Principles, Strategies, and Applications · CHEMOSENSORS Principles, Strategies, and Applications...

CHEMOSENSORS

Principles, Strategies, and Applications

Edited by

BINGHE WANGGeorgia State University

ERIC V. ANSLYNUniversity of Texas

A JOHN WILEY & SONS, INC., PUBLICATION

Innodata9781118019566.jpg9781118019566.jpg

CHEMOSENSORS

CHEMOSENSORS

Principles, Strategies, and Applications

Edited by

BINGHE WANGGeorgia State University

ERIC V. ANSLYNUniversity of Texas

A JOHN WILEY & SONS, INC., PUBLICATION

Copyright © 2011 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 eitherthe 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 forpermission should be addressed to the Permissions Department, John Wiley & Sons, Inc., 111 River Street, Hoboken, NJ 07030, (201) 748-6011, fax(201) 748-6008, or online at http://www.wiley.com/go/permission.

Limit of Liability/Disclaimer of Warranty: While the publisher and author have used their best efforts in preparing this book, they make no representationsor warranties with respect to the accuracy or completeness of the contents of this book and specifically disclaim any implied warranties of merchantabilityor fitness for a particular purpose. No warranty may be created or extended by sales representatives or written sales materials. The advice and strategiescontained herein may not be suitable for your situation. You should consult with a professional where appropriate. Neither the publisher nor author shallbe liable for any loss of profit or any other commercial damages, including but not limited to special, incidental, consequential, or other damages.

For general information on our other products and services or for technical support, please contact our Customer Care Department within the UnitedStates at (800) 762-2974, outside the United States at (317) 572-3993 or fax (317) 572-4002.

Wiley also publishes its books in a variety of electronic formats. Some content that appears in print may not be available in electronic formats. For moreinformation about Wiley products, visit our web site at www.wiley.com.

Library of Congress Cataloging-in-Publication Data:

Chemosensors : principles, strategies, and applications / edited by Binghe Wang, Eric V. Anslyn.p. cm. — (Wiley series in drug discovery and development ; 15)

Includes index.ISBN 978-0-470-59206-9 (hardback)

1. Biosensors. 2. Molecular recognition. 3. Chemical detectors. I. Wang, Binghe. II. Anslyn, Eric V., 1960–R857.B54C485 2011610.28—dc22

2011007559

Printed in Singapore

oBook ISBN: 978-1-118-01958-0ePDF ISBN: 978-1-118-01956-6epub ISBN: 978-1-118-01957-3

10 9 8 7 6 5 4 3 2 1

http://www.copyright.comhttp://www.wiley.com/go/permissionhttp://www.wiley.com

CONTENTS

Contributors vii

Preface ix

SECTION 1 FORCES GOVERNING EXCHANGEABLEINTERACTIONS 1

1 van der Waals Interactions and the Hydrophobic Effect 3Bruce C. Gibb

2 Ionic, Hydrogen Bond, and π –Cation Interactions 19Hector Adam Velazquez and Donald Hamelberg

3 Covalent Interactions in Chemosensor Design 25Yunfeng Cheng, Xiaochuan Yang, and Binghe Wang

4 Metal Chelation Chemistry 41Dongwhan Lee

SECTION 2 STRATEGIES TOWARD BUILDING THE DESIREDBINDING MOIETY 65

5 Scaffold Design Using Computational Chemistry 69Dale Drueckhammer

6 Combinatorial Search of Sensors 87Marc Vendrell, Suihan Feng, and Young-Tae Chang

7 Molecular Imprinting and Sensor Development 107Yagang Zhang and Ken D. Shimizu

8 Dendrimer-Based Sensors 121Lin Pu

v

vi CONTENTS

9 Nanoparticles and Sensors 163Yi-Cheun Yeh, Sarit S. Agasti, Krishnendu Saha, and Vincent M. Rotello

10 Aptamer Selection, Phage Display, and Sensor Development 191Hui Wang, Yan Chen, and Weihong Tan

11 Sensor Development Using Existing Scaffolds 211Hiroyasu Yamaguchi, Tomoki Ogoshi, and Akira Harada

SECTION 3 DETECTION METHODS IN CHEMOSENSING 227

12 Fluorescent Detection Principles and Strategies 229Raman Parkesh, Emma B. Veale, and Thorfinnur Gunnlaugsson

13 New Fluorophore Design 253Michael D. Heagy

14 Colorimetric Sensor Design 275Kevin L. Bicker, Sheryl L. Wiskur, and John J. Lavigne

15 Electrochemical Detection 297Simon R. Bayly, George Z. Chen, and Paul D. Beer

16 Surface Plasmon Resonance and Quartz Crystal MicrobalanceMethods for Detection of Molecular Interactions 329Yang Liu, Archana Jaiswal, Mark A. Poggi, and W. David Wilson

17 Array-Based Sensors 345Pavel Anzenbacher Jr. and Manuel A. Palacios

SECTION 4 CHEMOSENSORS: CASE STUDIES 369

18 Design of Cation-Selective Synthetic Fluorescent Indicators 371Christoph J. Fahrni

19 Anion Sensors 395Philip A. Gale and Claudia Caltagirone

20 Chemosensors: Case Studies of Indicators for Organic Molecules 429Oleksandr Rusin, Jorge O. Escobedo, and Robert M. Strongin

21 Molecular Recognition Elements for Toxin and Pathogen Detection 455Daniel M. Lewallen, Duane M. Hatch, and Suri S. Iyer

22 Chemical Sensing and Detection in Forensic Science 475Simon W. Lewis

Index 497

CONTRIBUTORS

Sarit S. Agasti, Department of Chemistry, University ofMassachusetts, Amherst, MA, USA

Pavel Anzenbacher, Jr., Department of Chemistry, Bowl-ing Green State University, Bowling Green, OH, USA

Simon R. Bayly, Inorganic Chemistry Laboratory, Univer-sity of Oxford, Oxford, UK

Paul D. Beer, Inorganic Chemistry Laboratory, Universityof Oxford, Oxford, UK

Kevin L. Bicker, Department of Chemistry and Biochem-istry, University of South Carolina, Columbia, SC, USA

Claudia Caltagirone, Dipartimento di Chimica Inorgan-ica ed Analitica, Universitá degli Studi di Cagliari,Monserrato (CA), Italy

Young-Tae Chang, Department of Chemistry & MedChemProgram of Life Sciences Institute, National Universityof Singapore, Singapore

George Z. Chen, Department of Chemical and Envi-ronmental Engineering, and Energy and SustainabilityResearch Division, Faculty of Engineering, Universityof Nottingham, Nottingham, UK

Yan Chen, Department of Chemistry and Department ofPhysiology and Functional Genomics, University ofFlorida, Gainesville, FL, USA

Yunfeng Cheng, Department of Chemistry and Centerfor Biotechnology and Drug Design, Georgia StateUniversity, Atlanta, GA, USA

Dale Drueckhammer, Department of Chemistry, StonyBrook University, Stony Brook, NY, USA

Jorge O. Escobedo, Department of Chemistry, PortlandState University, Portland, OR, USA

Christoph J. Fahrni, School of Chemistry and Biochem-istry, Petit Institute for Bioengineering and Bioscience,Georgia Institute of Technology, Atlanta, GA, USA

Suihan Feng, Department of Chemistry & MedChemProgram of Life Sciences Institute, National Universityof Singapore, Singapore

Philip A. Gale, School of Chemistry, University ofSouthampton, Southampton, UK

Bruce C. Gibb, Department of Chemistry, University ofNew Orleans, New Orleans, LA, USA

Thorfinnur Gunnlaugsson, School of Chemistry, TrinityCollege Dublin, Center for Synthesis and ChemicalBiology, University of Dublin, Dublin, Ireland

Donald Hamelberg, Department of Chemistry, GeorgiaState University, Atlanta, GA, USA

Akira Harada, Department of Macromolecular Science,Graduate School of Science, Osaka University, Osaka,Japan

Duane M. Hatch, Center for Sensors and Biosensors,Department of Chemistry, University of Cincinnati,Cincinnati, OH, USA

Michael D. Heagy, Department of Chemistry, New MexicoInstitute of Mining and Technology, Socorro, NewMexico, USA

Suri S. Iyer, Center for Sensors and Biosensors, Depart-ment of Chemistry, University of Cincinnati, Cincinnati,OH, USA

vii

viii CONTRIBUTORS

Archana Jaiswal, Q-Sense—A Division of Biolin Scien-tific, Linthicum Heights, MD, USA

John J. Lavigne, Department of Chemistry and Biochem-istry, University of South Carolina, Columbia, SC, USA

Dongwhan Lee, Department of Chemistry, Indiana Univer-sity, Bloomington, IN, USA

Daniel M. Lewallen, Center for Sensors and Biosensors,Department of Chemistry, University of Cincinnati,Cincinnati, OH, USA

Simon W. Lewis, Department of Chemistry, Curtin Uni-versity of Technology, Perth, Australia

Yang Liu, Department of Chemistry, Georgia State Uni-versity, Atlanta, GA, USA

Tomoki Ogoshi, Department of Chemistry and ChemicalEngineering, Graduate School of Natural Science andTechnology, Kanazawa University, Ishikawa, Japan

Manuel A. Palacios, Department of Chemistry, BowlingGreen State University, Bowling Green, OH, USA

Raman Parkesh, School of Chemistry, Trinity CollegeDublin, Center for Synthesis and Chemical Biology,University of Dublin, Dublin, Ireland

Mark A. Poggi, Q-Sense—A Division of Biolin Scientific,Linthicum Heights, MD, USA

Lin Pu, Department of Chemistry, University of Virginia,Charlottesville, VA, USA

Vincent M. Rotello, Department of Chemistry, Universityof Massachusetts, Amherst, MA, USA

Oleksandr Rusin, Department of Chemistry, Portland StateUniversity, Portland, OR, USA

Krishnendu Saha, Department of Chemistry, University ofMassachusetts, Amherst, MA, USA

Ken D. Shimizu, Department of Chemistry and Biochem-istry, University of South Carolina, Columbia, SC, USA

Robert M. Strongin, Department of Chemistry, PortlandState University, Portland, OR, USA

Weihong Tan, Department of Chemistry and Departmentof Physiology and Functional Genomics, University ofFlorida, Gainesville, FL, USA

Emma B. Veale, School of Chemistry, Trinity CollegeDublin, Center for Synthesis and Chemical Biology,University of Dublin, Dublin, Ireland

Hector Adam Velazquez, Department of Chemistry,Georgia State University, Atlanta, GA, USA

Marc Vendrell, Department of Chemistry & MedChemProgram of Life Sciences Institute, National Universityof Singapore, Singapore

Binghe Wang, Department of Chemistry and Centerfor Biotechnology and Drug Design, Georgia StateUniversity, Atlanta, GA, USA

Hui Wang, Department of Chemistry and Departmentof Physiology and Functional Genomics, University ofFlorida, Gainesville, FL, USA

W. David Wilson, Department of Chemistry, Georgia StateUniversity, Atlanta, GA, USA

Sheryl L. Wiskur, Department of Chemistry and Biochem-istry, University of South Carolina, Columbia, SC, USA

Hiroyasu Yamaguchi, Department of MacromolecularScience, Graduate School of Science, Osaka University,Osaka, Japan

Xiaochuan Yang, Department of Chemistry and Centerfor Biotechnology and Drug Design, Georgia StateUniversity, Atlanta, GA, USA

Yi-Cheun Yeh, Department of Chemistry, University ofMassachusetts, Amherst, MA, USA

Yagang Zhang, Department of Chemistry and Biochem-istry, University of South Carolina, Columbia, SC,USA

PREFACE

The field of chemosensing is experiencing a rapid growthin recent years. A large part of this is driven by the need inthe fields of medical diagnostics, environmental monitoring,and toxicological analysis, and the need to developprobes that allow for the in-depth understanding of therelationships between the presence of chemical or biologicalmarker and its biological implications. Chemosensingresearch requires knowledge in a wide range of subjectareas that span several disciplines. Therefore, we feel thata book addressing all key areas involved in chemosensingresearch will be very useful, especially to those new to thefield, such as graduate students and postdoctoral fellows.

This book covers the design, creation, and utilization ofchemosensors. A chemosensor in biology is broadly definedas a sensory receptor that transduces a chemical signal to anaction potential. However, in chemistry, and particularly insupramolecular chemistry, the term “chemosensor” is alsoassociated with the creation of receptors from nonbiologicalchemical entities. In contrast, a biosensor involves theuse of receptors based on biological structures, such aspeptides, proteins, and nucleic acids. The idea of creatingreceptors from nonbiological chemical entities most oftenmeans that chemical synthesis is used to create the unnaturalreceptors, which are therefore generally called “syntheticreceptors.”

A sensor in analytical chemistry is commonly definedto be a chemical indicator in an instrument that producesa signal indicative of the presence of an analyte. Insupramolecular chemistry, the sensor is often synonymouswith the receptor. In other words, if the color, emission,or redox potential of the receptor changes upon analytebinding, the receptor is defined as being a sensor. Of course,measuring and quantifying these changes still requiresinstrumentation, and, therefore, the two definitions are only

subtly different. In this book, the word sensor is usedin both ways, although, because supramolecular chemistryplays a large role in the majority of the chapters, the latterterminology is most prevalent.

The objectives of this book are to teach the novicesthe general thought processes, steps, and ultimate analysesperformed during the creation of both academically novel,as well as practically useful, chemosensors. The book givesseveral examples, along with details of the developmentprocesses that many of the leading practitioners in this areause. However, it is important to realize that the book wasintentionally written at a level that is meant to be read bysenior undergraduates and graduate students, and, therefore,can be used as a textbook. Our goal is to create a guide tothe creation of chemosensors, but to do so by highlightingcreative examples from leading research groups.

Corresponding to our goal of producing an educationalguide to chemosensor creation, we have organized ourbook to present the topic in a step-by-step progressionthat leads up to the sensors, and then culminates in thepractical utility of the sensors. The first section of thebook covers binding interactions used to create syntheticreceptors: solvophobic interactions, hydrogen bonding andother polar attractive forces, reversible covalent bonds,and metal chelation. Section 2 of the book illustrateshow to put these molecular recognition binding forcesinto receptors, either by creating new scaffolds or usingexisting scaffolds, along with computational chemistry,combinatorial methods, polymer imprinting, dendrimers,nanoparticles, and, finally, biological based receptors suchas nucleic acids and peptides. The book then turns toan examination of signal transduction, starting with twochapters that examine fluorescence techniques. A classicbook on the topic of chemosensors was published in

ix

x PREFACE

1993, and was specifically focused on this topic. It wasentitled Fluorescent Chemosensors for Ion and MolecularRecognition , by Anthony Czarnik. Our book also coverscolorimetric methods, electrochemical methods, surfaceplasmon resonance, quartz crystal microbalances, andarray-based methods, commonly referred to as differentialsensing. In the penultimate section of the book, we turn topractical utility, and present a series of case studies. Metalsensors for cellular imaging, anion sensors, small moleculeorganic sensors, bacterial analysis, and forensic science areall presented.

In summary, we believe that the future is bright forchemosensing using supramolecular chemistry principles.There should be no doubt that practical ramifications arewithin the grasp of the supramolecular community. Bybringing our knowledge of binding interactions, mecha-nistic organic and inorganic chemistry, electrochemistry,and photophysics to bear on analytical chemistry problemsin medical diagnostics, quality control, and environmentalanalysis, societal problems can be confronted and solved.Hopefully, this book will inspire chemists, young and old,to pursue this vision.

Eric V. AnslynBinghe Wang

SECTION 1

FORCES GOVERNING EXCHANGEABLE INTERACTIONS

Analytical chemists consider a sensor to be a chemicalindicator coupled with a device platform. The chemicalindicator undergoes either a covalent or noncovalentreversible reaction with the analyte. A pH indicatorinterrogated by a UV/vis spectrometer is a good example.When noncovalent interactions are involved, the approachfalls within the purview of supramolecular chemistry [1].Because supramolecular chemists focus on noncovalentbinding, they simply refer to the indicator as the sensor.Further, because supramolecular chemistry historically hasinvolved the creation of synthetic receptors, the receptorsthemselves are at the forefront of the efforts put intosensing protocols. Therefore, receptors that change opticalor electrochemical properties upon binding are referred toas the sensor by most supramolecular chemists [2].

At the heart of almost all mechanisms of molecularsensing is a binding event of some kind. Whether theseevents are protons on glass in a pH electrode, fleetinginteractions in a chromatography column, or a more discretemolecular recognition event such as an antibody–antigeninteraction, they all involve binding. Thus, it is notsurprising that the field of supramolecular chemistry is ofparamount importance to the field of chemical sensing.However, as discussed above, many sensors also involvecovalent bond formation with the analytes. For thisreason, a subfield of chemistry called SupramolecularAnalytical Chemistry is evolving [3], where the termsupramolecular is extended to include reversible covalentbonding, specifically when the dynamics of exchange aresuch that a signal is generated in response to an analytewithin the practical time scale. Therefore, when designingsensors, one should consider binding that is either reversiblecovalent or noncovalent, and these topics are the focusesof the first four chapters of this book.

There has been debate of whether or not ligand–metalbinding should be considered as supramolecular becauseof the strength of the bonding, and that many examplesare highly covalent. Yet, other forms of metal ligation areprimarily ionic, and can still be very strong. Irrespectiveof being covalent or noncovalent, or being classicallyconsidered supramolecular or not, these interactions clearlyinvolve binding that can be engineered to impart a signal forsensing purposes. Therefore, metal ligation is being widelyused in sensing protocols. Thus, we include a discussionon the basics of metal chelation chemistry as part of thetoolbox of forces for exchangeable interactions.

In summary, a focus of most molecular sensingapproaches is on binding the analyte. All imaginable bind-ing interactions can be exploited. Hence, hydrogen bond-ing, dipole interactions, electrostatic attraction, solvophobiceffects, metal chelation, cation–π forces, covalent bondswith low barriers of exchange, as well as any other bindingdriving force, can be exploited.

We start this book with an in-depth analysis of vander Waals interactions, dipole effects, and the hydrophobiceffect. The fundamental principles of these forces aregiven, followed by an analysis of the kinds of molecularreceptors that have been created to exploit these bindingforces. The next chapter moves to an analysis of bindinginteractions that have a higher ionic component: ion-pairing, hydrogen bonding, and cation–π interactions.Once again, the fundamental principles are presented, butthis is then followed by examples from biological systemsthat exploit these interactions. These first two chaptershighlight the classic noncovalent bonding forces used inthe creation of molecular receptors. The next two chaptersturn to reversible covalent bonding. Boronic acid–boronateester interchange is given considerable attention becauseit is arguably the most commonly used exchangeable

1

2

covalent bonding exploited in the creation of sensors.However, many others are rapidly gaining attention,including amine/carbonyl condensations, phosphorylations,and various nucleophilic additions to electrophiles suchas carbonyls and squaranes [4]. After covering theseinteractions in Chapter 3, the following chapter presents afocused treatment of metal chelation in sensor design. Onceagain, the basic principles are first discussed, followed byseveral practical examples of molecular sensors that exploitthis binding force.

REFERENCES

1. Lehn JM. Supramolecular chemistry - receptors, catalysts, andcarriers. Science 1985; 227(4689): 849–856.

2. Bell JW, Hext NM. Supramolecular optical chemosensors fororganic analytes. Chem Soc Rev 2004; 33(9): 589–598.

3. Anslyn EV. Supramolecular analytical chemistry. J Org Chem2007; 72(3): 687–699.

4. Hewage HS, Anslyn EV. Pattern-based recognition of thiolsand metals using a single squaraine indicator. J Am Chem Soc2009; 131(36): 13099–13106.

1VAN DER WAALS INTERACTIONSAND THE HYDROPHOBIC EFFECT

Bruce C. GibbDepartment of Chemistry, University of New Orleans, New Orleans, LA

1.1 INTRODUCTION

Within the field of supramolecular chemistry, two importantshifts in sensor development are currently under way.The first is the use of pattern-based recognition usingsensor arrays [1], the second is the continued shift awayfrom the nonaqueous phase to the aqueous realm. Theprincipal driving force for this second development hasbeen the goal of monitoring the biochemical milieu. Ittakes little imagination to envision an experiment in whichthe temporal resolution of the workings of a particularnetwork [2, 3] within a cell is attained by simultaneouslyaddressing a dozen small or macromolecule targets withspecific sensors. If not endless, the possibilities are at leastas complex as the human proteome.

Key to any functioning sensor is recognition of thetarget and the reporting of its complexation. The latteris dealt with elsewhere in this chapter. Here, the focusis on key design elements required for recognition inwater. A factor holding back the development of water-based sensors has been that the design principles forrecognition in water are very different from those in organicsolvents. In some respects, the knowledge gained from thestudy of supramolecular chemistry in nonaqueous solventshas acted as a solid foundation for studies in water.However, because many of the stalwarts of recognitionin organic solvents, for example, hydrogen bonding [4,5], are predominantly electrostatic in nature, they oftenstruggle to contribute to recognition in water; in the caseof hydrogen bonding, this is doubly so as competition fromhydrogen bonding to the solvent usually predominates over

Chemosensors: Principles, Strategies, and Applications, First Edition. Edited by Binghe Wang and Eric V. Anslyn.© 2011 John Wiley & Sons, Inc. Published 2011 by John Wiley & Sons, Inc.

hydrogen bonding between host and guest. Consequently,what has been learned from nonaqueous systems has toundergo a translation process if it is to be applied in aqueoussolutions.

If the polar environment of the aqueous solutiondiminishes noncovalent interactions such as hydrogenbonding or ion pairing, what can be relied upon forrecognition in water? In the simplest of terms, there aretwo important phenomena that can be utilized: van derWaals (vdW) forces [6–8] and the hydrophobic effect[9–14]. In the aqueous environment, vdW forces are notdiminished by competition, and as a result these weakforces can become increasingly important in the interactionsbetween host and guest. The other ally to complexationin water is the hydrophobic effect. Engendered by acombination of noncovalent forces—not least of which arevdW forces—the hydrophobic effect is most familiar asoil and water not mixing. Both these phenomena, vdWforces and the hydrophobic effect, are offshoots of nonpolarchemical structure and are key players in the ability to bringabout recognition in water.

1.2 CAVEATS

Two caveats need to be stated before we proceed further.First, in supramolecular chemistry the terms noncovalentforces and noncovalent interactions have been used eithersynonymously or to refer to specific phenomena, a fact thatcan lead to some confusion for the uninitiated. In regard tothe latter usage, the term forces is frequently used to refer

3

4 VAN DER WAALS INTERACTIONS AND THE HYDROPHOBIC EFFECT

Figure 1.1 Schematic representation of (left to right): monopole,dipole, quadrupole, quadrupole, octupole, and octupole.

to specific mathematical constructs defining the energybetween two chemical entities, that is, a “potential.” Thesepotentials vary considerably depending on the nature of theinteracting molecules. Figure 1.1 lists schematic represen-tations of the important poles found in chemical entities.

For reasons that will become clear, the focus here ison the different forces between monopoles and dipoles.Table 1.1 lists the different monopole and dipole forces thatoperate between molecules, and shows how the attractive,that is, long-range, component of the potential varies as afunction of distance (r).

As can be seen from the table, the attractive forcebetween two ions (monopoles) falls off proportionally with1/r . More precisely, this system is one that is describedby Coulomb’s law (Eq. 1.1), which relates the attractiveor repulsive energy (E) with the two charges q1 and q2,the distance between them (r), the relative permittivity (ordielectric constant) of the medium between them (εr), andthe vacuum permittivity (ε0). This equation applies strictlyto systems in which the electronic charges are static (henceelectrostatic). Thus, a pair of hard (nonpolarizable) ions isreadily modeled using only this interaction. That said, thisequation has also been used to describe systems in whichthe electronic charges on each entity are perturbed by theinteraction between them.

Note that in Eq. (1.1) the dielectric constant term is partof the denominator. As the dielectric constant increases,the interaction energy decreases. Thus, in water—which,with the exception of the small amide solvents andliquid HF, has the highest dielectric constant [15] (εr =78.3)—intermolecular forces that are electrostatic in natureare weakened. That said, this is the strongest of noncovalentforces. In a vacuum, two singly charged ions experiencean attraction less than the thermal energy (kT ) only whenthey are more than 52 nm apart! Put another way, when in

TABLE 1.1 The Major Classes of Monopole/Dipole Nonco-valent Forces, and How Their Interaction Energies Vary as aFunction of Distance (r)

Monopole Dipole Induced Dipole

Monopole 1/r 1/r2 1/r4

Dipole 1/r3 and 1/r6 a 1/r6

Induced dipole 1/r6

aFor fixed and freely rotating dipoles, respectively (see text).

contact, the interaction energy between the same two ionsis comparable to covalent bonding.

E = −q1q24πεrε0r

(1.1)

Contrastingly, the term interaction is often used to describestructurally well-defined supramolecular arrangements ofchemical entities such as the hydrogen bond [4, 5] orthe cation–π interaction [16]. Some of these interactions,such as the ion–ion interaction, are described by justone potential; the force or potential is the interactionand hence the synonymous use of the terms. However,most interactions require combinations of potentials tobe successfully modeled. To take one example, thecation–π interaction [16, 17] is composed mostly ofa monopole–quadrupole force; one estimation puts itscontribution at 60% of the overall energy [18]. However,the remaining 40% is a composition of other noncovalentforces such that the energy of interaction (E) varies not asa function of 1/r3 but as a function of 1/rn (where n < 2).In other words, in many respects the cation-π interactionresembles the Coulombic interaction. In such cases, theterms noncovalent interaction and noncovalent force havedifferent meanings and the interchangeability of the twoterms is less appropriate.

In short, when independently defined, the term nonco-valent interaction subsumes the term noncovalent force.The former are qualitative and quantitative summations ofthe many noncovalent forces that can exist between twochemical entities. However, the few cases when the twoterms are literally synonymous have resulted in the termsbeing used interchangeably irrespective of the nature of thesystem.

The second caveat to note concerns the hydrophobiceffect. As alluded to above, the hydrophobic effect betweenmolecules is just that, an effect. Nevertheless, it is stillfrequently referred to in the literature as a hydrophobicforce. There are two reasons as to why this is so. First, theterm, hydrophobic force arises from the notion that it is anapparent force. Second, research into the forces betweentwo hydrophobic surfaces reveals that the hydrophobiceffect can be felt by the two surfaces when they areas far apart as 100 nm [6]. This very long-range andapparently unique effect is a function of the size of eachsurface; the larger the surface area, the longer the rangeof the effect. For small molecules, however, it is arguablymisleading to describe the hydrophobic effect as a forcebecause it conjures the idea of there being this uniqueforce driving this phenomenon. This is not the case. Wateris a very special case, dare it be said unique, but itis becoming more and more apparent that the fact thatoil and water do not mix arises from a multitude of(normal) noncovalent forces, the most significant of which

CAVEATS 5

is hydrogen bonding between water molecules; betweentwo hydrophobic molecules in water, there is no utterlyunique force at work here, just the usual suspects workingin unusual ways. Finally, a related term hydrophobicinteraction also warrants a brief discussion. This near-synonymous term describes the unusually strong attractionbetween hydrophobic molecules (and surfaces) in water.Evidence has been accumulating for some time now [19]that this is just one specific example, albeit an extremeone, of solvophobicity. The solutes in question wouldrather “solvate” themselves rather than be solvated by thesolvent. As recognition between two or more moleculesin solution is always going to be in competition withrecognition of the individual molecules by solvent, froma supramolecular perspective the narrow term hydrophobicinteraction is not particularly helpful: all the more soin that it suffers from the same ambiguity discussedabove for well-defined supramolecular geometries thatare described accurately by the combination of severalpotentials.

1.2.1 van der Waals Forces

Intermolecular forces are loosely classified into three groups[6–8]. There are those that are purely electrostatic in originsuch as the just-discussed interaction between two ions.There are polarization forces, those that involve dipolesinduced by nearby ions or permanent multipoles. And thereare those that are quantum mechanical in nature such as theexchange interactions and the charge-transfer interactions.Even in such general terms, these classifications are neitherwell defined nor exhaustive. One of the main reasonsfor the overlap and “fuzziness” between these categoriesare the vdW forces. So what exactly are vdW forces?A perusal of the literature soon indicates that the answerto this question is not quite so clear-cut as may befirst assumed. We will return to this point very shortly,but suffice to say that they are among the weakest ofnoncovalent forces. Be that as it may, vdW forces areomnipresent in chemistry, and consequently are of immenseimportance to the bulk and supramolecular properties ofall molecules. The importance of vdW forces lies in thefact that even small molecules can make a large number ofvdW contacts and that each of these add up in a synergisticmanner. Thus, the heat of vaporization (�Hvap) of n-octane,n-nonane, and n-decane are comparable or greater thanthose of ethanol, methanol, and water, respectively. vdWforces are, however, frequently underappreciated. Whyis this so? Two culprits are apparent. First, in organicsolvents, other noncovalent forces often dominate theoverall supramolecular profile of a molecule; vdW forcesdisappear into the “background.” Second—and this pointwill be expanded upon below—vdW forces are viewed asundirected and difficult to control or model. There is some

truth to these two points, but they are not wholly true, andcertainly not so in water. Indeed, these preconceived notionsobtained from experiences in organic solvents in part standin the way of aqueous-based supramolecular chemistry.An appreciation of how disabling these assumptions areis essential to the goal of sensors that function in water.

Returning to the question of what exactly vdW forcesare, an examination of the literature reveals that the termhas been used in many different ways. One frequentlyutilized approach has seen it used interchangeably with“London dispersion forces” to mean noncovalent forcesthat arise between two induced dipoles. History supportseither of these terms. It was Johannes Diderik vander Waals who famously quantified these weakest ofinteractions to account for the fact that the noble gaseshave liquid and solid phases and show deviations fromideality, even though they are without multipole momentsand consequently cannot form dipole–dipole or dipole-induced dipole interactions. Subsequently, it was thetheoretical physicist Fritz Wolfgang London who firstshowed that the attractive vdW forces between suchatoms could be explained quantum mechanically. By thisnarrow definition, only intermolecular interactions betweennonpolar molecules such as alkanes involve vdW forces.

A second, slightly wider, definition is to classify allnoncovalent forces that are proportional to the inverse ofthe sixth power of the separation as vdW forces [6]. Bythis definition, there are (at least) three forces to consider:one that involves two freely rotating permanent dipoles (theso-called orientation force) and two that involves induceddipoles (the induction force and the dispersion force). Theemphasis here is on freely rotating ; as Table 1.1 shows,the distance dependence relationship for the energy ofinteraction of two dipoles is 1/r3 if the dipoles are fixedand 1/r6 if they are freely rotating. The former relationshippertains to interactions between stationary polar moleculessuch as in the solid or liquid-crystalline phases, and thelatter to the liquid and solution phases. This switch leadsto a minor semantic issue in supramolecular chemistrybecause, by this definition, dipole–dipole forces betweena host and guest are vdW forces if the guest can rotate, butnot so if it is fixed! However, using this broader definition,a much larger number of important noncovalent interactionsare classified as vdW forces.

In the next section, we review the three vdW forceswhose energy of interaction is inversely proportional to thesixth power of the distance. However, in order to paint abroader picture and put vdW forces in context, we will notlimit our discussion solely to these three forces.

1.2.2 A Brief Review of vdW Forces

Our understanding of noncovalent forces has come aboutby a combination of empirical data; the derivatization of


equations from first principles; and modeling based onempirical data, ab initio calculations, and first principles[20]. Although a comprehensive review of how ourunderstanding of noncovalent forces has been built up isbeyond the scope of this review, it is helpful to examineselected examples as they illustrate the important factorsdefining the interaction energy (E) between poles.

Figure 1.1 shows schematic representations of mono-poles through octupoles. Rather than examining all thepotential interactions between these poles, we focus hereon the three vdW forces (E ∝ 1/r6) from the list ofmonopole–dipole forces in Table 1.1: the so-called orien-tation, induction, and dispersion forces. In other words, wedo not specifically discuss quadrupoles, such as that foundin benzene, and octupoles typified by methane. This is notto diminish their importance. The quadrupole–quadrupoleforces are responsible for π –π stacking of aromatic rings.The large permanent quadrupole of the benzene ring iswell known to form both edge-to-face and slip-stackedsupramolecular structures. Furthermore, pairs of electron-ically dissimilar aromatic compounds, such as benzene andhexafluorobenzene, are also known to form (in register)stacked structures. However, many of the forces involvingthese poles do not come under the category of vdW forces(E ∝ 1/r6). For example, the monopole–quadrupole force,dipole–quadrupole force, and the quadrupole–quadrupoleforce vary as E ∝ 1/r3, 1/r4, and 1/r5, respectively.That said, other forces involving quadrupoles, such asthe quadrupole-induced dipole force, can be classified asvdW forces. Furthermore, some of the conceivable interac-tions between two fixed aromatic rings also demonstrate anE ∝ 1/r6 relationship. However, the complexity of theseinteractions is considerable, and so successful models ofthe likes of π –π stacking have focused on approximatingelectrostatic and vdW forces. We therefore do not developthis further except to direct readers to articles that con-sider the modeling of π –π stacking of aromatic rings [20,21]. The reader should note, however, that all this stated,aromatic rings are important recognition motifs in water.Although the quadrupole of benzene allows it to participatein many electrostatic interactions, for example, with cations,water, and ammonia, benzene and many aromatic com-pounds are nevertheless hydrophobic; by way of example,the association constants for the dimerization of benzeneand cyclohexane in water have been recorded as 2.0 and 2.7kcal/mol, respectively. We therefore see many examples ofwater-soluble hosts comprising aromatic rings (see below).

Looking again at Table 1.1, it is useful to remember thatthere is a relatively simple way to derive the relationshipbetween distance (r) and the interaction energy (E)between two permanent and freely rotating poles. Thus, ifan n-pole is defined as an array of point charges with an n-pole moment but no lower moment, that is, a monopole(n = 1) is a point charge; a dipole (n = 2) is an array

of charges with no monopole moment (overall charge);a quadrupole (n = 3) is built up from an array of pointcharges that has neither a dipole moment nor an overallcharge; and an octupole (n = 4) is an array of point chargeswith no quadrupole, dipole, or net overall charge, then thedistance–energy relationship between two permanent polesof pole moments n1 and n2 is given by Eq. (1.2):

V ∝ 1rn1+n2−1

(1.2)

We begin our review of vdW forces with that “borderline”case of the dipole–dipole force. As Table 1.1 shows, whentwo dipoles are fixed relative to one another, the interactionenergy between them varies with 1/r3. Figure 1.2 showsthe case at hand. Even in this straightforward system,constraints must be included; otherwise, the derivationsbecome quite complex. Thus, the dipole separation (r) mustbe much greater than the dipole length (d), and the dipoleskept in parallel. One angle, the angle θ , will be varied.

Equation (1.3) relates the distance–interaction energyrelationship for two such dipoles:

E = −μ1μ2(3 cos2 θ − 1)

4πεr ε0 r3(1.3)

where μ1 and μ2 are the dipole moments in question, εris the dielectric constant of the medium, and ε0 is thevacuum permittivity. As can be readily envisioned fromthe 3 cos2 θ − 1 term, the maximal attractive interactionoccurs when the two dipoles are in line (θ = 0◦). Thereadily derived Eq. (1.4) defines this case. Likewise, whenthe two dipoles are aligned such that θ = 90◦, the energyof interaction is half that of maximal (Eq. 1.5). Betweenthese two angles is the “magic angle” of 54.7◦, where theinteraction energy between any pair of dipoles is zero.

E = −2μ1μ24πεrε0r3

(1.4)

E = −μ1μ24πεrε0r3

(1.5)

+q1 −q1

+q2 −q2

d

θ

r

Figure 1.2 Spatial relationship between two dipoles.

CAVEATS 7

Equation (1.3) also tells us that when two molecules in avacuum possessing dipoles of 1D are held in line and 0.36nm apart, their interaction energy will equal kT . Hence,at normal temperatures this interaction alone is sufficientto maintain the liquid state only for very polar molecules.Furthermore, unlike ion–ion or ion–dipole interactions, inthe liquid state these forces alone are generally not sufficientto fix two molecules in a defined orientation (although wateris an exception to this).

Two freely rotating dipoles lead to the first of threevdW forces (as defined by E ∝ 1/r6), the orientationforce, or (Willem) Keesom interaction. The derivationof this equation is more complex. The angle-averagedor Helmholtz free energy is given by the potentialdistribution theorem and leads to Eq. (1.6). The E ∝ 1/r6relationship is a result of the inverse cube of the interactionpotential weighed by the Boltzmann term (which similarlyis proportional to the inverse cube of the separation).Additionally, the inverse dependence on the absolutetemperature (T ) arises because thermal motion overridesthe mutual orientation effects of the two dipoles at elevatedtemperatures. Put another way, under normal temperatures,the Keesom interaction is not strong enough to induce anyorientation effects in liquids.

E = −2μ21μ

22

3(4π εrε0)2 r6 kT(1.6)

The other two vdW forces are those that involve theinduction of dipoles. The first of these is the induction forceor Debye interaction, and comes about by the polarization(polarizability (α) is defined as the strength of the induceddipole moment (μ) acquired in a field of strength E, i.e.,μ = αE) of a nonpolar molecule by an adjacent moleculepossessing a permanent dipole. There are two kinds ofinduction force: electronic polarizability and orientationalpolarizability. Nonpolar molecules only demonstrate theformer, whereas dipolar molecules show the latter, whosetime-averaged dipole moment is zero; in effect, the externalfield changes the Boltzmann-averaged orientations of therotating dipole. When the inducing field is an ion, aninduced dipole as large as 1D can be generated. However,induction brought about by a permanent dipole of anotherinduced dipole results in a much weaker temporary dipole.

As Eq. (1.7) shows, the interaction energy (E) betweena dipole and an induced dipole also varies with inversedistance to the sixth power (1/r6). As with the Keesominteraction, the Debye interaction is not sufficiently strongto mutually orient the two molecules:

E = −μ21α

22

(4πεr ε0)2 r6(1.7)

where μ1 is the dipole moment of the permanent dipoleand α2 is the polarizability of the neutral molecule partner.

On a related note, a pair of dipolar molecules canmutually bring about induced dipoles in each other. In thisscenario, the net dipole-induced dipole energy is given bythe related equation (Eq. 1.8):

E = −μ21α2μ

22α1

(4πεr ε0)2 r6(1.8)

The third kind of vdW forces that demonstrate theE ∝ 1/r6 relationship are the dispersion forces or vdWforces. vdW forces, also known as London forces, charge-fluctuation forces, electrodynamic forces , or induced dipole-induced dipole forces , act between all atoms and molecules.The term dispersion force comes about from their involve-ment in the dispersion (the phenomenon in which the phasevelocity of a wave depends on its frequency) of light inthe visible and UV regions of the spectrum. Dispersionforces are found between molecules and surfaces. In theformer case, they are always attractive between identicalmolecules, but can be either attractive or repulsive betweendissimilar molecules. When dispersion forces are attrac-tive, they do not strongly align or orientate the two speciesinvolved. Overall, they are frequently the most importantof the three vdW forces because uniquely they are alwayspresent and are usually stronger than those forces aris-ing from dipoles (but see below). In vacuum, they arelong-range forces, influencing how two chemical entitiesinfluence one another from 0.2 nm out to more than 10nm, but their effect is greatly attenuated in the solutionphase. Importantly, however, in solution dispersion forcesare nonadditive; there is synergy in this force.

Although intuitively straightforward to understand—aninstantaneous dipole generates an electric field that polar-izes the atoms or molecules in its vicinity—the quantummechanical nature of these forces means that they are com-plex to model. A semiquantitative analysis based on theinteraction between two Bohr atoms leads to the Londonformula (Eq. 1.9):

E = −3α20hv

(4πε0)2 r6= −3α

20 I

(4πε0)2 r6(1.9)

where α0 is the electronic polarizability of the second atom,h is the Planck constant, ν is the orbiting frequency ofthe electron, and I is the ionization energy. This can berewritten for two dissimilar chemical entities in solution(Eq. 1.10):

E = −32

α1α2I1I2

(4πεrε0)2 r6(I1 + I2) (1.10)

It should be noted that this equation is a good approxi-mation for spherical entities of less than 0.5 nm diameter.Thus, for many receptors or hosts, that is, those that canbind relatively large (>10 non-hydrogen atoms) guests, this


equation breaks down. More importantly, Eq. (1.10) alsobreaks down for asymmetric (nonspherical) molecules: afact that discounts its application to every molecular hostever synthesized. Furthermore, simplifications that attemptto determine the overall vdW force in a system by deter-mining set values for defined moieties—for example, thetotal internal energy (U ) for a methylene group has beencalculated to be ≈ 6.9 kJ/mol—cannot be applied in solu-tion because of their nonadditivity. Consequently, they areusually modeled using empirical force fields such as theLennard–Jones (6–12) potential (Eq. 1.11) [20]:

V (r) = 4 ∈[(σ

r

)12 − (σr

)6](1.11)

where ∈ is the depth of the potential, σ is the (finite)distance where the interparticle potential is zero, and r isthe distance between the particles. As may be anticipated,the attractive dispersion term is defined by the 1/r6 term,whereas the repulsive 1/r12 term accounts for short-rangeexchange repulsion. It should be noted that exchangerepulsion can be fitted with a variety of 1/r terms.Exponential terms work equally as well, as do much“harder” terms. The reason for the popularity of the 6–12potential is that the repulsive factor can be calculated bysquaring the attractive 1/r6 term, which makes calculationswith the Lennard–Jones potential much quicker.

1.2.3 The Sum of the vdW Forces

In the previous section we highlighted three types of vdWforces: the orientation, induction, and dispersive forces. Inparticular, the quantum mechanical nature of the dispersionforce makes it difficult to model, and as a result Eq. (1.10)is a good approximation for spherical entities of less than0.5 nm diameter. Building on this, McLachlan formulateda more rigorous expression for dispersion forces [22], andcombined this with the other vdW forces to give an equation(Eq. 1.12) that relates the total vdW energy between twoidentical molecules (1) in a medium (3):

E ≈ −[

3kT

(ε1(0) − ε3(0)ε1(0) + 2ε3(0)

)2+

√3hve4

(n21 − n23)2(n21 + 2n23)3/2

]

×a61

r6(1.12)

where ve is the frequency of the (assumed to be) one strongabsorption peak of the solution, ε(0) and n are, respectively,the static dielectric constant and refractive index of thesolute and solvent, a is the radius of the solute (a1 � r),and k and h are the Boltzmann and Planck constants,respectively.

In Eq. (1.12), the first major term (populated withdielectric constant terms) corresponds to the contribution

from the induction and orientation vdW forces, and thesecond (populated by refractive index terms) correspondsto the contribution from the dispersion vdW forces. Thisequation highlights two important points about vdW forcesin solution. First, as hv e is much larger than kT, thedispersion term is usually greater than the inductiveand orientation components. Second, in the case oflow-molecular-weight alkanes associating in water, thedispersive term is very small (water and small alkanes havevery similar refractive indexes). Furthermore, because thelarge difference between the static dielectric constants ofwater and alkanes (≈ 80 verses ≈ 2), the first inductive andorientation term reduces to Eq. (1.13).

E ≈ −kT a61

r6(1.13)

Thus, the energy of interaction in this case is purelyentropic. This is indicative of an increase in the freedom ofwater molecules upon the dimerization of the alkanes, andis a hint of the hydrophobic effect. However, the valuesobtained from Eq. (1.13) are typically much smaller thanthose determined experimentally, which are usually twoorders of magnitude greater than predicted. This breakdownof the model is attributed to the excess polarizabilityinvolving the highly polar water.

1.3 HIGH-DEFINITION MOLECULARAND SUPRAMOLECULAR STRUCTURE

Key to the recognition of a guest molecule is a structurallywell-defined host that presents atoms and groups in acomplementary array. In organic solvents, this array isoften composed of hydrogen bonding groups, but thisnoncovalent force is often incapable of bringing aboutbinding in water. For the development of aqueous-basedsensors, one way to circumnavigate this recognition issueis to use hydrophobic groups such as alkyl or aryl groupsto define the binding site. We discuss some of the issuesassociated with this strategy here.

In the synthesis-dominated field of organic chemistry,saturated hydrocarbons are often viewed as being uninter-esting, primarily because historically they have representedsuch an inhospitable desert for synthetic chemists. Thisis now changing with new approaches to C—H bondactivation, but it is likely to take some time for textbooksto consider hydrocarbons as functional groups. Even insupramolecular chemistry, hydrocarbon chains are oftenviewed as linker groups joining more interesting parts of amolecule. However, in the aqueous solution, hydrocarbons,be they saturated or unsaturated, definitively take on afunctional role; they function via the hydrophobic effectand vdW forces to give a target or a target complex thedesired structure.

HIGH-DEFINITION MOLECULAR AND SUPRAMOLECULAR STRUCTURE 9

When thinking of structure, it is convenient to think ofthe degree of organization. Thus, micelles, liposomes, andtheir ilk [14, 23, 24] are partially ordered in so much asthey show well-defined structure at the micrometer scale,but little or no structure at the molecular level. The physicalorganic chemistry of such entities is relatively well devel-oped. In contrast, high-definition molecular/supramolecularstructure, that is, where there is good structural definitionat the nanometer scale and below, is less well understood.Nature provides a multitude of examples to be inspired by,but the noncovalent chemistry at the atomic level behindthe tertiary and quaternary structures of proteins [25] isrelatively uncharted.

So how is high-definition structure, at both the molecularand supramolecular level, attained? Obviously, covalentbonds and the constitution of a target have a role to play,but what about at the intra- or intermolecular level? Oneof the key issues to address is directionality. For example,hydrogen bonds or metal coordination is frequently usedto give a molecule, or a supramolecule, a well-definedstructure. This is particularly true in regard to self-assembly,where groups capable of hydrogen bonding [26–32] ormetal coordination [33–45] are powerful supramolecularmotifs. But some translation—from the organic phaseto the aqueous phase—is required here. When we talkabout a supramolecular motif, or more generally a moietycapable of forming well-defined noncovalent interactionsas having “directionality,” we are unconsciously thinkingabout enthalpy. For example, to instill high structuraldefinition within a molecule or between molecules, wemight utilize a hydrogen bond because as the angle betweendonor and acceptor is varied so the enthalpy of interactionchanges. It is the considerable variation in the enthalpyof interaction as the D—H · · · A bond angle is variedthat leads to the description of hydrogen bonds being“directional.” Thus, by careful design of the molecularstructure, supramolecular structure can be ensured becausethe molecule or molecules seek out the lowest freeenergy (with enthalpy playing a major role). In organicsolvents, this is intuitive because any kind of high-definitionstructure, be it the ordering of a chain of atoms or thebringing together of multiple molecules, is not entropicallyfavored. In short, in organic solvents we rely on “powerful”and “directional” noncovalent forces to drive assembly, andenthalpy is the underlying thermodynamic premise. Butwhat about entropy? Are there moieties that either throughentropy alone or with the aid of entropy lead to high-definition structure? In water, where the hydrophobic effectcomes into play, the answer must be a resounding yes! Evensmall proteins such as ubiquitin (76 residues) possess awell-defined structure with a hydrophobic core. Likewise,the crenellated profile of the hydrophobic strip along thelong axis of the leucine-zipper monomer leads to a coiled-coil dimer of precise structural definition. A third example

comes in the form of virus capsids. These exceedingly well-defined protective shells are perhaps the most inspirationalexamples of protein quaternary structure. Composed of, insome cases, thousands of copies of one (or more) protein,these macromolecules not only fold into precise shapesbut those shapes also define perfectly formed hydrophobicpatches at specific relative orientations to bring about self-assembly. In many, if not most of these cases, enthalpyplays a significant role. But it is also clear that entropy playsa very significant role, and perhaps in some examples even apredominant role. How does one define such precise shape-complementarity? And how is the entropy of associationand the hydrophobic effect influenced by concave andconvex curvature, or scale of these shapes, or indeed theway these shapes are connected to give an overall form?How is the entropy of association influenced by the atomtype or hybridization state? These are exceedingly difficultquestions to answer, and to do so will reveal much of whatis not known about the hydrophobic effect. We discuss whatis known about the hydrophobic effect next.

1.3.1 The Hydrophobic Effect

Water is a ubiquitous molecule. Nevertheless, many of itsproperties are still poorly understood. Indeed, the manymysteries and unknowns surrounding this solvent probablyhelp explain polywater [46] and memory water [47], twoexcellent examples of pathological science.

An example of a real (and emergent) phenomenonfrom water is the apparent force commonly known as thehydrophobic effect [9, 10, 12–14, 48, 49]. Nature utilizesthe hydrophobic effect in countless ways. A nonexhaustivelist includes the assembly of phospholipids to form vesicles,the folding and assembly of proteins, the assembly andstructure of duplex DNA, and of course the binding ofeffectors or substrates to proteins, enzymes, and ribozymes.How do these assembly and recognition phenomena comeabout? The short answer is that Nature has learned howto harness both enthalpy and entropy to define structure;the long answer is still being formulated by many researchgroups.

Although an intense area of study, why oil and water donot mix is still not fully understood. What is known aboutthe effect is that the two physical properties responsiblefor it are that liquid water is in close phase coexistencewith its vapor, and that water–water interactions are muchstronger than those between water and a hydrocarbon. Froma thermodynamic perspective, it is also known that thehydrophobic effect is more often than not promoted byentropy (although this is not always the case), and thatthe best thermodynamic signature for the manifestation is adrop in the heat capacity of the solution as the hydrophobicsolute is desolvated. More on these points below, but onefrustrating characteristic of the hydrophobic effect is the


large amount of seemingly contradictory thermodynamicdata that has been garnered to date. Fortunately, theamount of data gathered is now sufficiently large thatthis phenomenon can be explained; any conclusions froman experiment will depend on how and what is studied!Thus, in the first instance, the hydrophobic effect is sizedependent [11]. As we will see, small and large solutes (aswell as surfaces) do not behave the same way. Second,the shape of the solute also appears to be important tothe observed thermodynamics [50]. Third, the hydrophobiceffect is temperature dependent, with cold water leadinggenerally to an entropically driven hydrophobic effect andhot water leading to an enthalpically driven process. Thenet effect of these (and other) factors is that the admixtureof noncovalent forces behind the effect is marshaled indifferent ways. There are multiple ways in which thehydrophobic effect can be brought about, and the narrowwindow into the effect provided by any one experimentprovides a poor view of the overall vista.

Although there is a continuum of thermodynamic pos-sibilities driving the hydrophobic effect, that is, processespromoted by enthalpy and entropy (�H −ve/�S +ve),those promoted only by enthalpy (�H −ve/�S −ve), andthose promoted only by entropy (�H +ve/�S +ve), insupramolecular chemistry generally two kinds of hydropho-bic effect are referred to: the classical hydrophobic effect(entropically favored) and the nonclassical hydrophobiceffect (entropically penalized). Regarding the enthalpyaspects of the hydrophobic effect, when a hydropho-bic solute is expelled from the aqueous phase, strongwater–water interactions (the cohesive force) are increasedat the expense of weaker water–hydrocarbon interactions.This is the major enthalpic contribution to the hydropho-bic effect (a minor enthalpic role can also be played bythe polarizability of hydrocarbons vs the lack of polariz-ability of water). With its two hydrogens to donate andtwo lone pairs to accept, water simply prefers to stronglyhydrogen-bond to itself rather than weakly hydrogen-bondto the solute. But it is not just a matter of differencesbetween hydrogen bond strengths. When desolvating, say,a cyclohexane ring, each water molecule in the first solva-tion shell must either sacrifice some of its strong hydro-gen bonds to other waters, or “fix” its orientation toform the same number of hydrogen bonds as bulk solu-tion. Pushing the cyclohexane out of solution avoids thedilemma of either sacrificing hydrogen bonds or adoptinga fixed orientation.

There is another thermodynamic perspective to thejust-described situation of the solvation shell. In essence,the water molecules of this solvation shell are orderedaround a hydrophobic solute, and if the solute can besqueezed out of solution, that is, bind into a hydrophobicenvironment, then these waters are liberated to the bulk.This is believed to be the primary component of those

most frequent of cases when the hydrophobic effect ispromoted by entropy. As has been pointed out [10], thisnotion is correct to a first approximation, but one shouldtake care not to invoke the idea of some sort of clathratedwater structure around the solute. Clearly, such an extremeview is incorrect; the dynamics of water structure is on thefemtosecond timescale, and so the suggestion of clathratesaround hydrophobic solutes is one step toward polywaterand memory water. Nevertheless, a corollary of “ordered”water is that binding or self-assemblies in aqueous solutionneed not rely on enthalpy; entropy can also be harnessed.

How does the nature of the solute affect the overallthermodynamics of the hydrophobic effect? By modelingsolutes as spheres of different sizes, a good deal has beenlearned about how size influences the hydrophobic effect[11, 51]. Briefly, in terms of the number of hydrogen bondsthat each water molecule can form, water molecules takefar less “notice” of small solutes; they are able to forma solvation shell around the solute that is reminiscent ofbulk water but less fluxional, that is, more ordered. In thissituation, the solute surface is said to be wetted: that is,the water density immediately surrounding the hydrophobicsolute is greater (up to twice) than the bulk. In such cases,entropic changes are important in the desolvation process.In contrast, water cannot form a complete hydrogen bondnetwork around large solutes; each water molecule loseson average almost one hydrogen bond. As a result, thesurface of the large solute is dewetted, and the water densityimmediately surrounding the solute is less than that of thebulk. Hence, from a supramolecular perspective, a (large)host binding a (small) guest is favorable because of thedifference between the entropically dominated solvationfree energy of the guest, and the change in the enthalpicallydominated solvation free energy of the free host and thehost–guest complex.

Host and guests, like all molecules, are not of coursespherical. Consequently, the aforementioned effect of thesize of the molecule has layered upon it the issue ofshape. It is readily apparent that the shape of a moleculedrastically adjusts the solvation shell. Simply bending aflat surface into either a concave or convex surface willdrastically change its solvation. In this regard, models ofthe solvation of a convex surface, which corresponds to thesurface of a protein, have been studied in more detail thanthe solvation of convex surface [52]. There is, however,much still to learn about curved surfaces. Furthermore, ifyour imagination takes you to more complicated structures,perhaps twists to the surface or the addition of protrusionsand depressions, then you are in essentially unchartedterritory. And this is for a surface that is of homogeneoushydrophobicity!

So far, we have only mentioned one noncovalentinteraction, the hydrogen bond. What role do vdW forcesplay in the hydrophobic effect? In pure water, vdW forces


have little or no effect on the density of the bulk. Water isnot easily compressed and so these weak forces have littleeffect. Instead, it is hydrogen bonding that has a majorinfluence on the structure of the liquid. But vdW forces doinfluence the hydrophobic effect. The solvation shell arounda large solute has a lower density than bulk water, andthis shell is frequently described as being “soft” becauseextended interfaces or fluid structure can be translated withlittle change in free energy. As a consequence of thislow density and softness, vdW forces can pull the liquidinterface into contact with the hydrophobic surface. Thisleads to a partial wetting of the solute as the density ofthe water at the solute interface increases relative to theimmediate surroundings; hence, as a large solute moleculeis approached, water density is first seen to decrease, butthen increases slightly at the interface with the hydrocarbon.Large solutes are not, however, wetted in the same way assmall ones. At a wet interface, there is much less in the wayof density fluctuations. However, for the dewetted interfacethe slight increase in water density immediate to the solutedoes not prevent significant fluctuations in density fromoccurring.

Finally, a word must be said about the most reliablecharacteristic of the hydrophobic effect: the negative changein heat capacity as a solute moves out of the aqueous phase.Why does it take more energy to raise the temperature ofan aqueous solution of a hydrocarbon than to raise thetemperature of pure water? The basis of this observationagain appears to be the ordered water molecules arounda solute. At a lower temperature, this ordered solvationshell is populated mostly by waters in a low-energy state.At higher temperatures, a “melting” process occurs, whichresults in the water molecules adopting higher energy states.This process effectively acts as a heat storage mechanism,and so more energy must be added to the solution in orderto increase its temperature by a set amount. In contrast,bulk water molecules are not able to access higher energystates to the same degree and so warming the solution isless energy intensive.

In summary, our understanding of the hydrophobic effectcontinues to improve. The structure and dynamics of thesolvation shell of a hydrophobic solute is crucial to thebalance between enthalpy and entropy, which in turn isintimately tied to the structure of the solute. Nevertheless,a lot is still not known about the hydrophobic effect,a fact that is conveniently summarized in computationalwork. Thus, although many models faithfully reveal thefree energy change of the association of two hydrophobicentities, a lot fewer successes have been marked upin regard to accurately determining the enthalpy orentropy changes (first derivatives of the initial experiment).Additionally, success in modeling heat capacity changes (asecond derivative of the experiment) has been limited tospherical representations of methane [53–56].

1.3.2 General Principles for Receptor Design

What are the essential requirements for the complexationof guests in aqueous solution? An important rule for hostdesign is that the greater the degree of guest envelopment,the greater the observed selectivity. Sometimes a familyof molecules is the target and high selectivity is notsought. However, even if this is the case, for binding inwater guest envelopment is still essential for harnessingthe hydrophobic effect or desolvating a group that willhydrogen-bond to the bound guest. With this in mind, awater-soluble host needs three structural features. First,it must have a highly polar exterior to bestow it withwater solubility. Second, it must have a binding site thatis sufficiently large and suitably functionalized for thetarget. Our focus here is on relatively hydrophobic guests,which means that the host must be amphiphilic: not in thesense of (pseudo) one-dimensional long-chain fatty acidsor (pseudo) two-dimensional β-sheets in protein structure,but in a three-dimensional sense. The host must structurallymimic an enzyme and be water soluble on the outside, andhydrophobic on the inside. The third feature required of ahost is that it should be noncollapsing. A guest will alwayshave to compete with a pocket that spontaneously collapsesin aqueous solution, and binding will be nonexistent if theequilibrium between the collapsed and noncollapsed host istoo far toward the former.

In terms of synthesis, there are three general strategiesthat can be applied [57]. For all three, it is important toremember that a synthetic route should be as minimalisticas possible. The first approach is to design a polymerto fold to the requisite form possessing a binding site.This is precisely the same strategy that Nature uses toform proteins, and in fact α-amino acids could be usedas building blocks when the protein-folding problem hasbeen solved. It is also possible to use nonnatural subunitsto build a polymer (or foldamer) that possesses fewerdegrees of conformational freedom and so will fold morepredictably. One trade-off of the foldamer approach is thatthere is less control of the shape of the binding site, buta potentially important advantage with this strategy is thatthe dynamical nature of the host may be utilized in guestrecognition and signaling. A second strategy is to synthesizecavity-containing hosts. This approach might involve anadvanced starting material such as a cyclodextrin, or mayrequire synthesis from more basic starting materials. Eitherway, an advantage that this particular strategy has over thepolymer approach is atom economy. The third approach isa variation on the second, and aims to circumnavigate someof the synthetic steps in the synthesis of a host by utilizingself-assembly in the final step of a supramolecular species.This is a particularly useful strategy if the target for the hostor sensor is large, because the synthetic requirements of thecorresponding molecular host may be very demanding.


1.3.3 Common Classes of Water-Soluble Hosts

In this last section, we briefly review some examples ofwater-soluble hosts that are known to bind (relatively)nonpolar guests. The examples given are not intended tobe a comprehensive review of the literature, but rather givean idea of the structural variation available to designers ofsensors. Additionally, the focus is toward hosts that bindsizeable guests, by which we mean those with more thansix non-hydrogen atoms. Readers wishing to garner a fullerpicture of the different water-soluble hosts that are availableare directed to the many reviews cited in this section.

Foldamers, wholly nonnatural polymers that fold intowell-defined (tertiary) structures, constitute the first classof hosts under consideration. There is an expansive list offoldamers designed for a wide range of applications, includ-ing cell permeation, membrane disruption, discrete molec-ular recognition, and selective reaction with encapsulatedguests [58–63]. Likewise, a wide variety of subunits havebeen used to build foldamers, including steroids [64–66],aromatic oligoamides [67], β-peptides [58, 61, 62], andm-phenylene ethynylenes [63, 68–70]. There are manyexamples of foldamers that are water soluble [62, 71–74],but in terms of hosts designed to bind guest molecules, theyare relatively few. Indeed recent efforts in this area havefocused on the reverse: that is, the encapsulation of polarmolecules for solubilization in organic solvents or trans-port across bilayers [64, 67, 75, 76]. An example of thecreation of a hydrophobic pocket in a foldamer from theMoore group is the m-phenylene ethynylene 1. Compound1, although not soluble in pure water (analogous water-soluble derivatives have been reported [71]) is of interestbecause, by folding into a helical structure, it defines ahighly integral hydrophobic pocket (or bore) capable ofbinding bicyclic guests such as pinene. It has also provenpossible to add functionality (R’) to the bore so that gueststhat are of the correct size can enter the bore and can bechemically transformed at accelerated rates [69].

RO O

N3Et2Me3Si

R = –(CH2CH2O)3CH3

(1)

R′ n

A more common approach for hosts capable of bindingguests in aqueous solution is to synthesize nonpolymerichosts that possess, by dint of their constitution, enforcedcavities for guest binding. Like all strategies, the examples

most frequently used are those that are most easilyaccessible, and, not surprisingly then, the most popularfamily of hosts in this category is the cyclodextrins (2).These naturally occurring water-soluble oligosaccharidesare composed of a belt or torus of α-(1–4)-linked d-glucopyranose units. The most common examples, α-, β-,and γ -cyclodextrins, are respectively composed of six,seven, and eight sugars, and possess tapered cavities 8 Åin depth and 5.7, 7.8, and 9.5 Å in diameter. Because theyare open at both ends, their cavities can be described aspseudohydrophobic; the interior is akin to that found inthe butanol phase. Additionally, this toroidal architectureallows guest exchange to occur via either associative ordissociative mechanisms [77].

O

HO OH

OH

n

(2)

The rims of cyclodextrins are composed of 1◦ and 2◦

alcohols, which offer excellent points for functionalization.Countless cyclodextrin derivatives have been synthesizedby modification of either or both of these rims [78, 79],and regioselective chemistries for poly functionalization ofeither rim have been established [80]. Added functionalgroups generally point away from the cavity and thereforeit is possible to add functionality and not modify binding.However, frequently caps are added to the cyclodextrin tomodulate their complexation properties [80].

The range of guest molecules that can bind to cyclodex-trins is extensive [81–84]. The smallest, α-cyclodextrin,has been shown to bind acetonitrile, benzene derivatives,cyclohexanes, naphthalenes, alkyl derivatives up to 10 non-hydrogen atoms, sugars, adamantanes, and pyrene. For longguests such as 1,8-octanediol, the guest must necessarilyprotrude out of the two ends of the host. β-Cyclodextrinbinds the same type of guests, but the continuum ofguest sizes is shifted such that the smallest aforemen-tioned guests do not bind, while the larger ones generallybind more strongly. Likewise, for γ -cyclodextrin, whichdoes not bind acetonitrile and binds pyrene more stronglythan β-cyclodextrin. This wide binding profile has allowedcyclodextrins to be used as enzyme mimetics [85–90], incatenanes and rotaxanes [91], and of course as a componentto sensors [92–94].

A second class of water-soluble hosts belong to theextended family of calixarenes (3) [95–100] and resor-cinarenes (4) [96, 101–103]. Calixarenes come in many dif-ferent sizes (3, n = 4–20), but by far the most common are


the calixa[4]arene, calix[6]arene, and calix[8]arene (3, n =4, 6, and 8, respectively). There are two principal rea-sons for this. First, protocols for high-yielding synthesesof the cone conformation of these derivatives are avail-able. Second, the larger, ropey calixarenes have poorlydefined cavities. On a related note, rotation of the phenolrings around the chain of atoms defining the macrocycleis possible even for the smallest of calixarenes, and forthis reason the desired conformer is usually locked by thereplacement of the H atoms of the lower rim phenol groupswith alkyl or aryl groups. Protocols for replacing the t-butyl group of the upper rim (R = t-Bu) in the parentcalixarenes, which is a necessary blocking group during thebase-catalyzed condensation of the phenol and formalde-hyde, are also well developed. The variability in n, R, andthe phenol substituent provides a rich class of hosts withdiverse applications, including sensing [104–106].

OH

R

n

(3)

HO OH

Rn

(4)

Although a variety of different-sized resorcinarenes havebeen reported (n = 4, 5, 6, and 7), only the resorcin[4]arene4 (n = 4) is accessible in high yield by the acid-catalyzedcondensation of resorcinol and aldehydes. Predominantly,it is the cone conformation with the defined pocket thatis formed in these reactions, although the nature of the Rgroup in the aldehyde (and hence the R group in the resor-cinarene) can lead to other conformers [103]. HydrophilicR groups are required to make resorcinarenes appreciablywater soluble, but a number of the water-soluble hostsbased on resorcinarenes are obtained by building “up” fromthe core to engender hosts with deeper and more definedbinding pockets. Two examples of these bowl-shaped“cavitands” are shown below (5 and 6). Compound 5 isone example of a deep-cavity cavitand from the Rebekgroup, which has been shown to bind a variety of guests[107–109], whereas octa-acid 6 [110] or related dendritichosts [111] from the Gibb group has been shown to bindamphiphiles in a 1:1 manner, but is more predisposedto form 2:1 and 2:2 capsular complexes (see below). Asthese resorcinarene hosts possess bowl-like cavities, theyare by and large more hydrophobic than the cavity of

similarly sized β-cyclodextrin. In addition, the bowl-likecavity induces specific guest orientations more readily thanthat seen in cyclodextrins. For example, amphiphiles bindto both 5 and 6 with their polar head group at the portaland exposed to the aqueous phase, while the hydrocarbonportion of the guest resides deep within the binding site.

OO OO OO O

H H HH

O

N NN N N N N N

NN NN NN NN

CO2Na CO2Na CO2Na CO2Na

HH HH

(5)

OO O OO O

H H HH

O OH HH

O OO

O

O O

H

O O

R RR R

R1 R1R1 R1

(R = CO2H)

(6)

The third important class of water-soluble hosts isthe cucurbiturils (7) [112–114]. Originally possessing arather narrow range of structures, the range of cucurbiturilshas been greatly expanded by the efficient isolation ofcucurbiturils comprising between 5 and 10 glycoluril units(n = 5–10), partially inverted cucurbiturils [115], andcucurbituril analogs [116]. In particular, the replacement ofthe bridgehead methine hydrogen atoms with groups thatmodify the solubility properties of these hosts has been animportant development [112].

N N

N N

O

On

HH

(7)


The binding properties of these torus-shaped hostsare wide and varied [112]. The smaller cucurbiturilsstrongly bind cationic species at the two portals. Allof the cucurbiturils also bind hydrophobic guests withintheir pocket. The larger hosts are particularly interestingin this regard because they are capable of not justforming 1:1 complexes but ternary complexes also. Thus,cucurbit[8]uril (7, n = 8) is capable of binding paraquatand 2,6-dihydroxynaphthalene as a stable hetero-guest pair.When guests can take advantage of the hydrophobic pocketand the ring of carbonyls at each portal, associationscan be exceedingly strong (up to 2 × 1012 M−1) [112,117]. Cucurbiturils are currently being explored as ionchannels, and their properties in vesicles, on surfaces, andin polymers, investigated [112].

A fourth important type of water-soluble hosts is thecyclophanes [118, 119]. The structurally most straight-forward examples of cyclophanes are essentially belts ofaromatic rings, for example, 8 or 9, from the Diederich[120] and Dougherty laboratories [121]. Possessing aslot-like binding pocket, 8 is capable of binding aromaticguests and has revealed much concerning how π−πstacking can contribute to the hydrophobic effect. Theslightly more capacious 9 binds aromatic guests also, butadditionally more rotund saturated guests. Recently, moreelaborate and capacious examples, such as cyclophane10, have accomplished the difficult task of sequestering

O

N

O

N

O O

R

R

R

R

R R

R RClCl

n

n

R = OMe

(8)

CO2−

−O2C

−O2C

−O2C

−O2C

OO

CO2−

OO

CO2−

CO2−

(9)

N HO

NO

H

O

NH

RN H

O

NO

H

O

N RHNH

O

NO

H

O

NRH

NHO

N

O

H

O

NRH

NHO

NO

H

O

NRH

R =O

OO

CO2H

CO2H

CO2H

(10)

disaccharides from water [122]. In particular, sugars withall equatorial arrays of polar substituents and hence apolarfaces of axially directed CH groups show good affinity, afact that has been utilized in a slightly smaller host thatbinds O-linked β-N -acetylglucosamine (O–GlcNAc) andrelated derivatives [123].

Although the degree to which the hydrophobic effectis important in the desolvation of these types of guests isdebatable, the larger hydrophobic inner pocket of the hostis likely dewetted to a considerable degree. Additionally,the binding of polar guests such as sugars also raises theissue of the as-yet-unknown extent and manner by whichpolar groups modulate the solvation of nearby hydrophobicsurfaces.

Another approach to hosts utilizes self-assembly inthe final step of synthesis, a useful approach via whichvery large hosts can be synthesized from structurallysimpler subunits. In essence, the final—and usuallyquantitative—step in the synthesis is designed into earliersteps. In water, two self-assembly strategies have provenuseful. By far the most common approach is to use metalion coordination to direct assembly. This approach, pio-neered by the Fujita [33, 124, 125], Raymond [77, 126,127], and Stang groups [45, 128, 129] has led to an enor-mous range of both hosts and applications. Representativeexamples of hosts shown to bind guests are 11 and 12 (onlyone of six ligands shown).

CONCLUSIONS AND OUTLOOK 15

Pd

H2N

NH2

N

NN

N

N

NN

NN

N

N

N

N

N

N

N

N N

N

N

N

N

N N

12 NO3−

=

2+

(11)

Guest

O

O

O NH

NH

O

O

O

12_

(12)

= Al3+, Ga3+, In3+, Fe3+

Very briefly, capsule 11 is one of many examples fromthe Fujita laboratory [33] that have been shown to engendera nanoscale fluorous phase [36], bring about catalysis [35],and control self-assembly [130]. The Raymond tetrahedralcapsule 12 can be constructed from a variety of (apical)metal ions, and has be shown to be a consummate controllerof size- and shape-selective catalysis [37, 40, 127, 131,132]. It has also been shown to possess unusual memoryproperties [133] and can stabilize reactive intermediates[134].

These self-assembling systems are excellent examplesof what are usually thought of as highly directionalsupramolecular motifs. The metal coordination is, inenthalpic terms, highly directional. However, as discussedabove, Nature frequently brings about assembly by relyingon supramolecular motifs that are not, at least from anenthalpic viewpoint, very directional. Can “nondirectional”motifs be used to drive the assembly of synthetic systems?To the author’s knowledge, the only example to date thatsuggests the answer to this question is “yes” is host 7[135, 136]. The four “upper” aromatic rings in these types

of host constitute a wide hydrophobic rim around thebinding pocket. This, in combination with the hydrophobicpocket itself, predisposes the host to dimerize around guestsin aqueous solution. Rather than enthalpically powerfulmetal coordination, only weak π−π interactions can formbetween subunits, and it is the desolvation of the pocket[52], and rim of the host, and the desolvation of the guest orguests that are the important driving forces in the formationof the nanocapsule complexes. The importance of the guest(or guests) cannot be overstated. As discussed above, ifthe guest is amphiphilic, then the polar head group islocated at the entrance of the cavity in the 1:1 complexand the overall hydrophobicity of the dimerization interfaceis low. However, as the polarity of this head group islowered, dimerization becomes more thermodynamicallyfavored. Held together by the hydrophobic effect, thesecomplexes possess considerable thermodynamic and kineticstability [137], allowing the capsule to act as a yoctoliter reaction flask [138–141], affect the separation ofhydrocarbon gases [142], and modulate the conformational[143] and electrochemical [144] properties of entrappedguests. Although the subtleties of how the structureof either host or guest influences assembly are as yetundetermined, it is apparent that careful design with“nondirectional” supramolecular motifs can lead to well-defined supramolecular structure.

We have briefly reviewed the different classes andfamilies of large water-soluble hosts. As the many reviewscited highlight, there are countless other examples ofsmaller hosts that are also capable of binding moleculesof significance. In short, the choices of host “core” forsensor development are extensive. With many systems, adetailed understanding of the structure/binding relationship,and how this relates to the hydrophobic effect, has been wellestablished. However, there is still much to learn to fullyunderstand how the hydrophobic effect can be harnessed tomaximize binding. Thus, it is relatively straightforward tosuggest that a group strongly bind a dicarboxylic acid guestin chloroform; but using the hydrophobic effect, what is thebest motif for binding a pyridine ring?

1.4 CONCLUSIONS AND OUTLOOK

There is a wide range of host molecules that can either beused for inspiration or literally used to construct sensorsfor aqueous solution. Hosts that have been studied to dateprovide valuable information about the subtleties of vdWforces and the hydrophobic effect, and how these may beharnessed to control binding. However, much is still to beaccomplished if supramolecular chemistry is to “conquer”sensor design in water. That said, “much to learn” is also“much to discover,” and the application of intellect andperseverance can only lead to many exciting discoveries.


Acknowledgment

The author gratefully acknowledges the financial supportfrom the National Science Foundation (CHE-0718461) inpreparing this review.

REFERENCES

1. Anslyn EV. J Org Chem 2007; 72: 687–699.

2. Alon U. An introduction to systems biology: designprinciples of biological circuits. Boca Raton (FL): Chapman& Hall/CRC; 2007.

3. Palsson BØ. Systems biology: properties of reconstructednetworks. New York: Cambridge University Press; 2006.

4. Jeffrey GA. An introduction to hydrogen bonds. New York:Oxford University Press; 1997.

5. Desiraju GR, Steiner T. The weak hydrogen bond. Oxford:Oxford University Press; 1999.

6. Israelachvili J. Intermolecular & surface forces. San Diego(CA): Academic Press; 1991.

7. Anslyn EV, Dougherty DA. Modern physical organicchemistry. Sausalito (CA): University Science Books;2006.

8. Atkins P, de Paula J. Physical chemistry. New York: W. H.Freeman and Company; 2006.

9. Chandler D. Nature 2002; 417:491.

10. Chandler D. Nature 2005; 437: 640–647.

11. Lum K, Chandler D, Weeks JD. J Phys Chem B 1999; 103:4570–4577.

12. Dill KA. Biochemistry 1990; 29: 7133–7155.

13. Southall NT, Dill KA, Haymet ADJ. J Phys Chem 2002;106: 521–533.

14. Tanford C. The hydrophobic effect: formation of micelles& biological membranes. 2nd ed. New York: John Wiley &Sons, Inc.; 1980.

15. Reichardt C. Solvents and solvent effects in organicchemistry. 2nd ed. New York: VCH; 1990.

16. Ma JC, Dougherty DA. Chem Rev 1997; 97: 1303–1324.

17. Dougherty DA. Science 1996; 271: 163–168.

18. Caldwell JW, Kollman PA. J Am Chem Soc 1995; 117:4177–4178.

19. Diederich F, Smithrud DB, Sanford EM, Wyman TB,Ferguson SB, Carcanague DR, Chao I, Houk KN. ActaChem Scand 1992; 46: 205–215.

20. Leach AR. Molecular modelling: principles and applications.2nd ed. Harlow: Pearson Education.

21. Hunter CA, Sanders EB. J Am Chem Soc 1990; 112:5525–5534.

22. McLachlan AD. Discuss Faraday Soc 1965; 40: 239–245.

23. Gelbart WM., Ben-Shaul A., Roux D. Micelles, Membranes,Microemulsions, and Monolayers. New York: Springer-Verlag; 1994.

24. Dwars T, Paetzold E, Oehme G. Angew Chem Int Ed 2005;44: 7174–7199.

25. Petsko GA, Ringe D. Protein structure and function.Sunderland (MA): New Science Press; 2004.

26. Rebek J Jr. Angew Chem Int Ed 2005; 44: 2068–2078.

27. Dalgarno SJ, Tucker SA, Bassil DB, Atwood JL. Science2005; 309: 2037–2039.

28. MacGillivray LR, Atwood JL. Nature 1997; 389: 469–472.

29. Sijbesma RP, Meijer EJ. Chem Commun 2003: 5–16.

30. Sessler JL, Jayawickramarajah J. Chem Commun 2005:1939–1949.

31. Davis JT, Spada GP. Chem Soc Rev 2007; 36: 296–313.

32. Rebek J Jr. Acc Chem Res 1999; 32: 278–286.

33. Fujita M, Tominaga M, Hori A, Therrien B. Acc Chem Res2005; 38: 369–378.

34. Fujita M, Fujit

Date post:	28-Aug-2020
Category:	Documents
Upload:	others
View:	7 times
Download:	1 times

Principles, Strategies, and Applications · CHEMOSENSORS Principles, Strategies, and Applications...

Documents