Modern Aldol Reactions

R. Mahrwald (Ed.) Modern Aldol Reactions Vol. 1: Enolates, Organocatalysis, Biocatalysis and Natural Product Synthesis ...

Author: Rainer Mahrwald

454 downloads 1157 Views 6MB Size Report

This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!

Report copyright / DMCA form

DOWNLOAD PDF

R. Mahrwald (Ed.) Modern Aldol Reactions Vol. 1: Enolates, Organocatalysis, Biocatalysis and Natural Product Synthesis

Modern Aldol Reactions. Vol. 1: Enolates, Organocatalysis, Biocatalysis and Natural Product Synthesis. Edited by Rainer Mahrwald Copyright 8 2004 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 3-527-30714-1

Also of Interest Grubbs, R. H. (Ed.)

Handbook of Metathesis 3 Volumes 2003

ISBN 3-527-30616-1

Nicolaou, K. C., Snyder, S. A.

Classics in Total Synthesis II More Targets, Strategies, Methods 2003

ISBN 3-527-30685-4 (Hardcover with CD-Rom) ISBN 3-527-30684-6 (Softcover)

De Meijere, A., Diederich, F. (Eds.)

Metal-Catalyzed Cross-Coupling Reactions Second, Completely Revised and Extended Edition 2 Volumes 2004

ISBN 3-527-30518-1

Krause, N., Hashmi, A. S. K. (Eds.)

Modern Allene Chemistry 2 Volumes 2004

ISBN 3-527-30671-4

Rainer Mahrwald (Ed.)

Modern Aldol Reactions Vol. 1: Enolates, Organocatalysis, Biocatalysis and Natural Product Synthesis

PD Dr. Rainer Mahrwald Department of Organic Chemistry Humboldt University Brook-Taylor-Str. 2 12489 Berlin Germany

9 This book was carefully produced. Nevertheless, editor, authors and publisher do not warrant the information contained therein to be free of errors. Readers are advised to keep in mind that statements, data, illustrations, procedural details or other items may inadvertently be inaccurate. Library of Congress Card No.: Applied for British Library Cataloguing-inPublication Data: A catalogue record for this book is available from the British Library. Bibliographic information published by Die Deutsche Bibliothek Die Deutsche Bibliothek lists this publication in the Deutsche Nationalbibliograﬁe; detailed bibliographic data is available in the Internet at http://dnb.ddb.de ( 2004 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim All rights reserved (including those of translation in other languages). No part of this book may be reproduced in any form – by photoprinting, microﬁlm, or any other means – nor transmitted or translated into machine language without written permission from the publishers. Registered names, trademarks, etc. used in this book, even when not speciﬁcally marked as such, are not to be considered unprotected by law. Printed in the Federal Republic of Germany. Printed on acid-free paper. Typesetting Asco Typesetters, Hong Kong Printing Strauss Gmbh, M€ orlenbach Bookbinding Litges & Dopf Buchbinderei GmbH, Heppenheim ISBN 3-527-30714-1

v

Foreword Historically, the stimulus for the development of a particular reaction has been interconnected with a class of natural products whose synthesis would be greatly facilitated by the use of that particular bond construction. For example, the steroid synthesis challenges proved instrumental in the development of the Diels–Alder reaction. So too the synthesis challenges associated with the macrolide antibiotics have provided the motivation for the development of the full potential of the aldol addition reaction. R. B. Woodward’s 1956 quote on the ‘‘hopelessly complex’’ architecture of the erythromycins was probably stimulated, in part, by the fact that the aldol reaction existed in a completely underdeveloped state ﬁve decades ago. O Me

Me Me

OH

Me

OH Me

OH

O

The erythromycin-A structure, as viewed by Woodward in the ’50s

Me O

Et

O

H

O

O Me Me

H

O Me

MeO OH

NMe2 OH

‘‘Erythromycin, with all of our advantages, looks at present quite hopelessly complex, particularly in view of its plethora of asymmetric centers.’’ R. B. Woodward in Perspectives in Organic Chemistry; Todd, A. Ed.; WileyInterscience, New York, 1956, page 160.

The challenges associated with the development of this reaction are also embodied in the more general goals of acyclic stereocontrol that have been under active investigation for nearly twenty-ﬁve years. In these studies, the goal of understanding pi-face selectivity at trigonal carbon centers for a multitude of organic transformations has been the ultimate objective. From these research activities, a host of stereochemical models have evolved, such as the Felkin–Anh model for carbonyl addition and the Zimmermann– Traxler aldol stereochemical model for aldol diastereoselection.

vi

Foreword

The development of modern aldol reaction methods has evolved through a succession of pivotal discoveries that have advanced the whole ﬁeld of stereoselective synthesis: A. Development of enolization strategies for the formation of (E) and (Z) enolates. B. Development of kinetic diastereoselective aldol addition variants through the discovery of optimal metal architectures [B(III), Ti(IV), Sn(II)]. C. Discovery of aldol reaction variants such as the Lewis acid catalyzed addition of enolsilanes to aldehydes (Mukaiyama aldol variant). D. Development of chiral enolates exhibiting exceptional pi-face selectivities. E. Development of chiral metal complexes as Lewis acid aldol catalysts. This two-volume series on aldol addition reaction methodology brings together an up-to-date discussion of all aspects of this versatile process. The reader will gain an appreciation for the role of metal enolate architecture in aldol diastereoselectivities (Vol. I; Chapters 1–3) and for the utility of chiral metal complexes in the catalysis of the Mukaiyama aldol reaction (Vol. II; Chapters 1–3, 5). In Vol. II; Chapter 6, enantioselective catalytic processes incorporating both enolization and addition are surveyed, as is the exciting progress being made in the use of chiral amines as aldol catalysts (Vol. I; Chapter 4). This highly active area of research will continue to develop ever more versatile chiral catalysts and stereochemical control concepts. Students and researchers in the ﬁeld of asymmetric synthesis will greatly proﬁt from the contributions of this distinguished group of authors who have so insightfully reviewed this topic. May 2004

David A. Evans Harvard University

vii

Contents Volume 1 Preface

xvii

List of Contributors 1

1.1 1.2 1.3 1.4 1.4.1 1.4.1.1 1.4.1.2 1.4.2 1.4.3 1.4.4 1.5

2

2.1 2.2 2.3 2.3.1 2.3.2

xix

Fundamentals and Transition-state Models. Aldol Additions of Group 1 and 2 Enolates 1 Manfred Braun Introduction 1 The Acid or Base-mediated ‘‘Traditional’’ Aldol Reaction 2

The Aldol Addition of Preformed Enolates – Stereoselectivity and Transition-state Models 9 Stereoselective Aldol Addition of Lithium, Magnesium and Sodium Enolates 25 Addition of Chiral Enolates to Achiral Carbonyl Compounds 26 a-Substituted Enolates 26 a-Unsubstituted Enolates 32 Addition of Achiral Enolates to Chiral Carbonyl Compounds 41 Addition of Chiral Enolates to Chiral Carbonyl Compounds 49 Addition of Achiral Enolates to Achiral Carbonyl Compounds in the Presence of Chiral Additives and Catalysts 51 Conclusion 52 References 53 The Development of Titanium Enolate-based Aldol Reactions Arun K. Ghosh, M. Shevlin Introduction 63 Additions of Enolates to Ketones 65

63

Addition of Enolates Without a-Substituents to Aldehydes 66 Stereoselective Acetate Aldol Reactions Using Chiral Auxiliaries 67 Stereoselective Acetate Aldol Reactions Involving Chiral Titanium Ligands 69

viii

Contents

2.3.3 2.4 2.4.1 2.4.1.1 2.4.1.1.1 2.4.1.1.2 2.4.1.1.3 2.4.1.2 2.4.1.2.1 2.4.1.2.2 2.4.1.2.3 2.4.1.2.4 2.4.1.3 2.4.1.4 2.4.2 2.4.2.1 2.4.2.2 2.4.2.2.1 2.4.2.2.2 2.4.2.3 2.5 2.5.1 2.5.1.1 2.5.1.2 2.5.1.3 2.5.1.4 2.5.2 2.5.2.1 2.5.2.2 2.5.2.3 2.5.2.4 2.5.2.5 2.5.3 2.5.3.1 2.5.3.2 2.5.3.3 2.5.3.4 2.5.3.5 2.6 2.6.1

Alternative Approaches to Acetate Aldol Adducts 70 Addition of Enolates with a-Substituents to Aldehydes 72 Syn Diastereoselectivity 74 Synthesis of syn Aldols in Racemic Form 75 Reactions of Ketones 75 Reactions of Esters and Thiol Esters 77 Aldol Reactions of Aldehyde Hydrazones 78 Synthesis of Optically Active syn Aldols Using Chiral Auxiliaries 80 Amino Acid-derived Oxazolidinone and Related Auxiliaries 80 Camphor-derived Chiral Auxiliaries 84 Aminoindanol and Amino Acid-derived Chiral Auxiliaries 87 Other Chiral Auxiliaries 90 Synthesis of Optically Active syn Aldols Using Chiral Titanium Ligands 92 Synthesis of Optically Active syn Aldols with Chiral Enolates 95 Anti Diastereoselectivity 98 Synthesis of anti Aldols in Racemic Form 98 Synthesis of Optically Active anti Aldols by Use of Chiral Auxiliaries 99 Aminoindanol and Related Chiral Auxiliaries 99 Oxazolidinethione and Oxazolidineselone Chiral Auxiliaries 103 Synthesis of Optically Active anti Aldols by Use of Chiral Titanium Ligands 104 Natural Product Synthesis via Titanium Enolate Aldol Reactions 105 Lactone Natural Products 105 Tetrahydrolipstatin 106 Myxopyronins A and B 106 Callystatin A 107 AI-77-B 108 Macrolide Natural Products 110 Epothilone 490 110 Cryptophycin B 110 Amphidinolide T1 111 Rapamycin 112 Spongistatins 1 and 2 113 Miscellaneous Natural Products 114 Tautomycin 114 Crocacin C 115 Stigmatellin A 116 Denticulatin B 117 Membrenone C 119 Typical Experimental Procedures for Generation of Titanium Enolates 120 Experimental Procedures 120

Contents

2.6.2 2.7

Alternative Approaches to Titanium Enolate Generation Conclusion 121 References 122

3

Boron and Silicon Enolates in Crossed Aldol Reaction Teruaki Mukaiyama and Jun-ichi Matsuo Introduction 127

3.1 3.2 3.2.1 3.2.2 3.2.3 3.2.4 3.2.5 3.2.6 3.3 3.3.1 3.3.2 3.3.3 3.3.4 3.3.5 3.3.6 3.3.6.1 3.3.6.2

4

4.1 4.2 4.2.1 4.2.1.1 4.2.1.2 4.2.2 4.2.2.1 4.2.2.2 4.3 4.3.1 4.3.1.1 4.3.1.2 4.3.2 4.3.2.1

121

127

Crossed Aldol Reactions Using Boron Enolates 127 Discovery of Aldol Reaction Mediated by Boron Enolates 127 New Method for Direct Generation of Boron Enolates 129 Regioselectivity on Generation of Boron Enolates 130 Stereoselective Formation of (E) or (Z) Boron Enolates 131 syn-Selective Asymmetric Boron Aldol Reactions 134 anti-Selective Asymmetric Aldol Reaction 135 Crossed Aldol Reactions Using Silicon Enolates 137 Discovery of Silicon Enolate-mediated Crossed Aldol Reactions 137 Lewis Acid-catalyzed Aldol Reactions of Silicon Enolates 143 Non-catalyzed Aldol Reactions of Silicon Enolates 147 Lewis Base-catalyzed Aldol Reactions of Trimethylsilyl Enolates 148 Diastereoselective Synthesis of Polyoxygenated Compounds 149 Asymmetric Aldol Reactions Using Chiral Tin(II) Lewis Acid Catalysts 150 Stoichiometric Enantioselective Aldol Reaction 151 Catalytic Enantioselective Aldol Reaction 154 References 155 Amine-catalyzed Aldol Reactions Benjamin List Introduction 161

161

Aminocatalysis of the Aldol Reaction 162 Intermolecular Aldolizations 163 Aldehyde Donors 164 Ketone Donors 166 Intramolecular Aldolizations 167 Enolexo Aldolizations 167 Enolendo Aldolizations 171 Asymmetric Aminocatalysis of the Aldol Reaction Intramolecular Aldolizations 173 Enolendo Aldolizations 173 Enolexo Aldolizations 177 Intermolecular Aldolizations 179 Ketone Donors 179

173

ix

x

Contents

4.3.2.2

Aldehyde Donors References 196

5

Enzyme-catalyzed Aldol Additions Wolf-Dieter Fessner Introduction 201 General Aspects 202 Classiﬁcation of Lyases 202

5.1 5.2 5.2.1 5.2.2 5.2.3 5.3 5.3.1 5.3.2 5.3.3 5.3.4 5.4 5.4.1 5.4.2 5.4.3 5.4.4 5.4.5 5.4.6 5.5 5.6 5.7 5.8 5.9

6

6.1 6.2 6.2.1 6.2.2 6.2.2.1 6.2.2.2 6.2.2.3 6.3 6.3.1 6.3.2

193

201

Enzyme Structure and Mechanism 204 Practical Considerations 207 Pyruvate Aldolases 208 N-Acetylneuraminic Acid Aldolase 208 KDO Aldolase 216 DAHP Synthase 217 KDPG Aldolase and Related Enzymes 218 Dihydroxyacetone Phosphate Aldolases 221 FruA 222 TagA 224 RhuA and FucA 224 DHAP Synthesis 227 Applications 230 Aldol Transfer Enzymes 246 Transketolase and Related Enzymes 247 2-Deoxy-D-ribose 5-Phosphate Aldolase 250 Glycine Aldolases 254 Recent Developments 257 Summary and Conclusion 258 References 260 Antibody-catalyzed Aldol Reactions 273 Fujie Tanaka and Carlos F. Barbas, III Introduction 273 Generation of Aldolase Antibodies 273 Antibody as Catalyst Scaﬀold 273

Generation of Aldolase Antibodies that Operate via an Enamine Mechanism 274 Reactive Immunization with the Simple Diketone Derivative 275 Combining Reactive Immunization with Transition-state Analogs 277 Reactive Immunization with other Diketones 279 Aldolase Antibody-catalyzed Aldol and Retro-aldol Reactions 279 Antibody 38C2-catalyzed Aldol Reactions 280 Antibody 38C2-Catalyzed Retro-aldol Reactions and their Application to Kinetic Resolution 283

Contents

6.3.3 6.3.4 6.3.5 6.3.6 6.4 6.5 6.6 6.7 6.8 6.9 6.10

Aldol and Retro-aldol Reactions Catalyzed by Antibodies 93F3 and 84G3 285 Preparative-scale Kinetic Resolution Using Aldolase Antibodies in a Biphasic Aqueous–Organic Solvent System 288 Aldolase Antibody-catalyzed Reactions in Natural Product Synthesis 290 Retro-aldol Reactions in Human Therapy: Prodrug Activation by Aldolase Antibody 291 Aldolase Antibodies for Reactions Related to an Enamine Mechanism and the Nucleophilic Lysine e-Amino Group 293 Concise Catalytic Assays for Aldolase Antibody-catalyzed Reactions 297 Structures of Aldolase Antibodies and Reaction Mechanism of Nucleophilic Lysine e-Amino Group 298 Evolution of Aldolase Antibodies In Vitro 302 Cofactor-mediated Antibody-catalyzed Aldol and/or Retro-aldol Reactions 305 Summary and Conclusion 305 Experimental Procedures 306 Acknowledgments 307 References 307

7

The Aldol Reaction in Natural Product Synthesis: The Epothilone Story 311 Dieter Schinzer

7.1

History of Epothilones: Biological Source, Isolation, and Structural Elucidation 311 History of Epothilones: The Total Synthesis Race 311 Diﬀerent Strategies with Aldol Reactions: The Danishefsky Synthesis of Epothilone A Relying on Intramolecular Aldol Reaction 312 Diﬀerent Strategies with Aldol Reactions: The Nicolaou Synthesis of Epothilone A Using an Unselective Aldol Reaction 313 Diﬀerent Strategies with Aldol Reactions: The Schinzer Synthesis of Epothilone A with Complete Stereocontrol in the Aldol Reaction 314 Model Study via Chelation Control in the Aldol Reaction by Kalesse 319 Diﬀerent Aldol Strategies: Mulzer’s Total Syntheses of Epothilones B and D 320 Long-range Structural Eﬀects on the Stereochemistry of Aldol Reactions 322 Summary and Conclusion 326 References 326

7.2 7.2.1

7.2.2 7.2.3

7.3 7.3.1 7.4 7.5

Index

329

xi

xii

Contents

Volume 2 Preface

xvii

List of Contributors 1

1.1 1.2 1.3 1.4 1.5

2

2.1 2.1.1 2.1.2 2.1.3 2.1.4 2.2 2.2.1 2.2.2 2.2.3 2.3 2.3.1 2.3.2 2.3.3 2.3.4

3

3.1 3.2 3.3 3.3.1 3.3.1.1 3.3.1.2 3.3.2 3.3.2.1 3.3.2.2

xix

Silver, Gold, and Palladium Lewis Acids Akira Yanagisawa Introduction 1

1

Mukaiyama Aldol Reaction and Related Reactions 1 Asymmetric Aldol Reactions of a-Isocyanocarboxylates Summary and Conclusions 15 Experimental Procedures 18 References 21

8

Boron and Silicon Lewis Acids for Mukaiyama Aldol Reactions Kazuaki Ishihara and Hisashi Yamamoto Achiral Boron Lewis Acids 25 Introduction 25 BF3 Et2 O 26 B(C6 F5 )3 29 Ar2 BOH 30 Chiral Boron Lewis Acids 33 Introduction 33 Chiral Boron Lewis Acids as Stoichiometric Reagents 33 Chiral Boron Lewis Acids as Catalytic Reagents 40 Silicon Lewis Acids 53 Introduction 53 Lewis Acidity of Silicon Derivatives 54 Silicon Lewis Acids as Catalytic Reagents 55

25

Activation of Silicon Lewis Acids by Combination with Other Lewis Acids 60 References 65 Copper Lewis Acids 69 Jeﬀrey S. Johnson and David A. Nicewicz Introduction 69 Early Examples 69

Mukaiyama Aldol Reactions with Cu(II) Complexes 70 Enolsilane Additions to (Benzyloxy)acetaldehyde 70 Scope and Application 70 Mechanism and Stereochemistry 75 Enolsilane Additions to a-Keto Esters 80 Scope and Application 80 Mechanism and Stereochemistry 85

Contents

3.3.3 3.4 3.4.1 3.4.2 3.4.3 3.4.4 3.4.4.1 3.4.4.2 3.4.5 3.5

4

4.1 4.2 4.2.1 4.2.2 4.2.3 4.2.4 4.3 4.3.1 4.3.2 4.4 4.4.1 4.4.2 4.4.3 4.4.4 4.4.5 4.4.6

Enolsilane Additions to Unfunctionalized Aldehydes 88 Additions Involving In-Situ Enolate Formation 90 Pyruvate Ester Dimerization 90 Addition of Nitromethane to a-Keto Esters 91 Malonic Acid Half Thioester Additions to Aldehydes 94 Dienolate Additions to Aldehydes 96 Scope and Application 96 Mechanistic Considerations 97 Enantioselective Cu(II) Enolate-Catalyzed Vinylogous Aldol Reactions 99 Conclusions 101 References 102 Tin-promoted Aldol Reactions and their Application to Total Syntheses of Natural Products 105 Isamu Shiina Introduction 105 Tin-promoted Intermolecular Aldol Reactions 105 Achiral Aldol Reactions 105 The Reaction of Silyl Enolates with Aldehydes or Ketones 108 The Reaction of Silyl Enolates with Acetals 117 Reaction of Dienol Silyl Ethers 120 Tin-promoted Intramolecular Aldol Reactions 121 The Intramolecular Aldol Reaction of Silyl Enolates 121 Reaction of Dienol Silyl Ethers or g-Silyl-a,b-enones 123 Chiral Diamine–Sn(II) Complex-promoted Aldol Reactions 124 Asymmetric Aldol and Related Reactions of Sn(II) Enolates 125 Chiral Diamine–Sn(II) Complex-promoted Aldol Reactions 128 Asymmetric Aldol Reaction of Silyl Enolates 129 Catalytic Asymmetric Aldol Reaction 131 Asymmetric Synthesis of syn- and anti-1,2-Diol Groups 135

Enantioselective Synthesis of Both Enantiomers of Aldols Using Similar Diamines Derived from L-Proline 139 4.5 Asymmetric Total Syntheses of Complex Molecules Using Chiral Diamine–Sn(II) Catalysts 140 4.5.1 Monosaccharides 140 4.5.2 Leinamycin and a Part of Rapamycin 142 4.5.3 Sphingosine, Sphingofungins, and Khafrefungin 145 4.5.4 Febrifugine and Isofebrifugine 147 4.5.5 Altohyrtin C (Spongistatin 2) and Phorboxazole B 148 4.5.6 Paclitaxel (Taxol) 149 4.5.7 Cephalosporolide D 153 4.5.8 Buergerinin F 153 4.5.9 Octalactins A and B 154 4.5.10 Oudemansin-antibiotic Analog 155

xiii

xiv

Contents

4.6 4.7

Conclusions 157 Experimental 158 References 159

5

Zirconium Alkoxides as Lewis Acids 167 Yasuhiro Yamashita and Shu¯ Kobayashi Introduction 167

5.1 5.2 5.3 5.4 5.5 5.6 5.7 5.8

6

6.1 6.2 6.3 6.4 6.5 6.6 6.7 6.8

7

The Asymmetric Mukaiyama Aldol Reaction 169 Asymmetric Hetero Diels–Alder Reaction 175 Reaction Mechanism 180 Structure of the Chiral Zirconium Catalyst 184 Air-stable and Storable Chiral Zirconium Catalyst 187 Conclusion 190 Experimental 191 References 192 Direct Catalytic Asymmetric Aldol Reaction Using Chiral Metal Complexes 197 Masakatsu Shibasaki, Shigeki Matsunaga, and Naoya Kumagai Introduction 197 Direct Aldol Reactions with Methyl Ketones 198 Direct Aldol Reactions with Methylene Ketones 208 Direct Aldol Reaction with a-Hydroxyketones 210 Direct Aldol Reaction with Glycine Schiﬀ Bases 219 Other Examples 221 Conclusion 224 Experimental Section 225 References and Notes 226

Catalytic Enantioselective Aldol Additions with Chiral Lewis Bases Scott E. Denmark and Shinji Fujimori 7.1 Introduction 229 7.1.1 Enantioselective Aldol Additions 229 7.1.1.1 Background 230 7.1.2 Lewis Base Catalysis 233 7.1.3 Organization of this Chapter 235 7.2 Preparation of Enoxytrichlorosilanes 236 7.2.1 General Considerations 238 7.2.2 Preparation of Ketone-derived Trichlorosilyl Enolates 240 7.2.3 Preparation of Aldehyde-derived Trichlorosilyl Enolates 246 7.2.4 Preparation of Trichlorosilyl Ketene Acetals 248 7.3 Preparation of Chiral Lewis Bases 249 7.3.1 Preparation of Chiral Phosphoramides 250 7.3.2 Synthesis of Chiral bis-N-Oxides 251

229

Contents

7.4 7.4.1 7.4.2 7.4.3 7.5 7.5.1 7.5.1.1 7.5.1.2 7.5.2 7.5.2.1 7.5.2.2 7.5.3 7.6 7.7 7.8 7.9 7.9.1 7.9.2 7.9.3 7.10 7.11 7.11.1 7.11.2

8

8.1 8.2 8.2.1 8.2.2 8.2.3 8.2.4 8.3

Enantioselective Aldol Addition of Achiral Enoxytrichlorosilanes 253 Aldol Additions of Achiral Methyl Ketone-derived Enolates 254 Aldol Additions of Cyclic Trichlorosilyl Enolates 263 Addition of Acyclic Ethyl Ketone-derived Enolates 267 Diastereoselective Additions of Chiral Enoxytrichlorosilanes 272 Aldol Addition of Lactate-derived Enoxytrichlorosilanes 273 Methyl Ketone-derived Enolates 273 Ethyl Ketone-derived Enolates 277 Aldol Addition of b-Hydroxy-a-Methyl Ketone-derived Enoxytrichlorosilanes 280 Methyl Ketone-derived Enolates 280 Ethyl Ketone-derived Enolates 282 Addition of Enoxytrichlorosilanes with a b-Stereogenic Center 283 Aldol Additions of Aldehyde-derived Enoxytrichlorosilanes 288 Aldol Addition of Trichlorosilyl Ketene Acetal to Aldehydes and Ketones 294 Lewis Base Activation of Lewis Acids – Aldol Additions of Silyl Enol Ethers to Aldehydes 298 Toward a Uniﬁed Mechanistic Scheme 305 Cationic Silicon Species and the Dual-pathway Hypothesis 306 Uniﬁed Mechanistic Scheme 310 Structural Insights and Modiﬁcations 312 Conclusions and Outlook 315 Representative Procedures 316 Preparation of Enoxytrichlorosilanes 316 Aldol Addition of Ketone-derived Enoxytrichlorosilane 317 References 319 The Aldol–Tishchenko Reaction R. Mahrwald Introduction 327

327

The Aldol–Tishchenko Reaction 327 The Aldol–Tishchenko Reaction with Enolizable Aldehydes 327 The Aldol–Tishchenko Reaction with Ketones and Aldehydes 329 The Evans–Tishchenko Reduction 334 Related Reactions 339 Representative Procedures 341 References 342 Index

345

xv

xvii

Preface The aldol reaction was ﬁrst described by Kane in 1848. Thus it is high time to provide a comprehensive overview of the diﬀerent developments in aldol chemistry, especially those of the past few decades. Demands for this important method of CaCabond formation came and continue to come from every ﬁeld of synthetic chemistry, particularly from natural product synthesis. Here, challenging problems in regioselectivity, chemoselectivity, diastereoselectivity and enantioselectivity frequently arise, many of which are still awaiting a solution. Symptomatically the word ‘‘selectivity’’ in its various connotations occurs no fewer than 1,100 times in both volumes, i.e. an average of twice a page. This book examines the enormous variety of aldol chemistry from the view of both organic as well as inorganic and bioorganic chemistry. It presents a wide range of potent syntheses based on the discoveries from enolate chemistry or the catalysis of Lewis acids and Lewis bases, for instance. The important role of metal catalysis, organocatalysis and direct aldol addition is described, along with enzymatic methods. However, it was not our intention to simply list all existing publications about aldol chemistry. Instead, we wanted to point out fundamental and at the same time eﬃcient ways leading to deﬁned conﬁgured aldol products. Two of these are depicted on the cover: the metal catalysis and the enzymatic method. It is now my pleasure to express my profound gratitude to the 22 authors and co-authors, all belonging to the elite of aldol chemistry, for their outstanding contributions and their professional cooperation. Special thanks are due to Wiley-VCH, especially Elke Maase and Rainer Mu¨nz, for their ﬁne work in turning the manuscript into the ﬁnished book. Finally, I am indebted to my wife and my son for countless hours of assistance. Last but not least, this book is also a tribute to the works of Teruaki Mukaiyama, who has done tremendous work in the ﬁeld of aldol reaction and now celebrates his 77th birthday. Berlin, Germany May 2004

Rainer Mahrwald

xix

List of Contributors Editor PD Dr. Rainer Mahrwald Institut fu¨r Organische und Bioorganische Chemie der Humboldt-Universita¨t zu Berlin Brook-Taylor-Str. 2 12489 Berlin Germany Authors Prof. Dr. Carlos F. Barbas, III The Skaggs Institute for Chemical Biology and the Department of Molecular Biology The Scripps Research Institute 10550 North Torrey Pines Road La Jolla, CA 92037 USA Prof. Dr. Manfred Braun Institut fu¨r Organische Chemie und Makromolekulare Chemie I Heinrich-Heine-Universita¨t Du¨sseldorf Universita¨tsstr. 1 40225 Du¨sseldorf Germany Prof. Dr. Scott E. Denmark 245 Roger Adams Laboratory, Box 18 Department of Chemistry University of Illinois 600 S. Mathews Avenue Urbana, IL 61801 USA Prof. Dr. Wolf-Dieter Fessner TU Darmstadt Department of Organic Chemistry and Biochemistry Petersenstr. 22 64287 Darmstadt Germany

Shinji Fujimori 236 Roger Adams Laboratory, Box 91-5 Department of Chemistry University of Illinois 600 S. Mathews Avenue Urbana, IL 61801 USA

Prof. Dr. Arun K. Ghosh Department of Chemistry University of Illinois at Chicago 845 West Taylor Street Chicago, IL 60607 USA

Prof. Dr. Kazuaki Ishihara Graduate School of Engineering Nagoya University Chikusa Nagoya, 464-8603 Japan

Prof. Dr. Jeﬀrey S. Johnson Department of Chemistry University of North Carolina at Chapel Hill Chapel Hill, NC 27599-3290 USA

Prof. Dr. Shu¯ Kobayashi Graduate School of Pharmaceutical Sciences The University of Tokyo Hongo, Bunkyo-ku Tokyo 113-0033 Japan

xx

List of Contributors Naoya Kumagai Graduate School of Pharmaceutical Sciences The University of Tokyo Hongo 7-3-1, Bunkyo-ku Tokyo, 113-0033 Japan

Michael Shevlin Department of Chemistry University of Illinois at Chicago 845 West Taylor Street Chicago, IL 60607 USA

Prof. Dr. Benjamin List Max-Planck-Institut fu¨r Kohlenforschung Kaiser-Wilhelm-Platz 1 45470 Mu¨lheim an der Ruhr Germany

Prof. Dr. Masakatsu Shibasaki Graduate School of Pharmaceutical Sciences The University of Tokyo Hongo 7-3-1, Bunkyo-ku Tokyo, 113-0033 Japan

PD Dr. Rainer Mahrwald Institut fu¨r Organische und Bioorganische Chemie der Humboldt-Universita¨t zu Berlin Brook-Taylor-Str. 2 12489 Berlin Germany Prof. Dr. Shigeki Matsunaga Graduate School of Pharmaceutical Sciences The University of Tokyo Hongo 7-3-1, Bunkyo-ku Tokyo, 113-0033 Japan Dr. Jun-ichi Matsuo The Kitasato Institute Center for Basic Research (TCI) 6-15-5 Toshima Kita-ku, Tokyo 114-003 Japan Prof. Dr. Teruaki Mukaiyama The Kitasato Institute Center for Basic Research (TCI) 6-15-5 Toshima Kita-ku, Tokyo 114-003 Japan David A. Nicewicz Department of Chemistry University of North Carolina at Chapel Hill Chapel Hill, NC 27599-3290 USA Prof. Dr. Dieter Schinzer Otto-von-Guericke-Universita¨t Magdeburg Chemisches Institut Universita¨tsplatz 2 39106 Magdeburg Germany

Prof. Dr. Isamu Shiina Department of Applied Chemistry Faculty of Science Tokyo University of Science Kagurazaka, Shinjuku-ku Tokyo 162-8601 Japan Prof. Dr. Fujie Tanaka Department of Molecular Biology The Scripps Research Institute 10550 North Torrey Pines Road La Jolla, CA 92037 USA Prof. Dr. Hisashi Yamamoto Department of Chemistry The University of Chicago 5735 S. Ellis Avenue Chicago, IL 60637 USA Dr. Yasuhiro Yamashita Graduate School of Pharmaceutical Sciences The University of Tokyo Hongo, Bunkyo-ku Tokyo 113-0033 Japan Prof. Dr. Akira Yanagisawa Department of Chemistry Faculty of Science Chiba University Inage, Chiba 263-8522 Japan

1

1

Fundamentals and Transition-state Models. Aldol Additions of Group 1 and 2 Enolates* Manfred Braun 1.1

Introduction

In an aldol reaction, an enolizable carbonyl compound reacts with another carbonyl compound that is either an aldehyde or a ketone. The enolizable carbonyl compound, which must have at least one acidic proton in its aposition, acts as a nucleophile, whereas the carbonyl active component has electrophilic reactivity. In its classical meaning the aldol reaction is restricted to aldehydes and ketones and can occur between identical or nonidentical carbonyl compounds. The term ‘‘aldol reaction’’, in a more advanced sense, is applied to any enolizable carbonyl compounds, for example carboxylic esters, amides, and carboxylates, that add to aldehydes or ketones. The primary products are always b-hydroxycarbonyl compounds, which can undergo an elimination of water to form a,b-unsaturated carbonyl compounds. The reaction that ends with the b-hydroxycarbonyl compound is usually termed ‘‘aldol addition’’ whereas the reaction that includes the elimination process is denoted ‘‘aldol condensation’’. The ‘‘traditional’’ aldol reaction [1] proceeds under thermodynamic control, as a reversible reaction, mediated either by acids or bases. In contrast, modern aldol methods rely on the irreversible formation of ‘‘preformed enolates’’ which are added to aldehydes or ketones. In any case, the aldol reaction has proven itself by a plethora of applications to be one of the most reliable methods for carbon–carbon bond-formation yielding either carbon chains, with oxygen functionality in 1,3-positions, or alkenes, by a carbonyl oleﬁnation process [2, 3]. The ﬁrst example of this reaction, the acid-catalyzed self-condensation of acetone to give mesityl oxide, was reported more than one and a half centuries ago by Kane [4]. The condensation of an aromatic aldehyde with an aliphatic aldehyde or ketone, obviously the ﬁrst example of an aldol condensation under basic conditions, was reported by Schmidt [5] and by *

This chapter is dedicated to the memory of Ulrike Mahler (deceased 1995) and Ralf Devant (deceased 2002).

Modern Aldol Reactions. Vol. 1: Enolates, Organocatalysis, Biocatalysis and Natural Product Synthesis. Edited by Rainer Mahrwald Copyright 8 2004 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 3-527-30714-1

2

1 Fundamentals and Transition-state Models. Aldol Additions of Group 1 and 2 Enolates

Claisen and Clapare`de [6], and named after the inventors (‘‘Claisen– Schmidt condensation’’) [2]. Obviously, Wurtz ﬁrst recognized [7] the simultaneous presence of aldehyde and alcohol moieties in the ‘‘aldol’’ 1 resulting from the acid-induced reaction of acetaldehyde, and the reaction was named after the product it leads to later on (Eq. (1)). O H

H 3C

OH

HCl, H2O

O

ð1Þ

H

H 3C 1

1.2

The Acid or Base-mediated ‘‘Traditional’’ Aldol Reaction

Several reaction conditions feature in this ‘‘traditional’’ transformation. First, the reaction is run in protic solvents and can be mediated either by acid or by base. Second, the reaction is reversible, particularly under these conditions. Finally, the enol or the enolate, which acts as a nucleophile, is inevitably generated in the presence of the aldehyde or ketone that functions as an electrophile. The aldol reaction performed under these conditions (Scheme 1.1) was ‘‘the state of the art’’ until the early nineteenseventies, when the chemistry of ‘‘preformed enolates’’ emerged. The ‘‘traditional’’ aldol reaction has been the subject of several reviews, among which the summary by Nielsen and Houlihan in ‘‘Organic Reactions’’ in 1968 is a classical contribution and a very valuable survey [1]. The subject has also been treated in House’s monograph [2] and, more recently, by Heathcock [3]. Thus, only general features and few representative examples of ‘‘traditional’’ aldol addition will be given here; the reader is referred to the above-mentioned surveys for more details. Synthetically the reversibility of the aldol addition can cause substantial problems. Investigations performed in order to determine the relative energies of an enolate and an aldehyde on the one hand and the aldolate on the other revealed the outcome of the aldol reaction to be slightly exergonic [8]. The aldol formed by either acid- or base-catalyzed reaction is signiﬁcantly stabilized by a strong OH bond in the aldol 2 which arises either directly from acid-mediated addition or on protonation of the aldolate 3 in the base-catalyzed variant, as shown in Scheme 1.1. Alternatively, chelation of the counter-ion in aldolates resulting from preformed enolates in non-protic media serves as the driving force [9]. As a general rule, applicable for protic solvents, the equilibrium in an aldol addition is located on the product side when aldehydes react with each other (Eq. (2)), but on the side of the starting materials for ketones (Eq. (3)). OH

O 2

R

O

R H

H R

ð2Þ

1.2 The Acid or Base-mediated ‘‘Traditional’’ Aldol Reaction

acid catalysis H

O H

O +

R

O

R

R

H

H

O

H

H R

H (cat)

2

O

H2O

R

OH

H OH H2O +

R

O

O

O H

R

O

R

H

H

R 3

basic catalysis Scheme 1.1

Aldol addition: general acidic and basic catalysis.

OH O 2

R

O

R

R

R R

ð3Þ

R

As a consequence, self-addition of enolizable aldehydes is usually readily accomplished in aqueous basic media if suﬃcient solubility is not prevented by the extended length of the carbon chain. The aldol addition (Eq. (4)) and condensation (Eq. (5)) of butanal giving 2-ethyl-3-hydroxyhexanal [10] and 2ethylhexenal [11], respectively, can serve as illustrative examples. The selfaddition of enolizable ketones, on the other hand, does not, per se, lead to substantial amounts of the aldol product, because of the unfavorable equilibrium mentioned above. This equilibrium can, however, often be shifted by a subsequent elimination step, so the aldol condensation of ketones is more frequently applied than addition. There are also special procedures that enable shifting of the equilibrium in the aldol reaction of ketones towards the products. Among these the formation of ‘‘diacetone’’ 4 by heating acetone under reﬂux in a Soxhlet apparatus ﬁlled with calcium or barium hydroxide is a well-known procedure which avoids contact of the base with the non-volatile hydroxy ketone 4 thus preventing a retro aldol reaction from occurring (Eq. (6)) [12, 13].

3

4

1 Fundamentals and Transition-state Models. Aldol Additions of Group 1 and 2 Enolates

OH O

O

KOH, H2O 75%

H

86%

H

Ba(OH)2 Soxhlet, reflux

O

ð4Þ

H

ð5Þ

O

NaOH, H2O 80°C

O

H

OH

O

ð6Þ

75% 4

Mixed aldol reactions between diﬀerent aldehydes or ketones are usually plagued by formation of a mixture of products, because each component can function as a CH-acidic and carbonyl-active compound. Whereas the ‘‘directed aldol reaction’’ [14–16] is a rather general solution to this problem, the traditional aldol addition of non-identical carbonyl compounds is only successful when applied within the framework of a limited substitution pattern. Thus, a fruitful combination in mixed aldol reactions is that of an aldehyde with an enolizable ketone. Obviously, the aldehyde, having higher carbonyl reactivity, reacts as the electrophilic component, whereas the ketone, with comparatively lower carbonyl reactivity, serves as the CH-acidic counterpart. Because the self-aldolization of ketones is endothermic, this type of side reaction does not occur to a signiﬁcant extent, so the product of the mixed aldol condensation is obtained in fair yield, as illustrated by the formation of ketone 6 from citral 5 and acetone, a key step in the synthesis of b-ionone (Eq. (7)) [17]. O

O + H

5

O

NaOEt 45-49% 6

ð7Þ The most eﬃcient variant of this combination is based on reaction of an enolizable ketone with a non-enolizable aldehyde, so that self-condensation of the latter cannot occur. Several examples of this type of combination in aldol reactions are given in Scheme 1.2. Usually in situ elimination occurs, so a,b-unsaturated ketones result, in particular when aromatic aldehydes are condensed with ketones (‘‘Claisen–Schmidt reaction’’) [18–21]. The intramolecular aldol condensation of dialdehydes, ketoaldehydes, and

1.2 The Acid or Base-mediated ‘‘Traditional’’ Aldol Reaction

NaOH, H2O, EtOH 20–25°C

O

O +

2 Ph

H3C

H

H O

+

+ Ph

H

ref. [18] Ph

Ph

CH3 O

ref. [19]

O

NaOH, H2O, EtOH 15–30°C

O

O

60–66%

CH3

H3C

O

1. NaOH, H2O 2. H

O

O

Ph

90-94%

CH3

5

85%

CH3

O ref. [20] Ph

Ph

OCH3 OCH3

CH3

H3C

CH3

H + O

NaOH, H2O, EtOH 25°C

CH3 CH3 ref. [21]

88% O O

Scheme 1.2

Illustrative examples of the condensation of aromatic aldehydes with enolizable ketones (‘‘Claisen–Schmidt reaction’’).

diketones is one of the most eﬃcient means of synthesizing ﬁve, six, and seven-membered rings. There are numerous applications of this variant of the aldol reaction, in particular in the context of the Robinson annelation reaction, described in the literature for decades. Because this topic has been reviewed comprehensively [22, 23], a few illustrative examples only will be given here. An early and rather prominent example of the intramolecular aldol condensation is found in Woodward’s synthesis of cholesterol [24]. Because the precursor 7 is an unsymmetrical dialdehyde, the problem of regioselectivity arises. Nevertheless, the a,b-unsaturated aldehyde 8 is formed in excess and only minor amounts of the regioisomer 9 are obtained (Eq. (8)). Complete regioselectivity was observed, however, when dialdehyde 10 was submitted to an intramolecular aldol condensation. Thus, the enal 11 was obtained exclusively in the ﬁrst synthesis of genipin described by Bu¨chi and coworkers (Eq. (9)) [25].

1 Fundamentals and Transition-state Models. Aldol Additions of Group 1 and 2 Enolates

6

H

H

piperidine, AcOH, benzene 60°C

CHO CHO

66%

H O 7

CHO H

H

H

H

+ H

H

CHO

O

O 8

9

ð8Þ MeO2C

MeO2C

H

O OH

CHO

piperidine, AcOH

CHO

68%

H

ð9Þ

O

H

OH

10

H

CHO

11

When ketoaldehydes, compounds that are unsymmetrical per se, are submitted to intramolecular aldol condensation the ketone usually acts as the CH-acidic component whereas the aldehyde plays the role of the carbonyl active counterpart. This regiochemical outcome is also favored when the conditions of a thermodynamic control are used. Again, this type of aldol condensation has been used in a variety of natural products synthesis. A steroid synthesis, the aldolization step of which is given in Eq. (10), is an illustrative example [26]. O O H H

O

CHO

KOH, H2O 50–80°C

H

73%

H

O O

O

ð10Þ When this type of stereochemical outcome is prevented by steric hindrance, the two carbonyl groups can play opposite roles in the sense that the aldehyde, deprotonated in its a-position, functions as nucleophile whereas

1.2 The Acid or Base-mediated ‘‘Traditional’’ Aldol Reaction

the ketone acts as the carbonyl-active compound, as shown in the example given in Eq. (11) [27]. OCH2Ph

OCH2Ph O

Na2CO3, EtOH, H2O

ð11Þ

46% H

O

CHO

Although problems of regiochemistry are inherent, the aldol condensation of diketones has found wide application. Typical examples are syntheses of cyclopentenones and cyclohexenones from 1,4- and 1,5-diketones, respectively. The concept is illustrated by a synthesis of jasmone 12 (Eq. (12)) [28] and of the homosteroid derivative 13, the latter arising under thermodynamic control in a Robinson annelation reaction (Eq. (13)) [29]. O

O

ð12Þ

O 12 OMe MeO

+ O O OMe

OMe

O

O

O 13

ð13Þ Intramolecular aldol condensations also serve as the key step in ‘‘biomimetic’’ syntheses of polyketides, synthetic strategies that try to imitate in vitro a proven or an assumed biosynthetic pathway [30]. Although the ﬁrst attempts in this direction go back to the early 20th century [31], practical and eﬃcient syntheses based on this concept were elaborated much later.

7

8

1 Fundamentals and Transition-state Models. Aldol Additions of Group 1 and 2 Enolates

O O

OEt O

O

+

O

O

O O

O

O

2

O

O OEt

O

O

14

15

O

OH

16

OH

O

O O

O

O

OH

39%

O

O

OH 17 OH

OH

O

94%

OH

1. HCl 2. HI, HOAc

KOH, MeOH

O

OH

78%

O

OH

OH

HO 18

OH

O

OH

70% HO O 19 Scheme 1.3

Biomimetic synthesis of emocline 19 involving intramolecular aldol reactions.

The concept is illustrated in Scheme 1.3 – when intermediate 16 containing a carbon chain with six free keto groups and one protected keto group has been generated by twofold Claisen condensation of the diester 14 with the highly reactive dianion 15, it undergoes spontaneous aldol condensation followed by aromatization to give the naphthalene derivative 17. A further aldol addition, which leads to the formation of a third six-membered ring, needs treatment with potassium hydroxide. Finally, dehydration and deprotection lead to the anthrone 18, which is readily oxidized to the natural product emodine 19 [32]. Similar approaches based on intramolecular aldol reactions have been applied to the synthesis of naturally occurring anthracyclinones [33] and isoquinolines [34]; the biosynthesis of these is known to involve polyketone intermediates. The problem of stereochemistry has very rarely been addressed by tra-

1.3 The Aldol Addition of Preformed Enolates – Stereoselectivity and Transition-state Models

ditional aldol addition. The question of relative stereochemistry has been studied occasionally in the context of intramolecular, in particular transannular, aldolizations. There were, however, few diastereoselective variants. Control of enantioselectivity was achieved to a remarkable extent in the cyclization of 1,3-cyclopentanediones. Independent work by research groups at Hoﬀmann LaRoche and Schering AG in the early nineteen seventies revealed that highly enantioselective cyclization of triketones 20a, b can be accomplished by treatment with catalytic amounts of l-proline, as shown in Eq. (14) [35, 36]. Thus, the products 21a and 21b are obtained from the Hajos–Parrish–Eder–Sauer–Wiechert reaction, the intramolecular aldol condensation, in 93 and 99% ee, respectively. The method has been applied successfully to enantioselective steroid syntheses, and provided a route to the skeleton of several other natural products [37]. The source of stereoselectivity in this cyclization has been investigated carefully [38]. Very recently, proline catalysis has also been applied to enantioselective intermolecular aldol additions [39]. This promising approach will be discussed in detail in Chapter 4 of Part I. O N H

R

O

R

CO2H , DMF

O O

20a: R = Me 20b: R = Et

O

OH

21a: 93.4% ee 21b: 99.5% ee

ð14Þ Although the number of applications of the ‘‘traditional’’ aldol reaction ‘‘is legion’’, and despite its undoubted versatility, the reaction suﬀers from general lack of control of stereochemistry and from the diﬃculty of reliable determination of the carbonyl-active and CH-acidic components. Both problems have been solved by the technique of directed aldol addition based on preformed enolates.

1.3

The Aldol Addition of Preformed Enolates – Stereoselectivity and Transitionstate Models

The chemistry of preformed enolates emerged in temporal and causal coherence with the ‘‘LDA area’’. Although lithium and magnesium salts of diisopropylamine were ﬁrst developed in the nineteen-ﬁfties [40], lithium diisopropylamide (LDA) has been a widely used reagent since 1970, because of its behavior as a soluble, strong, and non-nucleophilic base [14]. LDA and related bases, for example lithium hexamethyldisilazane (LIHMDS) [41], lithium N-isopropylcyclohexylamide (LICA) [42], and lithium 2,2,6,6-

9

10

1 Fundamentals and Transition-state Models. Aldol Additions of Group 1 and 2 Enolates

tetramethylpiperidide (LITMP) [43] turned out to be the reagents of choice for conversion of a variety of carbonyl compounds into their enolates in an irreversible reaction which also enabled control of regiochemistry. This is illustrated in the kinetically controlled deprotonation of 2-methylcyclohexanone, 22, which leads to the formation of the enolate 23 with remarkable regioselectivity (Eq. (15)) [44]. Complementary routes that lead to the formation of the regioisomeric enolate 24 with a more substituted double bond have also been elaborated; they are based on a deprotonation under thermodynamic control or use of enol acetates, silyl enol ethers, or a,bunsaturated carbonyl compounds as precursors. In addition, procedures for formation of (E) and (Z) enolates were elaborated. The formation of preformed enolates has been reviewed comprehensively [45, 46]. In addition, the determination of enolate structures by crystal structure analyses, pioneered by the research groups of Seebach, Boche, and Williard, and NMR spectroscopic investigations and theoretical calculations led to insight into their reactivity [47]. O

22

LiN(iPr2), MeOCH2CH2OMe, -78°C

OLi

OLi +

23 99

24 :

1

ð15Þ Preformed enolates can be obtained not only from aldehydes and ketones, but also from carboxylic esters, amides, and the acids themselves. The corresponding carbonyl compound always acts irreversibly as the CH-acidic component. Thus, the term aldol reaction is no longer restricted to aldehydes and ketones but extended to all additions of preformed enolates to an aldehyde or a ketone. In contrast with the ‘‘traditional’’ aldol reaction, this novel approach is based on a three-step procedure (usually, however, performed as a one-pot reaction). First, the metal enolate 25 is generated irreversibly, with proton sources excluded, and, second, the compound serving as the carbonyl active, electrophilic component is added. The metal aldolate 26 thus formed is ﬁnally protonated, usually by addition of water or dilute acidic solutions, to give the aldol 27 (Scheme 1.4) [45, 46]. The principal aim in the development of the ‘‘modern’’ aldol reaction was stereochemical control, a ﬁeld that has been treated in a series of review articles [46, 48–60]. In stereochemical terminology, the topic is discussed in the terms of ‘‘simple diastereoselectivity’’ and ‘‘induced stereoselectivity’’ [61]. Except for relatively rare examples when R1 is identical with R2 and R3 is identical with R 4 , all aldol additions are stereogenic. If the carbonyl-active compound is either an aldehyde (except formaldehyde – R 4 ¼ H) or a prochiral ketone (R 3 0 R 4 ), addition of the enolate leads to formation of either

1.3 The Aldol Addition of Preformed Enolates – Stereoselectivity and Transition-state Models

O R

1

R2

R1

a

α

O

M

C C

X

R2

H

X 25

O

M

O

O

R4

R3 b

R4

R3

X R1

R2

26

OH H3O c

R4

R3

O X

R2 R 27

1

R1, R2: H, alkyl, aryl, OR, NR2 R3, R4: H, alkyl X: H, alkyl, aryl, OR, NR2, OM Scheme 1.4

Aldol reaction of preformed enolates. (a) irreversible enolate formation; (b) addition of the preformed enolate to aldehydes or ketones; (c) protonation.

one or two stereogenic centers. This depends on whether an enolate with identical a-substituents (mostly R1 ¼ R2 ¼ H) or an enolate with diﬀerent asubstituents is used. Under the latter conditions one of the substituents R1 and R2 is usually a hydrogen atom (Scheme 1.4). A general stereochemical pattern of the aldol addition is shown in Scheme 1.5. When a carbonyl compound 28 with an a-substituent R2 (which can be an alkyl or an aryl group, or a hetero substituent, for example alkoxy, or a protected amino group, but which is not identical with hydrogen) is converted into the ‘‘preformed’’ enolate 29 and added to an aldehyde, four stereoisomeric products 30a, 30b, 31a, 31b can result. When neither the enolate 29 nor the aldehyde contains stereogenic units, both reactants have enantiotopic faces and 30a and 30b are enantiomers. The same is true for the pair 31a and 31b. However, 30 and 31 form a pair of diastereomers. When an aldol addition leads to an excess of one of these diastereomers 30 or 31, it is said to exhibit simple diastereoselectivity. Several notations that assign descriptors to diastereomeric aldols are found in the literature. The classical erythro/threo nomenclature, which is based on Fischer projection formulas [62], will not be used in this chapter, because it can cause considerable confusion with branched carbon chains. Among the

11

12

1 Fundamentals and Transition-state Models. Aldol Additions of Group 1 and 2 Enolates

OH

O

OH R3

R1

R3

R1

R2 30b OH

R2 31a

O

OH R3

R1

O R3

R1

R2

R2

30a

31b R1CHO

R2 =/ H

O R2

O

CH2

C

O R2

CH

M

C R3

R3

28

29

R1CHO

OH

O

OH R3

R1

R2 = H

32a

O R3

R1 32b

Scheme 1.5

Stereochemical pattern of aldol addition.

diﬀerent alternatives proposed in the literature, the Prelog–Seebach notation deserves mention: their l (like) and u (unlike) descriptors are systematic and unambiguous, because they correlate strictly with Cahn–Ingold–Prelog nomenclature [63]. For practical reasons, however, the syn/anti notation, introduced by Masamune [64], will be used in this chapter. Thus, the carbon chain that contains the two stereogenic centers is drawn in a zigzag fashion. In the syn diastereomers both substituents at the stereocenters are directed either toward or away from the viewer. In the anti isomers one of the substituents is directed toward the viewer, the other one away from the viewer, or vice versa. According to this notation the stereoisomers 30a and 30b are termed syn whereas 31a and 31b are deﬁned as the anti isomers. When the topicity of an aldol addition is of interest the lk (like) and ul (unlike) notation [63] will be used in this chapter to describe the pathway by which the enolate approaches the carbonyl compound (aldehyde or ketone). The problem of simple diastereoselectivity does not arise in aldol additions when an a-unsubstituted enolate 29 (R2 ¼ H) or an enolate with two identical a-substituents reacts with an aldehyde or a prochiral ketone. The products 32a and 32b obtained from this combination are enantiomers, if neither the aldehyde nor the enolate is a chiral molecule.

1.3 The Aldol Addition of Preformed Enolates – Stereoselectivity and Transition-state Models

If an aldol addition is performed using either an enolate with stereogenic units, which can be located in the a-substituent R2 or in the ipso substituent R3 , or if a chiral aldehyde is used as the electrophilic component, the aldol products 30a, 31a, and 32a are diastereomers of 30b, 31b, and 32b. In these combinations not only is simple diasteroselectivity has to be achieved (except for the a-unsubstituted case leading to 32a/b) but also induced stereoselectivity. In detail, induced stereoselectivity is postulated for the following combinations of reactants: 1. reaction of chiral enolates with achiral or chiral aldehydes; 2. reaction of achiral enolates with chiral aldehydes; and 3. reaction of achiral enolates with achiral aldehydes, if they are mediated by use of a chiral catalyst (including enzymes or antibodies), chiral ligands at the metal M, or all kinds of chiral solvents. When, in reactions 1–3, the enolate does not have an a-substituent, e.g. 29 (R2 ¼ H), induced stereoselectivity is highly desirable. In addition, the corresponding reaction of a-substituted enolates 29 (R2 0 H) should not only provide induced stereoselectivity but also simple diastereoselectivity. In this chapter we will not use ‘‘induced diastereoselectivity’’ as a subdivision applicable to reactions 1 and 2 and ‘‘induced enantioselectivity’’ for reaction 3 (leading to enantiomeric aldol products), because, irrespective of the stereochemical relationship of the products (diastereomers or enantiomers), the transition states leading to the diﬀerent stereoisomeric compounds are always diastereotopic, even if the products are enantiomers (reaction 3). The term ‘‘induced stereoselectivity’’, which includes the diﬀerent variants 1–3, is therefore used here [57, 61]. Most aldol additions of preformed enolates are run under kinetic control. In some such kinetically controlled aldol reactions simple diastereoselectivity is related to the conﬁguration of the enolate. The seminal investigations of Dubois [65], then intensive studies by the research groups of OM R2

R3

1. R1CHO 2. H3O

OH

R3

R1 R2

H (Z )-enolate

OM H

R3 R

O

syn aldol

1. R1CHO 2. H3O

OH

R3

R1 R2

2

(E )-enolate

O

anti aldol

Scheme 1.6

Correlation between enolate geometry and aldol conﬁguration.

13

14

1 Fundamentals and Transition-state Models. Aldol Additions of Group 1 and 2 Enolates

House, Heathcock, and Ireland [48–51], revealed that (Z)-conﬁgured enolates furnish mainly syn-aldols whereas anti-b-hydrocarbonyl compounds arise predominantly from (E) enolates. In this context, the descriptors E and Z refer to the relative position of the a-substituent R2 and the oxygen–metal bond (Scheme 1.6). Because procedures for the selective generation of (Z) and (E) enolates have been elaborated for a variety of carbonyl compounds [45], the kinetically controlled aldol addition oﬀers a solution to the problem of simple diastereoselectivity. Representative examples of the Z/syn and E/anti correlation are given in Eqs. (16) and (17) [66, 67].

OLi Me

OH

O

Ph

R

H (Z )-enolate

OH

1. PhCHO THF, -78°C

R

2. H3O

+

O

Ph

R Me

Me syn

anti

syn

R

: anti

98.7 : 1.3

CMe3 Et

90

: 10

ð16Þ OLi H

O

Ph

X Me

OH

1. PhCHO THF, -78°C

OH X

2. H3O

+

Ph

X

Me

Me

syn

(E )-enolate

O

anti

syn : anti

X OMe

62

:

38

OCMe3 Me

51

:

49

12

:

88

2

:

98

O Me Me3C O

Me

Me3C

ð17Þ

1.3 The Aldol Addition of Preformed Enolates – Stereoselectivity and Transition-state Models

As shown by Eq. (16), (Z) lithium ketone enolates yield syn aldols, if the substituent R at the carbonyl group is sterically demanding. Because carboxylic amides and thioamides, like ketones, form (Z) enolates predominantly, their aldol addition also leads to the predominant formation of syn-b-hydroxycarboxylic acids (or their corresponding derivatives). In general, Z-conﬁgured boron and titanium enolates result in higher simple diastereoselectivity in favor of syn aldols than the corresponding lithium or magnesium enolates. Also, with regard to the induced stereoselectivity boron enolates are usually more selective than lithium enolates [49, 52]. This might be because the boron–oxygen bond in enolates is shorter than the lithium–oxygen bond, so cyclic transition states involving boron as the metal are tighter, and steric repulsion is more eﬀective and chiral information is transferred more eﬃciently (for chiral auxiliary groups or ligands). On the other hand, carboxylic esters and thioesters, which form predominantly (E) enolates, react with aldehydes with substantial anti selectivity. As shown in Eq. (18), this selectivity is, however, restricted to reactions in which bulky aromatic substituents form the alcoholic moiety of the ester. The anti diastereoselectivity of (E) enolates is usually lower then the syn selectivity of comparable (Z) enolates. One must also take into account the eﬀect of the asubstituent. As shown in Eq. (18) the Z–syn correlation can be completely reversed for a-substituents that are bulky, sterically demanding alkyl groups [68]. OLi R

CMe3 H

1. Me3CCHO Et2O, 20°C 2. H3O

OH

O

OH

O

CMe3 + Me3C

Me3C

CMe3

R

R

syn

anti R

syn : anti 100 : 100 :

0

Et n-Pr

98 :

2

CHMe2

29 :

71

Me

CMe3

0

0

: 100

ð18Þ In the Mukaiyama addition of the aldol reaction [16], silyl ketene acetals or silyl enol ethers are added to aldehydes in a reaction mediated by Lewis acids or ﬂuoride. Here again the Z–syn correlation is sometimes not observed [69, 70]. Thus, the Z–syn, E–anti correlation seems to be a rule with several exceptions [71]. The stereochemical outcome of the diﬀerent aldol additions of preformed enolates calls for plausible transition state models. Two kinds of explanation seem suitable for rationalizing the diﬀerent stereochemical results, which

15

16

1 Fundamentals and Transition-state Models. Aldol Additions of Group 1 and 2 Enolates

even seem to contradict each other: Thus, in a single type of transition state model a diﬀerent substitution pattern might cause the reactants to have diﬀerent orientations to each other so that formation of diﬀerent stereoisomers results. On the other hand, diﬀerent reaction conditions and reactants can be used, even if the way the aldol reaction is run is restricted to the use of preformed enolates. Thus, the latter might have very diﬀerent counter-ions, and the solvents might also be diﬀerent. It seems plausible that this can cause the reaction to occur via completely diﬀerent types of transition state. The most widely accepted transition state hypothesis for aldol additions is the Zimmerman–Traxler model. This was originally developed to explain the stereochemical outcome of the Ivanoﬀ reaction – addition of the dianion of carboxylic acids with magnesium counter-ions to aldehydes and ketones [72]. On the basis of investigation of the stereochemical outcome of the reaction of doubly deprotonated phenyl acetic acid to benzaldehyde (Eq. (19)), Zimmerman and Traxler proposed in a seminal paper a transition state model that involves a six-membered chair-like assembly of the reactants [73]. i-PrMgBr, Et2O

O Ph

OMgBr Ph

OH

OMgBr H

ð19Þ OH

OH

O

1. PhCHO Ph 2. H3O

OH Ph 69%

+

Ph

O OH

Ph 22%

This model oﬀers a plausible explanation of the (Z)–syn, (E)–anti correlation, as shown in Scheme 1.7. The diasteromeric transition states 33a and 33b, which emerge from addition of a (Z) enolate to an aldehyde, diﬀer in the position of the substituent R1 , which is equatorial in 33a and axial in 33b. By analogy with conformational analysis of the cyclohexane system [62], the transition state 33a is expected to have a lower energy than the diastereomeric alternative 33b. As a consequence the predominant formation of syn aldolates results from this kinetically controlled reaction. When the (E) enolate is chosen as the starting material, the analogous argument indicates the transition state 34a with R1 in an equatorial position to be favored compared with the alternative 34b, in which the substituent R1 occupies an axial position. Accordingly, the anti aldolate is expected to be the predominant product. Although ﬁrst developed for a magnesium enolate, the Zimmerman– Traxler model could be used very successfully to explain the stereochemical

1.3 The Aldol Addition of Preformed Enolates – Stereoselectivity and Transition-state Models

R2 O R1

O

M

R1

OM R2

R1

H

OH O

H H

H

O

17

33a

R3

R1

H

R2

R3 syn aldolate

R3

O

syn aldol

R2 OM

H

R3 (Z )-enolaye H

R2

O

O

M

O

H

R1

R1 R1

H

R3

33b

O

M

R1 R

H H

OH O

H

O

R2 anti aldol

OH O

H

2

R3

R2

O R3

R1 R2

R3

anti aldol

anti aldolate

34a

R3

R3

OM H

R1

O

R1

anti aldolate

H O R1

OM R2

H

H OM

R2 R3

(E )-enolate

H H

O

O R1

M

O

H R1

R2 34b

OM H

H R1

R3

OH O

R2

R3 syn aldolate

Scheme 1.7

Zimmerman–Traxler transition state models in the aldol additions of (Z ) and (E ) enolates.

outcome of aldol additions of boron and titanium enolates. This might be because they are monomeric, in contrast with the enolates of lithium and magnesium, known to form aggregates. Not only is (Z)–syn [49, 52] and (E)–anti [74, 75] correlation better for boron enolates, they also usually result in greater induced stereoselectivity. This also is easily explained by the Zimmerman–Traxler model if it is assumed that for boron and titanium enolates, stronger Lewis acids than lithium and magnesium, the sixmembered transition state is tighter, so steric eﬀects are maximized.

O R3

R1 R2 syn aldol

18

1 Fundamentals and Transition-state Models. Aldol Additions of Group 1 and 2 Enolates

The validity of the Zimmerman–Traxler model for alkali metal or magnesium enolates could be questioned, because these strongly electropositive metals might form ionic rather than covalent bonds to the enolate/aldolate oxygen atom. Even if there was a contact ion-pair of metal cation and oxygen anion, however, the geometry of a six-membered chair in the transition state would be very similar to that shown in Scheme 1.7. Even the aggregation of lithium enolates, well recognized today [47], does not severely contradict the Zimmerman–Traxler model – indeed, a six-membered transition state, postulated to occur at a tetrameric lithium pinacolone enolate [76], is very compatible with the ‘‘closed’’ model proposed by Zimmerman and Traxler. The Zimmerman–Traxler model has, however, been challenged by the frequent observation that (Z) enolates result in higher simple diastereoselectivity, giving syn aldols, compared with the lower anti selectivity of (E) enolates [49–51, 68]. Assuming the classical chair transition state, it has been remarked that for (E) enolates models 34a and 34b might both be plagued by unfavorable steric repulsion. Thus, the equatorial orientation of R1 in 34a avoids repulsion by R3 , but at the expense of a steric hindrance between R1 and R2 , which is enhanced because the torsional angle at the forming carbon–carbon bond is less than 60 . This has been postulated in skewed transition state models, proposed by Dubois [68] and Heathcock [51], in which the dihedral angle between the enolate double bond and the carbonyl group approaches 90 , as shown in Scheme 1.8 for the E–syn and E–anti correlation. The alternative, 34b, takes advantage of minimizing the R1 –R2 repulsion, but on the other hand, is disfavored by the axial R1 –R3 hindrance. As a result, both transition states 34a and 34b become similar in energy, so stereoselectivity is reduced. An important modiﬁcation of the classical Zimmerman–Traxler model, which still relies on the idea of a ‘‘pericyclic-like’’ transition state, considers H R

O

M

1

O

R2 H

OH

R2

R3

O

H

anti

M O

R2

R3

R1

(E )-enolate H

O

R1

R3 (E )-enolate Scheme 1.8

Skewed transition state model for (E) enolates.

OH

O R3

R1 R2 syn

1.3 The Aldol Addition of Preformed Enolates – Stereoselectivity and Transition-state Models

R3 H

OH

O

O M

H

O R2

R3

R1 R2

R1

(Z )-enolate

syn

R3 R1

OH

O

O M

H

O R2

H

(Z )-enolate

R3

R1 R2 anti

Scheme 1.9

Boat transition state models for (Z ) enolates.

boat conformations as alternatives. Thus, Evans [49, 77] has suggested boat transition states when there is substantial steric hindrance between groups R1 and R2 on the forming carbon–carbon bond. The model plausibly explains the results given in Eq. (18), which show that an increase of the asubstituent R2 in a (Z) enolate leads to a higher proportion of the anti aldol. The alternative orientations outlined in Scheme 1.9 for the Z–syn and Z– anti correlation show the latter to be a reasonable alternative for large substituents R2 . Hoﬀmann, Cremer and co-workers have proposed a transition state model for addition of enol borates to aldehydes [78]. The authors pointed out that a twist-boat 36b could easily be formed from the U-conformation of the boron enolate (E)-35b whereas the intermediate with the W-orientation, (Z)-35a, is a suitable precursor of the chair-type transition state 36a. The consequence of the assumption that (Z) enolates react via a chair transition state whereas (E) enolates react via a boat conformation is that syn aldols are formed irrespective of enolate geometry, as shown in Scheme 1.10 [79]. This type of stereochemical outcome has been observed in various examples of the Mukaiyama-type aldol addition [80] and in aldolizations of enol stannanes and zirconium enolates [81, 82]. In contrast, a clear (Z)–syn, (E)–anti correlation has been observed in the addition of allylboronates to aldehydes [83], a reaction that is closely related to the aldol addition. The stereochemical hypothesis shown in Scheme 1.10 has been underscored by the semi-empirical calculations of Gennari and coworkers [84] which show that, starting from (Z) enolates, the half-chair transition state leading to syn aldols is preferred. For (E) enolates, a preference for either syn or anti aldols is predicted, depending on whether or not the metal carries a bulky substituent. The diminished simple diastereoselectivity in aldol additions of (E) enolates can be interpreted as a result of competition between the chair

19

20

1 Fundamentals and Transition-state Models. Aldol Additions of Group 1 and 2 Enolates

R3

R3 O 1

M

H

O R

H

R

M

O H

O

2

R1

35a

O R3

R1 R

M

R1

H H

R

36a

OH

O

2

2

H H R2

O R3

R1

M O O

R2

35b

R3 36b

Scheme 1.10

Diﬀerent transition state models (chair and twist-boat) for (Z ) and (E ) enolates (M ¼ BL2 ).

and twist-boat transition states [85]. More recently, six-membered transition state hypotheses have also been proposed for ‘‘direct’’ aldol additions [86] and for the phosphoramide-catalyzed addition of trichlorosilyl enolates to aldehydes [87]. A diﬀerent cyclic transition state model which does, however, not incorporate the metal, has been proposed by Mulzer and coworkers. It was developed to explain the observation that in the addition of doubly deprotonated phenyl acetic acid to pivaldehyde the highest anti selectivity is obtained with the most ‘‘naked’’ enolate anions (e.g. K/18-crown-6). The hypothesis, which might explain this stereochemical result, assumes that the approach of the enolate to the aldehyde is dominated by the interaction of the enolate HOMO and the p orbital of the aldehyde that functions as the LUMO. The favored approach of the reactants occurs when the substituents of the enolate (phenyl) and the aldehyde (t-butyl) are oriented in a trans orientation at the forming carbon bond, so that their mutual steric repulsion is minimized (Scheme 1.11). The expected transition state 37 has some similarity to that of a 1,3-dipolar cycloaddition, although the corresponding cycloadduct 38 does not form, because of the weakness of the oxygen–oxygen bond. Instead, the doubly metalated aldol adduct 39 results [88]. In a similar

1.3 The Aldol Addition of Preformed Enolates – Stereoselectivity and Transition-state Models

H Me3C

OH OM

CHO + Ph

O OH + Me3C

Me3C

OM

OH

Ph syn M

anti

:

syn

58 70 97

: : :

42 30 3

R1 O

H

O R1

H

LUMO

O

R2

H O M

H HOMO R2

R1 H

O M

M 38

R2

OM

OM OM

H

37

O OH

Ph anti

½ Mg Li K/18-crown-6

21

O 39

R1 = C(CH3)3; R2 = C6H5 Scheme 1.11

Anti selectivity in the aldol addition of doubly deprotonated phenylacetic acid and 1,3-dipolar cycloaddition transition state model.

way, Anh and Thanh emphasized that frontier orbital interactions played an essential role in determining the stereochemical outcome of the aldol reaction [89]. A completely diﬀerent rationale for the stereochemical outcome of aldol additions relies on open-transition-state models. These involve antiperiplanar orientation of enolate and carbonyl group, in contrast with their syn-clinal conformation assumed in the six-membered cyclic transition states. Open-transition-state structures have been proposed to oﬀer a rationale for those aldol additions that give predominantly syn products, irrespective of enolate geometry [90]. This outcome has been observed in aldol reactions of tin and zirconium enolates and of ‘‘naked’’ enolates generated from enolsilanes by treatment with tris(diethylamino)sulfonium diﬂuoromethylsiliconate [70]. As shown in Scheme 1.12, the driving force for the

22

1 Fundamentals and Transition-state Models. Aldol Additions of Group 1 and 2 Enolates

O + R1

R2

OSiMe3 CH

(Et2N)3S

C

Me3SiF2

R3

H

Me3SiO

O

Me3SiO R3

R1

+

O R3

R1

R2

R2

syn

anti

enolate (Z ) : (E ) 99 : 9 :

1 91

product ratio syn : anti 95 94

: :

5 6

R1 = R3 = Ph; R2 = Me

δ− O R1

O R2

R3 H

O

H

δ−

O R2

H 41a (Z )

O δ−

δ−

δ− O R1 R1

R2 syn favored

O

R2 40b (E )

OH

R3

H

R3 R1

R2 anti disfavored

δ− O H

R1

R1

R3

40a (Z )

R3

δ− O H

OH

H

R3 H

R2

O δ−

41b (E )

Scheme 1.12

Formation of syn aldols irrespective of enolate geometry. Open-transition-state models.

open-transition-state model is the tendency of the negatively charged oxygen atoms to be as far apart from each other as possible. It is assumed that both transition states that lead to the formation of the anti aldol, i.e. 40a formed from the (Z) enolate and 40b from the (E) enolate, are disfavored, because of the steric repulsion of substituents R1 and R2 , which are oriented in a gauche conformation. This type of steric hindrance is avoided in the transition state structures 41a and 41b, so both (Z) and (E) enolates give the syn aldol predominantly. The basic assumption of open-chain transition-state models is the antiperiplanar orientation of the enolate and the carbonyl double bond. This

1.3 The Aldol Addition of Preformed Enolates – Stereoselectivity and Transition-state Models

H

C

H

H

O

Li +

O C C

H

H

C

H

Li O

O

C C

H

H

H

H 40

=/

Li O

O 106.9°

C H

H

O

C

H

41

O

C H

C H

Li

H

C C

H

H

H H 42

Scheme 1.13

Calculated reaction pathway for addition of acetaldehyde lithium enolate to formaldehyde.

type of model has also been used to explain the stereochemical outcome observed in aldol additions of thioester silylketene acetals [84]. An open transition has also been proposed in additions of silyl ketene acetals to aldehydes, mediated by chiral copper complexes [91]. Even if a particular enolate with a distinct geometry is reacted with an aldehyde, the question whether the transition state is ‘‘closed’’ or ‘‘open’’ cannot be answered by simple ‘‘either–or’’. More recent discussions have, instead, led to an ‘‘as well as’’, because the role of the counter-ion becomes more evident. Thus, ab-initio calculations of Houk and coworkers [92] predict an open-transition-state structure for metal-free, ‘‘naked’’ enolates and closed transition states for lithium enolates. For addition of acetaldehyde lithium enolate to formaldehyde, the lowest-energy reaction pathway (shown in Scheme 1.13) has been studied on the basis of on ab-initio (3–21 G) calculations [93]. The reactants ﬁrst reach the coordination complex 40, a local minimum on the energy hypersurface [94]. In this complex the OaLiaO angle can vary from 145 to 180 . The transition state of carbon–carbon bond formation is calculated to have the half-chair conformation 41. The angle of nucleophilic attack on the carbonyl group is 106.9 , consistent with the Bu¨rgi–Dunitz trajectory [95] and in accordance with calculations of Houk and coworkers [92]. The transition state structure 41 ﬁnally collapses to the aldolate 42 with the lithium atom coordinating the two oxygen atoms. The activation barrier of the reaction is calculated to be 1.9 kcal mol1 and the overall exothermicity is 40.2 kcal mol1 . Concerning the question of the conformation of the six-membered cyclic transition state, the diﬀerent possibilities, for example chair, half-chair, or

23

24

1 Fundamentals and Transition-state Models. Aldol Additions of Group 1 and 2 Enolates

twist boat, seem to be quite close in relative energy so the particular substitution pattern is assumed to determine the favored conformation. The key role of the counter-ion has been conﬁrmed by experimental results obtained from the intramolecular aldol addition of bicyclic keto aldehydes. Here again, enolates with a strongly coordinating metal counter-ion, for example Mg 2þ , have strong preference for a reaction via a closed transition state in which the metal counter-ion is coordinated both to the enolate and to the carbonyl oxygen atom. ‘‘Naked’’ enolates, on the other hand, have a pronounced tendency to react through an open transition structure with an anti-periplanar conformation of the enolate and the carbonyl moiety [96]. It is self-evident that the transition state hypotheses discussed above are exclusively relevant to kinetically controlled aldol additions. Although this type of reaction control is the rule when preformed enolates are used, one should be aware that the reversibility of aldol additions cannot be excluded a priori and in any instance. In aldol reactions of preformed enolates, reversibility becomes noticeable in equilibration of syn aldolates with anti aldolates rather than in an overall low yield as found in the traditional aldol reaction. Considering the chair conformations of the syn and the anti aldolates, the former seem to be thermodynamically less stable, because of the axial position of the a-substituent R2 . This situation is avoided in the anti adduct (Eq. (20)). Indeed, the anti diastereoisomer is favored in most aldol additions run under thermodynamic control. This has been observed, for example, in aldolates syn-43 and anti-43, which arise from addition of doubly lithiated phenylacetic acid to pivalaldehyde. Whereas the kinetically controlled reaction gives a syn/anti ratio of 1.9:1, equilibration occurring after several hours in tetrahydrofuran at 25 C leads to a 1:49 in favor of the anti products (Eq. (21)) [88]. R2

O

O

M

R1

R

OLi CO2Li Ph syn-43

O

M

ð20Þ

R2

3

H

syn

Me3C

H

R1

H H

O

R

3

anti Ph

OLi

H

OLi +

Me3C

OLi CO2Li

Me3C CHO

Ph anti-43

ð21Þ The enolate counter-ion has an important eﬀect on the rate of the reverse aldol reaction. Boron enolates usually undergo completely irreversible addition to aldehydes. The more ‘‘ionic’’ of the alkali metals, for example

1.4 Stereoselective Aldol Addition of Lithium, Magnesium and Sodium Enolates

sodium and potassium, have a greater tendency to undergo retro aldol reactions than lithium. Thus the potassium aldolate anti-44 formed from deprotonated ethyl mesityl ketone and benzaldehyde undergoes equilibration to syn-44 even at 78 C, whereas the corresponding lithium aldolate isomerizes at 0 C (Eq. (22)) [66]. That, in this reaction, the syn-aldolate is thermodynamically favored is possibly explained by the steric hindrance between the a substituent (methyl) and the bulky ipso substituent (mesityl) in the anti isomer. OM

O

Me

OM

Ph

O

Me

Ph Me Me syn-44

Me

Me Me

Me

anti-44

ð22Þ The inﬂuence of further counter-ions like ammonium, magnesium and zinc on the reversibility has been studied [65, 71]. Another inﬂuence comes from the stability of the enolate. As a rule, the rate of the retroaldol reaction correlates with the stability of the enolate. In stereoselective aldol addition, the reversibility is, in general, rather considered as a complication than a tool to obtain high selectivity. In particular, thermodynamically controlled aldol additions are usually not suitable to obtain non-racemic aldols.

1.4

Stereoselective Aldol Addition of Lithium, Magnesium and Sodium Enolates

Modern synthetic methods in organic chemistry are aimed at obtaining chiral products in a non-racemic, if at all possible, enantiomerically pure form. In addition, the products should be accessible as pure diastereomers. This section therefore focuses on aldol additions that provide pure stereoisomers. Amongst the enolate counter-ions of groups 1 and 2, only magnesium, sodium, and, particularly, lithium are important for this synthetic purpose. Although it should be remarked that most aldol additions leading to enantiomerically pure products rely on boron, tin, titanium, and zirconium enolates, topics that will be discussed in Chapters 1 and 3 of Part I of this book and Chapters 3–5 of Part II, there are also advantages of the more polar lithium and magnesium enolates. They are, in particular, highly reactive and can be added to aldehydes under mild conditions at low temperatures. Furthermore, their handling is easy, and they can be used on a large scale. As a consequence, a variety of useful aldol additions that rely on lithium and magnesium enolates have been developed and are applied fairly frequently.

25

26

1 Fundamentals and Transition-state Models. Aldol Additions of Group 1 and 2 Enolates

1.4.1

Addition of Chiral Enolates to Achiral Carbonyl Compounds a-Substituted Enolates A variety of carbohydrate-derived ketones have been converted into their corresponding lithium enolates and used as chiral nucleophiles in additions to aldehydes. Induced diastereoselectivity was, however, found to be moderate only. Chiral ketones, oxazolidinones, amides, and esters, on the other hand, performed amazingly well when used as boron, titanium, or tin enolates in aldol additions [49, 52, 55, 57]. The corresponding lithium enolates, however, resulted in substantially lower stereoselectivity. Interestingly, the lithium enolate generated from a-siloxy ketone 45 leads, stereoselectively, to the syn aldol 46a; the diastereomeric ratio (dr) exceeds 95:5, deﬁned as the ratio of the major isomer to the sum of all other isomers. In contrast, the corresponding boron enolate furnishes the diastereomeric syn aldol 46b, which results from the opposite induced stereoselectivity (Scheme 1.14). The diﬀerent behavior of the enolates is explained by a transition state model 48a with a chelated lithium counter-ion and a non-chelating boron atom in 48b. When ketone 45 is deprotonated by treatment with bromomagnesium tetramethylpiperidide, the anti diastereomer 47a is obtained in substantial excess relative to the minor product, the anti stereoisomer 47b. It turns out that the enolates with the diﬀerent counter-ions lithium, boron, and magnesium are, in a sense, complementary. The anti selectivity of the magnesium enolate is rationalized by assuming that it has the (E) conﬁguration (Scheme 1.14) [45, 97–100]. 1.4.1.1

Procedure: 6-Hydroxy-3-trimethylsilyloxy-4-alkanones 47a by Magnesiummediated Addition of (S)-5,5-Dimethyl-4-trimethylsiloxy-3-hexanone to Aldehydes [100]. An oven-dried 5-mL Wheaton vial is ﬂushed with nitrogen and 2,2,6,6-tetramethylpiperidine (0.26 mL, 1.5 mmol), dry THF (0.5 mL) and ethylmagnesium bromide (1.2 m in THF, 1.16 mL, 1.4 mmol) are added. The vial is capped securely and heated with stirring at 70 C for 24 h. The resulting solution is cooled to 0 C, and (S)-5,5-dimethyl-4-trimethylsiloxy-3hexanone (0.10 mL, 0.08 g, 0.50 mmol) in THF (0.25 mL) is added over 30– 60 min with a syringe pump. The solution is stirred for 1.5 h at 0 C after addition is complete. The enolate solution is then cooled to 78 C and the aldehyde (2 mmol) is added dropwise. After 30 min the reaction is quenched by pouring the mixture into satd aq. NaHCO3 (5 mL). The layers are separated, and the aqueous phase is extracted with diethyl ether (5 10 mL). The combined organic layers are washed with cold HCl (1%, 10 mL) and satd aq. NaHCO3 (10 mL). After drying, the solution is concentrated to yield the product as a clear oil. Ratios of diastereomers are determined by integration of the relevant peaks in the 1 H NMR spectra of the crude products. Puriﬁcation is accomplished by ﬂash chromatography (diethyl

1.4 Stereoselective Aldol Addition of Lithium, Magnesium and Sodium Enolates

1. LiN(i-Pr)2/THF -78°C 2. TMEDA 3. RCHO

OH

O

R

d.r. > 95 : 5 Me3SiO 46a 1. Bu2BOTf/(i-Pr)2NEt 2. RCHO 3. H2O2, OH

O

OH R

d.r. > 95 : 5 Me3SiO

O

46b

Me3SiO

1.

NMgBr

45

OH

O 2. RCHO

R

d.r. > 92 : 8 Me3SiO 47a NMgBr

1.

O

2. HMPA/ClTi(Oi-Pr)3 3. RCHO

OH R

d.r. > 95 : 5 Me3SiO 47b SiMe3

Me3C H

H

H O Li O

46a

R Me 48a

Me3SiO

H

O

CMe3 H H BBu2 O O

46b

R Me 48b

Scheme 1.14

Stereodivergent aldol addition of (S )-5,5-dimethyl-4-(trimethylsilyloxy)-3-hexanone (45).

27

1 Fundamentals and Transition-state Models. Aldol Additions of Group 1 and 2 Enolates

28

ether:hexanes, 5:95). The products 47a are thus obtained: R ¼ i-Pr, 70%; R ¼ t-Bu, 80%; R ¼ Ph, 80%. Because the titanium enolate of the ketone 45 aﬀords the stereoisomeric product 47b this completes a method of stereodivergent aldol addition. Starting from the identical chiral ketone 45 they lead to all of the diﬀerent stereoisomeric products in a controlled manner [45]. The chiral a-benzoyloxyketone 49, accessible from mandelic acid, also reacts stereoselectively with aldehydes to give the syn aldols 50a/b. Both the lithium [101] and the titanium [102] enolates lead to the predominant formation of the diastereomer 50a. However, the stereoselectivity obtained by use of the titanium enolate surpasses that of the lithium analog (Eq. (23)). O O

Ph

1. LiN(i-Pr)2 2. ClTi(Oi-Pr)3 3. RCHO

O

O

1. LiN(i-Pr)2/THF 2. RCHO or

O Ph

O

O +

R OH

49

O

Ph

R OH

O

50a

50b

Enolate counterion

R

Li Li Ti Ti

Et Me2CH Et Me2CH

ratio 50a : 50b 96 96 97 97.7

: : : :

4 4 3 2.3

ð23Þ Remarkably high stereoselectivity is obtained by means of the sodium enolate of a-N,N-dibenzylamino-substituted ketone 51, a counter-ion not very frequently used in stereoselective aldol additions. In this instance, however, the sodium enolate turned out to be more eﬃcient than the lithium analog. The predominant formation of the main diastereomeric product 52a rather than 52b is explained by an open transition state, assumed to be strongly favored over the cyclic transition state, when the more ‘‘ionic’’ sodium enolate is used rather than the corresponding lithium reagent (Eq. (24)) [103]. A large variety of propionic acid esters and higher homologs having a chiral alcohol moiety have been used in additions to aldehydes [56, 57]. It turned out, however, that the lithium enolates result in only moderate simple diastereoselectivity and induced stereoselectivity, in contrast with the corresponding boron, titanium, tin, or zirconium enolates and silyl ketene acetals, with which stereoselectivity is excellent. The same feature has been observed in enolates derived from chiral amides and oxazolidinones, as

1.4 Stereoselective Aldol Addition of Lithium, Magnesium and Sodium Enolates

1. LiN(i-Pr)2/THF or NaN(SiMe3)2

O R

O R Bn2N

51

OH

R

+

R'

2. R'CHO Bn2N

O

OH

R'

Bn2N 52a

52b

enolate counter ion

R

R'

Li Na

Me Me

Ph Ph

Li

CHMe2

Ph

63

: 11 : 6 : 37

Na

CHMe2

Ph

>95

: 5

ratio 52a : 52b 89 94

ð24Þ outlined in Chapter 2 and in Chapter 3 of Part I. The aldol additions of the chiral propanoate 53, generated from triphenylglycol, serve as an illustrative example. Whereas the lithium enolate 54 (M ¼ Li) gives the diastereomers 55a, b and 56a, b in a more or less stereo-random manner, acceptable simple diastereoselectivity in favor of the anti product combined with high induced stereoselectivity was obtained only after transmetalation of the lithium enolate into the zirconium species (Eq. (25)) [104, 105]. O Me

OM

Ph Me

OSiMe3

O Ph 53

Ph

Ph OSiMe3

O Ph

H

Ph

54

OH

OH

O

PhCHO Ph

OR*

+

OH OR*

+

ð25Þ

O

Ph

56a

OR* 56b

ratio enolate counterion anti-55 : syn-56 Li Cp2ZrCl

OR* 55b

O

Ph

O

Ph

55a OH

29

70 : 30 90 : 10

ratio 55a : 55b 68 : 32 >97 :

3

30

1 Fundamentals and Transition-state Models. Aldol Additions of Group 1 and 2 Enolates

The aldol reaction of chiral lactones, developed in the context of ‘‘selfreproduction of chirality’’ [106], takes advantage of the easy generation and high reactivity of lithium enolates. When the chiral cis dioxolanone 57, readily available from (S)-lactic acid and pivalaldehyde, is treated with LDA, the lactone enolate 58 is generated. Although the stereogenic center originating from lactic acid has vanished, because of enolate formation, the acetal carbon atom maintains the chiral information and the t-butyl residue directs the topicity in the addition to aldehydes and unsymmetrical ketones, so high stereoselectivity is achieved (Eq. (26)) [107]. H CMe3 LiN(i-Pr) 2 THF; -78°C O O H

O

57

Ph

O

O

H CMe3

O

H CMe3

Me

Ph

O

O O

OLi

HO

58

ð26Þ By analogy, a series of heterocyclic compounds 59 have been deprotonated to give the corresponding a-hetero-substituted cyclic enolates 60. On addition to aldehydes they are found to react with both high simple diastereoselectivity and induced stereoselectivity (Eq. (27)) [106–108]. H CMe3 X

Y

R

O

H CMe3 X

Y OLi

R

59

ð27Þ

60

X = O, NR', S; Y = O, NR'; R = H, Alkyl

The concept has also been extended to a six-membered homolog, the dioxanone 61, which furnishes the enolate 62 by deprotonation. Here again, subsequent aldol addition proceeds with high stereoselectivity and the diastereomer 63a results predominantly or almost exclusively (Eq. (28)) [109, 110]. With all heterocyclic enolates 58, 60, and 62 stereocontrol is relatively easily accomplished, because of the rigid structure of the heterocycle. Procedure: Aldol Addition of 2-tert-Butyl-6-methyl-one-1,3-dioxan-4-one to Propanal [110]. An ice-cold solution of (i-Pr)2 NH (11.13 mL, 79.4 mmol, 1.14 equiv.) in THF (160 mL) is treated with a solution of n-butyllithium in hexane (1.4 m, 53 mL, 79.4 mmol, 1.14 equiv.), kept at 0 C for 15 min, then cooled to 78 C. To this solution of LDA is added the dioxanone 61 (12.0 g, 69.7 mmol) in THF (80 mL) at such a rate that the temperature never ex-

1.4 Stereoselective Aldol Addition of Lithium, Magnesium and Sodium Enolates

CMe3 O

CMe3

LiN(i-Pr)2 THF; -78°C

O

RCHO

O

O

O

OLi

61

62

O

H

O

O

O

H

+

O

O R

ð28Þ

CMe3

CMe3

R

OH

OH 63b

63a

ceeds 70 C; the mixture is then maintained at 78 C for 45 min. To the resulting enolate solution are added (7.15 mL, 99 mmol, 1.42 equiv.) propanal in THF (80 mL), the temperature never being allowed to rise above 70 C. The reaction mixture is stirred at 78 C for 3 h then quenched at 78 C by the addition of satd aq. NH4 Cl (200 mL) then diethyl ether (200 mL). The two phases are separated, and the aqueous phase is extracted with diethyl ether (2 200 mL). The combined organic extracts are dried (MgSO4 ) and the volatile compounds removed by rotary evaporation and then with a high-vacuum pump. The crude product obtained, a 7:1 ratio of epimers at C(1 0 ), is (1 0 S,2R,5R,6R)-2-t-butyl-5-(1-hydroxypropyl)-6-methyl1,3-dioxan-4-one 63a (R ¼ Et), yield 9.6 g (60%). Eﬃcient stereochemical control is also provided by the chiral lithium ketone enolate 64, addition of which to a variety of aldehydes leads to the formation of the corresponding b-hydroxy ketones 65, usually as single products (Eq. (29)) [111]. O

Ph

O

H

N

H

LiN(i-Pr)2 THF Ph

Tos

O

LiO H

N

H

Tos 64 HO O

RCHO

Ph

O

H

N

H

R H

Tos 65 R = CHMe2, (CH2)3Me, CH=C(Me)-CO2Et

ð29Þ

31

32

1 Fundamentals and Transition-state Models. Aldol Additions of Group 1 and 2 Enolates

The deprotonation of chiral iron acyl complexes, which can be obtained as enantiomerically pure compounds, leads to the corresponding enolates, as shown by the research groups of Davies and Liebeskind [112–115]. The lithium enolate 67a, however, which originates from propanoate 66a, reacts stereoselectively with aldehydes or ketones only if it has been transmetalated into the corresponding copper or aluminum enolate (Eq. (30)) [116]. PPh3 OC

n-BuLi THF, -78°C

Fe

R

PPh3 OC

Fe

Me

R = Me OLi

O 66a: R = Me 66b: R = OBn

67a

n-BuLi R = OBn THF, -78°C

ð30Þ

PPh3 OC

Fe

OBn O

67b

PPh3

O

Li

OC

OBn

Fe

40% O

OH

68

Stereoselective aldol addition to the lithium enolates themselves has been achieved by reaction of the deprotonated benzyloxy-substituted iron complex 66b and subsequent reaction with symmetrical ketones. The enolate involved in this procedure is assumed to exist as a chelated species 67b. The aldol 68 is obtained in a diastereomeric ratio higher than 99:1. The reaction is, nevertheless, plagued by low chemical yield, because of deprotonation of acetone. Because a symmetric ketone is used as an electrophile, the reaction leads to the formation of just one new stereogenic center [117, 118]. a-Unsubstituted Enolates When, instead of an a-substituted enolate 29 (R2 0 H), the a-unsubstituted enolate 29 (R2 ¼ H) is used in an aldol addition, the stereochemistry is, at ﬁrst glance, simpliﬁed, as outlined in Scheme 1.5. Formation of the aldol products 32a and 32b shows that now only one new stereogenic center is formed. Nevertheless, aldol addition of a-unsubstituted chiral enolates (or achiral enolates, mediated by chiral ligands or additives) has been a problem, because of insuﬃcient induced stereoselectivity [53]. In fact, using the same chiral auxiliary group R3 transition from an a-substituted to an aunsubstituted enolate is often accompanied by complete loss of the ability to discriminate between the enantiotopic faces of an aldehyde, so that more or less equal amounts of the stereoisomers 32a and 32b result (cf. Scheme 1.4.1.2

1.4 Stereoselective Aldol Addition of Lithium, Magnesium and Sodium Enolates

1.5). The problem occurs in aldol additions of methyl ketones and all kind of acetic acid derivative, esters, amides, and thioamides [53, 54]. In a pioneering investigation, the lithium enolate derived from 3-methyl2-pentanone was added to aldehydes. Only moderate diastereoselectivity was obtained, however [119]. Exceptionally high induced stereoselectivity was observed when camphor-derived ketone 69 was converted into the lithium enolathe and subsequently added to aldehydes. a-Cleavage at the carbonyl group enabled the formation of b-hydroxy aldehydes and acids in high enantiomeric excess (Eq. (31)) [60]. The work on the aldol addition of methyl ketones led to a variety of stereoselective variants which rely mainly on boron enolates and will be discussed in Chapter 3 of Part I of this book. On the other hand, the transition from lithium to boron enolates did not signiﬁcantly improve the induced stereoselectivity of chiral acetamides [53]. Several highly stereoselective procedures based on tin and titanium enolates have been developed, however.

(31) Many attempts have been made to add chiral acetates to aldehydes or prochiral ketones, to obtain non-racemic b-hydroxycarboxylic esters. Here again, several variants based on boron and titanium enolates and on Mukaiyama aldol additions of silyl ketene acetals have been developed, and will be described in Chapter 2 (titanium enolates), Chapter 3 (boron enolates) and in Part II (Mukaiyama reaction). For enolates of group 1 and 2 elements the following fruitful approaches were elaborated. First, sulﬁnyl acetates 70 [120, 121], carrying their chiral information in the sulfoxide moiety were eﬃciently deprotonated with t-butylmagnesium bromide and added to aldehydes to give the aldol adducts 71. Removal of the sulﬁnyl residue is accomplished – in an immolative manner – by reduction with aluminum amalgam to furnish t-butyl b-hydroxycarboxylic esters 72. Remarkably, the method developed by Solladie´ and Mioskowski is not restricted to aldehydes as electrophilic components, but has been extended to prochiral ketones also (Scheme 1.15) [122]. The transition-state model 73, in which the magnesium atom is chelated by the enolate, sulfoxide, and carbonyl oxygen atoms, serves to explain the stereochemical outcome of the reaction. It is plausible that the aldehyde approaches the enolate from the side of the non-bonding electron pair of the sulfoxide (opposite to the aryl residue) and that the larger group R occupies a position anti to the sulﬁnyl substituent.

33

34

1 Fundamentals and Transition-state Models. Aldol Additions of Group 1 and 2 Enolates

1. t-BuMgBr THF, -78°C 2. O O Tol

O

S

OH R1 R2

R2

R1

O

OCMe3

OCMe3 S

O

Tol 70

71 OH

Al/Hg

O

R1

OCMe3

R2 72

Starting Material R1

R2

H

CH3

t-Butyl 3-Hydroxy Ester 72 e.e. (%)

Yield (%)

C7H15

86

80

C CC3H7

80

73

C CC6H13

70

53

C6H5

91

85

C7H15

95

88

C CC3H7

48

60

C CC6H13

36

72

8

80

C2H4OCOCH3

40

90

C6H5

68

75

CF3

20

75

COOC2H5

C6H5

L L Mg O O

70 Tol

S

O

RCHO Tol

S

L L Mg

O 71

O OCMe3

OCMe3 R

H 73 Scheme 1.15

Aldol addition of (R)-tolylsulﬁnylacetate 70 to aldehydes and prochiral ketones. Transition state model 73 (reaction of the magnesium enolate of 70 with RCHO).

1.4 Stereoselective Aldol Addition of Lithium, Magnesium and Sodium Enolates

Procedure: 3-Hydroxy carboxylic esters by addition of tert-butyl (B)-(R)-2-(4methylphenylsulﬁnyl)acetate to carbonyl compounds [123]. A solution of tBuMgBr (40 mL; prepared from 3 g Mg, 20 g t-BuBr, and 50 mL Et2 O) is added to a solution of sulﬁnyl ester 70 (1.5 g, 5.9 mmol) in THF (400 mL), at 78 C, over a period of 20 min, under argon. The mixture is then stirred for 30 min and the carbonyl compound (2 g) in THF (30 mL) is added. After 12 h at 78 C the mixture is hydrolyzed by addition of satd aq. NH4 Cl (50 mL) and extracted with CHCl3 (2 50 mL). The extract is dried with Na2 SO4 and concentrated. The residue, the b-hydroxy-a-sulﬁnyl ester, is diluted with THF (400 mL) and water (40 mL) and then treated with aluminum amalgam (4 5-g portions) while maintaining the temperature at 15–20 C. The solvent is evaporated and the residue, the b-hydroxy ester 72, is puriﬁed by column chromatography (silica gel; Et2 O–petroleum ether 20:80). The Solladie´ procedure has been successfully applied in a synthesis of maytansin [124]. In an analogous manner, chiral sulﬁnyl acetamides 70 (NMe2 instead of OCMe3 ) can be added to aldehydes and deliver b-hydroxy amides in high enantiomeric excess, again after reductive removal of the chiral auxiliary group [125]. More recently, the lithium enolate derived from the acetate 74 with axial chirality has been found to react with aldehydes in a highly stereoselective manner to give b-hydroxy esters 75 that can be converted into the corresponding carboxylic acids by alkaline hydrolysis (Eq. (32)) [126].

(32) Chiral acetyl iron complexes 76 also seem predestined to serve as reagents that enable introduction of a chiral acetate unit into aldehydes. In contrast with the benzyloxy-substituted derivative 66b (Eq. (30)), however, only marginal induced stereoselectivity is achieved when the lithium enolate of 76 is added to aldehydes, and the diastereomer 77a is formed in low preference compared with 77b (Eq. (33)). High diastereoselectivity is obtained only after transmetalation [112, 114, 115]. A signiﬁcant improvement – as far as chiral lithium enolates of acetyl iron complexes are concerned – came from the complex 78, which carries a (pentaﬂuorophenyl)diphenylphosphane ligand instead of the usual triphenylphosphane. Thus, the enolate 79, generated by treatment with LDA, gives the diastereomeric adducts 80a and 80b in a diastereomeric ratio of 98.5:1.5 on treatment with benzaldehyde. A donor–acceptor interaction between the enolate oxygen atom and the ﬂuorinated aromatic ring, supported

35

36

1 Fundamentals and Transition-state Models. Aldol Additions of Group 1 and 2 Enolates

PPh3 OC

1. n-BuLi THF, -78°C

Fe

2. PhCHO O 76

ð33Þ PPh3 OC

PPh3

Fe

Ph O

+

OC

Fe

OH

O

77a d.r.

Ph OH

77b :

57

43

by spectroscopic studies, in the boat like transition state model 81 is assumed to be responsible for the observed lk-topicity (i.e. the (S)-enolate 79 approaches the aldehyde from its Si face) of the reaction (Scheme 1.16) [127]. Despite the elegance of the concept of Davies–Liebeskind enolates, one should be aware that preparation of the acyl iron complexes deﬁnitely needs resolution. It is somewhat typical that a series of procedures has been elaborated by using the racemic iron complexes [115]. It is highly desirable that chiral a-unsubstituted enolates should be available by simple methods from enantiomerically pure starting materials that are inexpensive and readily accessible in both enantiomeric forms. This postulate seems to be fulﬁlled to a reasonable extent by (R)- and (S)-2hydroxy-1,1,2-triphenylethyl acetate 83 (‘‘HYTRA’’) [53, 128, 129]. It is readily prepared from methyl mandelate which is ﬁrst converted into triphenylglycol 82 and subsequently converted into the acetic ester 83 by treatment with acetyl chloride (Eq. (34)). Both enantiomers of the reagent are readily accessible, because both (R)- and (S)-mandelic acid are industrial products [130]. Diol 82 and acetate 83 are commercially available. PhMgBr Et2O, reflux

Ph HO

OMe

Ph OH

HO

77%

Ph

O (R)-methyl mandelate

Ph

82

ð34Þ

O Cl pyridine 92%

O

Ph O

OH

Ph Ph (R)-83

1.4 Stereoselective Aldol Addition of Lithium, Magnesium and Sodium Enolates

P(C6F5)(C6H5)2 OC

P(C6F5)(C6H5)2

LDA (2 equiv) THF; -42°C

Fe

OC

Fe

CH2

O

OLi

78

79 P(C6F5)(C6H5)2 OC

Fe

P(C6F5)(C6H5)2

C6H5 O

C6H5CHO THF; -78°C

+

Fe

OC

OH

C6H5 O

80a

OH

80b

F

F

F F

Ph Ph OC

F

P

O Fe Li O

H C6H5

81 Scheme 1.16

Diastereoselective aldol addition of lithiated (pentaﬂuorophenyl)diphenyl-substituted acetyl iron complexes 78. Transition state model 81.

Double deprotonation of the chiral acetate (R)-83 by treatment with 2 equiv. LDA enables generation of the enolate 84. Remarkably, the dilithiated reagent 84 dissolves in THF whereas the ester 83 is fairly insoluble. Bridging of the oxygen anions by the lithium cations in the enolate 84 might explain the enhanced solubility. When the lithium enolate is added to aldehydes at 78 C, the diastereomeric aldol adducts 85a are formed predominantly, the ratio of diastereomers 85a:85b ranging between 10:1 and 12:1. Enhancement of the induced stereoselectivity can be accomplished by transmetalation to the corresponding magnesium enolate and by performing the addition to the aldehyde at lower temperatures (110 C to 135 C, using 2-methylbutane as co-solvent). Thus diastereomeric ratios up to 50:1 can be achieved. A single recrystallization of the crude product mixture

37

38

1 Fundamentals and Transition-state Models. Aldol Additions of Group 1 and 2 Enolates

O

2 LiN(i-Pr)2 THF; -78°C to 0°C

Ph OH

O Ph 83

Ph

O 84

Ph

Ph

OH

O

Ph

1. MX 2. RCHO

LiO OLi

O

R

OH OH +

Ph

R

O

Ph OH

O Ph

Ph

85a

Ph

Ph

85b

Ph OH

− HO

NaOH

Ph Ph 82 OH

O

R

OH 86

Scheme 1.17

Stereoselective aldol additions of the chiral acetate 83.

usually gives the major diastereomer 85a in the pure form. Mild alkaline hydrolysis of the aldol adducts 85 furnishes the b-hydroxycarboxylic acids 86 and triphenylglycol 82, which is easily separated and can be reused (Scheme 1.17) [131]. When the lithium enolate 84 is added to propenal (without transmetalation), the diastereomeric esters 87a and 87b are formed in the ratio 92:8. In this reaction the crude mixture 87a/87b was hydrolyzed to give the carboxylic acid (R)-88 in 83.5% ee. To obtain the enantiomerically pure 3-hydroxy4-pentenoic acid, enrichment was performed by single recrystallization of the ammonium salt, formed from (S)-1-phenylethylamine. When the amine has been liberated from the salt the carboxylic acid (R)-88 is obtained in >99.8% ee and 41% overall yield (Scheme 1.18) [132]. The (S) enantiomer, but not the (R) enantiomer, of 3-hydroxy-4-pentenoic acid 88 (both prepared according to this procedure) has been shown to be a substrate for the enzyme 3-hydroxybutanoate dehydrogenase – another example of the diﬀerent biological activity of enantiomeric compounds [133]. Procedure: (R)-3-Hydroxy-4-pentenoic acid (88) by aldol addition of doubly deprotonated (R)-HYTRA (83) [132]. A 250-mL, two-necked, round-bottomed ﬂask is equipped with a magnetic stirrer, a septum, and a connection to a combined vacuum and nitrogen line. The air in the ﬂask is replaced by nitrogen and dry THF (100 mL) and diisopropylamine (37.7 mL, 0.264 mol)

1.4 Stereoselective Aldol Addition of Lithium, Magnesium and Sodium Enolates

1. 2 LiN(i-Pr)2 O 2. O

Ph

Ph 83 O

OH

H

OH

O

Ph

Ph

OH OH

O Ph

+

O

Ph OH

O Ph

Ph

Ph

87b

87a KOH MeOH, H2O

OH

O

Ph

OH OH

− HO

88 83.5% e.e.

Ph Ph 82 1.

NH2

Ph Me 2. recrystallization 3. NaOH

OH

O OH

88 99.8% e.e. Scheme 1.18

Synthesis of (R)-3-hydroxy-4-pentenoic acid 88 by stereoselective aldol addition of the chiral acetate (R)-83.

are injected via syringes via the septum. The mixture is cooled to 78 C and treated, while stirring, with a solution of n-butyllithium (15%, 168 mL, 0.269 mol) in hexane. The dry ice–acetone bath is replaced with an ice bath and stirring is continued for 30 min. A 2-L, three-necked, round-bottomed ﬂask equipped with a mechanical stirrer, a septum, and a connection to a combined vacuum and nitrogen line is charged with (R)-83 (40.0 g, 0.120 mol). The air in the ﬂask is replaced by nitrogen and dry THF (400 mL) is added through a cannula of 2 mm i.d., during which the ﬂask is slightly evacuated. The suspension is stirred at 78 C in a dry ice–acetone bath. The ice-cold solution of LDA is added via a cannula, with vigorous stirring, during which the 2-L ﬂask is slightly evacuated. The mixture is stirred at 0 C for 30 min to complete double de-

39

40

1 Fundamentals and Transition-state Models. Aldol Additions of Group 1 and 2 Enolates

protonation. A clear, orange solution forms. This is subsequently cooled to below 70 C (dry ice/acetone bath). The septum is removed cautiously (overpressure) and dry 2-methylbutane (900 mL) is poured into the ﬂask (alternatively, a low-boiling (30–37 C) fraction of petroleum ether can be used) during which a vigorous stream of nitrogen is maintained. The ﬂask is immediately closed with a septum and a thermocouple, connected to a resistance thermometer, is introduced via the septum. The reaction ﬂask is plunged into a liquid-nitrogen bath, the depth of immersion being 2–3 cm. When the temperature of the suspension has reached 125 C, a solution of propenal (acrolein; 19.2 mL, 0.285 mol) in dry THF (30 mL) is added dropwise via a syringe through the septum at such a rate that the temperature does not exceed 120 C. Stirring is continued for 30 min at 120 to –125 C, during which time the yellowish color turns to pale blue. The mixture is treated with satd aq NH4 Cl (250 mL) and left to warm to r.t. The organic solvents are removed in vacuo. The precipitate in the aqueous suspension is separated by suction ﬁltration, washed with several portions of water (total 400 mL), and transferred to a 4-L, round-bottomed ﬂask equipped with a magnetic stirrer and a condenser. Methanol (2.4 L), water (1 L), and potassium hydroxide (72 g) are added, and the mixture is heated under reﬂux for 3 h. After cooling to r.t. the organic solvent is removed in vacuo. The residual aqueous alkaline suspension is shaken with 200-mL portions of CH2 Cl2 . The aqueous solution is transferred to a 2-L, roundbottomed ﬂask, immersed in an ice bath, and acidiﬁed to pH 3 by cautious addition of hydrochloric acid (6 mol L1 ); the mixture is stirred vigorously with a magnetic stirrer and the pH is controlled carefully in order to avoid over-acidiﬁcation. The clear solution is saturated with NaCl and extracted with ethyl acetate (8 200 mL), during which the pH of the aqueous layer is monitored and, if necessary, readjusted to pH 3 by the addition of hydrochloric acid (6 mol L1 ). The combined organic layers are dried with MgSO4 , concentrated, and the oily residue is distilled under reduced pressure in a short-path distillation apparatus to aﬀord the colorless acid 88; yield 8.32 g (59.6%); b.p. 69 C/0.04 Torr (5.3 Pa); [a]D 17 21.7 (c ¼ 1.95% aq. ethanol); 83.5% ee. A solution of the acid (8.32 g, 0.072 mol; 83.5% ee) in dry diethyl ether (300 mL) at r.t. is placed in a 500-mL, round-bottomed ﬂask equipped with a magnetic stirrer. ()-(S)-1-Phenylethylamine (20 mL, 0.157 mol) is added in one portion with vigorous stirring. A white precipitate forms immediately and the ﬂask is closed with a drying tube ﬁlled with calcium chloride. After stirring for 30 min at 25 C the precipitate is separated by suction ﬁltration and washed with ice-cold diethyl ether (2 50-mL). Recrystallization from dry THF (240 mL) aﬀords 13.80 g of a colorless salt, m.p. 127 C; [a]D 20 9.7 (c ¼ 1.998, deionized water). The salt is treated with aq. sodium hydroxide (2%, 400 mL) and the mixture is washed with CHCl3 (3 70 mL). The acidiﬁed aqueous solution is extracted with ethyl acetate (as described above) and the combined organic extract is dried with MgSO4 and con-

1.4 Stereoselective Aldol Addition of Lithium, Magnesium and Sodium Enolates

centrated. To remove remaining traces of solvent the ﬂask containing the product is connected via a short, curved glass tube to a two-necked, liquidnitrogen-cooled ﬂask, which is connected to an eﬃcient oil pump; yield 5.77 g (69.4%, 41.4% relative to (R)-2-hydroxy-1,2,2-triphenylethyl acetate); [a]D 17 26 (c ¼ 0.996, 95% aqueous ethanol); >99.8% ee (determined by 1 H NMR measurement of the methyl ester in the presence of Eu(hfc)3 , no signals of the (S) enantiomer are detected). Since the ﬁrst report of stereoselective aldol additions of the chiral acetate 83, the reagent has been applied frequently in syntheses of natural products and biologically active compounds. Among these are g-amino-b-hydroxybutanoic acid (‘‘GABOB’’) [134], the enantiomeric naphthoquinones shikonin and alkannin [135], d- and l-digitoxose [136], desoxy and aminodesoxy furanosides [132], detoxinine [137], tetrahydrolipstatin and related pancreatic lipase inhibitors [138], statin [139] and statin analogs [140], compactin and mevinolin [141], ﬂuorooleﬁn peptide isoesters [142], the HMG-CoA synthase inhibitor F-(244) [143], epothilone A [144], the A-ring building block of 1a,25-dihydroxyvitamin D3 [145], intermediates for (23S)-hydroxyvitamin D3 derivatives [146], building blocks of lankacidin C [147], the synthetic statin NK-104 [148], the C1aC9 segment of bryostatin [149], the C20aC34 segment of the immunosuppressant FK-506 [150], and pyranoyl steroids having hypocholesterolemic properties [151]. Selected examples are shown in Scheme 1.19. Furthermore, a large variety of synthetic inhibitors of HMG-CoA reductase have been synthesized by including the HYTRA aldol procedure as one of the key steps [152–158]. The enzyme is responsible for reduction of hydroxymethylglutaryl CoA to mevalonic acid, a key step in cholesterol biosynthesis. HMG CoA reductase inhibitors contain a b,ddihydroxycarboxylic ester or carboxylate moiety or the corresponding dlactone. Several of these compounds, which can be regarded as synthetic analogs of compactin and mevinolin, have found their way to the marketplace, because of their hypocholesterolemic activity. Selected examples of the structures of these drugs are given in Scheme 1.20. As in the previous scheme, the stereogenic center generated by aldol addition of the chiral acetate 83 is marked with an asterisk. The (R)-conﬁgured reagent 83 always attacks the aldehyde predominantly from the Re side, the (S) acetate 83 correspondingly from the Si side. Thus, there is a predictable lk topicity in HYTRA aldol additions (Scheme 1.21). It has been reported that the induced stereoselectivity in aldol additions of the chiral lactate 83 can be improved by using an excess of base (LDA or lithium hexamethyldisilazane) in the deprotonation step [159]. 1.4.2

Addition of Achiral Enolates to Chiral Carbonyl Compounds

In a chiral aldehyde or ketone the two carbonyl faces are diastereotopic and the products resulting from either Re or Si face attack of the enolate are

41

42

1 Fundamentals and Transition-state Models. Aldol Additions of Group 1 and 2 Enolates

HO

O

HO

O

OH H3N

COO

*

* HO

O OH (R)-Shikonin

(R)-GABOB

OH

H N

* HO

O

O

* H

O OH (S)-Alkannin

O

O

OH HCO NH C11H23

OH

O *

*

H Tetrahydrolipstatin

Detoxinine

O

O NH3

HN

OCMe3

HN

OH

* OH

O

OH

R

CO2

OH O R = C6H11, Ph Statine analogs

epi-N-Boc-statine

O

*

OH

*

N-Boc-statine

OH O

O

OCMe3

*

CH2Ph

H O

pyroGluPheN

O LeuMetNH2

F R R = H: Compactin R = Me: Mevinolin Scheme 1.19

Natural products and biologically active compounds synthesized by use of the HYTRA aldol method.

Fluoroolefin peptide mimic

1.4 Stereoselective Aldol Addition of Lithium, Magnesium and Sodium Enolates

43

O S *

HO

O

OH

N

O

O

O

HMG-CoA synthase inhibitor

*

O OH Epothilone A

OR' *

R

*

CHCO2R

Me3CMe2SiO

O

O

OSiMe2CMe3 OR''

Vitamin D3 A-ring building block

23-Hydroxy vitamin D3 building blocks

Scheme 1.19 (continued)

diastereomers. The question of the stereochemistry of nucleophilic addition to chiral aldehydes or ketones has been addressed by the seminal studies of Cram and co-workers [160–162]. Indeed, the stereochemical outcome of this type of reaction is best rationalized by either the ‘‘Cram–Felkin–Anh model’’ [163] or ‘‘Cram’s cyclic model’’ when the carbonyl group is substituted by ether or amino residues. The latter type of stereochemical result has also been termed ‘‘chelation control’’, the former type of reaction accordingly as ‘‘non chelation control’’ [164]. Most of these investigations have been performed on racemic substrates. In view of the tendency towards enantiomerically pure products, emphasis is given here to those procedures that start from non-racemic aldehydes or ketones. In the aldol reaction ‘‘chelation control’’ is usually provided by those variants that use strong Lewis acidic enolate metals, for example boron, titanium, and tin. In particular, ‘‘chelation control’’ results very frequently in Mukaiyama aldol additions to a-oxygen- and a-nitrogen-substituted aldehydes [165]. Lithium enolates, in contrast, either give predominantly the product predicted by the ‘‘Cram–Felkin–Anh model’’ or react more or less nonstereoselectively. Thus, the favored formation of the syn-aldol product in the reaction of 2-phenylpropanal with the lithium enolates of acetone, pinacolone, methyl acetate, or N,N-dimethylacetamide is in accordance with Cram’s rule or the Felkin–Anh model (Eq. (35)). However, a rather moderate syn:anti ratio of 3:1 is typical of this type of reaction [51, 67].

44

1 Fundamentals and Transition-state Models. Aldol Additions of Group 1 and 2 Enolates

HO

F

O O

*

H

F

OH

OH

O

* N

OCMe3 F

N F ref. [141]

ref. [152]

OH

OH

CO2H

CO2Na

*

N

HO *

O

OH

O

N

F ref. [154]

ref. [153]

HO

HO

O *

O O

O

O

*

F

N R

H

H

R N

OH

N N Ph

HO

R

CONHPh

ref. [155]

ref. [157]

ref. [156]

R

R

OH *

N

N

R

R ref. [158] Scheme 1.20

Hypocholesterolemic-active drugs (HMGCoA reductase inhibitors) prepared by use of the HYTRA aldol method (selected examples).

*

OH CO2H

O

1.4 Stereoselective Aldol Addition of Lithium, Magnesium and Sodium Enolates

O

Ph

Ph OH

O

HO

Ph Ph (R)-83

MO MO

Ph Ph

O

Ph

O R

Si

O

O Ph (S)-83

Ph

Re

45

Ph Ph

OM OM

Ph

O

H

(R)-84

(S)-84

Scheme 1.21

Lk-topicity in aldol additions of (R) and (S ) acetate 83.

H

Ph

R

R

+ Ph

+ OLi

O

OH

O

OH

syn

O

anti 25

:

75

R

Ph

ð35Þ When the a-substituted (E) enolate 89 is added to 2-phenylpropanal, control of simple diastereoselectivity is provided in as far as the products with the 2,3-anti conﬁguration result exclusively. Induced stereoselectivity is lower, however, as indicated by the 80:20 ratio of diastereomeric b-hydroxy esters 90a and 90b (Eq. (36)) [51]. OLi + Ph

CHO 89

O

Ph OH

+

O 90a 80

O

Ph OH

:

O 90b 20

ð36Þ

46

1 Fundamentals and Transition-state Models. Aldol Additions of Group 1 and 2 Enolates

A single diastereomer 93, however, results from addition of the lithium enolate 92 derived of t-butyl thiopropanoate to the chiral, enantiomerically pure aldehyde 91. The transformation is a key carbon-chain-elongation step in Woodward’s synthesis of erythromycin A (Eq. (37)) [166]. Somewhat lower diastereoselectivity is observed in the aldol reaction between the lithium enolate 95 and the chiral aldehyde 94, a transformation used in a synthesis of maytansin (Eq. (38)). The diastereomeric adducts 96a and 96b result in a ratio of 90:10 [167].

H MeO

O

O

O

O

OAc

O

O

SCMe3

+

OLi 91

92

ð37Þ SCMe3 MeO

O

O

O

O

OAc

O

OH

O

93

O

H

O

Me3SiO

SLi +

H

SEt

94

O

95

H

OH

S

O SEt

Me3SiO 96a

+

:

H

OH

S SEt

Me3SiO 96b

ð38Þ A variety of a-alkoxy-substituted aldehydes have been submitted to aldol addition of lithium enolates. ‘‘Cram–Felkin–Anh’’ selectivity is usually observed, although often with rather low stereoselectivity. Exceptionally high diastereoselectivity results from the aldol reaction between the lithium enolate of pinacolone and isopropylidene glyceraldehyde. Thus, the b-hydroxy ketone 97 is obtained as a single product (Eq. (39)). Distinctly lower selectivity is observed when the same aldehyde is submitted to aldol additions of ester enolates, however [168].

1.4 Stereoselective Aldol Addition of Lithium, Magnesium and Sodium Enolates

O

O

CMe3

O

O

+

H

CMe3

OLi O

OH

O

97

ð39Þ The N-dibenzyl protecting group has been developed as a tool to provide non-chelate-controlled additions to a-amino aldehydes. Thus, anticonﬁgured aldol adducts are obtained predominantly when a-N-dibenzylprotected aldehydes 98 are submitted to aldol additions of lithium enolates as shown in Eqs. (40) and (41) [169, 170]. O R1

OLi R2

+

H

OMe R2

(PhCH2)2N 98 OH

O

1

R

OH OMe

(PhCH2)2N

R

+

R2 R 2

OMe

(PhCH2)2N

ð40Þ

R 2 R2 syn

anti

R1

R2

diastereomeric ratio anti : syn

yield

Me

H

95 : 5 90 : 10

82

97 :

84

PhCH2

H

Me

Me

(PhCH2)2N

O

1

83

3

O OLi H

+ OMe

O

O

(PhCH2)2N

(PhCH2)2N OMe OH

82%

d. r.:

>96

OMe

+

:

OH

<4

ð41Þ

47

48

1 Fundamentals and Transition-state Models. Aldol Additions of Group 1 and 2 Enolates

Further selected examples of diastereoselective aldol reactions between lithium enolates and chiral aldehydes are given in Eqs. (42) [171], (43) [172], and (44) [173]. In the last example, the salt-free generation of the lithium enolate was occasionally found to be crucial to stereoselectivity [174]. Me3CMe2SiO

O

OLi +

H

CMe3

Me3CMe2SiO

OH

Me3CMe2SiO

O CMe3

d. r.:

OH

O

+

CMe3

:

88

12

ð42Þ OR PhCH2 CH O 2 N H

OLi +

O

O

OR

OR

PhCH2 CH OH 2 N

O

PhCH2 O

CH2

OH

O

N

+

O

O

O

d. r.:

:

70

30

ð43Þ O H PhCH2O

H

OLi

H

OCMe3

+

O

87%

O

O OCMe3 PhCH2O d. r.:

OH 82

OCMe3

+ PhCH2O

O :

OH

O

18

ð44Þ

1.4 Stereoselective Aldol Addition of Lithium, Magnesium and Sodium Enolates

1.4.3

Addition of Chiral Enolates to Chiral Carbonyl Compounds

If a chiral aldehyde reacts with an achiral enolate the induced stereoselectivity is determined by the ‘‘inherent’’ preference of the aldehyde to be attacked from its Re or Si face. If, however, a chiral aldehyde is combined with a chiral enolate one must consider whether the inherent selectivities of the two reagents will be consonant in one of the combinations (‘‘matched pair’’), but dissonant in the other combination (‘‘mismatched pair’’). Thus, diﬀerent diastereoselectivity results from each combinations. The problem of insuﬃcient stereoselectivity in the ‘‘mismatched’’ combination can be solved by means of highly eﬃcient chiral enolates which can ‘‘outplay’’ the inherent selectivity of the aldehyde. The concept has been applied extensively in the context of boron enolates, a topic that has been reviewed comprehensively [52] and is discussed in detail in Chapter 3 of Part I of this book. The principle of ‘‘enolate controlled’’ stereochemistry can be demonstrated by use of the chiral acetate 83. When doubly deprotonated (R)- and (S)-HYTRA 83 reacts with enantiomerically pure 3-benzyloxybutanal 99 the (R)-conﬁgured acetate enolate attacks the aldehyde 99 (irrespective of its chirality) predominantly from the Re face so that, after hydrolysis, anti hydroxycarboxylic acid 100a results. On the other hand, the (S)-conﬁgured enolate of 83 attacks the enantiomerically pure aldehyde preferentially from the Si side to give syn carboxylic acids 100b with comparable selectivity, as shown in Scheme 1.22 [175]. In addition to this example of ‘‘acyclic’’ stereocontrol, the concept has also been applied to the cyclic chiral lithium enolate 101; this also resulted in high diastereoselectivity when the enantiomeric enolates were combined with the chiral aldehydes 102. As demonstrated by Eqs. (45) and (46), here again the stereochemical outcome is determined by the conﬁguration of the enolate 101 [176].

Me N

OLi H

+

Me3C N

O

Bz

ð45Þ

(R)-101 Me O

N Me3C N H

H

OBz

49

50

1 Fundamentals and Transition-state Models. Aldol Additions of Group 1 and 2 Enolates

O

Ph OH

O Ph

Ph

(R)-83 1. 2 LiN(i-Pr)2, THF, MgI2 2. PhCH2O O H

–125°C

99 3. KOH/MeOH, H2O

anti : syn :

:

90 PhCH2O

OH

10

O

PhCH2O

OH

O

+ OH

OH

anti-100a anti : syn :

syn-100b :

5

95

1. 2 LiN(i-Pr)2, THF, MgI2 2. PhCH2O O H 99 3. KOH/MeOH, H2O O

Ph OH

O Ph (S)-83 Scheme 1.22

‘‘Enolate controlled’’ addition of doubly deprotonated (R) and (S ) acetate 83 to (R)benzyloxybutanal.

Ph

–125°C

1.4 Stereoselective Aldol Addition of Lithium, Magnesium and Sodium Enolates

Me N

OLi H

+

Me3C N

O

Bz

ð46Þ

(S)-101 Me O

N Me3C N H

H

OBz

1.4.4

Addition of Achiral Enolates to Achiral Carbonyl Compounds in the Presence of Chiral Additives and Catalysts

Combination of achiral enolates with achiral aldehydes mediated by chiral ligands at the enolate counter-ion opens another route to non-racemic aldol adducts. Again, this concept has been extremely fruitful for boron, tin, titanium, zirconium and other metal enolates. It has, however not been applied very frequently to alkaline and earth alkaline metals. The main, inherent, drawback in the use of these metals is that the reaction of the corresponding enolate, which is not complexed by the chiral ligand, competes with that of the complexed enolate. Because the former reaction pathway inevitably leads to formation of the racemic product, the chiral ligand must be applied in at least stoichiometric amounts. Thus, any catalytic variant is excluded per se. Among the few approaches based on lithium enolates, early work revealed that the aldol addition of a variety of lithium enolates in the presence of (S,S)-1,4-(bisdimethylamino)-2,3-dimethoxy butane or (S,S)-1,2,3,4tetramethoxybutane provides only moderate induced stereoselectivity, typical ee values being 20% [177]. Chelation of the ketone enolate 104 by the chiral lithium amide 103 is more eﬃcient – the b-hydroxyl ketone syn-105 is obtained in 68% ee and no anti adduct is formed (Eq. (47)) [178]. Ph

O + CMe3

OMe N

Me2CH

+ LiN(i-Pr)2 Li

103

ð47Þ OLi CMe3 104

OH

O

PhCHO Ph

CMe3

105 92%, e.e.: 68%

51

52

1 Fundamentals and Transition-state Models. Aldol Additions of Group 1 and 2 Enolates

A variety of other chiral lithium amides, for example 106 and 108, have been applied more recently to bring about enantioselective aldol additions. As shown in Eqs. (48) [179] and (49) [180], both simple diastereoselectivity and induced stereroselectivity can be induced by these reagents. In the latter reaction, the enolate itself becomes chiral, because of desymmetrization of ketone 107 on deprotonation. Ph O LiN(i-Pr)2

+

N

+

O

N

OCMe3

Li O 106 OAc

OAc

O

1. PhCHO Ph

2. Ac2O

+

OCMe3

ð48Þ O

Ph

anti (76%) e.e.: 94%

OCMe3 syn (7%)

O Me O

O

Me

CMe3

+

Ph

+

N

LiCl

Li

107

108 CHO

95%

ð49Þ OH

H

O

O

O

Me

CMe3

single diastereomer e.e.: 90%

1.5

Conclusion

In a period longer than a century the aldol addition has proven itself to be an extremely useful ‘‘work horse’’ in organic synthesis. During the long history of this reaction the preformed-enolate technique was a breakthrough

References

enabling particular stereoselectivity to be achieved. Although boron and titanium become increasingly favored as counter-ions, lithium and magnesium enolates found their place in the repertoire of organic chemistry. The high reactivity of these reagents combined with easy handling has made them a useful tool in the hands of synthetic chemists.

References 1 A. T. Nielsen, W. J. Houlihan, Organic Reactions 1968, Vol.

16. 2 H. O. House, Modern Synthetic Reactions, 2nd ed., W. A.

Benjamin, Menlo Park, 1972. 3 C. H. Heathcock in Comprehensive Organic Synthesis, Vol. 2

(Ed.: B. M. Trost), Pergamon, Oxford 1991, Chapter 1.5. 4 R. Kane, J. Prakt. Chem. 1838, 15, 129. 5 J. G. Schmidt, Ber. Dtsch. Chem. Ges. 1880, 13, 2342; 1881, 14,

1459. 6 L. Claisen, A. Clapare`de, Ber. Dtsch. Chem. Ges. 1881, 14,

349. 7 ‘‘. . . dasselbe ist zugleich Aldehyd und Alkohol, . . . es ist der

8

9

10

11 12

13

14 15 16 17 18

erste Aldehyd eines Glycols’’: W. Wurtz, Ber. Dtsch. Chem. Ges. 1872, 5, 326. For the thermochemistry of the aldol reaction, see: J. P. Guthrie, Can. J. Chem. 1974, 52, 2037; 1978, 56, 962; E. M. Arnett, F. J. Fisher, M. A. Nichols, A. A. Ribeiro, J. Am. Chem. Soc. 1989, 111, 748; J. P. Richard, R. W. Nagorski, J. Am. Chem. Soc. 1999, 121, 4763. In the case of non-coordinating counter ions, the formation of the aldolate anion has been shown to be endothermic; cf. R. Noyori, I. Nishida, J. Sakata, M. Nishizawa, J. Am. Chem. Soc. 1980, 102, 1223. V. Grignard, A. Vesterman, Bull. Soc. Chim. Fr. 1925, 37, 425; C. Weizmann, S. F. Garrard, J. Chem. Soc. 1920, 117, 324. M. Ha¨usermann, Helv. Chim. Acta 1951, 34, 1482. J. B. Conant, N. Tuttle in R. Amus, Organische Synthesen, Vieweg u. Sohn, Braunschweig 1937, p. 192; cf. Organikum, Wiley–VCH, Weinheim, 21nd ed.; 2001, p. 522. For more recent applications of solid basic metal salts as catalysts in aldol reactions, see: J. F. Sanz, J. Oviedo, A. Ma´rquez, J. A. Odriozola, M. Montes, Angew. Chem. Int. Ed. Engl. 1999, 38, 506; G. Zhang, H. Hattori, K. Tanabe, Bull. Chem. Soc. Jap. 1989, 62, 2070; K. Tanabe, G. Zhang, H. Hattori, Appl. Catalysis 1989, 48, 63; G. Zhang, H. Hattori, K. Tanabe, Appl. Catalysis 1988, 36, 189. G. Wittig, Top. Curr. Chem. 1976, 67, 1. G. Wittig, H. Reiff, Angew. Chem. Int. Ed. Engl. 1968, 7, 7. T. Mukaiyama, Organic Reactions 1982, 28, 203. A. Russell, R. L. Kenyon, Org. Synth. Coll. Vol. III, 1955, 747. C. R. Conard, M. A. Dolliver, Org. Synth. Coll. Vol. II, 1943, 167.

53

54

1 Fundamentals and Transition-state Models. Aldol Additions of Group 1 and 2 Enolates 19 J. G. Leuck, L. Cejka, Org. Synth. Coll. Vol. I, 1944, 283. 20 E. P. Kohler, H. M. Chadwell, Org. Synth. Coll. Vol. I, 1944, 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

78. G. Kabas, Tetrahedron 1966, 22, 1213. M. E. Jung, Tetrahedron 1976, 32, 3. R. E. Gawley, Synthesis 1976, 777. R. B. Woodward, F. Sondheimer, D. Taub, K. Heusler, W. M. McLamore, J. Am. Chem. Soc. 1952, 74, 4223. ¨ chi, B. Gubler, R. S. Schneider, J. Wild, J. Am. G. Bu Chem. Soc. 1967, 89, 2776. G. I. Poos, W. F. Johns, L. H. Sarett, J. Am. Chem. Soc. 1955, 77, 1026. J. A. Marshall, A. E. Greene, J. Org. Chem. 1972, 37, 982. T.-L. Ho, Synth. Commun. 1974, 4, 265 and references given therein. W. S. Johnson, J. J. Korst, R. A. Clement, J. Dutta, J. Am. Chem. Soc. 1960, 82, 614. For a short survey, see: J. Mulzer, H.-J. Altenbach, M. Braun, K. Krohn, H.-U. Reissig, Organic Synthesis Highlights, VCH, Weinheim 1991, 232. J. N. Collie, J. Chem. Soc. 1907, 91, 1806. T. M. Harris, C. M. Harris, Tetrahedron 1977, 33, 2159. S. G. Gilbreath, C. M. Harris, T. M. Harris, J. Am. Chem. Soc. 1988, 110, 6172. G. Bringmann, Tetrahedron Lett. 1982, 23, 2009; G. Bringmann, J. R. Jansen, Tetrahedron Lett. 1984, 25, 2537. Z. G. Hajos, D. R. Parrish, J. Org. Chem. 1974, 39, 1615; Org. Synth. 1985, 63, 26. U. Eder, G. Sauer, R. Wiechert, Angew. Chem. Int. Ed. Engl. 1971, 10, 496. G. Sauer, U. Eder, G. Haffer, G. Neef, R. Wiechert, Angew. Chem. Int. Ed. Engl. 1975, 14, 417; P. Buchschacher, ¨ rst, Org. Synth. 1985, 63, 37. A. Fu C. Agami, C. Puchot, H. Sevestre, Tetrahedron Lett. 1986, 27, 1501; C. Agami, Bull. Soc. Chim. Fr. 1988, 499. B. List, R. A. Lerner, C. F. Barbas III, J. Am. Chem. Soc. 2000, 122, 2395; A. B. Northrup, D. W. C. MacMillan, J. Am. Chem. Soc. 2002, 124, 6798. F. C. Frostick, Jr., C. R. Hauser, J. Am. Chem. Soc. 1949, 71, 1350; M. Hammell, R. Levine, J. Org. Chem. 1950, 15, 162. M. W. Rathke, J. Am. Chem. Soc. 1970, 92, 3222; C. R. ¨ ger, E. G. Rochow, J. Organomet. Chem. 1964, 1, 476. Kru M. W. Rathke, A. Lindert, J. Am. Chem. Soc. 1971, 93, 2318. R. A. Olofson, C. M. Dougherty, J. Am. Chem. Soc. 1973, 95, 581, 582. H. O. House, M. Gall, H. D. Olmstead, J. Org. Chem. 1971, 36, 2361. C. H. Heathcock in Modern Synthetic Methods 1992, (Ed.: R. Scheffold), VHCA, VCH, Basel, Weinheim 1992, 1. C. H. Heathcock in Comprehensive Organic Synthesis, Vol. 2 (Ed.: B. Trost), Pergamon, Oxford 1993, Chapter 1.6. For comprehensive reviews, see: D. Seebach, Angew. Chem. Int. Ed. Engl. 1988, 27, 1624; C. Lambert, P. v. R. Schleyer in Houben Weyl, Methoden der Organischen Chemie, Vol. E 19d

References

48 49 50

51 52 53 54 55

56 57

58 59 60 61

62

63 64 65

66 67

68

(Ed.: M. Hanack), Thieme, Stuttgart 1993, p. 75; L. M. Jackman, J. Bortiatynski in Advances in Carbanion Chemistry, Vol. 1 (Ed.: V. Snieckus), JAI Press, Greenwich, Connecticut 1992, p. 45; P. G. Williard in Comprehensive Organic Synthesis, Vol. 1 (Ed.: B. M. Trost), Pergamon, Oxford 1991, Chap. 1.1. C. H. Heathcock, Science 1981, 214, 395. D. A. Evans, J. V. Nelson, T. R. Taber, Top. Stereochem. 1982, 13, 1. C. H. Heathcock, in Comprehensive Carbanion Chemistry, Part B (Eds.: E. Buncel, T. Durst), Elsevier, Amsterdam 1984, Chap. 4. C. H. Heathcock, in Asymmetric Synthesis Vol. 3, Part B (Ed.: J. D. Morrison), Academic Press, New York 1984, Chap. 2. S. Masamune, W. Choy, J. S. Petersen, L. R. Sita, Angew. Chem. Int. Ed. Engl. 1985, 24, 1. M. Braun, Angew. Chem. Int. Ed. Engl. 1987, 26, 24. M. Braun in Advances in Carbanion Chemistry, Vol. 1 (Ed.: V. Snieckus), JAI Press, Greenwich, Connecticut 1992, p. 177. B. M. Kim, S. F. Willians, S. Masamune in Comprehensive Organic Synthesis, Vol. 2 (Ed.: B. M. Trost), Pergamon, Oxford 1993, Chap. 1.7; M. W. Rathke, P. Weipert, ibid., Chap. 1.8.; I. Paterson, ibid., Chap. 1.9; C. Gennari, ibid., Chap. 2.4. M. Braun, H. Sacha, J. Prakt. Chem. 1993, 335, 653. M. Braun in Houben Weyl, Methoden der Organischen Chemie, Vol. E21b (Eds.: G. Helmchen, R. W. Hoffmann, J. Mulzer, E. Schaumann), Thieme, Stuttgart 1996 p. 1603. R. Mahrwald, Chem. Rev. 1999, 99, 1097. M. Sawamura, Y. Ito in Catalytic Asymmetric Synthesis (Ed.: I. Ojima), VCH, Weinheim 1993, chapter 7.2. C. Palomo, M. Oiarbide, J. M. Garcı´a, Chem. Eur. J. 2002, 8, 37; Chem. Soc. Rev. 2004, 33, 65. The stereochemical notation follows that proposed in the Houben–Weyl edition on ‘‘Stereoselective Synthesis’’, cf. G. Helmchen in Houben–Weyl, Methoden der Organischen Chemie, Vol. E21a (Eds.: G. Helmchen, R. W. Hoffmann, J. Mulzer, E. Schaumann), Thieme, Stuttgart 1996, p. 1. See textbooks on stereochemistry: e.g.: E. L. Eliel, S. H. Wiley, L. N. Mander, Stereochemistry of Organic Compounds, Wiley, New York 1994. D. Seebach, V. Prelog, Angew. Chem. Int. Ed. Engl. 1982, 21, 654. S. Masamune, S. A. Ali, D. L. Snitman, D. S. Garvey, Angew. Chem. Int. Ed. Engl. 1980, 19, 557. J. E. Dubois, M. Dubois, Tetrahedron Lett. 1967, 4215; J. Chem. Soc., Chem. Commun. 1968, 1567; Bull. Soc. Chim. Fr. 1969, 3120, 3553; J. E. Dubois, P. Fellmann, C. R. Acad. Sci. Se´r. C 1972, 274, 1307; Tetrahedron Lett. 1975, 1225. C. H. Heathcock, C. T. Buse, W. A. Kleschick, M. C. Pirrung, J. E. Sohn, J. Lampe, J. Org. Chem. 1980, 45, 1066. C. H. Heathcock, M. C. Pirrung, J. Org. Chem. 1980, 45, 1727; C. H. Heathcock, M. C. Pirrung, S. H. Montgomery, J. Lampe, Tetrahedron 1981, 37, 4087. P. Fellmann, J.-E. Dubois, Tetrahedron 1978, 34, 1349.

55

56

1 Fundamentals and Transition-state Models. Aldol Additions of Group 1 and 2 Enolates 69 T. H. Chan, T. Aida, P. W. K. Lau, V. Gorys, D. N. Harpp,

Tetrahedron Lett. 1979, 4029. 70 R. Noyori, I. Nishida, J. Sakata, J. Am. Chem. Soc. 1981,

103, 2106. 71 See, for example, the stereochemical outcome of aldol addition

72 73 74 75 76

77 78 79

80

81 82 83 84 85 86 87 88

89 90 91

92

performed with the lithium enolate of cyclohexanone, which deﬁnitely has (E) conﬁguration, H. O. House, D. S. Crumrine, A. Y. Teranishi, H. D. Olmstead, J. Am. Chem. Soc. 1973, 95, 3310; M. Majewski, D. M. Gleave, Tetrahedron Lett. 1989, 42, 5681. D. Ivanoff, A. Spassoff, Bull. Soc. Chim. Fr. 1931, 49, 371; D. Ivanoff, N. Nicoloff, Bull. Soc. Chim. Fr. 1932, 51, 1325. H. E. Zimmerman, M. D. Traxler, J. Am. Chem. Soc. 1957, 79, 1920. S. Masamune, T. Sato, B. M. Kim, T. Wollmann, J. Am. Chem. Soc. 1986, 108, 8279. E. J. Corey, S. S. Kim, J. Am. Chem. Soc. 1990, 112, 4976. R. Armstutz, W. B. Schweizer, D. Seebach, J. D. Dunitz, Helv. Chim. Acta 1981, 64, 2617; E. Juaristi, A. K. Beck, J. Hansen, T. Matt, T. Mukhopadhyay, M. Simson, D. Seebach, Synthesis 1993, 1271. D. A. Evans, L. R. McGee, Tetrahedron Lett. 1980, 25, 3975. R. W. Hoffmann, K. Ditrich, S. Froech, D. Cremer, Tetrahedron 1985, 41, 5517. R. W. Hoffmann, K. Ditrich, Tetrahedron Lett. 1984, 25, 1781; C. Gennari, L. Colombo, C. Scolastico, R. Todeschini, Tetrahedron 1984, 40, 4051. R. Noyori, S. Murata, M. Suzuki, Tetrahedron 1981, 37, 3899; T. Mukaiyama, S. Kobayashi, M. Tamura, Y. Sagawa, Chem. Lett. 1987, 491; E. Nakamura, M. Shimizu, I. Kuwajima, J. Sakata, K. Yokoyama, R. Noyori, J. Org. Chem. 1983, 48, 932. S. S. Labadie, J. K. Stille, Tetrahedron 1984, 40, 2329. D. A. Evans, L. R. McGee, J. Am. Chem. Soc. 1981, 103, 2876; T. Katuski, M. Yamaguchi, Tetrahedron Lett. 1985, 26, 5807. R. W. Hoffmann, Pure Appl. Chem. 1988, 60, 123 and references given therein. C. Gennari, R. Todeschini, M. G. Baretta, G. Favini, C. Scolastico, J. Org. Chem. 1986, 51, 612. I. Kuwajima, E. Nakamura, Acc. Chem. Res. 1985, 18, 181. R. Mahrwald, B. Ziemer, Tetrahedron Lett. 2002, 43, 4459. S. E. Denmark, X. Su, Y. Nishigaichi, J. Am. Chem. Soc. 1998, 120, 12990. ¨ ntrup, J. Finke, M. Zippel, J. Am. Chem. J. Mulzer, G. Bru ¨ntrup, J. Soc. 1979, 101, 7723; J. Mulzer, M. Zippel, G. Bru Senger, J. Finke, Liebigs Ann. Chem. 1980, 1108. N. T. Anh, B. T. Thanh, Nouv. J. Chim. 1986, 10, 681. Y. Yamamoto, K. Maruyama, Tetrahedron Lett. 1980, 21, 4607. D. A. Evans, M. C. Kozlowski, J. A. Murry, C. S. Burgey, K. R. Campos, B. T. Connell, R. J. Staples, J. Am. Chem. Soc. 1999, 121, 669. Y. Li, M. N. Paddon-Row, K. Houk, J. Org. Chem. 1990, 55, 481.

References 93 R. Leung-Toung, T. T. Tidwell, J. Am. Chem. Soc. 1990, 112,

1042. 94 Based on calculations, a similar coordination of the lithium

95

96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115

116

cation has been postulated previously, cf. E. Kaufmann, P. v. R. Schleyer, K. N. Houk, Y. D. Wu, J. Am. Chem. Soc. 1985, 107, 5560; S. M. Bachrach, A. Streitwieser, Jr., J. Am. Chem. Soc. 1986, 108, 3946; G. Stork, R. L. Polt, Y. Li, K. N. Houk, J. Am. Chem. Soc. 1988, 110, 8360; S. M. Bachrach, J. P. Ritchie, J. Am. Chem. Soc. 1989, 111, 3134; E. Kaufmann, S. Sieber, P. v. R. Schleyer, J. Am. Chem. Soc. 1989, 111, 4005; A. E. Dorigo, K. Morokuma, J. Am. Chem. Soc. 1989, 111, 4635; see also ref. 92. ¨rgi, E. Shefter, J. D. Dunitz, Tetrahedron 1975, 31, H. B. Bu ¨ rgi, J. D. Dunitz, E. Shefter, J. Am. Chem. 3089; H. B. Bu Soc. 1973, 95, 5065. S. E. Denmark, B. R. Henke, J. Am. Chem. Soc. 1991, 113, 2177. C. H. Heathcock, C. T. White, J. J. Morrison, D. VanDerveer, J. Org. Chem. 1981, 46, 1296. C. H. Heathcock, C. T. White, J. Am. Chem. Soc. 1979, 101, 7076. C. H. Heathcock, S. Arseniyadis, Tetrahedron Lett. 1985, 26, 6009. N. A. Van Draanen, S. Arseniyadis, M. T. Crimmins, C. H. Heathcock, J. Org. Chem. 1991, 56, 2499. A. Choudhury, E. Thornton, Tetrahedron 1992, 48, 5701. A. Choudhury, E. Thornton, Tetrahedron Lett. 1993, 34, 2221. J. B. Goh, B. R. Lagu, J. Wurster, D. C. Liotta, Tetrahedron Lett. 1994, 35, 6029. M. Braun, H. Sacha, Angew. Chem. Int. Ed. Engl. 1991, 30, 1318. ¨ller, M. Braun, Chem. Ber. 1994, 127, H. Sacha, D. Waldmu 1959. D. Seebach, A. R. Sting, M. Hoffmann, Angew. Chem. Int. Ed. Engl. 1996, 35, 2708. D. Seebach, R. Naef, G. Calderari, Tetrahedron 1984, 40, 1313. D. Seebach, T. Weber, Helv. Chim. Acta 1984, 67, 1650. W. Amberg, D. Seebach, Chem. Ber. 1990, 123, 2413. D. Seebach, J.-M. Lapierre, W. Jaworek, P. Seiler, Helv. Chim. Acta 1993, 76, 459. I. Hoppe, D. Hoppe, R. Herbst-Irmer, E. Egert, Tetrahedron Lett. 1990, 31, 6859. S. G. Davies, I. M. Dordor, P. Warner, J. Chem. Soc., Chem. Commun. 1984, 956. S. G. Davies, Pure Appl. Chem. 1988, 60, 13. L. S. Liebeskind, M. E. Welker, Tetrahedron Lett. 1984, 25, 4341. J. S. McCallum, L. S. Liebeskind in Houben–Weyl, Methoden der Organischen Chemie, Vol. E21b (Eds.: G. Helmchen, R. W. Hoffmann, J. Mulzer, E. Schaumann), Thieme, Stuttgart 1996, p. 1667. P. W. Ambler, S. G. Davies, Tetrahedron Lett. 1985, 26, 2129.

57

58

1 Fundamentals and Transition-state Models. Aldol Additions of Group 1 and 2 Enolates 117 S. G. Davies, M. Wills, J. Organomet. Chem. 1987, 328, C 29. 118 S. G. Davies, D. Middlemiss, A. Naylor, M. Wills,

Tetrahedron Lett. 1989, 27, 2971. 119 D. Seebach, V. Ehrig, M. Teschner, Liebigs Ann. Chem.

1976, 1357. 120 C. Mioskowski, G. Solladie´, J. Chem. Soc., Chem. Commun.

1977, 162. 121 For reviews, see: G. Solladie´ in Asymmetric Synthesis Vol. 2,

122 123

124 125

126 127 128 129 130

131 132 133 134 135 136 137 138

139

Part A (Ed.: J. D. Morrison), Academic Press, New York 1983, p. 184. C. Mioskowski, G. Solladie´, Tetrahedron 1980, 36, 227. G. Solladie´ in Houben Weyl, Methoden der Organischen Chemie, Vol. E21b (Eds.: G. Helmchen, R. W. Hoffmann, J. Mulzer, E. Schaumann), Thieme, Stuttgart 1996, p. 1808. E. J. Corey, L. O. Weigel, A. R. Chamberlin, H. Cho, D. H. Hua, J. Am. Chem. Soc. 1980, 102, 6613. R. Annunziata, M. Cinquini, F. Cozzi, F. Montanari, A. Restelli, J. Chem. Soc., Chem. Commun. 1983, 1138; Tetrahedron 1984, 40, 3815; for stereoselective aldol additions of 2-(arylsulﬁnyl)phenyl ketones, see: S. Nakamura, M. Oda, H. Yasuda, T. Toru, Tetrahedron 2001, 57, 8469. S. S. Keito, K. Hatanaka, T. Kano, H. Yamamoto, Angew. Chem. Int. Ed. 1998, 37, 3378. I. Ojima, H. B. Kwon, J. Am. Chem. Soc. 1988, 110, 5617. M. Braun, R. Devant, Tetrahedron Lett. 1984, 25, 5031. R. Devant, U. Mahler, M. Braun, Chem. Ber. 1988, 121, 397. M. Braun, S. Gra¨f, S. Herzog, Org. Synth. 1993, 72, 32. For an alternative acylation of 82, see: J. Macor, A. J. Sampognaro, P. R. Verhoest, R. A. Mack, Org. Synth. 2000, 77, 45; for a modiﬁed procedure of the preparation of 82, see: A. Millar, L. W. Mulder, K. E. Mennen, C. W. Palmer, Organic Preparations and Procedures Int. 1991, 23, 173; Chem. Abstr. 1991, 114, 163667. M. Braun, S. Gra¨f, Org. Synth. 1993, 72, 38. S. Gra¨f, M. Braun, Liebigs Ann. Chem. 1993, 1091. M. Hashmi, S. Gra¨f, M. Braun, M. W. Anders, Chem. Res. Toxicol. 1996, 9, 361. ¨ller, Synthesis 1989, 856. M. Braun, D. Waldmu M. Braun, C. Bauer, Liebigs Ann. Chem. 1991, 1157. M. Braun, J. Moritz, Synlett 1991, 750. W. R. Ewing, B. D. Harris, K. L. Bhat, M. M. Joullie´, Tetrahedron 1986, 42, 2421. P. Barbier, F. Schneider, U. Widmer, Helv. Chim. Acta 1987, 70, 1412; Eur. Patent 185359, 1986; P. Barbier, F. Schneider, U. Widmer, Hoﬀmann–La Roche and Co. AG; Chem. Abstr. 1987, 106, 102077; Eur. Patent 189577, 1986; Chem. Abstr. 1987, 106, 196249; US Patent 4931463, 1990; P. Barbier, F. Schneider, U. Widmer, Hoﬀmann–La Roche, Inc.; Chem. Abstr. 1991, 114, 82542; Eur. Patent 444482, 1991; R. Derunges, H. P. Maerki, H. Stalder, A. Szente, Hoﬀmann–La Roche AG; Chem. Abstr., 1992, 117, 7786. P. G. M. Wuts, S. R. Putt, Synthesis 1989, 951.

References 140 R. Devant, H. E. Radunz, Tetrahedron Lett. 1988, 2307;

141

142

143 144

145 146 147 148

149 150 151

152

153

154

German Patent 3743225, 1989; R. M. Devant, H. E. Radunz, Merck GmbH; Chem. Abstr. 1990, 112, 36458. J. E. Lynch, R. P. Volante, J. V. Wattley, I. Shinkai, Tetrahedron Lett. 1987, 28, 1385; Eur. Patent 251714, 1988; R. P. Volante, E. Corley, I. Shinkai, Merck and Co., Inc.; Chem. Abstr. 1988, 108, 150455; US Patent 4611081, 1986; J. E. Lynch, I. Shinkai, R. P. Volante, Merck and Co., Inc.; Chem. Abstr. 1987, 106, 18119. ¨ hler, Tetrahedron T. Allmendinger, E. Felder, E. Hungerbu Lett. 1990, 31, 7301; Eur. Patent 353732, 1990; T. ¨hler, R. Lattmann, S. Ofner, Allmendinger, E. Hungerbu W. Schilling, G. von Sprecher, E. Felder, Ciba–Geigy AG; Chem. Abstr. 1990, 113, 153046. S. Wattanasin, H. D. Do, N. Bhongle, F. G. Kathawala, J. Org. Chem. 1993, 58, 1610. D. Schinzer, A. Bauer, O. M. Bohm, A. Limberg, M. Cordes, Chem. Eur. J. 1999, 5, 2483; Int. Patent 9959985, 1999; K.-H. Altmann, A. Bauer, D. Schinzer; Novartis AG; Chem. Abstr. 2000, 132, 3284. K. Nagasawa, H. Ishihara, Y. Zako, I. Shimizu, J. Org. Chem. 1993, 58, 2523. Jap. Patent 08301811, 1996; S. Ri, M. Yamanashi, N. Shimizu; Kuraray Co; Chem. Abstr. 1997, 126, 131697. C. T. Brain, A. Chen, A. Nelson, N. Tanikkul, E. J. Thomas, Tetrahedron Lett. 2001, 42, 1247. M. Suzuki, Y. Yanagawa, H. Iwasaki, H. Kanda, K. Yanagihara, H. Matsumoto, Y. Ohara, Y. Yazaki, R. Sakoda, Bioorg. Med. Chem. Lett. 1999, 9, 2977. J. M. Weiss, H. M. R. Hoffmann, Tetrahedron: Asymmetry 1997, 8, 3913. S. Mills, R. Desmond, R. A. Reamer, R. P. Volante, I. Shinkai, Tetrahedron Lett. 1988, 29, 281. US Patent 5216015, 1993; D. G. McGarry, F. A. Volz, J. R. Regan, M. N. Chang; Rohne–Poulenc Rorer Pharmaceuticals Inc.; Chem. Abstr. 1993, 119, 226236. H. Jendralla, E. Baader, W. Bartmann, G. Beck, A. Bergmann, E. Granzer, B. v. Kerekjarto, K. Kesseler, R. Krause, W. Schubert, G. Wess, J. Med. Chem. 1990, 33, 61; H. Jendralla, E. Granzer, B. v. Kerekjarto, R. Krause, U. Schacht, E. Baader, W. Bartmann, G. Beck. A. Bergmann, K. Kesseler, G. Wess, L.-J. Chen, S. Granata, J. Herchen, ¨ ssler, K. Wagner, J. Med. Chem. 1991, H. Kleine, H. Schu 34, 2962; German Patent 3722807, 1989; G. Wess, W. Bartmann, G. Beck, E. Granzer, Hoechst AG; Chem. Abstr. 1989, 110, 231121; Eur. Patent 324347, 1989; E. Baader, H. Jendralla, B. Kerekjarto, G. Beck, Hoechst AG; Chem. Abstr. 1990, 112, 21003; German Patent 3826814, 1990; G. Beck, W. Bartmann, G. Wess, E. Granzer, Hoechst AG; Chem. Abstr. 1991, 114, 23976. O. Tempkin, S. Abel, C.-P. Chen, R. Underwood, K. Prasad, K.-M. Chen, O. Repic, T. J. Blacklock, Tetrahedron 1997, 53, 10659. D. V. Patel, R. J. Schmidt, E. M. Gordon, J. Org. Chem.

59

60

1 Fundamentals and Transition-state Models. Aldol Additions of Group 1 and 2 Enolates

155

156

157 158

159 160 161 162

163 164 165

166

167 168

1992, 57, 7143; German Patent 3805801, 1988; J. J. Wright, S. Y. Sit, Bristol–Myers Co.; Chem. Abstr. 1989, 110, 114836. B. D. Roth, C. J. Blankley, A. W. Chucholowski, E. Ferguson, M. L. Hoeﬂe, D. F. Ortwine, R. S. Newton, C. S. Sekerke, D. R. Sliskovic, C. D. Stratton, M. M. Wilson, J. Med. Chem. 1991, 34, 357; Eur. Patent 409281, 1991; B. D. Roth, Warner–Lambert Co.; Chem. Abstr. 1991, 115, 29107; see also: German Patent 3932887, 1990; N. S. Watson, C. Chan, B. C. Ross, Glaxo Group Ltd.; Chem. Abstr. 1991, 114, 164567. For a synthesis of the [14C] side chain-labeled compound, see: T. H. Lee, P. W. K. Woo, J. Labelled Compounds and Radiopharmaceuticals 1999, 42, 129. – For a review, see: B. D. Roth, Progress in Med. Chem. 2002, 40, 1. D. R. Sliskovic, C. J. Blankley, B. R. Krause, R. S. Newton, J. A. Picard, W. H. Roark, B. D. Roth, C. Sekerke, M. K. Shaw, R. L. Stanﬁeld, J. Med. Chem. 1992, 35, 2095. Eur. Patent 424929, 1991; H. Natsugari, H. Ikeda, Takeda Chemical Industries, Ltd.; Chem. Abstr. 1991, 115, 114373. Int. Patent, 1991; M. Matsuo, T. Manabe, H. Okumura, H. Matsuda, N. Fujii, Fujisawa Pharmaceutical Co. Ltd; Chem. Abstr. 1992, 116, 151782. K. Prasad, K. M. Chen, O. Repic, G. E. Hardtmann, Tetrahedron: Asymmetry 1990, 1, 703. D. J. Cram, F. A. Abd Elhafez, J. Am. Chem. Soc. 1952, 74, 5828. D. J. Cram, D. R. Wilson, J. Am. Chem. Soc. 1963, 85, 1245. For a discussion, see: E. L. Eliel in Asymmetric Synthesis, J. D. Morrison, Ed., Vol. 2, Part A, p. 125, Academic, New York 1983; J. Mulzer, H.-J. Altenbach, M. Braun, K. Krohn, H.-U. Reissig, Organic Synthesis Highlights, p. 3, VCH, Weinheim 1991. M. Che´rest, H. Felkin, N. Prudent, Tetrahedron Lett. 1968, 2199. Review: M. T. Reetz, Angew. Chem. Int. Ed. Engl. 1984, 23, 556. M. Braun in Houben–Weyl, Methoden der Organischen Chemie, Vol. E21b (Eds.: G. Helmchen, R. W. Hoffmann, J. Mulzer, E. Schaumann), Thieme, Stuttgart 1996, p. 1713. R. B. Woodward, E. Logusch, K. P. Nambiar, K. Sakan, D. E. Ward, B.-W. Au-Yeung, P. Balaram, L. J. Browne, P. J. Card, C. H. Chen, R. B. Cheˆnebert, A. Fliri, K. Frobel, H.-J. Gais, D. G. Garratt, K. Hayakawa, W. Heggie, D. P. Hesson, D. Hoppe, I. Hoppe, J. A. Hyatt, D. Ikeda, P. A. Jacobi, K. S. Kim, Y. Kobuke, K. Kojima, K. Krowicki, V. J. Lee, T. Leutert, S. Malchenko, J. Martens, R. S. Matthews, B. S. Ong, J. B. Press, T. V. Rajan Babu, G. Rousseau, H. M. Sauter, M. Suzuki, K. Tatsuta, L. M. Tolbert, E. A. Truesdale, I. Uchida, Y. Ueda, T. Uyehara, A. T. Vasella, W. C. Vladuchick, P. A. Wade, R. M. Williams, H. N.-C. Wong, J. Am. Chem. Soc. 1981, 103, 3210. A. I. Meyers, J. P. Hudspeth, Tetrahedron Lett. 1981, 22, 3925. C. H. Heathcock, S. D. Young, J. P. Hagaen, M. C. Pirrung, C. T. White, D. VanDerveer, J. Org. Chem. 1980, 45, 3846.

References 169 M. T. Reetz, M. W. Drewes, A. Schmitz, Angew. Chem. Int.

Ed. Engl. 1987, 26, 1141. 170 M. T. Reetz, Pure Appl. Chem. 1988, 60, 1607. 171 D. A. Evans, J. R. Gage, Tetrahedron Lett. 1990, 31, 6129. 172 E. J. Corey, G. A. Reichard, J. Am. Chem. Soc. 1992, 114,

10677. 173 J.-M. Escudier, M. Baltas, L. Gorrichon, Tetrahedron Lett.

1991, 32, 5345. 174 D. E. Ward, M. Sales, C. C. Man, J. Shen, P. Sasmal, C.

Guo, J. Org. Chem. 2002, 67, 1618. 175 U. Mahler, R. M. Devant, M. Braun, Chem. Ber. 1988, 121,

2035. 176 D. Seebach, E. Juaristi, D. D. Miller, C. Schickli, T.

Weber, Helv. Chim. Acta 1987, 70, 237. 177 D. Seebach, H. O. Kalinowski, B. Bastani, G. Crass, H.

Daum, N. Do¨rr, N. DuPreez, V. Ehrig, W. Langer, C. ¨sser, H.-A. Oei, M. Schmitt, Helv. Chim. Acta 1977, 60, Nu 301. For more recent applications of this method, see: Y. Nomura, M. Iguchi, H. Doi, K. Tomioka, Chem. Pharm. Bull. 2002, 50, 1131. 178 A. Ando, T. Shioiri, Tetrahedron 1989, 45, 4969. 179 M. Uragami, K. Tomioka, K. Koga, Tetrahedron: Asymmetry 1995, 6, 701. 180 M. Majewski, P. Nowak, J. Org. Chem. 2000, 65, 5152.

61

63

2

The Development of Titanium Enolate-based Aldol Reactions Arun K. Ghosh and M. Shevlin 2.1

Introduction

Asymmetric aldol reactions are very important carbon–carbon bondforming reactions in organic synthesis. Applications of aldol reactions have been widespread, particularly in the synthesis of complex and bioactive natural products. An aldol reaction can create two new stereogenic centers depending upon the choice of enolate substituent and the aldehyde or ketone. Furthermore, as shown in Scheme 2.1, the reaction can generate four possible diastereomers when X contains a chiral center – two syn (3 and 4) and two anti diastereomers (5 and 6). Signiﬁcant advances in asymmetric aldol reactions has led to control of aldol products with high enantio- and diastereoselectivity [1]. The ﬁeld of stereoselective synthesis of syn aldol products is particularly sophisticated. Important methods are also emerging for stereoselective anti aldol reactions. The use of chiral auxiliaries [2] has been an important concept in terms of generating one diastereomer in a stereopredictable manner. The ready availability of both enantiomers of a chiral auxiliary and easy removal and recovery are very important. syn and anti diastereoselectivity depends on several factors including enolate geometry, related metal ions, and reaction conditions. In general, Z enolates provide syn aldol products with high selectivity and E enolates aﬀord anti aldol products. This result can be rationalized on the basis of a cyclic Zimmerman–Traxler transition state [3]. As shown in transition state 7 for Z enolates (Scheme 2.2) the enolate substituent is generally pseudoaxial and the aldehyde substituent occupies a pseudoequatorial position, providing a syn product, 8. E enolates adopt transition state 9, in which the enolate and aldehyde substituents are pseudoequatorial and leads to the anti aldol product 10. Beside chair-like transition-state models, boat-like transition-state models have been proposed to explain stereoselectivity. It should be noted that strategies have been developed over the years to form the anti product from the Z enolate, as will be discussed in this chapter. Modern Aldol Reactions. Vol. 1: Enolates, Organocatalysis, Biocatalysis and Natural Product Synthesis. Edited by Rainer Mahrwald Copyright 8 2004 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 3-527-30714-1

2 The Development of Titanium Enolate-based Aldol Reactions

64

O

O

+

X

H

R2

R1

O

OH

X

O

R2

or

X

1

2

OH

O

R2

or

O

OH

X

R2

or

OH

X

R2

R1

R1

R1

R1

3

4

5

6

syn -aldol

anti-aldol

Scheme 2.1

Generation of four possible diastereomers when x contains a chiral center.

Diastereoselective aldol reactions have been extensively utilized in the synthesis of complex natural products, including macrolides [4] and ionophores [5]. In this context, iterative approaches are often exploited to append propionate units one at a time. This approach leads to double stereodiﬀerentiation [6] in which the reactant pairs can be either ‘‘matched’’ or ‘‘mismatched’’. The chirality of the two reactants reinforce each other if they are matched. As a result, the diastereoselectivity is often higher than would

R1 O O

R2

O O

R1 R2 M

X

X

9

7

O

O

OH

X

R2 R1

8 Scheme 2.2

Zimmerman–Traxler transition states.

OH

X

R2 R1

10

M

2.2 Additions of Enolates to Ketones

be expected from either reagent alone. In they are mismatched, the chirality of the reagents has a detrimental eﬀect, leading to a lower selectivity than would have been expected. The metal ion associated with the enolate has pronounced eﬀect on stereoselectivity. Numerous titanium enolate-based asymmetric aldol methods have provided convenient access to aldol products in enantiomerically pure form. The titanium enolate aldol reaction has tremendous synthetic potential, because titanium reagents are readily available and inexpensive. This chapter will focus on the development of a variety of titanium enolate aldol reactions. The ﬁrst titanium enolate aldol reactions were reported by De Kimpe and coworkers [7], who reported the self-condensation and elimination of achloroketones with TiCl 4 and pyridine.

2.2

Additions of Enolates to Ketones

Although aldol additions to aldehydes are robust methods and the corresponding theory is well developed, aldol additions to ketones are still largely unexplored, possibly because of the additional complexity of diﬀerentiating sterically between the two diﬀerent alkyl groups of ketones. The diﬃculty of aldol additions to ketones is apparent from the lack of asymmetric methods available for addition of titanium enolates to ketones. The notion of syn and anti products also tends to break down when dealing with non-symmetrical ketones with similar substituents. Tanabe and coworkers reported the ﬁrst instances of additions of titanium enolates to ketones [8]. They reported moderate to excellent yields for several reactions. When the two substituents on the ketone were suﬃciently sterically diﬀerentiated, good to excellent syn diastereoselectivity was observed. Representative examples of these crossed-aldol reactions are shown in Table 2.1. The initial method involved a catalytic (5 mol%) amount of trimethylsilyl triﬂate in the reaction mixture; it was suggested this generates a trichlorotitanium triﬂate species in situ which is more eﬀective at enolization. This theory was, however, called into question by reports of additions of enolates to ketones with trimethylsilyl chloride as catalyst, which also provides moderate to good yields [9]. Additions of enolates of phenyl esters or phenyl thioesters to ketones were also shown to proceed in good yield and with syn diastereoselectivity in the absence of silyl additive [10]. Oshima and coworkers reported reactions of aldehyde enolates with ketones [11]. They used enolates generated from silyl enol ether 14 with methyllithium and transmetalated with titanium tetrabutoxide. Representative examples are shown in Table 2.2. Selectivity was minimal, except for 1,1,1trichloroacetone (entry 5), for which the anti isomer was formed almost exclusively.

65

66

2 The Development of Titanium Enolate-based Aldol Reactions Tab. 2.1

Additions of enolates to ketones. O R1

O HO

TiCl4, base R2

R1

additive O then

11

R3

R3

O HO

+

R4

R1

R2

R4 R2

12

R4

R3

13

Entry R1

R2

R3

R4

Base

Additive

Yield (%) 12:13

Ref.

1 2 3 4 5 6 7 8 9 10 11

Me Me Me Me Et Et H Me Me Me Me

Ph Ph Ph Ph Ph n-C7 H15 Ph Et CH2 Cl Et Et

Me CH2 Cl Me Et Me Me CH2 Cl Ph Ph Et Ph

Bu3 N Bu3 N Bu3 N Bu3 N Bu3 N Bu3 N Bu3 N Et3 N Et3 N Et3 N Et3 N

TMSOTf TMSOTf TMSOTf TMSOTf TMSOTf TMSOTf TMSCl – – – –

95 91 84 92 72 60 94 83 77 77 98

8 8 8 8 8 8 9 10 10 10 10

Ph Ph Et Et Pr Pr TBSOCH2 PhO PhO PhS PhS

100:0 100:0 84:16 72:28 100:0 60:40 – 64:36 68:32 – 77:23

2.3

Addition of Enolates Without a-Substituents to Aldehydes

The development of asymmetric aldol reactions involving enolates with no a-substituent has been hindered by low asymmetric induction compared with the corresponding propionate aldol reactions. It is postulated that this diﬃculty is because of the lack of stereochemical constraints from the enolate portion of the Zimmerman–Traxler [3] transition-state model. Tab. 2.2

Reaction of aldehyde enolates with ketones. O OSiMe3 C8H17

MeLi

OLi C8H17

Ti(OBu)4

O

OTi(OBu)4 C8H17

R1

R2

OH

H C8H17

14

R2 R1

15

Entry

R1

R2

Yield (%)

Ratioa

1 2 3 4 5

HCcC PhCcC BuCcC Me3 SiCcC CCl3

Me Me Me n-C7 H15 Me

85 86 81 74 72

59:41 52:48 55:45 51:49 >99:1b

a Product b anti:syn

stereochemistry not assigned

2.3 Addition of Enolates Without a-Substituents to Aldehydes Tab. 2.3

Asymmetric acetate aldol reactions. R

O

R

O

OH

R

O

OH

a) BuLi, THF, 0° Ph

N H

16

b) Lewis acid c) PhCHO

Ph

Ph

N H

+

Ph

17

Ph

N H

18

R

Lewis Acid

7:8

Me CH2 OH

Ti(Oi-Pr)3 Cl TiCl 4

83:17 72:28

2.3.1

Stereoselective Acetate Aldol Reactions Using Chiral Auxiliaries

The early work in the ﬁeld of titanium enolate acetate aldol reactions was conducted by Braun in a general investigation of acetate aldol reactions [12]. The enolates were generated from chiral acetamide 16 by transmetalation of the lithium enolate with triisopropoxytitanium chloride or titanium tetrachloride, as shown in Table 2.3. They reported moderate selectivity for the reaction with benzaldehyde. Yan and coworkers developed titanium enolate acetate aldol reactions as an extension of their boron acetate enolate methodology [13, 14]. Good yields and diastereoselectivity were reported when using camphor-derived N-acyloxazolidinethione 19 (Scheme 2.3, Table 2.4, entries 1–7, and Figure 2.1). The high selectivities were attributed to additional chelation aﬀorded by the thiocarbonyl of the chiral auxiliary in transition state assembly 20, shown in Scheme 2.3. The corresponding camphor-derived oxazolidinone acetate imide provided no stereocontrol, supporting the chelation control hypothesis. Shortly thereafter, acetate aldol reactions using camphor-derived imidazolidinone 27 were reported by Palomo and coworkers [15]. They reported moderate yields and enantioselectivity for a variety of unsaturated and aliphatic aldehydes (Table 2.4, entries 8–12). Interestingly, enantioselectivity for unsaturated aldehydes was opposite that for aliphatic aldehydes. Also, enantioselectivity reported for titanium was completely opposite that of the corresponding lithium enolate reactions. Recently, Phillips and Guz reported a titanium enolate acetate aldol reaction based on a valine-derived N-acyloxazolidinethione 30 [16]. The titanium enolate is directly generated from TiCl 4 , ()-sparteine, and Nmethylpyrrolidone by use of Crimmins’ procedure, as discussed in Section 2.4.1.2.1. Their highly hindered chiral auxiliary can be synthesized in three steps from commercially available starting material. They reported good yields and good to excellent diastereoselectivity for a range of aldehydes,

67

68

2 The Development of Titanium Enolate-based Aldol Reactions Tab. 2.4

Survey of acetate aldol reactions employing chiral auxiliaries. O

O

Lewis acid

R1

R2CHO

O

OH

R1

R2

+

R1

24

23

OH R2

25

Entry

R1

R2

Lewis Acid

Yield (%)

24:25

Ref.

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

26 26 26 26 26 26 26 27 27 27 27 27 28 29

n-Pr i-Pr (E)-CH3 CHbCH Ph (E)-PrCHbCH (E)-Me3 SiCHbCH (E)-PhSCHbCH Ph (E)-PhCHbCH i-Pr t-Bu PhCH2 CH2 Ph Ph

TiCl 4 TiCl 4 TiCl 4 TiCl 4 TiCl 4 TiCl 4 TiCl 4 TiCl 4 TiCl 4 TiCl 4 TiCl 4 TiCl 4 Ti(Oi-Pr)3 Cl Ti(Oi-Pr)3 Cl

85 86 86 91 87 80 77 65 70 85 74 37 89 62

95:5 94:6 93:7 91:9 90:10 90:10 90:10 88:12 86:14 17:83 47:53 42:58 35:65 70:30

13 13 13 13 14 14 14 15 15 15 15 15 19 19

as shown in Table 2.5. Once again, the extra chelation aﬀorded by the thiocarbonyl moiety might be responsible for their excellent results. It has, furthermore, been noted that diastereoselectivity depends critically on the exact stoichiometry of the reagents. The stereochemical outcome has been rationalized on the basis of the coordinated chair model 33 or dipole-minimized boat transition model 34, shown in Figure 2.2.

Ph N O

O

N

N S

27

26

O

28

O

29

Fig. 2.1

Chiral auxiliaries used in the asymmetric acetate aldol reactions described in Table 2.4.

2.3 Addition of Enolates Without a-Substituents to Aldehydes

O

O TiCl4, CH2Cl2 i-Pr2NEt N O

S then RCHO

R

O O

H

OH

X

S

N H

O

R

L L

Ti

O

21 +

OH

L R

X

H

20

19

22

Scheme 2.3

Camphor-derived asymmetric acetate aldol reactions.

Urpı´ and Vilarassa investigated an asymmetric acetate aldol reaction in conjunction with the synthesis of Macrolactin A [17]. They reported good yield and excellent stereoselectivity for a handful of unsaturated aldehydes using a valine-derived thiazolidinethione chiral auxiliary. 2.3.2

Stereoselective Acetate Aldol Reactions Involving Chiral Titanium Ligands

Duthaler and coworkers demonstrated that stereoselectivity in the titanium enolate acetate aldol reaction could also be induced by chiral ligands on titanium [18]. Their cyclopentadienylbis(1,2:5,6-di-O-isopropylidene-a-dTab. 2.5

Sterically hindered oxazolidinethione-derived asymmetric acetate aldol reactions. S O

S

O N

TiCl4, (-)-sparteine NMP, CH2Cl2

Ph Ph

O

O

OH

S R

N

30

O

+

Ph

then RCHO

69

O

OH R

N

Ph Ph

Ph

32

31

Entry

R

Yield (%)

31:32

1 2 3 4 5 6 7 8 9 10 11

PhCH2 CH2 n-Pr Et n-Bu (CH3 )2 CHCH2 i-Pr n-C6 H13 (E)-CH3 CHbCH PMBOCH2 TBDPSOCH2 CH2 Ph

83 78 90 77 82 83 78 85 55 56 86

95:5 95:5 93:7 95:5 96:4 92:8 95:5 99:1 97:3 99:1 85:15

70

2 The Development of Titanium Enolate-based Aldol Reactions Ph

Ph Ph

O

H

S Cl Cl

Ph

O S

N

H N

H Ti

Cl

O O

R

H R

O

O

Cl Ti

H

H

34

33

S O

O

Cl Cl

Cl

OH R

N

Ph Ph

35 Fig. 2.2

Transition-state models of the asymmetric acetate aldol reaction.

glucofuranose-3-O-yl)chlorotitanate Lewis acid provided excellent enantioselectivity and moderate to good yields with tert-butyl acetate reacting with a wide range of aliphatic and unsaturated aldehydes (Table 2.6, entries 1–7, and Figure 2.3). The chiral ligands are, furthermore, commercially available and can be recovered after the reaction. This work was extended by Rutledge to an investigation of the eﬀects of the double stereodiﬀerentiation of chiral ligands on titanium and a chiral auxiliary derived from (þ)- or ()-menthol (45 and 46). (Table 2.6, entries 8–11, and Figure 2.4) [19]. Similar work by Fringuielli examined the eﬀects of chiral ligands [20] and auxiliaries [21] on the basis of (þ)-2-carane 43 and (þ)-3-carane 44 (Table 2.6, entries 12–17, and Figure 2.3). Although double stereodiﬀerentiation can result in impressive stereoselectivity, these methods suﬀer from requiring stoichiometric amounts of both chiral auxiliary and chiral ligands and yields are usually modest. 2.3.3

Alternative Approaches to Acetate Aldol Adducts

Because of the inherent diﬃculty of inducing chirality in the acetate enolate reaction, alternative approaches have been developed. A general approach is to synthesize a-substituted aldols and then reductively remove the asubstituent. Yan reported a one-step bromination–aldolization which provided a-bromo aldols in excellent yield and diastereoselectivity (Scheme 2.4) [22]. They then demonstrated that the a-bromo substituent could be reduc-

2.3 Addition of Enolates Without a-Substituents to Aldehydes Tab. 2.6

Stereoselective acetate aldol reactions involving chiral titanium Lewis acids.

37 Ti

OLi

O

Cl

BuLi

OR* OR*

R1

R1

36

Ti O

OR* OR*

R2CHO O

O

OH

R1

R2

+

R1

39

R1

OH

38

R2

40

Entry

R1

R2

OR*

Yield (%)

39:40

Ref.

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17

t-BuO t-BuO t-BuO t-BuO t-BuO t-BuO t-BuO 45 46 45 46 HO HO t-BuO t-BuO 45 47

n-Pr n-C7 H15 i-Pr t-Bu CH2 bCMe Ph 2-furyl Ph Ph Ph Ph Ph Ph Ph Ph Ph Ph

41 41 41 41 41 41 41 41 41 42 42 43 44 43 44 44 44

51 87 66 80 81 69 62 51 54 40 40 54 52 68 70 76 70

97:3 98:2 98:2 96:4 98:2 98:2 95:5 96:4 98:2 30:70 21:79 82:18 83:17 81:19 75:25 90:10 96:4

18 18 18 18 18 18 18 19 19 19 19 20 20 20 20 21 21

tively removed under mild conditions using aluminum amalgam to provide acetate aldols in good yield and enantioselectivity [23]. Urpı´ and co-workers reported aldol-like reactions of titanium enolates with acetals by use of a valine-derived N-acetylthiazolidinethione 52 [24]. They reported moderate to good stereoselectivity and moderate yield for a variety of aliphatic and unsaturated acetals, as shown in Table 2.7. In these reactions, the enolate presumably reacts with the oxycarbenium ion generated by Lewis acid activation of the acetals. Recently, Ghosh and Kim developed an alternative procedure for acetate aldol reactions utilizing ester-derived titanium enolate aldol reactions [25]. As shown in Scheme 2.5, reaction of chloroacetate 55 with TiCl 4 and diisopropylethylamine provided the titanium enolate, which upon reaction with monodentate aldehydes gave a highly anti-diastereoselective aldol product

71

72

2 The Development of Titanium Enolate-based Aldol Reactions

O O

Ph H

O

Ph

OO

Ph

Ph O

O

O O

41

O

42

O

O

H

H H

H

43

44

Fig. 2.3

Chiral ligands used in the stereoselective acetate aldol reactions described in Table 2.6.

56 in excellent yield. Addition of 2 equiv. acetonitrile is critical to the observed diastereoselectivity. Interestingly, when the aldol reaction is conducted with bidentate aldehydes, a variety of anti aldolates are obtained in excellent diastereoselectivity and yield. The reason for this reversal of diastereoselectivity will be discussed in Section 2.4.2.2.1. Reductive removal of chlorine provided convenient access to acetate aldol products. Representative aldol reactions with mono- and bidentate aldehydes are shown in Table 2.8.

2.4

Addition of Enolates with a-Substituents to Aldehydes

As described previously, when enolates have a-substituents, new a- and bchiral centers are formed. Thus, it is possible to form aldol products in which the a- and b-substituents are in a syn or anti relationship to each SiMe3O O H

O

O H

45

46

Fig. 2.4

Chiral auxiliaries used in the stereoselective acetate aldol reactions described in Table 2.6.

47

2.4 Addition of Enolates with a-Substituents to Aldehydes

O

O

b) Br2, i-Pr2NEt

N

R

L

Ti L

Br

O 48

OH

L

O O

H

c) RCHO

S

S

N H

a) TiCl4, i-Pr2NEt

49

O

OH

51 R

OH

a) Et3N, H2O, CH2Cl2

O

R

X

50

Br

b) Al / Hg

R

yield

R

yield

MeCH=CH

87%

MeCH=CH

90%

Ph

96%

Ph

94%

i-Pr

92%

i-Pr

91%

Scheme 2.4

Bromination–aldol reaction and reductive debromination–deprotection.

Tab. 2.7

Acetate aldol adducts from acetals. O

S S

N

O

S 1) TiCl4, i-Pr2NEt

S

N

OMe R

O

S

+

S

N

OMe R

2) RCH(OMe)2, LA

53

52

54

Entry

R

Lewis Acid

Yield (%)

53:54

1 2 3 4 5 6 7 8 9

Ph 4-MeOC6 H4 3-MeOC6 H4 4-ClC6 H4 (E)-PhCHbCH Me n-Pr i-Bu i-Pr

BF3 Et2 O BF3 Et2 O BF3 Et2 O BF3 Et2 O BF3 Et2 O SnCl 4 SnCl 4 SnCl 4 SnCl 4

77 87 75 81 77 57 62 70 60

88:12 93:7 88:12 88:12 82:18 76:24 73:27 79:21 71:29

73

2 The Development of Titanium Enolate-based Aldol Reactions

74

TsNH

TsNH

O

O

HO

RCHO

Cl

55

56

(monodentate)

OH

1) HCO2NH4, Pd/C

O

O

O

OH

TiCl4, i-Pr 2NEt

R

2) LiOH, H2O2

Cl

57

TiCl4, i-Pr 2NEt RCHO (bidentate) TsNH

O

OH

O

OH

1) HCO2NH4, Pd/C O

HO Cl

58

R

2) LiOH, H2O2

59

Scheme 2.5

Asymmetric chloroacetate aldol reactions.

other. Methods have been developed to provide both diastereoselectivity and enantioselectivity in the aldol products. These developments are not mere extensions of previous methods. In fact, many titanium enolate-based asymmetric transformations provide stereoselectivity and eﬃciency which cannot be obtained by use of other methods. 2.4.1

Syn Diastereoselectivity

In general, most enolates are formed in the Z conﬁguration. When the metal involved in the reaction is acidic enough to coordinate to both the enolate and the aldehyde of the aldol reaction, syn adducts result via a sixmembered cyclic Zimmerman–Traxler [3] transition state. This results in inherent preference for titanium enolate aldol reactions to form syn aldols,

Tab. 2.8

Chloroacetate aldol reactions. Entry

Aldehyde

Additive

Yield (%)

58:56

1 2 3 4 5 6 7 8

i-PrCHO i-PrCHO i-PrCHO i-BuCHO i-BuCHO PhCHO BnOCH2 CHO BnO(CH2 )2 CHO

– MeCN (2.2 equiv.) NMP (2.2 equiv.) MeCN (2.2 equiv.) NMP (2.2 equiv.) NMP (2.2 equiv.) – –

82 66 71 88 90 47 86 79

25:75 2:98 6:94 <1:99 <1:99 4:96 >99:1 96:4

2.4 Addition of Enolates with a-Substituents to Aldehydes

and several excellent methods have been developed to that end. One particular important feature of titanium enolate-based syn aldol reactions is the opportunity to develop chelation-controlled reactions. Several very interesting methods have been developed that exploit this unique ability of titanium. In asymmetric syn aldol reactions utilizing chiral oxazolidinone or oxazolidinethiones one can generate either syn aldol diastereomer from the same chiral auxiliary system. 2.4.1.1

Synthesis of syn Aldols in Racemic Form

Reactions of Ketones Aldol reactions of metal enolates containing a-substituents have been investigated extensively. Reetz reported syn-selective aldol reactions of titanium enolates of a variety of cyclic ketones and aldehydes providing diastereoselective syn aldols in the racemic form [26]. As shown in Table 2.9, formation of titanium enolates from ketones was achieved by transmetalation of the corresponding lithium enolates. They reported good to excellent selectivity for reactions of cyclic ketones with aldehydes. Whereas Z enolates generally furnish syn aldol products under kinetic conditions, formation of syn aldols from cyclic enolates seems to be diﬃcult. Interestingly, titanium enolates of cyclic ketones, which can only form Z enolates, provided high syn diastereoselectivity. Good selectivity in the reactions of acyclic ketones 2.4.1.1.1

Tab. 2.9

Aldol reactions of cyclic ketones with aldehydes. O

OTiL 3

OLi Ti(Oi-Pr)3Cl or

LDA (CH2)n

Ti(NEt2)3Br

(CH2)n

(CH2)n

60

RCHO

O

O

OH R

OH R

+

(CH2)n

(CH2)n

61

62

Entry

R

n

Lewis Acid

61:62

1 2 3 4 5 6 7

Ph Ph Ph Ph i-Pr Ph Ph

3 4 4 4 4 5 6

Ti(NEt2 )3 Br Ti(NEt2 )3 Br Ti(NMe2 )3 Br Ti(Oi-Pr)3 Cl Ti(Oi-Pr)3 Cl Ti(NEt2 )3 Br Ti(Oi-Pr)3 Cl

85:15 97:3 92:8 86:15 96:4 90:10 91:9

75

76

2 The Development of Titanium Enolate-based Aldol Reactions a) LDA

O

O

OH

64

b) Ti(Oi-Pr)3Cl R1

R1

R2

c) R2CHO

63

O R1

+

R2

R1

R2

64 : 65

Et

Ph

89 : 11

Et

t-Bu 81 : 19

t-Bu Ph

87 : 13

Ph

87 : 13

Ph

65

OH

Scheme 2.6

Aldol reactions of acyclic ketones with aldehydes.

with aldehydes has been observed, as shown in Scheme 2.6. Interestingly, syn aldol product preference occurred almost independently of enolate conﬁguration; stereochemical rationale has not been forthcoming. The aldol reactions of titanium enolates generated in situ were reported by Harrison [27] to give excellent yield and selectivity for syn aldol products, as shown in Table 2.10. However, methyl ketones tended to eliminate under the reaction conditions and provided a,b-unsaturated ketones. Reactions with propiophenone and benzaldehyde provided excellent yields of aldolates, with syn aldols being the major product (95:5 ratio). The stereochemical outcome was rationalized by Zimmerman–Traxler transition state model 67. Evans used a combination of titanium tetrachloride and diisopropylethylamine, which proved to be eﬃcient and general for many diﬀerent kinds of substrates [28]. The syn aldol diastereoselectivity was comparable with that of boron-mediated processes. Isolated yields with titanium enolates are considerably higher than from boron enolates. Furthermore, syn-selectivity Tab. 2.10

syn-Selective aldol reactions of directly generated titanium enolates.

O Ph

TiCl 4, Et 3N

Ar

H

O

O O

ArCHO H

TiLn Ph

O Ar

Ph

67

66

OH

68

+

OH Ar

Ph

69

Entry

Ar

Yield (%)

68:69

1 2 3 4 5

Ph p-MeC6 H4 p-MeOC6 H4 o-MeOC6 H4 p-NO2 C6 H4

91 94 95 96 98

95:5 96:4 89:11 87:13 87:13

2.4 Addition of Enolates with a-Substituents to Aldehydes O

O

OH 71

OH 72

O

a) TiCl4, i-Pr2NEt R b) i-PrCHO

R

70

i-Pr

R

+

i-Pr

R

Yield

71 : 72

Et

95

98 : 2

i-Pr 95

93 : 7

Scheme 2.7

Aldol reactions of directly generated titanium enolates.

depends on the size of the amine bases. The enolates exist as aggregated complexes, as was evidenced by NMR studies. Representative examples are shown in Scheme 2.7. Reactions of Esters and Thiol Esters Syn-selective aldol reactions involving directly generated thiol ester enolates were reported to give moderate to good selectivity and moderate yield, as shown in Table 2.11 [29]. Modest anti-selectivity was obtained in reactions of 2-pyridylthiopropionate (entry 7), suggesting the 2-pyridyl moiety could be coordinating with titanium. The reactions of a-thio-substituted ester enolates have been reported to give moderate to excellent selectivity and moderate to good yield, as shown in Table 2.12 (entries 1–6) [29]. Interestingly, for a-substituted propionates (entries 4–6), selectivity was good to excellent for the product in which the 2.4.1.1.2

Tab. 2.11

syn-Selective aldol reactions of thiol esters. O

O

O

OH

OH

a) TiCl4, Et3N R1

b) R2CHO

73

R1

R2

+

R1

R2

75

74

Entry

R1

R2

Yield (%)

74:75

1 2 3 4 5 6 7 8 9

PhS PhS PhS t-BuS t-BuS t-BuS 2-pyridyl-S o-MeOC6 H4 S C6 F5 S

Ph n-Pr i-Pr Ph n-Pr i-Pr i-Pr Ph Ph

70 68 65 77 72 75 69 71 40

85:15 78:22 89:11 86:14 69:31 71:29 29:71 87:13 95:5

77

78

2 The Development of Titanium Enolate-based Aldol Reactions Tab. 2.12

syn-Selective aldol reactions with a-heteroatomic substituents. O

O R3

R1

OH

R1

R2

76

R4

O

+

OH

R1

R4

R3 R 2

R3 R2

77

78

Entry

R1

R2

R3

R4

Additive

Yield (%)

77:78

Ref.

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20

EtO EtO EtO EtO EtO EtO MeO MeO MeO MeO MeO EtO EtO Me Me Me Me Me Et i-Pr

n-BuS n-BuS n-BuS n-BuS n-BuS n-BuS PhSe PhSe PhSe PhSe PhSe PhSe PhSe PhSe PhSe PhSe PhSe PhSe PhSe PhSe

H H H Me Me Me H H H H H Me Me H H H H H H H

Ph n-Pr i-Pr Ph n-Pr i-Pr i-Pr Ph PhCH2 CH2 (E)-PhCHbCH n-C5 H11 Ph i-Pr Ph i-Pr PhCHbCH PhCH2 CH2 n-C6 H13 CcC Ph PhCH2 CH2

– – – – – – Ph3 P Ph3 PO Ph3 PO Ph3 PO Ph3 PO Ph3 P Ph3 P – – – – – – –

80 74 73 90 71 76 86 92 92 81 83 93 92 91 75 74 94 97 71 92

94:6 76:24 84:16 95:5 88:12 88:12 90:10 97:3 88:12 >98:2 95:5 >98:2 >98:2 >98:2 >98:2 96:4 93:7 94:6 >98:2 >98:2

29 29 29 29 29 29 30 30 30 30 30 31 31 31 31 31 31 31 31 31

a-sulfur substituent and b-hydroxyl group are syn to each other. This might indicate additional chelation to titanium. Reactions of a-selenoacetate enolates have been reported to give excellent selectivity and yield in the presence of triphenylphosphine or triphenylphosphine oxide, as shown in Table 2.12 (entries 7–11) [30]. This work was extended to a-selenopropionate esters (Table 2.12, entries 12 and 13) and aselenoketones (Table 2.12, entries 14–20) [31]. As for a-sulfur substituents, the a-selenopropionate ester enolates gave excellent yield and selectivity for aldols in which the a-heteroatom and b-hydroxyl group were syn to each other. These papers highlighted the importance of the presence of 1 equiv. bulky phosphine or amine to coordinate to titanium to improve the yield and the selectivity. Aldol Reactions of Aldehyde Hydrazones In contrast with the aldol reactions of ketones or carboxylic acid derivatives, those of aldehyde enolates typically proceed with little to no selectivity. This 2.4.1.1.3

2.4 Addition of Enolates with a-Substituents to Aldehydes Tab. 2.13

syn-Selective aldol-like reactions of hydrazones with aldehydes. Me2NN

LDA

Me2NN

Li Ti(Oi-Pr)3Cl or

Me2NN

Ti(NEt2)3Br

H

H

H

R1

R1

TiL 3

R1 R2CHO

79 Me2NN

Me2NN

OH

H

R2

+

OH

H

R2

R1

R1

80

81

Entry

R1

R2

Lewis Acid

Yield (%)

80:81

1 2 3 4 5 6 7 8 9 10

Me Me Ph Ph Ph Ph Ph i-Pr Me Me

Ph Ph Ph p-NO2 C6 H4 Me Me i-Pr Ph Me t-Bu

Ti(Oi-Pr)3 Cl Ti(NEt2 )3 Br Ti(Oi-Pr)3 Cl Ti(Oi-Pr)3 Cl Ti(Oi-Pr)3 Cl Ti(NEt2 )3 Br Ti(Oi-Pr)3 Cl Ti(Oi-Pr)3 Cl Ti(Oi-Pr)3 Cl Ti(Oi-Pr)3 Cl

80 61 95 40 95 50 78 78 61 70

91:9 85:15 98:2 98:2 98:2 90:10 98:2 94:6 95:5 93:7

could be because of lack of steric bulk to inﬂuence the cyclic six-membered Zimmerman–Traxler transition state. For this reason there are few useful reactions that directly utilize aldehyde enolates, even though the term aldol originally referred to a b-hydroxyaldehyde. This issue was resolved by using titanium enolates generated from N,N-dimethylhydrazones by transmetalation with Ti(Oi-Pr)3 Cl or Ti(NEt2 )3 Br [32]. The reaction furnished racemic syn aldols with excellent selectivity and moderate to excellent yield, as shown in Table 2.13. Reactions with cyclic ketone hydrazones also provided excellent syn diastereoselectivity and yields, as shown in Scheme 2.8.

Me2NN

OH

Me2NN

Ph

>98:2 syn/anti 82

83

Scheme 2.8

syn-Selective aldol-like reaction of cyclic hydrazone.

79

80

2 The Development of Titanium Enolate-based Aldol Reactions

Synthesis of Optically Active syn Aldols Using Chiral Auxiliaries Asymmetric aldol reactions utilizing chiral auxiliaries or templates have emerged as one of the most reliable methods in organic synthesis. Both synand anti-selective aldol reactions have been developed over the years. The ﬁeld of asymmetric syn aldol reactions has been largely advanced by Evans since his development of dibutylboron enolate aldol chemistry based on amino acid-derived chiral oxazolidinones. This method requires expensive dibutylboron triﬂate, however, and the amino acid-derived chiral auxiliary is only readily available in one enantiomer and thus only provides one enantiomer of the syn aldol. Several methods developed on the basis of titanium enolates provide convenient access to both ‘‘Evans’’ and ‘‘non-Evans’’ syn aldol products. 2.4.1.2

Amino Acid-derived Oxazolidinone and Related Auxiliaries Oxazolidinone-, oxazolidinethione-, oxazolidineselone-, and thiazolidinethione-based enolates react with aldehydes via the well-established sixmembered Zimmerman–Traxler [3] chair-like transition state. Exhaustive studies and analysis by Crimmins have established the theoretical basis of these reactions [33]. These transition states can proceed without chelation between carbonyl or thiocarbonyl (84) or with an additional chelation to titanium (85), as shown in Scheme 2.9. To proceed via the chelated transition structure 85, one of the ligands on titanium (typically chloride) must be displaced by the carbonyl or thiocarbonyl group. Although these groups are not suﬃciently nucleophilic to completely displace this ligand on their own, 2.4.1.2.1

Y R1 LnTi

Y X

N H

R3 H

O O

R1

+L

Y

R2

OH R3

N

O O

H

84

O

L

R3

R2

X

X

N H

-L

O

"Evans" syn 86 Scheme 2.9

Asymmetric syn aldol transition-state models.

OH R3

N R2

R2 R1

L

85

X Y

L

Ti

R1

"non-Evans" syn 87

2.4 Addition of Enolates with a-Substituents to Aldehydes O

O

Ti(Oi-Pr) 3

R = i-Pr

PhCHO

N

O

O

O

Ph

N

O

O

O

OH

+

OH Ph

N

O

a) LDA

89

90

b) Ti(Oi-Pr)3Cl

89 : 90

eq. Ti(Oi-Pr)3Cl O

O O

N

3

3 : 92

1

77 : 16

OH

O

R

88 a) TiCl4

R = Bn

O

O

b) base O

N Bn

TiCl 3

O

O i-PrCHO

O

+

N Bn

O

OH

N

O

Bn

91 base

91 : 92

yield

i-Pr2NEt

94 : 6

84

TMEDA

98 : 2

83

92

Scheme 2.10

syn-Selective aldol reactions involving oxazolidinone chiral auxiliaries.

the ligand can be easily abstracted with a second equivalent of titanium. A consequence of this is that these substrates will occasionally give mixtures of products resulting from incomplete conversion of 84 to 85. Also, addition of chelating ligands, extra equivalents of amine bases, or even some solvent molecules tend to disfavor transition state 85, because of their preferential chelation to the titanium enolate. The initial adaptation of amino acid-derived oxazolidinone chiral auxiliaries to titanium enolate aldol reactions was conducted by Thornton [34– 36]. The procedure involved transmetalation of the lithium enolate with Ti(Oi-Pr)3 Cl in THF or ether. By varying the amount of titanium, the ratio of products 89:90 could be varied, as shown in Scheme 2.10. They attributed this observation to quenching of lithium interference, but it could also be a matter of switching between chelated and non-chelated transition structures with excess titanium. They observed that a stoichiometric amount of THF gave rise to titanium chelation [35]. By choosing the more powerfully chelating THF they could favor formation of the syn product 89 whereas the weakly chelating ether favored syn product 90. Because both ethereal solvents can chelate to titanium, however, their products were always complex mixtures of both syn and anti aldols. The ethereal solvents were necessary because of solubility problems. These complications were alleviated when Evans demonstrated that titanium enolates could be directly generated in dichloromethane with amine

81

82

2 The Development of Titanium Enolate-based Aldol Reactions

bases [28]. Stereoselection and reaction yields are comparable with those of boron enolate-based syn aldol reactions. Amine bases seemed to have a noticeable eﬀect on stereoselectivity. As shown in Scheme 2.10, the use of TMEDA improved selectivity compared with i-Pr2 NEt; this suggests the reaction proceeded through non-chelated transition state 84. Furthermore, stoichiometry of aldehydes (2 equiv.) is also critical for complete conversion. Crimmins et al. developed amino acid-derived oxazolidinethione [37] and thiazolidinethione [38] chiral auxiliaries and demonstrated their utility in titanium enolate aldol reactions. Depending upon the amount and nature of amine bases and the stoichiometry of TiCl 4 , ‘‘Evans’’ or ‘‘non-Evans’’ syn aldol products can be provided with excellent diastereoselectivity and isolated yields. Reactions involving oxazolidinethione auxiliaries led to ‘‘Evans’’ syn aldol product 94 when 1.0 equiv. TiCl 4 , 1.1 equiv. aldehyde, and 2.5 equiv. ()-sparteine were used, as shown in Table 2.14, entries 1–3. Rationalizing that the extra amine base or excess aldehyde could be acting as a ligand on titanium, they demonstrated that 1.05 equiv. TiCl 4 , 1.1 equiv. aldehyde, 1.0 equiv. ()-sparteine, and 1.0 equiv. N-methylpyrrolidinone also gave product 94, the latter conditions being more economical and simpler to work up. ()-Sparteine was discovered to have a dramatic rate-enhancement eﬀect on these aldol reactions, but it was demonstrated that its chiral architecture did not lead to signiﬁcant asymmetric induction. Using oxazolidinethione auxiliaries, the chelation-controlled ‘‘non-Evans’’ product 95

Tab. 2.14

Oxazolidinethione based syn aldol reactions. O

O a) 1.0 eq. TiCl4

R

N

S

OH

b) 2.5 eq. (-)-sparteine O

O S

A

c) RCHO

B

a) 2 eq. TiCl4

Bn

94 +

N

O

O Bn

93

S

OH R

N

b) 1.1 eq. i-Pr2NEt Bn

c) RCHO

95

Entry

Method

R

Yield (%)

94:95:anti

1 2 3 4 5 6

A A A B B B

i-Pr Ph MeCHbCH i-Pr Ph MeCHbCH

70 89 65 87 88 81

99:1:0 97:2:1 97:2:1 0:95:5 1:98:1 0:95:5

2.4 Addition of Enolates with a-Substituents to Aldehydes O MeNHOMe

MeO

N

OH R

imidazole O

O S

OH

(90%)

97

R

N

OH Bn

NaBH4

96

HO

R

EtOH (85%)

98

Scheme 2.11

Conversion of oxazolidinethione auxiliary to other functionality.

could be formed by using 2 equiv. TiCl 4 and only 1 equiv. diisopropylethylamine, as shown in Table 2.14, entries 4–6. In a typical procedure, 2 mmol TiCl 4 was added dropwise to a solution of 1 mmol oxazolidinethione 93 in 6 mL CH2 Cl2 at 0 C and stirred for 5 min. i-Pr2 NEt (1.1 mmol) was added dropwise and the dark red solution stirred for 20 min at 0 C and then cooled to 78 C. Aldehyde (1.1 mmol) was added dropwise and stirred for 1 h. One intrinsic feature of oxazolidinethiones is that the chiral auxiliary could be readily converted to Weinreb amide 97 by reaction with imidazole and methoxylamine salt, as shown in Scheme 2.11. Similarly, sodium borohydride reduction gave alcohol 98 and DIBALH reduction provided the corresponding aldehyde in excellent yield. To demonstrate the full utility of the ()-sparteine-mediated enolization, they also showed that oxazolidinone chiral auxiliaries could be reacted with either 1.1 equiv. TiCl 4 and 2.5 equiv. ()-sparteine or 1.05 equiv. TiCl 4 , 1.0 equiv. ()-sparteine, and 1.0 equiv. N-methylpyrrolidinone to yield product 94. Thiazolidinethione-derived chiral auxiliaries have similar reactivity and selectivity, as shown in Table 2.15. Because of the increased nucleophilicity of the thiazolidinethione ring, chelation-controlled reaction through transition state 85 enabled preferential formation of the ‘‘non-Evans’’ syn product 101. One equivalent of i-Pr2 NEt, TMEDA, or ()-sparteine and 1 equiv. TiCl 4 provided 101 diastereoselectively. Interestingly, when the reaction was performed with 2 equiv. of TMEDA or ()-sparteine, ‘‘Evans’’ syn aldol adduct 100 was obtained diastereoselectively. These aldol products can be converted to a variety of other functionality under mild conditions. Other oxazolidinethiones and thiazolidinethiones have resulted in comparable diastereoselectivity and yields. Similar reactions utilizing oxazolidineselone chiral auxiliaries were developed by Silks and coworkers [39]. They demonstrated good to excellent

83

84

2 The Development of Titanium Enolate-based Aldol Reactions Tab. 2.15

Thiazolidinethione-based syn aldol reactions. O

S S

a) 1 eq. TiCl4

S

A

Bn

100 +

c) RCHO

N

O

S

Bn

B

99

R

N

b) 1 eq. (-)-sparteine O

S

OH

a) 1 eq. TiCl4

S

OH R

N

b) 2 eq. (-)-sparteine Bn

c) RCHO

101

Entry

Method

R

Yield (%)

100:101

1 2 3 4

A A B B

i-Pr Ph i-Pr Ph

60 52 75 62

2:98 <1:99 97:3 >99:1

yields and excellent stereoselectivity in reactions of N-propionyl- and Nbenzyloxyacetyloxazolidineselone with a variety of aldehydes, as shown in Table 2.16. The acylated oxazolidineselone can be prepared in a one-pot procedure from oxazoline 102 via lithiation, addition of elemental selenium, and quenching with the appropriate acyl chloride. The selenocarbonyl also has utility as a chiral probe via 77 Se NMR. The selone chiral auxiliary can also be converted to other functionality similar to oxazolidinethione 96 and thiazolidinethione 99. Camphor-derived Chiral Auxiliaries After adaptation of amino acid-derived oxazolidinone chiral auxiliaries to titanium enolate aldol reactions, Thornton and coworkers went on to develop a camphor-derived chiral auxiliary [40]. Moderate to good selectivity was observed for reactions with a variety of aldehydes using camphorquinonederived N-propionyl oxazolidinone 106, as shown in Table 2.17. The carbonyl of the oxazolidinone is not a good enough nucleophile to completely displace chloride to form chelated transition structure 107, so the nonchelated transition state assembly 108 is always a competing pathway. Stereodiﬀerentiation experiments with (R)- and (S)-2-benzyloxypropanal were also investigated. syn Diastereoselectivity was, however, moderate (53% de) compared with lithium enolate-based reactions which provided high selectivity (85% de) for the (R) isomer. Reactions with (S)-benzyloxypropanal resulted in mismatched aldehyde and enolate selectivity and a 42:52 ratio of syn adducts. 2.4.1.2.2

2.4 Addition of Enolates with a-Substituents to Aldehydes

85

Tab. 2.16

Oxazolidineselone-based syn aldol reactions. O

Se

OH R3

N

O

R2 1) LiHMDS N

O

a) TiCl4

O

Se

R1

104 +

b) i-Pr2NEt

2) Se N

O

3) EtCOCl R1

R1

102

O

Se

c) R3CHO

R2

R3

N

O

103

OH

R2 R1

105 Entry

R1

R2

R3

Yield (%)

104:105

1 2 3 4

i-Pr Bn Bn Bn

BnO Me Me Me

i-Pr i-Pr MeCHbCH Ph

72 86 86 91

75:25 >99:1 >99:1 >99:1

Tab. 2.17

Camphor-derived syn aldol reactions.

O

O a) LDA N O O

b) Ti(Oi-Pr)3Cl

O

L L

Ti

c) RCHO L

O O

N H R

O O

R

H -L

106

N H

+L

O

R

109

L

108

O

OH

X

L

Ti

O O

H

107

L

L

+

OH R

X

110

Entry

R

109:110:anti

1 2 3 4

Et Ph i-Pr MeCHbCH

76:13:11 79:2:19 86:2:12 49:4:49

86

2 The Development of Titanium Enolate-based Aldol Reactions Tab. 2.18

Camphor-derived syn-selective aldol reactions.

O

O O

a) LDA

O

L N O

b) Ti(Oi-Pr)3Cl

L

c) RCHO

Ti L

O

N H

H

R

O O

N

+L H

-L

R

O O

H

O

111

O

OH

X

L

Ti L

113

112

O

L

L

R

+

OH

X

R

115

114 Entry

R

Yield (%)

114:115

1 2 3 4

Me i-Pr Ph MeCHbCH

84 70 79 70

>99:1 >99:1 >99:1 99:1

To improve the p-facial and syn selectivity of the camphor-derived chiral auxiliary, Ahn et al. developed a chiral oxazinone derived from ketopinic acid (111) [41]. The explanation for the low diastereoselectivity of camphorderived auxiliary 106 was that the steric inﬂuence of the proximal bridgehead methyl group was insuﬃcient for p-facial selectivity. It seemed that the syn-7-methyl group was too far away from the acyl moiety. In oxazinonebased chiral auxiliary 111 one enolate face is in proximity to the camphor skeleton and, as a result, steric bias was signiﬁcantly enhanced over the 106-derived system. Moderate to good yields and near complete syn diastereoselectivity for a range of aldehydes have been reported, as shown in Table 2.18. The steric inﬂuence of their auxiliary is apparently strong enough to overcome inherent weakness in titanium chelation; near complete selectivity for chelation-controlled product 114 arising from transition state assembly 112 was reported. Yan et al. developed camphor-derived oxazolidinethione chiral auxiliary 116 from ketopinic acid [42, 43]. The opposing location of the N-acyl group in this auxiliary when compared with 106 and 111 leads to formation of the other syn product 120 via chelation control. Good yields and excellent selectivity were reported for a range of aldehydes, as shown in Table 2.19. Be-

2.4 Addition of Enolates with a-Substituents to Aldehydes

87

Tab. 2.19

Camphor-derived syn-selective aldol reactions.

O a) TiCl4 O N

c) RCHO

L

Ti L

S

O

N H

L

L

b) i-Pr2NEt

O S

R

O O

H

116

R +L

H

L

Ti L

118

OH

X

L

O O

117

O

S

N H

-L

O

R

119

+

OH

X

R

120

Entry

R

Yield (%)

119:120

1 2 3

i-Pr Ph MeCHbCH

85 84 86

2:98 3:97 <1:99

cause the thiocarbonyl is a good ligand for titanium, the reaction proceeds through chelated transition state assembly 118. The reactions with the corresponding N-bromoacyl derivatives also provided excellent syn diastereoselectivity and isolated yields [22]. The observed syn stereochemistry is consistent with chelation-controlled model 118. Aminoindanol and Amino Acid-derived Chiral Auxiliaries Ghosh et al. developed highly diastereoselective ester-derived titanium enolate-based syn aldol reactions [44]. Chiral sulfonamide 121 was readily prepared by tosylation of optically active aminoindanol followed by reaction with propionyl chloride in pyridine. As shown in Scheme 2.12, the corresponding titanium enolate of 121 was generated with 1.2 equiv. TiCl 4 and 3.8 equiv. i-Pr2 NEt at 0 C to 23 C for 1 h. These conditions provided the Z enolate which, upon reacting with a variety of bidentate titanium-complexed aldehydes (2 equiv. aldehyde/2.2 equiv. TiCl 4 ), aﬀorded good to excellent yields of syn aldol products 125 with high diastereoselectivity, as shown in Table 2.20. On the other hand, when the reaction was carried out with monodentate aldehydes, anti aldol product 126 was obtained. The stereochemical outcome of these reactions has been rationalized by use of a chelation-controlled model. In this model, the titanium enolate 2.4.1.2.3

88

2 The Development of Titanium Enolate-based Aldol Reactions Ts

TsNH

N

O

Ti

O

a) TiCl4 O

O

b) i-Pr2NEt

122

121 O

( )n

H

O L L

Ti O

R1

O L

O L L Ti L N L

Bn

RCHO

OBn

O L

L R2 R1

O

Ti LL

S O O

124

123

TsNH

O L L Ti L N L O S O

O

OH R2

O

TsNH

O

R2

O

R1

125

OH

R1

126

Scheme 2.12

Aminoindanol-derived asymmetric aldol reactions.

is a seven-membered metallocycle and is assumed to have a chair-like conformation. It has been postulated that anti aldol diastereoselectivity was obtained from monodentate aldehydes via a Zimmerman–Traxler-like model 124 whereas reactions with bidentate aldehydes presumably proceed through 123. The oxyaldehyde side-chain is oriented pseudo-axially for effective metal chelation. As evidenced, this methodology provides excellent diastereoselectivity with a range of esters. The enhanced selectivity for benzyloxyacetaldehyde and benzyloxypropionaldeyde is because of ﬁve- or sixmembered chelation. Reaction of benzyloxybutyraldehyde results in slightly reduced the syn diastereoselectivity. This might be because of less favorable seven-membered chelation. On the basis of the possible transition state assembly 123, Ghosh and coworkers further speculated that the a-chiral center or the indane ring

2.4 Addition of Enolates with a-Substituents to Aldehydes Tab. 2.20

Aminoindanol-based syn aldol reactions. Entry

R1

R2

Yield (%)

125:126

1 2 3 4 5 6 7

Me Me Bn Bn Bn i-Bu i-Bu

BnOCH2 BnO(CH2 )2 BnOCH2 BnO(CH2 )2 BnO(CH2 )3 BnOCH2 BnO(CH2 )2

84 51 84 51 55 83 56

98:2 98:2 99:1 99:1 94:6 99:1 99:1

may not be necessary for syn selectivity for bidentate oxyaldehydes [45]. Indeed, the corresponding phenylalanine-derived chiral auxiliary 127 (R ¼ Bn) resulted in good syn selectivity with diﬀerent bidentate oxyaldehydes, as shown in Table 2.21. They also investigated the eﬀect of the b-chiral substituent and discovered that use of valine-derived chiral auxiliary 127 (R ¼ i-Pr) resulted in good yields and excellent syn selectivity for a wide range of mono- and bidentate aldehydes [46]. Reaction of enolate 124 with a variety of monodentate aldehydes provided aldol products with good to excellent anti diastereoselectivity; this will be Tab. 2.21

Amino acid-derived syn aldol reactions. OL

a) TiCl4

O

b) i-Pr2NEt

TsHN

L

O

O

c) R2CHO, TiCl4

R1

O

Ti

O L N

L R1

Bn

H

O

L

Ti S

L

L p-Tol

O

128

127

O TsHN

OH

O

R2

O TsHN

O

+

R1

OH R2

R1

130

129

Entry

R1

R2

Yield (%)

129:130

1 2 3 4

Bn Bn i-Pr i-Pr

BnOCH2 BnO(CH2 )2 i-Bu PhCHbCH

80 81 93 89

98:2 97:3 95:5 96:4

89

90

2 The Development of Titanium Enolate-based Aldol Reactions Tab. 2.22

Eﬀect of TiCl 4 stoichiometry on diastereoselectivity. TsNH

O

OH Ph

O a) TiCl 4

TsNH

O

132

b) i-Pr2NEt O

+

c) PhCH=CHCHO, TiCl4

131

TsNH O O

OH Ph

133 Entry

TiCl 4 Equiv.

Yield (%)

132:133

1 2 3 4 5 6

0.0 1.0 2.0 3.0 4.0 5.0

6 68 85 94 94 95

0:100 2:98 16:84 75:25 93:7 94:6

discussed in detail in Section 2.4.2.2. Subsequent investigation revealed that the stoichiometry of TiCl 4 required for aldehyde activation is critical to the observed selectivity [47]. As shown in Table 2.22, with increasing quantities of TiCl 4 , the syn:anti product ratio and reaction yields improved. There was, however, a dramatic reversal of diastereoselectivity when 3 equiv. TiCl 4 was used for complexation with cinnamaldehyde (2 equiv.). Reaction of 2 equiv. cinnamaldehyde precomplexed with 5 equiv. TiCl 4 provided diastereoselective syn aldol product 132 in excellent yield. The scope and generality of this methodology were then examined with a variety of aldehydes, as shown in Table 2.23. The stereochemical outcome can be rationalized by using open-chain transition state model 136, which is favored by use of increasing amounts of TiCl 4 and furnishes syn aldol 138 as the major product. One of the most important features of ester-derived titanium enolate aldol reactions is that, depending upon the choice of aldehyde and the stoichiometry of TiCl 4 , one can generate syn or anti aldol adducts diastereoselectively from the same chiral auxiliary. Ready availability of either enantiomer of cis-aminoindanol also provides convenient access to both diastereomers of syn or anti aldols in optically active form. Other Chiral Auxiliaries Other chiral auxiliaries used in syn aldol reactions are illustrated in Figure 2.5. Xiang et al. developed N-tosylnorephedrine-based chiral auxiliary 139 2.4.1.2.4

2.4 Addition of Enolates with a-Substituents to Aldehydes

91

Tab. 2.23

syn-Selective aminoindanol-derived aldol reactions. O a) TiCl4

TsNH

LnTi

L R2

O

O L L Ti L N L O S O

O

Ti

L

R1

LL

b) i-Pr2NEt O c) RCHO, TiCl4

Ti Ln

O

H

CH 3

H3C

H

O

O Ti Ln N

134

Ts

136

135

TsNH

TsNH

O

O

OH R2 +

O

OH R2

O R1

R1

138

137 Entry

R

Yield (%)

137:138

1 2 3 4 5

Et i-Pr t-Bu Ph Me 2 CbCH

80 82 65 91 65

13:87 13:87 1:99 7:93 11:89

[48]. It was shown that 139-based esters could enolize in the presence of TiCl 4 and amine bases. Although the auxiliary favored syn products, the reactions yielded complex mixtures of both syn and anti products. Ahn and coworkers developed stilbenediamine-derived 1,2,5-thiadiazolidine-1,1-dioxide-based chiral auxiliary 140 and demonstrated its utility in syn aldol reactions [49]. Excellent selectivity and yield were observed for a variety of aldehydes, as shown in Table 2.24, entries 1–3. The cyclic sulfamide auxiliary is novel in that it is bifunctional and C2 -symmetric, so only

Ph Ts

O

NH

O

Ph

N

O

N S

Ph

O O

139

140

Fig. 2.5

Other chiral auxiliaries used in syn aldol reactions.

Ph

N N

141

92

2 The Development of Titanium Enolate-based Aldol Reactions Tab. 2.24

syn Aldol reactions utilizing other chiral auxiliaries. O

O RCHO

X

R

X

142

O

OH

+

OH

X

R

144

143

Entry

X

R

Yield (%)

143:144

Ref.

1 2 3 4 5

140 140 140 141 141

Ph i-Pr (E)-MeCHbCH p-MeOC6 H4 t-Bu

91 93 89 93 70

96:4 97:3 95:5 96:4 75:25

49 49 49 50 50

half an equivalent of chiral material is needed. The ﬁrst aldol fragment can be removed by simple treatment with sodium methoxide in THF. To remove the second aldol fragment it was necessary to protect the free sulfamide with a Boc group. The second aldol product could then be removed by a method analogous to the ﬁrst. The favored product 143 is derived from chelation control, because of the excellent chelating ability of the sulfone in the chiral auxiliary. Hitchcock and coworkers developed ephedrine-derived 3,4,5,6-tetrahydro2H-1,3,4-oxadiazin-2-one-based chiral auxiliary 141 and demonstrated its utility in aldol reactions [50]. It was discovered that the aldehyde had to be present during enolization for reaction to occur, because of diﬃculties in enolization. Use of aromatic aldehydes resulted in good yield and good to excellent selectivity, as did aliphatic aldehydes without a-hydrogen atoms. This method is not useful for aldehydes bearing a-hydrogen atoms, because of self-condensation. Synthesis of Optically Active syn Aldols Using Chiral Titanium Ligands Duthaler and coworkers used carbohydrate–titanium complexes for synthesis of optically active syn-b-hydroxy-a-amino acids [51]. These syn-aaminoaldols were obtained in moderate yield and excellent syn diastereoselectivity, as shown in Table 2.25. Transmetalation of the lithium enolate of glycine ester derivative 145 with chiral titanium complex 146 provided a titanium enolate which upon reaction with a wide variety of aldehydes provided syn-b-hydroxy-a-amino esters 148. Subsequent hydrolysis and Nprotection gave a-aminoaldols 149. Duthaler and coworkers also reported asymmetric syn aldol methodology based on their titanium complex 146 [52]. Heathcock demonstrated the capacity of 2,6-dimethylphenyl propionate-derived lithium enolates to undergo addition to a range of aldehydes aﬀording racemic anti aldol adduct 151 2.4.1.3

2.4 Addition of Enolates with a-Substituents to Aldehydes Tab. 2.25

syn Aldol reactions of glycine derivatives. OLi

O

Ti LDA

EtO

EtO Si

Si

Ti O

OR* OR*

OR*

N

N Si

146 OR*

Cl

EtO

Si

N Si

Si

145 R1CHO

OH

O

R1

EtO

OH

O protection

NHR2

NH2

149

O O

H

R1

EtO

148

O O

R* = O O

147 Entry

R1

R2

Yield (%)

de (%)

1 2 3 4

Me CH2 bCMe Ph t-BuO2 C

Boc CHO Boc CHO

53 61 60 66

>98 99 >96 >96

with excellent diastereoselectivity. The E enolate presumably reacted with aldehydes via a cyclic transition state to form anti aldol adducts 151. Transmetalation of the lithium enolate of 150 with 146 and reaction with a variety of aldehydes provided syn aldols in moderate to good yield and excellent stereoselectivity, as shown in Table 2.26. It was observed that the kinetically generated E enolates were responsible for the observed stereoselectivity. Equilibration to the more stable Z-enolates gave anti products, as discussed in Section 2.4.2.3. To rationalize the stereochemical outcome of these syn aldol products, boat-like transition state 152 was proposed. There are also other examples of syn aldol product formation from titanium E-enolates [53]. Mahrwald reported aldol reactions of ketone enolates with aldehydes in which the reaction was conducted with equimolar amounts of titanium(IV) alkoxides and a-hydroxy acids [54]. This provided aldol products with high syn diastereoselectivity, as shown in Table 2.27. Among a variety of alkoxides and a-hydroxy acids examined, the use of Ti(Ot-Bu)2 -BINOL and (R)mandelic acid resulted in high syn diastereoselectivity and aldol products were obtained in enantiomerically enriched form. Other a-hydroxy acids such as tartaric acid or lactic acid resulted in very low enantioselectivity (18–24%). The inﬂuence of the chirality of BINOL

93

2 The Development of Titanium Enolate-based Aldol Reactions

94

Tab. 2.26

syn Aldol reactions using chiral titanium enolates. O

OLi

LDA

O

RCHO O

O

O

OH R

151

150 146 H H

TiLn

R

Ti

O

RCHO

O

O

OR* OR*

H 3C

O O

152

O

OH

O

R

153 Entry

R

Yield (%)

153:151

1 2 3 4

i-Pr CH2 bCMe Ph Pr

76 61 82 87

94:6 96:4 96:4 92:8

Tab. 2.27

syn-Selective aldol reactions. O RCHO

O

OH

BINOL=Ti(Ot-Bu)2

+

R (R)-mandelic acid

154 Entry

R

Yield (%)

syn:anti

ee (syn) (%)

1 2 3 4

Ph t-Bu PhCcC i-Pr

71 55 72 48

95:5 85:15 73:27 89:11

93 87 94 83

2.4 Addition of Enolates with a-Substituents to Aldehydes

95

Tab. 2.28

syn-Selective aldol reactions of chiral a-silyloxyketones. O

OH

R1 O

R1 H

H

R2

O O

155

OTBS

L

L

R1 OTBS

R2

H

TBSO

157

L

Ti

O L

+

OH

R1

R2 OTBS

156

158 Entry

R1

R2

Methoda

Yield (%)

157:158

Ref.

1 2 3 4 5 6

Cyclohexyl Cyclohexyl Cyclohexyl Cyclohexyl Me Bn

Ph i-Pr Ph i-Pr i-Pr i-Pr

A A B B C C

– – – – 90 85

99:1 >99:1 99:1 99:1 97:3 97:3

55 55 56 56 57 57

a Method

A: LDA, Ti(Oi-Pr)3 Cl, R2 CHO; Method B: LDA, Ti(Oi-Pr)4 , R2 CHO; Method C: TiCl 4 , i-Pr2 NEt, R2 CHO

was insigniﬁcant. Although the mechanism of this reaction is unclear, ligand exchange between t-BuOH and a-hydroxy acids is evident from NMR analysis and might be necessary for the syn diastereoselectivity and enantioselectivity observed. Synthesis of Optically Active syn Aldols with Chiral Enolates Thornton and Siegel have reported that reactions of the titanium enolates of chiral a-silyloxyketones resulted in excellent syn diastereoselectivity, as shown in Table 2.28, entries 1–4 [55, 56]. Use of tetrakisisopropoxytitanium enolates also aﬀorded excellent syn diastereoselectivity. In work similar to that with oxazolidinone chiral auxiliaries, these enolates were generated as lithium enolates then transmetalated with the appropriate titanium Lewis acid. Large excesses of titanium were necessary for good stereoselectivity. They also noted that carboxylic acid derivatives could be obtained by deprotection and oxidative cleavage. Chiral auxiliary methods are, however, more eﬃcient at providing adducts of this nature. Urpı´ and coworkers demonstrated that a variety of directly generated titanium enolates of a-silyloxyketones reacted with aldehydes to give syn aldols with excellent yield and selectivity, as shown in Table 2.28, entries 5–6 [57]. The selectivity of these reactions can be explained by the transition state assembly 156. Between the O-benzyl derivative and the OTBS protected ketones, the latter provided excellent yield and syn–syn diastereoselectivity. 2.4.1.4

2 The Development of Titanium Enolate-based Aldol Reactions

96

Tab. 2.29

Reactions of b-ketoimide enolates. O

O

O

OH

X O

Me O

O

O

O

H

b) i-Pr2NEt O

N

R

a) TiCl4

c) RCHO H

Bn

159

R Me

O O

Bn

L

H

N

161 L

Ti

O

O

+

O

OH

L O

160

R

N Bn

162 Entry

R

Yield (%)

161:162

1 2 3 4

i-Pr CH2 bCMe Et Ph

86 64 86 81

>99:1 95:5 >99:1 96:4

The steric bulk of the silyloxy group prohibits it from eﬀectively chelating titanium, so the products arise from the non-chelated transition state. Much work in the ﬁeld of aldol reactions of ketones was performed by Evans to enable the synthesis of polypropionate natural products. They demonstrated that b-ketoimides like 159 were selectively and completely enolized at the C4 position rather than the potentially labile methyl-bearing C2 position, most probably because steric factors prohibited alignment of the carbonyl groups necessary to activate the C2 proton. As shown in Table 2.29, it was demonstrated that these compounds would react with aldehydes to provide syn–syn product 161, via titanium enolates, with good yield and excellent selectivity, and the corresponding syn–anti product 162 could be favored by use of a tin enolate reaction [58]. They invoked the chelated transition state assembly 160 to explain the product stereochemistry observed, in which the C2 -methyl group directs diastereofacial selectivity. Interestingly, reduction with Zn(BH4 )2 provided the syn diol diastereoselectively. Evans also investigated the stereochemical inﬂuence of two adjacent stereogenic centers in the titanium enolate-based aldol reaction [28]. As shown in Scheme 2.13, asymmetric induction of the enolate resulted from the inﬂuence of the a-stereocenter; the b-stereogenic center has very little eﬀect. Subsequently, Evans examined double-stereodiﬀerentiating titanium enolate aldol reactions [59]. Both the aldehyde and the enolate contained a and b stereogenic centers. In aldol reactions between these substrates the enolate can adopt either a matched or mismatched relationship with the aldehyde. Several possible scenarios were investigated. As shown in Scheme

2.4 Addition of Enolates with a-Substituents to Aldehydes OTBS O

OTBS O

Ti Ln

OTBS O

i-PrCHO

a) TiCl4

97

OH

b) i-Pr2NEt

163

164 96 : 4 dr (96%)

OTBS O

OTBS O a) TiCl4

Ti Ln

OTBS O

i-PrCHO

OH

b) i-Pr2NEt

166 95 : 5 dr (82%)

165 Scheme 2.13

syn-Selective aldol reactions of chiral ketones.

2.14, when the ketone enolate and aldehyde were chirally fully matched (168) or partially mismatched (169 and 170), syn diastereoselectivity was quite good. The diastereoselectivity was poor only in the completely mismatched case (171). O

OPMB OTBS O

OTBS O

OH

OPMB

H

167

O

a) TiCl 4

OPMB

168

89 : 11 dr (86%)

OTBS O

OH

OPMB

H

b) i-Pr2NEt OTBS O

TiLn O

OPMB

169

87 : 13 dr (81%)

OTBS O

OH

H

170 O

OPMB

92 : 8 dr (85%)

OPMB OTBS O

OH

OPMB

H

171 Scheme 2.14

Double stereodiﬀerentiation in syn aldol reaction. ( a 28% of an anti diastereomer also isolated.)

37 : 35 : 28 dra (79%)

98

2 The Development of Titanium Enolate-based Aldol Reactions

2.4.2

Anti Diastereoselectivity

As can be seen from the developments described above, the control of both relative and absolute acyclic stereochemistry in a variety of syn aldol reactions can now be achieved highly stereoselectively. Both boron and titanium enolate-based syn aldol reactions have gained widespread popularity and are frequently used in synthesis. Whereas anti-a-alkyl-b-hydroxycarbonyl units are inherent to numerous bioactive natural products, there are relatively few eﬀective synthetic processes that are convenient, operationally simple, and aﬀord high diastereoselectivity for a wide range of aldehydes. Early examples of anti-selective aldol reactions, reported by Meyers in 1984, were based upon oxazoline-derived boron enolates [60]. Several other methods based upon metal enolates other than titanium have subsequently been developed. In this chapter, however, we will focus on titanium enolate-based methods. Synthesis of anti Aldols in Racemic Form Procter et al. reported a novel anti-selective aldol process using titanium enolates of N-propionylpyrrolidine [61]. The aldol products were obtained in good yield and anti diastereoselectivity. The aldolates were produced in racemic form, however. As shown in Table 2.30, formation of the titanium enolate of 172 was achieved by transmetalation of the lithium enolate with Cp2 TiCl2 in THF. The resulting titanium enolate reacted with a range of aldehydes. Interestingly, the lithium enolate of 172 has stereoselectivity and the corresponding zirconium enolate aﬀorded syn diastereoselectivity. 2.4.2.1

Tab. 2.30

anti-Selective aldol reactions of N-propionylpyrrolidine. O

OH

N

R

OTiCp 2Cl

O a) LDA N

N

173 RCHO

+

b) Cp2TiCl 2

O

172

OH

N

R

174 Entry

R

Yield (%)

173:174

1 2 3 4

Et i-Pr Ph MeCHbCH

65 64 68 77

79:21 87:13 98:2 90:10

2.4 Addition of Enolates with a-Substituents to Aldehydes Tab. 2.31

anti Aldol reactions of phenyl benzyloxythioacetate with respective aldehydes. OH

O H Bn OBn

177

O Ti O

L

175

OBn

SPh

O

PhS

L

R

PhS

R

O

+ OH

O

L

176

R

PhS OBn

178

Entry

R

Yield (%)

177:178

1 2 3 4

Me n-Pr t-Bu Ph

99 97 81 99

97:3 98:2 98:2 98:2

Cinquini and Cozzi reported the synthesis of racemic anti-a,b-dihydroxy thiolesters by aldol reaction of aldehydes with a-benzyloxythiolester-derived chlorotitanium enolates [62]. The thiolester was enolized by treatment with TiCl 4 and Et3 N in CH2 Cl2 at 78 C and the resulting enolate was reacted with a variety of aldehydes at 78 C providing anti aldol products with good to excellent anti diastereoselectivity, as shown in Table 2.31. The stereochemical outcome can be explained by boat-like model 176 in which the thiolester formed an E enolate presumably by chelation through the benzyloxy group. Kazmaier reported the synthesis of several racemic a-aminoaldols in which the a-amino group and the b-hydroxyl group are anti [63]. Reactions of N-benzyloxycarbonyl amino acid ester enolates with a variety of aldehydes aﬀorded moderate to good yields and good to excellent diastereoselectivity, as shown in Table 2.32. Titanium enolates were generated by deprotonation of 179 with LDA in THF followed by the addition of 2 equiv. Ti(Oi-Pr)3 Cl in THF at 78 C. 2.4.2.2

Synthesis of Optically Active anti Aldols by Use of Chiral Auxiliaries

Aminoindanol and Related Chiral Auxiliaries Ghosh and coworkers developed novel highly diastereoselective anti aldol methods based on ester-derived titanium enolate aldol reactions [64–66]. The utility of the cis-1-toluenesulfonamido-2-indanol-derived chiral auxiliary was demonstrated. This chiral auxiliary is readily prepared from commercially available enantiomerically pure cis-1-amino-2-indanol. Both enantio2.4.2.2.1

99

100

2 The Development of Titanium Enolate-based Aldol Reactions Tab. 2.32

Synthesis of sterically demanding a-aminoaldols. a) LDA

O

O t-BuO

OH

O

OH

b) Ti(Oi-Pr)3Cl

NHCO2Bn

t-BuO c) R2CHO

R1

R2 R1

t-BuO

NHCO2Bn

R2 R1

180

179

NHCO 2Bn

181

Entry

R1

R2

Yield (%)

180:181

1 2 3 4

i-Pr i-Pr Me Et

t-Bu i-Pr i-Pr i-Pr

70 87 87 85

97:3 92:8 92:8 95:5

mers of the chiral auxiliary are, furthermore, readily available, enabling synthesis of either enantiomer of the anti aldol. As shown in Scheme 2.15, ester 183 was readily enolized with titanium tetrachloride and diisopropylethylamine. First it was treated with TiCl 4 in CH2 Cl2 at 0 C. After stirring at 23 C for 15 min, i-Pr2 NEt was added. Evans reported that alkyl esters cannot be enolized with TiCl 4 and amine bases [28]. Smooth enolization of sulfonamidoesters is presumably because of internal chelation with the sulfonamido group, as was recently documented by Xiang et al. [48].

TsNH

NH2

O 1) TsCl OH

2) EtCOCl

182

O

183 a) TiCl4 b) i-Pr2NEt

H3 C L L

L O Ti L O

H H O H

N

L L

L O

S

O

CH3

L O Ti

O H

N S

O

184 Scheme 2.15

Enolization of N-tosylaminoindanol esters.

185

2.4 Addition of Enolates with a-Substituents to Aldehydes

101

TsNH L L O

L

184

Ti

Ti

L

L O

i-BuCHO

O

TiCl 4

L

N

O

O

OH

O

OH

O

L

187

L

+

S TsNH

O

O

186 188 Scheme 2.16

Aminoindanol-derived anti aldol reaction.

Although treatment of enolate 184 with isovaleraldehyde produced no reaction, addition of the enolate to a solution of isovaleraldehyde precomplexed with TiCl 4 provided the anti aldol product 187 in 97% yield as a single isomer (by HPLC and 1 H NMR) as shown in Scheme 2.16. Reaction with several other aldehydes also produced the anti isomer as the major product. It should be noted that of four possible diastereomers, only one syn and one anti product were observed in this reaction. The anti:syn selectivity ranged from 85:15 to >99:1 and yields ranged from 41–97%, as shown in Table 2.33. Whereas benzaldehyde resulted in very little to no selectivity under these conditions, in the presence of an additive there was a dramatic improvement in anti diastereoselectivity. Thus, aldol reaction of the enolate of 189 with benzaldehyde (2 equiv.) precomplexed with TiCl 4 (2.2 equiv.) in the presence of CH3 CN additive (2.2 equiv.) aﬀorded anti alTab. 2.33

anti-Selective asymmetric aldol reactions. TsNH O

a) TiCl4

TsNH

TsNH

O

O

OH

OH

b) i-Pr2NEt O

189

c) RCHO TiCl4

R +

O

190

O

191

Entry

R

Yield (%)

190:191

1 2 3 4 5 6 7

i-Bu Me Pr i-Pr MeCHbCH PhCHbCH Ph

97 50 74 91 41 63 85

>99:1 85:15 95:5 85:15 95:5 99:1 45:55

R

102

2 The Development of Titanium Enolate-based Aldol Reactions a) TiCl4

TsNH

TsNH

O

O

OH

b) i-Pr2NEt O

TiCl 4

192

a) TiCl4

TsNH

O

c) i-BuCHO

193 (20.5 : 5.8 : 1 dr) TsNH O

O

OH

b) i-Pr2NEt O

O

c) i-BuCHO TiCl4

194

a) TiCl4

MsNH

195 (1 : 1 dr) MsNH O

O

OH

b) i-Pr2NEt O

196

c) i-BuCHO TiCl4

O

197 (70 : 30 dr)

Scheme 2.17

Structure–reactivity relationship studies in anti aldol reactions.

dol product stereoselectively (96:4) in good yield. These conditions, however, did not improve the yield or selectivity for reactions with aliphatic aldehydes. The anti selectivity can be rationalized by using a novel Zimmerman– Traxler type transition state 186. In this transition state, both the methyl group of the enolate and the alkyl group of the aldehyde adopt pseudoequatorial positions, leading to the anti product. Several structural features have been shown to be critical for the high selectivities observed. First, reaction of dimethylphenylglycinol-derived auxiliary 192 with isovaleraldehyde provided a 20.5:5.8:1 ratio of isomers, as shown in Scheme 2.17. The major isomers are thought to be the two anti diastereomers. Removal of the indane aromatic ring resulted in complete loss of diastereoselectivity. Thus, reaction of the corresponding 1,2aminocyclopentanol derivative 194 with isovaleraldehyde resulted in a 1:1 mixture of syn and anti diastereomers. The use of the mesylate derivative 196 also resulted in a large loss of selectivity. The reaction of N-mesylaminoindanol derivative 196 with isovaleraldehyde provided a 70:30 mixture of anti and syn diastereomers. These results suggest a possible p-stacking interaction between the two aromatic rings. The result of this interaction would be to help stabilize the conformation shown in enolate 184, although this interaction has not been rigorously established. Double stereodiﬀerentiation experiments with matched chiral aldehyde 199 provided anti aldol 200 as a single diastereomer, as shown in Scheme 2.18 [67].

2.4 Addition of Enolates with a-Substituents to Aldehydes a) TiCl 4

TsNH O

TsNH

O

OH

b) i-Pr2NEt O

O c) TiCl4, O

198

200 (>99 : 1 dr)

H

199 Scheme 2.18

Double stereodiﬀerentiation in aminoindanol-derived anti aldol reaction.

On the basis of these structure–reactivity studies, Ghosh and Kim subsequently speculated that the planarity of the acenaphthene ring in conjunction with its aromaticity might further enhance p-stacking interactions with the arylsulfonamide functionality [68]. In this context, an eﬀective synthesis of both enantiomers of cis-2-amino-1-acenaphthenol from acenaphthylene was developed. Enolization of 202 using TiCl 4 and i-Pr2 NEt provided a single enolate, presumably the Z enolate. Reaction of this enolate with isovaleraldehyde precomplexed with TiCl 4 resulted in signiﬁcantly reduced anti diastereoselectivity (78:22), however. Similar to the aminoindanol chiral auxiliary, reactions with the acenaphthene-derived auxiliary produced only one anti and one syn diastereomer. Interestingly, aldol reaction of the enolate of 202 with aldehydes (2 equiv.) precomplexed with TiCl 4 (2.2 equiv.) in the presence of CH3 CN (2.2 equiv.) resulted in excellent anti diastereoselectivity and yield for a wide range of aldehydes, as shown in Table 2.34. Oxazolidinethione and Oxazolidineselone Chiral Auxiliaries Silks and coworkers reported anti-selective aldol reactions in conjunction with their investigations on chiral selenium compounds [39]. They discovered that bidentate oxyaldehydes precomplexed with TiCl 4 led to anti aldol products in good yield and with excellent stereoselectivity, as shown in Table 2.35, entries 1 and 2. Presumably, this chelation-controlled reversal of syn/anti selectivity is a phenomenon similar to that which Ghosh and coworkers exploited to produce syn aldols from aminoindanol chiral auxiliaries. Both benzyl and isopropyl oxazolidineselone chiral auxiliaries have comparable anti diastereoselectivity. Crimmins and McDougall demonstrated that protected glycolyloxazolidinethiones underwent aldol reactions with aldehydes to provide anti aldol products [69]. Moderate to good yields and good to excellent anti selectivity were observed for a wide range of reactions, as shown in Table 2.35, entries 3–8. To achieve anti selectivity, it is necessary to complex the aldehydes with TiCl 4 before addition of the enolate. Comparing their results with those of Silks, it seems that the a-oxy substituent on the enolate is a less powerful anti director than an oxygen substituent on the aldehyde. 2.4.2.2.2

103

104

2 The Development of Titanium Enolate-based Aldol Reactions Tab. 2.34

Acenaphthene-derived anti aldol reactions with respective aldehydes. TsNH OH

O TsNH O

a) TiCl 4

R

O

b) i-Pr2NEt

203

O c) RCHO, TiCl4, TsNH CH3CN

201

OH

O

202

R

O

204 Entry

R

Yield (%)

203:204

1 2 3 4 5 6

i-Bu Et i-Pr cyclohexyl Ph PhCH2 CH2

95 92 95 84 93 97

97:3 92:8 96:4 99:1 93:7 95:5

Synthesis of Optically Active anti Aldols by Use of Chiral Titanium Ligands Duthaler reported the synthesis of optically active anti aldols by use of the chiral CpTi(DAGO)2 complex [70]. They reported moderate yields and 2.4.2.3

Tab. 2.35

anti-Selective aldol reactions involving oxazolidinethione and oxazolidineselone auxiliaries. X O

O

X R2CHO

N

O

OH

O

R2 + O

N

R1

X

OH

O

R2

N R1

R1 Bn

Bn

Bn

205

R2 + O

N

R1

Bn

X

OH

O

206

207

208

Entry

X

R1

R2

Yield (%)

206:207:208

Ref.

1 2 3 4 5 6 7 8

Se Se S S S S S S

Me Me OCH2 CHbCH2 OCH2 CHbCH2 OCH2 CHbCH2 OBn OBn OMe

BnOCH2 BuOCH2 Me i-Pr Ph CH3 (CH2 )4 CHbCH2 Me

100 81 84 61 56 64 48 62

>99:0:1 >99:0:1 94:6:0 87:13:0 65:24:11 88:12:0 74:26:0 86:4:10

39 39 69 69 69 69 69 69

2.5 Natural Product Synthesis via Titanium Enolate Aldol Reactions Tab. 2.36

Stereoselective synthesis of anti aldols using chiral titanium Lewis acids. O

a) LDA

O

O

OH

O

b) CpTi(DAGO) 2Cl O

R

+

O

OH R

c) RCHO

209

211

210

Entry

R2

Yield (%)

210:211

1 2 3 4

Pr t-Bu CH2 bCMe Ph

74 59 50 73

89:11 83:17 54:46 23:77

good diastereoselectivity for a range of aldehydes when reacted with 2,6dimethylphenyl propionate derived titanium enolate, as shown in Table 2.36, entries 1 and 2. The method provided no selectivity (entry 3) or syn selectivity (entry 4), however, when the aldehyde was unsaturated and substituted. The corresponding N-propionyl-1,3-oxazolidin-2-one provided good selectivity with isobutyraldehyde (88:12 anti/syn) but poor yield, because of the formation of byproducts.

2.5

Natural Product Synthesis via Titanium Enolate Aldol Reactions

Numerous applications of titanium enolate aldol reactions have been reported for the synthesis of natural products and bioactive compounds of pharmaceutical interest. These syntheses were performed by utilizing the titanium enolate methods described above as key steps. There are many transformations that simply cannot be achieved by use of conventional aldol chemistry. Furthermore, the use of inexpensive titanium reagents, operational simplicity, and functional group tolerance make this chemistry very attractive in synthesis. The following applications in synthesis further exemplify the increasing importance of titanium aldol chemistry. 2.5.1

Lactone Natural Products

Asymmetric syn and anti aldol reactions have been used for synthesis of the following natural products containing lactone moieties. A titanium enolatebased aldol reaction has also been utilized in the aldol dehydration sequence during the synthesis of pyrone natural products mxyopyronin A and B. The following syntheses of lactone-containing natural products highlight the variety of aldol transformations.

105

106

2 The Development of Titanium Enolate-based Aldol Reactions

Tetrahydrolipstatin Ghosh and Fidanze reported a synthesis of tetrahydrolipstatin (217) in which the anti stereochemistry of the b-lactone moiety is set by using asymmetric anti aldol methodology, discussed in Section 2.4.2.2.1 [71]. Tetrahydrolipstatin is a saturated analog of the natural product lipstatin, isolated from Streptomyces toxytricini. It is a potent and irreversible inhibitor of pancreatic lipase and has been marketed in several countries as an anti-obesity agent under the trade name Xenical. Aldol reaction between the enolate of 212, generated with TiCl 4 and i-Pr2 NEt, and cinnamaldehyde precomplexed with Bu2 BOTf gave 213 in 60% yield as a 6.1:1 mixture of anti and syn diastereomers, as shown in Scheme 2.19. Aldol adduct 213 was converted to the benzyl ester 214, which was reduced selectively to the anti 1,3-diol by use of Evans’ procedure [72] and protected as the triisopropylsilyl ether 215. Hydrogenation of the O-benzyl group over Pearlman’s catalyst and lactonization with phenylsulfonyl chloride in pyridine gave the b-lactone 216 which, on deprotection and reaction with N-formyl-l-leucine under Mitsunobu conditions, aﬀorded ()-tetrahydrolipstatin. 2.5.1.1

Myxopyronins A and B Panek and coworkers developed syntheses of myxopyronin A (220a) and myxopyronin B (220b) in which an aldol reaction was used to append the 2.5.1.2

TsNH

TsNH

O

O

OH

i

TiCl 4, Pr2NEt O

O

212

C6H13

Bu2BOTf, PhCH=CHCHO

Ph C6H13

213

(60%) ref. 71

OH

O

OTIPS C11H23

BnO C6H13

215

OH

O

C11H23

BnO 2) TIPSOTf, 2,6-lutidine (96%)

C6H13

1) H2/Pd(OH)2 (99%)

O

O OTIPS

1) Bu4NF(70%)

O

O

2) PPh3, DIAD, N-formyl-L-leucine

216 Scheme 2.19

Synthesis of ()-tetrahydrolipstatin.

(90%)

214

H N

H

2) PhSO2Cl, pyr (74%) O

O

1) Me4NHB(OAc)3 (94%)

217

O

O

2.5 Natural Product Synthesis via Titanium Enolate Aldol Reactions O

OH

O

O

CO2Me

218 TiCl4, ,i-Pr 2NEt,

CHO

R

R = Pr : 58% R = Bu: 61% O

OH

O

O

R CO 2Me

219 ref. 73

O

OH

O

O

R

H N

CO 2Me

220a: R = Pr 220b: R = Bu Scheme 2.20

Myxopyronin A and B synthesis.

hydrophobic side chain [73]. The myxopyronins are natural products isolated from the gliding bacterium Myxococcus fulvus MX f50. These molecules are bacterial growth inhibitors, because of their capacity to selectively inhibit bacterial RNA polymerase. Aldol condensation of ketone 218-derived titanium enolate and subsequent elimination provided diene 219, as shown in Scheme 2.20. Attachment of the other side chain provided myxopyronins A and B. Callystatin A Crimmins and King reported a total synthesis of callystatin A (228) using their aldol methodology (as discussed in Section 2.4.1.2.1) to construct three of the four chiral centers in the molecule [74]. Callystatin A is a natural product, isolated from the marine sponge Callyspongia truncata, which has potent in-vitro toxicity against KB cell lines (IC50 ¼ 0.01 ng mL1 ). Phenylalanine-derived N-propionyloxazolidethione 221 was subjected to enolization with TiCl 4 and ()-sparteine using Crimmins’ procedure and reacted with (S)-2-methylbutanal to provide syn aldol adduct 222 in 83% yield with 98% selectivity, as shown in Scheme 2.21. Protection of the alco2.5.1.3

107

108

2 The Development of Titanium Enolate-based Aldol Reactions

Bn

1) TBSOTf

Bn TiCl4, (-)-sparteine

O

N

2) LiBH4 O

H

N 3) Swern ox.

O

S

221

H

S

O

OH

O

OTBS

(83%)

223

222

O (83%, 98:2)

TiCl4,(-)-sparteine

Oi-Pr

221 O

(81%, 98:2)

226

KOt-Bu

Bn

ref. 74

CHO BrBu3P

O TMSO

(90%) Oi-Pr

OTBS

N S

O

OH

OTBS

224

225

O

TMSO O

OTBS 1) PPTS, H2O

227

2) TPAP O

3) HF-pyr (43%)

228

O

OH

Scheme 2.21

Synthesis of ()-callystatin A.

hol as the TBS ether, reductive removal of the chiral auxiliary, and Swern oxidation furnished aldehyde 223. Aldol reaction under similar conditions with oxazolidethione 221 gave the all-syn product 224 in 81% yield with 98% selectivity; this was then converted to 225. Wittig oleﬁnation with fragment 226 (synthesized from (S)-glycidol) gave oleﬁn 227 which, on removal of the OTMS and isopropyl acetal protecting groups, perruthenate oxidation, and deprotection of the OTBS group gave ()-callystatin A. AI-77-B Ghosh and coworkers developed a convergent synthesis of ()-AI-77-B (237) in which all ﬁve stereogenic centers were set by asymmetric synthesis [75]. 2.5.1.4

2.5 Natural Product Synthesis via Titanium Enolate Aldol Reactions

109

As in their previous work [76], four of those chiral centers were set using the aminoindanol-based syn and anti aldol methods discussed in Sections 2.4.1.2.3 and 2.4.2.2.1, respectively. AI-77-B is a natural product isolated from Bacillus pumilus with potent antiulcerogenic and anti-inﬂammatory properties. Anti aldol reaction of 229-derived titanium enolate with 4,4,4trichlorobutryaldehyde gave adduct 230 in 90% yield with 90% de; this was subsequently transformed to isocoumarin fragment 231, as shown in Scheme 2.22. Aldol reaction of 232-derived titanium enolate with benzyloxyacetaldehyde

OMe O TsNH

TsNH

O

O

OH

TiCl4, i-Pr2NEt O

O

OH

TiCl4, i-Pr2NEt O

231

230 TsNH

O

OBn OBn

O

ref. 75 HO 2C

BnOCH2CHO

232

Boc

ref. 75

(90%, 90%de)

TsNH

H N

O

Cl3(CH2)2CHO

229

O

CCl3

233

(97%)

NBoc O

234 a) TFA

b) 231, EDCI, DMAP (72%) OMe O

OMe O O

OBn

H N

CO 2Bn

NBoc

2) H2 / Pd-C DOWEX 50-X8 (75%)

O O

OH

H N O

Scheme 2.22

AI-77B synthesis.

NH2 CO 2H

237

OH

3) CsCO3, BnBr (90%)

OBn

H N

NBoc O

O

235

236 1) MgI2 (93%)

OH

O 2) NaClO2 (99%)

O

O

1) O3, PPh3 (98%)

110

2 The Development of Titanium Enolate-based Aldol Reactions

provided syn adduct 233 in 97% yield as a single product; this was transformed into carboxylic acid 234 via Curtius rearrangement, stereoselective homologation after Dondoni’s procedure [77] and functional group manipulation. Deprotection of 231 and coupling with EDCI and DMAP gave amide 235. The allyl group was converted to benzyl ester 236 which, on demethylation with MgI2 and concomitant hydrogenolysis of the O-benzyl groups and isopropylidine deprotection, gave AI-77B. Key steps included stereoselective ester-derived asymmetric syn and anti aldol reactions, a regioselective Diels–Alder reaction, and Dondoni homologation. 2.5.2

Macrolide Natural Products

Titanium enolate aldol reactions have been used in the syntheses of several of important antitumor macrolides and of the immunosuppressive macrolide rapamycin. Duthaler’s titanium enolate acetate aldol reaction was used in the synthesis of epothilone 490. A variety of titanium enolate-based syn aldol reactions highlight the scope and utility of this technique. Epothilone 490 Danishefsky and coworkers reported a concise, convergent synthesis of epothilone 490 (243) with a key late-stage acetate aldol reaction [78]. Epothilone 490 is a natural product isolated from the cellulose-degrading bacterium Sorangium cellulosum which has potent Taxol-like microtubule-stabilizing induced cancer cytotoxicity. Known vinyl iodide 238 [79] was subjected to Stille coupling and acetylation to give acetate ester 239. Reaction with previously synthesized aldehyde 240 [80] using to Duthaler’s titanium enolate aldol method (discussed in Section 2.3.2) gave aldol 241 in 85% yield as a single diastereomer (Scheme 2.23). Deprotection of the OTroc group using zinc in acetic acid followed by ring closing metathesis using the Grubbs second generation catalyst 242 yielded epothilone 490 in 64% yield. 2.5.2.1

Cryptophycin B Ghosh and Bischoﬀ reported an eﬃcient and convergent synthesis of cryptophycin B (249) using their aminoindanol-derived syn aldol methodology (discussed in Section 2.4.1.2.3) [81]. Cryptophycin B (249), a marine natural product isolated from Nostoc sp. GSV 224, has potency at the picogram level against KB cells. Unsaturated ester 244-derived titanium enolate was subjected to an aldol reaction with 3-benzyloxypropionaldehyde to give syn product 245 in 98% yield as a single diastereomer, as shown in Scheme 2.24. Reductive removal of the chiral auxiliary then deoxygenation of the primary alcohol and protection of the remaining secondary alcohol furnished silyl ether 246. Conversion of the benzyloxy functionality to the a,bunsaturated ester moiety was accomplished by selective OBn deprotection, PCC oxidation, and Horner–Emmons oleﬁnation to yield ester 247 which 2.5.2.2

2.5 Natural Product Synthesis via Titanium Enolate Aldol Reactions 1) Pd2dba3, PPh3 S

S O

CH2=CHSnBu3 (96%)

OTES N

111

N O

2) Bu4NF (92%)

238

239

3) Ac2O, Et 3N, DMAP (98%) I

a) LDA, Et2O b) CpTi(OR)2Cl c) O

O

OTroc

H

240

(85%) O

O

S

OH

1) Zn, AcOH (86%) 2) MesN

N

Cl O OH

243

Cl

(64%) NMes Ru PCy3

O S

OH N

242 O

Ph

OTroc

241

Scheme 2.23

Epothilone 490 synthesis.

was converted to compound 249. Exposure to triﬂuoroacetic acid to remove the N-Boc and O-tert-butyl protecting groups, then Yamaguchi cycloamidation gave the macrocycle which, on selective epoxidation with dimethyldioxirane, gave cryptophycin B in 22% overall yield over fourteen steps. Amphidinolide T1 Ghosh and Liu reported the ﬁrst total synthesis of amphidinolide T1 (258), setting four of the seven stereogenic centers with their asymmetric aldol methodology (discussed in Section 2.4.1.2.3) [82]. Amphidinolide T1 is a marine natural product isolated from Amphidinium sp. with signiﬁcant antitumor properties against a variety of cell lines. Aldol reaction of ester 250 with 3-benzyloxypropionaldehyde gave syn adduct 251 in 90% yield as a single diastereomer, as shown in Scheme 2.25. Ester 251 was then converted into tetrahydrofuran derivative 252 via Wittig oleﬁnation and oleﬁn cross-metathesis. Aldol reaction of ester 253 with benzyloxyacetaldehyde gave exclusively syn aldol adduct 254 in 93% yield; this was transformed into alkene 255. Subsequent transformation yielded tetrahydrofuran derivative 256. This cyclic bromoether serves as masked functionality for the labile exocyclic 2.5.2.3

O

2 The Development of Titanium Enolate-based Aldol Reactions

112 TsNH

TsNH

O

OH

O TiCl4, i-Pr2NEt

O 98%

244

OBn

O

BnO(CH2)2CHO

245

Ph

Ph

1) LiAlH4 (92%) 2) PhLi, TsCl, LiAlH4 (96%) 3) TIPSOTf, 2,6-lutidine (99%) 1) K2CO3, BBr3 (83%) 2) PCC (98%) Ph

CO 2Et

3)

OTIPS

O

(EtO) 3PC

247

OBn

Ph OTIPS

CO 2Et

246

NaH (92%) ref. 81

Ph

O

Ph O

O

1) TFA

O

t-BuO HN

O

O

2) Yamaguchi

HN

O

O O

O

248

3) O OMe

O

(87%, 3:1)

HN

O

O

N H

O

OMe

249

Boc

Scheme 2.24

Cryptophycin B synthesis.

methylene group in the ﬁnal product. Stereoselective oxocarbenium ionmediated alkylation using a modiﬁcation of Ley’s procedure [83] with DTBMP and AlCl3 gave coupled product 257 in 73% yield as a single diastereomer. Deprotection of the alcohol and ester moieties followed by Yamaguchi macrolactonization, and treatment with zinc and ammonium chloride aﬀorded amphidinolide T1. Key steps included stereoselective aminoindanol-derived asymmetric aldol reactions, eﬃcient oleﬁn crossmetathesis, stereoselective oxocarbenium ion-mediated anomeric alkylation, and the use of the cyclic bromomethyl ether as a novel exo-methylene group surrogate. Rapamycin Danishefsky and coworkers published a total synthesis of rapamycin (260) using a novel aldol macrocyclization reaction as the key step [84]. Rapamycin is a natural product with immunosuppresive properties. The conclusion 2.5.2.4

2.5 Natural Product Synthesis via Titanium Enolate Aldol Reactions

SO2Ph

TsNH

TsNH

O

O

OH

TiCl4, i-Pr2NEt BnO(CH2)2CHO (90%)

250

O

O O

TiCl4, i-Pr2NEt

OTBS

OBn

OH ref. 82 BnO

255

254

(93%)

CO2Bn

252

O

BnOCH2CHO

253

BnO

251 TsNH

TsNH

O

ref. 82

O

O

113

OTIPS

ref. 81 Br Br

DTBMP O O

O 1) HF-pyr (87%)

TBSO

OTBS

AlCl3, 252 O

(73%) OTBS

CO2Bn 257

2) H2/Pd-C 3) Yamaguchi (71%) OH 4) Zn, NH4Cl (61%) O

O O

O

258 Scheme 2.25

Amphidinolide T1 synthesis.

of the synthesis treats late-stage intermediate aldehyde 259 with isopropoxytitanium trichloride in the presence of triethylamine to generate the cyclized product in 11% yield, with 22% of another isomer (possibly the syn aldol product), as shown in Scheme 2.26. Deprotection of the TIPS ether aﬀorded rapamycin. Spongistatins 1 and 2 Crimmins and coworkers published a convergent synthesis of spongistatin 1 (263a) and spongistatin 2 (263b), employing their phenylalanine-derived asymmetric aldol methodology (discussed in Section 2.4.1.2.1) to set one of the stereocenters [85]. Spongistatins 1 and 2 are natural products with subnanomolar growth inhibition of several NCI chemoresistant tumor types 2.5.2.5

256

114

2 The Development of Titanium Enolate-based Aldol Reactions OMe

OMe OTIPS

OH

H

N O

O H O

HO

H O

1) Ti(Oi-Pr)Cl3,

CHO

O MeO

N O

Et3N (11%) 2) HF-pyr (85%)

O

OMe

O

O

HO O

H

O H

O

O

OH MeO

O

OMe

H

259

260

Scheme 2.26

Rapamycin synthesis.

including human melanoma, lung, brain, and colon cancers. Aldol reaction of aldehyde 261 with N-propionyloxazolidinethione then reductive removal of the chiral auxiliary furnished diol 262 in 74% yield as a 96:4 ratio of stereoisomers, as shown in Scheme 2.27. This fragment was then incorporated into synthetic spongistatins 1 and 2. 2.5.3

Miscellaneous Natural Products

Several other natural products have been synthesized by using titanium enolate-based aldol methods. Many of these syntheses utilize ketone enolate aldol reactions to establish syn stereochemistry. Duthaler’s anti aldol reaction was used in the synthesis of tautomycin. Use of Evan’s ketone–aldol reaction was nicely exempliﬁed in syntheses of denticulatin B and membrenone C. Tautomycin Chamberlin and coworkers reported a convergent synthesis of tautomycin (270), employing Duthaler’s anti aldol methodology, as discussed in Section 2.4.2.3, to set four of its stereogenic centers [86]. Tautomycin, a serine/ threonine selective protein pyrophosphatase inhibitor is selective for PP1 over PP2A. Reaction of aldehyde 265 with the enolate of 264, generated with LDA and transmetalated with Duthaler’s reagent (R ¼ diacetoneglucose) aﬀorded compound 266 in 80% yield as an 8:1 mixture of anti and syn aldols, as shown in Scheme 2.28. Protection of the alcohol as the silyl ether and conversion of the ester gave iodide 267, which was converted to spirocyclic aldehyde 268. Aldehyde 268 which was reacted with the titanium 2.5.3.1

2.5 Natural Product Synthesis via Titanium Enolate Aldol Reactions Bn N

O BnO

BnO O

O S 1) TiCl4, (-)-sparteine

O

CHO

TESO

O

O

TESO

2) LiBH4

OH

(74% 96:4)

OH

OTES

OTES

262

261

ref. 85 OH HO HO H

O

H

H

O

O HO

O OH O

X

O H

H O

OH

OMe

O

O

H

AcO OAc OH

263a: X = CI 263b: X = H Scheme 2.27

Spongistatin 1 and 2 synthesis.

enolate of ester 264 using Duthaler’s procedure to give anti aldol adduct 269 in 67% yield as a 7:1 mixture of the anti and syn isomers. Reaction of 269 provided synthetic tautomycin. Crocacin C Chakraborty and coworkers reported a synthesis of crocacin C (276) using asymmetric aldol methodology developed by Crimmins (Section 2.4.1.2.1) to set two of the four chiral centers [87]. Crocacin C is a natural product isolated from myxobacterium Chondromyces crocatus with potent growth inhibition of Gram-positive bacteria, fungi, and yeasts. Titanium enolate aldol reaction of phenylalanine-derived N-propionyloxazolidinethione 271 with cinnamaldehyde gave the syn aldol adduct 272 in 89% yield as a single dia2.5.3.2

115

116

2 The Development of Titanium Enolate-based Aldol Reactions O

O

LDA, CpTi(OR)2Cl

O

O

OHC

266

265

264

OH

(80%, 8:1)

1) TBSOTf, 2,6-lutidine (94%) 2) DIBALH 3) I2, PPh3 (85%) OTBS

O

OHC H

ref. 86

H

I

O

268

267

LDA, CpTi(OR)2Cl

264 (67%, 7:1) O

OH

O

O H

H

O

269 ref. 86 O

OH

O

O O

O

OH

OMe

O

OH

O H

O

H O

270

Scheme 2.28

Tautomycin synthesis.

stereomer; this was converted to allylic alcohol 273, as shown in Scheme 2.29. Sharpless asymmetric epoxidation provided epoxide 274 which was opened regio- and stereoselectively with lithium dimethylcuprate to provide diol 275 in 86% yield. Diol 275 was then transformed to the ﬁnal product crocacin C. Stigmatellin A Enders and coworkers developed a synthesis stigmatellin A (281) using aldol reactions to set two of the four chiral centers [88]. Stigmatellin A is a natural product isolated from bacterium Stigmatella aurantica and is one 2.5.3.3

2.5 Natural Product Synthesis via Titanium Enolate Aldol Reactions Bn

117

Bn TiCl4, i-Pr 2NEt

N O

Ph

O

N

PhCH=CHCHO

S

OH

(89%)

O

O S

272

271

ref. 87 Ti(Oi-Pr)4, (-)-DIPT, Ph

OH OMe

t-BuOOH

O

Ph

OH OMe

(93%)

274

273

Me2CuLi (86%)

Ph

OH OMe OH

275

ref. 87

Ph

CONH2 OMe OH

276

Scheme 2.29

Crocacin C synthesis.

of the most powerful electron-transport inhibitors in chloroplasts and mitochondria. Hydrazone 277 was alkylated with 3-( p-methoxyphenoxy)propyl iodide using their previously developed SAMP/RAMP methodology to provide hydrazone 278 in 80% yield with 98% de, as shown in Scheme 2.30 [89]. The chiral auxiliary was removed and the resulting ketone was subjected to an aldol reaction with benzyloxyacetaldehyde to give syn aldol adduct 279 in 64% yield as a 2:1 mixture of isomers. Ketone 279 was subjected to selective anti reduction using Evans’ procedure [72] and methylated to yield compound 280, which upon removal of the p-methoxyphenyl and benzyloxy groups gave stigmatellin A. Denticulatin B Paterson and Perkins developed a synthesis of denticulatin B (285) and its isomer denticulatin A using a late-stage aldol reaction [90]. Denticulatin B is a marine natural product isolated from the mollusk Siphoneria denticulata. Aldol coupling of ketone 282 and aldehyde 283 provided the syn aldol adduct 284 in 90% yield as a mixture of diastereomers, as shown in Scheme 2.31. Swern oxidation of the hydroxyl groups and deprotection and cyclization with HF–pyridine aﬀorded denticulatin B in 20% overall yield over nine steps. 2.5.3.4

118

2 The Development of Titanium Enolate-based Aldol Reactions OMe N

OMe LDA

N

PMPO(CH2)3I

N PMP O

N

(80%, 98%de)

277

278

1) MMPP (87%) 2) TiCl4, i-PrNEt2 BnOCH2CHO (64%, 2:1)

OMe OMe

PMP O

O

1) Me4NHB(OAc)3 PMP O OBn (99%)

OH OBn

2) KH, MeI (77%)

280

279 ref. 88

OMe O OMe OMe MeO

O OH

281 Scheme 2.30

Stigmatellin A synthesis.

TiCl4, i-Pr2NEt

O

O

Si

t-Bu

O

H

OH

OH

283

t-Bu

282

O

284

O (90%, 83 : 17)

O t-Bu

Si

O t-Bu

1) Swern (100%) 2) HF-pyr (57%) OH H O O

O

285 OH Scheme 2.31

Denticulatin B synthesis.

OH

2.5 Natural Product Synthesis via Titanium Enolate Aldol Reactions TiCl 4, i-Pr2NEt BnO

OBn

BnO H

O

OBn O

286

O

287

OH

288 a) c-Hex 2BCl, Et3N

(70%, 95%de)

b) LiBH4 c) H2O2, pH 7 (88%, 95%de)

1) t-Bu2Si(OTf)2 2,6-lutidine (72%) H

H O

O

Si

t-Bu

O

O

t-Bu

290

2) H2/Pd-C (85%)

BnO

3) PCC (61%)

OBn OH

OH

289

TiCl 4, i-Pr2NEt 3-pentanone (90%, 90%de) O

1) Swern (100%) O 2) HF-pyr O

OH

O t-Bu

Si

O

OH

291 t-Bu

O

O

O

3) TFA (52%)

292

Scheme 2.32

Membrenone C synthesis.

Membrenone C Perkins and Sampson developed syntheses of membrenone C (292) and its isomers to establish the absolute conﬁguration of the natural product [91]. Membrenone C is a natural product isolated from the skin of a Mediterranean mollusk. Aldol reaction of the stereochemically matched chiral ketone 286 and protected (R)-3-hydroxy-2-methylpropionaldehyde 287 gave aldol adduct 288 in 70% yield and 95% de, as shown in Scheme 2.32. Synselective reduction of the ketone was accomplished by using a modiﬁcation of Narasaka’s procedure [92] to give diol 289 in 88% yield and 95% de. Protection of the diol as the cyclic silyl acetal followed by hydrogenolysis of the O-benzyl groups and PCC oxidation aﬀorded dialdehyde 290. Double aldol reaction with 3-pentanone gave compound 291 with nine contiguous chiral centers in 90% yield and 90% de. The synthesis was completed by Swern oxidation of the free hydroxyl groups, removal of the silyl protecting group, and cyclization in the presence of triﬂuoroacetic acid. 2.5.3.5

119

120

2 The Development of Titanium Enolate-based Aldol Reactions

2.6

Typical Experimental Procedures for Generation of Titanium Enolates 2.6.1

Experimental Procedures Titanium Enolate Formation by Transmetalation, Synthesis of syn Aldols (Thornton’s Procedure, Section 2.4.1.2.1). The lithium enolate was generated at 78 C with LDA in ether. Ti(Oi-Pr)3 Cl (1–3 equiv.) was added dropwise with stirring. The clear solution became brown–orange and was left to warm to 40 C over 1 h and then cooled to 78 C. Aldehyde (1.1 equiv.) was added rapidly by syringe and the reaction was left to warm to 40 C over 3 h. The reaction was quenched with saturated NH4 F and the layers were separated. The aqueous layer was extracted three times with ether. All organic layers were combined and dried over MgSO4 , followed by vacuum ﬁltration and rotary evaporation to dryness. Puriﬁcation by ﬂash chromatography provided the major product. Ester-derived Titanium Enolate by Transmetalation, Synthesis of syn Aldols (Duthaler’s Procedure, Section 2.4.1.2.2). BuLi (6.2 mmol) is added at 20 C to a solution of i-Pr2 NH (1 mL, 7.07 mmol) in 30 mL of ether under argon. After 15 min the reaction mixture is cooled to 78 C and a solution of 2,6-dimethylphenyl propionate (1.0 g, 5.61 mmol) in 10 mL ether is added dropwise and stirred for 1.5 h. An ethereal solution of chlorocyclopentadienylbis(1,2:5,6-di-O-isopropylidene-a-d-glucofuranos-3O-yl)titanium (0.088 m, 80 mL; 7.04 mmol, 1.25 equiv.) is added carefully via a cannula under argon pressure. After stirring for 24 h the aldehyde (7.29 mmol, 1.3 equiv.) is added and the reaction is monitored by TLC. The reaction mixture is quenched with 2 g NH4 Cl and 10 mL 1:1 THFaH2 O. After stirring for 2 h at 0 C the precipitated titanium salts are separated by ﬁltration and washed with ether. The ﬁltrate is washed with 20 mL 1 m HCl, 10 mL satd NaHCO3 , and brine. The aqueous washings are re-extracted with 2 50 mL EtOAc. The combined organic extracts are dried over MgSO4 and the solvent is removed. The crude products are either separated directly by chromatography or by ﬁrst stirring for 1 h with 200 mL 0.1 m HCl, extracted with ether (3 100 mL), and washed with sat. NaHCO3 and brine, removing glucose as the water soluble 1,2-acetonide. Ester-derived Titanium Enolate for anti Aldol Reactions (Ghosh’s Procedure, Section 2.4.2.2.1). N-tosylaminoindanol ester was reacted with TiCl 4 in CH2 Cl2 at 0–23 C for 15 min followed by addition of i-Pr2 NEt (4 equiv.) at 23 C and stirring of the resulting brown solution for 2 h. The titanium enolate was then added to the representative aldehyde (2 equiv.) precomplexed with TiCl 4 (2.2 equiv.) at 78 C and the mixture was stirred at 78 C for 2 h before quenching with aqueous NH4 Cl. The aqueous layer

2.7 Conclusion

was extracted with CH2 Cl2 . The combined organic extracts were washed with brine, dried over Na2 SO4 , and concentrated under reduced pressure. The crude products were puriﬁed by ﬂash chromatography on silica gel. Oxazolidinethione-derived Titanium Enolate for syn Aldol Reactions (Crimmins’s Procedure, Section 2.4.1.2.1). To a dry round-bottomed ﬂask under nitrogen was added 0.250 g (1.0 mmol) oxazolidinethione in 6 mL CH2 Cl2 . The solution was cooled to 0 C and TiCl 4 (1.05 mmol, 0.115 mL) was added dropwise and the solution was stirred for 5 min. To the yellow slurry or suspension was added ()-sparteine (2.5 mmol). The dark red enolate was stirred for 20 min at 0 C. Freshly distilled aldehyde (1.1 mmol) was added dropwise and the reaction stirred for 1 h at 0 C. The reaction was quenched with half-saturated NH4 Cl and the layers were separated. The organic layer was dried over Na2 SO4 , ﬁltered, and concentrated. Puriﬁcation of the crude material by column chromatography aﬀorded the main diastereomer. 2.6.2

Alternative Approaches to Titanium Enolate Generation

Although the direct generation of titanium enolates is typically the most useful method of generating titanium enolates for aldol reactions, other methods have been described. Grubbs and Stille reported that titanium enolates could be generated by reaction of biscyclopentadienyltitanium alkylidene complexes and acyl halides [93]. Oshima and coworkers reported the formation of titanium enolates from a-iodoketones with allylsilane and titanium tetrachloride [94]. Mukaiyama and coworkers reported the generation of titanium enolates from a-bromoketones on treatment with TiCl2 and copper powder [95].

2.7

Conclusion

Titanium enolate aldol reactions have been shown to be very eﬀective for control of relative and absolute stereochemistry in acetate aldol and both syn and anti aldol reactions. The use of readily available and inexpensive titanium reagents make these methods convenient for large-scale synthesis. The synthetic potential of a variety of aldol reactions has been demonstrated by highlighting the synthesis of numerous bioactive complex natural products. The signiﬁcance of enantio- and diastereoselection in synthesis, particularly in this pharmaceutical age, ensures that titanium enolate aldol reactions will remain an important part of organic synthesis for years to come. There is no doubt that unprecedented success has been achieved in the development of a variety of titanium enolate aldol reactions in the past decade. Much new potential and other exciting possibilities remain to be

121

122

2 The Development of Titanium Enolate-based Aldol Reactions

explored, however. We hope this chapter will stimulate further research and developments in titanium enolate-based aldol reactions and their use in organic synthesis. References 1 Lim, B.-M.; Williams, S. F.; Masamune, S. Comprehensive

2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19

20 21 22

23

Organic Synthesis; Trost, B. M.; Fleming, I. Eds; Pergamon Press: Oxford, 1991, Vol. 2, Heathcock, C. H., ed., Chapter 1.7, 239. The term ‘chiral auxiliary’ was ﬁrst introduced by Eliel. Eliel, E. L.; Tetrahedron 1974, 30, 1503 and references cited therein. Zimmerman, H. E.; Traxler, M. D. J. Am. Chem. Soc. 1957, 79, 1920. Masamune, S.; Bates, G. S.; Corcoran, J. W. Angew. Chem. Int. Ed. Engl. 1977, 16, 585. Westley, J. W. Adv. Appl. Microbiol. 1977, 22, 177. Masamune, S.; Choy, W.; Petereson, J. S.; Sita, L. R. Angew. Chem. Int. Ed. Engl. 1985, 24, 1. Verhe´, R.; De Kimpe, N.; De Buyck, L.; Thierie, R.; Schamp, N. Bull. Soc. Chim. Belg. 1980, 89, 563. Yoshida, Y.; Hayashi, R.; Sumihara, H.; Tanabe, Y. Tetrahedron Lett. 1997, 38, 8727. Yoshida, Y.; Matsumoto, N.; Hamasaki, R.; Tanabe, Y. Tetrahedron Lett. 1999, 40, 4227. Tanabe, Y.; Matsumoto, N.; Funakoshi, S.; Manta, N. Synlett 2001, 12, 1959. Yachi, K.; Shinokubo, H.; Oshima, K. J. Am. Chem. Soc. 1999, 121, 9465. Devant, R.; Braun, M. Chem. Ber. 1986, 119, 2191. Yan, T.; Hung, A.; Lee, H.; Chang, C.; Liu, W. J. Org. Chem. 1995, 60, 3301. Yan, T.; Hung, A.; Lee, H.; Liu, W.; Chang, C. J. Chinese Chem. Soc. 1995, 42, 691. Palomo, C.; Oiarbide, M.; Gonza´les, A.; Garcı´a, J.; Berre´e, F.; Linden, A. Tetrahedron Lett. 1996, 37, 6931. Phillips, A. J.; Guz, N. R. Org. Lett. 2002, 4, 2253. Gonza´lez, A´.; Aiguade´, J.; Urpı´, F.; Vilarrasa, J. Tetrahedron Lett. 1996, 37, 8949. Duthaler, R. O.; Herold, P.; Lottenbach, W.; Oertle, K.; Riediker, M. Angew. Chem. Int. Ed. Engl. 1989, 28, 495. Cambie, R. C.; Coddington, J. M.; Milbank, J. B. J.; Pausler, M. G.; Rustenhover, J. J.; Rutledge, P. S.; Shaw, G. L.; Sinkovich, P. I. Australian J. Chem. 1993, 46, 583. Fringuelli, F.; Martinetti, E.; Piermatti, O.; Pizzo, F. Gazz. Chim. Ital. 1993, 123, 637. Fringuelli, F.; Piermatte, O.; Pizzo, F.; Scappini, A. M. Gazz. Chim. Ital. 1995, 125, 195. (a) Wang, Y.; Su, D.; Lin, C.; Tseng, H.; Li, C.; Yan, T. Tetrahedron Lett. 1999, 40, 3577. (b) Wang, Y.; Su, D.; Lin, C.; Tseng, H.; Li, C.; Yan, T. J. Org. Chem. 1999, 64, 6495. Wang, Y.; Yan, T. J. Org. Chem. 2000, 65, 6752.

References 24 Cosp, A.; Romea, P.; Urpı´, F.; Vilarrasa, J. Tetrahedron Lett.

2001, 42, 4629. 25 Ghosh, A. K.; Kim, J. Manuscript in preparation, Univ. of.

Illinois–Chicago. 26 Reetz, M. T.; Peter, R. Tetrahedron Lett. 1981, 22, 4691. 27 Harrison, C. Tetrahedron Lett. 1987, 28, 4135. 28 Evans, D. A.; Rieger, D. L.; Bilodeau, M. T.; Urpı´, F. J. Am.

Chem. Soc. 1991, 113, 1047. 29 Annunziata, R.; Cinquini, M.; Cozzi, F.; Cozzi, P. G.;

Coslandi, E. Tetrahedron 1991, 47, 7897. 30 Toru, T.; Hayakawa, T.; Nishi, T.; Watanabe, Y.; Ueno, Y.

Phosphorus, Sulphur, and Silicon 1998, 136–138, 653. 31 Nakamura, S.; Hayakawa, T.; Nishi, T.; Watanabe, Y.; Toru,

T. Tetrahedron 2001, 57, 6703. 32 Reetz, M. T.; Steinbach, R.; Kesseler, K. Angew. Chem. Int.

Ed. Engl. 1982, 21, 864. 33 Crimmins, M. T.; King, B. W.; Tabet, E. A.; Chaudhary, K.

J. Org. Chem. 2001, 66, 894. 34 Nerz-Stormes, M.; Thornton, E. R. Tetrahedron Lett. 1986,

27, 897. 35 Shirodkar, S.; Nerz-Stormes, M.; Thornton, E. R.

Tetrahedron Lett. 1990, 31, 4699. 36 Nerz-Stormes, M.; Thornton, E. R. J. Org. Chem. 1991, 56,

2489. 37 Crimmins, M. T.; King, B. W.; Tabet, E. A. J. Am. Chem. Soc.

1997, 119, 7883. 38 Crimmins, M. T.; Chaudhary, K. Org. Lett. 2000, 2, 775. 39 Li, Z.; Wu, R.; Michalczyk, R.; Dunlap, R. B.; Odom, J. D.;

Silks, L. A. III. J. Am. Chem. Soc. 2000, 122, 386. 40 Bonner, M. P.; Thornton, E. R. J. Am. Chem. Soc. 1991, 113,

1299. 41 Ahn, K. H.; Lee, S.; Lim, A. J. Org. Chem. 1992, 57, 5065. 42 Yan, T.; Lee, H.; Tan, C. Tetrahedron Lett. 1993, 34, 3559. 43 Yan, T.; Tan, C.; Lee, H.; Lo, H.; Huang, T. J. Am. Chem.

Soc. 1993, 115, 2613. 44 Ghosh, A. K.; Fidanze, S.; Onishi, M.; Hussain, K. A.

Tetrahedron Lett. 1997, 38, 7171. 45 Ghosh, A. K.; Kim, J. Tetrahedron Lett. 2001, 42, 1227. 46 Ghosh, A. K.; Kim, J. Tetrahedron Lett. 2002, 43, 5621. 47 Ghosh, A. K.; Liu, C.; Xu, X. Unpublished results, University

of Illinois–Chicago. 48 Xiang, Y.; Olivier, E.; Ouimet, N. Tetrahedron Lett. 1992, 33,

457. 49 Ahn, K. H.; Yoo, D. J.; Kim, J. S. Tetrahedron Lett. 1992, 22,

6661. 50 Casper, D. M.; Burgeson, J. R.; Esken, J. M.; Ferrence, G.

M.; Hitchcock, S. R. Org. Lett. 2002, 4, 3739. 51 Bold, G.; Duthaler, R. O.; Riediker, M. Angew. Chem. Int.

Ed. Engl. 1989, 28, 497. 52 Duthaler, R. O.; Herold, P.; Wyler-Helfer, S.; Riediker,

M. Helv. Chim. Acta 1990, 73, 659. 53 (a) Reetz, M. T.; Peter, R. Tetrahedron Lett. 1981, 22, 4691.

(b) Harrison, C. R. Tetrahedron Lett. 1987, 28, 4135. (c) Panek, J. S.; Bula, O. A. Tetrahedron Lett. 1988, 29, 1661.

123

124

2 The Development of Titanium Enolate-based Aldol Reactions 54 Mahrwald, R. Org. Lett. 2000, 2, 4011. 55 Siegel, C.; Thornton, E. R. Tetrahedron Lett. 1986, 27, 457. 56 Siegel, C.; Thornton, E. R. J. Am. Chem. Soc. 1989, 111,

5722. 57 Figueras, S.; Martı´n, R.; Romea, P.; Urpı´, F.; Vilarrasa, J.

Tetrahedron Lett. 1997, 38, 1637. 58 Evans, D. A.; Clark, J. S.; Metternich, R.; Novack, V. J.;

Sheppard, G. S. J. Am. Chem. Soc. 1990, 112, 866. 59 Evans, D. A.; Dart, M. J.; Duffy, J. L.; Rieger, D. L. J. Am.

Chem. Soc. 1995, 117, 9073. 60 Meyers, A. I.; Yamamoto, Y. Tetrahedron 1984, 40, 2309. 61 Murphy, P. J.; Procter, G.; Russell, A. T. Tetrahedron Lett.

1987, 28, 2037. 62 Annunziata, R.; Cinquini, M.; Cozzi, F.; Borgia, A. L.

J. Org. Chem. 1992, 57, 6339. 63 Kazmaier, U.; Grandel, R. Synlett 1995, 945. 64 Ghosh, A. K.; Onishi, M. J. Am. Chem. Soc. 1996, 118, 2527. 65 Ghosh, A. K.; Fidanze, S.; Onishi, M.; Hussain, K. A. 66 67 68 69 70 71 72

73

74 75 76 77

78

79 80

81 82

Tetrahedron Lett. 1997, 38, 7171. Fidanze, S. Ph.D. Thesis. University of Illinois–Chicago, 2000. Ghosh, A. K.; Kim, J. Unpublished results. Ghosh, A. K.; Kim, J. Org. Lett. 2003, 5, 1063. Crimmins, M. T.; McDougall, P. J. Org. Lett. 2003, 5, 591. Duthaler, R. O.; Herlold, P.; Wyler-Helfer, S.; Riediker, M. Helv. Chim. Acta 1990, 73, 659. Ghosh, A. K.; Fidanze, S. Org. Lett. 2000, 2, 2405. (a) Evans, D. A.; Chapman, K. T.; Carriera, E. M. J. Org. Chem. 1988, 53, 3560. (b) Evans, D. A.; Chapman, K. T. Tetrahedron Lett. 1986, 27, 5939. Hu, T.; Schaus, J. V.; Lam, K.; Palfreyman, M. G.; Wuonola, M.; Gustafson, G.; Pankek, J. S. J. Org. Chem. 1998, 63, 2401. Crimmins, M. T.; King, B. W. J. Am. Chem. Soc. 1998, 120, 9084. Ghosh, A. K.; Bischoff, A.; Cappiello, J. Eur. J. Org. Chem. 2003, 821. Ghosh, A. K.; Bischoff, A.; Cappiello, J. Org. Lett. 2001, 3, 2677. (a) Dondoni, A.; Perrone, D.; Semola, T. J. Org. Chem. 1995, 60, 7927. (b) Dondoni, A.; Fantin, G.; Fogagnolo, M.; Medici, A.; Pedrini, P. Synthesis 1988, 685. (c) Dondoni, A.; Fogagnolo, M.; Medici, A.; Pedrini, P. Tetrahedron Lett. 1985, 26, 5477. Biswas, K.; Lin, H.; Njardarson, J. T.; Chappel, M. D.; Chou, T.; Guan, Y.; Tong, W. P.; He, Lifeng; Horwitz, S. B.; Danishefsky, S. J. J. Am. Chem. Soc. 2002, 124, 9825. Chappell, M. D.; Stachel, S. J.; Lee, C. B.; Danishefsky, S. J. Org. Lett. 2000, 2, 1633. (a) Lee, C. B.; Wu, Z.; Zhang, F.; Chappell, M. D.; Stachel, S. J.; Chou, T. C.; Guan, Y.; Danishefsky, S. J. J. Am. Chem. Soc. 2001, 123, 5249. (b) Wu, Z.; Zhang, F.; Danishefsky, S. J. Angew. Chem. Int. Ed. Engl. 2000, 39, 4505. Ghosh, A. K.; Bischoff, A. Org. Lett. 2000, 2, 1753. Ghosh, A. K.; Liu, C. J. Am. Chem. Soc. 2003, 125, 2374.

References 83 Ley, S. V.; Lygo, B.; Wonnacott, A. Tetrahedron Lett. 1989, 26,

535. 84 Hayward, C. M.; Yohannes, D.; Danishefsky, S. J. J. Am.

Chem. Soc. 1993, 115, 9345. 85 Crimmins, M. T.; Katz, J. D.; Washburn, D. G.; Allwein,

S. P.; McAtee, L. F. J. Am. Chem. Soc. 2002, 124, 5661. 86 Sheppeck, J. E. II; Liu, W.; Chamerlin, A. R. J. Org. Chem.

1997, 62, 387. 87 Chakraborty, T. K.; Jayaprakash, S.; Laxman, P. Tetrahedron

2001, 57, 9461. 88 Enders, D.; Geibel, G.; Osborne, S. Chem. Eur. J. 2000, 6,

1302. 89 (a) Enders, D. in Asymmetric Synthesis, Vol 3 (J. D. Morrison,

90 91 92

93 94 95

Ed.), Academic Press, Orlando, 1984, 275. (b) Enders, D. Chem. Scripta 1985, 65, 139. (c) Enders, D.; Fey, P.; Kipphardt, H. Org. Synth. 1987, 65, 173, 183. Paterson, I.; Perkins, M. V. Tetrahedron, 1996, 52, 1811. Perkins, M. V.; Sampson, R. A. Org. Lett. 2001, 3, 123. (a) Narasaka, K.; Pai, F. Tetrahedron 1984, 40, 2233. (b) Patterson, I.; Donghi, M.; Gerlach, K. Angew. Chem. Int. Ed. Engl. 2000, 39, 3315. Stille, J. R.; Grubbs, R. H. J. Am. Chem. Soc. 1983, 105, 1664. Maeda, K.; Shinokubo, H.; Oshima, K. J. Org. Chem. 1998, 63, 4558. Mukaiyama, T.; Kagayama, A.; Igarashi, K.; Shiina, I. Chem. Lett. 1999, 1157.

125

127

3

Boron and Silicon Enolates in Crossed Aldol Reaction Teruaki Mukaiyama and Jun-ichi Matsuo 3.1

Introduction

Metal enolates play an important role in organic synthesis and metal enolate-mediated aldol type reactions, in particular, are very useful synthetic tools in stereoselective and asymmetric carbon–carbon bond formation. Generation and reactions of diﬀerent metal enolates have been extensively studied and successful applications to the controlled formation of carbon– carbon bonds have been realized under mild conditions. The aldol reaction has long been recognized as one of the most useful synthetic tools. Under classical aldol reaction conditions, in which basic media are usually employed, dimers, polymers, self-condensation products, or a,b-unsaturated carbonyl compounds are invariably formed as byproducts. The lithium enolate-mediated aldol reaction is regarded as one useful synthetic means of solving these problems. Besides the well-studied aldol reaction based on lithium enolates, very versatile regio- and stereoselective carbon–carbon bond forming aldol-type reactions have been established in our laboratory by use of boron enolates (1971), silicon enolates– Lewis acids (1973), and tin(II) enolates (1982). Here we describe the ﬁrst two topics, boron and silicon enolate-mediated crossed aldol reactions, in sequence. 3.2

Crossed Aldol Reactions Using Boron Enolates 3.2.1

Discovery of Aldol Reaction Mediated by Boron Enolates

First, the background of how we ﬁrst conceived the idea of using boron enolate (vinyloxyboranes) in aldol reactions is described. At the beginning of the 1970s, several reactions were being screened by utilizing characteristics of alkylthioboranes based on the concept of elements in combination, i.e. two elements in combination, create a novel reactivity diﬀerent from when Modern Aldol Reactions. Vol. 1: Enolates, Organocatalysis, Biocatalysis and Natural Product Synthesis. Edited by Rainer Mahrwald Copyright 8 2004 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 3-527-30714-1

128

3 Boron and Silicon Enolates in Crossed Aldol Reaction

(CH3)2CO hν

OH O

H 2O

H2C C O + Bu2B SBu 1 2

SBu 4

SBu H2C C 3

SBu

Scheme 3.1

Unexpected formation of b-hydroxy thioester 4 on reaction of ketene 1 with thioborane 2.

they are used separately. When ketene 1 was mixed with two moles of butylthioborane 2, which we assumed would result in the formation of ketene thioacetal 3, S-butyl 3-hydroxy-3-methylbutanethiolate 4 was unexpectedly obtained (Scheme 3.1). It was diﬃcult at ﬁrst to discover the mechanism of this reaction, but the product soon indicated the participation of acetone in this reaction. In the experiment, ketene 1 is generated by degradation of acetone under irradiation; a small amount of acetone is, therefore, introduced into the reaction mixture. Thus b-hydroxy thioester 4 is aﬀorded by reaction of the three components acetone, ketene 1, and alkylthioborane 2. When a gaseous ketene 1, free form acetone, is introduced into the mixture of alkylthioborane 2 and carbonyl compound, the expected b-hydroxy thioesters 5 are obtained in high yield (Eq. (1)) [1].

RCHO +

1

OH O

H2O

H2C C O + Bu2B SBu

R

2

SBu

ð1Þ

5

Investigation of this mechanism reveals that the key intermediate of this reaction is vinyloxyborane (boron enolate) 7 generated from ketene 1 and alkylthioborane 6 (Eq. (2)) [2]. Thus, our original study on organothioboranes led us, unexpectedly, to discover the widely utilized aldol reactions via boron enolates [3]. OBBu2 Bu2B SPh + H2C C O 6 1

H 2C SPh vinyloxyborane 7

ð2Þ Me2CO

O

Bu Bu B O SPh

OH O SPh

3.2 Crossed Aldol Reactions Using Boron Enolates

3.2.2

New Method for Direct Generation of Boron Enolates

Direct generation of boron enolates from parent carbonyl compounds had been desired to expand the synthetic utility of the boron enolate-mediated aldol reaction. Although several synthetic methods to generate vinyloxyboranes (boron enolates) were reported [3–5] no useful ones for direct generation of boron enolates from parent carbonyl compounds were known until 1976. After discovering the above-mentioned aldol reaction via boron enolates we had been exploring a useful method for direct generation of boron enolates from parent carbonyl compounds. It was then thought that increasing the Lewis acidity of boron by introducing an excellent leaving group on to boron would result in an increase in acidity of the carbonyl compounds by coordination of a carbonyl group to the boron compound; the corresponding boron enolate would then be formed by abstraction of the a-proton of the carbonyl compound with a weak base such as tertiary amine. The triﬂuoromethanesulfonyloxy (triﬂate, TfO) group was chosen as the leaving group, and dibutylboryl triﬂate 8 was found to generate boron enolates 9 by the reaction with ketones in the presence of a weak base such as N-diisopropylethylamine or 2,6-lutidine (Eq. (3)) [6]. This is the ﬁrst example of the use of a metal triﬂate in synthetic chemistry; a variety of metal triﬂates are now known to be versatile Lewis acids in organic synthesis. Subsequent addition of aldehydes aﬀorded the corresponding aldols 10 in good yields. Thus, the crossed aldol reaction which starts from ketone and aldehyde is performed easily by applying dialkylboryl triﬂate under mild reaction conditions.

O R1

Bu2BOTf 8 i-Pr2NEt

OBBu2

O

R2CHO

OH

R1

R1 9

R2

ð3Þ

10

After our ﬁrst report, this aldol reaction has been investigated in detail by many research groups [3]. It is currently understood that the boron enolatemediated aldol reaction proceeds via a more rigid chair-like six-membered transition state (12 or 15) than those of alkali metal enolates, because of a shorter bond length between boron and oxygen (Figure 3.1), that is, dialkylboron enolates have relatively short metal–ligand and metal–oxygen bonds, which are suited to maximizing 1,3-diaxial (R3 –L) interactions in the transition states. This facilitates the formation of more stable transition states (12 and 15), where R3 occupies a pseudoequatorial position, when vinyloxyboranes (11 and 14) react with aldehydes to aﬀord aldol adducts (13 and 16) stereoselectively. Therefore, aldol reactions via boron enolates give aldol adducts more stereoselectively than those via alkali metal enolates such as lithium enolates. This stereoselective aldol reaction is thus an outstanding method for the stereoselective synthesis of acyclic compounds.

129

130

3 Boron and Silicon Enolates in Crossed Aldol Reaction

O OBL2 R

R1

2

H

H R

R

L O

R1 O

3 3

B

L

R2

R3

R1

O

11 (Z )-enolate

OH R2 13 syn

12

O OBL2

H

H

R3

R1 R2

2

R 14 (E )-enolate

R3

L R1 O

B

O L

OH R3

R1

O

R2 16 anti

15

Fig. 3.1

Stereoselective aldol reaction of (Z) or (E) boron enolates and aldehydes.

3.2.3

Regioselectivity on Generation of Boron Enolates

Regioselective formation of boron enolates is conducted by using a-halo[3–5], a-diazo- [3], or a,b-unsaturated ketones [3] but their direct generation from parent carbonyl compounds is a more important and synthetically useful method. The regioselectivity on generation of the boron enolate is controlled by the reaction conditions. The kinetic boron enolate of 2-pentanone 17 is formed by use of dibutylboryl triﬂate 8 and N-diisopropylethylamine at 78 C in a short reaction time [6a], whereas the thermodynamic enolate 18 is predominantly generated by the use of 9-BBNOTf 19 and 2,6lutidine at 78 C in a long reaction time (Figure 3.2) [6b]. Subsequent aldol reactions of boron enolates with aldehydes proceed without loss of regiochemical integrity, but the reactions with ketones proceed slowly and the regiochemical integrity of an aldol product does not reﬂect that of an enolate. Preparation of Dibutylboryl Triﬂate 8 (Eq. (4)) [6c, 7] Bu3B + TfOH

OB

18

Bu2BOTf 8

BOTf 19

ð4Þ

O

2,6-lutidine –78 °C, 16 h

Fig. 3.2

Regioselective generation of boron enolates.

Bu2BOTf 8 i-Pr2NEt –78 °C, 15 min

OBBu2

17

3.2 Crossed Aldol Reactions Using Boron Enolates

A small amount of triﬂuoromethanesulfonic acid (1.0 g) was added to tributylborane (15.16 g, 83.3 mmol) at room temperature under argon. The mixture was stirred and warmed to 50 C until evolution of butane began (there is an induction period). After cooling of the mixture to 25 C the remaining triﬂuoromethanesulfonic acid (11.51 g, total 83.3 mmol) was added dropwise at such a rate as to maintain a temperature between 25 and 50 C. The mixture was then stirred for further 3 h at 25 C. Distillation under reduced pressure gave pure dibutylboryl triﬂate 8 (19.15 g, 84%; bp 60 C/ 2 mmHg). Preparation of 9-BBN Triﬂate 19 (Eq. (5)) [6c] BH + TfOH

BOTf

ð5Þ

19

Triﬂuoromethanesulfonic acid (18.75 g, 125 mmol) was added to 9-BBN (15.33 g, 127 mmol) in hexane (100 mL) under argon. After overnight stirring the reaction mixture was concentrated and distilled in vacuo to aﬀord 9-BBNOTf 19 (38 C/0.03 mmHg, 28.84 g, 85%). 3.2.4

Stereoselective Formation of (E) or (Z) Boron Enolates

To obtain either syn or anti aldol adducts selectively it is important to generate boron enolates with the appropriate geometry (E or Z), that is, (Z) enolates 11 react with a variety of aldehydes to yield predominantly syn aldols 13, whereas (E) enolates 14 react somewhat less stereoselectively to give anti aldol adducts 16 as the major products (Figure 3.1 and Table 3.1). (E) Boron ketone enolates 20 are generated from a hindered dialkylboryl triﬂate (e.g. dicyclohexyl) and diisopropylethylamine at 0 C whereas the (Z) isomer 21 is prepared by using a less hindered boryl triﬂate (e.g. dibutyl) at 78 C (Figure 3.3) [8]. (Z) Ester boron enolates 22 [9] are selectively generated by using Bu2 BOTf and i-Pr2 NEt when the methyl or ethyl esters are employed, whereas (E) ester boron enolates 23 [9] are also selectively generated by using cHex2 BOTf and Et3 N when the tert-butyl ester is employed (Figure 3.4) [10]. Preparation of Dicyclohexylboryl Triﬂate (Eq. (6)) [11] c-Hex2BH + TfOH

c-Hex2BOTf

ð6Þ

A 250-mL round-bottomed ﬂask capped with a rubber septum, containing a magnetic stirring bar and a connecting tube attached to a mercury bubbler was kept at 0 C and charged with hexane (100 mL) and c-Hex2 BH (26.7 g,

131

132

3 Boron and Silicon Enolates in Crossed Aldol Reaction Tab. 3.1

Stereoselective formation of syn and anti aldols via (Z) and (E) boron enolates. O

OBBu2 R2

+

R1

Ph

OH O H

Ph

OH O R

1

+

R1

Ph

2

R2

R syn

Boron Enolate

syn/anti

Ref.

>95:5

7a

OBBu2 Me

Boron Enolate a

syn/anti

Ref.

>97:3

8a

>97:3

8a

OBBu2

Z OBBu2 Me

anti

t-Bu Z:E = >99:1 OBBu2

25:75

7a

Ph Z:E = 99:1

E OBBu2 OBBu2 Et Z :E = >97:3

8a

98:2

7c

O

Bu2BO

18:82

8a

Z:E = 19:81

N

O

Z:E = >97:3

OB(c-C5H9)2 c-Hex

10:90

Z:E = <5:95

OB(c-C5H9)2 i-Pr

St Bu

8a

>97:3

a The

14:86

7b

highest priority designation is assigned to the OBR2 group with enolate substituents.

Z:E = 12:88

150 mmol). Triﬂuoromethanesulfonic acid (13.3 mL, 150 mmol) was added dropwise using a syringe with constant stirring. Hydrogen is rapidly evolved and should be safely vented. The stirring was continued at 0 C for 2–3 h. All the suspended solid c-Hex2 BH dissolved and the homogeneous reaction mixture was left at 0 C for 1–2 h without stirring. Two layers were obtained and the top layer was transferred into a dry 250-mL round-bottomed ﬂask

OB(c-Hex)2 c-Hex2BOTf R (E )-20 R = i-Pr E:Z = 81:19

i-Pr2NEt 0 °C, 30 min

O R

Bu2BOTf i-Pr2NEt –78 °C, 30 min

Fig. 3.3

Stereoselective formation of (Z) or (E) boron enolates of ketones.

OBBu2 R (Z )-21 R = Et E:Z = <3:97

3.2 Crossed Aldol Reactions Using Boron Enolates

Bu2BOTf

OBBu2

OH

RO

RO

i-Pr2NEt –78 °C, 1 h

O

R'CHO

R = Me, Et

R' syn

(Z )-22 O RO

OB(c-Hex)2

c-Hex2BOTf Et3N –78 °C, 1 h

O

R'CHO

OH

RO

RO

R' anti

R = t-Bu (E )-23

Fig. 3.4

Stereoselective formation of (Z) or (E) boron enolates of carboxylic esters.

leaving the small yellow layer (approx. 2 mL) behind. Solid c-Hex2 BOTf was obtained by removing the solvent using a water aspirator (15–20 mm). It was then recrystallized from hexane. Mp 88 C, yield 80%. Stock solutions (1.00 m) in CCl 4 and in hexane were prepared and kept at 0 C for enolboration. Typical Experimental Procedure for Crossed Aldol Reaction via Boron Enolate (Eq. (7)) [6a] Bu2BOTf i-Pr2NEt

O

OBBu2

24

ð7Þ Ph

CHO

O

OH Ph 25

To a solution of dibutylboryl triﬂate (301 mg, 1.1 mmol) and diisopropylethylamine (142 mg, 1.1 mmol) in ether (1.5 mL) was added dropwise 2methyl-4-pentanone 24 (100 mg, 1.0 mmol) in ether (1.5 mL) at 78 C, under argon, with stirring. After stirring of the mixture for 30 min, 3phenylpropanal (134 mg, 1.0 mmol) in ether (1.5 mL) was added at the same temperature. The reaction mixture was allowed to stand for 1 h, then added to pH 7 phosphate buﬀer at room temperature and extracted with ether. After removal of ether, the mixture was treated with 30% H2 O2

133

3 Boron and Silicon Enolates in Crossed Aldol Reaction

134

(1 mL) in methanol (3 mL) for 2 h and H2 O was added. The mixture was concentrated to remove most of the methanol and extracted with ether. The organic layer was washed with 5% NaHCO3 solution and brine, dried over Na2 SO4 , and concentrated. The crude oil was puriﬁed by preparative TLC to give 3-hydroxy-7-methyl-1-phenyl-5-octanone 25 (192 mg, 82%). 3.2.5

syn-Selective Asymmetric Boron Aldol Reactions

Conventional asymmetric aldol reactions have been performed by using chiral enolates and achiral carbonyl compounds. A chiral boron enolate generated from a chiral oxazolidone derivative (26 and 28), dialkylboron triﬂate, and diisopropylethylamine reacts stereoselectively with aldehydes to aﬀord the corresponding syn aldol adducts (27 and 29) in good yields with excellent diastereoselectivity (Eqs. (8) and (9)) [12]. The opposite sense of asymmetric induction is achieved by changing the chiral auxiliary. Several other chiral auxiliaries have also been developed for highly diastereoselective synthesis of syn aldol adducts (Eqs. (10)–(13)) [13]. O

O

O

OBR2

R2BOTf O

N

O O

N

O

O

OH

R'CHO R'

N

i-Pr2NEt 27

26

ð8Þ O

O

O

OBR2

R2BOTf O

N

O i-Pr2NEt

Ph 28

O

O

OH

R'CHO O

N

R'

N

Ph

Ph

29

ð9Þ

O

OBR2

R2BOTf

O

R'CHO

OH R'

i-Pr2NEt

t-BuMe2SiO

O

1) HF-CH3CN 2) NaIO4

HO

t-BuMe2SiO

t-BuMe2SiO

OH R'

ð10Þ

3.2 Crossed Aldol Reactions Using Boron Enolates

Me Me

N

Ipc2BOTf

O

i-Pr2NEt

Me Me

MeO

2

N

OH

O

1) H2O2

RCHO

B

135

R

2) H3O+

O

syn/anti = 10/90~5/95 77~85% ee (anti)

3) CH2N2

ð11Þ OMe

Ph

N

9-BBNOTf

O

i-Pr2NEt

OMe

OH

MeO

N Ph

O

1) H2O2

RCHO

B

R

2) H3O+

O

syn/anti = 97/3~98/2 40~60% ee (syn)

3) CH2N2

ð12Þ

O

R2BOTf, i-Pr2NEt

OBR2

N

R'CHO

O

N CH2Cl2, -5°C

S O2

S O2

R'

XN

-78°C

OH

ð13Þ 3.2.6

anti-Selective Asymmetric Aldol Reaction

An anti-selective diastereoselective aldol reaction [14] has been performed by using enantiomerically pure carboxylic esters derived from ()- or (þ)norephedrine 30 [15]. This method is applicable to a wide range of aldehydes with high selectivity (both syn/anti and diastereoselectivity of anti isomer). It is proposed that (E) boron enolates 31 are formed by this procedure and aldol reaction proceeds via the six-membered transition state (Eq. (14)). The aldol products 32 are converted to the corresponding alcohols (LiAlH4 , THF) or carboxylic acids (LiOH, THFaH2 O) without loss of stereochemical integrity. Ph Me Bn

O

c-Hex2BOTf

O N

Et3N

SO2Mes 30

Ph Me Bn

OB(c-Hex)2

O N

SO2Mes

Me Bn

O N

SO2Mes

31

32

Me Me

Mes: Me

O

Ph RCHO

ð14Þ

OH R

3 Boron and Silicon Enolates in Crossed Aldol Reaction

136

It is also reported that addition of Lewis acids to the reaction of chiral boron enolates and aldehydes changes syn-selectivity to anti-selectivity (Eqs. (15) and (16)) [16]. The change of stereoselection is rationalized by considering a Lewis acid-mediated open transition state (e.g. 33). O

O

O Bu2BOTf

O

N

O

O

O

OH i-Pr +

N

O

O

OH

N

i-Pr

Et2AlCl

i-Pr2NEt

Bu Bu B O O

O

i-PrCHO

:

95

5

i

Pr Me

N

O H H i-Pr LA

ð15Þ

33 1) Et2BOTf, i-Pr2EtN

X

2) RCHO

R O

OH syn

ð16Þ SO2

O 1) Et2BOTf, i-Pr2EtN 2) RCHO, TiCl4

X

R O

OH anti

Thus, boron enolates prepared under mild conditions enable aldol-type reactions essentially under neutral conditions. Stereocontrolled synthesis of acyclic molecules has been achieved by employing boron enolate-mediated aldol reactions; this method has been extensively applied to the synthesis of natural products. Typical Procedure for syn-Selective Asymmetric Boron Aldol Reaction (Eq. (17)) [13] O

O

O

O

N

O

OBR2

Bu2BOTf

O

OH

i-PrCHO O

N

O

N

i-Pr

i-Pr2NEt 26

34 syn/anti = 497:1

ð17Þ

3.3 Crossed Aldol Reactions Using Silicon Enolates

To a 0.2–0.5 m solution of chiral imide 26 in CH2 Cl2 under argon (0 C) was added 1.1 equiv. dibutylboryl triﬂate followed by 1.2 equiv. diisopropylethylamine. After 30 min the reaction mixture was cooled (78 C) and 1.1 equiv. isobutyraldehyde was added and stirred for 0.5 h at 78 C and then for 1.5 h at room temperature. The reaction was quenched with pH 7 phosphate buﬀer and the boron aldolate complex was oxidized with 30% hydrogen peroxide–methanol (0 C, 1 h). The aldol product 34 was then isolated by ether extraction (syn/anti ¼ 497:1). Typical Procedure for anti-Selective Asymmetric Boron Aldol Reaction (Eq. (18)) [15c] Ph Me Bn

O

1) c-Hex2BOTf, Et3N

O N

SO2Mes

2) i-PrCHO

30

Ph Me Bn

O

OH

O N

SO2Mes

i-Pr

ð18Þ

35 anti:syn = >98:2 96%de (anti)

To a solution of chiral ester 30 (4.80 g, 10 mmol) and triethylamine (3.40 mL, 24 mmol) in dichloromethane (50 mL) at 78 C under nitrogen is added a solution of dicyclohexylboron triﬂate (1.0 m in hexane, 22 mL, 22 mmol) dropwise over 20 min. After the resulting solution has been stirred at 78 C for 30 min, isobutyraldehyde (1.08 mL, 12 mmol) is added dropwise. The reaction mixture is stirred for 30 min at 78 C then left to warm to room temperature over 1 h. The reaction is quenched by adding pH 7 buﬀer solution (40 mL), the mixture is diluted with methanol (200 mL), and 30% hydrogen peroxide (20 mL) is added. After the mixture has been stirred vigorously overnight, it is concentrated. Water is added to the residue, and the mixture is extracted with dichloromethane. The combined organic extracts are washed with water, dried over Na2 SO4 , ﬁltered, and concentrated. Puriﬁcation of the residue gives aldol adduct 35 (88%, anti/syn b 24:1).

3.3

Crossed Aldol Reactions Using Silicon Enolates 3.3.1

Discovery of Silicon Enolate-mediated Crossed Aldol Reactions

The driving force of the above-mentioned aldol reaction with boron enolate is considered to be the interconversion of enol ketones (boron enolates) to their more stable ketones (b-boryloxy ketones) [17]. When the boron

137

138

3 Boron and Silicon Enolates in Crossed Aldol Reaction

enolate-mediated aldol reaction was studied in our laboratory an investigation on the development of new reaction chemistry using titanium(IV) chloride was in progress [18]. A new and important idea was immediately came to mind that titanium(IV) chloride would eﬀectively generate active electrophilic species as a result of its strong interaction with carbonyl compounds, and the complex thus formed would react easily even with relativly weaker carbon nucleophiles to form a new carbon–carbon bond. Use of a stable and isolable silyl enol ether [19] was suggested as a weak nucleophile and, just as expected, aldol reaction between silyl enol ether of acetophenone 36 and benzaldehyde in the presence of titanium(IV) chloride aﬀorded the aldol product 37 in high yield (Eq. (19)) [20]. O Ph

OSiMe3 +

H

OH O

TiCl4 Ph

Ph

ð19Þ

Ph 37

36

It is reported that enol ethers react with acetals or ketals, promoted by Lewis acids, to give aldol-type adducts: these reactions of alkyl enol ethers are, however, often accompanied by undesired side reactions [21]. Further, is diﬃcult to perform crossed-aldol reactions selectively because conventional aldol reactions are conducted under equilibrium conditions using a basic or acidic catalyst in protic solvents [22]. Detailed studies of this new aldol reaction of silicon enolates, however, reveal a number of advantages over conventional methods. First, it not only gives a variety of aldol adducts in high yields but also a regioselective aldol adduct when the silyl enol ether of an unsymmetrical ketone is used. That is, the aldol reaction proceeds with retention of the regiochemical integrity of the starting silyl enol ethers to aﬀord the corresponding aldol regiospeciﬁcally. Starting silyl enol ethers can be conveniently prepared regioselectively under kinetically or thermodynamically controlled conditions. Second, functional group selectivity is observed – i.e. reactions with aldehydes proceed at 78 C whereas those with ketones proceed at elevated temperatures (ca. 0 C). Chemoselectivity is observed with acceptors having two diﬀerent kinds of carbonyl function, for example aldehyde and ketone or ester, in the same molecule. Treatment of phenylglyoxal with silyl enol ether 38 at 78 C aﬀords a-hydroxy-g-diketone 39 (Eq. (20)) [20b]. The reaction of ketoesters 40 other than b-ketoesters with silyl enol ether 38 gives hydroxyketoesters 41 as sole products (Eq. (21)) [23]. O

OSiMe3 H

Ph O

+

Ph

Me 38

O

TiCl4

Ph Me

Ph OH O 39

ð20Þ

3.3 Crossed Aldol Reactions Using Silicon Enolates

O

O

OSiMe3 +

OR

n

Ph

Me

Me

n

OR

Ph 41

38

40 n = 0, 2, 3 R = Me, Et

OH O

O

TiCl4

ð21Þ A directed aldol reaction between two ketones aﬀords thermodynamically unfavorable aldols in high yields, because of stabilization of the aldol adducts by their intramolecular chelation with titanium 42 or by their conversion to silyl ethers 43 (Eq. (22)). TiCl3 OSiMe3

O R2

R1

3 + R

O

TiCl4

R5

R

R4

O

Me3SiO

1

R5

R2 3 4 R R 42

HO

H2O R1

or

R

O

1

R2 3 4 R R

R5

43

O

R2 3 4 R R

R5

ð22Þ

Despite its remarkable power as a method for carbon–carbon bond formation, the level and sense of its stereoselectivity often vary. The syn/anti ratio for the aldol product is aﬀected by the stereochemistry of the aldehyde and silyl enolate, and the character of the Lewis acid catalyst (Figure 3.5 and Table 3.2) [24]. The mechanistic basis for the aldol reaction has not yet been ﬁrmly established. Apart from a limited number of exceptions, however [25], the stereochemical observations have been rationalized by considering acyclic transition states (44–47) [26]. Typical Procedure for Titanium(IV) Chloride-catalyzed Aldol Reaction of Silicon Enolates (Eq. (23)) [20b] OSiMe3

O

+ PhCHO 48

OH

TiCl4 Ph

ð23Þ

49

A solution of 1-trimethylsilyloxy-1-cyclohexene 48 (426 mg, 2.5 mmol) in dichloromethane (10 mL) was added dropwise into a mixture of benzaldehyde (292 mg, 2.75 mmol) and TiCl 4 (550 mg, 2.75 mmol) in dichloro-

139

140

3 Boron and Silicon Enolates in Crossed Aldol Reaction

O

OSiMe3 R2

+

1

R

R

OH O

Lewis Acid (LA)

3

H

Ph

OH O R

+

1

R1

Ph

R2

2

R anti

syn

LA O

H (Z )-enolate:

R

3

H

R1

LA O

R2 44

R

vs.

OSiMe3

R

3

R1 anti

LA

LA O

H

H

R2 R3

H

R1

45

H

Me3SiO

syn

(E )-enolate:

H

2

46

vs.

OSiMe3

O

R3

H

47

R1

Me3SiO

anti

R2

syn

Fig. 3.5

Transition-state models for Lewis acid-catalyzed aldol reaction of silicon enolates.

methane (20 mL) under an argon atmosphere at 78 C and the reaction mixture was stirred for 1 h. After hydrolysis at that temperature the resulting organic layer was extracted with ether, and the combined organic extract was washed with water, dried over Na2 SO4 , and concentrated. The residue was puriﬁed by chromatography to aﬀord a diastereo mixture of aldol adducts 49 (92%, syn/anti ¼ 75/25). As an extension of this new procedure for carbon–carbon bond formation, the reaction between silyl enol ethers and acetals 50, a typical protecting group of aldehydes, is performed to aﬀord b-alkoxy carbonyl compound 51 in the presence of titanium(IV) chloride (Eq. (24)) [27]. A variety of substituted furans are readily prepared by application of the TiCl 4 -promoted reaction of a-halo acetals 52 with silyl enol ethers (Eq. (25)) [27]. OSiMe3

R1 R2

OR OR 50

+

R3

R5 R4

R5 = H, Alkyl, Ar, OR

RO

TiCl4 R1

O

R2 3 4 R R 51

R5

ð24Þ

3.3 Crossed Aldol Reactions Using Silicon Enolates

141

Tab. 3.2

Stereochemical outcome of Lewis acid-catalyzed aldol reactions of a variety of silicon enolates. O

OSiMe3 R2

R1

+

Ph

OH O

OH O

Lewis acid H

Ph

R

1

+

2

R anti

syn

Silicon Enolate

Lewis acid syn/anti Ref.

Silicon Enolate a

OSiMe3 OEt Z:E = 100:0

33:67

24a

TiCl 4

26:74

24a

t-Bu

4:96

24c

BF3 aOEt2

47:53

24c

BF3 aOEt2

5:95

24d

BF3 aOEt2

4:96

24d

TiCl 4

14:86

24b

TiCl 4

77:23

24b

OSiMe3 Ph Z:E = 100:0 OSiMe3

TiCl 4

>92:8

St Bu

24b

OEt

Z:E = 10:90

OSiMe3

TiCl 4

OEt

TiCl 4

t-Bu Z:E = 100:0

Z:E = 15:85 OSiMe3

Lewis acid syn/anti Ref.

OSiMe3

TiCl 4

OSiMe3 OEt

R1

Ph

R2

>92:8

OSiMe2t Bu

24b

St Bu Z:E = >95:5

t-Bu OSiMe3

OSiMe3

TiCl 4

25:75

20b

OSiMe3 OSiMe3

Z:E = 0:100 t-Bu OSiMe3

TiCl 4

50:50

20b

BF3 aOEt2

60:40

24c

BF3 aOEt2

56:44

24c

OSiMe3

a The

highest priority designation is assigned to the OSiR3 group with enolate substituents.

Z:E = 0:100 OSiMe3 Et Z:E = 100:0 OSiMe3 i-Pr Z:E = 97:3

Br R1 R

OMe 2 OMe

OSiMe3 + R3CH

R4

TiCl4

Br

R3

R1 MeO R2 O

R4

toluene reflux

52

R1

O

R3

R2 53

ð25Þ

R4

142

3 Boron and Silicon Enolates in Crossed Aldol Reaction

OMe

OMe

1) TiCl4-Ti(Oi-Pr)4

OMe

+

OSiMe3

CHO

2) H2O

55 54

56

OH Vitamin A Scheme 3.2

Synthesis of vitamin A using aldol reaction of acetal 54 and silyl dienol ether 55.

In the presence of titanium(IV) chloride, silyl dienol ether 55 derived from an a,b-unsaturated aldehyde reacts with acetal 54 selectively at the gposition to give d-alkoxy-a,b-unsaturated aldehydes 56, albeit in low yields. Because titanium(IV) chloride is strongly acidic, polymerization of silyl dienol ether 55 proceeds. In these reactions addition of tetraisopropoxytitanium(IV) to titanium(IV) chloride increases the yield dramatically [28a] – vitamin A is successfully synthesized by utilizing this aldol reaction of silyl dienol ether 55 (Scheme 3.2) [28b]. When this reaction was further investigated using trimethylsilyltriﬂuoromethanesulfonate (Me3 SiOTf ) as Lewis acid some interesting results were reported. Me3 SiOTf-mediated aldol reaction of silicon enolates and acetals tends to give syn-b-methoxy ketones as a major products, irrespective of the stereochemistry of the silicon enolate double bond, except for the (Z) silicon enolate of tert-butyl ethyl ketone 57 (Table 3.3) [29]. Because they are even more nucleophilic than silyl enol ethers, silyl ketene acetals 58, derived from carboxylic esters, react with ketones and aldehydes in the presence of titanium(IV) chloride to give b-hydroxy esters 59 in high yields (Eq. (26)) [30, 31]. Although the Reformatsky reaction is well known as a good synthetic tool for synthesis of b-hydroxy esters, the titanium(IV) chloride-mediated reaction is a milder and more versatile method for synthesis of a-substituted b-hydroxy esters. OSiMe3

O R1

R2

+

R3

OR R4 58

HO

1) TiCl4

5

2) H2O

R

O

1

R2 3 4 R R 59

OR5

ð26Þ

After the discovery of aldol reactions of silyl enolates with carbonyl compounds or acetals, as described above, silyl enolates become one of the most popular carbon nucleophiles in organic synthesis and are also employed for other reactions such as Michael reaction [32], Mannich reaction [33], etc.

3.3 Crossed Aldol Reactions Using Silicon Enolates Tab. 3.3

Me3 SiOTf-catalyzed aldol reaction of acetals and silicon enolates. R1

OSiMe3 R3

MeO +

R2

Entry

Silyl Enol Ether

R4

O

Me3SiOTf (cat.)

OMe

R

R5

Acetal

Product O

OSiMe3

syn/anti

OMe

OMe

1 Ph

OMe R5 4 2 3R R R

1

Ph

92:8

OMe 86% O

OSiMe3

OMe

OMe

2 i-Pr

i-Pr

86:14

OMe 95% O OMe

OSiMe3

3 Ph

Ph

OMe

Ph

Ph

84:16

OMe 97% O

OSiMe3

4

Ph

OMe Ph

OMe

Ph

Ph

71:29

OMe 83% O

OSiMe3

5

t-Bu 57

OMe Ph

OMe

t-Bu

Ph

5:95

OMe 94%

[34]. Silyl enolates are superior to other metal enolates in isolation, regioselective formation, and unique reactivity under mild conditions. 3.3.2

Lewis Acid-catalyzed Aldol Reactions of Silicon Enolates

Initially, the titanium(IV) chloride-mediated aldol reaction of silyl enolates with aldehydes was investigated [20] and a catalytic amount of trityl salt 60 (e.g. trityl perchlorate) was found to promote the aldol reaction (Eq. (27)) [35]. Whereas the original reaction is performed by using a stoichiometric amount of titanium(IV) chloride, 5–10 mol% trityl salt is suﬃcient to drive the aldol reaction to completion. One interesting ﬁnding in this catalytic reaction is that the silicon enolate reacts with aldehydes to give the corresponding aldol adducts as their silyl ethers 61.

143

144

3 Boron and Silicon Enolates in Crossed Aldol Reaction

OSiMe3 1

R CHO

+

R

2

Me3SiO

TrX 60 (cat.)

R

R3 +

O R3

1

ð27Þ

2

R 61

–

TrX = Ph3C X X = ClO4, SbCl6, OTf, PF6, etc.

The aldol reaction of a variety of silyl enolates and acetals in the presence of a catalytic amount (1–10 mol%) of trityl perchlorate proceeded eﬃciently to aﬀord b-methoxy ketone in high yield. In the presence of trityl tetraﬂuoroborate catalyst, the reaction of dithioacetal 62 with silyl enol ethers aﬀords b-ethylthio ketones 63 (Eq. (28)) [36].

R1

OSiMe3

SEt +

R2

SEt

R

4

SEt O TrBF4

R1

R3

R4 R

62

2

ð28Þ

3

R 63

A speciﬁc combination of two weak acids, tin(II) chloride and chlorotrimethylsilane, is found to serve as an eﬀective catalyst for the aldol reaction [37]. Neither chlorotrimethylsilane or tin(II) chloride has any accelerating eﬀect at 78 C even when more than one equivalent is added. In the combined presence of catalytic amounts of chlorotrimethylsilane and tin(II) chloride, however, the aldol reaction gives the desired product in more than 90% yield (Eq. (29)). It is supposed that the cationic silyl species generated by coordination of the chloride to the tin(II) atom catalyzes the aldol reaction. Because a similar cationic silicon Lewis acid, trimethylsilyl triﬂate, is known to catalyze aldol reactions of silicon enolates [38], and several other unique Lewis acid catalysts have also been reported (e.g. lanthanide catalysts [39]). Water-stable Lewis acids such as lanthanide triﬂates, which catalyze the aldol reaction in aqueous solvent or pure water, have received much attention in the development of economical and environmentally benign synthetic methods [40]. 1) SnCl2-Me3SiCl (10 mol%)

OSiMe3 +

–78 °C PhCHO

O

OH Ph ð29Þ

2) H3O+

Silicon enolate-mediated aldol reactions are performed by other activation methods. The aldol reaction is performed by using a catalytic amount of tetrabutylammonium ﬂuoride (TBAF) or tris(diethylamino)sulfonium diﬂuorotrimethylsiliconate (TASF) (Table 3.4) [41]. Aldehydes undergo this

3.3 Crossed Aldol Reactions Using Silicon Enolates Tab. 3.4

TBAF-catalyzed aldol reaction of silicon enolates [41d]. R

1

OSiMe3 R3 R2

Entry

O +

H

Silyl Enol Ether (Geometric Purity, %)

O

Bu4NF (cat.) R4

THF, -72 °C

Aldehyde

R4 R2 R3

Product (Diastereomer Ratio)

O

OSiMe3

O

Ph

O

PhCHO

O +

Ph

O

O

O

OSiMe3

:

O

OSiMe3

O

:

O

O

(97)

Ph 36)

OH

H

Ph (35

OH

+

:

(64

OSiMe3

71) O

Ph

(97)

Ph

OH

PhCHO

OH

+

:

(29

OSiMe3

14) O

Ph

(83)

PhCHO

OH n-Pr

OH

PhCHO

H

3)

n-Pr + (86

8

i-Pr

OH

n-PrCHO

7

OH

i-Pr + (97

6

Ph 7)

OH

i-PrCHO

5

OH

Ph

:

(93

OSiMe3

56)

OH

Ph

OH Ph

:

(44

(93)

4

O +

Ph

OSiMe3

14)

OH

PhCHO (89)

Ph

:

(86

OSiMe3

OH

+

Ph

(97)

2

O

OH

PhCHO

1

3

OH

R1

O +

OH

H :

Ph 65)

145

3 Boron and Silicon Enolates in Crossed Aldol Reaction

146

R

1

OSiMe3 R3 R

2

O

F Me3SiF

R 64

O

O

R4CHO

R3

R1

R

R 65

O

R4 R R2 67

4

F

3

R

OSiMe3 R4

1 2

R R 66

3

H

3

O

H

O

Me3SiF R

R2

2

R1

O

1

R3 R1 O

R2

R4

68

Scheme 3.3

Reaction mechanism for ﬂuoride ion-catalyzed aldol reaction of silicon enolates.

type of aldol reaction quite readily (aromatic aldehydes gives aldol products more eﬀectively than aliphatic aldehydes), whereas ordinary aliphatic and aromatic ketones form no aldol products. This reaction is considered to proceed through a catalytic cycle involving reversible steps (Scheme 3.3). A naked enolate 64 generated by the reaction of silicon enolate and ﬂuoride ion reacts with aldehydes to aldol dianion 65. Rapid chemical trapping of unstable aldol anions 65 with Me3 SiF to form 66 serves to drive this reaction in the forward direction. Often, especially with sterically demanding enolates, syn aldol adducts are obtained irrespective of enolate geometry; this is ascribed to an extended open transition state 67. Recently, another open ‘‘skew’’ transition state 68 has been considered for less hindered enolates [42]. Preparation of Anhydrous Tetrabutylammonium Fluoride (TBAF) [41e]. Tetrabutylammonium ﬂuoride, which is commercially available as its trihydrate, was dried over P2 O5 (30–40 C, 0.5 mmHg, overnight). Typical Procedure for the TBAF-catalyzed Aldol Reaction (Eq. (30)) [41a] O

OSiMe3

OH

TBAF (10 mol%) + 48

Ph

PhCHO

ð30Þ

49

A mixture of the trimethylsilyl enolate 48 of cyclohexanone and benzaldehyde (1:1.1 mol ratio) was added in one portion to a solution of TBAF (10 mol%) in THF at 78 C under argon. The resulting mixture was stirred at the same temperature for 3.5 h and poured into hexane. The hexane solu-

3.3 Crossed Aldol Reactions Using Silicon Enolates

tion was washed with water, dried, and concentrated to give aldol silyl ether in 80% yield. The aldol silyl ether was dissolved in a 1:9 mixture of 1 m HCl and methanol and left to stand at room temperature for 5 min. The solution was diluted with 1:1 hexane–ether and treated with water. The organic layer was dried and concentrated to give the almost pure desired aldol adduct 49. Trimethylsilyl enolates are also activated by transmetalation to the corresponding metal enolates by use of MeLi [43], transition metal catalysts [44, 45], etc. [46]. 3.3.3

Non-catalyzed Aldol Reactions of Silicon Enolates

Specially designed silicon enolates which readily form hypervalent silicates react with aldehydes in the absence of catalysts. These include: trichlorosilyl enolates 69 (Eq. (31)) [47], dimethyl(triﬂoxy)silyl enolate 70 (Eq. (32)) [48], enoxysilacyclobutanes 71 (Eq. (33)) [49], and dimethylsily enolates 72 (Eq. (34)) [50]. Owing to the electron-withdrawing group (69 and 70), the less sterically demanding group (72), or angle strain (71) on silicon, Lewis acidity of the silicon is increased. It is also reported that O-silyl enol derivatives of amides 73 (O-silyl ketene N,O-acetals) undergo non-catalyzed aldol reaction with aldehydes (Eq. (35)) [51]. OSiCl3 OMe 69

OH O + PhCHO CH2Cl2, 0 °C

O

OSiMe2(OTf)

Me2Si(OTf)2

Ph

i-Pr2NEt

Ph

ð31Þ

OMe 98% O Ph

Ph

OH

PhCHO Ph

-78 °C 70

88% syn/anti = 91:9

ð32Þ R3 Si OH O

R3 Si O X

R2

+

R4CHO

OSiMe2H +

R4

X R 1 R2

R1 71 X = OR, SR, NR2

1) DMF, 50 °C, 48 h PhCHO

O

OH Ph

2) MeOH-HCl 72

ð33Þ

79% syn/anti = 58:42

ð34Þ

147

148

3 Boron and Silicon Enolates in Crossed Aldol Reaction

O

OSiEt3

CH2Cl2 + PhCHO

Me2N

Me2N

-30 °C, 4 h

73

OSiEt3 Ph

ð35Þ

80% syn/anti = 1:1.8

It is also known that non-catalyzed aldol reactions using silyl ketene acetals proceed at high temperature [52], or in H2 O [53], DMSO, DMF, and DME [54], or under high pressure [55]. 3.3.4

Lewis Base-catalyzed Aldol Reactions of Trimethylsilyl Enolates

Since 1973, many Lewis acid-catalyzed aldol reactions have been studied, as described above. There have, however, been few examples of Lewis basecatalyzed aldol reaction using special silyl enolates, e.g. phosphoramidecatalyzed aldol reactions of trichlorosilyl enolates 69 (Eq. (36)) [47] and CaCl2 -catalyzed reactions of dimethylsilyl enolates 72 in aqueous DMF (Eq. (37)) [56].

OSiCl3 OMe 69

OH O + t-BuCHO

OMe

ð36Þ

CD2Cl2 : 50% conversion (120 min) CD2Cl2+HMPA: 100% conversion (<3 min)

OSiMe2H + 72

t-Bu

CaCl2 (cat.) PhCHO

O

OH Ph

DMF 30 °C, 24 h

ð37Þ

90% syn/anti = 45:55

New challenges were then made to develop useful Lewis base-catalyzed aldol reactions of trimethylsilyl enolates, simple and the most popular silicon enolates. It has recently been found that aldol reactions of trimethylsilyl enolates with aldehydes proceed smoothly under the action of a catalytic amount of lithium diphenylamide or lithium 2-pyrrilidone in DMF or pyridine (Eq. (38)) [57]. This Lewis base-catalyzed aldol reaction of trimethylsilyl enolates [58] has an advantage over acid-catalyzed reactions in that aldol reaction of carbonyl compounds with highly-coordinative functional groups with Lewis acid catalysts are smoothly catalyzed by Lewis bases to aﬀord the desired aldol adducts in high yields.

3.3 Crossed Aldol Reactions Using Silicon Enolates

149

O OSiMe3 R1CHO

R

+

2

R

LiNPh2 or

(cat.)

LiN

4

DMF or pyridine

R3

HO R

O R4

1

R2

R

3

R4 = R, OR, SR

ð38Þ

Preparation of the 0.1 M Solution of Lithium Pyrrolidone in DMF. A solution of MeLi (1.14 m in ether, 0.52 mL, 0.59 mmol) was added to a solution of pyrrolidone (55.3 mg, 0.65 mmol) in THF at 0 C. The solvents were removed in vacuo and DMF (5.9 mL) was added to the residue. Typical Experimental Procedure for Lithium 2-Pyrrolidone-catalyzed Aldol Reaction (Eq. (39)) [57b,c] O O

OSiMe3 H +

OMe

Me2N

LiN

(10 mol%)

Me3SiO

O OMe

DMF, –45 °C, 1 h

74

Me2N

75 97%

ð39Þ A solution of ketene silyl acetal 74 (146 mg, 0.84 mmol) in DMF (0.8 mL) was added to a solution of lithium pyrrolidone in DMF (0.1 m, 0.6 mL, 0.06 mmol) at 45 C and a solution of aldehyde (89.5 mg, 0.60 mmol) in DMF (1.6 mL) was then added. The reaction mixture was stirred at the same temperature for 1 h and the reaction was then quenched by adding saturated NH4 Cl. The mixture was extracted with diethyl ether, and the combined organic extracts were washed with brine, dried over Na2 SO4 , ﬁltered, and concentrated. The crude product was puriﬁed by preparative TLC (hexane–ethylacetate, 3:1) to give the silylated aldol adduct 75 (187 mg, 96%). 3.3.5

Diastereoselective Synthesis of Polyoxygenated Compounds

Stereoselective aldol reaction of the trimethylsilyl enolate of methyl 2benzyloxyacetate 76 with enantiomerically pure trialkoxy aldehyde 77 is performed by using three equivalents of MgBr2 aEt2 O as activator to aﬀord an aldol adduct 78 in a high yield and excellent diastereoselectivity, whereas conventional Lewis acids such as TiCl 4 and SnCl 4 give the desired aldol

3 Boron and Silicon Enolates in Crossed Aldol Reaction

150

OBn

OBn

+ H

Me3SiO OMe

OSiMe2t-Bu O

OBn

OSiMe2t-Bu O

OH OPMB

77

BnO

78

O

AcO

O

TBSO

BzHN

O H HO BzO OAc

O OH

OBn

OH

O

Ph PMBO

OBn

MeO

toluene, –19 °C

OPMB

76

MgBr2-OEt2

Taxol® Scheme 3.4

Diastereoselective aldol reaction for preparing an acyclic polyoxy molecule 78.

product 78 in low yields [59, 60]. This method has been used as a key step in the construction of the B-ring system of an antitumor agent, Taxol (Scheme 3.4) [61]. Diastereoselective Aldol Reaction Mediated by Magnesium Bromide–Diethyl Ether Complex (Eq. (40)) [61c] OBn Me3SiO

OBn + H

OMe 76

OTBS O

OPMB 77

MgBr2-OEt2 toluene, –19 °C

OBn

OBn

MeO

OTBS O

OH OPMB 78

ð40Þ A solution of ketene trimethylsilyl acetal 76 (6.30 g, 24.9 mmol) in toluene (30 mL) and a solution of aldehyde 77 (8.10 g, 16.6 mmol) in toluene (30 mL) were successively added to a suspension of magnesium bromide– diethyl ether complex (12.9 g, 49.9 mmol) in toluene (100 mL) at 19 C. The reaction mixture was stirred for 1 h at 19 C then triethylamine (23 mL) and saturated aqueous NaHCO3 were added. The mixture was extracted with diethyl ether and the combined organic extracts were washed with brine, dried over Na2 SO4 , ﬁltered, and concentrated. The crude product was puriﬁed by column chromatography (silica gel, hexane–ethyl acetate, 9:1) to aﬀord the aldol adduct 78 (9.60 g, 87%) as a colorless oil. 3.3.6

Asymmetric Aldol Reactions Using Chiral Tin(II) Lewis Acid Catalysts

The asymmetric aldol reaction is one of the most powerful tools for the construction of new carbon–carbon bonds by controlling the absolute con-

3.3 Crossed Aldol Reactions Using Silicon Enolates

151

ﬁgurations of newly formed chiral centers [62]. Among asymmetric aldol reactions that mediated by silicon enolate has been extensively studied by many research groups over the past two decades. Stoichiometric Enantioselective Aldol Reaction Many aldehydes react with the (E) silicon enolate [63] derived from propionic acid thioester 79, to give syn aldol adducts in high yield and with perfect stereochemical control, by combined use of tin(II) triﬂate, chiral diamine 80, and dibutyltin acetate (Eq. (41)) [64–66] 3.3.6.1

OSiMe3 RCHO +

SEt

OH O

Sn(OTf)2 + 80 + n-Bu2Sn(OAc)2 R

CH2Cl2, –78 °C

79

N Me

SEt

R

syn

N H 80

OH O +

R2

R3 O O S CF3 TfO O O N

R1

N

anti 70–96% yield syn/anti = 100/0, >98% ee

Sn

O Sn nBu AcO n Bu 81 Assumed three-component Promoter

ð41Þ The formation of an active complex 81 consisting of three components, tin(II) triﬂate, chiral diamine 80, and dibutyltin acetate is assumed in these aldol reactions. The three-component complex would activate both aldehyde and silyl enolate (double activation), i.e. the chiral diamine-coordinated tin(II) triﬂate activates aldehyde while oxygen atoms of the acetoxy groups in dibutyltin acetate interact with the silicon atom of the silicon enolate. Because it has been found that the reaction does not proceed via tin(II) or tin(IV) enolates formed by silicon–metal exchange, silicon enolate is considered to attack the aldehydes directly [65]. The problem of this aldol reaction is that (Z) enolates [63] react with aldehydes more slowly, consequently aﬀording the aldols in lower yield and with lower diastereo- and enantioselectivity. Because optically active molecules containing 1,2-diol units are often observed in nature (e.g. carbohydrates, macrolides, polyethers), asymmetric aldol reaction of the silyl enolate of a-benzyloxythioacetate 82 with aldehydes has been investigated for simultaneous introduction of two vicinal hydroxy groups with stereoselective carbon–carbon bond-formation. It has, interestingly, been found that the anti-a,b-dihydroxy thioester derivatives 83 are

SEt

3 Boron and Silicon Enolates in Crossed Aldol Reaction

152

obtained in high yields with excellent diastereo- and enantioselectivity by combined use of tin(II) triﬂate, chiral diamine 84, and dibutyltin acetate (Eq. (42)) [67]. These results are unusual, because aldol reaction of simple silyl enolate 79 with aldehydes generally aﬀords syn aldol adducts as mentioned above (Eq. (41)). Consideration of the transition states of these aldol reaction leads us to postulate the coordination of the oxygen atom of the silyl enolate 82 by the tin atom of tin(II) triﬂate, which is essential for anti selectivity.

anti-selective aldol reaction

OSiMe3 RCHO

+

SEt

BnO

N N 84 Et Sn(OTf)2 , n-Bu2Sn(OAc)2 CH2Cl2, –78 °C

OH O R

82

SEt OBn syn-83

OH O +

R

SEt OBn anti-83

syn/anti = 2/98–1/99 95–98% ee (anti )

ð42Þ To examine this hypothesis silicon enolate 85, which has bulky tertbutyldimethylsilyl group, was prepared, to prevent coordination of the a oxygen atom to tin(II). As expected, syn aldol 86 is obtained in high stereoselectivity by reaction of the above-mentioned hindered silicon enolate 85, tin(II) triﬂate, a chiral diamine 87, and dibutyltin acetate (Eq. (43)) [68].

syn-selective aldol reaction N Pr OSiMe3 RCHO +

SEt

TBSO

85

N 87 OH O

Sn(OTf)2 , n-Bu2Sn(OAc)2 CH2Cl2, –78 °C

R

SEt OTBS syn-86

OH O +

R

SEt OTBS anti-86

syn/anti = 88/12–97/3 82–94% ee (syn)

ð43Þ It is, therefore, possible to prepare syn and anti aldols selectively when the appropriate protecting group is chosen for the alkoxy part of the molecule. This method has been applied to the synthesis of several monosaccharides including branched, deoxy, and amino sugars [69]. One example is shown below (Scheme 3.5) [69c].

CHO +

3.3 Crossed Aldol Reactions Using Silicon Enolates

153

O CH3

O

N N Me Sn(OTf)2 Bu2Sn(OAc)2

HO 1) OsO4 (cat.) NMO

OH O

+

SEt

OSiMe3

O

OBn 2) H2S gas 85% yield HO H 3C syn/anti = <2/>98 97% ee (anti )

SEt OBn 82

OH OBn 88

O

OH OBn 89 72% yield 88/89 = 72/28

DIBAL-H 88

H2, Pd/C

O CH3

O CH3

OH

CH2Cl2, –78 °C HO OH OBn 71% yield

EtOH quant.

HO

OH

OH OH

6-deoxy-L-talose Scheme 3.5

Stereoselective synthesis of 6-deoxy-l-talose.

Typical Procedure for Stoichiometric Enantioselective Aldol Reaction Using a Chiral Tin(II) Catalyst System (Eq. (44)) [67]

N Et OSiMe3 PhCHO

+

SEt

BnO 82

N 84 OH O

Sn(OTf)2 , n-Bu2Sn(OAc)2 CH2Cl2, –78 °C

Ph

SEt OBn syn-90

OH O +

Ph

SEt OBn anti-90

syn/anti = 1/99 96% ee (anti)

ð44Þ Dibutyltindiacetate (0.44 mmol) was added at room temperature to a solution of tin(II) triﬂate (0.4 mmol) and (S)-1-ethyl-2-[(piperidin-1-yl)methyl]pyrrolidine 84 (0.48 mmol) in dichloromethane (1 mL). The mixture was stirred for 30 min then cooled to 78 C. Dichloromethane solutions (0.5 mL) of the silyl enol ether of S-ethyl 2-benzyloxyethanethioate 82 (0.4 mmol) and benzaldehyde (0.27 mmol) were added successively. The reaction mixture was further stirred for 20 h then quenched with aqueous NaHCO3 . After the usual work up the desired aldol adduct 90 was obtained in 83% yield (syn/anti ¼ 1:99, 96% ee (anti)).

154

3 Boron and Silicon Enolates in Crossed Aldol Reaction

* N TfO

RCHO

N Sn

OTf O

+ OSiMe3 N 79

EtS

*

EtS

OSiMe3 R 92

N

TfO Sn O O EtS 91 + Me3SiOTf

R

* N

N

=

N Me

N H 80

Fig. 3.6

Proposed catalytic cycle.

Catalytic Enantioselective Aldol Reaction As described above, optically active aldol adducts are easily obtained by using a stoichiometric amount of chiral diamine, tin(II) triﬂate, and dibutyltin acetate. To perform the enantioselective aldol reaction by using a catalytic amount of the chiral catalyst, transmetalation of initially formed tin(II) alkoxide 91 to silyl alkoxide 92 with silyl triﬂate is an essential step (Figure 3.6). When the aldol reaction was conducted simply by reducing the amount of the chiral catalyst, aldol adducts were obtained with low stereoselectivity because SnaSi exchange occurs slowly and undesired Me3 SiOTf-promoted aldol reaction aﬀords racemic aldol adducts. To keep the concentration of trimethylsilyl triﬂate as low as possible during the reaction a dichloromethane solution of silyl enolate and aldehyde was added slowly to the solution of the catalyst 80 (20 mol%), and the aldol product 92 was obtained in good yields with high enantioselectivity (Eq. (44)) [70]. Selectivity is improved by using propionitrile as solvent instead of dichloromethane. The rate of metal exchange between SnaSi is faster in propionitrile than in dichloromethane [71]. After the ﬁrst reports of the above-mentioned highly eﬃcient catalytic enantioselective aldol reaction, some groups independently reported catalytic symmetric aldol reactions of silicon enolates with aldehydes using chiral boron [72], titanium [73], zirconium [74], and copper Lewis acids [75], or by transmetalation to chiral Pd(II) enolates [44]. Chiral phosphoramidepromoted aldol reactions of trichlorosilyl enol ethers have been reported as Lewis base-catalyzed asymmetric aldol reactions [76]. 3.3.6.2

References

Typical Procedure for Catalytic Enantioselective Aldol Reaction Using a Chiral Tin(II) Catalyst System (Eq. (45)) [71]

OSiMe3 + RCHO

EtS

N N H Me 80 Sn(OTf)2 (20 mol%) EtCN, –78 °C

79

O

OSiMe3

EtS

ð45Þ

R

92 syn/anti = 89/11~100/0 89~>98% ee (syn)

(S)-1-Methyl-2-[(N-naphthylamino)methyl]pyrrolidine 80 (0.088 mmol) in propionitrile (1 mL) was added to a solution of tin(II) triﬂate (0.08 mmol, 20 mol%) in propionitrile (1 mL). The mixture was cooled to 78 C and a mixture of silyl ketene acetal 79 (0.44 mmol) and an aldehyde (0.4 mmol) was then added slowly over 3 h. The mixture was further stirred for 2 h, then quenched with saturated aqueous NaHCO3 . After the usual work up the aldol-type adduct was isolated as the corresponding trimethylsilyl ether 92.

References 1 Mukaiyama, T.; Inomata, K.; Bull. Chem. Soc. Jpn. 1971, 44,

2

3

4 5

6

3215. The mechanism shown in this paper was corrected as mentioned in reference 2. (a) Mukaiyama, T.; Inomata, K.; Muraki, M.; J. Am. Chem. Soc. 1973, 95, 967. (b) Inomata, K.; Muraki, M.; Mukaiyama, T.; Bull. Chem. Soc. Jpn. 1973, 46, 1807. Review: (a) Mukaiyama, T.; Org. React. 1982, 28, 203. (b) Cowden, C. J.; Paterson, I.; Org. React. 1997, 51, 1. (c) Kim, B. M.; Williams, S. F.; Masamune, S.; in Comprehensive Organic Synthesis, (B. M. Trost, Ed.), Pergamon Press, London, 1991. (d) Evans, D. A.; Nelson, J. V.; Taber, T. R.; Topics Stereochem. 1982, 13, 1. From a-iodoketones and Et3 B: Aoki, Y.; Oshima, K.; Utimoto, K.; Chem. Lett. 1995, 463. From a-iodoketones with 9-BBN and 2,6-lutidine; (a) Mukaiyama, T.; Imachi, S.; Yamane, K.; Mizuta, M.; Chem. Lett. 2002, 698. (b) Mukaiyama, T.; Takuwa, T.; Yamane, K.; Imachi, S.; Bull. Chem. Soc. Jpn. 2003, 76, 813. Interestingly, some boron enolates were isolated by distillation. See: Hoffmann, R. W.; Ditrich, K.; Froech, S.; Tetrahedron 1985, 41, 5517, and 5b. (a) Mukaiyama, T.; Inoue, T.; Chem. Lett. 1976, 559. (b) Inoue, T.; Uchimaru, T.; Mukaiyama, T.; Chem. Lett. 1977,

155

156

3 Boron and Silicon Enolates in Crossed Aldol Reaction

7

8

9 10

11 12 13

14

15

16

17

153. (c) Inoue, T.; Mukaiyama, T.; Bull. Chem. Soc. Jpn. 1980, 53, 174. (a) Masamune, S.; Mori, S.; Van Horn, D. E.; Brooks, D. W.; Tetrahedron Lett. 1979, 1665. (b) Van Horn, D. E.; Masamune, S.; Tetrahedron Lett. 1979, 2229. (c) Evans, D. A.; Cherpeck, R. E.; Tanis, S.; unpublished results. (a) Evans, D. A.; Nelson, J. V.; Vogel, E.; Taber, T. R.; J. Am. Chem. Soc. 1981, 103, 3099. (b) Evans, D. A.; Vogel, E.; Nelson, J. V.; J. Am. Chem. Soc. 1979, 101, 6120. (c) Hirama, M.; Masamune, S.; Tetrahedron Lett. 1979, 2225. (d) Hirama, M.; Garvey, D. S.; Lu, L. D.-L.; Masamune, S.; Tetrahedron Lett. 1979, 3937. The degree of (E),(Z)-selection depends on the structure of carbonyl compounds to some extent. The highest priority designation is assigned to the OBR2 group concerning enolate substituents. (a) Abiko, A.; Liu, J.-F.; Masamune, S.; J. Org. Chem. 1996, 61, 2590. (b) Abiko, A.; J. Synth. Org. Chem. Jpn. 2003, 61, 26. (c) Abiko, A.; Org. Synth. 2002, 79, 116. (a) Brown, H. C.; Ganesan, K.; Dhar, R.; J. Org. Chem. 1993, 58, 147. (b) Abiko, A.; Org. Synth. 2002, 79, 103. Evans, D. A.; Bartroli. J.; Shih, T. L.; J. Am. Chem. Soc. 1981, 103, 2127. (a) Masamune, S.; Choy, W.; Kerdesky, F. A. J.; Imperiali, B.; J. Am. Chem. Soc. 1981, 103, 1566. (b) Meyers, A. I.; Yamamoto, Y.; J. Am. Chem. Soc. 1981, 103, 4278. (c) Meyers, A. I.; Yamamoto, Y.; Tetrahedron 1984, 40, 2309. (d) Oppolzer, W.; Pure Appl. Chem. 1988, 60, 39. (e) Oppolzer, W.; Blagg, J.; Rodriguez, I.; J. Am. Chem. Soc. 1990, 112, 2767. (a) Meyers, A. I.; Yamamoto, Y.; Tetrahedron 1984, 40, 2309. (b) Masamune, S.; Sato, T.; Kim, B.-M.; Wollman, T. A.; J. Am. Chem. Soc. 1986, 108, 8279. (c) Reetz, M. T.; Rivadeneira, E.; Niemeyer, C.; Tetrahedron Lett. 1990, 27, 3863. (d) Corey, E. J.; Kim, S. S.; J. Am. Chem. Soc. 1990, 112, 4976. (e) Gennari, C.; Hewkin, C. T.; Molinari, F.; Bernardi, A.; Comotti, A.; Goodman, J. M.; Paterson, I.; J. Org. Chem. 1992, 57, 5173. (f ) Gennari, C.; Moresca, D.; Vieth, S.; Vulpeti, A.; Angew. Chem., Int. Ed. Engl. 1993, 32, 1618. (a) Inoue, T.; Liu, J.-F.; Buske, D. C.; Abiko, A.; J. Org. Chem. 2002, 67, 5250. (b) Abiko, A.; Liu, J.-F.; Masamune, S.; J. Am. Chem. Soc. 1997, 119, 2586, and references are cited therein. (c) Abiko, A.; Org, Synth. 2002, 79, 116. (d) Abiko, A.; J. Synth. Org. Chem. Jpn. 2003, 61, 24. (a) Walker, M. A.; Heathcock, C. H.; J. Org. Chem. 1991, 56, 5747. (b) Oppolzer, W.; Lienard, P.; Tetrahedron Lett. 1993, 34, 4321. These phenomena had already been experienced in the former research of developing new reactions such as new-type 1,4addition polymerization. (a) Mukaiyama, T.; Fujisawa, T.; Hyugaji, T.; Bull. Chem. Soc. Jpn. 1962, 35, 687. (b) Mukaiyama, T.; Fujisawa, T.; Nohira, H.; Hyugaji, T.; J. Org. Chem. 1962, 27, 3337. (c) Mukaiyama, T.; Fujisawa, T.; Macromol. Synth. 1972, 4, 105. (d) Fujisawa, T.; Tamura, Y.; Mukaiyama, T.; Bull. Chem. Soc. Jpn. 1964, 37, 793. (e)

References

18 19 20

21

22 23

24

25

26 27 28 29 30 31 32

33

34 35

Nohira, H.; Nishikawa, Y.; Mukaiyama, T.; Bull. Chem. Soc. Jpn. 1964, 37, 797. (f ) Mukaiyama, T.; Sato, K.; Bull. Chem. Soc. Jpn. 1963, 36, 99. Mukaiyama, T.; Angew. Chem. Int. Ed. 1977, 16, 817. Brownbridge, P.; Synthesis 1983, 1, 85. (a) Mukaiyama, T.; Narasaka, K.; Banno, K.; Chem. Lett. 1973, 1011. (b) Mukaiyama, T.; Banno, K.; Narasaka, K.; J. Am. Chem. Soc. 1974, 96, 7503. (a) Isler, O.; Schudel, P.; Adv. Org. Chem. 1963, 14, 115. (b) Effenberger, F.; Angew. Chem. Int. Ed. Engl. 1969, 8, 295. House, H. O.; Modern Synthetic Reactions, 2nd edn, W. A. Benjamin, Menlo Park, 1972. (a) Banno, K.; Mukaiyama, T.; Chem. Lett. 1975, 741. (b) Banno, K.; Mukaiyama, T.; Bull. Chem. Soc. Jpn. 1976, 49, 2284. (a) Chan, T. H.; Aida, T.; Lau, P. W. K.; Gorys, V.; Harpp, D. N.; Tetrahedron Lett. 1979, 4029. (b) Dubois, J.-E.; Axiotis, G.; Bertounesque, E.; Tetrahedron Lett. 1984, 25, 4655. (c) Heathcock, C. H.; Hug, K. T.; Flippin, L. A.; Tetrahedron Lett. 1984, 25, 5973. (d) Gennari, C.; Beretta, M. G.; Bernardi, A.; Moro, G.; Scolastico, C.; Todeschini, R.; Tetrahedron 1986, 42, 893. (a) Myers, A. G.; Widdowson, K. L.; J. Am. Chem. Soc. 1990, 112, 9672. (b) Myers, A. G.; Widdowson, K. L.; Kukkola, P. J.; J. Am. Chem. Soc. 1992, 114, 2765. Denmark, S. E.; Lee, W.; J. Org. Chem. 1994, 59, 707. Mukaiyama, T.; Hayashi, M.; Chem. Lett. 1974, 15. (a) Mukaiyama, T.; Ishida, A.; Chem. Lett. 1975, 319. (b) Mukaiyama, T.; Ishida, A.; Chem. Lett. 1975, 1201. Murata, S.; Suzuki, M.; Noyori, R.; J. Am. Chem. Soc. 1980, 102, 3248. Saigo, K.; Osaki, M.; Mukaiyama, T.; Chem, Lett. 1975, 989. Saigo, K.; Osaki, M.; Mukaiyama, T.; Chem, Lett. 1976, 769. (a) Narasaka, K.; Soai, K.; Mukaiyama, T.; Chem. Lett. 1974, 1223. (b) Narasaka, K.; Soai, K.; Aikawa, Y.; Mukaiyama, T.; Bull. Chem. Soc. Jpn. 1976, 49, 779. (c) Saigo, K.; Osaki, M.; Mukaiyama, T.; Chem. Lett. 1976, 163. (d) Danishefsky, S.; Vaughan, K.; Gadwood, R. C.; Tsuzuki, K.; J. Am. Chem. Soc. 1980, 102, 4262. (e) Jung, M. E.; Pan, Y.-G.; Tetrahedron Lett. 1980, 21, 3127. (f ) Jung, M. E.; McCombs, C. A.; Takeda, Y.; Pan, Y.-G.; J. Am. Chem. Soc. 1981, 103, 6677. (g) Heathcock, C. H.; Norman, M. H.; Uehling, D. E.; J. Am. Chem. Soc. 1985, 107, 2797. (a) Ishitani, H.; Ueno, M.; Kobayashi, S.; J. Am. Chem. Soc. 1997, 119, 7153. (b) Kobayashi, S.; Ishitani, H.; Ueno, M.; J. Am. Chem. Soc. 1998, 120, 431. (c) Ishitani, H.; Ueno, M.; Kobayashi, S.; J. Am. Chem. Soc. 2000, 122, 8180, and references are cited therein. [4þ2]cycloaddition reactions, alkylations, acylations, hydroborations, oxidative processes, etc. (a) Mukaiyama, T.; Kobayashi, S.; Murakami, M.; Chem. Lett. 1985, 447. (b) Kobayashi, S.; Murakami, M.; Mukaiyama, T.; Chem. Lett. 1985, 1535.

157

158

3 Boron and Silicon Enolates in Crossed Aldol Reaction 36 Ohshima, M.; Murakami, M.; Mukaiyama, Y.; Chem. Lett.

1985, 1871. 37 Iwasawa, N.; Mukaiyama, T.; Chem. Lett. 1987, 463. 38 (a) Murata, S.; Suzuki, M.; Noyori, R.; Tetrahedron 1988, 44,

39

40

41

42

43 44

45 46 47 48 49

50 51 52

53 54 55

4259. (b) Muraki, C.; Hashizume, S.; Nagami, K.; Hanaoka, M.; Chem. Pharm. Bull. 1990, 38, 1509. (a) Vougioukas, A. E.; Kagan, H. B.; Tetrahedron Lett. 1987, 28, 5513. (b) Gong, L.; Streitwieser, A.; J. Org. Chem. 1990, 55, 6235. (c) Kobayashi, S.; Hachiya, I.; Takahori, T.; Synthesis 1993, 371. (d) Kobayashi, S.; Hachiya, I.; Ishitani, H.; Araki, M.; Synlett 1993, 472. (e) Kobayashi, S.; Hachiya, I.; J. Org. Chem. 1994, 59, 3590. (a) Kobayashi, S.; Nagayama, S.; Busujima, T.; J. Am. Chem. Soc. 1998, 120, 8287. (b) Kobayashi, S.; Hachiya, I.; J. Org. Chem. 1994, 59, 3590. (c) Loh, T.-P.; Chua, G.-L.; Vittal, J. J.; Wong, M.-W.; Chem. Commun. 1998, 861. (d) Manabe, K.; Kobayashi, S.; Synlett 1999, 547. (a) Noyori, R.; Yokoyama, K.; Sakata, J.; Kuwajima, I.; Nakamura, E.; J. Am. Chem. Soc. 1977, 99, 1265. (b) Noyori, R.; Nishida, I.; Sakata, J.; J. Am. Chem. Soc. 1981, 103, 2106. (c) Kuwajima, I.; Nakamura, E.; Acc. Chem. Res. 1985, 18, 181. (d) Nakamura, E.; Yamago, S.; Machii, D.; Kuwajima, I.; Tetrahedron Lett. 1988, 29, 2207. (e) Yamago, S.; Machii, D.; Nakamura, E.; J. Org. Chem. 1991, 56, 2098. (a) Nakamura, E.; Yamago, S.; Machii, D.; Kuwajima, I.; Tetrahedron Lett. 1988, 29, 2207. (b) Yamago, S.; Machii, D.; Nakamura, E.; J. Org. Chem. 1991, 56, 2098. Stork, G.; Hudrlik, P. F.; J. Am. Chem. Soc. 1968, 90, 4462. (a) Sodeoka, M.; Ohrai, K.; Shibasaki, M.; J. Org. Chem. 1995, 60, 2648. (b) Sodeoka, M.; Tokunoh, R.; Miyazaki, F.; Hagiwara, E.; Shibasaki, M.; Synlett 1997, 463. ¨ ger, J.; Carreira, E. M.; J. Am. Chem. Soc. 1998, 120, (a) Kru 837. (b) Fujimura, O.; J. Am. Chem. Soc. 1998, 120, 10032. Transmetalation to boron enolates: Wada, M.; Chem. Lett. 1981, 153. Also see ref. 41c. Denmark, S. E.; Stavenger, R. A.; Acc. Chem. Res. 2000, 33, 432. Kobayashi, S.; Nishido, K.; J. Org. Chem. 1993, 58, 2647. (a) Myers, A. G.; Kephart, S. E.; Chen, H.; J. Am. Chem. Soc. 1992, 114, 7922. (b) Denmark, S. E.; Griedel, B. D.; Coe, D. M.; Schnute, M. E.; J. Am. Chem. Soc. 1994, 116, 7026. Miura, K.; Sato, H.; Tamaki, K.; Ito, H.; Hosomi, A.; Tetrahedron Lett. 1998, 39, 2585. Myers, A. G.; Widdowson, K. L.; J. Am. Chem. Soc. 1990, 112, 9672. (a) Creger, P. L.; Tetrahedron Lett. 1972, 79. (b) Kita, Y.; Tamura, O.; Itou, F.; Yasuda, H.; Kishino, H.; Ke, Y. Y.; Tamura, Y.; J. Org. Chem. 1988, 53, 554. (a) Loh, T.-P.; Feng, L.-C.; Wei, L.-L.; Tetrahedron, 2000, 56, 7309. (b) Lubineau, A.; J. Org. Chem. 1986, 51, 2142. Ge´nisson, Y.; Gorrichon, L.; Tetrahedron Lett. 2000, 41, 4881. Yamamoto, Y.; Maruyama, K.; J. Am. Chem. Soc. 1983, 105, 6963.

References 56 Miura, K.; Nakagawa, T.; Hosomi, A.; J. Am. Chem. Soc.

2002, 124, 536. 57 (a) Fujisawa, H.; Mukaiyama, T.; Chem. Lett. 2002, 182. (b)

58

59 60

61

62

63 64 65 66 67 68 69

70 71 72

Fujisawa, H.; Mukaiyama, T.; Chem. Lett. 2002, 858. (c) Mukaiyama, T.; Fujisawa, H.; Nakagawa, T.; Helv. Chim. Acta 2002, 85, 4518. As another example, 20 mol% of tris(2,4,6trimethoxyphenyl)phosphine catalyzed the aldol reaction of trimethylsilyl ketene acetals. See: Matsukawa, S.; Okano, N.; Imamoto, T.; Tetrahedron Lett. 2000, 41, 103. Fujisawa, H.; Sasaki, Y.; Mukaiyama, T.; Chem. Lett. 2001, 190. Magnesium halide-mediated aldol reactions: (a) Takai, K.; Heathcock, C. H.; J. Org. Chem. 1985, 50, 3247. (b) Uenishi, J.; Tomozane, H.; Yamato, M.; Tetrahedron Lett. 1985, 26, 3467. (c) Bernardi, A.; Cardani, S.; Colonbo, L.; Poli, G.; Schimperna, G.; Scolastico, C.; J. Org. Chem. 1987, 52, 888. (d) Corey, E. J.; Li, W.; Reichard, G. A.; J. Am. Chem. Soc. 1998, 120, 2330. (a) Mukaiyama, T.; Shiina, I.; Iwadare, H.; Sakoh, H.; Tani, Y.; Hasagawa, M.; Saitou, K.; Proc. Jpn. Acad. 1997, 73B, 95. (b) Mukaiyama, T.; Shiina, I.; Iwadare, H.; Saitoh, M.; Nishimura, T.; Ohkawa, N.; Sakoh, H.; Nishimura, K.; Tani, Y.; Hasegawa, M.; Yamada, K.; Saitoh, K.; Chem. Eur. J. 1999, 5, 121. (c) Shiina, I.; Shibata, J.; Ibuka, R.; Imai, Y.; Mukaiyama, T.; Bull. Chem. Soc. Jpn. 2001, 74, 113. Review: (a) Carreira, E. M.; in Comprehensive Asymmetric Catalysis, (Jacobsen, E. N.; Pfaltz, A.; Yamamoto, H.; eds), Springer, Heidelberg, 1999, Vol. 3, p. 998. (b) Mahrwald, R.; Chem. Rev. 1999, 99, 1095. (c) Gro¨ger, H.; Vogl, E. M.; Shibasaki, M.; Chem. Eur. J. 1998, 4, 1137. (d) Nelson, S. G.; Tetrahedron: Asymmetry 1998, 9, 357. (e) Bach, T.; Angew. Chem. Int. Ed. Engl. 1994, 33, 417. The highest priority is assigned to OSiR3 group concerning enolate substituents. Kobayashi, S.; Mukaiyama, T.; Chem. Lett. 1989, 297. Kobayashi, S.; Uchiro, H.; Fujishita, Y.; Shiina, I.; Mukaiyama, T.; J. Am. Chem. Soc. 1991, 113, 4247. Mukaiyama, T.; Kobayashi, S.; J. Organomet. Chem. 1990, 382, 39. Mukaiyama, T.; Uchiro, H.; Shiina, I.; Kobayashi, S.; Chem. Lett. 1990, 1019. Mukaiyama, T.; Shiina, I.; Kobayashi, S.; Chem Lett. 1991, 1902. (a) Kobayshi, S.; Onozawa, S.; Mukaiyama, T.; Chem. Lett. 1992, 2419. (b) Mukaiyama, T.; Anan, H.; Shiina, I.; Kobayahi, S.; Bull. Soc. Chim. Fr. 1993, 130, 388. (c) Mukaiyama, T.; Shiina, I.; Kobayashi, S.; Chem, Lett. 1990, 2201. Mukaiyama, T.; Kobayashi, S.; Uchiro, H.; Shiina, I.; Chem. Lett. 1990, 129. Kobayashi, S.; Fujishita, Y.; Mukaiyama, T.; Chem. Lett. 1990, 1455. (a) Furuta, K.; Maruyama, T.; Yamamoto, H.; J. Am. Chem.

159

160

3 Boron and Silicon Enolates in Crossed Aldol Reaction

73

74 75

76

Soc. 1991, 113, 1041. (b) Parmee, E. R.; Tempkin, O.; Masamune, S.; Abiko, A.; J. Am. Chem. Soc. 1991, 113, 9365. (c) Furuta, K.; Maruyama, T.; Yamamoto, H.; Synlett 1991, 439. (d) Kiyooka, S.; Kaneko, Y.; Kume, K.; Tetrahedron Lett. 1992, 33, 4927. (a) Mikami, K.; Matsukawa, M.; J. Am. Chem. Soc. 1994, 116, 4077. (b) Mikami, K.; Matsukawa, M.; J. Am. Chem. Soc. 1993, 115, 7039. (c) Mikami, K.; Pure Appl. Chem. 1996, 68, 639. (d) Keck, G. E.; Krishnamurthy, D.; J. Am. Chem. Soc. 1995, 117, 2363. (e) Carreira, E. M.; Singer, R. A.; Lee, W.; J. Am. Chem. Soc. 1994, 116, 8837. (f ) Singer, R. A.; Carreira, E. M.; J. Am. Chem. Soc. 1995, 117, 12360. (g) Carreira, E. M.; Singer, R. A.; Lee, W.; J. Am. Chem. Soc. 1995, 117, 3649. Ishitani, H.; Yamashita, H.; Shimizu, H.; Kobayashi, S.; J. Am. Chem. Soc. 2000, 122, 5403. Evans, D. A.; Murry, J. A.; Kozlowski, M. C.; J. Am. Chem. Soc. 1996, 118, 5814. (b) Evans, D. A.; Kozlowski, M. C.; Tedrow, J. S.; Tetrahedron Lett. 1997, 42, 7481. (c) Evans, D. A.; Kozlowski, M. C.; Burgey, C. S.; MacMillan, D. W. C.; J. Am. Chem. Soc. 1997, 119, 7893. (d) Evans, D. A.; MacMillan, D. W. C.; Campos, K. R.; J. Am. Chem. Soc. 1997, 119, 10859. Denmark, S. E.; Winter, S. B. D.; Su, X.; Wong, K.-T.; J. A. Chem. Soc. 1996, 118, 7404. (b) Denmark, S. E.; Wong, K.-T.; Stavenger, R. A.; J. Am. Chem. Soc. 1997, 119, 2333.

161

4

Amine-catalyzed Aldol Reactions Benjamin List 4.1

Introduction

Aldolizations can be catalyzed by both Lewis and Brønsted acids and bases [1]. This catalytic diversity is possible because aldol reactions combine a nucleophilic addition, which is acid-catalyzed, with an enolization, which is catalyzed by both acids and bases (Scheme 4.1). Nature’s aldolases use combinations of acids and bases in their active sites to accomplish direct asymmetric aldolization of unmodiﬁed carbonyl compounds. Aldolases are distinguished by their enolization mode – Class I aldolases use the Lewis base catalysis of a primary amino group and Class II aldolases use the Lewis acid catalysis of a Zinc(II) cofactor. To accomplish enolization under essentially neutral, aqueous conditions, these enzymes decrease the pK a of the carbonyl donor (typically a ketone) by converting it into a cationic species, either an iminium ion (5) or an oxonium ion (8). A relatively weak Brønsted base co-catalyst then generates the nucleophilic species, an enamine- (6) or a zinc enolate (9), via deprotonation (Scheme 4.2). Although chemists also use acids and bases to catalyze aldolizations, the aldolase-like direct catalytic asymmetric aldol reaction remained an illusive challenge for a long time. Indirect aldol reactions utilizing preformed enolate equivalents provided the only viable strategy for catalytic asymmetric aldol synthesis [2–10]. Among the ﬁrst purely chemical direct asymmetric aldol catalysts were bifunctional Lewis acid-Brønsted base metal complexes that resemble class II aldolases [11–17]. Amino acids that mimic the enamine catalysis of class I aldolases have also been studied. Initially, these eﬀorts have concentrated on proline-catalyzed enantiogroup diﬀerentiating intramolecular aldolizations [18–28]. More recently the ﬁrst direct aminecatalyzed asymmetric intermolecular aldol reactions have been described [29–33]. This chapter reviews both asymmetric and non-asymmetric aminocatalytic intra- and intermolecular aldolizations. Not included here are

Modern Aldol Reactions. Vol. 1: Enolates, Organocatalysis, Biocatalysis and Natural Product Synthesis. Edited by Rainer Mahrwald Copyright 8 2004 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 3-527-30714-1

162

4 Amine-catalyzed Aldol Reactions

Brønsted Base Brønsted-or Lewis Acid

O

Brønsted-or Lewis Acid

O

O

OH

H

Lewis Base (Amine)

1

2

3

Scheme 4.1

Catalysis of the aldol reaction.

Class I Aldolases

O

H

4

Class II Aldolases

O

Enzyme B H

N

O

H

N

Enzyme BH

5

6

Enzyme Zn(II) B

Enzyme Zn(II) BH

O

H 7

8

9

Scheme 4.2

Two enzymatic strategies for enolization of carbonyl compounds.

other interesting organocatalysts of the aldol reaction, for example simple tertiary amine Brønsted bases and phase-transfer type quaternary ammonium salts [34–36]. Mechanistically related class I aldolases and catalytic antibodies have been fundamentally important in the development of the reactions described below and are discussed in another chapter of this book.

4.2

Aminocatalysis of the Aldol Reaction

Although aminocatalysis of the aldol reaction via enamine intermediates is an important enzymatic strategy and several bioorganic studies of the subject have appeared, applications in preparative organic synthesis, particularly in intermolecular aldol addition reactions, have been published only sporadically. Despite the often-used Mukaiyama-aldol reaction of enol ethers and Stork’s well-developed enamine chemistry [37, 38], aldolizations of preformed enamines are rare. One report describes Lewis acid-catalyzed aldolizations of preformed enamines with aldehydes that furnish aldol addition products [39]. Aldol condensation reactions of preformed enamines with aldehydes have also been described [40]. Only enamine-catalytic aldolizations, which are primary and secondary amine-catalyzed aldol reactions, will be discussed in this chapter, however.

4.2 Aminocatalysis of the Aldol Reaction

O

O

amine

O

OH

2 aq. buffer

H

H

10

(1)

H 11

O

OH

amine aq. buffer

13 O

O H 14

amine R

12 O 2

(2) 14 O

OH

15

(3)

R

aq. buffer 16

Scheme 4.3

Some amine-catalyzed aldolizations in aqueous buﬀers.

4.2.1

Intermolecular Aldolizations

Class I aldolase-like catalysis of the intermolecular aldol reaction with amines and amino acids in aqueous solution has been studied sporadically throughout the last century. Fischer and Marschall showed in 1931 that alanine and a few primary and secondary amines in neutral, buﬀered aqueous solutions catalyze the self-aldolization of acetaldehyde to give ‘‘aldol’’ (11) and crotonaldehyde (12) (Scheme 4.3, Eq. (1)) [41]. In 1941 Langenbeck et al. found that secondary amino acids such as sarcosine also catalyze this reaction [42]. Independently, Westheimer et al. and other groups showed that amines, amino acids, and certain diamines catalyze the retroaldolization of ‘‘diacetone alcohol’’ (13) and other aldols (Scheme 4.3, Eq. (2)) [43–47]. More recently Reymond et al. [48] studied the aqueous amine catalysis of cross-aldolizations of acetone with aliphatic aldehydes furnishing aldols 16 (Scheme 4.3, Eq. (3)) and obtained direct kinetic evidence for the involvement of enamine intermediates. It is believed that amine-catalyzed aldolizations, in a manner similar to class I aldolase-catalyzed reactions, proceed via a catalytic cycle that involves formation of (A) carbinolamine-, (B) iminium ion-, and (C) enamine intermediates (Scheme 4.4). Either the enamine generation or the subsequent CaC-bond-forming step (D) are rate limiting. Water addition to give a new carbinolamine (E) and its fragmentation (F) close the catalytic cycle. Houk and Bahmanyar have used density functional theory calculations to investigate transition states of the CaC-bond-forming step in aminecatalyzed aldolizations (Scheme 4.5) [49, 50]. Accordingly, the cyclic transition state of the primary amine-catalyzed aldol reaction has a half-chair conformation featuring NaHaO hydrogen-bonding for charge stabilization (A). In secondary amine-catalyzed aldol reactions the calculated zwitterionic transition state (B) is somewhat higher in energy and leads to an oxetane

163

164

4 Amine-catalyzed Aldol Reactions

+ H+ -H2O

N

R1

OH

N R1

B

R1 O

-H+ A

C N

HN R2

R1 OH

O

R2

R1 F

D

R1 E N R2 +H2O H OH OH O

+ R2CHO N

R1

Scheme 4.4

The enamine catalytic cycle for aldolizations.

d+

N H H

O H

N H

+

d

d-

O H

A

N H

d-

O H3O+ H

B

C

Scheme 4.5

DFT calculation-based transition states for primary (A) and secondary (B, C) aminecatalyzed aldolizations.

intermediate in an overall [2 þ 2] cycloaddition reaction. When an external proton source was included (C), the reaction was found to have no barrier and led directly to the iminium ion intermediate (Scheme 4.5). Aldehyde Donors Despite these mechanistic and theoretical studies, intermolecular aminecatalyzed aldolizations have only rarely been used on a preparative scale. A few noteworthy exceptions in which aldehydes are used as donors are shown in Scheme 4.6 [51–55]. These reactions are often performed neat or in the presence of small amounts of an organic solvent. The catalyst usually used is either a primary or secondary amine, a combination of an amine with a carboxylic acid, or simply an amino acid. These catalyst systems have previously been used in the Knoevenagel condensation and it is apparent that synthetic amine-catalyzed aldolizations originate from Knoevenagel’s chemistry [56]. A Mannich-type condensation mechanism involving an iminium ion electrophile similar to the aminocatalytic Knoevenagel reaction has recently been proposed for the amine-catalyzed self-aldolization of propionaldehyde (Eq. (6)) [55]. Although this mechanism is not unreasonable it should be 4.2.1.1

4.2 Aminocatalysis of the Aldol Reaction

2

O

Piperidine (1 mol%)

17

AcOH (1 mol%) EtOH

H Ph

O

19

20 1 eq

O

AcOH (10 mol%)

O

H

CO2Me

H 19

22

1-2 eq

1 eq

OH CO2Et

(2)

21 50% Dimethylamine (10 mol%)

O

(1)

Ph 18 35%

H

CO2Et

1.2 eq

Ph

H

Diethylamine (10 mol%)

O

H

O

O

OH

H

CO2Me

O

MeNHCH2CO2H (10 mol%),

(3)

23 56%

H

CO2Me

then Ac2O, ∆

(4)

24 84%

H 25 1 eq

Pyrrolidine (10 mol%)

O

O H C6H13

2

Ph

26 1.5 eq

AcOH (2 mol%)

O

Pyrrolidine (20 mol%)

19

PhCO2H (10 mol%) PhMe

H

O H

Ph C6H13 27 75%

(5)

O (6)

H 28 90%

Scheme 4.6

Preparative amine-catalyzed intermolecular aldolizations with aldehyde donors.

noted that aldol addition products are often isolated in such reactions (e.g. Eqs. (2)–(4)). Obviously, a Mannich-type mechanism has to be excluded in such cases. In general both, the iminium ion Mannich-type mechanism and the aldolase-type enamine mechanism are plausible and seem to be energetically comparable. Which of these two mechanisms actually operates in a given reaction depends on the donor component: In Knoevenagel-type reactions where easily enolized low-pK a malonate-type donors are used, primary and secondary amine catalysis most probably involves activation of the carbonyl acceptor as an iminium ion. In amine-catalyzed aldol reactions of aldehyde and ketone donors, which have higher pK a (i.e. acetone pK a ¼ 20, propionaldehyde pK a ¼ 17), activation of the donor as an enamine predominates.

165

4 Amine-catalyzed Aldol Reactions

166

O

O

O

Diethylamine (cat.)

NH

O

(1)

O EtOH, rt

N H 30 1 eq

Br 29 1 eq

Br Piperidine (cat.)

N H 33 1 eq O

32 1 eq

31 87%

OHC

O

OH

O (2) NH

EtOH, 175°C 5 min

34 61%

Piperidine, AcOH

O

O

O (3)

OHC reflux, 60 h 14

36

35 1 eq

Solvent O OHC O 37 1 eq

Ph

O Ph (4)

Ph C6H6

38 2 eq O

57%

Piperidine, AcOH (cat.)

39 O 40%

O CO2H

14 Solvent

40 1 eq

O

OH

(5)

CO2H

rt, 48 h

Morpholine (10 mol%)

O

O

Piperidine

41 O

O N (6)

H 26

42 1 eq O

Ph

1 eq OHC

F3C 46

45 2 eq

1 eq

Ph

i. C6H6,reflux, 10h ii. distillation

43

Piperidine, AcOH (cat.) THF, 0°C

Ph 44 8%

52% O

(7)

F3C 47 80%

Scheme 4.7

Some secondary amine-catalyzed intermolecular aldolizations with ketone donors.

Ketone Donors Ketone donors have only rarely been used in amine-catalyzed aldolizations. Selected examples that have been published over the last 70 years are shown in Scheme 4.7 [57–68]. The aminocatalysts used are the same that are also used when aldehydedonors are employed. Aldehydes and ketones can be used as acceptors and 4.2.1.2

4.2 Aminocatalysis of the Aldol Reaction

O

O

EnolendoAldolization

O

OH (1)

n

n

48 O O

O

49 OH

EnolexoAldolization

(2) n

n

50

51

Scheme 4.8

The two modes of intramolecular aldolization.

aldol addition and condensation products have been isolated. Apparently, aunbranched aliphatic aldehydes have never been used as acceptors in nonasymmetric reactions with ketone donors. Morpholine-catalyzed Aldolization of Cyclopentanone with Benzaldehyde to Give Enone 43 [40]. Cyclopentanone (37 g, 0.5 mol), morpholine (4.4 g, 0.05 mol, 10 mol%), and benzaldehyde (53 g, 0.5 mol) were heated under reﬂux in dry benzene (100 mL) for 10 h using a Dean–Stark trap. The resulting crude product was distilled to furnish enone 43 (45 g, 52%, Bp.0:05 ¼ 115– 120 C). 4.2.2

Intramolecular Aldolizations

There are two types of intramolecular aldol reaction, enolendo and enolexo aldolizations (Scheme 4.8, Eqs. (1) and (2), respectively). Both types can be catalyzed by amines and several examples have been published. Enolexo Aldolizations One of the earliest aminocatalytic intramolecular aldolizations was applied in Woodward’s 1952 total synthesis of steroids [69]. A 5-enolexo aldol condensation established the D-ring of the steroid skeleton. Hexanedial 52 on treatment with a catalytic amount of piperidinium acetate gave cyclopentenal 53 (Scheme 4.9). 4.2.2.1

CHO CHO

H H O

CHO

Piperidine (cat.) AcOH (cat.) C6H6, 1h, 60 °C 66%

52

Scheme 4.9

An intramolecular amine-catalyzed aldolization in Woodward’s total synthesis of steroids.

H H O

53

167

168

4 Amine-catalyzed Aldol Reactions

CHO H

OHC

H

Bn2NH.TFA (20 mol%)

OHC

OMEM O

THPO

OMEM 50°C, C6H6 64%

THPO

O

54

55 i. OsO4 ii. NaIO4

O

OHC

Pyrrolidine, AcOH Toluene, RT

S

57 OHC

S

58% OTBS 58

Piperidine, AcOH (cat.) C6H6, ∆

CHO

(3)

S

S

OHC

(2)

O

iii. Pyrrolidine, AcOH 69%

56

OHC

(1)

TBSO 59 OHC (4)

59%

MeOC 60 CHO

61 OAc

Piperidine AcOH (cat.) THF, H2O

MeOC

OHC

OAc (5)

HO 62

63

Scheme 4.10

Amine-catalyzed 5-enolexo aldolizations.

Since then amine-catalyzed 5-enolexo aldolizations have been applied in a variety of other synthetic contexts (Scheme 4.10) [70–75]. Typically, they constitute part of a strategy for the ring contraction of cyclohexenes to cyclopentenes via oxidative cleavage and aldolization (Eq. (2)). As in intermolecular aldol reactions, amine-catalyzed 5-enolexo aldolizations can either lead to aldol condensation (Eqs. (1)–(4)) or addition products (Eq. (5)). These reactions usually involve an aldehyde group as the aldol donor; the corresponding amine-catalyzed 5-enolexo aldolizations of ketone donors have so far not been realized. The acceptor carbonyl group can either be an aldehyde or a ketone. The regioselectivity of aminocatalytic aldolizations of ketoaldehydes such as 60 and 62 contrasts with that of the corresponding Brønsted base-

4.2 Aminocatalysis of the Aldol Reaction Piperidine, AcOH(cat.) C6H6, ∆

OHC (1)

CHO 65

O aq. KOH, Et2O, rt

64

O (2) 66

Scheme 4.11

Contrasting regioselectivity of amine- and base-catalyzed 5-enolexo aldolizations.

catalyzed processes. Similarly, ketoaldehyde 64 on treatment with piperidinium acetate in benzene gave only enal 65, whereas treatment with aq. KOH in ether furnished mainly its regioisomer 66 (Scheme 4.11) [76]. The regioselectivity of the amine-catalyzed process might reﬂect a high aldehyde enamine (67) to ketone enamine (68) ratio, because of steric hindrance (A 1;3-strain) in the ketone enamine (Scheme 4.12). In the base-

O Piperidine

CHO KOH

64 OK

N O

Me

O

Me

H

H 65

H

H 70

67

OK N O

H

H Me

68 Scheme 4.12

Rationalization of the contrasting regioselectivity of amine- and base-catalyzed aldolizations.

O

H

66

69

Me

169

170

4 Amine-catalyzed Aldol Reactions

Piperidine (10%.) p-TsOH (10%) C6H6, 16h, RT

CHO COMe

CHO (1)

62%

OTBS 71

OTBS 72 CHO

CHO Et2NTMS

CHO

O

Ph O 73

O

OAc

COMe CHO

MeO

74 OAc

Piperidine AcOH C6H6, ∆

Piperidine AcOH C6H6, ∆

CHO COMe

MeO

(4)

CHO

(5)

Piperidine AcOH toluene, ∆

O

(6)

41%

81

82 CHO CHO

83

OMe SiMe2Ph 78 CHO

THPO 80 H

H

O

AcO MeO2C

CHO

65%

77 SiMe2Ph + CHO Bn2NH2 CF3CO2 ° toluene,110 C TBSO CHO > 70% 79

(3)

76

OMe

THPO

COMe

25%

75

TBSO

(2)

OH Ph

O

40%

CHO

Bn2NH2+ CF3CO2benzene AcO MeO2C 54%

CHO

Scheme 4.13

Amine-catalyzed 6-enolexo aldolizations.

(7)

O OH 84

4.2 Aminocatalysis of the Aldol Reaction

CO2Et 2

CH2O

MeOC 85

i. Piperidine

MeOC

COMe CO2Et

EtO2C 87

86

(1)

O

O

ii. NaOEt 50%

CO2Et

CO2Et

EtO2C 88

89

OH

O CO2Et 2

S

Piperidine CO2Et

EtO2C

CHO

MeOC 85

171

>95% S

90 91

Scheme 4.14

Amine-catalyzed 6-enolendo aldolizations in the synthesis of Hagemann’s ester (89) and aldol 91.

catalyzed reaction the more stable and sterically relatively unhindered ketone enolate (69) leading to the opposite regioisomer dominates and, in addition, its aldolization might be faster than that of the less nucleophilic aldehyde enolate (70) (Scheme 4.12). 6-Enolexo aldolizations giving aldol addition or condensation products have also been described, examples are shown in Scheme 4.13 [77–82]. Typically, aldehydes function as the aldol donor in these reactions but ketones can also be used (Eq. (6)). Both ketones and aldehydes can play the role of the acceptor carbonyl group. Enolendo Aldolizations More than the corresponding enolexo-aldolizations, amine-catalyzed 6enolendo aldol reactions have frequently been used synthetically. An early example is the synthesis of Hageman’s ester (89) from formaldehyde and acetoacetate via piperidine-catalyzed Knoevenagel–Michael–aldol tandemreaction followed by decarboethoxylation (Scheme 4.14, Eq. (1)) [83]. Related product 91 was obtained when thiophene-carbaldehyde 90 was used (Eq. (2)) [84]. 4.2.2.2

Piperidine-Catalyzed Synthesis of Aldol 91. A mixture of thiophene-2carbaldehyde (90, 8 g, 71 mmol) and ethyl acetoacetate (85, 20 g, 154 mmol, 2.1 equiv.) was cooled to 0 C and treated with ﬁve drops of piperidine under stirring. After 48 h at room temperature a solid was formed, isolated by ﬁltration, and recrystallized twice from diethyl ether to furnish aldol 91 (ﬁne colorless needles Mp. ¼ 106 C) in almost quantitative yield.

(2)

172

4 Amine-catalyzed Aldol Reactions

CHO

HO

O COMe Piperidine BzOH

O

O O

HO O

O O

92 Piperidine AcOH (cat.)

CO2Et O O COMe 94 CO2Et

Pyrrolidine AcOH (cat.)

COMe

O O

(2)

O 95

CO2Et (3)

MeOH,H2O, ∆ 84

COMe 96

O 97 CO2Et

Piperidine AcOH (cat.)

CO2Et COMe

C6H6, ∆ 93%

COMe 98 COMe

H O 100 N

(4) O

Pyrrolidine AcOH (cat.)

99

THF, MeOH, 23°C 93%

H

Cl

EtO2C O COMe 102

93 CO2Et

C6H6, ∆ 78%

CHO

CHO

(1)

Pyrrolidine

O (5) O

101

Cl

EtO2C N

C6H6, ∆ 85

(6)

O 103

Scheme 4.15

Selected amine-catalyzed 6-enolendo aldolizations.

Other examples of amine-catalyzed 6-enolendo aldolizations are shown in Scheme 4.15 [85–90]. In enolendo aldolizations the aldol donor is always a ketone whereas both aldehyde and ketone carbonyl groups can serve as the acceptor. Probably the most important aminocatalytic 6-enolendo aldolizations are used in the syntheses of bicyclic ketones 106 and 109 (the Wieland– Miescher ketone) from triketones 104 and 108 via pyrrolidine-catalyzed cycloaldolization (Scheme 4.16) [91–93]. Enones 106 and 109 are important intermediates in natural product synthesis, particularly as AB- and CDfragments in the synthesis of steroids. It has been convincingly demon-

4.3 Asymmetric Aminocatalysis of the Aldol Reaction

O

O O

O

105

OH O

108

OH O

O

O 106 O

N H O

n

O

104 (n=1) 107 (n=2) 110 (n=3) O

111

OH

O 109

O

O 112

Scheme 4.16

Pyrrolidine-catalyzed 6-enolendo aldolizations.

strated by Spencer et al. that these reactions involve enamine intermediates [94–97].

4.3

Asymmetric Aminocatalysis of the Aldol Reaction

The ﬁrst example of an aminocatalytic asymmetric aldol reaction was the Hajos–Parrish–Eder–Sauer–Wiechert cyclization, a proline-catalyzed enantiogroup-diﬀerentiating 6-enolendo aldolization of some di- and triketones [18–21, 27, 28]. Discovered in the early 1970s this reaction was the ﬁrst example of a highly enantioselective organocatalytic process. Although the signiﬁcance of the Hajos–Parrish–Eder–Sauer–Wiechert reaction was recognized almost from the beginning, it was another 30 years until the ﬁrst proline-catalyzed intermolecular aldol reactions were described [29–33]. 4.3.1

Intramolecular Aldolizations Enolendo Aldolizations Hajos and Parrish at Hoﬀmann La Roche discovered that proline-catalyzed intramolecular aldol reactions of triketones such as 104 and 107 furnish aldols 105 and 108 in good yields and with high enantioselectivity (Scheme 4.17). Acid-catalyzed dehydration of the aldol addition products then gave condensation products 106 and 109 (Eqs. (1) and (2)). Independently, Eder, Sauer, and Wiechert at Schering AG in Germany directly isolated the aldol condensation products when the same cyclizations were conducted in the presence of proline (10–200 mol%) and an acid co-catalyst (Eqs. (3) and (4)). 4.3.1.1

173

174

4 Amine-catalyzed Aldol Reactions

O

O

O

(S )-Proline (3 mol%)

O p -TsOH (1)

DMF, rt 20h 100%

O 104 O

O

(S )-Proline (3 mol%)

O

DMF, rt 72h 52%

O 107

O

O

O

O 107

OH 105 93% ee O

O 106 O

p -TsOH O

C6H6 OH 108 74% ee

O (3) O 106 84% ee O

(S )-Proline (47 mol%) 1N HClO4 CH3CN, 80°C, 25h 83%

(2) O 109

(S )-Proline (47 mol%) 1N HClO4 CH3CN, 80°C, 20h 87%

O 104 O

C6H6

(4) O 109 71% ee

Scheme 4.17

Hajos–Parrish–Eder–Sauer–Wiechert reactions.

A proline-catalyzed cyclization of ketone 107 to give enone 109 directly without added acid was described by Fu¨rst et al. [28]. Proline-Catalyzed Cyclization of Triketone 107 to Give the Wieland–Miescher Ketone (109) [28]. Crude oily trione 107 (ca. 2 mol) in DMSO (2 L) was treated with (S)-proline (11.5 g, 0.1 mol, 5 mol%) and the resulting mixture was stirred under argon at 18 C for 5 days. The solvent was removed at 70 C (0.1 torr) and the resulting dark green oil (420 g) was adsorbed on a column charged with silica (3 kg) Elution with 17% ethyl acetate in hexanes aﬀorded 276 g product which was further puriﬁed by distillation (b.p. 108– 112 C/0.25 torr, bath temp. 150 C) to give the enantiomerically enriched ketone 109 (255 g, 1.43 mol, 72%) in 70% ee (½aD ¼ þ70 ). Proline-catalyzed enolendo aldolizations have been applied to a number of substrates, most often in steroid synthesis. Selected products from such Hajos–Parrish–Eder–Sauer–Wiechert reactions are shown in Scheme 4.18 [98–104]. Although several diﬀerent catalysts have been studied in such enolendo aldolizations, proline has typically been preferred. It can, however, be advan-

4.3 Asymmetric Aminocatalysis of the Aldol Reaction

O

O

O

O

OH 113 89%, 90% ee

PhS

O

O SPh

118

116 80%, 27% ee O

OH

60%, 57% ee

64%, 90% ee

OH

PhS

O

O

120

119

47%, 16% ee

O

O

117 35% (ee n.d.)

115 76%, 81% ee

O

O

O N

2

114 71%, >99% ee

F3C

O

O

OH

O

O

O

175

O

OH 122

121 49%, 34% ee

64%, 90% ee

Scheme 4.18

Selected products from proline-catalyzed 6-enolendo aldolizations.

O (S)-Proline 1N HClO4 N

O

O

O

123

CH3CN, 80°C, 10d 67% (S)-Phenylalanine 1N HClO4

125

(2)

O

32% ee

(3)

126

95% ee

(4)

DMSO, 65°C, 5d (S)-Phenylalanine 1N HClO4

O

86% ee 124

(S)-Proline O

(1)

N

CH3CN, 80°C, 40h 82%

O

27% ee O

CH3CN, 80°C, 40h

O

Scheme 4.19

Compared with proline, phenylalanine is a superior catalyst for some intramolecular aldolizations.

tageous to use primary amino acid catalysts such as phenylalanine, particularly when non-methyl ketones are employed. For example, Danishefsky et al. found that the proline-catalyzed cyclization of triketone 123 furnished product 124 in 27% ee whereas 86% ee was obtained when phenylalanine was used as the catalyst (Scheme 4.19, Eqs. (1) and (2)) [105]. Agami et al.

176

4 Amine-catalyzed Aldol Reactions

O (S)-Proline (10 mol%) DMF

O R

O

5d <30% yield

R

127

128

R

ee %

Ph n-C5H11

47 20

Me

42

i-Pr

8

t-Bu

0

Scheme 4.20

Enantiogroup-diﬀerentiating aldol cyclodehydrations of 4-substituted 2,6heptanedions 127.

O O N HO N OH H CO2

O2C

O O

O H H N

CO2

H O

O H N H O

N H O

O O

O Crystal Surface

A

B

Hajos Model

Agami Model

D Swaminathan Model

C Houk Model

Scheme 4.21

Mechanistic models proposed for the Hajos–Parrish–Eder–Sauer–Wiechert reaction.

made similar observations in the cyclization of ketone 125 (Eqs. (3) and (4)) [106]. Attempts have been made to expand the scope of the Hajos–Parrish– Eder–Sauer–Wiechert reaction to an enantiogroup diﬀerentiating aldolization of acyclic 4-substituted 2,6-heptandiones 127 to give cyclohexenones 128 (Scheme 4.20) [107, 108]. The yield and enantioselectivity of this reaction are relatively modest compared with the parent cyclo-aldolization. Several diﬀerent mechanisms have been proposed for the Hajos–Parrish– Eder–Sauer–Wiechert reaction (Scheme 4.21). These include the original Hajos model (A), which assumes activation of the acceptor carbonyl as a carbinolamine [21]; the Agami model (B), in which a side-chain enamine reacts with a ring carbonyl group, mediated by a second proline molecule [109–114]; Swaminathan’s model (C), proposing a heterogeneous mechanism [115]; and the Houk model (D), in which the side-chain enamine reacts with the ring acceptor carbonyl group, under concomitant activation via hydrogen-bonding to proline’s carboxylic acid group [25, 116, 117]. Model D has recently been supported experimentally [118]. It was shown that, in contrast to earlier reports [119], there are no non-linear (Scheme

4.3 Asymmetric Aminocatalysis of the Aldol Reaction

O

O

i. (S)-Proline (20 mol%) ii. H2SO4

O

O

O

104 O

106 O

O 107

O

(S)-Proline (20 mol%) O 109

Scheme 4.22

No non-linear eﬀects in proline-catalyzed Hajos–Parrish–Eder–Sauer–Wiechert reactions.

4.22) or dilution eﬀects in the asymmetric catalysis, and that prolinecatalyzed aldolizations are ﬁrst order in proline. Enolexo Aldolizations Although amine-catalyzed enolexo aldolizations (Section 4.2.2.1) are relatively common and catalytic asymmetric enolendo aldolizations have been known for three decades (Section 4.3.1.1), the ﬁrst catalytic asymmetric enolexo aldolizations were developed only very recently. It was discovered that a variety of achiral heptanedials (129) and 7-oxoheptanal on treatment with a catalytic amount of (S)-proline furnished anti aldols 130 with excellent enantioselectivity (Scheme 4.23) [120–122]. Whereas non-asymmetric amine-catalyzed enolexo-aldolizations often give 4.3.1.2

177

178

4 Amine-catalyzed Aldol Reactions

OH (S)-Proline (10 mol%)

OHC

OHC

R' R

R'

CH2Cl2, r.t. 8-16h

129

Dicarbonyl

OHC R 130

Yield %

ee%

dr

99

10:1

98

>20:1

97

>20:1

75 : 89 : >95 : 8

22 : 5 : 5:1

Products

OH OHC

OHC

95

OHC

130a

129a

OH OHC

OHC

Me Me

Me Me

OHC 74 130b

129b

OH OHC

OHC

OHC 75

Me Me

Me Me 130c

129c

OHC

OH

OHC

OHC 76

Me

Me

129d

130d

OH OHC

OHC 88

Me

Me

OHC Me

O

129f

1:1

99

2:1

Me 130e

meso-129e

OHC Me

99

HO Me 92

OHC

130f

Scheme 4.23

Proline-catalyzed enantioselective 6-enolexo-aldolizations.

aldol condensation products (Section 4.2.2.1), the corresponding prolinecatalyzed process selectively provides the aldol addition products. The corresponding 5-enolexo aldolizations are much less stereoselective. For example, treating hexanedial (131) with a catalytic amount of (S)-proline fur-

4.3 Asymmetric Aminocatalysis of the Aldol Reaction

CHO

OHC

(S)-Proline (10%) CH3CN, r.t., 7h

131

OH

OH

OHC

OHC

85%

132a 79% ee

2:1

132b 37% ee

Scheme 4.24

A proline-catalyzed 5-enolexo-aldolization.

nished aldol 132 with only modest diastereo- and enantioselectivity (Scheme 4.24). It can be expected that proline-catalyzed enolexo-aldolizations will ﬁnd utility in natural-product synthesis and that new catalysts will eventually be discovered that lead to improved selectivity in the potentially very useful 5-enolexo variant. 4.3.2

Intermolecular Aldolizations Ketone Donors The ﬁrst amine-catalyzed, asymmetric intermolecular aldol reactions were developed by List et al. in 2000 [29–33]. Initially it was found that excess acetone in DMSO containing sub-stoichiometric amounts of (S)-proline reacted with some aromatic aldehydes and isobutyraldehyde to give the corresponding acetone aldols (134) with good yields and enantioselectivity (Scheme 4.25). Particularly high ee were achieved with a-branched aldehydes. Similarly to the intramolecular enolendo variant, the only sideproduct in proline-catalyzed intermolecular aldol reactions are the condensation products (Scheme 4.25). 4.3.2.1

Proline-catalyzed Aldolization of Acetone with Isobutyraldehyde to Give Aldol 134f [29]. A mixture of (S)-proline (2.3 g, 0.02 mol, 20 mol%) and isobutyraldehyde (7.2 g, 0.1 mol) was stirred in acetone (40 mL) and dry DMSO (160 mL) at room temperature. The mixture became completely homogenous within ca. 1 h. After 36 h at room temperature (S)-proline (ca. 2.0 g, 87%) precipitated and was recovered for reuse via ﬁltration, washing with ethyl acetate, and drying. The combined organic solvents were washed with half-concentrated ammonium chloride solution and the aqueous layers were back-extracted with ethyl acetate. The organic layers were dried (MgSO4 ), ﬁltered, and concentrated. Column chromatography on SiO2 (20% ethyl acetate in hexanes) furnished (R)-4-hydroxy-5-methyl-hexan-2-one (134f, 12.0 g, 0.92 mol, 92%) as a colorless liquid. The ee was determined to be 96% by chiral stationary phase HPLC analysis (Chiralpak AS, 2% iPrOH in hexanes, 1 mL min1 , tðRÞ ¼ 18:0 min, tðSÞ ¼ 20:1 min).

179

180

4 Amine-catalyzed Aldol Reactions

(S)-Proline (20-30 mol%)

O RCHO

Aldehyde

Yield %

134 ee%a

Product O

OHC

R

DMSO, r.t.

133

14 excess

OH

O

OH 76

68 134a O OH

NO2

133a OHC

NO2 72b

62 133b OHC

O

134b OH 65

74 133c

134c

Br Cl

O

OH

OHC

Br Cl 69

94 134d 133d

O

OH

OHC 77

54 134e 133e O OHC

OH 96

97 134f

133f

O

OH

OHC 84

63 134g 133g OHC

O

OH >99

81 134h

133h OHC 133i

O

OH >99

85 134i

a)The enantiomeric excess was typically determined using chiral stationary phase HPLCanalysis (Chiralpak AS, not AD as originally reported). b) The originally reported ee of 134b (60%) has been corrected.

Scheme 4.25

Proline-catalyzed asymmetric intermolecular aldolizations of acetone with aromatic- and a-branched aldehydes.

4.3 Asymmetric Aminocatalysis of the Aldol Reaction

O RCH2CHO 14 excess Aldehyde

(S)-Proline (10-20 mol%)

Yield %

136

Product

67

135a

136a OH

O

73

35 135b

136b O

OHC

OH 72

34 135c

136c O

OHC

OH

34 135d

OH 36

22 135e

73

136d O

OHC

ee%

OH

31

OHC

OH R

Acetone or CHCl3, r.t.

135

O

OHC

O

136e

Scheme 4.26

Proline-catalyzed aldolizations of acetone with a-unbranched aldehydes.

a-Unbranched aldehydes (135) proved to be extremely challenging substrates under the original conditions. Self-aldolization of the aldehyde was a major side reaction. It was found later that by using acetone or acetone– CHCl3 mixtures instead of the commonly used DMSO as solvents, and 10– 20 mol% proline as catalyst, the cross-aldol products (136) could be isolated in modest yields and with acceptable enantioselectivity (Scheme 4.26) [123]. The new reaction conditions eﬀectively suppressed aldehyde selfaldolization. The main side product was now the corresponding acetone cross aldol condensation product, typically formed in comparable yields with the desired aldol addition product. Proline-catalyzed Aldolization of Acetone with n-Hexanal to Give Aldol 136b [31]. n-Hexanal (2.40 mL, 20 mmol) and (S)-proline (230 mg, 2 mmol, 10 mol%) were stirred in 100 mL dry acetone for 168 h. Silica gel (ca. 5 g) was added and the mixture was evaporated. The residue was poured on to a preloaded silica gel column and chromatographed with hexanes–ethyl acetate (4:1) to give (E)-non-3-en-2-one (1.37 g, 49%) and aldol 136b (1.11 g, 35%) in 73% ee (AS, 2% 2-propanol–hexanes, 1 mL min1 , tðRÞ ¼ 13:3 min,

181

182

4 Amine-catalyzed Aldol Reactions

O

1. TBSCl 2. ArNTf2

OH

OTf OTBS

57% 136d

137 SnBu3 Pd0 (cat.)

OH

TBAF

95%

OTBS

90% (S )-Ipsenol (139)

138

Scheme 4.27

Asymmetric synthesis of (S )-ipsenol.

O O

14

O (S)-Proline (30 mol%)

O H

O O

O

DMSO

O 140

OH O O O

O O

141 (>99% de)

Scheme 4.28

Application of a proline-catalyzed intermolecular aldolization in the synthesis of carbohydrate derivatives.

tðSÞ ¼ 15:7 min). Almost identical results were obtained using an aqueous work-up (phosphate buﬀered saline–ethyl acetate). At the time these relatively modest results represented the state-of-theart in direct catalytic asymmetric aldolizations with a-unbranched aldehyde acceptors. Neither Shibasaki’s nor Trost’s bimetallic catalysts, the only alternative catalysts available, gave superior results. Moreover, even nonasymmetric amine-catalyzed cross aldolizations with a-unbranched acceptors are still unknown. That the practicality of the process can compensate for the modest yield and enantioselectivity was illustrated by a straightforward synthesis of the natural pheromone (S)-ipsenol (139) from aldol 136d, featuring a high-yielding Stille coupling (Scheme 4.27). A highly stereoselective proline-catalyzed acetone aldol reaction has recently been used in the synthesis of sugar derivatives such as 141 (Scheme 4.28) [124]. Proline-catalyzed aldolizations with ketones other than acetone have also been described. Because the ketone component is typically used in large excess, one is limited to readily available and inexpensive smaller ketones such as butanone, cyclopentanone, cyclohexanone, and hydroxyacetone. Depending on the aldehyde component excellent enantio- and

4.3 Asymmetric Aminocatalysis of the Aldol Reaction O

O

(S)-Proline R'CHO

R

R 142

(10-30%) 143 Products

O

OH

144a 85% ee O OH

O

O

OH

O

OH 149a 97% ee OH Cl O

1:1

41

7:1

68

>20:1

77

3:1

62

>20:1

38

1.7:1

95

1.5:1

OH

147b 20% ee O

OH

OH

148a >99% ee OH

85

146b OH

OH

O

dr (anti/syn)

OH

O

147a 95% ee O

Yield %

145b 89% ee

OH

146a 97% ee OH

R' R 144-150

144b 76% ee O OH

145a 86% ee O

R

OH

148b O

OH

OH 149b 84% ee OH Cl O

OH

OH

150a 67% ee

150b 32% ee

Scheme 4.29

Examples of proline-catalyzed intermolecular aldolization with ketone donors other than acetone.

183

184

4 Amine-catalyzed Aldol Reactions

(anti)-diastereoselectivity can be achieved in such reactions (Scheme 4.29) [30, 31, 125]. Interestingly, the stereoselectivity of reactions of cyclohexanone with isobutyraldehyde and benzaldehyde were ﬁrst predicted by using density functional theory calculations on models based on Houk’s calculated transition state of the Hajos–Parrish–Eder–Sauer–Wiechert reaction [125]. The transition states of inter- and intramolecular aldol reactions are almost superimposable and readily explain the observed enantiofacial selectivity. Relative transition state energies were then used to predict the diastereo- and enantioselectivity of the proline-catalyzed reactions of cyclohexanone with isobutyraldehyde and benzaldehyde. The predictions are compared with the experimental results in Scheme 4.30. The good agreement clearly validates the theoretical studies, and provides support for the proposed mechanism. Additional density functional theory calculation also support a similar mechanism [126, 127]. The exceptionally high stereoselectivity found in proline-catalyzed aldol reactions involving hydroxyacetone has recently been used in a total synthesis of brassinolide (156), a steroidal plant-growth regulator [128]. Readily available aldehyde 151 underwent proline-catalyzed aldolization with hyO H

H O

NO O H R

H

O

OH

H O

O

HO N O OH R

R

OH

R

R

R = i-Pr

98.5% (>99%)

<1% (<1%)

R = Ph

46.5% (54.5%)

44% (44.5%)

O

Theoretical prediction and experimental veriﬁcation of diastereo- and enantioselectivity of proline-catalyzed aldolizations.

HO N O OH

H

H

R

OH

O

OH

R

<1% (<1%) 6% (0.1%)

Experiment (Density functional theory calculation) Scheme 4.30

OH R

R

NO O H H

O

(S)-Proline DMSO

O

R

1.5 (<1%) 3.5% (0.8%)

4.3 Asymmetric Aminocatalysis of the Aldol Reaction

OH O CHO H

H

H Proline (cat.) DMSO, rt, 12h

O

H H

H

84%,

OH OMe

H H

OH

153

OMe

152

151 OH OH H HO

H

OH

H

known H

OH

H

HO H

O

O Brassinolide (156)

H HO

H 154

Scheme 4.31

Formal total synthesis of brasinolide featuring a proline-catalyzed aldol reaction.

droxyacetone to give diol 153 in good yield and diastereoselectivity. This material was converted to derivative 154, a known precursor of brassinolide (Scheme 4.31). Other catalysts besides proline have also been investigated. A particularly large amount of data has been collected for the aldol reaction of acetone with p-nitrobenzaldehyde (Scheme 4.32). Simple primary a-amino acids and acyclic N-methylated a-amino acids are not catalytically active under standard reaction conditions. Of the simple cyclic amino acids studied, azetidine, pyrrolidine, and piperidine 2-carboxylate, proline is clearly the best catalyst. a-, a 0 -, and, in particular, N-methylation reduce the eﬃciency and whereas substitution of the 3- and 4-positions are tolerated without dramatic eﬀects. Proline amide is essentially catalytically inactive under the standard reaction conditions (DMSO, room temperature, 2 h) but after three days, the aldol could be isolated in good yields, albeit with very low enantioselectivity. Clearly, the carboxylic acid plays an important role in the catalysis and in determining enantioselectivity. That the enantioselectivity can be improved was shown with penicillamine derivative 168, proline derived diamine salt 169, and amide 170 [129–132]. By screening a wide variety of diamines Yamamoto et al. showed that diamine salt 169 catalyzes the aldol reaction of several diﬀerent ketones with p-nitrobenzaldehyde to give aldols in good yields and selectivities (Scheme 4.33) [130, 131]. Comparing results from reaction of cyclopentanone and cyclohexanone with p-nitrobenzaldhyde catalyzed by proline, 168, or diamine, 169, illus-

185

186

4 Amine-catalyzed Aldol Reactions

Catalyst

14

133a

NO2 Conditions

Catalyst

134a

Conditions

i-Pr H2N

CO2H i-Pr

N H

CO2H CO2H CO2H

CO2H N H CO2H N CO2H N H CONH2 N H

NO2

Yield %

ee %

157

30 mol% Cat., 2h, rt DMSO/Acetone (4:1)

<10

n.d.

158

30 mol% Cat., 2h, rt DMSO/Acetone (4:1)

<10

n.d.

159

30 mol% Cat., 2h, rt DMSO/Acetone (4:1)

<10

n.d.

160

30 mol% Cat., 2h, rt DMSO/Acetone (4:1)

55

40

161

30 mol% Cat., 2h, rt DMSO/Acetone (4:1)

68

76

162

30 mol% Cat., 2h, rt DMSO/Acetone (4:1)

<10

n.d.

163

30 mol% Cat., 2h, rt DMSO/Acetone (4:1)

26

61

164

30 mol% Cat., 2h, rt DMSO/Acetone (4:1) Same conditions, 72h

<10

n.d.

80

20

NH N H

OH

O

OHC

O

HO CO2H

165

30 mol% Cat., 2h, rt DMSO/Acetone (4:1)

85

78

CO2H

166

30 mol% Cat., 2h, rt DMSO/Acetone (4:1)

67

73

CO2H

167

30 mol% Cat., 24h, rt DMSO/Acetone (4:1)

<10

n.d.

CO2H

168

30 mol% Cat., 24h, rt DMSO/Acetone (4:1)

81

85

OTf H

169

3 mol% Cat., 2h, Acetone, 30°C

60

88

170

20 mol% Cat., 24h, Acetone, -25°C

66

93

N H S N H S N H S N H

N H

N O Ph

N H

HN Ph HO

Scheme 4.32

Aminocatalysts studied for reaction of acetone with p-nitrobenzaldehyde.

4.3 Asymmetric Aminocatalysis of the Aldol Reaction

H N H

O

R

N 169 (10%)

NO2

R

O

dr Yield % (anti/syn)

OH

OH

NO2

171b 5% ee O

172a 96% ee OH

-

88

0.75

97

2.85

81

1.17

NO2

NO2

OH

NO2 O

NO2

60a

OH

O

OH

173a 84% ee

NO2

Products

NO2 O

R 171-173

134a 88% ee

171a 84% ee

R

133a

O

O

OH

O

OHC

Solvent

OTf

172b 61% ee OH

173b 16% ee

NO2

a) Only 3 mol% of the catalyst was used.

Scheme 4.33

Diamine salt 169 catalyzes asymmetric aldol reactions of various ketones.

trates that it can be advantageous to use catalysts other than proline for selected reactions (Scheme 4.34). The best enantioselectivity in aminocatalytic intermolecular acetone aldolizations has recently been achieved by Tang et al. [132]. Study of selected proline-derived amides led this group to identify hydroxy amide 170 as a superb catalyst for reactions of acetone (solvent) with a variety of aldehydes (Scheme 4.35). Although the yields are typically slightly better when proline is used as the catalyst, enantioselectivity is invariably higher when amide 170 is used. The mechanism of catalysis by amide 170 has not yet been elucidated. Density functional theory calculations support activation of the aldehyde

187

188

4 Amine-catalyzed Aldol Reactions

O

R

A. B, or C R

NO2

R 133a

dr Yield (anti/syn) %

Products OH

O

171 O

OH

O

OHC

R 171-172

NO2

ee % (anti)

ee % (syn)

Conditions

73 63 88

1.7 1.6 0.8

69 63 85

63 60 5

A B C

65 65 97

1.7 1.7 2.8

89 90 96

67 69 61

A B C

NO2

OH

172

NO2

B

A CO2H N H 161 20 mol%, rt Ketone/DMSO (1:4) Solvent

S

CO2H N 168 H 20 mol%, rt Ketone/DMSO (1:4) Solvent

C H N H

N

OTf

169 10 mol%, 30°C Ketone Solvent

Scheme 4.34

Comparing yields, enantioselectivity, and diastereoselectivity of diﬀerent amino catalysts in the aldolization of cyclic ketones with p-nitrobenzaldehyde.

with two simultaneous hydrogen bonds from the amide-NH and the alcohol-OH in the transition state (A, Scheme 4.36). An alternative explanation involves a protonated oxazoline which might be formed in situ via dehydration of the hydroxy amide [133]. The corresponding transition state (B) might then be very similar to that proposed for the proline- and diamine salt-catalyzed reactions. Very recently N-terminal prolyl peptides have been suggested as another attractive class of aldol catalyst. Tang et al. [132] mention in a footnote that the dipeptide ProaThraOMe catalyzes the reaction of acetone with pnitrobenzaldehyde to give the corresponding aldol in 69% ee. Reymond et al. have studied a peptide library and also found that several N-terminal prolyl peptides catalyze the same asymmetric aldolization [124]. Independently, Martin and List showed that N-terminal prolyl peptides catalyze the direct asymmetric aldol reaction of acetone with p-nitrobenzaldehyde and the asymmetric Michael reaction between acetone and b-nitrostyrene (Scheme 4.37) [135]. Mechanisms have so not yet been proposed for these reactions.

4.3 Asymmetric Aminocatalysis of the Aldol Reaction

O Ph N H O RCHO 133

14 Solvent Aldehyde

HN Ph R

(20 mol%) -25°C, 1-2 d

Yield %

OH

O

170 HO

134

Product O

ee%

OH

OHC NO2

133a OHC

93 (76)

66 (68) 134a O OH

83 (72)

51 (62) 133b

NO2

134b OH

O

OHC

133c OHC

Br

77 (74)

Cl

O

Br

OH

Cl 85 (69)

83 (94)

133d OHC

90 (65)

134c

134d OH

O

84 (77)

93 (54) 134e 133e

OHC

133f OHC

O

OH

43 (97)

98 (96)

134f O OH 98 (84)

77 (63) 134g

133g OHC

O

51 (81)

OH

>99 (>99)

134h 133h OHC 135a

12 (31)

O

OH 136a

Scheme 4.35

Hydroxyamide 170-catalyzed aldolizations of acetone with diﬀerent aldehydes. Numbers in parentheses refer to the corresponding proline-catalyzed reactions.

86 (67)

189

190

4 Amine-catalyzed Aldol Reactions

Ph Ph

Ph

Ph

O H N HO

O N

O H Ph

NH O

H Ph

N B

A Scheme 4.36

Mechanistic considerations of the hydroxyamide-catalyzed aldolization.

Peptide

O NO2 14

OH

O

OHC

NO2

134a

133a Catalyst

Yield %

ee %

Ref.

H-Pro-Thr-OMe

n.r.

69

132

H-Pro-Gly-OH

99

46

134

H-Pro-Glu-Leu-Phe-OH

96

66

134

H-Pro-Aib-Glu-Phe-OH

94

37

134

H-Pro-Asp-Leu-Phe-OH

95

50

134

H-Pro-Aib-Asp-Phe-OH

97

12

134

H-Pro-Ala-OH

90

70

135

H-Pro-Trp-OH

77

65

135

H-Pro-Asp-OH

75

74

135

H-Pro-Glu-OH

72

68

135

H-Pro-Val-OH

89

70

135

H-Pro-Arg-OH

91

31

135

H-Pro-Ser-OH

87

77

135

H-Pro-LysHCl-OH

62

66

135

H-Pro-Gly-Gly-OH

68

53

135

H-Pro-His-Ala-OH

85

56

135

Scheme 4.37

N-terminal prolyl-peptides catalyze enantioselective aldolizations.

4.3 Asymmetric Aminocatalysis of the Aldol Reaction

(S)-Proline (30 mol%)

OHC

O

14

133b

O

176

(1) 134b 58%, 71% ee (62%, 72% ee in DMSO) O

178

176

[emim][OTf] 48h, rt

177

N

N

CF3

59%, 76% ee (S)-Proline (50 mol%)

CF3

OH (2)

[bmim][PF6] 24h, rt

177

OHC

OH

[bmim][PF6] 25h, rt

(S)-Proline (30 mol%)

OHC CF3

O

O

191

PF6

[bmim][PF6] (174)

O

OH

178

(3)

29%, > 98% ee

179

CF3

68%, 88% ee, dr = 99:1

N

N

CF3SO3

[emim][OTf] (175)

Scheme 4.38

Proline-catalyzed aldolizations in ionic liquids.

In addition to studying alternative catalysts, researchers have also investigated new reaction conditions for proline-catalyzed aldolizations. It has, for example, been found that the commonly used DMSO could be replaced with the room temperature ionic liquids [bmim][PF6 ] (174) or [emim][OTf ] (175) with comparable yields and selectivity (Scheme 4.38) [136–138]. The eﬀect of solvent structure on the regioselectivity can, apparently, be quite strong as is illustrated by the reaction of butanone with p-triﬂuorobenzaldehyde (Scheme 4.38, Eqs. (2) and (3)). Proline-catalyzed aldolizations can also be conducted under aqueous conditions, although rates and enantioselectivity generally decrease with increasing water concentration [48, 139]. An interesting observation has been made by Peng et al., who showed that a combination of proline with sodium dodecyl sulfonate can be used as a catalyst system, indicative of catalysis in aqueous micelles (Scheme 4.39) [140]. The enantioselectivity of the reaction studied has not been reported. Proline has also been attached to a polymer support. Takemoto et al. described the ﬁrst example of this approach [141]. Hydroxyproline was linked

192

4 Amine-catalyzed Aldol Reactions

O

OHC

OH

H2O

NO2 14

O

Catalyst

133a

134a

Catalyst

Conditions

Proline (40 mol%)

5d, rt

15

SDS (20%

5d, rt

12

Proline (40 mol%) + SDS (20%)

1d, rt

87

NO2

Yield %

SO3Na SDS =Sodium dodecyl sulfonate Scheme 4.39

Proline-catalyzed aldolizations in inverse micelles.

NaH

1.

HO

Cl

CO2Et N BOC

O CO2H

2. H+ 3. OH-

180

N H 181

O

O

O 181 34%

O 104

O

OH 105 18% ee

Scheme 4.40

The ﬁrst polymer-supported proline-catalyst.

to a polystyrene resin and the resulting material (181) was shown to catalyze the Hajos–Parrish–Eder–Sauer–Wiechert reaction, albeit in low yield and with low enantioselectivity (Scheme 4.40). Proline has also recently been attached to a poly(ethylene glycol) support. The resulting catalyst (183) eﬀectively catalyzes intermolecular aldolizations, a Robinson annexation, and Mannich reactions. It can be used under

4.3 Asymmetric Aminocatalysis of the Aldol Reaction

O OH

MeO-(CH2CH2O)n-CH2CH2 -O n = 110

O

182 O

O

MeO-(CH2CH2O)n-CH2CH2 -O 183

14

133f

N H

O

183

OHC

O

CO2H

O

DMSO 130 h

OH

134f 81%, >98% ee O

O

O

183 DMSO 130 h

O

O 109 55%, 75% ee

Scheme 4.41

Poly(ethylene glycol)-supported proline-catalysis.

homogenous conditions, then precipitated by solvent change and reused (Scheme 4.41) [142, 143]. Aldehyde Donors In addition to ketones, aldehydes can also be used as aldol donors in proline-catalyzed reactions [144]. Barbas et al. found that treating acetaldehyde solutions with proline provided aldehyde 185, an aldol trimer of acetaldehyde, in 84% ee and 4% yield (Scheme 4.42, Eq. (1)) [145, 146]. As shown by Jørgensen et al., other simple a-unbranched aldehydes can also be used as donors in proline-catalyzed cross aldolization with activated non-enolizable ketone acceptors to give aldols 188 in high enantioselectivity and yield (Scheme 4.42, Eq. (2)) [147]. Finally, Northrup and MacMillan found that proline to also catalyzed cross aldolizations of two diﬀerent aldehydes under carefully developed conditions using syringe pump techniques [148]. These reactions furnish anti aldols 191 in excellent enantioselectivity and good yield and diastereoselectivity (Scheme 4.43). 4.3.2.2

Proline-catalyzed Aldolization of Propionaldehyde with Isobutyraldehyde to Give Aldol 191e [148]. A solution of freshly distilled propionaldehyde (1.81 mL, 25 mmol) in DMF (12.5 mL) pre-cooled to 4 C was added slowly over

193

194

4 Amine-catalyzed Aldol Reactions

(S)-Proline (cat.)

O H

185 4% yield, ee = 84%

184

(S)-Proline (50 mol%.)

O

H

E

186 R 1 eq

(1)

H

THF, 4°C

O

E

O

OH

H

CH2Cl2

R

187 1 eq

R

OH

O

3

CO2Et CO2Et

(2)

188 Yield %

ee %

Me

(188a)

90

90

Et

(188b)

91

85

i-Pr

(188c)

88

85

CH2CH=CH2

(188d)

94

88

n-C6H13

(188e)

91

84

Ph

(188f)

97

0

Scheme 4.42

Proline catalyzed aldolreactions with aldehyde donors.

the course of 20 h to a stirred suspension of isobutyraldehyde (4.54 mL, 50 mmol), (S)-proline (288 mg, 2.5 mmol, 10 mol%), and DMF (12.5 mL) at 4 C. After 30 h the resulting solution was diluted with diethyl ether and washed successively with water and brine. The combined aqueous layers were back-extracted with three portions of dichloromethane. The organic layers were combined, dried over anhydrous MgSO4 , and concentrated in vacuo. Flash chromatography (20:7 pentane–diethyl ether) aﬀorded aldol 191e as a clear colorless oil (2.65 g, 20.6 mmol, 82%) in >99% ee and 96:4 anti:syn. ½aD ¼ 17:9 (c ¼ 1, CHCl3 ). The product ratios were determined by GLC analysis of the acetal derived from 2,2-dimethylpropane-1,3-diol on a Bodman Chiraldex b-DM column (110 C isothermal, 23 psig); (2S,3S) anti isomer t r ¼ 31:8 min, (2R,3R) anti isomer t r ¼ 33:9 min, (2R,3S) and (2S,3R) syn isomers t r ¼ 29:4, 29.8 min. The discovery of the direct enantioselective cross-aldolization of aldehydes, which not too long ago would have probably been regarded as impossible by many, can be viewed as a major breakthrough in the aldol ﬁeld.

4.3 Asymmetric Aminocatalysis of the Aldol Reaction

O

(S)-Proline (10 mol%)

O

H R

H

1

R2

4°C, DMF

OH

O

R2

H R

1

191 Product Me

H

O

%Yield

dr

%ee

80

4:1

99

88

3:1

97

87

14:1

99

81

3:1

99

82

24:1

>99

80

24:1

98

75

19:1

91

OH

O

Me 191a OH Me

H

Me Me 191b OH O H Me 191c OH O H Me 191d OH O H Me 191e OH O H Bu 191f OH O H Bn 191g

Scheme 4.43

Direct asymmetric cross-aldolization of aldehydes.

Finally, aldols are made in the most atom economical and practical way without use of stoichiometric bases, auxiliaries, or other reagents. Clearly, aminocatalytic aldolizations have great potential in academic and industrial synthesis and nothing other than frequent use and further development of these methods can be expected.

195

196

4 Amine-catalyzed Aldol Reactions

References 1 C. H. Heathcock, In Comprehensive Organic Synthesis; B.

2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17

18 19 20 21 22 23 24 25 26 27 28 29 30

M. Trost, I. Fleming, C. H. Heathcock, Eds.; Pergamon: Oxford, 1991; Vol. 2, pp 133. J. S. Johnson, D. A. Evans, Acc. Chem. Res. 2000, 33, 325. T. D. Machajewski, C.-H. Wong, Angew. Chem., Int. Ed. 2000, 39, 1352. B. Alcaide, P. Almendros, Eur. J. Org. Chem. 2002, 1595. S. E. Denmark, R. A. Stavenger, Acc. Chem. Res. 2000, 33, 432. S. G. Nelson, Tetrahedron: Asymmetry 1998, 9, 357. M. Shibasaki, H. Sasai, T. Arai, T. Iida, Pure Appl. Chem. 1998, 70, 1027. E. M. Carreira, R. A. Singer, Drug Discovery Today 1996, 1, 145. I. Peterson, Pure Appl. Chem. 1992, 64, 1821. S. Masamune, W. Choy, J. S. Peterson, L. R. Sita, Angew. Chem., Int. Ed. Engl. 1985, 24, 1. Y. M. A. Yamada, N. Yoshikawa, H. Sasai, M. Shibasaki, Angew. Chem., Int. Ed. Engl. 1997, 36, 1871. N. Yoshikawa, Y. M. A. Yamada, J. Das, H. Sasai, M. Shibasaki, J. Am. Chem. Soc. 1999, 121, 4168. N. Kumagai, S. Matsunaga, N. Yoshikawa, T. Ohshima, M. Shibasaki, Org. Lett. 2001, 3, 1539. B. M. Trost, H. Ito, J. Am. Chem. Soc. 2000, 122, 12003. B. M. Trost, E. R. Silcoff, H. Ito, Org. Lett. 2001, 3, 2497. M. Nakagawa, H. Nakao, K.-I. Watanabe, Chem. Lett. 1985, 391. Y. Yamada, K. I. Watanabe, H. Yasuda, Utsonomiya Daigaku Kyoikugakubu Kiyo, Dai-2-bu 1989, 39, 25. Some of these results could not be reproduced, see: M. A. Calter, R. K. Orr, Tetrahedron Lett. 2003, 44, 5699. Z. G. Hajos, D. R. Parrish, German Patent DE 2102623, 1971. U. Eder, R. Wiechert, G. Sauer, German Patent DE 2014757, 1971. U. Eder, G. Sauer, R. Wiechert, Angew. Chem., Int. Ed. Engl. 1971, 10, 496. Z. G. Hajos, D. R. Parrish, J. Org. Chem. 1974, 39, 1615. N. Cohen, Acc. Chem. Res. 1976, 9, 512–41. K. Drauz, A. Kleemann, J. Martens, Angew. Chem., Int. Ed. Engl. 1982, 21, 584–608. J. Martens, Chem. Ztg 1986, 110, 169. M. E. Jung, Tetrahedron 1976, 32, 3. E. R. Jarvo, S. J. Miller, Tetrahedron 2002, 58, 2481–2495. N. Harada, T. Sugioka, H. Uda, T. Kuriki, Synthesis 1990, 53. ¨ rst, Synthesis 1977, J. Gutzwiller, P. Buchschacher, A. Fu 167. B. List, R. A. Lerner, C. F. Barbas, III, J. Am. Chem. Soc. 2000, 122, 2395. W. Notz, B. List, J. Am. Chem. Soc. 2000, 122, 7386.

References 31 32 33 34 35 36 37 38 39 40

B. List, P. Pojarliev, C. Castello, Org. Lett. 2001, 3, 573. B. List, Tetrahedron 2002, 58, 5572. B. List, Synlett 2001, 1675. S. Hosokawa, K. Sekiguchi, M. Enemoto, S. Kobayashi, Tetrahedron Lett. 2000, 41, 6429. V. Pistara, P. L. Barili, G. Catelani, A. Corsaro, F. D’Andrea, S. Fisichella, Tetrahedron Lett. 2000, 41, 3253. T. Ooi, M. Taniguchi, M. Kameda, K. Maruoka, Angew. Chem. Int. Ed. 2002, 41, 4542. G. Storck, A. Brizzolara, H. Landesman, J. Szmuszkovicz, R. Terrell, J. Am. Chem. Soc. 1963, 85, 207. T. Mukaiyama, K. Banno, K. Narasaka, J. Am. Chem. Soc. 1974, 96, 7503. O. Takazawa, K. Kogami, K. Hayashi, Bull. Chem. Soc. Jpn. 1985, 58, 2427–2428. See for example: L. Birkhofer, S. M. Kim, H. D. Engels, Chem.

Ber. 1962, 95, 1495. 41 F.G. Fischer, A. Marschall, Ber. 1931, 64, 2825. 42 W. Langenbeck, G. Borth, Ber. 1942, 75, 951. 43 F. H. Westheimer, H. Cohen, J. Am. Chem. Soc. 1938, 60,

90. 44 R. P. Koob, J. G. Miller, A. R. Day, J. Am. Chem. Soc. 1951,

73, 5775. 45 J. C. J. Speck, A. A. Forist, J. Am. Chem. Soc. 1957, 79, 4659. 46 J. Hine, W. H. Sachs, J. Org. Chem. 1974, 39, 1937. 47 R. M. Pollack, S. Ritterstein, J. Am. Chem. Soc. 1972, 94,

5064. 48 J.-L. Reymond, Y. Chen, J. Org. Chem. 1995, 60, 6970. 49 S. Bahmanyar, K. N. Houk, J. Am. Chem. Soc. 2001, 123,

11273. 50 Also see: A. Sevin, J. Maddaluno, C. Agami, J. Org. Chem.

1987, 52, 5611. 51 W. Treibs, K. Krumbholz, Chem. Ber. 1952, 85, 1116. 52 J. Schreiber, C.-G. Wermuth, Bull. Soc. Chim. Fr. 1965, 8,

2242. 53 F. Merger, J. Frank, Eur. Pat. Appl. 0 421 271 A2, 1991. 54 A. Kuiterman, H. Dielemanns, R. Green, A. M. C. F.

Castelijns, US patent 5,728,892, 1998. 55 T. Ishikwa, E. Uedo, S. Okada, S. Saito, Synlett 1999, 4, 450. 56 For an excellent review, see: Tietze, L. F., In: Comprehensive

57 58 59 60 61 62 63

Organic Synthesis, Vol. 2; Trost, B. M., Ed.; Pergamon Press: New York, 1991, 341–394. H. G. Lindwall, J. S. Maclennan, J. Am. Chem. Soc. 1932, 54, 4739. G. Kobayashi, S. Furukawa, Chem. Pharm. Bull. 1964, 12, 1129. R. B. Van Order, H. G. Lindwall, J. Org. Chem. 1945, 10, 128. R. C. Blume, H. G. Lindwall, J. Org. Chem. 1946, 22, 185. A. L. Wilds, C. Djerassi, J. Am. Chem. Soc. 1946, 68, 1715. H. Schlenk, Chem. Ber. 1948, 81, 175. E. C. Constable, P. Harverson, D. R. Smith, L. A. Whall, Tetrahedron 1994, 26, 7799.

197

198

4 Amine-catalyzed Aldol Reactions 64 E. Buchta, G. Satzinger, Chem. Ber. 1959, 92, 449. 65 D. Mead, R. Loh, A. E. Asato, R. S. H. Liu, Tetrahedron Lett.

1985, 26, 2873. 66 A. H. Lewin, M. R. Nilsson, J. P. Burgess, F. I. Carroll,

Org. Prep. Proced. Int. 1995, 27, 621. ¨ . Sezer, O. Anac, Rev. Roum. Chim. 1997, 42, 67 A. Daut, O 1045. 68 H. Bayer, C. Batzl, R. W. Hartmann, A. Mannschreck, J.

Med. Chem. 1991, 34, 2685. 69 Woodward, R. B., Sondheimer, F., Taub, D., Heusler, K.,

MacLamore, W. M. J. Am. Chem. Soc. 1952, 74, 4223. 70 E. J. Corey, R. L. Danheiser, J. Am. Chem. Soc. 1978, 100,

8031. 71 B. B. Snider, K. Yang, J. Org. Chem. 1990, 55, 4392. 72 X.-X. Shi, Q.-Q. Wu, X. Lu, Tetrahedron: Asymmetry 2002, 13, 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

461. A. Srikrishna, N. C. Babu, Tetrahedron Lett. 2001, 42, 4913. B. Traber, H. Pfander, Tetrahedron Lett. 2000, 41, 7197. W. R. Roush, D. A. Barda, Tetrahedron Lett. 1997, 38, 8785. R. Lalande, J. Moulines, J. Duboudin, C. R. Bull. Soc. Chim. Fr. 1962, 1087. C.-Q. Wei, X.-R. Jiang, Y. Ding, Tetrahedron 1998, 54, 12623. H. Hagiwara, H. Ono, N. Komatsubara, T. Hoshi, T. Suzuki, M. Ando, Tetrahedron Lett. 1999, 40, 6627. M. Allard, J. Levisalles, H. Rudler, Bull. Soc. Chim. Fr. 1968, 303. Y. Ding, X.-R. Jiang, J. Chem. Soc., Chem. Commun. 1995, 16, 1693. A. Takahashi, M. Shibasaki, Tetrahedron Lett. 1987, 28, 1893. J. Wright, G. J. Drtina, R. A. Roberts, L. A. Paquette, J. Am. Chem. Soc. 1988, 110, 5806. P. Wieland, K. Miescher, Helv. Chim. Acta 1950, 33, 2215. N. P. Buu-Hoi, N. Hoan, D. Lavit, J. Chem. Soc. 1950, 230. A. Wettstein, K. Heusler, H. Ueberwasser, P. Wieland, Helv. Chim. Acta 1957, 40, 323. E. Plieninger, G. Ege, R. Fischer, W. Hoffmann, Chem. Ber. 1961, 94, 2106. A. L. Begbie, B. T. Golding, J. Chem. Soc. Perkin I 1972, 602. W. Kreiser, P. Below, Tetrahedron Lett. 1981, 22, 429. E. J. Corey, M. A. Tius, J. Das, J. Am. Chem. Soc. 1980, 102, 7612. B. Roy, H. Watanabe, T. Kitahara, Heterocycles 2001, 55, 861. S. Swaminathan, M. S. Newman, Tetrahedron 1958, 2, 88. S. Ramachandran, M. S. Newman, Org. Synth. 1961, 41, 38. R. Selvarajan, J. P. John, K. V. Narayanan, S. Swaminathan, Tetrahedron 1966, 22, 949. T. A. Spencer, H. S. Neel, T. W. Flechtner, R. A. Zayle, Tetrahedron Lett. 1965, 43, 3889. T. A. Spencer, K. K. Schmiegel, Chem. Ind. 1963, 1765. R. D. Roberts, H. E. Ferran, Jr., M. J. Gula, T. A. Spencer, J. Am. Chem. Soc. 1980, 102, 7054. T. A. Spencer, Bioorganic Chemistry, Vol. 1, Academic Press, New York San Francisco London, 1977.

References 98 S. Takano, C. Kasahara, K. Ogasawara, J. Chem. Soc. Chem.

Comm. 1981, 635. 99 J. C. Blazejewski, J. Fluorine Chemistry 1990, 46, 515. 100 S. Terashima, S. Sato, K. Koga, Tetrahedron Lett. 1979, 36,

3469. 101 A. Prze´zdziecka, W. Stepanenko, J. Wicha, Tetrahedron:

Asymmetry 1999, 10, 1589. 102 S. Kwiatkowski, A. Syed, C. P. Brock, D. S. Watt, Synthesis

1989, 818. 103 D. Rajagopal, R. Narayanan, S. Swaminathan, Tetrahedron

Lett. 2001, 42, 4887. 104 N. Ramamurthi, S. Swaminathan, Indian J. Chem. Sect. B

1990, 29, 401. 105 S. Danishefsky, P. Cain, J. Am. Chem. Soc. 1976, 98, 4975. 106 C. Agami, F. Meynier, C. Puchot, Tetrahedron 1984, 40,

1031. 107 C. Agami, N. Platzer, H. Sevestre, Bull. Soc. Chim. Fr. 1987,

2, 358. 108 C. Agami, H. Sevestre, J. Chem. Soc., Chem. Commun. 1984,

21, 1385. 109 C. Agami, J. Levisalles, H. Sevestre, J. Chem. Soc., Chem.

Commun. 1984, 418. 110 C. Agami, C. Puchot, H. Sevestre, Tetrahedron Lett. 1986, 27,

1501. 111 C. Agami, C. Puchot, J. Mol. Cat. 1986, 38, 341. 112 C. Agami, C. Puchot, Tetrahedron 1986, 42, 2037. 113 C. Agami, J. Levisalles, C. Puchot, J. Chem. Soc., Chem.

Commun. 1985, 441. 114 C. Agami, Bull. Soc. Chim. FR. 1988, 3, 499. 115 D. Rajagopal, M. S. Moni, S. Subramanian, S.

Swaminathan, Tetrahedron: Asymmetry 1999, 10, 1631. 116 S. Bahmanyar, K. N. Houk, J. Am. Chem. Soc. 2001, 123,

12911. 117 K. L. Brown, L. Damm, J. D. Dunitz, A. Eschenmoser, R.

Hobi, C. Kratky, Helv. Chim. Acta 1978, 61, 3108. 118 L. Hoang, S. Bahmanyar, K. N. Houk, B. List, J. Am. Chem.

Soc. 2003, 125, 16. 119 C. Puchot, O. Samuel, E. Dunach, S. Zhao, C. Agami,

H. B. Kagan, J. Am. Chem. Soc. 1980, 108, 2353. 120 C. Pidathala, L. Hoang, N. Vignola, B. List, Angew. Chem.

Int. Ed. Engl. 2003, 115, 2891. 121 Also see: R. B. Woodward, E. Logusch, K. P. Nambiar, K.

Sakan, D. E. Ward, et al., J. Am. Chem. Soc. 1981, 103, 3210. 122 C. Agami, N. Platzer, C. Puchot, H. Sevestre, Tetrahedron

1987, 43, 1091. 123 B. List, P. Pojarliev, C. Castello, Org. Lett. 2001, 3, 573. 124 I. Izquierdo, M. T. Plaza, R. Robles, A. J. Mota, F. Franco,

Tetrahedron: Asymmetry 2001, 12, 2749. 125 S. Bahmanyar, K. N. Houk, H. J. Martin, B. List, J. Am.

Chem. Soc. 2003, 125, 2475. 126 K. N. Rankin, J. W. Gauld, R. J. Boyd, J. Phys. Chem. A 2002. ´ , L. R. Domingo, Theor. Chem. Acc. 2002, 381. 127 M. Arno

199

200

4 Amine-catalyzed Aldol Reactions 128 L. Peng, H. Liu, T. Zhang, F. Zhang, T. Mei, Y. Li,

Tetrahedron Lett. 2003, 44, 5107. 129 K. Sakthivel, W. Notz, T. Bui, C. F. Barbas, J. Am. Chem.

Soc. 2001, 123, 5260. 130 S. Saito, M. Nakadai, H. Yamamoto, Synlett 2001, 8, 1245. 131 M. Nakadai, S. Saito, H. Yamamoto, Tetrahedron 2002, 58,

8167. 132 Z. Tang, F. Jiang, L.-T. Yu, Y. Cui, L.-Z. Gong, A.-Q. Mi,

Y.-Z. Jiang, Y.-D. Wu, J. Am. Chem. Soc. 2003, 125, 5262. 133 Suggested by Professor K. N. Houk, UCLA. 134 J. Kofoed, J. Nielsen, J.-L. Reymond, Biorg. Med. Chem. Lett.

2003, on-line article. 135 H. J. Martin, B. List, Synlett, 2003, 1901. 136 T. P. Loh, L.-C. Feng, H. Y. Yang, J.-Y. Yang, Tetrahedron

Lett. 2002, 43, 8741. 137 P. Kotrusz, I. Kmentova´, B. Gotov, S. Toma, E. Solca´niova´,

Chem. Comm. 2002, 2510. 138 T. Kitazume, Z. Jiang, K. Kasai, Y. Mihara, M. Suzuki, J.

Fluorine Chem. 2003, 121, 205. ´ rdova, W. Notz, C. F. Barbas, Chem. Commun. 2002, 139 A. Co 3024. 140 Y.-Y. Peng, Q.-P. Ding, Z. Li, P. G. Wang, J. P. Cheng,

Tetrahedron Lett. 2003, 44, 3871. 141 K. Kondo, T. Yamano, K. Takemoto, Macromol. Chem. 1985,

186, 1781. 142 M. Benaglia, G. Celentano, F. Cozzi, Adv. Synth. Catal.

2001, 343, 171. 143 M. Benaglia, M. Cinquini, F. Cozzi, A. Puglisi, Adv. Synth.

Catal. 2002, 344, 533. 144 Proline-mediated aldehyde self-aldolizations have been

145 146 147 148

mentioned by: F. Orsini, F. Pelizzoni, M. Forte, J. Heterocycl. Chem. 1989, 26, 837. A. Co´rdova, W. Notz, C. F. Barbas, J. Org. Chem. 2002, 67, 301. Also see: N. S. Chowdari, D. B. Ramachary, A. Co´rdova, C. F. Barbas, Tetrahedron Lett. 2002, 43, 9591. A. Bogevig, N. Kumaragurubaran, K. A. Jorgensen, Chem. Commun. 2002, 620. A. B. Northrup, D. W. C. MacMillan, J. Am. Chem. Soc. 2002, 124, 6798.

201

5

Enzyme-catalyzed Aldol Additions Wolf-Dieter Fessner 5.1

Introduction

As a supplement to classical chemical methodology, enzymes are ﬁnding increasing acceptance as chiral catalysts for the in-vitro synthesis of asymmetric compounds because they have been optimized by evolution for high selectivity and catalytic eﬃciency [1–11]. In particular, the high stereospeciﬁcity of aldolases in CaC bond-forming reactions gives them substantial utility as synthetic biocatalysts, making them an environmentally benign alternative to chiral transition metal catalysis of the asymmetric aldol reaction [12]. In contrast with most classical chemical operations, biocatalytic conversions can usually be performed under mild reaction conditions that are compatible with underivatized substrates, thus also obviating tedious and costly protecting group manipulation [13]. Several dozen aldolases have been identiﬁed in Nature [14, 15] and many of these are commercially available on a scale suﬃcient for preparative applications. Recent advances in molecular and structural biology have improved access to virtually any biocatalyst in large quantity and to its detailed functional topology which increasingly enables exploitation of its synthetic utility on rational grounds. Evidently, enzyme catalysis is thus most attractive for the synthesis and modiﬁcation of biologically relevant classes of organic compounds that are typically complex, multifunctional, and water soluble. Typical examples are those structurally related to amino acids [16, 17] or carbohydrates [18–24], which are diﬃcult to prepare and handle by conventional methods of chemical synthesis. Because of the multitude of factors that might be critical to the success of an enzymatic conversion, and because of the empirical nature of their development, it is mandatory in the design of new biocatalytic processes to become familiar with the scope and limitations of synthetically useful enzymes, both as a source of inspiration and for reference. Thus, this overview attempts to outline the current status of development for the most important aldolase biocatalysts and their preparative potential for asymmetModern Aldol Reactions. Vol. 1: Enolates, Organocatalysis, Biocatalysis and Natural Product Synthesis. Edited by Rainer Mahrwald Copyright 8 2004 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 3-527-30714-1

5 Enzyme-catalyzed Aldol Additions

202

O O

H

R

-

R'

X

OH

O

R

OH R'

R

X

OH

O R'

R

O

OH R'

R

X

X

O R'

X

Fig. 5.1

Generation of stereo-diversity by aldol addition.

OPO3=

O H

CO2H HH

H

CO2H H

pyruvate

phosphoenolpyruvate

O

O OPO3=

HO HH

dihydroxyacetone phosphate

H

O H

HH acetaldehyde

H2N

OH HH glycine

Fig. 5.2

Nucleophiles for preparatively useful aldolases.

ric synthesis. Catalytic CaC coupling is among the most useful synthetic methods in asymmetric synthesis because of its potential for stereoisomer generation by a convergent, ‘‘combinatorial’’ strategy [25]. Thus, attention is also paid to the feasibility of directed stereodivergent approaches by which multiple, diastereomeric products can be derived from common synthetic building blocks (Figure 5.1) [25, 26]. Obviously, such a synthetic strategy requires the prevalence of related, stereocomplementary enzymes which must have a similarly broad substrate tolerance.

5.2

General Aspects 5.2.1

Classiﬁcation of Lyases

Most enzymes used by Nature for carbon–carbon bond formation and cleavage (‘‘lyases’’) catalyze a crossed aldol reaction in the form of a reversible, stereocontrolled addition of a nucleophilic ketone donor to an electrophilic aldehyde acceptor. Synthetically the most useful and most extensively studied enzymes use aldol donors comprising 2-carbon or 3carbon fragments and can be grouped into four categories depending on the structure of their nucleophilic component (Figure 5.2): (i) pyruvate-

5.2 General Aspects

O B:

O X

R H

B—H

X

R

H

H R'–CHO

O

O B

H

O

R

X H

H R

B—H

R

O

X H

H

R

Fig. 5.3

Retention mechanism for deprotonation and CaC bond formation in enzymatic aldolization.

(and phosphoenolpyruvate-) dependent aldolases, (ii) dihydroxyacetone phosphate-dependent aldolases, (iii) an acetaldehyde-dependent aldolase, and (iv) glycine-dependent enzymes. Apart from possible mechanistic diﬀerences, members of the ﬁrst and third types generate a-methylene carbonyl compounds and thereby generate a single stereocenter, whereas members of the other types form a-substituted carbonyl derivatives that contain two new vicinal chiral centers at the new CaC bond, which makes them particularly appealing for asymmetric synthesis. Typically, lyases require highly speciﬁc nucleophilic donor components, because of mechanistic requirements. This includes the need for reasonably high substrate aﬃnity and the general diﬃculty of binding and anchoring a rather small molecule in a fashion that restricts solvent access to the carbanionic site after deprotonation and shields one enantiotopical face of the nucleophile to secure correct diastereofacial discrimination (Figure 5.3) [27]. Usually, approach of the aldol acceptor to the enzyme-bound nucleophile occurs stereospeciﬁcally following an overall retention mechanism, whereas facial diﬀerentiation of the aldehyde carbonyl is responsible for the relative stereoselectivity. In this manner the stereochemistry of the CaC bond formation is completely controlled by the enzymes, in general irrespective of the constitution or conﬁguration of the substrate, which renders the enzymes highly predictable. Few speciﬁc exceptions to this rule are currently known (vide infra). Most lyases tolerate reasonably broad variation of the electrophilic acceptor component, usually an aldehyde, on the other hand. This feature, which nicely complements the emerging needs of combinatorial synthesis,

203

204

5 Enzyme-catalyzed Aldol Additions 3-O

CHO HO

6 P 2O

NH2

OPO3=

N S

N pyridoxal phosphate

N N

thiamine diphosphate

N

H N

HO

N H

H2N N

tetrahydrofolate

HN R

Fig. 5.4

Types of cofactors for aldol and related reactions.

makes possible a stereodivergent strategy for synthesis of arrays of stereoisomeric compounds by employing stereo-complementary enzymatic catalysts to selectively produce individual diastereomers by design from a given starting material. Biochemical nomenclature of enzymes [28] usually follows historical classiﬁcations that are diﬀerent from the needs of synthetic organic chemists today. Whereas the common EC numbers are valuable for reference, in this overview an enzyme designation will be used that utilizes a three-letter code as reference to the natural substrate (from which usually stereochemistry is immediately obvious), followed by a single capital letter that indicates the conversion type (e.g., aldolases (A), synthases (S), isomerases (I), dehydrogenases (D), oxidases (O), kinases (K), (glycosyl)transferases (T), epimerases (E) etc.) [25]. In addition to the preparatively useful aldolases, several mechanistically distinct enzymes can be employed for synthesis of product structures identical with those accessible from aldolase catalysis. Such alternative enzymes (e.g. transketolase), which are actually categorized as transferases but also catalyze aldol-related additions with the aid of cofactors (Figure 5.4) such as pyridoxal 5-phosphate (PLP), thiamine pyrophosphate (TPP), or tetrahydrofolate (THF), are emerging as useful catalysts in organic synthesis. Because these operations often extend and/or complement the synthetic strategies open to aldolases, a selection of such enzymes and examples of their synthetic utility are included also in this overview. 5.2.2

Enzyme Structure and Mechanism

Mechanistically, enzymatic activation of the aldol donor substrates is achieved by stereospeciﬁc deprotonation along two diﬀerent pathways (Figures 5.5 and 5.6) [29]. Class I aldolases bind their substrates covalently via imine-enamine formation with an active site lysine residue to initiate bond cleavage or formation (Figure 5.5); this can be demonstrated by reductive interception of the intermediate with borohydride which causes irreversible inactivation of the enzyme as a result of alkylation of the amine [30, 31].

5.2 General Aspects

205

enzyme NH2

enzyme H

O

+

H 2O

NH

X

R

X

R H+

enzyme H+

X

R

A

H 2O

NH

H+

B

R'CHO

R'CHO

enzyme NH2 O X

R R'

enzyme

enzyme

OH

NH

H+

H 2O

X

R H+

R'

H 2O

OH

H+

C

Schematic mechanism for class I aldolases.

OPO3=

CO2 H

OPO3=

enzyme CO2H

O Zn His HO His His

O HO

D OPO3=

enzyme R

O CO2H

O HO H

X

R

Fig. 5.5

enzyme

NH

H+

Zn His His His

enzyme—O Fig. 5.6

Schematic mechanism for DHAP-dependent class II aldolases.

Zn His His His

R'

O

206

5 Enzyme-catalyzed Aldol Additions

In contrast, class II aldolases utilize transition metal ions as a Lewis acid cofactor, which facilitates deprotonation by bidentate coordination of the donor to give the enediolate nucleophile (Figure 5.6) [32]. This eﬀect is usually achieved by means of a tightly bound Zn 2þ ion but a few other divalent cations can act instead. Evidently, aldolases of the latter class can be eﬀectively inactivated by addition of strong complexing agents such as EDTA [29]. Mechanistic pathways for both classes of aldolases have been substantiated by several recent crystal structures of liganded enzymes that altogether provide a detailed insight into the catalytic cycles and the individual function of active site residues in the stereochemically determining events. For class I type enzymes, the (ba)8 -barrel structure of the class I fructose 1,6-bisphosphate aldolase (FruA, vide infra) from rabbit muscle was the ﬁrst to be uncovered by X-ray crystal-structure analysis [33]; this was followed by those from several other species [34–37]. A complex of the aldolase with non-covalently bound substrate DHAP (dihydroxyacetone phosphate) in the active site indicates a trajectory for the substrate traveling towards the nucleophilic Lys229 N e [38, 39]. There, the proximity of side-chains Lys146 and Glu187 is consistent with their participation as proton donors and acceptors in Schiﬀ base formation (A, B); this was further supported by sitedirected mutagenesis studies [40]. Complementary information was gained from structural investigations of the E. coli transaldolase (EC 2.2.1.2) at 1.87 A˚ resolution [41], including those on its covalent complex with dihydroxyacetone, which had been trapped at the active site Lys N e by Schiﬀ base reduction [42]. The three-dimensional structure of the neuraminic acid aldolase (NeuA, vide infra) from Escherichia coli has also been determined [43] after interception of the pyruvate Schiﬀ base at Lys165 N e by borohydride reduction [44]. Further insight came from high-resolution structures (up to 1.6 A˚) of the hemophilus inﬂuenzae enzyme in a complex with three substrate analogs [45]. Formation of a carbinolamine as the short-lived Schiﬀ base precursor (C) could be trapped by ﬂash freezing of 2-keto-3-deoxy-6-phosphogluconate aldolase in the presence of its natural donor substrate pyruvate [46]. Stereospeciﬁcity of the reaction seems to be ensured mostly by hydrophobic contact of the pyruvate methyl group in the active site. For class II aldolases, X-ray structures of the zinc-dependent fuculose 1-phosphate aldolase (FucA, vide infra) [47, 48] and the rhamnulose 1phosphate aldolase (RhuA, vide infra) [49] from E. coli have recently been solved and conﬁrm close similarity in their overall fold. Both enzymes are homotetramers in which subunits are arranged in C4 -symmetry. The active site is assembled in deep clefts at the interface between adjacent subunits and the catalytic zinc ion is tightly coordinated by three His residues. Phosphoglycolohydroxamate (PGH), a structural analog with the DHAP enediolate (D) that can be regarded as a mimic of an advanced intermediate or transition state, has been shown it to be a potent inhibitor of all currently accessible class II DHAP aldolases with K i in the nm range [32].

5.2 General Aspects

Its metal chelating binding mode has been determined by liganded structures of native protein and active-site mutants. In concert, these studies have enabled derivation of a conclusive blueprint for the catalytic cycle of metal-dependent aldolases that successfully rationalizes all key stereochemical issues [50]. Structural details are also available for the class II (ba)8 -barrel enzyme FruA from E. coli at excellent resolution (1.6 A˚) [51, 52]. The homodimeric protein requires movement of the divalent zinc cofactor from a buried position to the catalytically eﬀective surface position. Recent attempts to explore the origin of substrate discrimination of the structurally related E. coli aldolase with speciﬁcity for tagatose 1,6-bisphosphate by site-directed mutagenesis and structure determinations highlight the complexity of enzyme catalysis in this class of enzymes and the subtleties in substrate control [53–55]. 5.2.3

Practical Considerations

Although most of the aldolases attractive for synthetic applications arise from catabolic pathways in which they function in the degradative cleavage of metabolites, the reverse CaC bond-forming processes are often favored by thermodynamic relationships [56]. Because of the bimolecular nature of the reaction, the fraction of product at equilibrium might be increased in less favorable cases by working at higher substrate concentrations or by driving the reaction with a higher concentration of one of the reactants. Individual choice will certainly depend primarily on the cost of starting materials and ease of separation from the product when used in large excess; critical factors such as enzyme inhibition by the substrate(s) or product must also be considered. The latter factor will be more obvious if one recognizes that for most of the lyases both the donor and acceptor components contain strong electrophilic sites such as aldehyde or ketone carbonyl groups, and that many lyases, including class I aldolases, involve covalent binding of substrate and product at – but not necessarily restricted to – the active site. Most useful lyase families use substrates functionalized by anionically charged groups such as those present in pyruvate or dihydroxyacetone phosphate, which remain unaltered during catalysis. The charged group thereby introduced into products (phosphate, carboxylate) not only constitutes a handle for binding of the substrates by the enzymes but can also facilitate preparative isolation from an aqueous solution of the products and their puriﬁcation by salt precipitation or ion-exchange techniques. One problem arising as a result of the aﬃnity of the enzymes for anionic substrates is potential (competitive) inhibition by common buﬀer salts, e.g. inorganic phosphate. For routine practical application most aldolases are suﬃciently robust to enable their use in solution for an extended period of time, often several days. To enhance lifetime and to facilitate recovery of the biocatalysts after

207

208

5 Enzyme-catalyzed Aldol Additions

completion of the desired conversion, several options have been tested to immobilize the enzymes to or within insoluble matrices [57, 58], including cross-linking of enzyme crystals [59] or conﬁning them in membrane reactors [60–62]. To facilitate conversion of less polar substrates, reaction media containing up to 30% organic cosolvent and even highly concentrated waterin-oil emulsions have been tested successfully [63]. Applications of other current techniques are indicated with the individual enzymes, if available. Carbon–carbon bond construction is a pivotal process in asymmetric synthesis of complex molecular targets. Using enzyme catalysis, molecular complexity can be rapidly built up under mild conditions, without a need for protection of sensitive or reactive functional groups and with high chemical eﬃciency and usually uncompromised stereochemical ﬁdelity [12, 14]. Within a synthetic strategy for an intricate target structure, subsequent chemoselective or regioselective diﬀerentiation of functional groups might pose a considerable diﬃculty that can easily outbalance the economical nature of the biocatalytic transformation. This strongly suggests consideration of schemes suitable for introduction and handling of protective groups enzymatically [13]. For this purpose, a broad array of enzymes is available, particularly such with ester and amide-forming or cleaving behavior, and the feasibility of selective operation has been well investigated [1–11]. Such technology has not yet been adopted in routine applications at its full potential, but certainly would foster a more practical interface between biocatalytic and chemical synthesis.

5.3

Pyruvate Aldolases

In vivo, pyruvate-dependent lyases mostly serve a catabolic function in the degradation of sialic acids and KDO (2-keto-3-deoxy-manno-octosonate) and in that of 2-keto-3-deoxy aldonic acid intermediates from hexose or pentose catabolism. Because these freely reversible aldol additions often have less favorable equilibrium constants [29], synthetic reactions are usually driven by excess pyruvate to achieve a satisfactory conversion. A few related enzymes have been identiﬁed that use phosphoenolpyruvate in place of pyruvate which, by release of inorganic phosphate upon CaC bond formation, renders aldol additions essentially irreversible [14, 25]. Although attractive for synthetic applications, such enzymes have not yet been intensively studied for preparative applications [64]. 5.3.1

N-Acetylneuraminic Acid Aldolase

N-Acetylneuraminic acid aldolase (or sialic acid aldolase, NeuA; EC 4.1.3.3) catalyzes the reversible addition of pyruvate (2) to N-acetyl-d-mannosamine

5.3 Pyruvate Aldolases

HO HO

OH NHAc O

OH

OH

HO

+

O CO2H

S

CO2H

OH

1

OH

NeuA

O

209

OH

NHAc

2

HO

OH

AcHN HO OH

OH O

3

Fig. 5.7

Natural substrates of N-acetylneuraminic acid aldolase.

(ManNAc; 1) in the degradation pathway of the parent sialic acid (3) (Figure 5.7). The NeuA lyases found in both bacteria and animals are type-I enzymes that form a Schiﬀ base–enamine intermediate with pyruvate and promote si-face attack on the aldehyde carbonyl group with formation of a (4S) conﬁgured stereocenter. Enzyme preparations from Clostridium perfringens and Escherichia coli are commercially available, and the latter has been cloned for overexpression in E. coli [65, 66] or temperature-tolerant Serratia liquefaciens [67]. The enzyme has a broad pH optimum at approximately 7.5 and useful stability in solution at ambient temperature [68]. It has been used for synthetic applications in homogenous solution, in an immobilized form [68–72] or enclosed in a dialysis membrane [73, 74]. Because neuraminic acid is an important precursor to Zanamivir, an inhibitor of viral sialidases that is marketed for treatment of inﬂuenza infections, the large-scale synthesis of 3 has been developed as a prime example of an industrial aldolase bioconversion process at the multi-ton scale [75–77]. In this equilibrium-controlled bioconversion (equilibrium constant 12.7 m1 in favor of the retro-aldolization), the expensive 1 can be produced by integrated enzymatic in situ isomerization of inexpensive N-acetylglucosamine 4 by a combination of NeuA with an N-acylglucosamine 2epimerase (EC 5.1.3.8) catalyst in an enzyme membrane reactor [61, 78] (Figure 5.8). To facilitate product recovery, excessive 2 can be removed by formation of a separable bisulﬁte adduct, or by decomposition with yeast pyruvate decarboxylase into volatile compounds [75, 79]. The need for excess 2 might be circumvented altogether by coupling of the synthesis of 3 to a thermodynamically more favored process, e.g. combination with an irreversible sialyltransferase reaction to furnish sialyloligosaccharides [80, 81]. Extensive studies have indicated that only pyruvate is acceptable as the NeuA donor substrate, with the exception of ﬂuoropyruvate [82], but that the enzyme has fairly broad tolerance of stereochemically related aldehyde

CO2H

210

5 Enzyme-catalyzed Aldol Additions

OH

HO O

HO HO

4

OH NHAc

CO2H

O

HO OH AcHN

NH2

HN Zanamivir

NH

epimerase

steps HO HO HO

OH

1

HO

2, NeuA

NHAc O

OH

RCOHN HO OH

OH O

CO2H

3

Fig. 5.8

Industrial process for the production of N-acetylneuraminic acid as a precursor to an inﬂuenza inhibitor.

substrates as acceptor alternatives, for example several sugars and their derivatives larger than or equal in size to pentoses (Table 5.1) [62, 68, 83]. Permissible variations include replacement of the natural d-manno conﬁgured substrate with derivatives containing modiﬁcations such as epimerization, substitution, or deletion at positions C-2, C-4, or C-6 [20, 25]. Epimerization at C-2, however, is restricted to small polar substituents at strongly reduced reaction rates [84, 85]. As an example of continuous process design, KDN (6) has been produced on a 100-g scale from d-mannose and pyruvate using a pilot-scale enzyme membrane reactor (EMR) with a space–time yield of 375 g L1 d1 and an overall crystallized yield of 75% (Figure 5.9) [86]. Similarly, l-KDO (5) can be synthesized from l-arabinose [62]. The broad substrate tolerance of the catalyst for sugar precursors has recently been exploited in the equilibrium generation of sialic acid and analogs for in-situ screening of a dynamic combinatorial library [87]. Because of the importance of sialic acids in a wide range of biological recognition events, the aldolase has become popular for chemoenzymatic synthesis of a multitude of other natural and non-natural derivatives or analogs of 3 (Figure 5.10). Many examples have been reported for sialic acid modiﬁcations at C-5/C-9 [20, 25], for example diﬀerently N-acylated derivatives 8 [88–90], including amino acid conjugates [91], or 9-modiﬁed analogs 9 [92–94], in search for new neuraminidase (inﬂuenza) inhibitors. Most notably, the N-acetyl group in 1 can be either omitted [84, 85] or replaced by sterically demanding substituents such as N-Cbz (10) [95, 96] or even a non-polar phenyl group [84] without destroying activity. Large acyl substituents are also tolerated at C-6, as shown by the conversion of a Bocglycyl derivative 11 as a precursor to a ﬂuorescent sialic acid conjugate [97].

5.3 Pyruvate Aldolases Tab. 5.1

Substrate tolerance of neuraminic acid aldolase. OH R4 5 R R1 O R3 HO R2

NeuA OH

OH

R5 4

R

O

R1

pyruvate

CO2H

R3 OH R2

R1

R2

R3

R4

R5

Yield [%]

Rel. Rate [%]

Ref.

NHAc NHAc NHAc NHAc NHAc NHAc NHAc NHAc NHAc NHAc NHC(O)CH2 OH NHCbz OH OH OH OH H H Ph

H H H H H H H H H H H H H H H H F H H

OH OH OH OH OH OH OH OH OMe H OH OH OH H H OH OH OH OH

H H H H H H H H H H H H H H F H H H H

CH2 OH CH2 OAc CH2 OMe CH2 N3 CH2 OP(O)Me 2 CH2 O(l-lactoyl) CH2 O(Gly-N-Boc) CH2 F CH2 OH CH2 OH CH2 OH CH2 OH CH2 OH CH2 OH CH2 F H CH2 OH CH2 OH CH2 OH

85 84 59 84 42 53 47 22 70 70 61 75 84 67 40 66 30 36 76

100 20 – 60 – – – 60 – – – – 91 35 – 10 – 130 –

72, 68 72, 68 72 85 85 72, 85 97 85 72 100 72 96 79, 100 100 85 100 85 85, 100 100

OH

HO NeuA

HO

O

L-arabinose

pyruvate

HO

HO NeuA D-mannose

pyruvate

OH

CO2H

5

L-KDO

OH

HO HO OH

OH O

CO2H

6 Fig. 5.9

Sialic acids prepared on large scale.

D-KDN

211

5 Enzyme-catalyzed Aldol Additions

212

X HO HO

NHCOR O OH

OH

HO

NeuA

RCOHN HO OH

pyruvate

7

O

OH

X

OH CO2H

OH

AcHN HO OH

8

O

CO2H

9

NHBoc R = OtBu, CH2OH,

X = CH2OCH3, CH2OMOM, CH2OAc, CH2OBz

CH2Ph

O O

OH

HO

H O N HO OH

O

N H

OH

O

CO2H

10

O O

OH OH

AcHN HO OH

O

CO2H

11

Fig. 5.10

Neuraminic acid derivatives accessible by NeuA catalysis.

HO HO X

HO2C

NHAc O OH

HO

OH

AcHN HO HO

12

O

CO2H

13

X = N3, NH2, NHBoc O F HO HO HO

CO2H

15

HO O OH

NeuA

14

HO

OH

HO HO HO HO

O

CO2H F

16

Fig. 5.11

Sugar derivatives not accepted by NeuA in direction of synthesis or cleavage, and use of ﬂuoropyruvate for synthesis of ﬂuorosialates.

No 3-azido, 3-amino, or Boc-protected mannosamine analogs 12 were accepted by the enzyme (Figure 5.11) [98] which suggests that the presence of a 3-hydroxyl group is a necessary precondition for substrates of the aldolase. Likewise, conformationally inﬂexible acrylate 13 was not accepted in cleavage direction. By use of ﬂuoropyruvate 15 as the donor substrate a series of diastereomeric 3-deoxy-3-ﬂuoro ulosonic acids such as 16 has been prepared in good yields (> 49%) from pentoses or hexoses [82]. Such products are attractive for non-invasive in-vivo pharmacokinetic studies by NMR tomography ( 19 F derivatives) or positron-emission spectroscopy ( 18 F derivatives).

5.3 Pyruvate Aldolases

213

N-acetylL-mannosD-arabinose

L-mannose

amine

NeuA pyruvate

NeuA pyruvate

HO

OH O CO2H

HO

ent-5

OH

OH

OH

D-KDO

OH

OH

HO HO HO

ent-6

O CO2H OH

L-KDN

OH

HO AcHN HO

ent-3

O CO2H

L-NeuAc

OH

Enz O H α-4C1

OH

NeuA

HO OH

HO

H

3S Enz

Enz 3R

OH

HO Enz

O H Enz

si-face

normal attack

O H

H Enz

re-face

inverted attack

Fig. 5.12

NeuA-catalyzed synthesis of sialic acids bearing the unusual inverted (4R) conﬁguration, and three-point binding model for prediction of NeuA stereoselectivity based on conformational analysis.

In most of the examples investigated so far a high level of asymmetric induction for the (4S) conﬁguration is retained. Several carbohydrates were, however, also found to be converted with random or even inverse stereoselectivity for the C-4 conﬁguration [79, 83, 99]. Products generated by the unusual re-face attack with (4R) stereochemical preference include a number of related higher ulosonic acids of biological importance such as d-KDO (ent-5) [62, 79, 100, 101] or the enantiomers of naturally occurring sugars such as l-KDN (ent-6) [68, 71, 75, 86], or l-NeuNAc (ent-3) (Figure 5.12) [79]. Ready access to compounds of this type can be particularly valuable for investigation of the biological activity of sialoconjugates containing nonnatural sialic acid derivatives [102]. A critical and distinctive factor seems to be recognition by the enzyme catalyst of the conﬁguration at C-3 in the aldehydic substrate [62, 83]. The unusual outcome has been interpreted as a result of thermodynamic control or, alternatively, of an inverse conforma-

214

5 Enzyme-catalyzed Aldol Additions

OH H O N HO OH

O

CO2H

OH

pyruvate

OH

HO2C

OH O

CO2H

OH

HO

OH 12 OH

OH

OH

OH OH

O OH

OH

O

19

O

OH

HO HO OH

18

HO

OH

17

NeuA

O

N

HO

HO

OH

OH

HO

10

O

HO HO HO

HO

OH

HO

OH

O

20

amphotericin B

OH 16

CO2H

OR

Fig. 5.13

NeuA-catalyzed preparation of an intermediate for alkaloid synthesis, and of a synthetic precursor to the macrolide antibiotic amphotericin B.

tional preference of the diﬀerent substrate classes [62]. From an applicationoriented perspective a three-point binding model has been proposed for the conversion of substrate analogs in the direction of synthesis (Figure 5.12) [25]. On the basis of the (3S)-a- 4 C1 structure of the natural substrate and conformational analysis of its analogs, this model can predict the occasionnally observed compromise in, or even total inversion of, the facial stereoselectivity of CaC bond formation. Starting from the N-Cbz-protected aldolase product 10, azasugar 17 has been obtained stereoselectively by internal reductive amination as an analog of the bicyclic, indolizidine-type glycosidase inhibitor castanospermine (Figure 5.13) [96]. It has also been recognized that the C-12 to C-20 sequence of the macrolide antibiotic amphotericin B resembles the b-pyranose tautomer of 19. Thus, the branched-chain manno-conﬁgured substrate 18 was successfully chain-extended under the action of NeuA catalysis to yield the potential amphotericin B synthon 19 in good yield [103, 104]. A one-pot 13 C-labeling strategy has been developed for sialic acids by exploiting the reversible nature of the aldolase reaction (Figure 5.14) [105].

5.3 Pyruvate Aldolases

OH

HO

OH O

AcHN HO HO

NeuA

CO2H

3

OH

HO HO HO

OH

AcHN HO HO

1 O

L-LDH

CO2H

OH CO2H

O

20

O CO2H

NAD+

OH

HO

1. NPP'ase 2. NeuA

NHAc O

215

stage 2

21

CO2H

NADH

EtOH ADH

acetaldehyde

stage 1

Fig. 5.14

Process for [ 13 C]-labeling of neuraminic acid (and analogs) by controlled reversible aldolizations ( denotes [ 13 C]-label).

Nu*

O HO HO HO

OH

cell O

HN

NH

O OH

O

[NeuA, pyruvate]

CO2H

HO O OH OH O

22 Fig. 5.15

Cellular synthesis of a modiﬁed sialic acid by exposure of human cells to N-levulinoyl d-mannosamine 22. Resulting levulinoylated cell surface glycoproteins provide an opportunity for selective, covalent attachment of a variety of nucleophiles Nu*.

The procedure consists of stepwise enzymatic degradation of chemically prepared neuraminic acid derivatives (3) to give the corresponding modiﬁed ManNAc analogs (1), which is followed by subsequent neuraminic acid reconstruction in the presence of [3- 13 C] labeled pyruvate (21). To ensure complete conversion in the desired direction the ﬁrst step was promoted by cofactor-dependent reduction of pyruvate; the second, synthetic step, was uncoupled by destruction of the nucleotide cofactor before addition of labeled pyruvate. Sialic acids such as 20 with high label incorporation (> 87%) were obtained in good yields (46–76%). Acceptance of the non-natural N-levulinoyl d-mannosamine 22, containing a ketone moiety in the N-acyl group, by the cellular machinery in vivo has been used to produce cell-surface oligosaccharides modiﬁed in their neuraminic acid constituents (Figure 5.15) [106, 107]. The reactive ketone groups thus produced on the cell surface could then be used for versatile covalent cell redecoration under physiological conditions, for example by

O

cell

216

5 Enzyme-catalyzed Aldol Additions

O OH OH

O OH

HO

23

HO

HO HO

O CO2H

HO OH

F

HO

O CO2H

HO

25

OH

OH

O CO2H

R

OH

2

OH

24

HO

CO2H

OH OH

OH

KdoA

+

OH

HO OH O CO2H

HO

ent-5

OH

Fig. 5.16

Natural substrates of the 2-keto-3-deoxymanno-octosonic acid aldolase, and nonnatural sialic acids obtained by KdoA catalysis.

attaching functional nucleophiles (Nu*) such as ﬂuorescent hydrazine markers or toxin conjugates. 5.3.2

KDO Aldolase

The functionally related 2-keto-3-deoxy-manno-octosonate (KDO) aldolase (KdoA; correctly termed 3-deoxy-d-manno-octulosonic acid aldolase, EC 4.1.2.23) is involved in the catabolism of the eight-carbon sugar d-KDO ent5, a core constituent of the capsular polysaccharides (K-antigens) and outer membrane lipopolysaccharides (LPS, endotoxin) of Gram-negative bacteria [108], and of the cell wall of algae and a variety of plants [109]. The aldolase, which reversibly degrades ent-5 to d-arabinose 23 and pyruvate (Figure 5.16), has been isolated from Aerobacter cloacae, E. coli, and an Aureobacterium barkerei strain in which the enzyme seems to be located in the cell wall or membrane fraction [110]. Partially puriﬁed KdoA preparations have been studied for synthetic applications [101, 111]; these studies have shown that, similar to NeuA, the enzyme has a broad substrate speciﬁcity for aldoses (Table 5.2) whereas pyruvate was found to be irreplaceable. As a notable distinction, KdoA was also active on smaller acceptors such as glyceraldehyde. Preparative applications, e.g. that for the synthesis of KDO (ent-5) and its homologs or analogs 24/25, suﬀer from a less attractive equilibrium constant of 13 m1 in direction of synthesis [56]. The stereochemical course of aldol additions generally seems to adhere to re-face attack on the aldehyde carbonyl which is complementary to that of NeuA. On the basis of the results published so far it can be concluded that a (3R) conﬁguration is necessary (but not suﬃcient), and that stereochemical requirements at C-2 are less stringent

5.3 Pyruvate Aldolases Tab. 5.2

Substrate tolerance of 2-keto-3-deoxy-manno-octosonic acid aldolase [101]. OH

O

R3

H R

2

R

R1 R3

KdoA

R2 HO

pyruvate

1

O CO2H OH

Substrate

R1

R2

R3

Yield [%]

Rel. Rate [%]

d-Altrose l-Mannose d-Arabinose d-Ribose 2-Deoxy-2-ﬂuoro-d-ribose 2-Deoxy-d-ribose 5-Azido-2,5-dideoxy-d-ribose d-Threose d-Erythrose d-Glyceraldehyde l-Glyceraldehyde

OH H OH H F H H OH H H OH

H OH H OH H H H H OH OH H

(RR)-(CHOH)2 aCH2 OH (SS)-(CHOH)2 aCH2 OH (R)-CHOHaCH2 OH (R)-CHOHaCH2 OH (R)-CHOHaCH2 OH (R)-CHOHaCH2 OH (R)-CHOHaCH2 N3 CH2 OH CH2 OH H H

– 61 67 57 19 47 – – 39 11 –

25 15 100 72 46 71 15 128 93 23 36

[101]. For simple product recovery excess pyruvate can be decomposed with pyruvate decarboxylase. 5.3.3

DAHP Synthase

In the biosynthesis of aromatic amino acids in microorganisms and plants, 3-deoxy-d-arabino-heptulosonic acid 7-phosphate synthase (DAHP synthase or AroS; EC 4.1.2.15) is a pivotal enzyme for carbon framework construction. The enzyme transfers a pyruvate moiety from phosphoenolpyruvate (PEP) to d-erythrose-4-phosphate (Ery4P) in an aldol-like reaction [112, 113]. Use of the activated donor renders formation of DAHP (26), the key metabolic intermediate, practically irreversible (Figure 5.17). The AroS enzyme has been puriﬁed from E. coli [114] and Streptomyces rimosus [115] and has been cloned from E. coli, Salmonella typhimurium, and potato [116, 117]. The structure of an E. coli isozyme has been determined in a complex with PEP at high resolution [118]. The cytosolic AroS isoenzyme prevalent in higher plants has been reported to be remarkably highly tolerant of acceptor substrates, including glyoxalate, glycolaldehyde, and most 3–5-carbon sugars or their phosphates [119]. For synthesis of 26 (Figure 5.17), an immobilized enzyme system was set up to generate the labile acceptor aldehyde in situ by ketol transfer from dfructose 6-phosphate [120]. The twofold driving force from PEP used in the integrated scheme as both a phosphoryl and aldol donor secured complete

217

218

5 Enzyme-catalyzed Aldol Additions

D-fructose 6-phosphate

D-sedoheptulose 7-phosphate

transketolase

OH

D-ribose 5-phosphate

=O

3PO

H

OPO3=

Pi

PEP

O

HO HO

AroS

O

OH

OH

D-erythrose 4-phosphate

CO2H

HO CO2H

HO

OH OH

27

HO

26, DAHP

85%

CO2H

OH OH

CO2H

28

OH

OH

OH

OH

OH

29

30

Fig. 5.17

Multi-enzymatic scheme for synthesis of 3-deoxy-arabinoheptulosonic acid 7-phosphate based on the catalysis of DAHP synthase, and products generated consecutively in vivo by advanced microbial pathway engineering.

conversion of the starting material. Generally, synthesis of 26 on a larger scale is more eﬃcient and economical by using engineered whole cells that contain a plasmid coding for AroS [121]; thereby necessary substrates and enzymes are intrinsically provided by regular cell metabolism. Similar elaborate schemes have been designed to produce quinic acid (27) [122], shikimic acid (28) [123], cyclohexadiene trans-diols, for example 29 [208], or phenols such as 30 [124–126] by engineered microbial biosynthesis. Puriﬁed enzyme from Escherichia coli has been studied for its reactivity with homologous ﬁve-carbon phosphosugars to yield KDO derivatives, and with modiﬁed nucleophiles. Thus, the stereochemically distinct ﬂuoro-analogs of PEP (Z)- and (E)-31 could be separately condensed with d-erythrose 4-phosphate to yield, stereospeciﬁcally, the corresponding (3S)and (3R)-conﬁgured, ﬂuoro-substituted DAHP derivatives 32 and 33, respectively (Figure 5.18) [127]. This provides direct evidence that the enzyme catalyzes si face addition of the C-3 of PEP to the re face of the acceptor aldehyde. 5.3.4

KDPG Aldolase and Related Enzymes

An aldolase speciﬁc for cleavage of 2-keto-3-deoxy-6-phospho-d-gluconate (35) (KDPGlc aldolase or KdgA; EC 4.1.2.14) is produced by many species of bacteria for degradation of 6-phosphogluconate to give pyruvate and d-

5.3 Pyruvate Aldolases

OPO3=

HO

OPO3=

OPO3= F O

CO2H

HO

CO2H

F

F

HO

O

F

CO2H

OPO3=

E-31

(3R)-32

Z-31

219

OH

CO2H

HO

OH

(3S)-33

Fig. 5.18

Stereospeciﬁc conversion of 3-ﬂuoro-labeled PEP analogs to substituted DAHP derivatives by use of DAHP synthase.

KDPGlc aldolase

OH =

=O

3PO

CO2H

S

+

3PO

O

HO

2 OH

OH =O

KDPGal aldolase

3PO

R

OH

=

O CO2H

O3PO HO

35 O

Aldol reactions catalyzed in vivo by the 2-keto-3-deoxy-6-phospho-d-gluconate and 2-keto-3-deoxy-6-phospho-d-galactonate aldolases.

glyceraldehyde 3-phosphate 34 (Figure 5.19). The equilibrium constant favors synthesis (10 3 m1 ) [128]. Enzymes belonging to class I aldolases have been isolated [129–131] and cloned [132–135] from a variety of microbial sources. Genome sequencing of E. coli revealed that the enzyme is identical with 2-keto-4-hydroxyglutarate aldolase (KHG aldolase or KhgA; EC 4.1.3.16), which is involved in the catabolism of hydroxyproline [136, 137]. This activity is responsible for the retroaldolization of 4-hydroxy-2-oxoglutaric acid to pyruvate 2 and glyoxalate. In both cases, the enzyme recognizes an identical (4S) conﬁguration. Enzyme preparations from liver or microbial sources were reported to have rather high substrate speciﬁcity [138] for the natural phosphorylated acceptor d-34 and, but at much reduced reaction rates, to have rather broad substrate tolerance for polar, short-chain aldehydes (Table 5.3) [139–141]. Simple aliphatic or aromatic aldehydes are not converted. High stereoselectivity of the enzyme has been used in the preparation of compounds 39 and 40 and in a two-step enzymatic synthesis of 38, the N-terminal amino acid portion of nikkomycin antibiotics (Figure 5.20) [142]. Thermophilic KdgA from Thermotoga maritima has been shown to have substrate tolerance with

OH CO2H

R

Fig. 5.19

OH CO2H

S

OH

O H

34

O3PO

=O

O

36

5 Enzyme-catalyzed Aldol Additions

220

Tab. 5.3

Substrate tolerance of 2-keto-3-deoxy-6-phospho-d-gluconate aldolase [140]. O R

H

O COOH

Substrate

R

Rel. Rate [%]

d-Glyceraldehyde 3-phosphate 3-Nitropropanal Chloroethanal d-Glyceraldehyde d-Lactaldehyde d-Ribose 5-phosphate Erythrose Glycolaldehyde

d-CH2 OHaCH2 OPO3 ¼ CH2 aCH2 NO2 CH2 Cl d-CHOHaCH2 OH d-CHOHaCH3 d-ribo-(CHOH)3 aCH2 OPO3 ¼ erythro-(CHOH)2 aCH2 OH CH2 OH

100.0 1.6 1.0 0.8 0.2 0.04 0.01 0.01

H

NH3

pyruvate

CO2H

N

O

37

R

pyruvate

KDPGlc aldolase N

OH

KDPGlc aldolase

OH

O

NADH

CO2H

N

PheDH NAD+

38

OH

NH2

ee >99.7% CO2 OH O

CO2H

S

HO

OH

H 3C S

HO

39

O

formate DH HCO2–

OH CO2H

40

Fig. 5.20

Stereoselective synthesis of the amino acid portion of nikkomycin antibiotics and hexulosonic acids using KDPGlc aldolase.

a breadth that seems comparable to, but qualitatively distinct from, that of mesophilic enzymes, although with diminished stereoselectivity [135]. The aldolase from E. coli has been mutated for improved acceptance of non-phosphorylated and enantiomeric substrates towards facilitated enzymatic syntheses of both d and l sugars [143, 144]. Comparable with the situation for the sialic acid and KDO lyases (vide supra), a class I lyase complementary to the KDPGlc aldolase is known that has a stereopreference for the (4S) conﬁguration (Figure 5.19). The aldolase, which acts on 2-keto-3-deoxy-6-phospho-d-galactonate (36) (KDPGal aldolase; EC 4.1.2.21) and is less abundant [145, 146], has recently been studied for synthetic applications [147].

5.4 Dihydroxyacetone Phosphate Aldolases

O X

O

H

OPO3=

OH OH

=O

3PO

OH

O OPO3=

HO

HO

FruA

RhuA

3S,4R 3R,4S

3PO

OH

O

HO

O OPO3=

HO

OPO3= HO

OH H 3C

42

D-fructose 1,6-bisphosphate

=O

41

HO

43

L-rhamnulose 1-phosphate

TagA

FucA

OH

3S,4S 3R,4R

44

D-tagatose 1,6-bisphosphate

O OPO3=

H 3C HO

HO

45

L-fuculose 1-phosphate

Fig. 5.21

Aldol reactions catalyzed in vivo by the four stereo-complementary dihydroxyacetone phosphate-dependent aldolases.

5.4

Dihydroxyacetone Phosphate Aldolases

Whereas pyruvate aldolases form only a single stereogenic center, the aldolases speciﬁc for dihydroxyacetone phosphate (DHAP, 41) as a nucleophile create two new asymmetric centers at the termini of the new CaC bond. Particularly useful for synthetic applications is that Nature has evolved a full set of four unique aldolases (Figure 5.21) to cleave all possible stereochemical permutations of the vicinal diol at C-3/C-4 of ketose 1-phosphates 42–45 during the retro-aldol cleavage [26]. These aldolases have proved to be exceptionally powerful tools for asymmetric synthesis, particularly stereocontrolled synthesis of polyoxygenated compounds, because of their relaxed substrate speciﬁcity, high level of stereocontrol, and commercial availability. In the direction of synthesis this situation formally enables generation of all four possible stereoisomers of a desired product in building-block fashion [22, 25, 26]. In this manner the deliberate preparation of a speciﬁc target molecule can be addressed simply by choosing the corresponding enzyme and suitable starting material, aﬀording full control over constitution and absolute and relative conﬁguration of the desired product. Alternatively, the stereocomplementary nature of enzymes also enables a combinatorial, stereodivergent approach (Figure 5.1) to the generation of all stereoisomers in a

221

222 =

O3PO

5 Enzyme-catalyzed Aldol Additions

O HO HO

OH OPO3=

42

OH

O

FruA

OPO3=

HO

+

41

OPO3=

O

34

Fig. 5.22

Natural glycolytic substrates of fructose 1,6-bisphosphate aldolases.

small parallel library of diastereoisomers, e.g. for bioactivity screening [25]. Properties of the individual enzymes are therefore discussed separately from more general treatment of their preparative applications. The DHAP aldolases are quite speciﬁc for the phosphorylated nucleophile 41, which must therefore be prepared independently or generated in situ (cf. Section 5.4.4). Initial aldol products will thus contain a phosphate ester moiety, which facilitates product isolation, for example by barium salt precipitation or by use of ion-exchange techniques. The corresponding phosphate free compounds can be easily obtained by mild enzymatic hydrolysis using an inexpensive alkaline phosphatase at pH 8–9 [148], whereas baselabile compounds might require working at pH 5–6 using a more expensive acid phosphatase [149]. 5.4.1

FruA

The d-fructose 1,6-bisphosphate aldolase (FruA; EC 4.1.2.13) catalyzes, in vivo, the equilibrium addition of 41 to d-glyceraldehyde 3-phosphate (GA3P, 34) to give d-fructose 1,6-bisphosphate (42). The equilibrium constant for this reaction of 10 4 m1 strongly favors synthesis [56]. The enzyme occurs ubiquitously and has been isolated from a variety of prokaryotic and eukaryotic sources, both as class I and class II forms [29]. Corresponding FruA genes have been cloned from a variety of sources and overexpressed [25]. Typically, class I FruA enzymes are tetrameric whereas the class II FruA are dimers, with subunits of @40 kDa. The microbial class II aldolases are usually much more stable in solution (half-lives of several weeks to months) than their mammalian counterparts of class I (a few days) [149–151]. The class I FruA isolated from rabbit muscle (‘‘RAMA’’) is the aldolase used for preparative synthesis in the widest sense, because of its commercial availability and useful speciﬁc activity of @20 U mg1 . Its operating stability in solution is limiting, but recently more robust homologous enzymes have been cloned, e.g. from Staphylococcus carnosus [152] or from the extremophilic Thermus aquaticus [153], which promise to be unusually stable in synthetic applications [154]. Attempts at catalyst immobilization have been performed with rabbit muscle FruA, which has been covalently attached to microcarrier beads [58], cross-linked in enzyme crystals (CLEC) [59], or enclosed in a dialysis membrane [73]. It was recently shown that

5.4 Dihydroxyacetone Phosphate Aldolases Tab. 5.4

Substrate tolerance of fructose 1,6-bisphosphate aldolase. O R

OH

FruA H

DHAP

O OPO3=

R OH

R

Rel. Rate [%]

Yield [%]

Ref.

d-CHOHaCH2 OPO3 ¼ H CH3 CH2 Cl CH2 aCH3 CH2 aCH2 aCOOH CH2 OCH2 C6 H5 d-CH(OCH3 )aCH2 OH CH2 OH d-CHOHaCH3 l-CHOHaCH3 dl-CHOHaC2 H5 CH2 aCH2 OH CH2 aC(CH3 )2 OH dl-CHOHaCH2 F dl-CHOHaCH2 Cl dl-CHOHaCH2 aCHbCH2

100 105 120 340 105 – 25 22 33 10 10 10 – – – – –

95 – – 50 73 81 75 56 84 87 80 82 83 50 95 90 85

149, 177 149 149 149, 212 149 177 149 149 149, 177 149, 207 149 149 206, 207 205 207 206 206

less polar substrates can be converted as highly concentrated water-in-oil emulsions [63]. Literally hundreds of aldehydes have so far been tested successfully by enzymatic assay and preparative experiments as a replacement for 34 in rabbit muscle FruA catalyzed aldol additions [18, 25], and most of the corresponding aldol products have been isolated and characterized. A compilation of selected typical substrates and their reaction products is provided in Table 5.4, and further examples are indicated in Section 5.4.5. In comparison, metal dependant FruA enzymes are more speciﬁc for phosphorylated substrates and accept non-phosphorylated substrate analogs only with much reduced activity (< 1%). For chiral substituted aldehydes, racemic precursors are most often more readily available than enantiomerically pure material, but must be resolved either before or during the aldol step. Rabbit FruA discriminates between the enantiomers of its natural substrate with a 20:1 preference for d-GA3P (34) over its l antipode [149]. Assistance from anionic binding was revealed by a study on a homologous series of carboxylated 2-hydroxyaldehydes for which enantioselectivity was optimum when the distance of the charged group equaled that of 34 (Figure 5.23) [155]. The resolution of racemic substrates is not, however, generally useful, because the kinetic enantiose-

223

224

5 Enzyme-catalyzed Aldol Additions

O

O

O –O

–O

P

–O

OH

46 E > 10

O O OH

34 E > 10

Fig. 5.23

Kinetic enantiopreference of rabbit muscle FruA.

lectivity for non-ionic aldehydes is rather low [149]. 3-Azido substituents (95) can lead to up to ninefold preference for enantiomers in kinetically controlled experiments [156] whereas hydroxyl (usually some preference for the d antipodes) and derived functionality, or chiral centers at a larger distance, rarely enable more than statistical diastereomer formation. 5.4.2

TagA

The d-tagatose 1,6-bisphosphate aldolase (TagA; EC 4.1.2.n) is involved in the catabolism of d-galacto-conﬁgured carbohydrates and catalyzes the reversible cleavage of d-tagatose 1,6-bisphosphate (44) to d-glyceraldehyde 3phosphate (34) and dihydroxyacetone phosphate (41). Enzymes of class I that occur in a variety of Coccus species seem to have apparently no stereochemical selectivity with regard to distinction between 42 and 44 [157], whereas class II aldolases from Gram-positive microorganisms are highly stereoselective for the natural substrate in both cleavage and synthesis directions [158, 159]. The genes coding for TagA have been cloned from Coccus strains [160] and from E. coli [53, 161]. The reverse, synthetic reaction can be used to prepare ketose bisphosphates, as has been demonstrated by an expeditious multienzymatic synthesis of the all-cis (3S,4S)-conﬁgured d-tagatose 1,6-bisphosphate 44 from dihydroxyacetone including its cofactor-dependent phosphorylation, using the puriﬁed TagA from E. coli (Figure 5.24) [158, 159]. The aldolase also accepts a range of unphosphorylated aldehydes as substrates but produces diastereomeric mixtures only. This lack of stereoselectivity with generic substrate analogs, which makes native TagA enzymes synthetically less useful, has stimulated recent X-ray structure determination and protein engineering to improve its properties [53, 54, 162]. 5.4.3

RhuA and FucA

The l-rhamnulose 1-phosphate aldolase (RhuA; EC 4.1.2.19) and the lfuculose 1-phosphate aldolase (FucA; EC 4.1.2.17) are found in many microorganisms where they are responsible for the degradation of deoxysugars l-rhamnose and l-fucose to give 41 and l-lactaldehyde (Figure 5.25). RhuA is speciﬁc for cleavage and synthesis of a l-threo diol unit whereas FucA

5.4 Dihydroxyacetone Phosphate Aldolases

glycerol kinase

O HO

OH

O OPO3=

HO

47

41 ATP

=O PO 3

ADP

TagA

TPI pyruvate pyruvate kinase PEP

225

HO

O HO

OH

44

OH OPO3=

O

34

H Fig. 5.24

Enzymatic one-pot synthesis of tagatose 1,6-bisphosphate based on the stereoselective TagA from E. coli.

OPO3=

O

HO

OH

H 3C

OH

RhuA

DHAP

FucA

O H 3C

H OH

43

OPO3=

O

DHAP

OH

H 3C HO

OH

45

Fig. 5.25

Natural substrates of microbial deoxysugar phosphate aldolases.

recognizes the corresponding d-erythro conﬁguration. Both enzymes have been isolated from several sources [163–165], and the proteins from E. coli have been shown to be homotetrameric Zn 2þ -dependent aldolases with subunit molecular weights of @25 kDa and @30 kDa, respectively [150]. Cloning of RhuA [166, 167] and FucA [168] enzymes from several microorganisms has been reported. Eﬃcient overexpression [150, 160, 169] has set the stage for X-ray structure determinations of both E. coli proteins [48, 49]. Like a number of other aldolases, both the RhuA and FucA enzymes are commercially available. Overall practical features make the RhuA and FucA enzymes quite similar for synthetic applications. Both metalloproteins have very high stability in the presence of low Zn 2þ concentrations with half-lives in the range of months at room temperature, and the enzymes even tolerate the presence of large proportions of organic cosolvents (b30%) [150]. Both have very broad substrate tolerance for variously substituted aldehydes, which is very similar to that of the FruA enzymes, with conversion rates generally being usefully high (Table 5.5). Characteristically, of all the DHAP aldolases yet investigated the RhuA has the greatest tolerance of sterically congested acceptor substrates, as exempliﬁed by the conversion of the tertiary aldehyde 2,2-dimethyl-3-hydroxypropanal 48 (Figure 5.26) [25]. Aldehydes carrying an anionically charged group close to the carbonyl group, for example gly-

OPO3=

226

5 Enzyme-catalyzed Aldol Additions Tab. 5.5

Substrate tolerance of l-rhamnulose 1-phosphate and l-fuculose 1-phosphate aldolases [150, 170]. O R

H

DHAP

O

OH

aldolase

OH

O

OPO3= + R

R

OPO3= OH

OH

R

Rel. Rate [%]

RhuA Selectivity threo:erythro

Yield [%]

Rel. Rate [%]

FucA Selectivity threo:erythro

Yield [%]

l-CH2 OHaCH3 CH2 OH d-CHOHaCH2 OH l-CHOHaCH2 OH CH2 aCH2 OH CHOHaCH2 OCH3 CHOHaCH2 N3 CHOHaCH2 F H CH3 CH(CH3 )2

100 43 42 41 29 – – – 22 32 22

>97:3 >97:3 >97:3 >97:3 >97:3 >97:3 >97:3 >97:3 – 69:31 97:3

95 82 84 91 73 77 97 95 81 84 88

100 38 28 17 11 – – – 44 14 20

<3:97 <3:97 <3:97 <3:97 <3:97 <3:97 <3:97 <3:97 – 5:95 30:70

83 85 82 86 78 83 80 86 73 54 58

1. RhuA, DHAP 2. P'ase HO

CHO

48

O HO HO

49

OH

OH

Fig. 5.26

Substrate tolerance of RhuA for tertiary hydroxyaldehyde.

ceraldehyde phosphates, other sugar phosphates, or glyoxylic acid, are not converted by the deoxysugar aldolases [25, 148]. The stereospeciﬁcity of both enzymes for the absolute (3R) conﬁguration is mechanism-based (vide supra), and RhuA generally directs attack of the DHAP enolate to the re face of an approaching aldehyde carbonyl and is thereby speciﬁc for synthesis of a (3R,4S)-trans diol unit [150] whereas FucA controls si face attack to create the corresponding (3R,4R)-cis conﬁguration [150, 169]. This speciﬁcity for a vicinal conﬁguration is, however, somewhat substrate-dependent, in that simple aliphatic aldehydes can give rise to a certain fraction of the opposite diastereomer [25, 150]. In general, stereochemical ﬁdelity is usually higher, and diastereospeciﬁc results are observed more often with the FucA than with the RhuA enzyme. Stereocontrol is, in general, usually highly eﬀective with aldehydes carrying a 2- or 3-hydroxyl group (Table 5.5). In addition, both aldolases have strong kinetic

5.4 Dihydroxyacetone Phosphate Aldolases

RhuA

OPO3=

O

OH

R

O R

HO

O +

R

OH

D,L-50

de ≥ 90%

51

DHAP

OH

R HO

O +

R

H OH

OH

52

D-50

OPO3=

O

FucA

H OH

OH H +

227

de ≥ 90%

D-50

R = H3C–, H5C2–, H2C=CH–, H2C=CH–CH2–, FH2C–, N3CH2–, H3COCH2– Fig. 5.27

Kinetic enantiopreference of class II DHAP aldolases useful for racemic resolution of a-hydroxyaldehydes.

preference for l-conﬁgured enantiomers of 2-hydroxyaldehydes 50 (Figure 5.27), which facilitates racemate resolution [170, 171]. Essentially, this feature enables concurrent determination of three contiguous chiral centers in ﬁnal products 51 or 52 having an l conﬁguration (d.e. b95) even when starting from more readily accessible racemic material. 5.4.4

DHAP Synthesis

Apparently, all DHAP aldolases are highly speciﬁc for 41 as the donor component for mechanistic reasons [32, 50], which necessitates economical access to this compound for synthetic applications. Owing to the limited stability of 41 in solution, particularly at alkaline pH [172, 173], it is preferentially generated in situ to avoid high stationary concentrations. The most convenient method is formation of two equivalents of 41 by retro-aldol cleavage from commercially available fructose 1,6-bisphosphate (FBP, 42) by the combined action of FruA and triose phosphate isomerase (Figures 5.22 and 5.28 inset) [149]. This scheme has been extended into a highly integrated, multienzymatic scheme for eﬃcacious in-situ preparation of 41 from inexpensive glucose, fructose, or sucrose by using an ‘‘artiﬁcial metabolism’’ in vitro made up from up to seven inexpensive enzymes (Figure 5.28) [174]. Complications can arise from incomplete conversion, depending on the thermodynamic stability of the ﬁnal product relative to 42, because of the overall equilibrium nature of the system, and from diﬃculties separating the products from other phosphorylated components. Dihydroxyacetone 47 can be enzymatically phosphorylated using a glycerol kinase with ATP regeneration (see, for example, Figure 5.24) [175], or by transphosphorylation from phosphatidyl choline using phospholipases

228

5 Enzyme-catalyzed Aldol Additions

OH HO HO

O OH

HO OH O sucrose HOH2C

CH2OH

HO O

OH

O

HO HO

glucose OH

O OH HO

OH

OH OH

fructose

5 enzymes ATP, PEP

=

O3PO

O OH HO HO

OPO3=

42 FruA

OH OPO3=

O H

O

TPI

34

OPO3=

HO

41

Fig. 5.28

Multi-enzymatic ‘‘artiﬁcial metabolism’’ for in-situ generation of dihydroxyacetone phosphate from inexpensive sugars.

[176]. Alternatively, 41 can be formed from glycerol by successive phosphorylation and oxidation eﬀected by a combination of glycerol kinase and glycerol phosphate dehydrogenase, with an integrated double ATP/NADþ cofactor recycling system [148]. A more advanced technique for clean generation of 41 in situ is based on oxidation of l-glycerol 3-phosphate (74) catalyzed by microbial ﬂavinedependent glycerol phosphate oxidases (GPO; Figure 5.29, box) [177]. This method generates 41 practically quantitatively and with high chemical purity without a need for separate cofactor regeneration. Both oxygen from air or from a H2 O2 –catalase system can be used to sustain oxygenation. Because DHAP aldolases were found to be insensitive to oxygenated solutions, the oxidative generation of 41 can be smoothly coupled to synthetic aldol reactions [177]. This method has recently been extended to include reversible glycerol phosphorylation from inexpensive pyrophosphate (Figure 5.29).

5.4 Dihydroxyacetone Phosphate Aldolases OH

229

OH

HO

O

OH

OH 4 steps, 1 pot

glycerol

OH

PPi

Pi phytase pH 4.0

phytase pH 4.0

Pi

GPO pH 7.5

OH =

HO

OH

OH

O OPO3=

butanal

H 2O 2

OH

41

H 2O

OPO3=

HO

OPO3=

HO O2 1/2

FruA pH 7.5

O

OPO3

+

Fig. 5.29

Enzymatic in-situ generation of dihydroxyacetone phosphate for stereoselective aldol reactions using DHAP aldolases (box), and extension by pH-controlled, integrated precursor preparation and product liberation.

O HO

O O

41

PO3

=

53

HO

O S

54

PO3

=

HO

55

H N

O PO3

=

HO

PO3=

56

Fig. 5.30

Substrate analogs of dihydroxyacetone phosphate accessible by the GPO oxidation method.

Because phytase, an inexpensive acid phosphatase, is only active at low pH but virtually inactive at pH 7.5 in which aldolases have their catalytic optimum, this enables the independent staging of a one-pot synthetic cascade between (1) transphosphorylation, (2) aldolization, and (3) product dephosphorylation simply by switching the pH [178]. The GPO procedure can also be used for preparative synthesis of the corresponding phosphorothioate (54), phosphoramidate (55), and methylene phosphonate (56) analogs of 41 (Figure 5.30) from suitable diol precursors [179] which are used as aldolase substrates [177]. Such isosteric replacements of the phosphate ester oxygen were found to be tolerated by several class I and class II aldolases, and only some speciﬁc enzymes failed to accept the less polar phosphonate 56 [180]. Thus, sugar phosphonates (for example, 88 and 89) that mimic metabolic intermediates but are hydrolytically stable to phosphatase degradation can be rapidly synthesized (see Figure 5.42).

230

5 Enzyme-catalyzed Aldol Additions

O HO

OH O

47

HO

HX

OX

57 X = AsO3=

ketose

ketose X-ester

DHAP aldolase aldehyde

VO4=

Fig. 5.31

Spontaneous, reversible formation of arsenate and vanadate analogs of dihydroxyacetone phosphate in situ for enzymatic aldol additions.

Interestingly enough, dihydroxyacetone 47 in the presence of higher concentrations of inorganic arsenate reacts reversibly to form the corresponding arsenate ester 57 in situ which can replace 41 as a donor in enzymecatalyzed aldol reactions (Figure 5.31) [181, 182]. This procedure suﬀers from rather low reaction rates and the high toxicity of arsenates, however. Inorganic vanadate also spontaneously forms the corresponding vanadate ester analog under conditions that reduce its oxidation potential; so far only RhuA has been shown to accept the vanadate mimic 57 for preparative conversions [25]. Conversely, the dihydroxyacetone sulfate analog cannot be used in CaC bond-formation [183], and no replacement of the free hydroxy function in 41 for other electron-withdrawing substituents is tolerated [149]. Several alternative procedures have been developed for chemical synthesis of 41. Acetals of the dimer of 47 can be phosphoryated using phosphoroxy chloride [184] or diphenyl chlorophosphate [185], or by phosphitylation then oxidation [186]. After phosphate deprotection, free 41 is obtained, by acid hydrolysis, in overall yields of up to 60%. New procedures have recently been developed by controlled successive substitutions of dibromoacetone [187, 188] or by controlled hydrolysis of a cyclic phosphate diester of 47 [189]. 5.4.5

Applications

Four types of DHAP aldolase with distinct stereospeciﬁcity are available that have rather similar broad substrate tolerance (Figure 5.21). Despite having a rather unique physiological function in vivo, each enzyme will accept most diﬀerently substituted or unsubstituted aliphatic aldehydes in vitro as the acceptor component at a synthetically useful rate. Thus, from one common aldehyde substrate a full set of four possible diastereomeric

5.4 Dihydroxyacetone Phosphate Aldolases

aldolase DHAP

OH R

OH

OPO3=

OH

P'ase

R

CHO

OH

HO

O

aldolase DHAP

R

CHO

OH

OH

OH

ketofuranose

O

P'ase

R

O

R OH

3-hydroxyaldehyde

OH OH

2-hydroxyaldehyde

OH

O

R

231

OPO3=

OH OH OH

OH ketopyranose

Fig. 5.32

General scheme for formation of ketofuranoses and ketopyranoses from 2- and 3-hydroxyaldehydes.

products becomes synthetically accessible in a stereodivergent manner (Figure 5.1). This building block-type feature renders the DHAP aldolases particularly powerful catalysts for asymmetric synthesis [25]. The FruA from rabbit muscle has been most extensively applied in synthesis, because of its commercial availability but results discussed below should be considered as exemplary demonstrations only, and complementary conversions can probably also be achieved by use of one of the other enzymes with diﬀerent aldol stereospeciﬁcity. A vast number of aldehydes have so far been used as substrates of this set of aldolases in preparative experiments [20, 24, 25, 190]. With the exception of generic aldehydes, acceptor components must be prepared by chemical synthesis. In general, ozonolysis of suitable oleﬁns (with appropriate removal of the second fragment if this also is a substrate) or acid-catalyzed acetal deprotection are convenient routes for generation of aldehyde substrates under mild conditions. Chiral aldehydes require either asymmetric synthesis of the respective enantiomer or separation of diastereomeric products produced from racemic material. In speciﬁc cases racemate resolution can be eﬀected by the enantiomer selectivity of an aldolase (kinetic resolution; Figure 5.27) or when isomeric products have signiﬁcantly different stability (thermodynamic resolution; vide infra). Typical applications of the DHAP aldolases include the synthesis of monosaccharides and derivatives of sugars from suitable functionalized aldehyde precursors. High conversion rates and yields are usually achieved with 2- or 3-hydroxyaldehydes, because for these compounds reaction equilibria beneﬁt from the cyclization of the products in aqueous solution to give more stable furanose or pyranose isomers (Figure 5.32). For example, enantiomers of glyceraldehyde are good substrates, and stereoselective addition of dihydroxyacetone phosphate produces enantiomerically pure ketohexose

5 Enzyme-catalyzed Aldol Additions

232

1. AD-mix-β, 2. Pd(OH)2, H2

OH L

CHO

1. RhuA, DHAP 2. P'ase

OH

OH

OH

O OH

OH

O

OH

L-fructose

O 1. AD-mix-α, 2. Pd(OH)2, H2

58

OH D

CHO

OH

1. FruA, DHAP 2. P'ase

OH

OH

O OH

OH

OH

D-fructose

Fig. 5.33

Combination of metal-assisted asymmetric synthesis and enzymatic aldolization.

1-phosphates in high yield [149, 150, 170]; from these the free ketosugars are obtained by enzymatic dephosphorylation. Only TagA catalysis is an exception, because this enzyme is not erythroselective with unphosphorylated substrates [158, 159]. Using RhuA and FucA, the less common l-conﬁgured ketohexoses l-fructose and l-tagatose can be obtained as pure diastereomers from racemic glyceraldehyde also, because of the high kinetic enantiomer selectivity of these catalysts (Figure 5.27) [170]. Alternatively, non-racemic hydroxyaldehydes can be prepared for use in aldolase reactions by Sharpless asymmetric dihydroxylation from suitably protected precursors 58 (Figure 5.33) [191]. This general approach has been followed for the de novo synthesis of a multitude of diﬀerently substituted, unsaturated [192, 193], or regiospeciﬁcally labeled sugars [194]. Unusual branched-chain (60, 61) and spiroanulated sugars (63, 64) have been synthesized from the corresponding aldehyde precursors (Figure 5.34) [174]. 6-Substituted d-fructofuranoside derivatives such as aromatic sulfonamide 62 (a low nanomolar Trypanosoma brucei inhibitor) [197] are accessible via 6-azido-6-deoxyfructose from 3azido-2(R)-hydroxypropanal 95 by FruA catalysis [151, 196]. In an approach resembling the ‘‘inversion strategy’’ (vide infra) an a-C-mannoside 65 has been prepared from d-ribose 5-phosphate [216]. The synthesis of 6-C-perﬂuoroalkyl-d-fructose 66 met challenges from the strong hydrophobicity and electron-withdrawing capacity of a ﬂuorous chain, and the product’s potential surfactant properties [198]. The l-sorbo-conﬁgured homoC-nucleoside analog 67 has been synthesized as a structural analog of adenosine from an enantiomerically pure (S)-aldehyde precursor [195]. FruA-based, diastereoselective chain extension of chiral pool carbohydrates or their corresponding phosphates has been achieved in the synthesis of high-carbon ketose 1-mono- or 1,n-bisphosphates [199]. Such heptulose to nonulose derivatives are of biological importance as phosphorylated intermediates of the pentose phosphate pathway or as sialic acid analogs (for

5.4 Dihydroxyacetone Phosphate Aldolases

OH

OH

=

O3PO

O HO

OH OH

(CH3)2N

OH

65

HO OH

O OCH3 HO OH HO

HO HO

O

H3C

59

H N SO2

HO

OH

OPO3=

62

OH

HO

OPO3=

61

OH

O

63

OH

O HO

H3 C

60

O HO HO

H3 C

OPO3=

233

O

OH CH3

HO

OPO3=

OH

64 N

OH

OH O

HO

O

F17C8 OH

HO

N OH

OH

OH

66

HO

O HO

HO HO

67

Fig. 5.34

Examples of product structures accessible by enzymatic aldolization.

example, 59) that are diﬃcult to isolate from natural sources. Synthesis of sugar bisphosphates is rather unique for FruA catalysis, because the deoxy sugar aldolases (FucA, RhuA) will not usually accept anionically charged substrates. On the basis of FruA-catalyzed aldol reactions 3-deoxy-d-arabinoheptulosonic acid 7-phosphate (DAHP, 26), an intermediate of the shikimic acid pathway, has been synthesized from N-acetylaspartic semialdehyde 68 (Figure 5.35) [200]. Precursors to KDO (Section 5.3.2) and its 4-deoxy analog (71) have been prepared by FruA catalysis from aldehydes 70 that incorporate an acrylic moiety for further functionalization [201]. Pendant anionically charged chains have been extended from O- or Cglycosidic aldehydes (72) to furnish low-molecular-weight mimics of the sialyl Lewis X tetrasaccharide, for example 73 (Figure 5.36) [202]. Other higher-carbon sugar derivatives such as the bicyclic sugar 76 have been prepared by diastereoselective chain extension from simple alkyl galactosides (75; Figure 5.36) after their terminal oxidation in situ by use of a galactose oxidase (GalO; EC 1.1.3.9). The whole scheme can be conveniently eﬀected as a one-pot operation including the parallel generation of 41 by the GPO method [203]. Further bicyclic carbohydrate structures similar to 76 have also been prepared by uni- [152] and bidirectional extension of dialdehyde substrates (see Figure 5.48) [204]. The Zn 2þ -dependent aldolases facilitate eﬀective kinetic resolution of racemic 2-hydroxyaldehydes (rac-50) as substrates by an overwhelming preference (d.e. b95) for the l-conﬁgured enantiomers l-50; this enables control

NH2 N

N

5 Enzyme-catalyzed Aldol Additions

234

CO2CH3

H

OH

3PO

OPO3= COOH

(AcO)3BH–

NHAc

O

=O

FruA DHAP

OH

OH

O

HO HO

AcHN

CO2H

69

68 FruA DHAP

X H

OH

OH

OH

29 HO HO

X

OH O CO2H

X P'ase

O

O

COOH

HO

OH

COOH

X = OH KDO X=H 4-deoxy-KDO

71

70 Fig. 5.35

Synthetic approaches to DAHP and KDO by a ‘‘backbone inversion’’ strategy using FruA catalysis.

HO

HO

OH

O

O

DHAP, FruA

OH

=O

72

H

OH

OH

O

OH

O

OH

OH

3PO

73 OH

OH

O OPO3=

HO

GPO

=O

OPO3=

HO

3PO

OH

74 O2 HO

OCH3 OH

GalO

HO

H 2O 2

O

RhuA HO

OH O

HO

75

Cat

41

OH O OCH3

HO

CHO O

OH OCH3

HO OH

OH HO

=O

HO

O

76 Fig. 5.36

Sialyl Lewisx -related selectin inhibitor by chain extension of C-glycosyl aldehydes using enzymatic aldolization, and multienzymatic oxidation–aldolization strategy for synthesis of bicyclic higher-carbon sugars.

OH

O

3PO

OCH3

OH

5.4 Dihydroxyacetone Phosphate Aldolases

OH

OH

235

OH

O

OPO3=

S

H 3C

H 3C OH

O

H3C

OH H

77

OPO3=

OH

+ DHAP

OH

O

H

X 3C

80

78

HO

X = H, F

OPO3=

HO

1. FBP, FruA, TPI 2. P'ase X 3C

O OH

OH

H 3C R

OH

O

OH

FruA

H 3C

OH

O

OH OH

O

OPO3=

OH

79

HO

NaIO4 NaBH4

OH X 3C

O

OH

OH

81

Fig. 5.37

Diastereoselectivity in FruA catalyzed aldol additions to 3-hydroxyaldehydes under thermodynamic control, and synthesis of lfucose derivatives based on thermodynamic preference.

of three contiguous centers of chirality in the products (51 and 52; Figure 5.27) [170, 171]. This feature has been exploited in highly stereoselective syntheses of several rare ketose phosphates and of naturally occurring or related non-natural l sugars. For example, the kinetic enantioselectivity of FucA was not impaired by steric bulk or degree of unsaturation in the aldehyde substrates (de b90% for 50) in syntheses of higher homologs and unsaturated analogs of l-fucose 84 (see Figure 5.39) [193]. For comparison, the kinetic enantioselectivity of class I aldolases is limited to aldehydes carrying an anionic charge, and distinction of uncharged substrates is insigniﬁcant (Figure 5.23) [149, 155]. Under fully equilibrating conditions, however, diastereoselectivity of aldolase reactions can be steered by thermodynamic control to favor the energetically most stable product [149, 174, 205, 206]. Particularly strong discrimination results from utilization of 3-hydroxylated aldehydes such as 77, because of the cyclization of products to form pyranoid rings in water (Figure 5.37). The pronounced conformational destabilization by diaxial repulsion (79) strongly supports those diastereoisomers with a maximum of equatorial substituents [204, 206]. Thus, in FruA-catalyzed reactions (3S)conﬁgured hydroxyaldehydes are the preferred substrates giving the most stable all-equatorial substitution in the product (for example 78) with a de of up to 95%. Similarly, 2-alkylated aldehydes can be resolved, because of the high steric preference of an alkyl group for an equatorial position [205].

X = H, F

236

O

FruA DHAP

OR

H

5 Enzyme-catalyzed Aldol Additions

OR

O

OH

HO

P'ase

reduction

OR

deprotection

OR

OH

OH

82

OH

O

HO

H OH 2-deoxyaldose

O =O

O

OH

1. NaBH4

54

O

H

3PS

O

FruA

O =O

OH

3PS

2. H3O+

O O

OH

83

3. Ni-H2 4. HCl

H3C HO HO

O OH

D-olivose

Fig. 5.38

‘‘Inverted’’ approach for aldose synthesis using FruA catalysis, and application of the strategy for deoxysugar synthesis based on a phosphorothioate analog.

Under conditions of thermodynamic control the enantio-complementary nature of the FruA–RhuA biocatalyst pair enables construction of mirror image products 78 and ent-78 from racemic 3-hydroxybutanal 77 with similar selectivity, but preference for opposite enantiomers [25]. The allequatorial substitution in the predominant product can facilitate its separation by crystallization so that the remaining mixture can be re-subjected to further equilibration to maximize the yield of the preferred isomer 78 [177]. This general technique has recently found an application in a novel approach for the de novo synthesis of 4,6-dideoxy sugars such as 4-deoxy-lfucose 81 or its triﬂuoromethylated analog (Figure 5.37) [25]. Conversely, FucA will produce diastereomers at low de, because of the more balanced stability relationships for pyranoid products sharing a cis-diol substitution pattern. With racemic 2-hydroxylated aldehydes, thermodynamic control in FruA- or RhuA-catalyzed reactions favors (2R) conﬁgured enantiomers but also with lower discrimination (de up to 55%), because of the higher ﬂexibility of the corresponding ketofuranose rings [25, 206]. Because of the structure of the DHAP nucleophile, the enzymatic aldolization technique is ideal for direct synthesis of ketose monosaccharides and related derivatives or analogs (Figure 5.32). For an entry to aldoses an ‘‘inversion strategy’’ has been developed (Figure 5.38) which utilizes monoprotected dialdehydes (for example, 82 or 83) for aldolization and, after stereoselective ketone reduction, provides free aldoses on deprotection of the masked aldehyde function [209]. In this respect, the phosphorothioate analog of DHAP 54 makes terminally deoxygenated sugars accessible via a sequence of FruA-catalyzed aldolization followed by reductive desulfurization, as illustrated by the preparation of d-olivose along this ‘‘inversion strategy’’

5.4 Dihydroxyacetone Phosphate Aldolases

OH O R

50

1. FucA, DHAP 2. P'ase

OH

O R HO

FucI R

OH OH

O

OH OH

HO OH

237

84

R = CH3, C2H5, CH=CH2, C≡CH Fig. 5.39

Short enzymatic synthesis of l-fucose and hydrophobic analogs by aldolization–ketol isomerization, including kinetic resolution of racemic hydroxyaldehyde precursors.

[219]. Otherwise, deoxy sugars are usually only obtained when the deoxy functionality is introduced with the aldehyde. More general access to biologically important and structurally more diverse aldose isomers makes use of the enzymatic interconversion of ketoses and aldoses that in Nature is catalyzed by ketol isomerases. For full realization of the concept of enzymatic stereodivergent carbohydrate synthesis the stereochemically complementary l-rhamnose (RhaI; EC 5.3.1.14) and lfucose (FucI; EC 5.3.1.3) isomerases from E. coli have recently been shown to have broad substrate tolerance [25, 170, 193, 210]. Both enzymes convert sugars and their derivatives that have a common (3R)-OH conﬁguration but might deviate in stereochemistry or substitution pattern at subsequent positions of the chain [25, 26]. Because ketose products from RhuA- and FucA-catalyzed aldol reactions share the (3R) speciﬁcity, they can both be converted by the isomerases to corresponding aldose isomers; this enables access to a broad segment of aldose conﬁgurational space in a stereospeciﬁc, building block manner [26, 211]. This strategy has been illustrated by the synthesis of several l-conﬁgured aldohexoses using diﬀerent enzyme combinations, and by tandem FucA–FucI catalysis in the synthesis of new lfucose analogs 84 having tails with increased hydrophobicity and reactivity (Figure 5.39), starting from simple higher homologs of lactaldehyde and unsaturated analogs 50 [193, 210]. Similar results have been achieved by using a glucose isomerase (GlcI; EC 5.3.1.5) which is an industrially important enzyme for isomerization of d-glucose to d-fructose (Figure 5.40) but has a more narrow speciﬁcity. This enzyme also accepts derivatives and analogs of d-fructose and has been used in combined enzymatic syntheses, particularly of 6-modiﬁed d-glucose derivatives [207]. Further structural diversiﬁcation of FruA products has been investigated by enzymatic reduction to corresponding alditols using stereochemically complementary alditol dehydrogenases (Figure 5.40). Indeed, stereospeciﬁc (2S)- and (2R)-speciﬁc reduction of simple derivatives of d-fructose could be achieved by NADH-dependent catalysis of sorbitol dehydrogenase (EC 1.1.1.14) [209] or mannitol dehydrogenase (EC 1.1.1.67), respectively [26].

5 Enzyme-catalyzed Aldol Additions

238

OH

OH

O

OH

OH

OH

GlcI

OH

CHO OH

OH

OH

D-fructose

D-glucose

SorD OH

OH

OH

MtlD

OH

OH

OH

OH

OH OH

OH

OH

OH

sorbitol

OH

mannitol

Fig. 5.40

Stereospeciﬁc diversiﬁcation of aldol products by ketol isomerization or carbonyl reduction.

1. FruA, DHAP 2. P'ase

OH O2N

AcO O 2N

O HO

OH

CHO

AcO

steps

1. FruA, DHAP CHO

AcO

OPO= OH

O 2. P'ase

H3C

85

OH (Et2O)2OP CN

CN

NO2

OH

HO

(Et2O)2OP

OH

OH

OH OR

OH

HO

OR RO

86

87

CN

Fig. 5.41

Preparation of cyclitols by chemoenzymatic tandem reactions.

Several cyclitols (e.g. 85–87) have been prepared from aldol products carrying suitably positioned halogen, nitro, or phosphonate functionality by subsequent radical or nucleophilic cyclization reactions (Figure 5.41) [212– 214]. A cyclitol product was also found to be correctly conﬁgured to serve as a precursor to the spirocyclic Streptomyces metabolite sphydrofuran 63 [215].

5.4 Dihydroxyacetone Phosphate Aldolases

O HO FruA O

O

+

HO

H

HO

HO

PO3=

PO3=

O OH OH

PO3=

56

OH

88

HO

RhuA

FucA

OH OH

HO

(EtO)2P

OH

( )n

1. FruA/TPI FBP

H O

2. P'ase

90

O (EtO)2P

OH

( )n

O

89 PO3=

O

O

239

OH

OH

HO

91 n = 1,2

Fig. 5.42

Stereoselective synthesis of hydrolytically stable sugar phosphonates either from the bio-isosteric phosphonate analog of DHAP or from phosphonylated aldehydes.

Phosphonate analogs to phosphate esters, in which the PaO bond is formally replaced by a PaC bond, have attracted attention because of their stability toward the hydrolytic action of phosphatases, which renders them potential inhibitors or regulators of metabolic processes. Introduction of the phosphonate moiety by enzyme catalysis might, in fact, be achieved by two alternative pathways. The ﬁrst employs the bioisosteric methylene phosphonate analog 56, which gives products related to sugar 1-phosphates such as 88 and 89 (Figure 5.42) [177, 180]. This strategy is rather eﬀective because of the inherent stability of 56 as a replacement for 41 but depends on the individual tolerance of the aldolase for structural modiﬁcation close to the reactive center. The second option is a suitable choice for a phosphonylated aldehyde such as 90, which gives rise to analogs of sugar o-phosphates 91 [217, 218]. When placing thiol substitution in the aldehyde component for aldolase catalysis, the reaction products can cyclize to form rather stable cyclic hemiacetal structures. Such thiosugars are a structural variation of carbohydrates that have interesting biological properties, for example glycosidase inhibition. From 2- and 3-thiolated aldehydes stereochemical sets of furanoid or pyranoid thiosugars such as 92–94 have been prepared using different DHAP aldolases (Figure 5.43) [148, 220, 221]. It is worth noting that the observed unbiased stereoselectivity indicates the full equivalence of OH and SH substituents for correct substrate recognition. The structural resemblance of ‘‘azasugars’’ (1-deoxy sugars in which an

240

S HO R

HO

5 Enzyme-catalyzed Aldol Additions

1. FruA, DHAP 2. P'ase

OH OH

1. RhuA, DHAP 2. P'ase

HS

HO HO

CHO

S

S OH R

ent-92

92

S

OH

OH HS

CHO

1. FruA, DHAP 2. P'ase

1. FucA, DHAP 2. P'ase OH

OH S

HO

OH OH

S

HO HO

HO

93

OH

1. RhuA, DHAP 2. P'ase HO HO

S HO

OH

94

OH

OH

ent-93

Fig. 5.43

Stereodivergent enzymatic synthesis of thiosugars.

imino group replaces the ring oxygen) to transition states or intermediates of glyco-processing enzymes has made these compounds an attractive subject of research, because of their potential value as enzyme inhibitors for therapeutic applications. An important and ﬂexible synthetic strategy has been developed which consists in stereoselective enzymatic aldol addition to an azido aldehyde followed by azide hydrogenation with intramolecular reductive amination [222, 223]. Particularly noteworthy are the stereodivergent chemoenzymatic syntheses of diastereomers of the nojirimycin type from 3-azidoglyceraldehyde 95 that have been developed independently by several groups (Figure 5.44) [151, 196, 224–227]. Because of the low kinetic selectivity of FruA for 2-hydroxyaldehydes, use of enantiomerically pure aldehyde proved superior to the racemate for preparation of the parent 1-deoxy-d-nojirimycin. An extensive array of further 5-, 6- and 7-membered ring alkaloid analogs have since been made by following the same general strategy. For structural variation, as exempliﬁed by 96–98, diﬀerently substituted azido aldehydes of suitable chain length were converted by the distinct DHAP aldolases (Figure 5.45) [151, 156, 225, 228–230]. Stereocontrol during the reductive cyclization seems to be best eﬀected by Pd-catalyzed hydrogenation. The technique has been extended to the bifunctional class of azasugar phosphonic acids, for example 99, by exploiting the tolerance of rabbit FruA for the bioisosteric phosphonate nucleophile 56 [231]. The resulting heterocycles are a minimum structural motif of glycosyltransferase transition-state analogs. In a strategy inverse of that employed for compound 99, FucA

5.4 Dihydroxyacetone Phosphate Aldolases

HO

OH NH

HO HO

HO 1-deoxynojirimycin HO HO HN

OH OH

FruA

RhuA

a

a

RhuA

FruA

a

a

CHO

OHC

HO HO

HO HO HN

OH

OH NH

OH OH

FucA a

HN

OH OH

HO

HO HO

OH OH NH

S

R

HO

241

HO

HO N3

N3

D-95

L-95

TagA a

TagA

FucA

a

a

OH

HN OH OH

HO HO

Fig. 5.44

Stereodivergent synthesis of 1-deoxy azasugars of the nojirimycin type by twostep enzymatic aldolization/catalytic reductive amination (a: DHAP; P’ase; H2 /Pd–C).

and FruA were employed in the chemoenzymic synthesis of six-membered iminocyclitol phosphonic acids [232]. Another illustrative example of the azasugar synthetic strategy is the chemoenzymatic synthesis of the naturally occurring australine, 3-epiaustraline, and 7-epialexin (Figure 5.46) [233]. The bicyclic pyrrolizidine core structure resulted from twofold reductive amination of a linear precursor 101 in which the asymmetric hydroxylation sites had been installed during an aldolase-catalyzed chain extension from aminoaldehyde 100. A bidirectional aldolization approach furnished the Cglycosidically linked azadisaccharide 102 as an example of a disaccharide mimic. Ozonolysis of a racemic azido-substituted cyclohexenediol precursor was followed by tandem DHAP addition to both aldehydic termini to yield an intermediate azido-substituted dipyranoid 2,11-diulose which, when hydrogenated over Pd catalyst, gave the aza-C-disaccharide 102 highly selectively as a single diastereomer [25]. Such aldolase-catalyzed bi-directional chain elongation (‘‘tandem’’ aldolization) of simple, readily available dialdehydes has been developed into an eﬃcient method for generation of higher-carbon sugars (for example 104, 106, 108) by simple one-pot operations (Figure 5.47) [204, 234]. The furanoid (104) or pyranoid (106) nature of the products can be determined by use of a suitable hydroxyl substitution pattern in a corresponding cyclo-

OH OH NH

5 Enzyme-catalyzed Aldol Additions

242

HO

H N S HO HO

2. P'ase 3. H2/ Pd-C

96

N3

O

1. FucA, DHAP

O S

HO

OH

95

1. FruA, DHAP

OH

H 2. P'ase 3. H2/ Pd-C

N3

1. FruA, DHAP N3 2. P'ase 3. GlcI HO HO

O

Pd/C

PO3

PO3= N3

3P

OH

H2 Pd-C

98

H2 N+

O

56 FruA

H

OH

OH N H

=O =

OH

HO

OH OH

O

97

H2

O N3

OH

HO

H

HO

H HO S N HO HO

OH HO

99

Fig. 5.45

Stereoselective synthesis of 5- and 7membered ring azasugars and of novel azasugar phosphonates.

oleﬁnic precursor (e.g. 103 compared with 105). The overall speciﬁc substitution pattern in the carbon-linked disaccharide mimetics is deliberately addressable by the relative hydroxyl conﬁguration and choice of the aldolase. Single diastereomers can be obtained in good overall yield from racemic precursors if the tandem aldolizations are conducted under thermodynamic control (Figure 5.37). The thermodynamic advantage much favors a twofold trans (104) or equatorial connectivity (106), so that a C2 -symmetrical diastereomer 108 is obtained selectively, even from linear rac/meso diol mixtures 107. Similarly, highly complex structures like anulated (109) and spirocyclic (110) carbohydrate mimics can be obtained from appropriately customized precursors (Figure 5.48). In suitable circumstances use of two aldolases that aﬀord distinct selectivity can eﬀect kinetic regioselective discrimination of the independent addition steps, and thus even enable a stereochemical terminus diﬀerentiation [204]. DHAP aldolases typically yield carbohydrates or carbohydrate-derived materials, because of the nature of the reactive components, but can also be advantageously used in the construction of stereochemically homogenous fragments of non-carbohydrate natural products. An impressive illustration is the FruA-based chemoenzymatic syntheses of (þ)-exo-brevicomin (111), the aggregation pheromone of the Western pine bark beetle Dendroctonus

5.4 Dihydroxyacetone Phosphate Aldolases

OH

1. FruA, DHAP 2. P'ase

O

OH

OH

OH

O O

OH

H OHCHN

OHCHN

100

OH

OHCHN

101

HO

OH

australine HO 1. O3 2. FruA, DHAP 3. P'ase

OH

H N HO

HO

4. H2/ Pd-C (±)

HO

OH OH

HO

102

OH

N

3-epiaustraline HO

OH O

N3

OH

HO OH

N

OH

OH

H2 / Pd-C

NaCNBH3

HO

243

HO

Fig. 5.46

Synthetic route to oxygenated pyrrolizidine alkaloids, and an aza-C-disaccharide as glycosidase inhibitors.

brevicomis [235]. Addition of DHAP to 5-oxohexanal generated an enantiomerically pure vicinal syn-diol structure containing the only independent stereogenic centers of brevicomin. A complementary inverse approach to 111 (convergent at an intermediate from FruA/propanal) has also been followed using transketolase [236]. Application of an aldolase to the synthesis of the tricyclic microbial elicitor ()-syringolide (Figure 5.50) is another excellent illustration that enzyme-catalyzed aldolizations can be used to generate suﬃcient quantities of enantiomerically pure material in multi-step syntheses of complex natural and non-natural products [237]. Remarkably, the aldolase reaction established the absolute and relative conﬁguration of the only chiral centers that needed to be externally induced in 113; during the subsequent cyclization all others seemed to follow by kinetic preference. A FruA-mediated stereoselective DHAP addition to a suitable aldehyde precursor 114 (Figure 5.51) served as the key step in the synthesis of the ‘‘non-carbohydrate’’, skipped polyol C9aC16 chain fragment 115 of the macrolide antibiotic pentamycin [238, 239]. Using the same enzyme, compound 117 has been stereoselectively prepared as a synthetic equivalent to the C3aC9 fragment of (þ)-aspicillin, a lichen macrolactone (Figure 5.51) [240].

OH

5 Enzyme-catalyzed Aldol Additions

244

HO

OH (±)

HO

1. O3 2. FruA, DHAP 3. P'ase

O R HO

HO HO

103

HO

OH

104 1. O3 2. FruA, DHAP 3. P'ase

O

HO HO

105

HO

OH CH2

OH R

OH O

1. O3 2. FruA, DHAP 3. P'ase

OH

R

HO

106 OH

OH O

HO HO

n

(±)

OH HO

OH

OH (±)

OH

O HO

R

OH

107

CH2 n

R

OH

O

OH

R

OH

108

Fig. 5.47

Applications of bidirectional chain synthesis to the generation of disaccharide mimetics using tandem enzymatic aldol additions, including racemate resolution under thermodynamic control.

OH OH OHC

(±)

CHO

1. FruA, DHAP 2. P'ase

HO HO HO

OH

O O

109

OH OH OH

OH

single diastereomer

OH

HO

OH

1. O3 2. FruA, DHAP 3. P'ase

HO HO HO

O

O

110

OH OH

Fig. 5.48

Bidirectional chain extension for synthesis of anulated and spirocyclic oligosaccharide mimetics by DHAP aldolase catalysis.

OH OH

5.4 Dihydroxyacetone Phosphate Aldolases

OH

O

O

H

1. P'ase

OPO3=

FruA

O

O

O

2. H+

OH O

OH

+ DHAP

245

O O O

111 1. FruA, DHAP 2. P'ase

O H

O

acetone ZnI2

OH

HO

O HO

OH

O

O

Fig. 5.49

Complementary, inverse approaches for the FruA-based chemoenzymatic synthesis of the insect pheromone (þ)-exo-brevicomin.

O 1. DHAP, FruA H2O/DMF

O MPMO

H

112

CH3(CH2)6

O HO

2. P'ase HO

O

OH

4 steps

O

O

OMPM

OH

O

113

H+ 55% CH3(CH2)6

OH

O

O O

(-)-syringolide HO Fig. 5.50

Aldolase-based creation of independent centers of chirality in a short total synthesis of ()-syringolide, a structurally complex microbial elicitor.

O

246 OBn

MeO

O

HO

9

11 13

HO 15

O

BnO

116

OH

OH

O

115

HO 15 H3C

CH3 OH

pentamycin

OH OH

BnO

O

16

OH

1. FruA DHAP 2. P'ase

OH

OH

14

OH

H

OH

9

13

HO

OMe 16

O

HO

11

C5H11 HO 14

114

OBn

MeO

1. FruA DHAP 2. P'ase

OMe H

5 Enzyme-catalyzed Aldol Additions

O

OH

9

117

OH

O O

3

OH

(+)-aspicillin

Fig. 5.51

Stereoselective generation of synthetic precursors to the macrolide antibiotic pentamycin and the lichen macrolactone (þ)-aspicillin using FruA catalysis.

5.4.6

Aldol Transfer Enzymes

Transaldolase (EC 2.2.1.2) is an enzyme involved in the pentose phosphate pathway where it transfers a dihydroxyacetone unit between several phosphorylated metabolites [29]. The lyase belongs to class I aldolases and has been puriﬁed from several sources, cloned, and structurally characterized [41, 42]. In preparative studies yeast transaldolase, which is commercially available, has been shown to accept unphosphorylated aldehydes as the acceptor component [241–243]. The enzyme has also been employed in a multi-enzymatic scheme for conversion of starch into d-fructose; in this scheme transaldolase was used to formally dephosphorylate fructose 6-phosphate by transferring a dihydroxyacetone moiety to d-glyceraldehyde with formation of d-fructose and 34 [244]. The fructose 6-phosphate aldolase (FSA) from E. coli is a novel class I aldolase that catalyzes the reversible formation of fructose 6-phosphate from dihydroxyacetone and d-glyceraldehyde 3-phosphate; it is, therefore, functionally related to transaldolases [245]. Recent determination of the crystal structure of the enzyme showed that it also shares the mechanistic machinery [246]. The enzyme has been shown to accept several aldehydes as acceptor components for preparative synthesis. In addition to dihydroxyacetone it also utilizes hydroxyacetone as an alternative donor to generate 1-deoxysugars, for example 118, regioselectively (Figure 5.52) [247].

5.5 Transketolase and Related Enzymes

OH =

O

O OH

O3PO OH

=

OH

O HO

H

OH

OH OH

D-glyceraldehyde

O

O

HO

OH

OH

D-fructose

O H

34

transaldolase

O

HO

H OH

D-fructose 6-phosphate

HO

O3PO

OH

OH

FSA

+

O

HO OH

OH 1-deoxy-D-xylulose

118

Fig. 5.52

Transaldolase catalysis used for an apparent ‘‘transphosphorylation’’ strategy, as incorporated in a synthesis of fructose from starch, and deoxysugar synthesis by related FSA using a non-natural donor.

5.5

Transketolase and Related Enzymes

Transketolase (EC 2.2.1.1) is involved in the oxidative pentose phosphate pathway in which it catalyzes the reversible transfer of a hydroxyacetyl nucleophile between a variety of sugar phosphates. The enzyme, which requires thiamine diphosphate and divalent Mg as cofactors [248], is commercially available from baker’s yeast and can be readily isolated from many natural or recombinant sources [249, 250]. The yeast enzyme has been structurally well characterized [251], including protein with a carbanion intermediate covalently bound to the cofactor [252]. Large-scale enzyme production has been investigated for the transketolase from Escherichia coli [253–255]. Immobilization was shown to signiﬁcantly increase stability against inactivation by aldehyde substrates [256]. The enzyme is quite tolerant to organic cosolvent, and preparative reactions have been performed continuously in a membrane reactor [255], with potential in-situ product removal via borate complexation [257]. Enzymes from yeast, spinach, and Escherichia coli have been shown to tolerate a broad substrate spectrum, with the newly formed chiral center always having an absolute (S) conﬁguration as a result of re-face attack [258].

247

5 Enzyme-catalyzed Aldol Additions

248

Tab. 5.6

Substrate tolerance of transketolase. OH

OH

R

cosubstrate

O

R

+

H

R O

OH

Rel. Rate [%]

Yield [%]

Ref.

H CH2 OH CH2 OPO3 ¼ CH2 OCH3 CH2 OCH2 C6 H5 CH2 F CH2 N3 CH2 CN CH2 SH CH2 SCH3 CH2 SCH2 CH3 CH2 CH2 OH CHOHaCH3 CH3 CH2 CH3 CH2 CH2 CH3 C(CH3 )3 CHbCH2 CH2 CHbCH2 (S)-CHOHaCHbCH2 (R)-CHOHaCHbCH2

100 37 – 27 – 47 – – – 33 – <10 35 44 33 22 11 56 28 36 32

70 57 82 76 79 79 71 82 78 – 74 14 50 88 90 78 – 60 90 60 63

236, 262, 302 302 264 260, 261 260 260 196 260 220, 260 261 260 261 303 260–262 260, 261 261 261 261 261 261 261

OH

O H

OH

R

OH R

OH

O

transketolase

H

+

OH

HO2C

O

transketolase RR CO2

D,L-50

119

O

S

OH OH

+

H

R S

OH

O

120

L-50

Fig. 5.53

Kinetic resolution by transketolase, and non-equilibrium CaC bond formation by decomposition of hydroxypyruvate.

Although generic aldehydes are converted with full stereocontrol, and even a,b-unsaturated aldehydes are acceptable to some extent, hydroxylated acceptors are usually converted at higher rates (Table 5.6) [259]. In addition, antipodes of racemic 2-hydroxyaldehydes 50 with a (2R)hydroxyl group are discriminated with complete enantioselectivity; this enables eﬃcient kinetic resolution (Figure 5.53) [260, 261]. Vicinal diols of (3S,4R) conﬁguration are thereby generated with high stereocontrol. This

5.5 Transketolase and Related Enzymes

249

O =

=

O HO

O3PO HO

OH

O3PO

41

FruA OPO3=

OH TPI

42

OH =

TK

O3PO

34

OH =

O

OH

121 CO2

O –

O2C

O

O3PO

OH

119

Fig. 5.54

Multi-enzymatic scheme for stereoselective synthesis of two equivalents of xylulose 5phosphate from fructose 1,6-bisphosphate.

two-carbon chain-elongation method thus provides d-threo products 120 equivalent to those created by the respective three-carbon elongation reaction using FruA (or transaldolase) catalysis. For synthetic purposes hydroxypyruvate 119 can eﬀectively replace the natural donor components [258]. Its covalent activation occurs at a reduced rate of about 4% relative to xylulose 5-phosphate (121) but is accompanied by spontaneous decarboxylation [262]. Thus, loss of carbon dioxide renders synthetic reactions irreversible whereas alternative donors, for example lerythrulose, require coupling to cofactor recycling to shift the overall equilibrium [263]. The thermodynamic driving force from decarboxylation of 119 is particularly useful with equilibrating multi-enzyme systems such as that used in the gram-scale synthesis of two equivalents of 121 from 42 (Figure 5.54) [264]. Transketolase has been used for the key steps in chemoenzymatic syntheses of (þ)-exo-brevicomin 111 from racemic 2-hydroxybutyraldehyde [236], and of the azasugars 1,4-dideoxy-1,4-imino-d-arabinitol [196] or Nhydroxypyrrolidine 124 [265] from 3-azido (95) and 3-O-benzyl (122) derivatives, respectively, of glyceraldehyde (Figure 5.55). Such syntheses were all conducted with intrinsic racemate resolution of 2-hydroxyaldehydes and proﬁted from utilization of 119. Further preparative applications include the synthesis of valuable ketose sugars, particularly fructose analogs [258]. Transketolase has also been used for in-situ generation of erythrose 4phosphate from fructose 6-phosphate in a multi-enzymatic synthesis of DAHP (26; Figure 5.17) [131]. 1-Deoxy-d-xylulose 5-phosphate (125) is an intermediate in the recently discovered non-mevalonate pathway of terpene biosynthesis. This carbohy-

OH

5 Enzyme-catalyzed Aldol Additions

250

OH BnO

122

OH

transketolase H

O OH

BnO

O

123 119

steps

HO

OH OH

OH

N

124

OH

CO2

Fig. 5.55

Synthesis of a novel N-hydroxypyrrolidine on the basis of transketolase catalysis.

OH H2O3PO

DHAP

OH

DXS

CHO

34

OH

125

O CO2H

D-glucose

O

H2O3PO

CO2

NH2 O OH HO O P P O O O

HO HO

O

N

OH

126

HO

O

OH

Fig. 5.56

Preparation of intermediates in the non-mevalonate pathway of terpene biosynthesis.

drate derivative can be eﬃciently produced from pyruvate and 34 by the catalytic action of the thiamine diphosphate-dependent 1-deoxy-d-xylulose 5phosphate synthase (DXS) that has been cloned from E. coli [266]. The recombinant enzyme from Bacillus subtilis also has recently been applied in multi-enzymatic syntheses of diﬀerent 13 C- and 14 C-labeled isotopomers of 125 (Figure 5.56) and of labeled 4-diphosphocytidyl-2C-methyl-d-erythritol 126, a metabolic intermediate further downstream of this pathway [267, 268].

5.6

2-Deoxy-D-ribose 5-Phosphate Aldolase

The 2-deoxy-d-ribose 5-phosphate aldolase (RibA or ‘‘DERA’’; EC 4.1.2.4) is a class I enzyme that, in vivo, catalyzes the reversible addition of acetaldehyde to d-glyceraldehyde 3-phosphate (34; Figure 5.57) in the metabolic degradation of 127 from deoxyribonucleosides [269], with an equilibrium constant for synthesis of 2 104 m [56]. It is, therefore, unique among the aldolases in that it uses an aldehyde rather than a ketone as the aldol donor. RibA has been isolated from eukaryotic and prokaryotic sources [270, 271],

5.6 2-Deoxy-D-ribose 5-Phosphate Aldolase

O

OH =

O3PO

CHO

+

H 3C

OH

RibA

H 3C

H OH

O H

+

H3C

H 3C H3C

OH

RibA CH3

CH3

H 3C O H

+

H 3C

128 OH

RibA F

127

O

H3C

H3C

O

O

H2O3PO

H

34 O

251

O

H3C H3C

F

129

Fig. 5.57

Natural aldol reaction catalyzed by RibA, and acceptance of non-natural aldol donors.

and the enzyme from E. coli has been cloned [272] and overexpressed [273]. Spatial enzyme structures at high resolution have been determined for RibA from E. coli, including covalently bound intermediates [274], and from the hyperthermophilic archaea Aeropyrum pernix [275]. Interestingly, the enzyme’s relaxed acceptor speciﬁcity enables substitution of both cosubstrates, albeit at strongly reduced rates (< 1% of vmax ). Propionaldehyde, acetone, or ﬂuoroacetone can replace acetaldehyde as the donor in the synthesis of variously substituted 3-hydroxyketones for example 128 or 129 (Figure 5.28) [273, 276]. It is worthy of note that from reactions with propionaldehyde as the donor (which leads to formation of a second stereocenter) only a single diastereomer of absolute (2R,3S) conﬁguration results (for example 131; Figure 5.59) [276]. This is indicative not only of the high level of asymmetric induction at the acceptor carbonyl but also of stereospeciﬁc deprotonation of the donor. Aldehydes up to a chain length of four non-hydrogen atoms are tolerated as acceptors (Table 5.7). 2Hydroxyaldehydes 50 are relatively good acceptors, and the d isomers are preferred over the l isomers [276]. Reactions that lead to thermodynamically unfavorable structures can proceed with low stereoselectivity at the reaction center [277]. A single-point mutant aldolase was recently found to be 2.5 times more eﬀective than the wild type in accepting unphosphorylated glyceraldehyde [278]. E. coli RibA has been used in a multi-enzymatic commercial process for production of diﬀerent purine- or pyrimidine-containing deoxyribonucleosides, for example 130, in good yield (Figure 5.58) [279]. Similarly, 2-deoxyd-ribose, 13 C-labeled at diﬀerent positions, has been prepared from labeled acetaldehyde and 41 to serve as the precursor to isotopically labeled thymidine [280].

5 Enzyme-catalyzed Aldol Additions

252

Tab. 5.7

Substrate tolerance of deoxy-d-ribose-5-phosphate aldolase. O R

O

OH

RibA R

H

H

ethanal

OH

OH

R

Yield [%]

Rel. Rate [%]

Ref.

CH2 OPO3 ¼ H CH2 OH CH3 CH2 F CH2 Cl CH2 Br CH2 SH CH2 N3 C2 H5 CHbCH2 CHOHaCH2 OH CHN3 aCH2 OH CHOHaCH3 CHOHaCH2 aC6 H5 CH2 SCH2 aCHOHaCH2 OH

78 20 65 32 33 37 30 33 76 18 12 62 46 51 46 27

100 – 0.4 0.4 0.4 0.3 – – 0.3 0.3 – 0.3 – – – –

276 301 276 276 276 276 283 283 276 276 283 276 283 283 283 283

stage 1

DHAP FruA FBP

TPI O O H2O3PO

OH

H H

RibA

stage 2

O

H2O3PO

H OH

OH

34

127 N

adenine

PPM

O

HO

OPO3H2

HO

Pi

N

N

PNP HO

NH2

O

HO

Fig. 5.58

Two-stage aldolase-based technical process for deoxyribonucleoside synthesis.

N

130

5.6 2-Deoxy-D-ribose 5-Phosphate Aldolase

O

O +

H

N3

D-95

N3

O

OH

R

H 3C

H HO

OH

CH3

131

O

O H

D-95

N3

RibA

253

+

N3

RibA

O

OH

NH

HO

H2/ Pd-C

H

OH

HO

HO

Fig. 5.59

Azasugar precursors prepared by RibA catalysis.

O H 3C

OH

RibA H

H 3C

O

RibA

H3C R

H

O H 3C

OH

OH

O H

R

O H

H 3C

H

H3C

O OH

OH

133

Fig. 5.60

RibA-catalyzed sequential aldol addition.

Starting from azidoaldehydes, several azasugars containing a lower density of substituents (for example 132; Figure 5.59) have been prepared by sequential aldolization–hydrogenation [276]. More recently, pyranose synthons have been prepared as key intermediates in the synthesis of epothilones by using RibA catalysis (Chapter 7 in Part I of this book) [281]. When acetaldehyde is used as the only substrate, the initial aldol product can serve again as a suitable acceptor for sequential addition of a second donor molecule to give (3R,5R)-2,4,6-trideoxyhexose 133 (Figure 5.60) [282, 283]. Cyclization to stable hemiacetals masks the free aldehyde and thus eﬀectively precludes formation of higher-order adducts. When the ﬁrst acceptor is an a-substituted acetaldehyde, related aldol products from twofold donor additions can be prepared that are structurally related to mevinolactone. Combination of a RibA-catalyzed initial addition to other aldolases such as FruA or NeuA in a consecutive addition reaction has also been studied for synthesis of non-natural sugars [277, 282].

132

254

5 Enzyme-catalyzed Aldol Additions

NH2

O

SHMT

+

H

CO2H

H

NH2 CO2H

S

PLP

134

NH2 HO2C

CO2H

OH

OH

135

136

NH2 O

NH2 +

H 3C

H

CO2H

134

ThrA PLP

H 3C R

S

CO2H

NH2 +

H3C S

OH

137

L-Thr

S

CO2H

OH

138

L-allo-Thr

Fig. 5.61

Aldol reactions catalyzed in vivo by serine hydroxymethyl transferase and by threonine aldolases.

5.7

Glycine Aldolases

The metabolism of b-hydroxy-a-amino acids involves pyridoxal phosphatedependent enzymes, classiﬁed as serine hydroxymethyltransferase or threonine aldolases, that catalyze reversible aldol-type cleavage to aldehydes and glycine (134) [284]. In vivo serine hydroxymethyltransferases (SHMT; EC 2.1.2.1) catalyze the formation of l-serine 135 from addition of 134 to formaldehyde (Figure 5.61) in the presence of the cofactor tetrahydrofolate. SHMT have been puriﬁed and cloned from a large variety of organisms [285]. Eukaryotic enzymes have been shown to accept a range of other aldehydes (Table 5.8) to furnish the corresponding b-hydroxy-a-amino acid adducts [286, 287]. The enzymes are highly selective for the l-conﬁguration but will usually produce diastereomeric mixtures with low erythro/threo selectivity. In preparative reactions excess 134 can compensate for the unfavorable equilibrium constant [56], and economically viable yields of 135 have thus been obtained on a multi-molar scale at high product concentrations [288, 289]. Recombinant rabbit SHMT has been used for highly stereoselective synthesis of l-erythro2-amino-3-hydroxy-1,6-hexanedicarboxylic acid 136 as a potential precursor to carbocyclic b-lactams and nucleosides [290]. Threonine aldolases catalyze the reversible addition of 134 to acetaldehyde to give threonine (137), but the distinction between l-threonine aldolase (ThrA; EC 4.1.2.5) and l-allo-threonine aldolase (EC 4.1.2.6) is vague, because many catalysts have only poor capacity for threo/erythro discrimination (137/138; Figure 5.61) [284, 291, 292]. In fact, both l- and d-selective enzymes have been puriﬁed and cloned from a variety of organisms [284]. The ﬁrst stereospeciﬁc l-allo-threonine aldolase from Aeromonas jandaei has recently been cloned and characterized [293]. X-ray structures of Thermotoga

5.7 Glycine Aldolases

255

Tab. 5.8

Substrate tolerance of threonine-dependent aldolases. NH2

NH2 O R

R

H CH3 C3 H9 C5 H11 CH(CH3 )2 CH2 OBn CH2 CH2 OBn (CH2 )2 C6 H5 C6 H5 p-C6 H4 NO2 o-C6 H4 NO2 m-C6 H4 OH p-C6 H4 CH3 2-imidazolyl 2-furanyl 2-thienyl

aldolase H

R

serine

SHMT

CO2H OH

Yield [%]

– 2:98

94 –

60:40

25

60:40

22

33:67 50:50 56:44

10 20 11

CO2H OH

L-ThrA

Selectivity (threo:erythro)

+

R

D-ThrA

Ref.

Selectivity (threo:erythro)

Yield [%]

Selectivity (threo:erythro)

Yield [%]

9:91 24:76 37:63 46:53 2:98 47:53 28:72 60:40 47:53 58:42 73:27 55:45 66:34

40 21 16 15 88 53 10 9 53 93 43 17 40

53:47 67:33 61:39 84:16

60 37 31 26

87:13 74:26 55:45 72:28 74:26 43:57 61:39

16 11 88 89 54 25 60

288, 289 292 292 287, 292 292 297 297 292 287, 292 292 292 292 292 287, 292 287 287

maritima ThrA as apoenzyme and in complex with substrate and product have recently been determined [294]. Many ThrA enzymes have broad substrate tolerance for the aldehyde acceptor, notably including variously substituted aliphatic and aromatic aldehydes (Figure 5.63); however, a,b-unsaturated aldehydes are not accepted [291]. The erythro-selective l-threonine aldolase from the yeast Candida humicola has been used to catalyze the addition of 134 to d-glyceraldehyde acetonide 139 (Figure 5.62) [295]. The diastereomeric products 140 and 141, formed in a 1:2 ratio, were separated and further processed to furnish (S,S,R)- and (S,S,S)-3,4-dihydroxyprolines 143 and 144. The same enzyme has also been applied to the synthesis of nucleobase-modiﬁed amino acids of type 142 [296], and of compound 146 as an intermediate en route to the immunosuppressive lipid mycestericin D [297]. For the latter, benzyl protection of the acceptor (145) was found to increase the stereoselectivity, and the erythro product was obtained preferentially under kinetic control at low conversion. With benzyloxyacetaldehyde a cloned ThrA from E. coli resulted in higher diastereoselectivity than other enzymes under conditions of kinetic control; this was applied in the synthesis of novel sialyl Lewis x mimetics [298].

5 Enzyme-catalyzed Aldol Additions

256

OH CHO

ThrA

O O

O O

glycine

OH CO2H

NH2

139

CO2H

O

+

O

NH2

140

141

steps

steps

OH HO

CO2H N

N

N

N

NH2

OH

N H

142

HO

CO2H

OH

N H

143

CO2H

144

NH2 O H2N

134

+ CO2H

OBn

H

ThrA

OH OBn

H2N

PLP

CO2H

145

146

OH C6H13

H 2N HOH2C CO2H

mycestericin D

O

Fig. 5.62

Application of ThrA catalysis for stereoselective synthesis of dihydroxyprolines from glyceraldehyde, of an adenylamino acid in preparation of RNA mimics, and of a potential precursor to the immunosuppressive lipid mycestericin D.

More suitably, the speciﬁcity of ThrA enzymes can be used for kinetic resolution of diastereomer mixtures such as those produced by chemical synthesis (Figure 5.63). This is particularly promising for aryl analogs of threonine that are of interest as building blocks of pharmaceuticals, including vancomycin antibiotics. Thus, an l-ThrA from Streptomyces amakusaensis has been shown to be particularly useful for resolution of racemic threo-aryl serines 147 by retroaldolization under kinetic control to furnish enantiomerically pure d amino acids [290, 292, 299]. As a complementary example, the recombinant low-speciﬁcity d-ThrA from Alcaligenes xyloxidans has been used for resolution of DL-threo-b-(3,4-methylenedioxyphenyl)serine 148 by retro-aldol cleavage to furnish the desired l-threo isomer with a molar yield of 50% and almost 100% ee [300]. The latter compound serves as a synthetic intermediate en route to a parkinsonism drug.

5.8 Recent Developments

OH

OH CO2H

NH2

X

R

L-ThrA

S

CO2H

X

134 OH

CO2H NH2

O

NH2

+

D-threo-147

OH O

+

NH2

X

DL-threo-147

CHO

CO2H

257

DL-threo-148

D-ThrA

CO2H

O

NH2

O

+

piperonal + glycine

L-threo-148

Fig. 5.63

Resolution of diastereomer mixtures by retro-aldolization under kinetic control for preparation of enantiomerically pure phenylserines, and racemate resolution process for a synthetic intermediate of an antiparkinsonism drug.

5.8

Recent Developments

As a result of a recent almost explosive increase in knowledge of the structures of all important classes of aldolases, including their liganded complexes, detailed understanding of substrate recognition and catalytic machinery for stereospeciﬁc carbon–carbon bond formation is advancing rapidly (Section 5.2.2). This, with progress fueled by whole-genome sequencing, means the enzymes are now amenable to rational engineering or directed molecular evolution toward improved physical properties and catalytic performance. Tailoring of enzyme properties such as improved activity and stability has been amply demonstrated to be feasible on demand [304, 305], and improved substrate tolerance and optimized stereoselectivity for asymmetric syntheses are immediate targets [306]. For identiﬁcation of suitable mutant aldolases from engineering or evolutionary approaches, eﬃcient assay techniques have been developed for identiﬁcation of target catalysts in high-throughput mode based on chromogenic or ﬂuorogenic substrates [307–309]. The same is true in the search for novel enzymes with interesting synthetic capabilities from mass screening of natural or manmade biodiversity [310]. Most recently, ﬁrst attempts to redesign individual aldolase enzymes for an altered substrate spectrum have been reported [143, 144, 278], including a ﬁrst example of modiﬁcation of the stereospeciﬁcity of carbon–carbon bond creation by directed evolution [311]. Together with knowledge gained from the mechanistic proﬁles of catalytic antibodies as a complementary approach toward novel stereoselective aldolase biocatalysts (Chapter 6 in Part I of this book), such endeavors will probably also be instrumental in enabling better generic understanding of the mechanis-

258

5 Enzyme-catalyzed Aldol Additions

tic subtleties operating in the initiation and stereoelectronic steering of enzyme-catalyzed aldol reactions. Consequently, it is probable that current restrictions resulting from narrow donor speciﬁcity will soon be eliminated. In addition to the major types of aldolases discussed above, several biosynthetically important enzymes promote aldol-related Claisen additions of acyl thioesters of coenzyme A (CoA) to ketones, often to a-oxoacids. This includes some of the key enzymes in central metabolism such as the citric acid cycle, and many others involved in the biosynthesis of fatty acids, steroids, terpenoids, macrolides and other secondary metabolites [14, 15]. Members of this class of CaC bond-forming enzymes are certainly attractive for asymmetric synthesis, because the equilibrium constants proﬁt from a high driving force by ensuing thioester hydrolysis. High expense and limited CoA cofactor stability have so far precluded broader synthetic evaluation, but the technique might be more practical now that several costeﬃcient in-situ cofactor recycling schemes for acyl-CoA substrates have been developed [312–316]. Particular progress is also advancing in the ﬁeld of polyketides, a class of structurally diverse natural products with a broad range of biological activity. Polyketide biosynthesis is based on a fully modular scheme that involves distinct protein domains with ketosynthase (Claisen aldolization), or functional group processing activity, which iteratively build up the carbon backbone and an individually patterned polyketide structure. By choosing from known gene clusters of diﬀerent polyketides from diﬀerent microorganisms and plants that code for speciﬁc catalytic domains, new interspecies gene constructs can be assembled in a combinatorial fashion; this enables exploration of the feasibility of engineered biosynthesis of novel, non-natural polyketides with potentially useful bioactivity [317–320].

5.9

Summary and Conclusion

Progress in the synthetic use of aldol-active enzymes as eﬃcient and highly stereoselective catalysts has been remarkable. Biocatalytic CaC bond formation is eminently useful for asymmetric synthesis of complex multifunctional molecules. Process design and development, however, requires an intimate knowledge base covering subjects such as substrate tolerance for donor and acceptor components or the inﬂuence of substrate structure on enantio- and diastereoselectivity of CaC bonding. This overview demonstrates that many enzymes for carbon–carbon bond formation are now readily available and predictably useful for synthetic applications. Clearly, such enzymes bear signiﬁcant potential for application in asymmetric synthesis, particularly with regard to polyfunctionalized natural products and other biologically relevant classes of compounds. It is also evident that the technology is now well accepted in the chemical community as a powerful

Examples of Experimental Procedures

supplement to existing methodology of organic synthesis. Indeed, with the ﬁrst processes already at the state of industrial commercialization on a large scale, enzymatic aldol reactions are maturing rapidly. The scope of potential applications is broadening on all frontiers with increasing complexity of target structures. Limitations inherent in the properties of enzymes are appreciated and approaches are developed to resolve them. Hopefully, this compilation of the potential scope of the technique and its limitations will help to identify, and aid in the successful development of, future synthetic applications.

Examples of Experimental Procedures General

Dihydroxyacetone phosphate (41) was prepared enzymatically from glycerol 1-phosphate (74) by the action of glycerol phosphate oxidase [177]. Analytical thin-layer chromatography was performed on silica gel plates using a 1:1 mixture of sat. ammonia–ethanol for development, and anisaldehyde stain for detection. The aldolases are commercially available or can be puriﬁed in accordance with published procedures [150]. Activity of aldolases (1 unit catalyzes cleavage of 1 mmol l-ketose 1-phosphates (43/45) per minute at 25 C [150]) and amounts of 41 were determined photometrically by an assay coupled with glycerol phosphate dehydrogenase-catalyzed NADH oxidation [149]. L-Fructose 1-Phosphate from L-Glyceraldehyde by in-situ Formation of DHAP. GPO (70 U), catalase (1000 U), and RhuA (50 U) was added to a solution of l-glycerol 3-phosphate (74, 1.0 mmol) and l-glyceraldehyde (110 mg, 1.2 mmol) in 10 mL oxygen-saturated water at pH 6.8. The mixture was shaken at 20 C under an oxygen atmosphere at 100 rpm. Conversion was monitored by enzymatic assay for equivalents of 41 produced, and by 1 H and 31 P NMR spectroscopy. After complete conversion and ﬁltration through charcoal the pH was adjusted to 7.5 by addition of 1.0 m cyclohexylamine in ethanol and the solution was concentrated to dryness by rotary evaporation at a20 C in vacuo. The solid residue was dissolved in 0.5 mL water and the resulting solution was ﬁltered. Dry ethanol (2.5 mL) was added, then dry acetone until faint turbidity remained. Crystallization at 4 C furnished l-fructose 1-phosphate bis(cyclohexylammonium) salt as colorless needles; yield 370 mg (85%). L-Fuculose 1-Phosphate from DL-Lactaldehyde by Racemate Resolution [171]. Racemic lactaldehyde dimethylacetal (6.0 g, 50 mmol) in 50 mL water was hydrolyzed by treatment with cation-exchange resin (Dowex AG50W-X8, Hþ form) at 60 C for 8 h. After ﬁltration, an aqueous solution

259

260

5 Enzyme-catalyzed Aldol Additions

of 41 (80 mL, 10 mmol; prepared by GPO oxidation of 74 as above) was added. The mixture was adjusted to pH 6.8 with 1 m NaOH and incubated with FucA (300 U) at room temperature. After complete conversion of 41, as determined by TLC (RF (41) ¼ 0.39, RF (45) ¼ 0.30) and enzymatic assay, the mixture was ﬁltered through charcoal and passed through an anionexchange column (Dowex AG1-X8, HCO3 form, 100 mL). The column was washed with water (200 mL), and the product was eluted with 0.2 m triethylammonium hydrogen carbonate buﬀer. Repeated concentration from water (3 50 mL), ion exchange to the free acid (Dowex AG50W-X8, Hþ form, 100 mL), neutralization with cyclohexylamine, then crystallization from 90% aqueous ethanol provided the colorless bis(cyclohexylammonium) salt of 45 (3.76 g, 85%).

References 1 H. G. Davies, R. H. Green, D. R. Kelly, S. M. Roberts,

2 3 4 5 6 7 8

9 10 11 12 13 14

15 16

Biotransformations in Preparative Organic Chemistry, Academic Press, London 1989. W. Gerhartz, Enzymes in Industry: Production and Applications, VCH, Weinheim 1991. J. Halgas, Biocatalysts in Organic Synthesis, Elsevier, Amsterdam 1992. L. Poppe, L. Nova´k, Selective Biocatalysis. A Synthetic Approach, VCH, Weinheim 1992. C.-H. Wong, G. M. Whitesides, Enzymes in Synthetic Organic Chemistry, Pergamon, Oxford 1994. M. Sinnott, Comprehensive Biological Catalysis, Academic Press, San Diego 1998. W.-D. Fessner, Biocatalysis – From Discovery to Application (Top. Curr. Chem., Vol. 200), Springer, Heidelberg 1998. U. Bornscheuer, R. Kazlauskas, Hydrolases in Organic Synthesis – Regio- and Stereoselective Biotransformations, Wiley– VCH, Weinheim 1999. K. Faber, Biotransformations in Organic Chemistry, 4th Ed., Springer, Heidelberg 2000. A. Liese, K. Seelbach, C. Wandrey, Industrial Biotransformations, Wiley–VCH, Weinheim 2000. K. Drauz, H. Waldmann, Enzyme Catalysis in Organic Synthesis, 2nd Ed., Wiley–VCH, Weinheim 2002. T. D. Machajewski, C.-H. Wong, Angew. Chem. Int. Ed. 2000, 39, 1352–1374. H. Waldmann, D. Sebastian, Chem. Rev. 1994, 94, 911–937. D. F. Henderson, E. J. Toone, in Comprehensive Natural Product Chemistry, Vol. 3 (Ed.: B. M. Pinto), Elsevier Science, Amsterdam 1999, p. 367–440. D. Schomburg, M. Salzmann, Enzyme Handbook, Springer, Berlin 1990–1995. V. Schellenberger, H. D. Jakubke, Angew. Chem. Int. Ed. Engl. 1991, 30, 1437–1449.

References 17 D. Rozzell, F. Wagner, Biocatalytic Production of Amino Acids

and Derivatives, Hanser, Munich 1992. 18 E. J. Toone, E. S. Simon, M. D. Bednarski, G. M.

Whitesides, Tetrahedron 1989, 45, 5365–5422. 19 S. David, C. Auge´, C. Gautheron, Adv. Carbohydr. Chem.

Biochem. 1991, 49, 175–237. 20 H. J. M. Gijsen, L. Qiao, W. Fitz, C.-H. Wong, Chem. Rev.

1996, 96, 443–473. 21 G. M. Watt, P. A. Lowden, S. L. Flitsch, Curr. Opin. Struct.

Biol. 1997, 7, 652–660. 22 S. Takayama, G. J. McGarvey, C.-H. Wong, Ann. Rev.

Microbiol. 1997, 51, 285–310. 23 N. Wymer, E. J. Toone, Curr Opin Chem Biol 2000, 4, 110–

119. 24 W.-D. Fessner, V. Helaine, Curr Opin. Biotechnol. 2001, 12,

574–586. 25 W.-D. Fessner, C. Walter, Top. Curr. Chem. 1996, 184, 97–

194. 26 W.-D. Fessner, in Microbial Reagents in Organic Synthesis, Vol.

27 28 29

30 31 32

33 34

35

36 37 38 39 40 41

381 (Ed.: S. Servi), Kluwer Academic Publishers, Dordrecht, 1992, pp. 43–55. R. Kluger, Chem. Rev. 1990, 90, 1151–1169. E. C. Webb, Enzyme Nomenclature, Academic Press, San Diego 1992. B. L. Horecker, O. Tsolas, C. Y. Lai, in The Enzymes, Vol. VII (Ed.: P. D. Boyer), Academic Press, New York, 1972, pp. 213–258. G. Trombetta, G. Balboni, A. di Iasio, E. Grazi, Biochem. Biophys. Res. Commun. 1977, 74, 1297–1301. D. J. Kuo, I. A. Rose, Biochemistry 1985, 24, 3947–3952. W.-D. Fessner, A. Schneider, H. Held, G. Sinerius, C. Walter, M. Hixon, J. V. Schloss, Angew. Chem. Int. Ed. Engl. 1996, 35, 2219–2221. J. Sygusch, D. Beaudry, M. Allaire, Proc. Natl. Acad. Sci. USA 1987, 84, 7846–7850. S. J. Gamblin, G. J. Davies, J. M. Grimes, R. M. Jackson, J. A. Littlechild, H. C. Watson, J. Mol. Biol. 1991, 219, 573– 576. G. Hester, O. Brenner-Holzach, F. A. Rossi, M. StruckDonatz, K. H. Winterhalter, J. D. G. Smit, K. Piontek, FEBS Lett. 1991, 292, 237–242. H. Kim, U. Certa, H. Dobeli, P. Jakob, W. G. J. Hol, Biochemistry 1998, 37, 4388–4396. A. R. Dalby, D. R. Tolan, J. A. Littlechild, Acta Cryst. 2001, D 57, 1526–1533. N. Blom, J. Sygusch, Nature Struct. Biol. 1997, 4, 36–39. K. H. Choi, J. Shi, C. E. Hopkins, D. R. Tolan, K. N. Allen, Biochemistry 2001, 40, 13868–13875. A. Maurady, A. Zdanov, D. de Moissac, D. Beaudry, J. Sygusch, J. Biol. Chem. 2002, 277, 9474–9483. J. Jia, W. J. Huang, U. Scho¨rken, H. Sahm, G. A. Sprenger, Y. Lindqvist, G. Schneider, Structure 1996, 4, 715–724.

261

262

5 Enzyme-catalyzed Aldol Additions ¨ rken, Y. Lindqvist, G. A. Sprenger, G. 42 J. Jia, U. Scho Schneider, Protein Sci. 1997, 6, 119–124. 43 T. Izard, M. C. Lawrence, R. L. Malby, G. G. Lilley, P. M.

Colman, Structure 1994, 2, 361–369. 44 M. C. Lawrence, J. A. R. G. Barbosa, B. J. Smith, N. E.

45

46 47 48 49 50 51 52

53 54 55 56 57 58 59 60

61 62 63 64 65 66 67

Hall, P. A. Pilling, H. C. Ooi, S. M. Marcuccio, J. Mol. Biol. 1997, 266, 381–399. J. A. Barbosa, B. J. Smith, R. DeGori, H. C. Ooi, S. M. Marcuccio, E. M. Campi, W. R. Jackson, R. Brossmer, M. Sommer, M. C. Lawrence, J Mol Biol 2000, 303, 405–421. J. Allard, P. Grochulski, J. Sygusch, Proc. Natl. Acad. Sci. USA 2001, 98, 3679–3684. M. K. Dreyer, G. E. Schulz, J. Mol. Biol. 1996, 259, 458–466. M. K. Dreyer, G. E. Schulz, J. Mol. Biol. 1993, 231, 549–553. M. Kroemer, G. E. Schulz, Acta Cryst. 2002, D 58, 824–832. A. C. Joerger, C. Gosse, W.-D. Fessner, G. E. Schulz, Biochemistry 2000, 39, 6033–6041. N. S. Blom, S. Tetreault, R. Coulombe, J. Sygusch, Nature Struct. Biol. 1996, 3, 856–862. S. J. Cooper, G. A. Leonard, S. M. McSweeney, A. W. Thompson, J. H. Naismith, S. Qamar, A. Plater, A. Berry, W. N. Hunter, Structure 1996, 4, 1303–1315. S. M. Zgiby, G. J. Thomson, S. Qamar, A. Berry, Eur. J. Biochem. 2000, 267, 1858–1869. D. R. Hall, C. S. Bond, G. A. Leonard, C. I. Watt, A. Berry, W. N. Hunter, J. Biol. Chem. 2002, 277, 22018–22024. D. R. Hall, L. E. Kemp, G. A. Leonard, K. Marshall, A. Berry, W. N. Hunter, Acta Cryst. 2003, D 59, 611–614. R. N. Goldberg, Y. B. Tewari, J. Phys. Chem. Ref. Data 1995, 24, 1669–1698. A. Pollak, H. Blumenfeld, M. Wax, R. L. Baughn, G. M. Whitesides, J. Am. Chem. Soc. 1980, 102, 6324–6336. B. Bossow, W. Berke, C. Wandrey, BioEngineering 1992, 8, 12–19. S. B. Sobolov, A. Bartoszkomalik, T. R. Oeschger, M. M. Montelbano, Tetrahedron Lett. 1994, 35, 7751–7754. U. Kragl, U. Niedermeyer, M.-R. Kula, C. Wandrey, in Ann. N. Y. Acad. Sci., Vol. 613 (Eds.: H. Okada, A. Tanaka, H. W. Blanch), The New York Academy of Sciences, New York, 1990, pp. 167–175. U. Kragl, D. Gygax, O. Ghisalba, C. Wandrey, Angew. Chem. Int. Ed. Engl. 1991, 30, 827–828. U. Kragl, A. Go¨dde, C. Wandrey, N. Lubin, C. Auge´, J. Chem. Soc. Perkin Trans. I 1994, 119–124. L. Espelt, P. Clapes, J. Esquena, A. Manich, C. Solans, Langmuir 2003, 19, 1337–1346. W.-D. Fessner, M. Knorst, DE 10034586, 2002; Chem. Abstr. 136:166160. Y. Ohta, M. Shimosaka, K. Murata, Y. Tsukada, A. Kimura, Appl. Microbiol. Biotechnol. 1986, 24, 386–391. K. Aisaka, S. Tamura, Y. Arai, T. Uwajima, Biotechnol. Lett. 1987, 9, 633–637. K. Yamamoto, B. Kawakami, Y. Kawamura, K. Kawai, Anal. Biochem. 1997, 246, 171–175.

References 68 M. J. Kim, W. J. Hennen, H. M. Sweers, C.-H. Wong, J. Am.

Chem. Soc. 1988, 110, 6481–6486. 69 C. Auge´, S. David, C. Gautheron, Tetrahedron Lett. 1984, 25,

4663–4664. 70 C. Auge´, S. David, C. Gautheron, A. Veyrie`res, Tetrahedron

Lett. 1985, 26, 2439–2440. 71 C. Auge´, C. Gautheron, J. Chem. Soc. Chem. Commun. 1987,

859–860. 72 C. Auge´, S. David, C. Gautheron, A. Malleron, B. Cavaye´,

New. J. Chem. 1988, 12, 733–744. 73 M. D. Bednarski, H. K. Chenault, E. S. Simon, G. M.

Whitesides, J. Am. Chem. Soc. 1987, 109, 1283–1285. 74 E. S. Simon, M. D. Bednarski, G. M. Whitesides, J. Am.

Chem. Soc. 1988, 110, 7159–7163. 75 T. Sugai, A. Kuboki, S. Hiramatsu, H. Okazaki, H. Ohta,

Bull. Chem. Soc. Jpn. 1995, 68, 3581–3589. 76 M. Mahmoudian, D. Noble, C. S. Drake, R. F. Middleton,

77 78 79 80 81 82

83 84 85

86 87 88 89 90

91 92

D. S. Montgomery, J. E. Piercey, D. Ramlakhan, M. Todd, M. J. Dawson, Enzyme Microb. Technol. 1997, 20, 393–400. S. Blayer, J. M. Woodley, M. D. Lilly, M. J. Dawson, Biotechnol. Prog. 1996, 12, 758–763. I. Maru, J. Ohnishi, Y. Ohta, Y. Tsukada, Carbohydrate Res. 1998, 306, 575–578. C. H. Lin, T. Sugai, R. L. Halcomb, Y. Ichikawa, C.-H. Wong, J. Am. Chem. Soc. 1992, 114, 10138–10145. Y. Ichikawa, J. L. C. Liu, G. J. Shen, C.-H. Wong, J. Am. Chem. Soc. 1991, 113, 6300–6302. O. Blixt, J. C. Paulson, Adv. Synth. Catal. 2003, 345, 687– 690. J. Beliczey, U. Kragl, A. Liese, C. Wandrey, K. Hamacher, H. H. Coenen, T. Tierling, US 6355453, 2002; Chem. Abstr. 136:231340. W. Fitz, J. R. Schwark, C.-H. Wong, J. Org. Chem. 1995, 60, 3663–3670. C. Auge´, B. Bouxom, B. Cavaye´, C. Gautheron, Tetrahedron Lett. 1989, 30, 2217–2220. J. L. C. Liu, G. J. Shen, Y. Ichikawa, J. F. Rutan, G. Zapata, W. F. Vann, C.-H. Wong, J. Am. Chem. Soc. 1992, 114, 3901– 3910. C. Salagnad, A. Godde, B. Ernst, U. Kragl, Biotechnol. Prog. 1997, 13, 810–813. R. J. Lins, S. L. Flitsch, N. J. Turner, E. Irving, S. A. Brown, Angew. Chem. Int. Ed. 2002, 41, 3405–3407. W.-Y. Wu, B. Jin, D. C. M. Kong, M. von Itzstein, Carbohydrate Res. 1997, 300, 171–174. A. Kuboki, H. Okazaki, T. Sugai, H. Ohta, Tetrahedron 1997, 53, 2387–2400. A. J. Humphrey, C. Fremann, P. Critchley, Y. Malykh, R. Schauer, T. D. H. Bugg, Bioorg. Med. Chem. 2002, 10, 3175– 3185. C.-C. Lin, C.-H. Lin, C.-H. Wong, Tetrahedron Lett. 1997, 38, 2649–2652. M. Murakami, K. Ikeda, K. Achiwa, Carbohydrate Res. 1996, 280, 101–110.

263

264

5 Enzyme-catalyzed Aldol Additions 93 D. C. M. Kong, M. Vonitzstein, Carbohydrate Res. 1997, 305,

323–329. 94 M. J. Kiefel, J. C. Wilson, S. Bennett, M. Gredley, M. von

Itzstein, Bioorg. Med. Chem. 2000, 8, 657–664. 95 M. A. Sparks, K. W. Williams, C. Lukacs, A. Schrell, G.

96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121

Priebe, A. Spaltenstein, G. M. Whitesides, Tetrahedron 1993, 49, 1–12. P. Z. Zhou, H. M. Salleh, J. F. Honek, J. Org. Chem. 1993, 58, 264–266. W. Fitz, C.-H. Wong, J. Org. Chem. 1994, 59, 8279–8280. G. B. Kok, M. Campbell, B. L. Mackey, M. von Itzstein, Carbohydr. Res. 2001, 332, 133–139. S. David, A. Malleron, B. Cavaye, New J. Chem 1992, 16, 751–755. C. Auge´, C. Gautheron, S. David, A. Malleron, B. Cavaye´, B. Bouxom, Tetrahedron 1990, 46, 201–214. T. Sugai, G. J. Shen, Y. Ichikawa, C.-H. Wong, J. Am. Chem. Soc. 1993, 115, 413–421. S. Ladisch, A. Hasegawa, R. X. Li, M. Kiso, Biochemistry 1995, 34, 1197–1202. K. Koppert, R. Brossmer, Tetrahedron Lett. 1992, 33, 8031– 8034. A. Malleron, S. David, New J. Chem. 1996, 20, 153–159. T. Miyazaki, H. Sato, T. Sakakibara, Y. Kajihara, J. Am. Chem. Soc. 2000, 122, 5678–5694. L. K. Mahal, K. J. Yarema, C. R. Bertozzi, Science 1997, 276, 1125–1128. K. J. Yarema, L. K. Mahal, R. E. Bruehl, E. C. Rodriguez, C. R. Bertozzi, J. Biol. Chem. 1998, 273, 31168–31179. C. R. H. Raetz, J. Bacteriol. 1993, 175, 5745–5753. F. M. Unger, Adv. Carbohydr. Chem. Biochem. 1981, 38, 323– 388. B. R. Knappmann, M. R. Kula, Appl. Microbiol. Biotechnol. 1990, 33, 324–329. M. A. Ghalambor, E. C. Heath, J. Biol. Chem. 1966, 241, 3222–3227. P. R. Srinivasan, D. B. Sprinson, J. Biol. Chem. 1959, 234, 716–722. J. Hurwitz, A. Weissbach, J. Biol. Chem. 1959, 234, 710–712. P. H. Ray, J. Bacteriol. 1980, 141, 635–644. F. Stuart, I. S. Hunter, Biochim. Biophys. Acta 1993, 1161, 209–215. J. M. Ray, R. Bauerle, J. Bacteriol. 1991, 173, 1894–1901. L. M. Weaver, J. E. B. P. Pinto, K. M. Herrmann, Bioorg. Med. Chem. Lett. 1993, 3, 1421–1428. I. A. Shumilin, C. Zhao, R. Bauerle, R. H. Kretsinger, J. Mol. Biol. 2002, 320, 1147–1156. R. L. Doong, J. E. Gander, R. J. Ganson, R. A. Jensen, Physiol. Plant. 1992, 84, 351–360. L. M. Reimer, D. L. Conley, D. L. Pompliano, J. W. Frost, J. Am. Chem. Soc. 1986, 108, 8010–8015. K. M. Draths, J. W. Frost, J. Am. Chem. Soc. 1990, 112, 1657–1659.

References 122 K. M. Draths, T. L. Ward, J. W. Frost, J. Am. Chem. Soc.

1992, 114, 9725–9726. 123 D. R. Knop, K. M. Draths, S. S. Chandran, J. L. Barker, R.

124 125 126 127 128 129 130 131 132

133 134 135

136 137 138 139 140

141

142 143 144

145

von Daeniken, W. Weber, J. W. Frost, J. Am. Chem. Soc. 2001, 123, 10173–10182. K. M. Draths, J. W. Frost, J. Am. Chem. Soc. 1995, 117, 2395–2400. J. L. Barker, J. W. Frost, Biotechnol. Bioeng. 2001, 76, 376– 390. N. Ran, D. R. Knop, K. M. Draths, J. W. Frost, J. Am. Chem. Soc. 2001, 123, 10927–10934. A. K. Sundaram, R. W. Woodard, J Org Chem 2000, 65, 5891–5897. W. A. Wood, in The Enzymes, 3rd Ed, Vol. 7 (Eds.: P. D. Boyer), Academic, New York, 1972, pp. 281–302. R. H. Hammerstedt, H. Mo¨hler, K. A. Decker, D. Ersfeld, W. A. Wood, Methods Enzymol. 1975, 42, 258–264. R. K. Scopes, Anal. Biochem. 1984, 136, 525–529. M. C. Shelton, E. J. Toone, Tetrahedron Asymmetry 1995, 6, 207–211. T. Conway, R. Fliege, D. Jones-Kilpatrick, J. Liu, W. O. Barnell, S. E. Egan, Mol. Microbiol. 1991, 5, 2901– 2911. A. T. Carter, B. M. Pearson, J. R. Dickinson, W. E. Lancashire, Gene 1993, 130, 155–156. N. Hugouvieux-Cotte-Pattat, J. Robert-Baudouy, Mol. Microbiol. 1994, 11, 67–75. J. S. Griffiths, N. J. Wymer, E. Njolito, S. Niranjanakumari, C. A. Fierke, E. J. Toone, Bioorg. Med. Chem. 2002, 10, 545–550. S. E. Egan, R. Fliege, S. Tong, A. Shibata, R. E. Wolf, T. Conway, J. Bacteriol. 1992, 174, 4638–4646. T. S. M. Taha, T. L. Deits, Biochem. Biophys. Res. Commun. 1994, 200, 459–466. N. C. Floyd, M. H. Liebster, N. J. Turner, J. Chem. Soc. Perkin Trans. I 1992, 1085–1086. S. T. Allen, G. R. Heintzelman, E. J. Toone, J. Org. Chem. 1992, 57, 426–427. M. C. Shelton, I. C. Cotterill, S. T. A. Novak, R. M. Poonawala, S. Sudarshan, E. J. Toone, J. Am. Chem. Soc. 1996, 118, 2117–2125. I. C. Cotterill, M. C. Shelton, D. E. W. Machemer, D. P. Henderson, E. J. Toone, J. Chem. Soc. Perkin Trans. 1 1998, 1335–1341. D. P. Henderson, M. C. Shelton, I. C. Cotterill, E. J. Toone, J. Org. Chem. 1997, 62, 7910–7911. S. Fong, T. D. Machajewski, C. C. Mak, C.-H. Wong, Chem Biol 2000, 7, 873–883. N. Wymer, L. V. Buchanan, D. Henderson, N. Mehta, C. H. Botting, L. Pocivavsek, C. A. Fierke, E. J. Toone, J. H. Naismith, Structure 2001, 9, 1–10. H. P. Meloche, E. L. O’Connell, Methods Enzymol 1982, 90, 263–269.

265

266

5 Enzyme-catalyzed Aldol Additions 146 A. M. Elshafei, O. M. Abdel-Fatah, AFS, Adv. Food Sci.

2002, 24, 30–36. 147 D. P. Henderson, I. C. Cotterill, M. C. Shelton, E. J.

Toone, J. Org. Chem. 1998, 63, 906–907. 148 W.-D. Fessner, G. Sinerius, Bioorg. Med. Chem. 1994, 2, 639–

645. 149 M. D. Bednarski, E. S. Simon, N. Bischofberger, W.-D.

150

151

152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172

Fessner, M. J. Kim, W. Lees, T. Saito, H. Waldmann, G. M. Whitesides, J. Am. Chem. Soc. 1989, 111, 627–635. W.-D. Fessner, G. Sinerius, A. Schneider, M. Dreyer, G. E. Schulz, J. Badia, J. Aguilar, Angew. Chem. Int. Ed. Engl. 1991, 30, 555–558. C. H. von der Osten, A. J. Sinskey, C. F. Barbas, R. L. Pederson, Y. F. Wang, C.-H. Wong, J. Am. Chem. Soc. 1989, 111, 3924–3927. M. T. Zannetti, C. Walter, M. Knorst, W.-D. Fessner, Chemistry Eur. J. 1999, 5, 1882–1890. C. de Montigny, J. Sygusch, Eur. J. Biochem. 1996, 241, 243– 248. F. Go¨tz, S. Fischer, K.-H. Schleifer, Eur. J. Biochem. 1980, 108, 295–301. W. J. Lees, G. M. Whitesides, J. Org. Chem. 1993, 58, 1887– 1894. A. Straub, F. Effenberger, P. Fischer, J. Org. Chem. 1990, 55, 3926–3932. D. L. Bissett, R. L. Anderson, J. Biol. Chem. 1980, 255, 8750– 8755. W.-D. Fessner, O. Eyrisch, Angew. Chem. Int Ed. Engl. 1992, 31, 56–58. O. Eyrisch, G. Sinerius, W.-D. Fessner, Carbohydrate Res. 1993, 238, 287–306. E. Garcia-Junceda, G. J. Shen, T. Sugai, C.-H. Wong, Bioorg Med Chem 1995, 3, 945–953. A. Brinkkotter, A. Shakeri Garakani, J. W. Lengeler, Arch. Microbiol. 2002, 177, 410–419. G. J. Williams, S. Domann, A. Nelson, A. Berry, Proc. Natl. Acad. Sci. USA 2003, 100, 3143–3148. Y. Takagi, Methods Enzymol. 1966, 9, 542–545. T. H. Chiu, K. L. Evans, D. S. Feingold, Methods Enzymol 1975, 42, 264–269. M. A. Ghalambor, E. C. Heath, Methods Enzymol. 1966, 9, 538–542. P. Moralejo, S. M. Egan, E. Hidalgo, J. Aguilar, J. Bacteriol 1993, 175, 5585–5594. J. Nishitani, G. Wilcox, Gene 1991, 105, 37–42. Z. Lu, E. C. C. Lin, Nucleic Acids Res 1989, 17, 4883–4884. A. Ozaki, E. J. Toone, C. H. von der Osten, A. J. Sinskey, G. M. Whitesides, J. Am. Chem. Soc. 1990, 112, 4970–4971. W.-D. Fessner, J. Badia, O. Eyrisch, A. Schneider, G. Sinerius, Tetrahedron Lett. 1992, 33, 5231–5234. W.-D. Fessner, A. Schneider, O. Eyrisch, G. Sinerius, J. Badia, Tetrahedron Asymmetry 1993, 4, 1183–1192. S. A. Phillips, P. J. Thornalley, Eur. J. Biochem. 1993, 212, 101–105.

References 173 J. P. Richard, Biochem. Soc. Trans. 1993, 21, 549–553. 174 W.-D. Fessner, C. Walter, Angew. Chem. Int. Ed. Engl. 1992,

31, 614–616. 175 D. C. Crans, G. M. Whitesides, J. Am. Chem. Soc. 1985, 107,

7019–7027. 176 P. D’Arrigo, V. Piergianni, G. Pedrocchi-Fantoni, S.

Servi, J. Chem. Soc. Chem. Commun. 1995, 2505–2506. 177 W.-D. Fessner, G. Sinerius, Angew. Chem. Int. Ed. Engl. 1994,

33, 209–212. 178 R. Schoevaart, F. van Rantwijk, R. A. Sheldon, J Org Chem

2000, 65, 6940–6943. 179 H. L. Arth, G. Sinerius, W.-D. Fessner, Liebigs Ann. 1995,

2037–2042. 180 H. L. Arth, W.-D. Fessner, Carbohydrate Res. 1997, 305, 313–

321. 181 D. G. Drueckhammer, J. R. Durrwachter, R. L. Pederson,

182 183 184 185 186 187 188 189

190 191 192 193 194 195 196 197 198 199 200

D. C. Crans, L. Daniels, C.-H. Wong, J. Org. Chem. 1989, 54, 70–77. R. Schoevaart, F. van Rantwijk, R. A. Sheldon, J. Org. Chem. 2001, 66, 4559–4562. R. A. Periana, R. Motiu-DeGrood, Y. Chiang, D. J. Hupe, J. Am. Chem. Soc. 1980, 102, 3923–3927. F. Effenberger, A. Straub, Tetrahedron Lett. 1987, 28, 1641– 1644. S. H. Jung, J. H. Jeong, P. Miller, C.-H. Wong, J. Org. Chem. 1994, 59, 7182–7184. R. L. Pederson, J. Esker, C.-H. Wong, Tetrahedron 1991, 47, 2643–2648. M. L. Valentin, J. Bolte, Bull. Soc. Chimi. Fr. 1995, 132, 1167–1171. T. Gefflaut, M. Lemaire, M. L. Valentin, J. Bolte, J. Org. Chem. 1997, 62, 5920–5922. E. L. Ferroni, V. DiTella, N. Ghanayem, R. Jeske, C. Jodlowski, M. O’Connell, J. Styrsky, R. Svoboda, A. Venkataraman, B. M. Winkler, J. Org. Chem. 1999, 64, 4943–4945. W.-D. Fessner, Curr Opin Chem Biol 1998, 2, 85–97. I. Henderson, K. B. Sharpless, C.-H. Wong, J. Am. Chem. Soc. 1994, 116, 558–561. M.-J. Kim, I. T. Lim, Synlett 1996, 138–140. W.-D. Fessner, C. Gosse, G. Jaeschke, O. Eyrisch, Eur. J. Org. Chem. 2000, 125–132. C.-H. Wong, G. M. Whitesides, J. Org. Chem. 1983, 48, 3199–3205. K. K. C. Liu, C.-H. Wong, J. Org. Chem. 1992, 57, 4789–4791. T. Ziegler, A. Straub, F. Effenberger, Angew. Chem., Int. Ed. Engl. 1988, 27, 716–717. L. Azema, F. Bringaud, C. Blonski, J. Perie, Bioorg. Med. Chem. 2000, 8, 717–722. W. Zhu, Z. Li, Perkin Trans. 1 2000, 1105–1108. M. D. Bednarski, H. J. Waldmann, G. M. Whitesides, Tetrahedron Lett 1986, 27, 5807–5810. N. J. Turner, G. M. Whitesides, J. Am. Chem. Soc. 1989, 111, 624–627.

267

268

5 Enzyme-catalyzed Aldol Additions 201 D. Crestia, C. Guerard, J. Bolte, C. Demuynck, J. Mol.

Catal. B: Enzym 2001, 11, 207–212. 202 C.-H. Wong, F. Moris-Varas, S.-C. Hung, T. G. Marron,

203 204 205 206 207

208

209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226

C.-C. Lin, K. W. Gong, G. Weitz-Schmidt, J. Am. Chem. Soc. 1997, 119, 8152–8158. O. Eyrisch, M. Keller, W.-D. Fessner, Tetrahedron Lett. 1994, 35, 9013–9016. M. Petersen, M. T. Zannetti, W.-D. Fessner, Top. Curr. Chem. 1997, 186, 87–117. J. R. Durrwachter, C.-H. Wong, J. Org. Chem. 1988, 53, 4175–4181. K. K. C. Liu, R. L. Pederson, C.-H. Wong, J. Chem. Soc., Perkin Trans 1991, 1, 2669–2673. J. R. Durrwachter, D. G. Drueckhammer, K. Nozaki, H. M. Sweers, C.-H. Wong, J. Am. Chem. Soc. 1986, 108, 7812– 7818. D. Franke, V. Lorbach, S. Esser, C. Dose, G. A. Sprenger, ¨ ller, R. Takors, M. Mu ¨ ller, M. Halfar, J. Tho¨mmes, R. Mu Chem. Eur. J. 2003, 9, 4188–4196. C. W. Borysenko, A. Spaltenstein, J. A. Straub, G. M. Whitesides, J. Am. Chem. Soc. 1989, 111, 9275–9276. C.-H. Wong, R. Alajarin, F. Moris-Varas, O. Blanco, E. Garcia-Junceda, J. Org. Chem. 1995, 60, 7360–7363. W.-D. Fessner, GIT Fachz. Lab. 1993, 37, 951–956. W. Schmid, G. M. Whitesides, J. Am. Chem. Soc. 1990, 112, 9670–9671. W.-C. Chou, C. Fotsch, C.-H. Wong, J. Org. Chem. 1995, 60, 2916–2917. H. J. M. Gijsen, C.-H. Wong, Tetrahedron Lett. 1995, 36, 7057–7060. B. P. Maliakel, W. Schmid, J. Carbohydr. Chem 1993, 12, 415–424. F. Nicotra, L. Panza, G. Russo, A. Verani, Tetrahedron Asymmetry 1993, 4, 1203–1204. G. Guanti, L. Banﬁ, M. T. Zannetti, Tetrahedron Lett 2000, 41, 3181–3185. G. Guanti, M. T. Zannetti, L. Banﬁ, R. Riva, Adv. Synth. Catal. 2001, 343, 682–691. R. Duncan, D. G. Drueckhammer, J. Org. Chem. 1996, 61, 438–439. F. Effenberger, A. Straub, V. Null, Liebigs Ann. 1992, 1297–1301. W. C. Chou, L. H. Chen, J. M. Fang, C.-H. Wong, J. Am. Chem. Soc. 1994, 116, 6191–6194. G. C. Look, C. H. Fotsch, C.-H. Wong, Acc. Chem. Res. 1993, 26, 182–190. C.-H. Wong, R. L. Halcomb, Y. Ichikawa, T. Kajimoto, Angew. Chem. Int. Ed. Engl. 1995, 34, 412–432. R. L. Pederson, M.-J. Kim, C.-H. Wong, Tetrahedron Lett 1988, 29, 4645–4648. K. K. C. Liu, T. Kajimoto, L. Chen, Z. Zhong, Y. Ichikawa, C.-H. Wong, J. Org. Chem. 1991, 56, 6280–6289. P. Z. Zhou, H. M. Salleh, P. C. M. Chan, G. Lajoie, J. F.

References

227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252

253

Honek, P. T. C. Nambiar, O. P. Ward, Carbohydrate Res. 1993, 239, 155–166. W. J. Lees, G. M. Whitesides, Bioorg. Chem. 1992, 20, 173– 179. R. R. Hung, J. A. Straub, G. M. Whitesides, J. Org. Chem. 1991, 56, 3849–3855. F. Moris-Varas, X.-H. Qian, C.-H. Wong, J. Am. Chem. Soc. 1996, 118, 7647–7652. M. L. Mitchell, L. V. Lee, C.-H. Wong, Tetrahedron Lett. 2002, 43, 5691–5693. M. Schuster, W. F. He, S. Blechert, Tetrahedron Lett 2001, 42, 2289–2291. M. Mitchell, L. Qaio, C.-H. Wong, Adv. Synth. Catal. 2001, 343, 596–599. A. Romero, C.-H. Wong, J. Org. Chem. 2000, 65, 8264–8268. O. Eyrisch, W.-D. Fessner, Angew. Chem. Int. Ed. Engl. 1995, 34, 1639–1641. M. Schultz, H. Waldmann, H. Kunz, W. Vogt, Liebigs Ann. 1990, 1019–1024. D. C. Myles, P. J. I. Andrulis, G. M. Whitesides, Tetrahedron Lett 1991, 32, 4835–4838. R. Chenevert, M. Dasser, J Org Chem 2000, 65, 4529–4531. K. Matsumoto, M. Shimagaki, T. Nakata, T. Oishi, Tetrahedron Lett 1993, 34, 4935–4938. M. Shimagaki, H. Muneshima, M. Kubota, T. Oishi, Chem. Pharm. Bull. 1993, 41, 282–286. R. Chenevert, M. Lavoie, M. Dasser, Can. J. Chem. 1997, 75, 68–73. E. Grazi, M. Mangiarotti, S. Pontremoli, Biochemistry 1962, 1, 628–631. J. F. Williams, P. F. Blackmore, Int. J. Biochem. 1983, 15, 797–816. K. K. Arora, J. G. Collins, J. K. MacLeod, J. F. Williams, Biol. Chem. Hoppe Seyler 1988, 369, 549–557. A. Moradian, S. A. Benner, J. Am. Chem. Soc. 1992, 114, 6980–6987. ¨rmann, G. A. Sprenger, J Biol Chem 2001, 276, M. Schu 11055–11061. ¨rmann, G. A. Sprenger, G. S. Thorell, M. Schu Schneider, J. Mol. Biol. 2002, 319, 161–171. ¨rmann, M. Schu ¨ rmann, G. A. Sprenger, J. Mol. M. Schu Catal. B: Enzym. 2002, 19–20, 247–252. E. Racker, in The Enzymes, Vol. 5 (Eds.: P. Boyer, H. Lardy, K. Myrba¨ck), Academic Press, New York, 1961, pp. 397–406. G. A. Kotchetov, Methods Enzymol. 1982, 90, 209–223. U. Scho¨rken, G. A. Sprenger, Biochim. Biophys. Acta 1998, 1385, 229–243. U. Nilsson, L. Meshalkina, Y. Lindqvist, G. Schneider, J. Biol. Chem. 1997, 272, 1864–1869. E. Fiedler, S. Thorell, T. Sandalova, R. Golbik, S. Konig, G. Schneider, Proc. Natl. Acad. Sci. USA 2002, 99, 591– 595. C. French, J. M. Ward, Biotechnol. Lett. 1995, 17, 247–252.

269

270

5 Enzyme-catalyzed Aldol Additions 254 G. R. Hobbs, R. K. Mitra, R. P. Chauhan, J. M. Woodley,

M. D. Lilly, J. Biotechnol. 1996, 45, 173–179. ¨ rken, G. A. Sprenger, U. 255 J. Bongs, D. Hahn, U. Scho Kragl, C. Wandrey, Biotechnol. Lett. 1997, 19, 213–215. 256 S. Brocklebank, J. M. Woodley, M. D. Lilly, J. Mol. Catal.

B: Enzym. 1999, 7, 223–231. 257 R. P. Chauhan, J. M. Woodley, L. W. Powell, Ann. N. Y.

Acad. Sci. 1996, 799, 545–554. 258 N. J. Turner, Curr Opin Biotechnol 2000, 11, 527–531. 259 K. G. Morris, M. E. B. Smith, N. J. Turner, M. D. Lilly, R.

260 261 262 263 264

265 266

267

268

269 270 271 272 273 274 275

276 277 278

K. Mitra, J. M. Woodley, Tetrahedron Asymmetry 1996, 7, 2185–2188. F. Effenberger, V. Null, T. Ziegler, Tetrahedron Lett 1992, 33, 5157–5160. Y. Kobori, D. C. Myles, G. M. Whitesides, J. Org. Chem. 1992, 57, 5899–5907. J. Bolte, C. Demuynck, H. Samaki, Tetrahedron Lett 1987, 28, 5525–5528. L. Hecquet, C. Demuynck, G. Schneider, J. Bolte, J. Mol. Catal. B: Enzym 2001, 11, 771–776. F. T. Zimmermann, A. Schneider, U. Scho¨rken, G. A. Sprenger, W.-D. Fessner, Tetrahedron: Asymmetry 1999, 10, 1643–1646. A. J. Humphrey, S. F. Parsons, M. E. B. Smith, N. J. Turner, Tetrahedron Lett 2000, 41, 4481–4485. S. V. Taylor, L. D. Vu, T. P. Begley, U. Scho¨rken, S. Grolle, G. A. Sprenger, S. Bringer-Meyer, H. Sahm, J. Org. Chem. 1998, 63, 2375–2377. F. Rohdich, C. A. Schuhr, S. Hecht, S. Herz, J. Wungsintaweekul, W. Eisenreich, M. H. Zenk, A. Bacher, J. Am. Chem. Soc. 2000, 122, 9571–9574. S. Hecht, K. Kis, W. Eisenreich, S. Amslinger, J. Wungsintaweekul, S. Hertz, F. Rohdich, A. Bacher, J. Org. Chem. 2001, 66, 3948–3952. E. Racker, J. Biol. Chem. 1952, 196, 347–365. P. Hoffee, O. M. Rosen, B. L. Horecker, Methods Enzymol. 1966, 9, 545–549. P. A. Hoffee, Arch. Biochem. Biophys. 1968, 126, 795–802. P. Valentin-Hansen, F. Boetius, K. Hammer-Jespersen, I. Svendsen, Eur. J. Biochem. 1982, 125, 561–566. C. F. Barbas, Y. F. Wang, C.-H. Wong, J. Am. Chem. Soc. 1990, 112, 2013–2014. A. Heine, G. DeSantis, J. G. Luz, M. Mitchell, C.-H. Wong, I. A. Wilson, Science 2001, 294, 369–374. H. Sakuraba, H. Tsuge, I. Shimoya, R. Kawakami, S. Goda, Y. Kawarabayasi, N. Katunuma, H. Ago, M. Miyano, T. Ohshima, J. Biol. Chem. 2003, 278, 10799–10806. L. Chen, D. P. Dumas, C.-H. Wong, J. Am. Chem. Soc. 1992, 114, 741–748. H. J. M. Gijsen, C.-H. Wong, J. Am. Chem. Soc. 1995, 117, 2947–2948. J. Liu, G. DeSantis, C.-H. Wong, Can. J. Chem. 2002, 80, 643–645.

References 279 W. Tischer, H.-G. Ihlenfeldt, O. Barzu, H. Sakamoto,

280 281 282 283

284 285

286 287 288 289

290 291 292 293 294 295 296 297

298

299 300

301 302

E. Pistotnik, P. Marliere, S. Pochet, WO 0114566, 2001; Chem. Abstr. 134:208062. N. Ouwerkerk, J. H. Van Boom, J. Lugtenburg, J. Raap, Eur. J. Org. Chem. 2000, 861–866. J. Liu, C.-H. Wong, Angew. Chem. Int. Ed. 2002, 41, 1404– 1407. H. J. M. Gijsen, C.-H. Wong, J. Am. Chem. Soc. 1995, 117, 7585–7591. C.-H. Wong, E. Garcia-Junceda, L. R. Chen, O. Blanco, H. J. M. Gijsen, D. H. Steensma, J. Am. Chem. Soc. 1995, 117, 3333–3339. J. Q. Liu, T. Dairi, N. Itoh, M. Kataoka, S. Shimizu, H. Yamada, J. Mol. Catal. B: Enzym 2000, 10, 107–115. R. Contestabile, A. Paiardini, S. Pascarella, M. L. Di Salvo, S. D’Aguanno, F. Bossa, Eur. J. Biochem. 2001, 268, 6508–6525. B. T. Lotz, C. M. Gasparski, K. Peterson, M. J. Miller, J. Chem. Soc. Chem. Commun. 1990, 16, 1107–1109. A. Saeed, D. W. Young, Tetrahedron 1992, 48, 2507–2514. D. Ura, T. Hashimukai, T. Matsumoto, N. Fukuhara, EP 421477 1991; Chem. Abstr. 115:90657. D. M. Anderson, H.-H. Hsiao, in Biocatalytic Production of Amino Acids & Derivatives (Eds.: D. Rozzell, F. Wagner), Hanser, Munich, 1992, pp. 23–41. M. Bycroft, R. B. Herbert, G. J. Ellames, J. Chem. Soc. Perkin Trans. 1 1996, 2439–2442. V. P. Vassilev, T. Uchiyama, T. Kajimoto, C.-H. Wong, Tetrahedron Lett. 1995, 36, 4081–4084. T. Kimura, V. P. Vassilev, G. J. Shen, C.-H. Wong, J. Am. Chem. Soc. 1997, 119, 11734–11742. J. Q. Liu, T. Dairi, M. Kataoka, S. Shimizu, H. Yamada, J. Bacteriol. 1997, 179, 3555–3560. C. L. Kielkopf, S. K. Burley, Biochemistry 2002, 41, 11711– 11720. M. Fujii, T. Miura, T. Kajimoto, Y. Ida, Synlett 2000, 1046– 1048. T. Miura, M. Fujii, K. Shingu, I. Koshimizu, J. Naganoma, T. Kajimoto, Y. Ida, Tetrahedron Lett. 1998, 39, 7313–7316. K. Shibata, K. Shingu, V. P. Vassilev, K. Nishide, T. Fujita, M. Node, T. Kajimoto, C.-H. Wong, Tetrahedron Lett. 1996, 37, 2791–2794. S. H. Wu, M. Shimazaki, C. C. Lin, L. Qiao, W. J. Moree, G. Weitz-Schmidt, C.-H. Wong, Angew. Chem. Int. Ed. Engl. 1996, 35, 88–90. R. B. Herbert, B. Wilkinson, G. J. Ellames, Can. J. Chem. 1994, 72, 114–117. J. Q. Liu, M. Odani, T. Yasuoka, T. Dairi, N. Itoh, M. Kataoka, S. Shimizu, H. Yamada, Appl Microbiol Biotechnol 2000, 54, 44–51. H. J. M. Gijsen, C.-H. Wong, J. Am. Chem. Soc. 1994, 116, 8422–8423. V. Dalmas, C. Demuynck, Tetrahedron Asymmetry 1993, 4, 2383–2388.

271

272

5 Enzyme-catalyzed Aldol Additions 303 L. Hecquet, J. Bolte, C. Demuynck, Tetrahedron 1994, 50,

8677–8684. 304 U. T. Bornscheuer, M. Pohl, Curr. Opin. Chem. Biol. 2001,

5, 137–143. 305 K. A. Powell, S. W. Ramer, S. B. del Cardayre´, W. P. C.

306 307 308

309 310 311 312 313 314 315 316 317 318 319 320

Stemmer, M. B. Tobin, P. F. Longchamp, G. W. Huisman, Angew. Chem. Int. Ed. 2002, 41, 3949–3959. K. E. Jaeger, T. Eggert, A. Eipper, M. T. Reetz, Appl. Microbiol. Biotechnol. 2001, 55, 519–530. R. Perez Carlon, N. Jourdain, J. L. Reymond, Chem. Eur. J. 2000, 6, 4154–4162. E. Gonzalez-Garcia, V. Helaine, G. Klein, M. Schuermann, G. A. Sprenger, W.-D. Fessner, J.-L. Reymond, Chem. Eur. J. 2003, 9, 893–899. A. Sevestre, V. Helaine, G. Guyot, C. Martin, L. Hecquet, Tetrahedron Lett. 2003, 44, 827–830. D. Wahler, J.-L. Reymond, Curr. Opin. Chem. Biol. 2001, 5, 152–158. G. J. Williams, S. Domann, A. Nelson, A. Berry, Proc. Natl. Acad. Sci. USA 2003, 100, 3143–3148. S. S. Patel, H. D. Conlon, D. R. Walt, J. Org. Chem. 1986, 51, 2842–2844. T. Ouyang, D. R. Walt, S. S. Patel, Bioorg. Chem. 1990, 18, 131–135. T. Ouyang, D. R. Walt, J. Org. Chem. 1991, 56, 3752–3755. R. Jossek, A. Steinbuchel, FEMS Microbiol. Lett. 1998, 168, 319–324. Y. Satoh, K. Tajima, H. Tannai, M. Munekata, J. Biosci. Bioeng. 2003, 95, 335–341. C. Khosla, R. J. X. Zawada, Trends Biotechnol. 1996, 14, 335– 341. P. F. Leadlay, Curr. Opin. Chem. Biol. 1997, 1, 162–168. C. R. Hutchinson, Proc. Natl. Acad. Sci. USA 1999, 96, 3336–3338. T. A. Cropp, B. S. Kim, B. J. Beck, Y. J. Yoon, D. H. Sherman, K. A. Reynolds, Biotechnol. Gen. Eng. Rev. 2002, 19, 159–172.

273

6

Antibody-catalyzed Aldol Reactions Fujie Tanaka and Carlos F. Barbas, III 6.1

Introduction

Antibody catalysts have been designed to process a wide range of aldol reactions. Although aldol reactions can be catalyzed chemically and enzymatically, catalytic antibodies signiﬁcantly extend the range of aldol reactions. Natural aldolase enzymes have substrate speciﬁcity that are predetermined by nature (although they can be engineered); antibody catalysts designed by synthetic chemists can have substrate speciﬁcity diﬀerent from those of natural enzymes. In this chapter, we describe antibodies that catalyze aldol reactions (aldolase antibodies), their features, reactions and applications ranging from synthetic chemistry to the treatment of cancer.

6.2

Generation of Aldolase Antibodies 6.2.1

Antibody as Catalyst Scaﬀold

An antibody immunoglobulin G (IgG) molecule is represented schematically in Figure 6.1. One IgG is composed of four polypeptide chains, two identical light chains, and two identical heavy chains. The molecular weight of an IgG is @150,000 daltons and each IgG has two identical antigen-binding sites. The fragment antigen binding (Fab) portion of the IgG molecules have identical antigen binding sites, however; the molecular weight of this antibody fragment is @50,000 daltons. One antigen binding site is constructed by display of the six complementarity-determining regions (CDR) on a structurally conserved protein framework. The CDR regions are designated LCDR1, LCDR2, and LCDR3 in the light chain variable domain (VL) and HCDR1, HCDR2, and HCDR3 in the heavy variable domain (VL). The immune system has the potential to provide an almost Modern Aldol Reactions. Vol. 1: Enolates, Organocatalysis, Biocatalysis and Natural Product Synthesis. Edited by Rainer Mahrwald Copyright 8 2004 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 3-527-30714-1

274

6 Antibody-catalyzed Aldol Reactions

HCDR1 HCDR2 HCDR3

VH CH1

VL Fab

LCDR1 LCDR2 LCDR3

CL CH2 Fc CH3

Fig. 6.1

Schematic representation of an antibody (IgG) molecule. One antibody molecule has two identical active sites. A light chain is built up from one variable domain (VL) and one constant domain (CL), and a heavy chain from one variable domain (VH), followed by constant domains (CH1, CH2, and CH3). Complementarity determining regions, CDR1–CDR3, vary most signiﬁcantly among diﬀerent immunoglobulins and determine the speciﬁcity of the antigen–antibody interactions.

limitless array of diverse antibodies that can bind to their antigens with high speciﬁcity. The amino acid residues in VL and VH, especially in the six CDR, vary among diﬀerent immunoglobulins, and determine the speciﬁcity of antigen–antibody interactions, whereas constant domains support the structural stability of the antibody and provide connectivity to the rest of the immune system via conserved interactions. By taking advantage of the immune system and the highly speciﬁc binding of antibodies, antibodies that process chemical transformations have been developed [1]. Antigen binding sites of antibodies can serve as the catalytic active sites for catalysis of a wide variety of reactions. 6.2.2

Generation of Aldolase Antibodies that Operate via an Enamine Mechanism

Together with our colleagues, we have developed antibodies that catalyze aldol reactions via the covalent enamine mechanism of natural class I aldolases. Natural class I aldolases utilize the e-amino group of a lysine in their active site to form an enamine, a carbon nucleophile, in their catalyzed aldol reactions [2]. One of the most important issues for development of such antibody catalysts is the pK a of the active site lysine e-amino group. The pK a of the e-amino group of lysine in aqueous solution is 10.7 [3], and the e-amino group is protonated at neutral pH and thus is not nucleophilic

6.2 Generation of Aldolase Antibodies

under physiological conditions. For the e-amino group to be nucleophilic it must be in its uncharged form, i.e. the pK a must be perturbed as observed in class I aldolase enzymes. Although many types of antibody catalyst have been generated, selection of antibody catalysts has typically been based on binding to transition state analogs of the reactants or charged compounds designed using information from the reaction coordinate of a given chemical transformation [1]. Although this strategy has occasionally provided antibody catalysts that use a covalent mechanism, expecting the fortuitous is obviously not a reasonable way to gain access to aldolase antibodies that operate via a covalent enamine mechanism. Thus, together with our colleagues we have developed catalytic function and residue-based selections by using 1,3-diketones for immunization, a strategy later termed reactive immunization [4]. Reactive immunization provides a means of selecting antibody catalysts in vivo on the basis of their capacity to perform a chemical reaction [4, 5]. In this approach, a designed reactive immunogen is used for immunization and chemical reaction(s), for example formation of a covalent bond, occurs in the binding pocket of the antibodies during their induction. The chemical reactivity and mechanism integrated into the antibody by the covalent trap with the reactive immunogen are used in catalytic reactions with substrate molecules. The most highly proﬁcient and eﬃcient catalytic antibodies, aldolase antibodies, have a nucleophilic lysine residue in their active site and were generated by this reactive immunization strategy. Reactive Immunization with the Simple Diketone Derivative Historically, Frank Westheimer used 1,3-diketones as mechanistic probes of the enzyme acetoacetate decarboxylase [6]. The active site of acetoacetate decarboxylase has a nucleophilic lysine e-amino group and forms a Schiﬀ base in its catalytic cycle. In his studies, the active site lysine e-amino group of this enzyme reacted covalently with the 1,3-diketones. Later 1,3-diketones were used as modiﬁcation reagents for the nucleophilic lysine e-amino groups of many enzymes [7]. We used 1,3-diketone hapten 1-carrier protein conjugate for immunization in our experiments to generate aldolase antibodies (Scheme 6.1) [4]. Class I aldolases use the e-amino group of a lysine in their active site to form a Schiﬀ base with one of their substrates and this substrate becomes the aldol donor substrate. Schiﬀ-base formation reduces the activation energy for proton abstraction from the Ca atom and for subsequent enamine formation. The enamine, a carbon nucleophile, then reacts with an aldehyde substrate, the aldol acceptor, to form a new CaC bond. The Schiﬀ base is then hydrolyzed and the product is released [2]. The 1,3-diketone hapten acts as a mechanism-based trap of the requisite lysine residue in the active site and this lysine is necessary for formation of the essential enamine intermediate. The molecular steps involved in trapping the requisite lysine residue are essentially the same chemical steps as are involved in activating 6.2.2.1

275

Scheme 6.1

(a) Diketone 1 used for aldolase antibody induction and selection, and mechanism of trapping of the essential eamino group of a Lys residue in the antibody (AbaNH2 ) binding pocket by using 1. (b) Antibody-catalyzed aldoladdition reaction (AbaNH2 ¼ aldolase antibody). (c) UV spectra of non-catalytic and aldolase antibodies with 1 [1]. The aldolase antibody forms enaminone 2, readily detected at l ¼ 316 nm. (d) Kinetic data for aldolase antibodycatalyzed reactions.

6.2 Generation of Aldolase Antibodies

a substrate ketone to an enamine. These include: (1) catalysis of carbinolamine formation, (2) dehydration, and (3) a-deprotonation of the imminium intermediate to form enamine. This mechanistic mimicry enables selection of an active site that can adapt to the chemical and steric changes that occur during the course of the reaction. Further, the diketone structure provides appropriate binding sites for the two substrates of the intermolecular reaction and facilitates crossing of the entropic barrier intrinsic to this bimolecular reaction. The driving force for reaction of the 1,3-diketone hapten with the antibody is the formation of a stable covalent conjugated enaminone (2) between hapten 1 and e-amino group of lysine (Scheme 6.1). Antibodies prepared from hybridomas derived after immunization with 1,3-diketone hapten 1-carrier protein conjugate were screened for their capacity to form the stable enaminone (absorption maximum l max ¼ 316 nm) with hapten 1. Of the 20 antibodies screened two, 38C2 and 33F12, reacted to form a covalent bond with the 1,3-diketone and generated the enaminone absorption band in the UV. These two antibodies catalyzed aldol addition of acetone to aldehyde 3 to give b-hydroxyketone 4. The saturation kinetics of the antibody-catalyzed reactions were described by the Michaelis–Menten equation. Both antibodies catalyzed the addition of acetone to the si face of 3, irrespective of the stereochemistry at a-position of this aldehyde. These antibodies also catalyzed the retro-aldol reaction of 4. The other antibodies were unable to catalyze the aldol and retro-aldol reactions, indicating that only those that formed the critical enaminone intermediate were active. The catalytic activity of 38C2 and 33F12 was completely inhibited when either hapten 1 or 2,4-pentanedione was added before addition of acetone or aldehyde 3, indicating the antigen-binding site is also the catalytically active site. For the aldol reaction of acetone and aldehyde 3 at pH 7.5, aldolase antibodies 38C2 and 33F12 had ðkcat =Km Þ=k uncat A 10 9 . The eﬃciency of catalysis is largely because of an entropic advantage in the antibody-catalyzed reaction, which is reﬂected as a high eﬀective molarity, kcat =k uncat > 10 5 m. With the best substrates identiﬁed to date for the retro-aldol reaction [8–10], the catalytic proﬁciency ðkcat =Km Þ=k uncat A 6 10 10. The catalytic eﬃciency ðkcat =Km Þ of these aldolase antibodies with these substrates is only @20– 40-fold below that of the most studied aldolase enzyme, fructose-1,6bisphosphate aldolase 4.9 10 4 s1 m1 [2b]. The pK a of the active site lysine was determined to be 5.5 and 6.0 for antibodies 33F12 and 38C2, respectively, from the pH dependence of enaminone formation with 3-methyl2,4-pentanedione [11]. These perturbed pK a values are consistent with the catalysis and mechanism of these antibodies. Combining Reactive Immunization with Transition-state Analogs To improve on the initial concepts of reactive immunization, hapten 5 (Scheme 6.2) was used for induction of aldolase antibodies. Although 1,3diketone hapten 1 enabled generation of the eﬃcient aldolase antibodies 38C2 and 33F12, the hapten does not address the tetrahedral geometry of 6.2.2.2

277

278

6 Antibody-catalyzed Aldol Reactions

(a) O

O R

H

+

H2N Lys Ab R"

≠

δOH R

R'

OH O

R"

R'

δ+ N Lys Ab H

R

R" R'

Transition state (b)

Transition state analog

H N

HO O

O O S

O 5

H2N Lys Ab O O

O O S R O

N Lys Ab H

Reactive immunization Scheme 6.2

Diketone 5, used for generation of aldolase antibodies, includes conceptual elements derived from reactive immunization and transition state analog design. Ab-LysNH2 ¼ aldolase antibody.

the rate-determining transition state of the CaC bond-forming step. Hapten 5 contains features common to the transition state analog of the enzymecatalyzed reaction. The use of a transition state analog approach has been successfully used to generate antibodies that catalyze a diverse set of reactions whereas the 1,3-diketone functionality proved to be key for the reactive immunization strategy [12]. The tetrahedral geometry of the sulfone moiety in hapten 5 mimics the tetrahedral transition state of the CaC bondforming step and should therefore facilitate nucleophilic attack of the enamine intermediate on the acceptor aldehyde substrate. Nine antibodies out of 17 screened reacted with 2,4-pentanedione to form a stable enaminone, and all nine antibodies were catalytic. These results are consistent with a covalent catalytic mechanism in which a reactive amine is programmed in these antibodies. It is of practical interest that immunization with hapten 5 generated the two families of catalysts that operate on opposite optical isomers. Antibodies 93F3 and 84G3 generated with hapten 5 have reactivity antipodal to that of antibodies 38C2 and 33F12 generated with hapten 1, and antibodies 40F12 and 42F1 generated with hapten 5 have the same enantio preference as antibodies 38C2 and 33F12. The rate enhancement, kcat =k uncat , of the best of the nine antibodies (84G3) was 2.3 10 8 ; its catalytic proﬁciency ½ðkcat =Km Þ=k uncat was 7.7 10 13 , approximately 1000-fold higher than that reported for any other catalytic antibody [1c]. In this instance the catalytic eﬃciency ðkcat =Km Þ of 84G3, 3.3 10 5 s1 m1 , is slightly greater than that of the best characterized aldolase enzyme, fructose-1,6-biphosphate aldolase [2b].

6.3 Aldolase Antibody-catalyzed Aldol and Retro-aldol Reactions

O O

carrier protein N H H N carrier protein

N

O

O

O

N O

O

O 6

279

7

Fig. 6.2

Alternative diketones 6 and 7, used for generation of aldolase antibodies.

Reactive Immunization with other Diketones To expand the concept of reactive immunization with diketones, a mixture of diketones 6 and 7 (Figure 6.2) were used for generation of the aldolase antibody 24H6 that operates via an enamine mechanism to catalyze a distinct set of aldol reactions [13]. 6.2.2.3

6.3

Aldolase Antibody-catalyzed Aldol and Retro-aldol Reactions

Aldolase antibodies generated with any of the diketone haptens are typically very broad in scope, accepting a wide variety of substrates. The broad substrate speciﬁcity of aldolase antibodies are unprecedented in the ﬁeld of catalytic antibodies. Traditionally, antibody catalysts are very speciﬁc and only catalyze the reaction of a single substrate (or substrate combination). This is because, in normal immunization, a series of somatic mutations usually leads to highly speciﬁc binding of the inducing antigen [1, 14]. For aldolase antibodies selected with the diketones the usual process of somatic reﬁnement might be aborted because any clone that carries an antibody that has made a covalent bond with the antigen will be selected above clones containing only non-covalent antibody–antigen interactions. This is because no matter how many productive non-covalent interactions are generated by competing clones, they cannot equal the binding energy achieved by a formation of a single covalent bond. When a covalent bond is formed early in the process of antibody evolution, any selective pressure on the reﬁnement process may cease. Antibodies selected in this way are eﬃcient because, like Nature’s enzymes, they were selected on the basis of a chemical reaction. Unlike enzymes, however, they are broad in scope because the usual requirement for reﬁnement of the binding pocket has been circumvented.

280

6 Antibody-catalyzed Aldol Reactions

6.3.1

Antibody 38C2-catalyzed Aldol Reactions

Aldolase antibodies 38C2 and 33F12 generated by immunization with diketone 1 are capable of accelerating more than 100 diﬀerent aldol reactions [4, 8, 11, 15, 16]. Some examples of cross-aldol reactions are shown in Table 6.1. For cross-aldol reactions, a variety of ketones are accepted as donors, including aliphatic open-chain ketones (for example acetone to pentanone), aliphatic cyclic ketones (cyclopentanone to cycloheptanone), functionalized open-chain ketones (hydroxyacetone, dihydroxyacetone, ﬂuoroacetone), and functionalized cyclic ketones (2-hydroxycyclohexanone). As with the donors, the antibodies also accept diﬀerent kinds of aldehyde substrate, for example benzaldehyde derivatives 8–10, a,b-unsaturated aldehyde 11, and aliphatic aldehydes 12 and 13 with products as indicated in Table 6.1. These aldolase antibody-catalyzed reactions are often highly enantioselective. The enantioselectivity rules for 38C2-catalyzed aldol reactions are simple and general, although this selectivity was not directly programmed by the diketone hapten used for the immunization. Asymmetric induction is a consequence of the asymmetry of the active site that directs the attack of the enamine intermediate while stereochemically ﬁxing the face of the acceptor aldehyde. With acetone as the aldol donor substrate a new stereogenic center is formed by attack of the si face of the aldehyde, usually with >95% enantiomeric excess (ee). With hydroxyacetone as the donor substrate, the Z-enamine of hydroxyacetone reacts on the re face of the aldehyde. The antibody-catalyzed reactions of hydroxyacetone as the donor provided only the single regioisomer as shown in the formation of 16, 18, 20, 24, and 25 (Table 6.1). Lower enantioselectivity was achieved with acceptor aldehydes containing an sp 3 center in the a-position, for example the reaction with 12, although reactions with hydroxyacetone as the donor increased the enantioselectivity (for example, formation of 24 and 25 in Table 6.1). In the antibody-catalyzed reactions of 2-butanone with aldehydes, a mixture of both possible regioisomers of aldol products was formed with moderate selectivity by preferential bond formation at the most substituted carbon atom of the ketone. Although a variety of secondary aldols can be prepared by aldolase antibody 38C2-catalyzed cross-aldol reactions, tertiary aldols are typically not accessible via intermolecular cross-aldol reactions. For preparation of enantiomerically enriched tertiary aldols, aldolase antibody 38C2-catalyzed retroaldol reactions can be used (Section 6.3.2). Antibodies 38C2 and 33F12 also catalyzed self-aldol condensations of propionaldehyde and provided the aldol-elimination product 26, and the antibodies did not catalyze the consecutive aldol reaction of 26 with propionaldehyde (Scheme 6.3). The antibodies did not catalyze the self-aldol reactions of aldehydes bearing a longer alkyl chain (bvaleraldehyde, for example, aldehydes 3, 12, and 13). It might not be possible for the antibody to accept two

6.3 Aldolase Antibody-catalyzed Aldol and Retro-aldol Reactions Tab. 6.1

Antibody 38C2-catalyzed cross-aldol reactions. Acceptor

Donor

Product

O O

ee (%) OH O

O

H

O

98

(S)-14

N H

N H

8

OH O O O 15

N H

OH O O

O OH

OH

N H

16

O O

OH O O

H

O (S)-17

N H

>99

N H

9

OH O O

O OH

OH

N H O

18

OH O O

H

98 (S)-19

O2N

O 2N

10

OH O O OH O 2N

20

O

OH O O

H O2N

O

O

99 (S)-21

O 2N

11 H N

OH

H N

O

OH O

O

H

(R)-22

12 H N

O

OH O

O 23

20

281

282

6 Antibody-catalyzed Aldol Reactions Tab. 6.1 (continued)

Acceptor

Donor

Product

ee (%)

H N

OH O

O OH

O 24

77 (de >99)

OH

OH O

O

O H

98

OH

13

25 OH

Substrates

Product

kcat (minC1 )

Km (mM)

8 þ acetone 8 þ hydroxyacetone 9 þ acetone 9 þ hydroxyacetone 10 þ acetone 12 þ acetone

14 16 17 18 19 22

0.048 0.11 0.21 1.1 0.04 0.11

204 348 123 184 27 256

molecules of aldehyde with a larger substituent in the active-site cavity. Selfaldol reactions of acetone or cyclopentanone were also catalyzed by the antibodies in the absence of acceptor aldehyde for a cross-aldol reaction. In this instance the aldol product was also converted to the elimination product by the antibodies (for example, formation of 27). Antibodies 38C2 and 33F12 also catalyzed intramolecular aldol reactions and Robinson annulation (Scheme 6.4); 1,5-diketones 28 and 30 were converted into 29 and 31, respectively [15]. To explore the scope of Baldwin’s rules in antibody-catalyzed intramolecular aldol reactions, 38C2 was incubated with three diﬀerent aliphatic diketones – 2,4-hexanedione, 2,5-heptanedione and 2,6-octanedione. No catalysis was observed in the

O

O

H

H 26 O

O

O OH 27

Scheme 6.3

Antibody 38C2-catalyzed self-aldol reactions.

6.3 Aldolase Antibody-catalyzed Aldol and Retro-aldol Reactions

O

O (S)-28

O

O (R)-28

O

(S)-29

O

O

(R)-29 O

O

O 30

O (S)-31 >95% ee

Substrate

Product

kcat (min-1 )

K m (mM)

(S)-28

(S)-29

0.186

12.4

(R)-28

(R)-29

0.126

2.45

30

31

0.086

2.34

Scheme 6.4

Antibody 38C2-catalyzed intramolecular aldol reactions.

reaction pathway from 2,4-hexanedione to 3-methylcyclopent-2-enone, presumably because of the Baldwin disfavored 5(enolendo-trig) process involved in the attack of the enamine at C2 in 2,4-hexanedione. Although the corresponding ring closure (followed by water elimination) of 2,6-octanedione to give 3-methylcyclohept-2-enone is Baldwin favored (a 7(enolendo-trig) process) in this instance also no product formation was observed. In contrast, the Baldwin favored ring closure reaction of 2,5-heptanedione, (a 6(enolendotrig) process) followed by elimination of water and giving 3-methylcyclohex2-enone was catalyzed by antibody 38C2 [8]. 6.3.2

Antibody 38C2-Catalyzed Retro-aldol Reactions and their Application to Kinetic Resolution

The antibodies also catalyzed retro-aldol reactions of secondary [9, 11, 16] and tertiary aldols [10]. In these retro-aldol reactions antibodies 38C2 and 33F12 processed hydroxyketones whose stereochemistry was the same as that of the aldol reaction product. Kinetic resolution by the retro-aldol reaction therefore provided the opposite enantiomer from the forward aldol reaction (Scheme 6.5 and Table 6.2). For example, (R)-14 (> 99% ee) was obtained by the 38C2-catalyzed kinetic resolution of (G)-14 (Table 6.2) whereas

283

284

6 Antibody-catalyzed Aldol Reactions O

O +

R

H OH

38C2 or 33F12 Enantioselective aldol reaction

O

R Racemic

O

OH

O

R

38C2 or 33F12 Enantioselective retro -aldol reaction (Kinetic resolution)

OH

R

Scheme 6.5

Stereochemistry of antibody 38C2-catalyzed aldol reactions and kinetic resolution.

(S)-14 (98% ee) was formed in the 38C2-catalyzed forward-aldol reaction (Table 6.1). The recovered aldols were highly enantiomerically enriched by the aldolase antibody-catalyzed kinetic resolution; examples include secondary aldols 14, 17, 21, and 32–35, and tertiary aldols 36–40 (Tables 6.2 and 6.3). Kinetic resolution with antibody 38C2 usually aﬀorded higher enantiomeric excess for the recovered aldols than the forward aldol reactions, because one enantiomer was completely consumed in the resolution. For example, (S)-32 was obtained in >99% ee by 38C2-catalyzed kinetic resolution after 67% conversion whereas in the 38C2-catalyzed forward aldol reaction (R)-32 was obtained in 58% ee. In the aldol reactions to form secondary aldols the equilibrium constants favor the aldol product. For example, the aldol reaction of acetone with benzaldehyde has an equilibrium constant of 12 m1 [17]. Aqueous solutions containing 1 m acetone and 1 mm benzaldehyde reached equilibrium when 92% of the benzaldehyde had reacted to form the aldol. In the retroaldol reaction of a 1 mm solution of the aldol, equilibrium was reached at 99% conversion to benzaldehyde. Aldolase antibodies 38C2 and 33F12 eﬃciently provided enantiomerically enriched secondary aldols via aldol reactions or retro-aldol reactions, i.e. kinetic resolution of racemic aldols. The concentration of the substrates determined the reaction direction, aldol or retro-aldol, with these aldolase antibodies. In the aldol reaction of acetone and acetophenone, the equilibrium constant is 0.002 m1 [18]. A 1 mm solution of acetophenone would require the concentration of acetone to be 10,000 m (neat acetone ¼ 13.6 m) to reach 95% conversion. In the retro-aldol reaction, a 1 mm solution of this tertiary aldol would be converted almost completely to its constituent ketones at equilibrium. This is consistent with the fact that aldolase antibodies 38C2 and 33F12 can be used for kinetic resolution of tertiary aldols, but not for the forward aldol reactions to form tertiary aldols. While general access to optically active tertiary aldols is not available by traditional synthetic methods, a variety of enantiomerically enriched tertiary aldols can be prepared by aldolase antibody-catalyzed kinetic resolution. Signiﬁcantly, the aldolase antibodies process not only keto-aldols but also aldehyde-aldols for example, the reaction to provide (S)-40.

6.3 Aldolase Antibody-catalyzed Aldol and Retro-aldol Reactions Tab. 6.2

Secondary aldols prepared by antibody 38C2-catalyzed kinetic resolution. Product

Conversion

ee (%)

52%

>99

52%

>99

51%

>99

67%

>99

50%

95

54%

98

OH O O N H

(R )-14 OH O

O N H

(R)-17 OH O

(R)-21

O2 N

OH O O N H

(S)-32 OH O

MeO

(R )-33 OH O

MeO

(4S,5R)-34 OH O

MeO

(R)-35

Substrate

kcat (minC1 Þ

Km (mM)

kcat /kuncat

kcat /Km /kuncat (MC1 )

(S)-17 (G)-33 (G)-anti-34 (G)-35

1.4 2.2 1.4 1.0

270 16 93 14

1.7 10 7 1.0 10 5 1.9 10 4 1.0 10 6

6.2 10 10 6.3 10 9 2.1 10 8 7.1 10 10

6.3.3

Aldol and Retro-aldol Reactions Catalyzed by Antibodies 93F3 and 84G3

Aldolase antibodies 93F3 and 84G3 also have promiscuous active sites and catalyze reactions of a variety of substrates [12, 19]. These antibodies pro-

285

286

6 Antibody-catalyzed Aldol Reactions Tab. 6.3

Tertiary aldols prepared by antibody 38C2-catalyzed kinetic resolution. Product

Conversion (%)

ee (%)

50

>99

52

80

50

94

50

>99

50

95

OH O

MeO

(R)-36 OH O

O (S)-37

N H

OH O

(R)-38

MeO

OH O

MeO

(R)-39

O 2N

O N

OH O O

H (S)-40

Substrate

kcat (minC1 )

Km (mM)

kcat / kuncat

kcat /Km /kuncat (MC1 )

(S)-36 (G)-37 (G)-38 (G)-40

1.8 4.6 0.15 0.02

0.12 0.11 1.62 0.13

1.2 10 6 8.4 10 5

1.3 10 10 7.7 10 9

vide the opposite enantioselectivity to the 38C2 and 33F12-catalyzed reactions. Examples of antibody 93F3-catalyzed aldol reactions are shown in Table 6.4. Antibodies 93F3 and 84G3 provided (R)-17, (R)-19, and (R)-21 in the aldol reactions of acetone and aldehydes 9, 10, and 11, respectively, whereas antibodies 38C2 and 33F12 provided the corresponding (S) enantiomers in the aldol reactions (Table 6.1). When unsymmetrical ketones were used in antibody 84G3-catalyzed cross-aldol reactions with aldehydes the reactions occurred exclusively at the less substituted carbon atom of the ketones, irrespective of the presence of heteroatoms in the ketones: Aldol products (R)-41–44 were regioselectively and enantioselectively obtained. This is a notable feature, because antibodies 28C2 and 33F12 gave the corresponding regioisomer mixtures and because the background reaction fa-

6.3 Aldolase Antibody-catalyzed Aldol and Retro-aldol Reactions Tab. 6.4

Aldol reactions catalyzed by antibodies 93F3 and 84G3. Acceptor

Donor

Product

O O

OH O H

N H

O

O N H

9 O

O 2N

10

93F3

95

93F3

98

(R)-19 OH O

H

O

11

O2N

O

(R)-21 OH O

H O2N

>99

OH O

O

O2N

93F3 (R)-17

O

H O2N

Antibody ee (%)

O O 2N

10

84G3

95 (dr 100:1)

84G3

95 (dr 99:1)

84G3

95 (dr 100:1)

84G3

95 (dr 98:2)

(R)-41 OH O

O O2N

(R)-42 OH O

O

OMe OMe O2N

(R)-43 OH O

O

SMe SMe O2N

(R)-44

vored the formation of the other regioisomers. In contrast with antibodies 38C2 and 33F12, aldolase antibodies 93F3 and 84G3 were very poor catalysts for aldol reactions involving hydroxyacetone as donor with aldehydes as acceptors. When kinetic resolution of aldol (G)-17 was performed with antibody 93F3, (S)-17 (>99% ee) was obtained at 52% conversion (Table 6.5), whereas antibody 38C2 provided (R)-17 in the kinetic resolution (Table 6.2). The catalytic proﬁciency of antibodies 93F3 and 84G3 usually exceeded that of antibody 38C2, especially for retro-aldol reactions of 3-keto-5-hydroxy-type substrates (for example, 45 and 46). In contrast with antibodies 38C2 and

287

288

6 Antibody-catalyzed Aldol Reactions Tab. 6.5

Examples of kinetic resolution catalyzed by antibodies 93F3 and 84G3. Product

Conversion

ee (%)

52%

>99

50%

99

50%

>99

50%

96

OH O O N H

(S)-17 OH O

MeO

(S)-35 OH O

MeO

(S)-45 OH O

MeS

(S)-46 OH O

MeO

(S)-33

Substrate

Antibody

kcat (minC1 )

Km (mM)

kcat /kuncat

kcat /Km /kuncat (MC1 )

(G)-35

93F3 84G3 93F3 84G3 93F3 84G3

2.63 3.5 43.3 46.8 69.6 81.4

15 23 6.5 10.3 2.6 4.2

2.7 10 6 3.6 10 6 4.9 10 7 5.2 10 7 1.9 10 8 2.3 10 8

1.8 10 11 1.6 10 11 7.4 10 12 5.0 10 12 7.4 10 13 5.4 10 13

(G)-45 (R)-46

33F12, antibodies 93F3 and 84G3 cannot process retro-aldol reactions of tertiary aldols. 6.3.4

Preparative-scale Kinetic Resolution Using Aldolase Antibodies in a Biphasic Aqueous–Organic Solvent System

Antibody-catalyzed reactions are typically performed in aqueous buﬀer, because catalytic antibodies function ideally in an aqueous environment. Many organic molecules of interest are, however, poorly soluble in water. For the transformation of such molecules by aldolase antibodies, a biphasic

6.3 Aldolase Antibody-catalyzed Aldol and Retro-aldol Reactions Tab. 6.6

Preparative scale kinetic resolution in biphasic system. Product

Antibody

Time

Recovery (%)

ee (%)

1.55 g (49)

>97

OH O

N

(R)-47

38C2 (255 mg, 0.025 mol%)

88 h

84G3 (16 mg, 0.015 mol%)

340 h

154 mg (48)

95

38C2 (15.4 mg, 0.10 mol%)

144 h

25 mg (50)

97

91 h

469 mg (47)

97

172 h 259 h

441 mg (42) 458 mg (43)

97 97

193 h

22 mg (44)

99

OH O

N

(S)-47

OH O

MeO

(R)-35 OH O

MeO

(S)-35 OH O

MeO

84G3 (210 mg, 0.065 mol%) reuse (2nd round) reuse (3rd round) 38C2 (18 mg, 0.12 mol%)

(R)-36 OH O

84G3 (500 mg, 0.0086 mol%) MeO

65 h

10 g (50)

>99

(S)-45

aqueous–organic solvent system was especially useful in a large-scale reaction [20]. Although water-miscible solvents can be used to increase the solubility of the substrates, antibody 38C2-catalyzed kinetic resolution of (G)-47 in 20% CH3 CN–buﬀer resulted in reduced reactivity and a lower enantioselectivity than the reaction in 2–5% CH3 CN–buﬀer. The same reaction in toluene–phosphate buﬀer yielded (R)-47 in >97% ee with 49% recovery. Examples of kinetic resolution in biphasic systems are given in Table 6.6. In the biphasic system, racemic substrate (50–100 mm) in toluene or chlorobenzene is mixed with an antibody solution in buﬀer, for example, a 20-gscale reaction is performed in a reaction volume of @700 mL. After completion of the reaction (determined by monitoring the ee of the substrate by HPLC) the mixture is cooled (20 C) and the organic phase is easily separated from the frozen aqueous antibody solution. The product is puriﬁed by conventional column chromatography. The antibody solution is thawed and can be reused. Although the activity of the recycled catalysts was lower

289

290

6 Antibody-catalyzed Aldol Reactions

than on ﬁrst use and longer reaction time was necessary for reactions using recycling catalysts, the enantioselectivity of the reaction with the recycling catalysts was retained. 6.3.5

Aldolase Antibody-catalyzed Reactions in Natural Product Synthesis

Aldolase antibody-catalyzed aldol reactions and kinetic resolution are an eﬃcient means of synthesis of highly enantiomerically pure aldols. These processes have been used for the total synthesis of cytotoxic natural products epothilone A (48) and C (49) (Scheme 6.6) [21]. These compounds have

Scheme 6.6

Syntheses of epothilones using aldolase antibody-catalyzed reactions.

6.3 Aldolase Antibody-catalyzed Aldol and Retro-aldol Reactions

OH

OH N

N HO

S

O 55 99% ee (50% conversion) by 84G3 99% ee (50% conversion) by 93F3

S

MeO

O 56 95% ee (51% conversion) by 84G3 95% ee (51% conversion) by 93F3

OH

OH

N MeS

N

S

O 57

99% ee (55% conversion) by 84G3 99% ee (54% conversion) by 93F3

S

O F 96% ee (54% conversion) by 84G3 98% ee (52% conversion) by 93F3 58

Fig. 6.3

Chiral precursors for syntheses of epothilones and their derivatives, which were prepared by aldolase antibodycatalyzed reactions.

a taxol-like mode of action, functioning by stabilization of cellular microtubules. The structural moieties (þ)-syn-50 and ()-51 were prepared by antibody 38C2-catalyzed reactions and converted to the intermediates 52 and 53, respectively. The key compound ()-51 (98% ee) was also obtained by kinetic resolution using antibodies 93F3 and 84G3 [22]. Chiral precursors 55–58 (Figure 6.3) for the syntheses of other epothilones and their derivatives were also prepared by the aldolase antibody-catalyzed kinetic resolution [22]. Other syntheses of natural products have capitalized on the 38C2catalyzed aldol addition of hydroxyacetone to install 1,2-syn-diol functionality. Aldolase antibody 38C2 was also used for synthesis of brevicomins 59– 61 (Scheme 6.7) [23] and 1-deoxy-l-xylulose (62) (Scheme 6.8) [24]. Antibody 38C2-catalyzed kinetic resolution of a tertiary aldol was used for the synthesis of (þ)-frontalin (63) (Scheme 6.9) [10]. 6.3.6

Retro-aldol Reactions in Human Therapy : Prodrug Activation by Aldolase Antibody

An important application of aldolase antibodies is prodrug activation in chemotherapeutic strategies. Activation of a prodrug into an active drug at the tumor site enables selective destruction of those tumor cells. This type of site-speciﬁc targeting is known as antibody-directed enzyme prodrug therapy (ADEPT) [25]. The ADEPT complex serves two functions. The antibody portion of the complex enables delivery of the drug directly to the tu-

291

6 Antibody-catalyzed Aldol Reactions

292

O OH

O

38C2 O

O

CHO

55%, >99% ee

OH O

O

O

+ HO O

O O

60

OH O O

OH 38C2

O

O 59

OH

CHO

O

HO

O

O

OH

>99% ee

O O

HO

O

+ O

61

O

HO Scheme 6.7

Syntheses of brevicomins using aldolase antibody-catalyzed reactions.

O OH O O

H2 Pd(OH)2/C

OH O

38C2

OH HO

O

H 32%, 97% ee

OH

O

OH

1-Deoxy-L-xylulose (62) Scheme 6.8

Synthesis of 1-deoxy-l-xylulose using aldolase antibody-catalyzed reaction.

mor by recognition of antigen on the tumor cell surfaces. The enzyme part of the conjugate catalyzes the prodrug activation reaction at the tumor site. This system enhances the eﬃciency of anti-cancer drugs and reduces peripheral cytotoxicity because of the prodrug’s low toxicity. Most ADEPT sysOH O O

H

AcNH

44%, >99% ee (~50% conversion)

O OH

PO(OEt)2 O

LiOH

OH O

38C2 O

H

AcNH O

O H2 Pd(OH)2/C

AcNH Scheme 6.9

Synthesis of (þ)-frontalin using aldolase antibody-catalyzed reaction.

O (+)-Frontalin (63)

6.4 Aldolase Antibodies for Reactions Related to an Enamine Mechanism

tems incorporate a bacterial enzyme, and the problems of this system are: (1) identiﬁcation of an enzyme not already present in humans, and (2) the immunogenicity of such a bacterial enzyme. These problems can be circumvented by use of a humanized catalytic antibody which catalyzes the activation reaction of the prodrug selectively, in place of the foreign enzyme [26]. Immunogenicity of the enzyme component of ADEPT can therefore be solved using humanized catalytic antibodies. Antibody 38C2 catalyzes the activation reactions of prodrugs that incorporate a trigger portion designed to be released by sequential retro-aldolretro-Michael reactions (Scheme 6.10) [27, 28]. The retro-Michael step is also catalyzed by antibody 38C2, although this step occurs spontaneously in buﬀer. Because the retro-aldol reaction of the tertiary aldol in this reaction cascade is not catalyzed by any known natural enzymes, this masking of the anti-cancer drugs substantially reduces their toxicity. Combination of doxorubicin-prodrug 64 and antibody 38C2 strongly inhibited cell growth of cancer cell lines, whereas the same concentration of 64 alone was far less potent [27]. Camptothecin- and etoposide-prodrugs 65 and 66 were also activated by 38C2. The aldolase antibody-prodrug system has also proven to be eﬃcient in an animal model of cancer [28]. Incorporation of an additional 8.4-A˚ linker, as shown in prodrugs 65 and 66, enables a diverse group of drugs to be used for application of aldolase antibody-catalyzed prodrug activation. Aldolase antibody 33F12 has an active site lysine e-amino group at @10 A˚ depth in a narrow pocket (Section 6.6). The reaction sites of the masking linkers of 65 and 66 reach into the active site whereas the bulky drug molecules remain outside the active-site cavity. The prodrug activation strategy has also been demonstrated in the context of protein activation. Native insulin modiﬁed with aldol-terminated linkers at the primary amines crippled the biological activity of insulin (Scheme 6.11). The modiﬁed insulin was defective with regard to receptor binding and stimulation of glucose transport. Antibody 38C2 cleaved the linker, released insulin, and restored insulin activity in an animal model [29]. Antibody 38C2 was also used for the retro-aldol or retro-aldol-retro-Michael reactions of aldol functionality on dendrimers and polymers that have potential for drug encapsulation and delivery [30, 31].

6.4

Aldolase Antibodies for Reactions Related to an Enamine Mechanism and the Nucleophilic Lysine e-Amino Group

Because aldolase antibodies operate by an enamine mechanism, they also catalyze other reactions that proceed by a similar mechanism. For example, as described in the section on prodrug activation reactions, antibodies 38C2 and 33F12 catalyze b-elimination (retro-Michael) reactions (Section 6.3.6). These antibodies also catalyze decarboxylation of b-keto acids (Scheme 6.12)

293

N

OH

O

O

O

O

O

O

O

65

HO

N

66

8.4 Å

OCH3

O

O

O

O

64

O

O

OH

O

N O

HO

CO2

O

38C2

CO2

O

38C2

O

O

OH

OH

HOHN

O

O

N

O

O

O

O

O

N

N

O

OH

O

N

O

CH3O

Etoposide

OH

O

O

O

OCH3

O

O

NH

spontaneous

OH

O

O

O

CO2

O

38C2

CH3O

O

O

O O HO

OH

additional linker arm to enable eﬃcient catalysis of drug release.

O

O

CH3O

HO

O

3.7 Å

O

38C2

O

O

N

O

OH

Prodrug activation via a tandem retro-aldolretro-Michael reaction catalyzed by antibody 38C2. Prodrugs 65 and 66 incorporate an

Scheme 6.10

N

O

HOHN

O

OH

O

CH3O

O

O

O O HO

CH3O

OH

O

O

OH O

OH

O N

N

O

O

OH

OH O

OH

O

Camptothecin

N

Doxorubicin

HO NH2

O

O

294

6 Antibody-catalyzed Aldol Reactions

6.4 Aldolase Antibodies for Reactions Related to an Enamine Mechanism

295

NH2 Insulin

H 2N

NH2 O2N

O

38C2 O

OH O O

O O

O

OH O

O

NH Insulin

N H

OH O

O

HN

OH O O

Scheme 6.11

Modiﬁcation of insulin with retro-aldol-retroMichael linkers and its reactivation by aldolase antibody 38C2.

[32], allylic rearrangement of steroids (Scheme 6.13) [33], and deuteriumexchange reactions (Table 6.7) [34]. In the antibody 38C2-catalyzed decarboxylation of 67, incorporation of 18 O into product 68 was observed in the presence of H2 18 O, consistent with decarboxylation proceeding via an enamine intermediate [32]. Antibody 38C2 catalyzed deuterium-exchange reactions at the a-position of a variety of ketones and aldehydes (Table 6.7) [34]. Because aldehydes bearing a longer alkyl chain (bvaleraldehyde) were not substrates for the

O O N H

O

67

OH Lys Ab

H H2N Lys Ab

N O

O N H

O 18O

18

H2 O

O N H

68

Scheme 6.12

Antibody 38C2-catalyzed decarboxylation.

O–

Lys Ab

H

N+

CO2

N H

296

6 Antibody-catalyzed Aldol Reactions

38C2 O

O 38C2 O

O O

O 38C2

O

O

Scheme 6.13

Antibody 38C2-catalyzed allylic rearrangements.

38C2-catalyzed self-aldol reactions (Section 6.3.1), deuterium-exchanged aldehydes accumulated in the presence of 38C2 in D2 O. Aldolase antibodies can be covalently modiﬁed with cofactor derivatives at the active site lysine residue to enable catalysis of cofactor-dependent reactions [35, 36]. A variety of 1-acyl b-lactam derivatives formed the stable amide linkages to active site lysine, whereas 1,3-diketones bound reversibly (Scheme 6.14) [35]. Antibody 38C2 was modiﬁed with lactam 69, and the modiﬁed antibody catalyzed thiazolium-dependent decarboxylation of PhCOCOOH. Antibody 38C2 was also covalently modiﬁed with the succinic anhydride derivative bearing a bis-imidazole functionality that chelates Cu(II), and the modiﬁed antibody was used for Cu(II)-dependent ester hydrolysis [36]. The reactions of the active site lysine e-amino group of aldolase antibodies depend on the substrates – the catalytic lysine residue of antibody 38C2 did not react with polymer p-nitrophenyl ester [37]. An eﬀect of metal-cofactors on antibody 38C2-catalyzed aldol reactions has also been reported [38]. The active site lysine e-amino group of antibody 38C2 also functioned as general base for the Kemp elimination [39]. Over the years, we have studied a wide-variety of enamine based chemistries using aldolase antibodies. Several key synthetic reactions have been explored using aldolase antibodies. These include investigation of antibodycatalyzed additions to imines, i.e. antibody-based Mannich reactions that operate through an enamine intermediate, and the analogous reaction with nitrostyrene-derived electrophiles (Michael reactions) and Diels–Alder reactions wherein the antibody either generates an enamine derived diene or activates a dienophile through an iminium intermediate. With the exception of the Michael reaction, only recently shown to proceed exclusively with a maleimide electrophile [40], all other reactions failed, presumably because of steric constraints imposed by what is otherwise a promiscuous active site. Signiﬁcantly, these exploratory studies with catalytic antibodies set the stage

6.5 Concise Catalytic Assays for Aldolase Antibody-catalyzed Reactions Tab. 6.7

Antibody 38C2-catalyzed deuterium-exchange at the a-position. Substrate

kcat (minC1 )

O

kcat /kuncat

79

1.1 10 8

105

9.7 10 7

29.9

25

3.0 10 8

84.9

5

5.9 10 8

13.5

9.5 10 6

10.7

O

Km (mM)

9.7 3:5 ðCH2 Þ 0:6 ðCH3 Þ

O

O

O

0.95

O

19.0

0.7

27.6

0.6

H O H O

127

1

H O H

5.8

0.5

for application of these principles in the area of organocatalysis by our laboratory and former members of the laboratory familiar with these studies [41].

6.5

Concise Catalytic Assays for Aldolase Antibody-catalyzed Reactions

Spectroscopic or visible detection of antibody-catalyzed reactions enhances rapid characterization of catalysts on a small scale. Such detection systems are also useful for high-throughput screening for new aldolase antibody

297

298

6 Antibody-catalyzed Aldol Reactions

O

H2N Lys Ab

O

R

O

H

N

Lys Ab

R

R

O

O

O O H2N Lys Ab

N

Lys Ab R

N H

N H

O O

N S + Br –

N 69 Scheme 6.14

Cofactor introduction at a unique active site lysine of aldolase antibodies.

catalysts and for evolution of aldolase antibodies in vitro. Examples of substrates for spectroscopic or visible detection of the antibody-catalyzed reactions are shown in Scheme 6.15. Substrate 47 was the ﬁrst UV–visible-active aldol substrate designed for following retro-aldol reactions. Reaction results in liberation of the yellow product 70 [9] and this reagent served as the basis for the development of ﬂuorescent versions when ﬂuorescent aldehydes and ketones were later identiﬁed [42]. Substrates 35, 36, 71, and 72 liberate the ﬂuorescent products 73, 74, 75, and 76, respectively, by aldolase antibodycatalyzed retro-aldol reaction [42]. Thus the progress of the reactions with these ﬂuorogenic substrates can be followed by ﬂuorescence. Substrates 77– 79 are used for analysis of retro-aldol-retro-Michael reactions. Substrates 77 and 78 generate ﬂuorescent compounds resoluﬁn (80) and umbelliferone (81), respectively, after the reactions [42, 43]. Substrate 79 generates a 2naphthol derivative 82 that forms a visible colored azo dye with diazonium salts [44]. Maleimide derivative 83 is useful for detection of the carbon– carbon bond-formation catalyzed by aldolase antibodies [40]. It should be noted that all the other systems monitor carbon–carbon bond cleavage, not formation. The reaction of 83 with acetone provides 84, the ﬂuorescence of which is much greater than that of 83.

6.6

Structures of Aldolase Antibodies and Reaction Mechanism of Nucleophilic Lysine e-Amino Group

Antibodies 38C2 and 33F12, generated with diketone 1, are highly homologous with regard to sequence and have an essential lysine catalytic residue at the same position (H93) [11, 45]. Aldolase antibodies 40F12 and 42F1, generated with hapten 5, have the same enantio-preference as antibodies 38C2 and 33F12. They are also highly similar to 38C2 and 33F12 in their

O

OH O

N

N

47

H

fluorescence

λex 330 nm, λem 452 nm

fluorescence

λex 330 nm, λem 452 nm

fluorescence

λex 364 nm, λem 531 nm

MeO

fluorescence

λex 364 nm, λem 531 nm

73 O

OH O

MeO

36

74 O

OH O

H N

71

N

75 O

OH O

N

N

72

N O

76

N

OH O O

O

λmax 400 nm

O

35

MeO

UV/VIS

70

OH O

MeO

H

O

N

O O

O

O fluorescence

OH

O 80

77

λex 544 nm, λem 590 nm

OH O λex 360 nm, λem 460 nm

fluorescence O

O

O

O

O

78

OH 81

Br

CH3 N2 Br O

HO

1/2 ZnCl2

Br Cl

OH O

+Cl-

HO

N N CH3

82

79

Cl bright red dye precipitate O

H N

N N

+

O

O

H N

N N

83

O

84 fluorescence

O

O

λex 315 nm, λem 365 nm

Scheme 6.15

Substrates for ﬂuorescent and visible detection of aldolase antibody-catalyzed reactions.

300

6 Antibody-catalyzed Aldol Reactions

amino acid sequences and have an essential lysine catalytic residue at the same position (H93) [16]. On the other hand, aldolase antibodies 93F3 and 84G3, generated with hapten 5, have antipodal reactivity to antibodies 38C2 and 33F12 and have diﬀerent amino acid sequences from antibodies 38C2, 33F12, 40F12, and 42F12. The essential lysine residue is at position L89. X-ray crystal structures of the Fab fragments of 33F12 [11] and 93F3 [1d] are shown for comparison in Figures 6.4 and 6.5.

(a)

(b)

Fig. 6.4

X-ray structure of aldolase antibody 33F12. (a) Top view. Residues within 5 A˚ of the eamino group of the catalytic lysine (H93) are indicated. (b) Side view.

6.6 Structures of Aldolase Antibodies

Fig. 6.5

X-ray structure of aldolase antibody 93F3. The essential lysine residue (L89) is indicated.

The structure of 33F12 shows that the entrance of the antigen binding site of 33F12 is a narrow elongated cleft. The binding pocket is more than 11 A˚ deep and is comparable with combining sites of antibodies raised against other small haptenic molecules. At the bottom of the pocket LysH93 is found within a hydrophobic environment. Within this pocket, only one charged residue is within an 8-A˚ radius of the nitrogen of e-amino group of LysH93. No salt bridges or hydrogen bonds can be formed between LysH93 and any other residues in 33F12. A Hansch plot, a study of the free energy relationship for substrate partitioning into n-octanol and kcat =Km of the retro-aldol reactions, shows that the active site in 33F12 is 1.1 times more hydrophobic than n-octanol [11]. This environment perturbs the pK a of LysH93 and enables it to exist in its uncharged form, facilitating its function as a strong nucleophile. The hydrophobic environment used by these aldolase antibodies to tune the pK a of the e-amino group of the active site lysine residue is in contrast with the mechanism used by most natural class I aldolase enzymes; in the natural aldolases an electrostatic mechanism perturbs the pK a of the e-amino group of the lysine [2, 16, 46]. Notably, with the exception of a single residue, the residues in van der Waals contact with LysH93 are conserved in both antibodies and are encoded in the germline gene segments used by these antibodies. This conservation suggests that LysH93 appeared early in the process of antibody evolution in a germline antibody. The insertion of this residue into this hydrophobic microenvironment resulted in chemical reactivity that was eﬃcient enough to be selected. Once this covalent process appeared the

301

302

6 Antibody-catalyzed Aldol Reactions

binding pocket did not further evolve toward high speciﬁcity as would be indicated by the selection of somatic variants of the germline sequence. The broad substrate speciﬁcity of aldolase antibodies prepared by reactive immunization is likely to be the result of the special ontogeny of antibodies induced by immunogens that form covalent bonds within the binding pocket during the induction. Although all aldolase antibodies generated with 1,3-diketones use an enamine mechanism in their aldol and retro-aldol reactions, the micromechanisms of the reactions might diﬀer among these aldolase antibodies as they do among aldolase enzymes [2, 47]. In the 38C2-catalyzed retro-aldol reactions of a series of aldols prepared from acetone and p-substituted cinnamaldehyde, correlation between log kcat and the Hammett substituent s is not linear [13, 48]. A positive linear correlation between log kcat and s was, on the other hand, obtained for aldolase antibody 24H6, generated with a mixture of haptens 6 and 7; this indicates 38C2 and 24H6 diﬀer in their micromechanisms, including the rate-limiting steps. However, similar amino acid sequences of the aldolase antibodies reﬂect the broad similarity in substrate speciﬁcity, enantioselectivity, and micromechanism.

6.7

Evolution of Aldolase Antibodies In Vitro

Although aldolase antibodies are broad in scope, the eﬃciency with which any given aldol is processed can vary signiﬁcantly. To create aldolase antibodies with altered substrate speciﬁcity and turnover, phage libraries were screened using diﬀerent diketone derivatives [49]. In this approach, libraries were prepared by recombining the catalytic machinery of well-characterized aldolase antibodies with a naive V gene repertoire. In vitro selection systems enabled the use of multiple haptens without animal re-immunization and enabled the experimenter to combine insights gained by the study of existing catalytic antibodies with the diversity of the immune repertoire. This strategy was used to prepare catalysts that would eﬃciently process cyclohexanone-aldols 85, because retro-aldol reactions of 85 were relatively slow compared with those involving acetone-aldols using existing aldolase antibodies 38C2 and 33F12 (Scheme 6.16). The phage libraries [50] prepared by combination of active site residues of 38C2 and 33F12 with a naive V gene repertoire were selected with 86 and 1 as reactants. Fab 28 obtained from this selection catalyzed the retro-aldol reactions of anti-87, syn-87, and 35. The kcat values of Fab 28 were superior to those of parental antibodies for cyclohexanone-aldols anti-87 and syn-87 by approximately three- to tenfold [49]. In addition, Fab 28 catalyzed the reaction of acetone-aldol 35 with a kcat value similar to that of antibody 33F12. The stereochemistries of the preferred substrate enantiomers of Fab 28 were the same as those of the parental antibodies 38C2 and 33F12. On the basis of on the design of

6.6 Structures of Aldolase Antibodies O

OH

303

O retro-aldol

R

+ RCHO

85 O

O

O O

O

OH

O

O

N H

86

N H

O

carrier

OH

O

O

O

OMe 35

O Ph

Ph 88

N H

OH

OMe syn-87

O

N H

1

OMe anti-87 O

O

89

Scheme 6.16

Reactions and compounds for in vitro evolution of aldolase antibodies.

the library, Fab 28 retained speciﬁc sequence elements of the parental antibodies and the essential LysH93 of the catalytic mechanism. The remaining primary sequence of Fab 28 is not related to the parental antibodies. In addition, because a naive V gene library was generated using human bone marrow cDNA, Fab 28 is a human aldolase antibody. This strategy will be useful for providing human antibodies for catalytic antibodymediated prodrug activation described in Section 6.3.6 without the need for re-immunization and selection. A correlation was observed between the kcat of the antibody-catalyzed retro-aldol reaction and the apparent K d of the corresponding diketones (i.e. the reactivity to the diketones) within the family of aldolase antibodies 38C2, 33F12, Fab 28, and antibodies selected with Fab 28 in vitro. Stronger binding (lower K d value) to acetone-diketone 88 (i.e. higher reactivity with acetone-diketone 88) correlated with a higher kcat value for the reaction of acetone-aldol 35, and stronger binding to cyclohexanone-diketone 89 correlated with a higher kcat value for the reaction of cyclohexanone-aldol anti87 (Figure 6.6) [51]. Selection using a structure-altered diketone provided catalytic antibodies that had altered substrate speciﬁcity as directed by the structure of the selecting diketone. The correlation indicates that antibodies 38C2, 33F12, and the in vitro evolved aldolase antibodies share a similar micromechanism including the rate-limiting step in their catalyzed reactions. 1,3-Diketones have also been used for phage selections of small peptides that catalyze aldol and retro-aldol reactions via an enamine mecha-

carrier

304

6 Antibody-catalyzed Aldol Reactions

10

1

1 Fab 28 33F12

0.1

0.1

Fab 9 Fab 22 L87Phe

Fab 28

Fab 22 L87Phe

kcat (min-1)

kcat (min-1)

38C2

0.01

38C2

Fab 9

33F12

Fab 3

0.01

Fab 3

0.001 Fab 10

0.001

0.0001 0.1

1

(a)

10

100

0.1

1

(b)

K d (µM)

10

100

K d (µM)

Fig. 6.6

(a) Plot of k cat for the retro-aldol reaction of 35 against K d of diketone 88. (b) Plot of k cat for the retro-aldol reaction of anti-87 versus K d of diketone 89.

nism [52]. Selection with 1,3-diketones seems to be a general and eﬀective route to catalysts that operate via an enamine mechanism. Although in vitro selection strategies using 1,3-diketones can be used to obtain access to aldolase antibodies, direct selection of catalysis resulted in improved catalysts. Fluorogenic and chromogenic substrates described in Section 6.5 might be useful for the catalytic selection. Genetic selection is also useful for the selection of catalysts. Antibody-catalyzed reaction of substrate 90, a prodrug, generates p-aminobenzoic acid (PABA), an essential metabolite for E. coli. When an E. coli strain that cannot synthesize PABA, because of a genetic defect was provided with the gene for an aldolase antibody, expression of the aldolase antibody provided the E. coli with the ability to metabolize 90 and survive. Such a genetic selection has been shown in preliminary experiments to provide a growth advantage to a strain expressing aldolase antibody 38C2 (Scheme 6.17) [53]. Ideally, such a system will be used in the future to rapidly evolve aldolase antibodies based on both turnover eﬃciency and substrate speciﬁcity.

O HO CH3 O 90

COOH

COOH

O N H

H2N O

PABA

O CO2

Scheme 6.17

Reaction for genetic selection of aldolase antibodies.

6.9 Summary and Conclusion

305

O N H

linker

N+ H N

OH

O 91

H2 N

H N

O O H +

O

94 O

OH O

OH O

72D4 N H

92

N H

93

Scheme 6.18

Antibody-catalyzed reaction using cofactor primary amine.

6.8

Cofactor-mediated Antibody-catalyzed Aldol and/or Retro-aldol Reactions

Antibody 72D4, generated by immunization with 91, catalyzed the aldol reaction of acetone and aldehyde 92 and the retro-aldol reaction of 93, with some enantioselectivity, in the presence of a primary amine cofactor 94 (Scheme 6.18) [54]. The antibody did not catalyze the reactions in the absence amine 94, and evidence supported an enamine mechanism (enamine formation with 94) for this cofactor amine-mediated antibody-catalyzed reaction. An antibody that mimics threonine aldolase, which uses pyridoxal as the cofactor to catalyze the aldol reaction of glycine with aldehydes, has also been reported. Antibody 10H2 catalyzed the retro-aldol reaction of bhydroxy-a-amino acid in the presence of pyridoxal [55].

6.9

Summary and Conclusion

Class I aldolases function as Nature’s most fundamental carbon–carbon bond forming enzymes. Designer catalysts – aldolase antibodies – that mimic the aldolases have been created by using a reaction-based selection strategy with 1,3-diketones. The covalent reaction mechanism is a fundamental part of the selection. Broad scope, enhanced catalytic activity, and deﬁned chemical mechanism are three features of these aldolase antibodies that distinguish them from traditional antibody catalysts. The substrate speciﬁcities of aldolase antibodies are diﬀerent from those of naturally existing

306

6 Antibody-catalyzed Aldol Reactions

enzymes. Thus, applications of aldolase antibody-catalyzed reactions are very wide, from asymmetric synthesis to chemical transformations in cancer therapy [56, 57].

6.10

Experimental Procedures

Aldolase antibodies 38C2 and 84G3 are commercially available from Aldrich. Example 1. Kinetic Resolution in a Biphasic System [20]. A solution of antibody 84G3 (500 mg, 6.67 mmol active site, 0.0086 mol%) in phosphate buffered saline (PBS; pH 7.4, 87.5 mL) at room temperature was added to a solution of (G)-45 (20 g, 77.4 mmol) in toluene (600 mL) in a Teﬂon tube. The mixture was shaken at 250 rpm. under argon at the same temperature and the ee was monitored by chiral-phase HPLC. When ee > 99% was reached (65 h) the mixture was cooled to 20 C for several hours to enable the aqueous phase to freeze. The organic phase was decanted. The frozen aqueous phase was left to thaw and extracted with toluene (3 3 vol. equiv.) and 1:1 EtOAc–toluene (3 3 vol. equiv.). The combined organic phase was dried over Na2 SO4 , concentrated in vacuo, and puriﬁed by silica gel column chromatography (1:3 EtOAc–hexane) to aﬀord (S)-45 (10 g, 50%). Example 2. Kinetic Resolution [22]. A solution of antibody 84G3 (0.34 g, 0.00227 mmol) in PBS (27.2 mL) was added to a degassed solution of (G)-51 (16.8 g, 75 mmol) in PBS (1.55 L)–CH3 CN (140 mL). The mixture was incubated under an argon atmosphere at 37 C for 5 days. After consumption of more than 98% of one enantiomer, as judged by HPLC analysis, the mixture was dialyzed using Amicon membranes to recover the antibody and the ﬁltrate was passed through a reversed-phase column (C18 ). The column was ﬁrst washed with water and then the desired compounds were eluted with MeOH. The solvents were removed under vacuum and the residue was puriﬁed by silica gel column chromatography (EtOAc–hexanes, 9:1 to 2:1) to aﬀord ()-51 (7.6 g, 45%, >98% ee). Example 3. Aldol Reaction [8]. Antibody 38C2 (120 mm in PBS, 8.0 mL) was added at room temperature to a mixture of 11 (110 mg, 0.61 mmol) in DMF (15 mL), acetone (31 mL), and degassed PBS (571 mL). The ﬁnal concentrations were: 11, 1.0 mm; 38C2, 1.9 mm; and acetone, 5% (v/v). The reaction mixture was kept under argon at room temperature in the dark for 7 days. The reaction mixture was saturated with NaCl and extracted with EtOAc (3 150 mL). The organic phase was dried over MgSO4 , concentrated in vacuo, and puriﬁed by silica gel column chromatography (EtOAc–hexane, 1:2) to aﬀord (S)-21 (96 mg, 67%, 91% ee).

References

Acknowledgments

We would like to thank Richard A. Lerner, our long-term collaborator in the aldolase antibody projects performed at the Scripps Research Institute, for his contributions.

References 1 (a) Schultz, P. G.; Lerner, R. A. Science 1995, 269, 1835.

2

3 4 5 6 7

8

9

10

11

12

(b) Hilvert, D. Annu. Rev. Biochem. 2000, 69, 751. (c) Blackburn, G. M., Garcon, A., Catalytic antibodies. In Kelly, D. R. Ed. Biotechnology vol. 8b, Biotechnology Biotransformations II; Wiley–VCH: Weinheim, 2000; p. 403. (d) Schultz, P. G.; Yin, J.; Lerner, R. A. Angew. Chem. Int. Ed. 2002, 41, 4427. (e) Tanaka, F. Chem. Rev. 2002, 102, 4885. (a) Lai, C. Y.; Nakai, N.; Chang, D. Science 1974, 183, 1204. (b) Morris, A. J.; Tolan, D. R. Biochemistry 1994, 33, 12291. (c) Choi, K. H.; Shi, J.; Hopkins, C. E.; Tolan, D. R.; Allen, K. N. Biochemistry 2001, 40, 13868. Allard, J.; Grochulski, P.; Sygusch, J. Proc. Natl. Acad. Sci. U.S.A. 2001, 98, 3679. Heine, A.; DeSantis, G.; Luz, J. G.; Mitchell, M.; Wong, C.-H.; Wilson, I. A. Science 2001, 294, 369. Lide, D. R. Ed. CRC Handbook of Chemistry and Physics 83rd Edition; CRC Press: New York, 2002, p. 7-1. Wagner, J.; Lerner, R. A.; Barbas, C. F., III Science 1995, 270, 1797. Wirshing, P.; Ashley, J. A.; Lo, C.-H. L.; Janda, K. D.; Lerner, R. A. Science 1995, 270, 1775. Tagaki, W.; Guthrie, J. P.; Westheimer, F. H. Biochemistry 1968, 7, 905. Gilbert, H. F., III; O’Leary, M. H. Biochemistry 1975, 14, 5194. Gilbert, H. F., III; O’Leary, M. H. Biochem. Biophys. Acta 1977, 483, 79. Otwell, H. B.; Cipollo, K. L.; Dunlap, R. B. Biochem. Biophys. Acta 1979, 568, 297. Guidinger, P. F.; Nowak, T. Biochemistry 1991, 30, 8851. Hoffmann, T.; Zhong, G.; List, B.; Shabat, D.; Anderson, J.; Gramatikova, S.; Lerner, R. A.; Barbas, C. F., III J. Am. Chem. Soc. 1998, 120, 2768. Zhong, G.; Shabat, D.; List, B.; Anderson, J.; Sinha, R. A.; Lerner, R. A.; Barbas, C. F., III Angew. Chem., Int. Ed. 1998, 37, 2481. List, B.; Shabat, D.; Zhong, G.; Turner, J. M.; Li, A.; Bui, T.; Anderson, J.; Lerner, R. A.; Barbas, C. F., III J. Am. Chem. Soc. 1999, 121, 7283. Barbas, C. F., III; Heine, A.; Zhong, G.; Hoffmann, T.; Gramatikova, S.; Bjornestsdt, R.; List, B.; Anderson, J.; Stura, E. A.; Wilson, I. A.; Lerner, R. A. Science 1997, 278, 2085. Zhong, G.; Lerner, R. A.; Barbas, C. F., III Angew. Chem. Int. Ed. 1999, 38, 3738.

307

308

6 Antibody-catalyzed Aldol Reactions 13 Shulman, H.; Makarov, C.; Ogawa, A. K.; Romesberg, F.;

Keinan, E. J. Am. Chem. Soc. 2000, 122, 10743. 14 Fujii, I.; Tanaka, F.; Miyashita, H.; Tanimura, R.;

15

16

17 18 19

20 21 22

23 24 25 26

27 28

29

30

Kinoshita, K. J. Am. Chem. Soc. 1995, 117, 6199. Tanaka, F.; Kinoshita, K.; Tanimura, R.; Fujii, I. J. Am. Chem. Soc. 1996, 118, 2332. Patten, P. A.; Gray, N. S.; Yang, P. L.; Marks, C. B.; Wedemayer, G. J.; Boniface, J. J.; Stevens, R. C.; Schultz, P. G. Science, 1996, 271, 1086. Wedemayer, G. J.; Pattern, P. A.; Wang, L. H.; Schultz, P. G.; Stevens, R. C. Science 1997, 276, 1665. Yin, J.; Mundorff, E. C.; Yang, P. L.; Wendt, K. U.; Hanway, D.; Stevens, R. C.; Schultz, P. G. Biochemistry 2001, 40, 10764. Zhong, G.; Hoffmann, T.; Lerner, R. A.; Danishefsky, S.; Barbas, C. F., III J. Am. Chem. Soc. 1997, 119, 8131. List, B.; Lerner, R. A.; Barbas, C. F., III Org. Lett. 1999, 1, 59 and 353. Karlstrom, A.; Zhong, G.; Rader, C.; Larsen, N. A.; Heine, A.; Fuller, R.; List, B.; Tanaka, F.; Wilson, I. A.; Barbas, C. F., III; Lerner, R. A. Proc. Natl. Acad. Sci. U.S.A. 2000, 97, 3878. Guthrie, J. P.; Cossar, J.; Taylor, K. F. Can. J. Chem. 1984, 62, 1958. Guthrie, J. P.; Wang, X.-P. Can. J. Chem. 1992, 70, 1055. Maggiotti, V.; Resmini, M.; Gouverneur, V. Angew. Chem. Int. Ed. 2002, 41, 1012. Maggiotti, V.; Wong, J.-B.; Razet, R.; Cowley, A. R.; Gouverneur, V. Tetrahedron: Asymmetry 2002, 13, 1789. Turner, J. M.; Bui, T.; Lerner, R. A.; Barbas, C. F., III; List, B. Chem. Eur. J. 2000, 6, 2772. Sinha, S.; Barbas, C. F., III; Lerner, R. A. Proc. Natl. Acad. Sci. U.S.A. 1998, 95, 14603. Sinha, S.; Sun, J.; Miller, G.; Barbas, C. F., III; Lerner, R. A. Org. Lett. 1999, 1, 1623. Shinha, S. C.; Sun, J.; Miller, G. P.; Wartmann, M.; Lerner, R. A. Chem. Eur. J. 2001, 7, 1691. List, B., Shabat, D., Barbas III, C. F., and Lerner, R. A. Chem. Eur. J. 1998, 4, 881. Shabat, D.; List, B.; Lerner, R. A.; Barbas, C. F., III Tetrahedron Lett. 1999, 40, 1437. Niculesco-Duvanz, I.; Springer, C. J. Adv. Drug Delivery Rev. 1997, 26, 151. Wenthworth, P.; Datta, A.; Blakey, D.; Boyle, T.; Partridge, L. J.; Blackburn, G. M. Proc. Natl. Acad. Sci. U.S.A. 1996, 93, 799. Shabat, D.; Rader, C.; List, B.; Lerner, R. A.; Barbas, C. F., III Proc. Natl. Acad. Sci. U.S.A. 1999, 96, 6925. Shabat, D.; Lode, H. N.; Pertl, U.; Reisfeld, R. A.; Rader, C.; Lerner, R. A.; Barbas, C. F., III Proc. Natl. Acad. Sci. U.S.A. 2001, 98, 7528. Worrall, D. S.; McDunn, J. E.; List, B.; Reichart, D.; Hevener, A.; Gustafson, T.; Barbas, C. F., III; Lerner, R. A.; Olefsky, J. M. Proc. Natl. Acad. Sci. U.S.A. 2001, 98, 13514. Cordova, A.; Janda, K. D. J. Am. Chem. Soc. 2001, 123, 8248.

References 31 Gopin, A.; Pessah, N.; Shamis, M.; Rader, C.; Shabat, D.

Angew. Chem. Int. Ed. 2003, 42, 327. 32 Bjornestedt, R.; Zhong, G.; Lerner, R. A.; Barbas, C. F., III

J. Am. Chem. Soc. 1996, 118, 11720. 33 Lin, C.-H.; Hoffman, T. Z.; Wirshing, P.; Barbas, C. F., III;

34 35 36

37 38 39 40 41

42 43

44 45

46 47

48 49 50

51 52

Janda, K. D.; Lerner, R. A. Proc. Natl. Acad. Sci. U.S.A. 1997, 94, 11773. Shulman, A.; Sitry, D.; Shulman, H.; Keinan, E. Chem. Eur. J. 2002, 8, 229. Tanaka, F.; Lerner, R. A.; Barbas, C. F., III Chem. Commun. 1999, 1383. Nicholas, K. M.; Wentworth, P. Jr.; Harwig, C. W.; Wentworth, A. D.; Shafton, A.; Janda, K. D. Proc. Natl. Acad. Sci. U.S.A. 2002, 99, 2648. Satchi-Fainaro, R.; Wrasidlo, W.; Lode, H. N.; Shabat, D. Bioorg. Med. Chem. 2002, 10, 3023. Finn, M. G.; Lerner, R. A.; Barbas, C. F., III J. Am. Chem. Soc. 1998, 120, 2963. James, L. C.; Tawﬁk, D. S. Protein Science 2001, 10, 2600. Tanaka, F.; Thayumanavan, R.; Barbas, C. F., III J. Am. Chem. Soc. 2003, 125, 8523. Sakthivel, K.; Notz, W.; Bui, T.; Barbas, C. F., III J. Am. Chem. Soc. 2001, 123, 5260. Cordova, A.; Watanabe, S.; Tanaka, F.; Notz, W.; Barbas, C. F., III J. Am. Chem. Soc. 2002, 124, 1866. Pidathala, C.; Hoang, L.; Vignola, N.; List, B. Angew. Chem. Int. Ed. 2003, 42, 2785. List, B.; Barbas, C. F., III; Lerner, R. A. Proc. Natl. Sci. Acad. U.S.A. 1998, 95, 15351. Jourdain, N.; Carlon, R. P.; Reymond, J.-L. Tetrahedron Lett. 1998, 39, 9415. Carlon, R. P.; Jourdain, N.; Reymond, J.-L. Chem. Eur. J. 2000, 6, 4154. Tanaka, F.; Kerwin, L.; Kubitz, D.; Lerner, R. A.; Barbas, C. F., III Bioorg. Med. Chem. Lett. 2001, 11, 2983. The GenBank accession numbers are following: 33F12 VL, AF242212; 33F12 VH, AF242213; 38C2 VL, AF242214; 38C2 VH, AF242215; 40F12 VL, AF242216; 40F12 VH, AF242217; 42F1 VL, AF242218; 42F1 VH, AF242219. Westheimer, F. H. Tetrahedron 1995, 51, 3. Littlechild, J. A.; Watson, H. C. Trends Biochem. Sci. 1993, 18, 36. Morris, A. J.; Davenport, R. C.; Tolan, D. R. Protein Engineering 1996, 9, 61. Shulman, H.; Keinan, E. Bioorg. Med. Chem. Lett. 1999, 9, 1745. Tanaka, F.; Lerner, R. A.; Barbas, C. F., III J. Am. Chem. Soc. 2000, 122, 4835. Barbas III, C. F., Burton, D. R., Scott, J. K., and Silverman, G. J. Eds. Phage Display: A Laboratory Manual. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, New York, 2001. Tanaka, F.; Fuller, R.; Shim, H.; Lerner, R. A.; Barbas, C. F., III J. Mol. Biol. 2004, 335, 1007. Tanaka, F.; Barbas, C. F., III Chem Commun. 2001, 769. Tanaka, F.; Barbas, C. F., III J. Am. Chem. Soc. 2002, 124, 3510.

309

310

6 Antibody-catalyzed Aldol Reactions 53 Gildersleeve, J.; Janes, J.; Ulrich, H.; Yang, P.; Turner, J.;

54

55 56 57

Barbas, C.; Schultz, P. Bioorg. Med. Chem. Lett. 2002, 12, 1691 and 2789. Reymond, J.-L.; Chen, Y. Tetrahedron Lett. 1995, 36, 2575. Reymond, J.-L.; Chen, Y. J. Org. Chem. 1995, 60, 6970. Reymond, J.-L. Angew. Chem. Int. Ed. 1995, 34, 2285. Tanaka, F.; Oda, M.; Fujii, I. Tetrahedron Lett. 1998, 39, 5057. Barbas, C. F., III; Rader, C.; Segal, D. J.; List, B.; Turner, J. M. Advances in Protein Chemistry 2001, 55, 317. Tanaka, F.; Barbas, C. F., III J. Immunol. Methods 2002, 269, 67.

311

7

The Aldol Reaction in Natural Product Synthesis: The Epothilone Story Dieter Schinzer Dedicated to Clayton H. Heathcock, one of the pioneers of modern aldol reactions who ﬁgured out many of the import stereochemical principles of this fascinating chemistry

7.1

History of Epothilones: Biological Source, Isolation, and Structural Elucidation

The epothilones, a new class of macrocyclic compounds which show cytotoxic activity, have been isolated as secondary metabolites by Ho¨ﬂe and Reichenbach from myxo bacteria [1]. These are so-called gliding bacteria which cruise on slime tracks forming swarms. They grow preferentially on rodded material, like dung. Their morphogenetic potential is quite interesting, forming cylindrical vegetative cells, which under starvation conditions turn into fruiting bodies with solid walls, that further diﬀerentiate into myxo spores [2]. Besides the epothilones – which were isolated from a strain of Sorangium cellulosum – many other structural diverse natural products have been isolated from myxo bacteria [3].

7.2

History of Epothilones: The Total Synthesis Race

The publication of the structures of epothilone A and B in 1996 by Ho¨ﬂe et al. [4] in connection with the biological data of the tubulin assay was of great interest for the chemical community. The epothilones were identiﬁed by the Merck group in 1995 as tubulin-polymerizing natural products as a single active structural type out of a compound library of more than 60.000 molecules [5]. The cytotoxic eﬀect of Taxol2 via binding to tubulin and thus stabilizing the assembly, combined with the scope of its clinical application, prompted major activities around the world. Taxol2 is far from being an ideal drug: low water solubility requires formulation vehicles, such as Modern Aldol Reactions. Vol. 1: Enolates, Organocatalysis, Biocatalysis and Natural Product Synthesis. Edited by Rainer Mahrwald Copyright 8 2004 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 3-527-30714-1

312

7 The Aldol Reaction in Natural Product Synthesis: The Epothilone Story

Cremophore2 which create allergic side eﬀects and other risks [6]. Two different types of resistance arise during cancer therapy with Taxol2 ; an overexpression of the P-glycoprotein (P-gp) eﬄux system and tubulin mutation which both cause major drawbacks. Furthermore, the remarkable eﬃcacy of epothilone against MDR (multiple drug resistante) cell lines makes these compounds superior in comparison to known anticancer agents [7]. This started a race for the ﬁrst total synthesis in which many world leading chemists got involved. In addition, biologists, medicinal chemists, and clinicians started many eﬀorts along these lines. Besides various total syntheses, many partial solutions, a large variety of synthetic analogues have been synthesized and have created a tremendous number of publications and patent applications around the world [8]. Finally, diﬀerent large (and small) pharmaceutical companies took over the project and started clinical trials of epothilones or of their synthetic analogues in order to replace Taxol2 as a therapeutic anticancer agent [9]. The ﬁrst total synthesis was published by Danishefsky et al. [10], shortly followed by independent routes from Nicolaou et al. [11] and Schinzer et al. [12].

O S N

R 12

13 8

15

OH

7

O 1

3

5

OH

O

6

O

1 Scheme 7.1

Epothilone A: R¼H; epothilone B: R¼Me

7.2.1

Diﬀerent Strategies with Aldol Reactions: The Danishefsky Synthesis of Epothilone A Relying on Intramolecular Aldol Reaction

The ﬁrst total synthesis of epothilone A combines the two halves of the molecule via a B-alkyl Suzuki coupling [13] yielding the desired cyclization precursor. The critical cyclization was achieved by an intramolecular aldol reaction, utilizing KHMDS as base providing the natural conﬁguration with a 6:1 ratio, 51% yield. The selectivity of the macro condensation was best at 78 C and work-up of the potassium alkoxide at 0 C. Protonation at lower temperature yielded in more of (R)-conﬁgured material. Under special conditions, even the (R)isomer predominated over the ‘‘natural’’ conﬁguration. The reason for this surprising behavior is still unknown and under investigation.

7.2 History of Epothilones: The Total Synthesis Race

313

OTBS OTIPS

S

I

8

15

+

6

1. 9-BBN, THF, RT;

CH(OMe)2

7

N

PdCl2(dppf)2, CsCO3,

OAc

Ph3As, H2O, DMF; 71%

2

3

2. TsOH 3. KHMDS, THF, - 78 oC 51%

12

12

S

S

13

13

8 8

N 15

N

OTBS

15

7

O

5

O OK

O

5

1

6

3

1

3

OH

O

OTIPS

5

4 Scheme 7.2

First total synthesis of epothilone A by Danishefsky et al.

7.2.2

Diﬀerent Strategies with Aldol Reactions: The Nicolaou Synthesis of Epothilone A Using an Unselective Aldol Reaction

The Nicolaou strategy is based on RCM [14] as the ﬁnal ring closing step in order to establish the 16-membered macrolide. The linear subunit of C1aC12 was obtained by an intermolecular aldol reaction, in which an a-chiral aldehyde was coupled with an achiral enolate, yielding the desired aldol product as a 2:3 ratio of diastereomers 8 and 9. 2.3 eq. LDA, THF, - 78 oC

HO +

O

6

O

8

O

HO

7

+

O

O

OH

7

HO 6

O

8 Scheme 7.3

Model study for the total synthesis of epothilone A by Nicolaou et al.

O 9

OTBS

7

8

OH

6

OTIPS

7 The Aldol Reaction in Natural Product Synthesis: The Epothilone Story

314

2.3 eq. LDA, THF, - 78 oC

HO

3

+

O

8

OTBS O

O 7

10 HO

HO

7

3

O

OTBS O

+

8

6

O

OH

11

OTBS O

OH

12 S

1. EDC, 4-DMAP, 12 h; 52%

11 + 15

N

2. RuCl2(=CHPh)(PCy3)2, CH2Cl2, 20 h 79%

OH

13

12

S

13 8

N

15

OTBS

7

O 1

3

5 6

O

OTBS O 14

Scheme 7.4

Second total synthesis of epothilone A by Nicolaou et al.

This result is not surprising because only one chiral element was used in that reaction. A second chiral element should improve the selectivity. For that reason Nicolaou introduced a chiral enolate, which is shown in Scheme 7.4. In his second synthesis, Nicolaou used a chiral enolate of type 10 which contained the (S)-conﬁguration at C3 and the required oxidation level at C1. However, again almost no selectivity was observed in the aldol coupling [15]. 7.2.3

Diﬀerent Strategies with Aldol Reactions: The Schinzer Synthesis of Epothilone A with Complete Stereocontrol in the Aldol Reaction

The Schinzer synthesis of epothilone A is also based on a convergent strategy using three key fragments [16]. The 16-membered macrolide is closed via RCM to provide the key intermediate for the ﬁnal epoxidation.

7.2 History of Epothilones: The Total Synthesis Race

S N olefin metathesis, OH

13

O

epoxidation 7 O

S

12

13 8

N

OH aldol reaction

O esterification

7

15 1

3

O 1

O

5

OH

O

O

O

6

15 Scheme 7.5

Third total synthesis of epothilone A by Schinzer et al.

As shown before, the major stereochemical issue in epothilone chemistry is the absolute stereocontrol of the triad C6aC7aC8. In the Schinzer synthesis, a six-membered chiral acetonide of type 15 containing the (S)conﬁguration at C3 was used. Addition of a-chiral aldehyde 7 (X ¼ C) to the preformed enolate (generated with LDA in THF at 78 C) resulted in the formation of the (6R,7S)-diastereomer 16 as a the major isomer in ratio of 25:1 in 76% yield [17]. In a synthesis of oxa-epothilones, Schinzer used aldehyde 7 (X ¼ O) in a diastereoselective aldol reaction which exclusively yielded aldol product 16 (X ¼ O) [18]. Schinzer then performed the same reaction with a-chiral aldehyde 17. Remarkably, this reaction provided again a single isomer 18 with the ‘‘natural’’ epothilone conﬁguration in 49% yield [19]. The fully functionalized aldehyde 19 from Schinzer’s epothilone B synthesis was also used in the same aldol process. However, a 10:1 mixture of diastereomers was obtained with the ‘‘correct’’ stereochemistry as the major product 20 [20]. In the meantime, Schinzer’s chiral enolate 15 has been used extensively by other groups in academia [21] and in the pharmaceutical industry to synthesize a large number of analogues [22]. Most of the examples with highly func-

315

316

7 The Aldol Reaction in Natural Product Synthesis: The Epothilone Story

X

3

+

O

O

6

O

O

O

3

LDA, THF, - 78 oC

8

15

O

7

X

7 8

OH

O 16

X = C; O 3

7

LDA, THF, - 78 oC

15

+

O

OTBS

8

O

O

O

8

OTBS

OH

18

17

S 15

6

LDA, THF, - 78 oC

8

+

15

N

O

OTBS 19

7

3 6

O

O

O

S

8 15

OH

N 20

OTBS

Scheme 7.6

Selective double stereodiﬀerentiating aldol reactions

tionalized chiral aldehydes generated single isomers in high chemical yields. These outstanding results are connected to the nature of the acetonide, because the lithium counter ion attached to the enolate oxygen can chelate to one oxygen of the acetonide, as shown in Scheme 7.7. The chelated structure 21 was calculated with force ﬁeld (MMþ) and quantum mechanics (PM3) methods, both methods showed minima in favor of the chelated enolate 21 [19]. Thus, with these results he was able to explain the stereochemical results. There are two aspects to understanding these remarkable observations. The simple diastereoselectivity of the aldol reaction can be explained based on Heathcock’s studies [23], in which the enolate geometry is governed by the bulkiness of the group next to the ketone (in the Schinzer case a modiﬁed tbutyl group bearing an extra oxygen). This should result in the generation of a (Z)-lithium enolate, further stabilized by the oxygen at C3, forming a

7.2 History of Epothilones: The Total Synthesis Race

LDA, THF, - 78 oC

3

3

R

+ O

O

O

O

O

O

Li

8

O

Z-Enolate

15

22

21

R

H O

O Li

7

3

H

O H

6

H O

S

O

O

O

R 8

OH

24

S

317

Zimmerman-Traxler TS

23 Scheme 7.7

Chelated nucleophile 21 with Zimmerman-Traxler TS.

rigid bicyclic structure 21. The diastereofacial selectivity to explain the stereochemical outcome in aldol condensations with a-chiral aldehydes was ﬁrst proposed by Cram [24] and later on by Felkin [25]. Both models give the same products even though totally diﬀerent reactive conformations are used. In Schinzer’s case, the aldol addition occurred with the opposite sense of diastereofacial addition to the theoretical models predicted. Intrigued by these results the Schinzer group decided to study this particular aldol reaction in detail. First, they conducted experiments with socalled simple aldehydes containing no chiral center to check the inherent selectivity of the chiral enolate. The reaction generated only the two syn-type diastereomeric aldol products 26 and 27, as shown in table 1 and no anti-

Tab. 7.1

Aldol reactions with achiral aldehydes R

26

:

27

Yield [%]

Ph Pr i-Pr t-Bu 1-Heptenyl

4 5 5 3 5

: : : : :

1 1 1 1 1

81 63 59 49 74

318

R 15 +

7 The Aldol Reaction in Natural Product Synthesis: The Epothilone Story

7

3

LDA, THF, - 78 oC

R

R

6

O

O

O

+

O

25

O

OH

O

O

26

OH

27

Scheme 7.8

Inherent selectivity of nucleophile 15

aldol product was observed at all. In all cases studied, the 6(R),7(S)-isomer 26 was the predominant diastereomer. The inherent selectivity of the chiral enolate is in the range of about 4:1 [26]. Next, the double-stereodiﬀerentiating [27] aldol process with a-chiral aldehydes was examined, seperate experiments using both enantiomeric forms of phenyl propionaldehyde were undertaken. Use of (S)-phenylpropionaldehyde 28 gave two diastereomers 29 and 30 in a ratio of 2.5:1 and 68% yield. The major one is still the anti-Cram/Felkin compound 29, which is quite unusual because 28 usually provides high selectivity in favor of the Cram/ Felkin adduct 30 [28].

LDA, THF, - 78 oC

15 +

Ph

8

7

3 6

O

O

O

8

Ph

+

Ph

O

OH

O

O

OH

30

29

28

O

Scheme 7.9

Aldol reaction with (S)-phenylpropionaldehyde

This is a very good example where the inherent selectivity of the enolate overrides the inherent selectivity of the aldehyde with the 6(R),7(S)conﬁguration resulting from double stereodiﬀerentiation using the aldehyde with the (S)-conﬁguration. On the other hand, the same reaction with (R)phenylpropionaldehyde 31 yielded two diastereomers 32 and 33 in a ratio of 40:1 and 77% yield as shown in Scheme 7.10 [26].

3

LDA, THF, - 78 oC

15 +

8

Ph

7 6

O

O

O

8

OH

Ph

+

Ph O

O

O

O 31

32

Scheme 7.10

Aldol reaction with (R)-phenylpropionaldehyde

33

OH

7.3 Model Study via Chelation Control in the Aldol Reaction by Kalesse

319

8

O

O

O

6

O

LDA, THF, - 78 oC

3

7

3

35

O

O

36

8

OH

O 7

3 6

34

O 8

O

Ph

O

O

8

OH 37

28

Scheme 7.11

Double stereodiﬀerentiating aldol reactions with (R)-nucleophile 34

The major isomer was the Cram/Felkin product 32 with the 6(R),7(S),8(R)-conﬁguration which clearly indicated that in the (S),(R)combination of chirality both eﬀects work in the same direction and produce a matching case with very high selectivity. The same holds for the use of the (R)-enolate system as shown in Scheme 7.11 [26]. In contrast, a-chiral aldehydes of type 7 show the opposite behavior, favoring the anti-Cram/Felkin product as the major isomer. In these cases, the matched case is the (S),(S)- or (R),(R)-combination of chirality beneﬁting of both chiral elements. These correlation studies indicate a dominating inﬂuence of the chiral enolate 15 versus the chirality of the aldehyde. In the case of (S)-phenylpropionaldehyde 28 the chirality of enolate 15 overrides the directing effect of the aldehyde chirality, still producing a 2.5:1 ratio in favor of the 6(R),7(S),8(R)-stereochemical triad. The use of the (R)-phenylpropionaldehyde 31 forms exclusively the 6(R),7(S),8(R)-stereo chemical triad in a 40:1 ratio. In this case the matched pair is the (S)/(R)-combination of chirality. The (R)-enantiomer of the enolate reverses the stereochemical outcome. In connection with a synthesis of 6-desmethyl-epothilones Schinzer used methyl ketone 38 instead of the ethyl ketone 15 to study the inﬂuence of the missing methyl group. As seen in Scheme 7.12, the selectivity dropped to 1.7:1 still favoring diastereomer 39 with the (S)-conﬁguration at C7 [29]. The lower selectivity can be easily explained by the lack of 1,3-diaxial interactions in the chair-like Zimmerman-Traxler transition state model [30].

7.3

Model Study via Chelation Control in the Aldol Reaction by Kalesse

Kalesse et al. [31] described an aldol reaction where a silyloxy protected ethyl ketone 41 is coupled to an b-oxa-aldehyde 41 based on Roche acid [32] (Scheme

Ph

320

7 The Aldol Reaction in Natural Product Synthesis: The Epothilone Story LDA, THF, - 78 oC 3

+

O

O

O

67%

7

38

3

7

+

8

O

8

O

O

O

O

OH

O

O

OH 40

39 Scheme 7.12

Double stereodiﬀerentiating aldol reactions with methyl ketone 37

7.13). This is a very useful example in connection with the model studies to approach the epothilones. As a result of a chelation-controlled aldol reaction exclusive formation of compound 43 with the desired 6(R),7(S),8(R)-triad was observed. The structure has been conﬁrmed by X-ray analysis after transformation to a crystalline six-membered lactone. In their macro-lactonizsation strategy to synthesize epothilone B Nicolaou et al. used the same ketone 41 in an aldol reaction with aldehyde 19. The diastereoselectivity in the aldol addition was only moderate, giving a ratio of 3:1 in favor of the ‘‘natural’’ epothilone conﬁguration [33].

OBzl

3

+

LDA, THF, - 78 oC

8

O

OTBS OTBS O

42

41

3

OBzl

7 6

8

OH

OTBS OTBS O 43

Scheme 7.13

Chelation-controlled aldol reaction by Kalesse et al.

7.3.1

Diﬀerent Aldol Strategies: Mulzer’s Total Syntheses of Epothilones B and D

A series of similar aldol reactions were carried out in Mulzer’s lab. In his study towards the total synthesis of epothilone B and D Mulzer et al. used ﬁrst an achiral ethyl ketone of type 44 in an aldol reaction with the known (S)-aldehyde 45. In this particular case only the chiral element at 44 directed the addition giving a 4:1 mixture of diastereomers 46 and 47. The major di-

7.3 Model Study via Chelation Control in the Aldol Reaction by Kalesse

LDA, THF, - 78 oC

O

OPMB

7

OPMB +

321

8

6

8

+

OH

O

O

44

46

45

OPMB OH

O 47 Scheme 7.14

Chelation-controlled aldol reaction by Mulzer et al.

astereomer with the (R)-conﬁguration at C6 and the (S)-conﬁguration at C7 is formed as a result of a chelated Cram-type aldol reaction. In a later study Mulzer presented examples of double stereodiﬀerentiating aldol reactions with (S)-C3 protected nucleophiles [35]. The same double TBSO-protected (S)-ethyl ketone 41 used before by Kalesse et al. gave 6:1 ratio in an aldol reaction with an a-(S)-chiral aldehyde 48 as a result of matching chirality (70% yield). Again, the major isomer 49 had the ‘‘natural’’ epothilone conﬁguration at C6 and C7. A further modiﬁcation of the nucleophile 50 yielded the almost exclusive formation (19:1; 92% yield) of the ‘‘correct’’ diastereomer 52 in the presence of the sensitive epoxide function [36]. Both examples shown in Schemes 7.15 and 7.16 are aldol reactions with

O O

S

3

12

+

15

N

OTBS OTBS O

O

OTBS 48

41

O 6

OTBS OTBS O

O

7

3

12

8

15

OH 49

LDA, THF, - 78 oC

8

OTBS

Scheme 7.15

Double stereodiﬀerentiating aldol reactions by Mulzer et al.

S N

322

7 The Aldol Reaction in Natural Product Synthesis: The Epothilone Story

O S

3

13

+

LDA, THF, - 78 oC

8

O

15

N

OTBS O

12

OTBS 51

50

O

3

7 6

OTBS O

12

8

13 15

OH 52

S N

OTBS

Scheme 7.16

Double stereodiﬀerentiating aldol reactions in the presence of the epoxide functionality by Mulzer et al.

(S)-chiral nucleophiles protected at C3. The stereochemical outcome can most likely be explained by the cyclic transition states discussed in Scheme 7.7. All types of nucleophiles protected at C3 show a predominant eﬀect in driving the aldol reaction in the direction of favoring the anti-Cram/Felkin product. This characteristic seems to be independent of the nature of the aldehyde. The best selectivities were obtained with nucleophile 50 and even superior to that, nucleophile 15 protected as an acetonide. The ﬁnal part of this chapter will focus on a special type of aldehyde where a long range effect improves the stereochemical outcome of the aldol reaction.

7.4

Long-range Structural Eﬀects on the Stereochemistry of Aldol Reactions

In a study connected with an improved synthesis of epothilone B and deoxyepothilone F, Danishefsky et al. conducted aldol reactions with functionalized ethyl ketones and an g,d-unsaturated aldehyde. Two major studies were carried out in Danishefsky’s lab: One with an achiral system [37] and one with a chiral enolate bearing the correct oxidation level at C1 and a TBSprotected alcohol at C3 [38]. First, in a model study Danishefsky et al. investigated the reaction of a new type of enolate 53 with several aldehydes. The typical Cram/Felkindirecting (S)-phenyl-propionaldehyde 28 gave the expected all-syn orientated compound 54 (Scheme 7.17, 54/55 ¼ 11:1).

7.4 Long-range Structural Eﬀects on the Stereochemistry of Aldol Reactions LDA, THF, - 78 oC

t-BuO +

O

OTES O

8

t-BuO

Ph

Ph O

O

53

OTES O

+

OH

54

28

t-BuO

323

7 6

O

OTES O

8

Ph

OH

55

Scheme 7.17

Model study from Danishefsky et al.

A quite interesting trend was observed with other a-chiral aldehydes. (S)Methyl pentane aldehyde 56 (R ¼ ethyl) behaved only slightly more selectively generating a diastereomeric mixture 57 and 58 of 1:1.3. Unsaturated groups improved the selectivity to about 1:2–1:5.5 (see Scheme 7.18). Danishefsky explained these results by a special long range eﬀect connected to the distance between the formyl and the terminal oleﬁn group via a nonbonding interaction of the carbonyl group of the enolate and the oleﬁn group. The maximum eﬀect can be obtained with a g,d-unsaturated aldehyde 56 (R ¼ allyl; 1:5.5).

t-BuO

R +

O

OTES O

7

OTES O

OTES O 57

R 8

6

O

R + O

56

t-BuO

t-BuO

8

O

53

LDA, THF, - 78 oC

OH

58 R = ethyl, Ph, allyl, dimethylvinyl, OBzl

Scheme 7.18

Aldol reactions of 53 with various aldehydes

In a further series of experiments Danishefsky employed chiral ketones of type 59. Both enantiomers were available with high optical purity and could be involved in investigations in the double stereodiﬀerentiating aldol reaction. However, the lithium anion of 59 (R ¼ TBS) could not be eﬀected in useful yield due to the sensitivity of the b-silyloxy system to elimination. The less basic titanium enolate of 59 gave mixtures of diastereomers in moderate yields. The stereochemical outcome of these reactions showed that the conﬁguration at C3 rather than C8 had a larger eﬀect on the newly

OH

324

7 The Aldol Reaction in Natural Product Synthesis: The Epothilone Story

t-BuO

LDA, THF, - 78 oC

3

+

O

OR

O

O 60

59

t-BuO

3

t-BuO

6

O

8

OR 61

O

7 8

OH

+ O

OR

O

OH

62

Scheme 7.19

Double stereodiﬀerentiating aldol reaction by Danishefsky et al.

formed centers at C6 and C7. However, the selectivity of about 3.5:1 was quite low (even lower than the result shown in Scheme 7.18), indicating that neither of the enantiomers of 59 and 60 beneﬁted from matching chirality. The diﬃculty in the lithium series of these aldol reactions caused Danishefsky to investigate the lithium dianion 59 (R ¼ H) as the nucleophile. Indeed, these reactions did not induce b-elimination and provided quite unexpected results with a high level of stereoselectivity [38]. When the lithium dianion of (S)-59 (R ¼ H) was treated with (S)-60 a 2:3 mixture of diastereomers 61 and 62 was obtained in 53% yield, corresponding to the (S)-conﬁguration at C6 and the (R)-conﬁguration at C7 as the major diastereomer. Thus, the major isomer 62 corresponded to the ‘‘unnatural’’ stereochemistry of the epothilones. The minor isomer presented the required ‘‘natural’’ conﬁguration of the epothilones. When the same reaction was then performed with the lithium dianion of (R)-63 (R ¼ H) and the (S)-conﬁgurated aldehyde 60 it gave rise to a single diastereomer 64 that contained the ‘‘natural’’ conﬁguration of the epothilones at C6 and C7. This is a result of the matching chirality of the two chiral reaction partners, but represented a major drawback because the center at C3 proved to be (R)conﬁgurated. Therefore, it could not be used as a precursor for epothilone syntheses without a troublesome inversion at C3. Despite that, these undesired results from the dianion aldol reactions can be used to understand some principles of the double stereodiﬀerentiating reactions. The C3 protected series, such as Schinzer-type aldol reactions described before required the (S)-conﬁguration at C3 to establish the (R)conﬁguration at C6 and the (S)-conﬁguration at C7 with a similar a-chiral (S)-aldehyde. The matching chirality in the protected series corresponds to the mismatched case in the unprotected series of these double stereodiﬀerentiating aldol reactions. This disparity could be a result of diﬀerent transition states. Danishefsky proposed ﬁve factors governing such aldol reactions: (i) the chair-like transition state leading to a syn-aldol, (ii) chelation of the lithium counter ion by the b-oxygen at C3, (iii) a syn-relationship of the a-proton of the aldehyde and the methyl group at C6 of the enolate, (iv) anti-attack of

7.4 Long-range Structural Eﬀects on the Stereochemistry of Aldol Reactions 3

t-BuO

LDA, THF, - 78 oC +

O

OR

8

O

O 60

63

3

t-BuO

7 6

O

325

OR

8

OH

O

64 Scheme 7.20

Double stereodiﬀerentiating aldol reaction by Danishefsky et al.

H R

O R

O Li O

R

H

O

Li

R

L Me

O Li O

R

O O Li O

O Li O H

H

(i) chair

(ii) chelation

(iii) avoid syn

(iv) aldehde

(v) pentenyl

pentane interactions

attacks anti to R

aldehyde effect

Scheme 7.21

Factors governing the aldol reaction

the aldehyde versus the large group of the enolate, and (v) the special g,dunsaturated aldehyde eﬀect [38]. As a result of all these energy-lowering factors, Danishefsky came up with the following transition state model leading to 64 [38]. A ﬁnal experiment was designed to check the possibility of an internal kinetic resolution. For this purpose, a racemic mixture of the nucleophile

Ot-Bu H O O

t-BuO

3

7 6

O Li O H

O

OR 64

65 Scheme 7.22

Transition state of Danishefsky-type aldol reaction

O

8

OH

326

7 The Aldol Reaction in Natural Product Synthesis: The Epothilone Story

59 (R ¼ H) was reacted with homochiral (S)-60. The reaction generated only diastereomer 64 which supported the concept of matching and mismatching issues at the kinetic level [38]. 7.5

Summary and Conclusion

This short chapter gave some insight in the troublesome and quite complicated aldol reaction in the context of epothilone chemistry. All the examples presented were related to the major stereochemical issue around this fascinating class of natural products that is the absolute control of the stereochemical triad C6aC7aC8. Many groups have contributed to that particular stereochemical problem because of the importance of this class of natural products an and optimal solution to this problem obviously is highly desirable. In the meantime, many analogues of the natural epothilones have been synthesized by academic groups and pharmaceutical companies – some of these are even more potent in biological systems than the natural compounds. This could lead to important developments for new anti-cancer drugs in the near future. To our knowledge, the simple acetonide-protected ethyl ketone is the backbone of many industry-based analog programs to achieve the central aldol coupling. This ketone is known as a simple and robust coupling partner yielding exclusively the desired stereochemical outcome with a large number of chiral aldehydes – even on multi kilogram scale. Typical procedure [17]: Synthesis of (4R,5S,6S,4OS)-2-(2,2-dimethyl-1,3-dioxan-4-yl)-5-hydroxy-2,4,6-trimethyl-10-undecen-3-one 16: A solution of ethyl ketone 15 (1.17 g, 5.45 mmol) in THF (1.0 mL) was added to a freshly prepared solution of LDA [nBuLi (3.34 mL, 1.6 m solution in hexanes, 5.35 mmol, 0.98 equiv) was added to a solution of diisopropylamine (749 mL, 5.35 mmol) in THF (4.0 mL) at 0 C] dropwise at 78 C. The solution was stirred for 1 h at 78 C. Aldehyde 7 (688 mg, 5.45 mmol, 1.0 equiv) was added dropwise and stirring was continued for 45 min at 78 C. The reaction mixture was quenched by dropwise addition of saturated aqueous NH4 Cl solution at 78 C. The organic layer was separated and the aqueous layer was extracted with Et2 O. The combined extracts were dried over MgSO4 and concentrated in vacuo. Flash chromatography (pentane/Et2 O ¼ 10:1) of the residue aﬀorded antiCram aldol product 16 (1.36 g, 73%) and Cram aldol product (57 mg, 3%) as colorless oils. References ¨ ﬂe, N. Bedorf, H. Reichenbach, (GBF), DE-4138042, 1 G. Ho 1993 [Chem. Abstr. 1993, 120, 52841].

References ¨ ﬂe, H. Irschik, H. 2 K. Gerth, N. Bedorf, G. Ho Reichenbach, J. Antibiot. 1996, 49, 560–563. ¨ ﬂe, in: Biotechnical Advances, 1993, 3 H. Reichenbach, G. Ho 11, 219–277. ¨ ﬂe, N. Bedorf, H. Steinmetz, D. Schomburg, K. 4 G. Ho

5

6 7 8

9

10

11

12

13 14

15

16 17 18 19 20

21

Gerth, H. Reichenbach, Angew. Chem. 1996, 108, 1671– 1673; Angew. Chem. Int. Ed. Engl. 1996, 35, 1567–1569. a) D. M. Bollag, P. A. McQueney, J. Zhu, O. Jensens, L. Koupal, J. Liesch, M. E. Goetz, C. Lazarides, M. Woods, Cancer Res. 1995, 55, 2325–2333; b) R. J. Kowalski, P. Giannakakou, E. Hamel, J. Biol. Chem. 1997, 272, 2534– 2541. Moos, P. J.; Fitzpatrick, F. A. Proc. Natl. Acad. Sci. USA 1998, 95, 3896–3901. Pratt, W. B.; Buddon, R. W.; Ensminger, W. D.; Maybaum, J. The Anticancer Drugs, Oxford University Press, Oxford, 1994. D. Schinzer, Eur. Chem. Chron. 1996, 1, 7–10; b) M. Kalesse, Eur. Chem. Chron. 1997, 2, 7–11; c) L. Wessjohann, Angew. Chem. 1997, 109, 739–742; Angew. Chem. Int. Ed. Engl. 1997, 36, 739–742; d) K. C. Nicolaou, F. Roschangar, D. Vourloumis, Angew. Chem. 1998, 110, 2120–2153; Angew. Chem. Int. Ed. Engl. 1998, 37, 2014–2045. Wartmann, M.; Altmann, K.-H. Curr. Med. Chem. 2002, 2, 123; b) Harris, C. R.; Kuduk, S. D.; Danishefsky, S. J. Chemistry for the 21st Century 2001, 8. A. Balog, D. Meng, T. Kamenecka, P. Bertinato, D.-S. Su, E. J. Sorensen, S. J. Danishefsky, Angew. Chem. 1996, 108, 2976–2978; Angew. Chem. Int. Ed. Engl. 1996, 35, 2801–2803. Z. Yang, Y. He, D. Vourloumis, H. Vallberg, K. C. Nicolaou, Angew. Chem. 1997, 109, 170–172; Angew. Chem. Int. Ed. Engl. 1997, 36, 166–168. D. Schinzer, A. Limberg, A. Bauer, O. M. Bo¨hm, M. Cordes, Angew. Chem. 1997, 109, 543–544; Angew. Chem. Int. Ed. Engl. 1997, 36, 523–524. N. Miyaura, A. Suzuki, Chem. Rev. 1995, 95, 2457. K. C. Nicolaou, Y. He, D. Vourloumis, H. Vallberg, Z. Yang, Angew. Chem. 1996, 108, 2554–2556; Angew. Chem. Int. Ed. Engl. 1996, 35, 2399–2401. K. C. Nicolaou, Y. He, D. Vourloumis, H. Vallberg, F. Roschangar, F. Sarabia, S. Ninkovic, Z. Zang, J. I. Trujillo, J. Am. Chem. Soc. 1997, 119, 7960–7973. D. Schinzer, A. Limberg, O. M. Bo¨hm, Chem. Eur. J. 1996, 2, 1477–1482. D. Schinzer, A. Limberg, A. Bauer, O. M. Bo¨hm, Chem. Eur. J. 1999, 5, 2483–2491. F. Stuhlmann, unpublished results, University of Magdeburg. A. Limberg, PhD thesis, Technical University of Braunschweig, 1998. a) D. Schinzer, A. Bauer, J. Schieber, Chem. Eur. J. 1999, 9, 2492–2500; b) D. Schinzer, A. Bauer, J. Schieber, Synlett. 1998, 861–864. a) R. E. Taylor, Y. Chen, Org. Lett. 2001, 3, 2221–2224; b) R. E. Taylor, G. M. Galvin, K. A. Hilﬁker, Y. Chen, J. Org.

327

328

7 The Aldol Reaction in Natural Product Synthesis: The Epothilone Story

22

23

24

25 26 27

28 29 30 31 32 33

34 35 36 37

38

Chem. 1998, 63, 9580–9583; c) R. E. Taylor, J. D. Haley, Tetrahedron Lett. 1997, 38, 2061–2064; d) N. Yoshikawa, Y. M. A. Yamada, J. Das, H. Sasai, M. Shibasaki, J. Am. Chem. Soc. 1999, 121, 4168–4178; e) D. Sawada, M. Shibasaki, Angew. Chem. 2000, 112, 215–219; Angew. Chem. Int. Ed. Engl. 2000, 39, 209–213; f ) D. Sawada, M. Kanai, M. Shibasaki, J. Am. Chem. Soc. 2000, 122, 10521–10532. a) K.-H. Altmann, Mini-Reviews in Medicinal Chemistry 2003, 3, 149; b) U. Klar, W. Skuballa, B. Buchmann, W. Schwede, T. Bunte, J. Hoffmann, R. B. Lichtner, ACS Symposium Series 2001, 796, 131. a) C. H. Heathcock, Science, 1981, 214, 395; b) C. H. Heathcock, C. T. White, J. Am. Chem. Soc. 1979, 101, 7076– 7077; c) C. H. Heathcock, M. C. Pirrung, C. T. Buse, J. P. Hagen, D. S. Young, J. E. Sohn, J. Am. Chem. Soc. 1979, 101, 7077–7079; d) C. H. Heathcock, C. T. Buse, W. A. Kleschick, M. C. Pirrung, J. E. Sohn, J. Lampe, J. Org. Chem. 1980, 45, 1066–1081. a) D. J. Cram, F. A. Abd Elhafez, J. Am. Chem. Soc. 1952, 74, 5828–5835; b) D. J. Cram, K. R. Kopecky, J. Am. Chem. Soc. 1959, 81, 2748–2755. a) M. Cherest, H. Felkin, N. Prudent, Tetrahedron Lett. 1968, 2199; b) N. T. Anh, O. Eisenstein, Nouv. J. Chem. 1977, 1, 61. Muharram, planned PhD thesis, University of Magdeburg, 2004. a) S. Masamune, W. Choy, F. A. J. Kerdesky, B. Imperiali, J. Am. Chem. Soc. 1981, 103, 1566–1568; b) C. H. Heathcock, C. T. White, J. Am. Chem. Soc. 1979, 101, 7076–7077; c) S. Masamune, W. Choy, J. S. Petersen, R. L. Sita, Angew. Chem. 1985, 97, 1–31. C. H. Heathcock in Asymmetric Synthesis, Vol 3 (Ed. J. D. Morrison), Academic Press, New York, 1984, 111. O. M. Bo¨hm, University of Magdeburg, unpublished results. H. E. Zimmerman, M. D. Traxler, J. Am. Chem. Soc. 1957, 79, 1920–1923. E. Claus, A. Pahl, P. G. Jones, H. M. Meyer, M. Kalesse, Tetrahedron Lett. 1997, 38, 1359–1362. K. Gerlach, M. Quitschalle, M. Kalesse, Tetrahedron Lett. 1999, 40, 3553. K. C. Nicolaou, S. Nincovic, F. Sarabia, D. Vourloumis, Y. He, H. Vallberg, M. R. V. Finlay, Z. Yang, J. Am. Chem. Soc. 1997, 119, 7974–7991. ¨ hler, Tetrahedron Lett. 1998, J. Mulzer, A. Mantoulidis, E. O 39, 8633–8636. J. Mulzer, G. Karig, P. Pojarliev, Tetrahedron Lett. 2000, 41, 7635–7638. H. J. Martin, M. Drescher, J. Mulzer, Angew. Chem. 2000, 112, 591–593; Angew. Chem. Int. Ed. Engl. 2000, 39, 581–583. A. Balog, C. Harris, K. Savin, X.-G. Zhang, T.-C. Chou, S. J. Danishefsky, Angew. Chem. 1998, 110, 2821–2824; Angew. Chem. Int. Ed. Engl. 1998, 37, 2675–2678. C. B. Lee, Z. Wu, F. Zhang, M. D. Chappel, S. J. Stachel, T.-C. Chou, Y. Guan, S. J. Danishefsky, J. Am. Chem. Soc. 2001, 123, 5249–5259

1

1

Silver, Gold, and Palladium Lewis Acids Akira Yanagisawa 1.1

Introduction

Silver(I), gold(I), and palladium(II) salts have moderate Lewis acidity and have been exploited as catalysts in organic reactions in recent years. Among these salts, Pd(II) compounds are the most well-known reagents for catalyzing a variety of carbon–carbon bond-forming reactions such as allylic alkylations [1]. Ag(I) salts are also popular reagents for promoting transformations, including glycosylation, cycloadditions, and rearrangements, which make use of their halophilicity or thiophilicity [2]. There are, however, few examples of organic reactions employing Au(I) or Au(III) compounds as Lewis acid catalysts. This chapter focuses on aldol reactions catalyzed by silver(I), gold(I), or palladium(II) Lewis acids. The Mukaiyama aldol reaction of silyl enol ethers or ketene silyl acetals and related reactions using silver(I) and palladium(II) compounds are reviewed in Section 1.2. The next section covers the diastereo- and enantioselective aldol-type reactions of activated isocyanides with aldehydes.

1.2

Mukaiyama Aldol Reaction and Related Reactions

Silver(I) compounds are known to promote the aldol condensation between silyl enol ethers or ketene silyl acetals and aldehydes (the Mukaiyama aldol reaction). For example, the adduct 3 is obtained in 72% yield when ketene silyl acetal 1 is treated with a,b-unsaturated aldehyde 2 in the presence of a catalytic amount of Ag(fod) (Scheme 1.1). Eu(fod)3 or Yb(fod)3 catalyzes a hetero-Diels–Alder reaction of 1 and 2 [3]. A [2þ2] cycloaddition followed by a ring opening of the resulting oxetane is an alternative possible route to the adduct 3. A BINAP-silver(I) complex is a superior asymmetric catalyst for allylation of aldehydes with allylic stannanes [4]. The chiral phosphine-silver(I) Modern Aldol Reactions. Vol. 2: Metal Catalysis. Edited by Rainer Mahrwald Copyright 8 2004 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 3-527-30714-1

2

1 Silver, Gold, and Palladium Lewis Acids

OSiMe3

O +

MeO

OSiMe3

Ag(fod) (5 mol%)

OHC

Ph 2

OMe

Ph

MeO

CH2Cl2, r.t.

OMe 3 (dr 60:40)

72% yield

1 Scheme 1.1

Ag(fod)-catalyzed Mukaiyama aldol reaction of ketene silyl acetal.

catalyst is prepared simply by stirring a 1:1 mixture of BINAP and silver(I) compound in THF at room temperature. The BINAP-silver(I) complex can be also used as a chiral catalyst of asymmetric aldol reaction. Although a variety of beneﬁcial methods have been developed for catalytic asymmetric aldol reaction, most of these are chiral Lewis acid-catalyzed Mukaiyama aldol reactions using ketene silyl acetals or silyl enol ethers [5] and there has been no example on enol stannanes. Yanagisawa, Yamamoto, and their colleagues ﬁrst reported the enantioselective aldol addition of tributyltin enolates 4 to aldehydes catalyzed by a BINAP-silver(I) complex (Scheme 1.2) [6]. OSnBu3 2

R + R4CHO

R1 R

(R)-BINAP·AgOTf (10 mol%) THF, -20 °C, 8 h

3

4

O

OH

R1

R4 R2

R

3

5

Scheme 1.2

Enantioselective aldol reaction of tributyltin enolates catalyzed by BINAPsilver(I) complex.

The tributyltin enolates 4 are easily generated from the corresponding enol acetates and tributyltin methoxide without any solvent [7]. The tin compounds thus prepared exist in the O-Sn form and/or the C-Sn form. Although the tin reagents themselves have suﬃcient reactivity toward aldehydes [7c], under the inﬂuence of the BINAP-silver(I) catalyst the reaction advances faster even at 20 C. The results employing optimum conditions in the catalytic enantioselective aldol reaction of a variety of tributyltin enolates or a-tributylstannylketones with aromatic, a,b-unsaturated, and aliphatic aldehydes are summarized in Table 1.1. The characteristic features are: (i) all reactions occur to provide the corresponding aldol adducts 5 in moderate to high yield in the presence of 10 mol% (R)-BINAP-AgOTf complex at 20 C, and no dehydrated aldol adduct is formed; (ii) with an a,bunsaturated aldehyde, the 1,2-addition reaction is predominant (entry 3); (iii) use of a sterically hindered tin enolate results in an increase in the

1.2 Mukaiyama Aldol Reaction and Related Reactions

3

Tab. 1.1

Diastereo- and enantioselective aldol addition of tin compounds to aldehydes in the presence of 10 mol% of (R)-BINAPAgOTf complex in THF at 20 C. Entry Tin Compound

Aldehyde

O

O

1d

SnBu3

Ph

5a O

O t-Bu

OH

PhCHO

6

2d

Yield (%)a anti:syn b ee (%)c

Product

SnBu3

PhCHO

t-Bu

7

Ph

* Ph

5b

Ph

t-Bu

t-Bu

69

86

75

94

OH

t-Bu

4a

Ph

81

<1:99

95

77

<1:99

95

94

92:8

93i

95

93:7

94i

5e O

6e

Ph

5d O

PhCHO

95

OH

CHO

OSnBu3

5e

Ph

*

5c O

4d

78

OH

CHO

t-Bu

77

OH

O

3d

73

Ph

OH

CHO t-Bu

Ph 5f

OSnBu3

7f

O PhCHO

O Ph +

5g-anti

4b

8f,h

OH

PhCHO

OH Ph 5g-syn

a Isolated

yield. by 1 H NMR analysis. c The value corresponds to the major diastereomer. Determined by HPLC analysis with chiral columns. d OaSn:CaSn < 1:99. e OaSn:CaSn > 99:1. The E:Z ratio for the OaSn isomer was < 1:99. f OaSn:CaSn > 99:1. g The syn isomer: 25% ee. h 1 mol% catalyst was used. i The syn isomer: 33% ee. b Determined

enantioselectivity of the aldol reaction. For example, ee higher than 90% are observed when pinacolone and tert-butyl ethyl ketone-derived tin compounds 7 and 4a are treated with aldehydes (entries 2 and 4–6); (iv) addition of the enol tributylstannane 4b derived from cyclohexanone ((E)-enolate) to

4

1 Silver, Gold, and Palladium Lewis Acids

P

H R R2

O

E

R

Ag+

O

R3

P SnBu3

P

H

*

1

H

*

1

Ag+

O

R3

O

Z

P SnBu3

R2

H A

anti

B

syn

Fig. 1.1

Probable structures of cyclic transition states.

benzaldehyde in the presence of 10 mol% (R)-BINAP-AgOTf in THF at 20 C yields the non-racemic anti aldol adduct 5g selectively with an anti:syn ratio of 92:8, in contrast with the syn selectivity aﬀorded by representative chiral Lewis acid catalysts [5]. The anti isomer indicates 93% ee (entry 7). The amount of catalyst can be reduced to 1 mol% without losing the isolated yield or diastereo- and enantioselectivity (entry 8). In contrast, the (Z) enolate generated from tert-butyl ethyl ketone 4a produces the synaldol adducts 5e and 5f almost exclusively with 95% ee in the reaction with benzaldehyde and hydrocinnamaldehyde (entries 5 and 6). These results reveal unambiguously that the diastereoselectivity relies on the geometry of tin enolate, and that cyclic transition-state structures (A and B, Figure 1.1) are plausible models. Accordingly, from the (E) enolate, the anti-aldol product forms via a model A, and another model B for the (Z) enolate leads to the syn product. Analogous six-membered cyclic models including a BINAP-coordinated silver atom in place of a tributylstannyl group are also probable substitutes when the transmetalation to silver enolate is suﬃciently rapid. Although the above-mentioned reaction is a superior asymmetric aldol process with regard to enantioselectivity and diastereoselectivity, it has a disadvantage of requiring the stoichiometric use of toxic trialkyltin compounds. The same group has shown that the amount of trialkyltin compounds can be reduced to a catalytic amount when an enol trichloroacetate is employed as a substrate for the reaction [8]. For example, treatment of benzaldehyde with the enol trichloroacetate of cyclohexanone 8 under the inﬂuence of (R)BINAP-AgOTf complex (5 mol%), tributyltin methoxide (5 mol%), and MeOH (200 mol%) in dry THF at 20 C to room temperature for 20 h provides a 92:8 mixture of non-racemic anti and syn aldol adduct, 5g-anti and 5g-syn respectively, in 82% yield (Scheme 1.3). The anti isomer 5g-anti aﬀords 95% ee, a grade of enantiomeric excess similar to that obtained from a BINAP-silver(I)-catalyzed aldol reaction with enol tributylstannanes [6]. A suggested catalytic cycle of this asymmetric aldol reaction is shown in Figure 1.2. To start with, Bu3 SnOMe reacts with enol trichloroacetate 9 to yield trialkyltin enolate 4 and methyl trichloroacetate. The tin enolate 4 then adds enantioselectively to an aldehyde under the inﬂuence of BINAP-AgOTf

1.2 Mukaiyama Aldol Reaction and Related Reactions

(R)-BINAP·AgOTf (5 mol%) Bu3SnOMe (5 mol%) MeOH (200 mol%)

OCOCCl3 +

PhCHO

O

5

OH

O Ph +

Ph

THF, -20 °C (8 h) ~ r.t. (12 h) 82% yield

8

5g-anti (2S,1'R)

5g-syn (2R,1'R)

anti:syn = 92 (95% ee):8 Scheme 1.3

Enantioselective aldol reaction catalyzed by tin methoxide and BINAPsilver(I) complex.

as an asymmetric catalyst to furnish the tin alkoxide of non-racemic aldol adduct 10. Last, protonolysis of 10 by MeOH produces the optically active aldol product 5 and regenerates the tin methoxide. The rate of methanolysis is considered to be the key to success in the catalytic cycle. The BINAP-Ag(I)-catalyzed asymmetric Mukaiyama aldol reaction using trimethylsilyl enol ethers was ﬁrst developed by Yamagishi and co-workers, who found that the reaction was accelerated by BINAP-AgPF6 in DMF containing a small amount of water, to give the aldol product with high enantioselectivity [9] (Scheme 1.4). In the reaction with BINAP-AgOAc, much higher catalytic activity and opposite absolute conﬁguration of the aldol adduct were observed and ee was low [9]. Yanagisawa, Yamamoto, and their colleagues independently examined different combinations of BINAP-Ag(I) catalysts and silyl enol ethers and found that high enantioselectivity and chemical yields were obtained in the p-Tol-BINAP-AgF-catalyzed aldol reaction of trimethoxysilyl enol ethers in

O

(R)-BINAP·AgOTf R

OSnBu3

1

R4 R2 R 3 10

R4CHO

R1

OH

MeOH

OSnBu3 R2

O

3 4 R

Bu3SnOMe

MeOCOCCl3 R1

OCOCCl3 R2 9 R3

Fig. 1.2

A proposed catalytic mechanism for the asymmetric aldol reaction catalyzed by (R)-BINAPAgOTf and tin methoxide.

OH

R1

R4 R2 R 3 5

1 Silver, Gold, and Palladium Lewis Acids

6

OSiMe3

O

(S)-BINAP·AgPF6 (2 mol%)

+ Ph

Ph

H

O

HCl

DMF + 2% H2O, 25 °C, 2 h

THF

OH

Ph

11

Ph

12, 69% ee (S)

100% yield Scheme 1.4

BINAPsilver(I)-catalyzed asymmetric Mukaiyama aldol reaction.

methanol [10]. In addition, remarkable syn selectivity was observed for the reaction irrespective of the E:Z stereochemistry of the silyl enol ethers. For example, when the (Z)-trimethoxysilyl enol ether of t-butyl ethyl ketone 13 was treated with benzaldehyde the reaction proceeded smoothly at 78 to 20 C and syn-aldol adduct 5e was obtained almost exclusively with 97% ee (Scheme 1.5). In contrast, cyclohexanone-derived (E)-silyl enol ether gave the aldol adduct with an anti:syn ratio of 84:16 [10]. Use of a 1:1 mixture of MeOH and acetone as a solvent in the reaction of the trimethoxysilyl enol ethers resulted in higher enantioselectivity [10b]. O

OSi(OMe)3 +

OH

(R)-p-Tol-BINAP·AgF (10 mol%) PhCHO

t-Bu

MeOH, -78 ~ -20 °C, 6 h

13

84% yield

t-Bu

Ph

5e, syn:anti > 99 (97% ee):1

Scheme 1.5

Enantioselective aldol reaction of trimethoxysilyl enol ether catalyzed by p-Tol-BINAPAgF complex.

The BINAP-silver(I) complex was further applied to asymmetric Mannich-type reactions by Lectka and coworkers [11]. Treatment of silyl enol ether 11 with a solution of a-imino ester 14 in the presence of 10 mol% (R)-BINAP-AgSbF6 at 80 C leads the corresponding a-amino acid derivative 15 in 95% yield with 90% ee (Scheme 1.6). They showed that (R)BINAP-Pd(ClO4 )2 was also an eﬀective chiral Lewis acid for the reaction though it gave lower ee (80%). Asymmetric Mukaiyama aldol reactions can also be catalyzed by cationic BINAP-Pd(II) complexes. In 1995 Sodeoka, Shibasaki et al. ﬁrst reported Ts

OSiMe3

N

+ H

Ph 11

O

(R)-BINAP·AgSbF6 (10 mol%) CO2Et 14

THF, -80 °C, 24 h 95% yield

Ph

N

Ts

CO2Et 15, 90% ee

Scheme 1.6

Enantioselective Mannich-type reaction catalyzed by BINAPsilver(I) complex.

1.2 Mukaiyama Aldol Reaction and Related Reactions

(R)-BINAP·PdCl2 (5 mol%) AgOTf (5 mol%)

OSiMe3 + OHC

Ph

Ph 16

11

O

H+

7

OH

Ph

MS 4A, DMF-H2O, 23 °C

Ph 17, 73% ee (S)

86% yield Ar P

Ar

OH2 Pd2+ 2BF4– OH2 P Ar Ar 22, Ar = Ph 23, Ar = p-Tol

Ar P

H Ar

O

Ar

Ar P

Pd+ Pd+ O P P Ar Ar H Ar Ar 24, Ar = Ph 25, Ar = p-Tol

Scheme 1.7

Asymmetric Mukaiyama aldol reaction catalyzed by cationic BINAPPd(II) complexes.

that (R)-BINAP-PdClþ , prepared from a 1:1 mixture of (R)-BINAP-PdCl2 and AgOTf in wet DMF is an eﬀective chiral catalyst for asymmetric aldol condensation of silyl enol ethers and aldehydes [12]. For instance, when hydrocinnamaldehyde (16) is treated with trimethylsilyl enol ether of acetophenone 11 under the inﬂuence of 5 mol% of this catalyst followed by desilylation the desired aldol adduct 17 is obtained in 86% yield with 73% ee as shown in Scheme 1.7. Some examples of the aldol reaction are summarized in Table 1.2. The same group subsequently succeeded in generating chiral palladium diaquo complexes 22 and 23 from (R)-BINAP-PdCl2 and (R)- pTol-BINAP-PdCl2 , respectively, by treatment with 2 equiv. AgBF4 in wet acetone [13]. These complexes are inert toward air and moisture, and have similar reactivity and enantioselectivity in the aldol reaction of 11 with the aliphatic aldehyde 16. Sodeoka et al. have further developed catalytic asymmetric Mannich-type reactions of silyl enol ethers with imines employing binuclear m-hydroxo palladium(II) catalysts 24 and 25 generated from the diaquo complexes 22 and 23, respectively [14]. In the reaction, a chiral palladium(II) enolate is assumed to be formed from the corresponding silyl enol ether. They later developed palladium complexes with polymer-supported BINAP ligands and showed that these reusable complexes are good catalysts for the asymmetric aldol reactions and Mannich-type reactions mentioned above [15]. Doucet and coworkers have shown that the complex dppe-Pd(OAc)2 is an eﬃcient catalyst for Mukaiyama aldol addition of ketene silyl acetals to aldehydes and ketones under neutral conditions [16]. The reaction proceeds smoothly, even in the presence of 0.1 mol% of the catalyst. [Bis(diphenylphosphino)alkane]bis(propenyl)ruthenium complexes also catalyze the aldol addition, furnishing a variety of 3-hydroxymethyl esters in good yields, as do other late transition metal complexes; platinum(II) cationic complexes are known to act as Lewis acids. Fujimura reported the ﬁrst

2BF4–

8

1 Silver, Gold, and Palladium Lewis Acids Tab. 1.2

Catalytic asymmetric aldol reaction of silyl enol ethers with benzaldehyde in the presence of 5 mol% (R)-BINAPPdCl2 aAgOTf in wet DMF. Entry

Silyl Enol Ether

Product

OSiMe3

1c

O

Ph

ee (%)b

87d

71

80

73

58e

72f

OSiMe3

Ph 11

Yield (%)a

Ph 18

OSiMe3

O

OH Ph

2 20

19 OSiMe3

O

OH

3

Ph 21

5g

a Isolated

yield. by HPLC analysis with chiral columns. c Desilylation with acid was not done. d Desilylated product was formed in 9% yield with 73% ee. e The syn:anti ratio was 74:26. f The value corresponds to the major diastereomer. b Determined

example of platinum-catalyzed enantioselective aldol reaction of ketene silyl acetals with aldehydes [17]. The chiral catalysts are prepared by treatment of chiral bisphosphine-Pt acyl complexes with triﬂic acid. In the aldol reaction, a C-bound platinum enolate is assumed to be an intermediate, on the basis of on 31 P NMR and IR studies.

1.3

Asymmetric Aldol Reactions of a-Isocyanocarboxylates

In 1986, Ito, Sawamura, and Hayashi showed that chiral ferrocenylphosphine 26-gold(I) complexes catalyzed the aldol-type reaction of isocyanoacetate with aldehydes to provide optically active 5-alkyl-2-oxazoline-4carboxylate (Scheme 1.8) [18]. Since then, they have extensively studied the chiral gold(I)-catalyzed reaction [19] as have Pastor and Togni [20]. The gold complexes can be generated in situ by mixing bis(cyclohexyl isocyanide)gold(I) tetraﬂuoroborate and (R)-N-methyl-N-[2-(dialkylamino)ethyl]-1-[(S)-1 0 ,2-bis(diphenylphosphino)ferrocenyl]ethylamine (26). Examples of the reaction of methyl isocyanoacetate (27) and diﬀerent aldehydes in the presence of 1 mol% of 26c-Au(I) complex are summarized in

1.3 Asymmetric Aldol Reactions of a-Isocyanocarboxylates

Me

H N

Fe

NR'2 + [Au(c-HexNC)2]BF4

PPh2

Me PPh2 26

26·Au(I) complex

CH2Cl2

26a: NR'2 = NMe2 26b: NR'2 = N 26c: NR'2 = N

CO2Me + RCHO

O

26·Au(I) complex (1 mol%)

R

CH2Cl2, 25 °C

O

NC

CO2Me R

CO 2Me

+

27

N

O

trans-28

N

cis-28

H3O+ NH2 R

CO2H OH syn-29

NH2

NH2 OH

n-C13H27 OH D-syn-sphingosine

OH

n-C13H27 OH

(30)

D-anti-sphingosine

(31)

NHMe CO2H OH MeBmt [(4R)-4-((E)-But-2-enyl)-4,N-dimethyl-L-threonine, 32] Scheme 1.8

Asymmetric aldol reaction of methyl isocyanoacetate with aldehydes catalyzed by chiral ferrocenylphosphine 26–gold(I) complexes.

Table 1.3. Benzaldehyde and substituted aromatic aldehydes, except 4nitrobenzaldehyde, are transformed into the corresponding trans-oxazolines 28 with high enantio- and diastereoselectivity (entries 1–6). Secondary and tertiary alkyl aldehydes give trans-28 nearly exclusively with high ee (entries 9 and 10). The trans-oxazolines 28 can be readily hydrolyzed to threo-b-

9

1 Silver, Gold, and Palladium Lewis Acids

10 Tab. 1.3

Diastereo- and enantioselective aldol reaction of methyl isocyanoacetate (27) with aldehydes catalyzed by chiral ferrocenylphosphine 26cgold(I) complex. Entry Aldehyde

Yield, %a trans:cis b % ee c

Product Ph

1

CO2Me

Ph

PhCHO

O

N

O

trans-28a

N

CO2Me 2-MeC6H4

O Me

O

N

trans-28b 2-MeOC6H4

CHO

trans-28c CO2Me

O

4-ClC6H4

N

CO2Me 4-O2NC6H4

O

O

N

N

CO2Me

O CHO

O

+

N

O

83:17

86

86

95:5

96

85

87:13

92

99

89:11

89

99

96:4

87

N

CO2Me +

CHO

Me

O

N

CO2Me

N

cis-28g Me

CO2Me

+ O

N

O

trans-28h i-Bu i-BuCHO

80

cis-28f

trans-28g

9

94

n-Pr

O

MeCHO

94:6

CO2Me

O

CO2Me

8d

97

O

n-Pr n-Pr

92

cis-28e

trans-28f

7

92:8

CO2Me

+ O

O O

98

cis-28d

trans-28e

6

95

CO2Me

O

N

4-O2NC6H4 CHO

96:4

cis-28c

trans-28d

O 2N

N

+

CHO

98

CO2Me

O

N

4-ClC6H4 Cl

95

cis-28b

CO2Me 2-MeOC6H4

O OMe

5

N

+

3

95:5

CO2Me

+

2

93

cis-28a

2-MeC6H4

CHO

4

CO2Me

+

N cis-28h

CO2Me

i-Bu

CO 2Me

+ O

N

trans-28i

O

N cis-28i

1.3 Asymmetric Aldol Reactions of a-Isocyanocarboxylates

11

Tab. 1.3

(continued) Entry Aldehyde

CO2Me

t-Bu

10

Yield, %a trans:cis b % ee c

Product CO2Me

t-Bu +

t-BuCHO

O

N

O

trans-28j

N

94

cis-28j

a Isolated

yield. by 1 H NMR analysis. c Determined by 1 H NMR analysis with chiral shift reagent Eu(dcm) . 3 d 0.2 mol% of the catalyst was used. b Determined

hydroxy a-amino acids 29. The gold-catalyzed aldol reaction has been applied to the asymmetric synthesis of the biologically important compounds d-threo-sphingosine (30) [21], d-erythro-sphingosine (31) [21], and MeBmt (32) [22]. The enantioselective synthesis of ()-a-kainic acid has also been achieved using this aldol reaction [23]. A proposed transition-state model for the reaction is shown in Figure 1.3. The presence of the 2-(dialkylamino)ethylamino group in 26 is necessary to obtain high selectivity [24]. The terminal amino group abstracts one of the a-protons of isocyanoacetate coordinated with gold, and the resulting ion pair causes advantageous arrangement of the enolate and aldehyde around the gold. In contrast, Togni and Pastor proposed an alternative acyclic transition-state model [20d]. The chiral ferrocenylphosphine-gold(I)-catalyzed aldol reaction of a-alkyl a-isocyanocarboxylates 33 with paraformaldehyde gives optically active 4alkyl-2-oxazoline-4-carboxylates 34 with moderate to good enantioselectivity [25]. The absolute conﬁguration (S) of the product indicates that the reaction proceeds selectively at the si face of the enolate, as illustrated in Figure 1.3. These oxazolines 34 can be converted into a-alkylserine derivatives 35 (Scheme 1.9).

H

Ph Ph P

+ Au

Fe P

Ph

Me Me

Ph

O C

R N

O– OMe

+ NHR'2 H

N

Fig. 1.3

Transition-state assembly in the gold-catalyzed asymmetric aldol reaction.

>99:1

97

12

1 Silver, Gold, and Palladium Lewis Acids

CO2Me + (CH2O)n

R

L*·[Au(c-HexNC)2]BF4 (1 mol%) CH2Cl2, 25 °C

NC

MeO2C R O

33 R = Me, Et, i-Pr, Ph L* = 26a or 26b

N

(S)-34 63~81% ee H 3O + MeO2C R NH2 HO (S)-35

Scheme 1.9

Catalytic asymmetric synthesis of a-alkylserines.

This enantioselective aldol reaction using isocyanoacetate 27 is quite eﬀective for aromatic aldehydes or tertiary alkyl aldehydes, but not for sterically less hindered aliphatic aldehydes, as described above. Ito and coworkers found that very high enantioselectivity is obtained even for acetaldehyde (R ¼ Me) in the aldol reaction with N,N-dimethyl-a-isocyanoacetamide (36) (Scheme 1.10) [26]. Use of a-keto esters in place of aldehydes also results in moderate to high enantioselectivity of up to 90% ee [27]. The same group also developed an asymmetric aldol reaction of Nmethoxy-N-methyl-a-isocyanoacetamide (a-isocyano Weinreb amide) with aldehydes (Scheme 1.10). For instance, reaction of the Weinreb amide 37 with acetaldehyde in the presence of 26c-Au(I) catalyst gives the optically active trans-oxazoline 39 (E ¼ CON(Me)OMe; R ¼ Me) with high diastereoand enantioselectivity similar to those of 36 [28]. The oxazoline can be transformed into N,O-protected b-hydroxy-a-amino aldehydes or ketones. (Isocyanomethyl)phosphonate 38 is also a beneﬁcial pronucleophile that leads to optically active (1-aminoalkyl)phosphonic acids, which are

E +

RCHO

NC 36~38

L*·[Au(c-HexNC)2]BF4 (1 mol%)

R O

N

O

trans-39 E

R

L*

CONMe2 (36) CON(Me)OMe (37) PO(OPh)2 (38)

Me Me Ph

26b 26c 26b

Gold(I)-catalyzed asymmetric aldol reaction of isocyanoacetamides and (isocyanomethyl)phosphonate.

E

+

CH2Cl2, 25 °C

Scheme 1.10

R

E

N

cis-39

trans (% ee) : cis 91 (99) : 9 95 (97) : 5 >98 (96) : 2

1.3 Asymmetric Aldol Reactions of a-Isocyanocarboxylates

E +

R

E

O

N

26b·Ag(I) (1~2 mol%)

E

+

RCHO solvent, 25~30 °C

NC 27 or 40

R O

trans-39

E

R

Ag(I)

Solvent

CO2Me (27)a CO2Me (27)a SO2(p-Tol) (40) SO2(p-Tol) (40)

Ph i-Pr Ph i-Pr

AgOTf AgClO4 AgOTf AgOTf

ClCH2CH2Cl ClCH2CH2Cl CH2Cl2 CH2Cl2

N

cis-39

trans (% ee) : cis 96 (80) : 4 99 (90) : 1 >99 (77) : 1 >99 (86) : 1

a) slow addition of 27 over 1 h. Scheme 1.11

Asymmetric aldol reaction of methyl isocyanoacetate and tosylmethylisocyanide catalyzed by chiral ferrocenylphosphinesilver(I) complex.

phosphonic acid analogs of a-amino acids, via trans-5-alkyl-2-oxazoline-4phosphonates 39 (E ¼ PO(OPh)2 , Scheme 1.10) [29]. Ito and coworkers found that chiral ferrocenylphosphine-silver(I) complexes also catalyze the asymmetric aldol reaction of isocyanoacetate with aldehydes (Scheme 1.11) [30]. It is essential to keep isocyanoacetate at a low concentration to obtain a product with high optical purity. They performed IR studies on the structures of gold(I) and silver(I) complexes with chiral ferrocenylphosphine 26a in the presence of methyl isocyanoacetate (27) and found a signiﬁcant diﬀerence between the coordination numbers of the isocyanoacetate to the metal in these metal complexes (Scheme 1.12). The P 26a Au+ CNCH2CO2Me P

41 RCHO high ee

low ee

RCHO

RCHO P

P 26a Ag

+

+27 CNCH2CO2Me

P

42

-27

Scheme 1.12

A diﬀerence in the coordination number of methyl isocyanoacetate to metal between gold(I) and silver(I) complexes.

26a Ag+ P

CNCH2CO2Me CNCH2CO2Me 43

13

14

1 Silver, Gold, and Palladium Lewis Acids

gold(I) complex has a tricoordinated structure 41, which results in high ee, whereas the silver(I) complex is in equilibrium between tricoordinated structure 42 and tetracoordinated structure 43, which results in low enantioselectivity. Slow addition of isocyanoacetate 27 to a solution of the silver(I) catalyst and aldehyde eﬀectively reduces the undesirable tetracoordinated species and results in high enantioselectivity. The asymmetric aldol-type addition of tosylmethyl isocyanide (40) to aldehydes can also be catalyzed by the chiral silver(I) complex giving, almost exclusively, trans-5-alkyl-4-tosyl-2-oxazolines 39 [E ¼ SO2 ( p-Tol)] with up to 86% ee, as shown in Scheme 1.11 [31]. The slow addition method described above is not necessary for this reaction system. Soloshonok and Hayashi used chiral ferrocenylphosphine-gold(I) complexes in asymmetric aldol-type reactions of ﬂuorinated benzaldehydes with methyl isocyanoacetate (27) and N,N-dimethyl-a-isocyanoacetamide (36). Interestingly, successive substitution of hydrogen atoms by ﬂuorine in the phenyl ring of benzaldehyde causes a gradual increase in both the cis selectivity and the ee of cis oxazolines [32]. Cationic chiral palladium complexes are known to catalyze the aldol reaction of methyl isocyanoacetate (27) and aldehydes. For example, Richards et al. prepared cationic 2,6-bis(2-oxazolinyl)phenylpalladium(II) complex 44 from the corresponding bromopalladium(II) complex and AgSbF6 in wet CH2 Cl2 and showed that an increase in rate was observed for the aldol reaction of 27 with benzaldehyde in the presence of 1 mol% 44 and 10 mol% Hu¨nigs base [33]. Zhang and coworkers developed a palladium(II) complex of PCP-type chiral ligand 46 (PCP is the monoanionic ‘‘pincer’’ ligand [C6 H3 (CHMePPh2 )2 -2,6] ). Removal of the chloride with AgOTf produces an active cationic chiral Pd(II) catalyst for the asymmetric aldol reaction of aldehydes (Scheme 1.13) [34]. Several examples of the reaction under the inﬂuence of 1 mol% of catalyst 46 are summarized in Table 1.4. When the eﬀects of solvent on enantioselectivity were examined in the reaction with benzaldehyde, THF was found to be solvent of choice (trans-47a: 24% ee, cis47a: 67% ee, entry 1). The trans isomers were usually obtained as major products though with lower ee. In the reaction with aromatic aldehydes the enantioselectivity is almost constant (entries 1–5) and the trisubstituted aromatic aldehyde gives the highest ee (entry 5). It is noteworthy that higher enantioselectivity is observed with aliphatic aldehydes than with aromatic aldehydes with regard to their cis product (entries 6 and 7). l-Valine-derived NCN-type Pd(II) complex 48 (NCN is the monoanionic, para-functionalized ‘‘pincer’’ ligand [C6 H2 (CH2 NMe 2 )2 -2,6] ) synthesized by van Koten and coworkers is also an active catalyst for the aldol reaction after conversion into the corresponding cationic complex by treatment with AgBF4 in wet acetone [35]. Motoyama and Nishiyama have shown that excellent trans diastereoselectivity (> 99% trans) and moderate enantioselectivity (57% ee) was obtained in the asymmetric aldol-type condensation of tosylmethyl isocyanide with benzaldehyde employing cationic Pd(II) aqua complex 45 [36].

1.4 Summary and Conclusions

+ X– O

O N i-Pr

Pd

Ph2P

N

H H2O i-Pr 44 (X = SbF6) 45 (X = BF4)

CO2Me + RCHO

PPh2

46

46/AgOTf (1 mol%) (i-Pr)2NEt (10 mol%)

R

THF, 23 °C

O

NC

Pd Cl

H

27

CO2Me R

CO2Me

+ N

trans-47

O

N

cis-47

NMe2 Pd Br NH NMe2

HO2C 48 Scheme 1.13

Asymmetric aldol reaction of methyl isocyanoacetate with aldehydes catalyzed by cationic chiral palladium(II) complexes.

Other notable examples of the aldol-type reaction using a variety of palladium complexes have also appeared [37–41].

1.4

Summary and Conclusions

Described herein are examples of aldol reactions using silver(I), gold(I), or palladium(II) Lewis acids. The BINAP-silver(I) catalyst has made possible the aldol reaction of silyl enol ethers or trialkyltin enolates with high enantio- and diastereoselectivity. This silver catalyst is also eﬀective in Mannich-type reactions of silyl enol ethers with a-imino esters. The remarkable aﬃnity of the silver ion for halides is useful for accelerating chiral palladium-catalyzed asymmetric Mukaiyama aldol reactions. Isolated chiral palladium diaquo complexes and binuclear m-hydroxo palladium(II) complexes can catalyze asymmetric Mannich-type reactions and the aldol reaction. The chiral ferrocenylphosphine gold(I)-catalyzed asymmetric aldol reaction results in high stereoselectivity, although the substrates are restricted to a-isocyanocarboxylates and their derivatives, and has proven to be an excellent method for synthesizing optically active a-amino acid derivatives and

15

PhCHO

1

CH3

O 2N

3

4

2

Aldehyde

Entry

CHO

CHO

CHO

N

N

+

N

+

N

N

CO2Me

cis-47d

O

N

CO2Me

cis-47b

O

CO2Me

cis-47c

O

CO2Me 4-O2NC6H4

trans-47d

O

4-O2NC6H4

+

CO2Me 4-MeC6H4

trans-47c

O

4-MeC6H4

N

CO2Me

cis-47a

O

Ph

CO2Me N

+

trans-47b

O

CO2Me

trans-47a

O

Ph

Product

60

80

81

85

Yield, %a

74:26

82:18

81:19

78:22

trans:cis b

21

11

23

24

trans

% ee c

Asymmetric aldol reaction of methyl isocyanoacetate (27) with aldehydes catalyzed by a cationic Pd(II) complex generated from 46 and AgOTf.

Tab. 1.4

57

61

66

67

cis

16

1 Silver, Gold, and Palladium Lewis Acids

EtCHO

CHO

CHO

b Determined

yield. by 1 H NMR analysis. c Determined by GC analysis.

a Isolated

7

6

5

N

CO2Me

trans-47g

O

Et

N

+

CO2Me

trans-47f

O

c-C6H11

N

+

N

CO2Me

cis-47g

O

Et

N

N

CO2Me

cis-47e

O

CO2Me

cis-47f

O

c-C6H11

+

CO2Me 2,4,6-Me3C6H2

trans-47e

O

2,4,6-Me3C6H2

91

97

84

91:9

72:28

86:14

30

11

26

70

74

71

1.4 Summary and Conclusions 17

18

1 Silver, Gold, and Palladium Lewis Acids

amino alcohols. The examples given here unambiguously indicate that silver(I), gold(I), and palladium(II) compounds in combination with chiral ligands are chiral Lewis acid catalysts of great promise for asymmetric synthesis.

1.5

Experimental Procedures Typical Procedure for Asymmetric Aldol Reaction of Benzaldehyde with 3,3Dimethyl-1-tributylstannyl-2-butanone (7) Catalyzed by BINAP-AgOTf Complex. Synthesis of (R)-4,4-Dimethyl-1-hydroxy-1-phenyl-3-pentanone (5b, entry 2 in Table 1.1) [6]. A mixture of AgOTf (26.7 mg, 0.104 mmol) and (R)BINAP (64.0 mg, 0.103 mmol) was dissolved in dry THF (3 mL) under argon atmosphere and with direct light excluded and stirred at 20 C for 10 min. To the resulting solution was added dropwise a THF solution (3 mL) of benzaldehyde (100 mL, 0.98 mmol), and then 3,3-dimethyl-1tributylstannyl-2-butanone (7, 428.1 mg, 1.10 mmol) was added over a period of 4 h with a syringe pump at 20 C. The mixture was stirred for 4 h at this temperature and treated with MeOH (2 mL). After warming to room temperature, the mixture was treated with brine (2 mL) and solid KF (ca. 1 g). The resulting precipitate was removed by ﬁltration and the ﬁltrate was dried over Na2 SO4 and concentrated in vacuo. The crude product was puriﬁed by column chromatography on silica gel to aﬀord the aldol adduct 5b (161.7 mg, 78% yield as a colorless oil). TLC RF 0.22 (1:5 ethyl acetate–hexane); IR (neat) 3625–3130, 3063, 3033, 2971, 2907, 2872, 1701, 1605, 1495, 1478, 1455, 1395, 1368, 1073, 1057, 1011, 984, 914, 878, 760, 747, 700 cm1 ; 1 H NMR (CDCl3 ) d (ppm) 1.14 (s, 9 H, 3 CH3 ), 2.89 (d, 2 H, J ¼ 5.7 Hz, CH2 ), 3.59 (d, 1 H, J ¼ 3.0 Hz, OH), 5.13 (m, 1 H, CH), 7.29–7.39 (m, 5 H, aromatic); [a] 30 D þ61.5 (c 1.3, CHCl3 ). The enantioselectivity was determined to be 95% ee by HPLC analysis using a chiral column (Chiralcel OD-H, Daicel Chemical Industries, hexane–i-PrOH, 20:1, ﬂow rate 0.5 mL min1 ); t minor ¼ 17.7 min, t major ¼ 20.2 min. Typical Procedure for Silver(I)-catalyzed Asymmetric Mukaiyama Aldol Reaction of Benzaldehyde with Acetophenone Silyl Enol Ether 11. Synthesis of (S)-1Hydroxy-1,3-diphenyl-3-propanone (12, Scheme 1.4) [9]. Wet DMF containing 2% H2 O (206 mL) was added under nitrogen atmosphere to a mixture of AgPF6 (4.0 mg, 0.013 mmol) and (S)-BINAP (8.4 mg, 0.013 mmol) and the solution was stirred at room temperature for 10 min. Benzaldehyde (69 mL, 0.67 mmol) was added and stirring was continued for 10 min. Acetophenone silyl enol ether 11 (276 mL, 1.35 mmol) was added and the mixture was stirred for 2 h at 25 C. It was ﬁltered through a short silica gel column, concentrated in vacuo, and clear oil was obtained. The oil was hydrolyzed

1.5 Experimental Procedures

with 1 m HCl THF/H2 O (1:1) solution, extracted with ether, dried over anhydrous sodium sulfate, concentrated in vacuo to aﬀord a pale clear oil. This oil was puriﬁed by column chromatography on silica gel (1:1 ethyl acetate– hexane) to give the (S)-enriched aldol adduct 12 (100% yield, 69% ee). 1 H NMR (CDCl3 ) d (ppm) 3.38 (d, 2 H, J ¼ 5.9 Hz, CH2 ), 3.56 (d, 1 H, J ¼ 3.0 Hz, OH), 5.35 (dt, 1 H, J ¼ 3.0, 5.9 Hz, CH), 7.26–7.62 (m, 8 H, aromatic), 7.95 (d, 2 H, J ¼ 7.3 Hz, aromatic); FAB-MS calcd. for C15 H14 O2 , 226 (Mþ ); found 227 ((M þ 1)þ ), 209 ((M-OH)þ ); HPLC analysis using a chiral column (Chiralcel OB-H, Daicel Chemical Industries, hexane–i-PrOH, 9:1, ﬂow rate 0.9 mL min1 ); t minor ¼ 22.5 min (R), t major ¼ 33.5 min (S). Typical Procedure for Asymmetric Aldol Reaction of Trimethoxysilyl Enol Ether 13 with Benzaldehyde Catalyzed by (R)- p-Tol-BINAP-AgF Complex. Synthesis of 1-Hydroxy-2,4,4-trimethyl-1-phenyl-3-pentanone (5e, Scheme 1.5) [10]. A mixture of AgF (13.0 mg, 0.102 mmol) and (R)- p-Tol-BINAP (67.9 mg, 0.100 mmol) was dissolved in dry MeOH (6 mL) under an argon atmosphere and with direct light excluded, and stirred at 20 C for 10 min. Benzaldehyde (100 mL, 0.98 mmol) and t-butyl ethyl ketone-derived trimethoxysilyl enol ether 13 (236.9 mg, 1.01 mmol) were successively added dropwise, at 78 C, to the resulting solution. The mixture was stirred at this temperature for 2 h, then at 40 C for 2 h, and ﬁnally at 20 C for 2 h. It was then treated with brine (2 mL) and solid KF (ca. 1 g) at ambient temperature for 30 min. The resulting precipitate was removed by ﬁltration through a glass ﬁlter funnel ﬁlled with Celite and silica gel. The ﬁltrate was dried over Na2 SO4 and concentrated in vacuo after ﬁltration. The crude product was puriﬁed by column chromatography on silica gel (1:5 ethyl acetate–hexane as eluent) to aﬀord a mixture of the aldol adduct 5e (181.2 mg, 84% yield). The syn/anti ratio was determined to be >99/1 by 1 H NMR analysis. The enantioselectivity of the syn isomer was determined to be 97% ee by HPLC analysis using a chiral column (Chiralcel OD-H, Daicel Chemical Industries, hexane–i-PrOH, 40:1, ﬂow rate 0.5 mL min1 ); tsyn-minor ¼ 17.1 min, tsyn-major ¼ 18.0 min. Speciﬁc rotation of the syn isomer (95% ee) [a] 30 D 67.1 (c 1.3, CHCl3 ). Other spectral data (IR and 1 H NMR) of the syn isomer were in good agreement with reported data [6]. Typical Procedure for Asymmetric Mukaiyama Aldol Reaction of Silyl Enol Ethers with Aldehydes Catalyzed by Cationic BINAP-Pd(II) Complexes. Synthesis of (R)-1,3-Diphenyl-1-trimethylsiloxy-3-propanone (18, Scheme 1.7, and entry 1 in Table 1.2) [12, 13]. Wet DMF (8 mL dry DMF with 144 mL H2 O) was added to a mixture of (R)-BINAP-PdCl2 (160 mg, 0.20 mmol), AgOTf (51 mg, 0.20 mmol), and MS 4 A˚ (powder, 1.2 g) and the suspension was stirred at 23 C for 20 min. After cannula ﬁltration, benzaldehyde (410 mL, 4.0 mmol) and acetophenone silyl enol ether 11 (1.23 mL, 6.0 mmol) were added to the resulting orange solution and the mixture was stirred for 13 h at 23 C. Dilution of the reaction mixture with diethyl ether, ﬁltration

19

20

1 Silver, Gold, and Palladium Lewis Acids

through a short silica gel column, and concentration aﬀorded pale yellow oil. This crude product was puriﬁed by column chromatography on silica gel to give the (R)-enriched silylated aldol adduct 18 (1.04 g, 87% yield, 71% ee) and its desilylated product (82 mg, 9% yield, 73% ee). The enantioselectivity of the silylated product 18 was determined by HPLC analysis using Chiralcel OJ (hexane–i-PrOH, 9:1) after conversion to the corresponding desilylated product (1 m HClaTHF, 1:2). Speciﬁc rotation of the desilylated aldol adduct (70% ee) [a]D þ32.4 (c 0.74, MeOH). General Procedure for Asymmetric Aldol Reaction of Methyl Isocyanoacetate (27) with Aldehydes Catalyzed by Chiral Ferrocenylbisphosphine-Gold(I) Complexes (Scheme 1.8 and Table 1.3) [18]. Methyl isocyanoacetate (27, 5.0 mmol) was added to a solution of bis(cyclohexyl isocyanide)gold(I) tetraﬂuoroborate (0.050 mmol), chiral ferrocenylbisphosphine 26 (0.050–0.055 mmol), and aldehyde (5.0–5.5 mmol) in CH2 Cl2 (5 mL) and the mixture was stirred under nitrogen at 25 C until 27 was not detected by silica gel TLC (hexane–ethyl acetate, 2:1) or IR. Evaporation of the solvent followed by bulb-to-bulb distillation gave oxazoline 28. The trans/cis ratio was determined by 1 H NMR spectroscopy and the enantiomeric purity of trans-28 and cis-28, readily separated by MPLC (hexane–ethyl acetate), were determined by 1 H NMR studies using Eu(dcm)3 . The OCH3 singlet of the major enantiomer of trans-28 always appeared at a higher ﬁeld than that of the minor enantiomer. General Procedure for Asymmetric Aldol Reaction of Tosylmethyl Isocyanide (40) with Aldehydes Catalyzed by Chiral Ferrocenylbisphosphine-Silver(I) Complexes (Scheme 1.11) [31]. Aldehyde (1.5 mmol) was added to a solution of silver(I) triﬂate (0.011 mmol), chiral ferrocenylbisphosphine 26 (0.010 mmol), and tosylmethyl isocyanide (40, 1.0 mmol) in dry CH2 Cl2 (5 mL). The mixture was stirred under nitrogen at 25 C for 2 h. The catalyst was removed by passing the mixture through a bed of Florisil (17 mm 30 mm, EtOAc), and MPLC puriﬁcation (silica gel, CH2 Cl2 aEtOAc, 15:1) gave oxazoline 39. The trans/cis ratio was determined by 1 H NMR spectroscopy and the enantiomeric excess of trans-39 was determined by HPLC analysis with a chiral stationary phase after conversion to the corresponding anaphthylurea derivative of amino alcohol (1. LiAlH4 , 2. a-naphthyl isocyanate). Experimental Procedure for Pd(II)-Catalyzed Asymmetric Aldol Reaction of Methyl Isocyanoacetate (27) with Aldehydes (Scheme 1.13 and Table 1.4) [34]. Preparation of [(1R,1OR)-2,6-bis[1-(diphenylphosphino)ethyl]phenyl]chloropalladium(II) (46). PdCl2 (PhCN)2 (383 mg, 1.0 mmol) was added to a solution of (1R,1 0 R)-1,3-bis[1-(diphenylphosphino)ethyl]benzene (502 mg, 1.0 mmol) in CH2 Cl2 (15 mL). The resulting orange solution was stirred at room tem-

References

perature for 24 h. The reaction mixture was then reduced to one-third of its volume and absolute ethanol was added to precipitate the product. Filtration gave the desired product 46 as a yellow powder (418 mg, 85%), mp 249–252 C; [a]D 323.6 (c 1.0, CHCl3 ). 1 H NMR (360 MHz, CDCl3 ) d (ppm) 1.19 (m, 6H, CH3 ), 4.04 (m, 2H, CH), 7.10 (m, 3H), 7.34–7.45 (m, 12H), 7.70– 7.74 (m, 4H), 7.95–7.97 (m, 4H); 13 C NMR (90 MHz, CDCl3 ) d (ppm) 22.1 (s, 2C, CH3 ), 46.7 (m, 2C, CH), 122.5–156.8 (twelve diﬀerent aromatic carbon atoms); 31 P NMR (145 MHz, CDCl3 ) d (ppm) 46.5. HRMS calculated for C34 H31 P2 PdCl (Mþ ) 642.0624; found 642.0634. General Procedure for the Aldol Reaction. A solution of Pd complex 46 (7 mg, 0.011 mmol, 1.0 mol%) and AgOTf (3 mg, 0.011 mmol) in CH2 Cl2 (2 mL) was stirred for ca. 30 min at room temperature. The resulting cloudy solution was ﬁltered through Celite and the solvent was removed under reduced pressure to give the active catalyst. The catalyst was then dissolved in EtOAc and passed through a plug of silica gel to remove excess AgOTf. After removal of EtOAc the catalyst was dissolved in THF (6 mL), and methyl isocyanoacetate (27, 110 mL, 1.1 mmol) was added followed by introduction of diisopropylethylamine (19 mL, 0.11 mmol) and aldehyde (1.1 mmol). The reaction was monitored by TLC (EtOAc–hexane, 1:1, visualized with KMnO4 ). After removal of solvent, the pure product 47 was obtained by bulb-to-bulb distillation under reduced pressure (0.1 mmHg). The trans/cis ratio was determined by 1 H NMR spectroscopy by integration of the methyl ester protons and the enantiomeric excess for trans-47 and cis-47 were determined by GC analysis. References 1 (a) B. M. Trost, in Comprehensive Organometallic Chemistry,

Vol. 8, (ed.: G. Wilkinson), Pergamon Press, Oxford, 1982, Chapter 57. (b) S. A. Godleski, in Comprehensive Organic Synthesis, Vol. 4, (eds.: B. M. Trost, I. Fleming), Pergamon Press, New York, 1991, p. 585. (c) A. Pfaltz, M. Lautens, in Comprehensive Asymmetric Catalysis, Vol. 2, (eds.: E. N. Jacobsen, A. Pfaltz, H. Yamamoto), Springer, Heidelberg, 1999, Chapter 24, p. 833. (d) B. M. Trost, C. B. Lee, in Catalytic Asymmetric Synthesis, 2nd ed., (ed.: I. Ojima), Wiley– VCH, New York, 2000, Chapter 8E, p. 593. 2 (a) D. R. Rae, in Encyclopedia of Reagents for Organic Synthesis, Vol. 6, (ed.: L. A. Paquette), John Wiley & Sons, Chichester, 1995, p. 4461. (b) J. C. Lanter, in Encyclopedia of Reagents for Organic Synthesis, Vol. 6, (ed.: L. A. Paquette), John Wiley & Sons, Chichester, 1995, p. 4469. (c) L.-G. Wistrand, in Encyclopedia of Reagents for Organic Synthesis, Vol. 6, (ed.: L. A. Paquette), John Wiley & Sons, Chichester, 1995, p. 4472. (d) T. H. Black, in Encyclopedia of Reagents for Organic Synthesis, Vol. 6, (ed.: L. A. Paquette), John Wiley & Sons, Chichester, 1995, p. 4476.

21

22

1 Silver, Gold, and Palladium Lewis Acids 3 S. Castellino, J. J. Sims, Tetrahedron Lett. 1984, 25, 4059. 4 (a) A. Yanagisawa, H. Nakashima, A. Ishiba, H. Yamamoto,

5

6 7

8

9

10

11 12

13 14

15 16 17 18

19

J. Am. Chem. Soc. 1996, 118, 4723. See also: (b) C. Bianchini, L. Glendenning, Chemtracts–Inorg. Chem. 1997, 10, 339; (c) P. G. Cozzi, E. Tagliavini, A. Umani-Ronchi, Gazz. Chim. Ital. 1997, 127, 247. Reviews: (a) T. Bach, Angew. Chem. Int. Ed. Engl. 1994, 33, 417; (b) T. K. Hollis, B. Bosnich, J. Am. Chem. Soc. 1995, 117, 4570; (c) M. Braun, in Houben–Weyl: Methods of Organic Chemistry, Vol. E 21, (eds.: G. Helmchen, R. W. Hoffmann, J. Mulzer, E. Schaumann), Georg Thieme Verlag, Stuttgart, 1995, p. 1730; (d) S. G. Nelson, Tetrahedron: Asymmetry 1998, 9, 357; (e) H. Gro¨ger, E. M. Vogl, M. Shibasaki, Chem. Eur. J. 1998, 4, 1137. A. Yanagisawa, Y. Matsumoto, H. Nakashima, K. Asakawa, H. Yamamoto, J. Am. Chem. Soc. 1997, 119, 9319. (a) M. Pereyre, B. Bellegarde, J. Mendelsohn, J. Valade, J. Organomet. Chem. 1968, 11, 97; (b) I. F. Lutsenko, Y. I. Baukov, I. Y. Belavin, J. Organomet. Chem. 1970, 24, 359; (c) S. S. Labadie, J. K. Stille, Tetrahedron 1984, 40, 2329; (d) K. Kobayashi, M. Kawanisi, T. Hitomi, S. Kozima, Chem. Lett. 1984, 497. (a) A. Yanagisawa, Y. Matsumoto, K. Asakawa, H. Yamamoto, J. Am. Chem. Soc. 1999, 121, 892. (b) A. Yanagisawa, Y. Matsumoto, K. Asakawa, H. Yamamoto, Tetrahedron 2002, 58, 8331. (a) M. Ohkouchi, M. Yamaguchi, T. Yamagishi, Enantiomer 2000, 5, 71. (b) M. Ohkouchi, D. Masui, M. Yamaguchi, T. Yamagishi, J. Mol. Catal. A: Chem. 2001, 170, 1. (a) A. Yanagisawa, Y. Nakatsuka, K. Asakawa, H. Kageyama, H. Yamamoto, Synlett 2001, 69. (b) A. Yanagisawa, Y. Nakatsuka, K. Asakawa, M. Wadamoto, H. Kageyama, H. Yamamoto, Bull. Chem. Soc. Jpn. 2001, 74, 1477. D. Ferraris, B. Young, T. Dudding, T. Lectka, J. Am. Chem. Soc. 1998, 120, 4548. (a) M. Sodeoka, K. Ohrai, M. Shibasaki, J. Org. Chem. 1995, 60, 2648. Review: (b) M. Sodeoka, M. Shibasaki, Pure Appl. Chem. 1998, 70, 411. M. Sodeoka, R. Tokunoh, F. Miyazaki, E. Hagiwara, M. Shibasaki, Synlett 1997, 463. (a) E. Hagiwara, A. Fujii, M. Sodeoka, J. Am. Chem. Soc. 1998, 120, 2474. (b) A. Fujii, E. Hagiwara, M. Sodeoka, J. Am. Chem. Soc. 1999, 121, 5450. A. Fujii, M. Sodeoka, Tetrahedron Lett. 1999, 40, 8011. H. Doucet, J.-L. Parrain, M. Santelli, Synlett 2000, 871. O. Fujimura, J. Am. Chem. Soc. 1998, 120, 10032. (a) Y. Ito, M. Sawamura, T. Hayashi, J. Am. Chem. Soc. 1986, 108, 6405; (b) Y. Ito, M. Sawamura, T. Hayashi, Tetrahedron Lett. 1987, 28, 6215; (c) T. Hayashi, M. Sawamura, Y. Ito, Tetrahedron 1992, 48, 1999. Reviews: (a) M. Sawamura, Y. Ito, Chem. Rev. 1992, 92, 857; (b) M. Sawamura, Y. Ito, in Catalytic Asymmetric Synthesis, (ed.: I. Ojima), VCH, New York, 1993, p. 367.

References 20 (a) S. D. Pastor, Tetrahedron 1988, 44, 2883; (b) S. D. Pastor,

21 22 23 24 25

26 27 28 29

30 31 32

33 34 35 36 37 38 39

40 41

A. Togni, J. Am. Chem. Soc. 1989, 111, 2333; (c) A. Togni, S. D. Pastor, Helv. Chim. Acta 1989, 72, 1038; (d) A. Togni, S. D. Pastor, J. Org. Chem. 1990, 55, 1649; (e) A. Togni, R. Ha¨usel, Synlett 1990, 633; (f ) S. D. Pastor, A. Togni, Tetrahedron Lett. 1990, 31, 839; (g) A. Togni, S. D. Pastor, G. Rihs, J. Organomet. Chem. 1990, 381, C21; (h) S. D. Pastor, A. Togni, Helv. Chim. Acta 1991, 74, 905. Y. Ito, M. Sawamura, T. Hayashi, Tetrahedron Lett. 1988, 29, 239. A. Togni, S. D. Pastor, G. Rihs, Helv. Chim. Acta 1989, 72, 1471. M. D. Bachi, A. Melman, J. Org. Chem. 1997, 62, 1896. M. Sawamura, Y. Ito, T. Hayashi, Tetrahedron Lett. 1990, 31, 2723. (a) Y. Ito, M. Sawamura, E. Shirakawa, K. Hayashizaki, T. Hayashi, Tetrahedron Lett. 1988, 29, 235. See also: (b) Y. Ito, M. Sawamura, E. Shirakawa, K. Hayashizaki, T. Hayashi, Tetrahedron 1988, 44, 5253. Y. Ito, M. Sawamura, M. Kobayashi, T. Hayashi, Tetrahedron Lett. 1988, 29, 6321. Y. Ito, M. Sawamura, H. Hamashima, T. Emura, T. Hayashi, Tetrahedron Lett. 1989, 30, 4681. M. Sawamura, Y. Nakayama, T. Kato, Y. Ito, J. Org. Chem. 1995, 60, 1727. (a) A. Togni, S. D. Pastor, Tetrahedron Lett. 1989, 30, 1071; (b) M. Sawamura, Y. Ito, T. Hayashi, Tetrahedron Lett. 1989, 30, 2247. T. Hayashi, Y. Uozumi, A. Yamazaki, M. Sawamura, H. Hamashima, Y. Ito, Tetrahedron Lett. 1991, 32, 2799. M. Sawamura, H. Hamashima, Y. Ito, J. Org. Chem. 1990, 55, 5935. (a) V. A. Soloshonok, T. Hayashi, Tetrahedron Lett. 1994, 35, 2713; (b) V. A. Soloshonok, T. Hayashi, Tetrahedron: Asymmetry 1994, 5, 1091; (c) V. A. Soloshonok, A. D. Kacharov, T. Hayashi, Tetrahedron 1996, 52, 245. M. A. Stark, C. J. Richards, Tetrahedron Lett. 1997, 38, 5881. J. M. Longmire, X. Zhang, M. Shang, Organometallics 1998, 17, 4374. G. Guillena, G. Rodrı´guez, G. van Koten, Tetrahedron Lett. 2002, 43, 3895. Y. Motoyama, H. Kawakami, K. Shimozono, K. Aoki, H. Nishiyama, Organometallics 2002, 21, 3408. ¨ntener, M. Wo¨rle, Helv. R. Nesper, P. S. Pregosin, K. Pu Chim. Acta 1993, 76, 2239. C. Schlenk, A. W. Kleij, H. Frey, G. van Koten, Angew. Chem. Int. Ed. 2000, 39, 3445. A. W. Kleij, R. J. M. Klein Gebbink, P. A. J. van den Nieuwenhuijzen, H. Kooijman, M. Lutz, A. L. Spek, G. van Koten, Organometallics 2001, 20, 634. M. D. Meijer, N. Ronde, D. Vogt, G. P. M. van Klink, G. van Koten, Organometallics 2001, 20, 3993. G. Rodriguez, M. Lutz, A. L. Spek, G. van Koten, Chem. Eur. J. 2002, 8, 45.

23

25

2

Boron and Silicon Lewis Acids for Mukaiyama Aldol Reactions Kazuaki Ishihara and Hisashi Yamamoto 2.1

Achiral Boron Lewis Acids 2.1.1

Introduction

The classical boron Lewis acids, BX3 , RBX2 and R2 BX (X ¼ F, Cl, Br, I, OTf ) are now popular tools in organic synthesis. B(III) can act as a Lewis acid because there is an empty p-orbital on the boron. Enthalpy values indicate that when pyridine is the reference base, the Lewis acidities of Group IIIB halides increase in the order AlX3 > BX3 > GaX3 . The Lewis acidity of BX3 generally increases in the order ﬂuoride < chloride < bromide < iodide, i.e. the exact reverse of the order expected on the basis of relative sdonor strengths of the halide anions. The main reason for this anomaly is that in these BX3 compounds the BaX bonds contain a p-component which is formed by overlap of a ﬁlled p-orbital on the halogen with the empty porbital on the boron. Because the latter orbital is used to form a s-bond when BX3 coordinates with a Lewis base, this p-component is completely destroyed by complex formation. The strength of the p-component now increases in the order iodide < bromide < chloride < ﬂuoride, i.e. the amount of p-bond energy that is lost on complex formation increases as the atomic weight of the halogen decreases. Evidently, as far as the extent of complex formation is concerned, this is a more important factor than the corresponding decrease in the s-donor strength of the halogen. The BF3 and BCl3 complexes of diethyl ether are less stable than those of dimethyl ether, and the same order of stability is observed for complexes of diethyl and dimethyl sulﬁdes. As expected, steric interaction decreases as the distance between the metal and ligand atom is increased. Thus, it decreases when the metal atom is changed from boron to aluminum, and when the ligand atom is changed from oxygen to sulfur. The classical boron Lewis acids are used stoichiometrically in Mukaiyama Modern Aldol Reactions. Vol. 2: Metal Catalysis. Edited by Rainer Mahrwald Copyright 8 2004 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 3-527-30714-1

26

2 Boron and Silicon Lewis Acids for Mukaiyama Aldol Reactions

aldol reactions under anhydrous conditions, because the presence of even a small amount of water causes rapid decomposition or deactivation of the promoters. To obviate some of these inherent problems, the potential of arylboron compounds, Arn B(OH)n3 (n ¼ 1–3), bearing electron-withdrawing aromatic groups as a new class of boron catalyst has recently been demonstrated. For example, tris(pentaﬂuorophenyl)borane, B(C6 F5 )3 , is a convenient, commercially available Lewis acid of strength comparable with that of BF3 , but without the problems associated with reactive BaF bonds. Although its primary commercial application is as a co-catalyst in metallocenemediated oleﬁn polymerization, its potential as a Lewis acid catalyst for Mukaiyama aldol reactions is now recognized as being much more extensive. Diarylborinic acids and arylboronic acids bearing electron-withdrawing aromatic groups are also highly eﬀective Lewis acid catalysts [1]. 2.1.2

BF3 .Et2 O

Although TiCl 4 is a better Lewis acid at eﬀecting aldol reactions of aldehydes, acetals, and silyl enol ethers, BF3 .Et2 O is more eﬀective for aldol reactions with anions generated from transition metal carbenes and with tetrasubstituted enol ethers such as (Z)- and (E)-3-methyl-2-(trimethylsilyloxy)-2-pentane [2, 3]. One exception is the preparation of substituted cyclopentanediones from acetals by aldol condensation of protected fourmembered acyloin derivatives with BF3 .Et2 O rather than TiCl 4 (Eq. (1)) [2, 4]. Use of the latter catalyst results in some loss of the silyl protecting group. The pinacol rearrangement is driven by the release of ring strain in the four-membered ring and is controlled by an acyl group adjacent to the diol moiety.

O + R

R'

Me3SiO

OSiMe3

BF3•Et2O 89% yield

SiMe3 HO O O R R' O

(1)

R R' O

This reagent is the best promoter of the aldol reaction of 2-(trimethylsiloxy)acrylate esters, prepared by the silylation of pyruvate esters, to aﬀord galkoxy-a-keto esters (Eq. (2)) [5] These esters occur in a variety of important natural products.

2.1 Achiral Boron Lewis Acids

OSiMe3 OEt

OMe + Ph

27

OMe

O

ð2Þ

OMe O

BF3•Et2O

OEt

Ph

CH2Cl2 –78 °C to 0 °C 86% yield

O

BF3 .Et2 O can improve or reverse aldehyde diastereofacial selectivity in the aldol reaction of silyl enol ethers with aldehydes, to give syn adducts. For example, Heathcock and Flippin have reported that the reaction of the silyl enol ether of pinacolone with 2-phenylpropanal using BF3 .Et2 O results in enhanced Felkin selectivity (up to 36:1) compared with addition of the corresponding lithium enolate [6, 7]. When the a-substituents are more subtly diﬀerentiated, however, it is still diﬃcult to achieve acceptable levels of selectivity. Davis et al. have reported that use of triisopropylsilyl enol ether and i-Pr3 SiB(OTf )4 results in selectivity of ca. 100:1 with 2-phenylpropanal and a useful level of 7:1 with 2-benzylpropanal (Eq. (3)) [8]. Control experiments employing BF3 .Et2 O catalysis and 2-benzylpropanal as substrate results in lower selectivity (ca. 3:1) that does not depend substantially on the bulk of the silyl group in the enolate (Eq. (3)). In contrast, both levels of 1,2asymmetric induction in the i-Pr3 SiB(OTf )4 (5 mol%)- and the BF3 .Et2 O (1 equiv.)-promoted additions of silyl ketene thioacetals to a-asymmetric aldehydes are aﬀected by the bulk of the silyl group (Eqs. (4) and (5)) [8].

OTIPS + Bn

CHO

X

X=Me, t-Bu, OMe, Ot-Bu

Lewis acid

X

Bn TIPSO

i-Pr3SiB(OTf)4 BF3•Et2O

+

O

Cram 7 3

X

Bn TIPSO

: :

O

anti-Cram 1 1

ð3Þ

OR + Bn

CHO

i-Pr3SiB(OTf)4

St-Bu (5 mol%)

St-Bu + Bn

Bn

R=TIPS R=TBDMS

RO

O

Cram 5.5 3.6

St -Bu RO

:

O

anti-Cram 1 1

ð4Þ

28

2 Boron and Silicon Lewis Acids for Mukaiyama Aldol Reactions

OR

BF3•Et2O

+ Bn

CHO

St-Bu

(1 equiv)

St-Bu + Bn

Bn HO

R=TIPS R=TBDMS

St-Bu

O

Cram 13 5.8

HO

O

anti-Cram 1 1

:

ð5Þ

Addition of the tetrasubstituted selenoketene silyl acetal to b-benzyloxy aldehyde in the presence of Et2 BOTf leads to good yields and high ratios of products with 3,4-anti relative stereochemistry (Cram chelate model) (Eq. (6)). In contrast, reversed diastereoselectivity and a synthetically interesting ratio of 1:11 in favor of the 3,4-syn products is obtained when the bulky bsilyloxy aldehyde is used in the presence of the monodentate Lewis acid BF3 .OEt2 (Felkin–Anh model) (Eq. (7)) [9].

BnO

OSiMe3

O +

H

OMe SePh

Et2BOTf (1.2 equiv) CH2Cl2, –78 °C

BnO

OH O

OMe SePh 86% yield, syn:anti=<1:20

ð6Þ

TBDPSO

OSiMe3

O H

BnO

+

OMe SePh

BF3•OEt2 (1.5 equiv) CH2Cl2, –78 °C

TBDPSO

OMe SePh 84% yield, 2,3-(syn:anti )=11:1

BEt2OTf Me O

O Me

H H

enol ether

Cram chelate

enol ether

OH O

BF3

HH Felkin-Anh

OTBDPS

ð7Þ

The subsequent free-radical-based hydrogen transfer reaction of the products obtained from the Mukaiyama reactions occurs under the control of the exocyclic eﬀect. Excellent diastereoselectivity favoring the 2,3-anti relative stereochemistry is obtained from the reduction with Bu3 SnH in the presence of Et2 BOTf and N,N-diisopropylethylamine (Eq. (8)) [9]. The initial product of the aldol condensation is a b-hydroxy carbonyl compound, which is often transformed into the corresponding a,bunsaturated derivative. Interestingly, gaseous BF3 , which is a much stronger Lewis acid than BF3 .Et2 O promotes the fragmentation reaction from b-aryl-

2.1 Achiral Boron Lewis Acids

Et2BOTf (1.2~1.5 equiv) i-Pr2NEt (1.5 equiv) Et3B (0.2 equiv)

OR OH O

OR OH O

+ Bu3SnH OMe (2 equiv) CH2Cl2, –78 °C SePh

OMe 2,3-(syn:anti )=<1:20

RO TfOEt2B

Me

CO2Me

O

H

ð8Þ

HSnBu3 Exocyclic effect

b-hydroxyketones to (E)-arylalkenes and carboxylic acids in good yields (Eq. (9)) [10]. The combination of powerful Lewis acid and non-nucleophilic solvents, for example CCl 4 , are keys to this unexpected behavior. BCl3 and BBr3 are less eﬀective in this reaction. OH O Ph

BF3 (excess) R'

O

R

Ph

+ HO

CCl4, reflux

R BF3

Ph

–BF3

H O+ OB–F3 R' R

+

R'

(9)

HO R' OB–F3

Ph R

2.1.3

B(C6 F5 )3

B(C6 F5 )3 is an air-stable, water-tolerant Lewis acid catalyst, which can be readily prepared as a white solid by reacting BCl3 with C6 F5 Li [11, 12]. This compound does not react with pure oxygen [12]. It is very thermally stable, even at 270 C, and is soluble in many organic solvents [12]. Although B(C6 F5 )3 catalyzes reactions most eﬀectively under anhydrous conditions, B(C6 F5 )3 exposed to air is also available (not anhydrous grade). Mukaiyama aldol reactions of a variety of silyl enol ethers or ketene silyl acetals with aldehydes or other electrophiles proceed smoothly in the presence of 2 mol% B(C6 F5 )3 [13a,c]. The following characteristic features can be noted:

. the products can be isolated as b-trimethylsilyloxy ketones when crude adducts are worked-up without exposure to acid;

29

30

2 Boron and Silicon Lewis Acids for Mukaiyama Aldol Reactions

. the reaction can be conducted in aqueous media, so reaction of the silyl . .

enol ether derived from propiophenone with a commercial aqueous solution of formaldehyde does not present any problems; the rate of an aldol reaction is markedly increased by use of an anhydrous solution of B(C6 F5 )3 in toluene under an argon atmosphere; and silyl enol ethers can be reacted with chloromethyl methyl ether or trimethylorthoformate; hydroxymethyl, methoxymethyl, or dimethoxymethyl C1 groups can be introduced at the position a to the carbonyl group.

These aldol-type reactions do not proceed when triphenylborane is used (Eq. (10)).

R1CHO

OSiMe3 R

+

or other electrophiles

2

R4 R3

Examples (yield) OH O Ph

Et

1) B(C6F5)3 (2-10 mol%)

Ph

65%

Ph

O MeO

(10)

OMe O Ph

72%

OMe

96%

90%

Ph

R4 R2 R3

OH O Et

O HO

R1

2) HCl or TBAF

OH O

94%

OH O

MeO

Ph

65%

2.1.4

Ar2 BOH

Diarylborinic acids, Ar2 BOH, bearing electron-withdrawing aromatic groups are eﬀective catalysts for Mukaiyama aldol condensation and subsequent selective dehydration of b-hydroxy carbonyl compounds [14]. The catalytic activity of diarylborinic acids (C6 F5 )2 BOH and [3,5-(CF3 )2 C6 H3 ]2 BOH in Mukaiyama aldol reactions is much higher than that of the corresponding arylboronic acids. It is worthy of note that small amounts of E-isomeric dehydrated product has been isolated in reactions catalyzed by diarylborinic acids (C6 F5 )2 BOH and [3,5-(CF3 )2 C6 H3 ]2 BOH. In contrast, no dehydrated products have been isolated in the presence of (C6 F5 )3 B, despite its extremely high catalytic activity (Eq. (11)). Signiﬁcant features of these active borinic acid catalysts are that they are strong Lewis acids and have a hydroxy group on the boron atom. Dehydration is strongly favored in THF. The reaction usually proceeds smoothly, and a,b-enones are obtained in high yields as E isomers. In reactions of a-substituted b-hydroxy carbonyl compounds, a,b-enones are preferentially

2.1 Achiral Boron Lewis Acids

+

PhCHO

OSiMe3

ArnB(OH)3-n (2 mol%)

Ph

CH2Cl2, -78 °C

1) 1N NaOH 2) 1N HCl-THF

OH O Ph

O

+ Ph

Ph

Ph

C6F5B(OH)2

0%

(C6F5)2BOH

89%

7%

[3,5-(CF3)2C6H3]2BOH

89%

10%

B(C6F5)3

98%

0%

(11)

0%

obtained from anti aldols whereas most of the syn aldols are recovered. This dehydration is, therefore, a useful and convenient method for isolating pure syn aldols from syn/anti isomeric mixtures (Eq. (12)). Reaction of the bhydroxy function with the diarylborinic acid leads to a cyclic intermediate, which should be susceptible to dehydration. Subsequent transformation to a,b-enones occurs via an enolate intermediate resulting from selective abstraction of a pseudo-axial a-proton perpendicular to the carbonyl face. A cyclic intermediate formed from a syn aldol and a diarylborinic acid would be thermodynamically less stable than the cyclic intermediate. Thus, dehydration to (E)-a,b-enones occurs selectively for anti-aldols. (C6F5)2BOH (5-10 mol%)

OH O R1

R2 R3

O

THF, ambient temp.

R1

OH O

+ R1

R2 R3

R2 R3

Ar2BOH H 2 R

O Ar O B Ar

R3 R1

R1

O

R2 O B– Ar

R3 H

Ar

H

H+

(12)

Examples O Ph

O Ph

Ph

Ph

Ph

Bu

>99%

>99% OH O Ph

O

97% OH O

O Ph

syn:anti=71:29

Ph

Ph

35%

+

Ph

Ph

65% (>99% syn)

Diphenylborinic acid (10 mol%), which is stable in water, is an eﬀective catalyst for the Mukaiyama aldol reaction in the presence of benzoic acid (1

31

32

2 Boron and Silicon Lewis Acids for Mukaiyama Aldol Reactions

mol%) as a co-catalyst and sodium dodecyl sulfate (10 mol%) as a surfactant (Eq. (13)) [15]. Use of water as solvent is essential in this reaction. The reaction proceeds sluggishly in organic solvents such as dichloromethane and diethyl ether. Much lower yield than in water is obtained under neat conditions. Not only aromatic aldehydes but also a,b-unsaturated and aliphatic aldehydes give high syn selectivity. Although lower diastereoselectivity is observed when E enolates are used, reverse diastereoselectivity is observed when both stereoisomers of the silyl enolate derived from tert-butyl thiopropionate are used. Ph2BOH (10 mol%) C12H25SO3Na (10 mol%) PhCO2H (1 mol%)

OSiMe3

PhCHO +

H2O, 30 °C <24 h 72 h

St-Bu 98% E 97% Z

OH O Ph

St-Bu

62% yield, 96% syn 84% yield, 61% anti

ð13Þ This reaction can be explained by a mechanism via the boron enolate as reaction intermediate generated by SiaB exchange (Figure 2.1). That the diastereoselectivity is reversed by using the stereoisomers of the silyl enolate supports the hypothesis because this type of reversal has also been observed in the traditional boron enolate mechanism which involves a chair-like sixmembered transition-state. Furthermore, the trend that anti selectivity is poorer than syn selectivity in the reactions is also found in the traditional boron enolate-mediated aldol reactions. The mechanism is based on the hypothesis that Ph2 BOH can react with a silyl enolate to form the corresponding boron enolate under these conditions. When a Z enolate is used, an aldehyde and the boron enolate react via a chair-like six-membered tranOH O Ph

OSiMe3 St-Bu

St-Bu Ph2BOH

H2O Ph2BO Ph

Me3SiOH O

OBPh2 St-Bu

St-Bu

PhCHO Fig. 2.1

Proposed boron enolate mechanism.

2.2 Chiral Boron Lewis Acids

sition state to give the syn aldol product. The BaO bond of the initial aldol product is presumed to be easily cleaved by hydrolysis, and Ph2 BOH can be regenerated. In this mechanism benzoic acid might accelerate the SiaB exchange step, which is thought to be rate-determining.

2.2

Chiral Boron Lewis Acids 2.2.1

Introduction

Asymmetric aldol synthesis has recently been the focus of intense interest. Especially worthy of note is the development of homogeneous catalytic enantioselective Mukaiyama aldol reactions in which a small amount of chiral ligand can induce asymmetry for a given reaction. The possible applications depend on the selectivity of the homogeneous catalysts, which are of great interest because they provide simple methods for synthesizing complex molecules for which enantiocontrol is needed. This section addresses chirally modiﬁed boron Lewis acid complexes, in which there has been increased interest because of their capacity to induce chirality. They have been successfully used for Mukaiyama aldol reactions. 2.2.2

Chiral Boron Lewis Acids as Stoichiometric Reagents

This reaction of silyl ketene acetals with aldehydes using 1 as a stoichiometric chiral reagent was originally reported by Reetz et al. (Eq. (14)) [16]. The aldol addition of 1-(trimethylsiloxy)-1-methoxy-2-methyl-1-propene and 3-methylbutanal provides the aldol in only 57% yield, but with 90% ee.

Ph

i-BuCHO +

OTMS OMe

Ph B Cl 1 (1 equiv)

CH2Cl2 -78 °C to rt

OH i-Bu

(14) CO2Me

57% yield, 90% ee

The use of chiral acyloxyborane (CAB) as a chiral reagent seems to be more eﬀective for this reaction, which proceeds faster and with higher yields and enantiomeric excess. Kiyooka and his colleagues ﬁrst described the use of a variety of chiral oxaborolidines, derived from sulfonamides of a-amino acids and borane, in the course of the selective aldol reaction between silyl ketene acetals and aldehydes (Eq. (15)) [17a]. This reaction gives b-hydroxy esters in high enantioselectivity and yields.

33

34

2 Boron and Silicon Lewis Acids for Mukaiyama Aldol Reactions

O

i-Pr TsN

OTMS

+

R1CHO

OEt

O B 2a H (1 equiv)

R1 *

CH2Cl2 -78 °C to rt

(15)

OH CO2Et

77-87% yield, 83-93% ee

The role played by the trialkylsilyl group is unclear. Changing the trimethylsilyl group increases not only the selectivity but also the product of the reaction – b-hydroxy acetals are now obtained instead of b-hydroxy esters. They investigated the course of the reaction of a variety of tert-butyldimethylsilyl ketene acetals and aldehydes with 2a (Eq. (16)). The acetal is probably formed by hydride transfer to an intermediate ester. The tert-butyl group apparently stabilizes the second intermediate and consequently changes the course of the reaction. It should be noted that the ﬁrst cyclic intermediate is stabilized by coordination of the borane with the oxygen of the carbonyl. The results are outstanding when R2 ¼ Me (92–98%). Selectivity and yield decrease, however, when R2 ¼ H (45–62%) (Eq. (16)). R2 R1CHO +

OTBDMS

R2

OEt i-Pr TsN O Ph

B

2a (1 equiv) CH2Cl2, -78 °C

O

i-Pr

O TBDMS

TsN

H CO2Et

O

O

B

O OTBDMS

(16) OEt

Ph R2 R 2

R2 R 2 OH OTBDMS 1

OEt R * R2 R2 R2=H: 77-82% yield, 45-62% ee R2=Me: 79-85% yield, 92-98% ee

A stoichiometric amount of 2a catalyzed the asymmetric aldol reaction of aldehydes with enol silyl ethers and subsequent asymmetric reduction in one pot to aﬀord syn 1,3-diols with high enantioselectivity (Eq. (17)) [17b]. With a variety of aldehydes, 1,3-diols were obtained in moderate yields (53– 70%) with high syn diastereoselectivity. The syn 1,3-diols prepared from aliphatic aldehydes in the reaction (EtCN) are almost enantiomerically pure

2.2 Chiral Boron Lewis Acids

(96–99% ee). Propionitrile was the best solvent for reaction selectivity. The TBDMS substituent of enol silyl ether reduced the syn selectivity. OTMS

2a (1 equiv)

OH OH

OH O +

i-PrCHO + Ph

EtCN –78 °C, 3 h

i-Pr

Ph

i-Pr

65% yield syn:anti=97:3 99% ee

Ph

17% yield 60% ee (S)

ð17Þ Thus, good to excellent diastereo- and enantioselectivity are achieved simultaneously with 1,3-diols whereas the enantioselectivity for b-hydroxy ketones is substantially lower. These observations on the selectivity of the products suggest that syn-selective reduction of the reaction intermediate occurs after enantioselective aldol addition. On the basis of Kiyooka’s working hypothesis for the aldol reaction mechanism, the reduction proceeds via intramolecular hydride transfer, which is accelerated by a matching mode between the promoter’s chirality and that of the newly formed aldol (Eq. (18)). An alternative mechanism without chelation is also possible; this involves hydride delivery to the preferred Osilyl oxocarbenium ion conformer (Eq. (19)). i-Pr

i-Pr

O

i-Pr

Ts N B O O H i-Pr

H

O

SiMe3

O N O B H OTMS O

O

Ts

i-Pr

(18)

O

i-Pr

Ts N B O O H

Ts N O B O H Ph i-Pr

i-Pr

i-Pr

H

O

SiMe3 Ph

Ph

i-Pr

Ph

Ph

Ts N B O O OTMS

i-Pr O

O+TMS

i-Pr Ts N O O B O H + O TMS i-Pr Ph

O

Ts N B O O OTMS i-Pr

Ph

(19)

35

36

2 Boron and Silicon Lewis Acids for Mukaiyama Aldol Reactions

A very short asymmetric synthesis of an insect attractant, (1S,3S,5R)1,3-dimethyl-2,9-dioxabicyclo[3.3.1]nonane, a host-speciﬁc substance for the ambrosia beetle that infests the bark of the Norway spruce, has been realized with a 2a-mediated aldol reaction strategy; enantio- and diastereoselectivity were high (Eq. (20)) [17c].

O

O

O

EtCN –78 °C, 3 h

OH OTMS

O

2a (1 equiv)

OTMS

+

CHO

+

36% yield, 94% ee

OH OH

O

O

+

21% yield, 72% ee

O

O

OH O

24% yield, 43% ee

1. TBAF, THF 2. TsOH, CH2Cl2

ð20Þ O

O 78% yield

The chiral borane 2a-mediated aldol reaction proceeds with a-chiral aldehydes in a reagent-controlled manner. Both enantiomers are obtained almost optically pure from one racemic aldehyde (Eqs. (21) and (22)) [17d].

(S)-2a (1 equiv) CH2Cl2, -78 °C, 3 h

Ph

CO2Et

+

Ph

OH

CO2Et

ð21Þ

OH

44% yield, >99% ee

20% yield, >99% ee

OTMS + Ph

CHO

OEt

(R)-2a (1 equiv) CH2Cl2, -78 °C, 3 h

Ph

CO2Et OH

18% yield, >99% ee

+

Ph

CO2Et OH

(22)

41% yield, >99% ee

The reaction of b-chiral aldehydes with ketene silyl acetals gives both syn and anti aldols in similar yields without any Cram selectivity (Eq. (23)) [17d].

2.2 Chiral Boron Lewis Acids

OTMS CHO +

OEt

OBn

ð23Þ

(S)-2a (1 equiv) CH2Cl2, -78 °C, 3 h

CO2Et

+

OBn OH 46% yield, 98% ee

CO2Et OBn OH 42% yield, 82% ee

In Kiyooka’s approach to acetate aldols using a stoichiometric amount of 2a, a serious reduction (ca. 10–20%) in enantiomeric excess was observed in the reaction with silyl ketene acetals derived from a-unsubstituted acetates, compared with the high level of enantioselectivity (> 98% ee) in the reaction with 1-ethoxy-2-methyl-1-(trimethylsiloxy)-1-propene. Introduction of an eliminable substituent, e.g. a methylthio or bromo substituent, after aldol reaction at the a-position of chiral esters resolves this problem [17e]. Asymmetric synthesis of dithiolane aldols has been achieved in good yields by using the silyl ketene acetal derived from 1,3-dithiolane-2-carboxylate in the 2a-promoted aldol reaction; desulfurization of the dithiolane aldols produces the acetate aldols in high enantiomeric purity (Eq. (24)).

PhCHO +

S

OTMS

S

OEt

OH O

2a (1 equiv) Ph CH2Cl2, -78 °C, 3 h

S

S

OEt

88% yield Ni2B-H2

(24)

OH O Ph

OEt

85% yield, 98% ee (S)

A very short asymmetric synthesis of the bryostatin C1 aC9 segment has been achieved using three sequential 2a-promoted aldol reactions under reagent control [17f ]. This synthetic methodology is based on the direct asymmetric incorporation of two acetate and one isobutyrate synthones into a framework (Scheme 2.1). The 2a-promoted asymmetric aldol reaction of a variety of aldehydes with a silyl nucleophile derived from phenyl propionate (E isomer 98%) results in moderate anti diastereoselectivity with relatively low enantioselectivity. On the other hand, with pivalaldehyde and the silyl nucleophile derived from ethyl propionate (E/Z ¼ 85:15), the syn isomer is obtained as a major product (22:1) with 96% ee (Eq. (25)) [17g]. This unexpected switching of diastereoselectivity observed in the reaction of the bulky aldehyde can be explained by merging Corey’s hydrogen-bond model between the aldehyde hydrogen and the catalyst borane-ring oxygen [18e] and Yamamoto’s ex-

37

38

2 Boron and Silicon Lewis Acids for Mukaiyama Aldol Reactions

HO MeO2C

OAc

9

O

OH OH OH O

7

1

O

BnO OH

OH O

OEt

O O

AcO

OH CO2Me Bryostatin

CHO +

BnO

S

OTMS

1. (R)-2a (1 equiv) CH2Cl2, -78 °C, 8 h

S

OEt

2. Ni2B-H2

OH CO2Et

BnO >98% ee OTMS

S 1.

OTBDMS CHO

1. TBDMACl 2. DIBAH

BnO

TBDMSO

OH O

BnO

S OEt (S)-2a (1 equiv) CH2Cl2, -78 °C, 8 h 2. Ni2B-H2

TBDMSO

1. TBDMACl OEt

2. DIBAH

BnO

OTBDMS CHO

~100% de OTMS 1. OEt (R)-2a (1 equiv) CH2Cl2, -78 °C, 3 h 2. Ni2B-H2

TBDMSO

TBDMS O OH O

BnO

OEt ~100% de

Scheme 2.1

tended transition model 3 (see also Figure 2.4) [19] as depicted in Figure 2.2, where 4 is destabilized by gauche interaction between the methyl and tert-butyl groups. Important limitations have been observed with regard to reagent control in reactions with highly sterically hindered aldehydes involving a chiral hydroxy function at the b-position (Eq. (26)) [17g]. When (S)-2a is used for 5, the diastereo- and enantioselectivity are less satisfactory. When (R)-2a,

2.2 Chiral Boron Lewis Acids

Ts

BH O O Me H

H t-Bu O

Ts

i-Pr

N

O

>>

N

i-Pr

BH O O H H

Me t-Bu

O

O

O

O 4

3 Fig. 2.2

Kiyooka’s transition-state models.

OTMS t-BuCHO + OEt OH O

(S)-2a (1 equiv) CH2Cl2 –78 °C, 3 h

t-Bu

OH O OEt

+

ð25Þ

t-Bu

OEt

22:1 96% ee

95% ee

is used, however, the reaction proceeds more smoothly to give the corresponding aldols with moderate syn selectivity in 87% yield. Each of the isomers obtained is almost enantiomerically pure. The spatial orientation of the siloxy group at C-3, which is presumably ﬁxed by introduction of two methyl groups at C-2, aﬀects the entire conformation of the aldehydes, and when the chiral borane coordinates to the aldehyde an adequate ﬁt might be needed between the stereocenters of the reagent and the substrate (at C3) for the stereochemical outcome expected from reagent control. Reaction with (S)-2a loses reagent control because of stereochemically mismatched interactions. Even in such complex circumstances, however, the reaction with (R)-2a gives products with stereochemistry at C-3 similar to that expected on the basis of reagent control. Eﬀective approach of the silyl nucleophile might occur via a path similar to 11 in Figure 2.3.

Me TMSO Blocking

Me TMSO

Me H

O H B O 10

Fig. 2.3

Kiyooka’s transition-state models.

2a part

Me H

O H B O 11

Nu

39

40

2 Boron and Silicon Lewis Acids for Mukaiyama Aldol Reactions

TBDMSO

OTMS 2 CHO 3

2a (1 equiv)

OTMS +

1

OEt

CH2Cl2 –78 °C, 24 h

5 TMSO

OH O

TBDMSO

TMSO OEt

+

TBDMSO

OEt 7

6 TMSO +

OH O

TBDMSO

OH O

TMSO OEt

+

OH O

TBDMSO

8

OEt 9

(S)-2a

34% yield, syn(6+8):anti(7+9)=2:1, 6:8=7:5, 7:9=5:3

(R)-2a

87% yield, syn(6+8):anti(7+9)=4:1, 6:8=>50:1, 7:9=>50:1

ð26Þ

2.2.3

Chiral Boron Lewis Acids as Catalytic Reagents

CAB 12, R ¼ H, derived from monoacyloxytartaric acid and diborane, is an excellent catalyst (20 mol%) for the Mukaiyama condensation of simple enol silyl ethers of achiral ketones with a variety of aldehydes. The reactivity of aldol-type reactions can be improved without reducing enantioselectivity by using 10–20 mol% 12, R ¼ 3,5-(CF3 )2 C6 H3 , prepared from 3,5-bis(triﬂuoromethyl)phenylboronic acid (13) and a chiral tartaric acid derivative. Enantioselectivity could also be improved, without reducing the chemical yield, by using 20 mol% 12, R ¼ o-PhOC6 H4 , prepared from o-phenoxyphenylboronic acid and a chiral tartaric acid derivative. The 12-catalyzed aldol process enables preparation of adducts highly diastereo- and enantioselectively (up to 99% ee) under mild reaction conditions [19a,c]. These reactions are catalytic, and the chiral source is recoverable and reusable (Eq. (27)). The relative stereochemistry of the major adducts is assigned to be syn, and the predominant re-face attack of enol ethers at the aldehyde carbonyl carbon has been conﬁrmed when a natural tartaric acid derivative is used as Lewis acid ligand. The use of an unnatural form of tartaric acid as a chiral source gives the other enantiomer, as expected. Almost perfect asymmetric induction is achieved with the syn adducts, reaching 99% ee, although a slight reduction in both enantio- and diastereoselectivity is observed in reactions with saturated aldehydes. Irrespective of the stereochemistry of the starting enol silyl ethers generated from ethyl ketone, syn aldols are obtained with high selectivity in these reactions. The observed high syn selectivity, and its lack of dependence on the stereoselectivity of the silyl enol ethers, in 12-catalyzed reactions are fully consistent with Noyori’s TMSOTf-

2.2 Chiral Boron Lewis Acids

Oi-Pr O 1)

OTMS R1CHO + R2

HO

R3

O O B Oi-Pr 12 (10~20 mol%) R HO EtCN, -78 °C

HO Ph

(99%), 88% ee (12 (10 mol%), R=3,5-(CF3)2C6H3) HO

O

Ph

HO

Ph

Ph

O Et

(99%), 96% ee syn syn:anti=94:6 (12 (20 mol%), R=H)

O

Pr

(83%), 97% ee syn syn:anti=>95:5 (12 (20 mol%), R=3,5-(CF3)2C6H3)

HO

(92%), 96% ee syn syn:anti=99:1 (12 (10 mol%), R=3,5-(CF3)2C6H3)

O

R3 R2

O

Ph

O

R1

2) 1N HCl

O

Ph

CO2H O

HO Et

(27)

O Et

(61%), 88% ee syn syn:anti=80:20 (12 (20 mol%), R=H)

(95%), 93% ee syn syn:anti =94:6 (12 (20 mol%), R=3,5-(CF3)2C6H3)

catalyzed aldol reactions of acetals, and thus might reﬂect the acyclic extended transition state mechanism postulated in the latter reactions (Figure 2.4). Judging from the product conﬁgurations, 12 (from natural tartaric acid) should eﬀectively cover the si face of the carbonyl after its coordination, and selective approach of nucleophiles from the re face should result. This behavior is totally systematic and in good agreement with the results from previously described 12-catalyzed reactions for all of the aldehydes examined. A catalytic enantioselective aldol-type reaction of ketene silyl acetals with achiral aldehydes also proceeds smoothly with 12, R ¼ H; this can furnish erythro b-hydroxy esters with high optical purity (Eq. (28) [19b,c]. R2

TMSO

H H

TMSO

R

R1

O

CAB

<

anti Fig. 2.4

Extended transition-state model.

H

2

R

R H

syn

R1 O

CAB

41

42

2 Boron and Silicon Lewis Acids for Mukaiyama Aldol Reactions

OTMS R

1CHO

+ R2

HO

1) 12 (R=H, 20 mol%) EtCN, -78 °C

Ph

HO OPh

OR3 R2

O

Pr

O

R1

OR3 2) TBAF O

HO

HO OPh

Ph

O OPh

(28) 84% ee HO Pr

O

76% ee HO

O

OPh Pr

88% ee syn syn:anti=79:21

92% ee syn syn:anti=79:21 HO OPh

97% ee syn syn:anti=96:4

O OPh

94% ee syn syn:anti=95:5

A remarkable ﬁnding is the sensitivity of this reaction to the substituents of the starting silyl ketene acetals. The reactions of silyl ketene acetals derived from more common ethyl esters are totally stereorandom, and give a mixture of syn and anti isomers in even ratios with improved chemical yields. In sharp contrast, the use of silyl ketene acetals generated from phenyl esters leads to good diastereo- and enantioselectivity with excellent chemical yields. The reason for this ﬁnding is not clear, but a secondary interaction between electron-rich silyl ketene acetals derived from alkyl esters and Lewis acid might be responsible. Analogous to the previous results with silyl enol ethers of ketones, nonsubstituted silyl ketene acetals lead to lower levels of stereoregulation. On the other hand, propionate-derived silyl ketene acetals lead to high asymmetric induction. Reactions with aliphatic aldehydes, however, result in a slight reduction in optical yields. With phenyl ester-derived silyl ketene acetals, erythro adducts predominate, but the selectivity is usually moderate compared with the reactions of silyl enol ethers. Exceptions are a,bunsaturated aldehydes, for which diastereo-and enantioselectivity are excellent. The observed erythro selectivity and re-face attack of nucleophiles on the carbonyl carbon of aldehydes are consistent with the aforementioned aldol reactions of silyl enol ethers [19]. After the enantioselective aldol reaction using CAB 2a under stoichiometric conditions has been reported by Kiyooka and his colleagues in 1991 [17], Masamune [20], Kiyooka [21a], and Corey [22] and their co-workers all independently developed CAB-catalyzed systems of enantioselective aldol reactions (Eq. (29)). Masamune and colleagues examined several oxazaborolidines derived from a series of simple a-amino acid ligands derivatized as the corresponding N-p-toluenesulfonamides. A dramatic improvement in reaction enantio-

2.2 Chiral Boron Lewis Acids

R2 1

R CHO

2 (20 mol%) OTMS

+ R3

EtCN or EtNO2

X

i-Pr O 2b

TMSO

O

R1

X R2 R3

O

Ts N BH O

O BH N Ts

MeO MeO

(29)

2c

Masamune et al. (for X=OR4 or SR4) O

O

O

BH N SO2C6H4-p-NO2

O BBu N Ts

HN

2d

2e

Kiyooka et al.

Corey et al.

(for X=OR4 or SR4)

(for X=R4)

selectivity was observed when complexes prepared from a,a-disubstituted glycine arylsulfonamides were used. This suggests that the initial aldol adduct must undergo ring-closure to release the ﬁnal product 15 and to regenerate the catalyst 2 (Figure 2.5) [20]. Slow addition of the aldehyde to the reaction mixture (making enough time available for 14 to undergo ring closure) has often been beneﬁcial in improving the enantioselectivity of the reaction. Kiyooka and his colleagues have reported a straightforward improvement of this reaction to a catalytic version by using an N-p-nitro-

TMSO

O

O

X R1 * R2 R3 15

R

O BR" N SO2R' 2

R'SO2N

B O

O R"

TMS

O

X R1 * R2 R3 14 Fig. 2.5

The proposed catalytic cycle.

+

R2 OTMS

R3

O R

R1CHO

X

43

44

2 Boron and Silicon Lewis Acids for Mukaiyama Aldol Reactions

benzenesulfonyl-derived ligand and nitroethane instead of dichloromethane as solvent [21a]. Product enantioselectivity has also been optimized as a function of substitution of the arylsulfonamide. (Eq. (30)) [20]. Thus, for complexes with the general structure 2c, the enantiomeric excess of the benzaldehyde adduct varies along the series R 0 ¼ 3,5-bis(triﬂuoromethyl)phenyl (52% ee); mesityl (53% ee); 1-naphthyl (67% ee); 2-naphthyl (78% ee); 4-tertbutylphenyl (81% ee); phenyl (83% ee); 4-methoxyphenyl (86% ee); 4acetamidophenyl (86% ee). O

OSiMe3 PhCHO

+

O BH N SO2R'

MeO MeO

OH

(20 mol%)

OEt

EtCN

Ph

CO2Et

F3C Ar= X

CF3 52% ee

53% ee

67% ee

78% ee

X=t-Bu: 81% ee X=H: 83% ee X=MeO 86% ee X=AcNH 86% ee

ð30Þ An AM1-optimized structure of the chiral borane complex was used as the centerpiece of a model proposed by Kiyooka and co-workers to account for the stereochemical outcome of the reaction (Figure 2.6) [21a]. It was suggested the aldehydes coordinate with the boron on the face opposite the isopropyl substituent, thereby minimizing steric interactions. The Kiyooka model places the formyl-H over the ﬁve-membered ring chelate subtending an obtuse HaBaOaC dihedral angle. Analogous modes of binding have i-Pr O O O S N B O H

i-Pr O O O S N B O H

O

O

H

R O 2N The Kiyooka model

H

R O 2N The Corey model

Fig. 2.6

The proposed transition-state models.

2.2 Chiral Boron Lewis Acids

been proposed in other chiral acid boron compounds that have been ingeniously used for Diels–Alder cycloaddition reactions [23]. The preference for this orientation might result from presence for a stabilizing anomeric interaction. Alternatively, the bound aldehyde might be locked in the conformation invoked by Kiyooka as a result of a formyl CaH hydrogen bond to the acyloxy donor, in accordance with the bonding model proposed by Corey [18e, 21a]. Kiyooka et al. reported that the 2d-catalyzed aldol reaction of a silyl ketene acetal involving a dithiolane moiety with a b-siloxy aldehyde results in the production of syn- and anti-1,3-diols with complete stereoselectivity if the stereochemistry of the catalyst is chosen appropriately [21b]. This method has been applied to enantioselective synthesis of the optically pure lactone involving a syn-1,3-diol unit, which is known to be a mevinic acid lactone derivative of the HMG-CoA reductase inhibitors mevinolin and compactin (Scheme 2.2).

HO

O

HO

O

O

O

O

O Ph R R=H: Mevinolin R=Me: Compactin OTMS

S CHO +

Ph

S

OEt

1. 2d (20 mol%) EtNO2, -78 °C, 1 h

OH CO2Et

Ph

2. Ni2B-H2

>98% ee OTMS

S 1.

OH

1. TBDMACl Ph

2. DIBAH

S OEt 2d (20 mol%) EtNO2, -78 °C, 1 h CHO

2. Ni2B-H2 HO

TBDMSO

OH

100% de Scheme 2.2

O

1. Bu4NF CO2Et

Ph

O

2. TsOH Ph

45

46

2 Boron and Silicon Lewis Acids for Mukaiyama Aldol Reactions

Corey et al. used 2e in the conversion of aldehydes to 2-substituted 2,3dihydro-4H-pyran-4-ones by reacting them with 1-methoxy-3-trimethylsilyloxy-1,3-butadiene in propionitrile at 78 C for 14 h and then treating them with triﬂuoroacetic acid (Eq. (31)) [22]. OTMS R1CHO

2e (20 mol%)

+ EtCN, -78 °C OMe

R1

O

TMSO

(31) CF3CO2H

R1

O O

OMe

67-82% ee

Corey’s tryptophan-derived chiral oxazaborolidine 2e is highly eﬀective for not only the Mukaiyama aldol reaction of aldehydes with silyl enol ethers [22] but also the Diels–Alder reaction of a-substituted a,b-enals with dienes [23], although more than 20 mol% 2e is required for the former reaction. Other chiral oxazaborolidines that have been developed for enantioselective aldol reaction of aldehydes with relatively more reactive ketene silyl acetals also require large amounts (more than 20 mol%) to give aldol adducts in good yield [20, 21]. Yamamoto and his colleagues succeeded in enhancing the catalytic activity of CAB derived from 2,6-di(isopropoxy)benzoyltartaric acid and borane.THF by using 13 instead of borane.THF [19c]. In a similar manner they developed a new and extremely active Corey’s catalyst, 2f, using arylboron dichlorides bearing electron-withdrawing substituents as Lewis acid components [24]. A new chiral oxazaborolidine catalyst 2f has been prepared by treating N-( p-toluenesulfonyl)-(S)-tryptophan with an equimolar amount of 3,5bis(triﬂuoromethyl)phenylboron dichloride (15) in dichloromethane and subsequent removal of the resulting HCl and the solvent in vacuo (Scheme 2.3). Moisture-sensitive boron dichloride 15 and boron dibromide 16 are synthesized by dehydration of 13 to trimeric anhydride 14 and subsequent halogenation of 14 with 2 equiv. of BCl3 and BBr3, respectively [24]. The preparation of oxazaborolidines from arylboron dichlorides has been also reported by Reilly and Oh [25] and Harada and co-workers [26]. Although Bbutyloxazaborolidine 2e has been prepared from N-( p-toluenesulfonyl)-(S)tryptophan and butylboronic acid by dehydration [22], B-aryloxazaborolidine cannot be prepared from arylboronic acid, as observed by Nevalainen et al. [27] and by Harada et al. [26b]. In contrast, CAB derived from 2,6di(isopropoxy)benzoyltartaric acid in place of N-sulfonylamino acids has been easily prepared by adding an equimolar amount of the corresponding arylboronic acid at room temperature [19c]. According to Corey and co-workers [22], terminal trimethylsilyloxy (vinylidene) oleﬁns seem to be more suitable substrates for enantioselective Mu-

2.2 Chiral Boron Lewis Acids

CF3 benzene (HO)2B

trimer 14

azeotropic reflux (CaH2), 2~4 h

CF3 13

CF3 1. BX3 (2 equiv) hexane or heptane reflux, several hours 2. distillation

X2B CF3

15 (X=Cl) or 16 (X=Br) ca. 40~50% yield from 13 CF3 O HN

OH

1. 15, CH2Cl2, rt, 1 h

NH Ts

2. pump on 3. EtCN

O

O B N Ts 2f

HN

CF3

Scheme 2.3

kaiyama aldol coupling catalyzed by 2e than more highly substituted oleﬁns such as RCHbC(OSiMe3 )R 0 or R2 CbC(OSiMe3 )R 0 . In fact, reaction of the trimethylsilyl enol ether derived from cyclopentanone with benzaldehyde aﬀorded the aldol products in only 71% yield even in the presence of 40 mol% 2e [22]. According to Yamamoto et al. [24], reaction of benzaldehyde with the trimethylsilyl enol ether derived from acetophenone in the presence of 10 mol% 2e gives the trimethylsilyl ether of aldol and the free aldol in yields of only 38% and 15%, respectively (Eq. (32)). When the B-3,5-bis(triﬂuoromethyl)phenyl analog 2f is used, however, catalytic activity and enantioselectivity are increased to a turnover of 25 and 91–93% ee, respectively. The absolute conﬁguration of the aldol adducts is uniformly R.

PhCHO +

OSiMe3 Ph

cat. 2

Me3SiO

EtCN, –78 °C

Ph

2e (10 mol%): 2f (10 mol%): 2f (4 mol%):

O

HO

O

+ Ph

38% yield, 82% ee 91% yield, 93% ee 94% yield, 91% ee

Ph

Ph

ð32Þ

15% yield, 82% ee 4% yield, 68% ee 4% yield, 72% ee

These results indicate that introduction of an electron-withdrawing substituent such as the 3,5-bis(triﬂuoromethyl)phenyl group to the B atom of

47

48

2 Boron and Silicon Lewis Acids for Mukaiyama Aldol Reactions

chiral boron catalysts is an eﬀective method for enhancing their catalytic activity. The method is especially attractive for large-scale synthesis (Eq. (33)).

PhCHO + 10 mmol

OSiMe3 Ph 12 mmol

1. 2f (5 mol%), EtCN, –78 °C, 5 h 2. 1M HCl-THF

HO

O

Ph Ph 2.23 g, 99% yield 94% ee (R)

ð33Þ

CAB 2d is eﬀective for reaction not only with terminal trimethylsilyloxy oleﬁns but also trisubstituted (E)- and (Z)-trimethylsilyl enol ethers (Table 2.1). In the reaction of aromatic aldehydes such as benzaldehyde with the trimethylsilyl enol ether of cyclohexanone, both substrates should be sequentially added to a solution of 2f in propionitrile at 78 C according to Corey’s procedure (method A) [22]. The reaction proceeds quantitatively to give only the aldol products in a 78:22 syn/anti ratio, and the optical yield of the syn isomer 17 is 89% ee. Reaction of aliphatic aldehydes such as isobutyraldehyde with the same silyl enol ether does not proceed well, however, probably because of decomposition of isobutyraldehyde in the presence of the strong Lewis acid 2f before addition of the trimethylsilyl enol ether. On the other hand, sequential addition of silyl enol ethers and aldehydes to a solution of catalyst 2f (method B) gives the aldol adducts in higher yield, but the enantioselectivity is relatively low. High enantioselectivity is also observed in the reaction with acyclic (E)- and (Z)-silyl enol ethers. Reaction with (Z)-trimethylsilyl enol ethers also gives syn aldol adducts as major diastereomers. The syn preference and the absolute preference for carbonyl re-face attack observed in the reactions of aldehydes with (E)- and (Z)-trimethylsilyl enol ethers suggests that the reaction occurs via an extended-transition state assembly (Figure 2.7) [19, 22]. Anti preference has been observed in the reaction of aldehydes with (E)-ketene trimethylsilyl acetals catalyzed by other chiral oxazaborolidines [20, 21]. Harada and co-workers reported that arylboron complex 2g derived from N-tosyl-(aS,bR)-b-methyltryptophan [23] and ( p-chlorophenyl)dibromoborane is an excellent catalyst for enantioselective ring-cleavage reactions of 2substituted 1,3-dioxolanes with enol silyl ethers [26c]. Interestingly, chiral boron complexes prepared by reacting the sulfonamide ligands with BH3 THF do not have appreciable catalytic activity [26a,b]. Successful results have been obtained in the ring cleavage of 1,3-dioxolanes with aryl and alkenyl groups at the 2-position. Reaction of 2-alkyl derivatives is, however, very sluggish under these conditions. The 2-hydroxyethyl group in the ring-

2.2 Chiral Boron Lewis Acids Tab. 2.1

Mukaiyama aldol reaction of aldehydes with (E)- and (Z)-silyl enol ethers OSiMe3 R1CHO

R2

+

EtCN, –78 °C, 12 h

R3 OH O

R1

OH O R2

Ph

cat. 2f (5 or 10 mol%)

R2

+ Ph

R3

R3

17

18

Silyl Enol Ether

Methoda

Yield (%)b

17:18c

ee (%)d 17

18

OSiMe3

Ph i-Pr

Pr

OSiMe3

A Ae

>99 36

78:22 77:23

89 96

5 96

B

>99

48:52

95

93

A B B B Be B B

23 >99 >99 >99 94 85 92

>99:1 >99:1 62:38 97:3 83:17 95:5 89:11

96 >99 92 98 92 97 90

– – 77 – 91 – 58

f

Et

Pr Pr Pr i-Pr i-Pr (E)-MeCHbCH PhCcC

Ph (>99:1) Ph (>99:1) Et (97:3) Ph (>99:1) Et (97:3) Ph (>99:1) Ph (>99:1)

a Method A: A solution of silyl enol ether (0.96 mmol) in propionitrile (0.32 mL) was added over 2 min to a mixed solution of 2f (0.08 mmol) and an aldehyde (0.8 mmol) in propionitrile (0.65 mL). Method B: A solution of aldehyde (0.8 mmol) in propionitrile (0.32 mL) was added over 10 min to a mixed solution of the silyl enol ether (0.96 mmol) and 2f (0.08 mmol) in propionitrile (0.65 mL). b Isolated yield. c Determined by 1 H NMR analysis. d Determined by HPLC. e 5 mol% of 2d was used. f E:Z ¼ 70:30.

cleavage products can be removed simply by conversion to the iodides then treatment with zinc powder (Eq. (34)). Desymmetrization of meso-1,2-diols has been realized by chiral Lewis acid 2h-mediated enantioselective ring-cleavage of dioxolane derivatives [26d]. Transacetalization of 3,3-diethoxy-1-phenylpropyne with meso-2,3-butanediol gives a 86:14 mixture of syn- and anti-19 stereoselectively. Treatment of syn19 with 3 equiv. Me 2 CbC(OTMS)OEt and 1.0 equiv. 2h at 78 C gives the ring-cleavage product 20 (> 20:1 diastereoselectivity) in 72% yield with 94%

49

50

2 Boron and Silicon Lewis Acids for Mukaiyama Aldol Reactions

π-π stacking

HN

R1 H O

O

O B

S

Me3SiO R

H

O

O R2

1

R3

R3 re-face attack CF3

N O

OSiMe3

R2

CF3

CH3 Fig. 2.7

Proposed extended-transition state assembly.

O NH

1.

B N Ts 2g (10 mol%) CH2Cl2, –20 °C

R3 O

O

+

R2

R1

OSiMe3

Cl

2. TBAF

R2 HO

O R

O

O

1

OH

1. I2, PPh3 R3

R 2. Zn

R2 R2

O

1

(34) R3

R2 R2

Examples HO

O

HO

O

Ph

OEt

O O

HO

O

O

O StBu

Ph MeO

88% yield 86% ee

73% yield 93% ee

80% yield 85% ee

ee (Eq. (35)). A separate experiment using pure anti-19 showed that it is unreactive under these conditions. Boron complex 2h is also eﬀective in the ring-cleavage of other dioxolanes that can be prepared stereoselectively (syn:anti > 20:1) from the diols under kinetically controlled conditions. The

2.2 Chiral Boron Lewis Acids

results obtained by using other catalysts, for example 2i and 2j, suggest that the structure of the N-sulfonyl moiety aﬀects the enantioselectivity.

O

O

Ph

2h (1 equiv)

Ph OH + HO

O

O

OEt CO2Et

EtO2C

OTMS

Ph syn-19

20 (major)

ent-20 (minor)

(3 equiv)

Ph

O

2h: R1=tol → 20: 94% ee

O O2S N B R1 Ph

2i: R1=Me → 20: 48% ee 2j:

R1

ð35Þ

=CF3 → 20: 58% ee

Other examples

Et

Et

Ph

Ph O

OH

CO2Et 2h → 96% ee

Ph O

OH

CO2Et 2h → 93% ee

O

OH

CO2Et 2h → 85% ee

Itsuno et al. have developed novel polyaddition reactions based on the Mukaiyama aldol reaction of silyl enol ethers with aldehydes. Bis(triethylsilyl enol ether) and bis(triethylsilyl ketene acetal) are prepared as stable and isolable monomers. In the presence of Lewis acid catalysts these monomers react smoothly with dialdehydes to aﬀord the poly(b-hydroxy carbonyl) compounds. By asymmetric synthetic polymerization of such monomers with chiral modiﬁed Lewis acid it is possible to obtain optically active poly(b-hydroxycarbonyl) compounds with main-chain chirality [28]. For example, CAB 21, which is highly eﬃcient in the asymmetric aldol reaction of silyl enol ether with aldehyde, has been examined as a chiral catalyst for asymmetric aldol polymerization of 22 with 23. Unfortunately, CAB 21 is not suﬃciently active to polymerize these monomers at 78 C. Increasing the temperature made it possible to obtain the chiral polymer in low yield, accompanied by partial decomposition of the catalyst. The polymer obtained is optically active, however (Eq. (36)) [28a]. Silyl ketene acetals also react enantioselectively with aldehydes in the presence of a chiral Lewis acid. Several useful chiral Lewis acids have recently been developed for this reaction. Itsuo et al. found that Kiyooka’s catalyst 2a.SMe 2 acts as a chiral catalyst of asymmetric aldol polymerization

51

52

2 Boron and Silicon Lewis Acids for Mukaiyama Aldol Reactions

OMe O

CO2H O O

O O B•SMe 2 H 21 (50 mol%) OMe

OSiEt3 CHO +

CH2Cl2, –78 °C to –20 °C

OHC OSiEt3 22

23

O

OSiEt3

Et3SiO

O

5% yield, Mn=1300, Mw/Mn=1.67, [Φ]=–64

ð36Þ

OHC

OSiEt3 O

O

+

O

O

O

OSiEt3

O

24 O O 2a•SMe2 (200 mol%) CH2Cl2, –78 °C

CHO

25

O

OSiEt3

O

O O

O O

Et3SiO 62% yield, Mn=1900, Mw/Mn=2.21, [Φ]=–49

ð37Þ of 24 with 25 even at 78 C. The aldol polymer with optical activity is again obtained in 62% yield (Eq. (37)) [28a]. Itsuno and co-workers also reported that CAB 2d is a more eﬀective catalyst than other chiral oxazaborolidines 2 for asymmetric polymerization of bis(triethylsilyl enol ether)s and dialdehydes [28b]. The reactivity of dialdehydes containing ether linkages is quite low for formation of polymers, mainly because of the low solubility of dialdehyde monomers in propionitrile. Introduction of a silyl group into the monomeric structure of the dialdehyde dramatically improves the solubility. The asymmetric polymerization of silyl-containing dialdehyde 26 with 22 aﬀords the chiral polymer in high yield with high molecular weight (Eq. (38)). This polymer is soluble in

2.3 Silicon Lewis Acids

O

O

22 + H

2d (20 mol%)

H

EtCN, –78 °C to –20 °C

Si Me2 26

O

O

OSiEt3

Et3SiO

ð38Þ

Si Me2

71% yield, Mw=48200, Mw/Mn=10.3, [Φ]435=1670

common organic solvents such as THF, CH2 Cl2 , CHCl3 , DMF, and DMSO. All the chiral polymers obtained using 2d as catalyst have positive optical rotation.

2.3

Silicon Lewis Acids 2.3.1

Introduction

Silicon Lewis acids have advantages over traditional metal-centered activators. For example, silicon Lewis acids are compatible with many synthetically valuable C-nucleophiles, such as silyl enol ethers. Unlike metal halides, silicon Lewis acids are not prone to aggregation, which substantially simpliﬁes the analysis of the reaction mechanisms. Furthermore, the reactivity of silicon Lewis acids of R3 SiX structure can be ﬁnely controlled by varying the steric volume of alkyl substituents. The most advantageous circumstance is the opportunity to realize the processes in the presence of catalytic amounts of silicon Lewis acids if silicon Lewis acids and silyl enol ethers have identical trialkylsilyl fragments. Thus, depending on the type of electrophile, two mechanistically diﬀerent pathways can be considered (Scheme 2.4). For acetals and acetal-like compounds, silicon Lewis acids abstract the heteroatomic substituent, followed by reaction of electrophilic species formed with a nucleophile (left circle). When the substrates have a carbon–heteroatom double bond (e.g. carbonyl compounds, imines) silicon Lewis acids bind to their basic function leading, after carbon–carbon bond formation, to products containing the silyl group (right circle). The approach generalized in Scheme 2.4 was ﬁrst realized by Noyori and co-workers in the early eighties [29]. Subsequently silicon Lewis acid gained wide acceptance as mediators of a variety of transformations. This section

53

54

2 Boron and Silicon Lewis Acids for Mukaiyama Aldol Reactions

Y–E Y–E R3SiY +

E+X–

R3Si–Y+=E X–

Nu–SiR3

Nu–SiR3 R3SiX

E–Nu+–SiR3 X

R3Si–Y–E–Nu+–SiR3

–

X– E–Nu

R3Si–Y–E–Nu

Scheme 2.4

surveys data on the behavior of silicon Lewis acids of general formula R3 SiX in Mukaiyama aldol reactions [30]. 2.3.2

Lewis Acidity of Silicon Derivatives

In the last two decades, the problem of observation of trialkylsilyl cations R3 Siþ , apparently the strongest silicon Lewis acids, attracted considerable attention. According to the results of ab initio calculations [31] and experimental data [32] the equilibrium shown in Eq. (39) is substantially to the right. R3SiH + R3C+

R3Si+ + R3CH

(39)

Correspondingly, R3 Siþ can be readily formed in the gas phase where they can be characterized and studied [33]. Observation of these cations in the condensed state (in solution or in the crystalline state) is, however, very diﬃcult [34]. Nevertheless, Lambert demonstrated recently that silyl cations containing bulky substituents which hinder the approach of nucleophilic reagents to the silicon atom can be observed in solution. He succeeded in detecting Mes3 Siþ B(C6 F5 )4 ( 29Si NMR d (ppm) ¼ 225.5) [35], or Dur3 Siþ B(C6 F5 )4 ( 29Si NMR d ¼ 226:8) [36] (Mes ¼ 2,4,6-trimethylphenyl, Dur ¼ 2,3,5,6-tetramethylphenyl) in benzene, the chemical shifts being very close to the calculated value (d (Mes3 Siþ , calcd.) ¼ 230.1) [37]. Hence, covalent compounds of the type R3 SiX, where X is either the conjugated base of a strong acid (for example CF3 SO2 or ClO4 ) or a solvent molecule (for example, MeCN), generally serve as Lewis acids in carbon– carbon bond-forming reactions. Several approaches have been proposed for estimation of the Lewis acidity of R3 SiX. One of these assumes that the positive charge on the silicon atom is proportional to the chemical shift in 29Si NMR spectra. This scale can,

2.3 Silicon Lewis Acids

however, only be used as a reliable indication of the relative reactivity of compounds in which the silicon atom is bound to the same heteroatom. Another procedure for estimating the Lewis acidity of Me3 SiX, suggested by Hergott and Simchen, is based on comparison of the silylation rate constants of cyclopentanone and diisopropyl ketone with these reagents in the presence of triethylamine in dichloroethane [38]. Bassindale et al. have proposed estimating the strength of silicon Lewis acids from their ability to form the N,N-bis(trimethylsilyl)imidazolium cation in reactions with Ntrimethylsilylimidazole [39]. On the basis of results from studies of the kinetics and thermodynamics of this reaction, silicon Lewis acids were arranged Me3 SiCl < Me3 SiBr < Me3 SiI < Me3 SiOTf < Me3 SiClO4 in order of silyl-donating capacity. Although quantitative data on Me3 SiNTf2 and Me3 SiN(SO2 F)2 derivatives are lacking, the results of comparative experiments provide evidence that these reagents are much more reactive than Me3 SiOTf. The results obtained by diﬀerent research groups make it possible to arrange the most commonly used neutral silicon Lewis acids in the quantitative activity series: Me3 SiCl < Me3 SiOMs < Me3 SiOTs < Me3 SiBr f Me3 SiOTf A Me3 SiOSO2 F a Me3 SiI a Me3 SiClO4 < Me3 SiN(SO2 F)2 < Me3 SiNTf2 . Positively charged species such as MeCNaSiMe3 þ or complexes generated from neutral silicon Lewis acids and metal-centered Lewis acids might be even more reactive than Me3 SiNTf2 . 2.3.3

Silicon Lewis Acids as Catalytic Reagents

The reactions of carbonyl compounds with silyl enol ethers can be described by the mechanism shown in Scheme 2.5. Thus, the reaction of a substrate with a silicon Lewis acid initially aﬀords a ﬁve-coordinate complex 27 which can exist in equilibrium with cation 28 [40]. Subsequent nucleophilic attack on the carbon atom of complexes 27 or 28 is accompanied by formation of the carbon–carbon bond to give intermediate 29 or 30, respectively. The intermediate 30 is rapidly transformed into the ﬁnal product 31. The position of the equilibrium between RCHO þ Me3 SiX, 27, and 28 depends on the cation-stabilizing eﬀect of the substituents R1 and R2 and on the nature of the leaving group X. Attempts to observe complexes 27 or 28 generated from benzaldehyde and Me3 SiOTf by NMR spectroscopy have failed (only the starting component provided unambiguous evidence of a very small contribution of 27 and 28 to the equilibrium mixture). Hence, it can be tentatively assumed that cation 28 is not formed if the SiaX bond is suﬃciently strong, e.g. an SiaO bond. Neutral complex 27, in turn, is a much weaker electrophile than 28 or the oxocarbenium cation generated from acetals on elimination of the alkoxy group. Consequently, one would expect carbonyl compounds to be less reactive than the corresponding acetals in reactions with nucleophiles.

55

56

2 Boron and Silicon Lewis Acids for Mukaiyama Aldol Reactions

SiMe3

+

O

R2

R1 X–

O R2

R1

+

+ Me3SiX

–

O

28

X SiMe3 R2

R1

Nu–SiMe3 O R1

Nu–SiMe3 –

27

O R1

X SiMe3

SiMe3 +

R2 30

Nu SiMe3 X–

+

R2

Nu SiMe3

–Me3SiX

29 O R1

SiMe3

Nu R2 31

Scheme 2.5

The diﬀerent reactivity of the acetal and carbonyl groups is demonstrated by the bifunctional substrate 32, which reacts with 1-trimethylsiloxycyclohexene exclusively at the acetal fragment (Eq. (40)) [41]. OSiMe3

OMe +

O

MeO 32

Me3SiOTf

(40) O

OMe O

In the early 1980s, Noyori demonstrated that aldehydes and ketones do not react with 1-siloxycyclohexene in the presence of Me3 SiOTf in CH2 Cl2 at 78 C [30, 41, 42]. The reaction of benzaldehyde with 1-trimethylsiloxycyclohexene catalyzed by Me3 SiOTf proceeds only at room temperature to give the target silyl ether of aldol (in toluene the yield was 60%, syn:anti ¼ 49:51) or benzylidene-cyclohexanone (in CH2 Cl2 , 85%). Aliphatic aldehydes are not involved in this reaction. According to results from other studies benzaldehyde reacts smoothly with silyl enol ethers on catalysis by Me3 SiOTf (5 mol%) in CH2 Cl2 at 78 C to give the silyl ether of aldol in 89% yield in the ratio syn:anti ¼ 63:37 [43]. The latter reaction is probably catalyzed by traces of TfOH rather than by Me3 SiOTf itself. Reaction of trimethylsiloxycyclohexene with benzaldehyde or isobutyraldehyde in the presence of 5 mol% TfOH in CH2 Cl2 at 78 C is complete in 30 min to give aldol products in 86% (syn:anti ¼ 69:31) and 82% (syn:anti ¼ 73:27) yields, respectively [41, 44].

2.3 Silicon Lewis Acids

Me3 SiNTf2 , however, a considerably stronger silyl donor than Me3 SiOTf, eﬃciently catalyzes addition of silyl enol ethers to aldehydes and ketones (Eq. (41)). The reaction is best performed in diethyl ether as solvent at 78 C with as little as 0.5–1.0 mol% silicon Lewis acid, generated in situ from HNTf2 and silyl enol ether. OSiMe3 R2

3. HCl or Bu4NF

R1 R1,

1. 0.5–1.0 mol% HNTf2, Et2O, –78 °C 2. Addition of R3R4CO over 2 h

R2=H,

OH O R3

Alkyl, Ph

R4

R2 1

R

87–92% yield

R3, R4=H, Ph, (CH2)4 Examples (yield) OH O Ph

OH O Ph

Ph

92%

87%

OH O Ph

OH O

92% (syn:anti=70:30)

OH O Ph

88% (syn:anti=76:24) (silyl enol ether, 96% cis)

OH O Ph

87%

(41)

Ph

Ph

Ph 92%

To minimize the formation of side products it is necessary to add the carbonyl compound slowly [45]. The presence of excess molar amounts of carbonyl compounds per desired adducts produced in the reaction concurrently promotes at least three reactions (Scheme 2.6): (1) cyclic trimerization of the aldehyde (path a), (2) dimerization of the desired adducts (path b), and (3) acetalization of the desired adducts (path c). Slow addition of the carbonyl compound to a mixed solution of silyl enol ether and Me3 SiNTf2 is the best way to obtain the desired products selectively. The following mechanism has been proposed for this aldol reaction pathway [45b]. Electrophilic attack of silyl-activated aldehyde species on the silyl enol ether produces cationic species 33 which subsequently acts as a source of Lewis acidic silyl group without regeneration of Me3 SiNTf2 (Scheme 2.7). In accord with such a mechanism is the observation that the silylated aldol initially formed by coupling of benzaldehyde with silyl enol ethers derived from acetophenone contains the silyl group derived from the nucleophile, and not from R3 SiNTf2 (Eqs. (42) and (43) [45b]. In a similar experiment with Me3 SiOTf performed at 78 C a mixture of 34 and 35 is obtained (Eq. (44)), suggesting that after carbon–carbon bond formation the silyl triﬂate with a silyl group originating from the enol ether is generated.

57

58

2 Boron and Silicon Lewis Acids for Mukaiyama Aldol Reactions

Me3SiNTf2 (1 equiv)

OSit-BuMe2

PhCHO +

Ph

Me3SiO

O

Ph (43) 35 34:35=<1:99

Et2O, –100 °C

Ph

Me3SiOTf (1 equiv)

OSit-BuMe2

PhCHO + 1 mmol

Ph (42) 34 34:35=>99:1

t-BuMe2SiNTf2 (1 equiv)

Ph

O

Ph

Et2O, –100 °C

OSiMe3

PhCHO +

t-BuMe2SiO

34 + 35 Et2O

Ph

(44)

Et2O (50 mL), –100 °C, 0.5 h: 24% yield, 34:35=1:99 Et2O (12.5 mL), –78 °C, 5 h: 61% yield, 34:35=17:83

When, moreover, two enol ethers of diﬀerent ketones bearing diﬀerent silyl groups are used simultaneously scrambling of the silyl groups occurs [45b, 46]. These observations indicate that the ligand (X) of the silicon Lewis acid (R3 SiX) plays a crucial role in the Mukaiyama aldol reaction of trimethylsilyl enol ethers (Me3 SiNu). In the R3 SiOTf-induced reaction transfer of TfO from siloxocarbenium ion 36 is expected to occur by electrophilic attack of

–

O R

Me3SiNTf2

+

O

R

H

NTf2 SiMe3

R

O

R

path a O

O

H

R R R O

OSiMe3

Me3SiNu

O R

OSiMe3

Nu

path c

O

R

Nu

R

R Me3SiNu

path b R

O R Scheme 2.6

Nu

Nu Nu

Nu

2.3 Silicon Lewis Acids

–

O R1

+

O

Me3SiNTf2 R1

H

NTf2 SiMe3

NTf2 – Me3Si O

OSiMe3 R2

+O

SiMe3 R2

R1

H

59

33 O R

1

H

NTf2 – Me3Si O

OSiMe3

Me3 H +O Si O R1 –

R2

+

etc.

R2

R1 Scheme 2.7

–

OTf SiR3

OSiMe3

+

O

R1

R2

OTf − R3Si O

+

SiMe3

O

R3SiO

+ Me3SiOTf

R2

R1

R2

R1

H

O

36 Scheme 2.8

the ‘‘Me3 SiaOþ silicon’’ of 36 (Scheme 2.8). Me3 SiOTf would be generated by electrophilic attack of the ‘‘Me3 SiaOþ silicon’’ on the ‘‘SbO oxygens’’ or the ‘‘SaO oxygen’’ of OTf. In the R3 SiNTf2 -induced reaction, in contrast, less nucleophilicity and/or more bulkiness of NTf2 might suppress electrophilic attack of the ‘‘Me3 SiaOþ silicon’’ on the nitrogen or oxygen atoms of NTf2 , and might increase the Lewis acidity of siloxocarbenium ion 33 (Scheme 2.7). Unlike silyl enol ethers, silyl ketene acetals react with aldehydes and ketones on catalysis by Me3 SiOTf [47], and carbonyl compounds often seem to be more reactive in these reactions than their acetals (Eq. (45)) [47b]. Bis(silyl)methylacetoacetate is a synthetic equivalent of the corresponding dianion and its terminal carbon atom is involved in reactions with carbonyl compounds in the presence of Me3 SiOTf, ketones being more reactive than aldehydes (Eq. (46)) [48].

OSit-BuMe2 OEt

O +

+ Ph

MeO OMe

1. Me3SiOTf (10 mol%) CH2Cl2, –78 °C 2. work up

Ph

OH O

OMe O +

Ph

OEt 91% yield

Ph

OEt 0% yield

(45)

60

2 Boron and Silicon Lewis Acids for Mukaiyama Aldol Reactions

OSiMe3

Me3SiO

+

+ n-C5H11

OMe

1. Me3SiOTf (10 mol%) CH2Cl2, –78 °C

O

O

OH O n-C5H11

H

n-C7H15

2. work up OH O

CO2Me +

CO2Me

n-C7H15

52% yield

not detected

ð46Þ

2.3.4

Activation of Silicon Lewis Acids by Combination with Other Lewis Acids

Binding of silicon Lewis acid (R3 SiX) with another Lewis acid (LA) leads to the shift of the electron density from the silicon atom (conﬁrmed by 29Si NMR spectroscopic data). As a consequence, the resulting R3 SiX ! LA complexes are much stronger donors of the silyl group than the starting SLA. Olah and colleagues demonstrated that reaction of Me3 SiBr with AlBr3 produces the Me3 SiBr ! AlBr3 complex ( 29Si NMR, d ¼ 62:7) [49a]. Even SbF5 , one of the strongest Lewis acids, cannot abstract the ﬂuoride anion from Me3 SiF and gives the Me3 SiF ! SbF5 complex ( 29Si NMR, d ¼ 102) rather than the silyl cation [49b]. The possibility of using R3 SiX ! LA complexes as mediators in carbon– carbon bond-forming reactions was ﬁrst demonstrated by Mukaiyama et al. in 1987 [51]. While quite inactive separately, Lewis acids Me3 SiCl and SnCl2 taken together have properties of strong R3 SiX Lewis acids. Thus, aldehydes, a,b-unsaturated ketones, and acetals smoothly react with silyl enol ethers in the presence of this LA pair (Scheme 2.9). The Me3 SiClaZnCl2 system can function analogously although it is less eﬃcient than Me3 SiClaSnCl2 [50].

O

OH

O

Ph

O

OMe

O Ph

Ph

Ph PhCHO

Ph

OSiMe3 15–20 mol% Me3SiCl 8–12 mol% SnCl2 Ph CH2Cl2, –78 °C O Ph

OMe OMe

Scheme 2.9

HC(OMe)3

EtCH(OMe)2 Ph

O

2.3 Silicon Lewis Acids

It has also been reported that R3 SiCl can be activated by addition of InCl3 [51]. The reactivity of the R3 SiClaInCl3 mixture is highly dependent on the nature of the alkyl groups on the silicon atom. For example, the Me3 SiClaInCl3 system catalyzes the reactions of trimethylsilyl enol ethers both with aldehydes and acetals (Eq. (47)) yet only aldehydes react with tertbutyldimethylsilyl enol ethers in the presence of t-BuMe 2 SiClaInCl3 . This behavior enables selective nucleophilic addition at the carbonyl group in the presence of the acetal fragment (Eq. (48)). Me3SiO

O R1

H or

OSiMe3 +

OMe R1

R3

R3 R2

Me3SiCl/InCl3 (10 mol%) CH2Cl2, –43 °C or –23 °C

R2

O

R1 or MeO

R3

R1

OMe

(47)

O 2

R 68~93% yield O H

MeO

+

OSit-BuMe2

t-BuMe2SiCl/InCl3 (10 mol%) CH2Cl2, –78 °C

OEt OMe

t-BuMe2SiO

(48)

O OEt

MeO OMe 77% yield

Boron and aluminum compounds can also activate R3 SiX, leading to silicon species with very high catalytic activity. The high reactivity of these systems is probably associated with the complete transfer of the silyl group to the carbonyl oxygen atom to form the siloxycarbonium species RCHbOþ SiR3 . The exothermic reaction of B(OTf )3 with Me3 SiOTf gives the Me3 SiB(OTf )4 adduct ( 29Si NMR, d ¼ 62:0). The 11 B NMR spectrum (d ¼ 3:17, Dv1=2 ¼ 28 Hz) corresponds to the B(OTf )4 anion whereas the 13 C NMR spectrum shows the presence of only one triﬂuoromethyl group (d ¼ 118, q, 1 JC; F ¼ 318 Hz) [52a]. It is highly probable the triﬂuoromethyl group on this complex very rapidly migrates among all the triﬂate groups. Trace amounts of Me3 SiB(OTf )4 are suﬃcient for reaction of aldehydes with silyl enol ethers. In the presence of an asymmetric center adjacent to the carbonyl group, the diastereoselectivity of the process can be changed by varying the volumes of the substituents on the silicon atom (Eq. (49)). Apparently, an increase in the size of the silyl group bound to the carbonyl

61

62

2 Boron and Silicon Lewis Acids for Mukaiyama Aldol Reactions

oxygen leads to limitation of possible pathways of approach of the nucleophile, thereby improving the diastereoselectivity of the reaction (See also Eqs. (3) and (4) [8]) [52b]. A particularly useful property of B(OTf )3 is its ability to form complexes with chlorosilanes R3 SiCl, giving silylating reagents which compare favorably with the R3 SiOTf/B(OTf )3 system [52c]. The possibility of generating very strong silylating reagents based on sterically hindered chlorosilanes enables the use of these compounds instead of more expensive silyl triﬂates. OSiR3 Me3SiOTf/B(OTf)3 (5 mol%)

+ Ph

CHO

CH2Cl2, –80 °C

Ph

Ph

Ph R3SiO

R3Si

O

(49)

syn : anti 8:1 97 : 1

Me3Si i-Pr3Si

A combination of Me3 SiOTf and sterically hindered organoaluminum compounds MAD or MABR is another example of the formation of very active R3 SiX [53a]. As follows from Eq. (50), these organoaluminum compounds coordinate the triﬂate anion more eﬃciently than B(OTf )3 . The Me3 SiOTf/MABR system makes it possible to perform the reactions of silyl enol ethers even with poorly reactive carbonyl compounds, such as pivalaldehyde and methyl isopropyl ketone. The Me3 SiOTf/MAD or MABR/ PhCHO combinations are also useful for initiating cationic polymerization of silyl enolates [53b].

OSiMe3

+ Ph

CHO

Ph

t-Bu R

O t-Bu

1. cat. (5 mol%) CH2Cl2, –78 °C, 1 h 2. HCl

t-Bu O

Al Me

t-Bu

R=Me: MAD; R=Br: MABR

HO cat.

R

Ph

Ph

Me3SiOTf Me3SiOTf/B(OTf)3 Me3SiOTf/MAD Me3SiOTf/MABR MABR

O

yield, %

(50)

15 43 62 76 7

Study of complex formation between benzaldehyde, with MAD, and Me3 SiOTf by 13 C NMR spectroscopy at 50 C showed that addition of 2 equiv. Me3 SiOTf to the PhCHO ! MAD adduct aﬀorded a new electrophilic species of unknown nature. The 13 C NMR spectrum of the latter has a signal which is shifted downﬁeld by approximately 3 ppm compared with the signal of the PhCHO ! MAD complex. This species probably consists of benzaldehyde and two diﬀerent Lewis acids, and it behaves as a true electrophile, which attacks the double bond of silyl enol ether [53b].

Representative Experimental Procedures

In addition to these examples, it should be noted that activation of R3 SiX might cause undesirable transformations. Thus, it is diﬃcult to achieve high enantioselectivity when performing catalytic asymmetric aldol reactions of aldehydes with silyl enol ethers in the presence of chiral metalcentered Lewis acids. These diﬃculties are generally attributed to the eﬀect of R3 SiX, which is generated in the early steps of the process and then promotes carbon–carbon cross-coupling yielding a racemic product [45b, 46, 54]. Noyori also demonstrated that aldehydes do not react with silyl enol ethers under the action of Me3 SiOTf [29, 41, 42]. In this connection it is reasonable to assume that the low enantioselectivity observed might result from formation of a complex between R3 SiX and a chiral Lewis acid. Under the action of this complex the trialkylsilyl fragment can transferred to the carbonyl group, producing a racemic product.

Representative Experimental Procedures Typical Procedure for the Mukaiyama Aldol Reaction Catalyzed by B(C6 F5 )3 [13]. An anhydrous solution of B(C6 F5 )3 in toluene (81 mL, 0.02 mmol, 0.247 m) is added dropwise, at 78 C, under argon, to a solution of aldehyde (1.0 mmol) and silyl enol ether (1.2 mmol) in dichloromethane (2 mL). The mixture is stirred for several hours at the same temperature and then 1 m HCl (10 mL) and THF (10 mL) are added. The reaction mixture is stirred for 0.5 h, poured into NaHCO3 solution, extracted with diethyl ether, dried over MgSO4 and concentrated, and the residue is puriﬁed by column chromatography on silica gel to give the corresponding aldol in high yield. Preparation of 3,5-Bis(triﬂuoromethyl)phenylboron Dichloride (15) [24]. A solution of 13 (Lancaster Synthesis; 1 H NMR (C6 D6 , 300 MHz) d (ppm) 7.81 (s, 1H), 8.01 (s, 2H); 3.86 g, 15 mmol) in benzene (30 mL) is heated under reﬂux with removal of water (CaH2 in a Soxhlet thimble) for 2–5 h (oil bath: 100–105 C) then cooled to room temperature and concentrated in vacuo to give trimeric anhydride 14 as a white solid ( 1 H NMR (C6 D6 , 300 MHz) d 8.01 (s, 1H), 8.46 (s, 2H)). A 1 m solution of BCl3 (30 mL, 30 mmol) in hexane and a 1 m solution of BBr3 (30 mL, 30 mmol) in heptane are added separately to 14 at room temperature under argon. The two reaction mixtures are heated under reﬂux for 4 h (oil bath 100–105 C) and 56 h (oil bath 105–110 C), respectively, and the solvents are removed by distillation. Dichloroboron compound 15 is isolated as colorless oils by distillation under reduced pressure from the residues in ca. 40–50% yield: 38–40 C (0.05–0.06 torr); 1 H NMR (C6 D6 , 300 MHz) d 7.80 (s, 1H), 8.12 (s, 2H); 11 B NMR (C6 D6 , 96 MHz) d 53.2; 13 C NMR (C6 D6 , 75.5 MHz) d 123.1 (q, J ¼ 272:8 Hz, 2C), 127.1 (s, 1C), 131.0 (q, J ¼ 33:5 Hz, 2C), 134.8–135.2 (m, 1C), 135.5 (s, 2C); 19 F NMR (C6 D6 , 282 MHz) d 64.3.

63

64

2 Boron and Silicon Lewis Acids for Mukaiyama Aldol Reactions

Preparation of Chiral Oxazaborolidine Catalyst 2f [24]. 15 (22.1 mg, 0.075 mmol) is added at room temperature, under argon, to a solution of N-( ptoluenesulfonyl)-(S)-tryptophan [23a] (32.3 mg, 0.09 mmol) in dichloromethane (0.75 mL). The mixture is stirred for 1 h then concentrated in vacuo to give 2f as a white solid. This is dissolved in propionitrile and used for Mukaiyama aldol reactions. 1 H NMR (CD2 Cl2 , 300 MHz) d 2.37 (s, 3H), 3.56 (dd, J ¼ 2:6, 15.0 Hz, 1H), 3.83 (dd, J ¼ 4:5, 15.0 Hz, 1H), 4.56–4.59 (m, 1H), 7.08–7.30 (m, 4H), 7.26 (d, J ¼ 8:1 Hz, 2H), 7.54 (d, J ¼ 8:1 Hz, 2H), 7.82 (d, J ¼ 7:5 Hz, 1H), 7.95 (s, 1H), 8.04 (s, 2H), 8.22 (brs, 1H); 11 B NMR (CD2 Cl2 , 96 MHz) d 33.8; 19 F NMR (C6 D6 , 282 MHz) d 64.2. Representative Procedure for the Mukaiyama Aldol Reaction Catalyzed by 2f (Method A, Table 2.1) [24]. Propionitrile (1 mL) is added at room temperature to 2f (0.075 mmol, 6 mol%) prepared as described above. After being cooled to 78 C, benzaldehyde (127 mL, 1.25 mmol) is added and a solution of 1-phenyl-1-(trimethylsiloxy)ethylene (308 mL, 1.5 mmol) in propionitrile (0.5 mL) is subsequently added dropwise over 2 min. The reaction mixture is stirred at 78 C for 12 h and then quenched by addition of saturated aqueous NaHCO3 . The mixture is extracted with ether and the combined organic phases are dried over MgSO4 and evaporated. The residue is dissolved in THF (2 mL) and 1 m aqueous HCl (2 mL), and the resulting solution is left to stand for 30 min. Saturated aqueous NaHCO3 is added and the mixture is extracted with ether. The combined organic phases are dried over MgSO4 and evaporated to furnish an oily residue. Silica gel chromatography (hexane–ethyl acetate, 4:1) aﬀords 282 mg (> 99% yield) of the known aldol product. The enantiomeric ratio and the absolute conﬁguration are determined by HPLC analysis (Daicel OD-H column with hexane–iPrOH, 20:1, ﬂow rate 1.0 mL min1 ): tR ¼ 21:2 min ((S), minor enantiomer), 24.4 min ((R), major enantiomer). Preparation of Me3 SiOTf [30b]. Allyltrimethylsilane (1.6 g, 14 mmol) is added dropwise, with stirring, to a solution of TfOH (1.5 g, 10 mmol) in dry dichloromethane (8 mL) and the reaction temperature is maintained between 15 and 20 C for 1 h. The resulting mixture is transferred directly to a distillation apparatus and distillation under reduced pressure gives Me3 SiOTf as a colorless liquid (1.9 g, 85%), bp 52–53 C at 31 Torr; 1 H NMR (CDCl3 ) d 0.50 ppm. Preparation of Me3 SiNTf2 [30b]. HNTf2 (0.85 g) is reacted with a 3:1 molar excess of Me3 SiH in an FEP reactor. When the mixture is left to warm from 196 C an initial rapid reaction occurs near 22 C. After 4 h and occasional agitation, the homogeneous mixture is cooled slowly to 196 C. A quantitative amount of hydrogen is recovered and the excess silane is pumped away at 10 C, giving Me3 SiNTf2 (0.98 g, 92%) as a colorless liquid of low volatility. 19 F NMR (CFCl3 ) 77.47 (s) ppm; 1 H NMR 0.57 (s) ppm; major

References

m/e [Cl] 163 (Me3 SiOHþ ), 147 (TfNþ ), 77(?), 73 (Me3 Siþ ) with weak ions at 282 (Tf2 NH2 þ ) and 354 (Mþ ). Preparation of Me3 SiB(OTf )4 [30b]. TfOH (531 mL, 6 mmol) is added to BBr3 (distilled from Al powder; 190 mL, 2 mmol) at 0 C. After evolution of HBr has ceased the ﬂask is evacuated for 1 h to give B(OTf )3 as a viscous yellow liquid. In a separate ﬂask, a solution of Me3 SiOTf is prepared by addition of TfOH (177 mL, 2 mmol) to a solution of allyltrimethylsilane (320 mL, 2 mmol) in dichloromethane (5 mL), and the resulting mixture is left to stand for 10 min. Addition of this solution to the B(OTf )3 at 0 C results in evolution of heat and formation of a pale yellow solution of Me3 SiB(OTf )4 (0.4 m). Typical Procedure for the Mukaiyama Aldol Reaction Catalyzed by Me 3 SiNTf [45a]. Commercially available triﬂylimide (0.072 m solution in diethyl ether, 1.11 mL, 0.08 mmol) is added at 78 C under argon to a solution of silyl enol ether (8.8 mmol) in diethyl ether (2 mL). After stirring the mixture for 15 min, aldehyde or ketone (1.0 m solution in diethyl ether, 8.0 mL, 8.0 mmol) is added dropwise over a period of 2 h at 78 C. After stirring for 15 min at the same temperature, 1 m HCl (10 mL) and THF (10 mL) are added. The reaction mixture is stirred for 0.5 h, poured into NaHCO3 solution, extracted with diethyl ether, dried over MgSO4 and concentrated, and the residue is puriﬁed by column chromatography on silica gel to give the corresponding aldol in high yield. References 1 (a) Ishihara, K. In Lewis Acids in Organic Synthesis;

2 3 4 5 6 7 8 9 10

Yamamoto, H. Ed.; Wiley–VCH: Weinheim, 2000; Volume 1, pp. 89–190. (b) Ishihara, K. In Lewis Acids Reagents; Yamamoto, H. Ed.; Oxford University Press: Oxford, 1999; pp. 31–63. Wulf, W. D.; Gilbertson, S. R. J. Am. Chem. Soc. 1985, 107, 503. Yamago, S.; Machii, D.; Nakamura, E. J. Org. Chem. 1991, 56, 2098. Nakamura, E.; Kuwajima, I. J. Am. Chem. Soc. 1977, 99, 961. Sugimura, H.; Shigekawa, Y.; Uematsu, M. Synlett 1991, 153. Heathcock, C. H.; Flippin, L. A. J. Am. Chem. Soc. 1983, 105, 1667. Evans, D. A.; Gage, J. R. Tetrahedron Lett. 1990, 31, 5053. Davis, A. P.; Plunkett, S. J.; Muir, J. E. Chem. Commun. 1998, 1797. Guindon, Y.; Pre´vost, M.; Mochirian, P.; Gue´rin, B. Org. Lett. 2002, 4, 1019. Kalbaka, G. W.; Tejedor, D.; Li, N.-S.; Malladi, R. R.; Trotman, S. J. Org. Chem. 1998, 63, 6438.

65

66

2 Boron and Silicon Lewis Acids for Mukaiyama Aldol Reactions 11 A review of B(C6 F5 )3 : Piers, W. E.; Chivers, T. Chem. Soc.

Rev. 1997, 26, 345. 12 For preparation of B(C6 F5 )3 , see: (a) Massey, A. G.; Park, A. J.

13

14 15

16 17

18

19

20

21

22 23

24

J. Organomet. Chem. 1964, 2, 245. (b) Massey, A. G.; Park, A. J. J. Organomet. Chem. 1966, 5, 218. (a) Ishihara, K.; Hanaki, N.; Yamamoto, H. Synlett 1993, 577. (b) Ishihara, K.; Funahashi, M.; Hanaki, N.; Miyata, M.; Yamamoto, H. Synlett 1994, 963. (c) Ishihara, K.; Hanaki, N.; Funahashi, M.; Miyata, M.; Yamamoto, H. Bull. Chem. Soc. Jpn. 1995, 68, 1721. Ishihara, K.; Kurihara, H.; Yamamoto, H. Synlett 1997, 597. (a) Mori, Y.; Manabe, K.; Kobayashi, S. Angew. Chem. Int. Ed. 2001, 40, 2816. (b) Mori, Y.; Kobayashi, J.; Manabe, K.; Kobayashi, S. Tetrahedron 2002, 58, 8263. Reetz, M.; Kunish, F.; Heitmann, P. Tetrahedron Lett. 1986, 27, 4721. (a) Kiyooka, S.-I.; Kaneko, Y.; Komura, M.; Matsuo, H.; Nakano, M. J. Org. Chem. 1991, 56, 2276. (b) Kaneko, Y.; Matsuo, T.; Kiyooka, S. Tetrahedron Lett. 1994, 35, 4107. (c) Kiyooka, S.; Kaneko, Y.; Harada, Y.; Matsuo, T. Tetrahedron Lett. 1995, 16, 2821. (d) Kiyooka, S.; Kira, H.; Hena, M. A. Tetrahedron Lett. 1996, 37, 2597. (e) Kiyooka, S.; Hena, M. A. Tetrahedron: Asymmetry 1996, 7, 2181. (f ) Kiyooka, S.; Maeda, H. Tetrahedron: Asymmetry 1997, 8, 3371. (g) Kiyooka, S.; Maeda, H.; Hena, M. A.; Uchida, M.; Kim, C.-S.; Horiike, M. Tetrahedron Lett. 1998, 39, 8287. (a) Goodman, J. M. Tetrahedron Lett. 1992, 33, 7219. (b) Corey, E. J.; Rohde, J. J.; Fischer, A.; Azimioara, M. D. Tetrahedron Lett. 1997, 38, 33. (c) Corey, E. J.; Rohde, J. J. Tetrahedron Lett. 1997, 38, 37. (d) Corey, E. J.; BarnesSeeman, D.; Lee, T. W. Tetrahedron Lett. 1997, 38, 1699. (e) Corey, E. J.; Brans-Seeman, D.; Lee, T. W. Tetrahedron Lett. 1997, 38, 4351. (f ) Corey, E. J.; Lee, T. W. Chem. Commun. 2001, 1321. (a) Furuta, K.; Maruyama, T.; Yamamoto, H. J. Am. Chem. Soc. 1991, 113, 1041. (b) Furuta, K.; Maruyama, T.; Yamamoto, H. Synlett 1991, 439. (c) Ishihara, K.; Maruyama, T.; Mouri, M.; Gao, Q.; Furuta, K.; Yamamoto, H. Bull. Chem. Soc. Jpn. 1993, 66, 3483. (a) Parmee, E. R.; Tempkin, O.; Masamune, S. J. Am. Chem. Soc. 1991, 113, 9365. (b) Parmee, E. R.; Hong, Y.; Tempkin, O.; Masamune, S. Tetrahedron Lett. 1992, 33, 1729. (a) Kiyooka, S.; Kaneko, Y.; Kume, K. Tetrahedron Lett. 1992, 33, 4927. (b) Kiyooka, S.; Yamaguchi, T.; Maeda, H.; Kira, H.; Hena, M. A.; Horiike, M. Tetrahedron Lett. 1997, 38, 3553. Corey, E. J.; Cywin, C. L.; Roper, T. D. Tetrahedron Lett. 1992, 33, 6907. (a) Corey, E. J.; Loh, T.-P. J. Am. Chem. Soc. 1991, 113, 8966. (b) Corey, E. J.; Loh, T.-P.; Roper, T. D.; Azimioara, M. D.; Noe, M. C. J. Am. Chem. Soc. 1992, 114, 8290. (a) Ishihara, K.; Kondo, S.; Yamamoto, H. Synlett 1999, 1283. (b) Ishihara, K.; Kondo, S.; Yamamoto, H. J. Org. Chem. 2000, 65, 9125.

References 25 Reilly, M.; Oh, T. Tetrahedron Lett. 1995, 36, 221. 26 (a) Kinugasa, M.; Harada, T.; Fujita, K.; Oku, A. Synlett

27 28 29 30

31

32 33

34 35 36 37 38 39

40

41 42

43

1996, 43. (b) Kinugasa, M.; Harada, T.; Egusa, T.; Fujita, K.; Oku, A. Bull. Chem. Soc. Jpn. 1996, 69, 3639. (c) Kinugasa, M.; Harada, T.; Oku, A. J. Org. Chem. 1996, 61, 6772. (d) Kinugasa, M.; Harada, T.; Oku, A. J. Am. Chem. Soc. 1997, 119, 9067. (e) Kinugasa, M.; Harada, T.; Oku, A. Tetrahedron Lett. 1998, 39, 4529. (f ) Harada, T.; Egusa, T.; Kinugasa, M.; Oku, A. Tetrahedron Lett. 1998, 39, 5531. (g) Harada, T.; Egusa, T.; Oku, A. Tetrahedron Lett. 1998, 39, 5535. (h) Harada, T.; Nakamura, T.; Kinugasa, M.; Oku, A. Tetrahedron Lett. 1999, 40, 503. Nevalainen, V.; Mansikka, T.; Kostiainen, R.; Simpura, I.; Kokkonen, J. Tetrahedron: Asymmetry 1999, 10, 1. (a) Komura, K.; Nishitani, N.; Itsuno, S. Polym. J. 1999, 31, 1045. (b) Itsuno, S.; Komura, K. Tetrahedron 2002, 58, 8237. Noyori, R.; Murata, S.; Suzuki, M. Tetrahedron 1981, 37, 3899. (a) Dilman, A. D.; Loffe, S. Chem. Rev. 2003, 103, 733. (b) Oishi, M. In Lewis Acids in Organic Synthesis; Yamamoto, H. Ed.; Wiley–VCH: Weinheim, 2000; Volume 1, pp. 355–393. (c) Hosomi, A.; Miura, K. In Lewis Acids Reagents; Yamamoto, H. Ed.; Oxford University Press: Oxford, 1999; pp. 159–168. Maerker, C.; Kapp, J.; Schleyer, P. v. R. In Organosilicon Chemistry II; Auner, N., Weis, J.; VCH: Weinheim, 1996; pp. 329–359. Shin, S. K.; Beauchamp, J. L. J. Am. Chem. Soc. 1989, 111, 990. (a) Schwarz, H. In The Chemistry of Organic Silicon Compounds; Patai, S., Rappoport, Z., Eds.; Wiley: Chichester, 1989; Part 1, pp 445–510. (b) Chojnowski, J.; Stanczyk, W. A. Adv. Organomet. Chem. 1990, 30, 243. Lambert, J. B.; Zhao, Y.; Zhang, S. M. J. Phys. Org. Chem. 2001, 14, 370. Lambert, J. B.; Zhao, Y. Angew. Chem. Int. Ed. Engl. 1997, 36, 400. Lambert, J. B.; Lin, L. J. Org. Chem. 2001, 66, 8537. ¨ller, T.; Zhao, Y.; Lambert, J. B. Organometallics 1998, 17, Mu 278. Hergott, H. H.; Simchen, G. Liebigs Ann. Chem. 1980, 1718. (a) Bassindale, A. R.; Stout, T. J. Chem. Soc., Perkin Trans. 2 1986, 221. (b) Bassindale, A. R.; Lau, J. C.-Y.; Stout, T.; Tayor, P. G. J. Chem. Soc. Perkin Trans. 2 1986, 227. Mayr, H.; Gorath, G. J. Am. Chem. Soc. 1995, 117, 7862. (b) Kira, M.; Hino, T.; Sakurai, H. Chem. Lett. 1992, 555. (c) Prakash, G. K. S.; Wang, Q.; Rasul, G.; Olah, G. A. J. Organomet. Chem. 1998, 550, 119. (d) Prakash, G. K. S.; Bae, C.; Rasul, G.; Olah, G. A. J. Org. Chem. 2002, 67, 1297. Murata, S.; Suzuki, M.; Noyori, R. Tetrahedron 1988, 44, 4259. (a) Murata, S.; Suzuki, M.; Noyori, R. J. Am. Chem. Soc. 1980, 102, 3248. (b) Murata, S.; Suzuki, M.; Noyori, R. Tetrahedron Lett. 1980, 21, 2527. Mukai, C.; Hashizume, S.; Nagami, K.; Hanaoka, M. Chem. Pharm. Bull. 1990, 38, 1509.

67

68

2 Boron and Silicon Lewis Acids for Mukaiyama Aldol Reactions 44 Kawai, M.; Onaka, M.; Izumi, Y. Bull. Chem. Soc. Jpn 1988,

61, 1237. 45 (a) Ishihara, K.; Hiraiwa, Y.; Yamamoto, H. Synlett 2001,

46 47

48 49

50 51 52

53

54

1851. (b) Ishihara, K.; Hiraiwa, Y.; Yamamoto, H. Chem. Commun. 2002, 1564. Hollis, T. K.; Bosnich, B. J. Am. Chem. Soc. 1995, 117, 4570. (a) Ooi, T.; Tayama, E.; Takahashi, M.; Maruoka, K. Tetrahedron Lett. 1997, 38, 7403. (b) Chen, J.; Sakamoto, K.; Orita, A.; Otera, J. J. Org. Chem. 1998, 63, 9739. (c) Otera, J.; Chen, J. Synlett 1996, 321. Molander, G. A.; Cameron, K. O. J. Org. Chem. 1991, 56, 2617. (a) Olah, G. A.; Field, L. D. Organometallics 1982, 1, 1485. (b) Olah, G. A.; Heiliger, L.; Li, X.-Y.; Prakash, G. K. S. J. Am. Chem. Soc. 1990, 112, 5991. Iwasawa, N.; Mukaiyama, T. Chem. Lett. 1987, 463. Mukaiyama, T.; Ohno, T.; Han, J. S.; Kobayashi, S. Chem. Lett. 1991, 949. (a) Davis, A. P.; Jaspars, M. Angew. Chem. Int. Ed. Engl. 1992, 31, 470. (b) Davis, A. P.; Plinkett, S. J. J. Chem. Soc., Chem. Commun. 1995, 2173. (c) Davis, A. P.; Muir, J. E.; Plunkett, S. J. Tetrahedron Lett. 1996, 37, 9401. (a) Oishi, M.; Aratake, S.; Yamamoto, H. J. Am. Chem. Soc. 1998, 120, 8271. (b) Oishi, M.; Yamamoto, H. Macromolecules 2001, 34, 3512. (a) Carreira, E. M.; Singer, R. A. Tetrahedron Lett. 1994, 35, 4323. (b) Denmark, S. E.; Chen, C.-T. Tetrahedron Lett. 1994, 35, 4327. (c) For detailed mechanistic discussion of catalytic asymmetric Mukaiyama aldol reaction, see: Carreira, E. M. In Comprehensive Asymmetric Catalysis, Jacobsen, E. N., Pfaltz, A.; Yamamoto, H., Eds.; Springer: Heidelberg, 1999; Vol. 3, pp 997–1065.

69

3

Copper Lewis Acids Jeﬀrey S. Johnson and David A. Nicewicz 3.1

Introduction

Copper complexes serve as structurally diverse Lewis acids that promote additions of enolates and latent enolates to carbonyl compounds. The exact mode of activation depends on the complex: many copper(II) complexes are known to eﬀectively activate the electrophilic component in aldol additions whereas copper(I) complexes are implicated in aldol reactions that feature nucleophile activation (Scheme 3.1). Irrespective of the mechanistic details, when the metal complex carries stereochemical information in its ligand framework, chirality transfer to the nascent carbinol stereogenic center can be nearly complete. This review will survey nucleophilic addition of enolates and latent enolates to carbonyl compounds catalyzed by copper Lewis acids. Particular attention will be paid to stereoselective variants and the development of stereochemical models to account for observed enantiomeric enrichment. Applications to natural product synthesis will be highlighted. A distinction is drawn between carbonyl activation in a Mukaiyama aldol sense and nucleophile activation via a metalloenolate; because each of these reaction-types do involve Lewis acid–Lewis base interactions, however, both reaction families will be included in this chapter. Coverage will focus on catalytic examples.

3.2

Early Examples

The ability of Cu(II) ion to promote the addition of acetone to aromatic aldehydes in crossed-aldol condensation reactions was demonstrated by Iwata and Emoto in 1974 [1]. Subsequent extension to a regioselective crossed aldol reaction with 2-butanone was later described by Irie and Watanabe [2]. Both of these early examples employ more than one equivalent of Cu(II) source relative to the aldehyde. As a forerunner to his pioneering Au(I) work, Ito reported in 1985 that Modern Aldol Reactions. Vol. 2: Metal Catalysis. Edited by Rainer Mahrwald Copyright 8 2004 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 3-527-30714-1

70

3 Copper Lewis Acids

O

[Cu]

X

or

X

2

OSiR3 1 R 2

3

R R electrophile activation

O R

OSiR3 1 R

[Cu]

O R2 OH

3

R

3

R

X 1

O 1

R

X

O

[Cu]

R

[Cu] 1

R

X

O

nucleophile activation

2

3

R

R

[Cu] = Cu(I) or Cu(II) complex Scheme 3.1

Modes of activation for Cu-catalyzed aldol reactions.

catalytic quantities of a Cu(I) catalyst could be employed to promote addition of ethyl isocyanoacetate (1) to a,b-unsaturated aldehydes (Eq. (1)) [3]. The reactions are selective for formation of the trans-4,5-disubstituted oxazoline adducts (3). A footnote of that paper indicates that enantioselective variants of this reaction are possible employing ()-ephedrine as a scalemic additive. CO2Et

CO2Et C

5 mol % CuCl/Et 3N

O

R

N

N

H 1

R

CO2Et N O 3a 60% trans:cis 4:1

THF, r.t.

O 3

2 CO2Et Me

CO2Et Ph

N O 3b 75%

N

Me

O 3c ca. 100% trans:cis 2:1

CO2Et N

(1)

CH(Et)2

O 3d 75% trans only

3.3

Mukaiyama Aldol Reactions with Cu(II) Complexes 3.3.1

Enolsilane Additions to (Benzyloxy)acetaldehyde Scope and Application In 1996 Evans and coworkers reported highly enantioselective additions of latent enolates to (benzyloxy)acetaldehyde (7) catalyzed by enantiomerically pure pyridyl bis(oxazoline) Cu(II) complexes (4, hereafter (pybox)CuL n ) [4, 3.3.1.1

3.3 Mukaiyama Aldol Reactions with Cu(II) Complexes

5]. The reaction is a Mukaiyama aldol addition in which the aldehyde is activated toward nucleophilic addition by the electropositive Cu(II) center. The adduct is a b-silyloxy ester derivative that is readily desilylated under acidic conditions for the purpose of analyzing the enantiomeric enrichment of the product (Eq. (2)). Simultaneous investigations revealed that a bidentate C2 symmetric bis(oxazoline) ligand is also an eﬀective chiral control element, albeit with slightly reduced levels of enantiocontrol (Eq. (3)). The pendant phenyl substituent is optimal for the pybox ligand, and the tert-butyl group is most eﬀective among those surveyed for the bis(oxazoline) scaﬀold.

(2)

ð3Þ Catalyst preparation depends on the identity of the counter-anion, which has a marked eﬀect both on rate and selectivity. Bis(oxazoline)Cu(OTf )2 (5, hereafter (box)Cu(OTf )2 ) and (pybox)Cu(OTf )2 complexes are prepared simply by mixing equimolar quantities of the ligand and Cu(OTf )2 in CH2 Cl2 . The corresponding (ligand)Cu(SbF6 )2 complexes are synthesized via anion metathesis of the (ligand)CuCl2 complexes with two equivalents of AgSbF6 . Filtration of the resulting AgCl salt gives a clear blue or green solution of the active catalyst complex. (Ph-pybox)Cu(SbF6 )2 -catalyzed additions to (benzyloxy)acetaldehyde are highly enantioselective for several acetate-type nucleophiles derived from thio- and oxo-esters (Figure 3.1). Less ﬂexibility is possible with the electrophile. p-Methoxybenzyloxyacetaldehyde is an excellent substrate for the addition, but butoxyacetaldehyde is somewhat less selective. Enantiocontrol is signiﬁcantly less for aldehydes nominally incapable of chelation. The Chan diene (13) and dioxolanone-derived nucleophile (14) both serve as eﬀective acetoacetate nucleophile equivalents in asymmetric catalyzed additions to benzyloxyacetaldehyde (Scheme 3.2). The former example was optimized to employ only 2 mol% chiral catalyst to deliver multigram

71

72

O

OH

X

3 Copper Lewis Acids

8 X = SCMe3 99% ee 98% ee 10 X = SEt OBn 98% ee 11 X = OEt

O

OH

Me3CS

99% ee R = OBn 88% ee R = OBu R = OPMB 99% ee R = OTBS 56% ee R = CH2Ph <10% ee

12a 12b R 12c 12d 12e

Fig. 3.1

Enantioenriched aldol adducts derived from (Ph-pybox)Cu(SbF6 )2 -catalyzed reactions (Eq. (2)).

quantities of essentially optically pure material (15). On these scales the reaction must be initiated at a low temperature, because the reaction is highly exothermic. The d-hydroxy-b-keto ester product has been diastereoselectively reduced to aﬀord either the syn (16) or anti (17) diol product in good yield. The Chan diene addition product provides a useful entry into polyacetate building blocks. To this end, the enantiomeric (R,R-Ph-pybox)Cu(SbF6 )2 TMSO

O

OTMS

OBn

H

Me3CO 13

7 2+

O

N N Cu N 4 Ph 2 mol %

O

O

2 SbF6

OH OH OBn

Me3CO 16

Ph

O

O

OH

OBn 15 10 g, 85%, 99% ee

CH2Cl2, -93 →-78 °C; then PPTS, MeOH

Me3CO

O

OH OH OBn

Me3CO 17 2+ O Me O

N

Me O

OBn

H

MeO

Ph

O

14

7

N Cu N 4 5 mol %

O

Enantioselective aldol reactions of acetoacetate nucleophile equivalents catalyzed by (Ph-pybox)Cu(SbF6 )2 complexes.

Me

Me

O

O

Ph

CH2Cl2, -78 °C; then aq. HCl/THF

Scheme 3.2

2 SbF6

O

OH OBn

18 94%, 92% ee

3.3 Mukaiyama Aldol Reactions with Cu(II) Complexes

73

complex has been used to deliver the needed aldol enantiomer (ent-15) for ultimate transformation to the cytostatic natural product phorboxazole B [6], and the bryostatin family of antitumor agents (Scheme 3.3) [7]. It is noteworthy that in the former reaction a common asymmetric aldol product provides a common starting material for two diﬀerent pyran rings. A stereoselective catalyzed vinylogous aldol addition was developed for application to the asymmetric synthesis of callepeltoside A (Scheme 3.4) [8]. The reaction makes use of an air-stable hydrated catalyst, [((R,R)-Phpybox)Cu(OH2 )2 ](SbF6 )2 (22), to eﬀect the formation of the d-hydroxy-a,bunsaturated ester 21 with complete E selectivity and excellent enantiocontrol. Substituted (propionate-type) silylketene acetals also add to (benzyloxy)acetaldehyde with high diastereo- and enantiocontrol under the inﬂuence of (pybox)Cu(SbF6 )2 catalysis (Figure 3.2). A range of cyclic and acyclic nucleophiles participate in diastereo- and enantioselective aldol reactions to give the syn aldol diastereomer in all cases but one (23a–23e). The syn selectivity predominates irrespective of the geometry of the starting silylketene acetal. The only exception to this trend is 2-trimethylsilyloxyfuran, which aﬀords the anti diastereomer 23f in good chemical and optical yield. Kunieda and coworkers reported a modiﬁed catalyst system in 1999 that probes the eﬀect of the backbone spacer connecting the two oxazoline rings and steric congestion about the metal center [9]. Anthracene-based bis(oxazoline)Cu(II) complexes were prepared and tested in the addition of t-butyl thioester silylketene acetal to (benzyloxy)acetaldehyde (Eq. (4)). The methylene-bridged complex 24Cu(OTf )2 strongly favors formation of the R enantiomer, whereas extending the linking chain by one CH2 group (25Cu(OTf )2 ) results in a selectivity turnover to favor the S enantiomer. The authors propose that a change in aldehyde binding geometry could result from the structural perturbation. In neither reaction is selectivity superior to that of the Evans system.

ligand (10 mol %) O

OTMS

OBn

H

Me3CS

OH

O

Cu(OTf)2 (10 mol %) CH2Cl2, -78 °C

O OBn

Me3CS

R

O

O N

R

N

N

R

R 24a, R = H 24b, R = Me

ligand 24a 24b 25a 25b R

N

R

R

8R

O

O

25a, R = H 25b, R = Me

OBn

Me3CS

8S

R

OH

8S:8R 1:24 1:3.3 9:1 3.2:1

ð4Þ

H

O

Me

OBn

2 mol %

N Ph

O

Br

H O HO

*

OMe

O

N Me

H O

Me

O

O OH

O

Me O

O N

*

H O H H O

CH2

H

OH

* ent-15 OBn 85%, >99% ee

Me3CO

2 SbF6

phorboxazole B

CH2Cl2, -93 to -78 °C

Ph

N

N Cu

Application of enantioselective (Ph-pybox) Cu(SbF6 )2 -catalyzed aldol reactions to pyran-containing natural products.

HO

OTMS

MeO Scheme 3.3

Me3CO

TMSO

O

2+

Pr

MeO2C

OH

O

CO2Me

O OH H OH O O Me O Me OH

O

bryostatin 2

Me

* O

Me Me HO

74

3 Copper Lewis Acids

3.3 Mukaiyama Aldol Reactions with Cu(II) Complexes

2+ O

Me

TMSO

Ph

O

EtO

2 SbF 6

Ph

2.5 mol %

H OPMB 19

N N Cu N H2O OH2 22

O

CH2Cl2, -78 ° C; aq. HCl

20 O Me O MeO O

O

Me

EtO

OH

Me

*

Me OPMB

H MeO

21 93%, 95% ee

Me

O

H

O

NH Me

OH O O * Cl

callepeltoside A Scheme 3.4

Application of enantioselective (Ph-pybox) Cu(SbF6 )2 -catalyzed vinylogous aldol reactions to callepeltoside A.

Mechanism and Stereochemistry The proposed mechanism for the (pybox)Cu(SbF6 )2 catalyzed addition reaction involves activation of the chelating electrophile by the metal center (26), nucleophilic addition (26 ! 27), silylation of the metal aldolate (27 ! 28), and release of the neutral product (28 ! 29 þ 4) (Scheme 3.5). The silicon-transfer step of this mechanism has a signiﬁcant intermolecular component, as evidenced by double-labeling experiments (Scheme 3.6). The identity of the species responsible for intermolecular silyl transfer is not known, although the metal aldolate and Me3 SiSbF6 are potential candidates. What is apparent from the enantiomeric enrichment of the aldols in the crossover experiment is that no stereochemical ‘‘leakage’’ occurs in this process: any potential achiral aldol catalyst [10] is not competitive with the chiral cationic Cu(II) complex. The asymmetric aldol reaction catalyzed by (Ph-pybox)Cu(SbF6 )2 has a signiﬁcant positive non-linear eﬀect when complexes are prepared from enantioimpure pybox ligands. Experimental evidence points to a catalytically inactive heterochiral dimer as the source of this non-linear eﬀect (reservoir 3.3.1.2

75

76

3 Copper Lewis Acids

O

OH

O OBn

EtS

OH

EtS

Me

Me 23c

85% yield, 95% ee syn:anti 95:5 O

OH OBn

EtO

OBn

23b Me

90% yield, 97% ee syn:anti 97:3

H

OH

Me3CS

Me

23a

O

O OBn

86% yield, 99% ee syn:anti 85:15

OH

H

OBn

O

Me

OH OBn

O

23d

O

23e

60% yield, 87% ee syn:anti 84:16

23f

93% yield, 92% ee syn:anti 9:91

95% yield, 95% ee syn:anti 95:5

Fig. 3.2

Aldol adducts derived from enantio- and diastereoselective (Ph-pybox)Cu(SbF6 )2 catalyzed aldol reactions by use of substituted silylketene acetals.

N O Me3CS

OTMS O

O

2+ 2 SbF6

Bn

7

4

Bn

O

H

N Cu N

29 2+ N 2 SbF N 6 O Cu N O Bn H

2+ N 2 SbF 6

O Me3CS

N OTMS Cu N O Bn 28

26 OTMS 2+ N 2 SbF 6

TMSO Me3CS

N O Cu N O Bn 27

Scheme 3.5

Proposed mechanism for (Ph-pybox) Cu(SbF6 )2 -catalyzed enantioselective aldol reactions.

Me3CS 6

3.3 Mukaiyama Aldol Reactions with Cu(II) Complexes O OSiMe3

OX

O OBn

H

O

EtS

EtS

30 0.5 equiv

O OBn

Scheme 3.6

Crossover experiments to verify intermolecular silicon transfer in enantioselective (Ph-pybox)Cu(SbF6 )2 catalyzed aldol reactions.

eﬀect) [11]. Corroboration was obtained via crystallization of the heterochiral dimer 34, demonstrating that its formation is indeed feasible (Figure 3.3). Semiempirical calculations support the notion that the homochiral dimer is less stable than the heterochiral dimer, accounting for the positive non-linear eﬀect. Substantial insight into the mechanism of asymmetric induction has been obtained via crystallization of monomeric [(pybox)CuL n ](SbF6 )2 complexes. An X-ray structure of [(i-Pr-pybox)Cu(OH2 )2 ](SbF6 )2 (35) reveals square pyramidal geometry with one water molecule occupying the coordination site in the ligand plane and the second water molecule occupying the axial position (Figure 3.4). Neither counter-ion is within the coordination sphere of the metal. It is revealing that the CuaO bond length in the ligand plane is considerably shorter than the CaOaxial bond length (1.985(7) compared with 2.179(7) A˚). For maximum electrophile activation, aldehyde coordination should occur in the ligand plane. The presence of the axial binding site provides a second ‘‘contact point’’ for the chelating carbonyl compound and introduces an additional element of substrate organization.

2+

H

O

O

N N N Cu N H N Ph N Ph H O

O Ph

OBn 8 99% ee

OX

32 X = SiMe3; 23% 33 X = SiMe2Et; 34%

Ph

2 SbF6

H

34 Fig. 3.3

Structure of heterochiral dimer [((R,R)-Ph-pybox)Cu((S,S)-Ph-pybox)](SbF6 )2 (34).

OH

Me3CS

29 X = SiMe3; 30% 31 X = SiMe2Et; 13%

7 1.0 equiv

OSiMe2Et

O OBn

Me3CS

Me3CS 6 0.5 equiv

77

OH OBn

EtS 10 99% ee

78

3 Copper Lewis Acids

2+ O N

N Cu

Me2HC H2O OH2

O

2 SbF6

N CHMe2

35 Fig. 3.4

Structure of [(i-Pr-pybox)Cu(OH2 )2 ](SbF6 )2 (35).

Ultimate corroboration of this mode of activation was obtained via crystallization of the catalyst–substrate complex [(Ph-pybox)Cu(BnOCH2 CHO)](SbF6 )2 (36). The aldehyde coordinates to form a ﬁve-membered ring chelate, with the ether oxygen occupying the axial position (Figure 3.5). The aromatic ring of the benzyloxy group is ca. 3.5 A˚ removed from the aromatic pyridyl ring in an oﬀset face–face arrangement, a p–p interaction that might explain the superior selectivities observed for p-PMBOCH2 CHO and BnOCH2 CHO compared with n-BuOCH2 CHO. Thus coordinated, the aldehyde re face is shielded by the proximal phenyl ring of the pybox ligand; addition to the si face is predicted and experimentally observed. This model predicts that (S)- and (R)-a-benzyloxypropionaldehyde will behave as matched and mismatched substrates in the addition. In accord with this proposed transition state assembly, the (S) isomer (R1 ¼ Me, R2 ¼ H) undergoes a highly eﬃcient and diastereoselective addition (2 h, dr ¼ 98.5:1.5) whereas the (R) isomer (R1 ¼ H, R2 ¼ Me) is a sluggish reaction partner and poorly diastereoselective (12 h, dr ¼ 50:50).

O N O

N Ph

Cu O

O

H Bn R2

H

N

R1

H

Ph

X OSiR 3

Fig. 3.5

X-ray structure of [(Ph-pybox)Cu (BnOCH2 CHO)](SbF6 )2 (36) and stereochemical model for enantioselective additions.

3.3 Representative Experimental Procedures

O N O

N Ph

O H

N Cu

O

Ph

Me O H H Bn O H R' R

O

N N Ph

H

R'

R,R' = OTMS, SR

O OBn

EtS

Cu

H Bn O

OH

Me

H

N

Ph O Me H R

OH OBn

EtS Me

Fig. 3.6

Stereochemical models for syn-selective aldol reactions catalyzed by (Ph-pybox) Cu(SbF6 )2 .

The syn diastereoselectivity is accounted for by an open antiperiplanar transition structure that minimizes gauche, dipole, and other through-space eﬀects (Figure 3.6). Representative Experimental Procedures Preparation of (S,S)-Ph-pybox)Cu(SbF6 )2 (4). In a nitrogen atmosphere box (S,S)-bis(phenyloxazolinyl)pyridine (18.5 mg, 0.05 mmol) and CuCl2 (6.7 mg, 0.05 mmol) were placed in an oven-dried round-bottomed ﬂask containing a magnetic stirring bar. In a nitrogen atmosphere box AgSbF6 (34.4 mg, 0.10 mmol) was placed in an oven-dried round-bottomed ﬂask containing a magnetic stirring bar. The ﬂasks were ﬁtted with serum caps and removed from the nitrogen atmosphere box. The ﬂask containing the ligand–CuCl2 mixture was charged with CH2 Cl2 (1.0 mL). The resulting suspension was stirred rapidly for 1 h to give a ﬂuorescent green suspension. AgSbF6 (in 0.5 mL CH2 Cl2 ) was added via a cannula with vigorous stirring, followed by a 0.5 mL rinse. The resulting mixture was stirred rapidly for 3 h in the absence of light and ﬁltered through an oven-dried glass pipet tightly packed with cotton to remove the white AgCl precipitate, yielding active catalyst as a clear blue solution. Catalyzed Addition of Silylketene Acetals to Benzyloxyacetaldehyde Using (S,S)Ph-pybox)Cu(SbF6 )2 . Benzyloxyacetaldehyde (70.0 mL, 0.50 mmol), followed by a silylketene acetal (0.60 mmol), were added to a 78 C solution of 4 in CH2 Cl2 . The resulting solution was stirred at either 78 or 50 C until the aldehyde was completely consumed (15 min to 48 h) as determined by TLC (30% EtOAc–hexanes). The reaction mixture was then ﬁltered through a 1.5

79

3 Copper Lewis Acids

80

cm 8 cm plug of silica gel with Et2 O (50 mL). Concentration of the ether solution gave a clear oil, which was dissolved in THF (10 mL) and 1 m HCl (2 mL). After standing at room temperature for 15 min, this solution was poured into a separatory funnel and diluted with Et2 O (10 mL) and H2 O (10 mL). After mixing, the aqueous layer was discarded, and the ether layer was washed with saturated aqueous NaHCO3 (10 mL) and brine (10 mL). The resulting ether layer was dried over anhydrous MgSO4 , ﬁltered, and concentrated to provide the hydroxy esters. 3.3.2

Enolsilane Additions to a-Keto Esters Scope and Application Dialkylketones are typically poor electrophiles in traditional aldol bond constructions, but the presence of a strong electron-withdrawing group in aketo esters engenders reactivity that more closely resembles that of aldehydes. Evans and coworkers described the ﬁrst catalytic, enantioselective enolsilane addition to pyruvate esters [12, 13]. The most eﬀective catalyst with regard to yield and enantiocontrol is the (t-Bu-box)Cu(OTf )2 complex 5 (Eq. (5)) and its corresponding hydrated derivative 5 (H2 O)2 (Eq. (6)). The latter is an air-stable solid with identical reactivity when used in the presence of a desiccant. In contrast to the pybox system the cationic complex [(t-Bu-box)Cu](SbF6 )2 results in reduced enantioselectivity [14]. 3.3.2.1

Me

Me

N

N

O

O

OTMS Me3CS

Cu Me3C TfO OTf CMe3

O CO2Me

Me 6

5

O HO Me Me3CS

-78 °C; then aq. HCl

37

CO2Me

38 THF: 95%, 99% ee CH2Cl2: 94%, 99% ee

ð5Þ Me

O N

OTMS Me3CS

37

CO2Me

2 OTf

N

Cu Me3CH O OH CMe3 2 2 5·(H2O)2

O Me

6

2+

Me

O

MS 3Å -78 ° C; then aq. HCl

O HO Me Me3CS

CO2Me 38

THF: 97% ee CH2Cl2: 99% ee

ð6Þ

3.3 Representative Experimental Procedures

81

The addition reactions can be eﬀectively performed in a range of solvents, including THF, Et2 O, CH2 Cl2 , PhMe, hexane, and PhCF3 . The enantiomeric excess is >94% for addition of the tert-butyl thioacetate silylketene acetal to methyl pyruvate in all of these solvents. Catalyst loadings down to 1 mol% are feasible. The temperature–enantioselectivity proﬁle has been studied and shown to be relatively ﬂat (99% ee at 78 C; 92% ee at þ20 C). Interestingly, the catalytic reaction in THF, a relatively good donor solvent, is signiﬁcantly faster than the identical reaction in CH2 Cl2 . Control experiments with stoichiometric quantities of (t-Bu-box)Cu(OTf )2 demonstrate that the actual addition step is faster in CH2 Cl2 , a fact consistent with the predicted deactivation of the Lewis acidic center in THF via solvent coordination. Accordingly, THF must play a role in promoting catalyst turnover. One postulated role of THF in the catalytic cycle is to act as a silicon shuttle, forming a more reactive silylating species (e.g. [THF-SiMe3 ]OTf ). The ‘‘silicon shuttle’’ hypothesis predicts signiﬁcant intermolecular crossover, which is experimentally borne out by double-labeling experiments in analogy to those described above for benzyloxyacetaldehyde additions. Silylation of the putative metal aldolate by an exogenous Si(þ) source results in signiﬁcant rate accelerations. For example, a catalyzed pyruvate addition that requires 14 h in the absence of an additive is complete in 0.5 h in the presence of 1.0 equiv. TMSOTf (Eq. (7)). The presence of stoichiometric quantities of this Lewis acid does not erode the selectivity of the reaction. The Cu(II) complex again reacts to complete exclusion of the achiral complex. Me

Me

N

N

O

O

OTMS Me3CS

O Me

6

CO2Me 37

Cu Me3C TfO OTf CMe3 5 2 mol % n equiv TMSOTf CH2Cl2, -78 °C; then aq. HCl;

O HO Me Me3CS

CO2Me 38 n = 0; reaction time = 14 h; 97% ee n = 1; reaction time = 0.5 h; 97% ee

ð7Þ The scope of the reaction with regard to the carboalkoxy and acyl moieties (electrophile) includes a range of substituents (Figure 3.7). a-Branched substrates (e.g. i-PrC(O)CO2 Me) result in low p-facial selectivity (39e) but comprise the only subset of poorly selective a-keto esters. Enolsilanes derived from acetone and acetophenone are eﬀective and selective nucleophiles in additions to methyl pyruvate (39g–h). Propionate silylketene acetals are also usually eﬀective (39i). As in the [(pybox)Cu](SbF6)2 -catalyzed additions to benzyloxyacetaldehyde, good syn diastereoselectivity is observed. The only

3 Copper Lewis Acids

82

O HO Me Me3CS

39a

CO2Bn

95%, 99% ee O HO

Me3CS

39b

CO2CMe3

Me3CS

i

O HO Me

Pr

CO2Me 39e 36%, 36% ee

CO2Me 39f 97%, 97% ee

Ph

CO2Me 39g 77%, 99% ee

O HO Me

CO2Me

O HO Me

CO2Me

EtS

i

Bu 39j 88%, 93% ee syn:anti 90:10

Pr 39k 80%, 99% ee syn:anti 90:10

Et

39d

Bu CO2Me

94%, 94% ee O HO Me Me

CO2Me 39h 81%, 94% ee HO Me

CO2Me O 39l O 99%, 99% ee syn:anti 5:95

Et

Me

Me O 39m 85% yield syn:anti 93:7 97% ee

Me3CS

i

O

O HO Me EtS

CO2Me

EtS

i

Me 39i 88%, 99% ee syn:anti 97:3

O HO

O HO Me

EtS

O HO Me

39c

CO2Me

84%, 94% ee

91%, 99% ee

Me3CS

Me3CS

O HO Et

O HO Me

O

regioselectivity 98:2

Fig. 3.7

Aldol products derived from enantioselective additions catalyzed by (t-Bu-box) Cu(OTf )2 (5).

exception to this trend is again 2-trimethylsilyloxyfuran, for which anti diastereoselectivity is high (39l). 2,3-Pentanedione also participates in selective aldol reactions with silylketene acetals. In addition to diastereo- and enantioselectivity issues faced in other examples this electrophile contains a subtle regiochemical issue between two nominally similar carbonyl groups. In practice, the (t-Bu-box)Cu(OTf )2 complex performs the subtle discrimination between the two groups and eﬀects a highly regio- and stereoselective aldol reaction with the acetyl group to give 39m. Verdine has described the application of this aldol methodology to the enantio- and diastereocontrolled synthesis of a-hydroxy-a-methyl-b-amino acids (40) in a sequence that uses the carbothioalkoxy group as an amine surrogate via a Curtius rearrangement (Scheme 3.7) [15]. Thus, the desired protected b-amino acid can be obtained in four steps with the needed stereochemical relationships established in the (t-Bu-box)Cu(OTf )2 -catalyzed aldol addition. The asymmetric pyruvate addition can be eﬀected with a complex derived

3.3 Representative Experimental Procedures

Me

83

Me O

O N

O

OTMS

OEt

Me

EtS

O

Me

N Cu Me3C TfO OTf CMe3 5 THF, -78 °C 48 h, 85%

O Me OH EtS

CO2Et Me Me

Me

dr = 10-15:1; >91% ee Me OH

4 steps BocHN

CO2H Me Me 40

Scheme 3.7

Synthesis of a-hydroxy-a-alkyl-b-amino acids from enantioselective pyruvate aldol reactions catalyzed by (t-Bu-box)Cu(OTf )2 (5).

from Cu(OTf )2 and a polystyrene-bound bis(oxazoline) ligand (41) with selectivity approaching that of the solution reaction (Eq. (8)) [16]. As with many solid-supported complexes, catalytic activity was signiﬁcantly less than the soluble variant. Nonetheless, Salvadori and co-workers demonstrated that the ligand could be reused in multiple reaction cycles with no loss of activity provided that additional Cu(OTf )2 was added to the reaction mixture. In the absence of additional Cu(OTf )2 , recycling is still possible, with

O

Me O

O N Me3C O

OTMS

OMe

Me

Me3CS 6

O 37

N

ð8Þ

41 CMe3 12 mol %

O Me OR

Cu(OTf)2 (7 mol %) THF, 0 °C MS 3Å

Me3CS

CO2Me

38 R = H, TMS

7 cycles; reaction time = 1-4 h, ee = 88-93%

84

3 Copper Lewis Acids

the consequence of extended reaction times in subsequent cycles. It is interesting to note that the relative amounts of the silylated and unsilylated aldol products vary from run to run, but the enantioselectivity is relatively constant. A post-catalytic cycle desilylation seems most reasonable. Jørgenson and co-workers have extended the a-keto ester additions to keto malonate substrates (Eq. (9)) [17]. In these asymmetric additions, the tertiary carbinol is not a stereogenic center; in essence the chiral complex induces asymmetry on the nucleophile. For a range of enolsilane nucleophiles, enantiocontrol in the addition step is moderate to excellent. The optimal promoter for these additions is the (Ph-box)Cu(OTf )2 complex (42). In all instances but one the (E)-enolsilane was employed; the (Z)-enolsilane derived from propiophenone gave excellent results (43e). Me

Me O

O N

O

OTMS EtO2C

R

N Cu Ph TfO OTf Ph 42 10 mol % CO2Et

O

O

43a 82% 58% ee O Ph

OH CO2Et CO2Et

OH CO2Et CO2Et Me 43e 95% 90% ee

R'

OH CO2Et CO2Et

OH CO2Et CO2Et

O

OH CO2Et CO2Et

43b 91% 86% ee O

R

Et2O

R'

OH CO2Et CO2Et

O

43c 88% 93% ee O

O

OH CO2Et CO2Et

43d 90% 85% ee

OH CO2Et CO2Et

ð9Þ 43f 80% 60% ee

43g 26% 36% ee

Dalko and Cossy have employed the Danishefsky diene in additions to ethyl pyruvate catalyzed by an uncharacterized complex prepared by mixing enantiopure stilbene diamine 44 and cyclobutanone 45 (1:1), followed by complexation with Cu(OTf )2 (Eq. (10)) [18]. Cyclobutanone was optimal with regard to yield and enantioselectivity for the ketones and aldehydes surveyed. Reactant stoichiometry and premixing time were found to have a signiﬁcant eﬀect on enantioselectivity and reaction eﬃciency. The reaction aﬀords a mixture of both the silylated and desilylated acyclic aldol product

3.3 Representative Experimental Procedures

(48), in addition to the cyclized dihydropyrone (47). Whether the dihydropyrone is formed by a concerted or stepwise mechanism is yet to be determined. In practice, the acyclic aldols are easily cyclized to the dihydropyrone in the presence of triﬂuoroacetic acid for the purpose of determining the enantiomeric excess. This catalyst system is noteworthy for the simplicity with which the active catalyst is assembled (in situ). OMe

O OEt

Me TMSO

O 46

Ph Ph 44

37

NH2

O

NH2 (1:1)

45

OMe O

Cu(OTf)2 (10 mol %), MS 4Å THF, -72 °C

O

(10)

OH(TMS)

CO2Et Me

47 85%, >98% ee

O

CO2Et Me 48

F3CCO2H

Mechanism and Stereochemistry The mechanism of Cu(II)-catalyzed additions to a-keto esters is thought to proceed via a Mukaiyama aldol pathway, with the difunctional electrophile undergoing bidentate activation by the Cu(II) Lewis acid (49). This coordination event lowers the LUMO of the ketone to a point that facilitates addition of the silylketene acetal (49 ! 50). Silylation of the Cu(II) aldolate via an intra- or intermolecular silicon transfer gives the neutral metalcoordinated adduct (52) that decomplexes to regenerate the catalytically active Lewis acid and release the product, 53 (Scheme 3.8) [13]. A distorted square-planar metal center is implicated in all reactions involving (t-Bu-box)CuL n [19]. This is suggested both by X-ray crystallographic studies of the hydrated complex [(t-Bu-box)Cu(OH2 )2 ](SbF6 )2 and by PM3 calculations designed to probe the structure of activated intermediates. The X-ray structure reveals that the coordinated water molecules are tilted out of the ligand plane by approximately 30 (Figure 3.8). This is a steric eﬀect, as water molecules in the corresponding [(i-Pr-box)Cu(OH2 )](SbF6 )2 complex are nearly coplanar with the ligand (approximately 7 out of plane). By inspection, replacing the water molecules with the oxygen atoms of the pyruvate ester should result in a complex in which the enantiotopic faces of the carbonyl are signiﬁcantly diﬀerentiated. This has been con3.3.2.2

85

86

3 Copper Lewis Acids O

Me3CS

O R OTMS O

N

N Cu TfO OTf 5

OMe

53

O R OTMS N O Cu N Me3CS L 52 OMe

L TMS

Me3CS

N

N O R O Cu O 51

OMe 37

R

N O Cu O

49 OMe

N TMS

-L

2+ 2 OTf

N

2+ 2 OTf

2+ 2 OTf

O

R

N O R O Cu O

2+ 2 OTf

OTMS Me3CS 6

Me3CS +L 50 OMe OMe Scheme 3.8 Proposed mechanism for (t-Bu-box)Cu(OTf )2 -catalyzed enantioselective aldol reactions.

ﬁrmed by PM3 calculations. The pyruvate ester additions are all consistent with the stereochemical model shown in Figure 3.9. The bulky t-butyl group eﬀectively shields the re face of the ketone, directing nucleophilic addition to the si face. This complexation mode is now well established with this family of catalysts. The diastereoselectivity in additions of substituted enolsilanes to a-keto esters can be rationalized by an open, antiperiplanar transition structure that minimizes steric interactions between the enolsilane substituent and

Me

2+

Me O

O N

N

2 SbF6

Cu Me3CH O OH CMe3 2 2

Fig. 3.8

X-ray crystal structure of [(t-Bu-box)Cu(H2 O)2 ](SbF6 )2 .

3.3 General Experimental Procedure

Me O

O CMe3 N N Cu O O

Me

H CMe3 OMe

H

R

OTMS SR Fig. 3.9

Model for enantioselective addition to a-keto esters catalyzed by (t-Bu-box)Cu(OTf )2 .

the pendant ligand substituent (Figure 3.10). The disposition of the aOTMS and aSR groups are less important in this model, a point that is supported by the relative insensitivity of reaction diastereoselectivity as a function of enolsilane geometry.

General Experimental Procedure

In an inert atmosphere box, (S,S)-bis(tert-butyloxazoline) (15 mg, 0.050 mmol) and Cu(OTf )2 (18 mg, 0.050 mmol) were placed in an oven-dried

Me

Me

Me

2+ O

O N

N Cu CMe3 Me3C O Me O H Me

Me

N

N Cu CMe3 Me3C O H O Me Me

OMe R'

2+ O

O

R

OMe R' R

R,R' = OTMS, SR

O HO Me Me3CS

CO2Me

O HO Me Me3CS

CO2Me

Me

Me

anti disfavored

syn favored

Fig. 3.10

Stereochemical models for syn-selective aldol reactions catalyzed by (t-Bu-box)Cu(OTf )2 .

87

88

3 Copper Lewis Acids

10-mL round-bottomed ﬂask containing a magnetic stirring bar. The ﬂask was ﬁtted with a serum cap, removed from the inert atmosphere box, and charged with solvent (1.5–3.0 mL). The resulting suspension was stirred rapidly for 4 h with CH2 Cl2 to give a slightly cloudy bright green solution or 1 h with THF to give a clear dark green solution. The catalyst was cooled to 78 C, and the pyruvate (0.50 mmol) was added, followed by the silylketene acetal (0.60 mmol). The resulting solution was stirred at 78 C until the pyruvate was completely consumed (0.5 to 24 h) as determined by TLC (2.5% Et2 OaCH2 Cl2 ). The reaction mixture was then ﬁltered through a 2 cm 4 cm plug of silica gel with Et2 O (60 mL). Concentration of the Et2 O solution gave the crude silyl ether which was dissolved in THF (5 mL) and treated with 1 m HCl (1 mL). After being stirred at room temperature for 1–5 h this solution was poured into a separatory funnel and diluted with Et2 O (20 mL) and H2 O (10 mL). After mixing the aqueous layer was discarded and the ether layer was washed with saturated aqueous NaHCO3 (10 mL) and brine (10 mL). The resulting ether layer was dried over anhydrous Na2 SO4 , ﬁltered, and concentrated to provide the hydroxy esters. Puriﬁcation was achieved by ﬂash chromatography. 3.3.3

Enolsilane Additions to Unfunctionalized Aldehydes

In 1998, Kobayashi made the counterintuitive observation that the Lewis acid-catalyzed addition of enolsilanes to aldehydes could be conducted in wet organic solvents (e.g. 10% H2 O in THF) [20]. The initial study documented that a wide range of metal salts are eﬀective in promoting Mukaiyama aldol reactions in an aqueous environment. It is particularly relevant to this chapter that Cu(ClO4 )2 acts as a catalyst (Eq. (11)), but is not particularly eﬃcient (one turnover in 12 h at ambient temperature). The carbonyl addition pathway is clearly faster than Lewis or Brønsted acidcatalyzed decomposition of the enolsilane.

Ph

OTMS Me

O H

O

Cu(ClO 4)2 (20 mol %) h Ph

OH

Ph H2O/THF (1:9) 25 °C, 12 h

Ph

ð11Þ

Me 47%

This discovery led to the development of an enantioselective variant. Implicit requirements for aqueous enantioselective Mukaiyama aldol reactions include a strong association between the chiral ligand and the metal center that is not disrupted by water, and/or a ligand–metal complex that is considerably more active than the corresponding hydrated complex, M m (OH2 )n Xm . Given the documented activity of Cu(ClO4 )2 in water, attention was directed to bis(oxazoline) ligands, known to have strong aﬃnity for Cu(II). The optimal catalyst with regard to both chemical and optical yield was the (i-Pr-

3.3 General Experimental Procedure

box)Cu(OTf )2 complex; H2 OaEtOH (1:9) was identiﬁed as the best solvent. Under the optimized reaction conditions a range of substituted enolsilanes underwent asymmetric catalyzed addition to aromatic, heteroaromatic, alkenyl, and aliphatic aldehydes with moderate to good enantioselectivity (Eq. (12)) [21]. The absolute stereochemistry of these aldol adducts is unfortunately not known, so speculation about the mechanism of asymmetric induction is premature at this time. Me

Me

N

N

O

O

CHMe2 53 24 mol % Cu(OTf)2 (20 mol %)

Me2HC

1

R

OTMS Me

O H

R

O

2

R

H2O/EtOH (1:9)

R

OH

O

Me 54a 74%, syn/anti = 3.2/1 67% ee (syn) O

OH

O iPr

Me 54b 81%, syn/anti = 3.5/1 81% ee (syn)

Me 54c 95%, syn/anti = 4.0/1 77% ee (syn)

O

O

OH

Et

Et Me 54d 91%, syn/anti = 4.0/1 79% ee (syn) O

Me

Cl 54e 88%, syn/anti = 2.6/1 76% ee (syn) O

iPr

54f 87%, syn/anti = 2.9/1 75% ee (syn) O

OH

O

Me

OH

S Me

54j 78%, syn/anti = 5.7/1 75% ee (syn)

54i 86%, syn/anti = 4.0/1 76% ee (syn)

54h 97%, syn/anti = 4.0/1 81% ee (syn) O

OH

Et

O

Me

54g 77%, syn/anti = 4.6/1 42% ee (syn)

OH

Et

iPr Me

OH OMe

Et Me

OH

OH

Et

OH

ð12Þ

Me

-10 °C, 20 h O

OH

1

2

O

OH

Et Me 54k 56%, syn/anti = 1.6/1 67% ee (syn)

Cl

Me 54l 94%, syn/anti = 2.3/1 57% ee (syn)

Subsequent experiments demonstrated that pure H2 O, rather than mixtures of H2 O and organic solvent could be used as the solvent, either by the

89

90

3 Copper Lewis Acids

use of an additive (Triton-X100, Eq. (13)) or a lipophilic Cu(II) salt in conjunction with a fatty acid additive (Eq. (14)) [22, 23]. Me

Me

N

N

O

O

MeO

OTMS Me

53 Me2HC (20 mol %) CHMe2 Cu(OTf) 2 (20 mol %)

OHC

Triton X-100 (3 mol %) H2O

Me

O

OH

MeO Me Me 54m

0 °C, 24 h

86%, 53% ee

ð13Þ Me

Me O

O

N 53 Me2HC 24 mol % CHMe2 Cu(O 3SOC12H25)2 (20 mol %) N

Me

OTMS Me

OHC

CH3(CH2)9CO2H (10 mol %) H2O

O

OH

Me Me

54b 23 ° C, 20 h 76%, syn/anti = 74/26, 69% ee (syn)

ð14Þ

3.4

Additions Involving In-Situ Enolate Formation

A continuing goal of organic chemists is the development of ‘‘direct’’ reactions in which the compounds undergoing reaction are activated in situ. The Mukaiyama aldol reaction, despite its broad utility, is not an example of a direct reaction, because preformation of an enolsilane in a separate step is a necessary requirement. Direct enolization and subsequent aldol reaction have been achieved in a handful of asymmetric Cu(II)-catalyzed reactions. 3.4.1

Pyruvate Ester Dimerization

Additions to pyruvate esters without pre-activation of the nucleophilic reactant have been explored by Jørgenson and co-workers. Ethyl pyruvate is enantioselectively dimerized in the presence of a chiral Cu(II) Lewis acid and catalytic quantities of a trialkylamine base to aﬀord diethyl 2-hydroxy-2-

3.4 Additions Involving In-Situ Enolate Formation

methyl-4-oxoglutarate, 55 (Eq. (15)) [24]. Formation of the aldol was achieved with good enantiocontrol by use of (t-Bu-box)Cu(OTf )2 as catalyst in conjunction with a dialkylaniline base. Subtle interplay between the identity of the solvent, the counter-anion, and base were observed. The initial aldol adduct cyclizes in the presence of base and TBS-Cl to aﬀord a highly substituted g-lactone 56 in moderate yield and with high enantioselectivity (Eq. (16)). The scope of this reaction beyond use of ethyl pyruvate was not described. Me

Me O

O N

N Cu Me3C TfO OTf CMe3 5 10 mol %

O 2 Me

CO2Et 37

PhNBn2 (10 mol %) Et2O

O Me OH EtO2C

(15)

CO2Et 55

>80% conversion, 93% ee Me

Me O

O N

N Cu 1) Me3C TfO OTf CMe3 5 10 mol %

O 2 Me

CO2Et

PhNMe2 (5 mol %)

37 2) Et3N, TBSCl

O

O

Me

(16)

CO2Et TBSO

56 48%, 96% ee

3.4.2

Addition of Nitromethane to a-Keto Esters

The enantioselective addition of nitroalkanes to carbonyl compounds (Henry reaction) has been documented by Jørgenson and co-workers [25]. With nitromethane as solvent, a combination of the (t-Bu-box)Cu(OTf )2 complex 5 (20 mol%) and Et3 N (20 mol%) eﬀect room-temperature aldol reactions with a range of a-keto esters (Eq. (17)). The reaction is especially enantioselective with alkyl and aromatic groups on the ketone moiety. b,gUnsaturated-a-keto esters react with nitromethane to aﬀord aldol products in good yield, but enantiocontrol is signiﬁcantly lower than for aromatic or aliphatic substrates. Product partitioning is completely selective, however, for the 1,2-mode of addition compared with 1,4-conjugate addition. This selectivity is not observed when the reaction is conducted in the absence of (t-Bu-box)Cu(OTf )2 , leading the authors to propose that the nitronate anion might be coordinated to the metal center during the catalytic reaction. Triethylamine is uniquely suited as catalytic base in this reaction: N-

91

92

3 Copper Lewis Acids

methylmorpholine, dimethylaniline, tribenzylamine, pyridine, ethyl diisopropylamine, and potassium carbonate are all inferior with regard to both yield and enantiocontrol. Enantioselectivity is optimal for the t-Bu-box ligand and triﬂate counter-ion (as compared with hexaﬂuoroantimonate). Me

Me O

O N

N Cu Me3C TfO OTf CMe3 5 20 mol %

O CO2Et

R

Et3N (20 mol %) CH3NO2 25 °C

HO Me

HO Et

O2N

CO2Et 57a 95%, 92% ee

HO R O2N

CO2Et

Ph HO C6H13

HO

O2N

CO2Et 57b 46%, 90% ee

ð17Þ

57

O2N

O2N

CO2Et 57c 47%, 77% ee

CO2Et 57d 91%, 93% ee

Me Me

Me HO

HO O2N

CO2Et

57e 97%, 94% ee

O2N

HO CO2Et

57f 92%, 94% ee

O2N

CO2Et

57g 90%, 94% ee

Cl

HO CO2Et 57i 81%, 86% ee

O2N

CO2Et 57j 91%, 88% ee

CO2Me 57m 95%, 35% ee

OMe

HO

HO

O2N

CO2Et 57k 99%, 93% ee

Ph

O2N

CO2Et 57l 68%, 57% ee

Me

HO O2N

Me CO2Et 57h 99%, 92% ee

O2N

NO2

HO

O2N

HO

HO O2N

CO2Me 57n 95%, 30% ee

HO O2N

CO2Et 57o >96%, 60% ee

Reactant stoichiometry is critical in the catalyzed Henry reaction. With a 20% loading of (t-Bu-box)Cu(OTf )2 catalyst optimum enantiocontrol (92% ee for the reaction of nitromethane with ethyl pyruvate) is realized by use of 20 mol% Et3 N. In contrast, using 15 mol% Et3 N under otherwise identical conditions leads to a product enantiomeric excess of only 56%. Use

3.4 Additions Involving In-Situ Enolate Formation

of 25 mol% Et3 N leads to a product with 73% ee. The diminished enantioselectivity when employing excess base relative to Lewis acid can be accounted for by a racemic pathway – Et3 N alone catalyzes non-selective nitroaldol addition. The reason for reduced selectivity in the presence of excess Lewis acid (relative to base) is less clear. Another distinctive feature of the (t-Bu-box)Cu(OTf )2 -catalyzed Henry reaction is reversed p-facial selectivity relative to that normally obtained with this catalyst (vide supra). Indeed, the usual distorted square-planar arrangement (cf. Figures 3.8 and 3.9) cannot correctly account for the stereochemistry observed (the absolute stereochemistry of 57j was determined by X-ray crystallography). Jørgenson and co-workers instead propose that the ketone carbonyl and the nitronate anion coordinate to the Cu(II) center in the ligand plane, with the carboethoxy group bound in the axial position to complete the square pyramidal complex (Figure 3.11). The Zimmerman– Traxler chair-like transition structure is proposed on the basis of the observation of complete 1,2-selectivity in additions to b,g-unsaturated-a-keto Me

Me O

O

Me

CMe3 N N Cu O OO

H

O

N

O

H

Me3C

CMe3 Me

EtO

Me H O N Cu N O H OO Me CMe3 N O EtO

Me O

Me O

O

Me

H CMe3 N N Cu O CMe3 O R O N H O OEt

Me3C H R

O EtO

R OH O2N

CO 2Et

disfavored Fig. 3.11

Proposed stereochemical model for enantioselective Henry reactions catalyzed by (t-Bu-box)Cu(OTf )2 .

H N

N O Cu

O N O

Me

CMe3

HO R O2N

CO 2Et favored

O

93

94

3 Copper Lewis Acids

esters. One possible explanation of this selectivity is that the electropositive metal center directs addition to the carbonyl, rather than the CbC p bond. The strong preference for one of the two diastereomeric transition structures illustrated is less clear, because both require interaction of one reaction component (either a-keto ester or nitronate) with the bulky tert-butyl group of the ligand. 3.4.3

Malonic Acid Half Thioester Additions to Aldehydes

Shair and coworkers have developed a mild aldol addition based on the decarboxylative Claisen condensation that is the key step in polyketide biosynthesis [26]. Malonic acid half thioesters (MAHT) are employed as enolate equivalents in addition reactions with aldehydes catalyzed by Cu(2-ethylhexanoate)2 and 5-methoxybenzimidazole. This metal salt and additive were optimized after initial screening experiments that identiﬁed Cu(OAc)2 and imidazole as promising leads for the desired transformation. The decarboxylative aldol addition is performed at ambient temperature under ambient atmosphere in wet solvent to aﬀord moderate to excellent yields of the b-hydroxy thioesters (Figure 3.12). The substrate scope is good with regard to the electrophile. A common problem with direct aldol reactions is self-condensation of enolizable aldehydes; the reaction is not observed in this instance, demonstrating the mildness of the reaction. Good levels of diastereocontrol are observed when a-methyl MAHT are used. These diastereomer ratios are kinetic values, not thermodynamic, as judged by resubmission experiments. In the absence of aldehyde, no decarboxylation is observed, suggesting that a Cu(II)-thioacetate enolate is not on the reaction pathway. The authors propose that enolization of the MAHT by the amine base might be required for the reaction to proceed. No incorporation of a second aldehyde into an isolated aldol adduct is observed under the catalyzed reaction conditions – retro-aldol reactions do not occur.

General Experimental Procedure

5-Methoxybenzimidazole (0.11 mmol, 0.22 equiv.) and Cu(2-ethylhexanoate)2 (0.1 mmol, 0.2 equiv.) were added to a stirred solution of malonic acid half benzylthioester (0.5 mmol, 1.0 equiv.) in THF (5 mL; solvent stored in a vial without protection from the air before use) at 23 C. After the reaction became homogeneous (ca. 1 min) aldehyde (0.5 mmol, 1.0 equiv.) was added. The solution was stirred at 23 C for the prescribed time and quenched with 0.5 m HCl solution. The resulting solution was diluted with EtOAc and washed successively with 0.5 m HCl, saturated aq. NaHCO3 , and brine. The organic layer was dried over Na2 SO4 , ﬁltered through cotton,

H

Cu O

Cu O

Et 59

N H (22 mol %) wet THF

N

(20 mol %)

O

N H (22 mol %) wet THF

N

Et 59 (20 mol %)

OMe

2

2

BnS

BnS

O

R

OH

R

OH

Me

60

O

O

CO2Et

Me

Me

70%

81%

97%

85%

Ph 82%

61a R = (CH2)2Ph: 52%, 7:1 syn/anti 82%, 8:1 syn/anti 61b R = CO2Et:

Me

OMe

Me

Direct aldol additions of MAHT catalyzed by Cu(II) complexes.

O R

CO2H

R

CO2H

Me 58b

O

H

O

58a

Fig. 3.12

BnS

BnS

O

O

Ph

Ph OMOM

NO2

74%

22%

82%

65%

3.4 General Experimental Procedure 95

96

3 Copper Lewis Acids

and concentrated under reduced pressure. The product was puriﬁed by continuous gradient ﬂash column chromatography. 3.4.4

Dienolate Additions to Aldehydes Scope and Application Activation of silyl enolates toward aldol additions can be achieved by desilylation. Carreira and coworkers developed this approach in highly enantioselective additions of silyl dienolates to aromatic, heteroaromatic, and a,b-unsaturated aldehydes in the presence of an (S-Tol-BINAP)CuF2 catalyst [27]. The copper ﬂuoride catalyst is generated in situ by treatment of S-Tol-BINAP with Cu(OTf )2 , followed by addition of a crystalline, anhydrous ﬂuoride source, (Bu 4 N)Ph3 SiF2 (TBAT). When as little as 2 mol% is used, (S-Tol-BINAP)CuF2 (63) catalyzes enantioselective addition of a silyl dioxolanone-derived dienolate to a range of aldehydes at 78 C (Eq. (18)). Selectivity ranged from good to excellent (83–95% ee) for all substrates except a-methylcinnamaldehyde, for which the level of enantioinduction was somewhat lower (65% ee). Aliphatic aldehydes are also selective electrophiles, but alkylnals suﬀered from low yields (< 40%). 3.4.4.1

(pTol)2 P CuF2 P (pTol)2

RCHO

Me

Me

O

O

63 2 mol %

+

THF, -78 °C; then CF3CO2H

OSiMe3 62

Me

Me

O

O

OH R

O 64a-j

CHO

CHO

S

CHO

O

CHO

ð18Þ a: 92%, 94% ee

b: 86%, 93% ee

c: 98%, 95% ee

CHO

CHO

CHO OMe

MeO e: 93%, 94% ee

Me

d: 91%, 94% ee

f: 83%, 85% ee

Me

CHO

h: 48%, 91% ee

Me

CHO

i: 81%, 83% ee

g: 82%, 90% ee

Ph

CHO Me

j: 74%, 65% ee

3.4 General Experimental Procedure

(S-Tol-BINAP)CuF2 -catalyzed addition of silyl dienolates to aldehydes has found synthetic utility in the construction of the polyol subunit of amphotericin B (Scheme 3.9) [28]. Both key fragments of the polyol chain (C1 aC13 ) were derived from the same aldol reaction of furfural and the trimethylsilyl dienolate, diﬀering only in the antipode of (Tol-BINAP)CuF2 catalyst used. It is noteworthy that the furfural aldol adduct can be obtained in >99% ee after a single recrystallization of the aldol product. Chiral copper enolate methodology has also been employed for the total synthesis of leucascandrolide A [29]. Crotonaldehyde reacts with the aforementioned trimethylsilyl dienolate in the presence of 2 mol% (R-TolBINAP)CuF2 to aﬀord the allyl alcohol adduct in 91% ee and 42% yield (Scheme 3.10). Yields were hampered in this instance because of crotonaldehyde polymerization, as noted by the authors. Further elaboration of this aldol product furnished a highly convergent synthesis of leucascandrolide A. Mechanistic Considerations [30] The hard ﬂuoride anion is an eﬀective desilylating agent. When mismatched with a soft Cu(II) cation the ﬂuoride anion is designed to serve as an in-situ means of enolate formation. This is achieved by desilylation of the enolsilane, then by metalation. Supporting evidence for copper enolate formation was obtained by independent synthesis – when silyldienolate 62 is subjected to MeLi (10 mol%) followed by (S-BINAP)Cu(OTf )2 and benzaldehyde the expected aldol adduct is obtained in good yield and enantioselectivity. Similar results are obtained by employing (Bu 4 N)Ph3 SiF2 as the desilylating reagent. Mechanistic studies by Carreira and coworkers have concluded that the catalytically active species is the Cu(I) dienolate depicted in Scheme 3.11. Cu(II) complexes are known to undergo a one-electron reduction in the presence of enolsilanes. Thus desilylation of the silyl enolate via the hard ﬂuoride anion of the copper complex followed by copper metalation gives the putative Cu(I) dienolate 66. The same (S-Tol BINAP)Cu(dienolate) has been observed independently via desilylation of the trimethylsilyl dienolate by (Bu 4 N)Ph3 SiF2 , followed by treatment with (S-Tol-BINAP)Cu(ClO4 ). Formation of the (S-Tol BINAP)Cu(dienolate) species and its disappearance after addition of benzaldehyde was closely monitored by IR spectroscopy (ReactIR). Regeneration of the catalytically active species is achieved by desilylation of another molecule of silyldienolate by the resulting Cu(I) alkoxide complex (68 ! 69 þ 66). Reactivity is also observed in the presence of catalytic quantities of (S-Tol-BINAP)Cu(O t Bu), with identical yields and selectivity as with the corresponding Cu(I) or Cu(II) ﬂuoride catalysts, providing corroborating evidence for the proposed mechanism. 3.4.4.2

General Procedure. A mixture of Cu(OTf )2 (3.6 mg, 0.010 mmol, 2 mol%) and (S)-Tol-BINAP (7.5 mg, 0.011 mmol, 2.2 mol%) in 2 mL THF was stirred at 23 C in an inert gas atmosphere for 10 min to yield a clear yellow

97

Me

HO

Me

O

Me

THF, -78 °C

ent-63 2 mol%

Application of (S-Tol-BINAP)CuF2 -catalyzed additions of silyl dienolates to the polyol subunits of amphotericin B.

*

O

Me

O

Me BuPh2SiO

Scheme 3.9

t

O

*

OH

ent-64d

O

O O

Me

Me

1

O

(pTol)2 P CuF2 P (pTol)2

O

+

OSiMe3

* OH

THF, -78 °C

63 2 mol%

OH

13

OH

CO2H

OMycosamine

OH OH O

CHO

O

Me

amphotericin B

*

OH OH

O

Me

(pTol)2 P CuF2 P (pTol)2

H O

*

O

Me

O

Me

O

Me

O

Me t OSi BuMe2

O 64d 95%, > 99% ee *

OH O

98

3 Copper Lewis Acids

3.4 General Experimental Procedure

Me

Me

O

O

(pTol) 2 P CuF 2 P (pTol) 2 OTMS

62 +

OH O

Me O

*

Me

THF, -78 °C

O Me

Me ent-63 2 mol%

O

ent-64h 42%, 91% ee

H Me *

O

OM O

O

O

O

N O

HN

O

MeO

Me

O Me leucascandrolide A

Scheme 3.10

Application of an (S-Tol-BINAP)CuF2 catalyzed addition of a silyl dienolate to the leucascandrolide A.

solution. A solution of Ph3 SiF2 (Bu 4 N) (10.8 mg, 0.02 mmol, 4 mol%) in 0.5 mL THF was added via a cannula and stirring was continued for 10 min. The mixture was cooled to 78 C and the dioxenone-derived dienolate (0.16 mL, 0.75 mmol) was added dropwise, followed by a solution of the aldehyde (0.50 mmol) in 0.5 mL THF. The progress of the reaction was monitored by TLC (reaction times 0.5–8 h). On completion triﬂuoroacetic acid (0.2 mL) was added at 78 C and the solution was left to warm to 23 C. Stirring was continued an another hour. The reaction mixture was diluted with ether (5 mL) and a saturated aqueous solution of NaHCO3 was added dropwise until evolution of gas ceased. The organic layer was washed with brine (3 mL), dried over Na2 SO4 , and concentrated in vacuo. Puriﬁcation of the crude material by chromatography on silica gel with 3:1 ether– hexanes aﬀorded the aldol adduct. The enantiomeric excess of the alcohol products was determined by HPLC analysis using a racemic sample as reference. 3.4.5

Enantioselective Cu(II) Enolate-Catalyzed Vinylogous Aldol Reactions

The traditional challenge associated with vinylogous aldol reactions is competition between reactivity at the a or g positions of the vinyl enolate. For-

99

OSiMe3

-FSiMe3

62

O

O

Proposed catalytic cycle for (S-TolBINAP)CuF2 -catalyzed additions of silyl dienolates to aldehydes.

(pTol) 2 P Cu(OTf) 2 P (pTol) 2

Ph3SiF 2(NBu4)

Scheme 3.11

65

63

(pTol) 2 P CuF2 P (pTol) 2

Me

Me

O

Me

OCu

R

Me3SiO

66

O

Me

RCHO

69

O

Me

P

P

62

O

O O

R

Cu O

O

Me

68

OSiMe3

P

P

Me

Me

Me

O H 67

Me O O

Me

O

P R Cu O P

O

Me

O

100

3 Copper Lewis Acids

3.5 Conclusions

mation of the a-aldolate product has been suppressed by employing bulky Lewis acids such as aluminum tris(2,6-diphenyl)phenoxide (ATPH). Campagne and Bluet also discouraged a-aldolate formation and rendered the vinylogous aldol enantioselective by use of Carreira’s catalyst system (Eq. (19)) [31]. In the presence of 10 mol% (S-Tol-BINAP)CuF2 (63), the authors observed only the g-aldol products 70 in moderate enantioselectivity at ambient temperature. The authors propose that the catalytic cycle is analogous to that of the Carriera system, although no mechanistic studies of this system have yet been reported.

(pTol)2 P CuF2 P (pTol)2 63 10 mol%

OSiEt3 RCHO

+

OEt

OH R

THF, RT

OEt

Me OH

70a-d Me OH

O

O OEt

OEt 70a

O

Me

70b

80%, 70% ee

Me

OH

O

OH

O

OEt 70c

ð19Þ

70%, 48% ee

Me

35%, 56% ee

OEt 70d Me 68%, 77% ee

3.5

Conclusions

Copper complexes enable mechanistically diverse and synthetically useful approaches to the synthesis of b-hydroxy ketones. The Cu(II)-catalyzed asymmetric Mukaiyama aldol reaction developed by Evans provides facile stereocontrolled access to a range of aldol adducts derived from chelating electrophiles. Subsequent extension of this system in the context of ‘‘green’’ chemistry, described by Kobayashi, enabled access to aldols from unfunctionalized aldehydes. Recent eﬀorts have focused on the use of copper complexes to eﬀect direct aldol unions by way of in-situ enolization. Under this general mechanistic umbrella, reports of addition of nitromethane to pyr-

101

102

3 Copper Lewis Acids

uvate esters, malonic half thioester additions to aldehydes, and desilylative dienolate additions have been described. It is both remarkable and exciting that these mechanistically dissimilar reactions are all catalyzed by the same metal. Given that nearly all of the examples from this chapter were reported after 1995 it is reasonable to expect continued interest and development in these fundamental bond constructions.

References 1 M. Iwata, S. Emoto, Chem. Lett. 1974, 959–960. 2 K. Irie, K.-i. Watanabe, Chem. Lett. 1978, 539–540. 3 Y. Ito, T. Matsuura, T. Saegusa, Tetrahedron Lett. 1985, 26,

5781–5784. 4 D. A. Evans, J. A. Murry, M. C. Kozlowski, J. Am. Chem.

Soc. 1996, 118, 5814–5815. 5 D. A. Evans, M. C. Kozlowski, J. A. Murry, C. S. Burgey,

6 7

8 9 10 11 12 13 14

15 16 17 18 19 20

K. R. Campos, B. T. Connell, R. J. Staples, J. Am. Chem. Soc. 1999, 121, 669–685. D. A. Evans, D. M. Fitch, T. E. Smith, V. J. Cee, J. Am. Chem. Soc. 2000, 122, 10033–10046. D. A. Evans, P. H. Carter, E. M. Carreira, A. B. Charette, J. A. Prunet, M. Lautens, J. Am. Chem. Soc. 1999, 121, 7540– 7552. D. A. Evans, E. Hu, J. D. Burch, G. Jaeschke, J. Am. Chem. Soc. 2002, 124, 5654–5655. H. Matsunaga, Y. Yamada, T. Ide, T. Ishizuka, T. Kunieda, Tetrahedron: Asymmetry 1999, 10, 3095–3098. T. K. Hollis, B. Bosnich, J. Am. Chem. Soc. 1995, 117, 4570– 4581. C. Girard, H. B. Kagan, Angew. Chem. Int. Ed. 1998, 37, 2923–2959. D. A. Evans, M. C. Kozlowski, C. S. Burgey, D. W. C. MacMillan, J. Am. Chem. Soc. 1997, 119, 7893–7894. D. A. Evans, C. S. Burgey, M. C. Kozlowski, S. W. Tregay, J. Am. Chem. Soc. 1999, 121, 686–699. For recent modiﬁcations to the bis(oxazoline) ligand and application to the pyruvate addition, see: H. L. van Lingen, J. K. W. van de Mortel, K. F. W. Hekking, F. L. van Delft, T. Sonke, F. P. J. T. Rutjes, Eur. J. Org. Chem. 2003, 317–324. R. Roers, G. L. Verdine, Tetrahedron Lett. 2001, 42, 3563– 3565. S. Orlandi, A. Mandoli, D. Pini, P. Salvadori, Angew. Chem. Int. Ed. 2001, 40, 2519–2521. F. Reichel, X. M. Fang, S. L. Yao, M. Ricci, K. A. Jorgensen, Chem. Commun. 1999, 1505–1506. P. I. Dalko, L. Moisan, J. Cossy, Angew. Chem. Int. Ed. 2002, 41, 625–628. J. S. Johnson, D. A. Evans, Acc. Chem. Res. 2000, 33, 325– 335. S. Kobayashi, S. Nagayama, T. Busujima, J. Am. Chem. Soc. 1998, 120, 8287–8288.

References 21 S. Kobayashi, S. Nagayama, T. Busujima, Tetrahedron 1999,

55, 8739–8746. 22 K. Manabe, S. Kobayashi, Chem. Eur. J. 2002, 8, 4095–4101. 23 S. Kobayashi, K. Manabe, Acc. Chem. Res. 2002, 35, 209–217. 24 K. Juhl, N. Gathergood, K. A. Jorgensen, Chem. Commun.

2000, 2211–2212. 25 C. Christensen, K. Juhl, R. G. Hazell, K. A. Jorgensen, J.

Org. Chem. 2002, 67, 4875–4881. 26 G. Lalic, A. D. Aloise, M. S. Shair, J. Am. Chem. Soc. 2003,

125, 2852–2853. 27 J. Krueger, E. M. Carreira, J. Am. Chem. Soc. 1998, 120,

837–838. 28 J. Kruger, E. M. Carreira, Tetrahedron Lett. 1998, 39, 7013–

7016. 29 A. Fettes, E. M. Carreira, Angew. Chem. Int. Ed. 2002, 41,

4098–4101. 30 B. L. Pagenkopf, J. Kruger, A. Stojanovic, E. M. Carreira,

Angew. Chem. Int. Ed. 1998, 37, 3124–3126. 31 G. Bluet, J. M. Campagne, J. Org. Chem. 2001, 66, 4293–

4298.

103

105

4

Tin-promoted Aldol Reactions and their Application to Total Syntheses of Natural Products Isamu Shiina 4.1

Introduction

Stereoselective aldol reactions are frequently used for synthesis of complicated natural and unnatural oxygenated products, because b-hydroxy carbonyl groups are now easily prepared by several eﬀective methods. Among the three stable valences of tin, the stannic and stannous species are generally used for eﬀective formation of the desired aldol adducts from two starting materials. These tin-promoted reactions are divided into two types according to the principles: (i) directed Mukaiyama aldol reaction of silyl enolates with carbonyl compounds promoted by Sn(IV) or Sn(II) Lewis acids, and (ii) the crossed aldol addition of C- or O-enolates with Sn(IV) or Sn(II) to other carbonyl components. This review ﬁrst covers Sn(IV)mediated aldol reactions of enol silyl ethers (ESE) or ketene silyl acetals (KSA) with carbonyl compounds or acetals, which have been developed as powerful tools for stereoselective synthesis of b-hydroxy or b-alkoxy carbonyl groups. Chiral diamine–Sn(II) complex-promoted aldol and related addition reactions for preparation of a variety of optically active polyoxy compounds will be the second subject discussed. Finally, recent applications of the reactions to highly enantioselective syntheses of optically active natural products will be described.

4.2

Tin-promoted Intermolecular Aldol Reactions 4.2.1

Achiral Aldol Reactions

In 1973, Mukaiyama and Narasaka developed an acid-catalyzed aldol reaction of silyl enolates with electrophiles and revealed that Lewis acids such as TiCl 4 , SnCl 4 , AlCl3 , BF3 OEt2 , and ZnCl2 promoted the reaction quite Modern Aldol Reactions. Vol. 2: Metal Catalysis. Edited by Rainer Mahrwald Copyright 8 2004 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 3-527-30714-1

106

4 Tin-promoted Aldol Reactions and their Application to Total Syntheses of Natural Products

eﬀectively, aﬀording a variety of b-hydroxy ketones from ESE and carbonyl compounds (Eq. (1)) [1]. The synthetic capacity of KSA in the new aldol reaction was also reported in 1975, and the corresponding b-hydroxy- and b-siloxycarboxylic esters were obtained in good combined yields by use of TiCl 4 (Eq. (2)). OTMS

O + Ph

Ph

H

(1)

CH2Cl 2, -78 °C 83% OTMS

O + R

OH O

SnCl 4

H

R1

OR' O

TiCl 4

OR3

CH2Cl 2, -78 °C 84~99%

R2

OR3

R

(2)

R1 R2 R' = H or TMS

Although it is mentioned in their reports that TiCl 4 seems to be superior to other Lewis acids with regard to yield, SnCl 4 was also a popular reagent because of its mild activity and good chelation ability. For example, Wissner applied the SnCl 4 -mediated aldol reaction of tris(trimethylsiloxy)ethene with several aldehydes to the synthesis of a,b-dihydroxycarboxylic acids (Eq. (3)) [2] and Ricci and Taddai prepared a bicyclic g-lactone in good yield by aldol addition of 2,5-disiloxyfuran to two molar amounts of acetone using SnCl 4 (Eq. (4)) [3].

O R

2

OTMS

+

TMSO

H

O

+

TMSO

OTMS

O

OTMS

OH O

SnCl4 R

OH

58~82%

SnCl 4

(3)

OH O

O

(4) CH2Cl 2, -78 °C 70%

O

O

In 1983, Kuwajima and Nakamura reported a novel method for generation of a-stannylketones from ESE and SnCl 4 and studied the properties of the new metallic species in the reaction with carbonyl compounds giving aldol adducts (Eq. (5)) [4]. This facile method for preparing trichlorostannyl enolates was successfully employed in the regioselective synthesis of aldols, as shown in Eq. (6) [5]. Interestingly, syn selectivity was observed in this alternative method, in contrast with the anti selectivity obtained in the direct SnCl 4 -promoted aldol reaction of ESE with the electrophiles (Eqs. (5) and (7)) [1]. Therefore, dif-

4.2 Tin-promoted Intermolecular Aldol Reactions

O OTMS

O

SnCl 4

Ph

CI3Sn

α-stannylketone O + Ph

ð5Þ

Ph CH2Cl 2, -70 °C 80%

CH 2Cl 2, 20 °C

O

OH O

H

OH O

SnCl 4

TMS

syn/anti =93/7

Ph

H

CH2 Cl2 -78 to -50 ˚C 81%

Ph

(6)

Ph

ferent mechanisms were proposed for these reactions, and it was assumed that the silyl nucleophiles could directly attack the carbonyl compounds activated by the Lewis acid at low temperature. OTMS

SnCl4

O Ph

H

O

OH O

SnCl 4

(7) Ph

CH2Cl 2, -78 °C

Ph

H

activated aldehyde

CH2Cl 2, -78 °C 83%

syn/anti = 24/76

Structural features and reactivity of the Sn(IV) C- or O-enolates have been investigated [6, 7]. Yamamoto and Stille independently studied the aldol reaction of stannyl enolates derived from ketones with aldehydes, and showed that stereoselectivity depended on the substituents on the tin and the reaction temperature (Eqs. (8) and (9)) [8, 9]. O OLi

Ph3SnCl

OSnPh3

Ph

OH O

H Ph

rt Sn(IV) enolate

THF -70 °C 80%

(8)

syn/anti = 71/29

O OAc

Bu3SnOMe

OSnBu3

rt isolated Sn(IV) enolate

Ph

H

OH O

Ph ð9Þ THF -78 °C; 78% -78 °C; syn/anti = 20/80 45 °C; 86% 45 °C; syn/anti = 77/23

107

108

4 Tin-promoted Aldol Reactions and their Application to Total Syntheses of Natural Products

Mukaiyama and Iwasawa developed a facile method for the generation of Sn(II) enolates in situ from the corresponding carbonyl compounds with Sn(OTf )2 and a tertiary amine (Eqs. (10) and (11)) [10], and excellent syn selectivity of the aldol reaction was observed in the course of their studies of Sn(II) enolate chemistry (described in a later section). O O

OSnOTf

Sn(OTf) 2

Ph

OH O

H Ph

-78 °C 41%

NEt Sn(II) enolate

syn/anti =>95/5

CH2 Cl2, -78 °C

S

O N

(10) O

OTf Sn S O

Sn(OTf) 2

Ph

S N

NEt CH2Cl 2, -78 °C

S

-78 °C 94%

S

OH O

H Ph

N

S

syn/anti =97/3

Sn(II) enolate

ð11Þ

4.2.2

The Reaction of Silyl Enolates with Aldehydes or Ketones

Diastereoselective addition of ESE and KSA to aldehydes using SnCl 4 were systematically studied by Heathcock, Reetz, and Gennari, who produced a variety of synthetic intermediates. As shown in Eqs. (12)–(15), Heathcock and Reetz independently examined the stereoselectivity of the Mukaiyama aldol reaction of ESE with many kinds of aldehyde, promoted by SnCl 4 [11, 12]. Their results can be summarized: 1. good 2,3-anti or 2,3-syn asymmetric induction was observed in the reaction between achiral simple or a-heteroatom-substituted aliphatic aldehydes and ESE derived from ethyl ketones (Eqs. (12) and (13)); OTMS

O + R

Bu

OTMS

O BnO

t

H

Ph

(12)

syn/anti = 5/>95 OH O

SnCl 4 BnO CH2Cl 2, -78 °C >95% (conversion)

t Bu

R

CH2Cl 2, -78 °C 60~72%

+ H

OH O

SnCl 4

Ph

syn/anti = >95/5 [TiCl4 ; 90/10]

(13)

4.2 Tin-promoted Intermolecular Aldol Reactions

2. high 3,4-syn asymmetric induction was observed in the reaction between a-heteroatom-substituted aliphatic aldehydes and ESE derived from methyl ketones (Eq. (14)); and O

OH O

SnCl 4

OTMS +

H

Ph

OBn

Ph CH2Cl 2, -78 °C 68%

(14)

OBn syn/anti =>99/1

3. good 2,3-syn and high 3,4-syn asymmetric induction was observed in the reaction between a-heteroatom-substituted aliphatic aldehydes and ESE derived from ethyl ketones (Eq. (15)). O

OH O

SnCl 4

OTMS +

H

Ph

OBn

Ph CH2Cl 2, -78 °C 85%

BnO 2,3-syn/anti = 95/5 3,4-syn exclusively

(15)

The observed excellent 3,4-syn selectivity for the a-branched aldehyde was explained by the formation of a Lewis acid–aldehyde complex (so-called chelation model, Scheme 4.1). Good stereoselectivity was not achieved, however, when KSA was employed, even in the reaction with a-benzyloxypropionaldehyde, except when tetrasubstituted KSA were used (Eqs. (16)– (18)).

M O BnO

Nu-TMS

OH

β-face attack

H H Chelation Model

Nu BnO 3,4-syn selection

Me

M BnO Me

O

H R'

H TMSO

R' BnO R 2,3-syn-3,4-syn selection

OTMS

M O BnO R Me

OH O

favorable R H

OH O

unfavorable H R'

R' BnO R 2,3-anti-3,4-syn selection

Scheme 4.1

Chelation model for producing 3,4-syn aldols.

109

110

4 Tin-promoted Aldol Reactions and their Application to Total Syntheses of Natural Products

O H

O t Bu

OBn

O

CH2Cl 2, -78 °C 65%

OTMS +

H

CH2Cl 2, -78 °C >85%

OTMS +

H

syn/anti = 65/35

OH O

CH2Cl 2, -78 °C 93%

(17)

BnO syn/anti = >97/3

OH O

SnCl 4

OTMS

OBn

(16)

OBn

OMe

OMe

MeO

O t Bu

SnCl 4

OBn

O

OH O

SnCl 4

OTBS

+

OH OMe

BnO

(18)

2,3-syn/anti = 5/95 3,4-syn/anti =>99/1

Gennari further studied the Mukaiyama aldol addition of KSA to aldehydes and found that S- t Bu propanethioate or ethanethioate is a quite suitable precursor of the required KSA for stereoselective reactions [13]. Although the reaction of KSA derived from S- t Bu propanethioate with simple achiral aliphatic aldehydes gave poor 2,3-diastereoselectivity (Eq. (19)), it reacted with a-alkoxy aldehydes highly stereoselectively to aﬀord 2,3-syn-3,4syn isomers as shown in Eq. (20). O

OTMS +

H

St Bu

SnCl 4

OH O

S t Bu CH2Cl 2, -78 °C E/Z=93/7; 86% E/Z=93/7; syn/anti = 42/58 E/Z=10/90; 80% E/Z=10/90; syn/anti =42/ 58

ð19Þ O

OTBS H

OBn

SnCl 4

+ St Bu

CH2Cl 2, -78 °C

OH O S t Bu

OBn E/Z = >95/5; 89% E/Z=>95/5; 2,3-syn/anti = 97/3 E/Z = 5/>95; 90% 3,4-syn exclusively E/Z=5/>95; 2,3-syn/anti = 76/24 3,4-syn exclusively

ð20Þ Furthermore, KSA derived from S- t Bu ethanethioate also gave a 3,4-syn adduct preferentially in good yield (Eq. (21)). The highly 3,4-syn asymmetric induction by chelation control using SnCl 4 is quite eﬀective for the con-

4.2 Tin-promoted Intermolecular Aldol Reactions

111

struction of natural complex molecules (as is described at the end of this section). O H OBn

St Bu

OH O

SnCl4

OTBS +

CH2Cl 2, -80 °C 70%

St Bu

(21)

OBn syn/anti =>98/2

It is worthy of note that a-stannylthioesters, instantly generated from KSA by treatment with SnCl 4 , also react with aldehydes to aﬀord the corresponding aldol adducts, but the stereoselectivity of this reaction is sometimes very diﬀerent from that in the reaction involving KSA and an SnCl 4 – aldehyde complex (compare Eqs. (19) and (22), and Eqs. (20) and (23)). The order of addition of KSA and aldehydes to a solution of Lewis acid catalysts such as SnCl 4 should therefore be selected carefully in accordance with the stereoselectivity desired. O H

OTMS

O

SnCl4

S tBu CH2Cl 2, -78 °C

CH2 Cl2 -78 to 0 °C 45%

CI3Sn

St Bu

OH O St Bu syn/anti =>98/2

ð22Þ O H

α-stannylthioester OBn

OH O St Bu

CH2 Cl2 OBn -78 to -20 °C 2,3-syn/anti =10/90 75% 3,4-syn exclusively

(23) Further examples of the construction of multi-functional b-hydroxy carbonyl compounds using SnCl 4 are given in Eqs. (24)–(28). Although sulfursubstituted ESE derived from methylthioacetone reacted with a-alkoxy aldehyde to give a 3,4-syn adduct preferentially (Eq. (24)) [14], the reaction of KSA prepared from methyl a-methylthiopropionate aﬀorded an almost equimolar mixture of the corresponding 3,4-syn and anti diol groups (Eq. (25)) [15]. A 2,3-anti-3,4-syn a-amino carboxylic ester was stereoselectively prepared by reaction of amino-substituted KSA with a-benzyloxypropionaldehyde (Eq. (26)), and the adduct was converted to the corresponding g-lactone which is known as a synthetic intermediate of l-daunosamine and l-vancosamine [16].

112

4 Tin-promoted Aldol Reactions and their Application to Total Syntheses of Natural Products

O

OTMS H

OH O

SnCl4

+

OBn

CH2Cl 2, -78 °C OBn SMe E/Z =80/20; 58% E/Z =17/83; 53% E/Z=80/20; 3,4-syn/anti =80/20 E/Z=17/83; 3,4-syn/anti =82/18

SMe

(24) O

OTMS H

+ MeS

OBn

OMe

OH O

SnCl4 CH2Cl 2, -78 °C

OMe SMe

BnO

E/Z=25/75

(25)

OH O

i) NaIO 4

OMe

ii) ∆

BnO 65%, syn/anti =52/48

O

OTMS H

OBn

+

Bn 2N

OtBu

OH O

SnCl4

OtBu CH2 Cl2 -78 to -40 °C 50%

OBn NBn2 2,3-syn/anti =16/84 3,4-syn/anti =>98/2 HO

NHZ O

ð26Þ

O

a known intermediate in the synthesis of aminosugars NH2

NH2

HO

HO O

OH

L-Daunosamine

O

OH

L-Vancosamine

It is notable that the reaction of KSA derived from ethyl acetate with an a-amino aldehyde in the presence of SnCl 4 gave a 3,4-syn amino alcohol under chelation-control conditions (Eq. (27)) [17] whereas the opposite diastereoselectivity was observed when an a-phenylthio aldehyde was used as the electrophile (Eq. (28)) [18] and a 3,4-syn aldol adduct was obtained when TiCl 4 was employed for the latter reaction instead of SnCl 4 . The 3,4-anti selectivity in the reaction promoted by SnCl 4 was therefore explained, as an exception, by the non-chelation model.

4.2 Tin-promoted Intermolecular Aldol Reactions

O

OH O

SnCl4

OTMS +

H NHBoc

OEt

OEt

CH2Cl 2, -78 °C 60%

NHBoc syn/anti =91/9

ð27Þ O

OH O

SnCl4

OTMS +

H

OMe

CH2Cl 2, -78 °C 78%

SPh

OMe PhS syn/anti = 2/>98 (non chelation) [TiCl 4; 86%, 80/20 (chelation)]

ð28Þ The SnCl 4 -promoted diastereoselective aldol reaction has been applied to the synthesis of some parts of complex molecules such as oligopeptides, oligosugars, and polyoxyamides (Eqs. (29)–(36)). Joullie´ obtained a 2,3-syn3,4-syn thioester by reaction of KSA derived from S- t Bu propanethioate with an a-alkoxy aldehyde, as shown in Eq. (29) [19]. The prepared intermediate was successfully converted to the macrocyclic peptides didemnin A, B, and C in 1990. O H

+

OH O

SnCl4

OTBS

St Bu

St Bu CH2Cl 2, -78 °C 74%

OBn

OBn exclusively

O Me N

N

ð29Þ PMP O

O O

O

NH

R=

O

Me N

OH

O O O

O

HO

O

NHR NH

Didemnin C

Danishefsky employed a trisubstituted butadiene for reaction with an aldehyde connected to a ribonucleoside, with promotion by SnCl 4 (Eq. (30)); subsequent desilylation of the adduct aﬀorded the corresponding dihydropyran which was transformed to tunicamycins [20]. Cox and Gallagher employed a cyclic ESE as a nucleophile for reaction with ribosyl aldehyde, as depicted in Eq. (31); the aldol isomers formed could be used as precursors of tetracyclic hemiketals [21]. Akiyama and

113

114

4 Tin-promoted Aldol Reactions and their Application to Total Syntheses of Natural Products

OTBS BzO O

PMB N O N

MOMO O

O

PMB N O

OMe O

O SnCl 4

H

MOM TBS O O

O

N

OBz

CH2Cl 2, -78 °C

O

O

O

PMB N O

O HF

OMe

O

N

MeCN, 0 °C 61% (2 steps)

MOMO O

O

O

Bz O

5 4

O

O

with another 4,5-cis isomer

O

H N

O N HO

HO

OH

H O

O

O

H N

R O

O

NHAc

O

OH HO

OH OH

Tunicamycins, R = unsaturated alkyl chains

ð30Þ Ozaki developed a new chiral auxiliary group in the diastereoselective synthesis of optically active aldols with a tertiary hydroxyl group (Eq. (32)) [22]. For example, (R)-dimethyl citramalate was synthesized from an adduct produced by the SnCl 4 -accelerated aldol reaction of KSA derived from ethyl acetate with the chiral pyruvate. O O

O O

OTBS H

OBn

OH O SnCl4

O

O

+ O

CH2Cl 2, -78 °C 35%

O OBn syn/anti = 50/50 O

(31) H

H 2, Pd/C 50%

H O

O

H

O

O O

H

O

OH

4.2 Tin-promoted Intermolecular Aldol Reactions

115

OTBS O

OTBS O

O

OEt SnCl 4

O

O

OTBS HO O

O

OEt

O

O

O

O

CH2Cl 2, -20 °C 80%

O

O

O

>98% de O

HO MeO

OMe

O (R)-Dimethyl citramalate

ð32Þ Mukai and Hanaoka reported the formal synthesis of AI-77B in which they successfully used the stereoselective direct aldol reaction of KSA derived from S- t Bu ethanethioate with an a-alkoxy aldehyde (Eq. (33)) [23].

TBSO

O

Ph

OTBS H

OBn

+

S t Bu

SnCl 4

TBSO

OH O St Bu

Ph CH2Cl 2, -78 °C 53% (2 steps, from the corresponding thiol ester)

OBn

O O

OH NH2 O

H N

OH O

OH

AI-77B

ð33Þ

They also used an a-trichlorostannyl thioester generated from KSA with SnCl 4 in a reaction with an a-alkoxy aldehyde to produce a 2,3-anti-3,4-syn aldol group which was employed as an intermediate in the total synthesis of bengamide E, as shown in Eq. (34) [24]. Recently, Boeckman re-applied this methodology for the practical preparation of bengamide B, E, and Z (Eq. (35)) [25]. Mukai and Hanaoka also used SnCl 4 as catalyst in the reaction of the KSA derived from S- t Bu ethanethioate with an a,b-dibenzyloxy aldehyde possessing a Co complex part, to produce a new synthetic intermediate of

116

4 Tin-promoted Aldol Reactions and their Application to Total Syntheses of Natural Products

OTMS MeO

TBSO

O

SnCl 4 Cl 3Sn

St Bu CH Cl , -78 °C 2 2

St Bu

O H

+ OBn

OMe

E/Z=75/25 TBSO

OH O StBu

CH 2Cl2, -78 to 0 °C 73% (2 steps, from the corresponding thiol ester)

OBn OMe 2,3-syn/anti =8/92 3,4-syn exclusively OH OH O

OH OMe

NH

N H

O

ð34Þ

Bengamide E

OTMS MeO

SnCl 4

SPh CH Cl , -78 °C 2 2

TBSO

O Cl 3Sn

SPh

O

+

H OR

OMe TBSO

OH O

SPh CH2Cl 2, -78 °C OR OMe R = Bn; 73% R = Bn; 2,3-syn/anti =8/92 R = 2-naphthylmethyl; 73% R = 2-naphthylmethyl; 2,3-syn/anti =11/89 R2 OH OH O

OH OMe

N H

Bengamide B, E, Z

NR1 O

ð35Þ

bengamide E under direct Mukaiyama aldol-reaction conditions (Eq. (36)) [26]. In contrast, non-metalated a,b-dibenzyloxy aldehyde stereorandomly reacted with the same KSA to give a mixture of isomers, and a similar result was also observed in Liu’s recent research (Eq. (37)) [27].

4.2 Tin-promoted Intermolecular Aldol Reactions

BnO

BnO

O + MeO

OH O

SnCl 4

OTMS

H OBn Co(CO)3 Co (CO) 3

St Bu

E/Z =75/25

117

St Bu OBn OMe Co(CO)3 Co (CO) 3

CH2 Cl 2 -78 to 0 °C

BnO

CAN

OH O St Bu

MeOH, 0 °C 47% (3 steps, from the corresponding alcohol)

OBn OMe exclusively

OH OH O NH

N H

O

OH OMe Bengamide E

ð36Þ BnO

O H + MeO

SPh

OBn

BnO

SnCl 4

OTMS

OH O SPh

any Lewis Acid

OBn OMe no selectivity

ð37Þ 4.2.3

The Reaction of Silyl Enolates with Acetals

Mukaiyama reported that acetals are also activated by Lewis-acid catalysts and are eﬀectively coupled with nucleophilic ESE and KSA to give the desired b-alkoxy carbonyl compounds in high yields (Eq. (38)) [28]. Kuwajima showed that trichlorostannyl C- or O-enolates could be used as nucleophiles in reactions with acetals and aldehydes (Eq. (39)) [5]. OTMS

OEt + Ph

CH2Cl 2, -78 °C 95% O

CI

(38)

Ph

OEt

OMe

OEt O

TiCl 4

+

TMS

OMe Ph

SnCl 4 CH2 Cl2 -78 to -40 °C 72%

OMe O CI

(39) Ph

118

4 Tin-promoted Aldol Reactions and their Application to Total Syntheses of Natural Products

Some examples of reactions using ESE and KSA with acetals to produce the corresponding b-alkoxy carbonyl compounds are given in Eqs. (40)–(43) [29–32]. This method is useful for the synthesis of protected aldols directly from the silyl enolates and acetals. OMe +

OMe

N SO2Ph

OMe O

SnCl 4

OTMS Ph

Ph

N SO2Ph

CH2Cl 2, -70 °C 93%

(40)

Ph

N H

TMS

SnCl 4

OTMS

TMS

O

+ C5H 11

F F Br

O

OMe

Ph

OTMS

OEt

O

Ph

(41)

F F

SnCl 4

+

OEt O

Br

OEt

CH2Cl 2, -78 °C 83%

AcO + O

C5H 11

CH2Cl 2, -78 °C 92%

OEt

OTMS

SnCl 4 CH2Cl 2, -78 °C 87%

(42)

AcO

O O

ð43Þ Although little systematic research has been performed on the stereoselectivity of this reaction, some approaches to the use of chiral acetal parts for asymmetric synthesis have been reported. Kishi successfully accomplished the preparation of an optically active synthetic intermediate of aklavinone by reaction of a-stannylketone with a chiral aromatic aldehyde acetal as depicted in Eq. (44) [33]. A similar achiral acetal was also used for synthesis of racemic 4-demethoxydaunomycinone by Rutledge in 1986 (Eq. (45)) [34]. Rutledge’s and Yamamoto’s approaches are noteworthy for the stereoselective synthesis of optically active b-alkoxy carbonyl compounds in SnCl 4 promoted aldol reactions (Eqs. (46) and (47)) [35, 36].

4.2 Tin-promoted Intermolecular Aldol Reactions

OH O

OH

O

O O

OH O

OH O

+ MeCN, -20 °C 83%, ds 10/1

TMS O

OH O

SnCl 4

119

CO2Me

CO2Me

O OH O

OH OH OH

O

CO2Me

Aklavinone

ð44Þ O

OH OMe OMe

O

O OTMS

SnCl 4

O

MeCN, -23 °C 70%

OH OMe O

+

OMe CO2Me

O O

OMe CO2Me

O

OH OH OH

O OH O rac-4-demethoxydaunomycinone

ð45Þ MeO O Ph O OTMS

SnCl 4

O

MeO

MeO

H O

Cl 3Sn

MeCN, -23 °C

MeCN, -23 °C 73%

Ph

O

O MeO 3α/3β=64/36

ð46Þ TMS TMS

OTMS O O

O O

1 + R

R3 R2

O

SnCl 4 TMS

CH2 Cl2 -78, -97 or -125 °C 55~91%

R3

O R1 R2 64~94% ee

ð47Þ

120

4 Tin-promoted Aldol Reactions and their Application to Total Syntheses of Natural Products

Otera discovered that Sn(IV) Lewis acids such as Bu2 Sn(OTf )2 , Bu2 Sn(ClO4 )2 , Bu3 SnClO4 , and (C6 F5 )2 SnBr2 are suitable activators for the aldol reaction of ESE and KSA with aldehydes [37]. Interestingly, it was proved by elaborate research that reactions of ESE with ketones and of KSA with acetals do not proceed under the inﬂuence of these catalysts. It was also shown that a,b-unsaturated aldehydes are much more reactive than aromatic and saturated aliphatic aldehydes. They expanded this concept to ‘‘parallel recognition’’, in which some reaction patterns proceed exclusively, aﬀording the desired adducts only, on treatment of several diﬀerent nucleophiles and electrophiles in one pot. When an ambident electrophile is used in this new strategy ESE and KSA, respectively, react with one of the electrophilic points in the substrate to produce a single adduct in high yield with perfect selectivity (Eqs. (48) and (49)). TMSO t

OMe O

+

+

OTBS

MeO

Bu

TBS

(C 6F5 )2SnBr2

O MeO

O

O

OEt CH Cl , -78 °C t 2 2 Bu 73%

OEt

ð48Þ OTMS

OH O

+

O

Ph

(C 6F5 )2SnBr2

Ph

H OTBS +

O

OEt

CH2Cl 2, -78 °C 58%

CO2Et OTBS

ð49Þ

4.2.4

Reaction of Dienol Silyl Ethers

Siloxyheteroaromatic compounds function as nucleophilic dienol silyl ethers with carbonyl compounds or acetals under the inﬂuence of a Lewis acid catalyst. For example, Takei showed that 2-siloxyfurans react with aldehydes and acetals to give the corresponding g-substituted g-lactones in high yields (Eqs. (50) and (51)) [38]. O + R1

OTMS

R2

MeO OMe R1

O

R2

SnCl 4 CH2Cl 2, -78 °C 70~96%

+

O

OTMS

SnCl 4 CH2Cl 2, -78 °C 66~92%

R1 R2

OH

R1 R2

O

O

(50)

OMe O

O

(51)

4.3 Tin-promoted Intramolecular Aldol Reactions

2-Siloxypyrrole has also been used as a suitable nucleophile in the SnCl 4 promoted aldol reaction with aldehydes shown in Eq. (52). Several polyoxy a-amino acids and carbocyclic amines were synthesized by Rassu and Casiraghi using a stereoselective aldol reaction of 2-siloxypyrrole with protected glyceraldehyde in the presence of SnCl 4 [39]. O H

O

Boc N OTBS

+

O

OH Boc N O

SnCl 4 O O

Et 2O, -80 °C 80% >95% ds OH O

HO

NH2

HO OH

OH

NH2

(52) NH2

HO OH

HO

HO

OH

OH

Baldwin successfully used the reaction of functionalized 2-siloxypyrrole with 2-methylpropionaldehyde for synthesis of an intermediate of lactacystin, a natural g-lactam (Eq. (53)) [40]. Ph

Ph O

O H

+

O

SnCl4 N

OTBS

N

HO

Et 2O, -78 °C 61%

O

5

5α/5β=9/1

(53) AcHN O HO2C

S HO

H N

O

OH Lactacystin

4.3

Tin-promoted Intramolecular Aldol Reactions 4.3.1

The Intramolecular Aldol Reaction of Silyl Enolates

Not only intermolecular additions of ESE and KSA to acetals, but intramolecular reactions of acetals with a silyl enolate moiety are also quite

121

122

4 Tin-promoted Aldol Reactions and their Application to Total Syntheses of Natural Products

eﬀective, especially for synthesis of strained cyclic compounds. Kocienski fully studied the intramolecular reaction of ESE to give medium-sized compounds (Eq. (54)), and this method was even found to be applicable to the synthesis of an eight-membered carbocycle (Eq. (55)) [41]. Paquette also succeeded in preparing a bicyclic eight-membered ring compound as shown in Eq. (56) [42]. OSiMe 2Ph O

SnCl 4

(54) O

CH2Cl 2, -78 °C 30%

O

OH

O

cis/trans=50/50 OTMS O

O

SnCl 4

O

HO

O

(55)

CH2Cl 2, -40 °C 35% OSiMe 2Ph

O

SnCl 4 proton sponge

(56) O

CH2Cl 2, -78 °C 69%

O MeO

Tatsuta recently reported a total synthesis of pyralomicin 1c in which SnCl 4 -catalyzed cyclization was eﬀectively employed for formation of a 6membered ring core as shown in Eq. (57) [43].

O

TBS TBS O OMe

PhO2S TBSO

O SnCl 4

PhO2S

OTBS

OMe CH2Cl 2, -78 °C OTBS 71%

OTBS OTBS OH OH

HO

N

(57)

OH O

Cl Cl O

OH

Pyralomicin 1c

4.3 Tin-promoted Intramolecular Aldol Reactions

[1,3] rearrangement of ESE with anomeric carbon was developed by Ley in 1998, and several 6- and 5-membered C-glycosides have been prepared by SnCl 4 acceleration (Eqs. (58) and (59)) [44]. OTMS C6H13

O

R = Ph, C

O

CPh, tBu

CH2Cl 2, -30 °C 79~86%

SnCl 4 O

C6H13

R

OTMS O

OH

SnCl 4

C7H 15

O

R

(58)

O α/β=75/25 O

HO

O

CH2 Cl2 56%

(59)

C7H 15

4.3.2

Reaction of Dienol Silyl Ethers or g-Silyl-a,b-enones

In 1986, Kuwajima established a method for generation of g-stannyl-a,benones as nucleophilic species. It was found that these reactions with acetals proceeded smoothly to aﬀord the corresponding coupling products in good yields (Eq. (60)) [45].

O

O

SnCl 4 CH 2Cl2 , 0 °C

TMS

Cl 3Sn γ-stannylenone

O

OMe

γ-stannylenone O

OMe

(60) O

OMe

CH 2Cl2 , 0 °C 71%

Remarkable results from SnCl 4 -induced aldol cyclization using dienol silyl ethers have been observed in synthetic studies on taxane diterpenoids [46]. In the initial approach for construction of the basic skeleton of a taxane ring using a g-stannyl-a,b-enone generated from a g-silyl-a,b-enone by SiaSn metal exchange, the yield of the desired tricyclic compound was unsatisfactory (Eq. (61)). Cyclization of acetals with a dienol silyl ether moiety promoted by SnCl 4 occurred rapidly, however, to aﬀord the aromatic taxanes in high yields (Eqs. (62) and (63)). Kuwajima recently accomplished the total synthesis of paclitaxel (taxol), using this intermediate as the main component with the eight-membered ring core.

123

124

4 Tin-promoted Aldol Reactions and their Application to Total Syntheses of Natural Products

TMS

O

TIPSO H

OBn OBn

TIPSO HO

SnCl4

OMe

MeO

PhS

OMe

OMe OMe

OH

E/Z=60/40

CH2 Cl2, rt 15~17% [SnCl 4/TiCl 4; -23 °C; 73%]

OMe OMe

OH OMe

(61)

O

OMe

H

MeO

OMe

SnCl4

(62)

CH2Cl 2, -45 °C Z ; 77% E ; 66%

(MeBO)3 pyridine benzene, rt 90%

O H

i) SnCl4 CH2Cl2 -78 to -45 °C

OH OMe

PhS

ii) pinacol DMAP benzene, rt 76% (2 steps)

O HO

OBn

OH

AcO O Ph

NH

O

OH

O

Ph

O

OH

H O O Ac Bz Paclitaxel (Taxol®)

O

HO

ð63Þ

4.4

Chiral Diamine–Sn(II) Complex-promoted Aldol Reactions

Enantioselective aldol addition is one of the most powerful tools for construction of new carbon–carbon bonds with control of the absolute conﬁgurations of new chiral centers, and the utility of this reaction has been demonstrated by several applications to the synthesis of natural products such as carbohydrates, macrolide and polyether antibiotics, etc. In the asymmetric aldol reactions reported chiral auxiliary groups are usually attached to the reacting ketone-equivalent molecules. Until the early 1980s there had been no example of an aldol-type reaction in which two achiral carbonyl compounds were used to form a chiral molecule with the aid of a chiral ligand.

4.4 Chiral Diamine–Sn(II) Complex-promoted Aldol Reactions

Chiral auxiliaries derived from (S)-proline seemed to be particularly attractive, because they have conformationally rigid pyrrolidine rings. Chiral diamines derived from (S)-proline, especially, are successfully employed for creation of an eﬃcient chiral environment because almost all the main and transition metals having vacant d orbitals are capable of accepting a bidentate ligand. An intermediate derived from the chiral ligand and an organometallic reagent would have a conformationally restricted cis-fused ﬁve-membered ring chelate and would aﬀord optically active organic compounds by reaction with appropriate substrates. 4.4.1

Asymmetric Aldol and Related Reactions of Sn(II) Enolates

Enantioselective aldol reaction via Sn(II) enolates coordinated with chiral diamines was explored by Mukaiyama and Iwasawa in 1982 [10c, 47]. In the presence of chiral diamine 1a, various optically active aldol adducts were produced by reactions between aromatic ketones and aldehydes (Eqs. (64) and (65)). This is the ﬁrst example of the formation of crossed aldol products in high optical purity, using chiral diamines as chelating agents, starting from two achiral carbonyl compounds.

N

N Me Sn(OTf) 2

O

1a

OSnOTf

N N Me Sn O OTf

Ph

Ph NEt

Ph chiral Sn(II) enolate

CH2 Cl2, -78 °C

O R

OH O

H

-95 °C 57~78%

R * *

Ph

syn/anti=80/20~100/0 75~90% ee (syn)

ð64Þ O Sn(OTf) 2

O Ph

NEt CH2 Cl2, -78 °C

diamine 1a

Ph

H

-95 °C 35%

OH O Ph

Ph 75% ee

(65)

125

126

4 Tin-promoted Aldol Reactions and their Application to Total Syntheses of Natural Products

This procedure is successfully applied to the reactions of carboxylic acid derivatives such as thioamides and thione esters (Eqs. (66) and (67)) [48]. 3-Acetylthiazolidine-2-thiones are quite suitable substrates for the Sn(II) enolate-mediated asymmetric aldol reaction, and various optically active bhydroxy 3-acylthiazolidine-2-thiones are obtained by use of chiral diamine 1a (Eq. (68)) [49]. O LDA / THF

S NMe2

diamine 1a

Ph

OH S

H Ph

Sn(OTf) 2 -78 °C

CH2 Cl2 -20 °C 93%

NMe2

(66)

syn/anti =92/8 85% ee (syn)

O diamine 1a

Sn(OTf) 2

S

Ph

OH S

H Ph

OMe

OMe

(67)

73%

NEt

syn/anti =78/22 90% ee (syn)

CH2Cl 2, -78 °C O S

O N

diamine 1a

Sn(OTf) 2

R

S

-95 °C 63~81%

NEt

S

OH O

H R

N

S

(68)

65~>90% ee

CH2Cl 2, -78 °C

A complex formed from chiral diamine 2a with the Sn(II) enolate of 3acetylthiazolidine-2-thione reacts with some a-keto esters to aﬀord aldol adducts with tertiary hydroxyl groups and high ee, as shown in Eq. (69) [50].

N Me

S

O N

Sn(OTf) 2

N H

S

R

O

NEt

OMe

R CH2Cl 2, -78 °C

S

MeO2C OH O

2a

N

S

(69)

85~>95% ee

O 65~80%

When 3-(2-benzyloxyacetyl)thiazolidine-2-thione is treated under these reaction conditions the corresponding anti-diol groups are produced with

4.4 Chiral Diamine–Sn(II) Complex-promoted Aldol Reactions

good diastereoselectivity and high enantioselectivity by addition of chiral diamine 1a (Eq. (70)) [51]. O S

O BnO

N

Sn(OTf) 2

diamine 1a

S

R

OH O

H R

N S OBn syn/anti =19/81~7/93 87~94% ee (anti )

68~93%

NEt

S

CH2Cl 2, -78 °C

ð70Þ Because Sn(II) enolates of thioesters are generated by the reaction of Sn(II) thiolates with ketenes, the optically active b-hydroxy thioesters are also readily synthesized by way of an aldol reaction with aldehydes in the presence of Sn(OTf )2 and chiral diamine 1a (Eq. (71)) [52]. O O

Sn(StBu)2

Sn(OTf) 2

diamine 1a

CH2Cl2, -78 °C

R

OH O

H

-100 ˚C 55~70%

StBu

R

(71)

58~80% ee

Mukaiyama and Iwasawa also developed an enantioselective Michael Addition reaction using Sn(OTf )2 with chiral diamines [48, 53]. For example, Sn(II) enolate of methyl ethanedithioate reacts with benzalacetone in the presence of chiral diamine 2a and trimethylsilyl triﬂuoromethanesulfonate (TMSOTf ) to give the corresponding Michael adduct in 82% yield with good enantioselectivity (Eq. (72)). Sn(OTf) 2

S SMe

NEt

diamine 2a

TMSOTf

O

O

SMe * (72) 82%, 70% ee (after hydrolysis)

Ph CH2 Cl2, -78 °C

Ph

S

The Michael adducts were obtained from trimethylsilyl enethioate and a,b-unsaturated ketones in high yields with moderate to good enantioselectivity by use of a catalytic amount of chiral diamine–Sn(OTf )2 complex (Eq. (73)) [48, 54]. As shown in Scheme 4.2, SiaSn metal exchange occurs rapidly to generate chiral Sn(II) enolate and TMSOTf in situ, because the silicon–sulfur bond is rather weak and tin has high aﬃnity for sulfur. Activation of the a,b-unsaturated ketone by TMSOTf would lead to the Michael reaction, aﬀording the silyl enolate of the Michael adduct and regeneration of the chiral diamine–Sn(II) complex. To preclude the competitive

127

128

4 Tin-promoted Aldol Reactions and their Application to Total Syntheses of Natural Products

N NRR' Me Sn TfO OTf

STMS SMe

R2

TMSO

chiral Sn(II) complex

1

R

S SMe

* O R1

R2

N NRR' Me Sn S OTf SMe + TMSOTf Scheme 4.2

Catalytic asymmetric Michael reaction using chiral Sn(II) enolate.

direct reaction of KSA with a,b-unsaturated ketone under the inﬂuence of TMSOTf, the concentration of the KSA is kept low by slow addition of a solution of the KSA to the reaction mixture.

O

STMS

diamine 2a (0.11 eq) Sn(OTf) 2 (0.1 eq)

R2

O

S

+ R1

R2

SMe

CH2Cl 2, -78 °C 79~82% (after hydrolysis)

R1

SMe * 40~70% ee

(73)

These chiral enolate preparations and reactions with several electrophiles giving the optically active aldols have been applied to the synthesis of natural compounds such as b-lactam antibiotics. References in reviews by Mukaiyama et al. [10c, 48, 55] describe advanced studies on Sn(II) enolate chemistry. 4.4.2

Chiral Diamine–Sn(II) Complex-promoted Aldol Reactions

Chiral Sn(II) Lewis acids prepared in situ by coordination of chiral pyrrolidine derivatives to Sn(OTf )2 were developed by Mukaiyama and Kobayashi in 1989 to promote the asymmetric aldol reaction of ESE or KSA with carbonyl compounds. Some chiral Lewis acids had already been reported and fruitful results were observed in the ﬁeld of the Diels–Alder and related reactions, in particular, in the late 1980s. The chiral Lewis acids employed for

4.4 Chiral Diamine–Sn(II) Complex-promoted Aldol Reactions

Sn(OTf) 2 N

N Me

N N Me Sn TfO OTf

CH2 Cl2, rt 1a

chiral Sn(II) complex 1a-Sn Sn(OTf) 2

N H

N Me

N N Me Sn H TfO OTf

CH2 Cl2, rt

2a

chiral Sn(II) complex 2a-Sn

Scheme 4.3

Chiral Sn(II) Lewis acids generated from Sn(OTf )2 with diamines.

these reactions were rather strong and hard acidic metals such as aluminum and titanium. Chiral Sn(II) Lewis acids prepared in situ by chelation of a chiral diamine to Sn(OTf )2 , might, on the other hand, be eﬀective because Sn(II) is a soft metal and the complex has one vacant d orbital to be coordinated with oxygen in the carbonyl group of an aldehyde without losing the favorable asymmetric environment (Scheme 4.3). On this basis a variety of eﬃcient asymmetric aldol reactions between achiral silyl enolates and achiral carbonyl compounds have been developed. Some chiral diamines used in the asymmetric aldol reaction of KSA with carbonyl compounds promoted by Sn(OTf )2 are listed in Scheme 4.4. 4.4.3

Asymmetric Aldol Reaction of Silyl Enolates

Asymmetric aldol reaction of a KSA derived from S-Et ethanethioate with aldehydes achieves high ee by employing a chiral promoter, the combined use of Sn(OTf )2 coordinated with the chiral diamine (1a or 2a), and tributyltin ﬂuoride (Eq. (74)) [56]. diamine 1a or 2a Sn(OTf) 2 OTMS

O R

OH O

+ H

SEt

Bu3SnF CH2Cl 2, -78 °C 50~90%

R

SEt

(74)

78~>98% ee

The complex consisting of Sn(OTf )2 with 2a is quite eﬀective for reaction of KSA generated from S-Et propanethioate with aldehydes to aﬀord the corresponding syn aldol adducts with excellent diastereoselectivity and enantioselectivity (Eq. (75)) [56b, 57]. A highly enantioselective aldol reac-

129

130

4 Tin-promoted Aldol Reactions and their Application to Total Syntheses of Natural Products

N

N Me

N H

N Me 1a

2a N

N Et 1b

N Me 3a

N

N 6

N

N Et

1d

N

5 N

N Me

1c

N Pent

N Me

2b N

N

4

N H

N Me

N Pr

N Pr

N Me

N 7

3b

Scheme 4.4

Useful chiral diamines for asymmetric aldol reaction.

tion of the KSA of benzyl acetate with achiral aldehydes can be conducted using the chiral promoter formed from Sn(OTf )2 with 1b (Eq. (76)) [58].

OTMS

O + R

H

O + R

H

SEt

OTBS OBn

diamine 2a Sn(OTf) 2 Bu3 SnF [Bu 2Sn(OAc)2 ] CH2Cl 2, -78 °C 48~91% [70~96%]

OH O R

(75) syn/anti =>99/1 >98% ee (syn)

diamine 1b Sn(OTf) 2 Bu3SnF mesitylene CH2Cl 2, -95 °C 51~79%

SEt

OH O R

OBn

(76)

89~>98% ee

In the presence of a promoter including the chiral diamine 1d, the KSA of a thioester reacts with a-ketoesters to aﬀord the corresponding aldol-type adducts, 2-substituted malates, in good yields with excellent ee (Eq. (77)) [59].

4.4 Chiral Diamine–Sn(II) Complex-promoted Aldol Reactions

O

OTMS

diamine 1d Sn(OTf) 2

MeO 2C OH O

+ R

CO2Me

SEt

R = Me, iPr, Ph

R SEt 92~>98% ee

Bu3SnF CH2Cl 2, -78 °C 74~81%

(77)

In the course of developments in asymmetric synthesis using Sn(OTf )2 it has also been revealed that a tetrahydrothiophene ligand, an analog of the chiral diamine, also aﬀords an asymmetric environment around the Sn(II) metal suitable for promoting the enantioselective aldol reaction giving the desired adducts with high selectivity (Eq. (78)) [60].

S OTMS

O + R

H

SEt

N H OH O

Sn(OTf) 2 Bu 2Sn(OAc)2 CH2Cl 2, -78 °C 48~98%

R

(78) SEt

syn/anti =86/14~100/0 63~93% ee (syn)

4.4.4

Catalytic Asymmetric Aldol Reaction

Catalytic asymmetric synthesis is an extremely desirable method for producing optically active compounds from achiral substrates. If asymmetric ampliﬁcation could be realized by employing a catalytic amount of a chiral source, substantial amounts of optically active compounds could be synthesized in a convenient and rational way. The promoter, consisting of a chiral diamine and Sn(OTf )2 , has mild acidity which accelerates the asymmetric aldol reaction of KSA with aldehydes; these reactions always required a stoichiometric amount of the chiral diamines, however. From experimental examination of the mechanism of the stoichiometric asymmetric aldol reaction, Mukaiyama and Kobayashi considered the possibility of catalytic use of the chiral diamine–Sn(OTf )2 complex. Their hypothesis was:

. the reaction ﬁrst produces Sn(II) alkoxides and TMSOTf (Scheme 4.5); . if the substrates (KSA and aldehyde) are added quickly to a solution of .

the chiral diamine–Sn(OTf )2 complex, the initially formed TMSOTf promotes the achiral asymmetric process to produce the racemic aldol from the remaining substrates; and if the substrates are added slowly to a solution of the chiral diamine– Sn(OTf )2 complex, in accordance with the reaction rate of KSA with aldehyde, transmetalation from Sn(II) on the formed Sn(II) alkoxide to Si would occur by sequential reaction with TMSOTf in situ.

131

4 Tin-promoted Aldol Reactions and their Application to Total Syntheses of Natural Products

132

O R

N NRR' Me Sn TfO OTf

H +

TMSO

chiral Sn(II) complex

R

OTMS

O SEt

SEt

N NRR' Me Sn TfO O O R

SEt

+ TMSOTf Scheme 4.5

Catalytic asymmetric aldol reaction using chiral Sn(II) complex.

Actually, an optically active trimethylsilyl ether of the aldol was obtained by the reaction of KSA derived from S-Et propanethioate with aldehydes by slow addition of a mixture of substrates to a solution including a catalytic amount of the chiral diamine–Sn(OTf )2 complex (Eq. (79)) [61].

OTMS

O + R

H

SEt

diamine 2a or 2b (0.22 eq) Sn(OTf) 2 (0.2 eq)

OH O

R SEt CH2Cl 2, -78 °C 64~86% syn/anti =93/7~100/0 (after hydrolysis) 91~>98% ee (syn)

(79)

The rate of this transmetalation is aﬀected by the conditions, particularly the solvent. Propionitrile was found to be a suitable reaction medium for the catalytic process, and a variety of the optically active aldol adducts were prepared with high ee when a solution of aldehydes and KSA derived from S-Et propane- or ethanethioate was added to the catalyst consisting of a chiral diamine and Sn(OTf )2 in propionitrile (Eqs. (80) and (81)) [61, 62].

OTMS

O + R

H

SEt

diamine 2a (0.11~0.22 eq) Sn(OTf) 2 (0.1~0.2 eq)

OH O R

(80) SEt

EtCN, -78 °C 65~80% syn/anti = 89/11~100/0 (after hydrolysis) 89~>98% ee (syn)

4.4 Chiral Diamine–Sn(II) Complex-promoted Aldol Reactions

O

OTMS

diamine 2a or 2b (0.22 eq) Sn(OTf) 2 (0.2 eq)

OH O

(81)

+ R

H

SEt

EtCN, -78 °C 48~90% (after hydrolysis)

R

SEt 68~93% ee

A novel combined catalyst generated from SnO and TMSOTf was also developed for catalytic synthesis of the desired compounds, as shown in Eq. (82) [63]. Although the exact structure of the complex is unclear, SnO interacted with TMSOTf and formed an acidic species which functions as a chiral catalyst in this reaction. Kobayashi employed SnO as an eﬀective additive for aldol reaction of KSA with aldehydes, promoted by the chiral diamine–Sn(OTf )2 complex, that is, the optically active aldol adducts were synthesized with high stereoselectivity by using a new combination, a chiral diamine–Sn(OTf )2 aSnO (Eq. (83)) [64].

OTMS

O R

+

+ H

SEt R' R' = H, Me

OH O

R SEt CH2Cl 2, -78 °C 58~82% syn/anti = 91/9~98/2 (after hydrolysis) 67~94% ee (syn)

diamine 2a (0.22 eq) Sn(OTf) 2 (0.2 eq) OTMS SnO (0.2~0.4 eq)

O R

SEt

H

diamine 2a (0.5 eq) SnO (1.0 eq) TMSOTf (0.65 eq)

(82)

OH O

R SEt EtCN, -78 °C R' 50~85% syn/anti = 95/5~100/0 (after hydrolysis) 84~>98% ee (syn)

(83)

Evans recently designed an original chiral Sn(II) catalyst generated from bis(oxazoline) and Sn(OTf )2 [65]. Bis(oxazoline) functions as a bidentate ligand to Sn(OTf )2 and the complex formed might have a rigid C2 -symmetric structure creating an excellent asymmetric environment. This asymmetric aldol reaction proceeds with high diastereoselectivity and enantioselectivity to give the corresponding adducts when bis-functionalized electrophiles such as alkyl glyoxylates, a-ketoesters, and a-diketones are used, because of the formation of suitable complexes with bis(oxazoline) and Sn(OTf )2 . For example, the desired optically active anti aldols were obtained by reaction of KSA derived from thiol esters with ethyl glyoxylate using a catalytic amount

133

134

4 Tin-promoted Aldol Reactions and their Application to Total Syntheses of Natural Products

of the Bn/Box–Sn(II) complex. Anti-(2R,3S) selectivity is also observed in the construction of b-hydroxy-a,b-dimethyl thioesters by aldol addition of the KSA of S-Et propanethioate to an a-ketoester in the presence of a Bn/ Box–Sn(II) complex. A distinctive feature of this asymmetric aldol reaction is that an anti-(2S,3R) adduct, corresponding to the optical antipode of the above anti-(2R,3S) adduct, was produced on use of a Ph/PyBox–Sn(II) complex with the same chirality at the C4 position in the oxazoline moiety of Bn/Box (Eqs. (84) and (85)).

O

O N

N

Sn Bn TfO OTf Bn O R3O

OTMS 2

R

+

SR R1

O

OTMS

MeO

+

N

Ph

N Sn N TfO OTf

O

Ph

(85)

OH O

(0.1 eq) MeO

SR CH2Cl 2, -78 °C 81~94% (after hydrolysis)

R1

O

(84)

R3O

SR CH2Cl 2, -78 °C O R1 72~90% syn/anti =10/90~4/96 (after hydrolysis) 92~98% ee (anti )

O

O

HO R2 O

(0.1 eq)

SR O

R1

syn/anti =5/95~1/99 92~99% ee (anti )

KSA prepared from acetic acid and a,a-dialkyl-substituted acetic acid derivatives also reacted with ethyl glyoxylate to produced the desired aldol adducts with high enantioselectivity, as shown in Eqs. (86) and (87) [65, 66].

O

Bn O EtO

OTMS H

O

+

O N N Sn TfO OTf Bn (0.1 eq)

SPh CH2Cl 2, -78 °C 90% (after hydrolysis)

(86)

OH O EtO

SPh O 98% ee

4.4 Chiral Diamine–Sn(II) Complex-promoted Aldol Reactions

O

O N

Bn O EtO

OTMS +

H

N Sn TfO OTf Bn (0.1 eq)

OH O EtO

SEt

SEt CH2Cl 2, -78 °C 89% (after hydrolysis)

O

(87)

O 95% ee

4.4.5

Asymmetric Synthesis of syn- and anti-1,2-Diol Groups

Optically active 1,2-diol groups are often observed in nature as carbohydrates, macrolides, or polyethers, etc. Several excellent asymmetric dihydroxylation reactions of oleﬁns using osmium tetroxide with chiral ligands have been developed to give the optically active 1,2-diol groups with high enantioselectivity. Some problems remain, however, for example, preparation of the optically active anti-1,2-diols, etc. The asymmetric aldol reaction of a KSA derived from an a-benzyloxy thioester with aldehydes has been developed by Mukaiyama et al. to introduce two hydroxyl groups simultaneously with stereoselective carbon–carbon bond-formation using the chiral Sn(II) Lewis acid. First, a variety of optically active anti-a,b-dihydroxy thioester derivatives were obtained in good yield with excellent diastereoselectivity and enantioselectivity when the chiral diamine 1a or 1b, Sn(OTf )2 , and dibutyltin diacetate were employed together (Eq. (88)) [67]. By means of current aldol methodology, two hydroxyl groups can be stereoselectively introduced at the 1,2-position during formation of the new carbon–carbon bond.

OTMS

O + R

H

diamine 1a or 1b Sn(OTf) 2

SEt OBn

OH O R

Bu 2Sn(OAc)2 CH2Cl 2, -78 °C 59~88%

SEt OBn

(88)

syn/anti = 2/98~1/99 95~98% ee (anti )

On the other hand, several syn-aldol adducts are obtained under the same reaction conditions, i.e. in the presence of chiral diamine 1a, Sn(OTf )2 , and dibutyltin diacetate. The reaction of a KSA with a t-butyldimethylsiloxy group at the 2-position with achiral aldehydes smoothly proceeds to give the corresponding syn-a,b-dihydroxy thioester derivatives in high yield with good stereoselectivity. When a chiral diamine 1c, which is similar to 1a in possessing a propyl group on the nitrogen of the pyrrolidine ring, is used

135

136

4 Tin-promoted Aldol Reactions and their Application to Total Syntheses of Natural Products

the ee increases up to 94% (Eq. (89)) [68, 67b]. Now it becomes possible to control the enantiofacial selectivity of the KSA derived from a-hydroxy thioester derivatives just by choosing the appropriate protective groups of the hydroxy parts of the KSA, and the two diastereomers of the optically active a,b-dihydroxy thioesters can be synthesized. OTMS

O + R

H

SEt OTBS

diamine 1c Sn(OTf) 2

OH O R

Bu 2Sn(OAc)2 CH2Cl 2, -78 °C 46~93%

SEt OTBS

(89)

syn/anti = 88/12~97/3 82~94% ee (syn)

Kobayashi also introduced several new types of chiral diamine, for example 4 and 5, to obtain rather higher selectivity for the synthesis of syn-a,bdihydroxy thioester derivatives, as shown in Eq. (90) [69, 70]. OTMS

O + R

H

SEt OTBS

diamine 4 or 5 Sn(OTf) 2

OH O R

Bu 2Sn(OAc)2 CH2Cl 2, -78 °C 4; 69~89% 5; 61~86%

SEt OTBS

(90)

4; syn/anti = 94/6~99/1 86~96% ee (syn) 5; syn/anti = 98/2~>99/1 96~99% ee (syn)

Diastereoselective and enantioselective synthesis of both stereoisomers of a,b-dihydroxy-b-methyl thioester derivatives has also been achieved by reaction of KSA with a benzyloxy or t-butyldimethylsiloxy group at the 2position, promoted by an Sn(II) Lewis acid including chiral diamine 1c or 4 (Eqs. (91) and (92)) [69, 71].

OTMS

O + R

CO2Me

R = Me, Ph

OTMS +

R

CO2Me

R = Me, Ph

SEt OTBS

MeO 2C OH O R

SEt OBn

O

diamine 1c Sn(OTf) 2 Bu3SnF CH2Cl 2, -78 °C 66~93%

diamine 4 Sn(OTf) 2

(91)

syn/anti =13/87~7/93 91% ee (anti)

MeO 2C OH O R

Bu3SnF CH2Cl 2, -78 °C 76~89%

SEt OBn

SEt OTBS syn/anti = 84/16~94/6 87~88% ee (syn)

(92)

4.4 Chiral Diamine–Sn(II) Complex-promoted Aldol Reactions

KSA derived from phenyl esters have a unique capacity to promote remarkable stereoselectivity in asymmetric aldol reactions involving a chiral diamine–Sn(II) complex [72–74]. For instance, Kobayashi found that (E)KSA derived from p-methoxyphenyl (t-butyldimethylsiloxy)acetates reacts with aldehydes to aﬀord the corresponding anti-1,2-diol derivatives with high diastereoselectivity and enantioselectivity when promoted by an Sn(II) Lewis acid complexed with chiral diamine 1a (Eq. (93)) [74].

OTMS

O + R

H

OPMP OTBS

diamine 1a Sn(OTf) 2

OH O R

Bu 2Sn(OAc)2 CH2Cl 2, -78 °C 31~95%

OPMP OTBS

(93)

syn/anti = 31/69~2/98 84~95% ee (anti )

Reaction of (Z)-KSA derived from phenyl benzyloxyacetate with aldehydes, using chiral diamine 3b, also aﬀords the optically active anti-1,2aldols preferentially (Eq. (94)). However, the corresponding syn aldols were formed when the reaction was conducted in the presence of chiral diamine 6 (Eq. (95)) [75]. The latter reaction also proceeded when accelerated by a catalytic amount of chiral diamine 2a or 2b under Kobayashi conditions (Eq. (96)) [76].

O

OTMS +

R

H

+ H

BnO

OPh

OTMS

O R

OPh

OTMS

O R

BnO

+ H

BnO

OPh

diamine 3b Sn(OTf) 2

OH O R

Bu 2Sn(OAc)2 CH2Cl 2, -78 °C 51~80%

OPh OBn syn/anti =12/88~7/93 90~94% ee (anti )

diamine 6 Sn(OTf) 2

OH O R

Bu 2Sn(OAc)2 CH2Cl 2, -78 °C 85~90%

diamine 2a or 2b (0.2 eq) Sn(OTf) 2 (0.2 eq)

(94)

OPh

(95)

OBn syn/anti =94/6~>99/1 91~98% ee (syn)

OH

O

R OPh SnO (0.2 eq) OBn EtCN, -78 °C syn/anti =90/10~>98/2 68~87% 80~96% ee (syn) (after hydrolysis)

(96)

137

138

4 Tin-promoted Aldol Reactions and their Application to Total Syntheses of Natural Products

This method of producing chiral aldols is also applicable to the construction of an asymmetric quarternary carbon included in the 1,2-diol groups. In the presence of a chiral promoter consisting of the chiral diamine 1a, Sn(OTf )2 , and dibutyltin diacetate, optically active anti-a,b-dihydroxy-amethyl thioester and phenyl ester derivatives were synthesized in good yields with high stereoselectivity (Eqs. (97) and (98)) [72, 73].

OTMS

O + Ph

H

E/Z =12/88

OTMS +

Ar

H

OH O

SEt OBn

O

diamine 1a Sn(OTf) 2 Ph Bu 2Sn(OAc)2 CH2Cl 2, -78 °C 58% diamine 1a Sn(OTf) 2

OH O Ar

Bu 2Sn(OAc)2 CH2Cl 2, -78 °C 44~72%

(97)

syn/anti = 2/98 97% ee (anti )

OPh OBn E/Z =58/42

SEt OBn

OPh OBn

(98)

syn/anti = 26/74~8/92 73~95% ee (anti)

Another interesting phenomenon in which the corresponding syn-a,bdihydroxy-a-methyl ester derivatives were produced from similar KSA using a stoichiometric or catalytic amount of the chiral catalyst containing diamine 2a as shown in Eqs. (99) and (100) [73].

OTMS

O + R

H

OR OBn

R = Et, iPr, Ph E/Z =58/42~71/29

OTMS

O Ph

+ H

diamine 2a Sn(OTf) 2 Bu 2Sn(OAc)2 CH2Cl 2, -78 °C 52%~quant.

R OR BnO syn/anti = 81/19~98/2 80~97% ee (syn)

diamine 2a (0.24 eq) Sn(OTf) 2 (0.2 eq)

Oi Pr OBn E/Z =71/29

OH O

EtCN, -78 °C 60% (after hydrolysis)

(99)

OH O Ph BnO

Oi Pr

(100)

syn/anti = 90/10 96% ee (syn)

Similarly, it was found that KSA (E=Z ¼ 38 to 62) derived from p-methoxyphenyl a-benzyloxypropionate reacted with aldehydes in the presence of a Sn(II) catalyst containing diamine 3a to give the corresponding anti-aldol groups (Eq. (101)), whereas asymmetric aldol reaction of (E)-KSA, derived from p-methoxyphenyl a-benzyloxypropionate with a variety of aldehydes,

4.4 Chiral Diamine–Sn(II) Complex-promoted Aldol Reactions

promoted by Sn(OTf )2 coordinated by chiral diamine 2a, aﬀorded the stereoisomeric syn compounds with high ee (Eq. (102)) [77]. Tetrasubstituted KSA with an alkylthio group also reacted with aldehydes to produce the syn-aldol compounds preferentially [78]; these were used as synthetic intermediates of anti-b-hydroxy-a-methyl groups in the total synthesis of octalactins, described in a later section (Eq. (103)).

OTMS

O + R

H

BnO

OPMP

E/Z=38/62

O

OTMS +

R

H

OPMP OBn

OTMS

O + R

H

MeS

OEt

E/Z=12/88

diamine 3a Sn(OTf) 2

OH O R

Bu 2Sn(OAc)2 CH2Cl 2, -78 °C 53~69%

diamine 2a Sn(OTf) 2

OPMP OBn syn/anti = 6/94~3/97 88~92% ee (anti )

OH O

R OPMP Bu 2Sn(OAc)2 BnO CH2Cl 2, -78 °C syn/anti = 93/7~>99/1 64~79% 95~97% ee (syn) diamine 2a Sn(OTf) 2 Bu 2Sn(OAc)2 [Bu3SnF] CH2Cl 2, -78 °C 52~87%

(101)

(102)

OH O R MeS

OEt

(103)

syn/anti = 91/9~99/1 90~95% ee (syn)

4.4.6

Enantioselective Synthesis of Both Enantiomers of Aldols Using Similar Diamines Derived from L-Proline

Kobayashi recently reported remarkable results in the synthesis of optically active aldol compounds using new chiral diamine–Sn(II) complexes as promoters. Reaction of KSA derived from S-Et (t-butyldimethylsiloxy)ethanethioate with aldehydes using chiral diamine 6 mainly yielded the syn-(2R,3S) compounds which are optical antipodes of aldol adducts (syn(2S,3R)) prepared by the reaction using chiral diamine 1c (Eqs. (89) and (104)) [70, 79]. Optically active syn-(2R,3R) aldols were also prepared from propionic acid derivatives and promotion with an Sn(II) complex with chiral diamine 7, whereas syn-(2S,3S) aldols were produced if chiral diamine 2a was used in the same reaction (Eqs. (75) and (105)) [80]. Chiral diamines 1c, 2a, 6, and 7 were all prepared starting from l-proline and have identical chirality at the C2 position. Artiﬁcial switching of the enantiofacial selectiv-

139

140

4 Tin-promoted Aldol Reactions and their Application to Total Syntheses of Natural Products

ity of the aldol reaction by using only one chiral source could therefore be achieved by use of these methods.

OTMS

O + R

H

SEt OTBS

OTMS

O + R

H

SEt

diamine 6 Sn(OTf) 2

OH O R

Bu 2Sn(OAc)2 CH2Cl 2, -78 °C 63~86%

(104)

syn/anti =>99/1 98~>99% ee (syn)

diamine 7 Sn(OTf) 2 Bu 2Sn(OAc)2 CH2Cl 2, -78 °C 67~83%

SEt OTBS

OH O R

SEt

(105)

syn/anti =>99/1 80~92% ee (syn)

4.5

Asymmetric Total Syntheses of Complex Molecules Using Chiral Diamine–Sn(II) Catalysts

Enantioselective aldol reactions can be powerful tools for the stereoselective synthesis of complex molecules, especially for construction of optically active 1,2-diol groups in the carbon backbones of the target compounds. Recent progress in this area will be illustrated by means of successful methods for stereoselective synthesis of natural and unnatural polyoxy compounds. 4.5.1

Monosaccharides

In the last decade chemical synthesis of monosaccharides has made a great advance as a result of stereoselective addition reactions of 2,3-Oisopropylidene-d- or -l-glyceraldehyde or 4-O-benzyl-2,3-O-isopropylidene-lthreose with enolate components or allyl nucleophiles, and many examples of the eﬀective synthesis of sugars, both natural and unnatural, have been demonstrated [81]. In these syntheses one of the starting materials, glyceraldehyde or a threose derivative, is prepared from a natural chiral pool, mannitol and tartaric acid, respectively. In contrast, a general method has been developed for synthesis of a variety of sugars starting from both achiral KSA and a,b-unsaturated aldehydes (Scheme 4.6). Chiral induction can be accomplished by means of an asymmetric aldol reaction using a complex consisting of Sn(OTf )2 and an appropriate chiral diamine. Subsequent dihydroxylation or epoxidation of the double bond in the aldol adducts aﬀords several tetrahydroxy thioester derivatives which can be useful precursors for the syntheses of a variety of monosaccharides, including rare sugars.

4.5 Asymmetric Total Syntheses of Complex Molecules Using Chiral Diamine–Sn(II) Catalysts

R2

OSiMe 3

O

R1

H

+

R3

N N R Sn TfO OTf

SEt

R2 R1

OP

R2

OH O or

SEt OTBS

R3

2,3-syn

R1

HO R2 OH O

SEt R3

HO R2 OH O 2,3-anti

R1

HO R2 OH O H R1

R1

HO R2 OH O

H R1 H HO R3 OTBS HO R3 OTBS

HO R2 OH O H

OBn

2,3-anti

HO R2 OH O

H R1 H HO R3 OTBS HO R3 OTBS

OH O

R1

2,3-syn HO R2 OH O

141

R1

HO R2 OH O H R1

H

HO R3 OBn

HO R3 OBn

HO R3 OBn

HO R3 OBn

4,5-syn-3,4-anti

4,5-anti-3,4-anti

4,5-anti-3,4-syn

4,5-syn-3,4-syn

Scheme 4.6

Synthesis of monosaccharides by use of the asymmetric aldol reaction.

One example, synthesis of 6-deoxy-l-talose, is shown in Scheme 4.7 [82]. The asymmetric aldol reaction between crotonaldehyde and the KSA of a-benzyloxy thioester was carried out in the presence of Sn(OTf )2 , chiral diamine 1a, and dibutyltin diacetate, and the corresponding aldol adduct was obtained in 85% yield with >97% enantiomeric excess. Dihydroxylation of this chiral synthon, subsequent reduction of the resulted lactone, and deprotection of the benzyl group gave the desired 6-deoxy-l-talose in good yield.

OSiMe 3

O H

+

N N Me Sn TfO OTf

OH O

SEt OBn

SEt Bu 2Sn(OAc)2 CH2Cl 2, -78 °C 85%

OBn syn/anti=2/>98 >97%ee (anti ) O HO

OH OH

OH 6-Deoxy- L-talose Scheme 4.7

Synthesis of 6-deoxy-l-talose.

4 Tin-promoted Aldol Reactions and their Application to Total Syntheses of Natural Products

142

O

OH

4-C-Methy-D-ribose H

O

HO

OBn

R

OH

O

HO

OH

SEt

OH

HO

HO

OH O

D-Ribose

R = H, Me HO

OH O

OH

OH

H

O

OH

SEt OBn

HO NHAc N-Acetyl-L -fucosamine

AcHN

H

HO

O

AcHN OH 3-Acetamido-3,6-dideoxy-L -idose

OH OH

5-Acetamido-5,6-dideoxy-D-allose Scheme 4.8

Synthesis of d-ribose, 4-C-methyl-d-ribose and several amino sugars.

By use of this universal methodology, several monosaccharides including branched and amino sugars were synthesized as shown in Scheme 4.8 (d-ribose and 4-C-methyl-d-ribose (1990) [82], N-acetyl-l-fucosamine, 3acetamide-3,6-dideoxy-l-idose and 5-acetamide-5,6-dideoxy-d-allose (1993) [83]). Scheme 4.9 shows the syntheses of two stereoisomers of 6-deoxy-l-talose from the corresponding intermediates generated via asymmetric aldol reaction (6-deoxy-d-allose (1992) [84] and l-fucose (1993) [76]). Several 2-branched saccharine acid g-lactones, 2-C-methyl-d- or l-threono1,4-lactones and 2-C-methyl-d-erythrono-1,4-lactone have been eﬀectively prepared using this strategy, by enantioselective construction of asymmetric quaternary carbons developed in the former section (Schemes 4.10 and 4.11) [72, 73, 69]. Because the key asymmetric aldol reaction has wide ﬂexibility in controlling newly created chiral centers, these methods are expected to provide useful routes to the synthesis of a variety of monosaccharides from achiral KSA and aldehydes. 4.5.2

Leinamycin and a Part of Rapamycin

Fukuyama used the asymmetric formation of a 1,2-diol group for the total synthesis of leinamycin in which it was shown that KSA with a p-methoxybenzyloxy group at the C2 position functions as a suitable nucleophile for the multifunctional aldehyde (Scheme 4.12) [85]. White also reported that

4.5 Asymmetric Total Syntheses of Complex Molecules Using Chiral Diamine–Sn(II) Catalysts

N N Me Sn TfO OTf

OTMS

O +

H

SEt OBn

Bu 2Sn(OAc)2 CH2Cl 2, -78 °C 82%

143

OH O SEt OBn syn/anti=6/94 92% ee (anti) O

OH

HO

OH OH

6-Deoxy-D-allose

N N Me Sn H TfO OTf OTMS

O

BnO

+ H

OH O

(0.2 eq)

OPh

OPh SnO (0.2 eq) EtCN, -78 °C 87% (after hydrolysis)

OBn syn/anti=97/3 92% ee (syn) O

OH

HO

OH OH L-Fucose

Scheme 4.9

Synthesis of 6-deoxy-d-allose and l-fucose.

OTMS

O TIPSO

+ H

N N Me Sn H TfO OTf

O iPr OBn E/Z=71/29

Bu 2Sn(OAc)2 CH2Cl 2, -78 °C 45%

OH O TIPSO

OiPr BnO

syn/anti=91/9 90% ee (syn) O

O

OH HO 2-C-Methyl- D-threono-1,4-lactone Scheme 4.10

Synthesis of 2-C-methyl-d-threono-1,4-lactone.

144

4 Tin-promoted Aldol Reactions and their Application to Total Syntheses of Natural Products

N N Pr Sn TfO OTf

OTMS

O + CO 2Me

MeO2C OH O

SEt

SEt Bu3SnF CH2Cl 2, -78 °C 93%

OBn

OBn syn/anti=13/87 91% ee (anti ) O

O

HO OH 2-C-Methyl-L -threono-1,4-lactone

N N Pr Sn TfO OTf

OTMS

O + CO 2Me

SEt OTBS

MeO2C OH O SEt OTBS syn/anti=94/6 88% ee (syn)

Bu3SnF CH2Cl 2, -78 °C 89%

O HO

O

OH

2-C-Methyl-D-erythrono-1,4-lactone Scheme 4.11

Synthesis of 2-C-methyl-l-threono- and d-erythrono-1,4-lactones.

tBu

O

O

O

OTMS

H +

O O O

OMe O

N N Me Sn TfO OTf

SEt OPMB

Bu 2Sn(OAc)2 CH2Cl 2, -78 °C 92%

tBu

O

OH O

O

SEt OPMB

O O

OMe

O

O HO

H O O S

H N

S O O

Leinamycin Scheme 4.12

Synthesis of leinamycin.

O

N S

4.5 Asymmetric Total Syntheses of Complex Molecules Using Chiral Diamine–Sn(II) Catalysts

OTMS

OH O N N Me Sn TfO OTf

SEt O

O

SEt O

+

H

Bu 2Sn(OAc)2 CH2Cl 2, -78 °C 80%

OMe OMe

OMe OMe syn/anti =5/95 92% ee (anti )

OTMS

OTBS N N Me Sn TfO OTf

SEt OTBS

O

O H

OH O SEt O

+ Bu 2Sn(OAc)2 CH2Cl 2, -78 °C 80%

OMe OMe

HO O N

OMe OMe syn/anti =8/92 H

O O MeO

O MeO HO

H

O OH O O OMe Rapamycin

Scheme 4.13

Synthesis of a part of rapamycin.

reaction of the KSA generated from S-Et (3,4-dimethoxybenzyloxy)ethanethioate with a,b-unsaturated aldehydes proceeded smoothly to aﬀord the corresponding diol groups in high yields with excellent stereoselectivity, as shown in Scheme 4.13 [86]. 4.5.3

Sphingosine, Sphingofungins, and Khafrefungin

Kobayashi used asymmetric reactions for stereoselective synthesis of a variety of polyoxygenated natural compounds. Initially, a new method for the preparation of sphingosine was developed using the catalytic asymmetric aldol reaction of KSA with a,b-eynal as a key step (Scheme 4.14) [87].

145

146

4 Tin-promoted Aldol Reactions and their Application to Total Syntheses of Natural Products

N N Me Sn H TfO OTf O H

+

BnO

OPh

TMS

OH O

(0.2 eq)

OTMS

OPh SnO (0.2 eq) TMS OBn EtCN, -78 °C syn/anti =97/3 87% 91% ee (syn) (after hydrolysis) OH C13H 27

OH NH2

Sphingosine Scheme 4.14

Synthesis of sphingosine.

Sphingofungins B and F were also totally synthesized from small molecules by the asymmetric aldol strategy as shown in Scheme 4.15 [87b, 88]. Here the optically active polyol part was obtained by reaction of trisubstituted KSA using a chiral diamine ent-2a, and the sole asymmetric center in the side chain was synthesized by the reaction of KSA derived from an acetic acid derivative using the chiral diamine 2a. These segments were coupled to form the basic skeleton of sphingofungins in the total synthesis. A diastereoselective aldol reaction using an a-alkoxyaldehyde was also mentioned in this research (Eq. (106)) [88c]. TBSO

O

C13H 27

OTMS H

+

OBn

SEt TBSO

SnCl 4 C13H 27 CH2Cl 2, -78 ˚C 82% ds 100/0

ð106Þ

OH O SEt OBn

Total synthesis of khafrefungin and the determination of its stereochemistry was recently achieved by Kobayashi, who used chiral induction technology to give the optically active aldol compounds (Scheme 4.16) [89]. The asymmetric aldol reaction of KSA derived from S-Et propanethioate with aldehydes was applied not only to the ﬁrst step to aﬀord the corresponding thioester with high ee but also to the following stage to give the multifunctional linear thioester with excellent diastereoselectivity.

4.5 Asymmetric Total Syntheses of Complex Molecules Using Chiral Diamine–Sn(II) Catalysts

N N Me Sn H TfO OTf O H

+

BnO

OPh

TMS

OH O

(0.2 eq)

OTMS

OPh SnO (0.2 eq) TMS OBn CH 3CN, -78 °C syn/anti =97/3 87% 91% ee (syn) (after hydrolysis)

N N Me Sn H TfO OTf C6H 13CHO

+

(0.2 eq)

OTMS SEt

OH O

SnO (0.2 eq) CH2Cl 2, -78 °C 87% (after hydrolysis)

HO C6H 13

C6H 13

SEt

94% ee

OH OH O (CH2)6

OH HO

NH2

Sphingofungin B O C6H 13

OH OH O (CH2)6 HO

OH NH2

Sphingofungin F Scheme 4.15

Synthesis of sphingofungins B and F.

4.5.4

Febrifugine and Isofebrifugine

Kobayashi also reported the enantioselective total synthesis of febrifugine and isofebrifugine using the Sn(II)-mediated catalytic asymmetric aldol reaction giving the optically active diol groups (Scheme 4.17) [90]. The correct absolute stereochemistries of natural febrifugine and isofebrifugine were shown by comparison with spectral data and the sense of the optical rotations of four synthetic samples, including enantiomorphs.

147

148

4 Tin-promoted Aldol Reactions and their Application to Total Syntheses of Natural Products

N N Me Sn H TfO OTf OTMS C9H 19CHO

+

SEt

OPMB

O

C10H 21

H

OH O

(0.2 eq) SnO (0.2 eq) CH2Cl 2, -78 °C 83% (after hydrolysis)

C9H 19

SEt

syn/anti =97/3 94% ee (syn)

N N Me Sn H TfO OTf

OPMB

OH O

C10H 21 OTMS +

SEt

SEt

Bu 2Sn(OAc)2 CH2Cl 2, -78 °C 90% >98% ds OH

O

C10H 21

OH OH

O

OH

O OH O Khafrefungin

Scheme 4.16

Synthesis of khafrefungin.

4.5.5

Altohyrtin C (Spongistatin 2) and Phorboxazole B

Asymmetric aldol reaction accelerated by the chiral bis(oxazoline)– Sn(OTf )2 complex also provides a powerful means of construction of polyfunctionalized natural compounds. Evans succeeded in the total synthesis of altohyrtin C (spongistatin 2), a macrocyclic compound with many oxygenated functional groups (Scheme 4.18) [91]. Part of the tetrahydropyran segment (F ring) in altohyrtin C was stereoselectively obtained from the corresponding anti-b-hydroxy-a-methyl thioester generated by the asymmetric aldol reaction using the Ph/Box–Sn(II) complex catalyst. This asymmetric aldol reaction is eﬀective when using chelating electrophiles such as ethyl glyoxylate; therefore, a-oxazole aldehyde might be employed in the preparation of an optically active oxazole derivative. Indeed, the reaction of KSA generated from S- t Bu ethanethioate with a-oxazole al-

4.5 Asymmetric Total Syntheses of Complex Molecules Using Chiral Diamine–Sn(II) Catalysts

149

N N Me Sn H TfO OTf O TBSO

H

OPh

OH O

(0.2 eq)

OTMS

+ BnO

TBSO SnO (0.2 eq) EtCN, -78 °C 70% (after hydrolysis)

OPh OBn

syn/anti =95/5 >96% ee (syn) O TBSO

H OBn

OH O N H

N

O

N

N O Febrifugine

N

H O

N H

O Isofebrifugine

Scheme 4.17

Synthesis of febrifugine and isofebrifugine.

dehyde took place as expected to aﬀord the corresponding aldol with high ee, and the total synthesis of phorboxazole B was successfully achieved using the adduct as a part of the complex structure (Scheme 4.19) [92]. 4.5.6

Paclitaxel (Taxol)

Mukaiyama and Shiina accomplished the total synthesis of paclitaxel (taxol) by the strategy shown in Scheme 4.20, i.e. synthesis of the eight-membered B ring ﬁrst, starting from an optically active polyoxy precursor generated by the highly controlled enantioselective aldol reaction and subsequent construction of the fused A and C ring systems on to the B ring [93]. The optically active diol unit 9 was prepared by the asymmetric aldol reaction of a KSA with a benzyloxy group at the C2 position with an achiral aldehyde 8 using a chiral diamine–Sn(II) complex (Scheme 4.21). Synthesis of the eight-membered ring aldols from an optically active polyoxy-group 10 containing all the functionality necessary for the construction of taxol was performed by the intramolecular aldol cyclization using SmI2 . Subsequent acetylation of this mixture of isomeric alcohols and treatment with DBU gave the desired eight-membered enone 11 in good yield.

150

4 Tin-promoted Aldol Reactions and their Application to Total Syntheses of Natural Products

O

O N

Ph O EtO

OTMS H

+

OH O EtO

SPh

CH2Cl 2, -78 °C 97% (after hydrolysis)

O

TESO EtO

N Sn TfO OTf Ph (0.1 eq)

SPh O syn/anti=4/96 94% ee (anti ) O H O

OH O SEt

OMe

O

OTES OH

HO H O

O OH H O

F HO

O HO

HO

OMe

O

O

O O

O

AcO

OAc

OH Altohyrtin C (Spongistatin 2) Scheme 4.18

Synthesis of altohyrtin C (spongistatin 2).

As shown in Scheme 4.22, fully functionalized BC ring system 12 was then synthesized from the optically active eight-membered ring compound 11 by successive Michael addition and intramolecular aldol cyclization of ketoaldehyde. Intramolecular pinacol coupling of the diketone derived from the above BC ring system using a low-valent titanium reagent resulted in the formation of ABC ring system 13, a new taxoid, in good yield. 7-Triethylsilylbaccatin III was prepared from the above new taxoid 13 by oxygenation at the C13 position and construction of the oxetane ring. It was also shown that the asymmetric aldol reaction is useful for preparation of the chiral side chains of taxol (Scheme 4.23). Because reaction of the KSA derived from S-Et benzyloxyethanethioate with benzaldehyde aﬀorded the corresponding aldol adduct 14 in high yield with excellent selectivity, as shown in the last section, this adduct was successfully con-

4.5 Asymmetric Total Syntheses of Complex Molecules Using Chiral Diamine–Sn(II) Catalysts

O

O N

O H

O N

OTMS

+

StBu

Ph

N Sn Ph TfO OTf (0.1 eq)

OH O

Ph

StBu

O N

CH2Cl 2, -78 °C 91% (after hydrolysis)

Ph

94% ee

OH

O O

N

O

Br O

OMe O

HO

O

O

OH N O

OMe Phorboxazole B Scheme 4.19

Synthesis of phorboxazole B.

AcO

O

OH

B RO

A HO

TBS O O

BnO

C

H O O Ac Bz

PMBO

OBn

Scheme 4.20

Retrosynthesis of taxol.

OBn

OTBS OTBS

BnO

B

DO

OPMB

MeO

O

TBSO

+

MeO BnO

O H

OTBS

O H

+

OMe OBn

151

152

4 Tin-promoted Aldol Reactions and their Application to Total Syntheses of Natural Products

MeO MeO

N N Me Sn TfO OTf

OTBS

O +

H

MeO

OMe

MeO

OMe OBn syn/anti =20/80 9; 93% ee (anti )

Bu 2Sn(OAc)2 CH2Cl 2, -23 °C 68%

OBn 8

OH O

BnO

TBS O O

OTBS Br BnO

TBSO

1) SmI2 (70%)

OPMB

OBn 10

O

2) Ac2 O (87%) 3) DBU (91%)

PMBO

OBn 11

Scheme 4.21

Synthesis of the B ring of taxol.

BnO

O

BnO

TBSO

O

AcO

OH

O

OTES

TBSO

PMBO

OBn

PMBO

11

H O

O

OBn 12

O

Scheme 4.22

Synthesis of the ABC ring system of taxol.

O

OTMS H

+

N N Et Sn TfO OTf

OH O SEt

SEt Bu 2Sn(OAc)2 CH2Cl 2, -78 °C 96%

OBn

BzHN

OBn syn/anti=1/99 14; 96% ee (anti) Ph

O SEt OBn

Scheme 4.23

Synthesis of the side chain of taxol.

OH

BzN O PMP

O

15

13

4.5 Asymmetric Total Syntheses of Complex Molecules Using Chiral Diamine–Sn(II) Catalysts

AcO

HO HO

O

OTES

HO O H Bz

O

7-Triethylsilylbaccatin III

1) 15, DPTC DMAP (95%) 2) TFA (93%)

AcO BzHN

O

OH

O

Ph

O OH

HO

HO O H Bz

O

Paclitaxel (Taxol ®)

Scheme 4.24

Synthesis of taxol.

verted into the targeted b-amino acid 15 in good yield, with inversion of chirality at the b-position, by use of the Mitsunobu reaction. Introduction of N-benzoylphenylisoserine 15 to 7-triethylsilylbaccatin III was further studied, and dehydration condensation was found to proceed smoothly using DPTC (O,O-di(2-pyridyl)thiocarbonate) as a novel coupling reagent in the presence of DMAP to aﬀord the desired ester in 95% yield at 93% conversion (Scheme 4.24) [93e, 94]. Finally, deprotection of the intermediate gave the ﬁnal target molecule taxol in excellent yield. This established a new method for asymmetric synthesis of baccatin III by way of B to BC to ABC to ABCD ring construction and completion of the total synthesis of taxol by preparation of the side chain by asymmetric aldol reaction and subsequent dehydration condensation with 7-TES baccatin III using DPTC. This synthetic route would be widely applicable to the preparation of a variety of derivatives of taxol and related taxoids. 4.5.7

Cephalosporolide D

Shiina developed a method for preparation of cephalosporolide D, a natural eight-membered ring lactone, and the exact stereochemistry of this compound was determined through the ﬁrst total synthesis (Scheme 4.25) [95]. In this synthetic strategy two asymmetric carbon atoms were constructed by the asymmetric aldol reaction using the KSA derived from S-Et ethanethioate. It is also mentioned that the second diastereoselective aldol reaction aﬀorded the desired compound in 3:97 ratio when using the chiral diamine ent-2a–Sn(II) complex and that the ratio ranges from 97:3 to 59:41 when the 2a–Sn(II) complex or SnCl 4 was used as a catalyst. The desired eightmembered ring lactone moiety was constructed by cyclization of the seco acid via a novel mixed-anhydride method using (4-triﬂuoromethyl)benzoic anhydride (TFBA) with Hf(OTf )4 [96]. 4.5.8

Buergerinin F

The synthesis of buergerinin F, a natural compound consisting of a unique tricyclic skeleton, was achieved in the course of synthetic studies by Shiina

153

154

4 Tin-promoted Aldol Reactions and their Application to Total Syntheses of Natural Products

N N Me Sn H TfO OTf

OTMS

O

OH O

+ SEt

H

OTBS

O

SEt 96%ee

Bu3SnF CH2Cl 2, -95 °C 78%

N N Me Sn H TfO OTf

OTMS

OTBS

OH O

+ H

OH

SEt

Bu3SnF CH2Cl 2, -78 °C 89%

TFBA Hf(OTf) 4

OBn O OH

SEt ent-2a-Sn(OTf)2 ; syn/anti =3/97 2a-Sn(OTf)2; 62%, syn/anti =97/3 SnCl4; 62%, syn/anti =59/41

H2, Pd/C

O O

OH

MeCN reflux Chephalosporolide D

Scheme 4.25

Synthesis of cephalosporolide D.

on the utilization of the asymmetric aldol strategy [97]. The ﬁrst key step is producing the optically active a,b,g 0 -trioxy ester including an asymmetric quaternary carbon at the C2 position as shown in Scheme 4.26. It was also revealed that enantioselective aldol reaction of tetrasubstituted KSA with four oxygenated functional groups is quite eﬀective for preparation of this complex synthetic intermediate. Successive intramolecular Wacker-type ketalization and one-carbon elongation of the intermediate aﬀorded the optically active buergerinin F. On completion of the total synthesis using the asymmetric aldol reaction promoted by chiral diamine–Sn(OTf )2 as catalyst, the absolute stereochemistry of natural buergerinin F was determined. 4.5.9

Octalactins A and B

Shiina recently developed a new method for synthesis of octalactin A, an antitumor agent consisting of an eight-membered ring lactone (Scheme 4.29) [98]. Because the lactone moiety includes two pairs of anti-b-hydroxya-methyl groups, enantioselective addition of the KSA derived from ethyl 2methylthiopropanoate was eﬃciently used for construction of the required

4.5 Asymmetric Total Syntheses of Complex Molecules Using Chiral Diamine–Sn(II) Catalysts

OTMS

O

+

TBSO

H

OMe OBn 84/16

N N Me Sn H TfO OTf

O TBSO BnO

Bu 2Sn(OAc)2 EtCN, -78 °C 62%

OAc

HO H

OSiMe 2CMe 2Ph

OMe

OH H syn/anti =98/2 93% ee (syn)

PdCl2

HO

155

OAc

O O CuCl, O2

H

OSiMe 2CMe 2Ph

O O O H Buergerinin F Scheme 4.26

Synthesis of buergerinin F.

components, i.e. asymmetric aldol reaction of the tetrasubstituted KSA with aldehydes [78] and subsequent treatment of the optically active adducts formed with Guidon’s reduction [99] aﬀorded the desired two chiral segments 16 and 17 (Scheme 4.27). The optically active side chain 18 was also produced by means of the asymmetric aldol reaction of the KSA derived from S-Et ethanethioate with 2-methylpropionaldehyde (Scheme 4.28). A chiral linear precursor having repeated anti-b-hydroxy-a-methyl units was obtained by coupling segments 16 and 17, and the resulting seco acid was eventually cyclized to form the eight-membered ring lactone by a new quite eﬀective mixed-anhydride method using 2-methyl-6-nitrobenzoic anhydride (MNBA) with DMAP, as shown in Scheme 4.29 [100, 98b]. Finally, the side chain 18 was introduced to the eight-membered ring lactone moiety to aﬀord the targeted multioxygenated compounds, octalactins A and B. 4.5.10

Oudemansin-antibiotic Analog

Uchiro and Kobayashi recently reported the synthesis of b-methoxyacrylate antibiotics (MOA) and their analogs (Scheme 4.30) [101]. In accordance with their strategy for preparation of the related compounds, asymmetric aldol reaction of the KSA generated from S-Et propanethioate with cinnamyl aldehyde was used for stereoselective synthesis of the intermediate of an MOA analog, as shown in Scheme 4.30.

156

4 Tin-promoted Aldol Reactions and their Application to Total Syntheses of Natural Products

O +

MeS

OEt

H

TIPSO

N N Me Sn H TfO OTf

OTMS

E/Z=12/88

Bu3SnF CH2Cl 2, -78 °C 56%

OH O OEt

TIPSO MeS syn/anti =96/4 87% ee (syn) OBn TBSO

PPh 3I 16

O +

TIPSO

N N Me Sn H TfO OTf

OTMS MeS

OEt

H

E/Z=12/88

Bu3SnF CH2Cl 2, -78 °C 50%

OH O TIPSO

OEt SMe

syn/anti =87/13 69% ee (syn) PMP O

O

H O 17 Scheme 4.27

Synthesis of two chiral segments of octalactins A and B.

N N Me Sn H TfO OTf O

OTMS H

(0.2 eq.)

OH O

+ SEt

EtCN, -78 °C 48% (after hydrolysis)

SEt 90% ee OTBS I 18

Scheme 4.28

Synthesis of the side chain of octalactins A and B.

4.6 Conclusions

157

O O

OH

OBn

HO

MNBA OTBDPS

DMAP CH2 Cl2 rt, 13 h

TBDPSO

O

O 18 OH

O

OBn

O

O

OH

OH

O

O

O Octalactin A

Octalactin B Scheme 4.29

Synthesis of octalactin A and B.

4.6

Conclusions

In this chapter a variety of Sn(IV) or Sn(II) metallic species-promoted aldol reactions have been presented, with their application in syntheses of complicated molecules with high stereoselectivity. Alkoxy aldehydes were eﬀectively activated by SnCl 4 , and reactions with particular ESE or KSA are highly applicable to the generation of 3,4-syn aldol compounds with high diastereoselectivity. Intramolecular reaction of ESE and KSA with an acetal moiety is also quite attractive for preparation of medium-sized compounds which are generally not available by other methods. Sn(II)-promoted asymmetric aldol reaction could be now used as a general and powerful method for the construction of not only optically active small molecules but highly

OTMS

O + Ph

H

SEt

N N Me Sn H TfO OTf Bu 2Sn(OAc)2 CH2Cl 2, -78 °C 92%

OH O Ph

SEt 98% ee OMe O MeO

OMe

Oudemansin-type Analog Scheme 4.30

Synthesis of oudemansin-antibiotic analog.

OH

158

4 Tin-promoted Aldol Reactions and their Application to Total Syntheses of Natural Products

advanced multifunctional compounds. Progress in aldol reactions using tin reagents has contributed greatly to the syntheses of many useful substrates in the last decade, and this fruitful history might provide valuable information to organic and organometallic chemistry in the future.

4.7

Experimental Typical Procedure for Catalytic Asymmetric Aldol Reaction of a KSA with Simple Achiral Aldehydes (Eqs. (80) and (81)) [61d]. A solution of 2a (21.1 mg, 0.088 mmol) in EtCN (1 mL) was added to a solution of Sn(OTf )2 (33.4 mg, 0.080 mmol, 20 mol%) in EtCN (1 mL). The mixture was cooled to 78 C and a mixture of KSA (0.40 mmol) and an aldehyde (0.40 mmol) in EtCN (1.5 mL) was then added slowly over 3–4.5 h by means of a mechanical syringe. The mixture was further stirred for 2 h, and then quenched with saturated aqueous NaHCO3 . The organic layer was isolated and the aqueous layer was extracted with CH2 Cl2 (three times). The organic solutions were combined, washed with H2 O and brine, then dried over Na2 SO4 . After evaporation of the solvent the crude product was puriﬁed by preparative TLC on silica gel to aﬀord an aldol-type adduct as the corresponding trimethylsilyl ether. The trimethylsilyl ether was treated with THF–1 m HCl (20:1) at 0 C to give the corresponding alcohol. Typical Procedure for Catalytic Asymmetric Aldol Reaction of a KSA with Ethyl Glyoxylate (Eqs. (84) and (86)) [65a]. (S,S)-bis(Benzyloxazoline) (19.9 mg, 0.055 mmol) and Sn(OTf )2 (20.8 mg, 0.050 mmol) were placed, within an inert atmosphere box, in an oven-dried 8-mL vial containing a magnetic stirring bar. The ﬂask was ﬁtted with a serum cap, removed from the inert atmosphere box, and charged with CH2 Cl2 (0.8 mL). The resulting suspension was stirred rapidly for 1 h to give a cloudy solution. The catalyst was cooled to 78 C and the KSA (0.50 mmol) was added followed by distilled ethyl glyoxylate–toluene solution (8:2 mixture, 100 mL, 0.75 mmol). The resulting solution was stirred at 78 C until the KSA was completely consumed (0.1–2 h), as determined by TLC. The reaction mixture was then ﬁltered through a 0.3 cm 5 cm plug of silica gel, which was then washed with Et2 O (8 mL). Concentration of the ether solution gave the crude silyl ether which was dissolved in THF (2 mL) and 1 m HCl (0.2 mL). After standing at room temperature for 0.5 h this solution was poured into a separatory funnel and diluted with Et2 O (20 mL) and H2 O (10 mL). After mixing the aqueous layer was discarded and the ether layer was washed with saturated aqueous NaHCO3 (10 mL) and brine (10 mL). The resulting ether layer was dried over anhydrous Na2 SO4 , ﬁltered, and concentrated to furnish the hydroxy esters. Puriﬁcation by ﬂash chromatography provided the desired aldols.

References

Typical Procedure for Synthesis of an Optically Active Diol Group by Asymmetric Aldol Reaction (Eq. (88)) [67b, 93f ]. A solution of 1a or 1b (0.405 mmol) in CH2 Cl2 (0.5 mL), then a solution of Bu2 Sn(OAc)2 (131.1 mg, 0.373 mmol) in CH2 Cl2 (0.5 mL), were added successively to a suspension of Sn(OTf )2 (141.8 mg, 0.340 mmol) in CH2 Cl2 (1 mL). The mixture was stirred for 30 min at room temperature, then cooled to 78 C. A solution of 2-benzyloxy-1-ethylthio-1-(trimethylsiloxy)ethene (96.1 mg, 0.340 mmol) (Z=E ¼ 9:1, the E isomer has no reactivity) in CH2 Cl2 (0.5 mL), and a solution of aldehyde (0.228 mmol) in CH2 Cl2 (0.5 mL) at 78 C, were then added successively to the reaction mixture. The mixture was further stirred for 20 h then quenched with saturated aqueous NaHCO3 . The organic layer was isolated and the aqueous layer was extracted with CH2 Cl2 (three times). The organic solutions were combined, washed with H2 O and brine, then dried over Na2 SO4 . After evaporation of the solvent the crude product was puriﬁed by preparative TLC on silica gel to aﬀord the corresponding antia,b-dihydroxy thioester derivatives. References 1 (a) Mukaiyama, T.; Narasaka, K.; Banno, K. Chem. Lett. 1973,

2 3

4 5

6

7

8 9

1011. (b) Mukaiyama, T.; Banno, K.; Narasaka, K. J. Am. Chem. Soc. 1974, 96, 7503. (c) Saigo, K.; Osaki, M.; Mukaiyama, T. Chem. Lett. 1975, 989. (d) Mukaiyama, T. Org. React. 1982, 28, 203. (e) Mukaiyama, T.; Narasaka, K. Org. Synth. 1987, 65, 6. Wissner, A. Synthesis 1979, 27. (a) Lozzi, L.; Ricci, A.; Taddei, M. J. Org. Chem. 1984, 49, 3408. (b) Brownbridge, P.; Chan, T.-H. Tetrahedron Lett. 1980, 21, 3427. (a) Nakamura, E.; Kuwajima, I. Chem. Lett. 1983, 59. (b) Nakamura, E.; Kuwajima, I. Tetrahedron Lett. 1983, 24, 3347. (a) Kuwajima, I.; Inoue, T.; Sato, T. Tetrahedron Lett. 1978, 4887. (b) Inoue, T.; Sato, T.; Kuwajima, I. J. Org. Chem. 1984, 49, 4671. (a) Ponomarev, S. V.; Bankov, Y. I.; Dukukina, O. V.; Petrosyan, I. V.; Petrovskaya, L. I. J. Gen. Chem. USSR 1967, 37, 2092. (b) Noltes, J. G.; Creemers, H. M. J. C.; Van Der Kerk, G. J. M. J. Organomet. Chem. 1968, 11, 21. (c) Pereyre, M.; Bellegarde, B.; Mendelsohn, J.; Valade, J. J. Organomet. Chem. 1968, 11, 97. (d) Pereyre, M.; Quintard, J.-P.; Rahm, A. Tin in Organic Synthesis, Butterworths, London, 1987. (a) Yasuda, M.; Katoh, Y.; Shibata, I.; Baba, A.; Matsuda, H.; Sonoda, N. J. Org. Chem. 1994, 59, 4386. (b) Yasuda, M.; Hayashi, K.; Katoh, Y.; Shibata, I.; Baba, A. J. Am. Chem. Soc. 1998, 120, 715. Yamamoto, Y.; Yatagai, H.; Maruyama, K. J. Chem. Soc. Chem. Commun. 1981, 162. (a) Shenvi, S.; Stille, J. K. Tetrahedron Lett. 1982, 23, 627. (b) Labadie, S. S.; Stille, J. K. Tetrahedron 1984, 40, 2329.

159

160

4 Tin-promoted Aldol Reactions and their Application to Total Syntheses of Natural Products 10 (a) Mukaiyama, T.; Stevens, R. W.; Iwasawa, N. Chem. Lett.

11

12

13

14 15

16

17 18

19

20

21

22 23 24

1982, 353. (b) Mukaiyama, T.; Iwasawa, N. Chem. Lett. 1982, 1903. (c) Mukaiyama, T.; Iwasawa, N.; Stevens, R. W.; Haga, T. Tetrahedron 1984, 40, 1381. (a) Heathcock, C. H.; Hug, K. T.; Flippin, L. A. Tetrahedron Lett. 1984, 25, 5973. (b) Heathcock, C. H.; Davidsen, S. K.; Hug, K. T.; Flippin, L. A. J. Org. Chem. 1986, 51, 3027. (c) Heathcock, C. H.; Montgomery, S. H. Tetrahedron Lett. 1985, 26, 1001. (d) Montgomery, S. H.; Pirrung, M. C.; Heathcock, C. H. Carbohydrate Res. 1990, 202, 13. (a) Reetz, M. T.; Kesseler, K.; Jung, A. Tetrahedron Lett. 1984, 25, 729. (b) Reetz, M. T.; Kesseler, K.; Jung, A. Tetrahedron 1984, 40, 4327. (c) Reetz, M. T.; Kesseler, K.; Schmidtberger, S.; Wenderoth, B.; Steinbach, R. Angew. Chem. Int. Ed. Engl. 1983, 22, 989. (d) Reetz, M. T. Angew. Chem. Int. Ed. Engl. 1984, 23, 556. (e) Reetz, M. T.; Kesseler, K. J. Org. Chem. 1985, 50, 5434. (f ) Mahrwald, R. Chem. Rev. 1999, 99, 1095. (a) Gennari, C.; Bernardi, A.; Poli, G.; Scolastico, C. Tetrahedron Lett. 1985, 26, 2373. (b) Gennari, C.; Beretta, M. G.; Bernardi, A.; Moro, G.; Scolastico, C.; Todeschini, R. Tetrahedron 1986, 42, 893. (c) Gennari, C.; Cozzi, P. G. Tetrahedron 1988, 44, 5965. (d) Gennari, C. Comprehensive Organic Synthesis, Pergamon Press, 1991, Vol. 2, 629. Uenishi, J.; Tomozane, H.; Yamato, M. Tetrahedron Lett. 1985, 26, 3467. (a) Bernardi, A.; Cardani, S.; Gennari, C.; Poli, G.; Scolastico, C. Tetrahedron Lett. 1985, 26, 6509. (b) Bernardi, A.; Cardani, S.; Colombo, L.; Poli, G.; Schimperna, G.; Scolastico, C. J. Org. Chem. 1987, 52, 888. (a) Guanti, G.; Banﬁ, L.; Narisano, E.; Scolastico, C. Tetrahedron Lett. 1985, 26, 3517. (b) Banﬁ, L.; Cardani, S.; Potenza, D.; Scolastico, C. Tetrahedron 1987, 43, 2317. Mikami, K.; Kaneko, M.; Loh, T.-P.; Terada, M.; Nakai, T. Tetrahedron Lett. 1990, 31, 3909. (a) Annunziata, R.; Cinquini, M.; Cozzi, F.; Cozzi, P. G. Tetrahedron Lett. 1990, 31, 6733. (b) Annunziata, R.; Cinquini, M.; Cozzi, F.; Cozzi, P. G.; Consolandi, E. J. Org. Chem. 1992, 57, 456. (a) Ewing, W. R.; Harris, B. D.; Li, W.-R.; Joullie´, M. M. Tetrahedron Lett. 1989, 30, 3757. (b) Li, W.-R.; Ewing, W. R.; Harris, B. D.; Joullie´, M. M. J. Am. Chem. Soc. 1990, 112, 7659. (a) Danishefsky, S.; Barbachyn, M. J. Am. Chem. Soc. 1985, 107, 7761. (b) Danishefsky, S. J.; DeNinno, S. L.; Chen, S.; Boisvert, L.; Barbachyn, M. J. Am. Chem. Soc. 1989, 111, 5810. (a) Cox, P.; Mahon, M. F.; Molloy, K. C.; Lister, S.; Gallagher, T. Tetrahedron Lett. 1989, 30, 2437. (b) Cox, P.; Lister, S.; Gallagher, T. J. Chem. Soc. Perkin Trans. 1 1990, 3151. Akiyama, T.; Ishikawa, K.; Ozaki, S. Synlett 1994, 275. Mukai, C.; Miyakawa, M.; Hanaoka, M. J. Chem. Soc. Perkin Trans. 1 1997, 913. (a) Mukai, C.; Kataoka, O.; Hanaoka, M. Tetrahedron Lett.

References

25 26 27 28 29 30

31 32 33

34 35 36

37

38

39

40

1994, 35, 6899. (b) Mukai, C.; Kataoka, O.; Hanaoka, M. J. Org. Chem. 1995, 60, 5910. Boeckman Jr., R. K.; Clark, T. J.; Shook, B. C. Org. Lett. 2002, 4, 2109. Mukai, C.; Moharram, S. M.; Kataoka, O.; Hanaoka, M. J. Chem. Soc. Perkin Trans. 1 1995, 2849. Liu, W.; Szewczyk, J. M.; Waykole, L.; Repic, O.; Blacklock, T. J. Tetrahedron Lett. 2002, 43, 1373. Mukaiyama, T.; Hayashi, M. Chem. Lett. 1974, 15. Muratake, H.; Natsume, M. Heterocycles 1990, 31, 691. Linderman, R. J.; Graves, D. M.; Kwochka, W. R.; Ghannam, A. F.; Anklekar, T. V. J. Am. Chem. Soc. 1990, 112, 7438. Peng, S.; Qing, F.-L. J. Fluorine Chem. 2000, 103, 135. Cossy, J.; Rakotoarisoa, H.; Kahn, P.; Desmurs, J.-R. Tetrahedron Lett. 2000, 41, 7203. (a) Pearlman, B. A.; McNamara, J. M.; Hasan, I.; Hatakeyama, S.; Sekizaki, H.; Kishi, Y. J. Am. Chem. Soc. 1981, 103, 4248. (b) McNamara, J. M.; Kishi, Y. J. Am. Chem. Soc. 1982, 104, 7371. (c) Sekizaki, H.; Jung, M.; McNamara, J. M.; Kishi, Y. J. Am. Chem. Soc. 1982, 104, 7372. Cambie, R. C.; Larsen, D. S.; Rickard, C. E. F.; Rutledge, P. S.; Woodgate, P. D. Aust. J. Chem. 1986, 39, 487. Cambie, R. C.; Rutledge, P. S.; Watson, P. A.; Woodgate, P. D. Aust. J. Chem. 1989, 42, 1939. (a) Ishihara, K.; Nakamura, H.; Yamamoto, H. J. Am. Chem. Soc. 1999, 121, 7720. (b) Nakamura, H.; Ishihara, K.; Yamamoto, H. J. Org. Chem. 2002, 67, 5124. (a) Otera, J.; Chen, J. Synlett 1996, 321. (b) Chen, J.; Sakamoto, K.; Orita, A.; Otera, J. Synlett 1996, 877. (c) Chen, J.; Otera, J. Synlett 1997, 29. (d) Chen, J.; Otera, J. Tetrahedron 1997, 53, 14275. (e) Chen, J.; Otera, J. Angew. Chem. Int. Ed. Engl. 1998, 37, 91. (f ) Chen, J.; Otera, J. Tetrahedron Lett. 1998, 39, 1767. (g) Chen, J.; Sakamoto, K.; Orita, A.; Otera, J. Tetrahedron 1998, 54, 8411. (h) Shirakawa, S.; Maruoka, K. Tetrahedron Lett. 2002, 43, 1469. (a) Asaoka, M.; Sugimura, N.; Takei, H. Bull. Chem. Soc. Jpn. 1979, 52, 1953. (b) Asaoka, M.; Yanagida, N.; Ishibashi, K.; Takei, H. Tetrahedron Lett. 1981, 22, 4269. (a) Casiraghi, G.; Rassu, G.; Spanu, P.; Pinna, L. J. Org. Chem. 1992, 57, 3760. (b) Rassu, G.; Casiraghi, G.; Spanu, P.; Pinna, L.; Fava, G. G.; Ferrari, M. B.; Pelosi, G. Tetrahedron: Asymmetry 1992, 3, 1035. (c) Casiraghi, G.; Rassu, G.; Spanu, P.; Pinna, L. Tetrahedron Lett. 1994, 35, 2423. (d) Rassu, G.; Zanardi, F.; Cornia, M.; Casiraghi, G. J. Chem. Soc. Perkin. Trans. 1 1994, 2431. (e) Zanardi, F.; Battistini, L.; Rassu, G.; Cornia, M.; Casiraghi, G. J. Chem. Soc. Perkin. Trans. 1 1995, 2471. (f ) Rassu, G.; Auzzas, L.; Pinna, L.; Battistini, L.; Zanardi, F. J. Org. Chem. 2000, 65, 6307. (g) Rassu, G.; Auzzas, L.; Pinna, L.; Zambrano, V.; Zanardi, F.; Battistini, L.; Marzocchi, L.; Acquotti, D.; Casiraghi, G. J. Org. Chem. 2002, 67, 5338. Uno, H.; Baldwin, J. E.; Russell, A. T. J. Am. Chem. Soc. 1994, 116, 2139.

161

162

4 Tin-promoted Aldol Reactions and their Application to Total Syntheses of Natural Products 41 (a) Cockerill, G. S.; Kocienski, P. J. Chem. Soc. Chem.

42 43 44 45 46

47 48 49 50 51 52 53 54 55 56

57

58

59 60 61

Commun. 1983, 705. (b) Cockerill, G. S.; Kocienski, P. J. Chem. Soc. Perkin. Trans. 1 1985, 2093. (c) Cockerill, G. S.; Kocienski, P. J. Chem. Soc. Perkin. Trans. 1 1985, 2101. Friedrich, D.; Paquette, L. A. J. Chem. Soc. Perkin. Trans. 1 1991, 1621. Tatsuta, K.; Takahashi, M.; Tanaka, N. J. Antibiotics 2000, 53, 88. Dixon, D. J.; Ley, S. V.; Tate, E. W. Synlett 1998, 1093. Hatanaka, Y.; Kuwajima, I. J. Org. Chem. 1986, 51, 1932. (a) Horiguchi, Y.; Furukawa, T.; Kuwajima, I. J. Am. Chem. Soc. 1989, 111, 8277. (b) Furukawa, T.; Morihira, K.; Horiguchi, Y.; Kuwajima, I. Tetrahedron 1992, 48, 6975. (c) Seto, M.; Morihira, K.; Horiguchi, Y.; Kuwajima, I. J. Org. Chem. 1994, 59, 3165. (d) Nakamura, T.; Waizumi, N.; Tsuruta, K.; Horiguchi, Y.; Kuwajima, I. Synlett 1994, 584. (e) Morihira, K.; Hara, R.; Kawahara, S.; Nishimori, T.; Nakamura, N.; Kusama, H.; Kuwajima, I. J. Am. Chem. Soc. 1998, 120, 12980. (f ) Kusama, H.; Hara, R.; Kawahara, S.; Nishimori, T.; Kashima, H.; Nakamura, N.; Morihira, K.; Kuwajima, I. J. Am. Chem. Soc. 2000, 122, 3811. Iwasawa, N.; Mukaiyama, T. Chem. Lett. 1982, 1441. Iwasawa, N.; Yura, T.; Mukaiyama, T. Tetrahedron 1989, 45, 1197. Iwasawa, N.; Mukaiyama, T. Chem. Lett. 1983, 297. Stevens, R. W.; Mukaiyama, T. Chem. Lett. 1983, 1799. Mukaiyama, T.; Iwasawa, N. Chem. Lett. 1984, 753. Mukaiyama, T.; Yamasaki, N.; Stevens, R. W.; Murakami, M. Chem. Lett. 1986, 213. Yura, T.; Iwasawa, N.; Mukaiyama, T. Chem. Lett. 1988, 1021. Yura, T.; Iwasawa, N.; Narasaka, K.; Mukaiyama, T. Chem. Lett. 1988, 1025. Mukaiyama, T.; Kobayashi, S. Org. React. 1994, 46, 1. (a) Kobayashi, S.; Mukaiyama, T. Chem. Lett. 1989, 297. (b) Kobayashi, S.; Uchiro, H.; Fujishita, Y.; Shiina, I.; Mukaiyama, T. J. Am. Chem. Soc. 1991, 113, 4247. (c) Mukaiyama, T.; Kobayashi, S. Stereocontrolled Organic Synthesis, Blackwell Scientiﬁc Publications, 1994, 37. (a) Mukaiyama, T.; Uchiro, H.; Kobayashi, S. Chem. Lett. 1989, 1001. (b) Mukaiyama, T.; Uchiro, H.; Kobayashi, S. Chem. Lett. 1989, 1757. (a) Kobayashi, S.; Sano, T.; Mukaiyama, T. Chem. Lett. 1989, 1319. (b) Mukaiyama, T.; Kobayashi, S.; Sano, T. Tetrahedron 1990, 46, 4653. Kobayashi, S.; Fujishita, Y.; Mukaiyama, T. Chem. Lett. 1989, 2069. Mukaiyama, T.; Asanuma, H.; Hachiya, I.; Harada, T.; Kobayashi, S. Chem. Lett. 1991, 1209. (a) Mukaiyama, T.; Kobayashi, S.; Uchiro, H.; Shiina, I. Chem. Lett. 1990, 129. (b) Kobayashi, S.; Ohtsubo, A.; Mukaiyama, T. Chem. Lett. 1991, 831. (c) Mukaiyama, T.; Furuya, M.; Ohtsubo, A.; Kobayashi, S. Chem. Lett. 1991, 989. (d) Kobayashi, S.; Uchiro, H.; Shiina, I.; Mukaiyama, T. Tetrahedron 1993, 49, 1761.

References 62 (a) Kobayashi, S.; Fujishita, Y.; Mukaiyama, T. Chem. Lett.

63 64 65

66 67

68 69 70 71 72 73 74

75 76 77 78 79

80 81

82 83 84 85

86 87

1990, 1455. (b) Kobayashi, S.; Furuya, M.; Ohtsubo, A.; Mukaiyama, T. Tetrahedron: Asymmetry 1991, 2, 635. Mukaiyama, T.; Uchiro, H.; Kobayashi, S. Chem. Lett. 1990, 1147. Kobayashi, S.; Kawasuji, T.; Mori, N. Chem. Lett. 1994, 217. (a) Evans, D. A.; MacMillan, D. W. C.; Campos, K. R. J. Am. Chem. Soc. 1997, 119, 10859. (b) Johnson, J. S.; Evans, D. A. Acc. Chem. Res. 2000, 33, 325. Evans, D. A.; Wu, J.; Masse, C. E.; MacMillan, D. W. C. Org. Lett. 2002, 4, 3379. (a) Mukaiyama, T.; Uchiro, H.; Shiina, I.; Kobayashi, S. Chem. Lett. 1990, 1019. (b) Mukaiyama, T.; Shiina, I.; Uchiro, H.; Kobayashi, S. Bull. Chem. Soc. Jpn. 1994, 67, 1708. Mukaiyama, T.; Shiina, I.; Kobayashi, S. Chem. Lett. 1991, 1901. Kobayashi, S.; Horibe, M.; Saito, Y. Tetrahedron 1994, 50, 9629. Kobayashi, S.; Horibe, M. J. Am. Chem. Soc. 1994, 116, 9805. Kobayashi, S.; Horibe, M. Synlett 1994, 147. Kobayashi, S.; Shiina, I.; Izumi, J.; Mukaiyama, T. Chem. Lett. 1992, 373. Mukaiyama, T.; Shiina, I.; Izumi, J.; Kobayashi, S. Heterocycles 1993, 35, 719. (a) Kobayashi, S.; Kawasuji, T. Tetrahedron Lett. 1994, 35, 3329. (b) Kobayashi, S.; Horibe, M.; Hachiya, I. Tetrahedron Lett. 1995, 36, 3173. Kobayashi, S.; Hayashi, T. J. Org. Chem. 1995, 60, 1098. Kobayashi, S.; Kawasuji, T. Synlett 1993, 911. Kobayashi, S.; Horibe, M.; Matsumura, M. Synlett 1995, 675. Shiina, I.; Ibuka, R. Tetrahedron Lett. 2001, 42, 6303. (a) Kobayashi, S.; Horibe, M. Tetrahedron: Asymmetry 1995, 6, 2565. (b) Kobayashi, S.; Horibe, M. Tetrahedron 1996, 52, 7277. (c) Kobayashi, S.; Horibe, M. Chem. Eur. J. 1997, 3, 1472. Kobayashi, S.; Horibe, M. Chem. Lett. 1995, 1029. (a) McGarvey, G. J.; Kimura, M.; Oh, T.; Williams, J. M. J. Carbohydr. Chem. 1984, 3, 125. (b) Mukaiyama, T.; Suzuki, K.; Yamada, T.; Tabusa, F. Tetrahedron 1990, 46, 265. Mukaiyama, T.; Shiina, I.; Kobayashi, S. Chem. Lett. 1990, 2201. Mukaiyama, T.; Anan, H.; Shiina, I.; Kobayashi, S. Bull. Soc. Chim. Fr. 1993, 130, 388. Kobayashi, S.; Onozawa, S.; Mukaiyama, T. Chem. Lett. 1992, 2419. (a) Kanda, Y.; Fukuyama, T. J. Am. Chem. Soc. 1993, 115, 8451. (b) Fukuyama, T.; Kanda, Y. Synth. Org. Chem. Jpn. 1994, 52, 888. White, J. D.; Deerberg, J. Chem. Commun. 1997, 1919. (a) Kobayashi, S.; Hayashi, T.; Kawasuji, T. Tetrahedron Lett. 1994, 35, 9573. (b) Kobayashi, S.; Furuta, T. Tetrahedron 1998, 54, 10275.

163

164

4 Tin-promoted Aldol Reactions and their Application to Total Syntheses of Natural Products 88 (a) Kobayashi, S.; Hayashi, T.; Iwamoto, S.; Furuta, T.;

89 90

91

92

93

94 95

96

97

98

99

Matsumura, M. Synlett 1996, 672. (b) Kobayashi, S.; Matsumura, M.; Furuta, T.; Hayashi, T.; Iwamoto, S. Synlett 1997, 301. (c) Kobayashi, S.; Furuta, T.; Hayashi, T.; Nishijima, M.; Hanada, K. J. Am. Chem. Soc. 1998, 120, 908. Wakabayashi, T.; Mori, K.; Kobayashi, S. J. Am. Chem. Soc. 2001, 123, 1372. (a) Kobayashi, S.; Ueno, M.; Suzuki, R.; Ishitani, H. Tetrahedron Lett. 1999, 40, 2175. (b) Kobayashi, S.; Ueno, M.; Suzuki, R.; Ishitani, H.; Kim, H.-S.; Wataya, Y. J. Org. Chem. 1999, 64, 6833. (c) Okitsu, O.; Suzuki, R.; Kobayashi, S. J. Org. Chem. 2000, 66, 809. Evans, D. A.; Trotter, B. W.; Coleman, P. J.; Coˆte´, B.; Dias, L. C.; Rajapakse, H. A.; Tyler, A. N. Tetrahedron 1999, 55, 8671. (a) Evans, D. A.; Cee, V. J.; Smith, T. E.; Fitch, D. M.; Cho, P. S. Angew. Chem. Int. Ed. Engl. 2000, 39, 2533. (b) Evans, D. A.; Fitch, D. M. Angew. Chem. Int. Ed. Engl. 2000, 39, 2536. (c) Evans, D. A.; Fitch, D. M.; Smith, T. E.; Cee, V. J. J. Am. Chem. Soc. 2000, 122, 10033. (a) Mukaiyama, T.; Shiina, I.; Sakata, K.; Emura, T.; Seto, K.; Saitoh, M. Chem. Lett. 1995, 179. (b) Shiina, I.; Uoto, K.; Mori, N.; Kosugi, T.; Mukaiyama, T. Chem. Lett. 1995, 181. (c) Shiina, I.; Iwadare, H.; Sakoh, H.; Tani, Y.; Hasegawa, M.; Saitoh, K.; Mukaiyama, T. Chem. Lett. 1997, 1139. (d) Shiina, I.; Iwadare, H.; Sakoh, H.; Hasegawa, M.; Tani, Y.; Mukaiyama, T. Chem. Lett. 1998, 1. (e) Shiina, I.; Saitoh, K.; Fre´chard-Ortuno, I.; Mukaiyama, T. Chem. Lett. 1998, 3. (f ) Mukaiyama, T.; Shiina, I.; Iwadare, H.; Saitoh, M.; Nishimura, T.; Ohkawa, N.; Sakoh, H.; Nishimura, K.; Tani, Y.; Hasegawa, M.; Yamada, K.; Saitoh, K. Chem. Eur. J. 1999, 5, 121. Saitoh, K.; Shiina, I.; Mukaiyama, T. Chem. Lett. 1998, 679. (a) Shiina, I.; Fukuda, Y.; Ishii, T.; Fujisawa, H.; Mukaiyama, T. Chem. Lett. 1998, 831. (b) Shiina, I.; Fujisawa, H.; Ishii, T.; Fukuda, Y. Heterocycles 2000, 52, 1105. (a) Shiina, I.; Miyoshi, S.; Miyashita, M.; Mukaiyama, T. Chem. Lett. 1994, 515. (b) Shiina, I.; Miyashita, M.; Mukaiyama, T. Chem. Lett. 1994, 677. (c) Shiina, I. Tetrahedron 2004, 59, 1587. Shiina, I.; Kawakita, Y.; Ibuka, R. Abstracts of Papers, 83rd National Meeting of the Chemical Society of Japan, Tokyo, 2003, Vol. 2, 2C401; Ibuka, R. Ph.D. Thesis, Tokyo University of Science, Tokyo, Japan, 2003. (a) Shiina, I.; Oshiumi, H.; Hashizume, M.; Yamai, Y.; Ibuka, R. Tetrahedron Lett. 2004, 45, 543. (b) Shiina, I.; Kubota, M.; Oshiumi, H.; Hashizume, M. J. Org. Chem. 2004, 69, 1822. (a) Guindon, Y.; Faucher, A.-M.; Bourque, E´.; Caron, V.; Jung, G.; Landry, S. R. J. Org. Chem. 1997, 63, 9276. (b) Guindon, Y.; Jung, G.; Gue´rin, B.; Ogilvie, W. W. Synlett 1998, 213. (c) Hena, M. A.; Terauchi, S.; Kim, C.-S.;

References Horiike, M.; Kiyooka, S. Tetrahedron: Asymmetry 1998, 9, 1883. 100 (a) Shiina, I.; Ibuka, R.; Kubota, M. Chem. Lett. 2002, 286. (b) Shiina, I.; Kubota, M.; Ibuka, R. Tetrahedron Lett. 2002, 43, 7535. 101 Uchiro, H.; Nagasawa, K.; Kotake, T.; Hasegawa, D.; Tomita, A.; Kobayashi, S. Bioorg. Med. Chem. Lett. 2002, 12, 2821.

165

167

5

Zirconium Alkoxides as Lewis Acids ¯ Kobayashi Yasuhiro Yamashita and Shu 5.1

Introduction

Zirconium is a group 4 element, and its low cost and low toxicity are advantageous compared with other metals used in industry. The usefulness of zirconium is well known in organic chemistry. Zirconium compounds promote some organic reactions eﬃciently and play important roles resulting in interesting selectivity [1]. In aldol reactions zirconium compounds have often been used to realize high and unique selectivity by forming zirconium enolates [2, 3]. Bis(cyclopentadienyl) zirconium compounds, which form zirconium enolates via metal exchange from lithium enolates, have been successfully used in stereoselective aldol reactions, and high syn selectivity was obtained. Both (E) and (Z) enolates gave the same syn adducts predominantly. This methodology was also applied to highly diastereoselective asymmetric aldol reactions to aﬀord aldol adducts with excellent selectivity (Scheme 5.1) [3]. Zirconium alkoxides, especially zirconium tetra-t-butoxide, have been known to act as bases and directly deprotonate the a-hydrogen atoms of ketones to form zirconium enolates [4]. The enolates reacted with aldehydes to give aldol adducts (Scheme 5.2) [4c]. Following this report, asymmetric aldol and related reactions using chiral zirconium alkoxides as bases were investigated. Aldol–Tishchenko reactions are an eﬃcient method for synthesizing 1,3-diol derivatives from aldehydes and enolates. It was recently shown that a zirconium enolate generated by retro-aldol reaction of a b-keto-tert-alcohol and a catalytic amount of a zirconium t-butoxide–TADDOL complex, reacted with an aldehyde to aﬀord the corresponding 1,3-anti-diol via domino aldol–Tishchenko process in good yield with moderate enantioselectivity (Scheme 5.3) [5]. Similar to the aldol reaction of zirconium enolates, zirconium Lewis acidmediated aldol reactions of silicon enolates with aldehydes (Mukaiyama aldol reaction) have also been well explored [6]. Zirconium Lewis acids are comparatively mild and have been employed in several stereoselective reModern Aldol Reactions. Vol. 2: Metal Catalysis. Edited by Rainer Mahrwald Copyright 8 2004 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 3-527-30714-1

168

5 Zirconium Alkoxides as Lewis Acids

O

R1

LDA Cp2ZrCl2

R1 R2

OZr(Cl)Cp2 R2 O

(Z )-form

R3CHO

+

THF, –78 °C

R3

R1

R2 (E )-form

OH R3 R2

syn

R1

anti

syn/anti =52/48 to 98/2

MEMO

OZr(Cl)Cp2

RCHO +

O + R1

R2

OZr(Cl)Cp2

MEMO

OH

O

N

R

N

THF-hexane –78-0 °C

OH

69-77% > 96% de

Scheme 5.1

Stereoselective aldol reactions using zirconium enolates.

O

O

O +

H

OH

Zr(Ot Bu)4 (2.6 eq.) Ph

Ph

THF, –30 °C

(2.0 eq.)

77%

Scheme 5.2

Direct aldol reaction using a zirconium alkoxide.

Ph Ph O Zr(Ot Bu)4 + O

OH O H

+

(10 mol%) t Bu

OH OH

O

Ph Ph

O

CH2Cl2, 0 °C, 3 h

O

OH tBu

84%, 57% ee Scheme 5.3

Asymmetric aldol-Tishchenko reaction using a chiral zirconium catalyst.

actions. Among these, zirconium alkoxides have served as mild and stereoselective Lewis acids, especially in asymmetric catalysis. Several catalytic asymmetric reactions using chiral zirconium alkoxides as Lewis acids have been developed [4c]. In chiral modiﬁcation of zirconium catalysts, chiral alcohol derivatives or chiral phenol derivatives, especially 1,1 0 -binaphthalene2,2 0 -diol (BINOL) derivatives, have often been employed. In this chapter,

5.2 The Asymmetric Mukaiyama Aldol Reaction

aldol and related reactions catalyzed by zirconium alkoxides as chiral Lewis acids are discussed.

5.2

The Asymmetric Mukaiyama Aldol Reaction

A chiral zirconium catalyst prepared from zirconium alkoxide and BINOL derivatives has been successfully applied to the catalytic asymmetric Mukaiyama aldol reaction [7]. Among several types of zirconium catalyst, a catalyst prepared from zirconium tetra-t-butoxide and 3,3 0 -dihalogeno-1,1 0 bi-2-naphthol (3,3 0 -X2 BINOL) [8] was found to be eﬀective. The aldol reactions of benzaldehyde 1a with ketene silyl acetals proceeded smoothly in toluene at 0 C in the presence of an additional alcohol. It was found that the additional alcohol played important roles in this reaction (vide infra) [9]. Among the catalysts screened, a Zr catalyst containing 3,3 0 -I2 BINOL was the most eﬀective, and high level of enantiocontrol was achieved. The reproducibility of the reaction was not as good, however, and the enantioselectivity was sometimes poor. After several investigations to address this issue it was ﬁnally revealed that a small amount of water had an important eﬀect on enantioselectivity [10]. The eﬀect of water was signiﬁcant. Under strictly anhydrous conditions enantioselectivity was occasionally quite low. It was revealed that the presence of a small amount of water was essential to realize high enantioselectivity (Table 5.1). The eﬀect of alcohol additives was investigated in the reaction of benzaldehyde 1a with the ketene silyl acetal of S-ethyl ethanethioate (2a) (Table 5.2). In the presence of water reactions using normal primary alcohols such as ethanol (EtOH), propanol (PrOH), and butanol (BuOH) gave high yields and high enantioselectivity (entries 1–3). Other primary alcohols such as benzyl alcohol (BnOH) and 2,2,2-triﬂuoroethanol (CF3 CH2 OH) gave lower yields and selectivity (entries 4 and 5). Secondary and tertiary alcohols such as isopropanol ( i PrOH) and tert-butyl alcohol ( t BuOH) resulted in reduced yields and selectivity (entries 6 and 7). Phenol also resulted in lower yield and selectivity (entry 8). The best yields and enantioselectivity were obtained when 80–120 mol% PrOH was used (entries 9–12) and similar yields and selectivity were obtained when zirconium tetrapropoxide–propanol complex (Zr(OPr)4 aPrOH) was employed instead of Zr(O t Bu)4 (entry 14). The use of Zr(OPr)4 aPrOH is desirable economically. Other substrates were then examined; the results are shown in Table 5.3. The ketene silyl acetal derived from methyl isobutyrate (2b) also worked well. For aldehydes, whereas aromatic and a,b-unsaturated aldehydes gave excellent yields and selectivity, aliphatic aldehydes resulted in high yields but somewhat lower selectivity. Diastereoselective aldol reactions using this chiral zirconium catalyst were then examined (Table 5.4). First, the ketene silyl acetal derived from methyl

169

170

5 Zirconium Alkoxides as Lewis Acids Tab. 5.1

Eﬀect of water in asymmetric Mukaiyama aldol reactions using a chiral zirconium catalyst. OSiMe3

O Ph

H

R1

+

XR2 R1

Silicon Enolate

1a 2 3 4

Product

OSiMe3 SEt

Ph

ee (%)

SEt

0 0 10 20

28 42 86 94

16 3 83 88

OMe

0 20

94 88

10b 94

OH O

OSiMe3

2b

R 1 R1 4

Yield (%)

4aa

OMe

XR2

Ph

H2 O (mol%)

OH O

2a

5 6

OH O

toluene, 0°C, 14-18 h

2

1a

Entry

Zr(Ot Bu)4 (10 mol%) (R)-3,3'-I2BINOL (3a) (12 mol%) PrOH (50 mol%), H2O

Ph 4ab

a PrOH b The

was not used. product obtained was the opposite enantiomer. I OH OH I (R)-3,3'-I2BINOL (3a)

propionate (2c) was employed in the reaction with benzaldehyde. The reaction proceeded smoothly to aﬀord the desired anti-aldol adduct in high yield with high diastereo- and enantioselectivity when ethanol was used as a primary alcohol. The selectivity was further improved by use of the ketene silyl acetal derived from phenyl propionate (2d). Other aldehydes such as p-anisaldehyde (1b), p-chlorobenzaldehyde (1g), cinnamaldehyde (1d), and 3-phenylpropionealdehyde (1e), etc., were tested, and all the reactions proceeded smoothly, and the desired anti-aldol adducts were obtained in high yield with high diastereo- and enantioselectivity. In the reactions of the ketene silyl acetals derived from propionate derivatives, most chiral Lewis acids led to syn diastereoselectivity. Few catalyst systems giving anti-aldol adducts with high selectivity are known, so the general anti selectivity was a remarkable feature of the zirconium aldol reaction [11]. Although the high anti selectivity observed in these reactions is remarkable, examination of the eﬀect of the geometry of the ketene silyl acetals revealed further important information on the selectivity – when the (E) and (Z) ketene silyl acetals (2e and 2f ) derived from methyl propionate were employed in reactions with benzaldehyde, high anti selectivity was obtained for both, and it was conﬁrmed that selectivity was independent of the ge-

5.2 The Asymmetric Mukaiyama Aldol Reaction Tab. 5.2

Eﬀect of alcohol (additive). OSiMe3

O H +

Ph

SEt 2a

1a

Entry

Zr(Ot Bu)4 (10 mol%) (R)-3,3'-I2BINOL (3a) (12 mol%) PrOH , H2O (20 mol%)

OH O Ph

toluene, 0°C, 14 h

SEt 4aa

ROH (mol%)

Yield (%)

ee (%)

1 2 3 4 5

EtOH (50) PrOH (50) BuOH (50) BnOH (50) CF3 CH2 OH (50)

85 94 92 76 47

87 88 86 76 62

6 7 8

i

t

PrOH (50) BuOH (50) PhOH (50)

87 39 36

85 44 54

9 10 11 12 13

PrOH (30) PrOH (80) PrOH (100) PrOH (120) PrOH (160)

68 91 95 94 95

74 95 95 95 91

14a

PrOH (60)

98

92

a Zr(OPr)

4 -PrOH

was used instead of Zr(O t Bu)4 .

Tab. 5.3

Asymmetric aldol reactions (1). O R1

H 1

+

R3

Zr(Ot Bu)4 (10 mol%) OH O OSiMe3 (R)-3,3'-I BINOL (3a) (12 mol%) 2 1 XR2 XR2 PrOH (80 mol%), H2O (20 mol%) R R3 R 3 R3 toluene, 0°C, 18 h 2 4

Entry

Aldehyde (R1 )

Silicon Enolate

Yield (%)

ee (%)

1 2 3 4 5 6 7

Ph (1a) Ph (1a) p-MeOC6 H4 (1b) (E)-CH3 CHbCH (1c) (E)-PhCHbCH (1d) PhCH2 CH2 (1e) CH3 (CH2 )4 ð1f Þ

2a 2b 2b 2b 2b 2a 2a

91 95 92 76 94 92 93 (93)a

95 98 96 97 95 80 84 (87)a

a The

reaction was performed at 20 C. OSiMe3 SEt 2a

OSiMe3 OMe 2b

171

172

5 Zirconium Alkoxides as Lewis Acids Tab. 5.4

Asymmetric aldol reactions (2). OSiMe3

O R1

+

OR

H

2

Zr(Ot Bu)4 (10 mol%) (R)-3,3'-I2BINOL (3a) (12 mol%) ROH (80 mol%), H2O (20 mol%) toluene, 0°C, 18 h

2

1

OH O OR2

R1 4

Entry

Aldehyde (R1 )

Silicon Enolate

ROH

Yield (%)

syn/ anti

ee (%)

1 2 3 4 5 6 7 8 9

Ph (1a) Ph (1a) Ph (1a) Ph (1a) p-MeOC6 H4 (1b) p-ClC6 H4 (1g) (E)-CH3 CHbCH (1c) (E)-PhCHbCH (1d) PhCH2 CH2 (1e)

2c 2c 2d 2d 2d 2d 2d 2d 2d

PrOH EtOH PrOH EtOH PrOH PrOH PrOH PrOH PrOH

79 87 94 90 89 96 65 92 61

7/93 7/93 5/95 4/96 7/93 9/91 11/89 15/85 14/86

96 97 99 99 98 96 92 98 89

OSiMe3 OMe 2c

OSiMe3 OPh 2d

ometry of the ketene silyl acetals (Scheme 5.4). For the transition states of these reactions, acyclic pathways are assumed (details are discussed below). Further investigation of the eﬀect of the aldehyde structure was conducted (Table 5.5). Reactions of other aliphatic aldehydes were examined O

OSiEtMe2 H +

E/Z = 88/12 2e

1a

63% Yield syn/anti = 9/91 95% ee (Anti)

OMe

Zr(OtBu)4 (10 mol%) (R)-3,3'-I2BINOL (3a) (12 mol%) EtOH (80 mol%) H2O (20 mol%) toluene, 0 °C, 18 h

O H

+

OSiEtMe2 OMe E/Z = 7/93

1a

2f

Scheme 5.4

Eﬀect of geometry of the silicon enolates.

OH

O OMe

4ac

77% Yield syn/anti = 7/93 98% ee (Anti )

5.2 The Asymmetric Mukaiyama Aldol Reaction Tab. 5.5

Eﬀect of structures of aliphatic aldehydes. OSiMe3

O + R

OPh

H

Entry

Aldehyde

1

3 4

CHO Ph

CHO CHO CHO

OH O R

toluene, 0°C, 18 h

2d

1

2

Zr(Ot Bu)4 (20 mol%) (R)-3,3'-I2BINOL (3a) (24 mol%) PrOH (160 mol%), H2O (20 mol%)

OPh 4

Yield (%)

syn/anti

ee (%) (anti)

(1f )

64

12/88

85

(1e)

71

10/90

82

(1h)

71

15/85

81

(1i)

56

12/88

89

(1j)

52

14/86

78

(1k)

16

17/83

28

(1l)

14

21/79

31

(1m)

Trace

–

–

CHO

5 6 7 8

CHO

CHO CHO

using a chiral zirconium catalyst consisting of Zr(O t Bu)4 , (R)-3,3 0 -I2 BINOL (3a), PrOH, and water. For normal linear aliphatic aldehydes, for example hexanal and butanal, the reactions proceeded with high selectivity. gBranched aldehydes also reacted smoothly to aﬀord the desired anti adducts in good yield and with high diastereo- and enantioselectivity. The catalyst did not, however, work well in reactions of a-branched and b-branched aliphatic aldehydes, possibly because of by steric interaction between the BINOL parts (especially large di-iodo atoms at the 3,3 0 -positions) of the catalysts and the alkyl moieties of the aldehydes. These results indicated that the environment around the zirconium of the catalyst was crowded and that the catalyst recognized the structure of the aldehydes strictly. To create more eﬀective catalyst systems, improvement of catalyst activity is important. It has recently been revealed that introduction of stronger electron-withdrawing groups at the 6,6 0 -positions of the binaphthyl rings eﬀectively improved the Lewis acidity of the zirconium catalyst system [12]. The eﬀect of stronger electron-withdrawing groups at the 6,6 0 -positions of (R)-3,3 0 -I2 BINOL in this aldol system was therefore examined. Bromo, iodo, and pentaﬂuoroethyl groups were selected as the electron-withdrawing

173

174

5 Zirconium Alkoxides as Lewis Acids Tab. 5.6

Improvement of catalyst activity in the aldol reaction. OSiMe3

O + Ph

H 1a

SEt

Zr(Ot Bu)4 (5 mol%) (R)-BINOL 3 ROH (50 mol%), H2O (10 mol%)

OH O Ph

toluene, 0°C, 3 h

2g

SEt 4ag

Entry

BINOL Derivatives

Yield (%)

syn/anti

ee (%) (anti)

1

(R)-3,3 0 -I2 BINOL (3a) (6 mol%) (R)-3,3 0 -I2 -6,6 0 -Br2 BINOL (3b) (7.5 mol%) (R)-3,3 0 -I2 -6,6 0 -(C2 F5 )2 BINOL (3c) (7.5 mol%) (R)-3,3 0 ,6,6 0 -I4 BINOL (3d) (7.5 mol%)

38

5/95

96

61

4/96

98

71

7/93

96

70

4/96

98

2 3 4

X

I OH OH

X

I

X: H, Br, I, C2F5

(R)-3,3'-I2-6,6'-X2BINOL (3)

groups. In the aldol reaction of benzaldehyde with the ketene silyl acetal derived from S-ethyl propanethioate (2g), the new catalysts prepared from 6,6 0 -disubstituted-3,3 0 -I2 BINOL (3b–d) had greater activity and the reaction proceeded much faster than with the catalyst prepared from 3,3 0 -I2 BINOL. In particular, iodo and pentaﬂuoroethyl groups at the 6,6 0 -positions led to better results, giving the desired anti adducts in high yields with high diastereo- and enantioselectivity (Table 5.6). The new catalyst system was successfully applied to the reactions of aliphatic aldehydes. In the reactions of hexanealdehyde with the ketene silyl acetals derived from phenyl propionate (2d) and S-ethyl propanethioate (2g), the best results were obtained when (R)-3,3 0 ,6,6 0 -I4 BINOL (3d) was employed (Table 5.7). The electronwithdrawing substituents at the 3,3 0 - and 6,6 0 -positions of the BINOL derivatives were assumed to increase the Lewis acidity of the zirconium catalysts. By changing the chiral ligands, chemical yields were improved (38% to 71% in Table 5.6, entries 1 and 3; 9% to 92% in Table 5.7, entries 4 and 6) much more than enantioselectivity (80% ee to 87% ee in Table 5.7, entries 1 and 3). A chiral zirconium complex prepared from zirconium tetra-t-butoxide, 2.2 equiv. 6,6 0 -dibromo-1,1 0 -binaphthalene-2,2 0 -diol (6,6 0 -Br2 BINOL), and a small amount of water, which was originally developed in asymmetric Mannich-type reactions [12], was found to be an eﬀective catalyst in asymmetric aldol-type reactions using ethyl diazoacetate [13]. The chiral zirco-

5.3 Asymmetric Hetero Diels–Alder Reaction Tab. 5.7

Eﬀect of new BINOLs in the aldol reactions of hexanaldehyde 1f. OSiMe3

O H

C5H11

+

R

1f

Zr(Ot Bu)4 (10 mol%) OH O (R)-3,3'-I2-6,6'-X2BINOL (3) (12-15 mol%) ROH (80 mol%), H2O (20 mol%) R C5H11 toluene, 0°C, 18 h

2

Entry

X

Silicon Enolate

1 2 3

H (3a) C2 F5 (3c) I (3d)

OSiMe3

4 5 6

H (3a) C2 F5 (3c) I (3d)

OSiMe3

OPh 2d

SEt 2g X

4

Yield (%)

syn/anti

ee (%) (anti)

53 39 66

16/84 11/89 12/88

80 84 87

9 80 92

12/88 17/83 11/89

93 93 93

I OH OH

X

I

(R)-3,3'-I2-6,6'-X2BINOL (3)

nium complex deprotonates the a-hydrogen atom of the diazo ester directly (Scheme 5.5). Zr(OtBu)4 + 2(S)-6,6'-Br2BINOL O

O H

+

H

OEt N2

(20 mol%) H2O (20 mol%) DME, –35 °C, 72 h

OH O OEt N2 65%, 87% ee

Scheme 5.5

Asymmetric aldol reaction of an azoester.

5.3

Asymmetric Hetero Diels–Alder Reaction

Hetero Diels–Alder (HDA) reactions of aldehydes with 1-methoxy-3trimethylsiloxy-1,3-butadiene (Danishefsky’s diene) also proceeded in the presence of the chiral zirconium catalyst [14]. HDA reactions of aldehydes with Danishefsky’s dienes [15] mediated by Lewis acids, which provide 2,3dihydro-4H-pyran-4-one derivatives, are promising tools for construction of pyran ring systems [16]. Because Danishefsky’s dienes contain a silicon enolate moiety, an aldol-type reaction and subsequent cyclization process

175

176

5 Zirconium Alkoxides as Lewis Acids Tab. 5.8

Asymmetric hetero Diels–Alder reactions of benzaldehyde (1a) using a chiral zirconium catalyst. OSiR23

O Ph

H

+

OR1

1a

Zr(Ot Bu)4 (10 mol%) (R)-3,3'-I2BINOL (3a) (12 mol%) PrOH, H2O (20 mol%) TFA Ph

toluene, 18 h

2

O

O 5ah

Entry

Diene

PrOH (mol%)

Temp. (˚C)

Yield (%)

ee (%)

1 2 3 4 5 6 7 8 9 10a

2h 2h 2i 2i 2i 2i 2i 2j 2k 2j

50 80 50 80 120 80 80 80 80 80

0 0 0 0 0 20 45 20 20 20

39 35 50 65 44 70 Trace 80 24 Quant.

22 62 91 94 94 97 – 97 89 97

a Toluene/tBuOMe

(2:1) was used as solvent. OSiR3

OSiMe3 OMe 2h

Ot Bu 2i: SiR3 = SiMe3 2j: SiR3 = SiEtMe2 2k: SiR3 = Sit BuMe2

are a possible pathway in HDA reactions using Danishefsky’s dienes [17]. It was therefore assumed that the chiral zirconium catalyst eﬀective in aldol reactions might work well in HDA reactions of aldehydes with Danishefsky’s dienes. On the basis of this assumption a model HDA reaction of benzaldehyde with 1-methoxy-3-trimethylsiloxy-1,3-butadiene (2h) using a chiral zirconium catalyst prepared from Zr(O t Bu)4 , (R)-3,3 0 -I2 BINOL, PrOH, and water was examined (Table 5.8). The initial result was rather disappointing, however, and the corresponding pyranone derivative was obtained in lower yield and selectivity (Table 5.8, entry 1). The selectivity was improved to 62% ee by increasing the amount of PrOH, but the yield was even lower (entry 2). According to the reaction pathway the product was formed with elimination of methanol. It was suspected that this methanol might decompose the diene, reducing the yield, and that the selectivity might be reduced by interaction of the methanol with the zirconium catalyst. To prevent production of methanol, 1-tert-butoxy-3-trimethylsiloxy-1,3butadiene (2i) was used instead of 2h. As expected, yield and selectivity were improved, and the desired adduct was obtained in 50% yield with 91% ee (entry 3). Yield and selectivity were also found to be aﬀected by the amount of PrOH and the reaction temperature, and the desired product was ob-

5.3 Asymmetric Hetero Diels–Alder Reaction Tab. 5.9

Asymmetric hetero Diels–Alder reactions (1). O R

OSiEtMe2 H

+

Ot Bu 2j

1

Zr(Ot Bu)4 (10 mol%) (R)-3,3'-I2BINOL (3a) (12 mol%) PrOH (80 mol%), H2O (20 mol%)

O TFA R

solvent, –20°C, 18 h

O 5

Entry

Aldehyde; R

Solvent

Yield (%)

ee (%)

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

Ph (1a) Ph (1a) p-MeC6 H4 (1n) p-MeC6 H4 (1n) p-ClC6 H4 (1g) p-ClC6 H4 (1g) PhCH2 CH2 (1e) PhCH2 CH2 (1e) CH3 (CH2 )4 (1f ) CH3 (CH2 )4 (1f ) (E)-PhCHbCH (1d)

Toluene Toluene- t BuOMea Toluene Toluene- t BuOMea Toluene Toluene- t BuOMea Toluene Toluene- t BuOMea Toluene Toluene- t BuOMea Toluene- t BuOMea

80 Quant. 63 95 65 90 84 Quant. 69 98 97

97 97 95 95 84 84 90 90 91 93 90

a Toluene/ t BuOMe b Sc(OTf )

3

(2:1) was used. (10 mol%) was used instead of TFA.

tained in 70% yield with 97% ee when the reaction was performed using 80 mol% PrOH at 20 C (entry 6). The eﬀects of substituents on the silicon atoms of Danishefsky’s dienes and of the solvents in this HDA reaction were further examined. When diene 2j, with an ethyldimethylsilyloxy group, was employed the desired product was obtained in 80% yield with 97% ee (entry 8), although the more stable diene 2k, with a tert-butyldimethylsilyloxy group, did not work well under these conditions (entry 9). Finally, the desired pyranone derivative was obtained quantitatively with 97% ee by use of the mixed solvent system toluene– t BuOMe, 2:1 (entry 10) [18]. The reactions of diﬀerent aldehydes with diene 2j in toluene or toluene– t BuOMe, 2:1, were then tested (Table 5.9). Aromatic, a,b-unsaturated, and even aliphatic aldehydes reacted with the Danishefsky diene to aﬀord the desired HDA adducts in good to high yields with high enantioselectivity. It was noted that high stereocontrol was achieved even in reactions of aliphatic aldehydes. With regard to solvents, use of the toluene– t BuOMe system always resulted in higher yields than use of toluene. In these HDA pathways the reaction was quenched by adding saturated aqueous NaHCO3 , and after usual work-up the crude adduct was treated with triﬂuoroacetic acid (TFA) to accelerate formation of the pyranone derivative. In the reaction with cinnamaldehyde, however, the desired cyclic product was not obtained in good yield after treatment with TFA, because side-reactions occurred. When other work-up conditions were investigated it was found that treatment with a

177

178

5 Zirconium Alkoxides as Lewis Acids Tab. 5.10

Asymmetric hetero Diels–Alder reactions (2). Zr(Ot Bu)4 (10 mol%) OSiMe3

O R

H 1

+

Ot Bu

(R)-BINOL 3 (12-15 mol%) PrOH (80 mol%), H2O (20 mol%)

O TFA

Conditions

R

2l

O 5

Entry Aldehyde; R

BINOL Conditions

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

Ph (1a) Ph (1a) Ph (1a) p-MeC6 H4 (1n) p-MeC6 H4 (1n) p-ClC6 H4 (1g) p-ClC6 H4 (1g) p-ClC6 H4 (1g) (E)-PhCHbCH (1d) (E)-PhCHbCH (1d) PhCH2 CH2 (1e) PhCH2 CH2 (1e) PhCH2 CH2 (1e)

3a 3c 3c 3c 3c 3c 3c 3c 3c 3c 3c 3d 3d

14c 15c

CH3 (CH2 )4 (1f ) CH3 (CH2 )4 (1f )

3d 3d

Toluene, 0 C, 18 h Toluene, 20 C, 18 h Toluene, 40 C, 24 h Toluene, 20 C, 18 h Toluene, 40 C, 24 h Toluene, 20 C, 18 h Toluene, 40 C, 24 h Toluene, 40 C, 168 h Toluene, 10 C, 18 h Toluene, 20 C, 60 h Toluene, 0 C, 48 h Toluene, 20 C, 48 h Toluene/ t BuOMe (2:1), 20 C, 72 h Toluene, 20 C, 48 h Toluene/ t BuOMe (2:1), 20 C, 72 h

Yield (%)

cis/ trans

ee (%) (trans)

Trace Quant. 99 93 99 99 99 90 78 96 23 68 97

– 8/92 4/96 13/87 6/94 10/90 4/96 6/94 12/88 10/90 15/85 10/90 10/90

– 98 97 90 93 97 98 97 87 90 79 87 90

63 94

9/91 9/91

88 95

a2

mol% Zr catalyst. PrOH (80 mol%). b Sc(OTf )3 (10 mol%) was used instead of TFA. c PrOH (120 mol%). X

I OH (R)-3a: X=H (R)-3c: X=C2F5 OH (R)-3d: X=I

X

I

catalytic amount of scandium triﬂate (Sc(OTf )3 ) [19] gave a high yield of the desired product. The HDA reactions of 4-methyl-substituted Danishefsky’s diene, which include diastereo- and enantiofacial selectivity issues, were then investigated. 1-tert-Butoxy-2-methyl-3-trimethylsiloxy-1,3-pentadiene (2l) was selected as a model, and was reacted with benzaldehyde using a chiral zirconium catalyst with an (R)-3,3 0 -I2 BINOL moiety under the conditions shown in Table 5.9 (toluene was used as solvent). Unexpectedly, the reaction proceeded sluggishly (Table 5.10, entry 1), a result which clearly showed that 4substituted diene 2l was less reactive than 4-unsubstituted dienes 2h–2k. To increase the Lewis acidity of the zirconium catalyst, introduction of electronwithdrawing groups at the 6,6 0 -positions of the BINOL derivatives was

5.3 Asymmetric Hetero Diels–Alder Reaction

investigated. (R)-3,3 0 -I2 -6,6 0 -(C2 F5 )2 BINOL (3c) was chosen as chiral ligand and a chiral zirconium catalyst was prepared from Zr(O t Bu)4 , (R)-3c, PrOH, and water. It was remarkable that in the presence of 10 mol% of this chiral zirconium catalyst the reaction of benzaldehyde (1a) with 2l proceeded smoothly in toluene at 20 C to aﬀord the desired HDA adduct quantitatively. It was also of interest that the stereochemistry of the product was 2,3trans, and that the enantiomeric excess of the trans adduct was proved to be 98% (Table 5.10, entry 2). The trans selectivity was further improved when the reaction was performed at 40 C (entry 3). The reactions of other aldehydes, including aromatic, a,b-unsaturated, and aliphatic aldehydes, using this new chiral zirconium catalyst were examined. With aromatic and a,b-unsaturated aldehydes the reactions proceeded smoothly to give the desired pyranone derivatives in high yield with high trans selectivity; the enantiomeric excesses of the trans adducts were also high. Yield and selectivity were, however, lower in the reaction of an aliphatic aldehyde (entry 11). They were ﬁnally improved when (R)-3,3 0 ,6,6 0 -I4 BINOL ((R)-3d) was used instead of (R)-3c and the reaction was conducted in toluene– t BuOMe, 2:1, as a solvent; the desired products were obtained in high yield with high diastereo- and enantioselectivity. It should be noted this was the ﬁrst example of catalytic asymmetric trans-selective HDA reactions of aldehydes, and that a wide variety of aldehydes react with high yields and selectivity. HDA reactions using more functionalized Danishefsky’s dienes were also investigated. 3-Oxygenated 2-alkyl-2,3-dihydro-4H-pyran-4-one derivatives are important synthetic intermediates, aﬀording hexose derivatives [20]. As a new approach to hexose derivatives, an HDA reaction of an aldehyde with a Danishefsky’s diene having an oxy-substituent at the 4-position has already been developed [21]. As a catalytic asymmetric HDA reaction using this type of diene the reaction of benzaldehyde (1a) with 1-tert-butyldimethylsilyloxy-4-tert-butoxy-2-trimethylsilyloxy-1,3-butadiene (2m) was conducted using the Zr-3,3 0 ,6,6 0 -I4 BINOL catalyst system. The reaction proceeded sluggishly and it was speculated that the bulky substituent at the 4-position prevented the smooth progress of the reaction. Next, the reaction employing 1-benzyloxy-4-tert-butoxy-2-trimethylsilyloxy-1,3-butadiene (2n) as diene component was investigated. The HDA reactions of aldehydes with diene 2n using the zirconium catalyst were conducted under optimized conditions (Table 5.11). The reactions proceeded smoothly in toluene–tert-butyl methyl ether, 2:1, to aﬀord the desired cycloadducts in high yield with high diastereo- and enantioselectivity. It should be noted that the stereochemistry of the adduct obtained was 2,3-cis, completely opposite to the 2,3-trans selectivity obtained in the reaction with diene 2l (details of the selectivity are discussed in Section 5.4). In the reaction of other aldehydes, aromatic aldehydes and a,bunsaturated and aliphatic aldehydes reacted with the diene smoothly to aﬀord the desired products in high yield with high cis selectivity; the enantiomeric excess of the cis adducts was also high.

179

180

5 Zirconium Alkoxides as Lewis Acids Tab. 5.11

Asymmetric hetero Diels–Alder reactions (3). Zr(Ot Bu)4 (10 mol%) OSiMe3

O R

H 1

+ BnO

OtBu

(R)-3,3',6,6'-I4BINOL (3d) (12 mol%) PrOH (160 mol%), H2O (20 mol%) Sc(OTf)3

O BnO

toluene/t BuOMe (2:1) –20 °C, 96 h

2n

R

O 5

Entry

Aldehyde; R

Yield (%)

cis/trans

ee (%) (cis)

1 2 3 4 5 6

Ph (1a) p-MeC6 H4 (1n) p-ClC6 H4 (1g) p-NO2 C6 H4 (1o) (E)-PhCHbCH (1d) PhCH2 CH2 (1e)

95 90 Quant. 85 Quant. 54

97/3 95/5 97/3 93/7 85/15 92/8

97 94 97 90 92 81

OSiMe3 t

BuMe2SiO

Ot Bu 2m

5.4

Reaction Mechanism

In asymmetric aldol reactions the zirconium catalyst had anti selectivity irrespective of enolate geometry. This remarkable feature was in contrast with most syn-selective aldol reactions mediated by known chiral Lewis acids [22]. In the usual reactions aﬀording syn-aldol adducts the selectivity was explained in terms of steric repulsion between the alkyl groups of the aldehydes and the a-methyl groups of enolates in acyclic transition state models. In the zirconium-catalyzed reactions, anti-aldol adducts were obtained from both (E) and (Z) enolates, showing that acyclic transition states were most likely. It was speculated that the origin of the anti selectivity was steric interaction between the a-methyl groups of enolates and not the alkyl groups of aldehydes but chiral Lewis acids coordinated to carbonyl oxygen atoms (Figure 5.1) [23]. The asymmetric environment around the zirconium center seemed to be very crowded, because of the bulky iodo groups at the 3,3 0 -positions of the BINOL derivatives. Experiments in which this zirconium complex strictly recognized the structures of aliphatic aldehydes also seemed to be indicative of highly steric hindrance around the active site of the catalyst. Two mechanistic pathways have, on the other hand, been considered for HDA reactions of carbonyl compounds with Danishefsky’s diene catalyzed by Lewis acids (Scheme 5.6) [17]. One is a concerted [4þ2] cycloaddition pathway, the other a stepwise cycloaddition pathway (Mukaiyama-aldol reaction and cyclization). In most reports of HDA reactions of aldehydes with 4-substituted Danishefsky’s dienes catalyzed by chiral Lewis acids, fa-

5.4 Reaction Mechanism

L.A.

L.A. O

O

H

R2

R2

R1

H

R3

Me3SiO

OH

R1

R3

O R3

R1 R2

OSiMe3

Syn-adduct

L.A.

L.A. O

O

R2

R2

R3

R1

H

R1

H

O R3

R1

R3

Me3SiO

OSiMe3

OH

R2 Anti-adduct

Fig. 5.1

Origin of the anti-selectivity.

vored products were 2,3-cis-disubstituted pyranones [24]. In reactions with 4-methyl Danishefsky’s diene using the chiral zirconium complex, however, remarkable 2,3-trans selectivity was observed. This unique selectivity is difﬁcult to explain on the basis of the concerted [4þ2] cycloaddition mechanism, because of the disadvantage of the exo-type transition state. In almost

R2

OSiMe3 R3

R1

O

OR4

Lewis Acid H+

Concerted Pathway O R

1

O

H +

OSiMe3 R2

R3

R2 R1 OR4

R3

Stepwise Pathway

Lewis Acid

Me3SiO

H+

O OR4

R1 R

2

R

3

Scheme 5.6

Proposed reaction pathways in HDA reaction.

O

181

182

5 Zirconium Alkoxides as Lewis Acids

OSiMe3

O Ph

H

+

1a

Ot Bu

Chiral Zr Catalyst

HO

O

Ph

2l

Ot Bu

I syn/anti = 8/92 O TFA Ph

O 5al cis/trans = 8/92 98% ee (trans)

Scheme 5.7

Isolation of an intermediate of the HDA reaction.

all HDA reactions using the zirconium catalyst, it has been reported that the reaction mixture was simple, and that only one new product was observed by TLC analysis. In the reaction of benzaldehyde (1a) with 1-tertbutoxy-2-methyl-3-trimethylsiloxy-1,3-pentadiene (2l), the product was carefully isolated by use of deactivated silica gel column chromatography before treatment with TFA. It was revealed that the product isolated was the corresponding anti-aldol adduct (I) as a hydroxy-free form, and that high anti selectivity was observed (syn/anti ¼ 8:92), as shown in Scheme 5.7. In addition, the aldol adduct (I) readily cyclized quantitatively under acidic conditions to aﬀord the product with high selectivity (cis/trans ¼ 8:92, 98% ee (trans)). These facts, the observed 2,3-trans selectivity, and the isolation of the anti-aldol intermediate, indicate that the HDA reaction catalyzed by the chiral zirconium complex proceeds via a stepwise (Mukaiyama-aldol reaction and cyclization) pathway. This unique 2,3-trans selectivity can therefore be explained by the remarkable anti-selective Mukaiyama aldol reactions using the chiral zirconium catalyst system, which proceeded with anti preference irrespective of the E and Z geometry of the silicon enolates, as already mentioned. On this basis the lower reactivity of diene 2k would be understood by considering the greater stability of the tert-butyldimethylsilyloxy group than that of the trimethylsilyloxy or ethyldimethylsilyloxy group. The eﬀect of the tert-butoxy group of the dienes on the enantioselectivity could, moreover, be explained in terms of steric hindrance eﬀectively preventing the [4þ2] cycloaddition pathway. Remarkable cis selectivity obtained in the reaction with 1-benzyloxy-4-tert-butoxy-2-trimethylsilyloxy-1,3butadiene (2n) could also be accounted for by interaction of the benzyloxy group with the zirconium center of the catalyst. In the mechanism of zirconium-catalyzed aldol reactions steric repulsion between the methyl

5.4 Reaction Mechanism

Zr Me H

Zr O

O R1

OH O Me R1

H

Me anti-adduct

OSiMe3

Me3SiO

OR2

R1

O Me O R1 trans-product Zr

Zr O H

OBn R1 OSiMe3

Bn

O H

O

OH R1

Me3SiO

O OR2

R1 OBn syn-adduct O BnO

O R1 cis-product Fig. 5.2

origin of the trans- and cis-selectivity.

group of the enolate and the zirconium catalyst seemed to be an important factor explaining anti selectivity in an open-chain transition-state model. In reactions of the diene 2n coordination of the oxygen atom of the benzyloxy group would be more favored than steric repulsion, and the stereochemical outcome results in cis selectivity (Figure 5.2) [25]. An assumed catalytic cycle for this aldol and HDA reaction is shown in Scheme 5.8. First, the zirconium catalyst A is produced by mixing Zr(O t Bu)4 , (R)-3,3 0 -I2 BINOL, a primary alcohol, and H2 O. At this stage, the remaining t-butoxide groups are exchanged for the primary alcohols or H2 O. An aldehyde coordinated to this catalyst and a silicon enolate attack the carbonyl carbon of the aldehyde to generate intermediate B. The silyl group on the carbonyl oxygen is then removed by the primary alcohol, directly generating the aldol product and the original catalyst again, or moves to the most anionic atom in the same complex, the oxygen of the binaphthol, to form intermediate C. The SiaO and ZraO bonds of the intermediate C are also cleaved by the primary alcohols to form the aldol adduct and the catalyst again. This mechanism is supported by the observation that aldol adducts are obtained with free hydroxyl groups and that the trimethylsilyl ether of the alcohol and the mono trimethylsilyl ether of 3,3 0 -I2 BINOL are observed in the reaction system.

183

184

5 Zirconium Alkoxides as Lewis Acids

Zr(Ot Bu)4 + I

* O Zr

OSiMe3

O

R3

O OROR

OH OH

R1

R2 R4

H

I

+ ROH

* O I O

-

OR

RO O RO

OR

R1

Zr O

O

Zr

+O

SiMe3 R2

R3 R4

I

B

A OH

O R2

R1

ROH

R3 R 4 + ROSiMe3

* Me3SiO Zr

R1

O

O OROR COR2 R3 R4 C

Scheme 5.8

Assumed catalytic cycle of the aldol reaction.

5.5

Structure of the Chiral Zirconium Catalyst

NMR experiments were performed to clarify the structure of the chiral zirconium catalyst. The catalyst was prepared from 1 equiv. Zr(OPr)4 aPrOH, 1 equiv. 3,3 0 -I2 BINOL, and 1 equiv. H2 O in toluene-d8 . 1 H and 13 C NMR spectra were acquired at room temperature, and clear and simple signals were observed (Figure 5.3). It was revealed that this catalyst was stable in the presence of excess PrOH at room temperature and that almost the same spectra were obtained after one day. In the 13 C NMR spectrum two new kinds of signal corresponding to the naphthyl rings and two kinds of signal corresponding to the propoxide groups were observed in addition to the signals corresponding to the free BINOL. The presence of these two kinds of sharp signal suggested that the catalyst formed a dimeric structure. We

5.5 Structure of the Chiral Zirconium Catalyst

ppm

∗

∗

∗

ppm

a

The complex was prepared from Zr(OPr) 4-PrOH (1.0 equiv.), (R)-3,3'-I2BINOL (3a) (1.0 equiv.), and H2O (1.0 equiv.). ∗ : free 3,3'-I 2BINOL Fig. 5.3 1

H and

13

C NMR spectra of the zirconium complexa.

also observed characteristic signals of propoxide protons connected directly to the carbon atoms attached to the oxygen atoms at 3.8, 4.0, 4.8, and 5.2 ppm in the 1 H NMR spectrum (OaCH2 a). Integration of the proton signals indicated the presence of two kinds of propoxide moiety (one pair observed at 3.8 and 4.0 ppm and the other at 4.8 and 5.2 ppm) in the catalyst; the ratio was 2:1. The role of a small amount of water in this catalyst system was also revealed by NMR analysis [10, 26]. In the absence of PrOH and water a clear 13 C NMR spectrum was obtained from the combination of Zr(O t Bu)4 and 3,3 0 -I2 BINOL (Figure 5.4a). When PrOH was added to this system, rather complicated signals were observed (Figure 5.4b). Clear signals appeared once again when water was added to the catalyst system consisting of Zr(O t Bu)4 , 3,3 0 -I2 BINOL, and PrOH (Figure 5.4c). From these results, it was assumed that the role of water in this catalyst system was to arrange the structure of the catalyst, i.e. the desired structure was formed from the oligomeric structure by adding water.

185

186

5 Zirconium Alkoxides as Lewis Acids

a)

b) ∗

∗

∗

c) ∗

∗

∗

a) Zr(Ot Bu)4 (1.0 equiv.) + (R)-3,3'-I2BINOL (3a) (1.0 equiv.) b) Zr(Ot Bu)4 (1.0 equiv.) + (R)-3,3'-I2BINOL (3a) (1.0 equiv.) + PrOH (5.0 equiv.) c) Sample b) +PrOH (2.0 equiv.) + H 2O (1.0 equiv.) ∗ : free 3,3'-I 2BINOL

Fig. 5.4

Eﬀect of water.

Because of the dimeric structure of the catalyst, the possibility of a nonlinear eﬀect in the asymmetric aldol reaction was examined [27]. The reaction of benzaldehyde with the ketene silyl acetal derived from S-ethyl ethanethioate (2a) was chosen as a model, and the chiral Zr catalysts prepared from 3,3 0 -I2 BINOLs with lower enantiomeric excess were employed. It was found that a remarkable positive non-linear eﬀect was observed, as illustrated in Figure 5.5. After preparation of the chiral Zr catalysts from (R)3,3 0 -I2 BINOL and (S)-3,3 0 -I2 BINOL, respectively, they were combined and correlation between the ee of the zirconium catalyst and the ee of the product was investigated. A linear correlation between them was observed (Figure 5.6) [28]. These results also supported the dimeric structure of the cata-

5.6 Air-stable and Storable Chiral Zirconium Catalyst

Fig. 5.5

Correlation between the ee of the product and the ee of (R)-3,3 0 -I2 BINOL (3a) in the aldol reaction using the catalyst prepared from (R)-3,3 0 -I2 BINOLs with low ee.

lyst; on the basis of these experiments it was assumed the catalyst structure was as shown in Figure 5.7.

5.6

Air-stable and Storable Chiral Zirconium Catalyst

Although fruitful results have been obtained by use of chiral Lewis acids as catalysts in asymmetric synthesis, it has been known that Lewis acid catalysts are often sensitive to moisture and/or oxygen, even in air, and decompose rapidly in the presence of a small amount of water. Accordingly, most chiral Lewis acids must be prepared in situ under strictly anhydrous conditions just before use, often with tedious handling, and they cannot be stored for extended periods. This is also true for chiral zirconium catalysts. Development of air-stable and storable chiral Lewis acid catalysts is therefore desirable [29]. To address this issue, an air-stable, storable chiral zirconium catalyst (ZrMS) has been developed in catalytic asymmetric Mannich-type reactions [30]. This catalyst is stable in air at room temperature and is easy to handle

187

188

5 Zirconium Alkoxides as Lewis Acids

Fig. 5.6

Correlation between the ee of the product and the ee of the catalyst in the aldol reaction using the catalyst prepared by mixing (R)- and (S)-catalyst.

I O Zr O

I

H O

O

O Zr

O

O

O I

I

A Fig. 5.7

Assumed catalyst structure.

without requiring strict attention to levels of moisture and oxygen, etc. The key to this stability was combination of the catalyst and zeolite (5-A˚ molecular sieves (MS 5A)). This method should be generally applicable to zirconium catalysts. The zirconium catalyst with MS (3I-ZrMS) was prepared simply by combining the already prepared zirconium catalyst and MS 5A [31]. In the course of development of 3I-ZrMS, the amount of water in MS 5A was found to be important to achieving high enantioselectivity in the reaction

5.6 Air-stable and Storable Chiral Zirconium Catalyst Tab. 5.12

Eﬀect of storage time of 3I-ZrMS. OSiMe3

O Ph

(R)-3I-ZrMS (5 mol%)

OH O

PrOH (80 mol%) H

1a

+

OPh 2d

toluene, 0 °C, 18 h

Ph

OPh 4ad

Entry

Storage Time (weeks)

Yield (%)

syn/anti

ee (%) (anti)

1 2 3 4

0 2 6 13

Quant. Quant. Quant. Quant.

5/95 5/95 5/95 5/95

99 99 99 99

[10]. In practice 3I-ZrMS was prepared by ﬁrst combining a zirconium propoxide propanol complex (Zr(OPr)4 aPrOH) and 3,3 0 -I2 BINOL in toluene at room temperature for 3 h. MS 5A (2.5 g mmol1 ) containing 10% (w/w) H2 O was then added and the mixture was stirred for 5 min. After removal of the solvents under reduced pressure at room temperature for 1 h the 3I-ZrMS catalyst was formed. Compared with zirconium tert-butoxide Zr(OPr)4 aPrOH is an economical source of zirconium alkoxide. The aldol reaction of benzaldehyde (1a) with ketene silyl acetal 2d was then conducted using 5 mol% 3I-ZrMS catalyst in toluene at 0 C in the presence of PrOH (80 mol%). The reaction proceeded smoothly to aﬀord the desired product in high yield with high selectivity (quantitative yield, syn/anti ¼ 5/95, anti 99% ee). The result obtained by use of 3I-ZrMS was almost comparable with that obtained by use of the zirconium catalyst prepared in situ. It is worthy of note that this 3I-ZrMS catalyst was remarkably stable in air and moisture, and that the catalyst could be stored for at least 13 weeks at room temperature without loss of reactivity and selectivity (Table 5.12). The 3I-ZrMS catalyst was successfully applied to asymmetric aldol reactions of a variety of substrates; the results are summarized in Table 5.13. In reactions of benzaldehyde (1a) with other silicon enolates (2a and 2b) the 3I-ZrMS catalyst worked well and excellent yields and enantioselectivity were obtained (entries 1–3). In reactions with the ketene silyl acetal derived from phenyl propionate (2d), anti-aldol adducts were obtained with high diastereo- and enantioselectivity (entry 4). The reactions of p-methoxy- and p-chlorobenzaldehyde (1b, 1n) and a,b-unsaturated aldehyde with 2d also occurred with high diastereo- and enantioselectivity (entries 5–7). With 3phenylpropionaldehyde (1e), slight decreases of yield and selectivity were observed (entry 8). It is noted that high stereocontrol was achieved in reactions of several aldehydes and that anti-aldol adducts were obtained with excellent diastereo- and enantioselectivity. It was also found that the hetero Diels–Alder reaction of benzaldehyde (1a) with Danishefsky’s diene 2j proceeded smoothly in the presence of

189

190

5 Zirconium Alkoxides as Lewis Acids Tab. 5.13

Asymmetric aldol reactions using 3I-ZrMS. OSiMe3

O R1

1

+

R2

H

XR4 R3 2

OH O (R)-3I-ZrMS (5 mol%) PrOH (80 mol%) XR4 R1 R2 R 3 toluene, 0 °C, 18 h 4

Entry

Aldehyde; R1

Silicon Enolate

Yield (%)

syn/anti

ee (%) (anti)

1 2a 3 4 5 6 7 8

Ph (1a) Ph (1a) Ph (1a) Ph (1a) p-MeOC6 H4 (1b) p-ClC6 H4 (1n) PhCHbCH (1d) PhCH2 CH2 (1e)

2a 2a 2b 2d 2d 2d 2d 2d

Quant. 97 92 Quant. 80 Quant. 94 65

– – – 5/95 5/95 8/92 16/84 15/85

92 94 94 99 94 95 98 87

a 10

mol% catalyst. OSiMe3 SEt 2a

O Ph

OSiEtMe2 H

+

1a

Ot Bu 2j

OSiMe3

OSiMe3

OMe

OPh

2b

2d

(R)-3I-ZrMS (10 mol%) PrOH (80 mol%) toluene-tBuOMe

(2:1)

–20 °C, 18 h

O TFA Ph

O

5ah 96%, 96% ee

Scheme 5.9

Asymmetric hetero Diels–Alder reaction using 3I-ZrMS.

3I-ZrMS to aﬀord the desired product in high yield with high enantioselectivity (Scheme 5.9).

5.7

Conclusion

Asymmetric aldol reactions and hetero Diels–Alder reactions via the Mukaiyama aldol process using chiral zirconium alkoxides as Lewis acids have been discussed. Both reactions proceeded under milder conditions to aﬀord b-hydroxy esters and 2,3-dihydro-4H-pyran-4-one derivatives in high yield with high diastereo- and enantioselectivity. Addition of primary alcohols played an important role in catalyst turnover. It was discovered that a small amount of water aﬀected the structure of the catalyst and that the presence of water was essential for high enantioselectivity. The remarkable stereo-

5.8 Experimental

selectivity obtained would be ascribed to the unique steric features of the zirconium complex. Zirconium catalysts can be stored for a long time without loss of activity after stabilization on molecular sieves 5A. This methodology will contribute to practical asymmetric aldol chemistry.

5.8

Experimental Typical Experimental Procedure for Asymmetric Aldol Reactions Using Chiral Zirconium Catalyst Prepared from 3,3O-I2 BINOL (3a). Zr(O t Bu)4 (0.040 mmol) in toluene (1.0 mL) was added at room temperature to a suspension of (R)-3,3 0 -diiodo-1,1 0 -binaphthalene-2,2 0 -diol ((R)-3,3 0 -I2 BINOL, 0.048 mmol) in toluene (1.0 mL) and the solution was stirred for 30 min. Propanol (0.32 mmol) and H2 O (0.080 mmol) in toluene (0.5 mL) were then added and the mixture was stirred for 3 h at room temperature. After cooling to 0 C aldehyde 1 (0.40 mmol) in toluene (0.75 mL) and silicon enolate 2 (0.48 mmol) in toluene (0.75 mL) were successively added. The mixture was stirred for 18 h and saturated aqueous NaHCO3 (10 mL) was added to quench the reaction. After addition of dichloromethane (10 mL) the organic layer was isolated and the aqueous layer was extracted with dichloromethane (2 10 mL). The organic layers were combined and dried over anhydrous Na2 SO4 . After ﬁltration and concentration under reduced pressure the residue was treated with THF–1 m HCl (20:1) for 1 h at 0 C. The solution was then made alkaline with saturated aqueous NaHCO3 and extracted with dichloromethane. The organic layers were combined and dried over anhydrous Na2 SO4 . After ﬁltration and concentration under reduced pressure the crude product was puriﬁed by preparative thin-layer chromatography (benzene–ethyl acetate, 20:1) to aﬀord the desired aldol adduct 4. The optical purity was determined by HPLC analysis on a chiral column. For some compounds optical purity was determined after acetylation or benzoylation of the hydroxy group. Typical Experimental Procedure for Asymmetric Hetero Diels–Alder Reactions Using a Chiral Zirconium Catalyst Prepared from (R)-3,3O-I2 -6,6O-X2 BINOL (X: C2 F5 (3c), I (3d)). Zr(O t Bu)4 (0.040 mmol) in toluene (0.5 mL) was added at room temperature to a suspension of (R)-3,3 0 -diiodo-6,6 0 -disubstituted-1,1 0 binaphthalene-2,2 0 -diol ((R)-3,3 0 -I2 -6,6 0 -X2 BINOL (X: I, C2 F5 ), 0.048 mmol) in toluene (0.5 mL) and the solution was stirred for 3 h. A mixture of propanol (0.32 mmol) and H2 O (0.080 mmol) in toluene (0.3 mL) was added and the mixture was stirred for 30 min at room temperature. After cooling to 78 C, aldehyde 1 (0.40 mmol) in toluene (0.35 mL) and diene 2 (0.48 mmol) in toluene (0.35 mL) were successively added. The mixture was warmed to 20 C and stirred for 18 h. Saturated aqueous NaHCO3 (10 mL) was then added to quench the reaction. After addition of CH2 Cl2 (10

191

192

5 Zirconium Alkoxides as Lewis Acids

mL) the organic layer was isolated and the aqueous layer was extracted with CH2 Cl2 (2 10 mL). The organic layers were combined and dried over anhydrous Na2 SO4 . After ﬁltration and concentration under reduced pressure the residue was treated with TFA (0.5 mL) in CH2 Cl2 (8 mL) for 1 h at 0 C. For reactions of a,b-unsaturated aldehydes and reactions with diene 2n, scandium triﬂate (Sc(OTf )3 , 0.040 mmol, 10 mol% relative to the employed aldehyde) was used instead of TFA in CH2 Cl2 at room temperature for 12 h. The solution was made alkaline with saturated aqueous NaHCO3 (20 mL), the organic layer was isolated, and the aqueous layer was extracted with CH2 Cl2 (2 10 mL). The organic layers were combined and dried over anhydrous Na2 SO4 . After ﬁltration and concentration under reduced pressure, the trans and cis isomers were separated and puriﬁed by preparative thinlayer chromatography (benzene–ethyl acetate, 20:1). The optical purity of trans and cis isomers were determined by HPLC analysis on a chiral column (see following analytical data). For reactions with diene 2n diastereoselectivity was determined by 1 H NMR analysis of the diastereomixtures and enantioselectivity was determined by HPLC analysis of the diastereomixtures. Typical Experimental Procedure for Asymmetric Aldol Reactions Using 3IZrMS. PrOH (19.2 mg, 0.32 mmol) in toluene (0.3 mL) was added at room temperature to a suspension of 3I-ZrMS (74.5 mg, 5 mol%) in toluene (0.9 mL) and the mixture was stirred for 1 h at the same temperature. After cooling to 0 C aldehyde 1 (42.5 mg, 0.4 mmol) in toluene (0.4 mL) and silicon enolate 2 (107 mg, 0.48 mmol) in toluene (0.4 mL) were successively added and the mixture was stirred for 18 h at the same temperature. The reaction was quenched with saturated aqueous NaHCO3 , and dichloromethane (CH2 Cl2 ) was added. The organic layer was isolated, and the aqueous layer was extracted with CH2 Cl2 . The organic extracts were combined and dried over anhydrous sodium sulfate. After ﬁltration and concentration under reduced pressure the crude mixture was puriﬁed by preparative thin-layer chromatography (SiO2 , benzene–ethylacetate) to aﬀord the desired aldol adduct. The diastereomer ratio was determined by 1 H NMR analysis, and the optical purity was determined by HPLC analysis on a chiral column directly or after acetylation.

References 1 (a) Cardin, D. J.; Lappert, M. F.; Raston, C. L.; Riley, P. I.

In Comprehensive Organometallic Chemistry; Wilkinson, G.; Stone, F. G. A.; Abel, E. W. Eds.; Pergamon, New York, 1982, Vol. 3, 549; (b) Negishi, E.; Takahashi, T. Acc. Chem. Res. 1994, 72, 2591; (c) Takahashi, T.; Kotora, M.; Hara, R.; Xi, Z. Bull. Chem. Soc. Jpn. 1999, 72; for review of zirconium alkoxide in catalysis see: (d) Yamasaki, S.; Kanai, M.;

References

2

3

4

5

6

7

8 9

Shibasaki, M. Chem. Eur. J. 2001, 7, 4066; see also: (e) Krohn, K. Synthesis, 1997, 1115. (a) Evans, D. A.; McGee, L. R. Tetrahedron Lett. 1980, 21, 3975; (b) Yamamoto, Y.; Maruyama, K. Tetrahedron Lett. 1980, 21, 4607; (c) Sauve´, G.; Shwartz, D. A.; Ruest, L.; Deslongchamps, P. Can. J. Chem. 1984, 62, 2929; (d) Brown, D. W.; Campbell, M. M.; Taylor, A. P.; Zhang, X.-a. Tetrahedron Lett. 1987, 28, 985; (e) Panek, J. S.; Bula, O. A. Tetrahedron Lett. 1988, 29, 1661; (f ) Curran, D. P.; Chao, J.-C. Tetrahedron 1990, 46, 7325; (g) Yamago, S.; Machii, D.; Nakamura, E. J. Org. Chem. 1991, 56, 2098; (h) Wipf, P.; Xu, W.; Smitrovich, J. H. Tetrahedron 1994, 50, 1935. Asymmetric reactions see: (a) Evans, D. A.; McGee, L. R. J. Am. Chem. Soc. 1981, 103, 2876; (b) d’Angelo, J.; PecquetDumas, F. Tetrahedron Lett. 1983, 24, 1403; (c) Bernardi, A.; Colombo, L.; Gennari, C.; Prati, L. Tetrahedron 1984, 40, 3769; (d) Katsuki, T.; Yamaguchi, M. Tetrahedron Lett. 1985, 26, 5807; (e) Braun, M.; Sacha, H. Angew. Chem. Int. Ed. ¨ ller, D.; Braun, M. 1991, 30, 1318; (f ) Sacha, H.; Waldmu Chem. Ber. 1994, 127, 1959; (g) Vicario, J. L.; Badia, D.; Dominguez, E.; Rodriguez, M.; Carrillo, L. J. Org. Chem. 2000, 65, 3754; (h) Kurosu, M.; Lorca, M. J. Org. Chem. 2001, 66, 1205. (a) Stork, G.; Shiner, C. S.; Winkler, J. D. J. Am. Chem. Soc. 1982, 104, 310; (b) Stork, G.; Winkler, J. D.; Shiner, C. S. J. Am. Chem. Soc. 1982, 104, 3767; (c) Sasai, H.; Kirio, Y.; Shibasaki, M. J. Org. Chem. 1990, 55, 5306. (a) Mascarenhas, C. M.; Duffey, M. O.; Liu, S.-Y.; Morken, J. P. Org. Lett. 1999, 1, 1427; (b) Schneider, C.; Hansch, M. Chem. Commun. 2001, 1218; (c) Schneider, C.; Hansch, M. Synlett 2003, 837. (a) Hollis, T. K.; Robinson, N. P.; Bosnich, B. Tetrahedron Lett. 1992, 33, 6423; (b) Hollis, T. K.; Odenkirk, W.; Robinson, N. P.; Whelan, J.; Bosnich, B. Tetrahedron 1993, 49, 5415; (c) Cozzi, P. G.; Floriani, C.; Chiesi-Villa, A.; Rizzoli, C. Synlett 1994, 857; (d) Cozzi, P. G.; Floriani, C. J. Chem. Soc. Perkin Trans. 1 1995, 2557. (a) Ishitani, H.; Yamashita, Y.; Shimizu, H.; Kobayashi, S. J. Am. Chem. Soc. 2000, 122, 5403; (b) Yamashita, Y.; Ishitani, H.; Shimizu, H.; Kobayashi, S. J. Am. Chem. Soc. 2002, 124, 3292. Cox, P. J.; Wang, W.; Snieckus, V. Tetrahedron Lett. 1992, 33, 2253. Additional alcohol eﬀect in catalysis see: (a) Kawara, A.; Taguchi, T. Tetrahedron Lett. 1994, 35, 8805; (b) Kitajima, H.; Katsuki, T. Synlett 1997, 568; (c) Kitajima, H.; Ito, K.; Katsuki, T. Tetrahedron 1997, 53, 17015; (d) Yun, J.; Buchwald, S. L. J. Am. Chem. Soc. 1999, 121, 5640; (e) Evans, D. A.; Johnson, D. S. Org. Lett. 1999, 1, 595; (f ) Takamura, M.; Hamashima, Y.; Usuda, H.; Kanai, M.; Shibasaki, M. Angew. Chem. Int. Ed. 2000, 39, 1650; (g) Evans, D. A.; Scheidt, K. A.; Johnston, J. N.; Willis, M. C. J. Am. Chem. Soc. 2001, 123, 4480; (h) Onitsuka, S.; Matsuoka, Y.; Irie, R.; Katsuki, T. Chem. Lett. 2003, 32, 974.

193

194

5 Zirconium Alkoxides as Lewis Acids 10 In some metal catalyzed asymmetric reactions, water aﬀected

11

12

13 14

15 16

17 18

19

20

the yields and selectivities, Ribe, S.; Wipf, P. Chem. Commun. 2001, 299. Posner et al. and Mikami et al. also reported that a small amount of water aﬀected catalytic enantioselective ene reactions using a Ti–BINOL complex. See, (a) Posner, G. H.; Dai, H.; Bull, D. S.; Lee, J.-K.; Eydoux, F.; Ishihara, Y.; Welsh, W.; Pryor, N.; Petr Jr., S. J. Org. Chem. 1996, 61, 671; (b) Terada, M.; Matsumoto, Y.; Nakamura, Y.; Mikami, K. Chem. Commun. 1997, 281. (c) Terada, M.; Matsumoto, Y.; Nakamura, Y.; Mikami, K. J. Molecular Catalysis A: Chemical 1998, 132, 165. (d) Mikami, K.; Terada, M.; Matsumoto, Y.; Tanaka, M.; Nakamura, Y. Microporous and Mesoporous Materials 1998, 21, 461. (e) Terada, M.; Matsumoto, Y.; Nakamura, Y.; Mikami, K. Inorg. Chim. Acta 1999, 296, 267. Recent examples of anti-selective aldol reactions, (a) Parmee, E. R.; Hong, Y.; Tempkin, O.; Masamune, S. Tetrahedron Lett. 1992, 33, 1729; (b) Mikami, K.; Matsukawa, S. J. Am. Chem. Soc. 1994, 116, 4077; (c) Evans, D. A.; MacMillan, W. C.; Campos, K. R. J. Am. Chem. Soc. 1997, 119, 10859; (d) Yanagisawa, A.; Matsumoto, Y., Nakashima, H.; Asakawa, K.; Yamamoto, H. J. Am. Chem. Soc. 1997, 119, 9319; (e) Denmark, S. E.; Wong, K.-T.; Stavenger, R. A. J. Am. Chem. Soc. 1997, 119, 2333; (f ) Northrup, A. B.; MacMillan, D. W. C. J. Am. Chem. Soc. 2002, 124, 6798; (g) Yanagisawa, A.; Matsumoto, Y.; Asakawa, K.; Yamamoto, H. Tetrahedron 2002, 58, 8331; (h) Denmark, S. E.; Wynn, T.; Beutner, G. L. J. Am. Chem. Soc. 2002, 124, 13405; (i) Wadamoto, M.; Ozawa, N.; Yanagisawa, A.; Yamamoto, H. J. Org. Chem. 2003, 68, 5593. (a) Ishitani, H.; Ueno, M.; Kobayashi, S. J. Am. Chem. Soc. 1997, 119, 7153; (b) Ishitani, H.; Ueno, M.; Kobayashi, S. J. Am. Chem. Soc. 2000, 122, 8180. Yao, W.; Wang, J. Org. Lett. 2003, 5, 1527. (a) Yamashita, Y.; Saito, S.; Ishitani, H.; Kobayashi, S. Org. Lett. 2002, 4, 1221; (b) Yamashita, Y.; Saito, S.; Ishitani, H.; Kobayashi, S. J. Am. Chem. Soc. 2003, 125, 3793. Danishefsky, S.; Kitahara, T. J. Am. Chem. Soc. 1974, 96, 7807. (a) Danishefsky, S. J. Chemtracts: Org. Chem. 1989, 273; (b) Danishefsky, S. J. Aldrichimica Acta 1986, 19, 59; (c) Boger, D. L. in Comprehensive Organic Synthesis; Trost, B. M. Ed.; Pergamon Press: Oxford, 1991; Vol. 5, 451; (d) Waldmann, H. Synthesis 1994, 535. Danishefsky, S. J.; Larson, E.; Askin, D.; Kato, N. J. Am. Chem. Soc. 1985, 107, 1246. tert-Butyl methyl ether was used as an eﬃcient solvent in asymmetric HDA reactions previously. See: Schaus, S. E.; Bra˚nalt, J.; Jacobsen, E. N. J. Org. Chem. 1998, 63, 403. (a) Kobayashi, S. Synlett 1994, 689; (b) Kobayashi, S. Eur. J. Org. Chem. 1999, 15; (c) Kobayashi, S.; Sugiura, M.; Kitagawa, H.; Lam, W. W.-L. Chem. Rev. 2002, 102, 2227. Danishefsky, S. J.; Maring, C. J. J. Am. Chem. Soc. 1985, 107, 1269.

References 21 (a) Danishefsky, S. J.; Webb II, R. R. J. Org. Chem. 1984, 49,

22

23

24

25 26

27 28 29

30 31

1955; (b) Danishefsky, S. J.; Maring, C. J. J. Am. Chem. Soc. 1985, 107, 1269. Reviews of catalytic asymmetric aldol reactions, (a) Bach, T. Angew. Chem. Int. Ed. Engl. 1994, 33, 417; (b) Nelson, S. G. Tetrahedron: Asymmetry. 1998, 9, 357; (c) Groger, H.; Vogel, E. M.; Shibasaki, M. Chem. Eur. J. 1998, 4, 1137; (d) Mahrwald, R. Chem. Rev. 1999, 99, 1095; (e) Johnson, J. S.; Evans, D. A. Acc. Chem. Res. 2000, 33, 325; (f ) Machajewski, T. D.; Wong, C.-H. Angew. Chem. Int. Ed. 2000, 39, 1352; (g) List, B. Tetrahedron 2002, 58, 5573; (h) Alcaide, B.; Almendros, P. Eur. J. Org. Chem. 2002, 1595; (i) Palomo, C.; Oiarbide, M.; Garcia, J. M. Chem. Eur. J. 2002, 8, 37; ( j) Carreira, E. M. in Comprehensive Asymmetric Catalysis; Jacobsen, E. N.; Pfaltz, A.; Yamamoto, H. Eds.; Springer: Heidelberg, 1999. Vol. 3, p 998; (k) Carreira, E. M. in Catalytic Asymmetric Synthesis 2 nd Edition, I. Ojima Ed.; Wiley–VCH, New York, 2000, p 513; (l) Sawamura, M.; Ito, Y. in Catalytic Asymmetric Synthesis 2 nd Edition, I. Ojima Ed.; Wiley–VCH, New York, 2000, p 493. (a) Mukaiyama, T.; Kobayashi, S.; Murakami, M. Chem. Lett. 1985, 447. (b) Gennari, C.; Beretta, M. G.; Bernardi, A.; Moro, G.; Scolastico, C.; Todeschini, R. Tetrahedron 1986, 42, 893. Reviews of catalytic asymmetric hetero Diels–Alder reactions, (a) Ooi, T.; Maruoka, K. in Comprehensive Asymmetric Catalysis; Jacobsen, E. N.; Pfaltz, A.; Yamamoto, H. Eds.; Springer: Heidelberg, 1999. Vol. 3, p 1237; (b) Jørgensen, K. A. Angew. Chem. Int. Ed. Engl. 2000, 39, 3558. See also references in ref. 14b. Kobayashi, S.; Ueno, M.; Ishitani, H. J. Am. Chem. Soc. 1998, 120, 431. (a) Hanawa, H.; Hashimoto, T.; Maruoka, K. J. Am. Chem. Soc. 2003, 125, 1708; (b) Hanawa, H.; Uraguchi, D.; Konishi, S.; Hashimoto, T.; Maruoka, K. Chem. Eur. J. 2003, 9, 4405. Girard, C.; Kagan, H. B. Angew. Chem. Int. Ed. 1998, 37, 2922. Mikami, K.; Motoyama, T.; Terada, M. J. Am. Chem. Soc. 1994, 116, 2812. A stable chiral La catalyst has been reported: Y. S. Kim, S. Matsunaga, J. Das, A. Sekine, T. Ohshima and M. Shibasaki, J. Am. Chem. Soc. 2000, 122, 6506. M. Ueno, H. Ishitani and S. Kobayashi, Org, Lett, 2002, 4, 3395. Kobayashi, S.; Saito, S.; Ueno, M.; Yamashita, Y. Chem. Commun. 2003, 2016.

195

197

6

Direct Catalytic Asymmetric Aldol Reaction Using Chiral Metal Complexes Masakatsu Shibasaki*, Shigeki Matsunaga, and Naoya Kumagai 6.1

Introduction

The aldol reaction has established a position in organic chemistry as a remarkably useful synthetic tool which provides access to b-hydroxy carbonyl compounds and related building blocks. Intensive eﬀorts have raised this classic process to a highly enantioselective transformation employing only catalytic amounts of chiral promoters, as reviewed in this handbook [1]. Although many eﬀective applications have been reported, most methods necessarily involve the preformation of latent enolates such as ketene silyl acetals, by use of less than stoichiometric amounts of base and silylating reagents (Scheme 6.1, top). Because of an increasing demand for environmentally benign and atom-eﬃcient processes, such stoichiometric amounts of reagents, which inevitably result in waste, for example salts, should be excluded from the procedure. Thus, the development of a direct catalytic asymmetric aldol reaction (Scheme 6.1, bottom), which employs unmodiﬁed ketone as a donor, is desired. The clue for success in achieving the direct enolization of unmodiﬁed ketone with a catalytic amount of reagent is found in enzymatic reactions. (a) Mukaiyama-type reactions A: SiR3 or CH3

O

base (1 equiv.)

R2 (b) Direct reactions O R2

O

A

AO

chiral catalyst

R2

R1CHO

R2

R1

HO

chiral catalyst R1CHO

O

O

R1

Scheme 6.1

(a) Mukaiyama-type reactions and (b) direct reactions. Modern Aldol Reactions. Vol. 2: Metal Catalysis. Edited by Rainer Mahrwald Copyright 8 2004 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 3-527-30714-1

R2

198

6 Direct Catalytic Asymmetric Aldol Reaction Using Chiral Metal Complexes

O HO

OH

O

aldolase

OPO32-

OPO32OH

OH Tyr O H

H Glu

O

CO2–

H

O

His Zn His His O

H O

His Zn His O– His

OH O

OPO32HO OH

OH

Ser Thr

Ser

O

NH2

OPO32-

O HO

Asn

OH OH

NH2

OH

Asn

Ser Thr

Ser

Fig. 6.1

Mechanism of action of the class II aldolase fructose-1-phosphate aldolase.

Aldolases eﬃciently promote the direct aldol reaction under mild in-vivo conditions – fructose-1,6-bisphosphate and dihydroxyacetone phosphate aldolases are typical examples. Such enzymes function as a bifunctional catalyst, activating a donor (ketone) with Brønsted basic functionality and an acceptor (aldehyde) with an acidic functionality. As shown in Figure 6.1, the mechanism of class II aldolases (e.g. fructose-1-phosphate aldolase) is thought to involve co-catalysis by a Zn 2þ cation and a basic functional group in the enzyme’s active site [2]. The Zn 2þ functions as a Lewis acid to activate a carbonyl group, and the basic part abstracts an a-proton to form a Znenolate. Synthetic organic chemists regarded this bifunctional mechanism of the aldolases as a very promising strategy for achieving direct catalytic asymmetric aldol reactions with an artiﬁcial small molecular catalyst. This chapter focuses on notable advances recently achieved by use of metallic catalysis. This catalysis mimics that of the class II aldolases [3]. Important early contributions on direct catalytic asymmetric aldol reactions using Au as catalyst by Ito, Hayashi, and their coworkers [4] are described in Chapter 1 of Part II of this book. The other type of the direct aldol reaction, the organocatalysis as mimics of class I aldolases (amino acid based mechanism) [5] is discussed in Chapter 4 of Part I of this book.

6.2

Direct Aldol Reactions with Methyl Ketones

A possible catalytic cycle for achieving direct catalytic asymmetric aldol reaction by means of bifunctional metallic catalysis is shown in Scheme 6.2,

6.2 Direct Aldol Reactions with Methyl Ketones

R2

R2

O R1

H

II

R2

*

*

O M O LA I

O M O H O LA O

*

OH O R1

R2

O O M H O O LA

H R1

R2

LA : Lewis acid M : Metal of Brønsted base III

O : Chiral ligand

*

*

O

O M O H O LA

O

R1

IV

Scheme 6.2

Possible catalytic cycle for direct catalytic asymmetric aldol reactions.

which involves synergistic action of the Brønsted basic and the Lewis acidic moieties in the catalyst. The Brønsted base functionality (OM) in the catalyst I deprotonates an a-proton of a ketone to generate the metal enolate II. Lewis acid functionality activates an aldehyde to give III. The activated aldehyde then reacts with the metal enolate in a chelation-controlled asymmetric environment to aﬀord a b-keto metal alkoxide IV. Proton exchange between the metal alkoxide moiety and an aromatic hydroxy proton or an

La(O-i-Pr)3 +

3

OH OH

+

3 BuLi

(S)-BINOL Li O O O Li La O O O Li

Fig. 6.2

The structure of LaLi3 tris((S)-binaphthoxide) ((S)-LLB).

199

200

6 Direct Catalytic Asymmetric Aldol Reaction Using Chiral Metal Complexes

O

R1CHO

+

R2 2 (1.5–50 equiv.)

1

(S)-LLB (20 mol %) THF, -20 °C

OH O R2 3 y. 28-90% 44-94% ee

R1

Scheme 6.3

Direct catalytic asymmetric aldol reactions of methyl ketone 2 promoted by (S)-LLB.

a-proton of a ketone leads to the optically active aldol adduct, and at the same time leads to regeneration of the catalyst. In 1997, Shibasaki reported that a heterobimetallic (S)-LaLi3 tris(binaphthoxide) complex (Figure 6.2, (S)-LLB), prepared from La(O-i-Pr)3 , BINOL and BuLi, was eﬀective in the direct catalytic asymmetric aldol reaction of unmodiﬁed ketones (Scheme 6.3) [6]. As shown in Table 6.1, the LLB catalyst was eﬀective for methyl aryl ketones (2a and 2b) and methyl alkyl ketones

Tab. 6.1

Direct catalytic asymmetric aldol reactions of methyl ketone 2 promoted by (S)-LLBa. Entry

Aldehyde

Ketone (equiv.)

Product

1a

-Ph

2a (5)

3aa

2 3

1a 1a

-Ph -Ph

2a (1.5) 2a (10)

4

1a

5

1a

CHO

1

6

Ph

CHO

7b 8 CHO

9

CHO

10 11

Ph

Yield (%)

ee (%)

88

76

88

3aa 3aa

135 91

43 81

87 91

2b (8)

3ab

253

55

76

-CH3

2c (10)

3ac

100

53

73

1b

-Ph

2a (7.4)

3ba

87

90

69

1b 1b

-CH3 -CH2 CH3

2c (10) 2d (50)

3bc 3bd

185 185

82 71

74 94

1c

-Ph

2a (8)

3ca

169

72

44

1d

-Ph

2a (8)

3da

277

59

54

1e

-Ph

2a (10)

3ea

72

28

52

conditions: (S)-LLB (20 mol%), THF, 20 C. reaction was carried out at 30 C.

a Reaction b The

CHO

Time (h)

6.2 Direct Aldol Reactions with Methyl Ketones

O

O

H Me OMOM 4 (>99% ee)

O

+

OH O

(S)-LLB (10 mol %)

O

THF, –20 °C, 16 h TMS 5 (6 equiv.)

TBS OH O

Me OMOM 6 y. 65% α/β = 3.6/1

OTBDPS

O Me OTES

7

Scheme 6.4

Application of (S)-LLB to the formal total synthesis of fostriecin.

(2c and 2d). Excess ketone was necessary to achieve good yield (entries 1–3). Although the reactivity was moderate even using 20 mol% catalyst, the enantiomeric excess reached 94% ee. The results demonstrated that the concept shown in Scheme 6.2 was possible. As shown in Scheme 6.4, LLB was applicable to optically active functionalized aldehyde 4 and acetylenic ketone 5. The aldol reaction of 4 and 5 proceeded smoothly with 10 mol% (S)-LLB at 20 C to aﬀord 6 in 65% yield. Diastereoselectivity was 3.6:1 and the desired a-OH 6 was obtained as a major product. Compound 6 was successfully converted into known intermediate, 7, of fostriecin [7]. It is worthy of note that the aldol reaction did not proceed when a standard base, for example LDA, was used.

R1CHO

(S)-LLB (8 mol %) KHMDS (7.2 mol %) H2O (16 mol %)

O + 2

THF

R 2 (3–15 equiv.)

1

OH O R

1

R2 3 y. 50-91% 30-93% ee

KOH

(S)-LLB 1 equiv.

+

KHMDS 0.9 equiv.

+

H 2O

2 equiv.

Li O O O La Li O O O Li

(S)-LLB•KOH complex Scheme 6.5

Direct catalytic asymmetric aldol reactions of methyl ketone 2 promoted by (S)-LLB– KOH.

201

TMS

202

6 Direct Catalytic Asymmetric Aldol Reaction Using Chiral Metal Complexes

Tab. 6.2

Direct catalytic asymmetric aldol reactions of methyl ketone 2 promoted by (S)-LLB–KOHa. Entry

Aldehyde

Ketone (equiv.)

CHO

1 2 3 4 5

CHO

Ph

6b 7

CHO

BnO

8 CHO

9 10

CHO

11

CHO

12 13

2

Ph

CHO

Product

Temp (˚C)

Time (h)

Yield (%)

ee (%)

1a

-Ph

2a (5)

3aa

20

15

75

88

1b 1b 1b

-Ph -CH3 -CH2 CH3

2a (5) 2c (10) 2d (15)

3ba 3bc 3bd

20 20 20

28 20 95

85 62 72

89 76 88

1b

-Ph

2a (5)

3ba

20

18

83

85

1b

-Ph

2a (5)

3ba

20

33

71

85

1f

-Ph

2a (5)

3fa

20

36

91

90

1f

-Ph

2a (5)

3fa

20

24

70

93

1d

-Ph

2a (5)

3da

30

15

90

33

1d

-m-NO2 -C6 H4

2e (3)

3de

50

70

68

70

1g

-m-NO2 -C6 H4

2e (3)

3ge

45

96

60

80

1h

-m-NO2 -C6 H4

2e (5)

3he

50

96

55

42

1e

-m-NO2 -C6 H4

2e (3)

3ee

40

31

50

30

a Reaction

conditions: (S)-LLB (8 mol%), KHMDS (7.2 mol%), H2 O (16 mol%), THF, unless otherwise noted. b (S)-LLB (3 mol%), KHMDS (2.7 mol%), H O (6 mol%), THF. 2

Substantial acceleration of this reaction was achieved using a (S)-LLB– KOH catalyst prepared from (S)-LLB, KHMDS, and H2 O; this enabled reduction of the catalyst loading from 20 to 3–8 mol% with a shorter reaction time (Scheme 6.5) [8]. The results are summarized in Table 6.2. Aldol adducts were obtained in good yield and in moderate to good enantioselectivity (30–93% ee). The LLB–KOH complex was also applicable to the reaction between cyclopentanone and aldehyde 1b, aﬀording the aldol adduct 3bf syn selectively (syn/anti ¼ 93:7, syn:76% ee, anti:88% ee) in 95% yield (Scheme 6.6). Kinetic studies, including initial rate kinetics and isotope eﬀects, indicated that the rate-determining step was deprotonation of the ketone. Additional KOH is supposed to accelerate the rate-determining enolization step. Because high ee was achieved in the presence of additional KOH, self-assembly of the LLB complex and KOH was assumed. On the basis of mechanistic studies a working model was proposed for the direct

6.2 Direct Aldol Reactions with Methyl Ketones (S)-LLB (8 mol %) KHMDS (7.2 mol %) H2O (16 mol %) THF, 99 h, –20 °C

O CHO

Ph

+

1

2f (5 equiv.)

OH O Ph

3bf y. 95%, syn/anti = 93/7 syn: 76% ee, anti: 88% ee

Scheme 6.6

Direct catalytic asymmetric aldol reactions of cyclopentanone (2f ) promoted by (S)-LLB–KOH.

O

* O Li

O

Li O

O

La

O R1

Li

2

R

O

O Li K H

R2

O O

La

O O

O

*

O

O

Li

O

O

La

*

O

O

Li

*

K

*

R2

O

O

O La

1

H

R

O

R1

Li

*

proposed intermediate

* O Li O K R1

O

R1

O

Li O

*

H

O

H H

Li

H

* *

rate-determining step

Li

* H

R2

*

O

O KOH

H

O Li H H O

Li O O

La O

R2

O

*

O

O

H

*

Li K O

R2

Fig. 6.3

Working model for direct catalytic asymmetric aldol reactions promoted by the (S)-LLB–KOH complex.

catalytic asymmetric aldol reaction promoted by the LLB–KOH complex (Figure 6.3). KOH functions as a Brønsted base, generating an enolate from the ketone (rate-determining step); the lanthanum ion acts as a Lewis acid to activate the aldehyde. 1,2-Addition and the protonation of an alkoxide leads to the aldol adduct and regenerates the catalyst. As shown in Scheme 6.7, a key-intermediate of bryostatin 7 was synthesized by using the LLB– KOH complex twice. 3fa was prepared by the aldol reaction using (R)LLB–KOH. Baeyer–Villiger oxidation, followed by functional group manipulation, aﬀorded aldehyde 9. The aldol reaction of 9 with (S)-LLB–KOH proceeded in a catalyst-controlled manner and the anti adduct was obtained (anti/syn ¼ 7:1). The LLB–KOH complex was also applied to the total synthesis of epothilones; the LLB–KOH complex was eﬀectively applied for

203

204

6 Direct Catalytic Asymmetric Aldol Reaction Using Chiral Metal Complexes

OH O

OH O a

Ph

BnO

O

OPh

BnO

3fa

TBSO b

H

BnO

8

9 TBSO

O OH

BnO

O

(S)-LLB•KOH (20 mol %) ketone 2a (5 equiv.)

OH O

THF, –20 °C

Ph

BnO

10 y. 90% (anti/syn = 7/1)

11

(a) mCPBA, NaH2PO4, ClCH2CH2Cl, y. 73%; (b) i)TBSOTf, Hünig base; ii) DIBAL; iii) PCC y. 87% (3 steps). Scheme 6.7

Application to the synthesis of intermediate to bryostatin 7.

O RCHO

1

+

Ph 2a (2 equiv.)

OCH3 + Ba(O-i-Pr)2 OH

12

(R)-Ba-12 (5 mol %) DME, -20 °C

DME

OH O Ph 3 y. 77-99% 50-70% ee

R

CH3 O O Ba O O H3C

Ba-12 (proposed structure)

Scheme 6.8

Direct catalytic asymmetric aldol reactions of acetophenone (2a) promoted by (R)-Ba-12.

resolution of the racemic aldehyde [9]. Further details are given in Chapter 7 in Vol. 1. To reduce the amount of ketone Shibasaki prepared a Ba-12 complex from Ba(O-i-Pr)2 and 12 [10]. As shown in Scheme 6.8, the Ba-12 complex afforded aldol adducts in good yield (77–99%) from as little as 2 equiv. ketone 2a, although ee was modest (50–70% ee). In 2000 Trost reported a dinuclear Zn complex prepared from Et2 Zn and 13 [11]. 13 was easily synthesized from p-cresol in four steps. On the basis of ethane gas emission measurement and ESI-MS analysis the dinuclear complex Zn2 -13 was proposed (Scheme 6.9). As shown in Table 6.3, the Zn complex was eﬀective in direct catalytic asymmetric aldol reactions with a variety of methyl aryl ketones (2a, 2g–2j; 10 equiv.). Excellent enantioselectivity (up to 99% ee) was achieved by use of 5 mol% 13, although excess ketone was used and yields were occasionally moderate. It is worthy of note that high ee was achieved at relatively high reaction temperature

6.2 Direct Aldol Reactions with Methyl Ketones

205

Tab. 6.3

Direct catalytic asymmetric aldol reactions of methyl ketone 2 promoted by dinuclear Zn2 -13 complexa. Entry

Aldehyde

1

Ketone (Equiv.)

CHO

2 3

CHO CHO

4

CHO

5

6

Ph

CHO

Product

Temp (˚C)

Time (h)

Yield (%)

ee (%)

1i

-Ph

2a (10)

3ia

5

48

33

56

1i

-Ph

2a (10)

3ia

15

48

24

74

1j

-Ph

2a (10)

3ja

5

48

49

68

1d

-Ph

2a (10)

3da

5

48

62

98

1c

-Ph

2a (10)

3ca

5

48

60

98

1k

-Ph

2a (10)

3ka

5

48

79

99

1l

-Ph

2a (10)

3la

5

48

67 (dr: 2/1)

94

1m

-Ph

2a (10)

3ma

5

96

61

93

2g (10)

3dg

5

48

66

97

2h (10)

3dh

5

48

48

97

2i (5)

3di

5

48

36

98

2j (5)

3dj

5

48

40

96

Ph

7

8

Ph

CHO

TBSO

9

CHO

CHO

1d

O

OMe

10

1d

11

1d OMe

12

1d

a Reaction

conditions: ligand 13 (5 mol%), Et2 Zn (10 mol%), Ph3 PbS (15 mol%), THF.

(5 C). Bifunctional Zn catalysis is proposed, with one Zn acting as Lewis acid and another Zn-alkoxide functioning as a Brønsted base to generate a Zn-enolate (Figure 6.4). Trost also reported that modiﬁcation of the ligand occasionally led to better results. When ligand 14 was used instead of 13 the direct aldol reaction of acetone (10–15 equiv.) proceeded smoothly with good enantioselectivity (Scheme 6.10) [12]. The results are summarized in Table 6.4. Aldol adducts were obtained in good yield (59–89%) and ee (78–94%). In

206

6 Direct Catalytic Asymmetric Aldol Reaction Using Chiral Metal Complexes

O

R1CHO

+

HO

OH N

OH

Ph Ph

O

Ar 3 y. 24-79% 56-99% ee

Ph Ph

N

H O R1

MS 4A, THF

Ar 2 (5–10 equiv.)

1

Ph Ph

ligand 13 (5 mol %) Et2Zn (10 mol %) Ph3P=S (15 mol %)

Et O O Zn Zn N N O

Ph Ph

Et2Zn (2 equiv.)

THF CH3 13

CH3 Zn2-13 (proposed structure)

Scheme 6.9

Direct catalytic asymmetric aldol reactions of methyl ketone 2 promoted by dinuclear Zn2 -13 complex.

R

Ar O Ph Ph

Zn Zn N

H

O

O O

O

Ph Ph

N

CH3 Fig. 6.4

Proposed transition state for the direct aldol reaction of methyl ketone 2.

general, self-condensation of aldehydes should be suppressed to achieve good yield by using a-unsubstituted aldehydes. It is worthy of note that good yield and high ee were achieved even with a-unsubstituted aldehydes (entries 5–7, yield 59–76%, 82–89% ee). Noyori prepared a highly active Ca-15 complex from Ca[N{Si(CH3 )3 }2 ]2 and 15 (Scheme 6.11) [13]. As shown in Table 6.5, aldol adducts were obtained in good yield (75–88%) in moderate to good ee (66–91% ee) by using as little as 1–3 mol% catalyst loading. The reactivity of the Ca catalyst is higher than those of other catalysts. The enhanced reactivity is ascribed to the high basicity of the calcium alkoxide. Even with this active Ca-15 catalyst, however, excess ketone was essential to achieve good yield and ee. This problem would be solved in future research.

6.2 Direct Aldol Reactions with Methyl Ketones

ligand 14 (x mol %) Et2Zn (2x mol %)

O RCHO

1

+

2c (10-15 equiv.) MS 4A, THF, 5 °C, 48 h (x = 5-10 mol %)

OH O

HO

OH

Ph Ph

N

OH

3 y. 59-89% 78-94% ee

H3C

14

Direct catalytic asymmetric aldol reactions of acetone (2c) promoted by dinuclear Zn2 -14 complex.

Tab. 6.4

Direct catalytic asymmetric aldol reactions of acetone (2c) promoted by dinuclear Zn2 -14 complexa. Aldehyde

CHO

1

CHO

2

CHO

3

4

Ph

CHO

Product

Catalyst (mol%)

Yield (%)

ee (%)

1c

3cc

10

89

92

1d

3dc

10

89

91

1a

3ac

10

72

94

1k

3kc

10

84

91

1j

3jc

10

59

84

1e

3ec

10

76

82

1i

3ic

10

69

89

1n

3nc

5

78

83

1o

3oc

5

62

78

Ph

5 6

CHO Ph

7

CHO CHO CHO

8

CHO

9 O 2N a Reaction

N

R

Scheme 6.10

Entry

207

conditions: ligand 14 ( mol%), Et2 Zn (2 mol%), THF, MS 4A, 5 C, 48 h.

CH3

Ph Ph

208

6 Direct Catalytic Asymmetric Aldol Reaction Using Chiral Metal Complexes

O RCHO

+

C2H5CN/THF, –20 °C

Ph 2 (10 equiv.)

1

H O

(S,S)-Ca-15 (x mol %)

R

Ph 3 y. 75-88% 66-91% ee

(x = 1-3 mol %)

[(CH3)3Si]2N

O

OH Ca(thf)2

+

+

[(CH3)3Si]2N

Ca-15 complex

KSCN

OH

1 equiv.

15 3 equiv.

1 equiv.

Scheme 6.11

Direct catalytic asymmetric aldol reactions of acetophenone (2a) promoted by Ca-15 complex.

6.3

Direct Aldol Reactions with Methylene Ketones

The aldol reaction between aldehydes and methylene ketones or propionates should provide a powerful tool for construction of two continuous chiral centers and for formation of carbon–carbon bonds. Catalytic asymmetric syntheses of syn and anti aldols from latent enolates have already been well investigated [1]. In contrast, diastereo- and enantioselective synthesis of aldols, starting from methylene ketones, by means of the direct catalytic asymmetric aldol reaction is still immature. The bulkiness of methylene ketones was expected to make it more diﬃcult for the catalysts to abstract an a-hydrogen from the ketones. Shibasaki reported a strongly basic La-Li-16 complex with an Li alkoxide Tab. 6.5

Direct catalytic asymmetric aldol reactions of acetophenone (2a) promoted by Ca-15 complexa. Entry

Aldehyde

CHO

1 2 3

Ph

4

BnO

5

a Reaction

CHO

CHO

CHO

Product

Catalyst (mol%)

Time (h)

Yield (%)

ee (%)

1a

3aa

3

22

87

86

1a

3aa

1

20

79

82

1b

3ba

3

24

75

87

1f

3fa

3

24

76

91

1c

3ca

3

20

88

66

conditions: (S,S)-Ca-15 ( mol%), C2 H5 CN/THF, 20 C.

6.3 Direct Aldol Reactions with Methylene Ketones

CHO

BnO

O

+

1f (1.5 equiv.)

La-Li-16 complex (20 mol %) LiI (0–60 mol %)

2k (5 equiv.)

O OH OH

HO

toluene –20 °C, 6 d

OH O BnO

La(O-i-Pr)3 (1 equiv.) BuLi (3 equiv.)

HO HO

y. 5-38% 4-51% ee

3fk

O O La O O O Li LiO OLi La-Li-16 complex (proposed structure)

OH

16

209

Scheme 6.12

Direct catalytic asymmetric aldol reactions of 3-pentanone (2k) promoted by La-Li-16 complex.

moiety. The catalyst promoted the direct aldol reaction of 3-pentanone antiselectively; yield and ee were only modest, however (Scheme 6.12) [14]. A notable advance in the direct aldol reaction of methylene ketone was reported by Mahrwald in 2002 [15]. As shown in Scheme 6.13, a catalytic amount (10 mol%) of Ti-17-rac-BINOL complex was suitable for promoting the direct asymmetric aldol reaction of 3-pentanone with a variety of aldehydes. Interestingly, combination of chiral mandelic acid (17) and racBINOL aﬀorded a good chiral Ti-catalyst. As summarized in Table 6.6, aldol adducts were obtained syn-selectively (syn/anti up to 91:9) in moderate to good yield (43–85%) and with good ee (71–93%). Good selectivity was achieved at room temperature. The aldehyde/ketone ratio in this reaction is

rac-BINOL2Ti2(O-i-Pr)3 /(R)-17

O RCHO

1 (1.5 equiv.)

room temperature

2k(1 equiv.)

OH

+ Ph

R

3

y. 43-85% syn/anti =72/28–91/9 71-93% ee

OH OH

OH O

(10 mol %)

+

OH

+ Ti(O-i-Pr)4

O (R)-mandelic acid (17)

rac-BINOL Scheme 6.13

Direct catalytic asymmetric aldol reactions of 3-pentanone (2k) promoted by racBINOL2Ti2 (O-i-Pr)3 /(R)-17 complex.

rac-BINOL2Ti2(O-i-Pr)3 /(R)-17 complex

210

6 Direct Catalytic Asymmetric Aldol Reaction Using Chiral Metal Complexes Tab. 6.6

Direct catalytic asymmetric aldol reactions of 3-pentanone (2k) promoted by rac-BINOL2Ti 2 (O-i-Pr)3 /(R)-17 complexa. Entry

Aldehyde

1

PhCHO CHO

2 3 4 5

Ph

CHO CHO

CHO

Product

Catalyst (mol%)

dr (syn/ anti)

Yield (%)

ee (%)

1n

3nk

10

91/9

85

91

1a

3ak

10

88/12

71

93

1p

3qk

10

73/27

68

78

1d

3dk

10

79/21

43

71

1q

3qk

10

72/28

78

74

a Reaction

conditions: rac-BINOL2Ti 2 (O-i-Pr)3 /(R)-17 (10 mol%), room temperature.

also worthy of note. The aldol reaction proceeded smoothly with 1 equiv. ketone and 1.5 equiv. aldehyde.

6.4

Direct Aldol Reaction with a-Hydroxyketones

a-Hydroxyketones also serve as aldol donors in the direct catalytic asymmetric aldol reaction. In contrast with conventional Lewis acid-catalyzed aldol reactions, protection of the OH group is not necessary. The versatility of the resulting chiral 1,2-diols as building blocks makes this process attractive. The ﬁrst successful result was reported by List and Barbas with proline catalysis [16]. l-Proline catalyzed a highly chemo-, diastereo- and enantioselective aldol reaction between hydroxyacetone and aldehydes to provide chiral anti-1,2-diols. Trost and Shibasaki have made important contributions to metallic catalysis. Trost reported a direct aldol reaction with 2-hydroxyacetophenone and 2-hydroxyacetylfuran using the dinuclear Zn2 -13 catalyst [17] (Scheme 6.14). The aldol reaction between a variety of aldehydes and 1.5 equiv. ketone proceeded smoothly at 35 C with 2.5–5 mol% of catalyst, in the presence of MS 4A, to aﬀord the products syn-selectively (syn/anti ¼ 3:1 to 100:0) in 62–98% yield and 81–98% ee (Table 6.7). Strikingly, the absolute conﬁguration of the stereocenter derived from the aldehyde is opposite to that obtained with acetophenone (2a, Scheme 6.9) and acetone (2c, Scheme 6.10) as donors, possibly because of the bidendate coordination of a-hydroxyketone to the catalyst, as depicted in Figure 6.5. Occasionally the

6.4 Direct Aldol Reaction with a-Hydroxyketones

+

RCHO

Ar

MS 4A, THF, –35 °C

OH (x = 2.5–5 mol %) 18a: Ar = Ph 18b: Ar = 2-furyl (1.1-1.5 equiv.)

1

H O

ligand 13 (x mol %) Et2Zn (2x mol %)

O

Ph Ph

O

R

211

HO

OH N

OH

N

Ar

O H 19 y. 65-97% syn/anti = 4/1–100/0 86-98% ee

CH3 13

Scheme 6.14

Direct catalytic asymmetric aldol reaction of hydroxyketone 18 promoted by Zn2 -13 complex.

Tab. 6.7

Direct catalytic asymmetric aldol reaction of hydroxyketone 18 promoted by dinuclear Zn2 -13 complexa. Entry

Aldehyde

CHO

1 2 3 4

CHO

5 6 7

Ph

CHO

Ketone (Equiv.)

Product

Catalyst (Dmol%)

Yield (%)

dr (syn/ anti)

ee (%) (syn)

1c

18a (1.5)

19ca

2.5

83

30/1

92

1c 1c 1c

18a (1.5) 18b (1.3) 18b (1.1)

19ca 19cb 19cb

5 5 5

97 90 77

5/1 6/1 6/1

90 96 98

1d

18a (1.5)

19da

2.5

89

13/1

93

1d

18a (1.1)

19da

5

72

6/1

93

1k

18a (1.5)

19ka

2.5

74

100/0

96

1j

18a (1.5)

19ja

2.5

65

35/1

94

1j

18a (1.1)

19ja

5

79

4/1

93

1e

18a (1.5)

19ea

2.5

78

9/1

91

1r

18a (1.5)

19ra

5

89

5/1

86

1s

18a (1.5)

19sa

5

91

5/1

87

Ph

8

CHO

9 10

CHO

Ph

11 12 a Reaction

CHO 4

CHO 6

conditions: ligand 13 ( mol%), Et2 Zn (2 mol%), THF, MS 4A, 35 C, 24 h.

Ph Ph

212

6 Direct Catalytic Asymmetric Aldol Reaction Using Chiral Metal Complexes

Ar Ph Ph

R H O OH O Zn

O O Zn N

O

Ph Ph

N

CH3 Fig. 6.5

Proposed transition state for the direct aldol reaction of hydroxyketone 18.

reaction has been performed with a ketone/aldehyde ratio of 1.1:1.0 albeit at the expense of conversion, which comes closest in reaching the ideal atom economical process. The reaction with 18b resulted in higher ee and, moreover, the furan moiety is suitable for further conversion of the resulting chiral 1,2-diol. Oxidative cleavage of the furan ring was successfully used for asymmetric synthesis of (þ)-boronolide, as shown in Scheme 6.15 [18]. Shibasaki developed direct catalytic asymmetric aldol reactions of 2hydroxyacetophenones, providing either anti or syn chiral 1,2-diols, by using two types of multifunctional catalyst, (S)-LLB–KOH and an Et2 Zn/(S,S)linked-BINOL 20 complex [19]. (S)-LLB–KOH (5–10 mol%) promoted the direct aldol reaction of 2-hydroxyacetophenones 18 to aﬀord anti-1,2-diols in good yields and ee (up to 98% ee), although anti-selectivity was occasionally modest. (Scheme 6.16 and Table 6.8) [19, 20]. Enolizable aldehydes were successfully utilized without any self-condensation. The absolute conﬁguration was identical at the a-position of both anti and syn products, suggesting that the enantioface of the enolates derived from 2-hydroxyacetophenones

O +

H

18b 1.1 eq

1h: 16 mmol scale OAc OAc

O

a(a)

a

OH

OAc (+)-boronolide

TBSO O

O OH

OH

O

O

O

O

3

O 20

19bh TBSO MeO O

O

O b

O 21

3

5 mol % of Zn catalyst, MS 4A, THF, –35 °C, 12 h, 93% (syn/anti = 4.2/1, syn = 96% ee);

(b) RuCl3 (cat.), NaIO4, CCl4, CH3CN, H2O; CH2N2, Et2O, 70%. Scheme 6.15

Stereocontrolled total synthesis of (þ)-boronolide.

3

6.4 Direct Aldol Reaction with a-Hydroxyketones

213

KOH

(S)-LLB•KOH (5-10 mol %)

O R

RCHO +

THF, –40 or –50 °C

OH 1

Li O O O La Li O O O Li

OH O R

R OH

19 y. 50-96% anti/syn = 65/35–84/16

18 (2 equiv.)

anti : 84-97% ee

(S)-LLB•KOH

Scheme 6.16

Direct catalytic asymmetric aldol reaction of hydroxyketone 18 promoted by (S)-LLB–KOH.

18 was well diﬀerentiated. The conﬁguration at the b-position of the major anti diastereomer was opposite to that of the aldol product from the methyl ketone (Scheme 6.5) possibly because of bidendate coordination of a-hydroxyketone 18 to the catalyst. Proposed transition state models are depicted in Figure 6.6. Hydroxyketones would coordinate to La in a bidentate fashion, resulting in eﬃcient Si-face shielding of the Z enolate (Figure 6.6, a and b). In addition, this preferential bidentate coordination of 18 is supposed to suppress the self-condensation of aldehydes. Shibasaki reported another approach to the direct asymmetric aldol reac-

Tab. 6.8

Direct catalytic asymmetric aldol reaction of hydroxyketone 18 promoted by (S)-LLB–KOHa. Entry

Aldehyde

1

Ph

CHO

2 3 4 5 6

a Reaction

Product

Catalyst (mol%)

Temp (˚C)

Time (h)

Yield (%)

dr (anti/ syn)

ee (%) (anti/syn)

1t

H-

18a

19ta

10

50

24

84

84/16

95/74

1t 1t 1t 1t

H4-MeO2-Me4-Me-

18a 18c 18d 18e

19ta 19tc 19td 19te

5 10 10 10

50 40 40 40

40 35 35 35

78 50 90 90

78/22 81/19 77/23 83/17

92/70 98/79 84/57 97/85

4-Me-

18e

19ue

10

40

12

96

75/25

96/89

1v

H-

18a

19va

10

50

28

90

72/28

94/83

1j

H-

18a

19ja

10

50

24

86

65/35

90/83

CHO 1u CHO

7

8

Ketone (R-)

CHO

conditions: (S)-LLB ( mol%), KHMDS (0.9 mol%), H2 O (2 mol%), THF.

214

6 Direct Catalytic Asymmetric Aldol Reaction Using Chiral Metal Complexes

R

H

O H

Li H

O

H O

K

O Li H R

O

H O

La

K

La O

O

O

O Li

Li

a

b

OH O

OH O

R

Ph OH anti-19 (αR,βR)-dihydroxy ketone

R

Ph OH syn-19 (αR,βS)-dihydroxy ketone

(favored)

(disfavored)

Fig. 6.6

Transition states postulated for formation of anti diol and syn diol.

tion of 2-hydroxyacetophenones 18. The Et2 Zn/(S,S)-linked-BINOL 20 complex promoted the aldol reaction of 2-hydroxyacetophenones 18 to aﬀord syn-1,2-diols in good yield (Scheme 6.17) [19, 21, 22]. Reactivity and stereoselectivity depended on the substituent on the aromatic ring of the 2-hydroxyacetophenones 18. 2-Hydroxy-2 0 -methoxyacetophenone (18f ) gave the best result to aﬀord the product in 94% yield and high stereoselectivity (syn/anti ¼ 89:11, syn ¼ 92% ee, anti ¼ 89% ee) with as little as 1 mol% catalyst. The catalyst was applicable to a variety of aldehydes including aunsubstituted aldehydes (Scheme 6.17). As summarized in Table 6.9, the reaction reached completion within 24 h with 1 mol% catalyst to give syn-

O R1CHO +

(S,S)-linked-BINOL 20 (x mol %) 2 R Et2Zn (2x mol %)

OH 1

THF, –30 °C

O OH O R

R

OH OH

HO HO

OH

18 (2 equiv.)

19

(S,S)-linked-BINOL 20

R1 = PhCH2CH2, R2 = H (18a), X = 10: 48 h, y. 81%, syn/anti = 67/33, 78% ee(syn) R1 = PhCH2CH2, R2 = 2-MeO (18f), X = 3: 4 h, y. 94%, syn/anti = 90/10, 90% ee(syn) R1 = PhCH2CH2, R2 = 2-MeO (18f), X = 1: 16 h, y. 94%, syn/anti = 87/13, 93% ee(syn)

Scheme 6.17

Direct catalytic asymmetric aldol reactions of hydroxyketone 18 promoted by Et2 Zn/ (S,S)-linked-BINOL 20 complex.

6.4 Direct Aldol Reaction with a-Hydroxyketones Tab. 6.9

Direct catalytic asymmetric aldol reaction of hydroxyketone 18f promoted by Et2 Zn/ (S,S)-linked-BINOL 20 complexa. Entry

1

Aldehyde

CHO

Ph

2

CHO

3

CHO CHO

4

Product

Time (h)

Yield (%)

dr (syn/ anti)

ee (%) (syn/anti)

1e

19ef

20

94

89/11

92/89

1u

19uf

18

88

88/12

95/91

1j

19jf

18

84

84/16

93/87

1w

19wf

12

91

93/7

95/–

1v

19vf

24

94

86/14

87/92

1x

19xf

18

81

86/14

95/90

1y

19yf

16

84

72/28

96/93

1z

19zf

14

93

84/16

90/84

d

19df

24

83

97/3

98/–

1g

19gf

16

92

96/4

99/–

1c

19cf

18

95

97/3

98/–

O CHO

5

CHO

6

BnO

7

BnO

8

BOMO

9

10

11

CHO CHO

CHO

CHO

CHO

a Reaction

conditions: (S,S)-linked-BINOL 20 (1 mol%), Et2 Zn (2 mol%), 30 C, THF.

1,2-diols in excellent yield and stereoselectivity (yield 81–95%, syn/anti ¼ 72:28 to 97:3, syn ¼ 87–99% ee). Mechanistic investigations by kinetic studies, X-ray crystallography, 1 H NMR, and cold-spray ionization mass spectrometry (CSI-MS) analyses shed light on the reaction mechanism and the structure of the active species [22]. X-ray analysis of a crystal obtained from a 2:1 solution of Et2 Zn/(S,S)linked-BINOL 20 in THF revealed the complex consisted of Zn and ligand 20 in a ratio of 3:2 [trinuclear Zn3 (linked-BINOL)2 thf3 ] with C2 symmetry (Figure 6.7, 21). The CSI-MS analysis and kinetic studies revealed that the complex 21 was a precatalyst and that a oligomeric Zn-20-18f complex would work as the actual active species. The proposed catalytic cycle is shown in Figure 6.8. The product dissociation step is rate-determining.

215

216

6 Direct Catalytic Asymmetric Aldol Reaction Using Chiral Metal Complexes

Zn(1) Zn(3) Zn(2)

Zn3(linked-binol)2thf3 21 Fig. 6.7

X-ray structure of preformed complex Zn3 (linked-BINOL)2 thf3 21.

O

OMe

Zn3(linked-binol)2 preformed complex Et2Zn

OR

O

enolate formation oligomeric Zn-rich species –H+ (Ar*OZn-20) (I) Ar*OH = ligand 20

OMe

OR R = Zn or H OH O

MeO

slow

OMe

O O Zn

Zn Ar*OH/Zn-enolate (II)

exchange

+H+ protonation

R

Re face RCHO OH 19 H O

H HR

H

O Zn Zn Ar*OH/Zn-alkoxide (IV) MeO

R

1,2-addition Ar

O

Fig. 6.8

Postulated catalytic cycle for the direct aldol reaction with Et2Zn/(S,S)-linked-BINOL 20 complex.

fast

O Zn O (III)

O Zn

6.4 Direct Aldol Reaction with a-Hydroxyketones

aldehyde

(A)

OH O

favored

Re

R

OH (2R,3S)-syn-19 (major)

O O

Zn

OH O

O Zn Si

S

OMe

R

R

disfavored

OMe

R R

OH (2R,3R)-anti-19 (minor)

aldehyde

(B) H

H H

R

R

Ar O Zn O

O

Zn a (favored)

syn-19 (2R,3S)

H

Ar O Zn O

O

anti-19 (2R,3R)

Zn b (disfavored)

Fig. 6.9

Stereochemical course of direct aldol reaction of hydroxyketone 18f.

The identical absolute conﬁguration (R) was obtained at the a-position of both the syn- and anti-aldol products (Figure 6.9), suggesting that the catalyst diﬀerentiates the enantioface of the enolate well and aldehydes come from the Re face of the zinc enolate (Figure 6.9A). Syn selectivity is explained by the transition state shown in Figure 6.9B. The positive eﬀects of the ortho MeO group suggested the MeO group coordinated with one of the Zn centers in the oligomeric Zn complex aﬀecting the stereoselection step. The electron-donating MeO group has a beneﬁcial eﬀect on further conversion of the products into esters and amides via regioselective rearrangement as shown in Scheme 6.18 [21, 22]. Mechanistic studies suggested that additional Et2 Zn and MS 3A would accelerate the reaction rate. In the presence of MS 3A the second generation Zn catalyst, prepared from Et2 Zn/linked-BINOL 20 in a ratio of 4:1, promoted the direct aldol reaction of hydrocinnamaldehyde (1e) and 1.1 equiv. ketone 18f, smoothly and with reduced catalyst loading (0.25–0.1 mol%, Scheme 6.19) [22]. The practical utility of the reaction was demonstrated by a large-scale reaction performed on the 200-mmol scale by using 0.25 mol% 1 (0.5 mmol, 307 mg) to aﬀord 53.7 g product 19cf (yield 96%) in high dr (syn/anti ¼ 98:2) and ee (94% ee) after 12 h (Scheme 6.19). Considering that the standard catalyst loading for the direct catalytic asymmetric aldol reaction is 2.5 to 20 mol% the exceptionally low catalyst loading in this asymmetric zinc catalysis is remarkable.

217

6 Direct Catalytic Asymmetric Aldol Reaction Using Chiral Metal Complexes

218

O

O OH O

OMe O

R

Ph

S

O

O b

O

O

Baeyer-Villiger Ph OMe oxidation

Ph

OH 19ef

O y. 93%

Beckmann rearrangement y. 97% O O

H N

OMe

O

O

a

OH

OMe c

N OH H y. 94% a (a) mCPBA, NaH PO , CICH CH CI, 50 °C, 2h; (b) O-mesitylenesulfonylhydroxylamime, 2 4 2 2 CH2CI, rt, 4 h; (c) DIBAL, –78 °C to rt, 2h. Ph

Ph

O

OMe

Scheme 6.18

Transformations of aldol adduct via regioselective rearrangementa.

Et2 Zn/linked-BINOL 20, 4:1, with MS 3A enabled the direct aldol reaction of 2-hydroxy-2 0 -methoxypropiopheneone (23), leading to construction of a chiral tetrasubstituted carbon stereocenter (Scheme 6.20) [22]. Although a higher catalyst loading and 5 equiv. ketone 23 were required, a variety of a-unsubstituted aldehydes aﬀorded the product syn-selectively (syn/anti ¼ 59:41 to 71:29) in moderate to good yield and ee (yield 72–97%, syn ¼ 72– 87% ee, anti ¼ 86–97% ee) with (S,S)-linked-BINOL 20 (Table 6.10). Interestingly, the reaction using (S,S)-sulfur-linked-BINOL 22, a linked-BINOL analog including sulfur in the linker, instead of oxygen, resulted in the opO CHO

Ph

OMe

+

OH 1e

O

OMe

+

OH 1c 200 mmol

OH O Ph

OMe

R S

OH

MS 3A, THF, –20 °C

18f (1.1 equiv.)

CHO

Et2Zn (4x mol %) (S,S)-linked-BINOL 20 (x mol %)

19ef x = 0.25 mol %, 18 h yield: 90%, syn/anti = 89/11, syn = 96% ee x = 0.1 mol %, 36 h yield: 84%, syn/anti = 89/11, syn = 92% ee

Et2Zn (1 mol %) (S,S)-linked-BINOL 20 (0.25 mol %) MS 3A, THF, –20 °C, 12 h

18f (1.1 equiv.)

yield: 96%, syn/anti = 98/2, syn = 94% ee

Scheme 6.19

Direct catalytic asymmetric aldol reactions of hydroxyketone 18f promoted by Et2 Zn/ (S,S)-linked-BINOL 20 ¼ 4:1 complex with MS 3A.

OH O R S

OH 19cf 53.7 g

OMe

6.5 Direct Aldol Reaction with Glycine Schiﬀ Bases

O RCHO 1

OMe

+

Et2Zn (20 mol %) ligand 20 or 22 (5 mol %)

OH 23 (5 equiv.)

MS 3A, THF, – 30°C

OMe

R OH 24

X=O: (S,S)-linked-BINOL 20 y. 72-97% syn/anti = 59/41–71/29 syn: 68-87% ee, anti: 86–97% ee

X OH HO OH HO

X = O: 20, X = S: 22

OH O

219

X=S: (S,S)-sulfur-linked-BINOL 22 y. 56-82% syn/anti= 41/59–35/65 syn: 45-60% ee, anti: 81–93% ee

Scheme 6.20

Direct catalytic asymmetric aldol reactions of 2-hydroxy-2 0 -methoxypropiophenone (23) promoted by Et2 Zn/(S,S)-linked-BINOL complex.

posite diastereoselectivity (Scheme 6.20). Reactivity with 22 was somewhat lower than that with 20, and aldol adducts were obtained in moderate to good yield (56–82%) on use of 10 equiv. ketone 23. Major anti isomers were obtained in high ee (81–93% ee), although ee of minor syn isomers was rather low (48–60% ee) (Table 6.10).

6.5

Direct Aldol Reaction with Glycine Schiﬀ Bases

Catalytic asymmetric aldol reactions of glycine equivalents with aldehydes aﬀord eﬃcient and direct access to b-hydroxy-a-amino acid derivatives, which serve as useful chiral building blocks, especially in the pharmaceutical industry. A partially successful catalytic asymmetric aldol reaction of a glycine Schiﬀ base was reported by Miller in 1991 [23]. When Nbenzylcinchoninium chloride 26 was used as chiral phase-transfer catalyst, aldol reaction of Schiﬀ base 25 with aldehydes aﬀorded products synselectively (14–56% de) in good yield (46–92%) although ee was at most 12% ee (Scheme 6.21). After that work no eﬃcient artiﬁcial catalyst was reported for a decade, except for the chemoenzymatic process with glycinedependent aldolases [2]. In 2002 Shibasaki reported use of the heterobimetallic asymmetric catalyst La3 Li3 (binaphthoxide)3 .LiOH (LLB.LiOH) for anti-selective direct aldol reaction of glycine Schiﬀ bases with aldehydes (Scheme 6.22) [24]. Use of 20 mol% of the catalyst, Schiﬀ base 25b, and aldehydes (3 equiv.) gave the products anti-selectively (anti/syn ¼ 59:41 to 86:14) in moderate to good yield (71–93%) and with moderate ee (anti ¼ 19–76% ee).

6 Direct Catalytic Asymmetric Aldol Reaction Using Chiral Metal Complexes

220 Tab. 6.10

Direct catalytic asymmetric aldol reaction with 2-hydroxy-2 0 -methoxypropiophenone (23) promoted by Et2 Zn/ (S,S)-linked-BINOL complexa. Entry

1

Aldehyde

Ketone 23 (Equiv.) CHO

Ph

2 3

Ph

CHO

4

Product

Catalyst (Dmol%)

Temp. (˚C)

Yield (%)

dr (anti/ syn)

ee (%) (anti/syn)

1e

5

24e

20 (5)

30

97

62/38

87/96

1e

10

24e

22 (10)

20

82

35/65

60/92

1t

5

24t

20 (5)

30

72

64/36

78/90

10

24t

22 (10)

20

63

41/59

45/86

20 (5)

30

88

71/29

68/86

22 (10)

20

56

41/59

48/87

20 (5)

30

89

59/41

86/95

1t 0

5

24a

6

1a 0

10

24a 0

7

0

5

24b

0

1b 0

10

24b 0

22 (10)

20

73

41/59

58/93

1z

5

24z

20 (5)

30

92

69/31

87/97

1z

10

24z

22 (10)

20

72

39/61

52/81

1j

5

24j

20 (5)

30

80

68/32

72/87

1y

5

24y

20 (5)

30

80

65/35

85/92

5

1a

CHO

CHO

PMBO

8 9

CHO

BOMO

10 11 12

CHO BnO

a Reaction

CHO

1b

0

conditions: Et2 Zn (4 mol%), ligand ( mol%), MS 3A,

THF.

Maruoka recently developed an eﬃcient direct aldol reaction of glycine Schiﬀ base 25a using phase-transfer catalyst (R,R)-29 [25]. The aldol reaction of Schiﬀ base 25a with aldehydes was eﬃciently promoted by 2 mol% of 29 in toluene–aqueous NaOH (1%) at 0 C to give anti-b-hydroxy-a-amino acids in excellent ee (91–98% ee), moderate to good dr (anti/syn ¼ 1.2:1 to H HO N N N

Ph

CO2-t-Bu

+

Ph 25a

RCHO

1 (5 equiv.)

+ Cl

–

H

OH 26 10 mol %

CH2Cl2/NaOH (2 equiv., 5% w/v) yield: 46-92% de: 14-56% de ee: up to 12% ee

CO2-t-Bu

R N

Ph

27 Ph

Scheme 6.21

Direct catalytic asymmetric aldol reaction of 25a promoted by phase-transfer catalyst 26.

6.6 Other Examples

N

Ar Ar

CO2-t-Bu

i) (S)-LLB (20 mol %) LiOH (18 mol %) H2O (22 mol %), THF, –50 °C + RCHO

25b

1

ii) citric acid, THF-H2O 40 °C 2

221

OH R

CO2-t-Bu NH2

8 R = t-Bu: y. 71% anti/syn = 86/14 anti: 76% ee

Ar = 4-Cl-C6H4 Scheme 6.22

Direct catalytic asymmetric aldol reaction of 25b promoted by heterobimetallic catalyst (S)-LLB.LiOH.

20:1), and in good yield (58–78%) (Scheme 6.23 and Table 6.11). Low catalyst loading, operationally easy reaction conditions (0 C, two-phase reaction), and high ee are noteworthy. Use of (R,R)-29b as catalyst occasionally signiﬁcantly enhanced both diastereo- and enantioselectivity in this system.

6.6

Other Examples

Morken reported a catalytic asymmetric reductive aldol reaction in which aldehydes and acrylate 30 were converted into chiral syn-a-methyl-b-hydroxy esters under the catalysis of transition metal (Rh and Ir) complexes (Scheme 6.24) [26]. The method provided a means of catalytic synthesis of active

Ph

N Ph

CO2-t-Bu

+ RCHO 1

25a

i) 29a or 29b (2 mol %) toluene/aqueous NaOH (1%), 0 °C, 2 h ii) HCl (1 M)/THF

OH R

CO2-t-Bu

28 NH2 y. 40-78% anti/syn = 1.2/1–20/1 anti: 80-96% ee CF3

F3C

Ar Br–

+ N

CF3 Ar = CF3

CF3 Ar 29a Scheme 6.23

Direct catalytic asymmetric aldol reaction of 25a promoted by chiral phase-transfer catalyst 29 promoted by chiral phasetransfer catalyst 29a.

29b

CF3

222

6 Direct Catalytic Asymmetric Aldol Reaction Using Chiral Metal Complexes Tab. 6.11

Direct catalytic asymmetric aldol reaction of 25aa. Entry

Aldehyde

Product

Catalyst

Yield (%)

dr (anti/ syn)

ee (%) (anti)

1

Ph

1e

28e

29a

76

3.3/1

91

1e

28e

29b

71

12/1

96

1t

28t

29b

65

10/1

91

CHO

1c 0

28c 0

29b

72

20/1

98

CHO

1d 0

28d 0

29b

62

6.3/1

80

6

1d 0

28d 0

29a

71

2.4/1

90

7

1e 0

28e 0

29a

58

2.3/1

92

1c

28c

29a

40

2.8/1

95

1c

28c

29a

78

1.2/1

93

CHO

2 3

CHO

4

TIPSO

5

CH3CHO CHO

8 9b a Reaction

conditions: catalyst 29 (2 mol%), toluene/aqueous NaOH (1%), 0 C, 2 h. b Use of dibutyl ether as solvent.

[(cod)RhCl]2 (2.5 mol %) (R)-BINAP (6.5 mol %) Et2MeSiH (1.2 equiv.)

O 1

R CHO + 1

OPh

CHO + 1

OR2 30 (1.2 equiv.)

dichloroethane rt, 24 h

PPh2 PPh2

OPh

31 yield: 48-82% (R)-BINAP syn/anti = 1.8/1-5.1/1 syn = 45-88% ee, anti = 7-99% ee

[(cod)IrCl]2 (2.5 mol %) indane-pybox 32 (7.5 mol %) Et2MeSiH (1.2 equiv.)

O R1

R

dichloroethane rt, 24 h

30 (1.2 equiv.)

OH O

OH O O R

OR'

31 yield: 47-68% syn/anti = 2.7/1-9.9/1 syn = 82-96% ee

Scheme 6.24

Catalytic asymmetric reductive aldol reactions promoted by Rh–(R)-BINAP complex and Ir-32 complex.

O

N N

N 32

6.6 Other Examples Tab. 6.12

Catalytic asymmetric reductive aldol reaction promoted by Ir-32 complexa. Entry

Aldehyde

1b

PhCHO

2

BnO

CHO

3c 4

TBSO

5

BnO

CHO CHO

RO

Yield (%)

dr (syn/ anti)

ee (%) (syn)

1n

Et

68

6.6/1

94

1y

Me

49

9.9/1

96

1y

Me

59

9.5/1

96

1f 0

Me

47

8.2/1

96

1x

Me

65

2.7/1

82

a Reaction

conditions: [(cod)lrCl]2 (2.5 mol%), 32 (7.5 mol%), Et2 MeSiH (2 equiv.), rt, 24 h. b [(coe)lrCl] was used instead of [(cod)lrCl] . 2 2 c Reaction carried out on 35 mmol scale with 1 mol% [(cod)lrCl] and 2 3 mol% of 32.

enolate species from unmodiﬁed substrates. Combination of 2.5 mol% [(cod)RhCl]2 and 6.5 mol% (R)-BINAP promoted the reductive aldol reaction between aldehydes 1 and phenyl acrylate (30) in the presence of Et2 MeSiH at room temperature to aﬀord a diastereomixture (syn/anti ¼ 1.7:1 to 5.1:1) of b-hydroxy esters 31 in good to moderate yield (48–82%) and ee (45–88% ee) [26a]. The stereoselectivity was improved by using the Ir complex derived from [(cod)IrCl]2 and indane-pybox 32, aﬀording the product synselectively (up to 9.9:1) in high ee (up to 96% ee). The results obtained with Ir-32 are summarized in Table 6.12 [26b]. Use of H2 gas as reducing reagent instead of silanes was recently reported by Krische in an achiral reductive aldol reaction [27]. Evans reported a direct catalytic diastereoselective aldol reaction of Nacyloxazolidinones 33. In the presence of a catalytic amount of Mg salt (10 mol%) stoichiometric amounts of Et3 N (2 equiv.) and (CH3 )3 SiCl (1.5 equiv.) the enolate species was generated in situ and anti-aldol adducts were obtained in excellent stereoselectivity (yield 36–92%, dr 3.5:1 to 32:1; Scheme 6.25) [28a]. Stoichiometric amounts of silylating regents were essential to achieve eﬃcient catalyst turnover. Because mechanistic study revealed that the reaction does not involve an enol silyl ether as an intermediate, a Mukaiyama-type reaction pathway was reasonably excluded. By using N-acylthiazolidinethiones 35 instead of 33 the aldol adducts were obtained with diﬀerent stereoselectivity [28b]. MgBr2 aOEt2 (10 mol%) afforded the best results for 35 (Scheme 6.25). anti-Aldol adducts were obtained in excellent selectivity (yield 56–93%, dr 7:1 to 19:1). Catalytic asymmetric variants of these reactions on the basis of the Mg catalysis seem promising.

223

6 Direct Catalytic Asymmetric Aldol Reaction Using Chiral Metal Complexes

224

O O

O N R1 Bn 33

S S

+ R2CHO (1.1 equiv.) 1

O N R1 Bn 35

+ R2CHO (1.1 equiv.) 1

1) MgCl2 (10 mol %) (CH3)3SiCl (1.5 equiv.) Et3N (2 equiv.)

O O

OH

O

R2

N R1

EtOAc, 23 °C, 24 h 2) H+

34 Bn y. 36-94%, dr 3.5/1-32/1 dr ratio = (desired isomer)/Σ other isomers

1) MgBr2•OEt2 (10 mol %) (CH3)3SiCl (1.5 equiv.) Et3N (2 equiv.)

O O

EtOAc, 23 °C, 24 h 2) H+

OH

O

R2

N R1

36 Bn y. 56-93%, dr 7/1-19/1 dr ratio = (desired isomer)/Σ other isomers

Scheme 6.25

Direct catalytic diastereoselective aldol reactions promoted by Mg salts.

Shair recently reported an achiral direct aldol reaction starting from malonic acid half thioester 37 (Scheme 6.26). The active enolate species were generated under extremely mild conditions with Cu catalysis. Development of a direct catalytic asymmetric aldol reaction on the basis of this strategy is also promising [29].

6.7

Conclusion

Representative examples of direct catalytic asymmetric aldol reactions promoted by metal catalysis have been summarized. The ﬁeld has grown rapidly during past ﬁve years and many researchers have started to investigate this ‘‘classical’’ yet new ﬁeld using metal catalysts and organocatalysts. Application of these systems to direct Mannich reactions and Michael reactions has also been studied intensively recently. Further investigations will be needed to overcome problems remaining with regard to substrate generality, reaction time, catalyst loading, volumetric productivity, etc. The development of a direct catalytic asymmetric aldol reaction with unmodiﬁed esters as a donor is particularly required. O

O

BnS OH 37 (1 equiv.)

+ RCHO 1 (1 equiv.)

Cu(2-ethylhexanoate)2 (20 mol %) N

OMe

N (22 mol %) H wet THF, air, 23 °C, 2-24 h Scheme 6.26

Catalytic thioester aldol reactions prompted by Cu(II) salt.

O BnS

OH R

38 y. 22-97%

+ CO2

6.8 Experimental Section

6.8

Experimental Section Procedure for the Preparation of (S)-LLB Complex. A solution of La(O-i-Pr)3 (20.4 mL, 4.07 mmol, 0.2 m in THF, freshly prepared from La(O-i-Pr)3 powder and dry THF) was added to a stirred solution of (S)-binaphthol (3.50 g, 12.2 mmol) in THF (39.7 mL) at 0 C. (La(O-i-Pr)3 was purchased from Kojundo Chemical Laboratory, 5-1-28 Chiyoda, Sakado-shi, Saitama 350-0214, Japan; Fax: þ81-492-84-1351). The solution was stirred for 30 min at room temperature and the solvent was then evaporated under reduced pressure. The resulting residue was dried for 1 h under reduced pressure (ca. 5 mmHg) and dissolved in THF (60.5 mL). The solution was cooled to 0 C and n-BuLi (7.45 mL, 12.2 mmol, 1.64 m in hexane) was added. The mixture was stirred for 12 h at room temperature to give a 0.06 m solution of (R)-LLB which was used to prepare (S)-LLB–KOH catalyst. General Procedure for Direct Catalytic Asymmetric Aldol Reactions of Methyl Ketone 2 Using (S)-LLB–KOH. A solution of water in THF (48.0 mL, 0.048 mmol, 1.0 m) was added to a stirred solution of potassium bis(trimethylsilyl)amide (KHMDS, 43.2 mL, 0.0216 mmol, 0.5 m) in toluene at 0 C. The solution was stirred for 20 min at 0 C and then (S)-LLB (400 mL, 0.024 mmol, 0.06 m in THF, prepared as described above) was added and the mixture was stirred at 0 C for 30 min. The resulting pale yellow solution was cooled to 20 C and acetophenone (2a) (175 mL, 1.5 mmol) was added. The solution was stirred for 20 min at this temperature then 2,2-dimethyl-3-phenylpropanal (1b) (49.9 mL, 0.3 mmol) was added and the reaction mixture was stirred for 28 h at 20 C. The mixture was then quenched by addition of 1 m HCl (1 mL) and the aqueous layer was extracted with ether (2 10 mL). The combined organic layers were washed with brine and dried over Na2 SO4 . The solvent was removed under reduced pressure and the residue was puriﬁed by ﬂash chromatography (SiO2 , ether–hexane 1:12) to give 3ba (72 mg, 85%, 89% ee). General Procedure for Direct Catalytic Asymmetric Aldol Reaction of Methyl Ketone 2 Using Dinuclear Zn2 -13. The prepare the catalyst a solution of diethyl zinc (1 m in hexane, 0.2 mL, 0.2 mmol) was added to a solution of ligand 13 (64 mg, 0.1 mmol) in THF (1 mL) at room temperature under an argon atmosphere. After stirring for 30 min at the same temperature, with evolution of ethane gas, the resulting solution (ca. 0.09 m) was used as catalyst for the aldol reaction. To perform the aldol reaction a solution of the catalyst (0.025 mmol) was added, at 0 C, to a suspension of aldehyde (0.5 mmol), triphenylphosphine sulﬁde (22.1 mg, 0.075 mmol), powdered 4 A˚ molecular sieves (100 mg, dried at 150 C under vacuum overnight), and ketone 2 (2.5 or 5 mmol) in THF (0.8 mL). The mixture was stirred at 5 C for 2 days then poured on to

225

226

6 Direct Catalytic Asymmetric Aldol Reaction Using Chiral Metal Complexes

1 m HCl and extracted with ether. After normal work-up, the crude product was puriﬁed by silica gel column chromatography. General Procedure for Catalytic Asymmetric Aldol Reaction of Hydroxyketone 18f Promoted by Et2 Zn/(S,S)-linked-BINOL, 4:1, with MS 3A. MS 3A (200 mg) in a test tube was activated before use under reduced pressure (ca. 0.7 kPa) at 160 C for 3 h. After cooling, a solution of (S,S)-linked-BINOL (1.53 mg, 0.0025 mmol) in THF (0.6 mL) was added under Ar. The mixture was cooled to 20 C and Et2 Zn (10 mL, 0.01 mmol, 1.0 m in hexanes) was added to the mixture at this temperature. After stirring for 10 min at 20 C, a solution of 18f (182.8 mg, 1.1 mmol) in THF (1.1 mL) was added. Aldehyde 1e (1.0 mmol) was added and the mixture was stirred at 20 C for 18 h and then quenched by addition of 1 m HCl (2 mL). The mixture was extracted with ethyl acetate and the combined organic extracts were washed with sat. aqueous NaHCO3 and brine and dried over MgSO4 . Evaporation of the solvent gave a crude mixture of the aldol products. The diastereomeric ratios of the aldol products were determined by 1 H NMR of the crude product. After puriﬁcation by silica gel ﬂash column chromatography (hexane– acetone 8:1 to 4:1), 19ef was obtained (269.6 mg, 0.898 mmol, yield 90%, dr syn/anti ¼ 89:11, 96% ee (syn)). General Procedure for Catalytic Asymmetric Aldol Reaction of Glycine Schiﬀ Base 25a Promoted by Phase-transfer Catalyst 29. Aqueous NaOH (1%, 2.4 mL) was added at 0 C, under Ar, to a solution of Schiﬀ base 25a (88.6 mg, 0.3 mmol) and (R,R)-29b (9.9 mg, 2 mol%) in toluene (3 mL). Aldehyde 1e (79 mL, 0.6 mmol) was then introduced dropwise. The whole mixture was stirred for 2 h at 0 C, and water and diethyl ether were then added. The ether phase was isolated, washed with brine, dried over Na2 SO4 , and concentrated. The crude product was dissolved in THF (8 mL) and treated with HCl (1 m, 1 mL) at 0 C for 1 h. After removal of THF in vacuo the aqueous solution was washed three times with diethyl ether and neutralized with NaHCO3 . The mixture was then extracted three times with CH2 Cl2 . The combined extracts were dried over MgSO4 and concentrated. After puriﬁcation by silica gel column chromatography (CH2 Cl2 aMeOH 15:1) 28e was obtained (56.8 mg, 0.214 mmol, yield 71%, dr anti/syn ¼ 12:1, 96% ee (anti)).

References and Notes 1 Recent review: C. Palomo, M. Oiarbide, J. M. Garcı´a, Chem.

Eur. J. 2002, 8, 37. 2 Review: T. D. Machajewski, C.-H. Wong, Angew. Chem. Int.

Ed. 2000, 39, 1352. 3 Recent review for the direct catalytic asymmetric aldol

References and Notes

4 5 6 7 8 9

10 11 12 13

14 15 16

17 18 19

20 21 22

23 24 25 26

27 28

29

reactions: B. Alcaide, P. Almendros, Eur. J. Org. Chem. 2002, 1595. Review: M. Sawamura, Y. Ito, Chem. Rev. 1992, 92, 857. Review: B. List, Tetrahedron 2002, 58, 5573. Y. M. A. Yamada, N. Yoshikawa, H. Sasai, M. Shibasaki, Angew. Chem. Int. Ed. Engl. 1997, 36, 1871. K. Fujii, K. Maki, M. Kanai, M. Shibasaki, Org. Lett. 2003, 5, 733. N. Yoshikawa, Y. M. A. Yamada, J. Das, H. Sasai, M. Shibasaki, J. Am. Chem. Soc. 1999, 121, 4168. (a) D. Sawada, M. Kanai, M. Shibasaki, J. Am. Chem. Soc. 2000, 122, 10521; (b) D. Sawada, M. Shibasaki, Angew. Chem. Int. Ed. 2000, 39, 209. Y. M. A. Yamada, M. Shibasaki, Tetrahedron Lett. 1998, 39, 5561. B. M. Trost, H. Ito, J. Am. Chem. Soc. 2000, 122, 12003. B. M. Trost, E. R. Silcoff, H. Ito, Org. Lett. 2001, 3, 2497. T. Suzuki, N. Yamagiwa, Y. Matsuo, S. Sakamoto, K. Yamaguchi, M. Shibasaki, R. Noyori, Tetrahedron Lett. 2001, 42, 4669. N. Yoshikawa, M. Shibasaki, Tetrahedron 2001, 57, 2569. R. Mahrwald, B. Ziemer, Tetrahedron Lett. 2002, 43, 4459. (a) W. Notz, B. List, J. Am. Chem. Soc. 2000, 122, 7368; (b) K. Sakthivel, W. Notz, T. Bui, C. F. Barbas, III, J. Am. Chem. Soc. 2001, 123, 5260. B. M. Trost, H. Ito, E. R. Silcoff, J. Am. Chem. Soc. 2001, 123, 3367. B. M. Trost, V. S. C. Yeh, Org. Lett. 2002, 4, 3513. N. Yoshikawa, N. Kumagai, S. Matsunaga, G. Moll, T. Ohshima, T. Suzuki, M. Shibasaki, J. Am. Chem. Soc. 2001, 123, 2466. N. Yoshikawa, T. Suzuki, M. Shibasaki, J. Org. Chem. 2002, 67, 2556. N. Kumagai, S. Matsunaga, N. Yoshikawa, T. Ohshima, M. Shibasaki, Org. Lett. 2001, 3, 1539. N. Kumagai, S. Matsunaga, T. Kinoshita, S. Harada, S. Okada, S. Sakamoto, K. Yamaguchi, M. Shibasaki, J. Am. Chem. Soc. 2003, 125, 2169. C. M. Gasparski, M. J. Miller, Tetrahedron 1991, 47, 5367. N. Yoshikawa, M. Shibasaki, Tetrahedron 2002, 58, 8289. T. Ooi, M. Taniguchi, M. Kameda, K. Maruoka, Angew. Chem. Int. Ed. 2002, 41, 4542. (a) S. J. Taylor, M. O. Duffey, J. P. Morken, J. Am. Chem. Soc. 2000, 122, 4528. (b) C.-X. Zhao, M. O. Duffey, S. J. Taylor, J. P. Morken, Org. Lett. 2001, 3, 1829. Achiral reaction: (c) S. J. Taylor, J. P. Morken, J. Am. Chem. Soc. 1999, 121, 12202. H.-Y. Jang, R. R. Huddleston, M. J. Krische, J. Am. Chem. Soc. 2002, 124, 15156. (a) D. A. Evans, J. S. Tedrow, J. T. Shaw, C. W. Downey, J. Am. Chem. Soc. 2002, 124, 392. (b) D. A. Evans, C. W. Downey, J. T. Shaw, J. S. Tedrow, Org. Lett. 2002, 4, 1127. G. Lalic, A. D. Aloise, M. D. Shair, J. Am. Chem. Soc. 2003, 125, 2852.

227

229

7

Catalytic Enantioselective Aldol Additions with Chiral Lewis Bases Scott E. Denmark and Shinji Fujimori 7.1

Introduction 7.1.1

Enantioselective Aldol Additions

The aldol addition reaction is one of the most powerful carbon–carbon bond-construction methods in organic synthesis and has achieved the exalted status of a ‘‘strategy-level reaction’’. The generality, versatility, selectivity, and predictability associated with this process have inspired many reviews and authoritative summaries and constitute the theme of this treatise [1]. The primary objective in the evolution of the aldol addition is the striving for exquisite diastereo- and enantioselectivity from readily available enolate precursors. The ideal aldol reaction would provide selective access for all four isomers of the stereochemical dyad that make up the aldol products. This has given way to more ambitious investigation of the triads and tetrads that accrue from double and triple diastereoselection processes [2]. The solutions to these challenges have been imaginative and diverse, and have pioneered the contemporaneous development of asymmetric synthesis as a core discipline. A secondary and more recent objective is the development of ‘‘direct aldol additions’’ that mimic enzymatic processes (aldolases) and obviate the independent activation of the nucleophilic partner. The number and variety of inspired and elegant solutions for perfecting the aldol addition are expertly described in the accompanying chapters of this volume. This chapter diﬀers somewhat, however, in that it describes a conceptually distinct process that has been designed to address some of the shortcomings inherent in the more classic approaches involving chiral Lewis acid catalysis of aldol addition in its many incarnations. Thus, to assist the reader in understanding the distinctions and to provide the conceptual framework for invention of Lewis-base-catalyzed addition, the introduction will outline brieﬂy the stereocontrolling features of the main families Modern Aldol Reactions. Vol. 2: Metal Catalysis. Edited by Rainer Mahrwald Copyright 8 2004 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 3-527-30714-1

230

7 Catalytic Enantioselective Aldol Additions with Chiral Lewis Bases

Ph

O Me

O HO

Ph Ph

OH

1. LDA / THF / −78 °C 2. MgBr2 3. Aryl

CO2Me

Aryl 94% ee

CHO

4. NaOMe Me

Me

Me O N

S O2

Me O

1. LICA / THF −78 °C Me 2. PhCHO

N S O2

O

OH Ph

OH

+ Χ∗N

Ph Me

Me 85 / 8 / 7 (anti)

Scheme 7.1

Aldol additions with chirally modiﬁed lithium enolates.

of enantioselective aldol additions and thus, the basis for inventing a new process. Background Early examples of asymmetric aldol addition reactions involved lithium enolates of chiral carbonyl compounds that reacted with aldehydes to give good diastereoselectivity [3]. The chirality of the enolate translated to enantiomerically enriched products when the auxiliaries were destroyed or removed. Thus, using enolates of modiﬁed ketones [3a], esters [3b], and sulfonamides [3c], high enantioselectivity and diastereoselectivity can be achieved if enolates are generated in geometrically deﬁned form (Scheme 7.1). Although high selectivity is obtained, these reactions are not practical because they require stoichiometric amounts of covalently bound auxiliaries. In addition, the high reactivity of the lithium enolates did not ensure reaction via closed, organized transition structures, a feature crucial for stereochemical information transfer. A revolutionary advance in aldol technology was the use of less reactive metalloenolates (boron [4a], titanium [4b–d], and zirconium [4d]) that organize the aldehyde, enolate, and auxiliary in a closed transition structure (Scheme 7.2). Although these reagents are similar to those described above in that an auxiliary is needed in stoichiometric amounts, the use of boron and titanium enolates enable attachment of the modiﬁer by an acyl linkage or directly around the metal of the enolate. Geometrically deﬁned enolates react with aldehydes to give the syn or anti diastereomers with high enantiomeric excess. This variant is best exempliﬁed by the acyl oxazolidinone boron enolates [1a], the diazaborolidine derived enolates [5], titanium enolates derived from diacetone glucose [6], the diisiopinylcampheyl boron enolates for ketone aldolizations [2c], and proline-derived silanes for N,Oketene acetals [7]. 7.1.1.1

7.1 Introduction

O

O

Ln*M

MLn

R1CHO

O

or

H 3C

X R*

X

R

O

OH R1 or

RX

OH

RX

R1

CH3

CH3

231

syn

CH3

anti

Ph

n-Bu O H 3C

n-Bu B N

O O

Bn

SO2Aryl N Ph B diacetone glucoseO Ti N O O diacetone glucoseO ArylO2S RO XR CH3

CH3

CH3 B R

N

O CH3

CH3

O O Si

CH3

Scheme 7.2

Chirally modiﬁed boron, titanium, and silicon enolates.

The key stereocontrolling features common to these agents are:

. the organizational role of the metal center; . the close proximity of the electrophile, nucleophile and asymmetric mod.

iﬁer in coordination sphere of the metal assuring high stereochemical information transfer, and the high stereochemical inﬂuence of enolate geometry on product diastereoselectivity.

The major disadvantage of these variants is the inability to operate catalytically. Indeed, it is the high metal aﬃnity of the aldehyde, enolate, and chiral auxiliary that interferes with the turnover. Catalytic processes have, over the past decade, dominated the development of enantioselective aldol addition reactions [1j,k]. This category can be subdivided into ﬁve main classes:

. chiral-Lewis-acid-catalyzed aldol additions of silicon or tin enol ethers aldol addition); . (Mukaiyama in-situ generated metalloenolates from silicon or tin enol ethers; . in-situ generated metalloenolates directly from ketones; . in-situ generated enolate equivalents (enamines) directly from carbonyl and . compounds; enzyme- and antibody-catalyzed aldol additions.

7 Catalytic Enantioselective Aldol Additions with Chiral Lewis Bases

232

Me3SiO CH3 +

RX

R1

O

MXn*

O

R1

RX H

NH

O

Me Aryl N OH

O

MeO

Fe

N

H PPh2

Me

NMe2 Me O N

OMe PPh2

[Sn(OTf)2 / n-Bu3SnF]

R1 CH3

CO2H

HO2C

OH

+ RX

CH3

OH N Me

O

OH

[BH3]

[Ti(Oi-Pr)4]

[Au(I)]

Me O N t--Bu

t--Bu

[Cu(II)]

Scheme 7.3

Representative metal-based chiral Lewis acids.

The ﬁrst three only will be discussed in the context of the origins of stereoinduction. The use of chiral Lewis acids [8] has received by far the most attention and is amply discussed in the many chapters dedicated to various metals in this volume. Some of the more commonly used and selective chiral Lewis acids are shown here, for example diamine complexes of tin(II) triﬂate [9], borane complexes of a monoester of tartaric acid (CAB catalysts) [10], sulfonamido amino acid borane complexes [11], titanium binaphthol [12] and binaphthylimine complexes [13], ferrocenylphosphine–gold [5d] and BINAP– silver [14] complexes, and copper(II) bisoxazoline and pyridyl(bisoxazoline) complexes [15], (Scheme 7.3). These variants of the aldol reaction have several key features in common:

. the additions have been demonstrated for aldehydes and enol metal dewith sub-stoichiometric loading of the chiral Lewis acid; . rivatives the diastereo- and enantioselectivity are variable although they can be and . high; these reactions are not responsive to prostereogenic features – when the conﬁguration of the enolsilane nucleophile changes, the diastereoselectivity of the product does not change [16]. In these reactions, the metal center is believed to activate the aldehyde to addition and the enol addition subject primarily to steric approach control, i.e. it is lacking the pre-organization associated with the stoichiometric aldol addition reactions of the boron and titanium enolates. This problem has been addressed in part by the recently developed class of aldol additions that involve the use of chirally modiﬁed metalloids in a catalytic process [17]. In these reactions it is proposed a metal–phosphine

7.1 Introduction

complex undergoes transmetalation with TMS enol ethers or tributylstannyl ketones to provide chiral metalloid enolates in situ. The aldol addition then proceeds, with turnover of the metalloid species to another latent enol donor. In addition, in the third class the metalloenolate is generated in situ from either heterobimetallic (lanthanide/alkali metals) or chiral zinc phenoxide complexes and promotes the addition of unmodiﬁed ketones to aldehydes [18]. In these reactions it is postulated that the aldehyde is coordinated to the metal after generation of the metalloenolate. However, because the enolates are generated in situ, the enolate geometry is not known and geometry has not been correlated with product conﬁguration. Despite the power and clear synthetic applicability of these families, deﬁciencies are still apparent:

. lack of a catalytic variant of the boron or titanium enolate family; and . lack of controllable selectivity in the chiral Lewis acid family. Lewis base-activation provides a mechanism enabling devising of a class of aldol addition that addresses these concerns. This chapter describes, in detail, the formulation, development, and understanding of a Lewis-basecatalyzed aldol reaction process that embodies both the selectivity and versatility of the stoichiometric reactions in combination with the eﬃciency of the catalytic methods. 7.1.2

Lewis Base Catalysis

The design criteria for Lewis basic catalysis of the aldol addition are outlined in Figure 7.1. This approach diﬀers from Lewis acid catalysis of aldol addition in that it postulates activation of the enoxymetal derivative by preassociation with a chiral Lewis basic (LB) group bearing a non-bonding pair of electrons. This complex must be more reactive than the free enolate for ligand accelerated catalysis to be observed [19]. Next, association of this -ate complex with the Lewis basic carbonyl oxygen of the aldehyde produces a hyper-reactive complex in which the metal has expanded its valence by two. It is expected that this association complex between enolate, aldehyde, and the chiral Lewis basic group reacts through a closed-type transition structure to produce the metal aldolate product. For turnover to be achieved the aldolate must undergo the expulsion of the LB group with formation of the chelated metal aldolate product. Thus, Lewis base-catalysis involves simultaneous activation of the nucleophile and the electrophile within the coordination sphere of the metal. The reaction must occur in a closed array and be capable of releasing the activating group by chelation or change in the Lewis acidity. To realize this process selection of the appropriate enoxymetal and activator moieties is crucial. For the metal, the MXn subunit must be able expand

233

234

7 Catalytic Enantioselective Aldol Additions with Chiral Lewis Bases

O

Xn M *

O

O

MXn

LB (Lewis basic promoter)

* R

turnover event rapid

LB

LB MXn O O * * R

O

MXn

more reactive than free enolate

LB reaction through closed TS

O

MXn O H

O R

H

R

Fig. 7.1

Hypothetical catalytic cycle for Lewis-base-catalyzed aldol addition.

its valence by two and balance the nucleophilicity of the enolate with electrophilicity to coordinate both the Lewis basic aldehyde and the chiral LB group. To impart suﬃcient Lewis acidity to that metal group and accommodate the valence expansion such that two Lewis basic atoms may associate, the ligands (X) should be small and strongly electron-withdrawing. The criteria necessary for the chiral Lewis basic group LB are that it must be able to activate the addition without cleaving the OaMXn linkage and provide an eﬀective asymmetric environment. Candidates for the Lewis basic group include species with high donicity properties as reﬂected in solvent basicity scales [20]. The inspiration to propose the possibility of nucleophilic catalysis of aldol additions and guide selection of the appropriate reaction partners is found in the cognate allylation process by allyl- and 2-butenyltrichlorosilanes. Inspired by the pioneering observations of Sakurai [21] and Kobayashi [22] that allyltrihalosilanes can be induced to add to aldehydes in the presence of nucleophilic activators (ﬂuoride ion or DMF solvent) it was ﬁrst shown in 1994 that chiral Lewis bases (phosphoramides) are capable of catalyzing the addition of allyltrichlorosilanes [23]. Thus, by analogy, reducing this plan to practice required the invention of a new class of aldol reagent, trichlorosilyl enolates, in conjunction with one of the most Lewis basic neutral functional groups, the phosphoramide group, Figure 7.2. Trichlorosilyl enolates of esters had been reported in the literature [24] and (because of the electronwithdrawing chloride ligands on silicon) were expected to be highly electro-

7.1 Introduction

LB MXn O O

H

O

SiCl3

R1

R1 N O P R2 N N 2 R1 R

Fig. 7.2

Reaction components required for chiral Lewis-base-catalyzed aldol addition.

philic and thus able to stabilize the hypercoordinate silicon species necessary in such a process. The phosphoramides can be seen as chiral analogs of HMPA the Lewis basicity of which is well documented [20], especially toward silicon-based Lewis acids [25]. 7.1.3

Organization of this Chapter

On the basis of the design criteria outlined above, the ﬁrst, chiral-Lewisbase-catalyzed, enantioselective aldol addition was reported in 1996 [26]. This disclosure, which reported the reaction of the trichlorosilyl enolate of methyl acetate with a variety of aldehydes in the presence of several chiral phosphoramides, was signiﬁcant not so much for the results obtained (enantioselectivity was modest 20 to 62% ee) but rather as a proof of principle for this conceptually new approach to the aldol addition reaction (Scheme 7.4). This early success launched a broad-ranging program on the scope, synthetic application, and mechanistic understanding of chiral Lewis base catalysis of the aldol addition. A chronological recounting of the evolution of this program has already appeared [27]. For this chapter, a more comprehensive treatment of the various components of the process is presented, and thus, a more structurally based organization is employed. The main section begins with the preparation of the two new reaction components, namely the enoxytrichlorosilanes of ester, ketones, and aldehydes and the chiral Lewis basic catalysts (phosphoramides and N-oxides). Me N O P N N Ph Me 10 mol % Ph

OSiCl3 OMe

O + R

H

CH2Cl2, −78 °C 30 min - 3 h

HO R

O OMe

R = Ph, 87%, er 2.0/1 R = t-Bu, 78%, er 2.3/1

Scheme 7.4

The ﬁrst enantioselective, chiral-Lewis-base-catalyzed aldol addition.

235

236

7 Catalytic Enantioselective Aldol Additions with Chiral Lewis Bases

The bulk of the chapter is dedicated to describing the diversity of enolate structural subtypes in order of increasing structural complexity. Beginning with trichlorosilyl enolates of simple achiral methyl, ethyl, and cyclic ketones the survey then addresses chirally modiﬁed enolates of ketones and the phenomenon of double (1,n)-diastereoinduction. The next sections outline the use of trichlorosilyl enolates derived from aldehydes and esters and the features unique to these structures. The ﬁnal preparative section is dedicated to the newest variation on the theme, namely, the use of chiral Lewis bases to activate simple, achiral Lewis acids for enantioselective aldolization. To facilitate more fundamental understanding of the development of the reaction variants, the chapter ends with an overview of the current mechanistic picture. Although this aspect is still evolving, the basic features are well in hand and enable integrated understanding of the behavior of trichlorosilyl enolates under these conditions. Representative procedures for all the asymmetric processes described herein are provided at the end of the chapter.

7.2

Preparation of Enoxytrichlorosilanes

Silyl enol ethers (enoxysilanes) derived from carbonyl compounds are among the most important reagents in synthetic organic chemistry, because of their ability to form carbon–carbon bonds when combined with a myriad of carbon electrophiles [28]. The ﬁrst silyl enol ethers, reported in 1958, were obtained by hydrosilylation of unsaturated carbonyl compounds (Scheme 7.5) [29]. Since then silyl enol ethers have become particularly versatile synthetic intermediates, and a number of reviews on preparation and reactions of these compounds have appeared [30]. The synthetic utility of enoxysilanes was not fully recognized until pioneering work by Mukaiyama on Lewis acid-catalyzed aldol additions of trimethylsilyl enol ethers to diﬀerent carbonyl compounds (Scheme 7.6) [31]. The physical properties of simple enoxytrialkylsilanes were thoroughly investigated by Baukov and Lutsenko [30d]. Unlike conventional metal enolates, enoxytrialkylsilanes are stable and isolable covalent species. These

O + H

Et3SiH

cat. H2PtCl6 i -PrOH, reflux

O SiEt3

Me H

1 (62%) Scheme 7.5

First reported synthesis of a silyl enol ether.

7.2 Preparation of Enoxytrichlorosilanes

OTMS +

PhCHO

1. TiCl4 (1.1 equiv) CH2Cl2, −78oC 2. H2O

O

O

OH Ph

syn-2 (69%)

+

OH Ph

anti-2 (23%)

Scheme 7.6

Application of a silyl enol ether in a directed aldol addition.

species can be stored under non-acidic conditions for a long period of time but can also can be readily hydrolyzed to the parent carbonyl compounds under acidic conditions. Trialkylsilyl enol ethers were originally introduced as precursors for regioisomerically-deﬁned metal enolates. As enol derivatives they have reasonable nucleophilicity, although the most common use of these reagents involved regeneration of the metal enolate under basic conditions followed by reaction with electrophiles [30]. The nucleophilicities of a variety of enoxytrialkylsilanes have recently been correlated with other nucleophiles by Mayr [32]. The established order indicates that the nucleophilicity of enoxytrialkylsilanes is greater than that of allylic trialkylsilanes and less than that of commonly used enamines. Mayr also showed that silyl ketene acetals are much more nucleophilic than silyl enol ethers. Trialkylsilyl enol ethers are extensively utilized in chiral Lewis acidcatalyzed stereoselective aldol additions [33]. On the other hand, silyl enol ethers with other groups on the silicon have been less widely applied in synthesis [34]. Heteroatom-functionalized silyl enol ethers can be prepared by methods similar to those used to prepare their trialkylsilyl counterparts (Scheme 7.7). Walkup and coworkers reported a convenient procedure for syntheses of a variety of non-alkyl-substituted enoxysilanes such as 4 and 5 [34a]. Hydrosilylation of a,b-unsaturated carbonyl compounds is also a viable method for preparation of such enoxysilanes [35]. Although a variety of silicon-functionalized silyl enol ethers have appeared in the literature, their application in synthetically useful reactions is still limited. A recent exception disclosed by Yamamoto and coworkers is a Lewis-acid-catalyzed enantioselective aldol reaction of an enoxy(trimethoxy)silane (Scheme 7.8) [36]. Because development of an eﬀective Lewis-base-catalyzed aldol addition required access to electrophilic enolates, the preparation, properties and reactivity of enoxytrichlorosilanes became important areas of investigation. Pioneering studies by Baukov et al. ensured the possibility of generating trichlorosilyl ketene acetals, but it was subsequent studies by Denmark et al. that elevated these and related species to the status of useful synthetic reagents [30d, 37].

237

238

7 Catalytic Enantioselective Aldol Additions with Chiral Lewis Bases

O

Me Me Si O Cl

Li Me2SiCl2

t-Bu

Et2O

t-Bu

Me Me Si O OEt

EtOH, Et3N t-Bu

Et2O 4

3

O +

HSi(OEt)3

5 (79%) OSi(OEt)3

[Rh(OH)(cod)]2 (0.15 mol%) THF

6 (>99%) Scheme 7.7

Preparation of silicon-functionalized silyl enol ethers.

7.2.1

General Considerations

Trichlorosilyl enolates (enoxytrichlorosilanes) are typically viscous oils and can be obtained in the pure form by simple distillation. These silyl enolates can be stored under anhydrous conditions at low temperature for an appreciable time without decomposition. Exclusion of moisture is essential when working with trichlorosilyl enolates. Trace amounts of water leads to hydrolysis of the chlorosilane unit, and the resulting HCl is deleterious to the trichlorosilyl enolate. Degradation of the trichlorosilyl enolates can also be initiated by trace impurities such as metal salts and ammonium salts which promote formation of di- and polyenoxysilane species [37]. The thermal stability of trichlorosilyl enolates depends on their structure. For ketone- and aldehyde-derived trichlorosilyl enolates the O-silyl and Csilyl isomerism strongly favors the O-silyl species [37]. Although ketoneand aldehyde-derived trichlorosilyl enolates can be heated to 140 C, trichlorosilyl enolates can disproportionate into dienoxysilanes and silicon tetrachloride at higher temperatures [37]. On the other hand, trichlorosilyl ketene acetals are not as thermally stable as the other trichlorosilyl enolates O

OSi(OMe)3 8 (10 mol %) +

PhCHO

OH Ph 8=

MeOH, −78 oC 7

2 (78%) syn/anti , 5.3/1 er(syn), 14/1 Scheme 7.8

Asymmetric aldol addition of a trialkoxysilyl enol ether.

Ar Ar P Ag F P Ar Ar Ar = p-tolyl

7.2 Preparation of Enoxytrichlorosilanes

IR (C=C): 1659 cm-1 H NMR (ppm) E-HC(2): 4.42 (s) Z-HC(2): 4.55 (d) 13 C NMR (ppm) C(1): 152.62 C(2): 96.94

1

Cl3Si Me

O HZ

1

2

HE

9

Cl3Si Et

1

O Me

1

2

H (Z )-11

H NMR (ppm) HC(2): 5.01 (qt) 13 C NMR (ppm) C(1): 150.54 C(2): 105.18

Cl3Si

O

MeO

HZ

1

2

HE 10

Cl3Si Et

1

O 1

H 2

Me (E )-11

1

Cl3Si

O

H

1

Me 2

H (Z )-12

H NMR (ppm) HC(1): 6.25 (qd) HC(2): 4.95 (dq) 13 C NMR (ppm) C(1): 136 C(2): 112

IR (C=C): 1672 cm-1 NMR (ppm) E-HC(2): 3.41 (s) Z-HC(2): 3.65 (s) 13 C NMR (ppm) C(1): 152.62 C(2): 96.94

1H

H NMR (ppm) HC(2): 5.13 (qt) 13 C NMR (ppm) C(1): 150.54 C(2): 105.90 1

Cl3Si H

O 1

H 2

Me (E )-12

H NMR (ppm) HC(1): 6.27 (qd) HC(2): 5.40 (dq) 13 C NMR (ppm) C(1): 135 C(2): 111

Fig. 7.3

Spectroscopic properties of enoxytrichlorosilanes.

and tend to isomerize to the corresponding carbon-bound, a-trichlorosilyl esters upon heating [37]. Distillation of these reagents should therefore be performed under vacuum at a temperature as low as possible. The spectroscopic properties of several enoxytrichlorosilanes are summarized in Figure 7.3 [37, 38]. The 1 H NMR chemical shifts of the vinylic protons in enoxytrichlorosilanes are usually higher than those of the corresponding trimethylsilyl enol ethers. The vinylic protons for (E)- and (Z)-11 derived from ethyl ketones are suﬃciently diﬀerent that the E/Z ratios for these enolates are readily obtained by 1 H NMR analysis. The vinylic proton for the E enolate is typically found at lower ﬁeld than for the corresponding Z enolate. In the aldehyde-derived enoxytrichlorosilane the former aldehydic proton appears above 6 ppm. The IR stretching frequencies for the enol double bonds appear between 1630 and 1660 cm1 for ketone-derived enoxytrichlorosilanes and at 1677 cm1 for the acetate-derived trichlorosilyl ketene acetal. One of the major diﬀerences between trialkylsilyl enol ethers and trichlorosilyl enolates is their reactivity toward aldehydes. Trialkylsilyl enol ethers usually do not react with aldehydes in the absence of nucleophilic or electrophilic activators [39]. On the other hand, trichlorosilyl enolates undergo aldol additions spontaneously with aldehydes at or below ambient

239

240

7 Catalytic Enantioselective Aldol Additions with Chiral Lewis Bases

(1) Direct enolization: O R

SiCl4 Lewis base Et3N

O

SiCl3

R

(2) Metal exchange: OLi (a)

O SiCl4

R

R

OSnBu3 (b)

OTMS (c)

R

O SiCl4

R

SiCl4 MXn

SiCl3

SiCl3

R

O

SiCl3

R

Fig. 7.4

General methods for preparation of enoxytrichlorosilanes.

temperature to aﬀord aldol adducts in good yields. More importantly, trichlorosilyl enolates are susceptible to ligand-accelerated catalysis in the presence of Lewis bases [26]. The development, scope, utility, and mechanism of this process will be covered in subsequent sections. In this section, the preparation and properties of trichlorosilyl enolates, classiﬁed by enolate structure, are described. Preparations of enoxytrichlorosilanes can be generalized to several categories: direct enolization of parent carbonyl compounds, trapping of corresponding metal (lithium) enolates, and metathesis of tin(IV) or trialkylsilyl enol ether with silicon tetrachloride (Figure 7.4). The optimum method for a given diﬀerent class depends on the structure of the enolate. Synthetically viable methods only are discussed herein; other approaches are found in earlier review articles [30d]. 7.2.2

Preparation of Ketone-derived Trichlorosilyl Enolates

One of the earliest reports on the synthesis of a trichlorosilyl enolate described the reduction of an a-chloroketone using trichlorosilane and a tertiary amine. Benkeser employed a combination of trichlorosilane and tri-nbutylamine for reduction of polyhalogenated organic compounds [40]. For example, a-chloroketone 13 is smoothly converted to trichlorosilyl enolate 14 in good yield by use of this procedure (Scheme 7.9). Under similar con-

7.2 Preparation of Enoxytrichlorosilanes

O Cl Cl

THF Me + HSiCl3 + n-Bu3N

reflux,1 h

Cl

13

Cl

SiCl3 Me

Cl 14 (80%)

O

THF + HSiCl3 + n-Bu3N

Cl

O

Me

10 - 25 oC, 1 h

15

O

SiCl3 Me

9 (72%)

Scheme 7.9

Reductive silylation of a-chloroketones.

ditions the monochloro ketone 15 provides the corresponding trichlorosilyl enolate 9 in good yield. Surprisingly, the reaction cannot be eﬀected by triethylamine or diisopropylethylamine in place of tri(n-butyl)amine. The use of pentane as solvent was found to be superior to use of tetrahydrofuran because it enabled easier removal of solvent from these volatile trichlorosilyl enolates. These enolates are puriﬁed ﬁrst by vacuum-transfer of the reaction mixture to separate them from ammonium salt and then by redistillation to remove solvent. Despite the operational simplicity and high yields obtained by use of this procedure, the scope of the reaction is somewhat limited by the availability of the corresponding a-chloro ketone and the volatility of the resulting enolate, which is necessary for vacuum-transfer. For the acetone-derived enolate 9, however, this is the method of choice. Another useful method of preparation of trichlorosilyl enolates involves metathesis of the corresponding enol stannane with silicon tetrachloride (Scheme 7.10) [37, 41]. Enol stannanes can be prepared by treatment of enol acetates with tributylmethoxystannane at elevated temperature [42]. Reaction of enol stannanes with silicon tetrachloride at low temperature provides trichlorosilyl enolates in modest to good yield. Excess silicon tetrachloride is recommended to prevent formation of polyenoxysilanes. These two steps can be performed without puriﬁcation of the intermediate enol stannane, and this makes the method more practical. Trichlorosilyl enolates are eﬃciently generated from methyl and cyclic ketones by this method. For the propiophenone-derived enolate 18, excellent geometric selectivity for the Z isomer is observed. Although this method is general for the preparation of ketone-derived trichlorosilyl enolates, the use of a stoichiometric amount of the tin reagent and the limited availability of structurally homogeneous enol acetates make this procedure less practical. Distillation of the trichlorosilyl enolate in the presence of tin residues and excess chlorosilane during puriﬁcation is sometimes diﬃcult.

241

242

7 Catalytic Enantioselective Aldol Additions with Chiral Lewis Bases

OSiCl3

OSnBu3

OAc Bu3SnOMe

SiCl4

o

100 C

0 oC OSiCl3

OSiCl3 Me

OSiCl3 Me

Me Me 17 (67%)

18 (83%) Z/E >50/1

OSiCl3

OSiCl3

OSiCl3

19 (27%)

20 (78%)

16 (54%)

21 (63%)

Scheme 7.10

Preparation of enoxytrichlorosilanes from enoxystannanes.

To avoid the use of tin reagents several other methods have been developed for preparation of trichlorosilyl enolates. Among these the most general method for ketone-derived trichlorosilyl enolates is the metal-catalyzed trans-silylation of trimethylsilyl enol ethers. It is known that mercury(II) and tin(IV) salts react with trimethylsilyl enol ethers to generate a-mercurioand a-stannyl-ketones [43, 44]. Also, as shown previously, tin enolates can be readily converted into trichlorosilyl enolates by the action of silicon tetrachloride. From these observations, a metal-catalyzed process for conversion from trimethylsilyl enol ethers to trichlorosilyl enolates could be devised (Figure 7.5) [37]. A survey of diﬀerent metal salts revealed that soft Lewis acids such as Hg(OAc)2 and Pd(OAc)2 are eﬀective catalysts of this transformation [37]. Optimization studies indicated that the stoichiometry of the reagents and the reaction concentration are critical to the rate of trans-silylation and to control the amount of bisenoxysilane species formed. A bis(enoxy)dichlorosilane is a common impurity associated with many aspects of enoxytrichlorosilane chemistry. The formation of a bis(enoxy)dichlorosilane can be explained by disproportionation of a monoenoxytrichlorosilane. At

O

OTMS MXn - TMSX Fig. 7.5

Metal-catalyzed transsilylation.

OSiCl3 MXn-1

- MXn SiCl4

7.2 Preparation of Enoxytrichlorosilanes

243

the end of the reaction the crude reaction mixture usually contains 10–15% bis(enoxy)dichlorosilane species. The amount of bis(enoxy)silane depends on the metal catalyst used, and mercury(II) acetate is the most selective for production of enoxytrichlorosilanes. Although a slight excess of silicon tetrachloride can reduce the amount of bis(enoxy)dichlorosilane formed, use of a large excess (more than 3 equiv.) leads a signiﬁcant rate deceleration owing to catalyst deactivation. Optimum conditions are use of 2–3 equiv. of silicon tetrachloride and a concentration below 1.0 m in dichloromethane [37]. The loading of the metal catalyst can be as low as 0.25 mol%, but usually 1–5 mol% of the metal salt can be employed. Several trichlorosilyl enolates have been prepared by this method (Scheme 7.11). This transformation is general for a variety of enolate structures and it is synthetically appealing, because the precursor trimethylsilyl enol ether can be readily prepared in regiochemically pure form. This transformation is extremely facile, especially for preparation of methyl ketone-derived trichlorosilyl enolates, and enables complete conversion in less than 2 h with 1 mol% Hg(OAc)2 . The cyclic ketone-derived enol ethers require longer reaction times ranging from 18 to 24 h [37].

OTMS

OSiCl3

Hg(OAc)2 (1-5 mol %) SiCl4 (2 equiv) CH2Cl2, rt

OSiCl3

OSiCl3

OSiCl3

OSiHCl2

OSiCl3

OSiCl3

Me

n-Bu Me

19 (73%)

20 (68%)

OSiCl3

22 (59%) w/ HSiCl3 OSiCl3

Me

i-Bu

23 (83%)

21 (78%)

OSiCl3

24 (83%) OSiCl3

OSiCl3

TBSO

Me Me

25 (74%) OSiPhCl2 i-Bu 28 (45%) w/ PhSiCl3

26 (61%)

16 (81%)

27 (69%) OSiCl3

OSiCl3 Me

Me OTBS 29 (71%)

17 (71%) OSiCl3

Me OPiv 30 (78%)

Scheme 7.11

Preparation of enoxytrichlorosilanes by Hg(II)-catalyzed metathesis.

OBn 31 (60%)

244

7 Catalytic Enantioselective Aldol Additions with Chiral Lewis Bases Tab. 7.1

Metal-catalyzed transsilylation of 3-pentanone-derived trimethylsilyl ether 32. Me

OTMS SiCl4 (2 equiv) MX2 (5 mol %) 32

Me

Me

CH2Cl2, rt

OSiCl3 Me 11

Entry

MX2

32, E/ Z

Time, h

Yield, %a

11, E/Z b

1 2 3 4 5 6

Hg(OAc)2 Hg(OAc)2 Pd(OAc)2 Pd(OAc)2 Pd(TFA)2 Pd(TFA)2

3/1 1/4 3/1 1/4 3/1 1/4

5 5 5 5 15 15

72 60 76 69 70 43

1/2 1/2 1/6 1/6 1/7 1/6

a Yield

of distilled material. b Determined by 1 H NMR analysis.

Common functional and protecting groups can be tolerated under the reaction conditions and the resulting trichlorosilyl enolates are suﬃciently pure for use in the phosphoramide-catalyzed aldol addition (vide infra). The use of other chlorosilanes enables the preparation of diﬀerent classes of chlorosilyl enolate. For example, when trichlorosilane is used in place of silicon tetrachloride, dichlorohydridosilyl enolate 22 can be obtained. A major drawback to the metal-catalyzed trans-silylation is lack of control over the geometry of the resulting trichlorosilyl enolate. Starting from either E- or Z-enriched trimethylsilyl enol ether 32, the E/Z ratio of the enoxytrichlorosilane 11 is always 1:2 when Hg(OAc)2 is used as the catalyst (Table 7.1) [45]. Use of Pd(II) salts results in slightly higher Z selectivity, but again the E/Z ratio of the enoxytrichlorosilane does not mirror the E/Z ratio of the trimethylsilyl enol ether. The E/Z ratio also depends on the structure of enolates (Table 7.2). ZTrichlorosilyl enolates are always selectively formed. The general trend of Z/E selectivity is related to the size of the R group. For larger R groups, higher Z selectivity is observed. These trends are also observed in the transsilylation catalyzed by Pd(OAc)2 and Pd(TFA)2 . These observations can be rationalized by the following mechanism (Figure 7.6). The overall process consists of electrophilic attack of the metal salt to aﬀord a-metalloketone 34. Coordination of silicon tetrachloride to the carbonyl group of 34 and loss of metal salt gives the enoxytrichlorosilane. The initial formation of 34 is presumably reversible, and this event can account for the randomization of enolate geometry. The E/Z ratio of the resulting trichlorosilyl enolate is determined by the relative rate of breakdown of the two limiting conformers i and ii. The avoidance of steric interaction between Me and R in i makes this conformer more favorable, leading to the preferred formation of the Z enoxytrichlorosilane. This explanation is consistent with the trend observed in the relationship between steric

7.2 Preparation of Enoxytrichlorosilanes

245

Tab. 7.2

Hg(II)-catalyzed transsilylation of a variety of trimethylsilyl enol ethers. OTMS

SiCl4 (2 equiv) Hg(OAc)2 (5 mol %)

R

R

CH2Cl2, rt

Me

OSiCl3

OSiCl3 Me

+

R

Me (E )-33

(Z )-33

Entry

R

Time, h

Yield, %a

33, E/Z b

1 2 3 4 5 6c

H Me Et i-Pr t-Bu Ph

16 16.5 5 18 24 18

50 58 72 65 55 66

1/8 1/2 1/2 1/8 1/>20 1/99

of distilled material. b Determined by 1 H NMR analysis. mol% of Hg(OAc)2 was used.

a Yield c 10

demand of R and the E/Z ratio. In fact, the presence of bulky R groups enables highly selective preparation of Z enolates under these conditions. E-Conﬁgured trichlorosilyl enolates cannot, however, be obtained selectively by this method. A method has been developed that avoids the use of a metal catalyst to prepare geometrically deﬁned trichlorosilyl enolates. It involves generation of a lithium enolate, by treatment of an isomerically enriched trimethylsilyl enol ether with methyllithium, and subsequent capture of the conﬁgurationally deﬁned lithium enolate with silicon tetrachloride [46]. Both E and Z trichlorosilyl enolates can be prepared by means of this method, without loss of geometrical purity (Scheme 7.12). Addition of methyllithium to a trimethylsilyl enol ether leads to smooth conversion to the lithium enolate [46]. The E/Z ratio of the trichlorosilyl enolates mirrors the E/Z ratio of the starting trimethylsilyl enol ether.

MLn-1 SiCl4 R H

R

OTMS Me

MLn

O

MLn-1 34

(Z )-33

O Me SiCl3 i: favored

Me

-TMSL R

Cl

-MLn-1Cl

Cl SiCl4

Me R

H SiCl3 O MLn-1

ii: less favored Fig. 7.6

Proposed mechanism for metal-catalyzed transsilylation.

(E )-33 -MLn-1Cl

246

7 Catalytic Enantioselective Aldol Additions with Chiral Lewis Bases

1. MeLi 2. SiCl4

OTMS RZ

R

Et2O

RE OSiCl3 Me

TBSO

Me

OSiCl3 RZ

OSiCl3 Me

R RE

TIPSO Me

OSiCl3 Me

(Z )-35 (81%) Z/E, 50/1

OSiCl3

Me Me

Me

(E )-11 (34%) E/Z, 99/1

TIPSO

(Z )-36 (53%) Z/E, 32/1

(E )-36 (23%) E/Z, 15/1

Scheme 7.12

Preparation of geometrically deﬁned enoxytrichlorosilanes.

Numerous methods are used to prepare geometrically deﬁned trimethylsilyl enol ethers. For example, Z trimethylsilyl enol ethers can be prepared by using dibutylboron triﬂate and subsequent treatment of the boron enolate with trimethylsilyl chloride [47]. The E isomers are typically prepared by use of lithium tetramethylpiperidide as described by Collum [48]. The geometrically deﬁned trimethylsilyl enol ethers are converted into the corresponding trichlorosilyl enolates by the above-mentioned procedure. Unfortunately, the resulting enolate is often contaminated with the bis(enoxy)dichlorosilane thus reducing the overall yield of the process. Nonetheless, the ability to prepare geometrically deﬁned enoxytrichlorosilanes makes the metal-exchange method synthetically attractive. 7.2.3

Preparation of Aldehyde-derived Trichlorosilyl Enolates

In this section, three methods used to generate aldehyde-derived enoxytrichlorosilanes are described [49]. The ﬁrst is the metathetical route from the corresponding trimethylsilyl enol ether using a catalytic amount of Pd(OAc)2 and excess silicon tetrachloride (Scheme 7.13). In this method, the geometry of the resulting trichlorosilyl enolate is not dependent on the trimethylsilyl enol ether for the same reason as discussed above (Figure 7.6). Thus, only unsubstituted or symmetrically substituted enolates are suitable.

SiCl4 (2 equiv) Pd(OAc)2 (1 mol %)

n-C5H11 OTMS

n-C5H11

CH2Cl2

OSiCl3 37 (79%) Z/E, 3.5/1

Scheme 7.13

Preparation of an aldehyde-derived enoxytrichlorosilane.

7.2 Preparation of Enoxytrichlorosilanes

H

n-C5H11

Ph N O P N N

+

O

CDCl3

OSiCl3 38 (72%)

SiCl4 (i-Pr)2NEt

+ Me

O

Cl

n-C5H11

Ph

H

n-C5H11

SiCl4

247

N O

Me

n-C5H11

CH2Cl2

OSiCl3 37 (79%) Z/E, 3.5/1

Scheme 7.14

Direct silylation of aldehydes using SiCl4 and a Lewis base.

The second procedure is direct silylation from an aldehyde with phosphoramides or N-oxides and a base (Scheme 7.14). In the presence of silicon tetrachloride and a catalytic amount of a Lewis base aldehydes are rapidly transformed into a-chloro trichlorosilyl ethers. The formation of such intermediates has recently been documented and observed by means of 1 H NMR spectroscopic analysis [50]. Addition of an amine base promotes elimination of HCl to yield the trichlorosilyl enolate. Before use the resulting trichlorosilyl enolate must be distilled from the ammonium salt generated by the reaction. Although this procedure provides the trichlorosilyl enolate directly from a given aldehyde, the enol geometry cannot be controlled, thus limiting the utility of this process. Generation of stereodeﬁned trichlorosilyl enolates of aldehydes can also be accomplished by the direct O-to-O trans-silylation via lithium enolates (Scheme 7.15). The geometrically-deﬁned trimethylsilyl enol ethers of heptanal react with methyllithium to yield the conﬁgurationally stable lithium enolates. After trapping with a large excess of silicon tetrachloride the geometrically enriched trichlorosilyl enolates of aldehydes are prepared in good yield.

OTMS RZ

1. MeLi 2. SiCl4 Et2O

RE n-C5H11

OSiCl3 RZ

n-C5H11

OSiCl3

RE Me OSiCl3

OSiCl3 (Z )-37 (53%) Z/E, 99/1

(E )-37 (71%) E/Z, 30/1

(Z )-12 (29%) Z/E, 98/2

Scheme 7.15

Preparation of geometrically deﬁned enoxytrichlorosilanes.

Me

OSiCl3

(E )-12 (34%) E/Z, 99/1

248

7 Catalytic Enantioselective Aldol Additions with Chiral Lewis Bases

O SnBu3

MeO

0 oC

39 OSiCl2H MeO 40 (48%)

MeO

OSiCl3 MeO

OSiCl2Me MeO

10 (65%)

OSiCl2Ph MeO

OSiClRR'

RR'SiCl2

41 (57%)

OSiMe2Cl O

MeO

Si

Cl

MeO 42 (23%)

43 (18%)

44 (19%)

Scheme 7.16

Preparation of acetate-derived trichlorosilyl ketene acetals.

7.2.4

Preparation of Trichlorosilyl Ketene Acetals

The ﬁrst reported enoxytrichlorosilanes were derived from esters [38]. Those ketene acetals are prepared by reaction of a chlorosilane and an a-stannyl ester (Scheme 7.16). This method is still the most general preparation of acetate-derived trichlorosilyl ketene acetals. In the presence of excess silicon tetrachloride the stannyl ester 39 is smoothly converted to the ketene acetal 10. The ketene acetal can be distilled at ambient temperature under reduced pressure. These species cannot be heated because isomerization to a Ctrichlorosilyl ester occurs at higher temperatures [51]. This isomerization is also a problem when these ketene acetals are stored for a long time, because even at room temperature isomerization occurs in a month. Unlike the trichlorosilyl enolates derived from ketones and aldehydes, which exist exclusively as the O-silyl isomers, trichlorosilyl ketene acetals can isomerize to the thermodynamically more stable C-silyl isomer [38]. The C-silyl esters are not reactive in phosphoramide-catalyzed aldol reactions and these species do not revert to the corresponding trichlorosilyl ketene acetal under common reaction conditions. The use of diﬀerent chlorosilanes enables preparation of structurally diverse chlorosilyl ketene acetals (Scheme 7.16). Although this procedure is relatively simple, the method suﬀers from low yields because of the diﬃculty of separating the trichlorosilyl ketene acetal from tributylchlorostannane and from the C-trichlorosilylacetate. The purity of the chlorosilyl ketene acetal is critical because tin residues from the reaction promote oligomerization of the ketene acetal. In the reaction with tributylstannylpropanoates under similar conditions the major products obtained are, unfortunately, the C-trichlorosilyl prop-

7.3 Preparation of Chiral Lewis Bases

anoate derivatives. Thus, so far only acetate-derived trichlorosilyl ketene acetals have been prepared by this method. In summary, practical and eﬃcient methods are now available for preparation of enoxytrichlorosilanes. The most general method is the transition metal-catalyzed trans-silylation of trimethylsilyl enol ethers with silicon tetrachloride. Geometrically deﬁned enoxytrichlorosilanes are best prepared by silylation of lithium enolates. The conﬁguration of the lithium enolate precursor is preserved in this process. The metathesis of methyl tributylstannylacetate with silicon tetrachloride is the most eﬃcient route for preparation of trichlorosilyl ketene acetals.

7.3

Preparation of Chiral Lewis Bases

The structure of the Lewis base greatly aﬀects its catalytic activity and selectivity in the aldol addition. Moreover, diﬀerent types of trichlorosilyl nucleophile require diﬀerent types of chiral Lewis base catalyst. To examine a wide range of structures, general methods are needed for synthesis of phosphoramides and N-oxides [52]. The four most commonly used Lewis-base catalysts are shown in Chart 7.1. The phosphoramide 45 is the most general and selective catalyst for aldol addition of ketone-derived trichlorosilyl enol ethers to aldehydes [53]. In MeMe Me N O P N N

N

t-Bu

Me 45: addition of ketone-derived trichlorosilyl enol ethers

H H

O P N N Me

t-Bu

46: addition of trichlorosilyl ketene acetals to ketones

Me N O P N N Me Me

N

CH2

2

47: addition of allylic trichlorosilanes

N O O On-Bu n-BuO

CH2

2

48: addition of aldehyde-derived trichlorosilyl enol ether and addition of TMS enol ethers

Chart 7.1

Commonly used chiral Lewis bases for aldol additions.

249

250

7 Catalytic Enantioselective Aldol Additions with Chiral Lewis Bases

recent studies dimeric phosphoramide catalysts such as 47 and 48 have been shown to be highly selective in additions of allyltrichlorosilane and allyltributylstannane (with silicon tetrachloride) to aldehydes [54]. These catalysts are also eﬀective in the addition of aldehyde-derived trichlorosilyl enolates and in additions of trialkylsilyl ketene acetals and enol ethers to aldehydes [49, 50]. For addition of trichlorosilyl ketene acetal to ketones the bis-N-oxide 46 has proven to be the most selective catalyst [55]. Syntheses of these Lewis base catalysts are brieﬂy described in the following sections. 7.3.1

Preparation of Chiral Phosphoramides

The basic strategy for synthesis of chiral cyclic phosphoramides is to couple a chiral 1,2-, 1,3-, or 1,4-diamine to either a phosphorus(V) or phosphorus(III) reagent. There are three general routes (Scheme 7.17). Method A is the most straightforward strategy for preparation of chiral cyclic phosphoramides. A chiral diamine is combined with an aminophosphoric dichloride in the presence of triethylamine [56]. The reaction is typically conducted in a halogenated solvent under reﬂux. This method works well for preparation of sterically less bulky phosphoramides and for coupling aliphatic diamines. For sterically demanding coupling partners, elevated temperatures and longer reaction times are required. For example, phosphoramide 45 is obtained in good yield from (R,R)-N,N 0 -dimethyl-1,2-diphenylethylenediamine [57] by method A (Scheme 7.18). For less reactive diamines a more electrophilic phosphorus(III) reagent is needed to enhance the reaction rate (Methods B and C, Scheme 7.17) [58]. In these methods the diamine is ﬁrst lithiated by use of n-BuLi at low temperature. The lithiated diamine is combined with the mono-

NHR

A:

+

[G]* NHR

O Cl P NR12 Cl

+

[G]* NHR

CH2Cl2, reflux 1. n-BuLi 2. [O]

NHR

B:

Et3N

Cl P NR12 Cl 1. n-BuLi

C:

2. R12NH

NHR [G]*

+ NHR

Cl P Cl Cl

3. [O]

Scheme 7.17

General preparations of chiral phosphoramides.

R O N P NR12 N [G]* R R O N P NR12 N [G]* R R O N P NR12 N [G]* R

7.3 Preparation of Chiral Lewis Bases

Me NH + NH Me

O Cl P N Cl

Et3N CH2Cl2, reflux

251

Me N O P N N Me (R,R)-45 (50%)

Scheme 7.18

Preparation of monophosphoramide 45.

aminophosphorus(III) dichloride reagent (Method B, Scheme 7.17) or with phosphorus(III) chloride followed by treatment with an amine (Method C, Scheme 7.17). The resulting phosphorus(III) triamine species is oxidized with m-CPBA to give the desired phosphoramide. For example, the bisphosphoramide 48 is prepared in good yield from N,N 0 -dimethyl-1,1 0 binaphthyl-2,2 0 -diamine by a three-step sequence [59] (Scheme 7.19). In this example, the N,N 0 -dimethylpentanediamine (49) is used as the linker [60].

Me NH NH Me

Me N 49, Et3N P Cl N Me 49 = MeHN

1. n-BuLi 2. PCl3 THF

Me N P N N Me Me

m-CPBA CH2

NHMe

Me N O P N N Me Me

CH2

2

2

(R,R)-48 (72%, 3 steps) Scheme 7.19

Preparation of bisphosphoramide 48.

The bisphosphoramide 47 can be prepared from 2,2 0 -bispyrrolidine (Scheme 7.20) [61]. In the presence of triethylamine, enantiomerically pure bispyrrolidine reacts with phosphorus oxychloride to provide diaminophosphoryl chloride 50. The lithiated linker 49 is combined with (R,R)-50 to aﬀord the bisphosphoramide 47 in excellent yield.

252

7 Catalytic Enantioselective Aldol Additions with Chiral Lewis Bases

NH H H

+ POCl3 NH

Et3N

O P Cl N N

H H

Et2O

49, n-BuLi THF

O P N N Me N

H H

CH2 2

(R,R)-50 (67%)

R-(l,l )-47 (93%)

Scheme 7.20

Preparation of bisphosphoramide R-(l,l)-47.

7.3.2

Synthesis of Chiral bis-N-Oxides

N-Oxides are readily obtained by oxidation of tertiary amines [62]. Accordingly, chiral N-oxides are usually prepared by oxidation of chiral tertiary amines. Several axially chiral bis-N-oxides have been synthesized; these are known to promote addition reactions of chlorosilane species [62]. The chiral bis-N-oxide 46 contains both central and axial elements of chirality (Scheme 7.21). These two features are essential for the stereoselectivity observed in promoted aldol reactions and are also helpful in enantio- and diastereoselective synthesis of the catalysts. Introduction of the stereogenic center is achieved by reduction of the tert-butyl ketone by (Ipc)2 BCl [63]. The N-oxide obtained after etheriﬁcation and oxidation undergoes diastereoselective oxidative dimerization to aﬀord (P)-46 [55].

Me Br

1. n-BuLi 2. pivaloyl chloride (2 equiv) t-Bu

Et2O

N

Me

1. (-)-(Ipc)2BCl 2. (HOC2H4)2NH t-Bu

THF/ Et2O

N

Me

90% KOH, 18-c-6 n-BuBr DMF

N OH

O

84% (R/S = 97.8/2.2)

Me

Me m-CPBA

t-Bu

N On-Bu

CH2Cl2

n-BuO

84%

N O 84%

MeMe LiTMP; I2 THF, −73 oC - rt t-Bu n-BuO

t-Bu

N O

N O n-BuO

P-(R,R )-46 : 48% Scheme 7.21

Preparation of chiral bis-N-oxide P-(R,R)-46.

t-Bu

7.4 Enantioselective Aldol Addition of Achiral Enoxytrichlorosilanes

7.4

Enantioselective Aldol Addition of Achiral Enoxytrichlorosilanes

Trichlorosilyl enolates (enoxytrichlorosilanes) derived from ketones undergo additions to aldehydes spontaneously at or below ambient temperature without external activation [53]. The intrinsic reactivity of these reagents contrasts with that of trialkylsilyl enol ethers in the aldol addition, for which a promoter is usually required [64]. The reactivity of trichlorosilyl enolates is not because of the inherent nucleophilicity of the enolate but rather the high electrophilicity of the silicon atom [27]. The silicon atom of a trichlorosilyl enolate is highly electropositive, because of the eﬀect of chlorine ligands. Lewis basic functions, including aldehydes, can bind to the Lewis acidic silicon and form a hypercoordinate complex. On binding of an aldehyde to the Lewis-acidic silicon atom the aldehyde is electrophilically activated (Figure 7.7, iii). The enolate moiety is concurrently activated by increased polarization of the enolate SiaO bond. This dual activation results in the high reactivity of trichlorosilyl enolates. A similar rationale is also proposed for the aldol reaction of boron enolates [2c] and strained-ring alkyl silyl enolates [65]. Aldol additions of trichlorosilyl enolates are catalyzed by Lewis bases, most notably phosphoramides (Figure 7.8). It is believed that binding of a phosphoramide to a trichlorosilyl enolate leads to ionization of a chloride, forming a cationic silicon–phosphoramide complex [66]. The binding of an aldehyde to the silicon complex leads to aldolization through a closed transition structure. There are two catalyzed pathways – one involves the intermediacy of a pentacoordinate, cationic silicon complex in which only one phosphoramide is bound to silicon and the other involves a hexacoordinate, cationic silicon complex in which two phosphoramide molecules are bound to the silicon [66]. In the former pathway aldolization occurs through a boatlike transition structure, whereas in the latter pathway, the transition structure is chair-like (Figure 7.8, iv and v). This mechanistic duality in the catalyzed process is analyzed in more detail in Section 7.9.

nucleophilic activation OSiCl3

Cl O

+ RCHO

electrophilic activation

Cl Si

Cl

O

H

Cl3 Si O O

R iii: boat transition structure Fig. 7.7

Hypothetical assembly for uncatalyzed aldol addition of a trichlorosilyl enolate.

R

253

254

7 Catalytic Enantioselective Aldol Additions with Chiral Lewis Bases

Cl Cl O

OP(NR2)3 Si

Cl

O

Cl3 Si O

O HH one-phosphoramide pathway OSiCl3

(R2N)3PO

R'

R' iv: cationic, trigonal bipyramid boat

+ R'CHO Cl Cl two-phosphoramide pathway

HR'

O Si OP(NR2)3 OP(NR2)3 O H Cl

O

Cl3 Si O R'

v: cationic, octahedron chair Fig. 7.8

Divergent pathways for catalyzed aldol addition of a trichlorosilyl enolate.

The appeal of aldol additions of trichlorosilyl enolates is the selective and predictable diastereocontrol that probably arises from a closed transition structure. For substituted enolates the diastereomeric ratio of aldol products can be directly correlated with the enolate geometry as predicted by the Zimmerman–Traxler model [67]. Thus, the dominant reaction pathway in the catalyzed reactions of trichlorosilyl enolates involves a chair-like transition structure organized around the silicon. By employing chiral phosphoramides, enantioselection can be controlled. Thus highly stereocontrolled aldol addition can be envisaged in Lewis-base-catalyzed aldol addition of trichlorosilyl enolates. In this section aldol additions of achiral trichlorosilyl enolates derived from ketones are described. The inherent reactivity of these species and their potential use in asymmetric, catalytic processes will be discussed. 7.4.1

Aldol Additions of Achiral Methyl Ketone-derived Enolates

Trichlorosilyl enolates derived from methyl ketones are reactive toward aldehydes in the absence of Lewis base catalysts at ambient temperature (Scheme 7.22) [68]. Trichlorosilyl enolates bearing a broad range of non-participating substituents react with benzaldehyde to give excellent yields of the aldol

7.4 Enantioselective Aldol Addition of Achiral Enoxytrichlorosilanes

OSiCl3

O

OH

Me

OH

n-Bu

Ph

O

Ph

O

Ph

OH Ph

OH

TBSO

OH

i-Pr

Ph 54 (93%)

O

OH

Ph

Ph

56 (93%)

55 (91%)

O

53 (94%)

O

OH

Ph

i-Bu

52 (95%)

51 (92%)

OH

R

rt, 4 - 6 h

R O

O

CH2Cl2

PhCHO

+

57 (97%)

Scheme 7.22

Uncatalyzed aldol addition of methyl ketone-derived trichlorosilyl enolates.

products in several hours. The steric and electronic properties of the enolates do not have a large inﬂuence on the rate of aldol addition. Under similar conditions the trichlorosilyl enolate 24 undergoes aldol addition to a wide range of aldehydes at room temperature with excellent yields (Scheme 7.23). Aromatic and conjugated aldehydes are typically more reactive than aliphatic aldehydes, presumably because of their smaller size and higher Lewis basicity [69]. The structure of the aldehyde signiﬁcantly aﬀects the rate of aldol addition, however. Reactions with bulky aldehydes

OSiCl3

+

n-Bu

O

CH2Cl2

RCHO

rt, 4 - 14 h

OH

n-Bu

R

24 O

OH

n-Bu

Ph

O n-Bu

52 (95%) O

Ph

O n-Bu

Ph Me 59 (92%)

OH

O

Me MeMe

60 (93%)

OH

n-Bu

58 (91%)

OH

n-Bu

O

OH

61 (trace) (86%, w/ 10 mol % HMPA)

Scheme 7.23

Uncatalyzed aldol additions of 62 to a variety of aldehydes.

OH

n-Bu

Ph 62 (84%)

255

7 Catalytic Enantioselective Aldol Additions with Chiral Lewis Bases

256

OSiCl3

O

OHC

CH2Cl2

+

OH

n-Bu

+ 63 (95%)

−78oC, 2 h

n-Bu 24

64 (4%)

63 Scheme 7.24

Uncatalyzed aldol addition of 24 at 78 C.

are slower, as is evidenced by the reaction of pivalaldehyde. Addition of 10 mol% HMPA leads to a dramatic increase in the rate of addition, enabling isolation of 61 in good yield. The rate of reaction of trichlorosilyl enolates with aldehydes is greatly attenuated at low temperature. For example, in the addition of 24 to 4biphenylcarboxaldehyde, only 4% of 64 was isolated and 95% of unreacted aldehyde was recovered after 2 h (Scheme 7.24). This behavior is crucial for optimization of the asymmetric process, because suppression of the achiral background reaction is important for achieving high enantioselectivity in the catalyzed reaction. In the presence of several structurally diverse chiral phosphoramides the aldol addition proceeds smoothly at low temperature (Scheme 7.25). Although all of the phosphoramides are eﬀective in promoting the aldol addition of 24 to benzaldehyde, the stilbene-1,2-diamine-derived phosphoramide 45 is the most active and selective catalyst. The structure of the chiral diamine backbone clearly has a large eﬀect on the enantioselectivity of the process. Also, substitution of a diisopropylamino group for the piperdinyl group in 63 leads to dramatic drop in enantioselectivity. Correct choice of solvent is a critical aspect of obtaining high enantioselectivity in the aldol addition of trichlorosilyl enolates [68]. Dichloromethane is the most suitable solvent for the reaction, in terms of both reactivity and selectivity (Table 7.3). Other halogenated solvents such as trichloroethylene or the more polar propionitrile are good solvents for this reaction and proOSiCl3 n-Bu

PhCHO CH2Cl2, −78oC 2h

24

Ph Ph

Me N O P N N Me

45 (92%); er 12.5/1

O

cat. (10 mol %) +

Ph Ph

Me N O P i-Pr N N Me i-Pr

63 (79%); er 1.70/1

n-Bu

Me N O P N N Me 64 (76%); er 1/3.31

Scheme 7.25

Catalyzed aldol additions of 24 to benzaldehyde.

OH Ph

52 Me N O P N N Me 65 (71%); er 4.29/1

7.4 Enantioselective Aldol Addition of Achiral Enoxytrichlorosilanes Tab. 7.3

Eﬀect of solvent on aldol addition of 24 to benzaldehyde. OSiCl3

O

(S,S)-45 (10 mol %) +

PhCHO

n-Bu

solvent, −78 oC

OH

n-Bu

24

Ph

52

Entry

Solvent

ea

er

Yield, %

1 2 3 4 5 6b

CH2 Cl2 trichloroethylene Et2 O toluene THF EtCN

9.08 3.4 4.34 2.38 7.52 27.7

12.5/1 4.03/1 2.25/1 1.85/1 3.37/1 1/8.71

92 88 37 48 59 88

a Ref.

20a. b Performed with 10 mol% (R,R)-45.

vide good yield of the aldol product. Coordinating solvents such as ether and THF give low yields and attenuated selectivities. In ethereal solvents coordination of the Lewis basic oxygen with the silicon species might compete with binding of 45 or benzaldehyde, thus accounting for the lower yields and attenuated selectivity observed. In a non-polar medium such as toluene, the aldol product is formed in signiﬁcantly lower yield with only marginal enantioselectivity. The lack of reactivity observed in the non-polar solvent might reﬂect diﬃculties in generating the kinetically relevant, ionized silyl cation. Catalyst loading also has a large eﬀect on selectivity. Although high catalyst loadings (more than 10 mol%) do not improve enantioselectivity signiﬁcantly, reducing the catalyst loading to less than 3 mol% leads to a slower rate of reaction and attenuated selectivity. At low catalyst loadings the one-phosphoramide pathway that involves a poorly selective boat-like transition structure becomes competitive with the highly selective, twophosphoramide pathway that involves a chair-like transition structure (c.f. Scheme 7.6). Thus, a catalyst loading of 5 mol% or more is necessary for optimum enantioselectivity. The scope of the enolate structure in this aldol reaction is signiﬁcant (Scheme 7.26). Catalyzed reactions with benzaldehyde proceed in excellent yield with modest to good selectivity. The enolate structure does not have a signiﬁcant eﬀect on the rate of the reaction but greatly aﬀects the enantioselectivity of the process. Enolates with larger substituents on the nonparticipating side, for example phenyl and tert-butyl, result in signiﬁcantly lower enantioselectivity than those bearing smaller substituents such as methyl and n-butyl. In these reactions the lower enantioselectivity can be attributed to a lack of facial selectivity at the aldehyde. For this aldolization to be synthetically useful it is important for highly functionalized enolates to tolerate the reaction conditions. For example, ad-

257

258

7 Catalytic Enantioselective Aldol Additions with Chiral Lewis Bases OSiCl3 R

O

O

OH

Me

n-Bu

Ph

O

OH

O

Ph

OH

t-Bu

Ph

Ph O

OH

i-Bu

OH

Ph

O

Ph

Ph

54 (97%); er 9.75/1 OH

TBSO

55 (93%); er 2.92/1

OH

i-Pr

Ph

53 (95%); er 10.1/1

O

66 (95%); er 3.17/1

OH

R

0.5 M in CH2Cl2 −78oC, 2 h

52 (98%); er 12.0/1

51 (98%); er 14.6/1

O

(S,S)-45 (5 mol %) + PhCHO

Ph

56 (94%); er 13.5/1

Scheme 7.26

Addition of diﬀerent trichlorosilyl enolates to benzaldehyde.

dition of enolate bearing a TBSO group results in an excellent yield of 56 with high enantioselectivity. The scope of the reaction with regard to the aldehyde component is also broad. Good to high enantioselectivity can be achieved in the addition of 24 to a variety of aldehydes using 45 (Scheme 7.27). Under these reaction conditions uniformly excellent yields and good enantioselectivity are obtained OSiCl3 n-Bu

O

OH

n-Bu

O

Ph

O

OH

n-Bu

OH

n-Bu

OH

n-Bu

Ph Me 59 (95%); er 21.7/1

OH

n-Bu

67 (92%); er 13.1/1 O

O

58 (94%); er 11.5/1

52 (98%); er 12.0/1

R

OH

n-Bu

Ph

OH

n-Bu

0.5 M in CH2Cl2 −78oC, 2 - 6 h

24 O

O

(S,S)-45 (5-10 mol %) + RCHO

64 (95%); er 12.7/1 O

OH

O

OH

Me

n-Bu

n-Bu Me Me

60 (79%); er 17.5/1

61 (81%); er 24.0/1

Scheme 7.27

Catalyzed addition of 24 to a variety of aldehydes.

62, no reaction

Ph

7.4 Enantioselective Aldol Addition of Achiral Enoxytrichlorosilanes

259

Cl O Ph H

O

Cl Si

OP(NR2)3 OP(NR2)3

Cl n-Bu

vi

aldolization

Cl

O

Ph H

O

Cl Si

OP(NR2)3 OP(NR2)3

Cl vii n-Bu

no aldolization

Fig. 7.9

Reaction of aliphatic aldehydes under the action of phosphoramide catalysis.

from aromatic and unsaturated aldehydes. Aliphatic aldehydes, on the other hand, react signiﬁcantly more slowly than their unsaturated counterparts. When hydrocinnamaldehyde is used as acceptor no aldol product is isolated under standard conditions. Aldehyde structure clearly has a signiﬁcant eﬀect on enantioselectivity. Interestingly, sterically congested aldehydes, especially those with a-branching, result in higher enantioselectivity than unbranched substrates. Although the origin of this eﬀect is unclear, sterically bulky aldehydes presumably increase the energy diﬀerence between the competing chair- and boat-like transition structures. Several explanations have been proposed for the relatively low reactivity of aliphatic aldehydes. This behavior was originally rationalized by considering the low Lewis basicity of these aldehydes compared with related unsaturated aldehydes [69]. Another consideration was the competitive enolization of these substrates in the presence of phosphoramide catalysts [68b]. It has, however, now been shown that aliphatic aldehydes rapidly form a-chloro trichlorosilyl ethers under the reaction conditions and these species are unreactive towards nucleophiles (Figure 7.9) [50]. In the mechanism of phosphoramide-catalyzed aldol additions the cationic hexacoordinate silicon complex has a chloride counterion. When the aldehyde carbonyl is activated in this complex chloride is a potential nucleophile that competes with the enolate. Aldolization of aliphatic aldehydes is, apparently, slow compared with addition of chloride, resulting in formation of the a-chloro silyl ether. The role of ionized chloride in suppressing the aldolization is in agreement with the observation that aliphatic aldehydes are reactive under conditions without phosphoramide. Because these uncatalyzed reactions do not involve ionization, the formation of a-chlorosilyl ether is no longer competitive. Eﬀorts to construct a stereochemical model of absolute stereoselection in these aldol additions have so far been unsuccessful. The absolute conﬁguration of aldol adduct 52 has been determined by formation of the corresponding bromobenzoate then single-crystal X-ray analysis and the absolute

260

7 Catalytic Enantioselective Aldol Additions with Chiral Lewis Bases

conﬁgurations of other aldol products have been assigned by analogy. In aldol additions catalyzed by phosphoramide (S,S)-45, the nucleophile attacks the Si face of the complexed aldehyde, usually forming the S conﬁguration at the hydroxy-bearing carbon. The solution and solid-state structures of several Sn(IV)–phosphoramide complexes have been examined to obtain better understanding of the stereochemical environment crafted by the chiral phosphoramide [70]. Despite this eﬀort no transition structure has yet been proposed that can rationalize the absolute conﬁguration observed in these catalyzed aldol additions. One major concern in development of a model is understanding the mode of ligand binding around silicon. Among the challenges facing the formulation of reasonable transition structures are:

. multiple conﬁgurational possibilities with two phosphoramides around . silicon; conformational ﬂexibility of the phosphoramides; and . the deducing the reactive conformations of the ternary complex from the vast number of potential conﬁgurations and conformations. Computational solution of these problems is currently untenable. Because of the number of heavy atoms and rotatable bonds present in the complex, prediction of the reactive conformation in the ternary complex is diﬃcult at the current level of understanding. In general, performing aldol addition of a trichlorosilyl enolate involves two discrete steps – generation and isolation of the trichlorosilyl enolate then aldolization. In an eﬀort to streamline the process, a method for in situ formation of the reactive enolate has been developed (Scheme 7.28). As described in Section 7.2, trichlorosilyl enolates can be prepared from the corresponding trimethylsilyl enol ethers by use of silicon tetrachloride and a catalytic amount of transition metal salt. In this procedure the trichlorosilyl enolates generated in situ from trimethylsilyl enol ethers by Hg(OAc)2 catalyzed metathesis can, after removal of Me3 SiCl and excess SiCl 4 , be used directly in the subsequent aldol addition. The yield and enantioselectivity of the aldol reaction are not aﬀected by the presence of the mercury salt. This procedure obviates puriﬁcation and handling of the moisturesensitive trichlorosilyl enolate and enables use of shelf-stable trimethylsilyl

OTMS n-Bu

Hg(OAc)2 (1 mol %) SiCl4 (2.0 equiv) CH2Cl2, rt

OSiCl3 n-Bu

(S,S)-45 (5 mol %) RCHO CH2Cl2, −78oC

24

O

OH

n-Bu

Ph Me

59 (89%, 2 steps); er 23.4/1 Scheme 7.28

Aldol addition of 24 generated in situ.

7.4 Enantioselective Aldol Addition of Achiral Enoxytrichlorosilanes

O

OSiCl3 +

n-Bu

Me

H

O

CH2Cl2 rt

Me

n-Bu

OTBS

OTBS

24

68

+

n-Bu

69 (95%); syn/anti, 1/2.4

O

OSiCl3

Me

H

O

CH2Cl2 rt

OBn

24

OH

70

n-Bu

OH Me OBn

71 (92%); syn/anti, 1/2.7

Scheme 7.29

Uncatalyzed aldol additions to chiral aldehydes.

enol ethers. This in situ generation of trichlorosilyl enolates and their use has enhanced the synthetic utility of the phosphoramide-catalyzed aldol reaction. The possibility of substrate-controlled aldol additions of trichlorosilyl enolates has been investigated using lactate-derived chiral aldehydes [68b]. Uncatalyzed reactions of 24 with chiral aldehydes 68 and 70 proceed in high yield at room temperature (Scheme 7.29). The compatibility of common protecting groups on the aldehyde with trichlorosilyl enolates has been demonstrated in these examples. The internal diastereoselection (Section 7.5, Figure 7.13) exerted by the chiral aldehyde slightly favors the anti isomer, although the diastereomeric ratio obtained in this reaction is not synthetically useful [71]. The intrinsic selectivity can be rationalized by use of the Felkin–Ahn model, if boat-like transition structures are considered (Figure 7.10) [72]. The two possible approaches of the nucleophile can be envisaged with the oxygen anti to the incoming nucleophile, as suggested in the Heathcock model [73]. These two conformers viii and ix lead to the two diastereomeric transition structures. Addition of the nucleophile would occur on the sterically less hindered face (H rather than Me), which leads to the observed major anti diastereomer. The nature of the group on the a-oxygen seems to have little eﬀect on the selectivity of the aldolization. This indicates that stereoelectronic factors control the orientation of the oxygen atom in the stereochemistrydetermining step [74]. This observation is also consistent with Heathcock’s analysis of non-chelation-controlled additions to simple a-oxygenated aldehydes [73]. The aldol reaction of an achiral enolate with a chiral aldehyde in the presence of a chiral phosphoramide is an interesting opportunity for double diastereoselection [75]. Diastereoselectivity in aldol additions to chiral aldehydes can be signiﬁcantly enhanced when the sense of internal and external

261

262

7 Catalytic Enantioselective Aldol Additions with Chiral Lewis Bases

Cl

Cl

Si Cl O Me O RO H H viii: favored O

OSiCl3 +

n-Bu

Me

H

Cl Cl

OR

24

n-Bu

68 (R = TBS)

O n-Bu

n-Bu Me OR O H O Si Cl H Cl Cl

anti-69

Cl Si H O OR

Me H ix: disfavored

n-Bu RO H O Me O Cl Si H Cl Cl

syn-69

Fig. 7.10

Diastereoselection for addition of 24 to 68.

diastereoselection are matched (Scheme 7.30). When (S,S)-45 is used in the addition of 24 to 68, the aldol product obtained is the syn diastereomer, albeit with low selectivity. Use of (R,R)-45, on the other hand, provides anti-69 with good diastereoselectivity. In catalyzed additions to the chiral aldehyde 68 the conﬁguration of the catalyst dominates the stereochemical course of the addition. The inherent selectivity in the catalyzed pathway can be examined by employing an achiral catalyst. Modest anti selectivity is observed when 72 is used. The anti selectivity observed with phosphoramide (R,R)-45 is therefore a result of matching internal (from the chiral aldehyde) and external stereoinduction (from chiral catalyst) [71]. O

OSiCl3 +

n-Bu

Me

H

O

cat. (10 mol %)

OH Me

CH2Cl2, −78oC n-Bu

OTBS

OTBS

24

68 Me N O P N N Me

(S,S)-45 (47%) syn/anti, 2.7/1

69 Me N O P N N Me (R,R)-45 (50%) syn/anti, 1/15.6

Scheme 7.30

Catalyzed aldol additions to chiral aldehydes.

Me N O P N N Me 72 (37%) syn/anti, 1/6.7

7.4 Enantioselective Aldol Addition of Achiral Enoxytrichlorosilanes

OSiCl3

O

OH

CH2Cl2 Ph

+ PhCHO

0 oC n-5

n-5

O

OH

73 (90%) syn/anti, 22/1

O

OH

74 (92%) syn/anti, 49/1

n =5, 6, 7 O

OH

75 (90%) syn/anti, 19/1

Scheme 7.31

Uncatalyzed aldol additions of cyclic trichlorosilyl enolates.

7.4.2

Aldol Additions of Cyclic Trichlorosilyl Enolates

In the absence of a Lewis base catalyst, cycloalkanone-derived trichlorosilyl enolates undergo aldol additions to benzaldehyde even at 0 C (Scheme 7.31) [53, 76, 77]. Excellent yields of syn aldol products can be obtained under these conditions. The syn selectivity derived from the E-conﬁgured enolate suggests that the reaction proceeds through a closed, boat-like transition structure. The generality of uncatalyzed aldol additions with 20 has been demonstrated in reactions with a variety of aldehydes (Scheme 7.32). Additions with 20 provide high yields of aldol products with modest to high syn selectivity [76]. The rate of the reaction can be correlated with aldehyde structure. The reactions are signiﬁcantly slower for bulky aldehydes than for smaller ones. Aliphatic aldehydes are less reactive, presumably because of their attenuated Lewis basicity. The aldehyde structure also aﬀects the diastereoselectivity of the aldol reaction. The steric bulk around the aldehyde carbonyl group has a deleterious eﬀect on diastereoselectivity, as is illustrated by comparison of cinnamaldehyde and a-methylcinnamaldehyde. Aliphatic aldehydes are also poorly selective, as is exempliﬁed by aldol addition to cyclohexane carboxaldehyde. In such reactions the energy diﬀerence between the boat- and chair-like transition structures is assumed to be small (vide supra). From simple analysis of the transition structure model for the uncatalyzed process it is clear that the steric bulk of the aldehyde increases steric congestion in the favored boat transition structure (Figure 7.8, iv). The reactivity of 20 has been examined at low temperature to assess the rate of the background reaction in the context of the asymmetric, catalytic process. Aldol addition of 20 to benzaldehyde at 78 C results in 19% conversion after 2 h, indicating that 20 is more reactive than the methyl ketone-derived enolates (c.f. Scheme 7.24). Although the background reac-

263

264

7 Catalytic Enantioselective Aldol Additions with Chiral Lewis Bases

OSiCl3

O

OH

CH2Cl2 R

+ RCHO

0 oC, 1 - 36 h

20 O

O

OH

OH

O

OH

Ph Ph 74 (92%); syn/anti, 49/1 O

76 (83%); syn/anti, 49/1 O

OH

OH

Ph

77 (91%); syn/anti, 36/1 O

OH Ph

Me 78 (86%); syn/anti, 6/1

79 (92%); syn/anti, 1/1

80 (82%); syn/anti, 5.3/1

Scheme 7.32

Uncatalyzed addition of 20 to diﬀerent aldehydes.

tion is rather signiﬁcant in the aldol addition of 20 at low temperature, in the presence of a phosphoramide the catalyzed pathway becomes dominant. An extensive survey of catalysts for the addition of 20 to benzaldehyde showed that several chiral phosphoramides catalyzed the reaction eﬃciently (Scheme 7.33). The stereoselectivity was highly dependent on catalyst structure. Among the catalysts shown below, the stilbene-1,2-diamine-derived catalyst 45 is the most enantioselective. For other catalysts diastereo- and enantioselectivity were lower. The N,N 0 -diphenylphosphoramide 81 results in remarkably high syn diastereoselectivity, albeit with modest enantioselectivity. As shown in the reactions using 82, the phosphorus stereogenic center has no signiﬁcant eﬀect on stereoselectivity. The strong dependence of the diastereomeric ratio on catalyst structure again implies competition between chair- and boat-like transition structures. As discussed in Section 7.9, the primary pathway in the catalyzed aldol addition of trichlorosilyl enolates using 45 involves a chair-like transition structure organized around a cationic, hexacoordinate silicon atom bound by two phosphoramide molecules. Thus for an E-conﬁgured enolate the corresponding anti aldol product is expected. From this diastereoselectivity it is clear that in the reaction catalyzed by 45 the two-phosphoramide pathway is favored, because the anti product is obtained from the E enolate. The reaction with 81, on the other hand, proceeds predominantly through a boat-like transition structure involving one phosphoramide, enabling formation of the syn aldol product from the E enolate. For other catalysts the energy diﬀerence between the two transition structures is very small and so poor diastereoselectivity is obtained.

7.4 Enantioselective Aldol Addition of Achiral Enoxytrichlorosilanes

O

OSiCl3

OH

cat. (10 mol %) Ph

+ PhCHO CH2Cl2, −78oC 2h

20 Ph Ph

Me N O P N N

74 Me N O P N N Me

Me N O P N N

Me

Me

(S,S)-45 (94%) syn/anti, 1/50 er (anti), 27.6/1

(R,R)-64 (91%) syn/anti, 1/2.0 er (syn), 1.00/1 er (anti), 1/3.00

(R)-65 (87%) syn/anti, 3.2/1 er (syn), 1/3.00 er (anti), 3.55/1 Me

Ph Ph

Ph N O P N N Ph

(S,S)-81 (95%) syn/anti, 97/1 er (syn), 3.08/1

N N P O N

Me

(S,S)-82 (94%) syn/anti, 1/1.1 er (syn), 1.08/1 er (anti), 1/1.22

Me N N P N O

Me

(S,R)-82 (96%) syn/anti, 1/3.1 er (syn), 1.06/1 er (anti), 1/1.15

Scheme 7.33

Catalysts for the addition of 20 to benzaldehyde.

Analysis of the absolute conﬁgurations of the products enables construction of a crude transition state model that explains the overall arrangement of the components. The absolute conﬁguration of the aldol product 74 from the reaction using (S,S)-45 has been established to be (2R,1 0 S) by singlecrystal X-ray analysis of the corresponding 4-bromobenzoate. This observation, in conjunction with the results obtained for acyclic Z enolates, oﬀers important insights into the arrangement of the aldehyde relative to the enolate in the transition structure, and into factors that determine the absolute conﬁguration of the aldol adduct (Figure 7.11). The conﬁguration of the chiral phosphoramide determines the face of enolate that undergoes aldolization. On the other hand, the chair or boat transition structure determines the face of aldehyde to be attacked. For example, when (S,S)-45 is used as catalyst, the Si face of the enolate is blocked (Figure 7.11). Placement of aldehyde in a chair-like transition structure will then correctly predict the absolute conﬁguration of the major diastereomer. The phosphoramide 45 is also an eﬀective catalyst for the additions of other cyclic ketone-derived enolates (Scheme 7.34) [77]. Both cyclopentanoneand cycloheptanone-derived enolates provide anti products with good

265

266

7 Catalytic Enantioselective Aldol Additions with Chiral Lewis Bases

configuration of phosphoramide

chair or boat transition structure

Si-C(2) R

phosphoramide configuration O

Si

R

O H

Cl3SiO

OH

Re

Re-C(2)

chair or boat transition structure

Fig. 7.11

Factors leading to the observed conﬁguration.

enantioselectivity. It is interesting to note that diastereoselectivity for these reactions is sensitive to the rate of mixing; slow addition of aldehyde is therefore necessary to obtain reproducible and high diastereoselectivity. The synthetic utility of the aldol addition of 20 has been expanded by examining a wide range of aldehydes of diﬀerent structure (Scheme 7.35). Enantioselectivity is usually good to excellent for the anti diastereomers. For addition of 20 to diﬀerent aldehydes excellent anti selectivity is always obtained except for use of phenylpropargyl aldehyde. In this particular system there is no obvious relationship between stereoselectivity and aldehyde structure. Steric bulk around the aldehyde carbonyl seems to enhance diastereoselectivity and enantioselectivity in additions to benzaldehyde and to 1-naphthaldehyde. The lower diastereoselectivity observed in the addition to phenylpropargyl aldehyde can be attributed to the lack of facial diﬀerentiation for the aldehyde, because the substituents (H and acetylenic groups) are similar in size. Addition of 20 to aliphatic aldehydes does not, unfortunately, furnish the corresponding aldol products under catalysis by chiral phosphoramides. This may be caused by competitive enolization of the aldehyde by the basic O

OSiCl3

OH

(S,S)-45 (10 mol %) + PhCHO n-5

O

OH

74 (95%) syn/anti, 1/61 er (anti), 27.6/1

Ph

CH2Cl2, −78oC

n = 5, 6, 7

n-5

O

OH

73 (98%) syn/anti, 1/22 er (anti), 7.13/1

O

OH

75 (91%) syn/anti, 1/17 er (anti), 9.87/1

Scheme 7.34

Catalyzed aldol additions of a variety of cyclic enolates to benzaldehyde.

7.4 Enantioselective Aldol Addition of Achiral Enoxytrichlorosilanes

OSiCl3

O

OH

(S,S)-45 (10 mol %) R

+ RCHO CH2Cl2, −78oC

20

O

O

OH

O

OH

OH

Ph Ph 74 (95%) syn/anti, 1/61 er (anti), 27.6/1

O

76 (94%) syn/anti, <1/99 er (anti), 15.7/1

OH

77 (90%) syn/anti, 1/5.3 er (anti), 10.1/1

O

OH Ph Me

83 (94%) syn/anti, <1/99 er (anti), 65.7/1

78 (98%) syn/anti, <1/99 er (anti), 24.0/1

Scheme 7.35

Addition of 20 to a variety of aldehydes catalyzed by 45.

enolate/phosphoramide complex and/or formation of the a-chlorosilyl ether as discussed for addition of methyl ketone-derived enolates. 7.4.3

Addition of Acyclic Ethyl Ketone-derived Enolates

The control of enolate geometry is a major synthetic challenge in the addition of acyclic, substituted trichlorosilyl enolates, because the diastereomeric composition of the aldol products can reﬂect the E/Z ratio in the enolates [67]. Fortunately, the preparation of geometrically-deﬁned trichlorosilyl enolates has been described (Section 7.2) [45]. The strong preference for the chair-like transition structure has been demonstrated for aldol additions of cyclic ketone-derived enolates (E-conﬁgured) under suitable phosphoramide catalysis. A high degree of diastereocontrol can be achieved for cyclic enolate additions. Use of the E enolate correlates with formation of the anti aldol product. The initial study of an acyclic ketone-derived enolate was performed using (Z)-18 derived from propiophenone. At 0 C, aldol addition of 18 to a variety of aldehydes proceeds at an appreciable rate and provides aldol products in good yield (Scheme 7.36). All of the reactions provide anti aldol products as the major diastereomers, albeit with low diastereoselectivity. Correlation

267

268

7 Catalytic Enantioselective Aldol Additions with Chiral Lewis Bases

Ph

OSiCl3 Me

O Ph

+ RCHO

OH

Ph

R

0 oC, 10 - 16 h

(Z )-18 O

OH

CH2Cl2

O

OH

O

Ph

Ph

Ph Me

84 (97%) syn/anti, 1/2.3

85 (95%) syn/anti, 1/1.9

Ph

Me

86 (89%) syn/anti, 1/1.9

OH

O

Ph

OH

Me

Me

O

Me

OH

Ph Me 87 (89%) syn/anti, 1/2.2

Ph

Ph Me

Me

88 (64%) syn/anti, 1/2.2

Scheme 7.36

Uncatalyzed addition of (Z)-18 to a variety of aldehydes.

of the Z enolate with the anti aldol product suggests involvement of a boat transition structure. The modest selectivity of uncatalyzed addition of the Z enolate contrasts with uncatalyzed addition of cycloalkanone-derived enolates, which selectively aﬀord syn products. The low diastereoselectivity observed is because of the unfavorable interaction between the methyl substituent and an apical chloride in the preferred boat transition structure. Acyclic ethyl ketone-derived enolates are relatively unreactive toward aldehydes compared with the other types of enolate discussed above. The observed low reactivity of 18 might be because of steric interaction between the a-methyl substituent of the enolate. Aldol addition of 18 to benzaldehyde is much slower at low temperature. The uncatalyzed aldol addition gives only 1% of isolated aldol product after 6 h at 78 C [76] whereas in the presence of a catalytic amount of a phosphoramide aldol reactions of (Z)-18 are signiﬁcantly accelerated (Scheme 7.37). Although all phosphoramides give good yields of the benzaldehydederived aldol products, stereoselectivity is highly variable and depends on catalyst structure. Once again, the stilbene-1,2-diamine-derived phosphoramide 45 is highly selective, providing good syn selectivity combined with excellent enantioselectivity. The observed syn selectivity suggests that with 45, the predominant reaction pathway is through a chair-like transition structure. The scope of this reaction has been explored using phosphoramide 45. Increasing the catalyst loading from 10 to 15 mol% not only improves yields of aldol products but also leads to higher diastereoselectivity. Addition of 18

7.4 Enantioselective Aldol Addition of Achiral Enoxytrichlorosilanes

Ph

OSiCl3 Me

O + PhCHO

CH2Cl2, −78oC

Ph

Ph Me syn-84

18: Z/E >20/1 Me N O P N N Me

Ph Ph

OH

cat. (10 mol %)

Me N O P N N Me

Me N O P N N Me

(S,S)-45 (77%) syn/anti, 11.5/1 er (syn), 32.3/1

(R)-65 (61%) syn/anti, 1/1.5 er (syn), 1/2.23 er (anti), 1.13/1

(R,R)-64 (72%) syn/anti, 1/2.7 er (syn), 1/1.38 er(anti), 1.08/1

Scheme 7.37

Catalyzed addition of (Z)-18 to benzaldehyde.

to a variety of aldehydes is eﬃciently catalyzed by 45, with moderate to good syn relative diastereoselectivity and good to excellent enantioselectivity for the syn diastereomers (Scheme 7.38). Good syn selectivity has been observed for most conjugated aldehydes. Signiﬁcantly attenuated diastereoselectivity is observed for sterically de-

Ph

OSiCl3 Me

O CH2Cl2, −78oC

OH

Ph

Ph

Ph Me

O Ph

85 (97%) syn/anti, 9.4/1 er (syn), 24.0/1

Me

Br

89 (89%) syn/anti, 12/1 er (syn), 49.0/1

OH

Me

OH

O

Ph

84 (95%) syn/anti, 18/1 er (syn), 39.0/1

Ph

Me

OH

O

Me

O

R

Ph

+ RCHO

(Z )-18 Z/E >20/1 O

OH

(S,S)-45 (15 mol %)

90 (96%) syn/anti, 3/1 er (syn), 11.5/1 O

OH

Ph

Me Me

86 (94%) syn/anti, 7.0/1 er (syn), 21.2/1

Scheme 7.38

Catalyzed addition of (Z)-18 to various aldehydes.

OH

Ph Me

Ph

87, (92%) syn/anti, 1/3.5 er (syn), 3.76/1

269

270

7 Catalytic Enantioselective Aldol Additions with Chiral Lewis Bases

Et

O

OSiCl3 Me

Et

+ RCHO

OH

Et

R

CH2Cl2, −78oC

(Z )-11 Z/E, 16/1 O

OH

(R,R)-45 (15 mol %)

O

Ph

Me

OH

O

Et

O

Et

Me

Me

91 (84%) syn/anti, 16/1 er (syn), 21/1 O

OH

Me

92 (80%) syn/anti, 8/1 er (syn), 9/1

93 (79%) syn/anti, 4/1 er (syn), 5/1

O

OH

OH

Ph Et

Et Me 94 (85%) syn/anti, 5/1 er (syn), 3/1

Me 95 (45%) syn/anti, 1/2 er (syn), 1/1

Scheme 7.39

Catalyzed aldol additions of (Z)-11 to diﬀerent aldehydes.

manding aldehydes. For example, the diastereoselectivity in the addition to 1-naphthaldehyde is only modestly syn selective. In contrast, addition to phenylpropargyl aldehyde is surprisingly anti selective. The sense of stereoinduction enforced by the chiral catalyst (S,S)-45 is consistent with the model described previously (Figure 7.11). The absolute conﬁguration of the aldol product 84 has been unambiguously assigned as (2S,3S) by X-ray analysis of the corresponding 4-bromobenzoate ester. The phosphoramide (S,S)-45 blocks the Si face of the enolate and the resulting chair-like assembly of the enolate and the aldehyde leads to the stereoselectivity observed. The reactions of other ethyl ketone enolates have also been investigated. When catalyzed by 45 aldol additions of the 3-pentanone-derived trichlorosilyl enolate (Z)-11, Z/E ratio 16:1, give the syn aldol products selectively (Scheme 7.39) [45]. For addition to benzaldehyde the diastereoselectivity reﬂects the Z/E ratio of the starting enolate, indicating that the chair-like transition structure is strongly favored. The attenuated diastereoselectivity in additions to other aldehydes suggests that competitive boat transition structures become operative. Along with decreases in diastereoselectivity, enantioselectivity for other aldehydes is reduced signiﬁcantly. Unlike the perfect correlation between starting enolate geometry and aldol product conﬁguration observed in the addition of Z enolates, diastereo-

7.4 Enantioselective Aldol Addition of Achiral Enoxytrichlorosilanes

O

OSiCl3

OH

(R,R)-45 (15 mol %) Et

+ RCHO

Et

CH2Cl2, −78oC

Me (E )-11 E/Z, 15/1 O

271

OH

Et

O

Ph

OH

Et

Me

O

OH O

Et Me

91 (86%) syn/anti, 1/1 er (anti), 8/1

R Me

92 (82%) syn/anti, 1/8 er (anti), 12/1

Me 93 (76%) syn/anti, 1/2 er (anti), 3/1

Scheme 7.40

Catalyzed additions of (E)-11 to diﬀerent aldehydes.

selection is poor for addition of (E)-11 to benzaldehyde, although the enantioselectivity observed in the anti pathway is good (Scheme 7.40) [45]. In contrast, addition to 1-naphthaldehyde provides the anti aldol product with good diastereo- and enantioselectivity. The diﬀerence between results from the addition of E and Z enolates is striking. As discussed in the previous section, cycloalkanone-derived trichlorosilyl enolates (E enolates) furnish exclusively anti aldol products under the action of catalysis by 45. Z Enolates, on the other hand, produce predominantly syn aldol products. Acyclic E-trichlorosilyl enolates do not undergo selective aldol additions, however, furnishing aldol products with unpredictable diastereomeric ratios. This lack of diastereoselection is presumably because of competition between the chair and boat transition structures (Figure 7.12). In the structure xi the least sterically demanding group would be placed anti-periplanar to the enolate CaO bond to minimize the A 1; 3 allylic interaction. The nonparticipating group will then be in close proximity to the silicon center bearing two, large phosphoramide molecules. This steric congestion might ultimately cause the conformational change from the chair to the boat transition structure.

H Me HR

H Cl H O O

Cl

L Si

L Cl

Me

x (Z-enolate,chair)

Cl H H 1,3 H A Me SiLn O R Me O H xi (E-enolate, chair)

Fig. 7.12

Transition structures for aldol additions of (Z)-11 and (E)-11.

H R H Me

H Me O

Cl

O

SiLn H xii (E-enolate, boat)

272

7 Catalytic Enantioselective Aldol Additions with Chiral Lewis Bases

For these reasons, reactions of cycloalkanone-derived enolates are an exception in E enolate additions. Because of the absence of steric congestion on the non-participating side, the reaction can proceed predominantly through the chair transition structure, resulting in high anti diastereoselectivity. In the acyclic E enolate, unfavorable steric interactions in the chair transition structure enable competition from the boat transition structure. Therefore, only modest anti selectivity can be achieved with these enolates.

7.5

Diastereoselective Additions of Chiral Enoxytrichlorosilanes

The use of enoxytrichlorosilanes bearing a stereogenic center is an important extension of Lewis-base-catalyzed aldol additions in organic synthesis. If the reactants contain stereogenic centers the resulting aldol products are diastereomeric. Thus, in theory, stereoinduction can arise from both the resident stereogenic center and the external chiral catalyst. In additions of achiral trichlorosilyl enolates the stereochemical course of the reaction is governed by two factors (Figure 7.13) [71]. The ﬁrst is the relative diastereoselection which reﬂects the relative topicity (like or unlike) [78] of the combination of the two reacting faces (enolate and carbonyl group). In highly organized aldol additions this is often interpreted in terms of chair/boat selectivity in the transition structure. The relative diastereoselec-

(a) relative diastereoselection:

R

OSiCl3 Me

O + R'CHO

OH

R

R'

O +

OH

R

R'

Me

Me

syn

anti

(b) external enantioselection:

R

OSiCl3 Me

chiral phosphoramide + R'CHO

O

OH

R

R'

O +

OH

R

R'

Me

Me

syn

ent-syn

(c) internal diastereoselection:

R*

OSiCl3 Me

O + R'CHO

Fig. 7.13

Three types of stereoselection process.

OH

R*

R'

O

OH

+ R*

R'

Me

Me

syn-A

syn-B

7.5 Diastereoselective Additions of Chiral Enoxytrichlorosilanes

Me α

OSiCl3 R'

OR xiii

OR

Me β

OSiCl3 R'

RO

xiv

α

OSiCl3 R'

Me xv

Chart 7.2

Three classes of chiral trichlorosilyl enolate.

tion pertains to a-substituted enolates only, because the term refers to the relative conﬁguration of the two substituents (like or unlike) at the newly created stereogenic centers. The other is the absolute stereoselection (external stereoselection) determined by the chiral Lewis base catalyst. This term is used to describe the enantiofacial outcome at the newly created stereogenic centers. In additions of chiral enolates there is yet another stereoselection process that is controlled by the resident stereogenic center. The diastereoselection resulting from the eﬀect of the stereogenic centers in either of the reactants is referred as internal diastereoselection. When a chiral catalyst is used in conjunction with a chiral enolate there is a possibility of double diastereodiﬀerentiation [2c]. In a matched case in which the sense of external stereoinduction coincides with the internal stereoinduction, the diastereoselectivity of the reaction can be considerably enhanced [79]. Three classes of chiral trichlorosilyl enolate have been studied to investigate the eﬀect of the resident stereogenic center in the context of Lewisbase-catalyzed aldol addition (Chart 7.2). Two of these enolates (xiii and xiv) bear heteroatom-based stereogenic centers at the a- and b-carbon atoms on the non-participating side and one (xv) bears a carbon-based stereogenic center on the a-carbon of the non-participating side. 7.5.1

Aldol Addition of Lactate-derived Enoxytrichlorosilanes Methyl Ketone-derived Enolates Aldol additions of lactate-derived trichlorosilyl enolates in the absence of a Lewis base catalyst proceed at room temperature (Scheme 7.41) [80]. The aldol products derived from benzaldehyde are obtained in good yields as mixtures of diastereomers. Although aldol additions of these enolates are poorly selective, the anti diastereomers are always favored. Interestingly, the size of the protecting group has some eﬀect on diastereoselectivity; smaller protecting groups result in better selectivity. The observed anti selectivity can be explained by the model depicted in Figure 7.14. Initial coordination of the aldehyde with the Lewis acidic silicon center results in formation of a trigonal bipyramidal species (c.f. Figure 7.7, Section 7.4). Typically, uncatalyzed aldol additions of trichlorosilyl enolates proceed via boat-like transition structures [65, 72]. The oxygen sub7.5.1.1

273

7 Catalytic Enantioselective Aldol Additions with Chiral Lewis Bases

274

Hg(OAc)2 SiCl4

OTMS Me

OSiCl3

CH2Cl2

OR

O

PhCHO

Me

rt

OR

OH

Me

OH

Ph

+

Ph

OR

anti O

O

OH

Me

OH

Me

Ph

O Me

Ph OPiv

OTBS 96 (82%) syn/anti, 1/1.2

syn

OH Ph

OBn

97 (71%) syn/anti, 1/2.4

98 (75%) syn/anti, 1/3.4

Scheme 7.41

Uncatalyzed aldol additions of lactate-derived enolates.

Cl OSiCl3

PhCHO

Me

H

O

Me

H

Ph OR

H Ph

OR A

Cl3 Si O O

Cl Cl

O

Me OR

Si

anti

1,3

Fig. 7.14

Stereochemical model for uncatalyzed addition of lactate-derived enolates.

stituent on the enolate is expected to be antiperiplanar to the enol oxygen to minimize the net dipole [73]. In this conformation, the anti diastereomer is preferentially formed by approach of the aldehyde to the less hindered face of the enolate. This model is consistent with the observation that the bulkier protecting groups result in lower diastereoselectivity, because the A 1; 3-type interaction becomes substantial if the favored conformer contains a large protecting group. Uncatalyzed additions of 29 to other aldehydes have been examined (Scheme 7.42). Unfortunately, the substrate-induced diastereoselection is OTMS Me

Hg(OAc)2 SiCl4

OSiCl3

CH2Cl2

OTBS 99

O Ph

OTBS 100 (35%) syn/anti, 1/1

Me

rt

OTBS 29

OH

O Me

O RCHO

Me

OH R

O

OH

101 (55%) syn/anti, 1/3

R

syn

OH

Me OTBS

+

OTBS

anti

Me

OH

Me OTBS 102 (66%) syn/anti, 2.3/1

Scheme 7.42

Uncatalyzed aldol addition of 29 to a variety of aldehydes.

7.5 Diastereoselective Additions of Chiral Enoxytrichlorosilanes

Hg(OAc)2 SiCl4

OTMS Me

OSiCl3 Me

CH2Cl2

OTBS

O

OH

Ph +

Ph

OTBS

anti

syn

Me N O P N N Me

72 =

OH

Me

−78 oC, 4.5 h

OTBS 29

99

PhCHO 72 (5 mol %)

275

96 (81%) syn/anti, 1.2/1

Scheme 7.43

Addition of 29 to benzaldehyde catalyzed by 72.

only modest and the selectivity cannot easily be rationalized because syn/ anti selectivity depends on aldehyde structure. Also, for aliphatic aldehydes, signiﬁcantly reduced yields are obtained. In the presence of a catalytic amount of a phosphoramide the rate increases signiﬁcantly. The intrinsic selectivity induced by the stereogenic center has been examined for use of achiral catalyst 72 (Scheme 7.43). Although the product obtained is slightly enriched in the syn diastereomer, the syn/anti ratio is almost negligible, indicating that the resident stereogenic center has little eﬀect on diastereoselectivity. Double stereodiﬀerentiation using a chiral catalyst provides a dramatic matched/mismatched eﬀect in the aldol addition of 29, 30 and 31. (Scheme 7.44). When (S,S)-45 is used marginal improvement of syn selectivity is obHg(OAc)2 SiCl4

OTMS Me

OSiCl3 Me

CH2Cl2

OR

PhCHO 45 (5 mol %)

O

−78 oC, 2 h

OR

OH

Me

Ph +

O (R,R)-45

O

OH Ph

96 (85%) syn/anti, 73/1 O

O

OH

OTBS 96 (85%) syn/anti, 1.5/1

Me

anti

OH Ph

OBn

97 (78%) syn/anti, 20/1

Ph

(S,S)-45

Ph

O Me

OPiv

OTBS

Me

OH

Me

98 (77%) syn/anti, 11/1

OH Ph

OPiv 97 (78%) syn/anti, 3.4/1

Ph

OR

syn

Me

OH

O

OH

Me

Ph OBn

98 (78%) syn/anti, 1/1.1

Scheme 7.44

Catalyzed addition of lactate-derived trichlorosilyl enolates to benzaldehyde.

276

7 Catalytic Enantioselective Aldol Additions with Chiral Lewis Bases

Cl Cl (R,R)-45

(R2N)3PO (R2N)3PO

O O R Cl H H Me OTBS Si

O Me

OH R

OTBS

syn

xvi OSiCl3 Me

+ RCHO Cl

OTBS 29

Cl (S,S)-45

(R2N)3PO (R2N)3PO

O Si O R H Cl Me OTBS H xvii

O Me

OH R

OTBS

anti

Fig. 7.15

Stereochemical course of aldol addition of 29.

served. The use of (R,R)-45, on the other hand, results in the 1,4-syn aldol product with excellent diastereoselectivity. The diﬀerent protecting groups aﬀect the extent of diastereoselection and diastereoselectivity decreases in the order OTBS > OPiv > OBn. This order suggests that a pathway involving chelation of the cationic silicon might become possible with more coordinating oxygen functions (vide infra) [81]. The use of (R,R)-45 represents the matched case in which the sense of external stereoinduction is the same as that of internal stereoinduction, whereas in the use of (S,S)-45 the sense of stereoinduction is opposite, leading to attenuated selectivity. The predominant syn selectivity is rationalized by the model depicted in Figure 7.15. The preferred conformation of the resident stereogenic center again places the oxygen substituent in the plane of the enol double bond as explained above. The enolate faces are discriminated not only by the chiral phosphoramide but also by the substituents on the resident stereogenic center. In the chair-like transition structure xvi attack of the enolate on the Re face of the aldehyde leads to the syn diastereomer observed. In this model (R,R)-45 blocks the more sterically hindered face (syn to the methyl group) of the enolate, thus matching the internal and external stereoinduction. On the other hand (S,S)-45 prevents approach of the aldehyde from sterically less hindered face (anti to the methyl group) of the enolate (xvii, Figure 7.15). In this case, the internal and external stereoinductions oppose each other, resulting in signiﬁcantly attenuated diastereoselectivity.

7.5 Diastereoselective Additions of Chiral Enoxytrichlorosilanes

277

Cl OP(NR"2)3 OSiCl3 Me

OP(NR"2)3

Cl Cl

Si

O O

O Me

RO Me H H R'

+ R'CHO OR

OH R'

OR

xviii

anti

Fig. 7.16

Competitive, chelated transition structure for catalyzed aldol addition of lactate-derived trichlorosilyl enolates.

The attenuated diastereoselectivity when protecting groups other than OTBS are used might indicate the intervention of a competitive chelated transition structure (Figure 7.16). In the transition structure xviii, the oxygen on the non-participating side is coordinated to the Lewis acidic silicon. As the coordinating capacity of oxygen increases, the transition structure xviii might be favorable, and attenuated diastereoselectivity is observed. Catalyzed additions of 29 to oleﬁnic aldehydes have also been demonstrated. For example, addition of 29 to crotonaldehyde catalyzed by 45 gives the corresponding aldol product in good yield (Scheme 7.45). Although the diastereoselectivity obtained in these reactions is lower, the stereochemical trend remains the same. The (R,R)-45 catalyst is the matched case providing syn-102 with good diastereoselectivity. Ethyl Ketone-derived Enolates The stereochemical course of addition of the corresponding ethyl ketonederived enolates incorporates all three forms of stereoselection [82]. Aldol addition of (Z)-103 under the action of phosphoramide catalysis provides the syn,syn (relative, internal) aldol product with high selectivity (Scheme 7.46). A survey of chiral and achiral phosphoramides shows a remarkable 7.5.1.2

OTMS Me OTBS 99

Hg(OAc)2 SiCl4 CH2Cl2

Me

CHO OSiCl3 Me 45 (5 mol %) Me OTBS 29

Scheme 7.45

Catalyzed addition of 29 to crotonaldehyde.

−78 oC, 4.5 h

O

OH Me

OTBS

(S,S)-45

102 (80%) syn/anti, 1.2/1

(R,R)-45

102 (81%) syn/anti, 6.2/1

278

7 Catalytic Enantioselective Aldol Additions with Chiral Lewis Bases

OSiCl3 Me +

Me

O

cat. (15 mol %) PhCHO CH2Cl2, −78oC 2-8h

OTBS

Ph

Me N O P N N Me

(R,R)-45 (87%) rel. syn/anti, 16/1 int. syn/anti, >50/1

Ph

TBSO

Me

(syn,syn)-104

(Z )-103 Ph

OH

Me

Ph Ph

Me N O P N N Me

(S,S)-45 (80%) rel. syn/anti, 15/1 int. syn/anti, 30/1

Ph Ph

Ph N O P N N Ph

(R,R)-81 (65%) rel. syn/anti, 15/1 int. syn/anti, 3/1

Me N O Me P Me N N Me Me Me HMPA (79%) rel. syn/anti, 15/1 int. syn/anti, 30/1

Scheme 7.46

Aldol additions of (Z)-103 to benzaldehyde catalyzed by diﬀerent phosphoramides.

trend. All the phosphoramides yield the syn relative aldol product with perfect Z/E to syn/anti correlation, indicating that a chair-like transition structure is maintained. These results are intriguing considering that bulky phosphoramides such as (R,R)-81 favor the boat-like transition state in the addition of the cyclohexanone-derived trichlorosilyl enolate [76]. Internal stereoselectivity varies for diﬀerent catalyst structures. For the stilbene-1,2-diamine derived catalyst, (R,R)-45, excellent internal syn selectivity is obtained. Use of the enantiomeric catalyst (S,S)-45 does not reverse the sense of internal diastereoselection and the aldol product is again obtained with good internal syn selectivity. These observations indicate the overwhelming inﬂuence of the resident stereogenic center and the stereochemical course of the aldol addition is determined solely by this factor. To support this explanation, additions using a variety of achiral phosphoramides including HMPA demonstrate that the internal syn aldol product is preferentially formed under catalyzed conditions, irrespective of catalyst conﬁguration. Bulky phosphoramides such as (R,R)-81 result in attenuated internal diastereoselectivity compared with that resulting from other phosphoramides. The observed stereochemical outcome can be explained by the nonchelation model that places the OTBS substituent in the enolate in plane with the enolate double bond to minimize the dipole moment (Figure 7.17) [73, 80, 82]. The two enolate faces are diﬀerentiated by the size of the groups on the stereogenic center (H compared with Me), and the aldehyde approaches from the less hindered Si face of the Z enolate. The chair-like arrangement of the aldehyde in the transition structure leads to the formation of the observed syn,syn diastereomer. The low selectivity observed when bulky phosphoramides are used can be rationalized by the intervention of another transition structure, xx [82]. These phosphoramides are known to

7.5 Diastereoselective Additions of Chiral Enoxytrichlorosilanes

H H Cl OP(NR2)3 Me Si OP(NR2)3 O Ph O Cl

OSiCl3 Me

Me

O

TBSO

OTBS (Z )-103

279

OH

Me

Ph

TBSO

Me

Me

(R2N)3PO

(syn,syn)-104

xix

+ PhCHO

TBS O Cl H OP(NR2)3 O Si O Cl Ph Me

Me H

O

OH

Me

Ph

TBSO

Me

(syn,anti )-104

xx Fig. 7.17

Proposed transition structures for aldol addition of (Z)-103.

favor boat-like transition structures via a mechanism that involves only one phosphoramide in the stereodetermining step [66]. In this pentacoordinate species, it is possible that the silyloxy group could coordinate the Lewis acidic silicon to form an octahedral, cationic silicon intermediate. This internal coordination might favor the chair-like arrangement over the usual boat-like transition structure for these phosphoramides. Although the coordinating capacity of the TBS ether is modest at best [81], the proximity of the silyloxy group to the cationic silicon is believed to enhance the possibility of this type of chelation [83]. In this model, the aldehyde now approaches from the Re face of the enolate leading to the syn,anti diastereomer. The aldol additions of (Z)-103 to diﬀerent aldehydes illustrate the generality of this process (Scheme 7.47). The in-situ generation of trichlorosilyl enolate (Z)-103 from the corresponding TMS enol ether further demonstrates not only the synthetic utility of this reagent but also the improved OTMS Me

Me

Hg(OAc)2 SiCl4 CH2Cl2 rt

OTBS (Z )-103 O

TBSO

O

OH

CH2Cl2, −78oC

104 (87%) rel. syn/anti, 15/1 int. syn/anti, >50/1

TBSO

OH R

TBSO

Me O

O

Ph Me

O Me

syn,syn

OH

Me

Me

(R,R)-45 (5 mol %) RCHO

Me

105 (82%) rel. syn/anti, 15/1 int. syn/anti, >50/1

Scheme 7.47

Aldol addition of (Z)-103 to diﬀerent aldehydes.

O

OH

Me

Me TBSO

TBSO

OH

Me Me

Me

106 (85%) rel. syn/anti, 13/1 int. syn/anti, >50/1

107 (79%) rel. syn/anti, 15/1 int. syn/anti, >50/1

Ph

280

7 Catalytic Enantioselective Aldol Additions with Chiral Lewis Bases

yield and selectivity of the overall process. The addition can be catalyzed by either (R,R)-45 or HMPA, and syn,syn-aldol products can always be obtained selectively. The relative and internal diastereoselectivities are all perfect when (R,R)-45 is used as the catalyst. The diastereoselectivity and yield are slightly attenuated under the action of catalysis by HMPA. 7.5.2

Aldol Addition of b-Hydroxy-a-Methyl Ketone-derived Enoxytrichlorosilanes Methyl Ketone-derived Enolates The eﬀects of a-methyl and b-hydroxy groups on the stereochemical course of aldol additions with trichlorosilyl enolates have been investigated. This type of enolate structure is synthetically important because the resulting aldol product resembles the highly oxygenated structural motif for a variety of polypropionate natural products. Not surprisingly, diastereoselective aldol additions of this type of enolate have already been demonstrated for lithium, boron, and tin enolate aldol additions [1d, 79, 84]. The eﬀect of the a-methyl stereogenic center has been determined in aldol additions of the methyl ketone-derived trichlorosilyl enolate (S)-108 (Scheme 7.48) [85]. The addition of 108 using (R,R)-45 as catalyst provides the syn aldol product selectively. The use of (S,S)-45 enables formation of anti-109, albeit with attenuated selectivity. The intrinsic internal selectivity arising from an a-methyl stereogenic center is determined by examining the diastereoselectivity of the aldol addition using the achiral phosphoramide 72. The internal selectivity is low but slightly favors the 1,4-syn diastereomer. The inherent selectivity is rationalized by means of a transition structure model in which transition structure xxi (Figure 7.18) involves octahedral, cationic silicon in a chair-like arrangement of groups. To avoid steric interaction between the phosphoramide-bound silicon and the non-participating substituent on the enolate the least sterically demanding substituent (hydrogen) is placed in plane with the enolate CaO bond. This model predicts 7.5.2.1

TBSO

OSiCl3

cat. (5 mol %) + PhCHO

CH2Cl2, −78 oC

Me (S)-108

Ph Ph

Me N O P N N Me

(R,R)-45 (79%) syn/anti, 10/1

Ph Ph

TBSO

O

Ph Me syn -109

Me N O P N N

Me N O P N N

Me

Me

(S,S)-45 (76%) syn/anti, 1/6

OH

72 (51%) syn/anti, 5/1

Scheme 7.48

Aldol addition of 108 to benzaldehyde catalyzed by phosphoramides.

OH

+

Ph

anti-109

7.5 Diastereoselective Additions of Chiral Enoxytrichlorosilanes

TBSO

OSiCl3

H H Cl Me OP(NR2)3 RL Si OP(NR2)3 O Ph O Cl

72 + PhCHO

Me (S)-108

TBSO

281

O

OH Ph

Me

RL = CH2OTBS xxi

syn-109

Fig. 7.18

Stereochemical course of aldol addition of (S)-108 to benzaldehyde.

approach of benzaldehyde from the less sterically demanding methyl group side, leading to the syn diastereomer. The generality of this aldol addition has been investigated with a wide variety of aldehydes (Scheme 7.49). The trichlorosilyl enolate 108 generated in situ (from TMS enol ether 110) reacts with aromatic, conjugated, and TBSO

SiCl4 Hg(OAc)2

OTMS

CH2Cl2

Me (S)-110

OSiCl3

TBSO

RCHO TBSO 45 (10 mol %)

O

R

CH2Cl2, −78oC

Me

OH

OH

108

O

OH

TBSO

syn

O

OH

Ph

anti

TBSO

O

112 (34%) syn/anti, 10.1/1 TBSO

OH

Me Me

OH

Me

111 (81%) syn/anti, 8.0/1

OH

O

O

Ph

Me

109 (80%) syn/anti, 19.0/1 TBSO

TBSO Ph

Me

O

OH Me

Me

Me

Me

113 (78%) syn/anti, 24.0/1

Me

114 (85%) syn/anti, 4.9/1

Me Me

115 (73%) syn/anti, 27.9/1

Using (S,S)-45 TBSO

O

OH

TBSO

O

Ph Me

109 (75%) syn/anti, 1/7.3 O

TBSO

O

Me

113 (75%) syn/anti, 1/4.6

Ph 112 (22%) syn/anti, 1/2.5 TBSO

OH

Me Me

OH

Me

111 (82%) syn/anti, 1/4.3

OH

O

Ph

Me

TBSO

TBSO

OH

O

OH Me

Me Me 114 (83%) syn/anti, 1/2.9

Scheme 7.49

Aldol addition of 108 to diﬀerent aldehydes catalyzed by 45.

R

Me

Using (R,R)-45 TBSO

+

Me

Me Me

115 (78%) syn/anti, 1/6.5

282

7 Catalytic Enantioselective Aldol Additions with Chiral Lewis Bases Tab. 7.4

Catalyzed aldol additions of 35 and 116 to benzaldehyde. relative RO

OSiCl3 Me +

O H

cat. (10 mol %) Ph

Me 35 : R = TBS 116: R = TIPS

RO

CH2Cl2, −78 ˚C

OH

O

Ph Me

Me 117: R = TBS

internal 118: R = TIPS

Entry

Enolate

Z/E

Catalyst

Yield

Relative dr (syn/anti)

Internal dr a (syn/anti)

1 2 3 4 5 6 7

(Z)-35 (Z)-35 (E)-35 (E)-35 (Z)-116 (Z)-116 (Z)-116

50/1 50/1 1/50 1/50 50/1 50/1 50/1

(R,R)-45 (S,S)-45 (R,R)-45 (S,S)-45 (R,R)-45 (S,S)-45 72

72 82 72 72 84 82 81

9/1 12/1 1/4 1/2 53/1 32/1 27/1

10/1 1/7 6/1 2/1 24/1 1/8 5/1

a Ratio

of major relative diastereomer.

aliphatic aldehydes. Additions to aromatic aldehydes result in high yields and good diastereoselectivity. Additions to the oleﬁnic aldehydes always result in good yields, but selectivity is quite variable. Interestingly, steric bulk around the carbonyl group has a beneﬁcial eﬀect on diastereoselectivity. Unhindered aliphatic aldehydes are signiﬁcantly less reactive, resulting n only modest yields of the aldol products. Ethyl Ketone-derived Enolates When the corresponding ethyl ketone enolate reacts with aldehydes, an additional stereogenic center is formed (Table 7.4) [86]. The reactions of both (Z)-35 and (E)-35 have been examined, enabling the eﬀect of enolate geometry on diastereoselectivity to be probed. Several interesting trends can be noted from aldol additions of 35 and 116 to benzaldehyde. Additions of (Z)35 are generally syn (relative) selective, indicating that a chair-like transition structure is involved in the dominant pathway (Table 7.4, entries 1 and 2). The diastereoselectivity observed for (E)-35 is only marginal, however, and the anti (relative) diastereomer is preferred (Table 7.4, entries 3 and 4). In both reactions the E/Z ratio of the enolate does not translate strictly into the relative syn/anti ratio. This observation can be accounted for by the presence of competitive boat-like transition structures that lead to the minor diastereomers. Fortunately, the relative diastereoselectivity can be signiﬁcantly improved by changing the protecting group from TBS to the TIPS (Table 7.4, entries 5–7). The intrinsic selectivity has been determined using the achiral catalyst 72, and small preference for the internal syn diastereomer was observed (Table 7.4, entry 7). The internal diastereoselectivity is also largely determined by 7.5.2.2

7.5 Diastereoselective Additions of Chiral Enoxytrichlorosilanes

catalyst conﬁguration in the addition of Z enolates. Reactions with (R,R)-45 corresponds to matched cases wherein the sense of internal and external diastereoselection is the same. Thus, higher internal selectivity is obtained with (R,R)-45 than with (S,S)-45. In additions of (E)-35, internal selectivity is modest, and there is no dependence on catalyst conﬁguration. Additions of (Z)-116 to a variety of aldehydes furnish the syn (relative) diastereomers in good yields with good to excellent selectivity (Scheme 7.50). The structure of the aldehyde makes an important contribution to the relative diastereoselectivity. Bulky aldehydes such as 1-naphthaldehyde and tiglic aldehyde result in signiﬁcantly lower diastereoselectivity. The internal selectivity also depends on the aldehyde structure. Selectivity is signiﬁcantly higher for aromatic aldehydes than for oleﬁnic aldehydes. The internal diastereoselection is always determined by catalyst conﬁguration. When (R,R)45 is used high syn (internal) selectivity can be achieved, and anti (internal) diastereomers can be obtained by use of (S,S)-45, albeit with attenuated selectivity. The dramatic diﬀerence between the behavior of acyclic Z and E enolates in these aldol additions has already been discussed above (Figure 7.12, Section 7.4). Here again, addition of (Z)-35 results in good syn relative selectivity and addition of (E)-35 is only slightly anti relative selective. In the addition of (Z)-35 the diastereoselectivity observed can be better rationalized by the chair-like transition structure xxii (Figure 7.19). The transition structure xxii is consistent with the small internal diastereoselection exerted by the stereogenic center on the enolate. The conformation of the stereogenic center in xxii minimizes steric interaction between the substituents on the non-participating side of the enolate and the bulky ligands on the hypercoordinate silicon. This transition state model leads to the observed (syn,syn)-117. The poor selectivity observed for addition of (E)-35 is explained by the competitive transition structure models chair-xxiii and boat-xxiv. In chair-xxiii A 1; 3 strain between the equatorial methyl group and the nonparticipating substituent of the enolate is minimized [2c]; the disposition of the a-methyl and CH2 OTIPS groups toward the bulky silicon center can, however, cause severe steric congestion. This interaction can be signiﬁcant enough to make the boat-xxiv transition structure more favorable. The boatxxiv is easily accessed simply by placing the silicon group in the least crowded quadrant. The anti coordination of silicon to the aldehyde places the phenyl group of benzaldehyde in the pseudo-axial position, leading to the syn (relative) diastereomer. 7.5.3

Addition of Enoxytrichlorosilanes with a b-Stereogenic Center

Thus far the eﬀect of an a-stereogenic center on the stereochemical course of aldol additions of trichlorosilyl enolates has been described. Diastereo-

283

7 Catalytic Enantioselective Aldol Additions with Chiral Lewis Bases

284

TIPSO

RCHO 45 (10 mol %)

OSiCl3 Me + RCHO

TIPSO

OH R + diastereomers

CH2Cl2, −78oC

Me (Z )-116

O

Me Me syn,syn

Using (R,R)-45 TIPSO

O

OH

TIPSO

O

OH

TIPSO

OH

O

Ph Me

Ph

Me

Me

118 (84%) rel. syn/anti, 53/1 int. syn/anti, 24/1

Me

119 (71%) rel. syn/anti, 14/1 int. syn/anti, 89/1

TIPSO

O

120 (88%) rel. syn/anti, 9/1 int. syn/anti, 14/1

TIPSO

OH

O

OH Me

Me Me

Me

Me

Me

Me

121 (90%) rel. syn/anti, >50/1 int. syn/anti, 15/1

Me

Me

122 (85%) rel. syn/anti, 13/1 int. syn/anti, 13/1

Using (S,S)-45 TIPSO

O

OH

TIPSO

O

OH

TIPSO

O

OH

Ph Me

Ph

Me

Me

118 (82%) rel. syn/anti, 32/1 int. syn/anti, 1/8 TIPSO

Me

Me

119 (79%) rel. syn/anti, 14/1 int. syn/anti, 1/17 O

120 (75%) rel. syn/anti, 15/1 int. syn/anti, 1/6

TIPSO

OH

O

OH Me

Me Me

Me

121 (85%) rel. syn/anti, >50/1 int. syn/anti, 1/5

Me

Me

Me

Me

122 (80%) rel. syn/anti, 19/1 int. syn/anti, 1/5

Scheme 7.50

Aldol additions of (Z)-116 to diﬀerent aldehydes catalyzed by 45.

selectivity clearly depends on the nature of the a-substituent. The a-oxygen substituent of lactate-derived enolates has a strong eﬀect on the diastereoselectivity of the catalyzed aldol addition whereas the a-hydroxymethyl stereogenic center of hydroxybutyrate-derived enolates plays a minor role only in the diastereoselection, and catalyst conﬁguration primarily determines the stereochemical course of the aldol addition.

7.5 Diastereoselective Additions of Chiral Enoxytrichlorosilanes

L L L H Si O L O

H

(Z )-35 + PhCHO

(R2N)3PO

Me TBSOCH2 Ph H Me

Cl (syn,syn)-117

L = Cl, (R2N)3PO

xxii

OTBS H

H (E )-35 + PhCHO

(R2N)3PO

Cl

Me SiLn O O

Ph Me

(anti,anti )-117

H chair-xxiii

Me Ph H

OTBS Cl (syn,anti )-117

O O

H Me H

Si Ln

boat-xxiv Fig. 7.19

Proposed transition structures for addition of (Z)- and (E)-35.

The eﬀect of a remote stereogenic center on diastereoselection in aldol additions is also worth investigation. In the aldol addition of boron enolates it has been demonstrated that a b-oxygen stereogenic center can strongly inﬂuence the stereochemical course of the reaction [87]. This class of enolate is also important because these aldol products have a 1,3,5-oxygenated carbon chain, a common motif in a variety of natural products [2c]. The aldol reactions of 123 under phosphoramide catalysis are summarized in Table 7.5 [88]. The intrinsic selectivity determined using 72 is almost negligible, indicating that the b-stereogenic center does not exert signiﬁcant stereoinduction during addition of this enolate. Interestingly, use of chiral phosphoramides aﬀords only marginal improvement in diastereoselectivity. Additions of ethyl ketone-derived enolates (Z)- and (E)-36 are also catalyzed by 45 (Table 7.6). Good relative syn diastereoselectivity is observed for addition of (Z)-125. As previously observed for addition of (Z)-35, changing from the TBS protecting group to TIPS has a beneﬁcial eﬀect on the dia-

285

286

7 Catalytic Enantioselective Aldol Additions with Chiral Lewis Bases Tab. 7.5

Catalyzed aldol addition of 123 to benzaldehyde. TBSO

OSiCl3 + PhCHO

Me

cat. (10 mol %) CH2Cl2, −78 ˚C

(S)-123

TBSO

O

OH

Me

Ph 124

Entry

Catalyst

Yield, %

syn/anti

1 2 3

(R,R)-45 (S,S)-45 72

72 75 55

1/2.5 1.3/1 1/1.4

stereoselectivity and additions of (Z)-36 result in signiﬁcantly higher relative diastereoselectivity. Addition of (E)-36 is again unselective and, surprisingly, syn (relative) selective. The intrinsic internal diastereoselection is again almost negligible (Table 7.6, entry 6). Thus, internal diastereoselectivity is primarily controlled by catalyst conﬁguration. The match/mismatch eﬀect in these aldol additions is not signiﬁcant. This observation is in contrast to the strong 1,5-anti stereoinduction observed for boron enolate aldol additions [87]. The stereochemical model in this reaction should be analogous to that proposed for achiral enolate additions (Figure 7.11, Section 7.4). Aldol additions of (Z)-36 to a variety of aldehydes provide syn (relative) diastereomers selectively (Scheme 7.51). Excellent syn (relative) selectivity is obtained in the reaction with cinnamaldehyde. The internal selectivity is controlled by catalyst conﬁguration, enabling selective preparation of both syn (relative) diastereomers. Tab. 7.6

Catalyzed aldol additions of (Z)-125, (Z)-36, and (E)-36 to benzaldehyde. relative RO

OSiCl3 Me + PhCHO

Me 125: R = TBS 36: R = TIPS

cat. (10 mol %) CH2Cl2, −78 ˚C

RO

O

OH

Me

Ph Me 126: R = TBS 127: R = TIPS

internal

Entry

Enolate

Z/E

Catalyst

Yield, %

Relative dr (syn/anti)

Internal dr a (syn/anti)

1 2 3 4 5 6

(Z)-125 (Z)-125 (Z)-36 (Z)-36 (E)-36 (Z)-36

12/1 12/1 16/1 16/1 1/15 30/1

(R,R)-45 (S,S)-45 (R,R)-45 (S,S)-45 (S,S)-45 72

59 60 84 86 80 83

6/1 12/1 30/1 26/1 3/1 29/1

14/1 1/14 16/1 1/10 1/1 1.4/1

a Ratio

of major relative diastereomer.

7.5 Diastereoselective Additions of Chiral Enoxytrichlorosilanes

287

relative TIPSO

OSiCl3 Me + RCHO

TIPSO

cat. (10 mol %)

Me 36 (Z/E, 32/1)

O

OH

Me

CH2Cl2, −78 ˚C

R Me

internal

Using (R,R)-45 TIPSO

O

OH

Me

Ph

TIPSO

OH

O

Me

TIPSO Me

Me

O

Me

Ph

Me

127 (84%) rel. syn/anti, 30/1 int. syn/anti, 1/16

OH

Me 129 (80%) rel. syn/anti, >50/1 int. syn/anti, 1/9.6

128 (79%) rel. syn/anti, 28/1 int. syn/anti, 1/7

Using (S,S)-45 TIPSO

O

OH

Me

Ph Me

127 (86%) rel. syn/anti, 26/1 int. syn/anti, 10/1

TIPSO

O

OH

Me

TIPSO Me

Me 128 (83%) rel. syn/anti, 37/1 int. syn/anti, 6/1

O

OH

Me

Ph Me 129 (75%) rel. syn/anti, >50/1 int. syn/anti, 7.5/1

Scheme 7.51

Catalyzed aldol addition of (Z)-36 to diﬀerent aldehydes.

These aldol additions using three diﬀerent classes of chiral trichlorosilyl enolates are interesting examples of double stereodiﬀerentiating aldol additions. In the matched cases, high diastereoselectivity is obtained with the appropriate chiral phosphoramide catalyst. In aldol additions of lactate-derived enolates strong internal stereoinduction dominates the stereochemical course of the reaction. For the other two types of enolate, diastereoselection is primarily determined by catalyst conﬁguration (external diastereoselection), enabling access to two diastereomers. The eﬀect of the a and b stereogenic centers described above would be very important in the construction of a stereodyad or triad in a predictable manner. The compatibility of common protecting groups with trichlorosilyl reagents is clearly established, and the in-situ generation of trichlorosilyl enolate from the corresponding TMS enol ether further enhances the synthetic utility of this process. In the addition of the substituted enolates, the syn (relative) diastereomers can be obtained with high selectivity starting with Z enolates, although, because E enolates do not undergo selective aldol addition, the corresponding anti (relative) diastereomers cannot be accessed by these methods.

288

7 Catalytic Enantioselective Aldol Additions with Chiral Lewis Bases

7.6

Aldol Additions of Aldehyde-derived Enoxytrichlorosilanes

In the previous section the aldol addition of ketone-derived enolates was discussed and illustrated examples were used to document the synthetic utility of these reactions. This section deals with aldol additions of trichlorosilyl enolates derived from aldehydes. This type of aldol addition would be a particularly useful and practical approach to the construction of poly-propionate-derived natural products. The stereoselective aldol addition of an aldehyde-derived enolate and an aldehyde remains a challenging topic [3a]. The diﬃculties associated with this process arise from complications inherent in the self-aldol reaction of aldehydes:

. polyaldolization resulting from multiple additions to the aldol products; . Tischenko-type processes among the products; and . oligomerization of the aldol products. Only recently several approaches have been developed to address these problems [89]. Denmark et al. have achieved the ﬁrst catalytic, enantioselective crossed-aldol reaction of aldehydes utilizing the Lewis-base-catalyzed aldol addition of trichlorosilyl enolates [49]. More recent developments have been made in direct, catalytic crossed-aldol reactions of aldehydes using proline, although an excess of one component is needed [90]. In the Lewis base catalysis approach, the immediate aldol adduct obtained by addition of an aldehyde-derived trichlorosilyl enolate is protected as its a-chlorosilyl ether, which is less prone to further additions. The concept is illustrated in the addition of heptanal-derived enolate (Z)-37 to benzaldehyde in the presence of phosphoramide (S,S)-45 (Scheme 7.52). Low-temperature NMR analysis of this adduct revealed it exists in the form of the a-chlorosilyl ether 130. Because the aldolate occurs as a chelate complex, further reactions leading to a variety of side products are prevented. The chlorosilyl ether intermediate can be hydrolyzed to obtain either its aldehyde or acetal (Scheme 7.53). When 130 is quenched in a mixture of aqueous THF and triethylamine (basic conditions), the corresponding aldehyde is obtained in excellent yield. When dry methanol is used for quench-

(S,S)-45 (10 mol %)

OSiCl3 n-Pent

H

+ PhCHO CDCl3, −60oC

Cl H H Si O Ph Cl O Cl n-Pent

(Z )-37 Scheme 7.52

Catalyzed aldol addition of (Z)-37 to benzaldehyde.

130

7.6 Aldol Additions of Aldehyde-derived Enoxytrichlorosilanes

THF/H2O/Et3N (9/0.5/0.5)

OH

O

Ph (S,S)-45 (10 mol %)

OSiCl3 n-Pent

H

+ PhCHO

289

H n-Pent

Cl

Cl Si O O

131 (95%)

H Cl

CHCl3, −65 oC Ph n -Pent

(Z )-37

130 OH

OMe

Ph

MeOH

OMe n-Pent

132 (89%) Scheme 7.53

Quenching of the chlorosilyl ether intermediate 130.

ing, the chlorosilyl ether is converted to its dimethyl acetal 132 which can be isolated in excellent yield. High diastereoselectivity is observed in additions of geometrically deﬁned enolates (Z)- and (E)-37 (Scheme 7.54). The diastereomeric composition of the aldol product strictly mirrors the E/Z ratio of the enolate, suggesting the reaction proceeds exclusively via a closed transition structure. From the correlation of Z enolate with syn diastereomer and E enolate with anti diastereomer, a chair-like transition structure can be inferred. Although nearly perfect correlation is achieved between the E/Z ratio of the enolate and the syn/anti ratio of the aldol product, the enantiose-

OSiCl3 n-Pent

H

(S,S)-45 (10 mol %)

HCl

CHCl3, −65 oC

MeOH

+ PhCHO

(S,S)-45 (10 mol %)

H OSiCl3

OMe

Ph

OMe n-Pent

132 (89%) syn/anti, 99/1 er (syn), 1.7/1

37 (Z/E, 99/1)

n-Pent

OH

+ PhCHO CHCl3, −65 oC

OH

HCl MeOH

OMe

Ph

OMe n-Pent

37 (E/Z, 24/1) 132 (90%) syn/anti, 1/24 er (anti), 1.7/1 Scheme 7.54

Dependence of enolate geometry of aldol additions of 37.

290

7 Catalytic Enantioselective Aldol Additions with Chiral Lewis Bases

OSiCl3 n-Pent

H

(R,R)-48 (10 mol %)

HCl

CHCl3, −65 oC

MeOH

+ PhCHO

H OSiCl3

OMe

Ph

OMe n-Pent

(2S,3S)-132 (92%) syn/anti, 99/1 er (syn), 19/1

37 (Z/E, 99/1)

n-Pent

OH

(R,R)-48 (10 mol %)

HCl

CHCl3, − 65 oC

MeOH

+ PhCHO

OH

OMe

Ph

OMe n-Pent

37 (E/Z, 32/1) (2R,3S)-132 (91%) syn/anti, 1/32 er (anti), 10/1

(R,R)-48 =

Me N O P N N Me Me

CH2 2

Scheme 7.55

Aldol addition of 37 to benzaldehyde catalyzed by (R,R)-48.

lectivity obtained by the use of 45 is rather poor. Despite an extensive catalyst survey, no signiﬁcant improvement was achieved with a wide variety of monophosphoramides. A signiﬁcant improvement in the enantioselectivity is achieved by using dimeric phosphoramides. Among these catalysts the binaphthyldiamine-derived dimer 48 aﬀords the highest enantioselectivity for this transformation (Scheme 7.55). Additions of (Z)-12 to a variety of aldehydes in the presence of only 5 mol% (R,R)-48 provide the corresponding aldol products in excellent yield and with exclusive syn selectivity (Scheme 7.56). Enantioselectivity varies signiﬁcantly, good selectivity being observed for aromatic aldehydes only. There is no obvious correlation between aldehyde structure and enantioselectivity. It seems that the asymmetric induction provided by the catalyst (R,R)-48 is most eﬀectively transferred for benzaldehyde-like acceptors, and any structural change leads to erosion of enantioselectivity. Aliphatic aldehydes can also be used for aldol addition of (Z)-12. Additions to aliphatic aldehydes are, however, very slow at 65 C, so increased temperature (20 C) and longer reaction times are required for complete reaction. The absolute conﬁguration of syn-133 was assigned by conversion to the corresponding methyl ester, which was unambiguously assigned as the (2S,3S) isomer [91]. The corresponding E enolate also reacts with a variety of aldehydes to give anti b-hydroxy acetals (Scheme 7.57). This high relative diastereoselectivity

7.6 Aldol Additions of Aldehyde-derived Enoxytrichlorosilanes

(R,R)-48 (10 mol %)

OSiCl3 Me

H

OH

HCl

+ RCHO

R

MeOH

CHCl3, −65 oC 6h

OMe OMe

Me

(Z )-12 (Z/E, 99/1) OH

OMe

Ph

OH

OMe

OMe

Ph

Me

OMe Me

133 (95%) syn/anti, 49/1 er (syn), 9.5/1

134 (86%) syn/anti, 99/1 er (syn), 2.4/1

OH

OH

OMe OMe Ph

Ph

Me 136 (98%) syn/anti, 49/1 er (syn), 1.2/1

OMe

OMe Me (−20 oC, 20 h) 137 (47%) syn/anti, 19/1 er (syn), 1.2/1

OH

OMe

Ph

OMe Me

Me

135 (91%) syn/anti, 32/1 er (syn), 5.3/1 OH

OMe OMe

Me (−20 oC, 20 h) 138 (42%) syn/anti, 32/1 er (syn), 2.6/1

Scheme 7.56

Catalyzed addition of (Z)-12 to diﬀerent aldehydes.

contrast with the poor diastereoselectivity obtained from addition of E enolates derived from acyclic ketones. These trends in enantioselectivity are also diﬀerent from those observed for additions of (Z)-12. The highest enantioselectivity is obtained for addition to a-methylcinnamaldehyde whereas addition to benzaldehyde provides only modest enantioselectivity. Markedly higher yields are obtained for addition of (E)-12 to aliphatic aldehydes than for addition of (Z)-12, indicating that (E)-12 is more reactive than (Z)-12. The absolute conﬁguration of anti-133 was established by chemically by correlation with the TBS-protected aldehyde, which has been unambiguously assigned as the (2R,3S) isomer [92]. On the basis of the individual eﬀects of mono-substitution at the Z and E position it was envisaged that higher selectivity might be achieved by employing a disubstituted enolate. Aldol addition to benzaldehyde of adisubstituted trichlorosilyl enolate derived from isobutyraldehyde has been examined using 10 mol% of (R,R)-48 (Scheme 7.58) [93]. The enantiomer ratio of 140 is surprisingly low when compared with the results obtained from addition of (E)- and (Z)-12. For electron-rich aromatic aldehydes and electron-deﬁcient aromatic aldehydes, however, enantioselectivity improves signiﬁcantly (Scheme 7.59). Important mechanistic insights have been obtained from these observations [93]. It has been suggested that the divergence of enantioselectivity is because of the two diﬀerent factors determining enantioselectivity for electron-rich and electron-

291

292

7 Catalytic Enantioselective Aldol Additions with Chiral Lewis Bases

(R,R)-48 (10 mol %)

H Me

OSiCl3

OH

HCl

+ RCHO

R

MeOH

CHCl3, −65 oC 6h

OMe OMe

Me

(E )-12 (E/Z, 99/1) OH

OH

OMe

Ph

OMe

OMe

Ph

OMe

133 (97%) syn/anti, 1/99 er (anti), 3.9/1

134 (88%) syn/anti, 1/99 er (anti), 1.7/1

OH

OH

OMe OMe

Me 136 (99%) syn/anti, 1/49 er (anti), 7.3/1

Ph

OMe

Ph

Me

Me

Ph

OH

OMe Me

Me

135 (89%) syn/anti, 1/99 er (anti), 19/1 OH

OMe

OMe

OMe Me (−20 oC, 20 h)

OMe Me (−20 oC, 20 h)

137 (79%) syn/anti, 1/99 er (anti), 4.9/1

138 (69%) syn/anti, 1/99 er (anti), 1.5/1

Scheme 7.57

Catalyzed addition of (E)-12 to diﬀerent aldehydes.

deﬁcient aldehydes. For electron-poor aldehydes the event determining the stereochemistry is most probably the aldehyde binding process. For electron rich aldehydes, on the other hand, the stereocontrolling step is the aldolization. In both mechanistic extremes, high selectivity can be achieved. The diﬀerent electronic nature of the aldehydes not only aﬀects enantioselectivity but also reactivity. Additions of 139 to a variety of aldehydes result in modest to good enantioselectivity, albeit with no distinct trend (Scheme 7.60). Additions to aliphatic aldehydes also proceed with good yields and moderate selectivity although elevated temperatures and long reaction times are required. Problems associated with crossed-aldol reactions are successfully overcome by the Lewis-base-catalysis approach, and catalytic, enantioselective (R,R)-48 (10 mol %)

OSiCl3 Me

H Me 139

OH

OMe

HCl OMe

+ PhCHO CHCl3, −65 oC

Scheme 7.58

Catalyzed aldol addition of 139 to benzaldehyde.

MeOH

Me Me 140 (86%) er, 2.3/1

7.6 Aldol Additions of Aldehyde-derived Enoxytrichlorosilanes

(R,R)-48 (10 mol %)

HCl

CHCl3, − 65 oC

MeOH

OSiCl3 Me

+ RCHO

H Me 139 OH

OH

OMe

OMe Me Me

MeO OMe

OH

142 (80%) er, 7.0/1 OMe

OH

OMe

OMe

OMe Cl

OMe

MeO

Me Me

OMe

Me Me

OMe Me Me

141 (92%) er, 3.1/1

140 (86%) er, 2.3/1 OH

R

OMe MeO

OMe

OH

OMe

OMe Me Me

OH

F3C

143 (85%) er, 8.1/1

OMe

Me Me

Me Me

O2N

144 (86%) er, 9.0/1

145 (89%) er, 10/1

Scheme 7.59

Catalyzed addition of 139 to substituted benzaldehydes.

(R,R)-48 (10 mol %)

HCl

CHCl3, −65 oC

MeOH

OSiCl3 Me

+ RCHO

H Me 139

OH

OMe OMe

OMe Me Me

OH

OMe OMe

OMe

148 (85%) er, 4.4/1

OMe

147 (90%) er, 2.1/1

OMe

Me Me

OMe Me Me

Ph

146 (90%) er, 4.9/1

n-Bu

OMe

R

OH

Me Me

OH

OH

Ph

OH

OMe

Me

OMe

Me Me

Me Me

149 (82%) er, 1.3/1

150 (80%) er, 10/1

Scheme 7.60

Catalyzed aldol addition of 139 to a variety of aldehydes.

293

294

7 Catalytic Enantioselective Aldol Additions with Chiral Lewis Bases

crossed-aldol reactions of aldehydes have been achieved by use of dimeric phosphoramide 48. High diastereoselectivity can be achieved under the conditions described above by using geometrically deﬁned trichlorosilyl enolates. When the chiral bisphosphoramide 48 is used a variety of crossedaldol products are obtained with moderate to good enantioselectivity. The immediate aldol adduct can be recovered as the aldehyde or the acetal, depending on the quenching conditions. Thus the method discussed above will be extremely useful in enantioselective construction of a polypropionate chain.

7.7

Aldol Addition of Trichlorosilyl Ketene Acetal to Aldehydes and Ketones

For trichlorosilyl ketene acetals, enhanced nucleophilicity is expected, because of the additional oxygen substituent compared with ketone-derived enolates [32]. Trichlorosilyl ketene acetal 10 is, indeed, an extremely reactive nucleophile and reactions with a variety of aldehydes occur even at 80 C. Aromatic, conjugated and aliphatic aldehydes all aﬀord excellent yields of the aldol products within 30 min (Scheme 7.61). The compatibility with enolizable and sterically demanding aldehydes attests to the generality of this aldol addition. Several chiral phosphoramides from diﬀerent structural families have been examined for their capacity to induce enantioselectivity (Scheme 7.62). With 10 mol% of these phosphoramides aldol additions of 10 proceed rapidly at 78 C to give good to excellent yields of the aldol products. Unfortunately, the enantioselectivity obtained in these reactions is poor. Modiﬁcation of the reaction conditions did not signiﬁcantly improve enan-

OSiCl3

O

CH2Cl2

+

OMe

R

H

0

oC,

30 min

HO

O

R

OMe

10 HO Ph

HO

O OMe

151 (98%)

HO OMe

O

HO OMe

OMe 154 (96%)

OMe 153 (89%)

O

Ph

O

Ph

152 (94%) HO

HO

O

Ph

O

Me

OMe

Me Me 155 (96%)

Scheme 7.61

Uncatalyzed addition of 10 to a variety of aldehydes.

156 (99%)

7.7 Aldol Addition of Trichlorosilyl Ketene Acetal to Aldehydes and Ketones

OSiCl3

O

OMe 10

R

H

Me N O P N N Me

Ph Ph

(R,R)-64 R = Ph (88%); er 1/1.5 R = t-Bu (76%); er 1/1.8

HO

cat. (10 mol %)

+

CH2Cl2, −78 oC 30 min - 3 h

O

R

OMe

Me N O P N N Me

Me N O P N N Me

(S,S)-45 R = Ph (87%); er 2.0/1 R = t-Bu (78%); er 2.3/1

(R)-65 R = Ph (91%); er 1.6/1 R = t-Bu (91%); er 2.9/1

Scheme 7.62

Catalyzed additions of 10 to benzaldehyde and pivaldehyde.

tioselectivity. The poor enantioselectivity observed can be explained by the competitive, rapid background reaction between 10 and aldehydes. Although the addition of 10 to aldehydes gives modest enantioselectivity only, the extraordinary reactivity of 10 enables aldol addition to ketones. In the absence of Lewis basic promoters, ketene acetal 10 reacts with acetophenone sluggishly at 0 C; this background reaction can, however, be completely suppressed by reducing the temperature to 50 C. With 10 mol% HMPA the reaction gives almost quantitative yields of 157 [55]. In a survey of a variety of Lewis bases amine-N-oxides were found to be superior in promoting addition of 10 to acetophenone (Table 7.7) [55]. A variety of amine-N-oxides can promote this reaction and, among all the Lewis bases

Tab. 7.7

Survey of N-oxide promoters for addition of 10 to acetophenone. OSiCl3 OMe

promoter (100 mol %)

O + Ph

Me

CH2Cl2

HO

O

Ph Me

OMe 157

10

Entry

Promoter

Temp, ˚C

Time, min

Conv., %

1 2 3 4 5 6 7 8

none Me3 NO Me3 NO NMOa quinuclidine N-oxide pyridine N-oxide pyridine N-oxide pyridine N-oxideb

50 78 20 78 78 78 50 rt

240 240 50 70 70 70 50 120

0 10 76 25 35 37 97 100

a N-methymorpholine-N-oxide. b 10

mol% of promoter was used.

295

7 Catalytic Enantioselective Aldol Additions with Chiral Lewis Bases

296

10 mol % pyridine-N -oxide

O

OSiCl3

+

R1

CH2Cl2, rt

R2

R1

OMe

OH

O

R2

OMe

10 OH Ph

OH

O OMe

Me

Ph

157 (94%) OH

O

Me

OH

O OMe

Me

Ph

158 (92%)

O

OH OMe

161 (91%)

t-Bu

Me

O

OH OMe Ph

Me 159 (92%) OH

O OMe

162 (93%)

O

O OMe

Me 160 (94%) OH

O

OMe 163 (94%)

OMe 164 (45%)

Scheme 7.63

Addition of 10 to a variety of ketones, catalyzed by pyridine-N-oxide.

surveyed, pyridine-N-oxide resulted in the highest conversion in the aldol addition. With a catalytic amount of pyridine-N-oxide, complete conversion can be achieved within 2 h at room temperature. With catalytic amounts of pyridine-N-oxide addition of 10 to a variety of ketones has been achieved (Scheme 7.63). Excellent yields are obtained for a wide range of substrates, including highly enolizable ketones. Aldol addition to 2-tetralone is the only instance in which the reaction does not produce the expected aldol product in high yield. To provide enantiomerically enriched aldol products the use of structurally diverse chiral N-oxides has been studied (Scheme 7.64). In this process enantioselection is clearly enhanced by use of dimeric N-oxides with 6,6 0 stereogenic centers. Among these, the highest enantioselectivity is obtained by use of bis-pyridine-derived P-(R,R)-46. Interestingly, the M-(R,R)-46 is equally capable of catalyzing the aldol addition, although this reaction affords the enantiomeric product with slightly attenuated enantioselectivity. The generality of this catalyst system has been demonstrated in additions of 10 to a variety of ketones (Scheme 7.65). The enantiomeric ratio of the product ranged from modest to good, depending on the ketone structure. The crucial factor in obtaining high selectivity seems to be the size diﬀerential between the two substituents on the ketones. Catalytic, asymmetric aldol additions to ketones have been achieved by means of the extraordinary reactivity of trichlorosilyl ketene acetal combined with Lewis-base-catalysis. The axially chiral bipyridine-N-oxide bearing stereogenic centers at the 6,6 0 -positions has excellent catalytic properties and results in synthetically useful enantioselectivity. This process provides access to enantiomerically enriched tertiary alcohols catalytically. Enantio-

7.7 Aldol Addition of Trichlorosilyl Ketene Acetal to Aldehydes and Ketones OSiCl3

O

OMe

+

Ph

−20 oC, 12 h

N

N

Me O Me O

O O

N

(R)-165 er, 2.6/1

(S)-166 er, 1.7/1

t-Bu t-Bu

(R,R)-167 (92%) er, 3.4/1

N N O O On-Bu n-BuO

N N O O On-Bu n-BuO

MeMe

N N O O On-Bu n-BuO

CMe2Ph t-Bu

N N O O On-Bu n-BuO

t-Bu t-Bu

P-(R,R)-46 (94%) er, 12/1

P-(R,R)-169 (90%) er, 9.0/1

t-Bu

(R,R)-168 (94%) er, 4.6/1

MeMe

MeMe

PhMe2C

OMe 157

N N O O OMe MeO

t-Bu

O

Ph Me

CH2Cl2

Me

10

N

HO

cat. (10 mol %)

297

t-Bu

M-(R,R)-46 (89%) er, 1/2.5

Scheme 7.64

Catalyst survey for addition of 10 to acetophenone.

OSiCl3

P-(R,R)-46 (10 mol %)

O +

10 OH Ph

Me

OH

O

157 (96%) er, 10/1 OH

O

161 (87%) er, 2.9/1

OH OMe

170 (89%) er, 13/1

O

Me

O

Ph

OMe

OH OMe

Ph

R1

CH2Cl2, −20 oC 12 - 32 h

R2

R1

OMe

OH

160 (97%) er, 2.1/1

OMe

O

OH

Me

OH OMe

OMe 172 (90%) er, 9.1/1

O

Me 173 (91%) er, 1.9/1

Scheme 7.65

Addition of 10 to diﬀerent ketones catalyzed by P-(R,R)-46.

O

OMe

171 (89%) er, 3.5/1

O

Me

R2

O

OH OMe

t-Bu

Me

O OMe

162 (87%) er, 2.5/1

298

7 Catalytic Enantioselective Aldol Additions with Chiral Lewis Bases

selectivity, however, is not consistently high for diﬀerent substrates, so catalyst optimization is still needed if this reaction is to be truly practical.

7.8

Lewis Base Activation of Lewis Acids – Aldol Additions of Silyl Enol Ethers to Aldehydes

The aldol reactions of enoxytrichlorosilanes described in preceding sections all involve the use of a chiral Lewis base (phosphoramide or N-oxide) to activate the nucleophile and provide the chiral environment for CaC bond-formation [53, 55, 94]. These reactions all proceed by a common mechanistic pathway that involves a cationic, hypercoordinate silicon as an organizational center for the reactants and catalyst (Section 7.9). The ability of certain Lewis bases to induce the ionization of silicon Lewis acids has intriguing potential for a new concept in Lewis-acid catalysis of organic reactions. The possibility of activating a weak Lewis acid, for example silicon tetrachloride, with a chiral Lewis base and using the resulting complex as a chiral Lewis acid for a variety of reactions has recently been demonstrated [54b]. The Lewis acidity of SiCl 4 is relatively weak compared with typical Lewis acids such as TiCl 4 or BF3 [8]; in the presence of several diﬀerent Lewis bases, however, a highly Lewis acidic silyl cation is produced (Scheme 7.66) [95]. Formation of a cationic silicon complex has been experimentally demonstrated by Bassindale and coworkers in heterolysis of halosilanes with Lewis bases [95a]. For example, 1 H, 13 C, and 29Si NMR spectroscopic evidence suggested the formation of a cationic silicon complex from silyl halides and triﬂate in DMF. Although enhancement of Lewis acidity by Lewis bases is counter-intuitive, it can be explained by a set of empirical bond-length and charge-variation rules formulated by Gutmann [96]. Activation of silicon by ionization of a ligand has been proposed in several other systems [97]. The concept of Lewis base activation leads to an ideal opportunity for ligand-accelerated catalysis. Because the Lewis acid is active only when coordinated to the Lewis base, a stoichiometric amount of silicon tetrachloride can be used to assist rate and turnover [19]. The use of the [Lewis base–SiCl3 ]þ complex as a chiral Lewis acid was ﬁrst demonstrated in the opening of meso epoxides to obtain enantioenriched chlorohydrins [98]. A more relevant application of this concept has

SiCl4

+

O Me2N P NMe2 NMe2

Cl

OP(NMe2)3 Cl Si Cl OP(NMe2)3

Scheme 7.66

Formation of HMPA–trichlorosilyl cation complex.

Cl

7.8 Lewis Base Activation of Lewis Acids – Aldol Additions of Silyl Enol Ethers to Aldehydes

(R,R)-48 (5 mol %) SiCl4 (110 mol %) SnBu3 + PhCHO

CH2Cl2, −78 oC

299

OH Ph 174 (91%) er, 32/1

Scheme 7.67

Allylation of aldehydes using SiCl4 –bisphosphoramide complex.

been illustrated in asymmetric allylation of aldehydes (Scheme 7.67) [54b]. With allyltributyltin as an external nucleophile, the allylation of a variety of aldehydes proceeds in excellent yield; excellent enantioselectivity is obtained with the dimeric phosphoramide (R,R)-48. This reaction system can be applied to aldol additions of silyl ketene acetals (Scheme 7.68) [50]. Although the aldol addition of ketene acetal 175 to aldehydes does not proceed in the absence of the Lewis base catalyst, with 5 mol% (R,R)-48, the aldol products are obtained in excellent yields within 15 min. This behavior is strikingly diﬀerent from the addition of trichlorosilyl ketene acetal, which reacted spontaneously with aldehydes (Section 7.7). Not surprisingly, the enantioselectivity observed in these reactions is signiﬁcantly better than that obtained from the reaction of trichlorosilyl ketene acetal with aldehydes, which suﬀers from competitive background reaction. Excellent enantioselectivity was observed for most of the aldehydes surveyed. Sterically bulky aldehydes seem to aﬀord lower enantioselectivity, (R,R)-48 (5 mol %) SiCl4 (110 mol %)

OTBS OMe 175

OH

+ RCHO

O

OH

CH2Cl2, −78oC

O

OMe

OH OMe

151 (97%) er, 28/1

OH

O

O

R

OMe

OH

O

O

153 (95%) er, 32/1 OH

O

Me 178 (98%) er, 2.7/1

155 (98%) er, 2.7/1

O OMe

OMe

OMe

154 (94%) er, 14/1

Scheme 7.68

Aldol addition of 175 to a variety of aldehydes catalyzed by SiCl4 -(R,R)-48.

O OMe

OMe 177 (94%) er, 14/1

176 (98%) er, 9.0/1 OH

OH

300

7 Catalytic Enantioselective Aldol Additions with Chiral Lewis Bases

and there is no signiﬁcant electronic eﬀect. Additions to aliphatic aldehydes are slow, but good yields of the aldol products can be obtained after 6 h with good enantioselectivity. These observations are remarkable considering that addition of trichlorosilyl nucleophiles to aliphatic aldehydes have been problematic. The absolute conﬁguration of the benzaldehyde aldol product is R, which is consistent with the sense of asymmetric induction observed for the allylation. Aldol additions with substituted ketene acetals introduce the issue of relative diastereoselection. Unlike the reactions with trichlorosilyl nucleophiles which involve closed transition structures, the mechanism of Lewisacid-catalyzed aldol reactions usually involves an open transition structure. Under such conditions control of relative diastereoselection cannot be achieved simply by adjustment of enolate geometry. Interestingly, the reactions of propanoate-derived ketene acetals with benzaldehyde produce the anti aldol products with high diastereoselectivity (Scheme 7.69). A survey of ketene acetal structures indicated that larger ester groups aﬀord higher enantioselectivity. For the t-butyl propanoatederived ketene acetal, enantioselectivity is signiﬁcantly higher than for the addition of other ketene acetals. It is also important to note that the geometry of the ketene acetal does not aﬀect the stereochemical outcome of the reaction. Starting from either the E- or Z-enriched ketene acetal 183, the anti aldol product is obtained exclusively with excellent enantioselectivity. This suggests that these aldol additions do not proceed via a cyclic transition structure as alluded to above – an open transition structure can better account for the stereoconvergent aldol addition. This type of stereoconvergent anti aldol process is rare [16c] and the selectivity observed promises great synthetic utility for this aldol reaction. The broad scope of this reaction has been demonstrated with a variety of aldehydes (Scheme 7.70). Aromatic and conjugated aldehydes aﬀord high yields of the anti aldol product with excellent diastereoselectivity and mod-

(R,R)-48 (1 mol %) SiCl4 (110 mol%)

OTBS Me

OR

OH

+ PhCHO

OH

O

O

OMe Ph

Ph

CH2Cl2, −78 oC OH

OH

O

Ph

OR

Me OH

O

O

OPh Ph

OEt Ph

Ot-Bu

Me

Me

Me

Me

179 (98%) dr, 99/1 er, 6.1/1

180 (78%) dr, 49/1 er, 7.3/1

181 (98%) dr, 16/1 er, 16/1

182 (93%) dr, 99/1 er, >99/1

Scheme 7.69

Aldol addition of a variety of propionate-derived ketene acetals.

7.8 Lewis Base Activation of Lewis Acids – Aldol Additions of Silyl Enol Ethers to Aldehydes

SiCl4 (110 mol%) (R,R)-48 (1 mol %)

OTBS Ot-Bu + RCHO Me 183 OH

OH

Ot-Bu Me OH

O

Ot-Bu

O Ot-Bu

Ot-Bu

Me

Me

Me

182 (93%) dr, 99/1 er, >99/1 OH

O

R

CH2Cl2, −78 oC 3h

O

OH

185 (95%) dr, >99/1 er, >99/1

184 (98%) dr, 24/1 er, 32/1 O

OH

O

Ot-Bu

OH

O

Ot-Bu

Me

Me

Me

Ot-Bu Ph

Me

187 (90%) dr, >99/1 er, 24/1

186 (98%) dr, >99/1 er, >99/1

188 (92%) dr, 24/1 er, 5.3/1

Scheme 7.70

Aldol addition of 183 to a variety of aldehydes.

est to excellent enantioselectivity. Unfortunately, aliphatic aldehydes do not react with this particular ketene acetal under these conditions. The use of less sterically demanding ethyl ketene acetal 189 enables reactions with aliphatic aldehydes, however (Scheme 7.71). For example, combining 189 with hydrocinnamaldehyde aﬀords 190 in 71% yield with good

OTBS OEt

SiCl4 (110 mol %) (R,R)-48 (5 mol %) TBAI (10 mol %)

O +

Ph

H

Me

189

OEt Me

+

CH2Cl2, −78 oC 24 h

SiCl4 (110 mol %) (R,R)-48 (10 mol %) TBAI (10 mol %)

O

OTBS

H CH2Cl2, −40 C 24 h

189

Scheme 7.71

Aldol addition of 189 to aliphatic aldehydes.

o

301

OH

O OEt

Me 190 (71%) dr, 10/1 er, 16/1 OH

O OEt

Me 191 (49%) dr, 8.1/1 er, 2.1/1

302

7 Catalytic Enantioselective Aldol Additions with Chiral Lewis Bases

SiCl4 (150 mol %) (R,R)-48 (5 mol %) TBAOTf (10 mol %) i-Pr2EtN (10 mol %)

OTMS R O

OH Ph n-Bu

Me 51 (97%) er, 49/1

+ PhCHO

O

OH

CH2Cl2, −78 oC 3h O OH

Ph i-Bu 52 (99%) er, >99/1

O

OH

R O

Ph OH Ph Ph

Ph i-Pr 53 (98%) er, 99/1

54 (95%) er, >99/1

O

OH Ph

55 (98%) er, >99/1

Scheme 7.72

Catalyzed aldol additions of TMS enol ethers to benzaldehyde.

selectivity. The yield can be signiﬁcantly improved by addition of tetrabutylammonium iodide and extending the reaction time. In the addition to cyclohexanecarboxaldehyde a higher reaction temperature is also needed to achieve good conversion. The same catalyst system can eﬀect aldol additions of TMS enol ethers of ketones [99]. Additions of methyl ketone-derived TMS enol ethers to aldehydes proceed smoothly to aﬀord the corresponding aldol products in excellent yield and with excellent enantioselectivity (Scheme 7.72). The use of a catalytic amount of a tetraalkylammonium salt is important for achieving complete conversion in these reactions. These studies also revealed the compatibility of the reaction system with a small amount of diisopropylethylamine to scavenge adventitious HCl present in silicon tetrachloride. This modiﬁcation obviates distillation of silicon tetrachloride and makes the process more practical. The generality of this aldol addition is illustrated by the addition of 2hexanone-derived TMS enol ether 192 to a variety of aldehydes (Scheme 7.73). Aromatic aldehydes are among the best substrates in this reaction, aﬀording both high yields and high enantioselectivity. Heteroaromatic aldehydes are also compatible, and good yields and excellent enantioselectivity can be obtained. Sterically encumbered aldehydes react more slowly and less selectively. This eﬀect is most evident in the reaction with a-methylcinnamaldehyde. Unfortunately, aliphatic aldehydes are not reactive under these conditions. A highly regioselective vinylogous aldol reaction has also been achieved by use of the SiCl 4 –bisphosphoramide system [100]. Vinylogous ketene acetals have two nucleophilic sites, i.e. the C(2)- and C(4)-positions, and reaction at the C(2)-position is usually favored, owing to the higher electron density [101]. It has been a challenge to control the reactivity of these two sites to obtain the g-aldol adduct selectively [102]. In the presence of

7.8 Lewis Base Activation of Lewis Acids – Aldol Additions of Silyl Enol Ethers to Aldehydes

SiCl4 (150 mol %) (R,R)-48 (5 mol %) i-Pr2NEt (10 mol %)

OTMS + RCHO

n-Bu

CH2Cl2, −78 oC 3 -24 h

192 O

OH

O

n-Bu

O n-Bu

R

OH

O n-Bu

52 (99%) er, >99/1

193 (88%) er, 19/1

67 (95%) er, 24/1 O

OH S

n-Bu

OH O

n-Bu

O

OH

OH

n-Bu

O

OH

n-Bu Me 59 (54%) er, 3.5/1

194 (79%) er, 99/1

58 (98%) er, >99/1

Scheme 7.73

Aldol addition of 192 to diﬀerent aldehydes.

the bisphosphoramide (R,R)-48, crotonate-derived silyl ketene acetal 195 reacts with benzaldehyde to yield the g-aldol adduct exclusively in good yield with excellent enantioselectivity (Scheme 7.74). The exclusive g-selectivity is attributed to the steric diﬀerentiation between the a- and g-positions (substituted compared with unsubstituted). In this catalyst system the reaction occurs preferentially at less sterically demanding site. Under similar reaction conditions a variety of simple enoate-derived silyl ketene acetals undergo vinylogous aldol additions (Scheme 7.75). In the 2-pentenoate derived silyl ketene acetal a sterically bulky ester group is necessary for high regioselectivity. For example, reaction of the t-butyl esterderived dienol ether yields the g adduct 199 in good yield. The high re-

OH OTBS OEt

195

SiCl4 (110 mol %) (R,R)-48 (5 mol %)

O +

Ph

H

CH2Cl2, −78 oC 3h

Scheme 7.74

Aldol addition of dienol silyl ether 195 to benzaldehyde.

O OEt

196 (89%) γ / α, >99/1 er, 99/1

303

304

7 Catalytic Enantioselective Aldol Additions with Chiral Lewis Bases

R4

OTBS

R5

SiCl4 (110 mol %) (R,R)-48 (5 mol %)

O

OR2

+

R1

H

OH R

O OR2

1

CH2Cl2, −78 oC 3 - 24 h

R3

R4

R5

R3

Me Me O

OH Ph

OH O OMe

O

OH O

Ph

Ph

Me

Ot-Bu

Me

197 (93%) γ / α, >99/1 er, >99/1

OH

O

198 (92%) γ / α, >99/1 er, 6.7/1

O

Ph

OH OEt

199 (92%) γ / α, >99/1 dr (anti/syn), >99/1 er, 17/1

OH

O

Ph

OEt

Ph

O Ot-Bu

Me 200 (84%) γ / α, >99/1 er, 49/1

201 (68%) γ / α, >99/1 er, 19/1

202 (71%) γ / α, 99/1 dr (anti/syn), >99/1 er, 10/1

Scheme 7.75

Catalyzed aldol addition of a variety of dienol silyl ethers.

gioselectivity is complemented by high anti diastereoselectivity and excellent enantioselectivity. In terms of aldehyde scope, good yields and selectivity are obtained with aromatic and oleﬁnic aldehydes, and even aliphatic aldehydes can be employed in this reaction, although longer reaction times are needed. It is important to mention that the aldol products obtained by use of bisphosphoramide (R,R)-48 reveal the commonality of absolute conﬁguration at the hydroxyl center (Figure 7.20). When (R,R)-48 is used, nucleophiles attack the aldehyde Re face in xxv. Although stereochemical models need to be developed, the catalyst has created a highly deﬁned environment for the aldehyde acceptor. The bisphosphoramide–SiCl 4 complex has been successfully used as a chiral Lewis acid in highly eﬃcient catalytic, enantioselective aldol additions of silyl ketene acetals and silyl enol ethers. Compared with aldol additions of trichlorosilyl reagents, these systems are superior in terms of preparation and handling of the nucleophiles. In particular, additions of propanoate-derived ketene acetals are one of the most stereoselective anti aldol additions reported to date.

7.9 Toward a Uniﬁed Mechanistic Scheme

Nucleophilic attack OSiR3

Cl LB

O Si

LB

Cl H

Ph

Cl

X

OH O

R1

Ph R

Cl xxv

LB

X R1

OH

OH

O

O

= (R,R)-48

Ot -Bu

OMe

LB

Me 182

151 OH

O

OH n-Bu

52

Ph

174

Fig. 7.20

Commonality of absolute conﬁguration in a variety of aldol products.

7.9

Toward a Uniﬁed Mechanistic Scheme

Detailed discussion of the extensive kinetic, spectroscopic, and structural investigations that have provided the current mechanistic picture is beyond the scope of this chapter, the primary focus of which is preparative aspects of chiral Lewis base-catalyzed aldol reactions. Instead a summary of the important studies that have led to the current level of understanding will be presented, with the implications for catalyst design and reaction engineering. As originally formulated, the foundation of Lewis base activation of the aldol addition (and subsequent stereoinduction) was based on hypothetical ternary assembly of enolate, aldehyde, and chiral catalyst in a hexacoordinate arrangement about the silicon atom (Figure 7.21) [103]. When catalysis was successfully demonstrated, the hypothesis seemed correct – i.e. that the rate acceleration arose from dual activation of the enol and the aldehyde in close proximity. Two important aspects of the reactions seemed at odds with this picture, however – rate and stereoselectivity were both diﬃcult to rationalize. Although polarization of electron density away from the silicon atom was expected from the Gutmann analysis [96a], there was no basis for estimation of the magnitude of this eﬀect. From analysis of simple molecular models it was, furthermore, not at all clear how single-point binding could provide the highly dissymmetric environment that induced such high facial selectivity.

305

306

7 Catalytic Enantioselective Aldol Additions with Chiral Lewis Bases

R 2N

OSiCl3 R

RCHO

1

R2 (R2N)3P=O

R R

1

H

2

HR

NR2 P NR2

OSiCl3

O

O

R1

O Si Cl Cl O Cl

R R2

Fig. 7.21

Original hypothetical transition structure assembly.

7.9.1

Cationic Silicon Species and the Dual-pathway Hypothesis

The ﬁrst experimental evidence against the simple mechanistic picture in Figure 7.21 was the observation that the diastereoselectivity of aldolization of cyclohexanone-derived enolate 20 with benzaldehyde depended on catalyst loading (Scheme 7.76). The appearance of syn isomers from Econﬁgured enolates (at low catalyst loading) implied intervention of boat-like transition structures. Curiously, although the diastereomeric ratio changes dramatically the enantiomeric ratio of the anti isomer remains unchanged. This suggested that two independent pathways could be operating, one favoring the anti diastereomer (with high facial selectivity) and one favoring the syn isomer (with low facial selectivity). Quantitative support of this hypothesis was obtained from several studies. First, the diastereoselectivity of reactions promoted by the achiral phosphoramide 203 (Figure 7.22) is dramatically dependent on catalyst loading. Figure 7.22 depicts graphically the change in syn/anti ratio from 1.3:1 at 200 mol% loading to 130:1 at 2 mol% loading. The excellent correlation of diastereoselectivity with inverse phosphoramide concentration provided quantitative support for the dual pathway hypothesis, namely, one phosphoramide leads to syn and two phosphoramides lead to anti [66]. The second source of quantitative evidence is the divergent behavior of chiral catalysts (S,S)-45 and (S,S)-81 in studies of the dependence of enantioselectivity on catalyst composition. In contrast with the highly antiselective reactions promoted by (S,S)-45, diphenylphosphoramide catalyst

OSiCl3

1. (S,S)-45 (x mol %) PhCHO, –78 ˚C

O

OH

O

OH

Ph +

Ph

2. sat. aq. NaHCO3 (–)-syn

20

(–)-anti

10 mol % cat (94%) syn/anti, 1/50 (er anti, 21/1) 0.5 mol % cat (53%) syn/anti, 1/5 (er anti, 21/1) Scheme 7.76

Catalyst loading-dependent diastereoselectivity.

7.9 Toward a Uniﬁed Mechanistic Scheme

Ph N O P N N Ph 203

Fig. 7.22

Dependence on loading of selectivity with catalyst 203.

(S,S)-81 provided the syn aldol product in excellent diastereoselectivity (97:1), albeit with modest enantioselectivity (3.25:1 er) (Scheme 7.77) [66, 76]. With enantioselective catalysts now available for both syn and anti pathways, an important link between the steric demand of the catalyst and the resulting diastereoselectivity could be forged. According to the dual pathway hypothesis one (to syn) or two (to anti) catalyst molecules can be present in the stereochemistry determining transition structures, and that these diﬀerent pathways are also stereochemically divergent. This hypothesis could be tested by making use of non-linear eﬀects and asymmetric ampliﬁcation as pioneered by Kagan [104]. The dependence of enantiomeric excess (ee) of the aldol products on the enantiomer composition of the cata-

Ph Ph

Ph N O P N N Ph

OSiCl3

10 mol% (S,S)-81 +

PhCHO

O

OH Ph +

O

OH Ph

CH2Cl2, –78 ˚C 20

(94%) syn/anti 97/1 er (syn) 3.25/1

Scheme 7.77

syn-Selective aldolization catalyzed by (S,S)-81.

(+)-syn

(–)-anti

307

7 Catalytic Enantioselective Aldol Additions with Chiral Lewis Bases

Ph

Ph N O P N N

Ph Ph

Ph (S,S)-81 (

Me N O P N N Me

)

(S,S)-45 (

)

% ee of syn-adduct ( )

Ph

% ee of anti-adduct ( )

308

% ee of catalyst Fig. 7.23

Correlation of product and catalyst ee for (S,S)-45 (f) and (S,S)-81 (n).

lysts is illustrated in Figure 7.23. The linear relationship between catalyst ee and syn-adduct ee with phosphoramide (S,S)-81 (Figure 7.23, n) suggests this product arises from a transition structure involving only one chiral phosphoramide. In contrast, the obvious non-linear relationship between catalyst ee and anti-adduct ee with phosphoramide (S,S)-45 (Figure 7.23, f) suggests the participation of two phosphoramide molecules in the transition structure for aldolization [66]. The most compelling and direct evidence for the operation of dual pathways is provided by establishment of the order of the reaction in catalyst for (S,S)-45 and (S,S)-81 [66b]. The rate and sensitivity of these reactions required use of in-situ monitoring techniques such as ReactIR and rapid injection NMR (RINMR). First-order dependence on (S,S)-81 (R 2 ¼ 1.000) is observed for catalyzed aldol addition of 20 to benzaldehyde with typical catalyst loadings at 35 C. Importantly, the rate of reaction at very low catalyst loadings has pronounced curvature, indicative of a change in mechanism between the promoted and unpromoted pathways. RINMR analysis of the reaction catalyzed by (S,S)-45 at 80 C reveals aldol addition to have second order dependence on phosphoramide (plot of log kobs against log [catalyst]; m ¼ 2.113, R 2 ¼ 0.992). For Lewis-base-catalyzed aldol addition involving trichlorosilyl enolates the rate equations are rate ¼ k[cat][enolate][aldehyde] for catalyst (S,S)-81 and rate ¼ k[cat] 2 [enolate][aldehyde] for catalyst (S,S)-45. The experimentally determined reaction order is consistent with turnover-limiting com-

7.9 Toward a Uniﬁed Mechanistic Scheme

(a) O

1.000 OH (assumed)

Me

(b) O

1.000 OH (assumed)

Me 1.000 1.003 0.997 1.038 52

0.998 1.032 0.990 1.005 52

Fig. 7.24

(a) 13 C KIEs (k12 C /k13 C ) for a reaction taken to 5% conversion using limited aldehyde. (b) 13 C KIEs (k12 C / k13 C ) for a reaction taken to 5% conversion using limited enol ether.

plexation or aldolization, and whereas Arrhenius activation data suggest complexation is rate-limiting they do not discount the possibility that aldolization is the turnover-limiting step. Natural abundance 13 C kinetic isotope eﬀects (KIE) as pioneered by Singleton [105] provide a clear answer. 13 C NMR analysis of aldol product 52 from reaction of enolate 24 and benzaldehyde at 5% conversion aﬀorded excellent results (Figure 7.24). If binding or any other pre-equilibrium process not involving the reactive centers were turnover limiting, no isotope enrichment would be expected in the aldol product. The presence of signiﬁcant (1.038 and 1.032) [106] k12 C / k13C kinetic isotope eﬀects at the enolate carbon and the aldehyde carbonyl carbon clearly show, however, that rehybridization is occurring at both reactive centers in this transformation [107, 108]. These data, with results from the non-linear eﬀect studies above clearly support the conclusion that the aldolization step is both stereochemistry-determining and turnover-limiting. With evidence from a variety of sources that two phosphoramide molecules can be bound to the silicon atom of the enolate in the transition structure, formulating a picture of this assembly could be undertaken. It was reasonable to postulate that the aldehyde is also coordinated to silicon, because the stereochemical consequences of changing enolate geometry are strongly reﬂected in changing diastereoselectivity of the process. Thus, given the likelihood of a closed, silicon-centered transition structure, one of two possibilities arises:

. formation of a heptacoordinate silicon group; or . ionization of a chloride, forming a cationic, hexacoordinate silicon moiety. Support for the intermediacy of cationic silicon species is available from the eﬀects of ionic additives on the rate and selectivity of the reaction [66a]. For reactions with catalyst (S,S)-81 a clear trend emerges (Scheme 7.78). Addition of 1.2 equiv. tetrabutylammonium chloride causes marked deceleration and a diminution in enantioselectivity. Addition of 1.2 equiv. tetra-

309

310

7 Catalytic Enantioselective Aldol Additions with Chiral Lewis Bases

Ph N O P N N

Ph Ph

OSiCl3 +

PhCHO

Ph 10 mol% (S,S)-81

O

OH Ph

CH2Cl2, –78 ˚C 8 min

(+)-syn

44% conv. 8% conv. 92% conv.

er 3.26/1 er 1.70/1 er 3.44/1

20 no additive 1.2 equiv Bu4N+ Cl1.2 equiv Bu4N+ TfOScheme 7.78

Eﬀects of salts on the rate and selectivity of catalyzed aldolization.

butylammonium triﬂate results in moderate rate acceleration and an increase in the enantioselectivity of the overall process. The decrease in rate is consistent with a common-ion eﬀect wherein ionization of chloride precedes the rate-determining step. The corresponding increase in rate and selectivity with tetrabutylammonium triﬂate, which increases the ionic strength of the medium, conﬁrms the notion of ionization. 7.9.2

Uniﬁed Mechanistic Scheme

The available evidence from measurements of kinetics, additive eﬀects, nonlinear studies, and stereochemical information supports a revised picture of the mechanism of phosphoramide-catalyzed aldol additions. As originally proposed, ternary association of enolate, aldehyde, and Lewis base was believed to be suﬃcient for activation and selectivity. Whereas unpromoted additions of trichlorosilyl enolates to aldehydes probably involve simple combination of the two reactants in a trigonal bipyramidal assembly, the catalyzed process is clearly much more complex (Figure 7.25). On binding the Lewis basic phosphoramide the trichlorosilyl enolate undergoes ionization of chloride. Depending on the size and concentration of the phosphoramide two scenarios are possible [109]. With a bulky phosphoramide, or in the limit of insuﬃcient catalyst, aldehyde coordination and aldolization through a boat-like transition structure (with low facial selectivity) provides the syn aldol product (bottom pathway). Alternatively, with smaller phosphoramides or higher catalyst loading, a second molecule of catalyst can be bound to the cationic dichlorosilyl enolate to form an octahedral silicon cation [110]. On binding the aldehyde this intermediate undergoes aldolization through a chair-like transition structure organized around a hexacoordinate silicon atom (top pathway). This process occurs with a high level of facial selectivity, most probably because of the greater stereochemical inﬂuence of two chiral moieties in the assembly.

N

Ph

P N

O

N

O

O

Si Cl Cl

O +

P(NR2)3

coordination of phosphoramide displaces chloride

+

PhCHO

aldolization

Cl–

PhCHO

aldolization

(R2N)3P=O

Cl

–

O P(NR2)3

Cl Cl

O Si

O

P(NR2)3

Uniﬁed mechanistic scheme for phosphoramide-promoted aldolizations.

Fig. 7.25

N

Ph

P

Ph

N

Ph

Ph

one phosphoramide pathway

(R2N)3P=O

OSiCl3

2 (R2N)3P=O

two phosphoramide pathway

Me

N

Me

cationic tbp boat

Cl NR2 H O Si O P O Cl NR2 R2N

H

cationic octahedron chair

H

‡

‡ NR2 NR2 P O L O Si Cl O Cl H R2N

+

+

Ph

syn

O

anti

O

Ph

OSiCl3

Ph

OSiCl3

7.9 Toward a Uniﬁed Mechanistic Scheme 311

312

7 Catalytic Enantioselective Aldol Additions with Chiral Lewis Bases

7.9.3

Structural Insights and Modiﬁcations

The revised mechanistic picture provides a clearer understanding of the remarkable change in diastereoselectivity with catalyst size and loading, and of the origin of rate enhancement. The reasons for the high enantioselectivity observed remain obscure, however. Insights into the stereochemical consequences of catalyst binding are provided by the solution and solid-state structures of chiral phosphoramide complexes of tin(IV) Lewis acids [70, 111]. The uniﬁed mechanistic scheme suggests a preference for 2:1 complexation with (S,S)-45 and a preference for 1:1 complexation with (S,S)-81. Both scenarios are conﬁrmed crystallographically. Single-crystal X-ray structural analysis of the 2:1 complex, ((S,S)-45)2 aSnCl 4 reveals interesting features (Figure 7.26):

. 2:1 complexation is conﬁrmed, . cis geometry of the complex is preferred, . the piperidino nitrogen is planar and oriented orthogonal to the phospholidine ring

N

N

Cl P

Sn

Cl

Cl Cl

O P N

Fig. 7.26

X-ray crystal structure of ((S,S)-45)2 aSnCl4 .

7.9 Toward a Uniﬁed Mechanistic Scheme

Cl Cl

Cl Sn

Cl

O

O P N

N N

Fig. 7.27

X-ray crystal structure of (S,S)-81aSnCl4 aH2 O.

. the phospholidine nitrogen atoms are pyramidal, with the methyl groups away from the stilbene phenyl groups, and . disposed the PaOaSn unit is non-linear such that the tin moiety is oriented over the phospholidine ring. 119

Sn-solution NMR studies and analysis of 1JPaSn coupling constants corroborate the observation of 2:1 complexes (hexacoordinate chemical shift regime) favoring the cis conﬁguration. Crystallization of (S,S)-81 with SnCl 4 aﬀorded a 1:1:1 complex of (S,S)81aSnCl 4 with one molecule of water ﬁlling the sixth coordination site on the tin octahedron (Figure 7.27). This complexation stoichiometry is also obtained in solution, as veriﬁed by 119 Sn NMR studies that clearly show a doublet with ( 1JPaSn ) in the pentacoordinate chemical-shift region. The availability of an open coordination site in (S,S)-81aSnCl 4 suggested the possibility of incorporating a molecule of the substrate. Indeed, cocrystallization of (S,S)-81 with SnCl 4 and benzaldehyde aﬀorded a ternary complex, PhCHOa(S,S)-81aSnCl 4 (Figure 7.28). Both of these complexes had the same basic structural features as are found in the 2:1 complex ((S,S)-45)2 aSnCl 4 . Although these studies do indeed provide structural clues to the arrangement of groups around the central group 14 atom, there are still far too

313

314

7 Catalytic Enantioselective Aldol Additions with Chiral Lewis Bases

Fig. 7.28

X-ray crystal structure of PhCHO–(S,S)-81aSnCl4 .

many degrees of freedom to enable compelling depiction of the most favorable placement of reactive groups and alignment of combining faces. The structural insights available from these studies have, nevertheless, enabled important trends to emerge that facilitate the invention of new and better catalysts such as those that can enforce 2:1 binding by tethering and still accommodate the preferred arrangement of groups around the central atom. Such tethered dimeric phosphoramides have been prepared from several diﬀerent diamine subunits, for example those shown in Chart 7.3. In these cases the diamine subunits have been linked by aliphatic 1,n-diamines and have served admirably in a number reactions, for example catalytic enantioselective allylation with allylic trichlorosilanes [54c, 112] and activation of silicon tetrachloride for aldol and related reactions of trimethylsilyl enol ethers (Section 7.8). In the aldol addition of enoxytrichlorosilanes, the best results have been obtained from the use of the dimeric bis(phosphoramide) 48 for addition of aldehyde trichlorosilyl enolates (Section 7.6). A dimeric catalyst that promotes a highly enantioselective addition of trichlorosilyl enolates in general is still lacking [113].

7.10 Conclusions and Outlook

CH3 CH3 N O O N P P (CH2)n N N N N CH3CH3 CH3CH3

CH3 CH3 N O O N P P (CH2)n N N N N CH3 CH CH3 CH3 3 (R,R)-205: n = 3-6

(R)-(l,l )-204: n = 2-6

H H

N

O P

N

O

N

P (CH2)n N N N CH3 CH3

(R)-(l,l )-206: n = 4-6

315

H H

Ph Ph

CH3 CH3 N O O N P P (CH2)n N N N N CH3 CH3 CH3 CH3 (R)-(l,l )-207, n = 4-8

Chart 7.3

Tethered bisphosphoramides.

7.10

Conclusions and Outlook

The phenomenon of chiral Lewis base catalysis has been successfully demonstrated for a wide variety of aldol addition reactions. This represents a fundamentally new class of reactions that embody a conceptually novel and preparatively useful addition to the growing number catalytic, enantioselective processes. Design criteria for the invention of this new variant have been formulated and documented experimentally. Enoxytrichlorosilanes are a new class of aldolization reagents that are highly susceptible to catalysis by Lewis basic phosphoramides and N-oxides. A wide range of enolates have been prepared from simple cyclic and acyclic ketones, chiral ketones, esters, unsaturated esters, and aldehydes. Each of these classes of reagent has proven viable in aldol additions. The reactions are characterized by high yields, good functional group compatibility, excellent (and predictable) diastereoselectivity, and high enantioselectivity. There are, nevertheless, clearly identiﬁable limitations and shortcomings. For example, aliphatic aldehydes are a very important class of aldol partners that do not give generally acceptable results. In addition, the ability to generate substituted enolates with deﬁned geometry (both E and Z) is still limited. The concepts developed in this ﬁeld are also applicable (and have been applied) to other reactions such as allylation [112], imine addition, Michael addition, and epoxide opening [98]. Development of chiral Lewis base activation of Lewis acids is, furthermore, a powerful extension of these concepts that has enabled a broader range of carbon–carbon bond-forming processes to be executed under the action of enantioselective catalysis (e.g. the Passerini reaction [114]). In addition, Lewis base catalysis should ﬁnd applica-

Ph Ph

316

7 Catalytic Enantioselective Aldol Additions with Chiral Lewis Bases

tion in activation of processes associated with other main group elements capable of structural changes similar to silicon. Extensive kinetic and spectroscopic studies have revealed an unexpected mechanism involving the intermediacy of cationic silicon species. Elucidation of dual pathways proceeding via one or two-catalyst molecules has opened the door to the development of new dimeric catalysts that have proven useful in promoting faster, more selective reactions, but which have yet to ﬁnd application in the aldol process speciﬁcally. The synergistic evolution of synthetic utility and mechanistic understanding illustrates the fruitful interplay of synthesis, reactivity, and structure. These central activities constitute a chemical evergreen that will continue to address the challenges of the invention and development of new catalytic processes for years to come.

7.11

Representative Procedures 7.11.1

Preparation of Enoxytrichlorosilanes Transition Metal Catalyzed trans Silylation (Section 7.2, Scheme 7.11) – Preparation of Trichloro[(1-butylethenyl)oxy]silane (24). Silicon tetrachloride (9.18 mL, 80.0 mmol, 2.0 equiv.) was added quickly to a suspension of Hg(OAc)2 (127 mg, 0.40 mmol, 0.01 equiv.) in CH2 Cl2 (40 mL). During the addition the mercury salt dissolved. Trimethyl[(1-butylethenyl)oxy]silane (6.89 g, 40.0 mmol) was then added to the solution dropwise over 10 min and the solution was stirred at room temperature for an additional 50 min. During this time the reaction mixture became somewhat cloudy once again. Removal of a sample and 1 H NMR analysis indicated the reaction was complete. The mixture was concentrated at reduced pressure (100 mmHg) and the resulting oil was distilled twice through a 7.5 cm Vigreux column to give 7.76 g (83%) of the trichlorosilyl enolate 24 as a clear colorless oil. Metal Exchange via Lithium Enolate (Section 7.2, Scheme 7.11) – Preparation of (2Z,4S)-5-(tert-Butyl-dimethylsilyloxy)-4-methyl-3-trichlorosilyloxy-2-pentene ((Z)-35). (2Z,4S)-5-(tert-Butyl-dimethylsilyloxy)-4-methyl-3-trimethylsilyloxy2-pentene (908 mg, 3.00 mmol) was dissolved in 6 mL ether at 0 C. To this solution was slowly added MeLi (3.00 mL, 4.50 mmol, 1.5 equiv., 1.5 m in ether). The reaction mixture was stirred for 4.5 h at room temperature and then cooled to 78 C. The reaction mixture was transferred to a cold solution of silicon tetrachloride (3.45 mL, 30.0 mmol, 10 equiv.) in 6 mL ether by use of a cannula. The reaction mixture was stirred at 78 C for 1 h and then gradually warmed to room temp. The precipitate was left to settle at the bottom of the ﬂask and the supernatant was transferred to an-

7.11 Representative Procedures

other ﬂask by means of a cannula. The volatile compounds were removed under reduced pressure and the residue was distilled by means of a Kugelrohr apparatus to aﬀord 901 mg (2.48 mmol, 81%) (Z)-35 as a clear colorless oil. 7.11.2

Aldol Addition of Ketone-derived Enoxytrichlorosilane Aldol Addition of Achiral Enoxytrichlorosilane (Section 7.4, Scheme 7.26) – Preparation of (C)-S-1-Hydroxy-1-phenyl-3-heptanone (52). Trichlorosilyl enolate 24 (514 mg, 2.2 mmol, 1.1 equiv.) was added quickly to a cold (74 C) solution of (S,S)-45 (37.1 mg, 0.1 mmol, 0.05 equiv.) in CH2 Cl2 (2 mL). A solution of benzaldehyde (203 mL, 2.0 mmol) in CH2 Cl2 (2 mL) was cooled to 78 C and added quickly, via a short cannula, to the ﬁrst solution. During the addition the temperature rose to 68 C. The reaction mixture was stirred at 75 C for 2 h then quickly poured into cold (0 C) sat. aq. NaHCO3 solution. The slurry obtained was stirred for 15 min. The two-phase mixture was ﬁltered through Celite, the phases were separated, and the aqueous phase was extracted with CH2 Cl2 (3 50 mL). The organic extracts were combined, dried over Na2 SO4 , ﬁltered, and concentrated in vacuo. The crude product was puriﬁed by column chromatography (SiO2 , pentane–Et2 O, 4:1) to give 402.0 mg (98%) of ()-52 as a clear colorless oil. Aldol Addition of in-situ-generated Enoxytrichlorosilane (Section 7.5, Scheme 7.44) – Preparation of (1R,4S)-1-Hydroxy-4-[((dimethyl)-(1,1-dimethyl)silyl)oxy]1-phenyl-3-pentanone (syn-96). Trimethylsilyl enol ether 99 (548 mg, 2.0 mmol) was added dropwise over 2 min to a stirred solution of SiCl 4 (460 mL, 4.0 mmol, 2.0 equiv.) and Hg(OAc)2 (3.1 mg, 0.010 mmol, 0.005 equiv.) in CH2 Cl2 at room temperature. After complete addition the reaction mixture was stirred at room temperature for 1 h; volatile compounds were then removed under reduced pressure (0.3 mmHg) to give a cloudy residue. Dichloromethane (2.0 mL) was added and the mixture was cooled to 75 C. A solution of (R,R)-45 (37.0 mg, 0.1 mmol, 0.05 equiv., dried at 0.1 mmHg for 12 h) in CH2 Cl2 was then added over 1 min via a cannula. A solution of benzaldehyde (203 mL, 2.0 mmol) in CH2 Cl2 (1.0 mL) was then added over 1 min and the reaction mixture was stirred at 75 C for 3 h. The reaction mixture was then rapidly poured into cold (0 C) sat. aq. NaHCO3 solution (15 mL) and the mixture was stirred for 15 min. The heterogeneous mixture was ﬁltered through Celite, the organic phase was separated, and the aqueous phase was extracted with CH2 Cl2 (3 50 mL). The organic extracts were combined, dried over Na2 SO4 , ﬁltered, and concentrated to give a crude oil. Puriﬁcation by column chromatography (SiO2 , hexane–EtOAc, 8:1) aﬀorded 524.4 mg (85%) of a mixture of diastereomers 96 as a clear colorless oil. The diastereomeric ratio was determined by SFC analysis to be syn/anti 73:1.

317

318

7 Catalytic Enantioselective Aldol Additions with Chiral Lewis Bases

Aldol Addition of Aldehyde-derived Enoxytrichlorosilane (Section 7.6, Scheme 7.55) – Preparation of (1S,2S)-3,3-Dimethoxy-2-pentyl-1-phenyl-1-propanol (syn132). Trichlorosilyl enolate (Z)-37 (496 mg, 2.0 mmol, 1.0 equiv.) was added to a cold (78 C) solution of the bisphosphoramide (R,R)-48 (84 mg, 0.1 mmol, 0.05 equiv.) in CHCl3 aCH2 Cl2 , 4:1 (8 mL) and the mixture was stirred for 10 min. Freshly distilled benzaldehyde (0.205 mL, 2.0 mmol, 1.0 equiv.) was then added. After 6 h at 78 C, MeOH (32 mL) was added and the mixture was stirred at that temperature for 45 min. The cold bath was removed and reaction mixture was left to warm to room temperature (total time 0.5 h), then was poured into cold (0 C) sat. aq. NaHCO3 solution and the mixture was stirred for 4 h. The reaction mixture was ﬁltered through Celite and then washed with pentane–Et2 O, 1:1 (20 mL). The organic layer was separated and the aqueous layer was extracted once with pentane–Et2 O, 1:1 (20 mL). The combined extracts were dried over MgSO4 and then concentrated in vacuo. Column chromatography (SiO2 , hexane–EtOAc, 85:15) then bulb-to-bulb distillation gave 491 mg (92%) syn-132 as a clear, colorless, viscous liquid. Aldol Addition of Trichlorosilyl Ketene Acetal (Section 7.7, Scheme 7.65) – Preparation of Methyl 3-Hydroxy-3-phenylbutanoate (157). Trichlorosilyl ketene acetal 10 (380 mL, 2.4 mmol, 1.2 equiv.) was added to a solution of acetophenone (240 mL, 2.0 mmol) and chiral bis-N-oxide (P)-(R,R)-46 (101 mg, 0.20 mmol, 0.1 equiv.) in CH2 Cl2 (10 mL) at 20 C under nitrogen in a ﬂame-dried, round-bottomed ﬂask with magnetic stirrer. After stirring for 12 h at 20 C the reaction mixture was transferred dropwise to a cold (0 C) sat. aq. NaHCO3 solution (20 mL) with vigorous stirring. The mixture was further stirred for 30 min at room temperature. The silicate precipitate was removed by ﬁltration through Celite and the ﬁltrate was extracted with CH2 Cl2 (4 20 mL). The combined organic extracts were dried over MgSO4 and then were concentrated under reduced pressure. The crude aldol product was separated from the catalyst by distillation and was further puriﬁed by silica gel chromatography. Analytically pure 157 (364 mg, 94%) was obtained as a colorless liquid after bulb-to-bulb distillation. Aldol Addition of Propionate-derived Silyl Ketene Acetal (Section 7.8, Scheme 7.70) – Preparation of tert-Butyl (2S,3R)-3-Hydroxy-2-Methyl-3-Phenylpropanate (anti-182). A ﬂame-dried, 10-mL, 2-neck ﬂask containing a solution of bisphosphoramide (R,R)-48 (8.4 mg, 0.01 mmol, 0.01 equiv.) in CH2 Cl2 (5 mL) was cooled to 78 C under nitrogen and benzaldehyde (102 mL, 1.0 mmol, 1.0 equiv.) was then added. Silicon tetrachloride (123 mL, 1.1 mmol, 1.1 equiv.) was added to the resulting solution and the reaction mixture was stirred at 78 C for 5 min. (E)-1-[(tert-Butoxy)propenyl]-tert-butyldimethylsilane ((E)-183) (293 mg, 1.2 mmol, 1.2 equiv.) was then added dropwise to the reaction mixture over 5 min. The resulting mixture was stirred at 78 C (bath temperature) for 3 h whereupon the cold reaction mixture

References

was poured into 1:1 sat. aq. KF–1.0 m KH2 PO4 solution (20 mL) with rapid stirring. This two-phase mixture was stirred vigorously for 1 h before ﬁltration through Celite. The aqueous layer was washed with CH2 Cl2 (3 50 mL) and the combined organic extracts were washed with brine (50 mL), dried over Na2 SO4 (2 g), ﬁltered, and the ﬁltrate was concentrated in vacuo. The residue was puriﬁed by Kugelrohr distillation to yield 217 mg (93%) (2S,3R)-182 as a colorless oil.

References 1 (a) Evans, D. A.; Nelson, J. V.; Taber, T. R. In Topics in

Stereochemistry; Eliel, E. L., Wilen, S. H., Eds.; Wiley Interscience: New York, 1982; Vol. 13; pp 1–115. (b) Heathcock, C. H. In Comprehensive Carbanion Chemistry; Buncel, E., Durst, T., Eds.; Elsevier: New York, 1984; Vol. 5B, p 177–237. (c) Mukaiyama, T. Org. React. 1982, 28, 203– 331. (d) Braun, M. Angew. Chem., Int. Ed. Engl. 1987, 26, 24– 37. (e) Mukaiyama, T.; Kobayashi, S. Org. React. 1994, 46, 1–103. (f ) Norcross, R.; Paterson, I. Chem. Rev. 1995, 95, 2041–2114. (g) Carreira, E. M. In Modern Carbonyl Chemistry; Otera, J., Ed.; Wiley–VCH: Weinheim, 2000; Chapter 8. (h) Paterson, I.; Cowden, C. J.; Wallace, D. J. In Modern Carbonyl Chemistry; Otera, J., Ed.; Wiley–VCH: Weinheim, 2000; Chapter 9. (i) Carreira, E. M. In Comprehensive Asymmetric Catalysis, Vol. I–III; Jacobsen, E. N.; Pfaltz, A.; Yamamoto, H., Eds.; Springer: Heidelberg, 1999. Chapter 29. ( j) Carreira, E. M. In Catalytic Asymmetric Synthesis, 2nd Ed.; Ojima, I., Ed.; Wiley–VCH: Weinheim, 2000; Chapter 8B2. (k) Braun, M. In Stereoselective Synthesis, Methods of Organic Chemistry (Houben–Weyl); Edition E21; Helmchen, G.; Hoffman, R.; Mulzer, J.; Schaumann, E., Eds.; Thieme: Stuttgart, 1996; Vol. 3; pp 1603–1735. 2 (a) Masamune, S.; Choy, W.; Peterson, J. S.; Sita, L. R. Angew. Chem., Int. End. Engl. 1985, 24, 1–30. (b) Helmchen, G. In Stereoselective Synthesis, Methods of Organic Chemistry (Houben–Weyl); Edition E21; Helmchen, G.; Hoffman, R.; Mulzer, J.; Schaumann, E., Eds.; Thieme: Stuttgart, 1996; Vol. 1; pp 56–63. (c) Cowden, C. J.; Paterson, I. Org. React. 1997, 51, 1–200. 3 (a) Heathcock, C. H. In Comprehensive Organic Synthesis, Vol. 2, Additions to CaX Bonds, Part 2; Heathcock, C. H., Ed.; Pergamon Press: Oxford, 1991; Chapter 1.6. (b) Lynch, J. E.; Volante, R. P.; Wattley, R. V.; Shinkai, I. Tetrahedron Lett. 1987, 28, 1385–1388. (c) Oppolzer, W. Pure Appl. Chem. 1988, 60, 39–48. 4 (a) Kim, B. M.; Williams, S. F.; Masamune, S. In Comprehensive Organic Synthesis: Additions to CaX p-Bonds Part 2; Heathcock, C. H., Ed. Pergamon Press: Oxford, 1991; Chapt. 1.7. (b) Mahrwald, R. Chem. Rev. 1999, 99, 1095– 1120. (c) Reetz, M. T. In Organometallics in Synthesis;

319

320

7 Catalytic Enantioselective Aldol Additions with Chiral Lewis Bases

5 6 7

8

9 10

11

12

13

14 15

16

17

18

Schlosser, Ed.; John Wiley: New York, 1994; Chapter 3. (d) Paterson, I. In Comprehensive Organic Synthesis: Additions to CaX p-Bonds Part 2; Heathcock, C. H., Ed.; Pergamon Press: Oxford, 1991; Chapt. 1.9. Corey, E. J.; Kim, S. S. J. Am. Chem. Soc. 1990, 112, 4976– 4977. Duthaler, R. O.; Hafner, A. Chem. Rev. 1992, 92, 807– 832. (a) Myers, A. G.; Widdowson, K. L.; Kukkola, P. J. J. Am. Chem. Soc. 1992, 114, 2765–2767. (b) Myers, A. G.; Kephart, S. E.; Chen, H. J. Am. Chem. Soc. 1992, 114, 7922–7923. (a) Lewis Acids in Organic Synthesis; Yamamoto, H., Ed.; Wiley–VCH: Weinheim, 2001; Vols. 1 and 2. (b) Santelli, M.; Pons, J.-M. Lewis Acids and Selectivity in Organic Synthesis; CRC: Boca Raton, 1996. Kobayashi, S.; Uchiro, H.; Shiina, I.; Mukaiyama, T. Tetrahedron 1993, 49, 1761–1772. (a) Ishihara, K.; Maruyama, T.; Mouri, M.; Furuta, K.; Yamamoto, H. Bull. Chem. Soc. Jpn. 1993, 66, 3483–3491. (b) Ishihara, K.; Gao, Q.; Yamamoto, H. J. Am. Chem. Soc. 1993, 115, 10412–10413. (a) Parmee, E. R.; Tempkin, O.; Masamune, S.; Abiko, A. J. Am. Chem. Soc. 1991, 113, 9365–9366. (b) Kiyooka, S.-I.; Kaneko, Y.; Komura, M.; Matsuo, H.; Nakano, M. J. Org. Chem. 1991, 56, 2276–2278. (c) Corey, E. J.; Cywin, C. L.; Roper, T. D. Tetrahedron Lett. 1992, 33, 6907–6910. (a) Mikami, K.; Matsukawa, S. J. Am. Chem. Soc. 1994, 116, 4077–4078. (b) Sato, M.; Sunami, S.; Sugita, Y.; Kaneko, C. Chem. Pharm. Bull. Jpn. 1994, 42, 839–845. (c) Keck, G. E.; Krishnamurthy, D. J. Am. Chem. Soc. 1995, 117, 2363–2364. (a) Carreira, E. M.; Singer, R. A.; Lee, W. J. Am. Chem. Soc. 1994, 116, 8837–8838. (b) Carreira, E. M.; Singer, R. A. J. Am. Chem. Soc. 1995, 117, 12360–12361. Yanagisawa, A.; Matsumoto, Y.; Nakashima, H.; Asakawa, K.; Yamamoto, H. J. Am. Chem. Soc. 1997, 119, 9319–9320. (a) Evans, D. A.; Murry, J. A.; Kozlowski, M. C. J. Am. Chem. Soc. 1996, 118, 5814–5815. (b) Evans, D. A.; Kozlowski, M. C. Murgey, C. S.; MacMillan, D. W. C. J. Am. Chem. Soc. 1997, 119, 7893–7894. (c) Johnson, J. S.; Evans, D. A. Acc. Chem. Res. 2000, 33, 325–335. For examples of anti-selective Mukaiyama aldol additions see: For examples of primarily anti-selective catalytic enantioselective aldol additions, see: (a) Evans, D. A.; MacMillan, D. W. C.; Campos, K. R. J. Am. Chem. Soc. 1997, 119, 10859–10860. (b) Kobayashi, S.; Horibe, M.; Hachiya, I. Tetrahedron Lett. 1995, 36, 3173–3176. (c) Yamashita, Y.; Ishitani, H.; Shimizu, H.; Kobayashi, S. J. Am. Chem. Soc. 2002, 124, 3292–3302. (a) Sodeoka, M.; Tokunoh, R.; Miyazaki, F.; Hagiwara, E.; Shibasaki, M. Synlett 1997, 463–466. (b) Ref. 14. (c) Krueger, J.; Carreira, E. M. J. Am. Chem. Soc. 1998, 120, 837–838. (d) Wadamoto, M.; Ozasa, N.; Yanagisawa, A.; Yamamoto, H. J. Org. Chem. 2003, 68, 5593–5601. (a) Yamada, Y. M. A.; Yoshikawa, N.; Sasai, M.; Shibasaki,

References

19 20

21

22

23 24

25 26 27 28

29 30

31 32

M. Angew. Chem., Int. Ed. Engl. 1997, 36, 1871–1873. (b) Trost, B. M.; Ito, H.; Silcoff, E. R. J. Am. Chem. Soc. 2001, 123, 3367–3368. Berrisford, D. J.; Bolm, C.; Sharpless, K. B. Angew. Chem., Int. Ed. Engl. 1995, 34, 1059–1070. (a) Reichardt, C. Solvents and Solvent Eﬀects in Organic Chemistry, Second Edition; VCH: Weinheim, 1988; pp 17–27. (b) Bollinger, J.-C.; Faure, R.; Yvernault, T. Can. J. Chem. 1983, 61, 328–333. (c) Gritzner, G.; Ho¨rzenberger, F. J. Chem. Soc. Faraday Trans. 1995, 91, 3843–3850. (d) Gritzner, G. Zeitschrift fu¨ r Physikalische Chemie Neue Folge 1988, 158, 99–107. (e) Gritzner, G. Journal of Molecular Liquids 1997, 487–500. (a) Sakurai, H. Synlett 1989, 1–8. (b) Sakurai, H. In Proceedings of the 5th International Kyoto Conference on New Aspects of Organic Chemistry; Yoshida, Z.-I., Ohshiro, Y., Eds.; Kodansha Press: Tokyo, Japan, 1992; pp 129–157 and references cited therein. (c) Kira, M.; Zhang, L.; Kabuto, C.; Sakurai, H. Organometallics 1996, 15, 5335–5341. (a) Kobayashi, S.; Nishio, K. Tetrahedron Lett. 1993, 34, 3453– 3456. (b) Kobayashi, S.; Nishio, K. Synthesis 1994, 457–459. (c) Kobayashi, S.; Nishio, K. J. Org. Chem. 1994, 59, 6620– 6628. Denmark, S. E.; Coe, D. M.; Pratt, N. E.; Griedel, B. D. J. Org. Chem. 1994, 59, 6161–6163. (a) Burlachenko, G. S.; Khasapov, B. N.; Petrovskaya, L. I.; Baukov, Y. I.; Lutsenko, I. F. J. Gen. Chem. USSR (Engl. Transl.) 1966, 36, 532–537. (b) Ponomarev, S. V.; Baukov, Y. I.; Dudukina, O. V.; Petrosyan, I. V.; Petrovskaya, L. I. Zh. Obshch. Khim. 1967, 37, 2204–2207. (c) Benkeser, R. A.; Smith, W. E. J. Am. Chem. Soc. 1968, 90, 5307–5309. Bassindale, A. R.; Stout, T. Tetrahedron Lett. 1985, 26, 3403– 3406. Denmark, S. E.; Winter, S. B. D.; Su, X.; Wong, K.-T. J. Am. Chem. Soc. 1996, 118, 7404–7405. Denmark, S. E.; Stavenger, R. A. Acc. Chem. Res. 2000, 32, 432–440. Mekelburger, H. B.; Wilcox, C. S. ‘‘Formation of Enolates’’ in Comprehensive Organic Synthesis, Vol. 2, Additions to CaX p-Bonds, Part 2; Heathcock, C. H., Ed.; Pergamon Press: New York, 1991; pp 99–131. Petrov, A. D.; Sadykh-Zade, S. J. Dokl. Akad. Nauk. SSSR 1958, 121, 119–122. (a) Chan, T.-H. ‘‘Formation and Addition Reactions of Enol Ethers in Comprehensive Organic Synthesis, Vol. 2, Additions to CaX p-Bonds, Part 2; Heathcock, C. H., Ed.; Pergamon Press: New York, 1991; pp. 595–628. (b) Rasmussen, J. K. Synthesis 1977, 91–110. (c) Brownbridge, P. Synthesis 1983, 1, 85–104. (d) Baukov, Yu. I.; Lutsenko, I. F. Organometallic Chemistry Review A 1970, 6, 355–445. (a) Mukaiyama, T.; Banno, K.; Narasaka, K. J. Am. Chem. Soc. 1974, 96, 7503–7509. (b) Ref. 1c. Mayr, H.; Kempf, B.; Oﬁal, A. R. Acc. Chem. Res. 2003, 36, 66–77.

321

322

7 Catalytic Enantioselective Aldol Additions with Chiral Lewis Bases 33 For a comprehensive review see: Nelson, S. G. Tetrahedron

Asymm. 1998, 9, 357–389. 34 For examples of silicon-functionalized silyl enol ethers see

35 36 37 38

39

40 41

42 43

44

45 46

47 48 49 50

(a) Walkup, R. D. Tetrahedron Lett. 1987, 28, 511–514. (b) Walkup, R. D.; Obeyesekere, N. U. J. Org. Chem. 1988, 53, 920–923. (c) Walkup, R. D.; Obeyesekere, N. U.; Kane, R. R. Chem. Lett. 1990, 1055–1058. (d) Kaye, P. T.; Learmonth, R. A.; Ravindran, S. S. Synth. Commun. 1993, 23, 437–444. Mori, A.; Kato, T. Synlett 2002, 1167–1169. Yanagisawa, A.; Nakatsuka, Y.; Asakawa, K.; Kageyama, H.; Yamamoto, H. Synlett 2001, 69–72. Denmark, S. E.; Stavenger, R. A.; Winterm S. B. D.; Wong, K-T.; Barsanti, P. A. J. Org. Chem. 1998, 63, 9517–9523. Burlachenko, G. S.; Khasapov, B. N.; Petrovskaya, L. I.; Baukov, Yu. I.; Lutsenko, I. F. J. Gen. Chem. USSR (Engl. Transl.) 1966, 36, 532–537. Aldol addition can occur in aqueous solvent without Lewis acid catalyst: (a) Lubineau, A. J. Org. Chem. 1986, 51, 2142– 2144. (b) Loh, T.-P.; Feng, L.-C.; Wei, L.-L. Tetrahedron 2000, 56, 7309–7312. Benkeser, R. A.; Smith, W. E. J. Am. Chem. Soc. 1968, 90, 5307–5309. Panomarev, S. V.; Baukov, Yu. I.; Dudukina, O. V.; Petrosyan, I. V.; Petrovskaya, L. I. J. Gen. Chem. USSR (Engl. Transl.) 1967, 37, 2092–2094. Pereyre, M.; Bellegards, B.; Mendelsohn, J.; Valade, J. J. Organomet. Chem. 1968, 11, 97–110. Formation of C-mercurioketone: (a) House, H. O.; Auerbach, R. A.; Gall, M.; Peet, N. P. J. Org. Chem. 1973, 38, 514–522. (b) Yamamoto, Y.; Maruyama, K. J. Am. Chem. Soc. 1982, 104, 2323–2325. (c) Bluthe, N. Malacria, M.; Gore, J. Tetrahedron 1984, 40, 3277–3284. (d) Drouin, J.; Bonaventura, M.-A.; Conia, J.-M. J. Am. Chem. Soc. 1985, 107, 1726–1729. Formation of C-(trichlorostannyl)ketones: (a) Nakamura, E.; Kuwajima, I. Chem. Lett. 1983, 59–62. (b) Nakamura, E.; Kuwajima, I. Tetrahedron Lett. 1983, 24, 3347–3350. (c) Annunziata, R.; Benaglia, M.; Cinquini, M.; Cozzi, F.; Rainmondi, L. Tetrahedron 1994, 50, 5821–5828. (d) Kuwajima, I.; Nakamura, E. Acc. Chem. Res. 1985, 18, 181– 187. Denmark, S. E.; Pham, S. M. J. Org. Chem. 2003, 68, 5045– 5055. (a) Stork, G.; Hudrlik, P. F. J. Am. Chem. Soc. 1968, 90, 4464–4465. (b) House, H. O.; Trost, B. M. J. Org. Chem. 1965, 30, 2502–2512. Evans, D. A.; Nelson, J. V.; Vogel, E.; Taber, T. R. J. Am. Chem. Soc. 1981, 103, 3099–3111. Hall, P. L.; Gilchrist, J. H.; Collum, D. B. J. Am. Chem. Soc. 1991, 113, 9571–9574. Denmark, S. E.; Ghosh, S. K. Angew. Chem. Int. Ed. 2001, 40, 4759–4762. Denmark, S. E.; Wynn, T.; Beutner, G. L. J. Am. Chem. Soc. 2002, 124, 13405–13407.

References 51 (a) Baukov, Yu. I.; Lutsenko, I. F. Moscow Univ. Chem. Bull.

(Engl. Transl.) 1970, 25, 72. (b) Ref. 26. 52 Denmark, S. E.; Su, X.; Nishigaichi, Y.; Coe, D. M.; Wong,

53 54

55 56

57 58 59 60 61

62

63 64

65 66

67 68

69 70 71

72

K.-T.; Winter, S. B. D.; Choi, J. Y. J. Org. Chem. 1999, 64, 1958–1967. Denmark, S. E.; Stavenger, R. A.; Wong, K.-T.; Su, X. J. Am. Chem. Soc. 1999, 121, 4982–4991. (a) Denmark, S. E.; Fu, J. J. Am. Chem. Soc. 2001, 123, 9488– 9489. (b) Denmark, S. E.; Wynn, T. J. Am. Chem. Soc. 2001, 123, 6199–6200. (c) Denmark, S. E.; Fu, J. Chem. Commun. 2003, 167–170. Denmark, S. E.; Fan, Y. J. Am. Chem. Soc. 2002, 124, 4233– 4235. (a) Iseki, K.; Kuroki, Y.; Takahashi, M.; Kishimoto, S.; Kobayashi, Y. Tetrahedron 1997, 53, 3513–3526. (b) Peyronel, J.-F.; Samuel, O.; Fiaud, J.-C. J. Org. Chem. 1987, 52, 5320– 5325. Alexakis, A.; Aujard, I.; Mangeney, P. Synlett 1998, 873– 874. Holmes, R. R.; Chem. Rev. 1996, 96, 927–950. Miyano, S.; Nawa, M.; Mori, A.; Hashimoto, H. Bull. Chem. Soc. Jpn. 1984, 57, 2171–2176. Lesiak, T.; Seyda, K. J. Prakt. Chem. 1979, 321, 161–163. (a) Krajnik, P.; Ferguson, R. R.; Crabtree, R. H. New. J. Chem. 1993, 17, 559–566. (b) Denmark, S. E.; Fu, J.; Lawler, M. J. Org. Synth. 2004, 81, in press. (a) Nakajima, M.; Saito, M.; Shiro, M.; Hashimoto, S. J. Am. Chem. Soc. 1998, 120, 6419–6420. (b) Tao, B.; Lo, M.-C.; Fu, G. C. J. Am. Chem. Soc. 2001, 123, 353–354. Bolm, C.; Ewald, M.; Felder, M.; Schlingloff, G. Chem. Ber. 1992, 125, 453–458. For solvent-assisted reactions of alkylsilyl enol ethers under mild conditions, see: (a) Lubineau, A.; Ange, J.; Queneau, Y. Synthesis, 1994, 741–760. (b) Rajanbabu, T. V. J. Org. Chem. 1984, 49, 2083–2089. Denmark, S. E.; Griedel, B. D.; Coe, D. M.; Schnute, M. E. J. Am. Chem. Soc. 1994, 116, 7026–7043. (a) Denmark, S. E.; Su, X.; Nishigaich, Y. J. Am. Chem. Soc. 1998, 120, 12990–12991. (b) Denmark, S. E.; Pham, S. M. Helv. Chim. Acta. 2000, 122, 1846–1853. Zimmerman, H. E.; Traxler, M. D. J. Am. Chem. Soc. 1957, 79, 1920–1923. (a) Denmark, S. E.; Stavenger, R. A.; Wong, K.-T. J. Org. Chem. 1998, 63, 918–919. (b) Denmark, S. E.; Stavenger, R. A. J. Am. Chem. Soc. 2000, 122, 8837–8847. Maria, P.-C.; Gal, J.-F. J. Phys. Chem. 1985, 89, 1296– 1304. Denmark, S. E.; Su, X. Tetrahedron 1999, 55, 8727–8738. For deﬁnitions of these stereochemical terms, see: Denmark, S. E.; Almstead, N. G. In Modern Carbonyl Chemistry; Otera, J., Ed.; Wiley–VCH: Weinheim, 2000; Chapter 10. pp. 300– 301. Enoxysilanes that react via boat-like transition structure involving trigonal bipyramidal silicon species: (a) Ref. 7(b).

323

324

7 Catalytic Enantioselective Aldol Additions with Chiral Lewis Bases

73 74

75

76 77 78 79 80 81

82 83

84

85 86 87

88 89 90

91

(b) Gung, B. W.; Zhu, Z.; Fouch, R. A. J. Org. Chem. 1995, 60, 2860–2864. Lodge, E., P.; Heathcock, C. H. J. Am. Chem. Soc. 1987, 109, 3353–3361. (a) Reetz, M. T. Angew. Chem. Int. Ed. 1984, 23, 556–569. (b) Reetz, M. T. Acc. Chem. Res. 1993, 26, 462–468. (c) Reetz, M. T. Chem. Rev. 1999, 99, 1121–1162. For reviews on diastereoselective aldol additions using chiral aldehydes, see: (a) Gennari, C. In Comprehensive Organic Synthesis, Vol. 2, Additions to CaX p Bonds, Part 2; Heathcock, C. H., Ed.; Pergamon Press: Oxford, 1991; pp. 639–647. (b) Ref. 1(k), pp 1713–1722. Denmark, S. E.; Wong, K.-T.; Stavenger, R. A. J. Am. Chem. Soc. 1997, 119, 2333–2334. Denmark, S. E.; Stavenger, R. A.; Wong, K.-T. Tetrahedron 1998, 54, 10389–10402. Seebach, D.; Prelog, V. Angew. Chem. 1982, 21, 654–660. Paterson, I.; Goodman, J. M.; Isaka, M. Tetrahedron Lett. 1989, 30, 7121–7124. Denmark, S. E.; Stavenger, R. A. J. Org. Chem. 1998, 63, 9524–9527. For a discussion of the coordinating abilities of various ether substituents, see: (a) Ref. 74(b). (b) Chen, X.; Hortellano, E. R.; Eliel, E. L.; Frye, S. V. J. Am. Chem. Soc. 1990, 112, 6130– 6131. (c) Mori, S.; Nakamura, M.; Nakamura, E.; Koga, N.; Morokuma, K. J. Am. Chem. Soc. 1995, 117, 5055–5065. Denmark, S. E.; Pham, S. M. Org. Lett. 2001, 3, 2201–2204. For a discussion of chelation as a stereocontrol element in lactate-derived stannous enolate, see: Paterson, I.; Tillyer, R. Tetrahedron Lett. 1992, 33, 4233–4236. (a) Martin, V. A.; Murray, D. H.; Pratt, N. E.; Zhao, Y.; Albizati, K. F. J. Am. Chem. Soc. 1990, 112, 6965–6978. (b) Paterson, I.; Oballa, R. M. Tetrahedron Lett. 1997, 38, 8241– 8244. (c) Feutrill, J. T.; Lilly, M. J.; Rizzacasa, M. A. Org. Lett. 2002, 4, 525–527. Denmark, S. E.; Fujimori, S. Synlett 2001, 1024–1029. Denmark, S. E.; Fujimori, S. Org. Lett. 2002, 4, 3473– 3476. (a) Paterson, I.; Gibson, K. R.; Oballa, R. M. Tetrahedron Lett. 1996, 37, 8585–8588. (b) Paterson, I.; Collett, L. A. Tetrahedron Lett. 2001, 42, 1187–1191. (c) Paterson, I.; Oballa, R. M.; Norcross, R. D. Tetrahedron Lett. 1996, 37, 8581–8584. (d) Evans, D. A.; Coleman, P. J.; Cote, B. J. Org. Chem. 1997, 62, 788–789. Denmark, S. E.; Fujimori, S. Org. Lett. 2002, 4, 3477–3480. Alcaide, B.; Almendros, P. Angew. Chem. Int. Ed. 2003, 42, 858–860. (a) Northrup, A. B.; MacMillan, D. W. C. J. Am. Chem. Soc. 2002, 124, 6798–6799. (b) Kandasamy, S.; Notz, W.; Bui, T.; Barbas, C. F. III, J. Am. Chem. Soc. 2001, 123, 5260–5267. (a) Yan, T.-H.; Tan, C.-W.; Lee, H.-C.; Lo, H.-C.; Huang, T.Y. J. Am. Chem. Soc. 1993, 115, 2613–2621. (b) Heathcock, C. H.; White, C. T.; Morrison, J. J.; Van Derveer, D. J. Org. Chem. 1981, 46, 1296–1309.

References 92 Paterson, I.; Wallace, D. J.; Cowden, C. J. Synthesis 1998,

639–652. 93 Denmark, S. E.; Bui, T. Proc. Nat. Acad. Sci. 2004, 101, 5439–

5444. 94 Denmark, S. E.; Fu, J. J. Am. Chem. Soc. 2000, 122, 12021–

12022. 95 (a) Bassindale, A. R.; Stout, T. J. Organomet. Chem. 1982,

96

97

98 99 100 101

102

103

104

105 106

238, C41–C45. (b) Bassindale, A. R.; Glynn, S. J.; Taylor, P. G. In The Chemistry of Organic Silicon Compounds; Rappoport, Z.; Apeloig, Y., Eds.; Wiley: Chichester, 1998; Vol. 2, pp 495– 511. (a) Gutmann, V. The Donor–Acceptor Approach to Molecular Interactions; Plenum Press: New York, 1978. (b) Jensen, W. B. The Lewis Acid–Base Concepts; Wiley Interscience: New York, 1980; Chapter 4. (a) Chojnowski, J.; Cypryk, M.; Michalski, J.; Wozniak, J. J. Organomet. Chem. 1985, 288, 275–282. (b) Corriu, R. J. P.; Dabosi, G.; Martineau, M. J. Organomet. Chem. 1980, 186, 25–37. (c) Bassindale, A. R.; Lau, J. C.-Y.; Taylor, P. G. J. Organomet. Chem. 1995, 499, 137–141. Denmark, S. E.; Barsanti, P. A.; Wong, K.-T.; Stavenger, R. A. J. Org. Chem. 1998, 63, 2428–2429. Denmark, S. E.; Heemstra, J. R. Jr. Org. Lett. 2003, 5, 2303– 2306. Denmark, S. E.; Beutner, G. L. J. Am. Chem. Soc. 2003, 125, 7800–7801. (a) Fleming, I. Frontier Orbitals and Organic Chemical Reactions; Wiley–Interscience: New York, 1996; p. 40–47. (b) Herrmann, J. L.; Kieczykowski, G. R.; Schlessinger, R. H. Tetrahedron Lett. 1973, 14, 2433–2436. For examples of stereoselective vinylogous aldol reactions see: (a) Saito, S.; Shiozawa, M.; Ito, M.; Yamamoto, H. J. Am. Chem. Soc. 1998, 120, 813–814. (b) Ref. 13(b). (c) Bluet, G.; Campagne, J.-M. J. Org. Chem. 2001, 66, 4293–4298. (d) De Rosa, M.; Soriente, A.; Scettri, A. Tetrahedron Asymmetry 2000, 11, 2255–2258. (e) Evans, D. A.; Kozlowski, M. C.; Murry, J. A.; Burgey, C. S.; Campos, K. R.; Connel, B. T.; Staples, R. J. J. Am. Chem. Soc. 1999, 121, 669–685. For a general discussion of this position see: Denmark, S. E.; Stavenger, R. A.; Su, X.; Wong, K.-T.; Nishigaichi, Y. Pure & Appl. Chem. 1998, 70, 1469–1476. (a) Guillaneux, D.; Zhao, S.-H.; Samuel, O.; Rainford, D.; Kagan, H. B. Nonlinear Eﬀects in Asymmetric Catalysis. J. Am. Chem. Soc. 1994, 116, 9430–9439. (b) Kagan, H. B.; Fenwick, D. Asymmetric Ampliﬁcation. Top. Stereochem. 1999, 22, 257–296. (c) Avalos, M.; Babiano, R.; Cintas, P.; Jimenez, J. L.; Palacios, J. C. Nonlinear Stereochemical Eﬀects in Asymmetric Reactions. Tetrahedron: Asymmetry 1997, 8, 2997–3017. Singleton, D. A.; Thomas, A. A. J. Am. Chem. Soc. 1995, 117, 9357–9358. For a detailed discussion involving the determination of 13 C KIE from NMR integration and error analysis; Ref. 105.

325

326

7 Catalytic Enantioselective Aldol Additions with Chiral Lewis Bases 107 Pham, S. M. Ph.D. Thesis, University of Illinois, Urbana–

Champaign, 2002. 108 A similar study has been completed in the addition of

109 110

111

112 113 114

isobutyraldehyde trichlorosilyl enolate to benzaldehyde. Here as well, the aldolization step is shown to be rate-limiting; Ref. 93. The importance of the ordering of the subsequent steps is at present unknown. The conﬁguration around the octahedral silicon cation is unknown. Moreover, given the divergent criteria for which would be more stable and which more reactive, a deﬁnitive answer must await computational analysis. Attempts to identify stable complexes with silicon(IV) Lewis acids and phosphoramides have as yet been unsuccessful. However, a stable complex of bis(N-oxide) 46 with silicon tetrachloride has been analyzed crystallographically, Fan, Y. unpublished results from these laboratories. Denmark, S. E.; Fu, J. J. Am. Chem. Soc. 2003, 125, 2208– 2216. For recent studies on linked phosphoramides in the aldol addition of ethyl ketone trichlorosilyl enolates; Ref. 45. Denmark, S. E.; Fan, Y. J. Am. Chem. Soc. 2003, 125, 7825– 7827.

327

8

The Aldol–Tishchenko Reaction R. Mahrwald 8.1

Introduction

The Tishchenko reaction has been known for almost 100 years [1]. The importance of catalysis in this reaction – dimerization of aldehydes to the corresponding esters and the polymerization of dialdehydes to the expected polyesters – has grown in the last 50 years. This reaction has great potential in stereoselective synthesis of deﬁned stereocenters. Depending on the nature of substrates, reaction conditions, and catalysts, deﬁned diastereoselective and enantioselective stereogenic centers can be created.

8.2

The Aldol–Tishchenko Reaction

The aldol–Tishchenko reaction was ﬁrst studied at the beginning of the last century [2]. Although in many Tishchenko reactions the aldol–Tishchenko reaction is a competitive transformation, by use of the right reaction conditions and catalysts one can aﬀect which pathway is taken. In recent years interest in this area has increased substantially. This reaction can be performed either with enolizable aldehydes (resulting in trimerization of aldehydes) or with ketones (resulting in formation of 1,3-diol monoesters). 8.2.1

The Aldol–Tishchenko Reaction with Enolizable Aldehydes

The classic aldol–Tishchenko reaction is used to obtain 1,3-diol monoesters by self-addition of aldehydes with at least one a-hydrogen [3]. In the ﬁrst step of this reaction two molecules of aldehyde react by reversible aldol addition to give the expected aldol adduct; this is further reduced by a third molecule of aldehyde to give the 1,3-diol monoesters, 1 and 2 (Eq. (1)). Modern Aldol Reactions. Vol. 2: Metal Catalysis. Edited by Rainer Mahrwald Copyright 8 2004 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 3-527-30714-1

328

8 The Aldol–Tishchenko Reaction

3 R - CH2 - CHO catalyst

R OH

R O

R O 1 +

R O

R OH

R O 2

Equation 1

Catalysts: magnesium-2,4,6trimethylphenoxide [4], Cp2 Sm(THF)2 [5], SmI2 [6], LiO-iPr [7], BINOL-Li [8], Y2 O(OiPr)13 [9].

Merger et al. [10] reacted aldol adducts with aldehydes and isolated the 1,3-diol monoesters. These aldol–Tishchenko reactions were performed in the presence of metal alkoxides or without catalysts at higher temperatures. They showed that:

. the ester functionality does not come from the intermolecular combinaof two aldehydes; . tion hydride shift occurs in an intermediate equilibrium of hemiacetals and (Scheme 8.1); . dioxanolen aldols are hydride acceptors – the carbonyl function will be reduced; and . primary 1,3-diol monoesters are the thermodynamically stable products and they are formed by an acyl migration during the reaction. Results from mechanistic study of the stereoselective aldol–Tishchenko reaction support the mechanism depicted in Scheme 8.1. First, a rapid aldol reaction occurs and by reaction with a further molecule of aldehyde the hemiacetal 4 is formed. Subsequent hydride transfer (in the intermediate equilibrium of hemiacetals and dioxanolen) yields the 1,3-diol monoester 5. There are only two examples of enantioselective execution of the aldol– Tishchenko reaction of aldehydes. Loog and Ma¨eorg used chiral binaph-

8.2 The Aldol–Tishchenko Reaction

R2

H

R1

H

R1 3

O OH

O

R2

H

H

OM

O

R2 - CHO

OM

329

R2

R2 - CHO

R1 R2

O

5 H

R1 H

R1

O R2

H

O

M

O

O OM

O H

H

OM H R2

O

R2

O

R1

R1 R2 MO

R1 H

O

O

R2 R2

H

4

R2 H

Scheme 8.1

tholate catalysts to investigate the stereochemistry of the self-addition of 2-methylpropanal [8]. 1,3-Diol monoesters were obtained with low enantioselectivity (ee > 30%). Morken et al. recently published details of an asymmetric mixed aldol–Tishchenko reaction of aromatic aldehydes with 2methylpropanal catalyzed by salen complexes of yttrium [9]. The results are shown in Table 8.1. These are the ﬁrst examples of enantioselective catalytic aldol–Tishchenko reactions of two diﬀerent aldehydes. 8.2.2

The Aldol–Tishchenko Reaction with Ketones and Aldehydes

Ketones and aldehydes also undergo an aldol–Tishchenko reaction. Three adjacent stereogenic centers can be created by use of these reactants in the aldol–Tishchenko reaction whereas only two stereogenic centers can be formally produced by reacting enolizable aldehydes with aldehydes (Eq. (1)). This is a very eﬀective reaction sequence in terms of chiral economy [11]. The nomenclature used in Eq. (2) and Scheme 8.2 is used throughout the following sections.

330

8 The Aldol–Tishchenko Reaction Tab. 8.1

Enantioselective aldol–Tishchenko reaction in the presence of chiral yttrium complexes. Ph H

Ph H

N

N R - adamantyl

OH

HO R

R

2 mol% Y2O(OiPr)13 R - CHO +

CHO

13 mol% ligand

OH

OCOiPr

R

Entry

Substrate

Yield [%]

e.r. (Conﬁguration)

1 2 3 4 5

Phenyl 4-Bromphenyl Naphthyl 4-Methoxyphenyl 3-Phenyl-2-propenyl

70 55 50 21 50

87:13 (S) 85:15 (S) 82:18 86:14 55:45

O R'

O

2 R' - CHO +

R''

OH

O

O R'

OH

2 1

3

O

R'

2

R''

R'

1

3

R''

+ 6 1,3-diol 1-monoester

7 1,3-diol 3-monoester

Equation 2

Catalysts: nickel enolates [12], SmI2 [6, 13], zinc enolates [14], titanium ate complexes [15], LDA [16–18], Ti(OiPr)4 [19].

The 1,3-diol 1-monoester 6 and the corresponding 3-monoester 7 were prepared with high simple stereoselectivity. Only one of the four possible diastereoisomers has been formed in all examples described in the literature. On the basis of the transition state shown in Scheme 8.1 the 1,3-diol 1-monoester 6 was formed in the 1,2-anti, 1,3-anti conﬁguration (Scheme 8.2). The diol 3-monoester 7 was again formed by acyl migration during reaction. The extent of this migration usually depends on the steric bulkiness of the starting aldehydes. Heathcock et al. showed that isolated nickel ketone enolates react with

8.2 The Aldol–Tishchenko Reaction

O R1 2 R1 - CHO +

O

O

OH 6

R1

1,2-anti, 1,3-anti 1,3-diol 1-monoester R1

O R1

H O

O

R1

M OH R1

O

O 7

1,2-anti, 1,3-anti 1,3-diol 3-monoester

Entry

R1

Diastereoselectivity

1

Ph

99 : 1

2

tBu

98 : 2

3

iPr

98 : 2

4

nPr

97 : 3

Scheme 8.2

Reaction conditions: titanium ate complexes [15].

benzaldehyde to furnish products resulting from an aldol–Tishchenko reaction [12]. They also established the 1,2-anti, 1,3-anti conﬁguration of the isolated 1,3-diol monoester. These are the ﬁrst examples of an aldol– Tishchenko reaction of ketones with aldehydes. Later we found that 1,3-diol monoesters 6 and 7 were formed with high stereoselectivity by an one-pot aldol–Tishchenko reaction of ketones with aldehydes in the presence of substoichiometric amounts of titanium ate complexes [15]. An instructive example for the direction of stereochemistry during the aldol–Tishchenko reaction is the observation that the 1,2-anti, 1,3-anti conﬁguration of the isolated diol monoesters 6 and 7 is independent of the conﬁguration of the assumed starting aldol. To demonstrate this, we have reacted the pure syn aldol 8 of benzaldehyde and diethylketone with one equivalent of benzaldehyde in the presence of catalytic amounts of titanium ate complexes (Eq.

331

332

8 The Aldol–Tishchenko Reaction

(3)). Under these conditions the 1,2-anti, 1,3-anti conﬁgured diol monoesters 9 and 10 were formed exclusively. OH

O

Ph - CHO + Ph

8

Ph OH

O

O

Ph 9 + Ph O

O

OH

Ph 10 Equation 3

Reaction conditions: 10 mol% BuTi(OiPr)4 Li.

Other authors have also described achieving the same stereodirection by use of catalytic amounts of metal alkoxides [7] or LDA [16–18] (Eq. (2)). Three applications of this reaction are shown in Eqs. (4)–(6). A samarium ion-catalyzed aldol–Tishchenko reaction combined with a reductive cyclization process was reported by Curran and Wolin [13]. Only one isomer was detected during this transformation (Eq. (4)). The conﬁguration of the 1,3diol monoester 12 (1,2-anti, 1,3-anti) was the same as that shown in Scheme 8.2 and Eq. (3). O O

OH

O

Ph Ph

I 11 Equation 4

SmI2 , Ph(CH2 )2 CHO, 81%.

12

8.2 The Aldol–Tishchenko Reaction

333

Ph

Ph Ph O

O

O

OH

Ph 14

13 Equation 5

LDA, PhCHO, 56%.

O Br

OH

CHO Br

OH

Br

OTIPS

OTIPS

+ 15

16

17

Equation 6

1. SmI2 , MeCHO; 2. K2 CO3 , MeOH, 96%.

In another example, the keto epoxide 13 was reacted with LDA and benzaldehyde to give the hydroxyester 14 as a single isomer (Eq. (5)) [16]. Because of the missing stereogenic center in the epoxide 13 only one new stereogenic center was created. In terms of the stereochemistry only this example seems to lie between the Evans–Tishchenko reduction and the aldol–Tishchenko reaction. A samarium-catalyzed aldol–Tishchenko reaction has been used to synthesize the intermediate 17 in a highly convergent synthesis of luminacin D [20]. The diol 17 was obtained as a single isomer (Eq. (6)). The aldol–Tishchenko reaction has also been used in the synthesis of 1 0 branched chain sugar nucleosides. The 1 0 -hydroxymethyl group was introduced by an Sm2 -promoted aldol–Tishchenko reaction of 1 0 -phenylseleno2 0 -ketouridine 18 with aldehydes (Eq. (7)). This is the ﬁrst example of generation of an enolate by reductive cleavage of a CaSe bond by SmI2 [21].

O

N

O

NBOM

NBOM

NBOM RO

O

O

O

RO O

SmJ2

N

RO

O

O HO

R' - CHO

SePh RO

O 18

Equation 7

Synthesis of 10 -branched nucleosides.

RO

OSm3+

N

O R

OCOR'

RO 20

334

8 The Aldol–Tishchenko Reaction

O OH

OSiMe3

OH

OH

O

i

20

21 O

OH

OSiMe3

OH

OH

O

ii

22

23

Scheme 8.3

Reaction condition: (i)10 mol% Ti(OiPr)4 , EtCHO, 0 C, 91%; (ii) 10 mol% Ti(OiPr)4 , EtCHO, 0 C, 90%.

An interesting reaction was reported by Delas et al. [19]. They described aldol–Tishchenko reactions of enolsilanes (activated ketones) with aldehydes in the presence of Ti(OiPr)4 – a variation of the classic aldol– Tishchenko reaction (Scheme 8.3). They were able to obtain the stereosixtades 21 and 23 – compounds with six deﬁned adjacent stereogenic centers – in one reaction step. The diastereoselectivity observed was very high and the stereosixtades were obtained as single isomers. This example indicates that the stereodirection of the aldol–Tishchenko reaction can be aﬀected. Oxygen-containing functionality in the starting enolsilanes 20 and 22 has a useful stereodirecting eﬀect on this transformation and on the conﬁguration of the stereosixtades 21 and 23. This paper pioneered the ﬁeld of stereoselective aldol–Tishchenko reactions. The two examples given in Scheme 8.3 show the stereochemical potential of this process. Schneider et al. recently published an enantioselective approach to chiral 1,3-anti-diol monoesters. Although at ﬁrst glance this transformation seems to be a Tishchenko reduction of an acetate aldol (Section 8.2.3) inspection of the mechanism furnishes evidence of a retro-aldol/aldol–Tishchenko reaction. By using 10 mol% Zr(OtBu)4 -TADDOL the 1,3-diol monoesters were obtained with moderate enantioselectivity (Table 8.2) [22]. 8.2.3

The Evans–Tishchenko Reduction

The Evans–Tishchenko reduction is a special case of the aldol–Tishchenko reaction. The starting material is an aldol adduct, usually an acetate aldol. During the reaction the keto functionality of the starting aldol is reduced by

8.2 The Aldol–Tishchenko Reaction Tab. 8.2

Enantioselective aldol–Tishchenko reaction of aldehydes and ketones. R1 OH

10 mol% Zr(OtBu)4 TADDOL

O

R1 - CHO +

O

O

OH

R1

R2

R2

R1 - CHO

OZr(OtBu)3

R1 - CHO

OH

R2

O

R1

R2

Entry

R1

R2

Yield [%]

ee [%]

1 2 3 4

tBu tBu tBu tBu

nHex iPr cHex 2-ethylpropyl

88 84 75 69

42 57 50 47

reaction with an aldehyde in the presence of a Lewis acid. This reaction was ﬁrst described and elaborated on by Evans and Hoyveyda [23]. The reaction was performed in the presence of substoichiometric amounts of SmI2 . The 1,3-diols were isolated in excellent yields with high anti stereoselectivity (> 99:1). The transition structure proposed for the samarium-catalyzed reduction is given in Scheme 8.4. It is very close to those described in previous sections. R1 OH R1

O

R2 R2 +

O R3

H

R3 - CHO

O

O Sm

O R3

O O

OH

OH

O

R3

+ R1 Scheme 8.4

Catalysts: SmI2 [24–29] BuLi [30, 31], zirconocene complexes [32], Sc(OTf )3 [33], ArMgBr [34].

R2

R1

R2

335

336

8 The Aldol–Tishchenko Reaction

Several other metal compounds were subsequently found to induce this reduction (Scheme 8.4). Few authors have described the acyl migration as a result of this reduction that one could expect from the reaction mechanism [30, 34]. This is an unusual result. Five applications of the SmI2 -mediated reduction of hydroxy ketones in natural product synthesis are given in Scheme 8.5.

OH

O O

Ph

OAc

O O

Ph

O i

OH

C14H29

C14H29

24 OH TBSO ( ) 5

25

O

OBz OH ii

TBSO

( )5

27

26 OSEM O

iii

OSEM

OH

OH

OBz 29

28 iiii

PMBO O

PMBO OH

OH

OCOEt

30 OH

31 O

OTBS

OBz OH

Ph

v O

O

OTBDPS

32 Scheme 8.5

(i) CH3 CHO, SmI2 , 80% [25]; (ii) PhCHO, SmI2 , 70% [26]; (iii) PhCHO, SmI2 , 85% [27]; (iv) EtCHO, SmI2 , 97% [29]; (v) SmI2 , PhCHO, 95% [35].

OTBS

Ph O

O

OTBDPS

33

8.2 The Aldol–Tishchenko Reaction

The broad variety of functional groups which can be used in this reaction are represented in these substrates. An interesting example is reduction of the hydroxyketone 32 to the hydroxybenzoate 33. The authors obtained a single 1,2-syn, 1,3-anti-conﬁgured diastereoisomer [35]. Exclusive formation of the 1,2-syn, 1,3-anti-conﬁgured hydroxybenzoate 33 is observed, irrespective of the 1,2-anti conﬁguration of the starting hydroxyketone 32. This reaction could oﬀer an approach to natural products containing a 1,2-syn conﬁguration; these have previously been unattainable by contemporary aldol additions. Two spectacular examples of the use of the Evans–Tishchenko reduction in natural product synthesis are given in Scheme 8.6. In 1993 Schreiber et

OTBS PMBO

ODEIPS

O

OH

H

MeO 34

OMe

OTIPS CHO i,

NBoc

BocN H ODEIPS

OTBS PMBO

OH

O

O H

MeO 35

OMe

OTIPS

J

PMBO

36 OH

O

OTBS

ii PMBO

J OH

O

O

OTBS

Et Scheme 8.6

(i) SmI2 aPhCHO complex, 95%; (ii) SmI2 , EtCHO, 92%.

37

337

338

8 The Aldol–Tishchenko Reaction

al. used this reaction in the total synthesis of rapamycin to obtain the intermediate 35 [24]. This 1,3-diol monoester 35 was obtained with the correct and required stereochemistry as a single isomer by a Tishchenko reduction of ketone 34 in the presence of 30 mol% (PhCHO)SmIaSmI3 (Scheme 8.6). The formation of this complex was described by Evans and Hoyveyda [23]. Paterson et al. used the Tishchenko reduction successfully for several total syntheses. In the synthesis of callipeltoside they needed the intermediate 37 for the synthesis of the aglycone [28]. Again, Tishchenko reduction of ketone 36 with propionaldehyde in the presence of SmI2 yielded the diol monoester 37 with the required 1,2-anti, 1,3-anti conﬁguration (Scheme 8.6). Evans et al. [36] used this procedure in the total synthesis of bryostatin 2. Starting from the corresponding ketone 38 they obtained the ketone 39, with the required conﬁguration of the hydroxy group, by samariumcatalyzed Tishchenko reduction (Eq. (8)). These examples show the broad application of this highly stereoselective Tishchenko reduction process.

O PhO2S

OH

O

( )3 OPMB 38

O PhO2S

OR

OH

( )3 OPMB 39

Equation 8

Reaction conditions: SmI2 , p-NO2 C6 H4 CHO; ds > 95:5, 76%.

The Evans–Tishchenko reduction also provides an eﬃcient and practical solution for the oxidation of aldehydes containing sensitive electron-rich heteroatoms (e.g. aldehyde 40, Eq. (9)). Careful selection of the sacriﬁcial b-hydroxy ketone subsequently provides very ﬂexible access to the desired carboxylic acid 41 (Eq. (9)). This methodology was used in total synthesis of (þ)-13-deoxytedanolide [37].

8.2 The Aldol–Tishchenko Reaction

O

OH O

R S OHC

339

i

OH

S S

O

S

R

40 S ii HOOC

S

Equation 9

Reaction conditions: (i) 20 mol% SmI2 , THF, 10 C; (ii) LiOH, aqueous MeOH.

8.2.4

Related Reactions

The samarium-mediated coupling reaction of vinyl esters with aldehydes has been described [38]. These reactions were performed with enolizable and aromatic aldehydes. Unsymmetrical diesters such as 42 were formed by means of this transformation (Scheme 8.7). On the basis of labeling studies (reaction of vinyl acetate with PhCOD) the mechanism given in Scheme 8.7 seems plausible. An eight-membered alkoxy samarium species might be the key intermediate in this reaction. Subsequent intramolecular hydride shift, as known from the Tishchenko reaction, produces the diester. The pinacol–Tishchenko reaction has recently been described (Eq. (10)), Figure 8.1) [39]. a-Hydroxy epoxides 43 were reacted with SmI2 , eﬃciently forming the 2-quaternary 1,3-diol monoesters 44 and 45 with high diaster-

O R1

O + R2 - CHO

O

R1

O

R2 O

R2

O 42

Ph O D D O

O

D

Ph

O

Ph O

Sm O Scheme 8.7

Reaction conditions: 10 mol% Cp2 Sm(THF)2 .

O

D

O

Ph

41

340

8 The Aldol–Tishchenko Reaction

HO R2 O

R2 i

R1

2

OH R2 R1

1

R1

+

3

OCOR3

OH

1,2-anti, 1.3-anti-diol 1-monoester 44

43

OCOR3

1,2-anti,1.2-anti-diol 3-monoester 45

Equation 10

Reaction conditions: (i) 10–30 mol% SmI2 ; R1 ¼ R2 ¼ Ph, 95%; R1 ¼ Me, R2 ¼ Ph, 96%; R1 ¼ Et, R2 ¼ Ph, 92%; R1 ¼ iPr, R2 ¼ Ph, 92%.

eoselectivity. The reaction is similar to those already described. Rearrangement of the starting a-hydroxy epoxide 43 occurs under the reaction conditions described. Insertion into the MaO bond occurs on addition of the aldehyde R3 CHO, and as a consequence the hemiacetal is formed (Figure 8.1). Hydride transfer results in the formation of the 1,3-diol monoesters 44 and 45. Only the 1,2-anti, 1,3-anti conﬁgured diol monoester is formed. This conﬁguration is again independent of the conﬁguration of the starting ahydroxy epoxides. The corresponding 1,2-syn conﬁgured products were not detected.

O M R2

OH R1

R1

M

O

OH

R2 R3 - CHO

H R3

O O R2

R1 OH

R1

H R3

O R2

OH O M

Fig. 8.1

8.3 Representative Procedures

It is clear that the Tishchenko reaction and its variations are valuable additions to the repertoire of the chemist interested in stereoselective synthesis. The exceptionally high stereoselectivity obtained in this reaction is fascinating (de > 95:5 for all the examples described). Only one diastereoisomer was formed during the reaction in all other examples. Although much has been achieved there still is a need for control of the remaining directions of stereochemistry. This is true not only for the diastereoselectivity but also for the control of enantioselectivity. 8.3

Representative Procedures Typical Procedure for the Aldol–Tishchenko Reaction: rac(1S,2S,3R)-3-Hydroxy2-methyl-1-phenylpent-3-yl benzoate (9) and rac(1S,2R,3S)-1-Hydroxy-2-methyl1-phenylpent-1-yl benzoate (10) [15]. BuLi (0.64 mL, 1.0 mmol in hexane) was carefully added, under inert conditions, to a solution of titanium(IV) iso-propoxide (0.32 mL, 1.0 mmol) in 1-tert-butoxy-2-methoxyethane (1.5 mL). After stirring for 30 min at room temperature pentan-3-one (0.5 mL, 5 mmol) and then benzaldehyde (1.0 mL, 10 mmol) were added. The solution was stirred for further 24 h at room temperature. Diethyl ether (50 mL) was added and the organic phase was extracted with water until neutral. The organic layer was isolated, dried (Na2 SO4 ), ﬁltered, and evaporated under vacuum. The pure products 9 and 10 were separated by ﬂash chromatography with hexane–isoPrOH as eluent (95:5). Yield 63% (ratio of 9/10 ¼ 95:5). Typical Procedure for the Tishchenko Reduction: (7S,9R,10R)-1-Benzenesulfonyl-2,2-dimethyl-9-hydroxy-10-(para-methoxybenzyloxy)-7-(para-nitrobenzoyloxy)-undecan-3-one (39) [36]. Freshly prepared samarium diiodide (0.1 m in THF, 38 mL, 38 mmol, 0.24 equiv.) was added dropwise to a cooled (0 C) solution of ketone 38 (7.86 g, 15.6 mmol) and p-nitrobenzaldehyde (23.6 g, 156 mmol. 10 equiv.) in 200 mL THF. The reaction was stirred in the dark under an argon atmosphere for 5.5 h then quenched with 150 mL sat. aqueous NaHCO3 . The mixture was partitioned between 250 mL EtOAc and 250 mL aqueous sat. NaHCO3 and the aqueous phase was extracted with EtOAc (3 100 mL). The organic extracts were combined, washed with brine (1 75 mL), dried (MgSO4 ), ﬁltered, and evaporated in vacuo. A small sample was removed and ﬁltered through a plug of silica gel with 50% EtOAc–hexane as eluent. All the fractions containing the product were combined and assayed by HPLC (Zorbax silica gel, 0.7% EtOHaCH2 CL2 , 1.0 mL min1 , 254 nm UV cut-oﬀ ) to reveal the product was a 93:7 mixture of diastereoisomers (retention time major product 23.9 min; retention time minor product 32.7 min) which were shown to be from the previous aldol addition. The other diastereomer from the Tishchenko reduction could not be isolated and the only other compound visible in the HPLC chromatogram was present at <3%.

341

342

8 The Aldol–Tishchenko Reaction

The bulk of the residue was puriﬁed by ﬂash chromatography (8 cm 15 cm silica gel, elution with 1.5 L of 5% EtOAc in CH2 Cl2 and 1 L each 9, 10, 11, and 12% EtOAc in CH2 Cl2 ; the mixed fractions were re-puriﬁed (8 cm 12 cm silica gel with 9 and 10% EtOAc in CH2 Cl2 , then 6 cm 15 cm silica gel with 10% EtOAc in CH2 Cl2 ) to give 8.66 g (85% yield) of alcohol 39 as a rusty orange foam.

References 1 Tishchenko, W. Chem. Zentralbl. 1906, 77, 1309;

Tishchenko, W. Chem. Zentralbl. 1906, 77, 1552. 2 (a) Franke, A.; Kohn, L. Monatsh. Chem. 1889, 19, 354; (b)

3 4

5 6 7 8 9

10 11

12 13 14 15 16

17 18 19

Neusta¨dter, V. Monatsh. Chem. 1906, 27, 879; (c) Morgenstern, M. Monatsh. Chem. 1903, 24, 579; (d) Kirschbaum, M. A. Monatsh. Chem. 1904, 25, 249. For a review see: Peterson, J.; Timotheus, H.; Maeorg, U. Proc. Estonian Acad. Sci. Chem. 1997, 46, 93. (a) Casnati, G.; Pochini, A.; Salerno, G.; Ungaro, R. Tetrahedron Lett. 1974, 12, 959; (b) Pochini, A.; Salerno, G.; Ungaro, R. Synthesis 1975, 164; (c) Casnati, G.; Pochini, A.; Salerno, G.; Ungaro, R. J. Chem. Soc., Perkin Trans. I, 1975, 1527. Miyano, A.; Tashiro, D.; Kawasaki, Y.; Sakaguchi, S.; Ishii, Y. Tetrahedron Lett. 1998, 39, 6901. Lu, L.; Chang, H.-Y.; Fang, J.-M. J. Org. Chem. 1999, 64, 843. Mascarenhas, C. M.; Duffey, M. O.; Liu, S.-Y.; Morken, J. P. Org. Lett. 1999, 1, 1427. Loog, O.; Maeorg, U. Tetrahedron: Asymmetry 1999, 10, 2411. Mascarenhas, C. M.; Miller, S. P.; White, P. S.; Morken, J. P. Angew. Chem. 2001, 113, 621, Angew. Chem., Int. Ed. Engl. 2001, 40, 601. Fouquet, G.; Merger, F.; Platz, R. Liebigs Ann. Chem. 1979, 1591. For a discussion of ‘‘atom economy’’, see: a) Trost, B. Science 1991, 254, 1471; (b) Trost, B. M. Angew. Chem., Int. Ed. Engl. 1995, 34, 259. Burkhardt, E. R.; Bergman, R. G.; Heathcock, C. H. Organometallics, 1990, 9, 30. Curran, D. P.; Wolin, R. L. Synlett 1991, 317. Horiuchi, Y.; Taniguchi, M.; Oshima, K.; Utimoto, K. Tetrahedron Lett. 1995, 36, 5353. Mahrwald, R.; Costisella, B. Synthesis 1996, 1087. Baramee, A.; Chaichit, N.; Intawee, P.; Thebtaranonth, C.; Thebtaranonth, Y. J. Chem. Soc., Chem. Comm. 1991, 1016. Bodnar, P. M.; Shaw, J. T.; Woerpel, K. A. J. Org. Chem. 1997, 62, 5674. Abu-Hasanayan, F.; Streitweiser, A. J. Org. Chem. 1998, 63, 2954. Delas, C.; Blacque, O.; Moise, C. J. Chem. Soc., Perkin Trans. 1 2000, 2265.

Common Abbreviations Used in the Text 20 Shotwell, J. B.; Krygowski, E. S.; Hines, J.; Koh, B.;

21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36

37

38 39

Huntsman, E. W. D.; Choi, H. W.; Schneekloth, J. S.; Wood, J. L.; Crews, C. M. Org. Lett. 2002, 4, 3087. Kodama, T.; Shuto, S.; Ichikawa, S.; Matsuda, A. J. Org. Chem. 2002, 67, 7706. (a) Schneider, C.; Hansch, M. Synlett 2003, 837; (b) Schneider, C.; Hansch, M. Chem. Comm. 2001, 1218. Evans, D. A.; Hoyveyda, A. H. J. Am. Chem. Soc. 1990, 112, 6447. Romo, D.; Meyer, S. D.; Johnson, D. D.; Schreiber, S. L. J. Am. Chem. Soc. 1993, 115, 7906. Wild, R.; Schmidt, R. R. Tetrahedron: Asymmetry 1994, 5, 2195. Hulme, A. N.; Howells, G. E. Tetrahedron Lett. 1997, 47, 8245. Scho¨ning, K.-U.; Hayashi, R. K.; Powell, D. R. Kirschning, A. Tetrahedron: Asymmetry 1999, 10, 817. Paterson, I.; Davies, R. D. M.; Marquez, R. Angew. Chem. 2001, 113, 623, Angew. Chem., Int. Ed. Engl. 2001, 40, 603. Paterson, I.; Florence, G. D.; Gelach, K.; Scott, J. P.; Sereinig, N. J. Am. Chem. Soc. 2001, 123, 9535. Aoki, Y.; Oshima, K.; Utimoto, K. Chem. Lett. 1995, 463. Iwamoto, K.; Chatani, N.; Murai, S. J. Organomet. Chem. 1999, 574, 171. Umekawa, Y.; Sakaguchi, S.; Nishiyama, Y.; Ishii, Y. J. Org. Chem. 1997, 62, 3409. Gillespie, K. M.; Munslow, I. J.; Scott, P. Tetrahedron Lett. 1999, 40, 9371. Mellor, J. M.; Reid, G.; El-Sagheer, A. H.; El-Tamany, E. S. H. Tetrahedron 2000, 56, 10039. Enders, D.; Ince, S. J.; Bonnekessel, M.; Runsink, J.; Raabe, G. Synlett, 2002, 962. Evans, D. A.; Carter, P. H.; Carreira, E. M.; Charette, A. B.; Prunet, J. A.; Lautens, M. Angew. Chem. 1998, 110, 2526, Angew. Chem. Int. Ed. Engl. 1998, 37, 2354; Evans, D. A.; Carter, P. H.; Carreira, E. M.; Charette, A. B.; Prunet, J. A.; Lautens, M. J. Am. Chem. Soc. 1999, 121, 7540. (a) Smith, A. B. III; Lee, D.; Adams, C. M.; Kozlowski, M. C. Org. Lett. 2002, 4, 4539; (b) Smith, A. B. III; Adams, C. M.; Barbosa, S. A. L.; Degnan, A. P. J. Am. Chem. Soc. 2003, 125, 350. Takeno, M.; Kikuchi, S.; Morita, K.-I.; Nishiyama, Y.; Ishii, Y. J. Org. Chem. 1995, 60, 4974. (a) Fan, C. A.; Wang, B. M.; Tu, Y. Q.; Song, Z. L. Angew. Chem. 2001, 113, 3995, Angew. Chem., Int. Ed. Engl. 2001, 40, 3877; (b) Tu, Y. Q.; Sun, L. D.; Wang, P. Z. J. Org. Chem. 1999, 64, 629; (c) Wang, F.; Tu, Y. Q.; Fan, C. A.; Wang, S. H.; Zhang, F. M. Tetrahedron: Asymmetry 2002, 13, 395.

Common Abbreviations Used in the Text

Ac Ar

Acetate Aryl

343

344

8 The Aldol–Tishchenko Reaction

BINOL Bn Boc BOM Bu Bz Cp DEIPS Et Hex LDA Me Ph PMB Pr SEM TBDPS TBS Tf THF TIPS

1.1 0 -bi-2-Naphthol Benzyl tert-Butoxycarbonyl Benzyloxymethyl Butyl Benzoate Cyclopentadienyl Diethylisopropylsilyl Ethyl Hexyl Lithium diisopropylamide Methyl Phenyl p-Methoxybenzyl Propyl 2-(Trimethylsilyl)methyl tert-Butyldiphenylsilyl tert-Butyldimethylsilyl Triﬂuoromethanesulfonyl Tetrahydrofuran Triisopropylsilyl

345

Index Numbers in front of the page numbers refer to Volumes 1 and 2: e.g., 2/250 refers to page 250 in volume 2

a ab-initio calculation 1/23 acetal 1/142 acetoacetate decarboxylase 1/275 achiral enolates 1/51 acyclic transition state 1/139, 2/11 – model 2/180 acyclic E-trichlorosilyl enolate 2/271 acyl iron complexes 1/36 acyl migration 2/330 N-acyloxazolidinone 2/223 acyloxyborane (CAB) 2/33 1,2-addidion 2/2, 2/203 adjacent stereogenic center 1/96 Agami model 1/176 aklavinone 2/118 aldehyde 1/231 – chiral 1/231 aldohexose 1/237 aldol condensation 1/107, 1/162 aldol cyclization 2/150 – intramolecular 2/150 aldol reaction 1/282 – intramolecular 1/282 – reductive 2/221, 2/223 – regioselective vinylogous 2/302 aldolase 1/161 aldolase antibodies 1/275 aldolase antibody-catalyzed kinetic resolution 1/284 aldol-related Claisen addition 1/258 aldol-Tishchenko reaction 2/167, 2/334 allylic rearrangement 1/295 amino acid 1/80, 1/87, 1/164 – cyclic 1/185 a-aminoaldol 1/92

aminoindanol 1/87, 1/103, 1/109 amphotericin B 1/212, 2/97 anthracene-based bis(oxazoline)Cu(II) complex 2/73 anthracyclinones 1/8 (E)-anti correlation 1/16 anti diastereoselectivity 1/98, 1/101 antibody immunoglobulin 1/273ﬀ antibody-directed enzyme prodrug therapy (ADEPT) 1/291 anti-Cram/Felkin 1/322 antigen 1/292 aqueous micelle 1/191 l-arabinose 1/210 arylsulfonamide 2/44 aspicillin 1/243 asymmetric catalyst 2/219 – heterobimetallic 2/219 asymmetric Mannich-type reaction 2/187 asymmetric polymerization 2/52 asymmetric quarternary carbon 2/138, 2/142, 2/154 australine 1/241 azasugar 1/239f azasugar phosphonate 1/242 azido aldehyde 1/240

b B(OTf )3 2/62 Baeyer-Villiger oxidation 2/203 Baldwin enolendo-trig 1/283 Baldwin’s rules 1/282 bengamide 2/115 N-benzylcinchoninium chloride 2/219 benzyloxyacetaldehyde 1/88

Modern Aldol Reactions. Vol. 2: Metal Catalysis. Edited by Rainer Mahrwald Copyright 8 2004 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 3-527-30714-1

346

Index benzyloxypropionaldehyde 1/88 bidentate activation 2/85 bifunctional metallic catalysis 2/198 (R)-BINAP-AgOTf 2/2 (R)-BINAP-PdCl2 2/7 (R)-BINAP PdCl2 –AgOTf 2/8 BINAP-silver(I) 2/2ﬀ, 2/15 bis(oxazoline)–Sn(OTf)2 2/148 bis(oxazuoline) ligand 2/71 Bn/Box–Sn(II) 2/148 Bn/Box–Sn(II) complex 2/134 boat conformation 1/19 boat transition model 1/68 boat transition structure 2/271 boat-like structure 2/306 boat-like transition 1/93 boat-like transition state 2/278 – model 1/63 boat-like transition structure 2/261, 2/264, 2/273, 2/279, 2/282, 2/310 boron enolate 1/129ﬀ (E) boron ketone enolates 1/131 (þ)-boronolide 2/212 branched aldehydes 173 brevicokins 1/291 brevicomin 1/242f, 1/249 bryostatin 2/338 bryostatin 7 2/204 bryostatin C1 –C9 segment 2/37 (t-Bu-box) 2/82 (t-Bu-box)Cu(OTf )2 2/82 tert-butyldimethylsilyl enol ether 2/61

c Ca catalyst 2/206 Cahn-Ingold-Prelog nomenclature 1/12 calipeltoside 2/338 callepeltoside 2/73 camphor 1/67, 1/69, 1/84 camptothecin 1/293 (þ)-2-carane 1/70 carbohydrate-titanium complex 1/92 cationic silicon-phosphoramide complex 2/253 cephalosporolide D 2/153 chair-like transition 1/324 – structure 2/264, 2/268, 2/270, 2/276, 2/282f, 2/289, 2/310 chair-like Zimmermann-Traxler transition 1/319 Chan diene 2/71 chelating carbonyl compound 2/77 chelation 1/80, 1/86 – control 1/143, 1/92, 1/110

– model 2/109 chelation-controlled 1/103 – aldol reaction 1/320 – asymmetric environment 2/199 – model 1/87 chiral acetyl iron complexes 1/35 chiral aldehyde 1/231 chiral auyxiliary 1/83 chiral cyclic phosphoramide 2/250ﬀ chiral diamine 2/130 chiral enolate 1/315 chiral iron acyl complexes 1/32 chiral Lewis basic (LB) group 2/233 chiral lithium amides 1/52 chiral mandelic acid 2/209 chiral N-oxide 2/251, 2/296 chiral pyrrolidine 2/128 chiral zirconium catalyst 2/187 – air-stable 2/187 chlorosilyl ketene acetal 2/248 Claisen addition 1/258 – aldol-related 1/258 Claisen condensation 1/8, 2/94 Claisen-Schmidt reaction 1/4 class I aldolases 1/204, 1/274, 2/198 class II aldolases 1/206, 2/198 closed transition structure 2/230 compactin 2/45 competitive background reaction 2/299 concerted [4þ2] cycloaddition 2/181 (4S) conﬁguration 1/220 conformation 1/24 coordinating capacity 2/279 counter-ion 1/24 13 C NMR spectrum 2/184 Cram chelate model 2/28 Cram/Felkin 1/232 Cram/Felkin adduct 1/318 Cram/Felkin product 1/319 Cram’s cyclic model 1/43 Cram-Felkin-Anh model 1/43 cross-aldol reaction 1/65, 1/280, 1/286 Cu(ClO4 )2 2/88 Cu(OTf )2 2/82 Curtius rearrangement 1/110 cyclic amino acid 1185 cyclic enolates 1/75 cyclic transition 1/93 cyclic transition state 1/20 – structure 2/4 cyclitol 1/238 cyclization 1/176 – reductive 2/332 [4þ2] cycloaddition 2/180, 2/182

Index – concerted 2/181 cyclo-aldolization 1/176

d Danishefsky’s diene 2/84, 2/175ﬀ – functionalized 2/179 l-daunosamine 2/111 decarboxylation 1/295 4-demethoxydaunomycinone 2/118 density functional theory calculation 1/163, 1/184, 1/187 6-deoxy-d-allose 2/142 deoxyribonucleoside synthesis 1/252 6-deoxy-l-talose 2/141 13-deoxytedanolide 2/338 deprotonation 1/10 – kinetically controlled 1/10 desymmetrization 1/52, 2/49 diamine 2/130 – chiral 2/130 diarylborinic acid 2/30ﬀ diastereofacial selectivity 1/96 diastereoselection 2/261, 2/273, 2/273, 2/280, 2/283 – internal 2/261, 2/273, 2/280, 2/283 – simple 1/10, 1/12f, 1/18, 1/28 diastereoselectivity 2/282 – internal 2/282 diaxial repulsion 1/235 dibutyltin acetate 1/151ﬀ didemnin A 2/113 Diels-Alder reaction 1/296 – hetero- 2/1 dihydroxyacetone 1/221, 1/224 dihydroxyacetone phosphate 1/206, 1/221, 1/224, 1/229f dihydroxylation 2/140 diketones 1/7 1,3-diketone 1/275 1,3-diketone hapten 1/277 dimeric structure 2/186 (R)-dimethyl citramalate 2/114 dimethylphenylglycinol 1/102 dinuclear Zn complex 2/204 1,3-diol monoester 2/327ﬀ dioxolanone 1/30 dipeptide 1/188 direct aldol addition 1/20 direct catalytic 2/200ﬀ direct catalytic asymmetric aldol 2/208 directed aldol reaction 1/4 6,6 0 -disubstituted-3,3 0 -I2 BINOL 2/174ﬀ double activation 1/151

double deprotonation 1/37 double diastereoselection 2/261 double-stereodiﬀerentiating 1/96, 1/102, 1/318, 1/321, 1/323, 1/324, 2/275, 2/287 double-labeling experiment 2/75

e E/anti correlation 1/14 electronic nature 2/292 electron-withdrawing group 2/80, 2/173 electron-withdrawing substituent 2/47 elimination 3/2 emodine 1/8 enamine 1/161f enamine mechanism 1/274f, 1/293 enantioselective catalytic aldol-Tishchenko reaction 2/329 enantioselective cyclization 1/9 enantioselectivity 1/13, 1/224, 1/235 – induced 1/13 – kinetic 1/224, 1/235 enolate 1/14f, 1/315 – chiral 1/315 – preformed 1/9ﬀ (E) enolate 1/14 (Z) enolate 1/14f enol silyl ether (ESE) 2/105ﬀ enolate controlled stereochemistry 1/49 enolate geometry 1/13, 2/267, 2/289, 2/300 enolate oxygen 1/316 enolate structure 2/257 enolendo 1/171 enolendo aldolization 1/167 enolexo 1/171ﬀ 5-enolexo 1/168 enolexo aldol 1/167ﬀ enolexo aldolization 1/167 enolizable aldehyde 2/212, 2/327 ephedrine 1/92 epothilone 1/290 epoxidation 2/140 eptoposide 1/293 equatorial subsituent 1/235 equilibribrium 1/3 – constant 1/208, 1/284 equilibrium-controlled bioconversion 1/209 erythro/threo nomenclature 1/11 erythromycin 1/46 (Z) ester boron enolates 1/131

347

348

Index ethyl pyruvate 2/90 Evans-Tishchenko reduction 2/333ﬀ extended transition state 2/50 extended transition-state model 2/41 external stereoinduction 2/276 external stereoselection 2/273 E/Z ratio 2/244

f facial diﬀerentiation 2/266 facial selectivity 2/306 p-facial selectivity 1/86, 2/81 febrifugine 2/147 Felkin selectivity 2/27 Felkin-Anh model 2/28, 2/261 ferrocenylphosphine-gold(I) 2/8, 2/14 ferrocenylphosphine 26-gold(I) 2/8ﬀ ferrocenylphosphine-silver(I) 2/13 Fischer projection 1/11 ﬂuoride ion 1/146 ﬂuoropyruvate 1/209, 1/212 fostriecin 2/201 frontalin 1/291 frontier orbital 1/21 l-fucose 2/142 functionalized Danishefsky’s diene 2/179

g genetic selection 1/304 geometrically deﬁned trimethylsilyl enol ether 2/246 geometry 2/170, 2/172 glucose isomerase 1/237 glycine 2/220

h Hajos-Parrish-Eder-Sauer-Wiechert cyclization 1/173 Hajos-Parrish-Eder-Sauer-Wiechert reaction 1/9, 1/192 hemiacetal 2/328, 2/340 hetero Diels-Alder 2/175ﬀ heterobimetallic (S)-LaLi3 tris(binaphthoxide) complex 2/200 heterobimetallic asymmetric catalyst 2/219 hetero-Diels-Alder reaction 2/1 hexacoordinated cationic silicon complex 2/253 Hg(II)-catalyzed metathesis 2/243 1 H NMR 2/239 Houk Model 1/176 hydride shift 2/328

hydride transfer 2/328ﬀ – intramolecular 2/35 hydrosilylation 2/237 hydroxyacetone 1/184, 1/280, 2/210 a-hydroxy acids 1/93 hydroxyaldehyde 1/231ﬀ b-hydroxyaldehyde 1/79 a-hydroxy epoxide 2/340 hydroxyketone 1/251 hydroxyl ketone 2/336 hydroxymethyltransferase 1/254 hydroxyproline 1/219 hydroxypyrolidine 1/249 hydroxypyruvate 1/249 hypervalent silicate 1/147 HYTRA 1/36, 1/41 (S)-HYTRA 1/49

i iminium ion 1/161 in vitro evolution 1/303 indane-pybox 2/223 induced enantioselectivity 1/13 induced stereoselectivity 1/13 inherent selectivity 1/318 insulin 12/293 intermolecular silicon transfer 2/85 internal diastereoselection 2/261, 2/273, 2/280, 2/283 internal diastereoselectivity 2/282 internal stereoselectivity 2/278 intramolecular aldol cyclization 2/150 intramolecular aldol reaction 1/282 intramolecular hydride transfer 2/35 inversion strategy 1/236 ionic liquid 1/191 isopropylcyclohexylamide 1/9 isopropylidene glyceraldehyde 1/46 Ivanoﬀ reaction 1/16

j jasmone 1/7

k ketene silyl acetal (KSA) 2/1ﬀ, 2/105ﬀ ketofuranose 1/231, 1/236 ketol isomerase 1/237 ketopyranose 1/231 khafrefungin 2/146 kinetic conditions 1/75 kinetic control 1/13, 1/256 kinetic enantiopreference 1/227 kinetic enantioselectivity 1/224, 1/235 kinetically controlled condition 2/50

Index kinetically controlled deprotonation 1/10 Kiyooka’s-state model 2/39 Knoevenagel condensation 1/164 Knoevenagel-Michael-aldol tandemreaction 1/171

l labelled pyruvate 1/215 lactacystin 2/121 lactonization 106 lanthanide triﬂate 1/144 leucascandrolide A 2/97 Lewis acid catalyst 2/187 – air stable 2/187 Lewis acidity 1/147 Lewis base 1/148 Lewis-base-catalyzed aldol addition 2/234 ligand exchange 1/95 lithium amides 1/52 – chiral 1/52 lithium 2,2,6,6-tetramethylpiperidide 1/10 lithium dianion 1/324 lithium diisopropylamide (LAD) 1/9 (Z)-lithium enolate 1/316 lithium hexamethyldisilazane 1/9 low facial selectivity 2/310 luminacin D 2/333 2,6-lutidine 1/129f

m macro condensation 1/312 macro-lactonizsation 1/320 mandelic acid 2/209 – chiral 2/209 Mannich reaction 1/142, 1/192, 1/296 Mannich-type condensation 1/164 Mannich-type reaction 2/6f mannosamine analogs 1/212 d-mannose 1/210 match/mismatch eﬀect 2/286 matched 1/64, 1/96, 1/119 matched pair 1/49 matched substrate 2/78 maytansin 1/46 ()-menthol 1/70 Me3 SiNTf2 2/57ﬀ Me3 SiOTf 2/55ﬀ Me3 SiOTf/MAD 2/62 Me3 SiTf/MABR 2/62 metal enolate 2/199 metal exchange 2/167 – Si–Sn 2/127

metal-oxygen bond 1/129 metathesis 2/241, 2/260 – Hg(II)-catalyzed 2/243 N-methylpyrrolidinone 1/82f mevinolin 2/45 Michael reaction 1/142, 1/296 Michaelis-Menten equation 1/277 mismatched 1/64, 1/96f, 1/324, 2/97 mismatched pair 1/49 mismatched substrate 2/78 Mitsunobu reaction 2/153 Mukaiyama aldol reaction 2/1ﬀ mycestericin D 1/255

n neuraminic acid 1/209 nikkomycin antibiotics 1/219 nitroalkane 2/91 nitromethane 2/91ﬀ NMR experiments 2/184 nojirimycin 1/240 nomenclature of enzymes 1/204 non chelation control 1/43 non-chelated model 2/112 non-chelated transition 1/81f non-chelation-controlled addition 2/261 non-linear eﬀect 2/186

o octalactin 2/139 oligomeric structure 2/185 open antiperiplanar transition structure 2/79 open transition state 1/136 open transition state model 1/22 open transition structure 2/300 open-chain transition state model 1/90, 2/183 oxa-epothilone 1/315 oxazinone 1/86 oxazolidone 1/134 oxazoline 1/188 oxonium ion 1/161 oxygen stereogenic center 2/285

p paclitaxel 2/123, 2/149 pair 1/49 – matched 1/49 – mismatched 1/49 pentacoordinated cationic silicon complex 2/253 pentamycin 1/243 2,4-pentanedione 1/278

349

350

Index pericyclic-like transition 1/18 Ph/Pybox–Sn(II) complex 2/134 phage libraries 1/302 phenylalanine 1/175 phorboxazole B 2/73 phosphoenolpyruvate 1/208 phosphoramide 2/235, 2/249ﬀ – chiral cyclic 2/250ﬀ pinacol-Tishchenko reaction 2/339 polymer support 1/191 polymerization 2/52 – asymmetric 2/52 polypropionate 1/96 preformed enolate 1/9ﬀ Prelog-Seebach notation 1/12 primary amine cofactor 1/305 proline 1/173ﬀ l-proline 2/139, 2/210 (S)-proline 2/125 (promionate) silylketene acetal 2/73 l-propoline 1/9 pseudoaxial 1/63, 1/88, 2/31 pseudoequatorial 1/63, 1/102 pybox 2/71 (pybox)CuLn 2/70 pyralomicin 1c 2/122 pyrrolidine 2/128 – chiral 2/128 pyrrolizidine 1/241 pyruvate ester 2/89 pyruvate-dependent lyases 1/208ﬀ

q quarternary carbon 2/138, 2/142, 2/154 – asymmetric 2/138, 2/142, 2/154

r rabbit muscle (RAMA) 1/222f rapamycin 2/145, 2/338 rapid injection NMR (RINMR) 2/307 reductive aldol reaction 2/221, 2/223 reductive cyclization 2/332 reductive silylation 2/241 regiochemistry 1/7 regioselective vinylogous aldol reaction 2/302 regioselectivity 1/168, 2/303 retroaldeolization 1/256 retro-aldol 1/227, 1/293 retro-aldol cleavage 1/221 retro-aldol reaction 1/280, 1/283f retro-aldol-Tishchenko reaction 2/334 retroaldolization 1/219

retro-aldol-retro-Michael reaction 1/293, 1/298 retro-Michael reaction 1/293 reversibility 1/2 Robinson annelation 1/5, 1/7 Robinson annexation 1/192 Robinson annulation 1/282

s scandium triﬂate 2/178 Schiﬀ base 2/219f secondary aldol 1/280 secondary amine 1/164 selectivity 1/318 – inherent 1/318 – low facial 2/310 selenium 1/103 selfaddition 1/3 self-aldolization 1/4 self-reproduction of chirality 1/30 silicon complex 2/253 – hexacoordinated cationic 2/253 – pentacoordinated cationic 2/253 silicon transfer 2/85 – intermolecular 2/85 silyl enol ether 2/2, 2/26, 2/59, 2/236 silylation 2/247 – reductive 2/241 simple diastereoselectivity 1/10, 1/12f, 1/18, 1/28, 1/52 Si–Sn metal exchange 2/127 Sn(OTf )2 2/108 119 Sn-solution NMR 2/312 solvent 2/256 ()-sparteine 1/67, 1/82f, 1/107 sphingosine 2/9, 2/145 spirocyclic carbohydrate 1/242 stannyl enolate 2/107 g-stannyl-a,b-enone 2/123 a-stannylketone 2/106 stepwise cycloaddition 2/180 stereochemical model 2/86 stereodiﬀerentiation 1/84 stereodivergent aldol addition 1/28 stereogenic center 2/283 stereoinduction 2/270, 2/272 stereoselectivity 1/13 – induced 1/13 – internal 2/278 steric hindrance 1/22, 1/25, 2/182 steric repulsion 1/22, 2/183 stilbenediamine 1/91 strongly electron-withdrawing 2/234 substrate 2/78

Index – matched 2/78 – mismatched 2/78 substrate tolerance 1/219, 1/223, 1/226, 1/248, 1/252, 1/255 substrate-controlled aldol addition 2/261 sulﬁnyl acetates 1/33 sulfonyl acetamides 1/35 Suzuki coupling 1/312 Swaminathan’s model 1/176 Swern oxidation 1/108 (Z)-syn correlation 1/16 syringolide 1/243

triisopropoxytitanium chloride 1/67 trimethoxysilyl enol ether 2/6 trimethylsilyl enol ether 2/5ﬀ, 2/61 – geometrically deﬁned 2/246 trimethylsilyltriﬂuoromethanesulfonate 1/142 triphenylphosphine 1/78 triphenylphosphine oxide 1/78 tris(2,6-diphenyl)phenoxide (ATPH) 2/101 trityl perchlorate 1/143 tunicamycins 2/113 twist-boat 1/19

t

u

tandem aldolization 1/241 taxol 2/123, 2/149f N-terminal prolyl peptides 1/188 terpene biosynthesis 1/249 tertiary aldol 1/280, 1/283 tetrahydrothiophene ligand 2/131 thermodynamic control 1/24, 1/236 thiosugar 1/239 Ti(Oi-Pr)3 Cl 1/81 TiCl4 1/83 tin(II) chloride 1/144 titanium ate complex 2/331 titanium enolate 1/74 titanium tetrabutoxide 65 titanium tetrachloride 1/67, 1/76 TMEDA 1/83 (S-Tol-BINAP)Cu2 2/96 (S-Tol-BINAP)CuF2 2/97 N-tosylaminoindanol 1/100 tosylation 1/87 N-tosylnorephedrine 1/90 trans silylation 2/245 transition 1/81f – non-chelated 1/81f transition state 2/206 – model 2/44, 2/213 transition structure 2/272 transmetalation 1/75, 1/79, 1/93, 1/98, 2/132, 92ﬀ trialkylsilyl enol ether 2/237 trichlorosilyl enolate 1/148, 2/234ﬀ, 2/238, 2/253 7-triethylsilylbaccatin III2/150

UV-visible-active aldol substrate 1/298

v vanadate 1/230 vanadate ester 1/230 l-vancosamine 2/111 vinyl enolate 2/99 vinyloxyborane 1/128f

w water 2/169 Weinreb amide 2/12 Wieland-Miescher ketone 1/172 Wittig oleﬁnation 1/108

y Yamamoto’s extendend transition model 2/37

z Z/E selectivity 2/244 Z/syn correlation 1/14 Zimmermann-Traxler 1/80 – chair-like transition structure 1/319, 2/93 – model 1/16, 1/18ﬀ – transition 1/63 – transition state 1/74, 1/79, 1/317 – transition state model 1/17, 1/66, 1/76 – type transition 1/102 zirconium enolate 2/167 Zn2þ cation 2/198 Z-syn correlation 1/15

351