirements of the metal center in the two oxidation states. 2.2.2
Anion Translocation
2.2.2.1 Anion Translocation Between Metal Centers Driven by the Ni III/Ni II Couple Anions can be translocated within a ditopic system, by benefiting from a gradient of redox potential. In the case of previously considered metal translocation processes, the engine was the transported particle itself, through a metal-centered M(n þ 1)/Mn þ couple. This cannot be the case of anions, which can be redox active, for example, halides, through the X_/X couple, but oxidation makes the particle lose its electrical charge and its ionic nature. Thus, in the reported examples of redox-driven anion translocation, a determining role was played by a metal center, which acted both as an engine (through the M(n þ 1)/Mn þ couple) and as a receptor for the envisaged anion.13,14
System 11 consists of two tetramine ligating compartments, tren (tris(2-aminoethyl)amine)) and cyclam (1,4,8,11-tetraazacyclotetradecane), that display different coordinating tendencies toward metals and are separated by a 1,4-xylyl spacer.14
44
MOLECULAR MOTIONS DRIVEN BY TRANSITION METAL REDOX COUPLES
Consecutive reaction of 11 with 1 equivalent of NiII(ClO4)2 and 1 equivalent of Cu (ClO4)2 under proper conditions leads to the formation of a heterodimetallic complex in which the NiII ion is firmly encircled by the tetraaza macrocycle, whereas CuII is coordinated by the tripodal tetramine. If, in a 103 M solution of the dimetallic complex, 1 equivalent of tetraalkylammonium chloride is added, the Cl ion seeks the CuII center to form a five-coordinate ternary complex. In fact, the CuII(tren)2 þ subunit exhibits a well-defined tendency to form [CuII(tren)X] þ complexes of axially compressed trigonal bipyramidal geometry (X ¼ halide, pseudohalide). Preliminary spectrophotometric titration experiments had allowed to determine a binding constant log K ¼ 5.66 0.09. This means that, at 103 M concentration level, when 1 equivalent of anion is added, 95% is bound to the CuII(tren)2 þ subunit of the [CuII–NiII]4 þ conjugate system. The NiII(cyclam)2 þ moiety does not compete at all for the anion because the d8 metal center is in the low-spin state and is completely suitable for the square coordination provided by the cyclam macrocycle. On the other hand, the NiII center, when encircled by cyclam, undergoes one-electron oxidation to the NiIII state at a moderately positive potential.15 The so-formed d7 cation, in the low-spin state, displays a strong preference toward six-coordination, according to an axially elongated octahedral geometry. Thus, it is expected that on NiII-to-NiIII oxidation the chloride ion leaves the CuII center and moves to occupy one of the available axial positions of the NiIII macrocyclic subunits, while the other one is taken by a solvent molecule. The process is illustrated in Fig. 2.9 (11a K 11b). Occurrence of the reversible X translocation process is afforded by the following sequence of anion affinity: NiIII > CuII > NiII and can be monitored through voltammetric titration experiments. In particular, a differential pulse voltammetry (DPV) scan was carried out on a MeCN solution 103 M in the [CuII(11)NiII](ClO4)4 complex salt and 0.1 M in [Bu4N]ClO4, using a microsphere platinum electrode. The [CuII–NiII]4 þ system II
Figure 2.9 Redox-driven translocation of the anion X (e.g., chloride), based on the NiII/NiIII change. The nickel center acts both as an engine and as a receptor for the X anion (when in the NiIII state). Occurrence of the reversible X translocation is afforded by the following sequence of anion affinity: NiIII > CuII > NiII.
2.2
ION TRANSLOCATION
45
Figure 2.10 Differential pulse voltammetry profiles obtained at a platinum working electrode in a MeCN solution 5 104 in [Et3Bn]Cl and 0.1 M in [Bu4N]ClO4. Dashed line, 103 M in the [CuII(11)NiII](ClO4)4 complex salt; solid line, 103 M in both the [CuII(12)](ClO4)2 and [NiII(11)](ClO4)2 complex salts.
undergoes a one-electron oxidation at 0.74 V versus Fc þ /Fc, which corresponds to the NiII-to-NiIII change. On addition of Cl, a new peak develops at a much less positive potential (0.24 V), while the intensity of the peak at 0.74 V progressively decreases. In particular, the peak intensity at 0.24 V reaches a plateau with the addition of 1 equivalent of Cl; further anion addition does not induce the appearance of any other peak. Figure 2.10 (dashed line) displays the DPV profile recorded on addition of 0.5 equivalent of chloride. Such a voltammetric behavior can be explained on assuming that on NiII-to-NiIII oxidation a Cl ion leaves the CuII center and occupies an axial position on the NiIII center. In particular, the voltammetric response is based on the thermodynamic cycle illustrated in Fig. 2.11.
(2.1)
(2.2) Figure 2.11 Thermodynamic cycle that connects equilibrium (2.1) (the NiII-to-NiIII redox change and chloride ion translocation) and equilibrium (2.2) (simple NiII-to-NiIII oxidation) through chloride complexation equilibria involving CuII and NiIII, in the conjugate system [CuII(11)NiII]4 þ .
46
MOLECULAR MOTIONS DRIVEN BY TRANSITION METAL REDOX COUPLES
Note that the horizontal equilibrium (2.1) in the cycle corresponds to the translocation process pictorially illustrated in Fig. 2.9. Moreover, F ½E ðClÞE ¼ F DEðVÞ ¼ RT ½log KðNiIII --ClÞ--log KðCuII --ClÞ As DE ¼ 0.5 V and log K(CuIICl) ¼ 5.7, at 25 C, log K(NiIIICl) ¼ 8.5, which fulfills the first requirement for anion translocation. On the other hand, 23.08 DE(V) gives the DG value (in kcal/mol) for the “neat” translocation equilibrium (2.3): ð2:3Þ In equilibrium (2.3), the chloride ion is transferred from the CuII center to the already oxidized NiIII cation. Such an equilibrium is characterized by a very favorable free energy change DG ¼ 11.5 kcal/mol. At this stage, one could argue whether process (2.3) is authentically intramolecular (i.e., the chloride which has moved on NiIII is really coming from CuII) or intermolecular (the chloride leaves CuII for the solution, while a different chloride ion from the solution binds NiIII).
In this connection, it would be useful to consider a similar process that is unquestionably intermolecular. In particular, we considered the complex of CuII with “component” ligand 12 (a N-benzyl-substituted tren derivative) and the NiII complex with the other “component” ligand 13 (a N-benzyl-substituted cyclam derivative). The N-benzyl substituent is there in order to reproduce in the separated complexes the coordinating environments present in the conjugate system 11. Then, we carried out a voltammetric titration on a MeCN solution 103 M both in [CuII(12)]2 þ and in [NiII(13)]2 þ . Prior to chloride addition, an oxidation peak appeared at 0.74 V, which corresponds to the NiII/NiIII couple within macrocycle 13 and superimposes exactly with the peak observed for the [CuII(11)NiII]4 þ complex. On chloride addition, the peak intensity at 0.74 V decreases, while a new peak develops at 0.57 V. Figure 2.10 displays the DPV profile obtained for a solution containing 0.5 equivalent of [Et3Bn]Cl (solid line). It has to be noted that the NiII-to-NiIII oxidation takes place at a much more positive potential than in the case of the [CuII(11)NiII]4 þ complex. ð2:4Þ
2.2
ION TRANSLOCATION
47
From peak separation DE (0.17 V), it is possible to calculate the DG value corresponding to the intermolecular equilibrium (2.4), in which a chloride ion is detached from the [CuII(12)]2 þ complex and goes through the solution to coordinate the metal center in the [NiIII(13)]3 þ complex (DG ¼ 3.9 kcal/mol), a quantity significantly less negative than observed for the “neat” translocation equilibrium (2.3). In particular, the “intramolecular” anion translocation equilibrium (2.3) is favored by an extra free energy of 7.6 kcal/mol with respect to the intermolecular anion transfer between the separated components. An interesting question is whether such an energy advantage has an enthalpic or an entropic origin. Very conveniently, DS values associated with the “neat” anion translocation equilibria (2.3) and (2.4) can be determined by carrying out voltammetric investigations at varying temperature. In fact, DE ¼
DH DS þ T F F
ð2:5Þ
Thus, DE should vary linearly with the temperature and DS could be calculated from the slope of the linear plot DE versus T. DPV experiments on the conjugate system [CuII(11)NiII]4 þ ( þ 0.5 equivalent of Cl) and on system {[CuII(12)]2 þ [NiII(13)]2 þ } ( þ 0.5 equivalent of Cl) were carried out over the 20 to 30 C temperature range. Pertinent DE versus T plots are shown in Fig. 2.12. It is observed that DE(3), for the conjugate complex [CuII(11)NiII]4 þ (open triangles in Fig. 2.12), does not vary along the investigated temperature interval, indicating that DS ¼ 0. Thus, DG ¼ DH ¼ 11.5 kcal/mol, which indicates that the
Figure 2.12 Temperature dependence of DE for MeCN solutions of (i) [CuII(11)NiII]4 þ (open triangles) and (ii) an equimolar mixture of [CuII(12)]2 þ and [NiII(13)]2 þ (filled triangles), in the presence of 0.5 equivalent of Cl. DE is the peak separation in the DPV profiles in Fig. 2.10 and refers to the neat anion translocation equilibria (2.3) and (2.4). The slope of the DE versus temperature straight line gives DS /F for pertinent equilibrium.
48
MOLECULAR MOTIONS DRIVEN BY TRANSITION METAL REDOX COUPLES
favorable free energy change associated with the neat translocation equilibrium (2.6) is solely due to a very exothermic thermal contribution. This reflects the especially high energy of the NiIIICl interaction within the cyclam subunit compared to the CuIICl bond in the tren moiety of the conjugate system [CuII(11)NiII]4 þ . On the other hand, the DE value measured for the {[CuII(12)]2 þ þ [NiII(13)]2 þ } system shows a significant linear decrease with the increasing temperature (filled triangles in Fig. 2.12), which indicates a negative value of the entropy change. In particular, from the slope of the least squares straight line the following value was calculated: DS ¼ 28 cal/mol/K. The algebraic combination of DG (3.9 kcal/mol) and TDS (8.3 kcal/mol, at 25 C) gives DH ¼ 12.2 kcal/mol, a value comparable to that pertaining to the neat anion translocation equilibrium involving the conjugate system (DH ¼ 11.5 kcal/mol). The enthalpy changes typically reflect the balance of bonding energy terms: the closeness of DH values for equilibria (2.3) and (2.4) suggests that the balance of the bonding terms (breaking of the CuIICl bond and forming of the NiIIICl bond) is nearly the same for both the conjugate system [CuII(11)NiII]4 þ and the two separate components [CuII(12)]2 þ and [NiII(13)]2 þ . It derives that the significant disadvantage experienced by the neat anion translocation process between separate components solely results from a negative entropy effect. It is suggested that such an effect has a probabilistic nature. In this connection, one should consider that anion translocation results from the collision of the CuII(tren)(Cl) þ subunit with the NiIII center both in the conjugate and in the separate component system. It is intuitive that the process is more probable in the covalently linked system [CuII(11)NiII]4 þ than in the {[CuII(12)]2 þ þ [NiII(13)]2 þ } mixture. In fact, in the [CuII(11)NiII]4 þ complex, useful collisions are mainly intramolecular and result from the occasional folding of the covalently linked system, which brings CuII(tren)(Cl) þ and NiIII(cyclam)3 þ subunits into contact. On the other hand, for separate components, collisions involve complexes dispersed in the solution and the probability of their occurrence is related to the concentration. The greater the concentration, the higher the probability that the two complexes collide. A semiquantitative comparison can be attempted in the following way. The distance between copper and nickel centers in the [CuII(11)NiII]4 þ complex is ca. 7.5 A, as calculated through molecular modeling. Thus, we can assume that within a sphere whose center is represented by NiIII and the CuIICl fragment moves 3 whose volume is 1766 A , that is, 1.766 1024 L. Thus, the concentration of the active CuIICl fragment is 1.766 1024 molecule/L, or, in the more familiar molar scale, 0.94 M. This concentration is 103–104 higher than that of the [CuII(12)]2 þ complex in the solution electrochemically or spectrophotometrically investigated and accounts for the higher probability (and the lower loss of entropy) of the intramolecular chloride translocation in the [CuII(11)NiIII]5 þ system with respect to intermolecular anion transfer from [CuII(13)Cl] þ to [NiIII(13)]3 þ . There are not many other anions undergoing a CuII-to-NiIII translocation within the conjugate system 11. Reducing anions such as Br, I, and NCS undergo oxidation at a potential lower or comparable to that of the NiII/NiIII couple, which prevents any translocation. NO3 and HSO4 are resistant to oxidation, but show a moderate affinity toward the CuII center. In particular, in a solution containing equimolar
2.3
ASSEMBLING/DISASSEMBLING OF HELICATE COMPLEXES
49
amounts of the dimetallic system [CuII(11)NiII]4 þ and the anion, only a small fraction of the anions is bound to the CuII center. Among inorganic anions relevant to coordination chemistry, only NCO displays a “regular” behavior. In fact, it is quite resistant to the oxidation and gives a stable complex with the reduced form of the covalently linked system, [CuII(NCO)(11)NiII]3 þ , with log K ¼ 4.4 0.1. It derives that, in a solution 103 M both in [CuII(11)NiIII]4 þ and in NCO, 82% of the anions are bound to CuII. Then, over the course of a DPV titration experiment, a new peak develops at 0.27 V versus Fc þ /Fc and reaches a plateau after the addition of 1 equivalent of NCO (while the peak at 0.74 V decreases and disappears). Also in the present case the DE value (0.47 V) is remarkably larger than DE measured for a solution containing the separate components (0.26 V). This substantiates the observation that the intramolecular anion translocation from CuII to NiIII within the covalently linked system [CuII(11)NiIII]5 þ is distinctly favored with respect to the intermolecular translocation between the separate components. The engine of the investigated “machine” is provided by the NiII/NiIII redox change. One could ask whether the ancillary metal center (CuII) could be replaced by another one. In this connection, we considered that ZnII tetramine complexes exhibit a definite tendency to form five-coordinate complexes with a given anion X. In particular, they show a pronounced affinity toward the carboxylate donor group. Spectrophotometric titration experiments with the benzoate anion, BzO, in a solution of the [ZnII(11)NiII]4 þ system in MeCN/MeOH (1:1 v/v) indicated the formation of a rather stable [ZnII(BzO)(11)NiII]3 þ complex (log K ¼ 5.6 0.1). However, DPV studies showed that benzoate addition neither altered the intensity of the peak at 0.74 V versus Fc þ /Fc (pertaining to NiII-to-NiIII oxidation) nor induced a shift of the peak. This suggests that the benzoate anion forms a more stable complex with the ZnII(tren)2 þ subunit than with the NiIII(cyclam)3 þ fragment. Thus, the required sequence of anion binding tendencies (NiIII > MII > NiII) is not fulfilled and the redox-induced anion translocation process cannot occur.
2.3 ASSEMBLING/DISASSEMBLING OF HELICATE COMPLEXES DRIVEN BY THE CuI/CuII COUPLE Molecule 14 is a tetradentate ligand consisting of two bpy moieties linked in 3- and 30 -positions by a CH2OCH2 spacer; the plain bpy ligand forms complexes of formula [MI(bpy)2] þ of tetrahedral geometry with d10 metal ions, such as CuI and AgI, whose extra stability results from the donation of electron density from filled dp orbitals of the metal to empty p* molecular orbitals of bpy molecules. One would expect that mixing of 1 equivalent of 14 with 1 equivalent of [CuI(MeCN)4]ClO4 in a MeCN solution would give the 1:1 complex, in which 14 acts as a quadridentate ligand and fully chelates the d10 metal center. However, this does not occur due to the steric constraints, which prevents the achievement of the tetrahedral coordination geometry required by the CuI center. Instead, metal and ligand find it more convenient to give a complex of 2 : 2 stoichiometry, [Cu2I(14)2]2 þ ,
50
MOLECULAR MOTIONS DRIVEN BY TRANSITION METAL REDOX COUPLES
Figure 2.13 The dinucleating bis-bidentate ligand 14 forms with MI metal ions of electronic configuration d10 (e.g., CuI, AgI) dimetallic complexes of formula [M2I(14)2]2 þ , in which two molecules of 14 are intertwined to give a double helix. Ligands of the type 14 are named helicands and complexes such as 15 are called helicates. In this particular case, we have a double-strand helicate.
in which each CuI center is coordinated by one half of each bis-bidentate ligand and can achieve the preferred tetrahedral coordination geometry without not too serious strain.16 The structure of the dicopper(I) complex is outlined in Fig. 2.13 (15): the two molecules of 14 are arranged as a double helix. Following the language of coordination chemistry, molecule 14 has been named a helicand and complex 15 a helicate. Since the seminal papers by Lehn,17 the chemistry of helicates has greatly developed to give systems of higher complexity, involving a larger number of metals of different coordination numbers and geometry. However, the aspect of helicate chemistry relevant to this chapter refers to the redox behavior of dinuclear copper(I) double-strand helicate complexes, in particular to the consequences of the CuI-to-CuII oxidation process on a [Cu2I(L \ L)2]2 þ complex, where L \ L represents a bisbidentate ligand such as 14. This aspect has been electrochemically investigated by several authors.18 In this chapter, we will consider a homogeneous series of systems studied in our laboratory, in order to offer a complete view of the topic.19 To begin, the dinucleating helicand 16 will be considered.
16 gives stable complexes with CuI and CuII, both in MeCN solution and in the solid state. In particular, the two complex salts [Cu2I(16)2](CF3SO3)2, dinuclear, and [CuII(16)](CF3SO3)2, mononuclear, were isolated in the crystalline form and corresponding crystal and molecular structures were determined through X-ray diffraction
2.3
ASSEMBLING/DISASSEMBLING OF HELICATE COMPLEXES
51
Figure 2.14 The molecular structure of the [Cu2I(16)2]2 þ double-strand helicate complex cation. CuI metal centers are represented as spheres. The hydrogen atoms of the two strands have been omitted for clarity. Structure redrawn from data deposited at the Cambridge Crystallographic Data Centre: CCDC 118958.
studies. Figure 2.14 shows the double-strand helix arrangement of the bis-bidentate ligand 16 in the [Cu2I(16)2]2 þ complex cation. Each CuI center (gray ball) is bound to an imine and to a pyridine nitrogen atom from each strand and shows a rather distorted tetrahedral coordination geometry. On the other hand, Fig. 2.15, which displays the molecular structure of the [CuII(16)](CF3SO3)2 salt, shows that the CuII ion prefers to form a mononuclear complex species. In particular, the transition metal ion CuII finds it convenient to reach tetragonal coordination through the chelation by a single molecule of 16, in order to better benefit from ligand field stabilization energy terms. On assuming that the geometrical features described above are maintained in solution, the CuII/CuI redox change in solution would result in an assembling–disassembling equilibrium, as pictorially illustrated in Fig. 2.16. The occurrence of the redox-driven reversible assembling–disassembling process involving copper complexes of 16 has been verified through cyclic voltammetry experiments at a platinum electrode in a MeCN solution. Figure 2.17 shows the CV profile obtained with a solution of the double-strand helicate complex [Cu2I(16)2]2 þ . The starting potential has been set at 300 mV (versus Fc þ /Fc), where the dinuclear complex is stable (a). On increasing the potential, a poorly defined oxidation peak develops (b), corresponding to the CuI-to-CuII oxidation process. At 900 mV, the potential is reversed. However, going down to 500 mV and less, the reduction peak opposite to peak b is not observed. In fact, the oxidation to CuII is followed by a fast
52
MOLECULAR MOTIONS DRIVEN BY TRANSITION METAL REDOX COUPLES
Figure 2.15 The molecular structure of the [CuII(16)]2 þ . The hydrogen atoms of ligand 16 have been omitted. The CuII center experiences a rather distorted square coordination geometry. Structure redrawn from data deposited at the Cambridge Crystallographic Data Centre: CCDC 118957.
disassembling process to give two mononuclear [CuII(16)]2 þ complexes. The CuII mononuclear complex, due to the preferred tetragonal coordination geometry, is especially stable to the reduction, which takes place at a much less positive potential (peak d, at 0 mV). Following reduction, the two CuI mononuclear complexes quickly reassemble to give the double-strand helicate species, closing the cycle. The corresponding square scheme is outlined in Fig. 2.18. The high irreversibility of the profile depends upon the fact that the assembling– disassembling process is too fast with respect to the timescale of the CV experiment
Figure 2.16 The redox-driven disassembling of a dicopper(I) double-strand helicate complex to give two mononuclear copper(II) complexes, in which each strand behaves as a quadridentate ligand. On subsequent reduction, the two mononuclear complexes reassemble to give the helicate. The illustrated process fits well the behavior of copper complexes of 16 in a MeCN solution.
2.3
ASSEMBLING/DISASSEMBLING OF HELICATE COMPLEXES
53
Figure 2.17 Cyclic voltammogram of a MeCN solution of [CuII(16)](CF3SO3)2. Supporting electrolyte: 0.1 M [Bu4N]ClO4; scan rate: 0.1 V/s; internal reference electrode: Fc þ /Fc. Diagram adapted from Ref. 20.
(the profile in Fig. 2.17 has been taken at a potential scan rate of 100 mV/s). However, the same irreversible profile was observed also at a potential scan rate of 20 V/s. Considering that CV experiments of the type illustrated in Fig. 2.17 cover an interval of ca. 1 V, it derives that the transient species [Cu2II(16)]4 þ (upper right corner in the square scheme of Fig. 2.18) and [CuI(16)] þ (lower left corner of square scheme) have a lifetime lower than 50 ms.
Figure 2.18 A square scheme illustrating the disassembling of the [Cu2I(16)2]2 þ double helicate complex, following CuI-to-CuII oxidation, and the consequent assembling of two [CuII(16)]2 þ mononuclear complexes, following the CuII-to-CuI reduction. The process ultimately derives from the geometrical coordinative preferences of the two oxidation states: CuI prefers a tetrahedral coordination, which can be achieved with the double helicate arrangement; CuII prefers a square coordination geometry, which is fulfilled by the coordination of a single molecule of 16.
54
MOLECULAR MOTIONS DRIVEN BY TRANSITION METAL REDOX COUPLES
However, it has been recently demonstrated by Pallavicini et al. that the lifetime of the dicopper(II) double-strand helicate [Cu2II(16)]4 þ can be significantly increased by introducing hindering substituents on the framework of 16. In particular, this was shown to occur with the copper complexes of the bis-bidentate ligand 17.21
The bidentate ligand 17 forms bis with CuI a double-strand helicate complex, whose structure was elucidated through X-ray diffraction studies and is shown in Fig. 2.19. The [Cu2I(17)2]2 þ complex displays geometrical features very similar to those of the [Cu2I(16)2]2 þ analogue: as an example, the CuICuI distance is 3.71 A, to be compared to 3.67 A. The CV behavior of the [Cu2I(17)2]2 þ complex was investigated in a MeCN solution made 0.1 M in [Bu4N]ClO4, using a platinum microsphere as a working
Figure 2.19 The molecular structure of the [Cu2I(17)2]2 þ double-strand helicate complex cation. CuI metal centers are represented as spheres. Hydrogen atoms of the two strands have been omitted for clarity. Structure redrawn from data deposited at the Cambridge Crystallographic Data Centre: CCDC 641164.
2.3
ASSEMBLING/DISASSEMBLING OF HELICATE COMPLEXES
55
Figure 2.20 Cyclic voltammogram of a MeCN solution of [Cu2I(17)2]2 þ double-strand helicate complex. Supporting electrolyte: [Bu4N]ClO4; scan rate: 0.2 V/s; internal reference electrode: Fc þ /Fc. Diagram adapted from Ref. 21.
electrode. Figure 2.20 shows the pertinent profile: the potential was set at 0.00 V versus the reference couple (Fc þ /Fc) and was scanned toward positive values. Very interestingly, two consecutive waves were observed, with peaks located at 190 and 420 mV. The first wave originates from the one-electron oxidation of one CuI center (to CuII) and the second wave from the one-electron oxidation of the second CuI ion. Peak separation DE ¼ 230 mV originates from the repulsive electrostatic effect exerted by the CuII ion on the oxidation of the proximate CuI center and from the statistical effect (36 mV). On the reverse scan, two distinct and well-shaped reduction waves develop at 330 and 90 mV, respectively (peak separation 240 mV). This unambiguously indicates that on oxidation of the dicopper(I) helicate a dicopper(II) helicate forms according to two distinct one-electron oxidation steps. The [Cu2II(17)2]4 þ helicate complex is stable over the timescale of the CV experiment, even when performed at a potential scan rate as low as 20 mV/s. This indicates for the dicopper(II) helicate a lifetime of 50 s or more. Thus, the voltammetric behavior can be simply interpreted on the basis of the following stepwise equilibria, of mere electrochemical nature, involving intact double-strand helicates: ½Cu2 I ð17Þ2 2 þ K ½CuI CuII ð17Þ2 3 þ þ e
ð2:6Þ
½CuI CuII ð17Þ2 3 þ K ½CuII CuII ð17Þ2 4 þ þ e
ð2:7Þ
On the other hand, the CV investigation on a MeCN solution containing the mononuclear [CuII(17)]2 þ complex gave totally different results. Pertinent profile is shown in Fig. 2.21.
56
MOLECULAR MOTIONS DRIVEN BY TRANSITION METAL REDOX COUPLES
Figure 2.21 Cyclic voltammogram of a MeCN solution of [CuII(17)]2 þ mononuclear complex. Supporting electrolyte: 0.1 M [Bu4N]ClO4; scan rate: 0.2 V/s; internal reference electrode: Fc þ /Fc. Diagram adapted from Ref. 21.
The starting potential was again set at 850 mV and the potential was then scanned toward negative values. In particular, on moving toward cathodic potentials, a rather intense peak was observed at 150 mV, which, on the reverse scan, was not restored, indicating irreversible behavior. On the other hand, on moving further toward more positive potentials, two consecutive well-defined waves were observed to develop, with peaks at 160 and 380 mV, respectively (peak separation 220 mV). The voltammetric behavior can be explained as follows: (i) the reduction of the mononuclear [CuII(17)]2 þ complex to the corresponding mononuclear [CuI(17)] þ species is immediately followed by a fast assembling to give the dicopper(I) double-strand helicate. On scanning the potential to positive values, the [Cu2I(17)2]2 þ complex undergoes stepwise one-electron oxidation processes to give first [CuICuII(17)2]3 þ and then [CuIICuII(17)2]4 þ . The appearance of the two well-shaped one-electron peaks indicates that the double helix arrangement is maintained in the CuII oxidation state. The above evidence indicates that the OCH3 substituent in each quinoline moiety slows down the disassembling of the CuII double-strand helicate complex, which persists in solution on the timescale of the CV experiment (in the present case performed at a scan rate of 200 mV/s). This kinetic effect has probably to be related to the fact that methoxy substituents raise the energy of the transition state, which must involve a planarization of the tetrahedral coordinative arrangement within the CuII helicate complex. Any thermodynamic effect has to be ruled out, otherwise the mononuclear [CuII(17)]2 þ complex will tend to assemble spontaneously to the helicate complex [Cu2II(17)2]4 þ , even prior to the reduction. Double-strand dicopper helicate complexes are interesting systems in that they may show hysteresis (as observed with ligand 16), thus giving rise to a rare example
REFERENCES
57
of electrochemical bistability. It has also been shown that the hysteretic behavior can be controlled and modulated through simple synthetic modifications, as observed with system 17.
2.4
CONCLUDING REMARKS
In this chapter, we have considered different types of molecular movements, which are driven by metal-centered redox couples: FeIII/Fe/II and CuII/CuI. Due to the rich redox activity of transition metals, one would expect that the choice could be extended to other redox couples, within the d block of the periodic table. However, the simple occurrence of a fast and reversible one-electron redox process of a transition metal does not suffice for promoting molecular motions or conformational changes in the appropriate molecular system. In fact, it is required that the two oxidation states of the metal center exhibit distinctly different electronic and/or geometrical preferences. This is evident for the CuII-to-CuI change, in which the border is crossed between the realms of transition and posttransition metals, an almost unique opportunity (unless one would explore the d0/d1 change, perhaps TiIV/TiIII). The FeIII-to-FeII reduction process works well because it implies a change from a high-spin d5 cation, which cannot take advantage from LFSE effects and is therefore inclined to establish mainly electrostatic interactions, to a low-spin d6 cation, which benefits most from LFSE in an octahedral coordinative environment and, in the presence of p-acceptors ligands, can also exert back-donation. A d5-to-d6 change is observed in the RuIII/RuII couple, but in most complexes the second row RuIII center is present in the low-spin state, a condition that reduces its difference with respect to RuII. Thus, it seems probable that the design of electrochemically fueled engines based on metal centers will keep relying on the two couples described in the previous examples. This may further stimulate, rather than depress, the ingenuity of synthetic chemists.
REFERENCES 1. (a) V. Balzani, A. Credi, F. M. Raymo, J. F. Stoddart, Angew. Chem. 2000, 112, 3484–3530; Angew. Chem., Int. Ed. 2000, 39, 3348–3391. (b) Special issue on Molecular Machines (Ed.: J. F. Stoddart), Acc. Chem. Res. 2001, 34. (c) Special volume on Molecular Machines and Motors (Ed.: J.-P. Sauvage), Struct. Bond. 2001, 99. (d) A. H. Flood, R. J. A. Ramirez, W. Q. Deng, R. P. Muller, W. A. Goddard, J. F. Stoddart, Aust. J. Chem. 2004, 57, 301–322. (e) Special volume on Molecular Machines (Ed.: T. R. Kelly), Top. Curr. Chem. 2005, 262. (f) J.-P. Sauvage, Chem. Commun. 2005, 1507–1510. (g) K. Kinbara, T. Aida, Chem. Rev. 2005, 105, 1377–1400. (h) G. S. Kottas, L. I. Clarke, D. Horinek, J. Michl, Chem. Rev. 2005, 105, 1281–1376. (i) W. R. Browne, B. L. Feringa, Nat. Nanotechnol. 2006, 1, 25–35. (j) E. R. Kay, D. A. Leigh, F. Zerbetto, Angew. Chem. 2007, 119, 72–196;Angew. Chem., Int. Ed.2007, 46, 72–191;(k) Special issue on Molecular Machines and Switches (Eds.: A. Credi, H. Tian), Adv. Funct. Mater. 2007, 17. (l) V. Balzani, A. Credi, M. Venturi, ChemPhysChem 2008, 9, 202–220. (m) V. Balzani, A. Credi, M. Venturi, Molecular Devices and Machines, 2nd edition, Wiley-VCH, Weinheim, 2008.
58
MOLECULAR MOTIONS DRIVEN BY TRANSITION METAL REDOX COUPLES
2. R. A. Bissell, E. Co´rdova, A. E. Kaifer, J. F. Stoddart, Nature 1994, 369, 133–137. 3. A. Livoreil, C. O. Dietrich-Buchecker, J.-P. Sauvage, J. Am. Chem. Soc. 1994, 116, 9399–9400. 4. A. Tomita, M. Sano, Inorg. Chem. 1994, 33, 5825–5830. 5. L. Zelikovich, J. Libman, A. Shanzer, Nature 1995, 374, 790–792. 6. R. J. Motekaitis, Y. Sun, A. E. Martell, Inorg. Chem. 1991, 30, 1554–1556. 7. T. R. Ward, A. Lutz, S. P. Parel, J. Ensling, P. G€utlich, P. Buglyo´, C. Orvig, Inorg. Chem. 1999, 38, 5007–5017. 8. P. Belle, J.-L. Pierre, E. Saint-Aman, New J. Chem. 1998, 1399–1402. 9. J.-M. Lehn, Supramolecular Chemistry. Concepts and Perspectives, VCH, Weinheim, 1995, pp. 134–135. 10. V. Amendola, L. Fabbrizzi, C. Mangano, P. Pallavicini, Acc. Chem. Res. 2001, 34, 488–493. 11. C. Canevet, J. Libman, A. Shanzer, Angew. Chem., Int. Ed. 1996, 35, 2657–2670. 12. S. Zahn, J. W. Canary, Angew. Chem., Int. Ed. 1998, 37, 305–307. 13. G. De Santis, L. Fabbrizzi, D. Iacopino, P. Pallavicini, A. Perotti, A. Poggi, Inorg. Chem. 1997, 36, 827–832. 14. L. Fabbrizzi, F. Gatti, P. Pallavicini, E. Zambarbieri, Chem. Eur. J. 1999, 5, 682–690. 15. L. Sabatini, L. Fabbrizzi, Inorg. Chem. 1979, 18, 438–444. 16. J.-M. Lehn, A. Rigault, J. Siegel, J. Harrowfield, B. Chevrier, D. Moras, Proc. Natl. Acad. Sci. USA 1997, 84, 2565–2569. 17. J.-M. Lehn, Supramolecular Chemistry. Concepts and Perspectives, VCH, Weinheim, 1995, p. 146. 18. (a) J.-P. Gisselbrecht, M. Gross, J.-M. Lehn, J.-P. Sauvage, R. Ziessel, C. PiccinniLeopardi, J. M. Arrieta, G. Germain, M. Van Meerssche, Nouv. J. Chim. 1984, 8, 661–667. (b) Y. Yao, M. W. Perkovic, D. P. Rillema, C. Woods, Inorg. Chem. 1992, 31, 3956–3962. (c) K. T. Potts, K. M. Keshavarz, F. S. Tham, H. D. Abrun˜a, C. R. Arana, Inorg. Chem. 1993, 32, 4422–4435. (d) K. T. Potts, K. M. Keshavarz, F. S. Tham, H. D. Abrun˜a, C. Arana, Inorg. Chem. 1993, 32, 4436–4449. (e) K. T. Potts, K. M. Keshavarz, F. S. Tham, H. D. Abrun˜a, C. R. Arana, Inorg. Chem. 1993, 32, 4450–4456. (f) R. Ziessel, A. Harriman, J. Suffert, M. T. Youinou, A. De Cian, J. Fischer, Angew. Chem., Int. Ed. Engl. 1997, 36, 2509–2511. (g) M. Greenwald, M. Eassa, E. Katz, I. Willner, Y. Cohen, J. Electroanal. Chem. 1997, 434, 77–82. 19. (a) V. Amendola, L. Fabbrizzi, C. Mangano, P. Pallavicini, E. Roboli, M. Zema, Inorg. Chem. 2000, 39, 5803–5806. (b) V. Amendola, L. Fabbrizzi, P. Pallavicini, Coord. Chem. Rev. 2001, 435, 216–217. (c) V. Amendola, L. Fabbrizzi, L. Gianelli, C. Maggi, C. Mangano, P. Pallavicini, M. Zema, Inorg. Chem. 2001, 40, 3579–3587. (d) V. Amendola, L. Fabbrizzi, P. Pallavicini, E. Sartirana, A. Taglietti, Inorg. Chem. 2003, 42, 1632–1636. (e) V. Amendola, L. Fabbrizzi, E. Mundum, P. Pallavicini, Dalton Trans. 2003, 773–774. 20. V. Amendola, L. Fabbrizzi, L. Linati, C. Mangano, P. Pallavicini, V. Pedrazzini, M. Zema, Chem. Eur. J. 1999, 5, 3679–3688. 21. P. Pallavicini, M. Boiocchi, G. Dacarro, C. Mangano, New J. Chem. 2007, 31, 927–935.
CHAPTER 3
Molecular Encapsulation of Redox-active Guests ANGEL E. KAIFER Center for Supramolecular Science, Department of Chemistry, University of Miami, Coral Gables, FL, USA
3.1
INTRODUCTION
The dictionary definition of the verb encapsulate is “to place in or as if in a capsule.” In chemistry, molecular encapsulation is widely understood as the placing of a molecule inside a larger one. This definition immediately suggests the use of supramolecular interactions to achieve the encapsulated state, that is, we can take advantage of intermolecular forces to create some sort of self-assembled complex in which the smaller molecule (usually referred to as the guest) is held inside the cavity of the larger molecule (the host). If the encapsulated guest is held inside the host during an experimentally meaningful period of time, we can say that the host encapsulates the guest. While this is indeed widely accepted terminology, molecular encapsulation is also commonly used in reference to other situations, which may be quite different. For instance, it is entirely possible to covalently attach one or more branching polymeric structures to a functional group, resulting in a macromolecule (usually a dendrimer) containing the functional group at the core, partially or fully surrounded by dendritic material.1 The result is the site isolation of the functional group, that is, its relative isolation or protection from solvent molecules or other solutes in the medium. However, this situation is often also referred to as molecular encapsulation of the functional group. This general case, better described as covalent encapsulation, is treated in detail in a different chapter. Therefore, we will not consider it here in any significant detail. Recent advances in materials science have made possible the preparation of microscopic particles (in nanometer or micrometer size scales) containing molecules Electrochemistry of Functional Supramolecular Systems. Edited by Paola Ceroni, Alberto Credi, and Margherita Venturi Copyright 2010 John Wiley & Sons, Inc.
59
60
MOLECULAR ENCAPSULATION OF REDOX-ACTIVE GUESTS
coated with a different material, such as silica. However, this review will focus on well-characterized molecular materials and we will leave aside the study of these nano- or microparticle based materials, which are not discrete molecular systems. Thus, the primary focus of this chapter revolves around the idea of noncovalent encapsulation,2–7 in which a small molecule, the guest, is spontaneously included inside a larger host molecule that contains a preformed or guest-induced cavity. Guest inclusion by the host gives rise to a stable inclusion complex, which represents the encapsulated state. Of course, given the tenor of this volume, we will focus on redox-active guests. More specifically, we will pay particular attention to moieties capable of fast electron transfer reactions, since these are particularly amenable for electrochemical measurements.
3.2
THERMODYNAMIC AND KINETIC CONSIDERATIONS
At the microscopic level, the formation of a host–guest inclusion complex is a dynamic process. Usually, a combination of intermolecular forces will bring together the host and the guest to form the complex. These interactions may include ion–ion, ion–dipole, hydrogen boding, van de Waals, p–p stacking, and solvophobic forces, among others, and their combined effect affords the complex a degree of stability. In thermodynamic terms, the equilibrium between the host (H) and the guest (G) to form the complex (HG) is H þ G L HG
ð3:1Þ
The free energy change (DGo) associated with this process is a measure of the relative stability of the HG complex. For a complex that forms spontaneously, that is, self-assembles in the solution phase, DGo < 0. The free energy is related to the association equilibrium constant K by the well-known thermodynamic equation: DGo ¼ RT ln K
ð3:2Þ
Furthermore, the equilibrium constant is also related to the equilibrium activities of the species involved. For convenience, the activities are often replaced by concentrations to yield the more practical expression K¼
½HGeq ½Heq ½Geq
ð3:3Þ
In biochemistry, it is much more common to write the equilibrium as a complex dissociation process and use instead the dissociation equilibrium constant (Kd), whose value is simply the reciprocal of the association K value. The biochemists’ preference to handle Kd values has a pragmatic advantage, related to the fact that Kd has concentration units. Consider, for instance, a complex association equilibrium for
3.2
THERMODYNAMIC AND KINETIC CONSIDERATIONS
61
which K ¼ 106 M1. The same process can be represented by a dissociation equilibrium constant of Kd ¼ 106 M, which represents the lower end of host (and guest) concentration levels required for the equilibrium to be effective in the association direction. In other words, if Kd ¼ 106 M (or K ¼ 106 M1), at least micromolar concentrations of host and guest must be mixed for a significant fraction of complex to be formed. This equilibrium will not be effective at forming the HG complex when submicromolar concentrations of H and G are used. Therefore, it is common to speak of millimolar, micromolar, or nanomolar equilibrium constants depending on their magnitude. An association equilibrium constant K (also referred to as a binding constant) with a value of 5 103 M1 will not be effective at guest and host concentrations of 10 mM. In fact, simple equilibrium calculations reveal that ca. 95% of the host and the guest will remain unassociated under those conditions. These straightforward arguments are unfortunately some times forgotten when dealing with inclusion complex formation. In addition to this, the dynamic character of the equilibrium is also important. If we could follow the behavior of individual guest molecules randomly moving in solution, we would see that, every once in a while, they collide with a host molecule. Some times, these collisions are productive and lead to the formation of an inclusion complex, in which the guest is encapsulated by the host (Scheme 3.1). However, the inclusion complex is stable for a limited period of time and, eventually, undergoes dissociation to regenerate the free host and guest. The equilibrium association constant K can also be expressed as the ratio of the two kinetic rate constants shown in Scheme 3.1, that is, K¼
kIN kOUT
ð3:4Þ
The kinetic rate constant for the association process (kIN) has an upper limit set by diffusion. In other words, the rate of the fastest association processes cannot exceed the rate by which the host and the guest diffuse to encounter in solution. The maximum value of kD can then be estimated using the well-known Smoluchowski equation8: kD ¼ 1000 ½4prHG ðDH þ DH ÞNA
ð3:5Þ
Scheme 3.1 Definition of the two kinetic rate constants relevant to the formation of a host– guest inclusion complex.
62
MOLECULAR ENCAPSULATION OF REDOX-ACTIVE GUESTS
where rHG is the approach distance at which H and G react, DH and DG are the corresponding diffusion coefficients, and NA is Avogadro’s number. In aqueous solution, the upper limit for kD is in the range 109–1010 M1 s1. However, most host– guest complexation reactions are significantly slower since there are many relative molecular orientations in which host–guest collisions do not lend themselves to complex formation. Also, the host’s cavity entrance (portal) may be relatively constricted, in such a way that even with a suitable relative orientation, a collision may not be productive. Therefore, it is quite common to observe kIN values several orders of magnitude under the diffusion limit given by Equation 3.5. If the kIN value and the binding constant K are known, we can use Equation 3.4 to calculate the value of kOUT; the lifetime of the complex is simply the reciprocal of kOUT. Let us take as an example a host–guest complexation process with a typical kIN value of ca. 106 M1 s1. If the binding constant K is around 103 M1, the kOUT value should be around 103 s1, which translates to a complex lifetime of ca. 1 ms. These values can be considered representative of inclusion complexation by cyclodextrin hosts. On the other hand, if the binding constant reaches a larger value of 106 M1, the complex lifetime approaches a longer value of 1 s, assuming that kIN remains approximately the same. This would be the typical situation with a cucurbituril inclusion complex. In fact, as the stability of the inclusion complex increases, the steric fit and degree of complementarity between host and guest normally improve, which may lead to lower kIN values. Overall, higher thermodynamic stability often goes hand in hand with longer complex lifetimes. This may have important consequences on the mechanistic details of electron transfer reactions, as will be addressed later in this chapter.
3.3 CYCLODEXTRIN AND CUCURBITURIL COMPLEXATION OF REDOX-ACTIVE GUESTS One of the simplest forms of guest encapsulation is the formation of an inclusion complex with a host containing a well-defined, preformed cavity. In this regard, the cyclodextrins9–11 and the cucurbit[n]urils12–14 are probably the most relevant watersoluble hosts, and both are readily accessible with a variety of cavity sizes. While the extent of encapsulation that can be achieved with these host families is usually incomplete, their wide accessibility and aqueous solubility has led to their frequent use to complex redox-active guests. Therefore, it seems appropriate to provide here a brief description of both host families, emphasizing their relative advantages and disadvantages. Both cyclodextrins (CDs) and cucurbit[n]urils (CBns) are relatively water-soluble molecules that contain a rather rigid, well-defined cavity whose inner surface is best described as hydrophobic. Both host families are capable of forming inclusion complexes in aqueous media with a variety of guests. Although the guests are not limited to those containing hydrophobic moieties, the most stable inclusion complexes are typically formed by guests with a hydrophobic residue that fits well inside the host cavity. In spite of these general similarities between CDs and CBns, it is also important to establish clearly some of their most important differences. The CDs
3.3
CYCLODEXTRIN AND CUCURBITURIL COMPLEXATION OF REDOX-ACTIVE GUESTS
63
TABLE 3.1 Structures of Unmodified Cucurbit[n]uril (Left) and Cyclodextrin (Right) Receptors and Relevant Parameters for the Most Representative Members of These Host Families
Host
n
Molecular Weight
Cavity Diameter (A)
Aqueous Solution (mM)
CB6 CB7 CB8 a-CD b-CD g-CD
6 7 8 6 7 8
996 1163 1329 972 1135 1297
3.9–5.8 5.4–7.3 6.9–8.8 4.7–5.3 6.0–6.5 7.5–8.3
0.02 20 <0.01 149 16 178
are natural compounds and they are synthesized by the action of enzymes on starch. They are macrocyclic glucopyranose oligomers and have at least six units linked together by a-(1,4) linkages (see Table 3.1 for structures). The three most important unmodified (natural) CDs are a-cyclodextrin (a-CD), b-cyclodextrin (b-CD), and g-cyclodextrin (g-CD), composed of six, seven, and eight sugar units, respectively. In contrast, the CBns are synthetic compounds prepared by the condensation in acidic media of glycoluril with formaldehyde. This reaction produces a complex mixture of macrocyclic and acyclic oligomers from which the cyclic pentameric (CB5), hexameric (CB6), heptameric (CB7), octameric (CB8), and decameric (CB10) receptors can be separated in reasonable, albeit often low, yields. Trace amounts of other related compounds with less interesting binding properties can also be isolated from these reaction mixtures, but we will not try to give an exhaustive account of these phenomena here. Because of their range of applications as molecular container hosts and ease of isolation, we will focus our attention on the hosts CB6, CB7, and CB8 (see Table 3.1). While the cyclodextrins are chiral, the cucurbit[n]urils are not, as they exhibit a well-defined equatorial plane of symmetry and both cavity openings (portals) are indistinguishable. In CDs, the two cavity openings are different in size and chemical composition, with primary hydroxyls lining up the narrower cavity openings and secondary hydroxyls around the wider portal. Both cavity openings in CBn hosts are lined by the carbonyl oxygens from the glycoluril units. The cavity of the CDs approaches the shape of a truncated cone, reaching its maximum diameter at the wider opening. In contrast to this, the cavity of the CBns has the shape of a barrel and its maximum diameter is located at the molecular equator. The inner surface of the cavity has hydrophobic character in both host families, but
64
MOLECULAR ENCAPSULATION OF REDOX-ACTIVE GUESTS
negative charge density (from the carbonyl oxygens) accumulates on the CBn cavity openings.15 The hydroxyl groups on the CD portals are believed to be extensively hydrogen bonded, and electrostatic surface potential calculations show little net charge density accrued on the CD cavity openings. The differences on surface electrostatic potentials also constitute an important difference between both host families. The CDs form stable inclusion complexes in aqueous solution with guests containing hydrophobic residues that fit well inside the host cavity.9,11 The stability of the inclusion complex is the result of hydrophobic interactions and is usually enthalpically driven. Typically, the equilibrium association constant (K) between an excellent guest and the b-CD host is in the range 103–105 M1, with very few CD inclusion complexes exceeding this level of binding affinity. However, the formation of highly stable CBn inclusion complexes is usually driven by a combination of hydrophobic forces and ion–dipole interactions between strategically located positive charges on the guest and the rims of carbonyl oxygens on the cavity portals.12,13 As a result, CBn inclusion complexes in aqueous solution may reach equilibrium association constants as high as 1015 M1 (equivalent to that of the avidin–biotin host–guest pair16). Some reported examples suggest that CBn hosts are much more sensitive to the presence and nature of charges near the hydrophobic residue,15 giving rise to much higher binding selectivities than are possible with CDs. The functionalization chemistry of the CDs has been extensively developed at this time.17 Synthetic methodology is available to either monofunctionalize or perfunctionalize either the primary or secondary hydroxyl portal of these hosts. Derivatization is still a pending subject in the chemistry of cucurbit[n]uril hosts. Kim and coworkers have reported a procedure that allows the equatorial functionalization of the periphery of CB6, but the same procedure is highly inefficient with either CB7 or CB8.18 Further developments in this area may be of great significance to extend the range of applications of this host family. 3.3.1
Electrochemistry of CD Inclusion Complexes
Given the popularity of the CD hosts, the electrochemical behavior of their inclusion complexes with redox-active guests has received considerable attention for more than 20 years. The first work in this area was carried out by Evans and coworkers in 1985.19 Before their work was published, it was already well established that ferrocene was an excellent guest for inclusion complexation inside b-CD,20 as well as other unmodified CD hosts.21 However, the low solubility of ferrocene in aqueous solution (<50 mM) hampered quantitative investigations of its electrochemistry in the presence of CD hosts. Evans circumvented this problem by investigating the binding interactions of ferrocenecarboxylate (FcCOO) with b-CD in aqueous media buffered at pH 7.19 Formation of a stable inclusion complex was demonstrated by electronic absorption spectroscopy. In the presence of b-CD, the half-wave potential (E1/2) for the one-electron oxidation of FcCOO shifts to more positive values. E1/2 values are readily measured in voltammetric experiments and constitute an excellent approximation to the formal potential (Eo0 ) of the corresponding redox couple.22 The
3.3
CYCLODEXTRIN AND CUCURBITURIL COMPLEXATION OF REDOX-ACTIVE GUESTS
65
anodic half-wave potential shift observed by these authors is consistent with the differential stabilization of the reduced form of the redox couple. In other words, FcCOO is more stabilized by b-CD than its one-electron oxidized counterpart, Fc þ COO. An additional observation is that the current levels associated with the voltammetric wave decrease in the presence of b-CD. This is easily explained by the larger molecular weight and size of the b-CD . FcCOO complex compared to the free guest. Therefore, inclusion complexation tends to slow down the diffusional flow of guest to the electrode surface. This flow is driven by the low concentration of FcCOO near the electrode, where it is consumed (oxidized) by the electron transfer process. After a thorough analysis of the experimental current–potential curves, these authors concluded that FcCOO forms a stable inclusion complex with the host, while the oxidized, zwitterionic form of the guest does not interact strongly with b-CD. The electrochemical oxidation process is best rationalized by an electron transfer process preceded by a coupled chemical reaction (host–guest complexation/decomplexation equilibrium), which is referred to as a CE (chemical–electrochemical) mechanism in electrochemical jargon. The proposed mechanism is illustrated in Scheme 3.2. It is important to realize that the mechanism does not take into account the direct electrochemical oxidation of the inclusion complex. Evidence for this surprising finding was gathered at fast scan rates, at which the cyclic voltammetric wave for oxidation of FcCOO is flattened,19 because the complex dissociation mechanism becomes too slow to generate enough free guest to sustain the fast electrochemical oxidation. The reasons for the lack of electrochemical reactivity of the inclusion complex were not clear at the time this work was completed and still remain intriguing now. We can assume that the electrochemical kinetics of the inclusion complex is slower than that of the free guest and, because the inclusion complex is short lived, dissociation provides a bypass mechanism for the electron transfer to take place more quickly. Another take-home message from this finding is that to investigate the electrochemical behavior of included guests, we must prepare complexes of reasonable kinetic
Scheme 3.2 of b-CD.
CE mechanism for the one-electron oxidation of FcCOO in the presence
66
MOLECULAR ENCAPSULATION OF REDOX-ACTIVE GUESTS
stability, as kinetically labile complexes may easily find possible bypass conduits to avoid electron transfer to/from the included guest. In any instance, the electrochemistry of water-soluble ferrocene derivatives in the presence of CD hosts has received considerable attention. My own group23,24 and others25,26 have investigated several types of ferrocene derivatives, including dendronized ferrocenes27,28 and dendrimers with multiple ferrocene groups29 on their peripheries. The voltammetric data obtained with all these compounds seem to fit the mechanism shown in Scheme 3.2. Reinhoudt and coworkers have taken advantage of the multivalent interactions between dendrimers with multiple surface ferrocene residues and self-assembled monolayers containing CD binding sites for the development of “molecular printboards.”30 Interestingly, the electrochemistry of cobaltocenium in the presence of b-CD has also received attention.31 Cobaltocenium is a cationic, isolectronic analog of ferrocene, which undergoes reversible one-electron reduction to its neutral form, cobaltocene, at accessible potentials. Cobaltocenium, such as ferrocenium, does not interact strongly with b-CD, but the neutral cobaltocene forms a stable inclusion complex.31 As anticipated, the differential stabilization of the reduced form by the CD host leads to a shift of the E1/2 value to more positive potentials, that is, the reduction process is thermodynamically favored by b-CD. In this case, it is the oxidation wave for the electrogenerated b-CD . cobaltocene complex that flattens at fast scan rates, revealing that the inclusion complex does not undergo direct electrochemical oxidation (see Scheme 3.3). Therefore, this complex exhibits voltammetric behavior fundamentally similar to that of most b-CD inclusion complexes of ferrocene derivatives. These conclusions are strongly supported not only by the already mentioned flattening of CV waves at fast scan rates,31 but also by the fitting of the experimental current–potential waves using digital simulations techniques, which afford an excellent method to test different mechanistic proposals for electrochemical oxidation/reduction processes. My group also studied the binding interactions between b-CD and dendrimers containing multiple cobaltocenium units in their peripheries.32 Furthermore, we have
Scheme 3.3
Electrochemical reduction of CobCOO in the presence of b-CD.
3.3
CYCLODEXTRIN AND CUCURBITURIL COMPLEXATION OF REDOX-ACTIVE GUESTS
67
also investigated compounds containing both cobaltocenium and ferrocene centers,33 which are of interest because they exhibit binding interactions with CD hosts that have a pronounced redox switchable character. In their fully reduced state, these compounds present ferrocene and cobaltocene centers, and each of them becomes an independent CD-binding site. In strong contrast, upon full oxidation, the cobaltocenium and ferrocenium centers are not particularly good binding sites for b-CD. There is also a wide intermediate potential region, in which cobaltocenium and ferrocene centers coexist. In this potential window, the only effective CD binding sites are the ferrocene centers. Thus, these mixed ferrocene-cobaltocenium compounds exhibit modulated, redox-controlled interactions with CD hosts.33 We should also mention here that the binding interactions between dialkyl-N,N0 bipyridinium (viologen) derivatives and b-CD have also been investigated in detail by our group.34 Viologens undergo two consecutive one-electron reductions (Scheme 3.4). The original material, usually a dication (V2 þ ), is first reduced to a cation radical (V þ ), which in turn can be reduced to a neutral, quinonoid form (V) at more negative potentials. Not surprisingly, the dicationic form is not a guest for binding by b-CD, the cation radical forms a weak complex and the neutral form is strongly bound by the host. In other words, as electrons are pumped into the viologen residue, gradually removing its positive charge, the interaction with b-CD becomes stronger and the viologen residue becomes more strongly included inside the host cavity. In fact, digital simulations confirmed that the oxidation of the b-CD . V inclusion complex does not take place directly at the electrode surface.34 Like the rest of the CD inclusion complexes, oxidation takes places via a CE mechanism, and complex dissociation precedes the electron transfer process (V ! V þ ). Other groups have investigated the electrochemical reduction of viologens in the presence of g-CD
Scheme 3.4 b-CD.
The three oxidation states of viologens and their variable binding affinity for
68
MOLECULAR ENCAPSULATION OF REDOX-ACTIVE GUESTS
and found that cation radical dimerization is favored inside the larger cavity of this host.35 In summary, the cyclodextrins are readily accessible hosts that can form stable inclusion complexes with relatively hydrophobic redox-active guests in aqueous media. Most investigations of the electrochemical behavior of these complexes suggest that their labile kinetic nature allows relatively efficient electron transfer reactions to or from the free guest, following the fast dissociation from the host. Therefore, in spite of their accessibility and simplicity of use, these hosts do not form complexes of enough kinetic stability to allow the investigation of the electron transfer reactions involving the encapsulated guests. Furthermore, it can be argued that the truncated cone structure of CD hosts is far from ideal for encapsulation, as the cavity becomes gradually wider from one opening to the other. 3.3.2
Electrochemistry of CB Inclusion Complexes
The family of the cucurbituril hosts holds more promise for effective guest encapsulation. CBn cavities reach a maximum cross-section in the middle, while the two openings are identical and represent the narrowest points along the cavity. This type of “barrel” cavity shape is more suited to guest encapsulation than the truncated cone shape of the CD hosts. As a result of their cavity shapes, CBn inclusion complexes tend to be more kinetically stable than CD complexes with similar guests. This is manifested in 1 H NMR spectroscopic experiments, where the simultaneous observation of signals for the bound (included) and free guests is common with CBn complexes and rare with CD complexes. The proclivity of CBn complexes to fall in the slow exchange regime in NMR spectroscopic experiments reflects the generally longer lifetimes of these complexes. Since the CB6 cavity is too narrow to form stable inclusion complexes with most aromatic units, the electrochemical investigation of CBn complexes had to wait until the isolation of larger cavity CBn hosts by Kim and coworkers in 2000.36 In 2002, the Kim’s group37 and my own group38 reported the formation of a stable inclusion complex between the organic dication N,N0 -dimethyl-4,40 -bipyridinium (methyl viologen or paraquat, MV2 þ ) and CB7. In aqueous media, the K value for the formation of this complex is ca. 105 M1, although this value decreases as the medium’s ionic strength increases.39 The lifetime of the complex has been estimated to be 5.3 ms.37 In aqueous solution, one- or two-electron reduction of MV2 þ leads to less charged or neutral species, which are more hydrophobic and have a marked tendency to precipitate on the electrode surface. However, if we confine ourselves to the first one-electron reduction (MV2 þ ! MV þ ), we can see that the presence of CB7 decreases the current levels and shifts the half-wave potential slightly to more negative values (Fig. 3.1). This shows that CB7 differentially stabilizes the dicationic viologen form versus the cation radical, although the small magnitude of the CB7-induced, half-wave potential shift (ca. 30 mV) suggests only a moderate stability drop in the complex after the viologen guest loses one of its positive charges.38 The decrease in the current levels in the presence of CB7 reflects the larger size and lower diffusivity of the complex compared to the free guest.
3.3
CYCLODEXTRIN AND CUCURBITURIL COMPLEXATION OF REDOX-ACTIVE GUESTS
69
Figure 3.1 Cyclic voltammetric behavior on a glassy carbon electrode (0.072 cm2) of 1.0 mM MV2 þ in the absence (solid line) and in the presence (dotted line) of 1 equivalent CB7. Medium: 0.1 M NaCl. Scan rate: 0.1 V/s.
An interesting finding in the CB7-MV2 þ system is that, in clear contrast to host– guest systems involving CD hosts, the voltammetric data do not contain any indication that complex dissociation must precede any of the electron transfer processes. Furthermore, the electrochemistry of the inclusion complex is as fast—in the timescale accessible in these cyclic voltammetric experiments—as that of the free guest. This is clearly illustrated by the voltammograms depicted in Fig. 3.2, which show the comparative results of a scan-rate study on the MV2 þ /MV þ and CB7 . MV2 þ /CB7 . MV þ redox couples. In both cases, the observed anodic and cathodic peak potentials are basically invariant as the scan rate is increased up to 1.0 V/s, revealing the reversible character of both couples under the experimental conditions of the study. Therefore, we can conclude that CB7 inclusion of methyl viologen does not seem to affect its electrochemical kinetics in a pronounced way. We cannot unequivocally state that there is no change in the kinetics, since a small decrease in the standard rate constant (ko) may be possible and go undetected in these experiments. Although no crystal structure of the CB7 . MV2 þ complex is available, all the experimental and computational data are consistent with the structure shown in Fig. 3.3. The guest is centered inside the host, with the two positively charged nitrogens interacting with the carbonyl rims on the cavity portals and hydrophobic interactions further stabilizing the mid-section of the complex. Obviously, the guest is not fully encapsulated, as each of its N-methyl termini protrude through the host portals. However, the differences between the voltammetric behavior of this complex and CD inclusion complexes of redox-active guests are very pronounced. More recently, we have carried out a detailed comparison of the voltammetric behavior of methyl viologen and diquat (DQ2 þ ) in the presence of CB7.40 The latter is the common name of N,N0 -ethylene-2,20 -bipyridinium, a dication with a charge distribution very different from that of methyl viologen. The proximity between the two positively charged nitrogens in DQ2 þ decreases its binding affinity to CB7, and
70
MOLECULAR ENCAPSULATION OF REDOX-ACTIVE GUESTS
Figure 3.2 Scan-rate dependence of the cyclic voltammetric response on a glassy carbon electrode (0.072 cm2) of 1.0 mM MV2 þ in 0.1 M NaCl (a) in the absence and (b) in the presence of 1 equivalent CB7. Scan rates: 0.1, 0.2, 0.3, 0.4, 0.5, 0.6, 0.8 and 1.0 V/s.
the corresponding binding constant is just 350 M1. One-electron reduction of this guest increases the binding affinity to 1 104 M1. Detailed analysis of the voltammetric data using digital simulation techniques reveals further differences in the electrochemical reduction mechanism of these two dicationic guests. Mainly, no direct electron transfer to the CB7 . DQ2 þ complex is detected. In analogy to the case
Figure 3.3 Energy minimized structure (PM3 method) of the CB7 . MV2 þ complex. (See the color version of this figure in Color Plates section.)
3.3
CYCLODEXTRIN AND CUCURBITURIL COMPLEXATION OF REDOX-ACTIVE GUESTS
(a) DQ2+
+ e–
+CB7
MV2+
.
DQ+
+CB7 .
CB7 ·DQ2+ (b)
71
CB7 ·DQ+
+ e–
+CB7 CB7 ·MV2+
.
MV+
+CB7 +e
.
CB7 ·MV+
Scheme 3.5 Electrochemical and chemical reactions involved in the one-electron reduction of (a) MV2 þ and (b) DQ2 þ .
of CD inclusion complexes, the electron transfer takes place only upon complex dissociation. In contrast to this, the direct electrochemical reduction of the CB7 . MV2 þ complex is observed, as concluded by the fitting of the experimental current–potential curves to digital simulations. The actual mechanisms, as validated by the digital simulation analysis, are shown in both cases in Scheme 3.5. The differences between these two CB7 inclusion complexes are ascribed to the considerably lower thermodynamic and kinetic stability of the CB7-diquat complex. Kim and coworkers reported that MV2 þ also forms a 1:1 inclusion complex with the larger cavity host CB8.41 However, one-electron reduction of the guest gives rise to the formation of a ternary inclusion complex in which the methyl viologen cation radical (MV þ ) dimerizes inside the host cavity. The CB8-induced stabilization of this dimer is considerable and the formation of this ternary complex, CB8 . (MV þ )2 is essentially quantitative at millimolar concentration levels of the components. While dimerization of viologen cation radicals was well known, it was widely considered as a bothersome side reaction in the development of electrochromic applications involving viologen compounds. Upon Kim’s discovery, this dimerization reaction has become an important tool for CB8-mediated self-assembly.14 For instance, we have used it for the redox-controlled self-assembly of viologen-containing dendrimers.42,43 In 2003, we reported that both ferrocene and ferrocenium form inclusion complexes with CB7.44 The oxidized form is slightly preferred as the guest, but the selectivity is very moderate. Molecular modeling shows that the fit between the ferrocene residue and the CB7 host is very tight and, considering the hydrophobic character of both, hydrophobic forces play a very significant role in the stabilization of the complex. However, the very low aqueous solubility of ferrocene hampered the quantitative investigation of these complexation reactions. Therefore, we decided to move our efforts to the investigation of the CB7 complexation of a group of more water-soluble ferrocene derivatives (Fig. 3.4).
72
MOLECULAR ENCAPSULATION OF REDOX-ACTIVE GUESTS
+ N
OH Fe
Fe
FcOH
FcNMe3+ COO –
+ N
Fe
FcCOO–
Figure 3.4
Fe
N +
Fc(NMe3)22+
Structures of ferrocene derivatives investigated as guests for CB7.
The neutral guest FcOH was found to form a highly stable inclusion complex with CB715 (K ¼ 3 109 M1). However, the cationic guests reach extremely high binding affinities (K values of 4 1012 and 3 1015 M1 for FcNMe3 þ and Fc(NMe3)22 þ , respectively), which are similar to that observed between biotin and avidin,16 but the anionic guest FcCOO is not bound at all by CB7.15 These results highlight the pronounced binding differences between CBn and CD hosts. The best CD host for ferrocene-containing guests is b-CD, which reaches K values in the range 103–104 M1 with all the ferrocene guests in Fig. 3.4 regardless of charge and functional group. In strong contrast, the CB7 host develops a much higher binding affinity with guests FcOH, FcNMe3 þ , and Fc(NMe3)22 þ , while exhibiting considerable binding selectivity depending on the functional groups attached to the ferrocene residue. The cyclic voltammetric behavior of the cationic guest FcNMe3 þ in the presence of CB7 is interesting15 (Fig. 3.5). The decreased current levels induced by the presence of CB7 indicate again that the complex is bulkier and diffuses more slowly than the free guest. The CB7-induced E1/2 shift to more positive values reveals that (1) the electrochemical oxidation of the ferrocene residue is thermodynamically hindered by the CB7 cavity and (2) one-electron oxidation decreases the stability of the complex. Although this effect is pronounced (ca 1.5 orders of magnitude in the K value), the complex still maintains considerable stability after oxidation, due to the extremely high binding affinity of the CB7 . FcNMe3 þ complex. Given that this complex is kinetically and thermodynamically very stable, we were very interested in recording its voltammetric behavior as a function of the scan rate. The results are shown in Fig. 3.6. Clearly, increasing the scan rate causes an increased separation in the potential difference observed between anodic and cathodic peak potentials for the CB7 . FcNMe32 þ /CB7 . FcNMe3 þ redox couple. In the absence of CB7 the FcNMe32 þ /FcNMe3 þ redox couple is extremely fast, with measured ko values in the range 1–5 cm/s. Our measurements show that encapsulation inside CB7 leads to a considerable decrease of the ko value into the quasi-reversible regime,
3.3
CYCLODEXTRIN AND CUCURBITURIL COMPLEXATION OF REDOX-ACTIVE GUESTS
73
Figure 3.5 Cyclic voltammetric response on glassy carbon (0.072 cm2) of guest FcNMe3 þ (1.0 mM) in 0.1 M NaCl in the absence (solid line) and in the presence (discontinuous) of 1 equivalent CB7. Scan rate: 0.1 V/s.
where it can be easily determined in cyclic voltammetric experiments. We should contrast these results to those obtained with the MV2 þ guest, where we could not detect a measurable decrease in the ko value in similar voltammetric experiments. At this point, we can only speculate about the reasons for the measurable ko decrease in the CB7 . FcNMe3 þ complex and the lack of a measurable encapsulation effect with CB7 . MV2 þ . Perhaps, the poorer degree of encapsulation in the latter, with the two ends of the guest clearly piercing through the host portals (Fig. 3.3) plays a significant role. In the structure of the CB7 . FcNMe3 þ complex obtained by computational methods (Fig. 3.7), the redox-active ferrocene center appears to be better insulated by the host than the aromatic viologen residue in the case of CB7 . MV2 þ complex. Furthermore, Macartney and coworkers have measured the electron self-exchange rate constant for the CB7 . FcNMe3 þ complex and found a slight increase compared to the value determined for the free guest.45 Our electrochemical data are certainly at variance with these findings.
Figure 3.6 Cyclic voltammetric response on glassy carbon of guest 2 (1.0 mM) in 0.1 M NaCl in the presence of 1 equivalent CB7. Scan rates: 0.05, 0.10, 0.20, 0.50, and 1.0 V/s.
74
MOLECULAR ENCAPSULATION OF REDOX-ACTIVE GUESTS
Figure 3.7 Energy minimized structure (PM3 method) of the CB7 . 2 complex. (See the color version of this figure in Color Plates section.)
We are currently carrying out further investigations with neutral ferrocene derivatives in an attempt to resolve the apparent disconnection between the effects of CB7 encapsulation on homogenous and heterogeneous electron transfer reactions rates.
3.4 ELECTROCHEMISTRY OF INCLUSION COMPLEXES FORMED BY CAVITAND-TYPE HOSTS Cavitands are hosts formed in acidic condensation reactions between resorcinol derivatives and aldehydes.46 The resulting cyclic octol compounds are usually tetrameric and contain four aromatic units that form a relatively shallow bowl in the preferred C4v conformation. Further synthetic elaboration on the structure of the octols allows us to fix the conformation of these compounds in C4v symmetry with a well defined, albeit small cavity. Cram developed this chemistry further during the 1980s and 1990s and prepared compounds consisting of two connected cavitands facing each other. When the two cavitands are closely connected, with very small openings between them, the resulting host compounds were baptized as carcerands, as they could encapsulate (“incarcerate”) small solvent molecules.47 Of more interest to us was the preparation of larger compounds, in which the connecting branches between the two facing cavitands leave larger spaces between them, which effectively become portals into the host’s inside cavity. These host compounds were referred to as hemicarcerands, because the encapsulation of guest molecules was not complete.47 When the size (cross-section) of the guest is comparable to the opening of the cavity portals, direct inclusion of the guest inside the host does not take place at room temperature. However, heating of the guest and the hemicarcerand in solution provide enough kinetic energy so that some collisions lead to the formation of the inclusion complex or hemicarceplex. Upon cooling, the inclusion complex is trapped, as the guest does not have enough energy to escape through the tight portals on the equator of the host. Cram referred to this situation as constrictive binding. Conceptually, the situation is
75
3.4 ELECTROCHEMISTRY OF INCLUSION COMPLEXES FORMED BY CAVITAND-TYPE HOSTS
similar to that found in rotaxanes, in which the macrocyclic wheel cannot dissociate from the axle component because the latter is terminated in bulky stopper groups that prevent the sliding away of the wheel. 3.4.1
Electrochemistry of Redox-Active Hemicarceplexes
Inspired by the work of the Cram group, we decided a few years ago to prepare hemicarcerands with an internal cavity large enough to encapsulate ferrocene. We prepared two similar hemicarcerands (1 and 2) with octaimine connectivity between the two cavitand bowls and two different “feet” substituents on the polar ends: phenylethyl and pentyl.48 Scheme 3.7 shows the structure of hemicarcerand 1 (with phenylethyl feet). The size of the equatorial portals allows the inclusion of ferrocene when guest and host are heated together in tripiperidinephosphine oxide. After cooling down we isolated the ferrocene hemicarceplex (inclusion complex), which showed a very slow release of the ferrocene guest. Using 1 H NMR spectroscopic experiments, we determined that the half-life for guest release was >300 h. Obviously, this finding reveals a kinetic stability that provides a time window more than sufficiently long to carry out electrochemical experiments on the inclusion complex. While the Fc@1 hemicarceplex was found to be soluble only in tetrachloroethane, the Fc@2 hemicarceplex was also soluble in dichloromethane. In this solvent, the half-wave potential for ferrocene oxidation was found to shift from 0.45 V versus Ag/AgCl for the free guest to 0.57 V versus Ag/AgCl for Fc@2. This was rationalized as a result of the inability of the hemicarcerand walls to effectively solvate the positive charge on the oxidized ferrocenium guest. More interestingly, the standard rate constant for the heterogeneous electron transfer process (ko) decreased from 0.043 to 0.0041 cm/s.48 The data obtained with both hemicarceplexes (Fc@1 and Fc@2) in tetrachloroethane solution were entirely similar.
R' HO
RCHO HCl
R' HO HO
OH
R
R
R
R
R'
HO
OH OH
R'
octol Scheme 3.6
R' O
OH
R' HO
OH
O CH2BrCl, Cs2CO3 DMSO
O
R
R
R
R
O
R'
R' O O
O O
R'
cavitand General synthesis of octols and cavitands from resorcinol.
76
MOLECULAR ENCAPSULATION OF REDOX-ACTIVE GUESTS
Scheme 3.7 Incorporation of ferrocene inside hemicarcerand 1.
3.4.2 Electrochemistry of Guests Inside Resorcinarene Molecular Capsules Given the complexities in the synthetic preparation of the hemicarcerands described in the previous section, we were interested in using encapsulating molecules with much easier synthetic accessibility. In this regard, the seminal report by MacGillivray and Atwood49 in 1997 opened up the possibility of using self-assembled, hexameric capsules of resorcinarenes as the encapsulating hosts. The resorcinarenes are tetrameric macrocyclic octols prepared by the acid-catalyzed condensation of resorcinol in the presence of an aldehyde (see “octol” structures in Scheme 3.6). While their chemistry has been well known since the 1970s, MacGillivray and Atwood found that the simple resorcinarene 3 (see Fig. 3.8) crystallizes forming fascinating hexameric
HO
OH
HO
OH R
R
R
R
HO
OH
HO
OH
3, R = CH3 4, R = (CH2)10CH3
Figure 3.8 Structures of the resorcinarenes 3 and 4 used in the formation of self-assembled, hexameric molecular capsules.
3.4 ELECTROCHEMISTRY OF INCLUSION COMPLEXES FORMED BY CAVITAND-TYPE HOSTS
77
molecular capsules.49 Each capsule adopts a snug cube conformation with the 6 resorcinarene and 8 water molecules held together by a network of 60 hydrogen bonds. The internal volume estimated for each capsule is ca. 1375 A3. This groundbreaking report attracted attention from several groups and it was shown quickly that similar hexameric molecular capsules were formed by a number of resorcinarenes in the solution phase.50 Due to its solubility in low polarity solvents, resorcinarene 4 (Fig. 3.8) has been utilized extensively as the monomer for hexameric capsule formation in CHCl3 and CH2Cl2 solution. Since 46 is quite effective at trapping alkylammonium cations,50 we decided to investigate the encapsulation of the redoxactive cobaltocenium (Cob þ ). NMR spectroscopic data in CD2Cl2 solution verified the encapsulation of cobaltocenium inside a capsule composed of six molecules of resorcinarene 4. From NMR diffusion coefficient measurements, we concluded that the resorcinarene molecular capsules were preformed in CD2Cl2 solution51 in the absence of cobaltocenium or any other cationic species. Similar results in CDCl3 solution have been reported by Cohen and coworkers.52,53 It is now clear that cationic species are not required to “seed” the self-assembly of the capsules. The voltammetric behavior of cobaltocenium (as its hexafluorophosphate salt) in CH2Cl2 solution is reversible and the corresponding half-wave potential (E1/2) for the cobaltocenium/cobaltocene redox couple was found to be 1.06 V versus Ag þ /Ag. In the presence of 6 equivalents of resorcinarene 4, this voltammetric wave disappears completely and only the charging current baseline is observed in the potential range from 0.0 to 1.6 V.51 What factors could be responsible for the disappearance of the voltammetric response due to cobaltocenium reduction? To discard the passivation of the electrode surface by a layer of organic molecules, we carried out voltammetric experiments using two electroactive species: cobaltocenium and decamethylferrocene. The latter compound is considerable larger than the former (or ferrocene) due to the five methyl groups decorating the periphery of each of the cyclopentadienyl rings of the compound. In the absence of host 4, both reversible redox couples were observed in the cyclic voltammogram as anticipated. However, upon addition of 6 equivalents of 4, the wave corresponding to cobaltocenium disappears, while the wave corresponding to the reversible oxidation of decamethylferrocene remains essentially unchanged.54 Clearly, this finding reveals that the electrode surface is not passivated by the addition of host 4. The disappearance of the cobaltocenium voltammetric wave has to be caused by a substantial attenuation of the electrochemical kinetics of cobaltocenium reduction. Encapsulation of cobaltocenium inside a very large, closed capsular assembly-formed by six resorcinarene host molecules- must necessarily exert a strong effect on the kinetics of its one-electron reduction. Encapsulated cobaltocenium cannot approach the electrode surface as closely as free cobaltocenium to undergo heterogeneous electron transfer. In fact, cobaltocenium encapsulation inside 46 increases considerably the distance between the outer Helmholtz plane (OHP) and the electrode surface, leading to a much weaker coupling of electronic density levels between the redoxactive center and the electrode. Therefore, a substantial slowdown in the kinetics of heterogeneous electron transfer takes place, in agreement with the experimental
78
MOLECULAR ENCAPSULATION OF REDOX-ACTIVE GUESTS
Figure 3.9 Size comparison between cobatocenium (top left) and the hexameric molecular capsule formed by resorcinarene 3. Resorcinarene 4 would form a much larger capsule due to its longer undecyl “feet” compared to the methyl “feet” in 3. (See the color version of this figure in Color Plates section.)
observations. Given the size of the resorcinarene and its hexameric assembly (see Fig. 3.9), we estimate that, for encapsulated cobaltocenium, the OHP has to be at least 2 nm away from the electrode surface. There are several precedents for this rationalization of the observed voltammetric data. We observed a pronounced slowdown in the electrochemical kinetics for ferrocene oxidation upon encapsulation48 inside hemicarcerands 1 and 2. These hosts can be roughly considered to be about one-third of the size of the 46 capsule. In addition to this, our group55 and others56 have also observed the attenuation of electrochemical kinetics when redox centers are surrounded by covalently attached dendritic mass. Therefore, these kinetic encapsulation effects provide strong support to our interpretation. In an attempt to better understand the mechanism for encapsulation of cobaltocenium inside 46 we performed careful titrations of this redox-active cation with variable concentrations of host 4. To our surprise, we quickly realized that only 2–3 equivalents were necessary to fully shut down the electrochemical response of cobaltocenium.51 This result is in strong contrast to the NMR data, which clearly indicate a 6:1 [host/guest] stoichiometry for full encapsulation. There are two important differences between these two types of experiments. In the NMR experiments, solutions were prepared in pure, deuterated CD2Cl2 and the only solutes present are cobaltocenium hexafluorophosphate (ca. 1 mM) and host 4 (0–8 mM). In the electrochemical experiments, solutions were prepared in isotopically unenriched CH2Cl2 also containing 0.1 M tetradodecylammonium bromide as supporting electrolyte. The concentrations of cobaltocenium hexafluorophosphate and host 4 were similar to those used in the NMR experiments. It clearly became evident that the nature of the supporting electrolyte, especially the nature of its anion, was crucial to
3.4 ELECTROCHEMISTRY OF INCLUSION COMPLEXES FORMED BY CAVITAND-TYPE HOSTS
79
achieve encapsulation. The use as supporting electrolytes of tetra-alkylammonium bromides and chlorides, led to the observation of cobaltocenium encapsulation in voltammetric experiments, while the use of tetra-alkylammonium perchlorates, hexafluorophosphates or tetrafluoroborates led to no voltammetric evidence for the encapsulation. These observations strongly suggest that the cobaltocenium cation is trapped inside the assembly in the form of ion pairs, and larger anions may hinder the encapsulation. On the other hand, bromide and chloride are believed to participate in the network of hydrogen bonds holding together the resorcinarene molecules, probably assisting in the formation of aggregates smaller than the hexameric capsule. Mass spectrometric evidence for these bromide-containing assemblies has been reported by Schalley and coworkers.57 In contrast to the anions, the nature of the cation has only a minor influence on the type of voltammetric behavior observed for cobaltocenium in the presence of resorcinarene hosts. An interesting finding in our experiments is that the resorcinarene capsules show a strong preference to encapsulate cobaltocenium versus tetra-alkylammonium cations, as these did not interfere when present in large excess, as required for the cation of the supporting electrolyte. We also investigated the electrochemistry of ferrocene in the presence of host 4 and demonstrated that the positively charged, oxidized form (ferrocenium) is encapsulated by 46 capsules.54 The encapsulation of ferrocenium was affected by the nature of the supporting electrolyte anion in exactly the same ways that we have already discussed for encapsulation of cobaltocenium. In general terms, our voltammetric and NMR spectroscopic data show how the incorporation of a redox-active cation inside a large, hydrogen-bonded assembly can basically nullify its voltammetric response due to electrochemical kinetic attenuation effects. Furthermore, our data also afford clear indications that the assembly of resorcinarenes in low polarity solutions is a very complicated process and that the hexameric molecular capsule first crystallized by MacGillivray and Atwood is certainly associated to a Gibbs energy minimum, but the overall energetic landscape may be strongly affected by other system components, such as supporting electrolyte anions. 3.4.3 Electrochemistry of Cavitand-Encapsulated Guests in Aqueous Solution For quite some time most synthetic efforts to prepare cavitand-type hosts led to compounds that were only soluble in low polarity solvents. Because of their potential biological relevance, interest on the synthesis of water-soluble cavitands developed quickly, but only recently a number of accessible hosts has become available. We will describe here recent work done by us on Gibb’s octaacid, deep-cavity cavitand58 and Rebek’s water-soluble cavitand.59 The structures of these compounds are shown in Fig. 3.10. Gibb’s octaacid58 (compound 5) is soluble in mildly basic solutions, wherein its eight external carboxylic acid groups are ionized, conferring this host a substantial negative charge (8). It is well known that in the presence of hydrophobic guests of suitable size, two octaacid host molecules come together, assisted by hydrophobic
80
MOLECULAR ENCAPSULATION OF REDOX-ACTIVE GUESTS
COOH COO – N O
O O
O
O O
Et
Et
Et
Et
O
O
O
O
O
O O
O 4 Na+
N
O –OOC
5, R = CH2CH2COOH
N
O
COOH
O O
O O
O
R R R R
HOOC
HN O
O
O
COO–
HN
NH 6
NH N –
OOC
COOH
Figure 3.10 Structures of water-soluble cavitand-type hosts: Gibb’s octaacid (host 5) and Rebek’s cavitand (host 6).
effects, to yield a well-defined dimeric molecular capsule 52. We were intrigued by the possible formation of such a capsule around hydrophobic ferrocene, so we started the investigation of this system in collaboration with Gibb’s group. The first indication for a strong binding interaction is the observation that 1.0 mM Fc is soluble in aqueous solution in the presence of 2.0 mM 5. Note that we have already mentioned that Fc has very low solubility in water (<50 mM). The remarkable solubility increase measured in the presence of 2.0 equivalents of 5 reflects necessarily the presence of binding interactions. 1 H NMR spectroscopic experiments provide clear evidence that ferrocene is encapsulated inside 52. Not only because of the complexation-induced chemical shifts,60 which are very pronounced in the case of ferrocene (its protons are observed at 2.16 ppm!), but also because of diffusion coefficients measured using NMR techniques. In fact, in the presence of 2.0 equivalents of host, ferrocene was found to diffuse at the same rate as the 52 molecular capsule (1.6 106 cm2/s). This finding is very strong evidence for the formation of the Fc@52 molecular assembly. As soon as excess Fc is added to the solution, that is, if [Fc] > [5]/2, turbidity develops as insufficient host is available to encapsulate and solubilize all the ferrocene. The ferrocene voltammetric response of Fc@52 was found to be completely flat in the potential range where its oxidation is observed.60 In other words, no current wave for one-electron oxidation of Fc was observed either by cyclic voltammetry or by square wave voltammetry. However, the expected current wave emerged as soon as the concentration of Fc exceeded the value that can be encapsulated, that is, [Fc] > [5]/2. This suggests that there is no passivation of the electrode resulting from any accumulation or precipitation of host on its surface. In fact, our data clearly support the voltammetrically silent character of Fc@52 under the conditions of our experiments. Figure 3.11 shows an energy-minimized (PM3) structure of this assembly. While its size is much smaller than that of Cob þ @46 (Fig. 3.9) the encapsulation has
3.4 ELECTROCHEMISTRY OF INCLUSION COMPLEXES FORMED BY CAVITAND-TYPE HOSTS
81
Figure 3.11 Energy-minimized (PM3 method) structure of the Fc@52 molecular assembly. (See the color version of this figure in Color Plates section.)
a similar effect, basically eliminating the voltammetric response of the encapsulated redox center! Further work with this system is currently underway in our group. We have found that the negative charges around the 52 dimeric molecular capsule may play a very important role. For instance, hydrophobic cations, such as viologens,60 bind strongly to the outer surface of the capsule, due to the presence of four negative charges clustered at each polar end of the assembly and another eight negative charges gathered around the capsule’s equator. This opens interesting research possibilities related to mediated or photoinduced electron transfer reactions between species bound inside the capsule and tightly bound to its external surface. We are currently investigating these phenomena. Also very recently, we have reported on the voltammetric properties of inclusion complexes61 formed between ferrocene and Rebek’s deep-cavity cavitand 6 (Fig. 3.10). This compound is a tetra-acid, which forms dimeric capsules in the absence of suitable guests, but these capsules have no significant internal cavity.62 However, host 6 forms 1:1 inclusion complexes with hydrophobic guests, such as ferrocene.61 In these complexes, the four walls of the cavitand form a seam of intramolecular hydrogen bonds that give rise to a well-defined cavity in which ferrocene fits nicely. Again, 1 H NMR experiments verify the formation of this inclusion complex (Fc@6), which has substantial kinetic stability. To our surprise, this inclusion complex turned out to be voltammetrically silent! We also investigated the formation of inclusion complexes between host 6 and guests FcOH, FcNMe3 þ , and cobaltocenium. Our data show that all of these complexes are also voltammetrically silent.61 These results are particularly surprising, because an inclusion complex of 6 (see Fig. 3.12) has a molecular weight similar to those formed by CD or CBn hosts. With host 6, encapsulation of the redox group is not complete because the cavity has a permanent opening. All of these considerations make our voltammetric findings with host 6 relatively unexpected and hard to rationalize.
82
MOLECULAR ENCAPSULATION OF REDOX-ACTIVE GUESTS
Figure 3.12 Energy-minimized structure (PM3) of the Fc@6 inclusion complex. (See the color version of this figure in Color Plates section.)
3.5
CONCLUSIONS AND OUTLOOK
The results described in this chapter offer a global picture that can be rationalized only in part. In all the inclusion complexes formed by CD or CBn hosts, the redox guest is only partially encapsulated. Both of these host families contain two openings to their internal cavities. Furthermore, CD complexes have limited thermodynamic stability and short lifetimes (dissociation is kinetically very fast). As a result, the voltammetric behavior of redox guests in the presence of CD hosts leads to electron transfer upon complex dissociation. Direct electron transfer of the redox CD complex is usually not observed. CBn complexes tend to be much more kinetically stable and, as a result, direct electron transfer is readily observed. We have detected however major differences between the behavior of CB-viologen complexes and CB-ferrocene complexes. In the former, there is no detectable attenuation in the electron transfer kinetics upon encapsulation, while in the latter measurable kinetic attenuation is observed. This may be related to the more effective encapsulation of the redox-active center in the ferrocene complexes. Finally, the work with cavitand hosts shows that substantial kinetic attenuations in the electrochemical rates are possible, as it is the case with Cob þ @46. However, our data seem to point to the importance of solvation, as heavily solvated hosts, such as 5 and 6, are very effective at forming inclusion complexes that are voltammetrically silent. In fact, the self-assembled complex Fc@52 is comparable in molecular weight to the hemicarceplexes Fc@1 or Fc@2. However, the latter complexes show voltammetric response for the included ferrocene in low polarity solutions, while the former does not (in aqueous media). Without question, the most surprising finding is the lack of voltammetric response of redox-active complexes of 6, given the relative small
REFERENCES
83
molecular weight of these complexes. Unfortunately, it is difficult, if not impossible to investigate pH or ion strength effects on the behavior of redox-active guests encapsulated by 5 and/or 6 because these hosts offer limited aqueous solubility and do not permit significant variations in the solution pH or ionic strength. This highlights one of the main obstacles for quick progress in this field: the strong synthetic limitations associated with structural elaboration on these hosts. As more water-soluble hosts become available, our understanding of these phenomena should improve.
ACKNOWLEDGMENTS The author is grateful to the U.S. National Science Foundation for continued support of this research work. The author wishes to acknowledge the many contributions by members of his research group, whose names are listed in the corresponding publications. I am indebted to Dr. Suresh Gadde, Cui Lu, Dagmara Podkoscielny, and Wei Wang for providing figures for this chapter.
REFERENCES 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15.
16.
17.
C. B. Gorman, J. C. Smith, Acc. Chem. Res. 2001, 34, 60. S. M. Biros, J. Rebek Jr., Chem. Soc. Rev. 2007, 36, 93. A. L€utzen, Angew. Chem., Int. Ed. 2005, 44, 1000. M. R. Jonston, M. J. Latter, Supramol. Chem. 2005, 17, 595. J. Rebek Jr., Angew. Chem., Int. Ed. 2005, 44, 2068. D. M. Vriezema, M. C. Aragones, J. A. A. W. Elemans, J. J. L. M. Cornelissen, A. E. Rowan, R. J. M. Nolte, Chem. Rev. 2005, 105, 1445. J. L. Atwood, L. J. Barbour, A. Jerga, Perspect. Supramol. Chem. 2003, 7, 153. M. J. Pilling, P. W. Seakins, Reaction Kinetics, Oxford University Press, Oxford, 1995, Chapter 6. K. A. Connors, Chem. Rev. 1997, 97, 1325. J. Szejtli, Chem. Rev. 1998, 98, 1743. M. V. Rekharsky, Y. Inoue, Chem. Rev. 1998, 98, 1875. J. W. Lee, S. Samal, N. Selvapalam, H.-J. Kim, K. Kim, Acc. Chem. Res. 2003, 36, 621. J. Lagona, P. Mukhopadhyay, S. Chakrabartri, L. Isaacs, Angew. Chem., Int. Ed. 2005, 44, 4844. Y. H. Ko, E. Kim, I. Hwang, K. Kim, Chem. Commun. 2007, 1305. W. S. Jeon, K. Moon, S. H. Park, H. Chun, Y. H. Ko, J. Y. Lee, E. S. Lee, S. Samal, N. Selvapalam, M. V. Rekharsky, V. Sindelar, D. Sobransingh, Y. Inoue, A. E. Kaifer, K. Kim, J. Am. Chem. Soc. 2005, 127, 12984. M. V. Rekharsky, T. Mori, C. Yang, Y. H. Ko, N. Selvapalam, H. Kim, D. Sobransingh, A. E. Kaifer, S. Liu, L. Isaacs, W. Chen, S. Moghaddam, M. K. Gilson, K. Kim, Y. Inoue, Proc. Natl. Acad. Sci. USA 2007, 104, 20737. A. R. Khan, P. Forgo, K. J. Stine, V. T. D’Souza, Chem. Rev. 1998, 98, 1977.
84
MOLECULAR ENCAPSULATION OF REDOX-ACTIVE GUESTS
18. K. Kim, N. Selvapalam, Y. H. Ko, K. M. Park, D. Kim, J. Kim, Chem. Soc. Rev. 2007, 36, 267. 19. T. Matsue, D. H. Evans, T. Osa, N. Kobayashi, J. Am. Chem. Soc. 1985, 107, 3411. 20. B. Sigel, R. J. Breslow, Am. Chem. Soc. 1975, 97, 6869. 21. A. Harada, S. J. Takahashi, Chem. Soc., Chem. Commun. 1984, 6, 45. 22. A. J. Bard, L. R. Faulkner, Electrochemical Methods, Fundamentals and Applications, 2nd edition, Wiley, New York, 2001, Chapter 2. 23. R. Isnin, C. Salam, A. E. Kaifer, J. Org. Chem. 1991, 56, 35. 24. L.A. Godınez, S. Patel, C. M. Criss, A. E. Kaifer, J. Phys. Chem. 1995, 99, 17449. 25. E. Coutouli-Argyropoulou, A. Kelaidopoulou, C. Sideris, G. J. Kokkinidis, G. J. Electroanal. Chem. 1999, 477, 130. 26. D. Osella, A. Carretta, C. Nervi, M. Ravera, R. Gobetto, Organometallics 2000, 19, 2791. 27. C. M. Cardona, T. D. McCarley, A. E. Kaifer, J. Org. Chem. 2000, 65, 1857. 28. P. R. Ashton, V. Balzani, M. Clemente-Leon, B. Colonna, A. Credi, N. Jayaraman, F. M. Raymo, J. F. Stoddart, M. Venturi, Chem. Eur. J. 2002, 8, 673. 29. R. Castro, I. Cuadrado, B. Alonso, C. M. Casado, M. Moran, A. E. Kaifer, J. Am. Chem. Soc. 1997, 119, 5760. 30. M. J. W. Ludden, D. N. Reinhoudt, J. Huskens, Chem. Soc. Rev. 2006, 35, 1122. 31. Y. Wang, S. Mendoza, A. E. Kaifer, Inorg. Chem. 1998, 37, 317. 32. B. Gonzalez, C.M. Casado, B. Alonso, I. Cuadrado, M. Moran, Y. Wang, A. E. Kaifer, Chem. Commun. 1998, 2569. 33. B. Gonzalez, I. Cuadrado, B. Alonso, C. M. Casado, M. Moran, A. E. Kaifer, Organometallic 2002, 21, 2569. 34. A. Mirzoian, A. E. Kaifer, Chem. Eur. J. 1997, 3, 1052. 35. C. Lee, M. S. Moon, J. W. Park, J. Incl. Phenom. Molec. Recog. 1996, 26, 219. 36. J. Kim, I.-S. Jung, S.-Y. Kim, E. Lee, J.-L. Kang, S. Sakamoto, K. Yamaguchi, K. Kim, J. Am. Chem. Soc. 2000, 122, 540. 37. H.-J. Kim, W. S. Jeon, Y. H. Ko, K. Kim, Proc. Natl. Acad. Sci. USA 2002, 99, 5007. 38. W. Ong M. E. Go´mez-Kaifer, A. E. Kaifer, Org. Lett. 2002, 4, 1791. 39. W. Ong, A. E. Kaifer, J. Org. Chem. 2004, 69, 1383. 40. Y. Ling, J. T. Mague, A. E. Kaifer, Chem. Eur. J. 2007, 13, 7818. 41. Y. J. Jeon, H.-J. Kim, C. Lee, K. Kim, Chem. Commun. 2002, 1828. 42. J. Moon, J. Grindstaff, D. Sobransingh, A. E. Kaifer, Angew. Chem., Int. Ed. 2004, 43, 5496. 43. W. Wang, A. E. Kaifer, Angew. Chem., Int. Ed. 2006, 45, 7042. 44. W. Ong, A. E. Kaifer, Organometallics 2003, 22, 4181. 45. L. Yuan, D. H. Macartney, J. Phys. Chem. B 2007, 111, 6949. 46. L. M. Tunstad, J. A. Tucker, E. Dalcanale, J. Weiser, J. A. Bryant, J. C. Sherman, R. C. Helgeson, C. B. Knobler, D. J. Cram, J. Org. Chem. 1989, 54, 1305. 47. D. J. Cram, J. M. Cram, Container molecules and their guests, in J. F. Stoddart (Ed.), in Monographs in Supramolecular Chemistry, Vol. 4, Royal Society of Chemistry, Cambridge, 1994. 48. S. Mendoza, P. D. Davidov, A. E. Kaifer, Chem. Eur. J. 1998, 4, 864.
REFERENCES
85
L. R. MacGillivray, J. L. Atwood, Nature 1997, 389, 469. A. Shivanyuk, J. Rebek Jr., Proc. Natl. Acad. Sci. USA 2001, 98, 7662. I. Philip, A. E. Kaifer, J. Org. Chem. 2005, 70, 1558. L. Avram, Y. Cohen, Org. Lett. 2003, 5, 2329. L. Avram, Y. Cohen, J. Am. Chem. Soc. 2003, 125, 1618. I. Philip, A. E. Kaifer, J. Am. Chem. Soc. 2002, 124, 12678. C. M. Cardona, S. Mendoza, A. E. Kaifer, Chem. Soc. Rev. 2000, 29, 37. C. S. Cameron, C. B. Gorman, Adv. Funct. Mater. 2002, 12, 17. H. Mansikkamaki, C. A. Schalley, M. Nissinen, K. Rissanen, New J. Chem. 2005, 29, 116. C. L. D. Gibb, B. C. Gibb, J. Am. Chem. Soc. 2004, 126, 11408. S. M. Biros, E. C. Ullrich, F. Hoff, L. Trembelau, J. Rebek Jr., J. Am. Chem. Soc. 2004, 126, 2870. 60. D. Podkoscielny, I. Philip, C. D. L. Gibb, B. C. Gibb, A. E. Kaifer, Chem. Eur. J. 2008, 14, 4704. 61. D. Podkoscielny, R. J. Hooley, J. Rebek Jr., A. E. Kaifer, Org. Lett. 2008, 10, 2865. 62. R. J. Hooley, H. J. Van Anda, J. Rebek Jr., J. Am. Chem. Soc. 2007, 129, 13464. 49. 50. 51. 52. 53. 54. 55. 56. 57. 58. 59.
CHAPTER 4
Dendritic Encapsulation of Redox-Active Units CHRISTOPHER B. GORMAN Department of Chemistry, North Carolina State University, Raleigh, NC, USA
4.1
INTRODUCTION
The topology of dendrimers is a core (focal group) surrounded by layers of hyperbranching repeat units and a large percentage (at least 50%) of the units as “fingertips” (peripheral units). This description is simple and, in some important ways, imprecise. It suggests that primary structure is equivalent to tertiary structure, and that assumption is not necessarily so. Nevertheless, it does suggest that the hyperbranches and peripheral groups encapsulate the core (and groups either covalently or noncovalently held in the vicinity of the core). Furthermore, it seems reasonable that, as more and more “layers” of hyperbranching surround the unit, the unit will become less and less accessible to the outside. Indeed this type of behavior has been illustrated in several ways. This chapter will review and explore them. This book focuses on electrochemical phenomena, and this chapter concentrates on the effects of encapsulation of redox-active cores on electrochemical behaviors. However, before discussing this specific topic in detail, a short survey of the scope of encapsulation and of redox-active dendrimers is in order. Several behaviors have been explored for encapsulated units that extend beyond those discussed here. These are mentioned briefly to indicate the wide variety of roles that dendritic encapsulation can play. Fluorescence quenching of an encapsulated chromophore can be attenuated by reducing the relative accessibility of dioxygen (a small molecule)1–7 or larger units such as other chromophore-encapsulated
Electrochemistry of Functional Supramolecular Systems. Edited by Paola Ceroni, Alberto Credi, and Margherita Venturi Copyright 2010 John Wiley & Sons, Inc.
87
88
DENDRITIC ENCAPSULATION OF REDOX-ACTIVE UNITS
dendrimers8–10 and dendrimers with pendant donors and acceptors.11–15 Encapsulation may also be a key feature in light harvesting and energy funnels16–27and nonlinear optical effects.28–31 The reactivity of catalytic sites can also be tuned by dendritic encapsulation.32–36 This idea includes the synthesis of, encapsulation of, and catalysis by metal nanoparticles in dendrimers.37 Even within the scope of electrochemical behaviors in dendrimers, there are many interesting examples that do not strictly involve encapsulation. As above, some are mentioned briefly to indicate the wide scope of redox-active dendrimers. Many of these involve peripheral functionalization of the dendrimer with redox-active units.38–55 These are very interesting and potentially useful architectures and may, for example, serve as new types of redox mediators, sensors, or elements of supercapacitors. Other intriguing electrochemical behaviors of dendrimers include “redox sponges,”56–59 redox gradients for “one-way” radial charge transport,60,61 charge “trapping,”57 and dendritic insulation for molecular wires.62 Finally, a number of metal-containing dendrimers have been prepared, but their redox activity and encapsulation effects have been probed very little.63–66 These also could eventually be relevant to the unfolding story presented here. To explain dendritic encapsulation, one needs to know something about dendrimer conformations and dynamics. These can be complex and difficult to map out. This problem leads to a “chicken and egg” conundrum (to “crack” a bad encapsulating joke). Ideally, one has information about structure that can be used to rationalize encapsulation behavior. However, as it is difficult to get structural (particularly conformational) information about dendrimers, structures are often proposed based on observed encapsulation behaviors. Thus, a central theme of this chapter will be that the redox unit, as it is perturbed by the structure of the dendrimer, acts as both “embedded” reporter and performer. This approach runs the risk of generating a circuitous argument. However, in combination with other methods of assessing structure, it hopefully can iterate to a self-consistent model. In this chapter, a dendrimer will be generally characterized by its three main parts. The core is the topological center of the dendrimer (from which all the dendritic arms extend). Dotted lines show where the core attaches to the repeat units. For example, Core1 attaches to four repeat units. Hyperbranching occurs at every repeat unit. For example, Rpt1 attaches to the core through the amido linkage (the dotted line to the right) and hyperbranches into three branches (the three dotted lines to the left). Some repeat units are so common that they are further denoted by the principal author in which they were first described in a dendrimer. After hyperbranching “n” times (“n” is also referred to as the generation of the dendrimer), each site is capped by a peripheral unit. All peripheral units are shown with a single dotted line indicating the site at which the last of each of the hyperbranches is attached to it. This method of drawing dendrimers is abbreviated and thus somewhat cumbersome. However, to save space and to illustrate the commonality between structures of different dendrimers, they are drawn and classified in this way.
4.2
89
ENCAPSULATION MODULATES REDOX POTENTIAL
O OCH3
O
O N N
M
O
O
N
H N
O
N O
O
Periph1 O n
O O
Core 1 Rpt 1(Newkome)
Periph2
4.2 ENCAPSULATION MODULATES REDOX POTENTIAL: PORPHYRIN CORE DENDRIMERS AND BEYOND As larger and larger dendrimer units are placed around the core, they influence the environment around the core and thus its chemical (and redox) potential. This section gives several examples (and two counterexamples) of this behavior. Furthermore, a general model to explain how environment influences redox potential is provided. In 1994, Diederich’s group described the synthesis and properties of porphyrin core dendrimers of the type Core1–Rpt1(Newkome)–Periph1, M ¼ Zn, n ¼ 1 3.67 They observed that it became thermodynamically more difficult to reduce the core (e.g., a more negative reduction potential) as the generation increased. As the value of n increased from 1 to 3, the first reduction potential changed from 1.6 to 1.9 V (versus SCE (saturated calomel electrode) in CH2Cl2/100 mM Bu4NPF6, DME (dropping mercury electrode)), a 300 mV change. This behavior was explained by the relatively electron-rich environment provided by the dendrimer. Later discussions further suggested that, as reduction charged the core (from neutral to anionic), the relatively less polar dendrimer stabilized that charge less than the relatively more polar solvent.64 In water-soluble porphyrin core dendrimers68 of the type Core1–Rpt1(Newkome)– Periph2, M ¼ Zn and FeCl, n ¼ 1 2, a 420 mV more positive oxidation potential was required to reach the oxidized, more charged Fe(III) state when n ¼ 2 was compared to n ¼ 1 ( þ 190 mV versus 230 mV, versus SCE in H2O/100 mM Et4NClO4, GCE (glassy carbon electrode)). These observations were all consistent with the rationale that, with increasing generation, the dendrimer microenvironment (which was less polar) became dominant over the solvent microenvironment (which was more polar) and stabilized the more charged, oxidized state of the core less than did the solvent.
O
CH3
3
90
DENDRITIC ENCAPSULATION OF REDOX-ACTIVE UNITS
These early studies established the paradigm that the relative polarity of dendrimer versus solvent and the relative change in charge upon oxidation (or reduction) explained the shifts in redox potential. Significantly, these shifts are comparable to those found when comparing the redox potential of cytochrome c to small, solventexposed heme models. This model has been reviewed for several series of redoxactive core dendrimers.64 Finally, in these studies, it was noted that as the generation of the dendrimer increased, the shape of the CV wave flattened and broadened, as would be expected for reduced rates of heterogeneous electron transfer. This observation is expanded upon quantitatively in Section 4.5. This group went on to study dendritic iron porphyrins with tethered axial ligands69–71 and then a series of other, porphyrin core dendrimers, focusing on their behaviors as globular heme models.72 O
O
O
O
N
N
O n
Zn N
N O
O
O Rpt 2(Fréchet)
O
Periph3
O Core2
Frechet and Abrun˜a’s groups synthesized and studied a series of zinc tetraphenyl porphyrin core dendrimers of the form Core2–Rpt2(Frechet)–Periph3, n ¼ 0–4.73 Both the oxidation potential and the reduction potential increased within this series of molecules as the generation increased. Specifically, the first oxidation potential increased from 810 to 1120 mV as the generation increased from 1 to 3 (versus Ag/AgCl in 100 mM Bu4NClO4/CH2Cl2) and the first reduction potential increased in magnitude from 1122 to 1610 mV (versus Ag/AgCl in 100 mM Bu4NClO4/DMF). An increase in the separation between the anodic and cathodic peak potentials for a given redox transition (DE) was also observed. This behavior is relevant to discussions in Section 4.54.5 and is treated there. Later work in the Frechet group focused on modulation of photophysical properties,74–77 and this group developed extensive sets of data illustrating how these observations probed the effect of site isolation (encapsulation).78 Nierengarten et al. looked at the effects of pendant fullerenes on the electrochemical potentials of a porphyrin core. Molecule 1 had an oxidation potential 175 mV higher than meso-tetrakis(3,5-di-tert-butylphenyl)porphyrin (1195 mV compared
4.2
ENCAPSULATION MODULATES REDOX POTENTIAL
91
to 1020 mV versus SCE in CH2Cl2/100 mM Bu4NPF6, Pt electrode). This was attributed either to the electron-withdrawing effect of the C60 substituents or to solvation effects due to their presence.79 OC12H25 O O
O
OC12H25
R
R
O
N HN
NH
R=
N
O R
R
O
OC12H25
O O
1
OC12H25
Nierengarten et al. expanded upon their C60-based dendrimers to make a series of copper bis-phenanthroline core dendrimers of the form Core3–Rpt3–Periph4.80 In this series, the redox potential of the core did not vary notably with generation, but the kinetics slowed down as the generation increased. The CV wave of the second- and third-generation dendrimers could not be observed. OC8H17
O
O
O O
O
O
O
O
N
O
O
O
N Cu+
O
n
O
N N
O
O
O
O
O
O O
O Core3
OC8H17
O
O O
Rpt3
O O Periph4
OC8H17 OC8H17
V€ ogtle et al. illustrated this encapsulation behavior on a series of Ru(bpy)32 þ core dendrimers of the form Core4–Rpt2–Periph3 and Core4–Rpt2–Periph5.81 Comparison with the behavior of [Ru(Me2bpy)3]2 þ as a model compound showed that, as the generation of the molecule increased, oxidation occurred at a more positive potential, reduction occurred at less negative potential, and shape of the CV and differential pulse voltammograms indicated that the observed processes were not fully reversible. These effects were, however, relatively small (e.g., a 10–20 mV shift in going from the model compound to a second-generation dendrimer.
92
DENDRITIC ENCAPSULATION OF REDOX-ACTIVE UNITS
N N
N Ru N
O
N N
Periph5
Core4
The redox potentials of Co(II) bound tyrosine residues in dendrimers of the form Core5–Rpt2–Periph3 were assessed using CV.82 The value of E1/2 for the Co(II) to Co (I) reduction initially decreased in magnitude from 780 to 580 mV in going from n ¼ 0 to n ¼ 2 (versus Ag wire quasi-reference, 100 mM Bu4NClO4 in DMF (dimethyl formamide), GCE). It then was reported to increase in magnitude to 830 mV for n ¼ 3 and 880 mV for n ¼ 4. However, from the narrative, it was unclear why the authors were able to measure a thermodynamic redox potential as they commented on the broadness of their CV data. At least for the first few points, the trend in the redox potential follows the established paradigm. As the charge upon reduction decreases, the relatively less polar environment of the dendrimer compared to the solvent destabilizes the reduced form less than the oxidized form.
O
CH3
H2 N
O
N
Co N H2
N O
O
O Fe
O
Fe N
O O
N CH3
O
N
N
Periph6
CH3 Core5
Core6
Enomoto and Aida investigated the redox activities of a series of iron-oxo tricyclononane core dendrimers of the form Core6–Rpt2–Periph6. When the generation of the dendrimer increased from n ¼ 1 to n ¼ 4, the redox potential (E1/2) of the interior diiron(III) center became more negative from 319 to 393 mV (versus SCE using Fc/Fc þ as reference, 200 mM nBu4PF6 in CH3CN, Pt electrode) following the same paradigm as above. Liwporncharoenvong and Luck showed perhaps the most dramatic redox potential shift in a series of Mo2 tetrabenzoate cluster core dendrimers of the form
4.2
ENCAPSULATION MODULATES REDOX POTENTIAL
93
Core7–Rpt4–Periph7. In these, the value of E1/2 shifted from 588 mV for n ¼ 1 to 785 mV for n ¼ 3 (versus Ag/AgCl, 100 mM Bu4NPF6 in benzonitrile, GCE). The aromatic ester repeat unit and solvent are unique to this investigation and may be important considerations in addition to the environmental sensitivity of this core unit. O
O O
O O
O
O O
O
Mo
O
O
Mo
O
O
O
O
n
O
O
O
O
Rpt4
Core7
O
Periph7
Smith et al. expanded considerably on the variability in the relative polarity of dendrimer versus solvent in a study of several series of ferrocenyl core dendrimers including those of the form Core8–Rpt1(Newkome)–Periph8 and Core9–Rpt1(Newkome)–Periph8.83 A number of conclusions were drawn. The presence of the dendrimer shell made it more difficult to oxidize the ferrocenyl core. In CH2Cl2, the importance of the electrolyte concentration was nicely illustrated by varying the concentration and graphing the redox potential of ferrocene against the ET(30) solvent polarity index. A roughly linear correlation was observed. In methanol, little change in redox potential was observed when the generation of the dendrimer was increased. This behavior was attributed to facile interaction between the methanol and the dendritic shell. Thus, the simple model where the relative polarity of solvent versus dendrimer environment is compared has caveats/exceptions. Comparison to similar molecules reported by Kaifer et al. showed the opposite trend, a result that is still not well explained (cf. Section 4.5)
Fe
O
O
Core8
OCH3
Fe O
Core9
Periph8
Perhaps a most striking example of macromolecular control of redox potential was observed in molecule 3. In the DPV of this molecule, two oxidation peaks were
94
DENDRITIC ENCAPSULATION OF REDOX-ACTIVE UNITS
observed at 540 and 640 mV (SCE, 100 mM NaClO4, GCE) with an intensity ratio of approximately 1:2. These were assigned as the cisoid (3c) and transoid (3t) conformations of the molecule, respectively.84 GluO GluO GluO GluO GluO GluO
OH
H N H N
GluO GluO GluO
Fe
O
H N
O Fe
N H
O
OGlu OGlu
O OGlu =
OH OH
HO O
OGlu
O 3c
3t
Thus, these many examples establish the model in which a dendrimer can create a microenvironment around the redox unit that has a large effect on redox potential. No paradigm is fully formed without some counterexamples, however. Two are provided here. Gorman et al. showed that in Fe4S4 cluster core dendrimers with Frechet groupderived repeat units of the form Core10–Rpt5(Frechet)–Periph3, the change in redox potential in going from n ¼ 0 to n ¼ 3 was relatively small compared to the examples above.85 For n ¼ 0, E1/2 ¼ 970 mV and for n ¼ 3, E1/2 ¼ 1080 mV (versus Ag/ AgCl in 100 mM Bu4NPF6/DMF, Pt electrode). In this series, a dianion was being reduced to a trianion. Thus, the starting state was already highly charged, and the change in charge upon reduction was relatively smaller than the cases above.64 O
N O
S S S
S
Fe
Fe
S Fe
S
S
Fe
N O
S
N
Rpt 4(Fréchet)
O
Core4
n O
O
N
In a second example, Astruc et al. showed that the redox potential of FeCp(h6arene) core dendrimers (molecule 2 is the largest example studied) was similar to smaller molecules.86 The degree of hyperbranching is minimal in this molecule. Moreover, the authors also pointed out that the redox orbital is quite well buried at
4.3 CREATIVE EXPLORATIONS OF TOPOLOGY: FROM METAL BIPYRIDINE- AND TERPYRIDINE
95
the metal center in this redox unit. Thus, the environmental influences on redox potential for some units can be small.
O O O
O
O O
O Fe O
O O
O 2
O
4.3 CREATIVE EXPLORATIONS OF TOPOLOGY: FROM METAL BIPYRIDINE- AND TERPYRIDINE-CONTAINING DENDRIMERS TO OFF-CENTER FERROCENES At about the same time that these fundamental studies of molecular encapsulation were beginning to appear, other groups were exploring dendrimer topology broadly. Historically, at this point, there were few examples of metallodendrimers, so many creative example architectures were possible and were shown. Many different architectures and behaviors were being explored, so thematically these examples are diverse. Balzani’s group synthesized a series of ruthenium- and/or osmium-containing complexes of the general form 4 below.87–90 When all of the metal sites were the same, a single oxidation peak was observed in the CV with a size indicative of simultaneous and independent oxidation of all of the sites. When the metal sites were different, the oxidation behavior of the different sites could easily be distinguished. The group went on to study similar molecules in liquid SO2 solvent with a larger electrochemical potential window and observed additional oxidations and reductions not observable with conventional organic solvents.91 The idea of simultaneous oxidation of more or less chemically equivalent metal centers has been illustrated in a variety of metallodendrimers. This simple model, however, has been challenged by more careful voltammetry studies where modeling of multisite redox dendrimer voltammograms illustrated how closely spaced redox states could, in fact, be distinguished.92,93 Additional studies of these types of molecules have focused on their photophysical properties.94,95
96
DENDRITIC ENCAPSULATION OF REDOX-ACTIVE UNITS
Lp
Lp
Lp
Mp
Mp
Li
Mi
Li
N
Lp
N
N
N
Li or Lp Lp
Mp
Lp
Mc Li M i
Li
Li
Mi Li Li
Li
Mp
N
N
N
N
Lp Li
Lp
Mp Mp L p Lp Lp Lp Mc, Mi, Mp = Ru and/or Os 4
Another example of many interesting architectures prepared by the Newkome group are the isomeric metallodendrimers 5 and 6. The E1/2 values (versus Fc/Fc þ in 100 mM Et4NBF4/CH3CN, GCE) of the two waves of 5 were slightly (ca. 30 mV) more negative than that of 6. This difference might result from differences in solvent environment between the two isomers or differences in the relative interaction of the counterions with the electroactive moieties.96,97
Newkome et al. also prepared a series of ruthenium bis-terpyridine core dendrimers of the form Core11–Rpt6(Newkome)–Periph9.98 In this series of molecules, the two dendrons could be of different generations. They found, as expected, that the redox center became more difficult to oxidize as the total number of generations around the core increased. Thus, for the molecule in which both the dendrons
4.4
ESTABLISHING A GENERAL ESTIMATE OF ELECTRON
97
were n ¼ 1, E1/2 ¼ 747 mV (versus Fc/Fc þ , 100 mM Bu4NPF6 in CH3CN, GCE). For the molecule in which both dendrons were n ¼ 2, E1/2 ¼ 791 mV, and for the molecule in which one dendron was n ¼ 1 and the other dendron was n ¼ 4, E1/2 ¼ 880 mV.
N O
N Ru
N
N
O
H N
O O n
11 N
O
N
Core11
Rpt 6(Newkome)
Periph9
Astruc et al. have prepared dendrimers with ferrocenyl amido peripheries and have been very active in studying how the electrochemical behavior of these molecules can be used to recognize anions. Such molecules display large shifts in redox potential in the presence of H2PO4 and HSO4 anions.38 For example, in the presence of 1 equivalent of nBu4N HSO4, a dendrimer containing 3 peripheral ferrocenyl amido units showed a 30 mV shift, a dendrimer containing 9 peripheral ferrocenyl amido units showed a 65 mV shift, and a dendrimer containing 18 peripheral ferrocenyl amido units showed a 130 mV shift. Kaifer et al. showed similarly large shifts in a poly (propylene imine) dendrimer with 4, 8, 16, or 32 peripheral ferrocenyl urea terminal groups.99 These behaviors are nice examples of the potential for cooperative behavior in redox-active dendrimers that is not observed in small molecule analogues. Several additional examples have been presented with ferrocenyl-capped dendrimers of varying structures.39,100–105
4.4 ESTABLISHING A GENERAL ESTIMATE OF ELECTRON TRANSFER RATE ATTENUATION In the sections above, several mentions were made to the increase in DE between the anodic and cathodic waves in the CV with increasing dendrimer generation. This phenomenon is quite general (although there are the inevitable exceptions that are presented in Section 4.6). Of the examples presented above where a series of molecules of increasing generation were prepared, most, if not all, showed broadening CV waves and increasing DE values with increasing generation. In this section, several additional examples are reviewed. Fitting the CVor other type of voltammetry data in conjunction with the measure of the molecular diffusion coefficient in solution can provide quantitative values of the heterogeneous rate constant (k ). These types of investigations are presented in Section 4.5. Chow et al. prepared a set of dendrimers of the form Core12–Rpt7–Periph10, M ¼ Ru and observed a similar trend in redox potential changes to that reported by Newkome et al. above.106 They also graphed peak separation between the anodic
98
DENDRITIC ENCAPSULATION OF REDOX-ACTIVE UNITS
and cathodic waves (DE) versus molecular size (actually gel permeation chromatographic retention time) and were able to illustrate slowing electron transfer semiquantitatively in this way. N
N M
N
O N
O
N
N
n
O
O M = Ru or Fe
Cor e12
Rpt 7
Periph10
Chow et al. also prepared a series of iron bis-terpyridine core dendrimers of the form Core12–Rpt7–Periph10, M ¼ Fe.107 These molecules showed relatively little change in redox potential with increasing generation, however. The value of E1/2 for n ¼ 1 was 1020 mV (versus Fc/Fc þ , 100 mM Bu4NBF4 in CH2Cl2, Pt electrode) and the value for n ¼ 3 was 1120 mV (versus Fc/Fc þ , 100 mM Bu4NBF4 in THF (tetrahydrofuran), Pt electrode). Several electrochemical solvents were used across the series, so perhaps little should be concluded on this point. The peak separation (DE) over this series did change from 80 to 200 mV, however, indicating that the dendrimer did slow the electron transfer rate.
H3CO
O OCH3
O N N
N M
N
N N
N N H3CO
O O
Core13
OCH3
Kimura et al. showed that in cobalt phthalocyanine dendrimers of the form Core13– Rpt6–Periph9, M ¼ Co, n ¼ 1–2, the first-generation molecule showed a reversible redox wave but the third generation did not.108 The lack of a reversible wave was consistent with slow electron transfer kinetics. Majoral et al. showed that ferrocenyl core dendrimers of the form Core14–Rpt8–Periph11 ceased to show a discernable
4.5 ESTABLISHING A QUANTITATIVE MEASURE OF ELECTRON TRANSFER
99
redox wave at the third generation as well.109 Majoral et al. also showed that, as ferrocenyl units were placed at generations farther from the periphery, the redox kinetics qualitatively slowed down as evidenced by the shape and peak separation of CV waves.110 A similar qualitative picture was obtained by Newkome et al. on electroactive dendrimers containing diaminoanthraquinones as the electroactive groups buried within the dendrimers.111–113 S P N
N
S P N
Fe
N
N O
N
n
O
CHO
P S Core14
Periph11 Rpt 8
Gorman et al. showed this trend in iron–sulfur cluster core dendrimers. The separation between the anodic and cathodic peak potentials (DE) in a series of dendrimers of the form Core10–Rpt5(Frechet)–Periph3, n ¼ 0–4 increased from 130 to 370 mV as n increased from 0 to 3.85
4.5 ESTABLISHING A QUANTITATIVE MEASURE OF ELECTRON TRANSFER RATE ATTENUATION The electrochemical experiments on Fe4S4 cluster core dendrimers were expanded upon to arrive at a quantitative measure of electron transfer rate attenuation with increasing dendrimer generation.114 In addition to obtaining CV data, we also obtained Osteryoung square wave voltammograms (OSWV). OSWV voltammograms are useful for measuring rate attenuation as their shape continues to change over a larger range than for CV as the rate of heterogeneous electron transfer slows down. This behavior contrasts with that of CVs that, in the slower range of charge transfer rates, become broad and featureless and much more difficult to fit quantitatively. The shape of any voltammogram of a freely diffusing species in solution depends on both the rates of mass transfer and the rate of charge transfer at the electrode. Thus, to obtain the value of k from voltammetry, one must have an independent measure of mass transport—specifically the molecular diffusion coefficient (D0). In our experiments, values of D0 were obtained independently using both chronoamperometry and pulsed field gradient spin-echo NMR spectroscopy. Reasonably good correlation was obtained in this particular case. However, particularly for molecules containing multiple redox centers, these two values do not necessarily agree. Abrun˜a et al. have illustrated this behavior and explained it in detail.115 Readers intent upon performing quantitative electrochemical
100
DENDRITIC ENCAPSULATION OF REDOX-ACTIVE UNITS
measurements on redox-active molecules are encouraged to read their work. In this series of dendrimers, value of k decreased from 6.15 103 to 0.13 103 cm/s as the value of n increased from 1 to 4.
S S S
S
Fe
Fe
S Fe
S
S
Fe
n S
Core15
Rpt 9(Moore)
Periph12
In addition to the dendrimers of the form Core10–Rpt5(Frechet)–Periph3, n ¼ 0–4, a second series of phenyl acetylene-based dendrimers were prepared of the form Core15–Rpt9(Moore)–Periph12. Heterogeneous electron transfer rate constants indicated that this latter, more rigid series of dendrimers was more effective at attenuating the rate of electron transfer than were the former, more flexible series of dendrimers. For example, for n ¼ 2, the flexible dendrimer had a value of k ¼ 3.29 103 cm/s but the rigid dendrimer had a value of k ¼ 2.35 103 cm/s. Using a conformational searching/quenched molecular dynamics protocol, these results could be rationalized. An offset and mobile iron–sulfur core was depicted in the flexible series of molecules and a more central and relatively immobile iron–sulfur core was depicted in the rigid series of molecules. Offset and mobile results in a smaller effective electron transfer distance than central and immobile for a given molecular size. Since a computational model of the conformational manifold of the molecules was available, it tempted us to consider how the rate of electron transfer was diminishing with increased distance. In this case “distance” is some type of average value measuring the closest point to the molecular exterior from the core. However, even given that crude estimate, when the slope of ln(k ) versus distance was plotted, a value 1 (b) of 0.4 A was obtained for the flexible series of dendrimers. This value is smaller than that found for other largely saturated architectures such as those found in proteins. The nonspherical shape of the dendrimer models likely contributed to this lower value. When the same slope was obtained for the rigid series of molecules, b was found to be having an unreasonably low value of 0.16 A1. In addition to the aforementioned concern about nonsphericity of shape, this low value likely reflects more efficient coupling between the donor (the iron–sulfur cluster) and the acceptor (the electrode) due to the rigid, cross-conjugated architecture than was possible in the flexible architecture.
4.5
ESTABLISHING A QUANTITATIVE MEASURE OF ELECTRON TRANSFER
101
We wanted to probe further how dendrimer structural features could influence encapsulation. Furthermore, we wondered how subtle a structural feature could be and still have a measurable influence on electron transfer rate. To do this, a series of iron–sulfur cluster core dendrimer isomer pairs of the form [Fe4S4(Lig)4]2 were prepared. Most notably, one molecule in each pair had a more extended unit (a 3,5-linked phenyl group) and the other a back-folded group (a 2,6-linked phenyl group). In each case, the back-folded isomer was harder to reduce (e.g., had a more negative redox potential) and a had slower electron transfer rate. For example, [Fe4S4(Lig(2,2))4]2 had a value of E1/2 of 1516 mV (versus Ag/Ag þ , 100 mM Et4NBF4 in DMF, Pt electrode) and a value of k ¼ 1.58 103 cm/s, whereas [Fe4S4(Lig(2B,2B))4]2 had a value of E1/2 of 1589 mV and a value of k ¼ 0.45 103 cm/s. Note that these molecules are constitutional isomers of one another. S O
O
O
O
O
O
O
O
O
O
O
O
O O
O
O
O
O
O
S O
O
O O
O Lig(2,3)
S
O O
Lig(3,2)
Lig(2,2) S O
O
O
S
O
O
O O O
O
O O
O O
O
O O
O
O O
O O
O
O
O O
O
S Lig(2B,3)
Lig(3,2B)
Lig(2B,2B)
Using conformational searching/quench dynamics and T1 relaxation measurements, each back-folded isomer was determined to be smaller than its extended counterpart. Thus, the effective distance of electron transfer was not reflected in the hydrodynamic radius of the molecule. Rather, the back-folded isomers were argued to be less mobile with the iron–sulfur core buried more deeply within them. The extended isomers were more mobile with the iron–sulfur core more able to come closer to the molecular surface. By this reasoning, the back-folded isomers had a larger effective electron transfer distance than the extended isomers. What about the effect of coordination number on rate? The iron–sulfur cluster core dendrimers above are “four-armed” dendrimers. We compared these with
102
DENDRITIC ENCAPSULATION OF REDOX-ACTIVE UNITS
“six-armed,” metal tris(bipyridine) core dendrimers of the form Core4– Rpt5(Frechet)–Periph3.116 The same type of solution electrochemistry was explored for this series of molecules. Unsurprisingly, the rate of electron transfer decreased with increasing generation. For example, for M ¼ Fe and n ¼ 1, k ¼ 15.6 103 cm/ s but when n ¼ 3, k ¼ 2.26 103 cm/s. This change initially seemed larger than that for iron–sulfur cluster core dendrimers. Plotting ln(k ) versus the molecular weight difference between Gn and G0 put both the iron–sulfur cluster core dendrimers and these metal tris(bipyridine) core dendrimers on the same line. Thus, more arms make more weight and better encapsulation, but the coordination number itself does not appear to influence the conformational manifold and thus the encapsulation behavior in a way other than this. A number of other elegant, quantitative measures of dendrimer electron transfer rate attenuation have been reported. Cardona and Kaifer117 reported a series of asymmetric ferrocene core dendrimers of the form Core9–Rpt6(Newkome)–Periph9 and measured and modeled their CV behavior (200 mM Bu4NPF6 in CH2Cl2, GCE) to obtain heterogeneous rate constants (k ). Interestingly, the rate of electron transfer changed little with generation. The value of k for the first-generation molecule, n ¼ 1, was 80 103 cm/s, and this value decreased only about one order of magnitude to 5 103 cm/s for n ¼ 3. When two second-generation dendrons were attached (e.g., Core8–Rpt6(Newkome)–Periph9), k ¼ 9 103 cm/s. All of these changes thus resulted in only small changes in the value of k . However, it became easier to oxidize these dendrimers as the generation increased. For n ¼ 1, E1/2 ¼ 0.63 V (versus Ag/AgCl, 200 mM Bu4NPF6, GCE) and for n ¼ 3, E1/2 ¼ 0.54 V. This result is not consistent with any of those presented in Sections 4.2 and 4.3, particularly those reported by Smith et al. above.83 As stated above, no definitive explanation exists for this discrepancy. An intriguing result was obtained when Kaifer et al. cleaved the t-butyl esters of these molecules to give carboxylic acid-terminated dendrimers that were negatively charged at elevated pH.118 They then investigated the electrochemical behavior at a cystenamine-terminated self-assembled monolayer (SAM) on gold, which is positively charged (e.g. SCH2CH2NH3 þ ). At pH 7.4, the third-generation dendrimer (n ¼ 3) is expected to have deprotonated peripheral groups and had a measured k ¼ 6 106 cm/s. At low pH (where the molecule is expected to be uncharged), k ¼ 2 103 cm/s. These data suggest that the ferrocenyl dendron undergoes restricted rotation in the vicinity of the electrode surface at higher pH, tending to place the dendron between the redox unit and the electrode and slowing the electron transfer. The Kaifer group has also reported several other quantitative electrochemical investigations on these and similar dendrimers. Using electrochemistry and modeling the voltammetry to a CE process (electron transfer preceded by a chemical step), they obtained binding constants between the ferrocenyl groups and b-cyclodextrin of 950, 250 and 50 M1 for n ¼ 1, 2 and 3, respectively. These values also indicate an important feature of dendritic encapsulation—restriction of the interaction between a core and a relatively small molecule guest.119 Other investigations included electron transfer rate attenuation and binding between cobaltocenium core dendrons and
4.5
ESTABLISHING A QUANTITATIVE MEASURE OF ELECTRON TRANSFER
103
cucurbit[7]uril,120 and between viologen (4,4-bipyridinium) core dendrons and cucurbit[7]uril.121 Kraatz et al. also prepared and studied formally “off-center” ferrocene core dendrimers with glutamic acid-based repeat units of the form Core9–Rpt10–Periph13. They found modest changes in E1/2 by CV, from 633 mV for n ¼ 1 to 607 mV for n ¼ 6 (versus Ag/AgCl, 200 mM Bu4NClO4 in CHCl3, GCE), indicative of better stabilization of the oxidized ferricinium by the dendrimer than by the solvent. They found equally modest changes in heterogeneous electron transfer rate constants, from k ¼ 20.2 103 cm/s for n ¼ 1 to k ¼ 0.6 103 cm/s for n ¼ 6. Kraatz et al. also prepared and studied disubstituted ferrocene core dendrimers with glutamic acid-based repeat units of the form Core8–Rpt10–Periph13. In contrast to the electrochemical behaviors reported above, large changes in E1/2 and slowing of heterogeneous electron transfer rate were observed. The value of E1/2 changed from 872 mV for n ¼ 1 to 771 mV for n ¼ 3 and then increased slightly to 776 mV for n ¼ 4 (versus Ag/AgCl, 200 mM Bu4NClO4 in CHCl3, GCE). While the CV peak shapes were not fit to determine values of k , the shapes indicated slowing electron transfer kinetics. O
H N n
OC2H5
O Rpt10
Periph13
Chi et al. examined a series of phenyl acetylene-based dendrimers of the form Core16–Rpt11–Periph14 and calculated k from CV data.122 The value of k decreased from 7.5 103 cm/s for n ¼ 1 to 0.86 103 cm/s for n ¼ 4. These values are surprisingly low compared to those discussed above from the Kaifer group and the observations made on phenyl acetylene-based iron–sulfur cluster core dendrimers. Nevertheless, they do follow the established behavioral trend.
Fe
Si(CH3)3
H3CH2CO n
Core16
Rpt11
Periph14
Yamamoto et al. developed a nice system by which the redox units within a dendrimer could be installed in a radial, stepwise fashion.123 Dendrimers of the form
104
DENDRITIC ENCAPSULATION OF REDOX-ACTIVE UNITS
Struct1 bind Fe3 þ and Sn2 þ in a stepwise fashion with the first metal ions binding the most central nitrogens in the dendrimer and subsequent ions binding in layers that are progressively farther from the core.124 This stepwise radial complexation occurs because the nitrogen-based sites are progressively more basic as they are closer to the core. By simulating the CVs, they obtained an overall average rate constant. When two equivalents of ferric ion were added and only the center two imine sites were bound, the average rate constant was determined to be 1.5 103 cm/s (with an estimated average distance between the electroactive sites and the outer shell of 13.7 A). When 30 equivalents of ferric ion were added and all imine sites were bound, the average rate constant was determined to be 1.5 102 cm/s (with an estimated average distance between the electroactive sites and the outer shell of 2.40 A). In the former case, electron transfer had to occur to the topologically central sites. In the latter case, electron transfer could occur to any site, and thus the effective distance required for electron transfer decreased.
N N N N
N
N N N
N
N
N
N
N G1 (2 sites) G2 (4 additional sites, 6 total)
N N
G3 (8 additional sites, 14 total) Struct1
N
G4 (16 additional sites, 30 total)
Yamamoto also explored triphenylamine core dendrimers of the form Core17– Rpt12–Periph15.125 In this system, the redox process studied was the oxidation of this core moiety. They showed that as the generation of the dendrimer increased from 1 to 4, the shape of the CV broadened, indicating slowing electron transfer kinetics.
4.5
ESTABLISHING A QUANTITATIVE MEASURE OF ELECTRON TRANSFER
105
Interestingly, when the fourth-generation molecule was complexed with one equivalent of SnCl2, the redox current increased and the capacitance became smaller. It was concluded that SnCl2 complexation facilitated the electron transfer.
S
N
n
N
N S S
Rpt12
Core17
Periph15
Most recently, Yamamoto et al. prepared a series of porphyrin core dendrimers of the form Core18–Rpt12–Periph15.126 Using a Nicholson analysis of the CV data, they calculated a decrease in the value of k from 2.6 102 cm/s to 5.9 104 cm/s as n increased from 1 to 4. The oxidation potential decreased across this series from 340 to 260 mV (versus Fc/Fc þ in CHCl3CH3CN, GCE) and from 400 to 340 mV (versus Fc/Fc þ in THF, GCE). In this case, easier oxidation with increasing generation presumably reflects better stabilization of the oxidized state by the dendrimer than the solvent. This behavior is the opposite of that observed in more hydrophobic dendrimers (cf. Section 4.2).
N
N Zn
N
N
Core 18
The paradigm behind rate attenuation is based on a simple notion: the conformation of the dendrimer creates an effective distance for electron transfer, and the rate of electron transfer decreases exponentially with this distance. This behavior is what
106
DENDRITIC ENCAPSULATION OF REDOX-ACTIVE UNITS
would be expected for electron transfer via a superexchange tunneling mechanism. There have, however, been several computational efforts to explore the effect of dendrimer topology on electron transfer rate. Branching can create pathways that can, in principle, result in constructive or destructive interference of the superexchange pathways. Given the broad conformational manifold of dendrimers and the subsequent uncertainty over pathways and distance, a simpler, more rigid, branched (or hyperbranched) donor–acceptor model compound seems necessary to test these predictions properly.127–130 4.6 NOT ALWAYS ENCAPSULATING: VIOLOGEN CORE DENDRIMERS As the generation of encapsulating dendrons around a redox-active core increases, the rate of electron transfer decreases. In the section above, many examples conformed with this general model. However, one exception has been illustrated. The cause of this exception is mysterious, however. At about the same time, Kaifer et al.131 and V€ogtle et al.132 both reported the synthesis and electrochemistry of viologen core dendrimers of the form Core19–Rpt2(Frechet)–Periph3. The shape and appearance of CVs of these molecules did not change appreciably for n ¼ 1–3. Thus, these molecules had similar heterogeneous electron transfer rates. Somehow, the dendritic arms did not encapsulate the viologen core. Why?
N
N
O S
S S
2 PF6 Core 19
–
S O
Core 20
I wondered if perhaps the linear geometry of the viologen core might be important. However, Frechet et al.133 studied a series of oligothiophene core dendrimers of the form Core20–Rpt2(Frechet)–Periph3 and showed that for n ¼ 1–3, there was a large change in the shape of the CV. If anything the quaterthiophene core is longer and also linear. This example argues against the relevance of core geometry. Some electronic property of the viologen moiety must be important. Indeed, Kaifer has reported two other classes of viologen core dendrimers of the form Core21– Rpt6(Newkome)–Periph9131,134 and Core22–Rpt2(Frechet)–Periph3135, and both also show negligible change in rate with generation. However, the ability of these core units to associate with guests is influenced by generation in some cases. Garcia et al. reported that the association constant of anthracene with Core19–Rpt2(Frechet)– Periph3 changes from 126 M1 for methyl viologen to 50 M1 for n ¼ 2.136 Kaifer et al.137 showed that the association constant of 7 with Core21–Rpt6(Newkome)–Periph9
4.7
FROM THE SOLUTION TO THE SOLID STATE
107
in acetonitrile changes from 200 M1 for n ¼ 0 (methyl viologen) to 85 M1, but the association constant of Gn (n ¼ generation number) with Core23–Rpt2(Frechet)–Periph3 in acetonitrile changes only from 100 M1 for n ¼ 0 (methyl viologen) to 85 M1 for n ¼ 2. Clearly, we cannot regard the host, the guest, or their interaction in a general way for this system. This statement is true even when one partner is an electrode, and the measure of the interaction is a rate of electron transfer between them. N+
+
N
O
O +
N+
N+
N
O
–
N+
2 PF6
RR RR
O
O
N+
O
Core21
N+ O
O
N+
–
8 PF6 Core22 N+
+
N+
N 2 PF6
O
–
O
O
O
O
O
O
HO Core23
O
O
O
O
7
4.7 FROM THE SOLUTION TO THE SOLID STATE: ELECTROCHEMISTRY OF REDOX-ACTIVE DENDRIMER FILMS There is a great and interesting variety of investigations of redox-active dendrimers on surfaces.138 Not all are directly relevant to dendritic encapsulation and are thus mentioned only in passing here. These include illustration of organized monolayer
108
DENDRITIC ENCAPSULATION OF REDOX-ACTIVE UNITS
arrays of redox-active dendrimers by probe microscopy,139–142 electrochemical desorption of ferrocenyl-terminated dendrimers on cyclodextrin-modified surfaces,143–145 and other electrochemical modulation of dendrimer binding.146 Redox-active dendrimers have also been employed in several examples as mediators in biosensors.41,147–151 Extending work in which dendrimer/SAM mixed adlayers were used as permselective elements for electrochemical sensing,152–156 Crooks et al. showed that a ferrocene-terminated dendrimer/SAM adlayer can display electrochemical current rectification.157 Blackstock et al. showed that mixed arrays of electroactive dendrimers could display directional electron transport and had possible utility in charge trapping.158 Nonelectroactive dendrimers159,160 and dendron-based self-assembled monolayers161 have been explored as partial or selective blocking, surface-confined layers against redox processes by electroactive species in solution. Finally, dendrimers have become much studied electron transport and electroluminescent units in organic light emitting diodes 162–171 and solar cells.172–174 In this regard, the generation of the dendrimer can be important as there is a trade-off between chromophore isolation (to prevent intramolecular luminescence quenching) and chromophore interaction (to maximize the rate of carrier transport).9,175 Dendrimers are large molecules, and their solubility is often an issue. Thus, it is not surprising that dendrimers can be good candidates to precipitate upon and thus modify the surface of an electrode. Dendrimers with ferrocenyl peripheral units were shown to display well-defined redox waves with the ferrocenyl units oxidizing independently.176–178 These molecules oxidatively precipitated onto the electrode surface and were characterized electrochemically and by atomic force microscopy.44 It was later shown that cyclodextrin complexation increased the solubility of these molecules.179 Similar results were obtained with dendrimers containing pendant ruthenium tris (bipyridine) and bis(terpyridine) groups.180
N
N
H N
N n
Core 24
Rpt 13
Fe O Periph16
Silicon-based dendrimers 8 and 9 (Fc ¼ ferrocenyl) also showed oxidative precipitation onto electrodes to give idealized electrochemistry as films. 181 Specifically, the peak current was linear with scan rate and the potential difference between the anodic and cathodic waves was small (DE ¼ 10 mVat a scan rate of 100 mV/s). 182 This latter observation indicated that the rate of electron transfer was rapid. For molecule 9, the surface coverage was measured as GFc ¼ 2 1010 mol/cm2. These molecules were also explored as mediators in amperometric biosensors.183 In contrast, molecule 10 showed two redox peaks, indicative of interaction between the two ferrocenyl units at each peripheral site. 181 When oxidation of one of the two interacting redox units results in some electron sharing between the two units (Robin-Day class II mixed valence species), the second oxidation is naturally
4.7
109
FROM THE SOLUTION TO THE SOLID STATE
more difficult. From the difference in potential between these two peaks, an equilibrium constant can be calculated (e.g., Fc–Fc þ Fc þ –Fc þ fi 2Fc–Fc þ ). For molecule 10, K ¼ 1630 M1. Fc Si Fc
Fc
Si
Si
Si Si Fc
Fc
Si
Si
Si
Fc Fc Si
Si
Si
Si Si
Si
Si
Si
Fc
Fc
Fc Si Si
Fc
Fc 8
9 Fc Fc
Si
Fc Si Fc
Si Si Fc Si
Si
Si
Fc Si Fc Si Fc 10
Majoral et al. showed that phosphorus-based dendrimers with ferrocenyl units at the periphery precipitated onto an electrode and retained their electroactivity.109 They expanded upon this work by preparing dendrimers of the forms Core25–Rpt8–Periph17 and Core25–Rpt8–Periph18 with electroactive tetrathiafulvalene (TTF) groups.184 Dendrimers with TTF peripheries had been studied by Bryce et al.46,47,49–51 Both types of dendrimers formed films on electrodes that showed surface-confined waves. For a fifth-generation dendrimer with the (first shown) repeat unit, a coverage of GTTF ¼ 1.2 108 mol/cm2. For a dendrimer with the TTF crown (second shown) repeat unit, a linear shift of Epa was observed for a barium ion concentration of [Ba2 þ ] ¼ 1 104 to 8 104 M. No further shift in Epa was observed at higher
110
DENDRITIC ENCAPSULATION OF REDOX-ACTIVE UNITS
[Ba2 þ ]. This shift was similar to that observed in solutions of molecules similar in structure to that of the peripheral unit. 3 O
SCH3
O
S
S
O S S P
O
S
S
S S
Core25
O
S
H3CS
SCH3
S
SCH3
S S
S Periph17
Periph18
Kraatz et al. examined the peptide dendrimers described above of the form Core8–Rpt10–Periph13 adsorbed to a mercaptoundecanoic acid SAM.185 Electrostatic interaction between the positively charged ammonium groups in the dendrimers and the negatively charged carboxylate groups in the SAM was expected and found to anchor these molecules to the surface. Generations 1 through 5 of the dendrimer/SAM layers were studied (the sixth-generation dendrimer was examined but did not show any redox activity, presumably because the ferrocenyl group was too well encapsulated by the dendrimer). The values of DE (¼Eox – Ered as defined previously) for the fourth- and fifth-generation dendrimers were 15 and 23 mV, respectively. These values were substantially lower than those obtained for the first- through thirdgeneration dendrimers, indicating not only that these were surface-confined redox processes but also that these larger generation dendrimers were more uniformly organized on the surface. The values of DEfwhm (the width of the redox wave at half maximum) for the fourth- and fifth-generation dendrimers were close to the ideal value of 90 mV, suggesting the formation of uniform and homogeneous films, while values of generations 1–3 were larger, indicating disordered and potentially heterogeneous interactions between ferrocenyl centers. These examples illustrate a broad range of interesting behaviors for surfaceconfined, redox-active dendrimers. However, let us return to our original questions: How does the generation of a dendrimer influence the rate and redox potential for heterogeneous electron transfers? Now we can ask this question with regard to surface-confined dendrimers. We explored these issues on drop-coated (e.g., not monolayer) films of iron–sulfur cluster core dendrimers of the form Core10–Rpt5(Frechet)–Periph3.186 CV and chronoamperometry were employed to evaluate their electrochemical behaviors in propylene carbonate, a nonsolvent environment for the dendrimers. A dramatic change in redox potential with generation was found in these films. The firstgeneration dendrimer had E1/2 ¼ 1250 mV (versus Fc/Fc þ , Pt electrode), but the third-generation dendrimer had E1/2 ¼ 1744 mV, almost 500 mV more negative. We suggested that, in these films, solvent was excluded. In this scenario, the increasingly large fraction of more or less hydrophobic dendrimer resulted it more and more difficult to reduce the iron–sulfur clusters as the generation of the dendrimer increased. In DMF solution, the redox potential of these molecules were similar.
4.8
BACK TO SOLUTION AGAIN: HOMOGENEOUS ELECTRON TRANSFER
111
In solution, it was solvent, not dendrimer environment, that dominated the environment and thus the redox potential. The same conclusions were made when we studied similar films of iron–sulfur cluster dendrimer isomers of the form [Fe4S4(Lig)4]2 described above.187 To consider the effect of dendrimer generation on the rate of electron transfer in these films, we calculated the rate of charge hopping (kex) between oxidized and reduced sites in the film. This rate varied little with dendrimer generation and actually increased from 1.1 103 cm3/mol/s for the first-generation dendrimer to 5 103 cm3/mol/s for the third-generation dendrimer. Clearly, a much different paradigm is at work as this behavior is the opposite to that observed in solution. It turns out that the permeation of counterions into and out of the film is rate determining in this system.188 When the supporting electrolyte was varied from tetramethyl ammonium hexafluorophosphate to tetraethyl-, tetrabutyl-, and tetraoctyl ammonium hexafluorophosphate, the same dendrimer film monotonically showed about fourfold decrease in the value of kex. Plotting the redox potential versus the inverse square root of the ionic strength (the Debye length for charge screening) showed a linear increase in the magnitude of the redox potential with increasing Debye length. Thus, in films, choice and concentration of counterion play a key role in both rate and driving force for electron hopping.
4.8 BACK TO SOLUTION AGAIN: HOMOGENEOUS ELECTRON TRANSFER BETWEEN REDOX-ACTIVE CORE DENDRIMERS In the last section, electron transfer between dendrimers in films was mentioned. This type of hopping is important in thin-film organic devices (see above). How do dendrimers behave in solution? Are the paradigms the same? Actually, they are not. We took the metal tris(bipyridine) cluster core dendrimers of the form Core4– Rpt5(Frechet)–Periph3 and subjected solutions of them to varying amounts of a sacrificial oxidant.189 This process created solutions containing both oxidized and reduced forms of the dendrimers in varying mole fractions. Using NMR line broadening, the values of the self-exchange rate constant could be determined. A large decrease in rate constant was observed in going from n ¼ 0 (e.g., metal tris (dimethyl 2,20 -bipyridine) to n ¼ 1. For n ¼ 0, kex ¼ 2.6 107 M1s1 but, for n ¼ 1, kex ¼ 5.3 104 M1s1, almost 500 times slower. For even larger molecules, kex continued to decrease, albeit more slowly. By measuring the temperature dependence of kex, activation parameters (DHz and z DS ) could be calculated and were reported. However, I am not sure how to physically interpret these numbers. The temperature dependence of rate can be fit to other expressions, and here it is fit to the Marcus equation for nonadiabatic electron transfer in the case of degenerate electron transfer (e.g., DG0 ¼ 0) ket ¼
2p V 2 pffiffiffiffiffiffiffiffiffiffiffiffiffiffi el=4kT h 4plkT
112
DENDRITIC ENCAPSULATION OF REDOX-ACTIVE UNITS
TABLE 4.1 Activation Parameters and Fits to the Marcus Equation for Fe(Gnbpy)32 þ /3 þ Self-Exchange a Generation 0 1 2 3
l (eV) b
| V | (103 eV) c
0.70 (0.02) 1.47 (0.05) 1.28 (0.06) 1.11 (0.09)
1.1 (0.1) 2.6 (0.6) 0.7 (0.2) 0.2 (0.1)
a
Values in parentheses represent the magnitude of the 90% confidence interval. Values at 298K. c Calculated in a fit of the equation above. b
where k is the Boltzmann constant, T is absolute temperature, and l is the reorganization energy. The electronic coupling between the redox centers V can be expressed as V 2 ¼ V02 ebðrr0 Þ where V02 is the maximum coupling between the donor and the acceptor, r is the distance between the closest two atoms in the respective redox center, and r0 is the van der Waals contact distance. Note that both the pre-exponential term and the exponential term have temperature dependence. The results shown in Table 4.1 reflect more than one effect governing the magnitudes of l and |V| with increasing generation. The reorganization energy l increases substantially between the zeroth and first generation but then decreases. The value of |V| increases between the zeroth and first generation but then decreases. Four data points limit how much interpretation these data deserve. The intertangling of the dendrimers (e.g., preassociation), changes in solvation of the redox units with generation, physical diffusion, and other factors all could be important or dominant. More data on various types of dendrimers will really be interesting and key to understanding whether general encapsulation paradigms can be constructed for homogeneous self-exchange.
4.9
CONCLUSIONS
So, do we understand encapsulation effects on electron transfer in dendrimers? To date, I think an honest answer is yes, but with the caveat that general structural paradigms for encapsulation are not yet well developed. These paradigms could guide the synthesis of optimal candidates. Several of the examples provided in this chapter admittedly blend a lot of conjecture with only a little data. Studying a series of dendrimers of increasing generation lends itself to establishing trends, but since generation can only vary from “zero” to maybe five or six, those trends are necessarily based on a small number of points. However, several of the trends illustrated in this
REFERENCES
113
chapter persist across a range of dendrimer structures. This behavior suggests that we can, to a point, develop a structural model of dendrimer encapsulation. This structural model has several interrelated parts. Synthetically, we define structure at the primary level. That structure is the bonding between the units that is defined by the synthetic steps. This structure can also be viewed as the topological structure of the molecule. That structural information, however, is insufficient to define or exploit encapsulation. The missing associations are those that link structure at the primary level, structure at the tertiary level (e.g., conformations), and structure relevant to encapsulation (e.g., range of positions of the encapsulated unit with respect to the center and the exterior of the molecule). This latter structural model of encapsulation could then be correlated with encapsulation behaviors. With this information in hand, one might establish the map from molecular design to behavior. What is the state of progress in establishing this model? Like many overtures in the design of new supramolecular objects, dendrimers capable of efficient encapsulation started with simple and often only partially correct ideas. It seemed plausible that increasing the generation of the dendrimer would increase the degree of encapsulation. However, the relationship between dendrimer size and encapsulation behavior was unclear. Both experimental data and computational modeling190–192 have suggested that, rather than packing with a dense shell, dendrimers pack more with a dense core type of conformation. This generalization is particularly true for dendrimers dissolved in solvents of “poor” to “moderate” solvating power, which is often the case. However, this paradigm does not elucidate how, for a given size, different dendrimer architectures influence encapsulation. Only a few sets of data (cf. comparison of dendrimer isomers)193 touch on this. So, given the range of unexplored potential structures and the synthetic creativity of the chemical community, any of the paradigms elucidated here could be modified or overturned with a new, cleverly designed, hyperbranched molecule.
ACKNOWLEDGMENT We thank the National Science Foundation (Grant CHE- 0315311) for support.
REFERENCES 1. 2. 3. 4. 5. 6. 7.
R. Sadamoto, N. Tomioka, T. Aida, J. Am. Chem. Soc. 1996, 118, 3978. J. P. Collman, L. Fu, A. Zingg, F. Diederich, Chem. Commun. 1997, 193. V. Rozhkov, D. Wilson, S. Vinogradov, Macromolecules 35, 2002, 1991. S. Van Doorslaer, A. Zingg, A. Schweiger, F. Diederich, ChemPhysChem 2002, 3, 659. S. A. Vinogradov, L.-W. Lo, D. F. Wilson, Chem. Eur. J. 1999, 5, 1338. S. A. Vinogradov, D. F. Wilson, Chem. Eur. J. 2000, 6, 2456. O. Finikova, A. Galkin, V. Rozhkov, M. Cordero, C. H€agerh€all, S. Vinogradov, J. Am. Chem. Soc. 2003, 125, 4882.
114
DENDRITIC ENCAPSULATION OF REDOX-ACTIVE UNITS
8. C. Devadoss, P. Bharathi, J. S. Moore, Macromolecules 1998, 31, 8091. 9. J. M. Lupton, I. D. W. Samuel, R. Beavington, P. L. Burn, H. B€assler, Adv. Mater. 2001, 13, 258. 10. M. J. Frampton, S. W. Magennis, J. N. G. Pillow, P. L. Burn, I. D. W. Samuel, J. Mater. Chem. 2003, 13, 235. 11. O. S. Finikova, P. Chen, Z. P. Ou, K. M. Kadish, S. A. Vinogradov, J. Photochem. Photobiol. A Chem. 2008, 198, 75. 12. O. Finikova, A. Galkin, V. Rozhkov, M. Cordero, C. Hagerhall, S. Vinogradov, J. Am. Chem. Soc. 2003, 125, 4882. 13. F. V€ogtle, S. Gestermann, C. Kauffmann, P. Ceroni, V. Vicinelli, L. De Cola, V. Balzani, J. Am. Chem. Soc. 1999, 121, 12161. 14. C. M. Cardona, J. Alvarez, A. E. Kaifer, T. D. McCarley, S. Pandey, G. A. Baker, N. J. Bonzagni, F. V. Bright, J. Am. Chem. Soc. 2000, 122, 6139. 15. F. V€ogtle, S. Gestermann, C. Kauffmann, P. Ceroni, V. Vicinelli, V. Balzani, J. Am. Chem. Soc. 2000, 122, 10398. 16. G. M. Stewart, M. A. Fox, J. Am. Chem. Soc. 1996, 118, 4354. 17. M. R. Shortreed, S. F. Swallen, Z.-Y. Shi, W. Tan, Z. Xu, C. Devadoss, J. S. Moore, R. Kopelman, J. Phys. Chem. B 1997, 101, 6318. 18. S. L. Gilat, A. Adronov, J. M. J. Frechet, Angew. Chem., Int. Ed. Engl. 1999, 38, 1422. 19. F. V€ogtle, S. Gestermann, R. Hesse, H. Schwierz, B. Windisch, Prog. Polym. Sci. 2000, 25, 987. 20. M. Choi, T. Aida, T. Yamazaki, I. Yamazaki, Angew. Chem. 2001, 40, 3194. 21. M. Maus, R. De, M. Lor, T. Weil, S. Mitra, U. M. Wiesler, A. Herrmann, J. Hofkens, T. Vosch, K. M€ullen, F. C. De Schryver, J. Am. Chem. Soc. 2001, 123, 7668. 22. M. Kimura, T. Shiba, M. Yamazaki, K. Hanabusa, H. Shirai, N. Kobayashi, J. Am. Chem. Soc. 2001, 123, 5636. 23. T. H. Ghaddar, J. F. Wishart, D. W. Thompson, J. K. Whitesell, M. A. Fox, J. Am. Chem. Soc. 2002, 124, 8285. 24. O. P. Varnavski, J. C. Ostrowski, L. Sukhomlinova, R. J. Twieg, G. C. Bazan, T. GoodsonIII, J. Am. Chem. Soc. 2002, 124, 1736. 25. R. Gronheid, J. Hofkens, F. Kohn, T. Weil, E. Reuther, K. M€ ullen, F. C. De Schryver, J. Am. Chem. Soc. 2002, 124, 2418. 26. U. Hahn, M. Gorka, F. V€ogtle, V. Vicinelli, P. Ceroni, M. Maestri, V. Balzani, Angew. Chem., Int. Ed. 2002, 41, 3595. 27. K. R. J. Thomas, A. L. Thompson, A. V. Sivakumar, C. J. Bardeen, S. Thayumanavan, J. Am. Chem. Soc. 2005, 127, 373. 28. T. Le Bouder, O. Maury, A. Bondon, K. Costuas, E. Amouyal, I. Ledoux, J. Zyss, H. Le Bozec, J. Am. Chem. Soc. 2003, 125, 12284. ¨ rtegren, H. Ihre, U. W. Gedde, A. Hult, G. Andersson, A. Eriksson, 29. P. Busson, J. O M. Lindgren, Macromolecules 2002, 35, 1663. 30. H. Ma, A. K. Y. Jen, Adv. Mater. 2001, 13, 1201. 31. S. Yokoyama, T. Nakahama, A. Otomo, S. Mashiko, J. Am. Chem. Soc. 2000, 122, 3174. 32. D. Astruc, F. Chardac, Chem. Rev. 2001, 101, 2991. 33. L. J. Twyman, A. S. H. King, I. K. Martin, Chem. Soc. Rev. 2002, 31, 69.
REFERENCES
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.
115
A. M. Caminade, P. Servin, R. Laurent, J. P. Majoral, Chem. Soc. Rev. 2008, 37, 56. D. Mery, D. Astruc, Coord. Chem. Rev. 2006, 250, 1965. B. Helms, J. M. J. Frechet, Adv. Synth. Catal. 2006, 348, 1125. R. M. Crooks, M. Q. Zhao, L. Sun, V. Chechik, L. K. Yeung, Acc. Chem. Res. 2001, 34, 181. C. Valerio, J.-L. Fillaut, J. Ruiz, J. Guittard, J.-C. Blais, D. Astruc, J. Am. Chem. Soc. 1997, 119, 2588. D. Astruc, C. Ornelas, J. Ruiz, Acc. Chem. Res. 2008, 41, 841. M. Zhou, J. Roovers, Macromolecules 2001, 34, 244. H. C. Yoon, M. Y. Hong, H. S. Kim, Anal. Chem. 2000, 72, 4420. S. D. Holmstrom, J. A. Cox, Anal. Chem. 2000, 72, 3191. C. Amatore, Y. Bouret, E. Maisonhaute, J. I. Goldsmith, H. D. Abrun˜a, ChemPhysChem 2001, 2, 130. K. Takada, D. J. Dıaz, H. D. Abrun˜a, I. Cuadrado, C. Casado, B. Alonso, M. Moran, J. Losada, J. Am. Chem. Soc. 1997, 119, 10763. C.-F. Shu, H.-M. Shen, J. Mater. Chem. 1997, 7, 47. C. Wang, M. R. Bryce, A. S. Batsanov, L. M. Goldenberg, J. A. K. Howard, J. Mater. Chem. 1997, 7, 1189. C. A. Christensen, L. M. Goldenberg, M. R. Bryce, J. Becher, Chem. Commun. 1998, 509. M. R. Bryce, W. Devonport, L. M. Goldenberg, C. Wang, Chem. Commun. 1998, 945. W. Devonport, M. R. Bryce, G. J. Marshallsay, A. J. Moore, L. M. Goldenberg, J. Mater. Chem. 1998, 8, 1361. M. Bryce, P. de Miguel, W. Devonport, Chem. Commun. 1998, 2565. A. Beeby, M. R. Bryce, C. A. Christensen, G. Cooke, F. M. A. Duclairoir, V. M. Rotello, Chem. Commun. 2002, 2950. A. M. Caminade, J. P. Majoral, Coord. Chem. Rev. 2005, 249, 1917. J. Alvarez, T. Ren, A. E. Kaifer, Organometallics 2001, 20, 3543. B. Gonzalez, C. Casado, B. Alonso, I. Cuadrado, M. Moran, Y. Wang, A. Kaifer, Chem. Commun. 1998, 2569. J. B. Christensen, M. F. Nielsen, J. van Haare, M. Baars, R. A. J. Janssen, E. W. Meijer Eur. J. Org. Chem. 2001, 2123. S. Heinen, L. Walder, Angew. Chem., Int. Ed. 2000, 39, 806. S. Heinen, W. Meyer, L. Walder, J. Electroanal. Chem. 2001, 498, 34. N. Leventis, J. H. Yang, E. F. Fabrizio, A. M. M. Rawashdeh, W. S. Oh, C. SotiriouLeventis, J. Am. Chem. Soc. 2004, 126, 4094. F. Marchioni, M. Venturi, P. Ceroni, V. Balzani, M. Belohradsky, A. M. Elizarov, H. R. Tseng, L. F. Stoddart, Chem. Eur. J. 2004, 10, 6361. T. D. Selby, S. C. Blackstock, J. Am. Chem. Soc. 1998, 120, 12155. K. Bronk, S. Thayumanavan, J. Org. Chem. 2003, 68, 5559. A. P. H.J. Schenning, R. E. Martin, M. Ito, F. Diederich, C. Boudon, J.-P. Gisselbrecht, M. Gross, Chem. Commun. 1998, 1013. C. B. Gorman, Adv. Mater. 1998, 10, 295. C. B. Gorman, J. C. Smith, Acc. Chem. Res. 2001, 34, 60.
116
DENDRITIC ENCAPSULATION OF REDOX-ACTIVE UNITS
65. G. R. Newkome, E. He, C. N. Moorefield, Chem. Rev. 1999, 99, 1689. 66. M. Venturi, S. Serroni, A. Juris, S. Campagna, V. Balzani, Top. Curr. Chem. 1998, 197, 194. 67. P. J. Dandliker, F. Diederich, M. Gross, C. B. Knobler, A. Louati, E. M. Sanford, Angew. Chem. Int. Ed. Engl. 1994, 33, 1739. 68. P. J. Dandliker, F. Diederich, J.-P. Gisselbrecht, A. Louati, M. Gross, Angew. Chem. Int. Ed. Engl. 1995, 34, 2725. 69. P. Weyermann, F. Diederich, Helv. Chim. Acta 2002, 85, 599. 70. P. Weyermann, F. Diederich, J. P. Gisselbrecht, C. Boudon, M. Gross, Helv. Chim. Acta 2002, 85, 571. 71. P. Weyermann, J. P. Gisselbrecht, C. Boudon, F. Diederich, M. Gross, Angew. Chem., Int. Ed. 1999, 38, 3215. 72. A. Zingg, B. Felber, V. Gramlich, L. Fu, J. P. Collman, F. Diederich, Helv. Chim. Acta 2002, 85, 333. 73. K. W. Pollak, J. W. Leon, J. M. J. Frechet, M. Maskus, H. D. Abrun˜a, Chem. Mater. 1998, 10, 30. 74. A. Adronov, S. L. Gilat, J. M. J. Frechet, K. Ohta, F. V. R. Neuwahl, G. R. Fleming, J. Am. Chem. Soc. 2000, 122, 1175. 75. M. S. Matos, J. Hofkens, W. Verheijen, F. C. De Schryver, S. Hecht, K. W. Pollak, J. M. J. Frechet, B. Forier, W. Dehaen, Macromolecules 2000, 33, 2967. 76. S. Hecht, H. Ihre, J. M. J. Frechet, J. Am. Chem. Soc. 1999, 121, 9239. 77. E. M. Harth, S. Hecht, B. Helms, E. E. Malmstrom, J. M. J. Frechet, C. J. Hawker, J. Am. Chem. Soc. 2002, 124, 3926. 78. S. Hecht, J. M. J. Frechet, Angew. Chem.-Int. Edit. 2001, 40, 74. 79. J.-F. Nierengarten, C. Schall, J.-F. Nicoud, Angew. Chem. Int. Ed. Engl. 1998, 37, 1934. 80. N. Armaroli, C. Boudon, D. Felder, J. P. Gisselbrecht, M. Gross, G. Marconi, J. F. Nicoud, J. F. Nierengarten, V. Vicinelli, Angew. Chem., Int. Ed. 1999, 38, 3730. 81. F. V€ogtle, M. Plevoets, M. Nieger, G. C. Azzellini, A. Credi, L. De Cola, V. De Marchis, M. Venturi, V. Balzani, J. Am. Chem. Soc. 1999, 121, 6290. 82. T. R. Krishna, N. Jayaraman, J. Chem. Soc. Perkin Trans. 2002, 1, 746. 83. D. L. Stone, D. K. Smith, P. T. McGrail, J. Am. Chem. Soc. 2002, 124, 856. 84. P. R. Ashton, V. Balzani, M. Clemente-Leon, B. Colonna, A. Credi, N. Jayaraman, F. M. Raymo, J. F. Stoddart, M. Venturi, Chem. Eur. J. 2002, 8, 673. 85. C. B. Gorman, B. L. Parkhurst, K.-Y. Chen, W. Y. Su, J. Am. Chem. Soc. 1997, 119, 1141. 86. S. Nlate, Y. Nieto, J. C. Blais, J. Ruiz, D. Astruc, Chem. Eur. J. 2002, 8, 171. 87. G. Denti, S. Campagna, S. Serroni, M. Ciano, V. Balzani, J. Am. Chem. Soc. 1992, 114, 2944. 88. S. Serroni, G. Denti, S. Campagna, A. Juris, M. Ciano, V. Balzani, Angew. Chem. Int. Ed. Engl. 1992, 31, 1493. 89. S. Campagna, G. Denti, S. Serroni, M. Ciano, A. Juris, V. Balzani, Inorg. Chem. 1992, 31, 2982. 90. S. Campagna, G. Denti, S. Serroni, A. Juris, M. Venturi, V. Ricevuto, V. Balzani, Chem. Eur. J. 1995, 1, 211.
REFERENCES
117
91. P. Ceroni, F. Paolucci, C. Paradisi, A. Juris, S. Roffia, S. Serroni, S. Campagna, A. J. Bard, J. Am. Chem. Soc. 1998, 120, 5480. 92. S. Serroni, S. Campagna, F. Puntoriero, C. Di Pietro, N. D. McClenaghan, F. Loiseau, Chem. Soc. Rev. 2001, 30, 367. 93. M. Marcaccio, F. Paolucci, C. Paradisi, S. Roffia, C. Fontanesi, L. J. Yellowlees, S. Serroni, S. Campagna, C. Denti, V. Balzani, J. Am. Chem. Soc. 1999, 121, 10081. 94. V. Balzani, S. Campagna, G. Denti, A. Juris, S. Serroni, M. Venturi, Sol. Energy Mater. Sol. Cells 1995, 38, 159. 95. V. Balzani, S. Campagna, G. Denti, A. Juris, S. Serroni, M. Venturi, Acc. Chem. Res. 1997, 31, 26. 96. G. R. Newkome, E. He, L. A. Godinez, Macromolecules 1998, 31, 4382. 97. G. R. Newkome, E. He, L. A. Godinez, G. R. Baker, J. Am. Chem. Soc. 2000, 122, 9993. 98. G. R. Newkome, R. G€uther, C. N. Moorefield, F. Cardullo, L. Echegoyen, E. PerezCordero, H. Luftmann, Angew. Chem. Int. Ed. Engl. 1995, 34, 2023. 99. B. Alonso, C. M. Casado, I. Cuadrado, M. Moran, A. E. Kaifer, Chem. Commun. 2002, 1778. 100. J. Guittard, J.-C. Blais, D. Astruc, C. Valerio, E. Alonso, J. Ruiz, J.-L. Fillaut, Pure Appl. Chem. 1998, 70, 809. 101. C. Valerio, E. Alonso, J. Ruiz, J.-C. Blais, D. Astruc, Angew. Chem. Int. Ed. Engl. 1999, 38, 1747. 102. M. C. Daniel, J. Ruiz, D. Astruc, J. Am. Chem. Soc. 2003, 125, 1150. 103. M. C. Daniel, J. Ruiz, S. Nlate, J. C. Blais, D. Astruc, J. Am. Chem. Soc. 2003, 125, 2617. 104. J. Ruiz, M. J. R. Medel, M. C. Daniel, J. C. Blais, D. Astruc, Chem. Commun. 2003, 464. 105. D. L. Stone, D. K. Smith, Polyhedron 2003, 22, 763. 106. H. F. Chow, I. Y. K. Chan, P. S. Fung, T. K. K. Mong, M. F. Nongrum, Tetrahedron 2001, 57, 1565. 107. H.-F. Chow, I. Y.-K. Chan, D. T. W. Chan, R. W. M. Kwok, Chem. Eur. J. 1996, 2, 1085. 108. M. Kimura, Y. Sugihara, T. Muto, K. Hanabusa, H. Shirai, N. Kobayashi, Chem. Eur. J. 1999, 5, 3495. 109. C. O. Turrin, J. Chiffre, D. de Montauzon, J. C. Daran, A. M. Caminade, E. Manoury, G. Balavoine, J. P. Majoral, Macromolecules 2000, 33, 7328. 110. C. O. Turrin, J. Chiffre, D. de Montauzon, G. Balavoine, E. Manoury, A. M. Caminade, J. P. Majoral, Organometallics 2002, 21, 1891. 111. G. R. Newkome, V. V. Narayanan, L. Echegoyen, E. Perez-Cordero, H. Luftmann, Macromolecules 1997, 30, 5187. 112. G. R. Newkome, V. V. Narayanan, L. A. Godinez, E. Perez-Cordero, L. Echegoyen, Macromolecules 1999, 32, 6782. 113. G. R. Newkome, V. V. Narayanan, L. A. Godınez, J. Org. Chem. 2000, 65, 1643. 114. C. B. Gorman, J. C. Smith, M. W. Hager, B. L. Parkhurst, H. Sierzputowska-Gracz, C. A. Haney, J. Am. Chem. Soc. 1999, 121, 9958. 115. J. I. Goldsmith, K. Takada, H. D. Abrun˜a, J. Phys. Chem. B 2002, 106, 8504. 116. Y.-R. Hong, C. B. Gorman, Langmuir 2006, 22, 10506. 117. C. M. Cardona, A. E. Kaifer, J. Am. Chem. Soc. 1998, 120, 4023. 118. Y. Wang, C. M. Cardona, A. E. Kaifer, J. Am. Chem. Soc. 1999, 121, 9756.
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. 148.
DENDRITIC ENCAPSULATION OF REDOX-ACTIVE UNITS
C. M. Cardona, T. D. McCarley, A. E. Kaifer, J. Org. Chem. 2000, 65, 1857. D. Sobransingh, A. E. Kaifer, Langmuir 2006, 22, 10540. W. Ong, A. E. Kaifer, Angew. Chem., Int. Ed. 2003, 42, 2164. C. Y. Chi, J. S. Wu, X. H. Wang, X. J. Zhao, J. Li, F. S. Wang, Macromolecules 2001, 34, 3812. K. Yamamoto, M. Higuchi, S. Shiki, M. Tsuruta, H. Chiba, Nature 2002, 415, 509. R. Nakajima, M. Tsuruta, M. Higuchi, K. Yamamoto, J. Am. Chem. Soc. 2004, 126, 1630. N. Satoh, J. S. Cho, M. Higuchi, K. Yamamoto, J. Am. Chem. Soc. 2003, 125, 8104. T. Imaoka, R. Tanaka, S. Arimoto, M. Sakai, M. Fujii, K. Yamamoto, J. Am. Chem. Soc. 2005, 127, 13896. S. M. Risser, D. N. Beratan, J. N. Onuchic, J. Phys. Chem. 1993, 97, 4523. T. S. Elicker, J. S. Binette, D. G. Evans, J. Phys. Chem. B 2001, 105, 370. T. S. Elicker, D. G. Evans, J. Phys. Chem. A 1999, 103, 9423. C. Kalyanaraman, D. G. Evans, Nano Lett. 2002, 2, 437. R. Toba, J. M. Quintela, C. Peinador, E. Roman, A. E. Kaifer, Chem. Commun. 2001, 857. P. Ceroni, V. Vincinelli, M. Maestri, V. Balzani, W. M. M€ uller, U. M€ uller, U. Hahn, F. Osswald, F. V€ogtle, New J. Chem. 2001, 25, 989. J. J. Apperloo, R. A. J. Janssen, P. R. L. Malenfant, L. Groenendaal, J. M. J. Frechet, J. Am. Chem. Soc. 2000, 122, 7042. W. Ong, A. E. Kaifer, J. Am. Chem. Soc. 2002, 124, 9358. R. Toba, J. M. Quintela, C. Peinador, E. Roman, A. E. Kaifer, Chem. Commun. 2002, 1768. M. Alvaro, B. Ferrer, H. Garcia, Chem. Phys. Lett. 2002, 351, 374. W. Ong, J. Grindstaff, D. Sobransingh, R. Toba, J. M. Quintela, C. Peinador, A. E. Kaifer, J. Am. Chem. Soc. 2005, 127, 3353. D. C. Tully, J. M. J. Frechet, Chem. Commun. 2001, 1229. D. J. Diaz, G. D. Storrier, S. Bernhard, K. Takada, H. D. Abrun˜a, Langmuir 1999, 15, 7351. L. Latterini, G. Pourtois, C. Moucheron, R. Lazzaroni, J. L. Bredas, A. Kirsch-De Mesmaeker, F. C. De Schryver, Chem. Eur. J. 2000, 6, 1331. D. J. Diaz, S. Bernhard, G. D. Storrier, H. D. Abrun˜a, J. Phys. Chem. B 2001, 105, 8746. C. J. Xia, X. W. Fan, J. Locklin, R. C. Advincula, A. Gies, W. Nonidez, J. Am. Chem. Soc. 2004, 126, 8735. C. A. Nijhuis, J. Huskens, D. N. Reinhoudt, J. Am. Chem. Soc. 2004, 126, 12266. X. Y. Ling, D. N. Reinhoudt, J. Huskens, Chem. Mater. 2008, 20, 3574. C. A. Nijhuis, J. K. Sinha, G. Wittstock, J. Huskens, B. J. Ravoo, D. N. Reinhoudt, Langmuir 2006, 22, 9770. G. Cooke, J. Couet, J. F. Garety, C. Q. Ma, S. Mabruk, G. Rabani, V. M. Rotello, V. Sindelar, P. Woisel, Tetrahedron Lett. 2006, 47, 3763. C. M. Casado, I. Cuadrado, M. Moran, B. Alonso, B. Garcia, B. Gonzalez, J. Losada, Coord. Chem. Rev. 1999, 185–186, 53. L. Svobodova, M. Snejdarkova, K. Toth, R. E. Gyurcsanyi, T. Hianik, Bioelectrochemistry 2004, 63, 285.
REFERENCES
119
149. M. P. G. Armada, J. Losada, M. Zamora, B. Alonso, I. Cuadrado, C. M. Casado, Bioelectrochemistry 2006, 69, 65. 150. F. N. Crespilho, V. Zucolotto, C. M. A. Brett, O. N. Oliveira, F. C. Nart, J. Phys. Chem. B 2006, 110, 17478. 151. S. J. Kwon, E. Kim, H. Yang, J. Kwak, Analyst 2006, 131, 402. 152. M. Wells, R. M. Crooks, J. Am. Chem. Soc. 1996, 118, 3988. 153. Y. Liu, M. Zhao, D. E. Bergbreiter, R. M. Crooks, J. Am. Chem. Soc. 1997, 119, 8720. 154. H. Tokuhisa, R. M. Crooks, Langmuir 1997, 13, 5608. 155. M. Zhao, H. Tokuhisa, R. M. Crooks, Angew. Chem. Int. Ed. Engl. 1997, 36, 2596. 156. H. Tokuhisa, M. Zhao, L. A. Baker, V. T. Phan, D. L. Dermody, M. E. Garcia, R. F. Peez, R. M. Crooks, T. M. Mayer, J. Am. Chem. Soc. 1998, 120, 4492. 157. S. K. Oh, L. A. Baker, R. M. Crooks, Langmuir 2002, 18, 6981. 158. T. D. Selby, K. Y. Kim, S. C. Blackstock, Chem. Mater. 2002, 14, 1685. 159. H. Tokuhisa, M. Q. Zhao, L. A. Baker, V. T. Phan, D. L. Dermody, M. E. Garcia, R. F. Peez, R. M. Crooks, T. M. Mayer, J. Am. Chem. Soc. 1998, 120, 4492. 160. M. Q. Zhao, H. Tokuhisa, R. M. Crooks, Angew. Chem., Int. Ed. 1997, 36, 2596. 161. C. B. Gorman, R. L. Miller, K. Y. Chen, A. R. Bishop, R. T. Haasch, R. G. Nuzzo, Langmuir 1998, 14, 3312. 162. M. Halim, J. N. G. Pillow, I. D. W. Samuel, P. L. Burn, Adv. Mater. 1999, 11, 371. 163. J. N. G. Pillow, M. Halim, J. M. Lupton, P. L. Burn, I. D. W. Samuel, Macromolecules 1999, 32, 5985. 164. Y. Sakamoto, T. Suzuki, A. Miura, H. Fujikawa, S. Tokito, Y. Taga, J. Am. Chem. Soc. 2000, 122, 1832. 165. A. W. Freeman, S. C. Koene, P. R. L. Malenfant, M. E. Thompson, J. M. J. Frechet, J. Am. Chem. Soc. 2000, 122, 12385. 166. J. M. Lupton, I. D. W. Samuel, M. J. Frampton, R. Beavington, P. L. Burn, Adv. Funct. Mater. 2001, 11, 287. 167. J. P. J. Markham, S. C. Lo, S. W. Magennis, P. L. Burn, I. D. W. Samuel, Appl. Phys. Lett. 2002, 80, 2645. 168. H. Peng, L. Cheng, J. D. Luo, K. T. Xu, Q. H. Sun, Y. P. Dong, F. Salhi, P. P. S. Lee, J. W. Chen, B. Z. Tang, Macromolecules 2002, 35, 5349. 169. D. G. Ma, J. M. Lupton, I. D. W. Samuel, S. C. Lo, P. L. Burn, Appl. Phys. Lett. 2002, 81, 2285. 170. R. Beavington, M. J. Frampton, J. M. Lupton, P. L. Burn, I. D. W. Samuel, Adv. Funct. Mater. 2003, 13, 211. 171. Y. J. Pu, R. E. Harding, S. G. Stevenson, E. B. Namdas, C. Tedeschi, J. P. J. Markham, R. J. Rummings, P. L. Burn, I. D. W. Samuel, J. Mater. Chem. 2007, 17, 4255. 172. M. A. Saab, R. Abdel-Malak, J. F. Wishart, T. H. Ghaddar, Langmuir 2007, 23, 10807. 173. N. Satoh, T. Nakashima, K. Yamamoto, J. Am. Chem. Soc. 2005, 127, 13030. 174. D. N. Lee, J. K. Kim, H. S. Park, Y. M. Jun, R. Y. Hwang, W. Y. Lee, B. H. Kim, Synth. Met. 2005, 150, 93. 175. J. M. Lupton, I. D. W. Samuel, R. Beavington, M. J. Frampton, P. L. Burn, H. B€assler, Phys. Rev. B 2001, 63, 155206.
120
DENDRITIC ENCAPSULATION OF REDOX-ACTIVE UNITS
176. I. Cuadrado, M. Moran, C. M. Casado, B. Alonso, F. Lobete, B. Garcıa, M. Ibisate, J. Losada, Organometallics 1996, 15, 5278. 177. J.-L. Fillaut, J. Linares, D. Astruc, Angew. Chem. Int. Ed. Engl. 1994, 33, 2460. 178. S. Nlate, J. Ruiz, V. Sartor, R. Navarro, J. C. Blais, D. Astruc, Chem. Eur. J. 2000, 6, 2544. 179. R. Castro, I. Cuadrado, B. Alonso, C. M. Casado, M. Moran, A. E. Kaifer, J. Am. Chem. Soc. 1997, 119, 5760. 180. K. Takada, G. D. Storrier, M. Moran, H. D. Abruna, Langmuir 1999, 15, 7333. 181. I. Cuadrado, C. M. Casado, B. Alonso, M. Moran, J. Losada, V. Belsky, J. Am. Chem. Soc. 1997, 119, 7613. 182. B. Alonso, M. Moran, C. M. Casado, F. Lobete, J. Losada, I. Cuadrado, Chem. Mater. 1995, 7, 1440. 183. J. Losada, I. Cuadrado, M. Moran, C. M. Casado, B. Alonso, M. Barranco, Anal. Chim. Acta 1997, 338, 191. 184. F. Le Derf, E. Levillain, G. Trippe, A. Gorgues, M. Salle, R. M. Sebastian, A. M. Caminade, J. P. Majoral, Angew. Chem., Int. Ed. 2001, 40, 224. 185. F. E. Appoh, Y. T. Long, H. B. Kraatz, Langmuir 2006, 22, 10515. 186. C. B. Gorman, J. C. Smith, J. Am. Chem. Soc. 2000, 122, 9342. 187. T. L. Chasse, C. B. Gorman, Langmuir 2004, 20, 8792. 188. T. L. Chasse, J. C. Smith, R. L. Carroll, C. B. Gorman, Langmuir 2004, 20, 3501. 189. Y.-R. Hong, C. B. Gorman, Chem. Commun. 2007, 3195. 190. D. Boris, M. Rubinstein, Macromolecules 1996, 29, 7251. 191. P. Welch, M. Muthukumar, Macromolecules 1998, 31, 5892. 192. E. G. Timoshenko, Y. A. Kuznetsov, R. Connolly, J. Chem. Phys. 2002, 117, 9050. 193. T. L. Chasse, R. Sachdeva, Q. Li, Z. Li, R. J. Petrie, C. B. Gorman, J. Am. Chem. Soc. 2003, 125, 8250.
CHAPTER 5
Redox-Active Metal–Polypyridine Dendrimers as Light-Harvesting Antennae FAUSTO PUNTORIERO, SCOLASTICA SERRONI, FRANCESCO NASTASI, and SEBASTIANO CAMPAGNA Dipartimento di Chimica Inorganica, Chimica Analitica e Chimica Fisica, Universita di Messina, Messina, Italy
5.1
INTRODUCTION
Artificial light-harvesting antenna systems are supramolecular, multicomponent species capable of absorbing light energy and transferring the electronic energy collected toward a single subunit by means of a series of cascade energy transfer steps (antenna effect).1 In such a way, one of the key features of the natural photosynthetic process2 is answered. Design of artificial antenna systems is therefore an extensively investigated and growing research field, since it is quite related to the design of efficient systems capable of performing the photochemical conversion of solar energy into fuels, that is, artificial photosynthesis, one of the Holy Grails of modern science.3 Dendrimers made of suitable chromophoric units have been designed to play the role of artificial antennae by taking advantage of the topographic organization that is inherent to the dendritic structure.1b,4 Many of the light-harvesting antenna dendrimers prepared are based on polypyridine metal complexes as building blocks.5 For these species, of particular relevance is the metal-to-ligand charge transfer (MLCT) nature of the excited state(s) of the subunits, since both the absorption bands responsible for visible absorption and the lowest energy (emitting) excited states, which are also the excited states involved in the antenna effect, are of MLCT nature. The MLCT transitions (and excited states) are connected to the oxidation and reduction potentials of the metal complex,6 so a strict relationship exists between the various building blocks employed, with their relevant and peculiar redox properties, Electrochemistry of Functional Supramolecular Systems. Edited by Paola Ceroni, Alberto Credi, and Margherita Venturi Copyright 2010 John Wiley & Sons, Inc.
121
122
REDOX-ACTIVE METAL–POLYPYRIDINE DENDRIMERS
and their spatial organization in the supramolecular dendritic array, which governs the antenna properties of the final systems. Here, we review the basic features of the metal–polypyridine building blocks, their redox properties, and the way such properties are connected with the ideal topography of dendritic assembly that determines the desired antenna effect.
5.2
AN ARTIFICIAL ANTENNA: BASIC REQUIREMENTS
If an artificial system has to be developed, the obvious choice is probably to think of it as a modular system, as well as Nature did. This leads us to focus directly on the individual modules to be used (the building blocks). The requirements that the building blocks have to fulfill are the following: (i) stability in the ground and excited states, (ii) capability to significantly absorb solar light, and (iii) relatively long-lived excited state to allow the occurrence of energy transfer with minimal loss of energy. After the building blocks have been selected, the next step is to connect them together. This is not a trivial point in that a suitable spatial connection is crucial for fast energy transfer. Of course, the connections should be selected for assuring the necessary electronic coupling (through space or through bonds) between the building bocks. Once building blocks and connections have been selected, the next step is set up synthetic procedures that allow to efficiently build up the larger numbers of subunits together, suitably connected to one another, in a few synthetic steps. The dendritic structure appears quite suitable to this scope.1b Actually, for the construction of such branched structures, two main different synthetic approaches are used,7 as briefly explained in Figs. 5.1 and 5.2. In the divergent method (Fig. 5.1), buildup of the dendrimer starts at the core and proceeds toward the periphery. At each growing step a new layer of branching units is added, thus increasing by one the generation number. This approach is conceptually simple but requires very efficient reactions to avoid flaws in the dendrimer structure, as on increasing generation there are an exponentially increasing number of reactions that must take place simultaneously. Even if the presence of some structural defects is difficult to avoid in high-generation species, the divergent method is nowadays the most common way to prepare high-generation dendrimers. In the convergent method (Fig. 5.2), on the contrary, the synthesis starts at the periphery and proceeds toward the core. A small dendron is initially formed by connecting two peripheral groups to a branching unit. Then, two of these dendrons can again be connected to a branching unit to generate a higher generation dendron, and this coupling process can be repeated doubling each time the dendron size. Finally, a symmetrical dendrimer can be formed by connecting these dendrons to a polyfunctional core through their focal point. In the convergent approach, the number of reacting partners does not change on increasing generation, so that structural defects can be more easily avoided. However, the preparation of high-generation dendrons is hampered for steric reasons because coupling must occur at the focal point. The strategy used for the synthesis is
5.2
Figure 5.1
Figure 5.2
AN ARTIFICIAL ANTENNA: BASIC REQUIREMENTS
123
Schematization of the divergent synthetic approach to buildup dendrimers.
Schematization of the convergent synthetic approach to buildup dendrimers.
124
REDOX-ACTIVE METAL–POLYPYRIDINE DENDRIMERS
very important in organizing the redox-active building blocks on the supramolecular light-harvesting array in the correct way. It should be mentioned, however, that the three basic steps toward the preparation of artificial antennae, that is, (1) selection of building blocks, (2) connections between selected building blocks, and (3) synthetic strategies, are also strictly correlated to each other. In other words, the right connections (and synthetic procedure) have to be designed for the given building blocks.
5.3 REDOX PROPERTIES OF Ru(II) AND Os(II) POLYPYRIDINE COMPLEXES Probably the larger class of light-harvesting dendrimers investigated up to now is constituted of dendrimers based on Ru(II) and Os(II) polypyridine complexes, and the more extensive studies within these species involve the systems containing 2,3-(bis-pyridyl)pyrazine (2,3-dpp) as bridging ligands.5,8 The synthetic approach is based on protection/deprotection iterative steps of chelating ligands, developed within the frame of the so-called complexes as ligands/complexes as metals strategy.8 This chapter is therefore focused on the redox properties of dendrimers based on Ru(II) and Os(II) metal centers containing 2,3-dpp as bridging ligands. In some cases, also dendrimers containing the isomeric containing 2,5-(bis-pyridyl)pyrazine (2,5-dpp) as bridging ligand will be discussed. The redox properties of Ru(II) and Os(II) polypyridine species are well known and several reviews are available:9,10 here, we summarize the main properties of these species, starting from the mononuclear building blocks. In general, oxidation processes are reversible and metal centered, while reduction processes are reversible and ligand centered. Successive ligand reductions take place, even for homoleptic mononuclear compounds, and in this case reduction splitting is connected with ligand–ligand interaction as mediated by the metal-centered orbitals. In general, ligand–ligand interaction is larger for Os mononuclear species compared to homologous Ru species, as the metal-centered HOMO is closer in energy to the ligand-based LUMOs in the case of Os compounds. Each polypyridine ligand is usually reduced by one electron at mild potentials (typically, at potentials less negative than 2.00 V versus SCE). However, a second electron can be added to a single polypyridine ligand at more negative potentials: electron pairing takes place within 350–600 mV (e.g., is 600 mV in the prototype [Ru(bpy)3]2 þ ), depending on the size and delocalization of the polypyridine ligand.11 A direct consequence is that when first reduction of a polypyridine ligand occurs at less negative potential, a second electron can be added to the same ligand even before first reduction of the other ligands occurs. This is actually found for the multinuclear dendrimers containing 2,3-dpp as bridging ligand, since first reduction of the bridge occurs at about 0.5 V.12 Metal centers are usually mono-oxidized at mild potentials (e.g., at potentials less positive than þ 2.00 V versus SCE). However, at more positive potentials a successive oxidation, leading to a formally MIV metal center, can also occur. This has been particularly investigated by using liquid SO2 as solvent, as this special medium allows
5.4
Ru(II) AND Os(II) POLYPYRIDINE DENDRIMERS
125
to use an extended potential window.13–15 Generally, a second metal oxidation occurs at potentials more positive by at least 1.60 V compared to the first metal oxidation. Obviously, such second metal oxidation processes are easier to be detected for Os species. For dinuclear and larger nuclearity species, as dendrimers are, multiple redox processes become usual and the situation become more and more complex and interesting; for example, metal–metal interaction, mediated by bridging ligand, can lead to oxidation splitting for equivalent metal centers if enough interaction occurs. This leads us into the realm of mixed-valence chemistry.16 Moreover, since different positions in the dendritic array (e.g., terminal or inner sites) translate into different chemical environments, oxidation of metal centers can occur at different potentials. The same is also valid for the ligand-based reduction processes, so that the redox behavior of metal–polypyridine dendrimers is extremely rich. In general, the study of the redox properties of these species shows two aspects: (i) the redox-active nature of the various building blocks confers to the dendritic array quite interesting redox properties and (ii) detailed redox investigation allows to obtain useful pieces of information concerning the electronic coupling between the various subunits. In the following, these aspects will be discussed in more detail and some representative examples will be presented.
5.4 Ru(II) AND Os(II) POLYPYRIDINE DENDRIMERS: GENERAL REDOX PROPERTIES AND SELECTED EXAMPLES Knowledge of the electrochemical properties of metal-based dendrimers is of paramount importance in fully understanding the dendrimer properties and in foreseeing possible applications. For example, the study of successive metal-centered oxidations is important (i) for understanding the degree of metal–metal interaction within the dendritic array, (ii) for knowing which metal ion is easier to oxidize and therefore has the highest occupied d orbital, and (iii) for predicting the direction and the rate of a possible electron transfer. Furthermore, electrochemical data are useful for the photophysical characterization as they facilitate MLCT transitions on the basis of the known relation between spectroscopy and electrochemistry.9 In the dendritic species, each unit brings its own redox properties into the supramolecular dendritic array, more or less affected by intercomponent interactions. Metal–metal and ligand–ligand interactions are noticeable for metals coordinated to the same bridging ligand and for ligands coordinated to the same metal, whereas they are negligibly small for metals or ligands that are sufficiently far apart. It is known17 that supramolecular species containing a number of identical noninteracting centers exhibit current–potential responses having the same shape as that obtained with the corresponding molecule containing a single center. Only the magnitude of the current is enhanced by the presence of additional electroactive centers. Therefore, in this kind of dendrimers, equivalent and noninteracting units undergo electrochemical processes at the same potential. This allows us to control the number of electrons lost or gained at a certain potential by placing in the dendrimer the desired number of
126
REDOX-ACTIVE METAL–POLYPYRIDINE DENDRIMERS
suitable, equivalent, and noninteracting units. Many luminescent and redox-active Ru(II) and Os(II) dendrimers in which 2,3-bis(2-pyridyl)pyrazine (2,3-dpp) acts as the bridging ligand to assemble the various metal subunits have been synthesized and studied, with the larger systems containing as many as 22 metal subunits.18,19 As for any type of dendrimer, the possibility to prepare large systems having the desired properties resides in the availability of specific redox-active building blocks.19 Ru(II)–polypyridine complexes show an oxidation process metal centered and lead to Ru(II) compounds that are inert to ligand substitution. Only this first metal-centered oxidation process can be observed in the potential window available in the usual solvents (e.g., acetonitrile), but in SO2 solution at 70 C other processes involving ligand oxidation can be observed.15 It can be stated, however, that the Ru(III)/Ru(II) reduction potential in most complexes that contain only polypyridine-type ligands falls in a rather narrow range between þ 1.25 and þ 1.55 V (versus SCE, acetonitrile solution).12 Substitution of one bpy ligand by two Cl ions to give [Ru(bpy)2Cl2] lowers the potential to þ 0.32 V, whereas the strong p-acceptor CO causes an increase in the reduction potential above þ 1.9 V.9a 5.4.1
The Effective Models: Selected Dinuclear Complexes
Although the building blocks of the metal–polypyridine dendrimers are mononuclear species, the effective models for the high-nuclearity dendrimers are the dinuclear species. This is because the properties of the mononuclear and dinuclear compounds (absorption, luminescence, and redox properties) are significantly different, as a consequence of the bis-chelation of the dpp ligand.20–22 In the high-nuclearity dendrimers, dpp always plays the role of bridge, so the redox properties (and indeed also the spectroscopic properties) of the dendrimers are directly connected to the properties of the dinuclear species. Representative dinuclear species are discussed here. Very interesting and detailed oxidation patterns have been obtained for the dinuclear species [(bpy)2Ru(2,3-dpp)Ru(bpy)2]4 þ (Ru2) and [(bpy)2Ru(2,5-dpp) Ru(bpy)2]4 þ (Ru2a)21,22 (see Fig. 5.3 and Table 5.1), which contain a relatively small number of redox centers and can be considered as the most effective model compounds to understand the electrochemistry of the complexes of higher nuclearity. In the dinuclear species, the knowledge of the voltammetric behavior of the mononuclear building blocks allowed to establish the localization for each redox process and to evaluate the mutual interactions existing between the redox sites. The oxidation pattern obtained in acetonitrile showed two monoelectronic waves attributed at independent oxidations on the two metal ions. The oxidation splitting is due to electronic coupling across the bridging ligand. At this regard, it is useful to compare the oxidation splitting of the two Ru species21 to that of the corresponding Os dinuclear compounds [(bpy)2Os(2,3-dpp)Os(bpy)2]4 þ (Os2) and [(bpy)2Os(2,5-dpp)Os(bpy)2]4 þ (Os2a)23 (Fig. 5.3). Oxidation splitting is larger in the case of the osmium species, and this shows that the dominant pathway for metal–metal superexchangeassisted electronic coupling via the dpp bridging ligands is the electron transfer pathway.24
5.4
Ru(II) AND Os(II) POLYPYRIDINE DENDRIMERS
127
Figure 5.3 Schematic representation of the light-harvesting redox-active dendrimers and some models and/or building blocks. Notation is provided in the inset.
128
REDOX-ACTIVE METAL–POLYPYRIDINE DENDRIMERS
TABLE 5.1
Redox Data, in Deaerated CH3CN Unless Otherwise Stated
Compound Ru2
Ru2a
Os2
Oxidation
Reduction
a
þ 1.38; þ 1.55 þ 1.37; þ 1.60; þ 3.10 [2]; þ 3.30 [2]
b
0.67; 1.17; 1.57 [2]; 1.89 [2] 0.78; 1.17; 1.46; 1.55; 1.70; 1.82; 2.40; 2.51; 2.66; 2.83
a
þ 1.37; þ 1.54 þ 1.37; þ 1.60; þ 3.16 [2]; þ 3.40 [2]
b
0.53; 1.08; 1.50 [2]; 1.81 [2] 0.66; 1.11; 1.50; 1.55; 1.74; 1.82; 2.43; 2.70; 2.51; 2.81
a
þ 0.90; þ 1.20 þ 0.97; þ 1.34; þ 2.73; þ 2.88; þ 4.10
0.68; 1.10; 1.38; 1.62
Os2a
þ 0.92; þ 1.22
0.56; 1.00; 1.46; 1.71
Ru3 a
þ 1.48 [2]; þ 1.41 [2]; þ 1.89
b
0.55; 0.75; 1.17; 1.47 [2]; 1.75 [2] 0.72; 0.81; 1.15; 1.26; 1.49; 1.55; 1.67; 1.76; 1.85; 2.37; 2.46; 2.55; 2.69; 2.78
a
þ 1.45 [2]; þ 1.41 [2]; þ 1.89
b
Ru3a
0.48; 0.60; 1.10; 1.30; 1.52 0.61; 0.71; 1.12; 1.24; 1.50; 1.56; 1.74; 1.80; 1.87; 2.44; 2.49; 2.57; 2.72; 2.79
Cl2RuRu2
þ 0.82; þ 1.57 [2]
0.72; 0.88
Cl2RuOs2
þ 0.74; þ 1.16 [2]
0.72; 0.87
Cl2OsRu2
þ 0.57; þ 1.55 [2]
0.71; 0.96
Cl2OsOs2
þ 0.51; þ 1.16 [2]
0.69; 0.91
Ru4
þ 1.50 [3]
0.56; 0.63; 0.70; 1.20; 1.33; 1.48
OsRu3
þ 1.25; þ 1.55 [3]
0.55; 0.65; 0.77
þ 1.44 [4]
Ru6 a
Cl2RuRu6 Ru10
OsRu9 OsRu3Os6 Ru22 OsRu21
a
þ 1.46 [4]; þ 2.11 [2]; þ 3.18 [4]; þ 3.40 [4]; þ 4.00
0.55 (c) b
0.65; 0.70; 0.76; 0.81; 1.07; 1.14; 1.20; 1.27; 1.32; 1.40; 1.44; 1.48; 1.52; 1.57; 1.67; 1.71; 1.76; 1.81; 2.42; 2.46; 2.50; 2.54; 2.68; 2.73; 2.77; 2.82
þ 0.86 [1]; þ 1.56 [4]
0.51 [2]; 0.62 [2]; 0.76 [2]; 1.16 [2]
þ 1.43 [6] þ 1.46 [6]; þ 2.11 þ 2.44 [3]; þ 3.2 [6]; þ 3.50
0.73 [6]; 1.22 [3]
þ 1.17; þ 1.50 [6] þ 1.05 [6]; þ 1.39 þ 1.54 [12] þ 1.42; þ 1.54 [12]
E1/2 (V) values versus SCE are reported; all the processes are monoelectronic unless otherwise stated (the number in brackets). (a) In SO2 solution at 70 C; (b) in DMF solution at 54 C; (c) overlapping waves.
5.4
Figure 5.4
Ru(II) AND Os(II) POLYPYRIDINE DENDRIMERS
129
Cyclic voltammogram of Os2 in liquid SO2 at 70 C.
Interestingly, metal–metal interaction appears to be dependent on the oxidation state of the system. This is clearly evidenced by the results obtained for Os2 in liquid SO2 at 70 C, in that in these experimental conditions both the first and the second oxidation of each metal center can be seen (Fig. 5.4; Table 5.1).15 Potential splitting for the first oxidation process is 370 mV, whereas potential splitting for the second oxidation process is 150 mV. This result confirms the electron transfer pathway for the superexchange interaction: once mono-oxidized, the Os(III)-based orbitals are stabilized and their interaction with the bridge-based LUMO orbitals decreases. A similar result has been recently obtained in molecular grids based on similar polypyridine ligands.25
Figure 5.5 Cyclic voltammogram of Ru2 in DMF at 54 C and the schematic localization of the redox processes.
130
REDOX-ACTIVE METAL–POLYPYRIDINE DENDRIMERS
The reduction pattern of the dinuclear species is also quite rich. The cyclic voltammetric curves recorded at a sweep rate of 0.1 V/s of the dinuclear species Ru2 in highly pure DMF solution at low temperature is shown in Fig. 5.5.26 As it can be seen, an impressive number of reversible processes can be seen. In the figure, detailed attributions are also shown. The comparison of the redox series relative to the two dinuclear compounds (Table 5.1) shows that the first two one-electron voltammetric waves (peaks I and II), corresponding to sequential two-electron reduction of the bridging ligands, as also revealed by redox investigation in acetonitrile at room temperature,21 move toward less negative potentials on passing from Ru2 to Ru2a. At more negative potentials, two series of two close one-electron processes take place, each one involving the reduction of one peripheral bpy ligand of each metal center. The potential separation between each couple of processes, related to the electronic coupling between peripheral ligands, increases forsecond series of process. This is alsoevidenced by differential pulsevoltammetry experiments for Os2 and Os2a, particularly for the Os compounds.23 These findings suggest that the ligand–ligand electronic interaction in this class of multinuclear species increases on passing to the reduced states. Different sequences of reduction processes have also been obtained, when the peripheral bpy ligands are replaced by ligands such as 2,20 -biquinoline (biq), whose reduction potential is intermediate between the first and second reduction potentials of the bridging dpp ligands. Details can be found in the original reference 21. 5.4.2
Trinuclear Species: The Smallest Dendrons
The results obtained by the investigation of the trinuclear compounds allowed to gain further knowledge on the electronic interactions among the various components of the multicomponent-structured dendrimers: indeed, trinuclear compounds are the smallest species in which more than one bridging ligand is present, so that interaction between several bridging ligands can be evidenced. The prototype trinuclear compounds of this series are [(bpy)Ru{(2,3-dpp)Ru(bpy)2}2]6 þ (Ru3) and [(bpy)Ru{(2,5-dpp)Ru(bpy)2}2]6 þ (Ru3a) (bpy (2,2-bipyridine)) (Fig. 5.3).21,27 In these species, three metal centers are present, two of them peripheral (and equivalent to one another) and the third one of “inner” type, connected to two bridging units. Because of the different electron-withdrawing properties of the bpy and bridged dpp ligands, the peripheral metal centers are expected to be oxidized at less positive potentials. Actually, the oxidation pattern of the two trinuclear prototypes in acetonitrile is practically identical, showing a single, bielectronic process, related to the simultaneous one-electron oxidation of the two peripheral metal centers, at about þ 1.45 V versus SCE. The inner metal center is not oxidized within the potential window available in acetonitrile.21 Indeed, successive investigation of Ru3a in liquid SO2 at 70 C showed that the inner metal center is oxidized at þ 1.89 V in these latter conditions, confirming the previous results (note that in these conditions the first oxidation process, found at þ 1.45 V in acetonitrile room temperature, occurs at þ 1.41 V).15 On reduction, a series of monoelectronic processes are evidenced. The clearest result is given by Ru3a: for this species, two couples of close one-electron reduction
5.4
Ru(II) AND Os(II) POLYPYRIDINE DENDRIMERS
131
processes take place, the first couple occurring at 0.48 and at 0.60 V, assigned to first reduction of the two 2,5-dpp ligands, and the second couple, occurring at 1.10 and 1.30 V, related to the second reduction of the same sites.21 Electronic interaction between the 2,5-dpp bridges therefore translates into a 120 mV potential splitting, while electron pairing energy is of the order of 500 mV, assuming Koopmans’ theorem as valid. Qualitative similar results are obtained for Ru3. The reduction behavior of Ru3 and Ru3a has also been investigated in highly pured DMF solution at 54 C. In these conditions, up to 14 reduction processes have been identified.26 Trinuclear dendron-like species such as [Cl2Ru{(2,3-dpp)Ru(bpy)2}2]4 þ , Cl2RuRu228,29, and [Cl2Ru{(2,3-dpp)Os(bpy)2}2]4 þ , Cl2RuOs228,29, have also been prepared, essentially to build up most of the larger species of the series, through convergent or semiconvergent routes, for decanuclear dendrimers (see later).28,30 Comparisons among the potentials for oxidation processes of the various subunits in the series of the trinuclear Cl2MM2 dendrons (Table 5.1 and Fig. 5.6)29 evidence some intriguing features. For example, the value of the oxidation of {Cl2Os(2,3-dpp)2} subunit (i.e., the central osmium subunit) in complex Cl2OsOs2 is slightly less positive (60 mV) than that of the same subunit in the mixed-metal species Cl2OsRu2. This can be rationalized by taking into account that the peripheral subunits that are present in Cl2OsOs2, that is, {(2,3-dpp)Os(bpy)2}, have a better electron donating ability than those in Cl2OsRu2, that is, {(2,3-dpp)Ru(bpy)2}, because back donation (or p-donation) from Os(II) to the bridging ligand is more effective than from Ru(II). The same argument explains why the potential of the oxidation of the {Cl2Ru(2,3-dpp)2} subunit (i.e., the central ruthenium subunit) in Cl2RuOs2 is more positive (80 mV) than that of the same unit in Cl2RuRu2. It can be noted that the peripheral subunit oxidation in Ru3 occurs at a potential slightly less positive than that of the corresponding oxidation in analogous compounds of the Cl2MM2 series. It should be considered, however, that such oxidation processes in the latter compounds take place in the presence of an already oxidized central metal. 5.4.3 The First Effective Light-Harvesting Antenna: The Mixed Metal Os–Ru Tetranuclear Dendrimer As the energy of the excited states and the redox levels of each metal–polypyridine unit depend on metal and ligands in a predictable way, the simultaneous presence of different metals in a dendritic structures gives rise to intramolecular energy transfer processes as well to different redox patterns with multielectron processes. In particular, the tetranuclear [Os(2,3-dpp)3{(2,3-dpp)Ru(bpy)2}3]8 þ (OsRu3) shown in Fig. 5.3 has been designed to achieve an efficient antenna effect. This species can also be considered a first-generation mixed-metal dendrimers.31 On electrochemical oxidation, OsRu3 shows a monoelectronic, reversible wave at þ 1.25 V, followed by a trielectronic, almost reversible wave at þ 1.55 V. On the basis of the oxidation potentials in parent compounds (Table 5.1), the monoelectronic wave is referred to oxidation of the central Os(II) ion and the trielectronic wave to
132
REDOX-ACTIVE METAL–POLYPYRIDINE DENDRIMERS
Figure 5.6 Differential pulse voltammograms of Cl2RuRu2 (top), Cl2OsRu2, Cl2RuRu2, and Cl2OsOs2 (bottom).
oxidation of the three peripheral Ru(II) ions. Spectroelectrochemical experiments confirmed that the oxidation of OsRu3 involves two successive steps, leading to the disappearance of MLCT bands involving Os and Ru. On electrochemical reduction, a series of three waves are present in the cyclovoltammogram of OsRu3 (Table 5.1). The dpp ligand is easier to reduce than bpy, and it is even more so when it plays the role of a bridging ligand (as it was inferred from the redox studies of the dinuclear species). Therefore, the three reduction waves can be associated to the first one-electron reduction of the three dpp ligands. For potentials
5.4
Ru(II) AND Os(II) POLYPYRIDINE DENDRIMERS
133
Figure 5.7 Cyclic voltammogram of Ru4 in ACN. Fc indicate ferrocene, used as reference.
more negative than 1.35 V, several overlapping waves are present, corresponding to the second one-electron reduction of the dpp ligands and to the first reduction of peripheral bpy. The homologous compound [Ru(2,3-dpp)3{(2,3-dpp)Ru(bpy)2}3]8 þ (Ru4), Fig. 5.3, exhibits a qualitatively similar reduction pattern, but clear separation between the various reduction processes allowed to identify two couples of three sequential dpp reductions, the first and the second couple corresponding to the first and the second reduction of the bridging ligands, respectively.21 Electron pairing energy is confirmed to be around 500 mV as for the Ru3 species, whereas electronic coupling between the 2,3-dpp bridges results smaller for the first reduction process (about 70 mV) and larger for the second reduction process (about 120 mV) (see potential separation in Table 5.1 and Fig. 5.7). A larger reduction pattern for Ru4 has been investigated in highly pure DMF solution at 54 C. In these conditions, up to 18 reduction processes have been identified.26 As predicted, and indirectly confirmed by the redox behavior, the presence of two different metals in OsRu3 creates an energy gradient from the periphery to the Os(II) core.31 Indeed, independently of the excitation wavelength, luminescence is observed (875 nm at room temperature and 860 nm at 77 K) from a MLCT state involving Os(II) and the bridging ligand, whereas the Ru-to-dpp CT luminescence, occurring in Ru4 at 780 nm at room temperature and at 725 nm at 77 K, is totally absent, so indicating that a periphery-to-center energy transfer is effective.31 It is interesting to note that the excited-state energy of the central Os-containing core should not be substantially affected on replacing the peripheral Ru(bpy)22 þ subunits with Ru(2,3-dpp)22 þ . This has opened the way to higher nuclearity, luminescent dendrimers. 5.4.4
Second Generation: The Decanuclear Species
By taking advantage of the labile chloride ligands of the central (“inner”) metal, the above-described trinuclear dendrons allowed to obtain higher nuclearity systems, such as the prototype decanuclear [Ru{(2,3-dpp)Ru[(2,3-dpp)Ru(bpy)2]2}3]20 þ
134
REDOX-ACTIVE METAL–POLYPYRIDINE DENDRIMERS
compound (Ru10) (Fig. 5.3),28,32 in which the peripheral Ru-based units are expected to be oxidized at less positive potentials than the internal ones because the bpy ligands are better electron donors than the 2,3-dpp bridging ligands. Furthermore, the six peripheral Ru-based units are not expected to interact with one another because they are not directly connected. According to this expectation, the first oxidation process observed for this dendrimer ( þ 1.53 V) involves the exchange of six electrons at the same potential (Fig. 5.8). Successive oxidation of the other metal-based units cannot be observed in the potential window available in acetonitrile, presumably because the accumulation of the large positive charge, due to the oxidation of the peripheral units, displaces the oxidation process of the other units to more positive potentials outside the examined potential window. However, experiments in liquid SO2 allowed to observe such high-potential processes for Ru10.15 Actually, a monoelectronic oxidation occurs in these latter conditions at þ 2.11 V, and was assigned to the central Ru(II) core, followed by a trielectronic process, assigned to oxidation of the three intermediate metal centers, at þ 2.44 V.15 Dendrimers capable of showing distinct oxidation processes related to topologically different units can be designed by taking advantage of the fact that Os(II) is easier to oxidize than Ru(II) and that peripheral ligands are stronger electron donors than the bridging ligands. So, for the [Os{(2,3-dpp)Ru[(2,3-dpp)Ru(bpy)2]2}3]20 þ (OsRu9) dendrimer, which is made of an Os(II)-based core and nine Ru(II)-based units, a 1:6 pattern is predicted for the electrons exchanged on oxidation. In agreement with these expectations, the differential pulse voltammogram of this compound (Fig. 5.8) shows a shoulder at þ 1.35 V, assigned to the one-electron oxidation of the central Os(II) metal ion, superimposed to six times higher peak at þ 1.55 V, assigned to the simultaneous one-electron oxidation of the six peripheral noninteracting Ru(II) ions.28 Oxidation of the three intermediate Ru(II) ions is further shifted toward more positive potentials and cannot be observed in the accessible potential window. For the [Os{(2,3-dpp)Ru[(2,3-dpp)Os(bpy)2]2}3]20 þ compound OsRu3Os6,28 made of an Os(II)-based core, three Ru(II)-based units in the intermediate positions, and six Os(II)-based units in the peripheral positions, one expects a 6 : 1 oxidation pattern instead than the 1:6 one observed in the previous case. This is fully consistent with the differential pulse voltammetry results that show an oxidation peak at þ 1.05 V six times higher than a following peak at þ 1.39 V (Fig. 5.8).28 The reduction pattern of the decanuclear, second-generation dendrimers is quite complex, as the number of the reducible sites increases significantly. For example, the previously seen [Ru{(2,3-dpp)Ru[(2,3-dpp)Ru(bpy)2]2}3]20 þ compound Ru10 shows a differential pulse voltammogram with two broad peaks (at 0.73 and 1.22 V) followed by several other overlapping peaks.28 The first peak, which corresponds to the exchange of six electrons, is due to the oneelectron reduction of the six outer equivalent bridging ligands. The width of the peak, compared to that observed on oxidation, suggests a nonnegligible interaction between the two ligands coordinated to the same metal, resulting in two closely lying three-electron processes. The second broad peak, which involves three electrons, is assigned to the one-electron reductions of the three inner bridging
5.4
Ru(II) AND Os(II) POLYPYRIDINE DENDRIMERS
135
Figure 5.8 Differential pulse voltammograms of Ru10 (top), OsRu9 (center), and OsRu3Os6 (bottom). Fc indicate ferrocene, used as reference.
ligands, occurring at close potential values. Since the interaction between reduced ligands depends also on the nature of the metal, the electron exchange pattern can be modified by replacing Ru(II) with Os(II). For further details, see Ref. 28 and the references therein.
136
REDOX-ACTIVE METAL–POLYPYRIDINE DENDRIMERS
5.4.5
Hexanuclear Metal Complexes
Redox properties of metal–polypyridine dendrimers having various nuclearity, such as hexanuclear,33 heptanuclear,34 nonanuclear,35 and tridecanuclear species,36 not based on the {M(dpp)3}2 þ core and/or with different branching developments than that common to the tetranuclear, decanuclear, and (see later) docosanuclear species described here, as well as dendrimers with mixed bridging ligands,33b,37 have also been investigated. Among such systems, a particularly studied example is the hexanuclear dendrimer Ru6, shown in Fig. 5.3, that is, [{(bpy)2Ru(2,3-dpp)}2 Ru(2,3-dpp)Ru{(2,3-dpp)Ru(bpy)2}2]12 þ .33a This species, as well as other dendrimers of the same nuclearity, can be considered as made of two trinuclear dendrons attached to a central 2,3-dpp bridging ligand. Detailed redox experiments have been performed on Ru6 besides the common experiments in acetonitrile fluid solution. In particular, both oxidation in liquid SO215 and reduction in highly purified DMF26 at low temperatures have been investigated and this allowed to identify four oxidation states and 26 reduction states. The total of 30 different redox states (plus the ground state), almost all of them fully reversible in the experimental conditions used, makes most likely Ru6 one of the (supra)molecular systems with the largest number of identified redox states ever studied. Some details obtained from the above-mentioned investigations are as follows. The hexanuclear Ru6 species has four outer and two inner metal centers oxidation active. Both in acetonitrile at room temperature (E1/2 at þ 1.52 V) and in liquid SO2 at low temperature (E1/2 at þ 1.46 V), an oxidation process involving the practically simultaneous one-electron oxidation of the four outer Ru(II) centers is evidenced (Fig. 5.9 and Table 5.1). This confirms that the electronic interaction between metal centers that are not directly connected via a bridging ligand is negligible from an electrochemical viewpoint in the metal–polypyridine dendrimers. At more positive potentials, only recordable in liquid SO2 at low temperature (Fig. 5.9), a bielectronic process, related to the simultaneous one-electron oxidation of the two inner metal centers at þ 2.11 V, is found. This result was at a first sight surprising, since the redox
Figure 5.9 Cyclic voltammogram of Ru6 in liquid SO2 at 70 C.
5.4
Ru(II) AND Os(II) POLYPYRIDINE DENDRIMERS
137
behavior of formerly studied systems evidenced that two metal centers that are directly connected experience a significant electronic interaction, leading to splitting of their oxidation potentials. However, it is known that the extent of metal–metal interaction is a function of the peripheral groups.24 In Ru6, the oxidized peripheral metal centers stabilize the metal-centered orbitals of the inner Ru(II) and therefore their interaction via the bridging ligand LUMO, resulting into a smaller metal–metal interaction and in virtually coincident oxidation potentials.15 At potentials higher than 3.0 V, other multielectronic processes are seen and have been attributed to ligandbased oxidations. The reduction pattern of Ru6 is extremely rich (Fig. 5.10). The assignment of each process to specific sites has been made by using the information gained from the redox properties of the smaller dendrimers. The first peak comprises four oneelectron reduction processes, closely spaced (E1/2 between 0.65 and 0.81 V), which are assigned to first reduction of the four “outer” bridging ligands. The second peak comprises five processes between 1.07 and 1.32 V. The first out of the five processes is localized on the “inner” 2,3-dpp ligand, and the other four processes are assigned to second reduction of the four “outer” bridges. At more negative potentials,
Figure 5.10 Cyclic voltammogram of Ru6 in DMF at 54 C and schematic localization of the redox processes.
138
REDOX-ACTIVE METAL–POLYPYRIDINE DENDRIMERS
second reduction of the inner bridging ligand and reduction of the peripheral bpy ligands occur.26 The reduction pattern of Ru6 warrants some additional comments; it can be considered that the first electron probably enters the inner 2,3-dpp ligand, which should be the easiest site to be reduced; however, upon second electron addition to one of the outer bridging ligands, the first electron should move to one of the other outer bridges, for coulombic reasons. This sort of redox-induced “electron shift” is not rare in this class of metal–polypyridine dendrimers.38 5.4.6
The Third Generation: Docosanuclear Species
Two docosanuclear dendrimers have been prepared, an all-Ru species (Ru22)18 and a species having an Os(II) center as the core.19 For the docosanuclear dendrimer made of an Os(II)-based core and 21 Ru(II)-based units (OsRu21), a 1: 12 electron ratio pattern is obtained on oxidation, corresponding to the one-electron oxidation at þ 1.42 Vof the Os(II) ion, followed by the simultaneous oxidation at þ 1.54 Vof the 12 equivalent and noninteracting peripheral Ru(II)-based units.19 Of course, the oxidation at þ 1.42 mV is missing in Ru22.18b Because of the presence of many polypyridine ligands, each capable of undergoing several reduction processes, the electrochemical reduction of this type of dendritic compounds produces very complex electron exchange patterns that have not been investigated in detail. 5.4.7
The Dendritic Structure and the Energy Transfer Processes
The dendritic structure fulfill the requisite for artificial antenna, that is, the possibility to build up in a few synthetic steps a larger number of chromophores. When the dendrimer core is a trifunctional {M(dpp)3}2 þ unit, and each branching unit is of the same type, a topographical control is obtained, with different layers of chromophores emanating from the core. The synthetic procedure also allows to collocate identical subunits in each layer, but leaves the freedom to collocate different subunits at different layers. Even when the same unit is selected for different layers, anyway, the chemical environment determines a fine-tuning of the redox and spectroscopic properties of the individual units. As a consequence, energy gradients between the layers can be designed. An example is given by the photophysical properties of Ru22. In this species, energy transfer takes place from the inner layers to the peripheral layer chromophores. When full energy gradient is not obtained, anyway, as when highenergy subunits are interposed between lower lying subunits, the energy transfer process becomes less efficient. A way to overcome this problem has been recently overcome and will be discussed in the next section. 5.4.8 Electronic Energy Transfer in Metal–Polypyridine Dendrimers: A Two-Step Electron Transfer Pathway in a Heptanuclear Dendron The detailed investigation of the photophysical properties of metal-based dendrimers of high generation has clearly evidenced that5a,e (i) downhill (or even isoergonic)
5.4
Ru(II) AND Os(II) POLYPYRIDINE DENDRIMERS
139
photoinduced energy transfer is highly efficient (close to unity) between nearby, directly connected subunits, taking also advantage from singlet–singlet mechanisms and39 (ii) downhill photoinduced energy transfer between subunits having interposed building blocks with high-energy excited-state levels is strongly inefficient. This is the case for the inefficiency of energy transfer from the peripheral {(bpy)2Ru(2,3dpp)}2 þ subunits to the central {Os(2,3-dpp)3}2 þ core in the decanuclear OsRu928 and docosanuclear OsRu2119 dendrimers, in spite of the favorable driving force of the process (0.17 eV, calculated from the 77 K emission spectra of the low-generation tetranuclear dendrimers Ru4 and OsRu3,8bona fide models for Ru-based peripheral and Os-based central emitters contained in the higher generation species). In fact,
Figure 5.11 Energy migration pattern across dendritic shells in OsRu3, OsRu6, OsRu21, and Cl2RuRu6, see text.
140
REDOX-ACTIVE METAL–POLYPYRIDINE DENDRIMERS
interposed building blocks with higher energy excited states (from herein, highenergy subunit barriers) make direct energy transfer between donor and acceptor negligible and apparently do not play an appreciable role as superexchange mediators (Fig. 5.11). The limitation in the efficiency of energy transfer between “isolated” components has appeared for long time a serious drawback of high-generation dendrimers of this type as far as their possible use as light-harvesting antennas is concerned, therefore reducing the interest in larger systems. To overpass these problems, taking advantage of the knowledge on the redox properties of the different building blocks (e.g., see trinuclear species), an heptanuclear ruthenium(II) species, namely, [Cl2Ru{(2,3-dpp)Ru[(2,3-dpp)Ru(bpy)2]2}2]12 þ (Cl2RuRu6, see Fig. 5.3) was prepared.40 Interestingly, in Cl2RuRu6 energy transfer from the four peripheral {(bpy)2Ru(2,3-dpp)}2 þ subunits to the {Cl2Ru(2,3-dpp)2} core takes place, in spite of the presence of the interposed {Ru(2,3-dpp)3}2 þ subunits, which have higher energy excited-state levels (Fig. 5.11). Sequential, two-step electron transfer is operative in Cl2RuRu6, allowing to overcome the high-energy subunit barrier limitation. The heptanuclear species Cl2RuRu6 also exhibits interesting redox behavior: it undergoes two reversible oxidation processes and four reversible reduction processes in acetonitrile in the potential window investigated ( þ 1.90/1.40 V versus SCE) (Table 5.1).40 Internal comparison allowed to attribute the first oxidation to an one-electron process, involving the “core” {Cl2Ru(2,3-dpp)2}, and the second oxidation to four simultaneous one-electron processes involving the four peripheral Ru-based chromophore, while each one of the reduction processes involves two electrons, with the reduction process localized into the four “outer” bridging ligands. By allowing to calculate the redox potentials of the various subunits, the redox study has been essential to identify the two-step electron transfer mechanism.40 The results obtained for Cl2RuRu6 also evidenced another interesting property of the studied dendrimers: long-range electron transfer involving spatially separated subunits is not hampered by the presence of interposed high-energy building blocks. Details are reported in Ref. 40. This issue can be of relevance for designing functional dendrimers for applications based on long-range electron transfer processes.
5.5
CONCLUSION
Metal–polypyridine dendrimers exhibit a quite interesting redox behavior that is strictly connected to important spectroscopic and photophysical properties. The electrochemical data offer a fingerprint of the chemical and topological structure of the dendrimers. Furthermore, the knowledge of the electrochemical properties of the mononuclear subunits and the synthetic control of the supramolecular structure allow designing dendrimers with predetermined redox patterns. The made-to-order synthetic control of the energy migration processes and of the number of electrons exchanged at mild potentials makes such dendritic complexes very attractive in view of their possible application for the implementation of light-harvesting systems for photochemical energy conversion and/or as multielectron transfer catalysts.
REFERENCES
141
REFERENCES 1. (a) V. Balzani, M. Venturi, A. Credi, Molecular Devices and Machines, 1st edition, VCH-Wiley, Weinheim, 2008, Chapter 5. (b) S. Campagna, S. Serroni, F. Puntoriero, C. Di Pietro, V. Ricevuto, in V. Balzani (Ed.), Electron Transfer in Chemistry, Vol. 5, VCH-Wiley, Weinheim, 2001, p. 186, and refs therein. 2. (a) G. MacDermott, S. M. Prince, A. A. Freer, A. M. Hawthorntwwaite-Lawless, M. Z. Papiz, R. J. Cogdell, N. W. Isaacs, Nature 1995, 374, 517. (b) T. Pullerits, V. Sundstr€ om, Acc. Chem. Res. 1996, 29, 381. 3. See, for example: (a) J. H. Alstrum-Acevedo, M. K. Brennaman, T. J. Meyer, Inorg. Chem. 2005, 44, 6802. (b) R. Eisenberg, D. G. Nocera, Inorg. Chem. 2005, 44, 6799. (c) V. Balzani, N. Armaroli, Angew. Chem., Int. Ed. 2007. 4. See, for example: (a) C. Devadoss, P. Bharathi, J. S. Moore, J. Am. Chem. Soc. 1996, 118, 9635. (b) D.-L. Jiang, T. Aida, J. Am. Chem. Soc. 1998, 120, 10895. (c) S. L. Gilat, A. Adronov, J. M. J. Frechet, Angew. Chem., Int. Ed. 1999, 38, 1422. (d) F. Loiseau, S. Campagna, A. Hameurlaine, W. Dehaen, J. Am. Chem. Soc. 2005, 127, 11352. 5. See, for example: (a) V. Balzani, S. Campagna, G. Denti, A. Juris, S. Serroni, M. Venturi, Acc. Chem. Res. 1998, 31, 26. (b) S. Serroni, S. Campagna, F. Puntoriero, C. Di Pietro, F. Loiseau, N. D. McClenaghan, Chem. Soc. Rev. 2001, 30, 367. (c) S. Campagna, C. Di Pietro, F. Loiseau, B. Maubert, N. McClenaghan, R. Passalacqua, F. Puntoriero, V. Ricevuto, S. Serroni, Coord. Chem. Rev. 2002, 229, 67. (d) S. Serroni, S. Campagna, F. Puntoriero, F. Loiseau, V. Ricevuto, R. Passalacqua, M. Galletta, C. R. Chimie 2003, 6, 883. (e) V. Balzani, A. Juris, F. Puntoriero, S. Campagna, in R. De Jaeger, M. Gleria (Eds.), Inorganic Polymers, Nova Science Publishers, New York, 2007, Chapter 19. 6. V. Balzani, G. Bergamini, S. Campagna, F. Puntoriero, Top. Curr. Chem. 2007, 280, 1, and refs therein. 7. For further details, see also: (a) D. A. Tomalia, A. M. Naylor, W. A. Goddard III, Angew. Chem., Int. Ed. Engl. 1990, 29, 138. (b) J. M. J. Frechet, Science 1994, 263, 1710. (c) N. Ardoin, D. Astruc, Bull. Soc. Chim. Fr. 1995, 132, 875.(d) G. R. Newkome, C. N. Moorefield, F V€ogtle. Dendritic Molecules: Concepts, Synthesis, Perspectives, VCH, Weinheim, 1996. 8. V. Balzani, A. Juris, M. Venturi, S. Campagna, S. Serroni, Chem. Rev. 1996, 96, 759. 9. (a) A. Juris, V. Balzani, F Barigelletti, S. Campagna, P. Belser, A. von Zelewsky, Coord. Chem. Rev. 1988, 84, 85. (b) T. J. Meyer, Acc. Chem. Res. 1989, 22, 163. (c) K. Kalyanasundaram, Photochemistry of Polypyridine and Porphyrin Complexes, Academic Press, London, 1991. (d) S. Campagna, F. Puntoriero, F. Nastasi, G. Bergamini, V. Balzani, Top. Curr. Chem. 2007, 280, 117. 10. D. Kumaresan, K. Shankar, S. Vaidya, R. H. Schmehl, Top. Curr. Chem. 2007, 281, 101. 11. See, for example: (a) G. Farnia, S. Roffia, J. Electroanal. Chem., 1981, 122, 347. (b) S. Roffia, R. Casadei, F. Paolucci, C. Paradisi, C. A. Bignozzi, F. Scandola, J. Electroanal. Chem. 1991, 302, 157. 12. M. Venturi, S. Serroni, A. Juris, S. Campagna, V. Balzani, Top. Curr. Chem. 1998, 197, 193. 13. (a) J. G. Gaudiello, K. A. Norton, W. H. Woodruff, A. J. Bard, Inorg. Chem. 1984, 23, 3. (b) E. Garcia, J. Kwak, A. J. Bard, Inorg. Chem. 1988, 27, 4379. 14. P. Ceroni, F. Paolucci, S. Roffia, S. Serroni, S. Campagna, A. J. Bard, Inorg. Chem. 1998, 37, 2829.
142
REDOX-ACTIVE METAL–POLYPYRIDINE DENDRIMERS
15. P. Ceroni, F. Paolucci, C. Paradisi, A. Juris, S. Roffia, S. Serroni, S. Campagna, A. J. Bard, J. Am. Chem. Soc. 1998, 120, 5480. 16. (a) M. B. Robin, P. Day, Adv. Inorg. Radiochem. 1967, 10, 247. (b) E. Richardson, H. Taube, J. Am. Chem. Soc. 1983, 105, 40. (c) C. Creutz, Prog. Inorg. Chem. 1983, 30, 1. (d) N. S. Hush, Coord. Chem. Rev. 1985, 64, 135. 17. J. B. Flanagan, S. Margel, A. J. Bard, F. C. Anson, J. Am. Chem. Soc. 1978, 100, 4248. 18. (a) S. Serroni, G. Denti, S. Campagna, A. Juris, M. Ciano, V. Balzani, Angew. Chem., Int. Ed. Engl. 1992, 31, 1493. (b) S. Campagna, G. Denti, S. Serroni, A. Juris, M. Venturi, V. Ricevuto, V. Balzani, Chem. Eur. J. 1995, 1, 211. 19. S. Serroni, A. Juris, M. Venturi, S. Campagna, I. Resino Resino, G. Denti, A. Credi, V. Balzani, J. Mater. Chem. 1997, 7, 1227. 20. S. Campagna, G. Denti, G. De Rosa, L. Sabatino, M. Ciano, V. Balzani, Inorg. Chem., 1989, 28, 2565, and refs therein. 21. G. Denti, S. Campagna, L. Sabatino, S. Serroni, M. Ciano, V. Balzani, Inorg. Chem. 1990, 29, 4750. 22. (a) S. Roffia, M. Marcaccio, C. Paradisi, F. Paolucci, V. Balzani, G. Denti, S. Serroni, S. Campagna, Inorg. Chem. 1993, 32, 4003. (b) F. Loiseau, S. Serroni, S. Campagna, Collect. Czech. Chem. Commun. 2003, 68, 1677. 23. G. Denti, S. Serroni, L. Sabatino, M. Ciano, V. Ricevuto, S. Campagna, Gazz. Chim. Ital., 1991, 121, 37. 24. G. Giuffrida, S. Campagna, Coord. Chem. Rev. 1994, 135–136, 517. 25. D. M. Bassani, J.-M. Lehn, S. Serroni, F. Puntoriero, S. Campagna, Chem. Eur. J. 2003, 9, 5936. 26. M. Marcaccio, F. Paolucci, C. Paradisi, S. Roffia, C. Fontanesi, L. J. Yellowlees, S. Serroni, S. Campagna, G. Denti, V. Balzani, J. Am. Chem. Soc. 1999, 121, 10081. 27. S. Campagna, G. Denti, L. Sabatino, S. Serroni, M. Ciano, V. Balzani, Gazz. Chem. Ital., 1989, 119, 415. 28. G. Denti, S. Campagna, S. Serroni, M. Ciano, V. Balzani, J. Am. Chem. Soc., 1992, 114, 2944. 29. F. Puntoriero, S. Serroni, A. Licciardello, M. Venturi, A. Juris, V. Ricevuto, S. Campagna, J. Chem. Soc., Dalton Trans. 2001, 1035. 30. S. Serroni, S. Campagna, F. Puntoriero, A. Juris, G. Denti, V. Balzani, M. Venturi, Inorg. Synth. 2002, 40, 10. 31. S. Campagna, G. Denti, L. Sabatino, S. Serroni, M. Ciano, V. Balzani, J. Chem. Soc. Chem. Commun. 1989, 1500. 32. S. Serroni, G. Denti, S. Campagna, M. Ciano, V. Balzani, J. Chem. Soc. Chem. Commun. 1991, 944. 33. (a) S. Campagna, G. Denti, S. Serroni, M. Ciano, V. Balzani, Inorg. Chem. 1991, 30, 3728. (b) G. Denti, S. Serroni, S. Campagna, V. Ricevuto, A. Juris, M. Ciano, V. Balzani, Inorg. Chim. Acta 1992, 198–200, 507. 34. G. Denti, S. Campagna, L. Sabatino, S. Serroni, M. Ciano, V. Balzani, Inorg. Chim. Acta 1990, 176, 175. 35. J. Leveque, C. Moucheron, A. Kirsch-De Mesmaeker, F. Loiseau, S. Serroni, F. Puntoriero, S. Campagna, H. Nierengarten, A. Van Dorsselaer, Chem. Commun 2004, 877.
REFERENCES
143
36. S. Campagna, G. Denti, S. Serroni, M. Ciano, A. Juris, V. Balzani, Inorg. Chem. 1992, 31, 2982. 37. S. Serroni, S. Campagna, G. Denti, T. E. Keyes, J. G. Vos, Inorg. Chem. 1996, 35, 4513. 38. See, for example: A. Juris, M. Venturi, L. Pontoni, I. Resino Resino, V. Balzani, S. Serroni, S. Campagna, G. Denti, Can. J. Chem., 1995, 73, 1875. 39. J. Andersson, F. Puntoriero, S. Serroni, A. Yartsev, T. Pascher, T. Polivka, S. Campagna, V. Sundstr€om, Chem. Phys. Lett. 2004, 386, 336. 40. F. Puntoriero, S. Serroni, M. Galletta, A. Juris, A. Licciardello, C. Chiorboli, S. Campagna, F. Scandola, ChemPhysChem, 2005, 6, 129.
CHAPTER 6
Dendrimers as Multielectron Storage Devices PAOLA CERONI and MARGHERITA VENTURI Dipartimento di Chimica “G. Ciamician,” Alma Mater Studiorum, Universita di Bologna, Bologna, Italy
6.1
INTRODUCTION
Dendrimers1 are complex, repeatedly branched tree-like compounds that can be synthesized with well-defined composition and a high degree of order. By using suitable synthetic strategies, it is possible to prepare dendrimers that contain selected functional units in predetermined sites of their structure, namely, core, branches, and surface. Because of their tree-like multibranched architecture, they can form internal dynamic niches in which small molecules or ions can be hosted.2,3 Dendrimers can often exhibit remarkable chemical, physical, and biological properties, with a wide range of potential applications in such different fields as medicine, biology, chemistry, physics, and engineering.4 Incorporation of redox-active units in a dendritic architecture enable us to gain information on: (i) dendrimer structure and superstructure, (ii) self-assembly processes, (iii) degree of electronic interaction and communication between redox units located in different sites, and (iv) changes in conformation brought about by electron transfer processes. Electroactive dendrimers are attracting increasing interest in view of their possible application as sensors,5 catalysts,5,6 enzyme mimics,7 in which a redox center is buried inside the dendritic nanoenvironment, and, last but not least, multielectron storage devices.5a,8,9 A molecular multielectron storage device is able to store and exchange simultaneously a large and predetermined number of electrons under an electrochemical stimulus (potential difference). To perform such a task a molecule or, most likely, a supramolecular structure should consist of multiple identical and noninteracting
Electrochemistry of Functional Supramolecular Systems. Edited by Paola Ceroni, Alberto Credi, and Margherita Venturi Copyright 2010 John Wiley & Sons, Inc.
145
146
DENDRIMERS AS MULTIELECTRON STORAGE DEVICES
redox units able to reversibly exchange electrons with another molecular substrate or an electrode. The redox-active units should exhibit chemically reversible and fast electron transfer processes at easily accessible potential difference and chemical robustness under the working conditions. Multielectron storage devices can be used as (i) redox catalysts, also called electron mediators, for multielectron processes, (ii) electrochemical sensors with signal amplification, and (iii) molecular batteries that can be foreseen to power molecular machines in the future or that can be used to construct flexible rechargeable batteries.10 Dendrimers are ideal scaffolds to construct these devices since: (i) a large number of proper redox units can be placed in the periphery and/or branching points of their structure with the possibility of tuning their distance, (ii) the dendrimer skeleton can be designed to minimize electronic interaction between the redox centers, and (iii) the dendrimer periphery can be optimized to get the desired solubility properties and processability to eventually deposit the dendrimers on a surface. Other possible scaffolds are constituted by polymers11 and nanoparticles,12 which are easier to synthesize, but they do not enable control and tuning of the number, position, and distance of the active units. In this chapter, we will review recent advances in the field of dendrimers as multielectron storage devices that are dendrimers containing multiple electroactive units. Because of the very extensive literature and space reasons, only few selected examples, divided according to the chemical nature of the units used to functionalize the dendritic structure, will be described.
6.2 ELECTROCHEMICAL PROPERTIES OF THE REDOX-ACTIVE UNITS Dendrimers possessing a large number of chemically identical redox units can be defined as charge pooling devices only if they satisfy some conditions. Their redox units should be 1. electrochemically equivalent, that is, with the same half-wave potential; 2. chemical reversible and robust; 3. characterized by high values of the rate constants for heterogeneous electron transfer and through-space electron hopping among the redox units. The first condition is fulfilled if the extent of interactions between the redox sites, solvation changes, ion pairing, and structural changes of the macromolecule are negligible. The second prerequisite assures that charge can be stored for a long period of time without any competing chemical reaction involving the electrogenerated species and that an high turnover of the system can be reached. The third property guarantees that all the electroactive units can exchange electrons and that the process of storing and releasing electrons is fast. For large globular molecules, such as
6.2
ELECTROCHEMICAL PROPERTIES OF THE REDOX-ACTIVE UNITS
147
dendrimers, all the redox-active moieties cannot simultaneously be in contact with or in proximity to the electrode. Therefore, the dendrimer must rotate and/or electron hopping among identical redox units must occur to allow all these units to be oxidized or reduced in the timescale of the electrochemical experiment. For a cyclic voltammetric experiment with conventional electrodes and scan rates below 10 V/s, timescale is usually sufficiently long to enable dendrimer rotation and electron hopping between neighboring units, so that all the redox units can be addressed. Another desirable property for a charge storing device is that the multielectronic transfer process is the first oxidation or reduction process of the investigated species and that the corresponding half-wave potential corresponds to easy accessible potential values. These prerequisites should be verified experimentally by electrochemical techniques. In particular, a standard cyclic voltammetric experiment (with conventional electrodes) should present a single and reversible wave with the same shape of the corresponding curve for a species containing a single redox center, but higher peak current. The formal redox potentials of the electroactive units are distributed statistically along this wave, as shown by Bard and Anson for multiferrocenyl polymers.13 To estimate the number of electrons exchanged, further experiments are requested since peak current of the voltammetric wave depends not only on the number of electrons exchanged but also on the scan rate, the concentration, and the diffusion coefficient of the electroactive species, which should be measured independently.14 Analogously, the limiting current observed in a voltammetric experiment with a rotating disc electrode depends on the previously reported parameters and on the kinematic viscosity of the solvent. An independent measure of the diffusion coefficient can be obtained by nonelectrochemical techniques, such as pulse gradient stimulated echo (PGSE) NMR spectroscopy, taking care to perform the experiment in the same solvent/supporting electrolyte mixture or to correct the obtained diffusion coefficient values for the different viscosity, as illustrated by Kaifer.15 Otherwise, diffusion coefficients can be measured by other electrochemical techniques, such as bulk coulometry and/or spectroelectrochemistry in the same solvent/supporting electrolyte mixture. In both cases, the total current exchanged during the bulk electrolysis is measured. From this value, by knowing the exact concentration of the solution, the number of exchanged electrons per electroactive species can be estimated. The drawbacks of these techniques are related to the need of specialized cells and, most important, their long timescales, so that competing reactions involving the electrogenerated products may take place. Another alternative method has been proposed by Bard,16 in which the diffusion coefficient and the number of exchanged electrons can be determined by coupling steady-state cyclic voltammetry and chronoamperometry with ultramicroelectrodes. This procedure is explained with details in Ref. 16. It is worth noting that when determining diffusion coefficients by electrochemical methods, the measured current can be affected by factors not related to transport of the electroactive species. Nonfaradaic current may result from charging of the double-layer, adsorption of electroactive or nonelectroactive material, electron hopping between stationary redox sites. These parameters should be taken into account and experimentally verified in each investigated case.
148
6.3
DENDRIMERS AS MULTIELECTRON STORAGE DEVICES
METALLOCENE DENDRIMERS
Because of their reversible electrochemical properties, ferrocene [biscyclopentadienyl-iron(II), FeCp2] and cobaltocenium [biscyclopentadienyl-cobalt(III), CoCp2 þ ] are the most common electroactive units used to functionalize dendrimers. Both metallocene residues are stable, 18-electron systems, which differ on the charge of their most accessible oxidation states: zero for ferrocene and þ 1 for cobaltocenium. Ferrocene undergoes electrochemically reversible one-electron oxidation to the positively charged ferrocenium form, whereas cobaltocenium exhibits electrochemically reversible one-electron reduction to produce the neutral cobaltocene. Both electrochemical processes take place at accessible potentials in ferrocene- and cobaltocenium-containing compounds. Such units, together with methyl derivatives of ferrocene and cyclopentadienyl (h6-arene)-iron(II), [FeCp(h6-arene)] þ , have been extensively used as dendrimer peripheral units, but examples of dendrimers containing metallocene residues both as a core and as terminal units and as a core are also reported.5a,b,17 Because the related literature is very extensive, only selected and paradigmatic examples will be described. Dendrimers Containing Ferrocene-Type Units at Their Periphery By exploiting efficient synthetic approaches (dendrimer growth strategy pioneered by Newkome18 and “click” chemistry), several dendrimers containing 27 (compound 1, Fig. 6.1), 54, 81, and 243 ferrocene units in their periphery have been synthesized.19,20 Regardless to dendrimer generation and the chemical nature of the core and branches they show a similar electrochemical behavior characterized by the following features. 1. All the ferrocene units are oxidized at the same potential, with a number of exchanged electrons equal (within experimental error) to the number of peripheral ferrocene units (250 30 exchanged electrons for the dendrimer containing nominally 243 units20d). This finding not only shows that each ferrocene behaves independently but also shows that for these dendrimers the electron transfer between the terminal units and the electrode is fast. Several reasons have been invoked to rationalize such a remarkable electrochemical reversibility. First, the timescale of the standard electrochemical experiment (usually approximately 0.1 s) is larger than the rotation rate of the dendrimer, thereby all the ferrocene units have time to come close to the electrode.21 Second, even if the intramolecular distance between two redox units is large (most often more than 10 bonds), through-space electron hopping can occur because of the flexibility of the dendritic tethers that bring two redox units at the minimum distance in a very fast dynamic process.22 This stepwise electron transfer between the remote units and the ones that are located near the electrode contributes to the observed fast electron exchange with the electrode.
6.3
Figure 6.1
METALLOCENE DENDRIMERS
149
Structural formula of dendrimer 1 containing 27 ferrocene units at its periphery.20d
2. The potential value for the ferrocene-based process exhibits no or only minor dependence on dendrimer generation. 3. The ferrocene-based redox processes show also good chemical reversibility. The soluble orange-red ferrocene-functionalized dendrimers can be indeed chemically oxidized in CH2Cl2 to the insoluble deep-blue polyferrocenium dendrimers that are reduced back to the starting compounds without any decomposition. Taking advantage of the insolubility of their oxidized form the dendrimers carrying at the periphery a high number of ferrocene units are also useful to obtain stable modified platinum electrodes. Because of all the described properties, these dendrimers behave as molecular batteries that might find applications in molecular–electronic devices. A dendrimer containing 64 [FeCp(h6-C6Me6)] þ units at the periphery, compound 2 in Fig. 6.2, has also been prepared.23 Its electrochemical behavior is similar to that
150
DENDRIMERS AS MULTIELECTRON STORAGE DEVICES
Figure 6.2 Structural formula of dendrimer 2 containing 64 [FeCp(h6-C6Me6)] þ units at its periphery.23
observed for the ferrocene-functionalized dendrimers: a single wave involving the exchange of 63 6 electrons. Interestingly, the reduced form of this dendrimer, in which all the 64 peripheral units are present as 19-electron Fe(I) species, is capable to reduce fullerene molecule (C60). Under the experimental conditions used (MeCNtoluene, 30 C) a black precipitate is formed; it is made of the starting dendrimer with C60 anions most likely located at the dendrimer periphery and engaged in tight ion pairs.
6.3
METALLOCENE DENDRIMERS
151
Very recently, giant redox dendrimers were synthesized with ferrocene and pentamethylferrocene termini up to a theoretical number of 39 tethers (seventh generation) evidencing that lengthening of the tethers is a reliable strategy to overcome the bulk constraint at the dendrimer periphery.24 These redox metallodendrimers were investigated with a variety of techniques: (i) cyclic voltammetry has revealed a full chemical and electrochemical reversibility up to the seventh generation, (ii) chemical oxidation was used to isolate and characterize the blue 17-electron ferrocenium and deep-green mixed-valence Fe(III)/Fe(II) dendritic complexes, and (iii) atomic force microscopy, employed to study the behavior of the dendrimers on a mica surface, enabled to compare the size of the oxidized cationic form of the dendrimers with that of their neutral form. For the fifth-generation dendrimer, it was found that the average height of the oxidized species (6.5 0.6 nm) is much larger than that of its neutral form (4.5 0.4 nm). Thus, these giant redox metallodendrimers exhibit a “breathing mechanism” controlled by the redox potential. In CH2Cl2, a solvent that favors ion pairing, this breathing is accompanied by a displacement of the anions from the oxidant toward all the redox termini of the dendrimer, and the opposite is occurring upon reduction of the ferrocenium dendrimer. Because of the peculiar electrochemical behavior (a single cyclic voltammetric wave characterized by remarkable electrochemical and chemical reversibility), dendrimers terminated with ferrocene-type units can be profitable used as exoreceptors, provided that they contain a group able of interacting through noncovalent bonds with the species to be recognized. Furthermore, such a group has to be located near the ferrocene units to sufficiently perturb their electrochemical response as a consequence of the interaction with the guest species. The construction of dendrimers that behave as exoreceptors for anion sensing is of crucial importance from the biological and environmental viewpoints. In this regard, the first attempt20a,25 is represented by a family of amidoferrocene-terminated dendrimers that can recognize the HSO4 and H2PO4 oxoanions in CH2Cl2 solution. The behavior of these dendrimers containing 3, 9, and 18 amidoferrocene peripheral units was compared with that of the monoamidoferrocene compound by finding a clear positive dendritic effect. The results show indeed that (i) the monoamidoferrocene compound does not sense the oxoanions whereas the dendrimers do, and that (ii) for the dendrimers the effect is larger as the generation increases. The results also evidence that for HSO4 the interaction is of the weak type with only a potential shift of the ferrocene-based wave, while in the case of H2PO4 the interaction is of the strong type with the appearance of a new wave that increases at the expense of the initial one.26 More recently, recognition of H2PO4 and ATP2 anions has been observed with dendrimers obtained by condensation of ferrocene-terminated dendrons on octahedral Mo6 cluster27 (compound 3 in Fig. 6.3) or Au nanoparticles28 cores. Sensing of the oxoanions is evidenced in cyclic voltammetry carried out in CH2Cl2 with Pt electrode by the appearance of a new wave thereby indicating an interaction of the strong type. Large Au nanoparticle-cored and ferrocene-terminated dendrimers can also be used to prepare stable modified Pt electrode for ATP2 recognition and sensing, that contrary to most electrochemical sensors are recyclable.29
152
DENDRIMERS AS MULTIELECTRON STORAGE DEVICES
Figure 6.3 Structural formula of dendrimer 3 containing an octahedral Mo6 cluster as core and 36 ferrocene units in its periphery.27
Interestingly, a family of 1,2,3-triazole-linked dendrimers containing from 9 (compound 4, Fig. 6.4a) up to 243 peripheral ferrocene units give strong recognition of both oxoanions and metal cations with positive dendritic effect.30 Interaction with H2PO4 and ATP2 causes the appearance of a new wave at a potential less positive than the initial one because of the electron releasing properties of these anions. On the other hand, recognition of Cuþ , Cu2þ , Pd2þ , and Pt2þ , that have electron-withdrawing character, gives rise to a new wave that is shifted to potential more positive than the initial one (Fig. 6.4b). In the case of the interaction with Pd2þ cations, reduction by using NaBH4 or methanol leads to the production of Pd nanoparticles that are stabilized either by several dendrimers or by encapsulation inside a dendrimer.31 Very recently, it has been shown that new dendrimers containing up to 243 azidomethylferrocene peripheral units form modified Pt electrodes that are increasingly robust with increasing dendrimer generation. These electrodes enable
6.3
METALLOCENE DENDRIMERS
153
Figure 6.4 (a) Structural formula of the 1,2,3-triazole-linked dendrimer 4 containing nine ferrocene peripheral units; (b) cyclic voltammogram of 4 without (1) and in the presence of (2) H2PO4 (A) and (3) Pd2 þ (M þ ).30 Reproduced with permission from Ref. 30.
154
DENDRIMERS AS MULTIELECTRON STORAGE DEVICES
ATP2 sensing better than in solution and they exhibit a positive dendritic effect not evidenced in solution.32 Poly(propylene amine) dendrimers containing 4, 8, and 64 amidoferrocene peripheral units have also been incorporated in the highly ordered channels of mesoporous silica obtaining a novel type of redox-active materials. One significant feature of these new composite materials is that the ferrocene units of the guest dendrimers are easily accessible to electrochemical oxidation, as revealed by studies carried out in MeCN solutions by using Pt electrodes derivatized with films of such dendrimer-matrix complexes.33 While dendrimers containing ferrocene moieties are numerous, dendrimers decorated with permethylferrocene units have been relatively less explored.34 The main reason for this underdeveloped chemistry is the synthetic difficulties. However, organometallic compounds containing polymethylcyclopentadienyl ligands are interesting since they often exhibit significantly different properties than their nonmethylated analogs.35 As a result of the enhanced donor capability of the polymethylated cyclopentadienyl rings, polymethylferrocene derivatives are oxidized at lower redox potentials, and, as a consequence, electrode surfaces functionalized with these materials have shown to be of importance in electrocatalysis. For instance, electrode surfaces modified with polymethyl ferrocene derivatives can be used as sensors for cytochrome c36 and in the amperometric determination of hydrogen peroxide.37 To obtain ferrocene-based dendrimers having a redox potential more negative than that of ferrocene ones, new diaminobutane-based poly(propylene amine) dendrimers containing 4, 8, 16, and 32 (compound 5 in Fig. 6.5) octamethylferrocene peripheral units have been prepared.38 In agreement with expectations, it has been found that all generations of this dendritic family exhibit a single redox process in CH2Cl2 or tetrahydrofuran with a potential of approximately 0.04 V versus SCE. This potential is considerably more negative (approximately 0.45 V) than those of analogous nonmethylated ferrocenyl dendrimers in the same medium,39 and it results from the electron-donating effect of the eight methyl groups on the ferrocene rings. The fact that only a single redox process is observed implies simultaneous multielectron transfers at the same potential of the 4, 8, 16, and 32 octamethylferrocene peripheral units, and indicates that in the dendrimers such units are essentially noninteracting. These dendrimers also exhibit a tendency to adsorb on electrode surfaces, which is more pronounced for the higher generations. Pt or glassy carbon electrodes modified with films derived from the dendrimers, show a well-defined and persistent electrochemical response. Preliminary studies have shown that electrodes modified with these dendritic molecules containing redox-active octamethylferrocenyl moieties and amine NHCH2 hydrogen-bonding functionalities act as a new type of receptor system capable of electrochemically recognizing anionic guest species. As far as the electrochemical properties are concerned, a different behavior is exhibited by directly linked ferrocene oligomers in which metal oxidation occurs in separate steps because of a strong electronic interacting among the metal centers. The biferrocene (BFc) moiety, for example, can exist in three oxidation states, that is, the neutral Fe2(II,II), the mixed-valence cationic Fe2(II,III)þ , and the dicationic
6.3
METALLOCENE DENDRIMERS
155
Figure 6.5 Structural formula of dendrimer 5 containing 32 octamethylferrocene peripheral units.38
Fe2(III,III)2þ states. By considering that organometallic dendrimers based on conjugated ferrocene units are of special importance since mixed-valence states have interesting electrical, redox, and magnetic properties, recently three generations of poly(propylene amine) dendrimers, decorated at their periphery with 4, 8, and 16 (compound 6, Fig. 6.6a) BFc units, respectively, have been synthesized and the electrochemical behavior of the dendrimers complexed with b-cyclodextrins (b-CD) and adsorbed at self-assembled monolayers (SAMs) of heptathioether-functionalized b-CD on gold (molecular printboard) has been studied.40 The data obtained show that (i) the complexation of dendrimers to b-CD is sensitive to the oxidation state of the BFc unit; indeed oxidation of the neutral Fe2(II, II) state to the cationic, mixed-valence biferrocenium Fe2(II,III)þ species resulted in dissociation of the host–guest complexes and desorption of the dendrimers from the molecular printboard, (ii) for the dendrimer-b-CD complexes, in which all the BFc units are engaged in host–guest interactions, two oxidation waves are observed (oxidation of the ferrocene units interacting with b-CD and then oxidation of the remaining ferrocene units), and (iii) for the dendrimers on the molecular printboard, in which the number of interacting end groups is 3, 4, and 4 for dendrimer generations 1,
156
DENDRIMERS AS MULTIELECTRON STORAGE DEVICES
Figure 6.6 (a) Structural formula of dendrimer 6 containing at its periphery 16 biferrocene units; (b) schematic representation of the proposed oxidation mechanism (at low scan rate) of dendrimer 6 in solution and immobilized at the b-CD host surface.40 Reproduced with permission from Ref. 40. (See the color version of this figure in Color Plates section.)
2, and 3, respectively, three oxidation steps occur: oxidation of the surface-bCD-bound BFc moieties to give the mixed-valence state, Fe2(II,III)þ , then oxidation of the nonsurface-interacting BFc groups leading to the Fe2(II,III)þ state for all the BFc moieties, and finally oxidation of all the BFc moieties to the dicationic Fe2(III,III)2þ state (Fig. 6.6b).
6.3
METALLOCENE DENDRIMERS
157
Dendrimers Terminated with Cobaltocenium and Ferrocene-Cobaltocenium Units Like ferrocene, cobaltocenium is an excellent organometallic moiety to incorporate in or functionalize dendritic systems. As already discussed, it is indeed isoelectronic with ferrocene, highly stable, positively charged complex, which undergoes a reversible monoelectronic reduction to yield the neutral cobaltocene. The first family of polycationic cobaltocenium dendrimers reported in the literature is constituted by four generations of the diaminobutane-based poly(propylene amine) dendrimers containing 4, 8, 16, and 32 (compound 7 in Fig. 6.7a) cobaltocenium peripheral units, respectively.41
Figure 6.7 (a) Structural formula of the fourth-generation dendrimer 7 containing at its periphery 32 cobaltocenium units; (b) cyclic voltammograms of the first-, second-, third-, and fourth-generation (7) dendrimers of the same family from left to right.41 Reproduced with permission from Ref. 41.
158
DENDRIMERS AS MULTIELECTRON STORAGE DEVICES
From the electrochemical viewpoint all generations of this dendritic family show a single redox process ascribed to the simultaneous reduction of all the cobaltocenium peripheral moieties at the same potential in MeCN solution. Upon reduction, the dendrimers exhibit a tendency, more pronounced for the higher generations, to electrodeposit onto the electrode surface, as clearly evidenced by the shape of the anodic part of the cyclic voltammetric wave of cobaltocenium that becomes sharper by increasing dendrimer generation (Fig. 6.7b). This behavior is opposite to that previously observed for the analogous neutral ferrocene dendrimers, which become insoluble upon oxidation to the corresponding polycationic species.19a,42 A valuable feature of these organometallic dendrimers is their ability, related to the insolubility of the reduced species, to modify Pt or glassy carbon electrodes, resulting in electroactive material that remains persistently attached to the electrode surface. Interestingly, upon addition of b-CD to aqueous solution of the dendrimers with 4, 8, and 16 peripheral cobaltocenium units, the shape of the cyclic voltammetric wave changes becoming for all the analyzed dendrimers consistent with a fully reversible process. Such a finding clearly indicates the solubilization of the reduced form of the dendrimers by formation of inclusion complexes between the peripheral cobaltocene units and the freely diffusing b-CD hosts. These systems afford an example of high molecular weight supramolecular assemblies that undergo association upon “electrochemical activation” of the guest.43 Mixed ferrocene-cobaltocenium poly(propylene amine) dendrimers have also been prepared: they contain at their periphery a total number of metal centers equal to 4, 8, 16, and 32, respectively.44 Because of the presence of neutral ferrocene and cationic cobaltocenium units, these dendrimers exhibit an interesting solubility behavior that can be modified by varying the ratio of the peripheral organometallic moieties. As far as the electrochemical properties are concerned, all the four dendrimers show only one reversible oxidation wave, clearly assigned to the simultaneous one-electron oxidation of the multiple peripheral noninteracting ferrocene units, and only one reduction wave attributed to the simultaneous one-electron reduction of the multiple noninteracting terminal cobaltocenium units. Reduction causes the precipitation of the reduced neutral form of the dendrimers onto the electrode surfaces, and on the reverse scan the dendrimers redissolve as a consequence of their reoxidation. This feature has been exploited to obtain modified electrodes that are extremely durable and reproducible. The surface-confined heterometallic dendrimers are shown to exhibit a double function: while the ferrocene units act as mediators in enzymatic processes under anaerobic conditions, the cobaltocenium moieties take part in an electrocatalytic process in reaction conducted in the presence of oxygen. For example, the application of these dendrimer-modified electrodes on which glucose oxidase has been immobilized allows the reduction of oxygen to occur at a less negative potential and with higher intensity than with bare glassy carbon electrodes. Other Metallocene-Like Dendrimers Poly(propylene amine) dendrimers carrying at their periphery 9, 16, and 27 (compound 8 in Fig. 6.8) tetrairon [{CpFe(m3-CO)}4] clusters have been recently prepared
6.3
METALLOCENE DENDRIMERS
159
Figure 6.8 Structural formula of dendrimer 8 carrying at its periphery 27 tetrairon [{CpFe(m3-CO)}4] clusters.44
and their electrochemical behavior and oxoanion sensing properties have been investigated.45 The cyclic voltammograms of these dendrimer-clusters in CH2Cl2 resemble that of the monomeric tetrairon cluster,46 that is, the clusters are sufficiently remote from one another in the dendrimers to render the electrostatic factor almost nil. Therefore, all the sites undergo the redox change Fe4 ! Fe4þ in a single reversible wave. This wave has been used for redox recognition and titration of HSO4, H2PO4, and ATP2 oxoanions, added as their n-tetrabutylammonium salts to the electrochemical cell
160
DENDRIMERS AS MULTIELECTRON STORAGE DEVICES
containing a CH2Cl2 solution of the dendrimers. Remarkably, and for the first time, ATP2 is better recognized than H2PO4, whereas the opposite holds with ferrocene dendrimers. This specificity can be certainly attributed to the mutual nanosize of the Fe4 clusters and ATP2, which facilitates their interaction, whereas the smaller ferrocene groups do not exhibit this property. A Pt electrode modified with the 16-Fe4 or 27-Fe4 dendrimer provides selective ATP2 recognition. Furthermore, it is possible to wash the dendrimer with CH2Cl2 for recycling, and the quality of the modified electrode is optimum with the larger 27-Fe4 dendrimer due to its better adsorption. Finally, given the known properties of the monomeric tetrairon cluster as a selective hydrogenation catalyst for various functional groups,47 this family of metallodendrimers should be profitably used as recyclable catalysts.2d,48 By exploiting the good electrochemical behavior of the tetrairon cluster [{CpFe (m3-CO)}4] and the characteristic of gold nanoparticles (AuNPs) to be readily assembled with a large number of branch termini, two types of iron-cluster-derivatized AuNPs have been prepared and their oxoanion redox sensing properties have been investigated. The first type contains an average of 285 Au atoms in the core and 8 iron-cluster-functionalized thiolates per AuNP (Fig. 6.9), while the second one contains a core of 101 Au atoms and 58 thiolate ligands including 28 Fe4-clusterfunctionalized thiolate ligands per AuNP.49 The electrochemical properties of the two types of iron-cluster-derivatized AuNPs resemble that of the monomeric tetrairon cluster, except that they additionally show adsorption due to their large size, and evidence that all the Fe4 clusters are active at about the same potential, thereby indicating that they are sufficiently remote from one another to behave independently. The changes in the cyclic voltammetric pattern caused by the addition of H2PO4 and ATP2 oxoanions evidence recognition features that are very different from those obtained with dendritic ferrocene exoreceptors. In particular, the results show that with these iron-cluster-derivatized AuNPs (i) the redox recognition of oxoanions is easier than with ferrocene dendrimers or AuNPs and (ii) ATP2 anions cause a larger CV wave shift than H2PO4 contrary to titrations carried out with ferrocene-based exoreceptors. These features evidence the advantage of the use of both AuNPs and the large tetrairon cluster for sensing over dendritic and ferrocene-based exoreceptors. Recently, it has been reported the synthesis and the electrochemical characterization of the homometallic dendrimers 10 and 11 terminated with four and eight ferrocene units, respectively, and of the related heterometallic dendrimers 12 and 13 terminated with four ferrocene and four (h6-C6H5)Cr(CO)3 moieties, and with eight ferrocene and eight (h6-C6H5)Cr(CO)3 moieties, respectively (Fig. 6.10). Dendrimer 13 contains also 4 additional (h6-C6H5)Cr(CO)3 moieties nonbonded to the ferrocene units so that its total number of organometallic moieties is 20.50 An important aspect of this work is to evaluate the redox properties of the new multimetallic dendritic molecules, not only in homogeneous solution but also confined onto electrode surfaces (i.e., where the molecules serve as electrode modifiers). For the homometallic dendrimers 10 and 11 the electrochemical results obtained in CH2Cl2 solution unequivocally demonstrate that the single reversible
6.3
METALLOCENE DENDRIMERS
161
Figure 6.9 Schematic representation of a gold nanoparticle containing an average of 285 Au atoms in the core and 8 iron-cluster-functionalized thiolates.49
oxidation wave observed represents the simultaneous multielectron transfer of four and eight electrons, respectively, as expected for independent, reversible oneelectron process, at the same potential, of the four and eight noninteracting ferrocene redox centers incorporated at the periphery of the two dendrimers. Furthermore, the values of the diffusion coefficients Do increase as the number of peripheral ferrocene moieties increases, that is, Do increases with increasing
Figure 6.10 Structural formulas of homometallic dendrimers 10 and 11 terminated with four and eight ferrocene units, respectively, and of the related heterometallic dendrimers 12 and 13 terminated with four ferrocene and four (h6-C6H5)Cr(CO)3 moieties, and with eight ferrocene and eight (h6-C6H5)Cr(CO)3 moieties, respectively.50
6.4 DENDRIMERS BASED ON POLYPYRIDINE METAL COMPLEXES
163
dendrimer generation. A key feature concerning the two ferrocene dendrimers 10 and 11 is their ability to deposit onto electrode surfaces as they become oxidized. Thus, modification of electrodes with films of the two dendrimers has been successful, resulting in detectable electroactive material persistently attached to the electrode surfaces. The electroactive dendrimer films behave almost ideally with rapid electron- and charge-transfer kinetics. The heterometallic dendrimers 12 and 13 show an electrochemical behavior that is consistent with their heterometallic nature. The cyclic voltammetry of the smaller dendrimer 12, containing at its periphery four ferrocene and four (h6-C6H5)Cr(CO)3 moieties, reveals two diffusion-controlled, reversible oxidation processes. The first oxidation process can be ascribed to the simultaneous oxidation of the four iron centers, and the second one to the simultaneous oxidation of the four chromium centers. The potential of the first oxidation process is slightly higher than that found for the oxidation of the iron centers in the related homometallic ferrocene dendrimers, and this is reasonably due to the electron-withdrawing nature of the adjacent (h6-C6H5)Cr(CO)3 moieties, bonded through a bridging silicon atom. In addition, the chromium-centered oxidation shows a slight anodic shift with respect to that of the (h6-C6H5)Cr(CO)3 model compound as a result of the positive charges generated after the first ferrocene-based oxidation process. The electrochemical behavior of their higher heterometallic dendrimer 13, that contains 20 organometallic units, has been better interpreted by using differential pulse voltammetry. Two separated peaks of different heights are observed. The first peak, taking place at less positive potential, is assigned to the simultaneous oxidation of the eight ferrocene units, whereas the second peak, that occurs at more positive potential, is considerably broadened and suggests the presence of two overlapped redox processes. By comparison with model compounds, these two overlapping processes can be attributed to the oxidation of the isolated (h6-C6H5)Cr(CO)3 moieties linked to the Si-allyl groups followed by the oxidation of the remaining chromium centers of the (h6-C6H5)Cr(CO)3 moieties neighboring the ferrocene moieties already oxidized.
6.4
DENDRIMERS BASED ON POLYPYRIDINE METAL COMPLEXES
Polypyridine metal complexes have been extensively investigated from an electrochemical point of view.51 The most famous and widely used complex of this family is [Ru(bpy)3]2 þ , in which bpy is 2,20 -bipyridine ligand. This compound is thermodynamically stable, kinetically inert, and shows outstanding electrochemical52 and photochemical51c properties. It exhibits a metal-centered reversible oxidation (below 2 V versus SCE) process in MeCN at room temperature and six distinct reversible ligand-centered reduction processes in dimethylformamide at 219 K.53 By substituting either the metal (Os(II), Co(II)) or the ligand, (e.g., terpyridine tpy), a variety of metal complexes have been obtained and used as building blocks of dendrimers. As in the case of [Ru(bpy)3]2 þ , the first oxidation process is metal centered and reductions are ligand centered. Metal complexes can be placed at the
164
DENDRIMERS AS MULTIELECTRON STORAGE DEVICES
core, in the branching points, and/or at the periphery of the dendritic structure. In the present case we will review examples containing multiple metal complexes, placed in (i) the periphery, (ii) branching points, or (iii) all over the dendritic structure. Metal Complexes at the Periphery Poly(amido amine) dendrimers (PAMAM) decorated at the periphery with [Ru(bpy)3]2 þ or [Ru(tpy)2]2 þ complexes have been synthesized and extensively investigated by Abrun˜a and Amatore.54 The most interesting electron transfer process to be exploited in charge storing devices is the first metal-centered oxidation process, that is, Ru(III)/Ru(II) couple, or the first ligand-centered reduction. In all the reported example a single cyclic voltammetric wave is observed, so that an electrochemical equivalency of the redox sites is inferred. For example, the larger investigated dendrimer is the fourth-generation PAMAM dendrimer decorated at the periphery with 64 [Ru(bpy)3]2 þ or [Ru(tpy)2]2 þ complexes (compounds 14 and 15 in Fig. 6.11).54a,d These dendrimers show one oxidation, due to Ru(III)/Ru(II) couple, and two reduction processes, attributed to the first and second reduction of the ligands. Scanning the potential to more negative values, adsorption onto Pt electrode occurs and charge trapping peaks are observed. These peaks likely arise from redox centers that are electronically isolated from the electrode surface, so that their redox reactions are mediated by adjacent redox sites. Morphological changes of the deposited film have been observed upon application of a potential difference and they have been attributed to the deposition or dissolution of the dendrimer and/or to ejection or adsorption of counterions and/or solvent into the film. Ultrafast voltammetry has been applied to study dendrimer 15 adsorbed onto a Pt disk ultramicroelectrode (5.0 0.5 mm).54b,c At slow scan rates (v < 0.1 MV/s), the peak current is proportional to v, as expected for an adsorbed electroactive species. At higher scan rates (v > 1MV/s), the diffusion layer is smaller than the dimension of the adsorbed dendrimer, and the peak current is proportional to v1/2, as expected for a system undergoing semi-infinite diffusion. Apparent diffusion by electron hopping has been characterized and, by these data, topological information on the manner in which the dendrimer adsorb on the electrode surface has been gained: dendrimers do not keep their globular shape, but have a hemispherical shape on the electrode surface. Moreover, the rate of self-exchange electron transfer for the pendant [Ru(tpy)2]2 þ complexes has been measured; the value (k 109 M1 s1) is similar to that of sites in close contact in solution, demonstrating that the redox centers are extremely mobile around their equilibrium position. Dendrimer 15 was the object of a very interesting theoretical work by Amatore et al.55 discussing the possibility to observe a stochastic behavior of electrochemical events. In the limiting case of only one dendrimer 15 adsorbed on the electrode surface, the current measured in a chronoamperometric experiment would show a random and discontinuous succession of single electron transfer events, while for an array of 7800 dendrimers, corresponding to complete coverage of an electrode of 500 nm in radius, the stochastic nature of the phenomenon is no longer clearly discernable.
6.4 DENDRIMERS BASED ON POLYPYRIDINE METAL COMPLEXES
165
Figure 6.11 Structural formulas of the fourth-generation dendrimers terminated with 64 metal complexes: [Ru(bpy)3]2 þ (14),54a,d [Ru(tpy)2]2 þ (15),54 and [Co(tpy)2]2 þ (16).56
PAMAM dendrimers from the first up to the fourth generation (Compound 16 in Fig. 6.11) containing from eight up to 64 [Co(tpy)2]2 þ complexes appended at their periphery show two reversible redox processes at þ 0.30 and 0.75 V versus Ag/AgCl, ascribed to the Co(III)/Co(II) and Co(II)/Co(I) processes, respective-
166
DENDRIMERS AS MULTIELECTRON STORAGE DEVICES
ly.56 The presence of a single wave with a one-electron waveshape for both processes suggests that all the redox centers are electrochemically equivalent. The peak current of these two processes is the same only for a freshly polished electrode. On repeating scans, dendrimers are deposited onto the electrode and, when a stationary value of current is reached, the peak eights of the Co(II)/Co(I) redox process is much higher than those of Co(III)/Co(II). This behavior can be explained by a faster rate of electron self-exchange of the Co(II)/Co(I) couple than for Co(III)/Co(II). Dendrimer containing different types of redox-active units may be very interesting as charge storing devices because they can show both reduction and oxidation processes at easy accessible potentials by coupling good electron acceptor and good electron donor moieties. A peculiar example of this class is represented by a family of dendrimers containing a tris-viologen core appended with three or six [Ru(bpy)3]2 þ complexes (Compound 17 in Fig. 6.12) in the first and second dendrimer generation, respectively.57 Two reversible electron transfer processes are observed at þ 1.29 and 0.27 V versus Ag/AgCl for 17 in 0.1 M aqueous NaCl solution: the oxidation process is metal centered, while the reduction one is attributed to the first reduction of the viologen
Figure 6.12 Structural formula of dendrimer 17 containing a tris-viologen core appended with six [Ru(bpy)3]2 þ complexes.57
6.4 DENDRIMERS BASED ON POLYPYRIDINE METAL COMPLEXES
167
units. The peak current corresponding to the oxidation process is twice of that of the reduction process in the case of 17 in accordance with the number of Ru(II) complexes and viologen units. Therefore, the six [Ru(bpy)3]2 þ units are electrochemically equivalent, as well as the three viologen ones. Scanning the potential to more negative values resulted in the precipitation of the dendrimers on the electrode surface. Metal Complexes in the Branching Points Dendrimers containing [Ru(tpy)2]2 þ complexes as branching points have been extensively investigated.58 In particular, dendrimers having the same chemical formula and different position of the component units along their architecture have been synthesized by Newkome.58 They constitute the first example of dendritic constitutional isomers. Differences in internal density and void regions can be probed by cyclic voltammetry. For example, two dendritic constitutional isomers containing four [Ru(tpy)2]2 þ complexes in their branching points (Compounds 18 and 19 in Fig. 6.13) exhibit similar solubility and thermal stability, but slightly different electrochemical behavior.58b,c Dendrimer 18 exhibits two distinct oxidation processes with a 3:1 ratio of the height peak, while 19 shows a single oxidation process. Two reduction processes are observed for both isomers at similar potential values and involve the two terpyridine ligands of each complex. The observed differences for 18 and 19 in the case of the metal-centered oxidations suggest the presence of different chemical environments in the two isomers. In particular, a denser branched structure surrounds [Ru(tpy)2]2 þ complexes in dendrimer 18, causing a different solvent and counterion accessibility. The oxidation pattern (3:1) observed in 18 has been tentatively attributed by the authors to the fact that only three out of four Ru(II) centers are in close contact with the electrode surface, while the fourth Ru(II) center is oriented away from the electrode. However, the high flexibility of this dendritic structure and the slow scan rate (v ¼ 0.2 V/s) employed in the cyclic voltammetry should allow motion of the dendritic branches within the electrochemical timescale. Based on the reported results, dendrimer 19 is a better candidate for charge pooling devices since the four redox centers are electrochemically equivalent and four-electron transfers are observed both in the cathodic and anodic scans. Another interesting class of metallodendrimers are zwitterionic dendrimers in which the positive charge on the metal ion is balanced by carboxylate moieties present in the dendritic branches.58a,b Dendrimers 20 and 21 (Fig. 6.14) that contain four [Ru(tpy)2]2 þ complexes along the branches and eight carboxylate units either in the internal branching points or at the periphery are examples of this class of zwitterionic dendrimers. These carboxylate ions are weaker counterions than Cl and PF6, so that the corresponding dendrimers tend to ionize more easily and to give clearer signals in MALDI-TOF mass spectra. As to the electrochemical properties, their cyclic voltammetries show a reversible metal centered oxidation and two reversible ligand-centered reduction processes at potential values very similar to those of the corresponding dendrimers with Cl counterions. Therefore, the [Ru(tpy)2]2 þ complexes are electrochemically equivalent and can efficiently store charges.
Figure 6.13 Structural formulas of dendrimers 18 and 19 containing four [Ru(tpy)2]2 þ complexes in their branching points that represent the first example of dendritic constitutional isomers.58
6.4 DENDRIMERS BASED ON POLYPYRIDINE METAL COMPLEXES
169
Figure 6.14 Structural formulas of zwitterionic dendrimers that contain four [Ru(tpy)2]2 þ complexes along the branches and eight carboxylate units either in the internal branching points (20) or at the periphery (21).58a,b
Metal Complexes all Over the Dendritic Structure Within this family the more carefully investigated dendrimers are those containing Ru(II) and Os(II) as metal ions (M), 2,3- or 2,5-bis(2-pyridyl)pyrazine (2,3- and 2,5-dpp) as bridging ligands (BL), and 2,20 -bipyridine (bpy) or 2,20 -biquinoline (biq) as terminal ligands (L) (Fig. 6.15).59,8 The typical strategy used to prepare these dendrimers is the so-called “complexes as metals and complexes as ligands” approach,59 which has enabled the construction of species containing 4, 6, 10, 13, and 22 metal-based units. The knowledge of the electrochemical properties of the mononuclear component [M(L)n(BL)3n]2 þ units and the synthetic control of the supramolecular structure enable the design of dendrimers with predetermined redox patterns.
170
DENDRIMERS AS MULTIELECTRON STORAGE DEVICES
Figure 6.15 Structural formulas, abbreviations, and graphic symbols used to represent the components of dendrimers 22–25.
For each metal complex, only one, metal-based, process is observed in the anodic scan within the potential window usually investigated (< þ 2 V versus SCE). This process occurs at a potential which depends strongly on the nature of the metal ion (Os(II) is oxidized at less positive potentials compared to Ru(II)) and, less dramatically, on the nature of the coordinated ligands, whose electron donor power increases in the series m-2,5-dpp < m-2,3-dpp < biq < bpy (alternatively, one can say that the electron acceptor power decreases along the same series). The reduction processes are ligand localized. The reduction potential of each ligand depends on its electronic properties and, to a smaller extent, on the nature of the metal and the other ligands coordinated to the metal. The first reduction potential becomes more negative by increasing the electron donor power of the ligands. Each L ligand is reduced twice and each BL ligand is reduced four times, when coordinated to a metal ion, in the potential window 0.5/3.1 V versus SCE.60 In the dendritic species, each unit brings its own redox properties, more or less affected by intercomponent interactions. Metal–metal and ligand–ligand interactions are noticeable for metals coordinated to the same bridging ligand and for ligands coordinated to the same metal, whereas they are very small for metals or ligands that are sufficiently far apart. By placing in the dendrimer the desired number of suitable equivalent and noninteracting units it is possible to control the number of electrons lost or gained at a certain potential. To illustrate the most interesting electrochemical features of this family of dendrimers, it is reported the results obtained for three strictly related decanuclear compounds and the highest nuclearity compound of this dendrimer family, that is, a docosanuclear complex.
6.4 DENDRIMERS BASED ON POLYPYRIDINE METAL COMPLEXES
171
Figure 6.16 Redox patterns (differential pulse voltammetric peaks) for decanuclear dendrimers 22–24.61b,63 Peak Fc indicates the oxidation of ferrocene as an internal standard. The symbols used to represent the dendrimers are those displayed in Fig. 6.15.
Very interesting oxidation patterns have been obtained in MeCN for the series of decanuclear dendrimers shown in Fig. 6.16. In the case of dendrimer 22, [Ru{(m-2,3-dpp)Ru[(m-2,3-dpp)Ru(bpy)2]2}3]20 þ (Fig. 6.16a),61 the first oxidation process ( þ 1.53 V versus SCE) involves the exchange of six electrons at the same potential and is attributed to the peripheral Ru-based units which are oxidized at less positive potentials than the internal ones because the bpy ligands are worse electron acceptors than the bridging 2,3-dpp ligands. Furthermore, the six peripheral Ru-based units are not expected to interact with one another, because they are not directly connected. Successive oxidation of the
172
DENDRIMERS AS MULTIELECTRON STORAGE DEVICES
central Ru-based unit followed by oxidation of the three intermediate Ru-based units is observed in liquid sulfur dioxide.62 In the [Os{(m-2,3-dpp)Ru[(m-2,3-dpp)Ru(bpy)2]2}3]20 þ dendrimer 23 (Fig. 6.16b),61b,63 which is made of an Os(II)-based core and nine Ru(II)-based units, a one-electron process at þ 1.35 V and a six-electron process at þ 1.55 V (versus SCE) are observed. Indeed, the Os(II) ion is expected to be oxidized at less positive potentials than the nine Ru(II) ions. Furthermore, because of the different electron donor properties of the ligands, the six peripheral Ru(II) ions are expected to be oxidized at less positive potentials than the three intermediate Ru(II) ions. Since the core and the peripheral units are not directly connected, they are not expected to affect one another. Therefore, a 1:6 pattern is observed for the electrons exchanged on oxidation. Oxidation of the three intermediate Ru(II) ions is further shifted toward more positive potentials and cannot be observed in the potential window accessible in MeCN solution. For the [Os{(m-2,3-dpp)Ru[(m-2,3-dpp)Os(bpy)2]2}3]20 þ dendrimer 24 (Fig. 6.16c),61b made of an Os(II)-based core, three Ru(II)-based units in the intermediate positions, and six Os(II)-based units in the peripheral positions, a six-electron process at þ 1.05 V and a one-electron process at þ 1.39 V (versus SCE) is observed. Indeed, the first oxidation involves the six peripheral Os(II) ions (which contain the weaker electron acceptor bpy ligand in their coordination sphere), and then the central one, yielding for the electron exchange a 6:1 pattern instead than the 1:6 one observed in the case of 23. Oxidation of the intermediate Ru(II)-based units is not observed in the potential window accessible in MeCN solution. In larger dendrimers, the number of equivalent units becomes huge and a variety of electron exchanged patterns can be expected. In the docosanuclear dendrimer 25 (Fig. 6.17) made of an Os(II)-based core and 21 Ru(II)-based units, a one-electron
Figure 6.17 Structural formula of the docosanuclear dendrimer 25.63 The symbols used to represent the dendrimer are those displayed in Fig. 6.15.
6.5
VIOLOGEN-BASED DENDRIMERS
173
oxidation process at þ 1.42 V (versus SCE), assigned to the oxidation of the Os(II) ion, is followed by a 12-electron process at þ 1.54 V (versus SCE), due to the simultaneous oxidation of the 12 equivalent and noninteracting peripheral Ru(II)based units.63 The situation is even more complex as far as reduction is concerned, because each ligand can exchange at least two electrons. These dendrimers are therefore optimal candidates as charge storing devices because of the made-to-order synthetic control of the reversible exchange (storage and release) of a controlled number of electrons at a certain potential. It is worth noting that the electrochemical data offer also a fingerprint of the chemical and topological structure of the dendrimers.
6.5
VIOLOGEN-BASED DENDRIMERS
4,40 -Bipyridinium-type units (also known as viologens) are well-known electron acceptors64 extensively used in chemical and electrochemical redox processes,65 since they can undergo two reversible one-electron reduction processes. Because of these peculiar properties such units can be profitably used to functionalize the periphery of dendrimers, but examples of dendrimers containing a bipyridiniumtype unit as a core are also reported.66 4,40 -Bipyridinium Units All Over the Dendritic Structure Polyviologen dendrimers are particularly interesting because, in principle, they can behave as molecular batteries20d,23 being potentially capable of storing, at easy accessible potentials, a number of electrons twice that of the viologen units. In agreement with this expectation, it has been found that a series of dendrimers containing up to 45 electrochemically accessible 4,40 -bipyridinium units and terminated with methyl groups exhibit electron sponge properties. They can indeed store a number of electrons that is practically twice that of the viologen units present in the dendrimers.67 More recently, however, a different behavior has been observed for two families of very similar dendrimers (Fig. 6.18). These two families of dendrimers, both containing a 1,3,5-trisubstituted benzenoide core and 9 (A918 þ and B918 þ ) and 21 (A2142 þ and B2142 þ ) 4,40 -bipyridinium units in their branches, differ for the peripheral groups. The A-type dendrimers are terminated with tetraarylmethane bulky moieties while the B-type dendrimers contain in their periphery less bulky aryloxy groups.68 Electrochemical experiments performed on the two types of dendrimers in argonpurged MeCN solutions have revealed that in all cases only a fraction of the viologen units can be reduced, in agreement with previous results obtained for viologen groups appended at the periphery of poly(amido amine) dendrimers.69 Within experimental error, this fraction corresponds to the number of the viologen units present in the outer shell (6 for A918 þ and B918 þ , and 12 for A2142 þ and B2142 þ ). The fact that the number of reducible viologen units is smaller than that expected cannot be attributed to lack of branches in the structures of the dendrimers because the 1 H NMR spectroscopic and mass spectral characterization show clearly that the compounds
174
DENDRIMERS AS MULTIELECTRON STORAGE DEVICES
Figure 6.18 Structural formulas of two families of dendrimers containing a 1,3,5-trisubstituted benzenoide core and 9 (A918 þ and B918 þ ) and 21 (B2142 þ and A2142 þ ) 4,40 bipyridinium units in their branches.68 The A-type dendrimers are terminated with tetraarylmethane bulky moieties while the B-type dendrimers contain in their periphery less bulky aryloxy groups. For comparison purposes the structural formulas of dendrons A24 þ and B24 þ and model compound DoV2 þ are also reported.
6.5
VIOLOGEN-BASED DENDRIMERS
175
examined do contain 9 (A918 þ and B918 þ ) and 21 (A2142 þ and B2142 þ ) viologen units. The electrochemical reduction experiments have also shown that in each dendrimer, the first reduction process of all its reducible viologen units occurs at the same potential and this value is not affected by dendrimer generation. An interesting result is that the reduction potential of a reducible unit is not affected by the state of the other reducible units, which is an ideal property for a charge pooling system. Another result that emerges clearly from the data obtained is that the nature of the peripheral groups—tetraarylmethane and aryloxy units for the A-type and B-type families, respectively—affects neither the number of reducible viologen units nor the first reduction potential. Two concomitant effects can be taken into account to explain the lack of complete electrochemical reduction. They are: (i) upon reduction of the external viologen shell, the structure of the dendrimer shrinks, favored by dimerization of the reduced units ( a well-known process called pimerization64 and already observed for the viologen-based dendrimers previously studied.67,69), thereby preventing the internal viologen units “seeing” the electrode, and (ii) the internal viologens being engaged in tight ionic couples with the hexafluorophosphate counterions become more difficult to reduce, thereby preventing electron hopping from external to internal viologen units. Photosensitized reduction experiments, performed on the two families of dendrimers in CH2Cl2 by using 9-methylanthracene as a photosensitizer and triethanolamine as a sacrificial reductant, have shown that the numbers of viologen units reducible under the photochemical conditions are in reasonable agreement with those obtained by electrochemical experiments, confirming that only the viologens in the external shells can be reduced. The photosensitized reduction experiments also reveal that formation of the one-electron reduced viologen units is accompanied by their dimerization. The lack of reduction of all the viologen units in such experiments can be explained by considering (i) that the uncharged photosensitizer cannot displace the counterions assembled in the proximity of the highly charged core of the dendrimer and (ii) that the interaction between the photosensitizer and the internal viologen units can be prevented by the dimerization of the external reduced units, a phenomenon which shrinks the dendrimer structure. Interestingly, this kind of studies evidenced that dimerization is not a strongly favored process, as shown by the fact that it does not occur for the monoviologen DoV2 þ species (Fig. 6.18) and leads to only 15–20% of associated species in the case of dendrons B24 þ and A24 þ (Fig. 6.18) that are structurally preorganized to afford dimerization. In the case of the dendrimers, the fraction of dimerized viologen units is higher—by approximately 45% and 57% for the A-type and B-type families, respectively—than that expected because of the possibility to produce dimers also between reduced units belonging to different dendrons. Moreover, the results show that the bulky tetraarylmethane peripheral moieties disfavor the formation of dimers compared with the smaller aryloxy groups. It is also interesting to notice that these two families of polyviologen dendrimers can also act as polytopic receptors toward electron donor substrates, hosting a number of eosin dianions equal to the number of viologen units present in their branches. The data
176
DENDRIMERS AS MULTIELECTRON STORAGE DEVICES
Figure 6.19 Structural formula of dendrimer 26 containing a hexaphenylbenzene core and 24 C60 units.74
6.6
177
FULLERENE DENDRIMERS
obtained show clearly that the host–guest interactions, that drive complex formation in low polar media, are not affected by either the steric hindrance of the terminal groups or the electronic interactions that these groups establish with the viologen units. These results are of interest for the design and construction of dendrimers capable of performing functions related to drug delivery or sensing applications.
6.6
FULLERENE DENDRIMERS
Fullerene is an ideal candidate as a component of molecular batteries because it shows six chemically and electrochemically reversible, one-electron reduction70 and one oxidation process.71 In particular, the first reduction process occurs at easy accessible potentials (0.98 V versus Fc þ /Fc in MeCN/toluene solution at 263 K)70 and it is thus the most suitable process to exploit in charge storing devices. To covalently append fullerene to the dendritic structure, chemical functionalization of the buckyball is necessary. Fortunately, most of its derivatives keep the reversible electrochemical properties of C60, at least for the first reduction process, which usually occurs at more negative potentials than that of fullerene. Nierengarten et al. have reported many examples of fullerene-rich dendrimers, obtained by either covalent linking or self-assembly of proper units.72,73 A very interesting example is represented by a family of fullerodendrimers from the first to the third generation, containing a hexaphenylbenzene core and 6, 12, and 24 C60 units (compound 26 in Fig. 6.19) at the periphery, respectively.74 The cyclic voltammograms show three reduction waves in the potential window between 0 and 1.3 V versus SCE (Fig. 6.20). The first process is fully reversible, while the second one is chemically irreversible since the dianion is subject to a bond breaking of the cyclopropane ring, known as the retro-Bingel reaction.75 The third
(a)
(b) 0.03
1 I / μA
I / μA 0
0 –0.03 –0.06
–1
–0.09 –0.8
–0.4
E/V versus SCE
0
–1.2
–0.8
–0.4
0
E/V versus SCE
Figure 6.20 Cyclic voltammograms of the first (a) at 8000 V/s and second (b) at 140 V/s generation of dendrimers containing a hexaphenylbenzene core and 6 and 12 C60 units, respectively (THF in the presence of 0.3 M of tetrabutylammonium hexafluorophosphate as supporting electrolyte).74 Reproduced with permission from Ref. 74.
178
DENDRIMERS AS MULTIELECTRON STORAGE DEVICES
process corresponds to the reduction of the product formed after the second reduction of the starting compound. For the third-generation dendrimer 26 the first reduction wave is symmetrically bell-shaped and the peak current is proportional to the scan rate (for v < 400 V/s), demonstrating that 26 is adsorbed on the electrode surface. From the area under the peak and the dendrimer area, a bilayer is estimated to be deposited onto the electrode. Upon increasing scan rate (v > 104 V/s), only a fraction of the C60 units are reduced during the forward scan because only part of the fullerene moieties are close enough to the surface to allow direct electron transfer. On the other hand, at lower scan rates, all the fullerene redox sites are accessible and thus they can store charges.
Figure 6.21 Structural formula of dendron 27 functionalized with tris-isothiocyanate core to be anchored onto gold electrodes.76
6.6
FULLERENE DENDRIMERS
179
Compound 27 (Fig. 6.21) has been functionalized with tris-isothiocyanate core and anchored onto gold electrodes.76 Detailed cyclic voltammetric studies reveal that less than a monolayer is adsorbed and an efficient electron transfer from the electrode to the fullerene moieties takes place through space since they can get close to the electrode. Another example of fullerene-rich dendrimers consist of a Cu(I) bis-phenanthroline complex as core, linked to four dendrons, each one containing 1, 2 (compound 28 in Fig. 6.22), or 4 fullerene moieties appended at the periphery.77 The first reduction process occurring at 1.07 V versus Fc þ /Fc is reversible: the peripheral fullerene moieties behave independently from an electrochemical point of view in this dendrimer family. On the other hand, the first oxidation process is monoelectronic and it is centered on the Cu(I) complex: the rate of heterogeneous electron transfer decreases upon increasing dendrimer generation.
Figure 6.22 Structural formula of dendrimer 28 consisting of a Cu(I) bis-phenanthroline complex as core, linked to four dendrons, each one containing two fullerene moieties appended at the periphery.77
180
6.7
DENDRIMERS AS MULTIELECTRON STORAGE DEVICES
CONCLUSIONS
In this chapter, which for reasons of space is far to be exhaustive, it has been shown that dendrimers are ideal scaffolds to construct multielectron storage devices because of their peculiar features: (i) a large number of suitable redox units can be placed in the periphery and/or branching points of their structure with the possibility of tuning their distance, (ii) the dendrimer skeleton can be designed to minimize electronic interaction between the redox centers, and (iii) the dendrimer periphery can be optimized to get the desired solubility properties and processability to eventually deposit the dendrimers on a surface. For these features dendrimers containing multiple redox-active units become to find useful applications in catalytic processes, in functionalization of electrodes that can also work as sensors, and in fabrication of nanoparticles. They can also be used as electrochemical sensors with signal amplification, and molecular batteries that can be foreseen to power molecular machines in the future or that can be used to construct flexible rechargeable batteries. Another important aspect that emerges from the studies carried out on dendrimers is the power of electrochemistry as a tool for (i) elucidating their structure (and purity), a task that is not at all easy with these highly branched and sometimes highly charged compounds, (ii) evaluating the degree of electronic interaction among the various, chemically and/or topologically equivalent or nonequivalent moieties incorporated in the dendritic array, and (iii) investigating their endo- and exoreceptor capabilities.
REFERENCES 1. (a) F. V€ogtle, G. Richardt, N. Werner, Dendritische Molek€ ule, Teubner Studienb€ ucher Chemie, Wiesbaden, 2007. (b) G. R. Newkome, F. V€ ogtle, Dendrimers and Dendrons, Wiley-VCH, Weinheim, 2001. (c) J. M. J. Frechet, D. A. Tomalia (Eds.), Dendrimers and Other Dendritic Polymers, John Wiley & Sons, Chichester, 2001. 2. For reviews, see: (a) M. Gingras, J.-M. Raimundo, Y. M. Chabre, Angew. Chem., Int. Ed. 2007, 46, 1010. (b) C. A. Schalley, B. Baytekin, H. T. Baytekin, M. Engeser, T. Felder, A. Rang, J. Phys. Org. Chem. 2006, 19, 479. (c) T. Darbre, J.-L. Reymond, Acc. Chem. Res. 2006, 39, 925. (d) D. Astruc, F. Chardac, Chem. Rev. 2001, 101, 2991. (e) M. W. P. L. Baars, E. W. Meijer, Top. Curr. Chem. 2000, 210, 131. 3. For some recent examples, see e.g.: (a) V. Vicinelli, G. Bergamini, P. Ceroni, V. Balzani, F. V€ogtle, O. Lukin, J. Phys. Chem. B 2007, 111, 6620. (b) M. Chai, A. K. Holley, M. Kruskamp, Chem. Commun. 2007, 168. (c) R. van Heerbeek, P. C. J. Kamer, P. N. M. W. van Leeuwen, N. H. Reek, Org. Biomol. Chem. 2006, 4, 211. (d) R. M. Versteegen, D. J. M. van Beek, R. P. Sijbesma, D. Vlasspoulos, G. Fytas, E. W. Mejer, J. Am. Chem. Soc. 2005, 127, 13862. 4. (a) J.-P. Majoral (Ed.), New J. Chem. 2007, 31 (Special Issue on Dendrimers). (b) D. Mery, D. Astruc, Coord. Chem. Rev. 2006, 250, 1965. (c) D. A. Tomalia, J. M. J. Frechet (Eds.), Prog. Polym. Sci. 2005, 30 (3–4) (Special Issue on Dendrimers and Dendritic Polymers). (d) R. W. J. Scott, O. M. Wilson, R. M. Crooks, J. Phys. Chem. B 2005, 109, 692.
REFERENCES
5. 6. 7. 8. 9. 10. 11. 12.
13. 14. 15. 16. 17. 18. 19.
20.
21. 22.
23. 24. 25.
181
(e) P. A. Chase, R. J. M. Klein Gebbink, G. van Koten, J. Organomet. Chem. 2004, 689, 4016. (f) W. Ong, M. Gomez-Kaifer, A. E. Kaifer, Chem. Commun. 2004, 1677. (g) M. Ballauff, C. N. Likos, Angew. Chem., Int. Ed. 2004, 43, 2998. (h) A.-M. Caminade, J.-P. Majoral, Acc. Chem. Res. 2004, 37, 341. For some recent reviews, see: (a) D. Astruc, C. Ornelas, J. Ruiz, Acc. Chem. Res. 2008, 41, 841. (b) A. E. Kaifer, Eur. J. Inorg. Chem. 2007, 5015. P. A. Chase, R. J. M. K. Gebbink, G. van Koten, J. Organomet. Chem. 2004, 689, 4016. D. K. Smith, F. Diederich, Top. Curr. Chem. 2000, 210, 183. V. Balzani, S. Campagna, G. Denti, A. Juris, S. Serroni, M. Venturi, Acc. Chem. Res. 1998, 31, 26. H. D. Abrun˜a, Anal. Chem. 2004, 310A. H. Nishide, K. Oyaizu, Science 2008, 319, 737. For example, see: (a) K.Oyaizu, Y. Ando, H. Konishi, H. Nishide, J. Am. Chem. Soc. 2008, 130, 14459. (b) S. I. Cho, S. B. Lee, Acc. Chem. Res. 2008, 41, 699. For example, see: (a) A. Budny, F. Novak, N. Plumere, B. Schetter, B. Speiser, D. Straub, H. A. Mayer, M. Reginek, Langmuir 2006, 22, 10605. (b) X. Y. Ling, D. N. Reinhouldt, J. Huskens, Langmuir 2006, 22, 8777. J. B. Flanagan, S. Margel, A. J. Bard, F. C. Anson, J. Am. Chem. Soc. 1978, 100, 4248. A. J. Bard, L. R. Faulkner, Electrochemical Methods–Fundamentals and Application, 2nd ed., John Wiley & Sons, New York, 2000. H. Sun, W. Chen, A. E. Kaifer, Organometallics 2006, 25, 1828. G. Denuault, M. V. Mirkin, A. J. Bard, J. Electroanal. Chem. 1991, 308, 27–38. A. Juris,in V. Balzani (Ed.), Electron Transfer in Chemistry, Vol. 3, Wiley-VCH, 2001, p. 655. G. R. Newkome, Z. Yao, G. R. Baker, V. K. Gupta, J. Org. Chem. 1985, 50, 2003. (a) K. Takada, D. J. Dıaz, H. D. Abrun˜a, I. Cuadrado, C. Casado, B. Alonso, M. Moran, J. Losada, J. Am. Chem. Soc. 1997, 119, 10763. (b) R. Castro, I. Cuadrado, B. Alonso, C. Casado, M. Moran, A. E. Kaifer, J. Am. Chem. Soc. 1997, 119, 5760. (c) C. M. Casado, I. Cuadrado, M. Moran, B. Alonso, M. Barranco, J. Losada, Appl. Organometal. Chem. 1999, 13, 245. (d) I. Cuadrado, M. Moran, C.M. Casado, B. Alonso, J. Losada, Coord. Chem. Rev. 1999, 189, 123. (a) C. Valerio, J.-L. Fillaut, J. Ruiz, J. Guittard, J.-C. Blais, D. Astruc, J. Am. Chem. Soc. 1997, 119, 2588. (b) V. Sartor, L. Djakovitch, J.-L. Fillaut, F. Moulines, F. Neveu, V. Marvaud, J. Guittard, J.-C. Blais, D. Astruc, J. Am. Chem. Soc. 1999, 121, 2929. (c) S. Nlate, J. Ruiz, J.-C. Blais, D. Astruc, Chem. Commun. 2000, 417. (d) S. Nlate, J. Ruiz, V. Sartor, R. Navarro, J.-C. Blais, D. Astruc, Chem. Eur. J. 2000, 6, 2544. C. B. Gorman, J. C. Smith, B. L. Parkhurst, H. Sierzputowska-Gracz, C. A. Haney, J. Am. Chem. Soc. 1999, 121, 9958. A. K. Diallo, J.-C. Daran, F. Varret, J. Ruiz, D. Astruc, Fast electron transfer has been recently observed also for rigid dendrimers containing ferrocene units at the periphery, Angew. Chem., Int. Ed. 2009, 48, 3141. J. Ruiz, C. Pradet, F. Varret, D. Astruc, Chem. Commun. 2002, 1108. C. Ornelas, J. Ruiz, C. Belin, D. Astruc, J. Am. Chem. Soc. 2009, 131, 590. D. Astruc, C. Valerio, J.-L. Fillaut, J. Ruiz, J.-R. Hamon, F. Varret,in O. Kahn (Ed.), Supramolecular Magnetism, NATO ASAI Series, Kluwer, Dordrecht, 1996, p. 107.
182
DENDRIMERS AS MULTIELECTRON STORAGE DEVICES
26. S. R. Miller, D. A. Gustowsi, Z.-h. Chen, G. W. Gokel, L. Echegoyen, A. E. Kaifer, Anal. Chem. 1988, 60, 2021. 27. D. Mery, L. Plault, C. Ornelas, J. Ruiz, S. Nlate, D. Astruc, J.-C. Blais, J. Rodrigues, S. Cordier, K. Kirakci, C. Perrin, Inorg. Chem. 2006, 45, 1156. 28. D. Astruc, M.-C. Daniel, J. Ruiz, Chem. Commun. 2004, 2637. 29. M.-C. Daniel, J. Ruiz, S. Nlate, J.-C. Blais, D. Astruc, J. Am. Chem. Soc. 2003, 125, 2617. 30. C. Ornelas, J. Ruiz Aranzaes, E. Cloutet, S. Alves, D. Astruc, Angew. Chem., Int. Ed. 2007, 46, 872. 31. C. Ornelas, J. Ruiz Aranzaes, L. Salmon, D. Astruc, Chem. Eur. J. 2008, 14, 50. 32. J. Camponovo, J. Ruiz, E. Cloutet, D. Astruc, Chem. Eur. J. 2009, 15, 2990. 33. I. Dıaz, B. Garcıa, B. Alonso, C. M. Casado, M. Moran, J. Losada, J. Perez-Pariente, Chem. Mater. 2003, 15, 1073. 34. (a) J. Ruiz, J. Ruiz Medel, M. C. Daniel, J.-C. Blais, D. Astruc, Chem. Commun. 2003, 464. (b) M. C. Daniel, J. Ruiz, J.-C. Blais, N. Daro, D. Astruc, Chem. Eur. J. 2003, 9, 4371. 35. A. Hradsky, B. Bildstein, N. Schuler, H. Schottenberger, P. Jaitner, K.-H. Ongania, K. Wurst, J.-P. Launay, Organometallics 1997, 16, 392. 36. (a) Z. Chaofeng, M. S. Wrighton, J. Am. Chem. Soc. 1990, 112, 7578. (b) S. Chao, J. L. Robbins, M. S. Wrighton, J. Am. Chem. Soc. 1983, 105, 181. 37. (a) M. P. Garcıa-Armada, J. Losada, M. Zamora, B. Alonso, I. Cuadrado, C.M. Casado, Bioelectrochemistry 2006, 69, 65. (b) J. Losada, M. Zamora, M. P. Garcıa-Armada, I. Cuadrado, B. Alonso, C. M. Casado, Anal. Bioanal. Chem. 2006, 385, 1209. 38. M. Zamora, S. Herrero, J. Losada, I. Cuadrado, C. M. Casado, B. Alonso, Organometallics 2007, 26, 2688. 39. A. Salmon, P. Jutzi, J. Organomet. Chem. 2001, 637639, 595. 40. C. A. Nijhuis, K. A. Dolatowska, B. Jan Ravoo, J. Huskens, D. N. Reinhoudt, Chem. Eur. J. 2007, 13, 69. 41. K. Takada, D. J. DIaz, H. D. Abrun˜a, I. Cuadrado, B. Gonzalez, C. M. Casado, B. Alonso, M. Moran, J. Losada, Chem. Eur. J. 2001, 7, 1109. 42. I. Cuadrado, M. Moran, C. M. Casado, B. Alonso, F. Lobete, B. Garcıa, M. Ibisate, J. Losada, Organometallics 1996, 15, 5278. 43. B. Gonzalez, C. M. Casado, B. Alonso, I. Cuadrado, M. Moran, Y. Wang, A. E. Kaifer, Chem. Commun. 1998, 2569. 44. C. M. Casado, B. Gonzalez, I. Cuadrado, B. Alonso, M. Moran, J. Losada, Angew. Chem., Int. Ed. 2000, 39, 2135. 45. J. Ruiz Aranzaes, C. Belin, D. Astruc, Angew. Chem., Int. Ed. 2006, 45, 132. 46. J. A. Ferguson, T. J. Meyer, J. Am. Chem. Soc. 1972, 94, 3409. 47. E. Alonso, J. Ruiz, D. Astruc, C. R. Acad. Sci. Ser. IIc 2001, 3, 189. 48. G. E. Oosterom, J. N. H. Reek, P. C. J. Kamer, P. W. N. M. van Leeuwen, Angew. Chem., Int. Ed. 2001, 40, 1828. 49. J. Ruiz Aranzaes, C. Belin, D. Astruc, Chem. Commun. 2007, 3456. 50. M. Zamora, B. Alonso, C. Pastor, I. Cuadrado, Organometallics 2007, 26, 5153. 51. (a) A. A. Vlcek, Coord. Chem. Rev. 1999, 200–202, 979. (b) V. Balzani, A. Juris, M. Venturi, S. Campagna, S. Serroni, Chem. Rev. 1996, 97, 759. (c) A. Juris, V. Balzani, F. Barigelletti, S. Campagna, P. Belser, A. von Zelewsky, Coord. Chem. Rev. 1988, 84, 85.
REFERENCES
183
52. (a) A. A. Vlcek, Coord. Chem. Rev. 1982, 43, 39. (b) M. K. De Armond, C. M. Carlin, Coord. Chem. Rev. 1981, 36, 325. 53. (a) S. Roffia, R. Casadei, F. Paolucci, C. Paradisi, C. A. Bignozzi, F. Scandola, J. Electroanal. Chem. 1991, 302, 157. 54. (a) J. I. Goldsmith, K. Takada, H. D. Abrun˜a, J. Phys. Chem. B 2002, 106, 8504. (b) C. Amatore, Y. Bouret, E. Maisonhaute, J. I. Goldsmith, H. D. Abrun˜a, ChemPhysChem 2001, 5, 130. (c) C. Amatore, Y. Bouret, E. Maisonhaute, J. I. Goldsmith, H. D. Abrun˜a, Chem. Eur. J. 2001, 7, 2206. (d) G. D. Storrier, K. Takada, H. D. Abrun˜a, Langmuir 1999, 15, 872. 55. C. Amatore, F. Gr€un, E. Maisonhaute, Angew. Chem., Int. Ed. 2003, 42, 4944. 56. K. Takada, G. D. Storrier, J. I. Goldsmith, H. D. Abrun˜a, J. Phys. Chem. B 2001, 105, 2404. 57. M. A. Saab, R. Abdel-Malak, J. F. Wishart, T. H. Ghaddar, Langmuir 2007, 23, 10807. 58. (a) G. R. Newkome, K. Soo Yoo, H. Jin Kim, C. N. Moorfield, Chem. Eur. J. 2003, 9, 3367. (b) G. R. Newkome, E. He, L. A. Godınez, G. R. Baker, J. Am. Chem. Soc. 2000, 122, 9993. (c) G. R. Newkome, E. He, L. A. Godınez, Macromolecules 1998, 31, 4382. 59. (a) F. Puntoriero, F. Nastasi, M. Cavazzini, S. Quici, S. Campagna, Coord. Chem. Rev. 2007, 251, 536. (b) A. Juris, M. Venturi, P. Ceroni, V. Balzani, S. Campagna, S. Serroni, Collect. Czech. Chem. Commun. 2001, 66(1), 1. 60. S. Roffia, M. Marcaccio, C. Paradisi, F. Paolucci, V. Balzani, G. Denti, S. Serroni, S. Campagna, Inorg. Chem. 1993, 32, 3003. 61. (a) S. Campagna, G. Denti, S. Serroni, A. Juris, M. Venturi, V. Ricevuto, V. Balzani, Chem. Eur. J. 1995, 1, 211. (b) G. Denti, S. Campagna, S. Serroni, M. Ciano, V. Balzani, J. Am. Chem. Soc. 1992, 114, 2944. 62. P. Ceroni, F. Paolucci, C. Paradisi, A. Juris, S. Roffia, S. Serroni, S. Campagna, A. J. Bard, J. Am. Chem. Soc. 1998, 120, 5480. 63. S. Serroni, A. Juris, M. Venturi, S. Campagna, I. R. Resino, G. Denti, A. Credi, V. Balzani, J. Mater. Chem. 1997, 7, 1227. 64. P. M. S. Monk, The Viologens: Physicochemical Properties, Synthesis and Applications of the Salts of 4,40 -Bipyridine, John Wiley & Sons, Chichester, 1998. 65. V. Balzani, A. Credi, F. M. Raymo, J. F. Stoddart, Angew. Chem., Int. Ed. 2000, 39, 3348. 66. (a) P. Ceroni, V. Vicinelli, M. Maestri, V. Balzani, W. M. M€ uller, U. M€ uller, U. Hahn, F. Osswald, F. V€ogtle, New J. Chem. 2001, 25, 989. (b) R. Toba, J. M. Quintela, C. Peinador, E. Roman, A. E. Kaifer, Chem. Commun. 2001, 857. 67. (a) S. Heinen, L. Walder, Angew. Chem., Int. Ed. 2000, 39, 806. (b) S. Heinen, W. Meyer, L. Walder, J. Electroanal. Chem. 2001, 498, 34. 68. (a) F. Marchioni, M. Venturi, P. Ceroni, V. Balzani, M. Belohradsky, A. M. Elizarov, H.-R. Tseng, L. F. Stoddart, Chem. Eur. J. 2004, 10, 6361. (b) F. Marchioni, M. Venturi, A. Credi, V. Balzani, M. Belohradsky, A. M. Elizarov, H.-R. Tseng, J. F. Stoddart, J. Am. Chem. Soc. 2004, 126, 568. (c) C. M. Ronconi, J. F. Stoddart, V. Balzani, M. Baroncini, P. Ceroni, C. Giansante, M. Venturi, Chem. Eur. J. 2008, 14, 8365. 69. W. S. Baker, B. I. Lemon, III, R. M. Crooks, J. Phys. Chem. B 2001, 105, 8885. 70. Q. Xie, E. Perez-Cordero, L. Echegoyen, J. Am. Chem. Soc. 1992, 114, 3978. 71. Q. Xie, F. Arias, L. Echegoyen, J. Am. Chem. Soc. 1993, 115, 9818.
184
DENDRIMERS AS MULTIELECTRON STORAGE DEVICES
72. For reviews, see: (a) U. Hahn, F. Cardinali, J.-F. Nierengarten, New J. Chem. 2007, 31, 1128. (b) J.-F. Nierengarten, N. Armaroli, G. Accorsi, Y. Rio, J.-F. Eckert, Chem. Eur. J. 2003, 9, 37. 73. For example, see: (a) M. Gutierrez-Nava, G. Accorsi, P. Masson, N. Armaroli, J.-F. Nierengarten, Chem. Eur. J. 2004, 10, 5076. (b) D. Felder, J.-L. Gallani, D. Guillon, B. Heinrich, J.-F. Nicoud, J.-F. Nierengarten, Angew. Chem., Int. Ed. 2000, 39, 201. (c) D. Felder, H. Nierengarten, J.-P. Gisselbrecht, C. Boudon, E. Leize, J.-F. Nicoud, M. Gross, A. Van Dorsselaer, J.-F. Nierengarten, New J. Chem. 2000, 24, 687. 74. U. Hahn, E. Maisonhaute, C. Amatore, J.-F. Nierengarten, Angew. Chem., Int. Ed. 2007, 46, 951. 75. (a) R. Kessinger, J. Crassous, A. Hermann, M. R€uttimann, L. Echegoyen, F. Diederich, Angew. Chem., Int. Ed. 1998, 37, 1919. 76. J. A. Camerano, M. A. Casado, U. Hahn, J.-F. Nierengarten, E. Maisonhaute, C. Amatore, New J. Chem. 2007, 31, 1395. 77. N. Armaroli, C. Boudon, D. Felder, J.-P. Gisselbrecht, M. Gross, G. Marconi, J.-F. Nicoud, J.-F. Nierengarten, V. Vicinelli, Angew. Chem., Int. Ed. 1999, 38, 3730.
CHAPTER 7
Self-assembled Monolayers and Multilayers of Electroactive Thiols IBRAHIM YILDIZ and FRAN C ¸ ISCO M. RAYMO Department of Chemistry, University of Miami, Coral Gables, FL, USA MASSIMILIANO LAMBERTO Department of Chemistry, Medical Technology and Physics, Monmouth University, West Long Branch, NJ, USA
7.1 SELF-ASSEMBLED MONOLAYERS OF THIOLS ON NOBLE METALS Alkanethiols self-assemble on the surface of noble metals in the form of organized monolayers.1,2 The affinity of the thiol group for the metal and van der Waals contacts between adjacent oligomethylene chains are responsible for these supramolecular events. The resulting assemblies protect the metal substrate from the surrounding environment and, as a result, can be exploited to influence interfacial electron transfer.3,4 Specifically, metallic electrodes can be coated with these monolayers to prevent the direct contact of the underlying metal surface with electroactive species in solution. Under these conditions, the organization of the molecules within the monolayer and the film thickness affect the exchange of electrons between the electrode and the solution species and, hence, control the redox response of the latter. Indeed, a wealth of electrodes coated with self-assembled monolayers of alkanethiols has already been prepared with the ultimate goal of unraveling the fundamental factors regulating interfacial electron transfer5 as well as with the prospect of potential technological applications.6–8 The affinity of thiols for noble metals can also be exploited to anchor electroactive molecules to metallic electrodes.5,9,10 For example, compounds 1–4 (Fig. 7.1) have a redox center and an aliphatic tail terminated by a thiol group and they all adsorb on the Electrochemistry of Functional Supramolecular Systems. Edited by Paola Ceroni, Alberto Credi, and Margherita Venturi Copyright 2010 John Wiley & Sons, Inc.
185
186
SELF-ASSEMBLED MONOLAYERS AND MULTILAYERS OF ELECTROACTIVE THIOLS
Figure 7.1 Representative examples of electroactive compounds with pendant thiol groups able to self-assemble into monolayers on the surface of gold electrodes.
surface of gold to form a monolayer of electroactive species or coadsorb, together with alkanethiols, on the electrode surface to generate a mixed monolayer of electroactive and inert components.11–14 The attachment of the redox centers to the electrode surface can easily be confirmed by cyclic voltammetry.5 This convenient technique can also quantify the amount of electroactive species adsorbed on the electrode surface. However, the electrochemical response of surface-confined redox centers is significantly different from that of solution species.4,5 Indeed, the lack of physical diffusion imposes a linear scan-rate (n) dependence on the peak current (ip), according to Equation (7.1). The slope of the plot of ip against n is related to the number (n) of exchanged electrons, the Faraday constant (F), the area (A) of the electrode, the surface coverage (G), the gas constant (R) and the temperature (T). The value of G can be estimated from the charge (Q) passed on reduction or oxidation, according to Equation (7.2), and is equal to the moles of electroactive sites embedded in the film per unit area. ðnFÞ2 AGn ð7:1Þ ip ¼ 4RT Q ¼ nFAG
ð7:2Þ
As an example, the cyclic voltammograms (Fig. 7.2a) of 4 adsorbed on gold show the reversible oxidation of the tetrathiafulvalene unit to the corresponding radical
7.1
SELF-ASSEMBLED MONOLAYERS OF THIOLS ON NOBLE METALS
187
Figure 7.2 Cyclic voltammograms (a) of 4 adsorbed on a gold working electrode (0.1 M Bu4NPF6, MeCN, V versus Ag/AgCl, 100–500 mV/s) and scan-rate (n) dependence of the peak current (iP) for the anodic and cathodic waves (b).
cation.14 The redox waves for the oxidation and back reduction processes grow significantly with n. Plots (Fig. 7.2b) of the corresponding ip against n are both linear, consistently with Equation (7.1), and G is ca. 0.6 nmol/cm2, in agreement with the formation of a monolayer of 4 on the electrode surface. Similar monolayers have been prepared with a diversity of electroactive units with the ultimate goal of elucidating the subtle balance between the structural and the electronic factors that regulate interfacial electron transfer.5,9,10 In particular, these studies have focused their attention on the rationalization of the influence that the distance between the electrode surface and the redox centers as well as the nature of the linkers between them have on the rates of electron transfer. In parallel to these fundamental investigations, the ability of thiols to anchor electroactive units on metallic electrodes has also been exploited to fabricate a wealth of nanostructured materials with tailored functions and properties.6–8 Indeed, these convenient building
188
SELF-ASSEMBLED MONOLAYERS AND MULTILAYERS OF ELECTROACTIVE THIOLS
blocks have been integrated within miniaturized electronic devices and chemical sensors.
7.2 SELF-ASSEMBLED MONOLAYERS OF BISTHIOLS ON NOBLE METALS The spontaneous adsorption of thiols on noble metals is a convenient process to passivate electrode surfaces.1,2 This method, however, can only produce a single layer of molecules on the metal substrate. As a consequence, the resulting films are not particularly stable and offer only limited protection to the electrode surface. Compounds incorporating two thiol groups separated by a linear spacer can, instead, selfassemble into multiple layers on noble metals.15–27 Indeed, the formation of metal– thiolate bonds at one of their two ends anchors these molecules to the substrate in the form of a monolayer (Fig. 7.3a). In the presence of molecular oxygen, the subsequent and spontaneous formation of disulfide linkages at the other end encourages the deposition of additional molecular layers on the initial monolayer (Fig. 7.3b). The final result is the formation of robust and stable films consisting of multilayers of covalently linked building blocks. For example, compound 5 (Fig. 7.3) incorporates two thiol groups separated by an oligomethylene spacer.16 The immersion of a gold substrate into an ethanol solution of this particular bisthiol results in the adsorption of one of the two thiol groups on the metal surface with the initial formation of a monolayer. Over the course of 12 h, however, up to seven additional layers assemble on top of the initial one, as a consequence of the oxidative formation of disulfide linkages. In fact, X-ray photoelectron spectroscopic measurements confirmed the presence of disulfide bonds in these films and ellipsometric analyses demonstrated that the film thickness grows over time with an average thickness of 7 A per layer. Furthermore, these films survive, essentially unaffected, extensive washing with a variety of solvents in agreement with interlayer covalent bonding.
Figure 7.3 Gold–thiolate bonds encourage the adsorption of the bisthiol 5 on the surface of gold in the form of a monolayer (a). The subsequent and spontaneous formation of disulfide linkages promotes the deposition of additional molecules of 5 to produce multilayers (b).
7.4 SELF-ASSEMBLY OF ELECTROACTIVE MONOLAYERS AND MULTILAYERS
189
Figure 7.4 Electroactive compounds incorporating 4,40 -bipyridinium dications and either one (6) or two (7) thiol groups.
7.3 DESIGN AND SYNTHESIS OF BIPYRIDINIUM THIOLS AND BISTHIOLS Numerous bisthiols have been observed to form spontaneously multilayers on gold and silver on the basis of the oxidative formation of disulfides.15–27 Nonetheless, most of these compounds lack electroactive character, with few notable exceptions.23,27 In principle, the introduction of redox centers at the core of these molecules and their subsequent assembly into multilayers can be exploited to generate electroactive films. The concentration of redox centers within the resulting electrode coatings, as well as their thickness, can be significantly larger than those possible with electroactive thiols such as 1–4 (Fig. 7.1).11–14 In addition, the transition from electroactive monolayers to electroactive multilayers can translate into a significant enhancement in stability and a much more effective protection of the electrode surface. On the basis of these considerations, we designed the bipyridinium derivatives 6 and 7 (Fig. 7.4).27b,c Both incorporate two 4,40 -bipyridinium dications separated by a 1,4-dimethylphenylene spacer, but differ in the number of thiol groups at the end of their aliphatic tails. Compound 6 has only one thiol group and can be prepared from known precursors in four steps, involving the N-alkylation of the two heterocyclic fragments and the final deprotection of the thiol group. Compound 7 has two thiol groups and its synthesis requires only two steps, involving similar reagents and reaction conditions. The established electrochemical behavior of the 4,40 -bipyridinium dication imposes electroactive character on both compounds.28 The single thiol group of 6 can promote the self-assembly of this compound on gold in the form of a monolayer, while the two thiol groups of 7 can encourage the formation of multilayers.
7.4 SELF-ASSEMBLY OF ELECTROACTIVE MONOLAYERS AND MULTILAYERS The immersion of a polycrystalline gold electrode into a chloroform/methanol (2:1, v/v) solution of the tetrachloride salt of 6 (Fig. 7.4) results in the adsorption of this electroactive thiol on the metal surface.27c Consistently, the cyclic voltammogram (Fig. 7.5a), recorded after an immersion time of 24 h and extensive rinsing of the electrode surface, shows the reversible reduction of the bipyridinium dications to the corresponding radical cations. In addition, ip increases linearly with n (Fig. 7.6a),
190
SELF-ASSEMBLED MONOLAYERS AND MULTILAYERS OF ELECTROACTIVE THIOLS
Figure 7.5 Cyclic voltammograms (0.1 M KCl, H2O, V versus Ag/AgCl, 100 mV/s) recorded after the immersion of a polycrystalline gold working electrode in solutions of 6 for 24 h (a) and 7 for 1 h (b), 6 h (c), and 48 h (d) and extensive rinsing of the electrode surface.
consistently with Equation (7.1), and G is ca. 0.6 nmol/cm2, in agreement with the formation of a monolayer of 6 on the electrode surface. Thus, the terminal thiol group of this particular compound ensures the attachment of the pair of bipyridinium dications to the gold substrate and permits the assembly of an electroactive monolayer. However, the lack of a second thiol group prevents the subsequent adsorption of additional overlayers as a result of interlayer disulfide linkages. In fact, the cyclic voltammogram does not change when the immersion time is prolonged to 48 h. Even under these conditions, ip increases, once again, linearly with n and G remains essentially unaffected. Also the bisthiol 7 adsorbs on the surface of polycrystalline gold from chloroform/ methanol (2:1, v/v) solutions.27b,c The cyclic voltammogram (Fig. 7.5b), recorded after an immersion time of only 1 h and extensive rinsing of the electrode surface, reveals, once again, the characteristic response of the bipyridinium dications. Also in
7.4 SELF-ASSEMBLY OF ELECTROACTIVE MONOLAYERS AND MULTILAYERS
191
Figure 7.6 Scan-rate (n) dependence of the peak current (iP) in cyclic voltammograms recorded after the immersion of a polycrystalline gold working electrode in solutions of 6 for 24 h (a) and 7 for 1 h (b), 6 h (c), and 48 h (d) and extensive rinsing of the electrode surface.
this instance, ip increases linearly with n (Fig. 7.6b) and G is ca. 0.7 nmol/cm2, consistently with the formation of a monolayer on the gold substrate. However, this particular compound has a second thiol group available for disulfide bonding after the adsorption of the first on gold. Indeed, the cyclic voltammograms (Fig. 7.5c and d), recorded after immersion times of 6 and 48 h under otherwise identical conditions, reveal a significant increase in ip as a result of the adsorption of additional electroactive units on the electrode surface. Furthermore, the correlation between ip and n deviates from linearity (Fig. 7.6c and d), in agreement with multilayer formation, and G is ca. 2.5 and 8.4 nmol/cm2 after 6 and 48 h, respectively. These values are ca. 4 and 14 times, respectively, greater than what expected for a single layer of molecules on the electrode surface and, thus, confirm multilayer formation. Furthermore, the inability of 6 to form multilayers is indicative of the need to have a pair of thiol groups per molecule to encourage multilayer growth.
192
SELF-ASSEMBLED MONOLAYERS AND MULTILAYERS OF ELECTROACTIVE THIOLS
7.5 ELECTRON TRANSPORT THROUGH ELECTROACTIVE MULTILAYERS The nonlinear correlation between ip and n (Fig. 7.6c and d) is a consequence of the diffusive character of electron migration in multilayers of electroactive units.5,29 This process can be described quantitatively in the form of an apparent diffusion coefficient (DA). This parameter is a measure of the film’s ability to transport electrons and is related to two terms. One of them quantifies the physical diffusion of the redox sites within the film boundaries and the other is related to the electron self-exchange between adjacent redox sites within the film. The first of these two terms, however, is negligible for the multilayers prepared from 7. Indeed, the interlayer disulfide linkages restrict the mobility of the redox units and prevent their physical diffusion. Thus, the electron migration through these particular films is predominantly a result of self-exchange processes. The diffusion of electrons, however, must be accompanied by a flow of counterions to maintain charge neutrality. Specifically, the chloride anions associated with the bipyridinium dications must diffuse out of the film into the bulk solution upon reduction to permit the flow of electrons from the film/electrode to the film/solution interface. Chronoamperometry is a convenient technique for the determination of DA.5,29 It is based on the application of a potential step to the working electrode to encourage an oxidation or reduction process and the subsequent monitoring of the current (i) over time (t).4 The fitting of the resulting plot with Equation (7.3) provides an estimate of DA, after an independent determination of the concentration (c) of the redox units in the film. In principle, c can be calculated from G and the film thickness (f), according to Equation (7.4). Nonetheless, f cannot easily be determined under the same conditions of the electrochemical experiment. Furthermore, f can change during the oxidation or reduction of the film in the chronoamperometric measurement. As a pffiffiffiffiffiffiffi result, the term c DA is often reported in the literature instead of DA. pffiffiffiffiffiffiffi nFAc DA i ¼ pffiffiffiffiffi pt c¼
G f
ð7:3Þ ð7:4Þ
To assess the electron transport properties of our electroactive films, we applied a potential step from 0 to 0.65 V versus Ag/AgCl to an electrode coated with multilayers of 7.27c The fitting ofpthe ffiffiffiffiffiffiffilinear portion of the resulting plot (Fig. 7.7) with Equation (7.3) indicates c DA to be ca. 31 nmol/cm2/s1/2. This value is remarkably similar to those reported in the literature for redox polymers incorporating bipyridinium dications.27a Thus, the electron transport properties of these electroactive multilayers are essentially analogous to those of conventional redox polymers. The multilayer assembly of electroactive units protects the electrode surface and can prevent redox species in solution from contacting the underlying metal substrate. However, the electron transport properties of these electroactive units can be
7.5
ELECTRON TRANSPORT THROUGH ELECTROACTIVE MULTILAYERS
193
Figure 7.7 Chronoamperometric plot (0 ! 0.65 V versus Ag/AgCl, 0.1 M KCl, H2O) recorded after the immersion of a polycrystalline gold working electrode in a solution of 7 for 48 h and extensive rinsing of the electrode surface.
exploited to deliver electrons to species in solution. Indeed, the bipyridinium dications within the multilayer assembly can mediate the transfer of electrons from the electrode surface to species at the film/solution interface. For example, the cyclic voltammogram (Fig. 7.8a) of Ru(NH3)63 þ , recorded with a bare polycrystalline gold working electrode, shows the reversible reduction of the ruthenium centers.27b,c When the electrode is coated with a multilayer of 7, the ruthenium complex cannot approach the gold surface, but it can still be reduced through the bipyridinium dications in the film. Consistently, a comparison of the cyclic voltammogram of the film (Fig. 7.8b) with that of Ru(NH3)63 þ (Fig. 7.8c), recorded with a coated electrode, shows a significant increase in the current associated with the reduction of the bipyridinium dications. This trend demonstrates that the bipyridinium dications mediate the reduction of the ruthenium centers. However, the back scan of both cyclic voltammograms is essentially the same, suggesting that only the bipyridinium dications are regenerated, while the ruthenium centers remain in a reduced form. Thus, the electroactive multilayers impose rectification on the electron transport process. Electrons travel from the electrode through the film to the metal centers, where they remain trapped even when the electrode voltage is reset to the original value. This behavior is a result of the mismatch in energy between the bipyridinium and ruthenium orbitals. Specifically, electrons can travel through the bipyridinium lowest unoccupied molecular orbitals (LUMOs) to the singly occupied molecular orbital of the ruthenium complex, when the electrode voltage matches the bipyridinium reduction potential (Fig. 7.9a). After lowering the voltage to the original value, the electrons delivered to the ruthenium orbitals cannot be transferred back to the bipyridinium LUMOs, because of the higher energy of the latter relative to the former (Fig. 7.9b).
194
SELF-ASSEMBLED MONOLAYERS AND MULTILAYERS OF ELECTROACTIVE THIOLS
Figure 7.8 Cyclic voltammograms (0.1 M KCl, H2O, V versus Ag/AgCl, 100 mV/s) of Ru(NH3)6Cl3 (5 mM) recorded with a polycrystalline gold working electrode before (a) and after (c) immersion in a solution of 7 for 6 h and extensive rinsing of the electrode surface and of the same bipyridinium film (b) in the absence of Ru(NH3)6Cl3 in the electrolyte solution, under otherwise identical conditions. (d) Cyclic voltammogram (0.1 M KCl, H2O, V versus Ag/AgCl, 100 mV/s) of Ru(NH3)6Cl3 (5 mM) recorded in the presence of K4Fe(CN)6 (0.04 mM) with a polycrystalline gold working electrode after immersion in a solution of 7 for 6 h and extensive rinsing of the electrode surface.
The positive charges of the bipyridinium dications within the electroactive multilayers are balanced by chloride counterions. The addition of K4Fe(CN)6 to the electrolyte solution, however, results in the exchange of the chloride anions with ferrocyanide tetra-anions.27b,c The electrostatic trapping of the tetra-anions within the multilayers causes the appearance of waves for the reversible oxidation of the iron centers in the corresponding cyclic voltammogram in addition to those associated with the reversible reduction of the bipyridinium dications. Interestingly, the potentials associated with these two sets of waves are more positive and more negative,
7.5
ELECTRON TRANSPORT THROUGH ELECTROACTIVE MULTILAYERS
195
Figure 7.9 Electrons travel from the electrode to the ruthenium centers through the bipyridinium dications (a), when the gold voltage (V versus Ag/AgCl) matches the reduction potential of the bipyridinium dications, but cannot return from the ruthenium centers to the electrode (b), even when the voltage is lowered below the oxidation potential of the ruthenium centers.
respectively, than those for the reversible reduction of Ru(NH3)63 þ . It follows that the bipyridinium LUMOs can mediate the transfer of electrons from the electrode to the ruthenium centers (Fig. 7.10a), while the partially occupied orbitals of Fe(CN)63 can mediate the reoxidation of the ruthenium complexes (Fig. 7.10b). Indeed, the cyclic voltammogram (Fig. 7.8d) of Ru(NH3)63 þ , recorded with a polycrystalline gold working electrode coated with a multilayer of 7 and containing Fe(CN)64 anions, shows the oxidation (I) and back reduction (II) of the iron centers, the reduction of the bipyridinium dications together with that of the ruthenium centers (III), the back oxidation of the bipyridinium radical cations (IV), the back oxidation of the iron centers together with that of the ruthenium centers (V) and the back reduction of the iron centers (VI). The huge splitting between the reduction (III) and oxidation (V) waves associated with the ruthenium centers demonstrates that these processes are mediated by the bipyridinium and iron centers, respectively, and parallels the energy difference between the LUMOs of the bipyridinium dications and the highest occupied molecular orbitals (HOMOs) of the iron complex (Fig. 7.10). Thus, the introduction of electroactive anions in place of the chloride anions can be exploited to
196
SELF-ASSEMBLED MONOLAYERS AND MULTILAYERS OF ELECTROACTIVE THIOLS
Figure 7.10 Electrons travel from the electrode to the ruthenium centers through the bipyridinium dications (a), when the gold voltage (V versus Ag/AgCl) matches the reduction potential of the bipyridinium dications, and return from the ruthenium centers through the electroactive anions to the electrode (b), when the gold voltage matches the oxidation potential of the iron centers.
tune the electron transport properties of these bipyridinium assemblies and permit the transfer of electrons in two directions. Specifically, the cationic components of these films permit the transfer of electrons form the electrode to the film/solution interface, while the anionic components allow the transport in the opposite direction.
7.6
ELECTROCHROMIC MONOLAYERS AND MULTILAYERS
Bipyridinium dications do not absorb in the visible region of the electromagnetic spectrum.28 After their electrochemical reduction to the corresponding radical cations, however, a broad and intense absorption appears across the visible region. The process is fully reversible and the colorless state is regenerated, after the oxidation of the radical cations back to the original dications. In fact, bipyridinium dications are convenient building blocks for the realization of electrochromic materials, as a result of their spectroscopic response to electrochemical inputs.
7.7
CONCLUSIONS
197
To assess the electrochromic response of the bipyridinium dications embedded into multilayers of 7, we envisaged the possibility of assembling these films on optically transparent platinum electrodes.27d–f Specifically, we deposited an ultrathin platinum film on an indium-tin oxide substrate and then immersed the resulting assembly into a chloroform/methanol (2:1, v/v) solution of 7. As observed with the gold electrodes (Fig. 7.5), the corresponding cyclic voltammograms show waves for the reversible reduction of the bipyridinium dications with a significant increase in ip with the immersion time. In fact, G is 0.8, 1.5, and 3.1 nmol/cm2 after immersion times of 1, 6, and 72 h, respectively. Furthermore, the correlation between ip and n is linear after 1 h and deviates from linearity after 6 and 72 h. Thus, the bisthiol 7 can indeed form multiple electroactive layers also on platinum substrates. The indium-tin oxide/platinum assemblies are transparent in the visible region and permit the spectroscopic measurement of adsorbed species in transmission mode.27d–f The spectra of films prepared from 7 after immersion times of 1, 6, and 72 h do not show any significant absorbance in the visible region, when the voltage of the platinum electrode is maintained at 0 V versus Ag/AgCl. After the application of a potential step from 0 to 0.7 V, however, the bipyridinium dications are reduced to the corresponding radical cations and their characteristic absorption bands appear in the corresponding spectra (Fig. 7.11a–c) with a linear increase in absorbance with G. These spectral changes are fully reversible and the bands of the radical cations disappear after stepping the voltage back to 0 V. In fact, the absorbance in the visible region can be switched between low and high values repeatedly simply by switching the voltage of the platinum support between 0 and 0.7 V (Fig. 7.11d–f). Thus, the bipyridinium dications within these electroactive films retain their electrochromic response after their immobilization on the electrode surface in the form of multilayer assemblies.
7.7
CONCLUSIONS
Electroactive compounds can be anchored to the surface of metal electrodes, relying on the ability of thiols to adsorb on noble metals. The final result is the formation of a monolayer of electroactive molecules on the electrode surface. This strategy, however, is limited to the formation of a single molecular layer on the metal substrate and, therefore, produces films with limited stability and offers limited protection to the electrode. Compounds with two thiol groups separated by a linear spacer can instead assemble into multiple layers on noble metals, as a result of the spontaneous and oxidative formation of interlayer disulfide bonds. The covalent linkages holding the resulting assemblies translate into remarkable stability. Furthermore, these films protect the electrode surface more effectively than single monolayers. The insertion of redox units, in the form of pairs of bipyridinium dications, between the terminal thiols of these compounds can be exploited to impose electroactive character on the resulting films. These multilayer constructs can transport electrons with apparent diffusion coefficients comparable to those of conventional redox polymers. Furthermore, they can shuttle electrons from the electrode surface to species in solution, but
198
SELF-ASSEMBLED MONOLAYERS AND MULTILAYERS OF ELECTROACTIVE THIOLS
Figure 7.11 Absorption spectra (0.1 M, KCl, H2O, 0.7 V versus Ag/AgCl) recorded after the exposure of optically transparent platinum electrodes to solution of 7 for 1 h (a), 6 h (b), and 72 h (c) and extensive rinsing of the electrode surface. Absorbance changes at 540 nm observed for films prepared after the exposure times of 1 h (d), 6 h (e), and 72 h (f), when the electrode potential is switched between 0 and 0.7 V.
prevent the transfer of electrons in the opposite direction. Thus, they can be exploited to impose rectification on electron transport. In addition, the nature of the counterions associated with the bipyridinium dications can be regulated to control the electron transport properties of the films. Specifically, electroactive anions, in the form of ferricyanide tetra-anions, can be inserted within the multilayer structure and their electronic states can be exploited to mediate the transfer of electrons from species in solution to the electrode surface. In addition, these electroactive multilayers selfassemble also on the surface of optically transparent platinum films. Under these conditions, their spectral properties can be probed in transmission mode in response to electrochemical stimulations. In fact, the electrochemical reduction of the bipyridinium dications to the corresponding radical cations and their back oxidation result in
REFERENCES
199
the modulation of the absorbance in the visible region. Thus, the ability of bisthiols to form multilayers in combination with the electrochemical and spectral properties of bipyridinium dications can be exploited to generate functional materials with unique electron transport properties and electrochromic character. ACKNOWLEDGMENTS The authors thank the National Science Foundation (CAREER Award CHE-0237578 and CHE0749840) and the University of Miami for financial support.
REFERENCES 1. A. Ulman, An Introduction to Ultrathin Organic Films, Academic Press, Boston, 1991. 2. J. G. Vos, R. J. Forster, T. E. Keyes, Interfacial Supramolecular Assemblies, Wiley, New York, 2003. 3. V. Balzani (Ed.), Electron Transfer in Chemistry, Wiley-VCH, Weinheim, 2001. 4. A. J. Bard, L. R. Faulkner, Electrochemical Methods: Fundamentals and Applications, Wiley, New York, 2001. 5. R. W. Murray (Ed.), Molecular Design of Electrode Surfaces, Wiley, New York, 1992. 6. (a) J. M. Tour, Acc. Chem. Res. 2000, 33, 791–804. (b) J. Chen, W. Wang, J. Klemic, M. A. Reed, B. W. Axelrod, D. M. Kaschak, A. M. Rawlett, D. W. Price, S. M. Dirk, J. M. Tour, D. S. Grubisha, D. W. Bennett, Ann. NY Acad. Sci. 2002, 960, 69–99. 7. R. L. Carroll, C. B. Gorman, Angew. Chem., Int. Ed. 2002, 41, 4379–4400. 8. (a) B. H. Huisman, F. C. J. M. van Veggel, D. N. Reinhoudt, Pure Appl. Chem. 1998, 70, 1985–1992. (b) S. Flink, F. C. J. M. van Veggel, D. N. Reinhoudt, Adv. Mater. 2000, 12, 1315–1328. 9. (a) H. O. Finklea, Electroanal. Chem. 1996, 19, 109–335. (b) H. O. Finklea, Encyclopedia of Analytical Chemistry: Applications, Theory and Instrumentation, R. A. Meyers (Ed.), Wiley, Chichester, 2000, Vol. 11, pp. 10090–10115. 10. V. Chechik, C. J. M. Stirling, The Chemistry of Organic Derivatives of Gold and Silver, S. Patai, Z. Rappoport (Eds.), Wiley, Chichester, 1999, pp. 551–640. 11. C. E. D. Chidsey, C. R. Bertozzi, T. M. Putvinski, A. M. Mujsce, J. Am. Chem. Soc. 1990, 112, 4301–4306. 12. J. J. Hickman, D. Ofer, P. E. Laibinis, G. M. Whitesides, M. S. Wrighton, Science 1991, 252, 688–691. 13. H. O. Finklea, D. D. Hanshew, J. Electroanal. Chem. 1993, 347, 321–340. 14. E. J. Pacsial, D. Alexander, R. J. Alvarado, M. Tomasulo, F. M. Raymo, J. Phys. Chem. B 2004, 108, 19307–19313. 15. J. M. Tour, L. Jones II, D. L. Pearson, J. J. S. Lamba, T. P. Burgin, G. M. Whitesides, D. L. Allara, A. N. Parikh, S. V. Atre, J. Am. Chem. Soc. 1995, 117, 9529–9534. 16. P. Kohli, K. K. Taylor, J. J. Harris, G. J. Blanchard, J. Am. Chem. Soc. 1998, 120, 11962– 11968. 17. (a) S. W. Joo, S. W. Han, K. Kim, J. Phys. Chem. B 1999, 103, 10831–10837. (b) S. W. Joo, S. W. Han, K. Kim, Langmuir 2000, 16, 5391–5396. (c) S. W. Joo, S. W. Han, K. Kim, J.
200
18. 19. 20. 21. 22. 23. 24. 25. 26. 27.
28. 29.
SELF-ASSEMBLED MONOLAYERS AND MULTILAYERS OF ELECTROACTIVE THIOLS
Phys. Chem. B 2000, 104, 6218–6224. (d) S. W. Joo, S. W. Han, K. Kim, J. Coll. Interf. Sci. 2001, 240, 391–399. (e) S. W. Joo, S. W. Han, K. Kim, Mol. Cryst. Liq. Cryst. 2001, 371, 355–358. S. W. Chah, J. H. Fendler, J. H. Yi, Chem. Commun. 2002, 2094–2095. U. Weckenmann, S. Mittler, K. Naumann, R. A. Fischer, Langmuir 2002, 18, 5479–5486. T. L. Brower, J. C. Garno, A. Ulman, G.-Y. Liu, C. Yan, A. G€ olzh€auser, M. Grunze, Langmuir 2002, 18, 6207–6216. S. Rifai, M. Morin, J. Electroanal. Chem. 2003, 550, 277–289. D. L. Pugmire, M. J. Tarlov, R. D. van Zee, J. Naciri, Langmuir 2003, 19, 3720–3726. W. Haiss, H. van Zalinge, H. Hobenreich, D. Bethell, D. J. Schiffrin, S. J. Higgins, R. J. Nichols, Langmuir 2004, 20, 7694–7702. H. M. Zareie, A. M. McDonagh, J. Edgar, M. J. Ford, M. B. Cortie, M. R. Phillips, Chem. Mater. 2006, 18, 2376–2380. A. Shaporenko, M. Elbing, A. Baszczyk, C. von Hanisch, M. Mayor, M. Zharnikov, J. Phys. Chem. B 2006, 110, 4307–4317. K. H. A. Lau, C. Huang, N. Yakovlev, Z. K. Chen, S. J. O’Shea, Langmuir 2006, 22, 2968– 2971. (a) F. M. Raymo, R. J. Alvarado, Chem. Rec. 2004, 4, 204–218. (b) F. M. Raymo, R. J. Alvarado, E. J. Pacsial, D. Alexander, J. Phys. Chem. B 2004, 108, 8622–8625. (c) R. J. Alvarado, J. Mukherjee, E. J. Pacsial, D. Alexander, F. M. Raymo, J. Phys. Chem. B 2005, 109, 6164–6173. (d) S. Sortino, S. Di Bella, S. Conoci, S. Petralia, M. Tomasulo, E. J. Pacsial, F. M. Raymo, Adv. Mater. 2005, 17, 1390–1393. (e) S. Conoci, S. Petralia, P. Samorı`, F. M. Raymo, S. Di Bella, S. Sortino, Adv. Funct. Mater. 2006, 16, 1425–1432. (f) S. Sortino, S. Conoci, I. Yildiz, M. Tomasulo, F. M. Raymo, J. Mater. Chem. 2006, 16, 3171–3173. (g) I. Yildiz, J. Mukherjee, M. Tomasulo, F. M. Raymo, Adv. Funct. Mater. 2007, 17, 814–820. P. M. S. Monk, The Viologens: Physicochemical Properties, Synthesis and Applications of the Salts of 4,40 -Bipyridine, Wiley, New York, 1998. G. Inzelt, Electroanal. Chem. 1994, 18, 89–241.
CHAPTER 8
Electrochemistry of Carbon Nanoparticles LUIS ECHEGOYEN, AMIT PALKAR, and FREDERIC MELIN Department of Chemistry, Clemson University, Clemson, SC, USA
The discovery of fullerenes in 1985 led to the era of nanomaterials.1 The threedimensional geometry of these molecules as well as their unique properties distinguishes them from conventional molecules encountered in organic chemistry. Due to recent discoveries in this field, the horizons of this area have broadened to encompass various new molecules such as endohedral fullerenes, nanotubes, carbon nanohorns, and carbon nano-onions. This chapter discusses the electrochemical behavior of some of these carbon nanoparticles with special emphasis on endohedral fullerenes. Since a large number of fullerene derivatives have been prepared and their various electrochemical studies in different solvents and electrolytes have been reported, the electrochemistry of these derivatives is beyond the scope of this text.2,3 Among the other carbon nanoparticles, the electrochemistry of derivatives of carbon nanotubes has been reported. These studies have been highlighted in the final part of the chapter. 8.1 8.1.1
ELECTROCHEMISTRY OF EMPTY FULLERENES Electrochemistry of C60 and C70
The first members of the fullerene family to be discovered were C60 and C70. The electrochemical properties of these compounds have been well characterized. Both fullerenes show six reversible reductions and one oxidation by cyclic voltammetry.4,5 The reductions are almost equally spaced, with the first reduction occurring at 1.0 V versus ferrocene/ferrocenium couple. Successive reductions occur approximately 400 mV apart (Fig. 8.1). Electrochemistry of Functional Supramolecular Systems. Edited by Paola Ceroni, Alberto Credi, and Margherita Venturi Copyright 2010 John Wiley & Sons, Inc.
201
202
ELECTROCHEMISTRY OF CARBON NANOPARTICLES
C60 at –10ºC
(a)
10 μA
(b)
5 μA
–1.0
–2.0 Potential (volts versus Fc/Fe+)
–3.0
Figure 8.1 Cyclic voltammetry (a) and Osteryoung square wave voltammetry (b) of C60 (acetonitrile/toluene þ 0.1 M (n-Bu)4NPF6), using a glassy carbon electrode (GCE) working and ferrocene/ferrocenium (Fc/Fc þ ) couple as an internal reference. Reprinted with Permission from Ref. 4. Copyright 1992 American Chemical Society.
Empty h shell ( l = 5) s, p, d, f, g shells completely filled 2(l+1) = 11
t2u t1u
t2u t1u hu
hu
Figure 8.2 Partial molecular orbital diagram for C60.6
The reductions are a result of successive filling of low-lying t1u orbitals, while the oxidation is due to the removal of electrons from the hu orbitals, as shown in the partial molecular diagram in Fig. 8.2. 8.1.2
Electrochemistry of Larger Cages (C76–C92)
The electrochemistry of higher cages such as C76, C78, and C84 is, however, surprisingly different from that of the smaller ones. Figure 8.3 shows the electrochemistry of C76, C78, and C84.7 C76 and C78 show two reductions and two oxidations, respectively, whereas C84 shows three reductions and one oxidation. The reductions are shifted anodically with respect to C60 and C70 and the oxidations are shifted cathodically, and thus these molecules have a lower electrochemical gap than C60 or C70. The potential values obtained are reported in Table 8.1.
8.1 ELECTROCHEMISTRY OF EMPTY FULLERENES
203
+0.800 0.000 –0.800
Current (μA)
(a) C76 +0.600 0.000
(b) C78
–0.500
+0.600 0.000
(EOX –ERED)
(c) C84
“Electrochemical gap”
–0.600
+2.50
+1.50
+0.50
–0.50
–1.50
E (volts) versus ferrocene
Figure 8.3 Differential pulse voltammograms of C76, C78, and C84 (tetrachloroethane þ 0.1 M (n-Bu)4NPF6), using a GCE carbon electrode working and ferrocene/ferrocenium (Fc/Fc þ ) couple as an internal reference. Reprinted with permission from Ref. 7. Copyright 1995 American Chemical Society.
The electrochemistry of C86, C90, and C92 has been reported for isomeric mixtures of individual cages.8 As the cage size of the fullerene increases, the yield of fullerenes decreases significantly. In addition to this, the number of possible isomers also increases, which in turn makes separation of isomers difficult. Thus, no reports appear in the literature indicating the electrochemical behavior of a single isomer of empty TABLE 8.1 Electrochemical Potentials (in V Versus Fc þ /Fc) of Empty Fullerenes Obtained in Various Compounds Fullerene
E1/2 ox1
E1/2 red1
E1/2 red2
E1/2 red3
E1/2 red4
DEgap
23
1.26 (a) 1.20 (a) 0.81 0.81 0.95 0.72 0.93 0.74 0.43 0.78
0.98 (b) 0.97 (b) 0.83 0.94 0.77 0.69 0.67 0.59 0.49 0.46
1.37 (b) 1.34 (b) 1.12 1.26 1.08 1.04 0.96 0.87 0.78 0.7
1.87 (b) 1.78 (b) — 1.72 — 1.58 1.27 1.61 1.49 1.34
2.35 (b) 2.21 (b) — 2.13 — 1.94 — 1.97 1.83 1.69
2.24 2.17 1.64 1.75 1.72 1.41 1.60 1.33 0.92 1.24
C60 , C702,3 C767 (a) C768 C78 [C2v]7 (a) C828 (c) C847 (a) C868 (c) C908 (c) C928 (c)
(a) o-1,1,2,2-Tetrachloroethane þ 0.1 M (n-Bu)4NPF6, (b) acetonitrile/toluene (1/5) þ 0.1 M (n-Bu)4NPF6 at 10 C, and (c) o-dichlorobenzene, 0.1 M (n-Bu)4NPF6. Only redox potentials of major isomers are given; DEgap ¼ E1/2 ox1 E1/2 red1.
204
ELECTROCHEMISTRY OF CARBON NANOPARTICLES
fullerenes higher than C84. The electrochemistry of an isomer mixture of C86 shows four reductions and one oxidation, all of which are reversible. Half-wave potentials for the redox processes are reported in Table 8.1. The separation between the first and second reductions as well as between the third and the fourth indicates that the molecule possesses nondegenerate LUMO and LUMO þ 1. Based on studies performed by varying initial potential and equilibrium time, the electrochemistry of C86 appears to be a mixture of at least two isomers. The C90 cage has 46 possible constitutional isomers, out of which only five can be isolated. The electrochemistry of C90 shows two oxidations and six reductions. The redox potentials for C90 are given in Table 8.1. The first reduction potential appears at 0.49 V versus ferrocene/ferrocenium, thus making C90 the easiest to reduce among the empty cage fullerenes. The biggest fullerene isolated and studied using electrochemistry is C92, which shows eight reversible reductions and one broad irreversible oxidation. The intensities of the eight reduction peaks can be grouped into two distinct sets, thus indicating that the electrochemistry corresponds to a mixture of two isomers. The reduction and oxidation potentials are shown in Table 8.1. It can be easily observed from the fullerenes reported in the tables above that some empty fullerenes such as C72, C74, C80, and C88 are conspicuously missing from the list. Previously, it was assumed that these structures are not formed under the arcing condition as they were never observed in the soot extracts. The anomalous absence of empty fullerenes such as C74 or C80 has been attributed to the fact that these fullerenes have a very small HOMO–LUMO gap.9 The difference between the first oxidation and the first reduction potential as obtained from electrochemistry is regarded as a measure of the HOMO–LUMO gap. For the fullerenes reported above, almost all of them show a HOMO–LUMO gap of 1.0 eV or greater. Using density functional theory, C74 has been shown to have a bandgap of 0.05 eV, and thermal population of the triplet state produces a diradical.9 Therefore, C74 is kinetically unstable and so reactive that it can exist either only as a polymerized solid or as a charge transfer complex between a metal and the fullerene. Alford and Diener have reported the isolation of fullerenes up to C100 by reduction of sublimed raw carbon soot obtained after the arcing process.9 Upon reduction, the insoluble fullerenes are reduced to their anion state, which repel each other and therefore stay in solution. Upon reoxidation, these fullerenes can be selectively deposited on the working electrode as polymerized solids. Thus, many previously unobserved fullerenes exist in the soot as insoluble polymerized solids due to their small band gaps. Since reduction of these fullerenes increases their solubility, it can be anticipated that the incorporation of electronaccepting or electron-withdrawing species inside fullerenes cages can be a way to stabilize these small bandgap fullerenes.
8.2
ELECTROCHEMISTRY OF ENDOHEDRAL FULLERENES
Fullerenes can accommodate in their hollow core a large variety of atoms, clusters, and molecules.10–14 Inert rare gases15 and highly reactive moieties such as nitrogen,16
8.2
ELECTROCHEMISTRY OF ENDOHEDRAL FULLERENES
205
TABLE 8.2 Comparison of the Half-Wave Redox Potential (in V Versus Fc þ /Fc) of H2@C60 and C60 Obtained from Differential Pulse Voltammetry in Toluene–Acetonitrile þ 0.1 M (n-Bu)4NPF6 at 10 C under Vacuum Fullerene H2@C60 C604
20
E1/2 red1
E1/2 red2
E1/2 red3
E1/2 red4
E1/2 red5
E1/2 red6
0.95 0.95
1.37 1.37
1.89 1.88
2.39 2.35
2.95 2.88
3.5 3.35
phosphorus,17 or metallic atoms18 and trimetallic nitride clusters19 have been shown to be stabilized inside carbon cages. Therefore, these cages can be considered as a kind of ideal “chemical Faraday cages,” and unique cases of supramolecular structures, since covalent bonds are normally not involved between the cage and the encapsulated moiety. Frequently, endohedral fullerenes consist of carbon cages whose size or symmetry is not obtained in large abundance when empty. The supramolecular interaction between the encapsulated moiety and the carbon cage usually affects the electronic properties, the stability, and the chemical reactivity of both the encapsulated moiety and the carbon cage, and their electrochemical behavior reflects these intimate relationships between encaged species and cage. Only a few reports describe the redox behavior of molecules encapsulated in fullerenes cages. However, electrochemical studies on fullerenes encapsulating metallic atoms and clusters abound. 8.2.1
Encapsulation of Molecular Hydrogen: H2@C60
Molecular hydrogen has been successfully encapsulated by Komatsu et al.20 using an elegant “molecular surgery” approach initially proposed by Wudl21 and Rubin.22 A 13-membered ring orifice was dug on the surface of the C60 cage by chemical reaction, followed by the thermal insertion of molecular hydrogen, the subsequent reduction of the orifice size by chemical reaction, and the final restoration of the C60 cage by thermal annealing. The electrochemical behavior of H2@C60 was reported using the same conditions as used for C60 by Echegoyen et al.,4 that is, toluene–acetonitrile at 10 C under vacuum (see Table 8.2). The slightly electropositive character of molecular hydrogen shifts cathodically the last three reduction waves of the C60 cage. 8.2.2
Monometallofullerenes: M@C2n
The first metallofullerenes, La@C2n, were discovered by Smalley et al.18,23 after laser vaporization of composite targets made of graphite and lanthanum oxide or chloride. Because of their low-yield synthesis, laborious purification, and often air sensitivity and kinetic instability, studying the physical properties and chemical reactivity of these fascinating compounds was a serious challenge. Fortunately, the high sensitivity of the electrochemical methods was well adapted to study the microgram quantities in which these materials were usually available. The series of M@C82
206
ELECTROCHEMISTRY OF CARBON NANOPARTICLES
studied by Suzuki,24–30 Akasaka,30–32 Gu,33–36 Dunsch,37,38 and Wang et al.39 pointed out the influence of the metal on the electrochemical properties of these compounds. The series of Yb@C2n and Ca@C2n studied by Gu et al.40,41 described the influence of the cage size and structure on the electrochemical properties. 8.2.2.1 M@C82 Series: Influence of the Metal Following the pioneer work of Smalley et al. on La@C82, a lot of different M@C82 compounds have been synthesized in recent years10 with metals of the p block (Ca and Ba), d block (Sc and Y), and f block (La, Ce, Pr, Gd, Tm, Ho, Dy, and Yb), making the family of metallofullerenes with a C82 cage one of the richest among all metallofullerenes. These materials were usually produced by arcing graphite rods doped with a metal oxide or carbide, and purified by multistep HPLC methods. Two different redox behaviors have been observed when studying these compounds in solution (see Table 8.3). The electrochemistry of films of some of these metallofullerenes drop-cast on the surface of electrodes has also been reported43–48 but these results are beyond the scope of this review. A first group of compounds Ce@C82, Gd@C82, Y@C82, and the major [C2v] and minor [Cs] isomers of La@C82 and Pr@C82 showed two oxidation steps, the first reversible and the second irreversible, even at scan rates up to 1 V/s. The low potential of the first oxidation step, close to that of the ferrocene/ferrocenium couple (see Table 8.3 and Fig. 8.4), made these compounds rather good electron donors. These compounds could also be reduced in four to six distinct steps, most of them reversible, and their reducing ability was found even higher than that of C60 and similar to that of the major isomer of C82 (C2). Noticeably, all these compounds had a very low electrochemical HOMO–LUMO gap (DEgap < 0.50 V). In addition, similar UV/Visible spectra were obtained for all of them,28 suggesting also similar electronic structures. ESR showed that Y@C8229 and both isomers of La@C8249–52 are radical species and consequently that the formal oxidation state of the metal in these structures is probably þ 3. Therefore, their low HOMO–LUMO gap is probably a consequence of their open-shell electronic structure. The presence of the metal affects the number and shape of the cathodic waves (Fig. 8.4, voltammograms a and c). In particular, the second reduction step is either a two-electron process, or two close but successive one-electron processes. The metal also slightly influences the potentials of the anodic and cathodic steps. A good linear relationship was found between the first reduction and first oxidation potentials and the ionic radius of the metal (rM3 þ ).28 Presumably, the metal ion is not at the center of the cage and the SOMO of the compound has its largest electron density on the part of the cage close to the metal ion. The smaller metal ion of the series, Y, can come closer to the cage, and therefore the electrons on the SOMO are more tightly bound. As a consequence, Y@C82 has the lowest electron-donating ability and the highest electron-accepting ability of the series. In contrast, a very different redox behavior was found for Sm@C82, and all of the isolated isomers of Ca@C82, Yb@C82, and Tm@C82 (see Table 8.3). In particular, no oxidation was recorded for these compounds, indicating a larger HOMO–LUMO gap. It was concluded from their diamagnetism that the metal adopts a þ 2 formal oxidation
207
1.50 0.80
1.12 0.47
C6028 (c) C82 [C2]42 (f), (a)
1.95 1.42
1.52 1.58 1.33 1.56 1.30 1.55
1.53 2.01 1.46 1.99 1.53 2.22 2.05 2.22
E1/2 red3
2.41 1.84
1.88 1.81 1.73 1.90 1.70 1.90
2.17 2.47
2.26 2.40 2.21 1.99 1.79
E1/2 red4
2.03
2.35
2.25
2.46
E1/2 red5
2.32
2.58
2.50
E1/2 red6
2.24 >1.77
>1.08 >1.63 >1.90 >1.63 >1.89 >1.95
0.49 0.40 0.46 0.41 0.49 0.48 0.48 0.44
DEgap (b)
(a) differential pulse voltammetry peak potential; (b) DEgap ¼ E1/2 ox1 E1/2 red1; (c) o-DCB (1,2-dichlorobenzene) þ 0.1M (n-Bu)4 NPF6, (d) acetonitrile/toluene (1/ 4) þ 0.1 M (n-Bu)4 NPF6, (e) acetonitrile/toluene (1/4) þ 0.1 M (n-Bu)4 NClO4, or (f) pyridine þ 0.1M (n-Bu)4 NClO4; numbers inside square brackets represent the cage isomer labels.
1.12
0.07 0.07 0.07 0.07 0.08 0.09 0.20 0.10 0.63 0.65 0.76 0.67 0.74 0.96
1.07
1.07 1.08 1.08 1.05 1.08 1.08
0.28 0.33 0.60 0.33 0.59 0.65
1.05 1.05 1.00 1.00 1.00 0.95 0.95 0.90
Sm@C82 [C2v]35 (d) Yb@C82 [Cs]40 (e) Yb@C82 [C2]40 (e) Yb@C82 [C2v]40 (e) Ca@C82 [C2]41 (e) Ca@C82 [II]41 (e)
E1/2 ox1 1.37 1.40 1.35 1.39 1.41 1.38 1.25 1.34
E ox2 (a) 0.42 0.47 0.39 0.48 0.41 0.39 0.25 0.37
28
La@C82 [C2v] (c) La@C82 [Cs]31 (c) Pr@C82 [C2v]31 (c) Pr@C82 [Cs]31 (c) Ce@C82 [C2v]28 (c) Gd@C82 [C2v]28 (c) Gd@C82 [C2v]35 (d) Y@C82 [C2v]28 (c)
Fullerene
E1/2 red2
rM 3 þ
Half-wave Redox Potential Unless Otherwise Stated (in V Versus Fc þ /Fc) of the M@C82 Compounds E1/2 red1
TABLE 8.3
208
ELECTROCHEMISTRY OF CARBON NANOPARTICLES
0.4 μA
(c)
(b) 0.3 μA
(a) 0.3 μA
1.5
1
0.5
0
–0.5 –1 –1.5
–2
–2.5
E (V) versus Fc/Fc+
Figure 8.4 Differential pulse voltammetry of the major (a) and minor isomer (b) of La@C82 and Y@C82 (c) in o-DCB þ 0.1 M of (n-Bu)4NPF6; pulse amplitude: 50 mV, pulse width: 50 ms, pulse period: 200 ms, scan rate 20 mV/s. Reproduced from Ref. 26, with permission from Elsevier.
state in these compounds. Interestingly, their electron-accepting ability is also higher than that of most empty fullerenes, in spite of the 2 charge already present on the cage. From the studies of M@C82 compounds, we can therefore conclude that the formal oxidation number of the entrapped metal and its size, which determines its position inside the cage, have a clear influence on the redox properties of the metallofullerenes. 8.2.2.2 The Yb@C2n and Ca@C2n Series: Influence of the Cage The two isomers of La@C82 and Pr@C82 described previously showed a difference of electron-donating ability of 0.14 V (see Table 8.3) along with disparities in the shape and number of cathodic waves (Fig. 8.4, voltammograms a and b). Therefore, the electrochemistry of metallofullerenes is also highly dependent on the structure of the carbon cage. Similar conclusions were drawn by Gu et al. from the study of the Yb@C2n (37 n 42)40 and Ca@C2n (38 n 42)41 families (see Table 8.4). No anodic signal was observed for any of these compounds, which is an indication of a rather large HOMO–LUMO gap. Their redox behavior was interpreted assuming a þ 2 formal oxidation state for the Yb or Ca atoms and therefore a closed-shell electronic configuration. The electrochemical gap between the third and second cathodic events is larger than between both the second and first events, and the fourth and third events, which suggests that the LUMO of these compounds is nondegenerate.
8.2
ELECTROCHEMISTRY OF ENDOHEDRAL FULLERENES
209
TABLE 8.4 Half-Wave Redox Potential Unless Otherwise Stated (in V Versus Fc þ /Fc) of the Yb@C2n, Ca@C2n Compounds and Some Empty Fullerenes Obtained in Acetonitrile/Toluene (1/4) þ 0.1 M (n-Bu)4NClO4 Unless Otherwise Stated E1/2 red1
E1/2 red2
E1/2 red3
E1/2 red4
Yb@C74 [II] Yb@C76 [I]40 Yb@C76 [II]40 Yb@C7840 Yb@C8040 Yb@C82 [Cs]40 Yb@C82 [C2]40 Yb@C82 [C2v]40 Yb@C84 [II]40 Yb@C84 [III]40 Yb@C84 [IV]40 (a)
0.52 0.46 0.68 0.48 0.57 0.33 0.60 0.33 0.63 0.49 0.46
0.96 0.83 1.02 0.79 0.95 0.65 0.76 0.67 0.88 0.68 0.72
1.55 1.46 1.59 1.46 1.55 1.58 1.33 1.56 1.26 1.57 1.34
1.99 1.89 2.01 1.83 1.90 1.81 1.73 1.90 1.64 1.79 1.54
Ca@C7641 Ca@C82 [C2]41 Ca@C82 [II]41 Ca@C84 [II]41
0.61 0.65 0.59 0.64
0.99 0.96 0.74 0.90
1.57 1.55 1.30 1.27
1.97 1.90 1.70 1.65
C7653 (a) C78 [C2v]53 (a) C82 [C2] 42 (b), (c) C84 [D2d]54 (b), (c) C84 [D2]54 (b), (c)
0.83 0.72 0.47 0.52 0.61
1.17 1.08 0.80 0.84 0.97
1.68 1.79 1.42 1.30 1.20
2.10 2.18 1.84
Fullerene 40
(a) Acetonitrile/toluene (1/4) þ 0.1 M (n-Bu)4NPF6; (b) pyridine þ 0.1 M (n-Bu)4NClO4; (c) differential pulse voltammetry peak potential.
A first striking observation of this study was the dramatic difference in electronaccepting abilities of isomeric endohedral fullerenes. This difference reaches 0.22 V for the isomers of Yb@C76 and even 0.3 V between two isomers of Yb@C82. The cage size also influences the reduction potentials. The [Cs] and [C2v] isomers of Yb@C82 are the best electron acceptors of the family and the second isomer of Yb@C76 the weakest electron acceptor. In general, however, these metallofullerenes are better electron acceptors than the available empty cages. The metallofullerenes with bigger cages, Yb@C82 and Yb@C84, also accept more easily a fifth electron than the others. 8.2.2.3 Other Monometallofullerenes: Eu@C74 and Tm@C78 Dunsch et al. performed the electrochemistry of Eu@C7455 and Tm@C7856 in a glove box under inert conditions. Based on Raman measurements and the absence of an ESR signal, it was concluded that the formal oxidation number of the metal is þ 2 in Eu@C74. Interestingly, in addition to four reversible and monoelectronic reduction steps, two anodic events were also observed, contrary to all the other metallofullerenes with divalent metals described previously (Fig. 8.5). Therefore, Eu@C74 has a significantly
210
ELECTROCHEMISTRY OF CARBON NANOPARTICLES
7.000
Current i (nA)
5.000 3.000 1.000 –1.000 –3.000 –5.000 –2.250 –1.750 –1.258 –0.750 –0.250
0.250
0.750
1.250
1.75
Potential E (V)
Figure 8.5 Cyclic voltammogram of Eu@C74 in o-DCB þ 0.1 M (n-Bu)4NBF4; scan rate 100 mV/s. Reproduced from Ref. 55, with permission from Elsevier.
lower electrochemical HOMO–LUMO gap (close to 1 V). The four reduction events are almost equidistantly separated as those of C60, which suggests that this compound has a doubly degenerated LUMO. This is consistent with a transfer of two electrons from the metal to the cage, since the empty cage C74 has a nondegenerated LUMO and a doubly degenerated LUMO þ 1. The electrochemistry of Tm@C78 was quite similar with at least two reversible reduction steps and two anodic events. However, its HOMO–LUMO gap was even lower than that of Eu@C74, which could explain the lower kinetic stability of this compound. 8.2.3 Multimetallofullerenes M2@C2n and Metal Carbide Endohedral Fullerenes M2C2@C2n and M3C2@C2n Dorn et al. reported the electrochemistry of two isomers of Sc2@C82 and
[email protected] Hoffman et al. recognized the signature of Er3 þ ions in the near-infrared emission spectra of some erbium endohedral fullerenes,58 suggesting that six electrons are transferred from the bimetallic moiety to the cage in these structures. Considering the higher charge on the cage in these compounds, it was expected that their electron-accepting ability would be significantly lower than those of the empty fullerenes and monometallofullenes. However, these compounds (see Table 8.5) were better electron acceptors than anticipated, comparable to C60. The similarities in the redox properties of Er2@C82 and those of isomer [III] of Sc2@C82 suggest that these compounds share the same cage structure. La2@C80 described by Nagase et al.50,59 showed an even better electron-accepting ability, which was connected to its fair reactivity toward nucleophiles. Calculations showed that the LUMO of this
8.2
211
ELECTROCHEMISTRY OF ENDOHEDRAL FULLERENES
TABLE 8.5 Square Wave Voltammetry or Differential Pulse Voltammetry Peak Potentials (in V Versus Fc þ /Fc) of Multimetallofullerenes and Metal Carbide Fullerenes Fullerene 60
La2@C72 [D2] (a) La2@C80 [Ih]59 (a) Sc2@C82 [I]57 (b) Sc2@C82 [III]57 (b) Er2@C8257 (b) Sc2C2@C82 Sc2C2@C82 Sc3C2@C80 Sc3C2@C80
[III]57 (b) [III]64 (a) [Ih]65 (a) [Ih]66 (b)
C82 [C2]42 (b)
E red1
E red2
0.68 0.31 1.26 0.87 0.87
1.92 1.72 1.88 1.29 1.26
0.95 0.94 0.50 1.42
1.38 1.64 1.67
1.82
0.47
0.80
1.42
E red3
E red4
2.13 1.85
E ox1
E ox2
DEgap
0.24 0.56 0.12 0.07 0.19
0.75
0.92 0.87 1.14 0.94 1.06
0.16 0.47 0.06 0.32 1.84
0.10
0.04
1.11 1.42 0.44 1.38 >1.77
(a) o-DCB þ 0.1 M (n-Bu)4NPF6 or (b) pyridine þ 0.1 M (n-Bu)4NClO4; DEgap ¼ E ox1 E red1.
compound is localized on the bimetallic center. Recently, Nagase et al.60 also reported the electrochemistry of the non-IPR (isolated pentagon rule) metallofullerene
[email protected] This compound showed two reversible reduction waves and two reversible oxidation waves; its electron-accepting ability was weaker than that of La2@C80, but its electron-donating ability was stronger (see Table 8.5). The study of Sc2C2@C82 [C3v] and Sc3C2@C80 [Ih] began with a misunderstanding. Their redox properties were first interpreted as if these compounds were Sc2@C84 and Sc3@C82 before Nagase et al. showed that they were in reality metal carbide endohedral fullerenes.62,63 Whereas no clear relationship was found between the redox behavior of Sc2C2@C82 and the empty C84 [D2d], there were some similarities that could not be explained between the voltammograms of Sc2C2@C82 and Er2@C82 and those of isomer [III] of Sc2@C82 (see Table 8.5).57 These similarities are now well understood assuming that these species share the same C82 cage. This would be consistent with a separate 13 C NMR study that showed that Y2C2@C82 and Y2@C82 have the same C82 [C3v] cage.67 The electronic structure of this compound was established by DFT calculations64 to be (Sc2C2)4 þ C824. Sc3C2@C80 is paramagnetic and its very small electrochemical gap in 1,2-dichlorobenzene65 was ascribed to its open-shell electronic structure. The charge state of this compound was established by DFT calculations to be (Sc3 þ )3C23C806,68 and interestingly the redox state of the molecule does not seem to influence the formal oxidation state of the Sc atoms or the charge on the cage, but affects directly the formal charge on the C2 moiety. Therefore, the C2 moiety displays a remarkable flexibility of charge, varying from þ 2 (first oxidation) to 3 (third reduction). The contrasting redox behavior of Sc3C2@C80 in pyridine (see Table 8.5) could originate from a reduction of Sc3C2@C80 by pyridine yielding the electronic closed-shell species Sc3C2@C80 in this solvent.65
212
ELECTROCHEMISTRY OF CARBON NANOPARTICLES
8.2.4
Trimetallic Nitride Endohedral Metallofullerenes
Among all the endohedral fullerenes prepared and described so far, the trimetallic nitride templated endohedral metallofullerenes (TNT EMFs) hold special importance, due to their high-yield synthesis and high kinetic stability in air. The archetype of these new fullerene derivatives is Sc3N@C80 [Ih], which was discovered in 1999 by Dorn and coworkers19 and can now be prepared on a gram scale by arcing graphite rods packed with scandium oxide under a mixture of nitrogen/helium19,69 or ammonia/helium70 in a Kratschmer–Huffman reactor. Interestingly, neither the cluster Sc3N nor the [Ih] C80 cage has been isolated separately. The stability of the compound is ascribed to the supramolecular interaction between the cluster and the cage, which has been shown to be mainly electrostatic in nature. It is in effect now widely accepted that the cluster stabilizes the cage by transferring six electrons, and therefore the interaction between the encapsulated cluster and the cage can be reasonably well described by the ionic model: Sc3N6 þ C806.71 Without this electronic transfer, the C80 cage would have an open-shell configuration,72 and would be kinetically unstable, as discussed before by Alford and coworkers9 (see above). Since the discovery of Sc3N@C80, the family of TNT EMFs has been expanded considerably. The groups of Dorn, Echegoyen, and Dunsch have shown that two other metals of the group III, Y73,74 and La,75 and most of the lanthanide metals (Er,19,73 Ho,74,76,77 Lu,78 Tm,79,80 Gd,81–84 Dy,85–87 Tb,74,76,77,88,89 Nd,90,91 Pr,91 and Ce91) can also form trimetallic nitride endohedral fullerenes. In addition, mixed metal TNT EMFs have also been prepared.19,74,92–97 One interesting aspect displayed by this rich family of compounds is the influence of the cluster size on the cage-size distribution (Table 8.6). It has been shown that up to the size of Gd3N, the most abundant cage obtained is the C80 cage of [Ih] symmetry. In contrast, Nd3N, Pr3N, and Ce3N prefer to be encapsulated by a C88 cage. Finally, it was recently observed that La3N shows a preference for the C96 cage. However, in addition to the most abundant cage, usually a wide distribution of cage size is formed in each case. In the Sc3N@C80 synthesis, smaller cages such as C68,96,98 C70,99 and C78100,101 have also been isolated and studied. With Dy, Tb, and Gd, larger cages such as C82, C84, C86, and C88 have also been studied, and cages up to C98 have been detected. In the case of Nd and Pr, smaller cages such as C80, C84, and C86 and larger cages such as C96 have been reported.90,91 With Ce, larger cages than C88 are also formed and the C96 has been isolated and studied. Due to the large number of metals that can form TNT EMFs and the usually wide cage-size distribution obtained in each case, the number of different compounds that have been described and studied by electrochemistry is quite large. All of these new fascinating compounds usually show very rich redox properties and therefore have been considered promising candidates for molecular electronic applications, such as photovoltaics. The redox behavior of these compounds is influenced by the size and symmetry of the carbon cage and the nature of the metal. 8.2.4.1 Sc3N@C80: Influence of the Carbon Cage Symmetry Sc3N@C80 is the first TNT EMF historically prepared, and numerous electrochemical studies are now available for this compound.59,103–108. The purification by HPLC of the soot
213
a
3.6 C80 C82–C88
Y3N 3.0 C80 C78, C82–C98
Dy3N 3.70 C80 C82–C88
Tb3N 3.8 C80 C82–C98
Gd3N 3.9 C88 C80–C86 C90–C100
Nd3N
a
The size of the cluster was estimated from the size of the metallic cation (M3 þ )102 and a published formula.82.
3.0 C80 C68, C70, C78
Sc3N
Influence of the Size of the Cluster on the Cage-Size Distribution
Estimated diameter, A Most abundant cage Other cages obtained
Cluster
TABLE 8.6
4.0 C88 C80–C86 C90–C104
Pr3N
4.0 C88 C86 C90–C106
Ce3N
4.2 C96 C86, C88 C90–C94 C98–C110
La3N
214
ELECTROCHEMISTRY OF CARBON NANOPARTICLES
Figure 8.6 Cyclic voltammograms of the mixture of [Ih] and [D5h] isomers of Sc3N@C80 (a) and of the pure [Ih] isomer of Sc3N@C80 (b) obtained in o-DCB þ 0.05 M (n-Bu)4NPF6 (scan rate 100 mV/s). Reproduced from Ref. 117, with permission from Elsevier.
extract obtained upon arcing graphite rods packed with Sc2O3 yields a mixture of two isomers, the predominant [Ih] isomer and the less abundant [D5h]. The electrochemistry of this mixture of isomers has been reported (Fig. 8.6a). The first surprising observation for Sc3N@C80 was its ability to accept electrons despite the already high negative charge on the cage. Sc3N@C80 can be reduced in at least three irreversible reduction steps. The anodic part of the scan shows the superposition of the signals of the two isomers present in the mixture. Interestingly, the D5h isomer is easier to oxidize than the icosahedral one by about 0.3 V, which shows that the redox properties of these new endohedral fullerenes are also dramatically influenced by the symmetry of the carbon cage. This observation led Echegoyen et al. to carry out a separation of the two isomers, based on their different oxidation potentials.104 By choosing a chemical oxidant with a redox potential intermediate between that of the two isomers (the tris (p-bromophenyl)aminium hexachloroantimonate or “magic blue”), the [D5h] isomer could be preferentially oxidized and removed from the mixture by simple filtration over SiO2. This procedure allowed Echegoyen et al. to report the electrochemistry of the pure icosahedral isomer of Sc3N@C80 for the first time (Fig. 8.6b). The main isomer of Sc3N@C80 displays two oxidation steps, and the first one is reversible even at low scan rates (0.1 V/s) in o-DCB. In contrast, Sc3N@C80 can be reduced in at least three irreversible reduction steps in o-DCB at 0.1 V/s. The first reduction step displays two broad and anodically shifted
8.2
ELECTROCHEMISTRY OF ENDOHEDRAL FULLERENES
215
reoxidation waves. The separation between the second and third reductions is bigger than between the first and second reduction, as can be expected from the molecular orbital diagram published by Poblet et al., which shows nondegenerate LUMO orbitals.109 These reduction steps become reversible when scanning at faster scan rates, suggesting that a chemical step follows the electronic transfer (EC mechanism). Recently, it was also shown that reversibility can be improved by changing the solvent to a mixture of acetonitrile/toluene.107 The nature of the chemical step following the electronic transfer remains unclear at the time of writing this review. The bulk electrolysis of Sc3N@C80 at a potential corresponding to the first reduction, followed by a reoxidation at 0 V, yields the starting material, suggesting that this chemical step is reversible. The radical anion was generated either by chemical reduction with Na/K or by a one-electron bulk electrolysis. Its EPR spectrum consisted of 22 lines,104,110 assigned to 3 equivalent Sc nuclei with a large hyperfine splitting of 55.6 G. This suggests that the C3 symmetry of the endohedral cluster is conserved in the radical anion and that the spin is exclusively localized in the Sc3N cluster. One possible explanation is that the cluster pyramidalizes upon electron transfer, thus retaining the C3 symmetry. Further work is needed to identify the nature of the chemical process following the first electron reduction of Sc3N@C80. 8.2.4.2 The M3N@C80 series: Influence of the Metal All the group III and lanthanide metals from the size of scandium to the size of praseodymium have been shown to form M3N@C80 [Ih]. Up to the size of gadolinium, the M3N@C80 [Ih] is the most abundant species formed when arcing graphite rods packed with the respective metal oxide under a mixture of nitrogen/helium or ammonia/helium. The electrochemistry of all the M3N@C80 [Ih] studied so far in o-DCB is very similar to that of Sc3N@C80 with irreversible reduction steps and reversible oxidation steps. Similarly to Sc, a change in shape or position of the cluster inside the cage upon reduction could also explain the irreversible behavior, as discussed by Dunsch et al. in the case of Dy.85–87 The nature of the metal has a notable influence on the reduction potentials and a less marked influence on the oxidation potentials (see Table 8.7), suggesting a possible contribution of the trimetallic nitride cluster to the LUMO of the TNT EMF Ih and a less pronounced contribution to the HOMO. However, recent calculations suggest that only in the case of scandium, the cluster contributes notably to the LUMO.71 Therefore, to explain the effect of the metal on the redox potentials, it is proposed that the charge transfer between the cluster and the cage depends on the electronic properties and size of the metal. The effective metal valences that were determined by high-energy spectroscopic studies in some of these compounds show that scandium,111 the smallest and most electronegative metal of the series, seems to transfer less electrons to the cage than Tm112 or Dy.87 This could explain why Sc3N@C80 is the easiest of the series to reduce. In contrast, the more difficult to reduce Gd3N@C80, Nd3N@C80 and Pr3N@C80, are made from the largest and least electronegative metals of the series. Interestingly, Sc3N@C80 is also the easiest species to reduce in the D5h isomer family (see bottom of Table 8.7). However, as ScYErN@C8097 does not show a reduction ability
216
ELECTROCHEMISTRY OF CARBON NANOPARTICLES
TABLE 8.7 Relevant Redox Potentials (in V Versus Fc þ /Fc) of M3N@C80 Compounds Obtained in o-DCB þ 0.05 M (n-Bu)4NPF6 Metal Electronegativity102
Compound 105
Sc3N@C80 [Ih] Sc3N@C80 [Ih]103 Sc3N@C80 [Ih]59 Lu3N@C80 [Ih]108 Tm3N@C80 [Ih]80 Er3N@C80 [Ih]105 Y3N@C80 [Ih]105 Dy3N@C80 [Ih]113 Gd3N@C80 [Ih]84 Nd3N@C80 [Ih]91 Pr3N@C80 [Ih]91 ScYErN@C80 [Ih]97
1.36 1.36 1.36 1.27 1.25 1.24 1.22 1.22 1.20 1.14 1.13
Sc3N@C80 [D5h]108 Lu3N@C80 [D5h]108 Tm3N@C80 [D5h]80 Dy3N@C80 [D5h]113
1.36 1.27 1.25 1.22
Metal Effective Valency 111
2.4 2.4111 2.4111 2.9112
2.887
Epc red1
Epc red2
E1/2 ox1
1.29 1.24 1.22 1.40 1.43 1.42 1.41 1.37 1.44 1.42 1.41 1.55
1.56 1.62 1.59 1.78 1.80 1.83 1.86 1.86 1.89 1.84 1.97
0.59 0.62 0.62 0.64 0.65 0.63 0.64 0.70 0.58 0.63 0.59 0.64
1.85
0.34 0.45 0.39 0.40
1.33 1.41 1.45 1.40
intermediate between that of Sc3N@C80, Y3N@C80, and Er3N@C80 (see Table 8.7), the exact interplay between the nature of the cluster and the electrochemical behavior of these new compounds is far from obvious. 8.2.4.3 Decreasing the Size of the Cage: Sc3N@C68, Sc3N@C78, and Dy3N@C78 Another influence of cage size and symmetry on the redox properties of these compounds was demonstrated in a study of the smaller cages: C68 obtained in the case of scandium and C78 in the case of scandium and dysprosium (Table 8.8).
TABLE 8.8 Cathodic Peak Potentials and Anodic Half-Wave Potentials (in V Versus Fc þ /Fc) of Sc3N@C68, Sc3N@C78, and Dy3N@C78 Obtained by Cyclic Voltammetry in o-DCB þ 0.05 M (n-Bu)4NPF6 from the Cyclic Voltammograms; Comparison with M3N@C80 Compound Sc3N@C80 [Ih] Sc3N@C68114 Sc3N@C78115
103
Dy3N@C80 [Ih]113 Dy3N@C78113
E1/2 ox1
E1/2 ox2
1.62 2.05
0.62 0.33 0.12
0.85
1.86 1.93
0.70 0.47
Epc red1
Epc red2
1.24 1.45 1.54 1.37 1.54
8.2
ELECTROCHEMISTRY OF ENDOHEDRAL FULLERENES
217
Sc3N@C68 displays an interesting non-IPR cage [D3]96,98, and therefore its electronic structure can be expected to be very different from that of Sc3N@C80. Sc3N@C68 studied by Dunsch et al.114 upon reduction displayed a behavior similar to that of Sc3N@C80, with two irreversible steps, cathodically shifted, and comparable to that of Sc3N@C80 (see Table 8.8). The main difference between the two compounds could again be observed upon oxidation. The electron-donating ability of Sc3N@C68 was found to be significantly . greater than that of Sc3N@C80 [Ih]. The radical cation114 [Sc3N@C68] þ and 116 þ. [Sc3N@C68] were generated by electrolysis and both radical anion characterized by ESR. The ESR spectra obtained consisted of 22 lines and indicated 3 equivalent Sc with a small hyperfine splitting of 1.28 G for the cation and 1.75 G for the anion. Based on the smaller hyperfine splitting, as . compared to that of [Sc3N@C80] , a delocalization of the spin on both the cluster and the cage was proposed, which confirmed a major contribution of the cage to the HOMO of these compounds. Dunsch et al. also reported the redox behavior of Dy3N@C78113 and very recently Dorn et al. reported that of
[email protected] These TNT EMFs showed two irreversible reduction steps and one reversible oxidation step. Their electron-accepting ability was also significantly weaker and their electron-donating ability stronger than that of the parent TNT EMF with a C80 [Ih] cage. Yang et al.107 also reported the electrochemistry of Sc3N@C78, and, unlike Dorn et al., they found reversible waves in the cathodic region. The reversibility of their results however appears to be questionable as the faradic current is low, thus making it difficult to interpret the waves. 8.2.4.4 Increasing the Size of the Cage: The Gd3N@C2n (40 n 44), Nd3N@C2n (40 n 44), Pr3N@C2n (40 n 48), Ce3N@C2n (44 n 48), and La3N@C2n (44 n 48) Series An even more dramatic effect of the cage size was observed when studying the electrochemistry of the Gd3N@C2n family with cages ranging from C80 to C88 (Fig. 8.7 and Table 8.9).84,117 Gd3N@C82, Gd3N@C84, and Gd3N@C86 were found to have a redox behavior qualitatively similar to that of Gd3N@C80 with three irreversible reduction steps and at least one reversible oxidation step, in spite of dramatic differences in their cage structure. Gd3N@C80 [Ih] is a IPR fullerene81. In contrast, Gd3N@C84118 was recently shown to have the same non-IPR cage structure as Tb3N@C84119 and
[email protected] Gd3N@C88 showed a quite unexpected reversible behavior in both its reduction and oxidation. Remarkably, the size of the cage does not affect significantly the reduction ability of these compounds, which displayed very close first reduction potentials (see Table 8.9). In contrast, increasing the size of the cage pronouncedly eases the oxidation ability of these compounds. These results confirmed that the nature of the cage contributes significantly to the HOMO of the TNT EMFs, whereas the LUMO is more influenced by the encapsulated cluster. Consequently, a significant and systematic lowering of the HOMO–LUMO gap is observed when increasing the cage size.
218
ELECTROCHEMISTRY OF CARBON NANOPARTICLES
Figure 8.7 Cyclic voltammograms of the Gd3N@C2n (40 n 44) compounds obtained in o-DCB þ 0.05 M (n-Bu)4NPF6 (scan rate 0.1 V/s). Reproduced from Ref. 117, with permission from Elsevier.
Similar observations on the reversibility of the redox steps and on the lowering of the HOMO–LUMO gap (see Table 8.10) upon increasing cage size were made in the case of the Nd3N@C2n and Pr3N@C2n families.91 The characteristic features of Gd3N@C88, that is, its reversible behavior in both oxidation and reduction steps and its very low HOMO–LUMO gap were found to be a result of the trimetallic nitride clusters encapsulated in a C88 cage, since the Nd,90 Pr,91 and Ce91 counterparts showed the same behavior (see Fig. 8.8 and Table 8.11). This series of M3N@C88 does not display as many differences in their oxidation and reduction potentials as the M3N@C80 series. This is presumably due to TABLE 8.9 Relevant Redox Potential (in V Versus Fc þ /Fc) of the Gd3N@C2n Compounds Obtained in o-DCB þ 0.05 M (n-Bu)4NPF6 TNT EMF
Epc red1
Gd3N@C80 [Ih]84 Gd3N@C82117 Gd3N@C84 [Cs]84 Gd3N@C86117 Gd3N@C8884
1.44 1.53 1.37 1.39 1.43
DEgap ¼ E1/2 ox1 Epc red1.
Epc red2
E1/2 ox1
1.86 1.87 1.76 1.72 1.74
0.58 0.38 0.32 0.33 0.06
E1/2 ox2
DEgap
0.49
2.02 1.91 1.69 1.72 1.49
8.2
219
ELECTROCHEMISTRY OF ENDOHEDRAL FULLERENES
TABLE 8.10 Relevant Redox Potential (in V Versus Fc þ /Fc) of Nd3N@C2n and Pr3N@C2n Compounds Obtained in o-DCB þ 0.05 M (n-Bu)4NPF6 TNT EMF
Epc red2
E1/2 ox1
1.42 1.44 1.46 1.36
1.90 2.02 1.79 1.75
0.63 0.31 0.36 0.07
1.41 1.48 1.34
1.81 1.80 1.72
0.59 0.31 0.09
Epc red1
Nd3N@C80[Ih] Nd3N@C8491 Nd3N@C8691 Nd3N@C8891
91
91
Pr3N@C80[Ih] Pr3N@C8691 Pr3N@C8891
E1/2 ox2
DEgap
0.53
2.05 1.75 1.82 1.43
0.54
2.00 1.79 1.43
DEgap ¼ E1/2 ox1 Epc red1.
the very similar electronic properties of Gd, Nd, Pr, and Ce (see the electronegativity values for these metals in Table 8.7) and a main contribution of the cage to both the HOMO and LUMO of the compounds. It is notheworthy and surprising that such low HOMO–LUMO gap fullerenes are the most abundant species obtained in the soluble extract of the arcing soot in the case of Nd, Pr, and Ce! Only La3N@C88 showed a different behavior with irreversible reduction steps and a slightly larger HOMO– LUMO gap (see Table 8.11).75 Very recently, Echegoyen et al. were also able to report the electrochemistry of M3N@C96 (M ¼ Pr, Ce, and La).75 These compounds could be oxidized in two reversible steps, like their C88 counterparts, and reduced in at least three irreversible
Figure 8.8 Cyclic voltammograms of M3N@C88 obtained in o-DCB þ 0.05 M (n-Bu)4NPF6 (scan rate 0.1 V/s). Reproduced from Ref. 117, with permission from Elsevier.
220
ELECTROCHEMISTRY OF CARBON NANOPARTICLES
TABLE 8.11 Half-Wave Redox Potential (in V Versus Fc þ /Fc) of the M3N@C88 Compounds Obtained from Cyclic Voltammetry in o-DCB þ 0.05 M (n-Bu)4NPF6 Metal Electronegativity102 E1/2 red1 E1/2 red2
Compound 84
Gd3N@C88 Nd3N@C8890 Pr3N@C8891 Ce3N@C8891 La3N@C8875
1.20 1.14 1.13 1.12 1.10
1.40 1.33 1.31 1.30 1.36a
1.71 1.66 1.64 1.57 1.67
E1/2 ox1 0.06 0.07 0.09 0.08 0.21
E1/2 ox2 DEgap 0.50 0.53 0.54 0.63 0.66
1.46 1.40 1.40 1.38 1.57
DEgap ¼ E1/2 ox1 Epc red1. a Cathodic peak potential.
TABLE 8.12 Relevant Redox Potential (in V versus Fc þ /Fc) of the M3N@C96 Compounds Obtained in o-DCB þ 0.05 M (n-Bu)4NPF6 Compound
Metal electronegativity102 Epc red1 Epc red2 E1/2 ox1 E1/2 ox2 DEgap
Pr3N@C9675 Ce3N@C9675 La3N@C9675
1.13 1.12 1.10
1.51 1.50 1.54
1.86 1.84 1.77
0.14 0.18 0.14
0.53 0.67 0.53
1.65 1.68 1.68
DEgap ¼ E1/2 ox1 Epc red1.
steps. Within this family of C96 fullerenes, the corresponding redox potentials were also very similar and quite unexpectedly, the electrochemical bandgap was found larger than that of the M3N@C88 compounds (see Table 8.12). In conclusion, the electrochemistry of endohedral fullerenes is usually influenced by the nature of the encapsulated moiety and the size and symmetry of the cage. If the limit of kinetic stability of empty fullerenes, along with the large increase of isomers when the size of the cage increases, prevented the study of isomerically pure fullerenes larger than C84, the encapsulation of species as large as trimetallic nitride clusters allows the study of larger fullerenes such as C88, C96, and in future probably C104. However, giant fullerenes such as C240, C540, and so on still remain undetected. The study of larger carbon nanostructures, such as carbon nanotubes or carbon nanoonions, might be a way to approach the properties of these giant fullerenes. Unlike the fullerenes, the other allotropes of carbon such as carbon nanotubes possess larger aspect ratios and are therefore not soluble. Thus, their electrochemical behavior in solution cannot be studied. Functionalization of these nanoparticles results in soluble derivatives that can be analyzed by electrochemistry.
8.3
ELECTROCHEMISTRY OF CARBON NANOTUBES
Although carbon nanotubes were discovered soon after fullerenes, not many reports exist in the literature for the solution electrochemistry of these nanoparticles.
8.3
ELECTROCHEMISTRY OF CARBON NANOTUBES
221
Figure 8.9 Electrochemistry of functionalized nanotubes, 0.01 M tetrabutylammonium hexafluorophosphate, THF solution 1.67 mg/mL. V ¼ 0.5 V/s, T ¼ 25 C, and working electrode is Pt disk (r ¼ 62.5 mm); potentials measured versus silver quasi-reference electrode (approximately 0.05 V versus SCE). Reproduced with permission from Ref. 120. Copyright 2004 American Chemical Society.
This is mainly due to their laborious purification procedures and their required chemical modification for solubilization. Only recently, Prato et al. reported the electrochemistry of carbon nanotubes functionalized using the 1,3-dipolar cycloaddition reaction.120 The cyclic voltammogram obtained is shown in Fig. 8.9. The single-walled nanotubes undergo reduction around 0.5 V versus a Ag/Ag þ electrode. Unlike fullerenes, nanotubes show a continuum of diffusion-controlled current, which is attributed to electron transfer to empty electronic states. Since carbon nanotubes are predicted to have a high density of states, the successive filling of these states results in the continuum of current. Guldi et al. have also studied singlewalled carbon nanotubes by solubilizing derivatized nanotubes using electrostatic interactions with porphyrin salts. This leads to nanotubes that are soluble in aqueous media (Fig. 8.10). The cyclic voltammograms of these soluble assemblies show a behavior similar to that previously observed by Prato et al.121, that is, a continuum of cathodic current with the reduction onset voltage of 0.15 V versus SCE. Recently, Paolucci et al. were able to obtain the electrochemistry of reduced unfunctionalized carbon nanotubes that were solubilized by reduction with alkali metals to their respective polyelectrolyte salts.122 These salts were found to be soluble in polar organic solvents as well as in water. Their electrochemistry also shows a continuum of current due to the successive filling of electronic states of the nanotube. Figure 8.11 shows the cyclic voltammogram obtained for nanotubes reduced using sodium metal. Close observation reveals two broad waves around 0.81 and 1.01 V, which have been attributed to the more complex electronic structure of the pristine materials, as compared to derivatized nanotubes. This observation is in agreement with recent calculations indicating
222
ELECTROCHEMISTRY OF CARBON NANOPARTICLES
Figure 8.10 Soluble supramolecular complexes of carbon nanotubes, 0.01 M Na2SO4, aqueous solution. Scan rate ¼ 0.1 V/s; T ¼ 25 C; working electrode is Pt disc (r ¼ 0.05 cm). Potentials measured versus silver quasi-reference electrode (approximately 0.05 V versus SCE). Ref. 121, Reproduced by permission of the Royal Society of Chemistry.
Figure 8.11 Electrochemistry of nanotubes solubilized by direct sodium reduction. Background of the supporting electrolyte solution is shown with dashed line. The star indicates the irreversible anodic peak due to the oxidative stripping of the reduced alkali metal film. 2 mM tetrabutylammonium hydroxide/DMSO; working electrode Pt disk (r ¼ 25 mm); data recorded at 298K; scan rate 1 V/s. Potentials are referenced to SCE. Reproduced with permission from Ref. 122. Copyright 2008 American Chemical Society.
that functionalization significantly affects the low-lying electronic states of the nanotubes.
8.4
CONCLUSION
In conclusion, carbon-based nanomaterials generally exhibit a good electron-accepting ability. Empty fullerenes are capable of undergoing multiple, distinct one electron
REFERENCES
223
reductions. The reduction potentials shift anodically, as the size of the cage increases, thus indicating that the HOMO–LUMO gap for these fullerenes decreases with increasing size. However, the limit of kinetic stability of large fullerenes, along with the dramatic increase in the number of isomers, prevents the study of fullerenes larger than C92. The incorporation of moieties inside fullerene cages allows both the stabilization of larger carbon cages (such as C88, C96, C104, etc.) and the selection of only a few isomers of them. The electrochemistry of these fascinating endohedral fullerenes is influenced by the nature of the encapsulated species, and the size and symmetry of the cage. The larger carbon nanoparticles do not exhibit distinctive redox chemistry. In the case of nanotubes, the electrons successively fill the large number of shells, which results in a continuum of current in reduction. What will the electrochemistry of the appealing but still hypothetical giant fullerenes (C240, C540) be like? Would they still show discrete redox events, such as empty fullerenes and larger endohedral fullerenes, or would they exhibit a continuum of cathodic current, such as carbon nanotubes. This question remains open today.
ACKNOWLEDGMENT The authors wish to acknowledge Prof Dr. Lourdes Echegoyen for the ideas shared and the inputs provided by her.
REFERENCES 1. H. W. Kroto, J. R. Heath, S. C. Obrien, R. F. Curl, R. E. Smalley, Nature 1985, 318, 162–163. 2. L. Echegoyen, L. E. Echegoyen, Acc. Chem. Res. 1998, 31, 593–601. 3. L. Echegoyen, F. Diederich, L. E. Echegoyen,in K. M. Kadish, R. S. Ruoff (Eds.), Fullerenes, Chemistry, Physics, and Technology, Wiley–Interscience, New York, 2000, pp. 1–51. 4. Q. Xie, E. Perez-Cordero, L. Echegoyen, J. Am. Chem. Soc. 1992, 114, 3978–3980. 5. Q. Xie, F. Arias, L. Echegoyen, J. Am. Chem. Soc. 1993, 115, 9818–9819. 6. A. Hirsch, Fullerenes and Related Structures, Springer, Berlin/Heidelberg, 1999, pp. 1–65. 7. Y. Yang, F. Arias, L. Echegoyen, L. P.F. Chibante, S. Flanagan, A. Robertson, L. J. Wilson, J. Am. Chem. Soc. 1995, 117, 7801–7804. 8. W. Wang, J. Ding, S. Yang, X.–Y. Li, Proc - Electrochem. Soc. 1997, 97-14, 186–198. 9. M. D. Diener, J. M. Alford, Nature 1998, 393, 668–671. 10. H. Shinohara, Rep. Prog. Phys. 2000, 63, 843–892. 11. T. Akasaka, S. Nagase,(Eds.), Endofullerenes: A New Family of Carbon Clusters, Kluwer Academic Publishers, 2002. 12. S. Guha, K. Nakamoto, Coord. Chem. Rev. 2005, 249, 1111–1132. 13. L. Dunsch, S. Yang, Small 2007, 3, 1298–1320.
224
ELECTROCHEMISTRY OF CARBON NANOPARTICLES
14. L. Dunsch, S. Yang, Phys. Chem. Chem. Phys. 2007, 9, 3067–3081. 15. M. Saunders, H. A. Jimenez-Vazquez, R. J. Cross, R. J. Poreda, Science 1993, 259, 1428–1430. 16. H. Mauser, N. J.R. Van Eikema Hommes, T. Clark, A. Hirsch, B. Pietzak, A. Weidinger, L. Dunsch, Angew. Chem. Int. Ed. Eng. 1998, 36, 2835–2838. 17. J. A. Larsson, J. C. Greer, W. Harneit, A. Weidinger, J. Chem. Phys. 2002, 116, 7849– 7854. 18. J. R. Heath, S. C. O’Brien, Q. Zhang, Y. Liu, R. F. Curl, F. K. Tittel, R. E. Smalley, J. Am. Chem. Soc. 1985, 107, 7779–7780. 19. S. Stevenson, G. Rice, T. Glass, K. Harlch, F. Cromer, M. R. Jordan, J. Craft, E. Hadju, R. Bible, M. M. Olmstead, K. Maltra, A. J. Fisher, A. L. Balch, H. C. Dorn, Nature 1999, 401, 55–57. 20. M. Murata, Y. Murata, K. Komatsu, J. Am. Chem. Soc. 2006, 128, 8024–8033. 21. J. C. Hummelen, M. Prato, F. Wudl, J. Am. Chem. Soc. 1995, 117, 7003–7004. 22. Y. Rubin, Chem. Eur. J. 1997, 3, 1009–1016. 23. Y. Chai, T. Guo, C. Jin, R. E. Haufler, L. P.F. Chibante, J. Fure, L. Wang, J. M. Alford, R. E. Smalley, J. Phys. Chem. 1991, 95, 7564–7568. 24. T. Suzuki, Y. Maruyama, T. Kato, K. Kikuchi, Y. Achiba, J. Am. Chem. Soc. 1993, 115, 11006–11007. 25. T. Suzuki, Y. Maruyama, T. Kato, K. Kikuchi, Y. Achiba, K. Yamamoto, H. Funasaka, T. Takahashi, Proc - Electrochem. Soc. 1994, 94-24, 1077–1086. 26. T. Suzuki, Y. Maruyama, T. Kato, Synth. Met. 1995, 70, 1443–1446. 27. T. Suzuki, K. Kikuchi, Y. Nakao, S. Suzuki, Y. Achiba, K. Yamamoto, H. Funasaka, T. Takahashi, Proc - Electrochem. Soc. 1995, 95-10, 259–266. 28. T. Suzuki, K. Kikuchi, F. Oguri, Y. Nakao, S. Suzuki, Y. Achiba, K. Yamamoto, H. Funasaka, T. Takahashi, Tetrahedron 1996, 52, 4973–4982. 29. K. Kikuchi, Y. Nakao, S. Suzuki, Y. Achiba, T. Suzuki, Y. Maruyama, J. Am. Chem. Soc. 1994, 116, 9367–9368. 30. T. Akasaka, S. Okubo, T. Wakahara, K. Yamamoto, T. Kato, T. Suzuki, S. Nagase, K. Kobayashi, Proc - Electrochem. Soc. 1999, 99-12, 771–778. 31. T. Akasaka, S. Okubo, M. Kondo, Y. Maeda, T. Wakahara, T. Kato, T. Suzuki, K. Yamamoto, K. Kobayashi, S. Nagase, Chem. Phys. Lett. 2000, 319, 153–156. 32. K. Yamamoto, H. Funasaka, T. Takahashi, T. Akasaka, J. Phys. Chem. 1994, 98, 2008–2011. 33. T. Okazaki, H. Shinohara, Y. Lian, Z. Gu, K. Suenaga, Proc - Electrochem. Soc. 2000, 2000–2012, 378–387. 34. B. Sun, H. Luo, Z. Shi, Z. Gu, Electrochem. Commun. 2002, 4, 47–49. 35. B. Sun, M. Li, H. Luo, Z. Shi, Z. Gu, Electrochim. Acta 2002, 47, 3545–3549. 36. M. Li, J. Wang, B. Sun, N. Li, Z. J. Gu, J. Electrochem. Soc. 2004, 151, E271–E275. 37. U. Kirbach, L. Dunsch, Angew. Chem. Int. Ed. Engl. 1996, 35, 2380–2383. 38. L. Dunsch, U. Kirbach, P. Kuran, Proc - Electrochem. Soc. 1997, 97-42, 761–771. 39. W. Wang, J. Ding, S. Yang, X.-Y. Li, Proc - Electrochem. Soc. 1997, 97-14, 417–428. 40. J. Xu, M. Li, Z. Shi, Z. Gu, Chem. Eur. J. 2006, 12, 562–567. 41. Y. Zhang, J. Xu, C. Hao, Z. Shi, Z. Gu, Carbon 2006, 44, 475–479.
REFERENCES
225
42. P. B. Burbank, J. R. Gibson, H. C. Dorn, M. R. Anderson, J. Electroanal. Chem. 1996, 417, 1–4. 43. N. Nakashima, M. Sakai, H. Murakami, T. Sagara, T. Wakahara, T. Akasaka, J. Phys. Chem. B 2002, 106, 3523–3525. 44. L. Fan, S. Yang, S. Yang, Chem. Eur. J. 2003, 9, 5610–5617. 45. L.-Z. Fan, Y.-Q. Wu, S.-F. Yang, S.-H. Yang,S.-H. Huaxue Xuebao 2004, 62, 2213–2217. 46. L. Fan, S. Yang, S. Yang, J. Electroanal. Chem. 2005, 574, 273–283. 47. L. Fan, S. Yang, S. Yang, Thin Solid Films 2005, 483, 95–101. 48. Y. Wu, L. Fan, S. Yang, J. Phys. Chem. B 2005, 109, 17831–17836. 49. K. Kikuchi, S. Suzuki, Y. Nakao, N. Nakahara, T. Wakabayashi, H. Shiromaru, K. Saito, I. Ikemoto, Y. Achiba, Chem. Phys. Lett. 1993, 216, 67–71. 50. T. Suzuki, Y. Maruyama, T. Kato, K. Kikuchi, Y. Nakao, Y. Achiba, K. Kobayashi, S. Nagase, Angew. Chem. Int. Ed. Engl. 1995, 34, 1094–1096. 51. K. Yamamoto, H. Funasaka, T. Takahashi, T. Akasaka, J. Phys. Chem. 1994, 98, 2008–2011. 52. K. Yamamoto, H. Funasaka, T. Takahasi, T. Akasaka, T. Suzuki, Y. Maruyama, J. Phys. Chem. 1994, 98, 12831–12833. 53. J. P. Selegue, J. P. Shaw, T. F. Guarr, M. S. Meier, Proc – Electrochem. Soc. 1994, 94-24, 1274–1291. 54. P. L. Boulas, M. T. Jones, R. S. Ruoff, D. C. Lorents, R. Malhotra, D. S. Tse, K. M. Kadish, J. Phys. Chem. 1996, 100, 7573–7579. 55. P. Kuran, M. Krause, A. Bartl, L. Dunsch, Chem. Phys. Lett. 1998, 292, 580–586. 56. L. Dunsch, A. Bartl, P. Georgi, P. Kuran, Synth. Met. 2001, 121, 1113–1114. 57. M. R. Anderson, H. C. Dorn, S. A. Stevenson, Carbon 2000, 38, 1663–1670. 58. K. R. Hoffman, K. DeLapp, H. Andrews, P. Sprinkle, M. Nickels, B. Norris, J. Lumin. 1996, 66–67, 244–248. 59. Y. Iiduka, O. Ikenaga, A. Sakuraba, T. Wakahara, T. Tsuchiya, Y. Maeda, T. Nakahodo, T. Akasaka, M. Kako, N. Mizorogi, S. Nagase, J. Am. Chem. Soc. 2005, 127, 9956–9957. 60. X. Lu, H. Nikawa, T. Nakahodo, T. Tsuchiya, M. O. Ishitsuka, Y. Maeda, T. Akasaka, M. Toki, H. Sawa, Z. Slanina, N. Mizorogi, S. Nagase, J. Am. Chem. Soc. 2008, 130, 9129–9136. 61. H. Kato, A. Taninaka, T. Sugai, H. Shinohara, J. Am. Chem. Soc. 2003, 125, 7782–7783. 62. Y. Iiduka, T. Wakahara, T. Nakahodo, T. Tsuchiya, A. Sakuraba, Y. Maeda, T. Akasaka, K. Yoza, E. Horn, T. Kato, M. T.H. Liu, N. Mizorogi, K. Kobayashi, S. Nagase, J. Am. Chem. Soc. 2005, 127, 12500–12501. 63. Y. Iiduka, T. Wakahara, K. Nakajima, T. Tsuchiya, T. Nakahodo, Y. Maeda, T. Akasaka, N. Mizorogi, S. Nagase, Chem. Commun. 2006, 2057–2059. 64. Y. Iiduka, T. Wakahara, K. Nakajima, T. Nakahodo, T. Tsuchiya, Y. Maeda, T. Akasaka, K. Yoza, M. T.H. Liu, N. Mizorogi, S. Nagase, Angew. Chem., Int. Ed. 2007, 46, 5562–5564. 65. T. Wakahara, A. Sakuraba, Y. Iiduka, M. Okamura, T. Tsuchiya, Y. Maeda, T. Akasaka, S. Okubo, T. Kato, K. Kobayashi, S. Nagase, K. M. Kadish, Chem. Phys. Lett. 2004, 398, 553–556. 66. M. R. Anderson, H. C. Dorn, S. Stevenson, P. M. Burbank, J. R. Gibson, J. Am. Chem. Soc. 1997, 119, 437–438.
226
ELECTROCHEMISTRY OF CARBON NANOPARTICLES
67. T. Inoue, T. Tomiyama, T. Sugai, T. Okazaki, T. Suematsu, N. Fujii, H. Utsumi, K. Nojima, H. Shinohara, J. Phys. Chem. B 2004, 108, 7573–7579. 68. K. Tan, X. J. Lu, J. Phys. Chem. A 2006, 110, 1171–1176. 69. H. C. Dorn, S. A. Stevenson,(Virginia Tech Intellectual Properties, Inc., USA), US 99458289/6303760 2001. 70. L. Dunsch, P. Georgi, F. Ziegs, H. Zoeller,(Leibniz-Institut fuer Festkoerper- und Werkstoffforschung e.V., Germany), DE 2003-10301722, 2004. 71. M. N. Chaur, R. Valencia, A. Rodriguez-Fortea, J. M. Poblet, L. Echegoyen, Angew. Chem., Int. Ed. 2009, 48, 1425–1428. 72. S.-L. Lee, M.-L. Sun, Z. Slanina, Int. J. Quantum Chem. 1996, 60, 355–364. 73. H. C. Dorn, S. Stevenson, J. Craft, F. Cromer, J. Duchamp, G. Rice, T. Glass, K. Harich, P. W. Fowler, T. Heine, E. Hajdu, R. Bible, M. M. Olmstead, K. Maitra, A. J. Fisher, A. L. Balch, AIP Conference Proceedings 2000, 544, 135–141. 74. L. Dunsch, M. Krause, J. Noack, P. Georgi, J. Phys. Chem. Solids 2004, 65, 309–315. 75. M. N. Chaur, F. Melin, J. Ashby, B. Elliott, A. Kumbhar, A. M. Rao, L. Echegoyen, Chem. Eur. J. 2008, 14, 8213–8219. 76. M. Wolf, K.-H. Mueller, D. Eckert, Y. Skourski, P. Georgi, R. Marczak, M. Krause, L. Dunsch, J. Magn. Magn. Mater. 2005, 290–291, 290–293. 77. M. Wolf, K.-H. Mueller, Y. Skourski, D. Eckert, P. Georgi, M. Krause, L. Dunsch, Angew. Chem., Int. Ed. 2005, 44, 3306–3309. 78. S. Stevenson, H. M. Lee, M. M. Olmstead, C. Kozikowski, P. Stevenson, A. L. Balch, Chem. Eur. J. 2002, 8, 4528–4535. 79. M. Krause, J. Wong, L. Dunsch, Chem. Eur. J. 2005, 11, 706–711. 80. T. Zuo, M. M. Olmstead, C. M. Beavers, A. L. Balch, G. Wang, G. T. Yee, C. Shu, L. Xu, B. Elliott, L. Echegoyen, J. C. Duchamp, H. C. Dorn, Inorg. Chem. 2008, 47, 5234–5244. 81. S. Stevenson, J. P. Phillips, J. E. Reid, M. M. Olmstead, S. P. Rath, A. L. Balch, Chem. Commun. 2004, 2814–2815. 82. M. Krause, L. Dunsch, Angew. Chem., Int. Ed. 2005, 44, 1557–1560. 83. J. Lu, R. F. Sabirianov, N. Mei Wai, Y. Gao, C.-G. Duan, X. Zeng, J. Phys. Chem. B 2006, 110, 23637–23640. 84. M. N. Chaur, F. Melin, B. Elliott, A. J. Athans, K. Walker, B. C. Holloway, L. Echegoyen, J. Am. Chem. Soc. 2007, 129, 14826–14829. 85. S. Yang, L. Dunsch, J. Phys. Chem. B 2005, 109, 12320–12328. 86. S. Yang, S. I. Troyanov, A. A. Popov, M. Krause, L. Dunsch, J. Am. Chem. Soc. 2006, 128, 16733–16739. 87. H. Shiozawa, H. Rauf, T. Pichler, D. Grimm, X. Liu, M. Knupfer, M. Kalbac, S. Yang, L. Dunsch, B. Buchner, D. Batchelor, Phys. Rev. B: Condens. Matter Mater. Phys. 2005, 72, 195409/1-195409/5. 88. L. Feng, J.-X. Xu, Z.-J. Shi, X.-R. He, Z.-N. Gu, Gaodeng Xuexiao Huaxue Xuebao 2002, 23, 996–998. 89. T. Zuo, C. M. Beavers, J. C. Duchamp, A. Campbell, H. C. Dorn, M. M. Olmstead, A. L. Balch, J. Am. Chem. Soc. 2007, 129, 2035–2043. 90. F. Melin, M. N. Chaur, S. Engmann, B. Elliott, A. Kumbhar, A. J. Athans, L. Echegoyen, Angew. Chem., Int. Ed. 2007, 46, 9032–9035.
REFERENCES
227
91. M. N. Chaur, F. Melin, B. Elliott, A. Kumbhar, A. J. Athans, L. Echegoyen, Chem. Eur. J. 2008, 14, 4594–4599. 92. S. Yang, M. Kalbac, A. Popov, L. Dunsch, ChemPhysChem 2006, 7, 1990–1995. 93. S. Yang, A. A. Popov, L. Dunsch, J. Phys. Chem. B 2007, 111, 13659–13663. 94. S. Yang, A. Popov, M. Kalbac, L. Dunsch, Chem. Eur. J. 2008, 14, 2084–2092. 95. M. M. Olmstead, A. de Bettencourt-Dias, J. C. Duchamp, S. Stevenson, H. C. Dorn, A. L. Balch, J. Am. Chem. Soc. 2000, 122, 12220–12226. 96. S. Stevenson, P. W. Fowler, T. Heine, J. C. Duchamp, G. Rice, T. Glass, K. Harich, E. Hajdu, R. Bible, H. C. Dorn, Nature 2000, 408, 427–428. 97. N. Chen, E.-Y. Zhang, C.-R. Wang, J. Phys. Chem. B 2006, 110, 13322–13325. 98. S. Yang, M. Kalbac, A. Popov, L. Dunsch, Chem. Eur. J. 2006, 12, 7856–7863. 99. S. Yang, A. A. Popov, L. Dunsch, Angew. Chem., Int. Ed. 2007, 46, 1256–1259. 100. M. M. Olmstead, A. De Bettencourt-Dias, J. C. Duchamp, S. Stevenson, D. Marciu, H. C. Dorn, A. L. Balch, Angew. Chem., Int. Ed. 2001, 40, 1223–1225. 101. M. Krause, A. Popov, L. Dunsch, ChemPhysChem 2006, 7, 1734–1740. 102. D. R. Lide (Ed.), CRC Handbook of Chemistry and Physics, 81st edition, CRC Press, New York, 2000. 103. M. Krause, L. Dunsch, ChemPhysChem 2004, 5, 1445–1449. 104. B. Elliott, L. Yu, L. Echegoyen, J. Am. Chem. Soc. 2005, 127, 10885–10888. 105. C. M. Cardona, B. Elliott, L. Echegoyen, J. Am. Chem. Soc. 2006, 128, 6480–6485. 106. M. E. Plonska-Brzezinska, A. J. Athans, J. P. Phillips, S. Stevenson, L. Echegoyen, J. Electroanal. Chem. 2008, 614, 171–174. 107. L. Zhang, N. Chen, L. Fan, C. Wang, S. Yang, J. Electroanal. Chem. 2007, 608, 15–21. 108. T. Cai, L. Xu, M. R. Anderson, Z. Ge, T. Zuo, X. Wang, M. M. Olmstead, A. L. Balch, H. W. Gibson, H. C. Dorn, J. Am. Chem. Soc. 2006, 128, 8581–8589. 109. J. M. Campanera, C. Bo, M. M. Olmstead, A. L. Balch, J. M. Poblet, J. Phys. Chem. A 2002, 106, 12356–12364. 110. P. Jakes, K.-P. Dinse, J. Am. Chem. Soc. 2001, 123, 8854–8855. 111. L. Alvarez, T. Pichler, P. Georgi, T. Schwieger, H. Peisert, L. Dunsch, Z. Hu, M. Knupfer, J. Fink, P. Bressler, M. Mast, M. S. Golden, Phys. Rev. B: Condens. Matter Mater. Phys. 2002, 66, 035107/1-035107/7. 112. M. Krause, X. Liu, J. Wong, T. Pichler, M. Knupfer, L. Dunsch, J. Phys. Chem. A 2005, 109, 7088–7093. 113. S. Yang, M. Zalibera, P. Rapta, L. Dunsch, Chem. Eur. J. 2006, 12, 7848–7855. 114. S. Yang, P. Rapta, L. Dunsch, Chem. Commun. 2007, 189–191. 115. T. Cai, L. Xu, C. Shu, H. A. Champion, J. E. Reid, C. Anklin, M. R. Anderson, H. W. Gibson, H. C. Dorn, J. Am. Chem. Soc. 2008, 130, 2136–2137. 116. P. Rapta, A. A. Popov, S. Yang, L. Dunsch, J. Phys. Chem. A 2008, 112, 5858–5865. 117. M. N. Chaur, A. J. Athans, L. Echegoyen, Tetrahedron 2008, 64, 11387–11393. 118. T. Zuo, K. Walker, M. M. Olmstead, F. Melin, B. C. Holloway, L. Echegoyen, H. C. Dorn, M. N. Chaur, C. J. Chancellor, C. M. Beavers, A. L. Balch, A. J. Athans, Chem. Commun. 2008, 1067–1069.
228
ELECTROCHEMISTRY OF CARBON NANOPARTICLES
119. C. M. Beavers, T. Zuo, J. C. Duchamp, K. Harich, H. C. Dorn, M. M. Olmstead, A. L. Balch, J. Am. Chem. Soc. 2006, 128, 11352–11353. 120. M. Melle-Franco, M. Marcaccio, D. Paolucci, F. Paolucci, V. Georgakilas, D. M. Guldi, M. Prato, F. Zerbetto, J. Am. Chem. Soc. 2004, 126, 1646–1647. 121. D. M. Guldi, G. N.A. Rahman, J. Ramey, M. Marcaccio, D. Paolucci, F. Paolucci, S. H. Qin, W. T. Ford, D. Balbinot, N. Jux, N. Tagmatarchis, M. Prato, Chem. Commun. 2004, 2034–2035. 122. D. Paolucci, M. M. Franco, M. Iurlo, M. Marcaccio, M. Prato, F. Zerbetto, A. Penicaud, F. Paolucci, J. Am. Chem. Soc. 2008, 130, 7393–7399.
CHAPTER 9
Molecular Devices Based on Fullerenes and Carbon Nanotubes MATTEO IURLO, DEMIS PAOLUCCI, MASSIMO MARCACCIO, and FRANCESCO PAOLUCCI Dipartimento di Chimica “G. Ciamician”, Alma Mater Studiorum, Universit a di Bologna, Bologna, Italy
9.1
INTRODUCTION
A molecular device, that is, a device at the molecular level, is an assembly of a finite number of molecular components organized on the spatial, temporal, and energetic scales so that it may perform a useful and specific function.1 Soon after the discovery of this fascinating molecular structure and techniques for its large-scale preparation,2 fullerenes were demonstrated to be promising electronic materials and ideal candidates for the development of molecular and supramolecular devices.3 C60 is a single molecular unit, that is, not made up of independent subcomponents, with a unique three-dimensional delocalized p-electron system with a high Ih symmetry, interesting photophysical properties,4 and that may exist in solution in up to 10 different redox states, energetically spanning a potential range larger than 5 V.5 It displays a significant electron-accepting character and such a property adds to the many other unique properties of such a fascinating molecule, such as a high absorption throughout the visible spectral region and the ability of rapid photoinduced charge separation, and put the fullerene and its derivatives among the most important organic materials for photovoltaics applications and, particularly, in plastic solar cells.6 Over the past 15 years, we have assisted in a huge development of the covalent chemical functionalization of the C60 carbon sphere aimed at generating many new fullerene-based materials,6,7 in which the outstanding properties of the fullerenes would combine synergistically with those of other molecular materials, polymers, dendrons, liquid crystals, and more, in general, with photo-, electro-, or biologically active units (Scheme 9.1). Electrochemistry of Functional Supramolecular Systems. Edited by Paola Ceroni, Alberto Credi, and Margherita Venturi Copyright 2010 John Wiley & Sons, Inc.
229
230
MOLECULAR DEVICES BASED ON FULLERENES AND CARBON NANOTUBES
Scheme 9.1
The organic modification of the fullerene, carried out according to many and diverse synthetic strategies, also accomplished the important task of increasing fullerene solubility and processability, since its derivatives are much more soluble than the pristine species in polar solvents or in aqueous media. Though C60 has not (so far) found any major applications, the influence of C60 is now pervasive in chemistry and beyond.8 C60 is a kind of ideal nanoscale building block that can be picked up and manipulated with nanotechnological tools.9 Importantly, its curved, hollow structure has made us familiar with another view of carbon materials, different and complementary to that of flat sheets of carbon atoms in graphite. Ultimately, all the interest generated around such carbon allotropes has driven the research in the field and introduced the perhaps most notable representatives of the present nanoworld, carbon nanotubes (NTs). Carbon nanotubes show unique features that are proposed for the development of nanometer-scale materials with outstanding potential technological applications,10 in a process that is accompanied by a constantly increasing understanding of many of their properties. Issues that may greatly hamper their widespread technological use comprise the difficult processability of single-walled carbon nanotubes (SWNTs) in liquid phases and the currently unavoidable structural heterogeneity of as-synthesized SWNTs, and are now being overcome.11 Most common processing procedures involve dispersing the nanotubes by the help of a dispersant phase, a surfactant,12 a biomolecule,13 or a polymer,14 by dilution,15 or by chemical covalent or noncovalent functionalization,16 reduction,17 or protonation by superacids.18 Feasible nanotube applications for the near and long term comprise electronic devices and interconnects, field emission devices, electrochemical devices, sensors and biosensors, electromechanical actuators, polymer composites, and drug delivery systems.19 Because of the great expectations that these carbon nanostructures—fullerenes, carbon nanotubes, graphene, and related species20—have for potential applications in
9.2
FULLERENES
231
materials science, molecular electronics, biological and biomedical applications, an enormous amount of work has been carried out all over the world over the past 15– 20 years. A comprehensive coverage of all the aspects associated with the theme of this review is therefore necessarily beyond the scope of the authors. Rather, a few representative examples try to show the wide variety of forms and the exciting materials aspects that have been evolving in the recent past. The selected examples are in part taken from the authors’ own contribution to this research field or reflect otherwise their personal preferences. The reader is directed for further information to several excellent and comprehensive reviews and monographies21 covering many other aspects of this fascinating area. 9.2
FULLERENES
9.2.1
Donor–Acceptor Photoactive Dyads
Much of the chemistry and the applications of C60 are based on its electron-accepting capacity. Six electrons are readily accommodated in the threefold degenerate lowest unoccupied molecular orbital (LUMO), while six more can be introduced into the equally threefold degenerate LUMO þ 1.22 Li12C60 is stable and has long been described.23 The dynamics of reduction of C60 and its derivatives has also been thoroughly characterized in solution by electrochemical means. The facile reduction of C6024 contrasts with its difficult oxidation. In 1993, Echegoyen et al., employing scrupulously dried tetrachloroethane (TCE), tetrabutylammonium hexafluorophosphate (TBAPF6) as supporting electrolyte, and low temperatures, reported the first observation of electrochemical reversible one-electron oxidation of C60 at E1/2 ¼ 1.26 V (versus Fc þ /Fc).5a Furthermore, in condensed media, fullerenium radical cations—the first all-carbon carbocations—react immediately with any nucleophile present in solution leading to decomposition.5b Higher fullerenes are easier to oxidize than C60, and the second electrochemical oxidation (although irreversible) of C70, C76, and C78 in TCE/TBAPF6 was reported.25 By contrast, the further oxidation of C60 þ was only recently obtained by the adoption of suitable experimental conditions that comprise ultradry solvents and electrolytes with very high oxidation resistance and low nucleophilicity, thus obtaining the cyclic voltammetric reversible generation of C602 þ and C603 þ (Table 9.1).5c AsF6 was chosen as supporting counteranion, which is known for its high oxidation resistance and low nucleophilicity.26 TABLE 9.1
E1/2 (Volts Versus Fc þ /Fc) of Oxidation/Reduction of C60
Solvent
I
II
III
I
II
III
IV
V
VI
1.27
1.74
2.14
1.06 0.90
1.46 1.47
1.89 2.05
2.54
3.07
3.53
Oxidations
CH2Cl2 THF b a b
a
Reductions
Supporting electrolyte: tetrabutylammonium hexafluoroarsenate (0.05 M), T ¼ 60 C. W: Pt. Supporting electrolyte: tetrabutylammonium hexafluorophosphate (0.05 M), T ¼ 60 C. W: Pt.
232
MOLECULAR DEVICES BASED ON FULLERENES AND CARBON NANOTUBES
Interestingly, such a study very recently led to the serendipitous discovery that, when operating in the presence of very low levels of nucleophiles, C60 may be oxidatively electropolymerized, thus forming stable conducting polymeric films able to transport both holes and electrons, a feature that is not possessed by many organic systems.27 The very attractive electron acceptor properties of C60 combine synergistically with its unique three-dimensional structure in giving a highly performing partner in photoactive donor–acceptor dyads. Porphyrins and metalloporphyrins have been preferentially used as donor/chromophores in such systems, for the unique combination of favorable photophysical and redox properties of these compounds. Furthermore, these compounds, particularly appealing for their similarity with natural chromophores, are stable and allow high synthetic availability to many structural variations and therefore great flexibility in the design of D-S-A systems. Since 1994, a great number of different porphyrin–C60 dyads, triads, and more complex systems, also including supramolecular assemblies, have been investigated (see Scheme 9.2 for some examples). Other donors very often used in combination with fullerenes comprise ferrocene, phthalocyanine, transition metal complexes, aniline derivatives, tetrathiafulvalene and oligoacenes, carotenoids, oligoarylene, and oligothiophene and many examples are collected in recent reviews and books dedicated to this subject.3a,7e,28 Investigation of photoinduced intramolecular ET processes in such systems has in general shown that, compared to analogous porphyrin/quinone systems, i. charge separation occurs with higher efficiency and ii. charge-separated states have longer lifetimes (Scheme 9.4). Such a behavior is the result of the combination of two
Scheme 9.2
9.2
FULLERENES
233
favorable characteristics of fullerenes compared to quinones and other normal donor units, namely, i. the curvature of fullerene surface that allows a better electronic coupling between the acceptor unit and the hydrocarbon bridge and, ultimately with the D unit, ii. the smaller reorganization energy associated to fullerenes, with respect to most acceptors, as a consequence of their large size and rigid framework in the ground, excited, or reduced states. 29 Most photoinduced charge separation processes are only slightly exergonic; that is, the process generally occurs in the normal Marcus region; smaller reorganization energy therefore implies, for a given driving force, a faster process. Vice versa, charge recombination is often a highly exergonic process; this places the process in the Marcus inverted region and a smaller reorganization energy makes, for a given driving force, the kinetic constant of the charge recombination process smaller (Scheme 9.3).30 The strong donor tetrathiafulvalene (TTF) has emerged as an interesting class of compounds allowing the stabilization of the charge-separated state.31 This peculiarity is due to the gain of aromaticity as a result of the formation of the thermodynamically very stable heteroaromatic 1,3-dithiolium cation(s) upon oxidation(s) of the TTF molecule, which in its ground-state configuration is nonaromatic. TTF derivatives have been covalently linked to C60 through a flexible spacer using several methodologies. The search for new fullerene-based materials, whose properties could be of interest in nanotechnology (e.g., molecular switches, solar cells), has prompted the investigation of fullerene-containing thermotropic photoactive liquid crystals.32 Liquidcrystalline fullerenes were designed according to different synthetic routes (Scheme 9.4). The two classes of compounds shown in Scheme 9.4 differ at a fundamental level from the electrochemical point of view: whereas methanofullerenes undergo retro-Bingel reaction upon chemical and electrochemical reduction,33 fulleropyrrolidines are stable species. Dendritic liquid-crystalline fulleropyrrolidines would represent an interesting family of electroactive macromolecules, as they would combine the electrochemical behavior of C60 with the rich mesomorphism found for dendrimers. Ru tris-bipyridine complexes have been widely used in photoactive dyads to study photoinduced electron and energy transfer processes for their unique photophysical and redox properties.34 Furthermore, the chemistry of bipyridine is well developed and allows a wide range of possible functionalized derivatives.35 By combining the standard potential for the Ru(II)-based oxidation process (E þ 1.3 V) and its 3 MLCT excited-state energy (E00 2.0 eV), the standard potential of excited state E [Ru(III)/*Ru(II)] is approximately 0.7 V. The latter may therefore ignite the photoinduced electron transfer to fullerene derivatives (E 0.4 V). The nature and strength of the electronic interaction between a fulleropyrrolidine and the [Ru(bpy)3]2 þ moiety, and the influence of the bridging spacer on such an interaction, were investigated by electrochemistry, UV–vis–NIR absorption spectroscopy, and steady-state and time-resolved emission spectroscopy in a series of dyads where the units were covalently linked to each other by spacers showing different flexibility, that is, either a rigid androstane bridge that suppresses conformational freedom in the dyad
234
MOLECULAR DEVICES BASED ON FULLERENES AND CARBON NANOTUBES
Scheme 9.3
shown in Scheme 9.5,36 or flexible spacers37 that allow variable dimensional freedom upon temperature and/or solvent changes. A reversible one-electron transfer at þ 1.34 V was observed, assigned to the metalcentered Ru(II)/Ru(III) oxidation. The complex cathodic pattern of this species, comprising 10 successive reversible one-electron transfer, is the sum of that of its two subunits, the C60 core and the Ru(II) tris-bipyridyl complex, thus excluding any
235
Scheme 9.4
236
MOLECULAR DEVICES BASED ON FULLERENES AND CARBON NANOTUBES
Scheme 9.5
significant electronic interaction between the two moieties in the ground state. The first reduction process (at 0.47 V) involves the fullerene moiety. On the other hand, the driving force for the photoinduced ET process, from Ru 3 MLCT to fulleropyrrolidine, confirmed the hypothesis that electron transfer from the ruthenium complex to the fulleropyrrolidine may quench the photoexcited [Ru(bpy)3]2 þ MLCT state. Also, expectedly, the excited-state dynamics was strongly affected by the solvent polarity. Charge-separated state is therefore formed from the Ru 3 MLCT excited state, via a through-bond mechanism, with intramolecular rate constants of 0.69 109 s1 and 5.1 109 s1 in CH2Cl2–toluene (1:1, v/v) and CH3CN, respectively. 9.2.2
Organic Solar Cells
One of the most promising uses of C60 involves its potential application, when mixed with p-conjugated polymers, in polymer solar cells. Most often the so-called bulk heterojunction configuration is used, in which the active layer consists of a blend of electron-donating materials, for example, p-type conjugated polymers, and an electron-accepting material (n-type), such as (6,6)-phenyl-C61-butyric acid methyl ester (PCBM, Scheme 9.6).38 Organic solar cells with power conversion efficiencies (PCE) approaching 5% under AM1.5G illumination have recently been reported by using a regioregular poly(3-hexylthiophene) (rr-P3HT)/PCBM blend39 and by introducing morphology
9.2
FULLERENES
237
Scheme 9.6
improving procedures on the composite film 40. Tailoring bicontinuous electron donor (D) and acceptor (A) arrays in solution processable materials is an essential step for the realization of thin-film organic optoelectronics.41 Recently, a photoconductive liquid crystal having bicontinuous arrays of densely packed D and A components was reported, tailored from amphiphilic oligothiophene (OTP)–C60 dyad (Scheme 9.7).42 Together with contrasting results for a nonamphiphilic reference species (Scheme 9.7), a crucial role of the amphiphilic design both in structuring and in photoconductivity was highlighted. 9.2.3
Interlocked Architectures equipped with C60
A molecular-level machine is an assembly of a discrete number of molecular components designed to perform a specific function. Such an assembly may be organized in either a supramolecular or mechanically linked molecular architecture: each molecular component is capable of a single act, while the entire assembly performs a more complex function.1b In a mechanically linked system, the single components are covalently interlocked to each other. Conversely, noncovalent “weak” intercomponent interactions, such as hydrogen bonding, charge transfer interactions, solvophobic interactions, prevent instead dissociation of single components in supramolecular systems. Supramolecular chemistry represents an elegant alternative approach for the construction of functional systems by means of noncovalent bonding interactions. Supramolecular C60 and phthalocyanine dimers have been self-assembled via hydrogen bonding between complementary recognition motifs.43 In a wider context, hydrogen bonding p–p stacking, metal-mediated complexation, and electrostatic motifs have been investigated in order to incorporate fullerenes into well-ordered arrays (Scheme 9.8).9b A pseudorotaxane is the supramolecular product of the self-assembly of two components (i.e., a [2]pseudorotaxane), a molecular ring (macrocycle) and a molecular wire (thread) threaded into it. The occurrence of specific interactions between the two components lowers the free energy of the assembled system with respect to the separate components and drives the self-assembly of the pseudorotaxane (threading process). Threading occurs generally in solution and, upon changes of solvent
238
MOLECULAR DEVICES BASED ON FULLERENES AND CARBON NANOTUBES
Scheme 9.7
polarity or other external stimuli, dethreading of the pseudorotaxane may be induced: dethreading/rethreading of the wire and ring components is reminiscent of the way a piston moves within a cylinder. The supramolecular system in Scheme 9.9 comprises a phthalocyanine appended with dibenzo-24-crown-8 unit and a fullerene moiety carrying a quaternary ammonium cation. The pseudorotaxane self-assembles through hydrogen bonding, and the binding constant, calculated using steady-state fluorescence, is 1.4 104. The zinc phthalocyanine derivative exhibited an excited-state lifetime of approximately 3.1 ns. The fast decay of the excited state of 1 ZnPc upon complexation with the fullerene
9.2
FULLERENES
239
Scheme 9.8
ammonium cation suggested the intramolecular electron transfer from 1 ZnPc as the predominant quenching mechanism.44 Catenanes and rotaxanes represent examples of interlocked architectures. A catenane consists of two interlocked rings (it is called in this case a [2]catenane),
Scheme 9.9
240
MOLECULAR DEVICES BASED ON FULLERENES AND CARBON NANOTUBES
Scheme 9.10
while a rotaxane differs from a pseudorotaxane for the two bulky stoppers located at the end of the wire that prevent dethreading. In both systems, separation of the two components may only be obtained by the cleavage of at least one covalent bond. In such “molecular shuttles,” the translocation of the ring can be achieved by application of an external stimulus that ranges from electron transfer, either photoinduced45 or electrochemically, in solution46 and in self-assembled monolayers onto the electrode surface47, to change of the solvent polarity48 or pH change.49 Initially, fullerenes were introduced in catenanes50, pseudorotaxanes51, and rotaxanes43a,52 (Scheme 9.10) as a synthetic challenge, since fullerenes present a low solubility in many solvents and are hard to process. Once the synthetic methodologies were developed, the research efforts were focused on the preparation of photoactive systems where fullerenes had already shown potential applicability, such as organic photovoltaic materials. In these molecular-scale engineered systems, a fullerene electron acceptor contained in one submolecular fragment is coupled with an electron donor contained in the opposite component. Zinc porphyrins were thus coupled to C60 in many different architectures such as, for instance, in that shown in Scheme 9.11, where the ZnP was appended to
9.2
FULLERENES
241
Scheme 9.11
the macrocycle and C60 moieties were the stoppers on the thread around the central [Cu(phen)2] þ core.53 Intrarotaxane photoinduced charge transfer in such a species leads to a charge-separated state with lifetime of about 29 ms. Fullerene stoppers have also been introduced in rotaxanes as a way to probe the motion of the ring thanks to their well-defined photophysical and electrochemical properties.54 In the rotaxane shown in Scheme 9.12, the hydrogen bonding station (a glycylglycine template) was placed far away from the fullerene by a triethylene glycol spacer.55 As expected, in solvents such as CH2Cl2, CHCl3 and THF, the macrocycle stays preferentially on the peptidic station, far away from the fullerene. Instead, in solvents such as DMSO and DMF that weaken the hydrogen bonds, the interactions between the macrocycle and the fullerene are promoted, inducing a large positional change of the macrocycle. In the excited-state absorption measurements, nearly solvent-independent behavior was observed for the thread, while measurements carried out on rotaxane showed that the fluorescence of the fullerene is quenched 28% and 44%, respectively, by the proximity of the macrocycle with DMF and DMSO. The residual fullerene fluorescence in DMSO is 51% of that in CH2Cl2. These experiments provided not only a way to identify the interactions taking place between the macrocycle and the fullerene but also provide a much simpler way to monitor shuttling than with transient absorption measurements. An important feature of the rotaxane in Scheme 9.12 is that the translocation of the macrocycle is also achieved by the reduction of the fullerene to its trianion, which is both effected and observed by cyclic voltammetry. In DMSO, the proximity of the macrocycle to the fullerene stabilized substantially the electrogenerated trianion (DE1/2 ¼ 46 mV) through p–p interactions. Surprisingly, in THF where the macrocycle is preferentially positioned on the peptide station, a similar behavior was
242
MOLECULAR DEVICES BASED ON FULLERENES AND CARBON NANOTUBES
Scheme 9.12
observed (DE1/2 ¼ 40 mV). Although contradictory, these results can be easily rationalized in terms of electrochemically induced shuttling. A coconformational change takes place with the benefit of the stabilization of the negative charge present on the fullerene. The possibility of fine-tuning electron transfer processes through molecular shuttling was finally shown by introducing ferrocene electron donors on the macrocycle (Scheme 9.13).56 Steady-state and transient absorption photophysical measurements revealed through-space photoinduced electron transfer between the fullerene stopper and the ferrocenes on the macrocycle. In CH2Cl2, the radical ion pair state lifetime of 26.2 ns was measured. This value is consistent with the larger relative separation of the electroactive units that gives longer lifetimes. Addition of HFIP shortens the lifetime to 13.0 ns, a consequence imposed by weakening the hydrogen bonds that decreases the relative spatial separation between the donor and the acceptor, while increasing the shuttling rate.
9.3
CARBON NANOTUBES
243
Scheme 9.13
9.3
CARBON NANOTUBES
Carbon nanotubes represent a new kind of carbon materials that are superior to other kinds of carbon materials such as glassy carbon, graphite, and diamond. For their special structural features and unique electronic properties, NTs have found applications in electrocatalysis, direct electrochemistry of proteins, and electroanalytical devices, such as electrochemical sensors and biosensors. The electroanalytical applications of carbon nanotubes have been thoroughly described in a series of recent reviews.21f Recently, there has been a large interest in using SWNTs as electrodes for electrochemistry stemming from the prospect of using individual SWNTs as carbon nanoelectrodes or ensembles of SWNTs as large surface area carbon electrodes. Finally, for their redox properties, SWNTs represent unique building blocks for the construction of photofunctional nanosystems to be used in efficient light energy conversion devices.28e A single-walled carbon nanotube is often described as the result of seamless wrapping of a graphene sheet into a cylinder along the (n, m) roll-up vector,
244
MOLECULAR DEVICES BASED ON FULLERENES AND CARBON NANOTUBES
Scheme 9.14
v ¼ na1 þ ma2 (Scheme 9.14). The (n, m) indices fully define the SWNT radius and chirality and determine univocally its electronic structure.57 If (n m( ¼ 3q, where q is an integer, the SWNT is metallic, whereas for (n m ( 6¼ 3q, it is semiconducting with a bandgap in the density of states (DOS) whose size is inversely proportional to the diameter. As a consequence of the size-dependent quantization of electronic wave functions around the circumference of the SWNT, the DOS shows typical singularities, the so-called van Hove singularities, consisting of a singular increase in the DOS at energies evH followed by a (e evH)1/2 decrease (Scheme 9.15). About one-third of all the SWNTs are metallic and always have wider energy gaps between the first van Hove spikes than semiconducting ones with similar diameter. The presence of the van Hove singularities dominates the spectral features of these species58 as well as the electrochemical ones.59 Proper solution electrochemical experiments on pristine SWNTs were only possible after the recent discovery of an innovative way to form thermodynamically stable solutions of unmodified and uncut SWNTs.17,60 Upon reduction with alkali metals, SWNTs produce polyelectrolyte salts (Scheme 9.16) that are soluble in polar organic solvents without the use of sonication, surfactants, or functionalization. Polyelectrolyte SWNT salts were obtained by reacting arc-discharge samples (a-NT) or HiPco samples (h-NT) with different alkali metals (Na, K). The typical voltammetric curve for a solution of SWNTs in organic solvents displays a continuum of diffusion-controlled current, with onset, in both the negative and the positive potential region, that depends on the nanotube average diameter and ultimately on the NT preparation technique. In fact, SWNTs prepared according to the arc-discharge method display an anticipated onset with respect to HiPco ones
9.3
CARBON NANOTUBES
245
Scheme 9.15
indicative of a smaller energy gap, consistent with the larger average diameters of the arc SWNTs with respect to the latter.59 Covalent functionalization of the nanotubes also affects the voltammetric features: the richer curve morphology observed in the case of pristine nanotubes reflects the more complex electronic structure of the pristine materials, compared to functionalized ones (Scheme 9.17), in agreement with recent calculations that suggested that functionalization significantly affects the low-lying electronic states of the nanotubes.61 The knowledge of the oxidation and reduction potentials of SWNTs is instrumental to the rational design of electrochemical, photochemical, and electronic devices, such as sensors, photoactive dyads, light-emitting diodes, or other electroluminescent devices, in which the energetics of electrochemical and electric generation of electrons and holes in the SWNTs play an important role. The voltammetric response of individual SWNTs in solution reflects average collective properties rather than single SWNT ones and is therefore not a suitable technique to determine the standard redox potentials of individual semiconducting SWNTs as a function of the tube structure. This was instead achieved via an extensive vis–NIR spectroelectrochemical investigation of true solutions of polyelectrolyte SWNT (Scheme 9.16) solutions that was carried out in the –1500 mV range (versus SCE), where doping is expected to
246
MOLECULAR DEVICES BASED ON FULLERENES AND CARBON NANOTUBES
Scheme 9.16
modify the population of the first and second van Hove singularities of the electronic bands of SWNTs. The electronic transitions affected by charging are therefore the semiconducting S11 (v1s ! c1s ) and S22 (v2s ! c2s ) and possibly the metallic M11 (v1m ! c1m )58. As expected for heavily n-doped SWNTs, the starting solutions do not
Scheme 9.17
9.3
CARBON NANOTUBES
247
show significant absorption except for the plasmon band, typical of graphitic materials.16a,62 Spectral changes were recorded as long as the applied potential was increased above 1.2 V, associated with the progressive undoping of SWNTs: the intensity increased progressively until the maximum was reached for the complete depletion of the conduction bands. Scheme 9.18 displays a summary of the spectroelectrochemical investigation of h-NT’s S11. On both reduction and oxidation sides, the absorption bands have
Scheme 9.18 Top: Plots of optical absorption intensity as a function of wavelength and electrode potential in the S11 region for K[h-NT]. In all plots, raw electrochemical data, that is, uncorrected for ohmic drop, are referenced to SCE. Bottom: Chirality map displaying the average standard potentials associated to each SWNT. HiPco SWNTs are located inside the red line, while arc-discharge SWNT are inside the blue line. Starred values were extrapolated from the linear fitting equations given in the text. (See the color version of this Scheme in Color Plates section.)
248
MOLECULAR DEVICES BASED ON FULLERENES AND CARBON NANOTUBES
different onset potentials, thus suggesting that different redox potentials are associated with different tube structures. The analysis of the spectroelectrochemical results of h-NT and a-NT shown above allowed to obtain the standard redox potential as a function of the tube diameter of a large number of semiconducting SWNTs. The analysis was carried out assuming that the absorbance is due only to neutral SWNTs (and that they do it according to Lambert–Beer law) and that Nernst equation describes the ratios of activity (concentration) of the neutral-to-reduced (or -oxidized) SWNTs. Therefore, the absorbance (A) of the samples is described as a function of electrochemical potential by two equations for the n-doped and p-doped SWNTs63: eðF=RTÞðEEred;i Þ 0
Andop;i ¼
eN;i bc i
1þe
0 Þ ðF=RTÞðEEred;i
Apdop;i ¼ eN;i bc i
1 0 Þ ðF=RTÞðEEox;i
1þe
ð9:1Þ
where i is a single (n, m) nanotube that contributes to a specific transition band, eN is the extinction coefficient of the neutral (n, m) nanotube (for simplicity assumed to be the same for all tubes), c* is its bulk concentration, E is the electrochemical potential, b is the path of the spectroscopic cell, and Ered;i and Eox;i are the standard potentials of its reduction (n-doping) and oxidation (p-doping). The intensity of each band in the spectra was analyzed according to equations (9.1) to give a set of standard potentials for the oxidations and reductions of h-NT and a-NT. Finally, the E s vary linearly with the excitation energy
Ered ¼ 1:02 Eexc þ 0:26 ðR2 ¼ 0:989Þ
Eox ¼ 0:36 Eexc þ 0:22 ðR ¼ 0:704Þ 2
ð9:2Þ ð9:20 Þ
that can be employed in the design of devices mentioned above. The chirality map of Scheme 9.18, where color codes are used to gather the structures that share the same values, shows that, in analogy with optical transition frequencies, a simple dependence of redox properties of SWNTs on diameter is not strictly followed. 9.3.1
Donor–Acceptor Ensembles
Carbon nanotubes possessing electroactive, photoactive, catalytic, and biologically active functional components coupled to their surface with useful potential applications in medicine,19a sensors and biosensors,64 and light energy harvesting devices65 are a topic of great research interest. The presence of extended, delocalized p-electron systems renders most CNTs very useful for managing charge transfer and charge transport. In this respect, CNTs have been mostly considered, in analogy to fullerenes, as a good electron acceptor, with the additional property of quasiballistic electron transport along their tubular axis.66 Advances in the use of CNTs in photovoltaic devices for photocurrent generation have been very recently reviewed.67 Following our recent report on the first intramolecular electron transfer within a photoexcited SWNT and ferrocene moieties covalently linked to it (Scheme 9.19),68 a
9.3
CARBON NANOTUBES
249
Scheme 9.19
large number of photoactive dyads based on SWNTs have been reported.69 In such dyads, the electron donors—ferrocenyl derivatives, porphyrinoid and metalloporphyrinoid systems (Scheme 9.20), and extended-TTF—and in some cases the electron acceptor (e.g., a fullerene moiety) have been electronically coupled to the SWNT surface either by covalent attachment onto the sidewalls of SWNTs or, alternatively, by supramolecular van der Waals/electrostatic interactions. Typically, intramolecular electron transfer processes within these nanohybrids were probed by fluorescence and transient absorption spectroscopy70, while electrochemical techniques, mainly voltammetry and spectroelectrochemistry, were used to monitor the sizeable electronic interaction between donor and acceptor moieties in the ground state. As mentioned above, in the majority of these studies, the SWNTs acted as electron acceptors. However, SWNTs present much lower oxidation potentials than C60,5 thus making them, at variance with the latter, suitable for their alternative use as electron donors. In fact, supramolecular nanohybrids composed of single-walled carbon nanotubes and fullerene derivatives were constructed and studied. A C60bisadduct bearing a pyrene unit was used to solubilize the SWNTs, and the resulting nanohybrid (Scheme 9.21) was investigated by cyclic voltammetry, transmission electron microscopy, and photophysical experiments that confirmed the existence of p–p interactions and intrahybrid photoinduced charge transfer processes responsible for the quenching of fullerene emission.71 A similar supramolecular approach, in which both p–p stacking stacking of pyrene on the SWNT surface and alkyl ammonium–crown ether interactions were used in the self-assembly process of a fullerene derivative with SWNTs, was recently reported (Scheme 9.22).72 The nanohybrid integrity was probed with various spectroscopic techniques, TEM, and electrochemical measurements. Nanosecond transient absorption studies confirmed electron transfer as the quenching mechanism of the singlet excited state of C60 in the nanohybrid resulting in the formation of SWNT þ / Pyr-NH3 þ /crown-C60 charge-separated state.
250
MOLECULAR DEVICES BASED ON FULLERENES AND CARBON NANOTUBES
Scheme 9.20
The successful use of p–p interactions to anchor an electron-donating extTTF to the surface of SWNT was recently demonstrated (Scheme 9.23).73 Interaction between the concave hydrocarbon skeleton of exTTF and the convex surface of SWNT adds further strength and stability to the SWNT/pyrene-exTTF nanohybrid. Because of the close proximity of the exTTF to the electron acceptor SWNT, a very rapid intrahybrid electron transfer affords a photogenerated radical ion pair, whose lifetime is only a few nanoseconds. The present method for the preparation of SWNT/ exTTF nanohybrids nicely complements the covalent approach and bears a strong
9.3
Scheme 9.21
Scheme 9.22
CARBON NANOTUBES
251
252
MOLECULAR DEVICES BASED ON FULLERENES AND CARBON NANOTUBES
Scheme 9.23
promise for the preparation of systems for photoinduced energy conversion based on electroactive tweezers, in which the pyrene moiety acts as an efficient template for the supramolecular organization of SWNT-based donor–acceptor complexes. Among the many potential applications of SWNTs, field effect transistors (FETs) have been found to be sensitive to various gases—for example, oxygen, nitrogen dioxide, ammonia, and so on. For these applications, the SWNTs are generally decorated by means of noncovalent bonding interactions with bifunctional molecules that can be anchored, on the one hand, onto the nanotubes, and yet are able, on the other hand, to sense a particular biomolecule, thus permitting their detection with FET devices. Very recently, pyrene-modified b-cyclodextrin (pyrene-cyclodextrin)-decorated SWNT/FET devices have been reported (Scheme 9.24) that behave as chemical sensors in aqueous solution, detecting organic molecules as a consequence of their molecular recognition by the pyrene-cyclodextrin derivative.74 Other carbon nanostructures, like carbon nanohorns (CNHs), have also been integrated with photoactive electron donors, such as porphyrins and pyrenes. CNHs are characterized by very high purity and possess a unique morphology comprising a secondary superstructure forming physically inseparable spherical aggregates and are currently under intense investigation for practical technological applications such as photochemical water splitting and as catalysts for the reduction of CO2 to fuels. We recently described the functionalization of carbon nanohorns and the creation of new SWNH/porphyrin nanoconjugates.75 The electrochemical experiments revealed sizeable electronic interactions of porphyrins with SWNHs in the nanoconjugate. Transient absorption spectra permitted to highlight the electron transfer process between the porphyrins and the carbon nanostructures. New metallo-nanostructured materials of carbon nanohorns were recently prepared by the coordination of Cu(II)-2,20 :60 ,20 -terpyridine (CuIItpy) with oxidized carbon nanohorns (CNHs-COOH) and the resulted CNHs-COO-CuIItpy metallonanocomplexes have shown efficient fluorescence quenching, suggesting that electron transfer occurs from the singlet excited state of CuIItpy to CNHs.76 9.4
CONCLUSIONS AND PERSPECTIVES
Carbon is unique in nature. Besides its capability to form complex networks that is fundamental to organic chemistry, elemental carbon also shows unrivalled complexity
9.4
CONCLUSIONS AND PERSPECTIVES
Scheme 9.24
253
254
MOLECULAR DEVICES BASED ON FULLERENES AND CARBON NANOTUBES
among the elements, forming many allotropes, from 3D diamond and graphite to 0D fullerenes, through 1D nanotubes and the most recently obtained 2D graphene.77 Over the past two decades, we have assisted to the subsequent rise of the three different low-dimensional carbon allotropes. After being chemists’ superstar in the 1990s, C60 is still at the center of considerable attention in many different fields of science.6,28e,27,39d,78 Carbon nanotubes are perhaps the most notable representatives of the present nanoworld. However, they are very likely bound to share soon the stage with graphene, the “rapidly rising star on the horizon of materials science and condensed matter physics”.79 Graphene consists of a two-dimensional hexagonal lattice of sp2 carbon, through which electronic conduction can occur via the p-conjugated electron system, and is sometimes classified as a zero-gap semiconductor, since the density of states per unit area vanishes at the Fermi level. The unique properties predicted for graphene comprise a number of very peculiar electronic properties—from an anomalous quantum Hall effect to the absence of localization. As new procedures for the large-scale production of graphene are expected to be developed in the near future, most of such properties—and those still unknown—will be soon experimentally demonstrated, thus permitting the development of the many important technological applications foreseen for this material.
ACKNOWLEDGMENTS The authors gratefully acknowledge the contributions of their collaborators and coworkers mentioned in the cited references. This work was performed with partial support from the University of Bologna, the Fondazione Cassa di Risparmio in Bologna, and the Italian Ministry of University and Research.
REFERENCES 1. (a) V. Balzani, A. Credi, F. M. Raymo, J. F. Stoddart, Angew. Chem., Int. Ed. 2000, 39, 3348–3391.(b) V. Balzani, M. Venturi, A. Credi, Molecular Devices and Machines. A Journey into the Nanoworld Wiley-VCH, Weinheim, 2003. (c) K. Kinbara, T. Aida, Chem. Rev. 2005, 105, 1377–1400. (d) W. R. Browne, B. L. Feringa, Nat. Nanotechnol. 2006, 1, 25–35. (e) H. Tian, Q.-C. Wang, Chem. Soc. Rev. 2006, 35, 361–374. (f) E. R. Kay, D. A. Leigh, F. Zerbetto, Angew. Chem., Int. Ed. 2007, 46, 72–191. 2. (a) H. W. Kroto, J. R. Heath, S. C. O’Brien, R. F. Curl, R. E. Smalley, Nature 1985, 318, 162–163. (b) W. Kr€atschmer, L. D. Lamb, K. Fostiropoulos, D. R. Huffman, Nature 1990, 347, 354–358. 3. (a) F. Diederich, M. Go´mez-Lo´pez, Chem. Soc. Rev. 1999, 28, 263–277. (b) L. Sanchez, N. Martın, D. M. Guldi, Angew. Chem., Int. Ed. 2005, 44, 5374–5382. (c) D. M. Guldi, F. Zerbetto, V. Georgakilas, M. Prato, Acc. Chem. Res. 2005, 38, 38–43. (d) P. D. W. Boyd, C. A. Reed, Acc. Chem. Res. 2005, 38, 235–242. (e) D. Bonifazi, O. Enger, F. Diederich, Chem. Soc. Rev. 2007, 36, 390–414. (f) E. M. Perez, N. Martın, Chem. Soc. Rev. 2008, 37, 1512–1519.
REFERENCES
255
4. D. M. Guldi, P. V. Kamat,in K. M. Kadish, R. S. Ruoff (Eds.), Fullerenes: Chemistry, Physics and Technology Wiley–Interscience, New York, 2000, Chapter 5. 5. (a) Q. Xie, F. Arias, L. Echegoyen, J. Am. Chem. Soc. 1993, 115, 9818–9819. (b) C. A. Reed, K.-C. Kim, R. D. Bolskar, L. J. Mueller, Science 2000, 289, 101–104. (c) C. Bruno, I. Doubitski, I. M. Marcaccio, F. Paolucci, D. Paolucci, A. Zaopo, J. Am. Chem. Soc. 2003, 125, 15738–15739. 6. F. Langa, J.-F. Nierengarten, Fullerenes: Principles and Applications, RSC Publishing, Cambridge, 2007. 7. (a) A. Hirsch, The Chemistry of the Fullerenes, Thieme, Stuttgart, 1994. (b) F. Diederich, L. Isaacs, D. Philp, Chem. Soc. Rev. 1994, 23, 243–255. (c) R. Taylor (Ed.), The Chemistry of the Fullerenes, Advanced Series in Fullerenes, Vol. 4, World-Scientific, River Edge, NJ, 1995. (d) M. Prato, J. Mater. Chem. 1997, 7, 1097–1109. (e) Special issue on Functionalised Fullerene Materials, J. Mater. Chem. 2002, 12. 8. P. Ball, Chem. World 2005, 2, 28. 9. (a) D. M. Guldi, Chem. Commun. 2000, 321–327. (b) D. M. Guldi, N. Martın, J. Mater. Chem. 2002, 12, 1978–1992. (c) Y. Zhang, W. Liu, X. Gao, Y. Zhao, M. Zheng, F. Li, D. Ye, Tetrahedron Lett. 2006, 47, 8571–8574. 10. R. H. Baughman, A. A. Zakhidov, W. A. de Heer, Science 2002, 297, 787–792. 11. (a) M. S. Arnold, A. A. Green, J. F. Hulvat, S. I. Stupp, M. C. Hersam, Nat. Nanotechnol. 2006, 1, 60–65. (b) N. Grobert, Mater. Today 2007, 10 (1–2), 28–35. (c) L. Zhang, S. Zaric, X. Tu, X. Wang, W. Zhao, H. Dai, J. Am. Chem. Soc. 2008, 130, 2686–2691. 12. G. S. Duesberg, M. Burghard, J. Muster, G. Philipp, S. Roth, Chem. Commun. 1998, 435– 436. 13. M. Zheng, A. Jagota, E. D. Semke, B. A. Diner, R. S. Mclean, S. R. Lustig, R. E. Richardson, N. G. Tassi, Nat. Mater. 2003, 2, 338–342. 14. (a) R. Murphy, J. N. Coleman, M. Cadek, B. McCarthy, M. Bent, A. Drury, R. C. Barklie, W. J. Blau, J. Phys. Chem. B 2002, 106, 3087–3091. (b) A. Nish, J.-Y. Hwang, J. Doig, R. J. Nicholas, Nat. Nanotechnol. 2007, 2, 640–646. 15. (a) S. Giordani, S. D. Bergin, V. Nicolosi, S. Lebedkin, M. M. Kappes, W. J. Blau, J. N. Coleman, J. Phys. Chem. B 2006, 110, 15708–15718. (b) S. D. Bergin, V. Nicolosi, P. V. Streich, S. Giordani, Z. Sun, A. H. Windle, P. Ryan, N. P. P. Niraj, Z.-T. T. Wang, L. Carpenter, W. J. Blau, J. J. Boland, J. P. Hamilton, J. N. Coleman, Adv. Mater. 2008, 20, 1876–1881. 16. (a) S. Niyogi, M. A. Hamon, H. Hu, B. Zhao, P. Bhowmik, R. Sen, M. E. Itkis, R. C. Haddon, Acc. Chem. Res. 2002, 35, 1105–1113. (b) S. Banerjee, T. Hemraj-Benny, S. S. Wong, Adv. Mater. 2005, 17, 17–29. 17. A. Penicaud, P. Poulin, A. Derre, E. Anglaret, P. Petit, J. Am. Chem. Soc. 2005, 127, 8–9. 18. S. Ramesh, L. M. Ericson, V. A. Davis, R. K. Saini, C. Kittrell, M. Pasquali, W. E. Billups, W. W. Adams, R. H. Hauge, R. E. Smalley, J. Phys. Chem. B 2004, 108, 8794–8798. 19. (a) A. Bianco, K. Kostarelos, C. D. Partidos, M. Prato, Chem. Commun. 2005, 571–577. (b) J. Robertson, Mater. Today 2007, 10 (1–2), 36–43.(c) M. Endo, M. S. Strano, P. M. Ajayan, Carbon Nanotubes, Vol. 111, Springer-Verlag, Berlin, 2008, pp. 13–61. 20. J. L. Delgado, M. A. Herranz, N. Martın, J. Mater. Chem. 2008, 18, 1417–1426. 21. (a) P. L. Boulas, M. Go´mez-Kaifer, L. Echegoyen, Angew. Chem., Int. Ed. 1998, 37, 216–247. (b) L. Echegoyen, L. E. Echegoyen, Acc. Chem. Res. 1998, 31, 593–601. (c) B. S. Sherigara, W. Kutner, F. D’Souza, Electroanalysis 2003, 15, 753–772. (d) N. Armaroli, G.
256
22. 23. 24. 25. 26.
27.
28.
29.
30. 31. 32.
33.
34.
MOLECULAR DEVICES BASED ON FULLERENES AND CARBON NANOTUBES
Accorsi, F. Song, A. Palkar, L. Echegoyen, D. Bonifazi, F. Diederich, Chem. Phys. Chem. 2005, 6, 732–743. (e) X. Lu, Z. Chen, Chem. Rev. 2005, 105, 3643–3696. (f) K. Gong, Y. Yan, M. Zhang, L. Su, S. Xiong, L. Mao, Anal. Sci. 2005, 21, 1383–1393. (a) A. D. J. Haymet, Chem. Phys. Lett. 1985, 122, 421–424. (b) R. L. Disch, J. M. Schulman, Chem. Phys. Lett. 1986, 125, 465–466. J. W. Bausch, G. K. S. Prakash, G. A. Olah, D. S. Tse, D. C. Lorents, Y. K. Bae, R. Malhotra, J. Am. Chem. Soc. 1991, 113, 3205–3206. (a) Q. Xie, E. Perez-Cordero, L. Echegoyen, J. Am. Chem. Soc. 1992, 114, 3978–3980. (b) Y. Oshawa, T. Saji, J. Chem. Soc. Chem. Commun. 1992, 781–782. L. Echegoyen, L. E. Echegoyen,in H. Lund, O. Hammerich (Eds.), Organic Electrochemistry Marcel Dekker, New York, 2001, Chapter 7. (a) E. Garcia, J. Kwak, A. J. Bard, Inorg. Chem. 1988, 27, 4377–4382. (b) A. J. Bard, E. Garcia, S. Kukharenko, V. V. Strelets, Inorg. Chem. 1993, 32, 3528–3531. (c) P. Ceroni, F. Paolucci, C. Paradisi, A. Juris, S. Roffia, S. Serroni, S. Campagna, A. J. Bard, J. Am. Chem. Soc. 1998, 120, 5480–5487. C. Bruno, M. Marcaccio, D. Paolucci, C. Castellarin-Cudia, A. Goldoni, A. V. Streletskii, T. Drewello, S. Barison, A. Venturini, F. Zerbetto, F. Paolucci, J. Am. Chem. Soc. 2008, 130, 3788–3796. (a) H. Imahori, Y. Sakata, Adv. Mater. 1997, 9, 537–546. (b) N. Martın, L. Sanchez, B. Illescas, I. Perez, Chem. Rev. 1998, 98, 2527–2548. (c) S. Fukuzumi, D. M. Guldi,in V. Balzani (Ed.), Electron Transfer in Chemistry, Vol. 2, Wiley-VCH, Weinheim, 2001, pp. 270–337. (d) M. D. Meijer, G. P. M. van Klink, G. van Koten, Coord. Chem. Rev. 2002, 230, 140–162. (e) J. Mater. Chem.2008. 18 (Theme Issue Carbon Nanostructures). (a) M. N. Paddon-Row,in F. V€ogtle, J. F. Stoddart, M. Shibasaki (Eds.), Stimulating Concepts in Chemistry, Wiley-VCH, Weinheim, 2000, p. 267.(b) M. N. Paddon-Row,in V. Balzani (Ed.), Electron Transfer in Chemistry, Vol. 3, Wiley-VCH, Weinheim, 2001, p. 179. H. Imahori, K. Hagiwara, T. Akiyama, M. Aoki, S. Taniguchi, T. Okada, M. Shirakawa, Y. Sakata, Chem. Phys. Lett. 1996, 263, 545–550. M. Bendikov, F. Wudl, D. F. Perepichka, Chem. Rev. 2004, 104, 4891–4945. (a) T. Chuard, R. Deschenaux, Helv. Chim. Acta 1996, 79, 736–741. (b) S. Campidelli, R. Deschenaux, Helv. Chim. Acta 2001, 84, 589–593. (c) T. Chuard, R. Deschenaux, J. Mater. Chem. 2002, 12, 1944–1951. (d) S. Campidelli, R. Deschenaux, J.-F. Eckert, D. Guillon, J.F. Nierengarten, Chem. Commun. 2002, 656–657. (e) S. Campidelli, E. Vazquez, D. Milic, M. Prato, J. Barbera, D. M. Guldi, M. Marcaccio, D. Paolucci, F. Paolucci, R. Deschenaux, J. Mater. Chem. 2004, 14, 1266–1272. (f) S. Campidelli, J. Lenoble, J. Barbera, F. Paolucci, M. Marcaccio, D. Paolucci, R. Deschenaux, Macromolecules 2005, 38, 7915–7925. (g) S. Campidelli, M. Severac, D. Scanu, R. Deschenaux, E. Vazquez, D. Milic, M. Prato, M. Carano, M. Marcaccio, F. Paolucci, G. M. Rahman, D. M. Guldi, J. Mater. Chem. 2008, 18, 1504–1509. (a) V. V. Yanilkin, N. V. Nastapova, V. P. Gubskaya, V. I. Morozov, L. S. Berezhnaya, I. A. Nuretdinov, Russian Chem. Bull. 2002, 51, 72. (b) M. O¸cafrain, M. A. Herranz, L. Marx, C. Thilgen, F. Diederich, L. Echegoyen, Chem. Eur. J. 2003, 9, 4811–4819. (a) A. Juris, V. Balzani, F. Barigelletti, S. Campagna, P. Belser, A. von Zelewsky, Coord. Chem. Rev. 1988, 84, 85–277. (b) J. P. Sauvage, J. P. Collin, J. C. Chambron, S. Gillerez, C. Coudret, V. Balzani, F.Barigelletti, L. DeCola, L. Flamigni,Chem. Rev. 1994, 94,993–1019. (c) V. Balzani, A. Juris, M. Venturi, S. Campagna, S. Serroni, Chem. Rev. 1996, 96, 759–834.
REFERENCES
257
35. C. Kaes, A. Katz, M. W. Hosseini, Chem. Rev. 2000, 100, 3553–3590. 36. (a) M. Maggini, D. M. Guldi, S. Mondini, G. Scorrano, F. Paolucci, P. Ceroni, S. Roffia, Chem. Eur. J. 1998, 4, 1992–2000. (b) D. M. Guldi, M. Maggini, E. Menna, G. Scorrano, P. Ceroni, M. Marcaccio, F. Paolucci, S. Roffia, Chem. Eur. J. 2001, 7, 1597–1605. (c) C. Nasr, D. M. Guldi, M. Maggini, F. Paolucci, S. Hotchandani, Fullerenes, Nanotubes, and Carbon Nanostructures 2003, 11, 121–133. 37. (a) A. Polese, S. Mondini, A. Bianco, C. Toniolo, G. Scorrano, D. M. Guldi, M. Maggini, J. Am. Chem. Soc. 1999, 121, 3446–3452. (b) G. Possamai, E. Menna, M. Maggini, M. Carano, M. Marcaccio, F. Paolucci, D. M. Guldi, A. Swartz, Photochem. Photobiol. Sci. 2006, 5, 1154–1164. 38. (a) J. C. Hummelen, B. W. Knight, F. LePeq, F. Wudl, J. Yao, C. L. Wilkins, J. Org. Chem. 1995, 60, 532–538. (b) R. A. J. Janssen, J. C. Hummelen, K. Lee, K. Pakbaz, N. S. Sariciftci, A. J. Heeger, F. Wudl, J. Chem. Phys. 1995, 103, 788–793. 39. (a) G. Li, V. Shrotriya, J. Huang, Y. Yao, T. Moriarty, K. Emery, Y. Yang, Nat. Mater. 2005, 4, 864–868. (b) W. Ma, C. Yang, X. Gong, K. Lee, A. J. Heeger, Adv. Funct. Mater. 2005, 15, 16171622. (c) M. Reyes-Reyes, K. Kim, D. L. Carroll, Appl. Phys. Lett. 2005, 87, 083506. (d) C. Yang, J. Y. Kim, S. Cho, J. K. Lee, A. J. Heeger, F. Wudl, J. Am. Chem. Soc. 2008, 130, 6444–6450. 40. (a) N. Camaioni, G. Ridolfi, G. Casalbore-Miceli, A. Possamai, M. Maggini, Adv. Mater. 2002, 14, 1735–1738. (b) M. M. Wienk, J. M. Kroon, W. J. H. Verhees, J. Knol, J. C. Hummelen, P. A. van Hal, R. A. J. Janssen, Angew. Chem., Int. Ed. 2003, 42, 3371–3375. (c) D. Chirvase, Z. Chiguvare, M. Knipper, J. Parisi, V. Dyakonov, J. C. Hummelen, J. Appl. Phys. 2003, 93, 3376–3383. (d) F. Padinger, R. Rittberger, N. S. Sariciftci, Adv. Funct. Mater. 2003, 12, 85–88. unes, H. 41. (a) J. L. Segura, N. Martın, D. M. Guldi, Chem. Soc. Rev. 2005, 34, 31–47. (b) S. G€ Neugebauer, N. S. Sariciftci, Chem. Rev. 2007, 107, 1324–1338. (c) B. C. Thompson, J. M. J. Frechet, Angew. Chem., Int. Ed. 2008, 47, 58–77. 42. W.-S. Li, Y. Yamamoto, T. Fukushima, A. Saeki, S. Seki, S. Tagawa, H. Masunaga, S. Sasaki, M. Takata, T. Aida, J. Am. Chem. Soc. 2008, 130, 8886–8887. 43. (a) F. Diederich, L. Echegoyen, M. Go´mez-Lo´pez, R. Kessinger, J. F. Stoddart, J. Chem. Soc., Perkin Trans. 1999, 2, 1577–1586. (b) M. T. Rispens, L. Sanchez, J. Knol, J. C. Hummelen, Chem. Commun. 2001, 161–162. (c) J. J. Gonzalez, J. de Mendoza, S. Gonzalez, E. M. Priego, N. Martın, C. Luo, D. M. Guldi, Chem. Commun. 2001, 163–164. 44. M. V. Martınez-Dıaz, N. S. Fender, M. S. Rodrıguez-Morgade, M. Go´mezLo´pez, F. Diederich, L. Echegoyen, J. F. Stoddart, T. Torres, J. Mater. Chem. 2002, 12, 2095–2099. 45. (a) A. M. Brouwer, C. Frochot, F. G. Gatti, D. A. Leigh, L. Mottier, F. Paolucci, S. Roffia, G. W. H. Wurpel, Science 2001, 291, 2124–2128. (b) V. Balzani, M. Clemente-Leo´n, A. Credi, B. Ferrer, M. Venturi, A. H. Flood, J. F. Stoddart, Proc. Nat. Acad. Sci. USA 2006, 103, 1178–1183. 46. (a) H.-R. Tseng, S. A. Vignon, J. F. Stoddart, Angew. Chem. 2003. 115, 1529–1533;Angew. Chem., Int. Ed.2003, 42, 1491–1495. (b) A. Altieri, F. G. Gatti, E. R. Kay, D. A. Leigh, D. Martel, F. Paolucci, A. M. Z. Slawin, J. K. Y. Wong, J. Am. Chem. Soc. 2003, 125, 8644–8654. (c) F. Durola, J. P. Sauvage, Angew. Chem. 2007, 119, 3607–3610; Angew. Chem., Int. Ed.2007, 46, 3537–3540.
258
MOLECULAR DEVICES BASED ON FULLERENES AND CARBON NANOTUBES
47. G. Fioravanti, N. Haraszkiewicz, E. R. Kay, S. M. Mendoza, C. Bruno, M. Marcaccio, P. G. Wiering, F. Paolucci, P. Rudolf, A. M. Brouwer, D. A. Leigh, J. Am. Chem. Soc. 2008, 130, 2593–2601. 48. (a) T. Da Ros, D. M. Guldi, A. F. Morales, D. A. Leigh, M. Prato, R. Turco, Org. Lett. 2003, 5, 689–691. (b) D. A. Leigh, M. A. F. Morales, E. M. Perez, J. K. Y. Wong, C. G. Saiz, A. M. Z. Slawin, A. J. Carmichael, D. M. Haddleton, A. M. Brouwer, W. J. Buma, G. W. H. Wurpel, S. Leo´n, F. Zerbetto, Angew. Chem. 2005, 117, 3122–3127; Angew. Chem., Int. Ed. 2005, 44, 3062–3067. 49. (a) A. M. Elizarov, S.-H. Chiu, J. F. Stoddart, J. Org. Chem. 2002, 67, 9175–9181. (b) C. M. Keaveney, D. A. Leigh, Angew. Chem. 2004. 116, 1242–1244; Angew. Chem., Int. Ed.2004, 43, 1222–1224. 50. (a) A. Ikeda, S. Nobukuni, H. Udzu, Z. Zhong, S. Shinkai, Eur. J. Org. Chem. 2000, 3287–3293. (b) S. Campidelli, A. Mateo-Alonso, M. Prato,in F. Langa, J. F. Nierengarten (Eds.), Fullerenes: Principles and Applications, RSC, Cambridge, 2007. 51. T. Gu, C. Bourgogne, J.-F. Nierengarten, Tetrahedron Lett. 2001, 42, 7249–7252. 52. (a) P. R. Ashton, F. Diederich, M. Go´mez-Lo´pez, J.-F. Nierengarten, J. A. Preece, F. M. Raymo, J. F. Stoddart, Angew. Chem. 1997, 109, 1611–1614;Angew. Chem., Int. Ed. Engl.1997, 36, 1448–1451. (b) Y. Nakamura, S. Minami, K. Iizuka, J. Nishimura, Angew. Chem. 2003, 115, 3266–3270;Angew. Chem., Int. Ed.2003, 42, 3158–3162. (c) T. Haino, M. Yanase, Y. Fukazawa, Tetrahedron Lett. 2005, 46, 1411–1414. 53. (a) K. Li, D. I. Schuster, D. M. Guldi, M. A. Herranz, L. Echegoyen, J. Am. Chem. Soc. 2004, 126, 3388–3389. (b) K. Li, P. J. Bracher, D. M. Guldi, A. Herranz, L. Echegoyen, D. I. Schuster, J. Am. Chem. Soc. 2004, 126, 9156–9157. (c) D. I. Schuster, K. Li, D. M. Guldi, C. R. Chim. 2006, 9, 892–908. 54. (a) A. Mateo-Alonso, G. Fioravanti, M. Marcaccio, F. Paolucci, D. C. Jagesar, A. M. Brouwer, M. Prato, Org. Lett. 2006, 8, 5173–5176. (b) A. Mateo-Alonso, D. M. Guldi, F. Paolucci, M. Prato, Angew. Chem., Int. Ed. 2007, 46, 8120–8126. 55. A. Mateo-Alonso, G. Fioravanti, M. Marcaccio, F. Paolucci, G. M. A. Rahman, C. Ehli, D. M. Guldi, M. Prato, Chem. Commun. 2007, 1945–1947. 56. A. Mateo-Alonso, C. Ehli, G. M. A. Rahman, D. M. Guldi, G. Fioravanti, M. Marcaccio, F. Paolucci, M. Prato, Angew. Chem. 2007, 119, 3591–3595; Angew. Chem., Int. Ed.2007, 46, 3521–3525. 57. (a) R. Saito, M. Fujita, G. Dresselhaus, M. S. Dresselhaus, Appl. Phys. Lett. 1992, 60, 2204–2206. (b) C. Dekker, Phys. Today 1999, 52 (5), 22–28. 58. R. Saito, G. Dresselhaus, M. S. Dresselhaus, Phys. Rev. B, 2000, 61, 2981–2990. 59. (a) D. Paolucci, M. Melle Franco, M. Iurlo, M. Marcaccio, M. Prato, F. Zerbetto, A. Penicaud, F. Paolucci, J. Am. Chem. Soc. 2008, 130, 7393–7399. (b) M. Iurlo, D. Paolucci, M. Marcaccio, F. Paolucci, Chem. Commun. (Feature article) 2008, 4867–4874. DOI: 10.1039/b809285k. 60. J. Fischer, D. Petit, A. Penicaud,in M. Monthioux (Ed.), Metananotubes, Wiley, 2009, in press. 61. M. Melle-Franco, M. Marcaccio, D. Paolucci, F. Paolucci, V. Georgakilas, D. M. Guldi, M. Prato, F. Zerbetto, J. Am. Chem. Soc. 2004, 126, 1646–1647. 62. (a) A. M. Rao, P. C. Eklund, S. Bandow, A. Thess, R. E. Smalley, Nature 1997, 388, 257–259. (b) S. Kazaoui, N. Minami, R. Jacquemin, H. Kataura, Y. Achiba, Phys. Rev. B
REFERENCES
63. 64.
65. 66. 67. 68. 69.
70.
71. 72. 73. 74. 75.
76. 77. 78.
79.
259
1999, 60, 13339–13342. (c) P. Petit, C. Mathis, C. Journet, P. Bernier, Chem. Phys. Lett. 1999, 305, 370–374. A. J. Bard, L. R. Faulkner, Electrochemical Methods: Fundamentals and Applications, Wiley, New York, 2001, Chapter 17. (a) A. J. Downard, Electroanalysis 2000, 12, 1085–1096. (b) A. Callegari, M. Marcaccio, D. Paolucci, F. Paolucci, N. Tagmatarchis, D. Tasis, E. Vazquez, M. Prato, Chem. Commun. 2003, 2576–2577. (c) A. Callegari, S. Cosnier, M. Marcaccio, D. Paolucci, F. Paolucci, V. Georgakilas, N. Tagmatarchis, E. Vazquez, M. Prato, J. Mater. Chem. 2004, 14, 807–810. (d) J. Pinson, F. Podvorica, Chem. Soc. Rev. 2005, 34, 429–439. D. M. Guldi, G. M. A. Rahman, V. Sgobba, C. Ehli, Chem. Soc. Rev. 2006, 35, 471–487. D. M. Guldi, J. Phys. Chem. B 2005, 109, 11432–11441. V. Sgobba, D. M. Guldi, J. Mater. Chem. 2008, 18, 153–157. D. M. Guldi, M. Marcaccio, D. Paolucci, F. Paolucci, N. Tagmatarchis, D. Tasis, E. Vazquez, M. Prato, Angew. Chem., Int. Ed. 2003, 42, 4206–4209. (a) D. M. Guldi, G. M. A. Rahman, J. Ramey, M. Marcaccio, D. Paolucci, F. Paolucci, S. Qin, W. T. Ford, D. Balbinot, N. Jux, N. Tagmatarchis, M. Prato, Chem. Commun. 2004, 2034– 2035. (b) D. M. Guldi, G. M. A. Rahman, N. Jux, N. Tagmatarchis, M. Prato, Angew. Chem., Int. Ed. 2004, 43, 5526–5530. (c) M. Alvaro, P. Atienzar, P. de la Cruz, J. L. Delgado, H. Garcia, F. Langa, J. Phys. Chem. B 2004, 108, 12691–12697. (d) L. Sheeney-Haj-Ichia, B. Basnar, I. Willner, Angew. Chem., Int. Ed., 2005, 44, 78–83. (e) D. M. Guldi, H. Taieb, G. M. A. Rahman, N. Tagmatarchis, M. Prato, Adv. Mater. 2005, 17, 871–875. (f) C. Ehli, G. M. A. Rahman, N. Jux, D. Balbinot, D. M. Guldi, F. Paolucci, M. Marcaccio, D. Paolucci, M. Melle-Franco, F. Zerbetto, S. Campidelli, M. Prato, J. Am. Chem. Soc. 2006, 128, 11222– 11231. (g) D. M. Guldi, G. M. A. Rahman, S. Qin, M. Tchoul, W. T. Ford, M. Marcaccio, D. Paolucci, F. Paolucci, S. Campidelli, M. Prato, Chem. Eur. J. 2006, 12, 2152–2161. (a) T. A. Felekis, N. Tagmatarchis, Rev. Adv. Mater. Sci. 2005, 10, 272–276.(b) M. Terrones, A. G. Souza, A. M. Rao, Carbon Nanotubes Vol. 111, Springer-Verlag, Berlin, 2008, pp. 531–566. D. M. Guldi, E. Menna, M. Maggini, M. Marcaccio, D. Paolucci, F. Paolucci, S. Campidelli, M. Prato, G. M. A. Rahman, S. Schergna, Chem. Eur. J. 2006, 12, 3975–3983. F. D’Souza, R. Chitta, A. S. D. Sandanayaka, N. K. Subbaiyan, L. D’Souza, Y. Araki, I. Osamu, J. Am. Chem. Soc. 2007, 129, 15865–15871. Herranz, C. Ehli, S. Campidelli, M. Gutierrez, G. L. Hug, K. Ohkubo, S. Fukuzumi, M. A. M. Prato, N. Martın, D. M. Guldi, J. Am. Chem. Soc. 2008, 130, 66–73. Y.-L. Zhao, L. Hu, J. F. Stoddart, G. Gr€uner, Adv. Mater. 2008, 20, 1910–1915. C. Cioffi, S. Campidelli, C. Soambar, M. Marcaccio, G. Marcolongo, M. Meneghetti, D. Paolucci, F. Paolucci, C. Ehli, G. M. A. Rahman, V. Sgobba, D. M. Guldi, M. Prato, J. Am. Chem. Soc. 2007, 129, 3938–3945. G. Rotas, A. S. D. Sandanayaka, N. Tagmatarchis, T. Ichihashi, M. Yudasaka, S. Iijima, O. Ito, J. Am. Chem. Soc. 2008, 130, 4725–4731. M. I. Katsnelson, Mater. Today 2007, 10 (1–2), 20–27. (a)‘M. Nanjo, P. W. Cyr, K. Liu, E. H. Sargent, I. Manners, Adv. Funct. Mater. 2008, 18, 470–477. (b) R. Bakry, R. M. Vallant, M. Najam-Ul-Haq, M. Rainer, Z. Szabo, C. W. Huck, G. K. Bonn, Int. J. Nanomed. 2007, 2, 639–649. A. K. Geim, K. S. Novoselov, Nat. Mater. 2007, 6, 183–191.
CHAPTER 10
Functional Electroactive Biomolecules XIAOMIN BIN, PIOTR MICHAL DIAKOWSKI, KAGAN KERMAN, and HEINZ-BERNHARD KRAATZ Department of Chemistry, The University of Western Ontario, London, Ontario, Canada
Organometallic bioconjugates derived from amino acids, peptides, and nucleic acids have received significant attention. While the emphasis for amino acid and peptide conjugates is largely on structural design, exploiting inter- and intramolecular hydrogen bonding interactions, research efforts for redox-labeled nucleotides are directed toward the development of deoxyribonucleic acid (DNA) biosensor able to detect complementary sequences for identification purposes of single-nucleotide mismatches. Before discussing the properties of bioconjugates, we want to begin our discussion with a brief overview of the intrinsic electroactivity of amino acids and nucleic acids. These properties have recently been exploited and found applications in electrochemical biosensing schemes. However, we will not discuss such applications in detail, rather mention them in passing. This is followed by an overview of recent work on organometallic conjugates of amino acids, peptides, and nucleic acids with a focus on mono and 1,n0 -disubstituted ferrocene derivatives and their use as scaffolds for the design of foldamers and supramolecular assemblies, and on ferrocene-labeled nucleotides and oligonucleotides and their application in the biosensing systems.
10.1 INTRINSIC ELECTROACTIVITY OF AMINO ACIDS AND NUCLEIC ACIDS Five decades ago, Palecek’s report on the intrinsic electrochemical properties of DNA set the stage for more detailed explorations into its properties.1
Electrochemistry of Functional Supramolecular Systems. Edited by Paola Ceroni, Alberto Credi, and Margherita Venturi Copyright 2010 John Wiley & Sons, Inc.
261
262
FUNCTIONAL ELECTROACTIVE BIOMOLECULES
The oxidation of guanine (G) and adenine (A) follows a two-step mechanism involving the total loss of four electrons and four protons showing current peaks at approximately 0.9 and 1.2 V, respectively. However, the redox properties are dependent on the pH, the ionic strength of the electrolyte, and the electrode material.2 The reader is referred to a recent review by Palecek and coworkers for a more comprehensive discussion regarding the electrochemical mechanism of the oxidation and reduction of DNA bases on carbon and mercury electrodes.3,4 Guanine oxidation is irreversible and occurs in two consecutive steps (Fig 10.1).5 The first step is the oxidation to 8-oxoguanine (8-oxoG) and, at pH 4.5, occurs at þ 0.8 V versus Ag/AgCl, and is a two-electron/two-proton irreversible process6 and the peak at þ 0.95 V corresponds to the one-electron transfer reversible oxidation of the guanine dimers (G-dimer).7,8 The peak at þ 0.55 V corresponds to the reversible two-electron/two-proton oxidation of 8-oxoG, formed on the glassy carbon electrode surface, and is clearly observed after five scans. Adenine oxidation is also irreversible, and occurs in three steps.9 Interestingly, the oxidation peak for adenine splits at lower concentrations into two peaks (see Fig 10.2). The first peak, at þ 1.05 V, corresponds to adenine oxidation and the second, at þ 1.12 V, to the oxidation of adenine dimers. The electroactive adenine oxidation products 2,8-oxoadenine and 2-oxoadenine formed on the electrode surface are detected after 20 scans. Xie et al. have recently reported the electrocatalytic oxidation of guanine, guanosine, and guanosine monophosphate.10 Apilux et al. reported the electrochemical properties of native and thermally denatured fish DNA in the presence of cytosine (C) derivatives and porphyrin using cyclic voltammetry (CV).11 Potentially toxic compounds present in water were evaluated by changes in the electrochemical signal of guanine.12–15 The electrocatalytic oxidation of guanine and DNA on a carbon paste electrode modified with cobalt hexacyanoferrate films was reported.16 Koehne et al.17 developed a miniaturized multiplex label-free electronic chip for rapid nucleic acid analysis based on carbon nanotube-modified microarrays.
0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1 E (V)versus Ag/AgCl
G-dimer
50 pA
8-oxoG
8-oxoG
20 pA
(b)
G-dimer
(a)
263
Guanine
INTRINSIC ELECTROACTIVITY OF AMINO ACIDS AND NUCLEIC ACIDS
Guanine
10.1
0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1 E (V)versus Ag/AgCl
Figure 10.1 Differential pulse voltammetry (DPV) of guanine at pH 4.5 in 0.2 M acetate buffer at a glassy carbon microelectrode. (a) 0.5 mM guanine and (b) 50 mM guanine at a scan rate of 5 mV/s. 5 (Solid line: fifth scan; dotted line: first scan; dashed line: second scan.) Reprinted from Ref. 5a with permission from Elsevier.
Peptides and proteins containing cysteine (Cys) or cystine disulfides exhibit specific electrochemical signals on Hg electrodes, resulting in the formation of HgS bonds. In the presence of cobalt ions, catalytic evolution of hydrogen was observed (Brdicka reaction).18 Hydrogen evolution is also catalyzed at highly negative potentials in the absence of transition metal ions using mercury electrodes with proteins that contained or lacked sulfur-containing amino acids.18–20
Tyrosine (Tyr) exhibits strong adsorption at gold, platinum, and glassy carbon electrodes via the carboxylate group.21–25 At a platinum electrode, the hydrogen adsorption/desorption process of Tyr oxidation was partially blocked, indicating the adsorption of a species on the electrode surface. The surface coverage of tyrosine was estimated using the charge under the platinum oxide reduction peak measured
264
(b) Adenine
(a)
0.1 nA
1
2 - oxoadenine
2, 8 - oxoadenine
0.5 nA
20
A - dimer
Adenine
FUNCTIONAL ELECTROACTIVE BIOMOLECULES
1 1
0.2
0.4
0.6 0.8 1.0 1.2 E (V)versus Ag/AgCl
1.4
0.6
0.7
0.8 0.9 1.0 1.1 E (V)versus Ag/AgCl
1.2
Figure 10.2 Differential pulse voltammetry of adenine at pH 4.5 in 0.2 M acetate buffer at a glassy carbon microelectrode. (a) 1 mM adenine and (b) 10 mM adenine at a scan rate of 5 mV/s. 5 (Dotted line: first scan.) Reprinted from Ref. 5a with permission from Elsevier.
during CV studies by MacDonald and Roscoe.23 The surface coverage was relatively low at low potentials, whereas it increased at higher potentials. Recently combined electrochemical and infrared (IR) spectroscopic studies suggested that the products of the Tyr oxidation remain parallel to the electrode surface.24,25 At about þ 0.7 V, the oxidation of Tyr and tryptophan (Trp) at the electrode surface is similar to the Kolbe mechanism,26 with the loss of ammonium ions and carbon dioxide. Further oxidation at a potential of þ 1.1 V is responsible for the passivation of the electrode. Palecek recently reported on the utilization of the osmium tetroxide-2,20 -bipyridine (Os,bipy) complex as an electroactive marker for tryptophan residues in peptides.27 Os,bipy forms a stable adduct with the tryptophan indole moiety, which is similar to adducts formed by pyrimidine residues in Os,bipy treated nucleic acids. The osmium adducts with pyrimidines, as well as those with Trp in peptides exhibit a well-pronounced electrochemical activity. Os,bipy has been used as an electroactive DNA marker detectable with carbon and mercury electrodes. As is the case for Os,bipy-modified DNA, the Os,bipy-modified peptides give a remarkable catalytic signal at Hg electrodes that can be used for their highly sensitive determination. Cys is an amino acid of great interest and its redox behavior is influenced by the electrode materials, and buffer and electrolyte conditions. S-Based oxidation results in the formation of sulfinic or sulfonic acids. The electrochemical oxidation takes place in two irreversible steps at gold and platinum rotating disc electrodes at peak
10.1
INTRINSIC ELECTROACTIVITY OF AMINO ACIDS AND NUCLEIC ACIDS
265
positions of þ 0.8 and þ 1.1 V (versus saturated calomel electrode (SCE)) and þ 0.85 and þ 1.1 V (versus SCE) at platinum and gold electrodes, respectively.28–30 The initial redox event is presumably due to the oxidation of Cys to the disulfide cystine.29 The total charge for the Cys oxidation is dependent on the waiting time at potentials where no oxidation occurred, which is presumably due to thiol adsorption onto gold in acidic media. Since the amount of Cys that is chemisorbed, is also time dependent, there is a direct correlation between the waiting time and the oxidation peak area. These results indicate that the adsorption was rate-determining for the first oxidation. Johll et al.28 determined that the second anodic peak was due to the oxidation of adsorbed Cys molecules on the electrode surface as well as Cys in solution. Mass spectrometry studies of the products after 16 h of electrolysis show a presence of 70% cystine and 30% cysteic acid. Alternatively, the second peak could be the result of cystine oxidation to cysteic acid. Cys oxidizes irreversibly at þ 0.71 V (versus calomel electrode using 0.05 M KCl, 0.097 V positive to the SCE) at gold electrodes in 0.01 M HClO4. These experiments confirmed that the oxidation products remain adsorbed on the gold surface and are only desorbed during gold oxidation. Possari et al. determined the Cys concentration in basic samples using the reductive desorption of the chemisorbed Cys.30 This electrochemical method relies on the adsorption of Cys onto a gold electrode via its thiol moiety, followed by the reductive desorption and cleavage of the previously formed AuS bond. The adsorption of the aromatic ring of phenylalanine (Phe) on surfaces has been investigated by electrochemical means.31–33 Electrochemical quartz crystal microbalance, chronocoulometry, impedance spectroscopy, and IR spectroscopy were coupled with cyclic voltammetry measurements. The formation of Pt oxide was suppressed with the adsorption of phenylalanine at the platinum electrode.33 The surface charge density for Phe was estimated from the changes of the Pt oxide reduction between CV experiments in the absence and in the presence of amino acid. CV measurements of Phe at Au(111) single-crystal electrodes exhibited a threestep adsorption process with three pairs of quasi-reversible peaks at 0.1, þ 0.3, and þ 0.5 V.34 IR studies showed that the surface orientation of the adsorbed Phe molecules depended on the electrode surface potentials.35 IR-visible sum-frequency generation vibrational spectroscopy studies showed that the adsorption was potential dependent. Below the potential of zero charge, the molecules adopt a horizontal configuration with the amine group and the aromatic ring facing toward the electrode surface.31 Glycine (Gly) is the simplest amino acid with a methylene spacer separating the amino and carboxylic acid group. Adsorption of Gly at gold and platinum electrodes has been investigated in acidic, neutral, and basic media.35–37 Surface oxidation of Pt occurs in the same potential region as the amino acid oxidation. Roscoe compared the Pt surface oxide reduction in the absence and in the presence of Gly to estimate the amount of Gly adsorbed.36,37 Gly exhibited a high packing density and oxidation charge. The adsorbed Gly species was detected to decompose through a mechanism that is similar to the Kolbe cycle. There are a number of examples, in which the inherent redox activity of the protein is exploited for its detection and quantification. Wang’s insuline quantification is
266
FUNCTIONAL ELECTROACTIVE BIOMOLECULES
based on exploiting the redox activity of a Tyr residue present in the protein and will serve as an example for the tremendous bioanalytical application of this approach. Using chronoamperometric stripping analysis, it was possible to detect and quantify insulin at carbon paste electrodes. However, insulin adsorbs on the electrode surfaces and a decay of the electrochemical signal (chronoamperometric stripping) was observed. This problem was overcome by anodically pretreating the electrode and by using single-use disposable sensors.38,39 Redox reactions and adsorption of nucleic acids and proteins at the electrode surfaces provide label-free electrochemical detection possibilities. The detection requires the presence of redox-active nucleobases such as guanine and adenine and amino acids such as Cys, Tyr, or Trp within the biomolecule. But even nonredoxactive protein can be analyzed exploiting the ability of the protein to adsorb (or “foul”) surfaces. In essence, this disrupts the electrical double layer and provides a quantifiable change in the electrical signal. Although, these approaches are simple and allow a rapid and simple initial investigation into whether direct label-free detection is possible or not, they have profound limitations, including a lack of sensitivity and selectivity and more importantly, they cannot be used in complex sample matrices, where various protein molecules are present. “Label-free” protein detection is, therefore, commonly achieved by employing recognition elements with high affinity for the target protein. This ensures much improved specificity, especially when dealing with highly complex sample matrices such as urine, cerebral spinal fluid, and serum.
10.2
FERROCENE-CONJUGATED PEPTIDES
This section of the chapter summarizes some of the recent results in this area that is largely driven by ferrocene (Fc) acting as a scaffold that directs the structural properties of the conjugate and provides access to conjugates with well-defined structural and electrochemical properties. The particular choice of the Fc scaffold influences the ability of the conjugate to engage in interstrand hydrogen bonding and also influences the formation of supramolecular assemblies. As was reported before, conjugates of ferrocenecarboxylic acid (FcCOOH) often give rise to one-dimensional hydrogen-bonded chains, whereas peptide conjugates of ferrocene-1,10 -dicarboxylic acid, ferrocene-1,10 diamine, and 10 -aminoferrocene-1-carboxylic acid (Fca) can give rise to hydrogen-bonded b-sheet-like structures, in which the directionality of the peptide substituents can be altered allowing the design of a range of b-sheet models (Fig 10.3). Several synthetic methods are available for the preparation of Fc peptide conjugates, of which the “active ester” method is most compatible with biological environments.41 This method works under mild conditions with an isolable FcCOactive ester, in which reactive heterocyclic isolable esters are formed. These can be isolated or reacted in situ with suitable peptides to give the desired Fc-peptide conjugate. These active esters can be used as stoichiometric Fc delivery reagents, which make them suitable for automated solid-phase synthesis of Fc-peptide conjugates.
10.2
O
R N H
O
H N O
FERROCENE-CONJUGATED PEPTIDES
R
O
Fe
O
Fe
R R
O
H N O
ferrocenecarboxylic acid
O
H N
N H
R
267
N H
R
O
ferrocene-1,1′-dicarboxylic acid O
R
H N O
N H
N H
O
R Fe
Fe
O
R
H N
H N
O N H
ferroceneamine
H N R R
H N R
O
N H
ferrocene-1,1′-diamine R
O N H Fe
O
O N H
O
H N R R
H N R
O
N H
1′-aminoferrocene-1-carboxylic acid
Figure 10.3 The ferrocene-peptide conjugate family. Arrows point from the C to the N termini of the peptides. Copyright Wiley-VCH Verlag GmbH & Co. KGaA. Reproduced with permission from Ref. 40.
The acid chloride method was successfully employed to produce a number of 1,10 disubstituted ferrocenoyl peptide and amino acid systems.42–44 Active esters, initially employed by Degani and Heller for introducing the Fc-label into larger proteins at the e-amino group of lysines under mild conditions,45,46 were successfully employed in the synthesis of a range of Fc-peptide conjugates.41 Recently, a series of macrocyclic 1,n0 -ferrocenyl peptides conjugates Fc[CSA]2, Fc[Gly-CSA]2, Fc[Ala-CSA]2, Fc[Val-CSA]2, and Fc[Leu-CSA]2 (CSA ¼ cysteamine) were reported, which were formed by the reaction of ferrocenedicarboxylic acid with peptides-cystamine conjugates at high dilutions.47 These systems exhibit H-bonding involving the amide NH in solution as shown by their temperaturedependent nuclear magnetic resonance (NMR) spectra. With the exception of Fc [CSA]2, the Fc macrocycles display a intramolecular H-bonding in the solid state involving the amino acids proximal to the Fc that was postulated before by Herrick and coworkers.43 This work was extended to larger macrocycles with a designed
268
FUNCTIONAL ELECTROACTIVE BIOMOLECULES
Figure 10.4 Schematic representation of the formation of the pseudo-b-barrel formed from molecular building blocks by tiling. (a) Molecular structure of Fc[Gly-Val-CSA]2, (b) formation of b-sheets through intermolecular N(H)O¼C hydrogen bonding, (c) H-bonding interactions between four molecules to form a b-barrel, and (d) side view molecular-surface representation showing barrel. Copyright Wiley-VCH Verlag GmbH & Co. KGaA. Reproduced with permission from Ref. 49. (See the color version of this figure in Color Plates section.)
interface to foster controlled intermolecular H-bonding interactions. The self-assembly of Fc[Gly-Val-CSA]2 and Fc[Gly-Ile-CSA]2 leads to the first model for a b-barrel system, a ubiquitous structural motif in proteins.48 The b-barrel model has eight b-strands arranged almost parallel to the axis of the channel with an internal pore diameter of 8 A (Fig 10.4). Cyclic and acyclic Fc-His (Fc-histidine) conjugates are shown in Fig 10.5. These systems were prepared by high-dilution methods. The single-crystal structure of cyclo-Fc-His indicated that both proximal Fc carbonyl groups are syn with respect to each other, a new structural motif for Fc-amino acid conjugates. Interestingly, this conformation of the CO groups allows effective binding to alkali metal cations.
10.2
FERROCENE-CONJUGATED PEPTIDES 90
(a)
O1
80
C10 C6 C11
O4
C9
C12 N3
C18
C1
N1
C8
C7
C13
Fel C3
C2 N2
C19
C14 O3
70
C5 C17
C15
C4 C16
Current (mA)
O2
269
(b) 1 + 1+Na
60 50 40 30 20 10 0
–10 0.55
0.65
0.75 0.85 Potential (V)
0.95
1.05
Figure 10.5 (a) Molecular structure of cyclo-Fc[CO]2His showing the two carbonyl groups syn with respect to each other, allowing for effective metal ion binding and (b) Differential pulse voltammogram at a pulse width of 0.05 s of compound 1 before (bold curve) and after (dotted curve) the addition of NaClO4 to the solution (1 mM solute in MeCN) using tetrabuthylammonium perchlorate (TBAP) (0.1 M) as supporting electrolyte (glassy carbon, Ag/AgCl, Pt). Copyright Wiley-VCH Verlag GmbH & Co. KGaA. Reproduced with permission from Ref. 48.
Binding is evaluated by cyclic voltammetry monitoring the half-wave potential of the Fc group (Fig 10.7). Cation binding causes a shift to lower potential in the following order: Na þ > Li þ > K þ , Cs þ .49 1,10 -Bis(tert-butoxycarbonylamino)ferrocene, a protected derivative of 1,10 -diaminoferrocene, was synthesized by a very convenient method for the preparation of 1,10 -diaminoferrocene.50 Some initial amino acids conjugates of L- and D-Ala were reported. In these conjugates, the amino acids substituents engage in strong interstrand H-bonding generating 14-membered H-bonded rings, a motif previously unrealized in Fc-amino acid and peptide conjugates (Fig 10.6). As is the case for
Figure 10.6 The intramolecular interstrand H-bonding pattern in amino acid conjugates of disubstituted Fc derivatives results in the formation of 10-membered rings for ferrocene dicarboxylic acid conjugates, 12-membered rings for Fca-conjugates, and 14-membered rings for ferrocenediamine conjugates. Reproduced by permission from Ref. 50 of The Royal Society of Chemistry.
270
FUNCTIONAL ELECTROACTIVE BIOMOLECULES
054
C3 C4
C59B C61 C60
C5 C59A N52
052
058
053
C1
C2 C7B C6
N1 Fe1
C7A
C53
C52
C54 N2
C56 051
C51 C57A
01
C55
N51
C8
02 C57B
03
C9
Figure 10.7 Molecular structure of Boc-Ala-Ala-Fca-Ala-OMe showing interstrand Hbonding interactions. The Fc unit itself is in a rigid P-helical conformation stabilized by the H-bonding interactions. Copyright Wiley-VCH Verlag GmbH & Co. KGaA. Reproduced with permission from Ref. 40.
other Fc-peptide conjugates, the chiral organization of the conjugate is governed to a large extent by the chirality of the podant amino acid. The recent study of Fca conjugates of amino acids and peptides show that the Fca group acts as a strongly turn inducer, promoting H-bonding interactions between podant peptide chains (Fig 10.7). This opened an interesting problem. Could conjugates be synthesized that possess alternating amino acid and Fca sequences? The resulting oligomer would have a turn inducer adjacent to a H-bonding amino acid, which should result in the formation of a foldamer that may exhibit b-helical characteristics. Thus, a series of Ala (alanine) conjugates of Fca were prepared having alternating Fca-Ala sequence. The results clearly show that indeed the resulting foldamer adopts the expected structure and that the amino acid spacers engage in intrahelix H-bonding, which stabilize the overall structure of the foldamers. Figure 10.8 shows the structure of leftand right-handed pseudo-b-helical Fca-Ala foldamers.51 An entirely different approach to structural design of Fc-peptide conjugates was taken making use of amino acids with branch points that allow the synthesis of dendrimeric Fc-conjugates, in which the peptide dendrimer can encapsulate the Fc group and shield it from the exterior. Nonpeptidic dendrimeric systems are discussed in more detail in chapters by Gorman and Astruc. Generally, the dendrimer shell will influence the redox properties of a redox-active core. For high dendrimer generations, the central core can be completely encapsulated and isolated from the exterior by the surrounding dendrimer sheeth.52–58 However, different conformational changes that may occur in the dendrimers or the orientation effects between the
10.2
FERROCENE-CONJUGATED PEPTIDES
271
Figure 10.8 Molecular structure of (a) Boc[Ala-Fca]2-OMe (left) and its helical backbone (b) and (c) the molecular structure of Boc[Ala-Fca]3-OMe and its molecular backbone (d). The dashed lines represent H-bonds. It is important to note that the intra- and intermolecular H-bonding patterns are identical for both systems and that 12-membered H-bonding is observed. Importantly, Fca-conjugates of D-Ala are exact mirror images of the structures shown here. Copyright Wiley-VCH Verlag GmbH & Co. KGaA. Reproduced with permission from Ref. 51.
272
FUNCTIONAL ELECTROACTIVE BIOMOLECULES
dendrimer and the electrode surface influence this phenomenon.59,60 In general, the resulting dendrimer films possess a high mechanical stability and can be functionalized without any loss of dendrimer from the surface.61,62 The synthesis and characterization of Glu- (gluitamic acid) based peptide dendrimer generations 1–6 (G1–G6) (Fig 10.9), which possess an Fc group at the center of the dendrimer, was reported earlier. Two classes of dendrimers were prepared of the type Fc-Glu-dendrimer and Fc[Glu-dendrimer]2. Both types of dendrimers were ester protected, and thus had a hydrophobic exterior interface. Backfolding of the peptide dendrimer and intramolecular H-bonding increases with increasing dendrimer generation.63,64 The intramolecular H-bonding was found responsible for the formation of a globular structure of these peptide dendrimers in particular for the higher generations and the encapsulation of the redox core. In contrast, lower generations possess an open architecture, in which the Fc group is readily accessible to the environment. Although the redox properties were modulated by the peptide case, monosubstituted Fc-Glu-dendrimers dispayed redox activity up to G6, whereas Fc[Glu-dendrimer]2 systems were more effective in shielding the Fc core and no redox activity was observed at generations higher than G4. Interestingly, when studied at an interface modified with 1-mercaptoundecanoic acid (MUA), the modulation of the redox properties for Fc-Glu-dendrimers is clearly observed.65 G1–G5 films exhibited electron transfer from the Fc core to the electrode. G6 did not exhibit any redox activity when adsorbed on MUA-modified surfaces due to the encapsulation of the Fc-core by the peptide dendrimers. In addition, conjugated peptide dendrimer prevents access of counterions to the core of the dendrimer limiting the formation of Fc þ ClO4- ion pairs. Rigidity and distance effects may also contribute significantly to the loss of the redox signal. Interestingly, the formal potential, E0, of surface-immobilized Fc-peptide dendrimers is unaffected by dendrimer generation/size that is in contrast to solution measurements. More recently, Fc-peptides have been exploited in biosensing applications, in which the peptide plays the role of the recognition element and the Fc is the redox reporter. For this purpose, the Fc-peptide conjugate is attached in close proximity to a gold surface, which essentially will mean that any Fc redox process remains virtually unaffected by diffusive processes. The peptide, the recognition element, is pointing into the solution containing the analyte. The interaction between the recognition sequence and electrode surface will not be hindered by the interaction with the target analyte. This approach enhances the sensitivity and lowers the susceptibility to matrix effects. Examples have been reported for the detection of papain and human immunodeficiency virus (HIV) protease. For example, the conjugate Fc-Gly-Gly-Tyr-Arg contains the tetrapeptide Gly-Gly-Tyr-Arg as the recognition element. This peptide is an inhibitor for papain and has been used to purify papain by affinity chromatography.66 With increasing amounts of papain present in the assay buffer, the formal potential, E0, of the Fc group is shifted to anodic potential and there is a clear linear relationship between papain concentration and E0. A similar approach was used for the detection of HIV-I protease (see Fig 10.10). Pepstatin (Val-Val-Sta-Ala-Sta) a well-known potent inhibitor of aspartyl proteases contains the unusual amino acid, statine (Sta) or (3S,4S)-4-amino-3-hydroxy- 6-methylheptanoic
10.2
G1
FERROCENE-CONJUGATED PEPTIDES
273
G3
G2
G4
G6
G5 OO
HN O
Fe
O OC2H5 OC2H5
2
1
2
O
OH O
OH O
OHO
OH
1
Dendrimer
Washing
MUA
H
2
O
N
(a)
O
N
O H
O
O
(b)
H
(c) H
O
O
O
H O
O
1
Figure 10.9 Top: Fc-Glu-dendrimer from G1 to G6. Bottom: Preparation of the MUAdendrimer interface. (1) Formation of a MUA film on Au. (2) Immersion of the MUA interaction in deprotected Fc-Glu-dendrimer. This dendrimer has a hydrophilic carboxylate surface. H-bonding interactions between the dendrimers and the MUA film include (a) NH. . .OC interactions, (b) COOH. . .OC amide interactions, and (c) COOHOC ester interactions. Reprinted with permission from Ref. 65. Copyright 2006 American Chemical Society.
274
FUNCTIONAL ELECTROACTIVE BIOMOLECULES
Figure 10.10 (a) (1) Fc-pepstatin bioconjugate and the schematic representation of the diluted self-assembled peptide film on a gold electrode surface. (2) Binding of HIV-1 protease to Fc-pepstatin causes a significant shift of E0 and a modulation in the current intensity as shown in blue curve due to reduced accessibility of the counter ion to the Fc group. However, even at full binding, the redox signal of the Fc remains clearly visible. Reproduced by permission from Ref. 67 of The Royal Society of Chemistry. (b) (1) CV of a Au nanoparticle-modified gold electrode modified with the Fc-pepstatin conjugate in the presence of increasing concentration of HIV-I protease. (2) Linear relationship between the HIV-I protease concentration and the formal potential of the Fc redox group. Reprinted with permission from Ref. 68. Copyright 2006 American Chemical Society.
acid. Films of Fc-conjugated pepstatin on gold microelectrodes were used as an electrochemical biosensor for the detection of HIV-1 protease with good results.67 With the help of gold nanoparticle modified carbon nanotubes, it was possible to achieve extremely high detection sensitivity below the picomolar level.68 This section of the chapter summarizes some of the recent results in the synthesis of Fc-peptide conjugates. The aim was not to give a complete survey over this class of compounds. The interested reader is referred to more comprehensive review.41 These
10.3
Fc-LABELED NUCLEOBASES
275
conjugates are an attractive class of organometallic peptides that are conveniently obtained by solution or solid-phase synthetic strategies that have the potential to be tailored to bind to nucleic acid and protein targets. The structural properties displayed by these bioconjugates are interesting since they provide a higher degree of control over supramolecular assemblies. The biocomponent of the conjugate provides access to designing H-bonding and molecular recognition into the molecule that makes them interesting targets with potential bioanalytical applications. However, there is a lack of understanding of the driving forces giving rise to a particular supramolecular assembly. Currently, there are only a limited number of theoretical studies available on these and related urea peptide systems that might aid in molecular design and shed light on energetic differences between particular conformations.69 But at this point, the predictive value is still lacking.
10.3
Fc-LABELED NUCLEOBASES
Conjugations of redox-active species to oligodeoxynucleotide (ODN) have been extensively studied due to their potential application in quantitative and qualitative DNA analysis, the great interest into the charge transfer properties of DNA, and the development of electrochemical DNA biosensors.70–80 Fc conjugates take a prominent position since Fc-DNA conjugates are readily synthesized by a range of methods. In addition, the Fc group has a diverse substitution chemistry that allows to fine-tune the desired electrochemical properties of the probe. The preparation of Fc-DNA conjugates can be achieved by a range of synthetic methods that were reviewed by Zatsepin et al.,81 van Staveren,41a and Kraatz.41b Essentially, there are three main methods: (1) the conjugation of Fc with an ODN containing a reactive group at the 50 or 30 terminal, (2) the synthesis of an Fc-nucleoside triphosphate conjugate and its subsequent incorporation into the ODN by enzymatic methods, and (3) the synthesis of an Fc-nucleoside phosphoramidite derivative and its subsequent incorporation into an ODN by automatic solid-phase synthesis. Within these strategies, the latter method is most advantageous because it allows the introduction of the Fc group at specific positions within the ODN sequence. In this section of the chapter, we will mainly focus on the electrochemical properties of Fc-conjugated nucleobases, its application to electrochemical biosensors and provide a brief overview of synthesis of these compounds. The first work on Fc derivatives of nucleosides was reported by Gautheron and coworkers, who used a several of Pd-catalyzed CC coupling reactions involving 5-halogen-nucleoside with ethynylferrocene to synthesize Fc-conjugated uridine, deoxyuridine, and adenosine.82 Conjugate formation is also readily achieved by Sonigashira coupling involving ethynylferrocene and 5-iodouracil, iodouridine, 2-desoxyiodouridine, and bromoadenosine,83 yielding the substituted nucleobase 5-ethynylferrocenyluracil (1), 5-ethynylferrocenyluridine (2), and 5-ethynylferrocenyldeoxyuridine (3) as well as the ethynylferrocenyladenosine (7). The 5-ethynyluracil derivatives can be cyclized in the presence of base, for example, NEt3, yielding ferrocenyl derivatives 4–6 (see Scheme 10.1). It is important to point out that such a
276
FUNCTIONAL ELECTROACTIVE BIOMOLECULES
Fe O
O
1. Fc–C=CH (143), (Ph3P)2PdCl2 (cat.)
I
HN
2. Meoh, H2O
N R
O
O NEt3, D
HN O
Fe
N R 1 2 3
NH2
O
N R 4 5 6
NH2 N
N Me3SiO
HN
N
N
Br 1. Fc–C=CH, (Ph3P)2PdCl2 (cat.)
N
N HO
N
2. Meoh
O
N
Fe
O 7
Me3SiO
OH
OSiME3
OH
Scheme 10.1 Sonogashira coupling for ferrocenylated nucleosides. Reprinted with permission from Ref. 41a. Copyright 2006 American Chemical Society.
cyclization is not desirable and will essentially decrease the ability of this base to H-bond with its complementary base on the hybridized strand. Similar observation was reported by Yu et al.,83 and a mechanism has been suggested. Houlton and coworkers84 reinvestigated this reaction and it was reported that it is possible to suppress the cyclization by running the Sonogashira reaction at room temperature for 4 h. Houlton and coworkers85 also made use of trimethylammoniummethylferrocene iodide as a source of the ferrocenemethyl cation to form Fc-nucleobase conjugates. Fc derivatives of thymine, cytosine, and uracil and of N2-acetylguanine and 2-amino-6-chloropurine were prepared and fully characterized. Figure 10.11 shows the structure of 1-ferrocenylmethylcytosine. 0(2)
C(2)
C(18)
N(3)
C(13) C(12)
C(17)
N(1)
C(9)
C(14)
C(4) N(4)
C(8) Fe(1) C(16) C(15)
C(11)
C(5)
C(10) C(6)
Figure 10.11 Molecular structure of 1-ferroenylmethylcytosine. Reproduced by permission from Ref. 85 of The Royal Society of Chemistry.
10.3
Fc-LABELED NUCLEOBASES
277
Grinstaff and coworkers reported a facile on-column derivatization procedure allowing oligodeoxynucleotides to be labeled with an Fc derivative at specific sites.86,87 A Sonagashira cross-coupling reaction, using Pd(PPh3)4 and CuI, links Fc propargylamide to a halogenated base, such as 20 -deoxy-5-iodouridine or 8-bromo-20 -deoxyadenosine, which has been previously incorporated in the ODN strand (see Scheme 10.2). The structure and stability of the B-form DNA duplex are not significantly altered after site-specific labeling with ferrocene, as demonstrated by melting temperature study and circular dichroism (CD) spectroscopy. King and coworkers88,89 synthesized two Fc-labeled analogs of dTTP, 5-(3ferrocenecarboxamidopropenyl-1) 20 deoxyurindine 50 -triphosphate (Fc1-dUTP) and 5-(3-ferroceneacetamidopropenyl-1) 20 deoxyurindine 50 -triphosphate (Fc2-dUTP), Fc1-dUTP shows efficient incorporation into DNA synthesis and is also good substrate for polymerase chain reaction (PCR), however, Fc2-dUTP, which is one more methylene group in the linker between the Fc group and uridine base, acts predominately as a terminator for the polymerase reaction in the primer extension. Fc-carbodiimides carrying different redox potentials were designed and synthesized, these reagents can attach to DNA strands directly by reacting with the imino unit of thymine or guanine base on DNA or of uracil base on ribonucleic acid (RNA) under a basic buffer condition, yielding a labeled product quantitatively in a short period of time.90 Song and Kraatz91 reported syntheses, crystal structures, and electrochemical results for two Fc-modified pyrimidine nucleosides that could potentially be used for investigating electron transfer in DNA. For the first time, Fc was directly attached to the 5-position of deoxyuridine and deoxycytidine via Stille coupling without any linker group in between the redox center and the base, which in essence should enhance the electronic coupling between the two components (see Fig 10.12). The Fc-modified uridine was incorporated into DNA trinucleotides with standard solid-phase synthesis. An interesting Fc conjugate has also been reported recently having an Fc group linked to two nucleobases such as thymine/uracil. These conjugates form supramolecular assemblies by base paring directed self-assembly.92 As pointed out earlier, one of the main driving forces for the preparation of Fc-ODN conjugates is their application to biosensing and to study the basic charge transfer properties of DNA. In the following, we will discuss the redox properties of Fc nucleobase conjugates and its application in DNA biosensors. The Fc group is a valuable redox reporter, which exhibits reversible electrochemical properties that are tolerant to a wide range of environmental conditions. The introduction of electron donor groups into the Cp ring of the Fc group decreases the half-wave potential E1/2, while electron-withdrawing groups shift to more anodic potentials.93 The linker between the Fc group and the nucleobase will critically influence the redox properties and the effectiveness of the electronic coupling between the nucleobase and the Fc group. CV studies carried out by Houlton and coworkers84 were aimed at probing the redox properties of Fc-conjugated thymidine and evaluate the effects of different C2-linkers (CC, C¼C, C C) on the Fc oxidation. Table 10.1 gives an overview of redox potentials of literature reported Fc-nucleobase conjugates. Fc derivatives cover a wide range of redox potentials
278
FUNCTIONAL ELECTROACTIVE BIOMOLECULES
Scheme 10.2 On-column synthesis of Fc-labeled ODNs (see Ref. 21. (a) ODN synthesis, (b) Cross-coupling reaction conditions: Pd(PPh3)4, CuI, TEA (triethanolamine), DMF (dimethylformamide), and (c) Fc-ODN deprotection. Analogous procedure is used for 5-iodourindine modification. Reprinted from Ref. 87 with permission from Elsevier.
10.3
Fc-LABELED NUCLEOBASES
279
C22
C21
C23 C20
C21
C22
C24 Fel
C113 C23
C20 C24 C12
C112
N13
Fe1
C12
C11 C10
C11
C13
C16
C15
C13
C18 C10
C15
C111 O14
C19 N12
C15
N11 C18
C110
C14
C13 C14
C17
N12 O12 O14
C16 O11
C17 N11
O12
O11
C19
C113
C110
C112 C111
O13
Figure 10.12 ORTEP drawings of Fc-pyrimidines obtained by Stille coupling having the Fc directly attached to the nucleobase. Left: Fc-deoxyuridine; Right: Fc-deoxycytidine. Reproduced by permission from Ref. 91 of The Royal Society of Chemistry.
and display a range of linkers that clearly influence their properties. It is interesting to note that this should make it possible to design ODN-based electron transfer chains and explore their properties. However, only a very limited number of multi-Fcsystems have been reported and unfortunately these were not designed to study stepwise site-to-site electron transfer. Fc-nucleobases are usually incorporated into the ODN sequence either by automated solid-phase synthesis or by a PCR-based process and then immobilized on electrode surfaces in order to facilitate electrochemical biosensing. An excellent example of an innovative labeling strategy is provided by Willner and coworkers,94 who used Fc-conjugated uridine (dUTP) as electron transfer mediators between the redox enzymes and the electrode for the amplified bioelectrocatalytic detection of viral DNA. Viral DNA interacts with a thiolated ODN that chemisorbed on a gold surface and is complementary to viral DNA (M13F), thus enabling viral DNA to bind to ODN film. The partially double-stranded DNA (ds-DNA) assembly was interacted with the nucleotide triphosphate that included Fc-conjugated dUTP in the presence of Klenow fragment I polymerase (see Scheme 10.3). The polymerase will read the viral DNA and replicate viral DNA. In this process, approximately 350 Fc groups are incorporated into each DNA replica. The Fc groups are then exploited as redox mediator for the enzymatic oxidation of glucose to gluconic acid. Figure 10.13 shows a DPVof the Fc-labels as they are being incorporated into the viral DNA replica by the polymerase. As would be expected, the electrochemical signal saturates upon completion of the viral DNA replicas. In the presence of glucose oxidase and glucose, the Fc act as redox mediator and a catalytic current is observed due to the conversion of glucose to gluconic acid.
280
DMTrO
DMTrO
DMTrO
Compound
N
O
N
O
O
HN
OH
OH
O
O
HN
OH
O
O
HN
O
N
O
Fc
Fc
Fc
21 a
29 a
27 a
E1/2 (mV)
84
84
84
Reference
N
N
HO
HO
N
N
OH
O
N
O
N
HN O
OH
O
O
HN
NH2
Compound Fc
98
91
209
205 d
97
Reference
140 b, c
E1/2 (mV)
TABLE 10.1 Redox Potentials (Versus Ag/AgCl) of Conjugates of Nucleoside Derivatives with Fc in Aqueous Medium
Fe
Bn
DMTrO
DMTrO
DMTrO
O
HN
OH
OH
O O
N
N
N
O
O
HN
O
OH
O
O
HN
O
Fc
Fc
Fc
140 a
131 a
16 a
84
84
84
N
Fe
N
N
N
NH2
N
Bn
N
N
N
N N
NH2
N
N
Bn
Fe Fe
H2C
CH2
Bn
281
(continued)
98
98
165 d
43 d
98
210 d
282
N
OH
O
N
NH2
OH
O O
N
N
N
b
Solution in MeCN. Solution in THF. c Ag wire. d Refer to Fc/Fc þ . DMTr ¼ (4-MeOC6H4)2PhC.
a
HO
HO
O
(Continued)
HN
Compound
TABLE 10.1
N H
N H
O
O
Fc
Fc
48
262
a
E1/2 (mV)
87
86
Reference
N
Fe
N
N
N
N
H2C
CH2
CH2
Bn
CH2
NH2
N
NH2
Compound
N
N
Fe
Bn
98
98
35 d
62 d
Reference
E1/2 (mV)
10.3
Fc-LABELED NUCLEOBASES
283
Scheme 10.3 Amplified detection of Viral DNA by the generation of a redox-active replica and the bioelectrocatalyzed oxidation of glucose. Reprinted with permission from Ref. 94. Copyright 2002 American Chemical Society.
The full synthesis of two Fc-labeled bases 5-(3-ferrocenecarboxamidopropenyl-1) 20 -deoxyuridine 50 -triphosphate (Fc1-dUTP) and 5-(3-ferroceneacetamidopropenyl1) 20 -deoxyuridine 50 -triphosphate (Fc2-dUTP) were also reported by King and coworkers (see Scheme 10.4).88 These bases were incorporated into a DNA sequence by polymerases (Klenow and T4 DNA polymerases). In both cases, the Fc-labels display fully reversible redox behavior in solution with an E1/2 of 398 mV for Fc1-dUTP and 260 mV for the Fc2-dUTP (versus Ag/AgCl). King reports that the Fc-labeled DNA is detected at femtomolar levels by highperformance liquid chromatography (HPLC) using a coulometric detector. It has to be stressed at this point that the incorporation of the Fc-label by PCR and its facile and cost-effective electrochemical detection is very powerful and could ultimately replace more costly methods for nucleic acid analysis. King and Gooding also reported the incorporation of Fc-labeled triphosphates into an ODN using a single base extension reaction, which allowed the selective attachment of the Fc-label to the ODN.89 The extension of this work to RNA has also been described. Fc1-dUTP could
284
FUNCTIONAL ELECTROACTIVE BIOMOLECULES
01μA) 0
(b) 0.3
0.2
0.25
0.2
0
d
0.2 0 10 20 30 40 50 60 70
Time (min)
0.32 0.24
0.5 0.25
0.3
0.1
0.1 0.2 0.3 0.4 0.5
E(V)
0.4
II
0.4
a
b
+ +
0.15 0.1
b
0.05 0
I(μA)
0.48
I(μA)
0.56
I(μA)
0.5
I(μA)
I I(μA)
(a) 0.64
Δ 0
Δ
+ +
+
+
9 10 11 12 13 14
-log([M13]/M)
0.1
c
0.16 0.05
0.08
b
0 0
0.1
0.2
a
a
0.3
0.4
0 0
0.5
E(V)
0.1
0.2
0.3
0.4
0.5
0.6
E(V)
Figure 10.13 (a) DPV showing the increase in the Faradaic current as the Fc-labeled replica are being synthesized by the polymerase reaction; (b) CVs of Fc-labeled DNA in the presence of glucose oxidase and in the absence (a) and presence (b) of glucose. A large catalytic current is observed only when glucose is present. Reprinted with permission from Ref. 94. Copyright 2002 American Chemical Society.
be incorporated into RNA oligomers by two different RNA polymerases. Increasing the Fc1-dUTP ratio produced more heavily labeled transcripts, as shown by gel electrophoresis and an increased peak area in square wave voltammetry (SWV) of the RNA transcripts. These Fc-labeled RNA oligomers were then immobilized on Au electrodes (SAM), and the system was successfully exploited for the electrochemical detection of RNA. An electrochemical “two-color” assay is proposed by using Fc1-dUTP-modified RNA oligomers along with anthraquinone-labeled UTP monomers.95 O HN O– –O
O–
O–
P O P O P O O O O
O O
Fc1-dUTP
HN –O
O–
P O P O P O O O O
2"
N
O
O–
1" Fe
OH H
O–
7 8 9 CH CH CH2 NH CO
12 7 8 9 CH CH CH2 NH CO CH2
1" 2" Fe
O O OH H
N Fc2-dUTP
Scheme 10.4 Structures of Fc1-dUTP and Fc-dUTP. By permission from Ref. 88 of Oxford University Press.
10.4
ELECTROACTIVE DNA AND PNA
285
Takenaka’s group90 reported a direct Fc-conjugation to a DNA sequence. Fccarbodiimides react with the imino unit of thymine or guanine base on DNA or uracil base on RNA under a basic buffer condition, which then hybridized to the complementary sequences immobilized on a gold electrode surface. SWV shows that the peak current was proportional to the amount of the target DNA. When DNA was labeled with two Fc-carbodiimides with different formal potentials, the resulting DNA gave rise to competitive electrochemical signals, suggesting that electrochemical gene expression analysis might be possible with suitable detection limits of 0.05 mM DNA sample in 1 mL (i.e., 50 fmol). More recently, Tanaka’s group applied the same method for the activity detection of deoxyribonuclease I (DNase I).96 Multiple Fc units were incorporated into a thiolated DNA sequence followed by immobilization onto a gold electrode surface. The modified electrode showed a current response related to the surface concentration of Fc moieties. Expectedly the redox response decreased after digestion of the DNA with DNase I.
10.4
ELECTROACTIVE DNA AND PNA
Currently, there is a need for high-throughput determination of nucleic acid sequences. At present, detection systems most commonly employ fluorescence-based methods. However, wide spread applications of such methods are limited by low speed, high cost, size, and number of incubations steps, among other factors. Application of electrochemical methods in affinity DNA sensors presents likely a promising alternative, allowing miniaturization and cost reduction, and potentially allowing application in point-of-care assays. In recent years, numerous electrochemical DNA detection and sensing methods have been described in the literature. Oxidation of nucleobases, while possible (see Section 10.1), is not desirable under normal circumstances since it will result in the formation of reactive species that ultimately will lead to DNA decomposition. For example, guanine can be electrochemically oxidized,99 but practical application of guanine oxidation as detection mode is limited to the use of G-free capture strands, high oxidation potentials and irreversibility of the oxidation process. In response to this limitation, electroactive DNA intercalators, enzymatic signal amplification schemes, or redox-modified oligonucleotides have been developed as described earlier. The incorporation of such electroactive centers, in particular Fc, into ODNs to form redox-active conjugates and its subsequent surface immobilization made it possible to study single-stranded ODNs, hybridization events and even allowed the detection of mismatched DNA, as well as specific interactions with molecules. Often, covalent attachment of an Fc group to oligonucleotides is achieved by imine and amide formation with Fc-carboxaldehyde or Fc-carboxylic acid and also by Sonogashira coupling.82b,100 As described above, the Fc-label can be introduced either at the monomer stage, by a metal-catalyzed reaction, or after assembly of the oligomer sequence.101 Ferrocenylated monomers (nucleotides) can be conviniently used as building blocks in automated oligonucleotide synthesis.102 Ihara and coworkers obtained the desired ODN conjugates by covalently linking an Fc group to the
286
FUNCTIONAL ELECTROACTIVE BIOMOLECULES
50 -amino-hexyl-terminated synthetic ODN.103 It was demonstrated that these electroactive probes can be detected at femtomole levels by using HPLC with an ordinary electrochemical detector.104 Furthermore, the sensitivity of the electrochemical detection was improved by PCR amplification with the use of an Fc ODN primer, by two orders of magnitude (0.1 fM).105 Unfortunately, the introduction of the Fc-label into a DNA oligomer can decrease duplex stability. The stability effect depends on the position and method of substitution as well as on particular nucleobase to which label is substituted. Experimental melting temperatures for different Fc oligomers were compared in a recent review by Metzler-Nolte.41a To design functional DNA biosensor Fc-ODNs have to be immobilized on an electrode surface. Thin film formation is often accomplished by self-assembly of thiol or disulfide containing ODN onto a gold surface driven by facile gold-thiol bond formation. For example, Mirkin studied Fc-ODN films by CV and showed that such films exhibit reversible redox behavior.106 Fc-conjugated ODN films have also been studied to gain insight into electron transfer in DNA. The differences in potential and standard electron transfer constant can be used do differentiate modes of electron transfer, such as intrastrand and interstrand crossing mechanism, between the electroactive probe and the electrode surface.76 Since knowledge of the DNA duplex flexibility is of key importance in understanding of number of biological processes, such as DNA-protein recognition,107 Anne conducted an electrochemical study of DNA polymer strands flexibility by using ferrocene-labeled oligonucletide chains tethered to a gold electrode.76 It was shown that surface immobilized single-stranded DNA (ss-DNA) adopts an unordered flexible coil conformation, allowing for reversible electron transfer kinetics from the Fc to the electrode surface (see Fig 10.14). However, with increasing scan rate, v, the proportionality of anodic current to v was lost, indicating that bound diffusion of the ODN on the surface is operational.108 Introduction of the complementary strand to form ds-DNA resulted
Figure 10.14 (a) Schematic representation of hybridization between ferrocene-conjugated oligonucleotide immobilized on gold electrode and its complementary strand. (b) Cyclic voltammograms recorded at v ¼ 200 V/s and dependence of the anodic peak current on scan rate, normalized versus v1/2. Reprinted with permission from Ref. 76. Copyright 2003 American Chemical Society.
10.4
ELECTROACTIVE DNA AND PNA
287
in a conformational transformation that induced changes in the mobility of the Fc moiety, thus obstructing the electron transfer and decreasing voltammetric response. Yu and coworkers reported an Fc-conjugated DNA based sensor for detection and mismatch discrimination.71 In this research, automatic DNA/RNA synthesis techniques were used to insert the Fc group into oligonucleotides at various positions. The melting behavior of the Fc-ODN indicated that incorporation of the Fc probe has virtually no affect on duplex stability. Electrochemical studies by alternating current voltammetry (ACV) showed that the Fc-ODN presents suitable probe for nucleic acids detection. Furthermore, studies performed on CMS-DNA array chips showed that dual-signaling oligonucleotide probes (containing two electroactive moieties) facilitated detection of single-nucleotide mismatches. A slightly different strategy was employed in another study. Inouye et al. designed an electrochemical DNA sensor for detection of single-nucleotide polymorphism.109–111 A p-conjugated Fc-modified nucleoside analog was connected at the 50 end of single-stranded oligonucleotide. After hybridization to the complementary strand, the 30 end of the probe DNA strand was attached to gold electrode by Au-thiol chemistry (see Fig 10.15). The probe appears to behave as a molecular wire thanks to apparent hole transport through Fc-p-conjugated DNA probes. Consequently, the electrochemistry of the Fc marker can be observed, allowing for highly sensitive detection of complementary DNA (cDNA). The presence of single-nucleotide mismatch in the duplex causes, presumably, a blockage of the p-conduction pathway through the base stack at the position of the base-pair
Figure 10.15 Electrochemical discrimination of single-nucleotide mismatch with Fc-ODN. (a) probe hybridized to its complementary strand, (b) probe hybridized to single-nucleotide mismatched strand, and center: chemical structure of the HS-DNA-Fc probe. Reprinted with permission from Ref. 109. Copyright 2005 National Academy of Sciences, USA.
288
FUNCTIONAL ELECTROACTIVE BIOMOLECULES
mismatch. This results in a dramatic reduction of the electrochemical response. In addition, a comparison of different DNA probes containing an isomeric Fc-diamidopyridine conjugate for electrochemical mismatch detection was carried out by the same authors in a separate study.110 It was concluded that despite different stereochemistries of the Fc-label, all conjugated DNA probes were capable of providing satisfactory electrochemical response for mismatch discrimination. A different route toward design of a robust DNA biosensor was taken by Mirkin and coworkers.79 This involved preparation of polymers to which DNA oligonucleotides and Fc-groups side chains were attached. These polymer-DNA hybrids were characterized by larger binding constants and sharper melting points compared to the ODN used for their preparation. This original approach allowed obtaining better mismatch selectivity and target discrimination compared to single-nucleotide probes. Another interesting study demonstrated electronic nanoswitch based on the two-state conformational transition between single stranded random coils and four-stranded intermolecular G-quadruplexes (Fig 10.16).112 It is well known that DNA sequences rich in guanine are able to fold into G-quadruplex structures, consisting of stacked guanine tetrads that are stabilized by Hoogsteen-type H-bonds between the tetrads.113,114 Therefore, thiol terminated ODN sequences containing a single G-rich stretch per sequence were immobilized on the gold surface and labeled with an Fc derivative. In the presence of potassium ions, these surface-immobilized G-rich strands can change their conformations forming a four-stranded G-quadruplex with bound K þ , as indicated in Fig 10.16. The DNA conformational changes induce extending-shrinking like motion, which in effect induces changes in the electrochemical properties of the system. It was shown that proposed nanoswitch device could be used for selective detection of K þ over a wide concentration range. Repeated regeneration of the switch is also possible due to high reversibility of conformational changes.
Figure 10.16 Schematic depiction of the electronic nanoswitch and electrochemical detection of K þ .12 Reprinted from Ref. 112. Copyright 2008 with permission from Elsevier.
10.4
ELECTROACTIVE DNA AND PNA
289
Figure 10.17 Schematic representation of the systematic evolution of ligands by exponential (SELEX) enrichment process. (See the color version of this figure in Color Plates Section.)
A more recent application of redox labeled ODNs is redox-active aptamers that exploit molecular recognition between the aptamer and a target analyte. Briefly, aptamers are functional nucleic acids that selectively bind to a variety of targets. Due to a well-defined three-dimensional structure, aptamers can achieve selectivity comparable to that of antibodies but are readily accessible taking advantage of wellknown nucleic acid chemistry, polymeric chain reaction and contemporary separation methods, followed by aptamer selection from random pools of nucleic acids (DNA or RNA) by in vitro selection process called “systematic evolution of ligands by exponential enrichment” (SELEX)115–117 schematically depicted in Fig 10.17. First, a random pool of nucleic acids (typically 1015–1016) is incubated with the desired target molecule, followed by separation of bound and unbound molecules. Next, bound ODNs are eluted and amplified by PRC. After first enrichment cycle, sequences with variable affinity toward the target are used as a library for another selection cycle. Typically, after repeating this cycle 6–15 times for a given target the final library is cloned and sequenced. The high target affinity of aptamers originates from their ability to fold upon binding with a suitable target molecule. This induces significant conformational changes that makes them high selectivity biosensors. Furthermore, functional groups can be introduced during the aptamer synthesis at precise position for labeling or immobilization purposes. Due to these unique properties, aptamers rapidly found a number of applications in both fundamental research and biomedical diagnostics. It is not surprising that wide range of aptamer-based sensors (aptasensors) have been reported in the literature.118–120 Recently, electrochemical aptamer sensors gained considerable attention.121,122 Since, nucleic acids aptamers fold their structure upon binding to the target molecule, formation of the aptamer–target complex can be
290
FUNCTIONAL ELECTROACTIVE BIOMOLECULES
Figure 10.18
Detection schemes for electrochemical aptasensors.
detected electrochemically if the aptamer is labeled with redox-active moiety and immobilized on a conductive substrate. The ease of access of the redox reporter to the interfacial potential developed at the electrode depends on the conformation of the aptamer before and after target binding. Based on this strategy, several following scenarios are possible: (i) The target binding results in shielding of the redox probe from the electrode surface (see Fig 10.18a), thus impeding the electron transfer between the probe and electrode, signal-OFF assay. Based on this principle, an electrochemical thrombin aptasensor was proposed by Xiao et al.122 A methylene blue (MB) labeled aptamer was immobilized on an electrode surface. In unbound state, the flexible conformation of the aptamer oligonucleotide enabled electrical contact of the MB with the electrode, consequently a voltammetric response could be measured. In the presence of thrombin, the aptamer self-assembled into a G-quadruplex structures as a result of target binding. Formation of the Gquadruplex shielded the MB from the electrode surface, thus reducing the voltammetric signal and allowing for electrochemical detecting of thrombin with detection limit of approximately 20 nM. However, this detection scheme can be considered disadvantageous because the signal decreases upon target binding. To counter this problem, several signal-ON aptamer sensors were proposed. (ii) Signal-ON detection scheme relays on signal increase caused by the binding event as target binding causes aptamer to fold in such a way that electron transfer between redox label and an electrode becomes easier than in unbound state (Fig 10.18b).123–125 For instance, Radi and coworkers employed a bifunctionalized thrombin binding aptamer with a terminal Fc group and
10.4
ELECTROACTIVE DNA AND PNA
291
thiol group at the other end of the aptamer strand.123 Prior to target binding, the long and flexible aptamer prevented the electron transfer between the electrode and the redox-active Fc tag. The binding of the thrombin caused the modified aptamer chain to assemble into a rigid G-quadruplex. The Gquadruplex formation resulted in orientation of Fc group close to the electrode surface. This results in a positive amperometric signal due to the electron transfer between the electrode and redox label. This allowed for detection limit of approximately 5 nM of thrombin. (iii) In the cases described above, the flexible conformation of the aptamer chain will influence the electrochemical signal and can effectively result in a poor signal-to-noise ratio. This issue is addressed by the application of a duplexbased aptasensors. Since duplexes are more rigid structures than single strands this should overcome diffusion problems associated with highly flexible surface bound systems. For instance, a DNA-duplex consisting of Fc-labeled adenosine-50 -triphosphate (ATP) binding aptamer and its complementary strand was immobilized on the electrode surface (Fig 10.18c).126 Upon aptamer–target biding, the complementary strand was dissociated from the duplex and the aptamer chain formed a rigid structure that brought the Fc group into proximity of the electrode and allowed for facile electron transfer between the electrode and Fc group. The duplex-aptasensor allowed for nanomolar detection of ATP. A related approach employing surface bound DNA duplexes was proposed by Xiao and coworkers.127 In this case, two double strands were linked with noncomplementary nucleic acid bridge. The upper domain consisted of thrombin aptamer and its complementary MBconjugated strand. In the presence of the thrombin, the target binding duplex was liberated from its complementary strand and formed G-quadruplex. Consequently, the liberated MB-labeled strand folded in such a way that MB-based electrochemistry was observed. This method allowed the amperometric detection of thrombin with detection limit of about 3 nM. (iv) The three aptasensing scenarios described above employed surface bound aptamers as probes. A different general approach was recently proposed by Lu et al.128 In this case, oligonucleotides complementary to the aptamer are used (cDNA) (see Fig 10.18d). The cDNA probes are tailored to have complementary sequences at both ends. Also, redox species and thiol group are incorporated at both ends of cDNA for labeling and immobilization. The cDNA probes are hybridized with their aptamers to form double strands and immobilized on the gold electrode surface. Before target binding, the redox label is shielded from the electrode surface by the double strand, thus impeding the electron transfer. In the presence of target analyte the aptamer strands are liberated from their complementary strands and bound into target–aptamer complex. Since, the surface confined functionalized strands contain complementary sequences at both ends, they form stable hairpin structures. Formation of the hairpin brings the redox marker close to the electrode surface allowing for electrochemical detection of the target. The cDNA aptasensor performance was demonstrated by the authors for detection of thrombin and ATP.128
292
FUNCTIONAL ELECTROACTIVE BIOMOLECULES
Despite being relatively new technology, aptamers have a tremendous potential and can be envisioned to rival antibodies and other traditional recognition elements for molecular detection and recognition, due to their inherent affinity, selectivity, and ease of synthesis. In addition, the combination of aptasensors with electrochemical detection methods has the added advantage of further cost reduction and miniaturization of such systems. Peptide nucleic acid (PNA) is an artificial DNA analog in which the ribose phosphate ester backbone is replaced by pseudopeptide backbone.129 The various nucleobases are linked to the backbone by methylene carbonyl bonds. The PNA undergoes sequence-selective binding to RNA and DNA.130 Since the backbone of PNA contains no charged phosphate groups, there is no electrostatic repulsion between the backbones and the binding between PNA/DNA and PNA/PNA strands is stronger than between ordinary DNA/DNA strands. Also, the stability of PNA– DNA duplexes is almost unaffected by the ionic strength of the medium.131 These unique properties make PNA an interesting, although sometimes synthetically challenging and expensive, alternative for design of future biosensors. PNA-based biosensors promise better discrimination of single-nucleotide mismatches, higher specificity, and shorter hybridization times.132 Before electrochemical PNA sensor can be constructed conjugation of electroactive moiety to PNA is necessary. The preparation of Fc-PNA complexes was pioneered by Metzler-Nolte and coworkers.133,134 For example, Fc-labeled PNA was obtained by the reaction of Fccarboxylic acid with terminal amino group of a PNA heptamer.134 Another group used different strategy to obtain Fc-labeled PNA oligomers by inserting an Fc-labeled monomer into the PNA chain.135 Maiorana’s group also carried out detailed electrochemical characterization of the Fc-PNA monomers and dimers in various solvents that showed that the voltammetric response shifted toward positive potential in respect to Fc-methanol.135,136 The electrochemical signal intensity can be readily increased employing tris-Fc-derivatives.137 Metallocene-labeled PNA can be readily immobilized on gold surfaces using similar strategies developed for the formation of ODN films on gold.138 Briefly, a 10mer PNA was synthesized using standard solid-phase techniques. After synthesis of the oligomer, the last protecting group was removed and activated Fc-carboxylic acid or cobaltocenium carboxylic acid was added. Electrochemical characterization of the resulting metallocene PNA oligomers showed a quasi-reversible redox behavior in solution. Hybridization of the metallocene-PNAwith complementary ODN, followed by immobilization on gold electrodes, allowed the study of its redox properties. While the Fc-conjugates gave a satisfactory electrochemical signal, no signal was observed for cobaltocenium-PNA, presumably due to the high flexibility of the Ahx linker and the unfavorable and highly reducing redox conditions for the cobaltocinium probe. Li and coworkers communicated an interesting example of Fc-azeobenzene-conjugated PNA monomers.139 This original approach produced Fc-azobenzene labeled PNA monomer of thymine (Fc-Azo-T) that exhibit electrochemical and photochemical activities, allowing combined electrochemical and photophysical studies. Upon irradiation of Fc-Azo-T, the system undergoes an irradiation induced cis-trans izomerization that affects its electrochemical behavior.
10.4
ELECTROACTIVE DNA AND PNA
293
Figure 10.19 Schematic representation of the operational principle of the DNA sensor based on conformational flexibility change in the PNA probe structure stimulated by hybridization. (a) before hybridization, electron transfer between Fc and electrode is possible and (b) after hybridization, formation of the duplex rigidifies the probe structure, preventing efficient electron transfer. Reproduced by permission from Ref. 140 of The Royal Society of Chemistry.
Detection of DNA using Fc-PNA sensor based on signal-OFF principle was reported by Aoki and Tao (see Fig 10.19).140 The proposed sensor exploits the change in structural flexibility of the PNA chain stimulated by a hybridization event. A probe was designed that consisted of the PNA strand with an Fc and cysteine group at both ends of the strand for redox labeling and immobilization, respectively. Voltammetric experiments performed after immobilization of the probe on Au electrodes showed reversible oxidation behavior, as the electrode surface was easily accessible for the redox label due to the flexibility of the PNA chain. Upon hybridization of the complementary target DNA strand the flexibility of the probe decreased, thus impeding the electron transfer between Fc-label and electrode and decreasing observed current. Detection limit of approximately 14 pM was achieved. Also, the sensor could be used repeatedly by denaturation of the hybrid. The potential of the sensor for mismatch discrimination was also demonstrated. An interesting PNA based DNA electrochemical sensor was recently constructed by Reisberg et al.141 In this innovative approach instead of conjugating a metallocene to the PNA, the PNA chains were covalently attached onto an electrode coated with a quinine-based conductive polymer. The electrochemical behavior of the quinine group is highly sensitive to its chemical environment, such as ionic strength of the electrolyte solution. Since ODNs carry a high anionic charge density, they are able to affect the chemical environment of quinine groups upon hybridization to the probe
294
FUNCTIONAL ELECTROACTIVE BIOMOLECULES
PNA strand, resulting in an increase of the electrochemical signal (SWV peak current) as a result of probe–target hybridization. The PNA-functionalized conductive polymer sensor allowed for a detection limit of approximately 10 nM, and the feasibility for single-nucleotide mismatch detection was also demonstrated. The advantageous properties of PNA-based biosensors represent an appealing alternative to DNA and aptamer based sensors under some circumstances. They offer advantages of increased sensitivity, selectivity, and stability, therefore they present valuable alternative to DNA/aptamer based sensors but have problems associated with solubility and synthetic procedure.
CONCLUSIONS In recent years, organometallic bioconjugates ranging from conjugated amino acids to peptides and nucleic acids have attracted significant attention. The structural properties displayed by the bioconjugates are interesting in that they provide a high degree of control over the degree of supramolecular assembly, which is often critical for biorecognition events. The biocomponent of the conjugate allows the design of specific interfaces for molecular recognition events making them interesting targets with potential bioanalytical applications. Fc-bioconjugates are conveniently synthesized by well-established solution or solid-phase methods. In general, the synthetic approach is highly adaptive to a particular need and enables the incorporation of a range of binding sites. Similarly, Fc-nucleobases can be incorporated into the ODN sequence either by automated solid-phase synthesis or by a PCR-based process. The resulting labeled ODN conjugates display a high degree of selectivity for the interaction with a particular complementary strand and a number of bioanalytical assay methods are now reaching maturity for a range of applications (identification of species, mismatch identification, and phenotyping). Clearly bioorganometallic conjugates hold a great promise in this area that will ultimately be fully realized when coupled with miniaturization and microfluidics leading to significant cost reduction. Ultimately, handheld devices are feasible that might find entry in point-of-care applications.
REFERENCES 1. E. Palecek, Naturwissenschaften 1958, 45, 186. 2. S. Steenken, S. V. Jovanovic, J. Am. Chem. Soc. 1997, 119, 617. 3. E. Palecek, F. Jelen, Electrochemistry of nucleic acids, in E. Palecek, F. Scheller, J. Wang (Eds.), Electrochemistry of nucleic acids and proteins. Towards electrochemical sensors for genomics and proteomics, Elsevier, Amsterdam, 2005, p. 74. 4. J. Wang, S. Bollo, J. L. L. Paz, E. Sahlin, B. Mukherjee, Anal. Chem. 1999, 71, 1910. 5. (a) A. M. Oliveira-Brett, V. Diculescu, J. A. P. Piedade, Bioelectrochem. 2002, 55, 61. (b) A. M. Oliveira-Brett, J. A. P. Piedade, L. A. Silva, V. C. Diculescu, Anal. Biochem. 2004, 332, 321.
REFERENCES
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.
295
A. M. Oliveira-Brett, J. A. P. Piedade, S. H. P. Serrano, Electroanalysis 2000, 12, 969. R. N. Goyal, G. Dryhurst, J. Electroanal. Chem. 1982, 135, 75. P. Subramanian, G. Dryhurst, J. Electroanal. Chem. 1987, 244, 137. G. Dryhurst, J. Electroanal. Soc. 1969, 116, 1411. H. Xie, D. Yang, A. Heller, Z. Gao, Biohys. J. 2007, 92, L70. A. Apilux, M. Tabata, O. Chailapakul, Bioelectrochemistry 2007, 70, 435. K. Kerman, B. Meric, D. Ozkan, P. Kara, A. Erdem, M. Ozsoz, Anal. Chim. Acta 2001, 450, 45. G. Chiti, G. Marrazza, M. Mascini, Anal. Chim. Acta 2001, 427, 155. B. Meric, K. Kerman, D. Ozkan, P. Kara, A. Erdem, O. Kucukoglu, E. Erciyas, M. Ozsoz, J. Pharm. Biomed. Anal. 2002, 30, 1339. A. M. Tencaliac, S. Laschi, V. Magearu, M. Mascini, Talanta 2006, 69, 365. A. Abbaspour, M. A. Mehrgardi, Anal. Chem. 2004, 76, 5690. J. E. Koehne, H. Chen, A. M. Cassell, Q. Ye, J. Han, M. Meyyappan, J. Li, Clin. Chem. 2004, 50, 1886. R. Brdicka, Collect. Czech. Chem. Commun. 1933, 5, 112. M. Heyrovsky, Electroanalysis 2004, 16, 1067. M. Tomschik, L. Havran, E. Palecek, M. Heyrovsky, Electroanalysis 2000, 12, 274. V. Brabec, V. Mornstein, Biochim. Biophys. Acta 1980, 625, 43. X. Cai, G. Rivas, M. A. P. Farias, H. Shiraishi, J. Wang, E. Palecek, Anal. Chim. Acta 1996, 332, 49. S. M. MacDonald, S. G. Roscoe, Electrochim. Acta 1997, 42, 1189. C. F. Zinola, J. L. Rodriguez, M. C. Arevalo, E. Pastor, J. Electroanal. Chem. 2005, 585, 230. K. Ogura, M. Kobayashi, M. Nakayama, Y. Miho, J. Electroanal. Chem. 1999, 463, 218. A. K. Vijh, B. E. Conway, Chem. Rev. 1967, 67, 623. M. Fojta, S. Billova, L. Havran, H. Pivonkova, H. Cernocka, P. Horakova, E. Palecek, Anal. Chem. 2008, 80, 4598. M. E. Johll, D. G. Williams, D. C. Johnson, Electroanalysis 1997, 9, 1397. J. A. Reynaud, B. Malfoy, P. Canesson, J. Electroanal. Chem. 1980, 114, 195. R. Possari, R. F. Carvalhal, R. K. Mendes, L. T. Kubota, Anal. Chim. Acta 2006, 575, 172. H. Q. Li, A. Chen, S. G. Roscoe, J. Lipkowski, J. Electroanal. Chem. 2001, 500, 299. J. E. I. Wright, K. Fatih, C. L. Brosseau, S. Omanovic, S. G. Roscoe, J. Electroanal. Chem. 2003, 550–551, 41. C. L. Brosseau, S. G. Roscoe, Electrochim. Acta 2006, 51, 2145. J. Kim, K. C. Chou, G. A. Somorjai, J. Phys. Chem. B 2002, 106, 9198. D. G. Marangoni, R. S. Smith, S. G. Roscoe, Can. J. Chem. 1989, 67, 921. R. J. Petrie, S. M. MacDonald, K. L. Fuller, S. G. Roscoe, Can. J. Chem. 1997, 75, 1585. F. A. Armstrong, Curr. Opin. Chem. Biol. 2005, 9, 110. J. Wang, J. W. Mo, A. Erdem, Electroanalysis 2002, 14, 1365. X. Cai, G. Rivas, P. A. M. Farias, H. Shiraishi, J. Wang, E. Palecek, Anal. Chim. Acta 1996, 332, 49.
296
FUNCTIONAL ELECTROACTIVE BIOMOLECULES
40. L. Barisic, M. Cakic, K. A. Mahmoud, Y.-N. Liu, H.-B. Kraatz, H. Pritzkow, N. MetzlerNolte, S. I. Kirin, V. Rapic, Chem. Eur. J. 2006, 12, 4965. 41. (a) D. R. van Staveren, N. Metzler-Nolte, Chem. Rev. 2004, 104, 5931. (b) H.-B. Kraatz, J. Inorg. Organomet. Polym. Mater. 2005, 15, 83. 42. K. Severin, R. Bergs, W. Beck, Angew. Chem., Int. Ed. 1998, 37, 1634. 43. R. S. Herrick, R. M. Jarret, T. P. Curran, D. R. Dragoli, M. B. Flaherty, S. E. Lindyberg, R. A. Slate, L. C. Thornton, Tetrahedron Lett. 1996, 37, 5289. 44. A. Nomoto, T. Moriuchi, S. Yamazaki, A. Ogawa, T. Hirao, Chem. Commun. 1998, 1963. 45. Y. Degani, A. Heller, J. Phys. Chem. 1987, 91, 1285. 46. Y. Degani, A. Heller, J. Am. Chem. Soc. 1988, 110, 2615. 47. S. Chowdhury, G. Schatte, H.-B. Kraatz, Dalton Trans. 2004, 1726. 48. S. Chowdhury, G. Schatte, H.-B. Kraatz, Eur. J. Inorg. Chem. 2006, 5, 988. 49. S. Chowdhury, D. A. R. Sanders, G. Schatte, H.-B. Kraatz, Angew. Chem., Int. Ed. 2006, 45, 751. 50. S. Chowdhury, K. A. Mahmoud, G. Schatte, H.-B. Kraatz, Org. Biomol. Chem. 2005, 3, 3018. 51. S. Chowdhury, G. Schatte, H.-B. Kraatz, Angew. Chem., Int. Ed. 2006, 45, 6882. 52. C. M. Cardona, A. E. Kaifer, J. Am. Chem. Soc. 1998, 120, 4023. 53. C. M. Cardona, S. Mendoza, A. E. Kaifer, Chem. Soc. Rev. 2000, 29, 37. 54. C. M. Cardona, T. D. McCarley, A. E. Kaifer, J. Org. Chem. 2000, 65, 1857. 55. D. L. Stone, D. K. Smith, P. T. McGrail, J. Am. Chem. Soc. 2002, 124, 856. 56. K. N. Jayakumar, P. Bharathi, S. Thayumanavan, Org. Lett. 2004, 6, 2547. 57. K. N. Jayakumar, S. Thayumanavan, Tetrahedron 2005, 61, 603. 58. C. S. Camaron, C. B. Gorman, Adv. Funct. Mater. 2002, 12, 17. 59. T. L. Chasse, C. B. Gorman, Langmuir 2004, 20, 8792. 60. Y. Wang, C. M. Candona, A. E. Kaifer, J. Am. Chem. Soc. 1999, 121, 9756. 61. M. Wells, R. M. Crooks, J. Am. Chem. Soc. 1996, 118, 3988. 62. Y. Zhou, M. L. Bruening, D. E. Bergbreiter, R. M. Crooks, M. Wells J. Am. Chem. Soc. 1996, 118, 3773. 63. F. E. Appoh, D. S. Thomas, H.-B. Kraatz, Macromolecules 2005, 38, 7562. 64. F. E. Appoh, D. S. Thomas, H.-B. Kraatz, Macromolecules 2006, 39, 5629. 65. F. E. Appoh, Y.-T. Long, H.-B. Kraatz, Langmuir 2006, 22, 10515. 66. K. A. Mahmoud, H.-B. Kraatz, Chem. Eur. J. 2007, 13, 5885. 67. K. Kerman, K. A. Mahmoud, H.-B. Kraatz, Chem. Commun. 2007, 3829. 68. K. A. Mahmoud, S. Hrapovic, J. H. Luong, ACS Nano 2008, 2, 1051. 69. (a) K. Heinze, M. Beckmann, J. Organometal. Chem. 2006, 691, 5588–5596. (b) J. Lapic, D. Siebler, K. Heinze, V. Rapic, Eur. J. Inorg. Chem. 2007, 2014. (c) J. Lapic, G. Pavlovic, D. Siebler, K. Heinze, V. Rapic, Organometallics 2007, 27, 726. (d) S. Djakovic, D. Siebler, M. Cakic, K. Semencic, K. Heinze, V. Rapic, Organometallics 2008, 27, 1447. 70. T. Ihara, Y. Maruo, S. Takenaka, M. Takagi, Nucleic Acids Res 1996, 24, 4273. 71. C. J. Yu, Y. Wan, H. Yowanto, J. Li, C. Tao, M. D. James, C. L. Tan, G. F. Blackburn, T. J. Meade, J. Am. Chem. Soc. 2001, 123, 11155. 72. C. Fan, K. W. Plaxco, A. J. Heeger, Proc. Natl. Acad. Sci. USA 2003, 100, 9134.
REFERENCES
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.
297
J. Wang, J. Li, A. J. Baca, J. Hu, F. Zhou, W. Yan, D.-W. Pang, Anal. Chem. 2003, 75, 3941. C. E. Immoos, S. J. Lee, M. W. Grinstaff, ChemBioChem 2004, 5, 1100. H. H. Thorp, Trends Biotechnol 2003, 21, 522. A. Anne, A. Bouchardon, J. Moiroux, J. Am. Chem. Soc. 2003, 125, 1112. C. E. Immoos, S. J. Lee, M. W. Grinstaff, J. Am. Chem. Soc. 2004, 126, 10814. K. Kim, H. Yang, S. H. Park, D.-S. Lee, S.-J. Kim, Y. T. Lim, Y. T. Kim, Chem. Commun. 2004, 1466. J. M. Gibbs, S.-J. Park, D. R. Anderson, K. J. Watson, C. A. Mirkin, S. T. Nguyen, J. Am. Chem. Soc. 2005, 127, 1170. Y.-T. Long, C.-Z. Li, T. C. Sutherland, M. Chahma, J. S. Lee, H.-B. Kraatz, J. Am. Chem. Soc. 2003, 125, 8724. T. S. Zatsepin, S. Y. Andreev, T. Hianik, T. S. Oretskaya, Russ. Chem. Rev. 2003, 72, 537. (a) P. Meunier, I. Ouattara, B. Gautheron, J. Tirouflet, D. Camboli, J. Besancon, Eur. J. Med. Chem. 1991, 26, 351. (b) K. Sonogashira, Y. Tohda, N. Hagihara, Tetrahedron Lett. 1975, 4467. C. J. Yu, H. Yowanto, Y. Wan, T. J. Meade, Y. Chong, M. Strong, L. H. Donilon, J. F. Kayyem, M. Gozin, G. F. Blackburn, J. Am. Chem. Soc. 2000, 122, 6767. A. R. Pike, L. C. Ryder, B. R. Horrocks, W. Clegg, M. R. J. Elsegood, B. A. Connolly, A. Houlton, Chem. Eur. J. 2002, 8, 2891. A. Houlton, C. J. Isaac, A. E. Gibson, B. R. Horrocks, W. Clegg, M. R. J. Elsegood, J. Chem. Soc., Dalton Trans. 1999, 3229. A. E. Beilstein, M. W. Grinstaff, Chem. Commun. 2000, 509. A. E. Beilstein, M. W. Grinstaff, J. Organomet. Chem. 2001, 637–639, 398. W. A. Wlassoff, G. C. King, Nucleic Acids Res. 2002, 30, e58/1. D. A. Di Giusto, W. A. Wlassoff, S. Giesebrecht, J. J. Gooding, G. C. King, J. Am. Chem. Soc. 2004, 126, 4120. K. Mukumoto, T. Nojima, S. Takenaka, Tetrahedron 2005, 61, 11705. H. Song, X. Li, Y. Long, G. Schatte, H.-B. Kraatz, Dalton Trans. 2006, 4696. A. N. Patwa, S. Gupta, R. G. Gonnade, V. A. Kumar, M. M. Bhadbhade, K. N. Ganesh, J. Org. Chem. 2008, 73, 1508. M. D. Morris, Electroanal. Chem. 1974, 7, 79. F. Patolsky, Y. Weizmann, I. Willner, J. Am. Chem. Soc. 2002, 124, 770. D. A. Di Giusto, W. A. Wlassoff, S. Giesebrecht, J. J. Gooding, G. C. King, Angew. Chem., Int. Ed. 2004, 43, 2809. S. Sato, K. Fujita, M. Kanazawa, K. Mukumoto, K. Ohtsuka, M. Waki, S. Takenaka, Anal Biochem, in press. M. T. Carter, M. Rodriguez, A. J. Bard, J. Am. Chem. Soc. 1989, 111, 8901. M. Hocek, P. Stepnicka, J. Ludvik, I. Cisarova, I. Votruba, D. Reha, P. Hobza, Chem. Eur. J. 2004, 10, 2058. J. Wang, G. Rivas, J. R. Fernandes, J. L. Lopez Paz, M. Jiang, R. Waymire, Anal. Chim. Acta 1998, 375, 197. K. E. Dombrowski, W. Baldwin, J. E. Sheats, J. Organomet. Chem. 1986, 302, 281. A. Okamoto, K. Tainaka, I. Saito, Tetrahedron Lett. 2002, 43, 4581.
298
FUNCTIONAL ELECTROACTIVE BIOMOLECULES
102. E. Bucci, L. De Napoli, G. Di Fabio, A. Messere, D. Montesarchio, A. Romanelli, G. Piccialli, M. Varra, Tetrahedron 1999, 55, 14435. 103. T. Ihara, Y. Maruo, S. Takenaka, M. Takagi, Nucleic Acids Res. 1996, 24, 4273. 104. S. Takenaka, Y. Uto, H. Kondo, T. Ihara, M. Takagi, Anal. Biochem. 1994, 218, 436. 105. Y. Uto, H. Kondo, M. Abe, T. Suzuki, S. Takenaka, Anal. Biochem. 1997, 250, 122. 106. R. C. Mucic, M. K. Herrlein, C. A. Mirkin, R. L. Letsinger, Chem. Commun. 1996, 555. 107. M. E. Hogan, R. H. Austin, Nature 1987, 329, 263. 108. A. Anne, C. Demaille, J. Moiroux, Macromolecules 2002, 35, 5578. 109. M. Inouye, R. Ikeda, M. Takase, T. Tsuri, J. Chiba, Proc. Natl. Acad. Sci. USA 2005, 102, 11606. 110. R. Ikeda, J. Chiba, M. Inouye, e-J. Surf. Sci. Nanotechnol. 2005, 3, 393. 111. R. Ikeda, A. Akaishi, J. Chiba, M. Inouye, ChemBioChem 2007, 8, 2219. 112. Z.-S. Wu, C.-R. Chen, G.-L. Shen, R.-Q. Yu, Biomaterials 2008, 29, 2689. 113. N. Nagesh, D. Chatterji, J. Biochem. Biophys. Methods 1995, 30, 1. 114. F. M. Chen, Biochemistry 1992, 31, 3769. 115. C. Tuerk, L. Gold, Science 1990, 249, 505. 116. A. D. Ellington, J. Szostak, Nature 1990, 346, 818. 117. D. S. Wilson, J. W. Szostak, Annu. Rev. Biochem. 1999, 68, 611. 118. A. Warsinke, B. Nagel, Anal. Lett. 2006, 39, 2507. 119. A. Warsinke, Adv Biochem Engin/Biotechnol 2008, 109, 155. 120. N. de-los-Santos-Alvarez, M. Jesus Lobo-Castanon, A. J. Miranda-Ordieres, P. TunonBlanco, Trends Anal. Chem. 2008, 27. 121. I. Willner, M. Zayats, Angew. Chem., Int. Ed. 2007, 46, 6408. 122. Y. Xiao, A. A. Lubin, A. J. Heeger, K. W. Plaxco, Angew. Chem., Int. Ed. 2005, 44, 5456. 123. A.-E. Radi, J. L. Acero Sanchez, E. Baldrich, C. K. O’Sullivan, J. Am. Chem. Soc. 2006, 128, 117. 124. M. Mir, I. Katakis, Mol. Biosyst. 2007, 3, 620. 125. R. Y. Lai, K. W. Plaxco, A. J. Heeger, Anal. Chem. 2007, 79, 229. 126. X. Zuo, S. Song, J. Zhang, D. Pan, L. Wang, C. Fan, J. Am. Chem. Soc. 2007, 129, 1042. 127. Y. Xiao, B. D. Piorek, K. W. Plaxco, A. J. Heeger, J. Am. Chem. Soc. 2005, 127, 17990. 128. Y. Lu, X. Li, L. Zhang, P. Yu, L. Su, L. Mao, Anal. Chem. 2008, 80, 1883. 129. P. E. Nielsen, M. Egholm, R. H. Berg, O. Buchardt, Science 1991, 254, 1497. 130. E. Uhlmann, A. Peyman, G. Breipohl, W. W. David, Angew. Chem., Int. Ed. 1998, 37, 2796. 131. S. Tomac, M. Sarkar, T. Ratilainen, P. Wittung, P. E. Nielsen, B. Norden, A. Graslund, J. Am. Chem. Soc. 1996, 118, 5544. 132. J. Wang, Bioesens Bioelectron 1998, 13, 757. 133. A. Hess, N. Metzler-Nolte, J. Chem. Soc., Chem. Commun. 1999, 885. 134. J. C. Verheijen, G. A. van der Marel, J. H. van Boom, N. Metzler-Nolte, Bioconjugate Chem. 2000, 11, 741. 135. C. Baldoli, L. Falciola, E. Licandro, S. Maiorana, P. Mussini, P. Ramani, C. Rigamonti, G. Zinzalla, J. Organomet. Chem. 2004, 689, 4791.
REFERENCES
299
136. C. Baldoli, E. Licandro, S. Maiorana, D. Resemini, C. Rigamonti, L. Falciola, M. Longhi, P. R. Mussini, J. Electroanal. Chem. 2005, 585, 197. 137. C. Baldoli, C. Rigamonti, S. Maiorana, E. Licandro, L. Falciola, P. R. Mussini, Chem. Eur. J. 2006, 12, 4091. 138. A. Maurer, H.-B. Kraatz, N. Metzler-Nolte, Eur. J. Inorg. Chem. 2005, 3207. 139. J. Li, M. Chen, H. Zhang, S. Liu, Inorg. Chem. Commun. 2008, 11, 392. 140. H. Aoki, H. Tao, Analyst 2007, 132, 784. 141. S. Reisberg, L. A. Dang, Q. A. Nguyen, B. Piro, V. Noel, P. E. Nielsen, L. A. Leb, M. C. Pham, Talanta 2008, 76, 206.
CHAPTER 11
Functional Nanoparticles as Catalysts and Sensors BRIAN J. JORDAN, CHANDRAMOULEESWARAN SUBRAMANI, and VINCENT M. ROTELLO Department of Chemistry, University of Massachusetts, Amherst, MA, USA
Nanoparticles (NPs) provide important building blocks for the creation of functional materials due to their unique chemical, physical, and electronic properties.1 Their small size (typically 1–100 nm), variety of core materials, and tunable surface properties are advantageous for the creation of highly efficient catalysts and sensors. In addition to the properties conveyed by the core material, the NP surface can be covered with a variety of chemical moieties, including molecular receptors that bind analytes and generate a detectable signal for sensing applications. In addition, their tunable monolayers enable the NPs to be used as supramolecular building blocks to construct nanocomposite assemblies. The nanocomposite assemblies can be incorporated into electrochemical and biosensor devices, such as enzyme, immuno-, and DNA sensors, to improve sensitivity, signal transduction, and catalysis.2 NPs are typically composed of a metallic or semiconductor core stabilized by a self-assembled organic shell.3Monolayer protected clusters (MPCs) use the organic self-assembled monolayer to prevent aggregation of the cores and generate a scaffold to append functional molecular units at the nanoparticle periphery. Mixed monolayer protected clusters (MMPCs) allow for multiple functionalities to be appended to the NP surface. The resultant multifunctional MMPC monolayer can impart a variety of properties ranging from solubility to the introduction of redox-active or specific molecular recognition elements that enhance the versatility of these core–shell systems.4 Brust et al. have developed methods to fabricate MPC and MMPC systems that rely on the reduction of metal salts (Pd, Au, Ag, Pt) in the presence of capping ligands.5 The approach uses mild conditions and moderate reducing agents that are compatible
Electrochemistry of Functional Supramolecular Systems. Edited by Paola Ceroni, Alberto Credi, and Margherita Venturi Copyright 2010 John Wiley & Sons, Inc.
301
302
FUNCTIONAL NANOPARTICLES AS CATALYSTS AND SENSORS
with a variety of ligand functionality. In this method, the ratio of metal salt to the capping ligand controls the overall size of the NPs, yielding MPCs in the range of 1.5–11 nm in diameter.6 Multifunctional MMPCs can be prepared via place exchange reactions where initial capping ligands are replaced by different ligands.7 The ligand displacement can be repeated multiple times providing a method to produce both structurally diverse and tunable MMPCs. Many types of NPs are currently used for catalytic and sensing applications. Metallic NPs act as catalysts to increase electrochemical reactions and as “electronic wires” to enhance electron transfer between redox centers in proteins and electrode surfaces.8 Semiconductor NPs are commonly used as labels or tags for biosensors due to their inherent electrochemical9 and optical (usually fluorescent) properties.10 Oxide NPs are employed to immobilize biomolecules due to their relative biocompatibility.11 Polymeric materials, such as dendrimers12 or polystyrene beads,13 are incorporated into sensors, catalysts, and drug delivery agents due to their high density of active groups and good biocompatibility. Thus, NPs perform distinct roles within biosensors and catalysts based upon their unique chemical and physical properties. Several reviews have described the roles of NPs in catalytic and sensing applications, but only few have systematically addressed the functions of the NPs in these systems.14 It is the goal of this chapter to outline the specific electrochemical roles that NPs play as functional materials within catalysts and sensors. For sensors, we separate the NP roles into two broad categories based upon the location of their electrochemical signal or response: exo-active and core based. Exo-active surfaces describe NPs that generate an electrochemical response that occurs at the ligand–solution interface at the periphery of the NP (Fig. 11.1a). Exo-active surfaces are widely used for sensing applications due to the large number of molecular receptors and their accessibility to target molecules. Core-based materials differ from exo-active surfaces as they produce a signal or response that occurs predominantly at the NP core (Fig. 11.1b). Core-based materials typically can be used for sensing applications as they provide a suitable
Figure 11.1 Categories for functional NPs based upon location of electrochemical response. (a) Exo-active surfaces. (b) Core-based materials.
11.1
EXO-ACTIVE NANOPARTICLE SURFACES IN SENSING
303
electrochemical label to identify analytes and supply efficient signal transduction through activation of their core by biological events. For catalysts, the roles of NPs can be classified by their end-use function. Electrochemical applications of catalytic NPs focus on two main functions: enhancing an electrochemical reaction or assisting electron transfer at an electrode surface. Electrochemical reactions can occur at the NP core or the periphery, depending upon the design of the NP. However, electron transfer mechanisms can proceed through the core due to the conductivity of metallic NPs. While the variety of NPs used in catalytic and sensor applications is extensive, this chapter will primarily focus on metallic and semiconductor NPs. The term functional nanoparticle will refer to a nanoparticle that interacts with a complementary molecule and facilitate an electrochemical process, integrating supramolecular and redox function. The chapter will first concentrate on the role of exo-active surfaces and core-based materials within sensor applications. Exo-active surfaces will be evaluated based upon their types of molecular receptors, ability to incorporate multiple chemical functionalities, selectivity toward distinct analytes, versatility as nanoscale receptors, and ability to modify electrodes via nanocomposite assemblies. Core-based materials will focus on electrochemical labeling and tagging methods for biosensor applications, as well as biological processes that generate an electrochemical response at their core. Finally, this chapter will shift its focus toward the catalytic nature of NPs, discussing electrochemical reactions and enhancement in electron transfer.
11.1
EXO-ACTIVE NANOPARTICLE SURFACES IN SENSING
MMPCs are well suited for the generation of exo-active surfaces due to their facile fabrication techniques and high surface area. A variety of terminal functional groups can be appended to the NP core and used as molecular receptors to sense particular analytes through specific noncovalent interactions, such as hydrogen bonding, electrostatics, or p–p stacking. Thus, the NP core is used as a scaffold to decorate the periphery of the NP surface with molecular receptors. The binding event between the molecular receptor and a guest molecule induces an electrochemical signal or response at the ligand–solution interface that can be detected via voltammetric methods, cyclic voltammetry, square wave voltammetry, and so on. The molecular receptors on the periphery of the NP are either redox active themselves or are designed to bind redox-active guest molecules. The noncovalent interactions associated with the binding event offer a highly versatile and modular approach since the interactions are by definition thermodynamically reversible. Various redox functionalities have been appended to NP cores via thiolated redox molecules, place exchange reactions, and postfunctionalization of a MPC through amide15 or ester-forming reactions.16 The redox moieties include ferrocene,17 biferrocene,18 phenol,19 nitrobenzene,20 and anthraquinone,21 which commonly are synthesized using a gold core composition. The observed voltammetry of these redox-active units tend to exhibit similar electrochemical potentials to their free
304
FUNCTIONAL NANOPARTICLES AS CATALYSTS AND SENSORS
Figure 11.2 Cartoon representation of Au MMPC labeled with a redox-active group phenothiazine on the NP periphery. Adapted from Ref. 16 with permission.
ligand counterparts. However, the size of the electrochemical wave in relation to the number of copies of redox species present per NP demonstrates the enhancement of sensitivity for a highly functionalized NP.22 To a first approximation, the redox species at the periphery of the NP tend to react independently of each other. For example, a gold MPC functionalized with phenothiazine ligands16 demonstrates that the number of electrochemically oxidizable phenothiazine groups is identical to the average number present as measured by 1 H NMR (Fig. 11.2). The independent, electrochemically NP reactive surfaces act as nanoscale receptors that provide a larger reactive surface area for both catalytic and sensing applications. An example of an exo-active surface is provided by Astruc and coworkers. They have developed MMPC nanoscale receptors by appending amidoferrocene recognition elements to the periphery of the NP surface. The ferrocene receptors selectively bind oxo anions, demonstrating the interdependence between molecular recognition and redox events.23 In this study, a series of gold NPs with a mixed monolayer of amidoferrocene and dodecanethiol ligands was synthesized through place exchange reactions (Fig. 11.3). Different concentrations of amidoferrocene units (7–38%) and spacer lengths (C6 and C11) were used to evaluate the redox
Figure 11.3 Complex formation between amidoferrocene-functionalized Au MMPC and dihydrogen phosphate. Adapted from Ref. 23 a with permission.
11.1
EXO-ACTIVE NANOPARTICLE SURFACES IN SENSING
305
environment of ferreocene moieties within the monolayer. The electrochemical oxidations of the attached ferrocene units show a single reversible voltammetric wave (E1/2 665 mV versus internal standard) regardless of concentration or spacer length. The appearance of a single wave suggests near equivalence of ferrocene redox sites on the NP periphery. Addition of [NBu4][H2PO4] causes the emergence of a second voltammetric wave at a less positive potential (E1/2 465 mV). The intensity of the second wave appears to increase upon subsequent additions of salt, while the corresponding initial wave decreases in intensity. The complete disappearance of the initial wave is observed when the H2PO4 anion saturates all of the amidoferrocene units on the MMPC, suggesting that the stoichiometry between amidoferrocene host and phosphoanion guest is 1 : 1. The large shift (DE1/2 ¼ 220 mV) in potential indicates a strong hydrogen bonding-mediated association event between the phosphoanion and the amido groups adjacent to the ferrocene. The shift remains constant over variations in amidoferrocene concentrations and spacer lengths. Fine-tuning of host–guest binding in the oxidized state (potential shift) is observed via stereoelectronic effects. MMPCs with terminal amido decamethylferrocene units demonstrate a smaller potential shift of 125 mV as compared to their amidoferrocene counterparts. The methyl substituents on the decamethylferrocene introduce both steric constraints and electron-releasing groups that reduce the interactions between the amido group and the phosphoanion. In contrast, the electron-withdrawing effect of an acetyl group conjugated to the terminal amidoferrocene produces a larger potential shift (E1/2 ¼ 275 mV) than previously recorded for amidoferrocene. The larger shift indicates an enhancement in association between the electron-poor amidoferrocenyl and the anion. As nanoscale receptors, the amidoferrocene MMPCs demonstrate selectivity toward dihydrogen phosphate over other anions. A more significant shift in the half-wave potential (DE ¼ 220 mV versus DE ¼ 40 mV) of the ferrocene units is observed for H2PO4 versus HSO4. The stronger interaction between the amidoferrocenium group and the phosphoanion arises from the higher charge density on the oxygen atom. The apparent selectivity confirms that the dominant hydrogen bonding interaction is the result of the NH group of the amide and the terminal oxygen atom of the anions. To enhance the efficiency of these nanoscale receptors, Astruc and coworkers synthesized MMPCs featuring ferrocene units on dendritic tethers24 (Fig. 11.4). The dendritic tethers enhance recognition with respect to the generation number because redox centers are in close proximity for higher generations.25 This positive dendritic effect is observed through a comparison between assemblies with AB3 and AB9 dendritic tethers for the recognition of H2PO4. Gold NPs employing different numbers of amidoferrocene or silylferrocene groups of dendrons (AB3 or AB9) were synthesized by using place exchange reactions or Brust-type reactions. Progressive addition of [NBu4][H2PO4] generates a second voltammetric wave at a less positive potential as seen with the previous monomeric amidoferrocene NPs. Again, complete disappearance of the initial wave occurs with the concomitant increase of the second voltammetric wave, indicating a strong molecular recognition
306
FUNCTIONAL NANOPARTICLES AS CATALYSTS AND SENSORS
Figure 11.4 Cartoon representation of (a) AB3 dendronized Au MMPC and (b) AB9 dendronized Au MMPC. Adapted from Ref. 24 with permission.
event. The NPs dendronized specifically with AB9 silylferrocenyl groups exhibit a larger potential shift than the respective AB3 units or monomeric amidoferrocene. Thus, the positive dendritic effect is observed for the AB9 silylferrocenyl groups. The NPs dendronized with amidoferrocene group adsorb onto the electrode surface, which decreases the efficiency of the dendritic effect. The silylferrocene dendronized NPs selectively recognize H2PO4 and the biologically relevant adenosine-50 -triphosphate anion (ATP2). Other anions, such as H2PO4, Cl, Br, and NO3, do not generate a significant shift in the ferrocenyl wave upon addition of their respective salts. In fact, the NPs are so highly selective toward H2PO4 and ATP2 that no significant potential shifts are observed even in the presence of other anions, such as HSO4 or Cl.
11.1
EXO-ACTIVE NANOPARTICLE SURFACES IN SENSING
N
NH
O
Zn
N
307
HN N
N O HN NH
O O O N H
N H
N
S S
S
S S
S S
H N
N
S
S
N Zn
N
S
S
Au
O H N
O S S
S
S S
S S
S
S S S
O
S S
S
S
S S
O HN O
O
HN
NH O
N
N Zn
N
N
NH
Figure 11.5 Schematic of anion sensing porphyrin Au MMPC. Reprinted with permission from Ref. 26.
Beer and Davis report a similar enhancement in anion binding affinities to NP exo-active surfaces.26 Redox-active amide-disulfide zinc metalloporphyrins were appended to gold NPs using a place exchange reaction (Fig. 11.5). The combination of metal–ligand and hydrogen bonding interactions directs binding of the anion to the metalloporphyrin receptor. A series of different anions (e.g., Cl, Br, I, NO3, H2PO4, ClO4) were titrated into the solutions of both the zinc metalloporphyrinfunctionalized NPs and the free metalloporphyrins. The results demonstrate that the anions bind to the NPs more strongly than the free metalloporphyrin form, even within competitive solvents. For example, the association constant for the zinc metalloporphyrin-functionalized NP and chloride is two fold larger than the free metalloporphyrin in DMSO. The surface confinement of the zinc metalloporphyrin on the NP periphery reduces the receptor’s conformational flexibility and allows entropic contributions to be more favorable. Thus, the preorganization of the zinc
308
FUNCTIONAL NANOPARTICLES AS CATALYSTS AND SENSORS
metalloporphyrin receptors at the NP surface enhances the anion coordination and is a significant factor in controlling the binding affinity of the anions. Exo-active surfaces are not only limited to incorporating redox moieties onto the NP via synthetic schemes, but can also use molecular receptors to bind redox-active guest molecules and generate an electrochemical response. The Rotello group has explored the redox behavior of molecular recognition units at the periphery of NP surfaces. Most of their work has focused on a three-point hydrogen bonding interaction between 2,6-diamidopyridine (DAP) and flavin, where DAP serves as a molecular receptor that binds to the complementary flavin derivative.27 The nature of the noncovalent interactions dictates the redox activity for flavin. Thus, MMPCs are well suited to tune the electrochemical properties of flavin through host–guest interactions on the NP periphery. In their initial study, a mixed monolayer of octanethiol and DAP thiol derivative was appended on 2 nm gold NP through a place exchange reaction yielding MMPC 128 (Fig. 11.6). Aliquots of the resultant MMPC 1 were then added to the flavin solution and the binding was determined via 1 H NMR titration. The flavin N(3) imide proton steadily shifts downfield upon addition of MMPC 1, indicating a hydrogen bonding recognition event. The shifts in the NMR clearly generate a 1 : 1 binding isotherm, providing an association constant (Ka) of 196 M1. The binding constant for MMPC 1 and flavin is consistent with binding constants observed for free DAP (150–500 M1).29 Redox modulation of the host–guest interactions was evaluated by reducing the flavin in the presence of MMPC 1. MMPC 1 stabilizes the reduced flavin radical anion by 81 mV (1.85 kcal/mol). The corresponding association constant (Ka) for the [Flred-:MMPC 1] complex is about 20-fold (4500 M1 versus 196 M1) greater than the flavin oxidized neutral species. The significant binding enhancement allows
Figure 11.6 (a) Three-point hydrogen bonding recognition between MMPC 1 and flavin. (b) Equilibrium between oxidized flavin (on right) and flavin radical anion (on left). Adapted from Ref. 28 with permission.
11.1
EXO-ACTIVE NANOPARTICLE SURFACES IN SENSING
309
modulation of the flavin:MMPC 1 complex from a high-affinity binding event to a low-affinity binding event via an electrochemical stimuli. The redox modulation process is selective as the reduction of a methylated flavin analog (methylated at the N(3) imide position to disrupt hydrogen bonding) is relatively unaffected by the MMPC 1.26 The Rotello group has extended these studies to tune molecular recognition and redox properties by incorporating multivalent recognition elements into nanoscale receptors. They have developed exo-active surfaces consisting of MMPCs that facilitate both hydrogen bonding and aromatic stacking interactions within the mixed monolayer. Specifically, gold MMPCs were synthesized with three-point hydrogen bonding DAP moieties and pyrene groups on the NP periphery (Fig. 11.7). The electron-deficient flavin guest molecule interacts favorably with the electron-rich pyrene as observed via 1 H NMR experiments. An association constant of 323 M1 was recorded for the multivalent nanoscale receptor, twofold greater than the previous flavin:MMPC 1 complex (196 M1).30 However, in their studies, they observe that the distance between an array of functional groups on the NP periphery and the core has a marked effect on the multivalent binding to an external guest.31 The difference in binding affinity of external guests is attributed to the radial nature of MMPCs, where order decreases with increasing distance from the gold core.32 The apparent radial effect is quantified using a series of MMPC receptors featuring both mono and divalent recognition elements, as well as different spacer lengths (Fig. 11.7). MMPCs 2 and 3 feature only hydrogen bonding DAP recognition elements. The distance between the functional group and the core is 4 carbons for MMPC 2 and 10 carbons for MMPC 3. MMPCs 4 and 5 were synthesized with both hydrogen bonding DAP and pyrene aromatic stacking recognition elements. The distance between pyrene and the core is 11 carbons for MMPC 4 and 6 carbons for MMPC 5. The distance of the DAP is maintained at C10 and C4 for comparison to their monovalent MMPCs (C10 for MMPC 4 and C4 for MMPC 5, respectively). The divalent MMPCs 4 and 5 demonstrate a large radial effect when compared to their monovalent counterparts (MMPCs 2 and 3). MMPCs 2 and 3 reveal a similar binding constant to the oxidized flavin (196 versus 185 M1), while the divalent MMPCs 4 and 5 record an overall binding enhancement due to the influence of the aromatic stacking between pyrene and oxidized flavin. However, the shorter, more preorganized MMPC 5 binds flavin more efficiently than its longer chain equivalent MMPC 4 (930 versus 320 M1). The tightly packed arrangement of the pyrene on the NP periphery facilitates stronger aromatic stacking interactions between the pyrene and the oxidized flavin. Upon the reduction of flavin, the radial control over multivalent binding is even more pronounced. The monovalent receptors generally stabilize the flavin radical anion resulting in stronger binding between flavin and DAP. However, MMPC 3 demonstrates slightly stronger binding to the flavin radical anion than the shorter chain equivalent (MMPC 2). The stronger binding and hence stabilization of the flavin is attributed to reduced congestion at the NP periphery for the longer chain MMPC 3. In contrast, the variation of chain length for the divalent MMPCs causes
310
FUNCTIONAL NANOPARTICLES AS CATALYSTS AND SENSORS
Figure 11.7 Monovalent DAP and divalent DAP–pyrene-functionalized MMPCs. Adapted from Ref. 31 with permission.
a significant difference in flavin binding affinity. The reduction of flavin renders the molecule electron rich that results in unfavorable interactions with the likewise electron-rich pyrene. The repulsive interactions decrease the binding affinity for the flavin radical anion for both MMPC 4 and MMPC 5. In fact, the repulsive interactions are so large for MMPC 5, that the short-chain system preferentially binds the electrondeficient oxidized flavin form rather than the flavin radical anion. The redox behavior for the MMPCs is a direct consequence of the alkyl chain length. The short-chain
11.1
EXO-ACTIVE NANOPARTICLE SURFACES IN SENSING
311
system creates a denser environment incapable of relieving the repulsive interactions, while the longer chain 11-carbon spacer MMPC 4 possesses enough conformational flexibility to accommodate the electron-rich flavin guest and minimize the repulsive aromatic stacking interactions.31 Rotello and coworkers have applied this multivalent approach to generate model systems for flavoenzyme activity that function in aqueous solution. They have investigated the redox behavior of flavin mononucleotide (FMN) at the NP surface as an attempt to mimic the biological environment of flavoenzymes in water.33 In this work, a positively charged trimethylammonium-functionalized gold NP (MMPC 6) was fabricated to interact with negatively charged FMN (Fig. 11.8). Fluorescence titrations demonstrate that the addition of positively charged MMPC 6 to FMN quenches the FMN fluorescence dramatically. The fluorescence titrations were fit to a binding isotherm and the association constant between FMN and MMPC 6 is determined to be K ¼ 3.7 107 M1. The exceptional association constant for the FMN : MMPC 6 complex is due to multivalent electrostatic particle–FMN interactions and the interfacial properties of the MMPC surface.34 In contrast, negatively charged MMPCs do not show any significant quenching of the FMN fluorescence, demonstrating that the binding was selective. Cyclic voltammetry measurements demonstrate that the addition of MMPC 6 to FMN shifts the half-wave reduction potential from 492 to 394 mV. The resultant FMN half-wave potential is within the range of redox potentials for flavoenzymes such as microsomal NADPH-cytochrome P-450 reductase (390 mV) and p-hydroxybenzoate hydroxylase (412 mV).35 The addition of excess NaCl disrupts the electrostatic interaction of the FMN:MMPC 6 complex and shifts the FMN reduction to more negative potentials (443 mV). The reduced FMN is slightly stabilized by high salt concentrations, which prevents the resultant FMN potential from being completely restored to its original value (492 mV). Supramolcular interactions at the NP periphery can alternatively be used to control self-assembly for exo-active surfaces. Guest molecules can readily assemble or disassemble at the NP periphery due to competitive binding interactions. The respective process is usually coupled to an electrochemical response, which provides detection for multiple analytes. For example, Kaifer and coworkers use competitive binding interactions to tune the electrochemical response of ferrocenemethanol (1) at the NP periphery36 (Fig. 11.9). They generate exo-active surfaces that consist of inclusion complexes of cyclodextrin (CD), chosen based on the ability of CDs to form inclusion complexes in aqueous media.37 They modify CD by converting the primary hydroxyl group on the smaller rim of beta-CD into a thiol. This modified CD is readily immobilized on various surfaces such as platinum,38 palladium,39 silver,40 and gold41 colloids. In this strategy, the CD is appended in an orientation where the wider opening is pointing away from the particle surface. The resultant NP (MMPC 7) is soluble in aqueous media and offers an optimized environment for specific guest recognition. Through voltammetric experiments, the reversible assembly of guest 1 was monitored at the NP periphery. Addition of MMPC 7 to 1 lowers the overall current and shifts the half-wave potential to more positive values. The electrochemical
312
Figure 11.8 (a) Schematic representation of FMN molecules binding to MMPC 6 and their respective chemical structures. (b) Cyclic voltammograms for FMN, FMN þ MMPC 6, and FMN þ MMPC 6 in the presence of salt. Adapted from Ref. 33 with permission.
11.1
EXO-ACTIVE NANOPARTICLE SURFACES IN SENSING
313
Figure 11.9 Competitive binding of 1 and 2 to the CD hosts on MMPC 7. Adapted from Ref. 36 with permission.
response indicates that an inclusion complex forms between 1 and CD receptors on MMPC 7. Subsequent addition of 1-adamantanol (2) results in an increase in current and a shift in the half-wave potential back to its original value. The electroinactive guest 2 displaces the bound 1, resulting in the reemergence of the voltammetric signal for free 1 in the bulk solution. Thus, the introduction of adamantanol effectively controls the self-assembly of 1 on the MMPC 7 and results in a detectable electrochemical response. The versatility and reversibility of exo-active surfaces have been further explored by using pseudo-rotaxane architectures. Pseudo-rotaxanes generated from the electron-deficient cyclophane cyclobis(paraquat-p-phenylene) (CBPQT4 þ ) (3) and 1,5-dialkoxynaphthalene units lead to supramolecular systems with near binary recognition properties.42 The resultant electrochemical switch allows for potentially recyclable, reusable, catalytic, or diagnostic materials.43 Cooke and coworkers incorporated pseudo-rotaxanes on the periphery of a NP as a means of reversibly modifying NPs structure and function.44 MMPCs featuring 1,5-dialkyloxynaphthalene moieties were synthesized using a place exchange reaction (MMPC 8) (Fig. 11.10). Addition of the MMPC 8 to 3 results in a significant broadening of the proton resonances in 1 H NMR spectrum. The resonance broadening indicates that the formation of a pseudo-rotaxane architecture occurs on the NP periphery. Control over complex formation was achieved via the reduction of the cyclophane to its diradical dicationic state. Square wave voltammetry was used to measure the electrochemical response of the psuedorotaxane complex (Fig. 11.10).
314
FUNCTIONAL NANOPARTICLES AS CATALYSTS AND SENSORS
Figure 11.10 (a) Chemical structures for MMPC 8, cyclophane 3, and tetrathiafulvalene 4. (b) Square wave voltammogram of 3 (—) and upon addition of MMPC 8 (. . .). Adapted from Ref. 44 with permission.
Upon the first reduction, the voltammetric wave shifts by 20–30 mV, presumably due to donor–acceptor interactions resulting from the naphthalene and cyclophane moieties.39 The second reduction remains relatively unaffected, indicating that the complex disassembles prior to the second reduction of the cyclophane. Alternatively, the complex can disassemble through competitive binding interactions with a tetrathiafulvalene derivative. Thus, Cooke and coworkers demonstrate that selfassembly at exo-active surfaces can be reversibly controlled via an external electrochemical stimulus or competitive binding interactions. Controlled self-assembly allows exo-active surfaces to be viable supramolecular building blocks for constructing nanostructure assemblies. These nanostructure assemblies can be used to modify electrodes for sensing applications. Willner and coworkers have constructed nanostructure assemblies on electrodes through electrostatic cross-linking of citrate stabilized gold NPs and bipyridinium cyclophane (3, 5– 7) derivatives45 (Fig. 11.11). Anionic gold NPs were first assembled on an ammonium-functionalized indium tin oxide (ITO) conductive surface. A bipyridinium cyclophane layer was then deposited onto the gold NPs resulting in an electrostatically cross-linked film at the NP periphery. The assembly process was repeated in a stepwise manner to achieve layer-by-layer assembly of anionic gold NPs and bipyridinium cyclophanes. The resultant nanostructure assembly can detect p-donor substrates, such as p-hydroquinone, adrenaline, and dopamine.46 In this process, the bipyridinium cyclophanes act as receptors for the association of p-donor substrates in their cavities. The binding between the bipyridinium cyclophanes and the p-donor substrates generate an electrochemical response that is observed through the 3D conductive layer assembly at the electrode surface. The sensitivity of the resulting electrochemical sensor was tuned via control over the number of assembled layers on the conductive surface.47 For example, layered assemblies consisting of cyclophane 3 and gold NPs were produced on ITO surfaces. The number of layers was varied from one to five in the presence of an
11.1
EXO-ACTIVE NANOPARTICLE SURFACES IN SENSING
315
Figure 11.11 Cyclophane derivatives and anionic gold NPs generate nanocomposite assemblies on an electrode surface. Adapted from Ref. 45 with permission.
analyte, p-hydroquinone (8). The amperometric responses for both 3 and 8 increase with respect to the number of layers, indicating that the number of cyclophanes present in the assembly directly affects the concentration of the bound analyte (Fig. 11.12). The decrease in peak-to-peak separation for 8 is also evident, suggesting that the electron transfer kinetics greatly improve upon the buildup of the number of layers. The five-layered modified electrode was calibrated for 8 using various concentrations of the analyte. The calibration curve is approximately linear with respect to [8], demonstrating that the sensor is not easily saturated. The detection limit for 8 is as low as 1 106 M. For a modified electrode using N,N0 -diaminoethyl-4,40 -bipyridinium (7), p-hydroquinone is undetectable at low concentrations (105 to 106 M). Thus, the successful electrochemical detection of 8 by the five-layered electrode is the result of specific host–guest interactions between the cyclophane and the p-hydroquinone. Incorporation of different cyclophane receptors in the layered nanocomposite assembly controls selectivity on the electrode surface. Layer-by-layer assembly of a larger cyclophane 5 and anionic gold NPs was used to modify an electrode. The larger cavity for 5 enables detection of bis-dihydroxymethylferrrocene (9) but not 8. In contrast, 9 is not detected by the smaller cyclophane nanocomposite assembly
316
FUNCTIONAL NANOPARTICLES AS CATALYSTS AND SENSORS
Figure 11.12 (A) Cyclic voltammograms of a bare ITO electrode (a) and 1–5 layers of cyclophane 3/anionic gold NP nanocomposites (b–f) in the presence of 8 (1 105 M). (B) Cyclic voltammograms of the five-layered nanocomposite at various concentrations of 8. Inset: calibration curve for 8. Adapted from Ref. 47 with permission.
because the guest is too large to fit within the smaller cavity. Modification of an electrode with a nanocomposite assembly using a cationic Pd(II) complex (6) demonstrates detection of both 8 and 9 since its cavity associates with the former in a diagonal orientation and its cavity is large enough to fit 9.48 The lattice morphology of the nanocomposite assembly also affects selectivity between analytes (Fig. 11.13). Two nanocomposites consisting of six total layers of cyclophane and gold NPs assembled on an electrode surface. Nanocomposite B was assembled with three layers of cyclophane 3 closer to the electrode surface,
Figure 11.13 (A) Cyclic voltammograms of five-layered assembly of cyclophane 5 and anionic gold NPs at various concentrations of 9: (a) 0 M, (b) 0.7 106 M, (c) 1.3 106 M, (d) 2.0 106 M, (e) 2.6 106 M, and (f) 5.2 105 M. Inset: calibration plot corresponding to the electrochemical responses of the modified electrode at different concentrations of 9. (B) Cyclic voltammograms of electrodes corresponding to (a) nanocomposite A and (b) nanocomposite B. Adapted from Ref. 47 with permission.
11.1
EXO-ACTIVE NANOPARTICLE SURFACES IN SENSING
317
followed by three layers of cyclophane 5 at the outer layers exposed to the bulk solution. Nanocomposite A features the reverse deposition process, where cyclophane 5 layers were deposited onto the electrode surface first, followed by the deposition of the cyclophane 3 layers. Both analytes (8 and 9) are detected with nanocomposite B. However, only 8 is observed in the voltammograms for nanocomposite A. The difference in selectivity suggests that the nanocomposite B is more porous than nanocomposite A. The cyclophane 3 layers are impervious to the large 9 guest molecule that prevents 9 to interact with the cyclophane 5 at the bottom of the nanocomposite assembly. The above examples demonstrate that exo-active surfaces can be used to produce very sensitive and selective electrochemical sensors. An alternative approach is provided by ion-sensitive field effect transistors (ISFET) to detect charged species at a conductive surface through the change in polarization at the sensing interface. Nanocomposite assemblies can accumulate onto these conductive surfaces and the change in polarization is used to transduce complexation with redox-inactive guests. An ISFET-based sensor was constructed using a similar electrostatic cross-linked nanocomposite of gold NPs and bipyridinium cyclophane derivatives on Al2O3.49 Layers of polyethyleneimine, anionic gold NPs, and cyclophane were deposited stepwise onto the sensing interface (Fig 11.14). The resultant sensor was able to detect any charged molecules that bind to the cyclophane receptor, including electroinactive molecules such as serotonin. Plots of source–drain potentials or gate–source potentials provide reliable detection of adrenaline over six orders of magnitude concentration and reveal a binding constant of 200 M1 between cyclophane and adrenaline. Overall, there are a variety of nanoscale receptors that use exo-active surfaces to detect specific analytes. The exo-active surfaces employ a number of noncovalent interactions to generate efficient binding between host and guest. The binding results in an electrochemical signal that produces an efficient detection method for sensing applications.
Figure 11.14 Schematic representation of cyclophane/gold NP ISFET. Adapted from Ref. 49 with permission.
318
11.2
FUNCTIONAL NANOPARTICLES AS CATALYSTS AND SENSORS
CORE-BASED MATERIALS AS SENSORS
Core-based materials are defined as NPs that produce an electrochemical signal or response at the NP core. They are suitable as electrochemical labels or tags because the electrochemical signal from the core provides information about biomolecules within the sample. The general purpose of the NP label is to monitor complex biological reactions or detect specific biomolecule targets. The core is usually decorated with a monolayer of molecular receptors that binds to a specific biomolecule target, such as an antigen, antibody, or DNA. The monolayer serves a dual purpose as it protects the particle from aggregation and provides a binding motif for the labeling of biomolecules. The resultant NPs can be made biocompatible, with tagged biomolecules retaining their bioactivity without any inhibition from the NP. Core-based materials are generally composed of a metallic NP core since the metallic core is conductive and facilities efficient electron transfer to the electrode. Gold NPs are most frequently used as electrochemical labels due to their conductivity, biocompatibility, and ease of fabrication. The gold NPs remain as inert labels until they are dissolved to their trace metal ions and detected via stripping voltammetry. Stripping voltammetry is a very sensitive technique since the trace metal ions are directly proportional to bioanalyte concentrations in solution. Limoges and coworkers have developed core-based materials for a noncompetitive heterogeneous electrochemical immunoassay based upon an anodic stripping voltammetry (ASV) technique.50 Gold NPs (18 nm in diameter) were functionalized with a secondary antibody and used as an electrochemical label for the detection of goat immunoglobulin G (IgG) (Fig. 11.15). Primary antibodies for IgG are passively adsorbed onto the walls of a polystyrene microwell. The primary antibody captures the IgG, while the secondary antibody tagged with a gold NP creates a sandwich complex. Washing with buffer then reduces the amount of nonspecific adsorption from the labeled antibodies. The gold NPs are released from the immobilized secondary antibody via an acidic oxidative bromide–bromide solution. The gold NPs are oxidized to soluble AuIII ions and analyzed in solution via ASV. The electrochemical response generated at the electrode surface is then used to quantify the amount of IgG in the sample. The electrochemical immunoassay exhibits a dynamic range of 0.5– 100 ng/mL and a detection limit is estimated at 3 1012 M IgG. The picomolar detection limit is achieved due to the sensitivity of the stripping voltammetry technique as well as the relatively large number of gold ions produced from the dissolution of the gold NPs from a single recognition event. The Limoges group has adopted a similar electrochemical method for developing a DNA sensor by labeling an oligonucleotide with a 20 nm gold NP.51 The target 406-base human cytomegalovirus (HCMV) DNA sequence was immobilized on polystyrene microwells. The electrochemical NP probe hybridizes with the target and dissolution by an acidic bromide–bromide solution results in the release of AuIII ions. The ions were detected by anodic stripping voltammetry at sandwich-type screen printed microband electrodes. The microband electrodes improve the sensitivity of the AuIII detection in the small volumes of the nonagitated solution.
11.2
CORE-BASED MATERIALS AS SENSORS
319
Figure 11.15 Noncompetitive heterogeneous electrochemical immunoassay based on gold NP label and using ASV detection. Adapted from Ref. 50 with permission.
The electrochemical DNA sensor shows impressive sensitivity and selectivity. The detection limit, estimated to be 5 pM, is competitive with previously reported HCMV DNA hybridization assays based on an enzyme label.52 Introduction of a noncomplementary DNA sequence (human ETS2 DNA gene) records an electrochemical signal consistent with background noise. The electrochemical DNA sensor is therefore selective for the HCMV DNA sequence. Semiconductor NPs, such as CdS, are commonly used as labels for optical detection of bioanalytes due to their inherent fluorescent properties. Several reviews on semiconductor NPs as fluorescent labels for biosensors are currently available in the literature.53 However, since these fluorescent labels are beyond the scope of this chapter, only semiconductor NPs that involve electrochemical detection methods (stripping voltammetry or photoelectrochemical detection) will be discussed. Semiconductor NPs have been used as electrochemical labels in combination with stripping voltammetry techniques. Wang et al. immobilized three different kinds of nucleic acids on magnetic particles with different DNA targets.54 DNA probes labeled with ZnS, CdS, and PbS NPs were added to the solution. The probes hybridize with their complementary DNA targets, and dissolution of the semiconductor NPs yields trace metal ions for stripping voltammetry. The well-defined and resolved stripping peaks at 1.12 V (Zn), 0.68 V (Cd), and 0.53 V (Pb) (versus Ag/AgCl reference) thus enable simultaneous and sensitive electrochemical detection of different DNA targets.
320
FUNCTIONAL NANOPARTICLES AS CATALYSTS AND SENSORS –
SO3Na+
H2N S S CdS S S S – SO3Na+
O O S
N O
O
–
SO3–Na+
Na+
SO3
CdS
–
SO3Na+
NH2
SO3–Na+ SO3–Na+ SO3–Na+ O
H2N
O
H
CdS
H
N=CH
AChE
CHO
SO3–Na+ SO3–Na+ O
SO3–Na+ e
–
CdS h SO3–Na+ S S
N+(CH3)3
CH
hn
N
N+(CH3)3
S CH
N
AChE
O O–
+
SO3–Na+ e–
HS
N+(CH3)3
N+(CH3)3
Figure 11.16 Assembly of CdS NP/AChE modified electrode used for the photochemical detection of enzyme activity. Reprinted with permission from Ref. 55.
Core-based materials have applications in more direct sensing methods since their core properties can be activated in the presence of specific analytes. Enzyme–NP hybrid systems can utilize the biocatalytic process of an enzyme to activate the function of the NP.13d The core-based material generates an electrochemical or photoelectrochemical signal in response to the particular biocatalytic process, which results in the detection of a specific analyte. Semiconductor NPs have been used as photoelectrochemical labels for photocurrent generation and biosensor applications. Willner and coworkers have modified a gold electrode with an acetylcholine esterase (AChE)-functionalized CdS nanoparticle for the photoelectrochemical detection of AChE inhibitors55 (Fig. 11.16). The CdS nanoparticles were first capped with cystamine and mercaptoethane sulfonic acid, and then covalently attached to the gold electrode. Glutaric dialdehyde was then used as a bridging unit to covalently link the AChE to CdS. The surface coverage of AChE on the modified electrode was 3.9 1012 mol/cm2, which correlates to one AChE per 2.4 NPs. In the presence of acetylcholine, AChE catalyzes the hydrolysis reaction of acetylcholine to acetate and thiocholine. The thiocholine, in turn, donates electrons
11.2
CORE-BASED MATERIALS AS SENSORS
321
to the valence band holes upon photoexcitation of CdS. The scavenging of the valence band holes facilitates the generation of a photocurrent. The electrons accumulate in the conduction band and their transfer to the electrode initiates the photocurrent. Control experiments demonstrate that the photocurrent generation is active only in the presence of both acetylcholine and AChE–CdS modified electrode. Enzyme inhibitors, such as 1,5-bis(4-allyldimethylammoniumphenyl)pentan-3-one dibromide), block the biocatalytic process for AChE and decrease the photocurrent of the biosensor. However, the photocurrent was completely restored to the original intensity by washing the enzyme inhibitor from the sample cell. The versatile AChE–CdS modified electrode provides a functional interface for the active core CdS material to be tuned via a biocatalytic process and consequently serves as an efficient sensor for AChE inhibitors. Willner and coworkers have extended this approach to electron relay systems where core-based materials facilitate the electron transfer from redox enzymes in the bulk solution to the electrode.56 Enzymes usually lack direct electrical communication with electrodes due to the fact that the active centers of enzymes are surrounded by a thick insulating protein shell that blocks electron transfer. Metallic NPs act as electron “mediators” or “wires” that enhance electrical communication between enzyme and electrode due to their inherent conductive properties.47 Bridging redox enzymes with electrodes by electron relay systems provides enzyme electrode hybrid systems that have bioelectronic applications, such as biosensors and biofuel cell elements.57 Willner’s research with core-based materials, specifically gold NPs, is based upon their previous work using self-assembled monolayers to align both redox enzymes and intermediary electron relay units on the electrode to stimulate efficient communication between the enzyme redox centers and the electrode.58 They used a 1.4 nm gold NP functionalized with N6-(2-aminoethyl)-flavin adenine dinucleotide (FAD cofactor amino derivative) as an active core material. Reconstitution of apoflavoenzyme, apo-glucose oxidase (apo-GOx), with the NP allows for direct electrical communication with the electrode (Fig. 11.17). The enzyme–NP hybrid system was conjugated to the electrode through dithiols or alternatively the FADfunctionalized NP was first appended to the electrode, followed by subsequent addition of apo-GOx. The prepared enzyme electrodes reveal similar protein surface coverage of 1 1012 mol/cm2, regardless of their preparation method. The reconstituted GOx layer does not require any mediator for an efficient electrical communication with the electrode and the enzyme assembly induces the bioelectrocatalyzed oxidation of glucose. The enzyme–NP hybrid assembly provides excellent electrical connection with electrode, demonstrating an electron transfer rate of 5000 s1. This efficient electron communication helps to minimize the effects of interferants such as oxygen and ascorbic acid. The rate-limiting step in the electron transfer process is the charge transport across the dithiol molecular linker bridging the particle and the electrode. Thus, the enzyme electrode system exhibits highly efficient electrical communication with enhanced turnover rates and provides an effective sensor for glucose in the physiological concentration regime.
322
FUNCTIONAL NANOPARTICLES AS CATALYSTS AND SENSORS
Figure 11.17 (a) Fabrication of a GOx electrode via reconstitution of apo-GOx on a FADfunctionalized electrode. (b) Plot of current versus glucose concentration for modified electrodes. Adapted from Ref. 56 with permission.
11.3
CATALYTIC PROPERTIES OF NPS
NPs serve as highly efficient catalysts in a variety of applications, such as fuel cells, gas diffusion electrodes, and catalytic biosensors.59 They exhibit a large surface area to volume ratio that provides an advantage over 2D catalytic surfaces. Their size, shape, and capping ligands can be adjusted for specific catalytic applications providing control over their properties. Their ease of handling allows for facile NP recovery from the reaction medium, providing potentially reusable catalytic materials. Metallic NPs are most widely used in catalytic applications due to their inherent properties. Several examples of platinum and gold NPs are apparent in the literature. For example, electrodeposited platinum NPs on porous carbon substrates exhibit electrocatalytic activity for the oxidation of methanol.60 In another example, gold NPs catalyze the electrochemical oxidation of nitric oxide on modified electrodes.61 In general, catalytic NPs provide two distinct functions: enhancing an electrochemical reaction and/or increasing electron transfer to an electrode. The enhancement of electrochemical reactions usually occurs at modified electrode surfaces. The large NP surface area facilitates oxidation or reduction of small molecules at the electrode. For example, Niwa et al. modified a carbon film electrode with electrocatalytic platinum NPs to create a sensitive H2O2 sensor.62 Co-sputtering of carbon and platinum produces platinum NPs of 2.5 nm diameter embedded in a carbon matrix. The platinum NPs catalyze the oxidation of H2O2 at the electrode
11.3
CATALYTIC PROPERTIES OF NPS
323
surface and shift the H2O2 oxidation peak potential 170 mV lower than the corresponding platinum bulk electrode. Thus, a faster electron transfer rate and a higher electrocatalytic current were observed for the Pt modified electrode. Durand and coworkers modified a glassy carbon rotating disk electrode with platinum NPs to study the electrochemical reduction of oxygen in both acidic63 and alkaline64 media. They have discovered that a correlation exists between particle size and catalytic activity for platinum NPs. Larger platinum NPs demonstrated enhanced catalytic activity over smaller particles due to the stronger adsorption of oxygenated species. In contrast, only a slight correlation of particle size and catalytic activity was observed under basic conditions. Other metallic NPs have demonstrated extraordinary catalytic activity for the electrochemical reduction of oxygen over their bulk metal electrodes.65 Yet, Yang et al. observed that the type of metallic NP as well as the electrode preparation method influences the catalytic activity of the oxygen reduction reaction.66 Four different metallic NPs (Pd, Pt, Au, Ag) were electrodeposited onto screen printed carbon electrodes (SPE). The electrochemical response of the modified electrodes toward dissolved oxygen was determined via cyclic voltammetry (Fig. 11.18). Positive shifts in the peak potential were recorded for the NP modified electrodes as compared to bare SPE. Positive shifts indicated an enhancement in catalytic activity for the oxygen reduction reaction, resulting in the following order of enhancement from largest to smallest (Pd > Pt > Au > Ag). Furthermore, the preanodized SPE*-Pd exhibited a higher peak signal at the same potential ( þ 100 mV) as SPE-Pd. The higher peak signal suggested that the specific preparation method was a key factor in controlling electrocatalytic activity at the electrode surface. Besides enhancing electrochemical reactions, catalytic NPs assist in electron transfer. Typically, conductive NP films composed of catalytic NPs are deposited
Figure 11.18 Reduction of oxygen on various chemically modified SPEs at n ¼ 20 mV/s. Reprinted with permission from Ref. 66.
324
FUNCTIONAL NANOPARTICLES AS CATALYSTS AND SENSORS
Figure 11.19 (a) Schematic representation of silver NPs and myoglobin modified electrode. (b) Cyclic voltammograms of modified electrode (—) and bare electrode (. . .). Adapted from Ref. 68 with permission.
on electrode surfaces to enhance electron transfer from analytes in the bulk solution to the electrode. For example, colloidal gold NPs have been shown to facilitate electron transfer of cytochrome c in solution to an electrode. The gold NPs act as “electron antenna” that efficiently funnel electrons between the electrode and the electrolyte solution.67 Thus, metallic NPs serve as electron mediators for direct communication between the analytes and the electrode. The actual catalytic functions of NPs can occur either at the metallic/semiconductor core or on an active site at the NP periphery. Catalysis at the core takes advantage of the inherent conductive or reactive properties of the metallic surface. For example, Li et al. reports a process for enhancing electron transfer reactivity and catalytic activity of myoglobin (Mb) by silver NPs68 (Fig. 11.19). The silver NPs act as electron “wires” connecting the electron transfer process between the heme group deeply buried in the Mb and the electrode. Solutions of silver NPs and Mb were deposited onto pyrolytic graphite electrode in a specific ratio (VMb/VAgNPs 7 : 1) to achieve the best protein surface coverage and hence the highest electrochemical response. The modified electrode demonstrates well-defined redox peaks for Mb and a half-wave potential of 305 mV that is consistent with previous reports.69 The peak separation of 48 mV indicates a fast, quasi-reversible, one-electron, heterogeneous electron transfer process. No peaks were observed for a bare pyrolytic graphite electrode, suggesting that the silver NPs are directly responsible for the electrical communication between the Mb and the modified electrode. The modified electrode, in turn, enhances Mb to catalyze the reduction of H2O2. Addition of H2O2 increases in the cathodic peak intensity for the modified electrode. This electrochemical response is specific for the Mb:silver NP modified electrode because no signals are observed for either the bare electrode or an electrode with only a silver NP film. A Michaelis–Menten analysis demonstrates that the modified electrode shows higher affinity for H2O2 than other gold–cytochrome c and gold– horseradish peroxidase modified electrodes. The apparent enhanced catalytic activity
11.3
CATALYTIC PROPERTIES OF NPS
325
produces a sensitive H2O2 sensor with a detection limit of 1 106 M and the sensitivity of 0.205 mA per mM of H2O2. In addition to metallic NPs, semiconductor NPs demonstrate catalytic core properties that enable electron transfer to the electrode. Li has evaluated CdS NPs as effective electrical “wires” for biological processes.70 A mixture of hemoglobin and CdS NPs was immobilized onto a pyrolytic graphite electrode. The hemoglobin exhibits direct electrochemical behavior with quasi-reversible voltammetric waves and a half-wave potential of 288 mV. Both the cathodic and anodic peaks demonstrated a linear relationship with scan rate, supporting the immobilization of hemoglobin on the electrode. Again, the electron transfer process is indicative of the CdS electron relay system. No electrochemical signals were observed for the bare electrode or hemoglobin appended to the electrode surface in the absence of CdS. The effective enhancement of the electron transfer is thought to be dependent not only on the conductivity of NPs, but also on the arrangement between CdS and hemoglobin. Metallic/semiconductor NPs exhibit reactive core properties that result in both electrocatalytic activity and conductivity.71 For example, Ohsaka and coworkers have used the electrocatalytic activity and conductivity of gold NPs to selectively detect dopamine in the presence of ascorbate.72 Citrate stabilized gold NPs were immobilized onto a cystamine modified electrode. Cyclic voltammograms were obtained for the electrochemical oxidation of dopamine for the modified and bare electrodes. The modified electrode exhibited a smaller peak-to-peak separation (38 – 2 mV), indicating faster electron transfer kinetics than the bare electrode (60 – 3 mV at 100 mV/s). The electron transfer reaction is roughly about 5.7 times faster than the bare electrode, according to Nicholson theory.73 The faster reaction time was attributed to weakly adsorbed dopamine on the gold NPs of the modified electrode. The improved reversible redox behavior at the modified electrode surface generated a highly sensitive voltammetric sensor for dopamine. In the presence of ascorbate, a biological interferant, the modified electrode exhibits good sensitivity and selectivity for the detection of dopamine. Two well-defined voltammetric peaks, corresponding to the oxidations of ascorbate and dopamine, were visible in the cyclic and square wave voltammograms (Fig. 11.20). The ascorbate was readily oxidized by the electrocatalytic gold NPs before the oxidation potential of dopamine was reached, resulting in the large separation between ascorbate and dopamine voltammetric peaks (165 mV). Only a broad voltammetric wave was observed for the bare electrode, rendering the two compounds indistinguishable in the voltammograms. Electrode fouling coupled with catalytic oxidation of ascorbate by the oxidized dopamine produced poor selectivity and reproducibility for the bare electrode. Thus, the electrocatalytic activity and conductivity of gold NPs provide the selective determination of dopamine in the presence of ascorbate at relative biological concentrations. The reactive NP core provides an alternate use for catalytic NPs as sensitive electrocatalytic tags for biosensors. Brozik and coworkers have developed a reagentless electrochemical immunoassay by using electrocatalytic NP modified antibodies that are sensitive to the oxygen reduction reaction.74 Gold/palladium core–shell
326
FUNCTIONAL NANOPARTICLES AS CATALYSTS AND SENSORS
Figure 11.20 (A) Schematic representation of modified gold NP electrode. (B) Cyclic voltammograms for dopamine in the presence of ascorbate at bare electrode (a) and modified electrode (b). (C) Corresponding square wave voltammograms at bare electrode (a) and modified electrode (b). Adapted from Ref. 72 with permission.
particles were appended to antitumor necrosis factor (TNF-a) antibody. The resultant “sandwich” immunoassay demonstrated catalytic activity from the tagged TNF-a antibody with a peak potential at 160 mV and a slightly smaller peak potential at 570 mV. The peak potentials were assigned to the reduction of oxygen to H2O2 and the reduction of oxygen to H2O, respectively. No peaks were observed upon degassing with argon, suggesting that oxygen was necessary to provide the detectable electrocatalytic signal. The detection limit was calculated to be 1 ppt. Thus, catalytic NPs can be used as electrocatalytic tags that provide sensitive sensing devices without excess reagents or complicated device fabrication techniques. Catalysis at the NP periphery can proceed through active sites on the NP surface. The number, function, and spacing between active sites on the NP can be tuned via facile fabrication techniques. For example, Mirkin et al. used the catalytic amplification of silver on the periphery of a DNA-functionalized gold NP to detect DNA on microelectrodes75 (Fig. 11.21). Captured nucleic acids are immobilized onto a SiO2 20 mm gap microelectrode. The modified microelectrode is exposed to a target 27-mer nucleotide as well as a DNA-functionalized gold NP. The target 27-mer nucleotide hybridizes with the complementary shorter capture sequence immobilized on the microelectrode surface allowing the free 30 end of the target to be available for further hybridization with the gold NP. A solution of Ag þ ions provides electrostatic deposition of Ag þ to the active sites (phosphate groups on the DNA backbone) on the NP. Addition of hydroquinone reduces the bound Ag þ ions, resulting in the formation of metallic silver NPs along the DNA backbone. The deposited silver particles further catalyze the silver reduction and ultimately enhance the conductivity
11.4
CONCLUSIONS
327
Figure 11.21 Catalytic amplification of the DNA detection system. Adapted from Ref. 75 with permission.
as they continue to grow across the microelectrode gap. A buffer with the appropriate ionic strength was then used to wash off any nonspecific DNA adsorption from the microelectrode surface. The silver catalytic amplification technique provides a highly efficient method for the electrochemical detection of DNA. The detection limit for the system is 5 1013 M, which is less than conventional fluorescencebased systems that utilize confocal microscopy (5 1012 M). The silver catalytic technique is extremely sensitive since no polymerase chain reaction amplification is necessary to detect the target DNA concentrations between 50 nM and 500 fM. Mirkin has utilized this silver catalytic amplification technique to develop biobarcode assays that demonstrate ultrasensitive detection of proteins and nucleic acids.76
11.4
CONCLUSIONS
NPs provide highly efficient catalysts and sensors due to their unique chemical and physical properties. NPs can be used as exo-active surfaces where a multitude of molecular receptors can bind analytes and generate a signal. Alternatively, NPs can be used as core-based materials in which biocatalytic processes can activate their core or they provide a biologically inert electrochemical label. As catalysts, NPs utilize their large surface area to volume ratio and enhance either electrochemical reactions or electron transfer at an electrode. The use of NPs in catalysts and sensors will continue as these functional materials serve as active units within these applications.
328
FUNCTIONAL NANOPARTICLES AS CATALYSTS AND SENSORS
REFERENCES 1. (a) S. Laurent, D. Forge, M. Port, A. Roch, C. Robic, L. V. Elst, R. N. Muller, Chem. Rev. 2008, 108, 2064. (b) C. J. Murphy, A. M. Gole, J. W. Stone, P. N. Sisco, A. M. Alkilany, E. C. Goldsmith, S. C. Baxter, Acc. Chem. Rev. 2008, 41, 1721. (c) M. A. El-Sayed, Acc. Chem. Res. 2004, 37, 326. (d) J. Z. Zhang, Acc. Chem. Res. 1997, 30, 423. 2. (a) J. Wang, Anal. Chim. Acta 2003, 500, 247. (b) N. T. K. Thanh, Z. Rosenzweig, Anal. Chem. 2002, 74, 1624. (c) C. Zhang, Z. Y. Zhang, B. B. Yu, J. J. Shi, X. R. Zhang, Anal. Chem. 2002, 74, 96. 3. A. C. Templeton, M. P. Wuelfing, R. W. Murray, Acc. Chem. Res. 2000, 33, 27. 4. U. Drechsler, B. Erdogan, V. M. Rotello, Chem. Eur. J. 2004, 10, 5570. 5. M. Brust, M. Walker, D. Bethell, D. J. Schiffrin, R. Whyman, J. Chem. Soc. Chem. Commun. 1994, 801. 6. M. J. Hostetler, J. E. Wingate, C. J. Zhong, J. E. Harris, R. W. Vachet, M. R. Clark, J. D. Londono, S. J. Green, J. J. Stokes, G. D. Wignall, G. L. Glish, M. D. Porter, N. D. Evans, R. W. Murray, Langmuir 1998, 14, 17. 7. A. C. Templeton, M. J. Hostetler, E. K. Warmoth, S. W. Chen, C. M. Hartshorn, V. M. Krishnamurthy, M. D. E. Forbes, R. W. Murray, J. Am. Chem. Soc. 1998, 120, 4845. 8. (a) L. Wang, E. K. Wang, Electrochem. Commun. 2004, 6, 49. (b) L. Wang, E. K. Wang, Electrochem. Commun. 2004, 6, 225. (c) L. Pasquato, P. Pengo, P. Scrimin, J. Mater. Chem. 2004, 14, 3481. (d) H. Cai, C. Xu, P. G. He, Y. Z. Fang, J. Electroanal. Chem. 2001, 510, 78. 9. J. Wang, G. D. Liu, R. Polsky, A. Merkoci, Electrochem. Commun. 2002, 4, 722. 10. (a) A. P. Alivisatos, Science 1996, 271, 933. 11. (a) D. F. Cao, P. L. He, N. F. Hu, Analyst 2003, 128, 1268. (b) S. Q. Liu, Z. H. Dai, H. Y. Chen, H. X. Ju, Biosens. Bioelectron. 2004, 19, 963. (c) Y. Zhang, P. L. He, N. F. Hu, Electrochim. Acta 2004, 49, 1981. 12. (a) A. X. Li, F. Yang, Y. Ma, X. R. Yang, Biosens. Bioelectron. 2007, 22, 1716. (b) S. G. Wu, T. L. Wang, C. Q. Wang, Z. Y. Gao, C. Q. Wang, Electroanalysis 2007, 19, 659. (c) A. Myc, I. J. Majoros, T. P. Thomas, J. R. Baker, Biomacromolecules 2007, 8, 13. (d) A. Gomez-Escudero, M. A. Azagarsamy, N. Theddu, R. W. Vachet, S. Thayumanavan, J. Am. Chem. Soc. 2008, 130, 11156. 13. (a) A. N. Kawde, J. Wang, Electroanalysis 2004, 16, 101. (b) J. Wang, R. Polsky, A. Merkoci, K. L. Turner, Langmuir 2003, 19, 989. 14. (a) X. L. Luo, A. Morrin, A. J. Killard, M. R. Smyth, Electroanalysis 2006, 18, 319. (b) J. Wang, Analyst 2005, 130, 421. (c) Y. Joseph, I. Besnard, M. Rosenberger, B. Guse, H. G. Nothofer, J. M. Wessels, U. Wild, A. Knop-Gericke, D. S. Su, R. Schlogl, A. Yasuda, T. Vossmeyer, J. Phys. Chem. B 2003, 107, 7406. (d) H. Haick, J. Phys. D: Appl. Phys. 2007, 40, 7173. 15. A. C. Templeton, D. E. Cliffel, R. W. Murray, J. Am. Chem. Soc. 1999, 121, 7081. 16. D. T. Miles, R. W. Murray, Anal. Chem. 2001, 73, 921. 17. (a) S. J. Green, J. J. Stokes, M. J. Hostetler, J. Pietron, R. W. Murray, J. Phys. Chem. B 1997, 101, 2663. (b) S. J. Green, J. J. Pietron, J. J. Stokes, M. J. Hostetler, H. Vu, W. P. Wuelfing, R. W. Murray, Langmuir 1998, 14, 5612. (c) D. Li, Y. J. Zhang, J. G. Jiang, J. H. Li, J. Colloid Interface Sci. 2003, 264, 109.
REFERENCES
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.
329
M. Yamada, H. Nishihara, Langmuir 2003, 19, 8050. S. W. Chen, Langmuir 1999, 15, 7551. S. W. Chen, K. Huang, Langmuir 2000, 16, 2014. (a) R. S. Ingram, R. W. Murray, Langmuir 1998, 14, 4115. (b) J. J. Pietron, R. W. Murray, J. Phys. Chem. B 1999, 103, 4440. R. W. Murray, Chem. Rev. 2008, 108, 2688. (a) A. Labande, J. Ruiz, D. Astruc, J. Am. Chem. Soc. 2002, 124, 1782. (b) A. Labande, D. Astruc, Chem. Commun. 2000, 1007, M. C. Daniel, J. Ruiz, S. Nlate, J. C. Blais, D. Astruc, J. Am. Chem. Soc. 2003, 125, 2617. (a) C. Valerio, J. L. Fillaut, J. Ruiz, J. Guittard, J. C. Blais, D. Astruc, J. Am. Chem. Soc. 1997, 119, 2588. (b) C. Valerio, E. Alonso, J. Ruiz, J. C. Blais, D. Astruc, Angew. Chem., Int. Ed. 1999, 38, 1747. P. D. Beer, D. P. Cormode, J. J. Davis, Chem. Commun. 2004, 414. (a) V. M. Rotello, Curr. Org. Chem. 2001, 5, 1079. (b) M. D. Greaves, V. M. Rotello, J. Am. Chem. Soc. 1997, 119, 10569. (c) R. Deans, G. Cooke, V. M. Rotello, J. Org. Chem. 1997, 62, 836. (d) A. O. Cuello, C. M. McIntosh, V. M. Rotello, J. Am. Chem. Soc. 2000, 122, 3517. (e) V. M. Rotello, Curr. Opin. Chem. Biol. 1999, 3, 747. (f) A. Niemz, V. M. Rotello, Acc. Chem. Res. 1999, 32, 44. A. K. Boal, V. M. Rotello, J. Am. Chem. Soc. 1999, 121, 4914. E. Breinlinger, A. Niemz, V. M. Rotello, J. Am. Chem. Soc. 1995, 117, 5379. A. K. Boal, V. M. Rotello, J. Am. Chem. Soc. 2000, 122, 734. A. K. Boal, V. M. Rotello, J. Am. Chem. Soc. 2002, 124, 5019. (a) W. D. Luedtke, U. Landman, J. Phys. Chem. 1996, 100, 13323. (b) M. J. Hostetler, J. J. Stokes, R. W. Murray, Langmuir 1996, 12, 3604. A. Bayir, B. J. Jordan, A. Verma, M. A. Pollier, G. Cooke, V. M. Rotello, Chem. Commun. 2006, 4033. R. A., Robinson, R. H., Stokes, Electrolyte Solutions, Butterworths, London, 1968. T. Iyanagi, N. Makino, H. S. Mason, Biochemistry 1974, 13, 1701. J. Liu, W. Ong, E. Roman, M. J. Lynn, A. E. Kaifer, Langmuir 2000, 16, 3000. (a) M. V. Rekharsky, Y. Inoue, Chem. Rev. 1998, 98, 1875. (b) K. A. Connors, Chem. Rev. 1997, 97, 1325. (c) A. E. Kaifer, Acc. Chem. Res. 1999, 32, 62. (d) J. Liu, J. Alvarez, A. E. Kaifer, Adv. Mater. 2000, 12, 1381. J. Alvarez, J. Liu, E. Roman, A. E. Kaifer, Chem. Commun. 2000, 1151. (a) J. Liu, J. Alvarez, W. Ong, E. Roman, A. E. Kaifer, Langmuir 2001, 17, 6762. (b) L. Strimbu, J. Liu, A. E. Kaifer, Langmuir 2003, 19, 483. J. Liu, W. Ong, A. E. Kaifer, C. Peinador, Langmuir 2002, 18, 5981. J. Liu, S. Mendoza, E. Roman, M. J. Lynn, R. L. Xu, A. E. Kaifer, J. Am. Chem. Soc. 1999, 121, 4304. P. L. Anelli, P. R. Ashton, R. Ballardini, V. Balzani, M. Delgado, M. T. Gandolfi, T. T. Goodnow, A. E. Kaifer, D. Philp, M. Pietraszkiewicz, L. Prodi, M. V. Reddington, A. M. Z. Slawin, N. Spencer, J. F. Stoddart, C. Vicent, D. J. Williams, J. Am. Chem. Soc. 1992, 114, 193.
330
FUNCTIONAL NANOPARTICLES AS CATALYSTS AND SENSORS
43. (a) F. Ilhan, M. Gray, V. M. Rotello, Macromolecule 2001, 34, 2597. (b) S. Sivakova, S. J. Rowan, Chem. Soc. Rev. 2005, 34, 9. (c) H. Hofmeier, U. S. Schubert, Chem. Commun. 2005, 2423. 44. G. Cooke, J. F. Garety, S. G. Hewage, G. Rabani, V. M. Rotello, P. Woisel, Chem. Commun. 2006, 4119. 45. A. N. Shipway, M. Lahav, R. Blonder, I. Willner, Chem. Mater. 1999, 11, 13. 46. M. Lahav, A. N. Shipway, I. Willner, J. Chem. Soc.,Perkin Trans. 1999, 2, 1925. 47. E. Katz, I. Willner, J. Wang, Electroanalysis 2004, 16, 19. 48. M. Lahav, R. Gabai, A. N. Shipway, I. Willner, Chem. Commun. 1999, 1937. 49. A. N. Shipway, E. Katz, I. Willner, ChemPhysChem 2000, 1, 18. 50. M. Dequaire, C. Degrand, B. Limoges, Anal. Chem. 2000, 72, 5521. 51. L. Authier, C. Grossiord, P. Brossier, B. Limoges, Anal. Chem. 2001, 73, 4450. 52. (a) F. Azek, C. Grossiord, M. Joannes, B. Limoges, P. Brossier, Anal. Biochem. 2000, 284, 107. (b) O. Bagel, C. Degrand, B. Limoges, M. Joannes, F. Azek, P. Brossier, Electroanalysis 2000, 12, 1447. 53. (a) U. Resch-Genger, M. Grabolle, S. Cavaliere-Jaricot, R. Nitschke, T. Nann, Nat. Methods, 2008, 5, 763. (b) X. Michalet, F. F. Pinaud, L. A. Bentolila, J. M. Tsay, S. Doose, J. J. Li, G. Sundaresan, A. M. Wu, S. S. Gambhir, S. Weiss, Science 2005, 307, 538. 54. J. Wang, G. D. Liu, A. Merkoci, J. Am. Chem. Soc. 2003, 125, 3214. 55. V. Pardo-Yissar, E. Katz, J. Wasserman, I. Willner, J. Am. Chem. Soc. 2003, 125, 622. 56. Y. Xiao, F. Patolsky, E. Katz, J. F. Hainfeld, I. Willner, Science 2003, 299, 1877. 57. M. De, P. S. Ghosh, V. M. Rotello, Adv. Mater. 2008, 20, 1. 58. (a) I. Willner, V. Heleg-Shabtai, R. Blonder, E. Katz, G. L. Tao, A. F. Buckmann, A. Heller, J. Am. Chem. Soc. 1996, 118, 10321. (b) M. Zayats, E. Katz, I. Willner, J. Am. Chem. Soc. 2002, 124, 14724. (c) O. A. Raitman, E. Katz, A. F. Buckmann, I. Willner, J. Am. Chem. Soc. 2002, 124, 6487. (d) O. A. Raitman, F. Patolsky, E. Katz, I. Willner, Chem. Commun. 2002, 1936. 59. (a) D. Hernandez-Santos, M. B. Gonzalez-Garcia, A. C. Garcia, Electroanalysis 2002, 14, 1225. (b) V. S. Murthi, R. C. Urian, S. Mukerjee, J. Phys. Chem. B 2004, 108, 11011. (c) F. Maillard, G. Q. Lu, A. Wieckowski, U. Stimming, J. Phys. Chem. B 2005, 109, 16230. (d) M. C. Daniel, D. Astruc, Chem. Rev. 2004, 104, 293. 60. F. Gloaguen, J. M. Leger, C. Lamy, J. Appl. Electrochem. 1997, 27, 1052. 61. A. M. Yu, Z. J. Liang, J. H. Cho, F. Caruso, Nano Lett. 2003, 3, 1203. 62. T. Y. You, O. Niwa, M. Tomita, S. Hirono, Anal. Chem. 2003, 75, 2080. 63. (a) A. Gamez, D. Richard, P. Gallezot, F. Gloaguen, R. Faure, R. Durand, Electrochim. Acta 1996, 41, 307. (b) O. Antoine, Y. Bultel, R. Durand, J. Electroanal. Chem. 2001, 499, 85. 64. L. Genies, R. Faure, R. Durand, Electrochim. Acta 1998, 44, 1317. 65. (a) J. Maruyama, M. Inaba, Z. Ogumi, Electrochim. Acta 1999, 45, 415. (b) H. F. Cui, J. S. Ye, W. D. Zhang, J. Wang, F. S. Sheu, J. Electroanal. Chem. 2005, 577, 295. (c) X. R. Ye, Y. H. Lin, C. M. Wai, Chem. Commun. 2003, 642. (d) H. C. Ye, R. M. Crooks, J. Am. Chem. Soc. 2005, 127, 4930. 66. C. C. Yang, A. S. Kumar, J. M. Zen, Electroanalysis 2006, 18, 64. 67. K. R. Brown, A. P. Fox, M. J. Natan, J. Am. Chem. Soc. 1996, 118, 1154.
REFERENCES
331
68. X. Gan, T. Liu, J. Zhong, X. J. Liu, G. X. Li, ChemBioChem 2004, 5, 1686. 69. P. A. Loach, Handbook of Biochemistry Selected Data for Molecular Biology, 2nd edition, CRC Press, Cleveland, OH, 1970. 70. H. Zhou, X. Gan, T. Liu, Q. L. Yang, G. X. Li, J. Biochem. Biophys. Methods 2005, 64, 38. 71. M. M. Maye, Y. B. Lou, C. J. Zhong, Langmuir 2000, 16, 7520. (b) M. S. El-Deab, T. Ohsaka, Electrochem. Commun. 2002, 4, 288. 72. C. R. Raj, T. Okajima, T. Ohsaka, J. Electroanal. Chem. 2003, 543, 127. 73. R. S. Nicholson, Anal. Chem. 1965, 37, 1351. 74. R. Polsky, J. C. Harper, D. R. Wheeler, S. M. Dirk, J. A. Rawlings, S. M. Brozik, Chem. Commun. 2007, 2741. 75. S. J. Park, T. A. Taton, C. A. Mirkin, Science 2002, 295, 1503. 76. J. M. Nam, C. S. Thaxton, C. A. Mirkin, Science 2003, 301, 1884.
CHAPTER 12
Biohybrid Electrochemical Devices RAN TEL-VERED, BILHA WILLNER, and ITAMAR WILLNER Institute of Chemistry, The Hebrew University of Jerusalem, Jerusalem, Israel
12.1
INTRODUCTION
Biomolecules include encoded structural and functional information that was optimized by billions of years of chemical evolution. Nucleic acids hybridization, the specific binding of substrates by proteins or other receptors, and the organization of cellular assemblies with the confinement of enzymes, ion channels, or motor proteins, represent a few structural features of biomolecules. Representative structure-emerging functionalities of biomolecules include the specific recognition of substrates and biocatalytic transformations, gene replication and transcription, ion transport through membranes, motoric motility of biomolecules, signal-triggered reactions (e.g., the photosynthetic apparatus, the vision process, or neuronal responses), and more. Tremendous new scientific perspectives were achieved, however, by man-made intervention in natural systems. Genetic engineering of proteins, chemical modification of proteins, the selection of synthetic nucleic acids with specific recognition or catalytic properties, or the artificial integration of biomolecules with polymers, pave new opportunities for the design of new biomolecule-based systems of predesigned properties and functions. The field of supramolecular chemistry has advanced greatly in the past three decades, and the use of tailored supramolecular structures as biomimetic model systems gained substantial research efforts.1 Numerous examples of synthetic receptors,2 artificial enzymes,3 self-replicating systems,4 and more, were reported. The implementation of the structural and functional information encoded in biomolecules to tailor-made supramolecular structures holds, however, great promises as it may combine the extensive knowledge of chemists and physicists in synthesis, materials science, and surface science with the available expertise of biologists to biologically synthesize and engineer biomolecules to yield new hybrid supramolecular structures Electrochemistry of Functional Supramolecular Systems. Edited by Paola Ceroni, Alberto Credi, and Margherita Venturi Copyright 2010 John Wiley & Sons, Inc.
333
334
BIOHYBRID ELECTROCHEMICAL DEVICES
and architectures of designed properties. Indeed, recent studies have reported on the assembly of one-,5 two-,6 and three-dimensional7 supramolecular structures of nucleic acids through the self-organization of monomer oligonucleotides with appropriate base complementarity. Also, supramolecular poly-catenated nucleic acid chains,8 globular nucleic acids that specifically bind molecular substrates,9 and hybrid nanostructures consisting of nucleic acids and proteins10 were synthesized. Similarly, supramolecular nucleic acid structures were used to develop DNA machines,11 such as “tweezers”12 or “walkers”13 and DNA-based machines for amplified biosensing were reported.14 Nanotechnology provides new tools to manipulate surfaces, and materials of unique size-controlled optical, electronic, and catalytic properties. The comparable dimensions of biomolecules and nanoparticles suggest that their conjugation might yield new hybrid materials that combine the unique functions of biomolecules with the properties of nanoparticles. Indeed, the rapidly developing field of nanobiotechnology demonstrated the broad application of biomolecule–nanoparticle hybrid systems, and these new materials found extensive use for sensing, circuitry, and the design of devices.15 This chapter will address the recent advance in the preparation of electrical devices based on supramolecular biomolecular nanostructures. Figure 12.1 outlines the three elements that will be integrated to yield the electrical devices. One element includes an active surface that activates the electrical responses of the system, and it might comprise an electrode or a field-effect transistor. The biomolecules and nanoparticles are, then, integrated with the surfaces to yield the functional nanostructures. Among the emerging functions, as a result of the chemical engineering of the surfaces, the design of electrical biosensor, the synthesis of metallic nanocircuitry, and the fabrication of miniaturized devices will be discussed.
Figure 12.1 Emerging electronic functions of integrated biomolecule–nanoparticle systems assembled on surfaces.
12.2
335
ELECTRICAL CONTACT OF REDOX ENZYMES WITH ELECTRODES
12.2 ELECTRICAL CONTACT OF REDOX ENZYMES WITH ELECTRODES FOR AMPEROMETRIC BIOSENSING AND BIOFUEL CELL DESIGN The electrical contact of redox proteins is one of the most fundamental concepts of bioelectronics. Redox proteins usually lack direct electrical communication with electrodes. This can be explained by the Marcus theory16 that formulates the electron transfer (ET) rate, ket, between a donor–acceptor pair (Eq. 12.1), where do and d are the van der Waals and actual distances separating the donor–acceptor pair, respectively, and DG0 and l correspond to the free energy change and the reorganization enery accompanying the electron transfer process, respectively. "
ðDG0 þ lÞ2 ket / exp½bðddo Þ exp ð4RTlÞ
# ð12:1Þ
As the redox centers in proteins are embedded inside the protein matrices (exhibiting characteristic diameters in the range of 70–200 A), it was concluded that the spatial separation between the redox sites and the electrodes (that may be considered as electron acceptor–donor pairs) prohibits the electrical contact between the proteins and the electrode.17 Substantial efforts were directed in the past two decades to electrically contact redox proteins with electrodes by developing mechanisms to shorten the ET distances between the protein redox centers and the electrode surfaces.18 These included the application of diffusional electron transfer mediators,19 the tethering of redox relays, inside and on the periphery of the protein matrices,20 that enabled charge hopping between the protein redox centers and the electrodes, and the immobilization of the redox enzymes in redox-active polymer hydrogels.21 Besides the fundamental significance in transforming electrically insulated protein matrices into bioelectrocatalytically active biomaterials by chemical means, the electrical “wiring” of enzymes has important practical implications, as it provides the basis for the development of amperometric biosensors22 and biofuel cell devices.23,24 The tailoring of semisynthetic supramolecular biohybrid systems on electrodes introduced new paradigms for the electrical contact of redox enzymes with electrodes. Some recent advances in the electrical wiring of redox proteins with electrodes by tailoring relay-enzyme supramolecular structures on electrodes, by the design of nanoparticle (NP)/or carbon nanotube (CNT)–enzyme hybrids, and by the incorporation of enzymes in redox polymer hydrogels, will be highlighted with representative examples. Specifically, the use of the systems for amperometric biosensing and for the preparation of biofuel cell devices will be discussed. 12.2.1 Electrical Wiring of Redox Enzymes by Relay-Functionalized Monolayer Assemblies The reconstitution of apo-enzymes on relay-cofactor monolayer-functionalized electrodes was used to align redox enzymes and to establish electrical contact
336 Figure 12.2
12.2
ELECTRICAL CONTACT OF REDOX ENZYMES WITH ELECTRODES
337
between the biocatalyst and the electrode. According to this method, the native cofactor is extracted from the enzyme, and the resulting apo-protein is reconstituted on a relay-cofactor monolayer linked to the electrode (Fig. 12.2a). The mediated electron transfer between the redox center and the electrode by the relay unit activates, then, the bioelectrocatalytic function of the biocatalyst. This is exemplified in Fig. 12.2b with the electrical wiring of the flavoenzyme, glucose oxidase, GOx.25 The electron relay unit, pyrroloquinoline quinone (PQQ) (1), was covalently linked to a cysteamine monolayer bound to the Au electrode. The subsequent covalent attachment of 3-amino-phenyl boronic acid, (2), to the PQQ relay unit provided the boronic acid ligand for the “click-on” binding of the flavin adenine dinucleotide (FAD) (3), which yielded the relay cofactor monolayer structure. The reconstitution of apo-GOx on the FAD cofactor site yielded the electrically contacted biocatalyst. Figure 12.2c shows the electrocatalytic anodic currents generated by the electrode at different glucose concentrations, and the resulting calibration curve. Knowing the surface coverage of the enzyme, and the saturated anodic current generated by the electrode, the electron transfer turnover rate between the enzyme redox center and the electrode was estimated to be 700 s1. This high-value turnover rate that is comparable to the electron exchange rate between the redox center of GOx and its native acceptor oxygen (O2), yields oxygen-insensitive bioelectrocatalytic electrodes for the sensing of glucose. The reconstitution process was further implemented to electrically wire GOx on a functional molecular wire through the dynamic shuttling of the electron relay on a molecular wire26 (Fig. 12.2d). The bis-bipyridinium cyclophane (4) was threaded on a molecular wire that was assembled on a gold surface and contained a p-donor-bis-iminobenzene site. The cyclophane was stabilized on the wire by p-donor–acceptor interactions, and the supramolecular structure was stoppered by covalent attachment of aminoethyl flavin adenine dinucleotide (5) at the end
3 Figure 12.2 (a) Reconstitution of an apo-enzyme on a relay-cofactor monolayer for the alignment and electrical wiring of a redox enzyme (glucose oxidase). (b) Assembly of the PQQ/ FAD monolayer on a Au electrode via a boronic acid bridge, and the reconstitution of apo-GOx on the FAD cofactor sites. (c) Cyclic voltammograms corresponding to the bioelectrocatalyzed oxidation of variable concentrations of glucose by the reconstituted glucose oxidase-functionalized electrode according to (b). Glucose concentrations correspond to: (i) 0 mM, (ii) 5 mM, (iii) 10 mM, (iv) 15 mM, (v) 20 mM, (vi) 25 mM, (vii) 35 mM, (viii) 40 mM, and (ix) 50 mM; inset: calibration curve corresponding to the transduced electrocatalytic currents at different concentrations of glucose. (d) The reconstitution of apo-glucose oxidase on a FAD cofactor that “stoppers” the cyclophane (4) on the molecular wire. The redox enzyme is contacted with the electrode by means of the electrochemically shuttled redox unit along the wire. (e) Cyclic voltammograms corresponding to the bioelectrocatalyzed oxidation of different concentrations of glucose by the GOx-reconstituted electrode in the rotaxane structure: (i) 0 mM, (ii) 5 mM, (iii) 10 mM, (iv) 20 mM, (v) 30 mM, (vi) 50 mM, and (vii) 80 mM; inset: calibration curve derived from the cyclic voltammograms at 0.1 V (versus SCE). (b) and (c): Reproduced with permission from Ref. 25. Copyright 2002 American Chemical Society. (d) and (e): Reproduced with permission from Ref. 26. Copyright WileyVCH Verlag GmbH & Co. KGaA.
338
BIOHYBRID ELECTROCHEMICAL DEVICES
of the wire. The reconstitution of apo-GOx on the cofactor site led to the electrically contacted biocatalyst. Electrical wiring of the enzyme and the bioelectrocatalytic activation of GOx were accomplished by the reduction of the bis-cyclophone relay unit by the active site of the enzyme to the respective radical cation species. The reduced cyclophane lacked electron acceptor features, and the removal of the stabilizing interactions with the p-donor site enabled the dynamic shuttling of the mediator on the wire and its oxidation by the electrode. The bioelectrocatalytic anodic currents generated by the electrically contacted enzyme electrode and the derived calibration curves are depicted in Fig. 12.2e. The formation of supramolecular affinity complexes between redox proteins and cofactors on electrode surfaces provides an alternative approach to align enzymes on the electrodes, and to wire them with the surface toward effective ET. For example, cytochrome c (Cyt. C) was immobilized as a monolayer on a Au electrode. A Cyt. C/cytochrome oxidase (COx) affinity complex was, then, assembled on the electrode and was further cross-linked with glutaric dialdehyde to yield an integrated, electrically wired enzyme electrode (Fig. 12.3a).27 In the resulting electrode, the hemoprotein Cyt. C mediated the ET between COx and the electrode, and the wired protein-modified electrode acted as a cathode for the reduction of O2 to H2O (Fig. 12.3b). Similarly, the formation of affinity complexes between an NAD(P) þ cofactor and NAD(P) þ -dependent enzymes on monolayer-functionalized electrodes was used to organize electrically contacted enzyme electrodes25 (Fig. 12.3c). A pyrroloquinoline quinone, PQQ, was assembled as an electroactive monolayer on the electrode. The covalent anchoring of aminophenyl boronic acid to the PQQ site, and the subsequent ligation of the NAD þ cofactor (6) to the boronic acid ligand generated the functionalized surface for the electrical contact of lactate dehydrogenase (LDH). The biocatalyzed oxidation of lactate yields the reduced dihydro nicotinamide adenine dinucleotide (NADH) cofactor that is electrocatalytically oxidized by the PQQ relay, resulting in the wiring and the bioelectrocatalytic activation of the enzyme. This method was extended to tailor integrated, electrically wired NAD(P) þ dependent enzyme electrodes. Also, affinity complexes that align the enzyme in respect to the electrode surface and establish electrical communication between the redox center and the conductive support by means of molecular wires included the substitution of the molecular wire associated with the electrode with an enzyme inhibitor group that binds the active center to the electrode. For example, a thiolated diethylaniline oligo-phenyl acetylene conjugated molecular wire was used to electrically activate amine oxidase.28 The enzyme active center includes a topaquinone/Cu2 þ complex that catalyzes the oxidation of amines to aldehydes. The biocatalyst is inhibited by the diethylamine units, and hence, the binding of the enzyme to the inhibitor group tethered to the conjugated wire (associated with the electrode) resulted in the electrocatalytic activation of the enzyme. Conjugated aromatic polycycles may similarly act as molecular promoters that align and electrically wire redox enzymes with electrodes. For example, an anthracene monolayer was covalently linked to a pyrolytic graphite electrode, and the blue copper oxidase laccase was aligned on the surface, in a configuration that enabled the direct
Figure 12.3 (a) Assembly of the integrated bioelectrocatalytic Cyt.c/COx-electrode. (b) Cyclic voltammograms of the COx/Cyt.c electrode corresponding to the bioelectrocatalyzed reduction of O2 (i), and to the reference system, where O2 is excluded (ii). (c) Assembly of an integrated LDH electrode for the bioelectrocatalyzed oxidation of lactate by the surface crosslinking of an affinity complex formed between LDH and different structures of a boronatelinked PQQ-NAD monolayer. Parts (a) and (b): Reproduced from Ref. 27 by permission of the Royal Society of Chemistry (RSC). Part (c): Reproduced with permission from Ref. 25. Copyright 2002 American Chemical Society.
340
BIOHYBRID ELECTROCHEMICAL DEVICES
Figure 12.4 (a) Schematic configuration of a biofuel cell composed of an electrically contacted anode and an electrically contacted cathode. R1 and R2 correspond to relay units that electrically communicate the redox centers of the respective enzymes, E1 and E2, with the electrodes. (b) Schematic configuration of a noncompartmentalized biofuel cell employing glucose and O2 as fuel and oxidizer, and using PQQ/FAD/GOx and Cyt.c/COx-functionalized Au electrodes as biocatalytic anode and cathode, respectively. (c) Current–voltage behavior of the biofuel cell under different external loads. Inset: the electrical power extracted from the biofuel cell under different external loads. Parts (b) and (c): Reproduced with permission from Ref. 32. Copyright Elsevier, 1999.
electrical contact of the enzyme with the electrode, and the effective bioelectrocatalyzed reduction of O2.29 Besides the broad applications of electrically contacted enzyme electrodes as amperometric biosensors, substantial recent research efforts are directed to the integration of these functional electrodes as biofuel cell devices. The biofuel cell consists of an electrically contacted enzyme electrode acting as anode, where the oxidation of the fuel occurs, and an electrically wired cathode, where the biocatalyzed reduction of the oxidizer proceeds (Fig. 12.4a). The biocatalytic transformations occurring at the anode and the cathode lead to the oxidation of the fuel substrate and the reduction of the oxidizer, with the concomitant generation of a current through the external circuit. Such biofuel cells can, in principle, transform chemical energy stored in biomass into electrical energy. Also, the use
12.2
ELECTRICAL CONTACT OF REDOX ENZYMES WITH ELECTRODES
341
of organ-implantable miniaturized biofuel cells were suggested as potential bioelectronic devices that transform fuel substrates (e.g., glucose in blood) into electrical energy that activates implantable devices such as pacemakers or hearing aids.30 Common fuel substrates are glucose, alcohols, a-hydroxy acids, and others. The preferred oxidizer is O2, although some biofuel cells used hydrogen peroxide (H2O2) as an oxidizer.31 Figure 12.4b depicts the configuration of a noncompartmentalized glucose/O2 biofuel cell that was constructed by the electrical contact of the appropriate biocatalysts using supramolecular reconstitution procedures in monolayer assemblies.32 The anode consisted of an electrically contacted glucose oxidase, GOx, monolayer, generated by the reconstitution of apo-GOx on a pyrroloquinoline quinone-FAD monolayer linked to the electrode. The cathode consisted of an electrically wired COx/Cyt. C protein complex monolayer (cf. Fig. 12.3a) that catalyzed the four-electron reduction of O2 to water. The current/ voltage outputs of the cell at different external loads are shown in Fig. 12.4c, and the maximum power output corresponded to 4 mW at an external load of 0.9 kW. Other biofuel cells based on reconstituted enzyme electrodes were reported. For example, a lactate/O2 biofuel cell that contained an anode consisting of lactate dehydrogenase reconstituted on a pyrroloquinoline quinone-NAD þ monolayer and a COx/Cyt. C construct as cathode was reported.33 A further approach to construct supramolecular bioelectrocatalytic assemblies on electrodes included the layer-by-layer (LBL) deposition of the negatively charged enzyme (glucose oxidase) and the positively charged redox label Os(bisbipyridine) pyridine, Os(bpy)2Py3 þ , (7), tethered to the polyallylamine (PAA) polyelectrolyte, acting as an electron transfer mediating matrix (Fig. 12.5a).34 The bioelectrocatalytic currents generated by the electrode were dominated by the number of enzyme/polyelectrolyte layers linked to the conductive substrate (Fig. 12.5b). 12.2.2 Electrical Wiring of Redox Proteins by Supramolecular Nanoparticle or Carbon Nanotube Hybrid Systems Metallic nanoparticles and single-walled carbon nanotubes (SWCNTs) exhibit nanoscale dimensions comparable with the dimensions of redox proteins. This enables the construction of NP-enzyme or SWCNT-enzyme hybrids that combine the unique conductivity features of the nanoelements with the biocatalytic redox properties of the enzymes, to yield wired bioelectrocatalyts with large electrode surface areas. Indeed, substantial advances in nanobiotechnology were achieved by the integration of redox enzymes with nanoelements and the use of the hybrid systems in different bioelectronic devices.35 Au NPs (1.2 nm) that include a single N-hydroxysuccinimide-active ester functionality were modified with 2-amino-ethyl-flavin adenine dinucleotide, (5), and apoglucose oxidase was reconstituted on the FAD cofactor units to yield the Au NP-GOx hybrid (Fig. 12.6a). The resulting hybrids were linked to the Au surface by different dithiol bridging units (8), (9), and (10). The resulting NP-functionalized glucose oxidase, GOx, exhibited electrical contact with the electrode surface, and the Au NPs
342
BIOHYBRID ELECTROCHEMICAL DEVICES
Figure 12.5 (a) Layer-by-layer deposition of glucose oxidase and the polyallylamine Os3 þ /2 þ -polypyridine polyelectrolyte on the electrode. (b) Typical catalytic current responses for different glucose concentrations obtained by self-assembled nanostructured thin films based on different architectures: (i) PAH/Os/GOx, (ii) cysteamine/GOx/PAH-Os, (iii) PAH/GOx/ PAH-Os, and (iv) (PAH-Os)2/(GOx)1. All measurements were performed in 0.1 M tris buffer at pH 7.5. Part (b): Reproduced with permission from Ref. 34a. Copyright Wiley-VCH Verlag GmbH & Co. KGaA.
12.2
ELECTRICAL CONTACT OF REDOX ENZYMES WITH ELECTRODES
343
Figure 12.6 (a) (i) The assembly of a Au NP (1.4 nm)/electrically contacted glucose oxidase electrode by the reconstitution of apo-GOx on a FAD-functionalized Au NP, and the immobilization of the enzyme/nanoparticle hybrid on an electrode surface. (ii) A STEM image of GOx reconstituted with the Au-FAD hybrid NP. Arrows show Au clusters. (b) Cyclic voltammograms corresponding to the bioelectrocatalyzed oxidation of variable concentrations of glucose by the electrically contacted Au NP-reconstituted GOx-modified electrode. Glucose concentrations correspond to: (i) 0 mM, (ii) 1 mM, (iii) 2 mM, (iv) 5 mM, and (v) 10 mM; inset: calibration curve corresponding to the electrocatalytic currents at different glucose concentrations. (c) Synthesis of a 3D oligoaniline-cross-linked Au NP-reconstituted GOx composite by the electropolymerization of the thioaniline-modified Au NP-reconstituted GOx structure on the thioaniline monolayer-functionalized Au electrode. (d) Cyclic voltammograms corresponding to the bioelectrocatalyzed oxidation of glucose by the oligoaniline-cross-linked Au NPreconstituted GOx composite-modified Au electrode in the presence of variable concentrations of glucose: (i) 0 mM, (ii) 20 mM, (iii) 40 mM, (iv) 60 mM, (v) 80 mM, (vi) 100 , (vii) 120 mM, (viii) 140 mM, (ix) 160 mM, (x) 180 mM, and (xi) 200 mM; scan rate 5 mV/s; inset: calibration curve corresponding to the electrocatalytic currents measured at E ¼ 0.3 V versus SCE for different concentrations of glucose. Parts (a) and (b): Reprinted from Ref. 36 with permission from AAAS.
implanted into the biocatalyst wired the enzyme redox center with the electrode.36 The superior ET communication between the redox center and the electrode was observed with the fully conjugated benzene dithiol molecular bridge (10). Figure 12.6b depicts the bioelectrocatalytic anodic currents generated by the system in the presence of variable concentrations of glucose, and the resulting calibration curve (Fig. 12.6b, inset). From the value of the saturation anodic current, a turnover
344
BIOHYBRID ELECTROCHEMICAL DEVICES
rate of 5000 s1 was estimated. This value should be compared with the turnover rate of 700 s1 corresponding to the enzyme redox center and its native electron acceptor (O2). Thus, the Au NP plugged into the enzyme leads to an unprecedented effective electrical contact between the enzyme redox center and the electrode. This paradigm was implemented to electrically wire other redox enzymes. For example, glucose dehydrogenase (GDH) was electrically contacted with a Au electrode by the reconstitution of apo-glucose dehydrogenase on pyrroloquinoline quinone cofactor units linked to 1.2 nm-sized Au NPs, which were assembled on the electrode using dithiol bridges.37,38 Reconstituted enzymes on metallic NPs were used as functional hybrid structures for the fabrication of electrically contacted enzyme electrodes beyond the monolayer configuration (Fig. 12.6c). Au NPs (or Pt NPs) were functionalized with mercaptophenyl boronic acid, (11), and electropolymerizable thioanline, (12), units. The boronic acid ligands associated with the NPs provided ligation sites for the binding of the flavin adenine dinucleotide, FAD, cofactor. The reconstitution of apo-glucose oxidase on the FAD units resulted in a reconstituted GOx-metallic NPs conjugate. Electropolymerization of the enzyme-NPs hybrids resulted in a multilayer composite consisting of bis-aniline cross-linked metallic NPs that included the reconstituted enzymes as functional units. By the application of 60 electropolymerization cycles, a Au NP-GOx composite that revealed efficient electrical contact with the electrode was generated.38 The bioelectrocatalytic anodic currents generated by the Au NP-enzyme hybrid composite are depicted in Fig. 12.6d, and the extracted calibration curve is shown in Fig. 12.6d, inset. The turnover rate of electrons between the enzyme redox center and the electrode was estimated to be 2250 s1. A further approach to yield electrically contacted enzyme electrodes involved the coelectropolymerization of the metallic NPs and the enzyme on a modified electrode surface, and the construction of a bioelectrocatalytically active hybrid system on the electrode.39 Metallic NPs (Au or Pt) were synthesized with a mixed capping monolayer that consisted of mercaptoethane sulfonic acid as particle stabilizing groups, and thioaniline as electropolymerizable groups. In addition, the redox enzymes were also modified with the thioanline electropolymerizable groups (Fig. 12.7a). The electropolymerization of the modified NPs in the presence of the thioaniline-functionalized glucose oxidase led to a three-dimensional Au NPs/GOx hybrid structure, where the Au NPs provided the three-dimensional conductivity to the matrix, whereas the bis-aniline bridging units provided the electron relay components that wired the enzyme with the electrode. Figure 12.7b depicts the electrocatalytic anodic currents generated by the electrode in the presence of variable concentration of glucose, and the resulting calibration curve. Knowing the saturation current generated by the NP/enzyme hybrid electrode and the loading of the enzyme in the composite structure, the turnover ET rate between the enzyme redox center and the electrode was estimated to be 500 s1. The supramolecular bis-aniline cross-linked metallic NPs/enzyme composite does not only act as a conducting matrix that electrically contacts the redox center with the electrode, but the NPs may also provide catalytic sites that enhance the biocatalytic transformations at the enzyme active site. This has been demonstrated by the effective
12.2
ELECTRICAL CONTACT OF REDOX ENZYMES WITH ELECTRODES
345
Figure 12.7 (a) Modification of glucose oxidase with a polymerizable thioaniline functionality. (b) Cyclic voltammograms corresponding to the bioelectrocatalyzed oxidation of glucose by the oligoaniline-cross-linked GOx/Au NPs composite-modified Au electrode in the presence of variable concentrations of glucose: (i) 0 mM, (ii) 20 mM, (iii) 40 mM, (iv) 60 mM, (v) 80 mM, (vi) 100 mM, (vii) 120 mM, (viii) 140 mM, and (ix) 160 mM; scan rate 5 mV/s; inset: calibration curve corresponding to the electrocatalytic currents measured at E ¼ 0.3 V versus SCE for different concentrations of glucose. (c) Synthesis of the 3D oligoaniline-cross-linked Pt NPs/BOD composite by the electropolymerization of the thioaniline-modified Pt NPs and the thioanilinemodified BOD on the thioaniline monolayer-functionalized Au electrode. (d) (1) Cyclic voltammograms corresponding to the Pt NPs/BOD-modified Au electrode: (i) under Ar and (ii) under O2saturated electrolyte. The electrodes were prepared using 50 electropolymerization cycles and a 1 : 1 molar ratio of Pt NPs:BOD in the electropolymerizable mixture. (2) Cyclic voltammograms corresponding to the Pt NPs-modified Au electrode: (i) under Ar and (ii) under an O2-saturated electrolyte. The electrodes were prepared using 50 electropolymerization cycles. Data recorded in 0.1 M phosphate buffer (pH 7.0) at a scan rate of 5 mV/s. Parts (a) and (b): Reproduced with permission from Ref. 39. Copyright Wiley-VCH Verlag GmbH & Co. KGaA.
bioelectrocatalytic activation of bilirubin oxidase, and the preparation of an effective cathode that reduces O2 to H2O.40 The thioanline-functionalized Pt NPs and the thioanline-modified bilirubin oxidase, BOD, were coelectropolymerized on a roughened Au electrode to yield a Pt NPs/BOD multilayered composite structure (ca. 20
346
BIOHYBRID ELECTROCHEMICAL DEVICES
layers) (Fig. 12.7c). The Pt NPs in the composite structure provided the threedimensional electrical conductivity for the activation of the enzyme toward the reduction of O2 to H2O. The onset potential for the reduction of O2 by the Pt NPs/BOD composite was found to be 0.48 V versus SCE, a value that is ca. 180 mV more positive than the onset potential for the reduction of O2 by the bis-aniline-bridged Pt NPs composite that lacked the BOD (Fig. 12.7d). The enhanced bioelectrocatalytic properties of the Pt NPs/BOD composite were attributed to the cooperative catalytic function of the Pt NPs in the hybrid structure in decomposing the peroxide intermediate formed at the active site of BOD consisting of the Cu–ion cluster. The turnover rate of electrons from the electrode to the BOD active center was estimated to be 36 s1. CNTs provide an alternative conducting nanoelement for the electrical contact of redox proteins. The folding modes of the graphitic layers that form the SWCNTs control their electronic features, and SWCNTs exhibiting semiconductor or ballistic conductivity properties were reported.41,42 The integration of biomolecules and CNTs to form hybrid systems attracted substantial research efforts in recent years,42,43 and different applications of the hybrid nanostructures, such as electrical or optical sensors,44–46 nanocircuitry,47 and nanoscale devices48,49 were investigated. The dimensions of the SWCNTs (diameter ca. 1.8 nm) and their conductivity properties were used to electrically contact redox proteins with electrodes.50,51 The CNTs were subjected to oxidative cleavage by their treatment with an acid, a process that led to the formation of short CNTs with carboxylic acid functionalities at their ends. The resulting shortened CNTs were separated into fractions consisting of CNTs with lengths corresponding to 25–30, 40–50, 80–100, and 200–230 nm, which were used to assemble an electrically wired glucose oxidase electrode by the surface reconstitution process51 (Fig. 12.8a). For this purpose, the CNTs were covalently linked to a cysteamine-monolayer-functionalized Au electrode, and aminoethyl-FAD, (5), was covalently linked to the carboxylic acid residues associated with the free ends of the CNTs. Atomic force microscopy (AFM) images of the surface indicated that the CNTs adopt a perpendicular “standing” configuration in respect to the electrode. This enabled the reconstitution of apo-glucose oxidase on the FAD cofactor sites and the generation of the reconstituted GOx on the electrode surface. AFM images of the reconstituted GOx on the surfaces indicated the formation of globular nanostructures with a height of ca. 70–80 A, consistent with the formation of the reconstituted enzyme nanostructures (Fig. 12.8b). The reconstitution of the enzyme at the ends of the FAD-functionalized CNTs to yield the enzyme/CNT hybrid nanostructure was further confirmed by the modification of the two ends of the CNTs with amino-FAD cofactor units, in solution, and the reconstitution of apo-enzyme on the cofactor sites. AFM images (Fig. 12.8c), and TEM images (Fig. 12.8d), allowed the visualization of the hybrid nanostructures. The glucose oxidase reconstituted on the CNTs revealed electrical contact with the electrode, and the bioelectrocatalytic oxidation of glucose was activated by the hybrid systems. For example, the turnover rate for the bioelectrocatalyzed oxidation of glucose in the presence of the short CNTs as linkers (25–30 nm length) was
12.2
ELECTRICAL CONTACT OF REDOX ENZYMES WITH ELECTRODES
347
Figure 12.8 (a) Assembly of the CNT-electrically contacted GOx electrode. (b) AFM image of GOx reconstituted on the CNTs associated with the Au surface (length of CNTs ca. 50 nm). (c) AFM image of CNTs reconstituted at their ends with GOx units. (d) HRTEM image of a CNT modified at its end with GOx unit (enzyme was stained with uranyl acetate). (e) Cyclic voltammograms corresponding to the bioelectrocatalyzed oxidation of different concentrations of glucose by the GOx-reconstituted CNT-functionalized electrode: (i) 0 mM, (ii) 20 mM, (iii) 60 mM, and (iv) 160 mM; inset: derived calibration curve corresponding to the amperometric responses of the reconstituted electrode at 0.45 V versus SCE in the presence of different concentrations of glucose. Reproduced with permission from Ref. 51. Copyright Wiley-VCH Verlag GmbH & Co. KGaA.
348
BIOHYBRID ELECTROCHEMICAL DEVICES
estimated to be ca. 4200 s1. The electrical wiring efficiency was, however, found to be controlled by the length of the CNTs, and as the CNTs were longer, the bioelectrocatalytic oxidation (or electrical contact) turned to be less efficient (Fig. 12.8e). The dependence of the electrical contact efficiency on the length of the CNTs was attributed to defects introduced into the sidewalls upon their oxidative scission and introduction of the carboxylic acid functionalities to the edges of the tubes. These defect sites are generated by oxidative cleavage of graphitic bonds, and introduce charge transporting barrier along the CNTs. As the probability of occurrence of the defect sites increases with length, the electrical contact efficiency declines with the length of the CNT connectors. The carbon nanotubes also provide a high surface area conductive matrix for the immobilization and electrical contact of redox proteins by their attachment to the graphitic sidewalls.52 Polycyclic aromatic compounds bind strongly to the CNTs via p–p stacking interactions.53 Accordingly, Nile blue (NB þ ) was adsorbed onto the CNTs, and it exhibited a quasi-reversible two-electron redox process, E ¼ 0.35 V at pH 7. The modified CNTs were deposited onto GC electrodes and nicotinamide adenine dinucleotide (NAD þ ) cofactor was covalently tethered to the CNTs through a phenyl boronic acid bridging ligand (2) (Fig. 12.9a). The subsequent generation of an affinity complex between the alcohol dehydrogenase (AlcDH) and the NAD þ cofactor (or between glucose dehydrogenase and the NADP þ cofactor), followed by the cross-linking of the affinity complexes, resulted in the electrically contacted enzyme electrodes. The bioelectrocatalytic activation of alcohol dehydrogenase (or glucose dehydrogenase) was stimulated by the integrated enzyme/CNTs hybrids.52 The biocatalyzed oxidation of ethanol (or glucose) yielded the NAD(P)H cofactors that were oxidized by the NB þ , acting as an electron mediator. Besides the quantitative detection of ethanol (or glucose) by the enzyme-CNTs systems, the electrically wired AlcDH/CNT electrode was implemented as anode in an ethanol/O2 biofuel cell, where the electrically wired bilirubin oxidase (BOD) in the form of bis-aniline-cross-linked BOD/Pt NPs acted as a cathode (Fig. 12.9b). The maximum power output extracted from this biofuel cell corresponded to 200 mW/cm2 at an external load of 7.5 kW.40 A further approach to electrically wire redox enzymes by means of supramolecular structures that include CNTs as conductive elements involved the wrapping of CNTs with water-soluble polymers, for example, polyethylene imine or polyacrylic acid.54 The polymer coating enhanced the solubility of the CNTs in aqueous media, and facilitated the covalent linkage of the enzymes to the functionalized CNTs (Fig. 12.9c). The polyethylene imine-coated CNTs were covalently modified with electroactive ferrocene units, and the enzyme glucose oxidase (GOx) was covalently linked to the polymer coating. The ferrocene relay units were electrically contacted with the electrode by means of the CNTs, and the oxidized relay mediated the electron transfer from the enzyme-active center to the electrode, a process that activated the bioelectrocatalytic functions of GOx. Similar results were observed upon tethering the ferrocene units to polyacrylic acid-coated CNTs, and the covalent attachment of GOx to the modifying polymer.
Figure 12.9 (a) Assembly of an integrated, electrically contacted AlcDH-NAD þ -Nile blueSWCNT electrode. (b) Schematic representation of a biofuel cell employing bioelectrocatalytic electrodes composed of electrically contacted AlcDH-SWCNTs (anode), and oligoanilinecross-linked Pt NPs/BOD (cathode). (c) (1) Assembly of an integrated, electrically contacted, GC-CNTs/PEI-Fc-GOx electrode. (2) Assembly of an integrated, electrically contacted, GCCNTs/PAA-Fc-GOx electrode. Part (a): Reproduced with permission from Ref. 52. Copyright Wiley-VCH Verlag GmbH & Co. KGaA. Part (c): Reproduced with permission from Ref. 54. Copyright 2008 American Chemical Society.
350
BIOHYBRID ELECTROCHEMICAL DEVICES
Figure 12.9
12.2.3
(Continued)
Protein-Based Nanocircuitry and Devices
The self-assembly of proteins provides a means to organize templates for the synthesis of metallic nanowires, or for the fabrication of ordered linear or two-dimensional nanostructures. Aromatic short-chain peptides such as the Alzheimer’s diphenylalanine bamyloid (13) self-assemble into nanotubes andactas nanoreactors for the catalytic growth of silver nanowires55 (Fig. 12.10). The removal of the peptide coating resulted in micrometer-long metallic nanowires exhibiting a diameter of ca. 20 nm. The selfassembly of protein units into active templates for the growth of metallic nanowires, and even patterned metallic nanowires, was demonstrated by the use of actin filaments as nanostructuring templates.56 The G-actin protein monomer units were polymerized in the presence of ATP/Mg2 þ and functionalized with N-hydroxysuccinimide-modified Au NPs (diameter 1.4 nm) (Fig. 12.11a). The resulting Au NP-modified actin filament was dialyzed, and separated, to yield Au NP/G-actin hybrid units. The polymerization of the
Figure 12.10 Formation of a silver nanowire inside a channel of a short-chain diphenylalanine peptide tube.
Figure 12.11 (a) Polymerization of globular actin (G-actin) and modification of the filament with 1.4 nm Au NPs. (b) Atomic force microscope (AFM) image of the Au nanowire on the actin template. (c) AFM image of the hybrid system consisting of the actin filaments tethered to the Au nanowire. Part Adapted with permission from Ref. 56. Copyright Nature Publishing Group, 2004.
352
BIOHYBRID ELECTROCHEMICAL DEVICES
Au NP/g-actin hybrid monomers, in the presence of, ATP/Mg2 þ was followed by the electroless enlargement of the NPs by hydrazine/AuCl4 (Fig. 12.11a, path (1)), and resulted in micrometer-long Au nanowires exhibiting heights and widths in the range of 80–120 nm (Fig. 12.11b). The primary assembly of the Au NP-modified actin filaments by polymerization of the Au NP/g-actin units, followed by the polymerization of unlabeled G-actin units to the þ and ends of the Au NPs-actin filament resulted in a patterned template consisting of a Au NPs-functionalized wire tethered at its end to unlabeled actin filaments (Fig. 12.11a, path (2)). The electroless enlargement of the Au NPs by hydrazine/AuCl4 yielded the Au nanowires flanked between two actin filaments (Fig. 12.11c). The patterned Au-nanowire was used as a nanotransporting device. The actin-units associated with the patterned metallic nanowirewere deposited as filaments on a myosin-monolayer-modified glass surface. In the presence of added adenosine triphosphate (ATP), the motility of the nanowires was observed. Figure 12.12 depicts the fluorescence reflectance images of the nanowires upon addition of the ATP fuel. In this figure, the same area was imaged at time intervals of 5 s. The results suggest that such metal-actin nanostructures may act as nanotransporters for drug delivery, or eventually, the fuel-driven transport of the nanowires may be used to position nanocircuits on miniaturized electronic boards. Enzymes functionalized with metallic NPs were used as biocatalytic hybrids for the growth of metallic nanowires. The catalytic enlargement of metal nanoparticles by products generated by different enzymes was used to develop different optical sensors that follow the activities of enzymes and analyze their substrates.57 For example, hydrogen peroxide generated by the biocatalyzed oxidation of glucose by O2 in the
Figure 12.12 Images of the motility of the actin/Au-wire/actin filaments on a glass surface modified with myosin, upon the addition of ATP. Images were recorded by reflectance microscopy: (a), (b), (c), and (d) correspond to the same imaged frame at time intervals of 5 s. Adapted with permission from Ref. 56. Copyright Nature Publishing Group, 2004.
12.2
ELECTRICAL CONTACT OF REDOX ENZYMES WITH ELECTRODES
353
presence of glucose oxidase, acted as a reducing agent that allowed the reduction of AuCl4 on nanoparticle seeds, and resulted in enlarged Au NPs that monitored the enzyme activity optically and enabled the quantitative analysis of glucose.58 Accordingly, the enzyme glucose oxidase, GOx, was modified with Au NPs (1.4 nm), and the resulting GOx/Au NPs hybrid was deposited on Si surfaces using dip-pen nanolithography (DPN) as a surface nanopatterning tool (Fig. 12.13a). The resulting biocatalytic
Figure 12.13 (a) The generation of a Au nanowire by the biocatalytic enlargement of a Au NP-functionalized GOx line deposited on a silicon support by DPN. (b) AFM image of the Au nanowire generated by the Au NP-functionalized GOx “biocatalytic ink.” Reproduced with permission from Ref. 59. Copyright Wiley-VCH Verlag GmbH & Co. KGaA.
354
BIOHYBRID ELECTROCHEMICAL DEVICES
12.2
ELECTRICAL CONTACT OF REDOX ENZYMES WITH ELECTRODES
355
hybrid was then reacted with glucose in the presence of AuCl4, and the resulting H2O2 enlarged the Au NPs locally on the enzyme template.59 Au wires exhibiting heights and widths of 150–200 nm were generated by this method (Fig. 12.13b). Other oxidases that yield H2O2 were similarly used as “biocatalytic inks” for the deposition of metallic nanowires.59 An alternative biocatalytic ink for the growth of metallic nanowires consisted of alkaline phosphatase modified with Au NPs. This enzyme catalyzes the hydrolysis of p-amino-phenolphosphate, (14), to p-aminophenol, (15). The product reduces Ag þ to silver metal in the presence of Au NPs as catalytic seeds, thus leading to the formation of enlarged core-shell Au/Ag particles (Fig. 12.14a). Accordingly, the Au NP-functionalized alkaline phosphatase was used as “biocatalytic ink” for the DPN deposition of the hybrid on the surface. The subsequent interaction of the pattern template with Ag þ and (14) resulted in the formation of silver nanowires exhibiting heights and widths of 20–25 nm59 (Fig. 12.14b). The growth of the metallic nanowires by means of the enzymes reveals two major advantages for future nanotechnology applications: (i) As the enzyme yields the product that generates the nanowire, its coating by the enlarged particles introduces a self-inhibition mechanism, reflected by transport limitations of the substrates of the enzyme. Consequently, the complete coating of the enzyme blocks the formation of the reducing product, resulting in nanowires with dimensions comparable to the enzyme templates. (ii) The deposition of two (or more) different enzymes allows the orthogonal stepwise deposition of different metals. For example, by the DPN patterning of the Si surface with the Au NPsmodified GOx and the Au NPs-functionalized alkaline phosphatase, the orthogonal synthesis of Au and Ag nanowires was demonstrated (Fig. 12.14c). Specific binding properties of proteins, and specifically, multidentate binding sites in proteins, can be used for the cross-linking of nanoparticle circuits, and the synthesis of hybrid nanostructures. The homotetrameric protein streptavidin (SAv) includes four binding sites for biotin (Ka > 1014 M1). Accordingly, the interaction of biotinylated Au NPs with SAv led to the formation of 3D Au NPs aggregates, which were cross-linked by the protein units. Similarly, the two binding sites of antibodies were used to cross-link Au NPs. For example, 12 nm sized Au NPs were functionalized with the antidinitrophenyl antibody, and a cross-linked Au NP aggregate was formed upon interaction with the bifunctional antigen, bis-(N-2,4-dinitrophenyloctamethylene diamine).60 Systems of enhanced complexity were assembled by combining different protein receptors and NPs. For example, Au NPs networks were prepared by the functionalization of 5 nm sized Au NPs with SAv. Duplex nucleic acids functionalized
3 Figure 12.14 (a) Biocatalytic enlargement of Au NP-modified AlkPh with silver, and the fabrication of silver lines on a support using DPN and the “biocatalytic ink.” (b) AFM image of Ag nanowires generated on the Au NP-AlkPh template deposited on the silicon support by DPN after 40 min of enlargement in the Ag growth solution. (c) An AFM image of Au and Ag nanowires generated by deposition and enlargement of the Au NP-GOx template, followed by the passivation of the Au nanowire with mercaptoundecanoic acid, and the subsequent deposition of Au NP-AlkPh template and its enlargement to the Ag nanowire. Part (a): Reproduced with permission from Ref. 57. Parts (b) and (c): Reproduced with permission from Ref. 59. Copyright Wiley-VCH Verlag GmbH & Co. KGaA.
356
BIOHYBRID ELECTROCHEMICAL DEVICES
Figure 12.15 AFM images of nanoparticle networks prepared by using ds-DNA as spacer groups. The ds-DNA fragments contain two biotin binding sites attached to the two 50 -ends of the ds-DNA that allow cross-linking of the biotin binding protein streptavidin as a model nano-object. Reproduced with permission from Ref. 61. Copyright Wiley-VCH Verlag GmbH & Co. KGaA.
at their termini with biotin units acted as cross-linkers for the Au NPs, resulting in geometrically confined nanostructures61 (Fig. 12.15).
12.3 ELECTRICAL DEVICES BASED ON SUPRAMOLECULAR ASSEMBLIES CONSISTING OF NUCLEIC ACIDS The order of the base sequences in nucleic acids provides substantial structural and functional information. The complementarity of A-T and G-C base pairing provides the guides for hybridization and DNA duplex formation. The resulting DNA duplexes exhibit further encoded information, such as specific cleavage by endonucleases or nicking enzymes, the replication of duplex nucleic acids that include a single-stranded tether, the intercalation of intercalators in-between bases or in minor/major grooves, and more. Also, the base order in single-stranded nucleic acids enables the selfassembly of three-dimensional nucleic acid structures, such as G-quadruplexes. Particularly, interesting approaches to elicit nucleic acids that exhibit specific binding properties toward molecules, ions, or biopolymers (aptamers) and to prepare sequence specific nucleic acids with unique catalytic functions (DNAzymes or ribozymes) were developed.9,62 The route to elicit the nucleic acids that exhibit selective binding and catalytic properties involves the selection and amplification of the desired functional nucleic acid structure by the systematic evolution of ligands by exponential enrichment (SELEX) process from a library of 1015–1016 nucleic acids (Fig. 12.16). According to this procedure, the library of nucleic acids is interacted with a support modified with the substrate to which an aptamer needs to be selected. Minute quantities of nucleic acids exhibiting variable affinities to the substrate (or nonspecifically adsorbed nucleic acids) are, then, associated with the support. The removal of the adsorbed nucleic acids from the support, followed by the polymerase chain reaction (PCR), amplifies the nucleic acids, revealing different degrees of affinities to the target substrate. The subsequent interaction of the resulting nucleic acids with the support (modified with the respective substrate) followed by the removal of the bound
12.3 ELECTRICAL DEVICES BASED ON SUPRAMOLECULAR
Figure 12.16
357
Preparation of an aptamer by the SELEX protocol.
nucleic acids and their PCR amplification results in the selection of nucleic acids of higher affinities to the substrate. By repeating this process, a small number of nucleic acids exhibiting high binding affinity toward the substrate are formed. The separation of the nucleic acids and their sequencing yield well-defined nucleic acid structures with specific binding properties for the substrates (low-molecular-weight substrates or proteins)—aptamers. A similar selection procedure of nucleic acids against transition state analogs, or cofactor units, yields catalytic, enzyme-like nucleic acids (DNAzymes or ribozymes). The tailored recognition and catalytic functions of nucleic acids yield an arena of biomolecules that participate in the formation of supramolecular structures of predesigned functions. For example, the targeted association of lysozyme and thrombin onto micrometer-long nucleic acid wires consisting of alternate antithrombin and antilysozyme aptamer segments demonstrated the dictated self-assembly and precise positioning of two different proteins on a nucleic acid scaffold.63 The intercalation of hemin with a G-quadruplex structure yielded a supramolecular complex that mimicked the biocatalytic functions of horseradish peroxidase.64 Similarly, the formation of the supramolecular complex between cocaine and its aptamer triggered a “DNA machine” that enabled an ultrasensitive detection of cocaine.65 Nucleic acids can be conjugated to nano-objects, such as nanoparticles, nanowires, or nanotubes. The resulting nucleic acids–nano-objects hybrids combine the tailored recognition and catalytic properties of the nucleic acids with the electronic, optical, and catalytic features of the nano-objects. The forthcoming chapter will address the organization of nanoscale supramolecular structures of nucleic acids on
358
BIOHYBRID ELECTROCHEMICAL DEVICES
surfaces, and will highlight various applications of these systems for electrical sensing, synthesis of electrical nanocircuitry, and the directed fabrication of nanoscale electronic devices. 12.3.1 Supramolecular Nucleic Acid Monolayer Structures on Electrodes for Electrical Sensing The intercalation of polycyclic aromatic compounds into duplex DNA structures was used to develop nucleic acid-based electrochemical sensors.66 For example, the bisferrocene-tethered naphthalene diimide (16) was used as a redox-active intercalator to probe DNA hybridization.67 The thiolated probe was assembled on a Au electrode, and the formation of the duplex DNA with the complementary analyte nucleic acid was probed by the intercalation of (16) into the double-stranded nucleic acid structure and by following the voltammetric response of the ferrocene units (Fig. 12.17a). The method enabled the analysis of the target DNA with a sensitivity that corresponded to ca. 1 1020 mol. A different approach for the electrochemical analysis of DNA using monolayer nanostructures included the application of a redox-active hairpin structure, where a ferrocene-tethered hairpin nucleic acid, (17), was assembled on a Au electrode (Fig. 12.17b).68 The analyte nucleic acid is complementary to the single-stranded loop of the hairpin structure, and thus, upon hybridization with the analyte, the hairpin structure is opened and the ferrocene label is extended to a remote position in respect to the electrode. Thus, the efficient electrical contact of the ferrocene units with the electrode in the hairpin configuration is perturbed upon hybridization with the analyte, and the voltammetric responses of the electrode decrease upon increasing the extent of hybridization with the analyte (Fig. 12.17c). This concept was extended to analyze proteins,69 ions,70 or low-molecular-weight substrates.71 A nucleic acid that includes in its base structure the antithrombin aptamer structure and functionalized at its end with a redox label (ferrocene) was assembled on electrodes, and the flexible, remote, positioning of the redox label resulted in an inefficient electrical contact with the electrode (Fig. 12.18a). In the presence of thrombin, the aptamer sequence self-assembles into the G-quadruplex structure, which associates to thrombin. The self-organization of the aptamer–thrombin complex oriented the redox label close to the electrode surface, and effective electrical contact between the redox label and the electrode was attained. As a result, the voltammetric response of the system was intensified as the concentration of thrombin was elevated, allowing the analysis of thrombin with a detection limit that corresponded to 5 109 M. This concept was further applied to detect Pb2 þ ions.70 The analysis of cocainewas exemplified by the supramolecular assembly of an aptamer– cocaine complex on electrode surfaces.72 The anticocaine aptamer sequence was separated into two fragments (I and II) (Fig. 12.18b). Fragment I was assembled on the Au electrode through a thiolated functionality, while the redox label methylene blue (MB þ ) was tethered to the second aptamer fragment. In the presence of cocaine, (18), the supramolecular complex, consisting of the aptamer fragments I and II, and cocaine, was formed on the electrode surface, giving rise to the voltammetric response of the methylene blue redox label (Fig. 12.18c).
12.3 ELECTRICAL DEVICES BASED ON SUPRAMOLECULAR
359
Figure 12.17 (a) Electrochemical analysis of DNA using a bisferrocene-tethered naphthalene diimide (16) as a redox-active intercalator associated with the surface-confined ds-DNA. (b) Electrochemical DNA sensor based on a redox-active label-functionalized DNA hairpin selfassembled monolayer on a Au electrode. (c) Anodic linear sweep voltammograms of the DNA sensor in the presence of different concentrations of the complementary DNA: (i) 0 M, (ii) 30 pM, (iii) 500 pM, (iv) 30 nM, (v) 800 nM, and (vi) 5 mM; the hybridization time interval was 30 min; inset: the calibration curve of the anodic peak currents recorded at different concentrations of the DNA-analyte. Part (a): Adapted with permission from Ref. 67. Copyright 2000 American Chemical Society. Part (c): Reprinted from Ref. 68. Copyright 2003 National Academy of Sciences, USA.
360
BIOHYBRID ELECTROCHEMICAL DEVICES
Figure 12.18 (a) Electrochemical aptasensor for thrombin based on the control of electron transfer between the redox-labeled aptamer and the electrode: controlling the orientation of the redox label in respect to the electrode upon the formation of the thrombin/aptamer complex. (b) Electrochemical analysis of cocaine through the self-assembly of a redox-labeled analyte– aptamer subunits supramolecular complex. (c) Linear sweep voltammograms corresponding to the analysis of variable concentrations of cocaine: (i) 0 M, (ii) 1 105 M, (iii) 1 104 M, (iv) 5 104 M, and (v) 1 103 M. All experiments were performed in the presence of the redox-labeled aptamer, at 1 105 M; inset: calibration curve corresponding to the analysis of cocaine by the self-assembly of the redox-labeled analyte–aptamer subunits supramolecular complex. (d) Blocking the electrical response of methylene blue intercalated into the stem of a DNA hairpin as a result of the formation of the aptamer/thrombin complex. Parts (b) and (c): Reproduced from Ref. 72 by permission of the Royal Society of Chemistry (RSC). Part (d): Reprinted with permission from Ref. 66b. Copyright Wiley-VCH Verlag GmbH & Co. KGaA.
12.3 ELECTRICAL DEVICES BASED ON SUPRAMOLECULAR
361
Figure 12.19 Electrochemical detection of thrombin by the interaction between a nucleic acid and a redox-active oligothiophene polyelectrolyte. The electrical contact between the polyelectrolyte and the electrode by means of the aptamer/thrombin complex is blocked. Parts Reprinted with permission from Ref. 66b. Copyright Wiley-VCH Verlag GmbH & Co. KGaA.
The formation of aptamer–substrate complexes was also followed by the use of redox-active intercalators73 (Fig. 12.18d). A nucleic acid hairpin structure that contained in its single-stranded loop the antithrombin base sequence was assembled on a Au electrode, and methylene blue was intercalated as a redox label in the doublestranded “stem” of the hairpin structure. The hairpin was, then, opened in the presence of thrombin, by generating the respective G-quadruplex-thrombin complex, and as a result, the redox label was removed from the nucleic structure, showing a decrease in the voltammetric response with the increase in the concentration of thrombin. This method enabled the analysis of thrombin with a detection limit that corresponded to 1.1 108 M. Another method for the analysis of aptamer–protein complexes involved the use of a positively charged ferrocene-tethered polythiophene, (19), as redox label reporting unit (Fig. 12.19). The antithrombin aptamer was immobilized on an electrode surface, and the electrostatic binding of the redox polymer (19) to the aptamer monolayer resulted in a supramolecular complex that revealed electrical contact between the polymer and the electrode.74 The formation of the aptamer– thrombin complex removed the polymer from the surface and blocked the electrical contact between the polymer label and the electrode. As a result, higher concentrations of thrombin increased the surface coverage of the aptamer–thrombin complex on the electrode, and this decreased the amperometric responses of the sensing device. Amplified electrochemical detection of DNA in monolayer assemblies was accomplished by the conjugation of bioelectrocatalytic transformations to the DNA recognition events. This was exemplified with the amplified electrochemical analysis of M13 phage DNA (Fig. 12.20a).75 A capturing nucleic acid, (20), complementary to
362
BIOHYBRID ELECTROCHEMICAL DEVICES
Figure 12.20 (a) Amplified electrochemical detection of the viral M13mp18 DNA by the generation of a redox-active replica and the bioelectrocatalyzed oxidation of glucose. The amplified electrochemical analysis of thrombin using (b) the thrombin-catalyzed generation of the electroactive substrate p-nitroaniline and (c) enzyme-tethered antithrombin aptasensor as an amplifying label (GDH). Part (a): Reproduced with permission from Ref. 75. Copyright 2002 American Chemical Society. Parts (b) and (c): Reprinted with permission from Ref. 66b. Copyright Wiley-VCH Verlag GmbH & Co. KGaA.
a segment of the M13 phage DNA was immobilized on a Au electrode. The hybridization with the target M13 phage DNA was followed by the replication of the target in the presence of polymerase and a dNTPs mixture that included ferrocenelabeled dUTP (21). This latter process introduced into the duplex structure ferrocene units that acted as mediators for the bioelectrocatalytic activation of glucose oxidase toward the oxidation of glucose. This process involved the two-step amplification of the recognition process of DNA: the incorporation of numerous electron transfer mediator units as a result of a single duplex formation, and subsequent activation of the biocatalytic reaction, where numerous glucose substrate units are oxidized by the system during a fixed period of time. This route enabled the sensing of M13 phage DNA with a detection limit that corresponded to 1 1013 M. Several electrical aptamer biosensors implemented the biocatalytic hydrolytic activities of thrombin, or the fact that proteins (e.g., thrombin) often include several binding sites for the formation of supramolecular complexes with different aptamers. The bioelectrocatalytic detection of thrombin by an electrical aptasensor was demonstrated by formation of an aptamer–thrombin complex on the electrode, followed by a thrombin-mediated hydrolysis of the nitroaniline-functionalized peptide, (22), yielding the redox-active product nitroaniline, (23), which was analyzed electrochemically76 (Fig. 12.20b). A further bioelectrocatalytic aptasensors configuration is depicted in Fig. 12.20c, where the multidentate formation of aptamer–protein supramolecular complexes was used to analyze thrombin.76 Thrombin includes two different binding sites for aptamers.77 One of the thrombin aptamers (24) was linked to the electrode surface and associated thrombin by the formation of a
12.3 ELECTRICAL DEVICES BASED ON SUPRAMOLECULAR
363
G-quadruplex aptamer–thrombin complex. The association of the second biotinylated aptamer, (25), to the second thrombin binding site generated a three-component “sandwich-type” complex.78 The subsequent binding of biotinylated glucose dehydrogenase, GDH, to the complex by an avidin–biotin interaction led to the formation of the biocatalyst-labeled recognition complex. The electrical wiring of GDH with the diffusional 1-methoxyphenazine methosulfate (m-PMS) resulted in the amplified amperometric transduction of thrombin analysis. This method enabled the amperometric detection of thrombin with a linear response in the range of 4 to 10 108 M. 12.3.2 Supramolecular Nucleic Acid/Nanoparticle Hybrid Structures for Electrical Sensing The unique optical, electrical, and catalytic properties of metal or semiconductor quantum dots (QDs) were widely implemented in the development of nucleic acidbased sensors. Different nanoparticle-based nucleic acid optical sensors were developed, and these included the aggregation of Au NPs,79 the application of semiconductor QDs,80 or the surface plasmon resonance amplified detection of DNA hybridization.81 In this section we address the design and application of NP-nucleic acid hybrid nanostructures for electrical sensing. The electrocatalytic properties of metals were used to develop electrochemical DNA sensors or aptasensors. Pt NPs electrocatalyzed the reduction of H2O2, and the NPs were used as amplifying electochemical labels for the detection of DNA82 (Fig. 12.21a). The Au electrode was modified with the capture nucleic acid, (26), that hybridized with the analyte DNA (27). The subsequent hybridization of the nucleic acid (28)-modified Pt NPs with the single-stranded domain of the analyte, enabled, then, the electrocatalyzed reduction of H2O2. This method allowed the detection of the target DNA with a detection limit that corresponded to 1 1011 M. A related concept was applied to develop an electrochemical aptasensor (Fig. 12.21b). The antithrombin aptamer (29) was immobilized on the electrode, and Pt NPs were functionalized with a second antithrombin aptamer (30). The association of the Pt-modified aptamer to the surfaceconfined aptamer–thrombin complex yielded the supramolecular electrocatalytic hybrid nanostructure that enabled the amperometric analysis of thrombin with a detection limit that corresponded to 1 109 M. The composition of core-shell structures was extensively used to develop biosensors, and specifically, DNA-sensor systems.83 The amplified electrical detection of DNA by the catalytic enlargement of Au NPs and using conductivity as a readout signal was accomplished by the generation of conductivity paths in between microelectrodes through a catalytic process84 (Fig. 12.22). The probe nucleic acid (31) was immobilized in between two microelectrodes, and the target DNA, (32), was hybridized with the surface and further reacted with the nucleic acid (33)-functionalized Au NPs. The catalytic enlargement of the Au NP by Ag þ with hydroquinone yielded conductivity routes between the microelectrodes. The gap revealed high resistance, R > 2 108 W, prior to the enlargement process, and it dropped down to R ¼ 100 W, after 25 min of silver deposition. The method proved to be successful in detecting single-base mismatches in DNA. For example, exchange of the A-base in the capture nucleic
364
BIOHYBRID ELECTROCHEMICAL DEVICES
Figure 12.21 (a) The amplified electrochemical analysis of a DNA by nucleic acid-functionalized Pt NPs acting as electrocatalysts for the reduction of H2O2. (b) (1) Electrochemical analysis of thrombin by aptamer-labeled Pt NPs acting as electrocatalysts for the reduction of H2O2. (2) Voltammograms corresponding to the analysis of different concentrations of thrombin: (i) 1 106 M, (ii) 1 107 M, (iii) 1 108 M, (iv) 1 109 M, and (v) control sample in the absence of thrombin. Reproduced with permission from Ref. 82. Copyright 2006 American Chemical Society.
acid with a G-base enabled the temperature-controlled elimination of the hybridization of (32) due to a lower melting temperature. As a result, the insulating features of the gap were preserved. The conductivity of the gap was controlled by the concentration of the analyte DNA, since the content of the catalytic NPs that provided the origin of the conductivity paths was dominated by the surface coverage of (32). This technology to detect DNA has already been practically implemented, and DNA chips that contain electrode arrays for multiplexed analyses of DNAs, and conductivity analyzers that
12.3 ELECTRICAL DEVICES BASED ON SUPRAMOLECULAR
365
Figure 12.22 DNA analysis by the enhancement of the conductivity between electrodes due to the catalytic enlargement of Au nanoparticles associated with the DNA assembly.
probe the NPs-enlarged microelectrode gaps were fabricated.85 The method was further developed86 by employing Al microelectrodes on Al2O3 supports as the transducing elements. The functionalization of the oxide surface with an aminefunctionalized nucleic acid, followed by the hybridization of a biotin-modified nucleic acid to the surface-linked capture nucleic acid/analyte duplex resulted in the labeled conjugate. The subsequent binding of the conjugate, consisting of Au NPs linked to antibiotin Ab and the DNA complex, and the catalytic deposition of Ago on the Au NPs yield the conductive routes that traced the primary capturing of the analyte. Metal NPs, aggregated in the presence of the target DNA, were used as a conductive threedimensional structure for the amplified electrochemical analysis of DNA by the intercalation of a redox-active intercalator into the double-stranded matrix of the aggregated NPs87 (Fig. 12.23a). Two kinds of Au NPs functionalized with the nucleic acids (34 and 35) that are complementary to the 30 and 50 ends of the target DNA were reacted with the analyte (36). The hybridization resulted in the aggregation of the NPs. The collection of the Au NPs aggregated on a dithiol-functionalized electrode, followed by the intercalation of methylene blue, provided the electrochemical path for the voltammetric detection of (36) (Fig. 12.23b). The Au NP aggregates provided a three-dimensional structure for the electrical contact of numerous redox-active intercalator units with the electrode. The method enabled the detection of DNA with a detection limit that corresponded to 1 1013 M. Metal NPs were also used as labels to follow aptamer–substrate interactions. The supramolecular self-organization of the aptamer–substrate complexes on surfaces was implemented to develop different configurations of electrical aptasensors.88 The anticocaine aptamer was separated into two fragments (37 and 38) (Fig. 12.24a). The nucleic acid (37) was assembled on a Au electrode, whereas the nucleic acid (38) was
366
BIOHYBRID ELECTROCHEMICAL DEVICES
Figure 12.23 (a) Electrochemical detection of DNA through the aggregation of Au NPs and the voltammetric response of the redox-active MB þ intercalator in the duplex-aggregated NPs. (b) Differential pulse voltammograms corresponding to the analysis of different concentrations of (36): (i) 0 M, (ii) 1 1013 M, (iii) 1 1012 M, (iv) 1 1011 M, (v) 1 1010 M, and (vi) 1 109 M; inset: resulting calibration curve. In all experiments, the concentration of the probes (34)- and (35)-functionalized Au NPs was 6 109 M and the aggregation time interval prior to the deposition of the aggregates on the electrode surfaces was 20 min. Reproduced from Ref. 87 by permission of the Royal Society of Chemistry (RSC).
12.3 ELECTRICAL DEVICES BASED ON SUPRAMOLECULAR
367
Figure 12.24 (a) Amplified impedimetric analysis of cocaine by a supramolecular complex generated by anticocaine aptamer fragments, (37 and 38), using Au NPs-functionalized with one of the aptamer fragments, (38), as an amplifying label. (b) Faradaic impedance spectra (in the form of Nyquist plots) corresponding to the analysis of different concentrations of cocaine using the modified Au electrode and the (38)-functionalized Au NPs: (i) 0 M, (ii) 1 105 M, (iii) 1 104 M, (iv) 5 104 M, and (v) 1 103 M. The spectra were measured in the presence of 1 mM Fe(CN)63/4, at E ¼ 0.2 V versus saturated calomel electrode, and with an AC potential modulation amplitude of 10 mV. (c) Label-free, reagentless analysis of cocaine on an ISFET device through the formation of the cocaine-aptamer subunits complex on the gate surface. (d) Changes in the gate-to-source potentials upon the analysis of cocaine using (i) the aptamer subunit-functionalized Au nanoparticle as amplifying labels. (ii) The free aptamer subunit. Reproduced with permission from Ref. 88. Copyright Wiley-VCH Verlag GmbH & Co. KGaA.
used to modify Au NPs. While the (37)-monolayer electrode did not interact with the modified Au NPs, the addition of cocaine induced the formation of a tri-component supramolecular complex consisting of the aptamer fragments and the cocaine substrate. The formation of the complex on the surface resulted in a highly negatively charged interface composed of the nucleic acid-modified Au NPs. This process was, then, followed by impedimetric measurements using Fe(CN)63/4 as a redox label. As the surface coverage of the electrode was higher, the interfacial electron resistances increased88 (Fig. 12.24b). The detection of the tri-component aptamer–cocaine complex was, also, achieved by the use of ion-sensitive field-effect transistor (ISFET) transducers (Fig. 12.24c). The formation of the Au nanoparticle-modified aptamer– cocaine complexes on the gate surface of the ISFET device yielded a negative charge on the gate surface, thus altering its potential.88 As a result, the changes in the gate-tosource potentials, DVGS, to attain the constant current through the FET device, ISD, provided a quantitative measure for the recognition of events occurring on the gate (Fig. 12.24d).
368
12.3.3
BIOHYBRID ELECTROCHEMICAL DEVICES
DNA-Based Nanocircuitry and Devices
Nucleic acids (DNA) reveal unique properties that turn them into attractive biomolecular templates for the “bottom–up” construction of nanoengineered structures: (i) The base sequence in nucleic acids provides encoded information for specific hybridization, intercalation, or biocatalytic transformations (e.g., specific cleavage by endonucleases). (ii) Well-developed synthetic methodologies for the preparation of nucleic acids exist, including methods to incorporate artificial bases, or to tether to the DNA chemical functionalities that allow secondary modification reactions or conjugation processes. Methods to generate DNA structures of any shape (circular, square, or triangle) are available, and even selfassembly methods of DNA networks of long-scale periodicity are well established.89 (iii) A variety of biocatalysts that provide biomolecular nanotools to manipulate DNA are available. These include scission or nicking enzymes, replication (polymerase), ligation (ligase), and more. (iv) The negatively charged phosphate groups associated with the DNA can be used as anchoring sites for different metal ions, for example, Ag þ , Pd2 þ , and Cu2 þ . Similarly, the intercalation of molecular components into base pairs of duplex DNA enables the targeting of functional units to the DNA templates by means of the intercalators. The ionic or molecular coadducts allow the secondary modification of the DNA templates by chemical means. (v) Different proteins bind to sequence-specific domains of the DNA. This enables the selective insulation of duplex DNA domains, and thus, the patterning of the DNA using chemical methods. Indeed, substantial progress in the application of DNA as a template for the “bottom–up” construction of metallic NPs or nanowire circuitry, and the assembly of nanoscale miniaturized devices, was demonstrated in the past decade.90 A duplex DNA acted as a template for the synthesis of Au NPs wires using intercalator-modified Au NPs. The duplex biopolymer consisting of polyA/polyT, (39), was reacted with psoralen, (40)-functionalized Au NPs. The intercalation of the psoralen units into the A-T base pairs, followed by the photochemical cross-linking of the psoralen units to the thymine bases by a 2p þ 2p cycloaddition, resulted in the fixation of the NPs on the DNA template. The subsequent electroless enlargement of the NPs in the presence of NH2OH/AuCl4 resulted in Au nanowires exhibiting heights that range between 3 and 8 nm (Fig. 12.25a and b).91 The synthesis of conductive metallic nanowires that bridged two microelectrodes separated by a gap of 12–16 mm was demonstrated by the growth of a silver nanowire on a DNA template that bridged the gap (Fig. 12.26).92 Short thiolated nucleic acids (12 bases long) were attached to the microelectrodes, and these acted as “sticky” ends for the hybridization of l-DNA that bridged the gap. The association of Ag þ to the phosphate groups of the template, followed by their reduction with hydroquinone under basic conditions, resulted in the formation of Ago nanoclusters on the DNA template. The subsequent enlargement of the Ago seeds by the catalytic reduction of Ag þ by hydroquinone, under acidic conditions, yielded continuous Ag nanowires with a width of ca. 100 nm. The nanowires revealed nonlinear I–V
12.3 ELECTRICAL DEVICES BASED ON SUPRAMOLECULAR
369
Figure 12.25 (a) Assembly of a Au nanoparticles wire in the polyA/polyT template using Au nanoparticles functionalized with intercalator (psoralen) molecules. (b) AFM image of the Au nanoparticles wire in the polyA/polyT template. Reproduced with permission from Ref. 91. Copyright Wiley-VCH Verlag GmbH & Co. KGaA.
curves with a characteristic hysteresis. Although the charge transport mechanism through these nanowires is not fully understood, it was suggested that quantum effects or defects in the nanowires structure may lead to the observed conductivity features.92 The “bottom–up” synthesis of metallic nanowires was further applied to construct a nanotransistor device.93 The sequence-specific winding of the homologous nucleic acid carried by the RecA-protein into the duplex DNAwas used to address the nucleic acid/protein complex on the DNA scaffold (Fig. 12.27). The subsequent association of the anti-RecA antibody to the protein DNA complex, followed by the association of the biotinylated antiantibody, and the linkage of streptavidin-modified carbon nanotube deposited the tubes in the specific domain of the DNA scaffold. The further
370
BIOHYBRID ELECTROCHEMICAL DEVICES
Figure 12.26 Construction of a nanowire bridging two microelectrodes by the deposition of Ag þ ions on a bridging DNA strand, followed by their chemical reduction to the metallic agglomerate.
reduction of Ag þ ions exchanged on the free nucleic acid segments yielded catalytic Ago nanoclusters, and these were enlarged by the electrochemical deposition of Auo in the presence of AuCl4/SCN/hydroquinone to yield electrical contacts bridged by a carbon nanotube. The resulting nanoscale composite structures were, then, deposited on a Si substrate, where the gold contacts bridged by the CNT acted as a nanoscale field-effect transistor. The electron flow between the metallic nanocontacts, acting as source and drain electrodes, was controlled by the gate potential.
Figure 12.27
The construction of a DNA-templated SWCNT FET.
12.4
12.4
CONCLUSIONS AND PERSPECTIVES
371
CONCLUSIONS AND PERSPECTIVES
The significance of noncovalent interactions in biology led to the development of the fields of supramolecular chemistry and biomimetic chemistry. Ingenious supramolecular systems mimicking biological receptors, enzymes, ion channels and photosynthetic model systems were developed during the years. Also, principles of biological motor systems were duplicated by man-made molecular machines, and exciting applications of such systems for information storage and processing, slow release of molecules, vectorial transport of substrates, and more, were suggested. Also, the rapidly developing field of nanotechnology provides new materials in the form of nanoparticles, nanotubes, or nanowires that exhibit unique electronic, optical, or catalytic size-controlled properties. The conjugation of molecular functionalities to these nano-objects generated a new class of hybrid materials that enabled the formation of functional supramolecular structures by noncovalent interparticle interactions, or by the formation of supramolecular complexes on core nanoparticles or nanotubes. Such hybrid systems combine the functionalities of the supramolecular complexes with the unique electronic and optical properties of the nano-objects, and consequently, the use of systems such as sensors, nanoconductors, nanocontainers, or nanodevices was suggested. The tremendous progress in supramolecular chemistry and nanoscience provided intellectual concepts and guidelines to implement biomolecules and nano-objects as functional units for the self-assembly of biomolecular structures, or biomolecule– nanoparticle hybrid systems. Such biomolecular supramolecular complexes or hybrid biomolecular composites are anticipated to reveal properties and functions that emerge from the complexity of the structures. Biomolecules include encoded structural information that dictates their recognition properties and catalytic functions. Chemists have learnt to modify biomolecules with electroactive, photoactive, and surface linkable functionalities. Similarly, the site-specific modification of proteins by genetic engineering, or the synthesis of modified nucleic acids or nucleic acid analogs, establish a broad spectrum of new biomaterial-based molecules that provide the building blocks for biomolecular supramolecular science and nanobiotechnology. Indeed, substantial progress in these fields was accomplished in recent years. Man-made nucleic structures with specific recognition and catalytic properties were elicited (i.e., aptamer or DNAzyme) and their supramolecular complexes with low-molecular-weight substrates, ions, or proteins were used to develop sensors, biomaterial-based logic gates, and selforganized circuitry systems. Similarly, the area of nanobiotechnology is rapidly expanding; biomolecule–metal nanoparticle conjugates, biomolecule–semiconductor quantum dot (QD) hybrids, biomolecule–magnetic nanoparticle systems, and biomolecule–carbon nanotube conjugated structures were synthesized, and the composites were used to develop sensors for the in vivo imaging of cells, for nanomedicine, and for the fabrication of nanoscale devices. This chapter has focused on the use of biomolecular supramolecular structures and biomolecule–nanoparticle (or carbon nanotubes) composites as functional units for the construction of electrical devices.
372
BIOHYBRID ELECTROCHEMICAL DEVICES
Methods to electrically wire redox proteins with electrodes by the reconstitution of apo-proteins on relay-cofactor units were discussed. Similarly, the application of conductive nanoelements, such as metallic nanoparticles or carbon nanotubes, provided an effective means to communicate the redox centers of proteins with electrodes, and to electrically activate their biocatalytic functions. These fundamental paradigms for the electrical contact of redox enzymes with electrodes were used to develop amperometric sensors and biofuel cells as bioelectronic devices. Albeit the substantial progress in the bioelectrochemical activation of enzymes, one could identify two important future challenges in the field: (i) The active relay units wiring the redox centers of the enzymes with the electrodes could be generated by photoinduced electron transfer. This could pave the way to the photochemical wiring of enzymes and to the development of photobiofuel cells. (ii) DNA scaffolds provide unique templates for the ordered self-assembly of molecular or biomolecular units through dictated hybridization. The ordering of relay units and enzymes, or of relay units photosystems, on DNA templates associated with electrodes may yield attractive new supramolecular nanostructures for bioelectronics and optobioelectronics. We further addressed the use of the nucleic acids as biopolymers for the formation of supramolecular structures that enable the electronic or electrochemical detection of DNA. Specifically, we discussed the use of aptamer/low-molecular-weight molecules or aptamer/protein supramolecular complexes for the electrical analysis of the guest substrates in these complexes. Also, nucleic acid-NPs hybrid systems hold a great promise as sensing matrices for the electrical detection of DNA in composite three-dimensional assemblies. While sensitive and selective electrochemical sensors for DNA were fabricated, the integration of these sensor configurations in array formats (DNA chips) for the multiplexed analysis of many DNAs can also be envisaged. Finally, we discussed the possible application of biomolecules (proteins or DNA) as templates for the generation of electrically conducting nanowires for future nanoelectronics. Albeit substantial progress was achieved in this research field, the topic is yet at its infancy, and challenging goals are ahead of us. While most of the nanocircuitry systems consist of metallic nanowires, the future fabrication of semiconductor nanowires, and, even, insulating inorganic nanowires, on biomolecule templates will be essential to produce operating nanoelectronic platforms. Thus, the synthesis of patterned metallic/semiconductor/insulating nanowire is essential for the development of the field. A major drawback in the area of nanoelectronics involves the difficulties to precisely position and wire the nanostructures into operating devices. While DPN might be an appropriate approach to position these nanostructures, further methods for the parallel self-assembly of nanocircuits are essential. Furthermore, as conductivity measurements on metallic nanowires revealed nonlinear ohmic dependence, a phenomenon that was attributed to local defects in the nanostructures that perturb the charge transport, it seems that theoretical studies exploring the charge transport through such nanowires are important future challenges.
REFERENCES
373
The organization of supramolecular structures of biomolecules for electrochemical and electronic applications proved to be an exciting research area with numerous applications in bioelectronics and electrochemical sensing, energy conversion and storage (biofuel cells and batteries), bioelectrocatalysis, and nanoelectronics. We may anticipate, however, that interdisciplinary research efforts of chemists, physicists, biologists, materials scientists, and electrical engineers will lead to exciting new results in the field. REFERENCES 1. J.-M. Lehn, Supramolecular Chemistry: Concepts and Perspectives, VCH, Weinheim, 1995. 2. (a) J. Franke, F. V€ogtle, Top. Curr. Chem. 1986, 132, 135–170. (b) A. P. Davis, Chem. Soc. Rev. 1993, 22, 243–253. (c) F. V€ogtle, H. Sieger, W. M. M€ uller, Top. Curr. Chem. 1981, 98, 107–161. (d) J.-M. Lehn, Angew. Chem., Int. Ed. 1988, 27, 89–112. (e) J.-M. Lehn, Angew. Chem., Int. Ed. 1990, 29, 1304–1319. 3. (a) Y. Murakami, Top. Curr. Chem. 1983, 115, 107–155. (b) I. Tabushi, Acc. Chem. Res. 1982, 15, 66–72. 4. (a) M. M. Conn, J. Rebek, Curr. Opin. Struct. Biol. 1994, 4, 629–635. (b) E. A. Wintner, M. M. Conn, J. Rebek, Acc. Chem. Res. 1994, 27, 198–203. 5. (a) N. C. Seeman, J. Theor. Biol. 1982, 99, 237–247. (b) Y. Liu, C. Lin, H. Li, H. Yan, Angew. Chem., Int. Ed. 2005, 44, 4333–4338. 6. P. W. K. Rothemund, Nature 2006, 440, 297–302. 7. (a) A. Kuzuya, R. Wang, R. Sha, N. C. Seeman, Nano Lett. 2007, 7, 1757–1763. (b) C. Mao, W. Sun, N. C. Seeman, Nature 1997, 386, 137–138. 8. Y. Weizmann, A. B. Braunschweig, O. I. Wilner, Z. Cheglakov, I. Willner, Proc. Natl. Acad. Sci. USA 2008, 105, 5289–5294. 9. (a) A. D. Ellington, J. W. Szostak, Nature 1990, 346, 818–822. (b) M. Famulok, G. Mayer, M. Blind, Acc. Chem. Res. 2000, 33, 591–599. 10. Y. Weizmann, A. B. Braunschweig, O. I. Wilner, Z. Cheglakov, I. Willner, Chem. Commun. 2008, 4888–4890. 11. (a) M. K. Beissenhirtz, I. Willner, Org. Biomol. Chem. 2006, 4, 3392–3401. (b) J. Bath, A. J. Turberfield, Nat. Nanotechnol. 2007, 2, 275–284. (c) F. C. Simmel, B. Yurke, Phys. Rev. E 63, 041913 2001, 1–5. 12. (a) B. Yurke, A. J. Turberfield, A. P. MillsJr., F. C. Simmel, J. L. Neumann, Nature 2000, 406, 605–608. (b) Y. Chen, M. Wang, C. Mao, Angew. Chem., Int. Ed. 2004, 43, 3554–3557 (c) J. Elbaz, M. Moshe, I. Willner, Angew. Chem., Int. Ed. 2009, 48, 3834–3837. 13. (a) J.-S. Shin, N. A. Pierce, J. Am. Chem. Soc. 2004, 126, 10834–10835. (b) Y. Tian, Y. He, Y. Chen, P. Yin, C. Mao, Angew. Chem., Int. Ed. 2005, 44, 4355–4358. 14. (a) Y. Weizmann, M. K. Beissenhirtz, Z. Cheglakov, R. Nowarski, M. Kotler, I. Willner, Angew. Chem., Int. Ed. 2006, 45, 7384–7388. (b) D. Li, A. Wieckowska, I. Willner, Angew. Chem., Int. Ed. 2008, 47, 3927–3931. 15. (a) I. Willner, M. Zayats, B. Willner, Biosens. Bioelectron. 2007, 22, 1841–1852. (b) R. Gill, M. Zayats, I. Willner, Angew. Chem., Int. Ed. 2008, 47, 7602–7625.
374
BIOHYBRID ELECTROCHEMICAL DEVICES
16. R. A. Marcus, N. Sutin, Biochim. Biophys. Acta 1985, 811, 265–322. 17. (a) I. Willner, E. Katz, Angew. Chem., Int. Ed. 2000, 39, 1180–1218. (b) A. Heller, J. Phys. Chem. 1992, 96, 3579–3587. (c) A. Heller, Acc. Chem. Res. 1990, 23, 128–134. 18. M. Zayats, B. Willner, I. Willner, Electroanalysis 2008, 20, 583–601. 19. (a) P. N. Bartlett, P. Tebbutt, R. G. Whitaker, Prog. React. Kinet. 1991, 16, 55–155. (b) T. Matsue, N. Kasai, M. Narumi, M. Nishizawa, H. Yamada, I. Uchida, J. Electroanal. Chem. 1991, 300, 111–118. 20. (a) W. Schuhmann, T. J. Ohara, H. -L. Schmidt, A. Heller, J. Am. Chem. Soc. 1991, 113, 1394–1397. (b) Y. Degani, A. Heller, J. Am. Chem. Soc. 1988, 110, 2615–2620. (c) W. Schuhmann, Biosens. Bioelectron. 1995, 10, 181–193. (d) I. Willner, N. Lapidot, A. Riklin, R. Kasher, E. Zahavy, E. Katz, J. Am. Chem. Soc. 1994, 116, 1428–1441. 21. (a) R. Maidan, A. Heller, Anal. Chem. 1992, 64, 2889–2896. (b) B. A. Gregg, A. Heller, J. Phys. Chem. 1991, 95, 5970–5975. 22. (a) J. Wang, Chem. Rev. 2008, 108, 814–825. (b) I. Willner, B. Willner, Trends Biotechnol. 2001, 19, 222–230. (c) B. Willner, E. Katz, I. Willner, Curr. Opin. Biotechnol. 2006, 17, 589–596. 23. (a) S. C. Barton, J. Gallaway, P. Atanassov, Chem. Rev. 2004, 104, 4867–4886 (b) I. Willner, Y.-M. Yan, B. Willner, R. Tel-Vered, Fuel Cells, 2009, 9, 7–24. 24. (a) R. A. Bullen, T. C. Arnot, J. B. Lakeman, F. C. Walsh, Biosens. Bioelectron. 2006, 21, 2015–2045. (b) A. Heller, Phys. Chem. Chem. Phys. 2004, 6, 209–216. 25. M. Zayats, E. Katz, I. Willner, J. Am. Chem. Soc. 2002, 124, 14724–14735. 26. E. Katz, L. Sheeney-Haj-Ichia, I. Willner, Angew. Chem., Int. Ed. 2004, 43, 3292–3300. 27. V. Pardo-Yissar, E. Katz, I. Willner, A. B. Kotlyar, C. Sanders, H. Lill, Faraday Discuss. 2000, 116, 119–134. 28. C. R. Hess, G. A. Juda, D. M. Dooley, R. N. Amii, M. G. Hill, J. R. Winkler, H. B. Gray, J. Am. Chem. Soc. 2003, 125, 7156–7157. 29. C. F. Blanford, R. S. Heath, F. A. Armstrong, Chem. Commun. 2007, 1710–1712. 30. I. Willner, Science 2002, 298, 2407–2408. 31. I. Willner, G. Arad, E. Katz, Bioelect. Bioenerg. 1998, 44, 209–214. 32. E. Katz, I. Willner, A. B. Kotlyar, J. Electroanal. Chem. 1999, 479, 64–68. 33. E. Katz, A.F. B€ uckmann, I. Willner, J. Am. Chem. Soc., 2001, 123, 10752–10753. 34. (a) E. J. Calvo, A. Wolosiuk, ChemPhysChem 2004, 5, 235–239. (b) V. Flexer, E. S. Forzani, E. J. Calvo, Anal. Chem. 2006, 78, 399–407. 35. (a) E. Katz, I. Willner, Angew. Chem., Int. Ed. 2004, 43, 6042–6108. (b) R. Baron, B. Willner, I. Willner, Chem. Commun. 2007, 323–332. (c) I. Willner, B. Basnar, B. Willner, FEBS J. 2007, 274, 302–309. 36. Y. Xiao, F. Patolsky, E. Katz, J. F. Hainfeld, I. Willner, Science 2003, 299, 1877–1881. 37. M. Zayats, E. Katz, R. Baron, I. Willner, J. Am. Chem. Soc. 2005, 127, 12400–12406. 38. B. Willner, E. Katz, I. Willner, Curr. Upin. Biotechnol. 2006, 17, 589–596. 39. O. Yehezkeli, Y. Yan, I. Baravik, R. Tel-Vered, I. Willner, Chem. Eur. J. 2009, 15, 2674–2679. 40. Y. Yan, I. Baravik, R. Tel-Vered, I. Willner, Adv. Mater. 10.1002/adma.200900206. 41. J. E. Fischer, H. Dai, A. Thess, R. Lee, N. M. Hanjani, D. L. Dehaas, R. E. Smalley, Phys. Rev. B 1997, 55, R4921–R4924.
REFERENCES
375
42. E. Katz, I. Willner, ChemPhysChem 2004, 5, 1084–1104. 43. B. L. Allen, P. D. Kichambare, A. Star, Adv. Mater. 2007, 19, 1439. 44. J. J. Davis, K. S. Coleman, B. R. Azamian, C. B. Bagshaw, M. L. H. Green, Chem. Eur. J. 2003, 9, 3732–3739. 45. J. Wang, G. Liu, M. R. Jan, J. Am. Chem. Soc. 2004, 126, 3010–3011. 46. (a) K. Besteman, J. O. Lee, F. G. M. Wiertz, H. A. Heering, C. Dekker, Nano Lett. 2003, 3, 727–730. (b) E. S. Jeng, A. E. Moll, A. C. Roy, J. B. Gastala, M. S. Strano, Nano Lett. 2006, 6, 371–375. (c) P. W. Barone, M. S. Strano, Angew. Chem., Int. Ed. 2006, 45, 8138– 8141. 47. P. Avouris, Acc. Chem. Res. 2002, 35, 1026–1034. 48. (a) A. Star, J. -C. P. Gabriel, K. Bradley, G. GrQner, Nano Lett. 2003, 3, 459–463. (b) K. Bradley, M. Briman, A. Star, G. GrQner, Nano Lett. 2004, 4, 253–256. 49. R. J. Chen, S. Bangsaruntip, K. A. Drouvalakis, N. Wong Shi Kam, M. Shim, Y. Li, W. Kim, P. J. Utz, H. Dai, Proc. Natl. Acad. Sci. USA 2003, 100, 4984–4989. 50. J. J. Gooding, R. Wibowo, J. Liu, W. Yang, D. Losic, S. Orbons, F. J. Mearns, J. G. Shapter, D. B. Hibbert, J. Am. Chem. Soc. 2003, 125, 9006–9007. 51. F. Patolsky, Y. Weizmann, I. Willner, Angew. Chem., Int. Ed. 2004, 43, 2113–2117. 52. Y. Yan, O. Yeheskeli, I. Willner, Chem. Eur. J. 2007, 13, 10168–10175. 53. (a) R. J. Chen, Y. Zhang, D. Wang, H. Dai, J. Am. Chem. Soc. 2001, 123, 3838–3839. (b) J. Zhu, D. Kase, K. Shiba, D. Kasuya, M. Yudasaka, S. Iijima, Nano Lett. 2003, 3, 1033–1036. 54. Y. Yan, I. Baravik, O. Yehezkeli, I. Willner, J. Phys. Chem. C 2008, 112, 17883–17888. 55. M. Reches, E. Gazit, Science 2003, 300, 625–627. 56. F. Patolsky, Y. Weizmann, I. Willner, Nat. Mater. 2004, 3, 692–695. 57. I. Willner, R. Baron, B. Willner, Adv. Mater. 2006, 18, 1109–1120. 58. M. Zayats, R. Baron, I. Popov, I. Willner, Nano Lett. 2005, 5, 21–25. 59. B. Basnar, Y. Weizmann, Z. Cheglakov, I. Willner, Adv. Mater. 2006, 18, 713–718. 60. S. Mann, W. Shenton, M. Li, S. Connoly, D. Fitzmaurice, Adv. Mater. 2000, 12, 147–150. 61. C. M. Niemeyer, Angew. Chem., Int. Ed. 2001, 40, 4128–4158. 62. (a) I. Willner, B. Shlyahovsky, M. Zayats, B. Willner, Chem. Soc. Rev. 2008, 37, 1153–1165. (b) C. Tuerk, L. Gold, Science 1990, 249, 505–510. 63. Z. Cheglakov, Y. Weizmann, A. B. Braunschweig, O. I. Wilner, I. Willner, Angew. Chem., Int. Ed. 2008, 47, 126–130. 64. (a) P. Trvascio, P. K. Witting, A. G. Mauk, D. Sen, J. Am. Chem. Soc. 2001, 123, 1337–1348. (b) P. K. Witting, P. Travascio, D. Sen, A. G. Mauk, Inorg. Chem. 2001, 40, 5017–5023. (c) P. Travascio, Y. F. Li, D. Sen, Chem. Biol. 1998, 5, 505–517. 65. B. Shlyahovsky, D. Li, Y. Weizmann, R. Nowarski, M. Kotler, I. Willner, J. Am. Chem. Soc. 2007, 129, 3814–3815. 66. (a) S. Takenaka, K. Yamashita, M. Takagi, Y. Uto, H. Kondo Anal. Chem. 2000, 72, 1334–1341. (b) I. Willner, M. Zayats, Angew. Chem., Int. Ed. 2007, 46, 6408–6418. 67. S. Takenaka, K. Yamashita, M. Takagi, Y. Uto, H. Kondo, Anal. Chem. 2000, 72, 1334–1341. 68. C. Fan, K. W. Plaxco, A. J. Heeger, Proc. Natl. Acad. Sci. USA 2003, 100, 9134–9137.
376
BIOHYBRID ELECTROCHEMICAL DEVICES
69. (a) Y. Xiao, A. A. Lubin, A. J. Heeger, K. W. Plaxco, Angew. Chem., Int. Ed. 2005, 44, 5456–5459. (b) Y. Xiao, B. D. Piorek, K. W. Plaxco, A. J. Heeger, J. Am. Chem. Soc. 2005, 127, 17990–17991. 70. Y. Xiao, A. A. Rowe, K. W. Plaxco, J. Am. Chem. Soc. 2006, 129, 262–263. 71. B. R. Baker, R. Y. Lai, M. S. Wood, E. H. Doctor, A. J. Heeger, K. W. Plaxco, J. Am. Chem. Soc. 2006, 128, 3138–3139. 72. R. Freeman, Y. Li, R. Tel-Vered, E. Sharon, J. Elbaz I. Willner, Analyst, 2009, 134, 653–656. 73. G. S. Bang, S. Cho, B. -G. Kim, Biosens. Bioelectron. 2005, 21, 863–870. 74. (a) F. Le Floch, H. Ho, M. Leclerc, Anal. Chem. 2006, 78, 4727–4731. (b) F. Le Floch, H. A. Ho, P. Harding-Lepage, M. Bedard, R. Neagu-Plesu, M. Leclerc, Adv. Mater. 2005, 17, 1251–1254. 75. F. Patolsky, Y. Weizmann, I. Willner, J. Am. Chem. Soc. 2002, 124, 770–772. 76. M. Mir, M. Vreeke, I. Katakis, Electrochem. Commun. 2006, 8, 505–511. 77. (a) L. C. Bock, L. C. Griffin, J. A. Latham, E. H. Vermass, J. J. Toole, Nature 1992, 355, 564–566. (b) D. M. Tasset, M. F. Kubik, W. Steiner, J. Mol. Biol. 1997, 272, 688–698. (c) K. Padmanabhan, K. P. Padmanabhan, J. D. Ferrara, J. E. Sandler, A. Tulinsky, J. Biol. Chem. 1993, 268, 17651–17654. 78. K. Ikebukuro, C. Kiyohara, K. Sode, Biosens. Bioelectron. 2005, 20, 2168–2172. 79. N. L. Rosi, C. A. Mirkin, Chem. Rev. 2005, 105, 1547–1562. 80. F. Patolsky, R. Gill, Y. Weizmann, T. Mokari, U. Banin, I. Willner, J. Am. Chem. Soc. 2003, 125, 13918–13919. 81. L. He, M. D. Musick, S. R. Nicewarner, F. G. Salinas, S. J. Benkovic, M. J. Natan, C. D. Keating, J. Am. Chem. Soc. 2000, 122, 9071–9077. 82. R. Polsky, R. Gill, L. Kaganovsky, I. Willner, Anal. Chem. 2006, 78, 2268–2271. 83. (a) J. Wang, D. K. Xu, A. N. Kawde, R. Polsky, Anal. Chem. 2001, 73, 5576–5581. (b) J. Wang, R. Polsky, D. K. Xu, Langmuir 2001, 17, 5739–5741. 84. S. -J. Park, T. A. Taton, C. A. Mirkin, Science 2002, 295, 1503–1506. 85. M. Urban, R. M€oller, W. Fritzsche, Rev. Sci. Instr. 2003, 74, 1077–1081. 86. O. D. Velev, E. W. Kaler, Langmuir 1999, 15, 3693–3698. 87. D. Li, Y. Yan, A. Wieckowska, I. Willner, Chem. Commun. 2007, 3544–3546. 88. E. Sharon, R. Freeman, R. Tel-Vered, I. Willner, Electroanalysis, 2009, 21, 1291–1296. 89. (a) D. Liu, S. H. Park, J. H. Reif, T. H. LeBean, Proc. Natl. Acad. Sci. USA 2004, 101, 717–722. (b) H. Yan, S. H. Park, G. Finkelstein, J. H. Reif, T. H. LeBean, Science 2003, 301, 1882–1884. (c) H. Li, S. H. Park, J. H. Reif, T. H. LaBean, H. Yan, J. Am. Chem. Soc. 2004, 126, 418–419. 90. (a) M. G. Warner, J. E. Hutchison, Nat. Mater. 2003, 2, 272–277. (b) F. A. Aldaye, A. L. Palmer, H. F. Sleiman, Science 2008, 321, 1795–1799. (c) C. M. Niemeyer, U. Simon, Eur. J. Inorg. Chem. 2005, 3641–3655. (d) S. Chung, D. S. Ginger, M. W. Morales, Z. Zhang, V. Chandrasekhar, M. A. Ratner, C. A. Mirkin, Small 2005, 1, 64–69. 91. F. Patolsky, Y. Weizmann, O. Lioubashevski, I. Willner, Angew. Chem., Int. Ed. 2002, 41, 2323–2327. 92. E. Braun, Y. Eichen, U. Sivan, G. Ben-Yoseph, Nature 1998, 391, 775–778. 93. K. Keren, R. S. Berman, E. Buchstab, U. Sivan, E. Braun, Science 2003, 302, 1380–1382.
CHAPTER 13
Electroactive Rotaxanes and Catenanes ALBERTO CREDI and MARGHERITA VENTURI Dipartimento di Chimica ‘‘G. Ciamician’’, Alma Mater Studiorum, Universita di Bologna, Bologna, Italy
13.1
INTRODUCTION
Rotaxanes and catenanes are topologically intriguing chemical species that are currently the object of much interest. Their names derive, respectively, from the Latin words rota and axis for wheel and axle, and catena for chain. Rotaxanes1 are minimally composed (Fig. 13.1a) of a macrocyclic compound (the ‘‘ring’’) threaded by a dumbbell-shaped molecule terminated by bulky groups (‘‘stoppers’’) that prevent disassembly. Catenanes1 are made of (at least) two interlocked macrocycles or ‘‘rings’’ (Fig. 13.1b). Important features of these systems derive from noncovalent interactions between components that contain complementary recognition sites. Such interactions, which are also responsible for the efficient template-directed syntheses2 of rotaxanes and catenanes, include electron donor–acceptor ability, hydrogen bonding, hydrophobic–hydrophilic character, p–p stacking, electrostatic forces, and, on the side of the strong interaction limit, metal–ligand bonding. Rotaxanes and catenanes have, therefore, both molecular and supramolecular characters: molecular because the components are held together mechanically and can be unlinked only by breaking strong covalent bonds and supramolecular because of the presence of weak noncovalent interactions. As it will be shown later, the most interesting features of these systems derive from the intercomponent noncovalent interactions, that is, from their supramolecular nature. Besides their topology, rotaxanes and catenanes are also appealing systems for the construction of molecular machines because (i) the mechanical bond allows a large variety of mutual arrangements of the molecular components, while conferring stability to the system, (ii) the interlocked architecture limits the amplitude of the intercomponent motion in the three directions, (iii) the stability of a specific Electrochemistry of Functional Supramolecular Systems. Edited by Paola Ceroni, Alberto Credi, and Margherita Venturi Copyright 2010 John Wiley & Sons, Inc.
377
378
ELECTROACTIVE ROTAXANES AND CATENANES
Figure 13.1
Schematic representation of a rotaxane (a) and a catenane (b).
arrangement (coconformation)3 is determined by the strength of the intercomponent interactions, and (iv) such interactions can be modulated by external stimulation. The large-amplitude motions that can be achieved with rotaxanes and catenanes are represented schematically in Fig. 13.2. In the case of rotaxanes, two interesting molecular motions can be envisaged, namely, translation, that is, shuttling, of the ring along the axle (Fig. 13.2a), and rotation of the ring around the axle (Fig. 13.2b). Hence, rotaxanes are good prototypes for the construction of both linear and rotary molecular motors. Systems that are capable of performing translational movements are termed molecular shuttles and constitute the most common implementation of the molecular machine concept with rotaxanes. Interestingly, the dumbbell component of a molecular shuttle exerts on the ring motion the same type of directional restriction as imposed by the protein track for linear biomolecular motors (an actin filament for myosin and a microtubule for kinesin and dynein).4 It should also be noted that interlocked molecular architectures are largely present in natural systems—for instance, DNA catenanes and rotaxanes
Figure 13.2 Schematic representation of the intercomponent motions that can be obtained with simple interlocked molecular architectures: ring shuttling in rotaxanes (a), and ring rotation in rotaxanes (b) and catenanes (c).
13.1 INTRODUCTION
379
are known.1 Many processive enzymes, that is, enzymes that remain attached to their biopolymer substrates (DNA, RNA, or proteins) and perform multiple rounds of catalysis before dissociating, are thought to exhibit a rotaxane structure, as confirmed, for example, by the observation of the crystal structure of DNA l-exonuclease.5 In this case, the particular feature of the rotaxane architecture is utilized by Nature to enhance the activity of processive enzymes. When rotaxanes and catenanes contain redox-active units, electrochemical techniques are a very powerful means of characterization. They provide a fingerprint of these systems giving fundamental information on (i) the spatial organization of the redox sites within the molecular and the supramolecular structure, (ii) the entity of the interactions between such sites, and (iii) the kinetic and thermodynamic stabilities of the reduced/oxidized and charge-separated species. It should also be recalled that a full electrochemical, as well as spectroscopic and photophysical, characterization of complex systems such as rotaxanes and catenanes requires the comparison with the behavior of the separated molecular components (ring and thread for rotaxanes and constituting rings in the case of catenanes), or suitable model compounds. As it will appear clearly from the examples reported in the following, this comparison is of fundamental importance to evidence how and to which extent the molecular and supramolecular architecture influences the electronic properties of the component units. An appropriate experimental and theoretical approach comprises the use of several techniques that, as far as electrochemistry is concerned, include cyclic voltammetry, steady-state voltammetry, chronoamperometry, coulometry, impedance spectroscopy, and spectro- and photoelectrochemistry. Electrochemistry is, therefore, a powerful tool to ‘‘read ’’ the state of the system, but in suitably designed rotaxanes and catenanes, it can play a more important role. By causing the occurrence of endoergonic heterogeneous electron transfer processes, electrochemistry can provide the energy needed to modify the noncovalent interactions that stabilize a certain rotaxane and catenane structure. This change can promote mechanical movements that enable the systems to reach a new equilibrium state. The initial interactions can then be restored by using an opposite redox stimulus and, as a consequence, other mechanical movements occur that lead the system back to its initial state. In such cases, electrochemistry plays the dual role of ‘‘writing’’ and ‘‘reading’’ the system: by means of electrons and/or holes, it supplies the energy to make these systems work as molecular machines, and by means of the various electrochemical techniques, (e.g., voltammetry) it is used for controlling and monitoring the operation performed by the machine. It is also important to note that when molecular machines are powered by electrochemical energy inputs, there is the big advantage, for example, in comparison with chemical inputs that these systems work without formation of waste products, the accumulation of which would compromise their operation. A further remarkable advantage derives by the fact that electrodes represent one of the best ways to interface molecular machines to the macroscopic world, a feature that is essential for any future application. In this chapter, for space reasons, only a few paradigmatic examples of rotaxanes and catenanes based on donor–acceptor (charge transfer (CT)) and/or hydrogen bonding interactions (systems based on metal–ligand bonding are reviewed in another
380
ELECTROACTIVE ROTAXANES AND CATENANES
chapter) are reported. They are mainly taken from the authors’ own contribution to this research field and are divided according to the role played by electrochemistry, namely, electron exchange to either ‘‘read ’’ or ‘‘write and read ’’ the state of the system.
13.2
ELECTRON EXCHANGE TO ‘‘READ ’’ THE STATE OF THE SYSTEM
13.2.1 Rotaxanes Containing Identical Recognition Sites for the Ring in their Dumbbell Component If during the template-directed synthesis of a rotaxane, the location of two identical recognition sites (‘stations’) within its dumbbell component can be arranged (Fig. 13.3a and b), a degenerate, coconformational equilibrium state is obtained in which the macrocyclic component spontaneously shuttles back and forth between
Figure 13.3 (a) Operation of a two-station rotaxane as a degenerate molecular shuttle. (b) Idealized representation of the potential energy of the system as a function of the position of the ring relative to the axle. The number of circles in each potential well reflects the relative population of the corresponding coconformation in a statistically significant ensemble.
381
Figure 13.4 Structure formulas of (a) the two-station rotaxane 14 þ that behaves as a degenerate molecular shuttle and (b) its molecular components 2 and 34 þ and model rotaxane 44 þ .
382
ELECTROACTIVE ROTAXANES AND CATENANES
the two stations along the linear portion of the dumbbell.6 The ring spends most of its time surrounding a station; obviously, when one station is ‘‘engaged’’ with the ring, the other one is ‘‘free.’’ An example of rotaxane that behaves as degenerate molecular shuttle is represented by compound 14þ (Fig. 13.4a), which is made of ring 2 incorporating two electron donor dimethoxybenzene (DMB) units and dumbbell-shaped component 34þ (Fig. 13.4b) containing two identical electron acceptor bipyridinium units.7 Information on the supramolecular structure and the interactions that establish between the molecular components can be easily obtained by using electrochemical techniques, in particular by monitoring the cyclic voltammetric behavior of the bipyridinium unit, which is known to give two reversible and monoelectronic reduction processes.8 Rotaxane 14þ shows four distinct reduction processes, monoelectronic and reversible; their assignment can be made by a comparison with the behavior of dumbbell-shaped component 34þ and model rotaxane 44þ (Fig. 13.4b) in which both the bipyridinium units are engaged with macrocycle 2. Dumbbell 34þ exhibits two bielectronic and reversible processes that can be attributed to the simultaneous first and second reduction of the two bipyridinium units contained in its axle-like section. The bielectronic nature of the processes indicates, as expected, that the bipyridinium units are equivalent and behave independently. Also, model rotaxane 44þ shows two bielectronic and reversible processes that are straightforwardly assigned to the bipyridinium units contained in its dumbbell component; they are, however, shifted to more negative potentials compared to dumbbell 34þ . These shifts can be attributed to the CT interactions with the electron donor ring that make the electron acceptor bipyridinium units more difficult to reduce, whereas the bielectronic nature of the processes indicates the such units are noninteracting and equivalent—both of them are surrounded by a ring—in full agreement with the supramolecular structure of 44þ . On the basis of these observations, the four processes of rotaxane 14þ can be easily and unequivocally assigned as follows (Fig. 13.5). The first reduction, which occurs at the same potential as the first process of dumbbell 34þ , is related to the bipyridinium unit not surrounded by the ring; the second one, occurring at a potential that coincides with the first process observed for rotaxane 44þ , concerns the first reduction of the bipyridinium unit surrounded by the ring and therefore involved in CT interactions. Finally, the third and fourth processes can be assigned to the second reduction of the ‘‘free’’ and ‘‘engaged with the ring’’ bipyridinium unit, respectively. Structurally related to these species are the triply branched compound 56þ and its rotaxanes 66þ , 76þ , and 86þ (Fig. 13.6)9, in which one, two, or three acceptor units are encircled by the electron donor macrocyclic compound 2. Although these rotaxanes cannot behave as degenerate molecular shuttles because of their branched topology, they are nevertheless interesting from the electrochemical viewpoint. The triply branched compound 56þ shows a reversible three-electron reduction process and, at more negative potential, a process strongly affected by electrode adsorption. A comparison with the behavior of dumbbell 34þ enables to assign these processes to the simultaneous first and second reduction of its three bipyridinium units.
13.2
ELECTRON EXCHANGE TO ‘‘READ ’’ THE STATE OF THE SYSTEM
383
Figure 13.5 Correlation of the half-wave reduction potentials obtained for rotaxane 14þ , its dumbbell-shaped component 34þ , and model rotaxane 44þ . The two-electron processes are labeled with the number 2.
The branched rotaxanes 66þ , 76þ , and 86þ , besides the bipyridinium units of the triply branched backbone 56þ , contain macrocycle 2 whose two DMB units are oxidized at distinct potentials: the first oxidation process practically coincides with that of the model compound p-dimethoxybenzene, whereas the second one is displaced to a slightly more positive potential. In rotaxane 66þ , one of the three bipyridinium units is engaged in CT interactions with the ring and is therefore expected to be reduced at a more negative potential than the other ones. In accordance with these expectations, four reduction processes have been observed: (i) a bielectronic process at a potential very close to that found for the first reduction of 56þ assigned to the simultaneous reduction of the two free bipyridinium units, (ii) a monoelectronic process at more negative potential assigned to the first reduction of the bipyridinium unit encircled by the ring, (iii) a bielectronic process assigned to the simultaneous second reduction of the two free bipyridinium units, and (iv) a fourth process perturbed by adsorption phenomena, which can be assigned to the second reduction of the bipyridinium unit encircled by the ring. It is interesting to note that the same reduction pattern, except for the number of exchanged electrons, is exhibited by rotaxane 14þ . On oxidation, rotaxane 66þ shows a process, not present in 56þ , that can be assigned to oxidation of the ring. Because of the CT interaction, oxidation occurs at a more positive potential than the first oxidation process of the free macrocyclic compound. Furthermore, even if the process is not fully reversible, it seems that oxidation of the two DMB units of the ring occurs at the same potential, contrary to what is observed for the free compound. This different behavior is most likely due to the fact that in the rotaxane the
384
Figure 13.6
Structure formulas of triply branched compound 56þ and of its rotaxanes 66þ , 76þ , and 86þ .
13.2
ELECTRON EXCHANGE TO ‘‘READ ’’ THE STATE OF THE SYSTEM
385
two DMB units of the ring are prevented from interacting by the interposed bipyridinium unit. The electrochemical behavior of rotaxane 76þ can be straightforwardly explained on the basis of the above discussion. On reduction, a first monoelectronic process, assigned to the first reduction of the free bipyridinium unit, is followed by a bielectronic process, assigned to the first reduction of the two bipyridinium units encircled by the ring. Even the second reduction of the three bipyridinium units, which occurs at more negative potentials, occurs with the same 1:2 pattern. On oxidation, the behavior of rotaxane 76þ is again similar to that of rotaxane 66þ , with a more intense process in correspondence of oxidation of the DMB units. In rotaxane 86þ , the three bipyridinium units are expected to be electrochemically equivalent because each one is encircled by the ring. In agreement with this expectation, it has been found two trielectronic processes corresponding to the first and the second reduction of the bipyridinium units, a situation similar to that observed for compound 56þ . In the case of the rotaxane, however, the processes occur at more negative potentials because of the CT interaction. As far as oxidation is concerned, the behavior of 86þ is in line with that of 66þ and 76þ . The correlation of the electrochemical properties of 56þ and its rotaxanes is nicely illustrated in Fig. 13.7, where the differential pulse voltammetric peaks corresponding to the first reduction of the bipyridinium units are shown. On going from 56þ to 86þ , the areas of the first peak decrease with a decreasing number of free bipyridinium
Figure 13.7 Differential pulse voltammetric peaks (argon-purged MeCN/Et4NPF6 0.05 M, room temperature, glassy carbon electrode, scan rate 20 mV/s, pulse height and duration 75 mV and 40 ms) corresponding to the first reduction of the branched compound 56þ and its rotaxanes 66þ , 76þ , and 86þ .
386
ELECTROACTIVE ROTAXANES AND CATENANES
units and, in parallel, the areas of the second peak increase with an increasing number of encircled bipyridinium units. 13.2.2 Rotaxanes Containing Two Different Recognition Sites for the Ring in their Dumbbell Component When a rotaxane contains two different recognition sites in its dumbbell component, it can exist as two different equilibrating coconformations, the populations of which reflect their relative free energies as determined primarily by the strengths of the two different sets of noncovalent bonding interactions. In the schematic representation shown in Fig. 13.8, it has been assumed that the molecular ring
Figure 13.8 Schematic operation of a two-station rotaxane as a controllable molecular shuttle, and idealized representation of the potential energy of the system as a function of the position of the ring relative to the axle upon switching off and on station A. The number of dots in each position reflects the relative population of the corresponding coconformation in a statistically significant ensemble. Structures (a) and (c) correspond to equilibrium states, whereas (b) and (d) are metastable states. An alternative approach would be to modify station B through an external stimulus in order to make it a stronger recognition site compared to station A.
13.2
ELECTRON EXCHANGE TO ‘‘READ ’’ THE STATE OF THE SYSTEM
387
resides preferentially around station A (state 0), until a stimulus is applied that switches off this recognition site (A~A0 ). As a consequence, the system is brought into a nonequilibrium state that subsequently relaxes (equilibrates) according to the new potential energy landscape. This process implies the motion of the molecular ring to the second recognition site (station B) until a new equilibrium is reached (state 1). If station A is switched on again by an opposite stimulus, the original potential energy landscape is restored, and another coconformational equilibration occurs through the shuttling of the ring back to station A. In appropriately designed rotaxanes, the switching process can be controlled by reversible chemical reactions (protonation–deprotonation, reduction–oxidation, isomerization) caused by chemical, photochemical, or, as it will be discussed in detail later, electrochemical stimulation. After the first reported example,10 a remarkable number of controllable molecular shuttles have been described in the literature.11–14 Owing to their bistability, controllable molecular shuttles are also interesting for processing and storing binary information. An example of a rotaxane that behaves as a chemically controlled molecular shuttle is represented by compound 9H3þ (Fig. 13.9),15 which is made of a crown ether containing two electron donor dioxybenzene moieties and a dumbbell-shaped component that comprises in its rod section a secondary ammonium ion and the already seen electron acceptor bipyridinium unit. These units are two recognition centers, or in other words two possible stations, for the ring component since it can establish hydrogen bonding interactions with the ammonium ion and CT interactions with the bipyridinium unit. For the employed macrocyclic component, the hydrogen bonding interactions are much stronger than the CTones, and therefore the stable structure of the rotaxane is that in which the macrocycle surrounds the ammonium station. Such a structure can be, however, destabilized upon addition of a suitable base that, by deprotonating the ammonium ion, causes the complete displacement of the ring to the bipyridinium
Figure 13.9 Structure formula of rotaxane 9H3þ and representation of its operation as a pH controllable molecular shuttle.
388
ELECTROACTIVE ROTAXANES AND CATENANES
station. The back movement of the ring to the ammonium station can be subsequently obtained upon addition of a suitable acid capable of restoring the ammonium ion (Fig. 13.9). By successive additions of base and acid, it is, therefore, possible to induce the mechanical movements of the ring between the two stations and the system can be viewed as a pH controllable molecular shuttle (Fig. 13.9). Among the various techniques that can be employed to investigate the ring shuttling between the two stations, the electrochemical ones are very useful, particularly the cyclic voltammetry when it is used for monitoring the behavior of the bipyridinium unit, which is one of the two stations involved in the ring shuttling. In protonated rotaxane 9H3þ (Fig. 13.10), the first and second one-electron reduction
Figure 13.10 Cyclic voltammetric behavior on reduction of the protonated 9H3þ and deprotonated 92þ rotaxane shown in Fig. 13.9 and of its protonated and deprotonated dumbbell-shaped component (argon-purged MeCN/Et4NPF6 0.05 M, 298 K, glassy carbon electrode, scan rate 50 mV/s). The current intensity has been corrected to account for the differences in diffusion coefficients.
13.2
ELECTRON EXCHANGE TO ‘‘READ ’’ THE STATE OF THE SYSTEM
389
processes of such unit occur at the same potential values of the dumbbell-shaped component, clearly indicating that the bipyridinium unit is not engaged with the ring and demonstrating, again very clearly, that the ring resides on the other station. Remarkable changes are, however, observed upon deprotonation of the ammonium ion (Fig. 13.10). The reduction processes do not show any changes in the dumbbellshaped component, while in deprotonated rotaxane 92þ , the first reduction of the bipyridinium unit is strongly shifted toward more negative potential. This shift indicates that such a unit is now engaged in charge transfer interactions, thereby proving that the shuttling of the ring has occurred upon deprotonation of the ammonium ion. Interestingly, the electrochemical behavior of the rotaxane after reprotonation matches exactly that obtained before deprotonation, indicating the full reversibility of the acid/base-driven shuttling process. By using an incrementally staged strategy, the architectural features of switchable rotaxane 9H3þ (Fig. 13.9)15 were incorporated into those of a previously investigated16 triply threaded two-component supramolecular bundle. The result was the construction of two-component molecular devices like 10H39þ (Fig. 13.11) that behave like nanoscale ‘‘elevators.’’17 Compound 10H39þ , which measures about 2.5 nm in height and 3.5 nm in diameter, consists of a tripod component 11H39þ , containing two different notches—one ammonium center and one bipyridinium unit—at different levels in each of its three legs. Such legs are interlocked by a tritopic host that plays the role of a platform that can be stopped at the two different levels. Initially, the platform resides exclusively on the ‘‘upper’’ level, that is, with the three rings surrounding the ammonium centers (Fig. 13.11a, state 0). On addition of a phosphazene base to an acetonitrile solution of 10H39þ , deprotonation of the ammonium centers occurs giving the species 106þ and, as a result, the platform moves to the ‘‘lower’’ level, that is, with the three crown ether rings surrounding the bipyridinium units (Fig. 13.11b, state 1). The distance traveled by the platform is 0.7 nm, and it was estimated that a force of up to 200 pN, that is, more than one order of magnitude larger than that generated by natural linear motors,4a could be generated. Subsequent addition of acid restores the ammonium centers, and the platform moves back to the upper level. The acid/base controlled ‘‘up and down’’ elevator-like motion that corresponds to a quantitative switching and can be repeated many times and can be monitored by spectroscopic techniques (1 H NMR, absorption, and fluorescence) and, very conveniently, by electrochemistry. As already seen for 9H3þ , the cyclic voltammetric analysis focused on the behavior of the bipyridinium units is, indeed, a powerful and easy tool to evidence the machine operation. The voltammogram of 10H39þ shows the occurrence of two consecutive reversible reduction processes, each involving the exchange of three electrons (Fig. 13.12a), at potential values identical to those found for the tripod component 11H39þ . This observation indicates that the three bipyridinium units of 10H39þ (and of 11H39þ ) are equivalent, behave independently from one another, and are not engaged in electron donor–acceptor interactions with the platform, which consistently resides on the upper level with its three rings all surrounding the ammonium centers. After addition of three equivalents of base, two
390
Figure 13.11
Structure formula of 10H39þ and its operation as a molecular elevator controlled by acid–base inputs.
13.2
ELECTRON EXCHANGE TO ‘‘READ ’’ THE STATE OF THE SYSTEM
391
Figure 13.12 (a) Cyclic voltammetry of 10H39þ , showing the two reversible three-electron reduction processes of the three bipyridinium units (argon-purged MeCN/n-Bu4NPF6 0.04 M, room temperature, glassy carbon electrode, scan rate 100 mV/s). (b) Correlation diagram of the potential values, in V versus SCE, for the reduction processes of the bipyridinium units of 10H39þ (.) and 11H39þ (*) (top), upon addition of a base to afford 106þ and, respectively, 116þ (middle), and after reprotonation with an acid-stoichiometric amount with respect to the added base (bottom).
reversible three-electron reduction processes are still observed in cyclic voltammetric experiments, but they are displaced to more negative potential values (Fig. 13.12b). Such a change cannot be due simply to deprotonation of the ammonium centers because the potential values for reduction of the bipyridinium units in the deprotonated tripod component 116þ are identical to those of 11H39þ . Hence, the results obtained show that in the deprotonated 106þ the bipyridinium units are surrounded by the electron donor rings of the platform. The changes in reduction potential can be fully reversed by addition of acid and the cycle can be repeated without any apparent loss of reversibility. Electrochemical measurements also support the spectroscopic investigations showing that the platform operates by taking three distinct steps associated with each of the three deprotonation processes. Cyclic voltammetric data obtained for the species 10H39þ , 10H28þ , 10H7þ , and 106þ evidence that the numbers of the bipyridinium units engaged with the rings of the platform are, respectively, 0, 1, 2, and 3 (Fig. 13.13). The base–acid controlled mechanical motion in 10H39þ is associated with interesting structural modifications, such as the opening and closing of a large cavity (1.5 0.8 nm) and the control of the positions and properties of the bipyridinium legs. This behavior can, in principle, be used to control the uptake and release of a guest molecule, a function of interest for the development of drug delivery systems. 13.2.3 Catenanes Containing Identical Recognition Sites in Their Rings As already said, catenanes are minimally composed of two interlocked rings. If it is arranged during the template-directed synthesis to have two identical units, that is, recognition sites, located within two different macrocycles, then the resulting
392
Figure 13.13 (a) Cyclic voltammetric curves for the first three-electron reduction of (from top to bottom) 10H39þ , 10H28þ , 10H7þ , and 106þ (argon-purged MeCN/Et4NPF6 0.04 M, 298 K, glassy carbon electrode, 100 mV/s). (b) Changes in the absorbance at 276 nm on titrating a 5.0 mM 10H39þ acetonitrile solution with the phosphazene base at 298 K. (c) Stepwise motion of the platform down to the legs on successive deprotonation of its three ammonium centers. The structures shown were obtained by molecular mechanics calculations.
13.2
ELECTRON EXCHANGE TO ‘‘READ ’’ THE STATE OF THE SYSTEM
393
Figure 13.14 Dynamic processes associated with circumrotation of one ring in a catenane made of two different macrocycles, each incorporating two identical recognition sites. Asterisks are used to highlight the exchange of position of identical units.
catenane undergoes degenerate coconformational change when one of the macrocycles spontaneously circumrotates through the cavity of the other and vice versa, as illustrated in Fig. 13.14. As each ring spends most of its time surrounding a recognition site of the other ring, each macrocycle has one site located ‘‘inside’’ and the other one positioned ‘‘alongside’’ with respect to the other macrocycle. Electrochemical techniques can be very useful for distinguishing between these two topologically different positions, as already seen for rotaxanes containing two identical recognition sites in their dumbbell. Cyclophane 124þ (Fig. 13.15), which is a symmetric macrocycle extensively used in the construction of catenanes, contains two electron acceptor bipyridinium units. Its cyclic voltammogram shows two reversible bielectronic reduction processes straightforwardly attributed to the simultaneous first and second reduction of the two topologically equivalent bipyridinium units.18 Catenation of 124þ with symmetric macrocycle 2 containing two electron donor DMB units causes big changes in the electrochemical behavior. The resulting catenane 134þ (Fig. 13.15) shows, indeed, three reversible processes: the first two processes correspond to the separate monoelectronic reductions of the two bipyridinium moieties, while the third one, which is bielectronic, is assigned to the simultaneous second reduction of both moieties.18,19 The fact that two separate reductions are observed for the first redox processes is attributed to the topological difference between the inside and alongside positions. The alongside unit, which experiences the interaction with only one of the DMB units, is easier to reduce than the inside one, stabilized to a greater extent by the two electron donor units.
394
Figure 13.15 processes.
Structure formulas of cyclophane 124þ , catenanes 134þ and 144þ , and correlation of the potential values obtained for their reduction
13.2
ELECTRON EXCHANGE TO ‘‘READ ’’ THE STATE OF THE SYSTEM
395
Interestingly, the second reduction of the two bipyridinium units splits in the case of catenane 144þ (Fig. 13.15) obtained by interlocking cyclophane 124þ with a symmetric macrocycle containing two electron donor dimethoxynaphthalene (DMN) units.19 These observations suggest that when 134þ is bireduced the two bipyridinium units become equivalent, whereas they continue to be nonequivalent in 142þ . Electrochemical techniques can also give interesting information in the case of catenanes of higher complexity, as shown by the results obtained by investigating a series of catenanes made of up to seven interlocked rings.20 The three basic components of these catenanes are the tetracationic cyclophanes 124þ and 154þ , and macrocycle 16 containing three electron donor DMN units (Fig. 13.16). For space reasons, only the electrochemical behavior of catenanes 174þ , 188þ , 194þ , and 204þ (Fig. 13.17), compared to those of their molecular components, is reported. For cyclophane 154þ , in which the two bipyridinium units are connected by biphenylene spacers, the number of reduction processes, number of exchanged electrons, and potential values (Fig. 13.17) are very similar to those of cyclophane 124þ (Fig. 13.15). Macrocycle 16, containing three equivalent DMN electroactive units, shows three distinct oxidation processes (Fig. 13.17). Such a contrasting behavior between 16 and the tetracationic cyclophanes, in which the two incorporated bipyridinium units undergo simultaneous first and second reductions, can be interpreted considering that, in the cyclophanes, the rigidity of the structure prevents interaction between the two bipyridinium units, whereas the flexible structure of macrocycle 16 allows the three DMN units to approach one another. As already seen for catenanes 134þ and 144þ (Fig. 13.15), on going from separated molecular components 16, 124þ , or 154þ to their catenanes substantial changes in the electrochemical behavior are expected because the electroactive units incorporated in the cyclophanes and macrocycle are engaged in donor–acceptor interactions and occupy spatially different sites. Starting with catenane 174þ , obtained by interlocking macrocycle 16 with only one cyclophane 124þ , it is found, in agreement with these expectations, that the two bipyridinium units of 124þ undergo their first reduction in separated processes that are cathodically shifted with respect to the free cyclophane (Fig. 13.17a). Comparison with the previously investigated catenane 144þ (Fig. 13.15)19 shows that in 174þ the first reduction process, attributed to the alongside unit, occurs at a slightly more negative potential, whereas the second one, concerning the inside unit, is considerably displaced toward less negative potentials. This difference could result from the great flexibility of macrocycle 16, which (i) allows one of the DMN units to interact with the alongside bipyridinium unit of the cyclophane and, at the same time, (ii) does not force two DMN units to sandwich the inside bipyridinium unit. After the first reduction, the electron donor–acceptor interaction becomes weaker and the flexible nature of 16 could facilitate a fast interchange between monoreduced inside and alongside units, with a consequent lack of splitting of the second process. On the oxidation side, in 174þ the three DMN units of macrocycle 16 undergo distinct processes that are shifted to more positive potential values with respect to the free macrocycle (Fig. 13.17) because of the presence of charge transfer interactions. A detailed
396
Figure 13.16
Structure formulas and cartoons of molecular components 124þ , 154þ , and 16.
397
Figure 13.17 Correlation diagrams for the electrochemical behavior: (a) catenanes 174þ and 188þ , and their 124þ and 16 components; (b) catenanes 194þ and 204þ , and their 154þ and 16 components.
398
ELECTROACTIVE ROTAXANES AND CATENANES
Figure 13.18 Cyclic voltammograms for reduction of catenanes 188þ and 204þ (argon-purged MeCN/Et4NPF6 0.05 M, 298 K, glassy carbon electrode, scan rate 200 mV/s).
comparison between the two oxidation patterns is made difficult because, in the catenane structure, the donor–acceptor interactions between nonoxidized and oxidized DMN units are partially or completely prevented (Fig. 13.17a). For catenane 188þ —composed of 16 and two cyclophanes 124þ —simultaneous reduction of the alongside units of the two cyclophanes is followed by simultaneous reduction of the inside units (Fig. 13.18). Both processes are displaced to less negative potentials compared with 174þ (Fig. 13.17a) because the ratio between electron donor and electron acceptor units decreases from 3/2 to 3/4. The splitting of the second reduction process of the bipyridinium units is consistent with the crowded structure of the catenane, which presumably prevents fast interchange between inside and alongside units. As far as oxidation is concerned, in catenane 188þ , only two processes are observed, a feature that is consistent with the presence of an outside and two equivalent inside DMN units (Fig. 13.17a). On going to consider the catenanes made of macrocycle 16 and cyclophane 154þ , it is observed that in 194þ (Fig. 13.17b) the first reduction of the two bipyridinium units of the cyclophane component occurs, as expected, in two distinct processes. It is worth noting that the first process is not at all displaced compared to 154þ , indicating that the alongside unit is practically unperturbed by the presence of the macrocycle. This situation is presumably related to the large size and rigidity of the cyclophane, which, in the most stable catenane geometry, leaves one of the two bipyridinium units far away from the electron donor units of 16. The lack of splitting
13.2
ELECTRON EXCHANGE TO ‘‘READ ’’ THE STATE OF THE SYSTEM
399
of the second process (Fig. 13.17b), already observed for 174þ (Fig. 13.17a), and its very small displacement toward more negative potentials clearly indicate that, once the two bipyridinium units have been monoreduced, the large size of the two components prevents any further charge transfer interaction. The oxidation pattern of 194þ is fully consistent with that observed for 174þ . Catenane 204þ shows two bielectronic reduction processes because the two bipyridinium units of 154þ occupy equivalent positions (Figs 13.17b and 13.18). The strong displacement toward more negative potentials with respect to 154þ is obviously due to the charge transfer interaction with the electron donor units of the encircling macrocycle. The further displacement, compared to that observed for the first reduction of the inside unit of 194þ , can be attributed to the presence of two DMN units inside the cyclophane. This feature could favor a geometry where a close approach between the electron donor and electron acceptor units is sterically enforced, thereby optimizing the charge transfer interactions. As indicated by the noticeable displacement of the second reduction to more negative potentials (Fig. 13.17b), such a structure seems to be essentially maintained, even after oneelectron reduction of the two bipyridinium units. Concerning oxidation pattern (Fig. 13.17b), the first process can be assigned to the two DMN units not involved in stacking interactions, the second one to the two alongside units, and the subsequent ones to oxidation of the two (interacting) DMN units contained inside the cyclophane. Increasing attention has also been devoted to the incorporation of novel functions in catenane structures. Particularly interesting, for example, are the catenanes in which either the electron acceptor cyclophane ring or the electron donor macrocycle contains a 2,20 -bipyridine (bpy) ligand. Starting from cyclophane ligand 214þ , it has been possible to prepare the catenane ligands 224þ and 234þ , which were then employed to prepare several mononuclear catenane complexes (Fig. 13.19).21 These compounds exhibit an interesting electrochemical behavior characterized by several redox processes concerning (i) reduction of the bipyridinium- and bpy-type moieties of the catenane ligands, (ii) reduction of the bpy ligands (in the Ru complexes), (iii) oxidation of the metals, and (iv) oxidation of the DMN moieties of the macrocycle in the complexes containing the 234þ catenane ligand. For instance, complex 246þ exchanges up to a total of nine electrons. On reduction, it shows two monoelectronic and one bielectronic processes involving the bipyridinium units, and three monoelectronic processes concerning the bpy moieties (Fig. 13.20). On oxidation, two monoelectronic processes are observed; the first one, being reversible, is assigned to the oxidation of the metal center and the second one, not fully reversible, to the oxidation of the alongside DMN unit of the macrocycle interlocked with the cyclophane. 13.2.4 Catenanes Containing Different Recognition Sites in their Rings When one of the two rings of a catenane carries two different recognition sites, then the opportunity exists to control the dynamic processes (Fig. 13.21) in a manner reminiscent of the already discussed controllable molecular shuttles (Fig. 13.8).
400
ELECTROACTIVE ROTAXANES AND CATENANES
Figure 13.19 Structure formulas of cyclophane ligand 214þ , catenane ligands 224þ and 234þ , and mononuclear catenane complexes.
The two possible coconformational isomers of such catenanes can be interchanged by appropriate stimuli. In a diagram of potential energy against rotation angle of the asymmetric macrocycle, the two coconformations correspond to energy minima, provided by the intercomponent noncovalent bonding interactions. The
13.2
ELECTRON EXCHANGE TO ‘‘READ ’’ THE STATE OF THE SYSTEM
401
Figure 13.20 Cyclic voltammetric behavior on reduction of 246þ (argon-purged MeCN/ Et4NPF6 0.05 M, glassy carbon electrode, scan rate 50 mV/s).
initially populated coconformer is that associated with the most favorable energetic state (state 0). Stimulus S1 has the effect of destabilizing such isomer and leads to the other one (state 1), a change that can simply be viewed as a circumrotation of the asymmetric macrocycle. An opposite stimulus S2 restores the original situation. By switching off and on again the recognition properties of one of the two recognition sites of the asymmetric macrocycle, the relative populations of the two species can be controlled reversibly. It should be pointed out, however, that repeated switching between the two states does not need to occur through a full rotation. In fact, because of the intrinsic symmetry of the system, both the movement from state 0 to state 1 and that from state 1 to state 0 can take place, with equal
Figure 13.21 The two coconformational isomers associated with a catenane incorporating two different recognition sites within one of its two macrocyclic components can be interchanged by appropriate stimuli (S1 and S2).
402
ELECTROACTIVE ROTAXANES AND CATENANES
Figure 13.22 The circumrotation of the tetracationic cyclophane component of catenane 254þ can be controlled reversibly by adding–protonating n-hexylamine that forms a charge transfer adduct with the diazapyrenium unit of the catenane.
probabilities, along a clockwise or an anticlockwise direction. A full (360 ) rotation movement can only occur in ratchet-type systems, that is, in the presence of asymmetry elements that can be structural or functional in nature. Mechanical movements in suitably designed catenanes can be induced by chemical, photochemical, or, as it will be discussed in detail later, electrochemical stimulation.11–13 Catenane 254þ (Fig. 13.22) is an interesting example of a system in which the chemically controlled ring rotation can be monitored by using electrochemical techniques. It incorporates macrocycle 2 and a tetracationic cyclophane comprising a bipyridinium and a diazapyrenium unit. Its major isomer is that in which the diazapyrenium unit is located inside the cavity of the macrocycle and the bipyridinium unit is positioned alongside. In agreement with this coconformation, the differential pulse voltammogram (CH3CN, 298 K) of 254þ shows two peaks that can be easily assigned to the monoelectronic reductions of the alongside bipyridinium unit and of the inside diazapyrenium group (Fig. 13.23). The ability of n-hexylamine to form22–24 adducts with diazapyrenium has been exploited25 to displace the equilibrium in favor of the isomer having the diazapyrenium unit alongside the cavity of the macrocycle (Fig. 13.22). The electrochemical analysis is once again very useful to evidence this coconformational change. Indeed, after the addition of n-hexylamine, the first peak shifts by 60 mV to a potential that corresponds to the monoelectronic reduction of a bipyridinium unit encircled by macrocycle 2. Similarly, the second peak shifts by 20 mV to a potential that, by a comparison with the behavior of symmetric catenane 264þ , is associated with the monoelectronic reduction of a diazapyrenium unit interacting with n-hexylamine (Fig. 13.23). Upon addition of CF3SO2H, protonation of n-hexylamine occurs and the adduct formed between n-hexylamine and the diazapyrenium unit of the catenane is destroyed. Consequently, the original coconformation associated with 254þ is
403
Figure 13.23 Differential pulse voltammetry reduction pattern (argon-purged MeCN/n-Bu4NPF6 0.05 M, 298 K, glassy carbon electrode, scan rate 20 mV/s, pulse height 75 mV, pulse duration 40 ms) of 254þ (left) and 264þ (right). Curves a: starting solutions. Curves b: after addition of 10 equivalent of Me(CH2)5NH2. Curves c: after subsequent addition of 10 equivalent of CF3SO3H.
404
ELECTROACTIVE ROTAXANES AND CATENANES
restored, as revealed by the fact that the differential pulse voltammogram recorded after the addition of CF3SO2H is identical to that obtained before the addition of n-hexylamine (Fig. 13.23). 13.2.5
Catenanes and Rotaxanes Immobilized on Surfaces
Several examples of catenanes and rotaxanes have been constructed and investigated on solid surfaces.11a,d–f,12,13,26 If the interlocked molecular components contain electroactive units and the surface is that of an electrode, electrochemical techniques represent a powerful tool to study the behavior of the surface-immobilized ensemble. Catenanes and rotaxanes are usually deposited on solid surfaces by employing the Langmuir–Blodgett technique27 or the self-assembled monolayer (SAM) approach.28 The molecular components can either be already interlocked prior to attachment to the surface or become so in consequence of surface immobilization; in the latter setting, the solid surface plays the dual role of a stopper and an interface (electrode). In most instances, the investigated compounds are deposited on macroscopic surfaces, such as those of metal or semiconductor electrodes;26 less common is the case of systems anchored on nanocrystals.29 A series of investigations30,31 showed that cationic catenanes and rotaxanes such as those described in the previous sections can be used to obtain ordered films on the air–water interface when the counteranions are replaced by amphiphilic counterions such as the dihexadecylphosphate anion (DHP).31 For instance, films made of the acid–base switchable rotaxane 9H3þ (Fig. 13.9) and DHP anions were transferred onto ITO supports for spectroscopic and electrochemical investigations. Reversible switching of the potential values for the reduction processes of the bipyridinium unit of 9H3þ upon cyclic exposure of the films to vapors of HCl and NH3 was observed.31 A comparison of the behavior of the rotaxane-containing films with that of films containing its dumbbell-shaped component suggested that the switching process is mainly related to a rearrangement of the multilayer upon protonation–deprotonation of the ammonium center. However, the occurrence of a ring shuttling motion in the case of the rotaxane-containing films could not be excluded. Nonetheless, the examined thin films may be useful for sensing purposes because they provide a reversible electrical output signal in response to acid–base input stimuli. A rotaxane based on a benzylic amide macrocycle containing a pyridine moiety was grafted onto a SAM of 11-mercaptoundecanoic acid on gold.32 The grafting is achieved because of the formation of hydrogen bonds between the pyridine moiety of the rotaxane molecules and the carboxylic groups at the top of the Au-SAM. Cyclic voltammetric and electrochemical impedance spectroscopic measurements showed that the resulting SAM is densely packed and well ordered, and allowed the estimation of the average thickness of the monolayer. An interesting example of a system whose components become mechanically bound upon surface immobilization is rotaxane trans-27 (Fig. 13.24), consisting of a ferrocene-functionalized b-cyclodextrin (b-CD) macrocycle threaded on a molecule containing a photoisomerizable azobenzene unity and a long alkyl chain.33 A monolayer of trans-27 was self-assembled on a gold electrode. Therefore, the ring
13.2
ELECTRON EXCHANGE TO ‘‘READ ’’ THE STATE OF THE SYSTEM
405
Figure 13.24 Schematic representation of the electron transfer processes in the two states of the surface-bound photoswitchable rotaxane 27.
component is prevented from dethreading at either end of the axle by a bulky anthracene stopper group and the presence of the gold surface. The azobenzene unit in the trans configuration is complexed by b-CD; photoisomerization to the cis form renders complexation sterically impossible, so that the b-CD ring is displaced to the alkyl component. Back-photoisomerization restores the original trans configuration. The position of the b-CD-tethered ferrocene unit was determined by chronoamperometry. A fast current decay (k ¼ 65 s1) was observed for the trans isomer, implying that the ring component is close to the electrode surface (Fig. 13.24). Photoisomerization of the monolayer to the cis state resulted in a chronoamperometric transient characterized by a substantially lower electron-transfer rate (k ¼ 15 s1). This result indicates that in cis-27 the b-CD ring is more distant from the electrode
406
ELECTROACTIVE ROTAXANES AND CATENANES
surface. Owing to the reversibility of azobenzene photoisomerization, a cyclic pattern for the rate constant of the heterogeneous electron transfer process was observed. In this optoelectronic system, optical information is transduced by a mechanical shuttling to an electronic signal.33
13.3 ELECTRON EXCHANGE TO ‘‘WRITE AND READ’’ THE STATE OF A SYSTEM 13.3.1
Rotaxanes as Electrochemically Driven Molecular Shuttles
As discussed in Section 13.2.2, when a rotaxane contains two different recognition sites in its dumbbell component, it can behave as a controllable molecular shuttle, and, if appropriately designed by incorporating suitable redox units, it can perform its machine-like operation by exploiting electrochemical energy inputs. Of course, in such cases, electrons/holes, besides supplying the energy needed to make the machine work, can also be useful to read the state of the systems by means of the various electrochemical techniques. The first example of electrochemically driven molecular shuttles is rotaxane 284þ (Fig. 13.25) constituted by the electron-deficient cyclophane 124þ and a dumbbellshaped component containing two different electron donors, namely, a benzidine and a biphenol moieties, that represent two possible stations for the cyclophane.10 Because benzidine is a better recognition site for 124þ than biphenol, the prevalent isomer is that having the former unit inside the cyclophane. The rotaxane
Figure 13.25 Structure formula of rotaxane 284þ and its electrochemically controlled switching process.
13.3 ELECTRON EXCHANGE TO ‘‘WRITE AND READ’’ THE STATE OF A SYSTEM
407
Figure 13.26 Structure formula of rotaxane 294þ and the electrochemically induced shuttling of the cyclophane along the dumbbell-shaped component (CH3CN, 298 K).
exhibits two one-electron oxidations, both assigned to the benzidine recognition site. A comparison of the half-wave potentials of 284þ with those of a model compound incorporating a benzidine unit not encircled by the tetracationic cyclophane shows that the potential for the first oxidation is more positive in the rotaxane while that for the second oxidation is the same in both compounds. The shift of the first process indicates that the tetracationic cyclophane surrounds the benzidine station, making its first one-electron oxidation more difficult. The fact that the second oxidation is unaffected by the presence of the cyclophane can be explained by considering that once the benzidine is oxidized to the corresponding radical cation, the tetracationic cyclophane moves away from it. Upon reduction of the benzidine unit back to its neutral state, the original equilibrium between the two coconformations associated with the rotaxane is restored. After this first report, a remarkable number of electrochemically controllable molecular shuttles have been designed, constructed, and studied. Rotaxane 294þ (Fig. 13.26), for instance, incorporates the electron-deficient cyclophane 124þ and a dumbbell containing two kinds of electron-rich units, namely, one 2,6-dioxyanthracene and two 1,4-dioxybenzene moieties.34 In solution, the rotaxane is present as the isomer with the 2,6-dioxyanthracene unit inside the cyclophane, owing to the fact that this unit is a better station in comparison to the 1,4-dioxybenzene recognition sites. The cyclic voltammogram of 294þ shows a first process that corresponds to the oxidation of the 2,6-dioxyanthracene recognition site. This oxidation occurs, however, at a potential that is more positive than that of a model compound incorporating this unit not surrounded by the ring. As far as the oxidation of the two dimethoxybenzene units is concerned, two distinct processes are observed: the first one occurs at a potential that is almost identical to that of a model compound incorporating
408
ELECTROACTIVE ROTAXANES AND CATENANES
only this unit, while the second one occurs at a potential that is almost identical to that of a model rotaxane incorporating this unit encircled by the tetracationic cyclophane. These observations indicate that (i) initially the tetracationic cyclophane resides around the 2,6-dioxyanthracene recognition site, making its oxidation more difficult, and (ii) once this recognition site is oxidized, the tetracationic cyclophane moves away from it and encircles one of the two DMB units (Fig. 13.26).34 Another example is constituted by the previously described (Section 13.2.2) pH-controllable rotaxane 9H3þ (Fig. 13.9).15 As already discussed, the ring surrounds the ammonium ion but, after deactivation of such a station upon deprotonation, it moves to the electron acceptor bipyridinium unit, which is the other station present in the dumbbell component. The occurrence of this shuttling process is evidenced by the fact that the first reduction of the bipyridinium unit occurs at a potential more negative than that found for the dumbbell component (Fig. 13.10). Interestingly, this process exhibits a cyclic voltammetric pattern (a large and scan-rate-dependent separation between the anodic and cathodic peak), which is typical of systems in which the redox process is followed by a molecular rearrangement taking place on the timescale of the electrochemical experiments. In the case of the examined rotaxane, the nature of the occurring rearrangements can be elucidated by looking at the characteristics of the second reduction process, namely, a potential value that practically coincides with that of the protonated rotaxane (and the dumbbell-shaped component) and a behavior that is fully reversible. These two findings indicate that in the deprotonated rotaxane the monoreduced bipyridinium unit is no longer engaged in donor–acceptor interactions and that the system does not undergo any further rearrangement after one-electron reduction. It can be concluded, as shown in the square scheme reported in Fig. 13.27, that in the deprotonated rotaxane (i) the first reduction of the bipyridinium weakens the charge transfer interactions and promotes the displacement of the ring far from the monoreduced unit, and (ii) the reoxidation of such a unit, restoring its electron acceptor power, causes the back movement of the ring. From the electrochemical viewpoint, this system is particularly interesting because, by changing its protonation state, electrons can play different roles: in protonated rotaxane 9H3þ (see Section 13.2.2), they are simply used to ‘‘read ’’ the state of the system, whereas in deprotonated rotaxane 92þ , they play the dual role of ‘‘writing and reading’’ the system. Benzylic amide rotaxanes constitute another important class of interlocked compounds.1,11f Rotaxane 30, for example, consists of a benzylic amide macrocycle that surrounds an axle featuring two hydrogen bonding stations, namely, a succinamide and a naphthalimide unit, separated by a long alkyl chain (Fig. 13.28).35 Initially, the macrocycle resides onto the succinamide station because the naphthalimide unit is a much poorer hydrogen bonding recognition site. As the naphthalimide anion is a much stronger hydrogen bonding station compared to succinamide, electrochemical one-electron reduction of the naphthalimide unit causes the shuttling of the macrocycle from the succinamide to the naphthalimide station (Fig. 13.28). Subsequent reoxidation of the naphthalimide site to the neutral state restore the original coconformation. The thermodynamic
409
Figure 13.27 Dual-pathway square scheme mechanism that accounts for the rearrangements induced by the monoelectronic reduction of deprotonated rotaxane 92þ . The species A and C represent the stable structure of the deprotonated rotaxane and its monoreduced form, respectively, whereas B and D are metastable intermediates. Note that the exact position of the macrocycle along the axle in the reduced forms B and C is not known. From a simple digital simulation of the cyclic voltammetric patterns, the following values have been obtained: EAB ¼ 0.59 V, EDC ¼ 0.34 V, kAD 0.15 s1, 1 1 1 kDA 2.5 s , kBC 100 s , and kCB 1 s .
410
ELECTROACTIVE ROTAXANES AND CATENANES
Figure 13.28
The electrochemically induced shuttling in the benzylic amide rotaxane 30.
and kinetic aspects of the shuttling motion were investigated in detail by a cyclic voltammetric analysis on changing scan rate, temperature, and solvent, aided by computer simulations.35 For example, in tetrahydrofuran at room temperature, it was found that the relative ring binding affinities of the two stations can be switched by eight orders of magnitude and the redox-induced shuttling motion takes about 50 ms. Rotaxane 316þ was specifically designed36 to achieve photoinduced ring shuttling in solution,37 but it also behaves as an electrochemically driven molecular shuttle. This compound has a modular structure: its ring component is the electron donor macrocycle 2, whereas its dumbbell component is made of several covalently linked units. They are a Ru(II) polypyridine complex (P2þ ), a p-terphenyl-type rigid spacer (S), a 3,30 -dimethyl-4,40 -bipyridinium (A22þ ) and a 4,40 -bipyridinium (A12þ ) electron-accepting stations, and a tetraarylmethane group as the terminal stopper (T) (Fig. 13.29). Because of the presence of several redox-active units, the cyclic voltammogram of this rotaxane shows a complex redox pattern. However, the comparison to the electrochemical behavior of its molecular components and suitable model compounds (Fig. 13.29) enables to obtain useful information not only on its coconformational features, but also, and most importantly, on its machine-like operation. In dumbbell-shaped component 326þ , all the redox processes of the incorporated units are present at almost the same potentials as in the separated units (Fig. 13.30); this finding shows that there are no substantial intercomponent electronic interactions. On going from the dumbbell component to rotaxane 316þ , some processes are affected while others are not (Fig. 13.30). All the processes related to the Ru-based unit, namely, the metal-localized oxidation and the ligand-localized reductions, do not show any appreciable changes. The first reduction of the A12þ unit is displaced noticeably toward more negative potential values, indicating that it is surrounded by the electron donor macrocycle 2 (Fig. 13.31). Accordingly, the oxidation of the two DMB units of the macrocycle is displaced toward more positive potential values and occurs simultaneously, as already observed9 for other rotaxanes containing ring 2 (Fig. 13.31). The fact that the ring encircles the A12þ station, as confirmed by the NMR spectrum, is an expected result on the basis of the reduction potentials of A12þ and A22þ in component 326þ (or of separated model compounds). The second process of 316þ , which corresponds to the first reduction of the A22þ station, is also displaced
13.3 ELECTRON EXCHANGE TO ‘‘WRITE AND READ’’ THE STATE OF A SYSTEM
411
Figure 13.29 Structure formulas of rotaxane 316þ , its dumbbell-shaped component 326þ , ring 2, and model compounds 332þ , 344þ , 352þ , and 362þ of the units present in the dumbbell.
toward more negative potential values (Fig. 13.31), demonstrating that, at this stage, the A22þ unit is encircled by macrocycle 2. This behavior confirms that, when the better station (A12þ ) of the two has been ‘‘deactivated’’ upon reduction, the ring moves to the alternative A22þ station. Under these conditions, from the values of the first reduction potential of A22þ and the second reduction potential of A12þ in the dumbbell component, it can be estimated that the translational isomer with the ring surrounding A22þ is much more populated than that in which the ring encircles A1þ . When also the A22þ station has been reduced, the position of the ring is no longer controlled by strong CT interactions; from the electrochemical results, it seems that it resides close to A2 þ . The reversibility of the electrochemical processes involving the two stations shows that, after a two-electron reduction of rotaxane 316þ , one-electron oxidation relocated the ring on the A22þ station and a successive one-electron oxidation entices it back again onto the A12þ station. The electrochemical reversibility of these processes also indicates that the rates of the electrochemically induced ring movements are fast.
412
ELECTROACTIVE ROTAXANES AND CATENANES
Figure 13.30 Cyclic voltammetric patterns (argon-purged MeCN/Et4NPF6 0.05 M, 298 K, glassy carbon electrode, scan rate 200 mV/s) for rotaxane 316þ , its dumbbell-shaped component 326þ , ring 2, and model compounds 332þ , 344þ , 352þ , and 362þ .
More recently, the second-generation molecular shuttle 374þ (Fig. 13.32) was designed and constructed.38 The system is composed of two devices: a bistable redoxdriven molecular shuttle and a module for photoinduced charge separation. In the stable translational isomer, the electron-accepting cyclophane 124þ , which is confined in the region of the dumbbell delimited by the two stoppers T1 and T2, encircles the better electron donor tetrathiafulvalene (TTF) station.
13.3 ELECTRON EXCHANGE TO ‘‘WRITE AND READ’’ THE STATE OF A SYSTEM
413
Figure 13.31 Potential shifts caused by the donor/acceptor interaction between ring 2 and dumbbell-shaped component 326þ when they are assembled in rotaxane 316þ . Circles, squares, and triangles represent processes centered on DMB, A12þ and A22þ units, respectively.
Voltammetric experiments revealed remarkable electronic interactions between the various units of 374þ , pointing to the existence of folded conformations in solution.39 Interestingly, the TTF unit can be electrochemically oxidized only in a limited fraction of the rotaxane molecules; in these species, TTF oxidation causes the shuttling of the cyclophane 124þ away from this station. Most likely, 374þ occurs as conformations in which the electroactive TTF unit is buried inside a complex molecular structure and is, therefore, protected against oxidation performed by an electric potential applied externally. Such a behavior limits the efficiency for the operation of 374þ as a redox-driven molecular shuttle. The possibility of oxidizing TTF by an electric potential generated internally through intramolecular photoinduced electron transfer is currently under investigation. In general terms, these results indicate that, as the structural complexity increases, the overall properties of the system cannot be easily rationalized solely on the basis of the type and sequence of the functional units incorporated in the molecular framework—that is, its ‘‘primary’’ structure. Higher level conformational effects, which are reminiscent of those related to the secondary and tertiary structures of biomolecules,4d have to be taken into consideration. The comprehension of these effects constitutes a stimulating scientific problem and a necessary step for the design of novel artificial molecular devices and machines.
Figure 13.32 Structure formula of rotaxane 374þ .
414
13.3.2
ELECTROACTIVE ROTAXANES AND CATENANES
Catenanes as Electrochemically Driven Molecular Rotors
As discussed in Section 13.2.4, when one of the two rings of a catenane carries two different recognition sites, the dynamic processes of one ring with respect to the other can be controlled. In particular, if redox units are incorporated into the catenane structure, there is the possibility of controlling these processes upon electrochemical stimulation. Catenanes that exhibit such a behavior can be seen as electrochemically driven molecular rotors. An example is offered by catenane 384þ (Fig. 13.33a), which incorporates macrocycle 2 and a tetracationic cyclophane comprising one bipyridinium and one trans-1,2-bis(4-pyridinium)ethylene unit.19,40 In the major isomer, the bipyridinium unit is located inside the cavity of the macrocyclic polyether and the trans-bis(pyridinium)ethylene unit is positioned alongside, as confirmed by the electrochemical analysis. The cyclic voltammogram of the catenane shows four monoelectronic processes that, by a comparison with the data obtained for the free cyclophane, can be attributed as follows: the first and third processes to the first and second reductions of the bipyridinium unit, and the second and fourth ones to the first and second reductions of the trans-bis (pyridinium)ethylene unit. The comparison with the tetracationic cyclophane also evidences that all these reductions are shifted toward more negative potential values (Fig. 13.33b). The discussion can be restricted to the first and second reduction processes that are of particular interest in this context. The shift of the bipyridinium-based process is in agreement with the catenane coconformation in which the bipyridinium unit is located inside the cavity of the macrocyclic polyether (Fig. 13.33a); because of the CT interactions established with both the electron donor units of the macrocycle, its reduction is more difficult than in the free tetracationic cyclophane. The shift of the trans-1,2-bis(4-pyridinium)ethylene-based reduction indicates that, once the bipyridinium unit is reduced, the CT interaction that stabilize the initial coconformation are destroyed and, thereby, the tetracationic cyclophane circumrotates through the cavity of the macrocyclic polyether moving the trans-bis(pyridinium)ethylene unit inside, as shown by comparison of its reduction potential with that of a catenane model compound.19 The original equilibrium between the two coconformations associated with catenane 384þ is restored upon oxidation of both units back to their dicationic states. It is interesting to notice that for a machine-like performance, the presence of an asymmetric ring is a necessary but not sufficient requirement. This statement is clearly demonstrated by the behavior of catenane 394þ made by the same tetracationic cyclophane of 384þ and a macrocycle containing two DMN units.19,40 As in the case of 384þ , the major isomer is the one in which the bipyridinium unit is located inside the macrocycle. However, in contrast with the behavior of 384þ , for which the first reduction process concerns the inside bipyridinium unit, the first reduction of 394þ involves the alongside trans-bis(pyridinium)ethylene unit. This process has been undoubtedly attributed to this unit by performing spectroelectrochemical experiments and comparing the spectrum of the monoreduced catenane with that of a model compound containing only the trans-bis(pyridinium)ethylene
13.3 ELECTRON EXCHANGE TO ‘‘WRITE AND READ’’ THE STATE OF A SYSTEM
415
Figure 13.33 (a) The circumrotation of the tetracationic cyclophane component of catenane 384 þ can be controlled reversibly by reducing–oxidizing electrochemically its bipyridinium unit. (b) Correlation between the half-wave reduction potentials of catenane 384 þ and of its tetracationic ring component. Circles and squares correspond to the reduction of bipyridinium and trans-bis(pyridinium)ethylene units, respectively.
unit. The different behavior of the two catenanes, as far as the first reduction process is concerned, can be explained on the basis of the different strength of the CT interactions: in 394þ the bipyridinium unit is sandwiched between two DMN moieties that are stronger electron donor than the DMB moieties of the macrocyclic component of catenane 384þ . Because of these stronger interactions, the reduction
416
ELECTROACTIVE ROTAXANES AND CATENANES
Figure 13.34 Correlation between the half-wave reduction potentials of catenane 394þ and of its tetracationic ring component. Circles and squares correspond to the reduction of bipyridinium and trans-bis(pyridinium)ethylene units, respectively.
of such a unit becomes so difficult that it occurs at a potential more negative than that of the trans-bis(pyridinium)ethylene unit (Fig. 13.34). As a consequence of the fact that in 394þ the first reduction concerns the alongside unit, the CT interactions responsible of the initial coconformation are practically unaffected and no mechanical movement occurs in the monoreduced catenane. Catenane 404þ (Fig. 13.35) is another example of a system in which the coconformational motion can be controlled electrochemically.41 It is made of the symmetric tetracationic cyclophane 124þ and a nonsymmetric ring comprising two
Figure 13.35 Redox controlled ring rotation in solution for catenane 404þ , which contains the symmetric electron acceptor cyclophane 124þ and a nonsymmetric electron donor ring.
13.3 ELECTRON EXCHANGE TO ‘‘WRITE AND READ’’ THE STATE OF A SYSTEM
417
different electron donor units, namely, a tetrathiafulvalene group and a DMN unit. Because the TTF unit is a better electron donor than the DMN one, as witnessed by the potential values for their oxidation, the thermodynamically stable coconformation of the catenane is that in which the symmetric ring encircles the TTF unit of the nonsymmetric one (Fig. 13.35a, state 0). The cyclic voltammogram of the free macrocycle shows a reversible process attributed to the monoelectronic oxidation of the TTF unit. In the catenanes, such a process occurs at more positive potentials, in agreement with the fact that the TTF unit is located inside the cavity of the tetracationic cyclophane and, therefore, engaged in strong CT interactions. It is also interesting to notice that in the catenane this oxidation process is characterized by a large separation between the anodic and cathodic peaks that varies as the scan rate is changed. Upon increasing the scan rate, the anodic peak moves to more positive potentials, while the cathodic one shifts to less positive values. These observations indicate that the oxidation–reduction of the TTF unit is accompanied by the circumrotation of the macrocyclic polyether through the cavity of the tetracationic cyclophane and that this change is occurring on the timescale of the electrochemical experiment. Indeed, after oxidation, the newly formed monocationic tetrathiafulvalene unit (Fig. 13.35b) loses its electron donor power; as a consequence, it is expelled from the cavity of the tetracationic cyclophane and is replaced by the neutral DMN unit (Fig. 13.35c, state 1). After reduction, the original coconformation is restored as the neutral TTF unit replaces the DMN unit inside the cavity of the tetracationic cyclophane. Catenane 404þ was also incorporated into a solid-state device that could be used for random access memory (RAM) storage.42 In addition, this compound could be employed for the construction of electrochromic systems, because its various redox states are characterized by different colors.41,43 By an appropriate choice of the functional units that are incorporated into the catenane components, more complex functions can be obtained. An example is represented by catenane 41H5þ (Fig. 13.36), composed of a symmetric crown ether and a cyclophane ring containing two bipyridinium and one ammonium recognition sites.44 The electrochemical properties, as well as the absorption spectra, show that the crown ether surrounds a bipyridinium unit of the other ring both in 41H5þ (Fig. 13.36a) and in its deprotonated form 414þ (Fig. 13.36b), indicating that deprotonation–protonation of the ammonium unit does not cause any displacement of the ring (state 0). Electrochemical measurements also show that after one-electron reduction of both bipyridinium units of 41H5þ , the ring is displaced on the ammonium function (Fig. 13.36c, state 1), which means that an electrochemically induced coconformational switching does occur. Furthermore, upon deprotonation of the two-electron reduced form of the catenane (Fig. 13.36d), the crown ether moves to one of the monoreduced bipyridinium units (state 0). Therefore, in order to achieve the motion of the ring in the deprotonated catenane 414þ , it is necessary to both reduce (switch off) the bipyridinium units and protonate (switch on) the amine function. The mechanical motion in such a catenane occurs according to an AND logic,45 a function associated with two energy inputs of different nature.
418
ELECTROACTIVE ROTAXANES AND CATENANES
Figure 13.36 Switching processes of catenane 41H5þ in solution. Starting from the deprotonated catenane 414þ , the position of the ring switches under acid–base and redox inputs according to AND logic.
Controlled rotation of the molecular rings has also been achieved in catenanes composed of three interlocked macrocycles. For example, catenane 42H26þ (Fig. 13.37) is made up of two identical macrocycles 2 interlocked with a cyclophane containing two bipyridinium and two ammonium units.44 Because of the type of the macrocycles used, the stable coconformation of 42H26þ is that in which the two rings surround the bipyridinium units (Fig. 13.37a, state 0). Upon addition of one electron in each of the bipyridinium units, the two macrocycles move on the ammonium stations (Fig. 13.37b, state 1) and move back to the original position when the bipyridinium units are reoxidized. The electrochemical behavior of a catenane composed of two identical benzylic amide macrocycles was also investigated by cyclic voltammetry.46 Computer simulation of the voltammetric data, together with quantum chemical calculations, suggests that reduction of the macrocycles is followed by their irreversible chemical soldering, owing to the formation of a CC bond between two reduced carbonyl groups. Hence, the electrochemical stimulus can be used to prevent the mutual rotation of the two rings, although the irreversibility of the reaction limits further developments.
13.3 ELECTRON EXCHANGE TO ‘‘WRITE AND READ’’ THE STATE OF A SYSTEM
419
Figure 13.37 Redox controlled movements of the ring components in catenane 42H26þ composed of three interlocked macrocycles. These motions are obtained upon reduction– oxidation of the bipyridinium units of the cyclophane.
13.3.3 Catenanes and Rotaxanes Immobilized on Surfaces As pointed out in Section 13.2.5, the state of electroactive catenanes and rotaxanes attached on electrode surfaces can be conveniently monitored by electrochemical techniques. In appropriate cases, the control of the electrode potential can also be used to switch the state of the system; an interesting example of this kind is shown in Fig. 13.38. A monolayer of the rotaxane 434þ , consisting of the electron-accepting cyclophane 124þ threaded on a molecular axle that includes an electron-donating diiminobenzene unit and is stoppered by an adamantane moiety, was assembled on a gold electrode.47 The tetracationic ring, which is originally located on the diiminobenzene unit by virtue of electron donor–acceptor interactions, is displaced toward the electrode upon one-electron reduction of its two bipyridinium units at 0.53 V versus SCE, owing to disruption of the donor–acceptor interactions and electrostatic attraction to the electrode (Fig. 13.38). Reoxidation of the reduced cyclophane at 0.33 V versus SCE causes ring shuttling to the original diiminobenzene site. The position of the tetracationic and dicationic (reduced) cyclophane rings and the shuttling rate constants (80 s1 and 320 s1 at 298 K for reduction- and reoxidationinduced processes, respectively) were determined by chronoamperometry and impedance measurements.47a Investigation of the temperature dependence of the shuttling rates showed that the reduction-induced shuttling is an energetically downhill process with no measurable activation barrier, whereas reoxidation-induced shuttling requires thermal activation.47b The lack of an energy barrier in the former case is in agreement with the fact that the shuttling is mainly driven by coulomb attraction of the still positively charged cyclophane toward the negatively polarized electrode. A rotaxane related to 30 (Fig. 13.28), in which (i) the macrocycle bears pyridine moieties and (ii) the naphthalimide site on the axle is replaced by a naphthalene
420
ELECTROACTIVE ROTAXANES AND CATENANES
Figure 13.38 an Au-SAM.
The electrochemically driven ring shuttling of rotaxane 434þ incorporated into
diimide unit, was recently prepared and investigated.48 Three oxidation states of the naphthalene diimide unit could be electrochemically accessed in solution, each one with a very different binding affinity for the ring. Therefore, the rotaxane behaves as an electrochemically controlled molecular shuttle in solution. Interestingly, the reduction potential of the naphthalene diimide unit is sufficiently positive (0.68 V versus SCE) for the process to be compatible with operation in an Au-SAM. Hence, the rotaxane was grafted onto a SAM of 11-mercaptoundecanoic acid on gold (see Section 13.2.5) and the resulting layer was characterized by electrochemical impedance spectroscopy and double-potential step chronoamperometry.48 These experiments demonstrated that the redox-switched shuttling caused by one-electron reduction of the naphthalene diimide station also occurs on the surface, taking place on the millisecond timescale. The redox-controlled mechanical switching in SAMs of disulfide-functionalized bistable TTF–DMN rotaxanes consisting of cyclophane 124þ and a dumbbell-shaped component containing TTF and DMN stations was also extensively investigated.49
13.4
CONCLUSIONS
The described systems show the potentialities of electrochemical techniques (the various kinds of voltammetry, chronoamperometry, coulometry, impedance
REFERENCES
421
spectroscopy, spectroelectrochemistry, and photoelectrochemistry) in characterizing complex systems such as redox-active rotaxanes and catenanes. They provide, indeed, a fingerprint of these systems giving fundamental information on (i) the spatial organization of the redox sites within the molecular and supramolecular structures, (ii) the entity of the interactions between such sites, and (iii) the kinetic and thermodynamic stabilities of the reduced/oxidized and charge-separated species. Electrochemistry is, therefore, a powerful tool to ‘‘read’’ the state of the system. In suitably designed rotaxanes and catenanes, however, electrochemistry can play a more important role. By causing the occurrence of endoergonic heterogeneous electron transfer processes, electrochemistry can, indeed, provide the energy needed to modify the noncovalent interactions that stabilize a certain rotaxane and catenane structure promoting mechanical movements. In such cases, electrochemistry plays the dual role of ‘‘writing’’ and ‘‘reading’’ the system: by means of electrons and/or holes, it supplies the energy to make theses systems work as molecular machines, and by means of the various electrochemical techniques, it is used for controlling and monitoring the operation performed by the system. Although the examples described here evidence that electrochemists have learned how to deal with increasingly complex molecular and supramolecular structures, it is important to notice that electrochemistry is only a part of the game: as the complexity of the systems studied increases, the contribution from many disciplines in a joint and collaborative effort is needed. For the intriguing structures described here, this statement is particularly true. The goal of transforming molecular devices and machines into practically useful products requires, indeed, that people belonging to different fields, such as chemistry, solid-state physics, biology, computer science, mathematics, materials sciences, and so on, work together and learn a common language. ACKNOWLEDGMENTS We are grateful to all our collaborators and coworkers whose names appear in the reference list. Financial support from Ministero dell’Istruzione, dell’Universita e della Ricerca, Regione Emilia-Romagna (PROMINER), and Fondazione Cassa di Risparmio in Bologna is gratefully acknowledged.
REFERENCES 1. J.-P. Sauvage, C. Dietrich-Buchecker (Eds.), Catenanes, Rotaxanes and Knots, WileyVCH, Weinheim, 1999. 2. (a) C. A. Schalley, T. Weilandt, J. Bruggemann, F. V€ ogtle, Top. Curr. Chem. 2004, 248, 141. (b) F. Arico´, J. D. Badjic, S. J. Cantrill, A. H. Flood, K. C.-F. Leung, Y. Liu, J. F. Stoddart, Top. Curr. Chem. 2005, 249, 203. 3. M. C. T. Fyfe, J. F. Stoddart, Acc. Chem. Res. 1997, 30, 393. 4. (a) M. Schliwa, (Ed.), Molecular Motors, Wiley-VCH, Weinheim, 2003. (b) R. A. L. Jones, Soft Machines, Nanotechnology and Life, Oxford University Press, Oxford, 2004.
422
5. 6.
7.
8.
9.
10. 11.
12.
13. 14.
15.
ELECTROACTIVE ROTAXANES AND CATENANES
(c) D. S. Goodsell, Bionanotechnology—Lessons from Nature, Wiley, Hoboken, NJ, 2004. (d) S. Mann, Angew. Chem., Int. Ed. 2008, 47, 2. P. Thordason, R. J. M. Nolte, A. E. Rowan, Aust. J. Chem. 2004, 57, 323 and references therein. (a) P. L. Anelli, N. Spencer, J. F. Stoddart, J. Am. Chem. Soc. 1991, 113, 5131. (b) P. L. Anelli, M. Asakawa, P. R. Ashton, R. A. Bissell, G. Clavier, R. Go´rski, A. E. Kaifer, S. J. Langford, G. Mattersteig, S. Menzer, D. Philp, A. M. Z. Slawin, N. Spencer, J. F. Stoddart, M. S. Tolley, D. J. Williams, Chem. Eur. J. 1997, 3, 1113. P. R. Ashton, R. Ballardini, V. Balzani, M. Belohradsky, M. T. Gandolfi, D. Philp, L. Prodi, F. M. Raymo, M. V. Reddington, N. Spencer, J. F. Stoddart, M. Venturi, D. J. Williams, J. Am. Chem. Soc. 1996, 118, 4931. (a) A. Summers, The Bipyridinium Herbicides, Academic Press, New York, 1980.(b) P. M. S. Monk, The Viologens—Physicochemical Properties, Synthesis and Application of the Salts of 4,40 -Bipyridine, Wiley, Chichester, 1998. D. B. Amabilino, M. Asakawa, P. R. Ashton, R. Ballardini, V. Balzani, M. Belohradsky, A. Credi, M. Highuci, F. M. Raymo, T. Shimizu, J. F. Stoddart, M. Venturi, K. Yase, New J. Chem. 1998, 22, 959. R. A. Bissell, E. Co´rdova, A. E. Kaifer, J. F. Stoddart, Nature 1994, 369, 133. (a) V. Balzani, A. Credi, F. M. Raymo, J. F. Stoddart, Angew. Chem., Int. 2000, 39, 3348. (b) F. C. Simmel, W. U. Dittmer, Small 2005, 1, 284. (c) G. S. Kottas, L. I. Clarke, D. Horinek, J. Michl, Chem. Rev. 2005, 105, 1281. (d) K. Kinbara, T. Aida, Chem. Rev. 2005, 105, 1377. (e) W. R. Browne, B. L. Feringa, Nat. Nanotechnol. 2006, 1, 25. (f) E. R. Kay, D. A. Leigh, F. Zerbetto, Angew. Chem. Int. Ed. 2007, 46, 72. (g) J. Bath, A. J. Turberfield, Nat. Nanotechnol. 2007, 2, 275. (h) B. Champin, P. Mobian, J.-P. Sauvage, Chem. Soc. Rev. 2007, 36, 358. (i) A. Mateo-Alonso, D. M. Guldi, F. Paolucci, M. Prato, Angew. Chem. Int. Ed. 2007, 46, 8120. (a) J. F. Stoddart, (Ed.), Special issue on Molecular Machines, Acc. Chem. Res. 34 (6) 2001. (b) J.-P. Sauvage, (Ed.), Special volume on Molecular Machines and Motors, Struct. Bond. 2001, 99. (c) T. R. Kelly, (Ed.), special volume on Molecular Machines, Top. Curr. Chem 2005, 262. (d) A. Credi, H. Tian, (Eds.), special issue on Molecular Machines and Switches, Adv. Funct. Mater. 17, 5. 2007. V. Balzani, A. Credi, M. Venturi, Molecular Devices and Machines—Concepts and Perspectives for the Nanoworld, 2nd edition, Wiley-VCH, Weinheim, 2008. For recent examples see (a) F. Durola, J.-P. Sauvage, Angew. Chem. Int. Ed. 2007, 46, 3537. (b) L. Zhu, X. Ma, F. Ji, Q. Wang, H. Tian, Chem. Eur. J. 2007, 13, 9216. (c) W. Zhou, J. Li, X. He, C. Li, J. Lv, Y. Li, S. Wang, H. Liu, D. Zhu, Chem. Eur. J. 2008, 14, 754. (d) R. E. Dawson, S. Maniam, S. F. Lincoln, C. J. Easton, Org. Biomol. Chem. 2008, 6, 1814. (e) Y.-L. Zhao, W. R. Dichtel, A. Trabolsi, S. Saha, I. Aprahamian, J. F. Stoddart, J. Am. Chem. Soc. 2008, 130, 11294. (f) M. J. Barrell, D. A. Leigh, P. J. Lusby, A. M. Z. Slawin, Angew. Chem. Int. Ed. 2008, 47, 8036. (g) T. Umehara, H. Kawai, K. Fujiwara, T. Suzuki, J. Am. Chem. Soc. 2008, 130, 13981. (a) P. R. Ashton, R. Ballardini, V. Balzani, I. Baxter, A. Credi, M. C. T. Fyfe, M. T. Gandolfi, M. Go´mez-Lo´pez, M.-V. Martınez-Dıaz, A. Piersanti, N. Spencer, J. F. Stoddart, M. Venturi, A. J. P. White, D. J. Williams, J. Am. Chem. Soc. 1998, 120, 11932. (b) S. Garaudee, S. Silvi, M. Venturi, A. Credi, A. H. Flood, J. F. Stoddart, ChemPhysChem 2005, 6, 2145.
REFERENCES
423
16. V. Balzani, M. Clemente-Leon, A. Credi, J. N. Lowe, J. Badjic, J. F. Stoddart, D. J. Williams, Chem. Eur. J. 2003, 9, 5348. 17. (a) J. D. Badjic, V. Balzani, A. Credi, S. Silvi, J. F. Stoddart, Science 2004, 303, 1845. (b) J. D. Badjic, C. M. Ronconi, J. F. Stoddart, V. Balzani, S. Silvi, A. Credi, J. Am. Chem. Soc. 2006, 128, 1489. 18. P. L. Anelli, P. R. Ashton, R. Ballardini, V. Balzani, M. Delgado, M. T. Gandolfi, T. T. Goodnow, A. E. Kaifer, D. Philp, M. Pietrasziewicz, L. Prodi, M. V. Reddington, A. M. Z. Slawin, N. Spencer, J. F. Stoddart, C. Vincent, D. J. Williams, J. Am. Chem. Soc. 1992, 114, 193. 19. P. R. Ashton, R. Ballardini, V. Balzani, A. Credi, M. T. Gandolfi, S. Menzer, L. PerezGarcıa, L. Prodi, J. F. Stoddart, M. Venturi, A. J. P. White, D. J. Williams, J. Am. Chem. Soc. 1995, 117, 11171. 20. D. B. Amabilino, P. R. Ashton, V. Balzani, S. E. Boyd, A. Credi, J. Y. Lee, S. Menzer, J. F. Stoddart, M. Venturi, D. J. Williams, J. Am. Chem. Soc. 1998, 120, 4295. 21. P. R. Ashton, V. Balzani, S. E. Boyd, A. Credi, O. Kocian, D. Pasini, L. Prodi, N. Spencer, J. F. Stoddart, M. S. Tolley, M. Venturi, D. J. Williams, Chem. Eur. J. 1998, 4, 590. 22. R. Ballardini, V. Balzani, A. Credi, M. T. Gandolfi, S. J. Langford, S. Menzer, L. Prodi, J. F. Stoddart, M. Venturi, D. J. Williams, Angew. Chem. Int. Ed. Engl. 1996, 35, 978. 23. A. Credi, M. Montalti, V. Balzani, S. J. Langford, F. M. Raymo, J. F. Stoddart, New J. Chem. 1998, 22, 1061. 24. A. Credi, V. Balzani, S. J. Langford, J. F. Stoddart, J. Am. Chem. Soc. 1997, 119, 2679. 25. V. Balzani, A. Credi, S. J. Langford, F. M. Raymo, J. F. Stoddart, M. Venturi, J. Am. Chem. Soc. 2000, 122, 3542. 26. V. Balzani, A. Credi, M. Venturi, ChemPhysChem 2008, 9, 202 and references therein. (b) S. Silvi, M. Venturi, A. Credi, J. Mater. Chem. 2009, 19, 2279, and references therein. 27. M. C. Petty, Langmuir-Blodgett Films—An Introduction, Cambridge University Press, Cambridge, 1996. 28. A. Ulman, Characterization of Organic Thin Films, Butterworth-Heinemann, Boston, 1995. See also the chapter by F. M. Raymo et al. in this book. 29. (a) D. Ryan, S. N. Rao, H. Rensmo, D. Fitzmaurice, J. A. Preece, S. Wenger, J. F. Stoddart, N. Zaccheroni, J. Am. Chem. Soc. 2000, 122, 6252. (b) S. Saha, K. C.-F. Leug, T. D. Nguyen, J. F. Stoddart, J. I. Zink, Adv. Funct. Mater. 2007, 17, 685. 30. M. Asakawa, M. Higuchi, G. Mattersteig, T. Nakamura, A. R. Pease, F. M. Raymo, T. Shimidzu, J. F. Stoddart, Adv. Mater. 2000, 12, 1099, and references therein. 31. M. Clemente-Leo´n, A. Credi, M.-V. Martınez-Dıaz, C. Mingotaud, J. F. Stoddart, Adv. Mater. 2006, 18, 1291. 32. F. Cecchet, P. Rudolf, S. Rapino, M. Margotti, F. Paolucci, J. Baggerman, A. M. Brouwer, E. R. Kay, J. K. W. Wong, D. A. Leigh, J. Phys. Chem. B 2004, 108, 15192. 33. I. Willner, V. Pardo-Yssar, E. Katz, K. T. Ranjit, J. Electroanal. Chem. 2001, 497, 172. 34. R. Ballardini, V. Balzani, W. Dehaen, A. E. Dell’Erba, F. M. Raymo, J. F. Stoddart, M. Venturi, Eur. J. Org. Chem. 1999, 591. 35. A. Altieri, F. G. Gatti, E. R. Kay, D. A. Leigh, D. Martel, F. Paolucci, A. M. Z. Slawin, J. K. Y. Wong, J. Am. Chem. Soc. 2003, 125, 8644. 36. (a) P. R. Ashton, R. Ballardini, V. Balzani, A. Credi, R. Dress, E. Ishow, C. J. Kleverlaan, O. Kocian, J. A. Preece, N. Spencer, J. F. Stoddart, M. Venturi, S. Wenger, Chem. Eur. J. 2000,
424
37.
38. 39. 40. 41.
42.
43. 44. 45. 46. 47. 48.
49.
ELECTROACTIVE ROTAXANES AND CATENANES
6, 3558. (b) V. Balzani, M. Clemente-Leo´n, A. Credi, B. Ferrer, M. Venturi, A. H. Flood, J. F. Stoddart, Proc. Natl. Acad. Sci. USA 2006, 103, 1178. For a related example of a photochemically driven molecular shuttle, see A. M. Brouwer, C. Frochot, F. G. Gatti, D. A. Leigh, L. Mottier, F. Paolucci, S. Roffia, G. W. H. Wurpel, Science 2001, 291, 2124. S. Saha, A. H. Flood, J. F. Stoddart, S. Impellizzeri, S. Silvi, M. Venturi, A. Credi, J. Am. Chem. Soc. 2007, 129, 12159. T. Yamamoto, H.-R. Tseng, J. F. Stoddart, V. Balzani, A. Credi, F. Marchioni, M. Venturi, Collect. Czech. Chem. Commun. 2003, 68, 1488. P. R. Ashton, R. Ballardini, V. Balzani, M. T. Gandolfi, D. J.-F. Marquis, L. Perez-Garcıa, L. Prodi, J. F. Stoddart, M. Venturi, J. Chem. Soc. Chem. Commun. 1994, 177. (a) M. Asakawa, P. R. Ashton, V. Balzani, A. Credi, C. Hamers, G. Mattersteig, M. Montalti, A. N. Shipway, N. Spencer, J. F. Stoddart, M. S. Tolley, M. Venturi, A. J. P. White, D. J. Williams, Angew. Chem. Int. Ed. 1998, 37, 333. (b) V. Balzani, A. Credi, G. Mattersteig, O. A. Matthews, F. M. Raymo, J. F. Stoddart, M. Venturi, A. J. P. White, D. J. Williams, J. Org. Chem. 2000, 65, 1924. (a) C. P. Collier, G. Mattersteig, E. W. Wong, Y. Luo, K. Beverly, J. Sampaio, F. M. Raymo, J. F. Stoddart, J. R. Heath, Science 2000, 289, 1172. (b) Y. Luo, C. P. Collier, J. O. Jeppesen, K. A. Nielsen, E. Delonno, G. Ho, J. Perkins, H.-R. Tseng, T. Yamamoto, J. F. Stoddart, J. R. Heath, ChemPhysChem 2002, 3, 519. D. W. Steuerman, H.-R. Tseng, A. J. Peters, A. H. Flood, J. O. Jeppesen, K. A. Nielsen, J. F. Stoddart, J. R. Heath, Angew. Chem. Int. Ed. 2004, 43, 6486. P. R. Ashton, V. Baldoni, V. Balzani, A. Credi, H. D. A. Hoffmann, M.-V. Martinez-Diaz, F. M. Raymo, J. F. Stoddart, M. Venturi, Chem. Eur. J. 2001, 7, 3482. V. Balzani, A. Credi, M. Venturi, ChemPhysChem 2003, 4, 49. P. Ceroni, D. A. Leigh, L. Mottier, F. Paolucci, S. Roffia, D. Tetard, F. Zerbetto, J. Phys. Chem. B 1999, 103, 10171. (a) E. Katz, O. Lioubashevsky, I. Willner, J. Am. Chem. Soc. 2004, 126, 15520. (b) E. Katz, R. Baron, I. Willner, N. Richke, R. D. Levine, ChemPhysChem. 2005, 6, 2179. G. Fioravanti, N. Haraszkiewicz, E. R. Kay, S. M. Mendoza, C. Bruno, M. Marcaccio, P. G. Wiering, F. Paolucci, P. Rudolf, A. M. Brouwer, D. A. Leigh, J. Am. Chem. Soc. 2008, 130, 2593. (a) H.-R. Tseng, D. Wu, N. Fang, X. Zhang, J. F. Stoddart, ChemPhysChem 2004, 5, 111. (b) S. S. Jang, Y. H. Jang, Y.-H. Kim, W. A. Goddard III, A. H. Flood, B. W. Laursen, H.-R. Tseng, J. F. Stoddart, J. O. Jeppesen, J. W. Choi, D. W. Steuerman, E. DeIonno, J. R. Heath, J. Am. Chem. Soc. 2005, 127, 1563.
CHAPTER 14
Electrochemically Driven Molecular Machines Based on Transitionmetal Complexed Catenanes and Rotaxanes JEAN-PAUL COLLIN, FABIEN DUROLA, and JEAN-PIERRE SAUVAGE Laboratoire de Chimie Organo-Minerale, UMR 7177 du CNRS, Faculte de Chimie, Universite de Strasbourg, Strasbourg Cedex, France
14.1
INTRODUCTION
Molecules are highly dynamic objects that undergo very easily distortions in the gas phase, in solution, or even in the solid state. Processes such as nitrogen inversion in amines, rotation about a CC bond in a biphenyl derivative, or the chair-boat equilibrium in cyclohexane have been investigated for decades. They are now textbook notions. Very different is the situation as far as the elaboration and the study of compounds for which the motions are triggered and controlled from the outside, by sending one or several signals to the molecule. In such a case, the molecules behave as real “molecular” machines. They display one or several distinct geometries that can be interconverted in a reversible manner. This field of research has experienced a spectacular development in the course of the past 15 years as testified by the profusion of scientific papers and review article of books devoted to the field.1–4 The rapid growth of this research field is to a large extent related to the discovery of numerous biological systems behaving as machines, including in particular the protein motors, and to a better understanding of their role and of their functioning mode. The motions of these systems correspond to essential biological functions. These biological motors have been investigated in detail and, in a few instances, it has even been possible to visualize their movements while they are in action. Electrochemistry of Functional Supramolecular Systems. Edited by Paola Ceroni, Alberto Credi, and Margherita Venturi Copyright 2010 John Wiley & Sons, Inc.
425
426
ELECTROCHEMICALLY DRIVEN MOLECULAR MACHINES
The most classical examples are ATP synthase,5 a universal rotary motor responsible for the synthesis of ATP, the actin–myosin complex6 of the striate muscles, which behave as a linear motor, or kinesin,7 an essential protein that “walks” (or rather, runs) along a microtubule (playing the role of a rail) as it transports the most important molecular components of the cells over long distances. These fascinating biological motors represent attractive challenges for synthetic chemists, who are of course far from being able to fabricate artificial molecular systems of comparable complexity but who would like to synthesize primitive models fulfilling the most simple functions of the natural systems.8,9 The field of catenanes and rotaxanes occupies a central place in the area of molecular machines. These compounds have led to the deliberate fabrication of molecular machine prototypes from the mid-1990s.10 From a practical viewpoint, it is also easy to conceive that molecular systems whose constitutive elements are held together by mechanical bonds instead of covalent bonds are prompted to undergo large amplitude motions. In theory, a ring can rotate around the axis on which it is threaded (rotary motor) or move along this axis in a given direction or in the other, thus leading to a linear motor reminiscent of the behavior of a piston and a cylinder. More complex systems can also be envisaged, such as a molecular “muscle”.11 The field of molecular machines is not restricted to catenanes and rotaxanes. Outstanding work has been done in this field using photochemically isomerizable alkenes, affording light-driven rotary motors.12 Other interesting molecular systems have also been proposed, based on different principles.13–15 In this chapter, we will focus on transition metal-based catenanes and rotaxanes. We will restrict ourselves to compounds that are set in motion by an electrochemical signal. Indeed, the electrochemical techniques represent privileged methods for piloting these machines since they contain electroactive transition metal centers or complexes. In addition to triggering the motions, electrochemistry allows to investigate the dynamic properties of the compounds.
14.2 ELECTROCHEMICALLY DRIVEN MOTIONS IN COPPER-COMPLEXED CATENANES 14.2.1 The Archetype: Electrochemically Induced Ring Gliding Motions in a Two-Geometry Copper-Complexed Catenanes The first molecular motor elaborated and studied in our group was a catenane containing two different interlocking rings. Its principle is explained in Fig. 14.1.16 The actual system and the full-square scheme are indicated in Fig. 14.2. The starting copper(I) complex 1(4) þ is a four-coordinate species, whose high redox potential ( þ 0.63 V versus SCE in CH3CN) clearly indicates that the geometry of the system (tetrahedral or distorted tetrahedral) is well adapted to copper(I). This redox state being very stable with the environment provided by 1(4) þ , a relatively high redox potential will have to be applied for the monovalent copper center to be oxidized to the divalent state. Interestingly, the four-coordinate Cu(II) complex 1(4)2 þ is an
14.2
ELECTROCHEMICALLY DRIVEN MOTIONS IN COPPER-COMPLEXED CATENANES
427
Figure 14.1 Electrochemically triggered rearrangement of a copper-complexed [2]catenane containing two different rings. The stable four-coordinate monovalent complex (top left; the black circle represents Cu(I)) is oxidized to an intermediate tetrahedral divalent species (top right; the white circle represents Cu(II)). This compound undergoes a complete reorganization process to afford the stable five-coordinate Cu(II) complex (bottom right). Upon reduction, the five-coordinate monovalent state is formed as a transient (bottom left). Finally, the latter undergoes the conformational change that regenerates the starting complex.
intense green species, with a d–d absorption band at 670 nm (e 800 in CH3CN). This compound can be generated by either chemical (Br2 or NOBF4) or electrochemical oxidation. In the text, the subscript 4 or 5 indicates the number of nitrogen atoms coordinated to the metal. The changeover reaction converting 1(4)2 þ to the stable five-coordinate species 1(5)2 þ is quantitative. It is easily monitored by visible absorption spectroscopy since the product of the rearrangement reaction is only slightly colored (pale olive green;
Figure 14.2 Electrochemically induced molecular rearrangements undergone by the copper catenate 12 þ / þ . In the text, the subscript 4 or 5 indicates the number of nitrogen atoms coordinated to the metal. This number is explicitly shown on the figure.
428
ELECTROCHEMICALLY DRIVEN MOLECULAR MACHINES
lmax ¼ 640 nm; e 125). Both copper(II) complexes 1(4)2 þ and 1(5)2 þ have electronic spectra typical for four- and five-coordinate species, respectively, in accordance with previously reported complexes having analogous ligand sets. The same conversion process can also be monitored by electron paramagnetic resonance spectroscopy (EPR).17 It was demonstrated in an unambiguous fashion that 1(4)2 þ is a distorted tetrahedral complex and that the product of the changeover, 1(5)2 þ , is a square pyramidal compound. An interesting question is dealing with the rate of the ring gliding motion that transforms 1(4)2 þ into 1(5)2 þ or, after reduction of the latter, 1(5) þ into 1(4) þ . It was observed that this last process, involving Cu(I), is reasonably fast (a few seconds at room temperature, regardless of the solvent), whereas the copper(II) complex rearrangement 1(4)2 þ /1(5)2 þ is very slow and depends enormously on experimental conditions. This linkage isomerization reaction was shown to take place in a few minutes in anhydrous acetonitrile, but it requires hours or even days to go to completion in noncoordinating solvents or in the absence of coordinating counterions. The strong accelerating influence of CH3CN (over CH2Cl2) or Cl (over PF6) may give indications regarding the rearrangement mechanism. In the course of the changeover process, removal of a dpp (2,9-diphenyl-1,10-phenanthroline) unit from the copper(II) coordination sphere has to proceed before any interaction between the metal center and the entering terpy ligand is possible. This implies that the copper(II) atom is “half-naked” at some stage. If coordinating ions or solvent molecules are present in the medium, they could interact with the metal in this coordinatively unsaturated complex, in a transitory fashion, and thus lower the activation barrier of the rearrangement by stabilizing intermediate states. 14.2.2 A Copper-Complexed [2]Catenane in Motion with Three Distinct Geometries Multistage systems seem to be uncommon, although they are particularly challenging and promising in relation to nanodevices aimed at important electronic functions and, in particular, information storage.18–21 Among the many examples that have been reported in recent years, three-stage catenanes are particularly significant since they lead to unidirectional rotary motors.9 In the mid-1990s, our group has described a particular Cu-complexed [2]catenane that represents an example of such a multistage compound although no directional control was achieved in this particular example.22 The molecule displays three distinct geometries, each stage corresponding to a different coordination number of the central complex (CN ¼ 4, 5, or 6). The principle of the three-stage electrocontrollable catenane is represented in Fig. 14.3. Similarly to the very first and simpler catenane made in our group for which a large amplitude motion can deliberately be triggered by an external signal,16 the gliding of the rings in the present system relies on the important differences of stereochemical requirements for coordination of Cu(I) and Cu(II). For the monovalent state, the stability sequence is CN ¼ 4 > CN ¼ 5 > CN ¼ 6. On the contrary, divalent Cu is known to form stable hexacoordinate complexes, with pentacoordinate systems being often less stable and tetrahedral Cu(II) species being even more strongly disfavored.
14.2
ELECTROCHEMICALLY DRIVEN MOTIONS IN COPPER-COMPLEXED CATENANES
429
Figure 14.3 A three-configuration Cu(I) catenate whose general molecular shape can be dramatically modified by oxidizing the central metal (Cu(I) to Cu(II)) or reducing it back to the monovalent state. Each ring of the [2]catenate incorporates two different coordinating units: the bidentate dpp unit is symbolized by a U, whereas the terpy fragment (2,20 :60 ,200 -terpyridine) is indicated by a stylized W. Starting from the tetracoordinate monovalent Cu complex (Cu(I) N4 þ ; top left) and oxidizing it to the divalent state (Cu(II)N42 þ ), a thermodynamically unstable species is obtained that should first rearrange to the pentacoordinate complex Cu(II)N52 þ by gliding of one ring (left) within the other and, finally, to the hexacoordinate stage Cu(II)N62 þ by rotation of the second cycle (right) within the first one. Cu(II)N62 þ is expected to be the thermodynamically stable divalent complex. The double ring gliding motion following oxidation of Cu(I)N4 þ can be inverted by reducing Cu(II)N62 þ to the monovalent state (Cu(I)N6 þ ; top right), as represented on the top line of the figure.
The synthesis of the key catenate [Cu(I)N4] þ PF6 ¼ 2(4) þ (Fig. 14.8a) (one should notice that, as usual, the subscripts 4, 5, and 6 indicate the coordination number of the copper center) derives from the usual three-dimensional template strategy.23,24 The visible spectrum of this deep red complex shows a metal-to-ligand charge transfer (MLCT) absorption band (lmax ¼ 439 nm, e ¼ 2570 L/mol/cm, CH3CN). Cyclic voltammetry of a CH3CN solution shows a reversible redox process at þ 0.63 V (versus SCE). Both the cyclic voltammetry (CV) data and the UV–vis spectrum are similar to those of other related species.16,24 The reaction of 2(4) þ with KCN afforded the free catenand (not represented), which was subsequently reacted with Cu(BF4)2 to give 2(6)2 þ as a very pale green complex. The hexacoordinate structure of this species was evidenced by UV–vis spectroscopy and electrochemistry. The cyclic voltammogram shows an irreversible reduction at 0.43 V (versus SCE, CH3CN). These data are similar to the ones obtained for the complex Cu(diMetpy)2(BF4)2 (5,500 -dimethyl-2,20 :60 ,200 -terpyridine). When a dark red solution of 2(4) þ was oxidized by an excess of NO þ BF4, a green solution of 2(4)2 þ was obtained. The CV is the same as for the starting complex, and the visible absorption spectrum shows a band at lmax ¼ 670 nm, e ¼ 810 L/mol/cm, in CH3CN, typical of these tetrahedral Cu(II) complexes.16 A decrease of the intensity of this band was observed when monitoring it as a function of time. This fact is due
430
ELECTROCHEMICALLY DRIVEN MOLECULAR MACHINES
Figure 14.4 The three forms of the copper-complexed catenane, each species being either a monovalent or a divalent complex. (a) Four-coordinate complex, (b) five-coordinate
complex, and (c) six-coordinate complex. to the gliding motion of the rings to give the penta- and hexacoordinate Cu(II) complexes, whose extinction coefficients are lower as compared to the one for 2(4)2 þ (ca. 125 and 100, respectively). The final product is 2(6)2 þ , as indicated by the final spectro- and electrochemical data. A similar behavior was observed when a solution of 2(4) þ was electrochemically oxidized. When either the 2(6)2 þ solution resulting from this process or a solution prepared from a sample of isolated solid 2(6)2 þ (BF4)2 were electrochemically reduced at 1 V, the tetracoordinate catenate was quantitatively obtained. The cycle depicted in Fig. 14.3 was thus completed. The changeover process for the monovalent species is faster than the rearrangement of the Cu(II) complexes, as previously observed for the previously reported simpler catenate.16 In fact, the rate is comparable to the CV timescale, and three Cu species are detected when a CV of a CH3CN solution of 2(6)2 þ (BF4)2 is performed. The waves at þ 0.63 V and 0.41 V correspond, respectively, to the tetra- and hexacoordinate complexes mentioned above. By analogy with the value found for the previously reported copper-complexed catenane,16 the wave at 0.05 V is assigned to the pentacoordinate couple (Fig. 14.4b). 14.2.3 Intramolecular Motion Within a Heterodinuclear Bis-macrocycle Transition Metal Complex Wozniak and coworkers described recently the first heterodinuclear bismacrocyclic transition metal complex 34 þ (Fig. 14.5) that exhibits potential-driven intramolecular motion of the interlocked crown ether unit.25,26 Although the system contains transition metals, the main interaction between the various subunits, which also allowed to construct catenane 34 þ , is an acceptor–donor interaction of the charge transfer type. The reported heterodinuclear catenane should allow a controlled translocation of the crown ether unit back and forth between two different metal centers in response to an external stimulus, specifically a potential applied to the electrode (Fig. 14.6). The present system can be set in motion using two consecutive redox signals. The main feature of the machine-like catenane is that the preferred conformation
14.2
ELECTROCHEMICALLY DRIVEN MOTIONS IN COPPER-COMPLEXED CATENANES
Figure 14.5
431
Heterodinuclear [2]catenane 34 þ .
will be that for which the most electron-deficient transition metal macrocyclic complex will lie in between the two aromatic donor fragments of the crown ether. The bis-macrocyclic ring is positively charged because of the presence of Ni(II) and Cu(II). The crown ether and the bis-azamacrocyclic ring form a sandwich-like structure in such a way that one of the crown ether aromatic rings is located between the two metal-coordinated macrocyclic rings. The second aromatic ring is located almost parallel to the previous one outside the two linked macrocycles. Electrochemical characteristics of 34 þ show that the Ni(II)/Ni(III) oxidation peak appears at a more positive potential and the reduction Cu(III)/Cu(II) at a more negative potential than in the corresponding bis-macrocycle. These features result from the promoting effect of the crown ether activated by the d8–d8 structure. For such short linkers, these interactions can be observed. The nickel oxidation signals are split into two. The extent of the splitting is a function of time and temperature. It can be assumed that two different nickel centers, each with a different microenvironment, are the reason for the splitting. The donor properties of the first group of Ni(II) centers are affected by the vicinity of the electron-rich crown ether (Fig. 14.6a), while the other one is not (Fig. 14.6b). At lower scan rate, upon oxidation from Cu(II) to Cu(III), the crown ether has enough time to relocate from its initial position close to the nickel(II) center (Fig. 14.6a) toward the more positively charged copper(III) center (Fig. 14.6b). As a result of this relocation, the oxidation to nickel (III) appears at a more positive potential since it is free from the influence of the
Figure 14.6 Schematic representation of electrochemically controlled molecular motion.
432
ELECTROCHEMICALLY DRIVEN MOLECULAR MACHINES
p-electron-rich crown ether. This “frozen” interconversion within the molecule can be better observed at lower temperature or shorter timescales.25 14.3 ELECTROCHEMICALLY DRIVEN MACHINES BASED ON PIROUETTING COPPER-COMPLEXED ROTAXANES The electrochemical behavior of tetracoordinate Cu(I) complexes, that is, Cu(dpp)2based cores, is well established.27 The reversible redox potential for the Cu(II)/Cu(I) transition is around 0.6 – 0.7 V versus SCE. This relatively high potential underlines the stability of the four coordinate Cu(I) complexes versus the corresponding Cu(II) ones. The redox potential of pentacoordinate copper complexes16 is observed in a much more cathodic range. This potential shift when going from tetracoordinate to pentacoordinate copper systems is due to the better stabilization of the Cu(II) state thanks to the presence of five donor atoms in the coordination sphere. By taking avantage of this behavior, it has been possible to construct various electrochemically driven systems based on pirouetting copper-complexed rotaxanes. In this chapter, we would like to describe rotaxanes in which a new motion, pirouetting of the wheel around its axle, can be electrochemically triggered. The first distable copper-complexed rotaxanes 4(4) þ and 4(5)2 þ synthesized in our group are represented in Fig. 14.7. The wheel of the rotaxane is a bis-coordinating macrocycle
Figure 14.7
Copper-complexed rotaxanes 4(4) þ and 4(5)2 þ in motion.
14.3
ELECTROCHEMICALLY DRIVEN MACHINES BASED ON PIROUETTING
433
Figure 14.8 Cyclic voltammetry curves recorded using a Pt working electrode at a 100 mV/s sweep rate (CH3CN/CH2Cl2 (4:1), supporting electrolyte: NBu4BF4, 0.1 M, Ag wire pseudoreference). (a) Compound 4(4) þ ; (b) chemically prepared 4(5)2 þ . Curve (ii) refers to a second potential sweep following immediatly the first one (i).
containing both a bidentate moiety, a 1,10-phenanthroline, and a terdentate unit, a 2,20 ,60 ,200 -terpyridine (terpy). The axle incorporates only one 1,10-phenanthroline moiety with two stoppers linked through bis-ethoxy-ether spacers. The driving force of this motion is here again based on different geometrical preferences for Cu(I) and Cu(II). The electrochemical behavior of 4(4) þ in a CH2Cl2/CH3CN solution has been studied by CV and is represented in Fig. 14.8a. A reversible signal appears at 0.54 V. In the rotaxane 4(4) þ , where the metal is tetracoordinated, the signal occurring at 0.54 V corresponds to the tetracoordinate Cu(II)/Cu(I) couple. The ratio of the intensities of the anodic and cathodic peaks ipc/ipa is 0.95, showing that no transformation or reorganization of the coordination sphere of the tetracoordinate Cu(II) complex occurs in the timescale of the measurements (sweep rate of the potential is 100 mV/s).
434
ELECTROCHEMICALLY DRIVEN MOLECULAR MACHINES
The cyclic voltammetry behavior of the Cu(II) rotaxane, 4(5)2 þ (Fig. 14.8b), is very different from that of 4(4) þ . The potential sweep for the measurement was started at 0.9 V, a potential at which no electron transfer should occur, regardless of the nature of the surrounding of the central Cu(II) center (penta- or tetracoordinate). Curve i shows two cathodic peaks: a very small one, located at þ 0.53 V, followed by an intense one at 0.13 V. Only one anodic peak at 0.59 V appears during the reverse sweep. If a second scan ii follows immediately the first one i, the intensity of the cathodic peak at 0.53 V increases noticeably. The main cathodic peak at 0.15 V is characteristic of pentacoordinate Cu(II). Thus, in 4(5)2 þ prepared from the free rotaxane by metalation with Cu(II) ions, the central metal is coordinated to the terdentate terpyridine of the wheel and to the bidentate dpp of the axle. On the other hand, the irreversibility of this peak means that the pentacoordinate Cu(I) species formed in the diffusion layer when sweeping cathodically is transformed very rapidly and in any case before the electrode potential becomes again more anodic than the potential of the pentacoordinate Cu2 þ /Cu þ redox system. The irreversible character of the wave at 0.15 V and the appearance of an anodic peak at the value of þ 0.53 V indicate that the transient species, formed by reduction of 4(5)2 þ , has undergone a complete reorganization, which leads to a tetracoordinate copper rotaxane. The second scan ii, which follows immediately the first one i, confirms this assertion. These two complementary cyclic voltammetry experiments confirm that in this rotaxane, like in previously studied related systems, the tetracoordinate Cu(I) state is more stable than the pentacoordinate one and the pentacoordinate Cu(II) state is more stable than the tetracoordinate one. The determination of the kinetic rate constant k for the transformation of pentacoordinate Cu(I) into tetracoordinate Cu(I) has been realized by cyclic voltammetry following the method described by Nicholson and Shain.29 An average value of 17 s1 was found for k in this medium (CH3CN/CH2Cl2 (4/1), Bu4NBF4 0.1 M) at room temperature. The kinetic rate constant k0 was determined by an amperometric method for monitoring the rearrangement of tetracoordinate 4(4)2 þ into pentacoordinate 4(5)2 þ . The average value in this rearrangement is 0.007 s1. These experiments underline that the rearrangement rates from the less to the most stable geometries are drastically different for the two oxidation states of the metal. In order to increase the rate of the motions, a new rotaxane in which the metal center is as accessible as possible was prepared, the ligand set around the copper center being thus sterically little hindering compared to previous related systems. Ligand exchange within the coordination sphere of the metal is thus facilitated as much as possible. The two forms of the new bistable rotaxane, 5(4) þ and 5(5)2 þ , are depicted in Figure 14.9. The molecular axis contains a “thin” 2,20 -bipyridine motif, which is less bulky than a 1,10-phenanthroline fragment and thus is expected to spin more readily within the cavity of the ring. In addition, the bipy chelate does not bear substituents in a-position to the nitrogen atoms. 5(4) þ rearranges to the five-coordinate species 5(5)2 þ after oxidation and vice versa. The electrochemically driven motions were studied by cyclic voltammetry. A lower limit for the rate constant k of the
14.3
ELECTROCHEMICALLY DRIVEN MACHINES BASED ON PIROUETTING
Figure 14.9
435
Copper-complexed rotaxanes 5(4) þ and 5(5)2 þ in motion.
process can be estimated as >500 s1 or (t < 2 ms, with t ¼ k1).. k>500 s1
5 ð5Þ þ ! 5 ð4Þ þ The rearrangement rate for the four-coordinate Cu(II) complex is smaller than for the monovalent complex. It is nevertheless several orders of magnitude larger than in related catenanes or rotaxanes with more encumbering ligands: 5 s1
5 ð4Þ 2 þ ! 5 ð5Þ 2 þ This example shows that subtle structural factors can have a very significant influence on the general behavior (rate of the movement, in particular) of copper(II/I)based molecular machines. In order to make the copper central core as easy as possible to rotate, we thought that the bulky stoppers should be located far away from the central complex. We thus prepared and studied a new bistable rotaxane, depicted in Fig. 14.10, whose stoppers are indeed very remote from the copper center.31 This new dynamic system can indeed be set in motion more rapidly than the previously described systems. The results obtained by cyclic voltammetry clearly show that upon oxidation or reduction of the central metal copper, the macrocycle is set in motion. Upon oxidation of 6(4) þ , the resulting tetrahedrally coordinated Cu(II) is unstable as Cu(II) forms stable square planar complexes or higher coordination (five or six). Therefore, the macrocycle pirouettes around the axle permitting the restoration of a stable coordination, that is pentacoordination by the 2,20 ,60 200 -terpyridine and 20 ,20 -bipyridine
436
ELECTROCHEMICALLY DRIVEN MOLECULAR MACHINES
Figure 14.10 Copper-complexed rotaxanes 6(4) þ and 6(5)2 þ in motion.
moiety providing a stable coordination situation. The same phenomenon happens for the reduction of 6(5)2 þ : when the latter is reduced, the pentacoordinated situation 6(5) þ is unstable, the ring is set in motion leading to the preferred tetrahedral coordination of Cu(I). In order to determine the velocity of the rearrangement of the macrocycle, we applied higher scan rates to the system (2000 mV/s). With these results in hand we have shown that the macrocycle is indeed pirouetting very fast around the axle. The system moves faster than the scan rate, which implies that we can only estimate an upper limit for the half-life time of the species 6(4)2 þ .29 Using their reported equations, we find a reaction rate k ¼ 12 s1 for the rearrangement 6(4)2 þ ! 6(5)2 þ and a half-life time t1/2 ¼ 60 ms at 40 C. In agreement with previous studies on related compounds, it can be assumed that this step is about two to þ þ !6ð4Þ . We can thus three orders of magnitude slower than the rearrangement 6ð5Þ þ estimate that the half-life time of 6ð5Þ is in the range of 100 ms. Compared to previously þ , the motion of the macrocycle around the studied copper-complexed rotaxane 5ð5Þ axle is one order of magnitude faster. In conclusion, rotaxanes has been synthesized for which the stoppers attached at the ends of the axle are remote from the mobile central core and the chelate
14.4
ELECTROCHEMICALLY STEERED MACHINES
437
incorporated in the thread is sterically little hindering. These two features make the copper center very mobile and strongly facilitate ligand exchange within its coordination sphere. The mutual movements of the ring and the axle have been investigated by cyclic voltammetry, and it has been shown that the ring moves very fast around the axle (milliseconds) even at low temperature. The present system constitutes a prototype of a fast-moving molecular machine that will be used further to attach related species on electrode surfaces.
14.4 ELECTROCHEMICALLY STEERED MACHINES BASED ON TRANSLATING COPPER COMPLEXED ROTAXANES As seen in the previous part, by using a bis-coordinating macrocycles and monocoordinating threads, it has been possible to induce a rotating motion in copper-based rotaxanes. On the other hand, by using the same electrochemical principle but with mono-coordinating macrocycles and bis-coordinating threads, it is thus possible to induce a translation movement (Fig. 14.11). The reversible gliding motion of a ring along an axle leads to molecular systems called “molecular shuttles”.32 The first molecular shuttle 7(4) þ synthesized in our group33 is represented in Fig. 14.12. The cyclic part of this rotaxane is a classical 30-membered ring, containing one 1,10-phenanthroline coordinating unit, whereas its linear part contains one bidentate 1,10-phenanthroline and one tridentate terpyridine, linked together by a flexible butylene fragment. To avoid dethreading reactions, this thread is doubly ended by bulky stoppering groups.
Figure 14.11 Electrochemically induced molecular motion undergone by copper-based molecular shuttles. The stable four-coordinate monovalent complex (top left) is first oxidized to an intermediate tetrahedral divalent species (top right). This compound undergoes a complete reorganization process by translating the ring along the thread to afford the stable fivecoordinate Cu(II) complex (bottom right). Upon reduction, an unstable Cu(I) five-coordinate complex is formed (bottom left) and finally undergoes the inverse conformational change that regenerates the starting complex.
438
ELECTROCHEMICALLY DRIVEN MOLECULAR MACHINES
Figure 14.12
First copper-based molecular shuttle 7(4) þ and 7(5)5 þ .
The electrochemical and chemical behavior of rotaxane 7 þ was analyzed by CV and controlled potential electrolysis experiments.34,35 From the CV measurements at different scan rates (from 0.005 to 2 V/s) both on the copper(I) and on the copper(II) species, it could be inferred that the chemical steps (motions of the ring from the phenanthroline to the terpyridine and vice versa) are slow on the timescale of the experiments. As the two redox couples involved in these systems are separated by 0.7 V, the concentrations of the species in each environment (tetra- or pentacoordination) are directly deduced from the peak intensities of the redox signals. In Fig. 14.13 are displayed some voltammograms (curves a–e) obtained on different þ and at different times. oxidation states of the rotaxane 7ð4Þ þ in degassed Curve (a) displays the voltammogram of a red solution of 7ð4Þ acetonitrile. A reversible redox wave at 0.68 V (versus SCE) attests to the tetrahedral environment around the copper(I) center.16,17,22 During the potential scan, for rates between 0.005 and 2 V/s, no redox signal corresponding to the pentacoordination could be observed. This fact evidences the high kinetic stability of the four-coordinate copper(II) rotaxane generated at the electrode. At this stage, a controlled potential electrolysis (applied potential ¼ þ 1.0 V) was performed until 1 F was exchanged per mole of complex. During the electrolysis, the red color of the solution turned light green. Immediately after the coulometry, the voltammogram on the copper(II) species (curve (b)) showed the same redox couple at þ 0.68 Vand an additional small reversible couple at 0.33 V. These signals are characteristic of the CuIIð4Þ =CuIð4Þ and CuIIð5Þ =CuIð5Þ couples, respectively. After several hours at room temperature, without any visible color change, the progressive disappearance of the redox couple at þ 0.68 V (four-coordinate state) and the concomitant growth of the couple at 0.03 V (five-coordinate state) attest to the coordination change around the copper(II) ion (curves (c) and (d)). The analysis of the concentration of the two different copper(II)
14.4
ELECTROCHEMICALLY STEERED MACHINES
439
þ Figure 14.13 (a) CVof 7ð4Þ ; (b) CVafter electrolysis during 1 h at þ 1.0 V; (c and d) evolution 2þ of 7ð4Þ solution with time after 2 h (c) and after 4 h (d). (e) Cyclic voltammogram immediately after electrolysis of 72ð5Þþ solution at 0.3 V. Conditions: (0.1 M, n-Bu4NBF4), Pt electrode, 100 mV/s, 25 C.
species with time leads to a first-order rate constant for the following chemical reaction: 7 ð4Þ 2 þ !7 ð5Þ 2 þ ;
k ¼ 1:5 104 s1
According to the invariant shape of the signals with the scan rate of the copper(II) solution, the rate constant for the opposite movement can be maximized: 7 ð5Þ þ !7 ð4Þ þ ;
k < 102 s1
A second electrolysis at 0.3 V restores the initial red solution. The voltammogram (curve (e)) performed immediately after the reductive electrolysis of the redox couple of 7n þ is invariant with time. As all the pentacoordinate copper(I) species formed electrochemically are quantitatively transformed into tetracoordinate copper(I) species during the electrolysis, we can give a lower limit of 104 s1 for the rate constant of the chemical reaction. The residual signal at 0.03 V simply reflects an incomplete electrolysis. Due to the perfect chemical reversibility of these processes, the rotaxane 7n þ was a very promising compound in relation with electrochemically induced molecular motions. The weak point is the slowness of the chemical rearrangements, especially with the four-coordinate copper(II) complex, that is to say, the translation of the ring from the phenanthroline to the terpyridine. Considering the square scheme, depicted in Fig. 14.11, and since electron exchange is fast, it means that the kinetically limiting steps are ligand exchanges. A modification of these ligands is thus necessary to improve such a system.
440
ELECTROCHEMICALLY DRIVEN MOLECULAR MACHINES
Figure 14.14 Two different macrocycles containing a diimine chelating unit. The use of a dpp fragment as chelate, in macrocycle 8, leads to pronounced steric hinderance once a metal center is coordinated. On the other hand, by using a dpbiiq chelate, a sterically nonhindering macrocycle 9 is obtained.
The previously described studies of three pirouetting copper-based rotaxanes (4(4) þ , 5(4) þ , and 6(4) þ ), whose difference only concerns the steric hindrance of the coordinating unit on the thread, clearly demonstrate that having little hindering coordinating units is a key point to accelerate the ligand exchanges and thus make these molecular machines moving faster. The mobile ring 8 of rotaxane 7(4) þ contains a highly shielding and hindering 2,9-diphenyl-1,10-phenanthroline moiety as the complexing unit, which makes any ligand substitution within the coordination sphere of the metal center very difficult. In order to resolve this problem, a new family of chelating diimine ligands has been developed by our group. These 8,80 -diphenyl-3,30 -biisoquinoline (dpbiiq)-based ligands36,37 are nonsterically hindering, since there are no substituents a to the nitrogen atoms. They are also endocyclic, that is to say that, after cyclization of the ligand, the coordination site is located inside the ring. As shown in the Fig. 14.14, the dpbiiq-based macrocycle 9 thus obtained is larger than the former dpp-based macrocycle 8. A second molecular shuttle 10(4) þ , has been synthesized in our group38 and is represented in Fig. 14.15. The mobile ring is the nonhindering 39-membered ring 9 containing one 8,80 -diphenyl-3,30 -biisoquinoline chelate, and its linear part contains, as in previous molecular shuttle 7(4) þ , one bidentate 1,10-phenanthroline and one tridentate terpyridine units. Another difference is the phenylene group as a linker, which makes this thread more rigid. The electrochemically triggered translation of the copper-complexed ring between the dpp “station” and the terpyridine (terpy) unit was investigated by cyclic voltammetry, by analogy with the previously described copper-containing catenanes and rotaxanes. By modifying the potential scan rate, one can estimate the rate of the gliding motion undergone by the copper-complexed ring between the dpp and terpy units. A few representative cyclic voltammograms are represented in Fig. 14.16.
14.4
ELECTROCHEMICALLY STEERED MACHINES
441
Figure 14.15 Second copper-based molecular shuttle 10(4) þ , containing a sterically nonhindering macrocycle.
In a similar way to previous studies, the rearrangement rate was estimated from the shape of the CVs. In agreement with the other copper-based molecular machines made and investigated previously in our group, the unstable five-coordinate copper(I) complex 10(5) þ moves much faster than the other unstable species, namely, the fourcoordinate copper(II) complex 10(4)2 þ . This can easily be explained by considering that ligand substitution reactions are likely to be more facile around the singly charged metal center Cu(I) than around Cu(II). Whereas rearrangement of the Cu(II)complexed rotaxane is sufficiently slow to allow the gliding rate constant to be determined by the present technique, the opposite gliding motion experienced by the five-coordinate Cu(I) rotaxane is too fast to permit estimation of its rate constant. In this case, more sophisticated techniques would be required to afford a relatively precise value. We thus indicate the upper value of this rate constant only. The back-and-forth motion can be described by the following equations: 10 ð4Þ þ e !10 ð4Þ 2 þ 10 ð4Þ 2 þ !10 ð5Þ 2 þ
gliding;
k ¼ 2 s1
10 ð5Þ 2 þ þ e !10 ð5Þ þ 10 ð5Þ þ !10 ð4Þ þ
gliding;
k > 50 s1
442
ELECTROCHEMICALLY DRIVEN MOLECULAR MACHINES
Figure 14.16 Cyclic voltammetry study of rotaxane 10(4) þ . (a), (b), and (c): potential range: 0.4 V to 1.0 V, followed by 1.0 to 0.4 V. (d) Two consecutive scans. The electrochemical experiments have been performed at room temperature, in a 0.1 M solution of Bu4NBF4 in CH3CNCH2Cl2 (9:1), with a Pt working electrode, Ag wire as a pseudo-reference electrode, and Pt wire as a counterelectrode.
By comparing the electrochemical behavior of 10(4) þ and that of 7(4) þ , it is obvious that there exists a pronounced kinetic biisoquinoline effect. This ligand leads to a markedly more mobile electrochemically driven machine than the previous copper-based shuttle. The endocyclic but nonsterically protecting nor hindering nature of dpbiiq is with no doubt responsible for this spectacular improvement. In order to compare the dynamic properties of 10(4) þ and 10(5)2 þ to those of a molecule displaying as much similarity to 10(4) þ as possible in terms of chemical function, we also prepared and studied the rotaxane 11(4) þ , depicted in Fig. 14.17. 11(4) þ contains exactly the same axis and stoppers as 10(4) þ but the mobile ring is now the strongly shielding macrocycle 8. This allows to assess the effect of replacing the dpp-containing ring 8, a classical building block of our group, by the recently prepared dpbiiq-comprising macrocycle 9. The difference is remarkable.
14.5
Figure 14.17
CONCLUSION
443
Chemical structure of rotaxane 11(4) þ .
As discussed above, the unstable four-coordinate copper(II) complex 10(4)2 þ rearranges within less than 1 s. By contrast, after oxidation of 11(4) þ to 11(4)2 þ , the thermodynamically unstable form of the complex seems to be stable for several hours, also showing that the axis of 10(4) þ and 11(4) þ with its rigid purely aromatic connector between the phen and terpy fragments is much less favorable to fast gliding than the flexible axis originally used to prepare 7(4) þ . To sum up, a first copper-based molecular shuttle 7(4) þ has been synthesized and studied, revealing a very promising system thanks to the perfect chemical reversibility of the chemical and electrochemical processes. But gliding motions of the ring, that is to say ligand exchanges, were too slow (hours) due to the sterically hindering dpp-based macrocycle 8. A nonhindering dpbiiq-based macrocycle 9 has been developed and then used for the formation of a new rigid copper-based molecular shuttle 10(4) þ , which can move at the milliseconds to seconds timescale. A direct comparison with the former macrocycle 8, involved in rotaxane 11(4) þ , revealed that the biisoquinoline-based macrocycle 9 is about four orders of magnitude faster.
14.5
CONCLUSION
In this chapter, we have discussed a few examples of transition metal-containing molecular machines belonging to the catenae and rotaxane family. The compounds are set in motion using an electrochemical signal, the redox process being most of the time centered on the metal. It is still not sure whether the field will lead to applications in a short-term prospective, although spectacular results have been obtained in the course of the last few years in relation to information storage and processing at the molecular level.39 Apart from this important field, other applications could of course be envisaged: nanomachines for medical applications such as valves and pumps, molecular devices able to sort molecules, active molecular carriers able to transport selectively given molecules across a membranes, just to cite a few. No doubt that the electrochemical approach is particularly attractive and has already afforded spectacular examples of molecular machines and even practical devices. Nevertheless, other strategies based on other types of input (photonic, chemical,
444
ELECTROCHEMICALLY DRIVEN MOLECULAR MACHINES
magnetic, electric, etc.) are certainly worthwhile investigating. The photochemical approach has also produced highly promising devices based on noninterlocking compounds as well as on rotaxanes. From a purely scientific viewpoint, the field of molecular machines is particularly challenging and motivating: the fabrication of dynamic molecular systems, with precisely designed dynamic properties, is still in its infancy and will certainly continue to experience a rapid development during the next decades.
REFERENCES 1. E. R. Kay, D. A. Leigh, F. Zerbetto, Angew. Chem., Int. Ed. 2007, 46, 72–191. 2. J.-P. Sauvage (Ed.), Molecular Machines and Motors, Springer, Berlin, 2001. 3. V. Balzani, A. Credi, M. Venturi, Molecular Devices and Machines: A Journey into the Nano World, Wiley-VCH, Weinheim, 2003. 4. B. L. Feringa, Nature 2000, 408, 151–154. 5. J. E. Walker, Nobel Lecture, Angew. Chem., Int. Ed. 1998, 37, 2308–2319. 6. I. Rayment, H. M. Holden, M. Whittaker, C. B. Yohn, M. Lorenz, K. C. Holmes, R. A. Milligan, Science 1993, 261, 58–65. 7. N. Hirokawa, Science 1998, 279, 519–526. 8. V. Balzani, M. Clemente-Leon, A. Credi, B. Ferrer, M. Venturi, A. H. Flood, J. F. Stoddart, Proc. Natl. Acad. Sci. USA 2006, 103, 1178–1183. 9. D. A. Leigh, J. K. Y. Wong, F. Dehez, F. Zerbetto, Nature 2003, 424, 174–179. 10. V. Balzani, A. Credi, F. Marchioni, J. F. Stoddart, Chem. Commun. 2001, 1860–1861. 11. M. C. Jimenez, C. Dietrich-Buchecker, J.-P. Sauvage, Angew. Chem., Int. Ed. 2000, 39, 3284–3287. 12. N. Koumura, R. W. J. Zijistra, R. A. Van Delden, N. Harada, B. L. Feringa, Nature 1999, 401, 152–155. 13. V. Balzani, A. Credi, M. Venturi, Molecular Devices and Machines: A Journey into the Nano World, Wiley-VCH, Weinheim, 2003. 14. K. Tashiro, K. Konishi, T. Aida, Angew. Chem., Int. Ed. 1997, 109, 882–884. 15. A. Carella, G. Rapenne, J.-P. Launay, New J. Chem. 2003, 29, 288–290. 16. A. Livoreil, C. O. Dietrich-Buchecker, J.-P. Sauvage, J. Am. Chem. Soc. 1994, 116, 9399–9400. 17. F. Baumann, A. Livoreil, W. Kaim, J.-P. Sauvage, Chem. Commun. 1997, 35–36. 18. D. A. Parthenopoulos, P. M. Rentzepis, Science 1989, 245, 843–845. 19. I. Willner, R. Blonder, A. Dagan, J. Am. Chem. Soc. 1994, 116, 3121–3122. 20. M. Irie, O. Miyatake, K. Uchida, J. Am. Chem. Soc. 1992, 114, 8715–8716. 21. S. L. Gilat, S. H. Kawai, J.-M. Lehn, Chem. Commun. 1993, 1439–1442. 22. D. Cardenas, A. Livoreil, J.-P. Sauvage, J. Am. Chem. Soc. 1996, 118, 11980–11981. 23. C. O. Dietrich-Buchecker, J.-P. Sauvage, J.-P. Kintzinger, Tetrahedron Lett. 1983, 24, 5095–5098. 24. C. O. Dietrich-Buchecker, J.-P. Sauvage, Tetrahedron 1990, 46, 503–512.
REFERENCES
445
25. B. Korybut-Daszkiewicz, A. Wie¸ckowska, R. Bilewicz, S. Domagata, K. Wozniak, Angew. Chem., Int. Ed. 2004, 43, 1668–1672. 26. B. Korybut-Daszkiewicz, A. Wie¸ckowska, R. Bilewicz, S. Domagata, K. Wozniak, J.Am. Chem. Soc. 2001, 123, 9356–9366. 27. C. O. Dietrich-Buchecker, J.-M. Kern, J.-P. Sauvage, J. Am. Chem. Soc. 1999, 111, 7791– 7800. 28. L. Raehm, J.-M. Kern, J.-P. Sauvage, Chem. Eur. J. 1999, 5, 3310–3317. 29. R. S. Nicholson, I. Shain, Anal. Chem. 1964, 36, 706–723. 30. I. Poleschak, J.-M. Kern, J.-P. Sauvage, Chem. Commun. 2004, 474–476. 31. U. Letinois-Halbes, D. Hanss, J. M. Beierle, J.-P. Collin, J.-P. Sauvage Org. Lett. 2005, 7, 5753–5756. 32. R. A. Bissell, E. Co´rdova, A. E. Kaifer, J. F. Stoddart, Nature 1994, 369, 133–137. 33. P. Gavin˜a, J.-P. Sauvage, Tetrahedron Lett. 1997, 38, 3521–3524. 34. J.-P. Collin, P. Gavin˜a, J.-P. Sauvage, J. Chem. Soc., Chem. Commun. 1996, 2005–2006. 35. J.-P. Collin, P. Gavin˜a, J.-P. Sauvage, New J. Chem. 1997, 21, 525–528. 36. F. Durola, J.-P. Sauvage, O. S. Wenger, Chem. Commun. 2006, 171–173. 37. F. Durola, D. Hanss, P. Roesel, J.-P. Sauvage, O. S. Wenger, Eur. J. Org. Chem. 2007, 125–135. 38. F. Durola, J.-P. Sauvage, Angew. Chem., Int. Ed. 2007, 46, 3537–3540. 39. J. E. Green, J. W. Choi, A. Boukai, Y. Bunimovich, E. Johnston-Halperin, E. Delonno, B. A. Sheriff, K. Xu, Y. S. Shin, H.-R. Tseng, J. F. Stoddart, J. R. Heath, Nature 2007, 445, 414–417.
CHAPTER 15
Electroactive Molecules and Supramolecules for Information Processing and Storage GUANXIN ZHANG, DEQING ZHANG, AND DAOBEN ZHU Beijing National Laboratory for Molecular Sciences, Organic Solids Laboratory, Institute of Chemistry, Chinese Academy of Sciences, Beijing, China
15.1
INTRODUCTION
Two philosophically different approaches have been proposed for information processing at the molecular level. The first one is to mimic, at the nanometer scale, the operational principles of solid-state computers presently in use. This approach is based on molecular electronics, in which both input and output signals are electronic ones (electron fluxes). Molecular photonics based on photon fluxes can help in this way. The second approach that takes inspiration from information processing in living organisms is based on chemionics1 in which molecules and ions can be used as input/ output signals to process information by using a molecular substrate. Chemionics usually operates in solution and can be complemented by photonics since chemical and light input/output signals can be easily coupled. Within each aspect of molecular electronics, photonics, or chemionics, information processing takes place at logic gates and data manipulation relies on the binary digital (bit) nature of these input and output signals that are elaborated by means of the Boolean logic. Although significant progress has been achieved for molecular electronics in recent years, thanks to the advancement of nanotechnology, photonics and chemionics have also received more and more attention.1 In fact, information processing in nature is usually based on exchange of chemical signals, exemplified by neurons in our brain, which process signals relying on the behaviors of ions in solution. A number of functional molecules and supramolecules have been reported to mimic functions
Electrochemistry of Functional Supramolecular Systems. Edited by Paola Ceroni, Alberto Credi, and Margherita Venturi Copyright 2010 John Wiley & Sons, Inc.
447
448
ELECTROACTIVE MOLECULES AND SUPRAMOLECULES
of switches and logic gates. Various interesting molecular switches and logic gates have been described for information processing at the molecular level.1–5 Meanwhile, information storage has been becoming a central issue in this digital age. The past few decades has witnessed the explosive increasing of information and the remarkable miniaturization of electronic devices. Such trends continue to demand the increase of the bit area density and storage capacity to overcome the intrinsic physical limitations on memory device components. To fulfill this, vast efforts have been devoted to exploring and developing new recording technologies and materials that combine high density, fast response time, long retention time, rewriting capability, etc. Among the various recording media for high-density data storage, organic materials, due to their good stimuli-responsive properties and versatility in molecular design, are especially attractive in recent years and suggested as promising candidates. In principle, the basic requirements for recording materials are that they should possess at least two distinct stable states accessible via an external stimulus, where each state can represent “0” or “1” of a digital mode, and the states can be clearly distinguished during read-out. In this chapter, we will discuss the recent results regarding the development of molecular switches and logic gates toward information processing at the molecular level based on electroactive molecules and supramolecules. Also, we will illustrate the progress of application of electron donor–acceptor compounds in high-density information storage.
15.2
REDOX-CONTROLLED MOLECULAR SWITCHES
Photoinduced electron transfer processes involving electron donor (D) and acceptor (A) components can be tuned via redox reactions. Namely, the excitedstate properties of fluorophores can be manipulated by either oxidation of electron donors or reduction of electron acceptors. Also, the oxidized and the reduced species show different properties compared to the respective electron donors and acceptors. By making use of these properties of electron donors and acceptors, a number of molecular switches and logic gates have been described in recent years. In the following, we will introduce these redox-controlled molecular switches according to the redox centers. 15.2.1
Tetrathiafulvalene-Based Molecular Switches
Tetrathiafulvalene and its derivatives are electroactive and can be easily and revers. ibly oxidized to TTF þ and TTF2 þ . The TTF skeleton now occupies a critical position as far as switchable properties are concerned, and behaves as a key unit for a number of supramolecular concepts. For instance, the recent years have seen an increasing contribution of TTF to the preparation of interlocked compounds such as rotaxanes and catenanes. These systems are of particular importance as candidates for molecular machines.
15.2 REDOX-CONTROLLED MOLECULAR SWITCHES
Scheme 15.1 (D ¼ TTF).
449
Principle of a switchable, redox-controlled fluorescence in D–A systems
TTF-based D–A systems have been extensively used in recent years to play around photoinduced electron transfer processes. Typically, when an electron acceptor moiety that emits fluorescence intrinsically is linked to TTF (D), the fluorescence due to the A moiety may be quenched because of a photoinduced electron transfer process (Scheme 15.1). Accordingly, these molecular systems are potentially interesting for photovoltaic studies. For instance, efficient photoinduced electron transfer and charge separation were reported for TTF-fullerene dyads.6,7 An important added value provided by TTF relies on the redox behavior of this unit that can be reversibly oxidized according to two successive redox steps. Therefore, such TTF-A assemblies allow an efficient entry to redox fluorescence switches, for which the fluorescent state of the fluorophore A can be reversibly switched on upon oxidation of the TTF unit. Naturally, considering the intrinsic fluorescence properties of polyaromatic systems, substantial efforts can be found in recent literatures for their grafting to redox-active TTF. Anthracene emits in the range of 380–450 nm; thus, the fluorescence spectrum of anthracene has rather small overlap with the absorption spectrum . of TTF þ .8 It is anticipated that the fluorescence of the TTF-anthracene dyad can be reversibly modulated by redox reactions of TTF moiety. With this consideration in mind, we designed and synthesized the TTF-anthracene triad 1,9 which exhibit rather weak fluorescence. Following either chemical or electrochemical oxidation, large fluorescence enhancement is observed. As an example, Fig. 15.1a shows the fluorescence spectra of triad 1 after the solution was oxidized by applying a potential of 0.7 V versus Ag/AgCl for different periods. Moreover, the fluorescence intensity decreases gradually if the oxidized solution was subjected to reduction by applying
Figure 15.1 Variation of fluorescence spectra of the solution of triad 1 oxidized by applying an oxidation potential (a) and then followed applied a reduction potential (b).
450
ELECTROACTIVE MOLECULES AND SUPRAMOLECULES
a potential at 0.2 V (Fig. 15.1b). Thus, the fluorescence intensity of triad 1 can be reversibly modulated for several cycles. The fluorescence of the triad with TTF and pyrene moieties (2)10 can also be tuned depending on the redox state of the TTF moiety.
The TTF-porphyrin dyad 3 was described by the group of Odense.11 The fluorescence of 3 is significantly quenched by the photoinduced electron transfer process. Notably, the fluorescence intensity of dyad 3 increases largely after . . addition of Fe3 þ that oxidizes TTF into TTF þ . Successive reduction of TTF þ is not reported. Nevertheless, it is anticipated that the fluorescence of dyad 3 can be reversibly modulated by redox reactions. In fact, the fluorescence of the supramolecule 4, formed between Zn-tetraphenylporphyrin and a pyridinesubstituted TTF (TTF-Py), can be reversibly tuned by sequential oxidation and reduction of the TTF moiety in 4.12 It should be noted in this context that the synthetically challenging system associating a porphyrin ring fused to four TTFs (5) was also reported.13
The TTF-tetrachloride perylene diimide dyad 6 has been recently described.14,15 . The TTF moiety of dyad 6 is oxidized into the corresponding TTF þ and TTF2 þ by applying potentials of 0.85 and 1.25 V versus Ag/AgCl, respectively. The fluorescence intensity of tetrachloride perylene diimide at 545 nm decreases slightly . after oxidation of TTF into TTF þ , but interestingly increases gradually after . þ to TTF2 þ .14 Therefore, the fluorescence of dyad 6 is further oxidation of TTF directly dependent on the oxidation state of TTF moiety. Consequently, this dyad can be considered as a new reversible fluorescence-redox dependent molecular system. Similar tuning of fluorescence was also described for TTF-MPT
15.2 REDOX-CONTROLLED MOLECULAR SWITCHES
451
(5-methoxy-2-pyridylthiazole) dyad 7.16 Both TTF-MPT and TTF þ -MPT shows weak fluorescence, but significant fluorescence enhancement is observed for TTF2 þ MPT. Moreover, the variation of the fluorescence of TTF-MPT in response to metal ions was also investigated. The corresponding fluorescence variation behaviors mimic the functions of several logic gates. .
4-(N,N-Dimethylamino)benzonitrile (DMABN) and its derivatives, as a class of organic donor–acceptor compounds, exhibit dual fluorescence, one related to the local excited state (“B” band) and the other ascribed to the twisted intramolecular charge transfer (TICT) state (“A” band).17 As expected, compound 818 exhibits dual fluorescence, showing two fluorescence bands centered at 350 and 432 nm, which can be ascribed to the corresponding B band (from the local excited state) and A band (from the TICT state), respectively. After oxidation of TTF unit in 8, the fluorescence intensity of A band decreases while that of B band increases slightly. As expected, . further reduction of TTF þ into neutral TTF unit leads to the restoration of the fluorescence spectrum of 8. Therefore, the dual fluorescence spectrum of 8 can be reversibly modulated by redox reactions of TTF unit in 8. Such fluorescence switching behavior is based on the manipulation of the intramolecular energy transfer process upon redox reactions of TTF unit in 8. The neutral TTF unit shows strong absorption around 350 nm, while the corresponding radical cation absorbs in the range of 400–550 nm besides in the range of 600–1000 nm. Therefore, the B band of the fluorescence spectrum of 8 with a neutral TTF unit would be quenched to some extent due to the intramolecular energy transfer since there is a spectral overlap between the fluorescence B band and the absorption spectrum of TTF unit. On the other hand, the A band of the fluorescence spectrum . of 8 after oxidation (with TTF þ unit) would decrease because there is a spectral overlap between the fluorescence A band and the absorption spectrum of the radical cation of TTF unit and thus intramolecular energy transfer would take place. An influence of oxidation of TTF unit on the ground/excited state conformation of DMABN unit of 8 may also contribute to the dual fluorescence variation observed for 8 after oxidation of TTF unit.18 Oxidation of TTF and its derivatives induces the transformation from neutral . species into cationic ones, namely, cation radicals (TTF þ ) and dications (TTF2 þ ). . þ Moreover, TTF, TTF , and TTF2 þ exhibit different absorption spectra. Taking these advantages of TTF new TTF-based redox fluorescence switches and chiral switches have been recently reported. As stated above, the electron donating ability of TTF can be tuned by redox . reactions; compared to TTF, TTF þ , and TTF2 þ became poor electron donors. In fact, TTF2 þ can function as electron acceptor. Balzani, Williams, Stoddart,
452
ELECTROACTIVE MOLECULES AND SUPRAMOLECULES
and coworkers have reported an elegant three-pole supramolecular switch by making use of this property of TTF.19 This supramolecular switch is based on a three-component system consisting of tetracationic cyclophane cyclobis (paraquatp-phenylene) (CBPQT4 þ ) as electron accepting host, macrocyclic polyether 1,5dinaphtho-[38]crown-10 (1/5DN38C10) as electron donating host and TTF. As expected, neutral TTF (TTF0) can form a 1 : 1 inclusion complexes with CBPQT4 þ that can be dissociated/reassociated reversibly by cyclic oxidation/ reduction of TTF. UV absorption, 1 H NMR, and MS as well as X-ray structural analysis confirm that a 1 : 1 inclusion complex is formed between TTF2 þ as an . electron acceptor and 1/5DN38C10. By contrast, TTF þ cannot be bound by either . of the two hosts. Since the reversible interconversion among TTF, TTF þ , and 2þ TTF can be easily achieved by either redox reactions, a three-state switch can be constructed with this three-component supramolecular system as schematically shown in Scheme 15.2. The three states can be shifted reversibly by electrochemical stimuli. The three states show different absorption spectra and thus they can be easily distinguished. These results also suggest that this three-component mixture may form the basis of an electrochromic display system. Chiral binaphthalene shows strong CD signal and the intensity is strongly dependent on the dihedral angle of the two naphthalene rings. With the respective features of binaphthalene and TTF in mind, Zhang and coworkers designed and investigated binaphthalene-TTFs assemblies (S)-9a, (S)-9b, and (S)-9c.20 It is expected that the dihedral angle on the chiral binaphthalene framework can be tuned according to the redox state of TTF units. The CD signal intensity of (S)-9a at 220 and 240 nm successively increases and decreases after oxidation at the
Scheme 15.2 Cartoon displaying the electrochemically reversible three-pole supramolecular switch based on the three-component mixture CBPQT4 þ -1/5DN38C10-TTF. (See the color version of this figure in Color Plates section.)
15.2 REDOX-CONTROLLED MOLECULAR SWITCHES
453
corresponding cation radical and dication stages, respectively. For (S)-9b and (S)-9c, CD signal intensities at 235 and 250 nm decreases after sequential oxidation . of TTF into TTF þ /TTF2 þ ; but, reversible variation of CD spectrum is also observed. Chiral binaphthalene (R)-10 with two substituted TTF and trichloroquinone units was synthesized.21 It is expected from such system, that the variation of the D–A interaction in the presence of metal ions may induce a change of the dihedral angle between the two naphthalene rings. Both absorption and ESR spectral studies show that electron transfer occurs between TTF and trichloroquinone units in the presence of metal ions. Nevertheless, in contrast to expectation, the CD spectral change of (R)-10 is rather small after addition of metal ions. A chiral TTF-substituted poly(isocyanide) 11 was prepared by Amabilino et al.22 It is interesting to note that the CD spectrum of 11 can be tuned by changing the oxidation states of the TTF units in the side-chains of 11. Polymer 11 shows reversible interconversion between three univalent and two very broad mixed-valence redox states that have different chiroptical properties.
TTF and its derivatives are electroactive and, as strong p-donors, their redox potentials are sensitive to their environment. Heterocyclic compounds 12a-c containing TTF and azobenzene moieties were synthesized, aimed at modulating the electron donating ability of TTF unit by light irradiation.23 It is anticipated that the transformation of the trans-azobenzene to the corresponding cis-azobenzene unit would alter the steric strain within compounds 12a-c; as a result, the conformation of TTF unit and therefore its electron donating ability would be changed. For compounds 12a and 12b with short alkyl chain as the spacers the first oxidation potential of TTF unit can be reversible tuned by alternating UV and visible light irradiations. Conversely, the first oxidation potential of TTF unit keeps unchanged by UV and visible light irradiations for 12c, because of the lack of steric strain whatever the azo-benzene configuration, due to the longer spacers. Variation of the oxidation potential is also found to be negligible for compound 13 in which the azobenzene moiety is linked to vicinal positions on the TTF moiety.
454
ELECTROACTIVE MOLECULES AND SUPRAMOLECULES
Several TTF derivatives with photoresponsive moieties have been described in recent years with a view to studying the interplay of photochromic and electrochemical properties. Uchida, Irie, and coworkers have reported TTF derivatives with diarylethene moieties such as 14.24 The corresponding open- and closed-isomers of 14 can be reversibly interconverted by UV and visible light irradiations. The results show that the open- and closed-isomers of 14 show different electrochemical responses. 15.2.2
Ferrocene-Based Molecular Switches
Ferrocene has been widely investigated as an electron donor and its electron donating ability can be tuned by redox reactions. As anticipated, when a ferrocene unit is covalently connected to an electron acceptor moiety that shows intrinsic fluorescence, the fluorescence of the acceptor moiety would be largely quenched because of the photoinduced electron transfer between ferrocene and the fluorescent acceptor. For instance, triad 15 that contains perylene diimide flanked by two ferrocene moieties, shows rather weak fluorescence due to the photoinduced electron transfer between perylene diimide and ferrocene units. Either chemical or electrochemical oxidation of ferrocene unit lead to fluorescence enhancement. This is simply because the electron donating ability of ferrocene is reduced after oxidation and accordingly the photoinduced electron transfer is prohibited. In this way, the fluorescence intensity of 15 can be reversibly modulated by sequential electrochemical oxidation and reduction. Therefore, a new redox fluorescence switch can be established with triad 15.25 The efficient on/off switching of fluorescence from substituted zinc porphyrinferrocene dyads 16a and 16b is achieved through redox control of the excited-state electron transfer quenching.26 This redox fluorescence switch is based on the switching of the excited-state electron transfer from the ferrocene to the zinc porphyrin through the use of the ferrocene/ferrocenium (Fc/Fc þ ) redox couple. Similarly, the fluorescence intensity of the 1,4-disubstituted azine with ferrocene and pyrene units (17) can be reversibly modulated by sequential redox reactions of ferrocene moiety. In the neutral state, compound 17 displays weak fluorescence owing to the electron transfer from the ferrocenyl group to the excited pyrene unit or by energy transfer from the excited pyrene unit to the ferrocenyl unit. Oxidation of the ferrocenyl unit, however, leads to remarkable fluorescence enhancement. This is because the ferrocenium cation shows weak electron donating ability and also the corresponding spectral overlap becomes small.27
15.2 REDOX-CONTROLLED MOLECULAR SWITCHES
455
Ferrocenyl groups have also been linked to helical and chiral polyisocyanide 18 as side chains. Upon electrochemical oxidation of ferrocenyl groups in polyisocyanide 18 into the corresponding cations, the CD signal at 360 nm decreases and that at 250 nm almost disappears. Interestingly, the CD spectrum of the oxidized polyisocyanide 18 can be recovered by electrochemical reduction. Similar CD spectral changes can be achieved by chemical oxidation and reduction. As a result, a redox chiral switch can be realized with polyisocyanide 18.28
15.2.3
Quinone and Analogs-Based Molecular Switches
Fluorescent redox switches based on compounds with electron acceptors and fluorophores have been also reported. For instance, by making use of the quinone/ hydroquinone redox couple a redox-responsive fluorescence switch can be established with molecule 19 containing a ruthenium tris(bpy) (bpy ¼ 2,20 -bipyridine) complex.29 Within molecule 19, the excited state of the ruthenium center, that is, the triplet metal-to-ligand charge transfer (MLCT) state, is effectively quenched by electron transfer to the quinone group. When the quinone is reduced to the hydroquinone either chemically or electrochemically, luminescence is emitted from the ruthenium center in molecule 19. Similarly, molecule 20, a ruthenium (II) complex with hydroquinone-functionalized 2,20 :60 ,200 -terpyridine (tpy) and (40 -phenylethynyl2,20 :60 ,200 - terpyridine) as ligands, also works as a redox fluorescence switch.30 Another redox switchable system is based on dyad 21 in which 2-chloro-1,4naphthoquinone is covalently attached to 5-dimethyl-aminonaphthalene via a nonconjugated spacer. The intrinsic fluorescence of the dansyl excited state in dyad 21 is strongly quenched, due to the intramolecular electron transfer from the excited dansyl to the adjacent quinone acceptor. However, the fluorescence can be switched on by addition of a reducing agent. Apart from chemical switching, the fluorescence of dyad 21 can also be switched electrochemically. This can be realized using a photoelectrochemical cell, and the solution starts to fluoresce upon application of a reductive potential.31
456
ELECTROACTIVE MOLECULES AND SUPRAMOLECULES
The anthraquinone exhibits a similar redox behavior as benzoquinone. Thus, redox luminescence switch can also be constructed with fluorophore linked to anthraquinone. For example, the luminescence of molecule 22, a ruthenium complex with an appended anthraquinone moiety, can be reversibly tuned through the interconversion between the anthraquinone and the corresponding hydroquinone.32 Nicotinamide is an important redox moiety in biological system. The nicotinamide-perylene diimide dyad 23 can work as a redox-responsive fluorescence switch.33 Dyad 23, in which nicotinamide is on the oxidation state, exhibits strong fluorescence. However, it becomes nonfluorescent when nicotinamide is reduced due to the electron transfer from the reduced nicotinamide to the photoexcited perylene diimide. The fluorescence of dyad 23 can be reversibly switched off and on chemically by successive reduction with NaBH3CN and oxidation with tetrachlorobenzoquinone and switched electrochemically over 10 cycles without significant degradation. 15.2.4
Viologen-Based Molecular Switches
Methyl viologen (N,N0 -dimethyl-4,40 -bipyridinium dication, MV2 þ ) can function as an electron acceptor.34 When MV2 þ is linked to electron donor, photoinduced electron transfer would occur. For example, within molecule 24 the 3 MLCT excited state of [Ru(bpy)3]2 þ is quenched by MV2 þ through oxidative electron transfer . process. The excited state of [Ru(bpy)3]2 þ can also be quenched by MV þ and MV0. The transient absorption spectroscopic investigations show that the quenching of the . excited state of [Ru(bpy)3]2 þ by MV þ and MV0 is due to the reductive electron transfer process. Thus, the direction of the photoinduced electron transfer within molecule 24 is dependent on the redox state of MV2 þ , which can be switched by redox reactions induced chemically or electrochemically. This demonstrates the potential of molecule 24 as a redox switchable photodiode.35
15.2 REDOX-CONTROLLED MOLECULAR SWITCHES
Scheme 15.3 oxidation.
457
Electrochemically induced translocation of the CBPQT4 þ upon reduction or
The derivative of MV2 þ , cyclobis(paraquat-p-phenylene) (CBPQT4 þ ), has been widely used in construction of redox-responsive rotaxanes and catenanes for molecular machines.36 This is usually based on the following two facts: (1) CBPQT4 þ can form charge transfer complexes with suitable electron donors and (2) after either reduction of CBPQT4 þ or oxidation of the electron donor, the charge transfer complex would be destabilized and as a result the complex is dissociated and thus movement of CBPQT4 þ occurs. For example, Willner and coworkers reported the rotaxane monolayer on a Au electrode consisting of CBPQT4 þ threaded on a “molecular string” that includes a p-donor diiminobenzene unit and stoppered by an adamantine unit (Scheme 15.3).37 Initially, CBPQT4 þ forms charge transfer complex with diiminobenzene. Reduction of CBPQT4 þ to the corresponding biradical dication results in its dissociation from the diiminobenzene site, and the reduced cyclophane is translocated toward the electrode. Oxidation of the reduced cyclophane reorganizes the cyclophane on the diiminobenzene site. Accordingly, the position of CBPQT4 þ on the “molecular string” of rotaxane, corresponding to different chronoamperometric and impedance signals, can be switched by redox reactions. Moreover, the electrochemical shuttling of the redox-active cyclophane on the “molecular string” can also control the hydrophobic/hydrophilic properties of the interface. Another feature of MV2 þ is that its single-electron-reduced radical cation, referred to as violene, exhibits low-energy absorbing properties with lonset ¼ 800 nm. The chiral binaphthalene 26 fused with two MV2 þ units has been studied experimentally and theoretically for creating a new chiroptical switch.38,39 The CD signal intensities at 410 and 660 nm can be reversibly modulated by redox reaction of the MV2 þ moiety. Also, the optical rotation can be switched in the wavelength range of 1000–2000 nm upon redox cycles between its violene and tetracationic states. It should be noted that chiroptical switching materials are expected to find polarization-related photonics applications such as data storage, optical switches, and light modulators.
458
15.2.5
ELECTROACTIVE MOLECULES AND SUPRAMOLECULES
Redox/Photodual Mode Molecular Switches
1,2-Dithienylcyclopentene and its derivatives have received significant attentions in recent years because they show appealing photochromic properties such as thermal stability of the ring-open and ring-closed forms, and fatigue resistance in the photochemical ring-closing and ring-opening processes. As a result, they show promising applications in many areas such as optical recording. Branda et al. have recently discovered that the interconversion between the ring-open and the ringclosed forms of 1,2-dithienylethenes can also be triggered by redox reactions. UV light irradiation of the ring-open form of 27 leads to the formation of the ring-closed form of 27. Electrochemical studies indicate that oxidation of the ringclosed form of 27 can induce the ring opening process to generate the ring-open form. The ring opening process can be easily followed by 1 H NMR spectroscopy. The transformation of the ring-closed form into the open form can also be achieved by chemical oxidation as shown in Fig. 15.2, where the gradual color change of the solution of 27 containing 75% of the closed form is displayed after addition of [(4-BrC6H4)3N][SbCl6]. Therefore, it can be concluded that the interconversion between 27o and 27c can be realized by UV light irradiation and chemical or electrochemical oxidation. In fact, such dual-mode photochemical-ring-closing/ oxidation-ring-opening process can be recycled without any significant degradation for 10 cycles.40 Even more interesting is that this redox-opening process is catalytic. Only a small fraction of the ring-closed form 27c present in solution is required to undergo the redox reaction to completely drive the ring-opening reaction. As soon as the radical cation of 27c is generated, it will ring open to 27o, which will subsequently remove an electron from another molecule of 27c and regenerate the original radical cation.
Figure 15.2 Gradual color change of a CH2Cl2 solution of compound 27 containing 75% of the ring-closed isomer 27c when treated with a catalytic amount of [(4-BrC6H4)3N][SbCl6].
15.2 REDOX-CONTROLLED MOLECULAR SWITCHES
459
Such oxidation-driven ring-opening can also occur for the closed forms of 29, 30, and 31, but not for the closed form of 28. These results imply that the presence of aromatic rings connected onto the two carbon atoms of the thiophene heterocycles where the new CC single bond formed is the critical structural requirement for this ring-open reaction triggered by oxidation. Photochromic dithenylmaleimide 32 contains two ferrocene units.41 Electrochemical absorption spectral studies indicate that oxidation of ferrocenyl units can induce ring-opening reaction for the ring-closed isomer of 32.
Although the oxidation ring-opening reaction cannot occur for the ring-closed form of 33, the ring-closing reactions can be triggered by electrochemical oxidation for 33o (Scheme 15.4). The electrochemical ring-closing reaction can be monitored by absorption spectroscopy. When the solution of 33o is electrolyzed, the absorption band around 600 nm, corresponding to the absorption of the ring-closed form of 33c, increases gradually.42 1,2-Dithienylcyclopentene derivative 34, which contains cationic pyridinium units and a photochromic compound, was initially designed to demonstrate that the redox properties of 34 can be optically modulated. Branda et al. have recently shown that the electrochemical reduction of the ring-open isomer of 34 results in cyclization to generate the ring-closed isomer of 34. Both 1 H NMR and absorption spectroscopic studies of the solution of ring-open isomer of 34, which has been electrochemically reduced, followed by addition of Ag þ , indicate the formation of the ring-closed isomer of 34. Furthermore, the 1 H NMR results also suggest the formation of an additional ring-closed isomer, cis-34c as shown in Scheme 15.5.43 Dithienylethenes 35o2 þ -36o2 þ possess the structural features of 27 and 34.44 As discussed above, the introduction of N-methyl pyridinium group enables the reductive ring closing while aromatic groups such as thiophene on the C2 position results in oxidative ring opening. It is so rationally designed with a view to developing
Scheme 15.4 The reversible interconversion between the ring-open and ring-closed forms of 33 triggered by light irradiations and electrochemical oxidation.
460
ELECTROACTIVE MOLECULES AND SUPRAMOLECULES
Scheme 15.5 Illustration of the formation of the trans- and cis-ring-closed isomers of 34 by electrochemical ring-closing reaction.
a dithienylethene system in which both ring opening and ring closing can be triggered by redox reactions, thus exhibiting bidirectional electrochromism. The bidirectional electrochromism of 35o2 þ -36o2 þ can be visually demonstrated by subjecting a solution of 35o2 þ -36o2 þ to sequential electrolyses at potentials corresponding to the reduction and oxidation reaction determined from the cyclic voltammetry experiments. The absorption spectrum of 35o2 þ -36o2 þ can be reversibly modulated by applying an alternate reductive and oxidative stimuli. The above results clearly indicate that it is possible to maximize the performance of the photochromic dithienylethene derivatives by linking the appropriate functional groups to the dithienylethene backbone.
15.2.6
Molecular Redox Switches Based on Metal Ions
Redox couples such as Cu2 þ /Cu þ , Fe3 þ /Fe2 þ , Ni3 þ /Ni2 þ (Ni þ ), and Co2 þ /Co þ have been also employed to build molecular switches. Redox-responsive
15.2 REDOX-CONTROLLED MOLECULAR SWITCHES
461
fluorescence switches can be constructed with redox-active complexes 37–42. The copper thiacyclam complex is linked to an anthracene fluorophore in 37. When copper ion is in the Cu(I) state, which is a d10 metal ion and is inactive in interfering with the anthracene excited state, complex 37 exhibits relatively strong fluorescence. When Cu(I) is transformed into Cu(II), which has a d9 electronic configuration, the fluorescence of the units is quenched through electron transfer from the excited anthracene to Cu(II). Since the Cu(I) and Cu(II) states can be interconverted by redox reactions, a redox fluorescence switch can be constructed with complex 37.45
Similarly, the luminescence of complexes 38, 39, 40, 41, and 42 can be modulated by changing the redox states of the respective metal ions. Complexes 38 and 39 show emission in the Ni(II) state, whereas the emission is quenched in the Ni(III) state generated after oxidation.46 The fluorescence due to the naphthalene unit in complex 40 is observed in the Ni(II) state; after reduction to the corresponding Ni(I) state the naphthalene fluorescence is distinctly reduced.47 Water-soluble complexes 41 and 42 also works as redox-responsive fluorescence switches in a similar way.48 Fe(II)/Fe(III) is another well-known metal ion redox couple. By making use of the different coordination preferences of Fe(II) and Fe(III), molecular redox switches
462
ELECTROACTIVE MOLECULES AND SUPRAMOLECULES
based on chemical triggering of iron ion translocation in triple-stranded helical complexes have been reported.49 The triple-stranded helical metal complexes can accommodate a single metal ion in either their internal “hard” binding cavity or their external “soft” cavity. Reduction of Fe(III) to Fe(II) induces the metal ion translocation from the internal hydroxamate binding sites to the external bipyridyl sites with a simultaneous change of the color from light-brown (lmax ¼ 420 nm, characteristic of the Fe(III)-hydroxamate complexes) to deep purple (lmax ¼ 540 nm, characteristic of the Fe(II)-bipyridyl complexes). As mentioned above, chiroptical molecular switches have attracted substantial interest because of their stereochemical novelty and potential applications. Several examples of chiroptical switches based on the redox-active metal complexes have been described. These redox-active complexes exhibit the exciton-coupled circular dichroism (ECCD) signal, which is critically dependent upon the distance and dihedral angle between the chromophoric ligands. Variation of the metal ion state through redox reactions would alter the relative orientation of the ligands and accordingly the magnitude of the ECCD spectrum is expected to vary with the oxidation state of metal ion. The ligands in complexes 43 and 44 are arranged around Cu(I) and Cu (II) in a different way as schematically shown in Scheme 15.6. As a result, 43 and 44 show different CD spectra. Moreover, CD spectra can be reversibly modulated by sequential oxidation and reduction with ammonium persulfate and ascorbic acid, respectively.50 The chiroptical switching behavior has also been reported for the copper and cobalt complexes with (S)-N,N-bis(2-pyridylmethyl)methionine/(S)-N,N-bis(2quinolylmethyl)methionine as the ligands.51Scheme 15.7 shows the configurations of complexes 45 and 46 with Cu(I) and Cu(II), respectively. Complex 45 with Cu(I) exhibits a different configuration compared to complex 46 with Cu(II) because of the preference of Cu(I) for coordination with sulfur atom. Such configuration change leads to an ECCD couplet with the opposite signal. Similar CD spectrum variation is detected for the cobalt complex 47. After reduction with ascorbic acid the ECCD signal inverts to positive and it can be returned back to negative after oxidation by persulfate.
Scheme 15.6 complex 44.
Schematic representation of the configurations of Cu(I) complex 43 and Cu(II)
15.2 REDOX-CONTROLLED MOLECULAR SWITCHES
Scheme 15.7
463
Schematic representation of the Cu(II) complex 45 and Cu(I) complex 46.
Supramolecular chirality can also be tuned by redox reactions. Aggregation of the chiral regioregular polythiophene 48 in the presence of various poor solvents or metal salts in chloroform leads to induced circular dichroism (ICD) in the p–p* transition region. Oxidative doping of the main chain of polythiophene 48 with Cu(II) trifluoromethanesulfonate causes the disappearance of the ICD signal. Successive addition of amines such as triethylamine induces the undoping of the main chain of polythiophene 48 that results in the reappearance of the ICD signal. Therefore, the supramolecular chirality of the aggregate of polythiophene 48 can be reversibly controlled by doping and undoping the main chain of polythiophene 48 through redox reactions. It seems that the chirality switching through such a doping and undoping process of the main chain may be applicable to other chiral polythiophenes.52
15.2.7
Molecular Switches with Miscellaneous Electroactive Systems
Overcrowded alkenes have captured the interests of chemists for many years. Unidirectional light-driven molecular motors reported by Feringa and coworkers are based on overcrowded alkenes.53 Recently, this group has described a three-state
464
ELECTROACTIVE MOLECULES AND SUPRAMOLECULES
Scheme 15.8 The photo- and redox-controlled fluorescence switch based on bis-thiaxanthylidenes 49 and 50.
luminescence switch with compounds 49A and 50A. Compound 49A exhibits a fluorescence band approximately 440 nm. Electrochemical oxidation of 49A leads to the corresponding dication (492 þ ) that emits around 585 nm. Electrochemical reduction of 492 þ and further slight heating lead to the formation of 49B that is nonfluorescent. 49A can be also generated by light irradiation of 49B. Thus, on–off fluorescence switch can be achieved with 49A electrochemically or photochemically. In this manner, the fluorescence of 49A can be modulated by redox reaction, light irradiation, and heating. Similar fluorescence variation can be achieved with 50A (Scheme 15.8).54 Pyrene can be reduced to the radical anion and by making use of this feature a chiroptical molecular redox switch can be constructed. The chiral trans-cyclohexanediol bispyrene esters (1R,2R)-51a and (1S,2S)-51b can be reversibly reduced to the corresponding biradical anions 51a2 and 51b2 that show strong absorption bands at 510 nm. Compounds 51a2 and 51b2 exhibit strong CD signals around 500 nm. After oxidation of 51a2 and 51b2 to 51a and 51b, their CD spectra can be restored. In this manner, the CD spectra of 51a and 51b can be switched electrochemically. However, compounds 52a and 52b with amide as the linkers do not show such CD spectral variation by the same redox reactions.55
15.2 REDOX-CONTROLLED MOLECULAR SWITCHES
465
Naphthalimide can be reduced to the corresponding radical anion. Compared to naphthalimide, its radical anion exhibits high hydrogen binding affinity with hydrogen donors. By using these properties of naphthalimide, a molecular shuttle based on [2]rotaxane 53 containing succinamide and naphthalimide hydrogen-bonding stations for a benzylic amide macrocycle has been described recently by Paolucci, Leigh, and coworkers.561 H NMR spectroscopic data show that the benzylic amide macrocycle in [2]rotaxane 53 exhibits a positional selectivity for the succinamide station in a variety of solvents. Electrochemical reduction leads to the switching of the benzylic amide macrocycle to the naphthalimide position. The macrocycle will shift back to the succinamide position after subsequent oxidation. This shuttling process, that is reversible and cyclable, can be both effected and observed in cyclic voltammetry (Scheme 15.9). These results imply that the use of electrochemical stimuli to manipulate intercomponent interactions provides a powerful method for both controlling and observing submolecular motion in hydrogen-bonded molecular shuttles. Such molecular shuttle driven by redox reactions augurs well for the development of molecular devices.
Scheme 15.9 Reversible shuttling of the benzylic amide macrocycle in [2]rotaxane 53 by redox reactions.
466
ELECTROACTIVE MOLECULES AND SUPRAMOLECULES
Bis(catecholketal)s undergo a ready transformation to the folded (p-stacked) conformation upon oxidation. Optically active bis(catecholketal)s 54 and 55 have been studied for creation of a new class of chiroptical molecular switches based on the redox-induced conformational changes.57 Both experimental and theoretical studies indicate that CD spectral changes occur for 54 and 55 after chemical oxidation. The original CD spectra of 54 and 55 can be recovered by successive chemical reduction. It should be mentioned that the redox-induced chiroptical switching observed for 54 and 55 is based on the spontaneous conformational changes upon oxidation. Daub, Grimme, and coworkers have reported chiroptical switches based on binaphthyl boron dipyrromethane (BDP) conjugates. It is known that BDP dyes with appropriate functionalization can be reversibly oxidized and reduced.58 The CDspectroelectrochemical studies of R-56 show a decrease in the intensity of the Cotton effect at 501 nm by applying a reduction potential to the solution. The initial CD spectrum is restored completely after reoxidation; thus, the CD signal intensity at 501 nm for R-56 can be switched on and off electrochemically.59
Dihydro[5]helicenes 57, which contain two electron-donating spiro rings, can undergo reversible CC bond breaking to generate dications 572 þ . These transformations are accompanied by remarkable absorption spectral changes. Furthermore, the redox potentials of 57 and 572 þ are largely separated, thus endowing this system with high electrochemical bistability. The transformation of optically active 57 to the corresponding dications results in drastic CD spectral changes. Therefore, a redox
15.3
MOLECULAR LOGIC GATES BASED ON ELECTROACTIVE SYSTEMS
467
chiroptical switch can be constructed with dihydro[5]helicene in which helicity and the axial chirality of biaryls can be reversibly interconverted into a CC bond forming/breaking process controlled by redox reactions.60 Long p-conjugated molecules can function as molecular wires.61 Conjugated molecules containing redox active units such as 5862 and 5963 have been reported. It is anticipated that the conductivities of such molecular wires can be altered by switching of the redox active units. Compound 58 bearing an anthraquinone core can be reversibly switched electrochemically between a cross-conjugated and a linear-conjugated forms via two-electron reduction/oxidation processes. Theoretical calculations imply that such conjugation variation can induce conductivity change from low to high when the anthraquinone unit is switched from a cross- to a linear-conjugated state. Similarly, the linear-conjugated pathway in 59, an oligo(phenyleneethynylene)-tetrathiafulvalene, can be switched electrochemically to a cross-conjugated one via a two-electron oxidation of the flanking TTF moieties.
15.3 MOLECULAR LOGIC GATES BASED ON ELECTROACTIVE SYSTEMS Like molecular switches, molecular logic gates are also important components for future information processing. Molecular logic gates can transmit one or more input signals into one or more output signals. Integration of elementary logic gates leads to combinational logic circuits that can perform more complicated logic functions. In the following, we will describe representative logic gates and combinational logic circuit based on electroactive systems. 15.3.1 Logic Gates and Combinational Logic Circuit with TTF Derivatives “XNOR” Logic Gate Balzani, Stoddart, Venturi, and coworkers have reported an XNOR logic gate based on an electron donor–acceptor pseudorotaxane composed of cyclophane cyclobis (paraquat-p-phenylene) (CBPQT4 þ ) and a TTF derivative 60. Oxidation of 60 leads to the dethreading of the pseudorotaxane and further reduction induces the rethreading. Similarly, dethreading and rethreading processes can occur by sequential reduction and oxidation of CBPQT4 þ . If the oxidation of 60 (I1) and reduction of CBPQT4 þ (I2) are regarded as the input and output signals, respectively, the variation of the charge transfer absorption band of this pseudorotaxane at 830 nm (O1) mimics the function of an XNOR logic gate. This is explained as follows: (1) when no oxidation or reduction performed, namely, I1 ¼ 0 and I2 ¼ 0, the pseudorotaxane is stable showing CT band at 830 nm (O1 ¼ 1), (2) after either oxidation or reduction is carried out (I1 ¼ 1 and I2 ¼ 0 or I1 ¼ 0, and I2 ¼ 1), dethreading process occurs and thus the CT band in no longer present (O1 ¼ 0), and (3) if both oxidant and reductant are added
468
ELECTROACTIVE MOLECULES AND SUPRAMOLECULES
simultaneously, the pseudorotaxane is also stable and CT band can be detected again (O1 ¼ 1).64
Monomolecular “Half-Adder”Rational integration of relatively simple logic gates into high-level circuits is highly important not only for the potential application in molecular electronics but also for the understanding of complex mechanisms of important biological processes. To the best of our knowledge, there are only a few reports dealing with “half-adder” that, as the basis of electronic calculators and computer machines, requires two binary inputs and two binary outputs to perform simple arithmetic addition.The way that absorption spectra changes in association . with the interconversion between TTF, TTF þ , and TTF2 þ , mimics the function of a . 65 half-adder. TTF can be transformed into TTF þ or TTF2 þ , respectively, by either application of the correct oxidation potential or reaction with a suitable stoichiometric amounts of chemical oxidants such as NOPF6. Electrochemical oxidation at þ 0.65 V (versus Ag/AgCl) and addition of 1.2 equivalent of NOPF6 can be regarded as two input signals, I1 and I2, respectively. The absorptions at 350 nm (due to TTF2 þ ) and . 435 nm (due to TTF þ ) can be considered as two outputs, O1 and O2, respectively (Fig. 15.3). The absorption changes at 350 and 435 nm upon external inputs, I1 and I2,
Figure 15.3 Illustration of the construction of the AND (O1) and XOR (O2) gates by . detecting the absorption at 350 nm (O1) and 435 nm (O2) of TTF2 þ and TTF þ , respectively.
15.4
INFORMATION STORAGE WITH ELECTRON D–A MOLECULES
469
can be interpreted as “AND” and “XOR” logic gates, respectively. The “AND” and “XOR” gates can be “operated” in parallel since they are based on the entity of a single molecule of TTF under the same inputs. Therefore, a monomolecular half-adder is . realized. Because both TTF þ and TTF2 þ can be reduced to the neutral state of TTF, this new monomolecular half-adder shows resettable capability.65 It should be noted that rational design and development of new molecular half-adder with resettable character based on rather simple and easily accessible molecules still remain to be challenging.
15.4 INFORMATION STORAGE WITH ELECTRON D–A MOLECULES AND SUPRAMOLECULES According to the different external stimuli in data recording, data storage usually includes three types: magnetic data storage, optical data storage, and electrical data storage. Electrical data storage is a powerful approach to achieve high-density data recording. In electrical data storage, the data are stored based on the high- and lowconductivity response to an applied voltage (electrical bistability). SPM-based techniques represented by scanning tunneling microscopy (STM) and atomic force microscopy (AFM) have been demonstrated powerfully for conducting data storage at nanometer scale or molecular scale.66 Various electroactive molecules and supramolecules have been developed for information storage. Herein, we will just show a few examples of electron D–A molecules and supramolecules for information storage. Some conjugated electron donor–acceptor compounds have been found to be very promising materials for ultrahigh-density information storage. This is based on the finding that the electrical conductivity of the thin films of these D–A compounds can be reversibly switched by applying pulsed voltages with STM technique. Compound 6167 is one of these D–A molecules containing triphenylamine (as electron donor unit) and cyanovinyl (as electron acceptor unit) groups. When a voltage pulse (>2.3 V) is applied onto the thin film of 61 deposited on highly ordered pyrolytic graphite (HOPG) through STM tip, a typical recording pattern can be formed and the average size of the marks is approximately 2.5 nm in diameter. Interestingly, when a reversepolarity voltage pulse of more than 1.4 V is applied to the recorded region, the recorded marks can be substantially erased. Furthermore, a new data mark can be rewritten on the erased region of the thin film of 61 by applying another positive pulsed voltage. Further studies show that the conductivity of unrecorded regions on the film is in a high resistance state, while it is in a low-resistance state at the recorded regions. Thus, it can be concluded that the application of pulsed voltage induces the conductivity transition for the thin film of 61. A new absorption band centered at 448 nm appears in the absorption spectrum of the thin film of 61 after applying a pulsed voltage. This is assumed to be due to the intense intermolecular charge transfer interactions, which would lead to an increased number of the carriers and a delocalized state, resulting in a higher conductivity.
470
ELECTROACTIVE MOLECULES AND SUPRAMOLECULES
The [2]rotaxanes, which contain tetrathiafulvalene (TTF) and 1,5-dioxynaphthalene (DNP) as the two recognition stations and cyclobis(paraquat-p-phenylene) (CBPQT4 þ ) as the cyclic moiety, have been comprehensively investigated by Stoddart and coworkers.68 We have recently reported two new TTF-DNPCBPQT4 þ [2]rotaxanes 62 and 63. In these two [2]rotaxanes, the TTF moiety is the 4,40 (50 )-dialkylthiotetrathiafulvalene that is rather easily accessible based on the synthetic procedure developed by us previously,69 and two different spacers are used: the cyclohexyl and alkyl chains. Two stopper units are the G2-dendritic moieties.
15.4
INFORMATION STORAGE WITH ELECTRON D–A MOLECULES
471
[2]Rotaxanes 62 and 63 can form monolayers at the air–water interface and they can be transferred onto the substrates by using normal LB technique. In this way, multilayer LB films of [2]rotaxanes 62 and 63 can be fabricated. The conducting behavior of the LB films of 62 deposited onto the surface of HOPG has been studied with conducting AFM. When the applying voltage is lower than 1.4 V, the conductivity of the LB films is rather small. Interestingly, a sudden conductivity enhancement is observed when the voltage is higher than 1.4 V. The conducting behavior of the LB films is also studied with STM, and it is found that the conductivity increased largely after applying an electric field pulse (ca. 2.0 Vand 5 ms) to the LB films. These results indicate that LB films show electrical bistability behavior, and the low-conductivity state can be transformed into the high-conductivity state by applying an electrical voltage. The conducting behavior of the LB films of [2]rotaxanes 62 has been also studied with two-terminal junction device. The results also indicate that for the LB films of [2]rotaxane 62 conductance transition can be reversibly controlled by the external voltage. The conducting behavior of the LB films of [2]rotaxane 63 has been also studied with conducting AFM, STM, and two-terminal junction device. Similar electrical bistability behavior has been observed. Plenty of experimental results have provided support for the following mechanism explaining the switchable electrical bistability behavior observed for TTF-DNP-CBPQT4 þ [2]rotaxanes70,71: (1) By applying a . positive pulse voltage, the TTF unit would be oxidized into the TTF þ , and as a result 4þ would move to the DNP station owing to the electronic repulsion. CBPQT Theoretical calculation results indicated that the complex DNP-CBPQT4 þ exhibited higher conductivity than the complex TTF-CBPQT4 þ .72 Therefore, it can be understandable that the conductivity of the LB films of TTF-DNP-CBPQT4 þ [2] rotaxanes would be enhanced when CBPQT4 þ is shifted to the DNP station and (2) by . applying a negative pulse voltage TTF þ would be reduced to the neutral TTF unit 4þ would move back to the TTF station, and accordingly the lowand CBPQT conductivity state would be restored. Therefore, it is reasonable to assume that the same mechanism can also account for the electrical bistability exhibited by the LB films of [2]rotaxanes 62 and 63. The electrical bistable behaviors of the thin films of 62 and 63 demonstrate the potential applications of 62 and 63 for ultrahigh-density information storage. By applying voltage onto the thin films of 62 through STM tip, repeatable and rewritable nanorecording can be realized. Nanoscale dots can be written repeatedly with voltage pulses (ca. 2.0 V and 0.1–10 ms). It is interesting to note that the marks written on the thin films of 62 can be erased, rerecorded, and reerased on the same area.73 In the whole recording process, the dots remain a size of ca. 3.0 nm and are stable in air at room temperature for more than 12 h. The thin films of 63 can also be used for STM-based nanorecording.74 Similar recording dots can be written on the thin films of 63 by applying a positive voltage pulse of approximately 2.0 Vand 5.0 ms from the STM tip. But, erasing the recording dots cannot be achieved by applying a negative pulse on the thin film of 63. It is suggested that the intermolecular interactions for 63 are strong due to the soft alkyl spacer. This will make it difficult for CBPQT4 þ move back to TTF site, and as a result
472
ELECTROACTIVE MOLECULES AND SUPRAMOLECULES
the recording nanodots with the thin films of 63 are difficult to be erased. In comparison, the spacer in 62 is cyclohexyl group. Probably because of the steric hindrance the intermolecular interactions for 62 become weak. Accordingly, CBPQT4 þ can be switched back to the TTF site and thus the recording nanodots become erasable.75 Charge transfer complexes have also been investigated as the most attractive candidate materials for high-density electrical data storage.76,77 For instance, silver-tetracyanoquinodimethane (Ag-TCNQ) and copper-tetracyanoquinodimethane (Cu-TCNQ) have been studied for data recording since they exhibit electrical bistability. Additionally, attaching the electroactive molecules to an electroactive surface such as gold or silicon surface provides an attractive approach for molecular-based electrical recording.78 The electroactive molecules undergo reversible oxidation and reduction reactions after applying potentials, and thus information can be stored in the distinct oxidation states of the molecules. Also, the redox reactions induce the change of the conductivity, and as a result the information can be read based on the conductivity difference. Such writing, reading, and erasing information are accomplished electrically, and multiple bits of information can be stored by the molecules or molecular arrays that afford a set of distinct oxidation states.
15.5
SUMMARY AND OUTLOOK
In this chapter, we have discussed recent examples of electroactive molecules and supramolecules toward information processing and high-density information storage at molecular level. The main part of chapter is devoted to redox-controlled molecular switches based on various redox centers. A few examples of molecular logic gates and combinational logic circuit are also presented. It is demonstrated that electroactive molecules and supramolecules show potential applications in high-density information storage. For molecular switches and logic gates, these results still remain the rudimentary examples at the present stage. In the future, practical applications might emerge only after considerable fundamental studies on various aspects of this research area. One of the major challenges is in great need of more prototypes and strategies of integrating isolated molecular switches and logic gates to chemical networks, ensuring the fast information processing and transportation. Also, it is of great challenge to develop new methods to incorporate these electroactive molecules onto solid-state devices while maintaining signal transduction abilities. Great progress in development of electroactive systems as electrical recording media has been achieved, and nanometer even molecular scale data storage has been realized. The growing demand for high-density information storage is encouraging the exploration of novel recording technologies and innovative materials for future large-capacity memories. Compared to electrical-mode data storage, multimode data storage has received increasing attention in recent years. Multifunctional molecules,
REFERENCES
473
which can undergo different transformation depending on the type of external stimuli, are expected to realize higher density, multimode, and complex information processing. Unveiling the information-recording mechanism is another important issue. Besides, new techniques for the assembly of these functional molecules and fabrication of information recording the device are also crucial for realization of ultrahighdensity information storage.
REFERENCES 1. R. Ballardini, P. Ceroni, A. Credi, M. T. Gandolfi, M. Maestri, M. Semeraro, M. Venturi, V. Balzani, Adv. Funct. Mater. 2007, 17, 740–750, further references therein. 2. D. Margulies, G. Melman, A. Shanzer, Nat. Mater. 2005, 4, 768–771. 3. D. Margulies, G. Melman, A. Shanzer, J. Am. Chem. Soc. 2006, 128, 4865–4871. 4. A. P. de Silva, M. R. James, B. O. F. Mckinney, D. A. Pears, S. M. Weir, Nat. Mater. 2006, 5, 787–771. 5. D. C. Magri, G. J. Brown, G. D. McClean, A. P. de Silva, J. Am. Chem. Soc. 2006, 128, 4950–4951. 6. P. Hudhomme, C. R. Chim. 2006, 9, 881–891. 7. N. Martin, L. Sanchez, M. A. Herranz, B. B. Illescas, D. M. Guldi, Acc. Chem. Res. 2007, 40, 1015–1024. 8. H. Spanggaard, J. Prehn, M. B. Nielsen, E. Levillain, M. Allain, J. Becher, J. Am. Chem. Soc. 2000, 122, 9486–9494. 9. G. Zhang, D. Zhang, X. Guo, D. Zhu, Org. Lett. 2004, 6, 1209–1212. 10. X. Xiao, W. Xu, D. Zhang, H. Xu, L. Liu, D. Zhu, New. J. Chem. 2005, 29, 1291–1294. 11. H. C. Li, J. O. Jeppesen, E. Levillain, J. Becher, Chem. Commun. 2003, 846–847. 12. X. Xiao, W. Xu, D. Zhang, H. Xu, H. Lu, D. Zhu, J. Mater. Chem. 2005, 15, 2557–2561. 13. J. Becher, T. Brimert, J. O. Jeppesen, J. Z. Pedersen, R. Zubarev, T. Bjørnholm, N. Reitzel, T. R. Jensen, K. Kjaer, E. Levillain, Angew. Chem., Int. Ed. 2001, 40, 2497–2500. 14. S. Leroy-Lhez, J. Baffreau, L. Perrin, E. Levillain, M. Allain, M. J. Blesa, P. Hudhomme, J. Org. Chem. 2005, 70, 6313–6320. 15. S. Leroy-Lhez, L. Perrin, J. Baffreau, P. Hudhomme, C. R. Chim. 2006, 9, 240–246. 16. C. J. Fang, Z. Zhu, W. Sun, C. H. Xu, C. H. Yan, New. J. Chem. 2007, 31, 580–586. 17. (a) E. Lippert, W. L€uder, F. Moll, H. Nagele, H. Boos, H. Prigge, I. Siebold-Blankenstein, Angew. Chem., Int. Ed. 1961, 73, 695–706. (b) W. Rettig, Angew. Chem., Int. Ed. Engl. 1986, 25, 971–988, and references therein. 18. W. Tan, D. Q. Zhang, H. Wu, D. B. Zhu, Tetrahedron Lett. 2008, 49, 1361–1364. 19. P. R. Ashton, V. Balzani, J. Becher, A. Credi, M. C. T. Fyfe, G. Mattersteig, S. Menzer, M. B. Nielsen, F. M. Raymo, J. F. Stoddart, M. Venturi, D. J. Williams, J. Am. Chem. Soc. 1999, 121, 3951–3957. 20. Y. Zhou, D. Zhang, L. Zhu, Z. Shuai, D. Zhu, J. Org. Chem. 2006, 71, 2123–2130. 21. H. Wu, D. Zhang, D. Zhu, Tetrahedron Lett. 2007, 48, 8951–8955. 22. E. Gomar-Nadal, J. Veciana, C. Rovira, D. B. Amabilino, Adv. Mater. 2005, 17, 2095–2098.
474
ELECTROACTIVE MOLECULES AND SUPRAMOLECULES
23. G. Wen, D. Zhang, Y. Huang, R. Zhao, L. Zhu, Z. Shuai, D. Zhu, J. Org. Chem. 2007, 72, 6247–6250. 24. K. Uchida, G. Masuda, Y. Aoi, K. Nakayama, M. Irie, Chem. Lett. 1999, 1071–1072. 25. R. Zhang, Z. Wang, Y. Wu, H. Fu, J. Yao, Org. Lett. 2008, 10, 3065–3068. 26. J. Rochford, A. D. Rooney, M. T. Pryce, Inorg. Chem. 2007, 46, 7247–7249. 27. R. Martınez, I. Ratera, A. Tarraga, P. Molina, J. Veciana, Chem. Commun. 2006, 3809– 3811. 28. N. Hida, F. Takei, K. Onitsuka, K. Shiga, S. Asaoka, T. Iyoda, S. Takahashi, Angew. Chem., Int. Ed. 2003, 42, 4349–4352. 29. V. Goulle, A. Harriman, J.-M. Lehn, J. Chem. Soc., Chem. Commun. 1993, 1034–1036. 30. A. C. Benniston, G. M. Chapman, A. Harriman, S. A. Rostron, Inorg. Chem. 2005, 44, 4029–4036. 31. R. A. Illos, E. Harlev, S. Bittner, Tetrahedron Lett. 2005, 46, 8427–8430. 32. S. Arounaguiri, B. G. Maiya, Inorg. Chem. 1999, 38, 842–843. 33. P. Yan, M. W. Holman, P. Robustelli, A. Chowdhury, F. I. Ishak, D. M. Adams, J. Phys. Chem. B 2005, 109, 130–137. 34. P. M. S. Monk, The Viologens, John Wiley & Sons Ltd, Chichester, 1998. 35. R. Lomoth, T. H€aupl, O. Johansson, L. Hammarstr€ om, Chem. Eur. J. 2002, 8, 102–110. 36. (a) D. B. Amabilino, J. F. Stoddart, Chem. Rev. 1995, 95, 2725–2828. (b) D. Philp, J. F. Stoddart, Angew. Chem., Int. Ed. Engl. 1996, 35, 1154–1196. (c) V. Balzani, A. Credi, F. M. Raymo, J. F. Stoddart, Angew. Chem., Int. Ed. Engl. 2000, 39, 3348–3391. (d) A. C. Nijhuis, B. Jan Ravoo, J. Huskens, D. N. Reinhoudt, Coord. Chem. Rev. 2007, 251, 1761–1780. 37. E. Katz, O. Lioubashevsky, I. Willner, J. Am. Chem. Soc. 2004, 126, 15520–15532. 38. J. Deng, N. Song, Q. Zhou, Z. Su, Org. Lett. 2007, 9, 5393–5396. 39. J. Deng, N. Song, W. Liu, Q. Zhou, Z. Wang, ChemPhysChem 2008, 9, 1265–1269. 40. A. Peters, N. R. Branda, J. Am. Chem. Soc. 2003, 125, 3404–3405. 41. L. Sun, H. Tian, Tetrahedron Lett. 2006, 47, 9227–9231. 42. A. Peters, N. R. Branda, Chem. Commun. 2003, 954–955. 43. B. Gorodetsky, H. D. Samachetty, R. L. Donkers, M. S. Workentin, N. R. Branda, Angew. Chem., Int. Ed. 2004, 43, 2812–2815. 44. B. Gorodetsky, N. R. Branda, Adv. Funct. Mater. 2007, 17, 786–796. 45. G. De Santis, L. Fabbrizzi, M. Licchelli, C. Mangano, D. Sacchi, Inorg. Chem. 1995, 34, 3581–3852. 46. L. Fabbrizzi, M. Licchelli, G. De Santis, N. Sardone, A. H. Velders, Chem. Eur. J. 1996, 2, 1243–1250. 47. M. D. Casa, L. Fabbrizzi, M. Licchelli, A. Poggi, D. Sacchi, M. Zema, J. Chem. Soc. Dalton Trans. 2001, 1671–1675. 48. L. Fabbrizzi, M. Licchelli, S. Mascheroni, A. Poggi, D. Sacchi, M. Zema, Inorg. Chem. 2002, 41, 6129–6136. 49. L. Zelikovlch, J. Libman, A. Shanzer, Nature 1995, 374, 790–792. 50. S. Zahn, J. W. Canary, Angew. Chem., Int. Ed. 1998, 37, 305–307.
REFERENCES
475
51. D. Das, Z. Dai, A. Holmes, J. W. Canary, Chirality 2008, 20, 585–591. 52. H. Goto, E. Yashima, J. Am. Chem. Soc. 2002, 124, 7943–7949. 53. B. L. Feringa,(Ed.), Molecular Switches, Wiley-VCH Verlag Gmbh, 2001, further references therein. 54. W. R. Browne, M. M. Pollard, B. de Lange, A. Meetsma, B. L. Feringa, J. Am. Chem. Soc. 2006, 128, 12412–12413. 55. C. Westermeier, H. Gallmeier, M. Komma, J. Daub, Chem. Commun. 1999, 2427–2428. 56. A. Altieri, F. G. Gatti, E. R. Kay, D. A. Leigh, D. Martel, F. Paolucci, A. M. Z. Slawin, J. K. Y. Wong, J. Am. Chem. Soc. 2003, 125, 8644–8654. 57. M. Fukui, T. Mori, Y. Inoue, R. Rathore, Org. Lett. 2007, 9, 3977–3980. 58. (a) M. Kollmannsberger, T. Gareis, S. Heinl, J. Breu, J. Daub, Angew. Chem., Int. Ed. 1997, 109, 1391–1393. (b) A. Burghurt, H. Kim, M. B. Welch, L. H. Thoresen, J. Reibenspies, K. Burgess, F. Bergstr€om, L. Johansson, J. Org. Chem. 1999, 64, 7813–7819. 59. G. Beer, C. Niederalt, S. Grimme, J. Daub, Angew. Chem., Int. Ed. 2000, 39, 3252–3255. 60. J. Nishida, T. Suzuki, M. Ohkita, T. Tsuji, Angew. Chem., Int. Ed. 2001, 113, 3351–3354. 61. (a) J. Chen, M. A. Reed, A. M. Rawlett, J. M. Tour, Science 1999, 286, 1550–1552. (b) S. Kubatkin, A. Danilov, M. Hjort, J. Cornil, J.-L. Bredas, N. Stuhr-Hansen, P. Hedegard, T. Bjørnholm, Nature 2003, 425, 698–701. (c) A. R. Pease, J. O. Jeppesen, J. F. Stoddart, Y. Luo, C. P. Collier, J. R. Heath, Acc. Chem. Res. 2001, 34, 433–444. (d) J. M. Tour, A. M. Rawlett, M. Kozaki, Y. Yao, R. C. Jagessar, S. M. Dirk, D. W. Price, M. A. Reed, C.-W. Zhou, J. Chen, W. Wang, I. Campbell, Chem. Eur. J 2001, 7, 5118–5134. (e) R. McCreery, J. Dieringer, A. O. Solak, B. Snyder, A. M. Nowak, W. R. McGovern, S. DuVall, J. Am. Chem. Soc. 2003, 125, 10748–10758. (f) J. A. M. Dinglasan, M. Bailey, J. B. Park, A. -A. Dhirani, J. Am. Chem. Soc. 2004, 126, 6491–6497. 62. E. H. van Dijk, D. J. T. Myles, M. H. van der Veen, J. C. Hummelen, Org. Lett. 2006, 8, 2333–2336. 63. J. K. Sorensen, M. Vestergaard, A. Kadziola, K. Kilsa, M. B. Nielsen, Org. Lett. 2006, 8, 1173–1176. 64. M. Asakawa, P. R. Ashton, V. Balzani, A. Credi, G. Mattersteig, O. A. Matthews, M. Montalti, N. Spencer, J. F. Stoddart, M. Venturi, Chem. Eur. J. 1997, 3, 1992–1996. 65. Y. Zhou, H. Wu, L. Qu, D. Zhang, D. Zhu, J. Phys. Chem. B 2006, 110, 15676–15679. 66. (a) D. M. Eigler, E. K. Schweizer, Nature 1990, 344, 524–526. (b) A. Sato, Y. Tsukamoto, Nature 1993, 363, 431–432. (c) H. J. Gao, K. Sohlberg, Z. Q. Xue, H. Y. Chen, S. M. Hou, L. P. Ma, X. W. Fang, S. J. Pang, S. J. Pennycook, Phys. Rev. Lett. 2000, 84, 1780–1783. (d) D. X. Shi, Y. L. Song, D. B. Zhu, H. X. Zhao, S. S. Xie, S. J. Pang, H. J. Gao, Adv. Mater. 2001, 13, 1103–1105. (e) D. X. Shi, Y. L. Song, H. X. Zhang, P. Jiang, S. T. He, S. S. Xie, S. J. Pang, H. J. Gao, Appl. Phys. Lett. 2000, 77, 3203. (f) K. Yano, T. Ikeda, Appl. Phys. Lett. 2002, 80, 1067–1069. 67. Y. L. Shang, Y. Q. Wen, S. L. Li, S. X. Du, X. B. He, L. Cai, Y. F. Li, L. M. Yang, H. J. Gao, Y. L. Song, J. Am. Chem. Soc. 2007, 129, 11674–11675. 68. C. P. Collier, G. Mattersteig, E. W. Wong, Y. Luo, K. Beverly, J. Sampaio, F. M. Raymo, J. F. Stoddart, J. R. Heath Science 2000, 289, 1172–1175. 69. X. Guo, Y. Zhou, M. Feng, Y. Xu, D. Zhang, H. Gao, Q. Fan, D. Zhu, Adv. Funct. Mater. 2007, 17, 763–769.
476
ELECTROACTIVE MOLECULES AND SUPRAMOLECULES
70. (a) J. O. Jeppesen, K. A. Nielsen, J. Perkins, S. A. Vignon, A. Di Fabio, R. Ballardini, M. T. Gandolfi, M. Venturi, V. Balzani, J. Becher, J. F. Stoddart, Chem. Eur. J. 2003, 9, 2982–3007. (b) T. Yamamoto, H.-R. Tseng, J. F. Stoddart, V. Balzani, A. Credi, F. Marchioni, M. Venturi, Collect. Czech. Chem. Commun. 2003, 68, 1488–1514. (c) H.R. Tseng, S. A. Vignon, P. C. Celestre, J. Perkins, J. O. Jeppesen, A. Di Fabio, R. Ballardini, M. T. Gandolfi, M. Venturi, V. Balzani, J. F. Stoddart, Chem. Eur. J. 2004, 10, 155–172. (d) J. O. Jeppesen, J. Perkins, J. Becher, J. F. Stoddart, Angew. Chem., Int. Ed. 2001, 40, 1216–1221. (e) J. O. Jeppesen, S. A. Vignon, J. F. Stoddart, Chem. Eur. J. 2003, 9, 4611–4625. (f) H.-R. Tseng, D. Wu, N. X. Fang, X. Zhang, J. F. Stoddart, ChemPhysChem 2004, 5, 111–116. (g) J. O. Jeppesen, S. Nygaard, S. A. Vignon, J. F. Stoddart, Eur. J. Org. Chem. 2005, 196–220. (h) Y. Liu, A. H. Flood, P. A. Bonvallet, S. A. Vignon, B. H. Northrop, H.-R. Tseng, J. O. Jeppesen, T. J. Huang, B. Brough, M. Baller, S. Magonov, S. D. Solares, W. A. Goddard, C.-M. Ho, J. F. Stoddart, J. Am. Chem. Soc. 2005, 127, 9745–9759. (i) J. W. Choi, A. H. Flood, D. W. Steuerman, S. Nygaard, A. B. Braunschweig, N. N. P. Moonen, B. W. Laursen, Y. Luo, E. DeIonno, A. J. Peters, J. O. Jeppesen, K. Xu, J. F. Stoddart, J. R. Heath, Chem. Eur. J. 2006, 12, 261–279. 71. (a) P. M. Mendes, W. Lu, H.-R. Tseng, S. Shinder, T. Iijima, M. Miyaji, C. M. Knobler, J. F. Stoddart, J. Phys. Chem. B 2006, 110, 3845–3848. (b) K. Nørgaard, B. W. Laursen, S. Nygaard, K. Kjaer, H.-R. Tseng, A. H. Flood, J. F. Stoddart, T. Bjørnholm, Angew. Chem., Int. Ed. 2005, 44, 7035–7039. 72. (a) W.-Q. Deng, R. P. Muller, W. A. Goddard, III, J. Am. Chem. Soc. 2004, 126, 13562–13563. (b) Y. H. Jang, S. Hwang, Y. H. Kim, S. S. Gang, W. A. Goddard, III, J. Am. Chem. Soc. 2004, 126, 12636–12645. 73. M. Feng, L. Gao, Z. Deng, W. Ji, X. Guo, S. Du, D. Shi, D. Zhang, D. Zhu, H. Gao, J. Am. Chem. Soc. 2007, 129, 2204–2205. 74. M. Feng, X. Guo, X. Lin, X. He, W. Ji, S. Du, D. Zhang, D. Zhu, H. Gao, J. Am. Chem. Soc. 2005, 127, 15338–15339. 75. M. Feng, L. Gao, S. Du, Z. Deng, Z. Cheng, W. Ji, D. Zhang, X. Guo, X. Lin, L. Chi, D. Zhu, H. Fuchs, H. Gao, Adv. Funct. Mater. 2007, 17, 770–776. 76. (a) R. S. Potember, R. C. Hoffman, T. O. Poehler, Molecular electronics APL Tech. Dig. 1986, 7, 129. (b) J. Gong, Y. Osada, Appl. Phys. Lett. 1992, 61, 2787–2789. (c) Z. Q. Xue, M. Ouyang, K. Z. Wang, H. X. Zhang, C. H. Huang, Thin Solid Films 1996, 288, 296–299. (d) S. Yamaguchi, C. A. Vianda, R. S. Potember, J. Vac. Sci. Technol. B 1991, 9, 1129. (e) R. S. Potember, T. O. Poehler, D. O. Cowan, Appl. Phys. Lett. 1979, 34, 405–407. 77. (a) L. P. Ma, S. Pyo, J. Y. Ouyang, Q. F. Xu, Y. Yang, Appl. Phys. Lett. 2003, 82, 1419–1421. (b) L. D. Bozano, B. W. Kean, V. R. Deline, J. R. Salem, J. C. Scott, Appl. Phys. Lett. 2004, 84, 607–609. (c) J. Y. Ouyang, C. W. Chu, C. R. Szmanda, L. P. Ma, Y. Yang, Nat. Mater. 2004, 3, 918–922. (d) J. C. Scott, L. D. Bozano, Adv. Mater. 2007, 19, 1452–1463. 78. K. M. Roth, N. Dontha, R. B. Dabke, D. T. Gryko, C. Clausen, J. S. Lindsey, D. F. Bocian, W. G. Kuhr, J. Vac. Sci. Technol. B 2000, 18, 2359–2365.
CHAPTER 16
Electrochemiluminescent Systems as Devices and Sensors ANDRZEJ KAPTURKIEWICZ Institute of Physical Chemistry, Polish Academy of Sciences, Warsaw, Poland Institute of Chemistry, University of Podlasie, Siedlce, Poland
16.1 INTRODUCTION: ECL CONCEPT AND SHORT HISTORICAL OVERVIEW Electrochemically generated chemiluminescence (also called as electrochemiluminescence and abbreviated as ECL) is commonly defined as the emission of light resulting from the generation of electronically excited states formed in homogeneous electron transfer reaction between active precursors obtained by means of heterogeneous electron transfer (electrode) processes. After the first observation, ECL has come a long way in the past 45 years in terms of (i) finding new ECL luminophores, (ii) understanding the mechanisms associated with light generation, and (iii) developing practical applications. As in many other cases, development of ECL technique has passed some, below briefly described, milestones. The first modern observations of ECL phenomena were already reported in the mid-1960s in papers describing the electrochemical generation of light emission by electrolysis of polyaromatic hydrocarbons R in acetonitrile or N,N-dimethylformamide solutions.1–3 The light emission was produced by means of the radical ion annihilation reaction between an oxidized R þ and a reduced R species, both of which were generated at an electrode by alternate pulsing of the electrode potential. The potential of the working electrode was quickly changed between two different values, enough positive/negative to generate the oxidized/reduced species reacting further near the electrode surface with formation of the emissive state R. The applied triple-potential-step technique has been followed soon by theoretical description of the method.4,5 Within the next few years, the list of ECL emitters was extended to Electrochemistry of Functional Supramolecular Systems. Edited by Paola Ceroni, Alberto Credi, and Margherita Venturi Copyright 2010 John Wiley & Sons, Inc.
477
478
ELECTROCHEMILUMINESCENT SYSTEMS AS DEVICES AND SENSORS
include many more organic molecules with description of two main, single and mixed, types of ECL systems. In the first type, the annihilating R þ and R ions are produced from the same precursor, in contrary to the second one involving annihilating A and D þ reactants generated from different electron acceptor (A) and electron donor (D) precursors. At the same time, results from ECL experiments have been rationalized comparing the energetics of the annihilation reaction with energies of the excited states within A and D molecules. It was concluded that the annihilation energy equal or larger than the energy of available excited states is required to populate them efficiently. If the energy released in the A þ D þ annihilation is larger than the energy of the lowest excited singlet state ES, it is possible to directly generate 1 A and/or 1 D. When the annihilation energy was significantly smaller than ES, but larger than ET, the energy of the excited triplet states 3 A and/or 3 D, the observed ECL emission was attributed to the formation of triplets followed by triplet–triplet annihilation.6 The proposed mechanisms were supported by studying magnetic field effects.7,8 It was also found that ECL emission can be observed in systems with the annihilation energy not large enough to populate the excited singlet and triplet states.9 In such cases, light is produced by means of the radiative deactivation of electrochemically generated exciplex or excimer. Followed soon after the first ECL experiments, the phenomenon was interpreted in a theoretical paper explaining the fundamental reasons for the excited state formation in electron transfer reactions.10 Strongly exergonic formation of the ground-state products requires extremely rapid dissipation of a large amount of energy into the vibrational modes of the molecular frames. This is very difficult for the reacting system and only a limited part of the reacting species follow directly that pathway, leading to a direct formation of the stable ground-state products. The formation of the excited states is less exergonic and less thermodynamically favored. However, less amount of energy needs to be vibrationally dissipated and therefore the process may be kinetically preferred. In fact, ECL provided the first evidence of so-called inverted Marcus region, where generation of the excited state simultaneously takes place with the energetically more favored ground-state formation. Some years later, at the beginning of the 1970s, first ECL system based on the luminescent transition metal complex tris(2,20 -bipyridine)ruthenium(II)-Ru (bipy)32 þ -has been reported.11 It was shown that the excited state 3 RuðbipyÞ3 2 þ can be generated in aprotic media by annihilation of the reduced Ru(bipy)31 þ and oxidized Ru(bipy)33 þ ions. Due to many reasons (such as strong luminescence and ability to undergo reversible one-electron transfer reactions), Ru (bipy)32 þ later has become the most thoroughly studied ECL active molecule. Under certain conditions, the annihilation reaction between Ru(bipy)31 þ and oxidized Ru(bipy)33 þ producing the emissive metal-to-ligand charge transfer (MLCT) state occurs with an efficiency fes approaching 100%.12 Consequently, the overall, actinometrically determined ECL efficiency (fecl expressed in emitted photons produced per annihilation event) is close to the luminescence quantum yields fem of the excited 3 RuðbipyÞ3 2 þ . The fecl value (0.05 at room temperature) found for the Ru(bipy)31 þ þ Ru(bipy)33 þ annihilation is commonly used as the
16.1
INTRODUCTION: ECL CONCEPT AND SHORT HISTORICAL OVERVIEW
479
efficiency standard for other ECL processes. Many other studies concerning Ru (bipy)32 þ ions have further followed, with the first report of ECL in aqueous solution involved Ru(bipy)32 þ and the oxalate C2O42 ion.13,14 Subsequently, other species were shown to act as ECL coreactants, two species among them are peroxydisulfate S2O82 ion15 and tri-n-propylamine TPrA.16 The discovery of Ru(bipy)32 þ /TPrA system allowed efficient ECL not only in aqueous media but also at physiological pH. These developments have resulted in a wide range of ECL based analytical applications in which Ru(bipy)32 þ plays a crucial role. Further ECL active emitters, mostly Ru(bipy)32 þ , attached as a label or tag to molecules of biological relevance (such as an antibody)17 have found wide applications in clinical and biological analytics, such as immunoassay and DNA analyses. Within the next few years, it was found that many other organic compounds as well as metal chelates are characterized by combination of electrochemical and spectroscopic qualities required of ECL active materials. The search of new organic and inorganic compounds that produce ECL has recently been attainted by finding extremely efficient ECL systems (with fecl yield approaching 0.67) based on tris (2-phenylpyridine)iridium(III)-Ir(ppy)3 chelate.18 Broad range of “classical” organic and inorganic luminophores has been recently extended by observation of emission from semiconductive nanoparticles.19,20 The continued development of experimental methods applied in ECL investigations (e.g., designing of ECL cell allowing measurements at low temperatures or use photometric setups based on CCD cameras) has resulted in significant progress in understanding the ECL phenomenon. Especially, the application of single-photon counting technique has greatly enhanced the detection limit, allowing to record ECL signal from ultramicroelectrodes21 and to monitor the real time rate of ECL generation.22 Among many other achievements, the observation of ECL phenomena in polymer films23,24 has extended ECL systems from the liquid phase to the solid phase, enabling some practical applications as well. In particular, continued progress in light-emitting solid-state electrochemical cells can lead to practically applicable displays and light sources. Evidence for the laser action generated at the ECL conditions may be regarded as one of the most exciting examples of electricityto-light conversion in ECL processes.25 Continued progress in ECL investigations is nicely reflected in the number of yearly published papers, starting from a very few at the beginning of the ECL story with exponential growth up to more than 200 papers per year in 2006 or 2007. Up to ca. 50% of the recently published journal articles are devoted to bio- and medicinerelated applications, clearly indicating that ECL now becomes a powerful and widely used analytical technique. Of course, a considerable number of publications reviewing different ECL aspects are also available in the literature. Some of them have been published just after26–28 or shortly following29–31 the discovery of the ECL phenomenon. Despite their fundamental significance, they are, due to the progress in ECL field, somewhat outdated, but still attractive for the readers interested in the ECL studies. A recently published very comprehensive monograph32 edited by Bard addressing all intrinsic aspects of ECL, can be recommended for everyone interested in the fascinating marriage of electrochemistry and photochemistry. For publication
480
ELECTROCHEMILUMINESCENT SYSTEMS AS DEVICES AND SENSORS
appeared after 2003, readers can also be referred to consult excellent, recently published reviews.33–40 Of course, the whole list (above 20 positions) of the ECL reviews is much longer. References to these papers, mostly published between 1980 and 2000, can be simply found in the above-mentioned reviews. Readers able to read Cyrillic alphabet may also consult ECL monograph appeared in Russian language.41 Taking into account the number of the recently (and previously) published works reviewing different aspects of ECL phenomenon, it is rather difficult to write the next one especially with the limited available space of this chapter. This chapter should only be treated as a first step in acquiring ECL knowledge, as this is an abridged presentation of the most important ECL issues. The author hopes, however, that, his approach, despite of a somewhat oversimplified presentation, will be useful, at least for readers without prior experience in ECL. 16.2 16.2.1
ECL PROCESSES, TECHNIQUES, AND MATERIALS Energetics and Routes of ECL Emission
Extensive investigations of ECL processes have established a general, sometime quite complicated, scheme for ECL emission. In the electrochemical reactions, an electron acceptor A is reduced to A at the reduction potential Ered and an electron donor is oxidized to D þ at the oxidation potential Eox as follows: A þ e > A
De > D
þ
ð16:1Þ ð16:2Þ
Following the heterogeneous formation at the electrode, the primary redox products diffuse together to form an encounter complex [A D þ ] wherein homogeneous electron transfer takes place according to three principally possible reaction pathways given as ½A D þ ! A þ D
ð16:3Þ
½A D þ ! 3 A þ D or=and A þ 3 D
ð16:4Þ
½A D þ ! 1 A þ D or=and A þ 1 D
ð16:5Þ
where 3 A/3 D and 1 A/1 D are the excited triplet and singlet states of the initial reactant A and D, respectively. The preference of the given reaction pathway is related to the free energy difference DG between annihilating ions given as DG ¼ FðEred Eox Þ
ð16:6Þ
where F is the Faraday constant. If the ion’s annihilation exergonicity DG is not large enough, a rapid back electron transfer leads only to regeneration of the starting compounds A and D in the ground state. The reaction is less interesting from the ECL
16.2
ECL PROCESSES, TECHNIQUES, AND MATERIALS
481
point of view, although it has been shown that such systems may still produce light by means of the radiative electron transfer (CT emission) between A and D þ due to exciplex (A 6¼ D) or excimer (A ¼ D) character of an activated complex [A D þ ] (E-route). Usually, such emissions are characterized by broad featureless bands, red shifted from the singlet emission of the individual molecules (1 A or 1 D). CT nature of an ECL emission can be confirmed by a correlation of emission maxima n~em with the difference in the redox potential.42 Usually, this can be done by ECL studies for a homologous series of acceptor A molecules with the same donor D (or vice versa). Typically, the wavelength of CT emission varies also with solvent polarity and/or temperature.42 Although excimers or exciplexes can be formed through spectroscopic excitation, its formation is particularly likely in ECL processes because of the proximity of the reacting species in an activated complex [A D þ ]. Additionally, excimers or exciplexes that are not available directly with spectroscopic excitation can be generated in polar solvents through ECL42 CT emission not seen in spectroscopic excitation can even dominate in the ECL spectra.9,42 An increase in the ion annihilation exergonicity DG to values comparable to the excited triplet-state energies (DG þ ET 0) opens an additional electron transfer channel (T-route). In the simplest case, only one excited triplet 3 A or 3 D becomes accessible. Triplet emission can be directly observed from the ECL systems involving rare earth and transition metal complexes with allowed (due to extensive spin-orbit coupling) triplet–singlet electronic transition. Organic triplets, however, are in most cases nonemissive in liquid solutions (at least at room temperature) and their phosphorescence is usually not observed, but the presence of nonemissive triplets generated in course of ECL processes has been beyond any doubt supported by the triplet interceptor technique.43,44 Only in some organic ECL systems involving molecules with very high phosphorescence quantum yield (e.g., benzophenone), triplet phosphorescence in ECL experiments has been directly observed.45 Usually, nonemissive organic triplets with a lifetime in the range of seconds can effectively participate in bimolecular triplet–triplet annihilation followed by fluorescence from the excited singlet 1 A or 1 D expressed as 3
A þ 3 A ! 1 A þ A
or
3
D þ 3 D ! 1 D þ D
ð16:7Þ
On the other hand, interaction between a triplet (e.g., 3 A) and a doublet (A or D þ ) leads to triplet deactivation with dissipation of electronic excitation energy expressed as 3
A þ A ! A þ A
or
3
A þ Dþ ! A þ Dþ
ð16:8Þ
The two processes (triplet annihilation and triplet quenching) occur in parallel and usually lead to rather low (<103) fecl efficiencies in organic ECL systems following the T-route. For inorganic or metalorganic triplets, with emission lifetime in the range of microseconds, triplet annihilation and triplet quenching processes are usually negligible in the qualitative as well as in the quantitative description of ECL mechanism.
482
ELECTROCHEMILUMINESCENT SYSTEMS AS DEVICES AND SENSORS
Further increase of the ion annihilation exergonicity to values comparable to the excited singlet-state energies (DG þ ES 0) makes the direct formation of 1 A or 1 D species possible (S-route) and the observed ECL emission corresponds to the radiative deactivation of 1 A and/or 1 D. Taking into account the fact that in ECL systems following the S-route emission occurs from directly populated emitters, one can expect that the directly produced excited singlet states become the dominant light source, even if the excited triplet are much more effectively generated.46,47 It should be noted, however, that the number of organic ECL systems exhibiting intense emission according to the S-route is rather limited. Usually, energy released in the electron transfer annihilation, especially in the mixed ECL systems in which both oxidant and reductant are generated from different precursors, is not large enough to populate the excited singlet directly. Obviously, if the electron transfer reaction is energetically sufficient to populate the excited singlet directly, the formation of the lower energetically lying excited triplets must also take place (mixed ST-route). In the cases of organic ECL systems, however, the efficiency of light production is usually much higher for the S-route as compared with the T-route. The relative probabilities of the both reaction pathways leading to the excited-state products are directly related to the energetics of the reactions (16.4) and (16.5). Because the process (16.4), favored also by spin multiplicity statistical factor 3, is more exergonic than the process (16.5); one can expect that the formation of the excited triplets should prevail in most ECL systems. In some case, two or even three from the main routes of light generation in ECL processes are operative together resulting in quite broad (several hundred nanometers) and appearing white emission.48 Last but not least, it should be noted that the description of ECL processes as a simple superposition of the two or three electron transfer channels is somewhat oversimplified from the mechanistic point of view. In real cases, the electron transfer processes are preceded and followed by the diffusion of reactants from and electron transfer products into the bulk solution, respectively. Moreover, ECL reactants and products are species with distinctly different spin multiplicities, which causes an additional kinetic complication because of spin conservation rules. Correspondingly, the spin up-conversion processes (e.g., between two forms of an activated complex 1 ½A D þ > 3 ½A D þ ) cannot be a priori excluded from the kinetic considerations. Consequently, a kinetic description of an ECL system (similarly to other bimolecular reactions, e.g., electron transfer quenching reactions in solution) may be a very complex task (cf. Chapter 4 in the Bard’s ECL monograph32). 16.2.2
Experimental Techniques in ECL Investigations
Light generation in ECL processes is realized in electron transfer reactions involving strong oxidant and reductant. Principally, both reactants can be prepared in the common chemical way if reactive intermediates are stable enough,49–51 but usage of the chemically produced oxidants and/or reductants seems to be quite cumbersome, especially in quantitative works. The electrochemical way appears to be much more practical (in the cases of relatively unstable intermediates) and advantageous (due to
16.2
ECL PROCESSES, TECHNIQUES, AND MATERIALS
483
selectivity of oxidation and reduction at controlled potentials). Both ECL reactants can be prepared using a dual-electrolysis stopped-flow method with further mixing of the electrochemically generated species,52,53 but generation of both, oxidized as well as reduced intermediates in the same solutions is much more frequently applied in ECL experiments. It may be realized using two closely spaced electrodes with ECL generation occurring at steady-state conditions using thin-layer cells,54–56 the ring and the disk of a rotating ring-disc electrode,57,58 or double-band microelectrodes in the collector–generator mode.59 In contrary to steady-state conditions, transient methods for studying ECL ordinarily involve generation of both ECL reactants at the same single electrode. In the triple-potential-step technique, with alternate generation of oxidized and reduced species, the experiment starts with the working electrode held at a potential of no electroactivity (cf. Fig. 16.1). The electrode potential is changed to value at which generation of the first reactant occurs and next to values corresponding to formation of the second reactant. Subsequently, the electrode potential is again changed to the initial value. The experiment is performed with potential limits chosen to ensure production of the electrogenerated intermediates in the mass-controlled region and to minimize the influence of secondary electrochemical reactions. Light emission is usually observed during the second reactant generation step in the course of a triple-potential-step sequence. One can also employ an extension of the triple-potential-step experiment applying multicycle reactant generation to create a sequence of closely spaced light pulses for spectral study. Such attempt is frequently used in ECL experiment performed on microelectrodes with light detection by means of CCD spectrograph. From the integrated current passed through the ECL cell in the forward step, one can obtain the number of reactant species generated. Comparison of this value with the total number of photons produced (integrated absolute emission–time curve) provides information about the ECL emission efficiency fecl and the yields of the
Figure 16.1 Plots of potential, current and ECL emission intensity vs. time in an ECL experiment performed with the triple-potential-step technique.
484
ELECTROCHEMILUMINESCENT SYSTEMS AS DEVICES AND SENSORS
excited-state generation fes. The observed ECL efficiencies can be directly related to the yield of the excited-state generation fes and to the emission quantum yield fem of the populated emitter given as fecl ¼ fes fem
ð16:9Þ
The information obtained from the triple-potential-step experiment is also contained in the parameters governing luminescence decay. The appropriate relationship (usually called the Feldberg plots) can be derived after an analytical treatment of diffusion (to and from the electrode) and the overall electron transfer rate. In the simplest cases (i.e., in direct formation of an emissive state in electron transfer event or in the ECL systems producing light mainly according to E-route), relationship between I(DtR) and (tF/DtR)1/2 terms is expected to be linear with the slopeto-intercept ratio equal to 0.959 as follows: IðDtR Þ ¼ a
pffiffiffiffiffiffiffiffiffiffiffiffiffiffi tF =DtR b
ð16:10Þ
where tF is the duration time of the first (forward) step of the triple-step sequence and DtR is the delay time from the start of the second (reverse) step.60,61 The linearity I(DtR) versus (tF/DtR)1/2 criterion may be used to distinguish between possible mechanisms in the investigated ECL systems. For example, in the cases of intrinsic contribution from triplet–triplet annihilation the Feldberg plots are nonlinear.62 Similar deviations from linearity, however, are also caused by ECL reactants’ instability. The more unstable the electrogenerated species, the larger are the deviations from the linearity.63 Thus, it is very important to rule out the instability of both ECL reactants before discussing the recorded Feldberg plots in more detail. The analysis should be performed very carefully because other parasitic processes may additionally affect the ECL transients.64–66 Frankly speaking, only in the case of “pure” S- or E-routes with linear Feldberg plots, this analysis may be conclusive. For a more complicated ECL systems, some additional verifications such as influence of magnetic field7,8 or triplet interceptor technique43,44 may be necessary. ECL can also be generated with a single potential step or potential (one-directional as well as bidirectional) scanning at an electrode. Depending on the polarity of the applied potential, both the luminophore and the coreactant species (see below) can first be oxidized or reduced at the electrode to form active intermediates. Primary product generated from the coreactant undergo then decomposition with creation of a powerful reducing or oxidizing agent that reacts with the oxidized or reduced luminophores producing the emissive excited states. Parallel recording of the current flowing through the cell and emitted light intensity as function of time (in the single potential scan experiment) can be used in ECL mechanisms studies in the same way as in the triple-potential-step technique.67,68 Mechanistic information can be extracted from the current and the light intensity recorded as a function of applied potential (during potential scanning in cyclic voltammetry experiments) as well.15,17,69 ECL measurements can be also combined with other techniques. It greatly increases the difficulty of the experiment, supplying, however, more independent
16.2
ECL PROCESSES, TECHNIQUES, AND MATERIALS
485
information. For more detail concerning more sophisticated ECL techniques, the reader is recommended to consult Chapter 2 in the Bard’s ECL monograph32 and references cited therein. 16.2.3
ECL Materials: Some Most Important Examples
During ECL development, many inorganic and organic compounds have been tested as potential ECL luminophores. Extended lists can be found in the already cited reviews or in Chapters 6 and 7 in the Bard’s ECL monograph.32 Only some of them are mentioned in this chapter and only the most essential ones are briefly described. The selection has taken into account their significance in ECL progress and in potential or already realized applications. For many reasons, efficiency of the given ECL system is crucial for practical applications. Because the number of ECL systems with large fecl efficiencies, higher than 0.01 is rather limited, this criterion was also one of the most important one. 16.2.3.1 Inorganic ECL Systems As it was already mentioned above, inorganic ECL systems may produce light according to relatively simple mechanism with emission from the lowest excited, usually triplet state. Therefore, one may expect that such ECL systems may be very efficient. In fact, in some ECL systems involving transition metal complexes (with phosphorescence allowed by the spin-orbit coupling) the reported ECL yield fecl approaches the luminescence quantum efficiency fem indicating that the emitting state is produced with efficiency near unity. The most noticeable example is that concerning Ru(bipy)32 þ ions in acetonitrile solutions at a Pt electrodes with the reaction mechanism formulated as following. In the electrochemical reactions, the parent ions Ru(bipy)32 þ undergo70,71 one-electron reduction (with the added electron localized on individual ligand p*-orbitals) and oxidation (with removal of a metal t2g electron) followed by ion’s annihilation with the formation of the excited 3 RuðbipyÞ3 2 þ state and subsequent emission of light. RuðbipyÞ3 2 þ þ e > RuðbipyÞ3 þ
ð16:11Þ
RuðbipyÞ3 2 þ e > RuðbipyÞ3 3 þ
ð16:12Þ
RuðbipyÞ3 3 þ þ RuðbipyÞ3 þ ! 3 RuðbipyÞ3 2 þ þ RuðbipyÞ3 2 þ
ð16:13Þ
RuðbipyÞ3 3 þ þ RuðbipyÞ3 þ ! RuðbipyÞ3 2 þ þ RuðbipyÞ3 2 þ
ð16:14Þ
The found ECL efficiency12 strictly approaches (especially at lowered temperatures) the intrinsic luminescence efficiency of the populated emitter, allowing to conclude that (i) the formation efficiency of the excited 3 RuðbipyÞ3 2 þ is near unity and (ii) the thermodynamically favored direct formation of the ground-state products is kinetically inhibited. Other mononuclear ruthenium(II) chelates with ligands such as polypyridines,58,72,73 biquinolines,74 phenantrolines,75 or bipyrazine76,77 exhibit a similar behavior, but the measured fecl values have been found strongly depending on the ligand(s) nature. Usually, the observed ECL efficiencies are lower with respect to
486
ELECTROCHEMILUMINESCENT SYSTEMS AS DEVICES AND SENSORS
that found for Ru(bipy)32 þ what arises from differences in population efficiencies of the excited states emitting and their luminescence ability. Noteworthy, Ru(II) complexes with 4,40 -diphenyl-2,20 -bipyridine or 4,7-diphenyl-1,10-phenanthroline emit at ECL condition few times better than Ru(bipy)32 þ . Up till now, the maximum ECL efficiency among the mononuclear ruthenium chelates has been found for tris (4,7-diphenyl-1,10-phenantroline)ruthenium(II) ion, 0.24 in acetonitrile78 and 0.20 in butyronitrile79 solution at room temperature. At lowered temperatures, the observed ECL efficiency of this system is still higher79 in similar way as it was found for Ru(bipy)32 þ or tris(2,20 -bipyrazine)ruthenium(II) complexes.12,77 The observed increase in the fecl efficiencies is mostly related to the temperature induced enlargement in the fem yields of the given Ru(II) chelate. ECL investigations of dinuclear or polynuclear Ru(II) complexes have been recently performed with hope for developing more efficient electrochemiluminescent materials. Centrally80–82 or peripherally83 functionalized dendrimers with active RuL32 þ chelate units can produce higher (up to four to five times) ECL intensities as compared to their monomeric RuL32 þ precursors alone. It was also found that the ECL intensities of metallodendrimers become larger as the multiplicity of the involved Ru(II) units increases. Similar observations have been reported for binuclear Ru(II) complexes with weak interaction between both metallic centers.84–88 These results indicate that further studies in such direction may result in design of still more efficient ECL systems based on Ru(II) luminophores. Due to the many similarities between Ru(II) and Os(II) chelates, the later have also been investigated as ECL active compounds. The results obtained for Os (bipy)32 þ or Os(phen)32 þ ions89,90 clearly established that 3 OsðbipyÞ3 2 þ or 3 OsðphenÞ3 2 þ excited states are populated at ECL conditions but the obtained ECL efficiencies have been found to be distinctly smaller (most probably because their low emission efficiencies) as compared to their Ru(II) counterparts. Much higher ECL intensities were observed for Os(II) complexes of Os(bipy)2L2 þ or Os (phen)2L2 þ type with supplementary bidentate phosphine or arsine ligands L such as 1,2-bis(diphenylphosphino)ethane or 1,2-bis(diphenylarsino)ethane.89,90 Unlike Os (bipy)32 þ and Os(phen)32 þ , many of Os(bipy)2L2 þ or Os(phen)2L2 þ complexes show very intense photoluminescence with efficiencies two to three orders of magnitude higher than their parent derivatives. Consequently, the observed ECL efficiencies obtained in electron transfer annihilation between the electrochemically generated Os(phen)2L þ and Os(phen)2L3 þ ions were found to be close to that for Ru (bipy)32 þ . The ECL emission, much more intense than for the previously studied Os (bipy)32 þ and Os(phen)32 þ chelates, was observed in the ECL system based on tris (2,20 -bipyrazine)osmium(II)-Os(bprz)32 þ -with the emission from 3 OsðbprzÞ3 2 þ populated in electron transfer between the electrochemically generated Os(bprz)3 þ and Os(bprz)33 þ ions.91 ECL systems based on Os(II) complexes, although have some advantages such as their relatively low oxidation potentials, are pointedly less elaborated as compared to those based on Ru(II) derivatives. One can expect, however, that further development of the ligand(s) coordinating Os(II) kernel will result in new efficient ECL systems, similarly as it has taken place for Ir(III) chelates.
16.2
ECL PROCESSES, TECHNIQUES, AND MATERIALS
487
Figure 16.2 Cyclic voltammograms and ECL emission spectra of Ru(bipy)32 þ (right) and Ir (ppy)3 (left) chelates recorded in the author laboratory. (See the color version of this figure in Color Plates section.)
Tris(2-phenylpyridine)iridium(III)-Ir(ppy)3-molecule, formally isoelectronic with Ru(bipy)32 þ ion, exhibits many properties required for ECL reactants (Fig. 16.2). Ir(ppy)3 complex is strongly green emissive with phosphorescence occurring from the excited MLCT state 3 IrðppyÞ3 .92 Ir(ppy)3 can be also oxidized and reduced to quite stable Ir(ppy)3 þ cation and Ir(ppy)3 anion, respectively.93 Shortly after report describing application of Ir(ppy)3 complex in high-efficiency organic light-emitting devices based on electrophosphorescence,94 several authors have reported ECL emission from benzonitrile, acetonitrile, or acetonitrile–benzene solutions containing Ir(ppy)3 chelate.95–97 By analogy with Ru(bipy)32 þ complex the ECL mechanism was formulated as follows: IrðppyÞ3 þ e > IrðppyÞ3 þ
ð16:15Þ
IrðppyÞ3 e > IrðppyÞ3
ð16:16Þ
IrðppyÞ3 þ IrðppyÞ3 þ ! 3 IrðppyÞ3 þ IrðppyÞ3
ð16:17Þ
IrðppyÞ3 þ IrðppyÞ3 þ ! IrðppyÞ3 þ IrðppyÞ3
ð16:18Þ
Soon after these reports, ECL systems based on Ir(ppy)3 were quantitatively investigated in acetonitrile-dioxane solutions18 with reported ECL emission efficiency for the single Ir(ppy)3 þ Ir(ppy)3 system as high as 0.30. It has have also found that electron transfer between the radical anions of aromatic nitriles and ketones A and Ir(ppy)3 þ allows the direct population of the excited strongly emissive 3 IrðppyÞ3 with still higher yields.18 The fecl value of 0.67 (close to the excited 3 IrðppyÞ3 luminescence yield of 0.75) has been found for Ir(ppy)3 þ /2-cyanofluorene system.18 The fecl value of 0.67 is, to the best our knowledge, the highest ECL efficiency reported until now. Extremely high ECL efficiencies seem to be a common feature of the homolepticIrL3 as well as the heteroleptic-L2Ir(X) iridium(III) cyclometallated complexes. Extremely high ECL efficiencies (up to 0.55) were observed via ion annihilation between the electrochemically generated L2Ir(acac) þ or L2Ir(pico) þ cations (where
488
ELECTROCHEMILUMINESCENT SYSTEMS AS DEVICES AND SENSORS
Figure 16.3 Structural formulae of the ligands L and emission colors of their L2Ir(acac) complexes in the Commission Internationale de l’E’clairage (CIE) coordinates. (See the color version of this figure in Color Plates section.)
L are different organic ligands depicted in Fig. 16.3 and acac/pico are acetylacetone/ picolinic acid anions, respectively) and radical anions A of aromatic nitriles according to following mechanism98–100: L2 IrðXÞ þ e > L2 IrðXÞ þ
ð16:19Þ
Ae > A
ð16:20Þ
L2 IrðXÞ þ þ A ! 3 L2 IrðXÞ3 þ A
ð16:21Þ
L2 IrðXÞ þ þ A ! L2 IrðXÞ þ A
ð16:22Þ
The bright ECL emission can also be generated in annihilation of the electrochemically generated L2Ir(X) þ and L2Ir(X) species with ligand L being 2-phenylisoquinoline.101 Similar ECL behavior has been also reported for cationic (ppy)2Ir (bipy) þ and (ppy)2Ir(phen) þ chelates.102 The above presented results clearly indicate the cyclometallated Ir(III) chelates are especially interesting materials for designing new multicolor and extremely efficient ECL systems with precise tuning of ECL emission colors, from cyan (480 nm) to deep red (640 nm), that can be simply realized changing ligands L and X attached to Ir(III) kernel. Even more blue ECL emission (420 nm) can be generated from (3,5-F2-ppy)2Ir(pico) chelate.103 Thus, undoubtedly, one can expect that Ir(III) chelates, analogously to Ru(II) complexes, will find similarly wide applications. First reports101–104 describing generation of ECL emission from the cyclometallated Ir(III) complexes in presence of the coreactants such as TPrA or S2O82 point evidently to possible use of Ir(III) chelates in ECL based analytics. To finish listing of fairly efficient inorganic ECL systems one should also mention halide molybdenum and tungsten clusters ions such as Mo6Cl142 or W6Cl142.105–109 The cited works have clearly showed that quite efficient ECL generation is also possible for all-inorganic species. The ECL of Mo6Cl142 was produced by alternate oxidation and reduction in acetonitrile solution at a Pt electrode expressed as
16.2
ECL PROCESSES, TECHNIQUES, AND MATERIALS
489
Mo6 Cl14 2 þ e > Mo6 Cl14 3
ð16:23Þ
Mo6 Cl14 2 e > Mo6 Cl14
ð16:24Þ
Mo6 Cl14 3 þ Mo6 Cl14 ! 3 Mo6 Cl14 2 þ Mo6 Cl14 2 Mo6 Cl14 3 þ Mo6 Cl14 ! Mo6 Cl14 2 þ Mo6 Cl14 2
ð16:25Þ ð16:26Þ
Although the photoluminescence efficiency fem ¼ 0.19 of the excited 3 Mo6 Cl14 2 ion is quite high, the measured ECL efficiency fecl ¼ 0.012 in acetonitrile solution was rather low. Somewhat larger ECL efficiency fecl ¼ 0.031 was found in the electron transfer annihilation between Mo6Cl143 and tri-p-tolylamine radical cation. Much higher efficiencies fecl ¼ 0.095, however, have been found in less polar dichloromethane or 1,2-dichlorethane solvents clearly demonstrating the crucial role of the reaction medium in ECL processes.109 A considerable amount of effort has gone into studying the ECL mechanism of Mo6Cl142 clusters with special attention to probe the kinetic and mechanistic aspects of highly exergonic electron transfer reactions.107,108 The very low energy of the excited 3 Mo6 Cl14 2 ion (ET ¼ 1.9 eV) allowed for the observation of ECL from the annihilation of Mo6Cl143 or Mo6Cl14 ions with a variety of electroactive organic oxidants and reductants over a wide annihilation energy DG range by simply varying the redox potential of the electroactive organic species.107 The dependence of ECL efficiency on the driving force of the electron transfer annihilation has been investigated to probe the fecl versus DG relationship. ECL efficiencies fecl have been measured for series of donors D (aromatic amines forming stable radical cations D þ ): Mo6 Cl14 3 þ D þ ! 3 Mo6 Cl14 2 þ D
ð16:27Þ
Mo6 Cl14 3 þ D þ ! Mo6 Cl14 2 þ D
ð16:28Þ
acceptors A (quinones and aromatic nitrocompounds forming stable radical anions A): Mo6 Cl14 þ A ! 3 Mo6 Cl14 2 þ A
ð16:29Þ
Mo6 Cl14 þ A ! Mo6 Cl14 2 þ A
ð16:30Þ
and N-alkilpyridinium cations R þ (forming stable N-alkilpyridinium radicals R ): .
Mo6 Cl14 þ R ! 3 Mo6 Cl14 2 þ R þ .
.
Mo6 Cl14 þ R ! Mo6 Cl14
2
þR
þ
ð16:31Þ ð16:32Þ
It was found that over a narrow free energy range fecl rapidly increases with asymptotically approaching a limiting value smaller than fem. Although the observed behavior remains in accordance with the Marcus model prediction, it was, somewhat unexpectedly, found that not the electron annihilation energetics is a only factor
490
ELECTROCHEMILUMINESCENT SYSTEMS AS DEVICES AND SENSORS
governing the observed efficiencies of the electrochemical excitation. Both the threshold energy and the limiting fecl values have been found to be depending on the chemical nature of organic auxiliary reactants. It indicates that partitioning between the simultaneously occurring processes leading to the ground- and excited-state products may be much more complex than predicted by the Marcus model.10 Similar results were later reported for the Ru(II) chelates110,111 suggesting that such behavior is a general rule. Convincing explanation of the observed inconsistencies remains, however, an open question. Nearly all ECL systems presented above are transition metal chelates, what is not especially surprising taking into account their appropriate combination of electrochemical and spectroscopic properties required of ECL luminophores. Thus, one could expect that any adequate metal/ligand(s) combination should lead to similarly efficient ECL systems. At the present stage of ECL investigations, however, it is the only promising possibility. In most cases, ECL emission can be generated with very low efficiencies, much lower than expected from the luminescence quantum yields of the investigated luminophores, despite the fact that the energies released in the electron transfer annihilations are large enough for efficient excitation. It means that our understanding of the electron transfer excitation in the bimolecular reactions is rather crude and further quantitative investigations are absolutely necessary to clarify the roles of all factors affecting the process. 16.2.3.2 Organic ECL Systems Electrochemical excitation in organic ECL systems is somewhat more complicated as compared to inorganic ones resulting usually in relatively low ECL efficiencies. It is especially true for mixed organic ECL systems where both annihilating reactants, A and D þ , are generated from different precursors. Most of intense organic ECL systems are these with both annihilating partners generated from the same parent precursor. Aromatic hydrocarbons, with 5,6,11,12-tetraphenyltetracene, rubrene (RUB) and 9,10-diphenylanthracene (DPA) as the most noticeable examples among them, have played in the ECL studies the same key role as in the organic electrochemistry and the organic UV–vis spectroscopy. Stability of their radical ions, cations RUB þ or DPA þ and anions RUB or DPA,112 together with the luminescence quantum yields of 1 RUB or 1 DPA almost unity113 at room temperature have fascinated many ECL investigators just from the beginning of ECL story. The energy released during the annihilation of DPA ions is sufficiently large to directly populate both the excited singlet state (1 DPA with ES ¼ 3.06 eV) and the excited triplet state (3 DPA with ET ¼ 1.8 eV)114: DPA þ DPA þ ! 1 DPA þ DPA
ð16:33Þ
DPA þ DPA þ ! 3 DPA þ DPA
ð16:34Þ
DPA þ DPA þ ! DPA þ DPA
ð16:35Þ
The free energy for ground-state product formation is correspondingly so negative that this pathway lies in the Marcus inverted region. Consequently both, energetically
16.2
ECL PROCESSES, TECHNIQUES, AND MATERIALS
491
accessible, excited-state formation processes are dominating DPA þ DPA þ annihilation, yielding extremely bright ECL emission with an ECL efficiency that can approach 0.25 in 1,2-dimethoxyethane solution at low electrolyte concentrations.115,116 Because both pathways to form excited 1 DPA and 3 DPA states are governed by the spin statistics of the annihilating doublet radical ions, the reported ECL efficiency represents the theoretical maximum value for ECL from DPA singlets assuming that both pathways are diffusion-limited processes.115,116 Usually, however, ECL efficiency of DPA is smaller, not exceeding a few percent and varying with solvent polarity and ionic strength.114–116 Nevertheless, for any investigated solvent/ electrolyte compositions, DPA þ DPA þ annihilation is among the most efficient single organic ECL systems. The high ECL yields found for DPA indicate undoubtedly that light emission proceeds predominantly via direct formation of 1 DPA according to reaction (16.33) with only marginal contribution from TTA processes involving 3 DPA produced in reaction (16.34). Lack of magnetic field effect in this single ECL system supports additionally the dominance of the S-route.7,8,117 ECL behavior of rubrene corresponds also to concurrent generation of the excited singlet 1*RUB and triplet 3 RUB states.1,2,63,117–120 ECL behavior of rubrene, however, is somewhat more complicated than that found for 9,10-diphenylanthracene. The main difference arises from extremely low energy of the lowest excited 3 RUB state with ET as small as 1.15 eV,121,122 two times smaller than the energy of the lowest excited 1 RUB state with ES ¼ 2.29 eV.123 Therefore, population of two neighboring triplets, simultaneously with the formation of the excited singlet 1 RUB, may take place as follows: RUB þ RUB þ ! 3 RUB þ 3 RUB ! 1 RUB þ RUB
ð16:36Þ
This is leading to abnormally efficient triplet–triplet annihilation with the upconversion occurring before the triplets pair formed is separated.65,124 The three accessible ways of the excited 1 RUB state generation make ECL mechanism of rubrene quite complex with light production from a mixture of the S- and T-routes with their balance governed by solvent polarity and viscosity or temperature. Consequently, it results in strong variation of the observed ECL efficiencies with experimental conditions.119,124 Maximal fecl values, up to 0.04–0.05, have been found at room temperature in solvents such as tetrahydrofurane, 1,2-dimethoxyethane, or benzonitrile. Important contribution of the T-route in light production by means of RUB þ RUB þ annihilation leads furthermore to ECL efficiency of this single ECL system altered in the magnetic field.64,65,117,118 DPA and RUB can be regarded as model compounds for the single ECL systems based on polyaromatic hydrocarbons-PAHs (Fig. 16.4). Although PAHs exhibit many of the features required to achieve intense ECL emission, the observed ECL efficiencies are usually very low, sometime orders of magnitude lower than for DPA or RUB. In most cases, it may be rationalized taking into account the fact that aromatic hydrocarbons usually do not form enough stable radical cations. The best examples are DPA or RUB being much more efficient ECL luminophores than their parent aromatic hydrocarbons (anthracene or tetracene). Brighter ECL than their
492
ELECTROCHEMILUMINESCENT SYSTEMS AS DEVICES AND SENSORS
Figure 16.4 Structural formulae of polyaromatic hydrocarbons—brightness organic ECL luminophores: blue emissive 9,10-diphenylanthracene, green emissive 5,12-diphenylnaphthacene, orange emissive 5,6,11,12-tetraphenylnaphtacene (RUB), and red emissive bis-4,40 (7,12-diphenyl)-benzo[k]fluoranthene.
unsubstituted counterparts was also observed for 5,11-diphenyltetracene or 9,10dimethyl-7,12-diphenylbenzo[k]fluoranthene.125 Alkyl or aryl functionalization tends in formation of more stable radical cations that results brighter ECL emission. Due to sterical effects the functionalization may also results in weaker intermolecular interactions between R and R þ that marginalizes the formation of weakly emissive excimer 1 R2 and 3 R2 states with increasing efficiency of the excited 1 R populations.126 ECL emission has been also observed in the mixed ECL systems involving PAHs with reaction partners like aromatic amines or ketones forming radical cations D þ or radical anions A, respectively.114,127 Such approach solves the problems caused by the instability of ECL reactants but lowers distinctly the free energy available for the formation of an excited state. Usually, the energy released in electron transfer between A þ D þ ions is insufficient to populate emissive 1 A or 1 D states directly and the annihilation of the radical ions usually generates only nonemissive 3 A or 3 D triplets that produce light via triplet–triplet annihilation. Consequently ECL efficiencies in the mixed ECL systems are usually very low. Only in some cases, when radiative electron transfer between A þ D þ species is operative, relatively intense 1 ½A D þ exciplex emission can be observed. Despite the unproblematic selection of an appropriate combination of stable A and D þ for mixed organic ECL systems, their low ECL efficiencies limit very much their potential applications. Thus, the search of new organic ECL systems with high efficiencies should be rather focused on single ECL systems. It is, however, a quite difficult task because limited number of organic molecules forming both stable radical cation R þ and radical anion R. Appropriate electrochemical behavior can be anticipated for organic molecules possessing both donor D and acceptor A chromophores (though D and A are not necessarily different) bonded one to another. In the simplest case, it may be realized for A and D subunits directly linked by a formally single bond. Typical examples of asymmetric A–D and symmetric A–A derivatives are aryl and heteroaryl derivatives of aromatic amines66,128–133 or 9,90 bianthryl and 10,100 -dimethoxy-9,90 -bianthryl,134–136 respectively. Photoexcitation of such A–D or A–A compounds, intramolecular donor–acceptor systems (Fig. 16.5), leads usually to electron transfer from D to A with charge separation.137 The obtained intramolecular charge transfer singlet 1 ðA D þ Þ states are highly polar, strongly stabilized in polar solvents and lie energetically lower than the locally excited singlet states of both subunits A and D. The luminescence originating from the
16.2
ECL PROCESSES, TECHNIQUES, AND MATERIALS
493
Figure 16.5 Examples of the intramolecular donor–acceptor system and their emission spectra-aryl derivatives of N,N-dimethyl-aniline: BDMA: 4-(9-acridyl)-N,N-dimethylaniline, ADMA: 4-(9-anthryl)-N,N-dimethylaniline, PDMA: 4-(1-pyrenyl)-N,N- dimethylaniline, and NDMA: 4-(1-naphthyl)-N,N-dimethylaniline. (See the color version of this figure in Color Plates section.)
excited 1 ðA D þ Þ state corresponds to the intramolecular radiative electron transfer 1 ðA D þ Þ ! AD þ photon. The mechanism of the ECL processes of such systems may be formulated as follows.46,132–134,138 In electrochemical reactions the radical anion A–D and the radical cation A–D þ are formed from the parent A–D molecules: AD þ e > A D
ð16:37Þ
ADe > AD þ
ð16:38Þ
After electrochemical reduction electron is placed on the lowest unoccupied molecular orbital (LUMO) of the acceptor subunits of A–D molecule. In the electrochemical oxidation, an electron is correspondingly removed from the highest occupied molecular orbital (HOMO) of the donor moiety. In the diffusioncontrolled reaction electrochemically generated ions A–D and A–D þ form an activated complex A–D þ A–D for which the following reaction pathways are possible: AD þ A D ! AD þ AD
ð16:39Þ
AD þ A D ! 3 ðA DÞ þ AD
ð16:40Þ
AD þ A D ! 3 ðAD Þ þ AD
ð16:41Þ
AD þ A D ! 1 ðA D þ Þ þ AD
ð16:42Þ
AD þ A D ! 3 ðA D þ Þ þ AD
ð16:43Þ
where 1 ðA D þ Þ and 3 ðA D þ Þ represent excited intramolecular singlet and triplet states, respectively. The emissive 1 ðA D þ Þ state is deactivated with light emission with quantum efficiency fem), but the nonemissive 3 ðA D þ Þ undergoes nonradiative relaxation to the lower lying 3 ðA DÞ or 3 ðAD Þ and later to the
494
ELECTROCHEMILUMINESCENT SYSTEMS AS DEVICES AND SENSORS
ground-state A–D. The formation of 3 ðA D þ Þ was introduced in the above reaction pattern, because of the expected small energy gap between both forms of the excited charge transfer states in the approximation of a weakly interacting radical ion pair forming nearly degenerate 1 ðA D þ Þ and 3 ðA D þ Þ species. ECL processes of A–A molecules follow nearly the same reaction pattern46,136 with the annihilating species being A–A and A–A þ instead of A–D þ A–D þ ions and, due to symmetry reason, both locally excited triplets, 3 ðAD Þ and 3 ðA DÞ combined into 3 ðAA Þ one. Despite relatively complex mechanism of the electrochemical excitation of A–D or A–A molecules, fairly large ECL efficiencies, similar to that found for PAHs, have been observed (up to 0.05 and 0.03 for A–D133 and A–A136 molecules, respectively). High ECL emission efficiencies fecl as well as high yields of the 1 ðA D þ Þ state population fes have been found clearly establishing that the reaction (16.42) corresponding to the nearly isoenergetic (resonant) electron transfer is competitive with other processes occurring in parallel. The fes values have been interpreted in the context of Marcus’ model in which the necessary rate constants were evaluated from the pertinent electrochemical and spectroscopic data of the parent A–D molecules demonstrating also that the electron transfer model for ECL processes may be used to predict the ECL efficiency.46,133 It was also found that the ECL efficiency for A–D or A–A systems is mostly limited by the rates of reactions (16.40) or (16.41), that is, the formation of locally excited triplets 3 ðA DÞ or 3 ðAD Þ, which depend upon the reaction exergonicity in a manner consistent with the Marcus inverted region. Thus, because of the negligible slow direct formation of the ground-state reaction (16.39), extremely high ECL efficiencies are expected for the donor–acceptor system with all locally excited states, singlet and triplets lying energetically above the excited charge transfer states. In such cases, the expected formation efficiencies of the emissive singlet charge transfer 1 ðA D þ Þ and the nonemissive excited triplet charge transfer state 3 ðA D þ Þ are expected to be 0.25 and 0.75, respectively. Because direct relaxation 3 ðA D þ Þ ! AD is a spin forbidden process, it can be also expected that the electrochemically populated 3 ðA D þ Þ will be deactivated in the form of delayed fluorescence through the nearly isoenergetic singlet 1 ðA D þ Þ. Comparative study of the ECL behavior of two A–D systems: 4(3,6-di-tert-butylcarbazol-9-yl)benzonitrile-CBP and 4-(3,6-di-tert-butylcarbazol9-yl)terephthalonitrile-CTO in acetonitrile solutions have undoubtedly confirmed the above expectation.138 CBP molecule with a lowest lying locally excited 3 ðp; p Þ state (with the excitation localized in the donor—carbazole subunit) shows an ECL behavior resembling that previously reported for A–D compounds. On the contrary, the results point to considerable changes in the ECL mechanism for CTO system in which the lowest triplet state is of charge transfer nature. High ECL efficiencies, 0.027 for CBP and 0.011 for CTO were found at room temperature with an effective population yields of the fluorescent state 1 ðA D þ Þ markedly different for both compounds, 0.066 for CBP and unusually large, 0.64 for CTO. At low temperatures the fes efficiency of the fluorescent-state formation for CTO in ECL experiments was still higher, approaching unity.138 So high fes efficiency, the highest reported until now, is exceptional among organic ECL systems.
16.2
ECL PROCESSES, TECHNIQUES, AND MATERIALS
495
ECL investigations of molecules containing donor and acceptor subunits have been recently extended for systems where D and A are linked by more or less inert bridge such as alkyl, alkenyl, or alkinyl chain. The simplest case corresponds to intramolecular donor–acceptor systems in which A and D subunits are linked by CH2, C2H4, or longer aliphatic hydrocarbons. Both types of ECL luminophores, ACnH2nD128 and ACnH2nA135 have been investigated. ECL emissions have been attributed to intramolecular exciplexes or excimers generated via electron transfer annihilation between ACnH2nD ( ACnH2nA) and ACnH2nD þ (ACnH2nA þ ) radical ions, respectively. Luminophores such as ACH¼CHD or AC CD with acceptors and donors linked by a double or triple bond have recently been investigated in an attempt to understand the effect of structure on their spectroscopic, electrochemical and ECL behavior. The investigated AC CD derivatives were donor–acceptor systems with acceptors such as anthracene or acridine,139,140 quinoline or isoquinoline,141,142 and coumarin.143,144 ECL emission þ was observed via electron transfer annihilation of AC CD and AC CD radical ions. Vinyl bonded ACH¼CHD derivatives have been reported to exhibit 145–147 It should be ECL behavior similar to that found for AC CD molecules. noted, however, that ECL emission efficiencies observed for the above-listed linked donor–acceptor systems were smaller than reported for directly bounded A–D molecules. On the other hand, the mentioned examples shown clearly that the intramolecular donor–acceptor luminophores offer large flexibility in designing of new ECL systems. Report describing ECL behavior of A–D–A and A–D–D–A molecules nicely supports the above conclusion.148–150 16.2.3.3 ECL Coreactants ECL emission from the above-described ECL systems can be generated by annihilation reactions between independently produced oxidized and reduced species. Electron transfer annihilation between these species can be ECL operative if the energy released in the annihilation is sufficient to produce the excited state. Direct emission from these states can be only observed if their energies are as high as 2–3 eV. Consequently, annihilation reactions of interest are very energetic and the annihilating species are produced at relatively negative reduction and relatively positive oxidation potentials, respectively. Therefore, annihilation ECL processes are usually investigated in organic aprotic solvents because the potential windows for aqueous solutions are generally too narrow to allow convenient electrolytic generation of both the oxidized and the reduced ECL precursors. Generation of ECL emission in aqueous media is possible by adding certain species (called coreactants) into solutions containing luminophore species. Depending on the luminophore and the coreactant combination both species can first be oxidized or reduced at the electrode to form active intermediates. Typically, intermediates generated from luminophore are stable, whereas these formed from the coreactant undergo chemical decomposition with production of a powerful reducing or oxidizing agent. Highly reducing intermediate species are generated after an electrochemical oxidation of a coreactant, or highly oxidizing ones are produced after an electrochemical reduction, correspondingly. Strong reducing/oxidizing intermediates can further react with the oxidized/reduced luminophore to produce the excited
496
ELECTROCHEMILUMINESCENT SYSTEMS AS DEVICES AND SENSORS
states that emit light. Coreactants are useful in aqueous solutions, but their use makes ECL generation possible in organic aprotic solvent as well for luminophores that have only a reversible electrochemical reduction or oxidation. The main difference between “classical” annihilation ECL experiments (when all reactants can be principally regenerated) and those performed using coreactants is that the coreactant is consumed via electrochemical–chemical reactions and only luminophore species are restored. Although there are many examples of organic and inorganic luminophores that exhibit ECL using coreactant technique, the significant majority of published works concerned with generation of the excited 3 RuðbipyÞ3 2 þ , or its analogs, as the emitting species, because of their analytical application. As mentioned above, both reducing and oxidizing agents can be electrochemically generated from appropriate coreactants and the both agents can be applied in generation of ECL emission from the excited 3 RuðbipyÞ3 2 þ species. The reducing agents, generated by means of electrochemical oxidation, seem to be, however, much more important because potential window in aqueous solution is usually limited in the cathodic region due to hydrogen evolution. Amine-related coreactants,151,152 with tri-n-propylamine (TPrA)16 as the most important example allow generation of strong reductants by means of electrochemical oxidation. The Ru(bipy)32 þ /TPrA system, being starting point of ECL based analytics, has been already commercialized for immunoassay and DNA analyses. Generally, the ECL emission of this system is based on proton abstraction from . . the electrochemically generated TPrA þ cation (CH3CH2CH2)3N þ with further . . formation of TPrA radical (CH3CH2CH2)2NC HCH2CH3 (strong reductant with redox potential close to 1.5 V versus normal hydrogen electrode NHE153) reacting with Ru(bipy)33 þ oxidized species: TPrAe ! TPrA
.þ
! TPrA þ H þ .
ð16:44Þ
TPrA þ RuðbipyÞ3 3 þ ! ðCH3 CH2 CH2 Þ2 NC HCH2 CH3 þ 3 RuðbipyÞ3 2 þ .
.
ð16:45Þ Both reacting intermediates, TPrA and Ru(bipy)33 þ species, are produced simultaneously during electrochemical oxidation Actual ECL mechanism, however, is somewhat more complicated than expressed by the above reaction pattern with ECL emission from Ru(bipy)32 þ /TPrA system depending on the applied electrode potential. Usually, the direct oxidation of TPrA at the electrode occurs at more negative potentials than characteristic for the Ru(bipy)32 þ /Ru(bipy)33 þ redox couple. Generally, the ECL emission from Ru(bipy)32 þ /TPrA system as a function of applied potential consists of two emission waves (both associated with the emission from 3 RuðbipyÞ3 2 þ ) attributed to TPrA and Ru(bipy)32 þ oxidation, respectively.154 First . emission wave corresponds to annihilation of sufficiently stable TPrA þ (with halflife of 0.2 ms) and Ru(bipy)3 þ species with Ru(bipy)3 þ intermediate formed from . the reduction of Ru(bipy)32 þ by TPrA free radical: .
16.2
497
ECL PROCESSES, TECHNIQUES, AND MATERIALS
TPrA þ RuðbipyÞ3 2 þ ! ðCH3 CH2 CH2 Þ2 N þ C ¼ HCH2 CH3 þ RuðbipyÞ3 þ .
ð16:46Þ TPrA þ þRuðbipyÞ3 þ !ðCH3 CH2 CH2 Þ2 N þ C ¼ HCH2 CH3 þ 3 RuðbipyÞ3 2þ .
ð16:47Þ whereas the second emission wave follows mechanism expressed by Equations (16.44) and (16.45). It should be noted, however, that (additionally to direct . oxidation of TPrA at electrode) TPrA þ intermediate can be also produced via a “catalytic route” in which electrogenerated Ru(bipy)33þ reacts with TPrA: TPrAþRuðbipyÞ3 3þ !TPrA þ þRuðbipyÞ3 2þ .
ð16:48Þ
Consequently, the excited 3 RuðbipyÞ3 2þ state can be produced via three different . routes: (i) Ru(bipy)3 þ oxidation by TPrA þ cation radical, (ii) Ru(bipy)33þ reduction . by TPrA free radical, and (iii) the Ru(bipy)33þ and Ru(bipy)3 þ annihilation reaction. The ECL intensity for the first and second waves was found to be proportional to the concentration of both Ru(bipy)32þ and TPrA species in a very large dynamic range with reported detection limits as low as 0.5 pM155 for Ru(bipy)32þ and 10 nM156 for TPrA. In addition to Ru(bipy)32þ , many other metal chelates and aromatic compounds or their derivatives can produce ECL in the presence of TPrA as a coreactant upon electrochemical oxidation (cf. Chapter 4 in the Bard’s ECL monograph.32). Similarly to TPrA, a wide range of amine compounds have be examined as a coreactants taking part in ECL excitation of Ru(bipy)32 þ complex or its derivatives with finding that an amine should have at least one hydrogen atom attached to the acarbon that allows deprotonation process necessary to form a reducing free radical species. No other strict rules correlating the coreactant activity with the amine structure have been found, but ECL efficiency increases generally in the order primary < secondary < tertiary amines.151,152 The nature of substituents attached to nitrogen or a-carbon on tertiary amine molecule can also affect the ECL activity.157 It has been recently reported that tertiary amines such as tri-iso-butylamine67 or 2-(dibutylamino)ethanol158 are orders of magnitude more active than TPrA in commonly used Ru(bipy)32 þ /amine systems. Similarly to amines, many other compounds, exhibiting electrochemical oxidation followed by proton abstraction, can be applied as ECL active coreactants. Typical examples involve alcohols.159 The second important class of reducing agents is generated by means of oxidative decarboxylation of carboxylic acids. Electrochemical oxidation of oxalate ion C2O42 produces, in aqueous as well as in acetonitrile solutions containing Ru (bipy)32 þ , an extremely strong reductant, carbon dioxide radical anion CO2:13 C2 O4 2 e ! ½C2 O4 ! CO2 þ CO2
ð16:49Þ
The generated CO2 reductant, with redox potential close to 1.9 V versus NHE,160 reacts with the electrochemically formed Ru(bipy)33 þ ions with population of
498 3
ELECTROCHEMILUMINESCENT SYSTEMS AS DEVICES AND SENSORS
RuðbipyÞ3 2 þ emitter: CO2 þ RuðbipyÞ3 3 þ ! CO2 þ 3 RuðbipyÞ3 2 þ
ð16:50Þ
Optionally, Ru(bipy)22 þ can be reduced to Ru(bipy)3 þ that annihilate with Ru (bipy)33 þ to generate 3 RuðbipyÞ3 2 þ . ECL mechanism of the Ru(bipy)32 þ þ C2O42 system, however, may be somewhat more complicated than presented above, because the both, C2O42 and Ru(bipy)32 þ , electroactive species undergo electrochemical oxidation at similar electrode potentials. It may lead to indirect production of CO2 or Ru(bipy)33 þ intermediates involved in the ECL excitation expressed as C2 O4 2 þ RuðbipyÞ3 3 þ > C2 O4 þ RuðbipyÞ3 2 þ
ð16:51Þ
with the contribution of direct C2O42 and Ru(bipy)32 þ electrooxidation to the overall ECL behavior depending on the electrode potential applied as well as on other experimental conditions such as property of electrode or composition of the investigated solution.161–163 In acetonitrile solutions oxalate ion oxidizes easier than the Ru (bipy)32 þ complex69 and both active intermediates, Ru(bipy)33 þ and CO2, are generated in parallel without indirect formation of the oxidized Ru(bipy)33 þ species because transformation of the short-lived oxalate radical anion C2O4 into the carbon dioxide radical anion CO2 is an extremely rapid process making indirect production of the oxidized Ru(bipy)32 þ species unlikely. In an aqueous solution, however, Ru (bipy)32 þ is first oxidized at the electrode to Ru(bipy)33 þ cation capable to oxidize C2O42 in the diffusion layer close to the electrode surface. Similarly to C2O42 coreactant oxidative decarboxylation of organic acids . RCO2 produces a strong reductant R : RCO2 e ! ½RCO2 ! R þ CO2 .
.
ð16:52Þ
followed by recombination of R radical with Ru(bipy)33 þ cation. Typical examples are citric, tartaric, malic, and d-gluconic acid,164 gallic acid,165 ascorbic acid,166 or amino acids.167 Species that form strong oxidants upon reduction have been also served as coreactants. For example, the reduction of peroxydisulfate S2O82 ion15,168 proceeds as follows: .
S2 O8 2 þ e ! ½S2 O8 3 ! SO4 2 þ SO4 .
.
ð16:53Þ
with formation of strong oxidant sulfate radical anion SO4 with redox potential þ 2.9 V versus NHE.169 Ru(bipy)3 þ ions, generated in parallel upon electrochem. ical reduction of Ru(bipy)32 þ /S2O82 system, annihilate with an excess of SO4 2þ 3 intermediate with formation of the excited RuðbipyÞ3 species: .
SO4 þ RuðbipyÞ3 þ ! SO4 2 þ 3 RuðbipyÞ3 2 þ .
ð16:54Þ
16.3
PRACTICAL APPLICATIONS OF ELECTROCHEMILUMINESCENCE
499
The generated SO4 radical anion can also oxidize Ru(bipy)32 þ ions to Ru (bipy)33 þ with further formation of the 3 RuðbipyÞ3 2 þ species by means of Ru (bipy)33 þ þ Ru(bipy)3 þ annihilation. Due to many reasons (e.g., limited potential window of a Pt electrode in aqueous solutions), ECL emission from the Ru(bipy)32 þ / S2O82 system is usually generated in mixed CH3CN/H2O solutions.15 A strong ECL signal from the Ru(bipy)32 þ /S2O82 system, however, has been also detected at a carbon paste electrode in purely aqueous solution.170 ECL studies of many ruthenium chelates other than Ru(bipy)32 þ , including Ru(bprz)32 þ 171,172 or Ru(phen)32 þ 173 complexes, with use S2O82 as a coreactant were also reported. Peroxydisulfate coreactant has been also applied in ECL studies of several aromatic compounds in the mixed CH3CN/C6H6 solutions with finding that the formation of the emissive 1*R states is primarily caused by the R þ þ R annihilation with R þ generated via . oxidation of R by SO4 .174 Other examples of species that form strong oxidants upon reduction are peroxide system (such as benzoyl peroxide or H2O2) producing reactive oxidizing agents via electrochemical processes associated with OO bond cleavage. In both cases, oneelectron electrochemical reduction is followed by OO bond breaking with forma. . tion of strongly oxidizing species C6H5COO 175 and/or HO, 169 respectively. The above-presented examples clearly shown that application of coreactant does not require the direct electrochemical generation of both oxidized and reduced forms of a given luminophore. This can be a significant advantage because the use of a coreactant can make ECL possible even in solvents with a narrow potential window so that only a reduced or oxidized form of a luminophore can be produced. Additionally, it is still possible to generate ECL by using a coreactant for some fluorescent compounds that shown only a reversible electrochemical reduction or oxidation. Sometime, when the annihilation reaction between the oxidized and the reduced species is not efficient, the use of a coreactant may produce more intense ECL. .
16.3 PRACTICAL APPLICATIONS OF ELECTROCHEMILUMINESCENCE First ECL investigations have been mostly devoted to basic studies of the phenomenon as such with special attention to the use of the electrical energy for the generation of electronically excited states and the application of ECL for studying the kinetics and mechanism of strongly exergonic electron transfer reactions. From the earliest days of ECL story, however, there was also interest in using this phenomenon for any practical applications. Works in this area has been carried out by many scientists who have clearly established that ECL phenomenon can be really expanded from “pure science” to some practical applications. The intent of this chapter is to present shortly the possible applications of the ECL phenomenon. Some of them (such as detection of biologically important analytes) have been already realized or even commercialized. Other ones (e.g., lighting, displays, lasers, and energy up-conversion devices) are still remaining only very promising possibility, but one can expect that it is only a matter of time until continued progress in these fields will result in spectacular achievements.
500
16.3.1
ELECTROCHEMILUMINESCENT SYSTEMS AS DEVICES AND SENSORS
ECL Devices
16.3.1.1 Light-Emitting Electrochemical Cells From the beginning of ECL investigations, there was interest in using this phenomenon in practical application based on electricity-to-light conversion, motivated by the possibility of producing light-emitting devices. In the simplest case, it may be realized by inserting an appropriate solution between the two electrodes of a thin layer cell (solution electrochemiluminescent (SECL) cell). Of course, at least one of the electrodes must be transparent to allow observation of the emitted light. Optionally, both electrodes can be transparent with a see-through function. Usually, the SECL cells have a structure of glass/indium-tin oxide (ITO)/emitting solutions/ITO/glass with dissolved organic (e.g., DPA or RUB) as well as inorganic compounds (e.g., Ru(bipy)32 þ or Ir (ppy)3 chelates) applied as active luminophores.54,55,95,176–178 During operation of SECL cell reduced and oxidized species are produced at each electrode (cathode and anode, respectively) by injected electrons and holes. The generated, usually ionic, intermediates move due to ion conduction in the solution with excitation of emitting species produced by annihilation of both redox forms. SECL cells, with as well as without added supporting electrolyte, have been investigated with finding that longer lifetimes were obtained if the supporting electrolyte was absent. SECL cells using the electrolyte are less stable most probably because of side reactions of the impurities present in the electrolyte or the salt itself. As it could be expected distinctly voltage is required to operate the electrolyte-free devices because they are higher resistive as compared to their electrolyte containing analogs. Moreover, the mass transport mechanism in the electrolyte-free SECL cells is mostly convectional with very interesting phenomenon of patterned ECL emission.56,176,177,179 The theory of this interesting phenomenon of patterned ECL caused by electrohydrodynamic convection was later presented in more detailed studies.180–183 Considering the above phenomenon, one can conclude that ECL intensities in SECL cells may be largely affected by rereduction and/or reoxidation processes of electrogenerated species on counter electrodes with lowering collision chances between them. The increase in the intensity of ECL from cells consisting of nanopore array film (ZnO, Al2O3, or especially TiO2) sandwiched between two transparent conductive glasses have been recently reported.184,185 In such sandwiched structures, the electrohydrodynamic effects are distinctly smaller what results in higher probability of collision between electrogenerated species leading to more efficient electricity-to-light conversion. At the present state of the art, any practical applications of the SECL cells seem to be limited because to low stability of liquid ECL systems and technical problems with their tight encapsulating. Light-emitting cells involving light generation in a solid phase seem to be more stable because any effects ascribed to solvent impurities can be ruled out. Solid-state light emitting cells (SSLECs) (Fig. 16.6) are principally very similar to classic organic light-emitting diodes (OLEDs), but the presence of a (relatively) large concentration of mobile ions in LECs may lead to some important differences between them.186 The behavior of SSLECs is similar much more to that characteristic for typical ECL systems than to those of classical OLEDs, despite the
16.3
PRACTICAL APPLICATIONS OF ELECTROCHEMILUMINESCENCE
501
Figure 16.6 Schematic picture of a solid-state light-emitting electrochemical cell. (See the color version of this figure in Color Plates section.)
fact that in condensed phase the charge transport occurs mostly through electron hopping between fixed redox active sites while in solution phase the oxidized and reduced species move physically by diffusion and/or migration. The main characteristic of the SSLEC devices are (i) the insensitivity to the work function of the electrodes, (ii) a relatively low “turn-on” voltage, (iii) their symmetrical electrical and luminescence characteristics allowing SSLECs to be driven in the alternating current mode, and (iv) a large tolerance to the thickness of the emitting layer. Additionally, SSLECs have much more simple, single-layer architecture as compared to OLEDs, with multilayer structures that make their construction relatively easy. All the abovementioned advantages make LECs an interesting alternative to OLEDs, especially in the simplest cases where the emissive layer consists of only one substrate. Principally, SSLEC cells may be constructed by inserting any electroactive and luminescent materials directly between electrodes or by their immobilization in more or less inert matrixes. The immobilization of the active material in an inert matrix exhibits several attractive features allowing the variation of the emitter concentration as well as tuning the system’s properties by choosing different host materials. The highest possible emitter (polymeric materials or “small” molecules) concentration, however, can be achieved for the emissive layer with little amount (or even without) any host material. “Small” molecule devices for which LEC-type behavior has been reported are mostly based on ionic metal chelates.187 Inspired by results from the ECL solution studies, ruthenium(II) chelates RuL32 þ (with Ru(bipy)32 þ as the most noticeable and investigated example) have been extensively tested by many authors.188–198 Amorphous as well as crystalline films of RuL32 þ salts (especially with small counter ions such as BF4 or ClO4) exhibit reasonable ionic conductivity allowing observation of relatively large currents flowing through the constructed SSLEC cells with an external quantum efficiency of up to 0.034199 or 0.048.191 In the same way, polynuclear Ru(II) complexes have been also tested as active materials for the SSLEC devices.200 Additionally to ruthenium(II) chelates, osmium(II) related derivatives190,201 and rhenium(I) chelates202,203 have been also experienced. Recently, published papers describing the application of the cationic Ir(III) complexes in the SSLEC cells194,204–206 have shown that Ir(III) chelates are very promising candidates for electrochemically driven light-emitting devices of this kind. Blending the “small” molecule emitters with inert matrix increases sometimes the quantum efficiency, but higher operating voltages are usually required for the SSLEC cells operation. In such devices, charge injection occurs also via an electrochemical redox mechanism and the mechanism of light production is similar to solution ECL as it was found for solid-state light-emitting devices based on the tris-chelated Ru(II)
502
ELECTROCHEMILUMINESCENT SYSTEMS AS DEVICES AND SENSORS
complex with 4,7-diphenyl-1,10-phenanthroline disulfonate ligands immersed in poly(ethylene oxide) thin films blends.207 The SSELC cells with an active layer formed from chelate-based polymers exhibit similar behavior as it has been reported for the ester-substituted Ru(bipy)32 þ derivatives.208–211 The reported solid-state devices with the ester-substituted Ru(bipy)32 þ polymers have been found to be relatively stable.209 A new interesting class of the SSLEC cells based on a semiconducting polymer (noninert matrix) combined with phosphorescent complex have been recently reported.212 The active material was made of poly(2-(m-3,7-dimethyloctyloxyphenyl)-p-phenylenevinylene doped with a homogeneously dispersed Ru(II) ionic chelate playing roles of emitter and electron transfer mediator. Different colors of the emitted light were experienced from the constructed asymmetric ITO/(polymer Ru(II) complex)/(Al or Au) devices. At forward bias, the characteristic red emission from the excited triplet state of the Ru(II) complex was observed, whereas the opposite polarization resulted in green emission from the populated lowest excited singlet of the polymer host. Symmetric devices with Au as cathode and anode, however, have exhibited symmetric emission properties showing only red emission at both forward bias and reverse bias clearly indicating that an asymmetry in the charge injection process is needed to switch between both observed excitation mechanisms. At the present stage of investigations, the SSLEC devices seem to be uncompetitive with the best available electroluminescent OLEDs. The efficiencies and operating lifetimes of the SSLEC cells still need to be improved and their degradation mechanism clarified. One can hope, however, that further investigations focused on phosphorescent emitters will result in distinct improvement in the emission efficiency and stability of SSLEC devices. Particularly, the incorporation of electroluminescence and electrochemiluminescence in one light-emitting device may be a new way to improve the efficiency of light-emitting devices.213 Another interesting aspect of such hybrid-emitting cells is an easy way for color manipulation that may result in practical applications as signaling devices. 16.3.1.2 Electrochemically Driven Lasers Actions Potential application of ECL phenomenon in the light-emitting devices involves also construction of electrochemically pumped laser as it was already conceived and proposed approximately 35 years ago.214–217 Energy released in ECL processes can be used as a pump lasers action if electrochemical generation of the excited state with lasing ability would be performed in a laser resonator. This possibility seems to be particularly interesting because, in contrary to conventional dye lasers, electrochemically driven laser action would not require an additional laser source to pump optically the dye into its lasing excited state. Moreover, ECL based laser may also offer many advantages such as power, tunability, range of available wavelengths, and miniature dimensions. One can also expect that the lasing action achieved in a miniaturized ECL cell would result in making the family of dye lasers easier to maintain, cost effective, and user friendly. For any medium to act as a laser, it is essential that the number of molecules in the excited state exceeds the number of molecules in the ground state (condition called as population inversion). Usually, the excited states have a very short lifetime that results
16.3
PRACTICAL APPLICATIONS OF ELECTROCHEMILUMINESCENCE
503
in destroying the population inversion because the molecules in the excited state decay rapidly to the ground state. So, to hold a large number of molecules in the excited state, one is required to supply a lot of the excitation energy. The minimum energy required to invert the population in the given lasing system is called the threshold energy. The pumping rate achieved in a classical ECL experiment, however, is distinctly lower than necessary to overcome the threshold energy. The limitation arises mostly from to low recycling rate of the redox species involved in the excited-state population. Increased redox recycling can be achieved by reducing the space in which solution containing the ECL active dye flows between closely spaced and parallel oriented electrodes forming a Fabry–Perot cavity. This also reduces resistance, eliminating the need for a supporting electrolyte that may act as a quencher of the excited state populated in an ECL process. In addition, a Fabry–Perot cavity increases the stimulated emission rate. Such approach was succeeded at Basic Research Laboratories of Nippon Telegraph and Telephone Corporation in observation of laser action in ECL cell.25 The ECL active substance used in the experiment of Horiouchi et al. was 9,10-diphenylanthracene dissolved in N,N-dimethylformamide without any additional electrolyte. The continuously flowed DPA solution was electrolyzed in thin-layer electrochemical cell with platinum-film electrodes deposited on quartz substrates. Mirror and half-mirror platinum electrodes, prepared by controlling the deposition thickness, were positioned facing each other 2–7 mm apart. Light from the device was guided through multimode fiber and observed using the photonic multichannel analyzer. ECL spectra were recorded as a function of the faradaic current flowing through the cell with finding that the spectral distribution of the emitted light depends on the current density. The observed changes in the recorded spectra (narrowing effect similar to that observed in an optically pumped lasers) have been attributed to the stimulated emission from the electrochemically generated 1 DPA excited state. The results reported by Horiouchi et al.25 10 years ago have shown that the lasing action can be principally achieved in an electrochemiluminescence process. The results of Horiouchi et al.25 showed also that the construction of the electrochemically pumped dye laser is a rather difficult task because to all requirements that need to be fulfilled for an efficient ECL system one must add lasing ability of the applied ECL luminophore. This additional requirement drastically limits the selection of the ECL active substances potentially applied in the ECL laser devices. Undoubtedly, lack (to our best knowledge) of any further reports following the work of Horiouchi et al. agree with the later conclusion. It is rather difficult to propose ECL active compounds that can be applied in the ECL lasers but ECL systems exhibiting intra- or intermolecular charge transfer emission may be considered as potential candidates. Due to the high energies of the Franck–Condon states reached in the charge transfer emission, one can expect that the population inversion as well as lasing action (according to the four energy level scheme) in such ECL systems should be less difficult to achieve. 16.3.1.3 ECL Imaging Systems Electrogenerated chemiluminescence provides an excellent way for the direct observation of nonuniform current flows
504
ELECTROCHEMILUMINESCENT SYSTEMS AS DEVICES AND SENSORS
resulting from the ECL cell configuration as it was already mentioned in Section for the electrolyte-free SLEC cells with the patterned ECL emission. In a similar way, the geometry of irregularly shaped electrodes can be probed by ECL emission by appropriate changes of the electrode potential.218 Application of a high frequency square wave allows keeping the diffusion layer, associated with light-producing reaction layer, in proximity to the electrode surface that allows a quite high spatial resolution, potentially very useful for mapping the local mass transfer coefficient and localized hydrodynamic disturbances. The mapping is possible by means of changes in the ECL emission contrasts. ECL emission intensity has been applied as a measure of the local current density at rotating-disc electrodes for visualization of the edge effects with finding that the intensity of light, and thus the current density, is distinctly greater at the electrode edge than on the electrode interior.219,220 The imaging of hydrodynamic conditions at macroelectrodes has been also demonstrated in the case of shielding of the electrode surface by approaching gas bubbles.221 ECL has been also used to image ultramicroelectrodes and arrays electrodes.219–223 The ECL signal from the annihilation reaction between Ru(bipy)3 þ and Ru(bipy)33 þ species, generated at a platinum band microelectrodes separated by a micrometric insulating gap, has been observed with use of a confocal microspectrometer. It has allowed in situ photon detection of the ECL reaction with micrometric spatial resolution and the observation of the photon source distribution in the vicinity of the gap between the microelectrodes.224 High spatial resolution of light source based on ECL generated at an ultramicroelectrode has been applied in the scanning electrochemical microscopy (SECM) technique.225 The generation of ECL at a tip (with diameter of 1–10 mm) in the SECM moved in the vicinity of insulating and conductive area (an interdigitated array of Au band electrodes 30 mm wide spaced by 25 mm deposited on a glass substrate) has been recorded and an image was obtained by plotting the light detector response as a function of the tip XY position. It has been demonstrated resolution for this technique in the micrometer range limited by the tip size and tip-substrate spacing. The smaller is the tip size and tip-substrate distance the better is the resolution what seems to be understandable. Further studies performed with a submicrometers size tips have resulted in a distinct (better than 1 mm) improvement of the spatial resolution.226,227 Still better resolution of the SECM technique based on the ECL phenomenon seems to be principally possible, but it will require the generation of the excited states at extremely short distances from the electrode surface. Due to quenching of the excited states in the vicinity of metallic surfaces, one can expect the loss of the ECL signal for the emissive species generated at distances much closer than 100 nm.228 In the above-mentioned works, the ECL signal used for imaging was passed through the transparent substrate and collected by a photon detector (e.g., photomultiplier tube) located beneath the sample. Opposite approach uses selectively etched optical fibbers as working ultramicroelectrodes (Fig. 16.7).229,230 These ultramicroelectrodes are based on optical fibbers, which are coated by a thin layer of gold or other metals, insulated, and exposed at the tips. The metal surrounding an optical waveguide composes a microring electrode allowing the combination of optical and electrochemical methods in a very small system with simultaneous
16.3
PRACTICAL APPLICATIONS OF ELECTROCHEMILUMINESCENCE
505
Figure 16.7 Schematic diagrams illustrating the operating principles of ECL generation using SECM technique. (See the color version of this figure in Color Plates section.)
measurements using both the techniques. The exposed end of the fibber (conical or polished to a flat surface) collects the light from the ECL processes occurring on microring electrode. Microelectrodes based on optical fibbers (optoelectrodes) can be applied for concurrent electrochemical and optical measurements as an individual electrode as well as microelectrode arrays (imaging optical fibber bundles). In the latter case, the emitted light can be viewed using a microscope coupled with CCD camera that allows observation the light generated on each microelectrode from the assembled bundle. Single optoelectrodes can be used as a tip probe for scanning electrochemical/optical microscopy offering topographic imaging of nontransparent objects.231–233 Ordered microelectrode arrays based on optical fibber bundles allow the construction of structured microdevices collecting spatially resolved information without any mechanical movement.223,234–239 Potential applications of ECL-light generation on the optoelectrode arrays involve also the fabrication of nanosensors for analytical purposes with spatially resolved detection. 16.3.2
ECL Sensors
Electrogenerated chemiluminescence has rapidly gained importance as a sensitive and selective detection method in analytical science. During last decade, several review articles describing primarily different analytical applications of ECL have been published.33,35,38,39,240–245 The above-listed reviews start usually from description of ECL phenomenon from historical perspectives with a general introduction to spectroscopic and mechanistic considerations followed by a broad discussion of analytical application with more or less detailed classifications according to the type of analytes determined (including immunoassay and DNA probe assays). The reviews describe also the techniques that utilize ECL emission in analytical techniques, such as flow injection analysis (FIA), high-performance liquid chromatography (HPLC) or capillary electrophoresis (CE). Analytical aspects of ECL have been also extensively reviewed by Danielson (techniques) and Debad et al. (clinical and biological
506
ELECTROCHEMILUMINESCENT SYSTEMS AS DEVICES AND SENSORS
applications) in two chapters in the Bard’s ECL monograph.32 The following paragraphs will provide a general picture of ECL based detection. The intent of the following sections is only to present short state of the art of this particularly important aspect of ECL, by paying special attention to Ru(bipy)32 þ chelate, used mostly (at least in 90% of analytical applications published) as an active ECL reactant applied in ECL based analytics. 16.3.2.1 ECL Systems in Bulk Solutions Discovery of Ru(bipy)32 þ electrochemiluminescence was shortly followed by the use of the ECL phenomenon in number of analytical applications. Much interest has been paid to Ru(bipy)32 þ due to their excellent properties and compatibility with a wide range of analytes such as oxalate and organic acids, amines, amino acids and proteins, alkaloids, pharmaceuticals, pesticides, etc. Moreover, Ru(bipy)32 þ produces ECL with a high efficiency even in oxygen containing aqueous media what is from many points of view extremely important for practical applications. All ECL analytical applications Ru(bipy)32 þ are based on the production of Ru(bipy)33 þ oxidant followed by reaction of Ru(bipy)33 þ with an analyte species resulting in emission of light. ECL can be used in the quantitative analysis of various species, if the given analyte is able to react with Ru (bipy)33 þ as an active ECL coreactant. The analytical utility of ECL signal from the Ru(bipy)32 þ chelate is related to the ECL emission intensity being a function of the concentration of the analyte. As compared to other detection methods, ECL has the advantages of very low limits of detection and wide dynamic working ranges over several orders of magnitude. For example, ECL from Ru(bipy)32 þ has been used to measure the concentration of coreactants such as oxalate and peroxydisulfate to levels as low as 1013 M with linearity spanning over six orders of magnitude.17 Two procedures of electrochemical production of Ru(bipy)33 þ reagent can be principally applied as an analytical tool. In the first type, the reagent is generated in a cell that is remote from the site of its interaction with the analyte(s). In the second one, the reagent production, the subsequent ECL reaction, and light detection take place within the same electrochemical cell. The first approach was applied in the earliest analytical application of the electrochemically generated Ru(bipy)33 þ oxidant with the use of a steady-state chemiluminescence flow cell.246 Similar methodology was later applied in the ECL based detectors for FIA and HPLC applications.151,152 In these reports, the active Ru(bipy)33 þ oxidant was prepared by means of complete electrolysis of the Ru(bipy)32 þ precursor before mixing the reactant and the analyte containing solutions. Such approach seems to be somewhat disadvantageous because (i) large consumed volumes of Ru(bipy)32 þ solutions, (ii) extended time period is required to finalize analytical procedures, and (iii) limited stability of Ru(bipy)33 þ cation in the water containing media. Some limitations of the external generation mode can be overcome by the continual electrochemical oxidation of Ru(bipy)32 þ solution with the produced Ru(bipy)33 þ species pumped simultaneously into the connected detector.167,247 In the in situ technique, analyte and Ru(bipy)32 þ solutions are mixed before delivering them to the detection cell, which contains a working electrode.17,248–250 On the electrode surface, Ru(bipy)32 þ is oxidized to Ru(bipy)33 þ , which reacts with the
16.3
PRACTICAL APPLICATIONS OF ELECTROCHEMILUMINESCENCE
507
analyte. This method is more convenient experimentally and better reproducible than the external generation technique. Generation by in situ electrochemistry has been, due to a simpler instrumentation, the method most employed in the analytical applications published so far. Continuous development, with special attention paid to miniaturization of the ECL detectors,155,251 has resulted in the design of many different inline, flow-through electrochemical reactors including microfluidic sensors.252–255 For a more detailed description of the developed instrumentation, the reader is directed to the above-mentioned review articles. ECL analytical methods are typically based on electron transfer processes between active species generated electrochemically from ECL probe and analyte, both present in solutions. The detection of solution species not directly involved in the ECL process is possible as well, if the emission characteristics of active ECL luminophore change after interactions with an analyte of interest. It may be realized by attaching (directly or via linker) a recognition group to ECL active moiety. Such modifications can cause steric or electronic effects that can be modified in presence of an analyte affecting luminescence properties of an ECL luminophore. The sensitive nature of MLCT states in transition metal systems frequently used as ECL probes, make these types of complexes potentially useful in such applications as it was demonstrated for Ru (bipy)32 þ chelate linked with crown ether moiety.148,256,257 It has been shown that ECL efficiency from such systems is responsive to both the concentration and nature of the metal ion present in the investigated solutions. The obtained results clearly indicate the versatility of ECL for the detection of species not directly involved in the ECL excitation. The ECL activity of the simple crown ether derivatives of Ru (bipy)32 þ is, however, similarly affected by different metal ions making discrimination between different cations very difficult. 16.3.2.2 ECL from Modified Electrodes Despite the several advantages of the flow solution-phase ECL detection, its widespread application is rather limited. One of the most important disadvantages is the amount of consumed solution containing ECL probe. This limitation can be simply overcome by immobilizing the ECL probe on an electrode. The immobilization onto working electrode surface has some additional advantages over solution-phase ECL, such as simpler experimental design and (usually) enhancement of ECL sensitivity. Moreover, spatial control over the ECL reaction can be easy achieved as light emission is localized mostly on the electrode surface that simplifies collection of the emitted light leading to enhanced ECL signal. Similarly as it is in the case of solution phase, most of reported studies have been devoted to immobilization of Ru(bipy)32 þ due to universality of this chelate in ECL based analytical application. When electrochemical stimulation is applied to the system, Ru(bipy)32 þ immobilized on the electrode is oxidized to Ru(bipy)33 þ followed by further reduction of Ru(bipy)33 þ with a coreactant (e.g., TPrA) present in the bulk solution. Then ECL signal is obtained from the produced excited 3 RuðbipyÞ3 2 þ state, similarly as in the case of the solution-phase excitation. There are several methods allowing the immobilization such as (i) self-organization of Ru(bipy)32 þ derivatives by means of Langmuir–Blodgett (LB) or self-assembly
508
ELECTROCHEMILUMINESCENT SYSTEMS AS DEVICES AND SENSORS
monolayer (SAM) techniques, (ii) encapsulation of Ru(bipy)32 þ in an inert (e.g., polymer) matrix, or (iii) electroinitiated polymerization of appropriately substituted Ru(bipy)32 þ derivatives with formation of thin films of the redox polymer. All these three immobilization techniques have been applied for the construction of a solidstate ECL detector. The LB monolayer and SAM techniques have been applied for the creation of organized thin films on various electrode materials using systems consisting Ru (bipy)32 þ kernel with attached surfactant group(s) such as CH2NHOC (CH2)18CH3,258(CH2)12SH,259 or (CH2)12COO.260 The adsorbed LB or SAM monolayers, however, are easily detached from the surface, what limits their application as ECL based sensors. Another methodology involves covalent immobilization via substrate–silicon bond.261,262 In comparison to adsorption of the ECL reagent, covalent immobilization is advantageous as no leaching is observed, leading to increased reproducibility and sensor lifetime. Both adsorption and covalent mobilization methods produce thin films with relatively small amount of Ru(bipy)32 þ attached to the electrode that may result in correspondingly weak ECL signal because ECL intensity is generally proportional to the quantity of the emitting species. Moreover, one can recommend the semiconductor substrates such as ITO because the excited state of 3 RuðbipyÞ3 2 þ centers in the monolayers can be effectively quenched by the conductive metal substrate.228 To some extent, the above-mentioned limitations can be diminished by a more robust (multilayer) electrode surface coverage, for example, by using an electrostatic layer-by-layer method.263 The multilayer-film sensors based on the self-assembled Ru(bipy)32 þ layers on ITO electrodes seem to have higher sensitivity and stability than the monolayer ones.264,275 Another popular method is the immobilization of Ru(bipy)32 þ chelate in solid matrices. Since the early report13,14,266 of immobilization of Ru(bipy)32 þ on an electrode surface using Nafion cation-exchange polymer, quite a number of different methods and materials have been developed to immobilize Ru(bipy)32 þ . Though Ru (bipy)32 þ could be readily incorporated into Nafion films, the films suffered from slow charge transfer and long-term stability problems. Following Rubinstein and Bard reports, immobilization of Ru(bipy)32 þ chelate in the Nafion matrixes has been investigated by many researchers156,267–276 with distinct improvement of the ECL solid-state sensors characteristics. Particularly, application of carbon nanotubes (CNT) with development of CNT/Nafion composite films has resulted in ECL intensity 100–1000 times higher than that of the pure Nafion film-modified electrode.267 These findings can be explained by (i) more open structures of the composite film allowing diffusion of coreactant to R(bipy)33 þ sites and/or (ii) presence of conductive CNT facilitating the electron transfer from Ru(bipy)32 þ to the electrode surface. Similar results have been reported for the solid-state ECL from Ru(bipy)32 þ immobilized in TiO2-Nafion composite film.269,276 Eastman AQ55D polymer similar to Nafion has been proposed as an alternative in fabrication of solid-state ECL detectors.277–279 Some advantages of AQ55D (more hydrophilic character, more rapid response, antifouling properties) over Nafion make AQ55D very promising for further development of ECL sensors.
16.3
PRACTICAL APPLICATIONS OF ELECTROCHEMILUMINESCENCE
509
The incorporation of Ru(II) chelates in various types of silica sol–gel-based composite films is another mode of immobilization.262,280–284 Silica sol–gel materials provide some advantages over organic polymers such as the ease with which they can be prepared, modified, and doped with various reagents. Particularly, silica sol–gel films are stable upon continuous electrochemical oxidation or upon drying the gels in a high humidity environment. Directly synthesizing of polymeric materials at the electrode surface have also been applied in the fabrication of solid-state ECL sensors. Electrochemiluminescent polymer based on tris(4-vinyl-40 methyl-2,20 -bipyridine)ruthenium(II)-Ru 2þ (v-bipy)3 -can be easily obtained from acetonitrile solutions of the monomer by means of electrochemical polymerization.23,285,286 Electropolymerization allows simply adjusting of the layer thickness by controlling the charge passed through solution. Emission spectra from the electrodes modified with poly-Ru(v-bipy)32 þ were found to be similar to that characteristic for the Ru(bipy)32 þ precursor. The ECL signal, however, has a quite short lifetime, less than 30 min. Very similar materials, poly(vinylpyridine) polymers (PVP) incorporated with Ru(bipy)32 þ units, [Ru (bipy)2(PVP)10]2 þ have been synthesized and its ECL properties were studied using modified pyrolytic graphite,287–289 glassy carbon,290 and ITO163 electrodes. A quite interesting new class of polymers incorporating Ru(bipy)32 þ units has been synthesized by means of bipy ligand functionalized with polymerizable (CH2)5Si(OCH3)3 groups. The functionalized Ru(bipy)32 þ complex having two (CH2)5Si(OCH)3, on each bipy ligand was spin coated on the precleaned ITO surface from mixed iso-propanol-methanol solutions.291 The obtained thin film, after removing solvents’ traces, was cured to complete the gelation. This chemically immobilized gel films showed excellent stable cyclic voltammetry and ECL characteristics as well as long-term stability. In similar way, bipy ligand with two attached COONH(CH2)3Si(OC2H5)3 groups (Si-bipy) was applied in the fabrication of solidstate ECL sensor. The complex Ru(bipy)2(Si-bipy)2 þ was gelatinized and spin coated onto an ITO electrode to make modified ITO electrode with excellent ECL behavior.292,293 The above-listed examples show that many different immobilization procedures have been applied in fabrication of solid-state devices for ECL based sensing. Due to possible practical application one can expect, however, that this particular aspect of ECL will be object of further intensive studies and obtained results will allow construction of a still-better solid-state ECL detector. Especially promising seems to be the application of transition metal chelates other than mostly used Ru(bipy)32 þ that may lead to distinct improvement of this analytical tool. 16.3.2.3 Immunoassay Systems with ECL Active Labels Attaching of sensitive recognition groups to an ECL active kernel results in specific analytical tools based on ECL emission. The majority of ECL-based assays use a Ru(bipy)32 þ derivative as the electrochemiluminescent label induced to emit ECL at an oxidizing electrode in the presence of appropriate coreactant. Ru(bipy)32 þ molecule is served as a label for analysis in a manner analogous to radioactive or fluorescent labels by linking to biologically interesting molecules via a suitable group attached to the
510
ELECTROCHEMILUMINESCENT SYSTEMS AS DEVICES AND SENSORS
2,20 -bipyridine ligand.294–296 The Ru(bipy)32 þ ECL labels have several advantages over radioisotopes, such as easy disposal, long lifetime, sensitivity, and wide dynamic range. Due to high sensitivity and linear response of the ECL technique the Ru(bipy)32 þ ECL label can be straightforwardly determined at subpicomolar concentrations, along with an extremely wide dynamic range of greater than six orders of magnitude. Undoubtedly, the main interest in the application of ECL label is biochemical and clinical analysis with much attention devoted over the past decade to the determination of human biologically active species such as tumor and various pathologic conditions markers, hormones, proteins, enzymes, nucleic acid, and antibodies to name some more important of them. The more detailed listing of ECL-based assays developed for clinical analytes is beyond the scope of this chapter, and as such the reader is referred to the reviews34,36–40 and/or appropriate chapter in the Electrogenerated Chemiluminescence monograph32 edited by Bard dealing with this aspect. Here, only the main ideas of ECL labeling are shortly introduced to show the reader principles of this very significant methodology. Species labeled with Ru(bipy)32 þ tag are usually bound to a solid-phase support in proximity to an electrode because, due to the short lifetime of the electrogenerated species involved in the ECL process, only labels close to the electrode surface produce ECL. The ECL labels can be bound to the electrode itself or to magnetic particles that are collected on the surface of the electrode. In both cases, the most commonly used means of ECL detection is so-called sandwich type interaction.208,297–299 The sandwich approach employs two antibodies directed against the target molecule of interest. One of them is immobilized on a solid phase (electrode or magnetic particle surface), whereas the second one carries an ECL label. Binding of both antibodies to the analyzed moiety is resulting in the attachment of the ECL label to the solid-phase support (Fig. 16.8). Other approaches used in bioanalytical application of ECL involve (i) competition assay with immobilized antibody, (ii) competition assay with immobilized antigen, (iii) displacement assay with immobilized antibody, or (iv) displacement assay with immobilized antigen. Principally, all approaches already known from classical immunoassay with radioactive or fluorescent labels can be adopted for use with ECL label.
Figure 16.8 Schematic diagram illustrating the operating principles of ECL based immunoassay by means of the sandwich type interaction. (See the color version of this figure in Color Plates section.)
16.4
CONCLUSIONS AND REMARKS ABOUT POSSIBLE FUTURE DEVELOPMENT
511
Sandwich type recognition, the mostly applied probe–analyte bonding techniques, is used in many diagnostic tests in the medical and environmental fields including DNA characterization. Hybridization assay based on an immobilized single-strand DNA (ss-DNA) interacting with a labeled target ss-DNA297,298 or assay using integrase enzyme for bonding immobilized and free labeled double strand DNA (ds-DNA)300,301 can be used in similar way for DNA detection, quantification, or mismatch discrimination. Of particular note is the rapid development in such field in the last years resulted in the fabrication of a broad spectrum of the DNA biosensor and detectors based on the Ru(bipy)32 þ ECL tagging.245,292,293,299,302–306 To increase the ECL sensitivity, the multiple ECL labels technique has been developed. Instead single Ru(bipy)32 þ molecule the multi label technique apply the assembled set of ECL luminophores attached to probe used in molecular recognition. Representative examples involve Ru(bipy)32 þ chelate loaded in microsized polystyrene microspheres,297 silica nanoparticles,307 and carbon nanotubes308 or immobilized on gold nanoparticles.309 The recently published ECL detection of DNA using microchips307,310 open up new extremely attractive approach for further development in clinical analysis area. Although the field of clinical diagnostics is mostly benefited from the introduction of the highly sensitive ECL labeling method, there are many other examples of the use of electrochemiluminescence in the measurement and detection of biologically important compounds. In addition to detection of clinically important analytes, the ECL technique can be applied in life science research, water and food testing, and detection biological threat agents. The application of ECL to biological molecule detection in these and related fields will surely expand in the nearest future because the benefits of detection systems based on ECL have been already well recognized. It is especially true because ECL instrumentation has become commercially available (e.g., from Meso Scale Discovery or BioVeris companies) and adopted by users during last years.
16.4 CONCLUSIONS AND REMARKS ABOUT POSSIBLE FUTURE DEVELOPMENT Despite continuous progress in ECL investigations resulted in understanding the phenomena associated with it and finding applications, there is still place for intrinsic further development. There are still needs for new ECL systems and reactions as well as new applications to investigate and develop. Three fields of ECL investigations seem to be the most promising: (i) search for new efficient ECL systems together with studies of the phenomenon mechanism as such, (ii) use of the electrical energy for obtaining excited molecules and application of ECL systems in the light-emitting devices, and (iii) improvement of both sensitivity and specificity of ECL based analytical methods. The search of new ECL systems with high efficiencies and different emission colors seems to be crucial for all possible practical applications. It is, however, a difficult task because of the many factors affecting electrochemical excitation. The
512
ELECTROCHEMILUMINESCENT SYSTEMS AS DEVICES AND SENSORS
combined requirements of reductant and oxidant chemical stability and a lack of chemical complications following the initial electron transfer to and from the electrode still poses a problem. Further requirements such as high luminescence efficiency and photochemical stability of an excited state generated in ECL process, or chemical stability of an ECL luminophore in the presence of electrodes, electrolyte and solvent, additionally limit the number of the compounds suitable for ECL studies. Among many possible organic and inorganic compounds some transition metal chelates show an interesting combination of photophysical and electrochemical properties, which makes them potentially useful in ECL studies. In particular, the transition metal complexes with organic ligand(s) bonded to the metal core by the metal–carbon bond seem to be promising candidates for ECL investigations due to their extremely high luminescence yields. The quantitative studies of ECL systems are a still more difficult task. Only if all of the above-mentioned interferences are removed by the appropriate conditions of the experiment and if only simple electron transfer annihilation takes place during the ECL process without any competitive parasitic reactions, the obtained data (ECL efficiencies) allow for a more quantitative discussion of the electron transfer excitation. Extended Marcus model for ECL processes can be used for qualitative as well as quantitative descriptions of such kinds of electron transfer reactions. However, additional kinetic complications accompanying electron transfer excitation, such as diffusional limitations and presence of spin up-conversion processes, should also be taken into account in a more detailed discussion. It seems to be clear that, because of the large number of parameters going into the theoretical description, a quantitative understanding of the given ECL system may be rather difficult. This is also true for the relatively well understood ECL systems involving the most famous Ru(bipy)32 þ chelate. One can hope, however, that the more general and more quantitative interpretation than hitherto will be possible in view of new data. Further quantitative investigations (e.g., of temperature and solvent effects on ECL efficiencies) are necessary to give a decisive answer to remaining questions and doubts, especially concerning the adequacy of a simple ion-pair approximation for the ECL activated complex. Solid-state light-emitting electrochemical cells seem to an attractive alternative for organic-light emitting devices, but there are still many issues that need to be solved before development in light-emitting electrochemical cells based on polymer or solidstate systems could lead to light sources and displays. Generation of the electrochemically pumped laser action from miniaturized SSLEC cells seems to be principally possible but very challenging. Even though important progress has been achieved, the operating lifetime of SSLECs still needs to be improved to become competitive with the commercially available OLEDs. The same is true for SSLECs efficiencies. The search of new ECL luminophores can result in better performance in terms of efficiency and lifetime. Also, there is a need for a better understanding of the device behavior and characteristics including the clarification of the device degradation mechanism. One can also trust that further investigations will give details concerning similarities and differences in the electron transfer excitation between the solution and solid phases. Of course, comparative ECL studies in both liquid and
REFERENCES
513
solid media have to be performed for the same ECL luminophore(s) to clarify these issues. One can also expect that ECL phenomenon will be applied in further development of devices (scanning probe microscopes) for imaging and studying charge transport either in solution or in electroluminescent thin films with anticipated improvements in both spatial and temporal resolutions. ECL has also a promising future in analytical science. One can expect that further analytical applications will focus not just on Ru(bipy)32 þ and similar derivatives because use of ECL reagents other than Ru(bipy)32 þ seems to be another interesting and promising research objects. This nearly uncovered but very hopeful field of ECL investigations need unquestionably more attention because ECL systems emitting at wavelength different than characteristic for Ru(bipy)32 þ chelate offer a promising potential for parallel multianalyte detection. In this context one can mention the transition metal complexes as potential new ECL materials for studies in such direction. Cyclometallated Ir(III), Pt(II), or Os(II) chelates can be considered as possible candidates because their photophysical properties (emission color) are strongly dependent on the coordinated ligands nature allowing for precise tuning of their emission in required direction. On the other hand, their electrochemical and ECL properties are less affected by the central metal surroundings (see Section 16.2.3) making such complexes very suitable for further ECL studies, not only for analytical applications. Further progress of ECL probes immobilization methods should result in new robust, stable, reproducible ECL sensors. Especially, the use of electrochemiluminescent polymers may prove to be useful in this respect. There are also good prospects for ECL to be used as detection in miniaturized analytical systems particularly with a large increase in the applications of ECL immunoassay because high sensitivity, low detection limit, and good selectivity. One can believe that miniaturized biosensors based on ECL technology will induce a revolution in clinical analysis because of short analysis time, low consumption of reactants, and ease of automation. Summing up, ECL investigations seem to have a brilliant future because there are many places for the fundamental as well as applied research on this exciting border between electrochemistry and photochemistry. Exponential growth of published papers devoted to different aspects of ECL technique as observed in the past years seems to support the above conclusion.
REFERENCES 1. 2. 3. 4. 5. 6. 7.
D. M. Hercules, Science 1964, 145, 808. R. E. Visco, E. A. Chandross, J. Am. Chem. Soc. 1964, 86, 5350. K. S. V. Santhanam, A. J. Bard, J. Am. Chem. Soc. 1965, 87, 139. S. W. Feldberg, J. Am. Chem. Soc. 1966, 88, 390. S. W. Feldberg, J. Phys. Chem. 1966, 70, 3928. G. J. Hoytink, Discuss. Faraday Soc. 1968, 14. L. R. Faulkner, A. J. Bard, J. Am. Chem. Soc. 1969, 91, 209.
514 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.
ELECTROCHEMILUMINESCENT SYSTEMS AS DEVICES AND SENSORS
L. R. Faulkner, A. J. Bard, J. Am. Chem. Soc. 1969, 91, 6495. E. A. Chandross, J. W. Longworth, R. E. Visco, J. Am. Chem. Soc. 1965, 87, 3259. R. A. Marcus, J. Chem. Phys. 1965, 43, 2654. N. E. Tokel, A. J. Bard, J. Am. Chem. Soc. 1972, 94, 2862. W. L. Wallace, A. J. Bard, J. Phys. Chem. 1979, 83, 1350. I. Rubinstein, A. J. Bard, J. Am. Chem. Soc. 1981, 103, 512. I. Rubinstein, A. J. Bard, J. Am. Chem. Soc. 1981, 103, 5007. H. S. White, A. J. Bard, J. Am. Chem. Soc. 1982, 104, 6891. J. K. Leland, M. J. Powell, J. Electroanal. Soc. 1990, 137, 3127. D. Ege, W. G. Becker, A. J. Bard, Anal. Chem. 1984, 56, 2413. A. Kapturkiewicz, G. Angulo, Dalton Trans. 2003, 3907. Z. Ding, B. M. Quinn, S. K. Haram, L. E. Pell, B. A. Korgel, A. J. Bard, Science 2002, 296, 1293. A. J. Bard, Z. Ding, N. Myung, Struct Bond. 2005, 118, 1. M. M. Collinson, R. M. Wightman, Anal. Chem. 1993, 65, 2576. M. M. Collinson, R. M. Wightman, P. Pastore, J. Phys. Chem. 1994, 98, 11942. H. D. Abrun˜a, A. J. Bard, J. Am. Chem. Soc. 1982, 104, 2641. F. R. Fan, A. Mau, A. J. Bard, Chem. Phys. Lett. 1985, 116, 400. T. Horiuchi, O. Niwa, N. Hatekenaka, Nature 1998, 394, 659. A. Zweig, Adv. Photochem. 1969, 6, 425. D. M. Hercules, Chem. Res. 1969, 2, 301. E. A. Chandross, Trans. NY Acad. Sci. 2 1969, 31, 571. L. R. Faulkner, Int. Rev. Sci. Phys. Chem. 2 1976, 9, 213. L. R. Faulkner, A. J. Bard, Electroanal. Chem. 1976, 10, 1. S. M. Park, D. A. Tryk, Rev. Chem. Intermed. 1981, 4, 43. A. J. Bard (Ed.), Electrogenerated Chemiluminescence, Marcel Dekker, New York, 2004. X. B. Yin, S. J. Dong, E. Wang, Trends Anal. Chem. 2004, 23, 432. M. M. Richter, Chem. Rev. 2004, 104, 3003. H. Wei, E. K. Wang, Trends Anal. Chem. 2008, 27, 447. R. Pyati, M. M. Richter, Annu. Rep. Prog. Chem. Sect. C 2007, 103, 12. W. Miao,in C. G. Zoski (Ed.), Handbook of Electrochemistry, Elsevier, Amsterdam, 2007, p. 541. W. Miao, Chem. Rev. 2008, 108, 2506. C. A. Marquette, L. J. Blum, Anal. Bioanal. Chem. 2008, 390, 155. M. M. Richter,in F.S. Ligler, C. R. Taitt (Eds.), Optical Biosensors, 2nd edition, Elsevier, Amsterdam, 2008, p. 317. N. N. Rozhitskii, A. I. Bykh, M. E. Krasnogolovetz, Electrochemical Luminescence Kharkiv Technical University of Radio Electronics Kharkiv 2000. S. M. Park, A. J. Bard, J. Am. Chem. Soc. 1975, 97, 2978. D. J. Freed, L. R. Faulkner, J. Am. Chem. Soc. 1971, 93, 3565. D. J. Freed, L. R. Faulkner, J. Am. Chem. Soc. 1972, 94, 4790.
REFERENCES
45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55. 56. 57. 58. 59. 60. 61. 62. 63. 64. 65. 66. 67. 68. 69. 70. 71. 72. 73. 74. 75. 76. 77. 78. 79. 80.
515
S. M. Park, A. J. Bard, Chem. Phys. Lett. 1976, 38, 2571. A. Kapturkiewicz, Adv. Electrochem. Sci. Eng. 1997, 5, 1. E. L. Ritchie, P. Pastore, R. M. Wightman, J. Am. Chem. Soc. 1997, 119, 11920. A. F. Slaterbeck, T. D. Meehan, E. M. Gross, R. M. Wightman, J. Phys. Chem. B 2002, 106, 6088. A. Weller, K. Zachariasse, J. Chem. Phys. 1967, 46, 4984. A. Weller, K. Zachariasse, Chem. Phys. Lett. 1971, 10, 197. A. Weller, K. Zachariasse, Chem. Phys. Lett. 1971, 10, 590. M. Oyama, S. Okazaki, Anal. Chem. 1998, 70, 5079. M. Oyama, M. Mitani, S. Okazaki, Electrochem. Commun. 2000, 2, 363. D. Laser, A. J. Bard, J. Electrochem. Soc. 1975, 122, 632. G. H. Brilmyer, A. J. Bard, J. Electrochem. Soc. 1980, 127, 104. H. Schaper, H. Kostlin, E. Schnedler, J. Electrochem. Soc. 1982, 129, 1289. J. T. Maloy, K. B. Prater, A. J. Bard, J. Am. Chem. Soc. 1971, 93, 5959. N. E. Tokel-Takvoryan, R. E. Hemingway, A. J. Bard, J. Am. Chem. Soc. 1973, 95, 6582. J. E. Bartelt, S. M. Drew, R. M. Wightman, J. Electrochem. Soc. 1992, 139, 70. L. R. Faulkner, J. Electrochem. Soc. 1975, 122, 1190. L. R. Faulkner, J. Electrochem. Soc. 1977, 124, 1724. P. R. Michael, R. F. Faulkner, J. Am. Chem. Soc. 1977, 99, 7754. R. Bezmann, L. R. Faulkner, J. Am. Chem. Soc. 1972, 94, 6331. J. Kim, L. R. Faulkner, J. Electroanal. Chem. 1988, 242, 107. J. Kim, L. R. Faulkner, J. Am. Chem. Soc. 1988, 110, 112. A. Kapturkiewicz, J. Electroanal. Chem. 1990, 290, 135. P. Pastore, D. Badocco, F. Zanon, Electrochim. Acta 2006, 51, 5394. D. Badocco, F. Zanon, P. Pastore, Electrochim. Acta 2006, 51, 6442. M. M. Chang, T. Saji, A. J. Bard, J. Am. Chem. Soc. 1977, 99, 5399. P. A. Mabrouk, M. S. Wrighton, Inorg. Chem. 1986, 25, 526. H. E. Toma, P. R. Auburn, E. S. Dodsworth, M. N. Golorin, A. B. P. Lever, Inorg. Chem. 1987, 26, 4257. S. J. Park, D. H. Kim, D. H. Kim, H. J. Park, D. N. Lee, B. H. Kim, W. Y. Lee, Anal. Sci. 2001, 17, a93. M. Zhou, G. P. Robertson, J. Roovers, Inorg. Chem. 2005, 44, 8317. F. Bolletta, M. Vitale, Inorg. Chim. Acta 1990, 175, 127. B. H. Kim, D. N. Lee, H. J. Park, J. H. Min, Y. M. Jun, S. J. Park, W. Y. Lee, Talanta 2004, 62, 595. J. Gonzalez-Velasco, I. Rubinstein, R. J. Crutchley, A. B. P. Lever, A. J. Bard, Inorg. Chem. 1983, 22, 822. J. Gonzalez-Velasco, J. Phys. Chem. 1988, 92, 2202. P. McCord, A. J. Bard, J. Electroanal. Chem. 1991, 318, 91. A. Kapturkiewicz, Chem. Phys. Lett. 1995, 236, 389. D. N. Lee, J. K. Kim, H. S. Park, Y. M. Jun, R. Y. Hwang, W. Y. Lee, B. H. Kim, Synth. Met. 2005, 150, 93.
516
ELECTROCHEMILUMINESCENT SYSTEMS AS DEVICES AND SENSORS
81. D. N. Lee, H. S. Park, E. H. Kim, Y. M. Jun, J. Y. Lee, W. Y. Lee, B. H. Kim, Bull. Korean Chem. Soc. 2006, 27, 99. 82. D. N. Lee, B. K. Soh, S. H. Kim, Y. M. Jun, S. H. Yoon, W. Y. Lee, B. H. Kim, J. Organometal. Chem. 2008, 693, 655. 83. M. Zhou, J. Roovers, J. Macromol. 2001, 34, 244. 84. M. M. Richter, A. J. Bard, W. Kim, R. S. Schmehl, Anal. Chem. 1988, 70, 310. 85. S. Stagni, A. Palazzi, S. Zacchini, B. Ballarin, C. Bruno, M. Marcaccio, F. Paolucci, M. Monari, M. Carano, A. J. Bard, Inorg. Chem. 2006, 45, 695. 86. S. Zanarini, A. J. Bard, M. Marcaccio, A. Palazzi, F. Paolucci, S. Stagni, J. Phys. Chem. B 2006, 110, 22551. 87. M. Li, J. Liu, C. Zhao, L. Sun, J. Organometal. Chem. 2006, 691, 4189. 88. M. Li, J. Liu, L. Sun, J. Pan, C. Zhao, J. Organometal. Chem. 2008, 693, 46. 89. H. D. Abrun˜a, J. Electroanal. Chem. 1984, 175, 321. 90. H. D. Abrun˜a, J. Electrochem. Soc. 1985, 132, 842. 91. C. W. Lee, J. Ouyang, A. J. Bard, J. Electroanal. Chem. 1988, 244, 319. 92. K. A. King, P. J. Spellane, R. J. Watts, J. Am. Chem. Soc. 1985, 107, 1431. 93. I. M. Dixon, J. P. Collin, J. P. Sauvage, L. Flamigni, S. Encinas, F. Barigelletti, Chem. Soc. Rev. 2000, 29, 385. 94. M. A. Baldo, S. Lamansky, P. E. Burrows, M. E. Thompson, S. R. Forrest, Appl. Phys. Lett. 1999, 75, 4. 95. K. Nishimura, Y. Hamada, T. Tsujioka, K. Shibata, T. Fuyuki, Jpn. J. Appl. Phys. 2001, 40, L945. 96. E. M. Gross, N. R. Armstrong, R. M. Wightman, J. Electrochem. Soc. 2002, 149, E137. 97. D. Bruce, M. M. Richter, Anal. Chem. 2002, 74, 1340. 98. A. Kapturkiewicz, T. M. Chen, I. R. Laskar, J. Nowacki, Electrochem. Commun. 2004, 6, 827. 99. A. Kapturkiewicz, J. Nowacki, P. Borowicz, Electrochim. Acta 2005, 50, 3395. 100. A. Kapturkiewicz, J. Nowacki, P. Borowicz, Z. Phys. Chem. 2006, 220, 525. 101. L. S. Shin, J. I. Kim, T. H. Kwon, J. I. Hong, J. K. Lee, H. Kim, J. Phys. Chem. C 2007, 111, 2280. 102. J. I. Kim, I. S. Shin, H. Kim, J. K. Lee, J. Am. Chem. Soc. 2005, 127, 1614. 103. B. D. Muegge, M. M. Richter, Anal. Chem. 2004, 76, 73. 104. C. Cole, B. D. Muegge, M. M. Richter, Anal. Chem. 2003, 75, 601. 105. D. G. Nocera, H. B. Gray, J. Am. Chem. Soc. 1984, 106, 824. 106. R. D. Mussell, D. G. Nocera, Polyhedron 1986, 5, 47. 107. R. D. Mussell, D. G. Nocera, J. Am. Chem. Soc. 1988, 110, 2764. 108. R. D. Mussell, D. G. Nocera, Inorg. Chem. 1990, 29, 3711. 109. R. D. Mussell, D. G. Nocera, J. Phys. Chem. 1991, 95, 6919. 110. P. Szrebowaty, A. Kapturkiewicz, Chem. Phys. Lett. 2000, 328, 160. 111. A. Kapturkiewicz, P. Szrebowaty, J. Chem. Soc., Dalton Trans. 2002, 3219. 112. J. Perichon,in A. J. Bard, H. Lunds (Eds.), Encyclopaedia of the Electrochemistry of the Elements, Vol. 11, Marcel Dekker, New York, 1978, p. 71. 113. J. Saltiel, B. W. Atwater, Adv. Photochem. 1988, 14, 1.
REFERENCES
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. 148. 149. 150. 151. 152.
517
F. E. Beldeman, D. M. Hercules, J. Phys. Chem. 1979, 83, 2203. K. M. Maness, R. M. Wightman, J. Electroanal. Chem. 1994, 396, 85. K. M. Maness, J. E. Bartelt, R. M. Wightman, J. Phys. Chem. 1994, 98, 3993. H. Tachikawa, A. J. Bard, Chem. Phys. Lett. 1974, 26, 246. N. Periasamy, S. J. Shah, K. S. V. Santhanam, J. Chem. Phys. 1973, 58, 821. N. Periasamy, K. S. V. Santhanam, Proc. Indian Acad. Sci. Sect. A 1974, 80, 194. R. S. Glass, L. R. Faulkner, J. Phys. Chem. 1982, 86, 1652. W. G. Herkstroeter, P. B. Merkel, J. Photochem. 1981, 16, 331. A. P. Darmanyan, V. A. Kuzmin, Dokl. Akad. Nauk SSSR 1982, 260, 1167. S. J. Strickler, R. A. Berg, J. Chem. Phys. 1962, 37, 814. A. Kapturkiewicz, J. Electroanal. Chem. 1994, 372, 101. E. F. Fabrizio, A. Payne, N. E. Westlund, A. J. Bard, P. P. Magnus, J. Phys. Chem. A 2002, 106, 1961. S. M. Park, M. T. Paffett, G. H. Daub, J. Am. Chem. Soc. 1977, 99, 5393. H. A. Sharifian, S. M. Park, Photochem. Photobiol. 1982, 36, 83. K. Itaya, S. Toshima, Chem. Phys. Lett. 1977, 51, 447. M. Kawai, K. Itaya, S. Toshima, J. Phys. Chem. 1980, 84, 2368. A. Kapturkiewicz, Z. R. Grabowski, J. Jasny, J. Electroanal. Chem. 1990, 279, 55. A. Kapturkiewicz, J. Electroanal. Chem. 1991, 302, 131. A. Kapturkiewicz, Z. Phys. Chem. NF. 1991, 170, 87. A. Kapturkiewicz, Chem. Phys. 1992, 166, 259. A. Zweig, A. H. Maurer, B. G. Roberts, J. Org. Chem. 1967, 32, 1322. K. Itaya, A. J. Bard, M. Schwarc, Z. Phys. Chem. NF. 1978, 112, 1. A. Kapturkiewicz, J. Electroanal. Chem. 1993, 348, 283. Z. R. Grabowski, K. Rotkiewicz, W. Rettig, Chem. Rev. 2003, 103, 3899. A. Kapturkiewicz, J. Herbich, J. Nowacki, Chem. Phys. Lett. 1997, 275, 355. A. Elangovan, H. H. Chiu, S. W. Yang, T. I. Ho, Org. Biomol. Chem. 2004, 2, 3113. A. Elangovan, K. M. Kao, S. W. Yang, Y. L. Chen, T. L. Ho, Y. O. Su, J. Org. Chem. 2005, 70, 4460. A. Elangovan, T. Y. Chen, C. Y. Chen, T. I. Ho, Chem. Commun. 2003, 2146. A. Elangovan, J. H. Lin, S. W. Yang, H. Y. Hsu, T. I. Ho, J. Org. Chem. 2004, 69, 8086. A. Elangovan, S. W. Yang, J. H. Lin, K. M. Kao, T. I. Ho, Org. Biomol. Chem. 2004, 2, 1597. A. Elangovan, J. H. Lin, S. W. Yang, H. Y. Hsu, T. I. Ho, J. Org. Chem. 2005, 70, 1104. C. Y. Chen, J. H. Ho, S. L. Wang, T. I. Ho, Photochem. Photobiol. Sci. 2003, 2, 1232. X. Yang, X. Jiang, C. Zhao, R. Chen, P. Qina, L. Suna, Tetrahedron Lett. 2006, 47, 4961. X. Jiang, X. Yang, C. Zhao, K. Jin, L. Sun, J. Phys. Chem. C 2007, 111, 9595. R. Y. Lai, M. Chiba, N. Kitamura, A. J. Bard, Anal. Chem. 2003, 74, 551. R. Y. Lai, X. Kong, S. A. Jenekhe, A. J. Bard, J. Am. Chem. Soc. 2003, 125, 12631. K. T. Wong, T. H. Hung, T. C. Chao, T. I. Ho, Tetrahedron Lett. 2005, 46, 855. J. B. Noffsinger, N. D. Danielson, Anal. Chem. 1987, 59, 865. J. B. Noffsinger, N. D. Danielson, J. Chromatogr. 1987, 387, 520.
518 153. 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. 180. 181. 182. 183. 184. 185. 186. 187. 188. 189. 190.
ELECTROCHEMILUMINESCENT SYSTEMS AS DEVICES AND SENSORS
W. Miao, J. P. Choi, A. J. Bard, J. Am. Chem. Soc. 2002, 124, 14478. Y. Zu, A. J. Bard, Anal. Chem. 2000, 72, 3223. A. Arora, A. J. de Mello, A. Manz, Anal. Commun. 1997, 34, 393. T. M. Downey, T. A. Nieman, Anal. Chem. 1992, 64, 261. A. W. Knight, G. M. Greenway, Analyst 1996, 121, 101R. X. Liu, L. Shi, W. Niu, H. Li, G. Xu, Angew. Chem., Int. Ed. 2007, 46, 421. X. Chen, M. Sato, Y. Lin, Microchem. J. 1998, 58, 13. J. Butler, A. Henglein, Radiat. Phys. Chem. 1980, 15, 603. F. Kanoufi, A. J. Bard, J. Phys. Chem. B 1999, 103, 10469. F. Li, H. Cui, X. Q. Lin, Luminescence 2002, 17, 117. M. C. Lu, C. W. Whang, Anal. Chim. Acta 2004, 522, 25. X. Chen, W. Chen, Y. Jiang, L. Jia, X. Wang, Microchem. J. 1998, 59, 427. X. Q. Lin, F. Li, Y. Q. Pang, H. Cui, Anal. Bioanal. Chem. 2004, 378, 2028. M. Zorzi, P. Pastore, F. Magno, Anal. Chem. 2000, 72, 4934. S. N. Brune, D. R. Bobbitt, Talanta 1991, 38, 419. F. Bolletta, M. Ciano, V. Balzani, N. Serpone, Inorg. Chim. Acta 1982, 62, 207. R. Memming, J. Electrochem. Soc. 1969, 116, 785. G. Xu, S. J. Dong, Electroanalysis 2000, 12, 583. S. Yamazaki-Nishida, Y. Harima, K. Yamashita, J. Electroanal. Chem. 1990, 283, 455. K. Yamashita, S. Yamazaki-Nishida, Y. Harima, A. Segawa, Anal. Chem. 1991, 63, 872. H. Y. Wang, G. Xu, S. J. Dong, Microchem. J. 2002, 72, 43. W. G. Becker, H. S. Seung, A. J. Bard, J. Electroanal. Chem. 1984, 167, 127. E. A. Chandross, F. I. Sonntag, J. Am. Chem. Soc. 1966, 88, 1089. H. Schaper, E. Schnedler, J. Phys. Chem. 1982, 86, 4380. H. Schaper, E. Schnedler, J. Electroanal. Chem. 1982, 137, 39. K. Nishimura, Y. Hamada, T. Tsujioka, S. Matsuta, K. Shibata, T. Fuyuki, Jpn. J. Appl. Phys. 2001, 40, L1323. H. K€ostlin, H. Schaper, Phys. Lett. 1980, 76A, 455. M. Orlik, J. Rosenmund, K. Doblhofer, G. Ertl, J. Phys. Chem. B 1998, 102, 1397. M. Orlik, K. Doblhofer, G. Ertl, J. Phys. Chem. B 1998, 102, 6367. M. Orlik, J. Phys. Chem. B 1999, 103, 6629. D. Dini, K. Doblhofer, G. Ertl, Phys. Chem. Chem. Phys. 2000, 2, 1183. S. Okamoto, K. Soeda, T. Iyoda, T. Kato, T. Kado, S. Hayase, J. Electrochem. Soc. 2005, 152, A1677. K. Ide, M. Fujimoto, T. Kado, S. Hayase, J. Electrochem. Soc. 2007, 155, B645. J. C. deMello, N. Tessler, S. C. Graham, X. Li, A. B. Holmes, R. H. Friend, Synth. Met. 1997, 85, 1277. J. D. Slinker, J. Rivnay, J. S. Moskowitz, J. B. Parker, S. Bernhard, H. D. Abrun˜a, G. G. Malliaras, J. Mater. Chem. 2007, 17, 2976. E. S. Handy, A. J. Pal, M. F. Rubner, J. Am. Chem. Soc. 1999, 121, 3525. F. G. Gao, A. J. Bard, J. Am. Chem. Soc. 2000, 122, 7426. F. G. Gao, A. J. Bard, Chem. Mater. 2002, 14, 3465.
REFERENCES
191. 192. 193. 194. 195. 196. 197. 198. 199. 200. 201. 202. 203. 204. 205. 206. 207. 208. 209. 210. 211. 212. 213. 214. 215. 216. 217. 218. 219. 220. 221. 222. 223. 224.
519
H. Rudmann, M. F. Rubner, J. Appl. Phys. 2001, 90, 4338. H. Rudmann, S. Shimada, M. F. Rubner, J. Am. Chem. Soc. 2002, 124, 4918. M. Buda, G. Kalyuzhny, A. J. Bard, J. Am. Chem. Soc. 2002, 124, 6090. J. D. Slinker, A. A. Gorodetsky, M. S. Lowry, J. Wang, S. Parker, R. Rohl, S. Bernhard, G. G. Malliaris, J. Am. Chem. Soc. 2004, 126, 2763. Y. Chuai, D. N. Lee, C. Zhena, J. H. Minb, B. H. Kimb, D. Zou, Synth. Met. 2004, 145, 259. H. J. Bolink, L. Cappelli, E. Coronado, P. Gavin˜a, Inorg. Chem. 2005, 44, 5966. H. J. Bolink, L. Cappelli, E. Coronado, M. Gr€atzel, E. Orti, R. D. Costa, P. M. Viruela, M. K. Nazeeruddin, J. Am. Chem. Soc. 2006, 128, 14786. H. J. Bolink, L. Cappelli, E. Coronado, A. Parham, P. St€ ossel, Chem. Mater. 2006, 18, 2778. C. Liu, A. J. Bard, J. Am. Chem. Soc. 2002, 124, 4190. W. L. Jia, Y. F. Hu, J. Gao, S. Wang, Dalton Trans 2006, 1721. S. Bernhard, X. Gao, G. G. Malliaras, H. D. Abrun˜a, Adv. Mater. 2002, 14, 433. X. Gong, K. Ng Po, W. K. Chan, Adv. Mater. 1998, 10, 1337. W. K. Chan, K. Ng Po, X. Gong, S. Hou, Appl. Phys. Lett. 1999, 75, 3920. M. K. Nazeeruddin, R. T. Wegh, Z. Zhou, C. Klein, Q. Wang, F. De Angelis, S. Fantacci, M. Gr€atzel, Inorg. Chem. 2006, 45, 9245. S. J. Liu, Q. Zhao, Q. L. Fan, W. Huang, Eur. J. Inorg. Chem. 2008, 2177. L. He, L. Duan, J. Qiao, R. Wang, P. Wei, L. Wang, Y. Qiu, Adv. Funct. Mater. 2008, 18, 2123. C. H. Lyons, E. D. Abbas, J. K. Lee, M. F. Rubner, J. Am. Chem. Soc. 1998, 120, 12100. J. K. Lee, D. Yoo, M. F. Rubner, Chem. Mater. 1997, 9, 1710. C. M. Elliott, F. Pichot, C. J. Bloom, L. S. Rider, J. Am. Chem. Soc. 1998, 120, 6781. A. Wu, J. K. Lee, M. F. Rubner, Thin Solid Films 1998, 327–329, 663. A. Wu, D. Yoo, J. K. Lee, M. F. Rubner, J. Am. Chem. Soc. 1999, 121, 4883. S. Welter, K. Brunner, J. W. Hofstraat, L. De Cola, Nature 2003, 421, 54. C. Zhen, Y. Chuai, C. Lao, L. Huang, D. Zou, Appl. Phys. Lett. 2005, 87, 093508. R. M. Measures, Appl. Opt. 1974, 13, 1121. R. M. Measures, Appl. Opt. 1975, 14, 909. C. P. Keszthelyi, Appl. Opt. 1975, 14, 1710. C. A. Heller, J. L. Jernigan, Appl. Opt. 1977, 16, 61. R. M. Wightman, C. L. Curtis, P. A. Flowers, R. G. Maus, E. M. McDonald, J. Phys. Chem. B 1998, 102, 9991. R. C. Engstrom, K. W. Johnson, S. Desjarlais, Anal. Chem. 1987, 59, 670. R. C. Engstrom, C. M. Pharr, M. D. Koppang, J. Electroanal. Chem. 1987, 221, 251. S. Simison, A. Pellicano, B. Mathias, D. J. Schiffrin, J. Electroanal. Chem. 1999, 470, 89. P. Hopper, W. G. Kuhr, Anal. Chem. 1994, 66, 1966. S. Szunerits, J. M. Tam, L. Thouin, C. Amatore, D. R. Walt, Anal. Chem. 2003, 75, 4382. C. Amatore, C. Pebay, L. Servant, N. Sojic, S. Szunerits, L. Thouin, Chem. Phys. Chem. 2006, 7, 1322.
520 225. 226. 227. 228. 229. 230. 231. 232. 233. 234. 235. 236. 237. 238. 239. 240. 241. 242. 243. 244. 245. 246. 247. 248. 249. 250. 251. 252. 253. 254. 255. 256. 257. 258. 259. 260. 261. 262. 263.
ELECTROCHEMILUMINESCENT SYSTEMS AS DEVICES AND SENSORS
F. F. R. Fan, D. Cliffel, A. J. Bard, Anal. Chem. 1998, 70, 2941. Y. Zu, Z. Ding, J. Zhou, Y. Lee, A. J. Bard, Anal. Chem. 2001, 73, 2153. R. G. Maus, R. M. Wightman, Anal. Chem. 2001, 73, 3993. K. R. Kneten, R. L. McCreery, Anal. Chem. 1992, 64, 2518. L. S. Kuhn, A. Weber, S. G. Weber, Anal. Chem. 1990, 62, 1631. N. Egashira, H. Kumasako, K. Ohga, Anal. Sci. 1990, 6, 903. Y. Lee, A. J. Bard, Anal. Chem. 2002, 74, 3626. Y. Lee, Z. Ding, A. J. Bard, Anal. Chem. 2002, 74, 3634. H. Xiong, J. Guo, K. Kurihara, S. Amemiya, Electrochem. Commun. 2004, 6, 615. E. S. Jin, B. J. Norris, P. Pantano, Electroanalysis 2001, 13, 1287. S. Szunerits, D. R. Walt, Anal. Chem. 2002, 74, 1718. S. Szunerits, D. R. Walt, Chem. Phys. Chem. 2003, 4, 186. A. Chovin, P. Garrigue, N. Sojic, Electrochim. Acta 2004, 49, 3751. A. Chovin, P. Garrigue, P. Vinatier, N. Sojic, Anal. Chem. 2004, 76, 357. A. Chovin, A. Garrigue, G. Pecastaings, H. Saadaoui, N. Sojic, Meas. Sci. Technol. 2006, 17, 1211. W. Y. Lee, Mikrochim. Acta 1997, 127, 19. R. D. Gerardi, N. W. Barnett, S. W. Lewis, Anal. Chim. Acta 1999, 378, 1. A. W. Knight, Trends Anal. Chem. 1999, 18, 47. K. A. F€ahnrich, M. Pravda, G. G. Guilbault, Talanta 2001, 54, 531. B. A. Gorman, P. S. Francis, N. W. Barnett, Analyst 2006, 131, 616. H. Wei, E. K. Wang, Chem. Lett. 2007, 36, 210. W. K. Nonidez, D. E. Leyden, Anal. Chim. Acta 1978, 96, 401. S. N. Brune, D. R. Bobbitt, Anal. Chem. 1992, 64, 166. W. A. Jackson, D. R. Bobbitt, Anal. Chim. Acta 1994, 285, 309. A. W. Knight, G. M. Greenway, E. D. Chesmore, Anal. Proc. 1995, 32, 125. L. L. Shultz, J. S. Stoyanoff, T. A. Nieman, Anal. Chem. 1996, 68, 349. G. C. Fiaccabrino, N. F. de Rooij, M. Koudelka-Hep, Anal. Chim. Acta 1998, 359, 263. W. Zhan, J. Alvarez, R. M. Crooks, J. Am. Chem. Soc. 2002, 124, 13265. W. Zhan, J. Alvarez, R. M. Crooks, Anal. Chem. 2003, 75, 313. H. Qiu, J. Yan, X. Sun, J. Liu, W. Cao, X. Yang, E. Wang, Anal. Chem. 2003, 75, 5435. J. Yan, X. Xiurong Yang, E. Wang, Anal. Bioanal. Chem. 2005, 381, 48. B. D. Muegge, M. M. Richter, Anal. Chem. 2002, 74, 547. D. Bruce, M. M. Richter, Analyst 2002, 127, 1492. X. Zhang, A. J. Bard, J. Phys. Chem. 1988, 92, 5566. Y. S. Obeng, A. J. Bard, Langmuir 1991, 7, 191. M. Buda, F. G. Gao, A. J. Bard, J. Solid. State Electrochem. 2004, 8, 706. N. W. Barnett, R. Bos, H. Brand, P. Jones, K. F. Lim, S. D. Purcell, R. A. Russell, Analyst 2002, 127, 455. G. M. Greenway, A. Greenwood, P. Watts, C. Wiles, Chem. Commun. 2006, 85. L. Bi, H. Wang, Y. Shen, E. Wang, S. J. Dong, Electrochem. Commun. 2003, 5, 913.
REFERENCES
264. 265. 266. 267. 268. 269. 270. 271. 272. 273. 274. 275. 276. 277. 278. 279. 280. 281. 282. 283. 284. 285. 286. 287. 288. 289. 290. 291. 292. 293. 294.
295. 296. 297. 298.
521
Z. H. Guo, Y. Shen, M. K. Wang, F. Zhao, S. J. Dong Anal. Chem. 2004, 76, 184. X. Sun, Y. Du, L. Zhang, S. J. Dong, E. Wang, Anal. Chem. 2007, 79, 2588. I. Rubinstein, A. J. Bard, J. Am. Chem. Soc. 1980, 102, 6641. Z. H. Guo, S. J. Dong, Anal. Chem. 2004, 76, 2683. Z. H. Guo, S. J. Dong, Electroanalysis 2005, 17, 607. H. N. Choi, S. H. Cho, W. Y. Lee, Anal. Chem. 2003, 75, 4250. H. N. Choi, S. H. Cho, Y. J. Park, D. W. Lee, W. Y. Lee, Anal. Chim. Acta 2005, 541, 49. H. N. Choi, J. Y. Lee, Y. K. Lyu, W. Y. Lee, Anal. Chim. Acta 2006, 565, 48. H. N. Choi, Y. K. Lyu, W. Y. Lee, Electroanalysis 2006, 18, 275. S. N. Ding, J. J. Xu, H. Y. Chen, Electrophoresis 2005, 26, 1737. S. N. Ding, J. J. Xu, W. J. Zhang, H. Y. Chen, Talanta 2006, 70, 572. Y. F. Zhuang, H. X. Ju, Anal. Lett. 2005, 38, 2077. H. J. Song, Z. J. Zhang, F. Wang, Electroanalysis 2006, 18, 1838. H. Y. Wang, G. B. Xu, S. J. Dong, Anal. Chim. Acta 2003, 480, 285. Y. Du, B. Qi, X. R. Yang, E. K. Wang, J. Phys. Chem. B 2006, 110, 21662. L. H. Zhang, Z. H. Guo, Z. A. Xu, S. J. Dong, J. Electroanal. Chem. 2006, 592, 63. M. Sykora, T. J. Meyer, Chem. Mater. 1999, 11, 1186. M. M. Collinson, B. Novak, S. A. Martin, J. S. Taussig, Anal. Chem. 2000, 72, 2914. L. Armelao, R. Bertoncello, S. Gross, D. Badocco, P. Pastore, Electroanalysis 2003, 15, 803. Z. H. Guo, Y. Shen, F. Zhao, M. K. Wang, S. J. Dong, Analyst 2004, 129, 657. Y. Tao, Z. J. Lin, X. M. Chen, X. Chen, X. R. Wang, Anal. Chim. Acta 2007, 594, 169. H. D. Abrun˜a, P. Denisevich, M. Uman˜a, T. J. Meyer, R. W. Murray, J. Am. Chem. Soc. 1981, 103, 1. R. B. Lowry, C. E. Williams, J. Braven, Talanta 2004, 63, 961. L. Dennany, R. J. Forster, J. F. Rusling, J. Am. Chem. Soc. 2003, 125, 5213. L. Dennany, R. J. Forster, B. White, M. Smyth, J. F. Rusling, J. Am. Chem. Soc. 2004, 126, 8835. L. Dennany, C. F. Hogan, T. E. Keyes, R. J. Forster, Anal. Chem. 2006, 78, 1412. R. J. Forster, C. F. Hogan, Anal. Chem. 2000, 72, 5576. J. K. Lee, S. H. Lee, M. Kim, H. Kim, D. H. Kim, W. Y. Lee, Chem. Commun. 2003, 1602. H. Wei, Y. Du, J. Z. Kang, G. B. Xu, E. K. Wang, Chin. J. Chem. 2007, 25, 159. H. Wei, Y. Du, J. Z. Kang, E. K. Wang, Electrochem. Commun. 2007, 9, 1474. G. F. Blackburn, H. P. Shah, J. H. Kenten, J. Leland, R. A. Kamin, J. Link, J. Peterman, M. J. Powell, A. Shah, D. B. Talley, S. K. Tyagi, E. Wilkins, T. G. Wu, R. J. Massey, Clin. Chem. 1991, 37, 1534. J. H. Kenten, J. Casadel, J. Link, S. Lupold, J. J. Wiley, M. Powell, A. Rees, R. Massey, Clin. Chem. 1991, 37, 1626. J. H. Kenten, S. Gudibande, J. Link, J. J. Wiley, B. Curfman, E. O. Major, R. J. Massey, Clin. Chem. 1992, 38, 873. W. Miao, A. J. Bard, Anal. Chem. 2004, 76, 5379. W. Miao, A. J. Bard, Anal. Chem. 2004, 76, 7109.
522 299. 300. 301. 302. 303. 304. 305. 306. 307. 308. 309. 310.
ELECTROCHEMILUMINESCENT SYSTEMS AS DEVICES AND SENSORS
J. Zhang, H. L. Qi, Y. Li, J. Yang, Q. Gao, C. Zhang, Anal. Chem. 2008, 80, 2888. W. Y. Lo, A. J. Baeumner, Anal. Chem. 2007, 79, 1386. W. Y. Lo, A. J. Baeumner, Anal. Chem. 2007, 79, 1548. Y. T. Hsueh, R. L. Smith, M. A. Northrup, Sens. Actuators B 1996, 33, 110. Y. T. Hsueh, S. D. Collins, R. L. Smith, Sens. Actuators B 1998, 49, 1. D. Zhu, D. Xing, X. Shen, J. Liu, Q. Chen, Biosens. Bioelectron. 2004, 20, 448. C. Bertolino, M. MacSweeney, J. Tobin, B. O’Neill, M. M. Sheehan, S. Coluccia, H. Berney, Biosens. Bioelectron. 2005, 21, 565. X. Zhou, D. Xing, D. Zhu, L. Jia, Electrochem. Commun. 2008, 10, 564. Z. Chang, J. Zhou, K. Zhao, N. Zhu, P. He, Y. Fang, Electrochim. Acta 2006, 52, 575. Y. Li, H. Qi, F. Fang, C. Zhang, Talanta 2007, 72, 1704. H. Wang, C. Zhang, Y. Li, H. Qi, Anal. Chim. Acta 2006, 575, 205. A. M. Spehar-Deleze, L. Schmidt, R. Neier, S. Kulmala, N. de Rooij, M. Koudelka-Hep, Biosens. Bioelectron. 2006, 22, 722.
CHAPTER 17
Recent Developments in the Design of Dye-Sensitized Solar Cell Components STEFANO CARAMORI and CARLO ALBERTO BIGNOZZI Dipartimento di Chimica, Universita degli Studi di Ferrara, Ferrara, Italy
17.1
INTRODUCTION
Addressing the global energy issue according to the needs of the world countries defines the major science and engineering challenge of this century. Solutions will require large mobilization and investments in research and development in a number of energy-related technologies. World demand for energy is expected to more than double by 2050 and to even triple by the end of the century.1 Finding sufficient supplies of clean energy for the future is one of the most important challenges in our society. Sunlight provides by far the largest of all carbon-neutral energy sources. More energy is provided by sunlight striking the Earth in 1 h (4.3 1020 J) than the overall energy consumption on the planet for 1 year (4.1 1020 J). Yet solar electricity provides only approximately one millionth of the total electricity supply, and renewable biomass provides less than 0.1% of the total energy consumed. The huge gap between our present use of solar energy and its enormous undeveloped potential defines a grand challenge in energy research. The European Union roadmap on renewable energies assumes a 38% contribution from renewable sources to the total amount of electricity produced in 2020. Among the renewable energy sources, photovoltaic (PV) energy will have the highest growth rate (around 30%) during this time span and will cover about 9% of the electricity produced by renewable energy sources. Nevertheless, this goal will only be achieved if research will develop innovative solutions. Photovoltaic cells can be divided into three categories: (1) inorganic cells, based on solid-state inorganic semiconductors; (2) organic cells, based on organic Electrochemistry of Functional Supramolecular Systems. Edited by Paola Ceroni, Alberto Credi, and Margherita Venturi Copyright 2010 John Wiley & Sons, Inc.
523
524
DESIGN OF DYE-SENSITIZED SOLAR CELL COMPONENTS
semiconductors; and (3) photoelectrochemical (PEC) cells, based on interfaces between semiconductors and molecules. PV cells generally consist of a light absorber that will only absorb solar photons above a given minimum photon energy. This minimum energy threshold is called the “energy gap” or “bandgap” (Eg); photons with energies below the bandgap pass through the absorber, while those with energies above the bandgap are absorbed. As already mentioned, the light absorber in PV cells can be inorganic semiconductors, organic molecular structures, or a combination of both. Inorganic semiconductor materials, such as Si, have electronic states grouped within specific energy ranges called bands. The energy ranges, or bands, have energy gaps between them. The band containing electrons with the highest energies is called the valence band. The next band of possible electron energies is called the conduction band; the lowest electron energy in the conduction band is separated from the highest electron energy in the valence band by the bandgap. When all the electrons in the absorber are in their lowest energy state, they fill up the valence band, so that the conduction band is empty of electrons. This is the usual situation in the dark. When photons are absorbed, they transfer their energy to electrons in the filled valence band and promote them to higher energy states in the empty conduction band. There are no energy states between the valence and conduction bands, and thus the term bandgap indicates that only photons with energies equal or above this characteristic energy threshold can induce the transfer of electrons from the lower energy-state valence band into the higher energystate conduction band. When photons transfer electrons across the bandgap, they create negative charges in the conduction band and leave behind positive charges, called holes (h þ ) in the valence band. Accordingly, absorbed photons in semiconductors create pairs of negative electrons and positive holes. In a PV cell, electrons and holes formed upon absorption of light separate and move to opposite sides of the cell structure, where they are collected and flow through wires connected to the cell to produce a current and a voltage, thus generating electrical power (Fig. 17.1). All PV cells depend upon the absorption of light, the subsequent formation and spatial separation of electrons and holes, and the collection of the electrons and holes at different energies (called electrical potential). The efficiency of electron and hole formation, separation, and collection determines the photocurrent, whereas the energy difference between the electrons and the holes in their final state before leaving the cell determines the photovoltage. The cell delivers maximum power Pmax when operating at a point on the characteristic where the product IV is maximum. This is shown graphically in Fig. 17.2, where the position of the maximum power point represents the largest area of the rectangle shown. Figure 17.3 illustrates the various commercial large-area module efficiencies and the best laboratory efficiencies obtained for various materials and technologies. Although conventional solar cells based on silicon are produced from abundant raw materials, the high-temperature fabrication routes to single-crystal and polycrystalline silicon are energy intensive and expensive. The search for alternative solar cells has therefore focused on thin films composed of amorphous silicon and on other semiconductor heterojunction cells (e.g., cadmium telluride and copper indium
17.1 INTRODUCTION
Figure 17.1
525
Photoinduced charge separation in a p–n junction.
diselenide) that can be prepared by less energy intensive and expensive routes. A key problem in optimizing the cost/efficiency ratio of such devices is the need of relatively pure materials necessary to ensure that the photoexcited carriers are efficiently collected in conventional planar solar cell devices. Substantial research efforts have
Figure 17.2
I–V curve of a solar cell and maximum power delivered.
526
DESIGN OF DYE-SENSITIZED SOLAR CELL COMPONENTS
Figure 17.3
Best laboratory efficiencies obtained for various solar devices.
produced CdTe and CuInSe2 solar cells with efficiencies approaching 20%. Industrial efforts to manufacture cells made of these materials in high volumes are beginning to demonstrate success. Despite the progress achieved so far in the fabrication of PV cells, there is an intrinsic source of inefficiency reduction coming from the loss of sub-bandgap energy photons that are not adsorbed into the cell. One of the approaches to tackle this problem is the construction of multijunction cells, which consist of devices involving more than one solar cell, in such way that the higher energy radiation is adsorbed by a larger bandgap solar cell and the residual low-energy solar radiation is adsorbed by a smaller bandgap solar cell. Triple and quadruple junctions have a theoretical efficiency of 30–40% due to their broader spectral response.2 Also, organic molecular compounds can absorb photons of given energy: the photon absorption promotes the molecule to a higher energy electronic state, called excited state. These excited molecular states can eventually generate separated electrons and holes. Furthermore, some organic polymers and various molecular structures can form organic semiconductors that provide the basis for organic PV devices. One main difference between inorganic and organic PV cells is that in organic cells the electrons and holes are initially bound together in pairs called excitons that must be broken apart to separate the electrons and holes and generate electricity, whereas in inorganic PV cells, electrons and holes created by the absorption of light are not bound together but free to move independently in the semiconductor. Organic materials are chemically tunable (tailored molecules) to adjust physical properties such as bandgap, valence and conduction energies, charge transport, solubility, and morphological properties. In general, their processing is fairly easy and well established using both wet-processing techniques (spin coating,
17.1 INTRODUCTION
527
cast coating, ink-jet printing) as well as dry-processing techniques (thermal evaporation).3,4 Due to the small quantities needed for device preparation and the easy largescale production and purification, organic materials entail a further economic advantage with respect to the inorganic materials. Therefore, organic photovoltaic solar cells bear an important potential of development in the search for low-cost modules for the production of domestic electricity. Dye-sensitized solar cells (DSSCs) are photoelectrochemical solar devices, currently subject of intense research in the framework of renewable energies as a low-cost photovoltaic device. DSSCs are based upon the sensitization of mesoporous nanocrystalline metal oxide films to visible light by the adsorption of molecular dyes.5–7 Photoinduced electron injection from the sensitizer dye (D) into the metal oxide conduction band initiates charge separation. Subsequently, the injected electrons are transported through the metal oxide film to a transparent electrode, while a redoxactive electrolyte, such as I/I3, is employed to reduce the dye cation and transport the resulting positive charge to a counter electrode (Fig. 17.4). DSSCs efficiencies up to 10.4%8 have been reported for devices employing nanocrystalline TiO2 films. Several studies have addressed the use of alternative metal oxides including SnO2,9,10 ZnO,11,12 and Nb2O5.13 The performance of dyesensitized solar cells can be understood in view of the kinetic competition among the various redox processes involved in the conversion of light into electricity. Ultrafast electron injection (k2) has been observed in the femtosecond–picosecond (1015– 1012 s) time domain. Regeneration of the oxidized dye (k4) is typically characterized by rate constants of 107–109 s1. This is more than 100 times faster than recombination of injected electrons with the oxidized redox species (k6) and orders of magnitude
Figure 17.4
Electron transfer processes involved in a DSSC with the related rate constants.
528
DESIGN OF DYE-SENSITIZED SOLAR CELL COMPONENTS
faster than back transfer to the dye cation in the absence of a redox mediator (k3). As electron transport in the semiconductor electrode is generally one order of magnitude faster than recombination, the charge collection efficiency is near unit for optimized cells. Because of the prevailing role of electron transfer dynamics in DSSCs, the various processes have been widely studied in the last decade. While photoelectrochemical techniques have proved to be most adequate for the study of electron transport, time-resolved optical spectroscopy remains the leading tool for the study of interfacial electron transfer. Dye regeneration and back transfer reactions have been intensely studied by nanosecond laser spectroscopy. On the other hand, due to its astonishing rate, the forward electron transfer reaction remained unresolved for several years: the advent of femtosecond laser spectroscopy opened the door to the domain of ultrafast chemical processes. The functioning of DSSCs is the interlacing of several subsystems that should work in tandem: the dye sensitizer, that is, the dye adsorbed on the semiconductor surface is able to absorb the visible and near-IR photons and to pump electrons into the conducting band of semiconductor, the electron mediator for “hole” conduction, and the counter electrode catalytic materials. In this chapter, we report on the recent advances in the design of DSSC components and give an insight into their possible improvement. A few basic equations recalling the fundamentals of homogeneous and heterogeneous electron transfer process, along with the parameters governing cell efficiencies, are described in the next two sections. 17.2
ELECTRON TRANSFER PROCESSES
Electron transfer, regardless of the ground or excited-state nature of the reactants, can be treated with the same formalism. The electron transfer reaction can be schematized as
A þ B ! A þ þ B
where an excited donor ( A) transfers an electron to an acceptor B becoming an oxidized species A þ at the ground state. The general equation expressing the rate constant of a reaction is the Eyring equation ket ¼ kn exp
DG KT
ð17:1Þ
Where DG* is the activation energy of the process, K and T are the Boltzmann constant and the absolute temperature, respectively, n is the nuclear frequency factor, and k is the transmission coefficient, a parameter that expresses the probability of the system to evolve from the reactant to the product configuration once the crossing of the potential energy curves along the reaction coordinate has been reached (Fig. 17.5).
17.2
Figure 17.5
ELECTRON TRANSFER PROCESSES
529
Free energy curves and kinetic parameters for an electron transfer process.
In the case of electron transfer reaction, a modification of Equation 17.1 due to R. Marcus has been proposed.14–16 ! ðDG0 þ lÞ2 ð17:2Þ ket ¼ kn exp 4lKT where the activation energy DG* is expressed as DG ¼
ðDG0 þ lÞ2 4l
! ð17:3Þ
DG is the free energy variation characterizing the process, l is the total reorganization energy and represents the vertical energy necessary to transform the nuclear configurations of the reactant and of the solvent to those of the product state in a virtual isoergonic process (Fig. 17.5). It is the sum of inner li and outer lo components: li is due to the vibrational rearrangement consequent to the electron transfer, while lo is comprehensive of the reorganization of the solvation sphere (repolarization). Electron transfer reactions have also been treated from the quantum mechanical point of view in formal analogy to radiationless transitions, considering the weakly interacting states of a supermolecule AB: the probability (rate constant) of the electron transfer is given by a “golden rule” expression of the type17 ket ¼
4p2 el 2 HAB FCWD h
ð17:4Þ
530
DESIGN OF DYE-SENSITIZED SOLAR CELL COMPONENTS
el where HAB is the electronic coupling element between the two electronic states involved in the process and FCWD is the Franck–Condon factor (nuclear factor) weighted for the Boltzmann population of the vibrational energy levels. It can be shown that in the high-temperature limit (when hn < kT for the relevant nuclear vibrational frequencies, an adequate approximation for many cases at room temperature), the nuclear factor takes the form
h i FCWD ¼ ð1=4plkBTÞ1=2 exp ðDGo þ lÞ2 =4lkBT
ð17:5Þ
The exponential term of Equation 17.5 is the same as that predicted by the classical Marcus model, based on parabolic (free) energy curves for reactants and products such as those of Fig. 17.5, where the activation free energy is that required to go from the equilibrium geometry of the reactants to the crossing point of the curves. Both the classical and quantum results in Equations 17.2 and 17.5 contain an important prediction, namely, the existence of three typical kinetic regimes, depending on the driving force of the electron transfer reaction: (i) “normal” regime for small driving forces (l < DG < 0) where the process is thermally activated and is favored by an increase in driving force; (ii) “activationless” regime (l ¼ DG ) where no gain in rate can be obtained by changing the driving force; (iii) “inverted” regime for strongly exergonic reactions (l > DG ) where the process slows down with increasing driving force. Increasing l slows down the process in the normal regime, but accelerates it in the inverted regime. The three kinetic regimes are schematically shown, in terms of classical Marcus’ parabolas, in Fig. 17.6. The inverted region was initially predicted by Marcus and the decrease in the electron transfer rate constant with DG has been observed experimentally many times.18 This is an important and remarkable result both for natural and artificial photosynthesis and energy conversion: it predicts that, following electron transfer quenching of the excited A*–B, the back electron transfer in the inverted region for the charge-separated state A þ –B becomes slower as the energy stored increases.
Figure 17.6 Free energy curves for reactant and product states of an electron transfer process in the kinetic regimes of the Marcus model.
17.2
ELECTRON TRANSFER PROCESSES
531
The Marcus theory can also be applied to heterogeneous electron transfer, namely, electrode reactions, in close analogy to homogeneous reactions.19 The two main assumptions on the basis of this model are the following: (1) the electron transfer reaction, whether homogeneous or heterogeneous, is a radiationless transition occurring between isoenergetic states (the electron must move from an initial state on the electrode or in the reductant to a receiving state in the molecular species or on the electrode of the same energy); (2) reactants and products retain their configurations during the actual act of charge transfer because nuclear momenta and positions do not change on the timescale of electronic transitions, and thus the reactant and the product share a common nuclear configuration at the moment of the transfer. These assumptions lead to an expression relating the heterogeneous electron transfer (reduction) rate constant, kf, to the activation energy, DG*, of the type kf ¼ Knkel expðDG* =RTÞ
ð17:6Þ
where K is a precursor equilibrium constant representing the ratio of the reactant concentration in a reactive position (precursor state) with respect to the concentration in the bulk solution; for an heterogeneous electron transfer, the precursor state can be considered as a reactant molecule situated near the electrode at a distance where electron transfer is possible, thus allowing for a reduction reaction, K ¼ Co,surf/Co,bulk, where Co,surf is the surface concentration of electron acceptor in mol/cm2, n is the nuclear frequency factor (s1) that represents the frequency of attempts on the energy barrier, and kel is the electronic transmission coefficient (related to the probability of electron tunneling). The dimension of kf appearing in Equation 17.6 is, as required, cm/s. The act of electron transfer is usually considered as a tunneling of the electron between states in the electrode and those in the reactant. Tunneling effects are included in the transmission coefficient kel that falls off exponentially with the distance according to kel ¼ k0el expðbxÞ
ð17:7Þ
where x is the distance over which the tunneling occurs and b is a factor depending on the height of the energy barrier and on the nature of the intervening medium. kel depends on the electronic coupling between the reactant and the electrode: a strong coupling results in a splitting of the curves (Fig. 17.7a) at the crossing, which results in an higher probability for the system to evolve from the reagent to the product configuration once it has reached the transition state. A situation where the system stays on the lower curve passing from reagent to product is called adiabatic (kel ¼ 1). If the splitting is small compared to kT, the system tends to reside on the original reactant configuration and is said to be nonadiabatic (kel < 1).
532
DESIGN OF DYE-SENSITIZED SOLAR CELL COMPONENTS
Figure 17.7 Splitting of energy curves for a reduction reaction. In (a) the strong coupling leads to a considerable splitting and to a continuous curve connecting the reactant and the products configurations, favoring the evolution of the system toward the product state. In (b) the small splitting favors the tendency of the system to remain in the reactant curve (O þ e).
17.3
DSSCs CHARACTERIZATION
The performance of the cell can be quantified on a macroscopic level with parameters such as incident photon to current efficiency (IPCE), open-circuit photovoltage (Voc), and the overall efficiency of the photovoltaic cell (hcell). The parameter that directly measures how efficiently incident photons are converted to electrons is the IPCE. The wavelength-dependent IPCE term can be expressed as a product of the quantum yield for charge injection (F), the efficiency of collecting electrons in the external circuit (h), and the fraction of radiant power absorbed by the material or light harvesting efficiency (LHE), as represented by Equation 17.8: IPCE ¼ ðFÞðhÞðLHEÞ
ð17:8Þ
While F and h can be rationalized on the basis of kinetic parameters, LHE depends on the active surface area of the semiconductor and the cross section for light absorption of the molecular sensitizer. In practice, the IPCE measurements are performed with monochromatic light, and IPCE(l) values are calculated according to Equation 17.9. IPCEðlÞ ¼ 1:24 103 ðeV nmÞ Isc ðmA=cm2 Þ=lðnmÞIðmW=cm2 Þ
ð17:9Þ
In Equation 17.9, Isc is the photocurrent density produced by the cell, l the excitation wavelength, and I the incident photon flux. The light harvesting efficiency, or absorption factor, is related to the dye molar extinction coefficient (e(l), L/mol/cm) and to the surface coverage (G, mol/cm2)
17.3
DSSCs CHARACTERIZATION
533
by Equation 17.10. LHEðlÞ ¼ 110½1000eðlÞG
ð17:10Þ
An absorption factor above unity is ideal for a solar energy device as almost all the incident radiant power is collected. In the absence of photodecomposition reactions, the quantum yield for electron injection from the excited sensitizer to the semiconductor is given by Equation 17.11. F ¼ k2 =ðk1 þ k2 Þ
ð17:11Þ
In Equation 17.11, k1 is the sum of the radiative and nonradiative rate constants of the excited dye molecule (k1 ¼ kr þ knr) and k2 is the rate constant for the charge injection process. Charge injection from the excited dye will be activated if the donor energy is positive with respect to the conduction bandedge (ECB). Electron injection will be, on the contrary, activationless if the donor level has energy equal to or more negative than the conduction bandedge. This condition is met when EðS þ =S Þ þ l > ECB , where E(S þ /S*) represents the excited-state oxidation potential of the sensitizer (which is related to the ground-state oxidation potential and to the spectroscopic energy by E(S þ /S*) ¼ E(S þ /S) E00) and l represents the reorganization energy accompanying the charge injection process. The fraction of injected charges that percolate through the TiO2 membrane and reach the back contact of the photoanode is represented by the factor h. It has been shown that the electron diffusion pffiffiffiffiffiffiffiffiffiffi length (Ln), which is a key parameter for charge collection, given by Ln ¼ D0 t0 (D0 and t0 are the diffusion coefficient and lifetime of electrons in conduction bands), is of the order of 20–100 mm for the liquid electrolyte-based DSSCs, thus allowing to use relatively thick photoanodes for optimizing the light harvesting efficiency of the solar device. It is accepted that electron diffusion occurs with an ambipolar mechanism,20–22 which involves simultaneous motion of electrons and electrolyte cations. Although the diffusion coefficient of cations in the nanoporous network is normally small, high concentrations of cations, intercalated between the semiconductor nanoparticles, support a fast transport of the electrons as minority carriers, as shown by electron-density-dependent diffusion coefficients when high-ion-concentration electrolytes are used. The maximum open-circuit photovoltage, attainable in a dye-sensitized solar cell, is the difference between the Fermi level of the solid under illumination and the Nernst potential of the redox mediator. However, for these devices this limitation has not been achieved and Voc is in general much smaller. It appears that Voc is kinetically limited and the diode Equation 17.12 can be applied for an n-type semiconductor in a regenerative cell.23 Voc ¼
Iinj kT ln P n i ki ½Ai e
ð17:12Þ
534
DESIGN OF DYE-SENSITIZED SOLAR CELL COMPONENTS
In Equation 17.12, Iinj is the electron injection flux, n is the concentration of electrons in TiO2, and the summation is for all electron transfer rates to acceptors Ai. One successful strategy to increase Voc is by adding pyridine derivatives to the electrolyte to adsorb on the TiO2 surface inhibiting recombination of conduction band electrons with the electron mediator (k6). Finally, the overall efficiency of the photovoltaic cell, hcell, is given by Equation 17.13. hcell ¼
isc xVoc xff I
ð17:13Þ
In this equation, isc is the short-circuit photocurrent, ff is the “fill factor” of the cell, and I is the intensity of the incident light.
17.4
MOLECULAR SENSITIZERS
One of the most important DSSC component is the dye sensitizer that represents the electron pump of the device. It allows an independent electron injection into the semiconductor conduction band and conversion of visible and near-infrared (NIR) photons to electricity. Several organic and inorganic compounds have been investigated for semiconductor sensitization, such as chlorophyll derivatives,24 porphyrins,25 phthalocyanines,26,27 platinum complexes,28,29 fluorescent dyes,30 carboxylated derivatives of anthracene,31 polymeric films,32 coupled semiconductors with lower energy bandgaps,33 among others. Fruit extracts have also been used as natural sensitizers in photoelectrochemical solar cells.34,35 However, the best solar to electric power conversion efficiency has been achieved with polypyridyl complexes of ruthenium (II)36 and osmium(II)37 bearing carboxylated ligands, which are often employed as TiO2 sensitizers in such cells. These species give rise to intense visible metal-toligand charge transfer (MLCT) bands with a favorable energetics for activationless charge injection. Among this family of compounds, excellent results have been achieved with thiocyanate derivatives. In particular, the performances of the Ru(II)– NCS complexes (Fig. 17.8) in photoelectrochemical solar cells were found to be outstanding.36,38 The complexes show photoaction spectra dominating almost the entire visible region, short-circuit photocurrents exceeding 16 mA/cm2 in simulated A.M. 1.5 sunlight and open-circuit photovoltages of the order of 0.7 V in presence of I3/I as redox electrolyte. Their high efficiency is related to hole delocalization across the NCS ligands,39,40 which is responsible of an increased electronic coupling for the electron transfer reaction involving TiO2/RuIIINCS and I. This leads to an increase of the rate constant of the reductive process (k4, in Fig. 17.4) and, as a consequence, of IPCE. Based on the general structure of the N3 complex, mononuclear sensitizers with a higher extinction coefficient have been synthesized. An example is offered by the K19 complex38 (Fig. 17.8), which brings together the advantages of the
17.4
MOLECULAR SENSITIZERS
535
Figure 17.8 Structure of some of the most successful polypyridine Ru(II) dye based on the NCS ancillary ligand.
amphiphilic dyes with an improved light harvesting efficiency. In fact, the low-energy MLCT transition at 543 nm of this complex has a molar extinction coefficient of 18.2 103 M1 cm1, which is higher than that of the Z907 (12.2 103 M1 cm1) and N719 (14.0 103 M1 cm1) dyes due to the extended p-conjugation provided by the hexyloxystyryl substituents at the 4,40 -position of the bpy ligand. The K19 complex yielded an IPCE exceeding 70% in a broad spectral range from 400 to 650 nm and gave an overall efficiency of 7%. Moreover, it maintained its conversion efficiency when the TiO2 film thickness was reduced by a 20% amount. The photoelectrochemical performances given by the mononuclear complex [Ru(dcbH2)2(NCS)2] have been surpassed by an analogous species based on the terpyridine ligand.8 Indeed, TiO2 electrodes covered with the complex Ru(L)(NCS)3 (L ¼ 4,40 ,40 tricarboxy-2,20 :60 ,20 -terpyridine) in its monoprotonated form displayed very efficient panchromatic sensitization covering the whole visible spectrum and extending the spectral response at the near-IR region up to 920 nm, with maximum IPCE values comparable to that obtained with the dithiocyanate complex. The panchromatic efficiency has been further extended in a series of cationic dyes, corresponding to the general formulas [M(H3tcterpy)LY] þ 41: M ¼ Os(II) or Ru(II); (H3tcterpy) is the tridentate ligand, 4,40 ,40 -tricarboxy-2,20 :60 ,20 -terpyridine, and L is a bidentate ligand of the type bpy (2,20 -bipyridine) or pyq (2-20 -pyridylquinoline) (Fig. 17.9), which can be substituted in the both 4 and 40 positions by X ¼ H, CH3, COOH, or C(CH3)3, Y ¼ Cl, I, or NCS. These dyes show reversible metal oxidations and, in the Os
536
DESIGN OF DYE-SENSITIZED SOLAR CELL COMPONENTS
Figure 17.9 [M(H3tcterpy)LY] þ cationic dyes: M ¼ OsII, RuII; L ¼ bpy, pyq.
Figure 17.10 Photoaction spectra of Os dyes measured in sandwich-type solar cells (nanocrystalline TiO2 films, electrolyte made of 2 M LiI þ 0.1 M I2 in g-butyrolactone). OsH2dcpyCl ¼ [Os(H3tcterpy)(H2dcbpy)(Cl)] þ ; OstbbpyCl ¼ [Os(H3tcterpy)(tbbpy)(Cl)] þ ; OspyqCl ¼ [ Os(H3tcterpy)(pyq)(Cl)] þ . Red and black dye are N3 and Ru(H3Tcterpy)(NCS)3, respectively.
case, unprecedented photoconversion efficiency in the near infrared, approaching, with the best complexes, the value of 50% at 900 nm (Fig. 17.10).
17.5
ELECTRON MEDIATORS
DSSCs convert sunlight to electricity by a different mechanism than conventional p–n junction solar cell. Light is absorbed directly at the solid/liquid interface by a monolayer of adsorbed dye, and initial charge separation occurs without the need of exciton transport.42,43 Following the initial charge separation, electrons and holes are confined in two different chemical phases: electrons in the nanocrystalline
17.5
ELECTRON MEDIATORS
537
semiconductor and holes (oxidized redox species) in the electrolyte solution that permeates the solid phase. According to the widely accepted kinetic model, electron transport to the charge collector occurs by diffusion in an electric field-free regime.42 No electric field is created, essentially due to the fact that the sintered TiO2 nanoparticles are too small and too lightly doped to support a significant space charge and also due to a screening effect by the surrounding electrolytic solution. It is a chemical potential gradient that drives the electrons to the back contact, while the opposite electrical potential gradient, developed between the photoinjected electrons and the positive charges (either oxidized dye or oxidized electrolyte species) is screened by the electrolyte (according to some authors, without this effect, in the absence of band bending, most of photoinjected electrons would never be able to escape from their image charge and recombine before reaching the electron collector).44 Anyway, to contribute to the photocurrent, the photoinjected electrons must diffuse through hundreds of nanoparticles in close proximity to the oxidized redox species (electron acceptors) in solution and, in the absence of an electrostatic barrier, interfacial recombination may result as a major energy loss mechanism. So far the iodide/iodine electrolytes have been the most efficient and commonly used redox mediators, due to the fact that I allows for a fast regeneration of the oxidized dye, intercepting it on a nanosecond timescale45,46 (Fig. 17.11), whereas the reduction of I2 and I3 is a complex multistep reaction that involves the breaking of chemical bonds: some authors47,48 have studied it in great detail and suggested a first dissociation of I3 in I and I2, a subsequent reduction of I2 to I2 followed by the rate-limiting dismutation of two I2 to give I and I3. The same authors suggested that the reduction of I2 occurs only with adsorbed I2 molecules and the overall process is further slowed down by its relatively low concentration compared to I3 in solution. In summary, the electronic recapture involving I3 is kinetically so slow on both TiO2 and SnO2 surfaces that, under short-circuit conditions, most of electrons survive the transit through the mesoporous film and the SnO2 surface and appear in the
Figure 17.11
N3 recovery in presence of increasing concentration of NaI in acetonitrile.
538
DESIGN OF DYE-SENSITIZED SOLAR CELL COMPONENTS
external circuit. Thus, the I/I3 couple appears to have ideal kinetic properties that lead to an “asymmetric behavior” on the basis of the efficient functioning of the DSSC: the forward electron donation by I is a facile monoelectronic process that ensures an efficient dye recovery (k4 in Fig. 17.4), while the reduction I3 appears to be largely inefficient allowing for a minimization of the interfacial back recombination (k6 in Fig. 17.4). In other respects, the I/I3 redox couple is less than ideal for a number of reasons: I2 in equilibrium with I3 is volatile, complicating long-term cell sealing; I3 is darkly colored and limits the light harvesting efficiency of the dye, and DSSC cathodes require platinum coatings to effectively catalyze the I3 reduction. Most importantly, I/I3 is corrosive and will corrode most metals, posing a serious problem for the scaling up of the solar cells to large areas. Indeed, the sheet resistance of the conductive glass is relatively high, leading to a serious ohmic loss in cells of area exceeding few square centimeter.43 A way to solve this problem is based on the deposition of metallic grids, acting as electron collectors, on the transparent oxide: a solution that cannot be reliably applied in DSCs as long as the iodide/iodine couple is used. Besides these practical problems, the search for new electron transfer mediators potentially capable of replacing the iodide/iodine couple is attractive for at least two fundamental reasons: first, the investigation of new redox systems may improve our knowledge of the basic interfacial charge separation events and their dependence from the inherent structural and electrochemical properties of new mediators, and second, the discovery of new efficient electron mediators may open the way to new strategies for changing, optimizing, and improving DSSC design and performances. This type of research is challenging for a number of reasons related to the fact that an efficient electron mediator must simultaneously fulfill at least three relatively strict requirements: it has to intercept efficiently the oxidized form of the dye, recombine slowly with photoinjected electrons on both TiO2 and SnO2 substrates, and allow for an efficient mass transport in solution and into the TiO2 mesopores. As explained above, a sluggish electron recapture onto semiconductor oxide surfaces is believed to be crucial for the correct operation of the mediator. To our knowledge, there is no certain a priori indication that such requirement would be satisfied by a given chemical species on a specific substrate; however, in principle, every system characterized by an high reorganization energy associated to the electron transfer should exhibit a slow kinetic (see Sections 17.2 and 17.7). In the following, we report on the main attempts made by several international research groups to develop and improve new redox systems for DSC: these comprise quasi-solid-state gel electrolytes based on the I/I3, metal-organic couples based on coordination compounds, inorganic redox systems, conductive polymers, and inorganic solid-state hole conductors.
17.6 IONIC LIQUIDS AND GEL ELECTROLYTES BASED ON THE I/I3 REDOX COUPLE The presence of a redox couple dissolved in a liquid solvent creates stability problems for long-term operation of the DSSC, essentially related to cell sealing (solvent
17.6
IONIC LIQUIDS AND GEL ELECTROLYTES BASED ON THE I/I3 REDOX COUPLE
539
Figure 17.12 Imidazolium salts employed as a I source in quasi-solid-state electrolytes. R0 and R are usually methyl and propyl or ethyl and methyl groups, respectively.
evaporation and leaking). The main attempts for obtaining a redox-active gel have been based on the employment of the iodide/triiodide redox system: iodide couples its ideal kinetic properties to the small radii of the electroactive ions, ensuring an efficient diffusional transport even in a relatively viscous medium. Of course, even the use of a gel or of a ionic liquid does not overcome the drawbacks (corrosivity, etc.) related to the intrinsic properties of the I/I3 redox system. Room-temperature ionic liquids are attractive due to their chemical and thermal stability, negligible vapor pressure, high ionic conductivity, and ample electrochemical window. Their properties can be varied by a rational choice of the cations and of the anions and can represent an important iodide source for an I/I3-based electrolyte (Fig. 17.12). Gelation of liquid electrolytes to obtain a gel that effectively penetrates into the pores of the TiO2 can be obtained by addition of appropriate additives such as small molecules,49 polymers, inorganic nanoparticles,49,50 and carbon nanotubes.51 In one of the first studies, Yanagida et al.52 used a series of amphiphilic small molecules (Fig. 17.13) to obtain the gelation of an I/I2-based electrolyte composed by 0.6 M 1,2-dimethyl-3-propylimidazolium iodide, 0.1 M lithium iodide, 0.1 M
Figure 17.13 Chemical structures of the gelators employed by Yanagida et al. Reprinted with permission from Ref. 52. Copyright 2001 American Chemical Society.
540
DESIGN OF DYE-SENSITIZED SOLAR CELL COMPONENTS
iodine, and 1 M tert-butylpyridine in methoxypropionitrile. These gelators form thermoreversible physical gels from a variety of organic liquids, showing solution-togel transition temperatures from about 50 to 85 C, depending on the characteristics of the gelator molecules. Characteristics of the gelators are originated from their ability in the formation of intermolecular hydrogen bonds between the oxygen atom of the urethane group and the hydrogen of the amide group. The long alkyl chains also contribute to form aggregates via van der Waals forces forming a network encapsulating the liquid electrolyte. Other approaches by Graetzel and coworkers have privileged the use of photochemically stable fluorinated polymers such as poly(vinylidenefluoride-co-hexafluoropropylene) (PVDF-HFP)53 or 12 nm titanium dioxide or silica nanoparticles to induce the gelation of liquid electrolyte based on the 1-methyl-3-propylimidazolium iodide (MPII). When nanoparticles are used for the gelation, it is important that their size is much smaller than the pores of the TiO2 film, avoiding any problems for the quasi-solid-state electrolyte to permeate the mesoporous film. The performances of the gel electrolytes is well comparable to those of the liquid electrolyes, both under monochromatic and polychromatic irradiation: efficiencies as high as 7% have been reached under simulated solar irradiation, as well as maximum monochromatic photon to electron conversion of nearly 80%. Yanagida and coworkers reported that the conductivity of gel electrolytes was found to follow an Arrhenius-type behavior according to Equation 17.14 s ¼ s0 expðEa =kTÞ
ð17:14Þ
Where s is the conductivity, Ea is the activation energy, and k is the Boltzmann constant. Interestingly, in most cases the conductivity of the gel does not decrease significantly when compared to that of the equivalent liquid in the measured regions of temperature (20–60 C). This finding coupled to the fact that similar activation energy (11–12 kJ/mol) was found both for the liquid and for the quasi-solid-state electrolyte suggested that the channel for the charge transport in gel electrolytes is almost identical with that of the liquid electrolyte. The evidence that the I/I3 couple can diffuse freely in the liquid domains entrapped by the three-dimensional network of the gelators has also been found in the case of a PVDF-HFP gel via steady-state voltammetry at ultramicroelectrodes. Quite surprisingly the voltammogramms of the liquid and of the gel are almost perfectly superimposable (Fig. 17.14) and the diffusion coefficient of the redox ions could be calculated to be 3.6 106 cm2/s and 4.49 106 cm2/s for I and I3, respectively, using Equation 17.15, Il ¼ 4nFDCa
ð17:15Þ
where Il is the limiting current, n is the number of exchanged electrons, F is the Faraday constant, D is the diffusion coefficient, C is the concentration of electroactive species, and a is the microelectrode radius.
17.6
IONIC LIQUIDS AND GEL ELECTROLYTES BASED ON THE I/I3 REDOX COUPLE
541
Figure 17.14 Steady-state voltammetry of a liquid and polymer (PVDF-HFP) gel electrolyte at a Pt ultramicroelectrode. Scan speed 10 mV/s. Reprinted by permission from Mac Millan Publishers Ltd: Nature Materials, 2003, 2, 402.
Nanosecond transient absorption measurements provided a further indication of efficient mediator transport within the nanopores of the gel, showing basically no difference in dye regeneration using the liquid iodide/iodine precursor and the quasisolid-state polymeric electrolyte. More recently, the effect of addition of nanocomposite materials to ionic liquid electrolytes has been explored51: briefly, multiwalled carbon nanotubes (MWCNTs), graphite, single-walled carbon nanotubes, carbon fibers, and titanium dioxide nanoparticles have been dispersed in 1-ethyl-3-methylimidazolium bis(trifluoromethylsulfonyl)imide in the presence of 1.5 M 1-ethyl-3-methylimidazolium iodide, 0.1 m LiI, 0.15 M I2, and 0.5 M 4-tert-butylpyridine. After prolonged centrifugation and elimination of the excess liquid, a thick gel that could be spread onto the photoanode was obtained (Fig. 17.15). Despite an increased viscosity, the conductivity of the nanocomposite gel was found to be higher than the corresponding ionic liquid, and an improvement in the electron transport via an exchange reaction between I and I3 due to the adsorption of the 1-methyl-3-ethyl-imidazolium cations at the surface of carbon and TiO2-based nanomaterials was hypothesized: upon cation adsorption, the positive polarization of the surface attracts a local excess of I and I3 ions giving rise to a “nanoscale ordering,” which may improve the electron transport across the cell via a Grotthus-like mechanism.54 Indeed, the photovoltaic performances of the nanocomposite mediator were enhanced with respect to the liquid precursor, with the best results obtained in the case of the MWCNT (Fig. 17.15). The use of quasi-solid-state electrolytes usually reaches the goal of a superior thermal stability with respect to liquid electrolytes: cells can survive prolonged periods (accelerated aging of 1000 h) at relatively high temperatures (55–80 C) under
542
DESIGN OF DYE-SENSITIZED SOLAR CELL COMPONENTS
Figure 17.15 Photocurrent–voltage characteristics of dye-sensitized solar cells in the presence of various nanocomposite gel electrolytes. The inset shows the viscous MWCNT gel at the bottom of a test tube. Reprinted from Ref. 51. Copyright 2004 with permission from Elsevier.
light soaking without indication of performance loss. Thus, the quasi-solid-state cells offer specific benefits over the use of ionic liquids, enabling also the fabrication of flexible devices free of leakage and available in various geometries.
17.7 ELECTRON MEDIATORS BASED ON COORDINATION COMPOUNDS: MIMICKING I/I2 KINETICS The electrochemical properties of coordination compounds can be tuned through a rational choice of the metal and an appropriate design of the coordination sphere. These features might guarantee the necessary flexibility to project an electron transfer mediator capable of meeting the kinetic requirements on the basis of the functioning of the DSC. In principle, to minimize recombination kinetically slow couples have to be privileged. The relatively high concentration of electrolyte employed in photoelectrochemical cells (0.1–0.5 M in redox-active species) requires the use of substantial amounts of electron mediators: as a consequence, the choice and the design of the redox couple have to be done with consideration toward inexpensive and available metals such as the elements of the first transition row and easily synthesizable ligands. To date, the most successful attempts have been based on octahedral cobalt(II) and tetrahedral copper(I) complexes. The Co(II)/(III) couple is usually characterized by an high inner sphere reorganization energy associated with the electron transfer,
17.7
ELECTRON MEDIATORS BASED ON COORDINATION COMPOUNDS
543
Figure 17.16 Structure of the Co(II)(dbbip)22 þ . Reprinted with permission from Ref. 55. Copyright 2001 American Chemical Society.
essentially due to the involvement of a metal-centered eg redox orbital with antibonding characteristics. One of the first reports by Graetzel and coworkers was focused on the use of a oneelectron Co(II) redox mediator based on the Co(II)-bis-[2,6-bis(10 -butylbenzimidazol20 -yl)pyridine]55 (dbbip) complex (Fig. 17.16). The Co(III)(dbbip)23 þ /Co(II)(dbbip)22 þ potential was of 0.36 V versus SCE, a value comparable with that of the I/I3 couple. Used in conjunction with a compact blocking TiO2 underlayer, whose function is devoted to the minimization of the dark current (recombination with Co(III)), Co(dbbip)2 appeared to rival the iodide/ triiodide performances in terms of IPCE and global cell efficiency. Encouraging was also the weak coloration of the complex that exhibited a moderate absorption in the visible region with an extinction coefficient of the order of 102 L/mol/cm at 450 nm that allowed for the preparation of concentrated solutions without a serious competition in light absorption with the sensitizer. Laser photolysis experiments carried out at very low pulse energies (40 mJ/cm2/pulse at 510 nm) indicated a strong acceleration of the decay of the photooxidized dye (N719), monitored at 620 nm, according to a first-order kinetic. This evidence coupled to the fact that a variation of two order of magnitude in laser excitation intensity did not cause any significant variation in the decay rate allowed to infer that the interception of the oxidized dye was totally decoupled from the back electron transfer involving the photoinjected electrons and the dye cations (Fig. 17.17). The reciprocal of the half-lifetime yielded a value for the dye regeneration by Co (dbbip)22 þ of about 5 105 s1 against 1.7 106 s1 observed in the case of 0.1 M tetrabutylammonium iodide. Contrary to the case of iodide,56 a cooperative effect of adsorbed cations was not observed in the case of cobalt, for which an addition of Li þ 0.1 M did not cause any noticeable change in the interception kinetics; however, static interception by adsorbed Co(II) complexes partly compensated for the intrinsically slower reduction of the dye cation by Co(dbbip)22 þ resulting in a pseudo-first-order
544
DESIGN OF DYE-SENSITIZED SOLAR CELL COMPONENTS
Figure 17.17 Transient absorbance decay kinetics of the oxidized Ru(dcbpy)2(NCS)2 at 620 nm in the presence and in the absence of 0.1 M Co(II) in acetonitrile/ethylene carbonate 40:60 v/v. Reprinted with permission from Ref. 55. Copyright 2001 American Chemical Society.
rate constant not too far from that obtained for the iodide/triiodide system. Transient absorption spectroscopy also allowed for an evaluation of the back recombination involving the Co(III) species, which contributed with a residual long-lived (ca. 1 ms) transient positive absorption at 480 nm. From the decay rate of such signal, an apparent first-order rate constant of the order of 3 103 s1 could be obtained from which it could be inferred that recombination involving photoinjected electrons and Co(III) centers was only slightly faster than that in the case of triiodide. The best results with this type of cobalt mediators were obtained using heteroleptic complexes such as Z907 and N621 bearing long alkyl chains (Fig. 17.18). From a certain point of view, this finding could be surprising since the regeneration of N719 was found to be kinetically favored over the latter dyes. However, the adsorption of N719 onto the TiO2 nanoparticles imparts a negative polarization to their surface, which increases the coulombic attraction between the TiO2 nanocrystals and the cobalt complex, hindering the mobility of the redox species within the cell. Moreover, the electrostatic attraction is expected to be higher with Co(III), which tends to reside onto the surface of the TiO2, blocking the dye regeneration and favoring the back recombination. Under this viewpoint, the use of heteroleptic dyes57,58 brings two main advantages: reduction of the ion pairing effect due to a smaller negative z potential and suppression of the electron tunneling to Co(III) acceptors caused by the steric hindrance of the long alkyl chains that limit the access of the cobalt complex to the exposed TiO2 surface. The overall effect is a reduction of the probability of electron recapture, evident from the cell dark current (Fig. 17.19), the longer the alkyl chain the lower the probability of electron recombination with acceptor states of the electrolyte. For similar reasons, the cografting of long alkyl chain carboxylic acids like hexadecylmalonic acid allowed for an enhancement of the photovoltaic response of cobalt-mediated cells.58 One of the limitations of Co(II) benzimidazole complexes is their limited availability related to a multistep synthetic procedure for the ligand. Almost
17.7
Figure 17.18
ELECTRON MEDIATORS BASED ON COORDINATION COMPOUNDS
545
Heteroleptic dyes used in conjunction to cobalt electron transfer mediators.
contemporarily to the efforts made at the EPFL, it was found that particular cobalt polypyridine complexes formed from structurally simple ligands did function as efficient electron transfer mediators in DSSCs.59 The main focus of the investigation was thus to identify which structural and thermodynamic motifs generate the best mediators allowing for the assembly of cells with the closest match to the performance of the I/I3 mediator.
Figure 17.19 Dark currents of DSSCs measured with the four different dyes reported in Fig. 17.9 and [Co(dbbip)2]2 þ /3 þ 0.1/0.01 M, 0.2 M Li þ , and 0.1 M 4-tert-butylpyridine in acetonitirile/ethylene carbonate 1/1. From Herve Nusbaumer, EPFL, 2004.
546
DESIGN OF DYE-SENSITIZED SOLAR CELL COMPONENTS
Figure 17.20 Structure of polypyridilic Co(II) complexes: (a) X ¼ tert-butyl (DTB), methyl (DMB), H (bpy). (b) X ¼ H (phen). (c) X ¼ tert-butyl (tTbterpy), ethyl (tEterpy).
The cobalt complexes reported in Fig. 17.20 can be easily produced by mixing 1 equivalent of [Co(H2O)6]2 þ with 3 equivalents of a bidentate ligand or 2 equivalents of a tridentate ligand under magnetic stirring in refluxing methanol for 2 h. Addition of ethyl ether results in the precipitation and isolation the product that is then usually used without any further purification. All of the complexes under consideration exhibit similar UV–Vis absorption spectra. Each of the Co(II) complexes has a weak absorption band centered at ca. 440–450 nm. The onset of the ligand-based p–p* transition occurs in the UV above 350–380 nm for each of the ligands. The most intense visible absorption is for Co(tTbterpy)32 þ with e450 ¼ 1.4 103 M1 cm1. The remaining complexes all exhibit e440–450 values that are approximately an order of magnitude smaller. For the sake of comparison, the e440–450 value for I3 is ca. 2 103 M1 cm1; therefore, except Co(tTbterpy)32 þ that has a comparable absorbance, considerably less visible light is absorbed by all of the remaining cobalt complexes at similar concentrations. Electrochemical characterization of these complexes revealed an unexpected electrode surface dependence to the electron transfer kinetics. In the case of 4-40 substituted polypyridine Co(II) complexes, gold electrodes exhibit the most reversible and ideally shaped CVs. Glassy carbon and platinum electrodes also produce quasi-reversible voltammograms, although less reversible than gold. In general, the shapes of the quasi-reversible waves indicate that in cases where the heterogeneous electron transfer is slow, the transfer coefficient, a, is considerably greater than 0.5. In other words, for equivalent overpotentials the heterogeneous reduction of the Co(III) complex is considerably faster than the corresponding oxidation of the Co(II) species. The voltammetric results also suggest that while platinum is the cathode of choice for the I/I3 redox mediator, it should not be the optimal choice for cobalt complexbased mediators. Likewise, while carbon is a poor cathode with the I/I3 redox mediator system, it should be acceptable for any of the cobalt systems considered here. It must be noted that among the series of Fig. 17.20, one of the most promising cobalt complexes (Co(DTB)32 þ ) is nearly electrochemically inactive on FTO and ITO electrodes, meaning that the unwanted Co(III) to Co(II) reduction at the
17.7
ELECTRON MEDIATORS BASED ON COORDINATION COMPOUNDS
547
Figure 17.21 Cyclic voltammetry of approximately 1 103 M Co(DTB)32 þ recorded in ACN/TBAPF6 0.1 M on GC (solid line) and on ITO (dotted line).
photoanode requires high overpotentials to be effective (Fig. 17.21). The dramatic dependence of the Co(DTB)32 þ /3 þ heterogeneous electron transfer rate on the chemical nature of the electrode surface is most probably the consequence of specific interactions between the surface and the electroactive species that modify the electronic coupling and/or the activation barrier for the heterogeneous electron transfer reaction. The electrochemical behavior on metal oxide surfaces is in this case similar to the one of the I/I3 couple for which the electron transfer from the FTO substrate to I3 is a very slow multistep process.60 The performances of the photoelectrochemical cells are strongly dependent on the composition of the electrolyte solution (Fig. 17.22). A maximum conversion efficiency of ca. 80%, in correspondence to the metal-to-ligand charge transfer absorption maximum of N3 was obtained in the presence of 0.25 M LiI/0.025 M I2, whereas with the cobalt-based mediators, the best performances (ca. 50–55% of IPCE) were observed when solutions of Co(DTB)32 þ /3 þ and Co(tTBterpy)22 þ /3 þ were used. In the other investigated cases, Co(phen)32 þ /3 þ , Co(tEterpy)22 þ /3 þ , and Co(DMB)32 þ /3 þ mediators exhibited much lower conversions, with maximum IPCE values in the range of 10–20%. Generally, cobalt complexes of unsubstituted and methyl-substituted bipyridine or terpyridine ligands are poor electron transfer mediators in the type of DSSCs considered herein. In contrast, if the ligand contains a tertiary butyl substituent in the para position, the resulting cobalt-based cells yield quite good IPCE. The observed substituent effect cannot be related to an electronic perturbation of the Co(II) redox orbitals and to a consequent modification of the electrochemical properties of the coordination compound, since the electron-donating effect of all simple alkyl substitutents (e.g., methyl, ethyl, tert-butyl, etc.) is essentially the same61 and all of the complexes of a given ligand type (i.e., bipyridine, phenanthroline, or
548
DESIGN OF DYE-SENSITIZED SOLAR CELL COMPONENTS
Figure 17.22 Photoaction spectra of N3 bound to nanocrystalline TiO2 films in the presence of different electron mediators in MPN solutions: 0.25 M LiI/25 mM I2 (circles), 0.25 M Co(tTBterpy)22 þ /25 mM NOBF4 (squares), 0.25 M Co(DTB)32 þ /25 mM NOBF4 (diamonds), 0.25 M Co(phen)32 þ /25 mM NOBF4 (triangles), 0.25 M Co(tEterpy)22 þ /25 mM NOBF4 (upside down triangles), saturated (<0.15 M) Co(DMB)32 þ /15 mM NOBF4 (crosses). An amount of 0.25 M LiClO4 was added to all solutions containing a cobalt mediator.
terpyridine) were expected and found to have very similar E1/2 (100–200 mV versus SCE) values for the relevant Co(II/III) couple. Nanosecond time-resolved experiments allowed for the rationalization of the lower efficiency of the cobalt-based couples with respect to iodide/iodine and for the clarification of their structure-dependent performance. Figure 17.23a shows the decay of the photogenerated N3 dye cation, observed at 480 nm in the presence of Co (DTB)32 þ 0.1 M and TBAI 0.1 M: in both cases, a t2/3 of about 0.35 ms indicates that dye regeneration by iodide and by Co(II) occurs at a very similar rate. Upon Li þ
Figure 17.23 480 nm Ru(III)(H2DCB)2(NCS)2 þ recovery in the presence of (a) Tetrabutylammonium iodide (A) and Co(DTB)32 þ (B) 0.1 M; (b) Tetrabutyl-ammonium iodide (A) and Co(DTB)32 þ (B) 0.1 M þ LiClO4 0.1 M. Pulse energy 1 mJ/cm2. From Bignozzi et al., unpublished results.
17.7
ELECTRON MEDIATORS BASED ON COORDINATION COMPOUNDS
549
addition (Fig. 17.23b), an increased reduction rate by iodide is observed (t2/3 ¼ 0.12 ms), whereas regeneration of the oxidized dye by Co(II) is substantially unchanged. Indeed, TiO2 surface adsorption/intercalation by Li þ creates a positive polarization that attracts a surface excess of iodide in close proximity of the dye sites, allowing for a faster interception of the oxidized sensitizer. The same effect is obviously absent in the case of positively charged Co(II) complexes that do not experience any attraction with a positively polarized photoelectrode: the slower oxidized sensitizer reduction observed in the case of Co(II) electron mediators is thus believed to be one of the main reasons that limit Co(II) mediator performances with respect to iodide. Recombination of injected electrons with the oxidized mediator can also be one of the primary loss mechanism in DSSC. The problem could be particularly serious in the case of Co(II)/(III) electron transfer mediators in which the outer sphere monoelectronic Co(III) reduction at the photoanode could be considerably more efficient than the multistep I3 reduction. The process can be probed by means of transient spectroscopy: it must be recalled that, in the absence of electron-donating electrolytes, the only pathway of dye cation recovery is recombination with photoinjected electrons, if these were captured by the oxidized mediators, the lifetime of the dye cation would be substantially lengthened, since a fraction of electrons would not be available for direct recombination anymore. In general, the more relevant is the increase in dye cation lifetime, the more efficient is the electron recapture by the oxidized mediator. From Fig. 17.15a, it is evident that I3 gives rise to the smallest increase in t2/3 of the oxidized sensitizer, meaning that the CB electron interception has the lowest efficiency within the series of mediators explored so far. On the other hand, Co(III) complexes produce a consistent increase in the recovery time of N3 þ (Fig. 17.24a–b), showing that the back electron transfer from TiO2 (or from the exposed SnO2 back contact) is more efficient than in the case of I3. It is also evident that Co(bpy)33 þ , producing a more than fourfold increase in t2/3 (6.4 ms, Fig. 17.15b) with respect to the absence of oxidized mediators, is by far more effective than
Figure 17.24 480 nm Ru(III)(H2DCB)2(NCS)2 þ recovery in the presence of (a) neat solvent (g-butyrolactone) (A), 0.1 M Co(DTB)33 þ , (C) and tetrabutyl-ammonium triiodide (B); (b) Co(DTB)33 þ (B) and Co(bpy)33 þ (A). Pulse energy 1 mJ/cm2. From Bignozzi et al., unpublished results.
550
DESIGN OF DYE-SENSITIZED SOLAR CELL COMPONENTS
Co(DTB)33 þ in recombining with photoinjected electrons. This behavior is a convincing evidence that Co(II)/(III) mediators need bulky substitutents to minimize the undesired recombination with conduction band electrons, probably via reduction of the electronic coupling between the donor states of the semiconductor and the metal-centered Co(III) d orbitals. A direct explanation for the evidence of complexes characterized by very similar redox properties leading to very different photoconversion efficiencies depending upon the fact that sterically hindering substitutents were or were not present is thus provided. On the other hand, ligand-mediated recombination can be safely excluded, since their reduction potential is generally too negative (about 1 V versus SCE) to be accessible for TiO2 conduction band electrons. Cobalt mediator performance was found to be strongly dependent upon the presence of high charge density cations, with the best results in terms of both Jsc and Voc obtained in the presence of relatively high concentration of Li þ or Mg2 þ (0.3–0.5 M). Since Li þ is not instrumental in accelerating dye regeneration, the explanation must lie in the control of the parasitic back recombination. Indeed, there is evidence that the recombination reaction at the SnO2:F contact is affected by Li þ : on a bare SnO2:F electrode of the same type used as the TiO2 current collector, the overpotential for Co(DTB)33 þ reduction is several hundred millivolts more negative in the presence of 0.25 M Li þ and 0.10 M tetrabutylammonium ion (TBA þ ) than in 0.10 M TBA þ alone. It is also reasonable that the rate of the analogous reaction on TiO2 might respond to Li þ likewise, given the approximate similarity of their surfaces. Transient spectroscopy also supported in this case the electrochemical data: the t2/3 for oxidized dye recovery in presence of Co(DTB)33 þ and Li þ 0.1 M (1.52 ms) results only marginally superior to the lifetime obtained in the absence of oxidized mediators and Li þ (1.45 ms), meaning that electronic recapture by the oxidized cobalt complex is in this case largely suppressed thanks to the electrostatic repulsion between adsorbed Li þ and Co3 þ . Adsorption of Li þ reduces the back recombination in the case of Co(III)(bpy)33 þ too, but still the oxidized dye recovery is consistently longer (t2/3 ¼ 4.4 ms) than in the neat solvent, meaning that Co(III)(bpy)33 þ continues, also in these conditions, to effectively intercept conduction band electrons. The results that have been obtained indicate that, although for now less efficient than the iodide/iodine couple, cobalt complexes formed with relatively simple and commercially available alkyl-substituted polypyridines are promising electron transfer mediators for DSSCs. Their kinetic limitations reside essentially in a slower oxidized dye regeneration and in a faster recombination with photoinjected electrons: however, while the back recombination can be passivated by choosing an appropriate electrolytic composition (i.e., Li þ , Mg2 þ ions, tert-butylpyridine) and by the introduction of bulky substitutents in the coordination sphere of the dye and/or of the mediator, the slow electron donation is an intrinsic characteristic of polypyridine cobalt complexes. In the effort of improving the performance of such mediators, the possibility of using kinetically fast couples in conjunction with the best Co(II) mediators has been explored.62 Kinetically fast couples efficiently reduce the oxidized dye, but due to the fast recombination with injected electrons are totally unsuccessful as mediators in DSSC. However, when mixed with an excess of cobalt mediator, if their redox
17.7
ELECTRON MEDIATORS BASED ON COORDINATION COMPOUNDS
551
Figure 17.25 Sequence of electron transfer events involving the oxidized dye (S þ ), the comediator (D), and the cobalt complex (Co(II)).
potential is appropriate (i.e., more positive than the Co(III)/(II) couple), a cascade of electron transfer events allows to confine the hole on Co(III) (Fig. 17.16) that, by virtue of its very slow heterogeneous electron transfer on semiconductor oxides gives rise to a very inefficient electron recapture (Fig. 17.25). As a consequence, the large majority of Co(III) created in the second electron transfer step is free to diffuse to the counter electrode of the cell, whereupon is reduced. The use of a long alkyl chain dyes like Z907, for example, is beneficial to the system allowing to further reduce back recombination enhancing the electron collection efficiency. Phenothiazine (PTZ) and ferrocene (Fc), both of which have a small reorganization energy associated to the electron transfer, are the first comediators that have been considered. Each has a potential that falls between 0.22 and 0.75 V versus SCE, respectively, the potential of the Co(DTB)33 þ /2 þ and of the dye [Ru(H2DCB)(dnbpy)(NCS)2] þ /0. Because of the facile electron transfer, the photooxidized dye would be predominantly reduced by the comediator. Its oxidized form (PTZ þ and Fc þ ) can then be rapidly intercepted by Co(II), preventing the direct charge recombination between the oxidized comediator and the electrons in the TiO2. Nanosecond
Figure 17.26 Decay kinetic of photooxidized Z907 adsorbed on TiO2 in the presence of various electron mediators: (A) no mediator, (B) 0.1 M Co(DTB)32 þ , (C) 0.1 M PTZ, (D) 0.1 M Fc. Differential absorbance changes measured at 480 nm. Pulse energy: 5 mJ/cm2.
552
DESIGN OF DYE-SENSITIZED SOLAR CELL COMPONENTS
Figure 17.27 J–V curves obtained in the presence of Co(II)/PTZ (A), I/I3 (B), and Co(II)/Fc (C) in methoxypropionitrile. Inset: photoaction spectra. Incident irradiance 0.1 W/cm2.
transient absorption measurements confirmed a faster dye regeneration by both PTZ and Fc relative to Co(DTB)32 þ , despite their higher redox potential (Fig. 17.26). The best results were observed using a comediator/Co(II) ratio of 1:2 with 0.15 M Co(DTB)32 þ in acetonitrile. Addition of 0.5 M Li þ and 0.1 M Tbpy generally improved the open-circuit photovoltage. In the presence of the PTZ/Co(II) mixture cell IPCE% reached more than 80% (inset Fig. 17.27), a value well comparable to the best I/I3 cells. Under white light irradiation, the Jsc was also comparable, with the advantage of a better fill factor and higher Voc for the cobalt-based cell (Fig. 17.27). In general, the Co(II)/Fc mixture gave poorer overall photovoltaic performances than the equivalent Co(II)/PTZ system, notably a poorer Voc and fill factor. Since ferrocene was found to be faster in dye regeneration, an explanation can lie in a less efficient interception of Fc þ by Co(II) centers, resulting in a larger steady-state concentration of Fe(III) that undergoes recombination on the TiO2 surface. Indeed, chronocoulometry and cyclic voltammety experiments at FTO electrodes allowed to verify that PTZ þ was from 1.5 to 2 times more effective than Fc þ in the oxidation of Co(II), thus having a smaller probability of recombining with photoinjected electrons. An analogous behavior extends to other species having small reorganization energies and appropriate potentials such as the iron(II) complexes Fe(DMB)32 þ and Fe(DTB)32 þ (E1/2 0.95 V versus SCE). When used in the presence of an excess of Co(DTB)32 þ and in conjunction with suitable sensitizers like the heteroleptic dye Ru(dnbpy)(H2DCB)22 þ (E1/2 ¼ 1.25 V versus SCE) (Fig. 17.28), the iron(II) comediators clearly enhance the performance of the Co(DTB)32 þ and outperform the I/I3 redox couple, at least in terms of monochromatic photon to current conversion efficiency, with maximum values close to 85%.
17.7
ELECTRON MEDIATORS BASED ON COORDINATION COMPOUNDS
553
Figure 17.28 IPCE spectra of Ru(dnbpy)(H2DCB)22 þ obtained in the presence different electron transfer mediators.
Interestingly, the polypyridine Fe(II)/Co(II) mediator mixtures were rather insensitive to variations of the Fe(II) to Co(II) concentration ratio. Very similar performances were obtained over a Fe(II)/Co(II) ratio ranging from 0.1 to 0.5, with the optimum at 0.1. For higher Fe(II)/Co(II) ratios the cell performance dropped, ostensibly due to the formation of an excess of Fe(III) that was no longer effectively intercepted by Co(II) and recombined on the TiO2 surface. In the case of electron mediator mixtures characterized by the presence of a kinetically fast couple, the decrease in z potential and the blocking effect obtained by using heteroleptic dyes with hindering chains are more than ever important for controlling electron recapture: meaningfully, the performances of the Fe(II)/Co(II) mixtures used in conjunction with the homoleptic Ru(H2DCB)32 þ (where DCB is the 4,40 -dicarboxylic acid 2,20 bipyridine) were rather poor, not exceeding a maximum IPCE% of 30%, whereas the iodide/iodine couple produced almost identical conversions (ca. 40%) to those reported with the heteroleptic Ru(dnbpy)(H2DCB)22 þ . Along with cobalt complexes, some interesting results have been obtained with copper(I) coordination compounds. According to the Franck–Condon principle, electron transfer does not occur until the reactant is vibrationally excited to match the geometry of the product complex: usually this can be accomplished by a simple adjustments of bond lengths, but for copper complexes such process requires a large energy because Cu(I) and Cu(II) have different preferred coordination geometries (tetrahedral and tetragonal, respectively).63,64 In nature, electron transport systems based on blue copper proteins present optimized copper sites for a fast electron transfer, showing a distorted tetrahedral coordination geometry, intermediate between Cu(I) and Cu(II) geometries.65,66 Blue copper model complexes67 with a distorted tetrahedral geometry (Fig. 17.29) have been recently employed as efficient electron transfer mediators for DSSC. The assembly of solar cells based on copper complexes required TiO2 optimization and multilayered photoanodes with a compact TiO2 underlayer, necessary to suppress the back recombination from the exposed FTO contact. The photoelectrochemical characterization carried out with a 0.2 M Cu(I)/Cu(II) electrolyte (Cu(II) molar
554
DESIGN OF DYE-SENSITIZED SOLAR CELL COMPONENTS
11S
6R N N Cu S S NC
N
N
N N
Cu
Cu
N
N
N N
CN
[Cu(SP(mmt)]0/–
[Cu(dmp)2]2+/+
[Cu(phen)2]2+/+
Figure 17.29 Structure of distorted tetrahedral copper complexes. SP is sparteine-N,N0 , mmt is maleonitriledithiolate, dmp is 2,9-dimethyl-1,10-phenanthroline, and phen is 1,10-phenanthroline. Reprinted with permission from Ref. 67. Copyright 2005 American Chemical Society.
fraction 0.4), 0.5 M Li þ , and 0.2 M Tbpy revealed maximum IPCE ranging from 40% in the case of Cu(CuSP)(mmt) to 20% for Cu(phen)22 þ . The J–V curves (Fig. 17.30) obtained under simulated solar irradiation basically followed the same trend as observed under monochromatic light. Cu(phen)22 þ was the worst mediator due to the slow electron donation consistent with its small self-exchange rate constant (0.20 M1 s1). The more rigid coordination sphere of Cu(SP)(mmt) and Cu(dmp)22 þ allowed for a faster electron transfer and higher self-exchange rates (respectively, 40 and 100 times higher than Co(phen)22 þ ) beneficial for improving the electron transport within the electrolyte. The lower photocurrent observed for Cu(dmp)22 þ (E1/2 ¼ 0.66 V versus
Figure 17.30 J–V curves of DSSC containing (a) Cu(SP)(mmt)0/, (b) Cu(dmp)22 þ / þ , and (c) Cu(phen)22 þ / þ as the redox couples. Input power 0.1 W/cm2. Reprinted with permission Ref. 67. Copyright 2005 American Chemical Society.
17.7
ELECTRON MEDIATORS BASED ON COORDINATION COMPOUNDS
555
SCE) with respect to Cu(SP)(mmt) (E1/2 ¼ 0.29 V versus SCE) can be explained by a slower dye regeneration determined by its smaller driving force for dye reduction, but the drawback is compensated by a larger Voc (0.79 V) and a better fill factor. This is particularly relevant since the dmp ligand is commercially available and Cu(dmp)22 þ can be easily prepared in a single step. Thus, the use of copper complexes with a distorted tetragonal geometry, in which the structural change between copper(I) and copper(II) complexes is minimized, shows promise for developing alternative low-cost mediators for photoelectrochemical cells. The electrochemical properties of another series of Cu(I) complexes, based on substituted bipyridine and quinoline derivatives, have been also investigated68 (Fig. 17.31). To stabilize the Cu(I) oxidation state of Cu(I) polypyridine complexes, electron-withdrawing substitutents like esters have been considered. The same effect was also obtained with pyridyl-quinoline and biquinoline complexes, thanks to the increased p-accepting properties of the quinoline condensed aromatic ring. Oxidation of a tetrahedral Cu(I) complex to Cu(II) involves the removal of an electron from a metal-centered T2 orbital with a prevailing antibonding character. The electrochemical study showed that almost all of the Cu(I) complexes had a sufficiently negative oxidation potential to guarantee an ample driving force for reduction of most part of oxidized sensitizers (Table 17.1). The sole exception was constituted by [Cu(BQ)2] þ (6) that, due to the strong stabilization of metal-centered orbitals induced by the augmented back bonding to four quinoline rings, exhibited an E1/2 only 100 mV negative of the dye Z907. In general, Cu(I) complexes give rise to slow electron transfer processes, as shown by the peak separations, that, in all cases, largely
ROOC
+
+
+ COOR N
N
N
Cu N
N
N
Cu N
N
ROOC
N Cu
N
N
N
COOR [Cu(bpy-(COOEt)2)2]+ (1) [Cu(bpy-(COOnBut)2)2]+ (2) [Cu(bpy-(COOTbut)2)2]+ (3)
[Cu(PQ)2]+ (4)
[Cu(MeTbPQ)2]+ (5)
+
N
N Cu
N
N
[Cu(BQ)2]+ (6)
Figure 17.31 Structure of Cu(I) complexes employed as electron transfer mediators in photoelectrochemical cells.
556
DESIGN OF DYE-SENSITIZED SOLAR CELL COMPONENTS
TABLE 17.1 Electrochemical Properties of Cu(I) Mediators in ACN/LiClO4 0.1 M, 100 mV/s Recorded Using a Glassy Carbon Working Electrode Complex (1) (2) (3) (4) (5) (6) I/I3 Z907
E1/2 (mV versus SCE)
Anodic Peak (mV versus SCE)
445 350 350 350 280 688 466 710
576 516 998 410 342 750 716 757
Cathodic Peak (mV versus SCE) 214 187 302 292 226 627 216 663
Peak Separation (mV) 362 329 1300 120 120 123 500 94
exceed the 60 mV expected for an ideally reversible monoelectronic process. It is however clear from the data reported in Table 17.1 that pyridyl-quinoline-based complexes exhibit a markedly different electrochemical behavior with respect to the bipyridine-based ones. Indeed, the former showed well-defined diffusion-limited peaks, with a peak separation (without compensation of cell resistance) of the order of 120 mV, while the latter showed generally broad poorly resolved peaks with separations higher than 300 mV. Exceptionally slow is the quasi-reversible process associated with the oxidation of (3) with a peak separation of the order of 1.3 V, probably caused by the presence of bulky tert-butyl ester groups that contribute to decouple the metal center from the electrode. As a first approximation, neglecting specific interactions with the electrodic surface (which are however possible), such, quite unexpected, difference is ostensibly related to the geometry change that accompanies the oxidation of Cu(I) to Cu(II): in case of Cu(I) pyridyl-quinoline complexes, a smaller inner sphere reorganization energy probably contributes to limit the activation barrier for the electron transfer. As can be observed from Fig. 17.32, the best performances were obtained with mediators (1) and (5), generating an almost identical maximum IPCE% of the order of 35%, about one-half of that obtained with the classical I/I3 couple. Copper-based couples (2), (3), and (4) led to lower performances with maximum IPCE varying between 30% (4) and about 20% (3). The different IPCEs obtained with the various copper(I) mediators can only be explained in terms of different electron collection efficiencies (h): at the anodic compartment of the cell, losses due to electronic recombination with the oxidized sensitizer and with acceptor states of the electrolyte (namely, Cu(II) centers) limit the number of electrons that are able to flow out of the cell as a photocurrent. In addition, an inefficient Cu(I) regeneration at the counter electrode may lead to an increase of the steady concentration of Cu(II), ultimately increasing the probability of photoinjected electron recapture. First, we observe that within the pyridyl-quinoline-based Cu(I) series, the alkyl-substituted [Cu(MeTbPQ)2] þ complex allowed to obtain a maximum IPCE about 20% higher than the unsubstituted [Cu(PQ)2] þ mediator, indicating a possible beneficial effect of the alkyl chains on the mediator properties. Besides shifting the metal-centered
17.7
ELECTRON MEDIATORS BASED ON COORDINATION COMPOUNDS
557
Figure 17.32 Photoaction spectra obtained in presence of Cu(I)/Cu(II) electron mediators using regenerative sandwich cells equipped with gold counter electrodes. (5) Stars, (1) open circles, (4) open triangles, (2) solid triangles, (3) open squares, and (6) solid circles. Data compared with the I/I3 couple (black squares). Li+ 0.5 M was added to all copper-based electrolytes.
oxidation toward more negative values (280 mV for (5), 350 mV for (4)), slightly increasing the driving force for dye regeneration, the presence of bulky tert-butyl chains might partially block the back recombination, decreasing the electronic coupling between Cu(II) acceptors and electronically occupied TiO2 donor states. Considering the substituted bipyridine series, the performances decrease in the order (1) > (2) > (3), suggests that the increased steric hindrance of the ligands does not improve the overall mediator efficiency, and on the contrary, it has a detrimental effect. This finding could be surprising at first; however, the electrochemical study showed that 4,40 -disubstituted bipyridine complexes are intrinsically characterized by very slow electron transfer kinetics and bulky substitutents might not be able to control recombination any further. On the other hand, their presence might introduce an additional, unnecessary, barrier that may affect negatively both the regeneration of the oxidized dye and the electrochemical response of the counter electrode. Finally, the low conversion efficiencies obtained with [Cu(BQ)2] þ (6) can be quite easily explained by the exceedingly positive oxidation potential that does not allow for an efficient dye regeneration. In general, the modest efficiencies of Cu(I)/(II) redox couples compared to the iodide/triiodide can be explained by a slow dye reduction, which could be reasonably anticipated, given the slow kinetics that are associated to the Cu(I)/(II) redox chemistry. On the other hand, using a suitably built photoanode equipped with a compact TiO2 underlayer and an appropriate heteroleptic dye like Z907, the detrimental electronic
558
DESIGN OF DYE-SENSITIZED SOLAR CELL COMPONENTS
recapture by Cu(II) could be largely controlled, as demonstrated by copper mediator’s higher Voc’s (550 mV for (2) and (5), 400 mV for I/I3 under 510 nm monochromatic irradiation), lower dark current, and larger fill factors (63% versus 58%) with respect to the I/I3 redox couple. Monochromatic quantum conversion efficiencies of the order of 35–40% are however still too low to be considered of practical interest for replacing the classical I/I3 couple, but as the understanding of the relationship between mediator structure, redox properties, and interfacial electron transfer kinetics progresses, the rational design of new more efficient relays with the desired kinetic properties grows nearer.
17.8 MASS TRANSPORT LIMITATIONS USING COORDINATION COMPOUNDS AS REDOX MEDIATORS The transport of the electroactive ions is expected to play a significant role in determining DSSC efficiency: this is, particularly true under strong illumination when a large number of photooxidized dye molecules are simultaneously generated at the photoanode and an efficient turnover of electron-donating species is required to sustain the photocurrent. Iodide/triiodide species are relatively simple and small ions that can diffuse efficiently in solution and within the TiO2 mesopores, whereas larger complex ions like the polypyridine Co(II)/(III) can experience mass transport limitations due to their larger hydrodynamic radius and their positive charge, which leads to electrostatic attraction toward the TiO2 surface and reduces ion mobility. Chronoamperometry experiments at a platinum electrode under Cottrell conditions19 in acetonitrile/LiClO4 allowed to calculate a diffusion coefficient for Co(DTB)32 þ equal to 9.8 106 cm2/s. More recently, rotating disk experiments substantially confirmed the former value, yielding in similar conditions a value of 6.7 106 cm2/s. To make a comparison, iodide was found to have a substantially larger diffusion coefficient, of the order of 1.6 105 cm2/s. It must be noted that the diffusion coefficients have been obtained in different conditions (i.e., diluted Co(II) solutions) from those actually met in the concentrated electrolyte employed for the DSSC. Under working conditions, a large increase (11) in solvent (acetonitrile) viscosity was found for the cobalt based mediators, whereas iodide/iodine solutions accounted only for a 5 increase. Considering this effect, the bulk diffusion of I and I3 can be evaluated as about six times faster than Co(DTB)32 þ /3 þ . Furthermore, into the pores of the TiO2 the slow diffusion of the cobalt mediator can be exacerbated by channel constrictivity effects on the large cations and by electrostatic interactions with the surface. Indeed, when using Co(II)/(III) couples, mass transport limitations are evident from the lack of linearity of the photocurrent versus incident power plots (Fig. 17.33): in the presence of a spacer, the photocurrent density produced by the cobalt-mediated cell is very close to the iodide/iodine system at low irradiation intensities (0.018 and 0.033 W/cm2), but clearly approaches a plateau at incident light intensities superior to 0.07 W/cm2. On the other hand, the Jsc delivered by the iodide/iodine-mediated cells continues to increase in almost linear fashion with the incident power.
17.8 MASS TRANSPORT LIMITATIONS USING COORDINATION COMPOUNDS
559
Figure 17.33 Jsc versus incident irradiance plot: LiI/I2 0.3/0.03 M þ 0.1 M Tbpy in acetonitrile in conjunction with N3 (squares): LiI/I2 0.3/0.03 M þ 0.1 M Tbpy in acetonitrile in conjunction with Z907 (circles); Co(II)(DTB)32 þ 0.15 M þ 0.5 M Li þ þ 0.1 M Tbpy in acetonitrile in conjunction with Z907 (triangles); 120 mm spacer. From Bignozzi et al., unpublished results.
The limiting role of the diffusion in case of Co(II) electrolytes could be further investigated by recording photocurrent transients under strong illumination (ca. 0.1 W/cm2) (Fig. 17.34). The time-resolved experiment clearly shows an instantaneous rise of the photocurrent as soon as the cell is exposed to light, the appearance of a peak of about 6 mA/cm2, followed by a relaxation to a steady regime within about 10 s. A constant photocurrent
Figure 17.34 Photocurrent transients recorded on (a) Co(DTB)32 þ -mediated DSSC. Electrolyte composition: Co(II) ¼ 0.15 M, Li þ ¼ 0.5 M, and Tbpy ¼ 0.1 M. (b) LiI/I2 0.3/0.03 M 0.1 M Tbpy in acetonitrile. From Bignozzi et al., unpublished results.
560
DESIGN OF DYE-SENSITIZED SOLAR CELL COMPONENTS
density of nearly 3 mA/cm2 is then maintained as long as the cell is left under illumination. The appearance of a photocurrent peak followed by a decay to a steady value could be explained either by electron recombination after the initial charge injection, or by a slow Co(DTB)32 þ diffusion unable to sustain the amount of photocurrent initially produced. If the transients were determined by the former process, a much faster drop to a steady value would be observed since charge recombination usually occurs in the microsecond/millisecond time domain. Furthermore, time-resolved spectroscopy allowed to verify that with an appropriate dye and electrolytic composition charge recombination is not a limiting process with respect to the iodide/iodine couple. As a consequence, the observed behavior could be quite confidently assigned to a mass transport limitation of the cell. Indeed, the same phenomenon is not observed with iodide/triiodide-mediated cells, which show under identical irradiation a more ideal rectangular shape of the photocurrent transients (Fig. 17.34b). The limiting role of cobalt diffusion can be confirmed by comparing photocurrent transients obtained from cells equipped with an electrolyte-confining spacer (120 mm) to those recorded with cells assembled by pressing the TiO2 photoanode and the counter electrode in a direct contact (Fig. 17.35a and b). In the presence of the spacer (Fig. 17.35a), an initially high photocurrent value (6 mA/cm2) is achieved, but, due to the larger spacing between the two electrodes, the diffusion of the electron mediator is not fast enough to supply new reduced mediator to the TiO2/dye interface from which, under irradiation, is constantly depleted. Thus, a steady photocurrent value, significantly lower than the initial spike, is attained after a few seconds. In Fig. 17.35b, the reduced diffusional path for the electron mediator allows for a more effective mass transport that accounts for the generation of a stable photocurrent without the observation of photoanodic relaxation processes. Cell and TiO2 engineering are therefore required to avoid or reduce the mass transport limitations in DSSCs based on coordination compounds as redox
Figure 17.35 Photocurrent transients obtained with (a) spacer equipped cells and (b) without spacer. Electrolyte composition: 0.15 M Co(II)(DTB)32 þ þ 0.5 M Li þ þ 0.1 M Tbpy in acetonitrile. From Bignozzi et al., unpublished results.
17.8 MASS TRANSPORT LIMITATIONS USING COORDINATION COMPOUNDS
561
mediators: minimizing the spacing between the electrodes, choosing low viscosity solvents, and changing the morphology of the TiO2 substrate (i.e., ordered nanostructures with large pore size, TiO2 nanotubes, and nanorods) are expected and required to enhance mediator transport and cell efficiency. Thus, the search for simple small electroactive ions, free of serious diffusional limitations, is attractive, but so far there are only a few examples of successful species: Oskam et al.69 evaluated pseudo-halides like SCN/(SCN)2 and SeCN/(SeCN)2 in association to the N3 sensitizer; however, their performances were rather poor, with monochromatic photoconversion efficiencies of the order of 20%. More recently, a very efficient electron mediator based on a low viscosity ionic liquid, 1-ethyl-3-methylimidazolium selenocyanate (EMISeCN), has been reported70: such species exhibit an high conductivity, about 30 times higher than the most fluid iodide-based analog (PMII), it is transparent to the visible radiation, has an appropriate potential for dye reduction, and a satisfactory electrochemical behavior, resulting almost inactive on FTO and TiO2 surfaces, meaning that the unwanted electron recapture requires high overpotentials to occur (Fig. 17.36). Dye regeneration was also very satisfactory, being even faster than the electron donation from iodide ions in the PMII ionic liquid. Thus, the photovoltaic performance obtained with state-of-the art multilayered titania films and the Z907 dye appeared to be very promising, leading to an almost quantitative photon to electron conversion in the 470–580 nm region and to an overall conversion efficiency of about 8% under simulated sunlight, a result of absolute relevance, being comparable to the best optimized iodide-mediated cells. EMISeCN is however relatively expensive, being prepared by metathesis from silver salts, selenium has a low natural abundance and there can be issues related to its toxicity. To date, after 4 years from the initial
Figure 17.36 Cyclic voltammograms of (a) naked and (b) platinized conductive glass (FTO) in the presence of 5 mM EMISeCN and 5 mM K(SeCN3) in acetonitrile/0.1 M TBAPF6. Reprinted with permission from Ref. 70. Copyright 2004 American Chemical Society.
562
DESIGN OF DYE-SENSITIZED SOLAR CELL COMPONENTS
report, it does not appear to have replaced the iodide/iodine mediator to any significant extent both in DSSCs industrialization efforts and in other academic studies in the field of regenerative photoelectrochemical cells.
17.9
SOLID-STATE HOLE CONDUCTORS
In the most efficient dye-sensitized solar cell, a porous nanocrystalline TiO2 is in contact with a liquid electrolyte and, as already seen, the key processes take place at the interface between the nanostructured film and the liquid electrolyte. However, the electrolyte may degrade over a period of time due to intrinsic chemical instability of the redox species, solvent evaporation, or leaking caused by sealing imperfections. The search for suitable solid materials that can replace the liquid electrolyte is therefore an interesting and active area of research. In a solid-state DSC, the solid hole conducting material captures the holes (the positive charges left on the dye as a result of the photoinduced charge injection in the n-type material) and closes the circuit with the counter electrode. Solid hole conductors include conducting polymers, organic hole conductors and inorganic semiconductors such as CuI and CuSCN.71 Among these materials, the copper(I) compounds have shown the most promising results as hole conducting materials for their application to regenerative photoelectrochemical cells. CuI is a p-type semiconductor with a bandgap of 3.1 eV that can be deposited by spray or dip coating from an acetonitrile solution (typically, 3 g of CuI in 100 mL of acetonitrile) onto dye-sensitized substrates, using a low-temperature deposition technique that avoids the denaturation of the dye monolayer. The casting procedure can be repeated to ensure an optimal electrical contact and TiO2 pore filling by the p-type material. A recent detailed photoemission study72 showed that the anatase/CuI interface is a type II heterojunction with a substantial space charge that allows for an efficient unimpeded electron transfer from the p-type material to the n-type semiconductor. Indeed, in one of the first reports about solid-state DSC, Tennakone et al. obtained a global efficiency of about 0.8% using a natural cyanidine dye, CuI and gold or graphite counter electrodes (either directly evaporated over the hole conductor or simply pressed over it).71 A substantial drop in cell performance associated to an increased CuI film resistivity in the presence of a moisture-rich environment indicated CuI film degradation. In general, CuI-based DSC are not stable and, even stored dry in the dark, undergo a rapid decay. One of the reason of such instability seems to reside into the loosening of the electrical contact between the dyed TiO2 surface and CuI crystallites. CuI deposited from acetonitrile solution produces large crystallites (10 mm) that do not penetrate well into the TiO2 pores and form loose contacts. It has been found73 that the stability and the response of CuI-based DSC can be greatly enhanced by adding small quantities (103 M) of 1-methyl-3-ethyl-imidazolium thiocyanate (MEISCN) ionic liquid that acts as a powerful crystal growth inhibitor, probably through strong surface adsorption, leading to a 1000 reduction of CuI crystal size. Such effect permits a better permeation of the nanocrystalline TiO2 matrix and a strengthening of the coupling between the dye and the electron donor. In addition, the ionic liquid that remains at the CuI grain boundaries after solvent
17.9
SOLID-STATE HOLE CONDUCTORS
563
Figure 17.37 I–V characteristics of the cell made by deposition of CuI from an acetonitrile solution containing 9 103 M EMISCN. Inset: IPCE spectrum. Reprinted with permission from Chem. Mater.2002, 14, 3, 955. Copyright 2002 American Chemical Society.
evaporation seems to admit hole conductance, possibly thanks to the hole-accepting properties of SCN. The photoresponse obtained in the presence of CuI/EMISCN shows an overall efficiency of about 3% and a maximum monochromatic conversion of about 60%. After 10 days, however, the global efficiency of the EMISCN/CuI cell dropped to the 74% of the initial h value, and at the same time an equivalent cell without EMISCN dropped to 11% (Fig. 17.37). A secondary problem involving CuI may arise from Cu þ migration that tends to decrease the open-circuit photovoltage of the cell. For this reason, other high bandgap p-type materials like CuSCN may be more suitable. CuSCN couples transparency, good conductivity (103 W1 cm1), and fast hole injection kinetics. In the hexagonal CuSCN, each Cu(I) is surrounded by three sulfurs and one nitrogen atom, inducing a good stabilization of the Cu(II) form that leads to a facile electron donation to the oxidized dye. CuSCN can be electrodeposited onto the sensitized photoanode by electrochemical techniques74 (potentiostatic stepping from 0.3 to 0.8 V versus SCE in ethanolic Cu(ClO4)2 and LiSCN solution) or by chemical impregnation by using a solution of CuSCN in propylsulfide under moderate heating (ca. 80 C). The electrical back contact to CuSCN can be indifferently made by evaporated gold or pressed graphite (Fig. 17.38). O’Regan et al.75 analyzed the J–V characteristics of a series of solid-state cells built with commercially available TiO2 (Degussa P25) fabricated by casting the CuSCN over the N3-sensitized TiO2 using the wet impregnation method (CuSCN/propysulfide). The performances were found dramatically dependent on the drying of the cell
564
DESIGN OF DYE-SENSITIZED SOLAR CELL COMPONENTS
Figure 17.38 Schematic of a solid-state dye-sensitized photovoltaic cell. Reprinted with permission from Ref. 75. Copyright 2002 American Chemical Society.
and on their aging conditions (Fig. 17.39). Before vacuum treatment (drying), the cells showed high series resistance and modest short-circuit photocurrents. In most cases, several additional days (up to 14) of storage in dry conditions (under argon) were required to achieve optimal performances. Probably, the removal of remaining impurities of water and propylsulfide decreases the hole trap density in CuSCN and improves its conductivity. Another negative effect arising from hole trapping is the quenching of the excited state of dye by nearby holes, which indeed bear some similarities to Cu(II) ions in solution (Cu(II) is an effective quencher of the N3 luminescence). The stability in outdoor conditions has not been deeply investigated, but it seemed to be dependent upon drying history, moisture, and presence of UV light. More recently, TiO2/CuSCN heterostructures have been used in multilayer solidstate cells characterized by a broader spectral response. It has been shown that the electron injection from a photoexcited dye can occur through ultrathin layers
17.9
SOLID-STATE HOLE CONDUCTORS
565
Figure 17.39 Dark and photo I–V characteristics of a TiO2/N3/CuSCN cell under argon storage under 850 W/m2 irradiation. Reprinted with permission from Ref. 75. Copyright 2002 American Chemical Society.
(ca. 1 nm) of insulating or semiconducting materials via a ballistic or tunneling mechanism that circumvent the dissipative processes, which usually occurs when an electron transits across a surface. This ability to release energetic charge carriers can be exploited to build multijunction photovoltaic cells incorporating several pigments that enhance the spectral responsivity and cell efficiency. As an example, a solid-state CuSCN device comprising spectrally complementary organic dyes (fast green (FG), D1 and acridine yellow (AY), D2) show a marked enhancement over cells assembled with the individual pigments (Fig. 17.40). The anionic dye FG strongly chelates the TiO2 surface, absorbs in the red region of the spectrum, and injects an electron into the CB of the TiO2 and a hole into the CuSCN. AY readily adsorbs onto the CuSCN surface, absorbs in the blue region of the spectrum, and its ground and excited states allow for a hole and an electron injection into the second thick CuSCN layer and TiO2, respectively. Charge separation in such cell configuration involves tunneling of both holes and electrons through the CuSCN/ D2 barrier to reach the CuSCN anode. Indeed, the optimum thickness of the CuSCN barrier was found to be 2–3 nm. When it becomes too thick, exceeding 15 nm, the photovoltaic effect of the dye D2 becomes nearly undetectable. Low cost alternatives to inorganic p-type semiconductors can be found in organic species and conductive polymers. Organic hole conductors like the spiro-compound 2,20 ,7,70 -tetrakis(N,N-di-p-methoxyphenyl-amine)-9,90 -spirobifluorene (OMeTAD) (Fig. 17.41) have demonstrated some promise for application in dye-sensitized solar cells.76
566
DESIGN OF DYE-SENSITIZED SOLAR CELL COMPONENTS
Figure 17.40 Schematic energy-level diagram of the TiO2/D1/p-CuSCN/D2/p-CuSCN cell. Reproduced with permission from J. Phys. Chem. B 2003, 107, 13758. Copyright 2003 American Chemical Society.
The hole conductor has a spiro-center (a tetrahedral carbon linking two aromatic moieties) that is introduced in order to improve glass forming properties and prevent crystallization. Crystallization is undesirable since it impairs the formation of a good electrical contact between the TiO2 surface and the hole transporting
Figure 17.41
Structure of OMeTAD.
17.9
567
SOLID-STATE HOLE CONDUCTORS
material (HTM), which is cast over the dye-sensitized surface by spin coating from a chlorobenzene solution. A compact TiO2 underlayer is required to prevent the short circuiting between HTM and FTO collector, and an evaporated gold layer forms the cathode of the solid-state DSC, allowing for an intimate contact with the organic hole conductor. The oxidized dye Ru(III)(H2DCB)2(NCS)2 is efficiently regenerated by hole injection in the HTM layer on a nanosecond timescale, as demonstrated by the total disappearance of the Ru(III) bleaching, being compensated by the rise of the oxidized OMeTAD signal (lmax ¼ 530 nm) within the 6 ns laser excitation pulse. Monochromatic photon to electron conversion as high as 30% and overall efficiencies of 0.74% under ca. 0.01 W/cm2 were obtained by using additives such as N(PhBr)3SbCl6 and Li(CF3SO2)2N): N(PhBr)3SbCl6 acts as a dopant, introducing free charge carriers in the HTM by amine oxidation; the second additive, a lithium salt, is a source of Li þ cations, which are known to be potential determining ions for TiO2 and assist in electron injection while retarding back recombination. The lithium salt may also compensate for space charge effects: under illumination a net positive space charge is expected to be formed in the HTM that impairs current flow. The salt could screen this field, eliminating space charge control of the photocurrent, leading to Jsc of the order of 3 mA/cm2 under full A.M. 1.5 conditions. Conductive polymers based on polythiophenes and polypyrroles could be interesting candidates for replacing the liquid electrolyte in DSSC, due to their low cost, thermal stability, and good conductivity.77 Thanks to these properties, these systems have already found application in the OLED technology as charge transporting matrices. PEDOP (poly-ethylene-dioxy-pyrrole) can be deposited on a sensitized TiO2 photoanode via a photoassisted electropolymerization78 initiated by the purpose-built ruthenium complex [Ru(H2DCB)(pyrr-bpy)2]2 þ 79 (Fig. 17.42): the oxidized dye, [RuIII(H2DCB)(pyrr-bpy)2]3 þ , created after the charge injection into TiO2, is a sufficiently strong oxidizer, with a potential of 1.19 V/SCE, to oxidize EDOP monomers and induce their cationic polymerization, leading to the growth of
N
N
2+ pyrr-bpy = CO2H
N N
N
N
2PF6-
Ru N
CO2H
HO2C
N N
N
N
H2DCB = CO2H
N
N
[Ru(H2DCB)(pyrr-bpy)2]2+
Figure 17.42
Ru(II) complex uses for in situ photoassisted PEDOP growth.
568
DESIGN OF DYE-SENSITIZED SOLAR CELL COMPONENTS
Figure 17.43 J–V characteristic obtained on PEDOP-based DSSC in the presence of Li þ 0.3 M; 10 s polymerization time.
polymeric chains chemically attached to the dye with a possible advantage in terms of electronic coupling and charge transfer rate between electron donor (PEDOP) and acceptor (oxidized dye). The J–V characteristics of solid-state PEDOP cells indicate a strong recombination and an high serial resistance, affecting the fill factor (Fig. 17.43). The dark current corroborated the indication of an efficient recombination showing a marked increase in its magnitude as the direct bias approached 0.4 V, probably reaching the onset of PEDOP reduction, suggesting that as the potential increased toward negative values, higher lying states of the TiO2 were filled promoting recombination with PEDOP holes. The fact that oxidized/doped PEDOP was directly photoelectrochemically deposited onto TiO2 probably enhanced the electron/hole recombination pathways, since photoinjected electrons while diffusing to the back contact had a certain probability of reaching the surface of the nanoparticles in close proximity of PEDOP acceptor states. While the dark current suggested a possible photocurrent loss mechanism through PEDOP reduction, a relevant recombination with the oxidized dye could not be ruled out. Dark current simply could not provide information on the latter process since, in the dark, after equilibrium was attained, no oxidized Ru(III) centers should be present at the TiO2/dye/polymer interface. Indications of efficient recombination was also gained from cell photocurrent transients under short-circuit conditions, showing cathodic spikes due to recombination of photoinjected electrons with acceptor states at the TiO2 surface, namely, PEDOP holes and oxidized Ru(III) centers. The cathodic current dropped to zero on a very short timescale, suggesting that recombination was kinetically fast, at least for the timescale (0.1 s sampling time) of the experiment. Passivation of the TiO2 surface via the application of a nanometric insulating Al2O3 overlayer80 was instrumental in
17.9
SOLID-STATE HOLE CONDUCTORS
569
Figure 17.44 Photoaction spectra of a TiO2/dye/PEDOP/Au cell (white squares) compared to a TiO2/Al2O3/dye/PEDOP/Au system (black squares). From Bignozzi et al., unpublished results.
reducing the electronic recapture, allowing for an IPCE% doubling (from 3% to 6%) (Fig. 17.44). Cell Jsc and Voc measured under 0.05 W/cm2 white light irradiation were also significantly improved (from 50 to 100 mA/cm2 and from 0.4 to 0.7 V, respectively). Besides strong back recombination and despite the intimate contact between the dye and the hole transporting layer, the low efficiencies could also be explained by a nonoptimal hole injection into the PEDOP matrix, as evidenced by the slow oxidized dye recovery observed at 450 nm (Fig. 17.45). A better photovoltaic response has recently been obtained by using the excellent properties of PEDOT as a hole transporting material. As previously described, PEDOT can be in situ photoelectropolymerized from bis-EDOT monomers by exploiting the oxidizing power of a dye like Z907.81 The hydrophobic properties of Z907 allow for a good affinity for the scarcely polar polymeric matrix, ensuring a good electrical contact between the p-type material and the dye. The presence of a coadsorbate like deoxycholic acid (DCA) in an optimal 2:1 ratio with respect to the dye enhanced the global photovoltaic response by reducing dye aggregates. This function seems to be particularly relevant for polymer-mediated DSC, since aggregated dye clusters reduce the number of available polymer growth sites and leave less effective space for polymerization initiation and propagation. To optimize the contact between the hole transporting layer and the counter electrode, a PEDOT-functionalized FTO was used as the cathode, and to achieve optimal performances, a drop of BMImTFSI (1-butyl-3-methylimidazolium bis(trifluoromethanesulfonyl)imide) containing 0.2 M LiTFSI and 0.2 M Tbpy was cast onto the TiO2/dye/PEDOP junction. The ionic liquid may improve the charge transporting capabilities of the heterointerface through screening of space charge effects, lithium cations assist electron injection and percolation through the n-type semiconductor, and, as usual,
570
DESIGN OF DYE-SENSITIZED SOLAR CELL COMPONENTS
Figure 17.45 Ru(III) recovery in the presence of no mediator (C), photoelectropolymerized PEDOP (B), and electrolytic 0.1 M Co(II)(DTB)32 þ (A). Differential absorbance changes observed at 450 nm. Pulse energy 10 mJ/cm2, 532 nm excitation wavelength. From Bignozzi et al., unpublished results.
Tbpy increases the open-circuit photovoltage via suppression of the back recombination. With such a treatment, FTO/TiO2/dye/PEDOT–PEDOT/FTO sandwich cells afforded efficiencies of the order of 2.6%, one of the highest results so far recorded with solid-state DSCs based on hole conducting polymers (Fig. 17.46).
Figure 17.46 I–V curves of the solid-state TiO2/dye/PEDOT–PEDOT/FTO solar cells in the presence of the BMImTFSI electrolyte. The improvement of the cell performances over time is ostensibly due to a better percolation of the ionic liquid into the TiO2 mesopores. Reprinted from Ref. 81. Copyright 2008 with permission from Elsevier.
17.10
17.10
CATALYTIC MATERIALS FOR CATHODES OF DSSCs
571
CATALYTIC MATERIALS FOR CATHODES OF DSSCs
A kinetically fast reduction of the oxidized redox mediator at the cathodic compartment of the cell is required to maintain a sufficiently high concentration of electron donor for ensuring an efficient dye regeneration. With the I/I3 couple, the use of a catalytic platinum coating is almost mandatory to achieve a satisfactory electrochemical kinetics, but the use of redox mediators alternative to the I/I3 couple opens the way to the search for new, less expensive, more flexible and available catalytic materials and to their application in new cell configurations. Many electrode materials (carbon, gold, platinum) function adequately in the case of Co(II)/(III) mediators, but are generally opaque (Fig. 17.47). In contrast, optically transparent conductive oxide (TCO) are very poor cathodes for the Co (DTB)32 þ /3 þ couple, unless they are chemically modified by surface chemisorption of certain metal complexes. This allows to obtain optically transparent cathodes with a good electrochemical response that can be employed to build stacked cells
Figure 17.47 In parallel connected stacked cell based on the Co(II)/(III) couple. 1 is a transparent FTO cathode modified via chemisorption of an electroactive molecule. 2 is a conventional gold cathode. The red and the blue dyes, absorbing in complementary spectral regions, result in a panchromatic sensitization.
572
DESIGN OF DYE-SENSITIZED SOLAR CELL COMPONENTS
either serially or in parallel connected in which, for example, two spectrally complementary dye can work in their optimal absorption region, improving the spectral responsivity of the modules. In the first study,82 it was demonstrated that ITO and FTO electrodes modified by irreversibly adsorbing a monolayer of Fe(H2DCB)32 þ electrocatalyze efficiently the Co(DTB)32 þ oxidation via an EC0 mechanism. However, while electrogenerated Fe (III) has an ample driving force for oxidizing Co(II), Fe(II) is a thermodynamically too weak reductant for catalyzing Co(III) reduction, and the Fe(II)/(III) couple is of no value for building transparent cathodes. Nevertheless, it was found that the electrocatalysis of the Co(II)/(III) redox chemistry promoted by surface-bound electroactive species is a relatively general phenomenon and a number of molecular species with appropriate potential can promote the desired catalytic effect. For example, the osmium complex Os(H2DCB)2Cl2 can be strongly adsorbed on FTO and, with its potential (310 mV versus ferrocene) slightly negative of the Co(II)/(III) couple, proved to efficiently catalyze both the oxidation and reduction of Co(DTB)32 þ /3 þ . Indeed, the cyclic voltammetry of Co(II)/(III) (Fig. 17.48) recorded at osmium modified electrodes showed well-defined quasi-reversible waves (DEp 140 mV), nearly identical to those obtained on gold, whereas on the unmodified FTO or ITO, it was almost impossible to observe the analogous charge transfer process. Cyclic voltammetry results were substantially confirmed by thin-layer photoelectrochemical experiments that demonstrated that the osmium modified transparent cathodes gave rise to a very small overpotential (ca. 12 mVunder cell short-circuit conditions) for Co (III) reduction, indicative of a facile heterogeneous electron transfer.83 The stacked cell configuration, employing N3 and a red absorbing cyanine dye that absorbs photons in a region where N3 does not, showed a Jsc improvement of about
Figure 17.48 CVon gold (A), FTO (B), ITO (C), and osmium modified FTO (D) and ITO (E) working electrodes in the presence of ca. 0.1 mM Co(DTB)3ClO4. Potential referred to ferrocene.
17.10
CATALYTIC MATERIALS FOR CATHODES OF DSSCs
573
Figure 17.49 Cyclic voltammery of Co(DTB)32 þ at a transparent PEDOT modified FTO. Scan speed 100 mV/s, potential referred to SCE. The smaller peak centered at about 650 mV is determined by the intrinsic polymer electroactivity. PEDOT was grown by multiple scan deposition (three scans), cycling the potential between 0.7 and þ 1.5 V versus SCE with a scan speed of 20 mV/s. From Bignozzi et al., unpublished results.
15% over a single N3 cell. It must be considered that the blue dye is a relatively inefficient dye, generating very modest IPCE% (ca. 5–8%) due to less than ideal redox properties and electronic coupling with the TiO2. Nevertheless, the results confirmed the principle, and with a better performing dye, substantial performance improvements are expected. Conductive polymers such as PEDOP and PEDOT were also found effective in promoting the Co(II)/(III) electrochemical response (Fig. 17.49). Both PEDOP and PEDOT can be electrodeposited on a transparent conductive substrate by an anodic potentiostatic (ca. þ 1.3 V versus SCE) or potentiodynamic electrolysis of the appropriate precursor solution in the presence of a perchlorate containing supporting electrolyte. Co(II)/(III) appeared to have a clearly diffusion limited electrochemical behavior on both PEDOP and PEDOT-functionalized cathodes. PEDOT gave rise to a more ideal behavior associated with a smaller separation of the catalytic waves, with a peak separation of the order of 200 mV, substantially independent from Co(II) concentration, and the slight peak shift at higher cobalt concentration being essentially determined by uncompensated cell resistance. The electrodeposition offer the advantage of a fine control of the thickness of the catalytic layer, thus optimizing the electroactive area and the transparency of the cathode. A cathode that possesses both a remarkable catalytic activity and a partially porous structure with a large active area is important for limiting the concentration
574
DESIGN OF DYE-SENSITIZED SOLAR CELL COMPONENTS
2.2 2.0 1.8 1.6
2
J (mA/cm )
1.4 1.2
Au PEDOT
1.0 0.8 0.6 0.4 0.2 0.0 0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
V (V)
Figure 17.50 J–V curves of Co(II)/(III)-mediated cell employing a gold (dashed line) and a PEDOT counter electrode (solid line). Electrolyte composition: 0.15 M Co(II)/0.5 M Li þ / 0.1 M Tbpy in acetonitrile. Cell equipped with a 120 mm spacer. Photosensitizer: Z907. From Bignozzi et al., unpublished results.
overpotential that may develop at the counter electrode of the cell under strong illumination, allowing for the consequent reduction of the photocell series resistance. An improvement in cell fill factor and global efficiency is therefore expected and found (Fig. 17.50).
17.11
CONCLUSIONS
DSSCs are photoelectrochemical solar devices, currently subject of intense research in the framework of renewable energies as a low-cost photovoltaic device. Their functioning is based on the interlacing of subsystems working in tandem: the photoanode on which the dye sensitizer is adsorbed, the electron mediator, and the counter electrode. In this chapter, we have tried to give an overview on the recent advances in the design of these solar cell components. Research on dye sensitizers is mainly focused on transition metal complexes, but considerable work is now directed toward the optimization of organic sensitizers and of natural sensitizers extracted from fruits. Concerning the electron mediators, iodide/iodine has been so far the most efficient and commonly used redox system, due to the fact that I allows for a fast regeneration of the oxidized dye. Advances have been made in gelating such electrolyte, and introducing carbon nanotubes, thus reducing the problem of cell sealing and solvent evaporation. It is remarkable that despite an increased viscosity of nanocomposite electrolytes, their conductivity has been found to be higher than the corresponding
17.11
CONCLUSIONS
575
ionic liquid, most probably thanks to a nanoscale ordering that improves the electron transport. The use of alternative redox couples as electron mediators has been addressed, so far, by a limited number of research groups but the results are promising. The research has been triggered by the fact that the electrochemical properties of coordination compounds can be easily tuned through a rational choice of the metal and an appropriate design of the coordination sphere, and their electrochemical response is sensitive to the electrodic material. The choice and design of the redox couples has been done by considering inexpensive and available metals such as the elements of the first transition row and easily synthesizable ligands. To date, the most successful attempts have been based on octahedral cobalt(II) complexes and the best results obtained in combination with specific heteroleptic complexes. The performances of such mediators have been improved by using kinetically fast couples in conjunction with Co(II) complexes. It has been in addition observed that copper complexes with a distorted tetragonal geometry show promise for developing alternative low-cost mediators for photoelectrochemical cells. The transport of the electroactive ions is expected to play a significant role in determining DSSC efficiency: this is particularly true under strong illumination when a large number of photooxidized dye molecules are simultaneously generated at the photoanode and an efficient turnover of electron-donating species is required to sustain the photocurrent. Cell and TiO2 engineering are therefore required to avoid or reduce the mass transport limitations in DSSCs based on coordination compounds as redox mediators: minimizing the spacing between the electrodes, choosing low viscosity solvents, and changing the morphology of the TiO2 substrate with ordered nanostructures with large pore size or with TiO2 nanotubes and nanorods are expected to enhance mediator transport and cell efficiency. The search for suitable solid materials that can replace the liquid electrolyte is an additional interesting and active area of research. In a solid-state DSSC, the solid hole conducting material captures the holes and closes the circuit with the counter electrode. Solid hole conductors include conducting polymers, organic hole conductors, and inorganic semiconductors such as CuI and CuSCN. Among these materials, the copper(I) compounds have shown the most promising results as hole conducting materials for their application to regenerative photoelectrochemical cells. Low-cost alternatives to inorganic p-type semiconductors can be found in organic species and conductive polymers. Organic hole conductors such as spiro-compounds and conductive polymers based on polythiophenes and polypyrroles have demonstrated some promise for application in dye-sensitized solar cells. In particular, it has been reported that the presence of ionic liquids may improve the charge transporting capabilities of the heterointerface through screening of space charge effects. In the last part of the chapter, we have described some efforts toward the modification of the counter electrode with inexpensive and transparent materials. The interest being also related to the possible realization of stacked cells, either serially or in parallel connected, in which two spectrally complementary dyes can work in their optimal absorption region, improving the spectral responsivity of the modules. In these studies, it has been found that osmium complexes as well as
576
DESIGN OF DYE-SENSITIZED SOLAR CELL COMPONENTS
electrodeposited conductive polymers such as PEDOP an PEDOT are effective in promoting the electrochemical response of Co(II) electron mediators. The subjects discussed in this chapter are complementary and moves in parallel to the development of the semiconductor nanomaterial, mainly titanium dioxide, and to the cell design and engineering. So far several prototypes, containing I/I3, where the cell components are in series interconnected have been demonstrated. The development of efficient and noncorrosive electron mediators is considered to be of particular relevance since it may allow the building of large-area modules where the different components are in parallel interconnected, increasing the single module short-circuit photocurrent and allowing for a more flexible and solar panel production.
ACKNOWLEDGMENTS Financial support from the Polo Solare Organico, CHOSE, Regione Lazio is gratefully acknowledged.
REFERENCES 1. 2. 3. 4. 5. 6. 7. 8.
9. 10. 11. 12. 13. 14. 15. 16. 17. 18.
R. Eisenberg, D. G. Nocera, Inorg. Chem. 2005, 44, 6799. J. M. Olson, S. R. Kurtz, A. E. Kibbler, P. Faine, Appl. Phys. Lett. 1990, 56, 623. J. Simon, J.-J. Andre, Molecular Semiconductors, Springer Verlag, Berlin, 1985. T. Umeda, Y. Hashimoto, H. Mizukami, T. Shirakawa, K. Yoshino, Synth. Met. 2005, 152, 93. R. Memming, Prog. Surf. Sci. 1984, 17, 7. B. O’Regan, M. Graetzel, Nature 1991, 353, 737. M. Graetzel, Inorg. Chem. 2005, 44, 6841. M. K. Nazeeruddin, P. Pechy, T. Renouard, S. M. Zakeeruddin, R. Humphry-Baker, P. Comte, P. Liska, L. Cevey, E. Costa, V. Shklover, L. Spiccia, G. B. Deacon, C. A. Bignozzi, M. Graetzel, J. Am. Chem. Soc 2001, 123, 1613. C. Bauer, G. Boschloo, E. Mukhtar, A. Hagfeldt, Int. J. Photoenergy 2002, 4, 17. Y. Fukai, Y. Kondo, S. Mori, E. Suzuki, Electrochem. Commun. 2007, 9, 1439. M. Quintano, I. Edvinsson, A. Hagfeldt, G. Boschloo, J. Phys. Chem. C 2007, 111, 1035. M. Saito, S. Fujihara, Energy Environ. Sci. 2008, 1, 280. K. Sayama, H. Sugihara, H. Harakawa, Chem. Mater. 1998, 12, 3825. R. A. Marcus, Discuss. Faraday. Soc. 1960, 29, 21. R. A. Marcus, Annu. Rev. Phys. Chem. 1964, 15, 155. T. J. Meyer, H. Taube, Comprehensive Coordination Chemistry, Vol. 1, Pergamon Press, Oxford, 1987. J. H. Alstrum Acevedo, M. K. Brennaman, T. J. Meyer, Inorg. Chem. 2005, 44, 6802. C. A. Bignozzi, S. Schoonover, F. Scandola, Molecular Level Artificial Photosynthetic Materials, Vol. 44, J. Wiley & Sons, New York, 1997.
REFERENCES
577
19. A. J. Bard, L. R. Faulkner, Electrochemical Methods: Fundamentals and Applications, J. Wiley & Sons, Inc., 2001. 20. S. Nakade, S. Kambe, T. Kitamura, Y. Wada, S. Yanagida, J. Phys. Chem. B 2001, 105, 9150. 21. S. Kambe, S. Nakade, T. Kitamura, Y. Wada, S. Yanagida, J. Phys. Chem. B 2002, 106, 2967. 22. S. Nakade, M. Saito, W. Kubo, T. Kitamura, Y. Wada, S. Yanagida, J. Phys. Chem. B 2003, 107, 8607. 23. R. Argazzi, C. A. Bignozzi, T. A. Heimer, G. J. Meyer, F. N. Castellano, J. Phys. Chem. B 1997, 101, 2591. 24. A. Kay, M. Graetzel, J. Phys. Chem. 1993, 97, 6272. 25. F. Odobel, E. Blart, M. Lagree, M. Villieras, H. Boujtita, N. El Murr, S. Caramori, C. A. Bignozzi, J. Mater. Chem. 2003, 13, 502. 26. M. K. Nazeeruddin, R. Humphry-Baker, M. Graetzel, D. Wohrle, G. Schnurpfeil, G. Schneider, A. Hirth, N. Trombach, J. Porphyr Phthalocyanines 1999, 3, 230. 27. J. J. He, A. Hagfeldt, S. E. Lindquist, H. Grennberg, F. Korodi, L. C. Sun, B. Akemark, Langmuir 2001, 17, 2743. 28. W. Paw, S. D. Cummings, M. A. Mansour, W. D. Connick, D. K. Geiger, R. Eisenberg, Coord. Chem. Rev. 1998, 171, 125. 29. A. Islam, H. Sugihara, K. Hara, L. P. Singh, R. Katoh, M. Yanagida, Y. Takahashi, S. Murata, H. Arakawa, Inorg. Chem. 2001, 40, 5371. 30. J. M. Rehm, G. L. McLendon, Y. Nasagawa, K. Yoshihara, J.-E. Moser, M. Graetzel, J. Phys. Chem. 1996, 100, 9577. 31. P. V. Kamat, W. E. Ford, Chem. Phys. Lett. 1987, 135, 421. 32. S. Das, C. S. Rajes, C. H. Suresh, K. G. Thomas, A. Ajayaghosh, C. Nasr, P. V. Kamat, M. V. George, Macromolecules 1996, 28, 4249. 33. D. Liu, P. Kamat, J. Phys. Chem. 1993, 97, 10769. 34. C. G. Garcia, A. Sarto Polo, N. Y. Murakami Iha, J. Photochem. Photobiol. A: Chem. 2003, 160, 87. 35. G. Calogero, G. Di Marco, Sol. Energy Mater. Sol. Cells 2008, 92, 1341. 36. M. K. Nazeeruddin, A. Kay, I. Rodicio, R. Humphry-Baker, E. Mueller, P. Liska, N. Vlachopoulos, M. Graetzel, J. Am. Chem. Soc. 1993, 115, 6382. 37. G. Saue, M. E. Cass, S. J. Doig, I. Lauermann, K. Pomykal, N. S. Lewis, J. Phys. Chem. B 2000, 104, 3488. 38. P. Wang, C. Klein, R. Humphry-Baker, S. H. Zakeeruddin, M. Graetzel, J. Am. Chem. Soc. 2005, 127, 808. 39. A. Hagfeldt, M. Graetzel, Acc. Chem. Res. 2000, 33, 269. 40. F. Cecchet, A. M. Gioacchini, M. Marcaccio, F. Paolucci, S. Roffia, M. Alebbi, C. A. Bignozzi, J. Phys. Chem. B 2002, 106, 3926. 41. S. Altobello, R. Argazzi, S. Caramori, C. Contado, S. Da Fre, P. Rubino, G. Larramona, C. A. Bignozzi, J. Am. Chem. Soc. 2005, 127, 15343. 42. B. A. Gregg, Coord. Chem. Rev. 2004, 248, 1215. 43. B. A. Gregg, F. Pichot, S. Ferrere, C. R. Fields, J. Phys. Chem. B 2001, 105, 1422. 44. F. Pichot, B. A. Gregg, J. Phys. Chem. B 2000, 104, 6.
578 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.
DESIGN OF DYE-SENSITIZED SOLAR CELL COMPONENTS
T. A. Heimer, E. J. Heiweil, C. A. Bignozzi, G. J. Meyer, J. Phys. Chem. A 2000, 104, 4256. I. Montanari, J. Nelson, J. R. Durrant, J. Phys. Chem. B 2002, 106, 12203. G. Schlichthorl, S. Y. Huang, J. Sprague, A. J. Frank, J. Phys. Chem. B 1997, 101, 8141. S. Y. Huang, G. Schlichthorl, A. J. Nozik, M. Graetzel, A. J. Frank, J. Phys. Chem. B 1997, 101, 2576. P. Wang, S. M. Zakeeruddin, I. Exnar, M. Graetzel, Chem. Commun. 2002, 2972–2973. P. Wang, S. M. Zakeeruddin, P. Comte, I. Exnar, M. Graetzel, J. Am. Chem. Soc. 2003, 125, 1166. H. Usui, H. Matsui, N. Tanabe, S. Yanagida, Photochem. Photobiol. A: Chem. 2004, 164, 97. W. Kubo, K. Murakoshi, T. Kitamura, S. Yoshida, M. Haruki, K. Hanabusa, H. Shirai, Y. Wada, S. Yanagida, J. Phys. Chem. B 2001, 105, 12809. P. Wang, S. M. Zakeeruddin, J.-E. Moser, M. K. Nazeeruddin, T. Sekiguchi, M. Graetzel, Nat. Mater. 2003, 2, 402. R. Kawano, M. Watanabe, Chem. Commun. 2003, 330. H. Nusbaumer, J.-E. Moser, S. M. Zakeeruddin, M. K. Nazeeruddin, M. Graetzel, J. Phys. Chem. B 2001, 105, 10461. S. Pelet, J.-E. Moser, M. Graetzel, J. Phys. Chem. B 2000, 104, 1791. C. Klein, M. K. Nazeeruddin, D. Di Censo, P. Liska, M. Graetzel, Inorg. Chem. 2004, 43, 4216. P. Wang, S. M. Zakeeruddin, P. Comte, R. Charvet, R. Humphry-Baker, M. Graetzel, J. Phys. Chem. B 2003, 107, 14336. S. A. Sapp, C. M. Elliott, C. Contado, S. Caramori, C. A. Bignozzi, J. Am. Chem. Soc. 2002, 124, 11215. P. J. Cameron, L. M. Peter, J. Phys. Chem. B 2003, 107, 14394. L. G. Wade, Organic Chemistry, Prentice Hall, Englewood Cliffs, NJ, 1991. S. Cazzanti, S. Caramori, R. Argazzi, C. M. Elliott, C. A. Bignozzi, J. Am. Chem. Soc. 2006, 128, 9996. D. B. Rorabacher, Chem. Rev. 2004, 104, 651. B. J. Hathaway, Coord. Chem. Rev. 1981, 35, 211. E. T. Adman, Adv. Protein Chem. 1991, 42, 141. E. I. Solomon, M. D. Lowery, Science 1993, 259, 1575. S. Hattori, Y. Wada, S. Yanagida, S. Fukuzumi, J. Am. Chem. Soc. 2005, 127, 9648. M. Brugnati, S. Caramori, S. Cazzanti, L. Marchini, R. Argazzi, C. A. Bignozzi, Int. J. Photoenergy 2007, Id 80756, 1–10. G. Oskam, B. V. Bergeron, G. J. Meyer, P. C. J. Searson, J. Phys. Chem. B 2001, 105, 6867. P. Wang, S. M. Zakeeruddin, J. E. Moser, R. Humphry-Baker, M. Graetzel, J. Am. Chem. Soc. 2004, 126, 7164. K. Tennakone, G. R. A. Kumara, A. R. Kumarasinghe, K. G. U. Wijayantha, P. M. Sirimanne, Semicond. Sci. Technol. 1995, 10, 1689. A. R. Kumarasinghe, W. R. Flavell, A. G. Thomas, A. K. Mallick, D. Tsoutsou, C. Chatwin, S. Rayner, P. Kirkham, S. Warren, S. Patel, P. Christian, P. O’Brien, M. Graetzel, R. Hengerer, J. Chem. Phys. 2007, 127, 114703.
REFERENCES
579
73. G. R. A. Kumara, A. Konno, K. Shiratsuchi, J. Tsukahara, K. Tennakone, Chem. Mater. 2002, 14, 954. 74. B. O’Regan, D. T. Schwartz, S. M. Zakeeruddin, M. Graetzel, Adv. Mater. 2000, 12, 1263. 75. B. O’Regan, F. Lenzmann, R. Muis, J. Wienke, Chem. Mater. 2002, 14, 5023. 76. U. Bach, D. Lupo, P. Comte, J. E. Moser, F. Weissortel, J. Salbeck, H. Spreitzer, M. Graetzel, Nature 1998, 395, 583. 77. K. R. Murakoshi, R. Y. Kogure, Y. Wada, S. Yanagida, Sol. Energy Mater. Sol. Cells 1998, 55, 113. 78. S. Caramori, C. A. Bignozzi, L. Marchini, S. Cazzanti, R. Argazzi, P. C. Gros, D. Martineau, M. Beley, Inorganica Chimica Acta 2008, 361, 627. 79. D. Martineau, M. Beley, P. C. Gros, S. Cazzanti, S. Caramori, C. A. Bignozzi, Inorg. Chem. 2007, 46, 2272. 80. E. Palomares, J. N. Clifford, S. A. Haque, T. Lutz, J. R. Durrant, Chem. Commun. 2002, 1464. 81. Y. Kim, Y.-E. Sung, J.-B. Xia, M.-L. Cantu, N. Masaki, S. Yanagida, J. Photochem. Photobiol. Part A: Chem. 2008, 193, 77. 82. C. M. Elliott, S. Caramori, C. A. Bignozzi, Langmuir 2005, 21, 3022. 83. M. J. Scott, J. J. Nelson, S. Caramori, C. A. Bignozzi, C. M. Elliott, Inorg. Chem. 2007, 46, 10071.
INDEX
Acetylcholine esterase (AChE), 320, 321 CdS modified electrode, 321 CdS nanoparticle, 320 inhibitors, 320 Acid/base-driven shuttling process, 389 Acid chloride method, 267 Adenosine triphosphate (ATP), 352 Alcohol dehydrogenase (AlcDH), 348 Alkanethiols, self-assembled monolayers, 185 Alkyl chain dyes, use, 551 Alternating current voltammetry (ACV), 287 Ambipolar mechanism, 533 Amide-disulfide zinc metalloporphyrins, 307 Amidoferrocene-terminated alkyl thiols, 18 Amidoferrocene-terminated dendrimers, 151 Amino acids, intrinsic electroactivity, 261–266 10 -Aminoferrocene-1-carboxylic acid, 266 Amperometric biosensing, 335–356 biosensors, 335 redox enzymes electrical contact, 335 Amphiphilic counterions, 404 dihexadecylphosphate anion (DHP), 404 AND/XOR gates, absorption changes, 468 Anion sensing porphyrin, 307 Au MMPC, schematic, 307 Annihilation energy, 478 Anodic stripping voltammetry (ASV), 318 Antenna effect, 121 Anthracene fluorophore, 461 Anthraquinone, 456, 467 redox behavior, 456 Antithrombin aptamer, 363
Apo-enzyme, 335, 337 reconstitution, 335, 337 Apo-glucose oxidase (apo-GOx), 338, 321, 341, 346 reconstitution, 338, 341, 346 Apo-protein(s), 337, 372 reconstitution, 372 Aprotic solvents, 14, 22, 495 acetonitrile, 14 dimethylformamide, 14 tetrabutylammonium hexafluorophosphate, 14 Aptamer, 357 preparation by SELEX protocol, 357 protein complexes, analysis, 361 substrate complexes, formation of, 361 thrombin complex, self-organization, 358 Aptasensing scenarios, 291 Aptasensors, 289 Aptamer-based sensors, See Aptasensors Aromatic short-chain peptides, 350 Alzheimer’s diphenylalamine b-amyloid, 350 Arsine ligands, 486 1,2-bis(diphenylarsino)ethane, 486 1,2-bis(diphenylphosphino)ethane, 486 Artificial antenna, 122, 124 preparation, steps, 124 requirements, 122–124 Artificial light-harvesting antenna systems, 121, 122 design of, 121 modular system, 122 requirements, 122
Electrochemistry of Functional Supramolecular Systems. Edited by Paola Ceroni, Alberto Credi, and Margherita Venturi Copyright 2010 John Wiley & Sons, Inc.
581
582
INDEX
Arylimides, 22–23 Assembling-disassembling process, 52 Association constant, 107 Association process, 61 kinetic rate constant, 61 Atomic force microscopy (AFM), 346, 469 Avogadro’s number, 62 Azobenzene configuration, 453 Azobenzene photoisomerization, 406 Back-folded isomer, 101 Bandgap, 524, 526, 562 photons transfer electrons, 524 b-Barrel model, 268 Benzylic amide macrocycle, 465 reversible shuttling of, 465 Benzylic amide rotaxane, 408, 410 electrochemically induced shuttling, 410 Bielectronic process, 383, 385, 399 Bilirubin oxidase (BOD), 346, 348 Binaphthyl boron dipyrromethane (BDP) conjugates, 466 Biocatalytic inks, 355 Biofuel cell design, 335–356 electrically contacted enzyme electrode, 340 hydrogen peroxide (H2O2), use, 341 organ-implantable miniaturized, 341 redox enzymes electrical contact with electrodes, 335 schematic configuration, 340 Biohybrid electrochemical devices, 333 Biological systems, discovery of, 425 Biomolecular supramolecular complexes, 371 Biomolecules, 333, 372, 373 application, 372 structure-emerging functionalities, 333 supramolecular structures, 373 Biomolecules-nanoparticle systems, 334 electronic functions, 334 hybrid systems, application, 334 Biosensor devices, 301 Biotin-modified nucleic acid, hybridization, 365 Biotinylated aptamer, 363 Bipyridinium-based process, 414 Bipyridinium dications, 194–198 electrochemical reduction, 198
electrons traveling from electrode to ruthenium centers, 195, 196 LUMOs, 193, 195 Bipyridinium thiols, 189 design/synthesis, 189 Bis-thiaxanthylidenes, 464 photo/redox-controlled fluorescence switch, 464 Bisthiols, 188, 189 design/synthesis, 189 self-assembled monolayers, noble metals, 188 Boltzmann constant, 112, 528, 540 Bonafide models, 139 Bridging ligand, 125, 130, 136, 137 lowest unoccupied molecular orbital (LUMO), 137 Capillary electrophoresis (CE) process, 102 Carbon materials, 243 diamond, 243 glassy carbon, 243 graphite, 243 Carbon nanoparticles, 201 C60/C70, electrochemistry, 201 larger cages (C76-C92), electrochemistry, 202–204 Carbon nanostructures, 220, 230, 252 carbon nanohorns (CNHs), 252 carbon nanotubes, 220 nanoonions, 220 Carbon nanotube (CNT), 220, 222, 230, 243–252, 335, 346, 348, 370, 508 donor-acceptor ensembles, 248–252 enzyme hybrids, 335 electrochemistry, 220–222 length of, 348 Nafion composite films, 508 polycyclic aromatic compounds binding, 348 single-walled nanotubes, 221, 230 soluble assemblies, cyclic voltammograms, 221 soluble supramolecular complexes, 222 Cascade energy transfer, See Antenna effect Catalytic nanoparticles (NPs), 364 Catenane(s), 377, 391, 393, 395–399, 401, 402, 404, 414–418, 417, 419
INDEX
cartoons of, 396 circumrotation of, 393 coconformational isomers, 401 components of, 395 electroactive, 377, 419 electrochemical behavior, 418 correlation diagrams, 397 as electrochemically driven molecular rotors, 414–420 electrochemistry, role, 379 electron exchange, 380–406 to ‘‘read’’ the state of the system, 380, 406–420 to ‘‘write’’ state of the system, 406–420 half-wave reduction potentials, 383, 416 correlation, 383 immobilized on surfaces, 404–406, 419–420 molecular/supramolecular characters, 377 recognition sites in their rings, 391–404 redox-active units, 379 redox controlled ring rotation, 416 reduction, cyclic voltammograms, 398 ring rotation, 378 switching processes of, 418 template-directed synthesis, 391 tetracationic cyclophane component, circumrotation of, 402, 415 Cationic dyes, 536 Cavitands, 76 encapsulated guests, 79 electrochemistry, 79–81 general synthesis, 76 type hosts, 74 CB inclusion complexes, 68 electrochemistry, 68–74 CB7.2 complex, 74 energy minimized structure, 74 CB7-MV2 þ system, 69 CBPQT4 þ , 457 electrochemically induced translocation of, 457 CD2Cl2 solution, 77 NMR spectroscopic data, 77 Central phenolate oxygen atom, 40 Charge hopping, 111 Charge injection process, 533 rate constant, 533
583
Charge pooling devices, 146 redox units, 146 Charge storing device, 147, 166, 173 property, 147 Charge transfer (CT), 379 interactions, 414 Charge transport mechanism, 369 Chelating ligands, 124 protection/deprotection steps, 124 Chemical bonds, breaking, 537 Chemical-electrochemical mechanism, 65 Chemical Faraday cages, 205 Chiroptical switching materials, 457 behavior, 462 Chronoamperometric, 457 experiment, 164, 558 plot, 193 Circular dichroism (CD), 63 inclusion complexes, 64, 67, 70 electrochemistry, 64–68 signal, 452 intensities, 457 spectroscopy, 277 types, 63 Co(II) benzimidazole complexes, limitations of, 544 Co(II) electrolytes, 559 diffusion, limiting role, 559 Co(II)/(III)-mediated cell, 574 J-V curves, 574 Co(II)/PTZ system, 552 J-V curves, 552 Cobalt electron transfer mediators, 544, 545, 549, 550 heteroleptic dyes, 545 performance, 550 Cobaltocenium, 10, 16–17, 66, 158 anion receptors, 17 cobaltocene redox couple, 77 half-wave potential, 77 electrochemistry, 66 hexafluorophosphate, 78 reduction, 10 voltammetric wave, 77, 158 Cobalt phthalocyanine dendrimers, 98 Cocaine, 360, 367 amplified impedimetric analysis, 367 electrochemical analysis, 360 linear sweep voltammograms, 360
584
INDEX
Complementary DNA (cDNA), 287, 291 Complex systems, characterization of, 379 Conductive metallic nanowires, synthesis of, 368 Conductive polymers, 567, 573 PEDOP/PEDOT, 573 Conformational searching/quenched molecular dynamics protocol, 100, 101 Coordination compounds, 542–558 electrochemical properties, 542 electron mediators, 542 redox mediators, 558–562 mass transport limitations, 558 Copper-based molecular shuttles, 437, 438, 441, 443 electrochemically induced molecular motion, 437 Copper-complexed catenane, 426–432 electrochemically driven motions, 426 archetype, 426–428 electrochemically induced molecular rearrangements, 427 in motion with three distinct geometries, 428–430 Copper-complexed rotaxanes, 432, 435–437, 440 electrochemically steered machines, 437–443 mobile ring, 440 Copper(II)-complexed rotaxane, 438, 441, 434 cyclic voltammetry behavior, 434 rearrangement of, 441 schematic representation, 462, 463 voltammogram, 438 Copper-tetracyanoquinodimethane (Cu-TCNQ), 472 Copper thiacyclam complex, 461 Core-based materials, 318–322 definition, 318 gold NPs, 321 [Cu2(bis-bidentate ligand)2] þ double-strand helicate complex, 55 cyclic voltammetric behavior, 54 disassembling, square scheme, 53 MeCN solution, cyclic voltammogram, 55 molecular structure, 51 Cu(I) complex, 462, 555
electrochemical properties, 555 schematic representation of, 462 structure of, 555 Cu(I)/Cu(II) electron mediators, 557 efficiencies, 557 electrochemical properties, 556 photoaction spectra, 557 CuII double-strand helicate complex, 56 CuII mononuclear complex, 52 Cyanine dye, 572 Cyclic voltammetry (CV), 4, 8, 28, 167, 241, 262, 388, 435, 437 data, 6, 95, 97, 99, 102, 104, 105, 429 half-wave potential, 6 Nicholson analysis, 105 simulation software, 7 experiments, 40, 53, 69, 147, 552 patterns, 413 Cyclic voltammogram, 136, 137, 177, 186, 187, 189, 190, 193, 197 Cyclobis(paraquat-p-phenylene), 313, 457, 470 b-Cyclodextrins (b-CD), 155, 158 macrocycle, 404 ring, 405 self-assembled monolayers (SAMs), 155 Cyclodextrins (CDs), 62, 68, 311 inclusion complexes, 311 receptors, structures, 63 Cyclohexane, chair-boat equilibrium, 425 Cyclophane, 314, 316, 393, 457 chemical structures, 314 derivatives, 315 electron-deficient, 406 five-layered assembly, 316 cyclic voltammograms, 316 gold NP ISFET, schematic representation, 317 receptors, 315 Cytochrome c (Cyt. C), 90, 154, 338 polymethyl ferrocene derivatives, use, 154 redox potential, 90 cytochrome oxidase (COx), 338 assembly, 339 electrode, 339 Debye length, 111 Decanuclear species, 133–135 Delocalized p-electron system, 229, 248
INDEX
Dendrimer(s), 87, 88, 92, 93, 112, 121–123, 145, 146, 148–152, 155–157, 159, 163, 166, 168, 170, 174, 176, 180 array, 125 based on polypyridine metal complexes, 163–173 Co(II) bound tyrosine residues, 92 redox potentials, 92 components, 170 convergent synthetic approach, schematization, 122, 123 decanuclear, 171 divergent synthetic approach, schematization, 122, 123 docosanuclear, 138, 172 encapsulation effects, 87, 88, 112 on electron transfer, 112 ferrocene-type units, 148 structural formula, 149 fourth-generation, structural formula, 157 generation, 108, 111 hexaphenylbenzene core, structural formula, 176 molecular batteries, 180 nonspherical shape, 100 octahedral Mo6 cluster, structural formula, 152 polarity of solvent, 93 redox-active units, 145 role of, 121 scaffolds to, 146, 180 shells, energy migration pattern, 139 tetrairon [{CpFe(m3-CO)}4] clusters, 159 cyclic voltammograms, 159 structural formula, 159 topology of, 87 effect of, 106 tris-isothiocyanate core, structural formula, 178 tris-viologen core, structural formula, 166 Dendron, 178 based self-assembled monolayers, 108 tris-isothiocyanate core, structural formula, 178 Density of states (DOS), 244 Deoxycholic acid (DCA), 569 Deoxyribonucleic acid (DNA), 261, 365, 511 application, 368 base pairing, 1
585
based nanocircuitry/devices, 368–370 biosensor, 261, 285, 293, 363, 511 operational principle, schematic representation, 293 ECL detection of, 511 electrochemical analysis, 358, 359, 361, 364, 366 hybridization, 368 machine, 357 marker, 264 recognition process, 362 self-assembly methods, 368 templates, 368, 372 SWCNT FET construction, 370 DFT calculations, 211 4,40 (50 )-Dialkylthiotetrathiafulvalene, 470 2,6-Diamidopyridine (DAP), 308 pyrene-functionalized MMPCs, 310 recognition elements, 309 Diarylethene moieties, TTF derivatives, 454 Dicationic guests, electrochemical reduction mechanism, 70 1,2-Dithienylcyclopentene derivative, 458, 459 Dicopper(I) complex, structure, 50 double-strand helicate complex, 52 redox-driven disassembling, 52 Dicopper(II) double-strand helicate complex, 54 Differential pulse voltammetry (DPV), 44 experiments, 47 peaks, 385 profiles, 45, 46 studies, 49 voltammograms, 91, 132, 134, 135, 203, 402 Diffusion coefficient, 97, 99, 147, 161, 192, 533, 540 determination, 192 Digital simulations techniques, 66 Diluted self-assembled peptide film, 274 schematic representation, 274 Dimethoxybenzene (DMB) units, 382, 383, 393 Dimethoxynaphthalene (DMN) units, 395, 399, 417 electroactive units, 395 4-(N,N-Dimethylamino)benzonitrile (DMABN), 451
586
INDEX
5-Dimethyl-aminonaphthalene, 455 Dimethylaminophenyl group, 4 reversible oxidation, 4 Dimethylaminophenylurea systems, 9, 26 Dinuclear species, 126 [(bpy)2Ru(2,3-dpp)Ru(bpy)2]4 þ , 126–128 cyclic voltammogram, 129 [(bpy)2Ru(2,5-dpp)Ru(bpy)2]4 þ , 126–128 Dip-pen-nanolithography (DPN), 353, 372 1,4-Disubstituted azine, fluorescence intensity of, 454 Dithienylethene system, 459, 460 ring opening/closing, 460 DMF solution, 110, 241 DMSO, 17, 241, 307 Donor-acceptor systems, 106, 492–494 interaction, potential shifts, 413 luminophores, 495 Dpbiiq-based macrocycle, 443 Dpp-based macrocycle, 440 Dpp bridging ligands, 126, 132 Diquat (DQ2 þ ) complex, 70, 71 chemical reactions, 71 electrochemical reactions, 71 Dropping mercury electrode (DME), 89 Dry-processing techniques, 527 Dual-pathway square scheme mechanism, 409 Dye-sensitized solar cells (DSSCs), 527, 528, 533, 536, 538, 542, 545, 547, 554, 560, 562, 571, 574, 575 cathodes, catalytic materials for, 571–574 characterization, 532–534 components, developments in design, 523 dark currents, 545 design of, 528 efficiency, 527, 558, 575 electroactive ions, role, 558 role, 575 electron transfer processes, 527 electron transfer dynamics, role, 528 electron transfer mediators, 547 functioning, 528, 538 J-V curves, 554 mass transport limitations, 560, 575
open-circuit photovoltage, 533 performance, 527 photocurrent-voltage characteristics, 542 redox systems for, 538 Dye molar extinction coefficient, 532 Dye sensitizers, 534, 574 ECEC mechanism, 15 Electrical biosensor, 334 aptamer biosensors, 362 design, 334 Electrical devices, 356 based on supramolecular assemblies of nucleic acids, 356–370 Electrical potential, 524 Electrical sensing, 358–363 supramolecular nucleic acid, 358 monolayer structures on electrodes, 358 nanoparticle hybrid structures, 363–367 Electrically contacted enzyme electrodes, 340 applications, 340 Electroactive compounds, 186, 189, 197 4,40 -bipyridinium dications, 189 pendant thiol groups, 186 Electroactive DNA, 285–294 intercalators, 285 Electroactive films, 192 electron transport properties, 192 Electroactive monolayers/multilayers, 189 electron transport, 192–196 self-assembly, 189–191 Electroactive thiols, self-assembled monolayers/multilayers, 185 Electroanalytical devices, 243 biosensors, 243 electrochemical sensors, 243 Electrochemical aptasensors, detection schemes, 290 Electrochemical data analysis, 23 Electrochemical immunoassay, 318 Electrochemical ring-closing reaction, 459, 460 Electrochemical techniques, 147, 393, 395, 563 drawbacks of, 147
INDEX
Electrochemically controlled hydrogenbonding systems, 2–4, 12, 16, 19, 26 assemblies design, considerations, 8–15 electrolyte, 13–15 H-bonding vs. proton transfer, 15 host/guest structure, 12–13 redox couple, 8–12 characterization, 4–8 cobaltocenium, 16–17 detection, 4–8 ferrocene, 17–18 oxidation-based, 26–27 principles, 2–4 reduction-based, 19–26 Electrochemically controlled molecular motion, 33, 406, 430 Electrochemically driven lasers actions, 502–503 Electrochemiluminescence (ECL) system, 477, 478, 480–499, 503, 506 activated complex, 512 analytical methods, 507, 511 based assays, 509, 510 schematic diagram, 510 definition, 477 detectors, 506, 507 devices, 500–505 electrochemically driven lasers actions, 502–503 light-emitting electrochemical cells, 500–502 efficiency(ies), 485, 486, 491, 497, 512 emission, 480, 488, 505 intensity vs. time plots, 483 energetics, 480–482 excitation, 507 imaging systems, 503–505 investigations, 482, 510 experimental techniques, 482–485 labeling method, 511 luminophores, 491, 507, 512 materials, 480–499 ECL coreactants, 495–499 inorganic ECL systems, 485–490 organic ECL systems, 490–499 operating principles, schematic diagrams, 505 polymer, 509 practical applications, 499
587
probes immobilization methods, 513 sensors, 505, 509 signal, 507, 508 Electrochromic monolayers/multilayers, 196–197 Electrochemiluminescence sensors, 505 bulk solutions, 506–507 immunoassay systems, 509–511 from modified electrodes, 507–509 Electrohydrodynamic effects, 500 Electroluminescent devices, 245 Electron-accepting material, 236 (6,6)-phenyl-C61-butyric acid methyl ester (PCBM), 236 Electron-deficient flavin guest molecule, 309 Electron-donating species, 558 Electron donor-acceptor compounds, 448, 469 application of, 448 information storage, 469–472 interactions, 419 Electron fluxes, 447, 534 Electron/hole recombination pathways, 568 Electron mediator, 536–538, 542, 560, 574 chemical phases, 536 diffusion of, 560 Electron paramagnetic resonance spectroscopy (EPR), 215, 428 Electron shift, redox-induced, 138 Electron transfer process, 1, 35, 146, 164, 166, 193, 242, 249, 335, 379, 405, 477, 482, 494, 528–532, 555 annihilation, 486, 490, 495 driving force, 530 excitation, 512 free energy curves, 529, 530 golden rule expression, 529 kinetic parameters, 529 mediators, IPCE spectra, 553 rate attenuation, 97, 99 estimate of, 97–99 quantitative measure establishment, 99–106 Electronic absorption spectroscopy, 64 Electronic devices, 245 light-emitting diodes, 245 photoactive dyads, 245 sensors, 245 Electronic excitation energy, 481
588
INDEX
Electronic transmission coefficient, 531 Electrostatic layer-by-layer method, 508 Empty fullerenes, 201, 203 electrochemistry, 201–204 electrochemical potentials, 203 Encapsulated chromophore, 87 fluorescence quenching, 87 Encapsulating dendrons, generation, 106 Encapsulation process, 88, 113 structural model, 113 Endohedral fullerenes, 204–220 cage-size distribution, cluster size influence, 213 cage size decrement, 216 cage size increment, 217 electrochemistry, 204, 220 Gd3N@C2n, 218 characteristic features, 218 cyclic voltammograms, 218 redox potential, 218 La@C82/Y@C82, differential pulse voltammetry, 208 M@C82 compounds, 206–208 half-wave redox potential, 207 influence of metal, 206–208 metal carbide endohedral fullerenes, 210–212 differential pulse voltammetry, 211 square wave voltammetry, 211 M3N@C80 compounds, 216 anodic half-wave potentials, 216 cathodic peak potentials, 216 influence of metal, 215–216 redox potentials, 216 M3N@C88, cyclic voltammograms, 219 molecular hydrogen, encapsulation, 205 monometallofullerenes, 205–206, 209– 210 cyclic voltammogram, 210 Eu@C74/Tm@C78, 209–210 M@C2n, 205–206 multimetallofullerenes, 210, 211 differential pulse voltammetry, 211 square wave voltammetry, 211 M2@C2n, 210–212 Nd3N@C2n/Pr3N@C2n compounds, 219 redox potential, 219 Sc3N@C68, 217 donating ability, 217
non-IPR cage, 217 Sc3N@C80, 214, 215 bulk electrolysis, 215 cyclic voltammograms, 214 influence of carbon cage symmetry, 212–215 trimetallic nitride endohedral metallofullerenes, 212 Yb@C2n/Ca@C2n compounds, 208–209 half-wave redox potential, 209 influence of cage, 208 Energy gap, See Bandgap Energy migration processes, 140 Energy transfer processes, 138, 451 Enzyme-NPs hybrids, 321, 344 electropolymerization, 344 Enzyme electrode system, 321 1-Ethyl-3-methylimidazolium selenocyanate (EMISeCN), 561 Excimers, 481 Exciplexes, 481 Exciton-coupled circular dichroism (ECCD) signal, 462 Exo-active surfaces, 303, 308, 309, 313, 317 reversibility, 313 versatility, 313 Exo receptors, 18 Fabry–Perot cavity, 503 Faraday constant, 186, 480 Fast green (FG) dye, 565 Fc-azo-T system, 292 Fc-conjugated pepstatin, 274 Fc-conjugated uridine (dUTP), 279, 284 structures, 284 Fc-labeled DNA, 283 Fc-labeled nucleobases, 275–285 conjugates, 276 phosphoramidite derivative, 275 Fc-labeled ODNs, 278, 286, 287 conjugates, 277 on-column synthesis, 278 single-nucleotide mismatch, electrochemical discrimination, 287 Fc-peptides, 272 conjugates, 270 Fc-pyrimidines, 279 ORTEP drawings, 279
INDEX
Feldberg plots, 484 Ferrocene (Fc), 17–18, 26 conjugated oligonucleotide, 286 hybridization, schematic representation, 286 conjugated peptides, 266–275 dendrimers, 17, 25, 108, 150, 163 derivatives structures, 72 molecular switches, 454 electron acceptor/donor, 454 receptors, 304 voltammetric response, 80 Ferrocenylated nucleosides, Sonogashira coupling, 276 Field effect transistors (FETs), 252, 367 Flavin, 19–21 modified electrodes, 24 one-electron reduction, 19 Flavin adenine dinucleotide (FAD), 321, 337 cofactor, 341 units, 344 Flavin mononucleotide (FMN), 311 molecules binding, schematic representation, 312 Flavoenzymes, 19, 20 Fluorescence spectra, 449 Fluorescence switches, 461 behavior, 451 Four-coordinate copper(II) complex, 439 chemical rearrangements, 439 Fourth-generation dendrimers, 165 Franck–Condon factor, 530 Franck–Condon principle, 553 Franck–Condon states, 503 FT-IR spectroscopy, 16, 28 Fullerene(s), 203, 204, 229–242 cage size, 203 dendrimers, 177–179 donor-acceptor photoactive dyads, 231–236 electron acceptor properties, 232 interlocked architectures, 237–242 organic solar cells, 236–237 oxidation/reduction, 231 Functional nanoparticles, categories, 302 Gate-source potentials, plots, 317 Gelators, 539, 540
589
characteristics, 540 chemical structures, 539 Gel electrolytes, 538 I/I3 redox couple, 538–542 Giant redox dendrimers, 151 breathing mechanism, 151 investigation techniques, 151 Glassy carbon electrode (GCE), 69, 89, 401, 403 Globular actin (G-actin), 351 polymerization, 351 protein monomer units, 350 Globular nucleic acids, 334 Glucose dehydrogenase (GDH), 344, 363 electrical wiring, 363 Glucose/O2 biofuel cell, 341 Glucose oxidase (GOx), 337, 338, 341, 342, 345, 348 bioelectrocatalytic activation, 338 electrode, 322, 347 CNT-electrically contacted assembly, 347 fabrication, 322 layer-by-layer (LBL) deposition, 341, 342 modification, 345 Gold nanoparticles (AuNPs), 160, 161, 305, 343, 344, 353, 355, 365 assembly of, 343, 369 biocatalytic enlargement, 355 characteristic, 160 electrochemical properties, 160 g-actin hybrid monomers, 352 networks, 355 schematic representation, 161 thiolate bonds, 188 Gold nanowire, 352, 353 generation, 353 uses, 352 Grotthus-like mechanism, 541 Helicate complexes, 49 assembling/disassembling, 49–57 Heptanuclear dendron, 138, 140 redox behavior, 140 two-step electron transfer pathway, 138 Heterocyclic compounds, 453 Heterocyclic fragments, N-alkylation, 189 Heterodinuclear [2]catenane, 431 electrochemical characteristics, 431
590
INDEX
Heterodinuclear bis-macrocycle transition metal complex, 430 intramolecular motion, 430–432 Heteroditopic cylindrical ligand, translocation, 42 Heterogeneous electron transfer process, 75 Heteroleptic dyes, 544, 553, 557 advantages, 544 blocking effect, 553 Z907 dye, 557 Heterometallic dendrimers, electrochemical behavior, 163 High-generation dendrimers, drawback of, 140 High-generation dendrons, preparation of, 122 High-performance liquid chromatography (HPLC), 212, 283 Highest occupied molecular orbitals (HOMOs), 124, 195, 493 Highly ordered pyrolytic graphite (HOPG), 469 Hole transporting material (HTM), 567 HOMO-LUMO gap, 204, 206, 208, 210, 217–219, 223 M3N@C88 compounds, 220 half-wave redox potential, 220 M3N@C96 compounds, 220 redox potential, 220 Homometallic dendrimers, structural formulas, 162 Host-guest complex, 5, 8, 12, 155 structure, 12–13 Host-guest complexation process, 62 Host-guest inclusion complex, 60, 61 kinetic rate constants, definition, 61 Host-guest interactions, 177, 308 redox modulation, 308 Human cytomegalovirus (HCMV), 318 DNA hybridization assays, 319 Hybrid nanostructures, applications, 346 Hydrogen bonding, 3, 4, 20 vs. proton transfer, 15 shuttles, 23, 465 sites, 10 direct perturbation, 3–4 Hoogsteen-type, 288 indirect perturbation, 4 interaction, 2
strength, increasing methods, 3 supramolecular complexes, 8 Hydrogen donors, 17, 465 Hydrophobic dendrimers, 105 Incident photon to current efficiency (IPCE), 532, 556 electron collection efficiencies, 556 wavelength-dependent, 532 Inclusion complex, 61, 65, 74 electrochemical kinetics, 65 electrochemistry, 74–81 formation, 74–81 Indirect electrochemically controlled H-bonding systems, 27–29 Indium-tin oxide (ITO), 500 conductive surface, 314 electrode, 316 Induced circular dichroism (ICD), 463 Iodide/triiodide-mediated cells, 560 Ion-sensitive field effect transistor (ISFET) device, 317, 367 use of, 367 Ion translocation, 36–49 anion translocation, 43–49 NiIII/NiII couple, 43–49 metal translocation, 36–43 CuII/CuI couple, 41–43 FeIII/FeII couple, 36–41 Ionic liquids, 538 I/I3 redox couple, 538–542 IR-visible sum-frequency generation vibrational spectroscopy, 265 Iron center, 37, 40, 41 pendular motion, 40, 41 square scheme, 41 translocation, 37 Iron-oxo tricyclononane core dendrimers, 92 redox activities, 92 Iron-sulfur cluster core dendrimers, 99, 102 isomer, 101, 111 Jahn–Teller effect, 41 Koopmans’ theorem, 131 Labile chloride ligands, 133 advantage of, 133
INDEX
Langmuir–Blodgett technique, 404, 507 films, 471 monolayer techniques, 508 Laser photolysis experiments, 543 Ligand(s), 289, 488 systematic evolution, 289 Ligand approach, 169 Ligand-based reduction processes, 125, 167 Ligand field (LF) energy, 35 stabilization energy, 41, 51 Ligand-ligand interaction, 124 Light-emitting devices, 487, 502 Light-emitting electrochemical cells, 479, 500–502 Light-harvesting dendrimers, 124, 127 schematic representation, 127 Light harvesting efficiency (LHE), 532 Linkage isomerization reaction, 428 Liquid electrolytes, 539, 541 gelation, 539 Logic gates, 448 Lowest unoccupied molecular orbital (LUMO), 124, 129, 204, 210, 215, 231, 493 M13 phage DNA, 362 Macrocycle(s), 391, 395, 399, 402, 404, 408, 410, 436, 440 DMB units, 410 DMN units, 395 rearrangement, 436 shuttling of, 408 MALDI-TOF mass spectra, 167 Marcus equation, 111, 112 activation parameters, 112 fits, 112 for nonadiabatic electron transfer, 111 Marcus theory, 335, 490, 494, 530 assumptions, 531 prediction, 489 Mass spectrometry, 265 1-Mercaptoundecanoic acid (MUA)-modified surfaces, 272 Metal-based dendrimers, 88, 138 photophysical properties, 138 Metal bipyridine/terpyridine-containing dendrimers, 95–97 Metal-centered oxidation process, 164 Metal complexes, 163, 164, 167, 169
591
branching points, 167 dendritic structure, 169 Metal-polypyridine dendrimers, 125, 126, 136, 140 building blocks, features of, 122 redox properties, 125, 136 Metal-thiolate bonds, formation, 188 Metal-to-ligand charge transfer (MLCT), 133, 455, 478, 534, 547 absorption band, 429 state, 236 transitions, 121 Metallic nanowires, 355, 369 bottom-up synthesis, 369 growth, 355 Metallic nanoparticle, 321, 344, 352, 365 coelectropolymerization, 344 use, 365 Metallocene-labeled peptide nucleic acid, 292 Metallocene-like dendrimers, 148–163 Metallocene receptors, 9 Metallodendrimers, 95, 167, 486 class of, 167 ECL intensities, 486 Metalloporphyrins, 232 receptor, 307 Methylene blue (MB), 290, 358, 360 electrical response, blocking, 360 labeled aptamer, 290 redox label, 358 Methyl viologen (MV), 71, 456 cation radical, 71 electron acceptor, 456 one-electron reduction, 71 chemical reactions, 71 electrochemical reactions, 71 Mixed ferrocene-cobaltocenium poly (propylene amine) dendrimers, 158 prepartion, 158 Mixed monolayer protected clusters (MMPCs), 301, 303, 305, 308–310, 314 chemical structures, 314 nanoscale receptors, 304 redox behavior, 310 Mo2 tetrabenzoate cluster core dendrimers, 92 Molar extinction coefficient, 535
592
INDEX
Molecular batteries, 180. See also Dendrimers Molecular encapsulation, 59, 95 free energy change, 60 kinetic considerations, 60–62 thermodynamic considerations, 60–62 Molecular logic gates, 467 electroactive systems, 467 logic circuit with TTF derivatives, 467–469 Molecular machines, 33, 377 construction, 377 prototypes, 426 Molecular-scale engineered systems, 240 Molecular sensitizers, 534–536 Molecular shuttles, 23, 240, 437 Molecular surgery approach, 205 Molecular switches, 448 Monolayer protected clusters (MPCs), 301 Monomeric tetrairon cluster, properties, 160 Multicomponent-structured dendrimers, components, 130 Multielectron storage device, 145, 146 dendrimers, 145 use, 146 Multijunction photovoltaic cells, 565 construction, 526 Multiwalled carbon nanotubes (MWCNTs), 541 NAD(P) þ -dependent enzymes, 338 Nafion film-modified electrode, 508 Nanocrystalline TiO2 films, 548 photoaction spectra, 548 Nanoparticles (NPs), 301, 303 AFM images, 356 enzyme, construction, 341 enzyme hybrids, 335 roles, 302 Nanosecond time-resolved experiments, 548 Nanotechnology, 334, 447 advantages, 355, 447 applications, 355 Nanotubes, covalent functionalization, 245 Nanowire, construction, 370 Naphthalene fluorescence, 461 Naphthalimide, 465 shuttle, 23 NCS ancillary ligand, 534, 535
polypyridine Ru(II) dye, structure, 535 Near-infrared (NIR) photons, 534 Nernst equation, 248 N-hexylamine, protonation, 402 Nicotinamide perylene diimide dyad, 456 redox moiety, 456 Nicotinamide adenine dinucleotide (NADH) cofactor, 338 Nicotinamide adenine dinucleotide phosphate (NADP) cofactors, 348 Nitroaniline-functionalized peptide, 362 thrombin-mediated hydrolysis, 362 Nitrobenzenes, 9, 25 N,N0 -diaminoethyl-4,40 -bipyridinium, 315 N,N0 -dimethyl-4,40 -bipyridinium, 68 N,N0 -ethylene-2,20 -bipyridinium, 69 Noncompetitive heterogeneous electrochemical immunoassay, 319 Nuclear magnetic resonance (NMR) technique, 111, 267 experiments, 68, 75, 78, 80 spectroscopic data, 79 Nucleic acids, 357, 372 aptamers, 289 binding properties, 357 hybridization, 333 intrinsic electroactivity, 261–266 library of, 356 properties, 368 Nucleoside derivatives, 280 conjugates, redox potentials, 280–282 Octols, synthesis, 76 Oligodeoxynucleotide (ODN), 275 conjugates, 285 Optical absorption intensity, plots, 247 Organic dyes, 565 Organic light-emitting diodes (OLEDs), 500 technology, 567 Organic solar cells, 236 power conversion efficiencies (PCE), 236 Organometallic dendrimers, 155, 158 Os dyes, 536 sandwich-type solar cells, photoaction spectra, 536 Os(II) polypyridine complexes, 124, 126 dendritic structure, 138 dinuclear compounds, 126
INDEX
effective models, 126–130 dinuclear complexes, 126 energy transfer processes, 138 first effective light-harvesting antenna, 131–133 mixed metal Os-Ru tetranuclear dendrimer, 131 hexanuclear metal complexes, 136–138 metal-polypyridine dendrimers, 138–140 electronic energy transfer in, 138 oxidation process, 126 redox properties, 124–125 second generation, 133–135 decanuclear species, 133 third generation, 138 docosanuclear species, 138 trinuclear species, 130–131 Osteryoung square wave voltammetry, 99, 202 cyclic voltammetry, 202 Outer Helmholtz plane (OHP), 77 Oxidation-ring-opening process, 458, 459 Oxidized flavin, deprotonation, 21 Para-conjugated polymers, 236 Parallel connected stacked cell, 571 Para-substituted nitrobenzenes, 9 half-wave potential shift, 9 Pd nanoparticles, production of, 152 p donor substrates, 314 adrenaline, 314 dopamine, 314 p-hydroquinone, 314 PEDOP-based DSSC, 568 J-V characteristic, 568 matrix, 569 PEDOT, in situ photoelectropolymerization, 569 Pentacoordinate copper systems, 432, 434 Peptide nucleic acid (PNA), 292 based biosensors, 293, 294 DNA electrochemical sensor, 293 electroactive, 285–294 Phenothiazine (PTZ), 304, 551 ligands, 304 Phenyl acetylene-based dendrimers, 100, 103 Phosphorus-based dendrimers, 109 Photochemically stable fluorinated polymers, use, 540
593
Photoelectrochemical cells, 547 performances of, 547 Photoelectrochemical solar cells, 534, 574 Photoinduced charge separation process, 233 Photoinduced electron transfer process, 449, 456 Photon fluxes, 447 Photosensitized reduction experiments, 175 Photosynthetic model systems, 371 Photovoltaic cell, 523, 526, 534 efficiency of, 534 fabrication, 526 inorganic cells, 523, 526 light absorber, 524 organic cells, 523, 526 photoelectrochemical cells, 524 Photovoltaic energy, 523 renewable energy sources, 523 p-p interactions, 34, 241, 249, 250 Pimerization process, 175 Pirouetting copper-complexed rotaxanes, 432–437 electrochemically driven machines based, 432 Platinum film electrodes, 503 NPs, 346 p-n junction, photoinduced charge separation, 525 Polyallylamine (PAA) polyelectrolyte, 341 Poly(amido amine) (PAMAM) dendrimer, 164, 165 Polyaromatic hydrocarbons, 491, 492 Polyaromatic systems, 449 Polycationic cobaltocenium dendrimers, 157 Polycrystalline gold, 190, 191, 193, 194 Polymerase chain reaction (PCR), 356 Polymeric materials, 302 dendrimers, 302 polystyrene beads, 302 Poly(propylene amine) dendrimers, 154 donor capability, 154 Polypyridilic Co(II) complexes, structure, 546 Polypyridine ligands, 124, 138 Polypyridine metal complexes, 163 Poly(vinylidenefluoride-cohexafluoropropylene) (PVDF-HFP), 540, 541
594
INDEX
Poly(vinylpyridine) polymers (PVP), 509 Population inversion, definition, 502 Porphyrin core dendrimers, 89–95, 105 Porphyrin/quinone systems, 232 Positive dendritic effect, 17, 306 Potentiodynamic electrolysis, 573 Preorganization, 12 Probe DNA hybridization, 358 Protein based nanocircuitry/devices, 350–356 binding properties, 355 Pseudo-b-barrel formation, schematic representation, 268 Pseudorotaxane, 237, 238 P-type semiconductors, 565, 575 Pulsed field gradient spin-echo NMR spectroscopy, 99 Pulse gradient stimulated echo (PGSE) NMR spectroscopy, 147 Pyridyl-quinoline-based complexes, 556 Pyrrole unit, electrooxidation, 24 Pyrroloquinoline quinone (PQQ), 337, 338 Quantum dot (QD) hybrids, 371 semiconductor, catalytic properties, 363 Quasi-solid-state cells, 542 Quasi-solid-state electrolytes, 539–541 use, 541 O-Quinones, 24 Quinquedentate macrocycle, 42 Ratchet-type systems, 402 Redox-active core dendrimers, 111 Redox-active cyclophane, 457 electrochemical shuttling of, 457 Redox-active dendrimer, 97, 108 cooperative behavior, 97 films, 107–111 Redox-active guests, 62 cucurbituril complexation, 62–74 cyclodextrin complexation, 62–74 Redox-active hemicarceplexes, 75 Redox-active metal-polypyridine dendrimers, 121 artificial antenna, 122–124 Redox couple, 8–12
Redox-controlled fluorescence, 449 switchable, principle of, 449 Redox-controlled molecular switches, 448 analogs-based, 455–456 ferrocene-based, 454–455 with miscellaneous electroactive systems, 463–467 quinone-based, 455–456 redox/photodual mode, 458–460 tetrathiafulvalene-based, 448–454 viologen-based, 456–457 Redox-dependent host-guest binding, equilibria, 5 Redox-dependent receptors, 9, 17 Redox-driven intramolecular motion, 35 square scheme, 35 Redox-driven reversible assemblingdisassembling process, 51 Redox-driven translocation, 42, 44 Redox enzyme, 335 electrical wiring by, 335–341 relay-functionalized monolayer assemblies, 335–341 immobilization, 335 Redox fluorescence switch, 454, 455, 461 Redox luminescence switch, 456 Redox mediators, 537, 561, 575 Redox proteins, 335, 341 electrical contact, 335 electrical wiring by, 341–350 carbon nanotube-hybrid systems, 341 supramolecular nanoparticle, 341 Resorcinarene molecules, 76, 79 capsules, 76–79 structures, 76 Retro-Bingel reaction, 177, 233 Rotaxanes, 23, 377, 387–390, 404, 406, 407, 410, 411, 413, 419, 457, 470, 471 architectural features, 389 containing films, behavior of, 404 cyclophane shuttling, 407 deprotonated, 408 dumbbell-shaped component, 387, 410 electrochemically controlled switching process, 406 as electrochemically driven molecular shuttles, 406–413 electrochemistry, role, 379 electron exchange, 380–406
INDEX
to ‘‘read’’ the state of the system, 380, 406–420 to ‘‘write’’ the state of system, 406–420 immobilized on surfaces, 404–406, 419–420 molecular string, 457 molecular/supramolecular characters, 377 pH-controllable, 408 molecular shuttle, 387 recognition sites, 380–391 for ring in their dumbbell component, 380, 386 redox-active units, 379 ring rotation, 378 ring shuttling, 378 electrochemically driven, 420 photoinduced, 410 structure formula, 387, 390, 406, 407, 410, 411, 413 template-directed synthesis, 380 Ru(bipy)32 þ , 487 chelate, 506 cyclic voltammograms, 487 ECL emission spectra, 487 tag, 510 TPrA system, 479, 496 Ru(bipy)32 þ þ C2O42 system, 498 ECL mechanism, 498 Ru(dcbpy)2(NCS)2, 544 transient absorbance decay kinetics, 544 Ru(II) polypyridine complexes, 124, 126, 567 dendritic structure, 138 effective models, 126–130 dinuclear complexes, 126 energy transfer processes, 138 first effective light-harvesting antenna, 131–133 mixed metal Os-Ru tetranuclear dendrimer, 131 for in situ photoassisted PEDOP growth, 567 hexanuclear metal complexes, 136–138 metal-polypyridine dendrimers, 138–140 electronic energy transfer in, 138 second generation, 133–135 decanuclear species, 133 third generation, 138
595
docosanuclear species, 138 trinuclear species, 130–131 Ruthenium chelates, 499 Ru(bprz), 32, 171, 172, 499 Ru(phen), 32, 173, 499 Saturated calomel electrode (SCE), 89, 91, 177, 265 Scanning electrochemical microscopy (SECM) technique, 504 Scanning tunneling microscopy (STM), 469 based nanorecording, 471 Self-assembled monolayers (SAMs), 24, 102, 110, 155, 404, 508 cystenamine-terminated, 102 Semiconductor heterojunction cells, 524 Semiconductor nanoparticles, CdS, 319 Sensing, exo-active nanoparticle surfaces, 303–317 Sensitizer dye, 527 Shuttling process, 408 Silica sol-gel materials, 509 composite films, 509 Silver nanowire, formation, 351 Silver-tetracyanoquinodimethane (Ag-TCNQ), 472 Silylferrocene dendronized NPs, 306 Single-photon counting technique, 479 Single-stranded DNA (ss-DNA), 286, 511 Single-walled carbon nanotubes (SWCNTs), 230, 243, 245, 246, 249, 252, 341 applications, 252 dimensions, 346 enzyme hybrids, 341 exTTF nanohybrids, preparation method, 250 Van Hove singularities, 246 voltammetric response, 245 Smoluchowski equation, 61 Solar cells, 524, 525, 538, 553 assembly, 553 I-V curve, 525 Solar devices, 526, 533 laboratory efficiencies, 526 Solar energy, 121, 523 photochemical conversion, 121 use, 523 Solid-state cells, 563 J-V characteristics, 563
596
INDEX
Solid-state dye-sensitized photovoltaic cell, 564 Solid-state hole conductors, 562–570 Solid-state light emitting cells (SSLECs), 500–502, 512 efficiencies, 512 schematic picture, 501 Solution electrochemiluminescent (SECL) cell, 500 Solvent polarity index, 93 Sonagashira cross-coupling reaction, 275, 277, 285 Source-drain potentials, plots, 317 Space charge effects, 567, 569 screening, 569 Spacer equipped cells, 560 photocurrent transients, 560 Spin forbidden process, 494 Spin-orbit coupling, 481 SPM-based techniques, 469 Square coordination geometry, 42 Square wave voltammetry (SWV), 284, 313 Superexchange tunneling mechanism, 106 Supramolecular chemistry, 237 Supramolecules, 469 information storage, 469–472 nanohybrids, 249 poly-catenated nucleic acid chains, 334 Surface-confined heterometallic dendrimers, 158 Systematic evolution of ligands by exponential enrichment (SELEX) process, 289, 356 Tetrabutylammonium ion (TBA), 550 Tetracationic cyclophane, 407 Tetrachloride perylene diimide, 450 fluorescence intensity of, 450 Tetrachlorobenzoquinone, 456 Tetrachloroethane (TCE), 231 7,7,8,8-Tetracyanoquinonedimethane (TCNQ), 26 Tetrahedral copper complexes, 554 structure of, 554 Tetrairon cluster, 160 2,20 ,7,70 -Tetrakis(N,N-di-p-methoxyphenylamine)-9,90 -spirobifluorene (OMeTAD), 565
Tetrathiafulvalene (TTF), 26–27, 109, 233, 314, 412, 413, 417, 448–453, 470 anthracene dyad, 449 azobenzene moieties, 453 based D-A systems, 449 electron donating ability, 451 fullerene dyads, 449 MPT, fluorescence of, 450, 451 oxidation of, 451 station, 412 porphyrin dyad, 450 pyridine-substituted, 450 redox-active, 449 redox state of, 452 substituted poly(isocyanide), 453 tetrachloride perylene diimide dyad, 450 Tetrazines, 25–26 Thermodynamic cycle, 45, 56 effect, 56 Thin films, electrical bistable behaviors, 471 Thin-layer photoelectrochemical experiments, 572 Thioanline-functionalized Pt NPs, 345 Thiols, self-assembled monolayers, on noble metals, 185–188 Third-generation dendrimer, 178 Three-electron reduction processes, 392, 393 Thrombin, 360, 361 electrochemical aptasensor, 360 electrochemical detection, 361 Time-resolved spectroscopy, 560 TiO2-based nanomaterials, 541 TiO2-nafion composite film, 508 TiO2/N3/CuSCN cell, 565 dark/photo I-V characteristics, 565 Trans-azobenzene, 453 Trans/cis-ring-closed isomers, 460 Transient absorption spectroscopy, 549, 544 Transition metal-complexed catenane/ rotaxane, 425 electrochemically driven molecular machines, 425 Transition metal complexes, 485 Transition metal-containing molecular machines, 443 Transparent conductive oxide (TCO), 571 1,2,3-Triazole-linked dendrimer, 153 Tri-n-propylamine (TPrA), 496, 497 Triple-potential-step technique, 477, 483, 484
INDEX
Triple-stranded helical metal complexes, 462 Triplet interceptor technique, 481, 484 Triplet quenching, 481 Triplet-triplet annihilation, 478, 481, 492 Triply branched compound, 384 Tris (2-phenylpyridine)iridium(III), 479 Tunneling effects, 531 Twisted intramolecular charge transfer (TICT), 451 Two-color assay, 284 Two complementary cyclic voltammetry experiments, 434 Two-station rotaxane, 381, 386 Unmodified cucurbit[n]uril, 63 UV/visible light irradiations, 453, 458 UV-vis-NIR absorption spectroscopy, 233, 429, 490 van der Waals forces, 540 van Hove singularities, 244, 246 Viologen, 67
binding affinity, 67 oxidation states, 67 Viologen-based dendrimers, 173–177 Viologen core dendrimers, 106–107 Viral DNA, 279, 283 amplified detection, 283 Viral M13mp18 DNA, 362 amplified electrochemical detection, 362 Virtual isoergonic process, 529 Wang’s insuline quantification, 265 XNOR logic gate, 467 X-ray photoelectron spectroscopic measurements, 188 X-ray structural analysis, 452 Z907 dye, 551, 557, 561 decay kinetic, 551 Zinc phthalocyanine derivative, 238 Zinc porphyrin, 240, 450, 454 Zwitterionic dendrimers, 169
597
Figure 1.5 CVs of 1 mM 40 in 0.1 M NBu4PF6 with different ratios of CH2Cl2 and CH3CN: (a) 100% CH2Cl2, (b) 50% CH2Cl2, 50% CH3CN, (c) 20% CH2Cl2, 80% CH3CN, and (d) 100% CH3CN. 100 mV/s scan rate.67
Figure 3.3
Energy minimized structure (PM3 method) of the CB7 . MV2 þ complex.
Figure 3.7 Energy minimized structure (PM3 method) of the CB7 . 2 complex.
Figure 3.9 Size comparison between cobatocenium (top left) and the hexameric molecular capsule formed by resorcinarene 3. Resorcinarene 4 would form a much larger capsule due to its longer undecyl “feet” compared to the methyl “feet” in 3.
Figure 3.11
Energy-minimized (PM3 method) structure of the Fc@52 molecular assembly.
Figure 3.12
Energy-minimized structure (PM3) of the Fc@6 inclusion complex.
Figure 6.6 (a) Structural formula of dendrimer 6 containing at its periphery 16 biferrocene units; (b) schematic representation of the proposed oxidation mechanism (at low scan rate) of dendrimer 6 in solution and immobilized at the b-CD host surface.40 Reproduced with permission from Ref. 40.
Scheme 9.18 Top: Plots of optical absorption intensity as a function of wavelength and electrode potential in the S11 region for K[h-NT]. In all plots, raw electrochemical data, that is, uncorrected for ohmic drop, are referenced to SCE. Bottom: Chirality map displaying the average standard potentials associated to each SWNT. HiPco SWNTs are located inside the red line, while arc-discharge SWNT are inside the blue line. Starred values were extrapolated from the linear fitting equations given in the text.
Figure 10.4 Schematic representation of the formation of the pseudo-b-barrel formed from molecular building blocks by tiling. (a) Molecular structure of Fc[Gly-Val-CSA]2, (b) formation of b-sheets through intermolecular N(H)O¼C hydrogen bonding, (c) H-bonding interactions between four molecules to form a b-barrel, and side view molecular-surface representation showing barrel. Copyright Wiley-VCH Verlag GmbH & Co. KGaA. Reproduced with permission from Ref. 49.
Figure 10.17 Schematic representation of the systematic evolution of ligands by exponential (SELEX) enrichment process.
Figure 15.2 Gradual color change of a CH2Cl2 solution of compound 27 containing 75% of the ring-closed isomer 27c when treated with a catalytic amount of [(4-BrC6H4)3N][SbCl6].
Figure 16.2 Cyclic voltammograms and ECL emission spectra of Ru(bipy)32 þ (red lines) and Ir(ppy)3 (green lines) chelates recorded in the author laboratory.
Figure 16.3 Structural formulae of the ligands L and emission colors of their L2Ir(acac) complexes in the Commission Internationale de l’E’clairage (CIE) coordinates.
Figure 16.5 Examples of the intramolecular donor–acceptor system and their emission spectra—aryl derivatives of N,N-dimethyl-aniline: BDMA: 4-(9-acridyl)-N,N-dimethylaniline, ADMA: 4-(9-anthryl)-N,N-dimethylaniline, PDMA: 4-(1-pyrenyl)-N,N- dimethylaniline, and NDMA: 4-(1-naphthyl)-N,N-dimethylaniline.
Figure 16.6 Schematic picture of a solid-state light-emitting electrochemical cell.
Figure 16.7 Schematic diagrams illustrating the operating principles of ECL generation using SECM technique.
Figure 16.8 Schematic diagram illustrating the operating principles of ECL based immunoassay by means of the sandwich type interaction.