Recoverable and Recyclable Catalysts

Recoverable and Recyclable Catalysts

Recoverable and Recyclable Catalysts Edited by Maurizio Benaglia © 2009 John Wiley & Sons, Ltd. ISBN: 978-0-470-68195-4

Recoverable and Recyclable Catalysts Edited by Maurizio Benaglia

Department of Organic and Industrial Chemistry, University of Milan, Italy

This edition first published 2009 2009 John Wiley & Sons, Ltd Registered office John Wiley & Sons Ltd, The Atrium, Southern Gate, Chichester, West Sussex, PO19 8SQ, United Kingdom For details of our global editorial offices, for customer services and for information about how to apply for permission to reuse the copyright material in this book please see our website at www.wiley.com. The right of the author to be identified as the author of this work has been asserted in accordance with the Copyright, Designs and Patents Act 1988. All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, except as permitted by the UK Copyright, Designs and Patents Act 1988, without the prior permission of the publisher. Wiley also publishes its books in a variety of electronic formats. Some content that appears in print may not be available in electronic books. Designations used by companies to distinguish their products are often claimed as trademarks. All brand names and product names used in this book are trade names, service marks, trademarks or registered trademarks of their respective owners. The publisher is not associated with any product or vendor mentioned in this book. This publication is designed to provide accurate and authoritative information in regard to the subject matter covered. It is sold on the understanding that the publisher is not engaged in rendering professional services. If professional advice or other expert assistance is required, the services of a competent professional should be sought. The publisher and the author make no representations or warranties with respect to the accuracy or completeness of the contents of this work and specifically disclaim all warranties, including without limitation any implied warranties of fitness for a particular purpose. This work is sold with the understanding that the publisher is not engaged in rendering professional services. The advice and strategies contained herein may not be suitable for every situation. In view of ongoing research, equipment modifications, changes in governmental regulations, and the constant flow of information relating to the use of experimental reagents, equipment, and devices, the reader is urged to review and evaluate the information provided in the package insert or instructions for each chemical, piece of equipment, reagent, or device for, among other things, any changes in the instructions or indication of usage and for added warnings and precautions. The fact that an organization or Website is referred to in this work as a citation and/or a potential source of further information does not mean that the author or the publisher endorses the information the organization or Website may provide or recommendations it may make. Further, readers should be aware that Internet Websites listed in this work may have changed or disappeared between when this work was written and when it is read. No warranty may be created or extended by any promotional statements for this work. Neither the publisher nor the author shall be liable for any damages arising herefrom. Library of Congress Cataloging-in-Publication Data Recoverable and recyclable catalysts / edited by Maurizio Benaglia. p. cm. Includes bibliographical references and index. ISBN 978-0-4706-8195-4 (cloth : alk. paper) 1. Catalysts–Recycling. 2. Catalysis. I. Benaglia, Maurizio. QD505.R436 2009 541’.395–dc22 2009014028 A catalogue record for this book is available from the British Library. ISBN 978-0470-681954 (H/B) Typeset in 10/12pt Times by Thomson Digital, Noida, India. Printed and bound in Great Britain by CPI Antony Rowe, Chippenham, Wiltshire.

Contents Preface

xiii

Acknowledgements Contributors 1

xvii

The Experimental Assay of Catalyst Recovery: General Concepts John A. Gladysz

1

1.1 1.2 1.3 1.4

1 2 3 5 5 7 8

Introduction Catalyst Precursor vs Catalyst Catalyst vs Catalyst Resting State Catalyst Inventory: Loss Mechanisms 1.4.1 Catalyst Decomposition 1.4.2 Catalyst Leaching 1.5 Evaluation of Catalyst Recovery 1.5.1 Product Yield, Conversion, or TON as a Function of Cycle: Poor and Potentially Deceptive Criteria 1.5.2 Reaction Rate or TOF as a Function of Cycle 1.5.3 Gravimetric and Other Assays of Recovered Catalyst 1.5.4 Special Caveats when ‘Residues’ are Recycled 1.6 Prospective References 2

xv

Surface-functionalized Nanoporous Catalysts for Renewable Chemistry Brian G. Trewyn, Hung-Ting Chen and Victor S.-Y. Lin 2.1

2.2

Introduction 2.1.1 Homogeneous Catalysis vs Heterogeneous Catalysis 2.1.2 Multi-Site vs Single-Site Heterogeneous Catalysis Immobilization Strategies of Heterogeneous Catalysts 2.2.1 Supported Materials 2.2.2 Conventional Methods to Functionalize Silica Surfaces 2.2.3 Alternative Synthesis of Immobilized Complex Catalysts on a Solid Support 2.2.4 Techniques for Characterization of Heterogeneous Catalysts

8 9 12 13 13 14

15 15 16 16 17 17 18 25 26

vi

Contents

2.3

Efficient Heterogeneous Catalysts with Enhanced Reactivity and Selectivity with Functionality 2.3.1 Surface Interaction of Silica and Immobilized Homogeneous Catalysts 2.3.2 Introduction of Functionalities and Control of Silica Support Morphology 2.3.3 Selective Surface Functionalization of Solid Support for Utilization of Nanospace Inside the Porous Structure 2.3.4 Cooperative Catalysis by Multifunctionalized Heterogeneous Catalyst Systems 2.3.5 Mesoporous Mixed Metal Oxides for Heterogeneous Catalysts 2.4 Other Heterogeneous Catalyst Systems on Nonsilica Supports 2.5 Conclusion References 3

26 26 29 31 35 43 44 45 45

Insoluble Resin-supported Catalysts Gang Zhao and Zhuo Chai

49

3.1 3.2

49 50 50

Introduction Transition Metal Catalyzed CC Bond Formation Reactions 3.2.1 Pd-catalyzed Reactions 3.2.2 Asymmetric Additions of Organozinc Reagents to Aldehydes 3.2.3 Rh-catalyzed Asymmetric Intermolecular CH Activation 3.2.4 Cu-catalyzed Asymmetric Cyclopropanation 3.3 Oxidation 3.3.1 Oxidation of Sulfides to Sulfoxide 3.3.2 Oxidation of Alkanes, Alkenes and Alcohols 3.3.3 Epoxidation of Alkenes 3.3.4 Asymmetric Hydroformylation of Olefins 3.3.5 Asymmetric Dihydroxylation of Alkenes 3.4 Reduction 3.4.1 Asymmetric Reduction of Ketones 3.4.2 Reduction of Carboxamides to Amines 3.5 Organocatalyzed Reactions 3.5.1 Asymmetric Aldol Reaction and Aminoxylation 3.5.2 Asymmetric Tandem Reaction 3.5.3 Allylation of Aldehydes 3.5.4 Nucleophilic Substitution Reactions 3.6 Annulation Reactions 3.6.1 Cycloaddition 3.6.2 Intramolecular Hydroamination 3.7 Miscellaneous 3.8 Conclusion References

53 54 55 56 56 57 58 59 60 61 61 62 62 63 64 65 66 66 66 68 70 72 72

Contents

4

Catalysts Bound to Soluble Polymers Tamilselvi Chinnusamy, Petra Hilgers and Oliver Reiser

77

4.1 4.2 4.3

77 78 79 79 81 81

Introduction Soluble Supports – General Considerations Recent Developments of Soluble Polymer-supported Catalysts 4.3.1 Attachment of Catalysts to Polymer Supports 4.3.2 Polymer-bound Metal Catalysts – General Considerations 4.3.3 Polymer-bound Organocatalysts – General Considerations 4.4 Recent Examples for Reactions Promoted by Catalysts Bound to Soluble Polymers 4.4.1 Achiral Catalysts 4.4.2 Chiral Catalysts 4.5 Conclusion List of Abbreviations References 5

6

vii

81 81 88 98 98 98

Polymeric, Recoverable Catalytic Systems Qiao-Sheng Hu

101

5.1 5.2

Introduction Polymeric Catalyst Systems 5.2.1 1,10 -Bi-2-naphthol (BINOL)-based Polymeric Catalytic Systems 5.2.2 Bisphosphine-containing Polymeric Catalyst Systems 5.2.3 Salen-containing Polymeric Catalytic Systems 5.2.4 BINOL–BINAP-based Bifunctional Polymeric Catalytic Systems 5.2.5 Dendrimer Catalyst Systems 5.2.6 Dendronized Polymeric Catalytic Systems 5.3 Summary Acknowledgements References

101 102

Thermomorphic Catalysts David E. Bergbreiter

117

6.1 6.2 6.3 6.4 6.5 6.6

117 118 122 126 129 130

6.7

Introduction Thermomorphic Catalyst Separation Strategies Hydrogenation Reactions Under Thermomorphic Conditions Hydroformylation Reactions Under Thermomorphic Conditions Hydroaminations Under Thermomorphic Conditions Pd-catalyzed Reactions Under Thermomorphic Conditions 6.6.1 Pd-catalyzed Allylic Substitution Under Thermomorphic Conditions 6.6.2 Pd-catalyzed Cross-coupling Reactions Under Thermomorphic Conditions Polymerization Reactions Under Thermomorphic Conditions

102 103 108 108 110 111 114 115 115

130 131 138

viii

Contents

6.8 6.9

Organocatalysis Under Thermomorphic Conditions Cu(I)-catalyzed 1,3-Dipolar Cycloadditions Under Thermomorphic Conditions 6.10 Thermomorphic Hydrosilylation Catalysts 6.11 Thermomorphic Catalytic Oxidations 6.12 Conclusions References 7

9

144 144 145 147 147

Self-supported Asymmetric Catalysts Wenbin Lin and David J. Mihalcik

155

7.1 7.2

155

Introduction Self-supported Asymmetric Catalysts Formed by Linking Catalytically Active Subunits via Metal–Ligand Coordination 7.3 Self-supported Asymmetric Catalysts Formed by Post-synthetic Modifications of Coordination Polymers 7.4 Self-supported Asymmetric Catalysts Formed by Linking Multitopic Chiral Ligands with Catalytic Metal Centers 7.5 Conclusions and Outlook Acknowledgments References 8

142

156 163 168 172 174 174

Fluorous Chiral Catalyst Immobilization Tibor Soos

179

8.1 8.2 8.3

Introduction Fluorous Chemistry and its Basic Recovery Concepts Application of Fluorous Chiral Catalysts 8.3.1 Fluorous Nitrogen Ligands 8.3.2 Fluorous Oxygen Ligands 8.3.3 Phosphorous Ligands 8.4 Summary References

179 180 181 182 192 194 196 197

Biphasic Catalysis: Catalysis in Supercritical CO2 and in Water Simon L. Desset and David J. Cole-Hamilton

199

9.1 9.2 9.3

199 200 202 202 203 203 212 214 220 227 228

Introduction Biphasic Catalysis Aqueous Biphasic Catalysis 9.3.1 Introduction 9.3.2 Aqueous Biphasic Catalysis: Beyond Mass Transfer 9.3.3 Additives 9.3.4 Surface-active Ligands 9.3.5 Homogeneous Reaction with Biphasic Separation 9.3.6 Supported Aqueous Phase Catalysis (SAPC) 9.3.7 New Reactor Design 9.3.8 Conclusion

Contents

10

11

12

ix

9.4

Supercritical Carbon Dioxide 9.4.1 Introduction 9.4.2 Supercritical Carbon Dioxide for Catalyst Recycling 9.5 Conclusion References

229 229 230 246 247

Asymmetric Catalysis in Ionic Liquids Lijin Xu and Jianliang Xiao

259

10.1 Introduction 10.2 Metal-catalyzed Asymmetric Reactions in ILs 10.2.1 Asymmetric Hydrogenation 10.2.2 Asymmetric Transfer Hydrogenation 10.2.3 Asymmetric Oxidation 10.2.4 Asymmetric CC Bond Formation 10.2.5 Miscellaneous Reactions 10.3 Asymmetric Organocatalytic Reactions in ILs 10.3.1 Asymmetric Aldol Reactions 10.3.2 Asymmetric Michael Addition 10.3.3 Asymmetric Diels–Alder Reaction 10.3.4 Asymmetric Mannich Reaction 10.3.5 Asymmetric Baylis–Hillman Reaction 10.4 Concluding Remarks References

259 261 261 270 271 275 283 287 287 290 292 292 293 294 295

Recoverable Organic Catalysts Maurizio Benaglia

301

11.1 Introduction 11.2 Achiral Organic Catalysts 11.2.1 Oxidation Catalysts 11.2.2 Phase Transfer Catalysts 11.2.3 Miscellaneous Catalysts 11.3 Chiral Organic Catalysts 11.3.1 Phase Transfer Catalysts 11.3.2 Lewis Base Catalysts 11.3.3 Miscellaneous Catalysts 11.4 Catalysts Derived from Amino Acids 11.4.1 Proline Derivatives 11.4.2 Amino Acid-derived Imidazolinones 11.4.3 Other Amino Acids 11.5 General Considerations on Recyclable Organocatalysts 11.6 Outlook and Perspectives References

301 304 304 307 309 311 311 313 319 319 320 328 331 334 336 337

Organic Polymer-microencapsulated Metal Catalysts Jun Ou and Patrick H. Toy

341

12.1 Introduction

341

x

13

14

15

Contents

12.2 Non-cross-linked Polymer-microencapsulated Catalysts 12.2.1 Non-cross-linked Polystyrene 12.2.2 Non-cross-linked Polystyrene Derivatives 12.2.3 Polysulfone 12.2.4 Poly(xylylviologen dibromide) 12.3 Cross-linked Polymer-microencapsulated Catalysts 12.3.1 Divinyl Benzene Cross-linked Polystyrene 12.3.2 Oligo(ethylene glycol) Cross-linked Polystyrene 12.3.3 Urea Group Cross-linked Polyphenylene 12.4 Summary Table 12.5 Conclusions References

342 342 350 353 354 355 355 357 367 374 375 375

Organic Synthesis with Mini Flow Reactors Using Immobilised Catalysts Sascha Ceylan and Andreas Kirschning

379

13.1 Introduction 13.1.1 General Remarks 13.1.2 Batch versus Flow Processes 13.1.3 Micro versus Mini Flow Reactors 13.2 Catalysis in Mini Flow Reactors with Immobilised Catalysts 13.2.1 Solid Supports Based on Silica 13.2.2 Solid Supports Based on Polymers 13.2.3 Monolithic Supports 13.2.4 Immobilisation on Membranes 13.3 Miscellaneous Enabling Techniques for Mini Flow Systems 13.3.1 Ionic Liquids as Media for Immobilisation 13.3.2 Inductive Heating – a New Technique for Mini Flow Processes 13.4 Perspectives and Outlook References

379 379 380 381 382 382 387 392 401 404 404

Homogeneous Catalysis Using Microreactor Technology Johan C. Brandt and Thomas Wirth

411

14.1 Introduction 14.2 Acid-catalysed Reactions 14.3 Liquid–liquid Biphasic Systems 14.4 Photocatalysis 14.5 Asymmetric Catalytic Reactions 14.6 Unusual Reaction Conditions References

411 411 413 418 421 421 423

Catalyst Immobilization Strategy: Some General Considerations and a Comparison of the Main Features of Different Supports Franco Cozzi

427

15.1 Introduction

427

404 406 407

Contents

15.2 General Considerations on Catalyst Immobilization 15.2.1 Prerequisite Conditions for Immobilization 15.2.2 Reasons Justifying Immobilization 15.2.3 A General Discussion on the Practical Aspects of Immobilization 15.3 Comparison of Different Supports Employed for the Immobilization of Proline 15.3.1 Organic Supports 15.3.2 Inorganic Supports 15.4 Comparison of Different Supports Employed for the Immobilization of Bis(oxazolines) 15.4.1 Noncovalent Immobilization 15.4.2 Covalent Immobilization 15.5 Conclusions References Index

xi

428 428 433 437 442 442 450 452 452 453 458 458 463

Preface Why anyone should read a book about recoverable and recyclable catalysts? The reason is already contained in the three words of the title. Catalysis represents a frontier field of research, the concept of catalysis being one of the key features of modern chemistry where a substoiochiometric amount of compounds is used to produce large amounts of other compounds. Catalysis is one of the twelve principles of Green Chemistry: a process based on a catalytic methodology is already “green” by definition, since it is clearly stated that “catalysts are preferable to stoichiometric reagents.” Recoverable is also a fundamental word for the chemistry of the future, where terms like sustainability or low environmental impact will become more and more important. So a recoverable catalyst will generate less waste and will be a key step toward the development of more efficient processes. Recyclable clearly refers to a fundamental topic for the industry, the economy and the efficiency of the process, and it represents the link between the research in academy and industry. What is better for a company than the possibility to reuse again and again the same catalyst? The development of recyclable catalysts represents a fascinating challenge that may be tackled from several different points of view; it is indeed a real interdisciplinary field, where pure chemistry is deeply bound and connected to material science, or engineering and where even business and economy-related issues play an important role in determining the planning, the design and the realization of a project in the area. It is a field where many technologies and opportunities are offered to successfully realize an easy recoverable and, what is more important, reusable catalytic system. In this book the first chapter introduces the reader into the world of recyclable catalysts; written by a world wide recognized authority in the field it presents general principles which should be considered when a study in this area will be conducted and it offers a key to critically approach the reading of the other chapters where different methodologies used to realize a modern, efficient process involving recyclable systems will be discussed. After this introductory chapter in the first three chapters the somehow “traditionally” exploited techniques to realize heterogenized catalysts are presented, where different supports, inorganic and organic materials, soluble or insoluble polymers are discussed. In the following six chapters other methodologies, alternative to the traditional strategies, are illustrated, from the use of ionic liquids or supercritical CO2 until the perfluorous systems. Thermomorphic catalysts as well as polymeric, or self supported catalytic systems are all relatively new areas of investigation of potential enormous growth. In the last chapters some “new” topics in the field of recoverable catalysts are presented, such as the development of reusable organic catalysts, and the use of micro encapsulated catalysts or the employment of

xiv

Preface

new devices and “enabling technologies” like flow and membrane reactors, mini and micro reactors that represent the future of the synthetic chemist, which will have to be applied to catalysis as well in the next years. The final chapter represents an attempt to make a few general considerations on the immobilization process, and by using two case studies, one organometallic species and one organocatalyst, to try to compare the behaviour of the same catalytic species anchored to all different supports. I am really happy with the list of authors which have agreed to contribute to the volume and I am honoured to act as editor of a book whose chapters were delivered all by worldwide recognized experts in the area, which are universally considered in many cases the real pioneers in their own field. This gives a strong historical background of great authority to the chapters, which are anyhow principally devoted to discuss the recent, more important achievements in the field and in this way significantly leaning to the future perspectives and challenges in the area. For the neophyte in the field the book wants to be an easy to consult guide, that introduces the reader in the very complex topic of the recoverable and recyclable catalysts, giving an idea of how many possibilities and how many challenges are still out there to be tackled with new ideas and new approaches. On the other hand, for the experts in the field, the book wants to be the occasion to establish some general, important milestones, as starting points from which the future investigations will move towards new objectives but within a rigorous, well established scientific methodology, necessary for the design of new catalytic systems but also for the correct evaluation of the obtained results. In any case the real goal of the book is to promote new achievements in the field, to stimulate the interest of more scientists in entering and working in a very multidisciplinary area of research; the book wants to be a “catalyst” for the development of new recyclable catalysts. My personal wish is that the reader will enjoy the consultation of this book as much as I personally enjoyed in reading all chapters and editing a volume where a really multifaceted and stimulating chemistry is presented, where creativity, fantasy and courage emerge as qualities required to the modern chemist which wishes to successfully work in this field. Maurizio Benaglia

Acknowledgements I wish to acknowledge all the authors contributing to the book for their effort and their participation to this editorial adventure. A special thank goes to prof. Mauro Cinquini for the help and all the precious advice he offered to me in the editing work of the volume. Finally I wish to thank my wife, Michela, for the support that she always gives me and that gave me also in this project.

Contributors Maurizio Benaglia David E. Bergbreiter Johan C. Brandt Sascha Ceylan Zhuo Chai Hung-Ting Chen Tamilselvi Chinnusamy David J. Cole-Hamilton Franco Cozzi Simon L. Desset John A. Gladysz Petra Hilgers Qiao-Sheng Hu

Andreas Kirshning Victor S. Y. Lin Wenbin Lin David J. Mihalcik

Department of Organic and Industrial Chemistry, University of Milan, Milan, Italy Department of Chemistry, Texas A&M University, College Station, Texas, USA School of Chemistry, Cardiff University, Cardiff, UK Institut fur Organische Chemie, Leibniz Universitat Hannover, Hannover, Germany Shanghai Institute of Organic Chemistry, Shanghai, China Department of Chemistry, Iowa State University, Ames, Iowa, USA Institut f€ ur Organische Chemie, Universit€at Regensburg, Regensburg, Germany EaStCHEM, School of Chemistry, University of St. Andrews, St. Andrews, Fife, Scotland Department of Organic and Industrial Chemistry, University of Milan, Milan, Italy EaStCHEM, School of Chemistry, University of St. Andrews, St. Andrews, Fife, Scotland Department of Chemistry, Texas A&M University, College Station, Texas, USA Institut f€ ur Organische Chemie, Universit€at Regensburg, Regensburg, Germany Department of Chemistry, College of Staten Island and the Graduate Center of the City University of New York, Staten Island, New York, USA Institut fur Organische Chemie, Leibniz Universitat Hannover, Hannover, Germany Department of Chemistry, Iowa State University, Ames, Iowa, USA Department of Chemistry, University of North Carolina, Chapel Hill, North Carolina, USA Department of Chemistry, University of North Carolina, Chapel Hill, North Carolina, USA

xviii

Contributors

Jun Ou Oliver Reiser Tibor Soo´s

Patrick H. Toy Brian G. Trewyn Thomas Wirth Jianliang Xiao Lijin Xu Gang Zhao

Department of Chemistry, The University of Hong Kong, Hong Kong Institut f€ ur Organische Chemie, Universit€at Regensburg, Regensburg, Germany Institute of Biomolecular Chemistry, Chemical Research Center of the Hungarian Academy of Science, Budapest, Hungary Department of Chemistry, The University of Hong Kong, Hong Kong Department of Chemistry, Iowa State University, Ames, Iowa, USA School of Chemistry, Cardiff University, Cardiff, UK Department of Chemistry, Liverpool Centre for Materials and Catalysis, University of Liverpool, Liverpool, UK Department of Chemistry, Renmin University of China, Beijing, China Laboratory of Modern Synthetic Organic Chemistry, Shanghai Institute of Organic Chemistry, Chinese Academy of Sciences, Shanghai, China

Figure 7.2 (a) Schematic representation of the 3D framework of 2 showing the zigzag chains of [Cd(m-Cl)2]n along a axis. (b) space-filling model of 2 as viewed down the a axis, showing large 1D chiral channel (1.6 1.8 nm). (c) Schematic representation of the active (BINOLate)Ti (OiPr)2 catalytic sites in the open channels of 2. Reproduced with permission from C.-D. Wu, A. Hu, L. Zhang, W. Lin, A Homochiral Porous Metal–Organic Framework for Highly Enantioselective Heterogeneous Asymmetric Catalysis, J. Am. Chem. Soc., 2005, 127, 8940–8941. Copyright 2005 American Chemical Society


Figure 7.3 (a) The 2D square grid in 3. (b) Schematic representation of the 3D framework of 3. (c) Space-filling model of 3 as viewed down the c axis, the 2-fold interpenetrating networks are shown with blue and violet colors. (d) Schematic representation of the interpenetration of mutually perpendicular 2D grids in 4. (e) Space-filling model of 4 as viewed down the c axis. (f) Schematic representation of steric congestion around the chiral dihydroxyl group of L8 (orange sphere) arising from the interpenetration of mutually perpendicular 2D grids in 4. C.-D. Wu & W. Lin, Angew. Chem. Int. Ed., 2007, 46, 1075–1078. Copyright Wiley-VCH Verlag GmbH & Co. KGaA. Reproduced with permission

1 The Experimental Assay of Catalyst Recovery: General Concepts John A. Gladysz Department of Chemistry, Texas A&M University, College Station, Texas, USA

1.1 Introduction Congratulations. You have a well-placed interest in catalysis, and specifically in recoverable catalysts. Over the next several hundred pages of this book, you will encounter many types of recovery strategies and cleverly designed catalysts. You could have begun with these chapters, which many would say are more exciting and contain the ‘good stuff’. But instead, you have started here, where some thinly veiled sermonizing awaits. Please persevere, and ‘be converted’ to the ranks that strive for the highest quality in research. It is not difficult to carry out rigorous quantitative studies of recoverable catalysts. But many researchers fall far short. Serious deficiencies can even be found in some references cited in the individual chapters. A catalyst accelerates the rate of a chemical reaction, and is left unchanged by the reaction.1 Given this definition, the most obvious way to measure the efficacy of catalyst recovery is by the rate of a subsequent reaction cycle. If the catalyst is not completely recovered, the rate must be slower. The yields of products isolated from preparative reactions are most commonly determined gravimetrically, and catalyst recoveries can in principle be assayed analogously. However, this is not as general, as analyzed further below. For example, the very small amounts of catalysts used in many reactions virtually preclude


2


Figure 1.1 Yield of carbocycle 3 as a function of cycle, with catalyst recovery via precipitation

accurate determinations of mass. And we will see later that the yield of a reaction product as a function of cycle is nearly always an inappropriate measure. What is so difficult about putting a rate assay into practice? For one, there is a hard-core group of preparative chemists who will do everything possible to avoid anything to do with kinetics. However, we are not talking about a rate constant or law, but rather a less demanding series of ‘rate profiles’ or ‘reaction profiles’ or ‘conversion profiles’ for cycles conducted under identical conditions. These might involve spectroscopic or chromatographic monitoring. An example involving a carbon–carbon bond forming reaction catalyzed by the fluorous phosphine P((CH2)3(CF2)7CF3)3 (1) is illustrated in Figure 1.1.2 The highly temperature-dependent solubility of 1 allows efficient recovery by a simple solid/liquid phase separation (filtration or decantation) through five cycles, as shown by the superimposed traces. In any case, catalysis is an intricate science that features many complexities, and it is easy for beginning researchers to confuse various points. So let’s first take a detour through some of the basic background concepts that are important to all of the chapters in this book.

1.2 Catalyst Precursor vs Catalyst It is easy for any chemist to say ‘catalyst’ when he/she really means ‘catalyst precursor’. For example, Wilkinson’s catalyst (ClRh(PPh3)3) is familiar to everyone and has been studied in detail.3 Importantly, this species does not appear in the kinetically dominant cycle for the catalytic hydrogenation of alkenes. Rather, a PPh3 ligand must first dissociate. Thus, it is

The Experimental Assay of Catalyst Recovery: General Concepts

3

best regarded as a catalyst precursor. Realistically, it will never be practical to change the widely accepted name to ‘Wilkinson’s catalyst precursor’. Hence, the scientific community is stuck with imprecise usage. Particularly for homogeneous transition metal catalysts, additional steps are often needed to enter the catalytic cycle. Although this can be a hypothetically reversible ligand dissociation as with Wilkinson’s catalyst, fundamental and irreversible changes can also be involved. For example, the active catalyst might be generated via oxidizing or reducing a small amount of the catalyst precursor. This will often be evidenced by an induction period. If data are acquired rapidly enough, any induction period will be obvious in plots of the type in Figure 1.1. Some are illustrated below.

1.3 Catalyst vs Catalyst Resting State A catalyst can take many forms as it proceeds through the catalytic cycle, for which one step or energy maximum will be rate determining. The minimum with the lowest free energy will correspond to the catalyst resting (rest) state. In special cases, two minima in the cycle may be energetically similar, such that the catalyst rests as an equilibrium mixture of two states. Regardless, what is always being recycled is the catalyst resting state. Thus, the property being exploited for recovery (e.g., an affinity for a phase as conferred by a phase tag) must be associated with the resting state. Hence, the rational design of recoverable catalysts requires some consideration of mechanism. The catalyst resting state can also significantly depend upon reaction conditions. Perhaps one minimum consists of a metal species and an uncoordinated alkene (or other ligand), and another, close in energy, is a metal/alkene (ligand) complex. Given the entropic difference, the resting state could be a strong function of temperature. Furthermore, the resting state present after the limiting reactant has been consumed, such as in a batch reactor, may be different from the resting state present in a continuous reactor. Consider the rhodium-catalyzed hydrosilylation of ketones. The textbook mechanism when Wilkinson’s catalyst is employed is depicted in Figure 1.2.4 To the author’s knowledge, there have been no attempts to identify the resting state, although the formation of the rhodium alkyl complex 6 is believed to be rate determining. However, at the completion of the reaction, it would be more intuitive for the resting state to feature a ketone or ketone-derived ligand (e.g., 5) when the ketone is in excess. The system can be said to be ‘starved in silane’. On the other hand, it would be more intuitive for the resting state to be some type of rhodium/silane adduct (e.g., 4) when the silane is in excess (system ‘starved in ketone’). Importantly, there are many catalysts that can be recovered exactly as they were added to the reaction. In other words, these catalysts rest in their ‘native form’. There is no ‘catalyst precursor’, just the catalyst. Examples would include many reactions catalyzed by amines (Et3N, DMAP, DBU) and main-group Lewis acids. For instance, the Lewis acid might form only a very weak complex with the starting material that it activates, as well as the product. A case in point would be the fluorous phosphine catalyst 1 in Figure 1.1, which can be observed by NMR at the end of each cycle.2 Although the initial step involves a Michael addition of 1 to the a,b-unsaturated ketone 2, the equilibrium favors the reactants. The analogous addition of 1 to product 3, which is illustrated in Figure 1.3 (top) is

4

Recoverable and Recyclable Catalysts O

OSiR3 ClRh(PPh3)3 + R3SiH

+ R3SiH

H

Rh(PPh3)2Cl

Rh(PPh3)3Cl

Rh(PPh3)2Cl

– PPh3

R3Si

4 O

OSiR3 +

– H H

Rh(PPh3)2Cl

R3Si

Rh(PPh3)2Cl

R3SiO

O

6

5

Figure 1.2 Ojima’s mechanism for the hydrosilylation of ketones by Wilkinson’s catalyst

also energetically uphill. A counter example would be the familiar AlCl3 promoted Friedel–Crafts acylation. As pointed out in rigorous textbooks,5 the AlCl3 that activates the acyl chloride subsequently binds to the carbonyl group of the aryl ketone product, preventing turnover (Figure 1.3, middle). However, Friedel–Crafts alkylations, for which the products are much less Lewis basic, are catalyzed by AlCl3 (Figure 1.3, bottom). weak catalyst /product interaction (Figure 1.1): OH

OH

O–

O + P((CH2)3(CF2)7CF3)3

Ph 3

Ph +PR

1

strong "catalyst"/product interaction:

3

AlCl3

O

O + AlCl3

weak catalyst/product interaction: H AlCl3–

+ AlCl3 R

AlCl3 R

+

or R

Figure 1.3 Representative catalyst–product interactions

The Experimental Assay of Catalyst Recovery: General Concepts recovered catalyst or resting state

may be co-isolated

decomposed catalyst

5

by definition not co-isolated

leached catalyst

may be co-isolated

Figure 1.4 ‘Catalyst inventory’: what becomes of the initial charge?

1.4 Catalyst Inventory: Loss Mechanisms When studying any recoverable catalyst, it is a useful exercise to consider all possible modes of catalyst loss, with the goal of completely mass balancing the original catalyst charge. As diagrammed in Figure 1.4, this can be termed a ‘catalyst inventory’. The most obvious contributions are (a) catalyst decomposition and (b) catalyst leaching. When these are added to the recovered catalyst or catalyst resting state, the catalyst charge should be largely accounted for. 1.4.1 Catalyst Decomposition This can happen via a number of pathways. One intuitive and well-established example would involve oxygen sensitivity, or other oxidation modes. This is frequently observed with metal containing catalysts, and the phosphine catalyst 1 in Figure 1.1 becomes increasingly contaminated with the phosphine oxide O¼P((CH2)3(CF2)7CF3)3 with each cycle. However, many catalysts contain the seeds of their own destruction, and simply deactivate in the course of many turnovers. Any decomposition products that can be fished out can provide valuable guidance on how to design longer-lived catalysts that give higher turnover numbers. These in turn provide superior starting points for recoverable catalysts. Said differently, it is a waste of time to try to immobilize or recover a poor catalyst that easily deactivates. Well-studied contemporary examples include Grubbs’-type olefin metathesis catalysts, which in many cases are not particularly long lived.6 As shown in Figure 1.5 (top), Grubbs’ second generation catalyst 7 can decompose to give the novel diruthenium carbide complex 8. After the initial PCy3 dissociation that allows access to the catalytic cycle, formation of a RuCH2PCy3 species and dissociation of the ylide H2CPCy3 are believed to occur. Catalyst precursors with other types of labile ligands, or the addition of scavenging agents that can trap PCy3, may give longer-lived catalysts. Alternatively, well-defined polymer-supported metathesis catalysts are becoming increasingly available.7 When these are immobilized via the nondissociating ligands, as exemplified by the polystyrene-

6


N

N

N 0.023 M

Cl Ru CH2 Cl PCy3

C 6H 6 55 °C

Ru N

C Ru

H

Cl

N

7

PS

N Cl

+

CH3PCy3+ Cl−

Cl

8

9

O N

N

N

Cl Ru Cl O

10

O PS

O

PS = polystyrene/divinylbeneze copolymer

O X

F2 F2

F2

N Ru

O O

11 X = CF3C(=O)O

Figure 1.5 Decomposition of Grubbs’ second generation olefin metathesis catalyst (top), and polymer-immobilized, recyclable versions of the second generation Grubbs–Hoveyda catalyst (bottom)

anchored species 10 and 11 in Figure 1.5 (bottom), the ruthenium sites are kept at a distance. The probability of generating a diruthenium decomposition product such as 8 is then reduced. Both 10 and 11, which are variants of the Grubbs–Hoveyda catalyst and lack the problematic PCy3 ligand of 7, provide higher turnover numbers (TON) as compared to nonimmobilized analogs. Nonetheless, such efforts only ‘change the game’. There may be additional modes of catalyst decomposition, too slow to observe with 7, which can now manifest themselves. In the same vein, there has been much recent interest in iron tetramido catalysts of the type 12 (Figure 1.6).8 These catalyze a variety of oxidations by hydrogen peroxide and organic peroxides in aqueous solution. In a multi-year effort, the ligands were optimized (see 13–17) from the standpoints of oxidative and hydrolytic stability. This undertaking was facilitated by the careful analysis of ligand degradation products. The obvious starting point for a recoverable catalyst would be the geminal difluoride 17. Decomposed catalysts can (a) phase separate, such as via precipitation, (b) co-leach with leached catalyst, (c) be co-recovered with the catalyst or catalyst resting state, or (d) be removed at other stages of the recycling process. In particular (c) is a worrisome complication, and one example (the presence of some phosphine oxide in the recovered 1 in Figure 1.1) was noted above. Hence, some type of purity or activity assay for the recovered catalyst is essential (see below).


7

–

OH2 O N

X

X

O

N

Fe

N

N

R

O R

O

12 X

O

N

O

N

O

HH HH O

N

N

N

O

HH

N

O

N

N

O

OH HO

O

13 (1985) X

N

O

X

O

N HH HH

R

N

N

R

R O

O

X

N

O R

O

N HH HH

R

N

R

N

R

R O

O F

16 (1996)

O

15 (1989)

14 (1987) X

O

N HH HH

N

O

X

F

17 (2000)

Figure 1.6 Optimization of ligands for oxidation catalyst 12: increasing oxidative and/or hydrolytic stabilities

1.4.2 Catalyst Leaching Conceptually, catalyst leaching is usually associated with a phase boundary, and there are many types. For example, the active component of an insoluble or polymer-bound catalyst might slowly leach into solution by some mechanism, perhaps involving bond breaking. Alternatively, in liquid/liquid biphase catalysis, the catalyst might leach into the liquid phase that is not recycled, or that may contain the product. This tendency can be quantified by a partition coefficient. As foreshadowed above, fine distinctions can be made about ‘what actually leaches’. When the catalyst or catalyst resting state has leached into a product phase, the sample should exhibit some catalytic activity. Alternatively, an inactive decomposed catalyst might leach. There are many scenarios under which such species might cross a phase boundary more readily than the catalyst or catalyst resting state. Also, consider the case where a ligand must dissociate from a catalyst precursor before the catalytic cycle can be entered. The ligand might leach with little consequence for the recyclability of the catalyst. Various measures of catalyst leaching must be interpreted in these contexts. For example, atomic absorption spectroscopy and ICP–MS are very sensitive analytical

8


methods. The total rhodium leached from the hydrosilylation reaction in Figure 1.2 could easily be measured. However, the value would reflect the sum of the leached catalyst (resting state) and the leached decomposed catalyst. Other probes, such as NMR (a much less sensitive method) would be able to specifically measure the active catalyst.

1.5 Evaluation of Catalyst Recovery 1.5.1 Product Yield, Conversion, or TON as a Function of Cycle: Poor and Potentially Deceptive Criteria The majority of papers describing catalyst recycling reports one or more of these types of data as a function of cycle. Importantly, all are poor measures of recoverability. Suppose it has been arbitrarily decided to let each cycle run for 1 h. However, under the conditions of cycle 1, product formation is actually complete after 1 min. Then assume that half of the catalyst is lost in the first recycling operation. Cycle 2 could still give complete product formation in two minutes (given reasonable rate law assumptions, etc.). Suppose half of the catalyst is lost on each succeeding cycle. No major yield deterioration would be noted until cycle 8, despite the recurring losses. However, the rate of product formation or turnover frequency (TOF) would decrease markedly from cycle to cycle. This would be instantly evident from the types of plots in Figure 1.1. A case from the author’s laboratory illustrates ‘how to make a nonrecoverable catalyst look recoverable’. The fluorous palladacycle 18 in Figure 1.7, which features bridging acetate ligands, was evaluated as a catalyst precursor for the Heck reaction of phenyl iodide and methyl acrylate in DMF at 100–140 C. Recovery was effected by precipitation at room temperature and solid/liquid phase separation, similar to the protocol in Figure 1.1.9 Since an iodide salt is produced as a byproduct, the palladacycle was actually recovered with bridging iodide ligands. Screening experiments were conducted at somewhat high catalyst loadings for a Heck reaction (0.5 mol%). In duplicate runs, good yields were maintained over eight cycles, an impressive number. Data from a thesis, archived to illustrate precisely this point, are summarized in Table 1.1. Fortunately, there was not a rush to publish these results immediately.

DMF, 100 °C, 2 h I

+

CO2Me

+

Et3N

+ 0.5 mol% Rf8

(1.0 equiv)

(1.25 equiv)

(2.0 equiv) Rf8

Et3NH+ I–

CO2Me Rf8 N Pd

OAc

2 18 (Rf 8 = (CF2)7CF3)

(recovered as bridging iodide by precipitation/C8F17Br extraction)

Figure 1.7 A nonrecyclable palladium catalyst that can give deceptively high product yields over many cycles


9

Table 1.1 Data for the reaction in Figure 1.7 Trial 1 Cycle 1 2 3 4 5 6 7 8 a b

Conversion (%)a 95 95 96 94 91 90 77 60

Yield (%)b 86 85 98 83 75 75 50 56

Trial 2 TON

Conversion (%)a

Yield (%)b

TON

193 191 221 187 169 167 113 126

99 97 98 92 90 95 86 89

91 89 85 78 72 86 70 75

204 200 191 175 162 194 158 169

Conversion of phenyl iodide. Yield of methyl cinnammate.

The exceptionally talented coworker involved was well aware of the pitfalls outlined above. Experiments were next conducted at lower loadings, and the yields deteriorated over just 2–3 cycles. Rate profiles confirmed the loss of activity, and showed induction periods for every cycle (not just the first). Additional experiments indicated that 18 simply acted as a slow steady state-source of colloidal palladium(0) nanoparticles, which supplied the active catalysts. Only the undecomposed fluorous palladacycle was recycled, and when it was exhausted, activity ceased. A few investigators monitor conversion to product for the first cycle, and determine the ‘minimum time’ required to reach an arbitrary high conversion value. For example, perhaps 90% conversion is reached after 1 h (and only 85% after 0.9 h). In subsequent cycles, the conversion is assayed after 1 h. In such cases, the conversion would largely parallel the catalyst recovery. Nonetheless, this is a ‘poor man’s’ method; it should not be significantly more difficult to determine the complete rate profile as in Figure 1.1 and in the following section. 1.5.2 Reaction Rate or TOF as a Function of Cycle Such experiments give meaningful data, and represent the optimum ways to quantify catalyst recovery. However, they are absent in the majority of papers involving catalyst recycling. ‘Rate profiles’ as illustrated in Figure 1.1 above, or alternatively rate constants, will almost always accurately reflect catalyst recovery. If the recovered catalyst has partly decomposed, or been lost to leaching, this will automatically be mirrored by a lower rate. With systems that give induction periods, comparisons between the second and third (and subsequent) cycles are more meaningful than between the first and second. Figure 1.8 provides a representative case involving the hydrosilylation of ketones with the fluorous rhodium catalyst 19, which incorporates a lower homolog of the phosphine ligand 1 in Figure 1.1.10 Reactions were conducted under monophasic conditions at an elevated temperature, but cooling afforded biphasic systems with the catalyst and products in different solvents (fluorous and organic). Relevant to the discussion of the catalyst resting state in Figure 1.2 above, note that the silane is used in 10% excess. The yields of the

10


O PhMe2SiH

+

0.20 mol% 19 40 °C CF3C6F11/hexanes

OSiMe2Ph

(1.43 mmol)

(1.30 mmol)

P(CH2CH2(CF2)5CF3)3 19

Cl

Rh P(CH2CH2(CF2)5CF3)3 P(CH2CH2(CF2)5CF3)3

Reactants hexanes

40 °C

Reactants

cool

Products hexanes

19 19 CF3C6F11

Products

19 CF3C6F11

biphasic

monophasic

biphasic

separate product and catalyst phases and recycle

O 1.0

0.6 mmol

0.2

OSiMe2Ph 1.0 20 Cycle 1 Cycle 2 Cycle 3 Cycle 4

0.6 mmol

0.2

1.0 h

2.0 h

Figure 1.8 A recyclable fluorous rhodium ketone hydrosilylation catalyst that exhibits an induction period with the first cycle


11

Figure 1.9 A recyclable silica-immobilized palladium catalyst for the Heck reaction that exhibits an induction period with the first cycle. The graph is reproduced by permission of The Royal Society of Chemistry. J. H. Clark, D. J. Macquarrie, E. B. Mubofu, Preparation of a novel silica-supported palladium catalyst and its use in the Heck reaction, Green Chemistry, 2000, 2, 53–56

products and unreacted ketones were independently assayed by GLC vs an internal standard. There are moderate losses of activity between the second and third cycles, and between the third and fourth. However, not accounting for the induction period, product formation is faster in the second cycle than the first. The origin of the induction period is not obvious, but may be connected to a requirement for phosphine loss. Another example, involving a palladium catalyst precursor for the Heck reaction that has been immobilized on silica (20), is depicted in Figure 1.9.11 In contrast to the palladacycle 18 in Figure 1.7, 20 only exhibits an induction period for the first cycle. This might reflect a need to generate a small amount of palladium(0), which is generally viewed as the optimal oxidation state for initiating Heck reactions, and for which the Et3N could serve as a sacrificial reductant. Alternatively, it might be connected to the generation of palladium nanoparticles, which were the active species in Figure 1.7. However, in view of the rate profiles in Figure 1.10, the nanoparticles would have to be efficiently retained on the silica. Recyclable catalysts for the Heck reaction, and attendant controversies, have greatly helped to clarify and refine a number of mechanistic details.12 When rate profiles are generated by removing small aliquots from reaction mixtures, they can slightly underestimate catalyst recovery. If the aliquots are not returned, the amount of catalyst that can be recovered is decreased. If a sufficient volume% (mol%) is removed, the apparent efficiency of subsequent cycles will be significantly lowered. There are a number of ways to correct for this, such as by slightly decreasing the reactant and solvent charge in

12


Figure 1.10 Conversion as a function of time for the Heck reaction in Figure 1.7, conducted with 0.02 mol% 20

subsequent cycles. This was done for the experiments summarized in Figure 1.1.2 Note that the maximum yield of 3 in Figure 1.1 in 85%. Since all of the starting material is consumed, this indicates the formation of byproducts, as supported by the bright red color of the reaction mixtures.2 There is an analogous discrepancy between substrate consumption and product formation for the Heck reaction in Figure 1.7. Also, catalyst recovery is best assayed by comparing two consecutive cycles at partial conversion. There is little to be learned when the first data point in one cycle corresponds to 90% conversion and the first data point in the next cycle corresponds to 85% conversion. With measurements at lower conversions, there could be dramatic differences, reflecting poor catalyst recovery. In practice, most researchers address this need by lowering the temperature and/or the catalyst loading, such that the rate decreases. Figure 1.10 illustrates this point for the reaction in Figure 1.7 when carried out with a 0.02 mol% charge of 18.9 At 100 C, the induction periods and the rate difference between the first (&) and second (~) cycles are less pronounced. When the reaction is carried out at 80 C (^, ), the slower rates bring these features into sharper focus. As noted above, the persistence of an induction period in the second cycle indicates that the active catalyst is not being recycled. 1.5.3 Gravimetric and Other Assays of Recovered Catalyst In much bench-scale catalyst development, catalyst quantities are deliberately kept small. The whole point of catalyst design is to create a highly active species capable of thousands or millions of turnovers. For other desirable features, the reader is referred to essays on the ‘ideal catalyst’.13


13

In any event, catalyst quantities are often too small for reliable gravimetric recovery determinations. For any amount less than 10.0 mg, only two significant digits are possible on common analytical balances, and there is substantial uncertainty in the second digit (0.2 mg under the best circumstances). After determining the mass, the catalyst purity still must be assayed, as decomposed catalyst and other contaminants may be present. Finally, the catalyst resting state may not correspond to the species initially added, complicating mass comparisons. Of course, the reaction can be scaled up. But it would never be advisable to scale up solely the catalyst loading for the purpose of quantifying recovery. One simplifying variation is to gravimetrically assay catalyst recovery after x cycles (for example five), taking special care to do careful weight and purity determinations. An overall recovery of 90% would indicate ca. 98% recovery per cycle. Many examples of this approach are available.14,15 Other information can also be obtained. After the five cycles in Figure 1.1, the recovered phosphine catalyst 1 was found to be a 76:24 mixture of 1 and the corresponding phosphine oxide, corresponding to ca. 5% oxidation per cycle. Since separate experiments showed the phosphine oxide to be inactive, this accounts for nearly all of the loss of activity. The activity loss is best gauged from the data points taken at 3–5 h. There are a number of obvious alternatives to gravimetric assays. These would include quantitative spectroscopic assays, such as UV/visible measurements or NMR analyses in the presence of an internal standard. In some cases, chromatographic assays may also be feasible. 1.5.4 Special Caveats when ‘Residues’ are Recycled When liquid phases are recycled, intermediate purification steps, such as filtration, are possible. However, when solid residues are recycled, many species can ‘come along for the ride’, even when the sample is washed. Hence, the species that is actively being recycled may not be the true catalyst responsible for the rate profiles. One obvious contaminant would be finely divided metal particles or nanoparticles, which are extremely active heterogeneous catalysts for many reactions. Metal nanoparticles can be visualized by transmission electron microscopy, but the more relevant question is whether they are responsible for the observed reaction. Towards this end, it is common to add a small amount of elemental mercury. This amalgamates onto most metals, passivating the surface and poisoning any heterogeneous catalysis.16 If catalysis persists, this constitutes good evidence for a homogeneous molecular catalyst. If catalysis ceases, there still are mechanisms by which mercury could quench a homogeneous reaction. However, if the presence of nanoparticles can be confirmed by other means, a heterogeneous pathway is more likely. As a specific example, the hydrosilylation reactions in Figure 1.8 can also be conducted using fluoropolymer supports instead of fluorous solvents.9b Very similar rate profiles are obtained, with an induction period in the first cycle. When elemental mercury is added at the start of the second cycle, the rate profile attenuates only slightly. Hence, a homogeneous, molecular rhodium species must be largely responsible for the catalysis.

1.6 Prospective There may be a tendency among beginning researchers to mistakenly view projects involving recoverable catalysts as ‘all or nothing’ in terms of success. A dissertation or grant

14


application may seem to hang in the balance, and this may reinforce a tendency, also on the part of the project advisor, not to probe the catalytic system or data such as in Figure 1.7 too deeply. A higher catalyst loading for a reaction that is feasible at much lower loadings can be a danger sign. Thus, experienced researchers maintain a skeptical eye on the literature. However, with attention to the principles and quantitative details summarized above, the author would argue that there is little need for a coworker to fear that a careful study could not appear in the literature or a dissertation. Any effort that sets a reliable baseline, even if the efficiency of catalyst recovery is only marginal, provides a valuable reference point for future studies. In many cases, there are obvious approaches to improving and optimizing the initial data. With the above guidance, you are now fully armed to make a significant fundamental impact on the field of recoverable catalysts. If you are like many, you have some ‘designer catalyst’ or creative new protocol in mind. And if you ‘hit gold’, there are riches to be had. So proceed with vigor and rigor. The community awaits the creative input of the next generation of chemists in this field. The Welch Foundation is warmly thanked for financial support.

References 1. Rothenberg, G. Catalysis, Wiley/VCH, Weinheim, 2008, p 10. 2. Seidel, F. O.; Gladysz, J. A. Adv. Synth. Catal. 2008, 350, 2443. 3. Spessard, G. O.; Miessler, G. L. Organometallic Chemistry, Prentice Hall, Upper Saddle River, New Jersey, USA, 1997, pp 277–280. 4. Ojima, I.; Kogure, T. Organometallics 1982, 1, 1390, and references therein. 5. Streitweiser, A.; Heathcock, C. H.; Kosower, E. M. Introduction to Organic Chemistry, 4th edn; Macmillan, New York, USA, 1992, pp 696–701. 6. Hong, S. H.; Day, M. W.; Grubbs, R. H. J. Am. Chem. Soc. 2004, 126, 7414. 7. Buchmeiser, M. R. Chem. Rev. 2009, 109, 303. 8. (a) Collins, T. J. Acc. Chem. Res. 2002, 35, 782. (b) Polshin, V.; Popescu, D.-L.; Fischer, A.; Chanda, A.; Horner, D. C.; Beach, E. S.; Henry, J.; Qian, Y.-L.; Horwitz, C. P.; Lente, G.; Fabian, I.; M€unck, E.; Bominaar, E. L.; Ryabov, A. D.; Collins, T. J. J. Am. Chem. Soc. 2008, 130, 4497. 9. (a) Rocaboy, C.; Gladysz, J. A. New J. Chem. 2003, 27, 39. (b) Rocaboy, C. Doctoral Dissertation, University of Utah, 2002. 10. (a) Dinh, L. V.; Gladysz, J. A. New. J. Chem. 2005, 29, 173. (b) Dinh, L. V.; Gladysz, J. A. Angew. Chem., Int. Ed. 2005, 44, 4095; Angew. Chem.2005, 117, 4164. 11. Clark, J. H.; Macquarrie, D. J.; Mubofu, E. B. Green Chem. 2000, 2, 53. 12. Phan, N. T. S.; van der Sluys, M.; Jones, C. W. Adv. Synth. Catal. 2006, 348, 609. 13. Gladysz, J. A. Pure Appl. Chem. 2001, 73, 1319. 14. Some representative examples involving fluorous catalysts: (a) Ishihara, K.; Kondo, S.; Yamamoto, H. Synlett 2001, 1371. (b) Contel, M.; Villuendas, P. R.; Fernandez-Gallardo, J.; Alonso, P. J.; Vincent, J.-M.; Fish, R. H. Inorg. Chem. 2005, 44, 9771. (c) Sokeirik, Y. S.; Mori, H.; Omote, M.; Sato, K.; Tarui, A.; Kumadaki, I.; Ando, A. Org. Lett. 2007, 9, 1927. 15. Catalysis involving fluorous phases: fundamentals, directions for greener methodologies, Gladysz, J. A. in Handbook of Green Chemistry, vol. 1, Anastas, P.; Crabtree, R. H. Eds; Wiley-VCH, Weinheim, 2009, pp 17–38. 16. (a) Whitesides, G. M.; Hackett, M.; Brainard, R. L.; Lavalleye, J.-P. P. M.; Sowinski, A. F.; Izumi, A. N.; Moore, S. S.; Brown, D. M.; Stauch, E. M. Organometallics 1985, 4, 1819. (b) Widegren, J. A.; Finke, R. G. J. Mol. Catal. A 2003, 317.

2 Surface-functionalized Nanoporous Catalysts for Renewable Chemistry Brian G. Trewyn, Hung-Ting Chen and Victor S.-Y. Lin Department of Chemistry, U.S. Department of Energy Ames Laboratory, Iowa State University, Ames, Iowa, USA

2.1 Introduction The traditional chemical synthesis developed from the late 19th century mainly focused on the use of stoichiometric reagents through a sequence of functional group transformations converting starting materials to products. As the discovery of diverse reactive reagents, the chemical synthesis is widely used to produce a large number of fine chemicals leading to many materials and devices that are crucial to modern civilization. However, utilizing extremely reactive reagents typically requires strict reaction environments, such as oxygenor moisture-free conditions, to achieve the desired reaction kinetics and product yields. Moreover, the stoichiometric requirement of chemical reagents brings in a considerable amount of waste, bearing an environmental burden. To circumvent these drawbacks, the concept of catalysis was introduced to organic synthesis. Over the past few decades, many selective and efficient catalysts have been developed leading to an established field of research in the 1980s. Among a large variety of catalytic systems, the transition metal-mediated asymmetric catalysis, such as enantioselective hydrogenation and allylic epoxidation, represents two classic models at this period of time. Combining with modern spectroscopic methods, the studies of these homogeneous catalysts in terms of real active species, transition state, and detailed mechanistic aspects are


16


feasible. The thorough understanding of homogeneous catalytic centers on the molecular level provides much precious information, which made the rational design and improvement of a catalytic system possible. The unprecedented success of homogeneous catalysis represents a major breakthrough in chemistry and creates a new era in organic synthesis. 2.1.1 Homogeneous Catalysis vs Heterogeneous Catalysis Despite the impressive accomplishments of homogeneous catalysis, these sophisticated molecular catalytic systems are rarely applied to the petroleum and chemical industry. Today, most delicate transition metal complexes, especially the chiral auxiliaries applied in asymmetric reactions, are much more valuable than either their reactants or products. In addition, a considerable production cost is spent on the isolation of metal complex-based catalysts due to the pronounced toxicity of transition metal residues in solution. Obviously, recovering catalysts from reaction mixtures for recycling is important, but much more difficult in the homogeneous catalytic systems because of the high solubility of these molecules. Moreover, homogeneous catalysts are not designed to fit into continuous flowtype reactors, which are economically attractive to many industries. On the other hand, the simple properties of heterogeneous catalysts make separation of these materials from the reaction media much easier. Repeated usage of solid catalysts substantially reduces the production costs and chemical waste. For instance, replacing mineral acid with solid acid avoids the corrosion problems, extending the operation time of reactors, and reduces the cost of chemical neutralization. Heterogeneous catalysis research, burgeoning in the mid-1950s, was developed originally for oil refinement in the petrochemical industry. Compared to homogeneous catalysts in solution, transitional heterogeneous catalysis is closely related to surface science and solid-state chemistry. However, the active sites on the surface of heterogeneous catalysts such as metal or alloys are less sharply defined and possess a spectrum of energetic state and activity. It has been well documented that the surface defects on many metal and metal oxides, such as kicks or steps, indeed function as favored sites for the activation of CH, HH, O¼O. N:N, or C:O bonds. As far as those catalysts are concerned, precise engineering of active sites in the heterogeneous catalysis to sustain high catalytic performance becomes impractical. 2.1.2 Multi-Site vs Single-Site Heterogeneous Catalysis Heterogeneous catalysis falls into two main categories: multi- or connected-site heterogeneous catalysis and single-site heterogeneous catalysis. Multi- or connected-site heterogeneous catalysis is composed of closely packed atoms of metals, metal oxides, alloys and halides. The active sites of this type of heterogeneous catalyst are not spatially independent and have strong energetic interactions among them. These interactions make thermodynamic and kinetic analysis of the catalytic reactions very challenging. In contrast, single site heterogeneous catalysis is unique in that the active sites are spatially well separated, lack multiple active site interaction, and possess the same interaction energy between active sites and reactant. Investigation first began with Grasselli and coworkers1 over four decades ago and more recently, Professor John M. Thomas has done considerable investigations on single site heterogeneous catalysis and has defined the single site heterogeneous catalysis as simply spatially isolated homogeneous catalysts on the solid surface.2–4 This review of

Surface-functionalized Nanoporous Catalysts for Renewable Chemistry

17

renewable and reusable heterogeneous catalysis will discuss both multi-site and single-site heterogeneous catalysis in MCM-41 type mesoporous materials.

2.2 Immobilization Strategies of Heterogeneous Catalysts 2.2.1 Supported Materials The performance of heterogeneous catalysts is strongly dependent upon the choice of support material and the method by which the active site is immobilized on the surface. The activity of the active site will vary depending upon the nature of the surface. A few characteristics that a successful support material should have, but are not limited to, include: . . . .

Chemical, thermal, and mechanical stability; Large surface area and abundant surface functionalities; Rapid mass transport of reactants and products to and from the active sites; Interior and exterior surface areas that are independently functionalized.

Organic polymers, especially polystyrene-type resins, represent one of most popular commercialized support.5 This resin is commonly synthesized by copolymerization of active monomer, such as halogenated methyl styrene, in the presence of cross-linkers. With a low content of cross-linker, the resulting polymer requires compatible solvent swelling for the reactant to access internal active sites. For instance, low reactivity is typically observed for such supports in protic, high polar media such as water or alcohol. Increasing the amount of cross-linkers solidifies polymer resins, diminishing the swelling problems. However, they sometimes suffer from brittleness and poor loading capacity of active sites due to illdefined internal porous structures. Comparing with their organic counterparts, inorganic solid supports, such as silica, zeolite, alumina, zirconia, zinc oxide, clays etc., generally show thermal and mechanical resistance. Moreover, the relatively high surface area and appropriate pore sizes of the inorganic support maintain their competitive advantages over others. For reasons of cost, availability, mechanical robustness, and synthesis; silica is the most popular support matrix for inorganic solid catalytic support. Silica consists of fully condensed silanoxy bridges (:SiOSi:) in the framework and silanol functional groups (:SiOH) on the surface. Three types of silanol groups: isolated, vicinal (hydrogen bonded), and germinal, occur on the silica surface. The existence and relative amount of surface silanol groups can be easily determined by infrared (IR) spectroscopy. The density of silanol groups per gram of silica depends on surface area and method of silica formation. Generally, higher surface area results in more silanol groups. Silicas synthesized through the sol–gel route contain more silanol groups, due to incomplete condensation of molecular precursors. Thermal treatment of silica samples induces further condensation of silanol group to form complete silanoxy bridges. This condensation process has also been shown to be reversible and samples reconvert into silanols by hydration. Notably, the structure and morphology of silicate supports have a vital influence on the catalytic activity. Granular silica gel, widely used as desiccants, was frequently seen in the early literature serving as supports.6,7 Although several heterogeneous catalyst systems on

18


silica gel have been successfully demonstrated in publications and patents, the relative low surface area of these materials limit the applications. Later on, zeolite replaced silica gel as a popular support since its discovery, as Prof. Sanchez and coworkers have shown with several examples of heterogenizing catalysts on zeolites.8–12 Zeolites are crystalline alumisilicates with specific internal microporous (pore size 98% purity

34

O

Scheme 3.24 Combined use of insoluble resin-supported catalysts in multicomponent reactions

3.8 Conclusion In summary, the use of insoluble resins as catalyst carriers has found extensive applications in various organic reactions. In many cases, this strategy has demonstrated its great advantage in simplifying the separation of products and the recovery and reuse of the usually expensive catalysts while maintaining comparable catalytic performance to that of the unsupported parent catalysts. In addition, the development of microreactors with continuous flow reaction systems greatly added to its utility in organic synthesis. However, drawbacks such as diminished catalytic activity, low loading capacity, low mechanical stability and especially deterioration in enantioselectivity in asymmetric reactions still occurred in many cases. Therefore, there is still a large room for further improvements.

References 1. C. C. Leznoff, The use of insoluble polymer supports in organic chemical synthesis, Chem. Soc. Rev., 3, 65–85, (1974). 2. E. C. Blossey and W. T. Ford in Comprehensive Polymer Science. The Synthesis, Characterization, Reactions and Applications of Polymers, D. C. Sherrington and Hodge (Eds), Wiley, Chichester, 1988. 3. D. E. Bergbreiter, Organic polymers as a catalyst recovery vehicle, in Chiral Catalyst Immobilization and Recycling, D. E. De Vos, I. F. J. Vankelecom and P. A. Jacobs (Eds), Wiley–VCH, Weinheim, 2000. 4. C. A. McNamara, M. J. Dixon and M. Bradley, Recoverable catalysts and reagents using recyclable polystyrene-based supports, Chem. Rev., 102, 3275–3300 (2002).

Insoluble Resin-supported Catalysts

73

5. M. Benaglia, A. Puglisi and F. Cozzi, Polymer-supported organic catalysts, Chem. Rev., 103, 3401–3429 (2003). 6. A. Kirschning, H. Monenschein, R. Wittenberg, Functionalized polymer-emerging versatile tools for solution-phase chemistry and automated parallel synthesis, Angew. Chem. Int. Ed., 40, 650–679 (2001). 7. (a) R. Haag and S. Roller, Polymeric supports for the immobilization of catalysts, Top. Curr. Chem., 242, 1–42 (2004); (b) Q.-H. Fan, Y.-M. Li and A. S. C. Chan, Recoverable catalysts for asymmetric organic synthesis, Chem. Rev., 102, 3385–3466 (2002); (c) P. McMorn and G. J. Hutchings, Heterogeneous enantioselective catalysts: strategies for the immobilisation of homogeneous catalysts, Chem. Soc. Rev., 33, 108–122, (2004). 8. (a) D. C. Sherrington, Preparation, structure and morphology of polymer supports, Chem. Commun., 2275–2286 (1998); (b) B. M. L. Dioos, I. F. J. Vankelecom and P. A. Jacobs, Aspects of immobilization of catalysts on polymeric supports, Adv. Catala. Synth., 346, 1413–1446 (2006). 9. G. Gelbard, Organic synthesis by catalysis with ion-exchange resins, Ind. Eng. Chem. Res., 44, 8468–8498 (2005). 10. (a) F. Cozzi, Immobilization of organic catalysts: when, why and how, Adv. Catala. Synth., 348, 1367–1390 (2006); (b) G. Guillena, C. Najera, D. J. Ramon, Enantioselective direct aldol reactions: the blossoming of modern organocatalysis, Tetrahedron: Asymmetry, 18, 2249–2293 (2007). 11. (a) Y. Uozumi, Recent progress in polymeric palladium catalysts for organic synthesis, Top. Curr. Chem., 242, 77–112 (2004); (b) V. Farina, High-turnover palladium catalysts in cross-coupling and Heck chemistry: a critical review, Adv. Catala. Synth., 346, 1553–1582 (2006). 12. (a) R. Akiyama and S. Kobayashi, The polymer incarcerated methods for the preparation of highly active heterogeneous palladium catalysts, J. Am. Chem. Soc., 125, 3412–3413 (2003). (b) K. Okamoto, R. Akiyama and S. Kobayashi, Recoverable, reusable, highly active and sulfurtolerant polymer incarcerated palladium for hydrogenation, J. Org. Chem., 69, 2871–2873 (2004); (c) K. Okamoto, R. Akiyama and S. Kobayashi, Suzuki–Miyaura coupling catalyzed by polymer–incarcerated palladium, a highly active, recoverable, and reusable Pd catalyst, Org. Lett., 6, 1987–1990 (2004). 13. Y. Kobayashi, D. Tanaka, H. Danjo and Y. Uozumi, A combinatorial approach to heterogeneous asymmetric aquacatalysis with amphiphilic polymer-supported chiral phosphine–palladium complexes, Adv. Catala. Synth., 348, 1561–1566 (2006). 14. S. A. Testero and E. G. Mata, Prospect of metal-catalyzed C–C forming cross-coupling reactions in modern solid-phase organic synthesis, J. Comb. Chem., 10, 487–497 (2008). 15. (a) J.-H. Kim, B.-H. Jun, J.-W. Byun and Y.-S. Lee, N-Heterocyclic carbine–palladium complex on polystyrene resin surface as polymer-supported catalyst and its application in Suzuki crosscoupling reaction, Tetrahedron Lett., 45, 5827–5831 (2004); (b) J.-W. Kim, J.-H. Kim, D.-H. Lee and Y.-S. Lee, Amphiphilic polymer supported N-heterocyclic carbene palladium complex for Suzuki cross-coupling reaction in water, Tetrahedron Lett., 47, 4745–4748 (2006). 16. K. C. Y. Lau, H. S. He, P. Chiu and P. H. Toy, Polystyrene-supported triphenylarsine reagents and their use in Suzuki cross-coupling reactions, J. Comb. Chem., 6, 955–960 (2004). 17. S. Doherty, J. G. Knight and M. Betham, The first insoluble polymer-bound palladium complexes of 2-pyridyldiphenylphospine: highly efficient catalysts for the alkoxycarbonylation of terminal alkynes, Chem. Commun. 88–90 (2006). 18. L. Pu and H.-B. Yu, Catalytic asymmetric organozinc additions to carbonyl compounds, Chem. Rev., 101, 757–824 (2001). 19. (a) S. Degni, C.-E. Wilen and R. Leino, Asymmetric C–C bond formation with L-proline derived chiral catalysts immobilized on polymer fibers, Tetrahedron: Asymmetry, 15, 231–237 (2004); (b) L.-T. Chai, Q.-R. Wang and F.-G. Tao, The synthesis of supported prolinederived ligands and their application in asymmetric diethylzinc addition to aldehydes, J. Mol. Catal. A: Chem., 276, 137–142 (2007); (c) J. Mao, Z. Bao, J. Guo and S. Ji, Enantioselective alkylnylation of aromatic and heteroaromatic aldehydes catalyzed by resin-supported oxazolidine-titanium complexes, Tetrahedron, 64, 9901–9905 (2008); (d) G. Lesma, B. Danieli, D. Passarella, A. Sacchetti and A. Silvani, Chiral amino-amides as solution phase and immobilized

74

20. 21.

22. 23. 24. 25.

26.

27. 28. 29. 30.

31. 32.

33. 34. 35.

Recoverable and Recyclable Catalysts ligands for the catalytic asymmetric alkylation of aromatic aldehydes, Lett. Org. Chem. 3, 430–436 (2006). H. M. L. Davies and A. M. Walji, Asymmetric intermolecular C–H activation, using immobilized dirhodium tetrakis(s)-N-(dodecylbenzenesulfonyl)-prolinate) as a recoverable catalyst, Org. Lett., 5, 479–482 (2003). (a) D. Rechavi and M. Lemarie, Enantioselective catalysis using heterogeneous bis(oxazoline) ligands: which factors influence the enantioseletivity? Chem. Rev., 102, 3467–3493 (2002); (b) O. Reister, Polymer bound bis(oxazoline)ligands for asymmetric catalysis, Chem. Oggi, 20, 33–77 (2002). M. I. Burguete, J. M. Fraile, E. Garcıa-Verdugo, S. V. Luis, V. Martınez and J. A. Mayoral, Polymer-supported bis(oxazolines) and related systems:toward new heterogeneous enantioselective catalysts, Ind. Eng. Chem. Eng., 44, 8580–8587 (2005). S. L. Jain and B. Sain, An efficient approach for immobilizing the oxo-vanadium Schiff base onto polymer supports using Stautinger ligation, Adv. Catala. Synth., 350, 1479–1483 (2008). M. R. Maurya, M. Kumar, U. Kumar, Polymer-anchored vanadium (IV), molybdenum (VI) and copper (II) complexes of bidentate ligand as catalyst for the liquid phase oxidation of organic substrate. J. Mol. Catal. A: Chem., 273, 133–143 (2007). X.-Y. Yuan, H.-Y. Li, P. Hodge, M. Kilner, C. Y. Tastard and Z.-P. Zhang, An easy synthesis of robust polymer-supported chiral 1,10 -bi-(2-naphthol)s (BINOLs): application to the catalysis of the oxidation of prochiral thioethers to chiral sulfoxides, Tetrahedron: Asymmetry, 17, 2401–2407 (2006). (a) E. Burri, M., Öhm, K., C. Daguenet and K. Severin, Site-isolated porphyrin catalysts in imprinted polymers, Chem. Eur. J., 11, 5055–5061 (2005); (b) E. Burri, S. M. Leeder, K. Severin and M. R. Gagne, Modulating the activity and selectivity of an immobilized ruthenium–porphyrin catalyst using a fluorous solvent, Adv. Catal. Synth., 348, 1640–1644 (2006). V. B. Valodkar, G. L. Tembe, M. Ravindranathan and H. S. Rama, Catalytic oxidation of alkanes and alkenes by polymer-anchored amino acid–ruthenium complexes, J. Mol. Catal. A: Chem., 223, 31–38 (2004). C. -P. Du, Z.-K. Li, X.-M. Wen, J. Wu, X.-Q. Yu, M. Yang and R.-G. Xie, Highly diastereoselective epoxidation of cholest-5-ene derivatives catalyzed by polymer-supported manganese (III) porphyrins, J. Mol. Catal. A: Chem., 216, 7–12 (2004). E. Brule, Y. R. De Miguel and K. K. Hii, Chemoselective epoxidation of dienes using polymersupported manganese porphyrin catalysts, Tetrahedron, 60, 5913–5918 (2004). (a) F. Shibahara, K. Nozaki and T. Hiyama, Solvent-free asymmetric olefin hydroformylation catalyzed by highly cross-linked polystyrene-supported (R, S)-BINAPHOS-Rh(I) complex, J. Am. Chem. Soc., 125, 8555–8560 (2003). (b) S. Kinoshita, F. Shibahara and K. Nozaki, Comparison of two preparative methods: a polymer-supported catalyst by metal-complexation with a polymeric ligand or by polymerization of a metal complex, Green Chem. 7, 256–258 (2005). J. Achkar, J. R. Hunt, R. L. Beingessner and H. Fenniri, Synthesis and catalytic activity of TantaGel-supported asymmetric dihydroxylation (DHQ)2PHAL ligand, Tetrahedron: Asymmetry, 19, 1049–1051 (2008). (a) G. Wang, J. Hu and G. Zhao, Enantioselective reduction of a-keto esters to 1,2-diols using the NaBH4/Me3SiCl system catalyzed by polymer-supported chiral sulfonamide, Tetrahedron: Asymmetry, 15, 807–810 (2004); (b) J. Hu, G. Zhao and Z. Zhao, Enantioselective reduction of ketones catalyzed by polymer-supported sulfonamide using NaBH4/Me3SiCl (or BF3OEt2) as reducing agent, Angew. Chem. Int. Ed., 40, 1109–1111 (2001). S. Degni, C.-E. Wilen and A. Rosling, Highly catalytic enantioselective reduction of aromatic ketones using chiral polymer-supported Corey, Bakshi, and Shibata catalysts, Tetrahedron: Asymmetry, 15, 1495–1499 (2004). S. Itsuno, M. Chiba, M. Takahashi, Y. Arakawa and N. Haraguchi, Asymmetric hydrogenation of aromatic ketones using polymeric catalyst prepared from polymer-supported 1,2-diamine, J. Organomet. Chem. 692, 487–494 (2007). (a) Y. Motoyama, K. Mitsui, T. Ishida and H. Nagashima, Self-encapsulation of homogeneous catalyst species into polymer gel leading to a facile and efficient separation system of amine

Insoluble Resin-supported Catalysts

36.

37.

38. 39. 40.

41. 42. 43.

44. 45. 46.

47.

48.

75

products in the Ru-catalyzed reduction of carboxamides with polymethylhydrosiloxane (PMHS), J. Am. Chem. Soc., 127, 13150–13151 (2005). (b) S. Hanada, Y. Y. Motoyama and H. Nagashima, Dual Si–H effects in platinum-catalyzed silane reduction of carboxamides leading to a practical synthetic process of tertiary-amines involving self-encapsulation of the catalyst species into the insoluble silicone resin formed, Tetrahedron Lett. 47, 6173–6177 (2006). (a) D. Font, C. Jimeno and M. A. Pericas, Polystyrene-supported hydroxyproline: an insoluble, recyclable organocatalyst for the asymmetric aldol reaction in water, Org. Lett., 8, 4653–4655 (2006); (b) D. Font, A. Bastero, S. Sayalero, C. Jimeno and M. A. Pericas, Highly enantioselective a-aminoxylation of aldehydes and ketones with a polymer-supported organocatalyst, Org. Lett., 9, 1943–1946 (2007). (a) F. Giacalone, M. Gruttadauria, A. M. Marculescu and R. Noto, Polystyrene-supported proline and prolinamide: versatile heterogeneous organocatalysts both for asymmetric aldol reaction in water and a-selenenylation of aldehydes, Tetrahedron Lett. 48, 255–259 (2007); (b) F. Giacalone, M. Gruttadauria, A. M. Marculescu, F. D’Anna and R. Noto, Polystyrene-supported proline as recyclable catalyst in the Baylis–Hillman reaction of arylaldehydes and methyl or ethyl ketone, Catal. Commun., 9, 1477–1481 (2008). R. D. Carpenter, J. C. Fettinger, K. S. Lam and M. J. Kurth, Asymmetric catalysis: resin-bound hydroxyprolylthreonine derivatives in enamine-mediated reactions, Angew. Chem. Int. Ed., 47, 6407–6410 (2008). C. Ogawa, M. Suhiura and S. Kobayashi, Polymer-supported formamides as reusable organocatalysts for allylation of aldehydes with allyltrichlorosilane, Chem. Commun., 192–193 (2003). (a) D. W. Kim and D. Y. Chi, Polymer-supported ionic liquids: imidazolinm salts as catalysts for nucleophilic substitution reactions including fluorinations, Angew. Chem. Int. Ed., 43, 483–485 (2004); (b) D. W. Kim, H.-J. Jeong, S. T. Lim, M.-H. Sohn and D. Y. Chi, Facile nucleophilic fluorination by synergistic effect between polymer-supported ionic liquid catalyst and tert-alcohol reaction media system, Tetrahedron, 64, 4209–4214 (2008); (c) S. S. Shinde, B. S. Lee and D. Y. Chi, Polymer-supported protic functionalized ionic liquids for nucleophilic substitution reactions: superior catalytic activity compared to other ionic resins, Tetrahedron Lett., 49, 4245–4248 (2008). Y. Du, F. Cai, D.-L. Kong and L.-N. He, Organic solvent-free process for the synthesis of propylene carbonate from supercritical carbon dioxide and propylene oxide catalyzed by insoluble ion exchange resins, Green Chem., 7, 518–523 (2005). Y. Xie, Z. Zhang, T. Jiang, B. Han, T. Wu and K. Ding, CO2 cycloaddition catalyzed by an ionic liquid grafted onto a highly cross-linked polymer matrix, Angew. Chem. Int. Ed., 46, 7255–7258 (2007). (a) C. Girard, E. Önen, M. Aufort, S. Beauviere, E. Samson and J. Herscovici, Reusable polymersupported catalyst for the [3 þ 2] Huisgen cycloaddition in automation protocols, Org. Lett., 8, 1689–1692 (2006); (b) H.-F. Jiang, A-Z. Wang, H.-L. Liu and C.-R. Qi, Reusable polymersupported amine–copper catalyst for the formation of a-alkylidene cyclic carbonates in supercritical carbon dioxide, Eur. J. Org. Chem., 2309–2312 (2008). G. A. Slough, V. Krchňak, P. Helquist and S. M. Canham, Synthesis of readily cleavable immobilized 1,10-phenanthroline resins, Org. Lett., 6, 2909–2912 (2004). J. Zhao and T. J. Marks, Recyclable polymer-supported organolanthanide hydroamination catalysts: immobilization and activation via dynamic transamination, Organometallics, 25, 4763–4772 (2006). (a) K. Bravo-Altamirano and J.-L. Montchamp, Palladium-catalyzed dehydrative allylation of hypophosphorous acid with allylic alcohols, Org. Lett., 8, 4169–4171 (2006); (b) S. Deprele and J.-L. Montchamp, Environmentally benign synthesis of H-phosphinic acids using a watertolerant, recyclable polymer-supported catalyst, Org. Lett., 6, 3805–3808 (2004). (a) Y. Yin, G. Zhao and G.-L. Li, Synthesis of polystyrene-bound perfluoroalkyl sulfonic acids and the application of their ytterbium salts in multicomponent reactions (MCRs), Tetrahedron, 61, 12042–12052 (2005). (b) G. Li and G. Zhao, Allylation of aldehydes and imines: promoted by reusable polymer-supported sulfonamide of N-glycine, Org. Lett., 8, 633–636 (2006). G. A. Strohmeier and C. O. Kappe, Combinatorial synthesis of functionalized 1,3-thiazine libraries using a combined polymer-supported reagents/catch–release strategy, Angew. Chem. Int. Ed., 43, 621–624 (2004).

4 Catalysts Bound to Soluble Polymers Tamilselvi Chinnusamy, Petra Hilgers and Oliver Reiser Institut f€ ur Organische Chemie, Universit€ at Regensburg, Regensburg, Germany

4.1 Introduction The immobilization of small organic catalysts (metal–ligand complexes or small organic molecules acting themselves as catalysts) onto soluble supports has become a widely applied strategy in organic synthesis that promises to combine the advantages of homogeneous catalysis with facile recovery of the catalyst. The latter property is highly desirable not only from an economic point of view – many metals that are favorites in catalysis (rhodium, iridium, or more recently gold) and (chiral) ligands are generally expensive; organocatalysts often suffer from comparatively low turn over numbers and therefore have to be employed in larger quantities – but also from an ecological and toxicological point of view, requiring the rigorous removal of metal traces from products such as drugs in the ppm range. Taking the great success of heterogeneous inorganic catalysts as the lead, most commonly immobilization of organic catalysts is aimed at their heterogenization, e.g. attaching them to an insoluble inorganic support such as silica or to an organic polymer such as cross-linked polystyrene, which allows recovery of the catalyst by simple filtration. However, the unparalleled selectivity and reactivity many homogeneous organic catalysts offer can generally not be achieved in the corresponding heterogeneous systems. Diminished reaction rates, especially for chemical transformations that have to be run below room temperature, are often encountered, making such catalysts impractical for large-scale applications. Moreover, analysis of such immobilized catalysts by standard techniques (NMR, MS, TLC, etc.) is generally not possible, making it difficult to assess their integrity, which is especially important after recovery. Recoverable and Recyclable Catalysts Edited by Maurizio Benaglia © 2009 John Wiley & Sons, Ltd. ISBN: 978-0-470-68195-4

78


A viable alternative is therefore the immobilization of organic catalysts onto soluble supports, which allows a given transformation to be carried out under similar reaction conditions to those for the nonsupported catalyst, and comparable results with respect to reactivity and selectivity can be expected. Especially, concerns regarding the morphology and in particular the swelling properties of a heterogeneous support, being important for the accessibility of catalytic sites for substrates, do not have to be addressed. Likewise, mechanical destruction of the support by too vigorous agitation (mechanical stirring) is not an issue. On the other hand, efficient recovery of the catalyst now becomes a critical issue, which can be achieved by making use of special properties of the catalyst that are imposed by the support attached to it. Precipitation by a solvent in which the supported catalyst is insoluble is the most common approach. However, quite large quantities of solvents are often necessary to achieve quantitative recovery, making the problem of metal leaching more severe. Moreover, reaction or byproducts might co-precipitate, thus contaminating the recovered catalyst and reducing product yields. The topic of catalysts being covalently attached to soluble supports has been reviewed in a number of excellent articles;1–5 this contribution will therefore concentrate on recent developments within the last five years, focusing especially on applications in asymmetric catalysis.

4.2 Soluble Supports – General Considerations Despite the obvious analogies between catalysts immobilized on soluble supports and their nonimmobilized versions, there are nevertheless a number of differences to be considered. Attachment of a catalyst to the support increases the overall molecular weight, thereby potentially reducing the catalyst concentration and thus prolonging the reaction time for a given reaction. This problem can become especially severe for supports with low loading capacity, as exemplified for the widely used polyethylene glycols (PEGs) such as MeOPEG5000 (Mr ¼ 5000 g/mol), which allows the attachment of only 1 mol of catalyst per 5 kg of polymer (approx. 0.2 mmol/g). To overcome these limitations, hyperbranched polymers, starpolymers or dendrimers have been developed, which offer up to a 30 times higher loading capacity than PEG supports. However, for higher loading of a support with a catalyst the resulting hybrid increasingly resembles the properties of the latter, which often results in drastically reduced overall solubility. To recover the catalyst from the reaction mixture or to separate it from the products, most commonly a phase switch has to be performed at the end of the reaction. There is a wide range of different solubility properties available, spanning such supports as polyacrylates, proteinaceous supports, or polyamines, which are soluble in water and can be efficiently used in biphasic systems, polyethylene glycols, which are soluble in water, chlorinated and aromatic solvents and insoluble in ether and nonpolar solvents, or noncross-linked polystyrene, which is soluble in most organic solvents but can be precipitated from water or methanol. The major difference between the last two supports is the drastically reduced solubility of PEGs at low temperatures but also the ability of the polyether chains to bind to metal centers, two complications that are especially relevant in metal catalysis. Charge transfer complexes have also recently been utilized for catalyst immobilization, allowing their recovery due to their low solubility in pentane in contrast to dichloromethane.6

Catalysts Bound to Soluble Polymers

79

Changing the temperature profile during a reaction has also been used as a recovery strategy, e.g. some perfluorinated solvents are miscible at elevated but not at ambient temperature with organic solvents. This allows a reaction to be carried out in a perfluorinated/organic solvent mixture in a homogeneous phase upon heating, while upon cooling a phase separation occurs. A catalyst tagged with a perfluoroalkyl moiety, which can also be considered as a support although it is generally not labeled as such, is retained in the perfluorinated solvent phase.7 In this context, insoluble supports that can release a catalyst upon raising the temperature, thus allowing the catalyzed reaction to be carried out in a homogeneous phase, have been described.8–10 ROMP polymers of functionalized norbornenes,11 displaying a similar solubility profile as noncross-linked polystyrene, have become popular for the synthesis of a large variety of functional reagents; however, this approach has been used in only a few examples for creating supported catalysts but is gaining rapidly in popularity.12 The high functional group tolerance of the ruthenium–carbene catalysts (especially Grubbs second and third generation) allow even the oligomerization of preformed metal catalysts being attached to a suitable alkene that can undergo ring opening or ring expanding metathesis.13 Catalyst

Ring expanding

Ring opening

metathesis Catalyst

metathesis

Catalyst n

Catalyst n

Scheme 4.1 Metathesis strategies for oligomer/polymer bound catalysts

Last but not least, soluble dendrimers as a globular type of support can be separated by ultra- and nanofiltration through membranes, making use of volume differences between the dendritic catalyst and reagents and products. Originally limited to reactions being carried out in water using biomembranes, there is now a good range of membranes commercially available that are also resistant to organic solvents.14

4.3 Recent Developments of Soluble Polymer-supported Catalysts 4.3.1 Attachment of Catalysts to Polymer Supports The great impact that ligation reactions have made in recent years for medicinal chemistry and materials, allowing the reliable connection between two functional groups irrespective of other functionalities present, has also been recognized for attaching catalysts to supports. Especially the copper-catalyzed azide alkyne cycloaddition (CuAAC) has become popular for this purpose, having been first utilized with the immobilization of aza(bisoxazoline) ligands15 to MeOPEG5000 in 2005.16 Subsequently, complexation with copper(II) chloride

80


resulted in the corresponding copper(II)–azabox complexes, which were shown to efficiently catalyze the asymmetric benzoylation of diols. A potential drawback of this strategy for asymmetric metal catalysis was recognized in that the triazole moiety is able to act as a coordinating moiety for metals as well,17–19 bearing the danger of creating catalytically active metal centers in an achiral environment, thus lowering the enantioselectivity of a given reaction. To overcome this limitation, dendrimers as globular supports in which the triazole moieties are not as readily accessible inside its core have been synthesized, giving rise to improved enantioselectivities for the title reaction (Scheme 4.2).19

Undesired Ligand

N N N

N3

Metal

Ligand

complexation

Cu (I)

N N N Metal

Ligand

desired

N N N Ligand-Metal Complex

Scheme 4.2 Copper-catalyzed [3 þ 2]-azide-alkyne cycloaddition (CuAAC) as a ligation strategy towards polymer-bound catalysts

More recently, it has been demonstrated that not only ligands but also preformed metal complexes are amenable to the CuAAC, thus allowing the direct attachment of metal catalysts to supports,20–23 in this way also circumventing the problem of unselective metal complexation with the polymer backbone or linker. Likewise, azidofunctionalized supports also offer the possibility of utilizing the Staudinger ligation for attaching catalysts bearing a carboxylic acid handle to a support. Surprisingly, this strategy has only been demonstrated recently with the immobilization of cobalt and vanadium Schiff base complexes onto polystyrene and polyethyleneglycol (Scheme 4.3).20,24 O HN Ligand-Metal Complex

HO2C

Ligand-Metal Complex

Ligand-Metal Complex

Staudinger Ligation

N3

N N N

CuAAC Ligand-Metal Complex

Scheme 4.3 Application of azido-functionalized polymers for the immobilization of metal catalysts


81

4.3.2 Polymer-bound Metal Catalysts – General Considerations Upon recovery of a polymer-bound catalyst, it is generally necessary to perform a number of extractions with organic and aqueous solvents in order to remove products and salts formed in the course of the reaction. Given the nature of a metal catalyst, i.e. containing relatively weak metal–ligand bonds, partial removal of the metal – known as metal leaching – during these washing cycles is likely to occur, causing contamination of the products but also seriously hampering its reuse due to the necessity of performing another complexation step by adding metal anew. Consequently, metal complexes containing strongly coordinating ligands, especially those containing metal–carbon bonds as found with N-heterocyclic carbene (NHC) or pincer ligands, have recently received much attention for immobilization. Strongly chelating ligands, notably salens and porphyrins are also particularly attractive for designing polymer-bound metal complexes. Likewise, considerable efforts have been made to immobilize bis(oxazolines) and binaphthyls as these are especially successful ligands for homogeneous asymmetric catalysis. Besides the possibility of recovery and reuse, the activity and selectivity of the polymerbound catalyst during the reactions is of great importance. In asymmetric catalysis, metal leaching is especially detrimental with respect to enantioselectivity, since metal centers that have dissociated from the chiral ligand might still be catalytically active, although without the possibility of inducing optical activity in the products. In general it has been recognized that enantioselectivity in polymer-bound catalysts is similar to that in their nonimmobilized counterparts provided metal leaching does not occur. With respect to activity, catalysts bound to soluble polymers more or less retain the activity of their nonbound counterparts, as long as the loading onto the support is sufficiently high to avoid limitations of mass transport by too low a catalyst concentration. Strikingly, although not understood in detail, there are an increasing number of reports claiming to achieve even higher activities with the polymer-bound catalysts than with the nonpolymer-bound ones. 4.3.3 Polymer-bound Organocatalysts – General Considerations Currently, there is a spectacular development of many new chemical transformations promoted by organocatalysts. Nevertheless, many of these reactions occur with low turnover rates and frequencies making the use of high catalyst loading necessary. Obviously, recycling of such organocatalysts and thus their immobilization onto polymers is therefore an attractive strategy. Moreover, since there is no metal the problem of leaching as discussed above is nonexistent, provided that a stable linkage to attach the catalyst to the polymer is chosen. Nevertheless, the rate decrease that is often encountered by immobilizing catalysts on polymers can be especially problematic for organocatalytic reactions that often proceed with turn-over frequencies of one cycle per hour or even less.

4.4 Recent Examples for Reactions Promoted by Catalysts Bound to Soluble Polymers 4.4.1 Achiral Catalysts Two trends can be currently recognized in the development of immobilized, achiral catalysts on soluble supports. On the one hand, precious metals such as rhodium,25

82


ruthenium and palladium (following two sections) are in the focus both in academic and in industrial research, given the great impact of hydroformylations and hydrogenations, olefin metathesis and cross-coupling reactions that these metals can promote. On the other hand, the emerging importance of organocatalysts has also made their immobilization attractive, since these catalysts are more stable – no leaching of metals as encountered with metal complexes – but generally require higher loading, thus making recycling and reuse feasible and desirable at the same time. Phase transfer catalysts and TEMPO (see later sections) will be discussed in this review, since these catalysts have been studied with a broad variety of supports to allow a meaningful comparison. Ruthenium Catalysts for Olefin Metathesis With the discovery of robust and functional group tolerant ruthenium–carbene catalysts– commonly referred to as Grubbs catalysts – olefin metathesis has conquered organic synthesis in every area. Nevertheless, the catalyst loading is not optimal for many applications, and moreover, rigorous removal of organoruthenium species from the metathesis products has been problematic in some cases. Not surprisingly therefore, Grubbs catalysts have been attached to a broad range of soluble supports (Table 4.1, Scheme 4.4). As the two common strategies for immobilization, either NHC ligands (modification of Grubbs 2nd generation, 12–13) or, more popularly, the 2-iso-propoxy O

PEG

O

O Cl

O Cl

Ru

Cl

Cl

L 1a: L = PCy3 1b: L = SiMes

O

3

O

Cl

Ru

Cl 2

SIMes 2

R

Ru L

PEG 3: L = PCy3, R = (CH2)2C8F17 4: L = SiMes, R = (C(Me2)CH2)18H

SIMes = Mes

N

N Mes

5: L = SiMes, R = OCO(CH2)4, Poly(2-oxazoline) 6: L = SiMes, R = OCO–F13Polyacrylate 7: L = SiMes, R = OCO(CH2)n–Au

PF4 IMZ =

N

8: L = PCy3, R = O(CH2)4–IMZ

Me N

9: L = PCy3, R = (CH2)3–IMZ 10: L = PCy3, R = (CH2)4–IMZ n

O O

n

O

HN

O

O

N

O Cl Cl

Cl

MeOPEG

Ru L

Cl

O

12 m

N Mes

N Cl

Ph

O

N Mes Cl Ru

Cl

Ru PCy3

11

PCy3

Ph m

n

Ru

Cl Mes N

N

13: n/m = 2:1 14

11: n/m = 39:1

Scheme 4.4 Immobilized ruthenium–carbene complexes for olefin metathesis

c

b

a

1a 1b 12 2 3 6 4 5 7 8 9 10 11 13 14

Catalyst

Ru-catalyst

5 5 5 5 5 0.5–1 5 2 5 5 2.5 1 1 3–5 5

Loading (mol%) 15a 15a 15b 15c 15a 15a,15d 15b,15e 15f 15a,15e 15d 15a,15d 15a 15a 15f 15a

15

Y

n

2(98–63) 5(100–85) nrb(40–82) 17(100–94) 7 (98–99)c 20 (498–94) 5 (84–99)c 5 (90–9) 7 (98–20) 10 (498–90) 10 (498–95) 10 (498–96) 8 (498–75%) nrb (95) 4 (498)

Cycles(conversion (%))

16a-f

X

A ¼ Precipitation; B ¼ Fluorous chromatography; C ¼ Fluorous extraction; D ¼ Gravity-based extraction; E ¼ Ionic liquid. Not reported. Isolated yield.

Homooligomer

ROMP

PEG PEG MeOPEG PEG Perfluoroalkyl Poly(fluoroalkyl acrylate) Isobutylene oligomer Poly(2-oxayoline) Gold Nanoparticles Ionic liquid

Support

n

15a: X = NTs: n = 1: Y = H 15b: X = C(CO2Et): n = 1, 2: Y = H 15c: X = NTs: n = 2: Y = Me 15d: X = NTs: n = 3: Y = H 15e: X = NTs: n = 2: Y = H, Me 15f: X = C(CO2Et): n = 1: Y = H

Y

X

Table 4.1 Application of immobilized ruthenium–carbene complexes to ring closing metathesis

A (ether) A (ether) A (ether) A (ether) B C D A (ether) A (methanol) E E E A (ether/hexane) — A (ethylacetate)

Recovery/Separationmethoda

27 27 28 29 30 31 32 33 34 35 36 37 38 12 39

Ref.

Catalysts Bound to Soluble Polymers 83

84


moiety (modification of Grubbs–Hoveyda catalysts, 1–11), were chosen as anchor points for the synthesis of polymeric or oligomeric ruthenium catalysts using a wide variety of supports. Combining both strategies, the homooligomer 14 from the catalyst monomer itself has been synthesized as well. Conceptually intriguing are the catalysts 11 and 13, in which the polymer-bound catalyst has been self-generated by ring opening metathesis from the corresponding monomers. As a benchmark, all catalysts were tested in ring closing metathesis reactions with 15, and for most cases high activity and good recyclability has been reported. However, it must be kept in mind that simple substrates such as 15 leading to cycloalkenes with no further substituents on the double bond have been proved to be exceptionally active for ring-closing metathesis, requiring only 0.1 mol% of catalyst or even less.26 Therefore, some of the reaction conditions under which the polymer-bound catalysts were tested (2.5–5 mol%) are not limiting enough to allow an accurate comparison with their nonpolymer bound counterparts. Moreover, data on loading of the catalyst to the support are often not provided, making it difficult to judge the effectiveness of the preparation of the supported catalysts. Besides the attractive feature of recyclability, attaching a catalyst to a support also nicely allows solubility properties to be imposed on the catalyst. Thus, olefin metathesis could be run in the whole range of solvents, spanning water (5, 12), ionic liquids (8–10), fluorous solvents (3, 6) and nonpolar solvents like heptane (4). Besides ruthenium–carbene complexes for metathesis reactions, the immobilization of ruthenium–porphyrin complexes on polyethyleneglycol or dendritic structures for the cyclopropanation, epoxidation and aziridination of alkenes has also been successfully developed.40,41 Palladium Catalysts for Cross Coupling and Related Reactions Besides the alkene metathesis being discussed in the previous section, cross-coupling reactions have shown great versatility in forming both carbon–carbon and carbon–heteroatom bonds.42 Palladium plays a prominent role as a catalyst, allowing the connection of sp, sp2 and even sp3 carbon centers over a wide and growing range of combinations. Not surprisingly, considerable efforts have been made to create recoverable catalysts; however, very significant advances have also been reported in the development of highly active nonimmobilized palladium catalysts, which are often required in only ‘homeopathic’ amounts, in Suzuki or Heck reactions.43 Consequently, a careful analysis regarding the efficiency and added value of immobilized palladium catalysts with respect to recyclability must be made in light of these results. A common complication in palladium(0)-catalyzed reactions is the dissociation of the metal from the ligand followed by the formation of colloidal palladium (‘palladium black’). Consequently, in order to arrive at immobilized palladium complexes that allow efficient recovery and reuse, a strong palladium–ligand connection is necessary. As exemplified with complexes 17–20, 22–23, ligands that are able to coordinate via carbon to palladium are most frequently employed for this purpose (Table 4.2, Scheme 4.5). Nevertheless, these complexes might simply serve as a reservoir, slowly releasing palladium in very small quantities, which is sufficient to effectively catalyze the given reaction. Therefore, one could argue that only unreacted catalyst might be recovered.

Catalysts Bound to Soluble Polymers N

PS

Pd P

O

O

O

O

Pd

N

P PS

PG

N

N

PEG350

S R1

N N

Br Pd

85

Br

R

Pd Cl

N N

S R1 N

17

N

PEG350

18

R1 =

19: R = H, PEG350OMe 20: R1 = Ph, R = PNODAM 21: (PIB-PPh 2)4Pd

PG = Polyglycerol CONH(CH2)3 PIB = H3C 20

1 N R

N R R

Pd

Me

I N3P3

I

N

O

C N N P O H S

N C H

2

Ph2P

N R2

24

6

PdCl2

22: R = Me, R1 = CH2(CH2)n Poly(2-oxazoline), R2 = Me 23: R = R1 = R2 = Bn

Scheme 4.5 Immobilized palladium catalysts for cross-coupling and related reactions

Phase Transfer Catalysts Phase transfer catalysts are generally designed to act as a mediator between an aqueous and an organic phase, especially aiming at transporting inorganic bases and nucleophiles into the latter. The good solubility properties of polyethyleneglycol in many organic solvents as well as in water therefore makes it a logical support for phase transfer catalysts, which has been especially explored for quaternary ammonium salts. Thus, C-, N-, and O-benzylations as well as cyclopropanations can be effectively carried out with 25 or 26 (Scheme 4.6), both being readily recoverable from diethylether (Table 4.3). Moreover, 25 was found to be applicable for the epimerization of a-chiral aldehydes as demonstrated in the synthesis of 28 (Scheme 4.7), which is an advanced precursor in the synthesis of thrombin and

NBu3Br

MeOPEG

O

O

O

O PEG

O

C8F17(CH2)3 NBu3Br

25

2

N

N O

O

BrBu3N

26

27

Scheme 4.6 Immobilized achiral phase transfer catalysts

(CH2)3C8F17

18 19 20 21 21 21 22 22 23 24

Hyperbranched PG MeOPEG350 PNODAM Polyisobutylene oligomer

2–0.002 0.2 0.1 1 1 1 0.67 0.1 – 5

2.5 2.5

Loading (mol%)

Suzuki Heck Heck Heck Sonogashira Trost–Tsuji Heck Suzuki Heck Stille

Heck Suzuki

Type of reaction

A (ether) A (ether) B C C D D D D – E A (degassed ether)

4(499–97.3)d 4 (49–78) Multiple times 3(63–99) 3(48–99) 4(81–100) (80–68)d nrc(95–30) 10 (95) 2(98–86)d

c

b

Recovery methodb

1(98–60) nrc (88–78)

Cycles (yield (%))a

Isolated yield. A ¼ Precipitation; B ¼ Dialysis; C ¼ Thermomorphic condition; D ¼ liquid/liquid separation; E ¼ Solvent resistant nanofiltration. Not reported. d Yield based on conversion.

a

Imidazolylidene Phosphorous Dendrimer

Poly(2-oxazoline)

17 17

Catalyst

PS PS

Support

Table 4.2 Application of immobilized palladium complexes in cross-coupling reactions

49 50 51 52

46 47 48

44 44 45

Ref.

86 Recoverable and Recyclable Catalysts


87

tryptase inhibitors. Notably, other bases such as tributylammonium bromide or aqueous dimethylamine under various solvent and temperature conditions failed for the latter transformation. TEMPO The oxidation of alcohols using catalytic amounts of the stable nitroxyl radical 2,2,6,6tetramethylpiperidine-1-oxyl (TEMPO) in combination with safe and easy to handle primary oxidants has been widely applied in academia and industry due to the low toxicity of the reagent and to the good turnover rates and activities as well as chemoselectivities achieved. Nevertheless, TEMPO is expensive, which has prompted quite a number approaches for its immobilization on soluble supports (Scheme 4.8, Table 4.4). Polyethylene glycols (35, 38) and ROMP polymers (39) have been successfully used as soluble supports for TEMPO, which showed high activity and good recyclability in reactions with NaOCl or oxygen as the terminal oxidant. A number of fluorous-tagged TEMPO catalysts such as 36 were initially developed by Pozzi and coworkers, which could be recovered either on fluorous silica or by extraction with a perfluorinated solvent. In contrast, fluoroustagged 37 bearing a polar triazole moiety, showed little affinity to perfluorinated phases but could be readily adsorbed on silica, thus allowing facile separation from reaction products by chromatography. Based on the catalytic system CuBr2/TEMPO/2,2-bipyridine for the aerobic oxidation of alcohols discovered by Sheldon et al. a multipolymer approach was investigated: By immobilizing individually both bipyridine and TEMPO on MeOPEG better water solubility of the copper complex and good recyclability by precipitation with diethylether was achieved, although substantial loss of catalytic activity was observed within a few cycles.57 Table 4.3 Application of immobilized phase transfer catalysts Support PEG

MeOPEG

Perfluoroalkyl a

Catalyst

Loading (mol%)

25 25 25 25

3 1 1 1

25 26 26 26 26

2.5 5 10 0.1 4

26

4

27

2

Type of reaction

Cycles (yield (%))a

Recovery methodb

C-alkylation O-alkylation N-alkylation Cyclopropanation (CCl2) epimerization C-alkylation O-alkylation N-alkylation Halide displacement Cyclopropanation (CCl2) Halide displacement

2(90–91) nrc(95) nrc(98) 4(98–95)

A A A A

(ether) (ether) (ether) (ether)

53 53 53 53

2(95–90) 2(67)d 2(91)d 3(99)d nrc(75)d

A A A A A

(ether) (ether) (ether) (ether) (ether)

54 55 55 55 55

1(95,93)

A (ether)

55

5 (88–86)d

B

56

Isolated yield. A ¼ Precipitation; B ¼ Fluorous solid phase extraction. Not reported. d Yield based on conversion. b c

Ref.

88

Recoverable and Recyclable Catalysts H H

H H PMPO

CHO

3

25 (2.5 mol%)

CHO

3

CH2Cl2, rt, 18 h

N O

PMPO

N O

PMP

PMP

95% one stereoisomer 28

29

Scheme 4.7 Epimerization of b-lactam aldehyde 28

Scheme 4.8 Immobilized TEMPO reagents

4.4.2 Chiral Catalysts Because of the higher costs that are generally encountered with chiral catalysts, their facile recovery and reuse is even more desirable than that of achiral ones. However, asymmetric catalyses are adding another level of complexity due to the aim of obtaining products with high enantioselectivity. With heterogeneous chiral catalysts, an erosion of enantioselectivity is often observed in comparison with their homogeneous counterparts. Reduced mass transport to the catalytic active centers on the heterogeneous support can be Table 4.4 Oxidation of alcohols to aldehydes and ketones using supported TEMPO in the presence of a stoichiometric oxidant Support MeOPEG Fluorous tag Fluorous tag Fluorous tag PEG ROMP a

Oxidant O2 O2 NaOCl NaOCl O2 NaOCl

Catalyst c

35 36c 36 37 38e 39

Loading (mol%)

Cycles (yield (%))a

Recovery methodb

Ref.

5–10 1–10 1–10 1 10 1

5 (99–80) 5 (99–54)d 3 (99–64) 4 (94–90)d 5 (nr)f 2 (nr)f

A (ether) B B C A (ether) A (ether)

58 59 60 61 62 63

Yield based on conversion. A ¼ Precipitation; B ¼ Fluorous biphasic extraction; C ¼ Silica. Co-catalyst (Co(NO3)2/Mn(NO3)2. d Isolated Yield. e CuBr or CuCl as co-catalyst. f Not reported. b c


89

responsible, especially, if a given reaction is also proceeding in the absence of the catalyst (background reaction). In such cases, soluble supports might offer a significant advantage. Immobilized catalysts might also undergo metal leaching taking away metal centers from the chiral ligand, consequently still catalyzing a given transformation but outside of the chiral environment, thus leading to reduced optical induction. Ligands that bind strongly to the metal can provide a remedy for leaching, but at the same time might reduce its catalytic activity, e.g. by decreasing its Lewis acidity or by blocking coordination sites that need to be accessible for the substrates. There is a plethora of chiral ligands available that have been explored as metal catalysts. Nevertheless, only a few ligand classes – termed privileged – are broadly applicable to catalytic asymmetric transformations. Among these are the bis(oxazoline), salen, binaphthyl, and chincona alkaloid ligands, which will be in the focus of this review. Similarly, the emerging field of organocatalysis currently produces a wealth of metal-free entities that are able to promote organic reactions. Nevertheless, also in this area some universally applicable catalysts are recognized, such as proline, imidazolidinones (MacMillan catalyst), N-heterocyclic carbenes, thioureas or chincona alkaloids. Metal Bis(oxazoline) Complexes Among the most successful chiral ligands are bis(oxazolines)64,65 that combine ease of synthesis, structural flexibility, and the ability for high asymmetric discrimination of substrates. Although the first reports of covalently attaching these ligands to polymeric supports appeared only in the year 2000,66,67 immobilized bis(oxazoline) ligands quickly became a widely studied topic, which was the subject of a review shortly afterwards.68 The logical point of attachment of a support to these ligands has been uniformly identified as the central bridge connecting the two oxazoline units, which is either suitably functionalized to allow ligation with an appropriately functionalized support or which is introduced already connected to the support in the course of the synthesis of the ligand. For the former strategy, propargylated aza(bisoxazoline) 47 seems to be especially versatile, since this precursor allows facile ligation to any azide-functionalized support by a copper-catalyzed alkyne–azide cycloaddition. As a third strategy, units such as styrenes have been incorporated in the bis(oxazoline) bridge which can be subsequently polymerized. Noncovalently bound supports also have been utilized to arrive at recyclable bis(oxazoline) ligands: the charge transfer complex Cu(I)46 proved to be readily soluble in dichloromethane even at low temperature, but could be recovered quantitatively by precipitation from pentane (Scheme 4.9). Quite a number of standard transformations have been tested with immobilized copper– and palladium–bis(oxazoline) complexes (Scheme 4.9, Table 4.5). Despite the homogeneous reaction conditions, reactions that require low temperatures such as the Mukaiyama aldol or Diels–Alder reactions generally proceeded with inferior results in comparison with their nonimmobilized counterparts. Small molecular weight supports, as found in 46, seem to be advantageous in this respect: Diels–Alder reactions between cyclopentadiene and a,b-unsaturated acyloxazolidinones proceeded with excellent results even at –50 C. Perfluorinated bis(oxazolines) 40–42 seemed to be problematic in copper-catalyzed processes, while in palladium-catalyzed allylic alkylations very high yields and selectivities could be obtained. The results obtained in the

90


Scheme 4.9 Bis(oxazoline) ligands immobilized on soluble supports

carbonyl–ene reaction between 2-phenylpropene and ethylglyoxylate with gold-nanoparticle-supported Cu43 are remarkable (cf. the same reaction with Cu41 under identical reaction conditions), showing the great potential of nanoparticles as supports for catalysts. Metal–salen Complexes Metal–salen complexes have found broad application in asymmetric catalysis. Moreover, the tetradentate ligand offers strong binding to metals, thus making the resulting complexes promising candidates for immobilization aiming at their facile recovery. Indeed, a broad variety of polymeric supports have been used for this purpose, uniformly using the para position of the phenol in the salicyl aldehyde unit as the point of attachment (Scheme 4.10). Moreover, the strong metal–ligand association allows first the formation of the metal complex followed by attachment of the complex to the support, thus avoiding the generally difficult step of postmodification of already supported ligands. Most widely, the kinetic resolution of epoxides using water as the nucleophile has been investigated with cobalt–salen complexes, and norbornene ROMP polymers, polyglycerol, or polystyrene (51, 53, 56; Scheme 4.10) have been utilized with good results (Table 4.6). Also, using amphiphilic polyamides, thus creating a hydrophobic core for the catalytic center and reagents and a hydrophilic shell to enhance water solubility (termed nanoreactors), has resulted in the highly active, selective and recyclable catalyst 57. A notable exception seems to be the perfluorinated salen complex 58, giving inferior results for the title reaction. Obviously, an electron withdrawing moiety is not well tolerated in the salen

2 10 5 5 2.5 5 10 5 0.7 nrc 5 10

Cu40

Cu41 Cu41 Cu42 Pd41 Pd42

Cu43

Cu44 Cu48

Cu45

Cu49

Cu46

Fluorous

Fluorous Fluorous Fluorous Fluorous Fluorous

Gold Nanoparticle

MeOPEG MeOPEG

Dendrimer

Dendrimer

Charge Transfer complex

Diels–Alder

benzoylation

Mukaiyama–aldol

Mukaiyama–aldol benzoylation

cyclopropanation (styrene) carbonyl–ene-reaction allylic oxidation allylic oxidation allylic alkylation allylic alkylation carbonyl–ene- reaction

Type of reaction

c

b

Isolated yield, ee determined by either chiral HPLC or chrial GC. A ¼ precipitation; B ¼ dilution with hexane and centrifugation; C ¼ liquid/liquid extraction. Not reported. d Determined by GC.

a

Loading (mol%)

Catalyst

Support

Table 4.5 Application of immobilized bis(oxazoline) ligands

11(91–95, de 93–97, endo 84–94)

2(78–40, syn/anti 2:1, syn 64–25) 3(35–43, 99)

2(40–38, 50– 45) 5(37–49, 99–91)d

4(99–80, 86–84)

nrc(32, 5) 1(49–43, 50) 1(67–510, 61–nrc) 1(98d, 92) 2(100–98, 94–90)

4(79–68; E/Z ¼ 2:1 E: 80–70)

Cycles (yield, ee (%))a

A (pentane)

A (heptane)

A (methanol)

A (ether) A (ether)

B

— A (hexane) C C C

A (hexane)

Recovery methodb

6,75

19

74

73 16

72

70 70 71 71 71

69

Ref.


92


Scheme 4.10 Immobilized metal–salen complexes

moiety. Self-supported oligomeric salen complexes 59–60 performed impressively well with respect to selectivity and turn over cycles, although the possibility of recovery has not been tested with these compounds. Metal–Binaphthyl Complexes The binaphthyl moiety has been recognized as one of the most effective chiral inducers for catalytic asymmetric transformations. However, its synthetic accessibility in enantiopure form is more difficult compared withthe previously discussed bis(oxazoline) and salen ligands that are available from the chiral pool. This line of argument is even more true for suitably substituted derivatives needed for anchoring a support. Nevertheless, immobilization has been successful using various positions within the binaphthyl framework, but a few different approaches as exemplified with 65 have also been successfully explored (Table 4.7, Scheme 4.11). Globular supports, i.e. Frechet type dendrimers, and gold nanoparticles, were employed for titanium-catalyzed dialkyl zinc additions to aldehydes. Good enantioselectivities, even slightly higher than those with the nonimmobilized analogs were observed with Ti61–63. Similar results were obtained with the Ru–BINAP analog Ru64 for asymmetric hydrogenations. Very good results were also obtained with Rh65, which is conceptually

50 51 52 53 56 57 58 59 60 55 54

Catalyst 4 0.2 4 0.5–0.05 0.5 0.02–0.06 0.1 0.01–0.0003 0.01–0.1 10 0.1

Loading (mol%)

b

Yield based on conversion. A ¼ precipitation; B ¼ fluorous biphasic; C ¼ dialysis; D ¼ extraction. c Not reported.

a

ROMP ROMP Polyglycerol Polyglcerol Polystyrene Amphiphilic polyamide Fluorous Oligomer Oligomer MeOPEG MeOPEG

Support

Table 4.6 Application of immobilized salen complexes

epoxidation epoxide opening epoxidation epoxide opening epoxide opening epoxide opening epoxide opening epoxide opening epoxide opening diethyl zinc addition cyanosilylation

Type of reaction 3(100–85, 81–6) 3(57–5, 99–79) 4(98–75, 95–88) nrc(50–53, 92–85) 4(54–53, 99–498) 4(98–88) 3(37–18, 60–22) nrc(44–40, 499) nrc(48–42, 498) 3(91, 82) 4(95, 86–85)


A (ether) A (ether) A(hexane) — A(ether) D B — — A(ether) D

Recovery methodb

76 76 77 77 78 79 80 81,82 13 83 84

Ref.


94


Table 4.7 Application of immobilized binaphthyl ligands Support Dendrimer Dendrimer Gold nanoparticle Dendrimer MeOPEG

Catalyst

Loading (mol%)

Ti61 Ti62 Ti63 Ru64 Rh65

20 20 10 1 1


Recovery methodb

3(95–96, 87–89) 1(499c,86) 1(98–92,86–68) 3(100,92–91) 4(100,97–91)

A(methanol) A(methanol) A(ethanol) A(methanol) A(ether)

Ref. 85 86 87,88 89 90

a

Isolated yield based on conversion, ee determined by either chiral HPLC or chiral GC. A ¼ precipitation. c Yield based on conversion. b

O

R2

dendrimer O OH O

OH

NH R3

O

R1 PPh2

61: R2 = R3 = H; R1 = CH2O- dendrimer

PPh2

62: R1 = R2 = H; R3 = CH2 O- dendrimer

P O

MeOPEG

O R1

65

63: R1 = R3 = H; R2 = (CH2)nS- Au 64

NH m

Scheme 4.11 Immobilized binaphthyl ligands

intriguing since the parent 2,20 -BINOL, arguably the most easily accessible binaphthyl ligand, could be directly employed. Overall, the advantages of these catalysts were seen in their micellar structure, conferring good solubility properties and high activity. Organocatalysts During recent years spectacular progress has been made in the development of asymmetric transformations that are promoted by organocatalysts, especially for carbon–carbon bond forming reactions. Generally high enantioselectivities in a number of processes such as various types of aldol reactions, Mannich reactions, conjugate additions to alkenes, or cycloadditions have been achieved, rivaling or exceeding the performance of metal catalysts. However, reaction rates are far from optimal, requiring long reaction times and making high catalyst loading necessary. Immobilization of chiral organocatalysts aiming at their recovery and reuse is therefore an attractive strategy, especially since the problem of metal leaching as encountered frequently in immobilized metal complexes naturally does not pose a problem. Soluble supports play a prominent role for the development of supported organocatalysts in order not to reduce reaction rates further by heterogenization.

Catalysts Bound to Soluble Polymers O

R

O O

MeOPEG O

67: R =

HN

O O

N H

66

COOH

N H

COOH

O

68: R =

O

N H

O

O

OBn

=

N H HO

NH O

O

O Bn O Bn

O n=1

69

Polystyrene

O

O H N S

O MeOPEG O

O

N H

O

H N

S

95

70

n n=2 Dendrimer

OBn

Scheme 4.12 Immobilized proline-based catalysts

Proline as the prototype of an organocatalyst has been most widely studied for immobilization (Scheme 4.12), especially since cis- or trans-3-hydroxyproline derivatives are readily available which can be attached to a support via their hydroxy functionality. However, the carboxyl group of proline has also been used as an anchor point for immobilization as shown with 69, being one of the most effective immobilized organocatalysts for the aldol reaction between aromatic aldehydes and cyclic ketones described to date. Although in many cases at least initially similar yields and selectivities are achieved with the supported organocatalysts in comparison with the nonsupported analogs, often a rapid drop in the performance of the catalysts is observed upon its reuse (Table 4.8). Table 4.8 Application of immobilized proline-based catalysts Catalyst

Loading (mol%)

Type of reaction

Cycles (yield (%))a % ee

MeOPEG

66

30

Aldol

66

30

Mannich

MeOPEG

66

15

Enolate addition to nitroalkenes

4(68–51) 77–75 3(81–64) 96–97 4(60–18)

MeOPEG

92

MeOPEG

67

5

Enolate addition to nitroalkenes

35–20 4(94–24)

93

Polystyrene

68

5

Aldol

Dendrimer

69

10

Aldol

Dendrimer

70

30

Epoxidation of enones

Support

60 –510 5(70–59) 95–89 4(99–97)b 499 5(84–80) 74–72

a b

Isolated yield; catalyst recovered by precipitation. Catalyst precipitated from hexane/ethylacetate.

Ref. 91 91

94 95 96

96


This might be an indication that although complete recovery of catalysts based on mass recovery is generally claimed, the catalysts might undergo side reactions, which could explain their loss of activity. Considering the ketone enolate addition to nitroalkenes as an example in which both 66 and 67 showed considerably worse results after the first cycle, with related amine catalysts an irreversible formation of pyrrole between the catalyst, ketone and nitroketone was recently identified which accounts for the catalyst deactivation.97 It might be worthwhile to study more closely the fate of organocatalysts in given transformations, in particular when aiming at recyclable variants. Chincona alkaloids are gaining considerable attention as organocatalysts due to their multifunctionality. Their application as chiral phase transfer catalysts calls for quaternization of their nitrogen functionality, a step that can be combined with the introduction of a support, as demonstrated with catalysts 71 and 72 (Table 4.9, Scheme 4.13). Alternatively, the free hydroxyl group can be used for this purpose as seen in 73. The basic conditions that are employed in C-alkylation reactions require especially stable linkages, and in this respect 71 being connected to the polyethyleneglycol through amide bonds seem to be advantageous in the series of cinchona alkaloids that have been tested. Alternatively, the fluorous version of the binaphthylmodified tetrabenzylammonium bromide 74 also proved to be a highly efficient catalyst. Chiral imidazolidinones and thioureas are yet two more classes of organocatalysts that have proven to be most versatile for a variety of asymmetric processes (Table 4.10, Scheme 4.14). Nevertheless, immobilized variants of these catalysts are scarce; moreover, the examples reported so far seem to have room for improvement compared with their nonimmobilized analogs. Finally, chiral nucleophilic carbenes are efficient catalysts for Umpolung processes. First approaches to immobilizing such catalysts – so far achiral variants – on polyethyleneglycol have also been successful.102

Table 4.9 Application of immobilized chiral phase transfer catalysts Catalyst

Loading (mol%)

Type of reaction

Cycles (yield (%))a

PEG

71

10

C-Alkylation of Schiff bases

4(98–88)

98

MeOPEG

72

10


83–80 % ee 2(83–69)

99

MeOPEG

73

10


73–70% ee 2(75–11)

100

Fluorous

74

3


64–30 % ee 3(82–79) b

101

Support

90–92 % ee a b

Isolated yield; catalyst recovered by precipitation from ether. Catalyst recovered by FC-72 extraction.

Ref.


97

Cl N N H

R

Br

O

HO

N PEG NH H O

N

N

O MeOPEG

O OH

71

N

R

H

H

Cl

72

N

R = H or OMe

Si

Si Cl

Br N

N

O MeOPEG O

Si

Si

or Br

O

3

H Si

Si Si

Si

N

Si = SiMe2(CH2CH2C8F17)

73

74

Scheme 4.13 Immobilized chiral phase transfer catalysts Table 4.10 Application of immobilized imdazolidinone and thiourea catalysts Type of reaction

Cycles (yield (%))a % ee

Ref.

10

Diels–Alder reaction cyclohexadiene þ acrolein

4(67–38)

103

75

20

1,3-Dipolar CA of nitrone þ acrolein

92–85 3(59–26)

104

76

10

Conjugate addition of malonates to nitroalkenes

88–87 2(71–74)

105

Support

Catalyst

Loading (mol%)

MeOPEG

75

MeOPEG PEG

86–90 a

Isolated yield; catalyst recovered by precipitation from ether.

CF3 O

O Ar MeOPEG O Ar

S

3

HN

75 (Ar = p-C5H4)

N nBu

N NMe2 H

N H

O O

76

Scheme 4.14 Imidazolidinone- and thiourea-immobilized catalysts

PEG 2

98


4.5 Conclusion There are a number of reliable synthetic approaches available to immobilize catalysts on insoluble or, as covered in this review, on soluble supports. These strategies have been applied to just about every major catalyst and ligand class, and good results are generally obtained. However, the picture conveyed in the literature is that complete recovery and reuse of immobilized catalysts without loss of activity is possible in many cases, but such claims do not always hold up to close scrutiny. A high initial catalyst loading and few recovery cycles are often not testing vigorously the scope and limitation of the immobilized catalysts. While high selectivities, at least initially comparable with those of the nonimmobilized analogs, can be achieved, the biggest challenge appears to be to obtain and maintain high reaction rates. For the latter, it should be beneficial to investigate in more detail pathways of catalyst deactivation, since only entities with long lifetimes will be ultimately successful as highly recyclable catalysts.

List of Abbreviations AzaBOX BOX BINOL BINAP CuAAC MeOPEG PEG PG PIB PNODAM PS PTC ROMP TEMPO TOFs TON

Aza(bisoxazoline) Bis(oxazoline) 1,10 -Bi-2-naphthol 2,20 -Bis(diphenylphosphino)-1,10 -binaphthyl Copper-catalyzed azide alkyne cycloaddition Methoxy polyethylene glycol Polyethylene glycol Polyglycerol Polyisobutylene Poly(N-octadecylacrylamide) Polystyrene Phase transfer catalyst Ring opening metathesis polymerization 2,2,6,6-Tetramethyl- piperidin-1-oxyl Turnover frequencies Turnover number

References 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11.

D. Bergbreiter, Chem. Rev. 2002, 102, 3345. T. J. Dickerson, N. N. Reed, K. D. Janda, Chem. Rev. 2002, 102, 3325. J. i. Yoshida, K. Itami, Chem. Rev. 2002, 102, 3693. D. E. Bergbreiter, J. Li, Top. Curr. Chem. 2004, 242, 113. R. Haag, S. Roller, Top. Curr. Chem. 2004, 242, 1. G. Chollet, F. Rodriguez, E. Schulz, Org. Lett. 2006, 8, 539. J. A. Gladysz, D. P. Curran, Tetrahedron 2002, 58, 3823. L. V. Dinh, J. A. Gladysz, Angew. Chem., Int. Ed. 2005, 44, 4095. M. Wende, R. Meier, J. A. Gladysz, J. Am. Chem. Soc. 2001, 123, 11490. M. Wende, J. A. Gladysz, J. Am. Chem. Soc. 2003, 125, 5861. A. G. M. Barrett, B. T. Hopkins, J. Kobberling, Chem. Rev. 2002, 102, 3301.


99

12. W. J. Sommer, M. Weck, Adv. Synth. Catal. 2006, 348, 2101. 13. X. Zheng, C. W. Jones, M. Weck, J. Am. Chem. Soc. 2007, 129, 1105. 14. R. van Heerbeek, P. C. J. Kamer, P. W. N. M. van Leeuwen, J.N.H. Reek, Chem. Rev. 2002, 102, 3717. 15. H. Werner, R. Vicha, A. Gissibl, O. Reiser, J. Org. Chem. 2003, 68, 10166. 16. A. Gissibl, M. G. Finn, O. Reiser, Org. Lett. 2005, 7, 2325. 17. R. Rasappan, M. Hager, A. Gissibl, O. Reiser, Org Lett 2006, 8, 6099. 18. A. Bastero, D. Font, M. A. Pericas, J. Org. Chem. 2007, 72, 2460. 19. A. Gissibl, C. Padie, M. Hager, F. Jaroschik, R. Rasappan, E. Cuevas-Yanez, C. O. Turrin, A. M. Caminade, J. P. Majoral, O. Reiser, Org. Lett. 2007, 9, 2895. 20. S. Jain, O. Reiser, ChemSusChem 2008, 1, 534. 21. X.-Y. Wang, A. Kimyonok, M. Weck, Chem. Commun. 2006, 3933. 22. D. G. Cabrera, B. D. Koivisto, D. A. Leigh, Chem. Commun. 2007, 4218. 23. D.-M. Shen, C. Liu, Q.-Y. Chen, Eur. J. Org. Chem. 2007, 1419. 24. S. L. Jain, B. Sain, Adv. Synth. Catal. 2008, 350, 1479. 25. M. T. Zarka, M. Bortenschlager, K. Wurst, O. Nuyken, R. Weberskirch, Organometallics 2004, 23, 4817. 26. S. Maechling, M. Zaja, S. Blechert, Adv. Synth. Catal. 2005, 347, 1413. 27. S. Varray, R. Lazaro, J. Martinez, F. Lamaty, Organometallics 2003, 22, 2426. 28. J. P. Gallivan, J. P. Jordan, R. H. Grubbs, Tetrahedron Lett. 2005, 46, 2577. 29. Q. Yao, A. Rodriguez Motta, Tetrahedron Lett. 2004, 45, 2447. 30. M. Matsugi, D. P. Curran, J. Org. Chem. 2005, 70, 1636. 31. Q. Yao, Y. Zhang, J. Am. Chem. Soc. 2004, 126, 74. 32. C. Hongfa, J. Tian, H. S. Bazzi, D. E. Bergbreiter, Org. Lett. 2007, 9, 3259. 33. M. T. Zarka, O. Nuyken, R. Weberskirch, Macromol. Rapid Commun. 2004, 25, 858. 34. B. S. Lee, S. K. Namgoong, S.-g. Lee, Tetrahedron Lett. 2005, 46, 4501. 35. Q. Yao, Y. Zhang, Angew. Chem. Int. Ed. 2003, 42, 3395. 36. Q. Yao, M. Sheets, J. Organomet. Chem. 2005, 690, 3577. 37. N. Audic, H. Clavier, M. Mauduit, J. C. Guillemin, J. Am. Chem. Soc. 2003, 125, 9248. 38. S. J. Connon, A. M. Dunne, S. Blechert, Angew. Chem. Int. Ed. 2002, 41, 3835. 39. S.-W. Chen, J. H. Kim, H. Shin, S.-g. Lee, Org. Bio. Chem. 2008, 6, 2676. 40. J. L. Zhang, C. M. Che, Org. Lett. 2002, 4, 1911. 41. J.-L. Zhang, H.-B. Zhou, J.-S. Huang, C.-M. Che, Chem. Eur. J. 2002, 8, 1554. 42. in Metal-Catalyzed Cross-Coupling Reactions, 2 ed., (Eds.: A. de Meijere, F. Diederich), WileyVCH, Weinheim, 2004, p. 890. 43. A. Alimardanov, L. Schmieder-van de Vondervoort, A. H. M. de Vries, J. G. de Vries, Adv. Synth. Catal. 2004, 346, 1812. 44. F.-T. Luo, C. Xue, S.-L. Ko, Y.-D. Shao, C.-J. Wu, Y.-M. Kuo, Tetrahedron 2005, 61, 6040. 45. M. Meise, R. Haag, ChemSusChem 2008, 1, 637. 46. D. E. Bergbreiter, S. Furyk, Green Chemistry 2004, 6, 280. 47. D. E. Bergbreiter, P. L. Osburn, J. D. Frels, Adv. Synth. Catal. 2005, 347, 172. 48. D. E. Bergbreiter, J. Li, Chem. Commun. 2004, 42. 49. D. Schonfelder, O. Nuyken, R. Weberskirch, J. Organomet. Chem. 2005, 690, 4648. 50. D. Schonfelder, K. Fischer, M. Schmidt, O. Nuyken, R.-. Weberskirch, Macromolecules 2005, 38, 254. 51. D. Nair, J. T. Scarpello, I. F. J. Vankelecom, L. M. F. D. Santos, L. S. White, R. J. Kloetzing, T. Welton, A. G. Livingston, Green Chemistry 2002, 4, 319. 52. M. Koprowski, R. M. Sebastian, V. Maraval, M. Zablocka, V. Cadierno, B. Donnadieu, A. Igau, A. M. Caminade, J. P. Majoral, Organometallics 2002, 21, 4680. 53. M. Benaglia, M. Cinquini, F. Cozzi, G. Tocco, Tetrahedron Lett. 2002, 43, 3391. 54. R. Annunziata, M. Benaglia, M. Cinquini, F. Cozzi, F. Maggioni, A. Puglisi, J. Org. Chem. 2003, 68, 2952. 55. D. Albanese, M. Benaglia, D. Landini, A. Maia, V. Lupi, M. Penso, Ind. Eng. Chem. Res. 2002, 41, 4928. 56. A. M. Stuart, J. A. Vidal, J. Org. Chem. 2007, 72, 3735. 57. C. W. Y. Chung, P. H. Toy, J. Comb. Chem. 2007, 9, 115.

100


58. M. Benaglia, A. Puglisi, O. Holczknecht, S. Quici, G. Pozzi, Tetrahedron 2005, 61, 12058. 59. O. Holczknecht, G. Pozzi, QSAR & Combinatorial Science 2006, 25, 736. 60. O. Holczknecht, M. Cavazzini, S. Quici, I. Shepperson, G. Pozzi, Adv. Synth. Catal. 2005, 347, 677. 61. A. Gheorghe, E. Cuevas-Yanez, J. Horn, W. Bannwarth, B. Narsaiah, O. Reiser, Synlett 2006, 2767. 62. P. Ferreira, E. Phillips, D. Rippon, S. C. Tsang, Appl. Catal. B 2005, 61, 206. 63. C. Tanyeli, A. Gumus, Tetrahedron Lett. 2003, 44, 1639. 64. H. A. McManus, P. J. Guiry, Chem. Rev. 2004, 104, 4151. 65. G. Desimoni, G. Faita, K. A. Jorgensen, Chem. Rev. 2006, 106, 3561. 66. M. Glos, O. Reiser, Org. Lett. 2000, 2, 2045. 67. M. I. Burguete, J. M. Fraile, J. I. Garcia, E. Garcia-Verdugo, S. V. Luis, J. A. Mayoral, Org. Lett. 2000, 2, 3905. 68. D. Rechavi, M. Lemaire, Chem. Rev. 2002, 102, 3467. 69. J. Bayardon, O. Holczknecht, G. Pozzi, D. Sinou, Tetrahedron: Asymmetry 2006, 17, 1568. 70. J. Bayardon, D. Sinou, Tetrahedron: Asymmetry 2005, 16, 2965. 71. J. Bayardon, D. Sinou, J. Org. Chem. 2004, 69, 3121. 72. F. Ono, S. Kanemasa, J. Tanaka, Tetrahedron Lett. 2005, 46, 7623. 73. M. Benaglia, M. Cinquini, F. Cozzi, G. Celentano, Org. Bio. Chem. 2004, 2, 3401. 74. B.-Y. Yang, X.-M. Chen, G.-J. Deng, Y.-L. Zhang, Q.-H. Fan, Tetrahedron Lett. 2003, 44, 3535. 75. G. Chollet, F. Rodriguez, E. Schulz, Chem. Eur. J. 2007, 13, 992. 76. M. Holbach, M. Weck, J. Org. Chem. 2006, 71, 1825. 77. M. Beigi, S. Roller, R. Haag, A. Liese, Eur. J. Org. Chem. 2008, 2008, 2135. 78. X. Zheng, C. W. Jones, M. Weck, Chem. Eur. J. 2006, 12, 576. 79. B. M. Rossbach, K. Leopold, R. Weberskirch, Angew. Chem. Int. Ed. 2006, 45, 1309. 80. I. Shepperson, M. Cavazzini, G. Pozzi, S. Quici, J. Fluorine Chem. 2004, 125, 175. 81. D. E. White, E. N. Jacobsen, Tetrahedron: Asymmetry 2003, 14, 3633. 82. J. M. Ready, E. N. Jacobsen, Angew. Chem. Int. Ed. 2002, 41, 1374. 83. U. K. Anyanwu, D. Venkataraman, Tetrahedron Lett. 2003, 44, 6445. 84. U. K. Anyanwu, D. Venkataraman, Green Chemistry 2005, 7, 424. 85. L. Yin, R. Li, F. Wang, H. Wang, Y. Zheng, C. Wang, J. Ma, Tetrahedron: Asymmetry 2007, 18, 1383. 86. G.-H. Liu, W.-J. Tang, Q.-H. Fan, Tetrahedron 2003, 59, 8603. 87. S. Takizawa, M. L. Patil, K. Marubayashi, H. Sasai, Tetrahedron 2007, 63, 6512. 88. S. Takizawa, M. L. Patil, F. Yonezawa, K. Marubayashi, H. Tanaka, T. Kawai, H. Sasai, Tetrahedron Lett. 2005, 46, 1193. 89. G.-J.Deng,B.Yi,Y.-Y.Huang, W.-J.Tang,Y.-M.He,Q.-H.Fan,Adv.Synth.Catal.2004,346,1440. 90. X.-P. Hu, J.-D. Huang, Q.-H. Zeng, Z. Zheng, Chem. Commun. 2006, 293. 91. M. Benaglia, M. Cinquini, F. Cozzi, A. Puglisi, G. Celentano, Adv. Synth. Catal. 2002, 344, 533. 92. M. Benaglia, M. Cinquini, F. Cozzi, A. Puglisi, G. Celentano, J. Mol. Catal. A: Chem. 2003, 204–205, 157. 93. L. Gu, Y. Wu, Y. Zhang, G. Zhao, J. Mol. Catal. A: Chem. 2007, 263, 186. 94. Y.-X. Liu, Y.-N. Sun, H.-H. Tan, W. Liu, J.-C. Tao, Tetrahedron: Asymmetry 2007, 18, 2649. 95. Y. Wu, Y. Zhang, M. Yu, G. Zhao, S. Wang, Org. Lett. 2006, 8, 4417. 96. X. Liu, Y. Li, G. Wang, Z. Chai, Y. Wu, G. Zhao, Tetrahedron: Asymmetry 2006, 17, 750. 97. R. Rasappan, O. Reiser, Eur. J. Org. Chem. 2009, 1305. 98. X. Wang, L. Yin, T. Yang, Y. Wang, Tetrahedron: Asymmetry 2007, 18, 108. 99. D. Y. Kim, S. C. Huh, Bull. Korean Chem. Soc. 2004, 25, 347. 100. T. Danelli, R. Annunziata, M. Benaglia, M. Cinquini, F. Cozzi, G. Tocco, Tetrahedron: Asymmetry 2003, 14, 461. 101. S. Shirakawa, Y. Tanaka, K. Maruoka, Org. Lett. 2004, 6, 1429. 102. K. Zeitler, I. Mager, Adv. Synth. Catal. 2007, 349, 1851. 103. M. Benaglia, G. Celentano, M. Cinquini, A. Puglisi, F. Cozzi, Adv. Synth. Catal. 2002, 344, 149. 104. A. Puglisi, M. Benaglia, M. Cinquini, F. Cozzi, G. Celentano, Eur. J. Org. Chem. 2004, 2004, 567. 105. H. Miyabe, S. Tuchida, M. Yamauchi, Y. Takemoto, Synlett 2006, 3295.

5 Polymeric, Recoverable Catalytic Systems Qiao-Sheng Hu Department of Chemistry, College of Staten Island and the Graduate Center of the City University of New York, Staten Island, NY, USA

5.1 Introduction The recovery and reuse of catalyst systems, which are often expensive and/or difficult to obtain, have been of great research interest over the past decades.1,2 One approach to recover/reuse catalyst systems has been the immobilization of catalysts on polymer supports, such as polystyrenes. Such polymer-supported strategy offers advantages such as easy recovery by simple filtration and washing. However, because the environment of the catalytically active sites in polymer-supported catalyst systems is often altered from their monomeric counterparts and polymer-supported catalysts are often insoluble in the reaction media, such strategy often suffers from significant catalytic activity and/or enantioselectivity drop. One way to circumvent such drawbacks would be the development of polymeric catalyst systems through the polymerization of ligand-containing monomers, in which the environment of catalytically active sites might be preserved. The solubility of polymers produced by such strategy could also be tuned. Due to the size and solubility difference between the polymers and monomers, using polymeric chiral catalysts can simplify the reaction work-up, allow the recovery and reuse of the catalysts by simple filtration or precipitation, and may also make it possible to carry out reactions in a continuous mode. Polymeric catalysts generated by the polymerization of ligand-containing monomers can be generally divided into two types: linear polymeric chiral catalysts and dendritic chiral


102


catalysts. A large number of linear polymeric chiral catalysts have been developed and some of them exhibited enantioselectivities parallel to that of monomeric chiral catalysts. 3,4 Dendritic catalysts with catalytically active sites could be at the core, branch or periphery of a dendrimer with high solubility. Dendritic catalysts with catalytically active sites at the periphery of the dendrimer can have many catalytic sites around the dendrimer.5 They could potentially be very efficient polymeric catalysts. Many periphery-modified dendritic catalysts developed to date showed lower enantioselectivity than their corresponding monomeric catalysts. It is speculated that the lower enantioselectivity results from the highly condensed packing in the periphery of the dendrimers which causes a chiral environment different from that of the monomeric catalysts. Dendrimers with less condensed packing in the periphery have been developed and used for asymmetric catalysis. Results showed that they can mostly maintain the enantioselectivity of the monomeric catalysts. In this chapter, several polymeric catalyst systems, which were prepared through the polymerization of ligand-containing monomers, are summarized.

5.2 Polymeric Catalyst Systems 5.2.1 1,10 -Bi-2-naphthol (BINOL)-based Polymeric Catalytic Systems 1,10 -Bi-2-naphthol (BINOL, 1) and its derivatives have been established as one of the most extensively studied and used ligands.6 Incorporating this chiral unit into polymeric catalytic systems has been a subject of intense study. Optically active BINOL-containing polymers, which incorporate chiral 1,10 -binaphthol units into the polymer backbone and possess stable main chain chirality, have been prepared and studied (Chart 5.1, Schemes 5.1–5.5). 7–13 Studies of this new class of optically active polymers showed that properties of 3,30 -linked binaphthyl-based optically active polymers are very different from those of 6,60 -linked polymers. For example, while 3,30 -linked polymers such as 4 are excellent polymeric chiral catalysts for asymmetric diethylzinc addition with aldehydes and asymmetric hetero Diels–Alder reactions, 6,60 -linked polymers such as 2 only yield product with less than 40% ee.9,11 Comparison of the catalytic property of monomeric compound 5 with polymer 4 for the addition of diethylzinc with aldehydes revealed that 5 was a superior catalyst to 4,9 which suggested that the environment of the catalytically active sites in 5 was not completely preserved in 4. It was thus envisioned that the catalytic properties of a monomeric ligand could be maintained if they were incorporated in a sterically regular polymer backbone through a rigid linker. Based on this hypothesis, polymer catalyst 6 was designed and synthesized.10 It was shown that 6 was an excellent catalyst, with catalytic activity and enantioselectivity comparable with compound 5. It was also recoverable and reusable. BINOL-containing polymers 7 and 8 have also been prepared through the imine linkage by Chan and coworkers.14 They have been employed as ligands for Ti(Oi-Pr)4catalyzed diethylzinc addition with aldehydes. Up to 80% e.e. was observed. The polymers could be recovered and reused with similar catalytic properties to those of the original polymers.

Polymeric, Recoverable Catalytic Systems HO

OH

RO

OR

OR

n HO OR

RO

RO

OH

OH OH

OH

n

OH

RO

OR

n = 0, 2

R=C6H13

4

n = 1, R=C6H13, 3

1 HO

103

OH

OR RO RO OH

HO OR

OH RO

HO

OR

OH

OR OR

OH OR

R=C6H13

RO OR OH

RO

R=C6H13 OR

RO OH

6

OR

OR

5

Chart 5.1

5.2.2 Bisphosphine-containing Polymeric Catalyst Systems Bisphosphines such as 1,n-diphenylphosphinoalkanes (DPPA) and derivatives, 2,20 bis(diphenylphosphino)-1,10 -binaphthyl (BINAP), 1,10 -bis(diphenylphosphino)ferrocene (DPPF), and 2,20 -bis(diphenylphosphino)-1,10 -biphenyl (BIPHEP) represent a large group of bidentate ligands in transition metal catalysis.2,15 They have been demonstrated as very useful ligands in an array of bond forming reactions including catalytic hydrogenations and OH

Br OAc

1). Ni(1, 5-COD)2 1, 5-COD, Bipyridine, DMF

OAc

2). KOH, THF, H2O

OH

OH

OH

Br

2 OH

OH

OH

Scheme 5.1 Synthesis of BINOL polymer 2

OH

104

Recoverable and Recyclable Catalysts HO

Br

OH

OR OMOM OMOM

B(OH)2 1. Pd(PPh3)4 1M K2CO3, THF

+

RO

OR

2. HCl

(HO)2B OR

Br

OR

RO R=C6H13

3 HO

OH

Scheme 5.2 Synthesis of BINOL-containing polymer 3

I

OR

OMOM OMOM

OR

B(OH)2 1. Pd(PPh3)4 1M K2CO3, THF

+ (HO)2B

HO

OH

2. HCl

OH

OH

OH

RO

OR

I

RO

OH

OR R=C6H13

4

Scheme 5.3 Synthesis of BINOL-containing polymer 4 I OR

OR OMOM OMOM

+

B(OH)2

(HO)2B

1. Pd(PPh3)4, 1 M K2CO3, THF 2. HCl

RO

RO I

RO

RO HO

OH RO

HO

OR

OR

OR

RO

RO RO OR

OR OH

R=C6H13

OH OR

Scheme 5.4 Synthesis of BINOL-containing polymer 6

6

OH

Polymeric, Recoverable Catalytic Systems

105

NH2 CHO

OHC

H2N

+

HO

OH

HO

H2N

Polymer 6 or 7

or

NH2

OH

N

N

N

HO

N

N

N

N

N

7

HO

OH

8

HO

OH

OH

Scheme 5.5 Synthesis of BINOL-containing polymers 7 and 8

transition metal-catalyzed cross-coupling reactions.1,15 Since the 1970s, studies have been carried out to immobilize bidentate bisphosphines in polymer networks.1 While most of the early works focused on polymer-supported bisphosphines, which possess advantages such as easy recovery and reuse, and sometimes higher catalytic activities for derived catalysts compared with their monomeric counterparts, bisphosphine-containing polymers with bisphosphine-units in the polymer backbones have recently also been the subject of research. PPh2 nPPh 2

n = 0-3 DPPA

PPh2 PPh2

Fe

PPh2 PPh2

DPPF

BINAP

PPh2 PPh2 BIPHEP

Chart 5.2

BINAP, one of the most versatile and effective ligands for asymmetric catalysis, has been the most studied bisphosphine ligand for developing recoverable and soluble polymeric catalysts.2 Poly-BINAP 9, in which BINAP was incorporated in the polymer backbone via the Suzuki cross-coupling reaction followed by reduction (Scheme 5.6),16 has been used for the hydrogenation of ketones. This polymer can be easily separated from the reaction system after the completion of the reaction as it was soluble in common organic solvents such as methylene chloride, THF, chloroform, and toluene, but insoluble in methanol. BINAP-containing polymer 10, in which BINAP was incorporated into the polymer backbone through amide linkage (Scheme 5.7), has also been reported.17 Polymer 10 was employed as ligands for catalytic hydrogenation of b-ketoesters and amidoacrylic acids. It was found that the catalytic property of 10 compared favorably with that of BINAP.

106

Recoverable and Recyclable Catalysts PPh2

PPh2

O O

B

O

R

PPh2 PPh2

+ Br

1. Pd(dppf)Cl2 . CH.2Cl2 2 M K2CO3, THF

R

R

2. HSiCl3, Et3N, Xylene

O

O B

Br R

R

R

R=C6H13

9

O

PPh2

PPh2

Scheme 5.6 Synthesis of BINAP-containing polymer 9

O

NCO

OCN

H2N 1).

HN

PPh2 2).

PPh2

O

HN N H

OH

N H

10

H2 N PPh2

PPh2


Soluble chiral polyester-linked BINAP-containing polymers, 11, and 12, have been prepared through the condensation of 5,50 -diamino-BINAP, terephthaloyl chloride and (2S,4S)-pentanediol by Chan and coworkers (Scheme 5.8).18 These polymers were soluble in toluene, THF, and methylene chloride, and can be quantitatively recovered by H2N

NH2

O

+ Ph2P

O

Cl

+

Cl

Polymer 11 or 12

OH

HO

PPh2

R or S O

O

O N H

Ph2P

O

O

N H

O

O N H

Ph2P

PPh2

O

O

O

11

PPh2

N H

O

O

12

Scheme 5.8 Synthesis of BINAP-containing polymers 11 and 12


107

precipitation with methanol. These polymers have been employed for Ru-catalyzed hydrogenation of 2-(6-methoxyl-2-naphthyl)-propenoic acid. A noticeable solvent effect was observed for these polymer ligands; the soluble polymer catalyst was also observed to be more active than the corresponding parent homogeneous catalyst while retaining similar enantioselectivity. Such observations demonstrated the importance of the solubility property of the polymeric catalyst under catalytic reaction conditions. The polymeric catalyst has been demonstrated to be recoverable from the reaction mixture using solvent precipitation and could be reused for ten cycles without any loss of catalytic activity and enantioselectivity. BINAP-containing polymers with PEG portions, 13 and 14 (Schemes 5.9 and 5.10),19,20 have been reported to be more soluble in methanol. Application of these polymers for asymmetric hydrogenation showed that catalyst systems derived from these polymers gave H2N

NH 2

Ph2P

O

O +

+

Cl

Cl

CH2Cl2, 40-50 oC

O

HO

OH p-1

Pyridine

PPh2

O NH

O

O

O

O

O

N H

O p-1 n

m

Ph2P

13

PPh2


O H2N

NH2

Ph2P

PPh2

O

1). Cl

Cl DMAc, NEt3, 0 oC to r. t., 30 min

2). MeO-PEG-OH (Mn = 5,000) DMAc, NEt3, r. t. to 50 oC, 8 h

O

O

O NH

O O-PEG-OMe

N H

MeO-PEG-O

m Ph2P

PPh2

14


slightly higher catalytic activity and enantioselectivity that that of the parent homogeneous catalyst. The catalysts could be recovered and reused without loss of catalytic properties. Bisphosphine-containing polymers 15 and 16 have also been reported.21 These polymers were prepared via the Suzuki cross-coupling reaction (Scheme 5.11). They are soluble in common organic solvents such as THF, CH2Cl2, and toluene, but insoluble in methanol.

108

Recoverable and Recyclable Catalysts 1) (HO)2B

OR B(OH)2

RO

Fe

Br

Pd(PPh3)4, THF/K2CO3

Br

2). HPPh2

HO

OH

Polymer 15 or 16

OR

Fe

RO

OR

Fe PPh2

Ph2P

RO

PPh2 Ph2P R = C6H13

R = C6H13 15

16

Scheme 5.11 Synthesis of bisphosphine-containing polymers 15 and 16

They have been employed as efficient ligands for Ni-catalyzed cross-coupling reactions, and were found to be recoverable and reusable. 5.2.3 Salen-containing Polymeric Catalytic Systems Salen (17), one of the most extensively used ligands for asymmetric epoxidation of alkenes, has also been converted to the polymeric form for asymmetric catalysis.22,23 Salencontaining polymers 18 and 19 have been reported by Zheng and coworkers.24 The chiral polymer ligands 18 and 19 were synthesized by the polycondensation of (1R,2R)diaminocyclohexane with salicylaldehydes in ethanol (Scheme 5.12). The polymeric salen–Mn complexes, which were very soluble in CH2Cl2 but almost insoluble in hexanes and diethyl ether, were employed in the enantioselective epoxidation of olefins under homogeneous conditions. The results showed that the local C2 symmetry in the individual salen units in the polymer systems was maintained. The catalyst was recovered and reused several times by a simple catalysis separation method but the recovered catalyst was found to have lower enantioselectivity and activity after each cycle.

N

N

OH HO

17

Chart 5.3

5.2.4 BINOL–BINAP-based Bifunctional Polymeric Catalytic Systems Polymers containing more than one type of ligands in their backbones have also been reported. Such polymers were designed to function as ligands for sequential reactions to further heighten the efficiency of polymeric catalyst systems. For example, Pu and coworkers prepared a BINOL–BINAP-based bifunctional polymeric ligand, 20 (Scheme 5.13),25


O

O

HO

N

OH

109

N

OH HO

H2C

n

18 + NH2

H2N

or O

O O

HO

N

N

OH

OH HO

H2C O H2C

n

19

Scheme 5.12 Synthesis of salen-containing polymers 18 and 19

which contains two distinctively different catalytic sites, BINOL and BINAP. Its ruthenium diphenylethylenediamine complex was used in the one-pot enantioselective diethylzinc addition and enantioselective hydrogenation of ketoarylaldehydes, giving products in high yields and high stereoselectivities. This bifunctional polymer catalyst can also be used as

O O

OR

O

B

Br 1). Pd(dppf)Cl . CH Cl 2 2 2

B O

O PPh2 PPh2

OMOM

+

OMOM

K2CO3 (aq.), THF

OR

+

2). H+

OR

O

3). HSiCl3, NEt3

O B

B O

O

Br OR

O PPh2

PPh2

PPh2

PPh2

RO

OR

RO

OR

RO

OR RO OR' OR'

OR R=C6H13, R'=CH2OCH3

20

Scheme 5.13 Synthesis of BINOL–BINAP-containing polymer 20

110


either BINAP or BINOL catalyst for individual asymmetric reactions. (R,R)-20–Rucatalyzed asymmetric hydrogenation of acetophenone was studied. It was found that this soluble polymer catalyst showed higher catalytic activity and enantioselectivity than the insoluble polymer Ru(BINAP) catalysts.26 Another BINOL–BINAP-containing polymer 21, prepared via the direct condensation of (R)-3,30 -diformyl-1,10 -bi-2-naphthol with (R)-5,50 -diamino BINAP (Scheme 5.14), was reported by Fan and coworkers for asymmetric hydrogenation and diethylzinc addition.27 The BINOL and BINAP catalytic centers in this polymer were alternatively organized in a polymer chain. Polymer 21 was found to be effective with good enantioselectivity in the Ru (II)-catalyzed asymmetric hydrogenation of 2-arylacrylic acids and diethylzinc addition to benzaldehyde.

H2N

N H2

+

OHC HO

N

HO

OH CHO

OH

PPh2

N

N

PPh2

HO

OH

N

21 PPh2

PPh2

PPh2

PPh2

Scheme 5.14 Synthesis of BINOL–BINAP-containing polymer 21

5.2.5 Dendrimer Catalyst Systems Dendrimers are tree-like molecules with unique structure and solubility.5 The catalytic sites of a dendritic catalyst could be at the core, branch or periphery. Extensive research on the synthesis and preparation of dendrimer catalysts has been carried out.5 Although dendritic catalysts are generally more soluble than their linear polymeric counterparts, the enantioselectivities of many dendritic catalysts were observed to be lower than those of their parent chiral catalysts, which was most likely due to the dense packing of the catalytically active site at the periphery. Summarized below are two families of dendritic catalysts with enantioselectivities comparable with those of their corresponding parent catalysts. Dendritic b-amino alcohol-containing polymers 22 and 23, prepared from Sonogashira coupling (Scheme 5.15), have been reported for dialkylzinc addition reactions.28 It was found that both dendrimers showed high enantioselectivity (up to 86% e. e.) in the addition of dialkylzinc to aromatic aldehydes. Moreover, catalyst 22 could be recovered and reused without any loss of enantioselectivity. The enantioselectivities were lower than those using the corresponding monomer (1R,2S)-N-benzylephedrine.

Polymeric, Recoverable Catalytic Systems N

111

OH Ph

I I

I

+

HO

or

Ph

N

Pd(PPh3)4, CuI Ph

Dendrimer 22 or 23

OH

NEt3, DMF

N

Ph

HO

Ph

N

N

HO

Ph N

N

N

OH Ph

Ph HO

OH

OH

N

N

Ph

OH Ph

23

22 Ph N OH

Ph

N HO

Scheme 5.15 Synthesis of amino alcohol-containing dendrimers 22 and 23

Togni and coworkers reported an efficient synthesis of a family of dendritic chiral phosphines including 24 and 25 with Josiphos units (Scheme 5.16).29 These dendrimers were employed in the Rh-catalyzed hydrogenation of dimethyl itaconate and up to 98.6% e.e. was observed. This result was similar to that obtained with the monomeric Josiphos catalyst (99% e.e.). The dendritic catalysts could be separated from the reaction mixture through a nanofiltration membrane. Dendritic chiral phosphines with cyclophosphazene as the core, 26 and 27, were also reported by Togni and coworkers (Scheme 5.17).30 These dendrimers were also employed in the Rh-catalyzed hydrogenation of dimethyl itaconate and ca. 98% e.e. was observed. This result was similar to that obtained with the monomeric Josiphos catalyst (99% e.e.). 5.2.6 Dendronized Polymeric Catalytic Systems Dendronized polymers are macromolecules with dendritic side chains attached to polymeric cores.31–34 In dendronized polymers with rigid polymeric cores which can easily attain rod-like shapes, dendritic wedges are attached to the polymer backbone in a rotating fashion along the polymer axis. This unique structural feature accounts for a great potential to compete with linear polymers and dendrimers for asymmetric catalysis. Optically active dendronized polymers containing lower generation dendritic wedges with monomeric chiral units in their peripheries could overcome the condensed packing problem existing in the dendrimer-based chiral catalysts without sacrificing the advantage of easy recovery by either filtration or precipitation. They also possess multiple catalytic sites per repeat unit with higher solubility. The combination of features such as more catalytic sites, higher solubility, and nanoscopic dimensions of optically active dendronized polymers renders

112

Recoverable and Recyclable Catalysts PCy2

PCy2 Fe

PPh2

Cl

PPh2

Fe

Cl

O

O Cl

O

Si

Si

+ NH

COCl

HN

O

Dendrimer 24 or 25

or

O

ClOC

COCl

ClOC OH

Cy2P

Fe

Si

H N

Si

Ph2P

O

NH

O

Ph2P

Fe

Cy2P H N

H N O

O

O

O

Si O

O

Si

24

Fe Ph2P

Ph2P Cy2P

Fe

O

O

Cy2P

O

O PPh2

Cy2P

Fe

Si

Fe

Fe

HN

NH

Ph2P

Si

PCy2 PPh2

PCy2

PCy2 PPh2

Fe

Si

Si

HN

NH PCy2 PPh2

O

O

Fe

Si

Fe

Ph2P Si

Cy2P

CO2

O

O NH

HN

25

O2C CO2 NH

O2C

HN O

O

PCy2

Si

PPh2

O

O

Si

Fe

Fe NH

Si Ph2P Cy2P

HN

Ph2P

PCy2

Fe

Fe Ph2P Cy2P

Scheme 5.16 Synthesis of bisphosphine-containing dendrimers 24 and 25

Polymeric, Recoverable Catalytic Systems PCy2

PCy2 PPh2

Fe

P N Cl

Si

Si

Cl

Cl

PPh2

Fe

113

N

P

P N

Cl

+

Cl Cl

Dendrimer 26 or 27

or Cl

NH

Cl Cl N P Cl P N

HN

O

P Cl Cl

N

O

P N Cl Cl OH

PCy2 PPh2

PCy2 PPh2

Fe

Fe

Si

Si Cy2P

Ph2P

PCy2

Si Fe

PPh2 Fe

Cy2P

NH

Si

Ph2P

NH

NH O

Fe

PCy2

O

O

Si

PPh2

H N

Fe O

NH

O

O O

Si P

O

n

N

Si

NH

N

O

HN

O

O P

Fe Ph2P O

Si

O

O

Fe Ph2P

26: n = 1 27: n = 2

O

N

O

HN

Cy2P

P

HN

O O

HN

HN

Cy2P

HN Si

Fe PPh2 PCy2

Si

Si

Si Fe Ph2P Cy2P

Fe

Fe PPh2 PCy2

Ph2P Cy2P

Scheme 5.17 Synthesis of bisphosphine-containing dendrimers 26 and 27

them potentially more efficient than linear polymeric and dendritic chiral catalysts. These optically active dendronized polymers represent a new type of polymeric chiral catalysts. Optically active dendronized polymers 28 and 29, prepared by the Suzuki coupling polymerization (Scheme 5.18), have been reported by Hu and coworkers.35 These ephedrinecontaining polymers are soluble in common organic solvents such as THF, toluene, and dichloromethane, but insoluble in methanol. 28 and 29 have been employed as polymeric chiral catalysts for the asymmetric addition of diethylzinc to benzaldehyde and the catalytic

114

Recoverable and Recyclable Catalysts Ph

Ph

HO OH N Ph

OH N

N

RO (HO)2B

or

Br

n OR n = 0, 1

n B(OH)2

Polymer 28 or 29

Pd(PPh3)4, 2M K2CO3/ THF reflux, 24 h

Br

Br

Br

HO

Ph

Ph

N

N

OH Ph

OR

OR

OH N

OR

n

OR

n

OR

n

OR

OR

OR

n R = n-C6H13

R = C6H13

n = 0, 1: 29

28 Ph OH

N

N HO

N Ph

Ph

OH

Scheme 5.18 Synthesis of optically active dendronized polymers 28 and 29

properties of these dendronized polymers have been compared with those of their corresponding linear polymeric and dendritic chiral catalysts. It was found that 28 and 29 were more efficient than their corresponding linear polymeric and dendritic chiral catalysts. 28 and 29 can be easily recovered by filtration and reused without loss of enantioselectivity.

5.3 Summary The development of recoverable and reusable catalysts, especially enantioselective ones, has been the subject of intense study over the past decades. While the use of insoluble polymer-supported catalysts, which represented a significant part of this field, has shown high enantioselectivity and/or activity, in some cases were even superior to the parent homogeneous systems, most insoluble polymer-supported catalysts suffered from lower catalytic activity than the homogeneous analogues. Using soluble polymer or dendrimer as chiral catalyst support represents a relatively new approach to the problem of catalyst recycling and reuse. Such catalysts are expected to have the advantages of homogeneous catalysts during the reaction and be easily separated by precipitation methods after the reaction. Excellent catalytic efficiency has been achieved by using the soluble polymer catalysts such as BINOL- and BINAP-containing polymeric catalysts. Studies showed that it is possible to systematically fine-tune the structure, size,


115

shape, and solubility of a linear polymeric or dendritic chiral catalysts. Dendronized polymers are considered to be a unique type of soluble polymeric catalysts for organic synthesis, and might fill the gap between homogeneous and heterogeneous catalysis. They are thus expected to have a promising future. The development of highly active and enantioselective catalysts with broad applicability and the capability for recycle and reuse will continue to be an important goal and challenge in this field.

Acknowledgements Our work in this area was supported by the Department of Chemistry at the College of Staten Island – City University of New York, the US National Institutes of Health (1R15 GM69704), the donors of the Petroleum Research Fund, administered by the American Chemical Society (36348-G1 & 44428-AC1), the US National Science Foundation (CHE079311) and the PSC-CUNY Research Award Programs. I am also grateful for the contribution of the dedicated students and post-doctoral associates in my laboratory, especially Dr. Chaode Sun, Dr. Zhen-Yu Tang, Dr. Yong Lu, and Dr. Cheng-Guo Dong. I also thank Professor Lin Pu at the University of Virginia for encouraging me to work in this fascinating area and Dr. Hong-Bin Yu for her helpful discussion and editorial assistance.

References 1. (a) Itsuno, S. Polymeric Materials Encyclopedia; Synthesis, Properties and Applications. Salamone, J. C.; ed.; CRC Press, Boca Raton, FL, 1996, Vol. 10, P 8078. For recent comprehensive reviews: (b) J. Gladysz (Guest Editor), Chem. Rev. 2002, 102, 3215–3810. Also see: (c) S. V. Ley, I. R. Baxendale, R. N. Bream, P. S. Jackson, A. G. Leach, D. A. Longbottom, M. Nesi, J. S. Scott, I. Storer, S. J. Taylor, J. Chem. Soc., Perkin Trans. 1 2000, 3815–4195. 2. (a) Jacobsen, E. N.; Pfaltz, A.; Yamamoto, Y. Comprehensive Asymmetric Catalysis, Springer, 1999. (b) Ojima, I. Catalytic Asymmetric Synthesis; VCH, 1993. 3. For example: (a) Bolm, C.; Dinter, C. L.; Seger, A.; H€ ocker, H.; Brozio, J. J. Org. Chem. 1999, 64, 5730. (b) Dumont, W.; Poulin, J. C.; Dang, T. P.; Kagan, H. B. J. Am. Chem. Soc. 1973, 95, 8295. 4. (a) Pu, L. Chem. Eur. J. 1999, 5, 2227. (b) Yu, H.-B.; Hu, Q.-S.; Pu, L. J. Am. Chem. Soc. 2001, 122, 6500. (c) Fan, Q.-H.; Ren, C.-Y.; Yeung, C.-H.; Hu, W.-H.; Chan, A. S. C. J. Am. Chem. Soc. 1999, 121, 7407. 5. Recent reviews: (a) Oosterrom, G. E.; Reek, J. N. H.; Kamer, P. C. J.; van Leeuwen, P. W. N. M.; Angew. Chem. Int. Ed. Engl. 2001, 40, 1827. (b) Seebach, D.; Rheiner, P. B.; Greiveldinger, G.; Butz, T.; Sellner, H. Top. Curr. Chem. 1998, 197, 125. (c) Newkome, G. R.; Moorfield, C. N.; Vogtle, F. Dendritic Molecules: Concepts, Synthesis, perspectives. VCH, Weinheim, 1996. 6. For reviews: Pu, L. Chem. Rev. 1998, 98, 2405. 7. (a) Hu, Q.-S.; Zheng, X.-F.; Pu, L. J. Org. Chem. 1996, 61, 5200. (b) Hu, Q.-S.; Vitharana, D.; Zheng, X.-F.; Wu, C.; Kwan, C. M. S.; Pu, L. J. Org. Chem. 1996, 61, 8370–8377. 8. (a) Hu, Q.-S.; Huang, W.-S.; Vitharana, D.; Zheng, X.-F.; Pu, L. J. Am. Chem. Soc. 1997, 119, 12454. (b) Huang, W.-S.; Hu, Q.-S.; Zheng, X.-F.; Anderson, J.; Pu, L. J. Am. Chem. Soc. 1997, 119, 4313. 9. Huang, W.-S.; Hu, Q.-S.; Pu, L. J. Org. Chem. 1998, 63, 1364. 10. (a) Huang, W.-S.; Hu, Q.-S.; Huang, W.-S.; Pu, L. J. Org. Chem. 1999, 64, 7940. (b) Hu, Q.-S.; Huang, W.-S.; Pu, L. J. Org. Chem. 1998, 63, 2798. 11. Johannsen, M.; Jorgenson, K. A.; Zheng, X.-F.; Hu, Q.-S.; Pu, L. J. Org. Chem. 1999, 64, 299–301.

116


12. Simonsen, K. B.; Jorgenson, K. A.; Hu, Q.-S.; Pu, L. Chem. Comm. 1999, 811. 13. Yu, H.-B.; Zheng, X.-F.; Lin, Z.-M.; Hu, Q.-S.; Huang, W.-S.; Pu, L. J. Org. Chem. 1999, 64, 8149. 14. Dong, C.; Zhang, J.; Zjheng, W.; Zhang, L.; Yu, Z.; Choi, M. C. K.; Chan, A. S. C. Tetrahedron Asymm. 2000, 11, 2449. 15. (a) M. Beller, C. Bolm, Transition Metals for Organic Synthesis Wiley-VCH, Weinheim, 1998. (b) P. J. Stang, Transition Metal-Catalyzed Cross-Coupling Reactions. Wiley-VCH, New York, 1998. (c) J. Tsuji, Palladium Reagents and Catalysts. Wiley, Chichester, 1995. (d) J. P. Collman, L. S. Hegedus, J. R. Norton, R. G. Finke, Principles and Applications of Organotransition Metal Chemistry, University Science Books, Mill Valley, CA, 1987. 16. Yu, H.-B.; Hu, Q.-S.; Pu, L. Tetrahedron Lett. 2000, 41, 1681. 17. Halle, R.; Colasson, B.; Schulz, E.; Spagnol, M.; Leamire, M. Tetrahedron Lett. 2000, 41, 643. 18. Fan, Q. H.; Ren, C. Y., Yeung, C. H.; Hu, W. H.; Chan, A. S. C. J. Am. Chem. Soc. 1999, 121, 7407. 19. Fan, Q. H.; Den, G. J; Chen, X. M.; Xie, W. C.; Jiang, D. Z.; Liu, D. S.; Chan, A. S. C. J. Mol. Cat. A. Chem. 2000, 159, 37. 20. Fan, Q. H.; Deng, G. J.; Lin, C. C.; Chan, A. S. C. Tetrahedron Asymm. 2001, 12, 1241. 21. Lu, Y.; Plocher, E.; Hu, Q.-S. Adv. Synth. Catal. 2006, 348, 841. 22. McGarrigle, E. M.; Gillheany, D. G. Chem. Rev. 2007, 105, 1563. 23. Baleizao, C.; Garcia, H. Chem. Rev. 2006, 106, 3987. 24. Huang, W.; Song, Y.; Bai, C.; Cao, G.; Zheng, Z. Tetrahedron Lett. 2004, 45, 4763. 25. Yu, H.-B.; Hu, Q.-S.; Pu, L. J. Am. Chem. Soc. 2000, 122, 6500–6501. 26. ter Halle, R.; Schulz, E.; Spagnol, M.; Lemaire, M. SynLett 2000, 680. 27. Fan, Q. H.; Liu, G. H.; Deng, G. J.; Chen, X. M.; Chan, A. S. C. Tetrahedron Lett. 2001, 42, 9047. 28. Sato, I.; Shibata, T.; Ohtake, K.; Kodaka, R.; Hirokawa, Y.; Shirai, N. Soai, K. Tetrahedron Lett. 2000, 41, 3123. 29. Kollner, C.; Pugin, B.; Togni, A. J. Am. Chem. Soc. 1998, 120, 10274. 30. (a) Schneider, R.; Kollner, C.; Weber, I. Togni, A. Chem. Commun. 1999, 2415. (b) Togni, A.; Bieler, N.; Burckhardt, Y.; Kollner, C.; Pioda, G.; Shneider, R.; Schnyder, A. Pure Appl. Chem. 1999, 71, 1531. 31. (a) Hecht, S.; Frechet, J. M. J. Angew. Chem. Int. Ed. 2001, 40, 75. (b) Sch€ ulter, A. D.; Rabe, J. P. Angew. Chem. Int. Ed. 2000, 39, 864. (c) Frey, H. Angew. Chem. Int. Ed. 1998, 37, 2193. 32. Tomalia, D. A.; Kirchhoff, P. M. (Dow Chemical), U.S. Pat. 4694064, 1987. 33. For examples of dendronized polymers with flexible polymeric cores: (a) Shu, L.; Schl€ uter, A.; Macromolecules 2000, 33, 4321. (b) Yin, R.; Zhu, Y.; Tomalia, D. A. J. Am. Chem. Soc. 1998, 120, 1278. (c) Jahromi, S.; Coussens, B.; Meijerink, N.; Braam, A. W. M. J. Am. Chem. Soc. 1998; 120, 9753. (d) Percec, V.; Ahn, C.-H.; Ungar, G.; Reardley, D. J. P.; M€ oller, M.; Sheiko, S. S. Nature, 1998, 391, 161. 34. For examples of dendronized polymers with rigid polymeric cores: (a) Wyatt, S. R.; Hu, Q.-S.; Yan, X.-L.; Bare, W. D.; Pu, L. Macromolecules 2001, 34, 7983. (a) Marsitzky, D.; Vestberg, R.; Blainey, P.; Tang, B. T.; Hawker, C. J.; Carter, K. R. J. Am. Chem. Soc. 2001, 123, 6965. (b) Bo, Z.; Schl€uter, A. D. Chem. Eur. J. 2000, 6, 3235. (c) Sato, T.; Jiang, D.-L.; Aida, T. J. Am. Chem. Soc. 1999, 121, 10658. (d) Bao, Z.; Amundson, K. R.; Lovinger, A. L. Macromolecules 1998, 31, 8647. (e) Kaneko, T.; Horie, T.; Asano, M.; Aoki, T.; Oikawa, E. Macromolecules, 1997, 30, 3118. 35. (a) Hu, Q.-S.; Sun, C.; Monaghan, C. E. Tetrahedron Lett. 2002, 43, 927–930. (b) Hu, Q.-S.; Sun, C.; Monaghan, E. M. Tetrahedron Lett. 2001, 42, 7725.

6 Thermomorphic Catalysts David E. Bergbreiter Department of Chemistry, Texas A&M University, College Station, TX, USA

6.1 Introduction Thermomorphic catalysts are homogeneous catalysts that are used in a system where there is a phase change during a reaction as a consequence of a temperature change. Their importance lies in this phase separation process that affects the reactivity of catalyst or facilitates product catalyst separation. Most commonly, thermomorphic catalysts or thermomorphic catalyst–ligand complexes are separated from products after a reaction by virtue of their phase behavior. Ideally such recovered catalysts can then be reused. Such catalysts retain the virtue of traditional homogeneous catalysts in that they react with a substrate in a single phase as a homogeneous species. This minimizes potential reactivity problems associated with carrying out a reaction in a biphasic liquid–liquid or liquid–solid system where the catalyst and substrate are in different phases. Such situations are sometimes encountered with aqueous biphasic catalysis or with homogeneous catalysts immobilized on an always insoluble organic or inorganic support. In thermomorphic systems, the catalyst or catalyst ligand complex is designed so that the catalyst or the catalyst ligand complex separates into a different phase from the bulk of the product only after the reaction is complete. Examples of such systems include catalysts that separate as a solid, leaving products in solution; products that separate as solids, leaving catalysts in solution; or catalysts and products that separate into different liquid phases of a biphasic liquid–liquid mixture after a reaction. The general features of each of these three general


118


strategies are discussed briefly below. Then examples where these strategies are used in specific types of homogenous catalytic reactions are discussed in more detail.

6.2 Thermomorphic Catalyst Separation Strategies Thermomorphic catalysts that are separated from products as solids are most often separated from products by cooling. Typically the catalyst is soluble at elevated temperature and insoluble at a lower temperature where the separation is effected. This scheme, shown in Figure 6.1(a), uses materials whose solubility changes in a normal way and requires an eventual solid–liquid separation – typically filtration or centrifugation – to recover the catalyst. An alternative approach that also uses a solid–liquid separation is possible with catalysts that are supported on oligomers or polymers that that exhibit inverse temperature-dependent solubility. This is possible because polymers have a lower critical solution temperature (an LCST).1 This seemingly unusual temperature-dependent solubility is a consequence of the normal unfavorable entropy of dissolution of polymers. In most cases, polymer solubility requires solvent reorganization – a reorganization that is enthalpically favorable due to solvent–polymer interactions but can be entropically unfavorable due to the requirement that many solvating solvent molecules have to become associated with each of the many repeating units of a single polymer molecule. If catalysts, ligands, or substrates are attached to polymers at modest mol% loadings, the soluble polymer-bound catalyst will not significantly change the solubility characteristics of the polymer. Catalysts, ligands, or substrates on polymers that exhibit LCST behavior can thus be designed to be soluble in a cold solution but to become insoluble on heating because of the polymer support’s LCST behavior. By choosing an appropriate polymer, a catalyst can be used as a homogeneous solution near room temperature or below room temperature and can then be quantitatively separated by gentle heating (Figure 6.1b). Such lower critical solution temperature (LCST) behavior can also be used to affect a catalyst’s or substrate’s reactivity to prepare ‘smart’ catalysts, ‘smart’ substrates, ‘smart’ surfaces, or catalysts that exhibit hyper-Arrhenius reactivity.2–9 A second thermomorphic catalyst separation strategy uses the temperature-dependent miscibility of solvent mixtures and the phase selectivity of catalyst, ligand, or substrate to effect a homogeneous single phase catalytic reaction with a gravity-based separation afterwards.10–15 There are countless examples of binary or ternary solvent mixtures that

Figure 6.1 Thermomorphic changes that lead to (a) formation of a solid catalyst from a single homogeneous phase on cooling, or (b) formation of a solid separable catalyst phase from a single homogeneous phase on heating

Thermomorphic Catalysts

119

Figure 6.2 Schemes for recovering, separating and reusing a thermomorphic soluble polymerbound catalysts under liquid–liquid conditions from a solution containing a mixture of solvents

exhibit temperature-dependent miscibility.16 The most common scheme in these cases is to use a ligand that immobilizes the catalyst in either the polar or nonpolar phase after a reaction (Figure 6.2). The choice of which phase is most appropriate is determined by the solubility of the products. Ideally the catalyst will be essentially insoluble in the productcontaining phase. Partitioning of some product into the catalyst-containing phase is not desired but often occurs because low molecular weight products usually have some solubility in both phases of a biphasic mixture of two organic solvents. However, if a catalyst were recycled multiple times, any product solubility in the catalyst-containing phase will eventually reach a limiting value so that any loss of a small amount of product in the first few cycles averaged over multiple cycles will be of minimal consequence in the overall product yield. Product loss can also be minimized by an additional product extraction step in cases where the phase selective solubility of catalyst in the catalystcontaining phase is sufficiently high. Two general schemes are used most often. Usually the catalyst (the black solution in Figure 6.2) is separated from the product solution (the clear solution in Figure 6.2) by cooling the monophasic reaction solution. Depending on the nature of the catalyst and its support, the catalyst-containing phase separated solution can be the denser, usually more polar phase (as in Figure 6.2). Alternatively the catalyst can be in the less dense, usually less polar phase. In either case, the solvent mixture’s miscibility changes with temperature so that the catalysis occurs in a homogeneous monophasic solution with a subsequent biphasic separation. Several additional considerations affect the practicality of a thermomorphic liquid– liquid recycling scheme. First, the separation step produces a catalyst- and productcontaining phase each of which contain at least a small portion of the solvent that is predominant in the other phase. For example, a heptane–N,N-dimethylformamide (DMF) mixture that is miscible at 70 C and immiscible at 25 C can be separated, but the heptane phase contains some DMF and the DMF phase contains some heptane. The amount of heptane in the DMF or the amount of DMF in the heptane phase varies with temperature, but if the heptane phase were used to recover the catalyst and 10% heptane were present in the DMF phase, recycling requires that subsequent additions of a DMF-soluble substrate solution include enough heptane to compensate for the loss of heptane in the product DMF phase (otherwise the volume of the catalyst-containing phase will asymptotically approach zero). This loss of some solvent from the catalyst-carrying phase is only superficially a problem since any process using this strategy could avoid this solvent loss by using a substrate–product solvent phase that is saturated with the catalyst-containing phase solvent (e.g. heptane-saturated DMF in the example cited).

120


A second consideration is that reactions often produce byproducts. A Heck reaction with triethylamine, for example, produces a triethylammonium salt. A reaction that uses an inorganic carbonate as a base produces water and salt. If a thermomorphic liquid–liquid separation scheme were used where a polar phase soluble catalyst were separated from an insoluble nonpolar phase soluble product and if a byproduct like salt or water accumulated in the polar phase, the conditions required to achieve miscibility in subsequent cycles would gradually change. Eventually, accumulation of such byproducts would reach a point where it could become impossible to achieve full miscibility of the catalyst and substrate phases. Byproduct accumulation is a problem in any catalyst recovery scheme and is a general but often undiscussed issue for catalyst recovery schemes where catalysts are recovered in a polar phase and has to be considered on a case-by-case basis. A third consideration in a thermomorphic reaction scheme with two solvents is that the solvent medium during the catalysis is often a solvent mixture. One of these solvents could alter a catalyst’s activity either in a positive or negative way. A final sort of thermomorphic behavior is that seen in thermoregulated catalysis chemistry.10,17–19 In this case, a liquid–liquid biphasic system is present throughout the reaction but the polymer’s solubility changes due to heating so that the polymer and the catalyst bound to the polymer moves from one phase to the other. In the most common example of this approach to thermomorphic catalysis, a polymer that exhibits lower critical solution temperature (LCST) behavior in water is used in an aqueous–organic biphasic system. If the polymer phase separates as an oil-in-water emulsion and if the discontinuous hydrophobic polymer-containing phase that forms is soluble in the water-immiscible phase, a catalyst attached to the soluble polymer can migrate from the aqueous phase to the substrate containing organic phase with heating and return to the aqueous phase on cooling (Figure 6.3). The sorts of temperature-dependent phase behavior illustrated in Figures 6.1–6.3 above predate the use of the term thermomorphic as applied to catalysis. Indeed, the idea of using temperature changes to affect the solubility of a catalyst or to affect the miscibility behavior of two liquid phases has ample precedence. However, the use of these strategies as a clearly articulated approach in catalyst separation is a more recent development. The discussion below focuses on catalysts that have had the term thermomorphic used in their description. We also have included limited examples of what have been described as thermoregulated catalysts (Scheme 6.3) as those systems generally depend on the inverse temperature-

heating cooling

Figure 6.3 Thermomorphic changes in a biphasic reaction mixture that lead to a thermoregulated catalyst moving from a more polar, more dense aqueous phase to an organic phase on heating and returning to the lower denser aqueous phase on cooling


121

dependent solubility behavior of a substance (usually a polymer or oligomer with an LCST). Thermomorphic behavior in fluorous media is also known and select examples of this chemistry are discussed here. Fluorous chemistry has also been recently reviewed.20 Overall, the discussion in this review is largely limited to work that has been reported since 1998 when our group first used the term ‘thermomorphic’ in connection with homogeneous catalysts.21 A few historical antecedents of catalysts of thermomorphic catalysis chemistry are also discussed. However, it should be recognized that temperatureinduced phase changes – the essence of what makes thermomorphic catalysis interesting – have a very long history. This review makes no pretense of discussing this history in detail. Our first use of the term thermomorphic described the copolymer 1 as a support.21 This copolymer was prepared by reaction of isopropylamine with poly(N-acryloxysuccinimide) (Scheme 6.1).22,23 The copolymer 2 can also be prepared by copolymerization of N-isopropylacrylamide (NIPAM) and N-acryloxysuccinimide (NASI).

Scheme 6.1

Either product polymer contains reactive N-hydroxysuccinimide esters. Such esters had originally been used by Ringsdorf to prepare poly(acrylic acid) derivatives.24 These syntheses all rely on the fact that an N-hydroxysuccinimide ester is an active ester that readily reacts with water, alcohols, or amines to form carboxylic acids, esters, or amides. Thus, it was predictable that reaction of this copolymer and an amine-containing ligand or catalyst followed by quenching of any unreacted active esters with ammonia would be a viable route to a polymer-supported ligand or catalyst. Catalysts as well as other species attached to similar poly(N-isopropylacrylamide) (PNIPAM) and poly(alkene oxide) polymers had been studied earlier by our group.2–6,21,25–30 Those catalysts had LCST properties. Such LCST behavior in our earlier work led us to develop ‘smart’ catalysts or ‘smart’ substrates that phase separated from a substrate solution on heating because of the polymer support’s LCST properties. Such thermomorphic supports are useful because they autonomously control a catalytic reaction’s exotherm. However, while subsequent work showed that poly(alkene oxide)supported ligands and their catalysts that phase separate from water as oils on heating usefully lead to thermoregulated catalysts in aqueous––organic biphasic systems,17–19 our initial work wherein we heated a biphasic organic–aqueous solution of a PNIPAMsupported species did not lead to formation of a separate catalytically active phase with PNIPAM derivatives. Instead of dissolving in the organic phase, incomplete dehydration of the PNIPAM support on heating led to formation of a scummy hydrogel phase that precipitated at the organic– aqueous interface. However, this observation was nonetheless useful as it led us to develop and articulate the concept of liquid–liquid thermorphic catalysis. Specifically, it led us to use a ternary mixture of heptane, ethanol, and water that has temperature-dependent miscibility in place of an always biphasic toluene–water

122


mixture. This mixture of heptanes and aqueous ethanol had what we described as thermomorphic behavior. At room temperature an equivolume mixture of heptane and 90% ethanol–10% water is biphasic. At 70 C, this solvent mixture is a single phase. Studies with a dye-labeled PNIPAM 4 prepared as shown in Scheme 6.2 showed that a PNIPAMsupported dye was soluble exclusively in the polar phase at 25 C and that it was fully soluble in the miscible solvent mixture at 70 C.22,23 Since the studies with the dye-labeled polymer were carried out with the polymer-bound dye being present at a concentration similar to what would be appropriate for a homogeneous catalyst, we reasoned this phase selectively soluble polymer might a useful phase anchor for catalyst–product separations after a homogeneous single phase reaction. This proved to be the case.

Scheme 6.2

6.3 Hydrogenation Reactions Under Thermomorphic Conditions The reaction of 2 to form 5, an analog of Wilkinson’s catalyst (Scheme 6.3), and the use of this PNIPAM-supported Rh complex in alkene hydrogenation was the first reaction we studied to demonstrate how the temperature miscibility of a solvent mixture and the phase selective solubility of a polymer could be useful in thermomorphic catalysis.21

Scheme 6.3

Hydrogenations of 1-dodecene or 1-octene with the Rh(I) catalyst 5 occurred readily in a mixture of 90% EtOH–H2O and heptane when this reaction was carried out in this solvent mixture at 70 C. Under these conditions, the polymeric catalyst 5 was comparable but slightly less reactive than RhCl(PPh3)3. However, unlike RhCl(PPh3)3, the catalyst was >99.8% in the polar phase after the single phase present at 70 C was cooled to 25 C to form a biphasic mixture. Since the product dodecane (or octane) was exclusively


123

(or predominantly) in the nonpolar heptane phase, a simple gravity separation allowed us to recover and reuse this Rh(I) hydrogenation catalyst. Addition of fresh substrate in heptane to the recovered ethanol phase containing the catalyst allowed the catalyst 5 to be recycled four times with no changes in activity. While catalyst 5 was effective in hydrogenation and recyclable and while the initial use of this thermomorphic catalyst served as a proof in principle of the concept of a liquid–liquid thermomorphic separation, the scheme whereby a homogeneous catalyst is recovered and reused after cooling in the polar phase of a thermomorphic mixture of a polar and nonpolar solvent has some intrinsic problems that make this strategy not broadly useful. Specifically, even a casual consideration of the sorts of products of most interest for synthesis in batch-style catalytic processes reveals that most products will likely be decorated with polar functional groups and will thus be more soluble in the polar phase than in a nonpolar heptane-rich phase. In addition, many reactions produce byproducts and, as noted above, such byproducts are often polar. Their accumulation through multiple reaction cycles is likely to affect thermomorphic behavior of a solvent mixture. To address this issue, we decided to prepare nonpolar phase soluble supports so that the catalystcontaining phase would be the phase least likely to contain products or byproducts. To design a heptane-phase soluble support, we prepared poly(N-alkylacrylamide)s with different sized N-alkyl groups. These studies showed that small changes in the size of the N-alkyl group of these acrylamide polymers dramatically change the phase selective solubility of a dye-labeled polymer.22,23 While N-hexyl or N-octyl groups as N-alkyl substituents sufficed to produce heptane-soluble polymers in thermomorphic solvent mixtures, we opted to use poly(N-octadecylacrylamide) as the heptane soluble support in catalytic reactions. Using this support, we prepared a hydrogenation catalyst 7 by chemistry analogous to that used to prepare 5. The Rh(I) complex 7 formed in Scheme 6.4 was then successfully used and recycled as a heptane solution in hydrogenation of polar alkenes like acrylic acid.29

Scheme 6.4

About 20 years ago, we introduced the use of saturated hydrocarbon oligomers as catalyst supports.31–40 That work showed that polyethylene oligomers like 8 with a wide variety of terminal functional groups or ligands could be prepared by anionic polymerization of ethylene. Such functionalized oligomers were shown to function as thermomorphic

124


supports for recoverable catalysts. These low molecular weight versions of polyethylene are useful because their phase behavior is profoundly affected by temperature. Below about ca. 50 C, polyethylene oligomers with an Mn of ca. 1200 are completely insoluble in all solvents. Above about 70 C they are quite soluble in nonpolar or slightly polar solvents like toluene or dibutyl ether. Such supports can be prepared by anionic polymerization,41 by polymethylenation,42 or are commercially available43 and can be separated as solids at room temperature by centrifugation or filtration. During catalysis at elevated temperatures in the appropriate solvents they are soluble and terminally bound catalysts on these thermomorphic supports have reactivity like that of a low molecular weight homogeneous catalyst counterpart containing a dodecyl-containing ligand in place of a polyethylcontaining ligand. The thermomorphic solubility behavior seen for these polyethylene oligomer-bound ligands and the catalysts formed using these ligands is also seen with even simpler catalysts that just contain multiple octadecyl groups. Fan’s group has exploited this solubility behavior to prepare a recoverable, reusable catalyst for asymmetric hydrogenation of ketones using a tris(octadecyl) derivative of gallic acid (9) to attach so-called octadecyl ‘pony tails’ to the amines of (S)-5,50 -diamino-[2,20 -bis(diphenylphosphino)-1,10 binaphthyl] (10).44 The resulting bisphosphine formed from this chiral phosphine was allowed to react with [RuCl2(C6H6)]2 to form in situ a Ru complex (11) that was then used in asymmetric hydrogenation of b-ketoesters (Scheme 6.5). The reactions were carried out at 60 C in solvents like CH2Cl2 or in 3:1 mixtures of ethanol and 1,4-dioxane using 40 atm of H2. The 0.5 mol% of the catalyst present formed a red–brown solution on heating. On cooling to 0 C, the catalyst visually precipitated leaving a clear solution of the chiral b-hydroxyester that was separated from the catalyst by centrifugation. Four cycles were carried out without affecting the ca. 98% yield or the 97% enantioselectivity for reduction of tert-butyl acetoacetate to the S alcohol product. Yields and enantioselectivities mirrored those obtained for a similar BINAP-bound Ru species in EtOH under similar conditions. Ru

Scheme 6.5


125

leaching into the product was measured and was 1.8%. The slight decrease in catalyst reactivity and perhaps the small amount of Ru leaching was attributed to oxygen instability of the Ru hydride catalyst formed in the presence of hydrogen though small amounts of phosphine oxidation could lead to these low levels of Ru leaching too. Dendrimers too can exhibit thermomorphic behavior and dendrimer-bound mixed-metal Pd–Pt nanoparticles have been successfully recycled in hydrogenation chemistry using thermomorphic procedures.45 These mixed-metal nanoparticles were first prepared in dendritic ‘nanoreactors’ which were used to effect size control in the nanoparticle synthesis.46–48 The thermomorphically soluble dendrimers’ solubility was controlled by the introduction of peripheral N-n-hexylamide groups They were then used in hydrogenation of alkynes to Z-alkenes or in hydrogenation of 1,3-cyclooctadiene to cyclooctene (Scheme 6.6). When the dendrimer catalyst 12 was used in Scheme 6.6 in a mixture of heptane and DMF, a single phase formed at 65 C. Cooling to 25 C led to phase separation. Decantation served to remove the product-containing heptane phase from the catalystcontaining DMF-rich phase. In these experiments, yields of cyclooctene were 61%, 99%, 99%, and 99% in cycles 1–4.

Scheme 6.6

Hydrogenations with polymers that undergo phase transitions were used by us to illustrate the use of thermoresponsive polymers as ‘smart’ catalyst supports.2,4,5,27 Other thermally responsive polymers have also been reported where thermomorphic behavior is used both to separate products in a synthesis or for sequestration6,28 Thermomorphic solubility can also be used to separate a catalyst as a solid after a reaction. For example, Shaw’s group described a terminally functionalized poly(N-isopropylacrylamide)supported Rh(I) catalyst 13 that hydrogenated methyl methacrylate (Scheme 6.7).49 However, while the polymeric catalyst 13 could be separated from the soluble product, recovered and reused, the precipitation of the Rh species induced by PNIPAM’s LCST in this example did not shut down the catalyst’s activity.

Scheme 6.7

126


A recent report shows that thermomorphic hydrogenations of 1,5-cyclooctadiene to cyclooctene also occur with simpler PEG4000 ligated Pd nanoparticles at 70 C in a heptane–toluene mixture.50 In these cases, the ca. 90% selectivity was slightly less than the 99% selectivity seen in 1,3-cyclooctadiene hydrogenations studied by Kaneda.45 Wang’s group reported success though six cycles using this thermomorphic Pd nanoparticle catalyst. Thermomorphic hydrogenation catalysts have also been prepared from phosphines like 14 that are more often used under biphasic conditions in thermoregulated catalysis. For example, a Ru catalyst prepared in situ from reaction of 14 and Ru3(CO)12 was successfully used in hydrogenation of styrene at 90 C in toluene (Scheme 6.8).51 At the reaction temperature of 90 C, the reaction mixture was a single phase and a Ru catalyst ligated by 14 was soluble. However, this polar poly(alkene oxide)-bound catalyst was separable from the solvent on cooling. In this example, the catalyst separated from the toluene solution of ethylbenzene product as a viscous gum. A total of ten cycles were reported in this example of a hydrogenation with these Ru-bound PEG–phosphine catalysts. Thermomorphic hydrogenation of a broader range of alkenes using a catalyst formed from this same phosphine 14 with the trihydrate of RuCl3 in a mixture of H(OCH2CH2)3OCH3, and heptane is also successful. In this case, alkenes like 1-decene, 1-dodecene, cyclohexene, crotonaldehyde, and styrene were all successfully hydrogenated under monophasic conditions at 80 C. Ten cycles were carried out with 1-octene as a substrate. Ru leaching in the heptane phase was also measured by ICP-AES and was < 0.18 wt%.

Scheme 6.8

6.4 Hydroformylation Reactions Under Thermomorphic Conditions Some of the most elegant and detailed studies of thermomorphic catalysis systems have been carried out by Behr’s group.10,11 Catalytic hydroformylation has been at the forefront of these studies. For example, Behr’s group examined the isomerizing hydroformylation of E-4-octene to 1-nonanal under several conditions.52–54 These studies included hydroformylation in a toluene or propylene carbonate single phase system. They also examined this hydroformylation in a biphasic system and compared the results of biphasic catalysis with isomerizing hydroformylation in a thermomorphic system formed from a carbonate ester of a 1,2-diol, dodecane, and a third solvent (p-xylene or an N-alkyl-2-pyrrolidone) (Scheme 6.9). This study, like a number of other studies from this group,10,11 is much more detailed than most in that it not only reports on the catalysis and separation of the catalyst, but it analyzed the phase diagrams of these thermomorphic solvent systems. These studies also examined the effects of varying amounts of the components of various ternary solvent mixtures on the selectivity for the n-nonanal product and on the leaching level of the Rh catalyst. These authors noted that by adjusting the ratios of various system components, high conversions of the starting alkene to aldehyde could be achieved with very low 0.04% Rh and 0.51% phosphine ligand leaching though the selectivity for the n-nonanal product


127

Scheme 6.9

was only 80%. Higher selectivity toward the linear aldehyde product could be obtained but usually at the expense of higher levels of Rh and phosphine leaching. Other groups have also examined a propylene carbonate–dodecane–1,4-dioxane thermomorphic system for hydroformylation.55 They showed that the solubilities of hydrogen and carbon monoxide in this system are higher than those in many other hydroformylation solvents. This study also was able to predict the gas solubilities using thermodynamically modeling. Hydroformylation of either terminal or disubstituted alkenes under thermomorphic conditions using HRhCO(PPh3)3–P(OPh)3 in propylene carbonate and heptane was described by El Ali.56 For 1-octene which was the most studied substrate, linear/branched ratios were ca. 9:1. Recycling was also extensively studied with 1-octene with eight recycles showing little or no change in yields and selectivity. Hydroformylation of disubstituted alkenes (Scheme 6.10) was also described. In these cases the linear:branched ratio was generally higher (ca: 20:1) but the partitioning of the aldehydes product in the heptane and propylene carbonate phases was very similar.

Scheme 6.10

Fluorous phase chemistry has developed into a valuable approach for both synthesis and catalysis since Horvath and Rabai’s seminal publication appeared in 1994.57–60 While this chemistry is often carried out under biphasic conditions as in aqueous biphasic catalysis,61 the original exposition of the concept of fluorous biphasic catalysis by Horvath and Rabai emphasized the utility of the known temperature-dependent miscibility of mixtures of fluorous and organic solvents.16 This was graphically illustrated in the Horvath–Rabai paper with pictures of a deep-blue fluorous-labeled phthalocyanine dye that was selectively

128


soluble in the denser fluorocarbon phase of a liquid–liquid biphasic mixture of a fluorocarbon and an alkane at room temperature but fully miscible at elevated temperature. This temperature-dependent miscibility of fluorous solutions of a dye has been extended to fluorous-labeled dye-labeled polymers.62 It also has been extended to the use of fluoropolymers that have catalysts attached to them. Early examples of these catalytic studies used fluorous-tagged supports that were prepared using a strategy used earlier to make polymeric fluorous phase hydrogenation catalysts and metal sequestrants and did not use thermomorphic conditions.63–65 However, in subsequent work by Xiao’s group, the thermomorphic character of fluorous solvents and a polymeric hydroformylation catalyst that was prepared from a fluorous polymer containing a phosphine ligand prepared by free radical copolymerization of styryldiphenylphosphine with a fluoroacrylate monomer was successfully used.66 In this work, reaction of this polymeric phosphine 16 with [Rh(CO)2(acac)] in a hexane–toluene–perfluoromethylcyclohexane (40:20:40, vol) solvent mixture formed a Rh(I) hydroformylation catalyst (Scheme 6.11). While other studies with catalysts on polymeric fluorous supports used biphasic conditions or supercritical carbon dioxide as the reaction media,64,67–70 Xiao noted that the catalyst 17 in this solvent mixture formed a monophasic solution at 50 C which on cooling formed a liquid–liquid biphasic mixture. Xiao observed that these Rh catalysts exhibited both high activity (TOF 136 h1) and regioselectivity (99%) in the hydroformylation of 1-decene. However, catalyst recycling proved to be problematic because of the continuous loss of the fluorous phase into the organic solvents over three cycles.

Scheme 6.11

Thermomorphic processes are typically carried out as batch reactions. However, a benchscale continuous phase process using thermomorphic solvents has been reported.71 In this report, hydroformylation of 1-alkenes like octene, dodecene, and tetradecene can reportedly be carried out under monophasic conditions with recovery of the catalyst in the denser phase on cooling. Rhodium losses were reportedly below 0.1 ppm in this process as measured by atomic absorption spectroscopy. The process termed ‘thermoregulated phase-transfer catalysis’ (TRPTC) was invented by Jin’s group and first applied to Rh-catalyzed hydroformylation.17–19,72–75 As originally described, this process used an aqueous two phase system with a Rh catalyst bound to a poly(oxyalkene) phosphine ligand. This polymer-bound catalyst was designed to circumvent problems encountered in commercial aqueous biphasic hydroformylation that use a sulfonated phosphine ligated Rh catalyst. In those cases, larger alkenes are unsuitable because of solubility limitations in the aqueous phase. Such solubility considerations are not a problem in TRPTC systems because the polymer-bound catalysts’ inverse temperature-


129

dependent solubility makes them soluble in the organic phase where larger alkenes are soluble. In these cases, cooling changes the catalysts’ solubility so that it returns to the aqueous phase once the reaction mixture is cooled below the poly(alkene oxide) ligands’ LCST. An early example of this chemistry is the thermoregulated hydroformylation of styrene by a catalyst formed in situ using 18 and Rh(acac)(CO)2 in a mixture of heptane and water (Scheme 6.12).74 Later work showed that hydroformylation of alkenes like 1-decene using catalysts formed in situ with RhCl33H2O and 19 in a toluene–water system at 120 C were similarly effective. After 20 cycles, the aldehyde yield and the TOF of 99% and 198 h1 only decreased by about 5% indicating that the catalyst formed has excellent stability and recyclability.75

Scheme 6.12

6.5 Hydroaminations Under Thermomorphic Conditions Thermomorphic hydroamination of 1-octene to normal and branched derivatives of an N-nonyl morpholine can be accomplished using the same solvent mixtures that effect thermomorphic hydroformylations (Scheme 6.13).76 Conversions of 1-octene can be as high as 98%. Turnover numbers of ca. 900 are possible with Rh losses that are typically < 1% in reactions that use a solvent mixture consisting of propylene carbonate, dodecane, and a polar solvent (e.g. N-ethylpyrrolidone). However, while hydroformylation catalysts were generally recyclable, these reactions that proceed via enamines of the intermediate hydroformylation aldehyde products are complicated by the fact that the propylene carbonate is too reactive an electrophile to be fully compatible with the secondary amine.

Scheme 6.13

130


6.6 Pd-catalyzed Reactions Under Thermomorphic Conditions 6.6.1 Pd-catalyzed Allylic Substitution Under Thermomorphic Conditions The use of thermomorphic ethylene oligomer-supported catalysts is also effective in Pd catalysis. Our group had early on showed that oligomeric polyethyldiphenylphosphine was a very effective polymer in recycling a tetrakis(polyethyldiphenylphosphine)palladium(0) catalyst (20) in allylic substitution chemistry. If care were taken to avoid adventitious oxidation during the reaction, this catalyst could be used up to ten times in allylic substitution chemistry of allyl acetates by secondary amines. Pd catalysts like 20 were equally effective in decarboxylative rearrangements of allyl esters as shown in Scheme 6.14.36 Good yields of an allyl-substituted b-ketoester were even obtained when unactivated allyl alcohol was used as a substrate with ethyl acetoacetate. ICP analyses of the filtrates from product solutions after thermomorphic precipitation of the catalyst 20 showed that no detectable Pd was present in the product solution showing that < 0.001% of the Pd in the catalyst was lost through leaching.

Scheme 6.14

Pd(II) complexed by a polyethyldiphenylphosphine ligand has also been prepared as a thermomorphic catalyst for dimerization of dienes. When the Pd complex 21 was prepared and compared to (Ph3P)2Pd(OAc)2 in a dimerization of butadiene (Scheme 6.15), the thermomorphic catalyst was fully recoverable. While this catalyst was very susceptible to adventitious oxidation, it could be reused up to four times. Catalyst leaching in these reactions was not detected.36

Scheme 6.15

A nonpolar poly(N-octadecylacrylamide) (PNODAM)-bound Pd(0) catalyst 22 was prepared from 6 as shown in Scheme 6.4 using (Ph3P)4Pd as a Pd source. The Pd species formed in this way, like analogous PNODAM-bound Rh(I) hydrogenation catalyst 7, was used as a separable promoter in the thermomorphic allylic substitution chemistry shown in Scheme 6.16. This Pd catalyst was successfully used in five cycles of this allylic substitution reaction with a thermomorphic mixture of heptane and 90% EtOH water. In this case the catalytic species was fully soluble in the miscible solvent mixture at 70 C, but recoverable as a heptane solution at room temperature.29


131

Scheme 6.16

Kaneda’s group was the first to show that dendrimer-bound catalysts could be used and separated efficiently under thermomorphic conditions.77 This study used allylic substitution of E-cinnamyl acetate with dibutylamine in a heptane–DMF mixture catalyzed by a phosphinated poly(propylene imine) (PPI) dendrimer-bound Pd(0) catalyst (Scheme 6.17). Catalyses were carried out at 75 C and with a liquid–liquid separation of the dendrimerbound catalyst from the product. Less than 0.1 ppb Pd was in the heptane phase and the catalyst was recycled four times.

Scheme 6.17

Kaneda’s group used a third-generation PPI dendrimer whose peripheral amine groups were converted into a amides with decanoyl chloride to ionically bind 4-diphenylphosphinobenzoic acid.78 The internal ionically bound phosphines were then allowed to react with a Pd(II) salt to form a Pd-complex that was reduced to form Pd(0). The resulting catalyst was then used to convert a cinnamyl ester into a tertiary amine with piperidine. It was successfully recycled four times in the heptane phase of a thermomorphic heptane– DMF mixture. 6.6.2 Pd-catalyzed Cross-coupling Reactions Under Thermomorphic Conditions Pd-catalyzed cross-coupling reactions like the Heck, Suzuki, or Sonagishira reactions are the reactions that have been studied the most under thermomorphic conditions. Presumably this reflects the fact that these procedures are used in batch type reactions to prepare intermediates in the synthesis of specialty chemicals where catalyst–product separation is important.

132


The first example where Pd-catalyzed cross-coupling chemistry was effected under thermomorphic conditions used either the PNIPAM-bound Pd(0) complex 24, the PNIPAMbound SCS–Pd(II) complex 25, or the PEG-bound SCS–Pd(II) complex 26.29,30,79 These Pd complexes were all shown to be effective catalysts or precatalysts in Heck, Suzuki, and Sonogashira coupling reactions like those shown in Scheme 6.18.

Scheme 6.18


133

Chemistry that used the complex 26 as the source of Pd were equally successful in thermomorphic solvent systems consisting of either 90% EtOH–H2O–heptane or N,N-dimethylacetamide (DMA)–heptane. In the case of 90% EtOH–H2O, the reactions were carried in a miscible solution at 70 C and the polymer-bound Pd catalyst was recovered in the polar solvent phase at room temperature. In the DMA–heptane system, a higher reaction temperature (90 C) was needed to produce a single phase reaction mixture. Recycling was successful through three cycles in the case of the Heck and Suzuki coupling reactions in Scheme 6.18. The principle limitation in recycling experiments using the Pd complex 24 was adventitious oxidation of the phosphine ligand. Reactions with SCS–Pd complexes could be carried out in air and adventitious oxidation was not a problem. Later studies conclusively showed that the actual catalysts in reactions with 25 or 26 are not the SCS–Pd(II) species but rather Pd colloids or Pd nanoparticles formed in situ,80–82 both precatalysts 25 and 26 could be recycled under thermomorphic conditions. In the case of 25, thermomorphic recycling experiments that recovered the precatalyst required use of a solvent mixture consisting of 10% aqueous DMA and heptane. The reaction occurred under monophasic conditions at 95 C with a biphasic liquid–liquid separation of the product and recovered 25 at 25 C. No visual catalyst decomposition was seen in experiments where the precursor 27 had been completely converted to the SCS– Pd(II) complex 25. However, later work has since established that the actual catalyst is not the SCS–Pd(II) complex.80–82 The most effective recycling of 26 used PEG starting materials for synthesis of 26 that were first extracted using a continuous liquid–liquid extractor.79 In these cases, heptane extraction of a 50–50 EtOH–heptane (vol/vol) solution of the starting PEG removed lower molecular weight PEG species that are not as phase selectively soluble as higher molecular weight PEG derivatives. In the case where the PEG-bound Pd complex 26 was used, an equivolume mixture of heptane and DMA did not phase separate at room temperature. To use 26 under thermomorphic conditions required using a 1:2 mixture of heptane and DMA. Similar effects of substrates, products or polymers on the phase separability or critical solution temperature of a thermomorphic solvent mixture have also been noted by Behr.10,11 SCS–Pd(II) complexes containing PEG groups attached to the chelating sulfur groups of the SCS species have also been used under thermomorphic conditions. In this case, the Pd complex was recovered in the polar phase by cooling after a reaction in which microwave heating was used to produce monophasic conditions.30 Such heating was a more efficient way to generate a catalytically competent monophasic solution. Pd nanoparticles are again the actual catalyst. Pd leaching was nonetheless minimal with ICP analyses showing the Pd concentration of the heptane product phase to be very low in a reaction like that shown in Scheme 6.19. The temperature-dependent miscibility of fluorous-labeled materials has been used by Gladysz in reactions of the fluorous-labeled Pd complexes that are insoluble at room temperature in DMF but soluble on heating. Complexes like 29 and 30 were used in this way in Heck and Suzuki chemistry by Rocaboy and Gladysz.83,84 However, while these fluorous species have clear temperature-dependent solubility, the actual catalyst (as was true for species like 25 and 26) is believed to be palladium nanoparticles formed in situ. A thermomorphic solid–liquid Heck catalyst 31 containing a fluorous bipyridyl ligand was recently reported by Lu and used in Heck chemistry (Scheme 6.20).85 This species was

134


Scheme 6.19

Scheme 6.20

insoluble at room temperature in DMF but soluble at 140 C. Less than 0.02% of the charged Pd was lost in recycling this catalyst. This reaction, like those discussed above, appears to involve some Pd nanoparticle formation though the authors claimed that catalysis due to Pd nanoparticles was a minor effect. Polymeric phosphine ligands 32 useful in Pd-catalyzed cross-coupling chemistry can also be prepared by a ring-opening metathesis polymerization (ROMP) (Scheme 6.21).86 While the presence of excess phosphine ligands in this synthesis inhibits the Ru-catalyzed ROMP process, the polymeric ligand once formed was useful in immobilizing Pd catalysts for a Heck reaction of methyl acrylate and iodobenzene in DMF at 80 C. These ligands and the Pd catalyst could be recovered and recycled by simply cooling the reaction mixture as the polymeric catalyst was insoluble at room temperature. Though adventitious oxidation eventually interfered in thermomorphic solid–liquid recycling experiments, these catalysts were successfully reused five times.


135

Scheme 6.21

Thermomorphic conditions have also been used in a quite different fashion to prepare hydrophobic product libraries by Heck, Suzuki and Sonogashira reactions by the Chiba group (Scheme 6.22).87 In this chemistry, a phase tag like 33 that is soluble in cyclohexane or methylcyclohexane was used to prepare benzoate esters or thioether-bound aryl boronic acids. Cross-coupling chemistry on these substrates was carried out under thermomorphic conditions with a conventional Pd catalyst using DMF as the polar solvent. The products were then separated from the polar-phase soluble catalyst as cycloalkane solutions. Subsequent chemistry then afforded a typical organic product free of the cycloalkane phase tag. For example, a benzoate ester containing an aryl halide was used in a Sonogashira reaction (Scheme 6.22) to form an alkyne derivative that was subsequently cleaved from

Scheme 6.22

136


the cycloalkane-soluble phase tag to form a heterocycle. A similar approach using Heck or Suzuki coupling reactions (Scheme 6.22) led to cinnamate esters bound to the phase tag or biaryl thioethers on hydrocarbon phase tags. Since the products are separated from the DMF phase containing the catalyst, the catalyst is reusable. Plenio’s group has carried out an extensive set of studies on cross-coupling chemistry with a variety of Pd catalysts often under thermomorphic conditions. Much of this group’s earliest work in thermomorphic catalysis involved using DMSO as a polar phase with heptane as the nonpolar phase. These studies included chemistry where a Pd catalyst was ligated by an ordinary triarylphosphine-modified polar phase soluble PEG (37), where a hindered-phosphine attached to a polar phase soluble PEG polymer was used to ligate Pd (38), or cases where a nonpolar phase soluble poly(4-methylstyrene) polymer was used to support a hindered phosphine 39 and the Pd catalyst it forms. These polymersupported Pd complexes all were successfully used in Sonogashira alkyne–arene couplings (Schemes 23–25).88–90

Scheme 6.23

Scheme 6.24

Scheme 6.25

While Plenio’s group has often carried out monophasic reactions in DMSO with heptane extractions of product,91,92 Plenio’s group has also reported using the PEG-supported phosphine ligand 37 in Pd-catalyzed Sonogashira couplings in a thermomorphic solvent


137

mixture formed from a 5:2:5 (v:v:v) of CH3CN, Et3N, and heptane (Scheme 6.23).88 This catalyst had modest activity and only coupled aryl iodides and acetylenes. In this case Plenio’s group observed that the added Et3N affected the temperature-induced miscibility between CH3CN and heptane. By introducing 20 vol% Et3N, full miscibility was achieved by heating to 80 C. In the absence of this additive, CH3CN and heptane do not achieve full miscibility at this temperature. Since Et3N is converted to an ammonium salt which accumulates in the polar phase, this initial miscibility could not be replicated in recycling experiments because the accumulated ammonium salt affects the phase behavior of the solvent mixture. Such effects are well known and often affect biphasic catalyst systems, especially if byproducts accumulate in the catalyst-carrying phase. In this instance, the problem was solved by using K2CO3 as a base. Apparently the byproduct water and potassium iodide had a smaller effect on subsequent miscibility of the thermomorphic solvent mixture than the ammonium salt. Unlike others’ work that has often focused on the separation strategies, Plenio’s work often focused more on optimizing the catalysis chemistry than on optimizing the separation chemistry. An example of this is the use of hindered phosphine-ligated PEG-bound Pd catalysts to effect Sonogashira coupling on less reactive aryl bromides (Scheme 6.23).88 Unreactive aryl bromides were suitable substrates for Sonogashira coupling reactions conducted in DMSO–heptane with this more hindered polymer-bound phosphine ligand. This polymer-bound catalyst was successfully recycled through five cycles with overall yields of >90% with phenyl acetylene as the alkyne. Both leaching of ligand and Pd into the heptane phase was examined. The absence of PEG 1 H NMR spectra peaks and X-ray fluorescence analyses (XRF) for Pd and Cu showed the ligand 37, Pd, and Cu all stayed in the DMSO phase (Scheme 6.23). Plenio’s group also reported using poly(4-methylstyrene)-bound hindered phosphines 39 under thermomorphic conditions using mixtures of DMSO or nitromethane with cyclohexane in Sonogashira or Suzuki couplings (Scheme 6.25).89–90 In the Sonogashira chemistry, the aryl bromide substrates included –COCH3, –CH3, –Cl, –H and –OCH3 substituents with phenylacetylene or 1-octyene as the alkyne. In the Suzuki chemistry, Pd(OAc)2 was used as the Pd source, K3PO4 as the base, phenylboronic acid, and aryl bromides and chlorides with –COCH3, –H, –CN, and –OCH3 substituents. In either case, the Pd catalyst ligated by the poly(4-methylstyrene)-bound hindered phosphine was recovered in the cyclohexane phase through five cycles. Pd leaching into the polar phase was less than 0.2% of the charged Pd. In more recent work, Plenio’s group has focused its attention on catalytic reactions like those in Scheme 6.26 or 6.27 that are carried out in aqueous–alcohol solvent mixtures.93,94

Scheme 6.26

138


Scheme 6.27

Such solvent mixtures are separable liquid–liquid mixtures at room temperature but are miscible at elevated temperature.95 For example, n-butanol–water is biphasic at ambient temperature but becomes miscible at 49 C. In such separations, a catalyst bound to a sulfonated phosphine like 42 remains in the aqueous phase after cooling so it can be recycled. Such systems have a number of advantages as thermomorphic systems. First, solvents like 2-propanol or 1-butanol are biodegradable. Second, the more polar solvent mixture may have advantages in the synthesis of very polar heterocyclic compounds of interest to the pharmaceutical industry. As noted above, some examples of thermomorphic Pd catalysis of cross-coupling chemistry likely involve Pd nanoparticles as the actual catalytic species. The low losses of Pd in these examples suggest that thermomorphic systems containing polymer supports can be used to ligate, stabilize, and recycle catalytically active nanoparticles. This premise is supported both by work by the Beletskaya group and by Kanada’s work on dendrimerbound nanoparticle catalysts for hydrogenation and allylic substitution.45,77,78,96 The Beletskaya group, for example, prepared polymer-stabilized Pd(0) nanoparticle suspensions that were recyclable catalysts for both Heck coupling reactions and other Pd(0)-catalyzed reactions. In this work, the Pd nanoparticles were formed in a dendrimerlike polymeric micelle formed from a polystyrene–poly(ethylene oxide) block copolymer that contained a cetylpyridinium surfactant sorbed into the hydrophobic micellar core (41). The Pd nanoparticles in these micelles were stable for over a year and were shown to be competent catalysts for thermomorphic Heck couplings (Scheme 6.28) with activity like that of a conventional Pd(0) catalyst. These (PEO–PS–Pd(0)) colloids 41 were useful in a thermomorphic mixture of 90% aqueous DMA–heptane (1–2; v/v) through three cycles with yields of cinnamate products of 90, 86 and 94%. Heck chemistry with aryl bromides worked but longer reaction times and higher temperatures were required.

6.7 Polymerization Reactions Under Thermomorphic Conditions Separation of polymerization catalysts from a polymer is commonly not a problem because catalyst turnover numbers are often very high and the product polymer’s performance is not affected by trace catalyst residue. However, in some cases catalyst separation is an issue and thermomorphic separations have been employed in these cases. The thermomorphic solubility of catalysts ligated by multiple octadecyl hydrocarbon groups or of a catalyst bound to polyethylene have both been used to effect solid–liquid separations of catalysts and polymer products. The interest in catalyst–polymer separation in polymerizations arises when catalyst contamination interferes with a polymer’s eventual use or application. In the case of atom transfer radical polymerization (ATRP), the initial example of these controlled radical


139

Scheme 6.28

polymerizations required relatively high levels of a Cu(I) catalyst. In these cases, copper contamination of the products was a potential problem in cases where the ‘designer’ polymer products from these reactions were used in biological applications or where the products were used in electronic applications. ATRP chemistry has since advanced and Cu catalyst can now be reduced to very low levels with suitable ligands.97 However, the interest in reducing Cu contamination in polymer products led to several thermomorphic approaches for catalyst–product separation.98–101 The first example of this chemistry employed polyethylene oligomers that had previously been used as separable, recyclable catalysts in diene polymerization32 to support Cu(I)– amine catalysts for ATRP chemistry. The Brittain group showed that polyethylene-bound catalyst 42 could be separated from polyacrylate products in an ATRP polymerization.98 However, some of the control of the polymerization process that makes ATRP chemistry so attractive was absent. Subsequent work by Zhu’s group showed that a polyethylene oligomer-bound catalyst 43 that contained a PEG group afforded better control in an ATRP polymerization process.99 An even simpler thermomorphic approach was described by the Vincent group in 2004 who showed that a tren ligand with six octadecyl groups could be used to prepare the

140


Scheme 6.29

CuBr–tren ATRP catalyst 44.100 Vincent’s group showed that this catalyst afforded polymer products with excellent control of polymer dispersity (PDI for poly(methyl methacrylate) of 1.11) and with control of end group functionality. The catalyst complex 44 exhibited thermomorphic solubility behavior in 1,4-dioxane and was soluble under the polymerization conditions but insoluble on cooling. In these cases, a high level of catalyst recovery was reportedly attained.100 Strategies that use the thermomorphic differences in solubility of products in ATRP polymerization have also been described. In this chemistry, a soluble polyisobutylene (PIB)-bound Cu(I) catalyst complex 46 formed by Fokin–Huisgen ‘Click’ cyclization of an alkyne and polymeric azide 45 was used to form polystyrene in a mixture of styrene and heptane (Scheme 6.30). On cooling, the product polystyrene precipitated as a suspension, self separating from the solution of the PIB-bound catalyst.101

Scheme 6.30


141

Thermomorphic fluorous ligands for Cu(I) catalysts for ATRP chemistry have also been described102 using the fluorous ligand used to prepare a Pd catalyst (31) for Heck chemistry. This thermomorphic CuBr complex 47 catalyzed ATRP polymerization of methyl acrylate (Scheme 6.31). PDI values for the product polymer were in the 1.26–1.41 range. While the catalyst 47 was reportedly insoluble at room temperature, actual catalyst–product separations were carried out at 10 C. Residual Cu levels in the poly(methyl methacrylate) were in the 15–40 ppm range.

Scheme 6.31

‘Commercially viable’ thermomorphic polyethylene-supported Co(III)–porphyrin catalysts for polymerization of styrene or methacrylate have recently been described by DuPont.103 In this work (Scheme 6.32), the DuPont group was able to use functional polyethylene polymers available from Baker–Petrolite43 containing a terminal primary

Scheme 6.32

142


hydroxyl groups to form mesylates that were in turn used to alkylate a phenolic group of a porphyrin 48. Metal exchange then formed the Co(III) porphyrin catalyst 49 that was analogous to earlier Co–porphyrin catalysts but that contained four polyethylene groups. Such catalysts have thermomorphic solubility – they are soluble hot in the polymerization milieu but quantitatively precipitate on cooling. Such catalysts can be practically separated from the polymer product and the recovered catalyst is reusable. An important feature of this chemistry is that these catalysts, like those described by Vincent for Cu(I)-catalyzed ATRP chemistry, are not any different from the existing catalysts in terms of their ability to effect and control a polymerization. However, they are separable. This is important in the DuPont case because the chromogenic porphyrin catalyst contamination in product from a more conventional nonthermomorphic catalyst poses esthetic color problems in coatings (e.g. automotive paints) that use these polyacrylates.

6.8 Organocatalysis Under Thermomorphic Conditions Solid–liquid thermomorphic separations are also feasible in fluorous systems. Several strategies have been described. Gladysz’ group was the first to note that trialkylphosphines with heavily fluorinated ligands are insoluble in organic solvents at room temperature but soluble hot.104,105 Such catalysts mimic the solubility behavior of oligomeric alkanes noted above. As was true for those systems, fluorous systems can have dramatic solubility/ temperature dependence wherein greater than 100-fold differences are seen in solubility for a hot versus cold solution. Solubility changes or phase changes of this magnitude should be generally useful forcatalystrecoveryandGladysz’ grouphasused thisconcept invariousways since his group’s original discovery of this process. These studies were recently reviewed20 and selected examples of such thermomorphic catalysts are briefly summarized here. Gladysz has since expanded his work to use fluorinated solid supports to sorb a fluorous catalyst on cooling.106,107 In this scheme, the fluorous catalyst is soluble under the higher temperature reaction conditions, but on cooling, the catalyst is deposited on the always insoluble but inert fluorinated solid support. Introducing a desired amount of catalyst in such systems can be as simple as adding a specific length of Teflon tape since the catalyst is uniformly deposited on the fluorinated solid support. Recovery of the catalyst involves a simple decantation. Gladysz’ thermomorphic fluorous catalysts recoverable either as solids or on fluorinated solid supports have been used in various reactions. The original example was a phosphinecatalyzed organocatalytic reaction involving addition of alkoxy groups to methyl propiolate (Scheme 6.33).104 In an octane solution of methyl propiolate at room temperature, the fluorous phosphine 50 was insoluble. However, this phosphine dissolved in this solution on heating to 65 C at which point the reaction took place. Cooling to 30 C precipitated the catalyst 50. The catalyst 50 could also be sorbed onto Teflon beads or shavings. 31 P NMR analysis suggested that at room temperature ca. 99.6% of 50 was insoluble in octane.

Scheme 6.33


143

However, later studies indicated that 2–3% of the catalyst leached into octane in some form.105 Extension of the organocatalytic reactions of thermomorphic fluorous phosphines recoverable on fluorinated solvents to a variety of intramolecular cyclizations was recently reported. In these examples, the fluorous phosphine was again recovered this time by precipitation, with Teflon tape, or with Gore–Rastex fiber by cooling to 30 C after a homogeneous cyclization at 60–70 C (Scheme 6.34).107

Scheme 6.34

Thermomorphic polydimethylsiloxane supports that are useful in organocatalysis have also been described by our group.108 While the reported catalytic studies of quininecatalyzed Michael additions of thiols to a,b-unsaturated carbonyl compounds actually used a different catalyst recovery scheme, we were able to use dye-labeled siloxane polymers like 51 to establish that these nonpolar soluble polymer supports can be quantitatively recovered as heptane solutions in thermomorphic mixtures of heptane and DMF. Thermomorphic dendritic organocatalysts useful in Baylis–Hillman reactions of aromatic aldehydes and methyl vinylketone or acrylonitrile have been described by Yang.109 These catalysts are dendritic forms of 4-(N,N-dimethylamino)pyridine (DMAP). However, unlike DMAP, the dendrimer 52 has thermomorphic solubility. It is 97.5% phase selectively soluble in cyclohexane in a DMF–cyclohexane solvent mixture at room temperature. However, it is also soluble and active as a catalyst in reactions like those shown in Scheme 6.35. However, in recycling 52 in the reaction of 4-nitrobenzaldehyde and methyl

Scheme 6.35

144


Scheme 6.36

vinyl ketone, the reaction yields significantly decreased from 92.3% to 47.6% after two cycles. This catalyst deactivation was ascribed to formation of pyridinium salts. Treatment of the deactivated catalyst with 2 M NaOH at 60 C for 2 h regenerated the catalyst. Five reaction cycles were reported with yields of >90% per cycle when the catalyst was regenerated by treatment with base. Gladysz recently described using Teflon tape to recycle fluorous phase transfer catalysts.110 Such species behave like classical phase transfer catalysts with typical ionic reactions occurring even in a very nonpolar perfluorodecalin solution (Scheme 6.36). Reactions were effected both with fluorous–aqueous biphasic systems and in triphasic systems.

6.9 Cu(I)-catalyzed 1,3-Dipolar Cycloadditions Under Thermomorphic Conditions Cu(I)-catalyzed cycloadditions of azides and alkynes have developed into one of the most cited examples of ‘click’ chemistry since their discovery by Fokin. Thermomorphic schemes have been reported that effect efficient separation of the Cu(I) catalyst from the product. In this case,111 the Cu(I) catalyst 44 used in ATRP polymerization (Scheme 6.29) is insoluble cold but soluble hot. In solution, 44 catalyzes the azide–alkyne cycloaddition shown in Scheme 6.37 at elevated temperature. The Cu(I) complex can later be recovered by filtration for reuse by simply cooling the reaction mixture.

Scheme 6.37

6.10 Thermomorphic Hydrosilylation Catalysts The protocols Gladysz’ group used for organocatalysis were fluorous catalysts which are recovered using an inert fluorinated solid support and have also been used in transition metal catalysis.105,112 For example, the Wilkinson’s catalyst analog 53 containing a fluorous phosphine has been successfully recycled under thermomorphic conditions in hydrosilylation chemistry. In a reaction like that in Scheme 6.38 this catalyst was

Scheme 6.38


145

successfully recycled through three cycles by heating the reaction mixture up to 65 C for the reaction and then recovering 53 on some Teflon tape at room temperature. The catalyst deactivation in the fourth cycle was attributed to catalyst deactivation as opposed to catalyst leaching since the TON values for cycles 1–3 were essentially constant (658 8 mol of product/mol of catalyst). Some leaching of the phosphine was observed by a 19 F NMR analysis and AAS-ICP analysis for rhodium that indicated that leaching was as high as 5% in the second cycle.

6.11 Thermomorphic Catalytic Oxidations Catalytic oxidation reactions are very common but oxidation catalysts are less commonly bound to polymers and recycled. There are nonetheless a number of examples of oxidation catalysts have been used under thermomorphic conditions. As noted above, fluorous ligands for metal complexes can be used under biphasic or thermomorphic conditions. This has been used to advantage in oxidation chemistry in a collaborative effort of the Fish, Contel, and Vincent groups.113 Starting with a fluorinated carboxylate ligand, they prepared dimeric Cu(II) complexes that could be further modified with the addition of a chelating ligand to form a thermomorphic Cu(II) complex 54 that was useful as a catalyst in oxidation of allylic or benzylic alcohols using 2,2,6,6-tetramethylpiperidinoxyl (TEMPO) (Schemes 6.39 and 6.40). Using 54 they transformed 4-nitrobenzyl alcohol into a 4-nitrobenzaldehyde in a chlorobenzene–toluene mixture using oxygen as the penultimate oxidant. This Cu(II) complex 54 that was used as the catalyst exhibited thermomorphic character and had little or no solubility in chlorobenzene at room temperature but was soluble at 90 C. Three cycles were carried out successfully with recovery of ca. 80–90% of 54 after each cycle.

Scheme 6.39

Scheme 6.40

146


Fluorous bipyridyl ligands used in Heck catalysis and in ATRP chemistry have also been used in Cu-mediated catalytic oxidations of alcohols to ketones using TEMPO. In this report,114 quantitative yields in oxidations of primary alcohols were obtained in reactions carried out with 10 mol% TEMPO cocatalyst at 80 C in chlorobenzene using the Cu(I)–bipyridyl catalyst 47 (Scheme 6.41). Cooling precipitated this catalyst which was then separated and reused through up to eight cycles (recycling was only examined in oxidation of 4-nitrobenzyl alcohol).

Scheme 6.41

A thermomorphic solid–liquid catalyst based on poly(N-isopropylacrylamide) (PNIPAM) – a polymer that exhibits LCST behavior in water – has been described by Ikegami.115 This catalyst 55 was derived from a copolymer of PNIPAM that contained ca. 6 mol% of an amphiphilic ammonium salt. This ammonium salt was used to bind a phosphotungstic acid salt to the thermally responsive polymer. The resulting catalyst was shown to be to undergo LCST behavior and to form emulsions at 80–90 C in water. The emulsions that form were effective in catalyzing the oxidation of 1-phenyl-1-propanol to the corresponding ketone (Scheme 6.42). The catalyst was recycled three times with no change in yield of the product. Recycling was effective for a range of primary and secondary alcohols. The ca 0.1 mol% of the catalyst 55 was recoverable by cooling. Products were extracted from the product using diethyl ether.

Scheme 6.42

The temperature responsive oxidation catalyst 55 has also been used in ‘smart’ oxidative cyclization reactions using hydrogen peroxide as the oxidant.8 In these cases, the starting reaction mixture is a mixture of the ionically cross-linked polymer, the oily substrate, and water containing hydrogen peroxide. At 60 C, the ionically cross-linked polymer DD dehydrates and becomes more hydrophobic. This hydrophobic gel sorbs the substrate which


147

Figure 6.4 Thermomorphic gel phase oxidative cyclization using a thermally responsive polymer bound phosphotungstate oxidant and hydrogen peroxide

is oxidatively cyclized as shown in Figure 6.4. In this process, gel phase concentrates the substrate and catalyst for the oxidative cyclization reaction shown in Scheme 6.43. After the reaction is complete, the reaction mixture was cooled. The catalyst 55 rehydrates and releases the hydrophobic product which was recovered by a diethyl ether extraction. Three cycles of this oxidative cyclization reaction were successful using the substrate 4-penten1-ol. Studies with a variety of carboxylic acids, phenols, or alcohols containing g or d alkene groups showed that the oxidative cyclization catalyzed by 55 was a general reaction.

Scheme 6.43

6.12 Conclusions The idea of using small temperature changes to affect the solubility or phase location of a catalyst is a general and potentially useful way to increase the utility of many different sorts of homogeneous catalysts. In some cases, it may be possible to use such phase changes to effect new chemistry. More commonly, such phase changes like other multiphase strategies can make catalytic reactions simpler, faster, or can make desired catalyst–product separations more practical.116 Many of the approaches described to date have used polymer supports. However, in many other cases where a polymer is not needed because a simpler group can serve as a phase anchor to either separate a catalyst as a solid from a solution of product or to separate a catalyst in one of two immiscible liquid phases. In many cases, the limitation to the use of thermomorphic conditions is not inefficiency in the temperatureinduced phase change process but rather catalyst instability. Applications of these sorts of thermomorphic systems are only likely to increase as issues of green chemistry and sustainability become more and more important and as the use of multiphase systems in industry increases.

References 1. L.D. Taylor and L.D. Cerankowski, Preparation of films exhibiting a balanced temperature dependence to permeation by aqueous solution. Lower consolute behavior, J. Polym. Sci., Polym. Chem. Ed., 13, 2551–2570 (1975).

148


2. D.E. Bergbreiter, B.L. Case, Y.-S. Liu and J.W. Caraway, Poly(N-isopropylacrylamide) soluble polymer supports in catalysis and synthesis, Macromolecules, 31, 6053–6062 (1998). 3. H.N. Gray and D.E. Bergbreiter, Applications of polymeric smart materials to environmental problems, Environmental Health Perspectives Supplements, 105, 55–63 (1997). 4. D.E. Bergbreiter, V.M. Mariagnanam and L. Zhang, Polymer ligands that can regulate reaction temperature in ‘smart’ catalysts, Adv. Mater., 7, 69–71 (1995). 5. D.E. Bergbreiter, L. Zhang, and V.M. Mariagnanam, Smart ligands that regulate homogeneously catalyzed reactions, J. Am. Chem. Soc., 115, 9295–9296 (1993). 6. X. Huang, K.L. Witte, D.E. Bergbreiter, and C.-H. Wong, Homogenous enzymatic synthesis using a thermo-responsive water-soluble polymer support, Adv. Synth. Catal., 343, 675–681 (2001). 7. D.M. Davies and E.L. Stringer, Smart aqueous reaction medium, Langmuir, 19, 1927–1928 (2003). 8. H. Hamamoto, Y. Suzuki, H. Takahashi and S. Ikegami, A new solid-phase reaction system utilizing a temperature-responsive catalyst: oxidative cyclization with hydrogen peroxide, Adv. Synth. Catal., 349, 2685–2689 (2007). 9. K.-S. Liao, H. Fu, A. Wan, J.D. Batteas, and D.E. Bergbreiter, Designing surfaces with wettability that varies in response to solute identity and concentration, Langmuir, 25, 26–29 (2009). 10. A. Behr, G. Henze, and R. Schomaecker, Thermoregulated liquid/liquid catalyst separation and recycling, Adv. Synth. Catal., 348, 1485–1495 (2006). 11. A. Behr, B. Turkowski, R. Roll, R. Schobel and G. Henze, Multiphase catalysis in temperaturedependent multi-component solvent (TMS) systems, Top. Organomet. Chem., 23, 19–52 (2008). 12. D.E. Bergbreiter and S.D. Sung, Liquid/liquid biphasic recovery/reuse of soluble polymersupported catalysts, Adv. Synth. Catal., 348, 1352–1366 (2006). 13. D.E. Bergbreiter and J. Li, Applications of catalysts on soluble supports, Top. Curr. Chem., 242, 113–176 (2004). 14. D.E. Bergbreiter, Using soluble polymers to recover catalysts and ligands, Chem. Rev., 102, 3345–3383 (2002). 15. C. Hongfa, J. Tian and D.E. Bergbreiter, GGGG, Chem. Rev., 109, (2009) in press. 16. J.H. Hildebrand and D.R.F. Cochran, Liquid–liquid solubility of perfluoromethylcyclohexane with benzene, carbon tetrachloride, chlorobenzene, chloroform, and toluene, J. Am. Chem. Soc., 71, 22–25 (1949). 17. E. Hermanns, J. Hasenjaeger, B. Driessen-Hoelscher, PEG-modified ligands for catalysis and catalyst recycling in thermoregulated systems, Top. Organometal. Chem., 23, 53–66 (2008). 18. Z. Jin, Y. Wang, J. Jiang, and F. Wen, Biphasic catalysis with thermoregulated ligands, Curr. Top. Catal., 3, 15–32 (2002). 19. Z. Jin and X. Zheng, Thermoregulated phase-transfer catalysisin Aqueous-Phase Organometallic Catalysis, B. Cornils and W.A. Herrmann,Eds., Wiley, Weinheim, pp. 233–240 (1998). 20. J.A. Gladysz, and V. Tesevic, Temperature-controlled catalyst recycling: new protocols based upon temperature-dependent solubilities of fluorous compounds and solid/liquid phase separations, Top. Organometal. Chem., 23, 67–89 (2008). 21. D.E. Bergbreiter, Y.-S. Liu, and P.L. Osburn, Thermomorphic rhodium(I) and palladium(0) catalysts, J. Am. Chem. Soc., 120, 4250–4251 (1998). 22. D.E. Bergbreiter, R. Hughes, J. Besinaiz, C. Li, and P.L. Osburn, Phase-selective solubility of poly(N-alkylacrylamide)s, J. Am. Chem. Soc., 125, 8244–8249 (2003). 23. D.E. Bergbreiter, N.A. Aviles-Ramos, and D. Ortiz-Acosta, A combinatorial approach to studying the effects of N-alkyl groups on poly(N-alkyl and N,N-dialkylacrylamide) Solubility, J. Comb. Chem., 9, 609–617 (2007). 24. H.G. Batz, G. Franzmann, and H. Ringsdorf, Pharmacologically active polymers. 3. Model reactions for synthesis of pharmacologically active polymers by way of monomeric and polymeric reactive esters, Angew. Chem., Int. Ed. Engl., 11, 1103–1104 (1972). 25. M. Zhou, A. Sivaramakrishnan, K. Ponnamperuma, Krishan, W.-K. Low, C. Li, J.O. Liu, D.E. Bergbreiter and D. Romo, Synthesis, characterization, and utility of thermoresponsive natural/unnatural product macroligands for affinity chromatography, Org. Lett., 8, 5247–5250 (2006).


149

26. C. Li and D.E. Bergbreiter, Dye-labeled PNIPAM [poly(N-isopropylacrylamide)]-supported acylation catalysts, Chem. Ind., 89, 545–550 (2003). 27. D.E. Bergbreiter and J.W. Caraway, Thermoresponsive polymer-bound substrates, J. Am. Chem. Soc., 118, 6092–6093 (1996). 28. D.E. Bergbreiter, P.L. Osburn, and J.D. Frels, Nonpolar polymers for metal sequestration and ligand and catalyst recovery in thermomorphic systems, J. Am. Chem. Soc., 123, 11105–11106 (2001). 29. D.E. Bergbreiter, P.L. Osburn, A. Wilson, and E.M. Sink, Palladium-catalyzed C–C coupling under thermomorphic conditions, J. Am. Chem. Soc., 122, 9058–9064 (2000). 30. D.E. Bergbreiter and S. Furyk, Microwave-promoted Heck reactions using an oligo(ethyleneglycol)-bound SCS palladacycle under thermomorphic conditions, Green Chem., 6, 280–285 (2004). 31. D.E. Bergbreiter and J.R. Blanton, Functionalized ethylene oligomers as phase-transfer catalysts, J. Org. Chem., 50, 5828–5833 (1985). 32. D.E. Bergbreiter, L.-B. Chen and R. Chandran, Recyclable polymer-bound lanthanide diene polymerization catalysts, Macromolecules, 18, 1055–1057 (1985). 33. D.E. Bergbreiter and R. Chandran, Concurrent catalytic reduction/stoichiometric oxidation using oligomerically ligated catalysts and polymer-bound reagents, J. Am. Chem. Soc., 107, 4792–4793 (1985). 34. D.E. Bergbreiter and R. Chandran, Use of functionalized ethylene oligomers to prepare recoverable, recyclable nickel(0) diene cyclooligomerization catalysts, J. Org. Chem., 51, 4754–4760 (1986). 35. D.E. Bergbreiter and R. Chandran, Polyethylene-bound rhodium(I) hydrogenation catalysts, J. Am. Chem. Soc., 109, 174–179 (1987). 36. D.E. Bergbreiter and D.A. Weatherford, Polyethylene-bound soluble recoverable palladium(0) catalysts, J. Org. Chem., 54, 2726–2730 (1989). 37. D.E. Bergbreiter, B. Chen, and D. Weatherford, New strategies in using macromolecular catalysts in organic synthesis, J. Mol. Catal., 74, 409–419 (1992). 38. D.E. Bergbreiter and S.A. Walker, Soluble polyethylene- and polystyrene-bound tin halides as catalysts for reductions of alkyl and aryl bromides and iodides by sodium borohydride, J. Org. Chem., 54, 5138–5141 (1989). 39. J.C. Phelps, D.E. Bergbreiter, G.M. Lee, R. Villani and S.M. Weinreb, A polyethylene-bound ruthenium(II) catalyst for inter- and intramolecular Kharasch reactions, Tet. Lett., 30, 3915–3918 (1989). 40. D.E. Bergbreiter and D.R. Treadwell, Polyethylene carboxylate-bound triruthenium clusters as alcohol oxidation catalysts, Reac. Polym., 12, 291–295 (1990). 41. D.E. Bergbreiter, J.R. Blanton, R. Chandran, M.D. Hein, K.J. Huang, D.R. Treadwell, and S.A. Walker, Anionic syntheses of terminally functionalized ethylene oligomers, J. Polym. Sci., Part A: Polym. Chem., 27, 4205–4226 (1989). 42. K.J. Shea, Polyhomologation: the living polymerization of ylides, Chem.- Eur. J., 6, 1113–1119 (2000). 43. Baker–Petrolite sells terminally functionalized polyolefins with hydroxyl- and carboxylic acid groups under the tradenames UNILIN and UNICID. 44. Y.Y. Huang, Y.M. He, H.F. Zhou, L. Wu, B.L. Li, and Q.H. Fan, Thermomorphic system with non-fluorous phase-tagged Ru(BINAP) catalyst: Facile liquid/solid catalyst separation and application in asymmetric hydrogenation, J. Org. Chem., 71, 2874–2877 (2006). 45. M. Murata, Y. Tanaka, T. Mizugaki, K. Ebitani, K. Kaneda, Palladium–platinum bimetallic nanoparticle catalysts using dendron assembly for selective hydrogenation of dienes and their application to thermomorphic system, Chem. Lett., 34, 272–273 (2005). 46. R.M. Crooks, M. Zhao, L. Sun, V. Checkhik, L.K. Yeung, Dendrimer-encapsulated metal nanoparticles: synthesis, characterization, and applications to catalysis, Acc. Chem. Res., 34, 181–190 (2001). 47. N. Niu and R.M. Crooks, Dendrimer-encapsulated metal nanoparticles and their applications to catalysis, Comptes Rendus Chim., 6, 1049–1059 (2003).

150


48. R.W.J. Scott, O.M. Wilson, and R.M. Crooks, Synthesis, characterization, and applications of dendrimer-encapsulated nanoparticles, J. Phys. Chem., B, 109, 692–704 (2005). 49. W.J. Shaw, J.C. Linehan, A. Gutowska, D. Newell, T. Bitterwolf, J.L. Fulton, Y. Chen, and C.F. Windisch, Synthesis and characterization of a recoverable rhodium catalyst with a stimulus sensitive polymer ligand, Inorg. Chem. Commun., 8, 894–896 (2005). 50. T.S. Huang, Y.H. Wang, J.Y. Jiang, Z.L. Jin, PEG-stabilized palladium nanoparticles: An efficient and recyclable catalyst for the selective hydrogenation of 1,5-cyclooctadiene in thermoregulated PEG biphase system, Chin. Chem. Lett., 19, 102–104 (2008). 51. Y. Wang, X. Wu, F. Cheng, and Z. Jin, Thermoregulated phase-separable phosphine ruthenium complex for hydrogenation catalysis, J. Mol. Catal. A.: Chem., 195, 133–137 (2003). 52. A. Behr, D. Obst, C. Schulte, and T. Schosser, Highly selective tandem isomerization–hydroformylation reaction of trans-4-octene to n-nonanal with rhodium–BIPHEPHOS catalysis, J. Mol. Catal., A: Chem., 206, 179–184 (2003). 53. A. Behr, D. Obst, and B. Turkowski, Isomerizing hydroformylation of trans-4-octene to n-nonanal in multiphase systems: acceleration effect of propylene carbonate, J. Mol. Catal., A: Chem., 226, 215–219 (2005). 54. A. Behr, G. Henze, B. Turkowski, and D. Obst, Selection process of new solvents in temperaturedependent multi-component solvent systems and its application in isomerising hydroformylation, Green Chem., 7, 645–649 (2005). 55. M.S. Shaharun, H. Mukhtar, and B.K. Dutta, Solubility of carbon monoxide and hydrogen in propylene carbonate and thermomorphic multicomponent hydroformylation solvent, Chem. Eng. Sci., 63, 3024–3035 (2008). 56. J. Tijani and B. El Ali, Selective thermomorphic biphasic hydroformylation of higher olefins catalyzed by HRhCO(PPh3)3/P(OPh)3, Appl. Catal. A: General, 303, 158–165 (2006). 57. I.T. Horvath and J. Rabai, Facile catalyst separation without water: fluorous biphase hydroformylation of olefins, Science, 266, 72–75 (1994). 58. E. de Wolf, G. van Koten, B.-J. Deelman, Fluorous phase separation techniques in catalysis, Chem. Soc. Rev., 28, 37–41 (1999). 59. W. Zhang, Fluorous Synthesis of heterocyclic systems, Chem. Rev., 104, 2531–2556 (2004). 60. H.R. Hobbs and N.R. Thomas, Biocatalysis in supercritical fluids, in fluorous solvents, and under solvent-free conditions, Chem. Rev., 107, 2786–2820 (2007). 61. A. Behr, M. Urschey, and V.A. Brehme, Aqueous biphasic catalysis as a powerful tool for catalyst recycling in telomerization and hydrogenation chemistry, Green Chem., 5, 198–204 (2003). 62. D.E. Bergbreiter and J.G. Franchina, A soluble fluorous phase polymer support, Chem. Commun., 1531–1532 (1997). 63. D.E. Bergbreiter, J.G. Franchina, B.L. Case, L.K. Williams, J.D. Frels, and N. Koshti, Fluorousphase soluble polymeric supports, Comb. Chem. High Throughput Screening, 3, 153–164 (2000). 64. D.E. Bergbreiter, J.G. Franchina, and B.L. Case, Fluoroacrylate-bound fluorous-phase soluble hydrogenation catalysts, Org. Lett., 2, 393–395 (2000). 65. D.E. Bergbreiter, N. Koshti, J.G. Franchina, and J.D. Frels, Sequestration of trace metals using water-soluble and fluorous phase-soluble polymers, Angew. Chem., Int. Ed., 39, 1040–1042 (2000). 66. W. Chen, L. Xu, and J. Xiao, Fluorous soluble polymer catalysts for the fluorous biphase hydroformylation of olefins, Chem. Commun., 839–840 (2000). 67. I. Kani, R. Flores, J.P. Fackler, and A. Akgerman, Hydroformylation of styrene in supercritical carbon dioxide with fluoroacrylate polymer supported rhodium catalysts, J. Supercrit. Fluids, 31, 287–294 (2004). 68. R. Flores, Z.K. Lopez-Castillo, I. Kani, J.P. Fackler, Jr., and A. Akgerman, Kinetics of the Homogeneous catalytic Hydrogenation of olefins in supercritical carbon dioxide using a fluoroacrylate copolymer grafted rhodium catalyst, Ind. Eng. Chem. Res., 42, 6720–6729 (2003). 69. Z.K. Lopez-Castillo, R. Flores, I. Kani, J.P. Fackler, Jr., and A. Akgerman, Evaluation of polymer-supported Rhodium catalysts in 1-octene hydroformylation in supercritical carbon Dioxide, Ind. Eng. Chem. Res., 42, 3893–3899 (2003).


151

70. Z.K. Lopez-Castillo, R. Flores, I. Kani, J.P. Fackler, Jr., and A. Akgerman, Fluoroacrylate copolymer-supported Rhodium catalysts for hydrogenation reactions in supercritical carbon dioxide, Ind. Eng. Chem. Res., 41, 3075–3080 (2002). 71. D. Dietz, C. Ohman, and T. Scholten, Biphasic hydroformylation of higher olefins, in AIChE Spring National Meeting, Conference Proceedings, New Orleans, LA, US, Apr. 25–29, 2004, pp. 309–313. 72. Z. Jin, X. Zheng, and B. Fell, Thermoregulated phase transfer ligands and catalysis. I. Synthesis of novel polyether-substituted triphenylphosphines and application of their rhodium complexes in two-phase hydroformylation, J. Mol. Cat., A: Chem., 116, 55–58 (1997). 73. Y. Wang, J. Jiang, Q. Miao, X. Wu, and Z. Jin, Thermoregulated phase transfer ligands and catalysis. Part XIII. Use of nonionic tensioactive water-soluble phosphine ligands to effect homogeneous catalyst separation and recycling, Catal. Today, 74, 85–90 (2002). 74. R. Chen, X. Liu, and Z. Jin, Thermoregulated phase-transfer ligands and catalysis. Part VI. Twophase hydroformylation of styrene catalyzed by the thermoregulated phase-transfer catalyst OPGPP/Rh, J. Organomet. Chem., 571, 201–204 (1998). 75. C. Liu, J. Jiang, Y. Wang, F. Cheng, and Z. Jin, Thermoregulated phase transfer ligands and catalysis. XVIII: Synthesis of N,N-dipolyoxyethylene-substituted-2-(diphenylphosphino)phenylamine (PEO–DPPPA) and the catalytic activity of its rhodium complex in the aqueous–organic biphasic hydroformylation of 1-decene, J. Mol. Cat., A: Chem., 198, 23–27 (2003). 76. A. Behr and R. Roll, Hydroaminomethylation in thermomorphic solvent systems, J. Mol. Catal. A: Chem., 239, 180–184 (2005). 77. M. Ooe, M. Murata, T. Mizugaki, K. Ebitani and K. Kaneda, Supramolecular catalysts by encapsulating palladium complexes within dendrimers, J. Am. Chem. Soc., 126, 1604–1605 (2004). 78. T. Mizugaki, M. Murata, M. Ooe, K. Ebitani and K. Kaneda, Novel catalysis of dendrimer-bound Pd(0) complexes: sterically steered allylic amination and the first application for a thermomorphic system, Chem. Commun., 52–53 (2002). 79. D.E. Bergbreiter, P.L. Osburn, and Y.-S. Liu, Tridentate SCS Palladium(II) complexes: new, highly stable, recyclable catalysts for the Heck reaction, J. Am. Chem. Soc., 121, 9531–9538 (1999). 80. K.Q. Yu, W. Sommer, J.M. Richardson, M. Weck, and C.W. Jones, Evidence that SCS pincer Pd(II) complexes are only precatalysts in Heck catalysis and the implications for catalyst recovery and reuse, Adv. Synth. Catal., 347, 161–171 (2005). 81. W.J. Sommer, K. Yu, J.S. Sears, Y. Ji, X. Zheng, R.J. Davis, C.D. Sherrill, C.W. Jones, and M. Weck, Investigations into the stability of tethered palladium(II) pincer complexes during Heck catalysis, Organometallics, 24, 4351–4361 (2005). 82. D.E. Bergbreiter, P.L. Osburn, and J.D. Frels,Mechanistic studies of SCS–Pd complexes used in Heck catalysis, Adv. Synth. Catal., 347, 172–184. 83. C. Rocaboy and J.A. Gladysz, Highly active thermomorphic fluorous palladacycle catalyst precursors for the Heck reaction; evidence for a palladium nanoparticle pathway, Org. Lett., 4, 1993–1996 (2002). 84. C. Rocaboy and J.A. Gladysz, Thermomorphic fluorous imine and thioether palladacycles as precursors for highly active Heck and Suzuki catalysts; evidence for palladium nanoparticle pathways, New. J. Chem., 27, 39–49 (2003). 85. N. Lu, S.-C. Chen, T.-C. Chen, and L.-K. Liu, Palladium-catalyzed Heck reaction under thermomorphic mode, Tetrahedron Lett., 49, 371–375 (2008). 86. Y.C. Yang and T.Y. Luh, Polymeric phosphine ligand from ring-opening metathesis polymerization of a norbornene derivative. Applications in the Heck, Sonogashira, and Negishi reactions, J. Org. Chem., 68, 9870–9873 (2003). 87. S. Kim, K. Yamamoto, K. Hayashi, and K. Chiba, A cycloalkane-based thermomorphic system for palladium-catalyzed cross-coupling reactions, Tetrahedron, 64, 2855–2863 (2008). 88. A. Kollhofer and H. Plenio, Homogeneous catalysts supported on soluble polymers: Biphasic Sonogashira coupling of aryl halides and acetylenes using MeOPEG-bound phosphine–palladium catalysts for efficient catalyst recycling, Chem. Eur. J., 9, 1416–1425 (2003).

152


89. J. Hillerich and H. Plenio, Continuous biphasic catalysis: palladium catalyzed cross coupling reactions, Chem. Commun., 3024–3025 (2003). 90. A. Datta and H. Plenio, Nonpolar biphasic catalysis: Sonogashira and Suzuki coupling of aryl bromides and chlorides, Chem. Commun., 1504–1505 (2003). 91. H. Remmele, A. Koelhofer, and H. Plenio, Recyclable catalyst with cationic phase tags for the Sonogashira coupling of aryl bromides and aryl chlorides, Organometallics, 22, 4098–4103 (2003). 92. M. an der Heiden and H. Plenio, Homogeneous catalysts supported on soluble polymers: biphasic Suzuki–Miyaura coupling of aryl chlorides using phase-tagged palladium–phosphine catalysts, Chem. Eur. J., 10, 1789–1797 (2004). 93. C.A. Fleckenstein and H. Plenio, Aqueous/organic cross coupling: sustainable protocol for Sonogashira reactions of heterocycles, Green Chem., 10, 563–570 (2008). 94. C.A. Fleckenstein and H. Plenio, Efficient Suzuki–Miyaura coupling of (hetero)aryl chlorides with thiophene- and furanboronic acids in aqueous n-butanol, J. Org. Chem., 73, 3236–3244 (2008). 95. A.M. Halpern and S. Gozashti, Mutual solubilities of liquids in a binary two-phase system. A physical chemistry experiment, J. Chem. Educ., 65, 371–372 (1988). 96. I.P. Beletskaya, A.N. Kashin, A.E. Litvinov, V.S. Tyurin, P.M. Valetsky, and G. van Koten, Palladium colloid stabilized by block copolymer micelles as an efficient catalyst for reactions of C–C and C–heteroatom bond formation, Organometallics, 25, 154–158 (2006). 97. N.V. Tsarevsky and K. Matyjaszewski, ‘Green’ atom transfer radical polymerization: from process design to Preparation of well-defined environmentally friendly polymeric materials, Chem. Rev., 107, 2270–2299 (2007). 98. S. Liou, J.T. Rademacher, D. Malaba, M.E. Pallack, and W.J. Brittain, Atom transfer radical polymerization of methyl methacrylate with polyethylene-functionalized ligands, Macromolecules, 33, 4295–4296 (2000). 99. Y. Shen, S. Zhu, and R. Pelton, Soluble and recoverable support for copper bromide-mediated living radical polymerization, Macromolecules, 34, 3182–3185 (2001). 100. Barre, D. Taton, D Lastecoueres, and J.-M. Vincent Closer to the ‘ideal recoverable catalyst’ for atom transfer radical polymerization using a molecular non-fluorous thermomorphic system, J. Am. Chem. Soc., 126, 7764–7765 (2004). 101. D.E. Bergbreiter, P.N. Hamilton, and N.M. Koshti, Self-separating homogeneous copper(I) catalysts, J. Am. Chem. Soc., 109, 10666–10667 (2008). 102. N. Lu, C.-L. Lin, and T.-C. Chen, New fluorous bipyridine ligands for copper-catalyzed atom transfer radical polymerization of methyl methacrylate in the thermomorphic mode, J. Appl. Polym. Sci., 110, 2531–2537 (2008). 103. H.E. Bryndza, M.C. Grady, S.S. Kristjansdottir, C.M. Older, M.A. Page, J.C. Ritter, and W. Tam, E. I. du Pont de Nemours and Company, USA, Patent Application: WO 2007024634, (2007). 104. M. Wende, R. Meier, and J.A. Gladysz, Fluorous catalysis without fluorous solvents: a friendlier catalyst recovery/recycling protocol based upon thermomorphic properties and liquid/solid phase separation, J. Am. Chem. Soc, 123, 11490–11491 (2001). 105. M. Wende and J.A. Gladysz, Fluorous catalysis under homogeneous conditions without fluorous solvents: a ‘greener’ catalyst recycling protocol based upon temperature-dependent solubilities and liquid/solid phase separation, J. Am. Chem. Soc., 125, 5861–5872 (2003). 106. L.V. Dinh and J.A. Gladysz, ‘Catalyst-on-a-tape’-Teflon: A new delivery and recovery method for homogeneous fluorous catalysts, Angew. Chem., Int. Ed., 44, 4095–4097 (2005). 107. F.O. Seidel and J.A. Gladysz, Catalysis of intramolecular Morita–Baylis–Hillman and Rauhut–Currier reactions by fluorous phosphines; facile recovery by liquid/solid organic/ fluorous biphase protocols involving precipitation, Teflon tape, and Gore-Rastex fiber, Adv. Synth. Catal., 350, 2443–2449 (2008). 108. M.A. Grunlan, K.R. Regan, and D.E. Bergbreiter, Liquid/liquid separation of polysiloxanesupported catalysts, Chem. Commun., 1715–1717 (2006). 109. N.F. Yang, H. Gong, W.J. Tang, Q.H. Fan, C.Q. Cai, and L.W. Yang, Phase selectively soluble dendritic derivative of 4-(N,N-dimethylamino)pyridine: an easily recyclable catalyst for Baylis–Hillman reactions, J. Mol. Cat., A: Chem., 233, 55–59 (2005).


153

110. D. Mandel, M. Jurisch, C.S.M. Consorti, and J.A. Gladysz, Ionic transformations in extremely nonpolar fluorous media: easily recoverable phase-transfer catalysts for halide-substitution reactions, Chem. Asian J., 3, 1772–1782 (2008). 111. N. Candelon, D. Lastecoueres, A.K. Diallo, J. Ruiz Aranzaes, D. Astruc, and J.-M. Vincent, A highly active and reusable copper(I)–tren catalyst for the ‘click’ 1,3-dipolar cycloaddition of azides and alkynes, Chem. Commun., 741–742 (2008). 112. V.L. Dinh and J.A. Gladysz, Monophasic and biphasic hydrosilylations of enones and ketones using a fluorous rhodium catalyst that is easily recycled under fluorous–organic liquid–liquid biphasic conditions, New. J. Chem., 29, 173–181 (2005). 113. M. Contel, P.R. Villuendas, J. Fernandez-Gallardo, P.J. Alonso, J.-M. Vincent, and R.H. Fish, Fluorocarbon soluble copper(II) carboxylate complexes with nonfluoroponytailed nitrogen ligands as precatalysts for the oxidation of alkenols and alcohols under fluorous biphasic or thermomorphic modes: structural and mechanistic aspects, Inorg. Chem., 44, 9771–9778 (2005). 114. N. Lu and Y.-C. Lin, Efficient, recoverable, copper-catalyzed aerobic oxidation of alcohols under FBS and thermomorphic mode, Tetrahedron Lett., 48, 8823–8828 (2007). 115. H. Hamamoto, Y. Suzuki, Y.M.A. Yamada, H. Tabata, H. Takahashi, S. Ikegami, A recyclable catalytic system based on a temperature-responsive catalyst, Angew. Chem., Int. Ed., 44, 4536–4538 (2005). 116. H. Hugl and M. Nobis, Multiphase catalysis in industry, Top. Organometal. Chem., 23, 1–17 (2008).

7 Self-supported Asymmetric Catalysts Wenbin Lin and David J. Mihalcik Department of Chemistry, University of North Carolina, Chapel Hill, NC, USA

7.1 Introduction Asymmetric catalysis provides a powerful method for the synthesis of chiral molecules that are important ingredients for the pharmaceutical, agrochemical, and fragrance industries.1 Numerous highly selective chiral catalysts have been developed in the recent decades, and three pioneers in the field of homogeneous asymmetric catalysis were awarded the Nobel Prize in Chemistry in 2001.2 The practical applications of many homogeneous asymmetric catalysts in industrial processes are however hindered by their high costs as well as difficulties in removing trace amounts of toxic metals from the organic products.3 Heterogenization of homogeneous asymmetric catalysts represents a logical approach to overcome these problems.4 The heterogenized catalysts can potentially provide easily recyclable and reusable solid catalysts that have uniform and precisely engineered active sites similar to those of their homogeneous counterparts, and therefore combine the advantages of both homogeneous and heterogeneous systems. Many heterogenization approaches have been explored, including attaching the chiral catalysts to organic polymers, dendrimers, membrane supports, and porous inorganic oxides and immobilization via biphasic systems.5 This chapter provides an overview of recently developed, self-supported asymmetric catalysts that have the potential to provide the highest possible catalyst loading and also the most uniform catalytic sites.6 Self-supported asymmetric catalysts can in principle be prepared by incorporating catalytic sites as pendants on a linear organic polymer7 or in the backbone of an organic polymer.8 Many of these polymers have low molecular weights and tend to behave as


156


homogeneous catalysts due to their high solubility. As a result, their recovery during catalytic reactions typically requires large amounts of solvents of different polarity. Such polymer-immobilized asymmetric catalysts also tend to have diluted catalytic sites and are covered in other chapters of this volume. Instead, this chapter will focus on the heterogeneous catalyst systems recently prepared by three distinct strategies that do not rely on any support materials (Scheme 7.1). In the first approach, catalytically active subunits are linked by secondary metal centers to form solid catalysts. In the second approach, multitopic ligands containing orthogonal secondary functional groups can be lined by metal centers to form porous solids which are then treated with secondary metal centers to form solid catalysts. In the third approach, multitopic chiral ligands are linked with metal centers to form catalytically active polymeric solids. This chapter surveys recent advances in the synthesis and catalytic applications of self-supported asymmetric catalysts using these approaches. The advantages and disadvantages of each strategy are also discussed.

Scheme 7.1 Three distinct strategies for the synthesis of self-supported asymmetric catalysts.

7.2 Self-supported Asymmetric Catalysts Formed by Linking Catalytically Active Subunits via Metal–Ligand Coordination Lin et al. prepared a series of chiral, porous, hybrid solids based on zirconium phosphonates for highly enantioselective hydrogenation of b-keto esters. These hybrid materials were prepared based on Approach 1 (Scheme 7.2) and designed to combine the robust framework structure of metal phosphonates9 and enantioselectivity of metal complexes of the chiral bisphosphines.10 Enantiopure 2,20 -bis-(diphenylphosphino)-1,10 -binaphthyl-6,60 bis(phosphonic acid), L1-H4, was synthesized in three steps starting from the known 2,20 -dihydroxy-1,10 -binaphthyl-6,60 -bis(diethylphosphonate) in 47% overall yield.11 Treatment [Ru(benzene)Cl2]2 with 1 equiv of L1-H4 in DMF at 100 C led to the complex [Ru(L1-H4)(DMF)2Cl2] which was reacted with Zr(OtBu)4 in reluxing methanol to afford the zirconium phosphonate Zr–Ru–L1. The analogous solid Zr–Ru–L2 with the

Self-supported Asymmetric Catalysts O

O

HO P HO

HO P HO PPh2 [Ru(benzene)Cl2 ]2 PPh2

DMF

Ph Ph

Cl

t DMF Zr(O Bu)4 Zr Ru DMF MeOH P Cl Ph Ph

P

HO HO P

HO HO P O

O O P O

P O O P

Ph Ph

DMF DMF

Cl

O

Zr

Zr

O O O P

O Zr

Ph Ph P P

O

P O

Ph Ph

Cl

P P

DMF

Ru

DMF Zr Zr

P O O

Cl Ru

DMF DMF

Cl Ph Ph

O

Cl Ph Ph O

Cl Ru

P

O

Zr

Ph Ph

157

O P O

Zr Zr

(Zr-Ru-L1)

(Zr-Ru-L2)

O R1

O O

R 2 + H2

Zr-Ru-(R)-L1 or Zr-Ru-(R)-L2 CH 3OH

OH

O

R2 O up to 95.0% ee reused for >5 times R1

Scheme 7.2 Synthesis of Ru(BINAP)-containing zirconium phosphonates for asymmetric hydrogenation of b-ketoesters. Reproduced with permission from Hu, A., Ngo, H. L., Lin, W. (2003) Chiral, porous, hybrid solids for highly enantioselective heterogeneous asymmetric hydrogenation of b-keto esters. Angew. Chem. Int. Ed., 42 (48), 6000–6003

2,20 -bis-(diphenylphosphino)-1,10 -binapthyl-4,40 -bis(phosphonic acid) ligand, L2-H4 was similarly prepared. The chiral porous phosphonates were characterized by thermogravimetric analysis (TGA), nitrogen adsorption isotherms, X-ray diffraction, SEM, IR spectroscopy, and microanalysis. The compositions of Zr–Ru–L1 and Zr–Ru–L2 were determined by TGA and microanalysis while the IR spectra suggested the formation of zirconium phosphonate bonds. Zr–Ru–L1 exhibits a total BET surface area of 475 m2/g and a pore volume of 1.02 cm3/g, whereas Zr–Ru–L2 has a total BET surface area of 387 m2/g and a pore volume of 0.53 cm3/g. SEM images showed the presence of sub-micrometer particles in the solids while powder X-ray diffraction (PXRD) studies indicated the amorphous nature of the solids. The Zr–Ru–L1 and Zr–Ru–L2 solids were utilized for heterogeneous asymmetric hydrogenation of b-keto esters (Scheme 7.2). Zr–Ru–L1 catalyzed the hydrogenation of a wide range of b-alkyl-substituted b-keto esters with complete conversion and ee values ranging from 91.7% to 95.0%. Zr–Ru–L1 gave a turnover frequency (TOF) of 364 h1 with

158


0.1% solid loading compared to a TOF of 810 h1 for the homogeneous BINAP–Ru catalyst. In contrast, Zr–Ru–L2 catalyzed the hydrogenation of b-keto esters with only modest ee values. The supernatant solutions of Zr–Ru–L1 and Zr–Ru–L2 in MeOH did not catalyze the hydrogenation of b-keto esters, indicating the heterogeneity of the solid catalysts. In addition, analysis of the reaction solution by direct current plasma spectroscopy (DCP) indicated that less than 0.01% of the ruthenium leached into the organic solution during hydrogenations. The Lin group was able to reuse the Zr–Ru–L1 system for asymmetric hydrogenation of methyl acetoacetate without significant deterioration on enantioselectivity. The Zr–Ru–L1 system was used for five cycles of hydrogenation with complete conversion and ee values of 93.5, 94.2, 94.0, 92.4, and 88.5%, respectively. This synthetic strategy was modified by the Lin group to prepare zirconium phosphonates containing the Ru–BINAP–DPEN species (Scheme 7.3) that were shown by Noyori et al. to be highly active homogeneous catalyst for enantioselective hydrogenation of aromatic ketones.12 Reaction of L1-H4 with [Ru(benzene)Cl2]2 followed by (R,R)-DPEN afforded the phosphonic acid-substituted Ru–BINAP–DPEN intermediate which was treated with Zr(OtBu)4 under reflux conditions to give chiral porous Zr phosphonate of the approximate formula Zr[Ru(L1)(DPEN)Cl2]4H2O (Zr–Ru–L1–DPEN). The solid precatalyst Zr–Ru–L2–DPEN with a 4,40 -disubstituted BINAP was similarly prepared and also has an approximate formula of Zr[Ru(L2)(DPEN)Cl2]4H2O. The solid catalysts were characterized by a variety of techniques including TGA, adsorption isotherms, XRD, SEM, IR, and microanalysis.

HO HO

O

O

HO

P

HO

O

P Ph Ph Cl H2 N P Ru N P H2 Ph Ph Cl

PPh2 1) [Ru(benzene)Cl ] 2 2 2) (R,R)-DPEN

PPh2 O HO

O

P

HO

OH

Zr O O O P

O Ph Ph

Zr(OtBu)4

O

P

O

OH

Zr

Zr O O P O

Ph Ph

Zr O O

O

P O O

Ph Ph Cl H2 N P Ru N P H2 Ph Ph Cl

Zr

MeOH

Zr

Ph Ph Cl H2 N P Ru P N H2 Ph Ph Cl

O P Ph Ph

P O

Ph Ph Cl H2 N P Ru P N H Ph Ph Cl 2

Ph Ph

P O Zr

Zr Zr

(Zr-Ru-L2-DPEN)

(Zr-Ru-L1-DPEN)

Scheme 7.3 Synthesis of Ru(BINAP)(DPEN)-containing zirconium phosphonates. Hu, A., Ngo, H. L., Lin, W. (2003) Chiral porous hybrid solids for practical heterogeneous asymmetric hydrogenation of aromatic ketones. J. Am. Chem. Soc., 125 (38), 11490–11491

Self-supported Asymmetric Catalysts

159

Nitrogen adsorption measurements indicate that both Zr–Ru–L1–DPEN and Zr–Ru– L2–DPEN are highly porous and have wide pore size distributions. Zr–Ru–L1–DPEN has a surface area of 400 m2/g and a pore volume of 0.98 cm3/g, whereas Zr–Ru–L2–DPEN exhibits a total BET surface area of 328 m2/g and a pore volume of 0.65 cm3/g. SEM images showed that both solids were composed of submicrometer particles, while PXRD indicated that both solids were amorphous. Porous solids of Zr–Ru–L1–DPEN and Zr–Ru–L2–DPEN exhibited exceptionally high activity and enantioselectivity in the asymmetric hydrogenation of aromatic ketones.12a Acetophenone was hydrogenated to 1-phenylethanol with complete conversion and 96.3% ee in 2-propanol with 0.1mol% loading of the Zr–Ru–L2–DPEN solid. This level of ee is significantly higher than that observed for the parent Ru–BINAP–DPEN homogeneous catalyst which typically gives 80% ee for the hydrogenation of acetophenone under similar conditions.12b,12c In comparison, the Zr–Ru–L1–DPEN solid gives 79.0% ee for the hydrogenation of acetophenone under the same conditions. As shown in Table 7.1, the Zr–Ru–L2–DPEN solid has been used to catalyze a series of aromatic ketones with remarkably high ee values of 90.6–99.2% and complete conversions. In contrast, the ee for the hydrogenation of aromatic ketones is modest and similar to that of the parent Ru–BINAP–DPEN homogeneous catalyst. The Lin group later demonstrated that the enhanced ee values exhibited by the Zr–Ru–L2–DPEN solid is due to the beneficial steric influence of the bulky substituents in the 4,40 -positions of BINAP.13 Table 7.1 Heterogeneous Asymmetric Hydrogenation of Aromatic Ketonesa,b O Ar

Substrate Ar ¼ Ph, R ¼ Me Ar ¼ 2-naphthyl, R ¼ Me Ar ¼ 40 -t Bu-Ph R ¼ Me Ar ¼ 40 -MeO-Ph, R ¼ Me Ar ¼ 40 -Cl-Ph, R ¼ Me Ar ¼ 40 -Me-Ph, R ¼ Me Ar ¼ Ph, R ¼ Et Ar ¼ Ph, R ¼ cyclo-Pr Ar ¼ 1-naphthyl R ¼ Me

+ H2 R

Solid Loading 0.1% 0.1% 0.1% 0.1% 0.1% 0.1% 0.1% 0.1% 0.1% 0.02% 0.005% 0.005%

Zr-Ru-L1-DPEN orZr-Ru- L2-DPEN KOt Bu,IPA

KOtBu 1% 1% 1% 1% 1% 1% 1% 1% 1% 0.4% 0.02% 0.02%

OH Ar

R

Zr-Ru-L2-DPEN e.e.% 96.3(97.1)c 97.1 99.2 96.0 94.9 97.0 93.1 90.6 99.2 98.9 98.8(70%)d 98.6e

Zr-Ru-L1-DPEN e.e.% 79.0 (81.3)c 82.1 91.5 79.9 59.3 79.5 83.9 — 95.8

a Reproduced with permission from A. Hu , H. L. Ngo and W. Lin, Chiral Porous Hybrid Solids for Practical Heterogeneous Asymmetric Hydrogenation of Aromatic Ketones, J. Am. Chem. Soc., 2003, 125, 38, 11490–11491. Copyright 2003 American Chemical Society. b All of the reactions were carried out in 20 h and the e.e. values were determined by GC on a Supelco b-Dex 120 column. The absolute configurations of the products are identical to those obtained by the Ru-(R)-BINAP-(R,R)-DPEN catalyst. All the conversions were >99% as judged by the integrations of GC peaks. c homogeneous reactions. d 70% conversion. e 40 h reaction time.

160


The solid zirconium phosphonate catalysts were also extremely active for the hydrogenation of other aromatic ketones. For example, with only 0.02 mol% solid loading of Zr–Ru–L2–DPEN, 1-acetonaphthone was hydrogenated with complete conversion and 98.9% ee in 20 h. When the solid loading was decreased to 0.005 mol%, it took longer reaction time (40 h) for the hydrogenation of 1-acetonaphthone to complete (98.6% ee). The TOF was calculated to be 500 h1 at complete conversion and 700 h1 at 70% conversion. The solid catalysts were successfully reused for the asymmetric hydrogenation of 1-acetonaphthone without the deterioration of enantioselectivity. The Zr–Ru–L2–DPEN system was used for eight cycles of hydrogenation without any loss of enantioselectivity. The catalyst recycling and reuse experiments were conducted without rigorous exclusion of air, and the oxygen sensitivity of the ruthenium hydride complexes may have contributed to the loss of activity after multiple runs. Therefore, the loss of activity may not reflect the intrinsic instability of the Zr–Ru–L2–DPEN solid catalyst. DCP studies of the supernatant solution indicated less than 0.2% of Ru leached into the organic product during each round of hydrogenation. Nguyen, Hupp, and co-workers recently used a similar strategy to prepare self-supported catalysts for asymmetric epoxidation reactions.14 The [bis(catechol)-salen]Mn compound, L3, was designed to have catechol groups for crosslinking with metal centers to afford a series of coordination polymers (Scheme 7.4). Poly(Cu–L3) was prepared by treating L3 with CuII in the presence of triethylamine in DMF. After stirring, poly(Cu–L3) precipitated as brown solid and was thoroughly washed with DMF. Inductively couple plasma–mass spectrometric (ICP–MS) studies indicated that poly(Cu–L3) had a Mn/Cu ratio of 1 : 1.1, suggesting the formation of a quasi-one-dimensional structure where L3 is connected by bis(catecholate)copper linkages. The poly(Cu–L3) solid has a modest surface area of 72 m2/g and is amorphous as judged by SEM and PXRD.

N HO

O

N Mn

O

OH

Cl OH

HO crosslinkable functional group

catalytic part

L3

crosslinkable functional group

Scheme 7.4 Structure of the crosslinkable [bis(catechol)-salen]Mn compound (L3)

2,2-Dimethyl-2H-chromene was used as a model substrate for epoxidation reactions using 2-(tert-butylsulfonyl)iodosylbenzene as the oxidant. A slightly lower ee of 76% was observed for the heterogeneous poly(Cu–L3) as compared with an ee of 86% for the homogeneous counterpart, [bis(dimethoxyphenyl)salen]MnIIICl. As shown in Table 7.2, the poly(Cu–L3) catalyst was readily recovered by centrifugation and was reused up to ten times with little loss in activity (from 79% to 70% yield) and no loss in enantioselectivity (75–76% ee). After the first two cycles where 3.1% Mn and 4.7% Cu were determined to have leached from the sample, the metal loss slowed over the next four cycles and no metals were leached from the solid by the tenth run. The


161

Table 7.2 Recyclability of poly(Cu–L3) in the asymmetric epoxidation of 2,2-dimethyl-2Hchromene with 2-(tert-butylsulfonyl)iodosylbenznea O

poly(Cu-L3) or L3 2-(tert -butylsulfonyl)iodosylbenzene

O

CH2Cl2, r.t. O

Entry 1 2 3 4 5 6 7 8 9 10 11

Catalyst L3 poly(Cu–L3) poly(Cu–L3) poly(Cu–L3) poly(Cu–L3) poly(Cu–L3) poly(Cu–L3) poly(Cu–L3) poly(Cu–L3) poly(Cu–L3) poly(Cu–L3)

Reuse cycle st

1 1st 2nd 3rd 4th 5th 6th 7th 8th 9th 10th

Yield [%]

ee [%]

Leaching [%]d of Cu/Mn

87 79 74 78 80 80 73 79 72 69 70

86 76 76 75 76 75 76 76 75 75 76

not applicable 2.9 : 1.5 1.8 : 1.6 0.5 : 0.9 0.1 : 0.3 0.1 : 0.1 0.1 : 0.1 < 0.1 : 3 h and a total turnover number (TON) of >2000 in that period. In comparison, the homogeneous counterpart had a lifetime of 95 >95 70 >95 >95 95c 92c 96c 99c >99c >68c

e.e.% 55b 59 59 61 72 59 43c 45c 48c 29c 46c 0

The e.e. values were determined by GC on a Supelco g-Dex 120 column, while the conversions were determined by the integrations of 1 HNMR spectra and GC traces. All the reactions were carried out with 20 mol% solid loading in toluene at r.t. Reprinted from Journal of Molecular Catalysis A: Chemical, 215, H. L. Ngo, A. Hu and W. Lin, Molecular building block approaches to chiral porous zirconium phosphonates for asymmetric catalysis, 177–186. Copyright 2004, with permission from Elsevier. b The reaction is carried in DCM. c The reactions were carried out with 50 mol% solid loading at r.t. a

166


Figure 7.2 (a) Schematic representation of the 3D framework of 2 showing the zigzag chains of [Cd(m-Cl)2]n along a axis. (b) space-filling model of 2 as viewed down the a axis, showing large 1D chiral channel (1.6 1.8 nm). (c) Schematic representation of the active (BINOLate)Ti (OiPr)2 catalytic sites in the open channels of 2. Reproduced with permission from C.-D. Wu, A. Hu, L. Zhang, W. Lin, A Homochiral Porous Metal–Organic Framework for Highly Enantioselective Heterogeneous Asymmetric Catalysis, J. Am. Chem. Soc., 2005, 127, 8940–8941. Copyright 2005 American Chemical Society. (See plate section for a color representation)

using the solid derived from 2,20 -ethoxy-1,10 -binaphthyl-6,60 -bis(styrylphosphonic acid) (see control entry in Table 7.4) suggested that residual phosphonic acid protons in the solids can activate Ti(OiPr)4 for nonenantioselective ZnEt2 addition. This type of background reaction is probably responsible for modest ee values observed for the solid catalysts. In an effort to improve the enantioselectivity and also to obtain single-crystalline solids, Lin et al. designed BINOL-derived ligand (R)-6,60 -dichloro-2,20 -dihydroxyl-1,10 -binaphtylbipyridine (L8). The 3D homochiral MOF [Cd3(L8)3Cl6]4DMF6MeOH3H2O (2) was prepared by slow vapor diffusion of diethyl ether into the mixture of (R)-L8 and CdCl2 in DMF/ MeOH. 2 is built from linking 1D zigzag [Cd(m-Cl)2]n SBUs by the L8 ligands via pyridine coordination, and has a highly porous structure with the largest channel opening of 1.61.8 nm running along the a axis (Figure 7.2). One third of the chiral dihydroxy groups in 2 are facing the open channels and are accessible to secondary metal centers to generate active catalytic sites. 2 was pretreated by Ti(OiPr)4 to generate the grafted Ti–BINOLate species that efficiently catalyzed the diethylzinc addiction reactions in up to 93% ee. This level of ee is comparable with the homogeneous analogue (94% ee) (Table 7.5). The heterogeneous nature of this solid catalyst was indicated by the nonreactive supernatant from a mixture of 2 and Ti(OiPr)4. Moreover, Lin et al. carried out a set of control experiments using aldehydes with different size from 0.8 nm to 2.0 nm to show that the aldehydes are accessing the Ti active sites in the interior of crystals of 2. They observed decreased conversion when larger aldehyde was used. No conversion was observed for aldehyde G02 with size of 2.0 nm which is larger than the open channels of framework 2. While using the same ligand L8, Lin et al. obtained two different homochiral MOFs [Cd3(L3)4(NO3)6]7MeOH5H2O (3) and [Cd(L3)5(ClO4)2]DMF4MeOH3H2O (4) when Cd(NO3)2 and Cd(ClO4)2 were used as the metal sources, respectively. Compound 3 adopts twofold interpenetrating framework structure with each of the 3D frameworks constructed from linking2D grids with 1D zigzag polymeric chains. Large channels with dimensions of 13.5 13.5 A are present along the c axis (Figure 7.3c). Compound 4 adopts a 3D network from highly interpenetrated 2D grids, resulting in 1D channels with a size of 1.2 1.5 nm (Figure 7.3e).


167

Table 7.5 Ti(IV)-catalyzed ZnEt2 additions to aromatic aldehydesa,b O +ZnEt 2 Ar

H

(R)-2 Ar Ti(OiPr)4

OH Et H

BINOL/Ti(OiPr)4 Ar 1-Naph Ph 4-Cl-Ph 3-Br-Ph 40 -G0OPh 40 -G10 OPh 40 -G1OPh 40 -G20 OPh O

2*Ti

conv.%

ee%

conv.%

>99 >99 >99 >99 >99 >99 >99 95c

94 88 86 84 80 75 78 67c

>99 >99 >99 >99 >99 73 63 0

93 83 80 80 88 77 81 —

H

O O

R

ee%

R:

CH3 O

R

O O Dendritic aldehydes

Dendron: Est. size:

O

G0

0.8 nm

G1

1.45 nm

O

G1

G2'

1.55 nm

2.0 nm

a Reproduced with permission from C.-D. Wu, A. Hu, L. Zhang, W. Lin, A Homochiral Porous Metal–Organic Framework for Highly Enantioselective Heterogeneous Asymmetric Catalysis, J. Am. Chem. Soc., 2005, 127, 8940–8941. Copyright 2005 American Chemical Society. b All the reactions were conducted with 13 mol% of 2 or 20 mol% BINOL and excess amounts of Ti(OiPr)4 at room temperature for 12 h. Conversions were determined by GC or NMR, while ee% values were determined on chiral GC or HPLC for all the secondary alcohols except for 40 -G02 OPh whose ee% was determined by NMR spectrum of its Mosher’s ester. c With 40 mol% BINOL.

Compound 3 was treated with Ti(OiPr)4 to lead to an active heterogeneous asymmetric catalyst for the diethylzinc addition to aromatic aldehydes with up to 90% ee. However, under the same conditions, a mixture of 4 and Ti(OiPr)4 was inactive in catalyzing the diethylzinc addition to aromatic aldehydes. Careful investigation of the structure of 3 revealed the close proximity of the Cd(py)2(H2O)2 moiety in one 2D grid with the dihydroxyl groups of the other 2D grid. The dihydroxyl groups in 4 are thus inaccessible to Ti(OiPr)4. When treated with Ti(OiPr)4, compounds 3 and 4 had entirely different catalytic activities due to the subtle structure differences. The finding of such a drastic difference in catalytic activity is remarkable since 3 and 4 were built from exactly the same building blocks. This result points to the importance of the framework structure in determining the performance of self-supported asymmetric catalysts.

168


Figure 7.3 (a) The 2D square grid in 3. (b) Schematic representation of the 3D framework of 3. (c) Space-filling model of 3 as viewed down the c axis, the 2-fold interpenetrating networks are shown with blue and violet colors. (d) Schematic representation of the interpenetration of mutually perpendicular 2D grids in 4. (e) Space-filling model of 4 as viewed down the c axis. (f) Schematic representation of steric congestion around the chiral dihydroxyl group of L8 (orange sphere) arising from the interpenetration of mutually perpendicular 2D grids in 4. C.-D. Wu & W. Lin, Angew. Chem. Int. Ed., 2007, 46, 1075–1078. Copyright Wiley-VCH Verlag GmbH & Co. KGaA. Reproduced with permission. (See plate section for a color representation)

7.4 Self-supported Asymmetric Catalysts Formed by Linking Multitopic Chiral Ligands with Catalytic Metal Centers The third approach to self-supported asymmetric catalysts relies on directly linking multitopic chiral ligands with metal centers. Ding et al.18 and Sasai et al.19 independently demonstrated this concept by linking multitopic BINOL ligands with Al(III) and Ti(IV) centers to generate chiral Lewis acid catalysts. The bis(BINOL) derivatives L9a–d were synthesized by linking the BINOL units at the 6-position.20 Treatment of L9a–d with equal molar LiAlH4 in THF at 0 C resulted in spontaneous formation of a white precipitate which was reacted with 0.5 equivalents of BuLi to afford heterogeneous versions of the Al–Li–bis (binaphthoxide) (ALB) catalyst that was reported earlier (Scheme 7.7).21 With an equivalent of 20 mol% ALB catalyst loading, the self-supported catalysts 5a–d effectively catalyzed Michael reaction of 2-cyclohexanenone and dibenzyl malonate. While 5a and 5b gave very low ee values (6% and 17%, respectively), 5c and 5d gave the Michael product in 88% ee and 96% ee, respectively. Sasai et al. attributed this drastic increase in ee over the bent derivatives (5a and 5b) to the positioning of the dihydroxy groups at the opposite sides of the multitopic ligands. The enantioselectivity afforded by


169

Li

HO HO

OH

O

OH

O

O Al O

n L9a:1,2-disubstituted; L9a:1,3-disubstituted L9a:1,4-disubstituted, L9d:nobenzenering

5a-d

O

O

5a-d + CH2(CO2Bn)2

H CO2 Bn CO2 Bn

Scheme 7.7 Synthesis of self-supported Al–Li–bis(binaphthoxide) (ALB) catalysts for asymmetric Michael reactions

this immobilization technique is significant because previous attempts of immobilizing the ALB type catalyst onto polystyrene resin were unsuccessful and gave no enantioselectivity.22 The heterogeneity of catalyst 5d was confirmed by testing the clear supernatant solution which exhibited no catalytic activity. Sasai et al. also demonstrated the reusability of the self-supported catalysts. After removal of the supernatant and recharging of substrate under argon, it was determined that 5d could be reused three times with gradual deterioration of enantioselectivity (96% ! 85% ee). By linking L9d with Ti(IV) centers, Sasai et al. also demonstrated the synthesis of selfsupported catalysts for carbonyl–ene reactions.23 Two equivalents of Ti(O–iPr)4 in toluene and four equivalents of H2O were added to a solution of L9d in CH2Cl2 to generate a precipitate which was inferred to have an idealized structure of 6 (Scheme 7.8) by elemental analysis and IR spectra. 6 catalyzed the carbonyl–ene reaction of ethyl glyoxylate and a-methylstyrene to give the product 7 in 81% yield and 90% ee. Catalyst 6 was readily recovered in air and used for up to four times without loss of activity or enantioselectivity (88% yield, 88% ee). Ding et al. independently carried out the carbonyl–ene reaction with self-supported catalyst 6, and observed a similar level of enantioselectivity as Sasai et al. Ding et al. further showed that a variation of the self-supported catalyst 6 (i.e., Kagan–Uemura type catalysts24) was able to catalyze the asymmetric sulfoxidation of sulfides.25 In the presence of these self-supported catalysts, aryl methyl sulfides were oxidized by cumene hydroperoxide to chiral sulfoxides with excellent ee values (99.9–96.4%). This catalyst system was also very stable and reused for up to eight times without loss of enantioselectivity or activity. Ding et al. used a similar strategy to prepare self-supported Shibasaki’s BINOL/LaIII catalyst for enantioselective epoxidation of a,b-unsaturated ketones (Scheme 7.9).26 Catalytically active precipitates were obtained by treating multitopic BINOL ligands L10a–i with La(OiPr)3 and triphenylphosphine oxide in THF.27 The La/L10a system, for example, catalyzed the epoxidation of a,b-unsaturated ketone 8a by cumene oxide

170


O

O O

O

O O

Ti

O

O

O

O

O

OH

20mol% 6

+ EtO2C

O Ti

6

Ti O

Ti

Ph

H

81%,90%ee

Ph

EtO2 C

7

Scheme 7.8 Synthesis of the Ti/L9d self-supported catalyst for carbonyl–ene reactions

f:

a: spacer OH OH

HO HO

b: c: d:

L10a-i

single bond

i:

g: h:

e:

Scheme 7.9 List of bis(BINOL) ligands for the synthesis of self-supported Shibasaki’s catalysts

at >91% yield and up to 97.9% ee. The enantioselectivity of the La/L10 system seems to be sensitive to the spacer geometry and length. When a linear spacer was used, the ee increased as the length of the spacer increased (entries 1–4, Table 7.6). The ee dropped as the extension angle of the spacers was reduced (e.g., L10e and L10g). Ding et al. further showed that the La/L10a system was able to catalyze the epoxidation of a variety of a,b-unsaturated ketones with high ee values. The La/L10 system is also readily recoverable and reusable. For example, the La/L10a system was reusable for the epoxidation of 8a for six times to afford 9a in 99–83% yield and 96.5–93.2% ee. Furthermore, leaching of lanthanum was minimal and determined by ICP to be 99% conv.

R NHAc 11a-c

10a-c R = H (a), CH3 (b), Ph (c) spacer

N O

Rh/L11a-c (1 mol%)

O [Rh]

P

P

O

O

Ph

N

10d

NHAc

H2, 40 atm, toluene, RT >99% conv.

Ph

NHAc 11d

n

Rh/L11a-c a: single bond

b:

c:

Scheme 7.10 Self-supported Rh/MonoPhos catalysts for asymmetric hydrogenation of ß-substituted dehydro-a-amino acid 10a–c and enamide 10d. Reproduced with permission from X. Wang and K. Ding, Self-Supported Heterogeneous Catalysts for Enantioselective Hydrogenation, J. Am. Chem. Soc., 2004, 126, 34, 10524–10525. Copyright 2004 American Chemical Society

Rh/L11a–c system was supported by the lack of catalytic activity by the supernatant, low Rh leaching (99%). Recycling of the fluorous layer showed a drop in catalytic activity but the enantioselectivity was not affected. Next, Sinou and coworkers have investigated the asymmetric transfer hydrogenation of ketones using the above fluorous salen ligands 1, 4 and 6 (Scheme 8.4).20

OH

O R

[Ir(COD)Cl]2/1,4,6 or 7

R

iPrOH/D-100/KOH R = CH3; C2H4 or iPr

H H NH HN C8F17

C8F17 C8F17

C8F17 7

Scheme 8.4 Asymmetric transfer hydrogenation

In their study it was shown that the Ir complexes catalyze the reduction of acetophenones with moderate enantioselectivities under fluorous biphasic condition (Galden D-100, mainly perfluorooctane/isopropanol) using the organic solvent as a hydrogen donor. Recycling of the fluorous phase indicated an almost complete loss of catalytic activity

Fluorous Chiral Catalyst Immobilization

185

via decomposition of the fluorous Ir catalyst. Therefore, a more stable fluorous diamine 7 was synthesized without chelating OH functionalities and utilized in the same transfer hydrogenation of acetophenone.21 The reported enantioselectivity was higher (79% ee) than that with the former fluorous catalysts; furthermore, significantly improved recyclability of this fluorous catalyst system has also been claimed. The reuse of the catalytically active fluorous phase has been demonstrated for up to four cycles with a constant decrease of catalytic activity but with moderate loss of iridium (4% after first cycle, then 97%) and good to excellent diastereoselectivities (up to 29 : 1 dr) were obtained for the Michael adducts. Of greater importance is the efficient and convenient recovery of the catalyst using fluorous solid phase extraction. The recovered catalyst retained its capacity to promote a high level of enantioselectivity even after six cycles, although the activity of the recycled material is slowly decreased. Recently, a related fluorous dialkyl prolinol 19 was prepared and evaluated in a CBS reaction by Funabiki group.35 The catalytic activity and, to some extent, the ee values were lower than that of the diaryl counterpart 17, even though the catalyst can be recycled in a simple filtration with 84% efficiency. The development and applications of fluorous pyrrolidine sulfonamides 20 to promote enantioselective Michael additions and aldol reactions were described by Wang.36,37 An interesting and common aspect of their procedures is that water was used as a solvent. Using 10 mol% organocatalyst 20, the Michael addition of ketones and aldehydes to nitroolefins afforded adducts with high yields (>80%) and high enantio- and diastereoselectivity (ee up to 95% and dr up to 50 : 1). They also demonstrated that the catalyst could be recycled after fluorous solid phase extraction. Its catalytic performance remained excellent without marked loss of enantio- and diastereoselectivity until the fourth reuse, although a constant decrease of activity was reported. Recent extension of 20 toward aldol reactions by the same group further illustrated the utility of this catalyst (Equation 8.9). O

O + H

R R'

O

10 mol% 20 0 oC, H2O X

OH (eq. 9)

R R'

X

The reactions were conducted ‘on water’ to yield a variety of aldol adducts with high enantio and diastereomeric purity (up to 98% ee, dr >20). Notably, using again F-SPE, the recycling could be repeated up to seventh run with no apparent loss of selectivity. A novel group of ligands for asymmetric catalytic processes, especially enantioselective malonic ester allylation, was developed by Mino and coworkers using proline-based aminophosphines (Scheme 8.8).38 Because of the high potential of these ligands, they developed a fluorous version 21 having two ponytails. Favorable results were obtained with ee up to 97% using an optimized homogeneous reaction condition in diethyl ether. According to their results, the catalyst system recovery was effected by the precipitation of fluorous material from the reaction mixture using cold hexanes. It is worth mentioning that the recovered and recycled catalyst was actually a palladium enolate complex, with the proposed structure 22, which seemed to be enough stable to be reused up to five times.

190

Recoverable and Recyclable Catalysts OR

N

N MeO

PPh2

OR 21

R=CO(CF2)10CF3 EtOOC

COOEt

OAc Ph

Ph

+ EtOOC

COOEt

Ph

[Pd(η3-C2H5)Cl]2, base

EtO

Ph

OR O

EtO

O

Pd+

N

N Ph2P

OR

OMe

22

Scheme 8.8 Fluorous palladium catalyst for asymmetric allylic alkylation

Beside the above popular chiral amine scaffolds, several other fluorous amine-based ligands and organocatalysts were developed in the last 10 years. Asymmetric addition of dialkyl zinc reagent to aldehydes was of interest from a synthetic point of view (Equation 8.5). Historically, the first fluorous catalyst for that procedure was described by van Koten group.39 Using fluorous ethyl zinc aminothiolates 23a–c (Scheme 8.9), they could reach up to 94% ee running the reaction in hexanes. More interestingly, the reaction can be performed in a fluorous biphasic system with a high level of enantioselectivity (up to 92% ee). They also showed that both the activity and enantioselectivity were significantly better than that of the nonfluorous parent catalyst. Furthermore, the catalytic system could be recycled three times after phase separation. Ph Me2Si CnF2n+1(CH2)2

NR2 S Zn Et R = Me, n = 6 R = Me, n = 10 R = -(CH2)4-, n = 10 23

N

R1

OH

R1 = C(CH2CH2C6F13)3 R1 = Si(CH2CH2C6F13)3 24

Scheme 8.9 Fluorous catalysts for enantioselective diethyl zinc addition to aldehydes

Fluorous ephedrine 24a,b analogs were also developed by Nakamura and Takeuchi and evaluated in the same reaction.40 Although the reported enantioselectivity (up to 84% ee) was lower than that in the previous study, the ligand could be recycled ten more times without any deleterious effect on yield and enantioselectivity using reverse phase fluorous silica gel.


191

Curran and coworkers have investigated organocatalytic Diels–Alder reactions using the fluorous analog of MacMillan catalyst 25 (Scheme 8.10).41 Importantly, the efficiency of the organocatalytic process was not jeopardized by the presence of the perfluorinated tag. Even a slightly better catalyst performance (93% ee forendo adduct and 93 : 7 endo/exo ratio) was reported than that obtained with the parent catalytic system (90% ee for endo adduct and 90 : 10 endo/exo ratio). The organocatalyst was recovered by F-SPE in 84% yield with very high purity, therefore it can be directly reused in the next catalytic procedure. C8F17

N 25

Ph

(10 mol%) • HCl

H

+

N H

O

+ O

CH3CN-H2O, 25°C, 40 h

CHO

HO

Scheme 8.10 Fluorous secondary amine catalyst for enantioselective Diels–Alder reaction

Recently, a recyclable fluorous valine derivative 26 has been developed by Kocovsky and Malkov by appending a perfluorinated tag into a peripheral position (Scheme 8.11).42 This chiral fluorous Lewis base proved to be an effective catalyst for the asymmetric reduction of imines derived from acetophenone and its congeners with high enantioselectivity (up to 89% ee). The fluorous catalyst 26 was recovered in an undemanding fluorous solid phase extraction and recycled three more times without significant loss in activity. Since the nonfluorous parent catalyst was only negligible more active, the fluorous labeling could considerable enhance the potential of this catalytic procedure.

N R1

R2

Cl3SiH f-catalyst toluene, rt

HN

H N

R2 N

R1 H

O

O O

C6F13 26

Scheme 8.11 Fluorous chiral base-catalyzed asymmetric imine reduction

Finally, the extension of fluorous methodology to recover a chiral phase transfer catalyst has been reported by Maruoka and coworkers (Scheme 8.12).43 In their example, a fluorous binaphthyl-based spiro ammonium salt catalyst 27 was prepared and utilized in the asymmetric alkylation of a glycine derivative. The phase transfer catalyst was not fully soluble in the applied aqueous KOH/toluene biphasic system, nevertheless, it could promote the alkylation with high level of enantioselectivity (up to 93%). The separation of the catalyst followed by its reuse was achieved by fluorous liquid–liquid extraction using FC-72 as a fluorous solvent.

192

Recoverable and Recyclable Catalysts (C8F17H2CH2C)Me2Si (C8F17H2CH2C)Me2Si

SiMe2(CH2CH2C8F17) SiMe2(CH2CH2C8F17)

Br N

(C8F17H2CH2C)Me2Si (C8F17H2CH2C)Me2Si

SiMe2(CH2CH2C8F17) SiMe2(CH2CH2C8F17) 27

O Ph2C N

O

chiral catalyst OtBu

+

RX

PTC conditions

Ph2C N

OtBu

H R

Scheme 8.12 Enantioselective fluorous phase transfer catalysis

8.3.2 Fluorous Oxygen Ligands Fluorous BINOL derivatives 28, 29 were developed and tested by Takeuchi and coworkers for asymmetric addition of diethylzinc to aromatic aldehydes (Scheme 8.13).44,45 It was established that both fluorous BINOL compounds were effective in the titanium-catalyzed addition reactions (ee values up to 85%); however, different fluorous separation methods were applied for the recovery and recycling of fluorous ligands. The less fluorous ligand 28 O H

H OH

Ti(Oi Pr)4 (C2H5)2Zn R2 R1 OH OH R1 R2

28, R1 = Si(CH2CH2C6F13)3, R2 = H 29, R1 = Si(CH2CH2C8F17)3, R2 = H 30, R1 = C8F17, R2 = C8F17 31, R1 = C4F9, R2 =C4F9 32, R1 = C8F17, R2 = H

Scheme 8.13 Application of fluorous BINOLs


193

could only be efficiently separated by fluorous solid phase extraction, but it can be reused without further purification in four consecutive reactions. In contrast, the higher fluorine content of ligand 29 allowed its near quantitative recovery with not only fluorous SPE but also in fluorous biphasic system (toluene–hexane/FC-72). In this case, the fluorous layer was separated and reused showing only a slight activity decrease in the third cycle. At about the same time, Chan’s group described the synthesis and application of fluorous BINOL ligand 30 with only four perfluorooctyl groups directly attached to the chiral scaffold.46 In the same reaction (Scheme 8.13), their fluorous catalyst system promoted the formation of adducts with only moderate ee values (37–58% ee). However, the catalyst system could be recycled up to nine times in hexane/perfluoro(methyldecalin) due to its high fluorous phase affinity. Later, they evaluated other less fluorous ligands 31 in the same reaction.47 They concluded that the lower fluorine content has an advantageous effect on the enantioselectivity, although only the highly fluorous 30 could be effectively used in fluorous biphasic condition. As reported, the low fluorine content of fluorous BINOL 32 limited its utility in catalyst immobilization because of its significant leaching during fluorous solid phase extraction.48 To solve this problem, a simple and efficient solid phase separation of the catalysts system was achieved in the asymmetric allylation of benzaldehyde using allytributyl tin (Equation 8.10). Stuart and coworkers showed that a nonpolar fluorous BINOL–Sn complex formed in this reaction which accounts for the possibility to afford clean separation.49 After recovering fluorous BINOL 32 from its BINOL–Sn polymer complex, the BINOL ligand could be reused in three further catalytic runs with negligible fall of catalytic activity. O

OH H

SnBu3

+

Ti(OiPr)4/Ligand 32

(eq. 10)

0°C, 6h

In their systematic work, Ando and coworkers developed fluorous chiral diols with a different chiral scaffold for the above asymmetric dialkyl zinc addition reactions. Their intention was not only to phase label a known ligand, but also to exploit the electron withdrawing nature of the perfluorinated tag in catalyst design. Interestingly, the acidity of hydroxyl group of perfluoroalkyl carbinols is markedly higher (10 000 times) than a normal hydroxyl group. This fluorine substitution also implies a dramatic change of other chemical properties such as stability toward elimination and oxidation and forming more Lewis acidic metal complexes. On this basis, the biaryl 3350 and the fluorous TADDOL-like analog 3451,52 were prepared and evaluated in asymmetric additions to aldehydes (Scheme 8.14). C8F17 O

C7F15

H OH

HO H

C7F15

33

C8F17 OH

O C8F17 34

Scheme 8.14 Fluorous chiral diols

OH H

194


As reported, these ligands showed excellent catalytic activity in the asymmetric addition not only of Et2Zn but also of Me2Zn (up to 98% ee) to aromatic and aliphatic aldehydes. These results are encouraging since a rather limited number of catalyst can catalyze efficiently that challenging but synthetically important reaction. The fluorous biaryl ligands 33 were separated in repeated extractions of the quenched reaction mixture with perfluorohexane and reused without further purification. Based on its very low solubility in toluene, the highly fluorous diol 34 could be recovered and recycled up to four times without using a fluorous solvent or fluorous solid phase. 8.3.3 Phosphorous Ligands Phosphorous ligands are ubiquitous in homogeneous catalysis and also asymmetric transformations having a rather broad field of application. From an industrial perspective, poor recovery of the expensive chiral phosphine-based catalysts from the reaction mixture and also contamination of products by heavy metals often hinders their use. Therefore, the immiscibility of the fluorous phase with most organic solvents and products improves the prospect of easy separation of a fluorous phosphine-based metal catalysts from the reaction mixture. Several chiral phosphorous ligands with perfluorinated domain were developed for supercritical CO2 application, to obtain a higher solubility and better catalytic activity for chiral Rh- or Ir-based catalysts in scCO2.53,54 The first truly fluorous application of chiral phosphine was described by Pozzi and Sinou who reported the synthesis of the fluorous MOP analog 35 (Scheme 8.15).55 This ligand has been prepared by the coupling of three fluorous tags to the free hydroxyl groups of the chiral skeleton at the late stage of the synthetic route. The catalyst, generated by adding [Pd(C3H5)Cl]2 to the chiral phosphine, was active in asymmetric allylic alkylation (Equation 8.1) in toluene and also in benzotrifluoride: the best enantioselectivity was obtained for dimethyl malonate in toluene (87% ee, 88% yield). It has to be mentioned that the nonfluorinated MOP gave the same alkylated product in 95% yield and 99% ee under identical conditions. Although this fluorous ligand showed little affinity to the fluorous phase, complete removal of the fluorous ligand and its corresponding palladium complexes from the toluene phase was

R2

R1 P R3

2

R2 35, R1, R3 = O-CH2C7F15, R2 = H, 36, R1 = H, R2 = C8F17, R3 = OMe 37, R1 = H, R2 = Si(CH2CH2C6F13)3, R3 = OMe

Scheme 8.15 Fluorous MOP ligands


195

achieved with liquid–liquid extraction using perfluorooctane. Unfortunately, the recovered catalyst lost its activity probably due to the oxidation of the phosphine to phosphine oxide during work-up. Sinou and coworker have examined two other fluorous MOP analog 36, 37, having fluorous phase tags on the naphthyl ring.56 These fluorous functionalizations seem not to be advantageous, since the palladium complexes of 36 and 37 showed low enantioselectivities in the above reaction (ee values are lower than 37 and 24%). Moreover, application of strong bases was required to activate dimethyl malonate. The most interesting feature of their work was that the ‘pseudo enantiomer’ ligands 36 and 37 afforded the same configuration for the alkylated product. Two fluorous BINAP analogs 38, 39 were developed by Stuart and coworkers.57 The fluorous tag was attached to the naphthyl ring with or without insulating ethylene spacer (Scheme 8.16). These ligands formed active catalyst in methanol for asymmetric hydrogenation of dimethyl itaconate when combined with [RuCl2(C6H6)]. R1 20 bar H2 RuCl2(C6H6) 38 or 39

O O

O O

MeOH

PPh2 PPh2

O O

O O

R1 38, R1 = C6F13 39, R1 = CH2CH2C6F13

Scheme 8.16 Enantioselective reduction of dimethyl itaconate with fluorous BINAP RuCl2 complexes

The observed stereoinductions were practically the same as those obtained with BINAP (38, 39, and BINAP gave equally 95% ee). Although the enantioselectivities were not dependent on the fluorous substitution of the naphthyl ring, the electron withdrawing phase tag did affect the catalyst activity. The conversion of the substrate was measured after 15 min, and gave 42, 83 and 88% respectively. These results indicate that the catalytic activity can be almost restored applying an ethylene spacer between the perfluorinated tag and the naphthyl ring. The recovery and recycling of these fluorous ligands 38, 39 was achieved in a next study by fluorous solid phase extraction.58 However, an inert atmosphere was necessary during the separation to prevent the perfluoroalkylated BINAP ligands from oxidation. Sinou and coworkers have investigated the application of the fluorous BINAP analog 40 with four ponytails. In their test reaction, it was shown that its Pd complex efficiently catalyzed the asymmetric Heck reaction of 2,3-dihydrofuran with aryl triflates (Scheme 8.17).59 The conversion was complete using 4-chlorophenyl triflate, and the 2,3-dihydrofuran derivative was formed in very high selectivity, although the observed enantioselectivity was

196

Recoverable and Recyclable Catalysts R1

Cl Cl

OTf +

P(Ph-4-R2)2 P(Ph-4-R2)2

Pd(OAc)2, iPr2NEt, 40 or 41

+ O

O

Cl

R1 O

40, R1 = H, R2 = OCH2C7F15 41, R1 = Si(CH2CH2C6F13)3, R2 = H

Scheme 8.17 Fluorous BINAP analogs and application to an asymmetric Heck reaction

moderate (54% ee). This ligand still did not contain enough fluorine to be used with success in fluorous biphasic catalysis. However, the fluorous phosphine and its palladium complex could be quickly separated from the product at the end of the reaction using liquid–liquid extraction with a perfluorinated solvent. Nakamura reported the synthesis and application of 41 as a highly fluorinated analog of BINAP with six fluorous phase tags (Scheme 8.17).60 Synthesis of 41 relied upon the availability of a fluorous silyl group having three ponytails which was directly attached to the naphthyl ring. Ligand 41 was evaluated in the same asymmetric Heck reaction as depicted above. Although the selectivity was lower than for ligand 40, the enantioselectivity obtained was high. In addition, this ligand could be used in a fluorous biphasic system because of its higher fluorine content. However, attempts to recycle the fluorous phase containing the catalyst failed due to the oxidation of the ligand 41. All the above described fluorous BINAP analogs 38–41 have shown a significant sensitivity toward oxidation. This is an often-stated reason for the failure of recovery and reuse of these fluorous BINAP complexes. The prevention of catalyst oxidative degradation has recently been realized in a simple and elegant manner by Bannwarth.61 Adsorbing onto fluorous silica gel, the fluorous BINAP 41 Ru complex could be noncovalently immobilized and reused several times in catalytic asymmetric hydrogenations (up to five runs). This novel protocol prevents sensitive Ru–41 catalyst from air contact, thus minimizing its oxidation. Furthermore, the ruthenium leaching into the product was very low ranging between 1.6 to 4.5 ppm. Surprisingly, the silica gel support employed, having a 500 A pore size, had a favorable influence on the catalyst performance: it significantly increased the catalytic activity during dimethyl itaconate reductions.

8.4 Summary The above results highlight the increasing potential of fluorous chemistry in chiral catalyst immobilization but also demonstrate some existing difficulties and problems. As for other new and innovative methodologies with possible industrial potential, the development of fluorous chemistry and its foreseen advantages raise new questions and concerns, including the safety and toxicity of the fluorous materials employed, and the cost and scalability of fluorous compounds synthesis.


197

References 1. T. Horvath, J. Rabai, Science 1994, 266, 72. 2. T. Horvath, Acc. Chem. Rev. 1998, 31, 641. 3. Handbook of Fluorous Chemistry, Eds. J. A. Gladysz, D. P. Curran, I. T. Horvath, Wiley VCH, 2004. 4. M. Wende, R. Meier, J. A. Gladysz, J. Am. Chem. Soc. 2001, 123, 11490. 5. K. Ishihara, S. Kondo, H. Yamamoto, Synlett 2001, 1371. 6. M. Wende, J. A. Gladysz, J. Am. Chem. Soc. 2003, 125, 5861. 7. L. V. Dinh, J. A. Gladysz, Angew. Chem. Int. Ed. 2005, 44, 4095; Angew. Chem.2005, 117, 4164. 8. (a) C. C. Tzschucke, C. Markert, H. Glatz, W. Bannwarth, Angew. Chem. Int. Ed. 2002, 41, 4500; Angew. Chem.2002, 114, 4678. (b) C. C. Tzschucke, W. Bannwarth, Helv. Chim. Acta 2004, 87, 2882. (c) C. C. Tzschucke, V. Andrushko, W. Bannwarth, Eur. J. Org. Chem. 2005, 5248. 9. (a) Biffis, M. Zecca, M. Basato, Green Chem. 2003, 5, 170. (b) A. Biffis, M. Braga, M. Basato, Adv. Synth. Catal. 2004, 346, 451. 10. E. G. Hope, J. Sherrington, A. M. Stuart, Adv. Synth. Catal. 2006, 348, 1635. 11. F. O. Seidel, J. A. Gladysz, Adv. Synth. Catal. 2008, 350, early view. 12. Studer, S. Hadida, R. Ferrito, S. Y. Kim, P. Wipf, D. P. Curran, Science 1997, 275, 823. 13. G. Pozzi, I. Shepperson, Coord. Chem. Rev. 2003, 242, 115. 14. F. Fache, New J. Chem. 2004, 28, 1277. 15. Catalytic Asymmetric Synthesis, Ed. I. Ojima, Wiley VCH, 1993, pp. 159. 16. G. Pozzi, F. Cinato, F. Montanari, S. Quici, Chem. Commun. 1998, 877. 17. M. Cavazzini, A. Manfredi, F. Montanari, S. Quici, G. Pozzi, Eur. J. Org. Chem. 2001, 4639. 18. M. Cavazzini, S. Quici, G. Pozzi, Tetrahedron 2002, 58, 3943. 19. Shepperson, M. Cavazzini, G. Pozzi, S. Quici, J. Fluorine Chem. 2004, 125, 175. 20. D. Maillard, C. Nguefack, G. Pozzi, S. Quici, B. Valade, D. Sinou, Tetrahedron: Asymmetry 2000, 11, 2881. 21. D. Maillard, G. Pozzi, S. Quici, D. Sinou, Tetrahedron 2002, 58, 3971. 22. K. Ghosh, P. Mathivanan, J. Cappiello, Tetrahedron: Asymmetry 1998, 9, 1. 23. Bayardon, D. Sinou, Tetrahedron Lett. 2003, 44, 1449. 24. R. Annunziata, M. Benaglia, M. Cinquini, F. Cozzi, G. Pozzi, Eur. J. Org. Chem. 2003, 1191. 25. B. Simonelli, S. Orlandi, M. Benaglia, G. Pozzi, Eur. J. Org. Chem. 2004, 2669. 26. Bayardon, D. Sinou, J. Org. Chem. 2004, 69, 3121. 27. J. Bayardon, O. Holczknecht, G. Pozzi, D. Sinou, Tetrahedron: Asymmetry 2006, 17, 1568. 28. R. Kolodziuk, C. Goux-Henry, D. Sinou, Tetrahedron: Asymmetry 2007, 18, 2782. 29. Biffis, M. Braga, S. Cadamuro, C. Tubaro, M. Basato, Org. Lett. 2005, 7, 1841. 30. J. K. Park, H. G. Lee, C. Bolm, B. M. Kim, Chem. Eur. J. 2005, 11, 945. 31. Z. Dalicsek, F. Pollreisz, Á. G€om€ory, T. Soo´s, Org. Lett. 2005, 7, 3243. 32. Q. Chu, M. S. Yu, D. P. Curran, Org. Lett. 2008, 10, 749. 33. H. Cui, Y. Li, C. Zheng, G. Zhao, S. Zhu, J. Fluorine Chem. 2008, 129, 45. 34. Zu, H. Li, J. Wang, X. Yu, W. Wang, Tetrahedron Lett. 2006, 47, 5131. 35. S. Goushi, K. Funabiki, M. Ohta, K. Hatano, M. Matsui, Tetrahedron 2007, 63, 4061. 36. Zu, J. Wang, H. Li, W. Wang, Org. Lett. 2006, 8, 3077. 37. L. Zu, H. Xie, H. Li, J. Wang, W. Wang, Org. Lett. 2008, 10, 1211. 38. T. Mino, Y. Sato, A. Saito, Y. Tanaka, H. Saotome, M. Sakamoto, T. Fujita, J. Org. Chem. 2005, 70, 7979. 39. H. Kleijn, E. Rijnberg, J. T. B. H. Jastrzebski, G. van Koten, Org. Lett. 1999, 1, 853. 40. Y. Nakamura, S. Takeuchi, K. Okumura, Y. Ohgo, Tetrahedron 2001, 57, 5565. 41. Q. Chu, W. Zhang, D. P. Curran, Tetrahedron Lett. 2006, 47, 9287. 42. V. Malkov, M. Figlus, S. Stoncius, P. Kocovsky, J. Org. Chem. 2007, 72, 1315. 43. S. Shirakawa, Y. Tanaka, K. Maruoka, Org. Lett. 2004, 6, 1429. 44. Y. Nakamura, S. Takeuchi, Y. Ohgo, D. P. Curran, Tetrahedron Lett. 2000, 41, 57. 45. Y. Nakamura, S. Takeuchi, K. Okumura, Y. Ohgo, D. P. Curran, Tetrahedron 2002, 58, 3963. 46. Y. Tian, K. S. Chan, Tetrahedron Lett. 2000, 41, 8813.

198 47. 48. 49. 50. 51. 52. 53. 54. 55. 56. 57. 58. 59. 60. 61.

Recoverable and Recyclable Catalysts Y. Tian, Q. C. Yang, T. C. W. Mak, K. S. Chan, Tetrahedron 2002, 58, 3951. Y.-Y. Yin, G. Zhao, Z.-S. Qian, W.-X. Yin, J. Fluorine Chem. 2003, 120, 117. J. Fawcett, E. G. Hope, A. M. Stuart, A. J. West, Green Chem. 2005, 7, 316. Omote, N. Tanaka, A. Tarui, K. Sato, I. Kumadaki, A. Ando, Tetrahedron Lett. 2007, 48, 2989. Y. S. Sokeirik, H. Mori, M. Omote, K. Sato, A. Tarui, I. Kumadaki, A. Ando Org. Lett. 2007, 9, 1927. Y. S. Sokeirik, A. Hoshina, M. Omote, K. Sato, A. Tarui, I. Kumadaki, A. Ando, Chem. Asian J. 2008, 3, 1850. G. Franciò, W. Leitner, Chem. Commun. 1999, 1663. S. Kainz, A. Brinkmann, W. Leitner, A. Pfaltz, J. Am. Chem. Soc. 1999, 121, 6421. M. Cavazzini, S. Quici, G. Pozzi, D. Maillard, D. Sinou, Chem. Commun. 2001, 1220. D. Maillard, J. Bayardon, J. D. Kurichiparambil, C. Nguefack-Fournier, D. Sinou, Tetrahedron: Asymmetry 2002, 13, 1449. D. J. Birdsall, E. G. Hope, A. M. Stuart, W. Chen, Y. Hu, J. Xiao, Tetrahedron Lett. 2001, 42, 8551. E. G. Hope, A. M. Stuart, A. J. West, Green Chem. 2004, 6, 345. J. Bayardon, M. Cavazzini, D. Maillard, G. Pozzi, S. Quici, D. Sinou, Tetrahedron: Asymmetry 2003, 14, 2215. Y. Nakamura, S. Takeuchi, S. Zhang, K. Okumura, Y. Ohgo, Tetrahedron Lett. 2002, 43, 3053 J. Horn, W. Bannwarth, Eur. J. Org. Chem. 2007, 2058.

9 Biphasic Catalysis: Catalysis in Supercritical CO2 and in Water Simon L. Desset and David J. Cole-Hamilton EaStCHEM, School of Chemistry, University of St. Andrews, St. Andrews, KY16 9ST Fife, UK

9.1 Introduction Water and carbon dioxide are both ubiquitous and naturally occurring but the common grounds stops here. Water is present on 71% of the surface of the globe and is apparently abundant, but access to drinkable water is still a day-to-day struggle for millions of individuals.1 Carbon dioxide on the other hand is essential for photosynthesis and is emitted during respiration, but it is also produced everyday from the burning of fossil fuels and fermentation. Significant cutting of its emission constitutes one of today’s biggest scientific, technological and political challenges. From a physical chemistry point of view, water and CO2 also stand at the antipodes. Water, liquid under atmospheric conditions, displays a very large heat of evaporation while CO2 is gaseous and condenses only at 78 C (1 bar). More importantly, they lie at the two extremes of the polarity scale. Water is highly polar and readily dissolves salts and polar organic compounds while CO2 is nonpolar and is able to dissolve hydrophobic compounds, when pressurized. Nevertheless, for process engineers and chemists, these two compounds have two tremendous advantages; they are both cheap and safe. This has been the main driving force to develop their use as solvents for chemical production. The majority of


200


chemical reactions are currently carried out in organic solvents. However, many of these solvents are inflammable or explosive as well as volatile (VOCs) and hence subject to release into the atmosphere contributing to photochemical smog and to global warming. They may also cause health-related problems because of their noxious, toxic or carcinogenic nature. Emissions from solvents are now the largest sources of atmospheric VOCs in the UK, accounting for 25% of the 1.36 M tonnes emitted annually.2 Growing awareness and concern regarding the environmental impact of industrial activity have given impetus to the replacement of classical organic solvents. In this context, water and carbon dioxide have been developed as ‘green’, renewable and safe alternatives.3,4 While it may be desirable for a solvent to ensure intimate contact of all reagents present, an alternative strategy would be for a combination of solvents to ensure sequestration of different molecular species. This is especially true in organometallic catalysis were the catalyst, starting materials and products are typically dissolved in a single phase. This has been responsible for the success but also is the curse of homogeneous catalysis. Having all the reactive partners molecularly dispersed in the same medium, allows all catalytic sites to be accessible and hence ensures that the reaction is only limited by intrinsic kinetics. However, the homogeneous nature of the reaction mixture renders the separation of the products from the (expensive) dissolved catalyst a challenging operation.5 In some cases, the best (in terms of activity, selectivity and lifetime) catalysts cannot be used because the separation problem has not been solved.6 One approach to address this limitation is to carry out the reaction under biphasic conditions, where the catalyst is dissolved in one phase while the product is present in a second separated phase. Separation of the phases, either continuously or stepwise, allows the catalyst and the product to be easily and efficiently recovered separately. Being two extremes in terms of polarity and volatility, CO2 and water form biphasic mixtures with several other solvents. It is therefore in this context of biphasic catalysis that they have been extensively used and studied.

9.2 Biphasic Catalysis Numerous reactions catalysed by transition metals involve a reactive gas (hydrogenation, hydroformylation, carbonylation, etc.). Thus, while usually referred as ‘homogeneous reactions’ these systems are in fact gas–liquid (G–L) biphasic systems (Figure 9.1). However, if the rate of gas transfer into the liquid phase is fast compared with the intrinsic catalytic reaction rate, the reaction occurs in the bulk of the liquid phase. The liquid-phase concentrations of the gaseous species are given by Henry’s law and the kinetics of the reaction can solely be described in terms of intrinsic reaction rate. In some cases the slow step is the transfer of the gaseous reagent into the solution. These reactions are mass transport limited. In this chapter, for simplicity reasons, we will consider these G–L biphasic systems as homogeneous ones since the multiphasic nature of the reaction is the sole consequence of the physical state of one of the reactants under the reaction pressure and temperature. We will consider reactions as multiphasic when the presence of an additional phase serves the purpose of catalyst–product separation.

Biphasic Catalysis: Catalysis in Supercritical CO2 and in Water

201

Figure 9.1 Schematic representation of different multiphasic systems. Gas–liquid (left); liquid–liquid with only the product being soluble in the second phase (centre), liquid–liquid with both substrate and product soluble in the second phase (right). (Sn (X): substrate n in phase X, Pn (x): product n in phase X)

Among the different possible configurations between catalyst, substrates and products, the ideal situation is when both the catalyst and the starting material are soluble in one phase while the product of the reaction forms a separate phase (Figure 9.1). In this case, no mass transfer limitation occurs, the system behaves as a homogeneous one, with the extra advantage that the product is readily separated from the catalyst. Moreover, by forming a second distinct phase the product is less likely to react further with the catalyst and/or the starting material improving the overall selectivity of the reaction. Examples include the well-known Shell-SHOP process for the oligomerization of ethene7 and the dimerization of butenes using a nickel catalyst immobilized into an ionic liquid (Difasol process).8 This situation is rather rare and usually both starting material and product are preferentially soluble in a different phase from the catalyst. In such a configuration, the transfer of the substrate to the catalyst’s phase influences the reaction rate and can limit the practical application of the biphasic methodology. The aqueous biphasic hydroformylation of alkenes constitutes a typical example of such a system. In this process the catalyst is dissolved in an aqueous phase while the alkene and the produced aldehydes form a separated phase. The methodology is applied commercially for the hydroformylation of propene and butene (Ruhrchemie/Rhoˆne-Poulenc process, RCH/RP). Nevertheless, higher alkenes cannot be used as substrates since mass-transfer brings the reaction rate below any that can be economically viable. When considering multiphasic systems for catalyst recovery, several key parameters have to be carefully examined: .

.

The solvents used have to exhibit minimal mutual solubility to ensure efficient retention of the catalyst. If the solvent dissolving the catalyst displays some solubility in the product phase this would lead to solvent loss during the phase separation and, more seriously, it would carry away the dissolved catalyst. The biphasic system formed by the solvents has to separate quickly and completely during the separation step. Slow phase separation means that some of the catalyst must spend a long time in the separation unit, outside the reaction vessel. An important fraction of the catalyst would therefore be unproductive.

202 .


While the immobilization solvent circulates in a closed loop or ideally never leaves the reactor, the product phase solvent is removed together with the product from which it has to be separated. This separation is most of the time carried out by distillation. Therefore, care must be taken to choose solvents that do not form azeotropes with the products and that display low heat of vaporization.

Water and CO2 have generally opposite roles in biphasic systems. Most organic compounds are generally sparingly soluble in water. Water is therefore most of the time used as the catalyst supporting fluid, forming clean biphasic systems with the reaction products. ScCO2 on the other hand, being able to dissolve apolar compounds, is generally used as the product-dissolving phase. Moreover, products can readily be recovered from CO2 after reaction and separation from the catalyst by simple decompression, circumventing the need for distillation.

9.3 Aqueous Biphasic Catalysis 9.3.1 Introduction Today’s best know and most referred to biphasic process is the aqueous biphasic hydroformylation of alkenes, discovered at Rhone-Poulenc in the mid 1970s9 and developed by Ruhrchemie in the mid 1980s (Scheme 9.1).

Scheme 9.1 Aqueous biphasic hydroformylation of olefins

In this process, a rhodium catalyst is dissolved and immobilized in water via the watersoluble sodium salt of tris-sulfonated triphenylphosphine ligand (TPPTS), 1, while propene and the produced C4 aldehydes form a separated phase. A small portion of the crude reaction mixture is continuously taken out of the reactor via a simple overflow to a gravity separator. There, the mixture quickly splits into two phases, the bottom aqueous phase containing the catalyst is recirculated to the reactor while the products and unreacted starting material are send to downstream separation units.

This process produces virtually only aldehydes (99% selectivity to aldehydes) with very high linear selectivity (l/b ¼ 96 : 4) under moderate conditions (125 C, 50 bar). More


203

importantly, during the production of the first 2 000 000 t of butanal, only 2 kg of rhodium have been taken out with the product, in other words the leaching lies in the ppb range, clearly demonstrating the efficiency of the aqueous biphasic methodology to immobilize the catalyst.10 At present, five plants produce worldwide some 800 000 ton per year of aldehydes using the aqueous biphasic technology.11 This process constitutes the cornerstone of biphasic catalysis. It demonstrates that biphasic systems can be used on an industrial scale, and that they are economically viable and reliable. This has given a huge impetus for extending this methodology and developing new biphasic systems for transition metal-catalysed reactions. Several excellent reviews and books have appeared on aqueous biphasic-catalysis.5,12,13 A huge number of catalytic reactions have been investigated in the aqueous phase as an alternative to organic solvents. However, in many studies, catalyst recycling is not the prime concern. In the following section we will focus on systems were water is used as a catalyst immobilizing phase and allows catalyst recycling and reuse. More specifically, we will discuss the different strategies developed to enhance mass transfer in such systems since it greatly impedes the commercial use of this elegant technique. 9.3.2 Aqueous Biphasic Catalysis: Beyond Mass Transfer While aqueous biphasic hydroformylation technology is highly efficient for the transformation of propene and butene (see above), for less water-soluble substrates, mass transfer brings the reaction rate below any that would be economically viable. For example, under standard reaction conditions (100 C, 50 bar, [Rh] ¼ 1 103 mol dm3) hydroformylation of 1-octene only yields 8% of C9 aldehydes after 24 h of reaction.14 Several different strategies have been developed to extend the aqueous biphasic methodology for the transformation of hydrophobic substrates. 9.3.3 Additives One of the first methods developed for extending aqueous biphasic technology relies on the use of additives as mass transfer promoters. In order to preserve the inherent strength of the process, several parameters have to be carefully examined when considering additives. .

.

. . .

The additive has to be inert toward the catalyst, substrate and staring material. Care must be taken that it does not act as a substrate for the catalytic cycle or react with the starting material or product through catalysed and/or uncatalysed processes. While increasing the solubility of the substrate in the aqueous phase, care must be taken not reciprocally to increase the solubility of the catalyst in the organic phase, since this could result in significant leaching of the catalyst. The influence of the additive on the efficiency of the phase separation should be kept to a minimum. Interactions between the additive and the different catalytic species involved in the cycle should not result in a dramatic decrease in chemo- and/or regio-selectivity. Care must be taken regarding leaching of the additive into the product phase. This could complicate the separation of the different products and starting material. Moreover, if the additive is carried away together with the product, replenishment might be necessary to conserve the promoting effect.

204


The addition of co-solvents such as lower alcohols, acetone and acetonitrile to the water phase was one of the first strategies used to improve the solubility of the substrate in the aqueous phase.15 In this way, the lipophilicity of the catalytic phase is increased, enhancing the solubility of the substrate and hence the rate of mass tarnsfer and the reaction rate. Octene for example is estimated to be 104 times more soluble in 50% aqueous ethanol than in pure water.16 When this solvent mixture was used to carry out hydroformylation of 1-octene, it led to an effective increase of the reaction rate. However, acetal formation due to condensation of the aldehydes with the ethanol was observed together with a transfer of the co-solvent into the product phase.16 A different study, using methanol as a co-solvent also led to a rate increase but a decrease of the linear selectivity was observed.17 According to the authors, this lower linear selectivity is due to the presence of low coordinated phosphine species, i.e. [HRh(CO)2P], the formation of which is facilitated by the decrease of the ionic strength of water induced by the presence of the co-solvent.18 A more recent study comparing methanol, ethanol and n-butanol as cosolvent also showed an increase in activity together with a decrease in linear selectivity.19 In spite of their ability to increase the reaction rate through better solubilization of the substrate in the catalyst phase, the use of co-solvent is still subject to many drawbacks. The decrease of linear selectivity, the partitioning of the co-solvent in the product phase and, in the case of hydroformylation, their reaction with the product, hinders their use for commercial applications. Moreover, no data regarding the leaching of the catalyst in the presence of co-solvent has been reported. Leaching is likely to become important since co-solvents decrease the polarity gap between the two phases. Surfactants are another type of additive that has been widely studied to reduce mass transfer limitations in aqueous biphasic catalysis. By forming micelles above the critical micelle concentration (CMC), surfactants can solubilize organic compounds within the core of the micelles and increase the interfacial area in biphasic system hence promoting the transfer of the substrate between the phases. Interestingly, in some cases it has been demonstrate that, while anionic surfactants do not provide any promoting effects, cationic surfactants greatly increase the reaction rate.19,20 When conducting the hydroformylation of 1-dodecene in the presence of cetyltrimethylammoniumbromide (CTAB), 61% conversion to aldehydes was observed. On the other hand, when sodium dodecyl sulfate (SDS), an anionic surfactant, was used, no aldehyde was detected. According to the authors, both surfactants form micelles and solubilize the substrate. In the case of anionic surfactants, the external surface of the micelle is negatively charged repelling the negatively charged catalyst (Figure 9.2). However, when cationic surfactant is used, there is an electrostatic attraction between the positively charge micelle surface and the ionic catalyst, creating a high micro-concentration of catalyst in the vicinity of the micelle, which contains a high concentration of substrate, and hence increasing the reaction rate. Supporting this hypothesis, interactions between TPPTS and CTAB were evidenced by 31 P-NMR spectroscopy.21 Static light scattering experiments were also consistent with the binding of [HRh(CO)(TPPTS)3] and free TPPTS onto the positively charged micelle surface.22 Rhodium levels in the interfacial layer were found to be ca 90 times higher than that in the bulk of the water phase, also supporting the idea of an interaction between the complex and the micelles.22 Regarding the linear selectivity, the influence of surfactant addition is less marked and somehow less


205

Figure 9.2 Schematic representation of the interaction between [HRh(CO)(TPPTS)3] and micelles formed by cationic and anionic surfactants

consistent. Some authors reported an increase in the linear selectivity20,23 while others noticed a decrease.19 Since the addition of CTAB enables high reaction rates (TOF values up to 900 h1 have been reported24) several other cationic surfactants have been studied (Figure 9.3).25,26,27 The use of gemini surfactants with alkyl bridges led to slightly higher rates and linear selectivity than CTAB.25 Interestingly, the shorter the alkyl bridge, the more the linear selectivity increased. According to the authors, these surfactants tend to form more compact micelles in which the formation of the less crowded linear aldehyde is enhanced.

Figure 9.3 Structure of the different cationic surfactants used in aqueous biphasic hydroformylation of higher alkenes25–27

206


Among the different gemini surfactants based on piperazine, 2–4, the iodide salt, 2, performed the best.26 On the other hand, among the trimeric surfactant based on triazine, 5–7, the oxalate 6 and chloride 7 salts gave higher rates and selectivity. Double long-chain surfactants (DLCS) display the highest enhancement rates to date. TOF as high as 7500 h1 can be obtained under certain conditions.27 It has been demonstrate that the key parameter to obtain high reaction rates is the length of both chains. The activity jumps once a threshold (when n ¼ 8 if m ¼ 22 and when n ¼ 12 if m ¼ 16, see structures in Figure 9.3) chain length is reached. This is suspected to be due to the formation of vesicles instead of micelles once both chains are long enough.27 Regarding the leaching of the catalyst into the product phase and the efficiency of the phase separation after reaction, only rather few data were available until recently. For the hydroformylation of 1-dodecene in the presence of CTAB, a rhodium concentration of 0.04 ppm in the product phase has been reported.22 Worryingly the authors described the product phase as ‘saturated with water’. The extent of emulsification during hydroformylation promoted by CTAB has been studied as a function of different parameters.28 Emulsification is promoted by high stirring rates and high concentrations of surfactant and typically increases with the conversion. In the worst case, the mixture after reaction is severely emulsified and no separation could be observed at ambient temperature over 4 h. When using [CoCl2(TPPTS)2] instead of the rhodium-based catalyst in the presence of CTAB for the hydroformylation of C8 and C10 olefins, the phases have been reported to take up to 1 h to separate fully.29 Interestingly, in the presence of excess ligand, this time can be reduced to 5 min. When using a gemini surfactant with ethyl bridge, at least 10 min were required for the phases to separate.25 At a P/Rh ratio of 18, the rhodium level in the product phase was found to be 9.8 ppm. Rhodium was no longer detected in the product phase at a P/Rh ratio of 54. The aqueous phase containing the catalyst with the surfactant was recycled 4 times with little decrease in activity and selectivity. In the presence of the trimeric surfactant 6, the phases separated quickly.26 Leaching of the metal in the product phase was found to be 2.27 ppm and the catalyst-containing phase could be reused five times without obvious decrease in activity. When DLCS are used the phases separate forming three distinct layers, a colourless organic phase, a yellow aqueous phase and a brown interfacial emulsion layer.27 When both aqueous phase and interfacial layer were taken for recycling, the system could be recycled seven times with slight decrease in activity. However if only the aqueous phase was used for recycling, the conversion dropped after one recycle from 90% to 60%. This is consistent with the rhodium levels found in the different phases, 0.52 ppm in the product phase, 18.1 ppm in the bottom aqueous phase and 70.3 ppm in the interfacial emulsion layer. Whilst, under certain conditions, the addition of surfactants allows fast reaction with good phase separation and low leaching, those systems tend to be generally quite sensitive to stirring rate, conversion and concentration of the additive. All those parameters have to be carefully monitored during operation to avoid the detrimental formation of emulsions. The use of short chain surfactants based on the methylimidazolium cation may afford a solution to this drawback. Aqueous biphasic hydroformylation of 1-octene in the presence of 1-octyl-3-methylimidazolium bromide, [OctMIM]Br has been shown to proceed with initial TOF of 900 h1.30 The phases were fully separated by the time the autoclave was open, ca 10 min after the end of the reaction, and providing enough excess ligand was used, rhodium leaching as low as 0.49 ppm was recorded. Visual inspection of a biphasic system


207

Figure 9.4 Pictures of a biphasic system formed by 1-octene and water, without additive (top line) and with 0.5 mol dm3 of OctMImBr (bottom line) (octene dyed). (a) Before stirring; (b) after 2 s stirring; (c) after 6 s mixing; (d) after 12 s mixing (stirring is stopped); (e) 2 s after stirring has stopped; (f) 6 s after mixing has stopped; (g) 12 s after mixing has stopped; (h) 20 s after mixing has stopped. S. L. Desset, S. W. Reader and David J. Cole-Hamilton, Aqueous-biphasic hydroformylation of alkenes promoted by weak surfactants, Green Chemistry, 2009, 11, 630–637. Reproduced by permission of The Royal Society of Chemistry

composed of water and octene in the presence of the imidazolium salt shows that very efficient phase mixing is obtained on rapid stirring, but that the phases quickly separate once the mixing is stopped (Figure 9.4).30b The chain length of the alkyl group on the 1-alkyl-3-methylimidazolium cation has been shown to be very important. A hexyl chain gives almost no rate enhancement, whilst a decyl chain leads to the formation of a stable emulsion. The octyl chain seems to maintain a good balance between high rate enhancement and fast phase separation.30 Inverse phase transfer catalysts, compounds able to transfer lipophilic molecules from an organic phase to an aqueous phase, have been use to overcome mass transfer limitations in aqueous biphasic catalysis. The two main classes of such compounds that have been investigated are cyclodextrins and calixarenes. By formation of host–guest complexes, they transfer the substrate into the aqueous phase or into the interfacial layer where it can react with the water-soluble catalyst.31 The product is then released into the organic phase where the supramolecular carrier can trap another molecule of substrate (Figure 9.5). This methodology has received considerable attention and has been use in several aqueous biphasic transition metal-catalysed reactions such as the oxidation of alkenes (Wacker Process),32,33 hydrogenation of a,b-unsaturated acids and dienes,34 deoxygenation of allylic alcohols,35 Suzuki–Miyaura cross-coupling,36 Tsuji–Trost reaction of alkylallylcarbamate and carbonate,37,31b ruthenium-catalysed hydrogenation of aldehydes,38 hydrocarbonylation39 of alkenes and hydroformylation of long-chain alkenes.37e,40 Cyclodextrins (CD) are cyclic oligosaccharides composed of 6 (a-), 7 (b-) or 8 (g-) glucopyranose units. Those compounds are characterized by the shape of a truncated cone were the primary and secondary hydroxyl groups occupy the narrower and the wider rim of the cone respectively (Figure 9.6). It has been proposed that several conditions must be satisfied for CDs to promote catalysed reactions in aqueous biphasic systems.32b CDs

208


Figure 9.5 Schematic representation of the inverse-phase catalysis concept (S ¼ substrate; P ¼ product; Cat ¼ water-soluble catalyst). Reproduced with permission from L. Leclercq , F. Hapiot , S. Tilloy , K. Ramkisoensing , J. N. H. Reek , P. W. N. M. van Leeuwen and E. Monflier, Sulfonated Xantphos Ligand and Methylated Cyclodextrin: A Winning Combination for Rhodium-Catalyzed Hydroformylation of Higher Olefins in Aqueous Medium, Organometallics, 2005, 24, 2070–2075. Copyright 2005 American Chemical Society

have readily to form an inclusion complex with the substrate, this inclusion complex has to be soluble in water, the reactive function of the substrate must still be accessible to the catalyst once enclosed in the CD and the reaction product must dissociate easily from the CD. To fulfil these requirements the nature of the CD (a, b or g) together with the nature and the extent of its chemical modification play decisive roles. b-CDs have been shown to be more efficient than their narrower (a-CD) and wider (g-CD) in enhancing Wacker oxidation32b and the hydroformylation40b of higher alkenes. This has been attributed to a better recognition between the substrate and the CD in the case of the b-CD. Partial or complete substitution of the primary and secondary hydroxyl groups by various

Figure 9.6 Schematic representation of the shape of a- (n ¼ 6), b- (n ¼ 7) and g- (n ¼ 8) cyclodextrins. The protons H3 and H5 are situated inside the cavity whereas the protons H1, H2 and H4 point outwards


209

Table 9.1 Hydroformylation of 1-decene in an aqueous biphasic system in the presence of chemically modified cyclodextrinsa b-CD

Entry

1 2 3 4 5 6 7 8 9

t (h)

Rb

Nbc

S (g dm3)d

— H OMe OMe OMe OAc OAc iPrOH OSO3

— 0 12.6 14 21 14 21 6.3 9

— 18.5 570 (20) 570 (20) 570 (100 330 >400

6 8 8 8 8 8 8 8 8

Conversion (%)

Aldehyde selectivity (%)

l/b

10 19 76 75 30 46 6 32 7

60 78 91 91 57 57 66 84 69

2.7 2.1 1.8 1.9 2.5 2.6 2.6 2.0 2.8

a Adapted from T. Mathivet, C. Meliet, Y. Castanet, A. Mortreux, L. Caron, S. Tilloy & E. Monflier, Rhodium catalyzed hydroformylation of water insoluble olefins in the presence of chemically modified b-cyclodextrins: evidence for ligand–cyclodextrin interactions and effect of various parameters on the activity and the aldehydes selectivity, J. Mol. Catal. A: Chem, 2001, 176, 105–116. Copyright 2001, with permission from Elsevier.40c Conditions: [Rh(acac)(CO)2]: 0.16 mmol; TPPTS: 0.8 mmol; CD: 1.12 mmol; H2O: 45 cm3; decene: 80 mmol; P (CO/H2): 50 bar; T: 80 C. b Cyclodextrin modification. c Average number of substituted hydroxy groups. d Solubility of the CD in water at 25 C. Numbers in brackets correspond to the solubility at 80 C.

substituents (OMe, OAc, O-i-PrOH, OSO3) led to significant variations in yield and selectivity of CD-mediated biphasic reactions.32b–d,40b–c Among the different chemically modified CDs investigated, partially methylated b-CD (12.6 OH group on average replaced by OMe at positions 2,3 and 6) has been found to be the most efficient in enhancing the rate of the aqueous biphasic hydroformylation of decene (Table 9.1)40c This enhancement has been mainly attributed to the high solubility in both aqueous and organic phases of the methylated-b-CD. Supporting this, it is remarkable that while b-CD and sulfonated-b-CD display high solubility in water they are almost insoluble in the organic phase and have little effect on the reaction rate (Table 9.1, Entries 1 and 9). In contrast, permethylated and peracetylated b-CDs are highly soluble in the organic phase but sparingly soluble in the aqueous phase at 80 C. Consequently they display little or no enhancement effect on the reaction rate (Table 9.1, Entries 5 and 7). Interestingly, whereas partially methylated-b-CDs improve the reaction rate, they tend to decrease the linear selectivity (Table 9.1, Entries 1 and 3). Formation of an inclusion complex between TPPTS and b-CD has been observed but its influence on the reaction was not fully understood.41 Further studies showed this inclusion complex was responsible for the lower linear selectivity obtained.42 In the presence of b-CDs (native or randomly methylated), dissociation of TPPTS is facilitated, leading to the formation of low coordinated phosphine species such as [HRh(CO)2TPPTS], which is known to give low l/b ratios. To avoid this, different CD–catalyst combinations have been studied in order to maintain high linear selectivity. When using TPPTS ligand with rhodium or palladium for hydroformylation or Tsuji–Trost reactions of higher alkenes, methylated-a-CD, sulfobutyl ether-b-CD and 2,3-dimethyl-di-O-methyl-6-O-sulfopropylb-CD have all been shown to enhance the reaction rate while having little or no interaction with the catalytic system.40g,37d–e This avoids a decrease in the linear selectivity in hydroformylation reactions and avoids poisoning of the CDs by the ligand. When using

210


sulfonated xantphos 8, instead of TPPTS, in conjunction with methylated a or b CD, interactions between the ligand and the CD have been shown to have positive effects on the reaction.40e Formation of an inclusion complex between the phenyl ring of 8 and the CDs was observed by NMR spectroscopy, but those interactions did not promote the dissociation of the ligand from the metal. Therefore, improved reaction rate together with improved chemo-and regioselectivity were observed in the hydroformylation of 1-octene and 1-decene. Methylation on the ortho position of the phenyl ring of TPPTS also proved to be a fruitful methodology to avoid interaction between the ligand and the methylated-b-CDs.37f

a-CDs bearing ammonium groups have been shown to have a positive effect on the rate and selectivity of the hydroformylation of 1-decene.40f In this case, TPPTS and the CDs interact through ion pairing but do not form inclusion complexes. It has been postulated that the improved linear selectivity was due to the in situ formation of new catalytic supramolecular species by ion-exchange between the anionic ligand and the cationic CD. Regarding the phase separation and leaching of the catalyst, CDs were shown early on not to be detrimental to the process. In the hydroformylation of 1-decene with Rh/TPPTS catalyst in the presence of a large excess of methylated-b-CD, the phase separation was found to be excellent and the rhodium and phosphorus leaching into the product phase were found to be < 0.5 ppm and 1.2 ppm respectively.40b Similar results were reported later and the catalyst was recycled 4 times without noticeable loss in activity.39 Other authors reported an increase of activity upon recycling during the hydroformylation of 1-octene. They attributed this unexpected result to a gradual organization of the interface.40d Even when surface-active CDs such as 2,3-dimethyl-di-O-methyl-6-O-sulfopropyl-b-CD were used, the system was described as strictly biphasic.37e The phases readily separated and the catalyst could be recycled three times without loss in activity. In the hydrogenation of aldehydes, the ruthenium content in the organic phase was found to be 2000 PEG remains solid upon addition of CO2.


239

Preliminary studies have been carried out on using PEG900 in biphasic systems with scCO2 for the hydrogenation of styrene catalysed by [RhCl(PPh3)3].138 After the reaction, the product was extracted and the PEG phase containing the catalyst could be reused for a total of five cycles without noticeable decrease in activity. Rhodium leaching in the extracted product was found to be below 1 ppm. More recently, the selective hydrogenation of a,b-unsaturated aldehydes catalysed by [H2Ru(PPh3)4] was investigated in a PEG/CO2 biphasic system.139 Reduction of the carbonyl group of 3-methyl-2-butenal, citral and cinnamaldehyde proceed smoothly with almost perfect selectivity for the corresponding allylic alcohols in a PEG1000/CO2 biphasic system. Using higher molecular weight PEG seems to have virtually no influence on both conversion and selectivity below a molecular weight of 10 000. For PEG10000 and PEG12000 the conversion decreases but very high selectivity is still observed. The CO2 pressure also has a strong influence on the reaction yield up to 140 bar after which the yield is found to be independent of the CO2 pressure. The effect of both PEG molecular weight and CO2 pressure on the reaction yield seems to be related to an enhancement of the mass transfer with lower molecular weight PEG and high CO2 pressure. Batch recycling of the PEGimmobilized catalyst showed that steady deactivation of the catalyst occurred through four catalytic runs. The selectivity, however, remained constant throughout the recycling. As the ruthenium leaching was found to be very low (65 ppb), the authors suggested that a change in the complex structure may be responsible for the observed decrease in activity rather than the leaching of catalyst. Supporting this, a change in the coloration of the PEG phase was observed during recycling. Oxidation using molecular oxygen in CO2 is attracting increasing attention. The particularly attractive feature of such a combination is the increased safety gained when using CO2 as a diluent as a result of the stability of CO2 toward oxidation. Recently oxidation of alcohols by PEG -stabilized palladium nanoparticles has been studied in scCO2.140 Allylic, benzylic, secondary and some primary alcohols were selectively oxidized in scCO2 to their corresponding aldehydes by the giant palladium cluster [Pd561-phen60(OAc)180] (phen ¼ 1,10-phenanthroline) stabilized by PEG1500. Interestingly, when recycling the PEG-stabilized nanoparticles the conversion profile revealed that the induction period typically observed with fresh catalyst is almost completely suppressed. TEM images of the nanoparticles before and after reaction showed better dispersion of the used particles. According to the authors, the decrease of the PEG viscosity induced by the CO2 present during the reaction may be responsible for this observation. This system was used for the continuous flow oxidation of benzylic alcohol. During the 38 h on stream, the conversion to benzaldehyde showed a slow but steady increase presumably due to an increase in particle dispersion. Interestingly, decreasing the CO2 pressure (132 vs 155 bar) doubled the yield of aldehydes, allowing for ca 50% single-pass conversion. According to the authors, reducing the pressure results in a shift in partition coefficient of the substrate in favour of the PEG phase, thus increasing substrate availability in the catalytic phase. More recently, the same authors further developed this system for continuous flow operation in a fixed bed tubular reactor.141 Instead of liquid PEG, the nanoparticles were stabilized within PEG750 chains covalently attached to the surface of a silica support. The obtained supported catalyst was used for the oxidation of benzyl alcohol under continuousflow conditions over 30 h using scCO2 as a transport vector in a tubular reactor.

240


The conversion was found to be constant over time and TEM pictures of the spent catalyst showed very little aggregation of the particles. Interestingly, catalysts made of palladium particles supported directly on the silica or within a PEG phase supported on the silica surface (but not covalently bound to it) showed only slightly lower activity in batch experiments, but showed significant deactivation when used under continuous flow conditions. Wacker type oxidation of alkenes has recently been studied in a PEG300–scCO2 system.142 In this biphasic system, styrene could be efficiently oxidized by PdCl2 to acetophenone or to benzaldehyde in the presence or the absence of CuCl, respectively. Immobilization of the palladium catalyst into the PEG phase allowed a decrease in the palladium loading compared with other biphasic systems. The stability of the system was evaluated by recycling the PEG phase after extraction of the product by scCO2. The activity only slightly decreased over four batch recycles and the palladium leaching in the product phase was found to be 0.5 ppm. Ionic Liquid–scCO2 Biphasic Systems Ionic liquids (ILs) are salts made up of a bulky organic cation with an organic or inorganic anion melting below 100 C or ideally at room temperature. There has been and still is a tremendous interest in using ionic liquids as catalyst immobilizing phases. The main driving force is their excellent and tunable solvent properties together with their lack of a measurable vapour pressure, making them very interesting candidates for the replacement of volatile organic compounds. A detailed account of ionic liquids and their application for catalyst immobilization is the subject of Chapter 2.1 by J. Xiao. Ionic liquids are highly polar and nonvolatile while scCO2 is essentially nonpolar and has gas-like properties. These diametrically opposite properties make them almost ideal for combined use in a biphasic system. Important work by Brennecke, Beckmann and coworkers showed that, in contrast with permanent gases such as carbon monoxide and hydrogen, scCO2, is highly soluble in ionic liquids such as [BMIM]PF6 (up to 60 mol%).143 Ionic liquids, on the other hand, are essentially insoluble in scCO2. In addition, it was shown that organic compounds could be quantitatively extracted from ionic liquids by using scCO2.144 Those very interesting physico-chemical properties together with the increasing number of examples of efficient use of organometallic catalysts in ILs prompted researchers to investigate catalytic reactions in biphasic IL–scCO2 systems. Initial investigations were carried out in batch systems for the hydrogenation of tiglic acid (85–90 % ee),145 1-decene146 or CO2146 with the products being extracted using scCO2. Those repetitive batch systems demonstrated the viability of the concept and that careful catalyst design (most of the catalyst used were ionic to ensure high solubility in the IL phase while avoiding their extraction in the scCO2) could allow the metal leaching to be very low. Continuous flow operations with the catalyst dissolved in IL remaining in the reactor while the substrates and products are transported by scCO2 were developed almost simultaneously.147,148 The hydroformylation of medium chain alkenes (C8–C12) catalysed by rhodium complexes containing triphenylphosphine analogue ligands 49 (Scheme 9.7) and the asymmetric hydrovinylation of styrene catalysed by Wilke’s complex 50 (Scheme 9.8) were the two first systems investigated under continuous flow conditions in biphasic IL–scCO2 systems.


241

Scheme 9.7 Continuous flow hydroformylation of alkenes in an IL–scCO2 biphasic system

Scheme 9.8 Continuous flow hydrovinylation of styrene in an IL–CO2 biphasic system

These two systems emphasize that achieving continuous flow operation in an IL–scCO2 system requires careful design and tuning of all the parameters of the system. First of all, to ensure efficient retention of the catalyst in the IL phase, the use of a charged catalyst, either an ionic complex, as Wilke’s complex used for the hydrovinylation, or neutral complexes bearing ionic ligands as used for the hydroformylation, is required. The nature of the ionic liquids also plays a key role, so it cannot be solely viewed as a nonvolatile immobilization medium. The hydrovinylation reaction usually required a promoter in order to form the active nickel hydride catalyst (usually alkyl aluminium or alkali BARF salts). Interestingly, in the IL–CO2 system the IL itself could activate the catalyst. Moreover, a strong influence of the IL anion and cation structure on the activity, the chemo- and the enantioselectivity of the catalyst was observed. In the hydroformylation reaction the design of the ionic liquid also proved to be crucial.149 When using [BMIM]PF6 in the presence of phosphite ligand it appears that the PF6 anion reacts with traces of water generated in situ by the aldol condensation–dehydration of the produced aldehydes generating HF and O2PF2 which competes for rhodium coordination with the original ligand. The nature of the IL was also found to have an important effect on the performance of the system by influencing the partitioning of octene between the two phases. Reaction rates were found to be quite low when using [BMIM]PF6 presumably because of the poor solubility of octene in this IL.

242


Increasing the chain length of the alkly chain on the cation to C8 and switching to the bis-triflamide anion, improved the reaction rate by increasing the solubility of the octene in the catalytic phase. Engineering parameters have also been shown to have an important influence in these IL–CO2 systems. Especially in the case of hydroformylation, variation of substrate flow rate, CO/H2 partial pressure and temperature were shown to have complex effects on the performance of the system often because of changes in the partitioning of the substrate between the scCO2 and the IL (a more detailed description is beyond the scope of this chapter for further details see149). All this shows that the development of continuous flow homogeneous catalytic reactions in IL–CO2 biphasic system requires careful investigation leading to considerable improvements from the initial findings as a result of tuning the solvating power of the CO2, the ionic liquid and the catalyst together with the engineering design. Following this, successful systems have been developed as for hydrovinylation and hydroformylation where the catalysts were found to be stable and active at commercially interesting rates during more than 60 h of continuous reaction with low catalyst leaching (12 ppb in the case of hydroformylation).149

Further improvements regarding the hydroformylation of alkenes in IL–scCO2 under continuous flow conditions were recently reported. In order to enhance the selectivity of the reaction, modified xantphos ligand 51 was used instead of the PPh3-based ligand 49 and this allowed the reaction to proceed with very high linear selectivity (l/b ¼ 40) but at the expense of a decrease in reaction rate and an increase in the rhodium leaching (0.2 ppm vs 12 ppb with the PPh3-based ligand).150 In order to decrease the amount of ionic liquid used and to open up the possibility of working with a fixed bed tubular reactor a supported ionic liquid phase catalyst was used in conjunction with CO2.151 Similarly to the SAPC described in the previous section, an ionic liquid containing the active catalyst is impregnated on the surface of a porous material (usually silica but not exclusively) and can then be use as a solid catalyst in a fixed bed reactor. This methodology was earlier developed and used with success for gas phase and liquid phase reactions.152 However the hydroformylation using liquid phase substrates is less successful because of problems of slow diffusion of the gaseous reactants into the liquid filled pores of the catalyst. Only the gas initially dissolved in the liquid substrate is available for reaction and, once it is depleted, diffusion of further gas into the liquid filled pores is very slow. This limitation seemed to be overcome by using scCO2 as a transport fluid for the continuous flow hydroformylation of octene. It also appears that diffusion into the ionic liquid is not rate limiting since it was observed that the thickness of


243

the IL film on the surface of the silica showed little to no influence on the reaction rate. Under optimized conditions the system could be operated at a total pressure of 100 bar for 40 h with a constant 500 turnover per hour (significantly higher than the bulk IL –scCO2 previously described) and low metal leaching. It is probable that the reaction actually occurs in a CO2 expanded liquid phase. A supported ionic liquid phase with scCO2 as a transport vector was also used for the aerobic oxidation of alcohols by perruthenate (RuO4).153 The system was found to be very active, selective and stable. No ruthenium could be detected in the product phase, i.e. ruthenium leaching < 1 ppb, and the system could be reused under batch conditions without significant loss in activity (the small loss of activity observed is believed to be due to mechanical loss of part of the catalyst). A very similar system was recently developed but using cross-linked polymer supported IL instead of IL supported on silica.154 Since these pioneering studies on homogeneous catalysis in IL–scCO2 biphasic systems, other system have been developed. The synthesis of dimethylcarbonate, a low toxicity substitute for phosgene, from MeOH, CO2 and O2 was investigated in CO2–IL biphasic system.155 Here CO2 acts as substrate and solvent. The main problem associated with this transformation is that water formed during the reaction must be scavenged from the system. Without scavenger, dimethoxymethane was the main product, but 25 % selectivity to dimethylcarbonate could be obtained with a CO2:O2 ratio of 79 : 21. The reaction could be repeated three times with similar conversion, although the selectivity dropped slightly with reuse. For the production of cyclic carbonate from epoxides and CO2, the IL acts as solvent and catalyst.156 Among the different ionic liquids tested, [BMIM]BF4 proved to be the most suitable. However, reuse of the system led to a dramatic drop in selectivity. The asymmetric hydrogenation of imines already developed in scCO2 following the CESS approach116b(see above) was recently reinvestigated in an IL–scCO2 biphasic system (Scheme 9.9).157 Striking influences of the CO2 and the ionic liquid on the performance of this system were observed. The presence of CO2 was beneficial to the reaction by increasing the hydrogen solubility in the IL phase, thus increasing its availability for the catalyst. The ionic liquid allowed the use of in situ formed catalyst while this approach led to the formation of inactive complexes in classical molecular solvents. Moreover, the nature if the IL anion was shown to have a major effect on the enantioselectivity of the reaction. The catalyst immobilized in [BMIM]PF6 could be use for seven consecutive batch cycles showing constant activity and enantioselectivity. Leaching of the iridium into the product phase was found to be below the detection limit (99 95 95 92 99 99 69

11 12 12 13 13 13 14 15 16 16 17

S/C denotes substrate/catalyst molar ratio.

Dupont and coworkers studied the asymmetric hydrogenation of a-acetamidocinnamic acid with Rh–EtDuPHOS (see Figure 10.1 for ligand structures) as catalyst in IL/iPrOH mixtures, with the best enantioselectivity of 93% ee obtained in a [bmim][BF4]/iPrOH medium (Table 10.1, entry 1).11 It was found that the solubility of hydrogen in [bmim][BF4] is almost four times higher than in [bmim][PF6]. As a result, better results were observed in [bmim][BF4]/iPrOH. The IL-immobilized catalyst could be reused four times; thereafter, however, the catalytic activity dropped dramatically, presumably due to catalyst leaching during product extraction. Similarly, the asymmetric hydrogenation of enamides with Rh–MeDuPHOS in the biphasic [bmim][PF6]/iPrOH mixture afforded high product yields and enantioselectivities, comparable with those obtained in iPrOH (Table 10.1, entries 2–3).12 Recycling experiments showed that the immobilized catalyst retained its enantioselectivity after five cycles; but the yield decreased gradually. In contrast to hydrogenation in common organic solvents, all the operations could be conducted in air without significant loss of enantioselectivity, suggesting that the IL could stabilize highly air-sensitive catalysts. The same hydrogenation could also be carried out efficiently with ferrocene-based Taniaphos and Josiphos in ILs/water, affording up to >99% ee regardless of the nature of ILs used (Table 10.1, entries 4–5).13 However, lower enantioselectivities were obtained in pure ILs, IL/organic co-solvent biphasic media, or alcohols. The hydrogenation in wet ILs displayed a less pronounced hydrogen pressure effect than in alcohols, indicating that the better results observed in the wet ILs are not due to a lower H2 solubility, but more likely to differences in catalyst solvation in the two solvent systems. The immobilized catalyst could be reused up to seven times with complete conversions and more than 99% ee values. In the case of methyl a-acetamidoacrylate, a full conversion and 95% ee were demonstrated in wet [omim][BF4] at a high S/C ratio of 10 000 in 3 h (Table 10.1, entry 6).

Asymmetric Catalysis in Ionic Liquids CF3

Ligands: Et

Et P

PPh2

Me

Me

P

P

P PPh2

Et Et

(R,R)-MeDuPHOS

P

Fe

PCy2

Ar2P

Fe

PCy2 H N

Fe NMe2

O

Taniaphos OMe R1 = R2 = Ph R1 = R2 =

Josiphos

KO3S

P

R1

P

CF3

(R)-BINAP

R2

R2

F3 C F3C

Me Me

(R,R)-EtDuPHOS

R1

263

NTf2 N +

PPh2

N

PPh2 KO3S

1: Ar= 3,5-(CF3)2Ph

2

Me OMe

Chiral IL: CO2Me NH2

NTf2

CIL-1

Figure 10.1 Ligands and IL used in Table 10.1

To avoid using co-solvents, the Rh-catalyzed asymmetric hydrogenation of methyl a-benzamido cinnamate was investigated in neat ILs.14 At room temperature with Rh–EtDuPHOS or Ru–BINAP as catalyst, the hydrogenation in [bmim][BF4] led to no conversion. Changing to [emim][OTf], Rh–EtDuPHOS provided a 15% conversion and 85% ee, while Ru–BINAP afforded a similar conversion and better ee of 95% (Table 10.1, entry 7). The conversion in [emim][OTf] could be improved by increasing the reaction temperature or hydrogen pressure. For example, the conversion could be increased to 95% with 89% ee at 50 C. [emim][OTf] has a lower solubility than [bmim][BF4] in organic solvents, making product extraction easier. Asymmetric hydrogenation of (Z)-a-acetamidocinnamic acid and its methyl ester catalyzed by Rh–DIPAMP in ILs/iPrOH was also investigated by Bakos and coworkers.15 No reaction was observed in [bmpy][BF4]; but full conversion was reached in [bmim][BF4] and [bmim][PF6], with better enantioselectivities found in the former (Table 10.1, entry 8). The reaction was temperature dependent, the best results being obtained at 55 C and 70 C for the two substrates, respectively. The catalyst could be reused in three consecutive reactions with retained activity and enantioselectivity. Blaser et al. recently developed modified Josiphos ligands bearing an imidazolium tag.16 The ligands displayed similar catalytic performance to the parent ligands in hydrogenation in both organic solvents and IL/co-solvent mixtures. In particular, Rh–1 worked efficiently

264


to reduce methyl acetamidoacrylate with excellent enantioselectivities (99% ee) and full conversions in the biphasic [bmim][BF4]/TBME (t-butyl methyl ether) and [bmim][BF4]/ toluene (Table 10.1, entries 9–10). The imidazolium tag renders the catalysts preferentially soluble in the ionic phase, thereby reducing catalyst leaching. Indeed, the catalysts in [bmim][BF4]/TBME could be successively used eight times without significant loss of activity and enantioselectivity. The use of chiral ILs as both the solvent and only source of chirality in asymmetric catalysis has received increasing attention.9 In 2007, Leitner and coworkers reported asymmetric hydrogenation in chiral ILs with an achiral rhodium catalyst.17 The ability to transfer chirality from chiral ILs is seen in CIL-1 (Figure 10.1), which was tested in the Rh-catalyzed hydrogenation of methyl 2-acetamidoacrylate in the presence of a racemic tropisomeric ligand 2. Full conversion and an enantioselectivity of 49% ee were observed; a higher ee was obtained upon the addition of NEt3 (Table 10.1, entry 11). After extracting the product with scCO2, the catalyst was reused twice without significant loss of efficiency. Asymmetric Hydrogenation of Unsaturated Acids and Esters In 1997, Dupont and coworkers investigated the Ru-catalyzed asymmetric hydrogenation of 2-arylacrylic acids with (S)-BINAP as ligand in a mixture of [bmim][BF4] and alcoholic solvents, obtaining results comparable or better than those achieved in pure alcohols.18 It is interesting to note that the enantioselectivity was higher in the homogeneous mixture of [bmim][BF4]/MeOH (Table 10.2, entry 1; Figure 10.2) than in the biphasic mixture of [bmim][BF4]/iPrOH. After completion of the reaction in [bmim][BF4]/iPrOH, the product resided in the alcohol while the catalyst was in the IL phase, facilitating catalyst recycle. Under similar conditions, 2-(60 -methoxy-20 -naphthyl)acrylic acid was hydrogenated to provide an anti-inflammatory drug, Naproxen, in quantitative yield and 80% ee (Table 10.2, entry 2). Jessop and coworkers described enantioselective hydrogenation of tiglic acid with Ru–(R)-tolBINAP as catalyst in wet [bmim][PF6], where the product was extracted with scCO219 As with most hydrogenation catalysts, Ru–(R)-tolBINAP is insoluble in scCO2, ensuring easy extraction of the product. In this study, the recovered catalyst was reused five times with retained activity and even enhanced enantioselectivity. The effect of hydrogen pressure and co-solvent was also examined. At a low pressure, the hydrogenation ran smoothly in pure [bmim][PF6], affording a 88% ee. However, increasing the pressure Table 10.2 Asymmetric hydrogenation of unsaturated acids and esters in ILs Entry 1 2 3 4 5 6 7 8 9

Catalyst Ru–(S)-BINAP Ru–(S)-BINAP Ru–(R)-tolBINAP Ru–(R)-tolBINAP Ru–(R)-tolBINAP Ru–(R)-BINAP Rh–1 Rh–1 Rh–2

Solvent [bmim][BF4]/MeOH [bmim][BF4]/iPrOH [bmim][PF6]/H2O [emim][NTf2] [bmim][PF6]/MeOH [P(nC6H13)3(nC14H29)][Cl]/MeOH [bmim][BF4]/TBME [bmim][BF4]/iPrOH CIL-1

Sub. S/C P (bar) ee (%) A B C C A D D D D

40 40 40 30 30 6000 200 200 250

25 75 5 5 50 20 1 1 40

86 80 92 95 87 96 99 99 29

Ref. 18 18 19 20 20 21 16 16 17

Asymmetric Catalysis in Ionic Liquids

265

Substrates: O COOH

COOH

CO2H

Me

MeO A

O

O

Me

O B

C

D

Ligands:

(S)-BINAP

PPh2

PAr2

PPh2

PAr2

(R)-tolBNAP (Ar = tol)

Figure 10.2 Substrates and ligands used in Table 10.2

eroded the enantioselectivity. This is consistent with studies carried out in common organic solvents such as MeOH.1a In wet [bmim][PF6], the enantioselectivity reached 92% ee at a hydrogen pressure of 5 bar (Table 10.2, entry 3). In contrast, when switching the co-solvent from water to iPrOH, much lower enantioselectivities were noted. This might stem from a higher H2 concentration in IL/iPrOH than in IL/H2O under the same pressure. The asymmetric hydrogenation of tiglic acid and atropic acid was also investigated in other ILs.20 In the reduction of tiglic acid, enantioselectivities of up to 95% ee were obtained in viscous ILs (Table 10.2, entry 4), such as [bmim][PF6], [emim][NTf2] and [dmpim][NTf2]; in contrast, using less viscous ILs, IL/organic co-solvent, methanol, or CO2-expanded ILs gave lower ee values. This variation in ee presumably again results from variation in the hydrogen concentration in the solvents examined.1a However, atropic acid necessitates a high hydrogen concentration for good ee values. Thus, neat ILs are no longer a good choice because of their high viscosity that limits hydrogen diffusion, and so the addition of co-solvent is necessary. Indeed a high enantioselectivity of 87% ee was observed in a CO2-expanded [bmim][PF6]/MeOH mixture (Table 10.2, entry 5). The presence of compressed CO2 is believed to reduce the viscosity of the IL, thereby increasing the H2 solubility. Enhanced enantioselectivity and catalytic stability were observed by Livingston et al. in their study of Ru-catalyzed asymmetric hydrogenation of dimethyl itaconate in homogeneous ILs/methanol mixtures.21 Thus, changing the reaction medium from MeOH to [P(nC6H13)3(nC14H29)][Cl]/MeOH, [NnBu4][Cl]/MeOH or [PPh4][Cl]/MeOH resulted in a significant increase in enantioselectivity, e.g. from 75% ee in methanol to 96% ee in the [P(nC6H13)3(nC14H29)][Cl]/MeOH mixture (Table 10.2, entry 6). However, the enantioselectivity was sensitive to the nature of ILs, with increased ee values observed only in the ILs containing chloride. A high S/C ratio of 6000 was demonstrated, and the chiral catalyst could be recycled seven times with retained reactivity and enantioselectivity. The product was separated by using a nanofiltration membrane.

266


The catalyst derived from the modified Josiphos ligand 1 and [Rh(NBD)2][BF4] was also found to catalyze the hydrogenation of dimethyl itaconate, affording a 99% ee and full conversion in the biphasic [bmim][BF4]/TBME and [bmim][BF4]/iPrOH (Table 10.2, entries 7–8).16 Using the chiral IL CIL-1 as both solvent and source of chirality, the reduction was less efficient, however (Table 10.2, entry 9).17 Asymmetric Hydrogenation of Keto Esters Asymmetric hydrogenation of keto esters has been extensively studied, and many successful catalysts have been developed for use in organic solvents. Replacing these solvents with ILs has been the subject of a number of studies; selected examples are shown in Table 10.3. Lin and coworkers reported synthesis of the modified BINAP ligands 3 and 4, which were examined in the asymmetric hydrogenation of b-alkyl ketoesters in a homogeneous IL/MeOH (Figure 10.3).22 In the mixture of IL ([bmim][BF4], [bmim][PF6] or [dmpim] [NTf2]) and MeOH, Ru–4 afforded enantioselectivities comparable with those obtained in MeOH; but Ru–3 led to higher ee values. With the latter catalyst, ee values up to 99.3% were obtained regardless of the nature of ILs (Table 10.3, entries 1–2); but the enantioselectivity obtained with Ru–4 varied with the ILs, with the best ee (up to 98.9%) being attained in [bmim][BF4]/MeOH (Table 10.3, entry 3). The products were readily separated by extraction and the immobilized catalyst could be recycled 4 times. Ru–3 and Ru–5 also catalyzed efficient hydrogenation of b-aryl ketoesters in [bmim][BF4]/MeOH, affording up to 99.8% ee (Table 10.3, entries 4–7).23 The mixed solvent outperformed neat MeOH, and the presence of an IL makes the catalyst recyclable. a-Ketoesters were also reduced. An example is seen in the reduction of ethyl benzoyformate in ILs with a carborane-based rhodacarborane–BINAP catalyst (Table 10.3, entry 8).24 Full conversions and high enantioselectivities (>98%) were observed in [omim][BF4], [bmim][PF6], and a novel carborane–IL [bpy][CB10H12] at 50 C, requiring no co-solvent. In contrast, the conventional THF only offered lower conversions and ee

Table 10.3 Asymmetric hydrogenation of keto esters in ILs Entry 1 2 3 4 5 6 7 8 9 10 11 12 13

Catalyst Ru–3 Ru–3 Ru–4 Ru–3 Ru–3 Ru–5 Ru–3 Rh–(R)-BINAP Ru–(R)-P-Phos Ru–(R)-P-Phos Ru–6 Ru–7 Ru–(S)-BINAP

Solvent [bmim][PF6]/MeOH [dmpim][NTf2]/MeOH [bmim][BF4]/MeOH [bmim][BF4]/MeOH [bmim][BF4]/MeOH [bmim][BF4]/MeOH [bmim][BF4]/MeOH [bpy][CB10H12] [bmim][PF6]/MeOH [bmim][BF4]/MeOH [bpy][NTf2] [bpy][NTf2] Polymer-supported [bmim] [PF6]/iPrOH

Sub. A B B D E F G H I C C C A

S/C P (bar) 100 100 100 100 100 100 100 1000 100 100 1000 1000 140

105 105 105 90 90 90 90 12 70 70 40 40 40

ee (%)

Ref.

99.3 99.3 98.9 99.3 99.8 98.9 97.5 99.5 93 99 90 83 97

22 22 22 23 23 23 23 24 25 25 26 26 27

Asymmetric Catalysis in Ionic Liquids Substrates: O

O O

O

R1

O

R2

Ar

A: R1 = Me, R2 = Me; B: R1 = Et, R2 = Et; C: R1 = Me, R2 = Et.

O

O

O O

O

R O

D: Ar= Ph, R = Et; E: Ar = 2-ClPh, R = Me; F: Ar = 3-CF3Ph, R = Me; G: Ar = 2-CF3Ph, R = Me.

267

O

H

I

Ligands: SiMe3

PO(OH)2 (HO)2OP PPh2

PPh2

PPh2

PPh2 (HO)2OP

PPh2 SiMe3

PO(OH)2

5

4

3

PPh2

NH3Br

OMe

NH3Br

N MeO

PPh2

PPh2

MeO

PPh2

PPh2

PPh2 PPh2

N OMe (R)-P-Phos

NH3Br 6

NH3Br 7


values. The catalyst could be easily recovered and recycled 6 times without compromising the activity and enantioselectivity. Using [Ru((R)-P-Phos)Cl2] as catalyst, Chan and coworkers investigated the asymmetric hydrogenation of a- and b-ketoesters in [bmim][BF4] and [bmim][PF6].25 For methyl pyruvate, pure ILs provided low conversions (5–10%); but the addition of methanol as co-solvent led to higher enantioselectivities (83–86%) and moderate to good conversions (73–95%). On going to methyl benzoylformate, ee values up to 93% were achieved in [bmim][PF6]/methanol (Table 10.3, entry 9). Under similar conditions, b-ketoesters could be smoothly hydrogenated with up to 99% ee (Table 10.3, entry 10). The chiral catalyst could again be readily recovered and reused. Aiming to further improve the efficiency of catalyst immobilization in ILs, Lemaire et al. synthesized the ammonium-tagged BINAP 6 and 7.26 In the Ru-catalyzed asymmetric hydrogenation of ethyl acetoacetate in ILs derived from imidazolium, pyridium, and phosphonium salts in the absence of a co-solvent at 50 C, ee values up to 90% and 83% were recorded with Ru-6 and Ru-7, respectively (Table 10.3, entries 11–12). The catalyst

268


proved to be recyclable; but their efficiency depended on the nature of ILs, with those containing imidazolium and pyridinium cations producing good results. Further, better enantioselectivities were observed in water than in ILs. Supported ILs were also explored. Mixing [bmim][PF6] and a Ru-(S)-BINAP complex with poly(diallyldimethylammonium chloride) generated a heterogeneous catalyst.27 In the asymmetric hydrogenation of methyl acetoacetate with iPrOH as co-solvent, this polymersupported IL system displayed a higher activity than the biphasic [bmim][PF6]/iPrOH mixture, furnishing a high ee of 97% (Table 10.3, entry 13). Catalyst leaching into the organic phase during extraction was not detected, and the catalyst was reusable. Asymmetric Hydrogenation of Unfunctionalized Ketone Asymmetric hydrogenation of unfunctionalized ketones in ILs has been less studied, with only few reports available (Table 10.4). The aforementioned rhodacarborane–BINAP catalyst allows for the asymmetric hydrogenation of acetophenone in ILs.24 At 50 C in [omim][BF4], [bmim][PF6], or [bpy][CB10H12], full conversions and high enantioselectivities were obtained, with the best ee of 99.1% observed in [bpy][CB10H12] (Table 10.4, entry 1). When combined with Rh–DPEN (DPEN ¼ 1,2-diphenyl-1,2-ethylenediamine), the BINAP ligands 3 and 4 were also viable for the hydrogenation of simple aryl ketones in ILs/iPrOH.28 The catalyst activity varied with the IL employed, with full conversions only in [dmpim][NTf2]/iPrOH. In comparison with that containing 4, the catalyst Ru–3, i.e. [RuCl2(DPEN)(3)], led to higher enantioselectivities with various aryl ketones in Table 10.4 Asymmetric hydrogenation of unfunctionalized ketones in ILs O R1

R2

+

H2

OH chiral catalyst * solvent R1 R2

A: R1 = Ph, R2 = Me; B: R1 = 1-Naph, R2 = Me; C: R1 = 2-Naph, R2 =Me; D: R1 = 4-tBuPh, R2 = Me; E: R1 = 4-ClPh, R2 = Me; F: R1 = 4-MePh, R2 = Me; G: R1 = Ph, R2 = Et; H: R1 = Ph-CH=CH, R2 = Me.

Entry 1 2 3 4 5 6 7 8 9 10

Catalyst

Solvent

Sub.

Rh–(R)-BINAP Ru–3 Ru–3 Ru–3 Ru–3 Ru–3 Ru–3 Ru–3 Ru–TPPTS–DPENDS Ru–TPPTS–DPENDS

[bpy][CB10H12] [dmpim][NTf2]/iPrOH [dmpim][NTf2]/iPrOH [dmpim][NTf2]/iPrOH [dmpim][NTf2]/iPrOH [dmpim][NTf2]/iPrOH [dmpim][NTf2]/iPrOH [dmpim][NTf2]/iPrOH [bmim][Ts] [emim][Ts]/H2O

A A B C D E F G A H

S/C P (bar) ee (%) 1000 1000 1000 1000 1000 1000 1000 1000 112 112

12 49 49 49 49 49 49 49 5 5

99.1 96.1 98.5 96.7 98.1 93.7 96.5 92.9 84.8 75.9

Ref. 24 28 28 28 28 28 28 28 29 30


269

Ligands: SO3Na P 3

NaO3S

TPPTS

H2N

NH2

SO3Na

(1S,2S)-DPENDS

Figure 10.4 Ligands used in Table 10.4

[dmpim][NTf2] (Table 10.4, entries 2–8). The low conversions observed in [bmim][BF4] and [bmim][PF6] was attributed to in situ formation of Arduango carbenes, which could inhibit the catalysis. In contrast, the lack of acidic protons in [dmpim][NTf2] precludes the generation of such species. A series of ILs with [Rmim] (R ¼ ethyl, butyl, octyl, dodecyl) as cations and [Ts], [BF4] and [PF6] as anions were studied in the asymmetric hydrogenation of aryl ketones.29 The activity of the catalyst [RuCl2(TPPTS)2]2–(1S,2S)-DPENDS [DPENDS ¼ 1,2-diphenyl-1,2-ethylenediamine sulfonate disodium] (Figure 10.4) varied with the ILs used, with good catalytic performance found in hydrophilic ILs that contain [Ts] and [BF4]. The length of the alkyl group on the imidazole ring affected the enantioselectivity as well, the ee decreasing with increase in the alkyl chain length. Under optimized conditions, an 84.8% ee was observed in the hydrogenation of acetophenone in [bmim][Ts] (Table 10.4, entry 9). The same catalyst also enables the reduction of a,b-unsaturated ketones in wet ILs.30 For instance, benzalacetone was hydrogenated in 75.9% ee in wet [emim][Ts] (Table 10.4, entry 10). The catalyst could be reused seven times with catalytic performance retained. Asymmetric Hydrogenation of Imines and Enamines Asymmetric hydrogenation of imines and enamines in ILs has received still less attention. Table 10.5 shows examples reported. Trimethylindolenine was reduced in a variety of neat ILs with Ir–Xylphos as catalyst.31 Considering the high viscosity of ILs, the reaction was run at 50 C, affording a best ee of 86% in [C10mim][BF4] (Table 10.5, entry 1; Figure 10.5). The enantioselectivity is comparable with that in toluene; but the reaction rate in [C10mim] [BF4] is faster. In addition, the use of IL as reaction medium greatly reduced the airsensitivity of the catalytic system. Pfaltz and coworkers studied the asymmetric hydrogenation of imines in IL/CO2 mixtures, using a cationic Ir catalyst containing the phosphinooxazoline 8.32 The presence Table 10.5 Asymmetric hydrogenation of imines and enamines in ILs Entry 1 2 3

Catalyst Ir–Xylphos Ir–8 Ir–9

Solvent [C10mim][BF4] [emim][BARF]/scCO2 [bmim][SbF6]/iPrOH

Sub.

S/C

P (bar)

ee (%)

Ref.

A B C

250 500 100

40 30 1

86 78 97

31 32 33

270


Substrates: Ph

N

NHAc Ph

N A

B

C

Ligands:

Ph2P

O

P Fe

N

N

O

BF4 N

N

N

N

BF4

PPh2 N PPh2 PPh2

Xylphos

8

9


of compressed CO2 in the reaction system accelerated the reduction at low hydrogen pressure without deteriorating the enantioselectivity; this was attributed to the greatly increased H2 solubility in and reduced viscosity of ILs in the presence of CO2. The enantioselectivity was found to be sensitive to the anion of the IL, however. For example, on changing from [BF4] to the less coordinating [BARF], the ee values increased from 30% to 78% (Table 10.5, entry 2). An additional benefit is that the product could be easily isolated by using scCO2 extraction without recourse to organic solvents. To demonstrate the catalyst recyclability, the catalysis was repeated seven times, affording similar reactivity and enantioselectivity in each run. An example of enamine reduction is seen in N-acetylphenylethenamine (Table 10.5, entry 3).33 When complexed to rhodium, the chiral bidentate ligand 9, which bears two imidazolium salt tags, catalyzed the asymmetric hydrogenation of N-acetylphenylethenamine in [bmim][SbF6]/iPrOH, providing up to 97% ee. As might be expected, the catalyst was easily immobilized in the IL, and could be reused. 10.2.2 Asymmetric Transfer Hydrogenation Asymmetric transfer hydrogenation provides a powerful alternative to asymmetric hydrogenation for catalytic reduction because of its versatility and practical simplicity.1 However, the reaction has not been widely explored in ILs. Selected examples are shown in Table 10.6. In 2000 Andersen and coworkers reported the Rh-catalyzed asymmetric transfer hydrogenation of acetophenone in phosphonium tosylate ILs, with iPrOH as hydrogen source.34 Using a rhodium catalyst bearing ()-DIOP, acetophenone was enantioselectively reduced in [P(Octyl)3Et][OTs] at 120 C in the presence of KOH/iPrOH, furnishing a 50% conversion and 92% ee (Table 10.6, entry 1). These tosylate salts are liquid at the reaction temperature but solid at room temperature, thereby facilitating catalyst recovery.


271

Table 10.6 Asymmetric transfer hydrogenation in ILs

O

OH *

chiral catalyst IL

Entry

Catalyst

Solvent

S/C

ee (%)

Ref.

1 2 3

Rh–(-)-DIOP 10 Ru–11

[P(Octyl)3Et][OTs]/iPrOH [dmbim][PF6] [bmim][PF6]

100 200 100

92 98 93

34 35 36

Dyson and Geldbatch recently reported the reduction of acetophenone by iPrOH and the HCOOH–Et3N azeotrope in [dmbim][PF6], which can phase separate from iPrOH.35 With a modified Noyori–Ikariya catalyst 10, which bears an ‘IL-philic’ ionic tag, a quantitative conversion and high ee (up to 98%) were obtained in the reduction of acetophenone by HCOOH–Et3N (Table 10.6, entry 2; Figure 10.6). Using a similar tagging strategy, the ligand 11 was synthesized and when combined with [RuCl2(benzene)]2, it effected the transfer reduction of acetophenone by HCOOH–Et3N in [bmim][PF6], affording a 98% conversion and 93% ee (Table 10.6, entry 3).36 10.2.3 Asymmetric Oxidation Asymmetric Epoxidation of Olefins Asymmetric epoxidation of unfunctionalized olefins is one of the most powerful transformations in organic synthesis, since the resulting chiral epoxides are important building blocks in the preparation of biologically active compounds.1 Chiral Mn–Salen complexes developed by Jacobsen37 and Katsuki38 are effective catalysts, and many methods, including the use of ILs, have been developed to immobilize them.2,3 In 2000, Song et al. reported the asymmetric epoxidation of unfunctionalized olefins in [bmim][PF6] with Jacobsen’s Mn–Salen complex (R,R)-12 as catalyst (Scheme 10.3).39 The oxidation was carried out at 0 C with NaOCl as oxidant, affording a 96% ee in the example shown. Since the IL is solid at 0 C, it was necessary to add a co-solvent, CH2Cl2. The catalyst performed more effectively in the presence of the IL, displaying an enhanced

Ligands: N

O BF4 Cl Ru NTs H 2N Ph Ph 10

N

N

3

2[CF3CO2]

N O 2S

Ph Ph

N H 11

NH3

Figure 10.6 Catalyst and ligand used in Table 10.6

272


O

(R,R)-12 (4 mol%), NaOCl [bmim][PF6]/CH2Cl2, 0 oC

O up to 96% ee H N tBu

O

H N Mn Cl

O

tBu

tBu

tBu (R,R)-12

Scheme 10.3

activity as compared with that in pure CH2Cl2. The catalyst could be recycled up to five times; but the activity and enantioselectivity decreased gradually after each cycle. The same oxidation could also be effected with a Katsuki-type catalyst 13 in the presence of 4-phenylpyridine N-oxide in a [bmim][PF6]/CH2Cl2 or [bmim][PF6]/ethyl acetate mixture (Figure 10.7).40 A higher activity and comparable ee values (up to 100%) were recorded in comparison with values obtained in chlorinated solvents. In recycling runs, however, the catalyst again became less active and enantioselective. Solid-supported ILs have been explored to immobilize the Mn–Salen catalysts. Using [bmim][PF6] as an immobilizing phase adsorbed on mesoporous silicate MCM-48, the catalysts (S,S)-12 and 14 allowed a series of olefins, including styrene, a-methylstyrene, 1-phenylcyclohexene and indene, to be epoxidized, providing the corresponding products with moderate to excellent conversions (up to 99%) and enantioselectivities (up to 99%).41 The reaction was carried out at 0 C, with CH2Cl2 as co-solvent and m-CPBA/NMO (m-chloroperoxybenzoic acid/N-methylmorpholine N-oxide) as oxidant. The immobilized catalysts could be reused without loss of catalytic efficiency. Asymmetric Dihydroxylation of Olefins The Os-catalyzed Sharpless asymmetric dihydroxylation of olefins furnishes chiral vicinal diols, important intermediates for the preparation of chiral drugs and natural products.1 However, the high cost of osmium and the chiral ligand coupled with the toxicity and volatility of OsO4 have restricted its wider applications in industry. To address this problem, many attempts have been made to immobilize the catalyst, including microencapsulation,

H N

H N

O Ph

O Ph

Ph H N

Ph H N

Mn t Bu

O

Mn Cl

t Bu 13

O tBu

14

Figure 10.7 Epoxidation catalysts 13 and 14

t Bu


273

anchoring onto porous resins, or ionic exchange on various solids. However, none appears to endure a good number of reuses.2,3 Responding to the challenge, ILs have been explored for the dihydroxylation. Song et al. described the OsO4-catalyzed asymmetric dihydroxylation of mono- and di-substituted olefins in ILs/acetone/H2O mixtures using a bis-cinchona alkaloid ligand (QN)2PHAL and NMO as oxidant (Scheme 10.4).42 The catalyst worked in [bmim][PF6] or [bmim][SbF6] to provide high enantioselectivities (up to 97%) and good yields (up to 98%), and a remarkable rate acceleration was observed in the presence of the IL. Moreover, the IL suppressed overoxidation of the resulting diols, a commonly observed side reaction in organic solvents. The catalyst could be recovered and reused several times with retained enantioselectivity. Notably, a high turnover number of 2370 was obtained after three recycling runs using 0.1 mol% of OsO4. Interestingly, whilst changing from (QN)2PHAL to the well-known (DHQ)2PHAL did not alter the catalytic activity and enantioselectivity, severe leaching of the ligand and osmium occurred during extraction. This is probably due to the partial solubility of (DHQ)2PHAL in the ether used for extraction. In contrast, (QN)2PHAL was recyclable. This is attributed to the in situ dihydroxylation of its two double bonds, leading to a more polar ligand (QN)2PHAL(4-OH) (Scheme 10.4). A similar effect has also been observed by Sheldon and coworkers in the asymmetric dihydroxylation of trans-stilbene in [bmim][PF6]/H2O/acetone with the same ligand.43 R1 R2

OsO4 (1 mol%), NMO R3 (QN)2PHAL or (DHQ)2PHAL (2.5 mol%) [bmim][PF6]/acetone/H2O

R4

HO R1 R2

OH R3 R4

up to 97% ee

R1 = Ph, R2 = H, R3 = H, R4 = Ph, R1 = Ph, R2 = R3 = R4 = H R1 = Ph, R2 = Me, R3 = R4 = H R1 = Ph, R2 = H, R3 = H, R4 = Me, R1 = Ph, R2 = H, R3 = H, R4 = CO2Me, R1 = 4-MeOPh, R2 = H, R3 = H, R4 = CO2Me Et N

N

N N O

Et N

O

O

MeO

OMe N

N

N N O

MeO

OMe

N

N

(QN)2PHAL

N (DHQ)2PHAL

OH

N O

HO

N

N N

dihydroxylation

O

MeO

OMe N

HO N

N

N N O

MeO

N

OMe N

(QN)2PHAL

N

(QN)2PHAL(4-OH)

Scheme 10.4

OH

O

274

Recoverable and Recyclable Catalysts Et N

Et

O

OMe

O N

MeO N

N (DHQD)2PHAL

N

Ph O

O

MeO N

Et

Et N

N

N N

N

OMe

Ph

N (DHQD)2PYR

Figure 10.8 Dihydroquinidine ligands for asymmetric dihydroxylation

The asymmetric dihydroxylation of mono-, di-, and trisubstituted-olefins in ILs was also described by Afonso et al.44 With K2OsO2(OH)4–(DHQD)2PHAL or (DHQD)2PYR as catalyst and K3Fe(CN)6 as oxidant (Figure 10.8), the reaction in the biphasic [bmim][PF6]/ H2O or monophasic [bmim][PF6]/H2O/tBuOH mixture afforded yields (up to 98%) and enantioselectivities (up to 99%) comparable with or better than those achieved in the homogenous H2O/tBuOH mixture. The catalytic system could be used repeatedly over ten times without significant loss of activity and enantioselectivity, with more than 90% osmium retained in the IL phase. Further studies established that the reaction proceeded faster in [bmim][PF6] than in the homogeneous H2O/tBuOH, and NMO as oxidant led to higher enantioselectivities than K3Fe(CN)6.45 The asymmetric dihydroxylation could also be conducted in neat ILs.46 With K3Fe(CN)6 or NMO as oxidant, a range of ILs were screened. The catalyst aforementioned performed efficiently in [bmim][NTf2], [omim][BF4], [dmbim][PF6], [dmbim][BF4] and [dmbim] [NTf2] in the presence of NMO, affording high product yields (87–98%) and ee values (90–97%).46a It is noted that whilst NMO can be easily extracted from the IL phase together with the product, K3Fe(CN)6 necessitates extraction with H2O, increasing the possibility of osmium leaching. In a model dihydroxylation of 1-hexene in [bmim][NTf2] and [dmbim] [NTf2], the catalyst in both ILs could be reused 14 times following extraction with diethyl ether, with the activity and enantioselectivity maintained. Alternatively, scCO2 can be used for product extraction, circumventing the need for an organic solvent. This was demonstrated in the asymmetric dihydroxylation of methyl trans-cinnamate in neat ILs; subsequent scCO2 extraction provided isolated product without osmium contamination.46b To further enhance the affinity of catalysts for ILs, a mono-quaternized cinchona alkaloid ligand (QN þ )(QN)PHAL was prepared (Figure 10.9).47 When combined with

N

N

N N O

O

MeO

CH2Ph Br OMe

N

N (QN+)(QN)PHAL

Figure 10.9 Mono-quaternized quinidine ligand for asymmetric dihydroxylation


275

OsO4, the ligand led to efficient dihydroxylation of a series of olefins with NMO in a [bmim] [PF6]/H2O/acetone medium, providing good yields (up to 91%) and enantioselectivities (up to 99%).47 The catalyst was shown to be reusable without significant decrease of yield or enantioselectivity. It is also possible to conduct asymmetric dihydroxylation in chiral ILs without employing chiral ligands. Thus, using the chiral IL CIL-2 as solvent, K2OsO2(OH)4 catalyzed the dihydroxylation of styrene and 1-hexene by NMO, affording a 85% and 72% ee, respectively, with over 90% yields (Scheme 10.5).48 K2OsO2(OH)4 (0.5 mol%), NMO

HO

CIL-2

R

R

OH

R = Ph, nC4H9 HO CO2 NMe2 (C6H13)2N

N(C6H13)2

HO

OH OH

CIL-2

Scheme 10.5

10.2.4 Asymmetric CC Bond Formation Asymmetric Diels–Alder Reactions Lewis acid-catalyzed asymmetric Diels–Alder reactions have been well studied, including the use of ILs as alternative solvents.1 In 2003, Oh and coworkers reported the Cu–(S,S)-tBu-Box (Box ¼ bis(oxazoline))-catalyzed asymmetric Diels–Alder reaction of N-crotonoyloxazolidione with cyclopentadiene in [dbim][BF4] (Scheme 10.6).49 An enhanced enantioselectivity and endo/exo ratio were observed in the IL compared with those in CH2Cl2 (92% ee and 93/7 endo/exo ratio in IL vs 52% ee and 79/21 endo/exo ratio in CH2Cl2), showing that the IL is a better reaction medium than the conventional solvent.

O

O N

O

+

[Cu(OTf)2(S,S)-tBu-box] (10 mol%) [dbim][BF4] O

O N

O O

N

O

92% ee

N

tBu tBu (S,S)-tBu-box

Scheme 10.6

The same reaction was also studied by Doherty et al.50 Using Pt(II) catalysts containing (S)-BINAP or the conformationally flexible NUPHOS-type diphosphines 15 and 16 (Figure 10.10), the Diels–Alder reaction was found to proceed smoothly at room temperature regardless of the nature of ILs and in all the ILs studied, better enantioselectivities

276

Recoverable and Recyclable Catalysts R R

PPh2 PPh2

R

R 15: R = Ph 16: R = Me

Figure 10.10 Conformationally flexible diphosphines

(up to 95% ee) and faster reaction rates were observed than those in CH2Cl2. The better performance of the conformationally flexible NUPHOS diphosphines in ILs is partly attributed to significantly decreased rate of racemization. The Pt–BINAP and Pt–15 catalysts proved to be recyclable in [emim][NTf2] or [bmim][PF6], retaining the enantioselectivity and reactivity after three runs. When a Pd(II) catalyst containing the phosphinooxazolidine 17 was employed, N-crotonoyloxazolidione reacted with cyclopentadiene, furnishing good yields (52–89%) and ee values (76–96%) in a group of ILs (Scheme 10.7).51 The results were influenced by the nature of ILs, however, with the best results of 96% ee and 89% yield obtained in [bmim] [BF4], which are superior to those in CH2Cl2. The catalyst in the IL was reusable but became much less efficient after three cycles. O

O N

O

O

[Pd-17][SbF6]2 (5 mol%)

+

o

IL or IL/CH2Cl2, -40 C

IL: [bmim][BF4], [bmim][PF6], [bmim][SbF6], [bmim][OTf], [bmim][ClO4], [bmim][Tf2N], [dmbim][BF4]

O

N

O

up to 96% ee N O

Ph Ph

Ph2P 17

Scheme 10.7

The Diels–Alder reaction of both cyclic and open-chain dienes with 2-methacrolein and 2-bromoacrolein was shown to be feasible with In(III)–(S)-BINOL catalysis in neat [hmim] [PF6] at room temperature (Scheme 10.8).52 Enantioselectivities up to 98% ee were obtained, and the yields were much better than those achievable in CH2Cl2. In addition, the chiral In(III) catalyst could be reused six times with retention of catalytic efficiency. To minimize catalyst leaching, Doherty and coworkers developed the imidazoliumtagged Box ligands 18–20 (Figure 10.11), which were applied to Cu(II)-catalyzed Diels–Alder reactions in [emim][NTf2].53 In comparison with CH2Cl2, [emim][NTf2] provided much faster reaction rates and significantly increased enantioselectivities. For example, the reaction between N-acyloyloxazolidinone and cyclopentadiene in the IL led to

Asymmetric Catalysis in Ionic Liquids O R

H

+

diene

In-(S)-BINOL (20 mol%) allyltributyl stannane 4A MS, [hmim][PF6]

277

Diels-Alder Adduct up to 98% ee

R = Me, Br OH OH (S)-BINOL

Scheme 10.8

full conversion and up to 95% ee within 2 min, whilst it took 1 h to finish the reaction in CH2Cl2 with only 16% ee. Remarkably, the catalyst loading could be decreased to 0.5 mol% in the IL with retained reactivity and enantioselectivity. The introduction of the imidazolium tag into Box confers higher IL-affinity onto the catalyst, and as a result, the catalyst could be readily recycled at least 10 times without any decrease in efficiency. Asymmetric Allylic Substitution Pd-catalyzed asymmetric allylic substitution reactions represent one of the most useful methods for enantioselective synthesis of CC and CX bonds. It would be highly desirable to develop recyclable catalysts to maximize the total turnover numbers. Toma and coworkers described the enantioselective allylic substitution reaction of racemic (E)-1,3diphenyl-3-acetoxyprop-1-ene with dimethyl malonate-catalyzed by Pd–BPPFA and Pd–BPPFDEA in [bmim][PF6] (Scheme 10.9).54 A significant increase in enantioselectivities was noted in the IL compared with those obtained in THF, with the ligand BPPFDEA displaying better performance than BPPFA. The catalyst-containing IL was reusable after washing with water and drying over Na2SO4, but showing reduced yields and ee values. The ferrocene-based ligands (R,S)-BCyPFA, (S,S)-iPr-Phosferrox and (S,R)-iPr-Phosferrox were also examined for the same reaction (Figure 10.12), affording enantioselectivities comparable or better than with BPPFA and (R,S)-BPPFDEA.55 Optimizing of the reaction conditions using (S,R)-iPr-Phosferrox led to a 98% yield and 88% ee at 60 C.

N

N X

O

O

N N R R 18: R = nBu, X = Br 19: R = tBu, X = Br 20: R = tBu, X = NTf2

Figure 10.11 Imidazolium-tagged Box ligands

278

Recoverable and Recyclable Catalysts OAc Ph

CO2Me

+

CO2Me

Ph

CH(CO2Me)2

Pd(0)-BPPFA or BPPFDA (2 mol%)

Ph * Ph up to 74% ee

K2CO3, [bmim][PF6] OH

PPh2 NMe2

N OH

Fe PPh2

Fe

PPh2

PPh2

(R,S)-BPPFDEA

(S,R)-BPPFA

Scheme 10.9 NMe2

O Fe PPhN 2

Fe PCy2 PCy2 (R,S)-BCyPFA

(S,S)-iPr-Phosferrox

PPh2

O

Fe

N

(S,R)-iPr-Phosferrox

Figure 10.12 Ferrocenyl ligands for allylic substitution

New imidazolium-tagged ferrocene ligands have recently been introduced. In the Pd-catalyzed allylic substitution shown in Scheme 10.9, the ligands 21 and 23 gave rise to moderate enantioselectivities in [bmim][PF6]; but no reaction occurred with 22 (Figure 10.13).56 Using 24 led to a good yield (97%) and enantioselectivity (92% ee), and the same was true with 25. However, recycling of Pd–25 proved problematic; the recovered catalyst in the IL failed to reproduce the activity and enantioselectivity. Asymmetric allylic amination has also been demonstrated in ILs. A recent example is seen in the reaction of 1,3-diphenyl-2-propenyl acetate with di-n-propylamine in [dmbim] [BF4] (Scheme 10.10).57 Good conversions (up to 100%) and moderate ee values (up to 84%) were obtained with palladium catalysts ligated with the monodentate phosphorus ligands 26–28. The catalyst bearing 26 and 28 could be reused three times with retained enantioselectivity but decreased yields.

O N Fe E

Y 4

N

N

PF6

N Fe PPh2

O 4

Br

N Fe PPh2

PPh2 21: E = PPh2, Y = CO 22: E = SePh, Y = CO 23: E = PPh2, Y = CH2

24

Figure 10.13 Tagged ferrocenyl ligands

PPh2 25

4

N PF6

N

Asymmetric Catalysis in Ionic Liquids OAc Ph

+ Ph

L:

HN

N

N(C3H7)2

C3H7 [Pd(allyl)L2][BF4] (2 mol%)

P R

Ph * Ph up to 84% ee

[dmbim][PF6]

C3H7

279

O P N O

N Ph

26: R = OtBu 27: R = OAdamantyl

28

Scheme 10.10

Asymmetric Carbonyl–Ene Reactions The transition metal-catalyzed asymmetric carbonyl–ene reaction is an important method for the preparation of optically active homoallylic alcohols. Several efficient catalytic systems have been developed; but investigations into recovery of the chiral catalysts have been less documented.1 The Pt-catalyzed carbonyl–ene reactions of glyoxal derivatives with 1,10 -disubstituted alkenes in [emim][NTf2] was recently explored, using the conformationally flexible acyclic and cyclic NUPHOS diphosphines 15, 16 and 29 as ligands (Scheme 10.11).58 The reactions proceeded smoothly in IL, providing good yields (up to 90%) and high enantioselectivities (up to 95% ee). The catalyst was substrate dependent and also sensitive to the conformation of its precursor. Thus, the catalysts generated from d- and l-[Pt(S)-BINOL)(NUPHOS)] performed better in IL than in CH2Cl2, exhibiting marked enhancement in enantioselectivity. However, better performance was always observed with the catalyst bearing 29. Again, it was suggested that the increased enantioselectivity might be due to a decreased rate of racemization of the conformationally flexible ligands in the IL.

O R1

R2

+

H

R3

O R1 = R2 = -(CH2)4-, R3 = OEt R1 = R2 = -(CH2)4-, R3 = Ph R1 = H, R2 = Ph, R3 = OEt R1 = tBu, R2 = Me, R3 = Ph

Pt-15, 16 or 29 (5 mol%)

R1

[emim][NTf2]

R2

OH * R3

O up to 95% ee

Et PPh2 PPh2 Et 29

Scheme 10.11

These reactions can also be effected in [dmbim][NTf2] with Pd(II) catalysis using the conformationally restricted (R)-BINAP as ligand, affording up to 92% ee.59 The IL was found to stabilize the chiral catalyst, allowing it to be recycled 21 times with retained enantioselectivity, albeit with slowly decreasing activity.

280


Asymmetric Intramolecular Hydroacylation Asymmetric intramolecular hydroacylation was demonstrated in neat [bmim][NTf2] by Sato and coworkers (Scheme 10.12).60 Catalyzed with [Rh(R)-BINAP)][ClO4], 2-(prop-1en-2-yl)benzaldehyde underwent enantioselective hydroacylation in the IL, furnishing more than 99% yield and 99% ee. The catalyst could be repeatedly used for at least 5 times, with 99% ee in each cycle. O

O [Rh((R)-BINAP)][ClO4] (1 mol%)

H

[bmim][NTf2] 99% ee

Scheme 10.12

Asymmetric Friedel–Crafts Reaction An example of asymmetric Friedel–Crafts reactions in ILs is seen in the Ti-catalyzed reaction of anilines with ethyl glyoxylate in the pyridinium-based ILs [epy][CF3CO2] and [epy][BF4] (Scheme 10.13).61 The anion of the IL was found to influence both the yield and enantioselectivity, with better results recorded in [epy][CF3CO2] with Ti–(R)-BINOL-Br as catalyst. The combination of [epy][CF3CO2] and Ti–(R)-BINOL-Br was also effective for the alkylation of other aromatic amines, providing good yields (up to 89%) and ee values (up to 91%). O HO * O

R

H

R OEt

+

OEt

Ti-(R)-BINOL-Br (10 mol%) pyridinium IL

O

NMe2

NMe2

up to 91% ee

R = H, Cl, Br, Me, OMe IL: [epy][BF4], [epy][CF3CO2]

Br OH OH Br (R)-BINOL-Br

Scheme 10.13

Asymmetric Mukaiyama Aldol Reaction The immobilization of chiral catalysts for the asymmetric Mukaiyama aldol reaction has received increasing attention.2,3 Doherty and coworkers recently reported a Cu-catalyzed asymmetric Mukaiyama aldol reaction of methyl pyruvate under homogeneous conditions


281

in ILs and under heterogeneous conditions with the catalyst and IL supported on silica (Scheme 10.14).62 Using (S,S)-tBu-Box or 20 as ligand, the ILs allowed the reactions to finish in 2 min while CH2Cl2 required 1 h to reach moderate to good conversions (55–90%). However, the enantioselectivity varied with the nature of ILs, with the best value (90% ee) found in [emim][NTf2]. Further, the reactions in ILs suffered from inferior chemoselectivity as evidenced by the formation of a self-condensation product 3-hydroxy-1, 3-diphenylbutan-1-one. This could be circumvented by using the biphasic [emim] [NTf2]/Et2O, however. The catalysts supported on silica or IL-modified silica displayed a similar enantioselectivity but slightly decreased activity. The imidazolium-tagged ligand 20 allowed the catalyst to be recycled up to five times without reduction in conversion or ee, but as maybe expected, leaching was observed with (S,S)-tBu-Box.

Ph

OTMs

+

O

Me

[Cu(Box)](OTf)2 (10 mol%) IL or IL/Et2O

OMe

O

Ph

O HO Me OMe

O up to 94% ee

IL: [emim][NTf2], [bmpyrr][NTf2], [N3336][NTf2], silica-supported [emim][NTf2], imidazolium-modified silica-supported [emim][NTf2].

Scheme 10.14

Asymmetric Cyclopropanation The transition metal-catalyzed asymmetric cyclopropanation of olefins provides a convenient route to optically active cyclopropane compounds. A number of chiral catalysts have been reported, with Cu–Box complexes being one of the most efficient catalysts.1 In 2001, Mayoral et al. reported asymmetric cyclopropanation of styrene with ethyl diazoacetate catalyzed by a Cu–(S,S)-tBu-Box complex in ILs (Scheme 10.15).63 Among the ILs [emim] [NTf2], [emim][BF4] and [Oct3NMe][NTf2], [emim][NTf2] in combination with CuCl2 led to the best results, affording the products in a trans/cis ratio of 62/38 with similar ee values, at 85–86%. Interestingly, the catalyst was much less efficient in CH2Cl2. This is probably due to easy dissociation of the chloride in the IL. Ph

+ N2CHCOOEt

Cu-(S,S)-tBu-Box (1 mol%) IL

+ Ph

COOEt

Ph

COOEt

IL: [emim][NTf2], [emim][BF4], [Oct3NMe][NTf2]

Scheme 10.15

Further studies revealed that ILs allowed for more robust immobilization of cationic Cu–Box catalysts than anionic solids, such as Laponite, a synthetic clay. This arises from the high stability of the catalyst in the IL; in contrast, the solid-supported catalyst suffered from easy ligand dissociation and consequently decreased enantioselectivity. However, the catalyst performance is strongly dependent on the nature of IL, water content and the molar ratio of IL/catalyst. The catalyst immobilized in [emim][OTf] or [bmim][OTf] could be recycled five times with retained reactivity but decreased enantioselectivity.64 Switching

282

Recoverable and Recyclable Catalysts N

O N

O N

tBu

tBu

Azabis(oxazoline)

Figure 10.14 A modified BOX ligand

to the more electron-donating and hence more coordinating ligand azabis(oxazoline) improved the catalyst stability (Figure 10.14). Thus, the Cu–azabis(oxazoline) catalyst could be recycled up to eight times without loss of activity and enantioselectivity in the same cyclopropanation.65 More recently, supported IL films of nanometric thickness has been explored to immobilize Cu–Box complexes for recoverable asymmetric cyclopropanation.66 However, only low to moderate enantioselectivities and yields were obtained. Not unexpectedly, halide contaminants have adverse effects on the cyclopropanation. Thus, whilst styrene underwent smooth cyclopropanation with ethyl diazoacetate in [bmim][OTf], [bmim][PF6], [bmim][NTf2] or [bmim][BF4], affording yields up to 61% comparable with those obtained in CHCl3, the presence of chloride or bromide anions (5%) almost totally deactivated the catalyst.67 In [bmim][BF4], the Cu–(S,S)-tBu-Box catalyst was recyclable, affording a mixture of trans/cis isomers with ee values up to 98%, higher than those in [emim][NTf2].63 Asymmetric Allylation of Aldehydes and Ketones Asymmetric allylation of aldehydes and ketones provides easy access to optically active secondary and tertiary homoallylic alcohols, important intermediates in the enantioselective synthesis of bioactive compounds.1 The first, In(III)-catalyzed enantioselective allylation of aldehydes in ILs was recently reported (Scheme 10.16).68 The reaction was carried out at 60 C, necessitating a co-solvent. In the model reaction of benzaldehyde with allyltributylstannane, an 88% ee was observed in [hmim][PF6]/CH2Cl2 with (S)-iPrPybox as the ligand. Awide variety of different aldehydes were also allylated to provide the homoallylic alcohols in good yields and enantioselectivities (up to 94%). The In–(S)-iPrPybox catalyst was reusable. O +

In(III)-(S)-iPr-Pybox (20 mol%)

OH

SnBu3 [hmim][PF ]/CH Cl , -60 oC, TMSCl, 4A MS R * 6 2 2 up to 94% ee R = Ph, Naph, 4-ClPh, 6-MeFuran, PhCH=CH, PhCH2O(CH2)3, nC8H17 R

H

Ph Ph

O

O

N N

N

(S)-iPr-Pybox

Scheme 10.16

Ph Ph


283

The reaction could be run in neat ILs as well. Using an In–(S)-BINOL catalyst at room temperature, the allylation led to enantioselectivities up to 92% ee. However, catalyst recycling failed, presumably due to catalyst hydration leading to its deactivation.69 The In-catalyzed asymmetric allylation in ILs has been extended to ketones. With the chiral catalyst generated in situ from indium triflate and (S)-iPr-Pybox, the benchmark allylation of acetophenone with allyltributyl stannane ran smoothly in [hmim][PF6]/ CH2Cl2 at 0 C, affording a 82% yield and 65% ee (Scheme 10.17).70 Further studies showed that the catalyst also worked well for other ketones, leading to good yields (up to 82%) and ee values (up to 93%) and being recyclable. O

OH R * [hmim][PF6]/CH2Cl2, 0 C, TMSCl, 4A MS up to 93% ee In(III)-(S)-iPr-Pybox (20 mol%)

+

R

SnBu3

o

R = Ph, 4-MePh, Naph, PhCH2CH2, PhCH=CH

Scheme 10.17

Asymmetric Michael Reaction Sodeoka and coworkers investigated the asymmetric Michael reactions of b-ketoesters with methylvinyl ketone using [Pd(H2O)2((S)-BINAP)][OTf]2 or [Pd(H2O)2((S)-tolBINAP)] [OTf]2 as catalyst in ILs (Scheme 10.18).71 Good yields (up to 98%) and enantioselectivities (up to 88% ee) were observed, which are comparable with those obtained in common organic solvents. However, recycling of the catalyst immobilized in [hmim][BF4] or [bmim][OTf] was met with catalyst decomposition into palladium black. Interestingly, addition of the acidic benzenesulfonimide could activate the enone, resulting in a higher reaction rate, and further, enabled the catalyst to be reused at least five times with maintained catalytic efficiency. O

O CO2tBu

R1 R2

Pd-BINAP or tolBINAP (5 mol%) R1 +

O

IL, 0 oC

CO2tBu R2

O

up to 88% ee

R1 = R2 = Me R1 = R2 = -(CH2)3R1 = R2 = -(CH2)4IL: [bmim][OTf], [hmim][BF4]

Scheme 10.18

10.2.5 Miscellaneous Reactions Asymmetric Ring-opening of Epoxides Asymmetric ring-opening of meso-epoxides by TMSN3 catalyzed by Jacobsen’s Cr–Salen complex 30 in ILs was reported by Song et al. (Scheme 10.19).72 The reaction proceeded well in hydrophobic ILs, such as [bmim][PF6] and [bmim][SbF6], providing up to 97% ee.

284

Recoverable and Recyclable Catalysts Y

Y

TMSN3, 30 (3 mol%) IL

O

N3 OTMS up to 97% ee

Y = CH2, (CH2)2, O

IL: [bmim][PF6], [bmim][SbF6], [bmim][BF4], [bmim][OTf].

H N t Bu

O tBu

H N Cr Cl

tBu

O t Bu

30

Scheme 10.19

However, the hydrophilic IL [bmim][BF4] gave only a 5% yield and 3% ee, and no reaction took place in [bmim][OTf], although they were more effective in catalyst immobilization. Hydrolytic kinetic resolution of racemic epoxides was also demonstrated, as shown in Scheme 10.20.73 The Co(III)–Salen catalyst 31 afforded the epoxides with ee values as high as >99% in an ILs/THF mixture. UVand XPS analysis indicated that the oxidation state of cobalt remained þ III, not þ II, after completion of the reaction. When organic solvent is employed, the Co(III)–Salen complex is usually reduced to Co(II)–Salen during the reaction. Evidently, the IL could stabilize the Co(III) complex against reduction. The use of ILs also allowed for recycle of the chiral catalyst, showing no loss of catalytic activity and enantioselectivity for ten times. O

IL/THF

R

OH

O

H2O, 31

+

OH

R R up to >99% ee

R = CH2Cl, Me, nBu, Ph IL: [bmim][PF6], [bmim][NTf2]

H N t Bu

O tBu

H N Co OAc O t Bu 31

tBu

Scheme 10.20

Asymmetric Cyanosilylation of Aldehydes Asymmetric cyanosilylation of aldehydes using TMSCN in ILs catalyzed by a chiral vanadyl–Salen complex 32 was reported by Corma and coworkers (Scheme 10.21).74 Screening the reaction in [bmim][PF6], [emim][PF6], [bmim][Cl], and [bmim][BF4]


285

revealed a significant influence of counteranions. Thus, when carried out in [bmim][Cl] or [bmim][BF4], the reaction led to a low yield and enantioselectivity. In contrast, good results comparable with those achieved in CH2Cl2 were recorded with [bmim][PF6] and [emim] [PF6]. As with the Cr(III) and Co(III)–Salen catalysts above, 32 is recyclable. O

R

IL

H

R

OTMS

TMSCN, 32 (1 mol%)

H CN

up to 98% ee R = Ph, PhCH=CH, 2-FPh, nC5H11 IL: [emim][PF6], [bmim][PF6], [bmim][Cl], [bmim][BF4] H N tBu

O tBu

H N V O 32

O

tBu

tBu

Scheme 10.21

Asymmetric Fluorination Pd(II)-catalyzed asymmetric fluorination of b-ketoseters was reported by Sodeoka et al.71 With [Pd(OH)((R)-XylBINAP)]2[OTf]2 as catalyst, the fluorination with NFSI in [hmim] [BF4] afforded enantioselectivities up to 92% ee at room temperature (Scheme 10.22). After completion of the reaction, the product was separated from the IL by extraction with ether, and the catalyst was recyclable for more than ten times without loss of enantioselectivity. O

O CO2tBu

R1

+ R2

Pd-(R)-XylBINAP (2.5 mol%) PhO2S NF PhO2S IL

R1

NFSI R1 = Ph, R2 = Me R1 = Me, R2 = Et R1 = R2 = -(CH2)3-

CO2tBu

R2 F up to 92% ee

PAr2 PAr2

IL: [bmim][BF4], [bmim][OTf], [hmim][BF4].

(R)-XylBINAP Ar = 3,5-dimethylphenyl

Scheme 10.22

Asymmetric fluorination of b-keto phosphonates with NFSI in ILs was shown to be feasible with a related catalyst [Pd(H2O)(NCMe)((R)-BINAP)][BF4]2 (Scheme 10.23).75 Higher yields and enantioselectivities were obtained in [bmim][BF4] and [bmim][OTf]

286


than in [bmim][PF6], [bmim][SbF6] and [emim][BF4]. Using [bmim][BF4] as solvent, b-keto phosphonates were fluorinated with enantioselectivities up to 93% ee; the results are similar to those in MeOH. In the IL, however, the catalyst was reusable. O

O PO(OR)2 +

PhO2S NF PhO2S

n

* PO(OR)2 F

Pd-(R)-BINAP ( 2.5 mol%) IL n

NFSI

up to 93% ee

n = 1, R = Et n = 1, R = Me n = 1, R =iPr n = 0, R = Et IL: [bmim][BF4], [bmim][OTf], [emim][BF4], [bmim][SbF6], [bmim][PF6].

Scheme 10.23

Asymmetric Addition of Alkynes to Imines Alkynes add to imines under acid catalysis. An example in ILs is seen in the coppercatalyzed enantioselective addition of phenylacetylene to imines with the pyridinyl Box 33 as ligand (Scheme 10.24).76 Among the ILs examined, [bmim][NTf2] appears to be the best choice. Thus, under the catalysis of Cu–33 in [bmim][NTf2], the asymmetric addition of phenylacetylene to imines afforded propargylamines in high yields (up to 92%) and enantioselectivities (up to 99%). The catalyst-containing IL could be reused.

N Ar

Ph +

[Cu-33][OTf]2 (1 mol%)

Ph

IL

H

HN * Ar

Ph

Ph up to 99% ee

Ar = Ph, 4-MePh, 4-CF3Ph, 4-ClPh, 4-BrPh, 2-Naph IL: [bmim][NTf2], [bmim][PF6], [omim][PF6], [dmbim][NTf2], [dmbim][BF4]. O Ph

O

N

N

N 33

Ph

Scheme 10.24

Asymmetric Hydrovinylation of Styrene Leitner and coworkers described the asymmetric hydrovinylation of styrene with ethylene in IL/scCO2 using Wilke’s catalyst 34 (Scheme 10.25).77 On screening a group of ILs, the best enantioselectivity of 89.4% ee was observed in [emim][BARF]; the high ee may arise from the low coordination ability of the anion. The use of IL, which does not dissolve in scCO2, allowed the product to be continuously extracted with compressed CO2.

Asymmetric Catalysis in Ionic Liquids 34 (0.3 mol%)

+

287

*

IL/scCO2

IL: [emim][Y], [bmpy][Y] (Y = NTf2, BF4, BARF, Al(OC(CF3)2Ph)4)

Ni

Cl

P

N N

P Ni

up to 89.4% ee

Cl

34

Scheme 10.25

Asymmetric Sulfimidation of Sulfides Asymmetric sulfimidation of sulfides with Cu(II) catalysis was demonstrated in [bmim] [BF4] (Scheme 10.26).78 With the Box ligand 35, the sulfimidation proceeded smoothly in the IL, giving products in good yields but only moderate enantioselectivities (up to 50%). The product separation was easy, however.

R

S

Cu(acac)2/35 (6 mol%) PhI=NTs, [bmim][BF4]

R = 4-MePh, Ph, 4-BrPh

R up to 50% ee

O

O N Ph

NTs S*

N 35

Ph

Scheme 10.26

10.3 Asymmetric Organocatalytic Reactions in ILs Asymmetric catalysis with metal-free, organic molecules has become an important area of research, and a number of efficient organocatalysts have been developed in the past several years.1e As with metal-catalyzed reactions, recovery and reuse of the often expensive organocatalysts is necessary, so is the need to replace the commonly used organic solvents. 10.3.1 Asymmetric Aldol Reactions The asymmetric aldol reaction is one of the most useful processes for the preparation of optically active b-hydroxy carbonyl compounds. In 2002, Loh and coworkers first reported

288


the asymmetric aldol reaction in imidazolium ILs with L-proline as catalyst.79 A series of ILs were screened in the benchmark reaction of acetone with benzaldehyde, providing comparable or better enantioselectivities than those obtained in organic solvents. The reaction in [omim][Cl] proceeded faster than in other ILs but the yield was lower due to the formation of elimination product. With [bmim][PF6] as the reaction medium, ee values up to 89% were obtained in the addition of ketones to various aldehydes (Scheme 10.27). The reaction has been extended to other ketones and aryl aldehydes, with ee values up to 93% observed.80 Catalyst recycle in the imidazolium IL was shown to be feasible. More recently, guanidinium ILs have been explored for the L-proline-catalyzed asymmetric aldol reaction of acetone with aldehydes, affording better yields (up to 82%) and enantioselectivities (up to 99% ee) than those obtained in acetone or [bmim][PF6] at 25 C.81 O

O +

R1

R3

O

L-proline (30 mol%) H

R1

R3 R2 up to 89% ee

IL

R2 R1 = Me, R2 = H, R3 = Ph R1 = Me, R2 = H, R3 = Nap R1 = Me, R2 = H, R3 = 4-BrPh R1 = Me, R2 = H, R3 = cyclohexyl R1 = R2 = -(CH2)3-, R3 = Ph

OH

COOH N H L-proline

IL: [hmim][BF4], [omim][Cl], [omim][BF4], [bmim][PF6].

Scheme 10.27

A more effective catalyst appears to be the L-prolinamide 36. Thus, in the aldol reactions of acetone and butanone with various aldehydes at 0 C, 36 gave rise to moderate to good yields (up to 84%) and excellent ee values (up to >99%) in [bmim][BF4] (Scheme 10.28).82 Under similar conditions, use of L-proline as catalyst or organic solvents afforded lower O R1

O H

+

OH O

36 (20 mol%) IL, 0 oC

R2

R2 R1 up to >99% ee

R1 = 4-NO2Ph, 2-NO2Ph, 4-BrPh, 4-ClPh, 2-ClPh, Ph, 4-CNPh, 4-MePh, 3-NO2Ph, cyclohexyl, tBu R2 = Me, Et IL = [bmim][BF4], [bmim][PF6] O

Ph N H

N H

Ph HO

36

Scheme 10.28


289

ee values, and replacing [bmim][BF4] with [bmim][PF6] led to the same obervation. In the former IL, the catalyst loading could be reduced to 5 mol% without affecting the enantioselectivity, and the catalyst was reusable. Cross-aldol reactions were also demonstrated in ILs. For instance, in the presence of L-proline in [bmim][PF6], the self-aldol reaction of propionaldehyde led to a moderate yield (51%) and an excellent ee (>99%) (Scheme 10.29).83 The problematic oligomerization side reaction could be inhibited by adding DMF as co-solvent. Thus, smooth crossaldol reactions of a range of aldehydes took palce in [bmim][PF6]/DMF, affording b-hydroxy aldehydes in moderate to good yields (up to 78%) and excellent enantioselectivities (up to >99% ee).

H

R1

OH O

O

O +

L-proline (5 mol%) H

R2

[bmim][PF6]/DMF, 4 oC

R1 = Et, R2 = Me R1 = iPr, R2 = Me R1 = iBu, R2 = Me R1 = nhexyl, R2 = Me R1 = iPr, R2 = nBu

H R2 up to >99% ee

R1

Scheme 10.29

Attempting to enhance the catalyst recyclability, the tagging strategy has been applied to organocatalysts. An example is seen in the onium-tagged L-proline ligands 37–39 (Figure 10.15).84,85 These catalysts allow for faster aldol reactions than the parent proline in imidazolium ILs, and provide better ee values in the ILs than in organic solvents. For instance, 39 catalyzed the aldol reactions of acetone with aryl aldehydes in [bmim][BF4], providing good yields (up to 94%) and high ee values (up to 93%) at room temperature at a 10% catalyst loading.85 Whilst 37 and 38 were met with eroded activity and selectivity when reused, 39 in [bmim][BF4] worked in six repetitive cycles with maintained enantioselectivity but decreased yield. Silica gel-supported ILs have also been introduced as reaction media to improve the efficiency of catalyst recycling (Scheme 10.30).86–88 In these materials, the IL moiety is covalently attached as a monolayer to the surface of a silica gel support, and the organocatalyst is immobilized with or without additional IL. In comparison with unmodified silica gel, the supported ILs conferred better performace on L-proline in the reaction of acetone with benzaldehyde, and they also facilitated the reactions of acetone with other

NTf2

NTf2 N

O

N

O

O 37

N H

OH

Br O

BuMe2N

O

O 38

N H

N

O

N

OH

Figure 10.15 Onium-tagged L-prolines

39

O N H

OH

290


aldehydes, providing moderate to good yields (up to 95%) and high ee values (up to 96%). The catalysts so immobilized could be readily recovered by simple filtration and reused. For instance, the regenerated [SI1][BF4]/[bmim][BF4]/L-proline system functioned well in 13 consecutive cycles without losing efficiency.

N N

N

X

MeO Si OO [SI1][X] X = BF4, PF6, Cl.

N N

BF4

MeO Si OO

OMe O Si O

BF4

N

N

BF4

[SI4][BF4]

MeO Si OO

[SI2][BF4]

S

OMe O Si O

[SI3][BF4]

S

N

N

X

[SI5][X] X = BF4, PF6, Cl.

Scheme 10.30

10.3.2 Asymmetric Michael Addition Asymmetric Michael addition with chiral organocatalysts has attracted a great deal of recent attention.1e In ILs such as [bmim][PF6], [bmim][BF4], and [bpy][BF4], the Michael addition of dimethyl malonate to 1,3-diphenylprop-2-en-1-one was shown to be feasible in the presence of a quaternary ammonium salt 40 as phase transfer catalyst (Scheme 10.31).89 The reactions proceeded faster in the ILs, especially in the case of [bmim][PF6], than in the usual organic solvents, affording better yields (up to 99%) and comparable enantioselectivities (up to 50% ee). Surprisingly somehow, the product configuration was reversed on going from the organic solvents to the ILs. The phase transfer catalyst in [bmim][PF6] could be used in three consecutive runs without loss of efficiency.

O Ph

Ph

+

CO2Me 40 (8 mol%) CO2Me

IL, K2CO3

MeO2C CO2Me O * Ph Ph up to 50% ee

IL = [bmim][PF6], [bmim][BF4], [bpy][BF4]

N

HO

Br O S O

MeO N 40

Scheme 10.31


291

L-proline was also shown to catalyze the Michael addition in imidazolium ILs (Scheme 10.32).90 In the addition of cyclohexanone to nitrostyrene, the ILs [moemim] [OMs] and [bmim][Cl] led to better enantioselectivities (up to 75% ee) than those with organic solvents, whereas comparable results were observed in [bmim][BF4] and [bmim] [PF6]. Acyclic ketones were less reactive and less enantioselective, however. Similar observations were made in the addition of aldehydes and ketones to nitrostyrene and 2-(b-nitrovinyl)thiophene in a range of imidazolium ILs.91 O

NO2 +

R1 R2

O

Ph

L-proline (40 mol%) R1

IL

Ph

*

NO2

R2 up to 75% ee

R1 = Me, R2 = H; R1 = Ph, R2= H; R1 = Et, R2 = Me; R1 = Me, R2 = CO2Et; R1 = R2 = -(CH2)4-; R1 = R2 = -(CH2)3IL = [moemim][OMs], [bmim][BF4], [bmim][PF6], [bmim][Cl].

Scheme 10.32

Modified prolines have also been featured in the Michael addition in ILs. For instance, the morpholine-based 41 promoted the asymmetric 1,4-conjugate addition of unmodified aldehydes to methyl vinyl ketone in [bmim][PF6] (Figure 10.16).92 The IL was found to accelerate the reaction in comparison with organic solvents. Under the catalysis of 41, however, the conjugate addition afforded only modest results (up to 72% yield and 59% ee). The isoquinoline-modified pyrrolidines 42 and 43 were shown to be highly enantioselective in the Michael addition of cyclohexanone to nitroalkenes in ILs.93 Thus, in the model reaction of cyclohexanone with trans-b-nitrostyrene in [bmim][BF4], 93–95% yields and 81–99% ee values were observed in [bmim][BF4] with 42 and 43 as the catalysts. Other nitroalkenes proved to be viable as well, affording excellent yields (93–95%) and enantioselectivities (up to 100% ee) with no side products detected. In addition, 43 immobilized in [bmim][BF4] could be used in four consecutive runs without decrease in activity and enantioselectivity. Of further interest is that (R,R)-trans-1,2-diaminocyclohexane 44, a common ligand in metal catalysis, could catalyze the asymmetric Michael adition of ethyl cyclohexanone2-carboxylate to methyl vinyl ketone in ILs (Scheme 10.33).94 In [bmim][BF4], the reaction led to up to 92% ee but was slow, necessitating 6 days to completion at a catalyst loading of 5 mol%. Surprisingly, introduction of metal Lewis acids resulted in no improvement. Presumably, 44 facilitates the reaction by forming an active enamine intermediate. O N N H 41

N H

N X 42: X = Br 43: X = PF6

Figure 10.16 Modified L-proline ligands

292


O

O

O

44 (37.5 mol%)

OEt +

O * CO Et 2

IL

up to 92% ee

IL = [bmim][BF4], [dmbim][BF4], [bpy][BF4] NH2

H2N

44

Scheme 10.33

In related addition reactions, the asymmetric addition of aliphatic aldehydes to diethyl azodicarboxylate was shown to be feasible in ILs in the presence of L-proline.95 In particular, the catalyst in [bmim][BF4] allowed the reactions of various aldehydes bearing active a hydrogens to proceed smoothly, affording good yields (up to 85%) and enantioselectivities (up to 89% ee). With the same catalytic system, alkyl aldehydes and ketones also add to nitrosobenzene, resulting in asymmetric a-aminooxylation of the carbonyl compounds (Scheme 10.34).96,97 The results obtained in the IL are comparable with or better than those observed in CHCl3, with the additional advantage of easy catalyst recycle. O

N R1

R2

+

O

O L-proline (20 mol%) [bmim][BF4]

R1

O NHPh

R2 up to >99% ee

R1 = H, R2 = Me; R1 = H, R2 = Et; R1 = H, R2 = nBu; R1 = H, R2 = iPr; R1 = H, R2 = Bn; R1 = R2 = -(CH2)4-; R1 = Me, R2 = Me; R1 = R2 = -(CH2)2C(OCH2CH2O)CH2-

Scheme 10.34 96

10.3.3 Asymmetric Diels–Alder Reaction An example of an asymmetric Diels–Alder reaction catalyzed by an organocatalyst in ILs is shown in Scheme 10.35.98 The imidazolidin-4-one 45 catalyzed the reaction of cyclohexadiene with acrolein, furnishing good yields (up to 85%) and enantioselectivities (up to 93% ee) in [bmim][PF6] and [bmim][SbF6]. In surprising contrast, using [bmim][BF4] and [bmim][OTf] led to racemic products with very low yields, which may stem from the hydrophilic character of these ILs. Extending the reaction to cyclopentadiene and cinnamaldehyde was met with lower enantioselectivities, however. 10.3.4 Asymmetric Mannich Reaction Asymmetric Manich reactions catalyzed by L-proline in ILs were described by Barbas and coworkers (Scheme 10.36).99 In [bmim][BF4] in the presence of L-proline, the N-p-methoxyphenyl-protected a-imino ethyl glyoxylate reacted with cyclohexanone,

Asymmetric Catalysis in Ionic Liquids CHO 45 (5 mol%) IL

+

293

CHO up to 93% ee

IL = [bmim][PF6], [bmim][SbF6], [bmim][OTf], [bmim][BF4]. O N N H 45

Scheme 10.35

resulting in a quantitative yield and excellent enantioselectivity of >99% ee. After four successive runs in the IL, the catalyst retained its enantioselectivity, albeit with some decrease in activity. Similar results were observed in [bmim][PF6]. The catalysis was extended to other aldehydes and ketones, again affording high yields and ee values. Still further, the catalyst loading in the IL could be lowered to 1 mol% without affecting enantioselectivity. However, the Mannich reaction involving hydroxyacetone led to poorer results in [bmim][BF4] compared with those obtained in DMF or DMSO. O

PMP +

R1

H

R2

O

L-proline (5 mol%)

N CO2Et

IL

NHPMP

CO2Et R2 up to >99% ee

R1

R1 = H, R2 = nBu; R1 = H, R2 = nPent; R1= H, R2 = CH2CH=CH(CH2)4CH3; R1 = H, R2 = iPr; R1 = R2 = Me, R1 = R2 = -(CH2)4-; R1 = Me, R2 = H. IL: [bmim][BF4], [bmim][PF6]

Scheme 10.36

10.3.5 Asymmetric Baylis–Hillman Reaction The asymmetric Baylis–Hillman reaction has been investigated in chiral ILs. Scheme 10.37 shows an example of benzaldehyde reacting with methylacrylate in the presence of the O Ar

H

CO2Me DABCO (30 mol%)

+

CIL3-5

OH Ar *

CO2Me

up to 44% ee Ar = Ph, 4-MePh, 4-ClPh, 4-NO2Ph, 3-Py N C8H17

Z

X Ph

CIL3: X = OTf, Z = OH. CIL4: X = PF6, Z = OH. CIL5: X = OTf, Z = OAc.

Scheme 10.37

294


achiral DABCO (1,4-diazabicyclo[2,2,2]octane) as catalyst in the chiral ammonium ILs CIL3–5.100 Enantioselectivities of up to 44% ee were observed, indicative of asymmetric induction by the chiral IL. Presumably, the chirality is relayed from the IL via hydrogen bonding between the hydroxyl group of the IL cation and the carbonyl group of substrate. Consistent with this view, a poorer ee was seen in CIL5, in which the –OH is replaced with a –OAc. ‘‘Chiral-at-anion’’ ILs have also been explored. The methyltrioctylammonium (MtOA) IL CIL6, which contains a chiral dimalatoborate anion, was elaborated for the asymmetric aza-Baylis–Hillman reaction shown in Scheme 10.38.101 Considering the formation of zwitterionic intermediate during the reaction, a chiral IL capable of bifunctional stabilization is expected to suppress racemization and induce high enantioselectivity. With PPh3 as catalyst, the reaction between methyl vinyl ketone and N-(4-bromobenzylidene)-4-toluenesulfonate led to conversions up to 39% and enantioselectivities up to 84% ee. A 64% ee was obtained with N-(4-methoxybenzylidene)-4-toluenesulfonate; but the analogous, electron-deficient nitrobenzylidene yielded only a 10% ee. Replacing CIL6 with the structurally related CIL7 or CIL8 afforded much lower conversions and racemic products, indicating the necessity of a Brønsted acid functionality within the chiral ion. N

Tos

Tos O

H

+

PPh3 (10 mol%) CIL6-8

R R = Br, Me, NO2

NH O *

R up to 84% ee [MtOA] O

[MtOA] HO

O

O

OH

O B O

O CIL6

O

[MtOA] Ph O B O O CIL7

O

O O

O Ph O

B

O

O

O O

O CIL8

Scheme 10.38

10.4 Concluding Remarks The many examples featured in this chapter demonstrate that ILs are effective alternatives to the common organic solvents as reaction media for asymmetric catalytic reactions. The unique physicochemical propertites of ILs offer a number of advantages for asymmetric catalysis; these include easy and effective catalyst immobilization that often requires no ligand modification, facile product separation and catalyst recycling, compatibility with and stabilization of most chiral catalysts, and comparable or better catalyst activity and enantioselectivity in comparison with those in organic media. Of further interest is the use of chiral ILs as both the reaction medium and source of chirality. Considering the numerous possibilities in IL design, this opens a new door for asymmetric catalysis without traditional catalysts.


295

There is still much to be discovered and improved, however. For instance, common organic solvents are frequently used in reactions and in product extraction; this would diminish the value of using ILs. Further, given the cost of ILs, much higher S/C ratios or TON values in ILs than achievable in a normal solvent are necessary for any practical uses. Still further, detailed mechanistic understanding of reaction pathways in ILs and their effects on reactants and active species at the molecular level are lacking, which is needed if the full potential of ILs in asymmetric catalysis is to be seized.

References 1. (a) R. Noyori, Asymmetric Catalysis in Organic Synthesis, John Wiley & Sons, Inc., New York, 1994; (b) E. N. Jacobsen, A. Pfaltz and H. Yamamoto (Eds), Comprehensive Asymmetric Catalysis, Springer, Berlin, 1999; (c) I. Ojima (Ed.), Catalytic Asymmetric Synthesis, WileyVCH, New York, 2000; (d) G. Q. Lin, Y. M. Li and A. S. C. Chan, Principles and Applications of Asymmetric Synthesis, John Wiley & Sons, Inc., New York, 2001; (e) A. Berkessel, H. Groger, Asymmetric Organocatalysis—From Biomimetic Concepts to Powerful Methods for Asymmetric Synthesis, Wiley-VCH, Weinheim, 2005. 2. (a) C. J. Li and T. H. Chan, Organic Reaction in Aqueous Media, John Wiley & Sons, Inc., New York, 1997; (b) A. D. Pomogailo, Catalysis by Polymer-Immobilized Metal Complexes, Gordon & Breach, New York, 1998; (c) P. G. Jessop and W. Leitner (Eds.), Chemical Synthesis using Supercritical Fluids, Wiley-WCH, Weinheim, 1999; (d) D. E. De Vos, I. F. J. Vankelecom and P. A. Jacobs, Chiral Catalyst Immobilization and Recycling, Wiley-VCH, Weiheim, 2000; (e) F. Joo, Aqueous Organometallic Catalysis, Kluwer, Dordrecht, 2001. 3. (a) For a special issue on recoverable catalysts and reagents, see: Chem. Rev. 102, 3215–3892 (2002); (b) G. Pozzi and I. Shepperson, Fluorous chiral ligands for novel catalytic systems, Coord. Chem. Rev., 242, 115–124 (2003); (c) D. J. Cole-Hamilton, Asymmetric catalytic synthesis of organic compounds using metal complexes in supercritical fluids, Adv. Syn. Catal., 348, 1341–1351 (2006); (d) C. J. Li, Organic reactions in aqueous media with a focus on carbon–carbon bond formations: a decade update, Chem. Rev. 105, 3095–3165 (2005); (e) S. Liu and J. Xiao, Toward green catalytic synthesis – Transition metal-catalyzed reactions in nonconventional media, J. Mol. Catal. A: Chem., 270, 1–43 (2007); (f) H. Zhou, Q. Fan, Y. He, L. Gu and A. S. C. Chan, Application of liquid poly(ethylene glycol)s as green reaction media in organic reactions, Prog. Chem., 19, 1517–1528 (2007). 4. (a) R. Sheldon, Catalytic reactions in ionic liquids, Chem. Commun., 2399–2407 (2001); (b) J. Dupont, R.F. de Souza and P.A.Z. Suarez, Ionic liquid (molten salt) phase organometallic catalysis, Chem. Rev., 102, 3667–3692 (2002); (c) T. Welton, Ionic liquids in catalysis, Coord. Chem. Rev., 248, 2459–2477 (2004); (d) S. G. Lee, Functionalized imidazolium salts for task-specific ionic liquids and their applications, Chem. Commun., 1049–1063 (2006); (e) V. I. Parvulescu and C. Hardacre, Catalysis in ionic liquids, Chem. Rev., 107, 2615–2665 (2007). 5. (a) P. Wasserscheid and T. Welton, Ionic Liquids in Synthesis, Wiley-VCH, Weinheim, 2003; (b) P. J. Dyson and T. J. Geldbach, Metal Catalyzed Reactions in Ionic Liquids, Springer, The Netherlands, 2005. 6. (a) T. Welton, Room-temperature ionic liquids: solvents for synthesis and catalysis, Chem. Rev., 99, 2071–2083 (1999); (b) P. Wasserscheid and W. Keim, Ionic liquids – new solutions for transition metal catalysis, Angew. Chem. Int. Ed., 39, 3772–3789 (2000). 7. Y. Chauvin, L. Mussmann and H. Olivier, A novel class of versatile solvents for two-phase catalysis: hydrogenation, isomerization, and hydroformylation of alkenes catalyzed by rhodium complexes in liquid 1,3-dialkylimidazolium salts, Angew. Chem. Int. Ed., 43, 2698–2700 (1995). 8. (a) C. Baudequin, J. Baudoux, J. Levillain, D. Cahard, A. C. Gaumont and J. C. Plaqevent, Ionic liquids and chirality: opportunities and challenges, Tetrahedron: Asymmetry, 14, 3081–3093

296

9.

10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23.

Recoverable and Recyclable Catalysts (2003); (b) C. E. Song, Enantioselective chem- and bio-catalysis in ionic liquids, Chem. Commun., 1033–1043 (2004); (c) J. Durand, E. Teuma and M. Gomez, Ionic liquids as a medium for enantioselective catalysis, C. R. Chimie., 10, 152–177 (2007); (d) S. V. Malhotra, V. Kumar and V. S. Parmar, Asymmetric catalysis in ionic liquids, Curr. Org. Syn., 4, 370–380 (2007); (e) A. Paczal and A. Kotschy, Asymmetric synthesis in ionic liquids, Monat. Chem., 138, 1115–1123 (2007); (f) R. Sebesta, I. Kmentova and S. Toma, Catalysts with ionic tag and their use in ionic liquids, Green Chem., 10, 484–496 (2008). (a) C. Baudequin, D. Bregeon, J. Levillain, F. Guillen, J. C. Plaquevent and A. C. Gaumont, Chiral ionic liquids, a renewal for the chemistry of chiral solvent? design, synthesis and applications for chiral recognition and asymmetric synthesis, Tetrahedron: Asymmetry, 16, 3921–3945 (2005); (b) J. Ding and D. W. Armstrong, Chiral ionic liquids: synthesis and applications, Chirality, 17, 281–292 (2005); (c) A. Winkel, P. V. G. Reddy and R. Wilhelm, Recent advances in the synthesis and application of chiral ionic liquids, Synthesis, 999–1016 (2008); (d) K. Bica and P. Gaertner, Applications of chiral ionic liquids, Eur. J. Org. Chem., 3235–3250 (2008); (e) M. L. Patil and H. Sasai, Recent developments on chiral ionic liquids: design, synthesis and application, Chem. Record, 8, 98–108 (2008). (a) H.-U. Blaser, C. Malan, B. Pugin, F. Splindler, H. Steiner and M. Studer, Adv. Syn. Catal. 345, 103–151 (2003); (b) W. Tang and X. Zhang, Chem. Rev., 103, 3029–3069 (2003). A. Berger, R. F. De Souza, M. R. Delgado and J. Dupont, Ionic liquid-phase asymmetric catalytic hydrogenation: hydrogen concentration effects on enantioselectivity, Tetrahedron: Asymmetry, 12, 1825–1828 (2000). S. Guernik, A. Wolfson, M. Herskowitz, N. Greenspoon and S. Geresh, A novel system consisting of Rh–DuPHOS and ionic liquid for asymmetric hydrogenations, Chem. Commun., 2314–2315 (2001). B. Pugin, M. Studer, E. Kuesters, G. Sedelmeier and X. Feng, Mixture of ionic liquids and water as a medium for efficient enantioselective hydrogenation and catalyst recycling, Adv. Syn. Catal., 346, 1481–1486 (2004). K. L. Boyle, E. B. Lipsky and C. S. Kalberg, Asymmetric hydrogenation of methyl a-benzamido cinnamate in ionic liquid solvent, Tetrahedron Lett., 47, 1311–1313 (2006). T. Frater, L. Gubicza, A. Szollosy and J. Bakos, Enantioselective hydrogenation in ionic liquids: recyclability of the [Rh(COD)(DIPAMP)[BF4] catalyst in [bmim][BF4], Inorg. Chim. Acta, 359, 2756–2759 (2006). X. Feng, B. Pugin, E. Kusters, G. Sedelmeier and H.-U. Blaser, Josiphos ligands with an imidazolium tag and their application for the enantioselective hydrogenation in ionic liquids, Adv. Syn. Catal., 349, 1803–1807 (2007). M. Schmitkamp, D. Chen, W. Leitner, J. Klankermayer and G. Francio, Enantioselctive catalysis with tropos ligands in chiral ionic liquids, Chem. Commun., 4012–4014 (2007). A. L. Monteiro, F. K. Zinn, R. F. De Souza and J. Dupont, Asymmetric hydrogenation of 2-arylacrylic acids catalyzed by immobilized Ru–BINAP complex in 1-n-butyl-3-methylimidazolium tetrafluoroborate molten salt, Tetrahedron: Asymmetry, 8, 177–179 (1997). R. A. Brown, P. Pollet, E. McKoon, C. A. Eckert, C. L. Liotta and P. G. Jessop, Asymmetric hydrogenation and catalyst recycling using ionic liquid and supercritical carbon dioxide, J. Am. Chem. Soc., 123, 1254–1255 (2001). P. G. Jessop, R. R. Stanley, R. A. Brown, C. A. Eckert, C. L. Liotta, T. T. Ngo and P. Pollet, Neoteric solvents for asymmetric hydrogenation: supercritical fluids, ionic liquids and expanded ionic liquids, Green Chem., 5, 123–128 (2003). H. T. Wong, Y. H. See-Toh, F. C. Ferreira, R. Crook and A. G. Livingston, Organic solvent nanofiltration in asymmetric hydrogenation: enhancement of enantioselectivity and catalyst stability by ionic liquids, Chem. Commun., 2063–2065 (2006). H. L. Ngo, A. Hu and W. Lin, Highly enantioselective catalytic asymmetric hydrogenation of b-ketoesters in room temperature ionic liquids, Chem. Commun., 1912–1913 (2003). A. Hu, H. L. Ngo and W. Lin, Remarkable 4,40 -substituent effects on BINAP: highly enantioselective Ru catalysts for asymmetric hydrogenation of b-aryl ketoesters and their immobilization in room temperature ionic liquids, Angew. Chem. Int. Ed., 43, 2501–2504 (2004).


297

24. Y. Zhu, K. Carpenter, C. C. Bun, S. Bahnmueller, C. P. Ke, V. S. Srid, L. W. Kee and M. F. Hawthorne, (R)-Binap-mediated asymmetric hydrogenation with a rhodacarborane catalyst in ionic liquid media, Angew. Chem. Int. Ed., 42, 3792–3795 (2003). 25. K. Lam, L. Xu, L. Feng, J. Ruan, Q. Fan and A. S. C. Chan, Ruthenium catalyzed asymmetric hydrogenation of a- and b-ketoesters in ionic liquids using chiral P-Phos ligand, Can. J. Chem., 83, 903–908 (2005). 26. M. Berthod, J. M. Joerger, G. Mignani, M. Vaultier and M. Lemaire, Enantioselective catalytic asymmetric hydrogenation of ethyl acetoacetate in room temperature ionic liquids, Tetrahedron: Asymmetry, 15, 2219–2221 (2004). 27. A. Wolfson, I. F. J. Vankelecom and P. A. Jacobs, Co-immobilization of transition-metal complexes and ionic liquids in a polymeric support for liquid-phase hydrogenation, Tetrahedron Lett., 44, 1195–1198 (2003). 28. H. L. Ngo, A. Hu and W. Lin, Catalytic asymmetric hydrogenation of aromatic ketones in room temperature ionic liquids, Tetrahedron Lett., 46, 595–597 (2005). 29. W. Xiong, Q. Lin, H. Ma, H. Zheng, H. Chen and X. Li, Asymmetric hydrogenation of aromatic ketones in ionic liquid media by Ru–TPPTS–(1S,2S)-DPENDS complexes, Tetrahedron: Asymmetry, 16, 1959–1962 (2005). 30. J. Wang, R. Qin, H. Fu, J. Chen, J. Feng, H. Chen and X. Li, Asymmetric hydrogenation of a,b-unsaturated ketones catalyzed by Ru–TPPTS–(S,S)-DPENDS complex in ionic liquids, Tetrahedron: Asymmetry, 18, 847–851 (2007). 31. R. Giernoth and M. S. Krumm, Enantioselective hydrogenation of trimethylindolenine in ionic liquids, Adv. Syn. Catal., 346, 989–992 (2004). 32. M. Solinas, A. Pfaltz, P. G. Cozzi and W. Leitner, Enantioselective hydrogenation of imines in ionic liquid/carbon dioxide media, J. Am. Chem. Soc., 126, 16142–16147 (2004). 33. S. G. Lee, Y. J. Zhang, J. Y. Piao, H. Yoon, C. S. Song, J. H. Choi and J. Hong, Catalytic asymmetric hydrogenation in a room temperature ionic liquid using chiral Rh-complex of ionic liquid grafted 1,4-bisphosphine ligand, Chem. Commun., 2624–2625 (2003). 34. C. Comyns, N. Karodia, S. Zeler and J. Andersen, Clean catalysis with clean solvents – phosphonium tosylates for transfer hydrogeantion reactions, Catal. Lett., 67, 113–115 (2000). 35. T. J. Geldbach and P. J. Dyson, A versatile ruthenium precursor for biphasic catalysis and its application in ionic liquid biphasic transfer hydrogenation: conventional vs task-specific catalysts, J. Am. Chem. Soc., 126, 8114–8115 (2004). 36. I. Kawasaki, K. Tsunoda, T. Tsuji, T. Yamaguchi, H. Shibuta, N. Uchida, M. Yamashita and S. Ohta, A recyclable catalyst for asymmetric transfer hydrogenation with a formic acid–triethylamine mixture in ionic liquid, Chem. Commun., 2134–2136 (2005). 37. W. Zhang, J. L. Loebach, S. R. Wilson and E. N. Jacobsen, Enantioselective epoxidation of unfunctionalized olefins catalyzed by (Salen)manganese complexes, J. Am. Chem. Soc., 112, 2801–2803 (1990). 38. K. Noda, R. Ire, Y. Ito, N. Matsusumoto and T. Katsuki, Catalytic asymmetric epoxidation of unfunctionalized olefins, Tetrahedron Lett., 31, 7345–7348 (1990). 39. C. E. Song and E. J. Roh, Practical method to recycle a chiral (salen)Mn epoxidation catalyst by using an ionic liquid, Chem. Commun., 837–838 (2000). 40. K. Smith, S. Liu and G. A. El-Hiti, Use of ionic liquids as solvents for epoxdiation reactions catalyzed by a chiral Katsuki-type salen complex: enhanced reactivity and recovery of catalyst, Catal. Lett., 98, 95–101 (2004). 41. L. Lou, K. Yu, F. Ding, W. Zhou, X. Peng and S. Liu, An efficient approach for immobilization of chiral Mn(III) salen complexes through a supported ionic liquid phase, Tetrahedron Lett., 47, 6513–6516 (2006). 42. C. E. Song, D. U. Jung, E. J. Roh, S. G. Lee and D. Y. Chi, Osmium tetroxide-(QN)2PHAL in an ionic liquid: a highly efficient and recyclable catalyst system for asymmetric dihydroxylation of olefins, Chem. Commun., 3038–3039 (2002). 43. Q. Liu, Z. Zhang, F. Van Rantwijk and R. A. Sheldon, Osmium-catalyzed asymmetric dihydroxylation of olefins in ionic liquids: the effect of the chiral ligand sturcture on recyclability, J. Mol. Catal. A: Chem., 224, 213–216 (2004).

298


44. L. C. Branco and C. A. M. Afonso, Catalytic asymmetric dihydroxylation of olefins using a recoverable and reusable OsO42 in ionic lqiuid [bmim][PF6], Chem. Commun., 3036–3037 (2002). 45. L. C. Branco and C. A. M. Afonso, Ionic liquids as a convenient new medium for the catalytic asymmetric dihydroxylation of olefins using a recoverable and reusable osmium/ ligand, J. Org. Chem., 69, 4381–4389 (2004). 46. (a) L. C. Branco, A. Serbanovic, M. N. da Ponte and C. A. M. Afonso, Clean osmium-catalyzed asymmetric dihydroxyaltion of olefins in ionic liquids and supericritical CO2 product recovery, Chem. Commun., 107–109 (2005); (b) A. Serbanovic, L. C. Branco, M. N. da Ponte and C. A. M. Afonso, Osmimu catalyzed asymmetric dihydroxyaltion of methyl trans-cinnamate in ionic liquids, followed by supericritical CO2 product recovery, J. Organomet. Chem., 690, 3600–3608 (2005). 47. R. Jiang, Y. Kuang, X. Sun and S. Zhang, An improved catalytic system for recycling OsO4 and chiral ligands in the asymmetric dihydroxyaltion of olefins, Tetrahedron: Asymmetry, 15, 743–746 (2004). 48. L. C. Branco, P. M. P. Gois, N. M. T. Lourenco, V. B. Kurteva and C. A. M. Afonso, Simple transformation of crystalline chiral natural anions to liquid medium and their use to induce chirality, Chem. Commun., 2371–2372 (2006). 49. I. Meracz and T. Oh, Asymmetric Diels–Alder reactions in ionic liquids, Tetrahedron Lett., 44, 6465–6468 (2003). 50. S. Doherty, P. Goodrich, C. Hardacre, H. K. Luo, D. W. Rooney, K. R. Seddon and P. Styring, Marked enantioselectivity enhancements for Diels–Alder reactions in ionic liquids catalyzed by platinum diphosphine complexes, Green Chem., 6, 63–67 (2004). 51. K. Takahashi, H. Nakano and R. Fujita, Reuse of chiral cationic Pd–phosphinooxazolidine catalysts in ionic liquids: highly efficient catalytic asymmetric Diels–Alder reactions, Chem. Commun., 263–265 (2007). 52. F. Fan, Y. C. Teo and T. P. Loh, Catalytic enantioselectivity Diels–Alder reaction in ionic liquid via a recyclable chiral In(III) complex, Org. Lett., 8, 5999–6001 (2006). 53. S. Doherty, P. Goodrich, C. Hardacre, J. G. Knight, M. T. Nguyen, V. I. Parvulescu and C. Paun, Recyclable copper catalysts based on imidazolium-tagged bis(oxazoline): a marked enhancement in rate and enantioselectivity for Diels–Alder reactions in ionic liquid, Adv. Syn. Catal., 349, 951–963 (2007). 54. S. Toma, B. Gotov, I. Kmentova and E. Solecniova, Enantioselective allylic substitution catalyzed by Pd0–ferrocenylphosphine complexes in [bmim][PF6] ionic liquid, Green Chem., 2, 149–151 (2000). 55. I. Kmentova, B. Gotov, E. Solcaniova and S. Toma, Study of ligand and base effect on enantioselective allylation catalyzed by Pd(0) phosphine complexes in [bmim][PF6] ionic liquid, Green Chem., 4, 103–106 (2002). 56. R. Sebesta, M. Meciarova, V. Polackova, E. Veverkova, I. Kmentova, E. Gajdosikova, J. Cvengros, R. Buffa and V. Gajda, Imidazolium tagged ferrocene ligands, Collect. Czech. Chem. Commun., 72, 1057–1068 (2007). 57. S. E. Lyubimov, V. A. Davankov and K. N. Gavrilov, The use of an ionic liquid in asymmetric catalytic allylic amination, Tetrahedron Lett., 47, 2721–2723 (2006). 58. S. Doherty, P. Goodrich, C. Hardacre, H. K. Luo, M. Nieuwenhuyzen and R. K. Rath, Lewis acid platinum complexes of conformationally flexible NUPHOS diphosphines: highly efficient catalysts for the carbonyl–ene reaction, Organometallics, 24, 5945–5955 (2005). 59. H. K. Luo, L. B. Khim, H. Schumann, C. Lim, T. X. Jie and H. Yang, Enantioselective carbonyl–ene reactions of arylglyoxals with a chiral palladium(II)–BINAP catalyst, Adv. Syn. Catal., 349, 1781–1795 (2007). 60. Y. Oonishi, J. Ogura and Y. Sato, Rh(I)-catalyzed intramolecular hydroacylation in ionic liquids, Tetrahedron Lett., 48, 7505–7507 (2007). 61. S. V. Malhotra and Y. Xiao, Enantioselective Friedel–Crafts reactions of aromatic amines with ethyl glyoxylate in pyridinium-based ionic liquids, Aust. J. Chem., 59, 468–472 (2006). 62. S. Doherty, P. Goodrich, C. Hardacre, V. Parvulescu and C. Paun, Efficient heterogeneous asymmetric catalysis of the Mukaiyama aldol reaction by silica- and ionic liquid-


63. 64. 65. 66. 67. 68. 69. 70. 71. 72. 73. 74. 75. 76. 77. 78. 79. 80. 81. 82.

299

supported Lewis acid copper (II) complexes of bis(oxazolines), Adv. Syn. Catal., 350, 295–302 (2008). J. M. Fraile, J. I. Garcia, C. I. Herrerias, J. A. Mayoral, D. Carrie and M. Vaultier, Enantioselective cyclopropanation reactions in ionic liquids, Tetrahedron: Asymmetry, 12, 1891–1894 (2001). J. M. Fraile, J. I. Garcia, C. I. Herrerias, J. A. Mayoral, S. Gmough and M. Vaultier, Comparison of the immobilization of chiral bis(oxazoline)–copper complexes onto anionic solids and in ionic liquids, Green Chem., 6, 93–98 (2004). J. M. Fraile, J. I. Garcia, C. I. Herrerias, J. A. Mayoral, O. Reiser and M. Vaultier, The importance of complex stability for asymmetric copper-catalyzed cyclopropanations in [emim][OTf] ionic liquid: the bis(oxazoline)-azabis(oxazoline) case, Tetrahedron Lett., 45, 6765–6768 (2004). M. R. Castillo, L. Fousse, J. M. Fraile, J. I. Garcia and J. A. Mayoral, Supported ionic-liquid films (SILF) as two-dimensional nanoreactors for enantioselective reactions: surface-mediated selectivity modulation (SMSM), Chem. Eur. J., 13, 287–291 (2007). D. L. Davies, S. K. Kandola and R. K. Patel, Asymmetric cyclopropanation in ionic liquids: effect of anion and impurities, Tetrahedron: Asymmetry, 15, 77–80 (2004). J. Lu, S. J. Ji and T. P. Loh, Enantioselective allylation of aldehydes catalyzed by chiral indium (III) complexes immobilized in ionic liquids, Chem. Commun., 2345–2347 (2005). Y. C. Teo, E. L. Goh and T. P. Loh, Catalytic enantioselective allylation of aldehydes via a chiral indium(III) complex in ionic liquids, Tetrahedron Lett., 46, 4573–4575 (2005). J. Lu, S. Ji, Y. Teo and T. P. Loh, Asymmetric allyltributylstannane addition to ketones catalyzed by chiral PYBOX–In(III) complex immobilized in ionic liquid, Tetrahedron Lett., 46, 7435–7437 (2005). Y. Hamashima, H. Takano, D. Hotta and M. Sodeoka, Immobilization and reuse of Pd complexes in ionic liquids: efficient catalytic asymmetric fluorination and Michael reactions with b-ketoesters, Org. Lett., 5, 3225–3228 (2003). C. E. Song, C. R. Oh, E. J. Roh and D. J. Choo, Cr(salen) catalysed asymmetric ring opening reactions of epoxides in room temperature ionic liquids, Chem. Commun., 1743–1744 (2000). C. R. Oh, D. J. Choo, W. H. Shim, D. H. Lee, E. J. Roh, S. Lee and C. E. Song, Chiral Co(III) (salen)-catalyzed hydrolytic kinetic resolution of racemic epoxides in ionic liquids, Chem. Commun., 1100–1101 (2003). C. Baleizao, B. Gignate, H. Garcia and A. Corma, Ionic liquids as green solvents for the asymmetric synthesis of cyanohydrins catalysed by VO(salen) complexes, Green Chem., 4, 272–274 (2002). S. M. Kim, Y. K. Kang, K. S. Lee, J. Y. Mang and D. Y. Kim, Asymmetric electrophilic fluorination of b-keto phosphonates in ionic liquid media catalyzed by chiral palladium complexes, Bull. Korean Chem. Soc., 27, 423–425 (2006). J. N. Rosa, A. G. Santos and C. A. M. Afonso, Enantioselective addition of alkynes to imines in ionic liquids, J. Mol. Catal. A: Chem., 214, 161–165 (2004). A. Bosmann, G. Francioa, E. Janssen, M. Solinas, W. Leitner and Peter Wasserscheid, Activation, tuning and immobilization of homogenous catalysts in an ionic liquid/compressed CO2 continuous-flow system, Angew. Chem. Int. Ed., 40, 2697–2699 (2001). M. L. Kantam, B. Kavita, V. Neeraja, Y. Haritha, M. K. Chaudhuri and S. K. Dehury, Heterogeneous catalytic sulfimidation using immobilized Cu(acac)2, Adv. Syn. Catal., 347, 641–645 (2005). T. P. Loh, L. Feng, H. Yang and J. Yang, L-Proline in an ionic liquid as an efficient and reusable catalyst for direct asymmetric aldol reactions, Tetrahedron Lett., 43, 8741–8743 (2002). P. Kotrusz, I. Kmentova, B. Gotov, S. Toma and E. Solcaniova, Proline-catalysed asymmetric aldol reaction in the room temperature ionic liquid [bmim][PF6], Chem. Commun., 2510–2511 (2002). J. Shah, H. Blumenthal, Z. Yacob and J. Liebscher, Proline-catalyzed asymmetric aldol reaction in guanidine-derived ionic liquids, Adv. Syn. Catal., 350, 1267–1270 (2008). H. Guo, L. Cun, L. Gong, A. Mi, and Y. Jiang, Asymmetric direct aldol reaction catalyzed by an L-prolinamide derivative: considerable improvement of the catalytic efficiency in the ionic liquid, Chem. Commun., 1450–1452 (2005).

300


83. A. Cordova, Direct catalytic asymmetric cross-aldol reactions in ionic liquid media, Tetrahedron Lett., 45, 3949–3952 (2004). 84. M. Lambardo, F. Pasi, S. Easwar and C. Trombini, An improved protocol for the direct asymmetric aldol reaction in ionic liquids, catalysed by onium ion-tagged prolines, Adv. Syn. Catal., 349, 2061–2065 (2007). 85. L. Zhou and L. Wang, Chiral ionic liquid containing L-proline unit as a highly efficient and recyclable asymmetric organocatalyst for aldol reaction, Chem. Lett., 36, 628–629 (2007). 86. M. Gruttadauria, S. Riela, P. L. Meo, F. D’Anna and R. Noto, Supported ionic liquid asymmetric catalysis. A new method for chiral catalysts recycling. The case of proline-catalyzed aldol reaction, Tetrahedron Lett., 45, 6113–6116 (2004). 87. M. Gruttadauria, S. Riela, C. Aprile, P. L. Meo, F. D’Anna and R. Noto, Supported ionic liquids. New recyclable materials for the L-proline-catalyzed aldol reaction, Adv. Syn. Catal., 348, 82–92 (2006). 88. C. Aprile, F. Giacalone, M. Gruttadauria, A. M. Marculescu, R. Noto, J. D. Revell and H. Wennemers, New ionic liquid-modified silica gels as recyclable materials for L-proline or H-Pro-Pro-Asp-NH2-catalyzed aldol reaction, Green Chem., 9, 1328–1334 (2007). 89. R. T. Dere, R. R. Pal, P. S. Patil and M. M. Salunkhe, Influence of ionic liquids on the phase transfer-catalyzed enantioselective Michael reaction, Tetrahedron Lett., 44, 5351–5353 (2003). 90. M. S. Rasalkar, M. K. Potdar, S. S. Mohile and M. M. Salunkhe, An ionic liquid influenced L-proline catalyzed asymmetric Michael addition of ketones to nitrostyrene, J. Mol. Catal. A: Chem., 235, 267–270 (2005). 91. P. Kotrusz, S. Toma, H.-G. Schmaltz and A. Adler, Michael addition of aldehydes and ketones to b-nitrostyrenes in an ionic liquid, Eur. J. Org. Chem., 1577–1583 (2004). 92. H. Hagiwara, T. Okabe, T. Hoshi and T. Suzuki, Catalytic asymmetric 1,4-conjugate addition of unmodified aldehyde in ionic liquid, J. Mol. Catal. A: Chem., 214, 167–174 (2004). 93. D. Xu, B. Wang, S. Luo, H. Yue, L. Wang and Z. Xu, Pyrrolidine–pyridinium based organocatalysts for highly enantioselective Michael addition of cyclohexanone to nitroalkenes, Tetrahedron: Asymmetry, 18, 1788–1794 (2007). 94. V. Gallo, D. Giardina-Papa, P. Mastrorilli, C. F. Nobile, G. P. Suranna and Y. Wang, Asymmetric Michael addition promoted by (R,R)-trans-1,2-diaminocyclohexane in ionic liquids, J. Organomet. Chem., 690, 3535–3539 (2005). 95. P. Kotrusz, S. Alemayehu, S. Toma, H.-G. Schmatz and A. Adler, Enantioselective organocatalysis in ionic liquids: addition of aliphatic aldehydes and ketones to diethyl azodicarboxylate, Eur. J. Org. Chem., 4904–4911 (2005). 96. H. Guo, H. Niu, M. Xue, Q. Guo, L. Cun, A. Mi, Y. Jiang and J. Wang, L-proline in an ionic liquid as an efficient and reusable catalyst for direct asymmetric a-aminoxylation of aldehydes and ketones, Green Chem., 8, 682–684 (2006). 97. K. Huang, Z. Huang and X. Li, Highly enantioselective a-aminooxylation of aldehydes and ketones in ionic liquids, J. Org. Chem., 71, 8320–8323 (2006). 98. J. K. Park, P. Sreekanth and B. M. Kim, Recycling chiral imidazolidin-4-one catalyst for asymmetric Diels–Alder reactions: screeing of various ionic liquids, Adv. Syn. Catal., 346, 49–52 (2004). 99. (a) N. S. Chowdari, D. B. Ramachary and C. F. Barbas III, Organocatalysis in ionic liquids: highly efficient L-proline-catalyzed direct asymmetric Mannich reactions involving ketone and aldehyde nucleophiles, Synlett, 1906–1909 (2003); (b) W. Notz, S. I. Watanabe, N. S. Chowdari, G. Zhong, J. M. Betancort, F. Tanaka and C. F. Barbas III, The scope of the direct prolinecatalyzed asymmetric addition of ketones to imines, Adv. Syn. Catal., 346, 1131–1140 (2004). 100. B. Pegot, G. Vo-Thanh, D. Gori and A. Loupy, First application of chiral ionic liquids in asymmetric Baylis–Hillman reaction, Tetrahedron Lett., 45, 6425–6428 (2004). 101. R. Gausepohl, P. Buskens, J. Kleinen, A. Bruckmann, C. W. Lehmann, J. Klankermayer and W. Leitner, Highly enantioselective aza-Baylis–Hillman reaction in a chiral reaction medium, Angew. Chem. Int. Ed., 45, 3689–3692 (2006).

11 Recoverable Organic Catalysts Maurizio Benaglia Dipartimento di Chimica Organica e Industriale, Universit a degli Studi di Milano, Milano, Italy

11.1 Introduction The last 10 years have witnessed the explosion of the so-called ‘organocatalysis’.1 The possibility of using catalytic amounts of an organic compound of relatively low molecular weight and simple structure to promote reactions that previously required costly and possibly toxic transition metal-based catalysts, can be regarded as a significant step toward the development of a truly green chemistry.2 The organocatalytic approach fulfils many of requirements listed in the well-known 12 principles of green chemistry.3 A process based on a catalytic methodology is already ‘green’ by definition, since it is clearly stated that ‘catalysts are preferable to stoichiometric reagents’ because they minimize waste and increase energy efficiency; a catalyst often allows a reaction to be run under milder experimental conditions, once again improving the efficiency of a process from the economic and energetic point of view. Specifically organocatalysts may lead to the design of ‘safer chemicals and products’, as expected by modern synthetic chemists, also with the goal to use less hazardous solvents or reaction conditions. In this context the concept of ‘organocatalysis’ is highly attractive because it contains the ‘magic’ word of catalysis, which is one of the key features towards the development of modern and truly efficient chemical processes. Furthermore, the replacement of metalbased catalysts with equally efficient metal-free counterparts is very appealing in view of possible applications in the future of nontoxic, low cost, and more environmentally friendly promoters on an industrial scale with obvious advantages from the environmental and


302


economic points of view.4 The catalyst usually allows mild reaction conditions to be used; also the economic benefits of an efficient catalytic process are enormous since it is less capital intensive, has lower operating costs, produces higher purity products and fewer byproducts. In addition, a substoichiometric process provides important environmental benefits, since a really efficient technology minimizes the waste disposal problems that may represent a severe issue for large scale productions. In this framework chiral organocatalysts present other fundamental characteristics of potential enormous interest and, if possible, even more appealing for both academic research and specially for industrial scale application. Global sales of single-enantiomer compounds are estimated to have been of 9.5 billion dollars in 2005 and to reach 15 billion dollars by the end of 2009. Among the different possible available methodologies to synthesize enantiomerically enriched compounds the use of a chiral catalyst represents in principle the most attractive procedure, since in a catalytic process a small amount of a ‘smart’ molecule produces a large quantity of the desired chiral compound.5 Therefore there is an always increasing interest towards the discovery, the development and the large scale application of novel and efficient enantioselective catalytic methodologies. Demand for enantiomerically pure compounds is continuosly increasing, not only for use in pharmaceuticals but also in other fields such as agrochemicals, flavour and aroma chemicals, and speciality materials. Recent strict government regulations that require the individual evaluation of all the possible stereoisomers of a compound and the commercialization of a chiral product only as single enantiomer have called for further improvements in the stereoselective synthesis of chiral compounds. In this context it is surprising how relatively few enantioselective catalytic reactions are used on an industrial scale today.6 This is even more difficult to understand if one thinks of the impressive progress in the last few years in the field of enantioselective catalysts, where hundreds of catalytic transformations of great chemical and stereochemical efficiency have been developed. However, of the 7 billion dollars chiral products market realized in 2002, 55% of the enantiomerically pure (or enriched) target molecules were obtained through traditional technologies, such as resolution of the racemic mixture or synthesis starting from a chiral pool; only 35% of the desired chiral compounds were prepared by chemical catalysis and 10% by biocatalysis. It is true that a study has predicted things will be different in the future:7 the share of the market produced by traditional technologies would drop to 41% and catalysis will increase to 36% and biocatalysis to 22%, but the change is not really that impressive. It is even more difficult to fully understand the situation if one considers that the number of papers per year related to ‘chiral technologies’ has tripled over 1300 in 1994 to more than 4400 in 2003, and more than 70% of the works deal with stereoselective synthesis. So the obvious question is: why has the application of enantioselective catalysis to the fine chemicals industry, of potentially great economic and environmental interest, not been widely pursued on large scale? It is true that different issues must be addressed: first of all the cost of the chiral catalyst, but also other problems must be considered, such as general applicability; many of the very selective catalysts have been developed for reactions with selected model substrates but not tested on differently functionalized molecules. In addition, for many catalysts little information is available on catalyst selectivity, activity, productivity. The stability of the catalyst and the possibility of an easy separation and maybe recycling are also important aspects to be considered for an industrial asymmetric catalytic process.

Recoverable Organic Catalysts

303

The immobilization of the catalytic species on a solid support may represent a solution to some of the problems;8 not only the recovery and the possible recycling of a catalyst may be investigated and successfully realized through its immobilization, but also the studies of other issues like the stability, the structural characterization and the catalytic behaviour may be better conducted on a supported version of the enantioselective catalyst. These general considerations are true also for organic catalysts. Even if the transformation of a stoichiometric into a catalytic process can be regarded as a significant step towards the development of a truly green chemistry, catalytic reactions are amenable to a variety of improvements that can make them greener and greener. Among these, the replacement of metal-based catalysts with equally efficient metal-free counterparts, the so-called ‘organic catalysts’, can be extremely important. A number of chemically robust organocatalysts capable of displaying enzyme-like activity has thus become available, their application encompassing many fundamental reactions of organic chemistry that can now be run in the presence of nontoxic, cheap, and more environmentally friendly promoters. In this context, the term ‘organic’ is synonymous with ‘metal-free’ with all the advantages of performing a reaction under metal-free conditions. These advantages might include, inter alia, the possibility of working in wet solvents and under an aerobic atmosphere, dealing with a stable and robust catalyst and avoiding the problem of a (possibly) expensive and toxic metal leaching into the organic product. The immobilization of the catalyst on a support with the aim of facilitating the separation of the product from the catalyst, and thus the recovery and recycling of the latter, can also be regarded as an important improvement for a catalytic process.9 In this respect, the immobilization of organic catalysts seems particularly attractive, because the metal-free nature of these compounds avoids from the outset the problem of the leaching of the metal, which often negatively affects and practically prevents the efficient recycling of a supported organometallic catalyst.10 Furthermore a simple organic compound will be less affected by the connection to a support than more structurally complex (and somehow more ‘delicate’) enzymes, from which they are conceptually derived and with which they are often compared.11 In the very hectic area of organocatalysis the immobilization of chiral catalytic species may play (and in part it has already played) a decisive role in the further development of the field. In addition, it is important to note how the immobilization on the support can endow the catalyst with special properties (for instance: a different solubility profile or an enhanced catalytic activity) that can be fine-tuned by a careful selection of the support thus expanding the range of application of the catalyst. Based on recently published reviews covering the immobilization of achiral and chiral organic catalysts,10 the present chapter will cover the more relevant achievements reported in the field of immobilized organic catalysts, focusing on studies that appeared in the last few years, especially after 2003. Whenever possible, comparison between the behaviours of supported and nonsupported catalytic species will be discussed. Since ionic liquid-based technologies and recoverable fluorous systems, as well as thermomorphic catalysts are all topics discussed in other chapters in this book, only the more relevant cases of recoverable organic catalysts which make use of those methodologies will be discussed. The chapter does not intend to fully cover all the works in the field of supported organic catalysts, but to present the most representative examples of different classes of catalysts in order to discuss and eventually to compare different technologies and methodologies employed to develop recoverable and recyclable organocatalysts.

304


Particular attention will also be devoted to catalyst recovery and recycling. The discussion on the structure of the polymeric support will be limited to those examples for which its influence on the catalyst performance has clearly been demonstrated. After the first section dedicated to achiral catalysts, chiral organic catalysts will be discussed, with a separate section dedicated to amino acid-derived organocatalysts, which play a unique role in stereoselective metal-free catalysis. Finally a few considerations on the methodologies, the future and the problems related to chiral organic catalyst immobilization will be also briefly presented.

11.2 Achiral Organic Catalysts The main goal of the immobilization of a catalyst is the recovery and hopefully the recycling of the catalytic species, which in case of chiral species may be particular expensive. However, besides the recovery and the reuse of the catalyst, there may be other reasons to produce a supported version of a catalytic species. A fundamental feature of an immobilized organic catalyst that make its use extremely attractive is the possibility to simplify the reaction work-up. This extremely important aspect, specially from the large scale application point of view, was immediately recognized by the chemists of the 1970s and 1980s. However, if it is true that the development of immobilized versions of an organic catalyst has occurred shortly after the discovery of the catalyst itself, it is also safe to state that in the last 15 years the advent of combinatorial chemistry has given a decisive impulse to the discovery and implementation of new synthetic methods based on supported reagents and catalysts.12 11.2.1 Oxidation Catalysts Many of the organic catalysts identified in recent years have also been immobilized on different supports. Among these, oxoammonium ions employed for the selective oxidation of alcohols, and ketones for the epoxidation of alkenes occupy a preeminent position. A typical example of how a supported version of a catalyst may allow a great simplification of the reaction work-up is represented by the reactions promoted by TEMPO.13 The use of oxoammonium ions such as those derived from 2,2,6,6-tetramethyl-piperidine-1oxyl (TEMPO) in combination with inexpensive, safe, and easy-to-handle terminal oxidants in the conversion of alcohols into aldehydes, ketones, and carboxylic acids is a significant example of how it is possible to develop a safer and greener chemistry, by avoiding the use of environmentally unfriendly or toxic metals. However, separation of the products from TEMPO can be problematic, especially when the reactions are run on a large scale, and the immobilization on a solid support may offer a solution to this problem. The immobilization of the catalysts onto a solid support may represent a solution to the problem.10b The development of TEMPO or TEMPO analogues anchored on the soluble support Chimassorb 944 (average MW 3000 Da) was actively investigated with the aim of replicating as much as possible the remarkable features of the nonsupported catalytic system.14 Recently, examples of poly(ethylene glycol)-supported TEMPO have been reported. Soluble polymers have been a subject of an intense research activity; allowing the reaction


305

to be carried out in homogeneous solution would secure higher chemical and stereochemical efficiency than with an insoluble polymer. Among the soluble polymeric matrixes employed, poly(ethylene glycol)s (PEGs) are the most successful.15 These polymers with Mw greater than 2000 Da are readily functionalized, commercially available, inexpensive supports that feature convenient solubility properties being soluble in many common organic solvents and insoluble in a few other solvents, such as diethyl ether, hexanes, t-butyl-methyl ether. Therefore, the choice of appropriate solvent systems makes it possible to run a reaction under homogeneous catalysis conditions (where the PEG-supported catalyst is expected to perform at its best) and then to recover the catalyst under heterogeneous conditions, as if it were bound to an insoluble matrix. A PEG-supported TEMPO 1, soluble in organic solvents such as CH2Cl2 and acetic acid, but insoluble in ethers and hexanes was prepared and proved to be an effective catalyst for the selective oxidation of 1-octanol with various stoichiometric oxidants. When 1 was employed in 1 mol% as a catalyst in combination with KBr (10 mol%) a slight excess of buffered bleach (pH ¼ 8.6) as the terminal oxidant, partial overoxidation of 1-octanol to octanoic acid was observed (91% yield in octanal). This can be avoided either by using a stoichiometric amount of NaOCl or, more conveniently, by working under bromide-free conditions (Scheme 11.1, equation a). Although slightly decreased, the oxidation rate remains high even in the absence of KBr and the aldehyde was obtained in 95% yield after only 30 min of reaction.16 The PEG-supported TEMPO proved to be very efficient in the oxidation of 1-octanol to octanal not only with sodium hypochlorite, but also in combination with different terminal oxidants such as bis(acetoxy)iodobenzene and trichloroisocyanuric acid. This reaction could be extended to acyclic and cyclic primary and secondary alcohols with excellent results. It also was remarkable that the PEG-supported TEMPO maintained the good selectivity for primary vs secondary benzylic alcohol oxidation typical of nonsupported TEMPO. Catalyst recovery exploited the well-known different solubility of PEG in solvents of different polarity. In this case, where a PEG of relatively high MW was employed, addition of diethylether to the reaction mixture induced the precipitation of compound 1. Subsequent filtration allowed recovery of the supported reagent with less than 10% weight loss for each recovery. Catalyst recycling was demonstrated to be possible for seven reaction cycles in the oxidation of 1-octanol, which occurred with undiminished conversion and selectivity. The very positive performance of PEG-anchored TEMPO in water solution with a very cheap and safe oxidant agent is a clear example of how supported catalysts may be developed to be used in environmentally friendly or green solvents, as part of the drive towards developing more green chemistry. The beneficial effect of the spacer was also demonstrated in a subsequent work, where spacer-containing and spacer-lacking pre-catalysts (1–5 mol%) were employed for the oxidation of primary and secondary alcohols using oxygen as the terminal oxidant in the presence of 2 mol% of Co(NO3)2 hydrate and Mn(NO3)2 hydrate as co-catalyst.17 For instance, under these conditions oxidation of 1-octanol and cyclooctanol occurred in quantitative yield with 1 and in only about 65% yield with 2 (Scheme 11.1). Recycling of pre-catalyst 1 was demonstrated for the oxidation of 4-bromobenzyl alcohol for six reaction cycles occurring with slowly decreasing yields (>99% first cycle and 74% sixth cycle) (Scheme 11.1, equation b).

306


LINKER

N O

2

N O

1

= MeO-(CH 2CH2O)n-CH2CH2O .

LINKER =

Mw 5000 daltons

O

CH2O

O

O

1% mol cat 1 OH

Equation a

H NaOCl O 2% mol Co(NO3)2

OH

Equation b Br

H

1% mol cat 1 / O2

Br

N N N HO

Equation c

N O

O

N O

O

= PS

O

Equation d

O

3

O

O

H

n

n

N3

n

PVA-PEG

4

= PVA resin

Co O

N O

N3

O

N O

N N N O N

Co

N N N

Co n

Equation e

N O

= C/Co nanoparticles

O

5

Scheme 11.1 Supported achiral oxidation catalysts

Among several other supports investigated, a recent work where TEMPO was immobilized onto polystyrene through a copper (I)-catalysed azide-cycloaddition deserves special mention.18 A really high loading (more than 4 mmol/g) of PS-TEMPO was readily obtained by exploiting the so-called click-chemistry19 which allowed the supported


307

catalyst to be prepared in a sequence of only two steps (Scheme 11.1, equation c). The PSgrafted TEMPO 3 was employed in the oxidation of primary and secondary alcohols, showing a good chemical activity, using either sodium hypochlorite or oxygen as the terminal oxidant in the presence of catalytic amounts of Co(NO3)2 hydrate and Mn(NO3)2. The so-called PS-CLICK-TEMPO catalyst was shown to have a good recyclability; after a simple filtration the supported catalyst was reused up to five times with no appreciable loss of activity. Click chemistry was employed in another very recent contribution where new poly (vinyl alcohol)-graft-poly (ethylene glycol) resins (PVA-g-PEG) were prepared by anionic polymerization onto PVA beads of ethylene oxide.20 The new materials, having various PEG chain lengths, showed good swelling properties in both water and organic solvents; as demonstration of the remarkable characteristics of the new supports a polymer-anchored azide was reacted with a properly functionalized TEMPO derivative bearing a terminal alkyne (Scheme 11.1, equation d). The supported TEMPO 4 was successfully tested in the oxidation mediated by sodium hypochlorite and in the recycling study, after 12 runs, the time for the complete oxidation of benzyl alcohol to benzaldehyde increased from 5 to 18 min, indicating a reproducible catalytic performance of the immobilized species. Interestingly TEMPO was also immobilized onto magnetic nanoparticles, a support for catalysts that is receiving an increasing attention for the development of recyclable systems.21 Magnetic nanobeads have the very attractive feature of being readily separated from the reaction medium and allowing an easy recycling of the catalyst by magnetic separation. In particular it was demonstrated that magnetic C/Co-nanoparticles may be used as a support for the immobilization of a suitably functionalized TEMPO derivative (Scheme 11.1, equation e). Once again the cycloaddition between an alkyne and a supported azide was employed to immobilize a catalyst 5 that was successfully used in the oxidation of alcohols promoted by sodium hypochlorite.22 The separation of the magnetic beadanchored catalyst was performed in a few seconds by simply applying an external magnet and allowed a basically quantitative recovery of the supported TEMPO that was recycled six times with virtually no change in chemical efficiency.23 11.2.2 Phase Transfer Catalysts The use of supported phase transfer catalysts also offers the possibility of easier isolation of the reaction product, often without the necessity of any chromatographic purification. The multifaceted applications of phase transfer catalysis in organic synthesis decisively contributed to the establishment of organic catalysts as useful preparative tools.10 Polymer-supported phase transfer catalysis has been extensively examined but it was noted that the catalytic activity of the insoluble polystyrene-supported catalysts was strongly reduced in comparison with that of their nonsupported soluble counterparts. Since the 1960s soluble polymeric supports have been envisaged as possible alternatives to their insoluble counterparts for catalyst immobilization (Scheme 11.2). A quaternary ammonium salt was easily synthesized on a modified MeOPEG, and this supported catalyst was shown to be an efficient and recoverable promoter of several reactions carried out under phase transfer catalysis (PTC) conditions. Catalyst 6 showed a catalytic activity that was similar to, or even better than that of the nonsupported catalysts.24

308

Recoverable and Recyclable Catalysts N+Bu3 BrMeO

O

(CH3)3 O

= -(CH2CH2O)n-CH2CH2

-Br

Bu3+N

n = 100-125

N+Bu3 Br-

Bu3+N O

-Br

6

O

7 N+Bu3 Br-

= -CH2CH2-(O-CH2CH2)n- n = 104

Scheme 11.2 Polymer-supported phase transfer catalysts

Remarkably, the benzylation of phenol and pyrrole required only 0.01 equiv. of catalyst 5 to occur in 95% yield. Dichloromethane, in which the catalyst is readily soluble, was found to be the organic solvent of choice, but the reaction could also satisfactorily be carried out in the absence of organic solvent. Generally the use of solid/liquid conditions led to higher yields than those observed under liquid/liquid conditions. The PEG-supported ammonium salt could be recovered by precipitation and filtration, and recycled three times to run the same or even a different reaction without any appreciable loss of the catalytic activity. It is also worth mentioning that compound 6 favourably compares as a catalyst with other quaternary ammonium salts immobilized on insoluble polystyrene supports. The use of these catalysts generally required higher reaction temperatures and/or longer reaction times than those employed here. In addition, solid-supported phase transfer catalysts required a preliminary, long conditioning time (up to 15 h) to ensure bead swelling and optimum accessibility of substrate and reagent to the catalytic site. Finally, the high stirring rate necessary with these catalysts resulted in extensive mechanical degradation of the polymer beads, which were difficult to recover by filtration. It seems possible that the catalyst benefits from the involvement of the support, because the polyethylenoxy chain of PEG can complex the alkaline cation of the base helping the transfer of the HO counterion in the organic phase. In order to increase the number and to ensure a suitable spatial arrangement of the catalytic sites the loading expansion of PEG was carried out by exploiting the principles of dendrimer chemistry and led to the synthesis of the PEG-supported tetrakis ammonium


309

salt 7.25 This catalyst displayed a higher catalytic efficiency than 6, while retaining the solubility properties peculiar of the PEG support, which allowed simple catalyst recovery and recycling by precipitation and filtration (Scheme 11.4). The catalyst, which featured four quaternary ammonium groups located at the termini of the polymer backbone, displayed a remarkable efficiency in promoting different reactions carried out under PTC conditions at low catalyst concentrations (1–3 mol%) for short reaction times and under mild conditions. The polymer support allowed the catalyst to be easily recovered and recycled. 11.2.3 Miscellaneous Catalysts Catalyst instability can be another problem that may be tackled by developing an immobilized catalyst. Organic catalysts do exist that slowly decompose under the conditions necessary for their reaction and release trace amounts of byproducts that must be separated from the products. For instance in photo-oxygenation reactions catalysed by porphyrin the release of highly coloured materials derived from the photosensitizer made the product purification a real problem. Immobilization of the catalyst can solve this problem because also the decomposed material are supported and can be removed from the reaction medium in the work-up process. Starting from the commercially available monomethylether of PEG (MeOPEG) the commercially available, 5,10,15,20-tetrakis-(4-hydroxyphenyl)-porphyrin was immobilized to give the PEG-supported porphyrin 8 (MW ¼ 2000, Scheme 11.3). The irradiation of a 0.01 M methylene chloride solution of bisdialine 9 with a 100 W halogen lamp, in the presence of 3 mol% of poly(ethylene glycol)-supported tetrahydroxyphenyl-porphyrin (PEG-TPP, 7) as sensitizer gave a 82/18 mixture of supra and antara diastereoisomeric endoperoxides 10 in quantitative yield after 1 h (Scheme 11.3, equation a). The polymer-bound catalyst not only showed the same activity of the nonsupported species, but it greatly simplified the product isolation. At the end of reaction the solvent was concentrated in vacuum and diethyl ether was added to the PEG-supported porphyrin that was quantitatively recovered by filtration. From the concentrated filtrate solution the obtained endoperoxides were easily isolated by crystallization from ethanol.26 The one-pot oxidation of olefin to a,b-unsaturated ketones was applied on a gram scale to convert dicyclopentadiene into the corresponding dicyclopentadienone, a useful starting-material for the preparation of enantiopure diols; after filtration of the supported catalyst the product was isolated by simple loss of chemical or stereochemical efficiency. The design of immobilized catalysts may be seen as the attempt to develop a new ‘green’ evaporation of the organic solvent as an analytically pure compound, which did not require further purification (Scheme 11.3, equation b). It is noteworthy that a PEG-supported sensitizer was recycled six times with no appreciable protocols, to avoid the use of toxic and costly reagents or solvents. In this framework it was recently reported that polystyrenesupported BEMP (2-t-butylimino-2-ethylamino-1,3-dimethylperhydro-1,3,2-diazaphosphorine) 11 was shown to be an efficient catalyst able to promote the nitroalkane addition to a,bunsaturated carbonyl compounds27 (Scheme 11.3, equation c). The supported catalytic system worked efficiently in solvent-free conditions, but its recycling was hampered by the pulverization of the polymer beads due to high-speed magnetic stirring.

310

Recoverable and Recyclable Catalysts PEG

O

O N H

N

H N PEG

PEG

PEG

N

O

= MeO-(CH2CH2O)n-CH2CH2-; Mw 2000

PEG

O

8

Equation

O2, hν Sens. (cat.)

a

O O

CH2Cl2

Ac2O, Py, O2, hν

b

Sens. (cat.), DMAP (cat.) CH2Cl2

11

O c

R

O

N N P N N

11

Equation

10b antara

10a supra

9

Equation

O O

+

R'

Fe3O4

O2N R

CH3NO2

Si

N

N

O R'

12

SiO2

Scheme 11.3 Miscellaneous supported achiral catalysts

To overcome these difficulties a solvent-free cyclic continuous-flow reactor was developed, making the recovery and the reuse of the catalyst possible, and the isolation of the product feasible with a very small amount of organic solvent. Another attractive methodology alternative to the traditional techniques involving filtration and recovery of the catalyst is represented by the use of silica-coated magnetic particles. It was recently shown that the technology may also find efficient application in the case of organic catalysts.28 Magnetite nanoparticles were coated with a silica layer by


311

dispersion in aqueous quaternary ammonium hydroxide and then exposed to sodium silicate. Reaction of the silica-coated material with triethoxy silane bearing the N,N0 dialkylaminopyridine moiety in THF at refluxing temperature followed by a capping step with n-propyltrimethoxysilane afforded the immobilized catalyst 12 (Scheme 11.3). The loading was established to be 0.2 mmol/g by proton NMR and constantly monitored by the NMR technique. The supported catalyst showed high performances in different reactions, including esterification, peracetylation of glucose, BOC protection of indole and also proved to be extremely efficient with sterically hindered acylating agents. The separation of the catalyst was accomplished by simply applying an external magnet, followed by decantation of the reaction solution. No catalyst degradation or loss of chemical efficiency was observed after 30 consecutive catalytic cycles.

11.3 Chiral Organic Catalysts Besides the simplification of the reaction work-up, the recovery and hopefully the recycling of the precious chiral catalyst represent the obvious, more important aspects of the immobilization process of an enantiomerically pure organocatalyst. So far, in the field of organocatalysis research efforts have mainly focused on catalyst discovery; therefore the development of immobilized chiral catalysts has been less intensively pursued. Immobilization is obviously convenient if the catalyst is expensive, or has been obtained after a complex synthesis, or is employed in a relatively large amount. It is worth mentioning, however, that the synthesis of the supported catalyst should exploit a starting material comparable in cost and synthetic complexity with those of the compound used for the synthesis of the nonsupported catalyst. This is the case, for example, of the PEGimmobilized imidazolidinone (see Section 11.4.2); it was obtained from tyrosine (D 42 for 100 g), a starting material more convenient than phenylalanine (D 83 for 100 g) used for the synthesis of the nonsupported catalyst. Also in the present section dedicated to supported chiral organocatalysts special attention will be given to the more recent developments in the field, published after 2003. 11.3.1 Phase Transfer Catalysts Chiral phase transfer catalysis is a very interesting methodology that typically requires simple experimental operations, mild reaction conditions, inexpensive, environmentally benign reagents and is amenable to large-scale preparations.29 The advent in the early 1990s of the O’Donnell–Corey–Lygo protocol for the highly enantioselective alkylation of amino acid imines under PTC conditions catalysed by quaternized cinchona alkaloids is considered a real breakthrough in the field. The possibility to develop recoverable and recyclable versions of the chiral catalysts very soon attracted the interest of many groups. Indeed the immobilization of chiral phase transfer catalysts provided the first demonstrations of the feasibility of this approach and led to a series of investigations on the use of supported catalysts in these reactions.10a Insoluble polymer-anchored ammonium salts prepared from cinchona derivatives and tested in the the standard alkylation of tert-butyl glycinate benzophenone imine afforded often low enantioselectivities and somehow contradictory results.10c The soluble polymer-supported catalysts 13 and 14 (Scheme 11.4) were prepared by attaching two different MeOPEG5000/spacer fragments to the N-anthracenylmethyl salts of

312


nor-quinine and cinchonidine, respectively.30 The behaviour of the obtained catalysts, however, fell short of the expectations. However, while with 13 enantioselectivities lower than 12% ee were always obtained, 14 showed good catalytic activity promoting the benzylation reaction in 92% yield (solid CsOH, DCM, 78 to 23 C, 22 h), but with only 30% ee. Even if this was increased to 64% by maintaining the reaction temperature at 78 C and prolonging the reaction time to 60 h, the top level of stereoselectivity obtained with the nonsupported catalyst could not be matched. XP

LINKER

OH

Cl-

+ N P

+ N

HO H

LINKER O

N

N

13

14 X = Cl or Br P

= MeO-(CH2CH2O)n-CH2CH2 , n = ca 110 O BnBr / Base

O

N

H

PTC cat

O

N

O

ClHO

Cl-

N+

N+

HO

H

H COO

P N

15a

15b

Cl-

Cl-

N+

HO

HN N

N+

HO O

H

N

H

NH

O HN

H

O +N

N

N NH H

O +N

OH

Cl-

16a

P

COO

N

Cl-

= PEG2000

16b

Scheme 11.4 Chiral polymer-supported phase transfer catalysts

OH


313

PEG was considered to be responsible, at least in part, for these results because of the following effects. By increasing the polarity around the catalyst PEG prevents the formation of a tight ion pair between the enolate and the chiral ammonium salt, the formation of which is regarded as crucial for high stereocontrol. Moreover, PEG enhances the solubility of the inorganic cation in the organic phase leading to a competing nonstereoselective alkylation occurring on the achiral caesium enolate. To check the validity of this hypothesis control experiments were carried out by performing the reaction with the nonsupported catalyst in the presence of the bis-methylether of PEG 2000. The observed ee was 65%, a value that was in good agreement with that observed with catalyst 14 but markedly inferior to the >90% ee easily achieved with the nonsupported catalyst. In the course of this study, it was also found that both the supported and the nonsupported catalysts were quite unstable, thus preventing any possibility of recovery and recycling. In a more recent work, the synthesis of two catalysts where cinchonidine (15a) and cinchonine (15b) were connected through their bridgehead nitrogen atom to MeOPEG5000 by an ester linker was reported.31 These catalysts (10 mol%) afforded the (S)-benzylation product in 81% ee and the (R)-product in 53% ee, respectively (Scheme 11.4). The relatively large difference in the observed ee is quite surprising, since the quasi-enantiomeric structure of the alkaloid catalysts should secure virtually identical ee for both enantiomers. Equally surprising and without reasonable explanation was the strong variation of the stereoselectivity of the alkylation observed on changing the reaction solvent: toluene (81% ee), benzene (64% ee), xylene (58% ee), carbontetrachloride (65%ee), DCM 3% ee!). Attempts at recycling catalyst 15a led to a dramatic drop in enantioselectivity, ascribed to the instability of the catalyst’s ester linkage under the reaction conditions. More recently Wang reported a dimeric PEG-supported cinchona ammonium salt for the alkylation of Schiff bases in water as solvent.32 Soluble polymer-immobilized catalysts 16a and 16b were prepared by the reaction of diacetoamido-PEG2000 chloride with an excess of cinchonidine and quinine respectively (Scheme 11.4). The usual benzylation of the benzophenone imine derived from tert-butyl glycinate in 1 M aqueous solution of NaOH at 25 C afforded the (S)-product in 75% ee and the (R)-enantiomer in 80% ee, respectively. It is noteworthy that satisfactory enantioselectivity and yields were obtained by employing the recovered catalysts three times, showing a higher stability of these catalysts than the previously reported systems, probably due to the acetamido group connecting the alkaloid mojety to the polymer. 11.3.2 Lewis Base Catalysts The chemistry of penta and/or hexavalent silicon compounds has recently attracted much attention because of the possibility of developing organocatalysed enantioselective reactions in the presence of cheap, low toxic and environmental friendly species such as hypervalent silicates.33 The hypervalent silicon species involved in synthetically useful processes are generally formed in situ by reaction between a four-coordinated species and a Lewis base in what is often called the ‘activation step’.34 The so-formed five- or sixcoordinated silicon species is able to promote the desired reaction in a catalytic process if the base can dissociate from silicon after the product is formed. Among the metal-free methodologies recently developed, the use of trichlorosilane as reducing agent is particularly attractive. This cheap reagent is a colourless liquid, easily

314


prepared by the silicon industry, which has already been employed on a large scale for transforming phosphine oxide to phoshine and N-acyliminium ion to N-acylamine. Trichlorosilane needs to be activated by coordination with Lewis bases, such as N,Ndimethylformamide, acetonitrile, trialkylamines, to generate hexacoordinated hydridosilicate, the real active reducing agent that operates under mild conditions. In 2004 a successful methodology was developed, in which amino acid-derived chiral formamides behaved as organocatalysts for the stereoselective reduction of imines.35 On the basis of a screening of a series of amino acids L-valine-derived formamides were selected as more efficient systems (Scheme 11.5). R H N N

HSiCl3

R2

cat (10 mol%) toluene

R1

HN

R2

O O

N

R1

R

H non-supported catalyst

H N O O

N

H

O

N

spacer

O O

C6F13 N

H

O

H 17

18

Scheme 11.5 Chiral supported Lewis bases for trichlorosilane-mediated reductions

A recoverable version of this family of catalysts has also been realized (Scheme 11.5).36 The fluorous-tagged catalyst 17 was shown to be easily recoverable and recyclable: in addition the product isolation was easier and the high level of enantioselectivity preserved (for a more detailed discussion on the fluorous recoverable catalytic systems see the dedicated chapter in this book). Recently another study on the development of a polymer-anchored version of the same catalyst was reported; different supports were investigated, such as Merrifield and extended Merrifield, Wang, TentaGel, Marshall resins were all employed to immobilize the organocatalysts through an ethereal bond.37 The enantioselective reduction of N-aryl ketoimines in the presence of trichlorosilane was performed by employing typically 15–25 mol% amount of the supported catalyst, a higher loading than that used with the nonsupported system (typically 5–10 mol% cat). The immobilized catalysts showed a remarkable dependence on the reaction solvent; while the nonsupported organocatalysts work well


315

in toluene, the polymer-anchored species behave much better in chloroform. By operating under the best experimental conditions, with the Merrifield-anchored catalyst 18, the product was isolated in good yield and in 82% ee, which is, however, 10% lower than then the enantioselectivity obtained with the nonsupported catalyst. After filtration of the immobilized organocatalyst it was possible to reuse 18 five times always maintaining the same level of stereoselectivity, but a reactivation step was required. Chiral phosphoramides, developed by Denmark38 in the late 1990s are efficient catalysts for the allylation of aldehydes with allyl trichlorosilane38a or the aldol condensation of trichlosilyl-enolethers with aldehydes. 38b In 1994 Denmark reported the first enantioselective, noncatalytic, addition of allyltrichlorosilane to aldehydes promoted by the chiral phosphorotriamide 19 (Scheme 11.6).38a A series of detailed studies demonstrated that two pathways were possible; one involving an octahedral cationic silicon atom, coordinated by

N P N

O

N

N

N

O P

N

(CH2)5

N O

N P

N

20

19

OH O R

R'

+ H

AllSiCl3, i-Pr2NEt2

SiCl3 R''

R R' R''

cat (5 mol%)

LB LB

Cl Si Cl

O

R

Cl

H

Me N

O P

N Me

N

x

Me 21

x

y

21a

100

21b

71

29

21c

37

63

y

Scheme 11.6 Chiral supported Lewis bases

0

316


two Lewis base molecules leading to a good selectivity, and another less selective one where only one phosphoroamide was bound to a pentacoordinated silicon centre. In view of these mechanicistic considerations several chiral bidentate phosphoroamides were prepared and studied in the test allylation of benzaldehyde; a catalyst loading of as low as 5 mol% of compound 20 was found to promote the reaction affording the product in high yield and enantioselectivity up to 72% (Scheme 11.6).38a However, the first example of supported chiral phosphoramides on a polymeric matrix has been reported only in 2005.39 Polystyrene-anchored catalysts 21a–c of different active site contents were used as catalysts (10 mol%) to promote the allylation of benzaldehyde with allyl trichlorosilane in the presence of excess diisopropylethylamine (DCM, 78 C, 6 h), affording the product in 82–84% yield and 62–63% ee. Remarkably, the supported catalysts proved to be more efficient than the corresponding nonsupported derivatives featuring a benzyl group instead of the polymer residue both in terms of yield and of stereoselectivity. Since it has been shown that bis-phosphoramides are more efficient than mono-phosphoramides in promoting the allylation reaction, the better results obtained with 21a–c were regarded as suggestive that two phosphoramide groups of the supported catalysts could bind the hypervalent octahedrally co€ordinated silicon atom believed to be involved in the transition structure of the reaction. In other words, the polymer backbone apparently forces two catalyst’s sites into such a close proximity that they can behave as bisphosphoramides.40 Neither the recycling of 21a–c nor the extension of their use to the allylation of aldehydes different from benzaldehyde has been described. The use of a variety of chiral 4-N,N-dimethylaminopyridine (DMAP) analogues as organic catalysts is very well known,41 but only a couple of examples of supported versions of these compounds have been described so far. A family of chiral acylating species based on the N-40 -pyridinyl-a-methylproline structure capable of promoting the kinetic resolution of alcohols with a high level of enantioselectivity was recently reported.42 The ready availability of these compounds suggested the use of the easily functionalizable carboxy function to immobilize these catalysts on different polymer supports. Among others, derivatives 22a–c collected in Scheme 11.7 were prepared connecting N-40 -pyridinyl-amethylproline to low- and high-loading polystyrene (LLPS and HLPS), and to Wang resin by standard condensation methods. These compounds were tested as insoluble catalysts (5 mol %) in the kinetic resolution of cis-1,2-cyclohexanediol mono-4-dimethylaminobenzoate carried out with a deficiency of iso-butyrric anhydride in DCM (room temperature, 16 h, Scheme 11.7). By stopping the reaction at about 50% conversion, it was possible to recover the unreacted ()-alcohol in about 75% ee. This was increased to 93% by allowing the reaction to proceed to 67% conversion. No appreciable difference in chemical or stereochemical efficiency was observed as a function of the polymeric support. Catalyst 22b, recovered by filtration and thoroughly washed with DCM, was employed in three additional runs to afford the resolved product in slighltly higher ee at identical conversions. However, the activity of the recycled catalyst was somewhat lower than that of the fresh one, since longer reaction times were necessary to obtain the same conversions. Extension of the use of this catalyst to the kinetic resolution of other secondary alcohols was possible, although the immobilized catalyst performed constantly less efficiently than its best nonsupported analogue.43 Recently the enantioselective addition of ketones or activated nucleophiles to nitrostyrene has been succesfully accomplished by employing basically two kinds of organocatalysts: a

Recoverable Organic Catalysts N

O

N

N

5% mol/cat O

OH

O

+

O

(i-BuCO)2O

317

O

O O

OH

O

Me

22a = low loading polystyrene N 22b = high loading polystyrene 22c = Wang resin

H N

O

cat

N

Scheme 11.7 Supported chiral 4-N,N-dialkylaminopyridines

bifunctional catalyst able to activate both the ketone and the nitro derivative may be used, or a simple 2-substituted pyrrolidine with a sterically demanding side chain which is believed to exert stereochemical control on the reaction. In this context systems bearing tetrazoles or triazoles have been shown to be specially efficient and they have been the subject of recent studies aimed at the development of recyclable organocatalytic systems (Scheme 11.8). It has been reported that the 1,3-dipolar cycloaddition between a polystyrene-supported alkyne like 23 and an azidomethylpyrrolidine 24 may be employed to prepare the immobilized organocatalysts 25a and 25b, where the triazole ring plays the double role of grafting the chiral pyrrolidine onto the polymer and of providing the structural element necessary to obtain a high level of enantioselectivity in the reaction.44 Such catalysts represent an interesting entry in the scenario of recoverable organocatalysts because of the extraordinary efficiency in promoting ketone addition to nitrostyrene in water. Best results were obtained with trifluoroacetic acid as additive and by using 25b rather than 25a (87% ee vs 60% ee); the key to the high performances in water of the supported catalyst seems to be the presence in 25b of a p-phenylene group in the linker, which increases the hydrophobicity of the system, where a small hydrophilic mojety such as the pyrrolidine-triazole group is embedded in a extended hydrophobic region. By using the same Cu(I)-promoted cycloaddition chemistry the same organocatalyst has been immobilized onto silica gel.45 Catalyst 26 performed well in different solvents, but it gave significant results operating in solvent-free systems. As catalyst 25 a great excess of ketone (20 mol equiv.) was used, but with 26 longer reaction times were necessary to obtain high yields. The recycling of 26 in a 72 h long reaction at room temperature was investigated and the catalyst was shown to be reused four times. An analogous catalytic system was immobilized with a slightly different methodology; in this case a silica gel-supported pyrrolidine-based chiral ionic liquid 27 has been realized and tested in the usual model reaction, the Michael addition of ketones to nitrostyrene.46 Also in this case solvent-free reaction conditions allow the products to be obtained very often in almost quantitative yield after 36 h at room temperature.

318

Recoverable and Recyclable Catalysts O N

N

N

25a

N H O Linker

O

N3 +

N Boc

24

23

N

N H O

O Ar

+

25b

N

N

Ar NO2

NO2

N O N N H

N

Cl-

O Si

N

N+

O

2 26

R''' R'' R'

+

PS

SO2OH

PS

N H R

PS

N

27

N

*

H N

SO3

-

H2N H+ N

28a

PS

SO3

-

R''' H N - R'' + SO3 R' N R H

*

H2N H+ N

28b

Scheme 11.8 Recoverable chiral pyrrolidine-based organocatalysts

Finally it is worth mentioning that a new strategy for the preparation of heterogeneous chiral organocatalysts has recently been described, a polystyrene/sulfonic acid was combined in situ with chiral amines which were immobilized through acid–base interactions.47 The resin plays the dual role of support and modulator of chemical and stereochemical


319

efficiency. Indeed it was shown that the loading of the sulphonic acid moieties plays a key role in determining the stereochemical efficiency of the supported catalyst, 1% divinylbenzene cross-linked sulphonic acid resin with 1.39 mmol/g of SO3 group being the optimal support. By a combinatorial screening different amines were investigated and highly enantioselective heterogeneous catalysts 28a and 28b were identified for the stereoselective direct aldol and Michael reactions (Scheme 11.8). The approach may be considered a valid alternative to the ‘classical’ immobilization techniques based on the formation of a covalent bond between the organocatalyst and the resin; in this case a noncovalent linkage is proposed as an efficient way to heterogenize a catalytic species without the necessity to alter with sometimes multiple synthetic manipulations the chemical structure of the organocatalyst. Compared with other methods of noncovalent immobilization such as physical adsorption or biphasic technology, where the issue of recovery and recycling remained often problematic, the present strategy relies on a stronger linkage, such as acid–base interactions, and could ensure a quite efficient recovery and reuse of the catalyst, even if the regeneration of the recovered catalyst was required in order to achieve the same level of chemical efficiency of the initial run. 11.3.3 Miscellaneous Catalysts After the pioneering work of Jacobsen several groups have recently investigated the use of ureas and thioureas as chiral Brnsted acid catalysts.48 Chiral bifunctional catalysts incorporating the thiourea moiety have been shown to efficiently promote azaHenry reaction and Michael reactions.49 However the reactions suffer from difficulty in recovering the enantiomerically pure organocatalyst. Very recently Takemoto has studied the immobilization of a structurally modified version of the successful bifunctional catalyst 29.50 In order to attach the metal-free catalyst to different polymer supports the ester derivative 30 was prepared and its catalytic ability tested. In the diethyl malonate addition to trans-b-nitrostyrene promoted by 30 the product was obtained after 2 days in 88% yield and 91% ee (vs 86% yield and 93% ee with 29 (Scheme 11.9). On the basis of these results a PEG-bound thiourea 31 was prepared and tested in the same reaction to afford the product after 6 days in 71% yield and 86% ee. The recovery and recycling of the supported catalyst were also described.

11.4 Catalysts Derived from Amino Acids Amino acids and their derivatives represent an obvious source of chiral organic catalysts, which has been fully exploited by synthetic organic chemists,51 especially in the field of aminocatalysis, such as enantioselective catalytic processes promoted by enantiomerically pure amines.52 The ready availability of amino acidic compounds and their different functionalizations in the side chains enabled a number of applications in the field of supported catalysis. While the relatively low cost of many amino acids apparently does not seem to justify the preparation of supported catalysts derived from amino acids, other reasons, as mentioned above, may drive to the immobilization of chiral catalysts, such as to experiment with different solubility properties, easy separation of the products from the catalysts and the

320

Recoverable and Recyclable Catalysts NH

NH NH

S N Me Me

CF3

S N Me Me

CF3 29

NH

NH

COOEt

CF3

30

NH

S N Me Me

COO

HN

OOC

CF3

CF3

HN S Me N Me

31 = PEG8000

NO2

EtOOC +

EtOOC

COOEt

COOEt NO2

Scheme 11.9 Supported thiourea derivative as chiral bifunctional catalyst

catalyst’s recyclability. The immobilization of these compounds on a support can also be seen as an attempt to develop a minimalistic version of an enzyme, with the amino acid playing the role of the enzyme’s active site and the polymer that of an oversimplified peptide backbone not directly involved in the catalytic activity.53 It should be mentioned at this point that in principle amine-based catalysts also offer the possibility to be recovered by exploiting their solubility profiles in acids. 11.4.1 Proline Derivatives Proline represents a typical example of a chiral organic catalyst; it presents several positive features such as low molecular weight, simple structure, high stability, nontoxicity, and extremely low cost, which make the use of organocatalysts extremely attractive compared with organometallic species. In the last chapter of the book different supports employed for the immobilization of proline are examined and discussed, with special attention given to the identification of the relative merits of the various supports, particular interest being devoted to catalyst recovery and recycling; in order to avoid undesired overlaps between the two contributions, in this section only the most important examples will be discussed, trying to give a kind of historical perspective to the topic, and a general overview and state of the art in this area, by considering all different techniques explored to realize a recoverable and recyclable proline.54


321

In this light attention will be devoted in particular to the more recent achievements, especially regarding the supported versions of prolinamides or other proline derivatives rather than proline itself. Indeed the utility and the relevance of a supporting a catalyst as cheap as proline is questionable. Nevertheless, several reasons may make proline immobilization worth of investigation; the high catalyst loading (proline is often used up to 30 mol%, which can be regarded as a large amount of catalyst, especially if the reaction is carried out on a multigram scale) is a drawback that the use of a recyclable supported species may help to overcome; moreover immobilization of proline may be used to enhance its activity and stereoselectivity and, in general, to enable the exploration of new solubility profiles for the immobilized catalytic species. Finally, a modifications of the properties of the supported catalysts by employing specific characteristics of the support might be explored. It is also worth mentioning that if proline is cheap most of its derivatives are not; for example, some of the recently developed prolinamides contain chiral, expensive diamine scaffolds, which contribute to improve the stereochemical efficiency of the newly designed catalysts, but also make them more expensive; the possibility of recycling such proline-based catalytic systems is very appealing. In general two possibilities can be considered for organocatalyst immobilization. In one case the catalysts is covalently bounded to the support: L-proline, or a proline derivative, is covalently anchored to a soluble (e.g. PEG, dendrimer) or insoluble (e.g. MCM-41, polystyrene, magnetite) support. On the other hand a noncovalently supported catalyst can be designed: in this case the organocatalyst is adsorbed (onto IL-modified SiO2), dissolved (polyelectrolytes), included (cyclodextrins) or linked by electrostatic interactions (PS/SO3H, LDH) in several supports. Biphasic catalysis has been also explored by dissolving L-proline into ionic liquids and by extracting the product using an immiscible solvent. In the present section examples of recyclable prolines, realized through different methodologies and materials, will be presented first, followed then by the more significant achievements obtained with recoverable prolinamides. Recyclable Prolines Proline was immobilized very soon after the first, seminal works of List and Barbas.55 A soluble polymer-supported version of this versatile catalyst, 32 has been prepared by anchoring (2S,4R)-4-hydroxyproline to the monomethyl ether of PEG5000 by means of a succinate spacer.53 In the presence of 0.25–0.35 mol equiv. of this catalyst, acetone reacted with enolizable and nonenolizable aldehydes (Scheme 11.11, R ¼ H, equation a) in DMF at room temperature (40–60 h) to afford the corresponding aldol products in yield (up to 80%) and ee (up to >98%), comparable with those obtained using nonsupported proline derivatives as the catalysts (that however gave faster reactions). The condensation of hydroxyacetone with cyclohexanecarboxyaldehyde catalysed by 32 afforded the corresponding anti-a,bdihydroxyketone in 96% ee (anti/syn ratio >20/1) (Scheme 11.11, R ¼ OH, equation a). The doubly loaded catalyst 33 behaved similarly to 32 while allowing half the weight amount of catalyst to be used.53,56 Replacement of the aldehyde component of the aldol reactions with imines (either preformed or generated in situ) opened access to synthetically relevant b-amino- and syn-bamino-a-hydroxyketones, which were obtained in moderate to good yields (up to 80%) and good to high diastereo- and enantioselectivity (Scheme 11.10, equation b; ee up to 97%).56

322


O O HOOC

P--OMe

32

= MeO-(CH2CH2O)n-CH2CH2, n= ca 110

O N H

O O

N H

O HOOC

P--OMe

MeO- P--OMe

O O

O

O

O

O

MeO- P--OMe

33

COOH

= CH2CH2O-(CH2CH2O)n-CH2CH2, n= ca 100

OH O

O

O R

Equation a

N H

+

R'

DMF, 25°C H

R'

30% mol/cat

R N

O R

Equation b

+

R'

PMP

PMPHN

O

DMF, 25°C H

30% mol/cat

R' R

R = H, OH

R' = Ar, Alk

PMP = 4-methoxyphenyl

Scheme 11.10 PEG-supported (S)-proline derivatives

As far as catalyst recycling is concerned, it was shown that supported catalyst 32 could easily be recovered by exploiting its solubility properties and recycled 3/4 times in all of the above-mentioned reactions. These, however, occurred with slowly diminishing yields and virtually unchanged enantioselectivity. It is worth mentioning that the use of hydroxyproline anchored via a spacer to 1% DVB cross-linked chloromethylated polystyrene as the insoluble catalyst in the aldol step of a Robinson annulation led to the product in only 29% yield and 39% ee.57 Very recently a new strategy was reported to immobilize trans-4-hydroxyproline onto insoluble Merrifield-type polymer by exploiting Cu(I)-catalysed 1,3-dipolar cycloaddition


323

(click chemistry).58 The supported catalyst 34 was successfully employed in the aaminoxylation of ketones and aldehydes (Scheme 11.11). In the best reaction conditions (20% mol/cat, 2 equiv. of ketone, DMF, 23 C) the reaction of cyclohexanone with nitrosobenzene was catalysed by 34 in 3 h with 60% yield and 98% ee (Scheme 11.11, equation a). It is noteworthy that the reaction rates of cyclic ketones with supported catalyst are higher than those reported with (S)-proline. The use of a supported catalyst allowed for a simplification of the work-up procedure, since the product may be often obtained after simple filtration of the catalyst and evaporation of the solvents; furthermore 34 was recycled up to three times without decreasing of chemical and stereochemical efficiency.

N N N

O O O P

O COOH

N H

O

34

N N N

O

3

N N N

N H N H

COOH

COOH

36

35 = Merrifield resin

= magnetite

O

O N

+

Equation a

O

DMF, 23°C

O ONHPh

OH

20% mol/cat

O O Equation b

+

Ar

O

water, 23°C H

H N N

O O Si O

H N

Ar

10% mol/cat 10% mol DiMePEG

CuI / 36

Ar N

Cs2CO3

N

+ ArBr

Equation c

OH

H N O

N H

COOH

= MCM 41

37

Scheme 11.11 Insoluble material-supported (S)-proline derivatives

324


The same insoluble, recoverable catalyst was shown to be able to efficiently promote the stereoselective aldol reaction in water.59 The condensation of cyclohexanone with different aromatic aldehydes (Scheme 11.11, equation b) was efficiently catalysed in water by 34 in the presence of a suitable additive; a substoichiometric amount of water-soluble DiMePEG, MW 2000 increased the yields, without lowering the enantioselectivity of the process (ee >93% for aromatic and heteroaromatic aldehydes). The high hydrophobicity of the resin and the aqueous environment are believed to be decisive factors to ensure high stereoselectivity. After filtration, washing the resin with acetic acid and drying, the polymer-supported hydroxyproline was recycled three times without appreciable loss of performance. And in further studies on stereoselective reactions in water different polystyrene-supported proline resins were investigated in the aldol reaction between cyclohexanone and benzaldehyde in water. By employing 10 mol% of catalyst 35 the product was isolated in 74% yield, 92% diastereoselectivity and 98% ee after 24 h of reaction. Interestingly, with resin 35 a gel-like single phase, containing up to 24% in weight of water, was formed.60 This behaviour is believed to be due to the formation of a hydrogenbond network connecting the proline and 1,2,3-triazole unit. Catalyst 35 was used in the adol reactions between cyclohexanone or cyclopentanone and several aromatic aldehydes with high yields and stereoselectivities. In absence of water lower stereoselectivity was observed. Finally, cross-aldol reaction of propanal gave the product with a 97% ee for the major anti isomer even using only 1 mol% of 35. Such a resin was recycled and re-used five times without any appreciable loss in yield or in stereoselectivity. By following the same approach a magnetic nanoparticle-supported proline 36 was developed; azide-functionalized magnetite nanoparticles, prepared by coprecipitation of iron(II) and iron(III) ions in basic solution at 85 C, were reacted with suitably functionalized prolines to afford the supported catalyst with a loading of 2 mmol g1 as determined by elemental analysis.61 The immobilized proline was employed in CuI-catalysed Ullmanntype coupling reactions of aryl/heteroaryl bromides with various nitrogen heterocycles (Scheme 11.11, equation c). By exposure of the reaction mixture to an external magnet the catalyst was recovered and recycled four times, showing a slight decrease in chemical activity, from 98 to 93%. The modified proline 37 immobilized on the mesoporous siliceous material MCM-41 has been prepared, with an active site loading of 0.52 mmol g1.62 Large amounts of this catalyst (47–52 mol%) were then employed to promote the condensation between hydroxyacetone and iso-butyraldehyde (room temperature, 24 h) and benzaldehyde (90 C, 24 h) in DMSO or toluene to afford the products in yields that were only marginally higher than the amount of catalyst used (55–60%) (Scheme 11.11). The reaction times were significantly shortened to 10–30 min by the use of microwave irradiation. Catalyst 37 promoted the exclusive formation of anti diols of >99% ee in the condensation involving isobutyraldehyde, while the reaction with benzaldehyde led to syn diols in lower stereocontrol (syn/anti ratio 1.4/1; ee of syn 80%). Two recyclings of the catalyst recovered by filtration were shown to occur in slowly decreasing yield and unchanged diastereoselectivity, but with no information about the enantioselectivity. (S)-proline has been recently supported on the surface of modified silica gels with a monolayer of covalently attached ionic liquid with or without additional adsorbed ionic liquid.63 These material were able to catalyse the aldol reaction between acetone and several aldehydeswithgoodyieldsandeevalues,comparablewiththoseobtainedunderhomogeneous


325

conditions, especially with 4-methylpyridinium-modified silica gel 38 (Scheme 11.12). Moreover, these supported species were easily recovered by simple filtration and reused at least up to seven times. Exploring the same kind of approach proline was supported as an anion in a polystyrene-supported imidazolium resin which was prepared in few steps.64 The immobilized proline 39 was investigated in the CuI-catalysed N-arylation of heterocycles (ec. c, Scheme 11.11). In the best conditions (DMSO,120 C, 60 h, 2.4 equiv. of K2CO3, 10 mol% of CuI and 39, proline loading: 0.69 mmol g1, containing 20 mol% of proline unit) yieldsoften higherthan90% wereobserved; In the recyclingstudies,innine consecutive cycles of the reaction between imidazole and 4-bromobenzonitrile at 90 C, the yield was shown to decrease from 95 to 73%.

BF4OMe

+

O Si O

N

N

COO

N

38

N H

COOH

N H

39

+N O Si O OMe BF 4

= SiO2 = polystyrene

O

O

NO2 OH

H

+

O

NO2

O COOH

N H

= β-cyclodextrin X-

+N

H H

40

BF4-

X41

O

N N

n

+ - + N COO N

O

Bu N H

42

COOH

Tf2N-

O

N O

43

Scheme 11.12 Immobilized (S)-proline derivatives

N H

COOH

326


Another interesting noncovalent immobilization technique was exploited in the synthesis of catalyst 40 (Scheme 11.12). In this case the apolar phenyl ring of 4-phenoxyproline served as the handle for including the amino acid in the b-cyclodextrin cavity. Interestingly, the amount of catalyst actually included was found to be directly dependent on the temperature of the inclusion reaction, with higher temperatures leading to higher extent of inclusion. Thus, catalyst samples with different loadings could be obtained.65 In the aldol addition of acetone to 2-nitrobenzaldehyde it was shown that in the presence of the highest loaded catalyst (10% mol/cat) the product could be obtained in 90% yield and 83% ee after 16 h at room temperature. The catalyst was recovered by filtration and employed for three subsequent runs occurring in slowly decreasing yields (79%, 4th run) and unchanged enantioselectivity, a behaviour that almost perfectly paralleled that observed for PEGsupported proline over the same number of recycling experiments.56 In this context an (S)proline/polyelectrolyte system has also been recently reported;66 the amino acid was adsorbed on the solid support by simply mixing a poly-(diallyldimethylammonium) salt suspension in methanol with the organocatalyst solution in the same solvent (Scheme 11.12). The heterogeneous supported catalytic system 41 was shown to be able to promote the aldol reaction between acetone and aromatic aldehydes in yields and enantioselectivities comparable with those obtained in the proline-catalysed reactions. The immobilized catalyst was recovered by filtration and reused six times without appreciable loss of stereoselectivity. The use of ionic liquids or modified ionic liquids to immobilized proline is described elsewhere in this book, where a whole chapter is dedicated to the topic;54,67 here it is worth mentioning that ionic liquids may also be employed as support for the chiral catalyst. For example, proline was successfully anchored to an ionic liquid by reaction of N-benzyloxycarbonyl-(2S,4R)-4-hydroxyproline benzyl ester with a proper ionic liquid carboxylic acid followed by deprotection to afford supported catalyst 42.68 The supported catalyst performed better in the aldol condensation between acetone and different aldehydes by working in neat ketone as reaction solvent (60–87% ee). The recycling was investigated up to four cycles with good reproducibility of the enantioselectivity. Analogously it was reported that oniumtagged proline 43 was able to promote direct aldol condensation in 36–78% yield and 75–94% enantioselectivity.69 Recycling experiments showed a diminished yield after three cycles (from 75% to 30%) and slightly lower stereoselection (from 85% to 80% ee). Recyclable Prolinamides It should be mentioned at this point that in principle amine-based catalysts also offer the possibility of being recovered by exlpoiting their solubility profiles in acids, without the necessity of any functionalization in order to heterogenize the catalytic system. Recently several prolines derivatives have been prepared and tested in aldol reactions;70 in this context the synthesis of new organocatalysts obtained by connecting the proline moiety to a 1,10 -binaphthyl-2,20 -diamine scaffold, easily prepared in a few steps from inexpensive, commercially available, enantio-pure materials was reported.71 (S)-prolinamide 44 in the presence of stearic acid was able to promote the direct aldol condensation of cyclohexanone and other ketones with different aldehydes in the presence of a massive amount of water in very good yields, high diastereoselectivity and up to 95% ee (Scheme 11.13).72 A preliminary experiment to recover and recycle the catalytic system was attempted for the condensation between cyclohexanone and benzaldehyde catalysed by 44 and stearic

Recoverable Organic Catalysts O N H

327

H N

HN O

HN

44 O

10 mol % cat water

O +

R

H

O

OH R

20 mol % stearic acid

R = Aromatic, aliphatic

Scheme 11.13 Recoverable prolinamides for aldol reactions in water

acid. At the end of the reaction pentane was added and the biphasic system was separated; to the recovered ‘suspension’ in aqueous phase, new reagents were added and the reaction was run again. The product was obtained by simple evaporation of the pentane phase. Following this procedure the recycling of the stearic acid/44 catalytic system provided the product in 98/2 anti/syn ratio and 93% ee (compared with 99/1 anti/syn ratio and 93% ee of the first cycle), showing that the recycling of the catalyst is feasible. Following a more traditional approach polystyrene-supported prolinamide 45 was recently prepared and successfully tested in the direct aldol addition of acetone or cyclohexanone to several aromatic aldehydes.73 Best reaction conditions were found to be a 1 : 2 water/chloroform mixture, where the good swelling properties of chloroform could be compromised with the formation of a concentrated organic phase due to the presence of water (Scheme 11.14). It is noteworthy that the catalyst could be recovered and reused, after regeneration, for 12 reaction cycles.

Ph

Ph O

Ph OH

N H

H N N O

N

HN SO2

H N

HN C4F9 SO2

Linker 45

47

46

= polystyrene

O

O R

H

+

Ar

NO2

Ar NO2

H R R'

R'

Scheme 11.14 Recoverable immobilized prolinamides

H N

328


Immobilized versions of proline-sulphonamides have also been studied; for example, ionic liquid-supported pyrrolidine 46 was prepared in a few steps from (S)-2-aminomethylN-Boc-pyrrolidine and 3-chloropropanesulfonyl chloride.74 The catalyst was used for several Michael addition reactions of aldehydes with nitroolefins (Scheme 11.14). In the optimal conditions (cat. 20 mol%, 4 C, 6 d), yields ranging from 29 to 64% were obtained, with stereoselectivities syn: anti (89: 11)–(97: 3) and enantioselectivities ranging from 64% to 88%. The catalyst was recovered by precipitation with diethyl ether and recycled three times with no decrease in either yield or enantioselectivity. It must be noted that sulphonamide 47 featuring a C4 fluorous tail was employed to catalyse the addition of cyclohexanone to nitrostyrene in water, affording the product in 95% yield, basically single syn isomer and 90% ee.75 The recovery was performed using fluorous solid-phase extraction technique and the catalyst reused six times.76 11.4.2 Amino Acid-derived Imidazolinones Other chiral organic catalysts of major success are the protonated phenylalanine-derived imidazolidinones 48 developed by MacMillan,77 which have found widespread use in a number of relevant processes. Immobilized versions of these catalysts were developed both on soluble (PEG-supported catalyst 49,78) and on insoluble supports (catalysts 50 and 51,79) and employed in enantioselective Diels–Alder cycloadditions of dienes with unsaturated aldehydes (Scheme 11.15). In the reaction between acrolein and 1,3-cyclohexadiene (Scheme 11.15, equation a) under the best conditions, which involved the use of the trifluoroacetate salt of 49 (0.1 mol equiv.) in a 95/5 acetonitrile/water mixture at room temperature for 40 h, the product was obtained in 67% yield as a 94/6 mixture of endo/exo isomers having 92% and 86% ee, respectively. This result does not different very much from that obtained with 0.05 mol equiv. of the nonsupported catalyst (hydrochloride form), that afforded the product in 82% yield and 94% ee. Unfortunately, catalyst 49 (as well as its nonsupported, commercially available counterpart) proved to be rather unstable under the reaction conditions. Accordingly, catalyst recycling over four cycles showed a marked decrease in the chemical yield of the reaction (from 67 to 38%), while the ee was eroded more slowly (from 92 to 85%). A comparison with insoluble supported systems showed that the JandaJel-supported catalyst 50 performed better not only than its PEG and silica-supported analogues, which behaved almost identically to each other, but, quite surprisingly, also than the nonsupported compound. The PEG-supported imidazolidinone 49 was employed in 1,3-dipolar cycloadditions too.80 By reacting N-benzyl-C-phenylnitrone with acrolein, it was shown that the outcome of the reaction was strongly dependent on the nature of the acid employed to generate the catalyst, and that only the use of HBF4 allowed reproducible results to be obtained (Scheme 11.15, equation b). Under the best reaction conditions (20 mol% of catalyst, DCM, 20 C, 120 h) the product was obtained in 71% yield as a 85 : 15 trans/cis mixture of isomers, 87% ee for the trans isomer. The major difference between the PEG-supported and the nonsupported catalyst resides in the chemical rather than in the stereochemical efficiency. Indeed, while the supported catalyst gave trans/cis ratios almost identical to and ee only 3–6% lower than those obtained with the nonsupported catalyst, the difference in chemical yields was larger, ranging from 9 to 27%.

Recoverable Organic Catalysts O N

H

nBu Me

N

nBu

O N

H

N H

Me

329

Me Me

O O 49

48

O N

H

N H

= MeO-PEG5000

O Me

H

Me

N

Me = JandaJel

N H

= (CH2)3-silica 50

Equation

51

H

+

a

O O

Bn

Bn Equation

b

Ph + + N Bn O

N O

N O

H O

H

+ Ph O

H

trans

Ph O

H

cis

Scheme 11.15 Supported chiral imidazolinones

The PEG-supported catalyst was recycled twice to afford the product with constant level of diastereo- and enantioselectivity but with chemical yields diminishing from 71 to 38%. In order to explain this behaviour several experiments were carried out. After each recovery the supported catalyst was examined by 1 H NMR that showed degradation, increasing after each cycle, probably due to an imidazolidinone ring opening process. Indeed extensive catalyst degradation in the presence of acrolein, less degradation with crotonaldehyde, and

330


essentially no degradation with cinnamaldehyde was demonstrated by NMR analysis, while nitrone did not exert any effect on the catalyst stability. In agreement with the results of these experiments, the nonsupported catalyst also showed a marked instability and decrease in chemical efficiency when recycled.80 By employing (L)-tyrosine more recently a MacMillan type catalyst was anchored to novel siliceous mesocellular foams by two strategies.81 In one case allyl silanes bearing organocatalyst were reacted with the siliceous material that was then capped with hexamethylsilazane (catalyst 52a). In a second approach a polystyrene-coated catalyst 52b was prepared; capped foams were embedded with styrene, divinylbenzene and the catalyst featuring a styrene moiety (Scheme 11.16). The mixture was polymerized coating the walls of the siliceous material with a polystyrene resin. Both the supported catalysts were tested in the cinnamaldehyde cycloaddition with cyclopentadiene, showing a good chemical activity, even at 5% catalyst loading, but slightly lower enantioselectivities than the nonsupported catalyst (83% ee vs 93%). Only 52b was shown to maintain the same level of stereoselectivity in the second cycle.

O N

H

N H

Me

O N

H

Me

N H

OSiMe3 MCF

OSiMe3

PS

Me Me

OSiMe3

O MCF

OSiMe3

O Si

O

OSiMe3 52a

52b C8F17 O N

H

N H

Me Me

O H

N N H

Me Me H+

53

54 MCF

= siliceous mesocellular foams

= montmorillonite clay

Scheme 11.16 Silica-supported chiral imidazolinones


331

In another approach MacMillan type catalyst 53 bearing a perfluorinated tail was prepared.82 In the acrolein Diels–Alder reaction with cylopentadiene the catalyst showed excellent performances (93.4/6.6 endo/exo, 93.4% ee for the endo isomer) and it was recovered through the use of a fluorous silica phase. When recycled, it promoted the reaction with similar level of stereoselectivity.83 Finally it must be mentioned that lately a noncovalent strategy was explored in order to develop chiral recyclable imidazolinones. Montmorillonite clay was employed to immobilize the hydrocloride salt of MacMillan type catalyst through a cation-exchange reaction.84 The use of mont-entrapped organocatalyst 54 as heterogeneous catalyst was investigated in the Diels–Alder reaction. By working at 0 C and by often using an additive such as a carboxylic acid, the cycloadduct was obtained in the best conditions with 96/4 endo selectivity and 92% ee. It was noteworthy that after a simple filtration the catalyst was reused up to four times always maintaining the same level of enantioselectivity, apparently not showing any catalytst decomposition. The authors propose that the remarkable catalytic activity of the supported catalyst is due to the expansion of the interlayer space; the silicate sheet of montmorillonite would act as a counter anion able to entrap the organic molecule while keeping its catalytic efficiency. 11.4.3 Other Amino Acids In an attempt to investigate the use of polyamino acid derivatives as organocatalysts in the direct aldol reaction several di- and tripeptides were supported on Novasyn TG amino resin, an amine-terminated PEG polystyrene graft copolymer, and tested in the condensation between acetone and 4-nitrobenzaldehyde.85 By running the reactions in acetone at 20 C for 24 h the product was obtained in yields ranged from 13 to >99% and ee values from 22 to 77%. Higher enantioselectivities but lower yields were obtained with polymer-bound tripeptides. Supported proline–serine dipeptide was then selected for further studies in different solvents (acetone–water, DMSO–acetone, CH2Cl2–acetone at 20 C) and at several temperatures (acetone at 15, 25 and 45 C), the best result being achieved in acetone at 25 C for 40 h (98% conversion and 82% ee). Recently it has been demonstrated how tripeptide H–L-Pro–L-Pro–L-Asp–NH255 could act as very active catalyst for the aldol reaction between acetone and aldehydes.86 Employed at 1 mol% it afforded the condensation product in several cases with higher enantioselectivities than proline. In order to further explore the potentialities of the catalyst the immobilization on a solid support was studied.87 Tripeptide 55 was therefore anchored on different resins such as polystyrene, PEGA (polyethylene glycol–polyacrylamide), SPAR (polyacrylamide) and TentaGel (polyethylene glycol–polystyrene) (Scheme 11.17). The immobilized catalytic materials 56 were evaluated in the model reaction between acetone and 4-nitrobenzaldehyde. By performing the reaction at 25 C for 18 h with 1 mol% of the catalyst and N-methylmorpholine as base, TentaGel was identified as the support of choice. The reaction between acetone and five aldehydes, aromatic and aliphatic as well, in the presence of 56 gave comparable results with those obtained with nonsupported peptide. Recycling studies showed that enantioselectivity was maintained for at least eight cycles while activity decreased after three cycles. It must be added that recently tripeptide 55 was also adsorbed on modified-silica gels 57 (Scheme 11.17). The usual aldol condensation of acetone with different aldehydes gave the best results at 20 C. The recycling of the

332


H N N H

O

O

N H

COOH

O

55

O

COOH

56

OMe

N

O Si O

O H C N

H N

CONH2

N

S BF4-

N H

BF4O Si O OMe

N

H N O

CONH2

O

= SiO2

COOH

N

S

57 O H

Equation a H O O H

Equation b H

O H N

H

O

O NH 8

O

59

O

H N

H N

NH

O

H

454

8

58 H

H

H N

O

5

= MeO

O

N H

OMe

O

110 15

60 59-P

O 110

O Equation c

61

Ar

O H

(Leu)25.4-Trp-Trp-Aib-DPro-Pro / TFA

Ar

H

= PEG-PS

Scheme 11.17 Supported poly(amino acids)

immobilized catalyst was studied and it was shown that enantioselectivity was maintained over four cycle (83% ee) but with a marked decrease in conversion from 91% to 42%.88 Another reaction where amino acids play a key role is the Julia–Colonna epoxidation of a,b-unsaturated ketones,89 which involves the use of catalytic amounts of polymeric amino


333

acids, able to catalyse the Weitz–Scheffer epoxidation of chalcone using basic hydrogen peroxide, with high enantioselectivity (Scheme 11.17, equation a). An improved procedure for chalcone epoxidation was developed which involved the use of the urea–hydrogen peroxide complex as the oxidant and DBU as the base in anhydrous THF at room temperature.90 Under these conditions, fast (30 min), high yielding (85–100%), and highly stereoselective (>95% ee) reactions were observed. The immobilization of poly(amino acids) on polymeric supports is still attracting a lot of interest; however, recovery and recycling of these catalysts remained a problem; best results were obtained by immobilization of the poly(amino acid) on silica gel, which produced a very active catalyst recoverable by filtration and recyclable at least five times without any appreciable loss in activity and stereoselectivity.91 After extended work on insoluble polystyrene-supported polyamino acids later efforts focused on the immobilization on soluble PEGs, and proved the superiority of the latter support. The high molecular weight soluble catalyst 58 (Scheme 11.17; average MW of the PEG fragment 20 000 Da), featuring two terminal leucine octamers, was synthesized.92 This catalyst was employed to promote the epoxidation of chalcone (99% yield, 94% ee) in a continuously operated membrane reactor, where catalyst retention was achieved by means of a nanofiltration membrane. This equipment allowed 25 reaction cycles to occur with almost unchanged yield and selectivity, after which some decrease in both values were observed. A series of catalysts having general structure 59 (Scheme 11.17) were synthesized by attaching the C-terminus of leucine oligomers of different length to the PEG modified DVB cross-linked polystyrene TentaGel S NH2.92 These compounds were used to establish that a polymer-supported catalyst containing as few as five amino acid residues was able to catalyse the Julia–Colonna epoxidation of chalcone with up to 98% ee. By replacing the insoluble support of 59 with soluble MeOPEG, adduct 59-P was obtained (Scheme 11.17); also in this case a leucine pentamer was found to be the minimum structural requirement for achieving stereocontrol (>50% ee). Since at least four amino acid residues are required to form one turn of the a-helical structure, it was concluded that one complete turn was required for efficient stereoselectivity. Since in the case of nonsupported polyleucine catalyst a good level of enantioselectivity is observed only at the decamer level, the poly (ethylene glycol) portions of 59 type catalysts could act as a helix surrogate forcing the oligopeptide to adopt the helical arrangement at shorter chain lengths. More recently, by attaching a polyleucine chain to MeOPEG5000NH2 a new soluble catalyst 60 was prepared (Scheme 11.17),93 which promoted the epoxidation of chalcone with the urea/hydrogen peroxide complex in 95% conversion and 97% ee (DBU, THF, room temperature, 3 h). In this catalyst, CD measurements have shown that the peptide fragment exists mostly in the a-helical structure (86%); once again the existence of a direct relationship between content of a-helical structure and catalytic activity and, to a lesser extent, enantioselectivity was demonstrated. Further experiments showed that the minimum number of Leu residues necessary to achieve high stereoselectivity with these PEG-supported catalysts was six, in agreement with the observation that at least four amino acids are required to form a whole turn of the a-helical structure. Finally a polyleucine chain was a key structural element in the design of a novel resinsupported prolyl peptide 61 having a b-turn moiety as catalyst for the enantioselective reduction of a b-unsaturated aldehydes by transfer hydrogenation (Scheme 11.17, equation c).94 An essential element for obtaining high enantioselectivity in a 1/2

334


tetrahydrofuran–water mixture in the reduction with Hantsch ester of b-methyl-substituted a b-unsaturated aldehydes was shown to be the length of the hydrophobic polyleucine chain, in a system that is believed to simulate the behaviour of enzymes in the cell. In order to optimize the catalyst performance peptides including the D-pro-Aib (Aib: 2-aminoisobutyric acid) were selected, based on their ability to form peptidic b-turn structures, which were already demonstrated to work as efficient organocatalysts in apolar solvents. In the best conditions, 20 mol% of catalyst 61 was able to reduce different aldehydes with enantioselectivities often higher than 90%. The recovery and the recycling of the catalyst was not reported.

11.5 General Considerations on Recyclable Organocatalysts One of the goals of the present chapter was to demonstrate with several examples of achiral as well as chiral organocatalytic systems that different valid reasons may be considered for the development of easily recoverable catalysts. The recycling of the organocatalyst is maybe the more important and the more obvious aspect, especially in the case of chiral, precious molecules. But other issues can be tackled by designing a supported system, in order to have, for example, a simpler separation of the catalyst from the reaction mixture or an easier isolation of the reaction products; the possibility of using immobilization to change the catalyst’s solubility properties, to address the problem of the catalyst’s decomposition, or to develop new technologies to be used in environmentally friendly or green solvents are all equally interesting topics. In this context the choice of the support may be fundamental, but strongly dependent on the reaction or the problem under evaluation. It is also clear that different strategies of immobilization are available and often offer comparable results; so far, strictly concerning the recovery and recycling of chiral species it has been shown that covalently bonded supported organocatalysts have been more successfully employed than noncovalently linked systems. Other considerations may include the economic aspect of the immobilization technologies; the synthesis of the supported catalyst should be designed so that it could exploit as starting material a catalyst precursor comparable in cost and synthetic complexity with the compound used for the synthesis of the nonsupported catalyst. However, since in the last chapter of the book some general considerations on the immobilization process are made it’ is not the object of this contribution to further discuss the argument in detail, while in this section we would rather consider in particular two points which up to now have been underdeveloped and represent innovative areas of future certain interest. One topic is the application of immobilized systems to the discovery and optimization of catalytic systems. Several examples of the generation of libraries of catalysts have been reported in the literature.95 However, quite surprisingly, the use of supported catalysts in studies aimed at generating libraries of candidates and at finding the catalyst of choice has up to now remained largely unexplored. Besides a few cases recently described47,87 the more significant example still may be considered Jacobsen’s work;96 that represents a remarkable exception in which the development of the immobilized catalyst preceeded and was crucial to that of the nonsupported one. A fully organic catalyst for the Strecker reaction was developed using a thiourea based chiral Brnsted acid that turned out to be extremely chemically active, stereoselective, and broad in application. The optimization of the


335

catalyst structure was realized through a series of modifications of the salen-based structure carried out on an insoluble polystyrene support and using the principles of combinatorial chemistry for finding out the best amino acid, diamine, and diamine-amino acid linker combination. The screening of three successive libraries led to the identification of the supported thiourea catalyst 62 as the best one from which the nonsupported counterpart was derived (Scheme 11.18).

S 63

NH

t-Bu

N

= 1% cross-linked polystyrene

NH HO

N

OCOBu-t

O t-Bu

N

1% mol cat 63 HCN

H

CN N H

Scheme 11.18 Supported cinchona alkaloids for stereoselective b-lactams synthesis

The hydrocyanation of N-allyl or -benzyl imines derived from aromatic and aliphatic aldehydes and of some ketones was promoted by only 1 mol% of the catalyst in very high yield and almost complete stereoselectivity. It is interesting to note that the soluble and the resin-bound catalysts performed equally well. Recovery and recycling of the supported catalyst was demonstrated to occur without any erosion of chemical and stereochemical efficiency over ten reaction cycles. Given these excellent results, it is surprising that this approach has not been used more extensively for the discovery of chiral organic catalysts. The success of this methodology is even more significant if one considers that catalysts 62 are among the few chiral organocatalysts to be currently employed at the industrial level. The other topic worthy of special consideration is the recovery and recycling technology. In the context of catalyst separation and recycling, a system where a catalyst does not need to be removed from the reaction vessel is very attractive. An example comes from continuous flow methods,97 when the immobilized catalyst permanently resides in the reactor where it transforms the entering starting materials into the exiting products. The retention of the catalyst inside the reaction vessel can be achieved by different techinques ranging from ultrafiltration through an MW-selective membrane to immobilization on a silica gel column. Recently a few examples of organocatalysed reactions performed in continuous flow methodology have been reported but also in this case the area is open to further numerous studies.98 A whole chapter in the book is dedicated to organic synthesis with mini flow reactors; here it is worth mentioning only that in this field Lectka and co-workers obtained

336


spectacular results in a highly stereoselective synthesis of b-lactam.99 A process was devised which involves the use of solid-phase reagents and catalysts that constitute the packing of a ‘series of reaction columns’. In Scheme 11.19 the chemical steps in the catalytic asymmetric synthesis of b-lactams are illustrated, including a ketene generation step (SP base), the b-lactam formation (SP catalyst) and a purification step (SP scavenger). The catalyst involved is a quinine derivative, 63, anchored to Wang resin through an appropriate spacer, capable of yielding the products with very high stereoselectivity (>90% ee). It is noteworthy that carrying out the chemical reactions on sequential columns leads to an easy recovery of catalyst and reagents, and to simplified purification steps that avoid the need for chromatography.

O OH

N

= Wang resin

Supp. Cat. 62 N

R O

O

.

Supp. Base Cl

Ts

Supp. Cat.

R

COOEt

N

+ COOEt

R

O

Supp. Scav.

R

N Ts

O

COOEt N Ts

+ byproducts

Scheme 11.19 Supported Jacobsen’s thiourea-based organocatalyst

It is important to note that, under continuous flow conditions, product isolation, catalyst recovery and recycling are obtained in a single operation. The convenience of this approach is demonstrated by the fact that for this system one of the highest recycling numbers for a chiral organic catalyst has been accomplished. Quite interestingly, a few number of runs (five to ten) were necessary to obtain a catalyst aged enough to afford consistent results, since quinine ‘bleeding’ from the freshly prepared catalyst was found to occur. A properly aged resin performed with no erosion in yield or selectivity for as many as sixty cycles. Recently a comparison was made between the preparation of trans-1,2-cyclohexanediol in standard glassware, following a traditional batch production, and in microreactors, with a continuous flow production.100 In the microreactor the reaction could be carried out with up to three times higher reagent concentrations, leading to faster reaction rates, a reduced amount of waste solvent and a higher purity product.

11.6 Outlook and Perspectives In the last 20 years solid-supported organic catalysts have become powerful synthetic tools readily available to the chemical community; the reasons for developing an immobilized version of a chiral catalyst go well beyond the simple, still fundamental, aspect of the


337

recovery and the recycling of the precious catalytic species; stability, structural characterization, catalytic behaviour, new or different solubility properties, simplification of the reaction work-up, catalyst discovery and optimization, use in environmentally friendly or green solvents are all issues that may be conveniently addressed when working with supported systems. Almost totally unexplored areas of investigation are open avenues for researchers ready for new challenges; the use of libraries of supported catalyst in a combinatorial approach towards the optimization process of a catalyst, or the application of continuous flow methods and microreactors-based technologies are all fields of investigation of potential enornous growth.97c Significant progress could derive from the development of an interdisciplinary expertise with contributions from organic, polymer and material chemists. This could also allow some problems to be solved related to the cost and the commercial availability of the support, which is another issue of great importance in determining (and somehow limiting) the choice of the support. In conclusion, it is clear now that stereoselective organocatalysis has achieved the same relevance to asymmetric synthesis as organometallic catalysis, and every day novel fully organic methods are discovered and developed to perform in the future an even wider variety of reactions. A few chiral organocatalysts are already employed in large scale preparation at industrial level.101 In this context, the development of immobilized chiral organic catalysts will play an important role, in contributing to further expanding the applicability of organic catalysts and even to helping to discover new chiral organic catalytic species. The immobilization of chiral organic catalysts represents a relatively new field of research, in great expansion, open to the interdisciplinary contributions of organic and material chemists. It is a really multifaceted and stimulating chemistry where creativity, fantasy and courage are all qualities required by the modern chemist who wishes to work successfully in this field.

References 1. In: A. Berkessel and H. Groger, Editors, Asymmetric Organic Catalysis, Wiley-VCH, Weinheim (2005). See also in: P.I. Dalko, Editor, Enantioselective Organocatalysis: Reactions and Experimental Procedures, Wiley-VCH, Weinheim (2006). 2. (a) P.I. Dalko and L. Moisan, Angew. Chem. Int. Ed. 35 (2004), p. 5138. (b) Acc. Chem. Res. 37 (2004), thematic issue 8. (c) Adv. Synth. Catal. 346 (2004), thematic issues 8–9; (d) A. Dondoni and A. Massi, Angew. Chem. Int. Ed., 2008, 47, 4638–4660; (e) P. Melchiorre, M. Marigo, A. Carlone and G. Bartoli, Angew. Chem. Int. Ed., 2008, 47, 6138–6171. 3. P.T. Anastas and J.C. Warner, Green Chemistry: Theory and Practice, Oxford University Press, New York, 1998. 4. (a) P.T. Anastas, Green Chemistry Textbook, Oxford University Press, New York, 2004; (b) M. Lancaster, Green Chemistry, an Introductory Text, Royal Society of Chemistry, Cambridge, 2004. 5. (a) E.N. Jacobsen, A. Pfaltz and H. Yamamoto, Comprehensive Asymmetric Catalysis, Springer, Berlin, 1999. (b) R. Noyori, Asymmetric Catalysis in Organic Synthesis; Wiley, New York, 1994. 6. H.U. Blaser, Chem. Commun. 2003, 293–297. 7. A.M. Rouhi, Chem. Eng. News, 2004, 48, 47–62. 8. A. F. Trindade, P.M.P. Gois, and C.A.M. Afonso, Chem. Rev. 2009, 109, 418–514.

338


9. (a) R.A. Sheldon and H. van Bekkum, Fine Chemicals Through Heterogeneous Catalysis, Wiley-VCH, Weinheim, 2001. (b) D.E. De Vos, I.F.J. Vankelecom and P.A. Jacobs, Chiral Catalysts Immobilization and Recycling, Wiley-VCH, Weinheim, 2000. (c) K.J. Ding and F.JK. Uozomi, Handbook of Asymmetric Heterogeneous Catalysts, Wiley-VCH, Weinheim, 2008. 10. Reviews on supported organic catalysts: (a) M. Benaglia, A. Puglisi and F. Cozzi, Chem. Rev. 2003, 103, 3401–3429. (b) F. Cozzi, Adv. Synth. Catal. 2006, 348, 1367–1390. For a perspective on chiral supported metal-free catalysts see: (c) M. Benaglia, New J. Chem. 2006, 30, 1525–1533. 11. K. Drauz and H. Waldmann, Enzyme Catalysis in Organic Synthesis: A Comprehensive Handbook, Vol. 3, 2nd Edition, Wiley-VCH, Weinheim, 2002. 12. D.J. Diller, Curr. Opin. Drug Discovery Develop., 2008, 11, 346–355. 13. (a) P.L. Anelli, F. Montanari and S. Quici, Org. Synth. 1990, 69, 212 and references cited therein; (b) T. Vogler and A. Studer, Synthesis 2008, 1979–1993. 14. (a) A. Dijksman, I.W.C.E. Arends and R.A. Sheldon, Chem. Commun. 2000, 271–272; (b) T. Fey, H. Fischer, S. Bachmann, K. Albert and C. Bolm, J. Org. Chem. 2001, 66, 8154–8159; (c) R. A. Sheldon and I.W.C.E. Arends, Adv. Synth. Catal. 2004, 346, 1051–1071. 15. K.D. Janda, T.J. Dickerson and N.N. Reed, Chem. Rev., 2002, 102, 3325–3244. 16. G. Pozzi, M. Cavazzini, S. Quici, M. Benaglia and G. Dell’ Anna, Org. Lett., 2004, 6, 441–443. 17. G. Pozzi, S. Quici, M. Benaglia, A. Puglisi and O. Holczknecht, Tetrahedron, 2005, 61, 12058–12064. 18. A. Gheorghe, A. Matsuno and O. Reiser, Adv. Synth. Catal., 2006, 348, 1016–1020. 19. V.D. Bock, H. Hiemstra and J.H. van Maarseveen, Eu. J. Org. Chem., 2006, 51. 20. J. Luo, C. Parrdin, W.D. Lubell and X.X: Zhu, Chem. Commun., 2007, 2136–2138. 21. D. Astruc, F. Lu and J.R. Aranzaes, Angew. Chem. Int. Ed., 2005, 44, 7852. 22. A. Schatz, R.N. Grass, W.J. Stark and O. Reiser, Chem. Eur. J., 2008, 14, 8262–8266. 23. For another application of magnetic bead-supported catalysts see: M. Kawamura and K. Sato, Chem. Commun., 2006, 4718–4719. 24. M. Benaglia, R. Annunziata, M. Cinquini, F. Cozzi and G. Tocco, Org. Lett., 2000, 2, 1737–1739. 25. M. Benaglia, M. Cinquini, F. Cozzi and G. Tocco, Tetrahedron Lett., 2002, 43, 3391–3393. 26. M. Benaglia, T. Danelli, F. Fabris, D. Sperandio and G. Pozzi, Org. Lett., 2002, 4, 4229–4232. 27. R. Ballini, L. Barboni, L. Castrica, F. Fringuelli, D. Lanari, F. Pizzo and L. Vaccaro, Adv. Synth. Catal., 2008, 350, 1218–1224. 28. C.A. Dalaigh, S.A: Corr, Y. Gun’ko and S.J. Connon, Angew. Chem. Int. Ed., 2007, 46, 4329–4331. 29. Review on asymmetric phase-transfer catalysis: (a) A. Nelson, Angew. Chem., Int. Ed. Engl. 1999, 38, 1583. (b) T. Ooi and K. Maruoka, Aldrichim. Acta, 2007, 40, 77–86 (d) T. Ooi and K. Maruoka, Angew. Chem., Int. Ed. Engl., 2007, 46, 4222 For a perfluorous recyclable version of Maruoka’s type chiral phase transfer catalyst see:S. Shirakawa, Y. Tanaka and K. Maruoka Org. Lett., 2004, 6, 1429. 30. T. Danelli, R. Annunziata, M. Benaglia, M. Cinquini, F. Cozzi and G. Tocco, Tetrahedron: Asymmetry 2003, 14, 461. 31. B. Thierry, J.-C. Plaquevent and D. Cahard, Tetrahedron: Asymmetry 2003, 14, 1671–1677. 32. X. Wang, L. Yin, T. Yang and Y. Wang, Tetrahedron: Asymmetry 2007, 18, 108. 33. M. Benaglia, S. Guizzetti and L. Pignataro, Coord. Chem. Rev., 2008, 252, 492–512. 34. (a) C. Chuit, R.J.P. Corriu, C. Reye and J.C. Young, Chem. Rev., 1993, 93, 1371. (b) S.E. Denmark and G.L. Beutner, Angew. Chem. Int. Ed., 2008, 47, 1560–1638. 35. (a) A.V. Malkov, A. Mariani, K.N. MacDougal and P. Kocovsk y, Org. Lett., 2004, 6, 2253. (b) A.V. Malkov, S. Stoncius, K.N. MacDougal, A. Mariani, G.D. McGeoch and P. Kocovsk y, Tetrahedron 2006, 62, 264. 36. A.V. Malkov, M. Figlus, S. Stoncius and P. Kocovsk y, J. Org. Chem., 2007, 72, 1315. 37. A.V. Malkov, M. Figlous and P. Kocovsky J. Org. Chem., 2008, 73, 3985. 38. Allylation: (a) S.E. Denmark, J. Fu and M.J. Lawler, J. Org. Chem., 2006, 71, 1523–1536; S.E. Denmark, J. Fu, D.M. Coe, X. Su, N.E. Pratt and B. D. Griedel, J. Org. Chem., 2006,


39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55. 56. 57. 58. 59. 60. 61. 62. 63. 64. 65. 66. 67. 68. 69. 70. 71.

339

71, 1513–1522; and references cired therein. Aldol condensation: (b) S.E. Denmark and T. Bui, J. Org. Chem., 2005, 70, 10393–10399. T. Oyama, H. Yoshioka and M. Tomoi, Chem. Commun., 2005, 1857–1859. For a recent study describing the determination of site–site distance and site concentration within polymer beads of different polystyrene-type polymers see: R. Marchetto, E.M. Cilli, G.N. Jubilut, S. Schreier and C.R. Nakaie, J. Org. Chem., 2005, 71, 4561–4568. R. Murugan and E.F.V. Scriven Aldrichimica Acta 2003, 36, 21. G. Priem, B. Pelotier, S.J.F. Macdonald, M.S. Anson and I.B. Campbell, J. Org. Chem., 2003, 68, 3844–3848. G. Priem, B. Pelotier, S.J.F. Macdonald, M.S. Anson and I.B. Campbell, Synlett 2003, 679–683. E. Alza, X.C. Cambeiro, C. Jimeno and M.A. Perocas, Org Lett., 2007, 9, 3717–3720. Y.-B. Zhao, L.-W. Zhang, L-Y. Wu, X. Zhong, R. Li and J.-T. Ma Tetrahedron Asymm., 2008, 19, 1352–1355. P. Li, L. Wang, Y. Zhang and G. Wang Tetrahedron, 2008, 64, 7633–7638. S. Luo, J. Li, L. Zhang, H. Xu and J.-P. Cheng Chem. Eur. J., 2008, 14, 1273–1281. S.J. Connon, Chem. Eur. J., 2006, 12, 5418–5427. Y. Takemoto, Org. Biomol. Chem., 2005, 3, 4299–4306. H. Miyabe, S. Tuchida, M. Yamauchi and Y. Takemoto, Synthesis, 2006, 3295–3300. (a) E.R. Jarvo and S.J. Miller. Tetrahedron, 2002, 58, 2481–2495. (b) W. Notz, F. Tanaka and C.F. Barbas, Acc. Chem. Res., 2004, 37, 580–5914. B. List, Synlett, 2001, 1675. M. Benaglia, G. Celentano, M. Cinquini, A. Puglisi and F. Cozzi, Adv. Synth. Catal., 2002, 344, 149–152. For an excellent review on supported proline and its derivatives see: M. Gruttadauria, F. Giacalone and R. Noto, Chem. Soc. Rev., 2008, 37, 1666–1688. B. List, R.A. Lerner and C.F. Barbas III, J. Am. Chem. Soc., 2000, 122, 2395. M. Benaglia, M. Cinquini, F. Cozzi, A. Puglisi and G. Celentano, Adv. Synth. Catal., 2002, 344, 533–542. Proline supported on 1% cross-linked polystyrene: (a) K. Kondo, T. Yamano and K. Takemoto Makromol. Chem., 1985, 186, 1781. Proline supported on silica gel column: (b) K. Sakthivel, W. Notz, T. Bui and C.F. Barbas III, J. Am. Chem. Soc., 2001, 123, 5260–5267. D. Font, A. Bastero, S. Sayalero, C. Jimeno and M.A. Pericas, Org. Lett., 2007, 9, 1943–1947. See also E. Alza, X.C. Cambeiro, C. Jimeno and M.A. Pericas, Org. Lett., 2007, 9, 3717–3721. D. Font, C. Jimeno and M.A. Pericas Org. Lett., 2006, 8, 4653–4656. D. Font, S. Sayalero, A. Bastero, C. Jimeno and M.A. Pericas, Org. Lett., 2008, 10, 337–341. G. Chouhan, D. Wang and H. Alper, Chem. Commun., 2007, 4809. F. Caldero`n, R. Fernandez, F. Sanchez and A. Fernandez-Mayoralas Adv. Synth. Catal., 2005, 347, 1395–1403. See also E.G. Doyaguez, F. Caldero`n, R. Fernandez, F. Sanchez and A. Fernandez-Mayoralas J. Org. Chem., 2007, 72, 9353. M. Gruttadauria, S. Riela, C. Aprile, P. Lo Meo, F. D’Anna and R. Noto Adv. Synth. Catal., 2006, 348, 82–92. See also F. Giacalone, M. Gruttadauria, A.M. Marculescu and R. Noto Tetrahedron Lett., 2007, 48, 255–259. W. Chen, Y. Zhang, L. Zhu, J. Lan, R. Xie and J. You, J. Am. Chem. Soc., 2007, 129, 13879. Z. Shen, Y. Liu, C. Jiao, J. Ma, M. Li and Y. Zhang, Chirality, 2005, 17, 556–558. A.S. Kucherenko, M.I. Struchkova and S.G. Zlotin, Eur. J. Org. Chem., 2007, 2000–2004. For a very recent contribution where guanidine-derived ionic liquids are used see: J. Shah, H. Blumenthal, Z. Yacob and J. Liebscher, Adv. Synth. Catal., 2008, 350, 1267. W. Miao and T.H. Chan, Adv. Synth. Catal., 2006, 348, 1711. M. Lombardo, F. Pasi, S. Easwar and C. Trombini, Adv. Synth. Catal., 2007, 349, 2061. (a) Z. Tang, Z. Yang, X.-H. Chen, L. Cun, A. Mi, Y. Jiang and L.-Z. Gong, J. Am. Chem. Soc., 2005, 127, 9285. (b) S. Samanta, J. Liu, R. Dodda and C.-G. Zhao, Org. Lett., 2005, 7, 5321. (c) J.-R. Chen, H. Lu, X.-Y. Li, L. Cheng, J. Wan and W.-J. Xiao, Org. Lett., 2005, 7, 4543. (a) G. Guillena, M. Hita and C. Najera, Tetrahedron Asymmetry, 2006, 17, 1027; (b) D. Gryko, B. Kowalczyk and L. Zawadzki, Synlett, 2006, 1059. (c) S. Guizzetti, M. Benaglia, L. Pignataro and

340

72. 73. 74. 75. 76. 77. 78. 79. 80. 81. 82. 83. 84. 85. 86. 87. 88. 89. 90. 91. 92. 93. 94. 95. 96. 97. 98. 99. 100. 101.

Recoverable and Recyclable Catalysts A. Puglisi, Tetrahedron Asymmetry, 2006, 17, 2754. (d) G. Guillena, M Hita and C. Najera, Tetrahedron Asymmetry, 2006, 17, 1493. S. Guizzetti, M. Benaglia, L. Raimondi and G. Celentano, Org. Lett., 2007, 9, 1247–1250. See also ref. 71d. M. Gruttadauria, F. Giacalone, A.M. Marculescu and R. Noto Adv. Synth. Catal., 2008, 350, 1397. For the use of prolinamides in ionic liquids see for example: H.-M. Guo, L.-F. Cun, L.-Z. Gong, A.-Q. Mi and Y.-Z. Jiang, Chem. Commun., 2005, 1450. B. Ni, Q. Zhang and A.D. Headley, Green Chem., 2007, 9, 737. L.S. Zu, J. Wang, H. Li and W. Wang Org. Lett., 2006, 8, 3077. For the use of a perfluorous diphenylprolinol silyl ether see L.S. Zu, J. Wang, H. Li, X.H. Yu and W. Wang Tetrahedron Lett., 2006, 47, 5131. K.A. Ahrendt, C.J. Borths and D.W.C. MacMillan, J. Am. Chem. Soc., 2000, 122, 4243–4244. M. Benaglia, G. Celentano, M. Cinquini, A. Puglisi and F. Cozzi, Adv. Synth. Catal., 2002, 344, 149–152. S.A. Selk€al€a, J. Tois, P.M. Pihko and A.M.P. Koskinen, Adv. Synth. Catal., 2002, 344, 941–945. M. Benaglia, G. Celentano, M. Cinquini, A. Puglisi and F. Cozzi, Eur. J. Org. Chem., 2004, 567–573. Y. Zhang, L. Zhao, S.S. Lee and J.Y. Ying Adv. Synth. Catal., 2006, 348, 2027. Q.L. Chu, W. Zhang and D.P. Curran Tetrahedron Lett., 2006, 47, 9287. For the use of MAcMillan catalyst in ionic liquids see: J.K. Park, P. Sreekanth and B.M. Kim Adv. Synth. Catal., 2004, 346, 49. T. Mitsudome. K. Nose, T. Mizugaki, K. Jitsukawa and K. Kaneda Tetrahedron Lett., 2008, 49, 5464. M.R.M. Andreae and A.P. Davis, Tetrahedron: Asymmetry, 2005, 16, 2487. P. Krattiger, R. Kovasy, J.D. Revell, S. Ivan and H. Wennemers, Org. Lett., 2005, 7, 1101. J.D. Revell, D. Gantenbein, P. Krattiger and H. Wennemers, Biopolymers, 2006, 84, 105. C. Aprile, F. Giacalone, M. Gruttadauria, A. Mossuto Marculescu, R. Noto, J.D. Revell and H. Wennemers, Green Chem., 2007, 9, 1328. S. Colonna, H. Molinari, S. Banfi, S. Julia and J. Guixer, Tetrahedron, 1984, 40, 5207. P.A. Bentley, S. Bergeron, M.W. Cappi, D.E. Hibbs, M.B. Hursthouse, T.C. Nugent, R. Pulido, S.M. Roberts and L. Wu, Chem. Commun., 1997, 739. P.A. Bentley, R.W. Flood, S.M. Roberts, J. Skidmore, C.B. Smith and J.A. Smith, Chem. Commun., 2001, 1616–1617. A. Berkessel, N. Gasch, K. Glaubitz and C. Koch, Org. Lett., 2001, 3, 3839–3841. D.R. Kelly, T.T.T. Bui, E. Caroff, A.F. Drake and S.M. Roberts, Tetrahedron Lett., 2004, 45, 3885–3888. K. Akagawa, H. Akabane, S. Sakamoto and K. Kudo Org. Lett., 2008, 10, 2035. C. Gennari and U. Piarulli, Chem. Rev., 2003, 103, 3071–3100. M.S. Sigman and E.N. Jacobsen, J. Am. Chem. Soc., 1998, 120, 4901–4902. Reviews: (a) G. Jas and A. Kirschning, Chem. Eur. J., 2003, 9, 5708–5723. (b) C. Wiles and P. Watts Eur. J. Org. Chem., 2008, 1655–1671. (c) B. Ahmed-Omer, J.C. Brabdt and T. Wirth Org. Biomol. Chem., 2007, 5, 733. For examples of organocatalysed reactions carried out under continuous flow mode see ref 27. and: (a) K. Ishiara, A. Hasegawa and H. Yamamoto, Synlett, 2002, 1296–1298. (b) M. Br€ unjes, G. Sourkouni-Argirusi and A. Kirschning, Adv. Synth. Catal., 2003, 345, 635–642. A.M. Hafez, A.E. Taggi, T. Dudding and T. Lectka, J. Am. Chem. Soc., 2001, 123, 10853–10859; and references cited therein. A. Hartung, M.A. Keane and A. Kraft, J. Org. Chem., 2007, 72, 10235. M. Ikunaka, Org. Process Res. Dev., 2007, 11, 495–502.

12 Organic Polymer-microencapsulated Metal Catalysts Jun Ou and Patrick H. Toy Department of Chemistry, The University of Hong Kong, Hong Kong, People’s Republic of China

12.1 Introduction In addition to the other methods for metal catalyst recovery and reuse discussed elsewhere in this volume, the concept of microencapsulation has also been studied. Microencapsulation is a technique whereby a substance to be delivered is physically enveloped by a carrier material and it has been widely used in various fields, especially in the context of drug

delivery.1,2 It should be noted that ‘micro’ in microencapsulation refers to the scale of capsule size and has no implications regarding the nature of the entrapped species. In fact, many of the microencapsulated catalysts discussed in this review are actually metal nanoparticles. This review focuses on the examples where aromatic group bearing organic polymers were used as the encapsulating material.3–7 Various organic polymers have been used to microencapsulate numerous metal catalysts and such microcapsules are easily prepared using polymers prepared from aromatic monomers by homogeneously mixing the metal catalyst and polymer together so that an association can be formed between p-electrons of the polymer aryl rings and the empty orbitals of the metal catalyst being enveloped. If the polymer used is initially non-crosslinked, this procedure is followed by heterogenization of the mixture by techniques such

The Journal of Microencapsulation is published bimonthly.


342


as precipitation or cross-linking to form the microcapsules. Such formed microencapsulated metal catalysts seem to exhibit several advantageous features compared with heterogeneous metal catalysts immobilized by other methods, including: (1) metal leaching and loss is often avoided or is only negligible, (2) catalyst activity in most cases is comparable with the unsupported counterparts, and they sometimes even exhibit higher activity, and (3) recovery and reuse of the catalysts is simple and highly efficient. These features combine to make microencapsulation a promising method for the immobilization of metal catalysts and this chapter surveys the microencapsulated catalysts prepared using various organic polymers based on p-electron–metal orbital interactions, with special attention paid to the issues of metal leaching and recyclability. Other methods for immobilizing metal catalysts onto polymers are not discussed, and organization of this section is based on the organic polymers used to microencapsulate the various metal catalysts rather than on the catalysts themselves or the reactions in which they were used.

12.2 Non-cross-linked Polymer-microencapsulated Catalysts Initial research into the concept of using aromatic group bearing organic polymers to microencapsulate metal catalysts was focused on using linear, non-cross-linked polymers that are soluble in some solvents, but insoluble in others. This property allowed the microcapsules to be prepared via a homogeneous interaction between the polymer and the metal catalyst that was followed by heterogenization. 12.2.1 Non-cross-linked Polystyrene Linear non-cross-linked polystyrene (NCPS), which is soluble in solvents such as dichloromethane, tetrahydrofuran and toluene, but insoluble in acetonitrile, methanol and water, was the first aromatic organic polymer studied and has been used to encapsulate numerous metal catalysts (Figure 12.1). Metals encapsulated by NCPS include scandium, osmium, palladium, ruthenium, vanadium, copper and bismuth. In 1998, Kobayashi and coworkers first introduced the concept of metal catalyst microencapsulation using an organic polymer by immobilizing the Lewis acid Sc(OTf)3 with NCPS.8 The encapsulated catalyst 1 was prepared by dissolving NCPS in

Figure 12.1 Non-cross-linked polystyrene-encapsulated metal catalysts

Organic Polymer-microencapsulated Metal Catalysts

343

Table 12.1

Reaction cycle

1

2a

3a

4a

5a

6a

7a

Yield (%)

90

90

88

89

89

88

90

a

Recovered 1 was used.

cyclohexane with powdered Sc(OTf)3 (5 : 1 by weight) and heating at 50–60 C for 1 h, followed by cooling slowly to 0 C. The thus formed microcapsules of 1 were found at this stage to be soft, but washing them with hexane and acetonitrile, solvents in which NCPS is insoluble, hardened them and allowed them to be easily handled. By using this procedure approximately 60% of the Sc(OTf)3 was incorporated into 1. The microcapsules were studied by scanning electron microscopy and energy dispersive x-ray analysis and it was found that the Sc(OTf)3 was dispersed on the polymer surface. The importance of the polymer p-electrons for encapsulating the catalyst was confirmed by experiments using polybutadiene or polyethylene as the encapsulating material. Polybutadiene incorporated only 43% of the Sc(OTf)3 that NCPS did, while polyethylene did not incorporate any. Catalyst 1 was then used in a wide variety of Lewis acid-catalyzed reactions for which Sc(OTf)3 was a known catalyst, including aldol, Michael addition and Friedel–Crafts acylation reactions in batch processes, and imino aldol and Mannich-type three-component coupling reactions in flow systems, amongst many others.5,8 In all of these reactions, 1 afforded good yield of the desired product and its excellent recyclability was demonstrated in the imino aldol reaction. As can be seen in Table 12.1, there was no observable decrease in the activity of 1 upon recovery and reuse, even after five reuses. Later Oriyama and coworkers reported the use of 1 in alcohol silylation reactions with methallylsilanes (Scheme 12.1).9 Interestingly, 1 exhibited higher catalytic activity than Sc(OTf)3 in most cases, and only 2 mol% of it was required for the reactions to go to completion in room temperature propionitrile after only 15 min.

Scheme 12.1

Catalyst 1 has also been used in the synthesis of D,L-a-tocopherol in 88% yield starting from trimethylhydroquinone and isophytol in the polar solvent propylene carbonate (Scheme 12.2).10 Unfortunately 1 could not be recovered and reused from this reaction and some Sc(OTf)3 leaching was observed.

344


Scheme 12.2

Kobayashi next used his microencapsulation technology to prepare NCPS microcapsules of OsO4 as a catalyst for the dihydroxylation of olefins using NMO as the stoichiometric oxidant.11 Catalyst 2 was prepared using a procedure that was similar to the one used to prepare 1 in which 90% of the OsO4 was encapsulated, and it was used to dihydroxylate many cyclic and acyclic olefins in high yield. Cyclohexene was chosen as the substrate to test the recyclability of catalyst 2 (Table 12.2), (5 mol%), and quantitative recovery was possible through five reaction cycles with no decrease in catalytic activity or leaching observed. Catalyst 2 was subsequently used in the structure determination and synthesis of a variety of natural products.12–16 For example, Ohta and coworkers used 2 to dihydroxylate an exocyclic double bond of a synthetic intermediate in order to install a tertiary alcohol group in the first total synthesis of ()-linderol A (Scheme 12.3).15

Scheme 12.3

Table 12.2

Reaction cycle

1

2a

3a

4a

5a

Yield (%)

84

84

83

84

83

a



345

Another synthetic application of 2 was reported by Oshima and coworkers where it was used in the deprotection of an allyl ether that was impervious to more standard palladium-catalyzed cleavage conditions (Scheme 12.4).16 The reaction sequence of isomerization with Wilkinson’s catalyst and DABCO followed by oxidation with 2 and NMO removed the allyl ether group to afford furanodictine A.

Scheme 12.4

Next Kobayashi and coworkers described the preparation of NCPS-microencapsulated palladium catalyst 3.17 Catalyst 3 was prepared by mixing NCPS and Pd(PPh3)4 in a 5 : 1 mass ratio in cyclohexane at 40 C followed by cooling to 0 C and the addition of hexane. Washing of 3 with acetonitrile resulted in the recovery of three equivalents of PPh3, indicating that only one equivalent of PPh3 remained in the microcapsules of 3, which was confirmed by 31 P swollen-resin magic angle spinning (SR-MAS) NMR spectroscopy. Interestingly, when 3 was applied in an allylation reaction, no reaction occurred unless one equivalent of PPh3 was added (Scheme 12.5). The authors proposed that an initial 18-electron Pd(0) species in 3 is converted to a 14- or 16-electron Pd(0) species upon

Scheme 12.5

346


addition of the phosphine ligand that is catalytically active. In addition to allylation reactions, 3 was found to be a good catalyst for Suzuki–Miyaura cross-coupling reactions as well when the phosphine ligand P(o-Tol)3 was added. The recyclability of 3 was excellent in both types of reactions, and the requirement of an external ligand for catalytic activity provided the opportunity for conducting asymmetric catalysis by using chiral ligand. For instance, addition of chiral phosphine ligand A allowed for an allylation product to be obtained in 87% yield with 83% ee. More recently Kobayashi and coworkers prepared microencapsulated ruthenium complexes 4 and 5.18 The authors first treated [Ru(h6-C6H5CO2Et)Cl2]2 with PPh3 to afford Ru(h6-C6H5CO2Et)(PPh3)Cl2 quantitatively, and this was used to prepare encapsulated complex 4 by reaction with NCPS in a procedure similar to that which was used previously for the preparation of 1. In order to convert 4 into ring-closing methathesis (RCM) catalyst 5, it was reacted with PCy3, 1,1-diphenyl-2-propynol, and NaPF6 (Scheme 12.6). Catalyst 5 was then used in a variety of high yielding 5-, 6and 8-membered ring forming RCM reactions (Scheme 12.7). However, in order for it to be effectively recycled, 5 required reactivation with additional PCy3 and 1,1-diphenyl2-propanol prior to reuse. Importantly, no leaching of Ru was observed in any of the reactions according to x-ray fluorescence analysis. Additionally, 4 was used to effectively catalyze acetophenone reduction and a dienylalkyne cyclization reaction (Scheme 12.7).

Scheme 12.6

Lattanzi and Leadbeater have prepared microencapsulated VO(acac)26 using Kobayashi’s methodology and used it to catalyze a wide variety of allylic alcohol epoxidation reactions with tert-butyl hydroperoxide (TBHP) as the stoichiometric oxidant.19 The recyclability of 6 was examined with geraniol as the substrate (Table 12.3), and the activity of catalyst was maintained through three reuses. Metal leaching in these reactions was determined by ICP analysis and found to be < 0.1% for the first use of 6 and decreased with each reuse.


347

Scheme 12.7

In 2003 Kantam and coworkers reported the microencapsulation of Cu(acac)2 by NCPS to form of 7, which was applied in olefin aziridination reactions using PhI ¼ NTs.20 Catalyst 7 was found to exhibit similar activity to unsupported Cu(acac)2 (Scheme 12.8), and had the advantage that it could be recovered and reused at least four times without any observable decrease in catalytic activity. Metal leaching was measured by atomic absorption spectroscopy and found to be negligible (0.025 wt%) when PhI ¼ NTs was used as the nitrogen source, but very significant when chloramine-T (16.56 wt%) and bromamine-T (10.28 wt%) were used instead. Table 12.3

Reaction cycle

1

2a

3a

4a

Yield (%)

93

84

83

81

a


348


Scheme 12.8

Choudary and coworkers reported the preparation of NCPS-microencapsulated Bi(OTf)3 8 and used this as a heterogeneous catalyst for aldehyde allylation, Michael addition, alcohol acylation, Baeyer–Villiger oxidation and aldol condensation reactions (Scheme 12.9).21 All of the expected products could be isolated in high yield with essentially no Bi(OTf)3 contamination, and 8 could be easily recovered. Later, Sreedhar and coworkers found catalyst 8 to be efficient catalysts for methoxymethylation of alcohols and carboxylic acids using dimethoxymethane with less than 1 ppm contamination by bismuth (Scheme 12.9).22

Scheme 12.9

At almost the same time that Kobayashi reported the synthesis and use of 4 and 5, Gibson and Swamy independently described the microencapsulation of ruthenium complex B by


349

NCPS to form 9 (Scheme 12.10).23 The procedure for preparation of microencapsulated 9 involved the standard mixing of NCPS and B, followed by precipitation. Although 9 efficiently catalyzed the RCM of N,N-diallyl-p-toluensulfonamide, and several related dienes, its recyclability was poor, with the yield dropping from 92% in the first reaction to 40% in the fourth cycle.

Scheme 12.10

The first example of the microencapsulation of a chiral organometallic complex with NCPS was reported by Mayoral et al. where they immobilized Ru-pybox C with NCPS and used this as a catalyst in asymmetric cyclopropanation reactions (Scheme 12.11).24 A series of related microencapsulated catalysts 10 was prepared by using mixtures of cyclohexane and CH2Cl2 to dissolve C and NCPS. Subsequent solvent removal afforded 10. In this work 10 was used as a homogeneous catalyst and the cyclopropanation reactions were performed in CH2Cl2. After the completion of reaction, hexane was added to precipitate 10 for recycling. Using optimized reaction conditions, 60–68% yield could be achieved with enantioselectivities in the range of 75–85% ee for the major trans product. Unfortunately significant amounts of Ru leached in these homogeneous reactions using reversibly encapsulated 10. It was found that overall 15% of the Ru leached from 10 after four cycles. Finally, Naik et al. have used NCPS to microencapsulate a series of metallophthalocynines and metalloporphyrins to generate heterogeneous catalysts for numerous oxidation reactions, such as aerobic alcohol and alkene oxidation processes. The NCPS-microencapulated

350


Scheme 12.11

catalysts were found to be stable and more active than their nonencapsulated counterparts. A series of microencapsulated iron–, cobalt– and copper–phthalocyanine complexes were used to catalyze aerobic alcohol oxidation reactions25 and a microencapsulated manganese– porphyrin catalyst was used to oxidize several styrenes with NaIO4, KHSO5 or NaOCl as the stoichiometric oxidant.26 Furthermore encapsulated cobalt– and iron–porphyrin complexes were found to also be efficient at catalyzing aerobic alcohol oxidation reactions.27 12.2.2 Non-cross-linked Polystyrene Derivatives Initial attempts to use 2 with a chiral ligand in asymmetric Sharpless dihydroxylation reactions were not very successful. Thus Kobayashi and coworkers examined alternative polymers to NCPS for microencapsulating OsO4 and one promising material identified was poly(acrylonitrile-co-butadiene-co-styrene).28 Use of this polymer to microencapsulate OsO4 resulted in the formation of 11 (Figure 12.2). Initial studies with 11 and NMO resulted in good yields of dihydroxylated alkenes such as styrene (Scheme 12.12), with the catalyst being recovered quantitatively and readily reusable with no decrease in activity.

Figure 12.2 Poly(acrylonitrile-co-butadiene-co-styrene)-encapsulated OsO4

Scheme 12.12


351

Table 12.4

Reaction cycle 1 2a 3a 4a 5a a

Yield (%)

ee (%)

88 75 97 81 88

84 95 94 96 95

Recovered 11 and (DHQD)2PHAL were used.

With these promising results, 11 was then examined in asymmetric dihydroxylation of olefins according to Sharpless’s procedure using 1,4-bis(9-O-dihydroquinidinyl) phthalazine ((DHQD)2PHAL) as the source of chirality and NMO as the stoichiometric oxidant (Table 12.4). Optimized reaction conditions included adding the alkene substrate slowly over 24 h to a mixture of 5 mol% 11, (DHQD)2PHAL, and NMO, with the catalyst being recovered quantitatively by simple filtration, and the chiral ligand being recovered efficiently by acid–base extraction. The recovered 11 and (DHQD)2PHAL could be reused several times with an unexplained increase in enantioselectivity. These reaction conditions were applied to the asymmetric oxidation of a range of di- and tri-substituted alkenes with good results, and were even used successfully on a 100 mmol scale. One drawback of this procedure was the need to add the substrate slowly over 24 h. Thus the Sharpless two-phase dihydroxylation conditions, in which K3Fe(CN)6 serves as the oxidant, were studied using 11. Unfortunately this procedure resulted in some OsO4 leaching from 11. Furthermore, SR-MAS NMR analysis of recovered 11 indicated that the polymer backbone alkene groups, resulting from the polymerization of butadiene, were dihydroxylated during microcapsule preparation and this complicated recycling. Subsequently another NCPS derivative containing phenoxyethoxymethyl groups was synthesized and used to microencapsulate OsO4 (Scheme 12.13).29

Scheme 12.13

With this new polymer, microencapsulated catalyst 12 was prepared using the standard procedure used for the preparation of 1. Optimized reaction conditions for asymmetric

352


dihydroxylation reactions catalyzed by 12 were found to be 5 mol% of 12 and (DHQD)2PHAL with 2.0 equivalents of K3Fe(CN)6 and K2CO3 being added at both the beginning and the halfway point of the reaction (Scheme 12.14). Gratifyingly this catalyst and reaction procedure eliminated the need for slow alkene substrate addition, and a range of alkenes could be enantioselectively dihydroxylated in good yield with no metal leaching detected by x-ray fluorescence analysis.

Scheme 12.14

Subsequently, Kobayashi and coworkers further improved the use of 12 by conducting the dihydroxylation reactions in water.30 In these aqueous reactions the surfactant Triton X-405 was added to the reaction mixture, and styrene could be dihydroxylated in 76% yield and 74% ee (Scheme 12.15). Furthermore, under these reaction conditions 12 could be recovered and reused without loss of activity or enantioselectivity.

Scheme 12.15

More recently Toy et al. applied phosphine group-functionalized NCPS (D) to prepare encapsulated catalyst 13 for use in Suzuki–Miyaura cross-coupling reactions (Scheme 12.16).31 This work was inspired by the observation of Kobayashi that catalyst 3 required the addition of an external phosphine ligand for it to be effective in catalyzing Suzuki–Miyaura cross-coupling reactions. Since Toy had previously reported the synthesis of D, it was thought that it could be used to prepare a microencapsulated palladium catalyst that did not require any additional phosphine ligand for catalysis. Gratifyingly 13 showed high catalytic activity in Suzuki–Miyaura cross-coupling reactions of a variety of aryl halides and arylboronic acids. The biaryl compounds could be obtained in up to 98% yield and after completion of the reaction, the heterogeneous catalyst could be easily removed by simple filtration.


353

Scheme 12.16

Table 12.5

Reaction cycle No substrate 1 2a 3a 4a a

Time (h)

Yield (%)

Pd leached (ppm)

3 3 3 3 3

0 91 83 79 66

0.0186 0.2137 0.1097 0.0644 0.1375


The recyclability of catalyst 13 and metal leaching from it were also studied using the model cross-coupling of iodobenzene and phenylboronic acid. It was found that the catalytic activity of the catalyst did not decrease significantly even after three cycles and that only minimal leaching was detectable by ICP analysis (Table 12.5). 12.2.3 Polysulfone With Kobayashi’s previously reported work regarding encapsulated OsO4 in asymmetric dihydroxylation reactions using various polymers as inspiration, Venkateswarlu and coworkers described the use of a polysulfone as the encapsulating material.32 Unfortunately the exact structure of the polymer used was not presented in the manuscript. Nevertheless, the polysulfone-encapsulated OsO4 catalyst 14 was prepared and tested in asymmetric dihydroxylation reactions using 5 mol% of (DHQD)2PHAL. A variety of alkenes were oxidized with 14 to afford the corresponding chiral diol in 82–94% yield and 92–99% ee when NMO was used as the stoichiometric oxidant. Furthermore, the recyclability of the 14 was studied in the asymmetric dihydroxylation of trans-stilbene (Table 12.6). Results obtained indicated that 14 could be recovered and reused at least four times without loss of activity or enantioselectivity. A reported improvement of this catalyst system compared with previous ones is that the reactions only required 20–30 min in order to proceed to completion.

354


Table 12.6

Run 1 2a 3a 4a 5a a

Time (min)

Yield (%)

ee (%)

25 30 30 30 30

94 94 92 90 86

96 97 96 97 97


12.2.4 Poly(xylylviologen dibromide) Another non-cross-linked aromatic polymer used for metal catalyst microencapsulation is poly(xylylviologen dibromide), and this work was pioneered by Uozumi and coworkers. The first example of this concept was the preparation and use of a palladium catalyst dubbed nano-Pd-V (15) (Scheme 12.17).33 Poly(xylylviologen dibromide) (E) was mixed with PdCl2 to afford an insoluble palladium complex that was subsequently reduced to form 15.

Scheme 12.17

The synthetic utility of 15 was then investigated in reactions involving the a-alkylation of ketones with primary alcohols in the presence of Ba(OH)2, and ring-opening alkylation of cyclic 1,3-diketones with primary alcohols in the presence of Sr(OH)2 (Scheme 12.18).24 Good yield of the desired product was obtained in all of the reactions studied and the catalyst could be reused effectively.

Scheme 12.18


355

An extension to this concept involving microencapsulation of a tungsten catalyst was subsequently reported.35 Catalyst 16 was prepared by simply mixing amphiphilic pyridinium polymer F and H3PW12O40 in water for 3 days (Scheme 12.19). Catalyst 16 was applied in oxidative cyclization reactions of various alkenols with the expected products being isolated in excellent yield, and it was reported that the catalyst could be recovered and reused at least four times without loss of activity, even after prolonged storage.

Scheme 12.19

12.3 Cross-linked Polymer-microencapsulated Catalysts Although microencapsulation of metal catalyst by non-cross-linked polymers has achieved great successes, this strategy seems to have limited applications. It is generally assumed, and perhaps experimentally demonstrated, that the microcapsules must be used as heterogeneous catalysts in order to minimize metal leaching from them, and thus reaction solvent choice is somewhat limited. As a means to circumvent this issue, the use of crosslinked polymers that are permanently heterogeneous regardless of the solvent have been examined recently. It should be noted that some of these polymers are cross-linked to begin with, while others are cross-linked after catalyst microencapsulation, and both of these strategies are discussed. 12.3.1 Divinyl Benzene Cross-linked Polystyrene Kobayashi and coworkers reported the use of polystyrene resin cross-linked with divinyl benzene (DVB-PS) to encapsulate OsO4 by mixing these two materials together in warm THF, followed by the addition of MeOH to form 17.36 In this manner 85% of the OsO4 used could be microencapsulated by the polymer. The authors examined polystyrene crosslinked with varying amounts of DVB and found that 1% DVB-PS seemed to be optimal, since less DVB resulted in microcapsules that were not very physically stable and more DVB led to reduced catalytic activity of 17. When 17 was used to catalyze asymmetric alkene dihydroxylation reactions, good yields of products with high ee were obtained (Scheme 12.20). Importantly the substrate 1-phenyl-1-cyclohexene, for which catalyst 12 did not work not very well (26% yield, 86% ee), could be oxidized smoothly to 1-phenylcyclohexane-1,2-diol in 82% yield and 94% ee. Finally, the recyclability of 17 was examined with styrene as the substrate, and it was found that it could be used at least five times without loss of reactivity and no detectable metal leaching.

356


Scheme 12.20

Saladino and coworkers have also studied the use of DVB-PS as an encapsulating material in the preparation of immobilized methylrhenium trioxide (MTO) catalyst 18.37,38 This was found to be a highly active catalyst in H2O2-mediated alkene oxidation reactions. However selectivity was low in some cases, with an almost 1 : 1 ratio of trans-1,2cyclohexanediol being recovered with the corresponding epoxide (Scheme 12.21). Subsequently, 18 was used in the oxidation of cardanol derivatives,39 the oxidation of N,N-disubstituted hydroxylamines to nitrones,40 and glycal epoxidation–methanolysis

Scheme 12.21


357

reaction sequences.41 In the first two cases, H2O2 was the stoichiometric oxidant, while the carbohydrate reactions utilized urea hydrogen peroxide (UHP) in this role. In these reports, the issue of metal leaching from the microcapsules was not discussed. Tentagel, a DVB-PS resin containing poly(ethylene glycol) (PEG) grafts, is another heterogeneous polymer that has been used to encapsulate metal catalysts. In this case, metal nanoparticles were prepared by the reduction of polymer-supported metal complexes in such a way that the resulting metal particles were dispersed in the heterogeneous polymer. For the synthesis of encapsulated palladium nanoparticle catalyst 19, which Uozumi referred to as an amphiphilic resin-dispersion of nanoparticles of palladium (ARP-Pd), a dipyridyl complex was used as the starting material and benzyl alcohol was the reducing reagent (Scheme 12.22).42 Catalyst 19 had a loading level of 0.37 mmol of palladium per gram and was used in the aqueous aerobic oxidation of a variety of alcohols. Recycling of 19 was examined in the oxidation of cyclooctanol and no decrease in activity was observed through four reaction cycles. Subsequently 19 was used to catalyze aqueous alkene hydrogenation and chloroarene dehalogenation reactions, with good recyclability.43,44

Scheme 12.22

Uozumi and coworkers have subsequently reported the preparation and use of the analogous encapsulated platinum catalyst 20, which was prepared in a similar fashion (Scheme 12.23).45 This catalyst also showed good activity and recyclability in aqueous aerobic alcohol oxidation reactions.

Scheme 12.23

12.3.2 Oligo(ethylene glycol) Cross-linked Polystyrene Building upon their earlier success in developing NCPS-microencapsulated catalysts, Kobayashi and coworkers have more recently been extending this work to the methodology that they refer to as ‘polymer incarceration’ (PI). The PI concept is based on same principles as NCPS catalyst microencapsulation, but includes a final cross-linking step in microcapsule synthesis. For example, the first report of this concept involved the use of polymer G to prepare cross-linked microcapsules containing palladium (Scheme 12.24).46 After initial microcapsule formation the alcohol and epoxide groups of G were reacted together by heating at 120 C to form cross-links between the polystyrene chains in 21. Importantly, the incarceration of Pd(PPh4) by this method resulted in the isolation of 4 equivalents of PPh3 from the washings, indicating that it is a Pd(0) species that is microencapsulated.

358


Scheme 12.24

PI Pd(0) (21) was first used in the hydrogenation of benzalacetone where, the reduction product 4-phenylbutan-2-one was obtained in high yield, and the catalyst exhibited good recyclability, even after five uses. Allylic substitution reactions were also investigated using 21, but these required the addition of PPh3 as a ligand (Scheme 12.25).

Scheme 12.25

Later the utility of 21 in Suzuki–Miyaura cross-coupling reaction was also reported by the same group.47 Ligand, base and solvent screening identified a combination of P(o-Tol)3 and K3PO4 in a refluxing toluene–water mixture to be the optimal reaction conditions for using catalyst 21 in this application (Scheme 12.26). Using these conditions various aryl halides were cross-coupled with aryl boronic acids to produce the corresponding biaryl compounds in high yields with no leaching of Pd observed. The recyclability of 21 was also found to be excellent in the reaction between 2-methylbromobenzene and phenyl boronic acid (Table 12.7). When used in conjunction with Buchwald’s sterically hindered biphenyldicyclohexylphosphine, 21 was also demonstrated to catalyze aryl halide amination reactions.48

Scheme 12.26


359

Table 12.7

Reaction cycle

1

2a

3a

4a

5a

Yield (%)

83

88

85

85

83

a

Recovered catalyst 21 was used.

Although 21 proved to be an effective heterogeneous catalyst for hydrogenation reactions, it was assumed that the benzyl ether moieties of polymer G would be unstable under harsh reduction conditions, such as at high hydrogen pressure and temperature. Therefore analogue 22 was prepared using polymer H, which does not possess labile groups (Scheme 12.27), and this was used in a range of hydrogenation reactions.49 Significantly, catalyst 22 was found to be sulfur tolerant and in some cases, to be more active in hydrogenation reactions than 21. This catalyst was also able to reduce substrates such as phenanthrene and benzothiophene at elevated temperature and hydrogen pressure.

Scheme 12.27

A polymer-incarcerated palladium catalyst has also been applied in amidocarbonylation reactions.50 Initially 21 was examined as a catalyst in such reactions, however significant metal leaching was observed in the amide solvents that provided the best yields. Thus polymer I was designed that possessed solvent mimicking amide groups for catalyst

360


microencapsulation (Scheme 12.28). This was used to prepare catalyst 23 using a procedure similar to what was used to prepare 21 and 22. Gratifyingly 23 proved to be an excellent catalyst in the amidocarbonylation reactions with no leaching, thereby validating their hypothesis, and moderate to excellent yields were obtained of the desired N-acyl-a-amino acids.

Scheme 12.28

A subsequent report described how replacing polymer G with J (Figure 12.3), in which the epoxide groups are moved from the pendant aryl rings to the polymer backbone, in the incarceration process allowed for encapsulated palladium clusters on the sub-nanometer scale to be formed in catalyst 24.51 Locating both the relatively polar epoxide and tetraethylene glycol groups together allowed for polymer micelles to form, which in turn allowed for the metal clusters in 24 to average seven atoms in size. Catalyst 24 was subjected to a three-phase test involving a polymer-supported reaction substrate and the results indicated that 24 does indeed act as a heterogeneous catalyst, since the expected Heck reaction product was only produced in moderate yield by the

Figure 12.3 Polymer J used for polymer incarceration of palladium


361

homogeneous catalyst Pd(PPh3)4 (Scheme 12.29). When the reaction substrates were homogeneous, 24 was an efficient catalyst in hydrogenation and Heck cross-coupling reactions.

Scheme 12.29

The use of phosphinated polymers for the preparation of incarcerated palladium catalysts has also been explored. As mentioned previously, one of the drawbacks of 21 as a catalyst for Suzuki–Miyaura cross-coupling reactions is the requirement for the addition of an external phosphorous ligand. In order to avoid this issue polymer K was designed for the preparation of incarcerated catalyst 25 that would not need any external ligand (Scheme 12.30).52 A notable difference in the preparation of 25 compared with that of 21 is a reduction reaction using HSiCl3 after the cross-linking operation.

Scheme 12.30

As expected, catalyst 25 showed excellent activity in Suzuki–Miyaura cross-coupling reactions without the need for an external phosphine ligand (Scheme 12.31). Moreover no leaching of palladium was observed in these reactions, and the catalyst could be recovered quantitatively by filtration and reused without loss of activity (Table 12.8).

Scheme 12.31

362


Table 12.8

1

2a

3a

4a

5a

quant

quant

98

99

quant

Reaction cycle Yield (%) a


Catalyst 25 was also used in the partial hydrogenation of alkynes.53 Initially diphenylacetylene was chosen as substrate and it was observed that 25 could produce the semi-hydrogenated product in good yield with high cis selectivity (Scheme 12.32). On the other hand the use of 21 as the catalyst produced predominantly the corresponding alkane. It was proposed by the authors that the phosphine groups immobilized on the polymer may serve as weak poisoning agents to allow 25 to act as an analogue of the Lindlar catalyst system. Significantly, kinetic studies indicated that 25 could produce cis-alkenes without strict control of H2 consumption.

Scheme 12.32

Given the wide range of phosphines used in palladium catalysis, it is perhaps not surprising that in addition to K, other phosphine-functionalized polymers have been used to prepare incarcerated palladium catalysts, such as L, M and N (Figure 12.4).54,55 Starting

Figure 12.4 Phosphine-functionalized polymers used for polymer incarceration


363

Table 12.9

Reaction cycle

1

2a

3a

Yield (%)

90

92

81

a

Catalyst 26 was reduced with HSiCl3 after recovery and before reuse.

with these polymers, catalysts 26, 27 and 28 were prepared, respectively, using the same procedure outlined in Scheme 12.30. Catalyst 26, prepared from polymer L, exhibited high activity in aryl halide amination reactions (Scheme 12.33).54 Both aryl iodides and bromides with electron-donating groups afforded the desired amine products in good yields, and several types of secondary amines could be used. In all cases, no leaching of palladium was observed. For these reactions, it was found that treatment of recovered 26 with HSiCl3 prior to reuse was necessary for high yield with the recycled catalyst (Table 12.9). This was reportedly due to phosphine group oxidation and palladium cluster aggregation.

Scheme 12.33

Catalyst 27, prepared from polymer M, was observed to be an efficient catalyst in Sonogashira cross-coupling reactions (Scheme 12.34).54 Aryl iodides with electronwithdrawing groups afforded coupled products in good yields without leaching of palladium being detected. On the other hand, some metal leaching was observed in the reaction between 1-iodonaphthalene and phenylacetylene.

Scheme 12.34

364


Table 12.10

Reaction cycle

1

2a

3a

Yield (%)

87

89

95

a


As previously mentioned, 26 proved to be a good catalyst for amination reactions involving aryl bromides and iodides. However, when it was applied in reactions with aryl chloride starting materials, no desired amine product was produced. Therefore a catalyst 28, prepared from polymer N bearing biphenyldicyclohexylphosphine groups, was prepared and studied in these reactions.55 Gratifyingly, the use of the sterically hindered phosphine groups in 28 allowed it to function catalytically as desired (Scheme 12.35). The recyclability of 28 was also examined, and it could be recovered quantitatively by simple filtration and used at least three times (Table 12.10).

Scheme 12.35

Another major advancement in the preparation of incarcerated palladium catalysts was the switch to a relatively less expensive Pd(II) starting material to replace the original more expensive Pd(0) species previously used. It was found that Pd(NO3)2 was a suitable replacement for Pd(PPh3)4 that was used in the initial preparation of 21.56 Experimentation demonstrated that thermal decomposition of Pd(OAc)2 in the presence of alkali metal salts allowed for the preparation of 29 using polymer J without the need for any additional reducing agent (Scheme 12.36). It was found that NaOAc was the best salt for this purpose and 29 prepared with it was used to efficiently catalyze Suzuki–Miyaura cross-coupling reactions. The expected biaryl compounds could be obtained in high yields with no metal leaching observed.

Scheme 12.36


365

In addition to the various versions of incarcerated palladium catalysts mentioned, an incarcerated Sc(OTf)3 Lewis acid catalyst has been prepared with polymer J.57 It was found that the solvent choice for microencapsulation was critical, and a mixture of cyclohexane and THF was optimal (Scheme 12.37).

Scheme 12.37

The morphology of the resulting 30 was studied by transmission electron microscopic analysis, and it was observed that spherical micelles were formed when THF–cyclohexane was chosen as the solvent for catalyst preparation. Lewis acid 30 was used effectively to catalyze Mukaiyama aldol, Mannich and Michael addition reactions (Scheme 12.38).

Scheme 12.38

Polymers G and J were used for preparation of polymer-incarcerated ruthenium catalysts 31 and 32 by the procedure depicted in Scheme 12.39.58 These catalysts were used in the oxidation of alcohols and sulfides, and generally 32 was more active than the 31, most likely due to the greater surface area provided by polymer J compared with polymer G. Regardless, good recyclability with no metal leaching was observed with both of these catalysts. Subsequently, an analogue of 31 was prepared from polymer G and RuCl3 H2O as the metal source, and the resulting catalyst 33 was used in alcohol oxidation reactions (Scheme 12.40).59 It was observed that in the presence of a combination of 33 and TEMPO, alcohols were oxidized using molecular oxygen at atmospheric pressure to afford the corresponding aldehyde or ketone in good to excellent yield. Once again, the authors

366


Scheme 12.39

Scheme 12.40

reported that the catalyst could be recovered and reused numerous times without loss of activity. Incarceration of platinum was initially achieved in a manner similar to the original procedure for palladium incarceration by using Pt(PPh3)4.60 Thus catalyst 34 could be prepared using polymer G, and used in alkene and alkyne hydrosilylation reactions at low catalyst loading levels.

Scheme 12.41

Later, polymer J was used for the preparation of incarcerated platinum catalyst 35 by using PtCl2(COD) as the metal source and HSiEt3 as the reducing reagent (Scheme 12.42).61


367

This catalyst was successfully applied in a wide variety of catalytic hydrogenation reactions with no metal leaching detected.

Scheme 12.42

Finally and most recently, a method has been developed for the preparation of microencapsulated gold catalyst 36 for aerobic oxidation of alcohols by polymer incarceration.62 The preparation of 36 involved the use of polymer O as outlined in Scheme 12.43. A solution of AuClPPh3 in a small amount of diglyme was slowly added to a mixture of O and NaBH4 in diglyme at room temperature. Ether was then added to form microcapsules, which were filtered, washed, and dried. The microcapsules were next heated at 150 C for 5 h in order to cross-link the polymer side chains and then washed with THF and water to afford 36, which proved to be a good catalyst for aerobic oxidation of numerous alcohols in various solvents62 and hydroquinones.63

Scheme 12.43

12.3.3 Urea Group Cross-linked Polyphenylene In addition to the previously discussed cross-linked polymers that have been used to encapsulate metal catalysts, the Ley group has recently reported the utilization of polyphenylene cross-linked by urea groups for the preparation of such materials.6 These microcapsules were prepared using an interfacial polymerization technique that involves dispersing an organic phase containing the material being encapsulated and a polyphenylene polymer bearing multiple isocyanate groups into an aqueous phase containing colloid

368


stabilizers, dispersants, salts and chain extenders (Scheme 12.44). Upon dispersion of the organic phase into the aqueous phase, the reactive isocyanate groups at the oil–water interface undergo spontaneous in situ urea group formation, which cross-link the polyphenylene backbone and thus form the microcapsule walls that entrap the catalyst. The first such material prepared was a palladium(II) species that was referred to as Pd(II) EnCat (37).64

Scheme 12.44

Catalyst 37 was studied by various analytical techniques, including x-ray fluorescence and ICP, to determine that it possessed a loading level of 0.4 mmol of Pd/g, and it was tested in Suzuki–Miyaura cross-coupling reactions in which the desired biaryl compounds could be obtained in high yield (Scheme 12.45). The catalyst could be recovered and reused for four times without loss of activity and only less than 0.2% of metal leaching was observed.

Scheme 12.45

They next extended the applicability of 37 by demonstrating its utility in carbonylation, Heck, Suzuki–Miyaura, and Stille cross-coupling reactions in both conventional organic solvents and supercritical carbon dioxide (scCO2) (Scheme 12.46).65 In most cases, the 37-catalyzed reactions afforded the desired product in high yield with a low level of metal leaching into reaction mixture. The recycability of 37 was studied in the Stille cross-coupling of 4-nitrobromobenzene and phenyltrimethyltin (Table 12.11). Recovered 37 could be reused with minimal loss of reactivity, however the subsequent reactions required longer time to reach completion. In order to assess the practical utility of 37 in industrial production, continuous flow Suzuki–Miyaura cross-coupling reactions were studied.66 Reaction conditions were first optimized in batch mode reactions using a variety of ammonium salt bases, in both organic solvent and scCO2 (Scheme 12.47). The reactions were conducted using two methods, which included high temperature (110 C) for a relatively shorter reaction time (21 h), and


369

Scheme 12.46

at lower temperature (40 C) for a relatively longer reaction time (2.5 d). For the reactions performed in a toluene–methanol mixture both conditions furnished 4-methylbiphenyl in high yield, whereas when scCO2 was used as the reaction media, high yield could only be obtained at high temperature.

Scheme 12.47 Table 12.11

Reaction cycle 1 2a 3a 4a a


Yield (%)

Time (h)

99 98 99 97

3 4.5 12 24

370


Table 12.12

GC yield (%) Entry

n-Bu4NX

T ( C)

1 2 3 4 5

OAc OH OMe F OMe

55 55 55 55 70

Cycle 1

Cycle 2

Cycle 3

Cycle 4

Cycle 5

Cycle 6

4.7 3.1 12 36 100

7.5 1.3 48 48 —

11 70 85 55 —

12 — — — —

14 — — — —

15 — — — —

With this understanding, continuous flow Suzuki–Miyaura coupling reactions were performed by passing a stock reagent solution containing iodobenzene, p-tolylboronic acid and n-Bu4NX (X ¼ OAc, OH, OMe or F) through an HPLC column packed with the 37 (Table 12.12), and the cumulative yield of 4-methylbiphenyl was determined using gas chromatography. The best result was obtained when n-Bu4NOMe was used as the base at slightly elevated temperature, where 4-methylbiphenyl was formed in 100% yield simply by passing the reaction mixture through the HPLC column packed with the 37 a single time (entry 5). When a mixture of scCO2 and MeOH was used as the solvent, results were less satisfactory [68]. Next, microwave irradiation of Suzuki–Miyaura cross-coupling reactions catalyzed by 37 were studied.68,69 The methodology was used in batch mode for compound library preparation and in continuous-flow applications to produce multigram quantities of product. In the initial batch mode experiments, various biaryl compounds could be obtained in high yield. Encouraged by theses results, 11 boronic acids and 31 aryl halides and triflates were used to synthesize a potential array of 341 compounds. This resulted in 131 products being produced with greater than 98% purity, 40 products produced with between 80% and 98% purity and 55 produced with less than 80% purity. The remaining 115 possible compounds were only incompletely formed, or not produced at all. The purity levels reported were achieved by simple work-up and filtration, with no chromatography necessary. In the continuous flow reactions involving microwave irradiation, products were generally formed in higher yields and purities. An additional Suzuki–Miyaura cross-coupling protocol involving catalyst 37, that affords reaction products in high purity without the need for chromatography or aqueous work-up, was reported by Ley and coworkers.70 This methodology involved either thermal heating or microwave irradiation, and the use of a polymer-supported base (MP-carbonate) to initiate the reactions and a polymer-supported scavenging reagent (PS-DEAM) to remove any unreacted boronic acid (Scheme 12.48). A variety of aryl halides and boronic acids, including some bearing aldehyde, ketone and nitrile groups, could be smoothly converted into the expected biaryl compounds in high yield and purity, with the two different heating methods affording comparable results.


371

Scheme 12.48

As described the polyphenylene–polyurea-encapsulated catalyst 37 and related materials have been successfully applied in a variety of palladium-catalyzed cross-coupling reactions. However, the true nature of their catalysis was not established until a seminal report by Broadwater and McQuade.71 They described the study of a range of Pd EnCat materials in three-phase tests involving Heck and Suzuki–Miyaura cross-coupling reactions using a solid-phase aryl iodide (Scheme 12.49). The premise behind these experiments was that, if 37 acted as a heterogeneous catalyst, no product would be formed with a heterogeneous polymer-supported aryl iodide, since it would be difficult for two heterogeneous species to interact in a significant way. Interestingly, cross-coupled products were indeed formed in all reactions in virtually quantitative yield, indicating that the actual catalyst is a homogeneous species that is leached from the polyphenylene–polyurea matrix, and thus the polymer of 37 might act as a reservoir for some form of catalyst or precatalyst that is released into solution for homogeneous catalysis. Both ICP and transmission electron microscopy analysis of the Heck reaction solutions supported this notion, since palladium nanoparticles could be observed.

Scheme 12.49

Richardson and Jones subsequently provided additional evidence to support the previous findings that the Pd EnCat catalysts act as reservoirs for homogeneous catalytic species by using the metal scavenging resins Quadrapure TU (Q) and poly(4-vinylpyridine) (R) (Figure 12.5).72 Their experiments were designed so that, if catalysis did indeed occur homogeneously outside of the polyphenylene–polyurea support, the leached catalyst would be scavenged by either Q or R and catalysis would thus be inhibited. This strategy was tested in Heck reactions, where the addition of a large excess of R dramatically retarded product formation. When Q was used, no catalysis was observed at all.

372


Figure 12.5 Metal scavenging polymers Quadrapure TU (P) and poly(4-vinylpyridine) (Q)

In addition to the above mentioned cross-coupling reactions, urea group cross-linked polyphenylene-encapsulated palladium has also been used to catalyze the hydrogenation of alkenes, alkynes, imine and nitro groups.73 In these reactions 37 was pre-reduced to a Pd(0) species (38) under hydrogen (50 bar) for 2 d and the corresponding reduction products could be obtained in high yield (Scheme 12.50). Significantly the catalyst exhibited only negligible metal leaching.

Scheme 12.50

Catalyst 38 was also applied in epoxide reduction reactions. However, the sample prepared by hydrogen reduction was inactive in these reactions (Scheme 12.51).74 Thus, for this application, 38 was prepared by reduction of 37 with HCOOH. In this method the encapsulated Pd(OAc)2 of 38 undergoes anionic ligand exchange with formate, resulting in efficient local reduction to deposit fine nanoparticles dispersed in the support material with less aggregation. This notion was supported by electron microscopy. Initially, trans-stilbene oxide was chosen as a substrate and the hydrogenolysis reaction was complete after 5 h with 4 equivalents of Et3N and HCOOH in ethyl acetate at 23 C to give the corresponding alcohol in 99% isolated yield. It is worth noting that further hydrogenolysis of the alcoholic CO bond was not observed, even after a prolonged reaction time. Also, the recyclability of 38 was found to be excellent, even after ten successive uses. Furthermore, metal leaching was undetectable.

Scheme 12.51

Catalyst 38 generated by anionic ligand exchange with formate was also used to effectively catalyze the transfer hydrogenation of aryl ketones (Scheme 12.52).75,76 Acetophenone was chosen as substrate to examine the recyclability of 38 and it was found that under the optimized reaction conditions, the reduction proceeded to completion with excellent isolated yields (96–99%) through five reaction cycles, and importantly, there was no evidence for


373

compounds formed from over-reduction. The superior catalytic properties of 38, as compared with those of 10% Pd–C, were demonstrated by carrying out the reduction of propiophenone under identical conditions. It was presumed that the metal center of 38 is more electron-rich than Pd–C, and that this could account for its superior catalytic properties.

Scheme 12.52

Finally, 38 was used to catalyze the deprotection of phenoxy benzyl ethers in the presence of ammonium formate with the assistance of microwave irradiation.77 Under optimized reaction conditions, phenoxy benzyl ethers bearing various functional groups were deprotected to afford the corresponding phenol in high yield (Scheme 12.53). In these reactions, nitro groups were reduced to amines, and bromide groups were removed.

Scheme 12.53

An OsO4 analogue of 37 has also been reported.78 Catalyst 39 was prepared by a procedure similar to the one used to prepare 37, and was used in the dihydroxylation of olefins with NMO as the stoichiometric oxidant (Scheme 12.54). The recycability of 39 was tested in six successive reactions by using different substrates for each cycle. No diol product crosscontamination was observed, which indicated that the reagents and products are not retained in the microcapsules after the workup procedure. Experiments were carried out to test for catalyst leaching by stirring 39 suspended in solvent for 24 h. The microcapsules were filtered, and the solution was examined in catalytic oxidation reactions. No reaction occurred, suggesting that significant leaching of an active osmium species had not taken place.

Scheme 12.54

374


The utility of 39 was also demonstrated in the oxidative cleavage of alkenes (Scheme 12.55). Treatment of a range of alkene substrates with 39 and NaIO4 gave the expected carbonyl compounds in good yields.

Scheme 12.55

12.4 Summary Table Catalyst loading level (mmol/g)

Catalyst

Encapsulating polymer

Metal species used for microencapsulation

1 2 3 4 5

NCPS NCPS NCPS NCPS NCPS

0.20 0.42 0.14 0.06 0.06

8–10 11–16 17 18 18

6 7 8 9 10 11

NCPS NCPS NCPS NCPS NCPS poly(acrylonitrile-cobutadiene-co-styrene) phenoxyethoxymethylpolystyrene D polysulfone E F DVB cross-linked polystyrene DVB cross-linked polystyrene Tentagel Tentagel G H I J K L M N

Sc(OTf)3 OsO4 Pd(PPh3)4 Ru(h6-C6H5CO2Et)(PPh3)Cl2 [RuCl(PCy3)(¼C¼C¼CPh2)] þ [PF6] VO(acac)2 Cu(acac)2 Bi(OTf)3 Ru complex B Ru complex C OsO4

0.60 0.38 0.23 0.07 0.26 0.60

19 20 21,22 23 24 28,29

OsO4

0.60

30

Pd(OAc)2 OsO4 PdCl2 H3PW12O40 OsO4

0.20 0.66 1.67 0.30 0.57

31 32 33,34 35 36

MTO

0.44

37–41

Pd(OAc)2 K[PtCl3(CH2CH2)] Pd(PPh3)4 Pd(PPh3)4 Pd(PPh3)4 Pd(PPh3)4 Pd(PPh3)4 Pd(PPh3)4 Pd(PPh3)4 Pd(PPh3)4

0.37 0.29 0.11 0.71 1.04 0.62 0.36 0.42 0.32 0.23

42–44 45 46–78 49 50 51 52,53 54,55 54,55 54,55

12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28

Reference(s)

(continued)


375

(Continued) Table (continued)

Catalyst

Encapsulating polymer

Metal species used for microencapsulation

29 30 31 32 33 34 35 36 37 38 39

J J G J G G J O P P P

Pd(NO3)2 Sc(OTf)3 RuCl2(PPh3)3 RuCl2(PPh3)3 RuCl3 H2O Pt(PPh3)4 PtCl2(COD) AuClPPh3 Pd(OAc)2 Pd(0) OsO4

Catalyst loading level (mmol/g) 0.44 0.19 0.33 0.28 0.43 0.77 0.14 0.06–0.08 0.30–0.40 0.30–0.40 0.20

Reference(s) 56 57 58 58 59 60 61 62,63 64–72 73–77 78

12.5 Conclusions Hopefully this review has provided a comprehensive introduction to the concept of metal catalyst microencapsulation by aromatic group bearing organic polymers. As can be seen, a variety of such polymers have been used in this context, and the trend seems to be that crosslinked materials are preferred to non-cross-linked ones. Additionally, studies of how such materials catalyze reactions seem to indicate that the polymer matrix may act as a reservoir for a pre-catalyst or catalyst, especially in the case of the Pd EnCat materials such as 37. Regardless of the true nature of catalysis, it is widely accepted that such microencapsulated catalysis facilitate organic synthesis since they can be removed from reactions by filtration and they generally exhibit only negligible metal leaching into the desired reaction product. While much of the work described revolves around palladium catalysts, other microencapsulated metal catalysts have been reported as well. In recent years great advances have been made in the preparation of ruthenium, scandium, gold and osmium microencapsulated catalytic materials, and it is anticipated that future developments in polymer design and synthesis will lead to improved microencapsulated metal catalyst with broader utility.

References 1. Microcapsules and Nanoparticles in Medicine and Pharmacy, Donbrow, M., Ed., CRC Press, Boca Raton, 1991. 2. Microencapsulation of Drugs, Whateley, T. L., Ed., Harwood Academic Publishers, Philadelphia, 1992. 3. I. Yamada, R. Noyori, Chemtracts 2000, 13, 620–625. 4. S. Kobayashi, R. Akiyama, Pure Appl. Chem. 2001, 73, 1103–1111. 5. S. Kobayashi, R. Akiyama, Chem. Commun. 2003, 449–460. 6. D. A. Pears, S. C. Smith, Aldrichimica Acta 2005, 38, 23–33. 7. S. Kobayashi, M. Sugiura, Adv. Synth. Catal. 2006, 348, 1496–1504.

376 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55.

Recoverable and Recyclable Catalysts S. Kobayashi, S. Nagayama, J. Am. Chem. Soc. 1998, 120, 2985–2986. T. Suzuki, T. Watahiki, T. Oriyama, Tetrahedron Lett. 2000, 41, 8903–8906. F. Schager, W. Bonrath, Appl. Catal., A 2000, 202, 117–120. S. Nagayama, M. Endo, S. Kobayashi, J. Org. Chem. 1998, 63, 6094–6095. N. Murakami, M. Sugimoto, M. Kobayashi, Bioorg. Med. Chem. 2001, 9, 57–67. O. Okitsu, R. Suzuki, S. Kobayashi, J. Org. Chem. 2001, 66, 809–823. S. Aoki, K. Matsui, K. Tanaka, R. Satari, M. Kobayashi, Tetrahedron 2000, 56, 9945–9948. M. Yamashita, N. Ohta, I. Kawasaki, S. Ohta, Org. Lett. 2001, 3, 1359–1362. H. Kikuchi, Y. Saito, J. Komiya, Y. Takaya, S. Honma, N. Nakahata, A. Ito, Y. Oshima, J. Org. Chem. 2001, 66, 6982–6987. R. Akiyama, S. Kobayashi, Angew. Chem. Int. Ed. 2001, 40, 3469–3471. R. Akiyama, S. Kobayashi, Angew. Chem. Int. Ed. 2002, 41, 2602–2604. A. Lattanzi, N. E. Leadbeater, Org. Lett. 2002, 4, 1519–1521. M. L. Kantam, B. Kavita, V. Neeraja, Y. Haritha, M. K. Chaudhuri, S. K. Dehury, Tetrahedron Lett. 2003, 44, 9029–9032. B. M. Choudary, C. Sridhar, M. Sateesh, B. Sreedhar, J. Mol. Catal. A: Chem. 2004, 212, 237–243. B. Sreedhar, V. Swapna, C. Sridhar, Catal. Commun. 2005, 6, 293–296. S. E. Gibson, V. M. Swamy, Adv. Synth. Catal. 2002, 344, 619–621. A. Cornejo, J. M. Fraile, J. I. Garcıa, M. J. Gil, V. Martınez-Merino, J. A. Mayoral, Tetrahedron 2005, 61, 12107–12110. R. Naik, P. Joshi, R. K. Deshpande, Catal. Commun. 2004, 5, 195–198. R. Naik, P. Joshi, S. Umbarkar, R. K. Deshpande, Catal. Commun. 2005, 6, 125–129. R. Naik, P. Joshi, R. K. Deshpande, J. Mol. Catal. A: Chem. 2005, 238, 46–50. S. Kobayashi, M. Endo, S. Nagayama, J. Am. Chem. Soc. 1999, 121, 11229–11230. S. Kobayashi, T. Ishida, R. Akiyama, Org. Lett. 2001, 3, 2649–2652. T. Ishida, R. Akiyama, S. Kobayashi, Adv. Synth. Catal. 2003, 345, 576–579. H. S. He, J. J. Yan, R. Shen, S. Zhou, P. H. Toy, Synlett 2006, 563–566. S. M. Reddy, M. Srinivasulu, Y. V. Reddy, M. Narasimhulu, Y. Venkateswarlu, Tetrahedron Lett. 2006, 47, 5285–5288. Y. M. A. Yamada, Y. Uozumi, Org. Lett. 2006, 8, 1375–1378. Y. M. A. Yamada, Y. Uozumi, Tetrahedron 2007, 63, 8492–8498. Y. M. A. Yamada, H. Guo, Y. Uozumi, Org. Lett. 2007, 9, 1501–1504. T. Ishida, R. Akiyama, S. Kobayashi, Adv. Synth. Catal. 2005, 347, 1189–1192. R. Saladino, V. Neri, A. R. Pelliccia, R. Caminiti, C. Sadun, J. Org. Chem. 2002, 67, 1323–1332. R. Saladino, A. Andreoni, V. Neri, C. Crestini, Tetrahedron 2005, 61, 1069–1075. R. Saladino, E. Mincione, O. A. Attanasi, P. Filippone, Pure Appl. Chem. 2003, 75, 265–272. R. Saladino, V. Neri, F. Cardona, A. Gotib, Adv. Synth. Catal. 2004, 346, 639–647. A. Goti, F. Cardon, G. Soldaini, C. Crestini, C. Fiani, R. Saladino, Adv. Synth. Catal. 2006, 348, 476–486. Y. Uozumi, R. Nakao, Angew. Chem. Int. Ed. 2003, 42, 194–197. R. Nakao, H. Rhee, Y. Uozumi, Org. Lett. 2005, 7, 163–165. Y. Uozumi, R. Nakao, H. Rhee, J. Organomet. Chem. 2007, 692, 420–427. Y. M. A. Yamada, T. Arakawa, H. Hocke, Y. Uozumi, Angew. Chem. Int. Ed. 2007, 46, 704–706. R. Akiyama, S. Kobayashi, J. Am. Chem. Soc. 2003, 125, 3412–3413. K. Okamoto, R. Akiyama, S. Kobayashi, Org. Lett. 2004, 6, 1987–1990. R. Nishio, S. Wessely, M. Sugiura, S. Kobayashi, J. Comb. Chem. 2006, 8, 459–461. K. Okamoto, R. Akiyama, S. Kobayashi, J. Org. Chem. 2004, 69, 2871–2873. R. Akiyama, T. Sagae, M. Sugiura, S. Kobayashi, J. Organomet. Chem. 2004, 689, 3806–3809. K. Okamoto, R. Akiyama, H. Yoshida, T. Yoshida, S. Kobayash, J. Am. Chem. Soc. 2005, 127, 2125–2135. R. Nishio, M. Sugiura, S. Kobayashi, Org. Lett. 2005, 7, 4831–4834. R. Nishio, M. Sugiura, S. Kobayashi, Org. Biomol. Chem. 2006, 4, 992–995. R. Nishio, M. Sugiura, S. Kobayashi, Chem. Asian J. 2007, 2, 983–995. T. Inasaki, M. Ueno, S. Miyamoto, S. Kobayashi, Synlett 2007, 3209–3213.


377

56. H. Hagio, M. Sugiura, S. Kobayashi, Org. Lett. 2006, 8, 375–378. 57. M. Takeuchi, R. Akiyama, S. Kobayashi, J. Am. Chem. Soc. 2005, 127, 13096–13097. 58. H. Miyamura, R. Akiyama, T. Ishida, R. Matsubara, M. Takeuchi, S. Kobayashi, Tetrahedron 2005, 61, 12177–12185. 59. T. Matsumoto, M. Ueno, J. Kobayashi, H. Miyamura, Y. Mori, S. Kobayashi, Adv. Synth. Catal. 2007, 349, 531–534. 60. H. Hagio, M. Sugiura, S. Kobayashi, Synlett 2005, 813–816. 61. Y. Miyazaki, H. Hagio, S. Kobayashi, Org. Biomol. Chem. 2006, 4, 2529–2531. 62. H. Miyamura, R. Matsubara, Y. Miyazaki, S. Kobayashi, Angew. Chem. Int. Ed. 2007, 46, 4151–4154. 63. H. Miyamura, M. Shiramizu, R. Matsubara, S. Kobayashi, Chem. Lett. 2008, 37, 360–361. 64. C. Ramarao, S. V. Ley, S. C. Smith, I. M. Shirley, N. DeAlmeida, Chem. Commun. 2002, 1132–1133. 65. S. V. Ley, C. Ramarao, R. S. Gordon, A. B. Holmes, A. J. Morrison, I.. F. McConvey, I. M. Shirley, S. C. Smith, M. D. Smith, Chem. Commun. 2002, 1134–1135. 66. C. K. Y. Lee, A. B. Holmes, S. V. Ley, I. F. McConvey, B. Al-Duri, G. A. Leeke, R. C. D. Santos, J. P. K. Seville, Chem. Commun. 2005, 2175–2177. 67. G. A. Leeke, R. C. D. Santos, B, Al-Duri, J. P. K. Seville, C. J. Smith, C. K. Y. Lee, A. B. Holmes, I. F. McConvey, Org. Process Res. Dev. 2007, 11, 144–148. 68. I. R. Baxendale, C. M. Griffiths-Jones, S. V. Ley, Chem. Eur. J. 2006, 12, 4407–4416. 69. A. K. Sharma, K. Gowdahalli, J. Krzeminski, S. Amin, J. Org. Chem. 2007, 72, 8987–8989. 70. E. Vickerstaffe, A.L. Villard, M. Ladlow, S. V. Ley, Synlett 2007, 1251–1254. 71. S. J. Broadwater, D. T. McQuade, J. Org. Chem. 2006, 71, 2131–2134. 72. J. M. Richardson, C. W. Jones, Adv. Synth. Catal. 2006, 348, 1207–1216. 73. N. Bremeyer, S. V. Ley, C. Ramarao, I. M. Shirley, S. C. Smith, Synlett 2002, 1843–1844. 74. S. V. Ley, C. Mitchell, D. Pears, C. Ramarao, J. Q. Yu, W. Zhou, Org. Lett. 2003, 5, 4665–4668. 75. J. Q. Yu, H. C. Wu, C. Ramarao, J. B. Spencer, S. V. Ley, Chem. Commun. 2003, 678–679. 76. S. V. Ley, A. J. P. Stewart-Liddon, D. Pears, R. H. Perni, K. Treacher, Beil. J. Org. Chem. 2006, 2, 15. 77. M. Quai, C. Repetto, W. Barbaglia, E. Cereda, Tetrahedron Lett. 2007, 48, 1241–1245. 78. S. V. Ley, C. Ramarao, A. L. Lee, N. Østergaard, S. C. Smith, I. M. Shirley, Org. Lett. 2003, 5, 185–187.

13 Organic Synthesis with Mini Flow Reactors Using Immobilised Catalysts Sascha Ceylan and Andreas Kirschning Institut f€ ur Organische Chemie, Leibniz Universit€ at Hannover, Hannover, Germany

13.1 Introduction 13.1.1 General Remarks The past two decades have changed the nature of organic synthesis dramatically in several aspects. While catalysis1 and the development of stereoselective methods2 has given the synthetic chemist myriads of new chemical tools, combinatorial chemistry3 set the stage for focusing on new techniques that lead to automated organic synthesis. However, classical combinatorial chemistry has not fulfilled all expectations such that it is justified to say that its peak can be located in the near past. Nevertheless, the important outcome of combinatorial chemistry is that it gave the organic community a sense of the importance of automation and improved technologies. Consequently, so-called enabling technologies have emerged in the past decade and have influenced the way organic synthesis is conducted to a very great extent.4 We defined enabling technologies to be either traditional or new techniques whose purpose is to speed up synthetic transformations and importantly to ease workup as well as the isolation of products. Furthermore, they are expected to help close the gap between bench chemistry and chemical engineering by mimicking large-scale production in the laboratory. Enabling technologies can briefly be summarised as (a) solid phase assistance such as hetereogenised homogeneous catalysts,5 (b) new solvent systems like


380

Recoverable and Recyclable Catalysts solid phase assisted synthesis

new reactor design

new heating techniques

new solvent systems

Figure 13.1 Examples of enabling technologies in organic synthesis

ionic liquids,6 (c) new heating devices such as microwave (mw) irradiation7 and (d) new reactor designs such as continuous-flow and micro reactors (Figure 13.1).8 In fact, truly new synthetic technology platforms will not be based on the individual use of these enabling technologies but will require the combination of two or more of them. We wish to point out that those examples shown in Figure 13.1 have been chosen for obvious reasons; but the list can be complemented with additional techniques. 13.1.2 Batch versus Flow Processes Chemical synthesis in the laboratory is often carried out in standardised glassware which has commonly been in use since Justus Liebig’s time. Flow and continuous flow processes have traditionally been restricted to chemical and bio production processes within the food, chemical and pharmaceutical industries, where they have shown many advantages over batch processes.9 Flow reactors carry material as a flowing stream. Reactants are continuously fed into the reactor and emerge as a continuous stream of product. In batch production stoichiometry is defined by the concentration of chemical reagents and their volumetric ratio. In flow reactors this is defined by the concentration of reagents and the ratio of their flow rate. In batch production reaction time is determined by how long a vessel is held at a given temperature. In flow this is determined by the volume of the reactor, and the bulk flow rate. Advantageously, continuous flow reactors are generally smaller than batch reactors. Recently, flow processes have been minituarised and thus have appeared as bench-sized microstructured devices in the laboratory. Benefits of synthesis in such miniatuarised setups are that facile automation, reproducibility, safety and process reliability due to constant reaction parameters (efficient mixing, temperature, time, amount of reagents and solvent etc.)* can advantageously be assured. In some cases, the reaction temperature can be far above the solvent’s boiling point due to the ability of being operated under pressure. Importantly, multistep reactions can be arranged in a continuous sequence. This can be especially beneficial if intermediate compounds are unstable. It is also possible to arrange the system such that further reagents can be introduced into the flowing reaction stream at precisely the time that is required for the reaction. Chemistry in flow can be coupled with *By coupling the output of the reactor to a detector system, it is possible to create an unattended system which can sequentially investigate a range of possible reaction parameters and therefore optimise reactions with little or no intervention.

Organic Synthesis with Mini Flow Reactors Using Immobilised Catalysts

381

Micro flow reactor channel diameter: 50-500 μm

Mini flow reactor channel diameter: 500-2000 μm

Advantages

Advantages

- high heat transfer surface to product volume ratios

- improved flow capacities - lower pressure drop

- good heat transfer capabilities

Flow reactor - no blocking of channels

Disadvantages Disadvantages - micro channels suffer from restricted flow capacity

- lower heat transfer surface restricted flow capacity

- high pressure drop - poorer heat transfer capabilities - tendency to block - limited capabilities for a given diameter and length

Figure 13.2 Micro versus mini flow reactors

packed-beds or purification concepts using solid phase scavengers, immobilised reagents and catalysts, chromatographic separation or liquid/liquid extraction. Scale-up of a proven reaction can be achieved rapidly with little or no process development work, by either changing the reactor volume or by running several reactors in parallel, provided that flows are recalculated to achieve the same residence times. Although this has not been explicitly discussed in the literature, evidence can be collected that modern flow reactor devices need to be classified into micro flow reactors and mini flow reactors (Figure 13.2). Historically, miniaturisation focused on micro flow reactors but recently the development has been directed towards mini flow reactors for practical reasons. 13.1.3 Micro versus Mini Flow Reactors It is possible to have flow reactors operating at large scale; however, for use in the laboratory, channel/tube scale is likely to be in the region of 50 mm to a few mm. The channel diameter simply determines the category of the flow reactor: micro versus mini flow reactor. Nonetheless, this channel scale range is sufficiently broad to allow single experiments with approximately 10 mg of starting material to an annual production rate of several tens of tons of material. While micro reactors have found widespread use in the academic field as well as in analytical and diagnostic applications mini reactors have become more and more popular in the industrial context.10 In as far as synthetic efficiency is concerned, there are a number of benefits with respect to thermal and mass transfer, efficient mixing as well as mass transport that allow chemistry to be performed efficiently under flow conditions. The scale of micro flow reactors can make them ideal for process development experiments. From a chemical engineering point of view a high heat transfer capacity per unit volume of product is desirable. As far as flow reactors are concerned, the heat transfer capacity is heavily influenced by channel size since this determines the heat transfer area per unit volume. For mini reactors the common values are between 100 and 10 000 m2/m3 (depending on channel size) while for micro reactors

382


parameters are commonly between 5 000 and 50 000 m2/m3. Small diameter channels have the advantage of high heat transfer capacity favoring micro over mini flow reactors. However, the narrow channels can result in high pressure drops, limited flow capacity and a tendency to block. They are also often fabricated in a manner which makes cleaning and dismantling difficult or impossible. Thus, the fact that industry directs its attention to mini flow reactors is a result of combining chemical engineering requirements with practicability in the chemical production. However, a critical view on mini flow systems must include additional aspects that may hamper the realisation of flow-through processes. These include (a) inert properties of all materials in the flow-through system towards a large variety of different organic solvents, (b) efficient regeneration of reaction columns, (c) facilities to purify intermediates or final products and (d) problems associated with different kinetics of reactions and the necessity of different solvents when performing multistep syntheses under continuous flow conditions. In this overview we shall focus on flow processes, which are combined with enabling technologies based on functionalised solid supports and in some cases on new heating techniques. Most examples discussed will deal with mini flow reactors because we strongly believe that these reactors are the ideal compromise for industrial applications with miniaturised flow devices thus closing the gap between synthesis at the bench and industrial processes from a chemical engineering perspective. Indeed, one of the first reports on mini flow reactors filled with chemically functionalised monoliths came from our laboratories.11 In order to restrict the scope of this report we shall mainly provide examples using catalysts immobilised on several types of solid supports.12

13.2 Catalysis in Mini Flow Reactors with Immobilised Catalysts One ideal setup for continuous processes combines flow reactors with heterogenised catalysts.13 Indeed, in the future catalytic processes will become far more important, particularly when reactions that have been developed in the laboratory need to be rapidly transferred into large scale or production scale. This point is still one of the key bottlenecks in process development of chemicals, including fine chemicals. In fact, the broadest practical application of flow techniques using heterogeneous catalysis can be found for gas phase transformations in the automobile business.14 Several solid supports are suited for heterogeneous catalysis or for heterogenisation of homogeneous catalysts. Classically, inorganic materials based on clays, silica, alumina and carbon are used but more recently other supports such as ionic liquids or polymers are employed, too.15 Recently, research has been devoted towards creating inorganic or polymer-based monoliths as supports, a concept that has successfully been utilised in the automobile industry. In the following we shall categorise the use of heterogenised homogeneous catalysts inside flow reactors based on the supports chosen. Due to the scope of a book chapter we shall only describe representative examples. 13.2.1 Solid Supports Based on Silica The basic concept of using silica as a solid support is to utilise the free silanol groups for covalently attaching a spacer or functional group. Functional groups can be chosen so that a catalyst or ligand can be attached covalently, by ionic interactions, by coordinative means


383

or by van der Waals interactions. Importantly, these techniques are in principle applicable on most types of solid supports but this aspect will only be discussed in this paragraph in detail. An important property of silica supports is that they, unlike polymers, do not swell in organic solvents. However, the chemical stability towards acids and bases as well as mechanical stability is restricted giving rise to gradual degradation or fragmentation. Nevertheless, the problem of mechanical stability can be overcome when these supports are located inside flow reactors. A recent example of a coordinatively complexed copper(I) catalyst was disclosed by Ying and coworkers.16 They attached an aza(bisoxazoline) ligand via a short spacer on siliceous mesocellular foam (MCF). This procedure was followed by silylation thus capping remaining free silanol groups. Coordinative trapping of copper(I) ions yielded a catalytic support which was incorporated inside an HPLC column (50 mm length, 4.6 mm I.D.) in a packed bed fashion. Catalyst 1 was successfully employed in asymmetric cyclopropanation reactions giving yields and enantiomeric excesses superior to former systems (Scheme 13.1). As reported the catalyst could be reused more than 20 times without a decrease in reactivity. The advantage of MCF as support is associated with the large pores, providing a large surface area and guaranteeing only low back pressure inside the flow system. 10 °C, 0.2 mL/min, CH2Cl2

O

H OEt

CO2Et

N2

90 % yield, 91 % ee O O Si MeO

O N

1

Cu(L)n+

N O

Scheme 13.1 Cyclopropanation of isobutene under flow conditions

Another example of metal coordination to a covalently immobilised ligand was reported by Styring and Phan (Scheme 13.2).17 A nonsymmetrical salen-type Ni catalyst was tethered to hydrosiloxy-silica via a long alkyl chain. The ‘androgynous’ catalyst 2 was utilised in ligand-free Kumada cross-coupling reactions in a packed bed reactor (25 mm length, 3 mm I.D.). Kinetic studies revealed that the catalytic system performed comparatively well with respect to analogous batch reactions in which the reaction time decreased from 24 h to several minutes under otherwise identical conditions. However, after 5 h the catalyst showed some decrease in activity due to the precipitation of salts. Nonetheless, after washing and drying full activity of the catalytic system was restored. An example of a covalently bound base on silica and its use in catalysis under flow conditions was demonstrated by Macquarrie and Brophy (Scheme 13.3).18 The inner surface of two aluminium plates was coated with silica by using a solution of sodium silicate and 3-aminopropyltrimethoxysilane. This procedure resulted in a sandwich-like catalytic

384

Recoverable and Recyclable Catalysts rt, 13 μL/min, THF

MeO

MeO MgX

Br

X = Cl, Br

64 % yield + 36 % by-products O

O Si

(CH2)9

O

O Ni

2

N

N

Scheme 13.2 Ni-catalysed Kumada reaction under flow conditions

flowcell. The free amino groups in 3 were able to catalyse Knoevenagel condensations and Michael additions under solvent-free conditions. Importantly, the rate of conversion did not drop significantly even after several hours of continuous use. A comparative study revealed the superiority of the continuous flow system over the corresponding batch experiment. Thus, slow poisoning of the primary amino groups by the ester functionalities turned out to be minimised in the flow mode. It needs to be pointed out that the dimensions of the used flow cell were quite small (0.15 mm deep, 30 mm long, 17 mm wide).19 silica coating CO2Et O

NC

CO2Et 90 °C, 6.6 μL/min, neat

CN

top plate

or

or bottom plate

NO2

up to 80 % conversion

NO2

O O O O Si MeO

3

NH2

Scheme 13.3 The Knoevenagel condensation and Michael addition under flow conditions

Quaternary phosphonium salts which are loaded onto silica by ionic interactions inside a plug-flow reactor were utilised by Tundo et al. (Scheme 13.4).20 By heating silica gel with a solution of n-Bu4P þ Br in methanol a functionalised support 4 was generated which was able to catalyse the SN2 reaction of 1-butanol to the corresponding terminal butyl halide. The reaction was performed as a gas–liquid phase process in which a gas stream of hydrochloric acid and 1-butanol was pumped over a melt (170 C) of the supported catalyst. The catalyst showed high selectivity for a SN2-type reaction with minimal formation of rearranged and other by-products. The product was simply collected by separating the aqueous phase from the organic layer. The flow reactor had a total volume of 90 mL so that


385

150 °C, 30 mL/h n-BuOH

n-BuX

aq. HX

H2O

up to 95 % yield SiO2/ n-Bu4P+Br- (4) flowing gas phase:

liquid phase:

RCH2OH

RCH2OH

HX

RCH2X

Q+HX2-

RCH2X

H2O

H2O

solid support

Scheme 13.4 Substitution reactions using solid phase catalyst 4 under flow conditions

multigrams of the desired product were constantly formed within 1 h and without any decrease of the catalyst’s activity. Wasserscheid and Riisager thoroughly investigated a supported ionic-liquid-phase (SILP) for catalysis and exploited van der Waals/ionic interactions to impregnate a silica support with a rhodium catalyst and an ionic liquid (Scheme 13.5).21 Silica, bisphosphine ligand 6 and [Rh(acac)(CO2)] 7 were suspended in the ionic liquid [bmim][n-C8H17OSO3] 5 and volatile components were removed afterwards. This created silica gel covered with a thin film of the ionic liquid that contained the catalyst. This material was used inside a stainless-steel packed-bed tubular reactor for carrying out the selective hydroformylation of propene to n-propanal by passing a gas mixture consisting of propene, CO and H2 (1:1:1) under flow conditions. It was demonstrated that, beside the choice of the ionic liquid, the ligand and the compound ratio, the nature of the support’s surface was also crucial for best results. Thus, it was important to chemically remove most of the free silanol groups from the silica surface prior to use. In fact, these groups could bind irreversibly to the catalyst ligands thereby deactivating the catalyst. After this modification the catalytic system remained active after several days of use. Another example of silica-loaded catalysts inside flow reactors and their applications in asymmetric synthesis was reported by Leeuwen and coworkers (Scheme 13.6). Asymmetric transfer hydrogenations were achieved under flow conditions when covalently bound NH-benzyl-(1R,2S)-(-)-norephedrine ligands on silica gel were doped with {RuCl2(h6-pcymene}.22 Incorporated inside a reaction column (1.5 cm length, 0.7 cm I.D.) the supported catalyst 8 allowed to hydrogenate acetophenone with up to 95% conversion and in 90% ee without significant metal leaching (99 % conversion

13 O b. NaN3, MeOH/H2O, Δ then Cu(I), DIPEA, THF, 50 °C and MeO N

HO N N N MeO N

N

40 °C, toluene

OAc

Me

or

HN

O

N O

14

N 14

Scheme 13.10 Solid-phase assisted Knoevenagel condensation under flow conditions

toluene through the reactor. Conversion was determined to be 99% (93% ee) within minutes after one pass through the column. This set the stage for a continuous process preparing up to 13.0 mmol/h of product per gram of resin 16. In comparison to the corresponding batch experiment, the reaction times are 24-fold shorter under flow conditions. Although the authors did not state for how long the chiral support could be used, simple washing protocols allowed the reactor to be regenerated for several consecutive alkylations with different aldehydes. B(OH)2

100 °C, DMF/H2O 1:1, (iPr)2EtN, 6 μL/min

Br R

R

Omnifit glass column up to 77 % conversion

R = CHO, Me and others O N 15

O

Pd N

Me

O Me

Scheme 13.11 Suzuki–Miyaura cross-coupling reactions under flow conditions

Organic Synthesis with Mini Flow Reactors Using Immobilised Catalysts Me

10 °C, toluene, Et2Zn, 0.24 mL/min

OH

O

R

R up to 99 % conversion 93 % ee

R = e.g. F, CN, H

Ph

391

1. piperazine/LiClO4 O H 2. Merrifield resin

N N

Ph

Ph

Ph 16 H

Ph Ph OH

Scheme 13.12 Asymmetric alkylation of benzaldehydes under flow conditions

Based on the earlier work of Girard and coworkers31 Ley and Baxendale reported an application of a chelated copper(I) catalyst on basic Amberlyst-21, a dimethyl-aminomethyl-grafted polystyrene support. The catalyst 17 was prepared by simply mixing the resin with copper iodide in acetonitrile and subsequent evaporation of the solvent.32 By incorporating this heterogenised catalyst inside a packed-bed Omnifit glass column (6 mm I.D., length 150 mm) it was possible to synthesise various 1,2,3-triazoles in up to 93% yield with >95% purity (Scheme 13.13). The high degree of purity was guaranteed by use of rt, 30 μL/min, CH2Cl2 R1

N3

2 eq

R2 Omnifit glass column

1 eq 17

NMe2CuI

S 18

N H

NH2 N N

N R1

R2 70-93 % yield 19

PPh2

e.g.

N N

N

NO2

HO 93 %

Scheme 13.13 Copper(I)-catalysed ‘click reactions’ and purification concepts under flow conditions

392


additional columns connected in a linear mode which were filled with the commercially available QP-TU resin 18 (Quadrapure TU) for metal scavenging and a phosphine resin 19 (PS-PPh2) to capture excess of azide. In a continuous scale-up experiment 1.5 g of a triazole was obtained within 3 h. Interestingly, the flow mode led to suppression of the Glaser coupling, a side reaction which was observed in batch experiments. Importantly, this flow setup eases the handling of potentially dangerous organic azides and acetylenes. However, it must be noted that most of the continuous flow processes described so far utilise reactors with randomly packed catalytic beds. These commonly show uncontrolled fluid dynamics which result in stagnation zones and hot-spot formation, broad residence time distribution, low selectivity and in essence low process efficiency. 13.2.3 Monolithic Supports All examples presented so far relied on flow reactors that were equipped with randomly packed catalytic beds and thus had uncontrolled fluid dynamics. This concept, however, does not withstand requirements from a process and chemical engineering point of view. Disadvantages are stagnation zones and hot-spot formation, broad residence time distribution, low selectivity and in essence low process efficiency. Structured beds which are designed on a nano-scale up to the macro-geometry are considered to overcome these drawbacks. A monolith is the best structured material known for this purpose and in a broad sense is defined as a block of structured material which consists of continuous substructures or regular or irregular channels.33 In fact, these materials have found wide use in the automobile business as supports for catalysts. Monolithic materials have a high void volume and a large geometric surface area. This results in a low pressure drop during the passage of a gas or a fluid and a large contact area of the reagent or the catalyst with the fluid.34 Inorganic materials based on silica gel or carbon can ideally be prepared as monoliths with uniform mesopores and tuneable microchannels.35 Additionally, three concepts for creating monolithic materials with regular or irregular channels based on polymeric phases are currently known: (a) copolymerisation of different monomers in the presence of porogens,36 (b) preparation of block copolymers of which a well-defined cylindrical and degradable polymer is embedded inside the second polymer37 and (c) polymerisation of a monolithic polymeric phase wedged inside the microchannel pore system of an inert support such as glass and other preformed inorganic materials.38 Only monoliths of types (a) and (c) have found wider application in the context of this report (see above). A typical example of a covalently bound Ru-based catalyst 20 for cyclopropanations is depicted in Scheme 13.14.39 The ruthenium complex became part of a monolithic phase (type a) after radical copolymerisation of styrene, DVB and a dichlororuthenium(II) complex in toluene inside a stainless steel column (15 cm length, 6.4 mm I.D.) in the presence of dodecanol as porogen. This reactor was utilised for performing the asymmetric cyclopropanation of styrene with ethyldiazoacetate with yields up to 72%, trans/cis ratios up to 89:11 and ee values up to 89%. With the mechanical stability of a monolith at hand, it was also possible to elevate the impact of high pressure which led to the sustainable approach of using supercritical CO2 as solvent. This nonconventional solvent is regarded to be ‘green’ as it is nontoxic, nonflammable and environmentally benign. This strategy proved to be in principal feasible with the advantage that the productivity proved to be 7.7 times (mol product per mol ligand) better than the conventional solvent CH2Cl2. The


40 °C, 550 μL/min

O OEt N2

393

Ph

CO2Et Ph

CO2Et

Ph

CO2Et Ph

CO2Et

monolith yield

O

N N Ru Cl

trans/cis

trans ee

cis ee

sc CO2

29 %

89:11

59 %

83 %

neat

72 %

83:17

43 %

79 %

O Cl

N

20 AIBN (radical polymerisation)

O

O

N N

N

Scheme 13.14 Asymmetric cyclopropanation of alkenes under flow conditions

catalyst could be operated over a period of 6 h without any decrease in activity. With a slightly modified protocol the immobilisation of another bisoxazoline ligand for performing copper(II)-catalysed cyclopropanations could be also achieved with similar results.40 A similar approach, however this time for the asymmetric alkylation of benzaldehyde, was undertaken by Luis and Martens by searching for the best suited ligand to be immobilised.41 Initially, grafting concepts were investigated. However, it turned out that monolithic materials derived from radical polymerisation of the monomer styrene, DVB and toluene/dodecanol as porogenic mixture performed in a superior fashion under flow conditions. The results were rationalised with a better arrangement of the chiral cavities and the lack of diffusional problems. In this approach, however, it was required to circulate the reaction mixture through the mini reactor in order to achieve good conversions (up to 88% yields, up to 90% ee). Furthermore, the catalytic monolith could be reused at least four times. Luis and coworkers utilised monoliths to immobilise the enzyme Chirazyme L2-C2 (CALB) for the preparation of citronellyl propionate from citronellol, an important substrate for the fragrance industry (Scheme 13.15).42 The monoliths were prepared by radical polymerisation of vinylbenzyl chloride (VBC) and DVB in dodecanol and toluene

394


O

80 °C, sc CO2, 10MPa 0.01 - 0.04 mL/min cyclic mode

HO

O O

O 93 %

monolith

HN

Cl

Cl

N Bu

NH N Bu

N Bu

Cl

Cl

NH

NH

N Bu

CALB

21

Scheme 13.15 CALB-catalysed transesterification under flow conditions

followed by modification with an ionic liquid-type ligand. This ionic support was able to trap CALB via the butyl sidechains. Originally, the authors planned to apply this heterogenised enzyme 21 in scCO2. However, under these conditions deactivation of CALB took place rather rapidly, so that they had to use the ionic liquid as stabiliser. Incorporation of the monolithic support inside a mini reactor (1 mL) allowed to perform the transesterification in a recycling process to afford the product in up to 93.0% yield. Even after 5 days the enzymatic activity could still be maintained under the flow conditions. Furthermore, the immobilisation strategy was well suited for reloading enzymes on the monolith inside the reactor using simple washing protocols.43 Recently, the same setup served to immobilise basic anions 22 (e.g. OAc, Cl, OH) and apply them in the solvent-free Henry reaction of p-nitrobenzaldehyde and nitromethane (Scheme 13.16). With respect to backpressure and fluid issues the monolithic material turned out to be superior to polymeric beads under flow conditions.44 rt, neat, 0.1-0.5 mL/min recirculation

O Me

OH

NO2

O2N

monolith

NO2

O2N 99 %

HN N

X NH

Cl Cl

N

NH N

22 X =

OAc-,

OH-, Cl-

Scheme 13.16 Henry reaction under flow conditions


395

While the above mentioned examples relied on monolithic materials of type (a), Kirschning and Kunz developed hybrid monoliths of type (c).38 These materials relied on a chemically functionalised highly porous polymer–glass composite which was obtained after precipitation polymerisation of styrene, chloromethylvinyl benzene, vinyl pyridine, vinyl imidazole, vinyl pyrrolidinon or other monomers in the pore volume of highly porous glass. This procedure created a polymeric matrix inside the glass which consisted of small polymer-bridged beads (1–5 mm diameter). Basically, creation of a monolithic polymeric phase with a high surface area, wedged inside the microchannel pore system of the inert, inorganic support was achieved. Recently, this monolithic material shaped as a rod inside a metallic cylinder or as Raschig rings served as support functionalised with sulfonate groups for binding a tagged Hoveyda-type metathesis precatalyst 23 by ion exchange (Scheme 13.17).45 This material (named PASSflow for Polymer Assisted Solution Phase Synthesis) showed superior activity under batch conditions in metathesis reactions compared with Grubbs and classical Hoveyda-type precatalysts. The first reported olefin metatheses under flow conditions revealed that the lifetime of the catalytic system was decreased compared with batch conditions. The convective flow inside the reactor may have caused increased mechanistically based leaching of the catalyst whereas in the batch mode the resin more easily could recapture the ruthenium species. However, the tagging concept allowed facile regeneration and reloading of the Ru complex onto the solid support. 45 °C, CH2Cl2, 2 mL/min, recirculating

O

H

monolith

MeO2C

O

H MeO2C 99 % conversion

O3S NEt2

Mes NCl Ru

O Cl

N

23 Mes

Scheme 13.17 Ring closing metathesis under flow conditions using an ionically tagged Ru-complex 23

Based on the PASSflow concept the enantioselective kinetic resolution of terminal epoxides by addition of water was achieved using an immobilised cobalt salen complex (Scheme 13.18).46 The monolithic Merrifield resin was covalently functionalised with a nonsymmetrical salen ligand by nucleophilic substitution. The corresponding Co complex 24 transformed epibromohydrin into its corresponding alcohol under cyclic mode flow conditions and complete transformation was achieved after 20 h. The enantio-enriched (91–93% ee) alcohols were isolated in 76–87% yield. The same catalytic reactor was used

396


O

Br

rt, H2O(1.5 eq),THF, 2 mL/min OH recirculating R

Br 76-87 % yield 91-93 % ee

OH monolith

O

O O O

24

N N O Co AcO tBu O But

t

Bu

Scheme 13.18 Dynamic kinetic resolution of epibromohydrin under flow conditions

for four times but in between each run reactivation was necessary by washing with a mixture of toluene and acetic acid.47 In another attempt megaporous monoliths shaped as Raschig rings were incorporated inside a specially designed mini flow reactor instead of a rod-shaped glass–polymer monolith (Scheme 13.19).48 This type of reactor system has been used for transfer hydrogenations under flow conditions. Here, the solvent and reactants were forced to flow through these rings which contained palladium(0) nanoparticles on a monolithic anion exchange resin. The cationic ammonium groups were able to stabilise these particles.49 One major advantage of this setup is the low price of these functionalised monolithic rings and that these rings can easily be replaced if necessary and functionalised again. Interestingly, the whole immobilisation process was carried out under flow conditions and different factors such as the flow rate, substrate concentrations, degree of polymer cross-linking or concentration of Pd(0) particles 25 on the polymer were investigated in order to shed light on the factors that control particle size and how the particle size influences the catalytic performance of the flow system. It was found that small particles (1–4 nm) were better suited for conventional heating while larger particles (5–9 nm) gave better results for microwave-heated hydrogenations. In all cases studied flow conditions were found to be superior to hydrogenation in the batch mode (e.g. reaction time for the hydrogenation of ethyl cinnamate: 14 h vs 3 h). The catalytic reactor could be reused for more than ten times and only minor leaching of palladium was found (maximum of 0.1% of initial catalyst loading). It should be noted that also benzyl ethers and aromatic nitro groups could be reduced in the flow mode with this setup (see also Scheme 13.20, equation 1).50 Additionally, these Pd(0) nanoparticles 25 were also suitable for promoting Heck reactions (equation 4) as well as Suzuki–Miyaura (equation 2) and Sonogashira (equation 3) crosscoupling reactions (Scheme 13.20).50 All the reactions had to be conducted in a cyclic flow mode but afforded the desired products in high to quantitative yields. The C–C coupling reactions were performed without the presence of additional phosphine ligands, clearly


397

Scheme 13.19 Microwave-assisted transfer hydrogenation of cinnamate under flow conditions; bottom: Raschig rings (left), PASSflow reactor (centre and right)

demonstrating that the ammonium groups on the solid support were able to stabilise the Pd particles. This observation was further supported by the fact that the reactor could be reused for more than 20 times. Scale-up to a 2 molar scale was achieved without the necessity of tedious optimisations. A similarapproach was later pursued byLey et al. (Scheme 13.21).51 However, a monolithic support of type (a) instead of type (c) was prepared and functionalised with anionic groups and subsequently with palladium(0) particles. This approach yielded higher loadings because the monolith was exclusively composed of the polymeric phase. The material became part of a flow reactor (glass column, 70 mm length, 6.6 mm I.D.) and Heck reactions were performed in high yields (usually above 80%) in a single pass. It was found that instead of DMF ethanol was verywellsuited athigherpressure (100 psi,130 C).Themonolithicflow systemcouldbeused more than 25 times without loss of activity. Contamination of the product with palladium ( 290 nm)

CN H

NC

O

O +

O CH3CN 10

H

H

11

12

batch (180 min)

56%

17%

microflow (3.4 min)

55%

7%

Scheme 14.11 Intramolecular photocatalysed [2 þ 2] cycloaddition

Application of photochemically generated singlet oxygen is a powerful synthetic method in organic chemistry. Since microflow systems involve only small volumes of solvent, the explosive nature of oxygen-rich organic solvents, as well as of the products (e.g., endoperoxides) is substantially reduced. These factors prompted a number of research groups to use microreactors in singlet oxygen chemistry. Singlet oxygen has been used in [4 þ 2] cycloadditions as shown in Scheme 14.12. With Rose Bengal as sensitizer, the product 13 is formed in 85% conversion, whereas in a batch process a yield of 67% is obtained.35 Using cyclopentadiene as starting material, cycloaddition proceeds similarly well and an intermediate endoperoxide is reduced by thiourea to 14.36

O2 Rose Bengal

O O

microreactor

13

1. O2 Rose Bengal 2. thiourea

HO

OH

microreactor 14

Scheme 14.12 Singlet oxygen in [4 þ 2] cycloadditions

Photocatalytic reductions have been carried out using a titanium dioxide-coated microchannel device with a 365 nm UV-LED light source. The reduction of benzaldehyde to benzyl alcohol (11% yield) or from p-nitrotoluene to p-toluidine (46% yield) was carried out in a residence time of 1 min in the presence of ethanol.37 A similar, titanium dioxide-modified microreactor was used for the photocatalytic redox-combined synthesis of L-pipecolic acid from L-lysine. Although batch and microflow experiments gave similar enantioselectivities, the conversion for the microflow reactor was greatly enhanced.38 A photocatalytic alkylation of amines with alcohols is known to proceed by UV irradiation in the presence of a titanium dioxide photocatalyst. This reaction has also been performed in a microreactor leading to a mixture of N-ethyl benzylamine 15

420


and N,N-diethyl benzylamine 16, the ratio could be influenced by varying the residence time.39 By changing to a Pt-free photocatalyst, exclusive monoethylation to 15 was achieved in high yields.40 photocatalyst, hυ Ph

NH2

Ph

EtOH

Pt-free photocatalyst

5s 10 s 90 s

NHEt

+ Ph

NEt2

15

16

65% 0% 98%

40% 29% 0%

Scheme 14.13 Photocatalytic N-ethylation of benzylamine

The photocatalytic reaction of singlet oxygen or methanol with terpenes has also been investigated. Citronellol 17 can be oxidized in the presence of a photosensitizer ½Ruðt bpyÞ3 Cl2 to obtain the hydroperoxide 18. This has been subsequently converted by reduction of the hydroperoxide and acid-catalysed cyclization into rose oxide 19, an important compound in the perfume industry.41 The photocatalysed reaction of limonene with methanol has also been reported leading to slightly increased stereoselectivities in the addition products compared to the batch reaction.42

OH

OH

O2, h υ

1. Na2SO3 2. H2SO4

H O

OOH

Ru(t bpy)3Cl2 17

19

18

Scheme 14.14 Photosensitized oxidation of citronellol

Ryu et al. have shown that the Barton reaction of steroid 20 can be performed efficiently in a microreactor. Irradiation with black light yielded the product 21 in 71% yield. A serial connection of two microreactors demonstrated the versatility of this approach towards multigram synthesis being the first approach to scale up a photochemical reaction in a microreactor.43 O

ONO

O

HO O

O

hυ 71%

NOH

O

O 20

21

Scheme 14.15 Photochemical Barton reaction

Homogeneous Catalysis Using Microreactor Technology

421

14.5 Asymmetric Catalytic Reactions Only very few research groups have looked at asymmetric catalytic reactions performed in microreactors. High throughput kinetic investigations of asymmetric hydrogenations using rhodium catalysts in microdevices have been performed some time ago.44 More recently, a T-shaped, electroosmotic driven microreactor was used for the optimization of the enantioselective silylcyanation of benzaldehyde, catalysed by lanthanide–pybox complexes.45 Compared to a conventional batch procedure, higher conversion was observed within shorter reaction time. The microreactor process involving Lu(III) afforded essentially the same enantioselectivity as the batch process, whereas the enantioselectivity was lower in the microreactor for catalysts containing Yb(III). Ce(III)-based catalysts showed negligible selectivity in both processes.

OTMS CHO

TMSCN LnCl3

O

CN

4 mol % Ph-pybox Ln: Lu Yb Ce

N

N

Ph 22

O

N

Ph Ph-pybox

73% ee (batch: 76% ee) 84% ee (batch: 89% ee) 1% ee (batch: 11% ee)

HCl hydroxynitrile lyase Ar–CHO MTBE/buffer

OH CN

Ar 23

Scheme 14.16 Asymmetric catalysis of cyanohydrins 22 and 23

A similar reaction, the enzyme-catalysed asymmetric addition of hydrogen cyanide to aldehydes, has been described recently.46 Crude enzyme lysates containing hydroxynitrile lyase have been used to prepare cyanohydrins 23 with enantioselectivities from 85–100% depending on retention time and substituent Ar. A soluble polyethylene glycol poly-L-leucine catalyst has been used for the synthesis of chalcone epoxide using hydrogen peroxide as oxidant. Efficient pre-mixing of the catalyst and hydrogen peroxide was necessary to obtain about 90% ee in the product.47

14.6 Unusual Reaction Conditions Microreactors offer the advantage to operate under high pressure and also under high temperature reaction conditions. This allows performing reactions under reaction conditions, which usually cannot be achieved easily in batch chemistry. Reactions can be performed at temperatures that are above the boiling point of solvents. Higher reaction rates and higher selectivities have been observed in various reactions. A palladium-catalysed

422


aminocarbonylation was investigated using 7.9 bar carbon monoxide with a residence time of 4 min. The temperature dependency of the reaction using 3-iodoanisole and morpholine was examined. When heating the reaction mixture above the boiling point of toluene (110 C), a significant shift towards the synthesis of amide 25 rather than 24 was observed. Other arylbromides have been investigated in this transformation as well and similar results have been obtained.48

MeO

I +

H N

2 mol % Pd(OAc)2 2.2 mol % Xantphos CO (7.9 bar)

O MeO

100 ºC 150 ºC

O N

+ O

2 equiv DBU toluene 4 min

O

O MeO

N

O

24

25

25% 70%

75% 30%

Scheme 14.17 Palladium-catalysed aminocarbonylation of 3-iodoanisole

However, solvent properties such as the dielectric constant also change when these compounds are exposed to high temperatures and pressures. Water is able to dissolve organic compounds under high temperature and high pressure conditions as the dielectric constant of water at 5MPa and at 200 C is e ¼ 34.9 and is therefore similar to acetonitrile, N, N-dimethylformamide or N,N-dimethylacetamide (e ¼ 36–38). This advantage has been used in rapid and highly selective Sonogashira coupling reactions. Aryliodides 26 have been coupled with alkyne 27 to phenylarylethynes 28 as shown in Scheme 14.18. No copper catalyst is necessary under these reaction conditions and the reaction time can be reduced to 0.1 s.49

2 mol % PdCl2 2M NaOH

R I

26

+

R

Ph

Ph

27

H2O (16 MPa, 250 ºC) 0.1 s 7 examples, 83-99%

28

Scheme 14.18 Rapid and highly selective Sonogashira coupling reactions

The changed solvent properties under the extreme conditions even allow uncatalysed reactions to take place easily. O-Acylations of alcohols have been carried out with acetic anhydride in the absence of any catalyst. It is believed that water becomes Lewis acidic under the reaction conditions; the corresponding acetates 29 have been isolated in good yields.50 This approach is an environmentally benign process which might find application in industrial scale acylations. Reactions at high temperatures and pressures can evolve to reactions of general interest for the field of ‘green’ organic synthesis.

Homogeneous Catalysis Using Microreactor Technology O R OH

O

+ O

H2O (5 MPa, 200 ºC) no catalyst 11 examples, 82-99%

423

O R O 29

Scheme 14.19 Catalyst-free acetylation of alcohols

Substantial rate enhancements have also been observed for esterification reactions carried out with methanol under high pressure using supercritical carbon dioxide as a co-solvent. Pressures up to 110 bar have been investigated.51 Interesting studies also have appeared on the comparison of homogeneously and heterogeneously catalysed esterification reactions together with the design of novel miniaturized, fixed-bed reactors for that purpose.52 Acylations of ferrocene, catalysed by phosphoric acid, have been performed also in microreactors. It was found that such acylations proceed via a rapid and highly selective reaction while various anhydrides could be employed. The formation of diacylferrocenes, a typical byproduct in batch reactions, is not observed at all.53 Only few homogeneously catalysed oxidations performed in microreactors have been described in literature. The TEMPO (2,2,6,6-tetramethylpiperidine-1-oxyl)-catalysed oxidation of 2-butoxyethanol 30 to 2-butoxyacetaldehyde 31 has been optimized for a continuous process in a microreactor with an efficient suppression of the unwanted subsequent oxidation to 2-butoxyacetic acid.54 Catalytic oxidation of thiophene derivatives for desulfurization have been reported as well.55

O 30

OH

aq. NaOCl TEMPO NaBr CH2Cl2

O

O

31

Scheme 14.20 TEMPO-catalysed oxidation of 2-butoxyethanol 30

References 1. L. Ducry, D. Rohberge, Angew. Chem. Int. Ed. 2005, 44, 7972. 2. G. Dummann, U. Quitmann, L. Gr€oschel, D. W. Agar, O. W€ orz, K. Morgenschweis, Catal. Today 2002, 78–79, 433. 3. G. Panke, T. Schwalbe, W. Stirner, S. Taghavi-Moghadam, G. Wille, Synthesis 2003, 2827. 4. R. Halder, A. Lawal, R. Damavarapu, Catal. Today 2007, 125, 74. 5. A. A. Kulkarni, N. T. Nivangune, V. S. Kalyani, R. A. Joshi, R. R. Joshi, Org. Process Res. Dev. 2008, 12, 995. 6. J. R. Burns, C. Ramshaw, Trans IChemE 1999, 77, 206. 7. G. Dummann, U. Quittmann, L. Groschel, D. W. Agar, O. Worz, K. Morgenschweis, Catal. Today 2003, 79–80, 433. 8. (a) H. Audrain, Angew. Chem. Int. Ed. 2007, 46, 1772; (b) L. Cai, S. Lu, V. W. Pike, Eur. J. Org. Chem. 2008, 2853.

424


9. (a) J. M. Gillies, C. Prenant, G. N. Chimon, G. J. Smethurst, B. A. Dekker, J. Zweit, Appl. Rad. Isotopes 2006, 64, 333; (b) J. M. Gillies, C. Prenant, G. N. Chimon, G. J. Smethurst, B. W. Perrie, I. Hamblett, A. Dekker, J. Zweit, Appl. Rad. Isotopes 2006, 64, 325. 10. C.-C. Lee, G. Sui, A. Elizarov, C. J. Shu, Y.-S. Shin, A. N. Dooley, J. Huang, A. Daridon, P. Wyatt, D. Stout, H. C. Kolb, O. N. Witte, N. Satyamurthy, J. R. Heath, M. E. Phelps, S. R. Quake, H.-R. Tseng, Science 2005, 310, 1793. 11. C. J. Steel, A. T. O’Brian, S. K. Luthra, F. Brady, J. Label. Compd. Radiopharm. 2007, 50, 308. 12. Chemistry in Alternative Reaction Media, D. J. Adams, P. J. Dyson, S. J. Tavener, Wiley, 2004, 13. T. Ohkouchi, T. Kakutani, M. Senda, Bioelectrochem. Bioenerg. 1991, 25, 81. 14. R. Lahtinen, D. J. Fermin, K. Kontturi, H. H. Girault, J. Electroanal. Chem. 2000, 483, 81. 15. Chemistry in Alternative Reaction Media, D. J. Adams, P. J. Dyson, S. J. Tavener, Wiley, 2004. 16. B. Ahmed-Omer, J. C. Brandt, T. Wirth, Org. Biomol. Chem. 2007, 5, 733. 17. M. J. Earle, K. R. Seddon, Pure Appl. Chem. 2000, 72, 1391. 18. H. Zhao, S. V. Malhotra, Aldrichim. Acta 2002, 35, 75. 19. S. J. Tavener, J. H. Clark, J. Fluor. Chem. 2003, 123, 31. 20. X. Hao, O. Yamazaki, A. Yoshida, J. Nishikido, Green Chem. 2003, 5, 524. 21. C. de Bellefon, N. Tanchoux, S. Caravieihes, P. Grenouillet, V. Hessel, Angew. Chem. Int. Ed. 2000, 39, 3442. 22. K. Mikami, M. Yamanaka, N. Islam, K. Kudo, N. Seino, M. Shinoda, Tetrahedron 2003, 59, 10593. 23. M. Ueno, H. Hisamoto, T. Kitamori, S. Kobayashi, Chem. Commun. 2003, 936. 24. K. Tanaka, S. Motomatsu, K. Koyama, K. Fukase, Tetrahedron Lett. 2008, 49, 2010. 25. W. N. Lau, K. L. Yeung, R. Martin-Aranda, Micro. Meso. Mat. 2008, 115, 156. 26. (a) E. Angeletti, C. Canepa, G. Martinetti, P. Venturello, Tetrahedron Lett. 1988, 29, 2261; (b) E. Angeletti, C. Canepa, G. Martinetti, P. Venturello, J. Chem. Soc., Perkin Trans. 1 1989, 1, 105. 27. M. Fernandez-Suarez, S. Y. F. Wong, B. H. Warrington, Lab Chip 2002, 2, 170. 28. H. Song, D. L. Chen, R. F. Ismagilov, Angew. Chem. Int. Ed. 2006, 45, 7336. 29. D. L. Chen, L. Li, S. Reyes, D. N. Adamson, R. F. Ismagilov, Langmuir 2007, 23, 2255. 30. (a) B. Ahmed, D. Barrow, T. Wirth, Adv. Synth. Catal. 2006, 348, 1043; (b) B. Ahmed-Omer, D. Barrow, T. Wirth, Chem. Eng. J. 135S, 2008, S280. 31. B. Ahmed-Omer, D. A. Barrow, T. Wirth, Tetrahedron Lett. 2009, 50, 3352. 32. H. Lu, M. A. Schmidt, K. F. Jensen, Lab Chip 2001, 1, 22. 33. T. Fukuyama, Y. Hino, N. Kamata, I. Ryu, Chem. Lett. 2004, 33, 1430. 34. (a) H. Maeda, H. Mukae, K. Mizuno, Chem. Lett. 2005, 34, 66; (b) H. Mukae, H. Maeda, K. Mizuno, Angew. Chem. Int. Ed. 2006, 45, 6558. 35. R. C. R. Wootton, R. Fortt, A. J. de Mello, Org. Process Res. Dev. 2002, 6, 187. 36. K. J€ahnisch, U. Dingerdissen, Chem. Eng. Technol. 2005, 28, 426. 37. Y. Matsushita, S. Kumada, K. Wakabayashi, K. Sakeda, T. Ichimura, Chem. Lett. 2006, 35, 410. 38. G. Takei, T. Kitamori, H.-B. Kim, Catal. Commun. 2005, 6, 357. 39. Y. Matsushita, T. Ichimura, N. Ohba, S. Kumada, K. Sakeda, T. Suzuki, H. Tanibata, T. Murata, Pure Appl. Chem. 2007, 79, 1959. 40. Y. Matsushita, N. Ohba, S. Kumada, T. Suzuki, T. Ichimura, Catal. Commun. 2007, 8, 2194. 41. S. Meyer, D. Tietze, S. Rau, B. Sch€afer, G. Kreisel, J. Photochem. Photobiol. A 2007, 186, 248. 42. K. Sakeda, K. Wakabayashi, Y. Matsushita, T. Ichimura, T. Suzuki, T. Wada, Y. Inoue, J. Photochem. Photobiol. A 2007, 192, 166. 43. A. Sugimoto, Y. Sumino, M. Takagi, T. Fukuyama, I. Ryu, Tetrahedron Lett. 2006, 47, 6197. 44. C. de Bellefon, N. Pestre, T. Lamouille, P. Grenouillet, V. Hessel, Adv. Synth. Catal. 2003, 345, 190. 45. C. J€onsson, S. Lundgren, S. J. Haswell, C. Moberg, Tetrahedron 2004, 60, 10515. 46. K. Koch, R. J. F. van den Berg, P. J. Nieuwland, R. Wijtmans, H. E. Schoemaker, J. C. M. van Hest, F. P. J. T. Rutjes, Biotech. Bioeng. 2008, 99, 1028. 47. S.-P. Kee, A. Gavriiliidis, J. Mol. Catal. A 2007, 263, 156. 48. E. R. Murphy, J. R. Martinelli, N. Zaborenko, S. L. Buchwald, K. F. Jensen, Angew. Chem. Int. Ed. 2007, 46, 1734.

Homogeneous Catalysis Using Microreactor Technology

425

49. H. Kawanami, K. Matsushima, M. Sato, Y. Ikushima, Angew. Chem. Int. Ed. 2007, 46, 5129. 50. M. Sato, K. Matsushima, H. Kawanami, Y. Ikushima, Angew. Chem. Int. Ed. 2007, 46, 6284. 51. F. Benito-Lopez, R. M. Tiggelaar, K. Salbut, J. Huskens, R. J. M. Egberink, D. N. Reinhoudt, H. J. G. E. Gardeniers, W. Verboom, Lab Chip 2007, 7, 1345. 52. A. A. Kulkarni, K.-P. Zeyer, T. Jacobs, A. Kienle, Ind. Eng. Chem. Res. 2007, 46, 5271. 53. R.-J. Hu, M. Lei, H.-S. Xiong, X. Mu, Y.-G. Wang, X.-F. Yin, Tetrahedron Lett. 2008, 49, 387. 54. W. Ferstl, M. Schwarzer, S. Loebbecke, E. Fritz-Langhals, J. Stohrer, Chem. Eng. J. 135S, 2008, S292. 55. D. Huang, Y. C. Lu, Y. J. Wang, L. Yang, G. S. Luo, Ind. Eng. Chem. Res. 2008, 47, 3870.

15 Catalyst Immobilization Strategy: Some General Considerations and a Comparison of the Main Features of Different Supports Franco Cozzi Dipartimento di Chimica Organica e Industriale, Universita’ degli Studi di Milano and CNR-ISTM via C. Golgi 19, I-20133, Milano, Italy

15.1 Introduction The connection of a catalyst to a support, a procedure that, in a very broad sense, has been referred to with the expression ‘catalyst immobilization’, represents the simplest and more widely used approach to catalyst recovery and recycling. The design of a successful immobilization must take into account several factors, and can be directed by a variety of goals that, in principle, can be different from recovery and recycling. This can be the case for instance in the immobilization of a catalyst that is employed in a specific process at the industrial level. In this case, it is the nature of the process itself that dictates the design of the immobilization step in its various aspects. Thus, the choice of the support, the selection of the type of ‘bond’ connecting the catalyst to the support, the identification of a suitable method for catalyst separation from the product and for catalyst recycling are all predetermined by the peculiarity of the process, that is the chemical nature of reagents, products, and solvent, and the reaction conditions. On the other hand, when a catalyst is developed for the first time in its immobilized form, for instance at the academic level, versatility rather than specificity is the main goal. In this case, the identification of a


428


supported catalyst capable of satisfactorily promoting a variety of reactions while tolerating different substrates and reactants is considered a more coveted achievement than the development of a supported system exceptionally active in a single reaction. This dichotomy of interests can serve to exemplify the different role of academic and industrial research in the context of catalyst immobilization, with the former having the task of providing the widest range of options among which the latter can select for fulfilling its specific interests. Independently of the reasons behind the decision to immobilize a catalyst, however, some general considerations must precede and steer the immobilization process. The first part of this chapter will discuss these considerations, with the aim of providing some general guidelines to catalyst immobilization. In the second part, a comparison between different supports will be attempted discussing two examples of widely employed catalysts, namely proline derivatives and metal complexes of bis(oxazolines), for which a number of different supports have been employed. These catalysts have also been selected as paradigmatic representatives of immobilized organic and organometallic catalysts.

15.2 General Considerations on Catalyst Immobilization 15.2.1 Prerequisite Conditions for Immobilization The decision to immobilize a catalyst on a support should be taken only when some general prerequisite conditions have been satisfied. Since generally the ultimate goal of the immobilization of a catalyst is its recovery and recycling, catalyst stability appears to be of paramount importance. In this respect it is worth mentioning that very seldom the stability of a (newly discovered) nonsupported catalyst is thoroughly assessed under the conditions (including temperature, solvent, and the presence of reactants and products) to which the catalyst is exposed during its standard applications. It is important to mention that catalyst’s instability is very common and, interestingly, this is not only true for organometallic catalysts, in which the metal species is not covalently bound to the ligand and hence is per se more prone to be removed from the complex giving a catalytically inactive (or less active) ensemble, but is very frequent also in the case of apparently more stable organic catalysts. To illustrate this point let us consider protonated imidazolidin-4-one 1 and its supported analogue 2 (Figure 15.1). Compound 1 was originally proposed as an organic catalyst by MacMillan and coworkers,1 and had found widespread application in a number of relevant reactions.2 This fact led to the development of a version of this catalyst supported on the monomethyl ether of poly(ethylene glycol) having average molecular weight 5000 Da (MeOPEG5000, catalyst 2)3 and other supports.4 Studying the 1,3-dipolar cycloadditions of N-benzyl-C-phenylnitrone with acrolein (Figure 15.1, equation 1),5 it was shown that the two systems behaved quite similarly, the PEG-supported catalyst giving results slightly inferior to those observed with the nonsupported catalyst:6 chemical yield 71 vs 80%; trans/cis ratios 85/15 vs 86/14; enantiomeric excess (ee) of the trans isomer 87 vs 90%. After recovery of PEG-supported 2 by precipitation with diethylether and filtration, the dried catalyst was recycled twice in the same reaction to afford the products with constant levels of diastereo- and enantioselectivity but in clearly diminishing chemical yields (roughly 15% per cycle).5 With the goal of explaining this behaviour, the supported

Catalyst Immobilization Strategy Bun

Me O

N N H2

429

O

Me

N

Me

Me Me

N H2

X

BF4 OPEG5000OMe O

1

2 Bn

Ph

H

+ Bn

N

O

O

1 or 2

N

Bn

O +

N

O (Eq. 1)

Ph

Ph

CHO

CHO trans

cis

Figure 15.1 Structure of MacMillan’s catalyst and of its PEG-supported analogue

catalyst was checked by 1 H NMR after each recovery to show degradation, increasing after each cycle. The NMR analysis was suggestive of an imidazolidinone ring opening process. In order to check whether the instability of catalyst 2 was induced by the presence of the polymer, the triflate salt of the nonsupported imidazolidinone 1 was kept for 120 h at 24 C in the reaction solvent (a 95/5 CD3CN/D2O mixture) both in the presence and in the absence of the bismethyl ether of PEG2000. In neither case was degradation observed by NMR, thus suggesting that the polymer was not playing a leading role in inducing the instability of the supported catalyst. These findings pointed to catalyst degradation induced by the presence of the reagents, and indeed further NMR analysis showed extensive degradation of the supported catalyst in the presence of acrolein while the nitrone did not exert any effect on the catalyst stability (more hindered aldehydes, such as cinnamaldehyde, were less noxious than acrolein). In agreement with the results of these experiments, it was observed that the nonsupported catalyst also showed a marked instability and decreased chemical efficiency when recycled in reactions involving acrolein.5 Obviously, had the poor stability of 1 been established before immobilization, the development of a supported catalyst could have been considered not worth the effort. High catalytic efficiency seems to be another prerequisite condition for catalyst immobilization. This statement seems so obvious to be redundant, but its importance cannot be overemphasized once it is considered how strong an effect the attachment of the support can exert on the catalyst (as for instance in the event of the heterogenization of a soluble catalyst by attachment to an insoluble support). In general, the catalytic site of a supported system should be expected to be less accessible than that of a nonsupported one.7 As a result of the immobilization, a decrease in chemical activity should be expected and this is indeed what happens in the majority of cases.8 Thus, only very active catalysts have

430

Recoverable and Recyclable Catalysts NH2

Br

NBu3

H OPEG5000OMe

O

Si O O O

O H

O 3

4

Figure 15.2 Supported catalysts displaying a cooperative effect with the support

the best chances to preserve their reactivity or to have it just marginally depressed upon immobilization. For sake of completeness, however, it must be remembered that in some instances cooperativity between the catalyst and the support has been observed to take place and actually boost the activity of the supported catalyst. This is the case for instance of the tributylbenzylammonium catalyst 3 supported on MeOPEG5000 depicted in Figure 15.2. Catalyst 3 has been found to be an exceptionally active promoter for a variety of standard transformations occurring under phase transfer catalysis conditions.9 Its activity was shown to be superior to that of the structurally related nonsupported catalyst tributylbenzylammonium bromide in the reactions involving the use of solid NaOH under solid–liquid conditions (for instance, N- and O-benzylation of pyrrole and phenol; dichlorocyclopropanation of styrene). It seems possible that the catalyst benefits from the cooperation of the support, because the polyethylenoxy chain of PEG can complex the alkaline cation of the base helping the transfer of the HO counterion in the organic phase. A different type of activation by the support can be found in the aminocatalyst 4 supported on mesoporous silica nanoparticles (Figure 15.2).10 In this example, the free silanol groups residing on the surface of the support were found to facilitate the Henry addition of nitromethane to aldehydes catalysed by the immobilized primary amino group. Activation was hypothesized to occur by hydrogen bonding between the silanol donors and the aldehyde acceptor, an event that promoted the formation of an iminium ion more prone to attack of nitromethane than the aldehyde itself.11 The catalyst’s stereoselectivity is another frequent victim of immobilization, because it might happen that, because of the presence of the support, the substrate finds more difficulties in adopting the required orientation for the transformation to occur in a stereoselective fashion. Clearly, the immediate surroundings of the active site can deeply affect the approaching event, exerting undesirable stereochemical effects. In this respect, it must be said that it is more likely that a very active nonsupported catalyst will maintain its chemical activity upon immobilization than a highly stereoselective catalyst will preserve intact most of its stereodiscriminating ability. Still, the idea that the support can exert some sort of control on the relative orientation of the substrate/catalyst pair has been used in a positive sense to enhance the stereoselectivity of various catalytic reactions.12 Indeed, in 1999, Thomas and coworkers reasoned that spatial confinement of transition states formed around a chiral catalyst’s active site could provide a new method of boosting

Catalyst Immobilization Strategy Me Me

N

H

Me

Me

N Me

Ph2P

C a b o s i l

Fe

Cl2Pd

Me

N

Me

N Me

Ph2P

P Ph2

5b

H

Fe

Cl2Pd

P Ph2

5a

MCM-41

N

H

431

Me

N Me

Ph2P

Fe

Cl2Pd P Ph2

5c Ph Ph

OAc BnNH2 5a-5c

Ph

NHBn + 6

(Eq. 2) NHBn 7

Figure 15.3 Spatially confined catalyst 5c and its nonconfined analogues 5a,b

the enantioselectivity of a supported catalyst.13 In a revealing experiment, they compared the behaviour of ferrocenylphosphine/Pd(II) catalysts 5a (homogeneous), 5b (supported on the convex surface of the nonporous, high-surface area silica Cabosil), and 5c (supported on the concave surface of mesoporous silica MCM-41) in promoting the allylic amination reaction between cinnamyl acetate and benzylamine (Figure 15.3, equation 2). This reaction can afford the linear product 6 and the branched chiral adduct 7. It was discovered that: catalyst 5a led exclusively to the formation of amine 6; catalyst 5b to a largely unbalanced mixture of 6 and 7 (98:2 ratio); and catalyst 5c to an equimolar mixture of these compounds. More relevantly, while in the presence of catalyst 5b compound 7 was obtained in a moderate 43% ee, the use of spatially confined catalyst 5c led to 7 in a remarkable >99% ee, showing the ability of 5c to control both regio- and stereoselectivity of the process. In concluding this section, it can be said that once the prerequisite conditions for catalyst immobilization mentioned above (namely catalyst’s stability and efficiency) are properly

432


taken into account and the choice of the support can be as free as possible, the immobilization step rather than an unavoidable drawback can be considered an opportunity not only to improve the performance of the catalyst but also to open access to novel reactivity. The example of supported bifunctional catalysis depicted in Figure 15.4 can serve to demonstrate this point.14

H2N HN O

NH2

NH

NH

Si

Si O m O O

O

O

8a: m = 1, n = 4 8b: m = 1, n = 1 8c: m = 4, n = 1 8d: m = 1, n = 0 8e: m = 0, n = 1 O

n

= mesoporous silica NO2

N

H

O N 8a

NO2

Me

O H

H

N H2C

ArCHO

(Eq. 3)

Me2CO

Si

Si O

O

O

O

n

OH O

O m Me

O

9 HO

NO2 CN

CHO Eq. 5

O2N

O2N

Eq. 4

O2N

Figure 15.4 Cooperative catalysis in mesoporous silica-supported bifunctional catalysts

Catalyst Immobilization Strategy

433

In an attempt to reproduce the cooperative general acid and basic catalysis in the active sites of enzymes, in 2005 Lin and coworkers synthesized the mesoporous silica nanospheres 8a–e decorated with different loading of functional groups, reported in Figure 15.4.15 These compounds were obtained by a simple three-component co-condensation, a procedure that, by varying the relative amounts of the reagents, easily allowed these organic–inorganic hybrid materials to be obtained with well-defined structural features (macroscopic dimensions, surface area, pore size) and the indicated ratios of ureido and amino functional groups (as determined by 13 C and 29 Si NMR spectroscopy). The catalytically active residues were chosen with the aim of promoting the nucleophilic addition of acetone, nitromethane, and trimethylsilycyanide to 4-nitrobenzaldehyde (Figure 15.4, equations 3, 4, and 5, respectively). The role of the ureido group was that of activating the aldehyde’s carbonyl group by double hydrogen bonding. The secondary amine was responsible for activation of the nucleophile: by enamine formation in the case of acetone; by deprotonation in the case of nitromethane; by generation of a more reactive hypervalent silicon species in the case of trimethylsilycyanide. Cooperation between the catalytic sites of 8a–c was expected because of their proximity on the nanosphere surface, as depicted in model 9 in the case of the aldol reaction. The observation of enhanced rates with respect to the reactions catalysed by nanospheres 8d and 8e, decorated with only one of the two catalytic groups, was also anticipated. Catalyst 8a featuring a 1:4 ratio of secondary amine:ureido residues, proved to be the most effective in all three reactions. The aldol condensation carried out in acetone at 50 C in the presence of 5 mol% of bi-functional catalyst 8a occurred at a rate that was 2.0, 2.6, and 4.2 times faster than that observed with catalysts 8b, 8c, and 8d, respectively (compound 8e was not active in this process). The Henry reaction required only 1 mol% of catalyst 8a to proceed in high yield in nitromethane at 90 C. In this case the increases in reaction rate with respect to catalysts 8b, 8c, 8d, and 8e were 1.3, 1.9, 2.3, and 22 times, respectively. A reason for this less marked difference could be found in the fact that, contrary to the aldol process, in the nitroaldol condensation the mono-functionalized ureido catalyst 8e also showed some activity. The addition of trimethylsilycyanide gave results similar to the Henry reaction. Also in this case only a 1 mol% of 8a was enough for the reaction to proceed effectively in toluene at 50 C. The reaction catalysed by 8a was found to be 1.6, 2.5, 2.5, and 6.0 times faster than those observed with catalysts 8b, 8c, 8d, and 8e, respectively. Thus, in this example the support played a central role in securing the desired result, and it can be easily anticipated that future developments in this sense will follow. An obvious extension of this approach can be envisaged in the field of stereoselective catalysis, by assembling a system in which two chiral catalysts cooperate to enhance the stereocontrol of a given transformation. Incidentally, the synthesis of a chiral bi-functional catalyst by the co-condensation of different residues carrying mono-functional active sites to afford for instance a chirally modified version of silica nanospheres 8 can represent a simpler approach than a standard synthetic procedure to bifunctional catalysts. 15.2.2 Reasons Justifying Immobilization A number of reasons justify the immobilization of a catalyst. In addition to catalyst recovery and recycling, a simpler separation of the catalyst from the reaction mixture and an easier isolation of the reaction products are the most important.

434


In the context of separation, immobilization is often used to change the catalyst’s solubility properties. In this case, the support acts as a ‘solubility device’. Without entering for the time being in the discussion about the relative merits of soluble and insoluble catalysts, immobilization on an insoluble support such as a siliceous material or a crosslinked polystyrene can make the catalyst insoluble in the reaction medium and thus physically removable by filtration. In contrast, immobilization can allow the catalyst to be soluble in the reaction medium as long as the reaction proceeds, and then to become insoluble when the reaction is over. The switch in solubility can be induced by a change in the medium polarity as in the case of PEG16 or, even more simply, by changing the temperature of the reaction mixture, as in the case of thermomorphic polymers.17 Another possibility is offered by linear and soluble poly(4-t-butylstyrene). In this case, the reaction is carried out in an apolar solvent where the catalyst is made soluble by the support. At the end of the reaction the product is removed by extraction with a polar solvent while the catalyst remains in the apolar medium and is ready to be recycled. This technique has been dubbed ‘latent biphasic separation’.18 The possibility of designing an immobilization strategy that allows the catalyst to permanently remain in the reactor while unreacted materials, products, and byproducts are readily removed seems to be the best possible approach to catalyst recycling. This is the case for instance of continuous flow methods.19 The retention of the catalyst inside the reaction vessel can be achieved by different techniques ranging from ultrafiltration through a molecular weight-selective membrane to immobilization on a silica gel column. It is important to note that, under continuous flow conditions, product separation from the catalyst, catalyst recovery, and catalyst recycling are very conveniently performed in a single operation. The validity of this approach is demonstrated by the fact that one of the highest recycling numbers for an organic catalyst (60 cycles) has been reported for a spectacular enantioselective b-lactam synthesis carried out under the continuous flow mode.20 The possibility that the catalyst slowly decomposes under the reaction conditions releasing substances contaminating the product can be another reason for immobilization. In discussing this point, a distinction must be made between organometallic and organic catalysts. In the case of the former, the most common form of catalyst decomposition is metal leaching, that is the loss of the metal from the complex constituted by the supported ligand and the metal itself. In the case of leaching, even if a supported organometallic catalyst can be more stable than its nonsupported counterpart, immobilization is not a solution to catalyst decomposition. In the case of organic catalysts on the other hand, catalytic systems do exist that slowly release trace amounts of byproducts that must be separated from the products. If decomposition is slow and the catalyst is very chemically active, this phenomenon only marginally affects the efficiency of the catalyst and thus does not prevent recycling, but product purification remains a problem. This is the case for instance of the reactions involving the oxidation catalysts based on the oxo-ammonium ion derived from 2,2,6,6-tetramethylpiperidine-1-oxyl species (TEMPO), and of the catalysts embedding the porphyrin nucleus. These compounds are known to release highly coloured materials that are very difficult to dispose of. Immobilization of the catalyst can solve this problem, provided that also the decomposed material are tethered to the support and thus can be removed from the reaction medium during catalyst separation.


435

Immobilization can be an appealing option also if the catalyst is expensive, either per se or because it has been obtained after a complex synthesis. It is important, however, that the synthesis of the supported catalyst should not add further synthetic complexity (for instance by requiring protection/deprotection steps) or require as starting materials very expensive reagents. In this line, the synthesis of the supported catalyst should be designed so that it could exploit as starting material a catalyst precursor comparable in cost and synthetic complexity with the compound used for the synthesis of the nonsupported catalyst (for simplicity, the cost of the support and of the chemical entity employed to connect the catalyst to the support are not considered in this discussion). To explain this point let us consider the introduction of an hydroxy group on a nonsupported catalyst as an handle for its immobilization. This is a very common practice that relies on the use of easy-to-form ether- or ester-type bonds in the crucial immobilization step. In the case of the immobilization of MacMillan’s catalyst (Figure 15.1) the anchoring procedure required to replace the (S)-phenylalanine residue necessary for the synthesis of compound 1 with (S)-tyrosine as the starting material for the preparation of 2. Since the cost of tyrosine is about one half that of phenyalanine, the immobilization procedure could easily tolerate the introduction of the handle. In contrast, this was not the case for the first example 21 of immobilization of (S)-proline, in which this relatively cheap amino acid was replaced with the two-and-a-half times more expensive trans-(2S,4R)-4-hydroxyproline. Nevertheless, hydroxyproline remained a very popular starting material for proline immobilization (see below Section 15.3).22 This is even more surprisingly when one considers that supported versions of proline exist that have been shown to work very nicely as recoverable organocatalysts and the synthesis of which exploited unmodified proline as the starting material.23 The development of a supported catalyst is often considered convenient when the catalyst is employed in relatively large amounts ( 10 mol%). However, it must be remembered that generally the immobilization procedure largely increases the molecular weight of the catalyst and this has some impractical consequences that could make the immobilization effort meaningless. To illustrate this point, let us consider catalyst 3 (Figure 15.2) supported on a mono-functionalized PEG5000 sample. The related nonsupported catalysts tributylbenzyl ammonium bromide, has a molecular weight of 356 Da that becomes roughly 5500 Da in the immobilized species. If this catalyst must be used at the 10 mol% loading to promote the transformation of 1 mmol of a substrate whose molecular weight is 200 Da, then about 0.55 g of supported catalyst (compared with 0.072 g of the nonsupported catalyst) must be used to process 0.2 g of substrate, a clearly unpractical scenario. Several strategies have been used to circumvent this problem: among these, the use of supports with a high loading (expressed as mmol g1) of active sites seems to offer an obvious solution.24 However, it is not always true that a higher density of catalytic sites directly translates into a higher activity.25 Even if not central to the problem of catalyst recovery and recycling, it is worth mentioning in this section that immobilization of a catalyst can have a raison d’etre in facilitating the process of the catalyst’s optimization. Several examples of the generation of libraries of catalysts have been reported in the literature.26 The first example of the application of this methodology in the field of stereoselective organocatalysis by Jacobsen and coworkers remains a milestone highlighting the powerfulness of this approach.27

436


N

N Al

But

O

O

tBu

tBu But

10

Me R

But

N O

S N

N

H

H

N HO

But

OCOBut

11a, R = 1% cross-linked polystyrene 11b, R = H

Figure 15.5 Structure of catalyst 11a: an example of a catalyst immobilized to facilitate catalyst optimization

After having identified organometallic catalyst 10 (Figure 15.5) for the enantioselective Strecker-type hydrocyanation of the N-allyl imines of aromatic aldehydes in excellent yield (up to 91%) and ee (up to 95%), Jacobsen started a systematic modification of its salen structure to develop a thiourea-based Brønsted acid organic catalyst capable of the same accomplishment. The optimization of the catalyst structure began with a modification of the salen ligand by replacement of one of its imino functions with L-leucin containing dipeptide-like residues. The modifications were carried out on a catalyst precursor tethered to an insoluble polystyrene support and using the principles of combinatorial chemistry for sorting out the best aminoacid, diamine, and diamine-aminoacid linker combination. Even if the first obtained catalysts gave a low stereocontrol (19% ee) the screening of a small library of polymer-bound organic catalysts (48 members) led to the selection of a combination of aminoacid, chiral diamine, salicylaldehyde substituents, and aminoacid protecting group that improved the ee up to 55%. Based on this result, a third, larger library of 132 polymersupported members was prepared, the further screening of which led to the identification of the supported thiourea catalyst 11a and of its nonsupported counterpart 11b as the best ones. At a loading as low as 1 mol%, 11b promoted the hydrocyanation of N-allyl or N-benzyl imines derived from aromatic and aliphatic aldehydes and of some ketones in very high


437

yield and almost complete stereoselectivity. It is interesting to note that the soluble and the resin-bound catalyst performed equally well. Recovery and recycling of the supported catalyst was demonstrated to occur without any erosion of chemical and stereochemical efficiency over ten reaction cycles. The success of this methodology was made more complete by the fact that catalysts 11 are among the few chiral organic catalysts to be currently employed at the industrial level.2 15.2.3 A General Discussion on the Practical Aspects of Immobilization Several factors must be considered in practically dealing with the immobilization of a catalyst on a support in the best possible way. These include the choice of the support, the decision about where on the catalyst should be located the handle for the connection to the support, and the selection on what kind of linker, if any, must be inserted between the catalyst active site and the support. Given the countless examples of immobilized catalysts that have been reported in the literature, the viable options are far too many to allow the proposal of a comprehensive set of rules. Rather, the best immobilization procedure must be selected for each individual case, since this depends not only on the structure of the catalyst but also on the reaction the catalyst will promote, the physical and chemical properties of products and byproducts, and the mode of catalyst recovery and recycling. Still, some general considerations can be put forward as they are suggested by common sense and the results reported in the literature. The choice of the support is crucial, because the properties of the support influence every aspects of the catalyst behaviour. Relevant features of the support are: solubility profile, cost, commercial availability, degree of functionalization, and a possible involvement of the support backbone in the reaction (see above for examples discussing the latter issue). The solubility properties are the most important, the decision to develop a homogeneous or a heterogeneous catalytic system being the first to be made. However, even such a fundamental matter cannot be decided upon once and for all, and there cannot be a general consensus whether working with a soluble catalyst is better or worse than working with an insoluble one. On the basis of a far too simplistic and possibly na€ıve reasoning one can expect heterogeneous catalysts to be less reactive than homogeneous ones, but a number of results indicate that this is not always the case. On the other hand, insoluble catalysts seem to be more readily recovered and thus more easily recycled than soluble ones. By combining these two aspects, it can be anticipated that, if an insoluble system is selected, what is sacrificed in terms of reactivity can be gained in terms of simplicity of operation in recovery and recycling (and the other way around with a soluble system). Still, it must be remembered that catalyst recovery is not a mandatory condition for recycling, as demonstrated by the development of continuous flow methods in which the catalyst resides permanently in the reaction vessel. Moreover, a number of catalysts have solubility profiles that can be varied by changing medium polarity and reaction temperature, thus allowing to couple the advantages of homogeneous and heterogeneous systems and, in a sense, bypassing the problem of catalyst solubility. In evaluating the cost of the support one should always be aware of the fact that what is actually paid for a gram of support is not the only term to be considered. A better measure of

438


the cost of the support must take into account the balance between the cost and the number of functional groups present in a gram of support (the ‘loading’). It also seems reasonable that there should be some sort of correlation between the cost of the catalyst and that of the support, in the sense that the immobilization of an inexpensive catalyst on a very expensive matrix should be regarded as a nonsense. Also in this context, however, the goal of catalyst’s immobilization should guide the choice of the support. Thus, if the immobilization is required for simplifying the catalyst identification and optimization (for instance by a combinatorial approach, see above), the cost of the support is not a very relevant issue, given the relatively small amount of support actually employed. On the other hand, catalysts anchored to cheap supports have a better chance to be developed in multigram quantities and employed for practical applications. The importance of the commercial availability of the support goes well beyond practical reasons and often deeply influences the immobilization design. It is quite obvious that an organic chemist with limited or no experience in polymer or material chemistry will be very reluctant to tackle the synthesis of a complex support and, most of all, of its characterization, which often requires the use of nonconventional methods. On the contrary, most of the synthetic chemist’s efforts will be devoted to the identification of a good catalytic system to be anchored on one of the many commercially available supports. Moreover, it is generally believed that the commercial availability of the support is a prerequisite for the supported catalyst to have a chance to be adopted outside the laboratory of its inventor. Thus, the catalyst will be designed having in mind an available support and, as a consequence, the choice of the support has a fundamental impact on the design process. This attitude is extremely limiting, because on the one hand polymer and material chemists will tend to attach only known catalysts to custom-made and carefully crafted supports; on the other hand, synthetic chemists will anchor sophisticated catalysts to supports which, even if possibly not ideal, have the great advantage of being commercially available. While it is obvious that an interdisciplinary effort can provide a solution to this problem, it must be emphasized once again that the choice of the support should really be seen as a great opportunity for improving the catalyst properties rather than a limitation to imaginative catalyst design. The degree of functionalization is a very important factor in determining the choice of a support for catalyst immobilization. In principle, a high number of functional groups per gram should allow the introduction of a high number of catalyst sites and, accordingly, to decrease the mass of supported catalyst employed. As discussed in the previous section, the mass of immobilized catalyst necessary to promote a reaction is a problem with important practical consequences and must always be kept in mind. Moreover, it must be remembered that it is not always true that a higher density of catalytic sites directly translates into higher activity.25 First, the creation of a too crowded environment around the catalytic sites can prevent a facile approach of the reactants; secondly, the presence of the catalytic sites itself can have a strong influence on the polarity of the medium, again favouring or disfavouring reactants diffusion. Thus, in most cases the identification of the best loading is the result of a trial-and-error process, and no general rules can be proposed. Related to this topic is the fact that the support itself can possibly be involved in the reaction promoted by the supported catalyst. In this line, it is important to recognize the fact that generally the support largely outweighs the catalyst in molecular mass and thus its structure influences the environment around the active sites. For instance, polystyrene- and


439

Br N

O

R

H

N 12a, R = H O

12b, R =

PEG5000OMe

O Bn

O Ph

12 (10 mol%) N

Ph

Ph

OBut

N CsOH, BnBr

O OBut

(Eq. 6)

Ph

Figure 15.6 Cinchonium-derived chiral phase transfer catalyst and its PEG-supported analogue

silica-based supports are likely to create local environments of different polarity that can accelerate or slow down the catalysed reaction. In addition, as mentioned above, the support can also play a more decisive role interacting with the reaction partners in improving or diminishing the efficiency of the catalyst. While two different examples of support/catalyst cooperation have already been presented for systems 3 and 4 (Figure 15.2), an example of a negative effect exerted by the support is here discussed. On the basis of the high efficiency of the PEG-supported catalyst 3, the cinchona alkaloid-derived ammonium salt 12a employed by Corey and Lygo in the stereoselective alkylation of aminoacid precursors was immobilized on a modified PEG similar to that used in the case of 3 (Figure 15.6).28 The behaviour of the obtained catalyst 12b, however, fell short of the expectations. Indeed, while this catalyst (used in 10 mol% amount) showed good catalytic activity promoting the benzylation of the benzophenone imine derived from tert-butyl glycinate (equation 6) in 92% yield (solid CsOH, dichloromethane, 78 to 23 C, 22 h), the observed ee was only 30%. Even if this was increased to 64% by maintaining the reaction temperature at 78 C and prolonging the reaction time to 60 h, the top level of stereoselectivity obtained with the nonsupported catalyst could not be matched. The PEG support was considered to be, at least in part, responsible for these results. Indeed, it can be imagined that, by increasing the polarity around the catalyst, PEG prevented the formation of a tight ion pair between the enolate and the chiral ammonium salt, the formation of which is regarded as crucial for achieving high stereocontrol. Moreover, PEG enhanced the solubility of the inorganic cation in the organic phase leading to a competing nonstereoselective alkylation occurring on the achiral caesium enolate. To check the validity of these hypotheses control experiments were carried out by performing

440


the reaction with the nonsupported catalyst 12a in the presence of the bis-methylether of PEG2000. The observed ee was 65%, a value that was in good agreement with that observed with catalyst 12b but largely inferior to the >90% ee easily achieved in the absence of PEG. While in retrospect one can say that the behaviour of the chiral supported catalyst 12b could be at least in part anticipated on the basis of what observed with achiral 3, these results provided another demonstration of how dangerous generalizations can be in the identification of a good catalyst/support pair. As mentioned at the beginning of this section, the decision about where on the catalyst should be located the handle for the connection to the support is also quite relevant. A very common approach is based on the assumption (in principle quite reasonable) that the support should exert the minimum effect on the catalyst, and accordingly the longer the distance between the catalyst and the support, the better the chances for the supported catalyst to mimic the behaviour of the nonsupported one. This line of reasoning also suggests that bonds of relatively low conformational mobility should be preferred for the connection in order to avoid that the active part of the catalyst could fold back onto the support. The vast majority of supported catalysts, and especially chiral ones, have been developed following these principles. However, a too strict observance of this approach can entail some drawbacks. One can be that the immobilization requires very expensive starting materials. Another, that a relatively long sequence of synthetic modifications of the nonsupported catalyst could be necessary to get a compound suitable for immobilization. In addition, literature data suggest that the importance of the principle of maximum separation between the catalyst and the support should not be overestimated. An example in this line is provided by the chiral DMAP analogues reported in Figure 15.7 and employed in the kinetic resolution of secondary alcohols by enantioselective acylation.29 Researchers at GSK described a family of chiral acylating catalysts based on the Me

H N

N O

N

OH

OPDMAB

cis racemic

13a,

= low loading polystyrene

13b,

= high loading polystyrene

13c,

= Wang resin

(i PrCO)2O 13 (5 mol%)

OCOPri

OH

(Eq. 7)

+ OPDMAB

OPDMAB

PDMAB = 4-dimethylaminobenzoyl

Figure 15.7 Chiral supported catalysts for the kinetic resolution of secondary alcohols


441

N-40 -pyridinyl-a-methylproline structure capable of promoting the kinetic resolution of alcohols with high level of enantioselectivity. The ready availability of these compounds suggested exploiting the presence of the easily functionalizable proline carboxy function to immobilize these catalysts on different polymer supports. Among others, derivatives 13a–c were prepared connecting N-40 -pyridinyl-a-methylproline to low- and high-loading polystyrene, and to Wang resin by standard condensation methods. These compounds were tested as insoluble catalysts (5 mol%) in the kinetic resolution of racemic cis-1,2-cyclohexanediol mono-4-dimethylaminobenzoate (Figure 15.7, equation 7) carried out with a deficiency of iso-butyric anhydride in dichloromethane (room temperature, 16 h). By stopping the reaction at about 50% conversion, it was possible to recover the unreacted ()-alcohol in 75% ee. This was increased to 93% by allowing the reaction to proceed up to 67% conversion. No appreciable difference in chemical or stereochemical efficiency was observed as a function of the loading or the structure of the polymeric support. In this case the distance between the catalyst active site (the pyridine nitrogen) and the connection to the support is relatively short. This choice turned out to be beneficial to the success of the catalyst, since the amide proton is believed to play an active and decisive role in promoting the stereoselectivity of the reaction (by H-bonding the fast reacting alcohol) and the polymer backbone, residing in the proximity of the active site, can provide additional stereodiscriminating effects. The problem of securing a suitable separation between the active site and the support to enhance catalyst accessibility can also be approached by the insertion of a linker or spacer. The use of this topological device, originally proposed by Montanari and coworkers as a tool to improve the reactivity of phase transfer catalysts immobilized on cross-linked polystyrene,8 has become a general feature of supported catalysts ever since. It is interesting to note that the insertion of a spacer can solve problems of catalyst reactivity not only related to accessibility. For instance, a linker can help to create a microenvironment around the catalyst active site more suitable to the catalytic activity than that provided by the support. This was the case, for instance, for the just mentionedsupported ammonium and phosphonium salts,8 where the insertion of linear n-dodecyl linker chains between the catalyst and the support increased the lipophilicity of the catalyst enhancing its reactivity up to four times. Remarkably, and quite surprisingly, the beneficial effects of a separation between catalyst and support have been shown also in the case of soluble supports. For instance, the catalytic activity of the PEG-supported phase transfer catalyst 3 (Figure 15.2) was shown to be more than twice that of the catalyst in which the benzylammonium group was directly connected to the PEG chain.9 Finally, it must be mentioned that the accessibility of the catalyst active sites can also be enhanced by moving them to the surface of the solid support without making recourse to a linker. This possibility was recently demonstrated in the case of some polystyrene-supported phase transfer catalysts by changing the procedure for the synthesis of the backbone.30 Indeed, the delayed addition of 4-chloromethylstyrene to the co-polymerization mixture containing styrene and divinylbenzene led to an increase of the surface-exposed chloromethyl groups that were eventually exploited for catalyst attachment. By doubling the number of the accessible catalyst sites, a doubled catalyst activity was observed. However, attempts at increased accessibility did not result in further improvement in the catalytic activity.

442


15.3 Comparison of Different Supports Employed for the Immobilization of Proline Proline can be regarded as the paradigmatic organic catalyst. This is essentially due to two main reasons. Firstly, the rediscovery of proline as catalyst for the direct stereoselective aldol reaction in the year 2000 (and the subsequent extension of proline and proline-derived catalysts to a number of other reactions)31 is generally considered the starting point of the current gold rush in organocatalysis.2 Secondly, thanks to its low molecular weight, simple structure, high stability, nontoxicity, and extremely low cost, proline displays all the peculiar features that concurred to define an organic catalyst as a highly preferable alternative to an organometallic catalyst. Even if proline was first immobilized on 1% cross-linked polystyrene as early as in 1985,32 more frequent and extensive immobilization efforts began to appear in the literature only after 2001, when a MeOPEG5000-supported version of proline was described and employed as catalyst for the cross-aldol reaction.33 The relevance of supporting a catalyst as cheap as proline (which, moreover, enjoys a solubility profile that facilitates its recovery also when nonimmobilized) has often been questioned. Nevertheless, the fact that in several instances proline is employed at high catalyst loading and the hope of modifying and possibly improving the catalytic behaviour of this important and versatile catalyst, were (and still are) invoked to justify the immobilization. Accordingly, research devoted to supported proline enjoys a continuing interest.22 In the following sections different examples of supported prolines will be examined. This survey, however, will not cover every example of application of these organocatalysts, but will rather be focused on the identification of the relative merits of the various supports with particular interest being devoted to catalyst recovery and recycling. To obtain a more meaningful comparison, the use of differently immobilized prolines in the same benchmark reaction (that is, the aldol addition of ketones to aldehydes) will be discussed. 15.3.1 Organic Supports Polymers As mentioned above, a MeOPEG5000-supported version of proline was described and employed as catalyst back in 2001.21,33 The soluble catalyst 14 (Figure 15.8), obtained by connecting trans-4-hydroxyproline to PEG by a succinate spacer, promoted the reaction of acetone with some aromatic aldehydes (equation 8) affording the products in yields and ee similar to those observed with proline itself under the same experimental conditions (25–30% mol of catalyst, DMSO or DMF as solvent, room temperature, long reaction times). Extension of the use of this catalyst to hydroxyacetone as the aldol donor and to aliphatic aldehydes and imines as the acceptors was also possible. When the crucial issue of recycling was addressed, however, some problems became apparent. Indeed, when the catalyst, readily recovered by precipitation with diethylether, was employed in subsequent cycles almost constant ee were observed, but the chemical yield of the reaction was appreciably eroded (for the reaction of equation 8 performed on 4-nitrobenzaldehyde, the yield decreased from 70 to 50% in three cycles). The trend observed upon recycling in the aldol and aza-aldol processes was also a common feature of conjugate addition reactions promoted by 14, which, however, afforded the


443

O O

O O

COOH N H = MeOPEG5000

14,

OH O

O

14

+ 4X-C6H4CHO

(Eq. 8) X

R

R

R = H, OH; X = NO2, H, Br O O

X O

H N COOH N H

15a,

= MeOPEG2000; X = none

15b,

= MeOPEG5000; X = none

16,

= MeOPEG5000; X = 4NH-C6H4-SO2

Figure 15.8 PEG-supported proline-derived catalysts

products in lower ee than those observed with proline as the catalyst.34 In an attempt to bypass this problem, the soluble PEG-supported prolines 15a,b and 16 (Figure 15.8) were synthesized and tested in the conjugate addition of cyclohexanone to nistrostyrene.35 The use of these catalysts was beneficial in that yields and ee higher that those found with 14 were observed. However, the problem of a decrease of chemical yield upon recycling remained unsolved. As a whole, the results obtained with the PEG-supported catalysts showed that these suffered from an inability to preserve catalytic efficiency upon recycling. This negative behaviour was observed also in the case of other PEG-bound organic catalysts.36 After the pioneering work of Takemoto in 1985, insoluble polystyrene-supported prolines have recently become extensively studied catalysts, especially by the groups of Pericas and Gruttadauria. Starting from N-Boc–trans-4-hydroxyproline, catalyst 17a (Figure 15.9) was obtained in four steps by Pericas and coworkers and shown to effectively promote the reaction between cyclohexanone and benzaldehyde (equation 9) in water as the solvent and in the presence of the bismethylether of PEG2000 as a diffusion-facilitating additive (75% yield, 96:4 anti:syn ratio, 96% ee for the anti isomer, room temperature, 60 h).37 Catalyst 17a, recovered by filtration and reactivated by drying, was used in three cycles without observing any decrease in chemical yield or stereoselectivity. The role of the support was investigated in a subsequent paper in which the polystyrene support was modified to obtain catalysts 19–21. The behaviour of these catalysts was compared with that of 17 and of its nonsupported counterpart 18 in the reaction of equation 9.38 First, it was

444


N

N

N

O

17a,

COOH N H = 1% cross-linked polystyrene

17a,

= 2% cross-linked polystyrene

18,

= Ph

19,

= PEG-grafted polystyrene

20,

= Macroporous polystyrene N

N

N

O

21,

COOH N H = 1% cross-linked polystyrene

O

O cat (10 mol%)

OH Ph

+ PhCHO

(Eq. 9)

H2O, RT, 24 h

S O COOH N H 22,

= 1% cross-linked polystyrene H N

spacer O

COOH N H

= linear polystyrene 23a, spacer = NH-CO-(CH2)223b, spacer = NH-CO-(CH2)423c, spacer = 4O-C6H423d, spacer = 4NH-CO-C6H4-

Figure 15.9 Insoluble and soluble polystyrene-supported proline-derived catalysts


445

discovered that 17 performed much better than 18, thus suggesting that the polystyrene backbone in 17 can act as a sort of hydrophobic pocket, as that present in aldolase I-type enzymes. Support to this hypothesis was collected when it was found that neither the insertion of an hydrophilic PEG fragment between the polymer an the catalyst (as in 19), nor the use of a macroprorous resin (as in 20) led to any improvement with respect to 17. Eventually, direct connection of the spacer’s triazole moiety to the proline nucleus, as in 21, afforded a very active and stereoselective catalyst (70% yield, 93:7 anti:syn ratio, 97% ee for the anti isomer, room temperature, 12 h), that could be used also at a very low loading (1 mol%). An explanation for this observation was found in the fact that, whereas with catalysts 17, 19, and 20 the reaction was a multiphase system, with 21 a gel-like single phase was formed with the resin showing a high water content. The strong tendency of 21 to swell in water (and thus to become more accessible to reagents) was ascribed to the formation of a water-mediated hydrogen-bond network between the triazole moiety and the proline nucleus. Remarkably, catalyst 21 could be recycled five times without any appreciable loss in yield and stereoselectivity. In 2007 Gruttadauria and coworkers developed the insoluble catalyst 22 (Figure 15.9) starting from mercaptomethyl-substituted 1% cross-linked polystyrene and a styrene derivative of trans-4-hydroxyproline.39 In the standard addition of acyclic and cyclic ketones to aromatic aldehydes, carried out in water at room temperature for long reaction times, this catalyst (10 mol%) performed nicely, affording the aldol products in up to 98% yield, high anti selectivity ( 96:4), and up to 98% ee (for the major anti isomer). As the catalysts described by Pericas, also derivative 22 could simply be recovered by filtration and recycled five times affording the products with constant levels of chemical yield and stereocontrol. Thus, independently of the nature of the spacer connecting proline to the polystyrene backbone, all of the insoluble supported catalysts depicted in Figure 15.9 displayed recovery and recycling properties superior to those observed for the related soluble PEG-supported ones. The structural gap between PEG- and cross-linked polystyrene-supported catalysts was bridged when cis-4-aminoproline was anchored on linear polystyrene to afford the soluble catalysts 23a–d reported in Figure 15.9.40 In the standard reaction between cyclohexanone and aromatic aldehydes (carried out at room temperature in a DMF:water 15:1 mixture as solvent) the use of these catalysts (5 mol%) allowed to isolate the products in high yield ( 94%) and stereoselectivity (anti:syn ratios up to 95:5; up to 96% ee for the major isomer). When recycling was studied, it was found that the catalysts, recovered by precipitation with diethylether followed by filtration, could be recycled for a few cycles, promoting the reactions in constantly high ee but slowly decreasing chemical yields. Thus, the behaviour of the soluble catalysts 23a–d was more similar to that of their PEG-supported analogues than to that of their insoluble counterparts supported on cross-linked polystyrene. It seems possible that the different techniques employed for catalyst recovery, (precipitation and then filtration for soluble catalysts vs simple filtration for insoluble ones) can play a role in determining the recycling efficiency. To conclude this section on organic polymer-supported prolines, it must be remembered that several catalysts featuring a proline unit as the N-terminus of short peptide chains were synthesized on insoluble supports containing the polystyrene skeleton.22 These prolinamidic catalysts performed nicely in the standard aldol reaction between acetone and aldehydes. Recovery and recycling of these catalysts was studied only in the case of

446

Recoverable and Recyclable Catalysts Ph

H N

O

N O

H N

H 4HO-Ph

O

N H

24 COOH O

H N

N O

O N H N

H 25

=Polystyrene-PEG- (Tentagel)

Figure 15.10 Supported proline-containing peptidic catalysts

catalysts 24 and 25 (Figure 15.10). Catalyst 24 (20 mol%) could be recycled for five cycles in the reaction between acetone and 4-nitrobenzaldehyde virtually without erosion of chemical or stereochemical efficiency (first cycle: 100% yield, 71% ee; fifth cycle: 96% yield, 71% ee).41 In contrast, catalyst 25 (1 mol%) proved less robust when repeatedly employed in the same reaction. Indeed, while the first cycle afforded the product in 93% yield and 80% ee, the yield of the fifth cycle was reduced to 60% even if the ee remained high (80%). Further recycling led to a dramatic decrease of both yield and ee.42 Ionic Liquids Properly functionalized ionic liquids were recently identified as suitable supports for organic catalysts, when it was realized that the solubility profile of ionic liquids could provide a simple entry to catalyst recovery and recycling via product removal by extraction with a solvent in which the ionic liquid material was not soluble. In a short period of time four different proline catalysts covalently connected to ionic liquids have been described.22 At high catalyst loading (30 mol%) imidazolium-bound proline 26 (Figure 15.11), obtained by Miao and Chan in 2006,43 performed well in the addition of acetone to aldehydes affording the products in good yield (up to 92%) and ee (up to 87%) using acetone as the solvent. The catalyst was recovered by evaporation of the volatile materials present in the reaction mixture, followed by addition of dichloromethane and centrifugation. The insoluble catalyst, dried under vacuum, was recycled four times to promote the aldol addition of acetone to 4nitrobenzaldehyde in virtually constant yield and ee. One year later, the strictly related catalyst 27 was shown to perform equally well at a lower catalyst loading (10 mol%) but in the expensive solvent 1-n-butyl-3-methylimidazolium tetrafluoroborate (bmim)(BF4).44 A similar solvent (bmim)[(CF3SO2)2N] was also employed to study the behaviour of catalyst 28 (10 mol%) in the reaction of acetone with


447

X2 N

N

O

COOH NH

Y 26; X2 = O, Y = BF4 27; X2 = H, Y = Br O MeBu2N

O

COOH NH

(CF3SO2)2N 28 O *

N

N

O

COOH

4

11

NH PF6 29

Figure 15.11 Ionic liquid-supported proline-derived catalysts

aromatic aldehydes.45 It is interesting to note that while standard good yields and stereoselectivity were observed in these reactions, the recycling of catalyst 28 was less successful than that of the related species 26 and 27, a marked decrease in chemical yield (from 75 to 30%) being observed after as few as three reaction cycles. Given the similarity between the structures of these catalysts, this difference in recyclability is puzzling and difficult to rationalize. Finally, the hydrophobic catalyst 29 was prepared 46 with the idea that it should perform particularly well ‘on water’.47 Indeed, a high loading of this catalyst (30 mol%) promoted the reaction of cyclohexanone with 4-methoxybenzaldehye in the presence of water (10 h, room temperature) in 86% conversion and excellent stereocontrol (anti: syn ¼ 97:3, ee of anti isomer >99%). These values remained constant up to the fifth reaction cycle (the supported catalyst having been recovered by extraction with diethylether and evaporation). Dendrimers Dendrimeric structures offer a valuable option for immobilization, recovery and recycling of a catalyst, especially because the dendritic skeleton can offer at the same time the handle for recovery and the proximity of the catalytic sites which can be used to develop new catalytic systems capable of displaying cooperation effects. A couple of examples of proline supported on dendrimers have been reported in the literature. In 2005, Kokotos and

448


Bellis synthesized a few poly(propyleneimine)-based dendrimers to employ them as catalysts for the aldol reaction between acetone and electron-poor aromatic aldehydes.48 Among these, the catalyst derived by treatment of the second generation dendrimer 30 with triethylamine (Figure 15.12) performed better than its first, third, fourth or fifth generation counterparts. In particular, at the relatively low catalyst loading of 6.5 mol%, the products were obtained in short reaction times and good yield and ee. The reaction had to be carried out in the presence of triethylamine to release the secondary amine function of proline, the precatalyst 30 being totally inactive in its fully protonated form. Unfortunately, no recycling results were reported. More recently, Kehat and Portnoy prepared a series of first, second and third generation dendrons supported on the insoluble, polystyrene-type Wang resin and covalently attached NHR

RHN

H RHN

N

N

N

H RHN

NHR H

H

H

N

N

N

NHR

H 6Cl

NHR

RHN

30, R =

O O

COOH N

O

H

H

Cl

X O X X

O O

X

N N 31;

= Wang resin, X =

N

O

COOH NH

Figure 15.12 Proline-derived catalysts anchored on dendrimeric supports


449

to them different number of proline residues through a triazole-containing spacer.49 Using the standard addition of acetone to aromatic aldehydes as model reaction, it was shown that the second generation species 31 (Figure 15.12) gave better yield and ee values than both the other dendritic catalysts and the ad hoc prepared nondendritic catalyst similarly supported on Wang resin. However, while the latter was fully recyclable (as expected on the basis of the work of Pericas and Gruttadauria, see section Polymers), catalyst 31 performed very poorly when recycled, especially in term of chemical efficiency. Cyclodextrin The hydrophobic cavity of b-cyclodextrin (CD) was used for preparing recyclable proline-based catalyst by immobilization through noncovalent interactions between the cavity itself and an hydrophobic residue ad hoc inserted on the proline nucleus. In a first report,50 inclusion of cis-4-phenoxyproline into the CD cavity was simply achieved by refluxing an ethanol–water solution of the reagents. Evaporation of the solvent afforded the included catalyst 32 (Figure 15.13) as shown by NMR and UV spectroscopy. When employed at a 10 mol% loading this catalyst promoted the addition of acetone (used as reaction solvent) to electron-poor aromatic aldehydes in up to 90% yield and 83% ee. Remarkably, the insoluble noncovalently supported catalyst could be recovered by filtration and recycled four times affording the products in undiminished ee and slightly decreasing yield. A similar strategy was followed to include cis-4-(1-adamantanoylamido) proline into CD to afford adduct 33.51 This was used at the 10 mol% loading to promote the reaction of cyclohexanone with aromatic and heteroaromatic aldehydes in water. The reaction worked well especially with electron-poor aldehydes, to afford the product in up to 97% yield, complete anti stereoselectivity and >99% ee. Three recycling of the catalyst occurred without any erosion of the efficiency of the process. Later it was demonstrated that the inclusion step could be performed in the reaction mixture provided that a very hydrophobic proline derivative was employed and water used as the reaction solvent.52 Thus, reaction of equimolar amounts of cyclohexanone and aromatic aldehydes carried out at room temperature in water in the presence of proline 34 HOOC NH

COOH

COOH

NH O

NH O

NH

O

But 32

33

34

Figure 15.13 Proline-derived catalysts noncovalently anchored on b-cyclodextrin

450


(2 mol%) and sulfated CD (10 mol%) allowed the aldol products to be isolated in excellent diastereo- (anti:syn ratios up to 99:1) and enantioselectivity (ee 96–>99%). As usual, electron-poor aldehydes gave better yields (up to 100%) than electron-rich ones. The fact that the ketone can be used in a stoichiometric amount in these reaction is particularly relevant, as in these type of processes the ketone donor is always used in large excess. Unfortunately, no recycling experiments of this in situ generated catalyst were reported. 15.3.2 Inorganic Supports Silica Chandrasekaran and coworkers were the first to report the synthesis of a proline catalyst supported on mesoporous silica MCM41 and its use in the aldol reaction of acetone with two aromatic aldehydes.53 Even if the results were not particularly encouraging (moderate yields and ee, poor recyclability of the catalyst), the same approach was investigated a couple of years later by Fernandez-Mayoralas and his group.54 Catalyst 35 (Figure 15.14) was obtained in several steps from the expensive N-Cbz protected trans 4-hydroxyproline by introduction on this amino acid of the trimethoxysilyl-substituted chain necessary for grafting the catalytic active species on MCM41. The inorganic–organic hybrid catalyst was

O O O

Si

H N

(CH2)3

H N COOH

O

N H

35 O

O

OH

35 (20 mol%)

(Eq. 10)

+ RCHO

R

OH

OH R = iPr, cC6H11, Ph

O O O

Si

H N

(CH2)3 36

O O

COOH N H

+ 4NO2C6H4CHO O

O

OH

O

O

(Eq. 11)

35 or 36 (30 mol%) O2N

O

O

= MCM41

Figure 15.14 Proline-derived catalysts supported on silica


451

then used at high loading (20 mol%) to promote the reaction between hydroxyacetone and three different aldehydes reported in equation 10. The reaction required harsh condition to proceed satisfactorily (90 C, in DMSO), and the supported catalyst worked as efficiently as nonsupported proline only in the reaction with isobutyraldehyde. The use of microwave drastically reduced the reaction times. Yield erosion upon catalyst recycling (two cycles) was also observed (the effect of recycling on ee was not described). A couple of years later, the same group studied the role exerted by the solvent when the same promoter 35 and the related catalyst 36 (Figure 15.14) were employed in the reaction of equation 11.55 It was discovered that the reaction proceeded nicely in hydrophilic dipolar aprotic solvents (such as formamide or DMF) and satisfactorily in wet apolar solvent (toluene containing up to 5 equivalent of water). As for the catalysts, 35 performed better than 36, but a thorough study of the influence of the different configuration at the C-4 stereocenter of the proline nucleus or of the chemical nature of the spacer separating the support from the catalyst was not carried out. A single recycling experiment showed a very minor decrease in chemical yield and ee. Layered Double Hydroxides Proline has also been supported on other inorganic supports, such as magnetite (Fe3O4) nanoparticles and layered double hydroxides (LDH). While the magnetite-supported proline was employed only in the N-arylation of heterocyclic aromatic compounds,56 proline intercalated in LDH was used in the aldol reaction of acetone with benzaldehyde and therefore must be briefly discussed here.57 LDH are synthetic anionic layered clays and several methods are available to introduce anions between the layers. In the case of LDH-supported proline, sodium prolinate was employed at different concentration to afford catalysts with proline contents ranging from 0.9 to 1.8 mmol g1. After the materials were tested for structural and configurational stability, they were used as heterogeneous catalysts (35 mol%) in the above mentioned reaction carried out in refluxing acetone leading to the product in excellent yield (89%) and ee (94%). While the dependence of the outcome of the reaction on the yield of the intercalation process and the exposure of the catalyst to high temperature was studied, no mention of the use of this catalyst in other reactions or of recycling experiments were provided. In concluding this section on the use of supported prolines, while any attempt to the identification of the ‘best’ support appears scientifically meaningless, a few general trends can be pointed out: (i) Independently of the nature of the support, most of the immobilized catalysts, when used for the first time, secured yield and stereoselectivity comparable to those obtained with nonsupported proline. However, efficient catalyst recovery and recycling was observed in way less instances. When recycled, insoluble catalysts performed generally better than soluble ones; this seems to hold true independently of the nature of the support. (ii) The covalent attachment of the catalyst to the support does not seem a mandatory condition for catalyst stability and hence recyclability, with obvious benefits in term of reduced synthetic manipulation for attachment to the support. (iii) An active role of the support has not been investigated systematically, and a wide margin of improvement under this aspect seems possible.

452


15.4 Comparison of Different Supports Employed for the Immobilization of Bis(oxazolines) Although not as central to organometallic catalysis as proline to organocatalysis, chiral bis(oxazoline) ligands (Box) have played a fundamental role in the development and the establishment of stereoselective organometallic catalysis as a reliable and versatile approach to the synthesis of enantiomerically enriched compounds.58 Accordingly, it is no wonder that, after Mayoral’s pioneering work in 1997,59 these ligands have been the subject of several efforts devoted to their immobilization on a variety of supports.60 In the following sections different examples of supported Box and their metal complexes will be examined. As in the case of supported prolines, this survey will not cover every example of application of supported Box ligands, but rather it will be focused on the identification of the relative merits of the various catalyst–support combinations with particular interest to catalyst recovery and recycling. To obtain a more meaningful comparison, only the use of differently immobilized Box/copper complexes in the benchmark cyclopropanation reaction of styrene with ethyl diazoacetate will be considered. 15.4.1 Noncovalent Immobilization As mentioned above, in 1997 Mayoral and coworkers reported the noncovalent immobilization of Box/Cu(II) 37a–c (Figure 15.15) on different clays, including laponite, montmorillonite and bentonite, by cation exchange.59 When employed in the reaction of Me O

O N

R

Me

Me

N

N

X = Cl, OTf

R

CuX2

N

O

a, R = Bn b, R = Ph c, R = But

But

37a-c

O N

CuOTf2

But

38 + Ph

Ph

COOEt

+

COOEt

cat.

Ph trans

COOEt (Eq. 12)

N2

+ Ph

Ph

COOEt

COOEt

cis F CF2CF2 n

CF2

C O

CF3 C F2

CF

CF2CF2SO3Na m

Na-Nafion

Figure 15.15 Box/Cu complexes noncovalently supported on laponite and Nafion


453

equation 12 (Figure 15.15) the laponite-supported catalysts gave the best results. These, however, fell short of those obtained with the nonsupported catalyst under homogeneous conditions. This was true in terms of chemical yield, trans:cis stereoselectivity, and enantiomeric excess of the produced isomeric cyclopropane derivatives. Both yields and ee slightly decreased when the catalysts were recycled in a subsequent reaction. Further recycling was not described. A couple of years later the same group reported the immobilization of catalysts 37 on an anionic solid organic support.61 Exchange of the triflate anions of 37b/Cu(OTf)2 with Na–Nafion (Figure 15.15) afforded a supported catalyst which had a copper content similar to that of its laponite-supported counterpart, and performed similarly to the homogeneous catalyst in the cyclopropanation reaction. Remarkably, the catalyst maintained its stereochemical efficiency upon recycling, showing only a moderate decrease in chemical activity. Unfortunately, immobilization of complex 37c, a catalyst more stereoselective than 37b under homogeneous conditions, afforded a supported catalyst very prone to leaching of both the metal and, more importantly, the ligand. While covalent immobilization of the ligand was envisaged as the obvious solution to this problem (see Section 4.2 below), the use of aza-bis(oxazolines), which are known to afford more stable and catalytically more active complexes than Box, was also studied.62 Indeed, immobilization of the aza-analogue of 37c, (38, Figure 15.15) on laponite afforded a supported catalyst that allowed to perform the cyclopropanation reaction of styrene with ethyl diazoacetate in yield (46%), trans diastereoselectivity, and ee (83% for the trans isomer; 76% for the cis isomer) only slightly inferior to those observed with the nonsupported catalyst. A single recycling was found to occur in virtually identical yield and stereocontrol. 15.4.2 Covalent Immobilization Organic Supports Several groups have reported the covalent immobilization of Box ligands on polystyrene.60 This was accomplished essentially by two alternative techniques: (i) grafting of a preformed, suitably functionalized Box on a commercial polymer; or (ii) radical-initiated polymerization involving ad hoc prepared Box monomers, and, if necessary, styrene and cross-linking agents. The desired insoluble catalysts were then obtained by addition of an excess of copper salts (the unreacted material being removed with methanol). By this procedure a Cu/Box ratio < 1 was always obtained, for obvious reasons of accessibility of the ligand on the surface of the polymer beads. Typical catalyst loadings were low, ranging from 0.1 to 0.3 mmol(Cu)/g polymer. Among the catalysts prepared by Mayoral’s group,63 homopolymeric compound 39 (Figure 15.16), employed at different loadings, performed better, affording the product of the standard cyclopropanation reaction (equation 12, Figure 15.15) in 36% yield, 63:37 cis:trans ratio, and 78 and 72% ee for the major and minor diastereoisomer, respectively. As far as stereoselectivity is concerned, these results were very similar to those obtained with the structurally related nonsupported catalyst 40 (Figure 15.16), which however gave a slightly better yield (46%). Mayoral also reported the first immobilization of 2,6-bisoxazoline-substituted pyridines (Py-Box) and their use in the Ru (II) promoted cyclopropanation reaction.64 Also in this

454

Recoverable and Recyclable Catalysts ** *

n*

n

O

O N

Ph

O

O

N

N

N

CuOTf2

Ph

Ph

CuOTf2 40

39

*

*

7

2

91

O

O

N N

N RuCl2

Pri

Ph

*

*

2

Pri 41

*

* 4

54

42

*

O

* 54

Me O

O N But

N

CuOTf2

But

42

Figure 15.16 Box/metal complexes supported on insoluble polystyrene matrices


455

case both grafting and co-polymerization approach were followed and, as in the previous case, the catalysts obtained by polymerization performed generally better. In particular, the use of 3 mol% of the Ru(II) catalyst 41 (Figure 15.16) in dichloromethane at room temperature allowed the cyclopropane products to be obtained in 60% yield, with good trans selectivity (trans:cis ratio 88:12) and high ee (90% for anti isomer). Only after recycling the catalyst three times did the chemical yield decrease, whereas the stereoselectivity remained high.65 Pini, Salvadori, and coworkers also investigated extensively the use of Box ligands immobilized by co-polymerization on insoluble polystyrene in enantioselective catalysis.60 In their paper dedicated to the use of these systems in the cyclopropanation reaction of alkenes with ethyl diazoacetate,66 highly cross-linked catalyst 42 (Figure 15.16) was synthesized with the hope that the macroporous nature of the polymer and the reduced steric hindrance at the carbon atom bridging the two oxazoline units would help to mimic at best the structure of the nonsupported catalyst. Using only 3.6 mol% of 42 (dichloromethane, 0 C, 3 h) a 67:33 mixture of trans and cis cyclopropanes was formed in 60% yield. The products showed very high ee of 93 and 90%, for the trans and cis adducts, respectively. These results were remarkably similar to those observed with the nonsupported catalyst featuring two methyls at the bridging carbon atom: 61% yield, 71:29 diastereoisomer ratio, 94 and 92% ee. To assess the possibility of recycling the supported catalyst, the polymeric material was filtered, rinsed with dichloromethane, dried and directly re-used in other runs. Following this procedure five recyclings were shown to occur in virtually identical yield and stereoselectivity. Soluble catalysts 4367 and 44,68 immobilized on MeOPEG5000 (Figure 15.17) were also reported and employed in the cyclopropanation of alkenes and in other reactions. In the benchmark reaction of equation 12 (Figure 15.15) the aza-Box-derived catalyst 43 performed slightly better than the traditional Box 44 in terms of yield (69 vs 63%), whereas trans selectivity (71:29 vs 70:30), and ee (91 vs 91%) were identical. However, catalyst 43, taking advantage of the high reactivity of the aza-Box ligand, could be employed at a much lower loading than 44 (1 vs 10 mol%). After precipitation with diethylether, 43 was recycled without any addition of metal salt or other activators for nine cycles occurring with constant yields and stereoselectivity. Inorganic Supports Several Box ligands have been immobilized on siliceous materials 20,69 or other inorganic supports,20,70 and the catalysts obtained by their metallation have been employed as recyclable promoters in a number of different reactions (mostly Diels–Alder cycloadditions). Considering only the application of these catalysts to the cyclopropanation reaction (equation 12, Figure 15.15), the results obtained by Ying’s71,72 and Mayoral’73 groups seem particularly relevant. In particular, Ying grafted commercially available indanol-derived Box on siliceous mesocellular foam (a material featuring extra large pores), and studied the effect of capping the free silanol groups present on the surface of the support with trialkylsilanes. It was discovered that the capping procedure enhanced the activity and the enantioselectivity of the resulting catalyst 45 (Figure 15.18) with respect to those of the uncapped material, affording a catalytic system comparable in efficiency with its nonsupported counterpart.71

456

Recoverable and Recyclable Catalysts MeOPEG5000O

N

O

O

N But

N

CuOTf2

But

43 MeOPEG5000O

O

Me O

O N

N

But

But CuOTf

44

Figure 15.17 PEG-supported Box/and aza-Box/Cu complexes

Working with a low catalyst loading (1 mol%) at room temperature, chemical yields as high as 91%, and ee of 82 and 84% for the trans and cis isomers, respectively, were obtained. The catalyst could be recycled two times with very minor erosion in yield and stereoselectivity. In a more recent work,72 the favourable effect of capping on similar aza-Box-derived catalysts was confirmed. In addition, it was shown that catalysts connected to the mesocellular foam with a rigid spacer (such as 46a) performed better than the one featuring a more flexible spacer (such as 46b), likely because in the former case minor support/ catalyst interactions could be present (these interactions could be particularly noxious since the aza-Box ligand features the coordinating and basic bridging nitrogen). In addition, the heterogenized catalyst was successfully applied to a circulating flow-type packed bed reactor with excellent productivity and enantioselectivity, which remained virtually constant for as many as 20 reaction cycles. Almost simultaneously to the first report by Ying, Mayoral73 and coworkers reported a mesoporous silica-supported catalyst obtained by the co-condensation rather than by the grafting method. The co-condensation procedure can be considered a sort of ‘inorganic polymerization’ and can open access to interesting inorganic–organic hybrid materials featuring different catalytic sites (Figure 15.4). The catalyst obtained by this method however (47, Figure 15.18) was less active and stereoselective than its counterparts obtained through grafting by Ying. Under the best conditions, moderate yield (33%) and ee (53%) were obtained. These values remained constant upon a single recycling experiment.


Siliceous mesocellular foam OSiR3 O O

MeO

O O

Si

Si

O

OSiR3

OMe

O N

N

CuOTf2 45 Siliceous mesocellular foam OTMS

O

OTMS

R

O Si OMe

N

O

O

N But

OTMS OTMS

N

CuOTf2

But

46a, R =

46b, R =

(CH2)3

mesoporous silica O O O Si

O O O Si

NH

NH

O

O O

O

O

O N

Ph

N

CuOTf2

Ph

47

Figure 15.18 Box/Cu complexes supported on different silaceous materials

457

458


As in the case of supported prolines, also in concluding this section on the use of immobilized Box-containing catalysts any attempt to the identification of the ‘best’ support appears meaningless. However, it can be said that: (i) Several supported Box ligands appeared well suited to recovery and recycling, to the point that, in a few cases, a high number of reaction cycles could be performed with a recovered catalyst. It is remarkable that efficient recycling was possible with both soluble 67 and insoluble 72 catalytic systems. (ii) In general, Box-based catalysts appeared very sensitive to the covalent or noncovalent nature of the connection to the support, the former appearing generally safer than the latter. (iii) A relatively limited number of supports have been tested in combination with Boxderived catalyst, likely because the presence of the metal forced a very conservative choice among possible supports.

15.5 Conclusions It was the aim of this chapter to illustrate the prerequisite conditions, the justifying reasons, and the practical aspects of the immobilization of a catalyst on a support, providing at the same time a few guidelines to compare the merits of different supports. In concluding this discussion, it can be said that the research in this field has shown that the development of robust, efficient, and recyclable supported catalysts is indeed possible in several cases. In addition, recent results have also demonstrated that the immobilization procedure should not be considered an unavoidable obstacle en route to the development of a recyclable catalyst, but rather should be regarded as a fruitful opportunity to improve, modify, and discover new properties for the supported catalytic system.74 It can be easily anticipated that the continuous discovery of new materials will definitely contribute to further developments in this field, as will the relentless search for new catalysts. Obviously, the combination of the efforts of material and polymer chemists with those of the researchers active in the discovery of new catalysts will provide the winning strategy for this endeavour.

References 1. K. A. Ahrendt, C. J. Borths and D. W. C. MacMillan, J. Am. Chem. Soc., 122, 4243–4244 (2000). 2. A. Berkessel and H. Groger, Asymmetric Organic Catalysis, Wiley-VCH, Weinheim, 2005. 3. M. Benaglia, G. Celentano, M. Cinquini, A. Puglisi and F. Cozzi, Adv. Synth. Catal., 344, 149–152 (2002). 4. S. A. Selk€al€a, J. Tois, P. M. Pihko and A. M. P. Koskinen, Adv. Synth. Catal., 344, 941–945 (2002). 5. M. Benaglia, G. Celentano, M. Cinquini, A. Puglisi and F. Cozzi, Eur. J. Org. Chem., 567–573 (2004). 6. W. S. Sen, J. J. M. Wiemer and D. W. C. MacMillan, J. Am. Chem. Soc., 122, 9874–9875 (2000). 7. For a review on the general problem of reactant accessibility within a polymeric support, see: P. Hodge, Chem. Soc. Rev., 26, 417–424 (1997). 8. To circumvent this problem the insertion of a spacer separating the catalyst from the support has been proposed. For pioneering results see: H. Molinari, F. Montanari, S. Quici and P. Tundo, J. Am. Chem. Soc., 101, 3920–3927 (1979); For a discussion on the influence of the spacer on


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.

459

catalyst reactivity see: P. L. Anelli, B. Czech, F. Montanari and S. Quici, J. Am. Chem. Soc., 106, 861–869 (1984), and references cited therein. R. Annunziata, M. Benaglia, M. Cinquini, F. Cozzi and G. Tocco, Org. Lett., 2, 1737–1739 (2000). See also: K. Shibatomi, T. Nakahashi and Y. Uozumi, Synlett, 1643–1645 (2000). J. D. Bass, A. Solovyov, A. J. Pascal and A. Katz, J. Am. Chem. Soc., 128, 3737–3747 (2006). For a review describing the enhancement of heterogeneous catalysis through cooperative hybrid organic–inorganic interfaces see: J. M. Notenstein and A. Katz, Chem. Eur. J., 12, 3954–3965 (2006). For a review describing the exploitation of nanospace confinement for asymmetric catalysis see: J. M. Thomas and R. Raja, Acc. Chem. Res., 41, 708–720 (2008). B. F. G. Johnson, S. A. Raynor, D. S. Shepard, T. Mashmeyer, J. M. Thomas, G. Sankar, S. Bromley, R. Oldroyd, L. Gladden and M. D. Mantle, Chem. Commun., 1167–1168 (1999). For a recent review see: E. L. Margelefsky, R. K. Zeidan and M. E. Davis, Chem. Soc. Rev., 37, 1118–1126 (2008). This review collects a few examples of cooperative supported bi-functional catalysts reported by the Davis’s group. S. Huh, H.-T. Chen, J. W. Wiench, M. Pruski and V. S.-Y. Lin, Angew. Chem. Int. Ed., 44, 1826–1830 (2005). D. J. Gravert and K. D. Janda, Chem. Rev. 97, 489–509 (1997). D. E. Bergbreiter, Chem. Rev. 102, 3345–3383 (2002). D. E. Bergbreiter and C. Li, Org. Lett., 5, 2445–2447 (2003). For a review on continuous flow methods see: G. Jas and A. Kirschning, Chem. Eur. J., 9, 5708–5723 (2003). A. M. Hafez, A. E. Taggi, T. Dudding and T. Lectka, J. Am. Chem. Soc., 123, 10853–10859 (2001). M. Benaglia, M. Cinquini, F. Cozzi, A. Puglisi and G. Celentano, Adv. Synth. Catal., 344, 533–542 (2002). For an excellent review on supported proline and proline derivatives as catalyst see: M. Gruttadauria, F. Giacalone and R. Noto, Chem. Soc. Rev., 37, 1666–1688 (2008). G. Sz€oll€osi, G. London, L. Balaspiri, C. Somlai and M. Bartok, Chirality, 15, S90–S96 (2003). M. Benaglia, M. Cinquini, F. Cozzi and G. Tocco, Tetrahedron Lett., 43, 3391–3393 (2002). For pioneering work demonstrating that an increased loading of active sites does not necessarily result into an increased catalytic activity see: P. L. Anelli, F. Montanari and S. Quici, J. Org. Chem., 51, 4910–4914 (1986). For an excellent review see: C. Gennari and U. Piarulli, Chem. Rev., 103, 3071–3100 (2003). M. S. Sigman and E. N. Jacobsen, J. Am. Chem. Soc., 120, 4901–4902 (1998); M. S. Sigman, P. Vachal and E. N. Jacobsen, Angew. Chem. Int. Ed., 39, 1279–1281 (2000); P. Vachal and E. N. Jacobsen, J. Am. Chem. Soc., 124, 10012–10014 (2002). T. Danelli, R. Annunziata, M. Benaglia, M. Cinquini, F. Cozzi and G. Tocco, Tetrahedron: Asymmetry, 14, 461–467 (2003). For the synthesis of the nonsupported catalyst, see: G. Priem, B. Pelotier, S. J. F. Macdonald, M. S. Anson and I. B. Campbell, J. Org. Chem., 68, 3844–3848 (2003); For that of the supported systems, see: G. Priem, B. Pelotier, S. J. F. Macdonald, M. S. Anson and I. B. Campbell, Synlett, 679–683 (2003). T. Balakrishnan and E. Murugan, J. Polym. Sci., Part A: Polym. Chem., 41, 347–364 (2003). Reviews: B. List, Tetrahedron, 58, 5573–5590 (2002); E. R. Jarvo and S. J. Miller, Tetrahedron, 58, 2481–2495 (2002); W. Notz, F. Tanaka and C. F. Barbas, Acc. Chem. Res., 37, 580–591 (2004). K. Kondo, T. Yamano and K. Takemoto, Makromol. Chem., 186, 1781–1785 (1985). M. Benaglia, G. Celentano and F. Cozzi, Adv. Synth. Catal., 343, 171–173 (2001). M. Benaglia, M. Cinquini, F. Cozzi, A. Puglisi and G. Celentano, J. Mol. Catal. A: Chem., 204–205, 157–163 (2003). L. Gu, Y. Wu, Y. Zhang and G. Zhao, J. Mol. Catal. A: Chem., 263, 187–194 (2007). M. Benaglia, A. Puglisi and F. Cozzi, Chem. Rev., 103, 3401–3430 (2003); F. Cozzi, Adv. Synth. Catal., 348, 1367–1390 (2006). D. Font, C. Jimeno and M. A. Pericas, Org. Lett., 8, 4653–4655 (2006). D. Font, S. Sayalero, A. Bastero, C. Jimeno and M. A. Pericas, Org. Lett., 10, 337–340 (2008).

460


39. F. Giacalone, M. Gruttadauria, A. Mossuto Marculescu and R. Noto, Tetrahedron Lett., 48, 255–258 (2007); F. Giacalone, M. Gruttadauria, A. Mossuto Marculescu, S. Riela and R. Noto, Eur. J. Org. Chem., 4688–4695 (2007). 40. Y. X. Liu, Y. N. Sun, H. H. Tan, W. Liu and J. C. Tao, Tetrahedron: Asymmetry, 18, 2649–2656 (2007); Y. X. Liu, Y. N. Sun, H. H. Tan and J. C. Tao, Catal. Lett., 120, 281–287 (2008). 41. K. Akagawa, S. Sakamoto and K. Kudo, Tetrahedron Lett., 46, 8185–8187 (2005). 42. J. D. Revell, D. Gantenbein, P. Krattiger and H. Wennemers, Biopolymers, 84, 105–113 (2006). 43. W. Miao and T. H. Chan, Adv. Synth. Catal., 348, 1711–1718 (2006). 44. L. Zhou and L. Wang, Chem. Lett., 36, 628–629 (2007). 45. M. Lombardo, F. Pasi, S. Easwar and C. Trombini, Adv. Synth. Catal., 349, 2061–2065 (2007). 46. D. E. Siyutkin, A. S. Kucherenko, M. I. Struchkova and S. G. Zlotin, Tetrahedron Lett., 49, 1212–1216 (2008). 47. N. Mase, Y. Nakai, N. Ohara, H. Yoda, K. Takabe, F. Tanaka and C. F. Barbas, J. Am. Chem. Soc., 128, 734–735 (2006); Y. Hayashi, T. Sumiya, J. Takahashi, H. Gotoh, T. Urushima and M. Shoji, Angew. Chem., Int. Ed., 45, 958–961 (2006). 48. E. Bellis and G. Kokotos, J. Mol. Catal. A: Chem., 241, 166–174 (2005). 49. T. Kehat and M. Portnoy, Chem. Commun., 2823–2825 (2007). 50. Z. Shen, J. Ma, Y. Liu, C. Jiao, M. Li and Y. Zhang, Chirality, 17, 556–558 (2005). 51. K. Liu, D. Haussinger and W.-D. Woggon, Synlett, 2298–2300 (2007). 52. J. Huang, X. Zhang and D. W. Armstrong, Angew. Chem., Int. Ed., 46, 9073–9077 (2007). 53. D. Dhar, I. Eadham and S. Chandrasekaran, Proc. Ind. Acad. Sci. (Chem. Sci.), 365–372 (2003). 54. F. Calderon, R. Fernandez, F. Sanchez and A. Fernandez-Mayoralas, Adv. Synth. Catal., 347, 1395–1403 (2005). 55. E. G. Doyaguez, F. Calderon, F. Sanchez and A. Fernandez-Mayoralas, J. Org. Chem., 72, 9353–9356 (2007). 56. G. Chouhan, D. Wang and H. Alper, Chem. Commun., 4809–4811 (2007). 57. Z. An, W. Zhang, H. Shi and J. He, J. Catal., 241, 319–327 (2006). 58. J. S. Johnson and D. A. Evans, Acc. Chem. Res., 33, 325–335 (2000); F. Fache, E. Schulz, M. L. Tommasino and M. Lemaire, Chem. Rev., 100, 2159–2232 (2000). 59. J. M. Fraile, J. I. Garcia, J. A. Mayoral and T. Tarnai, Tetrahedron: Asymmetry, 8, 2089–2092 (1997); see also: J. M. Fraile, J. I. Garcia, J. A. Mayoral and T. Tarnai, Tetrahedron: Asymmetry, 9, 3997–4008 (1998); P. J. Alonso, J. M. Fraile, J. Garcia, J. I. Garcia, J. I. Martinez, J. A. Mayoral and M. C. Sanchez, Langmuir, 16, 5607–5612 (2000). 60. D. Rechavi and M. Lemaire, Chem. Rev., 102, 3467–3494 (2002). 61. J. M. Fraile, J. I. Garcia, J. A. Mayoral, T. Tarnai and M. A. Harmer, J. Catal., 186, 214–221 (1999). 62. J. M. Fraile, J. I. Garcia, C. I. Herrerias, J. A. Mayoral, O. Reiser, A. Socuellamos and H. Werner, Chem. Eur. J., 10, 2997–3005 (2004). 63. M. I. Burguete, J. M. Fraile, J. I. Garcia, E. Garcia-Verdugo, C. I. Herrerias, S. V. Luis and J. A. Mayoral, J. Org. Chem., 66, 8893–8901 (2001). 64. A. Cornejo, J. M. Fraile, J. I. Garcia, M. J. Gil, S. V. Luis, V. Martinez-Merino and J. A. Mayoral, J. Org. Chem., 70, 5536–5544 (2005). 65. For recent applications of other metal complexes of Py-Box ligands related to that present in 41, see: M. Tilliet, S. Lundgren, C. Moberg and V. Levacher, Adv. Syn. Catal., 349, 2079–2084 (2007). 66. A. Mandoli, S. Orlandi, D. Pini and P. Salvadori, Chem. Commun., 2466–2467 (2003). 67. M. Glos and O. Reiser, Org. Lett., 2, 2045–2048 (2000). 68. R. Annunziata, M. Benaglia, M. Cinquini, F. Cozzi and M. Pitillo, J. Org. Chem., 66, 3160–3166 (2001). 69. For examples not included in ref. 60, see: D. Rechavi and M. Lemaire, Org. Lett., 3, 2493–2496 (2001); J. K. Park, S. W. Kim, T. Hyeon and B. M. Kim, Tetrahedron: Asymmetry, 12, 2931–2935 (2001); A. Lee, W. Kim, J. Lee, T. Hyeon and B. M. Kim, Tetrahedron: Asymmetry, 15, 2595–2598 (2004); P. O’Leary, N. P. Krosveld, K. P. De Jong, G. van Koten and R. J. M. K. Gebbink, Tetrahedron Lett., 45, 3177–3180 (2004).


461

70. For a zeolite Y-supported Box, see: N. A. Caplan, F. E. Hancock, P. C. Bulman Page and G. J. Hutchings, Angew. Chem., Int. Ed., 43, 1685–1688 (2004). 71. T. M. Lancaster, S. S. Lee and J. Y. Ying, Chem. Commun., 3577–3579 (2005). 72. J. Lim, S. N. Riduan, S. S. Lee and J. Y. Ying, Adv. Synth. Catal., 350, 1295–1308 (2008). 73. J. M. Fraile, J. I. Garcia, C. I. Herrerias and J. A. Mayoral, Chem. Commun., 4669–4671 (2005). 74. For a very recent review proposing a new approach to the development of supports capable of improving the efficiency of a supported catalyst see: N. Madhavan, C. W. Jones and M. Weck, Acc. Chem. Res., 41, 1153–1165 (2008).

Index Note: page number in italic refer to figures and schemes. active sites on metal and alloys 16 principle of maximum separation 440–1 aldol condensation bifunctional catalyst on silica 41–2, 43 in ionic liquid 280–1 microencapsulated catalyst 348, 365 microreactor 415–16 organocatalysed 63–4, 287–90, 324, 326–7, 331–2 insoluble resin-supported catalyst 63–4 allylation amination on silica 33, 34 fluorous chiral catalyst 185, 189 insoluble resin-supported catalyst 65–6 in ionic liquids 282–3 microencapsulated catalyst 345–6 Amberlyst-21 support 67, 391 amidocarbonylation 359–60 amination, allylic 33, 34 amino acid derived catalysts 319–34 see also proline amino alcohol ligands 54 aminocarbonylation 422 aminopropyl functionality 36, 40–1 aminoxylation 63 amphiphilic PS–PEG resins 50 annulation reactions 66–9 see also cyclization aqueous phase catalysis, supported 220–7 see also biphasic catalysis arylaldehydes, alkylation of 389–90, 391, 393 assay of catalyst see catalyst recovery atom transfer radical polymerization 138–42

Baeyer–Villiger reaction 348, 415 Barton reaction 420 batch process 380–1 Baylis–Hillman reaction insoluble resin-supported catalyst 64, 143–4, 293–4 in ionic liquids 293–4 silica supported catalyst 34–5 theromomorphic 143–4 bentonite 452 benzylation 308, 312, 313 1,10 -Bi-2-naphthol see BINOL bifunctionalized silica catalyst 35–8, 432–3 BINAP-containing polymers fluorous BINAP analogs 96, 195–6 linear polymeric chiral catalysts 103, 105–8 mixed BINAP–BINOL 108–10 ruthenium-sulfonated catalyst 226 self-supporting 156–8, 172 binaphthyl complexes 81, 92, 94, 223 BINOL-containing polymers 57 fluorous chiral catalysts 192–3 linear polymeric chiral catalysts 102–3, 104 mixed BINOL–BINAP 108–10 self-supporting 163–72 biphasic catalysis 199–229, 413–17 additives/mass transfer promotors co-solvents 204 inverse phase transfer catalysts calixarenes 210–11 cyclodextrins 207–10 styrene copolymers 211–12 surfactants 204–7


464

Index

biphasic catalysis (Continued ) commercial hydroformylation of alkenes 202–3 fluorous solvents/catalysts 179–81, 215 homogeneous reaction with biphasic separation 214–20 pH dependant solubility 218–20 thermoregulated phase transfer 216–17 water–oil microemulsions 215–16 new reactor design 227–8 non-aqueous systems 413–17 see also ionic liquids; supercritical carbon dioxide overviews 200–2, 228–9, 413–15 supported aqueous phase catalysis 220–7 palladium TPPTS catalyst 226–7 rhodium TPPTS catalyst 221–3, 224–6 rhodium/polyacrylic acid catalyst 226 ruthenium-sulfonated BINAP catalyst 223–4 sulfoxantphos catalyst 226 surface-active ligands 212–14 BIPHEP ligands 103–4 2,20 -bis(diphenylphospino)-1,10 -binaphthyl see BINAP bis(oxazolines) see BOX ligands bisphosphine-containing polymers 103–7 dendrimers 111, 112, 113 polyethylene oligomers 124 bis(phosphonic acids) 163, 164 BOX ligands 452 clay/laponite supports 452–3 fluorous chiral catalysts 185–7 insoluble resin-supported 55–6, 453–4 silaceous supports 27–8, 455–8 soluble polymers 79–80, 81, 89–90 CALB enzyme 393–4 calixarenes 210–11 capping procedures 455–6 carbon dioxide 199–200 in aqueous biphasic reactions 219–20 cycloaddition reactions 66–8, 243 see also supercritical carbon dioxide carbonylation 368 carbonyl–ene reactions 279–80 carbon–carbon bond formation in ionic liquids 275–83 organocatalysts 94 insoluble resin-supported catalysts 50–6

carboxamides 62 catalysis, defined 1 catalyst efficiency and immobilization 429–30 leaching 7–8, 81 optimization 436–7 stability/degradation 5–7, 309, 426–7, 428, 429, 434 catalyst immobilization see immobilization catalyst precursor 2–3 catalyst recovery, assay of 1–2 catalyst precursor vs catalyst 2–3 catalyst resting state 3–5 loss mechanisms decomposition 5–6 leaching 7–8 measures of recoverability gravimetric analysis 12–13 product yield/conversion and turnover frequency 8–12 Cavitron reactor 227 cetyltrimethylammonium bromide 18, 23, 24, 204–6 chalcone epoxidation 333 chincona derived catalysts 89, 96, 311–13, 336 a-chloro esters 388, 390 Chyrazyme L2-C2 393–4 citronellol 393–4, 420 clay supports 452–3 co-condensation method 23–5 and control of silica support morphology 29–31 commercial supports 438 connectivity handle 437, 440 continuous flow processes 380–1, 434, 437 microencapsulated catalysts 368–9 supercritical carbon dioxide systems 232–3, 236, 241, 245 see also microreactor technology; mini flow reactors conversion profile 2 conversion rate 8–12 cooperative catalysis 35–43, 45, 430, 432–3 Corey–Bakshi–Shibata catalyst 61 costs 435 covalently bonded catalysts 20–1 critical solution temperature 118 cross-coupling reactions and biphosphine containing polymers 108

Index microencapsulated catalysts 346, 352–3, 358, 360–1, 363, 368 in mini flow reactor 383, 384, 389, 390 CTAB 18, 23, 24, 204–6 CuAAC reactions 79–80 cyanohydrins 421 cyanopropyl group 36–7 cyanosilylation 41–2, 284–5 cyclization microencapsulated catalyst 346 in mini flow reactor 385, 387, 395, 400–1 resin-supported catalysts 66–9 cycloaddition 66–8 organocatalysed 328–31 photocatalysed 418–19 soluble polymer-supported 79–80 thermomorphic 144 cyclodextrin supports 207–10, 449–50 cyclopropanation fluorous chiral catalyst 186, 187 insoluble resin-supported 55–6 in ionic liquids 281–2 microencapsulated catalyst 349, 350 mini flow reactor 383 decaffeination 230 decomposition/degradation, catalyst 5–7, 309, 426–7, 429, 434 dendrimer catalyst systems 78, 80 b-amino alcohol-containing polymers 110, 111 optically active dendronized polymers 111–14 and Pd nanoparticles 138 and proline immobilization 447–9 thermomorphic catalysis 125, 131, 143 dialkylaminopyridine catalysts 316–17 dialkylzinc see organozinc Diels–Alder reactions fluorous chiral catalyst 191 in ionic liquids 275–7 Difasol process 201 dihydroamino acids, hydrogenation of 261–4 dihydroxylation insoluble resin-supported catalyst 60 in ionic liquids 272–5 microencapsulated catalysts 344, 350–2, 353, 354, 355, 373 dimethylaminopyridine (DMAP) ligand 34–5, 37, 45, 440–1

dimethylcarbonate 243 divinyl benzene copolymer DPEN ligands 158–60 DPPA ligands 103, 105 DPPF ligands 103, 105

465

356–7

enamines, asymmetric hygrogenation of 269–70 enzymatic catalytic behaviour 433 epibromohydrin 395–6 epoxidation fluorous chiral catalysts 182–3, 188 insoluble-resin supported catalyst 58–9 in ionic liquids 271–2 microencapsulated catalyst 346, 347, 356 organocatalysed 332–3 self-supporting catalysts 160–2, 170, 171 silica supported catalyst 26 ethylphenylsulfonic acid 40–1 flow process see continuous flow fluorination 285–6 fluorous chiral catalysts 179–97 background 179–82 recovery concepts 127–8, 170–81 fluorous nitrogen ligands 133–4, 141, 146, 182–92 bis(oxazoline) ligands 185–7 miscellaneous amine-based 190–2 proline-based 187–9 pyrrolidine sufonamides 188, 189–90 salen ligands 182–5 fluorous oxygen ligands 192–4 BINOLs 192–3 diols 193–4 fluorous phosphorus ligands 3–4, 142–3, 194–6 BINAP analogs 96, 195–6 MOP ligands 194–5 supercritical carbon dioxide systems 234–5 fluorous solvents 215, 415, 416 formamide catalysts 65–6 Friedel–Crafts reactions 4, 280, 343 FSM-type silica 18 gatekeeping effect 38–40 gel type resins 50 gravimetric analysis 12–13 Grubbs/Grubbs–Hoveyda catalysts 82–4

5–6, 79,

466

Index

handles 437, 440 Hartwig–Buchwald reaction 397, 399 Heck reaction and catalyst assay 8–9, 11–12 fluorous chiral catalysts 195–6 microencapsulated catalyst 360–1, 368, 371 in microflow reactor 417 in mini flow reactor 396, 397, 398, 399 Henry reaction 39–40, 41–2 in mini flow reactor 394 heterogeneous catalysts characterization 26 multi-site vs single site 16–17 vs homogeneous 16 homogeneous catalysis 16 reactions with biphasic separation 214–20 homogeneous catalysts 16 Huisgen cycloaddition 67–8 hydroamination intramolecular on resin-support 68–9 thermomorphic 129 hydroboration 233 hydrocyanation 335 hydroformylation aqueous biphasic systems 203–28 commercial process 201, 202–3 in mini flow reactor 385, 386 resin-supported catalysts 59–60 silica supported catalyst 20 supercritical carbon dioxide systems 232–5, 237–8, 240–3 thermomorphic catalysis 126–9 hydrogen bonding 27 hydrogen peroxide 6 hydrogenation BINAP containing polymers 107 fluorous chiral catalysts 184–5 in ionic liquids 261–70 microencapsulated catalysts 357, 359, 362, 367, 372 self-supporting catalysts 159–60, 170–1 silica-supported catalysts 19, 28 supercritical carbon dioxide systems 233, 235, 236, 237, 239, 243–5 thermomorphic catalysts 122–6 hydrosilylation and catalyst assay 3, 4, 9–11, 13 microencapsulated catalysts 366 thermomorphic reactions 141–5

hydrovinylation in ionic liquids 286–7 in supercritical carbon dioxide 235, 241 hydroxyproline catalysts 63–4 hydroxyprolylthreonine catalysts 64–5 hypervalent silicon compounds 313–15 imidazole salts 66 imidazolidinones 89, 96, 97, 328–31 imines, asymmetric hydrogenation of 269–70 immobilization strategies 427–61 practical aspects choice of support 437–40 linker/spacer selection 441 location of connection handle 440–1 prerequisite conditions catalyst stability/degradation 426–7 catalytic efficiency 429–30, 432–3 stereoselectivity 430–1, 433 reasons for immobilization continuous flow processes 434 cost factors 435 optimization of catalysis 435–7 organocatalysed reactions 303 stability and revovery of catalyst 434 silica supports see mesoporous silica functionalization soluble polymer supports 79–80 supports compared 17–18 for bis(oxazoline) ligands 452–8 for proline 442–50 incarceration, polymer 357–60, 387 induction period 3, 9, 10 insoluble resin-supported catalysts 49–75 background 49–51 carbon–carbon bond formation Cu-BOX cyclopropanation 55–6 organozinc addition to aldehydes 53–4 Pd-catalysed allylic substitution/cross coupling 50–3 Rh-catalysed intermolecular C–H bond activation 54–5 organocatalyzed reactions 62 allylation of aldehydes 65–6 asymmetric aldol reactions and aminoxylation 63 asymmetric tandem reaction 64–5 nucleophilic substitutions 66 oxidation of alkanes, alkenes and alcohols 57–8

Index dihydroxylation of alkenes 60 epoxidation of alkenes 58–9 hydroformylation of olefins 59–60 sufides to sufoxide 56–7 reduction carboxamides to amines 62 of ketones 61 ring formation cycloaddition 66–8 intramolecular hydroamination 68–9 use in mini flow reactors 387–92 intermolecular C–H activation 54–5 inverse phase transfer catalysts 210–11 inverse temperature-dependent solubility 118 ionic liquids 259–61, 446–7 aldol reactions copper catalysed Mukaiyama 280–1 organocatalysed 287–90 alkyne addition to imines 286 Baylis–Hillman reaction 293–4 carbon dioxide cycloadditions 66–7 common ionic liquids listed 260 cyanosilylation of aldehydes 284–5 C–C bond formation, metal catalysed allylation of aldehydes and ketones 282–3 allylic substitution 277–9 carbonyl–ene reactions 279–80 cyclopropanation 281–2 Diels–Alder reactions 275–7 Friedel–Crafts reaction 280 Michael reaction 283 Mukaiyama aldol reaction 280–1 fluorination 285–6 hydrogenation acetophenone by transfer hydrogenation 270–1 dihydroamino acids 261–4 imines and enamines 269–70 keto esters 266–8 unfunctionalized ketones 268–9 unsaturated acids and esters 264–6 hydrovinylation of styrene 286–7 Mannich reaction 292–3 Michael reaction metal catalyzed 273 organocatalyzed 290–2 and mini flow reactor 385

467

oxidation dihydroxylation of olefins 272–5 epoxidation of olefins 271–2 and proline immobilization 20, 326, 446–7 pyrollidine based organocatalysts 317–18 ring-opening of epoxides 283 sulfimidation of sufides 287 supercritical carbon dioxide systems 240–4 iron oxide supports 44–5, 451 isophorone 245 isotope-labelled tracers 413, 414 keto esters, hydrogenation of 266–8 ketones, hydrogenation of 268–9 Knoevenagel condensation 384, 388 Kumada cross-coupling 383, 384 b-lactams 336, 388, 389 lanthanum catalysts 69, 169–71 laponite 452–3 layered double hydroxide supports 451 leaching, catalyst 7–8, 81 Lewis acids 3 microencapsulated catalyst 342–4 Lewis base, organocatalysts 313–19 limonene 420 linear polymeric chiral catalysts 101–15 background 101–2 1,10 -Bi-2-naphthol(BINOL)-based systems 102–3, 103 biphosphine-containing systems 103, 105 BINAP 105–8 mixed BINAP–BINOL systems 108–10 dendritic systems 111–14 linkers/spacers 437, 441 liquid–liquid biphasic systems 413–17 see also biphasic catalysis MacMillan’s catalyst 89, 328–31, 428–9, 435 macroporous resins 50 magnetic nanoparticles 44–5, 307, 310–11, 324 magnetite supports 44–5, 451 manganese–porphyrin catalysts 58–9 Mannich reaction 292–3, 343, 365 mass transfer promoters 203–12 MCM41 mesoporous silica 450–1 MeOPEG supports 78, 430, 442, 455 mercury amalgam 13 Merrifield resin 389, 395

468

Index

mesocellular silica foam 383, 455–6, 457 mesoporous mixed metal oxides 43–4 mesoporous silica functionalization 15–47 background 17–18 homogeneous vs heterogeneous catalysts 16 multi-site vs single site catalysis 16–17 support materials 17–18 and BOX ligands 456, 457 characterization of heterogenous catalysts 26 conventional functionalization 18–19 co-condensation method 23–5 covalent bonding/silylation reagents 20–1 noncovalent binding 19–20 post-synthetic grafting silylation 21–3 metalosalen complexes 25–6 mini flow reactors 382–7 multifunctionalization and cooperative catalysis 35–8, 430, 432–3 gatekeeping effect and selectivity 38–40 synergic acid/base catalysis 40–3 particle/pore morphology control nanovoid/internal pore utilization 31–5 polar/nonpolar precursors and particle size 29–31, 32 surface interactions 26–7, 27–9 reduced by spacers 27–9 metal leaching 7–8, 81 metal phosphonates 156 metal scavenging polymers 371, 372 metal–binaphthyl complexes 92, 94 metal–salen complexes see salen complexes metathesis 5–6, 82–4 ring-closing 385, 387, 395, 400–1 ring-opening 79, 134, 385, 387 Michael reactions 168–9 in ionic liquids 273, 283 microencapsulated catalyst 343, 348, 365 mini flow reactor 384 microemulsions 215–16 microencapsulated metal catalysts 341–77 catalysts and polymers summarized 374–5 cross-linked polystyrene with divinyl benzene 356–7 with oligo(ethylene glycol) 350–65 noncross-linked polystyrene 344–50 derivatives 350–2 polysulphone 353–4

poly(xylylviologen dibromide) 354–5 urea group cross-linked polyphenylene 367–74 microporous resins 50 microreactor technology acid-catalysed reactions isotope-labelled tracers 413, 414 nitration 411–12 asymmetric catalytic reactions 421 high temperature/pressure conditions 421–3 liquid-liquid biphasic systems 413–16 segmented flow systems 416–17 mini flow reactors vs micro flow reactors 381–2 photocatalysis 418–20 mini flow reactors 335–6, 379–410 batch vs flow processes 380–1 catalyst immobilization 382 monolithic supports 392–401 polymer supports 387–92 silica supports 382–7 enabling technologies 379–80 micro vs mini flow reactors 381–2 monolithic supports 392–401 monophosphine (MOP) ligands 194–5 montmorillonite 452 MSM-type silica see mesoporous silica MSU-type silica 18 Mukaiyama reaction 280–1, 365, 415–16 multi-site catalysis 16–17 multifunctionized mesoporous silica 35–43, 430, 432–3 multiphasic systems 201–2 N-heterocyclic carbene (NHC) ligands 52, 81, 82, 89 nanoparticles encapsulated palladium 357 magnetic 44–5, 307, 310–11, 324 silaceous see mesoporous silica thermomorphic catalysis 125, 138 nitration reactions 411–12 nitroaldol reaction 39–40 nitroalkane addition 309–10 noncovalent immobilization BOX ligands 452–3 on mesoporous silica support 19–20 organocatalysts 319 norbornenes, functionalized 79

Index oligo(ethylene glycol) copolymer 350–65 optimization, catalyst 436–7 organic peroxides 6 organoalkoxysilanes 29–31 organocatalyzed reactions 62–6, 301–40 achiral catalysts miscellaneous 309–11 oxidation 304–7 phase transfer 307–9 allylation of aldehydes 65–6 aminoxylation 63 asymmetric aldol 63 asymmetric tandem reaction 64–5 background/general considerations 81, 310–13, 334–7 chiral catalysts amino acid based 319–34 imidazolinone derivatives 238–331 miscellaneous amino acids 331–4 proline derivatives 320–8 Lewis Base 313–19 miscellaneous 319 phase transfer 311–13 fluorous systems 142–3, 187–90 in ionic liquids 290–2 nucleophilic substitutions 66 thermomorphic catalysts 142–4 organosilanes 23–5 see also silylation organozinc reagents 53–4, 165, 167, 190, 192–4 ouabain hexaacetate 417 oxidation and catalyst decomposition 6, 7 insoluble resin-supported catalysts 56–60 in ionic liquids 272–5 microencapsulated catalysts 348, 349–50, 356–7, 365–6 in mini flow reactor 387–8 organocatalyzed achiral 304–7, 309, 374 singlet oxygen 419–20 TEMPO 87–8, 145–6, 304–7, 365–6, 423, 434 thermomorphic reactions 145–7 palladium catalysts insoluble resin-supported 50–3, 70 silica-supported 32–4 soluble polymer-bound 84–5, 86, 89, 91 thermomorphic reactions 126, 130–8

469

Paraxon 45 PASSflow matrix 395, 398, 400 PEG support 439 amphiphilic PS–PEG 50 for Box ligands 456 for proline 95, 442–3, 445 soluble polymer systems 78, 95, 97 TEMPO catalyst 305 thermomorphic catalysis 132–3, 136–7 PEPPSITM catalyst 397, 399 pH dependent solubility 218–20 phase transfer catalysts fluorous 191–2 microreactor alkylation 416 organocatalysis achiral 307–9 chiral 311–13 soluble polymer-bound 85, 87, 96, 97 phosphine functionalized polymers 362–3 dendritic chiral 111, 112, 113 phosphinic acids 70 phosphoramide catalysts 315–16 photocatalysis 418–20 PNIPAM support 121–3, 125 polyacrylic acid catalyst 226 polydimethylsiloxane support 143 polyethylene glycol supports see PEG polyisobutylene support 140 polymer incarceration 50–1, 357–60, 387 polymer-microecapsulation see microencapsulated catalysts polymer-supports see dendrimer; insoluble resin-supported; soluble polymer-bound polymeric chiral catalysts see linear polymeric chiral catalysts polymerization reactions supercritical carbon dioxide systems 235 thermomorphic 138–42 poly(N-alkylacrylamide) supports 121–3, 125, 132, 146 polyphenylene, cross-linked 367–74 polystyrene supports 49–50, 78 amphiphilic PS–PEG resins 50 BOX ligands 452, 453, 455 proline 443, 444, 445 polysulphones 353 poly(xylylviologen dibromide) 354–5 porphyrin catalyst 309 precursor, catalyst 2–3

470

Index

proline-based catalysts 320–8 applications tabulated 95 cyclodextrin supported 326, 449–50 dendrimer supported 447–9 fluorous chiral catalysts 187–9 insoluble resin-supported 54–5, 63–5, 322–4 and ionic liquids 20, 287–90, 291, 292–3, 326, 446–7 magnetite and layered double hydroxides 451 polymer supports compared 442–6 prolinamides 326–8 silica supported 324–6, 450–1 soluble polymer–bound 89, 95–6 2-pyridyldiphenylphosphine 53 pyrollidine-based organocatalysts 317–18 pyrrolidine sufonamides 188, 189–90 radiopharmaceuticals 413, 414 rate/reaction profile 2, 9 reactor design 227–8 recovery, catalyst assay techniques see catalyst recovery strategies 79, 428 reduction fluorous chiral catalyst 188, 191 insoluble resin-supported catalysts 61–2 microencapsulated catalyst 246, 372 organocatalysed 314–15, 333–4 see also hydrogenation resting state, catalyst 3–5 ring-closing metathesis 385, 387, 395, 400–1 ring-opening epoxides in ionic liquids 283 metathesis polymerization (ROMP) 79, 134, 385, 387 Ruhrchemie/Rhoˆne-Poulenc process 201, 202 salen complexes fluorous chiral catalysts 182–5 in mini flow reactor 383 salen-containing polymers 108 self-supported catalyst 160 silica supported 26, 383 soluble polymers 26, 90–2, 93, 383 SBA-type silica 18 segmented flow systems 416–17 self-encapsulation 62 self-supported asymmetric catalysts 155–77

background and approaches to 155–6 multitopic chiral ligands linked with metal centres 168–72 post-synthetic modifications of coordination polymers 163–8 subunits linked via metal-ligand coordination 156–63 Shell-SHOP process 201 Shibasaki’s catalyst 169–70 shock-wave reactor 227 silanes see organosilanes silazane derivatives 21 silica supports and bis(oxazolines) immobilization 455–8 described 17–18 mesocellular foam 383, 455–6, 457 for mini flow reactors 382–7 and proline immobilization 450–1 see also mesoporous silica functionalization silicone resin 62 silylation 20–1 co-condensation method 23–5 and microencapsulated catalyst 343 post-synthetic grafting method 21–3 silica support morphology and internal pore surface 31–5 particle size 29–31, 32 silylcyanation 421 singlet oxygen 419–20 smart catalyst supports 125 soluble polymer-bound catalysts 77–100 achiral catalysts 81–2 palladium for cross coupling 84–5, 86 phase transfer catalysts 85, 87 ruthenium–carbene for olefin metathesis 82–4 TEMPO oxidation 87–8 chiral catalysts 88–9, 100–2 metal–binaphthyl complexes 92, 94 metal–bis(oxazoline) complexes 89–90, 91 metal–salen complexes 90, 92, 93 organocatalysts 94 chicona alkaloids 96 imidazolidinones and thioureas 96, 97 phase transfer 96, 97 proline-based 95 general considerations/ background 77–9, 98 immobilization strategies 79–80

Index

471

metal catalysts 81, 89 organocatalysts 81, 89, 94 see also dendritic chiral catalysts; linear polymeric chiral catalysts Sonogashira reaction 363, 396, 398, 422 spacers/linkers 437, 441 PEG-supported TEMPO 305 and silica support 27–9 Staudinger ligation 56, 80 stereoselectivity 430–1, 435–6 Stille reaction 368 styrene copolymers 211–12 sulfimidation of sufides 287 sulfoxantphos catalyst 226 supercritical carbon dioxide systems 229–47 catalyst recycling approaches 230–2 carbon dioxide–water biphasic systems 236–40 ionic liquid-carbon dioxide biphasic systems 240–4 microencapsulated catalysts 368–70 solid–carbon dioxide biphasic systems 245–6 solvent and extraction 232–5 critical parameters and phase diagram 228–9 overview 229–30, 246–7 supported aqueous phase catalysis 220–7 supports see under immobilization strategies support–catalyst cooperation 439 surface-active ligands 212–14 surface-functionalization see immobilization surfactants 204–7, 212–16 Suzuki–Miyaura reactions microencaspsulated catalysts 346, 352–3, 358, 361, 364, 368, 370–1 in mini flow reactor 389, 390, 396, 397, 398, 399

organocatalyzed reactions 304–7 Tentagel resin 357 tetraethoxysilane 18, 23, 24 thermomorphic catalysts 117–53 background 117–18 poly(N-isopropylacrylamide) supports 121–2 separation strategies 118–22, 181 1,3-cycloadditon 144 hydroamination 129 hydroformylation 126–9 hydrogenation 122–6 hydrosilylation 144–5 organocatalysis 142–4 oxidation 145–7 palladium-catalyzed reactions allylic substitution 130–1 cross-coupling reactions 131–8 polymerization 138–42 thermoregulated catalysis 120–1, 126, 129 phase transfer 216–17 thiazine synthesis 71–2 thiourea-based organocatalysts 320, 334–5 soluble polymers 96, 97 TPPTS ligand 212–17 and commercial hydroformylation 202 supported aqueous phase catalysis 221–4, 226 transesterification 393–4 transfer hydrogenation in ionic liquids 270–1 microencapsulated catalyst 372–3 mini flow reactor 385, 386, 396, 397 trialkylsilanes 455–6 triazole synthesis 67–8, 391 trichlorosilane 314 triphenylphosphine trisulfonate see TPPTS turnover frequency 8–12

TantaGel 60 temperature-dependent miscibility/ solubility 118–19 TEMPO oxidation 87–8, 145–6, 434 microencapsulated catalyst 365–6 microreactor 423

Wilkinson’s catalyst

2–3, 4

yield, product 8–12 zeolites 18 zirconium phosphonates

156–61, 163–5